Biomedical Nanotechnology 2011

Methods in Molecular Biology™
Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For other titles published in this series, go to

www.springer.com/series/7651
wwwwwwwwwwwwwww
Biomedical Nanotechnology
Methods and Protocols
Edited by
Sarah J. Hurst
Center for Nanoscale Materials, Argonne National Laboratory, Argonne, IL, USA
Editor
Sarah J. Hurst, Ph.D.
Center for Nanoscale Materials
Argonne National Laboratory
9700 S. Cass Avenue
Argonne, Illinois 60439
USA
shurst@anl.gov
ISSN 1064-3745 e-ISSN 1940-6029

ISBN 978-1-61779-051-5 e-ISBN 978-1-61779-052-2
DOI 10.1007/978-1-61779-052-2
Springer New York Dordrecht Heidelberg London
Library of Congress Control Number: 2011922747
© Springer Science+Business Media, LLC 2011

All rights reserved. This work may not be translated or copied in whole or in part without the written permission of
the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013,
USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified
as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of going to press, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may
be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Printed on acid-free paper
Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface
Nanoscience and technology focuses on synthesizing structures that have at least one dimen-
sion on the sub-100 nm length scale. It deals with investigating the fundamental proper-
ties of such structures, which usually differ significantly from that of the bulk material, and
taking advantage of these qualities in constructing novel materials and devices or develop-
ing unique applications. Owing to widespread interest and investment, biomedical nano-
technology, or the use of nanostructures in medicinal applications, is an area of intense
research that is growing and progressing at a rapid pace. This rapid development is driven
by the fact that nanomaterials often offer superior capabilities when compared with con-
ventionally used materials for the detection, diagnosis, and treatment of disease. Further,
they have the potential to enable real-time disease detection and therapy and to advance
point-of-care systems.
The goal of this volume is to provide an overview of biomedical nanotechnology, from
the conception of novel materials in the laboratory to the application of such structures in
the clinic. After a brief introductory chapter, the first section consists of protocol chapters
which provide practical information on the synthesis of a variety of solution-phase and
surface-bound nanomaterials and their application in sensing, imaging, and/or therapeu-
tics; most chapters provide step-by-step instructions and insight into overcoming possible
pitfalls and challenges. The chapters are written by leading researchers in biology, chemistry,
physics, engineering, and medicine from academia, industry, and the national laboratories.
The second section consists of a series of case study/review chapters that discuss the toxi-
cology of nanomaterials, the regulatory pathways to US Food and Drug Administration
(FDA) approval of these materials, their patenting, marketing, and commercialization, and
the legal and ethical issues surrounding their use. These are written by experts in the science,
social science, business, law, and ethics communities. Nanotechnology looks not only to
revolutionize medical care but the fundamental property differences associated with nano-
materials and the potential for their use as multicomponent/multifunctional structures are
also transforming the aspects of these fields that take part in mediating their introduction to
the public.
This volume is a useful reference for scientists and researchers at all levels who are
interested in working in a new area of nanoscience and technology or in expanding their
knowledge base in their current field. The case study/review chapters are meant to inform
scientists of routes they can take in moving their research beyond the bench, so they can
design their systems with real-world considerations in mind. In turn, this volume also will
be of interest to the social scientist, lawyer, or businessperson who wants to learn about
the salient points of nanotechnology as they are applied to their field.
I would like to thank Prof. John M. Walker, series editor for Methods in Molecular
Biology, for his guidance and the authors for sharing their expertise and producing such
high-quality chapters. It was a pleasure to work with them on this project. I would also
like to acknowledge Dr. Haley D. Hill for her support and helpful discussions.
Argonne, IL Sarah J. Hurst
v
wwwwwwwwwwwwwww
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 Biomedical Nanotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Sarah J. Hurst
Part I Using Nanomaterials in Sensing, Imaging, and Therapeutics

2 Multiplexed Detection of Oligonucleotides with Biobarcoded Gold
Nanoparticle Probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Jae-Seung Lee
3 Molecular Detection of Biomarkers and Cells Using Magnetic
Nanoparticles and Diagnostic Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . 33
Jered B. Haun, Tae-Jong Yoon, Hakho Lee, and Ralph Weissleder
4 Real-Time Quantum Dot Tracking of Single Proteins . . . . . . . . . . . . . . . . . . . . . 51
Jerry C. Chang and Sandra J. Rosenthal
5 Titanium Dioxide Nanoparticles in Advanced
Imaging and Nanotherapeutics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Tijana Rajh, Nada M. Dimitrijevic, and Elena A. Rozhkova
6 Surface Modification and Biomolecule Immobilization
on Polymer Spheres for Biosensing Applications . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Chris R. Taitt, Lisa C. Shriver-Lake, George P. Anderson,
and Frances S. Ligler
7 Multivalent Conjugation of Peptides, Proteins, and DNA
to Semiconductor Quantum Dots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Duane E. Prasuhn, Kimihiro Susumu, and Igor L. Medintz
8 A Single SnO2 Nanowire-Based Microelectrode . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Jun Zhou, Yaguang Wei, Qin Kuang, and Zhong Lin Wang
9 Biosensing Using Nanoelectromechanical Systems . . . . . . . . . . . . . . . . . . . . . . . . 119
Ashish Yeri and Di Gao
10 Nano “Fly Paper” Technology for the Capture of Circulating Tumor Cells . . . . . 141
Shutao Wang, Gwen E. Owens, and Hsian-Rong Tseng
11 Polymeric Nanoparticles for Photodynamic Therapy . . . . . . . . . . . . . . . . . . . . . . 151
Yong-Eun Koo Lee and Raoul Kopelman
12 Hydrogel Templates for the Fabrication of Homogeneous
Polymer Microparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Ghanashyam Acharya, Matthew McDermott, Soo Jung Shin,
Haesun Park, and Kinam Park
vii
viii Contents
13 Antibacterial Application of Engineered Bacteriophage

Nanomedicines: Antibody-Targeted, Chloramphenicol Prodrug
Loaded Bacteriophages for Inhibiting the Growth
of Staphylococcus aureus Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Lilach Vaks and Itai Benhar
14 Viruses as Nanomaterials for Drug Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Dustin Lockney, Stefan Franzen, and Steven Lommel
15 Applications of Carbon Nanotubes in Biomedical Studies . . . . . . . . . . . . . . . . . . 223
Hongwei Liao, Bhavna Paratala, Balaji Sitharaman,
and Yuhuang Wang
16 Electrospun Nanofibrous Scaffolds for Engineering Soft Connective Tissues . . . . 243
Roshan James, Udaya S. Toti, Cato T. Laurencin,
and Sangamesh G. Kumbar
17 Peptide Amphiphiles and Porous Biodegradable Scaffolds
for Tissue Regeneration in the Brain and Spinal Cord . . . . . . . . . . . . . . . . . . . . . 259
Rutledge G. Ellis-Behnke and Gerald E. Schneider
18 Computational Simulations of the Interaction of Lipid Membranes
with DNA-Functionalized Gold Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
One-Sun Lee and George C. Schatz
Part II Translating Nanoscience and Technology

from the Lab to the Clinic
19 Cytotoxic Assessment of Carbon Nanotube Interaction

with Cell Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Hanene Ali-Boucetta, Khuloud T. Al-Jamal,
and Kostas Kostarelos
20 Nanoparticle Toxicology: Measurements of Pulmonary
Hazard Effects Following Exposures to Nanoparticles . . . . . . . . . . . . . . . . . . . . . 313
Christie M. Sayes, Kenneth L. Reed, and David B. Warheit
21 Nanoparticle Therapeutics: FDA Approval, Clinical Trials,
Regulatory Pathways, and Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Aaron C. Eifler and C. Shad Thaxton
22 Legislating the Laboratory? Promotion and Precaution
in a Nanomaterials Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Robin Phelps and Erik Fisher
23 Navigating the Patent Landscapes for Nanotechnology: English
Gardens or Tangled Grounds? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
Douglas J. Sylvester and Diana M. Bowman
24 Scientific Entrepreneurship in the Materials and Life Science Industries . . . . . . . . 379
Jose Amado Dinglasan, Darren J. Anderson, and Keith Thomas
25 Applying the Marketing Mix (5 Ps) to Bionanotechnology . . . . . . . . . . . . . . . . . 393
Michael S. Tomczyk
26 Managing the “Known Unknowns”: Theranostic Cancer
Nanomedicine and Informed Consent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Fabrice Jotterand and Archie A. Alexander
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Contributors
Ghanashyam Acharya • Departments of Biomedical Engineering and Pharmaceutics,
Purdue University, West Lafayette, IN, USA
Archie A. Alexander • Adjunct Instructor of Health Administration (Health Law
and Ethics)
Hanene Ali-Boucetta • Centre for Drug Delivery Research, The School of Pharmacy,
University of London, London, UK
Khuloud T. Al-Jamal • Centre for Drug Delivery Research, The School of Pharmacy,
Darren J. Anderson • Vive Nano, Inc., Toronto, ON, Canada
George P. Anderson • Center for Bio/Molecular Science and Engineering,
U.S. Naval Research Laboratory, Washington, DC, USA
Itai Benhar • Department of Molecular Microbiology and Biotechnology, Tel Aviv
University, Ramat Aviv, Israel
Diana M. Bowman • Melbourne School for Population Health, University
of Melbourne, Victoria, Australia; Department of International and European Law,
KU Leuven, Leuven, Belgium
Jerry C. Chang • Department of Chemistry, Vanderbilt University, Nashville,
TN, USA
Nada M. Dimitrijevic • Center of Nanoscale Materials, Argonne National Laboratory,
Argonne, IL, USA
Jose Amado Dinglasan • Vive Nano, Inc., Toronto, ON, Canada
Aaron C. Eifler • Feinberg School of Medicine, Northwestern University,
Chicago, IL, USA
Rutledge G. Ellis-Behnke • Department of Brain and Cognitive Sciences,
Massachusetts Institute of Technology, Cambridge, MA, USA
Erik Fisher • School of Politics and Global Studies, Consortium for Science,
Policy & Outcomes, Center for Nanotechnology in Society, Arizona State
University, Tempe, AZ, USA
Stefan Franzen • Department of Chemistry, North Carolina State University,
Raleigh, NC, USA
Di Gao • Department of Chemical and Petroleum Engineering, University
of Pittsburgh, Pittsburg, PA, USA
Jered B. Haun • Center for Systems Biology, Massachusetts General Hospital,
Boston, MA, USA
Sarah J. Hurst • Center for Nanoscale Materials, Argonne National Laboratory,
Argonne, IL, USA
Roshan James • Department of Orthopaedic Surgery, Department of Chemical,
Materials and Biomolecular Engineering, University of Connecticut Health Center,
Farmington, CT, USA
ix
x Contributors
Fabrice Jotterand • Division of Ethics and Health Policy, Department of Clinical

Sciences and Psychiatry, UT Southwestern Medical Center, Dallas, TX, USA
Raoul Kopelman • Department of Chemistry, University of Michigan,
Ann Arbor, MI, USA
Kostas Kostarelos • Centre for Drug Delivery Research, The School of Pharmacy,
Qin Kuang • School of Materials Science and Engineering, Georgia Institute
of Technology, Altanta, GA, USA
Sangamesh G. Kumbar • Department of Orthopaedic Surgery, Department
of Chemical, Materials and Biomolecular Engineering, University of Connecticut
Health Center, Farmington, CT, USA
Cato T. Laurencin • Department of Orthopaedic Surgery, Department
of Chemical, Materials and Biomolecular Engineering, University of Connecticut
Hakho Lee • Center for Systems Biology, Massachusetts General Hospital,
Boston, MA, USA
Jae-Seung Lee • Department of Materials Science and Engineering,
Korea University, Seoul, Republic of Korea
One-Sun Lee • Department of Chemistry, Northwestern University,
Evanston, IL, USA
Yong-Eun Koo Lee • Department of Chemistry, University of Michigan,
Ann Arbor, MI, USA
Hongwei Liao • Department of Chemistry and Biochemistry,
University of Maryland, College Park, MD, USA
Frances S. Ligler • Center for Bio/Molecular Science and Engineering,
Dustin Lockney • Department of Chemistry, North Carolina State University,
Raleigh, NC, USA
Steven Lommel • Department of Plant Pathology, North Carolina State University,
Raleigh, NC, USA
Matthew McDermott • Akina Inc., West Lafayette, IN, USA
Igor L. Medintz • Center for Bio/Molecular Science and Engineering,
Gwen E. Owens • Department of Molecular and Medical Pharmacology, Crump
Institute for Molecular Imaging, Institute for Molecular Medicine,
University of California, Los Angeles, Los Angeles, CA, USA
Bhavna Paratala • Department of Biomedical Engineering, State University of New
York at Stony Brook, Stony Brook, NY, USA
Haesun Park • Departments of Biomedical Engineering and Pharmaceutics,
Purdue University, West Lafayette, IN, USA
Kinam Park • Departments of Biomedical Engineering and Pharmaceutics,
Purdue University, West Lafayette, IN, USA; Akina Inc., West Lafayette,
IN, USA
Robin Phelps • School of Public Affairs, University of Colorado, Denver,
Denver, CO, USA
Duane E. Prasuhn • Center for Bio/Molecular Science and Engineering,
Contributors xi
Tijana Rajh • Center for Nanoscale Materials, Argonne National Laboratory,

Argonne, IL, USA
Kenneth L. Reed • DuPont Haskell Global Centers for Health and Environmental
Sciences, Newark, DE, USA
Sandra J. Rosenthal • Department of Chemistry, Department of Pharmacology,
Department of Chemical and Biomolecular Engineering, Department of Physics
and Astronomy, Vanderbilt University, Nashville, TN, USA
Elena A. Rozhkova • Center for Nanoscale Materials, Argonne National
Laboratory, Argonne, IL, USA
Christie M. Sayes • Department of Veterinary Physiology & Pharmacology,
Texas A&M University, College Station, TX, USA
George C. Schatz • Department of Chemistry, Northwestern University,
Evanston, IL, USA
Gerald E. Schneider • Department of Brain and Cognitive Sciences,
Massachusetts Institute of Technology, Cambridge, MA, USA
Soo Jung Shin • Akina Inc., West Lafayette, IN, USA
Lisa C. Shriver-Lake • Center for Bio/Molecular Science and Engineering,
Balaji Sitharaman • Department of Biomedical Engineering, State University of New
York at Stony Brook, Stony Brook, NY, USA
Kimihiro Susumu • Division of Optical Sciences, U.S. Naval Research Laboratory,
Washington, DC, USA
Douglas J. Sylvester • Sandra Day O’Connor College of Law, Center for the Study
of Law, Science and Technology, Arizona State University, Tempe, AZ, USA
Chris R. Taitt • Center for Bio/Molecular Science and Engineering, U.S. Naval
Research Laboratory, Washington, DC, USA
C. Shad Thaxton • Department of Urology, Feinberg School of Medicine,
Institute for BioNanotechnology and Medicine, and the International Institute
for Nanotechnology, Northwestern University, Chicago, IL, USA
Keith Thomas • Vive Nano, Inc., Toronto, ON, Canada
Michael S. Tomczyk • The Wharton School, The University of Pennsylvania,
Philadelphia, PA, USA
Udaya S. Toti • Department of Orthopaedic Surgery, Department of Chemical,
Materials and Biomolecular Engineering, University of Connecticut
Hsian-Rong Tseng • Department of Molecular and Medical Pharmacology,
Crump Institute for Molecular Imaging, Institute for Molecular Medicine,
Lilach Vaks • Department of Molecular Microbiology and Biotechnology, Tel Aviv
University, Ramat Aviv, Israel
Shutao Wang • Department of Molecular and Medical Pharmacology,
Crump Institute for Molecular Imaging, Institute for Molecular Medicine,
Yuhuang Wang • Department of Chemistry and Biochemistry,
University of Maryland, College Park, MD, USA
Zhong Lin Wang • School of Materials Science and Engineering,
Georgia Institute of Technology, Altanta, GA, USA
xii Contributors
David B. Warheit • DuPont Haskell Global Centers for Health and Environmental
Sciences, Newark, DE, USA
Yaguang Wei • School of Materials Science and Engineering,
Georgia Institute of Technology, Altanta, GA, USA
Ralph Weissleder • Center for Systems Biology, Department of Systems Biology,
Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
Ashish Yeri • Department of Chemical and Petroleum Engineering,
University of Pittsburgh, Pittsburg, PA, USA
Tae-Jong Yoon • Center for Systems Biology, Massachusetts General Hospital,
Boston, MA, USA
Jun Zhou • School of Materials Science and Engineering, Georgia Institute
of Technology, Altanta, GA, USA
Chapter 1
Sarah J. Hurst
Abstract
This chapter summarizes the roles of nanomaterials in biomedical applications, focusing on those
highlighted in this volume. A brief history of nanoscience and technology and a general introduction to
the field are presented. Then, the chemical and physical properties of nanostructures that make them ideal
for use in biomedical applications are highlighted. Examples of common applications, including sensing,
imaging, and therapeutics, are given. Finally, the challenges associated with translating this field from the
research laboratory to the clinic setting, in terms of the larger societal implications, are discussed.
Key words: Nanoparticles, Nanodevices, Biomedical nanotechnology, Biodetection, Nanothera-

peutics, Implant materials
1. Introduction
Nanoscience and technology is a field that focuses on developing

new synthetic and analytical tools for building and studying struc-
tures with submicrometer, and more typically sub-100 nm dimen-
sions (1 nm = 1 × 10−9 m) (see Fig. 1) (1, 2). Moreover, it is
concerned with the study of the chemical and physical properties
of such structures and how these properties change as the size of
a material is scaled down from the bulk to a collection of several
atoms. Finally, nanoscience and technology centers on utilizing
the capabilities and the fundamental property differences associ-
ated with such highly miniaturized structures to construct novel
functional materials and devices and to develop ground-breaking
applications. The nanoscale is a length scale that falls between
that of traditional chemistry and physics, which deals with the
manipulation of atomic bonds (~10−10 m) and that of biology,
where most structures are on the order of ~10−6 to 10−7 m in
Sarah J. Hurst (ed.), Biomedical Nanotechnology: Methods and Protocols, Methods in Molecular Biology, vol. 726,
DOI 10.1007/978-1-61779-052-2_1, © Springer Science+Business Media, LLC 2011
1
2 Hurst
Fig. 1. Length scales.
diameter (e.g., cells, viruses, and bacteria), and as a consequence

the field is highly interdisciplinary, encompassing aspects of
physics, chemistry, biology, engineering, and medicine.
Although a boom in nanoscience and technology research
occurred only recently, nanomaterials, in particular metallic nano-
particles, have been used for centuries in art (i.e., the Lycurgus
cup), architecture (i.e., stained glass windows), photography (i.e.,
the developing process), and medicinal remedies (3, 4). In the
late 1800s, Michael Faraday discovered and developed reliable
syntheses for pseudo-spherical gold nanoparticles and later (in
1908) a theoretical framework for understanding of their optical
properties was put in place by Gustav Mie. In the mid-1900s,
enabling technologies for the imaging and manipulation of atoms
and nanostructures were pioneered as atomic force, and electron
microscopes were designed and put to use. It was also during this
time that biologists were unraveling cellular structure and discov-
ering a multitude of biological species (e.g., DNA and proteins)
that have nanoscale dimensions. Further, the advent of novel
electronic components such as transistors was ushering in the age
of high-powered computing. Such findings led prominent scien-
tists such as Richard Feynman and others to speculate that nano-
science and technology would be a revolutionary new research
direction for many fields of science (5).
Today, one of the main thrusts of nanoscience research is the
synthesis of novel nanoparticle materials and devices. Solution-
based syntheses exist for making monodisperse samples of both
isotropic and anisotropic metallic, semiconducting, and polymeric
nanoparticles of a variety of shapes (6) including spheres (7–10),
disks (11–13), prisms (14, 15), cubes (16–18), wires (19), rods
Biomedical Nanotechnology 3
Fig. 2. Electron microscopy images of some common nanostructures. (b) Adapted with permission from Nano Lett.
(2009), 9, 3116. Copyright 2009 American Chemical Society. (c) Adapted with permission from J. Phys. Chem. B (2007),
111, 1249. Copyright 2007 American Chemical Society. (d) Adapted with permission from J. Am. Chem. Soc. (2008), 130,
14958. Copyright 2008 American Chemical Society. (e) Adapted with permission from J. Am. Chem. Soc. (2005), 127,
5312. Copyright 2005 American Chemical Society.
(20–23), and branched (24–26) structures (see Fig. 2). Techniques

such as arc discharge, laser ablation, and chemical vapor deposition
(CVD) are being used to create carbon-based nanomaterials (i.e.,
fullerenes and carbon nanotubes) (27, 28) (see Chapter 15).
Surface-based techniques (e.g., electrospinning (29), lithography
(30–32), and templation (20)) are being utilized to make surface-
bound nanostructures (33) (see Chapters 8 and 10), nanostruc-
tured scaffolding materials (see Chapters 15–17) (34–36),
nanoelectromechanical devices (NEMS) (see Chapter 9) (37, 38),
and nanofluidic devices (39). The nanomaterials produced using
these processes are being applied in electronics (40), catalysis (41,
42), energy storage and generation (43, 44), environmental
remediation (45, 46), security (47), and especially in biology and
medicine (48–50).
In particular, this volume focuses on the application of the
above types of nanomaterials in biomedicine. This chapter pro-
vides a brief introduction to this field. It highlights the common
properties of such structures and the main advantages that they
can offer compared to conventionally used materials. It then dis-
cusses the key applications in which these structures are used,
4 Hurst
namely, sensing, imaging, and therapeutics. Finally, it gives a

future perspective of this field, not only in terms of new research
directions, but also in terms of the larger societal implications of
and challenges associated with transitioning nanoscience and
technology products and strategies from the research laboratory
to the clinical setting.
2. Nanomaterials
in Medicine
The types of nanoparticles and nanodevices that are utilized in
biomedical applications are chemically and physically diverse, but
despite this diversity, they share several commonalities that make
them advantageous compared to conventionally used structures.
First, nanomaterials are small in size, having at least one
dimension (e.g., particle diameter or feature size) between 1 and
100 nm (1). Due to their small size, these nanostructures have
high surface-to-volume ratios and hence are very reactive both
during and after their synthesis. This property in part makes them
highly tailorable and since their chemical and physical properties
depend on their size, shape, and composition, important param-
eters including their charge, hydrophobicity, solubility, and stabil-
ity can be easily tuned. For example, a given nanostructure could
be designed to be either structurally robust (51) or easily biode-
gradable (52) over a certain period of time in a biological envi-
ronment. Also, their small size allows nanostructures to readily
interact with biological entities, which have similar dimension.
Nanoparticles have been shown to be taken into cells through the
pores of their membranes (53) and even localized in particular
areas of the cell (54). They also have been known to cross the
blood–brain-barrier through tight biological junctions unlike
larger macrosized objects (55).
In addition, the nanomaterials employed in biomedical appli-
cations usually are multicomponent in nature (20, 49). Often, a
nanoparticle or nanostructure is conjugated to one or more types
of chemical and/or biological species such as oligonucleotides
(short, synthetic DNA strands), proteins, drugs, or lipids through
techniques including chemical conjugation, encapsulation, infu-
sion, or adsorption (56, 57) (see Chapters 6 and 7). For instance,
the gold nanoparticles discussed in Chapter 2 are functionalized
with oligonucleotides of more than one sequence (58) while the
bacteriophages discussed in Chapter 13 are modified with anti-
bodies as well as the hydrophobic drug, chloramphenicol (59).
Further, the nanomaterial itself is often composed of two or more
inorganic components or a combination of inorganic and organic
components in an alloy, core-shell, multishell, or dumb-bell
arrangement (60, 61). In Chapter 3, for instance, the utilized
nanomaterial is a particle with an iron-oxide core and a cross-linked

dextran shell (62, 63). These multicomponent nanostructures
retain the chemical and physical properties of their individual
parts but also are imparted with a synergistic set of properties
resulting from their combination.
These primary considerations make nanomaterials advanta-
geous compared to conventional materials in many cases. As a
result of their unique nanoscale properties, quantum dots, for
example, have higher quantum yields, narrower emission wave-
lengths, and better photostability than conventional fluorescence
dyes; their emission also can be easily tuned by varying their com-
position (64). As a result, detection systems employing quantum
dots can be more sensitive than those utilizing conventional dyes,
pushing them toward single molecule level detection (see Chapter 4).
In another example, the high density of oligonucleotides on the
surface of a metal nanoparticle in part allows them to more read-
ily enter cells (53, 65), be more resistant to degradation (66),
have higher target binding properties (67), and produce a weaker
immune response compared to systems using traditional gene
delivery agents where DNA is not concentrated on a nanostruc-
tured surface (68) (see Chapter 2). The details underpinning such
differences are being investigated computationally (see Chapter
18) as well as experimentally.
3. Applications
of Nanomaterials
in Medicine
The main applications of nanomaterials in biology are in the areas of
sensing, imaging, and therapeutics. Because of their multicomponent
structure, many nanoparticles can carry out more than one of these
tasks simultaneously (i.e., multifunctionality) (see Subheading 3.3).
3.1. Sensing Most of the sensing and imaging applications of nanomaterials in

and Imaging biomedicine are based on a two-part process: (1) a self-assembly
event and (2) a readout event (see Fig. 3a) (48). Typically, the
first step is a chemical or biological recognition event which takes
place between the molecule attached to the nanomaterial and a
target species in a sample of interest. Some common interactions
that are exploited are oligonucleotide base pairing, protein–protein,
and peptide–protein interactions as well as the interactions of
oligos, proteins, and peptides with small molecules (69). For
example, Chapter 4 describes a protocol in which proteins of
interest in a cell sample are labeled with antibodies (against pro-
teins including glycine receptors, nerve growth factors, kinesin
motors, or GABA receptors, for instance) possessing a biotin tag;
these are then exposed to quantum dots modified with streptavi-
din moieties. The biotin and streptavidin units participate in a
6 Hurst
Fig. 3. General scheme for nanoparticle-based (a) biodetection and (b) therapeutics.
highly specific binding interaction and the quantum dot label is

bound to the protein of interest. In another case, silicon sub-
strates possessing nanopillar features were modified with epithe-
lial cell adhesion antibodies (see Chapter 10) (70). When solutions
of cells then were exposed to these substrates, the antibodies
interact with proteins on the surface of the cell, trapping them on
the substrate where they can be further analyzed.
Once the target species has been recognized, the next step in
the application is typically a readout event, which often involves the
nanomaterial. The readout can be accomplished using colorimetry
(71, 72), fluorescence (73), light scattering (74), radiochemistry- (75)
or Raman-based approaches (76, 77), magnetic resonance (78), or
electrical signal (79) to name a few. The detection of the surface
plasmon resonance (SPR) or fluorescence emission signal of a metal
or semiconducting particle probe, respectively, could be an indica-
tor of the presence of a given target in a sample. A change in the
electronic properties (e.g., conductance) of a fabricated microelec-
trode linked with a semiconducting nanowire could be indicative of
target binding at its surface (see Chapter 8) (79). The difference in
magnetic relaxitivity of superparamagnetic particles could be used
to detect whether they are dispersed (and not bound to the target
molecule) and aggregated (bound to the target molecule of interest)
(see Chapter 3).
In these detection strategies, selectivity and sensitivity are
important considerations (48, 80). Nonspecific binding should
be minimized and care should be taken to ensure that the mole-
cules are oriented on the nanomaterial in such a way that they
maintain their biological function (i.e., the ability to participate in
subsequent binding interactions). The level of sensitivity attain-
able in a readout method varies; some methods are more sensitive
than others. Many nanomaterial-based biodetection strategies
have been shown to rival or even surpass the current limits of
detection of conventional techniques [e.g., enzyme-linked

immunoabsorbant assay (ELISA), polymerase chain reaction
(PCR)]. In addition to advantages in terms of sensitivity and
selectivity, nanomaterial-based detection schemes can be easier,
faster, and cheaper than conventional strategies. Also, they have
the potential to be more readily portable, making them useful in
field deployable or point-of-care systems.
3.2. Therapeutics Nanotherapeutic strategies are being used to treat diseases rang-
ing from genetic disorders to cancer (50, 81). Similar to the
detection strategies, these schemes often involve two major steps
(see Fig. 3b). In the first step, the nanomaterial is targeted to
specific cells or tissues. This targeting step is vital to ensure that
the treatment is not delivered to healthy cells, causing unwanted
side effects. Targeting can be accomplished using passive or active
methods (50). Passive methodology involves tuning the physical
and chemical properties (i.e., size and charge) of the material to
trap the nanomaterials inside tumor sites, for instance, by exploit-
ing the physiological differences between tumor cells and healthy
cells (e.g., vasculature and pH). In active targeting schemes, the
material is functionalized with species capable of participating in
chemical and biological recognition events with the cells being
targeted This volume presents one example of active targeting in
Chapter 14, where the role of surface-bound peptides in target-
ing of viruses is discussed.
After the material is targeted to the cells or tissues of interest,
a therapeutic payload can be released or a non-drug-based therapy
can be initiated. Small molecule drugs, doped inside polymer par-
ticles can be gradually released as the particle degrades, for example
(82) (see Chapter 12). The inherent assembly/disassembly prop-
erties of a plant virus nanoparticle capsid can be exploited to load
and then unload drug molecules at targeted locations (see Chapter
14). Nanoparticle-based schemes are also replacing traditional
gene therapy approaches and, in these cases, the payload is plasmid
DNA, oligonucleotides, or siRNA (83). Other methods use light
or heat (e.g., photodynamic or photothermal therapy) (84, 85) to
initiate cytotoxic effects. In Chapter 5, titanium dioxide and semi-
conducting nanoparticles (~5 nm in diameter), modified with
cancer-targeting proteins, are discussed (86). These particles are
capable of generating cytotoxic reactive oxygen species (ROS)
when excited by light of a particular wavelength (87).
In another kind of therapeutic scheme, nanostructured
implant materials are utilized in the regeneration of organ and
nervous system tissue (e.g., brain and spinal cord) (88, 89).
Here, the nanomaterial can deliver a drug cargo but primarily
provides a scaffold for the mitigation of cell growth and develop-
ment. These materials can be synthesized and then surgically
placed at the site of injured tissue or administered as nanoscale
precursor materials, capable of assembling into scaffolds in vivo.
8 Hurst
Three examples of nanostructured scaffolds are presented in this

volume: carbon nanotube-reinforced polymer nanocomposites,
polymeric, nanofiberous scaffolds, synthesized using electrospin-
ning, and porous biodegradable scaffolds, synthesized via the
self-assembly of peptide amphiphiles (see Chapters 15–17,
respectively).
3.3. Combined Current research is focusing more and more on the development
Sensing/Imaging of multicomponent nanomaterials that perform both sensing
and Therapy and/or imaging and therapy concomitantly (90, 91). These
nanomaterials can potentially contain multiple types of targeting
moieties, drugs, and dyes or contrast agents all on one surface
(see Fig. 4). Several examples of the use of such particles are pre-
sented in Chapter 11, which highlights photodynamic therapies.
Inorganic, photon-upconverting particles made of NaYF4:Yb3+,
Er3+ and coated with the polymer polyethyleneimine (PEI) can be
loaded with photosensitizers (92). Not only did these particles
demonstrate the destruction of specific cells with near-IR excita-
tion but they also produced red/green emission allowing them to
be imaged both in vitro and in vivo. Using these types of systems,
one could imagine designing structures that are capable of pro-
viding real-time feedback and diagnostic information as a treat-
ment which is being administered. In this way, treatments and
dosages could be tailored on a patient-by-patient basis, moving
towards personalized medicine.
As these applications are being further developed and begin-
ning to enter and navigate the US Food and Drug Administration
(FDA) approval process (see Chapter 21), it is becoming impor-
tant to assess not only their limits of detection or therapeutic
effectiveness but also the toxicology profiles of the novel nano-
materials used (64, 93, 94). Due to their structural complexity,
they must be assessed on a case-by-case basis. The same particle
functionalized with different ligands could have different toxicity
profiles, for example, or be metabolized differently in the body;
Fig. 4. General schematic of a multicomponent nanoparticle showing targeting moieties,

dyes, and drugs on one surface.
the components of the degradation of these complex materials

also warrant study. In addition, it is being discovered that
nanomaterials can interfere with components of traditional assays
that are used to assess nanoparticle toxicology and safety,
necessitating the development of novel or modified methods of
analysis. Strides are being made in this direction (see Chapters 19
and 20). Chapter 19 assesses cell injury caused by exposure to
carbon nanotubes. Chapter 20 introduces a protocol for assessing
the pulmonary hazards of nanoparticles (both fine and ultrafine,
pigmentary and nano-sized) using biochemical and histopatho-
logical evaluations.
4. Future
Perspectives
The field of biomedical nanotechnology is growing at an unprec-
edented rate (4, 95, 96). More research dollars are being spent
on this industry than ever before both in the USA and around
the world. It is the source of numerous statewide and global col-
laborations. The patenting of intellectual property in this area is
immense. Government and university programs are being devel-
oped and user facilities geared toward nanotechnology research
have been opened in national laboratory settings in the USA.
Nanotechnology startups are being spun out of university
research laboratories and larger companies are adopting nano-
materials into their everyday processes. Nanomaterials are already
used in products ranging from sunscreen to golf clubs, and the
first wave of nanotechnology-based drugs is being approved by
the FDA.
The rapid growth of research in nanoscience and technology
not only looks to revolutionize the scientific community, but also
the business, social science, and the legislative and legal communi-
ties, responsible in part for disseminating nanoscience knowledge,
products, and applications into society. The multicomponent and
multifunctional nature of the nanomaterials that are used in bio-
medicine leads to key advantages; however, this complexity also
leads to complications for decision-makers in these fields. Business
people are wrangling with how to most beneficially expand
companies around (see Chapter 24) and market (see Chapter 25)
nanotechnology products. Politicians are determining how to
legislate them such that ground-breaking technology will be
quickly moved from the laboratory to the public arena but also in
a manner that is responsible and beneficial to both the general
public and the environment (see Chapter 22). Lawyers are working
to protect the intellectual property of nanoscience researchers (see
Chapter 23) and ethicists struggle with the potential moral issues
that these novel types of medicines bring up (see Chapter 26).
10 Hurst
The continued growth of the field of nanoscience and

technology relies almost as much on the actions of these decision-
makers as on the success of the research projects carried out by
scientists in the laboratory. These types of decisions are partially
responsible for shaping public opinion and, as mentioned repeat-
edly in this volume, a single negative incident has in the past been
sufficient to cause a general public mistrust of a particular field of
science. Therefore, this volume incorporates a section of chapters
written by experts in the above mentioned fields highlighting
some of the major challenges that they are facing, in order to
foster the translation of knowledge between researchers from the
natural science and social science fields. In this way, this volume
intends to present a complete picture of the current state of
biomedical nanoscience and technology from the laboratory to
the clinic.
Acknowledgements
S.J.H. would like to thank Dr. H. Christopher Fry for his careful
editing of this chapter and Dr. Tijana Rajh for being an outstand-
ing mentor and advisor. S.J.H also acknowledges Argonne
National Laboratory for a Director’s Postdoctoral Fellowship.
Work at the Center for Nanoscale Materials was supported by the
US Department of Energy, Office of Science, Office of Basic
Energy Sciences, under contract no. DE-AC02-06CH11357.
References
1. Mirkin, C. A. (2005) The Beginning of a Spherical Nanocrystals. Angew. Chem. Int. Ed.
Small Revolution. Small 1, 14–16. 46, 4630–4660.
2. Mirkin, C. A. (1999) Tweezers for the 8. Smith, D., Pentzer, E., and Nguyen, S. T.
Nanotool Kit. Science 286, 2095–2096. (2007) Bioactive and Therapeutic ROMP
3. Daniel, M.-C. and Astruc, D. (2004) Gold Polymers. Polym. Rev. 47, 419–459.
Nanoparticles: Assembly, Supermolecular 9. Cushing, B. L., Kolesnichenko, V. L., and
Chemistry, Quantum-Size-Related Properties, O’Connor, C. J. (2004) Recent Advances in
and Applications Toward Biology, Catalysis, and the Liquid-Phase Syntheses of Inorganic
Nanotechnology. Chem. Rev. 104, 293–346. Nanoparticles. Chem. Rev. 104, 3893–3946.
4. Ratner, M. and Ratner, D. (2003) 10. Oh, M. and Mirkin, C. A. (2005) Chemically
Nanotechnology: A Gentle Introduction to the Tailorable Colloidal Particle from Infinite
Next Big Idea. Prentice Hall, Upper Saddle Coordination Polymers. Nature 438, 651–654.
River, NJ. 11. Maillard, M., Giorgio, S., and Pileni, M.-P. (2002)
5. Feynman, R. (1999) The Pleasure of Finding Silver Nanodisks. Adv. Mater. 14, 1084–1086.
Things Out. Perseus Books, New York, NY. 12. Hao, E., Kelly, K. L., Hupp, J. T., and Schatz,
6. Burda, C., Chen, X., Narayanan, R., and G. C. (2002) Synthesis of Silver Nanodisks
El-Sayed, M. (2005) Chemistry and Properties Using Polystyrene Mesospheres as Templates.
of Nanocrystals of Different Shapes. Chem. J. Am. Chem. Soc. 124, 15182–15183.
Rev. 105, 1025–1102. 13. Chen, S. and Carroll, D. L. (2002) Synthesis
7. Park, J., Joo, J., Kwon, S. G., Jang, Y., and and Characterization of Truncated Triangular
Hyeon, T. (2007) Synthesis of Monodisperse Silver Nanoplates. Nano Lett. 2, 1003–1007.
14. Hulteen, J. C. and Van Duyne, R. P. (1995) 28. Dai, H. J. (2002) Carbon Nanotubes:
Nanosphere Lithography: A Materials General Synthesis, Integration and Properties. Acc.
Fabrication Process for Periodic Particle Array Chem. Res. 35, 1035–1044.
Surfaces. J. Vac. Sci. Technol. A 13, 29. Lu, X., Wang, C., and Wei, Y. (2009) One-
1553–1558. Dimensional Composite Nanomaterials:
15. Millstone, J. E., Hurst, S. J., Metraux, G. S., Synthesis by Electrospinning and Their
Cutler, J. I., and Mirkin, C. A. (2009) Applications. Small 5, 2349–2370.
Colloidal Gold and Silver Nanoprisms. Small 30. Gates, B. D., Xu, Q., Stewart, M., Ryan, D.,
5, 646–664. Willson, C. G., and Whitesides, G. M. (2005)
16. Ahmadi, T. S., Wang, Z. L., Green, T. C., New Approaches to Nanofabrication:
Henglein, A., and El-Sayed, M. (1996) Molding, Printing, and Other Techniques.
Shape-Controlled Synthesis of Colloidal Chem. Rev. 105, 1171–1196.
Platinum Nanoparticles. Science 272, 31. Ginger, D. S., Zhang, H., and Mirkin, C. A.
1924–1926. (2004) The Evolution of Dip-Pen Nano
17. Sun, Y. and Xia, Y. (2002) Shape-Controlled lithography. Angew. Chem. Int. Ed. 43,
Synthesis of Gold and Silver Nanoparticles. 30–45.
Science 298, 2176–2179. 32. Euliss, L. E., DuPont, J. A., Gratton, S., and
18. Yu, D. and Yam, V. W.-W. (2004) Controlled DeSimone, J. (2006) Imparting Size, Shape,
Synthesis of Monodisperse Silver Nanocubes and Composition Control of Materials for
in Water. J. Am. Chem. Soc. 126, Nanomedicine. Chem. Soc. Rev. 35,
13200–13201. 1095–1104.
19. Hu, J., Odom, T. W., and Lieber, C. M. 33. Hayes, C. L. and Van Duyne, R. P. (2001)
(1999) Chemistry and Physics in One Nanosphere Lithography: A Versitile
Dimension: Synthesis and Properties of Nanofabrication Tool for Studies of Size-
Nanowires and Nanotubes. Acc. Chem. Res. Dependent Nanoparticle Optics. J. Phys.
32, 435–445. Chem. B 105, 5599–5611.
20. Hurst, S. J., Payne, E. K., Qin, L., and Mirkin, 34. Tang, Z., Wang, Y., Podsiadlo, P., and Kotov,
C. A. (2006) Multisegmented One- N. A. (2006) Biomedical Applications of
Dimensional Nanorods. Angew. Chem. Int. Layer-by-Layer Assembly: From Biomimetics
Ed. 45, 2672–2692. to Tissue Engineering. Adv. Mater. 18,
21. Peng, X., Manna, L., Yang, W., Wickham, J., 3203–3224.
Scher, E., Kadavanich, A., et al. (2000) Shape 35. Wei, G. and Ma, P. X. (2009) Nanostructured
Control of CdSe Nanocrystals. Nature 404, Biomaterials for Regeneration. Adv. Funct.
59–61. Mater. 18, 3568–3582.
22. Jana, N. R., Gearheart, L., and Murphy, C. J. 36. Yoo, P. J., Nam, K. T., Qi, J. F., Lee. S.-K.,
(2001) Wet Chemical Synthesis of High Park, J., Belcher, A. M., et al. (2006)
Aspect Ratio Cylindrical Gold Nanorods. Spontaneous Assembly of Viruses on
J. Phys. Chem. B 105, 4065–4067. Multilayered Polymer Surfaces. Nat. Mater.
23. Kim, F., Song, J. H., and Yang, P. (2002) 5, 234–240.
Photochemical Synthesis of Gold Nanorods. 37. Craighead,H.G.(2000)Nanoelectromechanical
J. Am. Chem. Soc. 124, 14316–14317. Systems. Science 290, 1532–1535.
24. Manna, L., Milliron, D. J., Meisel, A., Scher, 38. Li, M., Tang, H. X., and Roukes, M. L.
E. C., and Alivisatos, A. P. (2003) Controlled (2007) Ultra-Sensitive NEMS-Based
Growth of Tetrapod-Branched Inorganic Cantilevers for Sensing Scanned Probe and
Nanocrystals. Nat. Mater. 2, 382–385. Very High-Frequency Applications. Nat.
25. Chen, S., Wang, Z. L., Ballato, J., Foulger, S. H., Nanotechnol. 2, 114–120.
and Carroll, D. L. (2003) Monopod, Bipod, 39. Sparreboom, W., van den Berg, A., and Eijkel,
Tripod, and Tetrapod Gold Nanocrystals. J. C. T. (2009) Principles and Applications of
J. Am. Chem. Soc. 125, 16186–16187. Nanofluidic Transport. Nat. Nanotechnol. 4,
26. Hao, E., Bailey, R. C., Schatz, G. C., Hupp, J. 713–720.
T., and Li, S. (2004) Synthesis and Optical 40. Lu, W. and Lieber, C. M. (2007)
Properties of “Branched” Gold Nanocrystals. Nanoelectronics from the Bottom Up. Nature
Nano Lett. 4, 327–330. 6, 841–850.
27. Zhou, W., Bai, X., Wang, E., and Xie, S. 41. Bonnemann, H. and Richards, R. M. (2001)
(2009) Synthesis, Structure, and Properties of Nanoscopic Metal Particles – Synthetic
Single-Walled Carbon Nanotubes. Adv. Mater. Methods and Potential Applications. Eur.
21, 4565–4583. J. Inorg. Chem. 2001, 2455–2480.
12 Hurst
42. Bell, A. T. (2003) The Impact of Nanoscience Synthesis, Properties, and Applications.
on Heterogeneous Catalysis. Science 299, Angew. Chem. Int. Ed. 43, 6042–6108.
1688–1691. 57. Hermanson, G. T. (2008) Bioconjugate
43. Zhong, C.-J., Luo, J., Fang, B., Wanjala, B. Techniques (2nd ed.). Academic Press, New
N., Njoki, P. N., Rameshwori, L., et al. (2010) York, NY.
Nanostructured Catalysts in Fuel Cells. 58. Stoeva, S. I., Lee, J.-S., Thaxton, C. S., and
Nanotechnology 21, 062001. Mirkin, C. A. (2006) Multiplexed DNA
44. Talapin, D. V., Lee, J.-S., Kovalenko, M. V., Detection with Biobarcoded Nanoparticle
and Shevchenko, E. V. (2010) Prospects of Probes. Angew. Chem. Int. Ed. 45, 3303–3306.
Colloidal Nanocrystals for Electronic and 59. Yacoby, I., Bar, H., and Benhar, I. (2007)
Optoelectronic Applications. Chem. Rev. 110, Targeted Drug Carrying Bacteriophages as
389–458. Antibacterial Nanomedicines. Antimicrob.
45. Savage, N. and Diallo, M. S. (2005) Agents. Chemother. 51, 2156–2163.
Nanomaterials and Water Purification: 60. Cozzoli, P. D., Pellegrino, T., and Manna, L.
Opportunities and Challenges. J. Nanopart. (2006) Synthesis, Properties and Perspectives
Res. 7, 331–342. of Hybrid Nanocrystal Structures. Chem. Soc.
46. Pradeep, T. and Anshup (2009) Nobel Metal Rev. 35, 1195–1208.
Nanoparticles for Water Purification: A 61. Zou, H., Wu, S., and Shen, J. (2008)
Critical Review. Thin Solid Films 517, Polymer/Silica Nanocomposites: Preparation,
6441–6478. Characterization, Properties and Applications.
47. Golightly, R. S., Goering, W. E., and Natan, Chem. Rev. 108, 3893–3957.
M. J. (2009) Surface-Enhanced Raman 62. Josephson, L., Tung, C. H., Moore, A., et al.
Spectroscopy and Homeland Security: A (1999) High-Efficiency Intracellular Magnetic
Perfect Match? ACS Nano 3, 2859–2869. Labeling with Novel Superparamagnetic-Tat
48. Rosi, N. L. and Mirkin, C. A. (2005) Peptide Conjugates. Bioconjug. Chem. 10,
Nanostructures in Biodiagnostics. Chem. Rev. 186–191.
105, 1547–1562. 63. Harisinghani, M. G., Barentsz, J., Hahn, P.
49. Gao, J., Gu, H., and Xu, B. (2009) F., Deserno, W. M., Tabatabaei, S., Hulsbergen
Multifunctional Magnetic Nanoparticles: van de Kaa, C., et al. (2003) Noninvasive
Design, Synthesis and Biomedical Applications. Detection of Clinically Occult Lymph-Node
Acc. Chem. Res. 42, 1097–1107. Metastases in Prostate Cancer. N. Engl.
50. Peer, D., Karp, J. M., Hong, S., Farokhzad, J. Med. 348, 2491–2499.
O. C., Margalit, R., and Langer, R. (2007) 64. Smith, A. M., Duan, H., Mohs, A. M., and
Nanocarriers as an Emerging Platform for Nie, S. (2008) Bioconjugated Quantum Dots
Cancer Therapy. Nat. Nanotechnol. 2, for In Vivo Molecular and Cellular Imaging.
751–760. Adv. Drug Deliv. Rev. 60, 1226–1240.
51. Wu, H., Zhu, H., Zhuang, J., Yang, S., Liu, 65. Giljohann, D. A., Seferos, D. S., Patel, P. C.,
C., and Cao, Y. C. (2008) Water-Soluble Millstone, J. E., Rosi, N. L., and Mirkin, C. A.
Nanocrystals with Dual-Interaction Ligands. (2007) Oligonucleotide Loading Determines
Angew. Chem. Int. Ed. 47, 3730–3734. Cellular Uptake of DNA-Modified Gold
52. Pridgen, E. M., Langer, R., and Farokhzad, Nanoparticles. Nano Lett. 7, 3818–3821.
O. C. (2007) Biodegradable, Polymeric 66. Seferos, D. S., Prigodich, A. E., Giljohann, D.
Nanoparticle Delivery Systems for Cancer A., Patel, P. C., and Mirkin, C. A. (2009)
Therapy. Nanomedicine 2, 669–680. Polyvalent DNA-Nanoparticle Conjugates
53. Rosi, N. L., Giljohann, D. A., Thaxton, C. S., Stabilize Nucleic Acids. Nano Lett. 9,
Lytton-Jean, A. K. R., Han, M. S., and Mirkin, 308–311.
C. A. (2006) Oligonucleotide-Modified Gold 67. Lytton-Jean, A. K. R. and Mirkin, C. A.
Nanoparticles for Intracellular Gene (2005) A Thermodynamic Investigation into
Regulation. Science 312, 1027–1030. the Binding Properties of DNA Functionalized
54. Patel, P. C., Giljohann, D. A., Seferos, D. S., Gold Nanoparticle Probes and Molecular
and Mirkin, C. A. (2008) Peptide Antisense Fluorophore Probes. J. Am. Chem. Soc. 127,
Nanoparticles. Proc. Natl. Acad. Sci. U.S.A 12754–12755.
105, 17222–17226. 68. Massich, M. D., Giljohann, D. A., Seferos, D.
55. Silva, G. A. (2006) Neuroscience S., Ludlow, L. E., Horvath, C. M., and
Nanotechnology: Progress, Opportunities, and Mirkin, C. A. (2009) Regulating Immune
Challenges. Nat. Rev. Neurosci. 7, 65–74. Response Using Polyvalent Nucleic Acid-
56. Katz, E. and Willner, I. (2004) Integrated Gold Nanoparticle Conjugates. Mol. Pharm.
Nanoparticle-Biomolecule Hybrid Systems: 6, 1934–1940.
69. Bloomfield, V. A., Crothers, D. M., and Microparticles. J. Control. Release 141,
Tinoco, Jr., I. (2000) Nucleic Acids: Structures, 314–319.
Properties, and Functions. University Science 83. Ditto, A. J., Shah, P. N., and Yun, Y. H.
Books, Sausalito, CA. (2009) Non Viral Gene Delivery Using
70. Wang, S., Wang, H., Jiao, J., Chen, K.-J., Nanoparticles. Exp. Opin. Drug Deliv. 6,
Owens, G. E., Kamei, K.-I., et al. (2009) 1149–1160.
Three-Dimensional Nanostructured Substra 84. Koo, Y.-E., Reddy, G. R., Bhojani, M.,
tes Toward Efficient Capture of Circulating Schneider, R., Philbert, M. A., Rehemtulla,
Tumor Cells. Angew. Chem. Int. Ed. 48, A., et al. (2006) Brain Cancer Diagnosis and
8970–8973. Therapy with Nanoplatforms. Adv. Drug
71. Mirkin, C. A., Letsinger, R. L., Mucic, R. C., Deliv. Rev. 58, 1556–1577.
and Storhoff, J. J. (1996) A DNA-Based 85. Jain, P. K., Huang, X., El-Sayed, I. H., and
Method for Rationally Assembling El-Sayed, M. A. (2008) Nobel Metals on the
Nanoparticles into Macroscopic Materials. Nanoscale: Optical and Photothermal
Nature 382, 607–609. Properties and Some Applications in Imaging,
72. Storhoff, J. J., Elghanian, R., Mucic, R. C., Sensing, Biology and Medicine. Acc. Chem.
Mirkin, C. A., and Letsinger, R. L. (1998) Res. 41, 1578–1586.
One-Pot Colorimetric Differentiation of 86. Rozhkova, E. A., Ulasov, I., Lai, B.,
Polynucleotides with Single Base Imperfections Dimitrijevic, N. M., Lesniak, M. S., and Rajh,
Using Gold Nanoparticle Probes. J. Am. T. (2009) A High-Performance Nanobio
Chem. Soc. 120, 1959–1964. Photocatalyst for Targeted Brain Cancer
73. Zhong, W. (2009) Nanomaterials in Therapy. Nano Lett. 9, 3337–3342.
Fluorescence-Based Biosensing. Anal. 87. Dimitrijevic, N. M., Rozhkova, E., and Rajh,
Bioanal. Chem. 394, 47–59. T. (2009) Dynamics of Localized Charges in
74. Taton, T. A., Mirkin, C. A., and Letsinger, R. Dopamine-Modified TiO and their Effect on
L. (2000) Scanometric DNA Array Detection the Formation of Reactive Oxygen Species.
with Nanoparticle Probes. Science 289, J. Am. Chem. Soc. 131, 2893–2899.
1757–1760. 88. Hartgerink, J. D., Beniash, E., and Stupp, S.
75. Schipper, M. L., Cheng, Z., Lee, S.-W., I. (2001) Self-Assembly and Mineralization of
Bentolila, L. A., Iyer, G., Rao, J., et al. (2007) Peptide-Amphiphile Nanofibers. Science 294,
Micro-PET-Based Biodistribution of 1684–1688.
Quantum Dots in Living Mice. J. Nucl. Med. 89. Zhang, L. and Webster, T. J. (2009)
48, 1511–1518. Nanotechnology and Nanomaterials: Promises
76. Cao, Y. C., Jin, R., and Mirkin, C. A. (2002) for Improved Tissue Regeneration. Nano
Nanoparticles with Raman Spectroscopic Today 4, 66–80.
Fingerprints for DNA and RNA Detection. 90. Cheon, J. and Lee, J.-H. (2008) Synergistically
Science 297, 1536–1540. Integrated Nanoparticles as Multimodal
77. Schlucker, S. (2009) SERS Microscopy: Probes for Nanobiotechnology. Acc. Chem.
Nanoparticle Probes and Biomedical Res. 41, 1630–1640.
Applications. Chemphyschem 10, 1344–1354. 91. Dhar, S., Daniel, W. L., Giljohann, D. A.,
78. Na, H. B., Song, I. C., and Hyeon, T. (2009) Mirkin, C. A., and Lippard, S. J. (2009)
Inorganic Particles for MRI Contrast Agents. Polyvalent Oligonucleotide Gold Nanoparticle
Adv. Mater. 21, 2133–2148. Conjugates as Delivery Vehicles for Platinum
79. Cui, Y., Wei, Q. Q., Park, H. K., and Lieber, C. (IV) Warheads. J. Am. Chem. Soc. 131,
M. (2001) Nanowire Sensors for Highly Sensitive 14652–14653.
and Selective Detection of Biological and 92. Zhang, P., Steelant, W., Kumar, M., and
Chemical Species. Science 293, 1289–1292. Scholfield, M. (2007) Versatile Photosensitizers
80. Giljohann, D. A. and Mirkin, C. A. (2009) for Photodynamics Therapy at Infrared
Drivers of Biodiagnostic Development. Excitation. J. Am. Chem. Soc. 129, 4526-4527.
Nature 462, 461–464. 93. Lewinski, N., Colvin, V., and Drezek, R.
81. Fang, X. and Tan, W. (2010) Aptamers (2008) Cytotoxicity of Nanoparticles. Small
Generated from Cell-SELEX for Molecular 4, 26–49.
Medicine: A Chemical Biology Approach. 94. Maurer-Jones, M. A., Bantz, K. C., Love, S.
Acc. Chem. Res. 43, 48–57. A., Marquis, B. J., and Haynes, C. L. (2009)
82. Acharya, G., Shin, C. S., McDermott, M., Toxicity of Therapeutic Nanoparticles.
Mishra, H., Park, H., Kwon, I.-C., et al. Nanomedicine 4, 219–241.
(2010) The Hydrogel Template Method for 95. http://www.clinicaltrials.gov (2010).
Fabrication of Homogeneous Nano/ 96. http://www.nano.gov (2010).
Part I
Using Nanomaterials in Sensing, Imaging, and Therapeutics

Chapter 2
Multiplexed Detection of Oligonucleotides

with Biobarcoded Gold Nanoparticle Probes
Jae-Seung Lee
Abstract
Applications in a variety of fields rely on the high-throughput ultrasensitive and multiplexed detection of
oligonucleotides. However, the conventional microarray-based techniques that employ fluorescent dyes
are hampered by several limitations; they require target amplification, fluorophore labeling, and compli-
cated instrumentation, while the fluorophore-labeled species themselves exhibit slow binding kinetics,
photo-bleaching effects, and overlapping spectral profiles. Among the emerging nanomaterials that are
being used to solve these problems, oligonucleotide–gold nanoparticle conjugates (Oligo-AuNPs) have
recently been highlighted due to their unique chemical and physical properties. In this chapter, a detec-
tion scheme for oligonucleotides that utilize Oligo-AuNPs is evaluated with multiple oligonucleotide
targets. This scheme takes advantage of the sharp melting transitions, intense optical properties, catalytic
properties, enhanced binding properties, and the programmable assembly/disassembly of Oligo-AuNPs.
Key words: Oligonucleotide, Multiplexing, Gold nanoparticle, Detection, Sensing, Biomolecule,

DNA, RNA
1. Introduction
The dramatic development of genomics and proteomics in the

last several decades has attracted biologists to investigate how
the two fields merge together at the molecular level. For example,
the desire to understand how proteins are expressed and what
genes cause the expression has strongly motivated researchers to
investigate proteins or genes in vivo or in vitro under different
experimental conditions. To accomplish this task, they have taken
advantage of current molecular biology techniques such as mass
spectrometry (1), tap-tags (2), ELISAs (3), microarrays (4), and
blottings associated with gel electrophoresis. However, the need
to develop alternative assays that are rapid, sensitive, and selective
17
18 Lee
is growing significantly since the progress of the current research is

hampered by the low-throughput, high cost, slowness, and, in cer-
tain cases, limited reproducibility of the conventional methods.
Emerging nanotechnologies are considered to be promising
to solve these problems. Due to their unique and characteristic
properties that have not been observed in macroscopic worlds
previously and that excel the conventional materials in efficiency,
nanomaterials have been intensively investigated and utilized in a
variety of applications in diagnostics and therapeutics (5–7).
Among the large number of nanomaterials that have been synthe-
sized, oligonucleotide–gold nanoparticle conjugates (Oligo-
AuNPs) have been used as probes in numerous detection schemes
for biomolecule targets (8). The unique chemical and physical
properties of Oligo-AuNPs that stem from their “inorganic nano-
particle core and densely packed oligonucleotide shell” structure
offer high sensitivity and selectivity in such applications. To begin
with, a brief introduction of oligonucleotides, AuNPs, and Oligo-
AuNPs is given as follows.
1.1. Oligonucleotide An oligonucleotide is a short nucleic acid, or polymer of nucle-

otides, with typically fewer than 100 bases. Although it is possible
to cleave longer DNA or RNA strands to obtain them, oligonu-
cleotides often refer to “synthetic” ones synthesized by polymer-
izing individual nucleotide precursors utilizing phosphoramidite
chemistry. Because their chemical identity is the same as that of
longer DNA or RNA except in terms of length, oligonucleotides
have most of the same chemical and physical properties; they dis-
play a UV absorbance at 260 nm, form duplexes via base pairing,
have binding interactions with proteins, and form monomolecu-
lar structures, such as G-quadruplexes. In addition, a variety of
modified bases and nucleotide linkages (e.g., 2¢-O-methylated
RNA bases, phosphorothioate-modified phosphate backbones,
and locked nucleic acids (LNA)) have been developed to control
their properties, especially to enhance their stability.
One of the most important properties of oligonucleotides is
reversible duplex formation. Based upon the pairing of the four
bases (adenine (A)-thymine (T) and guanine (G)-cytosine (C))
via hydrogen bonds, an oligonucleotide sequence can reversibly
bind or hybridize to the complementary sequence. During the
hybridization process, its absorbance at 260 nm decreases due to
hypochromism. When heated, however, the duplexed oligonucle-
otides dehybridize into single strands with an increase in absor-
bance at 260 nm. If this dehybridization, or “melting” process, is
monitored as a function of temperature using UV–vis spectros-
copy (at 260 nm), a broad curve (full width at half maximum
(FWHM) ~10°C) is observed, whose midpoint is considered to
be the melting temperature (Tm) of the duplex system. The melting
temperature, which is the representative value of the stability of
Multiplexed Detection of Oligonucleotides with Biobarcoded Gold Nanoparticle Probes 19
the duplex pair, can be controlled by changing the salt concentration

of the solution, the length of the sequence, or the number of
mismatched bases.
Synthetic oligonucleotides, terminally modified with func-
tional chemical moieties such as thiols, amines, carboxyls, fluoro-
phores, quenchers, or biotin groups, are getting more and more
widely utilized in nanotechnology applications, including microar-
rays (9), DNA origami (10), nanomaterials assembly (7, 8, 11),
and so on. In this chapter, we cover how oligonucleotides that are
conjugated with nanomaterials, specifically gold nanoparticles,
play a significant role as a “smart” material based upon the afore-
mentioned properties.
1.2. Gold Colloidal gold nanoparticles (AuNPs) have been investigated by

Nanoparticles: many researchers because of their unusual properties that include
Chemical and Physical relatively high stability, low toxicity, catalytic activity, surface plas-
Properties mon resonance (SPR), enhanced Raman signal, and facile chemi-
cal tailorability. Also, these structures can support multiple
functionalities on their surface. AuNPs with diameters between
2 and 250 nm can be prepared in aqueous media in relatively
monodisperse forms using a variety of synthetic methods (12, 13).
The unique and intense colors associated with these AuNPs are
due to their SPR. This SPR is observed in the visible region of the
spectrum and is due to the collective oscillations of the free elec-
trons in the conduction band interacting with the electromagnetic
field of the incoming light. In solution, monodisperse AuNPs
(~15 nm in diameter) exhibit a red color indicative of a surface
plasmon absorption centered at 520 nm. Since AuNPs, especially
those stabilized with citrate, are charged particles, they are exceed-
ingly sensitive to changes in solution dielectric. For example, the
addition of sodium chloride (NaCl) shields the surface charge of
these AuNPs and leads to a concomitant decrease in interparticle
distance and eventual particle aggregation. A solution containing
coalesced AuNPs appears purple or blue in color, corresponding
to a characteristic dampening and red-shifting of the SPR of these
particles, to around 600 nm. This distance-dependent optical
property forms the basis of certain colorimetric detection schemes
that employ nanoparticle–biomolecule conjugates as probes.
1.3. General Properties An oligonucleotide–gold nanoparticle conjugate (Oligo-AuNP) is

of Oligo-AuNPs a two-component system comprised of thiolated-oligonucleotide
strands and their nanoparticle scaffold. The properties of these
inorganic nanoparticle core/organic biomolecular ligand hybrids
stem from not only each component of the hybrids but also syn-
ergistically from the combination of those two materials. The
deep red color exhibited by Oligo-AuNPs in aqueous media is
from the SPR of the gold nanoparticles, which shows a narrow
absorption in the visible range around 520 nm. However, the
20 Lee
chemical recognition abilities of these structures are derived from

the oligonucleotide shell. When two complementary Oligo-
AuNPs are combined, they form three-dimensional networks
through DNA duplex interconnects and an effective shift of the
SPR (50 nm or more) takes place with a concomitant red-to-
purple color change (14). Due to the reversible hybridization
properties of oligonucleotides, however, the assembled Oligo-
AuNPs disassemble at an elevated temperature (or reduced salt
concentration) returning back to red. Importantly, melting analy-
ses of the hybridized particle aggregates show sharper melting
transitions (FWHM ~2°C) than free DNA duplexes (15). The
sharpness of this melting transition is explained by a cooperative
mechanism that originates from the presence of multiple DNA
interconnects between the Oligo-AuNPs and the melting cascade
that results as counter ions are rapidly released from the aggregate
(15, 16). Further, Oligo-AuNPs exhibit particle size-dependent
melting properties and enhanced binding properties, which are
the representative synergetic results of cooperative interactions
between the AuNP and surrounding oligonucleotides (17, 18).
In addition, Oligo-AuNPs are chemically, physically, and bio-
logically stable, and nontoxic. While the inertness of AuNPs is
well known, the biological in vivo or in vitro stability of Oligo-
AuNPs had remained a question until recently. Unlike free
oligonucleotide strands, those on AuNPs are resistant to enzy-
matic degradation by DNase (19). This high biological stability is
due to the densely packed oligonucleotides on the AuNP surface,
which cause steric inhibition of the enzymatic cleavage reaction.
This increased resistance to nuclease degradation plays a signifi-
cant role in the therapeutic applications of these structures.
1.4. Biobarcoded Biobarcode assays are characterized by their ultrahigh sensitivity

Oligonucleotide–Gold and multiplexing capability for the detection of oligonucleotides
Nanoparticle or proteins. In case of oligonucleotide targets, Oligo-AuNP
Conjugate Probes probes are hybridized to oligonucleotide-functionalized magnetic
microparticle (MMP) probes using the target sequence as a linker;
these complexes are then separated magnetically for subsequent
release of the oligonucleotides from the Oligo-AuNP probes (see
Fig. 1.). These released oligonucleotides or “biobarcodes” are
quantitatively analyzed by the scanometric assay. The high sensi-
tivity of the biobarcode assays is believed to originate from four
reasons: (1) Because the biobarcode assay can be considered as a
pseudo-homogeneous assay, the binding equilibrium of the tar-
get and the particle probes can be increased by increasing the
probe concentration. (2) The dense loading of oligonucleotides
(biobarcodes) on the AuNPs was found to increase the equilib-
rium binding constant of the Oligo-AuNP up to two orders
of magnitude higher than that of unmodified free DNA (18).
(3) Due to the dense loading of oligonucleotides (biobarcodes)
on AuNPs, one binding event of the target sequence to the MMPs
Fig. 1. Multiplexed biobarcode assay for oligonucleotide target detection. Reproduced from ref. 23 with permission from
John Wiley & Sons, Inc.
and the AuNPs generates hundreds of biobarcode strands, lead-

ing to literally “hundreds of times” the target amplification.
(4) Finally, these biobarcodes, as surrogate targets, are further
amplified by silver enhancement during the scanometric assay.
Except the second reason (2), biobarcode assays for protein targets
also have the same reasons for the ultrahigh sensitivity, where anti-
bodies on the particle probes play a role in binding the targets.
This chapter will discuss the general protocol of the biobar-
code assay for oligonucleotide targets in a multiplexed form.
In brief, it covers the synthesis of gold nanoparticles and their
conjugates with oligonucleotides, the preparation of MMP
probes, the main biobarcode assay procedure, and final scanomet-
ric assay. As mentioned above, the fundamental principles of the
biobarcode assays for the detection of protein targets is very simi-
lar to what is described in this chapter for the oligonucleotide
targets, except for the probe preparation that includes the conju-
gation of the particles with antibodies.
2. Materials
2.1. Gold Nanoparticle 1. Gold (III) chloride trihydrate.

Synthesis 2. Ultrapure water (>18.2 MW cm).
22 Lee
3. Trisodium citrate.
4. Magnetic stirrer.
5. Stir bar (egg-shaped).
6. Round bottom flask.
7. Reflux condenser.
8. Heating mantle.
9. Temperature controller.
10. Concentrated hydrochloric acid (HCl, 37%).
11. Concentrated nitric acid (HNO, 70%).
12. Nylon filter (pore diameter = 0.45 mm).
2.2. Oligo-AuNP 1. Thiol-oligonucleotide of a given sequence (HPLC-purified)

Conjugate Synthesis (see Note 1).
2. NAP-5 Sephadex column.
3. Dithiothreitol (DTT).
4. UV–vis spectrometer.
5. Sodium dodecyl sulfate (SDS) solution: 1 wt %
6. Phosphate buffer: 100 mM sodium phosphate
(Na2HPO4 + NaH2PO4), pH = 7.4.
7. Sodium chloride solution: 2.0 M NaCl.
2.3. MMP Preparation 1. Amino-functionalized MMPs (2.8 mm in diameter; Invitrogen,

Carlsbad, CA).
2. Succinimidyl-4-(p-maleimidophenyl) butyrate (SMPB).
3. Coupling buffer: 0.2 M NaCl, 100 mM sodium phosphate,
pH = 7.0.
4. Passivation buffer: 0.15 M NaCl, 150 mM sodium phosphate,
pH = 8.0.
5. Separation magnets (sample volume ≥1 mL).
6. Sulfosuccinimidyl acetate (Sulfo-NHS-acetate).
7. Ammonium sulfate ((NH4)2SO4).
8. Oligonucleotides (3¢-thiol MMP-sequences).
9. Anhydrous dimethyl sulfoxide (DMSO).
2.4. Multiplexed 1. AuNPs (30 nm in diameter, Ted Pella, Inc., Redding, CA).
Biobarcode Assay 2. Assay buffer: 0.2 M NaCl, 0.1% Tween 20, and 10 mM
for DNA Targets sodium phosphate, pH = 7.2.
3. Separation magnets (sample volume ≥1 mL).
4. Scanometric buffer: 0.5 M NaCl, 0.01% Tween 20, and
10 mM sodium phosphate, pH = 7.2.
5. Verigene Reader System (Nanosphere, Northbrook, IL).

6. A 96-well-plate magnet.
7. Oligonucleotides (target sequences, 5¢-thiol AuNP-
sequences).
2.5. Scanometric 1. N-Hydroxysuccinimide (NHS) ester-activated Codelink

Detection of DNA slides (SurModics Inc., Eden Prairie, MN).
Targets 2. Printing buffer: 150 mM sodium phosphate, 0.01% SDS,
pH = 8.5.
3. Capture oligonucleotide (amine-modified) and target oligo-
nucleotide (unmodified).
4. Passivating solution: 0.2% SDS.
5. Spin dryer.
6. Two-well manual hybridization chambers (Nanosphere,
Northbrook, IL).
7. Temperature-controllable incubator.
8. Scanometric buffer: 0.5 M NaCl, 0.01% Tween 20, 10 mM
sodium phosphate, pH = 7.2.
9. Verigene Reader System (Nanosphere, Northbrook, IL).
10. Silver enhancing solutions (Nanosphere, Northbrook, IL).
11. Formamide.
12. Microarrayer.
13. GenePix Pro 6 software (Molecular Devices).
3. Methods
3.1. Gold Nanoparticle 1. Clean all glassware with aqua regia (3:1 HCl:HNO3, see Note 2).
Synthesis Rinse the glassware with 18.2 MW cm ultrapure water and dry in
an oven prior to use.
2. Bring an aqueous solution of HAuCl4 (1 mM, 500 mL) to
reflux while stirring.
3. Rapidly add 50 mL 38.8 mM trisodium citrate solution to the
refluxing solution.
4. Observe that the solution quickly changes color from pale
yellow to deep red.
5. After waiting 15 min, allow the mixture to cool to room tem-
perature and subsequently filter it through a 0.45 mm nylon
filter.
6. Characterize the colloid using UV–vis spectroscopy (see Note 3).
24 Lee
3.2. Oligo-AuNP 1. Dissolve the lyophilized thiol-modified oligonucleotides

Conjugate Synthesis (~40 nmol) in 0.1 M dithiothreitol (DTT) solution in phos-
phate buffer (0.17 M sodium phosphate, 100 mL) for 30 min
(see Note 4).
2. Purify the deprotected oligonucleotides from excess DTT
using a NAP-5 column. Apply less than 300 mL of the oligo-
nucleotide–DTT mixture to a NAP-5 column, and elute
about 1 mL, collecting about 4–5 drops at a time in separate
microtubes.
3. Measure the amount of DNA from each aliquot by UV–vis
spectroscopy and add it to AuNP solution (see Note 5).
4. Bring the solution to 0.3 M NaCl, 10 mM sodium phos-
phate, and 0.01% SDS, pH = 7.4 (see Note 6).
5. Typically, the system is allowed to equilibrate at least for a
couple of hours.
6. Remove the excess DNA by repeated centrifugation (16,000 × g,
25 min for 13 nm AuNPs; 12,000 × g, 10 min for 30 nm
AuNPs) and wash with a buffer solution (the concentration
of NaCl that you will use in the subsequent assay, 10 mM
phosphate, 0.005% Tween 20, pH = 7.4) (see Note 7).
7. Store the Oligo-AuNP probes at 4°C prior to use (see Note 8).
3.3. MMP Probe 1. Wash the MMPs (30 mg/mL, 1 mL) twice with anhydrous
Preparation dimethyl sulfoxide (DMSO, 1 mL).
2. Prepare a solution of SMPB (50 mg) in DMSO (15 mL) prior
to the reaction (see Note 9).
3. Add the SMPB/DMSO solution to the magnetic beads.
4. Allow the reaction between the primary amino group of
MMPs and the N-hydroxysuccinimide (NHS) ester of SMPB
to proceed for 4 h with gentle shaking at room temperature.
5. Separate the beads magnetically and wash them three times
with DMSO (10 mL) and two times with coupling buffer
(10 mL).
6. Reduce the disulfide bonds in all 3¢-thiol MMP-sequences
using DTT prior to mixing with the SMPB-activated MMPs
(see Note 10).
7. Prepare solutions (5 mM) of each of the freshly cleaved oligo-
nucleotides in coupling buffer.
8. Add 300 mL of each oligonucleotide to each of the washed
SMPB-activated magnetic beads.
9. Allow the reaction between the maleimide group and the
thiol-oligonucleotide to proceed at 20°C for 1 h under con-
stant vortex.
10. Place the oligo-functionalized MMPs on a magnet. Remove

the supernatant, and wash the beads twice with coupling buf-
fer and then twice with passivation buffer.
11. Use the supernatant to determine the coupling efficiency.
Measure the absorbance at 260 nm and compare it with that
before the oligonucleotide-functionalization of MMPs (see
Note 11).
12. Passivate the surface of the MMP probes by adding a freshly
prepared solution of sulfo-NHS-acetate (100 mg) in passiva-
tion buffer (40 mL) (see Note 12). Allow the passivation pro-
cess to proceed for 30 min at room temperature with mild
shaking.
13. Wash the MMP probes twice with passivation buffer, twice
with assay buffer, and store them at 4°C in assay buffer at a
concentration of 30 mg/mL.
3.4. Multiplexed 1. Prepare Oligo-AuNPs by conjugating oligonucleotides with

Biobarcode Assay the 30 nm-AuNPs (see Note 13) and finally redisperse them
for DNA Targets in assay buffer ([Oligo-AuNP] ~ 5 nM).
2. Prepare MMP multiplexing solution by combining all four
MMP probes in assay buffer (final total MMP probe concen-
tration is 1.56 mg/mL).
3. Begin the assay by mixing assay buffer (150 mL), MMP mul-
tiplexing solution (40 mL), and the appropriately mixed tar-
get solution (10 mL).
4. Heat the mixture at 45°C for 30 min and 25°C for 3 h under
constant vortex to allow hybridization between the MMP
probes and the target oligonucleotides (see Note 14).
5. Plate the reaction tube on a 96-well-plate magnet and wash
the MMP–target complexes twice with assay buffer
(200 mL).
6. Prepare Oligo-AuNP probe solution composed of all four NP
probes (NP multiplexing solution) by diluting equal volumes
of each NP probe (5 nM) in assay buffer containing forma-
mide (15%) to a final total NP concentration of 200 pM (see
Note 15).
7. Add the Oligo-AuNP multiplexing solution (50 mL) to the
MMP–target complexes and allow hybridization to proceed
at 25°C for 75 min under vortex.
8. Wash the mixture seven times with assay buffer (200 mL)
using a 96-well-plate magnet to remove nonspecifically bound
Oligo-AuNPs and any free barcode oligonucleotides.
9. Release the barcode oligonucleotides from the Oligo-AuNP
probes by the addition of DTT (75 mL, 0.5 M) in scanomet-
ric buffer.
26 Lee
10. Collect the released barcode oligonucleotides from the mixture

by removing MMPs using a magnet and analyze them quantita-
tively by the scanometric DNA detection method (see Note 16).
3.5. Scanometric 1. Dissolve amine-terminated capture DNA sequence (5¢ H2N-

Detection of DNA DNA sequence 3¢) in printing buffer. The final concentration
Targets of the capture DNA sequence is 100 mM (see Note 17).
2. Print the DNA spots on a Codelink slide using a microar-
rayer. Try to keep the humidity in the arrayer as low as rea-
sonably possible (see Note 18).
3. Hydrolyze the unreacted NHS groups of the slides or “passi-
vate” the slides with passivating solution at 50°C for 20 min.
4. Spin-dry the slides.
5. Inject target oligonucleotides (20 mL) into hybridization
chambers attached to a microarray slide prepared as above.
6. Hybridize the oligonucleotides to the capture strands on the
slide in an incubator at 60°C for 15 min, at 37°C for 30 min,
and at 25°C for 15 min with mild shaking (see Note 19).
7. Disassemble the chambers and rinse the slides copiously with
scanometric buffer, spin them dry, and reassemble them with
an unused hybridization chamber.
8. Introduce the Oligo-AuNP chip-probe solution (20 mL,
[Oligo-AuNP] = 500 pM) into the hybridization chambers
and allow the probes to hybridize at 37°C for 45 min (see
Note 20).
9. Disassemble, wash three times using scanometric buffer, and
spin-dry the hybridization chambers.
10. Perform the silver enhancement of the AuNPs by applying
the silver enhancing solution (2 mL; Nanosphere, Northbrook,
IL) on the dried chips for 3 min. Terminate the reaction by
washing the slides with ultrapure water (18.2 MW cm) (see
Note 21).
11. Spin-dry the slides and image them with the Verigene Reader
System (Nanosphere, Northbrook, IL), which records the
scattered light from the developed spots. Alternatively use a
flatbed scanner to read out the scattering signal. Examples of
the data obtained during readout are shown in Fig. 2.
4. Notes
1. Four exemplary sequences (Au1–Au4) for the AuNP probes

used in biobarcode assays are given as an example.
Au1: 5¢ HS-TACGAGTTGAGAATC-TACCACATCATC
CAT 3¢
T4
Target is
T3
present
T2
T1
a b c d e
T4
T3
Target is
T2
not present
T1
f g h i j
Fig. 2. Scanometric detection of the biobarcode-oligonucleotides released from the

30-nm Oligo-AuNP probes for ten different samples. (a) All targets are present; (b) T1;
(c) T2; (d) T3; (e) T4; (f) T1 and T2; (g) T1 and T3; (h) T1 and T4; (i) T1, T3, and T4; and
(j) T1, T2, and T4. The gray-scale images from the Verigene Reader System were con-
verted into colored ones using GenePix Pro 6 software (Molecular Devices). The diameter
of the spots on the chip is 200 mm. Reproduced from ref. 23 with permission from John
Wiley & Sons, Inc.
Au2: 5¢ HS-TACGAGTTGAGAATC-CTGATTACTATT
GCA 3¢
Au3: 5¢ HS-TACGAGTTGAGAATC-TTGTTGATACTG
TTC 3¢
Au4: 5¢ HS-TACGAGTTGAGAATC-TGCATCCAGGTC
ATG 3¢
2. Aqua regia is an extremely powerful oxidizing solution which
is the only acidic solution that can be used to dissolve gold.
Use extreme caution when working with aqua regia, because
it generates harmful chlorine (Cl2) and nitrogen oxide (NOx)
gases and can cause severe tissue damage.
3. The characteristic SPR band of monodispersed particles is
located around 520 nm. Particles with a more uniform size
distribution have narrower absorption bands. Note that
aggregated 13 nm gold nanoparticles, as discussed above, dis-
play a flattened and red-shifted absorption peak at approxi-
mately 600 nm.
4. Usually, the DNA sequences are 15- to 20-base pairs in
length, complementary to a DNA target of interest, and
modified with either 3¢ or 5¢ terminal sulfur-containing groups
in the form of monothiols, cyclic disulfides (20), or trithiols
(21). The terminal monothiol group of each oligonucleotide
28 Lee
is protected in the oxidized form (disulfide), which needs to

be reduced by DTT to the monothiol form before reaction
with gold nanoparticles.
5. The final oligonucleotide and AuNP concentrations are
~4 mM and ~6 nM, respectively.
6. The final concentrations of sodium phosphate (10 mM),
sodium chloride (0.3 M), and SDS (0.01%) are achieved by
adding the appropriate amounts of the concentrated solu-
tions described in Subheading 2.2 to the oligonucleotide–
gold nanoparticle mixture. The number of oligonucleotides
per gold nanoparticle (15 nm in diameter) can be increased
(~50–100 strands per particle) by increasing the NaCl con-
centration (from 0.1 to 1 M NaCl, respectively). Sodium ions
screen the repulsions between the negatively charged oligo-
nucleotide strands. The oligonucleotide loading also increases
when the sequence has a polyethylene glycol (PEG) unit near
the thiol group, or the mixture of oligonucleotides and gold
nanoparticles is sonicated (22).
7. The removal of the excess DNA based upon centrifugation is
repeated typically four times. More centrifugation steps can
be performed when high purity is required, depending on the
experimental purposes.
8. Oligo-AuNPs are stable and biochemically active for about
2 months when stored at 4°C.
9. Caution should be taken to prevent exposure of the solution
to light.
10. The reduction of the disulfide bonds by DTT and the purifi-
cation of the thiol-oligonucleotides are performed follow-
ing the procedure described in Subheading 3.2. Four
thiolated sequences (M1–M4) are given as an example; the
sequence of these strands could be modified according to
the experimental purposes [23). The A10 portion near the
thiol group is a spacer and does not participate in
hybridization.
M1: 5¢ ATAACTGAAAGCCAA-A10-SH 3¢
M2: 5¢ TCTTCCGTTACAACT-A10-SH 3¢
M3: 5¢ TCCAACATTTACTCC-A10-SH 3¢
M4: 5¢ TTATTCCAAATATCTTCT-A10-SH 3¢
11. Typically, there are on average 3 × 105 oligonucleotides per
MMP.
12. Sulfo-NHS acetate acylates primary amino groups in basic media.
13. The conjugation of the oligonucleotide with the AuNP probe
is performed following Subheading 3.2, except that the final
NaCl concentration is 0.2 M. Four thiolated sequences

(Au1–Au4) are given as an example; these sequences could be
modified according to the experimental purposes (23). The
“TACGAGTTGAGAATC” portion is common to all the
sequences and is used for hybridization with the Oligo-AuNP
chip-probes in the scanometric assay.
Au1: 5¢ HS-TACGAGTTGAGAATC-TACCACATCATCCAT 3¢
Au2: 5¢ HS-TACGAGTTGAGAATC-CTGATTACTATTGCA 3¢
Au3: 5¢ HS-TACGAGTTGAGAATC-TTGTTGATACTGTTC 3¢
Au4: 5¢ HS-TACGAGTTGAGAATC-TGCATCCAGGTCATG 3¢
14. Four unmodified target sequences (T1–T4) are given as an
example; the sequence of these strands could be modified
according to the experimental purposes. The initial reaction
temperature (45°C) was determined to completely dehy-
bridize the target sequences from the oligonucleotides on
MMPs, leading to better duplex formation at 25°C in the
next step.
T1: 5¢ TTGGCTTTCAGTTAT-ATGGATGATGTGGTA 3¢
T2: 5¢ AGTTGTAACGGAAGA-TGCAATAGTAATCAG 3¢
T3: 5¢ GGAGTAAATGTTGGA-GAACAGTATCAACAA 3¢
T4: 5¢ AGAAGATATTTGGAATAA-CATGACCTGGAT
GCA 3¢
15. Formamide increases the stringency at which hybridization
occurs since it competes with the complementary DNA strand
for hydrogen bonding interactions. The addition of 15% for-
mamide to the reaction mixture resulted in optimal selectivity
of the system.
16. The scanometric detection of DNA is performed following
the procedure in Subheading 3.5.
17. Only use HPLC-purified oligonucleotides. Amine contami-
nants can reduce coupling.
18. Codelink Slides must be stored desiccated, because the NHS
leaving group on the Codelink slide also reacts with H2O,
resulting in a decrease in the coupling efficiency of the amine-
modified oligonucleotides. Before and after each spotting
run, always wash the pins 3–4 times to be sure that there is no
salt build-up on them.
19. The incubation temperatures were controlled to completely
dehybridize the biobarcode sequences from the capture
sequences on the chip at 60°C and to enhance the hybridiza-
tion by gradually decreasing the temperatures (37 → 25°C).
This process is called “annealing.”
30 Lee
20. The hybridization temperature is determined empirically not

to dehybridize the probes at too high temperature, nor to
hybridize the probes nonspecifically at too low temperature.
21. The staining time needs to be optimized depending on the
number of hybridized AuNP probes. Longer staining time
than enough saturates the signal intensity, while too short
staining time cannot enhance the low signal intensity enough
to be visualized.
References
1. Vlahou, A. and Fountoulakis, A. (2005) 11. Alivisatos, A. P., Johnsson, K. P., Peng, X.,
Proteomic approaches in the search for dis- Wilson, T. E., Loweth, C. J., Bruchez, M. P.,
ease biomarkers. J. Chromatogr. B 814, et al. (1996) Organization of “nanocrystal
11–19. molecules” using DNA. Nature 382,
2. Rigaut, G., Shevchenko, A., Rutz, B., Wilm, 609–611.
M., Mann, M., and Seraphin, B. (1999) A 12. Grabar, K. C., Freeman, R. G., Hommer, M.
generic protein purification method for pro- B., and Natan, M. J. (1995) Preparation and
tein complex characterization and proteome characterization of Au colloid monolayers.
exploration. Nat. Biotechnol. 17, Anal. Chem. 67, 735–743.
1030–1032. 13. Frens, G. (1973) Controlled nucleation for
3. Seubert, P., Vigopelfrey, C., Esch, F., Lee, M., regulation of particle-size in monodisperse
Dovey, H., Davis, D., et al. (1992) Isolation gold suspensions. Nat. Phys. Sci. 241, 20–22.
and quantification of soluble Alzheimers beta- 14. Daniel, M. C. and Astruc, D. (2004) Gold
peptide from biological-fluids. Nature 359, nanoparticles: assembly, supramolecular
325–327. chemistry, quantum-size-related properties,
4. DeRisi, J. L., Iyer, V. R., and Brown, P. O. and applications toward biology, catalysis,
(1997) Exploring the metabolic and genetic and nanotechnology. Chem. Rev. 104,
control of gene expression on a genomic scale. 293–346.
Science 278, 680–686. 15. Jin, R., Wu, G., Li, Z., Mirkin, C. A., and
5. Rosi, N. L. and Mirkin, C. A. (2005) Schatz, G. C. (2003) What controls the melt-
Nanostructures in biodiagnostics. Chem. Rev. ing properties of DNA-linked gold nanopar-
105, 1547–1562. ticle assemblies? J. Am. Chem. Soc. 125,
6. Wang, H., Yang, R., Yang, L., and Tan, W. 1643–1654.
(2009) Nucleic acid conjugated nanomateri- 16. Storhoff, J. J., Lazarides, A. A., Mucic, R. C.,
als for enhanced molecular recognition. ACS Mirkin, C. A., Letsinger, R. L., and Schatz, G.
Nano 3, 2451–2460. C. (2000) What controls the optical proper-
7. Niemeyer, C. M. and Simon, U. (2005) ties of DNA-linked gold nanoparticle assem-
DNA-based assembly of metal nanoparticles. blies? J. Am. Chem. Soc. 122, 4640–4650.
Eur. J. Inorg. Chem. 18, 3641–3655. 17. Lee, J.-S., Stoeva, S. I., and Mirkin, C. A.
8. Mirkin, C. A., Letsinger, R. L., Mucic, R. C., (2006) DNA-induced size-selective separa-
and Storhoff, J. J. (1996) A DNA-based tion of mixtures of gold nanoaparticles.
method for rationally assembling nanoparticles J. Am. Chem. Soc. 128, 8899–8903.
into macroscopic materials. Nature 382, 18. Lytton-Jean, A. K. R. and Mirkin, C. A.
607–609. (2005) A thermodynamic investigation into
9. Taton, T. A., Mirkin, C. A., and Letsinger, R. the binding properties of DNA functionalized
L. (2000) Scanometric DNA array detection gold nanoparticle probes and molecular fluo-
with nanoparticle probes. Science 289, rophore probes. J. Am. Chem. Soc. 127,
1757–1760. 12754–12755.
10. Anderson, E. S., Dong, M., Morten, M. N., 19. Rosi, N. L., Giljohann, D. A., Thaxton, C. S.,
Kasper, J., Subramani, R., Mamdouh, W., Lytton-Jean, A. K. R., Han, M. S., and Mirkin,
et al. (2009) Self-assembly of a nanoscale C. A. (2006) Oligonucleotide-modified gold
DNA box with a controllable lid. Nature 459, nanoparticles for intracellular gene regulation.
73–76. Science 312, 1027–1030.
20. Letsinger, R. L., Elghanian, R., Viswanadham, 22. Hurst, S. J., Lytton-Jean, A. K. R., and Mirkin,
G., and Mirkin, C. A. (2000) Use of a ste- C. A. (2006) Maximizing DNA loading on a
roid cyclic disulfide anchor in constructing range of gold nanoparticle sizes. Anal. Chem.
gold nanoparticle-oligonucleotide conjugates. 78, 8313–8318.
Bioconjug. Chem. 11, 289–291. 23. Stoeva, S. I., Lee, J.-S., Thaxton, C. S., and
21. Li, Z., Jin, R. C., Mirkin, C. A., and Letsinger, Mirkin, C. A. (2006) Multiplexed DNA
R. L. (2002) Multiple thiol-anchor capped detection with bio-barcoded nanoparticle
DNA-gold nanoparticle conjugates. Nucleic probes. Angew. Chem. Int. Ed. 45,
Acids Res. 30, 1558–1562. 3303–3306.
Chapter 3
Molecular Detection of Biomarkers and Cells Using

Magnetic Nanoparticles and Diagnostic Magnetic
Resonance
Jered B. Haun, Tae-Jong Yoon, Hakho Lee, and Ralph Weissleder
Abstract
The rapid and sensitive detection of molecular targets such as proteins, cells, and pathogens in biological
specimens is a major focus of ongoing medical research, as it could promote early disease diagnoses and
the development of tailored therapeutic strategies. Magnetic nanoparticles (MNP) are attractive candi-
dates for molecular biosensing applications because most biological samples exhibit negligible magnetic
susceptibility, and thus the background against which measurements are made is extremely low. Numerous
magnetic detection methods exist, but sensing based on magnetic resonance effects has successfully been
developed into a general detection platform termed diagnostic magnetic resonance (DMR). DMR tech-
nology encompasses numerous assay configurations and sensing principles, and to date magnetic nano-
particle biosensors have been designed to detect a wide range of targets including DNA/mRNA, proteins,
enzymes, drugs, pathogens, and tumor cells with exquisite sensitivity. The core principle behind DMR is
the use of MNP as proximity sensors that modulate the transverse relaxation time of neighboring water
molecules. This signal can be quantified using MR imagers or NMR relaxometers, including miniaturized
NMR detector chips that are capable of performing highly sensitive measurements on microliter sample
volumes and in a multiplexed format. The speed, sensitivity, and simplicity of the DMR principle, cou-
pled with further advances in NMR biosensor technology should provide a high-throughput, low-cost,
and portable platform for large-scale parallel sensing in clinical and point-of-care settings.
Key words: Magnetic nanoparticles, Molecular biosensing, Diagnostic magnetic resonance,

Magnetic relaxation switches
1. Introduction
Magnetic nanoparticles (MNP) offer unique advantages for

molecular detection of biological targets compared to other plat-
forms. For example, MNP are inexpensive to produce, physically
and chemically stable, biocompatible, and environmentally safe.
33
34 Haun et al.
In addition, biological samples exhibit virtually no magnetic

background, and thus measurements can be performed in turbid
or otherwise visually obscured samples. To date, numerous meth-
ods have been developed to sense biomolecules using magnetic
labels (1), but one that has achieved considerable success in bio-
medicine is based on magnetic resonance. In this case, the MNP
serve as proximity sensors that accelerate the relaxation rate of
neighboring water molecules following exposure to a magnetic
field. This phenomenon has been exploited for magnetic reso-
nance imaging to obtain detailed anatomical information.
When magnetic resonance is used for molecular detection of
biomolecules and cells outside of the body, it is referred to as
diagnostic magnetic resonance (DMR) (2, 3). DMR assays employ
affinity molecule-conjugated MNP to bind molecular targets and
effect a change in proton relaxation rate by one of two methods
(see Fig. 1). The first makes use of the phenomenon of magnetic
relaxation switching (MRSw), in which molecular targets are used
to self-assemble MNP into clusters and thereby cause a corre-
sponding change in the bulk relaxation rate (4, 5). The second
mode involves tagging large structures such as whole cells (2, 3, 6).
In both cases, the binding interactions are performed homoge-
neously in solution and make use of the built-in amplification that
magnetic resonance offers through its effect on billions of neigh-
boring water molecules. These factors make DMR faster than
techniques that require solid-phase immobilization, diffusion of
nanoparticles to sensing elements, or discrete amplification steps.
The magnetic resonance signal can be quantified by NMR or
MRI as a decrease in longitudinal (spin–lattice, T1) or transverse
(spin–spin, T2) relaxation times. Typically T2 is used because the
transverse relaxivity (r2) is significantly greater than the longitudi-
nal relaxivity (r1) for most MNP.
To date, DMR has successfully been used to detect a wide
variety of biomolecular targets with high sensitivity and specific-
ity, including DNA, mRNA, proteins, enzyme activity, metabo-
lites, drugs, pathogens, and tumor cells (see Table 1). Notably,
detection thresholds in the femtomolar biomarker concentration
and near single-cell range have been documented in complex
media such as cell lysates, sputum, and whole blood (3, 4, 6). The
magnetic resonance signal can be read-out using MRI scanners or
benchtop NMR systems. Recently, however, a miniaturized, chip-
based NMR (mNMR) detector system was developed that houses
all components for DMR detection in a handheld, portable device
(2, 3, 6). The mNMR chip device is capable of performing mea-
surements on microliter sample volumes and contains multiple
detection coils to enable parallel sensing of numerous biomarkers.
The mNMR chip thus represents a critical advance in DMR technol-
ogy that could facilitate detection of disease markers in sample-
limited clinical specimen.
Molecular Detection of Biomarkers and Cells Using Magnetic Nanoparticles 35
Fig. 1. DMR sensing principles. (a) Magnetic relaxation switching (MRSw) involves the
assembly of MNP into clusters or disassembly of preformed clusters by the action of a
target biomolecule. Clustered MNP dephase the nuclear spins of neighboring water
molecules more efficiently than evenly dispersed MNP, shortening the bulk transverse
relaxation time (T2). (b) Tagging cells with MNP imparts a magnetic moment that is
proportional to the number of nanoparticles bound. Following washing procedures to
remove unbound MNP, the magnetic moment can be measured as a decrease in T2
relaxation time. (c) Representative NMR output depicting the shortening of T2 relaxation
time that accompanies MNP clustering (MRSw) or cellular tagging. Modified with per-
mission from ref. 2, copyright (2008) Nature Publishing Group.
2. Materials
2.1. Preparation 1. Superparamagnetic nanoparticles: Synthesized (7, 8) or

for Bioconjugation purchased commercially (Miltenyi Biotec, Auburn, CA; Ocean
of Affinity Molecules NanoTech, Springdale, AK) with primary amine functional
to MNP groups or conjugated with proteins such as avidin or protein A.
2. Affinity molecule: Lyophilized or dissolved in buffer. This
sample should be free of contaminants that would interfere
with coupling interactions (see Note 1).
36 Haun et al.
Table 1
DMR biosensors/applications to date
Class Target Mode Magnetic nanoparticle sensor References
DNA Telomeres MRSw (A) CLIO-(CCCTAA)3 (24)

RNA GFP MRSw (A) CLIO-ATTTGCCGGTGT (4)
and CLIO-TCAAGTCGCACA
Protein GFP MRSw (A) CLIO-antibody (anti-GFP) (4)
Avidin MRSw (A) CLIO-biotin (4)
b-HCG MRSw (A) CLIO-antibody (anti-b-HCG) (25)
Telomerase MRSw (A) CLIO-antibody (anti-telomerase) (26)
CA-125 MRSw (A) CLIO-antibody (anti-CA-125) (2)
VEGF MRSw (A) CLIO-antibody (anti-VEGF) (2)
a-Fetoprotein MRSw (A) CLIO-antibody (anti-a-fetoprotein) (2)
Enzyme Caspase 3 MRSw (D) CLIO-GDEVDG-biotin; CLIO-avidin (4)
activity
BamH1 MRSw (D) CLIO-TTA-CGC-CTAGG-ATC-CTC (27)
and CLIO-AAT-GCG-GGATCC-
TAC-GAG
DNA methylase MRSw (D) Methylated BamH1 sensor (27)
Renin MRSw (D) Biotin-IHPFHLVIHTK-biotin; (28)
CLIO-avidin
Trypsin MRSw (D) Biotin-(G)4RRRR(G)3K-biotin (28)
or biotin-GPARLAIK-biotin;
CLIO-avidin
MMP-2 MRSw (D) Biotin-GGPLGVRGK-biotin; (28)
CLIO-avidin
Telomerase MRSw (A) CLIO-AATCCCAATCCC and (26)
CLIO-AATCCCAATCCC
Peroxidases MRSw (D) CLIO-phenol or CLIO-tyrosine (29)
Small Drug MRSw (D) CLIO-antibody (anti-d-phenylalanine) (30)
molecule Folate MRSw (D) CLIO-antibody (anti-folate) (31)
Glucose MRSw (D) CLIO-Concavalin (31)
HA peptide MRSw (D) CLIO-antibody (anti-HA) (31)
Calcium MRSw (A) CLIO-Calmodulin and CLIO-M13 (32, 33)
or CLIO-chelator
Organism Herpes virus MRSw (A) CLIO-antibody (anti-HSV1) (34)
Adenovirus-5 MRSw (A) CLIO-antibody (anti-adenovirus-5) (34)
MAP MRSw (A) CLIO-antibody (anti-MAP) (35)
Staphylococcus MRSw (A) CLIO-Vancomycin (2)
aureus
MTB/BCG Tagging CLIO-antibody or CB-antibody (6)
(anti-BCG)
Human cell Tumor cell lines Tagging CLIO-antibody (anti-Her2/neu, (2)
EGFR, EpCAM)
Murine tumor Tagging Mn-MNP-antibody (anti-Her2/neu, (3)
biopsies EGFR, EpCAM)
MRSw (A): MRSw assay using assembly of MNP into clusters. MRSw (D): MRSw assay based on disassembly of
pre-clustered MNP
3. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM

KCl, 10 mM Na2HPO4, and 1.8 mM KH2PO4 in de-ionized
water, adjust pH to 7.4 with NaOH. PBS can also be pur-
chased, but be sure that divalent cations are not present (see
Note 2).
4. Ethylenediamine tetraacetic acid (EDTA) buffer: 10× Solution
with 100 mM EDTA in the PBS buffer listed in item 3,
Subheading 2.1. EDTA is necessary for thiol-based chemis-
tries to prevent formation of disulfide bonds between free
thiol groups.
5. Bicarbonate buffer: 0.1 M NaHCO3, adjust pH to 10 with
NaOH.
6. Dimethylformamide (DMF) [if desired, dimethylsulfoxide
(DMSO) can be used interchangeably].
7. Thiol chemistry bioconjugations will require one or more of
the following reagents:
(a) Sulfosuccinimidyl-4-[N-maleimidomethyl]cyclohexane-
1-carboxylate (sulfo-SMCC, available in convenient no-
weigh format from Thermo Fisher, Waltham, MA).
(b) N-Succinimidyl-S-acetylthioacetate (SATA, Thermo
Fisher).
(c) b-Mercaptoethylamine (MEA, Sigma Aldrich, St. Louis,
MO).
8. Deacetylation buffer: 0.5 M Hydroxylamine⋅HCl, 25 mM
EDTA in the PBS buffer listed in item 3, Subheading 2.1,
adjust pH to 7.4 with NaOH. This reagent is required for
point (b) of item 7, Subheading 2.1 only.
9. Click chemistry bioconjugations can be performed using
commercially available reagents including succinimidyl ester
or iodoacetamide conjugated azide and alkyne moieties
(Invitrogen, Carlsbad, CA).
10. Amicon Ultra centrifugal filtration device, 15 mL volume,
100,000 MWCO (Millipore, Billerica, MA).
11. Sephadex G-50 (GE Healthcare, Piscataway, NJ) or similar
gel filtration media.
12. PD-10 (GE Healthcare) or Zeba (5 mL, Thermo Fisher)
desalting columns.
13. Coomassie protein stain (Thermo Fisher).
2.2. Coupling, 1. Copper catalyst buffer: 10× Solution with 50 mM sodium
Processing, and ascorbate, 10 mM Cu(II)SO4, 50 mM of a polytriazole ligand
Characterization (see Note 3).
2. Sephadex G-100 Superfine (GE Healthcare) or similar gel
filtration media.
38 Haun et al.
3. Fe digestion buffer: 6 M HCl, 0.3% H2O2 in de-ionized

water.
4. Micro BCA™ Protein Assay Kit (Thermo Fisher).
5. 96-Well plate (Becton Dickinson, Franklin Lakes, NJ).
2.3. MRSw Assay 1. Analyte samples can be obtained from numerous sources,
for Soluble Analyte including unpurified biological samples such as cell culture
Sensing supernatants, cell lysates, milk, sputum, and whole blood. For
disassembly assays, the MNP are typically pre-clustered using
purified cross-linkers such as oligonucleotides or proteins
prior to addition of analyte.
2.4. Magnetic 1. PBS with bovine serum albumin (BSA) (PBS+): 1 g/L BSA
Tagging of Cells added to the PBS buffer listed in item 3, Subheading 2.1,
sterile filter.
2. Cell samples can be obtained from any source of interest,
such as in vitro cultures or in vivo specimen, but should be
suspended as single cells. This can be accomplished using
various enzymes (trypsin, collagenase, etc.) and/or EDTA
(see Note 4). See Note 5 for total cell requirements.
2.5. Detection 1. 5 or 10 mm NMR tubes (Thermo Fisher).

Using NMR 2. Benchtop NMR relaxometer (i.e., Minispec, Bruker Optics,
Billerica, MA). Alternatively, custom devices such as miniaturized
NMR (mNMR) detectors have been used (2, 3, 6). The indicated
options both operate at approximately 20 MHz and 0.5 T.
2.6. Detection 1. 384-Well plate (Becton Dickinson).

Using MRI 2. Clinical or experimental MRI scanner operating at 1–11 T
[i.e., ClinScan (7 T) or PharmScan (4.5 T), Bruker BioSpin,
Billerica, MA].
3. Software to analyze T2 images (i.e., OsiriX Viewer, available
at http://www.osirix-viewer.com).
3. Methods
The MNP used for magnetic resonance detection must be super-

paramagnetic, meaning that they become magnetized when
placed in an external magnetic field but lose their magnetic
moment when the field is removed (no magnetic remanence).
The bulk of DMR assays have been performed using iron oxide
nanoparticles, in particular cross-linked iron oxide (CLIO) (7, 9).
CLIO has a magnetic core composed of a monocrystalline iron
oxide nanoparticle (MION), and the MION core is caged in
cross-linked dextran. Such iron oxide nanoparticles can be

obtained in various forms from commercial sources. Aside from
standard iron oxide options, it has been demonstrated that dop-
ing iron oxide with metals such as manganese, cobalt, and nickel
can increase magnetization, and thus detection sensitivity (10).
Likewise, a core–shell nanoparticle composed of an elemental
iron core and protective iron oxide shell has also been shown to
possess greater magnetic susceptibility than iron oxide (11).
Manganese-doped iron oxide nanoparticles (Mn-MNP) and ele-
mental Fe core nanoparticles (cannonballs, CB) have been vali-
dated for DMR applications and significantly improve detection
sensitivity (3, 6). Micron-sized particles composed of many iron
oxide cores embedded in a polymer matrix have been used for
DMR assays as well (12, 13), but their extremely high magnetiza-
tion is offset by rapid settling out of solution, limiting utility.
Regardless of the type of superparamagnetic core used, a
polymer coating is required to render the core water soluble, pre-
vent aggregation, and provide chemical functional groups for
bioconjugation. Most often the functional groups are primary
amines, but carboxylic acids or thiols are also used. The protocols
listed below detail the attachment of affinity molecules to amine-
terminated MNP; treatments of MNP with other surface-based
functional groups can be found elsewhere (14). Subheading 3.1
or 3.2 can also be followed to first attach avidin, proteins A or G,
secondary antibodies, or fluorescent molecules (see Note 6) to
the MNP or alternatively MNP can be purchased with these moieties
already conjugated to their surface. In these cases, skip to
Subheading 3.3 for attachment of affinity molecules modified
with biotin or antibody Fc domains. The choice of affinity mole-
cule is dependent on the target and application of interest. To
date, strategies have included the attachment of oligonucleotides,
antibodies, peptides, small molecules, metal chelators, and natu-
ral biological binding partners, as listed in Table 1.
3.1. Bioconjugation 1. These instructions are intended for the attachment of affinity
Using Thiol Chemistry molecules that contain a free thiol group, such as peptides
with a terminal cysteine residue and proteins that are appro-
priately modified to bear a thiol group (mild reduction of
disulfide bonds or conversion of a primary amine to a thiol
group).
2. Dilute amino-MNP with PBS to approximately 0.5 mg Fe/
mL concentration and adjust the pH to 8.0 using the bicar-
bonate buffer. Lower Fe concentrations can be used but a
different purification process is required (see Note 7).
3. Add at least 0.1 mg sulfo-SMCC per mg Fe (approximately
30-fold molar excess for CLIO containing 30 amines/MNP,
see Note 8). The addition of more sulfo-SMCC will not
40 Haun et al.
egatively affect the reaction. Incubate the reaction mixture

n
while shaking at room temperature for 1–2 h (see Note 9).
4. Concentrate the MNP by adding to a centrifugal filtration
device and centrifuging for 10 min at 2,000 × g. The final vol-
ume is usually about 0.2 mL.
5. Remove excess sulfo-SMCC reagent using a desalting column
packed with Sephadex G-50. The total column volume should
be at least 50 times greater than the sample volume to ensure
effective separation. The resulting maleimide-MNP should
be used as soon as possible, but can be stored at 4°C for up to
24 h if necessary.
6. Prepare the affinity molecule by dissolving the sample in or
diluting it with the 10× EDTA buffer, and additional PBS if
necessary, to obtain a final sample volume of 1 mL containing
EDTA at 1× concentration. If the affinity molecule contains a
free thiol functional group, it can be used directly (skip to
Subheading 3.3).
7. For proteins that contain a disulfide bond that can be reduced
without affecting binding function (such as most antibodies),
add 1 mg of MEA (26 mM final concentration) and incubate
for 2 h at 37°C. Alternatively, the primary amine groups of
certain proteins can be converted to protected sulfhydryls
using SATA; these groups can then be deprotected to yield
free sulfhydryls. In this case, dissolve the SATA in DMF and
add it to the protein sample at two- to fivefold molar excess
(see Note 10). Incubate the mixture for 1–2 h at room tem-
perature (see Note 9), then add 0.1 mL deacetylation buffer,
and incubate for another 1.5 h.
8. Purify these protein affinity molecules by desalting with a
gravity (PD-10) or centrifugal column (Zeba). First, equili-
brate the column with PBS containing 1× EDTA buffer
before applying the sample. For PD-10 columns, collect
~1 mL fractions and check for protein by mixing 50 mL sam-
ple with 50 mL Coomassie reagent (a color change will be
observed if protein is present). Pool positive fractions. For
the Zeba columns, use the flow-through directly (skip to
Subheading 3.3).
3.2. Bioconjugation 1. These instructions are intended as a general method to attach

Using “Click” affinity molecules to amine-modified MNP using bioorthog-
Chemistry onal “click” chemistries (15). The most widely used “click”
reaction is the copper-catalyzed azide-alkyne cycloaddition
(16, 17), which has been used to conjugate small molecules
and antibodies to MNP (18, 19), along with many other
applications. Recently, a new catalyst-free cycloaddition
between a 1,2,4,5-tetrazine and trans-cyclooctene dienophile
(20) was successfully used in the same capacity (21). Regardless

of the reaction pair employed, they can be used interchange-
ably on the affinity molecule of choice and MNP, and
therefore will generically be referred to as Click Reactants 1
and 2. They can be attached using amine- or thiol-reactive
moieties, such as a succinimidyl ester or iodoacetamide/
maleimide, respectively.
2. Dilute amino-MNP with PBS containing 5% DMF to a con-
centration of 0.5 mg Fe/mL. Adjust the pH to 8.0 using the
bicarbonate buffer.
3. Dissolve amine-reactive Click Reactant 1 with DMF and add
a tenfold molar excess relative to the total amine content per
MNP. For example, 2 mg of CLIO with 30 amines/particle
would require a minimum 1.4 mmol amine-reactive Click
Reagent 1 (see Note 8). Incubate the reaction mixture while
shaking at room temperature for 2 h.
4. Concentrate and desalt as indicated in steps 4 and 5,
Subheading 3.1.
5. Prepare the affinity molecule by dissolving the sample in or
diluting it with PBS to 0.8 mL. Add 0.1 mL sodium bicarbon-
ate buffer. Dissolve Click Reactant 2 (amine- or thiol-reactive)
in DMF at 10 mg/mL and add to the reaction solution at
two- to fivefold molar excess (see Note 10). Include additional
DMF to achieve a total concentration of 10% if necessary.
Incubate at room temperature for 1–2 h (see Note 9).
6. Purify the affinity molecule as described in step 8,
Subheading 3.1, but omit EDTA from the wash solution
because it is only necessary if free thiol groups are present in
the sample. For small molecules or peptides, HPLC purifica-
tion may be required.
3.3. Coupling, 1. Combine the appropriate MNP (conjugated with maleimide,

Processing, and Click Reactant 1, avidin, protein A/G, or secondary anti-
Characterization body) with the appropriate affinity molecule (conjugated with
a free thiol group, Click Reactant 2, biotin, or antibody Fc
domain) samples. If necessary, adjust the reaction volume
with PBS such that the final MNP concentration is less than
0.5 mg Fe/mL to prevent cross-linking of nanoparticles (see
Note 10). The affinity molecule should be added in excess to
increase MNP valency, and thus binding activity, and prevent
cross-linking of MNP for cases in which the affinity molecule
contains multiple coupling domains. In general, a fivefold
molar excess is sufficient, but a tenfold excess is preferred. For
the copper-catalyzed azide-alkyne cycloaddition, the copper
catalyst buffer must also be added to the reaction mixture at
1× concentration.
2. Incubate for 4 h at room temperature or overnight at 4°C.
42 Haun et al.
3. For thiol conjugations, add ammonium chloride or cysteine

at 1 mM final concentration to cap unreacted maleimide
groups and prevent nonspecific binding to biological samples
(see Note 11). Incubate 30 min at room temperature.
4. Concentrate by adding to a centrifugal filtration device and
centrifuging for at least 10 min at 2,000 × g.
5. Remove excess affinity molecules using a size-exclusion col-
umn packed with Sephadex G-100. Again, the total column
volume should be at least 50 times greater than the sample
volume to ensure effective separation.
6. Quantify the MNP concentration based on Fe content by
digesting with hydrochloric acid to yield FeCl3 and determin-
ing the Fe absorbance at 410 nm. Dilute the MNP solution
with at least nine volumes of Fe digestion buffer and incubate
at 50–60°C for 1 h. Determine the absorbance at 410 nm and
convert to Fe concentration using an extinction coefficient of
1,370 M−1 cm−1 and the appropriate dilution factor.
Alternatively, a quick estimate of the MNP concentration can
be obtained by measuring the absorbance at 410 nm and cali-
brating with a known standard (i.e., the original MNP stock
solution).
7. For protein and some peptide affinity molecules, successful
conjugation can readily be determined using the Micro
BCA™ Protein Assay. Combine equal volumes (minimum
50 mL) of the MNP sample solution and the Micro BCA
working reagent in a 96-well plate. Incubate for 2 h at 37°C,
cool to room temperature, and measure the absorbance of
the reacted BCA reagent at 562 nm. The absorbance can be
converted to a concentration using a calibration curve pre-
pared from a stock sample of the affinity molecule with known
concentration, or BSA protein as an estimate. The number of
molecules per MNP can then be calculated using the molecu-
lar weight, the MNP concentration from step 6,
Subheading 3.3, and the molecular weight of the MNP (see
Note 8).
8. Use dynamic light scattering, if desired, to ensure that the
MNP did not aggregate during conjugation procedures.
3.4. MRSw Assay 1. This section describes the detection of soluble molecules
for Soluble Analyte based on the principle of MRSw, in which the molecular tar-
Sensing get induces assembly or disassembly of MNP into clusters and
causes a change in the bulk T2 relaxation time (see Fig. 2).
Disassembly MRSw assays are the preferred method for the
detection of enzymes (cleavage of specific sites in the cross-
bridges) and small molecules (competitive binding). In this
case, MNP must first be clustered, which has been accom-
plished using several different strategies (see Table 1).
a
70
Mismatch Match
T2 (ms)
50
0 0.5 1.0 1.3 1.7 2.7 30

0 0.5 1.0 1.5
Oligo added (fmol) Oligo added (fmol)
b Protein sensing c Enzyme sensing

GFP Caspase 3
High T2
High T2 Low T2
Low T2
60
BSA 80
60
Caspase 3
50 50
T2 (ms)
60
T2 (ms)
T2 (ms)
40
30
40 0 10 20 30 40
GFP (fmol) Inhibitor
GFP 20
30
0 10 20 30 40 0 10 20
Time (min) Time (min)
Fig. 2. DMR detection of biomolecules using magnetic relaxation switching (MRSw). (a) Detection of an oligonucleotide
target using MNP conjugated with complementary oligonucleotide sequences. The left panel displays a T2-weighted MR
image of a 384-well plate containing varying amounts of target or mismatched oligonucleotide and constant levels of
MNP. Hybridization to the target causes the MNP to cluster, resulting in a corresponding decrease in T2 relaxation time.
(b) Detection of GFP protein using MNP conjugated with a polyclonal antibody specific for GFP. T2 relaxation time, which
was determined using a benchtop NMR relaxometer, decreased linearly with GFP concentration, but was not affected by
the concentration of a control protein (BSA). (c) Detection of caspase 3 enzymatic activity using MNP that were clustered
using a linker containing the peptide sequence DEVD. Introduction of caspase 3 resulted in rapid cleavage of the peptide
sequence and an increase in T2 relaxation time, which was abrogated by the addition of a caspase 3 inhibitor. Reproduced
with permission from ref. 4, copyright (2002) Nature Publishing Group.
2. Dilute the affinity molecule-MNP sample with PBS to

10–20 mg/mL total Fe concentration (see Note 12).
3. For preclustering MNP prior to a disassembly assay, add the
cross-linking agent at the optimal concentration to induce
aggregation as determined by experimentation (see Note 13).
Incubate for at least 1 h at room temperature. Longer incubations
44 Haun et al.
may produce more pronounced aggregation depending on

the rate of cluster formation. Confirmation of MNP aggrega-
tion can be obtained by dynamic light scattering at this point.
Then, combine equal volumes of the aggregated MNP solu-
tion and analyte sample.
4. For aggregation assays, add an equal volume of analyte
sample to the MNP solution.
5. Incubate for at least 1 h at room temperature and proceed to
analysis (Subheading 3.6 or 3.7).
3.5. Magnetic 1. This section describes the detection of biomarkers on the sur-
Tagging of Cells face of intact, suspended cells by tagging with MNP to impart
magnetic susceptibility (see Fig. 3). Similar methods can be
used to label adherent cells, followed by disruption and sus-
pension. In addition, intracellular markers can be tagged if
the cells are first permeabilized (see Note 14).
2. Wash the cell sample by centrifuging at 300 × g for 5 min, aspi-
rating the supernatant, and resuspending in 0.5 mL PBS+.
3. Add the affinity molecule-MNP sample to the cell suspen-
sion. When using antibodies as the affinity molecule, a final
MNP concentration of 100 nM (~45 mg/mL total Fe con-
centration for CLIO, see Note 8) should be sufficient (see
Note 15).
4. Incubate for 30 min at room temperature on an orbital
shaker.
5. Remove unbound MNP using two rounds of centrifugation
as described in step 2, Subheading 3.5, but use 1 mL of
ice-cold PBS+ each time.
a b Current DMR technology

0.8 Mn-MNP (new) Cytology
R 2 > 99% 100 Histology, cell block
75
∆R 2 (s−1)
0.4
∆T (%)
CLIO (old) 50
2
25
0.0
0
0 1 2 3 100 101 102 103 104
Cell counts (x103 cells) Cell counts
Fig. 3. DMR detection of tumor cells using the tagging method. (a) Her2/neu was detected on breast cancer cells (BT474)
using an anti-Her2/neu antibody that was conjugated to CLIO and Mn-MNP nanoparticles. Transverse relaxation rate
(r2 = 1/T2) was measured using a miniaturized NMR (mNMR) detector, and varied proportionally with cell number and the
magnetic susceptibility of the nanoparticle employed. (b) As few as two cells could be detected using the Mn-MNP nano-
particle, well above the detection threshold of established clinical methods (cytology and histology). Reproduced with
permission from ref. 3, copyright (2009) National Academy of Sciences, USA.
6. Resuspend in the minimum volume PBS+ necessary for the

detection method of choice (see Subheadings 3.6 and 3.7).
3.6. Detection 1. Magnetic resonance signal from MRSw or tagged cell samples
Using NMR can be detected using NMR relaxometers operating at low
frequency and magnetic field strength. Benchtop systems
such as the Bruker Minispec (20 MHz, <1 T) measure T1 and
T2 relaxation times of samples in NMR tubes (see Fig. 2b, c).
Alternatively, mNMR devices can be used to detect T1 or T2
within microfluidic channels (see Fig. 3).
2. For benchtop NMR systems, load sample into 5 (>0.3 mL) or
10 mm (>0.5 mL) NMR tubes. For mNMR devices, sample
volumes as low as 1 mL have been achieved using microfluidic
elements (3).
3. Measure T2 relaxation time (and T1 if desired). For both
benchtop and miniaturized systems, T1 relaxation time is
measured using inversion recovery pulse sequences. For T2
measurement, Carr–Purcell–Meiboom–Gill (CPMG) spin-
echo pulse sequences are employed to compensate for the
spatial inhomogeneity of the external magnetic field. The
typical echo time is 2–5 ms and the repetition time is »5 × T1
to ensure full recovery of nuclear spins (2).
3.7. Detection 1. Clinical or experimental MRI scanners can be used to detect

Using MRI magnetic resonance signals from samples loaded into multi-
well plates (see Fig. 2a). MRI scanners employ strong mag-
netic fields (1–11 T) generated by superconducting magnets,
and use sophisticated data acquisition schemes.
2. Load 50 mL sample into a 384-well plate.
3. Obtain a T2 map using T2 spin-echo sequences with variable echo
times. Typically, echo time is varied between 25 and 1,000 ms,
and the repetition time ranges between 2,000 and 3,000 ms.
4. Analyze the T2 image using the appropriate software.
4. Notes
1. If the MNP or affinity molecule storage buffer contains con-

taminants that will interfere with downstream processing
(i.e., Tris or carrier protein for amine reactions, sodium azide
for click chemistry), they must first be removed by desalting
or chromatography.
2. It is important to avoid divalent cations when working with
nanoparticles because they can crosslink the nanoparticles,
leading to aggregation. Care should also be taken that chemi-
cal modifications to the nanoparticles, including conjugation
46 Haun et al.
of proteins or peptides, do not alter the surface chemistry

such that aggregation results.
3. A polytriazole ligand such as bathophenanthroline disulfonic
acid should be added to the copper catalyst buffer to stabilize
the reduced form of copper [Cu(I)] that is required for cata-
lytic activity (22).
4. When using an enzyme to disrupt tissues, caution should be
taken that the biomarker of interest is not affected by the
activity of the enzyme.
5. The total cell requirement for tagging assays is dependent on
the target expression level and the detection sensitivity of the
magnetic resonance sensor platform. For NMR measure-
ments, detection thresholds using CLIO are in the range of
10,000 cells using a benchtop relaxometer (Minispec) and
1,000 cells using the miniaturized NMR (mNMR) for a high
expression level marker (millions of copies per cell, see Fig. 3).
Use of higher magnetization MNP can decrease the detection
threshold to near single-cell for the mNMR.
6. Fluorophores can aid significantly in optimizing the binding
of the MNP conjugates to cells using fluorescence microscopy
or flow cytometry and in tracking the MNP concentration
during processing steps using a fluorometer. Fluorophores
should be attached prior to affinity molecule conjugation.
Since excess reagents are not required for this conjugation, it
is easier to control the degree of modification and MNP
cross-linking is not an issue. Care should be taken that func-
tional groups remain for affinity molecule conjugation, how-
ever. After dye conjugation, the free dye can be removed by
desalting and the amount of dye attached can be determined
by absorbance (known extinction coefficient) or fluorescence
(compared to a standard) measurement. Following dye con-
jugation, the nanoparticles are typically referred to as mag-
neto-fluorescent nanoparticles (MFNP).
7. Iron oxide MNP are brown and visible to the naked eye at
concentrations above 100 mg Fe/mL. Therefore, when work-
ing with samples above this concentration, column purifica-
tions can be performed manually. For lower concentration
samples, purifications should be performed using an auto-
mated system (i.e., AKTA fPLC from GE Healthcare) so that
absorbance (410 nm) can be monitored.
8. The molecular weight of the MNP can be calculated based on
the number of Fe atoms per core and the molecular mass of
Fe (55.85 g/mol). For example, CLIO have approximately
8,000 Fe atoms/core (23), thus the estimated average molec-
ular weight is 447,000 g/mol.
9. For modifications of MNP and affinity molecules using amine-
reactive N-hydroxy-succinimidyl (NHS) esters, reaction times
of 1–2 h are common since the NHS groups hydrolyze in

water. However, longer incubations may increase the reaction
yield somewhat and should not adversely affect the final prod-
uct provided that stability is not an issue.
10. MNP aggregation can result if the affinity molecule contains
multiple coupling sites (e.g., such as multiple thiol, click
reagent, or biotin molecules on a single macromolecular pro-
tein). For this reason, the affinity molecule modifications call
for a minimal excess of amine-reactive reagent (i.e., two- to
fivefold). These recommended values may need to be adjusted
to yield approximately 1–2 coupling moieties per affinity mol-
ecule. Also of note, excess thiols can self-react to form disul-
fide bonds, and therefore should be capped with a reagent
such as iodoacetamide.
11. Unreacted maleimide and thiol groups remaining after thiol
couplings can potentially react with biomolecules or cells,
increasing background adhesion and thus decreasing detection
specificity. In some cases, capping of free thiol groups using
iodoacetamide can help improve binding specificity. However,
this should be performed following the purification steps
described in steps 4 and 5, Subheading 3.3. Capping is not
required when using a bioorthogonal click coupling chemistry.
12. For CLIO, the 10–20 mg/mL Fe recommended for MRSw
assays equates to approximately 20–40 nM and typically reg-
isters a T2 relaxation time of 50–100 ms. Due to the extra
dilution factor associated with disassembly MRSw assays, the
initial MNP concentration used is usually 20 mg/mL.
13. MNP clustering is governed by the equivalence principle,
which dictates that clustering is greatest when a multivalent
binder (i.e., MNP) and cross-linking agent are present at
equimolar concentrations (13). In this case, the pertinent
molar concentration is the number of affinity molecules per
MNP, not the number of MNP.
14. Intracellular markers can be detected by adding 0.1% saponin
to the PBS+ buffer. In this case, longer incubations are rec-
ommended (1 h) to increase MNP penetration into the cell,
as well as extended washes at room temperature to allow
unbound MNP to diffuse out of the cells.
15. The concentration of affinity molecule-MNP used for cell
tagging can be lowered below 100 nM if the binding kinetics
of the affinity molecule is sufficiently high. For example, a
concentration of 10 nM would be sufficient if the equilibrium
dissociation constant (KD) of the interaction is subnanomo-
lar. Likewise, the concentration should be increased if the
binding kinetics is poor (i.e., micromolar KD). The ideal con-
centration should be determined experimentally for each
affinity molecule-MNP of interest.
48 Haun et al.
Acknowledgments
The authors thank Nikolay Sergeyev for technical guidance

regarding MNP synthesis and characterization and Dr. Neal
K. Devaraj for assistance with click chemistry protocols.
References
1. Tamanaha, C. R., Mulvaney, S. P., Rife, J. C., 11. Peng, S., Wang, C., Xie, J., and Sun, S. (2006)
and Whitman, L. J. (2008) Magnetic labeling, Synthesis and stabilization of monodisperse
detection, and system integration. Biosens. Fe nanoparticles. J. Am. Chem. Soc. 128,
Bioelectron. 24, 1–13. 10676–10677.
2. Lee, H., Sun, E., Ham, D., and Weissleder, R. 12. Koh, I., Hong, R., Weissleder, R., and
(2008) Chip-NMR biosensor for detection Josephson, L. (2008) Sensitive NMR sensors
and molecular analysis of cells. Nat. Med. 14, detect antibodies to influenza. Angew. Chem.
869–874. Int. Ed. Engl. 47, 4119–4121.
3. Lee, H., Yoon, T. J., Figueiredo, J. L., Swirski, 13. Koh, I., Hong, R., Weissleder, R., and
F. K., and Weissleder, R. (2009) Rapid detec- Josephson, L. (2009) Nanoparticle-target
tion and profiling of cancer cells in fine-needle interactions parallel antibody-protein interac-
aspirates. Proc. Natl Acad. Sci. USA 106, tions. Anal. Chem. 81, 3618–3622.
12459–12464. 14. Sun, E. Y., Josephson, L., Kelly, K. A., and
4. Perez, J. M., Josephson, L., O’Loughlin, T., Weissleder, R. (2006) Development of nano-
Hogemann, D., and Weissleder, R. (2002) particle libraries for biosensing. Bioconjug.
Magnetic relaxation switches capable of sensing Chem. 17, 109–113.
molecular interactions. Nat. Biotechnol. 20, 15. Kolb, H. C., Finn, M. G., and Sharpless, K. B.
816–820. (2001) Click chemistry: diverse chemical
5. Josephson, L., Perez, J. M., and Weissleder, function from a few good reactions. Angew.
R. (2001) Magnetic nanosensors for the Chem. Int. Ed. Engl. 40, 2004–2021.
detection of oligonucleotide sequences. 16. Rostovtsev, V. V., Green, L. G., Fokin, V. V.,
Angew. Chem. Int. Ed. Engl. 40, 3204–3206. and Sharpless, K. B. (2002) A stepwise huis-
6. Lee, H., Yoon, T. J., and Weissleder, R. (2009) gen cycloaddition process: copper(I)-cata-
Ultrasensitive detection of bacteria using core- lyzed regioselective “ligation” of azides and
shell nanoparticles and an NMR-filter system. terminal alkynes. Angew. Chem. Int. Ed. Engl.
Angew. Chem. Int. Ed. Engl. 48, 5657–5660. 41, 2596–2599.
7. Josephson, L., Tung, C. H., Moore, A., and 17. Tornoe, C. W., Christensen, C., and Meldal,
Weissleder, R. (1999) High-efficiency intrac- M. (2002) Peptidotriazoles on solid phase:
ellular magnetic labeling with novel super- [1,2,3]-triazoles by regiospecific copper(I)-
paramagnetic-Tat peptide conjugates. catalyzed 1,3-dipolar cycloadditions of termi-
Bioconjug. Chem. 10, 186–191. nal alkynes to azides. J. Org. Chem. 67,
8. Lu, A. H., Salabas, E. L., and Schuth, F. 3057–3064.
(2007) Magnetic nanoparticles: synthesis, 18. Sun, E. Y., Josephson, L., and Weissleder, R.
protection, functionalization, and application. (2006) “Clickable” nanoparticles for targeted
Angew. Chem. Int. Ed. Engl. 46, 1222–1244. imaging. Mol. Imaging 5, 122–128.
9. Harisinghani, M. G., Barentsz, J., Hahn, P. 19. Thorek, D. L., Elias, D. R., and Tsourkas, A.
F., Deserno, W. M., Tabatabaei, S., van de (2009) Comparative analysis of nanoparti-
Kaa, C. H., et al. (2003) Noninvasive detec- cle-antibody conjugations: carbodiimide
tion of clinically occult lymph-node metasta- versus click chemistry. Mol. Imaging 8,
ses in prostate cancer. N. Engl. J. Med. 348, 221–229.
2491–2499. 20. Devaraj, N. K., Upadhyay, R., Haun, J. B.,
10. Lee, J. H., Huh, Y. M., Jun, Y. W., Seo, J. W., Hilderbrand, S. A., and Weissleder, R. (2009)
Jang, J. T., Song, H. T., et al. (2007) Fast and sensitive pretargeted labeling of can-
Artificially engineered magnetic nanoparticles cer cells through a tetrazine/trans-cyclooctene
for ultra-sensitive molecular imaging. Nat. cycloaddition. Angew. Chem. Int. Ed. Engl.
Med. 13, 95–99. 48, 7013–7016.
21. Haun, J. B., Devaraj, N. K., Hilderbrand, S. A., protease assays. Angew. Chem. Int. Ed. Engl.
Lee, H., and Weissleder, R. (2010) Bioor 42, 1375–1378.
thogonal chemistry amplifies nanoparticle 29. Perez, J. M., Simeone, F. J., Tsourkas, A.,
binding and enhances the sensitivity of cell Josephson, L., and Weissleder, R. (2004)
detection. Nat. Nanotechnol. 5, 660–665. Peroxidase substrate nanosensors for MR
22. Chan, T. R., Hilgraf, R., Sharpless, K. B., and imaging. Nano Lett. 4, 119–122.
Fokin, V. V. (2004) Polytriazoles as copper(I)- 30. Tsourkas, A., Hofstetter, O., Hofstetter, H.,
stabilizing ligands in catalysis. Org. Lett. 6, Weissleder, R., and Josephson, L. (2004)
2853–2855. Magnetic relaxation switch immunosensors
23. Reynolds, F., O’Loughlin, T., Weissleder, R., detect enantiomeric impurities. Angew. Chem.
and Josephson, L. (2005) Method of deter- Int. Ed. Engl. 43, 2395–2399.
mining nanoparticle core weight. Anal. Chem. 31. Sun, E. Y., Weissleder, R., and Josephson, L.
77, 814–817. (2006) Continuous analyte sensing with mag-
24. Grimm, J., Perez, J. M., Josephson, L., and netic nanoswitches. Small 2, 1144–1147.
Weissleder, R. (2004) Novel nanosensors for 32. Atanasijevic, T., Shusteff, M., Fam, P., and
rapid analysis of telomerase activity. Cancer Jasanoff, A. (2006) Calcium-sensitive MRI
Res. 64, 639–643. contrast agents based on superparamagnetic
25. Kim, G. Y., Josephson, L., Langer, R., and iron oxide nanoparticles and calmodulin. Proc.
Cima, M. J. (2007) Magnetic relaxation Natl Acad. Sci. USA 103, 14707–14712.
switch detection of human chorionic gonado- 33. Taktak, S., Weissleder, R., and Josephson, L.
trophin. Bioconjug. Chem. 18, 2024–2028. (2008) Electrode chemistry yields a nanopar-
26. Perez, J. M., Grimm, J., Josephson, L., and ticle-based NMR sensor for calcium. Langmuir
Weissleder, R. (2008) Integrated nanosensors 24, 7596–7598.
to determine levels and functional activity of 34. Perez, J. M., Simeone, F. J., Saeki, Y.,
human telomerase. Neoplasia 10, 1066–1072. Josephson, L., and Weissleder, R. (2003)
27. Perez, J. M., O’Loughin, T., Simeone, F. J., Viral-induced self-assembly of magnetic nano-
Weissleder, R., and Josephson, L. (2002) particles allows the detection of viral particles
DNA-based magnetic nanoparticle assembly in biological media. J. Am. Chem. Soc. 125,
acts as a magnetic relaxation nanoswitch 10192–10193.
allowing screening of DNA-cleaving agents. 35. Kaittanis, C., Naser, S. A., and Perez, J. M.
J. Am. Chem. Soc. 124, 2856–2857. (2007) One-step, nanoparticle-mediated bac-
28. Zhao, M., Josephson, L., Tang, Y., and terial detection with magnetic relaxation.
Weissleder, R. (2003) Magnetic sensors for Nano Lett. 7, 380–383.
Chapter 4
Real-Time Quantum Dot Tracking of Single Proteins

Jerry C. Chang and Sandra J. Rosenthal
Abstract
We describe a single quantum dot tracking method that can be used to monitor individual proteins in the
membrane of living cells. Unlike conventional fluorescent dyes, quantum dots (fluorescent semiconductor
nanocrystals) have high quantum yields, narrow emission wavelengths, and excellent photostability, making
them ideal probes in single-molecule detection. This technique has been applied to study the dynamics
of various membrane proteins including glycine receptors, nerve growth factors, kinesin motors, and
g-aminobutyric acid receptors. In this chapter, a basic introduction and experimental setup for single
quantum dot labeling of a target protein is given. In addition, data acquisition and analysis of time-lapse
single quantum dot imaging with sample protocols are provided.
Key words: Quantum dot, Biological labeling, Biophysics, Protein trafficking, Single-particle tracking,
Fluorescence microscopy
1. Introduction
Over the past decade, research utilizing semiconductor quantum

dots (qdots) has continuously shown huge potential for in vitro
and in vivo biological imaging (1–6). The great interest among
researchers is due to the unique photophysical properties inherent
to qdots, such as narrow emission wavelengths, excellent photo-
stability, and exceptionally high quantum yields (7). These unique
properties allow qdots to serve as ideal probes in biological detec-
tion. Another interesting property which makes qdots a near-perfect
candidate for single-molecule studies is the fluorescent intermit-
tency, or blinking, phenomenon (8). Taking advantage of this
unique property, conventional biochemical studies such as Western
blot analysis (9) and antibody affinity assays (10) are being suc-
cessfully pushed forward toward the molecular scale.
51
52 Chang and Rosenthal
In 2003, the Dahan group published the first study using

single qdots for the detection of glycine receptors in living cells
(11). Since that time, a few groups have successfully expanded
single qdot tracking for several different targets including nerve
growth factors (12, 13), kinesin motors (14), and g-aminobutyric
acid (GABA) receptors (15). More recently, single qdot detection
for an in vivo mouse tumor model has been demonstrated (16).
This chapter describes the principles, methodologies, and
basic experimental protocols for the generation of single quantum
dot tracking in living cells. The first part describes the protocol
for imaging system calibration using spin-coated single qdots.
The second part describes a protocol for targeted labeling of single
qdots in living cells. However, site-specific labeling of target pro-
tein is a difficult task and it is not possible to provide a universal
protocol for this purpose. Therefore, only the protocol derived in
our laboratory is provided. The third part describes the protocols
for the image processing and data analysis of single quantum dot
tracking (see Fig. 1). In general, time-lapse images of live cells
labeled with single qdots are first acquired using an optical micro-
scope [e.g., epifluorescence, confocal, or total internal reflection
fluorescence (TIRF) microscope]. The spot positions of single
qdots are subsequently derived from the time-lapse imaging data.
Fig. 1. Approach to single quantum dot tracking.

Real-Time Quantum Dot Tracking of Single Proteins 53
The latter step allows for the generation of single-molecule

trajectories of the target proteins. After trajectory data analysis,
diffusion dynamic behaviors of target proteins can be obtained.
2. Materials
2.1. Reagents 1. Biotinylated small molecule or antibody against an extracel-

lular epitope.
2. Qdot streptavidin conjugate (1 mM, Invitrogen Corporation,
Carlsbad, CA).
3. 35-mm cell culture dishes with coverglass bottom (MatTek
Corporation, Ashland, MA).
4. HeLa Cells.
5. Dulbecco’s Modified Eagle’s Medium (DMEM) (GIBCO,
Invitrogen Corporation, Carlsbad, CA).
6. Phenol red-free DMEM (GIBCO, Invitrogen Corporation,
Carlsbad, CA).
7. Penicillin–streptomycin antibiotic mixture (100×) (GIBCO,
Invitrogen Corporation, Carlsbad, CA).
8. Fetal bovine serum (FBS) (GIBCO, Invitrogen Corporation,
Carlsbad, CA).
9. l-Glutamine (GIBCO, Invitrogen Corporation, Carlsbad, CA).
2.2. Equipment, 1. Fluorescent microscope (see Subheading 3.1).

Software, and 2. Microscope mounted heating chamber.
Accessories
3. Microscope image acquisition and analysis software
(MetaMorph 7.6, Molecular Devices, Sunnyvale, CA).
4. Technical computing software for numerical analysis (Matlab
R2008b, MathWorks, Natick, MA).
5. Spin coater.
3. Methods
The methods described below outline (1) the preparation of the

nanoconjugate probes, (2) time-lapse imaging for single quantum
dot tracking, and (3) trajectory data construction and analysis. The
general principle provided in this protocol can be applied to differ-
ent optical imaging systems including epifluorescence, confocal,
and TIRF microscopes. Readers seeking a more detailed comparison
of epifluorescence, confocal, and TIRF microscopy for single-
molecule experiments are advised to consult Lang et al. (17).
3.1. Imaging System A general strategy to identify whether the optical system is able to
Calibration Using detect individual qdots is to carry out time-lapse imaging of a very
Spin-Coated Single dilute qdot solution to avoid interparticle reactions. Individual
Qdots qdots are characterized by their unique blinking properties (8, 18).
As an example, the emission intensity of a single quantum dot is
shown in Fig. 2, in which a single quantum dot blinks completely
on and off during a time-lapse sequence of 60 s at a 20-Hz frame
rate. When the fluorescence is produced by an aggregate structure
consisting of several qdots, such blinking effects are completely can-
celed out. The protocol described below is based on a custom-built
Zeiss Axiovert 200M inverted fluorescence microscope together
with a charge-coupled device (CCD) camera (Cool-SnapHQ2, Roper
Scientific, Trenton, NJ). To track single qdots on the fly, the acqui-
sition rate should be set at 10 Hz or higher. However, the imaging
rate is usually limited by the frame readout time of the camera. This
particular CCD is chosen due to its decent 60% quantum efficiency
(QE) throughout the entire visible spectrum (450–650 nm) with a
frame rate >20 at 512 × 512 pixels (see Photometrics CCD specifica-
tion document at http://www.photomet.com). A more advanced
back-illuminated electron multiplying CCD (EMCCD) with sub-
millisecond temporal resolution will be a much better choice since
the EMCCD can achieve single-photon sensitivity with exception-
ally high frame rates at 10 MHz (19). Imaging should be performed
with a high-resolution (63× or 100×) oil-immersion objective lens
with a numerical aperture of 1.30 or greater.
1. Prepare a clean microscope glass slide coverslip (or 35-mm
culture dish with coverslip in the bottom).
2. Add one drop (20 mL) of 1 nM Qdot® Streptavidin conjugate
solution onto the coverslip.
Fig. 2. Intensity of fluorescence as a function of time measured from a single quantum

dot (data taken with a Zeiss LSM 5 Live Confocal Microscope at a 20-Hz frame rate). As
displayed in the figure, the intensity trajectory of a single qdot displays two dominant
states: an “on” state and an “off” state, termed blinking.
3. Spin cast the qdot solution on the coverslip for 30 s at

2,000 rpm or less (see Note 1).
4. Mount the coverslip on the microscope stage.
5. Acquire time-lapse images (10 Hz, 60 s).
In order to obtain a well-dispersed nanoparticle sample and
minimize the number of nanoparticle aggregates, the volume and
concentration of Qdot® Streptavidin conjugate (strep-qdot) solu-
tion in step 2 will need to be adjusted depending on the adhesion
property between the coverslip and the qdots. In our experiments
using Qdot® 655 Streptavidin conjugate solution with MatTek
dishes, less than 2% of the detected fluorescent spots observed
were aggregates.
3.2. Cell Culture HeLa cells are cultured in DMEM supplemented with 10% FBS,
2 mM l-glutamine, 100 units/mL penicillin, and 100 mg/mL
streptomycin and maintained at 37°C with 5% CO2. For single
quantum dot labeling studies, cells are plated at a density of
1 × 105 cells/mL in 35-mm coverslip-buttoned culture dish as
viable at the single-cell level.
3.3. Single Quantum Single quantum dot labeling can be prepared through either a
Dot Labeling and direct labeling (one step) procedure or an indirect (two step) pro-
Real-Time Imaging tocol (see Note 2). In the direct labeling procedure, the target-
in Live Cells specific biotinylated probe (small-molecule ligand or antibody) is
premixed with the strep-qdots to make ligand-qdot nanoconju-
gates. Therefore, the cellular labeling strategy could be performed
in one step in which the live cell sample is treated with a target-
specific nanoconjugate prior to fluorescent imaging. However,
this procedure might lead to multivalent quantum dot–protein
interactions (see Note 3). Other covalent conjugation strategies
are also available for the immobilization of different surface mod-
ified functional groups, such as carboxylic acids, on the surface of
water-soluble quantum dots. For alternative conjugation methods,
refer to Note 4.
In the two-step procedure, the cell sample is first incubated
with biotinylated ligand or antibody. This incubation should yield
the desired specific binding between the target protein and
biotinylated ligand or antibody. After appropriate washing steps,
strep-qdot is added as the fluorescent tag for the single-molecule
imaging. For the labeling of a transfected cell sample, the stan-
dard protocol given below should be followed:
1. Prepare a 35-mm coverslip-buttoned culture dish with
cells that have reached about 50% confluence (see
Subheading 3.2).
2. Wash the cells gently three times with phenol red-free culture
medium by repeatedly pipetting out.
3. Incubate cells with a biotinylated small molecule (0.5–2 mM)

or antibody (1–10 mg/mL) in red-free DMEM for 20 min at
37°C.
4. Wash cells gently three times with phenol red-free culture
medium by repeatedly pipetting out.
5. Incubate the cells with Qdot Streptavidin conjugate
(0.5–1 nM) in phenol red-free culture medium for 5 min at
37°C (see Note 5).
6. Wash the cells at least three times with phenol red-free culture
medium.
7. Place the culture dish on the microscope stage with mounted
heating chamber.
8. Acquire time-lapse images at room temperature or 37°C (see
Note 6).
3.4. Single Quantum After gathering a series of time-lapse images of live cells labeled
Dot Localization with single qdots (see Fig. 3), trajectories of individual target pro-
and Trajectory teins can be extracted by postimage processing and analysis. Data
Construction analysis can by complicated by the fact that many laboratories
develop custom-made routines for single-molecule data analysis.
The following methods describe not only basic principles but also
rather available software routines which we believe are suitable for
single qdot tracking.
3.4.1. Estimation of 2D In optical microscopy, the observed intensity distribution from a

Position with Subpixel single point source is described by the point spread function
Accuracy (PSF). The theoretical PSF can be calculated:
 2 J ((2p / l ) NAr )
2
PSF(r ) = C ×  1  , (1)
 (2p / l ) NAr 
Fig. 3. Example images of membrane proteins labeled with single qdots (a: bright field image, b: fluorescence image).
Note that the blinking phenomenon in the time-lapse images should be used as a signature to confirm if the individual
spots represent single molecules.
where r is the distance from the origin, NA is the numerical aperture,

C defines the intensity value at r = 0, J1 is the first-order Bessel
function, and l is the wavelength of light. As a subwavelength
nanostructure, fluorescent images of single qdots are known
to fulfill the PSF and have been fit well with 2D Gaussians (20).
In order to calculate a subpixel estimate of single qdot position,
the general method is to consider the intensity distribution of a
single qdot as an elliptical Gaussian intensity function and calculate
the local maximum intensity with Gaussian interpolation:
ln( z x −1 ) − ln( z x +1 )
x0 = , (2)
2[ln( z x +1 ) − 2ln( z x ) + ln( z x −1 )]
where zx is the local maximum intensity, and zx−1 and zx+1 are the
two neighbors. The maximum of the interpolated curve is
located at x0 with respect to the index X of the maximum sample
(see Fig. 4).
Fig. 4. (a) Schematic of 2D Gaussian regression of a single quantum dot fluorescent image. (b) Example of a direct Gaussian
fit along the x- and y-axis to the intensity distribution of a single quantum dot fluorescent image. Note that a much more
accurate localization in the center can be obtained by matching the Gaussian function to fit the experimental intensity data.
A tracking routine based on the least-square 2D Gaussian

regression for subpixel position estimation in single-particle tracking
(SPT) was initially developed by Crocker and Grie (21). The
source code and an excellent introduction can be found at http://
www.physics.emory.edu/~weeks/idl/. Matlab versions of these
routines have been made by Blair and Dufresne and are available
at http://physics.georgetown.edu/matlab/ as detailed, step-by-step
tutorials. These routines have been successfully used to calculate
the diffusion dynamics of single qdots at silica surfaces in static
and flow conditions (22).
3.4.2. Trajectory After x and y coordinates of the selected single qdots were deter-
Construction and 2D mined in each frame, step displacements of each quantum dot,
Displacement Generation which represent the time course of the corresponding moment
are determined by the following formula (see Fig. 5):
{ }
(1 / 2 )
∆ dn = [ x(n + 1 ) − x(n)]2 + [ y(n + 1 ) − y(n)]2 , (3)
Fig. 5. (a) Schematic of XY trajectory and 2D step displacement of single quantum dot tracking. (b) An example of the XY
trajectories of individual transmembrane proteins labeled with single qdots in living cells.
where x (n) and y (n) denote the position in frame (n), Ddn indicates
that a single step takes place during a single lag time Dt from time
point Dt × (n) to Dt × (n + 1).
3.5. Diffusion In addition to the standard XY trajectory and 2D displacement

Behavior Analysis plots, mean-square displacement (MSD) for individual trajecto-
ries should also be used for probing the single protein diffusion
dynamics in a plasma membrane. MSD can be obtained according
to the formula given below:
2
N −1−n  [x( j∆t + n∆t ) − x( j∆t )] 
MSD (nd t ) = (N − 1 − n)−1 ∑  , (4)
j =1  − [y( j∆t + n∆t ) − y( j∆t )]2 
 
where x(t) and y(t) are the position of particle at time t. x(jDt + nDt)
and y(jDt + nDt) are the position following a time interval nDt, N is
total number of frames, Dt is time-resolution, n is the number of
time intervals, and j is a positive integer.
Once the MSD values for individual trajectories were obtained,
the diffusion coefficient could be calculated through the best fit
of the MSD with the theoretical model listed below (see Fig. 6).
For an excellent overview with details of the theoretical models
of membrane diffusion dynamics as well as the principle of SPT,
see Saxton and Jacobson (23).
MSD (t ) = 4D Dt Normal diffusion, (5)

MSD (t ) = 4D Dt a Anomalous diffusion, (6)
MSD (t ) = 4D Dt + (nDt )2 Directed motion with diffusion, (7)

  4 A D ∆t  
MSD (t ) = rc2 1 − A1 exp  − 2 2   Corralled motion, (8)
  〈 rc 〉  
Fig. 6. Different types of diffusion behavior and their corresponding mean-square displacement (MSD) curve for (a) normal
diffusion, (b) direct motion, and (c) corralled motion. The insets to the plots schematically show expected 2D trajectories.
Note that the diffusion behavior and MSD curve of anomalous diffusion are expected between normal diffusion and
corralled motion.
where D denotes the diffusion coefficient, Dt is time-resolution, a

is the anomalous diffusion exponent, n is velocity, 〈 rc 〉 indicates
2
the corral size, and A1 and A2 are constants determined by the

corral geometry.
Due to its simplicity and ability to produce rapid results, the
random walk model (normal diffusion, Eq. 5) is commonly used
to estimate the diffusion coefficient in single quantum dot track-
ing experiments (11). However, in biological interpretation,
anomalous diffusion (Eq. 6) was reported as the dominant model
of diffusion dynamics in the plasma membrane (24, 25). The
standard method to obtain the diffusion coefficient from the
anomalous diffusion model is through a linear fit, as indicated
below, since this fit converges more consistently than the variable
power-law fit (25):
log[ MSD (t )] = a log( Dt ) + log(4 D). (9)
4. Notes
1. The spinning force is not critical in this step. In most cases,

a rotational speed as low as 500 rpm is sufficient to achieve
a uniformly spread.
2. It must be emphasized that the key parameter for preparation
of the functionalized qdot probe for single-molecule tracking
is to use a target-specific small molecule or antibody with low
nonspecific binding. All labeling protocols should employ
three negative control experiments to demonstrate binding
specificity prior to the single qdot tracking experiments:
(1) labeling of a parental cell lacking target protein expression,
(2) labeling of a target protein-expressing cell preblocked
with an inhibitor or antibody specific to the intended target,
and (3) incubation of a target protein-expressing cell with
quantum dots only (without a target-specific small molecule
or antibody). These negative controls will provide an inde-
pendent assessment of whether the observed single qdot
labeling was the result of nonspecific interactions arising at
the cell membrane–qdot interface, or the result of specific
interactions with the intended target.
3. For one-step labeling, it is important to remember that each
strep-qdot is typically conjugated with 5–10 streptavidins
(Qdot® Streptavidin Conjugates User Manual, Invitrogen,
CA). Using a nanoconjugate probe prepared from the one-
step procedure may result in more than one protein binding
to a single qdot, which leads to multivalent quantum dot–
protein interactions where the movement of single quantum
dot would not be able to reflect individual membrane protein
trafficking. In this case, the concentration of the ligand in the

reaction mixture is important in determining the final ligand/
quantum dot ratio of the coupling. Incubation of biotinylated
ligand with strep-qdot at a 1:1 molar ratio was typically used.
Recently, Howarth et al. published a modified protocol which
provide a much better solution in the case where monovalent
instead of multivalent streptavidin-conjugated qdots were
prepared (26).
4. In contrast to qdot probes prepared by using Qdot®
Streptavidin Conjugate, Qdot® Antibody Conjugation Kits,
and ITK™ Carboxyl Quantum Dots can also be used for
direct conjugation. The experimental procedures used to per-
form direct conjugation and purification are described in
detail elsewhere (27, 28). Other general protocols are also
available on Invitrogen’s web site (http://www.invitrogen.com).
5. An average of 10–20 qdot-labeled proteins/cell is expected
to be presented.
6. The experiment is carried out at room temperature to settle
lower imaging rate with decreasing particle velocity, however,
imaging at 37°C is closer to the physiological condition.
Acknowledgments
The authors wish to thank their colleagues in the group, espe-

cially Dr. James McBride, Dr. Michael Schreuder, Albert Dukes,
and Oleg Kovtun, for all the helpful discussions and suggestions.
We thank Dr. David Piston for helpful advice with single quantum
dot tracking experimental setup. This work was supported by
grants from National Institutes of Health (R01EB003778).
References
1. Bruchez, M., Moronne, M., Gin, P., Weiss, S., 5. Rosenthal, S. J., Tomlinson, I., Adkins, E. M.,
and Alivisatos, A. P. (1998) Semiconductor Schroeter, S., Adams, S., Swafford, L., et al.
nanocrystals as fluorescent biological labels. (2002) Targeting cell surface receptors with
Science 281, 2013–2016. ligand-conjugated nanocrystals. J. Am. Chem.
2. Chan, W. C. and Nie, S. (1998) Quantum dot Soc. 124, 4586–4594.
bioconjugates for ultrasensitive nonisotopic 6. Liu, W., Choi, H. S., Zimmer, J. P., Tanaka, E.,
detection. Science 281, 2016–2018. Frangioni, J. V., and Bawendi, M. (2007)
3. Alivisatos, P. (2004) The use of nanocrystals Compact cysteine-coated CdSe(ZnCdS) quan-
in biological detection. Nat. Biotechnol. 22, tum dots for in vivo applications. J. Am. Chem.
47–52. Soc. 129, 14530–14531.
4. Kim, S., Lim, Y. T., Soltesz, E. G., De Grand, 7. McBride, J., Treadway, J., Feldman, L. C.,
A. M., Lee, J., Nakayama, A., et al. (2004) Pennycook, S. J., and Rosenthal, S. J. (2006)
Near-infrared fluorescent type II quantum Structural basis for near unity quantum yield
dots for sentinel lymph node mapping. Nat. core/shell nanostructures. Nano Lett. 6,
Biotechnol. 22, 93–97. 1496–1501.
8. Nirmal, M., Dabbousi, B. O., Bawendi, M. 18. Yao, J., Larson, D. R., Vishwasrao, H. D.,
G., Macklin, J. J., Trautman, J. K., Harris, T. D., Zipfel, W. R., and Webb, W. W. (2005) Blinking
et al. (1996) Fluorescence intermittency in and nonradiant dark fraction of water-soluble
single cadmium selenide nanocrystals. Nature quantum dots in aqueous solution. Proc. Natl
383, 802–804. Acad. Sci. U.S.A. 102, 14284–14289.
9. Scholl, B., Liu, H. Y., Long, B. R., McCarty, 19. Zhang, L., Neves, L., Lundeen, J. S., and
O. J., O’Hare, T., Druker, B. J., et al. (2009) Walmsley, I. A. (2009) A characterization of
Single particle quantum dot imaging achieves the single-photon sensitivity of an electron
ultrasensitive detection capabilities for Western multiplying charge-coupled device. J. Phys. B
immunoblot analysis. ACS Nano 3, At. Mol. Opt. Phys. 42, 114011.
1318–1328. 20. Michalet, X., Fabien, P., Thilo, D. L., Maxime,
10. Temirov, J. P., Bradbury, A. R., and Werner, J. D., Marcel, P. B., Alivisatos, A. P., et al. (2001)
H. (2008) Measuring an antibody affinity dis- Properties of fluorescent semiconductor
tribution molecule by molecule. Anal. Chem. nanocrystals and their application to biologi-
80, 8642–8648. cal labeling. Single Mol. 2, 261–276.
11. Dahan, M., Levi, S., Luccardini, C., Rostaing, 21. Crocker, J. and Grier, D. (1996) Methods of
P., Riveau, B., and Triller, A. (2003) Diffusion digital video microscopy for colloidal studies.
dynamics of glycine receptors revealed by sin- J. Colloid Interface Sci. 179, 298–310.
gle-quantum dot tracking. Science 302, 22. Rife, J. C., Long, J. P., Wilkinson, J., and
442–445. Whitman, L. J. (2009) Particle tracking single
12. Cui, B., Wu, C., Chen, L., Ramirez, A., protein-functionalized quantum dot diffusion
Bearer, E. L., Li, W.-P., et al. (2007) One at a and binding at silica surfaces. Langmuir 25,
time, live tracking of NGF axonal transport 3509–3518.
using quantum dots. Proc. Natl Acad. Sci. 23. Saxton, M. J. and Jacobson, K. (1997) Single-
USA. 104, 13666–13671. particle tracking: applications to membrane
13. Rajan, S. S., Liu, H. Y., and Vu, T. Q. (2008) dynamics. Annu. Rev. Biophys. Biomol. Struct.
Ligand-bound quantum dot probes for study- 26, 373–399.
ing the molecular scale dynamics of receptor 24. Saxton, M. J. (2008) A biological interpreta-
endocytic trafficking in live cells. ACS Nano tion of transient anomalous subdiffusion. II.
2, 1153–1166. Reaction kinetics. Biophys. J. 94, 760–771.
14. Courty, S., Luccardini, C., Bellaiche, Y., 25. Martin, D. S., Forstner, M. B., and Kas, J. A.
Cappello, G., and Dahan, M. (2006) Tracking (2002) Apparent subdiffusion inherent to
individual kinesin motors in living cells using single particle tracking. Biophys. J. 83,
single quantum-dot imaging. Nano Lett. 6, 2109–2117.
1491–1495. 26. Liu, W., Howarth, M., Greytak, A. B., Zheng,
15. Bouzigues, C., Morel, M., Triller, A., and Y., Nocera, D. G., Ting, A. Y., et al. (2008)
Dahan, M. (2007) Asymmetric redistribution Compact biocompatible quantum dots func-
of GABA receptors during GABA gradient tionalized for cellular imaging. J. Am. Chem.
sensing by nerve growth cones analyzed by Soc. 130, 1274–1284.
single quantum dot imaging. Proc. Natl Acad. 27. Warnement, M. R., Faley, S. L., Wikswo, J. P.,
Sci. USA. 104, 11251–11256. and Rosenthal, S. J. (2006) Quantum dot
16. Tada, H., Higuchi, H., Wanatabe, T. M., and probes for monitoring dynamic cellular
Ohuchi, N. (2007) In vivo real-time tracking response: reporters of T cell activation. IEEE
of single quantum dots conjugated with Trans. Nanobioscience 5, 268–272.
monoclonal anti-HER2 antibody in tumors of 28. Warnement, M. R., Tomlinson, I. D., Chang,
mice. Cancer Res. 67, 1138–1144. J. C., Schreuder, M. A., Luckabaugh, C. M.,
17. Lang, E., Baier, J., and Köhler, J. (2006) and Rosenthal, S. J. (2008) Controlling the
Epifluorescence, confocal and total internal reactivity of ampiphilic quantum dots in bio-
reflection microscopy for single-molecule exper- logical assays through hydrophobic assembly
iments: a quantitative comparison. J. Microsc. of custom PEG derivatives. Bioconjug. Chem.
222, 118–123. 19, 1404–1413.
Chapter 5
Titanium Dioxide Nanoparticles in Advanced

Imaging and Nanotherapeutics
Tijana Rajh, Nada M. Dimitrijevic, and Elena A. Rozhkova
Abstract
Semiconductor photocatalysis using nanoparticulate TiO2 has proven to be a promising technology for
use in catalytic reactions, in the cleanup of water contaminated with hazardous industrial by-products,
and in nanocrystalline solar cells as a photoactive material. Metal oxide semiconductor colloids are of
considerable interest because of their photocatalytic properties. The coordination sphere of the surface
metal atoms is incomplete and thus traps light-induced charges, but also exhibits high affinity for oxygen-
containing ligands and gives the opportunity for chemical modification. We use enediol linkers, such as
dopamine and its analogs, to bridge the semiconductors to biomolecules such as DNA or proteins.
Nanobio hybrids that combine the physical robustness and chemical reactivity of nanoscale metal oxides
with the molecular recognition and selectivity of biomolecules were developed. Control of chemical
processes within living cells was achieved using TiO2 nanocomposites in order to develop new tools for
advanced nanotherapeutics. Here, we describe general experimental approaches for synthesis and charac-
terization of high crystallinity, water soluble 5 nm TiO2 particles and their nanobio composites, methods
of cellular sample preparation for advanced Synchrotron-based imaging of nanoparticles in single cell
X-ray fluorescence, and a detailed experimental setup for application of the high-performance TiO2-based
nanobio photocatalyst for targeted lysis of cancerous or other disordered cells.
Key words: TiO2, Nanoparticles, Surface reconstruction, Photocatalysis, Charge transfer complex,
Hybrid composites, DNA, Antibody, Targeted cancer therapy, Synchrotron X-ray fluorescence
1. Introduction
The impact of nanoscience and nanotechnology on cell manipula-

tion and actuation is critically dependent on the creation of new
classes of functionally and physically integrated hybrid materials
that incorporate nanoparticles and biologically active molecules.
These hybrid bioinorganic composites integrate both inorganic
materials and biological entities via multivalent lock-and-key
63
64 Rajh, Dimitrijevic, and Rozhkova
interactions, offering opportunities to impact diverse applications

ranging from quantum computation, energy transduction and
site-selective catalysis to advanced nanomedicine. Hybrid materi-
als that combine collective properties of crystalline materials and
localized properties of biomolecules are of special interest because
they present basic functional units capable of carrying out site-
selective redox chemistry within intracellular machinery.
In order to develop new tools for biomedicine and biotech-
nology that carry out redox chemistry, we focus on these special
classes of functional nanomaterials that mimic the exquisite con-
trol over energy and electron transfer that occurs in natural energy
transducing processes (1). Nanobio hybrids that combine the
physical robustness and chemical reactivity of nanoscale metal
oxides with the molecular recognition and selectivity of biomol-
ecules were developed. DNA and monoclonal antibodies were
utilized to direct TiO2 nanoparticles to the specific cells with tar-
get molecules. Photo-induced charge separation was then
employed to control and manipulate processes within the cells
and to alter their functioning. These hybrid TiO2–DNA nano-
composites form a basis for the development of novel artificial
restriction enzymes and single nucleotide polymorphism (SNP)
genotyping with high specificity in vitro or in vivo (2, 3).
Recently, we developed polychromatic visible-light inducible
nanobio hybrid systems based on 5 nm TiO2 nanocrystals cova-
lently tethered to a biological vehicle (4) capable of selective rec-
ognition of cancer-associated antigens (see Figs. 1–3). Similar to
“classic” photodynamic therapy (PDT), our approach includes
three main components: light, oxygen, and a photoreactive mate-
rial. The hybrid semiconductor particles absorb energy from light,
which is then transferred to molecular oxygen, producing cyto-
toxic reactive oxygen species (ROS) (4, 5). The advantages of
nanoscale photosensitization compared to “classical” PDT are
the result of a synergistic combination of the advanced physical
properties of inorganic materials and the targeting abilities of
biomolecules and the multiple functions of drugs and imaging
payloads in one ideal therapeutic system. Furthermore, nanopar-
ticles may overcome biological barriers.
Approaches for tethering biomolecules to the surface of TiO2
particles utilize the ability of oxygen-containing functional groups,
such as carboxy-, hydroxyl-, and phosphate, to bind to the surface
of nanoparticles. Our strategy to construct bio-TiO2 hybrids is
based on dihydroxybenzenes, for example, dopamine (DA) or its
natural metabolite 3,4-dihydroxyphenylacetic acid (DOPAC), as
linkers (see Fig. 1) (6–8). As a result of the presence of two OH–
groups in the ortho position, the catecholate group forms a strong
bidentate complex with the coordinatively unsaturated Ti atoms
at the surface of nanoparticles (7). Furthermore, it has been
Titanium Dioxide Nanoparticles in Advanced Imaging and Nanotherapeutics 65
Bulk Nanoparticle Hybrid nanoparticle
1 2 3 4 5 6 7
d = 8 nm
100 nm
200 nm
Fig. 1. TiO2 nanoscale materials. Top: Surface-modified 45-Å TiO2 nanoparticles with different bidentate ligands: (1) bare
TiO2, (2) salicylic acid, (3) dihydroxycy-clobutenedione, (4) vitamin C, (5) alizarin, (6) dopamine, and (7) tert-butyl catechol.
Reprinted with permission from ref. 6. Bottom: TEM images of a library of TiO2 nanoscale materials. Modified with per-
mission from Dimitrijevic, N. M., Saponjic, Z. V., Rabatic, B. M., Poluektov, O. G., and Rajh, T. (2007) Effect of size and
shape of nanocrystalline TiO2 on photogenerated charges. An EPR study. J. Phys. Chem. C. 111, 14597–14601.
shown that when DNA or proteins are covalently bound to DA,

DA acts as a conductive bridge between the biomolecules and
TiO2 nanocrystals, allowing transport of photogenerated holes to
the biomolecules (9, 10). Chemisorption of DA or DOPAC
serves two important purposes: first, it “heals” the semiconductor
surface, enabling absorption in a visible part of solar spectrum
(see Fig. 1), and, second, DA or DOPAC modify the particle sur-
face with amino- or carboxylic functional groups, which are useful
for further covalent tethering to a biomolecule (see Fig. 2).
Readily available water-soluble reagents for carbodiimide cou-
pling allow tethering of biological molecules to the particle sur-
face with retained photo-physical and biological recognition
properties in the resulting hybrid system.
bare 5 nm TiO2 particles:

bandgap: 3.2 eV,lmax <380 nm
DOPAC
SulfoNHS
EDAC
DOPAC-coated TiO2 particle:
bandgap: 1.6 eV,
lmax > 450 nm (up to 750 nm)
D.
Ab
TiO2 75
37
TiO2
TiO2 DNA TiO2 1 2 3 4 5
Transmittance, Arbitrary Units

DNA Ab 1540 Amide II
C.
1638 Amide I
TiO2 Water
TiO2
5 nm IL13
5 nm IL13-TiO2
4000 3500 3000 2500 2000 1500 1000
Wavelength (cm-1)
Fig. 2. Synthesis of TiO2 nanobio hybrids. Top: General scheme. Bottom left and middle: Examples of TiO2-biomolecule
hybrids (AFM images). Bottom right: Complete conversion of a free antibody in the course of the carbodiimide coupling
to the TiO2–DOPAC is confirmed by SDS-PAGE (denaturizing conditions) analysis of the final conjugate (4), reaction mix-
ture supernatant (2) and all washing solutions (3). (1) and (5) are free antibody and Precision Plus Protein Standards
(BioRad), respectively. FTIR spectrum of the conjugate (blue line) contains amide bands typical of immunoglobulins.
Modified with permission from ref. 4.
The ability to manipulate molecules and materials with nano-

meter resolution is crucial in the field of nanotechnology. In order
to understand and control complex behavior of nanoscale materi-
als within biological system, such as cellular machinery, advanced
imaging techniques are required. Thus, interactions of 5 nm TiO2
nanoparticles functionalized with DNA or an antibody with cel-
lular compartments (nucleus and mitochondria) or membrane
proteins were studied by X-ray fluorescence microscopy (XFM)
using the Advanced Photon Source (4, 11). Third-generation
synchrotrons with spatially coherent high-brilliance X-rays allow
elemental mapping of biological specimens in near-native envi-
ronments with submicrometer to dozens of nanometers spatial
resolution, which provides valuable complementary information
to light microscopy. An exposure of a sample to hard X-rays results
in the emission of characteristic “secondary” (or fluorescent) X-rays
or distinctive “signatures” of each element comprising the sample.
Fig. 3. General concept of the TiO2-based photocatalytic cell lysis. Left: Nanobiocomposites consisted of 5 nm TiO2 and
IL13R-recognizing antibody linked via DOPAC linker to recognize and bind exclusively to surface IL13R. Visible light
photo-excitation of the nanobio hybrid in an aqueous solution results in the formation of various ROS. ROS, mainly super-
oxide, cause cell membrane damage, permeability changes, and cell death. Right: Phototoxicity of the TiO2–mAb toward
A172 glioblastoma cells (top) and same cytotoxicity in the presence of various ROS quenchers (bottom). Isotype-matched
negative control antibody immunoglobulin IgG1, either conjugated or unconjugated, did not recognize isolated or cellular
IL13a2R and did not show photo-induced toxicity. Modified with permission from ref. 4.
XFM allows imaging of tiny nanoparticles composed of nonbiogenic

elements including titanium in single cell and, therefore, these
can be used to label and map cellular structures of interest (see
Fig. 4). Furthermore, this method makes possible imaging of the
redistribution of cellular endogenous trace elements (e.g., iron,
manganese, calcium) as response to external stimulus, for exam-
ple, intracellular redox elements relocalization during photo-
induced apoptosis.
Semiconductor TiO2 is well known as a photocatalyst in the
degradation of organic substrates and the deactivation of micro-
organisms and even viruses ((4) and references therein). Under
ultraviolet light (UV) excitation, TiO2 nanoparticles of various
sizes, morphologies, and solubility have been reported to exhibit
cytotoxicity toward some tumor cells. Owing to the significantly
a 10µm 10µm 10µm
Cell Culture
b 20µm 20µm 20µm
c 50µm 50µm 50µm

Confocal Microscopy
Beam Max: 91.5 Max: 3.2 Max: 0.27

P Min: 0
mg/cm2
Ti Min: 0
mg/cm2
Zn Min:
mg/cm
0
2
Max: 109 Max: 9.2 Max: 0.19

Min: 0 Min: 0 Min: 0
20mm
Optical Images,  20 Max: 0.91 Max: 0.022 Max: 0.161

Min: 0 Min: 0 Min: 0
min concentration max
Fig. 4. Advanced X-ray imaging of the TiO2 nanobio hybrid within a single cell. Top left: Cells are grown in Petri
dishes with attached carbon/formvar-coated gold-finder grids. Bottom left: Optical micrograph of cells attached to the
grid. Top right: Laser confocal microscopy images of cell undergoing various stages of photo-induced apoptosis. Control
cells with TiO2–mAb, but no light applied (a), and after 30 min (b) or 90 min (c) following light exposure. Right bottom:
X-ray fluorescence imaging of the TiO2–mAb binding to the single glioblastoma cell (representative images of the high
antigen overexpressing A172 line). Elemental distribution of biogenic phosphorus and zinc are used to sketch cells and
nucleus. The intensity of the elemental images was displayed using a prism color table in logarithm scale, which is shown
in the bottom right. The maximum and minimum threshold values in micrograms per square centimeter are given above
each frame. Scans were obtained by using 10.0-keV incident energy with dwell times of 1 s per pixel and 1-mm steps
through the sample. Modified with permission from ref. 4.
improved photoreactivity the catecholate-modified TiO2

nanoparticles represent promising materials for nanotherapeutics
as they can be induced by the visible part of the solar spectrum
closely approaching the optimal spectral window for biological
tissue penetration of 800 nm (9, 12). We recently demonstrated
successful applicability of TiO2–DA (DOPAC)-based nanobio
composites in targeted photo-induced lysis of brain cancer or
pathogenic T-cells associated with psoriasis (4, 13).
Here, we describe general experimental approaches for syn-
thesis and characterization of high crystallinity, water soluble
5 nm TiO2 (14) particles and their nanobio composites for

functional integration with biomolecules (4, 14). Methods of
cellular sample preparation for advanced Synchrotron-based
imaging of nanoparticles in single cell X-ray fluorescence are also
described (4, 15, 16). Finally, a detailed experimental setup for
application of the high-performance TiO2-based nanobio photo-
catalyst for targeted lysis of cancerous or other disordered cells is
presented (4, 13).
2. Materials
All chemicals of the highest grade available were obtained from

Sigma-Aldrich and used without further purification. Milli-Q
water was used in all experiments (see Note 1).
2.1. Cell Culture The human malignant glioma cell lines A172MG and U87MG
(American Type Culture Collection, Manassas, VA) and normal
human astrocytes (Cambrex-Clonetics, East Rutherford, NJ) are
routinely grown in Dulbecco’s Modified Eagle’s Medium
(DMEM) with 4.5 g/L glucose and l-glutamine, supplemented
with 10% fetal bovine serum (FBS; Mediatech, Herndon, VA) in
a humidified atmosphere with 5% CO2 at 37°C.
2.2. Synthesis of High 1. Milli-Q water, out-gas with argon or nitrogen to remove
Crystallinity “Bare” oxygen.
5 nm TiO2 Particles 2. Titanium tetrachloride.
3. 2,000 MW Cutoff dialysis cassettes (Pierce, Rockford, IL).
4. 0.2 M LiOH.
5. Isopropyl glycidyl ether (also called 1,2-epoxy-3-
isopropoxypropane).
2.3. TiO2-Biomolecule 1. 10 mM Sodium phosphate buffer (PBS), pH = 6.2.

Synthesis and 2. Monoclonal anti-human-IL13Ralpha2 or Mouse IgG1
Characterization Isotype Control (R&D Systems, Minneapolis, MN).
Antibodies (500 mg) are reconstituted in 50 mL sterile PBS
pH = 7.4 right before conjugation to the pre-activated nano-
particles to final concentration 10 mg/mL.
3. 11 mM DOPAC in 10 mM PBS, pH = 6.2.
4. 44 mM N-Hydroxysulfosuccinimide (Sulfo-NHS) (Pierce,
Rockford, IL) in 10 mM PBS, pH = 6.2.
5. 50 mM 1-Ethyl-3-[3-dimethylaminopropyl]carbodiimide
hydrochloride (EDAC) (Sigma-Aldrich, St. Louis, MO) in
10 mM PBS, pH = 6.2.
6. 2-Mercaptoethanol.
7. 2,000 MW Cutoff dialysis cassettes (Pierce, Rockford, IL).

8. 10 mM PBS, pH = 7.4.
9. 25 mM Glycine.
10. 4–20% Tris–HCl precast gel and prestained molecular weight
markers “Kaleidoscope markers” (Bio-Rad, Hercules, CA).
2.4. Photo-Induced 1. 10 mM PBS, pH = 7.4.

Cell Lysis 2. Petri dishes (12-well cell culture cluster, tissue culture treated,
Costar ®).
3. Cermax® PE300BFM 300 W Xenon Lamp, UV filter (Perkin
Elmer Optoelectronics). An IR filter was custom-made in the
Argonne glass blowing shop, but any cylinder of sufficient
size filled with water can be used.
4. Standard LDH test: the CytoTox-ONE™ Homogeneous
Membrane Integrity Assay (Promega, Madison, WI).
5. Stock solutions of radical scavengers: 10,000 U/mL catalase,
1,000 U/mL superoxide dismutase (SOD), 5 M mannitol,
200 mM NaN3, and 1 M histidine (His).
2.5. Cell Sample 1. TEM 100-mesh, carbon/formvar-coated gold-finder grids

Preparation for X-Ray (Electron Microscopy Sciences, PA).
Fluorescence 2. Glass bottom 35 × 10 mm clear wall cell culture dishes
(PELCO ®).
3. 4% Paraformaldehyde solution in 10 mM phosphate-buffered
saline, pH = 7.2 (Affymetrix/USB). The pH point is not well
buffered, and tricky to reach. If necessary, adjust solution pH
with concentrated HCl or with 1 M NaOH. To get 1% para-
formaldehyde, dilute 1 mL of the 4% paraformaldehyde to
4 mL with 10 mM PBS.
4. 0.5% Dodecyltrimethylammonium chloride (DOTMAC)/1%
paraformaldehyde. Dilute 4 mL 4% paraformaldehyde and
0.5 mL 10% DOTMAC in 5.5 mL 10 mM PBS, pH = 7.4.
5. 1% BSA in 10 mM PBS, pH = 7.4.
6. 10 mM PBS, pH = 7.4.
7. 20 mM Piperazine-N,N¢-bis(2-ethanesulfonic acid)
(PIPES)/200 mM sucrose buffer. Prepare about 50 mL of
this buffer by dissolving PIPES and sucrose in Milli-Q-water.
Adjust pH to 7 only by adding acetic acid. Do not use any
other base or acid.
2.6. Laser Confocal 1. Glass bottom 35 × 10-mm clear wall cell culture dishes
Imaging (PELCO ®).
2. MitoTracker Red CMXRos (Molecular Probes, Invitrogen).
3. Zeiss LSM 510 Meta Confocal Microscope.
3. Methods
(See Note 2)
3.1. Synthesis of High 1. Add 5.0 mL of prechilled (~4°C) titanium tetrachloride drop-
Crystallinity Bare 5 nm wise to 200 mL ice water under vigorous stirring until the white
TiO2 Particles fog disappears. Attention! This step requires chemical fume hood.
2. Dialyze this solution against 2 L Milli-Q water using
2,000 MW cutoff dialysis cassettes at 4°C for 3 days, chang-
ing water daily, allowing slow growth of the particles, until
the pH of the colloid reaches ~3.5 (see Note 3).
3. Dilute the colloid ~20 times (final colloid concentration
~0.015 M) and add 100 mL isopropyl glycidyl ether.
4. Inject 1 mL LiOH and, mix vigorously to adjust pH to
~9–12.
5. Dialyze this solution against 2 L against a buffer with desired
pH values (e.g., for carboxylic group activation an optimal
buffer with pH = 6.3).
6. Keep colloidal solutions of particles at 4°C for at least
6 months before use for perfecting crystallinity by aging.
3.2. TiO2-Biomolecules 1. Mix 320 mL 11 mM TiO2 particles with 16 mL 11 mM DOPAC
Synthesis and (final concentration 550 mM) in 10 mM phosphate buffer
Characterization pH = 6.2, to reach a final particle/DOPAC ratio of 1/100.
The slightly yellow TiO2–DOPAC complex immediately
formed was monitored by observing the characteristic ligand-
to-particle charge-transfer (LPCT) band at 420 nm (shoul-
der) as a result of chemisorption of DOPAC molecules on the
TiO2 surface. Optical absorption spectra were recorded using
“Nanodrop ND-1000” or Perkin Elmer Lambda 950 UV/
Vis and LS55 spectrometers.
2. Mix 6 mL 44 mM Sulfo-NHS and 4 mL 50 mM EDAC in the
same buffer with the TiO2–DOPAC. Incubate the reaction
mixture for 1 h at room temperature with continuous gentle
shaking in the dark, and then add 1 mL 2-mercaptoethanol to
quench excess EDAC.
3. Place the reaction mixture into 2,000 MW cutoff dialysis bags
(each 1 mL in size) and dialyze against 1 L 10 mM phosphate
buffer, pH = 7.4 for 1 h.
4. Mix 500 mg of an antibody (mAb) (see Note 4) in 50 mL
10 mM phosphate buffer, pH = 7.4 with the pre-activated
TiO2–DOPAC particles (TiO2–DOPAC–N-hydroxysulfosucc
inimide ester) and incubate for 4–6 h at room temperature with
continuous gentle mixing in the dark.
5. Add to the reaction mixture 25 mL 25 mM glycine and incu-
bate for 15 min to quench the remaining active sites on the
particle surface. Remaining unquenched active sites were

reported to reduce an antibody’s biorecognition activity.
6. Dialyze the reaction mixture for 6 h using the same dialysis
system.
7. Spin-wash the final TiO2–mAb conjugate four times with
100 mL 10 mM phosphate buffer, pH = 7.4, to remove any
unbound protein. Resuspend the final slightly yellowish pellet
in 300 mL of the same buffer. The resulting colloidal suspen-
sion is stable at 4°C for up to 1 month.
8. Verify complete conversion of the free mAb by comparative
SDS/2-mercaptoethanol-denaturation polyacrylamide gel
electrophoresis of the final TiO2–mAb conjugate and all spin/
washing solutions (see Fig. 2). AFM TiO2–mAb imaging
revealed an average TiO2 particle/mAb ratio as ~1.5/1 (see
Fig. 2). The FTIR spectrum of the TiO2–mAb conjugate
contained new bands at 1638 (amide I) and 1540 (amide II)
cm−1 typical for immunoglobulins (see Fig. 2).
3.3. Photo-Induced 1. Grow cells in 12-well plates to reach 105 cells per well and
Cell Lysis and Cell then wash them three times with PBS (see Note 5).
Viability Examination 2. Add aliquots of 6–600 ng/mL TiO2–mAb or the antibody-
free TiO2 particles to cells and incubate for 1 h at the same
culture condition (see Note 6). After 1 h, wash the cells thor-
oughly (six times) with PBS to eliminate any unbound
nanomaterial.
3. Illuminate the plates for 5 min by focused polychromatic
visible light with Cermax® PE300BFM 300 W Xenon Lamp. A
UV filter (Orion Lighting Systems) should be used to cut wave-
lengths below 380 nm off. The intensity of incident light focused
on the Petri dish (D = 3.5 cm) reached 60 mW/cm2 as measured
by a power energy meter (Scientech 372, Boulder, CO). A water
filter (D 3.5 × H 20 cm) is used to remove infrared radiation and
exclude hyperthermia cytotoxic effects (see Note 7).
4. After light treatment, leave the cells to recover in a humidi-
fied atmosphere with 5% CO2 at 37°C. Assess cell viability in
6, 24, and 48 h by standard LDH testing.
5. Perform control experiments to verify the origin of photo-
induced cytotoxicity in cultures exposed to the TiO2–mAb.
Control experiments include cell cultures exposed to (1)
TiO2–mAb but not illuminated (negative control), (2) free
TiO2 particles with illumination (negative control), (3)
TiO2–IgG1 isotype antibody conjugates with illumination
(negative control to demonstrate specific binding of the
TiO2–mAb nanoparticles to cancer cells), and (4) focused
light without nanoparticles (a background control to esti-
mate nanoparticles-driven phototoxicity). ROS-scavengers
of H2O2 (final concentration: 100 U/mL catalase), SOD
(10 U/mL), hydroxyl radical (50 mM mannitol), and singlet

oxygen (2 mM NaN3 or 10 mM His) are used in control
experiments for the identification of ROS involved in the
cytotoxicity. An example of the results produced is shown in
Fig. 3.
3.4. Cell Samples 1. Culture cells at the standard conditions in growth medium in
Preparation for X-Ray glass bottom, 35 × 10 mm, clear wall cell culture dishes
Fluorescence (PELCO ®) with 100-mesh, carbon/formvar-coated gold-
Elemental Analysis finder grids (Electron Microscopy Sciences, PA) attached to
their bottom until a sufficient amount of cells are attained on
the grid (~10 K per grid) (see Note 8).
2. Simultaneously fix and permeabilize cells by incubation in
0.5% DOTMAC/1% paraformaldehyde in 10 mM PBS for
5 min (see Note 9).
3. Add 1% paraformaldehyde in 10 mM PBS, incubate cells for
20 min.
4. Block the cells for 90 min in 1% BSA to reduce nonspecific
interaction with the antibody.
5. Add 600 ng/mL TiO2–mAb conjugate and incubate for 1 h.
6. Add dyes at this stage to stain cellular compartments, if
desired.
7. Wash specimens six times with sterile PBS followed by the
removal of residual PBS by several washes in 20-mM pipes,
pH 7.2–200 mM sucrose (see Note 10). Gently remove
excess liquid using Kimwipes, air-dry.
8. Capture optical images of the cells using optical microscope
(e.g., Zeiss LSM 510 Meta Confocal Microscope) with 40×
or 20× objectives before X-ray analysis.
9. Keep samples dry in a desiccator for use in up to 3–4 months.
An example of sample preparation and the results produced
are shown in Fig. 4.
4. Notes
1. All solutions should be prepared in water that has a resistivity

of 18.2 MW cm and total organic content of less than five parts
per billion. This standard is referred to as “water” in this text.
2. Since nanoscale materials can exhibit properties different
from bulk materials with the same chemical composition and
not all their properties and health effects are yet known,
nanoscale materials must be considered of unknown toxicity.
Therefore, as all new compounds, or those of unknown toxic-
ity, nanoscale materials should be considered as materials both
acutely and chronically toxic. Habitual safe practices should be

directed to minimizing of exposure by avoiding skin contact
and inhalation through proper clothing and ventilation (17).
3. If allowed to dry out, the solution gradually turns into a trans-
parent soft gel, which continues to change with time, from
translucent to slight white, shrinking to split in the end. In step
2, the pH of the colloid solution should not reach higher than
3.5– 4 to avoid the TiO2 isoelectric point at pH = 5.0. Rapid
injection of LiOH then allows the adjustment of the pH to
10–12, avoiding isoelectric point and TiO2 precipitation.
4. This protocol can be adapted for other biomolecules, includ-
ing DNA and PNA functionalized with an amino-group. For
tethering biomolecules containing carboxylic groups
TiO2–DA should be used instead of TiO2–DOPAC.
5. This protocol can be adapted for many other cell culture lines.
6. Cells can also be incubated with nanoparticles or nanobio
conjugates at lower temperatures of 4°C to avoid possible
internalization of particles.
7. In all experiments, the cell culture solution temperature could
be remotely monitored with an infrared camera (e.g., model:
ICI 7320 with 0.038 K temperature sensitivity, Infrared
Cameras Inc. or similar). When water filter is applied, the
temperature variations should not be more than 1–2°C during
the light exposure. This excludes hyperthermia as a possible
mechanism of cell damage.
8. The carbon/formvar-coated gold-finder grids are gently
attached to a bottom of a dish using scotch-tape. Blunt-ended
tweezers are very convenient for handling very fragile grids.
Glass window area should be avoided as it will be used for
laser confocal imaging.
9. This method of cell fixation and preparation can be applied for
any advanced microcopy, including X-ray fluorescence. This
technique minimizes disruption of plasma membrane micro-
structures and cellular compartments as well as metal ion topol-
ogy and it does not significantly alter typical cellular trace element
content relative to fixation by plunge freezing (4, 15, 16).
10. For a final washing any buffers and solutions containing
sodium, potassium, chloride, etc. must be avoided.
Acknowledgments
Work at the Center for Nanoscale Materials was supported by the

US Department of Energy, Office of Science, Office of Basic
Energy Sciences, under Contract No. DE-AC02-06CH11357.
We are thankful to our colleagues Drs. B. Lai, L. Finney, S. Vogt

and J. Maser from Advanced Photon Source, Argonne National
Laboratory and collaborators Drs. M. S. Lesniak and I. V. Ulasov
from the University of Chicago, Brain Tumor Center.
References
1. Thurnauer, M. C., Dimitrijevic, N. M., nanocrystalline-DNA interface: probing DNA

Poluektov, O. G., and Rajh, T. (2004) recognition. Nano Lett. 4, 1017–1023.
Photoinitiated charge separation: from photo- 10. Dimitrijevic, N. M., Saponjic, Z. V., Rabatic,
synthesis to nanoparticles. Spectrum 17, B. M., and Rajh, T. (2005) Assembly and
10–15. charge transfer in hybrid TiO2 architectures
2. Liu, J., de la Garza, L., Zhang, L., Dimitrijevic, using biotin-avidin as a connector. J. Am.
N. M., Zuo, X. B., Tiede, D. M., et al. (2007) Chem. Soc. 127, 1344–1345.
Photocatalytic probing of DNA sequence by 11. Paunesku, T., Rajh, T., Wiederrecht, G.,
using TiO2/dopamine-DNA triads. Chem. Maser, J., Vogt, S., Stojicevic, N., et al. (2003)
Phys. 339, 154–163. Biology of TiO2-oligonucleotide nanocom-
3. Liu, J., Saponjic, Z. V., Dimitrijevic, N. M., posites. Nat. Mater. 2, 343–346.
Luo, S., Preuss, D., and Rajh, T. (2006) 12. de la Garza, L., Saponjic, Z. V., Dimitrijevic,
Hybrid TiO2 nanoparticles: an approach for N. M., Thurnauer, M. C., and Rajh, T. (2006)
developing artificial restriction enzymes. in: Surface states of titanium dioxide nanoparti-
Colloidal Quantum Dots for Biomedical cles modified with enediol ligands. J. Phys.
Applications (Osinski, M., Yamamoto, K., and Chem. B 110, 680–686.
Jovin, T. M. eds.). Bellingham, WA. SPIE Proc 13. Rajh, T., Dimitrijevic, N. M., Elhofy, A., and
Series 6096, 60960F. Rozhkova, E. A. (2010) Hybrid TiO2 based
4. Rozhkova, E. A., Ulasov, I. V., Lai, B., nanocomposites-new tools for biotechnology
Dimitrijevic, N. M., Lesniak, M. S., and Rajh, and biomedicine. in: Handbook of Nanophysics:
T. (2009) A high-performance nanobio pho- Functional Nanomaterials (Sattler, K. D. ed.).
tocatalyst for targeted brain cancer therapy. CRC, Boca Raton, FL.
Nano Lett. 9, 3337–3342. 14. Rajh, T., Ostafin, A. E., Micic, O. I., Tiede,
5. Dimitrijevic, N. M., Rozhkova, E. A., and D. M., and Thurnauer, M. C. (1996) Surface
Rajh, T. (2009) Dynamics of localized charges modification of small particle TiO2 colloids
in dopamine-modified TiO2 and their effect with cysteine for enhanced photochemical
on the formation of reactive oxygen species. reduction: an EPR study. J. Phys. Chem. 100,
J. Am. Chem. Soc. 131, 2893–2899. 4538–4545.
6. Rajh, T., Chen, L. X., Lukas, K., Liu, T., 15. Nakamura, F. (2001) Biochemical, electron
Thurnauer, M. C., and Tiede, D. M. (2002) microscopic and immunohistological observa-
Surface restructuring of nanoparticles: an effi- tions of cationic detergent-extracted cells:
cient route for ligand-metal oxide crosstalk. detection and improved preservation of
J. Phys. Chem. B 106, 10543–10552. microextensions and ultramicroextensions.
7. Redfern, P. C., Zapol, P., Curtiss, L. A., Rajh, BMC Cell Biol. 2, 10.
T., and Thurnauer, M. C. (2003) 16. Finney, L., Mandava, S., Ursos, L., Zhang,
Computational studies of catechol and water W., Rodi, D., Vogt, S., et al. (2007) X-ray
interactions with titanium oxide nanoparti- fluorescence microscopy reveals large-scale
cles. J. Phys. Chem. B 107, 11419–11427. relocalization and extracellular translocation
8. Saponjic, Z. V., Dimitrijevic, N. M., Tiede, D. of cellular copper during angiogenesis. Proc.
M., Goshe, A. J., Zuo, X., Chen, L. X., et al. Natl. Acad. Sci. USA. 104, 2247–2252.
(2005) Shaping nanometer-scale architecture 17. Committee on Prudent Practices for Handling,
through surface chemistry. Adv. Mater. 17, Storage, and Disposal of Chemicals in
965–971. Laboratories, National Research Council.
9. Rajh, T., Saponjic, Z., Liu, J., Dimitrijevic, (1995) Prudent Practices in the Laboratory
N. M., Scherer, N. F., Vega-Arroyo, M., Handling and Disposal of Chemicals. National
et al. (2004) Charge transfer across the Academy Press, Washington, DC.
Chapter 6
Surface Modification and Biomolecule Immobilization

on Polymer Spheres for Biosensing Applications
Chris R. Taitt, Lisa C. Shriver-Lake, George P. Anderson,
and Frances S. Ligler
Abstract
Microspheres and nanospheres are being used in many of today’s biosensing applications for automated
sample processing, flow cytometry, signal amplification in microarrays, and labeling in multiplexed analyses.
The surfaces of the spheres/particles need to be modified with proteins and other biomolecules to be
used in these sensing applications. This chapter contains protocols to modify carboxyl- and amine-coated
polymer spheres with proteins and peptides.
Key words: Microspheres, Nanoparticles, Immobilization, Latex spheres
1. Introduction
While encapsulation of enzymes in hydrophilic microspheres has

long been the standard procedure for industrial bioprocessing
(e.g., food, pharmaceuticals, and cosmetics) and a variety of
commercial products are available (1), microspheres and nano-
spheres with recognition molecules on the surface have only
become widely utilized over the last decade. These structures are
being used in analytical and biosensing, drug delivery, and affinity
purification applications. As integral components of analytical
methods, recognition molecules are currently being immobilized
on magnetic spheres for target separation (2–6), on fluorescent
microspheres for use in flow cytometric analysis (7–11), on latex
spheres for dynamic light scatter-based assays (12–14), and on
gold colloids for the generation of colored signals (15, 16).
Fluorescent or metallic spheres have been used for signal amplifi-
cation in optical and electrochemical assays (17–20).
77
78 Taitt et al.
The methods used for immobilization vary, depending on the

composition of the recognition molecule and the chemistry of
the bead surface (bead will be used interchangeably with micro-
sphere and/or nanoparticle throughout this chapter). In general,
there are three basic approaches for immobilization of molecules
on spheres, depending on the chemistry of the bead surface. If a
metal is exposed on the surface, organic molecules containing a
binding moiety are preferred (i.e., a thiol group to bind to a gold
surface or a histidine oligomer to bind to exposed nickel (16, 21, 22)).
If the surface is glass or silicon, organosilanes are bound to
activated silanol groups on the surface and serve as anchors for
cross-linkers and/or biomolecules (23, 24). Finally, if the spheres
are composed of a synthetic polymer such as polystyrene or
polymethylmethacrylate (PMMA) [7, 25–28), liquid crystals (29),
or natural biopolymers such as chitosan or collagen (30, 31),
chemical reactions compatible with both the complex surface and
the molecule being immobilized must be used.
Thus, a variety of approaches have been developed to attach
biomolecules to polymer spheres of nanometer (nm) and micron
diameter. In general, the goals are (1) to provide a bond that is
stable, (2) to prevent secondary adsorption of the biomolecule to
the bead with attendant denaturation, (3) to expose the active site
of the biomolecule to the environment away from the bead
surface, and (4) to immobilize the biomolecule at high density.
The primary choices for achieving a stable bond are to
covalently attach the biomolecule to the bead directly or through
an intermediate layer or to attach it through a high affinity
interaction such as avidin–biotin linkage or oligonucleotide
hybridization. The trapping of biomolecules onto microspheres/
nanospheres using electrostatic layers such as polyethyleneimine
and polystyrene has also been documented (32–34), but this
method may inhibit interactions between some recognition
molecules and their target ligands, especially if the targets are large.
Secondary adsorption is not a problem if the polymer spheres
are composed of a hydrophilic polymer, as is the case with natural
biopolymers, but can be a problem if they are composed of hydro-
phobic polymers, such as PMMA, nylon, and polystyrene (latex).
In the latter case, hydrophilic molecules such as protein, dextran,
or polyethylene glycol (PEG) can be linked to the surface of the
bead, and the recognition molecule can be linked to this interme-
diate layer. In addition to preventing the denaturation of the
attached recognition molecule, these hydrophilic layers also help
prevent nonspecific adsorption of nontarget molecules from
complex samples (35–37). When the recognition molecules are
relatively small (e.g., antimicrobial peptides, haptens, or oligosac-
charides), a spacer can be placed between them and the bead
surface (or the intermediate layer) to improve the access of the
Surface Modification and Biomolecule Immobilization on Polymer Spheres 79
active site to its target in the surrounding fluid. In many cases, an

intermediate layer that prevents nonspecific adsorption can also
serve as the spacer.
Finally, in most assay procedures, it is helpful to immobilize
the recognition molecules at relatively high density in order to
maximize target binding (38). If direct binding or binding
through a spacer or intermediate layer does not provide sufficient
density, molecular brushes or dendrimers can be used to increase
the number and density of attachment sites on the surface of the
spheres (39–41). In this chapter, we use the term “scaffold” to
refer to the intermediate layer between the bead and the recog-
nition molecules that either prevents nonspecific adsorption or
increases the density of reactive groups at the bead surface.
Several companies (Luminex Corporation (42), Spherotech (43),
Bangs Lab (44), Invitrogen (45), and Sigma-Aldrich (46)) sell
microspheres and nanospheres with reactive groups on the surface
as well as fluorescent dyes or ferromagnetic material in the bead
core. The established protocols provided by the manufacturers
are usually for direct immobilization of large molecules exhibiting
amine groups (see Fig. 1a, analogous to the manufacturers’
protocols). We have found that this is the simplest and most direct
method to bind a protein to a bead functionalized with a carboxyl
group. For large proteins, there are sufficient lysines with terminal
amino groups that the majority of molecules will be oriented so
that sufficient active sites are available to bind the targets. This
protocol can also be used for the attachment of avidin and subse-
quent noncovalent attachment of a biotinylated recognition
molecule (29). Biotinylation may provide better control over
the orientation of the molecule than direct immobilization,
particularly for small molecules. Moreover, a long-chain biotin
(e.g., EZ-Link NHS-LC-LC-biotin, EZ-Link maleimide-PEG11-
biotin, or EZ-Link biotin-PEO-LC-amine) (47) can add a short
spacer to move the recognition molecule further from the avidin
binding pocket and into the surrounding fluid. Low molecular
weight haptens have been attached to large proteins or dextran
and then subsequently attached to the carboxyl groups on the
bead surface through the protein using this scheme (48, 49).
A convenient variation on this same basic chemistry can be
performed in which the carboxyl groups on the microsphere/
nanoparticle surface are modified by the attachment of diamine
scaffolds to create a surface layer of amino groups (see Fig. 1b).
Recognition molecules can then be attached through the
carboxyl groups rather than through terminal amino groups.
In some cases, this reverses the molecular orientation at the
surface to provide better exposure for the active sites. For
recognition molecules of less than 1,000 molecular weight,
attaching the appropriate end to the bead is frequently critical
to its activity (50, 51).
80 Taitt et al.
Fig. 1. Three schemes (a–c) for generating biomolecule-coated microspheres/nanoparticles.
We also show that an attachment protocol for spheres that

have amine groups on the surface rather than carboxyl groups
(see Fig. 1c). This protocol can also be employed after carboxy-
lated spheres have been modified with a protein [such as bovine
serum albumin (BSA)], dextran, or a dendrimer (52) to create a
scaffold of amino groups on the surface. Once reactive carboxyl
groups are attached to the amino groups, the same procedures
are employed as in Fig. 1a.
These three protocols, described in detail below, are specified

for spheres with micron diameters (microspheres) and a centrifu-
gation step is used to separate the spheres from the reagents/
solutions. These protocols work equally well with nm-diameter
spheres. However, for smaller spheres (nanospheres), bead sepa-
ration can be accomplished with dialysis, desalting gel-filtration
chromatography, centrifugation in Centricon columns, or simple
quenching of the reaction. For example, the activity of sulfo-
N-hydroxysuccinimide can be quenched by the addition of
2-mercaptoethanol or terminated by desalting (53).
2. Materials
2.1. Attachment 1. Microspheres: Luminex’s xMAP-compatible or other micro-

of Amine-Containing spheres can be purchased from the following vendors:
Molecules to COOH- MiraiBio, Spherotech, Invitrogen, and Sigma-Aldrich.
Microspheres 2. Activation buffer: 0.1 M sodium phosphate buffer, pH = 6.
( See Fig. 1a) 3. EDC solution: 50 mg/mL 1-ethyl-3-[3-dimethylaminopro-
pyl]carbodiimide hydrochloride (EDC), in dimethyl sulfoxide
(DMSO). This solution must be used within 2 h of prepara-
tion (see Note 1).
4. Sulfo-NHS Solution: 50 mg/mL N-hydroxysulfosuccinimide
(sulfo-NHS) in deionized H2O. Prepare the same quantity of
solution as used for EDC solution, above.
5. Biomolecule solution (e.g., protein, peptide, ethylenediamine,
and aminodextran): 10 mg/mL to 1 mg/mL (depending on
the MW and availability of the amine-containing biomolecule)
in phosphate-buffered saline (PBS), pH = 7.4 (see Note 2).
Generally, a large excess of the biomolecule relative to the
bead assures the attachment of a higher density of molecules,
but lower amounts may also be effective.
6. PBS, pH = 7.2–7.4.
7. PBST: PBS supplemented with 0.05% Tween-20.
8. PBSTB: PBS supplemented with 0.05% Tween-20 and 1 mg/
mL BSA.
2.2. Attaching 1. Dialysis tubing: Dialysis tubing with an appropriate molecular

COOH-Containing weight cutoff can be obtained from many manufacturers
Molecules to NH2- and should be fully rehydrated before use per manufac-
Microspheres turer’s instructions. Use of dialysis clips (versus tying the
(“Upside-Down EDC tubing) greatly simplifies the protocol and allows the user to
Reaction”) ( See perform EDC/NHS activation of proteins within the same
Fig. 1b) dialysis bag.
2. Biomolecule solution: 10 mg/mL to 1 mg/mL carboxyl-
containing protein/biomolecule in PBS, pH = 7.4. Higher or
82 Taitt et al.
lower concentrations of biomolecule can be used as needed.

Total protein solution volume should be 1 mL or less.
3. Activation buffer: 0.1 M phosphate buffer, pH = 6.
4. EDC solution: 50 mg/mL EDC in DMSO. This solution
must be used within 2 h of preparation (see Note 1).
5. Sulfo-NHS solution: 50 mg/mL sulfo-NHS in deionized
H2O. Prepare the same quantity as for the EDC solution.
6. 100× 2-Mercaptoethanol: 35 mL 2-mercaptoethanol in 965 mL
deionized water (0.5 M 2-mercaptoethanol) (see Note 3).
7. Amine-decorated microspheres: Microspheres treated with
an appropriate di- or multivalent amine-containing molecule
(e.g., ethylenediamine and aminodextran) as described in
Subheading 3.1 or ones commercially available (43).
8. NHS coupling buffer: 100 mM sodium borate, pH = 8.0.
Other basic or neutral buffers may work as well or better
depending on the nature of the ligand.
9. PBS, pH = 7.2.
10. PBST: PBS supplemented with 0.05% Tween-20.
11. PBSTB: PBS supplemented with 0.05% Tween-20 and 1 mg/
mL BSA.
12. Small molecule possessing carboxyl moiety, e.g., fumonisin,
aflatoxin, or peptide.
2.3. Converting 1. Amine-decorated microspheres: From step 10 of Sub

NH2-Microspheres heading 3.1.3, coated using aminodextran or ethylenediamine.
into COOH- 2. Conversion buffer: 100 mM sodium borate, pH = 9.0.
Microspheres
3. Succinic anhydride.
( See Fig. 1c)
4. PBS, pH = 7.2.
3. Methods
3.1. Attachment Attachment of amine-containing molecules to carboxyl-decorated

of Amine-Containing microspheres is simple and straightforward when using an EDC/
Molecules (Including NHS linkage. The presence of NHS stabilizes the amine-reactive
Scaffolds) to COOH- intermediate (O-acylisourea) created by the reaction of EDC with
Microspheres carboxyls by converting the intermediate to an amine-reactive
(See Fig. 1a) sulfo-NHS ester. We have adapted the protocol used by Luminex
and other commercial sources for microspheres (42–46) to
simplify the physical manipulations and to minimize the loss of
microspheres during preparation. Further, this protocol can be
used to convert carboxyl-decorated microspheres into amine-
decorated microspheres (to be used in further coupling reactions)
if the biomolecule being attached possesses multiple amine

moieties, including ethylene diamine (H2N–CH2–CH2–NH2),
polyoxyethylene bis(amine) (H2N–[–CH2–CH2–O–]n–CH2–
CH2–NH2), or aminodextran. Likewise surface carboxyls can be
converted to thiols using cystamine (HN2–CH2–CH2–S–S–CH2–
CH2–NH2) with subsequent reduction of the disulfide by a
reducing agent such as tris(2-carboxyethyl) phosphine (TCEP).
3.1.1. Microsphere In this step, microspheres are washed and exchanged into an
Preparation: Washing appropriate buffer for EDC chemistry.
1. Resuspend microspheres in stock bottle by vortexing and
brief bath sonication.
2. Pipet 100 mL of beads into Eppendorf tube (~1.5 × 106 micro-
spheres; scale as desired).
3. Centrifuge for 4 min at 18,400 × g or 14,000 rpm in an
Eppendorf centrifuge.
4. Remove the supernatant and put it into a secondary tube
(referred to as the “chase tube”). The use of this chase tube
minimizes the loss of microspheres during handling.
5. Resuspend microspheres in 100 mL Activation buffer in the
primary tube.
6. Repeat centrifugation of both the primary and chase tubes.
7. Discard supernatant from the chase tube, move the superna-
tant from the primary tube to the chase tube.
8. Resuspend microspheres in 100 mL Activation buffer in the
primary tube.
9. Repeat centrifugation of the primary and chase tubes.
10. Discard supernatant from the chase tube, move the superna-
tant from the primary tube to the chase tube.
11. Repeat centrifugation of the chase tube.
12. Discard the supernatant from the chase tube.
13. Resuspend microspheres in the chase tube in 40 mL Activation
buffer.
14. Resuspend microspheres in the primary tube in 80 mL
Activation buffer.
15. Combine the solution in the chase tube with that in the
primary tube.
3.1.2. Microsphere In this section, the protocol describes the activation of the
Activation with EDC/ carboxyl groups on the microspheres with EDC/sulfo-NHS
Sulfo-NHS to form an NHS-ester. This intermediate is more stable than
the O-acylisourea intermediate that is formed in the absence
of NHS.
84 Taitt et al.
1. Add 15 mL EDC solution and 15 mL sulfo-NHS solution to

the microspheres in the primary tube. The final concentra-
tions of each will be ~5 mg/mL (25 mM).
2. Mix well by briefly vortexing and bath sonicating. Keep in the
dark for 20 min.
3. Centrifuge for 4 min at 18,400 × g.
4. Remove the supernatant from the primary tube and pipette
into the chase tube.
5. Resuspend microspheres in the primary tube in 100 mL
Activation buffer.
7. Discard supernatant from the chase tube and transfer super-
natant from the primary tube to the chase tube.
8. Resuspend the microspheres in the primary tube in 100 mL
Activation buffer. Repeat centrifugation of the primary and
chase tubes.
9. Discard supernatant from the chase tube and transfer super-
natant from the primary tube to chase tube.
10. Resuspend the microspheres in the primary tube in 100 mL PBS.
12. Discard the supernatant from the chase tube.
13. Resuspend microspheres in the chase tube with PBS (~50 mL)
and vortex/sonicate.
14. Transfer the contents of the chase tube to the primary tube
and vortex/sonicate to mix.
3.1.3. Cross-linking In this step, the NHS-ester active group on the bead binds to
of Activated Microspheres amine-containing biomolecules in solution. The protocol
to Amine-Containing described is appropriate for creating a single type of “decorated”
Biomolecules/Linkers microsphere. Alternatively, the resuspended microspheres can be
or Scaffolds aliquoted and each aliquot is incubated with a different biomol-
ecule to create batches of microspheres functionalized with
different biomolecules. For immobilization of proteins, we typically
use 100 mg (~1 mg/mL) protein for 1.5 × 106 microspheres. Less
can be used, but using an excess of the biomolecule at a high
concentration helps to ensure good coupling efficiency. To
calculate the minimum amount of protein to use, determine the
surface area of the microspheres to be coated and the footprint
occupied by each protein; for most couplings, a tenfold excess of
protein or more should be used. At the end of the reaction,
the biomolecule-decorated microspheres are resuspended and
washed in an appropriate buffer. The composition of this final
buffer will depend on whether or not additional coupling
reactions are desired.
1. Add 100 mL biomolecule solution to the resuspended

microspheres (see Subheading 3.1.2, step 15).
2. Vortex and bath-sonicate the microspheres to ensure they are
well suspended.
3. Incubate for at least 2 h in the dark. Rotate the tube to keep
the microspheres suspended or alternatively, mix frequently
(every 10–15 min). Alternatively, an overnight incubation
can be performed.
5. Remove supernatant and discard.
6. Resuspend microspheres in 100 mL (a) PBST or PBSTB if no
additional couplings are desired, (b) Activation buffer if
additional EDC-mediated coupling is desired, and (c) other
type of coupling buffers as required for additional coupling
reactions.
8. Remove supernatant and discard.
9. Wash microspheres once by resuspending in 100 mL appro-
priate buffer (see step 6) and centrifuging for 4 min at
18,400 × g.
10. Resuspend microspheres in 100 mL buffer and proceed with
additional coupling reactions. Alternatively, store micro-
spheres at 4°C in the dark for several days to several years or
more, depending on the stability of the immobilized biomol-
ecule. Larger storage volumes may be desirable depending on
the application.
3.2. Attaching COOH- Two protocols are described for the attachment of protein
Containing Molecules (see Subheading 3.2.1) or small molecules (see Subheading 3.2.2)
to NH2-Microspheres to microspheres possessing pendant amine moieties. These
(“Upside-Down” EDC amine-decorated microspheres can be prepared by treating
Reaction) (See Fig. 1b) carboxyl-decorated microspheres (as supplied by the manufacturer)
with a di- or multivalent amine-containing “scaffold.” The protocols
below are essentially the inverse reaction from that detailed in
Subheading 3.1 – essentially an “upside-down” EDC reaction.
3.2.1. Attachment Direct attachment to activated microspheres requires that the

of Proteins to NH2- protein solution not contain any other amine/carboxyl reactive
Decorated Microspheres groups (e.g., glycine, Tris, and BSA). For protein solutions
supplied in an amine/carboxyl-containing buffer, the original
buffer must be removed prior to coupling. For this reason, a
dialysis step is included to exchange the buffer and remove any
interfering components. If the protein is supplied or can be
satisfactorily diluted into the activation buffer it may be possible
to proceed to step 3.
86 Taitt et al.
1. Place protein solution (~1 mg/mL) into a dialysis membrane

having a suitable MW cutoff for the protein.
2. Dialyze against >100 times the sample volume of 10%
Activation buffer for ~1 h. If the protein has previously been
in an amine-containing buffer, it will be necessary to change
the buffer at least three times (after 1 h, after 3 h, and after
overnight incubation) (see Note 2).
3. Add 1/50 volume each of EDC and sulfo-NHS solutions to
the protein solution in the dialysis bag. The final concentra-
tion of each will be 1 mg/mL (5 mM).
4. Dialyze for 1 h against fresh 10% Activation buffer.
5. After 1 h, add 1/100 volume 100× 2-mercaptoethanol
solution to the protein solution; the final concentration of
mercaptoethanol in the sample will be 5 mM (see Note 4) to
stop the reaction.
6. During protein dialysis, prepare the microspheres either pur-
chased as NH2-microspheres or prepared using step 10 of
Subheading 3.1.3. Wash the amine-decorated microspheres
twice by resuspending in 100 mL NHS coupling buffer and
centrifuging for 4 min at 18,400 × g.
7. After quenching of the EDC reaction (see step 5) or removing
the unreacted EDC (see Note 4), add an equal volume of
NH2-decorated microspheres to the target protein.
8. Incubate the microspheres with the protein for at least 2 h.
9. Centrifuge the mixture for 4 min at 18,400 × g.
10. Discard the supernatant and wash the protein-coated micro-
spheres twice by resuspension in 100 mL PBST or PBSTB and
centrifugation for 4 min at 18,400 × g.
11. Store the protein-coated microspheres at 4°C in the dark.
3.2.2. Homogeneous Many small molecules such as peptides, phycotoxins, and myco-
Protocol for Small toxins possess carboxyl moieties. If these molecules do not
Carboxyl-Containing possess disulfide bridges and are not in the presence of amine-
Molecules based buffers, they can be attached to amine-coated microspheres
without dialysis.
1. Dissolve or dilute the molecule of interest into an appropriate
volume of Activation buffer (typically, 0.1–1 mg/mL in
£1 mL).
2. Add 1/50th volume each of EDC and sulfo-NHS solutions.
The final concentration of each will be 1 mg/mL (5 mM).
3. Incubate for 1 h at room temperature.
4. During EDC activation, prepare the amine-decorated micro-
spheres. Wash the amine-decorated microspheres twice by
resuspension in 100 mL NHS coupling buffer and centrifugation

for 4 min at 18,400 × g. Resuspend the microspheres in NHS
coupling buffer.
5. Add 1/100th volume of 100× 2-mercaptoethanol solution to
the EDC/sulfo-NHS reaction (see Note 3). The final concen-
tration of mercaptoethanol will be 5 mM.
6. Mix the resuspended microspheres with the EDC-activated
small molecule.
7. Incubate the microspheres with the activated small molecule
for at least 2 h at room temperature.
8. Centrifuge the mixture for 4 min at 18,400 × g.
9. Discard the supernatant and wash the coated microspheres
twice by resuspension in 100 mL PBST or PBSTB and
centrifugation for 4 min at 18,400 × g each time.
10. Store the small molecule-coated microspheres at 4°C in the dark.
3.3. Making Protein– In some cases, the user may desire to create a multivalent conju-
Small Molecule gate comprising a protein scaffold and many attached small
Conjugates molecules before its attachment to microspheres. These conju-
gates may be useful because the conjugate orients the small
molecules correctly, provides a higher avidity surface for subse-
quent immunoassays, or presents the antigen in a highly immuno-
genic form for antibody generation. There are a number of
methods for conjugating small molecules with proteins, particu-
larly with regard to creating immunogens (49). In our laboratory,
we have coupled mycotoxins and explosives to proteins, such as
BSA or keyhole limpet hemocyanin (KLH). We have also
employed the EDC chemistry described above with small mole-
cules, using the protein in place of the microspheres. Once these
small molecule–protein conjugates have been formed, the protein
can be attached to the microspheres through an amine as described
in the protocols above. It is often also convenient to couple a
protein, such as ovalbumin, BSA, or hen egg lysozyme, to the
microsphere as the intermediate layer and then link the small
molecule to the immobilized protein using EDC chemistry
(see Subheading 3.1). In that case, the use of chase tubes is not
typically necessary.
3.4. Converting Sometimes a user desires a higher density of carboxyl moieties

NH2-Microspheres than is provided on commercially available microspheres, or alter-
into COOH- natively, the attached moiety may be unstable if directly immobi-
Microspheres lized to the microsphere surface using EDC-based, zero-length
(See Fig. 1c) attachment chemistry. In these cases, the user may attach an inter-
vening layer of ethylenediamine (to add a short linker) or a mul-
tivalent scaffold such as aminodextran (to increase the density of
88 Taitt et al.
available sites for immobilization) using EDC chemistry (see

Subheading 3.1). The resulting amine-decorated surfaces can
then be converted back into carboxylated surfaces using the pro-
tocol below. Once converted, the standard EDC/sulfo-NHS
protocol described in Subheading 3.1 can be used to attach an
amine-containing biomolecule to the diamine or multivalent
scaffold. This can easily be accomplished by the use of succinic
anhydride.
1. Remove approximately 50–100 mL amine-decorated micro-
spheres from step 10, Subheading 3.1.3.
2. Centrifuge microspheres for 4 min at 18,400 × g and discard
supernatant.
3. Resuspend microspheres in 0.5 mL Conversion buffer and
centrifuge again.
4. Resuspend microspheres in 0.5 mL Conversion buffer.
5. Add ~1 mg solid succinic anhydride directly to the resus-
pended microspheres.
6. Sonicate and vortex the sample until all the succinic anhy-
dride is dissolved.
7. Incubate for 30 min with intermittent mixing.
8. Centrifuge the microspheres for 4 min at 18,400 × g and
discard supernatant.
9. Wash once by resuspending in 0.5 mL Conversion buffer and
centrifuging.
10. Repeat steps 4–8 twice more for a total of three treatments
with succinic anhydride.
11. Wash once by resuspending in 0.5 mL PBS or desired buffer
and centrifuging.
12. Finally, resuspend the microspheres in the desired buffer: PBS if
storing or Activation buffer if further coupling chemistry is going
to be performed (see Subheading 2.2). The carboxy-coated
microspheres can then be exposed to proteins/biomolecules
using the protocols described above (see Subheading 3.1).
3.5. Experimental Many immunoassays have been developed for the detection of
Results small antigens including mycotoxins, phycotoxins, and explosives
that possess either a single epitopic site or several overlapping
epitopes. For example, while ochratoxin possesses a single carboxyl
moiety and must be immobilized using that group, fumonisin
possesses both amine and carboxyl groups and can be immobi-
lized in multiple orientations (see Fig. 2). In these assays, appro-
priate presentation of the small molecule is critical for optimal
assay performance. Furthermore, when a competitive format is
utilized, the density of the immobilized species is also important.
If the density of immobilized species is too high, assay sensitivity
o o
o o
O
C OH N o o o
O OH O
N O
H CH3 o o o
Cl
o o
o o
15,000
median fluorescence
10,000
5,000
Fum
onisin
A Ochr B
BS e atoxin
ym an ) A
oz xtr ine
lys od
e m
in t (a
am ec
dir
Fig. 2. Top: Chemical structures of ochratoxin A (left ) and fumonisin B1 (right ). Gray arrows point to carboxyl moieties that
can be linked to amine groups through EDC chemistry. The black arrow points to the unique primary amine on fumonisin
B1; this amine can be linked to carboxyl moieties using EDC chemistry. Lower panel: Median fluorescence of labeled
anti-fumonisin monoclonal antibodies (10 mg/mL, gray bars) and anti-ochratoxin antibodies (10 mg/mL, white bars)
bound to microspheres coated with fumonisin or ochratoxin, respectively, that were immobilized through amine-rich
scaffolds (BSA, lysozyme, and aminodextran) or directly.
will be poor; if too low, the signals generated will be too low to
produce a robust assay.
In order to develop sensitive and robust assays utilizing
immobilized small molecules, microspheres were decorated with
immunogens or capture reagents using different methods, and
their performance in immunoassays was used to compare the
presentation and functionality of the immobilized species. Each
mycotoxin (ochratoxin and fumonisin) was immobilized onto
samples of microspheres decorated with three different types of
amine-rich scaffolds (i.e., BSA, lysozyme, and aminodextran).
First, the method outlined in Subheading 3.1 was used to attach
the amine-rich scaffolds to carboxy-functionalized microspheres,
then the method outlined in Subheading 3.2 was used to attach
90 Taitt et al.
the carboxyl groups of the mycotoxins to these scaffolds. Also,

fumonisin B1 was coupled directly to the carboxy-functionalized
microspheres via its single amine group. Anti-fumonisin antibod-
ies bound poorly to microspheres on which fumonisin had been
immobilized by its single primary amine, indicating that that
amine may be in a critical part of the epitopic region (see Fig. 2).
Interestingly, however, anti-fumonisin and anti-ochratoxin anti-
bodies bound very well to microspheres on which the respective
mycotoxins were immobilized through lysozyme or aminodex-
tran scaffolds, although poor binding was observed when they
were immobilized through BSA. Since the molar proportion of
lysines (through which the mycotoxins were attached) is roughly
equivalent with BSA and lysozyme, and only half as high in the
aminodextran used here, the relative affinities of the antibodies
for aminodextran- or lysozyme-immobilized mycotoxins is not a
simple density issue.
Small molecules have also been used as affinity capture
reagents in assays designed to complement immunoassays; such
capture biomolecules include aptamers (54–57), carbohydrates
and glycolipids (58–64), and antimicrobial peptides (50, 65–67).
Here too, both the presentation and density of immobilized cap-
ture molecules are critical for assay performance. We immobilized
several antimicrobial peptides (polymyxin B, melittin, cecropin A,
and magainin II) as biological recognition elements for Escherichia
coli; these molecules have previously been used to detect E. coli
and other Gram-positive and Gram-negative bacteria in biosensor
assays (see Fig. 3). Each peptide was immobilized onto micro-
spheres coated with various scaffolds or linkers. Scaffolds offering
the possibility of higher density linkages included BSA, lysozyme,
aminodextran, and poly-d-lysine. After immobilization onto the
microspheres, pendant amines on these scaffolds were converted to
carboxyl moieties using succinic anhydride (see Subheading 3.4),
and subsequently linked to the peptides using standard EDC
chemistry (see Subheading 3.1.2). Additional linkers including
ethylenediamine and PEG diamine were also immobilized using
standard EDC chemistry, converted to carboxyl-terminated linkers,
and finally linked to the peptides’ amines via standard EDC
chemistry. The final linker tested was a glycine tetrapeptide;
immobilization of antimicrobial peptides using this (gly)4 linker
was accomplished through two rounds of direct EDC coupling.
In general, E. coli exhibited the highest binding to melittin-
coated microspheres independent of the scaffold or linker used
with most effective immobilization on poly-d-lysine scaffolds
(see Fig. 3). Cecropin A also exhibited optimal binding when
immobilized through poly-d-lysine. However, we observed that
E. coli bound best to polymyxin B when aminodextran was used as
scaffold. Overall, no clear patterns of preferential immobilization
were observed among the different peptides, indicating the impor-
tance of optimizing the attachment chemistry for each system.
10,000
8,000 Polymyxin B
Melittin
Cecropin A
Median Fluorescence
Magainin-II
6,000
No peptide
4,000
2,000
0
BSA lysozyme aminodextran poly-D-lysine PEG-diamine ethylene (gly)4
diamine
Scaffold or Linker
Fig. 3. Escherichia coli assays using antimicrobial peptides as capture reagents. Biotinylated E. coli was incubated for
30 min with beads coated with various antimicrobial peptides (polymyxin B, melittin, cecropin A, and magainin II), washed,
and then interrogated with streptavidin–phycoerythrin conjugate. Shown are median fluorescence values for each bead
set. The antimicrobial peptides were immobilized onto beads indirectly using amine-rich scaffolds (BSA, lysozyme, amin-
odextran, and poly-d-lysine), amine-terminated linkers (PEG diamine and ethylenediamine), or a tetrapeptide possessing
both a single carboxyl and single amine [(gly)4]. Nonspecific binding of E. coli to the scaffold/tether material is shown as
white bars (“no peptide”).
4. Notes
1. EDC is highly unstable in the presence of water. It should be

stored in a dessicator at −20°C. The dessicator and vial inside
should be allowed to warm to room temperature before
opening to avoid water condensation.
2. This solution must have no amine-containing components
other than the biomolecule itself; amine-containing buffers
such as Tris and glycine should not be used. If Tris or another
unwanted amine is present, it will be necessary to dialyze or
remove by gel filtration prior to reacting with the micro-
spheres or poor coupling would result.
3. 2-Mercaptoethanol is toxic. Further, it has a distinctive
unpleasant odor and causes irritation to nasal passageways
and respiratory tract upon inhalation. For these reasons, it
should be used in a chemical fume hood and unused/waste
solutions should be disposed of as hazardous waste.
92 Taitt et al.
4. 2-Mercaptoethanol forms a stable complex with unreacted

EDC and quenches the activation reaction. However,
mercaptoethanol may also inactivate certain proteins. If
inactivation of the protein is problematic, activated protein
can be purified away from unreacted EDC using a Centricon
(Millipore) or gel filtration, provided that these methods are
performed as rapidly as possible.
Acknowledgments
This work was supported by the Defense Threat Reduction

Agency and ONR/NRL 6.2 work unit 6336. The views are those
of the authors and do not represent opinion or policy of the Naval
Research Laboratory, US Navy, or Department of Defense.
References
1. Brady, D. and Jordaan, J. (2009) Advances in 9. Muldrew, K. L. (2009) Molecular diagnostics
enzyme immobilisation. Biotechnol. Lett. 31, of infectious diseases. Curr. Opin. Pediatr. 21,
1639–1650. 102–111.
2. Bi, S., Zhou, H., and Zhang, S. S. (2009) 10. Tait, B. D., Hudson, F., Cantwell, L., Brewin,
Bio-bar-code functionalized magnetic nano- G., Holdsworth, R., Bennett, G., et al. (2009)
particle label for ultrasensitive flow injection Review article: Luminex technology for HLA
chemiluminescence detection of DNA hybrid- antibody detection in organ transplantation.
ization. Chem. Commun. 5567–5569. Nephrology 14, 247–254.
3. Liu, C. X., Stakenborg, T., Peeters, S., and 11. Vignali, D. A. A. (2000) Multiplexed particle-
Lagae, L. (2009) Cell manipulation with mag- based flow cytometric assays. J. Immunol.
netic particles toward microfluidic cytometry. Methods 243, 243–255.
J. Appl. Phys. 105, 102014-1–102014-11. 12. Du, B. A., Li, Z. P., and Cheng, Y. Q. (2008)
4. Ugelstad, J., Stenstad, P., Kilaas, L., Prestvik, Homogeneous immunoassay based on aggre-
W. S., Herje, R., Berge, A., et al. (1993) gation of antibody-functionalized gold nano-
Monodisperse magnetic polymer particles – particles coupled with light scattering detection.
New biochemical and biomedical applications. Talanta 75, 959–964.
Blood Purif. 11, 349–369. 13. Liu, X., Dai, Q., Austin, L., Coutts, J.,
5. Yang, H., Li, H. P., and Jiang, X. P. (2008) Knowles, G., Zou, J. H., et al. (2008) A one-
Detection of foodborne pathogens using step homogeneous immunoassay for cancer
bioconjugated nanomaterials. Microfluid. biomarker detection using gold nanoparticle
Nanofluid. 5, 571–583. probes coupled with dynamic light scattering.
6. Safarik, I. and Safarikova, M. (2009) Magnetic J. Am. Chem. Soc. 130, 2780–2782.
nano- and microparticles in biotechnology. 14. Sanchez-Martinez, M. L., Aguilar-Caballos, M. P.,
Chem. Pap. 63, 497–505. and Gomez-Hens, A. (2009) Homogeneous
7. Dunbar, S. A. (2006) Applications of immunoassay for soy protein determination in
Luminex (R) xMAP (TM) technology for food samples using gold nanoparticles as labels
rapid, high-throughput multiplexed nucleic and light scattering detection. Anal. Chim.
acid detection. Clin. Chim. Acta 363, Acta 636, 58–62.
71–82. 15. Thaxton, C. S., Rosi, N. L., and Mirkin, C. A.
8. Lehmann, S., Dupuy, A., Beaudeux, J. L., (2005) Optically and chemically encoded
and Lizard, G. (2009) Multiplexed analysis nanoparticle materials for DNA and protein
for identification and evaluation of novel detection. MRS Bull. 30, 376–380.
biomarkers in biological fluids, tissue and 16. Wilson, R. (2008) The use of gold nanopar-
cell extracts. Ann. Biol. Clin. (Paris) 67, ticles in diagnostics and detection. Chem. Soc.
381–393. Rev. 37, 2028–2045.
17. Cha, B. H., Lee, S. M., Park, J. C., Hwang, K. S., the preparation of micro- and nanoparticulates
Kim, S. K., Lee, Y. S., et al. (2009) Detection of biomaterials from electrostatic and ultra-
of hepatitis B virus (HBV) DNA at femtomolar sonic methods. Biomed. Eng. Appl. Basis
concentrations using a silica nanoparticle- Commun. 21, 343–353.
enhanced microcantilever sensor. Biosens. 31. Nobs, L., Buchegger, F., Gurny, R., and
Bioelectron. 25, 130–135. Allemann, E. (2004) Current methods for
18. Seydack, M. (2005) Nanoparticle labels in attaching targeting ligands to liposomes and
immunosensing using optical detection meth- nanoparticles. J. Pharm. Sci. 93, 1980–1992.
ods. Biosens. Bioelectron. 20, 2454–2469. 32. Al-Saadi, A., Yu, C. H., Khurtoryanskiy, V. V.,
19. Wang, J. L., Munir, A., Li, Z. H., and Zhou, Shih, S. J., Crossley, A., and Tsang, S. C.
H. S. (2009) Aptamer-Au NP conjugates- (2009) Layer-by-layer electrostatic entrap-
enhanced SPR sensing for the ultrasensitive ment of protein molecules on superparamag-
sandwich immunoassay. Biosens. Bioelectron. netic nanoparticle: A new strategy to enhance
25, 124–129. adsorption capacity and maintain biological
20. Wu, Y. F., Liu, S. Q., and He, L. (2009) activity. J. Phys. Chem. C 113, 15260–15265.
Electrochemical biosensing using amplification- 33. Kato, N. and Caruso, F. (2005) Homogeneous,
by-polymerization. Anal. Chem. 81, competitive fluorescence quenching immuno-
7015–7021. assay based on gold nanoparticle/polyelectro-
21. Kuramitz, H. (2009) Magnetic microbead- lyte coated latex particles. J. Phys. Chem. B
based electrochemical immunoassays. Anal. 109, 19604–19612.
Bioanal. Chem. 394, 61–69. 34. Subramani, C., Ofir, Y., Patra, D., Jordan, B.
22. Nagaoka, T., Shiigi, H., and Tokonami, S. J., Moran, I. W., Park, M. H., et al. (2009)
(2007) Highly sensitive and selective chemical Nanoimprinted polyethyleneimine: A multi-
sensing techniques using gold nanoparticle modal template for nanoparticle assembly and
assemblies and superstructures. Bunseki immobilization. Adv. Funct. Mater. 19,
Kagaku 56, 201–211. 2937–2942.
35. Choi, S. W., Kim, W. S., and Kim, J. H. (2003)
23. Weetall, H. H. (1993) Preparation of immo-
Surface modification of functional nanoparti-
bilized proteins covalently coupled through
cles for controlled drug delivery. J. Dispersion
silane coupling agents to inorganic supports.
Sci. Technol. 24, 475–487.
Appl. Biochem. Biotechnol. 41, 157–188.
36. Rossi, S., Lorenzo-Ferreira, C., Battistoni, J.,
24. Marquette, C. A., Cretich, M., Blum, L. J., and
Elaissari, A., Pichot, C., and Delair, T. (2004)
Chiari, M. (2007) Protein microarrays enhanced
Polymer mediated peptide immobilization onto
performance using nanobeads arraying and
amino-containing N-isopropylacrylamide-
polymer coating. Talanta 71, 1312–1318.
styrene core-shell particles. Colloid Polym. Sci.
25. Chen, Z. J., Peng, K., and Mi, Y. L. (2007) 282, 215–222.
Preparation and properties of magnetic poly- 37. Sheng, Y., Liu, C. S., Yuan, Y., Tao, X. Y.,
styrene microspheres. J. Appl. Polym. Sci. 103, Yang, F., Shan, X. Q., et al. (2009) Long-
3660–3666. circulating polymeric nanoparticles bearing a
26. Yalcin, G., Elmas, B., Tuncel, M., and Tuncel, combinatorial coating of PEG and water-solu-
A. (2006) A low, particle-sized, nonporous ble chitosan. Biomaterials 30, 2340–2348.
support for enzyme immobilization: Uniform 38. Iannone, M. A. and Consler, T. G. (2006)
poly(glycidyl methacrylate) latex particles. Effect of microsphere binding site density on
J. Appl. Polym. Sci. 101, 818–824. the apparent affinity of an interaction partner.
27. Bake, K. D. and Walt, D. R. (2008) Cytometry A 69A, 374–383.
Multiplexed spectroscopic detections. Annu. 39. Mouaziz, H., Braconnot, S., Ginot, F., and
Rev. Anal. Chem. 1, 515–547. Elaissari, A. (2009) Elaboration of hydrophilic
28. Kang, K., Kan, C., Yeung, A., and Liu, D. aminodextran containing submicron magnetic
(2006) The immobilization of trypsin on latex particles. Colloid Polym. Sci. 287,
soap-free P(MMA-EA-AA) latex particles. 287–297.
Mater. Sci. Eng. C 26, 664–669. 40. Shen, Y., Kuang, M., Shen, Z., Nieberle, J.,
29. Spillmann, C. M., Naciri, J., Anderson, G. P., Duan, H. W., and Frey, H. (2008) Gold nano-
Chen, M. S., and Ratna, B. R. (2009) Spectral particles coated with a thermosensitive hyper-
tuning of organic nanocolloids by controlled branched polyelectrolyte: Towards smart
molecular interactions. ACS Nano 3, temperature and pH nanosensors. Angew.
3214–3220. Chem. Int. Ed. Engl. 47, 2227–2230.
30. Kuo, S. M., Chang, S. J., Yao, C. H., and 41. Shi, X. Y., Thomas, T. P., Myc, L. A.,
Manousakas, I. (2009) A perspective view on Kotlyar, A., and Baker, J. R. (2007) Synthesis,
94 Taitt et al.
characterization, and intracellular uptake of element: The case of HIV-1 Tat protein.
carboxyl-terminated poly(amidoamine) den- Biosens. Bioelectron. 20, 1149–1156.
drimer-stabilized iron oxide nanoparticles. 57. Cheng, A. K. H., Sen, D., and Yu, H. Z.
Phys. Chem. Chem. Phys. 9, 5712–5720. (2009) Design and testing of aptamer-based
42. Luminex Corporation. (2009) Microspheres/ electrochemical biosensors for proteins and
Reagents. small molecules. Bioelectrochemistry 77, 1–12.
43. Spherotech Inc. (2009) Microspheres. 58. Spangler, B. D. and Tyler, B. J. (1999) Capture
44. Bangs Laboratories Inc. (2009) Polymer agents for a quartz crystal microbalance-
Microspheres. continuous flow biosensor: Functionalized
self-assembled monolayers on gold. Anal.
45. Invitrogen Corporation. (2009) Beads and
Chim. Acta 399, 51–62.
Microspheres.
59. Ngundi, M. M., Taitt, C. R., and Ligler, F. S.
46. Sigma-Aldrich Company. (2009) Latex Beads.
(2007) Crosslinkers modify affinity of immo-
47. ThermoScientific – Pierce. (2009) Biotin bilized carbohydrates for cholera toxin. Sensor
Labeling of Antibodies and Other Proteins. Lett. 5, 621–624.
48. Anderson, G. P., Lamar, J. D., and Charles, P. T. 60. Ngundi, M. M., Taitt, C. R., McMurry, S. A.,
(2007) Development of a Luminex based Kahne, D., and Ligler, F. S. (2006) Detection
competitive immunoassay for 2,4,6-trinitro- of bacterial toxins with monosaccharide arrays.
toluene (TNT). Environ. Sci. Technol. 41, Biosens. Bioelectron. 21, 1195–1201.
2888–2893.
61. Rowe-Taitt, C. A., Cras, J. J., Patterson, C.
49. Anderson, G. P. and Goldman, E. R. (2008) H., Golden, J. P., and Ligler, F. S. (2000) A
TNT detection using llama antibodies and a ganglioside-based assay for cholera toxin using
two-step competitive fluid array immunoas- an array biosensor. Anal. Biochem. 281,
say. J. Immunol. Methods 339, 47–54. 123–133.
50. Kulagina, N. V., Shaffer, K. M., Ligler, F. S., 62. Song, X. D., Shi, J., and Swanson, B. (2000)
and Taitt, C. R. (2007) Antimicrobial pep- Flow cytometry-based biosensor for detection
tides as new recognition molecules for screen- of multivalent proteins. Anal. Biochem. 284,
ing challenging species. Sens. Actuators B 35–41.
Chem. 121, 150–157.
63. Moran-Mirabal, J. M., Throckmorton, D.,
51. Ngundi, M. M., Kulagina, N. V., Anderson, Singh, A. K., and Craighead, H. G. (2004)
G. P., and Taitt, C. R. (2006) Nonantibody- Micropatterning of functional lipid domains
based recognition: Alternative molecules for for toxin detection. Mater. Res. Soc. Symp.
detection of pathogens. Expert Rev. Proteomics Proc. EXS-1, H3.20.21–H23.20.23.
3, 511–524.
64. Chen, H., Zheng, Y., Jiang, J.-H., Wu, H.-L.,
52. Jain, K. K. (2006) Nanoparticles as targeting Shen, G.-L., and Yu, R.-Q. (2008) An ultrasen-
ligands. Trends Biotechnol. 24, 143–145. sitive chemiluminescence biosensor for cholera
53. Hermanson, G. T. (2009) Bioconjugate toxin based on ganglioside-functionalized sup-
Techniques (2nd ed.). Elsevier Science and ported lipid membrane and liposome. Biosens.
Technology Books, Boulder, CO. Bioelectron. 24, 684–689.
54. Porschewski, P., Grattinger, M. A. M., 65. Kulagina, N. V., Anderson, G. P., Ligler, F. S.,
Klenzke, K., Erpenbach, A., Blind, M. R., and Shaffer, K. M., and Taitt, C. R. (2007)
Schafer, F. (2006) Using aptamers as capture Antimicrobial peptides: New recognition
reagents in bead-based assay systems for diag- molecules for detecting botulinum toxins.
nostics and hit identification. J. Biomol. Screen. Sensors 7, 2808–2824.
11, 773–781. 66. Kulagina, N. V., Lassman, M. E., Ligler, F. S.,
55. Cheng, A. K. H., Ge, B., and Yu, H. Z. (2007) and Taitt, C. R. (2005) Antimicrobial
Aptamer-based biosensors for label-free volta- peptides for detection of bacteria in biosensor
mmetric detection of lysozyme. Anal. Chem. assays. Anal. Chem. 77, 6504–6508.
79, 5158–5164. 67. Taitt, C. R., North, S. H., and Kulagina, N. H.
56. Minunni, M., Tombelli, S., Gullotto, A., Luzi, (2009) Antimicrobial peptide arrays for detec-
E., and Mascini, M. (2004) Development of tion of inactivated biothreat agents. Methods
biosensors with aptamers as bio-recognition Mol. Biol. 570, 233–255.
Chapter 7
Multivalent Conjugation of Peptides, Proteins,

and DNA to Semiconductor Quantum Dots
Duane E. Prasuhn, Kimihiro Susumu, and Igor L. Medintz
Abstract
Semiconductor nanocrystals or quantum dots (QDs) have become well-established as a unique nanoparticle
scaffold for bioapplications due to their robust luminescent properties. In order to continue their devel-
opment and expand this technology, improved methodologies are required for the controllable function-
alization and display of biomolecules on QDs. In particular, efficient routes that allow control over ligand
loading and spatial orientation, while minimizing or eliminating cross-linking and aggregation are needed.
Two conjugation approaches are presented that address these needs: (1) polyhistidine-based metal-
affinity self-assembly to QD surfaces and (2) carbodiimide-based amide bond formation to carboxy-
functionalized polyethylene glycol or PEGylated QDs. These approaches can be successfully employed in
the construction of a variety of QD-biomolecule constructs utilizing synthetic peptides, recombinant
proteins, peptides, and even modified DNA oligomers.
Key words: Quantum dots, Semiconductor nanocrystal, Bioconjugation, Nanoparticle, Self-

assembly, EDC coupling, Peptide, Protein, Polyhistidine, Metal affinity, Fluorescence, Biosensing
1. Introduction
1.1. Biomolecule- One of the most prominent research areas in nanotechnology

Functionalized, continues to be the development of nanoparticle systems for use in
Luminescent biomedical and bioinorganic nanoscale engineering applications.
Semiconductor These systems offer several advantages over those that are conven-
Quantum Dots tionally used; for example, nanoparticles can be imbued with multiple
molecular functionalities, allowing them to perform both a therapeutic
and a diagnostic function simultaneously (1). Luminescent semi-
conductor nanocrystals, or quantum dots (QDs), have now become
well-established fluorophores and are one of the most popular types
of nanomaterials being explored for use in in vitro and in vivo
bioapplications (2–5). This utility derives from their unique
95
96 Prasuhn, Susumu, and Medintz
photophysical properties which include broad absorption spectra

with large molar extinction coefficients (e ~ 105–106 M−1 cm−1);
narrow, symmetrical photoluminescence (PL) spectra (35–45 nm
full width at half-maximum), which can be tuned from the UV to
near-IR as a function of QD core size and constituent material
(the PL peaks red-shift as the radius of the nanoparticle is
increased); high quantum yields of up to 40–60%, effective Stokes
shifts of 300–400 nm; large multiphoton action cross-sections
with typical values of 8,000 to >20,000 Goeppert-Mayer (GM)
units at 800 nm (6); access to facile multiplexing configurations;
and unparalleled resistance to photo- and chemical degradation as
compared to conventional organic fluorophores and fluorescent
proteins (3, 7–9). Further, these properties allow QDs to function
as excellent fluorescence resonance energy transfer (FRET) donors
in a variety of biosensing configurations (7).
Core/shell QD materials are typically used in biological applica-
tions, where a wider band-gap semiconductor shell functions to
passivate the core material, prevent leaching, and improve fluo-
rescence quantum yield (3, 8–11). These nanomaterials are com-
monly synthesized from binary combinations of semiconductor
materials, including ZnS, CdS, CdSe, InP, CdTe, PbS, PbTe, and
PbSe and can have a range of emissions from the UV to near-IR
depending upon the constituents used (3, 8–11). The reactions
used to synthesize these structures are usually carried out in
organic solution at high temperatures using pyrophoric precursors
and the resulting particles are stabilized with hydrophobic organic
ligands that lack intrinsic aqueous solubility (10, 11). This presents
an issue for the use of such structures in biological applications,
since water-soluble antibodies, proteins, peptides, or DNA must
be attached to the QD surface to achieve a conjugate structure
capable of targeting and biorecognition in an aqueous environ-
ment. Therefore, the QD surfaces require further modification
with bifunctional molecular ligands that imbue them with aqueous
solubility (3, 8) and that sometimes display pertinent groups
(such as carboxyls or amines) for subsequent chemical coupling
to biomolecules of interest. The two major strategies for accom-
plishing this are either the use of bifunctional amphiphilic polymers
that interdigitate with the native organic surface ligands or hydro-
philic molecules that replace the native layer via “cap” exchange
to provide a hydrophilic particle surface (3, 8, 12).
Unfortunately, most strategies for functionalizing the surface
of QDs with biomolecules provide little control over their loading
and/or spatial orientation. For instance, although biotin-Avidin
chemistry is commonly used, it is always constrained by the need
to label both participants appropriately, their susceptibility to
cross-linking, and the presence of large multivalent Avidin protein
(64 kDa) intermediaries in the conjugate (13). Also, with very
few exceptions, adsorption or noncovalent association (through
Multivalent Conjugation of Peptides, Proteins, and DNA 97
electrostatic or hydrophobic interactions) of natural/engineered

proteins or nucleic acids generally yields poor control over loading
and results in heterogeneity in spatial orientation and mixed func-
tional avidity (3, 4, 8). Functional groups commonly targeted for
biomodification, such as carboxyls and amines, are ubiquitous in
biological molecules and the resulting functionalization with
QDs, or other nanoparticles, can be heterogeneous unless appro-
priate issues are considered. Reviews expanding on many of these
points are available for the interested reader (3, 4, 8–10).
Thus, alternative QD conjugation chemistries that are effi-
cient, targeted, utilize functionalities orthogonal to common bio-
logical moieties (i.e., do not cross-react), minimize heterogeneity
in attachment, and allow for control over conjugate loadings can
greatly increase the utility of these nanomaterial fluorophores.
Two such chemistries for functionalizing QDs are presented here:
(1) polyhistidine-based self-assembly which is applicable to proteins,
peptides, and modified DNA and (2) carbodiimide-based cova-
lent coupling, which is more suited to peptides and DNA. When
appropriately implemented, these chemistries can provide high-
affinity attachments with control over the valence and orientation
of biomolecules attached to a QD particle.
1.2. Self-Assembly Many proteins are recombinantly engineered to express N- or

of Polyhistidine- C-terminal hexahistidine (His)6 sequences so that they can be
Appended purified over Ni 2+ -nitrilotriacetic acid (NTA) chelate media (14, 15).
Biomolecules This technology is based on the high-affinity interactions of the
to Quantum Dots histidine side-chain imidazolium groups with chelated divalent
cations, including those of Ni, Cu, Fe, Cr, Co, Zn, and Mn. Over
the last few years, the same type of approach has been used to
synthesize and characterize biomolecule–semiconductor QD
conjugates where the polyhistidine-appended biomolecule is
attached to the surface of the semiconductor QD by taking advan-
tage of the metal-affinity interaction between, for example, the
Zn-rich surface of a CdSe/ZnS core/shell QD and the histidine
side-chain of a biomolecule of interest (see Fig. 1a) (17). The self-
assembly that occurs rapidly (on the order of seconds to minutes
after mixing), is equally efficient when using QDs capped with
small, negatively charged dihydrolipoic acid (DHLA) or neutral
polyethylene glycol (PEG) ligands, has a high-affinity equilibrium
constant in solution (Ka ~1 nM−1), and allows for control over the
number of substrate molecules arrayed on a single QD through
modulation of the molar ratios of the participants used (i.e., number
of nanoparticles to number of biomolecules) (17). In some cases,
such self-assembly can also provide control over the spatial orien-
tation of the biomolecule on the QD surface (18). This metal-
affinity driven strategy has already been demonstrated for
assembling proteins, peptides, and appropriately modified peptide–
DNA conjugates on QDs to create a variety of sensor systems that
a Polyhistidine self-assembly
DH LA
O
H
Functional biomolecule N S OH
S
S
HN S
N
Zn+2 dative thiol
OC bonds
Protein, peptide, DNA QD DH LA-PEG
HN S O
Zn+2
N S S O R
OC N O
S H n
N
H
R = OH, OMe, N3, NH2, NHOCCH2CH2COOH
b Assembled valence: # of MBP/QD

Poisson
Observed 0 .1 .2 .5 1 2 4
valence: - distribution
>2
2
1
0
+
MBP = N-maltose binding protein - His5 - COOH
c O
SO3Na
H2N
Cl- HO N
N N+ Primary amine(s)
C H containing biomolecule
N O
EDC Sulfo-NHS SO3Na
O
H Cl-
N N N+ N
H O O
QD COOH
O O QD CONH
O
O QD-biomolecular
QD SO3Na
conjugate
QD HO N
O
o-Acylisourea Sulfo-NHS ester
Fig. 1. (a) General representation of a decorated QD nanocrystal (CdSe core in red and ZnS overcoat in blue), not to scale.
The QDs are made more hydrophilic by cap-exchanging with dihydrolipoic acid (DHLA) or a PEG derivative that can display
a variety of terminal functionalities (DHLA-PEG, NHC0 is an amide bond). A polyhistidine-appended functional biomolecule
(only two histidine residues shown) coordinated to the QD surface by metal-affinity driven self-assembly is also displayed.
It should be noted that although two adjacent histidines are depicted as involved in coordination to two differing Zn 2+ ions,
the exact nature of the complexation probably involves multiple, varying coordination configurations. (b) An agarose gel
showing the separation of self-assembled QD-protein bioconjugates with different numbers of proteins per conjugate. In
this example, maltose binding protein (MBP, MW ~ 44 kDa) with a C-terminal pentahistidine sequence was self-assembled
to DHLA-QDs. At small ratios, samples show several mobility shift bands due to the Poisson distribution as expected. The
white arrow indicates a sample valence of 1 which can be expected to manifest three types of ratios: ~33% with 0 mol-
ecules/QD, ~33% with 1 molecule/QD, and ~33% with >1 molecule/QD (mostly 2/QD). These conjugates merge into a
single band indicative of a homogeneous distribution of conjugate sizes as the average protein to-QD ratio increases.
Figure adapted from ref. 16 with permission from the American Chemical Society. (c) General schematic for the conjuga-
tion of aminated biomolecules to PEGylated-QDs (PEG molecules terminate in carboxyls) through amide bond formation
using an EDC-mediated reaction with a sulfo-NHS ester intermediary step (13).
target nutrients, explosives, and other small molecules (3, 7, 19–22)

and is being steadily adopted by many research groups (23, 24).
Several factors should be considered before undertaking this
type of bioconjugation with QDs. First, the relevant biomolecules
must display clearly available polyhistidine sequences (i.e., not
internal or sterically hindered). As mentioned, proteins are commonly
expressed with recombinant (His)6-sequences and nascent peptides

can be synthesized to include them. DNA needs to be chemically
linked to the (His)6-sequences and several strategies for achieving
this have been demonstrated with terminally-modified oligonu-
cleotides (17, 19–21, 25). Similar to protein purification, four to
six consecutive His residues are sufficient for high-affinity attach-
ment to QDs (17). Due to the small ligand size, proteins, peptides,
and DNA can be assembled to dihydrolipoic acid (DHLA)-
functionalized QDs, while the larger size of the PEGylated ligands
may sterically preclude assembly of larger globular proteins
(see Fig. 1a). When these types of strategies were demonstrated
with commercial QD preparations, the addition of small amounts of
Ni2+ ions that presumably interacted with surface functional groups
(principally carboxyls), was necessary to form a functional Ni2+
chelate structure (26). Further, QDs have been covalently modi-
fied with NTA groups on their surface ligands to facilitate similar
attachments (27).
In this approach, the number of (sometimes fluorescently
labeled) biomolecules attached to the QDs can usually be deter-
mined through two methods. The first is through physical separa-
tion of the conjugates using an agarose gel (16). Under appropriate
conditions, even mono-labeled QD species can be easily resolved
(see Fig. 1b) (28). In the second method, changes in FRET
between the central QD donor and increasing numbers of accep-
tor dye-labeled biomolecules attached to the QD are monitored
and the number of self-assembled biomolecules is then determined
by comparison to a known or previously established “calibration
curve” (see Fig. 2) (19). Poisson distribution kinetics determines
the average number of biomolecules attached per QD since self-
assembly is a random process (29). This may only be an issue when
working at low valences of <4 biomolecules per QD as unlabeled
QDs can be assumed to be present. At valences of >4, the actual
distributions will more closely match predictions. For example,
when assembling a nominal ratio of 1 biomolecule per QD, the
resulting bioconjugates can be expected to manifest three types of
ratios: ~ 33% with 0 molecules/QD, ~ 33% with 1 molecule/QD,
and ~33% with >1 molecule/QD (mostly 2/QD) (see Fig. 1b).
1.3. Carbodiimide Amide bond formation using 1-ethyl-3-(3-dimethylaminopropyl)

Coupling to carbodiimide (EDC) chemistry is one of the most common bio-
Polyethylene Glycol conjugation chemistries employed for modifying biomolecules
Functionalized including peptides, proteins, and DNA (see Fig. 1c) (13). The
Quantum Dots popularity of this chemistry arises from the fact that the two targeted
groups (amines and carboxyls) are ubiquitous to proteins and
peptides and can be chemically added to DNA; the same func-
tional groups can be displayed on most nanoparticle surfaces in a
relatively facile manner. Also, the reagents for performing this
chemistry can be purchased in large quantities at relatively cheap
a
Ratio of Cy5-peptide
40000
per QD
590 nm QD 0
Photoluminescence (AU)
1
2
30000
4
6
8
20000 Cy5 10
15
10000
0
600 650 700 750
Wavelength (nm)
b 2000
Equivalent Cy5’s
1
Photoluminescence (AU)
1500 2
4
Cy5 6
8
1000 10
15
500
0
600 650 700 750
Wavelength (nm)
c 1.0
Relative Emission, Efficiency
0.8
0.6 QD PL
FRET efficiency E
FRET E corrected
0.4
0.2
0.0
0 2 4 6 8
n, # conjugates/QD
Fig. 2. (a) Representative FRET spectral results for a QD–Cy5-labeled acceptor system (QD donor emission maxima at
590 nm and Cy5 acceptor emission maxima at 670 nm) for an increasing number of Cy5-labeled His6-peptides self-
assembled per QD. The QD donor PL shows the characteristic losses in magnitude as the Cy5-acceptor emission rises.
(b) The direct-acceptor excitation spectrum for equivalent amounts of Cy5-labeled His6-peptides alone. Note the significant
increase in FRET-sensitized Cy5 emission at ~670 nm when coordinated to the QDs. All spectra were taken using 350 nm
excitation. (c) Normalized QD PL loss vs. n (acceptor valence), corresponding FRET efficiency, and Poisson-corrected FRET
efficiency for the QD–Cy5-labeled His6-peptide system. The QD-dye separation distance or r value determined for this
system using Eq. 7.3 is ~43 Å and the corrected value using Eq. 7.4 is 45 Å. This type of data would, for example, be useful
to quantitatively monitor enzymatic cleavage of the peptide while attached to the QD in a protease sensor configuration (19, 22).
prices. Indeed, the QDs used in some of the seminal biomedical

demonstrations where functionalized with proteins for cellular
uptake by reacting to the thiol-alkyl-COOH ligands utilizing
EDC chemistry (30). Subsequent studies, however, showed that
QDs capped with DHLA or other similar short-chain thiol-alkyl-
COOH ligands often encounter a loss of solubility and macro-
scopic aggregation at the neutral to acidic pHs required for this
chemistry. This aggregation results from the protonation of the
terminal carboxyl groups responsible for maintaining colloidal
stability (31). The more recent development of pH stable, PEG-
based QD solubilizing ligands displaying a variety of terminal
functionalities (including carboxylic acids) circumvent many of
the previous issues (12).
Again, several factors are important for consideration when
undertaking this type of bioconjugation with QDs. The chemistry
a priori should work equally well with QDs displaying carboxyls
and targeting amines on the biomolecules as well as in the converse
configuration. The best results to date, however, have been con-
sistently achieved utilizing QDs displaying carboxyls and targeting
either peptides or DNA site-specifically modified with a unique,
terminal amine. This type of set-up provides for control over
biomolecule orientation on the QD since the attachment can
only be in one configuration. EDC chemistry for attaching pro-
teins to QDs will usually result in mixed orientation/avidity and
cross-linking due to the multiple possible molecular conforma-
tions. These issues can be addressed by testing several different
ratios of proteins to QDs to optimize the outcome. Although
EDC can be used directly in the reaction with both participants,
better results are achieved in combination with sulfo-NHS to first
activate the carboxylic acids (13). EDC has a very short usable
half-life in aqueous solution and it has been found that the reac-
tions can be “recharged” with additional aliquots of fresh EDC
without deleterious effects to provide higher conjugation yields.
The number of moieties attached per QD can be controlled
empirically through the molar excess of biomolecule used and
usually requires several iterative attempts if a specific ratio is
required.
Successful bioconjugation can be confirmed by monitoring
changes in QD mobility on agarose or polyacrylamide gels, by
using Förster resonance energy transfer (FRET) interactions, or
by UV absorbance if the biomolecule is dye-labeled (see Figs. 2
and 3) (12) although there are intrinsic limitations to each. In
fact, we commonly utilize side-by-side reactions where a dye-
labeled peptide or DNA is attached as a positive control. The
difference in appearance of the QD-protein conjugate profiles
relative to control QD samples in gel analysis are used as inferen-
tial evidence to verify or confirm conjugation (see Fig. 3). This
data does not, however, tell us the actual ratio of proteins conju-
gated/QD nor if they are still functional. Further, as QD absorbance
a Agarose gel:
QD - 1 2 3 4 - QD
_
+
b PAGE gels:
_
QD 4 3 2 1 QD QD 4 3 2 1 QD
UV Commassie stain +
Fig. 3. Representative agarose (a) and polyacrylamide gels (b) for PEGylated QDs derivatized with a ~25 kDa protein via
EDC coupling chemistry. QD designates the nanoparticles alone. QDs were reacted with a 3×, 8×, 16×, and 32× excess
molar ratio of proteins as designated by samples 1–4, respectively. In the agarose gel, the differences in migration and
intensity correlate with the formation of a much larger molecular weight species. In the PAGE gels, the staining of proteins
by Coomassie, along with the QD fluorescence, correlate to the newly formed species, especially the slower moving,
higher molecular weight QD-protein composites.
dramatically increases as a continuum towards the UV, this will

most often mask the commonly used DNA/protein absorptions
at 260 and 280 nm, respectively, severely limiting this portion of
the spectrum for analysis of component absorptions.
2. Materials
2.1. Self-Assembly 1. Dissolve approximately 1 mg of peptide (here, a peptide

of Polyhistidine- sequence CSTRIDEANQRATKLP9SH6 that terminates with
Appended a C-terminal His6 sequence was utilized as an example) in
Biomolecules 100 mL of 1× phosphate buffered saline (PBS; 137 mM NaCl,
to Quantum Dots 10 mM phosphate, 2.7 mM KCl; Fisher), pH = 7.4 (see Note 1).
This example is similar to that described in ref. 22.
2. Two vials of Cy5-maleimide monoreactive dye (GE
Healthcare; Piscataway, NJ) dissolved in 100 mL dimethyl
sulfoxide (DMSO) total volume (see Note 2).
3. Three mini-columns containing Ni-NTA Agarose resin
(Qiagen; Valencia, CA) (see Note 3).
4. 300 mM imidazole dissolved in 1× PBS, pH = 7.4.
5. Oligonucleotide purification cartridge (OPC; Invitrogen;

Carlsbad, CA) equilibrated with 3 mL acetonitrile and 3 mL
2 M triethylammonium acetate (TEAA, Invitrogen; Carlsbad,
CA).
6. 70% acetonitrile/water (v/v).
7. 100 mM TEAA.
8. 1 mM aqueous solution of QDs surface functionalized with
dihydrolipoic acid (DHLA) or DHLA-PEG600 (see Note 4).
9. 96-Well microtiter plates (polystyrene with nonbinding surface,
Corning; Corning, NY).
2.2. EDC Coupling 1. 100 mM Lissamine rhodamine B ethylenediamine dye

to PEGylated QDs (Invitrogen; Carlsbad, CA) (for this example) in 1× PBS,
for Bioconjugation pH = 7.4.
2. 100 mM N-hydroxysulfosuccinimide (sulfo-NHS; Pierce
Biotechnology, Rockford, IL) in 1× PBS, pH = 7.4.
3. 500 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
(EDC; Pierce Biotechnology) in 1× PBS, pH = 7.4.
4. 1 mM aqueous solution of QDs (emission at ~590 nm) capped
with DHLA-PEG750-OMe (methoxy)/DHLA-PEG600-
COOH (19:1 ratio of Methoxy:COOH terminated ligand –
equivalent to QDs functionalized with 5% surface carboxyls)
(see Note 5).
5. Disposable PD-10 desalting gel columns (GE Healthcare,
Piscataway, NJ) equilibrated with 1× PBS, pH = 7.4.
3. Methods
3.1. Self-Assembly 1. Mix peptide solution with reactive dye solution in a 1.5-mL
of Polyhistidine- Eppendorf tube and incubate overnight (16–18 h) in the
Appended dark at room temperature or alternatively at 4°C with con-
Biomolecules to stant agitation using a benchtop vortexer.
Quantum Dots and 2. Load the peptide–dye mixture over three consecutive columns
Förster Resonance of Ni-NTA agarose using 1-mL syringes (see Note 6).
Energy Transfer 3. Wash each column with 10 mL 1× PBS to remove excess
Characterization of the dye.
Resulting Conjugates
4. Elute the Cy5-peptide with ~2 mL 300 mM imidazole per
column.
5. Load the dye-modified peptide eluent (in imidazole) into an
OPC containing reverse phase media (see Note 7).
6. Rinse the cartridge with 50 mL of 100 mM TEAA and 50 mL
deionized water (see Note 8).
7. Elute the substrate from the OPC with a 70% acetonitrile/

water solution (1–2 mL).
8. Concentrate the sample as necessary. In these experiments, a
DNA120 speed-vacuum centrifuge (GMI; Ramsey, MN) was
used.
9. Measure the absorption spectra of the Cy5-peptide and quan-
tify the amount of labeled peptide (see Note 9).
10. Aliquot the dye–peptide into 1.5-mL Eppendorf tubes, then
dry and store the samples as a pellet at −20°C in a dessicator
(can be stored >6 months).
11. When ready to use, resuspend the dye–peptide in 10% DMSO/
water (v/v) yielding a final concentration of 10 mM.
12. To multiple 1.5-mL Eppendorf tubes, add 20 mL 1 mM
QD-DHLA-PEG600 particles (equivalent to 20 pmols of QD
per sample in 100 mL or 0.2 mM).
13. To each tube, add the necessary volume of dye–peptide solu-
tion to yield the desired ratio of dye–peptide to QD (i.e.,
2.0 mL for 1 peptide/QD, 4.0 mL for 2 peptides/QD, etc.)
in a final volume of 100 mL of 1× PBS buffer. Mix/vortex and
briefly centrifuge to collect the samples in the bottom of the
tubes. Table 1 shows an example of the volumes of reagents
and concentrations necessary to achieve desired peptide
loadings.
14. Incubate the samples at room temperature for at least
15 min.
15. Transfer 100 mL of each reaction into a microtiter 96-well
plate and measure the emissions on a plate reader with an
excitation at 350 nm. Here, readings were taken on a Tecan
Safire Dual Monochromator Multifunction Microtiter
Plate Reader (Tecan; Durham, NC). Representative emission
Table 1
Reagent volumes utilized for QD-peptide self-assembly
Desired ratio of peptide/QD

Reagent (mL) 0 1 2 4 6 8 10 15
QD (1 mM) 20 20 20 20 20 20 20 20
Peptide (10 mM) – 2 4 8 12 16 20 30
10× PBS 10 10 10 10 10 10 10 10
H2O 70 68 66 62 58 54 50 40
Total 100 100 100 100 100 100 100 100
results are presented in Fig. 2 (see Note 10). Most laboratory

benchtop fluorometers can also be used.
16. Analyze the resulting FRET interactions (Fig. 2) as follows:
First, determine the Förster separation distance R0
(defined as the spacial separation distance between donor and
acceptor corresponding to 50% energy transfer efficiency) for
the particular QD-donor dye-acceptor pair that is being uti-
lized using (32):
 [9, 000 × (ln10) ]k p2

1/6

R0 =  QD I  , (1)
 128p nD N A
5 4


where, nD is the refractive index of the medium, Q D is the
QD PL quantum yield (determined through comparison of
QD sample emission with the emission of a standard com-
pound of known quantum yield; typical values for the
reported QDs are 20–30%), I is the integral of the spectral
overlap function, k p2 is the dipole orientation factor
( k p2 = 2/3 is used for self-assembled donor–acceptor pairs
with random dipole orientations) and NA is Avogadro’s
number (18, 19, 32–34).
Then, extract the average energy transfer efficiency E
from the fluorescence data for the biomolecule-QD conju-
gates using the expression:
( FD − FDA )
E= , (2)
FD
where, FD and FDA are, respectively, the fluorescence intensi-
ties of the QD donor alone and donor in the presence of the
acceptor(s) (i.e., the labeled biomolecules) (32) (Fig. 2).
Then, if analyzed within the Förster dipole–dipole formal-
ism, the energy transfer efficiency data can be fit to the expres-
sion (34):
nR06
E= , (3)
nR06 + r 6

where, R0 was determined using Eq. 1, n is the average num-
ber of fluorescently labeled biomolecules per QD and r is the
QD-donor dye-acceptor center-to-center separation distance.
The metal-His driven self-assembly of biomolecules to these
nanocrystals provides conjugates with a centrosymmetric dis-
tribution of acceptors around a QD. For QDs self-assembled
with dye-labeled proteins and peptides, this distribution is
usually characterized by a constant average center-to-center

separation distance, r (19, 34). For conjugates having small
numbers of acceptors (n < 5), heterogeneity in the conjugate
valence can be accounted for by using a Poisson distribution
function, p(N,n), when fitting the efficiency data (29):
e− N
E = ∑ p ( N , n) E ( n) p ( N , n) = N n , (4)
n n!
where, n designates the exact numbers of acceptors (valence)

for conjugates with a nominal average valence of N (see
Note 11).
3.2. EDC Coupling 1. Here, an amine-containing dye is conjugated to a QD. To a

to PEGylated QDs 1.5-mL Eppendorf tube, add 118.8 mL of a QD-DHLA-
for Bioconjugation PEG750-OMe/DHLA-PEG600-COOH (19:1) stock solution.
This is similar to the example highlighted in ref. 12.
2. Add EDC (77 mL), sulfo-NHS (46.2 mL), and 1× PBS
(81 mL).
3. Incubate the reaction at room temperature in the dark with
orbital shaking for ~10 min.
4. Lissamine rhodamine B ethylenediamine (77 mL) is then added
to yield a final reaction volume of 400 mL with final concentra-
tions of 1.93 mM, 19.3 mM, 57.8 mM, and 19.3 mM for QD,
EDC, sulfo-NHS, and the dye, respectively (see Note 12).
5. Allow the mixture to react at room temperature in the dark
with orbital shaking for ~2 h (see Note 13). The EDC and
sulfo-NHS in the reaction can be recharged through the addi-
tion of further aliquots as needed.
6. Load the mixture onto a PD-10 desalting gel column and
elute the QD conjugate with 1× PBS (see Note 14).
7. Concentrate the sample as needed and run on a UV-visible
spectrophotometer to determine conjugate loading per QD if
the biomolecule is dye-labeled. If a protein is used, agarose or
polyacrylamide gels can provide inferential evidence that the
QDs have been conjugated (see Fig. 3), but not exact loading
numbers or verification of activity.
4. Notes
1. The presented work utilizes a peptide as an example substrate

for conjugation to QDs. These peptides are commonly pre-
pared by standard solid-phase synthesis on Rink amide resin and
can be subsequently modified with a dye molecule using
cysteine–maleimide chemistry. The methodology, however, works

in conjunction with a variety of bioconjugation chemistries and
can be extended to other polyhistidine-appended biomolecules,
such as recombinantly expressed proteins and peptides, and
DNA oligomers (17, 21, 25).
2. The labeling kit, as purchased, contains enough dye to label
1 mg of protein. Thus, to ensure complete labeling, use excess
dye (i.e., two or more vials).
3. Mini-columns with syringe attachments at both ends were
prepared by loading ~1 mL of Ni-NTA Agarose resin into
empty oligonucleotide purification cartridges (OPC, Applied
Biosystems).
4. CdSe/ZnS core/shell QDs capped with hydrophobic, organic
ligands (generally a mixture of trioctylphosphine/trioc-
tylphosphine oxide (TOP/TOPO) and hexadecylamine)
were synthesized using organometallic procedures as previ-
ously reported (11, 31, 35). The nanocrystals were made
water-soluble by exchanging the TOP/TOPO ligands with
polyethylene glycol-appended dihydrolipoic acid (DHLA) or
DHLA-PEG600 (PEG with MW ~ 600 or MW ~ 750) through
standard methods (12, 31, 36, 37). The above procedures
can be extended to cap particles of various sizes (diameters of
~4 to >10 nm) with DHLA and recent publications have also
utilized commercially available QDs (23, 38, 39).
5. Mixed surface CdSe/ZnS core/shell QDs were prepared as
mentioned in Note 4. The nanocrystals were made water-
soluble by exchanging the TOP/TOPO ligands with a 19:1
mixture of methoxy- and carboxy-terminated polyethylene
glycol-appended dihydrolipoic acid (DHLA-PEG750-OMe/
DHLA-PEG600-COOH, respectively) through standard
methods (12). Using QDs prepared with higher percentages
of carboxylated ligands on their surface will result in corre-
spondingly higher surface functionalization efficiencies.
6. To facilitate purification, the reaction mixture is diluted with
200–400 mL buffer. Then, the entire reaction mixture is
passed through a single column 10–15 times to ensure maxi-
mum binding. The remaining filtrate is sequentially passed
through a second column 10–15 times and then a third col-
umn 10–15 times. If necessary, more columns could be used
to ensure maximum product isolation, but generally all sub-
strate binding has occurred within the initial passes through
the two columns since ~1 mL of NTA resin can bind ~1 mg
of labeled protein.
7. Generally, a single loading through an OPC is sufficient. If,
however, there is still color in the filtrate (or one is working
with a colorless substrate), a second/third loading onto an
equilibrated OPC may be desired/necessary for maximum

product isolation.
8. The function of this step is twofold: it removes the imidazole
(which can quench QD photoluminescence), and it desalts/
lyophilizes the peptide prior to long-term storage.
9. This reactive dye has an extinction coefficient of
250,000 M−1 cm−1 at 649 nm.
10. Some biomolecules, such as large proteins, may have diffi-
culty accessing the QD surface for self-assembly due to steric
interactions with the PEG ligands. Recombinant addition of
a flexible, linker sequence between the polyhistidine residues
and the main protein structure can lead to enhanced coordi-
nation. QDs cap-exchanged with far smaller DHLA do not
experience such steric issues.
11. Due to the nature and scope of this publication, only a cur-
sory discussion is presented on the FRET analyses of these
QD systems. A more thorough explanation and discussion of
this topic can be obtained in refs. 7, 19, 34.
12. The 10 min incubation prior to dye addition is expected to
allow some pre-formation of the amine-reactive NHS-ester
intermediate. These are example reaction concentrations;
optimized reagent concentrations can vary. The reader is
referred to refs. 12, 13, 37 for a discussion of considerations
regarding reaction optimization.
13. At times, a nonfluorescent precipitate may form after reaction.
The exact nature of this material is unknown, but it is believed
to be some sort of salt by-product. Also, longer reaction times
are possible, but gradual photoluminescence loss can occur if
extended for much longer than a few hours (i.e., overnight).
14. As reported, this reaction is done at a QD concentration and
volume that allows monitoring of the emission using a UV
lamp during column elution. If less material is required, frac-
tions (1 mL or smaller) can be collected and the fraction con-
taining the desired eluent can be identified using a UV-visible
spectrophotometer. Samples can be concentrated using a
speed vacuum as well.
Acknowledgments
I. L. M. and K. S. acknowledge DTRA/ARO, ONR, NRL, and

the NRL-NSI for financial support. D. E. P. acknowledges an
ASEE fellowship through NRL.
References
1. Niemeyer, C. M. (2001) Nanoparticles, 14. Hochuli, E., Bannwarth, W., Dobeli, H.,
proteins, and nucleic acids: Biotechnology Gentz, R., and Stuber, D. (1988) Genetic
meets materials science. Angew. Chem. Int. approach to facilitate purification of recombi-
Ed. 40, 4128–4158. nant proteins with a novel metal chelate adsor-
2. Alivisatos, P. (2004) The use of nanocrystals bent. Biotechnology 6, 1321–1325.
in biological detection. Nat. Biotechnol. 22, 15. Schmitt, J., Hess, H., and Stunnenberg, H. G.
47–52. (1993) Affinity purification of histidine-tagged
3. Medintz, I., Uyeda, H., Goldman, E., and proteins. Mol. Biol. Rep. 18, 223–230.
Mattoussi, H. (2005) Quantum dot biocon- 16. Pons, T., Uyeda, H. T., Medintz, I. L., and
jugates for imaging, labeling and sensing. Mattoussi, H. (2006) Hydrodynamic dimen-
Nat. Mater. 4, 435–446. sions, electrophoretic mobility and stability of
4. Michalet, X., Pinaud, F. F., Bentolila, L. A., hydrophilic quantum dots. J. Phys. Chem. B
Tsay, J. M., Doose, S., Li, J. J., et al. (2005) 110, 20308–20316.
Quantum dots for live cells, in vivo imaging, 17. Sapsford, K. E., Pons, T., Medintz, I. L.,
and diagnostics. Science 307, 538–544. Higashiya, S., Brunel, F. M., Dawson, P. E.,
5. Parak, W. J., Pellegrino, T., and Plank, C. et al. (2007) Kinetics of metal-affinity driven
(2005) Labelling of cells with quantum dots. self-assembly between proteins or peptides
Nanotechnology 16, R9–R25. and CdSe-ZnS quantum dots. J. Phys. Chem.
6. Clapp, A. R., Pons, T., Medintz, I. L., C 111, 11528–11538.
Delehanty, J. B., Melinger, J. S., Tiefenbrunn, 18. Medintz, I. L., Konnert, J. H., Clapp, A. R.,
T., et al. (2007) Two-photon excitation of Stanish, I., Twigg, M. E., Mattoussi, H., et al.
quantum-dot-based fluorescence resonance (2004) A fluorescence resonance energy trans-
energy transfer and its applications. Adv. fer derived structure of a quantum dot-protein
Mater. 19, 1921–1926. bioconjugate nanoassembly. Proc. Nat. Acad.
7. Medintz, I. L. and Mattoussi, H. (2009) Sci. USA 101, 9612–9617.
Quantum dot-based resonance energy trans- 19. Medintz, I. L., Clapp, A. R., Brunel, F. M.,
fer and its growing application in biology. Tiefenbrunn, T., Uyeda, H. T., Chang, E. L.,
Phys. Chem. Chem. Phys. 11, 17–45. et al. (2006) Proteolytic activity monitored by
8. Klostranec, J. M. and Chan, W. C. W. (2006) fluorescence resonance energy transfer
Quantum dots in biological and biomedical through quantum-dot-peptide conjugates.
research: Recent progress and present chal- Nat. Mater. 5, 581–589.
lenges. Adv. Mater. 18, 1953–1964. 20. Berti, L., D’Agostino, P. S., Boeneman, K., and
9. Alivisatos, A. P., Gu, W., and Larabell, C. A. Medintz, I. L. (2009) Improved peptidyl link-
(2005) Quantum dots as cellular probes. ers for self-assembling semiconductor quantum
Annu. Rev. Biomed. Eng. 7, 55–76. dot bioconjugates. Nano Res. 2, 121–129.
10. Murray, C. B., Kagan, C. R., and Bawendi, M. G. 21. Medintz, I. L., Berti, L., Pons, T., Grimes, A.
(2000) Synthesis and characterization of F., English, D. S., Alessandrini, A., et al.
monodisperse nanocrystals and close-packed (2007) A reactive peptidic linker for self-
nanocrystal assemblies. Annu. Rev. Mater. Sci. assembling hybrid quantum dot-DNA bio-
30, 545–610. conjugates. Nano Lett. 7, 1741–1748.
11. Dabbousi, B. O., Rodriguez-Viejo, J., 22. Sapsford, K. E., Pons, T., Medintz, I. L., and
Mikulec, F. V., Heine, J. R., Mattoussi, H., Mattoussi, H. (2006) Biosensing with lumi-
Ober, R., et al. (1997) (CdSe)ZnS core-shell nescent semiconductor quantum dots. Sensors
quantum dots: synthesis and optical and struc- 6, 925–953.
tural characterization of a size series of highly 23. Dennis, A. M. and Bao, G. (2008) Quantum
luminescent materials. J. Phys. Chem. B 101, dot-fluorescent protein pairs as novel fluores-
9463–9475. cence resonance energy transfer probes. Nano
12. Susumu, K., Uyeda, H. T., Medintz, I. L., Lett. 8, 1439–1445.
Pons, T., Delehanty, J. B., and Mattoussi, H. 24. Ipe, B. I., Lehnig, M., and Niemeyer, C. M.
(2007) Enhancing the stability and biological (2005) On the generation of free radical species
functionalities of quantum dots via compact from quantum dots. Small 1, 706–709.
multifunctional ligands. J. Am. Chem. Soc. 25. Pons, T., Medintz, I. L., Sapsford, K. E.,
129, 13987–13996. Higashiya, S., Grimes, A. F., English, D. S., et al.
13. Hermanson, G. T. (2008) Bioconjugate (2007) On the quenching of semiconductor
Techniques (2nd Ed.) Academic Press, San quantum dot photoluminescence by proximal
Diego. gold nanoparticles. Nano Lett. 7, 3157–3164.
26. Yao, H., Zhang, Y., Xiao, F., Xia, Z., and Rao, based on quantum dot FRET donors. Nat.
J. (2007) Quantum dot/bioluminescence Mater. 2, 630–638.
resonance energy transfer based highly sensi- 34. Clapp, A. R., Medintz, I. L., Mauro, J. M.,
tive detection of proteases. Angew. Chem. Int. Fisher, B. R., Bawendi, M. G., and Mattoussi,
Ed. 46, 4346 –4349. H. (2004) Fluorescence resonance energy
27. Kim, J., Park, H. -Y., Kim, J., Ryu, J., Kwon, transfer between quantum dot donors and
D. Y., Grailhe, R., et al. (2008) dye-labeled protein acceptors. J. Am. Chem.
Ni-nitrilotriacetic acid-modified quantum Soc. 126, 301–310.
dots as a site-specific labeling agent of histi- 35. Peng, Z. A. and Peng, X. (2001) Formation
dine-tagged proteins in live cells. Chem. of high-quality CdTe, CdSe, and CdS nano-
Commun. (16), 1910–1912. crystals using CdO as precursor. J. Am. Chem.
28. Sperling, R. A., Pellegrino, T., Li, J. K., Soc. 123, 183–184.
Chang, W. H., and Parak, W. J. (2006) 36. Mei, B. C., Susumu, K., Medintz, I. L.,
Electrophoretic separation of nanoparticles Delehanty, J. B., Mountziaris, T. J., and
with a discrete number of functional groups. Mattoussi, H. (2008) Modular poly(ethylene
Adv. Funct. Mater. 16, 943–948. glycol) ligands for biocompatible semicon-
29. Pons, T., Medintz, I. L., Wang, X., English, ductor and gold nanocrystals with extended
D. S., and Mattoussi, H. (2006) Solution- pH and ionic stability. J. Mater. Chem. 18,
phase single quantum dot fluorescence reso- 4949–4958.
nant energy transfer sensing. J. Am. Chem. 37. Mei, B. C., Susumu, K., Medintz, I. L., and
Soc. 128, 15324–15331. Mattoussi, H. (2009) Polyethylene glycol-
30. Chan, W. C. W. and Nie, S. (1998) Quantum based bidentate ligands to enhance quantum
dot bioconjugates for ultrasensitive noniso- dot and gold nanoparticle stability in biologi-
topic detection. Science 281, 2016–2018. cal media. Nat. Protoc. 4, 412–423.
31. Mattoussi, H., Mauro, J. M., Goldman, E. R., 38. Dif, A., Henry, E., Artzner, F., Baudy-Floc’h,
Anderson, G. P., Sundar, V. C., Mikulec, F. M., Schmutz, M., Dahan, M., et al. (2008)
V., et al. (2000) Self-assembly of CdSe-ZnS Interaction between water-soluble peptidic
quantum dot bioconjugates using an engi- CdSe/ZnS nanocrystals and membranes:
neered recombinant protein. J. Am. Chem. Formation of hybrid vesicles and condensed
Soc. 122, 12142–12150. lamellar phases. J. Am. Chem. Soc. 130,
32. Lakowicz, J. R. (2006) Principles of 8289–8296.
Fluorescence Spectroscopy (3rd Ed.) Springer, 39. Liu, W., Howarth, M., Greytak, A. B., Zheng,
New York. Y., Nocera, D. G., Ting, A. Y., et al. (2008)
33. Medintz, I. L., Clapp, A. R., Mattoussi, H., Compact biocompatible quantum dots func-
Goldman, E. R., Fisher, B., and Mauro, J. M. tionalized for cellular imaging. J. Am. Chem.
(2003) Self-assembled nanoscale biosensors Soc. 130, 1274–1284.
Chapter 8
A Single SnO2 Nanowire-Based Microelectrode

Jun Zhou, Yaguang Wei, Qin Kuang, and Zhong Lin Wang
Abstract
SnO2 nanowires are synthesized via the chemical-vapor-deposition process using gold as the catalyst and
characterized by X-ray powder diffraction, field-emission scanning electron microscopy, high-resolution
transmission electron microscopy, and cathodoluminescence. Finally, a new type of microelectrode based
on a single SnO2 nanowire is fabricated. This microelectrode is expected to have promising applications
in various chemical and biomedical nanosensors.
Key words: SnO2, Nanowire, Microelectrode, Chemical sensor, Biosensor
1. Introduction
One-dimensional (1D) nanowires are ideal building blocks for

constructing nanosized devices due to their high surface to vol-
ume ratio and their special physical and chemical properties (1).
One of the most active research areas in nanoscience is the design
and fabrication of chemical and biosensing devices based off of
these wires due to their diverse practical and potential applica-
tions (2–5). To improve the sensing characteristics, a general
route is to make chemical and biomedical sensors at the nano-
scale, taking advantage of the large surface areas of nanoscale
structures. Chemical nanosensors based on carbon nanotubes
(6, 7), silicon nanowires (1), and ceramic (8) nanostructures are
of particular interest.
SnO2 is an n-type semiconductor and nanostructures made
from this material possess many unique optical and electrical
properties. SnO2 has a wide band gap of 3.6 eV at 300 K, and
possesses remarkable receptivity variation in gaseous environ-
ments, high-optical transparency in the visible range (up to 97%),
111
112 Zhou et al.
low resistivity, and excellent chemical stability. These properties

make SnO2 nanowires well suited for chemical sensors and trans-
parent conducting electrodes.
In this chapter, we focus on the synthesis of SnO2 nanowires
and the fabrication of a single SnO2 nanowire-based microelec-
trode (9). The SnO2 nanowires are synthesized via the chemical
vapor deposition (CVD) process using gold as the catalyst and
characterized by X-ray powder diffraction (XRD), scanning elec-
tron microscopy (SEM), high-resolution transmission electron
microscopy (HRTEM), and cathodoluminescence (CL). Then,
we describe the fabrication of a single SnO2 nanowire-based
microelectrode.
2. Materials
2.1. Synthesis 1. Growth system: a homemade CVD experimental apparatus

of SnO2 Nanowires including a horizontal quartz reaction chamber (F 50 mm
× 400 mm), a DC controller, a gas supply control system, and
a rotary pump system. The details of such an apparatus can be
found in (10).
2. Temperature detector: infrared radiation thermometric indi-
cator (Thermalert TX, Raytek).
3. Precursor gas: SnH4 mixed with N2 at a fixed ratio of 8% and
stored in an air pocket (SY-42, Sanhe Medical Instrument
Co. Ltd).
4. Substrates: n-type Si (100) wafers (20 mm × 5 mm × 0.3 mm).
5. Catalyst: 5-nm Au film was deposited on the Si substrates by
means of ion sputtering using Au as the target.
6. Ultrasonic system (Branson 2510).
7. Cleaning solution: acetone (Analytical Reagent from Alfa
Aesar) and ethanol.
8. Deionized (DI) water.
2.2. Fabrication 1. Equipment for patterning of Au electrodes: hotplate, Solitec

of Microelectrodes Spin Coater, Karl Suss MJB-3 Mask Aligner, CVC E-beam
evaporator.
2. Materials for patterning of Au electrodes: Au (99.99% purity),
AZ5214 photoresist, strip 315 developer, acetone, ethanol,
deionized water.
2.3. Single Nanowire 1. Probe station including two tungsten probes, X–Y stage, and
Microelectrode microscopy (Cascade Microtech, Inc., Beaverton, OR).
Fabrication 2. Function generator (Stanford research DS345, Sunnyvale, CA).
A Single SnO2 Nanowire-Based Microelectrode 113
3. 50-mL glass bottle.

4. Scalpel.
5. Wire bonder (West Bond 7476D079, West Bond Inc.,
Anaheim, CA).
6. Gold wire (F 25 mm).
7. Ethanol.
8. Platinum.
9. Pt deposition: FEI Nova Nanolab 200 FIB/SEM.
10. Support chip.
3. Methods
3.1. Synthesis 1. Clean n-type Si (100) wafers (20 mm × 5 mm × 0.3 mm) (see

of SnO2 Nanowires Note 1) with acetone, ethanol, and DI water sequentially for
5 min each in an ultrasonic bath.
2. Deposit 5 nm Au on the top surface by thermal evaporation
process. The gold was used as catalyst for the growth of SnO2
nanowires.
3. Prepare the precursor SnH4 by mixing with N2 at a fixed ratio
of 8% and store in an air pocket.
4. Place the Si substrates at the center of the horizontal quartz
reaction chamber.
5. Pump down the reaction chamber to 1–2 Torr using a rotary
pump.
6. Heat the Si substrates from room temperature to 770°C
within several seconds by adjusting the DC through the Si
substrates. The temperature of the Si substrates was mea-
sured using an infrared radiation thermometric indicator
(see Note 2).
7. Introduce the SnH4/N2 mixture to the reaction chamber
with the flow rate of 20 sccm (see Note 3). The deposition
usually takes 5–10 min.
8. Cool the reaction chamber to room temperature naturally.
3.2. Characterization 1. Characterize the structure of the SnO2 nanowires by XRD

of SnO2 Nanowires (Panalytical X-pert) using Cu Ka radiation (see Fig. 1).
2. Characterize the morphology of the SnO2 nanowires using
field-emission scanning electron microscopy (FESEM) and
HRTEM equipped with energy dispersion spectroscopy
(EDS) (see Fig. 2 and Note 4).
114 Zhou et al.
JCPDS:041-1445
(110)
Intensity (a.u.)
(101)
(211)
(200)
(220)
(310)
(210)
(301)
(112)
(111)
(002)
20 30 40 50 60
2Theta (degree)
Fig. 1. XRD pattern of the as-synthesized SnO2 nanowires. The diffraction peaks agree
well with that of bulk SnO2 of rutile structure (JCPDS: 041-1445). Reproduced from ref.
9 with permission from the American Chemical Society.
Fig. 2. (a) Typical SEM image of SnO2 nanowires. (Top right ) TEM image and (bottom
right ) corresponding SAED pattern of a single SnO2 nanowire. Reproduced from ref. 9
with permission from the American Chemical Society.
3. Characterize the crystalline quality and presence of the defect

structure of the SnO2 nanowires using a cathodolumines-
cence (CL) system at room temperature (see Note 5).
3.3. Fabrication of 1. Ultrasonically clean the SiO2/Si wafer using acetone, etha-
Microelectrodes nol, and DI water for 5 min each in sequence then blow dry
the wafer with pure nitrogen gas (see Note 6).
2. Spin coat the photoresist onto the wafer at f376 ´ g (4,000
rpm) for 40 s then bake the film at 80°C for 2 min.
3. Pattern the photoresist coated wafer using a Karl Suss MJB-3

Mask Aligner (UV patterning).
4. Develop the patterned photoresist using strip 315 developer
for 2 min.
5. Deposit 50 nm of Au on the patterned substrate using a CVC
E-beam evaporator.
6. Remove the photoresist by immersing the substrate into ace-
tone for 10 min.
3.4. Fabrication of a 1. Pour 10 mL ethanol into the glass bottle. Scratch the
Single SnO2 Nanowire as-synthesized SnO2 nanowires from the substrate into the
Microelectrode bottle using a scalpel. Then, ultrasonicate the sample for 15 min
to disperse the nanowire bundles into individual nanowires
(see Note 7). Finally, the SnO2 nanowire suspension is formed.
2. Place the Au microelectrode on the plate of the probe station.
By adjusting the X–Y stage, contact the two tungsten probes
to the two 500 mm × 500 mm Au pads, which are 10 mm
apart.
3. Add a 1-mL droplet of SnO2 nanowire suspension (nanowire
concentration ~1 mg/mL) between the two electrodes.
4. Turn on the function generator and apply a 5-V and 1-MHz
AC signal between the two tungsten probes (see Note 8).
5. Allow the solution to dry, then deposit two Pt
(4 mm × 1 mm × 0.2 mm) pads by FIB on the two ends of the
SnO2 nanowire on the Au microelectrodes to improve the
electrical contact (see Note 9 and Fig. 3a).
6. Place the device onto a supporting chip and connect the two
electrodes to the supporting chip by Au wire bonding (see
Fig. 3b).
Fig. 3. (a) SEM image of a single SnO2 nanowire which was fixed on the Au microelec-
trodes by Pt pads that were deposited using FIB. (b) Optical image of a real device con-
nected on a support chip. Reproduced from ref. 9 with permission from the American
Chemical Society.
116 Zhou et al.
4. Notes
1. These wafers can be used as substrates to grow SnO2 nano-

wires since their melting points are higher than 770°C.
2. For this application, 770°C is an ideal temperature for the
growth of high-quality SnO2 nanowires. When the deposition
temperature is less than 700°C, the grown SnO2 nanowires
are too large in diameter (around 300–500 nm). When the
deposition temperature is over 850°C, the grown SnO2 nano-
wires are too long and easily twisted with each other.
3. The chemical reaction of this process is as follows:
SnH4 + 2O2 → SnO2 + 2H2O.
4. Typically, the diameter and length of the SnO2 nanowires
range from 50 to 300 nm and up to tens of micrometers,
respectively (Fig. 2a). The TEM image (Fig. 2b) shows that
there is an Au nanoparticle on the tip of each SnO2 nanowire,
revealing the vapor–liquid–solid (VLS) growth mechanism.
The selective area electron diffraction (SAED) pattern
(Fig. 2c) indicates that the SnO2 nanowires are single crystal-
line with a [001] growth direction.
5. The near band edge emission, which was expected around
320 nm, was not detected. But, a broad blue luminescent
peak centered at approximately 470 nm was detected, indi-
cating that the as-synthesized SnO2 NWs possess a large num-
ber of oxygen vacancies in the crystal.
6. It is critical to keep the wafer wet when moving it from one
solution to another, otherwise residue will attach to the wafer,
which will be difficult to remove.
7. The ultrasonication time should not be longer than 15 min.
Otherwise, the SnO2 nanowires will break into short pieces
and then be unable to bridge the gap between the Au
electrodes.
8. This signal generates an alternating electrostatic force on the
SnO2 nanowires in the solution. Under the electrical polariza-
tion force, the SnO2 nanowires are placed in contact with the
two electrodes. This method is called the “dielectrophoresis
technique.” By precisely controlling the concentration of
the SnO2 nanowires in the solution, a circuit can be made
where only a single SnO2 nanowire bridges the two electrodes.
The 5-V and 1-MHz AC signal is optimized for fabricating
single nanowire devices by the dielectrophoresis method.
9. The deposition current should be very low (~10 pA).
Otherwise the whole SnO2 nanowire will be contaminated by
the Ga vapor and cause the device to short circuit.
Acknowledgments
The work was partially sponsored by the NSF, DARPA, and

the NIH.
References
1. Cui, Y., Wei, Q. Q., Park, H. K., and Lieber, 6. Kong, J., Franklin, N. R., Zhou, C. W.,
C. M. (2001) Nanowire nanosensors for Chapline, M. G., Peng, S., Cho, K. J., et al.
highly sensitive and selective detection of bio- (2000) Nanotube molecular wires as chemical
logical and chemical species. Science 293, sensors. Science 287, 622–625.
1289–1292. 7. Qi, P. F., Vermesh, O., Grecu, M., Javey, A.,
2. Hahm, J. and Lieber, C. M. (2004) Direct Wang, Q., and Dai, H. J. (2003) Toward large
ultrasensitive electric detection of DNA and arrays of multiplex functionalized carbon nan-
DNA sequence variations using nanowire otube sensors for highly sensitive and selective
nanosensors. Nano Lett. 4, 51–54. molecular detection. Nano Lett. 3, 347–351.
3. Patolsky, F., Zheng, G. F., Hayden, O., 8. Kim, I., Rothschild, A., Lee, B. H., Kim, D.
Lakadamyali, M., Zhuang, X. W., and Lieber, Y., Jo, S. M., and Tuller, H. L. (2006)
C. M. (2004) Electrical detection of single Ultrasensitive chemiresistors based on elec-
viruses. Proc. Natl. Acad. Sci. U.S.A. 101, trospun TiO2 nanofibers. Nano Lett. 6,
14017–14022. 2009–2013.
4. Zheng, G. F., Patolsky, F., Cui, Y., Wang, W. 9. Kuang, Q., Lao, C. S., Wang, Z. L., Xie, Z.
U., and Lieber, C. M. (2005) Multiplexed X., and Zheng, L. S. (2007) High-sensitivity
electrical detection of cancer markers with humidity sensor based on a single SnO2 nano-
nanowire sensor arrays. Nat. Biotechnol. 23, wire. J. Am. Chem. Soc. 129, 6070–6071.
1294–1301. 10. Kuang, Q., Jiang, Z. Y., Xie, Z. X., Lin, S.
5. Patolsky, F., Timko, B. P., Yu, G. H., Fang, Y., C., Lin, Z. W., Xie, S. Y., et al. (2005)
Greytak, A. B., Zheng, G. F., et al. (2006) Tailoring the optical property by a three-
Detection, stimulation, and inhibition of neu- dimensional epitaxial heterostructure: a case
ronal signals with high-density nanowire tran- of ZnO/SnO2. J. Am. Chem. Soc. 127,
sistor arrays. Science 313, 1100–1104. 11777–11784.
Chapter 9
Biosensing Using Nanoelectromechanical Systems

Ashish Yeri and Di Gao
Abstract
Nanoelectromechanical systems (NEMS) correlate analyte-binding events with the mechanical motions
of devices in nanometer scales, which in turn are converted into detectable electrical or optical signals.
Biosensors based on NEMS have the potential to achieve ultimate sensitivity down to the single-molecule
level, provide rapid and real-time detection signals, be operated with extremely low power consumption,
and be mass produced with low cost and high reproducibility. This chapter reviews fundamental concepts
in NEMS fabrication, actuation and detection, and device characterization, with examples of using NEMS
for sensing DNA, proteins, viruses, and bacteria.
Key words: Biosensors, Microelectromechanical systems, Micromachining, Microfabrication,

Nanofabrication, MEMS
1. Introduction:
Evolution of MEMS
to NEMS for
Biosensing The development of the atomic force microscope (AFM) by
Binnig, Quate, and Gerber in 1986 gave rise to the field of force
spectroscopy whereby intermolecular forces such as van der Waals
forces, Casimir’s forces, dissolution forces in liquid, and even
single-molecule rupture forces on the order of piconewtons could
be measured by a tiny mechanical cantilever. When applied to the
study of biomolecules, force spectroscopy can be used to measure
the attractive forces between biomolecules and the discrete steps
in the rupture of these bonds. Inspired by AFM, researchers
started to use microelectromechanical systems (MEMS), where
the displacement of a mechanical component is driven and sensed
by electronic signals, for biosensing. Using MEMS devices, the
presence of a biological analyte can be detected either based on
the deflection of a mechanical component such as a cantilever,
where analyte binding to a “functionalized” cantilever produces a
119
120 Yeri and Gao
deflection of the cantilever beam, or from the change of the resonant

frequency of a mechanical component in the resonant mode,
where analyte binding causes a decrease in the resonant frequency
of vibration of the cantilever beam (1).
Although MEMS biosensors have been well studied for over a
decade, they are plagued by a number of problems. For example,
cantilever deflection-based MEMS devices require a large amount
of molecules to be bound by the device to transduce the change
into measurable quantities. Also, these devices are unable to respond
to the rapidly changing forces of biomolecule interactions that have
time scales on the order of microseconds; this could be critical
when trying to distinguish between specific and nonspecific inter-
actions. These and other factors including the potential for increased
sensitivity, reduced power consumption, and reduced cost (2) have
motivated the reduction in size of existing MEMS devices.
Miniaturization also might allow these devices to be implanted in
the human body for medical diagnostic applications.
2. NEMS
Fabrication
2.1. Top-Down The fabrication processes of NEMS biosensors can be broadly
Approach classified into two categories: top-down and bottom-up
for Fabrication approaches. The top-down approach is directly borrowed from
of NEMS the microfabrication technologies. The only difference between
top-down fabrication processes on micrometer and nanometer
scales is the lithography process. For fabrication of MEMS,
photolithography using either visible or ultraviolet light is typically
employed to pattern micrometer-scale device features; for fabrication
of NEMS, more advanced lithography techniques, using electron
beams, focused ion beams, atomic force microscope tips, or pre-
synthesized nanostructures, for example, are employed to define
the device features.
The two most common top-down microfabrication tech-
niques are bulk micromachining and surface micromachining. In
bulk micromachining, fabrication of mechanical elements is
achieved by etching the substrate. Figure 1a presents a schematic
of a typical bulk micromachining process. The structural material
such as Si3N4, SiC, or GaAs is deposited via chemical vapor depo-
sition (CVD), for instance, onto the substrate and then patterned
by lithography. The mechanical elements then are released from
the substrate by removing the underlying substrate material. In sur-
face micromachining, a sacrificial material is deposited and pat-
terned between the structural material and the substrate (see
Fig. 1b). Instead of etching away the substrate material, the sacri-
ficial layer is then removed by etching to release the structure.
Advanced and innovative lithography techniques are required
to transfer the top-down approaches used in microfabrication to
Biosensing Using Nanoelectromechanical Systems 121
a b
Substrate
Substrate
Sacrificial layer
Passivating layer
Mask layer
Substrate
Device layer
Substrate
Device layer
Lithography to Sacrificial layer
define the structure Mask layer Passivating layer
Device layer Substrate
Substrate
Device layer
Lithography to define device layer
Etching of device layer Sacrificial layer
Mask layer
Passivating layer
Device layer
Substrate
Substrate
Removing the sacrificial layer to release the device
Removal of photoresist and wet etching of the substrate Free standing mechanical
Free standing mechanical structure
structure Passivating layer
Substrate
Substrate
Fig. 1. Basic depiction of (a) bulk micromachining and (b) surface micromachining processing steps leading to suspended
devices.
nanofabrication strategies. Electron beam (e-beam) lithography,

which is capable of defining features of almost any geometry down
to tens of nanometers in size, is the most commonly employed
technique to fabricate NEMS using the top-down approach.
Figure 2 shows an example where e-beam lithography is used to
fabricate NEMS devices consisting of a Fabry–Perot interferometer
and a doubly clamped beam array on a silicon-on-insulator (SOI)
substrate (3). After coating the silicon with poly(methyl meth-
acrylate) (PMMA), e-beam lithography was used for patterning.
Aluminum was evaporated (35 nm) on the resist as a mask for CF4
reactive ion etching (RIE), and then the silicon oxide was etched
using buffered hydrofluoric acid. The thickness of the mechanical
element and the distance between the mechanical element and the
substrate are defined by the thickness of the silicon layer and
the thickness of the insulator layer of the SOI substrates, respec-
tively. The lateral dimension of the mechanical element is defined by
e-beam lithography. In another example, the critical dimension of
the mechanical element in NEMS is defined by presynthesized
nanostructures such as nanofibers. Figure 3 shows such an example,
where a silicon nitride nanowire less than 200 nm wide is fabri-
cated using a PMMA nanofiber synthesized by electrospinning
(4). After patterning the electrodes, the nanofiber is deposited
between them, serving as a mask to define the width of the silicon
nitride nanowire during the etching process.
2.2. Bottom-Up In the bottom-up approach, the mechanical element of the NEMS
Approach (e.g., nanowires, nanotubes) is synthesized through chemical
for Fabrication reactions, and the dimension of these nanostructures is defined
of NEMS during the synthesis process instead of by lithography techniques.
a Resist patterning RIE Etch Aluminum lift-off Oxide Etch
PMMA
Silicon
Oxide
Silicon
b c
5 mm 5 mm
Fig. 2. Example of NEMS device fabrication using e-beam lithography through the top-down approach. (a) Schematic of
fabrication steps. (b) A 200 nm thick square Si pad suspended with silicon nanowires 100 nm wide. The distance between
the pad and the silicon substrate is 400 nm. (c) 50 nm thick and 200 nm (left) and 120 nm (right) wide suspended Si
nanowires of lengths varying from 7 to 16 mm. Reprinted with permission from ref. 3.
i ii iii
Si3N4
SiO2
Si
iv v vi
300 nm
Fig. 3. Example of NEMS device fabrication using presynthesized nanostructures. A resonator based on a silicon nitride
nanowire is fabricated using a polymer nanofiber synthesized by electrospinning. (a) Schematic of the processing steps.
(b) SEM images of a 15 mm long, less than 200 nm wide, doubly clamped silicon nitride nanowire. Reprinted with permis-
sion from refs. 6 and 4, respectively.
In order for the NEMS device to function, these nanostructures

must be positioned between electrodes which are used to drive
and sense their mechanical motion. These electrodes are typically
on the micrometer scale and fabricated by micromachining pro-
cesses. It is a challenging task to position nanostructures between
micrometer-sized electrodes; it is typically accomplished using
one of the following two routes.
The first route involves placing presynthesized nanostructures
on prefabricated electrodes, either by random dispersion or by
using micromanipulators, and then welding the nanostructure to
the electrode by e-beam lithography or focused ion beam deposi-
tion. This route has been widely used for fabricating electronic
sensors based on 1D nanomaterials, where there is no mechanical
motion of the nanostructures. However, because the anchor, by
which the nanostructure is attached to the electrode, is not rigid
enough for high-frequency resonators, this technique is rarely
used for fabrication of NEMS devices. A variant method of this
approach is to calcinate the presynthesized nanostructure after it
is placed on the electrode, during which the chemical composi-
tion of the nanostructure may change. Figure 4 shows such an
example, where a silica nanofiber with a resonant frequency of ~
10 MHz is made by first placing a polymer nanofiber, presynthe-
sized by electrospinning, on top of the electrodes and then calci-
nating the fiber at 850°C (4–6).
The second route involves direct chemical synthesis of nano-
structures at a specific location on the device, preferably in a pre-
determined direction and architecture (7, 8). The advantages
of this route are that the construction of the NEMS devices is
simplified, and the arduous task of postsynthesis assembly and the
a Composite Nanofiber b
Silicon Chip
Electric Field
Rotating
Direction
Rotating Counter electrode
SiO2 Nanofiber
10 µm
Fig. 4. Example of NEMS device fabrication by placing presynthesized nanostructures onto prefabricated electrodes
through the bottom-up approach. (a) Diagram showing steps involved in fabricating a silica nanofiber with a resonant
frequency of ~10 MHz by first depositing a polymer nanofiber synthesized by electrospinning on top of the electrodes and
then calcinating the fiber at 850°C. (b) SEM image of a silica nanofiber anchored to the electrodes. Reprinted with per-
mission from ref. 5.
124 Yeri and Gao
Fig. 5. Scanning electron microscopy (SEM) image of a Si nanowire resonator grown by

chemical vapor deposition. The doubly clamped device is suspended over a trench fab-
ricated on a silicon-on-insulator (SOI) substrate. Reprinted with permission from ref. 9.
potential deterioration of the nanostructure during the process

are eliminated (7). Figure 5 shows an example of NEMS fabri-
cated through this route, in which a doubly clamped single crys-
talline Si nanowire is suspended between two microelectrodes
(9). The microelectrodes are fabricated by top-down microfabri-
cation techniques, while the Si nanowire is directly synthesized
between the electrodes by a metal-catalyzed CVD process, which
is a bottom-up approach. The location and the diameter of the
nanowire are controlled by the location and size of the Au cata-
lytic nanoparticle, which is deposited onto the sidewall of one
electrode prior to the growth of the wire.
3. NEMS Actuation
and Detection
NEMS devices convert electrical inputs to the sensor into mechanical
motion and vice versa. Similar to MEMS devices, a typical NEMS
device consists of an input transducer and an output transducer;
the input transducer converts the electrical inputs into mechanical
motion and the output transducer senses the motion or displace-
ment of the sensor and converts it into electrical or optical signals.
Although the configuration of NEMS is similar to that of MEMS,
actuation and detection methods used for MEMS devices do not
always work well with NEMS devices. In some cases, optical tech-
niques are not compatible with the small length scale of NEMS
devices; in others, parasitic impedances make electrical detection
challenging. This section summarizes some of the actuation and
detection methods that have been reported in the literature.
3.1. Actuation Methods When a current is passed through a NEMS resonator, it induces
surface heating and thermal expansion of the structural material.
3.1.1. Thermal Actuation
The NEMS device may be actuated by this process utilizing
customized structural designs (e.g., by using a bilayer structure

consisting of two layers of materials with different thermal expan-
sion coefficients). Superimposing an AC voltage onto a constant
DC voltage leads to sinusoidal heating of the mechanical structure.
Because of the very small dimension of the NEMS device, the
time scale of the heating and cooling of the mechanical structure
could be small enough (on the order of ns to ps) for actuation of
high-frequency NEMS resonators. When it is not convenient to
use the resonator itself as the heating source, thermal actuation
may also be carried out by fabricating on-chip Joule-heating resis-
tors near the resonators (8).
3.1.2. Magnetomotive When an alternating current i(w) is passed through the NEMS
Actuation cantilever beam in the presence of a strong magnetic field B,
the beam experiences a Lorentz force of f (w ) = lBi (ω ) , where
l is the length of the beam. The direction of the force is perpen-
dicular to both the AC and the magnetic field. Cleland et al. (10)
have demonstrated magnetomotive actuation of a doubly
clamped beam; AC is passed through the beam and a magnetic
field perpendicular to the direction of the AC induces out-
of-plane beam vibrations.
3.1.3. Piezoelectric Piezoelectric actuation can be carried out by either exciting

Actuation the NEMS device using an external (off-chip) piezoelectric
device or embedding piezoelectric materials into the NEMS
device. Off-chip piezoelectric actuators have been used to
detect thiolated SAMS (11) and viruses (12), but such actua-
tors typically have a low Q-factor (see Subheading 4.2.2), and
the requirement of an external piezoelectric device is undesir-
able for miniaturized devices. Direct piezoelectric actuation
can be accomplished by embedding the piezoelectric material
inside the mechanical structure. For example, Lee et al. have
actuated Si3N4 cantilever beams at frequencies on the order of
104 Hz with very high mass sensitivities (13, 14) by depositing
lead zirconate titanate (PZT, a piezoelectric material) on
them.
3.1.4. Electrostatic When a voltage is applied across two plates of a capacitor, an

Actuation attractive force develops between them. This attractive force
increases nonlinearly as the separation between the plates decreases.
Since the separation between the plates in MEMS and NEMS
devices is very small, the electrostatic force is sufficiently large to
actuate the mechanical elements of the devices. For example, a
doubly clamped beam can be actuated by applying a potential
between the beam and a nearby gate electrode (15); an out-of-
plane resonance of a single crystalline Si mesh suspended by nano-
beams has been induced by an electric force perpendicular to the
plane (16).
126 Yeri and Gao
3.1.5. Other Actuation Actuation can also be carried out by thermal noise and the fluctuations
Methods in ambient air. Ilic et al. has reported detection of Escherichia coli
in both air and vacuum (17, 18) using the transverse vibrations
resulting from thermal noise. This technique is simple, requiring
no external oscillators or complex fabrication techniques. A can-
tilever beam also has been excited at its resonance frequency by
heating from a laser source (19). Here, a 670-nm diode laser was
employed to actuate a cantilever beam, and optical interferometry
was used to detect its deflection.
3.2. Detection Methods Focused laser beams are commonly used to detect the deflection
of cantilevers in MEMS sensors. When the analyte of interest
3.2.1. Optical Detection
binds to the cantilever and produces a deflection, a laser beam
focused on the cantilever detects this deflection. This simple
detection scheme does not scale down well to NEMS devices due
to the diffraction of the laser at these smaller length scales. Another
common optical detection technique is based on interferometry,
where interference of light is used to sense very minute changes
in deflection of the mechanical element in the NEMS device. For
example, Michelson interferometry-based methods use the inter-
ference of the laser beam reflected from the structure with a stable
reference laser beam; Fabry–Perot interferometry-based methods
use the refractive index difference between the structural material
of the NEMS device and the substrate (3, 20, 21). Although
interferometry is able to detect subwavelength movement of
mechanical structures, the strong diffraction effects associated
with the very small dimensions of NEMS devices are still prohibi-
tive in making optical interferometry applicable for detecting
motions of NEMS devices. In addition, micrometer-scale optic
fiber cables are generally used for detection with MEMS devices;
these would be too large for use with NEMS devices. Thus, when
faced with the practical problems associated with focusing lasers
onto subwavelength structures, the high power losses associated
with scattering by surface defects (22), and the troublesome posi-
tioning of the optic fiber, optical detection systems are limited for
use with NEMS devices.
3.2.2. Electrostatic The deflection of a charged resonator, acting as one of the plates
or Capacitive Detection of a capacitor, causes a change in the capacitance, which is
inversely proportional to the distance from the substrate. This
change of capacitance can be detected by custom-designed cir-
cuitry. This method offers very high detection sensitivity for both
MEMS and NEMS devices. One problem associated with capaci-
tive detection though is the parasitic capacitance, which is in par-
allel to the actuating capacitance; this reduces the efficiency of
detection at high frequencies. Balanced bridge techniques (15),
for example, can be used to overcome the effects of parasitic
capacitance. Capacitive detection schemes are very sensitive to
the dielectric constant of the medium, and changes in this param-

eter can have unfavorable effects.
3.2.3. Piezoresistive Piezoresistive materials, such as doped Si, undergo changes in

Detection electrical resistance when a strain is applied on them or when
mechanical motion imposes a strain. Piezoresistive detection is
usually carried out with a Wheatstone network (e.g., with two of
four resistors placed on a cantilever). Piezoresistive materials have
great potential to be used in NEMS devices because such devices
can be operated without the aid of external equipment, can be
made in reduced size, and have the ability to operate in opaque
liquids such as blood. However, care must be taken to prevent
overheating of the device due to Joule heating, which would
cause undesirable mechanical deformations.
4. NEMS Based
on Cantilever
and Doubly
Clamped Beams Beams with at least one dimension in the nanometer scale, including
both cantilevers (with one end anchored) and doubly clamped
(with both ends anchored) beams, have simple geometries and
are relatively easy to fabricate and model. Therefore, they are
widely used as mechanical elements in NEMS sensors. This sec-
tion introduces some fundamental concepts for cantilever- and
doubly clamped beam-based NEMS.
4.1. Cantilever Beams The deflection of cantilever beams upon deposition of metals was
Operated in Deflection first studied by G.G. Stoney (23) who gave the relationship
Mode (known as Stoney’s formula) between the radius of curvature of
the beam (R) and the surface stress (s) as
Et 2
R= , (1)
6s (1 − u )
where E is the Young’s modulus, t is the thickness of the sub-
strate, and n is the Poisson’s ratio of the substrate. The same prin-
ciples are applied to the deflection of cantilever beams used as
sensors. Analyte binding to the surface of the cantilever beam
causes a surface stress which decreases the radius of curvature of
the beam (see Fig. 6a). Typically, only one side of the cantilever
beam is functionalized to trap the analyte, which causes the beam
to bend in one direction. Although, it is debated if the surface
stress s from analyte binding is the sole cause of the bending of
the cantilever. Also, the location of analyte binding could in fact
increase the stiffness of the material, causing the radius of curva-
ture to increase. Cantilever beams operated in the static mode or
deflection mode have the advantage that they can be operated in
liquid media without a significant loss of sensitivity (see Note 1).
128 Yeri and Gao
Fig. 6. Schematic showing the two different modes of operation for cantilever-based
biosensors. The cantilever is functionalized on only the top side with biomolecules that
capture the target molecule. (a) Deflection-based sensing: the cantilever on the left is the
reference cantilever and analyte binding is shown on the right; the deflection with respect
to the reference is due to the surface stress induced by the analyte binding. (b) Resonance-
based sensing: the shift in the resonance frequency of the cantilever on the right with
respect to the reference on the left is related to the analyte binding.
4.2. Cantilever NEMS resonators typically can achieve frequencies on the order
and Doubly Clamped of MHz with some in the GHz range. By treating a NEMS reso-
Beams Operated nator as a harmonic oscillator, its resonant frequency (f0), in the
in Resonant Mode absence of external forces and damping, is given by:
4.2.1. Frequency 1
f0 = , (2)
2p k / meff
where k is the spring constant and meff is the effective mass of the
cantilever. The spring constant is proportional to the flexural
rigidity, D, of the resonator, which is given by the product of the
Young’s modulus, E, and the moment, I. Accordingly, the funda-
mental out-of-plane resonance frequency (f0) of a cantilever or a
doubly clamped beam has the following relationship to the geom-
etry and the material properties of the beam:
E t
f0 ∝ , (3)
r L2
where t is the thickness and L is the length of the beam and r is

the mass density. Based on Eq. 3, materials with a high ratio of E
to r are preferred for NEMS devices to obtain high frequencies.
To improve the electrical conductance of the beam, a metallic
layer is often added onto the semiconducting material. In such
cases, the above relationship becomes:
1 E1 I1 + E2 I 2
f0 ∝ , (4)
L2 r1 A1 + r2 A2
where the subscripts 1 and 2 refer to the semiconducting and

metallic layers, respectively.
Upon mass addition (Dm) to the resonator, f0 changes by Df, which
based on Eq. 2 can be written by a first order approximation as:
− Dm
Df = f0, (5)
2meff
assuming that Df is attributed only to Dm, k remains unchanged,

and the added mass is uniformly distributed over the beams.
According to Eq. 5, by measuring the resonant frequency before
and after loading, the mass of the analyte added to the resonator
may be calculated (see Fig. 6b).
Because their effective mass is very small, NEMS devices
have the potential to be very sensitive mass sensors. However,
the interpretation of experimental data could be very compli-
cated due to the fact that both the added mass and the binding
of analyte to the resonator may induce surface stress. Further,
the distribution of mass and the induced surface stress may not
be uniform across the resonator. The frequency shift associated
with the surface stress induced by analyte binding has been ana-
lyzed by treating it as a one dimensional axial force; however, this
treatment has been the source of much debate. When nonuni-
form binding occurs (e.g., the analyte binds only to the tips of a
cantilever), the relationship between the frequency shift and the
position of the bound analyte becomes very complex. Depending
upon the position of the adsorbed mass, the NEMS device could
experience a negative frequency shift (at the tip of the cantilever
beam) or a positive frequency shift (near the clamped end). This
result has been shown experimentally with adsorbed bacteria (see
Fig. 7) (24, 25) and proteins (26). A theoretical model was dis-
cussed by Tamayo et al. in ref. 27. When the analyte concentra-
tion in solution is very low, the resonant frequency shift may be
enhanced by the addition of mass labels in certain biosensor
applications.
4.2.2. Q-Factor The quality factor (Q-factor) is defined as the ratio of the reso-
nance frequency of the oscillator to the bandwidth. It is inversely
proportional to the energy dissipation factor D, which is defined
in terms of energy dissipated versus energy stored in the oscil-
lator as:
1 Energy dissipated per oscillation (6)

D= = .
Q 2p Energy stored in the oscillator
The Q-factor can be used to characterize the performance of

NEMS devices. When the NEMS devices are used as sensors, the
sensitivity of the sensor is directly determined by the Q-factor;
130 Yeri and Gao
Fig. 7. Dependency of frequency shift on the location of the mass addition on cantilever beams. (a) Optical micrographs
of bacteria deposited at three different locations on the cantilever (500 mm long, 100 mm wide, and 1 mm thick) which
are 390, 200, and 73 mm, respectively, away from the position of clamping (shown by dotted line). (b) Frequency
response of the cantilevers where the location of bacteria is as shown in (a). The dotted line is before the deposition of
bacteria and the continuous line is after bacteria adsorption. The measurement was performed in air and the resonant
frequencies are in the range of 7–8.5 kHz. The resonant frequency decreased by 4%, remained the same, and increased
by 3.6% after the deposition of bacteria near the end, near the center, and near the clamping, respectively. Reprinted with
permission from ref. 24.
the least mass (Dmmin) that could be detected with resonant sensors
is given by (28, 29):
B
Dmmin = 2meff , (7)
− DR / 20
wQ
where meff is the effective mass of the resonator, B is the measure-
ment bandwidth, w is the fundamental resonant frequency, and
DR is a dynamic range term.
Q-factors of NEMS devices strongly depend on the design of
the devices and process conditions. Typically, the Q-factors of
NEMS resonating devices are in the range of 103–105 in vacuum
(30). NEMS devices fabricated from silicon (mono or polycrys-
talline), GaAs, and SiC have all shown very high Q-factors. As the
dimensions of the resonator decrease, the Q-factor decreases
because the higher surface-to-volume ratio will cause more surface
losses (see Note 2). This enables the NEMS resonator to operate
at a very low power level with very high force sensitivity. Efforts
to increase the Q-factor of submicron resonators include high-
temperature annealing (31–33) and chemical treatments to allevi-
ate surface loss mechanisms due to the oxidation of the substrate.
Also, Q-factors may be improved by increasing the crystallinity
of the materials and generating high internal residual stress.
For example, devices fabricated from Si3N4 having an internal stress
of about 1,200 MPa show a Q-factor on the order of a million
(4, 28); a comparison between the doubly clamped nanowires

fabricated from the high-stress Si3N4 and low-stress Si3N4 shows
that the high stress may improve Q-factors.
4.2.3. Power Level A rough estimate of operating power levels can be given by the
ratio of the thermal energy to the characteristic time scale of energy
exchange between the mode and the surroundings at frequency,
wo. The thermal energy is given by kBT, where kB is the Boltzmann’s
constant and T is temperature. The time scale t is given by Q/wo.
Therefore, the minimum operating power level of NEMS devices
can be estimated by kBTwo/Q , which is on the order of aW to fw
(10−18 to 10−15 W) for devices with Q-factors of about 1,000 to
10,000 and resonance frequencies in the MHz range.
5. Biosensor
Examples Enabled
by NEMS (See
Note 3) Many biodetection strategies (e.g., for DNA, proteins, and cells)
rely heavily on fluorescent labels. Label-free detection techniques,
enabled by technologies such as NEMS, surface plasmon reso-
nance (SPR), and quartz crystal microbalances (QCM), have
been very useful not only as sensors but also as tools for the study
of the binding processes of biomolecules. Compared to SPR and
QCM, NEMS devices have additional advantages that they are
extremely small and have great potential to be mass-produced
with high reproducibility. In this section, we look into some of
the outstanding applications of NEMS devices for biosensing.
5.1. Biosensing Deflection-based cantilever beams have been utilized for the
by Cantilever-Based detection of cells, proteins, and DNA.
NEMS in the Deflection
Mode
5.1.1. Detection of Proteins Microcantilever-based detection of prostate specific antigen
(PSA), a tumor marker for prostate cancer, was demonstrated by
Mazumdar et al. (34, 35). Using gold-coated Si3N4 cantilever
beams (200 × 20 × 0.5 mm, L × W × T) and an optical detection sys-
tem which measured the deflection of the cantilever tip, they suc-
cessfully detected two forms of PSA at concentrations ranging
from 0.2 ng/mL to 60 mg/mL in the presence of human serum
albumin (HSA) and human plasminogen (HP), each at a concen-
tration of 1 mg/mL. Their work also showed that surface stress,
which is responsible for the deflection of the cantilever, is a func-
tion of the PSA concentration in the solution.
Arntz et al. (36) used an array of microcantilevers to detect
two cardiac markers, creatin kinase and myoglobin, at a sensitivity
of about 20 mg/mL. The antibodies to these cardiac markers were
immobilized onto the cantilever surface, and antigen–antibody
132 Yeri and Gao
binding was detected, as the absolute deflection compared to a

reference cantilever beam, by optical reflection. The reference
beam also was used to eliminate thermal noise and turbulence
which arose due to injection of the samples.
Also, Wee et al. (37) used a piezoresistive detection method
to detect PSA and C-reactive proteins (CRP), cardiac markers, at
a sensitivity of 10 ng/mL using deflection of cantilever beams.
Fluorescently tagged PSA was used to verify that the change in
cantilever resistance upon PSA binding was dependent only on
the concentration of the PSA in solution.
5.1.2. Detection of DNA Huber et al. (38) used gold-coated Si cantilever beams to study
the binding of two DNA transcription factors, with a sensitivity of
80–100 nM, to double-stranded DNA functionalized onto the
beam surface.
Fritz et al. (39) have been able to differentiate between the
binding of a 12-mer single-stranded DNA (ssDNA) and a 12-mer
ssDNA with a single base mismatch to the complementary ssDNA
strand tethered to gold-coated Si cantilever beams, where the
detection was made using optical reflection techniques. McKendry
et al. (40) have demonstrated rapid DNA detection based on
hybridization of a target DNA to strands on a cantilever beam.
Here, only one of the cantilever beams out of a total of eight is
functionalized with the complementary base sequence with
respect to the analyte ssDNA. The resultant deflection of this
beam was monitored by optical reflection, and the absolute deflec-
tion due to analyte binding was noted by subtracting the deflections
of the other beams which were treated as reference beams.
The detection of a single-nucleotide polymorphism (SNP)
was also demonstrated by Mukhopadhyay et al. (41) where the
surface stresses resulting from hybridization of a fully comple-
mentary strand to a probe immobilized onto the cantilever sur-
face was compared to the hybridization of a strand with a single
base mismatch. In this case, the resistance change of the piezore-
sistive cantilever upon binding of the analyte was used to measure
the deflection of the cantilever beam.
5.2. Biosensing The binding of thiolated self-assembled monolayers (SAMs) on

by Cantilever-Based gold-coated silicon cantilevers has been detected with a sensitivity
NEMS in the Resonant of 5.5 fg by using photothermal actuation of the cantilever beam
Mode and optical interferometry to measure the resonant frequency
(19). Ilic et al. have detected thiolated SAMS on a gold pad
5.2.1. Detection of SAM located near the cantilever tip and studied the effect of spatial
Coating mass loading on the decrease in its resonant frequency, enabling
an accurate determination of the mass bound (11).
5.2.2. Detection of Proteins PSA was also detected electrically with a sensitivity of 10 pg/mL
(42) in air and of 1 ng/mL (13) in liquid by Kim and coworkers.
Piezoelectric material (PZT) 0.5 mm thick was sandwiched
between two Pt electrodes and was resonated with an AC voltage,

which eliminates the use of an external oscillator. Similarly, Kwon
et al. (13) detected activated cyclic adenosine monophosphate
(cyclic AMP)-dependent protein kinase (PKA), with a sensitivity
of 6.6 pM, with a fragment of PKI protein, PKI (5–24), immobi-
lized onto the piezoelectric cantilever beam. Protein kinase is an
enzyme which modifies other proteins, regulating their activity
and coordinating central cellular processes. The measurement of
C-reactive protein which provides a good indicator for the pre-
diction of heart attack and stroke in postmenopausal women was
also done by Lee et al. (43) on a similar format employing a PZT
embedded cantilever.
Burg et al. (44) used a suspended microfluidic channel in
place of a regular cantilever to address the problem that NEMS
devices have low Q-factors when operated in fluids (see Fig. 8).
The biomolecules bind on the inner walls of the suspended
microchannel where the microchannel itself is resonated in air or
vacuum. This way, Q-factors of the order of 15,000 are achieved
with zeptogram mass sensitivity in vacuum. The thickness of the
microchannel walls and the fluid layer was 800 nm and 1.2 mm,
respectively. A continuous delivery of analyte solution was made
possible by a polydimethylsiloxane (PDMS) microfluidic network.
The device is actuated electrostatically, and the deflection of the
microchannel is measured by optical reflection. Biotinylated
a b
c buffer avidin bBSA avidin
0
Frequency shift (Hz)
−1
−2
−3
−4
0 2 4 6 8 10 12 14
Time (minutes)
Fig. 8. Embedding microfluidic channels in a suspended resonant cantilever for biosensing.

(a) Schematic showing the flow direction in the suspended microchannel. (b) Schematic of
the cross-section of the microchannel showing analyte binding within the microchannel.
The cantilever with the microchannel is resonated in vacuum. (c) Frequency decrease of
the resonating microchannel in vacuum as a function of time due to binding of avidin,
biotinylated BSA, and avidin in sequence. Reprinted with permission from ref. 44.
134 Yeri and Gao
bovine serum albumin (bBSA) (44) was detected by functionalizing

the surface of the microchannel with avidin. Then, bBSA was
flowed through the channel followed by avidin again (see Fig. 8).
With this system operated in air, the surface mass sensitivity was
10−17 g/mm2. The detection of goat anti-mouse IgG (45) was
also performed in a microchannel where biotinylated anti-goat
antibodies were linked to the surface through neutravidin on a
layer of poly(ethyleneglycol)-biotin grafted poly-l-lysine (PLL-
PEG-biotin). Real-time binding of goat anti-mouse IgG was
observed for concentrations ranging from 0.7 nM to 0.7 mM.
A distinct advantage of this geometry is the ability to weigh mol-
ecules that pass through the microchannel without binding.
5.2.3. Detection Ilic et al. (46) detected a 1,587 bp dsDNA using an array of reso-
of Double-Stranded DNA nant NEMS cantilever beams. Si3N4 beams 90 nm thick and between
3.5 and 5 mm long were excited, and their resonant frequencies
were detected by optical interferometry (see Fig. 9). A gold
nanodot was deposited on the free end of the cantilever to which
the thiolated dsDNA binds, providing a reliable calibrated
response of frequency decrease with mass adsorption (since the shift
in resonant frequency is dependent on the location of binding). The
binding of the dsDNA was performed in solution, and the resonant
frequencies were noted before and after the binding in vacuum
(3 × 10−7 Torr). Q-factors in the range of 104 were reported, and the
mass of a single dsDNA molecule 1,587 bps long was detected.
5.2.4. Detection of Viruses Resonating silicon cantilever beams of nanoscale thickness were
used to detect Vaccinia virus (average mass ~9.5 fg) by thermal
and ambient noise actuation of the beams (47, 48). The detection
of the resonant frequency was made by using a microscope scan-
ning laser Doppler vibrometer.
The detection of an insect baculovirus was performed by Ilic
et al. (11) using a Si3N4 cantilever beam 150 nm thick, 0.5 mm
wide, and 6, 8, or 10 mm long with paddles of 1 × 1 mm. The
beams were excited piezoelectrically, and signal transduction was
achieved by optical interferometry. The beams were immersed in
a solution of AcV1 antibody (antibody to the baculovirus)
followed by immersion in the virus solution; they then were dried
in nitrogen. The resonant frequencies were noted before and after
the immobilization of the virus in vacuum (4 × 10−6 Torr). In
calculating the mass adsorbed, it was assumed that there was no
change in the flexural rigidity and that there was a linear relation-
ship between the decrease in frequency and the mass adsorbed.
Mass sensitivities of the order of 10−15 g/Hz were reported, and
virus concentrations ranging from 105 to 107 pfu/mL were
detected. Control experiments determined that only about 3.5%
of the mass change resulted from nonspecific binding, showing
that the assay had high specificity.
1.0
1.0
Optical Detector Output (A.U.)
0.9
0.8
0.6
11.835 11.840
0.4
0.2
0.0
11.82 11.83 11.84 11.85 11.86
Frequency (MHz)
Fig. 9. Detection of dsDNA using an array of resonant cantilever beams. (a) SEM image showing the gold nanodot on
the cantilever tip 300 nm away from the free end. (b) Schematic of the thiolated dsDNA binding to the Au nanodot
(top). Schematic showing the actuating laser (415 nm) with the detecting laser (HeNe) in the foreground (bottom).
(c) Frequency spectra in vacuum of the cantilever beam before and after binding of approximately 2 dsDNA molecules.
Reprinted with permission from ref. 46.
5.2.5. Detection of Bacteria A highly sensitive detection of E. coli (17, 18) was performed
using Si3N4 cantilever beam of varying dimensions (see Fig. 10).
The beams were excited to their resonance frequencies by ther-
mal mechanical noise and ambient fluctuations in air and were
detected by an optical deflection system. The antibodies to E. coli
cells were bound to the cantilever beams by soaking them in solu-
tion and drying them. The resonance frequency was measured
before and after the beam was bound by the E. coli cells. Again, it
was assumed that there was a linear relationship between the
decrease in frequency and the mass adsorbed and that there were
no changes in the flexural rigidity of the beam. These assump-
tions are justified as the mass of E. coli adsorbed for varying cell
concentrations is linear with the frequency change. Escherichia
coli cells (ranging from 106 to 109 in number) were detected,
136 Yeri and Gao
Fig. 10. Detection of Escherichia coli using resonating cantilever beams. (a) Scanning electron micrograph of a single
E. coli O157:H7 cell bound to the immobilized antibody layer on top of the oscillator. Length and width of the cell were
estimated to be 1.43 mm and 730 nm and the thickness determined from tapping AFM to be 350 nm. (b) Measured
frequency shift as a function of the number of E. coli cells. Reprinted with permission from ref. 17.
showing that the NEMS device could be operated in air without

the use of external oscillator. The high dampening of oscillations
in air gave rise to a very low Q-factor as compared to that for the
operation of the device in vacuum, which is of the order of 104.
6. Notes
1. Most biosensing applications require that the sensor be operated

in a liquid environment. In the resonant mode, viscous damping
in liquid significantly reduces the Q-factor. When NEMS sen-
sors are operated in liquid media, issues concerning fluid flow
and mass transport also need to be taken into account.
2. Although size reduction of NEMS devices may lead to higher
frequencies and improvement in mass sensitivity, as we scale
down to the point where the size of the analyte becomes com-
parable to the size of the device, device modeling and experi-
mental data interpretation may become considerably complex.
3. Most NEMS sensors require that the surface of the device be
functionalized with a biomolecule. However, most bioreceptors
are chemically and structurally complex and may lose their
“biosensing” ability due to the conformation changes that
they undergo when immobilized onto substrates. Also, the
limited stability of the bioreceptors can adversely affect the
long-term storage possibilities and reusability of the biosensor.
In addition, functionalization of NEMS with biomolecules
could result in losses in Q-factor, which would be detrimental
to the success of the biosensor. For example, a common

way to attach biomolecules such as DNA and proteins to
surfaces is by thiol-chemistry, which requires the mechanical
sensing element to be coated with gold. Experimental studies
(49, 50) have shown that even a very thin layer of gold,
~100 nm thick, causes a considerable decrease in the Q-factor
of microcantilevers.
Acknowledgments
The authors acknowledge financial support from the National

Science Foundation (CBET 0747164).
References
1. Waggoner, P. S. and Craighead, H. G. (2007) 9. He, R., Feng, X. L., Roukes, M. L., and Yang,
Micro- and nanomechanical sensors for envi- P. (2008) Self-transducing silicon nanowire
ronmental, chemical, and biological detec- electromechanical systems at room tempera-
tion. Lab. Chip. 7, 1238–1255. ture. Nano Lett. 8, 1756–1761.
2. Hierold, C. (2004) From micro- to nanosys- 10. Cleland, A. N. and Roukes, M. L., (1996)
tems: mechanical sensors go nano. J. Micromech. Fabrication of high frequency nanometer scale
Microeng. 14, S1–S11. mechanical resonators from bulk Si crystals.
3. Carr, D. W. and Craighead, H. G. (1997) Appl. Phys. Lett. 69, 2653–2655.
Fabrication of nanoelectromechanical systems 11. Ilic, B., Craighead, H. G., Krylov, S.,
in single crystal silicon using silicon on insula- Senaratne, W., Ober, C., and Neuzil, P. (2004)
tor substrates and electron beam lithography. Attogram detection using nanoelectrome-
J. Vac. Sci. Technol. B 15, 2760–2763. chanical oscillators. J. Appl. Phys. 95,
4. Verbridge, S. S., Parpia, J. M., Reichenbach, 3694–3703.
R. B., Bellan, L. M., and Craighead, H. G. 12. Ilic, B., Yang, Y., and Craighead, H. G. (2004)
(2006) High quality factor resonance at room Virus detection using nanoelectromechanical
temperature with nanostrings under high ten- devices. Appl. Phys. Lett. 85, 2604–2606.
sile stress. J. Appl. Phys. 99, 124304.
13. Lee, J. H., Hwang, K. S., Park, J., Yoon, K.
5. Kameoka, J., Verbridge, S. S., Liu, H.,
H., Yoon, D. S., and Kim, T. S. (2005)
Czaplewski, D. A., and Craighead, H. G.
Immunoassay of prostate-specific antigen
(2004) Fabrication of suspended silica glass
(psa) using resonant frequency shift of piezo-
nanofibers from polymeric materials using a
electric nanomechanical microcantilever.
scanned electrospinning source. Nano Lett. 4,
Biosens. Bioelectron. 20, 2157–2162.
2105–2108.
6. Czaplewski, D. A., Verbridge, S. S., Kameoka, 14. Kwon, H., Han, K., Hwang, K., Lee, J., Kim,
J., and Craighead, H. G. (2004) Nanomechanical T., Yoon, D., et al. (2007) Development of a
oscillators fabricated using polymeric nanofiber peptide inhibitor-based cantilever sensor assay
templates. Nano Lett. 4, 437–439. for cyclic adenosine monophosphate-depen-
dent protein kinase. Anal. Chim. Acta 585,
7. Wu, Y., Cui, Y., Huynh, L., Barrelet, C. J., Bell,
344–349.
D. C., and Lieber, C. M. (2004) Controlled
growth and structures of molecular-scale silicon 15. Ekinci, K. L. (2005) Electromechanical trans-
nanowires. Nano Lett. 4, 433–436. ducers at the nanoscale: Actuation and sens-
8. Lange, D., Hagleitner, C., Hierlemann, A., ing of motion in nanoelectromechanical
Brand, O., and Baltes, H. (2002) systems (nems). Small 1, 786–797.
Complementary metal oxide semiconductor 16. O’Shea, S. J., Lu, P., Shen, F., Neuzil, P., and
cantilever arrays on a single chip: mass-sensitive Zhang, Q. X. (2005) Out-of-plane electro-
detection of volatile organic compounds. static actuation of microcantilevers.
Anal. Chem. 74, 3084–3095. Nanotechnology 16, 602–608.
138 Yeri and Gao
17. Ilic, B., Czaplewski, D., Zalalutdinov, M., and 31. Yasumura, K. Y., Stowe, T. D., Chow, E. M.,
Craighead, H. G. (2001) Single cell detection Pfafman, T., Kenny, T. W., Stipe, B. C., et al.
with micromechanical oscillators. J. Vac. Sci. (2000) Quality factors in micron- and submicron-
Technol. B 19, 2825–2828. thick cantilevers. J. Microelectromech. Syst. 9,
18. Ilic, B., Czaplewski, D., Craighead, H. G., 117–125.
Neuzil, P., Campagnolo, C., and Batt, C. 32. Haucke, H., Liu, X., Vignola, J. F., Houston,
(2000) Mechanical resonant immunospecific B. H., Marcus, M. H., and Baldwin, J. W.
biological detector. Appl. Phys. Lett. 77, (2005) Effects of annealing and temperature
450–452. on acoustic dissipation in a micromechanical
19. Lavrik, N. V. and Datskos, P. G. (2003) silicon oscillator. Appl. Phys. Lett. 86,
Femtogram mass detection using photother- 181903.
mally actuated nanomechanical resonators. 33. Yang, J., Ono, T., and Esashi, M. (2000)
Appl. Phys. Lett. 82, 2697–2699. Surface effects and high quality factors in
20. Carr, D. W. and Craighead, H. G. (1998) ultrathin single-crystal silicon cantilevers.
Measurement of nanomechanical resonant Appl. Phys. Lett. 77, 3860–3862.
structures in single-crystal silicon. J. Vac. Sci. 34. Wu, G., Datar, R. H., Hansen, K. M.,
Technol. B 16, 3821–3824 Thundat, T., Cote, R. J., and Majumdar, A.
21. Kouh, T., Karabacak, D., Kim, D. H., and (2001) Bioassay of prostate-specific antigen
Ekinci, K. L. (2005) Diffraction effects in (psa) using microcantilevers. Nat. Biotechnol.
optical interferometric displacement detection 19, 856–860.
in nanoelectromechanical systems. Appl. Phys. 35. Majumdar, A. (2002) Bioassays based on
Lett. 86, 013106. molecular nanomechanics. Dis. Markers 18,
22. Foresi, J. S., Black, M. R., Agarwal, A. M., 167–174.
and Kimerling, L. C. (1996) Losses in poly- 36. Arntz, Y., Seelig, J. D., Lang, H. P., Zhang,
crystalline silicon waveguides. Appl. Phys. Lett. J., Hunziker, P., Ramseyer, J. P., et al. (2003)
68, 2052–2054. Label-free protein assay based on a nanome-
23. Stoney, G. G. (1909) The tension of metallic chanical cantilever array. Nanotechnology 14,
films deposited by electrolysis. Proc. R. Soc. 86–90.
Ser. A 82,172–175. 37. Wee, K. W., Kang, G. Y., Park, J., Kang, J. Y.,
24. Ramos, D., Tamayo, J., Mertens, J., Calleja, Yoon, D. S., Park, J. H., et al. (2005) Novel
M., and Zaballos, A. (2006) Origin of the electrical detection of label-free disease marker
response of nanomechanical resonators to bac- proteins using piezoresistive self-sensing
teria adsorption. J. Appl. Phys. 100, 106105. micro-cantilevers. Biosens. Bioelectron. 20,
25. Craighead, H. (2007) Nanomechanical sys- 1932–1938.
tems: Measuring more than mass. Nat. 38. Huber, F., Hegner, M., Gerber, C.,
Nanotechnol. 2, 18–19. Guntherodt, H., and Lang, H. (2006) Label
26. Gupta, A. K., Nair, P. R., Akin, D., Ladisch, free analysis of transcription factors using
M. R., Broyles, S., Alam, M. A., et al. (2006) microcantilever arrays. Biosens. Bioelectron.
Anomalous resonance in a nanomechanical 21, 1599–1605.
biosensor. Proc. Natl. Acad. Sci. USA 103, 39. Fritz, J., Baller, M. K., Lang, H. P., Rothuizen,
13362–13367. H., Vettiger, P., Meyer, E., et al. (2000)
27. Tamayo, J., Ramos, D., Mertens, J. and Translating biomolecular recognition into
Calleja, M. (2006) Effect of the adsorbate nanomechanics. Science 288, 316–318.
stiffness on the resonance response of micro- 40. McKendry, R., Zhang, J., Arntz, Y., Strunz,
cantilever sensors. Appl. Phys. Lett. 89, T., Hegner, M., Lang, H. P., et al. (2002)
224104. Multiple label-free biodetection and quantita-
28. Verbridge, S. S., Craighead, H. G., and Parpia, tive dna-binding assays on a nanomechanical
J. M. (2008) A megahertz nanomechanical cantilever array. Proc. Natl. Acad. Sci. USA
resonator with room temperature quality fac- 99, 9783–9788.
tor over a million. Appl. Phys. Lett. 92, 41. Mukhopadhyay, R., Lorentzen, M., Kjems, J.,
013112. and Besenbacher, F. (2005) Nanomechanical
29. Ekinci, K. L., Yang, Y. T., and Roukes, M. L. sensing of dna sequences using piezoresistive
(2004) Ultimate limits to inertial mass sensing cantilevers. Langmuir 21, 8400–8408.
based upon nanoelectromechanical systems. 42. Hwang, K. S., Lee, J. H., Park, J., Yoon, D.
J. Appl. Phys. 95, 2682–2689. S., Park, J. H., and Kim, T. S. (2004) In situ
30. Ekinci, K. L. and Roukes, M. L. (2005) quantitative analysis of a prostate-specific antigen
Nanoelectromechanical systems. Rev. Sci. (psa) using a nanomechanical pzt cantilever.
Instrum. 76, 061101. Lab Chip 4, 547–552.
43. Lee, J. (2004) Label free novel electrical detec- 47. Gupta, A., Akin, D., and Bashir, R. (2004)
tion using micromachined pzt monolithic thin Single virus particle mass detection using
film cantilever for the detection of c-reactive microresonators with nanoscale thickness.
protein. Biosens. Bioelectron. 20, 269–275. Appl. Phys. Lett. 84, 1976–1978.
44. Burg, T. P. and Manalis, S. R. (2003) 48. Johnson, L., Gupta, A. K., Ghafoor, A., Akin,
Suspended microchannel resonators for D., and Bashir, R. (2006). Characterization of
biomolecular detection. Appl. Phys. Lett. 83, vaccinia virus particles using microscale silicon
2698–2700. cantilever resonators and atomic force micros-
45. Burg, T. P., Godin, M., Knudsen, S. M., Shen, copy. Sens. Actuators, B 115, 189–197.
W., Carlson, G., Foster, J. S., et al. (2007) 49. Sandberg, R., Mølhave, K., Boisen, A., and
Weighing of biomolecules, single cells and Svendsen, W. (2005) Effect of gold coating
single nanoparticles in fluid. Nature 446, on the q-factor of a resonant cantilever.
1066–1069. J. Micromech. Microeng. 15, 2249–2253.
46. Ilic, B., Yang, Y., Aubin, K., Reichenbach, R., 50. Dohn, S. R., Sandberg, R., Svendsen, W., and
Krylov, S., and Craighead, H. G. (2005) Boisen, A. (2005) Enhanced functionality of
Enumeration of dna molecules bound to a nano- cantilever based mass sensors using higher
mechanical oscillator. Nano Lett. 5, 925–929. modes. Appl. Phys. Lett. 86, 233501.
Chapter 10
Nano “Fly Paper” Technology for the Capture of Circulating

Tumor Cells
Shutao Wang, Gwen E. Owens, and Hsian-Rong Tseng
Abstract
Some efficient diagnosis and therapy systems require the isolation and quantification of circulating tumor
cells (CTCs), since these species are important “biomarkers” for monitoring cancer metastasis and prog-
nosis. Existing techniques for isolating/counting CTCs include immunomagnetic-bead-based separation
and microfluidic capture. However, some of these techniques have low capture efficiency and low speci-
ficity. Through the use of a three-dimensional (3D) nanostructured substrate – specifically, a silicon-
nanowire (SiNW) array coated with epithelial-cell-adhesion-molecule antibodies (anti-EpCAM) – we
show that CTCs can be captured efficiently and specifically. Unlike conventional methods for isolating
CTCs that depend on collision frequency and contact duration, nanoscaled local topographic interac-
tions between the CTCs and the substrate increase their binding and markedly enhance capture
efficiency.
Key words: Circulating tumor cells, Silicon-nanopillar array, Cell capture, Cancer diagnostic
1. Introduction
Most cancer-related deaths from solid tumors are caused by

metastasis (1). Although the molecular mechanisms of cancer
metastases remain largely unknown, tumor cells shed from a pri-
mary tumor mass, which enter the blood stream and travel to
distant tissues, signal the earliest stage of malignant progression
(2, 3). These cells are known as circulating tumor cells (CTCs)
(4). Conventional diagnostic imaging and serum marker detec-
tion, which can enumerate and characterize CTCs within blood,
provide valuable information (4–7) for evaluating early stage can-
cer metastases, predicting patient prognosis, and monitoring
therapeutic interventions and outcomes (8). Over the past decade,
several techniques for isolating and counting CTCs have been
141
142 Wang, Owens, and Tseng
Nanoscaled cellular Enhanced cell

Cancer cells
surface components capture efficiency
SiNPs
Anti-EpCAM-coated Enhanced local Anti-EpCAM-coated 5 µm

SiNP substrate topographic interactions SiNP substrate
Fig. 1. A nano “fly paper” substrate for capturing CTCs: a 3D-nanostructured substrate – specifically, a silicon-nanopillar
(SiNP) array coated with anti-EpCAM – enhances the local topographic interactions between nanoscaled cell surface
components and the SiNP substrate, resulting in improved CTC-capture efficiency. As shown in the SEM micrograph
(right ), many interdigitated cellular protrusions (with diameters of ca. 100–200 nm) were observed on the SiNP sub-
strates, validating the working mechanism of our SiNP-based cell-capture approach. Copyright Wiley-VCH Verlag GmbH
& Co. KGaA. Reproduced from ref. 13 with permission.
developed (e.g., immunomagnetic beads (9), flow cytometry

(10), microfluidics (11, 12)). These techniques allow reproducible
detection of CTCs in blood when used in a clinical setting.
However, the challenges that remain include improving CTC-
capture efficiency, reducing the cost of measurement, and
simplifying the sequential molecular analysis of captured cells.
In this chapter, we describe nano “fly paper” technology (13)
(see Fig. 1) that can be employed to capture CTCs. Silicon nano-
pillar (SiNP)-covered substrates coated with epithelial-cell-adhesion-
molecule antibodies (anti-EpCAM) exhibit outstanding efficiency
for isolating viable CTCs from whole blood samples. Using a
simple stationary device, CTCs are immobilized on the SiNP sub-
strates thanks to strong topographic interactions between the
substrate-grated SiNPs and the nanoscale cell surface structure.
2. Materials
2.1. Making SiNP 1. Oriented prime grade silicon wafers, p-type, resistivity of ca.
Substrates Using Wet 10–20 ohm cm (Silicon Quest Int’l, Santa Clara, CA). Store
Chemical Etching at room temperature.
2. Acetone (ACS reagent, SpectroGrade, 99.5%). Store at room
temperature.
3. Ethanol, >99.5%. Store at room temperature.
4. Piranha solution: 4:1 (v/v) sulfuric acid/hydrogen peroxide.
Store at room temperature.
5. RCA solution: 1:1:5 (v/v/v) ammonia/hydrogen peroxide/
H2O.
6. Etching solution: 4.6 M Hydrofluoric acid (HF), 0.2 M silver
nitrate in deionized water. Store at room temperature.
7. Aqua regia solution: 3:1 (v/v) hydrochloric acid/nitric acid.
Nano “Fly Paper” Technology for the Capture of Circulating Tumor Cells 143
2.2. Making 1. 4% (v/v) 3-Mercaptopropyl trimethoxysilane in ethanol

Streptavidin-Coated (95%) (Sigma-Aldrich, St. Louis, MO). Store at room
SiNP Substrates temperature.
2. 0.25 mM N-g-maleimidobutyryloxy succinimide ester
(GMBS, 98% HPLC) (Sigma-Aldrich). Store at room
temperature.
3. Streptavidin (10 mg/mL) (Invitrogen, Carlsbad, CA). Store
in single-use aliquots at −20°C.
4. 1× Dulbecco’s phosphate-buffered saline (PBS) (Invitrogen).
Store at 4°C.
2.3. Scanning Electron 1. 4% Glutaraldehyde (E.M. grade, Polysciences, Warrington,

Microscopy of SiNP PA) in 0.1 M cacodylic acid sodium salt trihydrate (Sigma-
Substrates Aldrich). Store at room temperature.
2. 1% Osmium tetroxide (ACS reagent, >98%, Sigma-Aldrich).
Toxic! Store at room temperature.
3. 1% Tannic acid (Electron Microscopy Sciences, Hatfield, PA).
4. 0.5% Uranyl acetate (Electron Microscopy Sciences). Store at
room temperature.
5. Alcohol (97%, Sigma-Aldrich).
6. Hexamethyldisilazane (HDMS) (Sigma-Aldrich). Toxic!
7. Gold plate.
2.4. Preparation 1. Breast cancer cell line, MCF7 (American Type Culture
of Artificial Blood Collection, Manassas, VA).
Samples 2. Reagents for cell culture: Dulbecco’s Modified Eagle’s
Medium (DMEM, 1×) (Invitrogen), fetal bovine serum (FBS)
(Fisher Scientific, Pittsburg, PA), penicillin–streptomycin
(100×) (Fisher Scientific), and trypsin. Store at −20°C.
3. Citrated whole rabbit blood (Colorado Serum Company,
Denver, CO).
4. Vybrant® DiD cell-labeling solution (Invitrogen). Store at
4°C.
5. Dulbecco’s phosphate-buffered saline (PBS) (Invitrogen)
Store at 4°C.
2.5. Capture and 1. Lab-Tek chamber slides, 4-well glass, sterile (Thermo Fisher
Immunochemistry Scientific). Store at room temperature.
of CTCs from Artificial 2. 10 mg/mL Biotinylated anti-human EpCAM/TROP1 anti-
Blood Samples body (Goat IgG, R&D Systems, Minneapolis, MN), 1% (w/v)
bovine serum albumin (BSA, Sigma-Aldrich), 0.09% (w/v)
sodium azide in 1× PBS. Store in single-use aliquots at −20°C.
3. 1× PBS.
4. 4% Paraformaldehyde in 1× PBS. Store at 4°C.
5. 0.2% Triton X-100 in 1× PBS. Store at 4°C.
6. 1% DAPI in 1× PBS. Store at 4°C.
2.6. Capture and 1. Lab-Tek chamber slides, 4-well glass, sterile (Thermo Fisher
Immunochemistry for Scientific). Store at room temperature.
CTCs Captured from 2. 10 mg/mL Biotinylated anti-human EpCAM/TROP1 anti-
Patient Samples body (Goat IgG, R&D Systems, Minneapolis, MN), 1%
(w/v) bovine serum albumin (BSA, Sigma-Aldrich), 0.09%
(w/v) sodium azide in 1× PBS. Store in single-use aliquots
at −20°C.
3. 1× PBS.
4. Vacutainer tubes containing the anticoagulant ethylenedi-
aminetetraacetic acid (ETDA).
5. 4% Paraformaldehyde in 1× PBS. Store at 4°C.
6. 0.3% Triton X-100 in 1× PBS. Store at 4°C.
7. Blocking solution: 5% normal goat serum, 0.1% Tween 20,
3% BSA in 1× PBS.
8. 1% DAPI in 1× PBS. Store at 4°C.
9. 20 mg/mL Anti-cytokeratin conjugated with phycoetythrin
(CAM5.2 conjugated with PE) (BD Biosciences, San Jose,
CA) in 1× PBS. Store in single-use aliquots at −20°C.
10. 20 mg/mL Anti-human CD45 conjugated with FITC (Ms
IgG1, clone H130) (BD Biosciences) in 1× PBS. Store in
single-use aliquots at −20°C.
3. Methods
Over the past decade, techniques for isolating and counting CTCs
have been developed that are based on different working mecha-
nisms (4–7). Recently, we have developed a novel 3D SiNP-based
CTC-capture technology and we call this as nano “fly paper.” To
test the cell-capture performance of the SiNPs substrates, we pre-
pared a cell suspension solution of an EpCAM-positive breast
cancer cell line (i.e., MCF7) in cell culture medium (DMEM).
After optimization of experimental parameters (e.g., CTC cap-
ture, sensitivity, and reproducibility), a clinical study with the
SiNP-based CTC-capture technology was conducted using blood
samples collected from patients with metastatic prostate cancer in
collaboration with the Department of Urology UCLA under
UCLA IRB approval (IRB #09-03-038-01). After CTC capture
by the SiNP-based system, a 3-parameter immunocytochemistry
protocol (for parallel staining of DAPI, FITC-labeled anti-CD45,

and PE-labeled anti-cytokeratin (CAM2.5)) was applied to stain
the immobilized cells. Based on the signal thresholds and size/
morphology features established for model cells, the CTCs were
clearly distinguishable from background immune cells. Since only
1.0 mL of blood is required for each CTC-capture study, we
could perform as many as three measurements on each blood
sample. Figure 2 presents an example of one successful CTC-
capture study on the SiNP-based substrate using blood samples
that, when tested using CellSearchTM technology, had failed to
detect any CTCs.
3.1. Making SiNP 1. Cut silicon wafer into pieces 1×2 cm in dimension.
Substrates Using Wet 2. Ultrasonicate the cut silicon substrates in acetone for 10 min
Chemical Etching at room temperature and dry under nitrogen gas. Next, ultra-
sonicate the substrates in ethanol for 5 min at room tempera-
ture and dry them under nitrogen gas. These steps remove
contamination (such as organic grease) from the silicon
substrate.
Fig. 2. A scatter plot summarizes the quantitative immunocytochemistry measured on

both cytokeratin and CD45 expression levels of patient CTCs and lymphocytes captured
on a 1 × 2 cm SiNP substrate. The size of each dot reflects the footprint of each cell. In
this case, we were able to identify 6 CTCs in a 1-mL blood sample from a prostate
cancer patient. The insert is a typical fluorescent image of CTCs captured on SiNP
substrates.
3. To etch the silicon wafer surface, first heat the silicon substrates
in boiling piranha solution for 1 h. Then, heat the substrates in
boiling RCA solution for 1 h. Rinse the substrates five times
with DI water. Next, place the silicon substrates in a teflon ves-
sel and etch them with etching solution at 50°C (see Note 1).
4. Immerse the substrates in boiling aqua regia for 15 min.
5. Rinse the substrates with DI water and dry them under
nitrogen gas.
3.2. Making 1. Place the substrates in 4% (v/v) 3-mercaptopropyl trimethyl-

Streptavidin-Coated silane in ethanol for 45 min at room temperature.
SiNP Substrates 2. Treat the substrates with 0.25 mM N-g-maleimidobutyryloxy
succinimide ester (GMBS) and incubate for 30 min at room
temperature.
3. Treat the substrates with streptavidin (SA) (10 mg/mL) and
incubate for 30 min at room temperature.
4. Flush the substrates with 1× PBS to remove excess
streptavidin.
5. Store modified substrates at 4–8°C for up to 6 months.
3.3. Scanning Electron 1. Allow cells to incubate on the substrates for 24 h.
Microscopy 2. Fix cells with 4% glutaraldehyde buffered in 0.1 M sodium
Observation of SiNP cacodylate and incubate cells for 1 h at 4°C. Afterward, post-
Substrates fix cells using 1% osmium tetroxide for 1 h, using 1% tannic
acid as a mordant.
3. Dehydrate samples through a series of alcohol concentrations
(30, 50, 70, and 90%). Stain samples with 0.5% uranyl acetate
and then further dehydrate the samples through a series of higher
alcohol concentrations (96, 100, and 100%). Dehydrate the
samples for the final time in hexamethyldisilazane (HMDS).
4. Air-dry the samples.
5. Once dry, sputter-coat the samples with gold (3–5 nm in
thickness) and examine with a field emission SEM (accelerat-
ing voltage of 10 keV).
3.4. Preparation 1. Stain MCF7 cells (in 1 mL DMEM without serum, 106 cells/
of Artificial mL) with DiD red fluorescent dye (5 mL, for 20 min).
Blood Samples 2. Prepare control samples by spiking stained MCF7 cells into
rabbit blood with cell densities of 1,000–1,250, 80–100, and
5–20 cells/mL.
3.5. Capture 1. Place substrates into a size-matched 4-well Lab-Tek Chamber

and Immunochemistry Slide. Drop 25 mL of biotinylated anti-EpCAM (10 mg/mL
of CTCs from Artificial in 1× PBS with 1% (w/v) BSA and 0.09% (w/v) sodium
Blood Samples azide) onto a 1×2 cm substrate. Incubate for 30 min. Wash
(See Fig. 3) with 1× PBS.
Fig. 3. A simple and convenient procedure of cell-capture experiments by using the nano “fly paper” technology. Copyright
Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced from ref. 13 with permission.
2. Load 1 mL control sample (see Subheading 3.4) onto each

substrate.
3. Incubate for 45 min (37°C, 5% CO2).
4. Gently wash the substrate with 1× PBS at least five times.
5. Fix cells captured on the substrates with 4% paraformalde-
hyde (PFA) in 1× PBS for 20 min.
6. To stain and visualize captured cells, treat the substrates with
0.2% Triton X-100 in 1× PBS and incubate for 10 min.
Incubate the substrates with a DAPI solution (1% DAPI in
1× PBS) for 5 min. Wash the substrates three times with 1×
PBS. Invert the substrates onto a standard cover glass.
7. Image and count cells using a Nikon TE2000 fluorescence
microscope and a hemocytometer, respectively (see Note 2)
(see Subheading 3.7).
3.6. Capture and 1. Place substrates into a size-matched 4-well Lab-Tek Chamber
Immunochemistry Slide. Drop 25 mL of biotinylated anti-EpCAM (10 mg/mL
for CTCs Captured in 1× PBS with 1% (w/v) BSA and 0.09% (w/v) sodium
from Patient Samples azide) onto a 1 × 2 cm substrate. Incubate for 30 min. Wash
(See Fig. 3) with 1× PBS.
2. Blood samples drawn from patients with advanced solid-stage
tumors (as approved by IRB) are collected into vacutainer
tube containing the anticoagulant ETDA. Samples should be
processed immediately after collection.
3. To capture cells, load 1 mL patient sample onto each SiNP
substrate. Incubate for 45 min (37°C, 5% CO2) and then
gently wash the substrates with 1× PBS at least five times.
4. Fix cells captured on the substrate by loading 200 mL of 4%

paraformaldehyde (PFA) in 1× PBS onto each substrate for
20 min at room temperature. Wash each substrate three times
with PBS.
5. Permeabilize cells by treating each substrate with 200 mL of
0.3% Triton X-100 in PBS for 30 min at room temperature.
Subsequently, wash each substrate three times with 1× PBS.
6. Add 200 mL of blocking solution to each substrate and incu-
bate for 1 h at room temperature. Next, add 200 mL of each
fluorophore-labeled antibody solution (CAM2.5-PE and
CD45-FITC) to each substrate and incubate the substrates in
the dark at 4°C overnight. Wash with 200 mL of 1× PBS three
times (first wash: 15 min at room temperature, second and
third washes: 5 min at room temperature). Incubate sub-
strates with 1% DAPI solution for 5 min. Wash each substrate
three times with 1× PBS.
7. Gently invert the substrates, using tweezers, onto a cover
glass to prepare for imaging and counting (see Note 3).
3.7. Identification of 1. Select fluorescent microscope settings. Optimized exposure

CTCs by Fluorescence times: DAPI (blue) filter: 50 ms exposure time (background:
Microscopy ~1,300), FITC (green) filter: 300 ms exposure time (back-
(See Note 4) ground: ~1,600), PE (red) filter: 100 ms exposure time
(background: ~1,300). Set the digitizer to 1 MHz. Other
settings will yield a very high background signal when using a
green filter.
2. Place a sample on the microscope and focus on the edge of
the substrate. Once focused, switch to the blue filter.
3. Starting in the upper right corner of the substrate, scan for
nuclei that are approximately 7–20 mm in diameter at 4× or
10× magnification.
4. Increase the magnification to 10× or 20× when a putative cell
has been located. Check the fluorescence intensity under
blue, red, and green filters. Score a sample fluorescence inten-
sity >2× the background fluorescence intensity as a positive
result.
4. Notes
1. The nanopillar length depends on the duration of the etching

step; nanopillars 10 mm in length are optimal for these
experiments.
2. Cells that show dual stains (red: DiD+ and blue: DAPI+) and
meet the phenotypic morphological characteristics (e.g., cell
size, shape, and nucleus size) are scored as CTCs. Cells stained
by DAPI+ only are counted as nonspecific cells.
3. Cells that demonstrate dual staining (red: PE+ and blue:
DAPI+) and meet standard phenotypic and morphological
characteristics should be scored as CTCs. Cells that show
dual staining (green: FITC+ and blue: DAPI+) should be
excluded as lymphocytes/nonspecific cells (see Fig. 2). Species
that demonstrate staining in all three filters (green+ and red+
and blue+) should be excluded as cellular debris.
4. Care should be taken to ensure that the substrates are never
allowed to dry.
Acknowledgments
The authors appreciate the helpful discussions with Dr. Hao

Wang, Dr. Jing Jiao, Dr. Ken-ichiro Kamei, Dr. Jing Sun, Kuan-Ju
Chen, Gwen E. Owens, and David J. Sherman. This research was
supported by NIH-NCI NanoSystems Biology Cancer Center
(U54CA119347).
References
1. Steeg, P. S. (2006) Tumor metastasis: mecha- biological properties of disseminating tumour

nistic insights and clinical challenges. Nat. cells. Nat. Rev. Cancer 8, 329–340.
Med. 12, 895–904. 8. Cristofanilli, M., Budd, G. T., Ellis, M. J.,
2. Bernards, R. and Weinberg, R. A. (2002) A Stopeck, A., Matera, J., Miller, M. C., et al.
progression puzzle. Nature 418, 823. (2004) Circulating tumor cells, disease pro-
3. Pantel, K. and Brakenhoff, R. H. (2004) gression, and survival in metastatic breast can-
Dissecting the metastatic cascade. Nat. Rev. cer. New Engl. J. Med. 351, 781–791.
Cancer 4, 448–56. 9. Talasaz, A. H., Powell, A. A., Huber, D. E.,
4. Budd, G. T., Cristofanilli, M., Ellis, M. J., Berbee, J. G., Roh, K.-H, Yu, W., et al. (2009)
Stopeck, A., Borden, E., Miller, M. C., et al. Isolating highly enriched populations of circu-
(2006) Circulating tumor cells versus imag- lating epithelial cells and other rare cells from
ing--predicting overall survival in metastatic blood using a magnetic sweeper device. Proc.
breast cancer. Clin. Cancer Res. 12, Natl. Acad. Sci. USA 106, 3970–3975.
6403–6409. 10. Allan, A. L., Vantyghem, S. A., Tuck, A. B.,
5. Allard, W. J., Matera, J., Miller, M. C., Chambers, A. F. Chin-Yee, I. H., and Keeney,
Repollet, M., Connelly, M. C., Rao, C., et al. M. (2005) Detection and quantification of
(2004) Tumor cells circulate in the peripheral circulating tumor cells in mouse models of
blood of all major carcinomas but not in human breast cancer using immunomagnetic
healthy subjects or patients with nonmalig- enrichment and multiparameter flow cytome-
nant diseases. Clin. Cancer Res. 10, try. Cytometry A 65, 4–14.
6897–6904. 11. Nagrath, S., Sequist, L. V., Maheswaran, S.,
6. Zieglschmid, V., Hollmann, C., and Böcher, Bell, D. W., Irimia, D., Ulkus, L., et al. (2007)
O. (2005) Detection of disseminated tumor Isolation of rare circulating tumour cells in
cells in peripheral blood. Crit. Rev. Clin. Lab. cancer patients by microchip technology.
Sci. 42, 155–196. Nature 450, 1235–1239.
7. Pantel, K., Brakenhoff, R. H., and Brandt, B. 12. Adams, A. A., Okagbare, P. I., Feng, J.,
(2008) Detection, clinical relevance and specific Hupert, M. L., Patterson, D., Gottert, J., et al.
(2008) Highly efficient circulating tumor cell 13. Wang, S., Wang, H., Jiao, J., Chen, K.-J.,
isolation from whole blood and label-free Owens, G. E., Kamei, K.-I., et al. (2009)
enumeration using polymer-based microfluid- Three-dimensional nanostructured substrates
ics with an integrated conductivity sensor. toward efficient capture of circulating tumor
J. Am. Chem. Soc. 130, 8633–8641. cells. Angew. Chem. Int. Ed. 48, 8970–8973.
Chapter 11
Polymeric Nanoparticles for Photodynamic Therapy

Yong-Eun Koo Lee and Raoul Kopelman
Abstract
Photodynamic therapy is a relatively new clinical therapeutic modality that is based on three key components:
photosensitizer, light, and molecular oxygen. Nanoparticles, especially targeted ones, have recently
emerged as an efficient carrier of drugs or contrast agents, or multiple kinds of them, with many advan-
tages over molecular drugs or contrast agents, especially for cancer detection and treatment. This paper
describes the current status of PDT, including basic mechanisms, applications, and challenging issues
in the optimization and adoption of PDT; as well as recent developments of nanoparticle-based PDT
agents, their advantages, designs and examples of in vitro and in vivo applications, and demonstrations
of their capability of enhancing PDT efficacy over existing molecular drug-based PDT.
Key words: Photodynamic therapy, Nanoparticle, Photosensitizer, Optical penetration depth,

Polymer
1. PDT Basic Facts
Photodynamic therapy (PDT) is a minimally invasive localized

treatment modality based on three key components: photosensi-
tizer, light, and molecular oxygen. It is a relatively new therapy –
tested in cells and animal models in the 1960s and 1970s, tested
in clinical trials in the 1980s, and approved clinically in 1995 in
the USA (1) – although its discovery dates back to the early 1900s
(2, 3). Currently, PDT is clinically approved for treatment of
several medical conditions, including cases of skin actinic kerato-
sis, several forms of cancer, blindness due to age-related macular
degeneration and localized bacterial infection (3–5). New medi-
cal applications of PDT continue to be discovered and clinically
tried (4, 5). The PDT procedure involves the following steps:
(1) the photosensitizer, after being administered systemically, or
topically, preferentially accumulates in target tissues, and (2) light of
151
152 Koo Lee and Kopelman
Excited Singlet States

Formation of Cytotoxic Species
Excited Triplet States

Intersystem
Crossing Type I: Oxygen independent
formation of radicals and ions
of photosensitizers or
Non-radiative singlet decay
substrates by electron or
hydrogen transfer
Apoptosis, Cell Death
Fluorescence
Necrosis, (PDT)
Non-radiative triplet decay

Absorption
Autophagy
Phosphorescence
Type II: Oxygen dependent
formation of singlet oxygen
by energy transfer reaction
with molecular oxygen
Ground Singlet State
Fig. 1. Reaction kinetics of photosensitization leading to PDT (Types I and II).
the appropriate wavelength is directed onto the target tissue to

induce a photodynamic reaction between an optically excited
photosensitizer and surrounding oxygen (or other) molecules,
so as to produce reactive oxygen species (ROS) (see Fig. 1).
Type I reactions involve a direct electron/hydrogen transfer
from the photosensitizer, producing ions, or electron/hydrogen
abstraction from a substrate molecule, to form free radicals. These
radicals then react rapidly, usually with oxygen, resulting in the
production of ROS such as superoxide anion radicals, hydrogen
peroxide, and hydroxyl radicals. Type II reactions involve an
energy transfer reaction between the excited triplet state photo-
sensitizer and an oxygen molecule, resulting in the formation of
“singlet oxygen,” the first electronically excited, and highly reac-
tive, state of oxygen. The reactive species thus produced, via type
I or II reactions, then attack plasma membranes or subcellular
organelles such as the mitochondria, plasma membrane, Golgi
apparatus, lysosomes, endosomes, and endoplasmic reticulum,
leading to cell death by apoptosis, necrosis, and autophagy (6).
The relative contribution from type I and II reactions depends on
the type of photosensitizer used, the available tissue oxygen and
the type of substrate (e.g., a cell membrane or a biomolecule), as
well as the binding affinity of the photosensitizer for the substrate
(3). Direct and indirect evidence indicates that type II reactions,
hence singlet oxygen, play a dominant role in PDT (7, 8). The
reported lifetime (t) of singlet oxygen in vivo ranges from 0.03 to
0.18 ms (8), and the estimated diffusion distance of singlet oxy-
gen, based on a diffusion coefficient, D, for oxygen in water of
2 × 10−9 m2/s (9) and the root-mean-square diffusion distance,
(2Dt)1/2, is only 11–27 nm. The other cytotoxic reactive species
generated by the type I reactions also have short lifetimes and
Polymeric Nanoparticles for Photodynamic Therapy 153
diffusion distances (10). This indicates that the PDT effects are
confined to the illuminated area containing photosensitizers. This
localization minimizes side effects related to PDT.
The PDT efficiency depends on the photosensitizer’s ability to
produce ROS, the oxygen availability, and the light dose (intensity
and duration), as well as the photosensitizer concentration at
the treated area. The PDT action on the tumor tissue level can
be described by three different processes: (1) direct photodynamic
tumor cell killing, (2) indirect tumor cell killing by photody-
namic damage to, or shutdown of, the tumor supporting vascula-
ture, with loss of oxygen and nutrients to the tumor, and (3)
additional antitumor contributions from the inflammatory and
immune responses of the host (11).
Since the photosensitizers used are of inherently low systemic
toxicity and become pharmacologically active only when exposed
to light, PDT can be applied selectively to the diseased tissue
without significantly affecting connective tissues and underlying
structures. There is less morbidity, less functional disturbance,
and better cosmetic outcome than with other therapies. PDT has
a low mutagenic potential and, except for sustained skin photo-
toxicity, few adverse effects. PDT can also be repeated at the same
site without lifetime dose limitation, unlike radiotherapy. PDT
can be used as a stand-alone modality or in combination with
other therapeutic modalities, wherein it can be applied either
before or after other cancer therapies, without compromising
these treatments or being compromised itself. For example, PDT
is employed for the treatment of glioblastoma multiforme (GBM)
as adjuvant therapy after surgical resection of the glioblastoma
mutiforme tumor, resulting in a significant survival advantage
without added risk to the patient (12).
The one challenging issue for PDT is that its efficiency is
inherently limited by the optical tissue penetration depth. The
effective excitation light magnitude is determined by the combi-
nation of optical absorption and scattering properties of the tissue.
The optical absorption properties of tissue vary significantly with
wavelength (see Fig. 2). The optical scattering of tissue decreases
with wavelength, and the reduced scattering coefficient ( µ′s )
follows the scaling law, µ s′=λ − b , where b = 0.5–2 over the visible and
IR spectral region (13). For the spectral range of 450–1,750 nm,
tissue scattering is, in general, more prevalent than absorption,
although for the range of 450–600 nm, melanin and hemoglobin
provide significant absorption, while water plays a similar role
for l > 1,350 nm (13). Therefore, the optimal optical window
for PDT, as well as for optical imaging, is in the near-infrared
(NIR) spectral region (600–1,300 nm), where the scattering and
absorption by tissue are minimized and, therefore, the longest
penetration depth can be achieved. Within this optical window, the
longer the wavelength is, the deeper is the penetration depth.
106
105
Absorption Coefficient µa [cm−1]

Collagen
Protein
104
103 Melanin
102
Hb
101 HbO2
100
Water
0.1 0.3 1 3 10
Wavelength λ [µm]
Fig. 2. Optical absorption coefficients of principal tissue chromophores in the 0.1–12 mm

spectral region. Reproduced from ref. 13 with permission from the American Chemical
Society.
Moreover, the optical penetration depth varies significantly with

tissue type (14–16) (see Table 1). Here, the optical penetration
depth into tissue is defined as the distance over which the light
intensity drops to 1/e or 37% of the initial value. It should be
noted that beyond this depth there remains light at a lower inten-
sity, which still may be sufficient for PDT. The effective depth of
PDT-induced necrosis, a more practical term for therapeutic
depth of PDT activity, is approximately three times the penetra-
tion depth (17, 18). For instance, the optical penetration depth at
630 nm, the wavelength for PDT using Photofrin, is 0.2–2.9 mm,
depending on organ (see Table 1), which limits PDT efficacy to
less than 9 mm below the illuminated surface.
PDT has also other challenging issues regarding its clinical
therapeutic efficacy, similar to challenges faced by chemotherapy
and radiotherapy. Those include improvement of PDT drug effi-
ciency, efficient delivery of the PDT drugs to the treatment area,
and development of methods to maintain the therapeutic effi-
ciency even at low oxygen levels, especially in the case of tumors.
2. Factors
Controlling PDT
Effectiveness
The PDT efficiency depends on multiple factors: (1) the photo-
chemical properties of the photosensitizer such as its singlet
oxygen production efficiency and tissue penetration depth of its
excitation light, (2) the delivery system, (3) the biological state of
the tumor including the tumor type and its microvasculature, as
well as the tissue oxygenation level, (4) the physical localization
Table 1
Optical penetration depth (mm) of selected tissues
Wavelength of light
Tissue 630 nm 632.8 nm 665 nm 675 nm 780 nm 835 nm 1,064 nm References
Blood 0.19 0.28 0.42 0.51 [14]
Mammary tissue 2.59 2.87 3.12 3.54 [14]
Mammary carcinoma 2.87 3.14 3.62 4.23 [14]
Mammary carcinoma 2.0 2.3 3.7 [15]

in C3H/HEJ mice
Brain (postmortem) 0.92 1.38 2.17 2.52 [14]
Brain 1.6 [16]
Brain tumor (glioma) 3.1 [16]
Colon 2.48 2.73 2.91 [14]
Lung 0.81 1.09 1.86 2.47 [14]
Lung carcinoma 1.68 2.01 2.82 3.89 [14]
and the amount of photosensitizer in treated tissue, and (5) the

light parameters such as light dose, light fluence rate, and the
interval between drug and light administration. The success of
PDT depends on improvements in the combination of all these
factors, so as to circumvent the aforementioned challenging
issues. The chemical, biological, and clinical progresses for those
factors made so far are described below.
2.1. PDT Efficiency The hematoporphyrin derivative, Photofrin, is the first clinically
and Optical Tissue approved, and currently the most widely used, PDT drug. It is
Penetration Depth approved for tumors (esophageal and endobronchial cancers) and
of Photosensitizer for high-grade dysplasia in Barrett’s esophagus. This so-called
“first-generation” PDT drug has several drawbacks. Photofrin
lacks a definite molecular structure as it is a mixture of oligomers
formed by ether and ester linkages of up to eight porphyrin units
(19); this makes the interpretation of its pharmacokinetic data
difficult. Photofrin requires excitation light of 630 nm to achieve
sufficient tissue penetration, though the absorption peak at
630 nm corresponds to the longest wavelength but weakest
absorption peak of Photofrin, resulting in low PDT efficiency.
Furthermore, its accumulation in skin and slow clearance cause
prolonged skin photosensitivity lasting for at least 30 days but
often up to 90 days or more (20). Therefore, there has been a
strong need for producing photosensitizers with significantly
improved properties, in comparison to Photofrin.
Ideally, a photosensitizer should (1) be a single chemical

compound of definite molecular structure in a highly purified and
stable form, (2) possess high photosensitizing ability, (3) possess
high photostability, (4) have a relatively high absorbance at
longer wavelengths, for greater tissue penetration, (5) exhibit no
dark toxicity, (6) exhibit no long-term cutaneous phototoxicity,
(7) accumulate selectively in the targeted tissue, and (8) undergo
fast clearance from the body to avoid prolonged photosensitivity.
To date, second-generation photosensitizers that have definite
structures and/or a major absorption band in the wavelength
range of 650–800 nm have been developed (see Table 2).
In addition to the improvements in conventional one-photon
excitation PDT as shown above, two-photon absorption
induced excitation of photosensitizers has received attention lately
as an alternative approach to achieve long tissue penetration depth.
This approach can achieve enhanced tissue penetration depth, in
comparison to the one-photon excitation approach, by combining
the energy of two photons – in the range of 780–950 nm – where
tissues have maximum transparency to light but where the energy
of one photon is not sufficient enough to produce singlet oxygen
(21). Several such dyes have been tested for use in vitro (22) and
in vivo (21) but not yet for use in a clinical setting.
2.2. Localization Selective delivery of drug or imaging contrast agents to diseased

of the Photosensitizer target tissue or cells is one of the most important factors in
in Diseased Tissue determining the efficacy of the intravenously injected therapy
or Tumor or imaging modalities. Higher drug localization in target tissue
can increase the therapeutic efficacy, even with a reduced dose.
Second-generation photosensitizers have been developed to
exhibit improved selectivity and body clearance rate compared to
ones in the first-generation so as to reduce post-PDT photosensi-
tivity. However, developing a way to enhance tumor selectivity
is still one of the major challenges to be solved in PDT. This
problem can be addressed by “third-generation” photosensitizers –
photosensitizers conjugated with a targeting component that
is specific toward the tumor site/cells and that can increase the
selective accumulation of the photosensitizer in sites of interest
as well as minimize healthy cell localization and concomitant
damage (23).
2.3. Administration The PDT drug can be administered to the infected area either
of PDT Drug and Light topically or by systemic intravenous injection (see Table 2). For
topical administration, there is no need for delay in the adminis-
tration of light, unless the delivered PDT drug is a prodrug such
as Levulan that requires 15–60 min of activation time for the
treatment of actinic keratoses (24). For systemic administration,
however, the location and accumulation amount of the photosen-
sitizers depend on postinjection time. The drugs, after a short
Table 2
Second-generation photosensitizers that are clinically approved or under clinical trial
Excitation Method of
Photosensitizer Trade name wavelength (nm) administration Clinical use
Temoporfin (meta-tetrahydroxy- Foscan 652 Intravenous (i.v.) Approved for the palliative treatment of patients with
phenylchlorin, i.e., m-THPC) advanced head and neck cancer (European Union, Norway,
and Iceland)
Talaporfin (mono-l-aspartyl Laserphyrin 664 i.v. Approved for early endobronchial carcinoma (Japan)
chlorine e6)
Verteporfin (Benzoporphyrin Visudyne 689 i.v. Approved for the treatment of predominantly classic
derivative-monoacid ring) (liposomal subfoveal choroidal neovascularization (CNV) due to
formulation) age-related macular degeneration (AMD), pathologic
myopia, or presumed ocular histoplasmosis (USA)
HPPH (2-(1-hexyloxyethyl)-2- Photochlor 665 i.v. Phases 1 and 2 clinical studies in patients with Barrett’s
devinyl pyropheophorbide- esophagus with high-grade dysplasia, obstructive
alpha) esophageal cancer, early or late-stage lung cancer,
or basal cell carcinoma
Lutetium texaphyrin (Motexafin Antrin 732 i.v. Phase I trial for locally recurrent prostate cancer; Phase I for
lutetium) Lutex phototherapy (PT) in patients undergoing percutaneous
Lutrin coronary intervention with stent deployment
ALA (5-aminolevulinic acid Levulan 630 Topical Prodrug for Protoporphyrin IX (PpIX). Approved for a
HCl) precancerous skin condition called actinic keratosis (USA)
Methylene blue Periowave 665 Topical Approved for periodontal diseases, and nasal decolonization
Polymeric Nanoparticles for Photodynamic Therapy
of methicillin-resistant Staphylococcus aureus (Canada)

157
postinjection time that is less than the plasma half-life of the PDT
drugs, predominantly stay in the vascular compartment of the
tumor. At longer postinjection times, drugs may accumulate in
the extravascular compartment of the tumor due to its relatively
slow leakage from the vasculature and interstitial diffusion.
Therefore, the interval between the PDT drug administration and
light exposure (drug-to-light interval) may play an important role
in the PDT outcome. The drug-to-light interval used for current
clinical PDT is quite long – for example, on the order of 40–50
and 90–110 h for Photofrin (19) and Foscan (25), respectively.
There have been efforts to enhance the PDT efficacy by
changing the light or drug administration protocols. Fractionated
drug-dose PDT (i.e., administration of photosensitizers at mul-
tiple time intervals before light activation) was reported to achieve
a high localization of photosensitizers in both vascular and tumor
cell compartments, resulting in a superior PDT effect compared
to single-dose PDT (26). The light fluence rate is an important
item to modulate the PDT outcome. The fluence rate during the
administration of light can significantly affect photobleaching and
antitumor effectiveness, by controlling tissue oxygenation during
light delivery (27, 28).
3. Why
Nanoparticles
for PDT?
Nanoparticles have been utilized as a delivery vehicle for drugs,
image contrast agents or for both, and have shown their ability to
improve the efficacy of existing imaging and therapy methods,
especially in cancer detection and treatment, including PDT (29).
The success can be attributed to the many advantages resulting
from the nanoparticles’ inherent properties including their size,
inert and nontoxic matrices, as well as their flexible engineerability
that allows various modifications of their properties, improved
targetability, and possible multifunctionality. Specifically, the
advantages of using nanoparticles for PDT are described below.
First, each nanoparticle can carry a large amount of photo-
sensitizers, offering a coherent, critical mass of destructive power
for intervention. Most photosensitizers are hydrophobic and
poorly water-soluble, thereby tending to form aggregates under
physiological conditions. Proper loading of photosensitizers into
nanoparticles of higher aqueous solubility and longer plasma resi-
dence time can make intravenous administration very easy and
enhance the PDT efficiency of the system by preventing the
aggregation of the photosensitizers. Note that for most photo-
sensitizers, their dimers or aggregates have a very low singlet oxygen
production efficiency (30, 31). The drug-loading procedure
can be done by a variety of methods, such as encapsulation, covalent
linkage, or postloading by physical adsorption through electrostatic

or hydrophobic interactions. The loading degree can be con-
trolled by the input amount of photosensitizers per nanoparticle, as
well as by engineering or modifying the nanoparticles’ properties,
such as size, composition, charge, and hydrophobicity.
Second, the size of the nanoparticles offers certain distinct
advantages for drug delivery. When drug molecules or carriers are
in the blood stream, they can extravasate across the endothelium
by transcapillary pinocytosis, as well as by passage through inter-
endothelial cell junctions, gaps, or fenestrae (32). The molecular
weight or the size of the drug molecules or carriers has been shown
to be a critical factor for the endothelial permeability (32, 33). The
size of fenestrae varies with the organ type – from a few angstroms
for the blood–brain barrier (BBB) to 150 nm for the liver and
spleen (34). However, in pathological situations, such as cancer or
inflammatory tissues, the vasculature becomes leaky, with larger
fenestrae, up to 780 nm in size (34). Small molecular drugs freely
diffuse in and out of most normal tissues as well as tumor tissues,
without any discrimination. However, nanoparticles escape the
vasculature preferentially through abnormally leaky tumor blood
vessels and are then subsequently retained in the tumor tissue for
a long time, because tumor tissue has a very poor lymphatic drain-
age system, unlike normal or inflammatory tissue. Because of this
penetration phenomenon, called the enhanced permeability and
retention (EPR), nanoparticles can efficiently deliver therapeutic
agents to tumors (35, 36). Nanoparticles between 10 and 100 nm
are believed to provide the best option, because nanoparticles with
a size greater than 10 nm can avoid kidney clearance, resulting in
prolonged and elevated levels in the blood stream, and those
with a size smaller than 100 nm can penetrate deep into tissues,
without being trapped by the phagocytes (37, 38).
Third, the nanoparticles can be surface-modified to make
more efficient and selective delivery to target tissues or cells. Drug
localization in target tissue is known to be determined by vascular
permeability and interstitial diffusion, which depend not only on
the size of the drugs as described above but also on the configura-
tion, charge, and hydrophilic or lipophilic properties of the com-
pound, as well as the physiological properties of blood vessels
(39). These properties can be relatively easily modified for the
nanoparticle-based agents. The charge, hydrophilicity, or lipophi-
licity can be changed either by introducing polymers or mono-
mers with characteristic functional groups (charged, hydrophilic, or
lipophilic) during synthesis or by coating the nanoparticles with
polymers of such characteristics. For instance, nanoparticles can
be coated with polyethylene glycol (PEG) for longer plasma
circulation time (40). The nanoparticle surface can also be conju-
gated with tumor-specific ligands, such as antibodies, aptamers,
peptides, or small molecules that bind to antigens that are present
on the target cells or tissues, for “active targeting.” Some of the

targeting moieties even help NPs to penetrate into tumor cells
(41–43), or help get across physiological drug barriers such as
the blood–brain barrier (BBB) (44), for the treatment of brain
tumors and other central nervous system (CNS) diseases. Targeted
NPs were reported to have enhanced binding affinity and specificity
over targeted molecular drug and contrast agents, due to the
multiple numbers of targeting ligands packed on their surface, a
“multivalency effect” (45, 46).
Fourth, the NP matrix or surface coating may be able to
prevent the PDT drugs from degrading in, or interacting with,
the living biological environment. This is done by blocking the
diffusion of large enzyme molecules into the NPs. For example,
methylene blue is clinically approved for topical application to
diseased tissue (see Table 2) but has not been used for systemic
delivery as it gets reduced into an inactive form by plasma enzymes.
By being loaded inside nanoparticles, methylene blue can be pro-
tected from enzymatic reduction (see Fig. 3) and thus becomes
useful for intravenously delivered PDT applications (47).
Fifth, nanoparticles can reduce immunogenicity and side
effects of the drug. The maximum tolerated dose of the drug can
be increased, as the nontoxic (biocompatible) polymer reduces
toxicity.
Sixth, nanoparticles can also alleviate the problem posed by
the multidrug resistance (MDR) of cancer cells – which reduces the
1.0 a
0.9
Normalized Fluorescence
0.8 H
N N
N S N N S N
0.7
0.6
0.5 b
0.4
0 600 1200 1800 2400 3000 3600

Time (s)
Fig. 3. Fluorescence emission at 680 nm vs. time (after subtraction of photobleaching for
both curves) for: (a) 3 mg/mL MB-containing NPs and (b) 1 mg/mL MB, respectively,
when mixed in PBS (pH 7.2) with 0.45 mmol NADH and 0.05 mg diaphorase. The inset
shows the reduction chemistry of MB to its nonphotoactive form. Reproduced from
ref. 47 with permission from Elsevier Inc.
effectivity of most drugs – by masking the drug (i.e., entrapping it

within the NPs). This feature may significantly enhance the delivery
of drugs that are normally excluded from tumors by MDR.
Seventh, a single nanoparticle can be made multifunctional
by being loaded with multiple components such as drugs, contrast
agents, targeting ligands, or other active components, enabling
synergistic multimodal therapy, multimodal imaging, as well as
integrated imaging and therapy that can provide an individual
patient-based, image-guided therapy, by imaging the diseased tis-
sue and then treating the tissue on-site (29, 43). Multifunctionality
can be introduced to molecular agents, but this usually involves a
very complicated synthetic procedure (48, 49). Sometimes such
prepared molecular agents are not very useful as the dose for
imaging and the dose for therapy can be quite different. Note
that the nanoparticles can be loaded with multiple agents in
different ratios by using standard methods that are available for
each nanoplatform.
Due to the advantages listed above, nanoparticles have the
potential to improve many aspects of PDT. High amounts of drug
can be effectively accumulated at target tissue, thus reducing the
drug dose and the possible concomitant side effects but enhanc-
ing the PDT efficiency (50, 51). Furthermore, the drug delivery
time required for PDT or drug-light interval can be significantly
shortened (43, 52).
4. Nanoparticle-
Based PDT Agents:
Designs, Materials,
Preparation, and A variety of nanoparticle matrices and photosensitizers have been
Characterization utilized to develop nanoparticle-based PDT agents. The investi-
gated photosensitizers include single photon absorption photo-
sensitizers, such as porphyrins (e.g., Photofrin and verteporfin),
chlorins, m-THPC, phthalocyanines, methylene blue and hypericin,
as well as two-photon absorption PDT dyes, such as porphyrin
tetra(p-toluenesulfonate). The investigated nanoparticles are of
both synthetic and natural origins. Most of them are polymeric
(i.e., polymer nanoparticles or polymer-coated nonpolymeric
nanoparticles), although there are nanoparticles made of photo-
sensitizers (53) or nonpolymeric nanoparticles with conjugated
photosensitizers. The polymer nanoparticles include synthetic
polymer nanoparticles made of polylactide–polyglycolide copoly-
mers (PLGA) (50, 54–60), polyacrylamide (29, 43, 47, 61–66),
silica (61) and organically modified silica (ormosil) (61, 67–76),
and natural polymer nanoparticles made of natural proteins and
polysaccharides such as albumin (77), xanthane gum (78), colla-
gen (78), and alginate (79, 80). Nonpolymeric nanoparticles
include gold nanoparticles (52, 81), iron oxide (82–86), photon
upconverting nanoparticles made of NaYF4 nanocrystals doped

with Yb3+ and Er3+ (87, 88), quantum dots (89), and fullerenes
(90) as well as nanosized biological entities such as low-density
lipoproteins (22 nm) (91) or plant viruses (28 nm) (92). The
polymers used for surface coating include silica/ormosil (82–84,
87), dextran (85), chitosan (86), and polyethyleneimine (PEI)
(88). The nanoparticles can be prepared from monomers, by
polymerization, as in the case of polyacrylamide, silica, or ormosil,
or from preformed polymer, by precipitation, solvent evaporation,
or covalent cross-linkage, as in the case of PLGA and natural
polymer nanoparticles.
The designs of the agents are very versatile (see Fig. 4). The
most common and simplest design is the nanoparticle, as an inert
platform, loaded with photosensitizers. There are a couple of
designs utilizing FRET for two-photon absorption PDT. In one
FRET-based design, polymer-coated upconverting nanoparticles
were loaded with photosensitizers (88), and in another design,
photosensitizers and two-photon absorption dyes were loaded
into inert nanoparticles (74). The two-photon dyes or upconvert-
ing nanoparticles absorb NIR irradiation and emit visible light to
excite the photosensitizing molecules so as to avoid the tissue
penetration depth limitations. There are also other designs where
the nanoparticles are made completely of photosensitizers them-
selves (53) or where the photosensitizers are conjugated to gold
particles (52, 81). Multifunctional nanoparticle agents have been
also designed to perform not only PDT but also image contrast
enhancement or other therapies. For instance, a PDT drug and
chemotherapy drug were loaded into inert nanoplatforms for
simultaneous PDT and chemotherapy (80), PDT drugs and NIR
fluorescent dyes for PDT and fluorescence imaging (67), and
PDT drugs and iron oxide nanocrystals or gadolinium chelate for
PDT and MRI (43, 62, 65, 93). In other designs, magnetic iron
oxide nanoparticles are coated with polymer shells containing
photosensitizers (82–86) or conjugated with photosensitizers
(94) for PDT and MRI, or PDT and hyperthermia therapy,
respectively. In another design, X-ray scintillation nanoparticles
are conjugated with porphyrins for simultaneous radiotherapy
and PDT (95, 96). When the nanoparticle–photosensitizer con-
jugates are stimulated by X-rays during radiotherapy, the particles
generate visible light that can activate the photosensitizers for
PDT. These nanoparticle agents can be used for deep tumor treat-
ment as X-rays can penetrate through tissue. The scintillation
nanoparticles are made of several doped inorganic materials such
as LaF3:Ce3+, LuF3:Ce3+ CaF2:Mn2+, CaF2:Eu2+, BaFBr:Eu2+,
BaFBr:Mn2+, CaPO4:Mn2+ (95), and LaF3:Tb3+ (96).
Some of these nanoparticles are biodegradable, over a time
scale of several months, and some are not. It should be noted
that, unlike chemotherapy, the PDT efficacy does not depend on
a b
Excitation
Excitation
Polymer nanoparticle Core
FRET
Non-polymeric
nanoparticle
Core
FRET
One-photon or two-photon
photosensitizer
Polymer Shell
Excitation
Two-photon absorption dye
Non-polymeric Targeting moiety

nanoparticle
Core Crosslinker or PEG
FRET
Singlet oxygen
MRI Contrast Agent
Chemotherapy drug
Fig. 4. Various designs of nanoparticle-based PDT agents: (a) polymer nanoparticles loaded with photosensitizers and
other active components, (b) nonpolymeric nanoparticles with surface-conjugated photosensitizers, and (c) nonpolymeric
nanoparticles with surface polymer shell loaded with photosensitizers.
the drug’s release from the nanoparticles, because molecular oxygen

can enter the nanoparticles and the produced singlet oxygen can
diffuse out of the nanoparticle matrix to induce photodynamic
reaction. In such cases, the PDT efficiency of the agents depends on
the nanoparticle type (size and oxygen permeability of the matrix),

as shown by a study done with three nanoparticles of different
nondegradable matrices loaded with the same photosensitizer
(methylene blue) (61). Nondegradable nanoparticles, however,
may confront bioelimination issues for in vivo applications.
Loading of photosensitizers into nanoparticles are performed
typically by one of the three methods: encapsulation, covalent
linkage, or postloading. The encapsulation is a simple method
wherein the photosensitizers are mixed with the nanoparticle-
producing reaction mixture and are physically entrapped inside
the nanoparticle matrix during synthesis – either by steric constric-
tion or by physical interaction, such as electrostatic interactions,
hydrogen bond formation, and hydrophobic interactions, between
the photosensitizers and the nanoparticle matrix. Here, the pho-
tosensitizers may be inactivated during nanoparticle synthesis and
they may form aggregates within the nanoparticle matrix.
However, the encapsulated photosensitizers may be leached out
from the nanoparticles before the in vivo nanoparticle dose
reaches its target area, thus reducing efficiency of treatment and
possibly resulting in side effects. In the postloading method, the
photosensitizers are added to the already formed nanoparticle, in
an appropriate solvent, and the loading mechanism is governed
by the physical interactions. It is simple to do, causing no
chemical damage to the photosensitizers, and can give a high
drug-loading efficiency. However, the nanoparticle agents with
postloaded photosensitizers are more prone to premature leach-
ing than those prepared by encapsulation, with potentially bad, or
good, consequences for in vivo applications. The covalent linkage
method requires a relatively complicated preparation procedure,
compared to the above two methods. The photosensitizers are
covalently linked either with monomer molecules that are then
polymerized into nanoparticles or with already-prepared nano-
particles. This prevents any leaching-related problems during
systemic in vivo circulation. Any aggregation of photosensitizers
can be avoided by conjugating each photosensitizer to a well-
separated location within the nanoparticle matrix. It should be
noted that high loading of photosensitizers does not always lead to
high PDT efficacy due to the increased probability for drug aggre-
gation (50, 61). This is especially true for agents that are prepared
by encapsulation or physical adsorption of the photosensitizers.
The above nanoparticle-based agents have been tested in
solution, in cells, and in in vivo animal models, demonstrating
their high potential for successful future clinical applications.
The solution tests determine if the photochemical and oxygen per-
meable properties of the photosensitizer-loaded nanoparticles
are sufficiently good for PDT, by measuring the singlet oxygen
production rate, either by chemical probes (43, 83, 84, 97) or
by NIR phosphorescence of singlet oxygen (8, 77). The most
commonly used chemical probe is the water-soluble, anthracene-9,

10-dipropionic acid disodium salt (ADPA) (97). However, these
chemical probe molecules may enter the nanoparticle matrix,
leading to overestimates of the singlet oxygen production by the
nanoparticle agents. A nanoparticle-based probe has been made
to detect only the singlet oxygen exiting the photodynamic nan-
oplatforms, which is critical for determining the singlet oxygen
production efficiency of the nanoparticle-based PDT agents (98).
The in vitro tests can demonstrate the cell-penetration and/or
intracellular-localization properties of the PDT agents as well as
their effectiveness in killing the target cells. The PDT cell kill is
determined typically by a live/dead cell staining assay (61), a
clonogenic assay (99), or a MTT assay (100) (see Fig. 5).
Tumor cells
or bacteria
Petri dish
with cells
Add Calceine-AM and PI

Incubation with
Photosensitizer
Illumination of light
Fluorescence Imaging
(Green fluorescence for live cells;
Red fluorescence for dead cells) Add MTT and incubate 2-4 hours
In vitro culture
(See Note 1)
Growing into
Remove cell media; add DMSO; a colony
and shake the mixture slowly (1-14days)
overnight to dissolve
water-insoluble formazan
Counting
colonies
Measure absorbance of the
dissolved formazan at 550 nm.
Reproduced from ref.61 (See Note 2)
with permission
Higher the intensity, lower the PDT
from Wiley
efficiency (Healthier cells)
LIve/Dead Cell Staining Assay MTT Assay Clonogenic Assay
Fig. 5. In vitro tests for PDT efficacy determinations. Note 1: The nonfluorescent Calcein-AM is converted into calcein by
esterase in a live cell, emitting strong green fluorescence (excitation: 490 nm, emission: 515 nm). PI enters the dying or dead
cells through disrupted membranes and intercalates with nuclear DNA, emitting red fluorescence (excitation: 535 nm, emis-
sion: 617 nm). The optimal concentrations of Calcein-AM and PI should be determined for each tested cell line. Note 2: The
colored formazan solution has an absorption peak around 500–600 nm. The absorption maximum varies with the solvent
employed – for instance, 550 nm for DMSO. The MTT assay depends on reductase enzymes whose activity can be changed
upon experimental conditions and therefore sometimes produce erratic results.
The in vivo tests demonstrate the effectiveness of the PDT

agents. This efficacy is governed by many additional factors, such
as the pharmacokinetics and tissue distribution of the nanopar-
ticles. The in vivo PDT efficiency is determined by monitoring
tumor volume reduction, or animal survival, which is typically
displayed by a Kaplan–Meier plot (43).
In subsequent sections, recent PDT applications of nanopar-
ticles, especially of polymeric nanoparticles, are shown.
5. Examples
of Polymeric
Nanoparticle-
Based PDT Agents The polyacrylamide (PAA) nanoparticle is a hydrogel that is
produced through microemulsion polymerization. Historically, it
5.1. Polyacrylamide was first produced as a “nano-PEBBLE” intracellular fluores-
Nanoparticles cence-based sensor, which produces singlet oxygen while sensing
(being quenched by) oxygen (101). It can be functionalized with
amine or carboxyl groups, for surface modification with a target-
ing moiety and/or with PEG, and loaded with a combination of
actuating molecules that make it suitable for multifunctional
tasks, including PDT (102). The PAA nanoparticle is highly
soluble in water, offering easy systemic administration without
aggregation. The PAA nanoparticle is also biologically inert and
nontoxic, which was demonstrated by in vivo toxicity studies,
showing no evidence of alterations in histopathology or clinical
chemistry for the PAA nanoparticle dose of 10 mg/kg–1 g/kg
(29), as well as by a report on safety assessment of PAA (103).
It can also be engineered to have controllable biodegradability by
introducing different amounts of biodegradable cross-linkers (29).
PAA nanoparticles have been utilized as a platform to pro-
duce PDT agents as well as multifunctional agents for tumor-
selective PDT and MRI. The PAA nanoparticles typically have
the size range of 30–70 nm and have been loaded with methyl-
ene blue (47, 61, 66), Photofrin (43, 62, 65), and two-photon
absorption photosensitizers including 5,10,15,20-tetrakis(1-methyl
4-pyridinio) porphyrin tetra(p-toluenesulfonate) (TMPyP) (63).
The PAA nanoparticles were prepared with either biodegradable
(43) or nonbiodegradable cross-linkers (47, 61–65), and
both have shown an effective photodynamic activity in cancer
cells such as 9L glioma, C6 glioma, and MBA-MD-435 cells.
Moreover, targeted PAA nanoparticles with encapsulated
Photofrin and surface-conjugated F3 peptide were demon-
strated to kill target cells selectively in vitro (43). The same PAA
nanoparticles showed successful in vivo PDT therapy in 9L
glioma bearing rats when injected in the rat tail vein (43).
Treatment of the tumor-bearing rats with F3-targeted nanopar-
ticles, followed by 7.5 min of red light irradiation through an
inserted optical fiber, showed a significant improvement in

survival rate compared to various sets of control rats: (1) those
that received the same light treatment but with nontargeted
nanoparticles or Photofrin, (2) those that only received the light
treatment, and (3) those that received no treatment. Indeed,
60 days later, 40% of the animals treated with F3-targeted
Photofrin nanoparticles were found to be tumor free, and this
was still the case 6 months later. In contrast, all control rats were
dead within 2 weeks. It should be noted that this demonstrates
that the surface-conjugated targeting moieties do make a sig-
nificant difference in animal survival. These nanoparticles also
showed enhanced in vivo MRI tumor contrast due to coencap-
sulated iron oxide, thus enabling image-guided therapy.
The PAA nanoparticles with encapsulated methylene blue
also showed good potential as a PDT agent to eliminate bacterial
infections (66). The PDT with the nanoparticles was performed
on bacterial cells – of both gram-positive strains (i.e., Staphylococcus
aureus strain) and gram-negative strains (i.e., Escherichia coli,
Pseudomonas aeruginosa, and Acinetobacter sp.) – in suspension
as well as in preformed biofilms. The data revealed that the nano-
particles can inhibit biofilm growth and eradicate almost all of the
early age biofilms that are formed by all of the bacteria examined.
However, the killing efficiency for gram-positive strains was much
higher than that for gram-negative strains. PAA nanoparticles
were also prepared with ultrasmall size (2–3 nm), for easy renal
clearance from the body, and were successfully loaded with a
hydrophobic photosensitizer (m-THPC), despite the hydrophilic
nature of PAA (64). These ultrasmall nanoparticle agents were as
effective in killing cultured cancer cells as free m-THPC photo-
sensitizer molecules.
5.2. PLGA Poly (d,l-lactide-co-glycolide) (PLGA) is a biodegradable polymer

Nanoparticles that is usually prepared using either emulsion solvent evaporation
or solvent displacement techniques (104). PLGA particles are
made of a mixture of a less hydrophilic lactic acid polymer (PLA)
and the more hydrophilic glycolic acid polymer (PGA). They
undergo hydrolysis, in the body, to form lactic acid and glycolic
acid, which are eventually removed from the body by the citric acid
cycle. Owing to their biodegradability and ease of formulation to
carry a variety of drugs, these have been the most extensively
investigated polymer nanoplatforms for drug delivery (105).
PLGA nanoparticles loaded with a variety of photosensitiz-
ers including porphyrins, chlorins, phthalocyanines, and hyper-
icin were studied for PDT efficacy in vitro and in vivo. These
PLGA nanoparticles exhibited a higher photoactivity than free
PDT drugs in cells such as EMT-6 mammary tumor (54)
and NuTu-19 ovarian cancer cells (50), and in an in vivo
chick embryo chorioallantoic membrane (CAM) model (56).
The meso-tetraphenylporpholactol loaded 50:50 PLGA nanoparticles

that were 98 nm in diameter showed complete eradication of
subcutaneous implanted U587 gliomas in nude mice at 27 days
(191 mW/cm2, 57.3 J/cm2) after the light therapy, which was
performed at 24 h after i.v. injection of the nanoparticles (60).
It should be noted that these nanoparticles themselves have very
low fluorescence and singlet oxygen production efficiency prob-
ably due to aggregation of the incorporated dye, indicating that
the dye released from the nanoparticles contributed to such a
PDT outcome.
The PDT efficacy of the PLGA nanoparticle-based agents was
found to depend on the lipophilicy of the incorporated photosen-
sitizers, and the properties (the copolymer molar ratio and the
size) of the PLGA nanoparticles. A study with PLGA nanoparti-
cles loaded with three photosensitizers of different lipophilicity –
meso-tetraphenylporphyrin (TPP), meso-tetra-(4-carboxyphenyl)
-porphyrin (TCPP), and chlorin e6 (Ce6) – showed that the
higher the photosensitizer’s lipophilicity, the higher the incorpo-
ration efficiency, the lower the amount of the extravasation (i.e.,
longer residence time in the blood vessels) and, therefore, the
higher their photothrombic efficiency, when tested in an in vivo
CAM model (58). Note that the developing chicken embryo is
surrounded by a chorioallantoic membrane (CAM), which
becomes vascularized as the embryo develops. The CAM is a con-
venient model for monitoring the modifications of the vascula-
ture as the transparency of its superficial layers allows an
examination of structural changes of each blood vessel in real
time (56). The molar ratio between PLA and PGA determines
the biodegradation rate (106) and the lipophilicity of the parti-
cles, affecting the in vitro phototoxicity of incorporated photo-
sensitizers, meso-tetra(hydroxyphenyl)porphyrin (m-THPP), in
the order of 50:50 PLA:PGA > 75:25 PLA:PGA > PLA (54). The
PLGA particle size seems to play an important role on PDT as
demonstrated by a couple of studies. In one study (55), 75:25
PLGA nanoparticles of two different sizes (167 and 370 nm in
diameter) were loaded with verteporfin and tested in vitro and
in vivo. The smaller nanoparticles exhibited greater photocyto-
toxicity on EMT-6 mammary tumor cells compared to the larger
NPs or to the free drug. The smaller particles also released
m-THPP faster than the free photosensitizer in DMSO/PBS. In
vivo PDT with mice bearing rhabdomyosarcoma (M1) tumor in
the right dorsal area indicated that the i.v. administered PLGA
particles (167 nm) with incorporated verteporfin effectively
reduced tumor growth for 20 days in mice with short drug-to-
light intervals (15 and 30 min). In another study (107), the 50:50
PLGA nanoparticles loaded with m-THPP, with three different
mean diameters of 117, 285, and 593 nm, respectively, were
prepared. The nanoparticles of 117 nm exhibited the highest rate
of reactive oxygen species production and the fastest m-THPP

release in vitro. Similarly, the 117-nm-sized nanoparticles exhib-
ited the highest in vivo photodynamic activity in the CAM model.
The amount of residual poly(vinyl alcohol) (PVAL) stabilizing
agent during the particle synthesis, at the particle surface, were
shown to affect both the release and activity of the photosensitiz-
ers for the bigger particles (285 and 593 nm) (57). In the other
study, four batches of PLA nanoparticles containing meso-
tetra(carboxyphenyl)porphyrin (TCPP), with different mean sizes
ranging from 121 to 343 nm, were prepared (59). The extravasa-
tions of each TCPP-loaded nanoparticle formulation from blood
vessels were measured, as well as the extent of photochemically
induced vascular occlusion in the CAM model, both revealing
size-dependent behaviors. The smallest nanoparticles (121 nm)
exhibited the greatest extent of vascular thrombosis as well as the
lowest extravasation.
5.3. Silica and Ormosil Silica, an inorganic polymer, and organically modified silica
Nanoparticles (ormosil) nanoparticles are inert but nonbiodegradable. The sur-
face can be easily coated with additional functionalized silica or an
ormosil layer for further surface modifications. The size of silica
or ormosil nanoparticles used to prepare PDT agents has been
reported in the range of 100–200 nm when produced by the
Stober process (61, 68), a two-step acid/base-catalyzed proce-
dure, (61, 108) or a surfactant template procedure (76), but the
size range has been 20–40 nm when prepared in a micelle (67, 70,
73, 109).
The silica or ormosil nanoparticles have been loaded with
various photosensitizers such as protoporphyrin IX (PplX), elsino-
chrome A, HPPH, m-THPC, and methylene blue by encapsula-
tion (61, 67–69, 109) and covalent linkage (70–72, 76). The
silica/ormosil nanoparticles with both encapsulated and cova-
lently linked photosensitizers produced singlet oxygen and
showed in vitro PDT efficiency. However, the leaching of encap-
sulated photosensitizers was observed for the small (20–30 nm)
ormosil nanoparticles containing m-THPC and HPPH (67, 73)
in the presence of organic solvent or serum. The ormosil nano-
particles of 20-nm size with covalently linked iodobenzylpy-
ropheophorbide (70) and the mesoporous ormosil nanoparticles
of ~100-nm size with PplX (76) showed no leaching but high
PDT cell-killing efficiency, indicating high loading of monomeric
photosensitizers without aggregation. It also demonstrates the
high singlet oxygen permeability of the ormosil matrix.
The ormosil nanoparticles have also been utilized to prepare
various PDT agents of synergistic designs by loading another
component in addition to the photosensitizers. In one design, the
ormosil nanoparticles were loaded with HPPH and an excess of
9,10-bis [4¢-(4″-aminostyryl)styryl]anthracene (BDSA), a highly
two-photon-active molecule (74). HPPH absorption in nanopar-

ticles has significant overlap with the fluorescence of BDSA aggre-
gates, enabling indirect two-photon excitation (850 nm) of
HPPH. Drastic changes in the morphology of the cells were
observed, which were indicative of impending death as a result of
two-photon PDT with these nanoparticles. In another design
(75), ormosil nanoparticles containing covalently loaded iodine as
a second component that induces the intraparticle external heavy-
atom effect on the encapsulated photosensitizer molecules signifi-
cantly enhanced the efficiency of singlet oxygen generation and,
thereby, the in vitro PDT efficacy. Furthermore, the ormosil
nanoparticles were loaded with fluorescence dyes (67) and target-
ing moieties (72) for multitasking (PDT and fluorescent imaging)
and for target-specific PDT, respectively.
5.4. Dendrimers Dendrimers are polymeric materials that are highly branched and
monodisperse macromolecules. Dendrimers have been adopted
as a platform to deliver 5-Aminolevulinic acid (ALA). ALA is a
precursor of protoporphyrin IX (PpIX) that is used as an endog-
enous photosensitizer and is FDA-approved for topical PDT
(see Table 2). However, due to the hydrophilic nature of ALA,
ALA–PDT may be limited clinically by the rate of ALA uptake
into neoplastic cells and/or penetration into tissue (110). A sec-
ond-generation dendrimer with conjugated ALA, containing 18
aminolevulinic acid residues attached via ester linkages to a mul-
tipodent aromatic core, showed enhancement of porphyrin syn-
thesis in vitro (111) and in vivo (112). In vitro, the dendrimer-ALA
was more efficient than ALA for porphyrin synthesis in trans-
formed PAM 212 murine keratinocyte and A431 human epider-
moid carcinoma cell lines, up to an optimum concentration of
0.1 mmol/L (111). In an in vivo study with male BALB/c mice,
the porphyrin kinetics from ALA exhibited an early peak between
3 and 4 h in most tissues, whereas the dendrimer induced sus-
tained porphyrin production for over 24 h, and basal values were
not reached until 48 h after administration (112). Integrated
porphyrin accumulation from the dendrimer and that from an
equal drug equivalent dose of ALA was comparable, showing that
the majority of ALA residues were liberated from the dendrimer.
The porphyrin kinetics appears to be governed by the rate of
enzymatic cleavage of ALA from the dendrimer, which is consistent
with the in vitro results.
5.5. Polymer NPs Nanoparticles made of cross-linked natural polymers such as

Based on Natural proteins and polysaccharides have been prepared and their PDT
Polymers efficacy was investigated. The investigated proteins include human
serum albumin (HSA) and bovine serum albumin (BSA), a most
abundant protein in blood plasma. The polysaccharides include
xanthan gum and alginate, both of which have been used as a
food additive and rheology modifier.
For example, cross-linked HSA nanoparticles (100–300 nm

in diameter) were loaded with photosensitizers, pheophorbide
(77). The singlet oxygen quantum yield of the HSA nanoparticles
was very low probably due to the attachments of the photosensi-
tizers to proteins and their aggregation within the nanoparticles.
When the Jurkat cells were incubated with these nanoparticles,
the photosensitizer was released due to nanoparticle decomposi-
tion in the cellular lysosomes, which was illustrated by fluores-
cence lifetime imaging and confocal laser scanning microscopy.
After a 24-h incubation period, the nanoparticles showed a
higher phototoxicity but lower dark toxicity than free pheophor-
bide, demonstrating their potential as a PDT agent. BSA nano-
particles containing the photosensitizer silicon (IV) phthalocyanine
(NzPc) (452 nm) were reported to show in vitro phototoxicity
in human fibroblasts (113). The same nanoplatform containing
both magnetic nanoparticles and NzPc were also prepared for
potential use in PDT combined with hyperthermia. Both BSA
nanoparticles reveal no dark cytotoxicity (113).
Marine atelocollagen/xanthan gum microcapsules with encap-
sulated photosensitizer, 5,10,15-triphenyl-20-(3-N-methylpyri-
dinium-yl)porphyrin, were prepared (78). The size of these
polymeric formulations was relatively large compared to other
nanoplatforms (i.e., 100–1,000 nm in diameter even after size
reduction by ultrasonic processing in the presence of Tween 20
surfactant). These polymer particles showed about four times
more phototoxicity than the respective phosphatidylcholine lipidic
emulsion but negligible cytotoxicity toward HeLa cells.
Aerosol OT (AOT)-alginate nanoparticles were also devel-
oped into PDT agents by encapsulating methylene blue. These
nanoparticles were determined to be ~80 nm in size by dynamic
light scattering (DLS) and showed significantly enhanced overall
cytotoxicity, following PDT, in comparison to free methylene
blue, in two cancer cell lines, MCF-7 and 4T1 (79). Such enhanced
in vitro PDT efficiency results from increased ROS and singlet
oxygen production by the nanoparticles and significant nuclear
localization of the methylene blue released from the nanoparti-
cles. The same nanoplatforms, loaded with both methylene blue
and doxorubicine (with measured diameter of 39 nm by AFM
and of 62 nm by DLS), were also developed for a combination
therapy of chemotherapy and PDT (80). Both methylene blue
and doxorubicin were released from the nanoparticles, which
resulted in a significantly higher nuclear accumulation of doxo-
rubicin and methylene blue than that with the free drugs.
Furthermore, methylene blue serves not only as a PDT drug but
also as a P-glycoprotein inhibitor, enhancing doxorubicin-induced
cytotoxicity in drug-resistant NCI/ADR-RES cells. Nanoparticle-
mediated combination therapy resulted in a significant induction
of both apoptosis and necrosis and improved cytotoxicity in
drug-resistant tumor cells.
5.6. Polymer-Coated Two kinds of inorganic nanoparticles, photon upconverting and

Inorganic iron oxide nanoparticles, have been developed as PDT agents after
Nanoparticles being coated with a polymer layer containing photosensitizers.
Photon upconverting nanoparticles with surface polymer
coatings have been developed for deep tissue penetrating PDT,
where the upconverting nanoparticles were made of NaYF4:Yb3+,
Er3+. Through excitation with multiple photons, these nanopar-
ticles convert lower-energy (NIR) light to higher-energy (visible)
light, which then excites the photosensitizing molecules loaded in
the surface-coated polymer. The silica-coated upconverting nano-
particles were loaded with photosensitizers, merocyanine, and
also functionalized with a monoclonal antibody to target breast
cancer cells (87). The resultant nanoparticles were 60–120 nm in
diameter and performed efficient PDT after infrared (974 nm)
irradiation. The PEI-coated upconverting nanoparticles were
loaded with zinc phthalocyanine photosensitizers and conjugated
with folic acid (88). These 50-nm nanoparticles produced green/
red emission with near-infrared (NIR) excitation, both in vitro
and in vivo, demonstrating that these particles could be excited
after deep intramuscular injection in rats. The particles showed
the ability of significant cell destruction with NIR excitation after
targeted binding to cancer cells (HT29 cells).
Iron oxide nanoparticles (typical core size 5–10 nm) have
been coated with silica (82–84), dextran (85), and chitosan (86)
to produce 20–100 nm nanoparticles. Dextran is a branched
polysaccharide and chitosan is a linear polysaccharide. The photo-
sensitizers were loaded into the polymer shell, to serve for PDT,
and sometimes for fluorescence imaging, while the iron oxide
particle core provides magnetic properties for MRI contrast
enhancement. The investigated photosensitizers included an
iridium complex (82), a porphyrine (PHPP) (83) and methylene
blue (84) encapsulated in the silica shell, PHPP encapsulated in a
chitosan shell (86), and a chlorin (TPC) covalently linked to the
dextran shell (85). These nanoparticles showed PDT capability in
solution (84), in vitro on SW480 carcinoma cells (83, 86), Hela
cells (82), murine and human macrophages (85), as well as in vivo
(86) on SW480 carcinoma bearing mice. Interestingly, in the
in vivo study, magnetic targeting was performed before PDT,
which improved tumor accumulation and retention of the nano-
particles (86).
6. Summary
and Perspectives
PDT is a rapidly evolving treatment with significant advantages
over other therapeutic modalities. PDT has been approved for
several cancer applications (see Table 2). However, it has not yet
been used for mainstream oncologic practice (i.e., treatment

of patients with solid tumors). This results from (a) PDT’s depen-
dence on (1) tissue penetration depth by light and (2) oxygen
levels in the target tissue, which is usually a problem in solid
tumors, as well as (b) inefficient delivery of the PDT drugs to the
target tissue, in current clinical practice. Nanoparticles have many
advantages as a delivery vehicle, with potential to improve the
efficacy of existing imaging and therapy methods, especially for
cancer detection and treatments. A variety of nanoparticle
matrices, photosensitizers, other active chemicals and targeting
components have been utilized to make nanoparticle-based PDT
agents that can improve current PDT efficacy. Many of these
nanoparticles have excellent optical and tumor-targeting quali-
ties, qualities that are not achievable with molecular PDT drugs.
So far, no nanoparticle-based PDT agents have been used in the
clinic, although the in vitro and in vivo data from nanoparticle-
based PDT agents seem to promise a very high potential for
clinical applications of these nanoparticles, provided that several
challenging issues are resolved. The bioelimination profile is an
important issue to be considered, as it is liable to have a kinetic
pattern that is quite different from that of currently used molec-
ular-type drugs and imaging agents. The other challenging issue
is the need for a new clinical (FDA) approval procedure. The
latter is expected to be “complicated,” as these NP-based agents
are made of multiple active ingredients. It should be noted that
these issues are common to all nanoparticle-based chemothera-
peutic agents. In spite of this, several of them have already reached
the clinic for chemotherapy (38). Therefore, nanoparticle-based
PDT agents are also expected to advance toward clinical applica-
tions in the near future. Moreover, with these advanced nano-
particle agents, PDT may become useful in other potential
clinical applications.
Acknowledgments
This article is partially supported by funding from NIH grants

1R01EB007977, R33CA125297 03S1, and R21/R33CA125297.
References
1. Siegel, M. M., Tabei, K., Tsao, R., Pastel, M. mass spectrometry and laser desorption/
J., Pandey, R. K., Berkenkamp, S., et al. jet-cooling photoionization mass spectrometry.
(1999) Comparative mass spectrometric J. Mass Spectrom. 34, 661–669.
analyses of photofrin oligomers by fast atom 2. Mitton, D. and Ackroyd, R. (2008) A brief
bombardment mass spectrometry, UV and overview of photodynamic therapy in
IR matrix-assisted laser desorption/ionization Europe. Photodiagn. Photodyn. Ther. 5,
mass spectrometry, electrospray ionization 103–111.
3. Dolmans, D. E. J. G. J., Fukumura, D., and ICP, and light penetration in the brain.
Jain, R. K. (2003) Photodynamic therapy for Photochem. Photobiol. 46, 929–935.
cancer. Nat. Rev. Cancer 3, 380–387. 17. Dougherty, T. J., Weishaupt, K. R., and
4. Brown, S. B., Brown, E. A., and Walker, I. Boyle, D. G. (1985) Photodynamic sensitiz-
(2004) The present and future role of photo- ers. in: Cancer: Principles and Practice of
dynamic therapy in cancer treatment. Lancet Oncology (DeVita, Jr., V. T. and Rosenberg,
Oncol. 5, 497–508. S. A., eds.) pp. 2272–2279, J. B. Lippincott
5. Wilson, B. C. and Patterson, M. S. (2008) Company, Philadelphia.
The physics, biophysics and technology of 18. Kavar, B. and Kaye, A. H. (2007)
photodynamic therapy. Phys. Med. Biol. 53, Photodynamic therapy. in: High-Grade
R61–R109. Gliomas: Diagnosis and Treatment (Barnett,
6. Buytaert, E., Dewaele, M., and Agostinis, P. G. E., Ed.) pp. 461–484, Humana Press,
(2007) Molecular effectors of multiple cell Totowa.
death pathways initiated by photodynamic 19. http://www.accessdata.fda.gov/drugsatfda_
therapy. Biochim. Biophys. Acta 1776, docs/label/2008/020451s019lbl.pdf.
86–107. 20. http://www.photofrin.com/pdf/patient-
7. Foote, C. S. (1991) Definition of type I and guide.pdf.
type II photosensitized oxidation. Photochem. 21. Starkey, J. R., Rebane, A. K., Drobizhev, M.
Photobiol. 54, 659. A., Meng, F., Gong, A., Elliott, A., et al.
8. Niedre, M., Patterson, M. S., and Wilson, B. (2008) New two-photon activated photody-
C. (2002) Direct near-infrared luminescence namic therapy sensitizers induce xenograft
detection of singlet oxygen generated by tumor regressions after near-IRLaser treat-
photodynamic therapy in cells in vitro and ment, through the body of the host mouse.
tissues in vivo. Photochem. Photobiol. 75, Clin. Cancer Res. 14, 6564–6573.
382–391. 22. Mir, Y., Houde, D., and van Lier, J. E. (2008)
9. Tsushima, M., Tokuda, K., and Ohsaka, T. Photodynamic inhibition of acetylcholinest-
(1994) Use of hydrodynamic chronocou- erase after two-photon excitation of copper
lometry for simultaneous determination of tetrasulfophthalocyanine. Lasers Med. Sci.
diffusion-coefficients and concentrations of 23, 19–25.
dioxygen in various media. Anal. Chem. 66, 23. Josefsen, L. B. and Boyle, R. W. (2008)
4551–4556. Photodynamic therapy: novel third-genera-
10. Pryor, W. A. (1986) Oxy-radicals and related tion photosensitizers one step closer? Br.
species: their formation, lifetimes, and reac- J. Pharmacol. 154, 1–3.
tions. Ann. Rev. Physiol. 48, 657–667. 24. http://www.dusapharma.com/levulan-pre-
11. Morgan, J. and Oseroff, A. R. (2001) scribing-information.html.
Mitochondria-based photodynamic anti-cancer 25. http://www.biolitecpharma.com/public/
therapy. Adv. Drug Deliv. Rev. 49, 71–86. smpc.asp?s=foscan.
12. Eljamel, M. S., Goodman, C., and Moseley, 26. Dolmans, D. E. J. G. J., Kadambi, A., Hill, J.
H. (2008) ALA and Photofrin® fluorescence- S., Flores, K. R., Gerber, J. N., Walker, J. P.,
guided resection and repetitive PDT in glio- et al. (2002) Targeting tumor vasculature and
blastoma multiforme: a single centre Phase cancer cells in orthotopic breast tumor by
III randomised controlled trial. Lasers Med. fractionated photosensitizer dosing photody-
Sci. 23, 361–367. namic therapy. Cancer Res. 62, 4289–4294.
13. Vogel, A. and Venugopalan, V. (2003) 27. Seshadri, M., Bellnier, D. A., Vaughan, L. A.,
Mechanisms of pulsed laser ablation of bio- Spernyak, J. A., Mazurchuk, R., Foster, T.
logical tissues. Chem. Rev. 103, 577–644. H., et al. (2008) Light delivery over extended
14. Stolik, S., Delgado, J. A., Perez, A., and time periods enhances the effectiveness of
Anasagasti, L. (2000) Measurement of the PDT. Clin. Cancer Res. 14, 2796–2805.
penetration depths of red and near infrared 28. Henderson, B. W., Busch, T. M., and Snyder,
light in human “ex vivo” tissues. J. Photochem. J. W. (2006) Fluence rate as a modulator of
Photobiol. 57, 90–93. PDT mechanism. Laser Surg. Med. 18,
15. Svaasand, L. O., Gomer, C. J., and Profio, 489–493.
A. E. (1989) Laser-induced hyperthermia of 29. Koo, Y. L., Reddy, G. R., Bhojani, M.,
ocular tumors. Appl. Opt. 28, 2280–2287. Schneider, R., Philbert, M. A., Rehemtulla,
16. Muller, P. J. and Wilson, B. C. (1987) A., et al. (2006) Brain cancer diagnosis and
Photodynamic therapy of malignant primary therapy with nano-platforms. Adv. Drug
braintumours: clinical effects, postoperative Deliv. Rev. 58, 1556–1577.
30. Redmond, R. W., Land, E. J., and Truscott, nuclei of tumor cells and tumor endothelial
T. G. (1985) Aggregation effects on the cells in vivo. Proc. Natl. Acad. Sci. U S A 99,
photophysical properties of porphyrins in 7444–7449.
relation to mechanisms involved in photody- 42. Allen, T. M. (2002) Ligand-targeted thera-
namic therapy. Adv. Exp. Med. Biol. 193, peutics in anticancer therapy. Nat. Rev.
293–302. Cancer 2, 750–763.
31. Severino, D., Junqueira, H. C., Gugliotti, 43. Reddy, G. R., Bhojani, M. S., McConville P.,
M., Gabrielli, D. S., and Baptista, M. S. Moody, J., Moffat, B. A., Hall, D. E., et al.
(2003) Influence of negatively charged inter- (2006) Vascular targeted nanoparticles for
faces on the ground and excited state proper- imaging and treatment of brain tumors. Clin.
ties of methylene blue. Photochem. Photobiol. Cancer Res. 12, 6677–6686.
77, 459–468. 44. Costantino, L., Gandolfi, F., Tosi, G., Rivasi,
32. Takakura, Y., Mahato, R. I., and Hashida, M. F., Vandelli, M. A., and Forni, F. (2005)
(1998) Extravasation of macromolecules. Peptide-derivatized biodegradable nanopar-
Adv. Drug Deliv. Rev. 34, 93–108. ticles able to cross the blood–brain barrier.
33. Weissleder, R., Bogdanov Jr., A., Tung, C. J. Controlled Release 108, 84–96.
H., and Weinmann, H. J. (2001) Size opti- 45. Hong, S., Leroueil, P. R., Majoros, I. J., Orr,
mization of synthetic graft copolymers for B. G., Baker, J. R., and Holl, M. M. B.
in vivo angiogenesis imaging. Bioconjug. (2007) The binding avidity of a nanoparti-
Chem. 12, 213–219. cle-based multivalent targeted drug delivery
34. Gaumet, M., Vargas, A., Gurny, R., and platform. Chem. Biol. 14, 107–115.
Delie, F. (2008) Nanoparticles for drug 46. Montet, X., Funovics, M., Montet-Abou, K.,
delivery: the need for precision in reporting Weissleder, R., and Josephson, L. (2006)
particle size parameters. Eur. J. Pharm. Multivalent effects of RGD peptides obtained
Biopharm. 69, 1–9. by nanoparticle display. J. Med. Chem. 49,
35. Maeda, H. (2001) The enhanced permeabil- 6087–6093.
ity and retention (EPR) effect in tumor vas- 47. Tang, W., Xu, H., Park, E. J., Philbert, M.
culature: the key role of tumor-selective A., and Kopelman, R. (2008) Encapsulation
macromolecular drug targeting. Adv. Enzyme of methylene blue in polyacrylamide nano-
Regul. 41, 189–207. particle platforms protects its photodynamic
36. Matsumura, Y. and Maeda, H. (1986) A new effectiveness. Biochem. Biophys. Res. Commun.
concept for macromolecular therapeutics in 369, 579–583.
cancer chemotherapy: mechanism of tumori- 48. Chen, Y., Gryshuk, A., Achilefu, S.,
tropic accumulation of proteins and the anti- Ohulchansky, T., Potter, W., Zhong, T., et al.
tumor agent smancs. Cancer Res. 46, (2004) A novel approach to a bifunctional
6387–6392. photosensitizer for tumor imaging and pho-
37. Vinogradov, S. V., Bronich, T. K., and totherapy. Bioconjug. Chem. 8, 1105–1115.
Kabanov, A. V. (2002) Nanosized cationic 49. Li, G., Slansky, A., Dobhal, M. P., Goswami,
hydrogels for drug delivery: preparation, L. N., Graham, A., Chen, Y., et al. (2005)
properties and interactions with cells. Adv. Chlorophyll-a analogues conjugated with
Drug Deliv. Rev. 54, 135–147. aminophenyl-DTPA as potential bifunctional
38. Davis, M. E., Chen, Z., and Shin, D. M. agents for magnetic resonance imaging and
(2008) Nanoparticle therapeutics: an emerg- photodynamic therapy. Bioconjug. Chem. 16,
ing treatment modality for cancer. Nat. Rev. 32–42.
Drug Discov. 7, 771–782. 50. Zeisser-Labouebe, M., Lange, N., Gurny, R.,
39. Jain, R. K. (2001) Delivery of molecular and and Delie, F. (2006) Hypericin-loaded nano-
cellular medicine to solid tumors. Adv. Drug particles for the photodynamic treatment
Deliv. Rev. 46, 149–168. of ovarian cancer. Int. J. Pharm. 326,
40. Moffat, B. A., Reddy, G. R., McConville, P., 174–181.
Hall, D. E., Chenevert, T. L., Kopelman, R. 51. Saxena, V., Sadoqi, M., and Shao, J. (2006)
R., et al. (2003) A novel polyacrylamide Polymeric nanoparticulate delivery system
magnetic nanoparticle contrast agent for for Indocyanine green: biodistribution in
molecular imaging using MRI. Mol. Imaging healthy mice. Int. J. Pharm. 308, 200–204.
2, 324–332. 52. Cheng, Y., Samia, A. C., Meyers, J. D.,
41. Porkka, K., Laakkonen, P., Hoffman, J. A., Panagopoulos, I., Fei, B., and Burda, C. (2008)
Bernasconi, M., and Ruoslahti, E. (2002) A Highly efficient drug delivery with gold
fragment of the HMGN2 protein homes to the nanoparticle vectors for in vivo photodynamic
therapy of cancer. J. Am. Chem. Soc. 130, platforms for in vivo MRI enhancement and
10643–10647. photodynamic therapy of a rat brain cancer.
53. Hu, Z., Pan, Y., Wang, J., Chen, J., Li, J., and J. Magn. Magn. Mater. 293, 404–410.
Ren, L. (2009) Meso-tetra (carboxyphenyl) 63. Gao, D., Agayan, R. R., Xu, H., Philbert, M.
porphyrin (TCPP) nanoparticles were inter- A., and Kopelman R. (2006) Nanoparticles
nalized by SW480 cells by a clathrin-mediated for two-photon photodynamic therapy in
endocytosis pathway to induce high photocy- living cells. Nano Lett. 6, 2383–2386.
totoxicity. Biomed. Pharmacother. 63, 64. Gao, D., Xu, H., Philbert, M. A., and
155–164. Kopelman R. (2007) Ultrafine hydrogel par-
54. Konan, Y. N., Berton, M., Gurny, R., and ticles: synthetic approach and therapeutic
Allemann, E. (2003) Enhanced photody- application in living cells. Angew. Chem. Int.
namic activity of meso-tetra(4-hydroxyphe- Ed. 46, 2224–2227.
nyl)porphyrin by incorporation into 65. Ross, B., Rehemtulla, A., Koo, Y. L., Reddy,
sub-200 nm nanoparticles. Eur. J. Pharm. G. R., Behrend, C., Buck, S., et al. (2004)
Sci. 18, 241–249. Photonic and magnetic nanoexplorers for
55. Konan-Kouakou, Y. N., Boch, R., Gurny, R., biomedical use: from subcellular imaging to
and Allemann, E. (2005) In vitro and in vivo cancer diagnostics and therapy. Proc. SPIE
activities of verteporfin-loaded nanoparticles. 5331, 76–83.
J. Controlled Release 103, 83–91. 66. Wu, J. F., Hao, X., Wei, T., Kopelman, R.,
56. Vargas, A., Pegaz, B., Debefve, E., Konan- Philbert, M. A., and Xi, C. (2009) Eradication
Kouakou, Y., Lange, N., Ballini, J.-P., et al. of bacteria in suspension and biofilms using
(2004) Improved photodynamic activity of methylene blue-loaded dynamic nanoplat-
porphyrin loaded into nanoparticles: an forms. Antimicrob. Agents Chemother. 53,
in vivo evaluation using chick embryos. Int. 3042–3048.
J. Pharm. 286, 131–145. 67. Compagnin, C., Bau, L., Mognato, M.,
57. Vargas, A., Lange, N., Arvinte, T., Cerny, R., Celotti, L., Miotto, G., Arduini, M., et al.
Gurny, R., and Delie, F. (2009) Toward the (2009) The cellular uptake of meta-tetra
understanding of the photodynamic activity (hydroxyphenyl)chlorin entrapped in organi-
of m-THPP encapsulated in PLGA nanopar- cally modified silica nanoparticles is mediated
ticles: correlation between nanoparticle by serum proteins. Nanotechnology 20,
properties and in vivo activity. J. Drug 345101.
Targeting 17, 599–609. 68. Yan, F. and Kopelman, R. (2003) The embed-
58. Pegaz, B., Debefve, E., Borle, F., Ballini, J., ding of metatetra(hydroxyphenyl)-chlorin
van den Bergh, H., and Kouakou-Konan, Y. into silica nanoparticle platforms for photo-
N. (2005) Encapsulation of porphyrins and dynamic therapy and their singlet oxygen
chlorins in biodegradable nanoparticles: the production and pH dependent optical prop-
effect of dye lipophilicity on the extravasa- erties. Photochem. Photobiol. 78, 587–591.
tion and the photothrombic activity. A com- 69. Zhou, L., Dong, C., Wei, S. H., Fenga, Y. Y.,
parative study. J. Photochem. Photobiol. B 80, Zhoua, J. H., and Liu, J. H. (2009) Water-
19–27. soluble soft nano-colloid for drug delivery.
59. Pegaz, B., Debefve, E., Ballini, J., Konan- Mater. Lett. 63, 1683–1685.
Kouakou, Y. N., and van den Bergh, H. 70. Ohulchanskyy, T. Y., Roy, I., Goswami, L.
(2006) Effect of nanoparticle size on the N., Chen, Y., Bergey, E. J., Pandey, R. K.,
extravasation and the photothrombic activity et al. (2007) Organically modified silica
of meso(p-tetracarboxyphenyl)porphyrin. nanoparticles with covalently incorporated
J. Photochem. Photobiol. B 85, 216–222. photosensitizer for photodynamic therapy of
60. McCarthy, J. R., Perez, J. M., Bruckner, C., cancer. Nano Lett. 7, 2835–2842.
and Weissleder, R. (2005) Polymeric nano- 71. Rossi, L. M., Silva, P. R., Vono, L. L. R.,
particle preparation that eradicates tumors. Fernandes, A. U., Tada, D. B., and Baptista,
Nano Lett. 5, 2552–2556. M. S. (2008) Protoporphyrin IX nanoparticle
61. Tang, W., Xu, H., Kopelman, R., and carrier: preparation, optical properties, and
Philbert, M. A. (2005) Photodynamic char- singlet oxygen generation. Langmuir 24,
acterization and in vitro application of meth- 12534–12538.
ylene blue-containing nanoparticle platforms. 72. Brevet, D., Gary-Bobo, M., Raehm, L.,
Photochem. Photobiol. 81, 242–249. Richeter, S., Hocine, O., Amro, K., et al.
62. Kopelman, R., Koo, Y. L., Philbert, M., (2009) Mannose-targeted mesoporous silica
Moffat, B. A., Reddy, G. R., McConville, P., nanoparticles for photodynamic therapy.
et al. (2005) Multifunctional nanoparticle Chem. Commun. 12, 1475–1477.
73. Kumar, R., Roy, I., Ohulchanskyy, T. Y., system in magnetic resonance imaging,
Goswami, L. N., Bonoiu, A. C., Bergey, E. J., luminescence imaging, and photodynamic
et al. (2008) Covalently dye-linked, surface- therapy. Small 4, 218–224.
controlled, and bioconjugated organically 83. Chen, Z., Sun, Y., Huang, P., Yang, X., and
modified silica nanoparticles as targeted Zhou, X. (2009) Studies on preparation of
probes for optical imaging. ACS Nano 2, photosensitizer loaded magnetic silica nano-
449–456. particles and their anti-tumor effects for
74. Kim, S., Ohulchanskyy, T. Y., Pudavar, H. E., targeting photodynamic therapy. Nanoscale
Pandey R. K., and Prasad, P. N. (2007) Res. Lett. 4, 400–408.
Organically modified silica nanoparticles co- 84. Tada, D. B., Vono, L. L. R., Duarte, E. L.,
encapsulating photosensitizing drug and Itri, R., Kiyohara, P. K., Baptista, M. S., et al.
aggregation-enhanced two-photon absorb- (2007) Methylene blue-containing silica-
ing fluorescent dye aggregates for two-pho- coated magnetic particles: a potential mag-
ton photodynamic therapy. J. Am. Chem. Soc. netic carrier for photodynamic therapy.
129, 2669–2675. Langmuir 23, 8194–8199.
75. Kim, S., Ohulchanskyy, T. Y., Bharali, D., 85. McCarthy, J. R., Jaffer, F. A., and Weissleder,
Chen, Y., Pandey, R. K., and Prasad, P. N. R. (2006) A macrophage-targeted theranos-
(2009) Organically modified silica nanopar- tic nanoparticle for biomedical applications.
ticles with intraparticle heavy-atom effect on Small 2, 983–987.
the encapsulated photosensitizer for 86. Sun, Y., Chen, Z., Yang, X., Huang, P.,
enhanced efficacy of photodynamic therapy. Zhou, X., and Du, X. (2009) Magnetic
J. Phys. Chem. C 113, 12641–12644. chitosan nanoparticles as a drug delivery
76. Tu, B. H., Lin, Y., Hung, Y., Lo, L., Chen, system for targeting photodynamic therapy.
Y., and Mou, C. (2009) In vitro studies of Nanotechnology 20, 135102.
functionalized mesoporous silica nanoparti- 87. Zhang, P., Steelant, W., Kumar, M., and
cles for photodynamic therapy. Adv. Mater. Scholfield, M. (2007) Versatile photosensitiz-
21, 172–174. ers for photodynamic therapy at infrared
77. Chen, K., Preuß, A., Hackbarth, S., Wacker, excitation. J. Am. Chem. Soc. 129,
M., Langer, K., and Röder, B. (2009) Novel 4526–4527.
photosensitizer-protein nanoparticles for 88. Chatterjee, D. K. and Yong, Z. (2008)
photodynamic therapy: photophysical char- Upconverting nanoparticles as nanotrans-
acterization and in vitro investigations. ducers for photodynamic therapy in cancer
J. Photochem. Photobiol. B 96, 66–74. cells. Nanomedicine 3, 73–82.
78. Deda, D. K., Uchoa, A. F., Carita, E., 89. Yaghini, E., Seifalian, A. M., and MacRobert,
Baptista, M. S., Toma, H. E., and Araki, K. A. J. (2009) Quantum dots and their poten-
(2009) A new micro/nanoencapsulated por- tial biomedical applications in photosensiti-
phyrin formulation for PDT treatment. Int. zation for photodynamic therapy.
J. Pharm. 376, 76–83. Nanomedicine 4, 353–363.
79. Khdair, A., Gerard, B., Handa, H., Mao, G., 90. dos Santos, L. J., Alves, R. B., de Freitas, R.
Shekhar, M. P. V., and Panyam, J. (2008) P., Nierengarten, J.-F., Magalhaes, L. E. F.,
Surfactant-polymer nanoparticles enhance Krambrock, K., et al. (2008) Production of
the effectiveness of anticancer photodynamic reactive oxygen species induced by a new
therapy. Mol. Pharm. 5, 795–807. [60]fullerene derivative bearing a tetrazole
80. Khdair, A., Handa, H., Mao, G., and Panyam, unit and its possible biological applications.
J. (2009) Nanoparticle-mediated combina- J. Photochem. Photobiol. A 200, 277–281.
tion chemotherapy and photodynamic ther- 91. Song, L. P., Li, H., Sunar, U., Chen, J.,
apy overcomes tumor drug resistance in vitro. Corbin, I., Yodh, A. G., et al. (2007)
Eur. J. Pharm. Biopharm. 71, 214–222. Naphthalocyanine-reconstituted LDL nano-
81. Wieder, M. E., Hone, D. C., Cook, M. J., particles for in vivo cancer imaging and treat-
Handsley, M. M., Gavrilovic, J., and Russell, ment. Int. J. Nanomedicine 2, 767–774.
D. A. (2006) Intracellular photodynamic ther- 92. Suci, P. A., Varpness, Z., Gillitzer, E.,
apy with photosensitizer-nanoparticle conju- Douglas, T., and Young, M. (2007) Targeting
gates: cancer therapy using a “Trojan horse.” and photodynamic killing of a microbial
Photochem. Photobiol. Sci. 5, 727–734. pathogen using protein cage architectures
82. Lai, C., Wang, Y., Lai, C., Yang, M.-J., Chen, functionalized with a photosensitizer.
C.-Y., Chou, P.-T., et al. (2008) Iridium- Langmuir 23, 12280–12286.
complex-functionalized Fe3O4/SiO2 core/ 93. Xu, H., Buck, S. M., Kopelman, R., Philbert,
shell nanoparticles: a facile three-in-one M. A., Brasuel, M., Ross, B. D., et al. (2004)
Photo-excitation based nano-explorers: 104. Jain, R. A. (2000) The manufacturing

chemical analysis inside live cells and photo- techniques of various drug loaded biode-
dynamic therapy. Isr. J. Chem. 44, 317–337. gradable poly(lactide-co-glycolide) devices.
94. Gu, H., Xu, K., Yang, Z., Changa, C. K., and Biomaterials 21, 2475–2490.
Xu, B. (2005) Synthesis and cellular uptake of 105. Panyama, J. and Labhasetwar, V. (2003)
porphyrin decorated iron oxide nanoparticles- Biodegradable nanoparticles for drug and
a potential candidate for bimodal anticancer gene delivery to cells and tissue. Adv. Drug
therapy. Chem. Commun. 34, 4270–4272. Deliv. Rev. 55, 329–347.
95. Chen, W. and Zhang, J. (2006) Using nano- 106. Shive, M. S. and Anderson, J. M. (1997)
particles to enable simultaneous radiation Biodegradation and biocompatibility of PLA
and photodynamic therapies for cancer treat- and PLGA microspheres. Adv. Drug Deliv.
ment. J. Nanosci. Nanotech. 6, 1159–1166. Rev. 28, 5–24.
96. Liu, Y. F., Chen, W., Wang, S. P., and Joly, A. 107. Vargas, A., Eid, M., Fanchaouy, M., Gurny,
G. (2008) Investigation of water-soluble R., and Delie, F. (2008) In vivo photody-
X-ray luminescence nanoparticles for photo- namic activity of photosensitizer-loaded
dynamic activation. Appl. Phys. Lett. 92, nanoparticles: formulation properties, admin-
043901. istration parameters and biological issues
97. Moreno, M. J., Monson, E., Reddy, G. R., involved in PDT outcome. Eur. J. Pharma.
Rehemtulla, A., Ross, B. D., Philbert, M., Biopharma. 69, 43–53.
et al. (2003) Production of singlet oxygen by 108. Hah, H. J., Kim, J. S., Jeon, B. J., Koo, S. M.,
Ru(dpp(SO3)2)3 incorporated in polyacryl- and Lee, Y. E. (2003) Simple preparation
amide PEBBLEs. Sens. Act. B: Chem 90, of mono-disperse hollow silica particles with-
82–89. out using templates. Chem. Commun. 14,
98. Cao, Y., Koo, Y.-E. L., Koo, S. M., and 1712–1713.
Kopelman, R. (2005) Ratiometric singlet 109. Qian, J., Gharibi, A., and He, S. L. (2009)
oxygen nano-optodes and their use for moni- Colloidal mesoporous silica nanoparticles with
toring photodynamic therapy nanoplatforms. protoporphyrin IX encapsulated for photody-
Photochem. Photobiol. 81, 1489–1498. namic therapy. J. Biomed. Opt. 14, 014012.
99. Shinohara, E. T., Cao, C., Niermann, K., 110. Peng, Q., Warloe, T., Berg, K., Moan, J.,
Mu, Y., Zeng, F., Hallahan, D. E., et al. Kongshaug, M., Giercksky, K.-E., et al.
(2005) Enhanced radiation damage of tumor (1997) 5-Aminolevulinic acid-based photo-
vasculature by mTOR inhibitors. Oncogene dynamic therapy, clinical research, and future
24, 5414–5422. challenges. Cancer 79, 2282–2308.
100. Gill, Z. P., Perks, C. M., Newcomb, P. V., 111. Battah, S., Balaratnam, S., Casas, A., O’Neill,
and Holly, J. M. P. (1997) Insulin-like S., Edwards, C., Batlle, A., et al. (2007)
growth factor-binding protein (IGFBP-3) Macromolecular delivery of 5-aminolaevu-
predisposes breast cancer cells to pro- linic acid for photodynamic therapy using
grammed cell death in a non-IGF-dependent dendrimer conjugates. Mol. Cancer Ther. 6,
manner. J. Biol. Chem. 272, 25602–25607. 876–885.
101. Clark, H. A., Barker, S. L. R., Brasuel, M., 112. Casas, A., Battah, S., Di Venosa, G., Dobbin,
Miller, M. T., Monson, E., Parus, S., et al. P., Rodriguez, L., Fukuda, A., et al. (2009)
(1998) Subcellular optochemical nanobio- Sustained and efficient porphyrin generation
sensors: probes encapsulated by biologically in vivo using dendrimer conjugates of 5-ALA
localised embedding (PEBBLEs). Sens. Act. for photodynamic therapy. J. Controlled
B 51, 12–16. Release 135, 136–143.
102. Harrell, J. A. and Kopelman, R. (2000) 113. Rodrigues, M. M. A., Simioni, A. R., Primo,
Biocompatible probes measure intracellular F. L., Siqueira-Moura, M. P., Morais, P. C.,
activity. Biophotonics Int. 7, 22–24. and Tedesco, A. C. (2009) Preparation, char-
103. Anderson, F. A. (2005) Amended final report acterization and in vitro cytotoxicity of BSA-
on the safety assessment of polyacrylamide based nanospheres containing nanosized
and acrylamide residues in cosmetics. Int. J. magnetic particles and/or photosensitizer.
Tox. 24(Suppl. 2), 21–50. J. Magn. Magn. Mater. 321, 1600–1603.
Chapter 12
Hydrogel Templates for the Fabrication of Homogeneous

Polymer Microparticles
Ghanashyam Acharya, Matthew McDermott, Soo Jung Shin,
Haesun Park, and Kinam Park
Abstract
Nano/microparticulate drug delivery systems with homogeneous size distribution and predefined shape
are important in understanding the influence of the geometry and dimensions of these systems on blood
circulation times and cellular uptake. We present a general method using water dissolvable hydrogel
templates for the fabrication of homogeneous, shape-specific polymer/drug constructs in the size range
of 200 nm to 50 mm. This hydrogel template strategy is mild, inexpensive, and readily scalable for the
fabrication of multifunctional drug delivery vehicles.
Key words: Hydrogel, Gelatin, Template, PLGA, Microparticle
1. Introduction
Polymeric nano/microparticulate drug delivery systems have

been developed for the delivery of various pharmaceutical agents
to targeted areas of the human body (1). The currently available
particulate drug delivery systems include liposomes, polymer
micelles, and nano/microspheres (2–4). Most nano/microparti-
cles have been prepared by emulsion methods (5, 6). These deliv-
ery systems, however, can only be used to prepare particles which
carry small amounts of drug compared with the total polymer
content. Furthermore, these systems produce particles which are
polydisperse in size and so it is difficult to correlate the effect of
the size and shape of the particles on biological responses.
Recently, various nanofabrication methods have been developed
for making nano/microparticles with homogeneous size distribu-
tion (7–9). However, nanofabrication of particulate drug delivery
systems for drug delivery applications has been a challenging task
179
180 Acharya et al.
mainly due to the lack of general fabrication methods that can

produce purified particles in large scale.
Hydrogels are being developed as coating materials on bio-
implants to function as interfaces between implants and biological
tissues (10, 11) and as reversible thermosensitive templates for
biomimetic mineralization (12, 13) and for nanoparticle fabrica-
tion. In terms of nanofabrication, the hydrogel template strategy
addresses various issues associated with the nanofabrication of
particles for use in drug delivery systems. This strategy is broadly
applicable and mild and can be used to synthesize nano/micropar-
ticles composed of many different polymer/drug combinations
ranging in size from 200 nm to greater than 50 mm. Since hydro-
gels are water soluble, the formed polymer particles (which are
not soluble in water) can be easily collected by simply dissolving
the template. Further, certain biopolymer hydrogels, such as gel-
atin, form phase-reversible elastic and mechanically strong hydro-
gels that can withstand physical manipulation during template
preparation and filling and that can maintain the polymer or poly-
mer–drug mixture inside the cavities of the template, preventing
its diffusion. Templates for nanoparticle fabrication also can be
made from other polymers, such as poly(dimethyl siloxane)
(PDMS), however, removing the formed particles from PDMS
templates is very difficult since PDMS is not soluble in water.
We herein describe a strategy based on dissolvable hydrogel
templates that can be utilized to produce homogeneous polymeric
drug delivery systems with a predefined size and shape. Briefly,
solutions of hydrogel-forming biopolymer (i.e., gelatin) are used
for imprinting nano/microsized features from a PDMS template
master. Upon gelation, the hydrogel template retains the exact
negative imprint of the features present on the PDMS template
(i.e., wells). These wells then are filled with polymer solution and
the hydrogel template is dissolved with warm water to release free
polymer/drug delivery particles of homogeneous size (14).
2. Materials
2.1. Fabrication 1. PDMS templates with predefined patterns (Akina, Inc., West
of Hydrogel Templates Lafayette, IN).
2. Gelatin from porcine skin, Type A, 300 bloom (Sigma, St.
Louis, MO).
3. Nanopure water.
2.2. Fabrication 1. Poly(lactic-co-glycolic acid) (PLGA) (MW 65,000, IV

of PLGA Microparticles 0.82 dL/g).
2. Fluorescent dye: Nile Red.
3. Dichloromethane.
Hydrogel Templates for the Fabrication of Homogeneous Polymer Microparticles 181
2.3. Collection of PLGA 1. Microfilters (Sterlitech Corporation, Kent, WA).

Microparticles 2. Eppendorf Centrifuge 5804, Rotor A-4-44, used at 4,500 × g
(Eppendorf, Hauppauge, NY).
3. Olympus Spinning Disk Confocal Imaging Microscope
BX61-DSU (Center Valley, PA) equipped with Intelligent
Imaging Innovations Slide Book 4.0 software for automated
Z-stack and 3D image analysis.
3. Methods
The overall process of the hydrogel template method of making

nano/microparticles is described in Fig. 1. Hydrogel templates
are prepared by pouring a warm gelatin solution onto a PDMS
template containing circular posts (i.e., 20 mm in diameter and
20 mm tall). The PDMS template is then cooled for 10 min in a
refrigerator and the formed hydrogel template is carefully peeled
off (gelatin undergoes a sol–gel phase transition at 40–45°C).
The hydrogel template contains the exact negative imprint of the
features present on the PDMS master template (i.e., wells 20 mm
in diameter and 20 mm deep). The as-synthesized hydrogel tem-
plates then are filled with PLGA polymer solution. Finally, the
PLGA-filled, hydrogel template is dissolved in warm water, releas-
ing free homogeneous PLGA microstructures. Using hydrogel
templates, homogeneous polymer particles of 200 nm to 50 mm
and larger have been prepared (see Fig. 2). Drug loaded PLGA
microparticles can be prepared by mixing the PLGA polymer
solution with a drug. The drug loaded particles of nanoscale
Fig. 1. Schematic diagram for fabrication of particles by the hydrogel template method. (a) A PDMS template having vertical
posts is prepared from a silicon wafer master template. (b) A warm gelatin solution is poured onto the PDMS template.
(c) The formed hydrogel template is peeled off from the PDMS template (wells are denoted by white spots). (d) The
microwells in the hydrogel template are filled with a PLGA polymer solution (containing a drug) by swiping with a blade.
(e) Homogeneous free PLGA microstructures are obtained by simply dissolving the hydrogel template in water.
182 Acharya et al.
Fig. 2. Fabrication of homogeneous 20 mm PLGA microstructures with a hydrogel template. (a) Bright field image of a
gelatin hydrogel template. (b) A fluorescence image of a hydrogel template filled with PLGA solution containing Nile Red.
(c, d) Bright field and fluorescence images, respectively, of homogeneous free PLGA microstructures obtained by dissolving
the hydrogel templates. Scale bars correspond to 40 mm.
dimension (e.g., 200 nm) are useful for targeting tumor sites by

intravenous administration, while those of microscale dimension
(e.g., 20 mm) are ideal for long-term delivery ranging from weeks
to months after implantation.
3.1. Preparation 1. Weigh gelatin powder (30 g) and transfer it into a 250-mL
of Gelatin Solution Pyrex bottle.
2. Add 100 mL Nanopure water to the Pyrex bottle and mix
thoroughly.
3. Cap the bottle to prevent evaporation and place in an oven at 65°C
for 2 h or until the formation of a clear solution (see Note 1).
4. Use this as-synthesized clear gelatin solution (30% w/v in
water) in the fabrication of hydrogel templates (see Note 2).
3.2. Fabrication 1. Transfer the warm gelatin solution (5 mL) with a pipette onto
of Hydrogel Templates a PDMS template (3² diameter) containing circular pillars
(i.e., 20 mm diameter and 20 mm height).
2. Spread the gelatin solution evenly to form a thin film com-
pletely covering the PDMS template and cool it to 4°C for
10 min by keeping it in a refrigerator (see Note 3).
3. Peel the hydrogel template from the PDMS master template
(see Note 4).
4. The hydrogel template thus prepared will be ~3 in. in diam-
eter and contain circular wells (i.e., 20 mm diameter and
20 mm depth).
5. Examine the hydrogel template under a bright field micro-
scope (see Note 5).
3.3. Preparation 1. Transfer 200 mg of PLGA polymer into a 5-mL glass vial.
of PLGA Solution 2. Add 1 mL of dichloromethane (CH2Cl2) to the vial with a
micropipette (see Note 6).
3. Seal the glass vial with Parafilm and place on a flask rocker
until all the PLGA is completely dissolved.
4. A clear, thick PLGA solution of 20% w/v concentration will
be formed.
5. For fluorescence imaging purposes, the PLGA solution thus

prepared may be doped with a fluorescent dye (e.g., 0.001%
Nile Red).
3.4. Fabrication 1. Transfer 200 mL 20% PLGA solution w/v in dichloromethane

of PLGA Microparticles with a pipette onto a 3 in. diameter hydrogel template contain-
ing circular wells 20 mm in diameter and depth (see Note 7).
2. Evenly spread the PLGA solution on the hydrogel template
by swiping with a razor blade at a 45° angle (see Note 8).
3. Keep the PLGA-filled gelatin template on the table at room
temperature (~25°C) for 5–10 min to evaporate CH2Cl2. For
the complete removal of dichloromethane, longer drying
time may be required.
4. Examine the hydrogel template under a bright field micro-
scope (see Note 9).
3.5. Collection 1. Dissolve a batch of ten hydrogel templates in a 100-mL bea-

of the PLGA ker containing 50 mL of water at 40°C.
Microparticles 2. Gently swirl the beaker for 2 min to completely dissolve the
hydrogel templates.
3. Upon complete dissolution of the hydrogel templates, free
PLGA microparticles will be released into the water.
4. Transfer the solution into 15-mL conical centrifuge tubes
and centrifuge at 4,500 × g for 5 min.
5. Discard the supernatant liquid and collect the pellets.
6. Combine all the pellets, redisperse them in water and filter
through a 25-mm filter holder equipped with a 25-mm filter
(see Note 10).
7. Spot a drop of this dispersion on a glass slide and examine
under a bright field microscope (see Note 11).
8. Transfer the solution into a 15-mL conical centrifuge tube
and centrifuge at 4,500 × g for 5 min.
9. Freeze dry the pellet obtained upon centrifugation and store
it in a refrigerator (see Note 12).
10. This material will be stable for up to 1 year when stored in a
freezer.
4. Notes
1. Gelatin solution should not be kept hot (>65°C) for long

periods of time (>10 h). High temperatures may denature the
protein and cause the solution to lose its ability to form
hydrogel templates.
184 Acharya et al.
2. Elastic and mechanically strong hydrogel templates will be

formed with a 30% gelatin solution.
3. Cooling results in the formation of a mechanically stable
hydrogel template that can be easily peeled away from the
PDMS template.
4. Gelatin templates soften and melt at room temperatures.
They need to be stored in the refrigerator.
5. The presence of homogeneous circular wells in the hydrogel
template will be clearly visible (see Fig. 2a).
6. PLGA solutions prepared from ethyl acetate and tetrahydro-
furan solvents can also be used for hydrogel template
filling.
7. Solutions of other polymers, such as polycaprolactone, poly-
styrene, and poly(vinyl chloride), can also be used for filling
the hydrogel templates.
8. Swiping with a razor blade minimizes formation of the PLGA
film (i.e., scum layer) on the hydrogel template surface.
Gentle pressure has to be applied to force the PLGA solution
to completely fill the wells without deforming the hydrogel
template. However, avoid pressing the razor blade too hard
as it might slough off the template.
9. The circular wells filled with polymer solution in the hydrogel
template appear slightly darker compared to an unfilled
hydrogel template under a bright field microscope.
Alternatively, the filled templates can be observed with a fluo-
rescence microscope if a polymer/dye mixture, such as
PLGA/Nile Red, is used (see Fig. 2b).
10. Filtration removes larger aggregates or pieces of polymer film
formed during filling of the hydrogel template with polymer
solution.
11. Look for the presence of homogeneous circular microstruc-
tures dispersed in water (see Fig. 2c). Alternatively, the nano-
particles can be viewed with a fluorescence microscope if a
polymer/dye mixture, such as PLGA/Nile Red, is used (see
Fig. 2d).
12. This process yields approximately 1 mg of PLGA microstruc-
tures. Homogeneous PLGA microparticles containing drugs
(e.g., felodipine, progesterone, and paclitaxel) can be pre-
pared by using a solution of PLGA polymer and drug to fill
the templates. The stored pellets can be redispersed in water
for further use by vortexing.
Acknowledgment
This project was supported in part by the Showalter Trust Fund

from Purdue Research Foundation.
References
1. Peer, D., Karp, J. M., Hong, S., Farokhzad, based fabrication of shape-specific, enzymati-
O. C., Margalit, R., and Langer, R. (2007) cally-triggered smart nanoparticles. J. Control.
Nanocarriers as an emerging platform for can- Release 125, 263–272.
cer therapy. Nat. Nanotechnol. 2, 751–760. 9. Huang, K.-S., Lu, K., Yeh, C.-S., Chung,
2. Torchilin, V. P. (2005) Recent advances with S.-R., Lin, C.-H., Yang, C.-H., et al. (2009)
liposomes as pharmaceutical carriers. Nat. Microfluidic controlling monodisperse
Rev. Drug Discov. 5, 145–160. microdroplet for 5-fluorouracil loaded geni-
3. Mundargi, R. C., Babu, V. R., Rangaswamy, pin-gelatin microcapsules. J. Control. Release
V., Patel, P., and Aminabhavi, T. M. (2008) 137, 15–19.
Nano/micro technologies for delivering 10. Peppas, N. A., Hilt, J. Z., Khademhosseini, A.,
macromolecular therapeutics using poly and Langer, R. (2006) Hydrogels in biology
(d,l-lactide-co-glycolide) and its derivatives. and medicine: From molecular principles
J. Control. Release 125, 193–209. to bionanotechnology. Adv. Mater. 18,
4. Blanco, E., Kesssinger, C. W., Sumer, B. D., 1345–1360.
and Gao, J. (2009) Multifunctional micellar 11. Park, K., Shalaby, W. S. W., and Park, H.
nanomedicine for cancer therapy. Exp. Biol. (1993) Biodegradable Hydrogels for Drug
Med. 234, 123–131. Delivery. Technomic Publishing, Lancaster,
5. Danhier, F., Lecouturier, N., Vroman, B., PA.
Jérôme, C., Marchand-Brynaert, J., Feron, 12. Liu, G., Zhao, D., Tomsia, A. P., Minor, A.
O., et al. (2009) Paclitaxel-loaded PEGylated M., Song, X., and Saiz, E. (2009) Three-
PLGA-based nanoparticles: In vitro and in vivo dimensional biomimetic mineralization of
evaluation. J. Control. Release 133, 11–17. dense hydrogel templates. J. Am. Chem. Soc.
6. Heslinga, M. J., Mastria, E. M., and Eniola- 131, 9937–9939.
Adefeso, O. (2009) Fabrication of biodegrad- 13. Du, B., Cao, Z., Li, Z., Mei, A., Zhang, X.,
able spheroidal microparticles for drug delivery Nie, J., et al. (2009) One-Pot preparation of
applications. J. Control. Release 138, 235–242. hollow silica spheres by using thermosensitive
7. Euliss, L. E., DuPont, J. A., Gratton, S., and poly(N-isopropylacrylamide) as a reversible
DeSimone, J. (2006) Imparting size, shape, template. Langmuir 25, 12367–12373.
and composition control of materials for 14. Acharya, G., Shin, C. S., McDermott, M.,
nanomedicine. Chem. Soc. Rev. 35, Mishra, H., Park, H., Kwon, I. C., et al.
1095–1104. (2010) The hydrogel template method for
8. Glangchai, L. C., Caldorera-Moore, M., Shi, fabrication of homogeneous nano/micropar-
L., and Roy, K. (2008) Nanoimprint lithography ticles. J. Control. Release 141, 314–319.
Chapter 13
Antibacterial Application of Engineered Bacteriophage

Nanomedicines: Antibody-Targeted, Chloramphenicol
Prodrug Loaded Bacteriophages for Inhibiting
the Growth of Staphylococcus aureus Bacteria
Lilach Vaks and Itai Benhar
Abstract
The increasing development of bacterial resistance to traditional antibiotics has reached alarming levels,
thus there is an urgent need to develop new antimicrobial agents. To be effective, these new antimicrobials
should possess novel modes of action and/or different cellular targets compared with existing antibiotics.
Bacteriophages (phages) have been used for over a century as tools for the treatment of bacterial infec-
tions, for nearly half a century as tools in genetic research, for about two decades as tools for the discovery
of specific target-binding proteins and peptides, and for almost a decade as tools for vaccine development.
We describe a new application in the area of antibacterial nanomedicines where filamentous phages can
be formulated as targeted drug-delivery vehicles of nanometric dimensions (phage nanomedicines) and
used for therapeutic purposes. This protocol involves both genetic and chemical engineering of these
phages. The genetic engineering of the phage coat, which results in the display of a target-specificity-
conferring peptide or protein on the phage coat, can be used to design the drug-release mechanism and
is not described herein. However, the methods used to chemically conjugate cytotoxic drugs at high
density on the phage coat are described. Further, assays to measure the drug load on the surface of the
phage and the potency of the system in the inhibition of growth of target cells as well as assessment of
the therapeutic potential of the phages in a mouse disease model are discussed.
Key words: Peptide phage display library, Phage display, Single-chain antibodies, BirA biotin ligase,
ZZ domain, IgG, Fc antibody fragment
1. Introduction
The increasing development of bacterial resistance to traditional

antibiotics has reached alarming levels (1), forcing scientists to
develop new antimicrobial approaches. In both traditional and
newly developed antibiotics, the target selectivity lies in the potency
187
188 Vaks and Benhar
of the drug itself as well as in its ability to affect a mechanism

which destroys or hinders the target microorganism and not its
host. In fact, a vast number of potentially potent drugs have been
excluded from use as therapeutics due to low selectivity (i.e., toxicity
to the host as well as to the pathogen) (2). This brings to mind
the limited selectivity of anticancer drugs and recent efforts to
overcome this low selectivity through the development of novel-
targeted, drug-delivery strategies.
In this chapter, we introduce a novel application of filamentous
bacteriophage (phage) as targeted drug carriers for the eradica-
tion of pathogenic bacteria. In nature, bacteriophages are viruses
that selectively invade the specific bacteria cells causing growth
inhibition and death. As an efficient immunotherapeutic, it con-
tains the following three components: a drug carrier, a targeting
moiety, and a cytotoxic drug, which are responsible for payload,
specificity, and efficacy, respectively. In this protocol, the geneti-
cally and chemically modified filamentous M13 filamentous bac-
teriophage takes the role of a nanometric, modular, high-capacity
drug-carrying platform, while an antibody provides targeting drug
and chloramphenicol serves as a cytotoxic drug. These bacterio-
phages have dimensions on the nanometer scale (~1 mm in length
but less than 10 nm in diameter); hence, we refer to them as
phage nanoparticles. The M13 filamentous bacteriophage (see
Fig. 1) refers to Escherichia coli, male-specific bacteriophage Ff
class that is able to infect and replicate in E. coli bacteria. However,
in this study, the phages are targeted against bacterial pathogens
by specific antibodies displayed on its coat; therefore, their natural
host specificity is not relevant for the target specificity. We use
these phages to deliver a large chloramphenicol payload to patho-
genic bacteria such as Staphylococcus aureus (3, 4) (see Note 1). In
this protocol, chloramphenicol, a hydrophobic drug, is chemically
Fig. 1. Structure of the filamentous bacteriophage. The display of proteins and peptides
is achieved by in frame fusion of its coding sequence with the sequence of the chosen
phage coat protein.
Antibacterial Application of Engineered Bacteriophage Nanomedicines 189
modified to contain an esterase-cleavage linker that enables its

slow release from the surface of the phage by serum esterases. The
high density of chloramphenicol on the surface of the phage (104
drugs/phage) is made possible by linking it to the phage coat
through aminoglycoside antibiotics that serve as solubility-
enhancing, branched linkers.
This protocol describes genetically engineered fUSE5 phages
that display a 15-mer peptide AVITAG (GLNGLNDIFEAQK
IEWHE) (5, 6) on the N-terminus of the p3 minor coat protein
(g3p-AVITAG-fUSE5). The AVITAG peptide undergoes efficient
biotinylation in vivo by the BirA biotin ligase enzyme (7) which
enables coupling with a specific biotinylated antibody via the
biotin–streptavidin bridge (see Note 2).
The preparation of drug-carrying phages then can be divided
into three major steps:
1. Phage propagation in bacteria and purification by PEG/NaCl
precipitation.
2. Prodrug preparation by sequential chemical reactions.
3. Conjugation of the prodrug to the bacteriophage through a
hydrophilic aminoglycoside linker via an 1-ethyl-3-(3-dimethy
laminopropyl)-carbodiimide hydrochloride (EDC) coupling
reaction.
Further, we describe how to quantify the number of drug
molecules on the surface of the phage and how to carry out a
bacterial growth inhibition assay to determine the effect that the
drug-conjugated phages have on retarding the growth of the target
bacteria. Finally, we establish the in vivo therapeutic potential of
this system using a disease model in BALB/c mice with a lethal
systemic infection of S. aureus bacteria. The colony-forming units
(CFU) quantification method (8) or radio-labeled phage particles
(9) can be used to learn about the targeting ability of drug-carrying
phage pharmacokinetics and biodistribution.
2. Materials
(See Note 3)
2.1. Phage Preparation 1. Phage vector g3p-AVITAG-fUSE5 (see Note 4).
2. Bacteria strains: E. coli DH5a and DH5aF¢ (GibcoBRL, Life
Technologies, MD, USA) are used for phage preparation and
for phage titration.
3. Growth medium: Yeast extract-tryptone x2 (2YT) and Tryptic
Soy Broth (TSB) (see bacteria growth media).
4. Antibiotics: tetracyclin and chloramphenicol (see Sub
heading 2.4).
190 Vaks and Benhar
5. Vectors: pBirAcm (Avidity, LLC, http://www.avidity.com/)

is a biotin ligase birA expression vector, in which the expres-
sion of the birA gene is controlled by an isopropyl b-d-1-
thiogalactopyranoside (IPTG) inducible promoter.
6. d-biotin: 50 mM d-biotin in water, add 10 N NaOH drop-
wise to dissolve biotin completely. Filter 0.22 mm to sterilize.
Store at 4°C for up to 3 months.
7. IPTG: 1 M of IPTG in sterile double-distilled (MilliQ) water
(SDDW) is stored in 1 mL aliquots at −20°C. Use 0.1 mM
for birA biotin ligase overexpression.
8. PEG/NaCl: 20% PEG6000 and 2.5 M NaCl in MilliQ water.
Sterilize by autoclaving.
9. Vacuum filtration device (0.45 mm) (Amicon, USA).
10. Magnetic beads: Dynabeads M-280 streptavidin (Invitrogen
Dynal, http://www.invitrogen.com/site/us/en/home/
brands/Dynal.html).
11. Antibodies: biotinylated mouse-anti-S. aureus IgG (Abcam,
USA, http://www.abcam.com). Any biotinylated IgG that
does not bind S. aureus as a negative control.
12. Avidin: 10 mg/mL stock solution in DDW. Store at 4°C.
2.2. Prodrug Chemicals and solvents were either A.R. grade or purified by
Preparation standard techniques.
1. Tetrahydrofurane (THF).
2. Glutaric anhydride.
3. Triethylamine (Et3N).
4. Dimethylaminopyridine.
5. Ethylacetate (EtOAc).
6. Hexane.
7. Hydrochloric acid (HCl).
8. Magnesium sulfate (MgSO4).
9. Silica gel Merck 60 (particle size 0.040–0.063 mm).
10. Dichloromethane (DCM).
11. N, N ¢-Dicyclohexylcarbodiimide (DCC).
12. N-Hydroxysuccinimide (NHS).
13. Argon gas.
14. Thin layer chromatography (TLC): silica gel plates Merck
60F254. Compounds were visualized by irradiation with
UV light.
2.3. Drug Conjugation 1. NaHCO3: pH = 8.5 buffer that provides basic conditions for
drug conjugation.
2. Dimethyl sulfoxide (DMSO): used for dissolving chloram-
phenicol–N-Hydroxysuccinimide (CAM–NHS) prodrug to a
final stock concentration of 44 mg/mL.
3. For high-performance liquid chromatography (HPLC): a reverse
phase C-18 column and 80% acetonitrile solution (in water w/w).
4. Citrate buffer: pH = 5.0 solution that is composed of citric
acid 1 M and sodium citrate 1 M at appropriate dilution. For
100 mL citrate buffer, pH = 5.0, use 41 mL citric acid and
59 mL sodium citrate.
5. Neomycin: an aminoglycoside antibiotic, which serves as a
branched, hydrophilic linker that conjugates the chloramphen-
icol prodrug to the phage coat proteins via EDC chemistry.
6. EDC, a zero-length crosslinking agent, was used to couple car-
boxyl groups to primary amines. Store the powder at −20°C.
Make a fresh solution in DMSO immediately prior to use.
7. For dialysis: SnakeSkin-Pleated Dialysis tubing (10 kDa cutoff)
supplied by PIERCE (Rockford, Illinois, USA).
2.4. General Buffers 1. Phosphate-buffered saline (PBS): 8 g NaCl, 0.2 g KCl, 1.44 g
and Reagents Na2HPO4, and 0.24 g KH2PO4 per 1 L, pH = 7.4.
2. Chloramphenicol: 34 mg/mL in 100% ethanol. Store at −20°C.
3. Tetracyclin: 12.5 mg/mL in 50% ethanol. Store at −20°C.
4. Normal rabbit serum. Store at 4°C.
2.5. Bacteria Growth Any supplier of bacterial growth medium components or pre-
Media prepared media. We use products of Becton-Dickinson (http://
www.bd.com/).
1. 2YT: 16 g Bacto-Tryptone, 10 g Yeast extract, and 5 g
NaCl/L water.
2. TSB: 30 g of Bacto-TBS/L water.
3. To prepare solid media, Bacto-agar at the final concentration
of 1.8% was added to the solutions. Following autoclaving,
the media were supplemented with 0.4 or 1% glucose and
antibiotics. The final concentrations of antibiotics used in this
study were as follows: tetracycline (12.5 mg/mL) and
chloramphenicol (34 mg/mL).
2.6. Bacterial Strains In our studies, we used domestic isolates of target bacteria (model
pathogens). Such bacterial strains can be obtained from the
Global Bioresource Center (ATCC) (http://www.atcc.org). The
model bacterial strain used in this protocol is S. aureus COL from
our laboratory collection (see Note 1).
192 Vaks and Benhar
2.7. Animal Studies Female BALB/c mice 8–10 weeks old, ~20 g, at least five mice in
group. During the experiment, monitor mice weight, behavior,
fur condition, and vitality (see Note 5).
3. Methods
This protocol provides a detailed description of antibacterial appli-

cation of engineered bacteriophage nanomedicines as they were
carried out in the laboratory of the authors. Ideally, such a tar-
geted drug-delivery system should home and bind to the target
cells (bacteria) before the drug release is triggered. Basically, fila-
mentous bacteriophages are first equipped with a targeting moiety
that is displayed or linked to a phage coat protein. Variations
to the phage display theme can be found in the phage display
literature (10–13). We provide examples of peptide-displaying
phages that were isolated from a peptide phage display library by
affinity selection on S. aureus (3). A similar approach can be used
to isolate phages that have specificity to any target cell. It is also
possible to use phages that display antibody fragments or other
target-specificity-conferring proteins, each should be carefully
evaluated for how well it tolerates the drug-conjugation chemistry
(see Note 6). The biotinylated phage we designed (g3p-AVITAG-
fUSE5) is optimal in that it can be complexed with targeting
antibodies after completion of the drug-conjugation chemistry.
We provide a description of a chloramphenicol-based prodrug
which has an esterase-cleavable linker with a terminal NHS leaving
group to facilitate its conjugation to amine groups (see Note 7).
Therefore, we began using aminoglycosides (such as neomycin)
as solubility-enhancing, branched linkers that provide larger drug-
loading capacity, better solubility, longer residence in the blood
following i.v. injection into mice, and reduced immunogenicity of
the targeted drug-carrying phage nanomedicines (ref. 4 and
unpublished data). Based on similar concepts, it is possible to
design other means of drug conjugation and release, as we did in
a related study, where a protease-based drug-release mechanism
was engineered into the phage coat (14). Since in a targeted drug-
delivery system, the drug selectivity is replaced with the selectivity
conferred by the targeting moiety, a slew of chemically conjugatable
toxic compounds can be recruited to serve as potent antimicrobial
drugs or as drugs that target other cells that are bearers of disease.
3.1. Phage Preparation Phage g3p-AviTag-fUSE5 displays a 15-amino-acid long peptide

(called AVITAG: with the amino-acid sequence (single letter
3.1.1. Preparation
code) GLNDIFEAQKIEWHE) (Avidity, LLC) at the N-terminus
of Biotinylated g3p-
of the p3 minor coat protein. The peptide undergoes biotinyla-
AVITAG-fUSE5 Phage
tion by the enzyme BirA biotin ligase. The plasmid pBirAcm carries
Vector (See Note 8)
an IPTG inducible (see Note 9) copy of birA.
1. G3p-AviTag-fUSE5 should be co-transformed with the

pBirAcm plasmid into DH5a E. coli cells. As an alternative,
a stock of competent cells that already carry pBirAcm may
be transformed with phage DNA. Plate the transformed
cells on a 2YT-agar plate containing: 12.5 mg/mL tetracy-
cline and 34 mg/mL chloramphenicol. Leave for 16 h at
37°C until colonies of transformed bacteria are clearly
visible.
2. Prepare a starter culture by inoculating 3 mL of 2YT medium
containing: 12.5 mg/mL tetracycline and 34 mg/mL chloram-
phenicol in a 13 mL test tube with a single colony of trans-
formed bacteria. Grow in an incubator-shaker for 16 h
at 37°C.
3. Transfer the 3 mL starter into 200 mL 2YT containing:
12.5 mg/mL tetracycline, 34 mg/mL chloramphenicol, 50 mM
d-biotin, and 0.1 mM IPTG. Under these conditions, and
during the subsequent incubation, the phages are produced
and the AVITAG peptide is biotinylated before the
phages are released from the producing bacteria into the
culture medium.
4. Grow for 24 h shaking (250 rpm) at 30°C (see Note 10).
5. Add 0.1 mM IPTG.
6. Grow for 24 h shaking (250 rpm) at 30°C.
7. Precipitate the phage particles with PEG/NaCl (see
Subheading 3.1.2 and Note 11).
3.1.2. PEG/NaCl Phage 1. To separate the phages from the bacteria, centrifuge the culture
Precipitation (Sorvall centrifuge, GSA rotor, 8,000 × g for 20 min at 4°C)
and filter the supernatant (to eliminate remaining bacteria)
using a 0.45 mm vacuum filtration device (Amicon, USA) (see
Note 12).
2. To precipitate the phages and separate them from the growth
medium (that includes components that unless removed, may
interfere with the subsequent chemical conjugation of drug
to the phages), add one-fifth of the volume of PEG/NaCl to
the supernatant, mix well, and incubate for 2 h or more on ice
(see Note 13).
3. Collect the phage-precipitates by centrifugation at 8,000 × g
for 30 min at 4°C and carefully discard the supernatant.
4. Re-suspend the phages in sterile water, filter at 0.45 mm, and
store at 4°C (see Notes 14 and 15).
5. Determinate the phage concentration (see Subheading 3.1.3).
3.1.3. Phage Quantification 1. Based on optical absorbance: add 50 mL of the phage to
(See Note 16) 450 mL PBS and measure the absorbance at 269 and 320 nm.
194 Vaks and Benhar
Use special UV-transparent or quartz cuvettes. Quantify the

phage concentration according to this formula (15):
(O.D.269nm − O.D.320 nm )× 6 × 1016 × 10 Dilution factor = Phage/ml

( )
Phage genome size (b.p.)

2. Live titration (see Note 17):
(a) Grow DH5aF¢ bacteria in 2YT to A600nm = 0.6–0.8.
(b) Using a multichannel pipette, fill a lane in a sterile 96-well
plate with 90 mL of bacteria, add to the first well 10 mL
of phage, and make serial dilutions by transferring 10 mL
to the next well, making sure to change the tips after
each dilution step. Incubate for 1 h at 37°C.
(c) Plate the 10 mL drops of incubated bacteria onto agar plates
supplemented with tetracycline and grown at 37°C over-
night to develop colonies of resistant (phage infected) cells.
(d) Calculate the phage quantity according to the resistant bac-
teria colonies number multiplied by the dilution factor.
3.1.4. Quantification Biotinylated phages should be readily captured on streptavidin-

of Phage Biotinylation coated magnetic beads. The calculation of biotinylation efficiency
Efficiency is based on capturing the phages on such beads followed by deter-
mination of the uncaptured (presumably unbiotinylated) phages
that are left in the supernatant.
1. Prepare phage stock for reading, by dilution of 33 mL
1–5 × 1013 in 960 mL of PBS.
2. Divide phage solution into two separate tubes.
3. Using a spectrophotometer, read the absorbance of the phage
suspension at 269 nm.
4. Place 100 mL of the magnetic streptavidin beads into a
new tube.
5. Wash the magnetic beads twice with 500 mL PBS by placing
the tube in a magnetic rack for 1 min and then removing the
bead-free liquid with a pipette.
6. Block the beads by adding 500 mL of 1% BSA in PBS solution
for 1 h at 37°C.
7. Discard the supernatant by pipetting it out of the tube while
still on the magnetic rack.
8. Remove the tube from the magnetic rack.
9. Add 500 mL of phages, as prepared earlier.
10. Rotate slowly (10 rpm) on a benchtop tube rotator for 15 min.
11. Place the tube back into the rack for 1 min.
12. Take supernatant into new tube.
13. Read at 269 nm.
14. Compare the concentration of the phages before and after

incubation with beads and determine the exact concentration
of biotinylated phages (see Note 18).
3.2. Drug Conjugation It is highly recommended that the synthesis of such drugs is car-
ried out by an experienced organic chemist.
3.2.1. Preparation
and Conjugation of the 1. Dissolve 1 g chloramphenicol (6.2 mmol) in dry THF.
Chloramphenicol Prodrug 2. Add glutaric anhydride (800 mg, 6.82 mmol), Et3N (1.0 mL,
6.82 mmol), and a catalytic amount of DMAP.
3. Incubate the reaction stirring at room temperature overnight.
4. Check the compound by TLC (EtOAc:Hex = 9:1) to receive
acid-like running.
5. Stop the reaction by adding a large volume (~50 mL) of
EtOAc. The mixture becomes milky during this step.
6. Add the same volume of 1 N HCl. The mixture becomes
partially clear and separates into two layers during this step.
7. Collect the organic layer, dry with magnesium sulfate, and
remove the solvent under reduced pressure.
8. Purify the crude product by column chromatography on silica
gel (EtOAc:Hex = 4:1). The resulting product is a viscous gel
(~2 g weight).
9. Carry out NMR of the product (see Fig. 2): 1H NMR
(200 MHz, CD3OD): d = 8.17 (2H, d, J = 8); 7.65 (2H, d,
J = 8); 6.22 (1H, s); 5.08 (1H, d, J = 2); 4.44–4.41 (2H, m);
4.24 (1H, d, J = 2); 2.40–2.32 (4H, m); 1.92 (2H, t, J = 7).
10. If needed, repeat the wash and purification steps 5–8. If there
are problems with dissolving a product in EtOAc, then add a
small amount of methanol.
11. Dissolve the resulting product (2 g, 4.57 mmol) in DCM.
12. Add DCC (1.4 g, 6.86 mmol) and NHS (790 mg, 6.86 mmol).
13. Incubate the reaction stirring at room temperature overnight.
14. Check the reaction progress by TLC (EtOAc:Hex = 9:1) to
receive a polar compound.
15. Filter the reaction and remove the solvent under reduced
pressure.
16. Purify the crude product by column chromatography on silica
gel (EtOAc:Hex = 4:1) to yield a white solid powder (~1.5 g,
62% yield).
17. Carry out NMR analysis of the CAM–NHS prodrug (see
Fig. 2a): 1H NMR (200 MHz, CD3OD): d = 8.17 (2H, d, J = 8);
7.65 (2H, d, J = 8); 6.22 (1H, s); 5.08 (1H, d, J = 2); 4.44–4.41
(2H, m); 4.24 (1H, d, J = 2); 3.02 (4H, s); 2.91 (2H, t, J = 7);
2.68 (2H, t, J = 7), 2.20 (2H, t, J = 7); 1.43 (1H, t, J = 7).
196 Vaks and Benhar
Fig. 2. Schematic representation of the chemical reactions used to prepare neomycin–chloramphenicol adduct for con-
jugation. (a) Two chemical steps were used to modify chloramphenicol (CAM) for conjugation to amine groups. In the first
step, the chloramphenicol primary hydroxyl group was reacted with glutaric anhydride to create an ester linkage, result-
ing in chloramphenicol-linker. In the second step, the free carboxyl group of the chloramphenicol-linker was activated
with NHS to allow subsequent linkage to amine groups. (b) The chloramphenicol–NHS was reacted with neomycin in a
solution of 0.1 M NaHCO3, pH = 8.5, resulting in neomycin–chloramphenicol adduct. The six primary amine groups of
neomycin are circled. (c) The resulting neomycin–chloramphenicol adduct is conjugated to free carboxyl groups of the
phage coat by the EDC procedure.
18. Store as dried powder at −20°C under argon. To dissolve, use

DMSO (see Note 19).
3.2.2. Conjugation 1. Mix solid neomycin and 100 mM chloramphenicol prodrug in

of Neomycin to the DMSO within 0.1 M NaHCO3, pH = 8.5, at a molar ratio of
Chloramphenicol Prodrug 1:2 for the chloramphenicol prodrug:neomycin (see Note 20).
2. Leave on stirrer overnight at room temperature.
3. Determine the prepared Neo–CAM adduct by reverse-phase
HPLC. Use reverse phase C-18 column on a Waters machine
with a gradient 0–100% of acetonitrile (stock solution of 80%
in water w/w) and water (100% water to 0%) in the mobile
phase, at 1 mL/min flow rate.
4. The Neo–CAM adduct should elute 18 min after sample
injection while the intact CAM prodrug should elute 24 min
after sample injection. An example of the results produced is
shown in Fig. 3 (see Note 21).
Fig. 3. Reverse-phase HPLC analysis of the Neo–CAM adduct. (a) HPLC analysis of
chloramphenicol–NHS prior to conjugation to neomycin. The chloramphenicol–NHS
prodrug was separated using a gradient of acetonitrile in water on a Waters HPLC
machine (RP; C-18 column). CAM–NHS was eluted 25 min postinjection. (b) HPLC analy-
sis of Neo–CAM adduct. The Neo–CAM adduct was separated using a gradient of ace-
tonitrile in water on a Waters HPLC machine (RP; C-18 column). The Neo–CAM adduct
was eluted at 18–19 min postinjection.
5. Conjugate Neo–CAM to biotinylated or antibody-complexed

phage nanoparticles by the EDC procedure.
3.2.3. EDC Conjugation In our system, drug conjugation with EDC is between exposed
Chemistry carboxyl side chains on the phage coat [most of those would be on
the major coat protein-p8 that contains four carboxylic amino-acid
residues at its exposed N-terminus (Glu2; Asp4; Asp5; Glu12)]
and neomycin that contains six primary amines (see Fig. 2).
1. Prepare a conjugation mix in a total volume of 1 mL containing:
0.1 M citrate buffer, pH = 5.0; 0.75 M NaCl; 2.5 × 10−6 mol
Neo–CAM; and 5 × 1012 g3p-AVITAG-fUSE5 phage particles
(see Note 22).
198 Vaks and Benhar
2. Add 2.5 × 10−6 mol of EDC and leave the reaction at room

temperature for 1 h while gently rotating (10 rpm). (see
Note 23).
3. Add the same amount of EDC and leave the reaction rotating
for 1 h.
4. Perform two-step dialysis against 0.3 M NaCl (see Notes 24
and 25).
3.3. Complexing The biotinylated phages that were prepared according to

g3p-AVITAG-fUSE5 Subheading 3.1.1 and conjugated to drug according to
Phages with IgG Subheading 3.2.3 are complexed through an avidin bridge with a
biotinylated antibody to confer them with target specificity.
1. Add 0.1 mg avidin to 1012 Neo–CAM conjugated phage in
1 mL of 0.3 M NaCl (see Note 26).
2. Incubate for 1 h at room temperature with gentle rotation at
10 rpm.
3. Add 0.3 mg biotinylated IgG to the phage–avidin complex.
Prepare a batch of phages in complex with the target-specific
IgG and a second batch in complex with a negative control
antibody.
4. Incubate for 1 h at room temperature and gentle rotation at
10 rpm on a benchtop tube rotator.
5. Store IgG-complexed phages at 4°C for up to 2 months. A
scheme of the complete targeted drug-conjugated phages is
shown in Fig. 4.
3.4. Growth Inhibition In the described experiment, S. aureus bacteria are treated with
Experiments (See chloramphenicol-carrying, antibody-targeted phages (the treat-
Note 27) ment group is in complex with target-specific IgG and the control
group is the phages conjugated to IgG that does not bind the
target bacteria) and growth rate is recorded. Normal rabbit serum
is added as a source of esterases to facilitate drug release.
1. Grow S. aureus bacteria culture overnight at TSB (see Note 1).
2. Collect 0.1–1 mL aliquot of bacteria culture by centrifugation
for 1 min at 15,000 × g in a microfuge at 4°C, and wash twice
by re-suspension and re-centrifugation in ice-cold PBS.
3. Re-suspend the bacteria in an equal volume of ice-cold PBS.
4. Incubate 10 mL of washed bacteria (~107 cells) with 100–
300 mL of targeted chloramphenicol-carrying phage nano-
particles (~1–3 × 1011 particles) for 1 h on ice.
5. Add an equal volume (100–300 mL) of normal rabbit serum
(see Note 28).
Fig. 4. A scheme of the complete targeted drug-conjugated phages. Each phage is con-
jugated to about 10,000 drug molecules on its coat. On the phage tip, the AVITAG peptide
that undergoes biotinylation is displayed on all (3–5) copies of the g3p minor coat pro-
tein. The biotin is bound by an avidin tetramer which through the unoccupied three other
biotin binding sites binds biotinylated antibodies (IgG). While for simplicity, the biotin–
avidin-biotinylated IgG is shown only once; in theory, every drug-carrying phage may be
targeted by many (up to 15) targeting antibodies if all binding sites are occupied. The
scheme is not drawn to scale.
6. Incubate for 3 h at 37°C.

7. Dilute 100–300 mL, respectively, of this mixture in 3 mL TSB
in 13 mL tubes.
8. Grow at 37°C by shaking at 250 rpm and monitor the absor-
bance at 600 nm. Plot the OD at 600 nm against time to
monitor bacterial growth. An example of the results produced
is shown in Fig. 5.
200 Vaks and Benhar
2.5
OD at 600nm
1.5
0.5
0
60 160 260 360 460 560 660
Time (min)
Fig. 5. Effect of drug-carrying antibody-targeted phages on the growth of S. aureus.

Growth curves of S. aureus cells treated with 3 × 1011 Neo–CAM carrying fUSE5-ZZ
phages, displaying specific anti-Staphylococcus antibodies bound via ZZ (filled squares)
or with nontargeted (Fc displaying) Neo–CAM carrying phages (open triangles).
Staphylococcus aureus grown with PBS (filled triangles) or with naked fUSE5-ZZ (open
circles) represents bacteria grown without any inhibitor.
3.5. In Vivo There are no existing “text book” about small animal disease
Experiments models for pathogenic bacteria. A model has to be designed care-
fully to allow efficient evaluation of the drug-delivery platform.
3.5.1. Staphylococcus
In the described experiment, mice are injected with S. aureus
aureus Disease Model
bacteria and are treated with chloramphenicol-carrying, antibody-
targeted phages. The mice are followed for signs of toxicity (such
as weight loss, apathic behavior, or death). This is a fatal disease
model, and untreated mice succumb to death within a few days.
Efficacy of the treatment is indicated by delaying of symptoms or
prolonging (or preventing) death.
1. Grow a S. aureus starter culture overnight in 3 mL of TSB in
a 13 mL culture test tube at 37°C by shaking at 250 rpm in a
shaking incubator (see Note 29).
2. Dilute the bacteria 1:100 in 200 mL of fresh TSB in 1 L
Erlenmeyer flask and grow at 37°C by shaking at 250 rpm in
a shaking incubator until the culture reaches A600nm = 1 (equiv-
alent to 5 × 108 bacteria/mL).
3. Collect the bacteria by centrifugation and wash the bacteria
twice in sterile PBS. Each wash consists of thoroughly re-
suspending the bacteria in PBS and collecting the cells by
centrifugation. Bring the bacteria to a final concentration of
5 × 109/mL in PBS.
4. Inject 109 bacteria per mouse into the tail vein of female
BALB/c mice (see Notes 30 and 31).
5. Monitor the survival and weight loss for 3 weeks (see Note 32).
3.5.2. In Vivo Therapeutic 1. Grow S. aureus culture overnight at TSB (follow steps 1–3 of
Activity of Targeted Subheading 3.5.1).
Drug-Carrying
2. Dilute 1:100 in TSB and grow at 37°C until the culture
Bacteriophages
reaches A600nm = 1 (equivalent to 5 × 108 bacteria/mL).
3. Wash the bacteria twice in sterile PBS and bring to a final
concentration of 5 × 109/mL.
4. Perform a number of comparable injections of S. aureus and
targeted drug-carrying phage with different time intervals
between bacteria and phage injections:
(a) Incubate 1 mL aliquot of washed bacteria with 1 × 1011
anti-S. aureus Neo–CAM carrying bacteriophage for 1 h
on ice. Wash in sterile-cold PBS. Re-suspend in 1 mL of
cold sterile PBS and inject the mice with 200 mL (=1 × 109
bacteria) into the tail vein.
(b) Precipitate 1 mL of 5 × 109/mL of S. aureus and re-suspend
in 1 mL of anti-S. aureus Neo–CAM carrying bacterio-
phage (1 × 1011). Inject the mice with 200 mL (=1 × 109
bacteria) into the tail vein.
(c) Inject 109 bacteria per mouse into the tail vein. 30 min,
1, 2, 4, 8, and 12 h later inject 1 × 1011 anti-S. aureus
Neo–CAM carrying bacteriophages into the tail vein.
(d) Inject 109 bacteria per mouse into the tail vein. 30 min,
1, 2, 4, 8, and 12 h later inject the mice with 1 × 1011
anti-S. aureus Neo–CAM carrying bacteriophage
intraperitonealy.
(e) Inject 109 bacteria per mouse into the tail vein. For 24
and 48 h, monitor the behavior and health of the mice.
When the therapy is examined, inject the mice with
1 × 1011 anti-S. aureus Neo–CAM carrying bacteriophage
into the tail vein/intraperitonealy. In subsequent experi-
ments, you may test different time gaps between the
injection of the pathogen and the phages.
5. Monitor the survival and weight loss of the mice for up to for
3 weeks.
4. Notes
1. During the research that was carried out in our laboratory, in

addition to S. aureus described here in detail, we showed the
possibility to target a variety of pathogenic bacteria, such as
Streptococcus pyogenes and avian pathogen E. coli O78. The
targeting ability depends exclusively on the targeting moiety
displayed on or linked to the phage coat.
202 Vaks and Benhar
2. Different targeting molecules, displayed on the phage coat

may be used to confer target specificity to the drug-carrying
phages. While the protocol describes in detail phages that dis-
play the AVITAG peptide and are linked to targeting anti-
bodies through an avidin–biotin bridge, three alternative
targeting moieties have been developed in our laboratory and
can be used:
(a) A specific anti-S. aureus 12-mer peptide VHMVAGPGREPT
that is displayed as an N-terminal fusion to the p8 (g8p)
major coat protein of the fth phage (A12C phage (3)).
This S. aureus binding peptide was isolated from a
disulfide-bond constrained 12-mer phage display library,
designed on the fth “type 88” expression vector (16).
The library was constructed and kindly provided by Prof.
Jonathan Gershoni’s group at Tel-Aviv University.
(b) A single-chain-specific antibacterial antibody (scFv) dis-
played on the N-terminus of the p3 minor coat protein of
the fUSE5 phage (scFv-fUSE5). The scFv display and the
two following display methods are based on the fUSE5
vector system developed for polyvalent display on p3 by
Smith and co-workers (5) (http://www.biosci.missouri.
edu/smithgp/PhageDisplayWebsite/vectors.doc).
(c) The ZZ domain displayed on the N-terminus of the p3
fUSE5 phage coat protein (17, 18). The ZZ domain is a
modified S. aureus protein fragment that specifically binds
the Fc region of the antibody. The FUSE5-ZZ bacterio-
phage is able to form a stable complex with target-specific
IgGs. The main disadvantage of the above display methods
is the fact that the targeting component may lose its binding
activity following the drug-conjugation chemistry.
3. Most of the materials and reagents that are listed may be
obtained from several vendors. We listed the vendors from
whom we routinely purchase, which does not mean that we
endorse the products of those particular vendors.
4. The bacteriophages from our laboratory collection: g3p-AVI-
TAG-fUSE5, as well as fUSE5-ZZ, scFv-fUSE5, and A12C
can be obtained from the authors upon request (3, 4).
5. The mice behavior and vitality should be monitored daily
during the experiment. Healthy animals usually are energetic
with shiny fur. One should mention a light increase in body
weight (up to 2–3% weekly). However, mice in significant
pain or distress typically display a lack of activity, sunken eyes,
ruffled fur, and weight decrease.
6. Care should be taken to avoid using phages that have chemi-
cally modifiable amino-acid residues at key contact residues of
the displayed peptides, as these may lose their target-binding
ability upon chemical conjugation of the drug.
7. Early experiments we carried out, in which this drug was

linked directly to free amine groups of the phage coat, pro-
vided evidence that it is not possible to link a large payload of
a hydrophobic drug such as chloramphenicol to the phages
that precipitate as a result.
8. The preparation of fUSE5-ZZ, scFv-fUSE4, and A12C bacte-
riophages is a routine protocol that can be found in ref. (19).
9. The plasmid pBirA carries a copy of birA biotin ligase down-
stream to the lac promoter that enables the birA protein effi-
cient expression following the IPTG addition to the growth
media. IPTG is widely used as an inducer for overexpression
of the cloned proteins.
10. For efficient growth and biotinylation of g3p-AVITAG-
fUSE5, we grow it for approximately 48 h supplied twice
with 0.1 mM IPTG. Following the first step of overnight
growth, the bacteria culture would not reach growth satura-
tion and would show low optical density. Add an additional
dose of IPTG (according to the protocol) and continue the
growth for the next 24 h.
11. Phage precipitation step with PEG/NaCl solution is aimed to
separate the phages from the bacterial growth media, concen-
trate them, and re-suspend in the desired solvent.
12. Use only 0.45 mm cutoff filters. Up to 50% of filamentous
bacteriophages (that reach 1 mm in length) are trapped and
lost in 0.22 mm filter.
13. For large volumes (0.5 L), we recommend to incubate the
phage mixture overnight at 4°C.
14. Re-suspend the phage pellet in distilled water (not PBS)
because it is preferable for the following chemistry steps.
15. We do not recommend freezing phage stocks because they
lose their infectivity probably due to structural instability. It is
best to keep it at 4°C.
16. The two approaches are usually used for phage quantification:
“live” titration, a process of infecting bacteria with diluted
phages followed by counting the resulting infected bacterial
colonies that grow on selective plates, and measuring the optical
absorbance at 269 nm. The live titration method counts viable
infective virus particles, while the optical method quantifies
everything that absorbs at 269 nm. In theory, the values
should be identical; however, in practice, the absorbance
method gives us up to ten times higher value. The difference
comes from impurities and noninfective phage particles.
17. Most phages that are used in research laboratories lyse the
infected bacteria and form plaques on bacterial lawns which
can be counted to calculate their number. However, it is more
convenient to count bacterial colonies than to count plaques.
204 Vaks and Benhar
Fortunately, many genetically engineered phages carry antibiotic

resistance genes and thus infected cells form colonies on the
appropriate selective media. In our protocol, phage quantifica-
tion by live titration is based on the fact that these specific bacte-
riophages carry in their genome a tetracycline resistance cassette
that provides tetracycline resistance to the infected bacteria. The
infection of tetracycline sensitive (tets) DH5aF¢ bacteria with the
above phages will result in tetracycline-resistant (tetR) bacteria.
The number of tetR bacteria colonies is proportional to phage
particle quantity the bacteria were infected with.
18. When you use streptavidin-coated magnetic beads, you also
should check PBS only to ensure that the magnetic beads are
not broken. Use nonbiotinylated phage as a control to deter-
mine the stickiness of the beads.
19. Divide small aliquots of CAM–NHS (for approximately ten
reactions) into tubes and keep under argon at −20°C. It is a
highly hygroscopic substance! Before use, bring a vial of
powder to room temperature and then open. Dissolve in
DMSO only before usage.
20. Add CAM–NHS last to the reaction.
21. It is possible to collect purified Neo–CAM using preparative
HPLC. However, according to our experience, a small
amount of unconjugated chloramphenicol does not interfere
with subsequent chemistry steps.
22. You can alternatively use 1012 fUSE5-ZZ phages previously
complexed with 0.1–0.3 mg IgG (for 1 h at room tempera-
ture) (3, 4). The scFv-fUSE5 and A12C phages are already
targeted and should be directly used for drug conjugation
(1012 phage particles) (3, 4). We found that EDC chemistry
harms the ZZ domain. Therefore, we recommend to com-
plex targeted IgGs to fUSE5-ZZ before drug conjugation.
Biotin, on the other hand, is impervious to EDC chemistry,
hence it is possible to conjugate the drug to biotinylated
phages before forming a complex with targeting antibodies.
23. EDC conjugation reaction is accompanied with a rise in pres-
sure; therefore, avoid using tubes that can easily be opened.
24. We perform dialysis using 10 kDa snakeskin dialysis tubing
(see Subheading 2) while during this step we lose some
amount of phage particles. However, if a smaller cutoff is
used, some phage particles can precipitate on the snakeskin
membrane and clog it.
25. Following the EDC chemistry, bacteriophages partially lose
their infectivity and change their absorbance at 269 nm;
therefore, it is impossible to determine the exact drug-carrying
phage concentration. According to our calculation, the final
dialyzed drug-conjugated bacteriophage is ~1 × 1012/mL.
26. It is important not to add a too high concentration of avidin,

because free avidin can trap the biotinylated IgG added at the
next step and can eliminate its binding to phages.
27. According to our calculation, we can conjugate up to 10,000
chloramphenicol molecules per phage.
28. Pay attention that serum is clean and noncontaminated.
Filter-sterilize it if needed.
29. Bacteria dosage for injection may differ among the strains.
We worked with fairly pathogenic bacteria; thus, the amounts
that were used for lethal model were very high. When highly
pathogenic bacteria are available, a lower concentration of
bacteria may be injected with the same drug-carrying phage
dosage. This can extremely improve the therapeutic results.
30. We injected the bacteria and the drug directly to the tail vein
of the mice. As an alternative, intraperitoneal administra-
tion can be performed. It allows using larger injection
volumes, and more easily performed by inexperienced
experimenters.
31. It is highly recommended to warm the animal in an incubator
or under an incandescent light. This procedure makes the tail
veins more visible for injection. Use 28 gauge needles. Be sure
there are no air bubbles in the solution to be injected, as this
can harm the mice. Before injection, wipe the injection site
clean with a disinfecting gauze to avoid unintended infection.
32. In this disease model, expect a massive decrease in body
weight of infected animals leading to at least 80% mortality
within 1 week from injection of the S. aureus bacteria.
Acknowledgments
Studies of targeted drug-carrying phage nanomedicines at the

author’s laboratory received a grant from the Israel Public
Committee for Allocation of Estate Funds, Ministry of Justice,
Israel and by the Israel Cancer Association.
References
1. Yoshikawa, T. T. (2002) Antimicrobial resis- agents by using drug-carrying filamentous
tance and aging: beginning of the end of the bacteriophages. Antimicrob. Agents Chemother.
antibiotic era? J. Am. Geriatr. Soc. 50, 50, 2087–2097.
S226–S229. 4. Yacoby, I., Bar, H., and Benhar, I. (2007)
2. Forget, E. J. and Menzies, D. (2006) Adverse Targeted drug-carrying bacteriophages as
reactions to first-line antituberculosis drugs. antibacterial nanomedicines. Antimicrob.
Expert Opin. Drug Saf. 5, 231–249. Agents Chemother. 51, 2156–2163.
3. Yacoby, I., Shamis, M., Bar, H., Shabat, D., 5. Smith, G. P. (1985) Filamentous fusion phage:
and Benhar, I. (2006) Targeting antibacterial novel expression vectors that display cloned
206 Vaks and Benhar
antigens on the virion surface. Science 228, 13. Sidhu, S. S., Weiss, G. A., and Wells, J. A.
1315–1317. (2000) High copy display of large proteins on
6. Beckett, D., Kovaleva, E., and Schatz, P. J. phage for functional selections. J. Mol. Biol.
(1999) A minimal peptide substrate in biotin 296, 487–495.
holoenzyme synthetase-catalyzed biotinyla- 14. Bar, H., Yacoby, I., and Benhar, I. (2008)
tion. Protein Sci. 8, 921–929. Killing cancer cells by targeted drug-carrying
7. Scholle, M. D., Kriplani, U., Pabon, A., phage nanomedicines. BMC Biotechnol. 8, 37.
Sishtla, K., Glucksman, M. J., and Kay, B. K. 15. Berkowitz, S. A. and Day, L. A. (1976) Mass,
(2006) Mapping protease substrates by using length, composition and structure of the fila-
a biotinylated phage substrate library. mentous bacterial virus fd. J. Mol. Biol. 102,
Chembiochem 7, 834–838. 531–547.
8. Zou, J., Dickerson, M. T., Owen, N. K., 16. Enshell-Seijffers, D., Smelyanski, L., and
Landon, L. A., and Deutscher, S. L. (2004) Gershoni, J. M. (2001) The rational design of
Biodistribution of filamentous phage peptide a “type 88” genetically stable peptide display
libraries in mice. Mol. Biol. Rep. 31, 121–129. vector in the filamentous bacteriophage fd.
9. Molenaar, T. J., Michon, I., de Haas, S. A., Nucleic Acids Res. 29, E50–E60.
van Berkel, T. J., Kuiper, J., and Biessen, E. A. 17. Nilsson, B., Moks, T., Jansson, B., Abrahmsen,
(2002) Uptake and processing of modified L., Elmblad, A., Holmgren, E., et al. (1987) A
bacteriophage M13 in mice: implications for synthetic IgG-binding domain based on staph-
phage display. Virology 293, 182–191. ylococcal protein A. Protein Eng. 1, 107–113.
10. Benhar, I. (2001) Biotechnological applica- 18. Nilsson, J., Larsson, M., Stahl, S., Nygren, P.
tions of phage and cell display. Biotechnol. A., and Uhlen, M. (1996) Multiple affinity
Adv. 19, 1–33. domains for the detection, purification and
11. Cortese, R., Felici, F., Galfre, G., Luzzago, immobilization of recombinant proteins.
A., Monaci, P., and Nicosia, A. (1994) Epitope J. Mol. Recognit. 9, 585–594.
discovery using peptide libraries displayed on 19. Enshell-Seijffers, D. and Gershoni, J. M.
phage. Trends Biotechnol. 12, 262–267. (2002) Phage display selection and analysis of
12. Hoogenboom, H. R., de Bruine, A. P., Ab-binding epitopes. in: Current Protocols in
Hufton, S. E., Hoet, R. M., Arends, J. W., Immunology (Coligan, J. E., Bierer, B. E.,
and Roovers, R. C. (1998) Antibody phage Margulies, D. H., Shevach, E. M., and
display technology and its applications. Strober, W., eds) pp. 9.8.1-9.8.27. John Wiley
Immunotechnology 4, 1–20. & Sons, Inc., USA.
Chapter 14
Viruses as Nanomaterials for Drug Delivery

Dustin Lockney, Stefan Franzen, and Steven Lommel
Abstract
Virus delivery vectors are one among the many nanomaterials that are being developed as drug delivery
materials. This chapter focuses on methods utilizing plant virus nanoparticles (PVNs) synthesized from
the Red clover necrotic mosaic virus (RCNMV). A successful vector must be able to effectively carry and
subsequently deliver a drug cargo to a specific target. In the case of the PVNs, we describe two types of
ways cargo can be loaded within these structures: encapsidation and infusion. Several targeting approaches
have been used for PVNs based on bioconjugate chemistry. Herein, examples of such approaches will be
given that have been used for RCNMV as well as for other PVNs in the literature. Further, we describe
characterization of PVNs, in vitro cell studies that can be used to test the efficacy of a targeting vector,
and potential routes for animal administration.
Key words: Bioconjugation, Capsid, Drug delivery, Encapsidation, Infusion, Nanomaterial, Peptides,
Plant virus nanoparticle, Targeting
1. Introduction
The methods and strategies we use to synthesize a plant virus

nanoparticle (PVN), as a drug delivery platform, are related to
the structure and function of Red clover necrotic mosaic virus
(RCNMV). Therefore, it is important to review the basic structure
of RCNMV and how it functions as a virus before we discuss the
methods for PVN synthesis.
RCNMV has an icosahedral protein capsid that contains its
highly organized bipartite genome, named RNA-1 and RNA-2
(Fig. 1). RCNMV is ~36 nm in diameter and is constructed from
180 chemically equivalent 37 kDa capsid proteins (CPs) (1).
These CPs assume one of three conformations in an ABC trimer,
called an icosahedral asymmetric unit (IAU) (Fig. 1). Sixty of
these IAU’s are needed to form the T = 3 capsid (1). The T = 3
207
208 Lockney, Franzen, and Lommel
Fig. 1. Cryoelectron microscopy image of RCNMV and the quasi-threefold axis of the IAU with and without divalent
cations. The capsid shows the icosahedral symmetry. Open: RCNMV free of Ca2+ and Mg2+ (a) and closed : RCNMV in the
presence of Ca2+ and Mg2+ (b).
value is the triangulation number (T) and refers to the number of

CPs in the IAU. The size of the virus can also be related to the
T number. A T = 1 virus has a single protein as the subunit for viral
assembly and is the smallest virus. Depending on the size of the
capsid subunit, T = 1 viruses can be as small as 20 nm in diameter.
The most common structure in the field of plant virus nanoparti-
cles (NPs) is the T = 3. Brome mosaic virus (BMV), Cowpea chlorotic
mottle virus (CCMV), Cowpea mosaic virus (CPMV), Hibiscus chlo-
rotic ringspot virus (HCRSV), Tomato bushy stunt virus (TBSV),
and RCNMV are all T = 3 viruses which have diameters in the range
from 29 to 37 nm (2–5).
RCNMV, like other plant viruses, does not use an endosomal
mechanism for entry into its host. Plant viruses must use a vector,
such as an insect, fungus, or animal, to create a mechanical
opening in the cell for virus entry. Once the virus has entered the
cell, the virus disassembles and delivers its genome. The mecha-
nisms of virus assembly and disassembly comprise an entire field
of research. It is important to realize that capsid assembly and
disassembly must be energetically favorable in the same cell. In
addition, viral proteins and nucleic acids are able to assemble into
virions containing only viral proteins and/or nucleic acids.
Investigation into these mechanisms has elucidated some features
of RCNMV that are important in PVN synthesis.
Viruses as Nanomaterials for Drug Delivery 209
Investigation into the possible disassembly mechanism (also

known as uncoating) of RCNMV, using cryoelectron microscopy,
revealed conformational changes that are sensitive to the presence
of Ca2+ and Mg2+ (1). Figure 1a shows that when Ca2+ and Mg2+
are removed, the CPs rotate and move away from the center of the
quasi-threefold axis (1). This conformational transition leads to
the opening of a 10–13 Å channel at the center of each trimer axis.
This indicates that when RCNMV is in a Ca2+ and Mg2+ rich environ-
ment (e.g., soil), it will be in a closed conformation (Fig. 1b). When
RCNMV is in a cytosolic environment, where the concentration
of Ca2+ is ~100 nM, the capsid will be in an open conformation.
Further work on understanding how RCNMV functions has
led to the identification of genomic sequences/structures that are
important for viral replication and genome packaging. In particular
the transactivator (TA), located on RNA-2, is a hairpin structure
that hybridizes to the transactivator binding sequence (TABS)
on RNA-1. The TA is necessary for CP expression and is hypoth-
esized to be the origin of assembly (OAS). The OAS is thought
to act as a nucleation site for capsid assembly.
It is most convenient to formulate strategies and methods,
when engineering a plant virus nanoparticle that exploit the
intrinsic properties of the plant virus. Figure 2 shows an overview
Express Manipulate capsid Load cargo

RCNMV proteins
In Vitro or Purify and Attach targeting

in vivo characterize peptides
assay
Fig. 2. Representation of the general protocol for synthesis of a PVN. The sequence involves isolation of purified plant
viruses followed by capsid protein manipulation. Then a therapeutic or other molecule of interest is infused. Targeting
peptides are added to the surface using bioconjugate chemistry methods. This is followed by purification and character-
ization. Finally the PVN is tested in vitro or in vivo.
of the procedures used to synthesize PVNs for testing both

in vitro and in vivo. The first step is expression and purification
of RCNMV from a host. Once we have our plant virus we manip-
ulate the capsid proteins in preparation for cargo loading. Cargo
loading can be done by either encapsidation or infusion. For
encapsidation, the virus is completely disassembled and then reas-
sembled around the cargo. Infusion exploits the swelling properties
of RCNMV and is the method used to infuse cargo (e.g., doxo-
rubicin) into the virus. Next, the virus is functionalized with the
desired targeting moieties; we use small peptides. Once the nano-
particle is synthesized it is very important to purify and properly
characterize the product. Finally, the PVN is tested in vitro and
then in vivo. Herein, we will discuss these procedures.
2. Materials
2.1. Purification 1. 10 mM sodium phosphate buffer, pH = 7.0.

of Red Clover 2. 200 mM sodium acetate buffer, pH = 5.3.
Necrotic Mosaic Virus
3. Carborundum.
4. Cheesecloth.
5. DNA plasmids: RC-169 and RC-2.
6. Homogenizing buffer: 200 mM sodium acetate pH = 5.3,
1,000-fold dilution of b-mercaptoethanol.
7. High capacity ultracentrifuge.
8. Nicotiana clevelandii plants.
9. Miracloth.
10. Mortar and pestle.
11. Polyethylene glycol (PEG) 8000.
12. Sucrose.
13. MEGAshortscript™ (Applied Biosciences).
14. Viral RNA transcripts: RNA-1 and RNA-2.
15. Deionized water.
2.2. Encapsidation 1. Appropriate NPs less than 15 nm in diameter. This can include
Au, CdSe, Fe3O4, etc. We will use Au NPs coated with bis-
sulfonatophenyl phenylphosphine (BSPP) as an example.
2. An oligonucleotide with an appropriate chemical group for
attachment to the NP (e.g., alkane thiolate linker for Au).
3. RCNMV, prepared as discussed in Subheading 3.1.
4. RNA-1 transcript (1,539 bp).
5. Micro Bio-Spin P-30 columns (Biorad).
6. 100 mM dithiothreitol (DTT).
7. 10 mM sodium phosphate buffer, pH = 7.0.

8. 0.1 M NaCl, 10 mM sodium phosphate, pH = 7.0.
9. 50 mM Tris–HCl buffer, pH = 5.5.
10. 200 mM sodium ethylenediaminetetraacetic acid (EDTA),
pH = 10.
11. Sephadex G75 columns.
12. Slide-A-Lyzer dialysis cassette (10 kDa molecular weight
cut-off (MWCO)).
13. Sucrose cushion.
14. 100 mM glycine–NaOH, pH = 10.
2.3. Infusion 1. 1× Dulbecco’s phosphate buffered saline (1× DPBS).

2. 6,000–8,000 MWCO dialysis tubing. A higher MWCO
(100 kDa) may be substituted to remove possible proteolytic
enzymes or degraded coat proteins that are present from the
purification process.
3. Conjugation buffer 1: 50 mM sodium phosphate buffer,
pH = 7.25.
4. Conjugation buffer 2: 50 mM 4-(2-hydroxyethyl)-1-pipera-
zineethanesulfonic acid (HEPES), 50 mM NaCl, pH = 7.25.
5. Dimethyl sulfoxide (DMSO, cell grade).
6. Infusion buffer: 50 mM Tris base, 50 mM sodium ethylene-
diaminetetraacetic acid (EDTA), 50 mM sodium acetate,
pH = 7.0.
7. Infusion dye.
8. NAP 25 column.
9. RCNMV, prepared as discussed in Subheading 3.1.
10. Sephadex G25 column (30 cm L × 1.1 cm I.D.). The sephadex
G25 may be replaced by a higher molecular weight exclusion
resin (sephadex G200 is very common) if it provides better
resolution between the eluted virus peak and infusion dye.
2.4. Decoration of the 1. 1× Dulbecco’s phosphate buffered saline.

PVN Surface with 2. Dimethyl sulfoxide (DMSO, cell grade).
Targeting Peptides:
3. NAP 25 column.
Conjugation Protocols
4. Peptides.
5. Sephadex G25 column (30 cm × 1.1 cm I.D.). A higher
molecular weight exclusion resin may be used if it provides
better resolution of the virus and excess peptides.
6. Sulfosuccinimidyl 4-[N-maleimidomethyl]cyclohexane-1-
carboxylate (Sulfo-SMCC). There are derivatives containing
polyethylene glycol spacer arms that can be substituted.
7. 50 mM sodium phosphate, 50 mM NaCl, pH = 7.0.
3. Methods
3.1. Purification Plant virus nanoparticle production begins with the simple
of Red Clover mechanical inoculation of the virus or a clone (or transcript) of
Necrotic Mosaic Virus the virus to the leaf of a susceptible plant. The virus infection
spreads from the inoculated cells through the plasmodesmata
(cell-to-cell junctions) and the vascular system of the plant, thus
resulting in a large percentage of the cells comprising the plant
becoming infected. After 1 week, the plant reaches its maximum
capacity for production of virus and can be harvested. The bio-
mass of the virus can be as high as 1% of the plant tissue. Plant
virus yields vary depending on the species of plant virus and the
host combination, but on the high end can yield 2–10 mg of virus
per gram of wet weight tissue. Plant bioreactors represent well
developed and robust technology that can be employed for the
production of large scale quantities of the virus NPs under phar-
maceutical and good manufacturing practice (GMP) conditions.
Plant bioreactors as compared to cell bioreactors offer both distinct
advantages and drawbacks. The high yields in plants are not always
reproducible in cells in a bioreactor. However, plant tissue pres-
ents greater problems for purification.
1. Transcribe the genome of RCNMV from DNA templates
RC-169 and RC2 using MEGAshortscript™ (T7-polymerase
kit).
2. Mix 1 mL of RNA-1 and RNA-2 transcripts together and
dilute to 110 mL with 10 mM phosphate buffer, pH = 7.0.
3. Gently rub carborundum, an abrasive powder, on the leaves
of Nicotiana clevelandii plants and then wash using dH2O.
4. Inoculate the leaves with 27 mL RNA transcript mixture by
supporting the underside of the leaf with one hand and gently
rubbing the drops of inoculum over the surface of the leaf
with the other. Wash away the inoculum with dH2O and
allow the plants to incubate at 18–26°C for 7–10 days.
5. Combine three infected leaves with 5 mL dH2O and grind
with a mortar and pestle.
6. For a large preparation of RCNMV, rub this pulverized plant
material on carborundum treated plants using a sponge. Gently
rinse off inoculum and carborundum with a water hose.
7. Incubate these plants in a greenhouse 7–10 days. Symptoms
should appear in 4–5 days as little ring spots.
8. Harvest the whole plants by cutting the main stems and weigh
the tissue.
9. Combine plant tissue with homogenizing buffer (one part
tissue: two parts buffer) and homogenize using a blender
(blender capacity: 350 g tissue, buffer capacity: 700 mL).

Blend for 30 s (three times) at low speed.
10. Gradually pour slurry through four layers of cheesecloth into
a beaker to remove plant debris.
11. Twist cheesecloth into a ball and squeeze through remaining
liquid.
12. Incubate on ice for 10 min with stirring.
13. Pour above slurry into 250-mL centrifuge bottles and spin at
6,600 × g for 25 min.
14. Pour supernatant through a sheet of miracloth into a gradu-
ated cylinder.
15. Measure the filtrate volume and pour it into a beaker.
16. Add ¼ the volume of 40% PEG 8000/NaCl to clarified
filtrate.
17. Incubate filtrate on ice for 1 h with stirring at low speed.
18. Pour precipitated samples into centrifuge bottles and centri-
fuge at 6,600 × g for 20 min at 4°C.
19. Discard supernatant to waste container and keep pellet by
draining inverted tubes on a paper towel.
20. Resuspend each pellet in centrifuge tube with 30 mL 200 mM
sodium acetate buffer, pH = 5.3, and transfer to 30-mL
centrifuge tubes. Centrifuge at 7,800 × g for 20 min at 4°C.
21. Save supernatant at 4°C for later use or continue purification.
22. Add 3 mL 20 % sucrose to supernatant.
23. Ultracentrifuge above supernatant at 163,000 × g for 2 h at
5°C.
24. Discard supernatant.
25. Add 500 mL 200 mM sodium acetate buffer, pH = 5.3 to
surface of pellet and incubate at 4°C overnight.
26. Collect suspension and portion to a microcentrifuge tube.
Centrifuge at 9,800 × g for 5 min, three times, each time col-
lecting supernatant and moving it to a fresh microcentrifuge
tube.
27. Determine A260 and A280. Calculate the concentration of
RCNMV using e260 = 6.46 mL/mg/cm, and assess purity
compared to a standard value of A260/A280 = 1.69.
28. Viruses can be stored at 4°C or frozen.
3.2. Encapsidation Encapsidation refers to inclusion inside the virus capsid protein
shell using the viral RNA OAS as scaffolding. If the RNA is not
present, the protein shell can still be used to encapsulate NPs by
various routes (6–8) with a strong contribution due to electro-
static interactions (9, 10). Several groups have explored the use of
plant virus capsid proteins to encapsidate gold NPs up to 15 nm

in diameter (11, 12), collections of smaller particles including
quantum dots (13), or metal oxide particles (9, 14).
The encapsidation protocol consists of packaging NPs and
proteins with radii ranging from 4 to 15 nm. This protocol is
demonstrated using a particle bioconjugated with a short hairpin,
DNA-2, a thiolated DNA analog of the RCNMV OAS (Fig. 3a)
(9, 12, 15). An RNA-1 fragment, containing the TABS, is hybrid-
ized with the OAS (Fig. 3b) to facilitate in vitro self-assembly of
the virus (Fig. 3c). Based on these studies, it is evident that many
plant viruses have a cargo capacity equivalent to a sphere of 10 nm
or greater.
1. Synthesize 5¢-thiol deoxyuridine modified DNA oligonucle-
otides, DNA-2, with the sequence: 5¢-SH-AGAGGUAUCG
CCCCGCCUCU-3¢. In order to deprotect the thiol group,
add 1 mL 100 mM DTT to DNA-2 and allow to react for
30 min at room temperature.
2. Remove excess DTT using a Micro Bio-Spin P-30 column.
3. Synthesize Au NPs coated with bissulfonatophenyl phe-
nylphosphine (BSPP).
4. Incubate a 1:300 mole ratio of Au NPs with DNA-2 at 37°C
for 8 h in 100 mL 10 mM phosphate buffer, pH = 7. Dilute to
500 mL with 0.1 M NaCl, 10 mM phosphate buffer, pH = 7
and incubate for an additional 40 h. Unattached DNA should
be removed by centrifugation at 14,000 rpm in a SS-34 rotor
(23,000 × g) for 25 min. Remove the supernatant, which con-
sists of the unreacted DNA. Resuspend the precipitate in
500 mL of 10 mM phosphate buffer, pH = 7 and recentrifuge.
This step should be repeated twice. Finally, suspend the
DNA/Au conjugates in 10 mM phosphate buffer, pH = 7.
5. Obtain full-length RCNMV RNA-1 used for encapsidation
by transcription from a Sma1 linearized plasmid vector in
RNA-1
a b c
RNA-1
Coat protein
DNA-2
Gold nanoparticle Encapsidated

gold nanoparticle
Fig. 3. Schematic representation of encapsidation strategy using DNA-2, an analog of the TA of RNA-2, as the origin of
assembly, combined with synthetic RNA-1 to capture CPs.
which the bacteriophage T7 RNA polymerase promoter was

used to transcribe full-length RCNMV cDNA clones.
6. Prepare RCNMV CPs by suspending RCNMV at a final con-
centration of 0.5 mg/mL in 200 mM EDTA, pH = 10.0 at
room temperature. Pellet aggregates at 10,000 × g for 10 min
and collect supernatant. EDTA causes the opening of the 60
pores in the capsid.
7. Separate the CP from the genome by size-exclusion chroma-
tography (Sephadex G 75). The concentration of CP can be
estimated using a Bradford assay.
8. Dissolve 20 mL of RCNMV CP (0.5 mg/mL) in a total vol-
ume of 100 mL with glycine–NaOH, pH 10 at room tem-
perature. Use of the glycine buffer minimizes capsid protein
aggregation.
9. Dialyze using Slide-A-Lyzer Dialysis Cassette (10 kDa
molecular weight cutoff) against 50 mM Tris buffer at three
different pH’s (5.5, 6.0, and 6.5) overnight at room tem-
perature. A sample at each pH should be analyzed by dynamic
light scattering (DLS) to observe appropriate PVN diameter
of ~34 nm.
10. Add DNA/nanoparticle conjugates to 1 mL T7 RCNMV
RNA-1 transcripts (4 mg/mL). Incubate this mixture for
10 min, and then add 5 mL purified RCNMV CP (10 mg/mL)
(see steps 6–9).
11. Carry out the encapsidation reaction by dialysis of the sample
against 50 mM Tris–HCl, pH = 5.5 overnight at room tem-
perature, using a Slide-A-Lyzer Dialysis Cassette (10 kDa
MWCO). Separate unencapsidated virus by ultra centrifuga-
tion through a sucrose cushion at 218,000 × g for 20 min
using a SW-55 rotor.
3.3. Infusion The swelling and pore opening of the PVN can be used to incor-
porate small molecules, peptides, or oligonucleotides. The infu-
sion process exploits a natural mechanism employed by the virus
to release its genome upon entry into a newly infected cell. First,
the virus is opened using ethylenediamine tetraacetic acid
(EDTA) which removes Ca2+ and Mg2+ from the solution.
Second, molecules are infused into the opened virions. Initially,
this infusion protocol was demonstrated for rhodamine (positive
charge), luminarosine (neutral) fluorescein (negative charge),
and doxorubicin (positive charge) (15). The PVN was exposed
to a 1,000-fold excess of the molecule. After an incubation
period, the PVNs were closed by addition of Ca2+ and Mg2+.
Subsequent experience has shown that 900–1,000 doxorubicin
molecules can reproducibly be infused into the RCNMV capsid
by these methods.
1. Prepare dialysis tubing by soaking for 15 min and washing

thoroughly with dH2O, inside and out, to remove sodium
azide preservative.
2. Dilute RCNMV stock sample to 5 mg/mL with distilled
water. Use e260 = 6.46 mL/mg/cm when determining
concentration.
3. Place the desired amount of RCNMV in dialysis tubing and
seal using clips. Allow to dialyze over night at 4°C.
4. Dissolve dye/chemotherapeutic in 100% DMSO to a
concentration ³ 5mM.
5. Remove RCNMV from dialysis tubing and aliquot 1 mL
volumes into 1.5-mL microcentrifuge tubes. Add dye/
chemotherapeutic to the RCNMV aliquots at a dye:virus
mole ratio of 1,000:1 (see Note 1). The infusion buffer should
be at a final concentration of £10% DMSO, 50 mM Tris base,
50 mM sodium acetate, 50 mM EDTA, pH 7.1. A higher
concentration of DMSO may be used if it is shown that the
virus is stable.
6. Wrap the dye/chemotherapeutic with virus in aluminum foil
and incubate at room temperature for a minimum of 4 h and
maximum of 24 h.
7. Remove excess dye/chemotherapeutic from the suspension/
solution using a NAP 25 column preequilibrated with buffer
of choice (see Note 2). If proceeding to conjugation of
peptides, equilibrate the NAP 25 column with 30 mL of
Conjugation buffer 1 or 2 (see Note 3).
8. Trace Ca2+ in buffers used for gel filtration is sufficient to close
the virus.
9. Pellet aggregates at 10,000 × g for 10 min at room temperature.
3.4. Decoration The virus capsid provides a highly organized array of amino acids
of the PVN Surface on the icosahedral structure of the capsid. Common methods of
with Targeting achieving structural modifications in PVN synthesis use a combi-
Peptides: Conjugation nation of bioconjugation techniques and/or mutagenesis.
Protocols Bioconjugation strategies are used for attaching fluorophores,
biotin, folic acid, chemotherapeutic drugs, polyethylene glycol
(PEG), antibodies, and peptides. The most useful amino acid
residues for chemical modification are lysines and cysteines, and
to a lesser extent aspartic and glutamic acids (16). Lysines are
abundant in proteins and will react with N-hydroxysuccinimidyl
esters (NHS-esters). Glutamic and aspartic acids are also abundant
and can be modified using 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide (EDC) to form a reactive species that will undergo
an amidation reaction with a primary amine (17). Reaction of
aspartic and glutamic acids may decrease the colloidal stability
of the viral suspension and is an uncommon practice in viral
nanotechnology. Cysteines are the least abundant amino acid on

the surface, but provide a point for selective conjugation using
maleimides. In our lab, we use succinimidyl-4-(N-maleimidom-
ethyl) cyclohexane-1-carboxylate (SMCC) for the orthogonal
conjugation of cysteine terminated peptides to the surface lysines
of the RCNMV (Fig. 4).
1. Prepare RCNMV (1–5 mg/mL) in 50 mM sodium phos-
phate, 50 mM sodium chloride buffer, pH = 7.0. The phos-
phate buffer may be substituted with HEPES or MES, if
necessary.
2. Portion RCNMV into 1 mL aliquots.
3. Dissolve 1–2 mg of sulfo-SMCC in 100 mL 100% DMSO.
4. Slowly add the sulfo-SMCC solution to the 1 mL virus aliquot,
vortex immediately.
5. Check pH. The pH should be 7.2 ± 0.1. Ideally, the ionic
strength of the buffer should be strong enough such that the
pH does not need to be adjusted (see Note 4).
O
O O
C N
+ N O
−NH2 O
SMCC
O
O
O
C N +
N peptide
H O HS
peptide
O
S
O
C N
N
H
O
RCNMV
Fig. 4. Bioconjugate coupling by SMCC is shown schematically. In this scheme a surface

lysine on the CP is conjugated to a terminal cysteine of a targeting peptide.
6. Allow the reaction to incubate at room temperature on a

rocker plate for 30–45 min. This is enough time to function-
alize the surface of RCNMV with maleimides.
7. Remove excess sulfo-SMCC using a NAP-25 column pre-
equilibrated with 50 mM sodium phosphate buffer, 50 mM
sodium chloride, pH = 7.0 (or HEPES).
8. Dissolve peptides at a concentration of 1–2 mg/mL in ~50 mL
100% DMSO. Make sure the solution is clear, indicating the
peptide has fully dissolved.
9. Slowly add the peptide solution to the maleimide activated
virus and check the pH. The pH should be between 6.8 and
7.3.
10. Incubate for at least 6 h to overnight at room temperature. If
stored at lower temperature (i.e., 4°C), the reaction must be
allowed to continue for 24 h.
11. Pellet aggregates at 10,000 × g for 10 min at room
temperature.
12. Remove excess peptide using size-exclusion chromatography
(i.e., 30 × 1.1 cm sephadex G 25, 50, 75, or 200 column) (see
Note 5).
13. Concentrate samples using centrifugal filters (100 kDa
MWCO is best) at ~7,000 × g, if necessary. Do not use poly-
styrene filters as this can cause severe aggregation.
3.5. Characterization One major objective of PVN synthesis is to specifically deliver a

chemotherapeutic to cancerous tissue. However, success depends
upon accurate understanding of the mechanism in which a PVN
targets and enters a cell. Any testable hypothesis that is made of
the mechanism is highly dependant on purity and careful charac-
terization of the nanomaterial. For the most part, purification of
PVNs is predominantly done by size-exclusion chromatography,
but anion exchange can be used as well. Density gradient cen-
trifugation, sucrose or iodixanol, can also be used to determine
whether infusion is successful. DLS and transmission electron
microscopy (TEM) can be used to verify if the synthesized particle
is intact and has the correct dimensions.
It is important to realize that any manipulation of the PVN
will result in some level of aggregation. The amount of aggrega-
tion varies depending on the procedure performed and is not
always noticeable by eye. Aggregation that is not noticeable by
eye can be detected using DLS. DLS is best measured at a con-
centration of 0.5 mg/mL. Removal of these aggregates can be
done by low speed centrifugation (~10,000 × g for 10 min).
For TEM analysis, we use either carbon type A or B copper
grids from Ted Pella with a 2% uranyl acetate stain. It is important
to not overload the grid. We have found that a 20 mL drop of
10–20 mg/mL sample, placed on a grid for 30 s, is sufficient for

particle detection. The excess drop can be wicked away using a
paper towel. The virus that remains can then be stained with
20 mL of 2% uranyl acetate for 30 s and the excess solution can be
wicked away in the same manner. Allow the virus sample to dry
before placing it into the grid holder.
3.6. In Vitro Cell The PVN platform can be tested for efficacy using in vitro cell
Studies of PVN culture. The cells of interest are grown to 80% confluence in
Delivery appropriate plates (12- or 96-well depending on the signal-to-
noise ratio needed in a plate reader measurement). Subsequently,
the PVN formulation is introduced to the plated cells in serial
dilutions. The PVN formulation at the end of the purification
process is suspended in buffer (1× DPBS, 10 mM HEPES
pH = 7.1, or 10 mM phosphate buffer pH = 7.1), and not in
growth media. For delivery, the sample must be filter sterilized.
We use 0.2-mm syringe filters. It is acceptable to add the formulation
directly to the well; however this will dilute the growth media.
Control lanes for mock deliveries should be included to test for
alteration of cell survival. To minimize buffer effects, one should
avoid dilutions greater than ~10% of the volume.
The efficacy of the delivery can be determined using a surviv-
ability assay. There are several standard kits that can be used to
determine the percent survival in a 12- or 96-well plate format.
The adenosine triphosphate (ATP) and 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays are most
widely used. The ATP assay monitors the level of ATP, which is a
measure of cell viability. The assay is based on the ATP-dependent
reaction of luciferin with oxygen catalyzed by the enzyme
luciferase. The chemiluminescent signal from the reaction is pro-
portional to the number of living cells. This value is determined
at an initial time and then the value is used as a reference for sub-
sequent measurements as a function of time. The MTT assay is
a colorimetric test that measures the enzymatic activity of the
cell. This test also relies on the cell density as a reporter of the
cell viability. These assays can be used to determine whether
chemotherapeutic agents such as a doxorubicin have been deliv-
ered to cells.
3.7. Routes Animal trials testing the immunogenicity and clearance rates of
of Administration PVNs use a parenteral route of administration. Injection into the
of PVNs saphenous vein is the common route in a murine model. Rapid
clearance is often observed in studies of PVNs and other NPs.
As mentioned above, surface attached PEG provides a method
to prolong circulation, presumably by avoiding uptake by the
reticuloendothelial system (RES). This has been demonstrated
in diverse nanoparticle platforms including liposomes, Au NPs,
polymers, and PVNs.
4. Notes
1. To increase loading, a higher dye:virus mole ratio can be used.

However, the stability of the virus can be compromised by
overloading.
2. The choice of buffer greatly depends on the solubility of the
dye/chemotherapeutic agent. Otherwise, the conjugation
chemistry works best in a phosphate buffer.
3. If one wishes to calculate quantum yields or use sample in a
drug delivery experiment, then extra purification methods
must be used to remove the excess cargo. Fluorophores such
as rhodamine or doxorubicin self-quench when infused.
4. If the pH exceeds 8 during this step, the sample should be
discarded. RCNMV is not stable at pH > 8.0 and hydrolysis
of the NHS-ester is faster at higher pH.
5. Peptide conjugation can be validated using several methods.
The most accurate method for quantization and verification
of peptide conjugation is mass spectrometry. Gel electropho-
resis can be used to verify peptide conjugation and spectros-
copy can be used to quantify the number of peptides attached
if they have a spectroscopic handle (fluorophore or chro-
mophore). However, the extinction coefficient of the spec-
troscopic handle can change significantly when attached to a
peptide on the surface of a virus. Moreover, the folding of
peptides can be disrupted when attached to the surface of a
virus. Validation of peptide conjugation by these methods
does not necessarily indicate that a peptide will retain its
targeting capabilities after conjugation. Molecular dynamics
simulations may give insight into the nature of peptide folding
on virus surfaces. However, retention of targeting function
can only be validated in a cell based assay.
Acknowledgments
We thank NanoVector, Inc. for funding that supported D. L.
References
1. Sherman, M. B., Guenther, R. H., Tama, F., 2. Speir, J. A., Munshi, S., Wang, G., Baker, T. S.,
Sit, T. L., Brooks, C. L., Mikhailov, A. M., et al. and Johnson, J. E. (1995) Structures of the
(2006) Removal of divalent cations induces native and swollen forms of cowpea chlorotic
structural transitions in Red clover necrotic mottle virus determined by X-ray crystallogra-
mosaic virus, revealing a potential mechanism phy and cryoelectron microscopy. Structure 3,
for RNA release. J. Virol. 80, 10395–10406. 63–78.
3. Lucas, R. W., Larson, S. B., and McPherson, A. for loading of polyacids. J. Gen. Virol. 87,
(2002) The crystallographic structure of 2749–2754.
brome mosaic virus. J. Mol. Biol. 317, 11. Chen, C., Kwak, E. S., Stein, B., Kao, C. C.,
95–108. and Dragnea, B. (2005) Packaging of gold
4. Lin, T. W., Chen, Z., Usha, R., Stauffacher, particles in viral capsids. J. Nanosci.
C. V., Dai, J. B., Schmidt, T., et al. (1999) Nanotechnol. 5, 2029–2033.
The refined crystal structure of cowpea mosaic 12. Loo, L., Guenther, R. H., Basnayake, V. R.,
virus at 2.8 angstrom resolution. Virology Lommel, S. A., and Franzen, S. (2006)
265, 20–34. Controlled encapsidation of gold nanoparti-
5. Hopper, P., Harrison, S. C., and Sauer. R. T. cles by a viral protein shell. J. Am. Chem. Soc.
(1984) Structure of tomato bushy stunt virus. 128, 4502–4503.
V. Coat protein sequence determination and 13. Dixit, S. K., Goicochea, N. L., Daniel, M. -C.,
its structural implications. J. Mol. Biol. 177, Murali, A., Bronstein, L., De, M., et al. (2006)
701–713. Quantum dot encapsulation in viral capsids.
6. Aniagyei, S. E., DuFort, C., Kao, C. C., and Nano Lett. 6, 1993–1999.
Dragnea, B. (2008) Self-assembly approaches 14. Shtykova, E.V., Huang, X., Gao, X., Dyke, J.
to nanomaterial encapsulation in viral protein C., Schmucker, A. L., Dragnea, B., et al.
cages. J. Mater. Chem. 18, 3763–3774. (2008) Hydrophilic monodisperse magnetic
7. Douglas, T. and Young, M. (1998) Host- nanoparticles protected by an amphiphilic
guest encapsulation of materials by assembled alternating copolymer. J. Phys. Chem. C 112,
virus protein cages. Nature 393, 152–155. 16809–16817.
8. Dragnea, B., Chen, C., Kwak, E. -S., Stein, 15. Loo, L., Guenther, R. H., Lommel, S. A., and
B., and Kao, C. C. (2003) Gold nanoparticles Franzen, S. (2008) Infusion of dye molecules
as spectroscopic enhancers for in vitro studies into Red clover necrotic mosaic virus. Chem.
on single viruses. J. Am. Chem. Soc. 125, Commun. (Camb) (1), 88–90.
6374–6375. 16. Strable, E. and Finn, M. G. (2009) Chemical
9. Loo, L., Guenther, R. H., Lommel, S. A., and Modification of Viruses and Virus-Like
Franzen, S. (2007) Encapsidation of nanopar- Particles. In: Viruses and Nanotechnology
ticles by red clover necrotic mosaic virus. J. Am. (Manchester, M. and Steinmetz, N. F., eds.)
Chem. Soc. 129, 11111–11117. pp. 1–21, Springer, Berlin, Heidelberg.
10. Ren, Y. P., Wong, S. M., and Lim, L.Y. (2006) 17. Hermanson, G. T. (2008) Bioconjugate
In vitro-reassembled plant virus-like particles Techniques. Academic Press, San Diego.
Chapter 15
Applications of Carbon Nanotubes in Biomedical Studies

Hongwei Liao, Bhavna Paratala, Balaji Sitharaman, and Yuhuang Wang
Abstract
Carbon nanotubes (CNTs) are novel, one-dimensional nanomaterials with many unique physical and
chemical properties that have been increasingly explored for biological and biomedical applications. In
this chapter, we briefly summarize the intrinsic properties of single-walled carbon nanotubes (SWNTs), a
special class of CNTs, and their corresponding applications in these fields. SWNTs have been utilized for
the ultrasensitive detection of biological species, providing a label-free approach. SWNT-Raman tags
have achieved detection sensitivity down to 1 fmol/L. SWNT-based drug delivery systems have shown
promising potential based on preliminary in vitro and in vivo studies. Also, the remarkable optical prop-
erties of SWNTs have made them promising candidates as contrast agents for imaging in cells and ani-
mals. Moreover, due to their excellent mechanical strength, SWNTs have been used to improve the
mechanical properties of solid polymeric nanocomposites and porous scaffolds. Sample preparation pro-
cedures for the use of SWNTs as fluorescent imaging labels and in biological composites will be
discussed.
Key words: Single-walled carbon nanotubes, Surface functionalization, Biomedical applications, Drug
delivery, Biomedical imaging, Nanocomposite
1. Introduction
Carbon nanotubes (CNTs) are seamless cylinders of graphene

sheets, exhibiting a wide variety of remarkable chemical and phys-
ical properties, which have drawn tremendous interest in the past
decade (1, 2). Depending on the number of graphene layers,
CNTs are classified as single-walled carbon nanotubes (SWNTs)
or multiwalled carbon nanotubes (MWNTs). Applications of
CNTs span many fields; they can be used as nanoelectronics (2),
field-effect emitters (3), and composite materials (4), for example.
In recent years, due to their interesting size, shape, structure, and
their unique physical properties, efforts have been devoted to
223
224 Liao et al.
exploring the potential biological applications of CNTs (5–9).

In this chapter, we will give a brief review of the applications of
SWNTs in biomedical studies, followed by protocols describing
the use of SWNTs in fluorescent imaging and biocomposites as
specific examples.
1.1. What Properties SWNTs are one-dimensional (1D) hollow nanomaterials with
Make SWNTs Useful diameters of 0.4–4 nm and lengths ranging from 50 nm up to
in Biological 1 cm. With all atoms on the surface, SWNTs have ultrahigh surface
Applications? area (1,300 m2/g), permitting the loading of multiple molecules
onto the nanotube sidewall. These functionalized nanotubes can
bend to interact with one cell at multiple binding sites, resulting
in improved binding affinity.
Due to quantum confinement of the electronic density of
states (DOS) along the circumference, SWNTs exhibit sharp
electronic DOS at the van Hove singularities, giving rise to fasci-
nating optical properties uniquely associated with each SWNT
structure (see Fig. 1) (10). Depending on the structure, SWNT
can be metallic or semiconducting. Semiconducting SWNTs
exhibit strong optical absorption and photoluminescence in the
NIR range with emission between 800 and 1,600 nm (11). Since
biological tissues are transparent within part of this range, SWNTs
are therefore suitable for use in biological imaging. SWNTs also
have several distinctive Raman scattering features including the
radial breathing mode (RBM) and tangential mode (G-band)
(15), which are sharp, strong peaks that can be easily distin-
guished from fluorescence background and are thus suitable for
Raman detection/imaging (16). Other examples of SWNT appli-
cations include photothermal therapy (12, 13) and photoacoustic
imaging (14).
SWNTs due to their excellent mechanical strength (Young’s
modulus ~1 TPa) are also being investigated for use as reinforc-
ing agents, to enhance the mechanical properties of solid poly-
meric nanocomposites and porous scaffolds for biomedical
applications. The properties of SWNT and the corresponding
biomedical applications are schematically shown in Fig. 2.
1.2. Examples Motivated by various chemical and physical properties of SWNTs,

of Biological many efforts have been devoted to apply SWNTs in biomedical
Applications applications. SWNT-based sensors have been developed to detect
biological species including proteins and DNA (6, 17, 18).
SWNTs can be utilized as optical tags or contrast agents in bio-
logical imaging techniques (9, 14, 16, 19). Properly functional-
ized SWNTs are able to deliver biological or molecular cargo into
cells (7, 8, 20–22). Recently, SWNTs have shown promise for
in vivo cancer treatment in a mouse model (23). In this section,
examples of applications of SWNTs in biomedical studies are
briefly reviewed.
a b
.8
Energy (eV)
Abs. (O.D.)
.6
E11 E22
.4
.2
DOS(E) (eV+cm−1) 400 600 800 1000 1200 1400 1600

Wavelength (nm)
c 20000
G-Band
15000
RBM
Intensity
10000
G-Band
5000
D-Band
0
0 500 1000 1500 2000 2500
Raman Shift (cm−1)
Fig. 1. Optical properties of SWNTs. (a) Schematic density of electronic states of a single SWNT structure. The sharp fea-
tures of the DOS are attributed to van Hove singularities. E11 and E22 are optical transitions correspond to photon absorp-
tion in the NIR and visible (vis) ranges, respectively. (b) Absorption spectrum of an aqueous solution of SWNTs. Peaks in the
spectrum are due to SWNTs with different structures. (c) Raman spectrum of SWNTs. The peaks at 200–300, ~1,340,
~1,590, and ~2,700 cm−1 are the radial breathing modes (RBM), D-band mode, G-band mode, and G¢-band mode, respec-
tively. (d) Contour plot of fluorescence intensity versus excitation and emission wavelengths for a sample of semiconducting
HiPCo SWNTs. SWNTs with different structures emit at different wavelengths under different excitations.
226 Liao et al.
Single Walled Carbon

Nanotube
1D hollow Strong resonant High optical High Young’s

Semiconducting Photo luminescence
structure Raman Scattering absorption Modulus
Drug Delivery
MRI Contrsat Agents
Raman Imaging
Raman Tags
Surface Ehnanced
Electrical Detection
Imaging and Detection

Photoluminescent
Photothermal Therapy
Photoacoustic Imaging
Composite
Fig. 2. SWNT properties and the corresponding biological applications.
1.2.1. Electrical Detection Field-effect transistors (FET) based on semiconducting SWNTs

have been utilized for biomolecule detection. SWNTs conju-
gated with biotin, Staphylococcal protein A, and U1A antigen
(6, 24) have been reported to impart specific binding of strepta-
vidin, immunoglobulin G, and the monoclonal mouse antibody
10E6, respectively. This technique has achieved in situ direct
detection of these analytes in the nmol/L range via electrical
read out. A variety of SWNT devices have been demonstrated for
selective detection of oxidase and dehydrogenase activity, as well as
for other biomolecules of interest, in a label-free fashion (25, 26).
1.2.2. Photoluminescent The interesting optical properties of semiconducting SWNTs

Detection on open another route for sensitive and selective detection of biomol-
Semiconducting SWNT ecules. Two mechanisms, charge transfer and fluorescence
quenching (27, 28), have been used for biomolecule detection
based on SWNT band-gap fluorescence. The band-gap fluores-
cence is sensitive to the local dielectric environment around the
SWNT and this property can be exploited in chemical sensing.
DNA conformational polymorphism, induced by divalent metal
cations that bind to DNA and stabilize the Z-form, has been
detected by monitoring the SWNT band-gap fluorescence red
shift (29). This strategy has also been utilized to detect the metal
ions. Strano and co-workers have developed an array of SWNT
sensors for detecting H2O2 molecules, generated upon growth
factor stimulation in living A431 human epidermal carcinoma
cells, which stochastically absorb and quench the SWNT fluores-
cence with spatial and temporal resolution (30). SWNT NIR
fluorescence does not photobleach, has negligible autofluores-
cence from other assay components in the NIR range, and
Applications of Carbon Nanotubes in Biomedical Studies 227
demonstrates a large Stokes shift compared to that of traditional

fluorophores. These properties make SWNTs a novel class
of fluorophores that allow the use of a range of excitation energies
and the real time tracking of biological processes. This chapter
highlights one method that can be used to synthesize highly fluo-
rescent SWNTs.
1.2.3. SWNT-Raman Tags SWNTs show intense Raman scattering cross-sections and high
scattering efficiencies (16). The Raman scattering spectra of
SWNTs are simple, with strong, well-defined Lorentzian peaks
which are easily distinguishable from noise (18). SWNT-Raman
tags do not photobleach even under high laser powers.
Surface-enhanced Raman spectroscopy (SERS) (31) can be
used to increase the intensity of Raman active molecules in prox-
imity to surface plasmons associated with gold, silver, and copper
nanostructures (32). By coupling the intense Raman scattering
efficiency of SWNTs with SERS substrates, the limit of detection
of traditional fluorescence assays can be extended from approxi-
mately 1 pmol/L (33) to the femtomolar level or below.
1.2.4. In Vitro Delivery Properly functionalized SWNTs are able to enter cells by endocy-
of Biomolecules tosis without obvious toxicity (7) and they are chemically stable
in biological environments. Owing to these properties, function-
alized SWNTs have been used to efficiently deliver various bio-
logical cargos such as drugs, proteins, and DNA/RNA into cells.
Once taken up by cells via endocytosis, functionalized SWNTs are
able to exit cells through exocytosis (28).
Drug molecules can be covalently or noncovalently conju-
gated to SWNTs for in vitro delivery. SWNTs covalently tethered
with the platinum (IV) complexes are taken into cancer cells. The
platinum (II) core complex is released in reducing pH environ-
ments, thus killing the cancer cells. When attached to SWNTs,
the cytotoxicity of the free platinum (IV) complex increases 100-
fold (34). Aromatic molecules can be noncovalently loaded onto
functionalized SWNTs via p–p stacking. Doxorubicin, a com-
monly used cancer chemotherapy drug, has been noncovalently
loaded in high amounts onto the surface of PEGylated SWNTs
(up to 4 g drug/1 g nanotube). The loading/binding is pH
dependent and favorable for drug release in tumor microenviron-
ments with acidic pH (35).
1.2.5. In Vivo Tumor In addition, SWNTs can be used for in vivo tumor targeting and
Targeting and Cancer cancer therapy. Dai and co-workers have conjugated paclitaxel
Therapy (PTX), a commonly used chemotherapy drug, to branched PEG
functionalized SWNTs via a cleavable ester bond (23). The
SWNT–PTX conjugate exhibits improved treatment efficacy over
the clinical Cremophor-based PTX formulation, Taxol®, in a 4T1
murine breast cancer model in mice.
228 Liao et al.
1.2.6. Biological Imaging The intrinsic optical properties of SWNTs make them useful as
Using Carbon Nanotubes optical probes. Owing to their unique 1D structure, SWNTs
exhibit strong resonance Raman scattering, high optical absorp-
tion, and photoluminescence in the NIR range. All of these prop-
erties can be utilized for in vitro and in vivo imaging in biological
systems. Individual semiconducting SWNTs have photolumines-
cence in the NIR range between 800 and 1,600 nm, depending
on their structure. This is useful for biological imaging, due to the
high optical transparency of biological tissue near 800–1,000 nm
and the inherently low autofluorescence from tissue in the NIR
range (36). Biological imaging using SWNTs also benefits from
the reduced background from autofluorescence. Using the intrinsic
NIR photoluminescence of SWNTs, Jin et al. can track endocyto-
sis and exocytosis of SWNTs in NIH-3T3 cells in real time (28).
SWNTs exhibit strong resonance Raman scattering that can
be easily distinguished from fluorescence background. Raman
microscopy has been utilized to image SWNTs in liver cells and
tissue slices, using either their RBM or G-band peaks (16, 37–39).
Also, Raman signals of SWNTs do not photobleach. They can be
used for long-term imaging and tracking (16, 39).
SWNTs have strong optical absorption in the visible and NIR
range which can be utilized in photoacoustic imaging.
Photoacoustic imaging has higher spatial resolution than tradi-
tional ultrasound and deeper tissue penetration than fluorescence
imaging (40). RGD-conjugated SWNTs have been used as the
contrast agent for photoacoustic molecular imaging of cancer in
living mice (14). The RGD-SWNT conjugate showed eight times
greater photoacoustic signal than nontargeted SWNTs.
Recently, SWNTs conjugated with Gd3+ have been used in
magnetic resonance imaging (MRI) (5). The aquated Gd3+ ion
clusters within ultrashort SWNTs were found to be superpara-
magnetic, with a MRI efficacy of 40 times greater than the Gd3+-
based contrast agents in current clinical use.
1.2.7. SWNT-Based The field of tissue engineering and regenerative medicine requires
Composite and Porous the development of biomaterials with superior mechanical and
Scaffolds for Tissue bioactive properties. Metallic- or ceramic-based materials used in
Engineering Applications the development of implants and devices can support significant
functional loads. However, their limitations include a weak tissue
interface, potential for corrosion and fatigue, and poor bioactivity.
The past decade has seen a significant increase in research on
polymers as materials to overcome some or all of the above limita-
tions. The mechanical properties of polymers can be improved by
modifying the processing conditions or composition and by
incorporating nanomaterials as reinforcing agents.
Recently, CNTs (SWNTs and MWNTs) have been demon-
strated to substantially improve the mechanical and structural
properties of polymer composites (4, 41–43). These porous
bioscaffolds provide structural support, guide cell growth, and

transport nutrients and waste products, essential for tissue regen-
eration. The presence of less than 0.5 wt% of SWNTs in a poly-
propylene fumarate (PPF, a linear biodegradable, biocompatible
polyester) matrix has been shown to significantly enhance the
compressive and flexural properties of the matrix up to threefold
against the polymer alone (43, 44). Recent studies also show that
the presence of small nontoxic amounts of SWNTs in polymeric
scaffolds enhances cellular adhesion and proliferation and may
induce bioactive properties into the scaffolds (41, 42, 44). The
present chapter provides protocols for preparing SWNT-reinforced
polymer nanocomposites and porous scaffolds. In addition, it
elaborates on the techniques and methodologies that are used
to characterize the structural and mechanical properties of rein-
forced polymer nanocomposites and porous scaffolds.
1.3. Five Degrees Similar to other nanomaterials, the intrinsic toxicity level of CNTs
of Control to Nontoxic depends on their physicochemical characteristics (e.g., size distri-
Nanotubes bution, metal catalyst residual, and surface modifications). There
are a wide variety of end products with different physicochemical
characteristics given the combination of different synthesis meth-
ods and purification processes that can be used. A SWNT sample
suitable for biomedical applications should be: (a) metal free, (b)
water soluble, (c) length controlled, (d) structurally sorted, and
(e) architecturally controlled (surface chemistry, defect density,
and 3D assembly). For most applications, (a) and (b) are essen-
tial; for the most demanding applications, one may need all five
degrees of control.
For instance, only semiconducting SWNTs are useful for
photoluminescent and electrical detection; the existence of metal
particles, metallic SWNTs, and other carbonaceous particles in
the raw SWNTs may significantly affect the sensitivity of using
nanotubes as fluorescent imaging tags and cause toxic side effects
to the biological system. Purification, separation, and isolation of
these semiconducting SWNTs from other by-products are there-
fore required for biomedical applications. Because pristine SWNTs
are hydrophobic, surface functionalization is required in order to
improve their aqueous solubility. Recent results have shown that
while nonfunctionalized, hydrophobic SWNTs can be toxic (45–48),
those with biocompatible coatings (9, 12, 34, 35, 49–54) are
harmless to cells in vitro and in vivo, at least to mice within tested
dose ranges (38, 39).
1.4. Materials Since their discovery, synthesis has been the main challenge in the
Chemistry Toward Five basic and applied research of CNTs. A variety of techniques have
Degrees of Control been developed and improved for SWNT production, and com-
mercial SWNT materials are now available. Typically, SWNTs are
1.4.1. Synthesis of SWNTs grown by heating a carbon-containing feedstock at elevated
230 Liao et al.
temperature in the presence of a transition metal catalyst such as

iron or cobalt. Carbon feedstocks used to date include bulk
graphite, hydrocarbons, and carbon monoxide. The most com-
mon processes are arc discharge (55), laser ablation (56), and
chemical vapor deposition (CVD) (57). Among the CVD meth-
ods, high-pressure CO (HiPco) developed at Rice University (57)
and CoMoCat (Southwest Nanotechnologies, Inc.) (58) are two
major commercial approaches used to produce large-scale, high-
quality SWNTs. The comparison of those methods is summarized
in Table 1.
1.4.2. Functionalization As-grown SWNTs are insoluble in organic solvents. For biomedical
of SWNTs applications, surface chemistry or functionalization is required to
solubilize SWNTs and to render biocompatibility and low toxicity.
Surface functionalization of SWNTs can be covalent or noncova-
lent. Covalent methods involve addition of functional groups to
SWNT sidewalls by chemical bonds. In noncovalent approaches,
the SWNT sidewalls are functionalized by, for example, aromatic
compounds, surfactants, and polymers, employing p–p stacking
or hydrophobic interactions. Because noncovalent modifications
of SWNTs can preserve their desired properties while imparting
high water solubilities, this approach to synthesizing aqueous
soluble nanotubes is widely used.
1.4.3. Noncovalent Noncovalent functionalization of SWNTs can be carried out by

Functionalization of SWNTs coating SWNTs with aromatic small molecules, biomacromole-
cules, polymers, and amphiphilic surfactant molecules. Since the
chemical structure of the p-network of the CNTs remains, the
physical properties of CNTs are essentially preserved.
Consequently, aqueous solutions of noncovalently functionalized
SWNTs are promising for multiple biomedical studies including
imaging.
The polyaromatic graphitic surface of SWNTs is accessible to
the binding of aromatic molecules, such as pyrene, porphyrin,
and their derivatives, via p–p stacking (24, 59, 60). Chen et al.
showed that proteins can be tethered on SWNTs functionalized
Table 1
Comparison of synthesis methods of SWNT
Methods Quantity Quality Yield
Arc discharge Grams Good Up to 30%

Laser ablation Grams Good Around 70%
Chemical vapor deposition Large scale Excellent Up to 95%
Source: http://en.wikipedia.org/wiki/Carbon_nanotube#Synthesis
with an amine-reactive pyrene derivative (24). Dai and co-workers

have shown that fluorescein (FITC)-terminated PEG chains are
able to solubilize SWNTs through the aromatic FITC domain
p–p stacked on the nanotube surface. The obtained SWNT con-
jugates have visible fluorescence which is useful for biological
detection and imaging (61).
Amphiphiles can also be used to solublize SWNTs in aqueous
solutions, with hydrophobic domains attached to the SWNT sur-
face via van der Waals forces and hydrophobic effects (62–64).
Cherukuri et al. used Pluronic tri-block polymer to solubilize
SWNTs for in vivo experiments (62). Surfactants, such as sodium
dodecyl sulfate (SDS), have also been used to suspend SWNTs in
water (11). PEGylated phospholipids (PL-PEG) have been devel-
oped to noncovalently functionalize SWNTs (12, 23, 35, 65).
The hydrocarbon chains of the lipid strongly anchor onto the
nanotube surface while the hydrophilic PEG chain imparts water
solubility and biocompatibility. Biological molecules can be con-
jugated onto PEGylated SWNTs by using a functional group
(e.g., amine) at the PEG terminal.
1.4.4. Covalent Various methods have been developed to covalently functionalize

Functionalization of SWNTs SWNTs. Oxidation is one of the most common (66). During the
oxidation process, carboxyl groups are formed at the ends of
tubes as well as on the sidewalls. After oxidation, sp2 carbon atoms
can be turned into sp3 which can be covalently conjugated with
amino acids (67).
Cycloaddition reactions are widely used to covalently func-
tionalize SWNTs. The cycloaddition reaction occurs on the
aromatic sidewalls, instead of the nanotube ends and defect sites as
with oxidation. A 1,3-dipolar cycloaddition reaction on SWNTs
developed by Prato et al. is now a commonly used reaction (68, 69).
An azomethine-ylide generated by condensation of an a-amino
acid and an aldehyde is added to the graphitic surface, forming a
pyrrolidine ring coupled to the SWNT sidewall. Functional
groups introduced via a modified a-amino acid can be used for
further conjugation of biological molecules (22, 70).
Billups and co-workers applied the Birch reduction (71) to
functionalize SWNTs (72, 73). The Billups reaction is particularly
effective as lithium and other alkali metals can intercalate nano-
tube ropes in liquid ammonia, permitting the exfoliated nano-
tubes to react homogeneously with alkyl or aryl radicals produced
from dissociation of corresponding halide precursors (72, 73).
Various functional groups, including carboxylic acids (74), can be
covalently added to SWNT sidewalls via this chemistry to afford
predominantly individual nanotubes in water.
After chemical reactions, the intrinsic physical properties of
CNTs such as photoluminescence and Raman scattering are likely
destroyed due to the disrupted nanotube structure. The intensities
232 Liao et al.
of Raman scattering and photoluminescence of SWNTs are

drastically decreased after covalent functionalization, reducing
their potential for their use in optical applications in biomedical
studies. Recently, the Wang group developed an approach to
selectively oxidize the outer wall of double-walled carbon nano-
tubes using a combination of oleum and nitric acid (75). This
double-wall chemistry enabled high water solubility through
carboxylic acid functional groups introduced to the outer wall,
while leaving the inner tube intact. This provides the opportunity
to covalently tether molecules of interest onto the outer wall of
the double-walled nanotube while using the optical properties
of the inner wall of the nanotube for detection and imaging.
2. Materials
2.1. Preparation 1. High-pressure CO converted (HiPco) SWNTs (see Note 1).

of Water-Soluble, 2. Sodium cholate.
Brightly Fluorescent
3. 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine–N-
SWNT Conjugates
[methoxy(polyethylene glycol) 5000] (DSPE-mPEG5k).
4. 3,500 Molecular weight cut-off (MWCO) membranes.
2.2. Preparation 1. PPF, prepared and purified as described previously (76), and
of SWNT-Reinforced propylene fumarate di-acrylate (PPF-DA).
Polymers and Porous 2. Purified HiPco SWNTs (iron content approximately 2%)
Scaffolds (77).
3. Benzoyl peroxide (BP), diethyl fumarate, and N,N-dimethyl-
p-toluidine (DMT) NaCl (300–500 mm crystal size) sieved
with USA Standard Testing Sieves.
2.3. Characterization 1. Gold wire.

2. Methylene chloride.
3. Methods
3.1. Preparation 1. Add 1 mg raw HiPco SWNTs and 40 mg sodium cholate to
of Water-Soluble, 4 mL of water.
Brightly Fluorescent 2. Bath sonicate the mixture for 1–6 h.
SWNT Conjugates
3. Ultracentrifuge the resulting black suspension at 300,000 × g for
(Adapted from Ref. 19) 1 h to remove large aggregates and bundled nanotubes, leaving
a dark supernatant of predominantly individual SWNTs.
4. Add 1 mg/mL DSPE–mPEG5k to the supernatant and sonicate

briefly (<1 min) to ensure that the DSPE–mPEG5k is fully
dissolved.
5. Dialyze the solution against water using a 3,500 MWCO
membrane over a period of 4–5 days with multiple water
changes per day. This process slowly removes the sodium
cholate from the solution, allowing the DSPE–mPEG5k to
coordinate to the surface of the nanotubes.
6. Following dialysis, ultracentrifuge the solution again at
300,000 × g for 1 h to remove any bundles that may have
formed during the exchange process.
3.2. Preparation 1. Disperse SWNTs in chloroform by high shear mixing for

of SWNT-Reinforced 5 min and then sonication for 15 min.
Polymers 2. Mix PPF and PPF-DA (cross-linking agent of PPF) in chloro-
form (1 g PPF and 2 g PPF-DA per 3 mL chloroform).
3. Add the SWNTs (from step 1) immediately to the PPF/
PPF-DA mixture to achieve SWNT concentrations ranging
between 0 and 2 wt% (see Note 2).
4. Sonicate this mixture for 15 min, then remove the chloro-
form by rotary evaporation and vacuum-drying to obtain the
uncross-linked, SWNT-reinforced polymer nanocomposites.
5. To achieve cross-linking, trigger the thermal polymerization
reaction by adding 1 wt% BP (a free radical initiator, 0.1 g/
mL dissolved in diethyl fumarate) and then 0.15 wt% DMT
(accelerator) to the SWNTs under vigorous stirring.
6. Centrifuge the specimens at 721 × g for 5 min to remove any
air bubbles.
7. Add the polymeric mixture into cylindrical glass vials. Vials
6.5 mm in diameter and 40 mm in length or 3 mm in diam-
eter and 150 mm in length work best for compressive and
flexural testing, respectively.
8. Cure them at 60°C for 24 h.
9. Recover the specimens by breaking the glass container.
Specimens can also be cut into appropriate shapes and lengths
with a diamond saw.
3.3. Preparation 1. Perform steps 1–4 in Subheading 3.2.

of SWNT-Reinforced 2. Mix the uncross-linked nanocomposites with 1 wt% free radical
Porous Scaffolds initiator (BP), then add the appropriate amount of NaCl to
(“Thermal Cross- achieve the desired porosity (see Note 3). NaCl is used as the
Linking Particulate water-soluble porogen (78).
Leaching”) 3. Cast the mixtures and thermally cross-link them at 100°C (see
Note 4) for 24 h in cylindrical Teflon molds (4 mm diameter
234 Liao et al.
and 8 mm height) or cylindrical glass molds (6.5 mm diameter

and 40 mm in length).
4. Soak the cross-linked samples in water (change the water
every 8 h) on a shaker table (80 rpm) at room temperature
for 3 days to leach out the NaCl porogen. Blot them with
absorbent paper and dry them under vacuum for 24 h.
3.4. Characterization 1. Sputter coat cross-sections of cut disks with gold using a sputter
of Structural coating system.
Properties 2. Examine the characteristics, such as the dispersion of SWNTs,
3.4.1. SWNT-Polymer the pore structure of the scaffold, the pore size, the morphology,
Interactions and Scaffold and the interconnectivity, by observing the samples above
Pore Structure Analysis under a field emission scanning electron microscope at an
Using SEM accelerating voltage of 15 kV.
3.4.2. Sol Fraction Analysis In order to assess the influence of the SWNTs on cross-linking
density in the PPF polymer matrix, sol fraction analysis can be
carried out on uncross-linked, SWNT-reinforced polymers.
1. Weigh 0.5 g of the sample (Wi, accuracy = 0.001 g) into a vial
with 20 mL of methylene chloride.
2. Seal the vial and place it on a shaker table (80 rpm) at room
temperature for 7 days.
3. Filter the solid sample with a weighed filter paper (Wp). Dry the
retained material on the filter paper at 60°C for 1 h and keep it at
room temperature for another 1 h. Then, weigh it again (Wp+s).
4. Calculate the sol fraction using the following equation for
each group (n ³ 5):
Wi − (Wp + s − Wp )
Sol fraction = × 100%. (1)
Wi
3.4.3. Porosity Micro-CT can be used to nondestructively and quantitatively

Measurements Using measure the three-dimensional (3D) porosity and porous inter-
Micro-CT and Mercury connectivity of SWNT-reinforced scaffolds.
Porosimetry
1. Scan 4 mm × 8 mm cylindrical samples of each scaffold type
3.4.3.1. Micro-CT Analysis with a micro-CT imaging system at 10 mm resolution using
a voltage of 40 kV and a current of 250 mA.
2. Conduct image reconstruction and analysis. First, reconstruct
the raw images of scaffolds to serial coronal-oriented tomo-
grams using a 3D cone beam reconstruction algorithm.
3. Perform a threshold analysis to determine the threshold value
for which grayscale tomograms of scaffolds are most accu-
rately represented by their binarized counterparts in terms
of porosity. Apply the optimal threshold value for all 3D
reconstructions and quantitative analysis.
4. Generate representative 3D reconstructions (top and side

views at a camera viewing angle of 10°) of porous scaffolds
based on binarized tomograms.
5. In order to eliminate potential edge effects, select a cylindrical
volume of interest (VOI) with a diameter of 3 mm and a
height of 6 mm in the center of a scaffold. Perform a shrink-
wrap process between two 3D measurements to shrink the
outside boundary of the VOI in a scaffold through any openings
whose size is equal to or larger than the threshold value.
6. Calculate scaffold porosity as:
Porosity = 100% − Vol% of binarized object. (2)
7. Interconnectivity is quantified as the fraction of the pore vol-
ume in a scaffold that is accessible from the outside through
openings of a certain minimum size (79). Calculate intercon-
nectivity as follows:
V − Vshrink - wrap
Interconnectivity = × 100%, (3)
V − Vm
where V is the total volume of the VOI, Vshrink-wrap is the VOI

volume after shrink-wrap processing, and Vm is the volume of
scaffold material.
3.4.3.2. Mercury Intrusion 1. Evacuate the sample chamber of a mercury intrusion poro-
Porosimetry simeter and fill it with mercury until an initial pressure of
~0.6 psi.
2. Weigh each 4 mm × 8 mm cylindrical samples for each scaf-
fold type (n ³ 3) and place it into the sample chamber.
3. Increase the chamber pressure at a rate of 0.01 psi/s to 50 psi,
and record the intruded volume of the mercury. The intruded
mercury volume per gram of the sample is assumed to be
equal to the pore volume (Vpore).
4. Calculate the porosity, e, using the formula:
Vpore
e= × 100%, (4)
Vpore + (1/ r)
where r is the density of the nanocomposites (see Note 3).
5. Make pore size measurements using the Washburn equation
4g cos q
D= , (5)
P
where D is the pore diameter, g is the surface tension of mer-
cury, q is the contact angle between mercury and the scaffold
material (see Note 5), and P is the pressure.
236 Liao et al.
3.5. Characterization 1. Test viscoelastic properties of the uncross-linked, SWNT-

of Bulk Properties polymer nanocomposites with a rheometer in oscillatory shear
mode at 25°C.
3.5.1. Viscoelastic Testing
Using Rheometer 2. Melt the nanocomposite samples (weight percentages up to
0.2 wt%) (see Note 6) and place them between the base plate
and cone geometry (60 mm diameter, 59 min cone angle,
and 26 mm truncation).
3. Perform the measurements as a function of the oscillatory
strain frequency (w) of 0.001–30 Hz using 0.01–0.1 strain
amplitude (see Note 7).
4. Record the complex viscosity magnitude, storage modulus,
and loss modulus as measures of viscoelastic properties of the
polymer nanocomposites.
3.5.2. Mechanical Testing: Mechanical properties under compression and flexion are tested
Compressive and Flexural at room temperature using a mechanical testing machine.
Properties (See Note 8)
3.5.2.1. Compressive 1. Compress the prepared specimens (4 mm × 8 mm) along their

Testing long axis until failure, and record the force and displacement
throughout the compression.
2. Generate stress–strain curves based on the initial specimen
dimensions (see Note 9).
3.5.2.2. Flexural Testing 1. The testing specimens are placed on a three-point bending
apparatus with two supports ~40 mm from each other.
2. Load a nose midway between the supports until the specimen
fails.
3. Record the force and displacement and convert to a stress–
strain curve (see Note 9).
4. Notes
1. As-grown HiPco materials contain both iron catalysts and

carbon nanoparticles. The majority of the iron can be removed
in the ultracentrifugation step. Alternatively, these impurities
can be removed by wet chemistry (80) or other purification
methods.
2. The same basic procedure can be followed to disperse SWNTs
into other hydrophobic polymers.
3. VNaCl
e = × 100%, (6)
VNaCl + Vnano
e r
WNaCl = × NaCl × WNano , (7)
1 − e rNano
where e is the apparent porosity (volume percent of porogen

in a scaffold), VNaCl and VNano are the volumes of NaCl and
the nanocomposite in a scaffold, WNaCl and WNano are the weights
of NaCl and the nanocomposite in a scaffold, and rNaCl is the
density of NaCl (2.17 g/mL). The density of the nanocom-
posite (rNano) is calculated by measuring the mass and volume
of five solid, cross-linked nanocomposite cylinders; found
here to be 1.25 g/mL (41).
4. The curing temperature of 100°C is applied to ensure com-
plete cross-linking of the scaffold materials (81).
5. 140° is the reported contact angle (q) between mercury and
this scaffold material (82).
6. Rheological measurements are performed only with compos-
ites up to 0.2 wt% CNT since previous studies (4, 83) already
confirm that solid-like behavior commences at very low
SWNT weight percentages (0.05–0.2 wt%).
7. The 0.01–0.1 strain amplitude is chosen as it allows for rheo-
logical measurement in the linear dynamic range (4). Use the
low end of the reported strain amplitude range for the nano-
composite melts; use the high end for the uncross-linked
polymer melts.
8. Compressive and flexural testing should follow the American
Society of Testing Materials (ASTM) Standard D695-02a and
ASTM Standard D790-03, respectively.
9. The slope of the initial linear portion of the curve gives the
compressive modulus and a line drawn parallel to the curve
defining the modulus, beginning at 1.0% strain (offset) gives
the offset compressive yield strength (the stress at which the
stress–strain curve intersects the line). The flexural modulus
can be calculated from the stress–strain curve using similar
methods. The compressive and flexural strength are defined as
the maximum stress carried by the specimen during compres-
sion or flexural testing, respectively
Acknowledgments
The authors would like to thank Jarrett Leeds for helping in the
preparation of soluble SWNT. This work was supported by
Office of the Vice President of Research at Stony Brook
University (SB).
238 Liao et al.
References
1. Liu, Z., Tabakman, S., Welsher, K., and Dai, multifunctional biological transporters and
H. (2009) Carbon nanotubes in biology and near-infrared agents for selective cancer cell
medicine: in vitro and in vivo detection, destruction. Proc. Natl. Acad. Sci. USA 102,
imaging and drug delivery. Nano Res. 2, 11600–11605.
85–120. 13. Chakravarty, P., Marches, R., Zimmerman, N.
2. Cao, Q. and Rogers, J. A. (2008) Random S., Swafford, A. D. E., Bajaj, P., Musselman, I.
networks and aligned arrays of single-walled H., et al. (2008) Thermal ablation of tumor
carbon nanotubes for electronic device appli- cells with anti body-functionalized single-
cations. Nano Res. 1, 259– 272. walled carbon nanotubes. Proc. Natl Acad.
3. Fan, S. S., Chapline, M. G., Franklin, N. R., Sci. USA 105, 8697–8702.
Tombler, T. W., Cassell, A. M., and Dai, H. J. 14. De la Zerda, A., Zavaleta, C., Keren, S.,
(1999) Self-oriented regular arrays of carbon Vaithilingam, S., Bodapati, S., Liu, Z., et al.
nanotubes and their field emission properties. (2008) Photoacoustic molecular imaging in
Science 283, 512–514. living mice utilizing targeted carbon nano-
4. Shi, X., Hudson, J., Spicer, P., Tour, J., tubes. Nat. Nanotechnol. 3, 557–562.
Krishnamoorti, R., and Mikos, A. (2005) 15. Rao, A. M., Richter, E., Bandow, S., Chase,
Rheological behaviour and mechanical char- B., Eklund, P. C., Williams, K. A., et al. (1997)
acterization of injectable poly (propylene Diameter-selective Raman scattering from
fumarate)/single-walled carbon nanotube vibrational modes in carbon nanotubes.
composites for bone tissue engineering. Science 275, 187–191.
Nanotechnology 16, 531. 16. Heller, D. A., Baik, S., Eurell, T. E., and
5. Sitharaman, B., Kissell, K. R., Hartman, K. Strano, M. S. (2005) Single walled carbon
B., Tran, L. A., Baikalov, A., Rusakova, I., nanotube spectroscopy in live cells: towards
et al. (2005) Superparamagnetic gadonano- long-term labels and optical sensors. Adv.
tubes are high-performance MRI contrast Mater. 17, 2793–2799.
agents. Chem. Commun. 31, 3915. 17. Tang, X. W., Bansaruntip, S., Nakayama, N.,
6. Chen, R. J., Bangsaruntip, S., Drouvalakis, K. Yenilmez, E., Chang, Y. L., and Wang, Q.
A., Kam, N. W. S., Shim, M., Li, Y. M., et al. (2006) Carbon nanotube DNA sensor and
(2003) Non-covalent functionalization of car- sensing mechanism. Nano Lett. 6,
bon nanotubes for highly specific electronic 1632–1636.
biosensors. Proc. Natl. Acad. Sci.USA 100, 18. Chen, Z., Tabakman, S. M., Goodwin, A. P.,
4984–4989. Kattah, M. G., Daranciang, D., Wang, X.,
7. Kam, N. W. S., Jessop, T. C., Wender, P. A., et al. (2008) Protein microarrays with carbon
and Dai, H. (2004) Nanotube molecular nanotubes as multi-color Raman labels. Nat.
transporters: internalization of carbon nano- Biotechnol. 26, 1285–1292.
tube-protein conjugates into mammalian cells. 19. Welsher, K., Liu, Z., Sherlock, S., Robinson,
J. Am. Chem. Soc. 126, 6850–6851. J., Chen, Z., Daranciang, D., et al. (2009) A
8. Bianco, A., Kostarelos, K., Partidos, C. D., route to brightly fluorescent carbon nano-
and Prato, M. (2005) Biomedical applications tubes for near-infrared imaging in mice. Nat.
of functionalised carbon nanotubes. Chem. Nanotechnol. 4, 773–780.
Commun. 571–577. 20. Singh, R., Pantarotto, D., McCarthy, D.,
9. Cherukuri, P., Bachilo, S. M., Litovsky, S. H., Chaloin, O., Hoebeke, J., Partidos, C. D.,
and Weisman, R. B. (2004) Near-infrared et al. (2005) Binding and condensation of
fluorescence microscopy of single-walled car- plasmid DNA onto functionalized carbon
bon nanotubes in phagocytic cells. J. Am. nanotubes: toward the construction of nano-
Chem. Soc. 126, 15638–15639. tube-based gene delivery vectors. J. Am.
10. Tans, S. J., Devoret, M. H., Dai, H., Thess, Chem. Soc. 127, 4388–4396.
A., Smalley, R. E., Geerligs, L. J., et al. (1997) 21. Kam, N. W. S. and Dai, H. (2005) Carbon
Individual single-wall carbon nanotubes as nanotubes as intracellular protein transport-
quantum wires. Nature 386, 474–477. ers: generality and biological functionality.
11. O’Connell, M. J., Bachilo, S. M., Huffman, J. Am. Chem. Soc. 127, 6021–6026.
C. B., Moore, V. C., Strano, M. S., Haroz, E. 22. Pantarotto, D., Briand, J. P., Prato, M., and
H., et al. (2002) Band gap fluorescence from Bianco, A. (2004) Translocation of bioactive
individual single-walled carbon nanotubes. peptides across cell membranes by carbon
Science 297, 593–596. nanotubes. Chem. Commun. 16–17.
12. Kam, N. W. S., O’Connell, M., Wisdom, J. 23. Liu, Z., Chen, K., Davis, C., Sherlock, S.,
A., and Dai, H. (2005) Carbon nanotubes as Cao, Q., Chen, X., et al. (2008) Drug delivery
with carbon nanotubes for in vivo cancer 36. Aubin, J. E. (1979) Autofluorescence of viable
treatment. Cancer Res. 68, 6652–6660. cultured mammalian cells. J. Histochem.
24. Chen, R. J., Zhang, Y. G., Wang, D. W., and Cytochem. 27, 36–43.
Dai, H. (2001) Non-covalent sidewall func- 37. Liu, Z., Winters, M., Holodniy, M., and Dai, H.
tionalization of single-walled carbon nano- (2007) siRNA delivery into human T cells and
tubes for protein immobilization. J. Am. primary cells with carbon nanotube transport-
Chem. Soc. 123, 3838–3839. ers. Angew. Chem. Int. Ed. 46, 2023–2027.
25. Kim, S. N., Rusling, J. F., and 38. Schipper, M. L., Nakayama-Ratchford, N.,
Papadimitrakopoulos, F. (2007) Carbon nan- Davis, C. R., Kam, N. W. S., Chu, P., Liu, Z.,
otubes for electronic and electrochemical et al. (2008) A pilot toxicology study of sin-
detection of biomolecules. Adv. Mater. 19, gle-walled carbon nanotubes in a small sample
3214–3228. of mice. Nat. Nanotechnol. 3, 216–221.
26. Wang, J. (2005) Carbon-nanotube-based 39. Liu, Z., Davis, C., Cai, W., He, L., Chen, X.,
electrochemical biosensors: a review. and Dai, H. (2008) Circulation and long-
Electroanalysis 17, 7–14. term fate of functionalized, biocompatible
27. Barone, P. W., Parker, R. S., and Strano, M. S. single-walled carbon nanotubes in mice
(2005) In vivo fluorescence detection of glu- probed by Raman spectroscopy. Proc. Natl
cose using a single-walled carbon nanotube Acad. Sci. USA 105, 1410–1415.
optical sensor: design, fluorophore properties, 40. Xu, M. H. and Wang, L. H. V. (2006)
advantages, and disadvantages. Anal. Chem. Photoacoustic imaging in biomedicine. Rev.
77, 7556–7562. Sci. Instrum. 77, 041101.
28. Barone, P. W., Baik, S., Heller, D. A., and 41. Shi, X., Sitharaman, B., Pham, Q., Liang, F.,
Strano, M. S. (2004) Near-infrared optical Wu, K., Billups, W. E., et al. (2007) Fabrication
sensors based on single-walled carbon nano- of porous ultra-short single-walled carbon nan-
tubes. Nat. Mater. 4, 86–92. otube nanocomposite scaffolds for bone tissue
29. Heller, D., Jeng, E., Yeung, T., Martinez, B., engineering. Biomaterials 28, 4078–4090.
Moll, A., Gastala, J., et al. (2006) Optical 42. Shi, X., Sitharaman, B., Pham, Q., Spicer, P.,
detection of DNA conformational polymor- Hudson, J., Wilson, L., et al. (2008) In vitro
phism on single-walled carbon nanotubes. cytotoxicity of single-walled carbon nano-
Science 311, 508–511. tube/biodegradable polymer nanocompos-
30. Jin, H., Heller, D., Kalbacova, M., Kim, J., ites. J. Biomed. Mater. Res. A 86, 813–823.
Zhang, J., Boghossian, A., et al. (2010) 43. Sitharaman, B., Shi, X., Tran, L., Spicer, P.,
Detection of single-molecule H2O2 signaling Rusakova, I., Wilson, L., et al. (2007)
from epidermal growth factor receptor using Injectable in situ cross-linkable nanocompos-
fluorescent single-walled carbon nanotubes. ites of biodegradable polymers and carbon
Nat. Nanotech. 5, 302–309. nanostructures for bone tissue engineering.
31. Nie, S. and Emory, S. R. (1997) Probing sin- J. Biomater. Sci. Polymer Ed. 18, 655–671.
gle molecules and single nanoparticles by 44. Sitharaman, B., Shi, X., Walboomers, X. F.,
surface-enhanced Raman scattering. Science Liao, H., Cuijpers, V., Wilson, L. J., et al.
275, 1102–1106. (2008) In vivo biocompatibility of ultra-short
32. Jeanmaire, D. L. and Van Duyne, R. P. (1977) single-walled carbon nanotube/biodegrad-
Surface Raman spectroelectrochemistry part 1. able polymer nanocomposites for bone tissue
Heterocyclic, aromatic, and aliphatic amines engineering. Bone 43, 362–370.
adsorbed on anodized silver electrode. J. Elec- 45. Cui, D. X., Tian, F. R., Ozkan, C. S., Wang,
troanal. Chem. 84, 1–20. M., and Gao, H. J. (2005) Effect of single
33. Espina, V., Woodhouse, E. C., Wulfkuhle, J., wall carbon nanotubes on human HEK293
Asmussen, H. D., Petricoin, E. F., and Liotta, cells. Toxicol. Lett. 155, 73–85.
L. A. (2004) Protein microarray detection 46. Lam, C. W., James, J. T., McCluskey, R., and
strategies: focus on direct detection technolo- Hunter, R. L. (2004) Pulmonary toxicity of
gies. J. Immunol. Methods 290, 121–133. single-wall carbon nanotubes in mice 7 and
34. Feazell, R. P., Nakayama-Ratchford, N., Dai, 90 days after intratracheal instillation. Toxicol.
H., and Lippard, S. J. (2007) Soluble single- Lett. 77, 126–134.
walled carbon nanotubes as longboat deliv- 47. Warheit, D. B., Laurence, B. R., Reed, K. L.,
ery systems for platinum(IV) anticancer drug Roach, D. H., Reynolds, G. A. M., and Webb,
design. J. Am. Chem. Soc. 129, 8438–8349. T. R. (2004) Comparative pulmonary toxicity
35. Liu, Z., Sun, X., Nakayama, N., and Dai, H. assessment of single-wall carbon nanotubes in
(2007) Supramolecular chemistry on water- rats. Toxicol. Lett. 77, 117–125.
soluble carbon nanotubes for drug loading 48. Poland, C. A., Duffin, R., Kinloch, I.,
and delivery. ACS Nano 1, 50–56. Maynard, A., Wallace, W. A. H., Seaton, A.,
240 Liao et al.
et al. (2008) Carbon nanotubes introduced longlived intracomplex charge separation.
into the abdominal cavity of mice show asbes- Adv. Mater. 17, 871–875.
tos-like pathogenicity in a pilot study. Nat. 60. Chen, J., Liu, H. Y., Weimer, W. A., Halls, M.
Nanotechnol. 3, 423–428. D., Waldeck, D. H., and Walker, G. C. (2002)
49. Wu, W., Wieckowski, S., Pastorin, G., Non-covalent engineering of carbon nanotube
Benincasa, M., Klumpp, C., Briand, J. P., surfaces by rigid, functional conjugated poly-
et al. (2005) Targeted delivery of amphoteri- mers. J. Am. Chem. Soc. 124, 9034–9035.
cin B to cells by using functionalized carbon 61. Nakayama-Ratchford, N., Bangsaruntip, S.,
nanotubes. Angew. Chem. Int. Ed. 44, Sun, X. M., Welsher, K., and Dai, H. (2007)
6358–6362. Non-covalent functionalization of carbon
50. Dumortier, H., Lacotte, S., Pastorin, G., nanotubes by fluorescein-polyethylene glycol:
Marega, R., Wu, W., Bonifazi, D., et al. (2006) supramolecular conjugates with pH-depen-
Functionalized carbon nanotubes are noncyto- dent absorbance and fluorescence. J. Am.
toxic and preserve the functionality of primary Chem. Soc. 129, 2448–2449.
immune cells. Nano Lett. 6, 1522–1528. 62. Cherukuri, P., Gannon, C. J., Leeuw, T. K.,
51. Chen, X., Lee, G. S., Zettl, A., and Bertozzi, Schmidt, H. K., Smalley, R. E., Curley, S. A.,
C. R. (2004) Biomimetic engineering of car- et al. (2006) Mammalian pharmacokinetics of
bon nanotubes by using cell surface mucin carbon nanotubes using intrinsic near-infrared
mimics. Angew. Chem. Int. Ed. 43, fluorescence. Proc. Natl Acad. Sci. USA 103,
6111–6116. 18882–18886.
52. Chen, X., Tam, U. C., Czlapinski, J. L., Lee, 63. Richard, C., Balavoine, F., Schultz, P.,
G. S., Rabuka, D., Zettl, A., et al. (2006) Ebbesen, T. W., and Mioskowski, C. (2003)
Interfacing carbon nanotubes with living cells. Supramolecular self-assembly of lipid deriva-
J. Am. Chem. Soc. 128, 6292–6293. tives on carbon nanotubes. Science 300,
53. Chin, S. F., Baughman, R. H., Dalton, A. B., 775–778.
Dieckmann, G. R., Draper, R. K., Mikoryak, 64. Wang, H., Zhou, W., Ho, D. L., Winey, K. I.,
C., et al. (2007) Amphiphilic helical peptide Fischer, J. E., Glinka, C. J., et al. (2004)
enhances the uptake of single-walled carbon Dispersing single-walled carbon nanotubes
nanotubes by living cells. Exp. Biol. Med. 232, with surfactants: a small angle neutron scat-
1236–1244. tering study. Nano Lett. 4, 1789–1793.
54. Yehia, H. N., Draper, R. K., Mikoryak, C., 65. Liu, Z., Cai, W. B., He, L. N., Nakayama, N.,
Walker, E. K., Bajaj, P., Musselman, I. H., et al. Chen, K., Sun, X. M., et al. (2007) In vivo
(2007) Single-walled carbon nanotube interac- biodistribution and highly efficient tumour
tions with HeLa cells. J. Nanobiotech. 5, 8. targeting of carbon nanotubes in mice. Nat.
55. Journet, C., Maser, W. K., Bernier, P., Loiseau, Nanotechnol. 2, 4–52.
A., Delachapelle, M. L., Lefrant, S., et al. 66. Niyogi, S., Hamon, M. A., Hu, H., Zhao, B.,
(1997) Large-scale production of single Bhowmik, P., Sen, R., et al. (2002) Chemistry
walled carbon nanotubes by the electric-arc of single-walled carbon nanotubes. Acc. Chem.
technique. Nature 388, 756–758. Res. 35, 1105–1113.
56. Thess, A., Lee, R., Nikolaev, P., Dai, H., Petit, 67. Zeng, L., Alemany, L. B., Edwards, C. L., and
P., Robert, J., et al. (1996) Crystalline ropes Barron, A. R. (2008) Demonstration of cova-
of metallic carbon nanotubes. Science 273, lent sidewall functionalization of single wall car-
483–487. bon nanotubes by NMR spectroscopy: side
57. Nikolaev, P., Bronikowshi, M. J., Bradley, R. chain length dependence on the observation of
K., Rohmund, F., Colbert, D. T., Smith, K. the sidewall sp3 carbons. Nano. Res. 1, 72–88.
A., et al. (1999) Gas-phase catalytic growth of 68. Georgakilas, V., Kordatos, K., Prato, M.,
single-walled carbon nanotubes from carbon Guldi, D. M., Holzinger, M., and Hirsch, A.
monoxide. Chem. Phys. Lett. 313, 91–97. (2002) Organic functionalization of carbon
58. Resasco, D. E., Alvarez, W. E., Pompeo, F., nanotubes. J. Am. Chem. Soc. 124, 760–761.
Balzano, L., Herrera, J. E., Bitiyanan, B., 69. Tagmatarchis, N. and Prato, M. (2004)
et al. (2002) A scalable process for production Functionalization of carbon nanotubes via
of single-walled carbon nanotubes (SWNT) 1,3-dipolar cycloadditions. J. Mater. Chem.
by catalytic disproportionation of CO on a 14, 437–439.
solid catalyst. J. Nanopart. Res. 4, 131–136. 70. Pastorin, G., Wu, W., Wieckowski, S., Briand,
59. Guldi, D. M., Taieb, H., Rahman, G. M. A., J. P., Kostarelos, K., Prato, M., et al. (2006)
Tagmatarchis, N., and Prato, M. (2005) Double functionalisation of carbon nanotubes
Novel photoactive single-walled carbon nano- for multimodal drug delivery. Chem. Commun.
tube-porphyrin polymer wraps: efficient and 1182–1184.
71. Birch, A. J. and Smith, H. (1958) Reduction from the gas-phase decomposition of CO
by metal–amine solutions: applications in syn- (HiPco process). J. Phys. Chem. B 105,
thesis and determination of structure. Q. Rev. 8297–8301.
Chem. Soc. 12, 17–33. 78. Fisher, J. P., Holland, T., Dean, D., Engel, P.,
72. Liang, F., Sadana, A. K., Peera, A., and Mikos, A. (2001) Synthesis and proper-
Chattopadhyay, J., Gu, Z., Hauge, R. H., ties of photocross-linked poly (propylene
et al. (2004) A convenient route to function- fumarate) scaffolds. J. Biomater. Sci. Poly. Ed.
alized carbon nanotubes. Nano Lett. 4, 12, 673–687.
1257–1260. 79. Moore, M., Jabbari, E., Ritman, E., Lu, L.,
73. Liang, F., Alemany, L. B., Beach, J. M., and Currier, B., Windebank, A., et al. (2004)
Billups, W. E. (2005) Structure analyses of Quantitative analysis of interconnectivity of
dodecylated single-walled carbon nanotubes. porous biodegradable scaffolds with micro-
J. Am. Chem. Soc. 127, 13941–13948. computed tomography. J. Biomed. Mater. Res.
74. Deng, S. L., Brozena, A. H., Zhang, Y., Piao, 71, 258–267.
Y.-M., Wang, Y. H. (2011) “Diameter- 80. Wang, Y., Shan, H., Hauge, R., Pasquali, M.,
Dependent, Progressive Alkylcarboxylation and Smalley, R. E. (2007) A highly selective,
of Single-Walled Carbon Nanotubes.” Chem. one-pot, green chemistry for carbon nanotube
Comm. 47, 758–760. purification. J. Phys. Chem. B 111, 1249–1252.
75. Brozena, A., Moskowitz, J., Shao, B., Deng, 81. Timmer, M., Carter, C., Ambrose, C., and
S., Liao, H., Gaskell, K., et al. (2010) Outer Mikos, A. (2003) Fabrication of poly (propyl-
wall selectively oxidized, water-soluble dou- ene fumarate)-based orthopaedic implants by
ble-walled carbon nanotubes. J. Am. Chem. photo-crosslinking through transparent sili-
Soc. 132, 3932–3938. cone molds. Biomaterials 24, 4707–4714.
76. Shung, A., Timmer, M., Jo, S., Engel, P., and 82. Lowell, S., Shields, J., Thomas, M., and
Mikos, A. (2002) Kinetics of poly (propylene Thommes, M. (2004) Characterization of
fumarate) synthesis by step polymerization of porous solids and powders: surface area, pore
diethyl fumarate and propylene glycol using size, and density. Springer, The Netherlands.
zinc chloride as a catalyst. J. Biomater. Sci. 83. Shi, X., Hudson, J., Spicer, P., Tour, J.,
Polym. Ed. 13, 95–108. Krishnamoorti, R., and Mikos, A. (2006)
77. Chiang, I., Brinson, B., Huang, A., Willis, P., Injectable nanocomposites of single-walled
Bronikowski, M., Margrave, J., et al. (2001) carbon nanotubes and biodegradable poly-
Purification and characterization of single- mers for bone tissue engineering. Biomacro
wall carbon nanotubes (SWNTs) obtained molecules 7, 2237–2242.
Chapter 16
Electrospun Nanofibrous Scaffolds for Engineering

Soft Connective Tissues
Roshan James, Udaya S. Toti, Cato T. Laurencin,
and Sangamesh G. Kumbar
Abstract
Tissue-engineered medical implants, such as polymeric nanofiber scaffolds, are potential alternatives to
autografts and allografts, which are short in supply and carry risks of disease transmission. These scaffolds have
been used to engineer various soft connective tissues such as skin, ligament, muscle, and tendon, as well as
vascular and neural tissue. Bioactive versions of these materials have been produced by encapsulating mole-
cules such as drugs and growth factors during fabrication. The fibers comprising these scaffolds can be designed
to match the structure of the native extracellular matrix (ECM) closely by mimicking the dimensions of the
collagen fiber bundles evident in soft connective tissues. These nanostructured implants show improved bio-
logical performance over the bulk materials in aspects of cellular infiltration and in vivo integration, and the
topography of such scaffolds has been shown to dictate cellular attachment, migration, proliferation, and dif-
ferentiation, which are critical steps in engineering complex functional tissues and crucial to improved biocom-
patibility and functional performance. Nanofiber matrices can be fabricated using a variety of techniques,
including drawing, molecular self-assembly, freeze-drying, phase separation, and electrospinning. Among
these processes, electrospinning has emerged as a simple, elegant, scalable, continuous, and reproducible
technique to produce polymeric nanofiber matrices from solutions and their melts. We have shown the ability
of this technique to be used to fabricate matrices composed of fibers from a few hundred nanometers to several
microns in diameter by simply altering the polymer solution concentration. This chapter will discuss the use of
the electrospinning technique in the fabrication of ECM-mimicking scaffolds. Furthermore, selected scaffolds
will be seeded with primary adipose-derived stromal cells, imaged using scanning electron microscopy and
confocal microscopy, and evaluated in terms of their capacity toward supporting cellular proliferation over time.
Key words: Nanofiber scaffolds, Electrospinning, Tissue engineering, Soft tissue, Skin, Tendon,
Extracellular matrix, Biodegradable scaffold, Cell behavior, Stem cells
1. Introduction
The increasing demand for biologically compatible donor tissue

and organ transplants (allografts) far outstrips the availability,
leading to an acute shortage. The available allografts have the
243
244 James et al.
potential to elicit an immune response and carry risk of disease

transmission. Autografts from an individual are associated with
issues such as donor site morbidity and very limited availability.
Tissue-engineered implants, such as biodegradable three-dimensional
(3D) porous scaffolds, have emerged as a viable alternative to these
materials to repair/regenerate damaged tissues and restore func-
tionality. Isolated cells alone cannot reassemble into complex 3D
functional tissues; however, these materials can be used to fill the
tissue void and provide anchorage for cells to attach, infiltrate,
populate, and differentiate to create functional tissue (1, 2). These
scaffolds of both natural and/or synthetic origin are being designed
to mimic the structure and functions of the native tissue (i.e.,
extracellular matrix (ECM)) morphologically, mechanically, and
dimensionally (3, 4). By virtue of their small fiber size, these
materials show exceptionally high surface area and porosity and
superior mechanical and degradation properties compared to
micro-sized porous materials, and as a result have shown improved
tissue regeneration capabilities (5). Additionally, biodegradable
nanofiber scaffolds can be used to deliver drugs, proteins, and
growth factors locally, which can further accelerate or modulate
the in vivo response (6, 7). For these reasons, nanofiber scaffolds
are a popular choice to repair and regenerate various soft tissues
such as skin, blood vessel, nerve, tendon, and cartilage (5, 8–10).
Nanofiber matrices can be fabricated using a variety of tech-
niques including electrospinning (7), molecular self-assembly
(11), phase separation (12), and drawing. The choice of the fab-
rication method is largely based on the properties of the chosen
matrix material. For instance, molecular self-assembly is a pre-
ferred technique to produce peptide nanofiber scaffolds (13),
while temperature-induced phase separation is the popular choice
to fabricate highly crystalline poly(l-lactic acid) (PLLA) nanofi-
ber scaffolds. All these processes have advantages and disadvan-
tages in terms of their fabrication, controllability, reproducibility,
and desired end application (Table 1). Drawing is a simple pro-
cess where a micropipette, a few micrometers in diameter, is
dipped into a polymer liquid and withdrawn at a fixed speed to
produce fibers. This labor-intensive process can produce only
fibers with diameters in the micrometer size regime, and an
additional step such as weaving is needed to produce scaffolds
for tissue engineering applications. This process is also limited
by the cohesive properties of the material which must be able
to support the stress produced under deformation/pulling. The
phase separation technique utilizes the physical incompatibility of
two materials and their tendency to separate into two phases to
fabricate nanofiber scaffolds. In brief, for example, flash-frozen
highly crystalline PLLA solution in dimethylformamide will
undergo gelation and is washed with water followed by freeze-
drying to remove the solvent phase to obtain highly porous
Electrospun Nanofibrous Scaffolds for Engineering Soft Connective Tissues 245
Table 1
Comparison of nanofiber fabrication techniques that can be used to synthesize
tissue-engineered implants
Fabrication
technique Advantages Disadvantages
Drawing • Simple equipment • Discontinuous process

• Not scalable
• No control on fiber
dimensions
Temperature-induced • Simple equipment • Limited to specific polymers
phase separation • Convenient fabrication process • Not scalable
• Mechanical properties of the fiber • No control on fiber
matrices can be varied by changing dimensions
polymer composition
Molecular • Only smaller nanofibers of few nanometer • Complex process involving
self-assembly in diameter and few microns in length can intermolecular forces
be fabricated • Not scalable
• Complex functional structures • Weak nanofiber strength
Electrospinning • Simple instrument • Jet instability
• Continuous process • Toxic solvents
• Cost effective compared to other existing • Packaging, shipping, and
methods handling
• Scalable
• Ability to fabricate fiber diameters few
nanometer to several microns
• Aligned and random-oriented fibers
nanofiber scaffolds (14). Though the technique is simple and

requires minimum instrumentation, it can be used only for certain
specific polymer–solvent combinations. Nanofiber scaffolds fabri-
cated from phase separation processes are highly porous and lack
the mechanical properties suitable for load-bearing applications.
Molecular self-assembly is based on spontaneous organization of
molecules and components into patterns or structure without
human intervention. This technique involves building nanoscale
fibers using small molecules which interact by intermolecular
forces such as van der Waals forces, hydrogen bonds, and electro-
static forces. The system is highly flexible with the possibility of
self-assembling innumerable structural shapes by modifying the
structure of the molecule. Self-assembly is ideally suited to design
short structures composed of ECM peptide or bioactive groups.
Electrospinning, or the electrostatic spinning process, is a ver-
satile platform that can be used to fabricate nanofiber scaffolds
from polymer solutions. Polymers of both natural and synthetic
origin as well as their blends have been fabricated into nanofiber
246 James et al.
scaffolds for a variety of biomedical applications (6, 15, 16). This

process requires relatively simple instrumentation and is a continu-
ous, scalable, and highly reproducible technique. Electrospinning
is mediated by the application of high electric potential to a
polymer solution. A typical electrospinning apparatus has three
essential components including a power source, a polymer solution
delivery system, and a collector (see Fig. 1). In brief, a program-
mable syringe pump delivers polymer solution at the desired rate
a Conducting
stationary target
Insulated grounded
(+ve voltage) target
Syringe
Syringe Pump
Polymer Jet
Power Source
b
Servo motor
Controller
Conducting
rotating target
Insulated grounded
(+ve voltage) target
Syringe
Syringe Pump Conducting

Polymer Jet
pointed target
Pivot
Fig. 1. Schematics of the electrospinning process using (a) a stationary target such as aluminum foil and (b) moving target
such as a rotating mandrel. A high-voltage power source is used to apply an electric potential of a few kV to a pendant
polymer droplet at the end of a blunt needle. The polymer solution is pumped out of the syringe at a controlled flow rate.
The charged polymer solution undergoes a series of bending and stretching instabilities across the air-gap distance, moving
toward the grounded target. The solvent rapidly evaporates and ultra-thin polymeric fibers are deposited.
and, when the applied electric potential exceeds the forces of

surface tension acting in the opposite direction, a thin polymer jet
ejects from the needle tip of the syringe and travels toward the
grounded target. During this journey, the jet undergoes a series of
electrically driven bending and stretching instabilities which result
in looping and spiraling motions. The jet elongates and stretches
to minimize this instability due to repulsive electrostatic forces,
and ultra thin fibers are deposited on the target. Various process
parameters, such as needle diameter, solution flow rate, applied
electric potential, and working distance (distance between needle
and target), and system parameters, including polymer molecular
weight and solution viscosity, surface tension, and conductivity,
affect the electrospinning and hence need to be optimized to pro-
duce continuous fibers of desired morphology and mechanical
properties. The type of collector target also plays an important
role in fiber orientation. For instance, the use of a stationary target
results in random nanofiber deposition, while the use of a rotating
mandrel-like target results in oriented nanofiber deposition (5).
It is possible to achieve any nanofiber orientation by manipulating
rotation speed and the local electrical field at the target. Several
efforts are also being made to align nanofibers by using special
collector configurations where fibers experience a local electrical
field that forces them to align.
This chapter will only emphasize electrospinning of nanofiber
matrices and their related soft tissue regeneration applications in
skin and tendons. In this chapter, electrospun nanofiber morphology,
pore structure, and diameters are analyzed by scanning electron
microscopy (SEM). SEM micrographs are also used to study cell
morphology, infiltration, and behavior following in vitro and
in vivo experimentation. Cell proliferation and differentiation on
nanofiber scaffolds are determined using a calorimetric assay to
quantify cellular metabolic activity; a standard curve correlating
absorbance of cell metabolic activity with cell number was used to
calculate cell number. Confocal microscopy analysis was used to
analyze fluorescently stained cells and provides valuable informa-
tion on cell infiltration and cell viability. Detection of live and
dead cells on the nanofiber scaffolds may not be possible using
light microscopy due to scaffold opacity.
2. Materials
2.1. Electrospinning 1. Organic solvents: Tetrahydrofuran (THF) and N,N-

Polymer Sheets dimethylformamide (DMF).
2. Synthetic polymer: poly(lactic-co-glycolic acid) 65:35
(PLAGA 65:35) (SurModics Pharmaceuticals, Birmingham,
AL). Store at −20°C (see Note 1).
248 James et al.
3. Vial: Glass vial (28 × 95 mm) with screw thread.

4. Paraffin film.
5. Shaker: Vortex mixer.
6. Syringe: 10-mL Luer-Lok tip.
7. Dispensing needle: 18-gauge, 1.0″ length, blunt-end, stainless
needle with threaded cap.
8. DC power: HV power supply (Gamma High Voltage Research,
Ormond Beach, FL).
9. Pump: Aladdin-1000 syringe.
10. Lab jack.
11. Grounded target: heavy duty aluminum foil.
12. Fine tip tweezers.
2.2. Scanning Electron 1. Double-sided carbon tape (8 mm × 20 mm).

Microscopy 2. Aluminum SEM specimen mount stubs.
3. Sharp blade or scalpel.
4. Sputter coater with gold foil.
2.3. Absorbance Assay 1. Ethyl alcohol 200 proof. Working solution is prepared by
for Proliferation Using diluting to 70% ethyl alcohol using sterile double-distilled
Primary Adipose- deionized water.
Derived Stromal Cells 2. UV light source (cell culture hood).
3. Sterile gauze.
4. Sterile fine tip tweezers.
5. Dulbecco’s phosphate-buffered saline 1× (PBS).
6. Dulbecco’s modified Eagle’s medium (DMEM) low glucose 1×.
7. Fetal bovine serum (FBS).
8. Penicillin streptomycin (P/S) having 10,000 units/mL penicillin
and 10,000 mg/mL streptomycin.
9. 0.5% Tryspin–EDTA (10×).
10. Sterile T-75 cell culture flasks.
11. Aseptic cell culture hood.
12. Aseptic cell culture incubator.
13. Sterile non-tissue culture-treated plate, 24-well (Non-TCP).
14. CellTiter 96 AQueous One Solution Cell Proliferation Assay
(Promega, Madison, WI).
15. 10% Sodium dodecyl sulfate (SDS).
16. 96-Well cell culture plate.
2.4. Confocal 1. Dulbecco’s PBS 1×.

Microscopy of Live 2. Live/Dead Viability/Cytotoxicity kit for mammalian cells
and Dead Cells (Molecular Probes, Eugene, OR).
3. Fine tip tweezers.

4. Lab-Tek two-well glass chamber slide (Nalge Nunc,
Naperville, IL).
3. Methods
Electrospinning system and process parameters vary with the use

of different polymers, which have different physical and chemical
properties. Solution viscosity is one property that we have found
to affect the morphology of the nanofiber scaffolds greatly. For
instance, a lower molecular weight copolymer solution (low viscosity)
will result in electrospray (i.e., bead formation resulting from
breakage of the polymer jet as it is deposited on the target), while a
higher molecular weight copolymer solution will result in electro-
spun fibers (17). Furthermore, as the viscosity increases, the fiber
diameter increases. This methodology exhibits very high reliability
and reproducibility as evidenced by the consistent fiber diameter
distribution, porosity, pore size, and degradation properties.
Herein, electrospun scaffold morphology (e.g., bead formation
and fiber diameter) is evaluated as a function of varying process
parameters. Fiber diameter is quantified using imaging software.
Cell proliferation response to nanofiber scaffolds of various
fiber diameters is determined by evaluating cellular metabolic
activity or DNA content. A cell proliferation assay that can be
used to quantify cell number as a function of metabolic activity is
discussed. This assay gives results that are indicative of the inhibi-
tory or stimulatory effect on cell growth over the time course of
cell culture. Furthermore, it is reported in the literature that cells
exhibit recognition of the nanofiber dimensions and orientation
(18). Fibroblastic cells are known to exhibit a round or a flat well-
spread morphology depending on the surface they are seeded
onto (19). We use confocal microscopy to determine the localiza-
tion of live and dead cells on the scaffold and the shape of the cells
present on the nanofibers.
3.1. Electrospun 1. Prepare a polymeric solution of PLAGA 65:35, by weighing

PLAGA Sheets accurately 2.2 g of PLAGA 65:35 polymer pellets in a glass
vial and then adding 10 mL of an organic solvent composed
of 3:1 THF:DMF. Tightly cap the vial immediately and wrap
in paraffin film. Shake the vial containing the polymer and
organic solvent vigorously overnight at room temperature to
dissolve the polymer completely, which will yield a 22% w/v
polymer solution (see Note 2).
2. Place aluminum foil of dimension 10 cm × 10 cm in the center
of a plastic sheet (¼ in. thickness and such that it fits inside
the plastic chamber) with a ½ in. diameter hole in the middle.
250 James et al.
Place the plastic sheet upright inside the electrospinning

chamber and fix its position. Pass the grounded cable through
the hole in the middle of the plastic sheet and attach it to the
now vertically oriented aluminum foil target (see Note 3).
A schematic of the electrospinning setup is shown in Fig. 1.
3. Place the syringe pump on the lab jack placed inside the
chamber. Place an empty 10-mL syringe with an 18-G needle
attached onto the syringe pump with the blunt end of the
needle facing the vertically oriented aluminum foil. Adjust
the height of the lab jack such that the needle tip is aligned
with the center of the plastic sheet. Adjust the horizontal dis-
tance between the needle tip and the grounded aluminum
foil to be 30 cm (this is referred to as the air-gap distance).
Remove the empty syringe from the syringe pump and detach
the needle.
4. Fill the 10-mL syringe with 8 mL of the polymer solution and
attach an 18-G blunt end needle to it. Remove all air bubbles
from the polymer solution contained in the syringe (see Note 4).
5. Place the syringe–needle assembly containing the polymer
solution onto the syringe pump and set the flow rate to
4.0 mL/h.
6. Attach the DC voltage source to the tip of the stainless steel
blunt needle. Turn on the DC voltage and adjust it to 20 kV
(see Note 5). Switch off the power supply and the syringe
pump when all the polymer solution has been electrospun.
7. Remove the aluminum foil containing the electrospun scaf-
fold from the plastic sheet. Peel off the aluminum foil using
tweezers to separate the scaffold, and store at least overnight
under vacuum or until further use. The vacuum will remove
any residual solvent. An example of the structural changes
resulting from the use of varying polymer concentrations is
shown in Fig. 2. Scaffolds can be fabricated into various
shapes and size as shown in Fig. 3 (see Note 6).
3.2. SEM 1. Place double-sided carbon tape onto the aluminum specimen
mount stubs.
2. Cut the electrospun scaffolds using a sharp blade into squares
of side 0.5 cm. Attach the scaffolds to the carbon tape and
sputter-coat with gold for 150 s at 60 mA current and below
10−1 mbar vacuum. Sputter-coating deposits a conductive
metal on the scaffold to enable imaging using the electron
beam current (see Note 7).
3. Prepare SEM images of the scaffolds and determine fiber
diameter using NIH Image J software available freely at
http://rsbweb.nih.gov/ij/.
Fig. 2. SEM image of electrospun scaffolds fabricated from (a) 12%, (b) 14%, (c) 16%, (d) 18%, (e) 20%, (f) 22%, and (g)
24% w/v PLAGA 65:35 polymer in 3:1 THF:DMF pumped at a flow rate of 4 mL/h from a 18-G blunt-end needle with an
applied electric potential of 20 kV across an air-gap distance of 30 cm. At 12% w/v, the SEM image shows beads of
polymer attached by a thin fiber of polymer. This is referred to as beads-on-a-string morphology. With increasing concen-
tration/viscosity of the polymer solution, the rounded beads begin to flatten out with more fibers being fabricated. At
concentrations of 20, 22, and 24% w/v polymer, beads are minimal or not present at all, and the fabricated scaffold is
composed completely of nano-diameter fibers.
3.3. Cell Proliferation 1. The following instructions assume familiarity with primary
Assay cell isolation and aseptic cell culture techniques. Isolate adipose-
derived stromal cells from the inguinal fat pad of wild-type
Fischer 344 rats and maintain in culture using low glucose
DMEM supplemented with 10% FBS and 1% P/S at 37°C in
a humidified incubator. DMEM provides the nutrient media
for the cells, in which they survive, grow, and divide. FBS
contains a rich variety of proteins which helps maintain
cultured cells in media. P/S is an antibiotic to prevent bacterial
252 James et al.
Fig. 3. Electrospun PLAGA 65:35 nanofiber scaffolds fabricated as sheets and tubes of
varying diameter. The electrospinning technique is highly versatile and can be readily
modified to fabricate scaffolds of different shapes and sizes. Based on the type of col-
lector/target and its motion, it is possible to shape the scaffold and align the deposited
fibers. The use of a rotating mandrel/drum results in tubular scaffolds as shown.
contamination during culture. Passage the cells when they are

70% confluent. Detach the cells using 0.05% trypsin to provide
new maintenance cultures in T-75 flasks.
2. Cut the electrospun scaffolds into squares of side 1 cm and
sterilize them for 5 min in 70% ethanol and then through UV
sterilization in a cell culture hood for 30 min each on both
sides. Soak the sterile scaffolds overnight in low glucose
DMEM at 37°C in a humidified incubator. Incubation over-
night in media will wet the scaffolds and allow the media to
infiltrate into the porous structure (see Note 8).
3. Ensure that the cells have reached 70% confluence in the T-75
flasks. Rinse the flasks gently with warmed PBS and lift the
monolayer of cells using 0.05% trypsin for 5 min. Neutralize
trypsin activity with low glucose DMEM supplemented with
10% FBS and gently pellet the cells for 5 min at 500 × g.
Resuspend the cell pellet in DMEM and dilute using media

to 30,000 cells/100 mL.
4. Place the wetted scaffolds into wells of a 24-well non-TCP
plate.
5. Place 100 mL of the cell suspension onto each scaffold and
incubate for 2 h at 37°C in a humidified chamber.
6. Add 0.5 mL of supplemented DMEM into each well and incu-
bate at 37°C in a humidified chamber. Replace with fresh sup-
plemented low glucose DMEM every 3 days (see Note 9).
7. The following instructions assume the availability of and
familiarity with an absorbance spectrophotometer. Aspirate
the media present in the wells. Gently rinse the cell-seeded
scaffolds three times with 1 mL of warmed PBS solution.
Transfer the scaffold using forceps to a new well plate, incu-
bate with 1 mL of DMEM, and add 200 mL of MTS reagent
(provided in the cell proliferation assay kit). MTS reagent will
be metabolized in the mitochondria of live cells into a col-
ored product, which is then quantified by absorbance
spectroscopy.
8. Incubate the scaffolds for 2 h in a humidified chamber at
37°C. Stop the reaction with the addition of 250 mL of 10%
SDS.
9. Transfer 250 mL of the incubated solution from each scaffold
into a single well of a 96-well plate. Measure the absorbance
at 490 nm (see Note 10). At 490 nm, the absorbance of
the metabolized product of the MTS reagent is quantified.
10. Create a standard curve using samples with known numbers
of primary adipose-derived stromal cells to correlate absorbance
to cell number. An example result of cellular proliferation of
MSCs on a 22% w/v PLAGA 65:35 electrospun nanofiber
scaffold is shown in Fig. 4. Proliferation of human skin
fibroblasts on PLAGA 50:50 electrospun nanofiber scaffolds
composed of varying fiber diameters is depicted in Fig. 5.
3.4. Confocal 1. Prepare PBS solution of 4 mM ethidium homodimer-1 (EthD-1)

Microscopy and 2 mM calcein AM (EthD-1 and calcein AM are reagents
supplied in the assay kit) using the stock solution reagents.
Cover and protect from light as the reagents are sensitive to
light (see Note 11).
2. Rinse the cell–scaffold constructs with warmed PBS three
times and then transfer them to a glass chamber slide.
3. Add PBS reagent solution to the wells to cover the cell–scaf-
fold constructs sufficiently. Incubate at room temperature
and protected from light for up to 45 min.
4. Wash the scaffolds with warmed PBS and transfer to clean
glass chamber slide. Keep the scaffolds wet using PBS.
254 James et al.
Fig. 4. Cellular proliferation of primary adipose stromal cells over 28 days in vitro culture on PLAGA 65:35 electrospun
scaffold sheets fabricated from a 24% w/v polymer solution in 3:1 THF:DMF using process parameters of 20 kV potential,
30 cm air gap, 4.0 mL/h flow rate, and 18-G blunt needle. Cellular proliferation was significantly upregulated at 7 days
and at 21 days in culture. Initial seeding density was 30,000 cells/scaffold (n = 4, p < 0.05, asterisk indicates significant
difference). Stromal cells seeded on two-dimensional PLAGA 65:35 flat sheets proliferated very rapidly, but easily
detached during media changes and washing with PBS (data not shown).
7000 0.6
Fiber diameter
6000 14 day cell number 0.5
Cell count (110 ^ 6)
5000
Fiber dimeter (nm)
0.4
4000
0.3
3000
0.2
2000
1000 0.1
0 0
1 2 3 4 5 6 7
Sample identification
Fig. 5. Cellular proliferation of human skin fibroblasts seeded onto PLAGA 50:50 electrospun scaffolds at 14 days.
Scaffolds of varying polymer concentrations were electrospun to fabricate meshes of different fiber diameters. Cell
numbers were significantly higher on scaffolds composed of fibers 350–1,100 nm in diameter (i.e., Matrix 3–5). The
proliferation trend was consistent throughout all the time points studied up to 28 days in culture. Fiber diameter range
for the electrospun scaffolds is measured from SEM images: Matrix 1 and Matrix 2: 200–300 nm (same diameter range),
Matrix 3: 250–467 nm, Matrix 4: 500–900 nm, Matrix 5: 600–1,200 nm, Matrix 6: 2,500–3,000 nm, and Matrix 7: 3,250–
6,000 nm. Reprinted from ref. 19 with permission from Elsevier.
Fig. 6. Confocal microscopy image (×20) of aMSCs on a PLAGA 65:35 electrospun nano-
fiber scaffold sheet at 21 days in vitro culture. Live cells have been stained using the
Live/Dead Viability/Cytotoxicity kit for mammalian cells. The aMSCs show a well-spread
morphology with extending cellular processes both at the surface and at various depths
into the PLAGA 65:35 nanofiber scaffold.
5. View the constructs using confocal microscopy. Excitation at

515 nm induces calcein AM emission (green emission,
em = 495 nm) for the detection of live cells present on the scaf-
fold, while excitation at 635 nm induces EthD-1 (red emission,
em = 495 nm) for the detection of dead cells on the scaffold.
Software can be used to overlay the fluorescence images.
Examples of primary adipose stromal cells present on the elec-
trospun PLAGA 65:35 nanofiber scaffolds are shown in Fig. 6.
4. Notes
1. Long-term storage of PLAGA is best at −20°C. Additionally,

it is best to aliquot smaller amounts of the polymer pellets
soon after the sealed pouch is opened. Prepare the aliquots
and wrap the capped end of the tubes or vials in paraffin film.
The polymer can then be stored long term until needed
without repeated opening, thus avoiding moisture condensa-
tion. Additionally, allow the vials/tubes containing the stored
polymer to come to room temperature before opening.
256 James et al.
When possible, use the same batch of the polymer. There may
be variations in the polydispersity index between different
manufactured batches. These differences may affect the elec-
trospinning processing parameters, which then may require
modification.
2. Always use adequate protection (organic solvent-resistant
gloves and protective eye wear) and work with organic sol-
vents inside a chemical fume hood. Prepare all the pipettes
and glass vials, and weigh out the polymer before measuring
the organic solvent since the solvent evaporates rapidly. Delay
in capping the vials will lead to solvent evaporation altering
the final polymer solution concentration. Always prepare a
fresh solvent (i.e., THF:DMF 3:1) just before adding it to the
polymer pellets. Store protected with similar organic solvents
in a fireproof cabinet. Ideally, use the polymer solution as
soon as it is dissolved. Do not keep the polymer solution in
glass vials for long durations as the solvent may evaporate in
small amounts from the capped vials.
3. Electrospinning setup must be housed inside a transparent
plastic-walled chamber exhausted to a chemical fume hood.
The inside of the plastic box or chamber must be accessible as
necessary and the chamber designed to be completely closed
to prevent solvent fumes from leaking out during the fabrica-
tion process and to prevent accidental contact with the high-
voltage power source.
4. Prepare the electrospinning setup in advance. Transferring
the polymer solution to the syringe, removing air bubbles
from the syringe, and placing the syringe onto the syringe
pump should be the penultimate step before attaching the
high-voltage power source, closing the chamber, and turning
the power on.
5. When charged, ensure that the polymer droplet forming at
the needle end is moving toward the target. When sufficient
electric potential is applied, the rounded polymer droplet
coming out of the needle end becomes conical in shape. This
phenomenon is referred to as a Taylor cone. Due to solvent
evaporation occurring at the needle end, some polymer may
deposit and clog the needle orifice. In such a scenario, turn
off the voltage source and then open the chamber to access
the needle. Stop the syringe pump and clean the needle tip
using Kimwipes.
6. When about 10–20 mL of polymer solution is deposited onto
a 10 cm × 10 cm aluminum foil target, the polymer sheet can
be readily separated without tearing. The electrospinning
parameters can be readily modified to accommodate other
polymer–solvent systems. Additionally, the target can be
modified to include different shapes and rotational movement
such as a rotating mandrel. For any new polymer, it is best to

prepare different polymer concentrations in suitable solvents.
Electrospinning small volumes onto smaller targets will enable
one to narrow down a working range for the system
parameters.
7. Quantification of the fiber diameter data is done by taking
measurements of 100 individual fibers within each image and
averaging the results.
8. Immersion in ethanol may cause shrinkage of some electro-
spun scaffolds. Thicker and bigger scaffolds may accommo-
date these dimensional changes.
9. Media should be added gently along the sides of the well fol-
lowing incubation of the wetted scaffold with the cell
suspension.
10. Make sure no bubbles are present in the incubated solution
transferred to the 96-well plate prior to absorbance measure-
ment. Gently pop any bubbles using a sharp needle.
Additionally run a blank sample, which is a wetted scaffold
without any cells. This control sample will determine any
interference of the polymer with the assay reagents.
11. For economy, only prepare enough PBS solution with 4 mM
EthD-1 and 2 mM calcein AM to just cover the scaffold.
Staining is done in chamber wells where the gasket around
the samples will retain the solutions.
Acknowledgments
The authors gratefully acknowledge funding from the NIH (R01

EB004051 and R01 AR052536), the NSF-(EFRI-0736002),
and the Coulter Foundation (526203-FY10-01).
References
1. Almarza, A. J., Yang, G., Woo, S. L., Nguyen, DM, DZ, EC, EE, EG, ES, FI, GB, GD,
T., and Abramowitch, S. D. (2008) Positive GE, GH, GM, GT, HN, HR, HU, ID, IL,
Changes in Bone Marrow-Derived Cells in IN, IS, JP, KE, KG, KM, KN, KP, KR, K
Response to Culture on an Aligned Bioscaffold. (TRUNCATED).
Tissue Eng. Part A 14, 1489–1495. 3. Laurencin, C. T., Nair, L. S., Bhattacharyya,
2. Li, S., Patel, S., Hashi, C., Huang, N. F., and S., Allcock, H. R., Bender, J. D., Brown, P.
Kurpinski, K. (2007) Biomimetic Scaffolds W., and Greish, Y. E. (2005) Polymeric
made from Biodegradable Nanofibers Nanofibers for Tissue Engineering and Drug
Covalently Linked with Extracellular Matrix Delivery. Main IPC: A61L027-18; Secondary
and/or Animal Growth Factors and IPC: A61L027-50, PCT Designated States:
Entrapped Cells. PCT Designated States: Designated States W: AE, AG, AL, AM, AT,
Designated States W: AE, AG, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BW, BY, BZ, CA,
AU, AZ, BA, BB, BG, BR, BW, BY, BZ, CH, CN, CO, CR, CU, CZ, DE, DK, DM,
CA, CH, CN, CO, CR, CU, CZ, DE, DK, DZ, EC, EE, EG, ES, FI, GB, GD, GE, GH,
258 James et al.
GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, 12. Li, X. T., Zhang, Y., and Chen, G. Q. (2008)
KP, KR, KZ, LC, LK, LR, L Nanofibrous Polyhydroxyalkanoate Matrices
(TRUNCATED). as Cell Growth Supporting Materials.
4. James, R., Kesturu, G., Balian, G., and Biomaterials 29, 3720–3728.
Chhabra, A. B. (2008) Tendon: Biology, 13. Stupp, S. I., Hartgerink, J. D., and Niece, K.
Biomechanics, Repair, Growth Factors, and L. (2004) Self-Assembling Peptide-Amphiphiles
Evolving Treatment Options. J. Hand Surg. and Self-Assembled Peptide Nanofiber Networks
[Am] 33, 102–112. for Tissue Engineering. Main IPC: C07K, PCT
5. Kumbar, S. G., James, R., Nukavarapu, S. P., Designated States: Designated States W: AE,
and Laurencin, C. T. (2008) Electrospun AG, AL, AM, AT, AU, AZ, BA, BB, BG, BR,
Nanofiber Scaffolds: Engineering Soft Tissues. BY, BZ, CA, CH, CN, CO, CR, CU, CZ,
Biomed. Mater. 3, 034002. DE, DK, DM, DZ, EC, EE, EG, ES, FI, GB,
6. Kumbar, S. G., Nair, L. S., Bhattacharyya, S., GD, GE, GH, GM, HR, HU, ID, IL, IN, IS,
and Laurencin, C. T. (2006) Polymeric JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, L
Nanofibers as Novel Carriers for the Delivery (TRUNCATED).
of Therapeutic Molecules. J. Nanosci. 14. Liu, X. and Ma, P. X. (2009) Phase Separation,
Nanotechnol. 6, 2591–2607. Pore Structure, and Properties of Nanofi-
7. Kumbar, S. G., Nukavarapu, S. P., James, R., brous Gelatin Scaffolds. Biomaterials 30,
Hogan, M. V., and Laurencin, C. T. (2008) 4094–4103.
Recent Patents on Electrospun Biomedical 15. Jiang, H., Hu, Y., Li, Y., Zhao, P., and Zhu,
Nanostructures: An Overview. Recent Pat. K. (2005) Preparation of Core-Shell Nano/
Biomed. Eng. 1, 68–78. Micro-fiber or Capsule for Drug Sustained
Release. Main IPC: A61K009-00, Patent
8. He, W., Yong, T., Teo, W. E., Ma, Z., and
Application Country: Application: CN; Patent
Ramakrishna, S. (2005) Fabrication and
Country: CN.
Endothelialization of Collagen-Blended
Biodegradable Polymer Nanofibers: Potential 16. Andersen, E., Smith, D., and Reneker, D.
Vascular Graft for Blood Vessel Tissue (2005) A Medical Device with Nanofiber Outer
Engineering. Tissue Eng. 11, 1574–1588. Surface Layer Incorporating Nitric Oxide and
Poly(Ethylenimine) Diazeniumdiolate for
9. Janjanin, S., Li, W. J., Morgan, M. T., Shanti,
Insertion in to Vascular System. Main IPC:
R. M., and Tuan, R. S. (2008) Mold-
A61L029-00, PCT Designated States:
Shaped, Nanofiber Scaffold-Based Cartilage
Designated States W: AE, AG, AL, AM, AT,
Engineering using Human Mesenchymal
AU, AZ, BA, BB, BG, BR, BW, BY, BZ, CA,
Stem Cells and Bioreactor. J. Surg. Res.
CH, CN, CO, CR, CU, CZ, DE, DK, DM,
149, 47–56.
DZ, EC, EE, EG, ES, FI, GB, GD, GE, GH,
10. Koh, H. S., Yong, T., Chan, C. K., and GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP,
Ramakrishna, S. (2008) Enhancement of KR, KZ, LC, LK, LR, L (TRUNCATED).
Neurite Outgrowth using Nano-Structured 17. Kumbar, S. G., Bhattacharyya, S., Sethuraman,
Scaffolds Coupled with Laminin. Biomaterials S., and Laurencin, C. T. (2007) A Preliminary
29, 3574–3582. Report on a Novel Electrospray Technique
11. Park, Y. and Choo, J. E. (2007) Self-Assembly for Nanoparticle Based Biomedical Implants
Nanocomposites Comprising Hydrophilic Coating: Precision Electrospraying. J. Biomed.
Bioactive Peptides and Hydrophobic Materials Mater. Res. B 81, 91–103.
for Surface Treatment of Biomaterials and 18. Li, W. J., Jiang, Y. J., and Tuan, R. S. (2006)
Tissue Regeneration Therapy. PCT Designated Chondrocyte Phenotype in Engineered
States: Designated States W: AE, AG, AL, AM, Fibrous Matrix Is Regulated by Fiber Size.
AT, AU, AZ, BA, BB, BG, BR, BW, BY, BZ, Tissue Eng. 12, 1775–1785.
CA, CH, CN, CO, CR, CU, CZ, DE, DK, 19. Kumbar, S. G., Nukavarapu, S. P., James, R.,
DM, DZ, EC, EE, EG, ES, FI, GB, GD, GE, Nair, L. S., and Laurencin, C. T. (2008)
GH, GM, GT, HN, HR, HU, ID, IL, IN, IS, Electrospun Poly(Lactic Acid-Co-Glycolic
JP, KE, KG, KM, KN, KP, KZ, L Acid) Scaffolds for Skin Tissue Engineering.
(TRUNCATED). Biomaterials 29, 4100–4107.
Chapter 17
Peptide Amphiphiles and Porous Biodegradable Scaffolds

for Tissue Regeneration in the Brain and Spinal Cord
Rutledge G. Ellis-Behnke and Gerald E. Schneider
Abstract
Many promising strategies have been developed for controlling the release of drugs from scaffolds, yet
there are still challenges that need to be addressed in order for these scaffolds to serve as successful treat-
ments. The RADA4 self-assembling peptide spontaneously forms nanofibers, creating a scaffold-like tissue-
bridging structure that provides a three-dimensional environment for the migration of living cells. We
have found that RADA4: (1) facilitates the regeneration of axons in the brain of young and adult hamsters,
leading to functional return of behavior and (2) demonstrates robust migration of host cells and growth
of blood vessels and axons, leading to the repair of injured spinal cords in rats.
Key words: CNS regeneration, Spinal cord injury, Tissue repair, Self-assembling peptide,
Nanofiber scaffold, Schwann cell, Neural progenitor cell, Surgery, Trauma, Nanomedicine
1. Introduction
Nanotechnology has ushered in a new era of possibilities in tissue

and organ reconstruction. The ability of researchers to establish
fine control of the nanodomain is making way for increased
targeting of cell placement and therapeutic delivery, amplified by
cell encapsulation and implantation. In particular, scaffolds play a
central role in organ regeneration (1) and repair of central nervous
system (CNS) tissues (2, 3). Also, drug delivering scaffolds may
need to be combined with cells to obtain functional recovery in
treating traumatic brain injury (TBI) or spinal cord injury (SCI) (4).
Acting as a template, scaffolds can guide cell proliferation,
cell differentiation, and tissue growth, influencing the survival of
transplanted cells or the invasion of cells from the surrounding
tissue, by providing a surface for cell adhesion and migration. In
addition, they can be used to deliver drugs at a rate designed to
259
260 Ellis-Behnke and Schneider
match the physiological need of the tissue (5). Yet, there are still
challenges to be addressed in order for these scaffolds to serve as
successful treatments (2, 4, 6–8). First, the cells must be precisely
placed into the scaffold before it is implanted to prevent their
migration. Second, in order for the tissue to be reconstituted
from the surrounding area, the scaffold must be able to allow cells
to migrate into it. Third, it is important to prevent the acidic
breakdown of the cell scaffold, which, if allowed, results in an
adverse environment for cell growth.
Scaffolds can be synthesized from a variety of different materials
and each type of material has distinct advantages and disadvan-
tages. Natural scaffolds, including alginate (9–11), chitosan
(12–14), collagen (15, 16), fibrin (17–23), and hyaluronan (6,
24–26), impart intrinsic signals within the structure that can
enhance tissue formation (27), contain sites for cell adhesion, and
allow for cell attachment; they also exhibit similar properties to the
tissues they are replacing (28). Some, but not all, natural materials
also allow for cell infiltration (28). However, since these materials are
obtained from natural sources, homogeneity of the product
between batch preparations can be an issue (28), and purification
is essential to prevent a foreign body response after implantation.
For example, chitosan can cause an allergic reaction (29) while
fibrins from blood products and collagen from animal products
have been known to cause an immune response as well as transfer
infectious agents from donor to recipient (28). In addition, the
modulus of these natural scaffolds may be very different from that
of the tissue in which they are implanted (30). This difference can
cause expression patterns of genes in the cell that are not condu-
cive to repair and that could lead to the development of tissues
unable to function like the original tissues (31). In a pulsatile envi-
ronment, these materials could shear away from the surrounding
tissue, causing physical damage or even increased rates of cellular
compression (30). Many flexible scaffold systems have been devel-
oped, utilizing both a natural (i.e., collagen) and a synthetic com-
ponent that can degrade by both hydrolysis and collagenase
degradation pathways, as well as support cell growth. This type of
scaffold could possibly be used in soft tissue applications (32).
Unlike natural scaffolding materials, synthetic materials,
including PEG [poly (ethylene glycol)] (33–36), PLA [poly (lactic
acid)]/PGA [poly (glycolic acid)]/PLGA [poly (lactic-co-glycolic)
acid] (37–41), and pHEMA-MMA [poly (2-hydroxyethyl meth-
acrylate-co-methyl methacrylate)] (42–46), have known compo-
sitions and can be custom designed with specific mechanical or
degradation properties or to minimize the immune response (27).
Also, these polymers can be conjugated to produce materials with
properties that exhibit only the beneficial aspects of the individual
components (28). PLAs, PLGAs, and MMAs have been used to form
scaffolds pre-impregnated with cells (47) which then can be implanted
Peptide Amphiphiles and Porous Biodegradable Scaffolds 261
in the damaged tissue area (48). Unfortunately, the by-products

of these degraded synthetic materials can be absorbed by the body
and may cause pH changes around the implantation site, leading
to necrosis, delayed apoptosis, and, in rare cases, pain (28).
Designed self-assembling peptides (SAPs) that spontaneously
form nanofibers are the next generation of cell scaffolds to emerge
(2, 6, 8, 49, 50). These peptides create a scaffold-like tissue-
bridging structure that provides a three-dimensional (3D) envi-
ronment for cell growth and migration, similar to the native
extracellular matrix (6, 49, 51, 52). The scaffolds are assembled
through the weak ionic bonds and van der Waals forces of the
self-complementary peptides (2, 6, 8, 49, 50). Typically, these
peptides are composed of alternating positive and negative l-amino
acids that assemble into highly hydrated structures in the presence
of physiological concentrations of salts, tissue culture media, and
human body fluids, such as cerebrospinal fluid (2, 8, 53). The
benefits of SAPs over some natural and/or synthetic scaffold
materials are that they: (1) pose a minimized risk of carrying bio-
logical pathogens or contaminants (53, 54), (2) elicit no immune
response, and (3) show excellent physiological compatibility as
well as minimal cytotoxicity (55, 56). In addition, SAPs allow for
high cell implantation densities and enable cells to migrate freely
in and out of the scaffold and the surrounding tissue. Further,
these materials can be designed to match the modulus of the sur-
rounding tissue (30) and they can either be prebuffered or allowed
to be buffered during implantation by any physiological fluid
available from the surrounding sites (30). Finally, the pH in the
local environment does not decrease during breakdown of this
type of scaffold.
The RADA4 peptide is one family of amphiphilic molecules
that contain both a hydrophobic and a hydrophilic face. These
peptides spontaneously self-assemble into nanofibers about 10 nm
in diameter when the ionic strength of the solvent is increased to
above its acidity level or pH values are raised to neutrality (e.g.,
physiological salt concentrations, culture media, and buffers).
The nanofibers are highly hydrated, with greater than 99% water
content (16, 34, 35, 53); however, they can be mixed at higher
concentrations to reduce their water content and change their
physical properties. We have utilized this peptide scaffold to cre-
ate tissue-bridging structures that provide a 3D environment for
the migration of living cells that can be used to: (1) regenerate
axons in the brain of hamsters, leading to functional return of
behavior (55) and (2) repair the injured spinal cord of rats, dem-
onstrating robust migration of host cells, and growth of blood
vessels and axons into the scaffolds (57). This chapter outlines in
detail the materials and methods that are needed to: (1) repair the
brain and (2) repair the spinal cord using this self-assembling
nanopeptide scaffold system.
2. Materials
2.1. Preparation 1. 1% RADA4 solution: 10 mg RADA4 powder (obtained from

of the RADA4 Solution the Massachusetts Institute of Technology Center for Cancer
Research Biopolymers Laboratory, Cambridge, MA) in 1 mL
deionized (DI) water (Millipore Corp, Billerica, MA) (see Fig. 1).
2.2. Animals – Brain 1. 53 P2 Syrian hamster pups and adult Syrian hamsters
(Mesocricetus aurotus) (see Note 1).
2. Sodium pentobarbital (50 mg/kg).
3. Gelfoam.
4. Wound clips (or suturing materials).
5. Knife.
6. Head holder.
7. Isotonic saline (typically used in any surgical procedure).
2.3. Animal Behavior 1. Sunflower seeds.

Testing – Brain 2. Small black rubber ball or block, 1–1.5 cm in diameter.
3. White wire.
4. Cages (26 cm × 43 cm).
5. Video camera.
2.4. Preparation for 1. Sodium pentobarbital (50 mg/kg).

Tracing Regenerated 2. Glass micropipette with a tip diameter (~10 mm) attached to
Axons – Brain a Pico Spritzer (General Valve, Fairfield, NJ).
Fig. 1. Self-assembling peptide nanofiber scaffold RADA4. (Left) Molecular model of the RADA4 (arginine, alanine, aspar-
tate, and alanine) building block. (Right) This peptide is a liquid when dissolved in deionized (DI) water (shown in a vial).
When applied to a physiological environment, the material becomes a gel (data not shown).
3. 1% Cholera-toxin subunit B conjugated with FITC (CTB-

FITC).
4. 0.9% NaCl.
5. 0.25% NaNO3.
6. 2% Paraformaldehyde in phosphate-buffered saline (PBS)
(pH = 7.4).
7. 30% Sucrose.
8. Gelatin-coated slides.
2.5. Immunolabeling 1. 0.1 M PBS (pH = 7.4).

of the Optic Tract 2. 2% Triton 100, 2% normal rabbit serum, 2.5% bovine serum
Axons – Brain albumin (BSA) in 0.1 M PBS (pH = 7.4).
3. Goat anticholeragenoid (List Biological Laboratories,
Campbell, CA) (1:8,000 dilution), 2% Triton 100, 2% nor-
mal rabbit serum, 2.5% BSA in 0.1 M PBS (pH = 7.4).
4. Fluorescent donkey anti-IgG antibodies Alexa-488 (secondary
antibody from Invitrogen – Molecular Probes) (1:200 dilution).
5. DAKO mounting medium (DAKO, Carpinteria, CA).
6. Fluorescence microscope.
7. Digital camera.
2.6. Animals – Spinal 1. Adult female wild-type Sprague–Dawley rats (220–250 g).

Cord (See Note 1) 2. Adult female green fluorescent protein (GFP)-transgenic
Sprague–Dawley rats [“green rat CZ-004” SD TgN (act-
EGFP) OsbCZ-004] (220–250 g).
2.7. Pretreatment for 1. Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12
RADA4 – Spinal Cord (DMEM/F12) with 10% fetal bovine serum (FBS) (Gibco,
Grand Island, NY).
2.8. Isolation 1. 10% FBS in DMEM/F12.

and Culture 2. 1.25 U/mL Dispase.
of Schwann Cells
3. 0.05% Collagenase.
(ScCs) – Spinal Cord
4. 15% FBS in DMEM/F12.
5. Poly-l-lysine (0.01%, Sigma, St. Louis, MO).
6. 20 mg/mL Pituitary extract (Sigma) and 2 mM forskolin
(Sigma).
7. Ca2+- and Mg2+-free Hanks balanced salt solution (CMF-
HBSS, Gibco).
8. 0.05% Trypsin in CMF-HBSS (Gibco).
9. 0.02% Ethylenediaminetetraacetic acid (EDTA) in CMF-
HBSS (Gibco).
2.9. Isolation 1. Ca2+- and Mg2+-free Hanks balanced salt solution (CMF-
and Culture of Neural HBSS, Gibco).
Precursor Cells 2. DMEM/F12 supplemented with B27 (2%, Gibco).
(NPCs) – Spinal Cord
3. N2 (1%, Gibco).
4. Epidermal growth factor (EGF, 20 ng/mL, Gibco).
5. Basic fibroblast growth factor (bFGF, 20 ng/mL, Sigma).
6. Penicillin (100 U/mL).
7. Streptomycin (100 mg/mL).
8. Poly-l-lysine (0.01%, Sigma).
2.10. ScCs and NPCs 1. 10% FBS in DMEM/F12.

in 3D Culture Within 2. Two-photon confocal microscope (Zeiss LSM510 META,
the RADA4 Scaffold – Jena, Germany).
Spinal Cord
2.11. Immunocyto- 1. 4% Paraformaldehyde.

chemistry – 2. Poly-l-lysine (0.01%, Sigma).
Spinal Cord
3. 30% Sucrose.
5. 1% BSA, 10% normal goat serum, 0.3% Triton X-100 in 0.1 M
PBS (pH = 7.4).
6. Rabbit antip75 (1:200 dilution, Promega, Madison, WI).
7. Mouse anti-nestin (1:2,000 dilution, BD Biosciences,
Cambridge, MA).
8. Rabbit anti-glial fibrillary acidic protein (anti-GFAP, 1:1,000
dilution, Chemicon, Temecula, CA).
9. Mouse anti-Rip (1:50 dilution, BD Biosciences, Cambridge,
MA).
10. Mouse anti-b-tubulin type III (1:1,000 dilution, Sigma).
1 1. Fluorescent Alexa 568 goat anti-mouse or anti-rabbit
secondary antibody (1:400 dilution, Molecular Probes,
Eugene, OR).
12. DAKO mounting medium containing 4¢,6-diamidino-2-
phenylindole (DAPI, 2 mg/mL).
2.12. Surgical 1. Ketamine (80 mg/kg).

Procedures – 2. Xylazine (10 mg/kg).
Spinal Cord
3. 11-0 suture.
4. Microscoop (suction can also be used).

5. Isotonic saline.
2.13. Preparation 1. 4% Paraformaldehyde in 0.1 M PBS (pH = 7.4).

of Tissue Sections – 2. Pentobarbital anesthesia.
Spinal Cord
3. 10% Sucrose and 4% paraformaldehyde in 0.1 M PBS
(pH = 7.4).
4. 30% Sucrose in 0.1 M PBS (pH = 7.4).
6. Optimum cutting temperature compound (OCT).
2.14. Immunohisto- 1. 0.1 M PBS (pH = 7.4).

chemistry on 2. 1% BSA, 10% normal goat serum, and 0.3% Triton X-100 in
Tissue Sections – 0.1 M PBS (pH = 7.4).
Spinal Cord
3. Mouse anti-nestin (1:2,000 dilution).
4. Rabbit anti-GFAP (1:1,000 dilution).
5. Mouse anti-Rip (1:50 dilution).
6. Mouse anti-b-tubulin III (1:500 dilution).
7. Mouse anti-NF200 (1:400 dilution, Sigma).
8. Rabbit anti-5HT (serotonin, 1:200 dilution, Sigma).
9. Rabbit anticalcitonin gene-related peptide (CGRP, 1:200
dilution, Sigma).
10. Mouse anti-ED1 (1:1,000 dilution, Serotec, Raleigh, NC).
11. Rabbit antip75 (1:200 dilution, Promega) for SCs.
12. Mouse anti-myelin basic protein (anti-MBP, Ipswich, MA,
1:1,000 dilution).
13. Fluorescent Alexa 568 goat anti-mouse or anti-rabbit secondary
antibody (1:400 dilution).
14. DAKO mounting medium containing DAPI (2 mg/mL).
15. Two-photon confocal microscope.
2.15. Hematoxylin 1. H&E and native endothelial alkaline phosphatase (AP).

and Eosin (H&E) 2. Collagen-coated or -charged slides.
Staining and Alkaline
3. Distilled water.
Phosphatase (AP)
Histochemical 4. Alum hematoxylin.
Staining – Spinal Cord 5. 0.3% Acid alcohol.
6. Scott’s tap water substitute.
3. Methods
3.1. Preparation 1. Mix 10 mg of RADA4 powder in 1 mL of DI water, sonicate

of the RADA4 for 30 s, and filter (see Note 2).
Solution – Brain
3.2. Young Animals – 1. Anesthetize 53 P2 Syrian hamster pups hypothermically using

Brain whole-body cooling. Open the scalp and sever the optic tract
within the superior colliculus (SC) with a sharp surgical knife
through a slot cut in the cartilaginous skull, extending 1.5 mm
below the surface, from the midline to a point beyond the
lateral margin of the SC.
2. Treat animals by injection into the brain wound with 10 mL of
1% RADA4 or 10 mL isotonic saline (control) (see Fig. 2). The
cut and untreated controls receive no injection.
Fig. 2. Montage of parasagittal sections of a cut filled with saline solution (control), 72 h
survival time postlesion. Scale bars = 100 mm.
Fig. 3. Dorsal view reconstruction of the hamster brain with cortex removed. Rostral is
to the left and caudal is to the right. The heavy black line depicts the location of the
transection of the optic tract (brachium of the SC). The locations of the superior colliculus
(SC), pretectal area (PT), lateral posterior nucleus (LP), medial geniculate body (MGB),
lateral geniculate body (LGB), and inferior colliculus (IC) are shown.
3.3. Adult Animals – 1. Anesthetize adult Syrian hamsters with an intraperitoneal

Brain (See Note 3 injection of sodium pentobarbital (50 mg/kg) and then fit
and Ref. 58) them in a head holder. Expose and aspirate the overlying
cortex to reveal the rostral edge of the SC and the brachium of
the SC on the left side. Completely transect the brachium
of the SC of each animal (see Fig. 3 and Note 4).
2. With the aid of a sterile glass micropipette, inject either 30 mL
of 1% RADA4 solution or 30 mL isotonic saline (control) into
the site of the lesion. The cut and untreated control group
receives no injection.
3. Fill the surgical opening in the skull with saline-soaked gelfoam
and close the overlying scalp with wound clips.
4. Perform behavior testing at several time points after the date
of surgery (see Subheading 3.4).
3.4. Animal Behavior 1. Video-record animal behavior testing trials for analysis.
Testing – Brain 2. Allow the animal to wait, relatively motionless, on the plat-
(See Note 5) form for several seconds (see Note 6).
3. Test visually elicit orienting movements by presenting sun-
flower seeds (see Note 7) to part of the hamster’s visual field
(temporal or nasal, upper or lower), avoiding the nasal-most
45° (see Fig. 4) first by hand, and later with the aid of a white
metal wire, on the end of which there is a small black rubber
ball, slotted for holding a seed (see Note 8).
4. Make a minimum of ten presentations on each side, with random

choice of side. Animals should be tested two to three times
per week by two different investigators independently.
3.5. Preparation for 1. Anesthetize animals with an intraperitoneal injection of
Tracing Regenerated sodium pentobarbital (50 mg/kg).
Axons – Brain 2. Using a glass micropipette (tip diameter, ~10 mm) attached to
a Pico Spritzer, deliver intraocular injections of 1 mL of 1%
cholera-toxin subunit B conjugated with FITC into the vitreous
humor of the right eye of each animal (see Fig. 5).
3. Return the animals to their cages, place them under a heat
lamp, and monitor them until they recover.
4. Four days after intraocular injection, sacrifice the animals with
an overdose of anesthetic (see Note 9) and perfuse transcardi-
ally with 0.9% NaCl and 0.25% NaNO3 (pH = 7.4), followed
by 2% paraformaldehyde in 0.1 M PBS (pH = 7.4).
Fig. 4. Behavior. This adult animal turns toward the stimulus in the affected left visual field in small steps, prolonged here
by movements of the stimulus away from him. Each frame is taken from a single turning movement, at times (a) 0,
(b) 0.27, (c) 0.53, and (d) 0.80 s from movement initiation. The animal reached the stimulus just after the last frame. This
is about 0.20 s slower than most turns by a normal animal.
Fig. 5. The location of the eyes and location of the injection of CTB-FITC. This figure also illustrates how the eyes are
connected to the SC.
5. Remove the brains and eyes and postfix in 2% paraformaldehyde

at 4°C for 4 days. For cryoprotection, place brains in 30%
sucrose at 4°C until they sink. Cut 30 mm parasagittal sec-
tions on a cryostat and mount them directly on gelatin-coated
slides.
3.6. Immunolabeling 1. Air-dry the mounted sections, wash three times with 0.1 M
of the Optic Tract PBS (pH = 7.4) at 10 min intervals and preblock in 0.1 M PBS
Axons – Brain (pH = 7.4) containing 2% Triton 100, 2% normal rabbit serum,
and 2.5% BSA for 30 min at room temperature (see Note 10).
2. Incubate the slides with goat anticholeragenoid (1:8,000
dilution), 2% Triton 100, 2% normal rabbit serum, 2.5% BSA
for 48 h at room temperature (goat anticholergenoid labels
CTB-FITC, previously injected into the eye to trace the optic
pathway).
3. Wash the slides again three times in 0.1 M PBS (pH = 7.4)
and incubate with fluorescent donkey anti-IgG antibodies
Alexa-488 (secondary antibody) (1:200 dilution) for 1.5 h at
room temperature in a light-protected chamber.
4. Wash the slides four to five times in 0.1 M PBS (pH = 7.4) at
5 min intervals and coverslip with DAKO to mount the sections.
5. Examine the sections under a fluorescence microscope and
take pictures with a digital camera (see Fig. 6).
6. Analyze serial sections to reconstruct the locations of regen-
erated axons in the SC.
3.7. Pretreatment 1. Because untreated RADA4 peptide has a very low pH (about
of RADA4 – Spinal 3–4) and will damage the host tissue, either in the dorsal column-
Cord (See Note 11) transected or right-hemisected spinal cord, neutralize the
RADA4 in culture medium before transplantation. Gently and
quickly plate 1% RADA4 peptide to a dish in DMEM/F12 with
10% FBS. The self-assembly of the peptide into a scaffold material
will occur as soon as the material contacts the medium.
2. Change the medium twice at 1, 10, and 30 min after plating,
and once every 3 days during the following days.
3. On day 7, use the medium-treated RADA4 material for
transplantation.
3.8. Isolation 1. Isolate the ScCs by cutting the sciatic nerves from adult GFP-
and Culture of transgenic Sprague–Dawley rats into 1-mm3 explants and
Schwann Cells place in culture dishes with DMEM/F12 supplemented with
(ScCs) – Spinal Cord 10% FBS.
2. When the outgrowth of migratory cells (predominantly fibro-
blasts) reach a near-confluent monolayer around the explants
(about 7 days), transfer the explants to new culture dishes
with fresh medium.
Fig. 6. Brain of 8-month old hamster. Dark-field photo of a parasagittal section from the
brain treated with 1% RADA4 at the time of surgery in the lesion site. Rostral is left and
caudal is right. Arrows show the location of the lesion. The axons have grown through
the site of lesion and are reinnervating the SC. Note the lack of tissue disruption.
3. After three to five such passages (3–5 weeks), the cells that

emerge from the explants will be primarily ScCs. Transfer the
explants to a 35 mm dish containing 1.25 U/mL dispase,
0.05% collagenase, and 15% FBS in DMEM/F12 for incuba-
tion overnight at 37°C in 5% CO2.
4. On the following day, dissociate the explants and plate the
cells onto poly-l-lysine (0.01%)-coated dishes in DMEM/
F12 with 10% FBS.
5. Later, re-feed the cultures with the same medium supple-
mented with 20 mg/mL pituitary extract and 2 mM forskolin
for dividing.
6. When the ScCs reach confluence rinse in Ca2+- and Mg2+-free
Hanks balanced salt solution (CMF-HBSS) and briefly treat
with 0.05% trypsin and 0.02% EDTA in CMF-HBSS.
7. Wash cells twice in DMEM/F12 with 10% FBS and transfer
into new dishes at a density of 2 × 106 cells per 100 mm dish.
8. When the cells reach confluence again, collect for trans
plantation.
3.9. Isolation 1. Dissect the hippocampi of embryonic day 16 (E16) embryos

and Culture of of GFP-transgenic Sprague–Dawley rats in cooled CMF-
NPCs – Spinal Cord HBSS and dissociate mechanically.
2. Collect the cells after centrifugation and re-suspend in
DMEM/F12-B27 supplement (2%) N2 (1%), EGF (20 ng/
mL), bFGF (20 ng/mL), penicillin (100 U/mL), and strep-
tomycin (100 mg/mL).
3. Adjust the cells to 1 × 105 cells/mL and plant into culture flasks.
4. Replace the medium by one-half every 3 days.
5. Typically, the cells grow in suspending neurospheres (indi-
vidual clones of spheres derived from neural stem/progenitor
cells); mechanically dissociated them approximately once each
week and re-seed at approximately 1 × 105 cells/mL.
6. Use the second neurospheres for transplantation. To assess
the purification of the NPCs, dissociate some of the neuro-
spheres and seed onto poly-l-lysine-coated plates with the
same medium as above.
3.10. ScCs or NPCs 1. Before transplantation, culture the ScCs or NPCs within the
in 3D Culture Within RADA4 scaffold.
the RADA4 – Spinal 2. Collect the ScCs or NPCs and finally adjust to
Cord (See Note 12) 5 × 105 cells/mL.
3. Then, mix 1 mL cell suspension with 9 mL RADA4 peptide;
gently and quickly plate the mixture to a dish in DMEM/F12
with 10% FBS.
4. Change the medium twice at 1, 10, and 30 min after plating,
and once every 3 days during the following days.
5. On day 7, the cultures can be used for transplantation (see
Subheading 3.12).
6. Maintain some of the cultures for 4 weeks, take images of liv-
ing cells using a two-photon confocal microscope, to deter-
mine cell-material viability and perform immunostaining.
3.11. Immunocyto- 1. For immunostaining, directly fix the ScCs or dissociated cul-
chemistry – Spinal tured NPCs with 4% paraformaldehyde.
Cord (See Note 13) 2. Plate the neurospheres of the NPCs on poly-l-lysine-coated
coverslips and grow in culture medium. After the neuro-
spheres are attached on the coverslips (about 1 h), fix with 4%
paraformaldehyde.
3. Cryoprotect the fixed cultures of RADA4 with ScCs or NPCs
in 30% sucrose, then cut the cryostat sections (10 mm) and
mount on gelatin-coated slides.
4. Prior to exposure to antibodies, preblock all samples with 1%
BSA, 10% normal goat serum, and 0.3% Triton X-100 in
0.1 M PBS (pH = 7.4) for 1 h at room temperature (25°C);
then apply the following primary antibodies: (1) rabbit anti-p75

(1:200) for identifying ScCs; (2) mouse anti-nestin (1:2,000)
for NPCs; (3) rabbit anti-glial fibrillary acidic protein (anti-
GFAP, 1:1,000) for astrocytes; (4) mouse anti-Rip (1:50) for
oligodendrocytes; and (5) mouse anti-b-tubulin type III
(1:1,000; Sigma) for neurons.
5. Incubate the sections with the primary antibody in 0.1 M
PBS (pH = 7.4) plus 1% BSA and 0.3% Triton X-100 over-
night at 4°C and treat with fluorescent Alexa 568 goat anti-
mouse or anti-rabbit secondary antibody (1:400), respectively,
for 2 h at room temperature.
6. Finally, mount the sections with mounting medium containing
DAPI to counterstain the nuclei (see Figs. 7 and 8).
3.12. Surgical 1. Anesthetize the wild-type Sprague–Dawley rats with ketamine

Procedures – (80 mg/kg) and xylazine (10 mg/kg).
Spinal Cord 2. Perform a dorsal laminectomy on the sixth (C6) and seventh
cervical (C7) segments. Incise the dura longitudinally and
pull laterally. Make a spinal cord dorsal column between C6
and C7, followed by removing 1 mm dorsal column tissue
(see Fig. 9).
3. With the assistance of a microscoop, transfer 5 mL precul-
tured RADA4 scaffold or 5 mL cultured mixture of RADA4
with ScCs or NPCs (see Subheading 3.10) from the culture
dish into the lesion cavity.
Fig. 7. Images of live neural precursor cells (NPCs) in 1% RADA4 that survived in cell culture for 1 month. (Left) Merged
bright field and fluorescent images showing that the cells only grow where the RADA4 is and not in other parts of the
culture well. (Right) Two-photon microscope picture 4 weeks after the cells were cultured within RADA4 in vitro.
Fig. 8. RADA4 pretreated with Schwann cells (ScCs) added by culture medium before transplantation. The implants of
RADA4 integrate very well with host tissue, and no obvious cavities or gaps are observed between the implants and host.
Moreover, many cells migrated into the surrounding tissue from the implants, as shown by the GFP expressing cells
(white) that have migrated beyond the boundaries of the original injury site. With this type of pretreatment, the cells
survive very well in the implants. There is no evidence of RADA4 material in the implants after 8 weeks. Also, there is no
sign of a barrier at the tissue implant interface.
Fig. 9. Rat spinal cord. (Left ) Intact rat spinal cord. (Center ) Complete transection of the spinal cord. (Right ) Transected
spinal cord treated with RADA4 only. The molecule is 5 nm in length; the spinal cord is 2 mm.
4. As a control, place 5 mL uncultured RADA4 or saline in the

lesion cavity.
5. After transplantation, close the dura by suturing with an 11-0
suture. The muscle layers and skin should also be closed with
a suture.
3.13. Immunohisto- 1. Perfuse the animals transcardially with 4% paraformaldehyde

chemistry on Tissue in 0.1 M PBS (pH = 7.4) after an overdose of pentobarbital
Sections – Spinal Cord anesthesia at a certain time point after transplantation.
(See Note 14)
2. Postfix the spinal cords overnight in the perfusing fixative

plus 10% sucrose at 4°C.
3. Cryoprotect the tissues in 30% sucrose in 0.1 M PBS (pH = 7.4)
for 12 h at 4°C.
4. Separate a 1 cm length of the spinal cord, centered at the
injury site and embed it in OCT. Cut horizontal cryostat sec-
tions (30 mm) and mount onto gelatin-subbed slides (see
Note 15). Store at –20°C.
5. Air-dry the frozen slides at room temperature for 30 min and
wash with 0.1 M PBS (pH = 7.4) for 10 min. Then, preblock
with 1% BSA, 10% normal goat serum, and 0.3% Triton X-100
in PBS for 1 h at room temperature.
6. Afterward, apply the following primary antibodies overnight
at 4°C: mouse anti-nestin (1:2,000) for NPCs; rabbit anti-
GFAP (1:1,000) for astrocytes; mouse anti-Rip (1:50) for
oligodendrocytes; mouse anti-b-tubulin III (1:500) for neu-
rons; mouse anti-NF200 (1:400) for axons; rabbit anti-5HT
(serotonin, 1:200) for raphespinal axons; rabbit CGRP
(1:200) for primary sensory axons; mouse anti-ED1 (1:1,000)
for macrophages; rabbit antip75 (1:200) for SCs; and mouse
anti-myelin basic protein (anti-MBP 1:1,000) for myelin.
7. Wash the slides in 0.1 M PBS (pH = 7.4) three times and incu-
bate with fluorescent Alexa 568 goat anti-mouse or anti-rabbit
secondary antibody (1:400) for 2 h at room temperature.
8. Coverslip the slides with DAKO mounting medium containing
DAPI to counterstain the nuclei.
9. Take images using a confocal microscope.
3.14. Hematoxylin 1. Mount fixed sections at 10–50 mm on collagen-coated or

and Eosin (H&E) -charged slides.
Staining and Alkaline 2. To show the morphology and the angioregeneration around
Phosphatase (AP) the grafts, use H&E and native endothelial alkaline phos-
Histochemical phatase (AP).
Staining – Spinal Cord 3. Bring sections to distilled water.
(See Note 16)
4. Stain nuclei with the alum hematoxylin. The level and type of
fixation may affect the duration required for reaction. Four to
five minutes is a good start.
5. Rinse in running tap water.
6. Differentiate with 0.3% acid alcohol. Differentiation will take
between 3 and 5 min and could be longer depending on the
thickness of the tissue.
7. Rinse in running tap water.
8. Rinse in Scott’s tap water substitute, used for blueing, to
obtain optimal contrast for cell differentiation.
9. Rinse in tap water.
Fig. 10. Spinal cord with quantification grid. The grid is 200 by 200 mm with a line
bisecting the square at the center and the grid is placed in the center of the lesion on
every fourth section. The neurofilaments (white) are passing through the center of the
grid; each fiber is counted whenever it crosses any of the white lines. The data is com-
pared to a nonoperated control; the nonoperated control is a series of age-matched
controls where the average number of filaments is set at 100%. Even though 100% is
not needed for behavioral return, this is a way to measure the amount of reinnveration
and the continuity of the fibers reconnecting the disconnected area.
10. Stain with eosin 2 min. If the staining is too intense, it can be
reduced by rinsing.
11. Dehydrate, clear, and mount.
3.15. Quantification 1. Quantify NF-200-, CGRP-, and 5HT-positive axons that

of Axons – Spinal Cord regenerated within the graft on immunostaining sections by
using a fluorescent microscope. Only immunolabeled cells of
interest need to be quantified.
2. Superimpose three lines at intervals of 0.5 mm onto the graft
perpendicularly to the longitudinal axis with the middle one
through the center of the graft (see Fig. 10).
3. Count the axons intercepted with the superimposed lines.
The mean number of axons obtained from five sections of
each animal is defined as the number of axons.
4. Notes
1. In this note, we discuss the number of animals needed to

perform the experiments described herein.
In all of the experiments, the majority of the animals are
allowed to survive the maximum amount of time. For example,
we usually sacrifice 10, 10, 10, 20, and 50% of the animals at
each subsequent time point, for an experiment including five

time points; for experiments with three time points, we sacri-
fice 10, 30, and 60% of the animals at each subsequent time
point. This allows for a comparison of each of the time points,
and during these experiments the most regeneration is usually
seen at the later time points. Typical time points are: 30, 45,
60, and 90 days from the date of surgery.
In the case of the hamster pups, four groups of animals are
necessary for a valid comparison: (1) uncut controls, (2) cut
and untreated controls, (3) cut and saline-treated controls,
and (4) cut and treated controls. There should be six animals
in each group, for each time point. For example: 4 groups × 6
animals/group = 24 animals plus 5 time points × 6 animals for
each time point = 30 animals, for a total of 54 animals needed
for an experiment with 5 time points. For the adult hamsters,
five groups of animals are necessary for a valid comparison:
(1) uncut controls, (2) enucleated controls (where one eye is
removed to determine the rate of spontaneous turning toward
the blind side), (3) cut and untreated controls, (4) cut and
saline-treated controls, and (5) cut and treated controls. In
addition, behavior controls will be needed to determine spon-
taneous turning. Again, there should be six animals in each
group, for each time point. The group identity of an animal
should be unknown to the investigator during testing periods.
The uncut controls should be injected with the same carrier
fluid used to mix the peptide material.
For the wild-type rats, five groups of animals are necessary
for a valid comparison: (1) a saline control group, (2) an
uncultured RADA4 group, (3) a precultured RADA4 only
group, (4) a precultured RADA4 seeded with NPCs group,
and (5) a precultured RADA4 seeded with ScCs group. There
should be six animals in each group, for each time point.
We note that these are the minimum numbers of animals
needed and it is always better to have a few extras, especially
as experiments can last for up to a year. If your animal han-
dling and surgery skills are good and consistent, fewer ani-
mals could be used; if you have not handled animals very
often, you may want to increase the number of extra animals
by up to 50% of the total.
2. Take extra care to ensure that the material is pure by mixing
the powder into the liquid and letting it sit at room tempera-
ture for 1 month. An indication of its purity is that it remains
clear and odorless. The material can also be analyzed using
HPLC to assess purity.
3. It is important to use both young and adult animals. In young
animals, the optic tract is still growing (our intention was to see if
a permissive environment for growth could be created). Further,
with young animals, there is no need for growth factors.
4. Pay special attention to ensure that there is a complete transection

of the brachium of the SC from the lateral edge to the midline.
In many CNS regeneration models, the lesions are shallow.
To drive regeneration through the center of the lesion, use a
2 mm deep cut. The increased depth reduces the possibility of
axons growing around the bottom of the cut. Look specifi-
cally for regeneration growing through the center of the
lesion location since that shows the creation of a permissive
environment.
5. The purpose of animal behavior testing is to show that there
is functional return of vision driven by behavior. Many regen-
eration papers refer to action potentials in the axons but if
there is no return of functional behavior then regeneration is
useless. Only test adult animals for behavior.
6. The cage bottom is a convenient and useful test platform,
because the odors are very familiar to the animal. Investigators
should wear clean laboratory gowns and gloves (disposable
ones work well) because novel odors will interrupt the hamster’s
movements.
7. Sunflower seeds are a favorite food of hamsters, and under
the right conditions they will respond to visual presentations
of a seed, turn toward it, take it into their mouth, and transfer
it into a cheek pouch.
8. Count trials where the response occurs within 2 s of stimulus
presentation. In both normal and blind animals, a turning
response can also be elicited by touching the whiskers; there-
fore, care should be taken to avoid whisker contact during
visual presentations. Do not count a trial if the animal turns
before the visual stimulus presentation commences or if the
seed comes into contact with the whiskers. The completion of
a turn is signaled by the animal’s head coming to a stationary
position within 5° of the stimulus for at least one-third of a
second. If the animals orient in the direction of the stimulus
more than 70% of the time, vision is deemed as successfully
restored. Because an enucleated animal generally fails to
respond when the stimulus was placed outside the intact
visual field, its whiskers were often touched on the blind side
to elicit turns in that direction.
9. An overdose typically requires more than 60 mg/kg, but the exact
amount is dependent upon local animal handling requirements.
10. The preblocking solutions are used to reduce nonspecific
binding of the primary and secondary antibodies and are specific
for each primary antibody used. Care must be taken when
using multiple primary antibodies in the same section that the
preblock eliminates all nonspecific binding.
11. Neutralization is necessary in the spinal cord for culturing
ScCs prior to implantation into the lesion site because, unlike
the brain, the spinal cord has less cerebral spinal fluid that
allows for instantaneous buffering of the material.
12. ScCs are used because they are myelinating cells and can provide
growth factor support; NPCs are used to replace the lost neu-
ronal connections.
13. These steps are used to determine if the scaffold material
causes the cells to be transformed before they are implanted
(i.e., if the cells are healthy and viable after they have been
mixed with the material).
14. These steps are used to assess the regeneration in the spinal
cord and reconnection of the axons after lesion. Look for
everything from inflammation to numbers of astrocytes, neu-
rons, oligodendrocytes, raphespinal axons, macrophages, and
ScCs.
15. Take care to maintain the orientation of the tissue during
embedding and sectioning. Section the tissue longitudinally
and mount directly on coated slides.
16. The tissue structure is being labeled to look at the overall
distribution of collagen and cell structure. Immunostaining
reveals various cell types in the sections; H&E is a general
stain used to reveal the general appearance of the tissue so the
distribution of immunostained cells can be precisely located.
AP is a standard used for the visualization of vasculature.
Acknowledgments
The authors gratefully acknowledge Dr. David K. C. Tay in the

Department of Anatomy at the University of Hong Kong Faculty
of Medicine for reviewing the chapter and for taking pictures.
References
1. Ellis-Behnke, R. G., Teather, L. A., Schneider, 4. Willerth, S. M. and Sakiyama-Elbert, S. E.

G. E., and So, K. F. (2007) Using nanotechnol- (2008) Cell therapy for spinal cord regen-
ogy to design potential therapies for CNS regeneration. Adv. Drug Deliv. Rev. 60,
eration. Curr. Pharm. Des. 13, 2519–2528. 263–276.
2. Zhang, S., Holmes, T., Lockshin, C., and 5. Langer, R. (1998) Drug delivery and target-
Rich, A. (1993) Spontaneous assembly of a ing. Nature 392, 5–10.
self-complementary oligopeptide to form a 6. Holmes, T. C., de Lacalle, S., Su, X., Liu, G.,
stable macroscopic membrane. Proc. Natl Rich, A., and Zhang, S. (2000) Extensive
Acad. Sci. USA 90, 3334–3338. neurite outgrowth and active synapse forma-
3. Takenaga, M., Ohta, Y., Tokura, Y., tion on self-assembling peptide scaffolds. Proc.
Hamaguchi, A., Suzuki, N., Nakamura, M., Natl Acad. Sci. USA 97, 6728–6733.
et al. (2007) Plasma as a scaffold for regenera- 7. Schmidt, C. E. and Leach, J. B. (2003) Neural
tion of neural precursor cells after transplanta- tissue engineering: strategies for repair and
tion into rats with spinal cord injury. Cell regeneration. Annu. Rev. Biomed. Eng. 5,
Transplant. 16, 57–65. 293–347.
8. Zhang, S., Holmes, T. C., DiPersio, C. M., 19. Sakiyama-Elbert, S. E. and Hubbell, J. A. (2000)
Hynes, R. O., Su, X., and Rich, A. (1995) Controlled release of nerve growth factor from a
Self-complementary oligopeptide matrices heparin-containing fibrin-based cell ingrowth
support mammalian cell attachment. matrix. J. Control. Release 69, 149–158.
Biomaterials 16, 1385–1393. 20. Sakiyama-Elbert, S. E. and Hubbell, J. A.
9. Kataoka, K., Suzuki, Y., Kitada, M., (2000) Development of fibrin derivatives for
Hashimoto, T., Chou, H., Bai, H., et al. controlled release of heparin-binding growth
(2004) Alginate enhances elongation of early factors. J. Control. Release 65, 389–402.
regenerating axons in spinal cord of young 21. Taylor, S. J., McDonald III, J. W., and Sakiyama-
rats. Tissue Eng. 10, 493–504. Elbert, S. E. (2004) Controlled release of
10. Kataoka, K., Suzuki, Y., Kitada, M., Ohnishi, neurotrophin-3 from fibrin gels for spinal cord
K., Suzuki, K., Tanihara, M., et al. (2001) injury. J. Control. Release 98, 281–294.
Alginate, a bioresorbable material derived 22. Taylor, S. J., Rosenzweig, E. S., McDonald III,
from brown seaweed, enhances elongation of J. W., and Sakiyama-Elbert, S. E. (2006)
amputated axons of spinal cord in infant rats. Delivery of neurotrophin-3 from fibrin
J. Biomed. Mater. Res. 54, 373–384. enhances neuronal fiber sprouting after spinal
11. Prang, P., Muller, R., Eljaouhari, A., cord injury. J. Control. Release 113, 226–235.
Heckmann, K., Kunz, W., Weber, T., et al. 23. Taylor, S. J. and Sakiyama-Elbert, S. E. (2006)
(2006) The promotion of oriented axonal Effect of controlled delivery of neurotro-
regrowth in the injured spinal cord by alg- phin-3 from fibrin on spinal cord injury in a
inate-based anisotropic capillary hydrogels. long term model. J. Control. Release 116,
Biomaterials 27, 3560–3569. 204–210.
12. Crompton, K. E., Goud, J. D., Bellamkonda, 24. Gupta, D., Tator, C. H., and Shoichet, M. S.
R. V., Gengenbach, T. R., Finkelstein, D. I., (2006) Fast-gelling injectable blend of
Horne, M. K., et al. (2007) Polylysine- hyaluronan and methylcellulose for intrathe-
functionalised thermoresponsive chitosan cal, localized delivery to the injured spinal
hydrogel for neural tissue engineering. cord. Biomaterials 27, 2370–2379.
Biomaterials 28, 441–449. 25. Hahn, S. K., Jelacic, S., Maier, R. V., Stayton,
13. Freier, T., Koh, H. S., Kazazian, K., and P. S., and Hoffman, A. S. (2004) Anti-
Shoichet, M. S. (2005) Controlling cell adhe- inflammatory drug delivery from hyaluronic
sion and degradation of chitosan films by acid hydrogels. J. Biomater. Sci. Polym. Ed.
N-acetylation. Biomaterials 26, 5872–5878. 15, 1111–1119.
14. Freier, T., Montenegro, R., Shan Koh, H., 26. Tian, W. M., Hou, S. P., Ma, J., Zhang, C. L.,
and Shoichet, M. S. (2005) Chitin-based Xu, Q. Y., Lee, I. S., et al. (2005) Hyaluronic
tubes for tissue engineering in the nervous acid-poly-d-lysine-based three-dimensional
system. Biomaterials 26, 4624–4632. hydrogel for traumatic brain injury. Tissue
15. Archibald, S. J., Krarup, C., Shefner, J., Li, Eng. 11, 513–525.
S. T., and Madison, R. D. (1991) A collagen- 27. Segura, T., Anderson, B. C., Chung, P. H.,
based nerve guide conduit for peripheral Webber, R. E., Shull, K. R., and Shea, L. D.
nerve repair: an electrophysiological study of (2005) Crosslinked hyaluronic acid hydro-
nerve regeneration in rodents and nonhuman gels: a strategy to functionalize and pattern.
primates. J. Comp. Neurol. 306, 685–696. Biomaterials 26, 359–371.
16. Mahoney, M. J., Krewson, C., Miller, J., and 28. Willerth, S. M. and Sakiyama-Elbert, S. E.
Saltzman, W. M. (2006) Impact of cell type (2007) Approaches to neural tissue engineering
and density on nerve growth factor distribu- using scaffolds for drug delivery. Adv. Drug
tion and bioactivity in 3-dimensional collagen Deliv. Rev. 59, 325–338.
gel cultures. Tissue Eng. 12, 1915–1927. 29. Retrieved from HemCon Medical Technologies,
17. Herbert, C. B., Bittner, G. D., and Hubbell, Inc. on 1/20/2011: HemCon bandage FAQ/
J. A. (1996) Effects of fibinolysis on neurite Shellfish allergy study at www.hemcon.com/
growth from dorsal root ganglia cultured Products/HemConBandageOverview.aspx.
in two- and three-dimensional fibrin gels. 30. Ellis-Behnke, R. G., Liang, Y. X., Tay, D. K.,
J. Comp. Neurol. 365, 380–391. Kau, P. W., Schneider, G. E., Zhang, S., et al.
18. Sakiyama, S. E., Schense, J. C., and Hubbell, (2006) Nano hemostat solution: immediate
J. A. (1999) Incorporation of heparin-binding hemostasis at the nanoscale. Nanomedicine 2,
peptides into fibrin gels enhances neurite 207–215.
extension: an example of designer matrices 31. Leach, J. B., Brown, X. Q., Jacot, J. G., Dimilla,
in tissue engineering. FASEB J. 13, P. A., and Wong, J. Y. (2007) Neurite outgrowth
2214–2224. and branching of PC12 cells on very soft
s ubstrates sharply decreases below a threshold of biocompatibility of poly(2-hydroxyethyl

substrate rigidity. J. Neural Eng. 4, 26–34. methacrylate-co-methyl methacrylate) nerve
32. Guan, J., Stankus, J. J., and Wagner, W. R. conduits. Biomaterials 26, 1741–1749.
(2006) Development of composite porous 43. Carone, T. W. and Hasenwinkel, J. M. (2006)
scaffolds based on collagen and biodegradable Mechanical and morphological characteriza-
poly(ester urethane)urea. Cell Transpl. 15 tion of homogeneous and bilayered poly(2-
Suppl. 1, S17–S27. hydroxyethyl methacrylate) scaffolds for use
33. Burdick, J. A., Ward, M., Liang, E., Young, in CNS nerve regeneration. J. Biomed. Mater.
M. J., and Langer, R. (2006) Stimulation of Res. B, Appl. Biomater. 78, 274–282.
neurite outgrowth by neurotrophins delivered 44. Nomura, H., Katayama, Y., Shoichet, M. S.,
from degradable hydrogels. Biomaterials 27, and Tator, C. H. (2006) Complete spinal
452–459. cord transection treated by implantation of a
34. Krause, T. L. and Bittner, G. D. (1990) Rapid reinforced synthetic hydrogel channel results
morphological fusion of severed myelinated in syringomyelia and caudal migration of the
axons by polyethylene glycol. Proc. Natl Acad. rostral stump. Neurosurgery 59, 183–192;
Sci. USA 87, 1471–1475. discussion 183–192.
35. Mahoney, M. J. and Anseth, K. S. (2006) 45. Tsai, E. C., Dalton, P. D., Shoichet, M. S., and
Three-dimensional growth and function of Tator, C. H. (2004) Synthetic hydrogel guid-
neural tissue in degradable polyethylene gly- ance channels facilitate regeneration of adult rat
col hydrogels. Biomaterials 27, 2265–2274. brainstem motor axons after complete spinal
36. Piantino, J., Burdick, J. A., Goldberg, D., cord transection. J. Neurotrauma 21, 789–804.
Langer, R., and Benowitz, L. I. (2006) An 46. Tsai, E. C., Dalton, P. D., Shoichet, M. S.,
injectable, biodegradable hydrogel for trophic and Tator, C. H. (2006) Matrix inclusion
factor delivery enhances axonal rewiring and within synthetic hydrogel guidance channels
improves performance after spinal cord injury. improves specific supraspinal and local axonal
Exp. Neurol. 201, 359–367. regeneration after complete spinal cord
37. Aubert-Pouessel, A., Venier-Julienne, M. C., transection. Biomaterials 27, 519–533.
Clavreul, A., Sergent, M., Jollivet, C., Montero- 47. Newman, K. D. and McBurney, M. W. (2004)
Menei, C. N., et al. (2004) In vitro study of Poly(d,l lactic-co-glycolic acid) microspheres
GDNF release from biodegradable PLGA as biodegradable microcarriers for pluripotent
microspheres. J. Control. Release 95, 463–475. stem cells. Biomaterials 25, 5763–5771.
38. Kim, D. H. and Martin, D. C. (2006) 48. Lloyd, D. A., Ansari, T. I., Gundabolu, P.,
Sustained release of dexamethasone from Shurey, S., Maquet, V., Sibbons, P. D., et al.
hydrophilic matrices using PLGA nanoparti- (2006) A pilot study investigating a novel sub-
cles for neural drug delivery. Biomaterials 27, cutaneously implanted pre-cellularised scaffold
3031–3037. for tissue engineering of intestinal mucosa.
39. Lam, X. M., Duenas, E. T., and Cleland, J. L. Eur. Cell Mater. 11, 27–33; discussion 34.
(2001) Encapsulation and stabilization of 49. Zhang, S. (2003) Fabrication of novel bioma-
nerve growth factor into poly(lactic-co-gly- terials through molecular self-assembly. Nat.
colic) acid microspheres. J. Pharm. Sci. 90, Biotechnol. 21, 1171–1178.
1356–1365. 50. Zhang, S., Marini, D. M., Hwang, W., and
40. Patist, C. M., Mulder, M. B., Gautier, S. E., Santoso, S. (2002) Design of nanostructured
Maquet, V., Jerome, R., and Oudega, M. biological materials through self-assembly of
(2004) Freeze-dried poly(d,l-lactic acid) peptides and proteins. Curr. Opin. Chem. Biol.
macroporous guidance scaffolds impregnated 6, 865–871.
with brain-derived neurotrophic factor in the 51. Garreta, E., Genove, E., Borros, S., and Semino,
transected adult rat thoracic spinal cord. C. E. (2006) Osteogenic differentiation of mouse
Biomaterials 25, 1569–1582. embryonic stem cells and mouse embryonic
41. Teng, Y. D., Lavik, E. B., Qu, X., Park, K. I., fibroblasts in a three-dimensional self-assembling
Ourednik, J., Zurakowski, D., et al. (2002) peptide scaffold. Tissue Eng. 12, 2215–2227.
Functional recovery following traumatic spi- 52. Zhang, S., Gelain, F., and Zhao, X. (2005)
nal cord injury mediated by a unique polymer Designer self-assembling peptide nanofiber
scaffold seeded with neural stem cells. Proc. scaffolds for 3D tissue cell cultures. Semin.
Natl Acad. Sci. USA 99, 3024–3029. Cancer Biol. 15, 413–420.
42. Belkas, J. S., Munro, C. A., Shoichet, M. S., 53. Holmes, T. C. (2002) Novel peptide-based
Johnston, M., and Midha, R. (2005) Long- biomaterial scaffolds for tissue engineering.
term in vivo biomechanical properties and Trends Biotechnol. 20, 16–21.
54. Lupi, O., Madkan, V., and Tyring, S. K. factor 1 (IGF-1) delivery with biotinylated
(2006) Tropical dermatology: bacterial tropi- peptide nanofibers improves cell therapy for
cal diseases. J. Am. Acad. Dermatol. 54, myocardial infarction. Proc. Natl Acad. Sci.
559–578. USA 103, 8155–8160.
55. Ellis-Behnke, R. G., Liang, Y. X., You, S. 57. Guo, J., Su, H., Zeng, Y., Liang, Y. X., Wong, W.
W., Tay, D. K., Zhang, S., So, K. F., et al. M., Ellis-Behnke, R. G., et al. (2007) Reknitting
(2006) Nano neuro knitting: peptide nano- the injured spinal cord by self-assembling
fiber scaffold for brain repair and axon peptide nanofiber scaffold. Nanomedicine 3,
regeneration with functional return of 311–321.
vision. Proc. Natl Acad. Sci. USA 103, 58. Schneider, G. E., Ellis-Behnke, R. G., Liang,
5054–5059. Y. X., Kau, P. W., Tay, D. K., and So, K. F.
56. Davis, M. E., Hsieh, P. C., Takahashi, T., (2006) Behavioral testing and preliminary
Song, Q., Zhang, S., Kamm, R. D., et al. analysis of the hamster visual system. Nat.
(2006) Local myocardial insulin-like growth Protoc. 1, 1898–1905.
Chapter 18
Computational Simulations of the Interaction of Lipid

Membranes with DNA-Functionalized Gold Nanoparticles
One-Sun Lee and George C. Schatz
Abstract
We develop a shape-based coarse-grained (SBCG) model for DNA-functionalized gold nanoparticles
(DNA-Au NPs) and use this to study the interaction of this potential antisense therapeutic with a lipid
bilayer model of a cell membrane that is also represented using a coarse-grained model. Molecular dynam-
ics simulations of the SBCG model of the DNA-Au NP show structural properties which coincide with
our previous atomistic models of this system. The lipid membrane is composed of 30% negatively charged
lipid (1,2-dioleoyl-sn-glycero-3-phosphoserine, DOPS) and 70% neutral lipid (1,2-dioleoyl-sn-glycero-
3-phosphocholine, DOPC) in 0.15 M sodium chloride solution. Molecular dynamics (MD) simulations
of the DNA-Au NP near to the lipid bilayer show that there is a higher density of DOPS than DOPC
near to the DNA-Au NP since sodium counterions are able to have strong electrostatic interactions with
DOPS and the DNA-Au NP at the same time. Using a steered MD simulation, we show that this
counterion-mediated electrostatic interaction between DNA-Au NP and DOPS stabilizes the DNA-Au
NP in direct contact with the lipid. This provides a model for interaction of DNA-Au NPs with cell
membranes that does not require protein mediation.
Key words: DNA, Gold, Nanoparticle, Lipid, DOPS, DOPC, Molecular dynamics simulation,
Charge–charge interaction, Sodium ion
1. Introduction
The delivery of inorganic nanoparticles into cells has been one of

the most exciting recent developments in therapeutics or imaging
(1–4). Among the many different nanoparticles being considered,
gold nanoparticles have emerged as an attractive candidate for the
delivery of various drug molecules into targeted cells (5–9).
In particular, Mirkin and coworkers have developed a method
for delivering DNA or RNA-functionalized gold nanoparticles
into cells for therapeutic purposes such as antisense treatment (10).
283
284 Lee and Schatz
According to recent work in the Mirkin lab, these DNA-

functionalized gold nanoparticles (DNA-Au NPs) are readily
taken up into cells through endocytosis. However, this is some-
what surprising since the surface of the gold nanoparticle is
densely covered by negatively charged ss-DNA (~19 pmol/cm2),
and the cell membrane is also negatively charged. According to a
study by Giljohann et al., the DNA-Au NPs adsorb a large number
of positively charged serum proteins on the nanoparticle surface
(7), resulting in a zeta potential for the DNA-Au NP which is
changed from −21 ± 4 to −13 ± 1 mV after adsorption. Based on
these experiments, they proposed that the interaction of DNA-Au
NP with proteins is a possible mechanism for recognition of the
nanomaterial by the cell and subsequent cellular internalization.
However, even though this explains many aspects of cellular inter-
nalization of the DNA-Au NPs, there still remain many questions
about the interaction between these negatively charged nanopar-
ticles and the cell membrane. Indeed, even after the adsorption
of positively charged serum proteins, the DNA-Au NP still has a
negative charge. Moreover, according to recent experiments,
DNA-Au NPs are internalized by cells without serum proteins,
albeit a smaller amount of internalized nanoparticles (11).
Therefore, the fundamental nature of the interaction between
these negatively charged nanoparticles and a cell membrane has
yet to be fully elucidated, and this has proven to be a hindrance in
understanding how cells interact with nanoparticles.
Computer simulation is an emerging and promising tool for
investigating large systems composed of bio- and nanomaterials
as a result of recent progress in the development of molecular
theories and computer technologies. In our earlier work, our
group has developed an atomistic model of DNA-Au NPs which has
provided deeper understanding concerning a variety of structural
features including the effective radius of the DNA-Au NP, the
interaction between DNA strands on the surfaces of gold parti-
cles, and their local salt concentration (12–14). However, even
though our atomistic model correctly describes many details of
this system, this model is not appropriate for studying the interac-
tions between a nanoparticle and a cell membrane due to its high
demand for computational resources. Therefore, a new computa-
tional approach such as coarse-grained model is needed for the
study of DNA-Au NPs and their interactions with membranes.
Recently, the Schulten group has developed the shape-based
coarse-grained (SBCG) model of cell membranes for the study of
membrane curvature that is induced by membrane proteins. Such
models are essential for understanding endocytosis pathways that
are important for many processes in cell biology (15). The SBCG
model is simpler than previous CG lipid models in terms of the
number of beads, but it still describes the properties of mem-
branes effectively. This suggests that the SBCG lipid model will
Computational Simulations of the Interaction of Lipid Membranes 285
be appropriate for studying the interaction of DNA-Au NPs with

membranes, and since the required computational resources are
reasonable, it is a logical place to start such studies.
In Subheading 2, we develop a SBCG model for the DNA-Au
NP based on an atomistic model we previously reported (12). In
addition, we show that the properties of this newly developed
SBCG model are very close to what we obtained with our atom-
istic modeling. In Subheading 3, molecular dynamics (MD) sim-
ulations are performed to study the interaction between the
DNA-Au NPs and a lipid bilayer model of a cell membrane. In
addition, we perform steered MD simulations to provide a deeper
understanding of the interactions between the negatively charged
DNA-Au NP and the lipid membrane.
2. Development
of a SBCG Model
for DNA-
Functionalized Force field parameters for the SBCG model of the DNA-Au NP
Gold Nanoparticles (see Fig. 1) are developed based on atomistic simulations described
in our previous report (12). In this work, we considered a 2-nm
gold nanoparticle functionalized with four single stranded oligo-
nucleotides as representation of the smallest DNA-Au NP system
that has been studied in the Mirkin group. In the present applica-
tion, this is approximated by one bead for the gold nanoparticle
and two for each DNA strand (see Fig. 1c).
Fig. 1. (a) Atomistic model of a DNA-functionalized gold nanoparticle. (b) Coarse-grained

model of the same structure, with the atomistic model superimposed. (c) Name assign-
ments of the CG beads.
286 Lee and Schatz
Equations 1–3 are used for the bond, angle, and van der Waals
interaction energy calculations, respectively, the parameters for
which are listed in Table 1. Each bead for DNA (D1 and D2) in
Fig. 1 corresponds to 5.5 adenine bases, and the bead for gold
nanoparticle corresponds to 201 gold atoms. Each bead that
comprises the DNA strand has −5.5 charges, whereas the bead for
the gold nanoparticle is electrically neutral.
Ebond = kd (d − d0 )2 . (1)
Eangle = kq (q − q 0 )2 . (2)
 R  12  R 
6
ELJ = e   − 2    . (3)
 r   r  
To compare the dynamic properties of the SBCG model with

the atomistic model, we performed MD simulations of the SBCG
Table 1
Force field parameters for the CG model of DNA-Au NP
and membrane
Bond kd (kcal/mol/Å2) d0 (Å)
G–D1 10.0 28.0

D1–D2 10.0 25.0
PCH–DOT* 0.2 12.0
PSH–DOT* 0.2 12.0
Angle kq (kcal/mol/rad2) q0 (°)
D1–G–D1 200.0 109.5

G–D1–D2 100.0 140.0
Atom (LJ) e (kcal/mol) R (Å) Charge (–e)
G 20.0 18.0 0.0

D1 0.1 13.6 −5.5
D2 0.1 13.6 −5.5
PCH* 0.1 13.6 0.0
PSH* 0.1 13.6 −2.2
DOT* 10.0 13.6 0.0
NAB* 0.1 14.0 2.2
CLB* 0.1 14.0 −2.2
See Figs. 1 and 3 for the name assignment of the CG beads. *Adapted from ref. 15
model. Periodic boundary conditions were used corresponding

to a box of dimensions of 100 × 100 × 100 Å3 (16). To neutralize
the system, 20 NAB ions are added. The NAB ion corresponds
to 2.2 sodium ions and has +2.2 charges. In addition to these
sodium ions, another 123 NAB and CLB ions (CLB ion corre-
sponds to 2.2 chlorine ions, and has −2.2 charge) are added to
make the concentration of sodium ions in the box to be 0.52 M.
This choice matches what has been used previously in the atom-
istic simulations.
The system is simulated for 12 ns using the NVT ensemble
(17) and Langevin dynamics at a temperature of 300 K with a
damping coefficient of g = 2/ps (18, 19). According to the work of
Arkhipov et al. (15), a value of 2/ps for the damping coefficient
in Langevin dynamics reproduces the viscosity of water for the
coarse-graining level used in our study. No atomic coordinates
were constrained during the production period. Atomic coordi-
nates were saved every 1 ps for the trajectory analysis. MD simu-
lations were carried out using NAMD2 (20).
To compare the flexibility of the SBCG DNA-Au NP model
with the atomistic results, the distribution of the angles qG–D1–D2
and qD1–G–D1 (see Fig. 1) obtained from MD are compared with
our previous atomistic MD simulations. As shown in Table 2,
the average value of qD1–G–D1 obtained from the SBCG model is
109 ± 8°, whereas it is 108 ± 24° for the atomistic model. Also, the
average qG–D1–D2 in SBCG is 123 ± 12° and it is 127 ± 20° for
atomistic model.
To calculate the effective radius of the DNA-NPs, the radius
of gyration (RG) is introduced. The radius of gyration is defined
in Eq. 4, where ND is the number of beads in the DNA, 〈〉
denotes a time average, rD is the position vector of the Dth DNA
bead, and rG is the position vector of the center of the gold particle
bead.
N D   2
RG2 =
1
∑
N D + 1 D =0
rD − rG ( . ) (4)
Table 2
Average angle qD1–G–D1 and qG–D1–D2 obtained from a 12-ns MD
simulation for the CG model of DNA-Au NP
qD1–G–D1 (°) qG–D1–D2 (°)
CG model 109 ± 8 123 ± 12

Atomistic simulation 108 ± 24 127 ± 20
The average angle from the atomistic simulation is also shown for comparison
288 Lee and Schatz
During the last 2 ns of the simulation, the value of RG is

31 ± 1 Å, whereas it is 28.4 ± 2.3 Å in our previous simulation at the
atomistic level. All statistical uncertainties are ±1s (one standard
deviation).
We calculated the number density of sodium ions within
r = 31 Å from the center of the gold particle. This distance was
chosen to be close to RG, so the number density refers to ions that
are in the volume occupied by the DNA. The number density of
sodium ions r(r) is calculated using Eq. 5.
〈ni (ri )〉
r (r ) = . (5)
v
Here v is the volume of a sphere with a radius 31 Å minus the
volume of the gold particle. 〈 n i (ri )〉 is the number of particles
averaged over time as shown in Eq. 6, where T is the total time
for the sampling.
1 T
〈ni (ri )〉 = ∑ ni (ri (t )). (6)
T t =1
The distribution of number density of sodium atoms
(normalized to its bulk value) during the last 2 ns of the MD
simulation is shown in Fig. 2. Here, the number density of sodium
ions around the gold nanoparticle is normalized relative to that
for the 0.52 M bulk concentration. Figure 2 shows that the con-
centration of sodium ions around the gold nanoparticle is about
22% higher than the bulk concentration. This is consistent with
our previous atomistic simulation results (20%).
Fig. 2. Distribution of the relative concentration of sodium ions within 31 Å of the gold
nanoparticle in the DNA-Au NP complex. The Na+ concentration around the gold particle
is about 22% higher than the bulk concentration.
3. Nanoparticle–
Membrane
Interaction
We performed 1 ms MD simulations for the SBCG DNA-Au NP
interacting with a lipid bilayer to study the nanoparticle/mem-
brane interaction. Force field parameters for the chosen lipids are
listed in Table 1 as adapted from the work of Arkhipov et al. (15).
Two kinds of lipid molecules, 1,2-dioleoyl-sn-glycero-3-phos-
phoserine (DOPS) and 1,2-dioleoyl-sn-glycero-3-phosphocho-
line (DOPC) (21), are used in the simulations. As shown in Fig. 3,
both DOPS and DOPC are composed of two beads: one for the
head group and one for the tail group. Note that the DOPS or
DOPC correspond to 2.2 lipid molecules, and the head group of
DOPS has a −2.2 charge, whereas DOPC has zero charge (15).
The total number of lipid molecules is 2,738 (37 × 37 × 2) and
consists of 820 DOPS and 1,918 DOPC (see Fig. 3). Therefore,
~30% of our lipids have a negative charge. Periodic boundary
conditions are used, corresponding to a box of dimensions of
464 × 464 × 440 Å3. To neutralize the system, 840 NAB ions are
Fig. 3. Side and top views of coarse-grained model of a lipid bilayer composed of 70%
DOPC (PCH) and 30% DOPS (PSH). The surface of the lipid is composed of a 37 × 37 × 2
array.
290 Lee and Schatz
added. In addition to these sodium ions, another 2,714 NAB and

CLB ions are added to make the concentration of sodium ions to
be 0.15 M.
The distance between the gold nanoparticle and the surface
of the lipid is taken to be ~75 Å in the starting structure. A 1-ns
MD simulation at 5,000 K with a NVT ensemble is performed
to equilibrate the system. During the equilibration, the position
of the gold and lipid is fixed with harmonic constraints. In the pro-
duction period, the system is simulated for 1 ms using the NVT
ensemble and Langevin dynamics at a temperature of 310 K with
a damping coefficient g = 2/ps (18, 19). A time step of 100 fs is
used for the simulation. The long-range interaction cutoff is taken
to be 35 Å. No atomic coordinates were constrained during the
production period. Atomic coordinates were saved every 1 ns for
the trajectory analysis. MD simulations were carried out using
NAMD2 (20).
A snapshot of the system after a 1-ms MD simulation is shown in
Fig. 4. Even though the distance between the gold nanoparticle
and the surface of the membrane is about 75 Å at the starting
position, the DNA-Au NP then approaches to the membrane
Fig. 4. Side and top views of a DNA-Au NP on a membrane after a 1-ms MD simulation.
The DNA-Au NP is adsorbed on the membrane, but penetration of the membrane is not
observed.
such that the distance after 200 ns is ~40 Å. At this distance, the
surfaces of the DNA-Au NP and the membrane are almost touch-
ing each other as shown in Fig. 4. During the 1-ms MD simula-
tion, penetration of the DNA-Au NP through the membrane
(leading to endocytosis) is not observed.
Radial distribution functions (RDFs) between sodium ions
and the head group of DOPS or DOPC are presented in Fig. 5.
The first peak in g DOPS− Na (r) appears at ~10 Å and is very intense,
+
whereas the first peak in g DOPC − Na (r) appears at ~12 Å and is much
+
weaker. This indicates that sodium ions are localized to DOPS, as

it makes sense based on electrostatic effects. A snapshot of the
membrane, shown in Fig. 5c, shows the distribution of DOPS
and DOPC, while Fig. 5d shows the distribution of the sodium
ions superimposed on the data of Fig. 5c. Both figures show dis-
tinct dark gray patches, indicating that most of the sodium ions
are closely associated with DOPS.
RDFs between the gold nanoparticle in the DNA-Au NP and
DOPS or DOPC have been generated in order to scrutinize the
position of the DNA-Au NP during the MD simulation. Figure 6
shows the results, and we see that the first peak in gAu–DOPS(r)
appears at 47 Å, whereas the first peak in gAu–DOPC(r) appears at 52 Å
(and is less intense). Therefore, the DNA-Au NP is predominantly
Fig. 5. RDFs between (a) DOPS and sodium ion and (b) DOPC and sodium ion obtained from a 1-ms MD simulation.
Sodium ions are localized around DOPS even though DOPS is only 30% of the lipid. (c) Snapshot of a DNA-Au NP on a
membrane after 1 ms MD simulation. DOPS is shown in light gray and DOPC is shown in dark gray. (d) The distribution of
sodium ions is shown. Sodium ions are preferentially localized around the DOPS and the DNA-Au NP.
292 Lee and Schatz
Fig. 6. RDFs between the gold nanoparticle of the DNA-Au NP and (a) DOPS and (b)
DOPC during a 1-ms MD simulation. This shows that the DNA-Au NP is in closer contact
with DOPS than DOPC.
localized near DOPS even though both the DNA-Au NP and

DOPS are negatively charged and DOPS represents only 30% of
the lipid composition (15). Therefore, we conclude that sodium
ions interacting with both the DNA-Au NP and DOPS induce
the DNA-Au NP to be closer to DOPS than to DOPC.
To examine the interaction between the DNA-Au NP and
DOPS or DOPC, we performed a steered MD simulation.
A force is exerted on the gold nanoparticle of the DNA-Au NP
along the path shown in Fig. 7a. Note that the path is parallel to
the bilayer, so the effect of the steered MD is to pull the NP
through regions of varying lipid composition. The pulling velocity
is 150 Å/ms, and a time step of 50 fs is used. The interaction ener-
gies of the DNA-Au NP with the membrane, and with the sodium
and chloride ions along the path of the steered MD simulation
are shown in Fig. 7b. The total interaction energy of the DNA-Au
NP is the sum of these individual interaction energies.
E total = E NP − lipid + E NP − Na + + E NP − Cl− . (7)
Fig. 7. (a) The trajectory used for the DNA-Au NP during the steered MD simulation is
shown with an arrow. A force is exerted on the gold nanoparticle of the DNA-Au NP and
the total distance of the steered MD is 150 Å. (b) Interaction energies of the DNA-Au NP
with sodium ion, chlorine ion, and membrane during the steered MD simulation. The
total energy that is the sum of the individual interaction energies is shown in the inset.
This energy is shown in the inset of Fig. 7b. Note that all
energies in Fig. 7b are normalized relative to the absolute value of
the interaction energy of the DNA-Au NP with its environment
in 0.15 M sodium chloride solution. To obtain the normalization
factor, we performed a MD simulation for the DNA-Au NP in
0.15 M sodium chloride solution for 0.5 ms with periodic boundary
conditions of 406 × 406 × 406 Å3. The average interaction energy of
the DNA-Au NP with the sodium and chloride ions is calculated
during the last 0.1 ms of this simulation.
As shown in Fig. 7b, the interaction energy between the
DNA-Au NP and lipid (ENP–lipid) is negligible compared with that
of E NP − Na and E NP − Cl . Etotal is overall negative, meaning that
+ −
interaction of the functionalized nanoparticle with Na+ is more

294 Lee and Schatz
important than with Cl−. Its value for d = 0 is close to −1.00,
meaning that the energy at that point matches that from bulk
solution conditions and the particle is not strongly bound to
the surface. However, we see that Etotal decreases to −1.04 as the
DNA-Au NP is dragged across the surface. This means that
the DNA-Au NP is stabilized by 4% of the bulk interaction
energy as a result of moving it across the surface and exposing it
to more DOPS. For a deeper understanding of this stabiliza-
tion of the DNA-Au NP close to the membrane, the number of
DOPS molecules near the DNA-Au NP (within 35 Å) has been
calculated during the steered MD calculation. This number is
plotted in Fig. 8a, and we see that this number starts at 15 but
increases to ~20 at d = 130 Å. This happens while the interaction
energy in Fig. 7 is decreasing, so it is apparent that more DOPS
Fig. 8. (a) The number of DOPS within 35 Å of the DNA-Au NP during the steered MD
simulation is shown. (b) Snapshot from the steered MD simulation. Sodium ions (NAB)
are intercalated between the DNA-Au NP and the head of DOPS (PSH). All other lipids are
shown with a stick model for clarity.
bind to the cluster as the steered MD proceeds (somewhat like the

function of a vacuum cleaner), and this causes the functionalized
nanoparticle to be attracted to the surface. A snapshot of the
DNA-Au NP during the steered MD simulation is shown in
Fig. 8b. We see that the sodium ions are interposed between
DOPS and bead D2 of the DNA-Au NP, essentially providing the
“glue” to bind the nanoparticle to the DOPS-coated surface.
4. Conclusion
An important accomplishment of this chapter is the development

of a SBCG model for DNA-Au NP, and the demonstration of its
use for simulating the interaction of this potential therapeutic
material with a lipid bilayer composed of 30% DOPS and 70%
DOPC in 0.15 M sodium chloride solution. MD simulations for
several microseconds are possible for this system, which makes it
possible to determine the thermodynamic and structural proper-
ties, providing an explanation for why the negatively charged
functionalized nanoparticle is attracted to the negatively charged
membrane. Indeed, we find that counterion-mediated electro-
static attractive interactions are sufficient to enable the particle to
strongly bind to the negatively charged surface. This indicates
why protein mediation is not essential for getting the DNA-Au
NP to stick to the surface, but it leaves open the question of how
endocytosis works for this system. This latter point will be the
subject of future studies.
Acknowledgments
This research was supported by National Science Foundation

(grant CHE-0843832), and by the Northwestern Center for
Cancer Nanobiotechnology Excellence (1 U54 CA119341-01).
References
1. Alemdaroglu, F. E., Alemdaroglu, N. C., 3. Nel, A. E., Madler, L., Velegol, D., Xia, T.,
Langguth, P., and Hermann, A. (2008) DNA Hoek, E. M. V., Somasundaran, P., et al.
block copolymer micelles – A combinatorial (2009) Understanding biophysicochemical
tool for cancer nanotechnology. Adv. Mater. interactions at the nano-bio interface. Nat.
20, 899–902. Mater. 8, 543–557.
2. Ghosh, P., Han, G., De, M., Kim, C. K., and 4. Thurn, K. T., Brown, E. M. B., Wu, A., Vogt,
Rotello, V. M. (2008) Gold nanoparticles in S., Lai, B., Maser, J., et al. (2007) Nanoparticles
delivery applications. Adv. Drug Deliv. Rev. for applications in cellular imaging. Nanoscale
60, 1307–1315. Res. Lett. 2, 430–441.
296 Lee and Schatz
5. Chithrani, B. D. and Chan, W. C. W. (2007) 13. Lee, O.-S. and Schatz, G. C. (2009)
Elucidating the mechanism of cellular uptake Interaction between DNAs on a gold surface.
and removal of protein-coated gold nanopar- J. Phys. Chem. C 113, 15941–15947.
ticles of different sizes and shapes. Nano Lett. 14. Lee, O.-S. and Schatz, G. C. (2009) Molecular
7, 1542–1550. dynamics simulation of ds-DNA on a gold
6. Chithrani, B. D., Ghazani, A. A., and Chan, W. surface at low surface coverage. MRS Proc.
C. W. (2006) Determining the size and shape 1177, 209–233.
dependence of gold nanoparticle uptake into 15. Arkhipov, A., Yin, Y., and Schulten, K.
mammalian cells. Nano Lett. 6, 662–668. (2008) Four-scale description of membrane
7. Giljohann, D. A., Seferos, D. S., Patel, P. C., sculpting by BAR domains. Biophys. J. 95,
Millstone, J. E., Rosi, N. L., and Mirkin, C. A. 2806–2821.
(2007) Oligonucleotide loading determines 16. Allen, M. P. and Tildesley, D. J. (1987)
cellular update of DNA-modified gold nano- Computer simulation of liquids. Oxford
particles. Nano Lett. 7, 3818–3821. University Press, Inc., New York.
8. Giljohann, D. A., Seferos, D. W., Prigodich, 17. McQuarrie, D. A. (1976) Statistical mechanics.
A. E., Patel, P. C., and Mirkin, C. A. (2009) Harper Collins Publishers, New York.
Gene regulation with polyvalent siRNA-nano- 18. Feller, S. E., Zhang, Y. H., Pastor, R. W., and
particle conjugates. J. Am. Chem. Soc. 131, Brooks, B. R. (1995) Constant-pressure
2072–2073. molecular-dynamics simulation – The
9. Rosi, N. L., Giljohann, D. A., Thaxton, C. S., Langevin piston method. J. Chem. Phys. 103,
Lytton-Jean, A. K. R., Han, M. S., and Mirkin, 4613–4621.
C. A. (2006) Oligonucleotide-modified gold 19. Martyna, G. J., Tobias, D. J., and Klein, M. L.
nanoparticles for intracellular gene regulation. (1994) Constant-pressure molecular-dynamics
Science 312, 1027–1030. algorithms. J. Chem. Phys. 101, 4177–4189.
10. Patel, P. C., Giljohann, D. A., Seferos, D. S., 20. Kale, L., Skeel, R., Bhandarkar, M., Brunner,
and Mirkin, C. A. (2008) Peptide antisense R., Gursoy, A., Krawetz, N., et al. (1999)
nanoparticles. Proc. Natl. Acad. Sci. U.S.A. NAMD2: Greater scalability for parallel
105, 17222–17226. molecular dynamics. J. Comput. Phys. 151,
11. Patel, P.C. at Northwestern University (2009) 283–312.
Personal communication. 21. Becker, W. M., Reece, J. B., and Poenie, M. F.
12. Lee, O.-S. and Schatz, G. C. (2009) Molecular (1996) The world of the cell. The Benjamin/
dynamics simulation of DNA-functionalized Cummings Publishing Company, Inc., Menlo
gold nanoparticles. J. Phys. Chem. C 113, Park.
2316–2321.
Part II
Translating Nanoscience and Technology

from the Lab to the Clinic
Chapter 19
Cytotoxic Assessment of Carbon Nanotube Interaction

with Cell Cultures
Hanene Ali-Boucetta, Khuloud T. Al-Jamal, and Kostas Kostarelos
Abstract
The field of nanotoxicology recently has emerged out of the need to systematically study the biocompat-
ibility and potential adverse effects of novel nanomaterials. Carbon nanotubes (CNT) are one of the most
interesting types of nanomaterials, and recently, their use in applications has dramatically increased. Their
potential adverse impact on human health and the environment, however, have caused them to be viewed
with apprehension in certain cases so further studies into their toxicology are justified. Current method-
ologies using cell culture (in vitro) models are unreliable and are not yet able to offer conclusive results
about the toxicity profile of CNT. The need for reliable and rapid toxicity assays that will allow high
throughput screening of nanotube materials is a prerequisite for the valid assessment of CNT toxicity. The
assay described here was developed based on the pitfalls and drawbacks of traditionally used cytotoxicity
assays. A methodological description of the main problems associated with the MTT and the LDH assays
is offered to illustrate the advantages of this novel assay for the study and determination of the cytotoxic
profile of CNT. Most importantly, a thorough account of this novel assay which is considered to be rapid,
reliable, and suitable for broad-spectrum cytotoxicity screening of different types of CNT is
described.
Key words: Nanotechnology, Nanotoxicology, MTT, LDH, Fluorescence, Cell death, Apoptosis
1. Introduction
Carbon nanotubes (CNT), novel cylindrical nanostructures, have

already exceeded many expectations in terms of widespread usage
and large-scale manufacturing due to their extraordinary proper-
ties which include high electrical and thermal conductivity and
robust mechanical properties (1). CNT also can be utilized as
components in a variety of biomedical applications ranging from
probes in biosensing strategies to drug delivery vectors in thera-
peutic schemes (1–6). However, before such applications are used
299
300 Ali-Boucetta, Al-Jamal, and Kostarelos
in a clinical setting systematic toxicological assessment of CNT is

needed and warranted.
Despite the increased usage of CNT, general conclusions
about their in vitro cytotoxicity have proven difficult because of
the wide variety of available CNT materials (e.g., in terms of their
manufacturing, colloidal dispersion, and chemical purity) and the
variability in the assays used to determine their toxicity. The inter-
action between CNT and the molecules used in traditional and
well-established toxicity assays is one key factor attributing to this
variability (7–12). Most traditional assays are based on colorime-
try and fluorescence, and the CNT strongly interact with the
chromophore molecules used. CNT also intrinsically interact
through supramolecular stacking and assembly with species, such
as macromolecules (e.g., polymers, proteins, and nucleic acids)
and small molecules (e.g., doxorubicin).
Worle-Knirsch et al. (7) have previously indicated that the
MTT assay is unreliable to use with CNT due to the commonly
occurring false-positive results caused by the strong interaction
between the CNT and the insoluble formazan crystals. They sug-
gested using alternative cytotoxicity assays (such as LDH and flow
cytometry with Annexin V/PI staining) and other tetrazolium-
based assays (WST-1, INT, XTT). Casey et al. confirmed through
spectroscopic analysis that single-walled carbon nanotubes (SWNT)
interact with the dyes (Coomassie Blue, Alamar Blue, Neutral Red,
MTT, and WST-1) used in cytotoxicity assays and concluded that
most are not suitable for the quantitative toxicity assessment of
CNT (8). In an attempt to propose a reliable cytotoxicity assay that
would not rely on light absorbance, the same group described a
novel approach using the clonogenic assay (13). Although reliable
this assay is too time-consuming to allow for rapid screening. No
other report has used the clonogenic assay since this work was pub-
lished, and despite the reported inaccuracies most in vitro toxicity
studies are still carried out using colorimetry-based methodologies.
More recently, Monteiro-Riviere et al. (11) studied the reliability of
a range of widely used viability and cytotoxicity assays with many
types of nanoparticles including SWNT. They also found that
SWNT interfered to varying degrees with the results of most estab-
lished toxicity assays. The need for reliable toxicity assays that would
allow rapid screening of nanomaterials has now become a serious
obstacle toward safety validation of the myriad types of CNT as
well as other types of nanoparticles.
Here, we propose a modified version of one of the most
widely used and established cytotoxicity assays, the lactate dehy-
drogenase (LDH) assay, that circumvents all interactions between
CNT and the fluorophore molecules leading to a reliable and
technically straightforward methodological solution. A compari-
son with other established assays like the MTT and the original
Cytotoxic Assessment of Carbon Nanotube Interaction with Cell Cultures 301
LDH assay is also provided to illustrate the pitfalls and problems

with those methodologies compared to the proposed modified
LDH (mLDH) method that offers a trustworthy and reproduc-
ible determination of cellular toxicity following their interaction
with CNT.
2. Materials
2.1. CNT Preparation 1. Pristine multiwalled carbon nanotubes (MWNT) (Nanocyl,

Belgium) (see Note 1).
2. Pluronic F127 copolymer (Sigma, UK).
3. Sterile deionized water.
4. Glass vials.
5. Water bath.
6. Bath sonicator (Ultrasonic cleaner, VWR).
2.2. Cell Culture 1. Adherent cells, such as the lung epithelial cell line A549
(CCL-185, ATCC, UK), or others.
2. 0.05% Trypsin with 0.53 mM ethylenediaminetetraacetic acid
(EDTA) tetrasodium salt (Gibco, Invitrogen, UK).
3. Culture media appropriate for the cell line being studied. F12
Ham media supplemented with 10% fetal bovine serum (FBS),
50 U/mL penicillin, and 50 mg/mL streptomycin (all from
Gibco, Invitrogen, UK) was used with A549 cells.
4. 96-Well flat bottom plate (Corning Costar Corporation®, USA).
5. 10% v/v Dimethyl sulfoxide (DMSO) (>99.7%, Hybri-Max™,
sterile filtered, hybridoma tested) in complete cell culture
medium.
6. 1, 5, and 25 mL serological pipettes.
7. Incubate at 37°C with 5% CO2.
8. Trypan blue dye exclusion assay kit.
2.3. In Vitro 1. MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

Cytotoxicity Assays bromide) powder (Sigma, UK). Stored at 4°C before
reconstitution.
2.3.1. MTT Assay
2. Sterile phosphate-buffered saline (PBS) (1×) (Gibco,
Invitrogen, UK).
3. Sterile filter (0.22 mm).
4. DMSO, 100%.
5. Plate reader.
2.3.2. LDH Assay (Original 1. LDH kit: CytoTox 96 non-radioactive cytotoxicity assay
and Modified) (Promega, UK) containing substrate mix (five vials), assay
buffer (60 mL), LDH positive control (25 mL), lysis solution
(3 mL) (see Note 2), and stop solution (65 mL). Store sub-
strate mix and assay buffer at −20°C, protected from light
until use. Store LDH positive control, lysis solution, and stop
solution at 4°C.
2. 96-Well flat bottom plate (Corning Costar Corporation®, USA).
3. Phenol-free media (e.g., RPMI media) (Gibco, Invitrogen, UK).
4. 9% v/v Triton X-100.
5. Plate reader.
3. Methods
3.1. Preparation 1. In a glass vial, hydrate the Pluronic F127 to a final concentra-
of CNT Dispersions tion of 1% w/v with sterile deionized water.
2. Place the vial in a water bath (37°C) for 30–45 min or until
the Pluronic F127 flocculates disappear.
3. Disperse 1 mg/mL pristine MWNT powder in 1% (10 mg/mL)
Pluronic F127 by bath sonication for 30–45 min (see Note 3).
4. Store the stock MWNT:F127 dispersion at 4°C until further
use. When ready to use, sonicate for 15 min.
5. Use the 1% F127 stock solution for the controls in the toxi-
cological assessments. Store at 4°C until further use.
3.2. Cell Culture The following protocol describes the incubation of the
MWNT:F127 dispersions and control samples with A549 lung
epithelial cell lines. If desired, the cells can be treated with various
inhibitors and the incubation times of the samples with the cells
and CNT concentration in the samples can be varied. DMSO is
used as a positive control for cytotoxicity.
1. Passage A549 cells when they reach 70–80% confluency to main-
tain exponential growth. Use for a maximum of ten passages.
2. To trypsinize the monolayer, rinse with 1× PBS then incubate
with trypsin–EDTA at 37°C for 5 min. Detach the cells by
vigorous up and down pipetting.
3. Centrifuge the cells at 240 × g for 5 min at 4°C and resuspend
in complete media.
4. Count cells and determine cell viability by Trypan blue dye
exclusion assay.
5. Seed 10,000 cells per well (150 mL/well) in a 96-well plate
and incubate for 24 h at 37°C in a humidified atmosphere
(5% CO2) incubator.
6. Dilute the stock MWNT:F127 dispersion or stock 1% F127

in complete cell culture medium to reach the desired concen-
trations (see Note 4).
7. Sonicate the diluted MWNT:F127 dispersion in media for
2 min prior to its addition to cells.
8. Incubate the cells with the CNT dispersions or control
samples (1% F127 or 10% DMSO) for 24 h at 37°C in a
humidified atmosphere (5% CO2) incubator. Also, leave some
cells untreated as controls.
9. Proceed with the cytotoxicity assessments (see Subheading 3.3).
3.3. In Vitro The colorimetric MTT assay is used to measure cell viability (14).
Cytotoxicity Assays The yellow tetrazolium salt (MTT) is reduced by mitochondrial
reductase in living and metabolically active cells to purple, water-
3.3.1. MTT Assay
insoluble formazan crystals, which can then be dispersed using
DMSO or other detergents. A decrease in absorbance at 570 nm
compared to untreated control cells is then a measure of the cell
viability or the amount of apoptosis or necrosis that has been
caused by the test material (see Fig. 1).
1. Aspirate the media after the incubation period is over.
2. Prepare the MTT solution by reconstituting the MTT pow-
der in 1× sterile PBS to a final concentration of 5 mg/mL and
or
A549 Yellow MTT
or DMSO 10%
cells
24 h
Aspirate media
MTT solution
3½h
DMSO
solubilization
Purple
Formazan
Absorbance 570 nm
Fig. 1. Schematic of the MTT assay.
subsequently filter and sterilize this solution using a 0.22-mm

sterile filter. Store in 2 mL aliquots at −20°C protected from
light until use (stable for at least 6 months after
reconstitution).
3. Dilute the MTT solution with complete media containing FBS
(at a ratio of 1:6) and add 120 mL of this solution to each well.
4. Incubate the cells for 3.5 h at 37°C in a humidified atmosphere
(5% CO2) incubator.
5. Remove the MTT solution by gently inverting the 96-well
plate into a paper tissue.
6. Add 150 mL DMSO (100%) to solubilize the formazan
crystals and incubate the plate for 15 min at 37°C to remove
air bubbles.
7. Read the absorbance at 570 nm in a plate reader and express
the results as the percentage cell viability (n = 8 ± S.D.) com-
pared to untreated control cells (see Fig. 2 and Note 5). The
percentage cell viability is calculated using this formula:
A 570 nm of treated cells
% Cell viability = × 100.
A 570 nm of untreated cells
3.3.2. Original LDH Assay LDH is a stable cytosolic enzyme that is released from the cell
(See Note 6) upon cell lysis. The LDH assay is based on quantitatively measur-
ing released LDH using a coupled enzymatic assay, in which LDH
plays a role in the conversion of a tetrazolium salt (INT) into a
red soluble formazan product which then can be measured colo-
rimetrically. The amount of LDH released is proportional to the
number of lysed cells (15) (see Figs. 3 and 4).
1. Transfer 50 mL media containing released LDH from all wells
into a fresh 96-well plate.
2. For maximum LDH release: Add 10 mL lysis solution (10×)
for every 100 mL of fresh media. Keep the cells after treat-
ment at 37°C for 45–60 min.
3. Centrifuge the plate at 240 × g for 4 min and supernatant into
the fresh 96-well plate.
4. Thaw the assay buffer and warm to room temperature, while
keeping it protected from light.
5. Transfer 12 mL assay buffer into one vial of substrate mix.
Gently mix to dissolve the substrate mix, while keeping it
protected from light. Both the assay buffer and non-used
reconstituted substrate mix can be stored again at −20°C (see
Note 7).
6. Add 50 mL reconstituted substrate mix to each well containing
the transferred aliquots. Cover the plate with foil and incubate
for 30 min at room temperature.
a 120
100
Percentage Cell Viability 80
60
40
20
0
0 DMSO 10% 1.9µg/ml 7.8µg/ml 31.25µg/ml 125µg/ml
MWNT:F127 Pluronic F127
b 3
2.5
Absorbance @ 570 nm
1.5
0.5
0
0 0.97 3.9 15.6 62.5 250
MWNT concentration (µg/ml)
Fig. 2. (a) Percentage cell viability of A549 assessed by the MTT assay for varying MWNT and Pluronic F127 concentra-
tions. DMSO is used as a positive control; untreated cells are the negative control (0). The MWNT:F127 dispersions show
concentration-dependent toxicity after 24 h exposure. The values listed are the concentrations of MWNT in solution, the
corresponding F127 control for each concentration is at a concentration ten times higher (e.g., 1.9 mg/mL MWNT, 19 mg/
mL F127) (see Note 4). (b) Absorbance (at 570 nm) of insoluble formazan mixed with MWNT:F127 dispersions. The spik-
ing experiment shows that the intrinsic absorbance of MWNT can interfere with the results of the MTT assay.
7. Add 50 mL stop solution to each well and pop any large bubbles
using a syringe needle.
8. Read the absorbance of the solutions at 490 nm in a plate
reader and express the results as the percentage LDH released
(n = 4 ± S.D.) compared to maximum LDH released from the
untreated control cells (see Fig. 5). The percentage LDH
released (% cytotoxicity) is calculated using this formula:
A490 nm of treated and untreated

cells − A490 nm of media alone
% LDH released = × 100.
A490 nm of maximum of untreated
cells − A490 nm of media alone
A549 or
cells or DMSO 10%
Damaged cells after treatment
24 h
50 µL cell lysate is added

into a new 96 w/p
LDH
50 µL substrate mix added
30 min @ 37 8C
Detailed
reaction
50 µL stop solution in Fig. 4
added
Absorbance 490 nm
Red Formazan
Fig. 3. Schematic of the original LDH assay.
NAD+ Formazan
Lactate
LDH Diaphorase
NADH
Pyruvate
Iodonitrotetrazolium
(INT)
Fig. 4. LDH-mediated conversion of the INT salt into formazan.

120
Percentage Cell Survival

100
80
60
40
20
0
0 DMSO 1.9 7.8 31.25 125 125
10%
24hrs 48hrs
MWNT:F127 F127
Fig. 5. Percentage cell survival with varying concentrations of MWNT:F127 and Pluronic
F127. The values listed are the concentrations of MWNT in solution (in mg/mL), the cor-
responding F127 control for each concentration is at a concentration ten times higher
(e.g., 1.9 mg/mL MWNT, 19 mg/mL F127) (see Note 4). DMSO was used as a positive
control; untreated cells (0) were the negative control. MWNT:F127 showed a clear dose-
dependent toxicity after 24 h of exposure, which was potentiated after 48 h due to
Pluronic F127 toxicity. The Pluronic F127 did not, however, cause any cytotoxicity after
24 h incubation at the concentrations used.
The absorbance data at 490 nm also can be shown without

converting it into percentage LDH release to highlight the
interference of CNT with the results of the assay (see Fig. 6
and Note 8).
3.3.3. The “Modified LDH” The original colorimetric LDH assay was modified to avoid inter-
Assay ference of the components used in the assay with the CNT. The
survived cells after treatment are artificially lysed with Triton
X-100, and the cell lysate is centrifuged in order to precipitate the CNT.
The released LDH is therefore an indication of the number of
viable cells that survived treatment with CNT (see Figs. 4 and 7).
1. Replace the media with 100 mL per well phenol and serum-
free media (RPMI, phenol-free media) (see Note 6).
2. Add 10 mL 9% v/v Triton X-100 per 100 mL added phenol
and serum-free media.
3. Incubate the plate at 37°C for 45–60 min (see Note 9).
4. Transfer the cell lysate into tubes and centrifuge at 16,000 × g
for 5 min to pellet the uptaken CNT (see Note 10).
5. Transfer 50 mL cell lysate, avoiding the CNT pellet, into a
fresh 96-well plate.
6. Add 50 mL reconstituted substrate mix to each well containing
the transferred and centrifuged cell lysate. Cover the plate
1.2
Absorbance @ 490 nm
0.8
0.6
0.4
0.2
0
0 DMSO 10 % 7.8µg/ml 31.25µg/ml 125µg/ml
LDH :MWNT-F127 MWNT-F127 alone F127
Fig. 6. Percentage LDH release (absorbance at 490 nm) after treatment of cells with
different concentrations of MWNT:F127 and Pluronic F127. The values listed are the con-
centrations of MWNT in solution (in mg/mL), the corresponding F127 control for each
concentration is at a concentration ten times higher (e.g., 7.8 mg/mL MWNT, 78 mg/mL
F127) (see Note 4). MWNT:F127 dispersions (no assay) were used as controls (no
assay). The absorbance of the released LDH in the MWNT-treated wells (LDH:MWNT:F127)
is identical to the intrinsic absorbance of the MWNT:F127 which indicates that the
observed LDH readings are attributed in part to the intrinsic absorbance of CNT. The LDH
enzyme might also be inhibited by the presence of the positive control (DMSO 10%) as it
shows low absorbance compared to the untreated control.
or Lysed (Survived)
A549 cells
cells or DMSO 10%
24 h
Aspirate the media
Lyse the cells (lysis solution)
Transfer to eppondorfs LDH

Centrifuge to pellet the CNT
50 µL cell lysate is added Detailed

into a new 96 w/p reaction in
Fig. 4
50 µL substrate mix added
15 min @ 37 8C
50 µL stop solution
added
Red Formazan
Absorbance 490 nm
Fig. 7. Schematic of the modified LDH assay.

140
120
% Cell Viability/ Survival

100
80
60
40
20
0
5%
l
%
M
M
M
%
tro
M
m
25
10
m
m
m
20
m
on
SO
75
1.
5
12
24
03
06
48
SO
SO
C
01
00
M
SO
0.
0.
0.
0.
0.
0.
M
D
0.
D
D
MTT Modified LDH
Fig. 8. Different concentrations of cationic liposomes (0.0075–0.48 mM) and DMSO

(1.25–20%) were used to assess the reliability of the modified LDH version compared
to the MTT assay. Very similar toxicity patterns were observed with both assays,
emphasizing the reliability of using the modified LDH assay for the in vitro cytotoxicity
assessment of CNT.
with foil and incubate for 15 min at room temperature.

(Follow steps 4 and 5 from Subheading 3.3.2 for the prepara-
tion of the reconstituted substrate mix).
7. Add 50 mL stop solution to each well and pop any large bubbles
using a syringe needle.
8. Read the absorbance at 490 nm in a plate reader and express
the results as the percentage cell survival (n = 4 ± S.D.)
compared to untreated control cells (see Fig. 8 and Note 11).
The percentage cell survival is calculated using this formula:
A 490 nm of treated cells
% Cell survival = × 100.
A 490 nm of untreated cells
4. Notes
1. Noncovalently functionalized CNT are used as an example.

If desired, different dispersing agents or CNT that are modi-
fied with different surface molecules also can be used. If cova-
lently functionalized CNT are used, disperse them in 5%
dextrose by bath sonication for 30–45 min and store at 4°C
until further use.
2. 9% v/v Triton X-100 in deionized water can also be used as a
lysis solution.
3. A well-dispersed sample of MWNT should contain no pre-
cipitates or aggregates. If desired, SWNT can be used, however,
they are not as easily dispersed in F127 as MWNT; longer

sonication times may be required. Concentrations higher
than 1 mg/mL MWNT are not easily dispersed.
4. We used MWNT:F127 dispersions with MWNT concentrations
ranging from 0 to 125 mg/mL. As you dilute the MWNT:F127
stock in cell culture media, the concentration of the Pluronic
F127 is also diluted. So, in these samples, the F127 concen-
tration varies as the CNT concentration is varied. As a result,
the corresponding F127 controls should have concentrations
in the range between 0 and 1,250 mg/mL.
5. The concentration-dependent decrease in cell viability (see
Fig. 2a) should be viewed with apprehension. According to
published data, the MWNT might adsorb to the formazan
crystals (through a strong, p–p stacking interaction) and not
allow them to dissolve when the solubilizing agent (DMSO)
is added. This would cause a falsely low reading in the absor-
bance at 570 nm and falsely low percentage cell viability.
Further, we believe that the intrinsic absorbance of CNT can
alter the results of the assay when high cellular uptake and
internalization of CNT occurs. We spiked the insoluble for-
mazan with different concentrations of the MWNT:F127
dispersion and read the absorbance at 570 nm (see Fig. 2b).
An increase in absorbance was observed as the concentration
of MWNT increased indicating that the intrinsic absorbance
of the MWNT is also contributing significantly to the absor-
bance at 570 nm. This effect results in falsely high cell viabil-
ity, explaining why during many assays cell viability over
100% is obtained. The systematic effects of these two differ-
ent types of interference can have the opposite effect on the
results of the assay causing unreliable and irreproducible
results. In conclusion, the use of the MTT assay for the
assessment of CNT cytotoxicity should be avoided due to
this unpredictable balance between false-positive and -nega-
tive readings.
6. Media containing phenol and serum (FBS) can contribute to
background absorbance. This background should be sub-
tracted from all results before calculating the percentage
LDH released. In order to reduce this background without
affecting cell viability, use phenol-free media with a reduced
serum concentration (5%). If a positive control is desired for
this LDH assay, gently vortex the LDH positive control and
mix 2 mL into 10 mL PBS + 1% bovine serum albumin (BSA)
(1:5,000 dilution). This stock should be prepared fresh before
each use and triplicate or quadruplicate wells are recom-
mended. This control should have an absorbance of 1.39 ± 25%
at 490 nm.
7. Reconstituted substrate mix can be stored for 6–8 weeks
at − 20°C without loss of activity. Upon storage, a precipitate
may occur in the assay buffer which can be removed by cen-

trifugation at 300 × g for 5 min and does not affect the assay
performance (15).
8. The original LDH assay, similar to the MTT assay, is a colo-
rimetry-based method therefore significant CNT interference
is possible. As can be seen from Fig. 6, the media containing
the released LDH showed exactly the same absorbance at
490 nm as MWNT:F127 dispersions diluted in media.
Therefore, it is not possible to attribute the trend seen in
Fig. 6 (as the concentration of the MWNT was increased,
there was an increase in the LDH released) to actual cytotox-
icity. Note that the positive control (DMSO 10%) showed
low LDH release (low absorbance) compared to the other
controls. This could be due to the inhibition of LDH enzyme
by the DMSO in the media.
9. This step can be replaced by a freeze–thawing cycle. Incubate
the plate at −70°C for approximately 30 min followed by
thawing at 37°C for 15 min and then proceed to step 4.
10. It can be difficult to precipitate the CNT from the media due
to their high dispersability. If necessary, the centrifugation
time can be increased depending on the amount of CNT
uptaken by the cells. A discernible pellet should be observed
at the end of the centrifugation step. The centrifugation speed
does not seem to affect the release of LDH over a range of
speed from 300 to 16,000 × g. Centrifugation at 4°C is pref-
erable since the LDH enzyme is stable at 4°C.
11. The toxicity of cationic liposomes, which do not interact with
the chemicals used in such assays, was analyzed using this
modified protocol and the MTT assay. Results of both assays
showed the same trends, proving that the modified version of
the LDH assay is in fact promising and reliable (see Fig. 8).
Acknowledgments
This work was partially supported by the European Union FP7

ANTICARB (HEALTH-2007-201587) program. H.A.-B.
wishes to acknowledge the Ministére de l’Enseignement Supèrieur
et de la Recherche Scientifique (Algeria) for a full Ph.D.
scholarship.
References
1. Kostarelos, K., Bianco, A., and Prato, M. 2. Bianco, A. and Prato, M. (2003) Can carbon
(2009) Promises, facts and challenges for car- nanotubes be considered useful tools for bio-
bon nanotubes in imaging and therapeutics. logical applications? Adv. Mater. 15,
Nat. Nanotechnol. 4, 627–633. 1765–1768.
3. Prato, M., Kostarelos, K., and Bianco, A. Probing the interaction of single walled
(2008) Functionalized carbon nanotubes in carbon nanotubes within cell culture medium
drug design and discovery. Acc. Chem. Res. as a precursor to toxicity testing. Carbon 45,
41, 60–68. 34–40.
4. Ali-Boucetta, H., Al-Jamal, K. T., McCarthy, 10. Monteiro-Riviere, N. A. and Inman, A. O.
D., Prato, M., Bianco, A., and Kostarelos, K. (2006) Challenges for assessing carbon nano-
(2008) Multiwalled carbon nanotube-doxo- material toxicity to the skin. Carbon 44,
rubicin supramolecular complexes for cancer 1070–1078.
therapeutics. Chem. Commun. 4, 459–461. 11. Monteiro-Riviere, N. A., Inman, A. O., and
5. Podesta, J. E., Al-Jamal, K. T., Herrero, M. Zhang, L. W. (2009) Limitations and relative
A., Tian, B., Ali-Boucetta, H., Hegde, V., utility of screening assays to assess engineered
et al. (2009) Antitumor activity and pro- nanoparticle toxicity in a human cell line.
longed survival by carbon-nanotube-mediated Toxicol. Appl. Pharmacol. 234, 222–235.
therapeutic siRNA silencing in a human lung 12. Davoren, M., Herzog, E., Casey, A.,
xenograft model. Small 5, 1176–1185. Cottineau, B., Chambers, G., Bryne, H. J.,
6. Pastorin, G., Wu, W., Wieckowski, S., Briand, et al. (2007) In vitro toxicity evaluation of
J. -P., Kostarelos, K., Prato, M., et al. (2006) single walled carbon nanotubes on human
Double functionalization of carbon nanotubes A549 lung cells. Toxicol. In Vitro 21,
for multimodal drug delivery. Chem. Commun. 438–448.
11, 1182–1184. 13. Herzog, E., Casey, A., Lyng, F. M., Chambers,
7. Worle-Knirsch, J. M., Pulskamp, K., and G., Byrne, H. J., and Davoren, M. (2007) A
Krug, H. F. (2006) Oops they did it again! new approach to the toxicity testing of car-
Carbon nanotubes hoax scientists in viability bon-based nanomaterials – The clonogenic
assays. Nano Lett. 6, 1261–1268. assay. Toxicol. Lett. 174, 49–60.
8. Casey, A., Herzog, E., Davoren, M., Lyng, F. 14. Mosmann, T. (1983) Rapid colorimetric assay
M., Byrne, H. J., and Chambers, G. (2007) for cellular growth and survival: Application to
Spectroscopic analysis confirms the interac- proliferation and cytotoxicity assays. J. Immunol.
tions between single walled carbon nanotubes Methods 65, 55–63.
and various dyes commonly used to assess 15. CytoTox 96 Non-Radioactive Cytotoxicity Assay-
cytotoxicity. Carbon 45, 1425–1432. Technical Bulletin. http://www.promega.
9. Casey, A., Davoren, M., Herzog, E., Lyng, F. com/catalog/catalogproducts.aspx?categoryn
M., Byrne, H. J., and Chambers, G. (2007) ame=productleaf_405. Accessed 2010.
Chapter 20
Nanoparticle Toxicology: Measurements of Pulmonary

Hazard Effects Following Exposures to Nanoparticles
Christie M. Sayes, Kenneth L. Reed, and David B. Warheit
Abstract
Health risks following exposures to nanoparticle types are dependent upon two primary factors, namely,
hazard and exposure potential. This chapter describes a pulmonary bioassay methodology for assessing
the hazardous effects of nanoparticulates in rats following intratracheal instillation exposures; these pul-
monary exposures are utilized as surrogates for the more physiologically relevant inhalation route of
exposure. The fundamental features of this pulmonary bioassay are dose–response evaluations and time-
course assessments to determine the sustainability of any observed effect. Thus, the major endpoints of
this assay are the following: (1) time course and dose–response intensity of pulmonary inflammation and
cytotoxicity, (2) airway and lung parenchymal cell proliferation, and (3) histopathological evaluation of
lung tissue. This assay can be performed using particles in the fine (pigmentary) or ultrafine (nano) size
regimes.
In this assay, rats are exposed to selected concentrations of particle solutions or suspensions and lung
effects are evaluated at 24 h, 1 week, 1 month, and 3 months postinstillation exposure. Cells and fluids
from groups of particle-exposed animals and control animals are recovered by bronchoalveolar lavage
(BAL) and evaluated for inflammatory and cytotoxic endpoints. This protocol also describes the lung
tissue preparation and histopathological analysis of the lung tissue of particle-instilled rats. This assay
demonstrates that instillation exposures of particles produce effects similar to those previously measured
in inhalation studies of the same particulates.
Key words: Pulmonary toxicity, Particulate materials, Particles, Fine particles, Ultrafine particles,
Nanoparticles, Nanomaterials, In vivo, Rat, Lung, Intratracheal instillation, Lung hazards, Pulmonary
bioassay
1. Introduction
There is a great need for the development of rapid and reliable

short-term bioassays to evaluate the pulmonary toxicity of novel
nanoparticles and materials. Although inhalation is the only rele-
vant physiological route of administration that simulates human
313
314 Sayes, Reed, and Warheit
exposure, it is in most cases not a practical method for conducting

lung hazard studies. One critical requirement of an inhalation
bioassay is the use of adequate exposures to ensure that sufficient
amounts of material deposit in the lung so that pulmonary effects
can be appropriately assessed. If sufficient amounts of material are
not available, then predictive studies via inhalation are of little
value. Also, it is not a practical way to examine the effects of mul-
tiple particulates simultaneously.
To this end, several investigators have proposed methods of
administration using intratracheal instillations of particles into the
lungs of experimental animals to simulate high-dose exposures
(1–4). Intratracheal instillation procedures allow for a direct route
of entry and exposure to the lung using a small amount of mate-
rial in the liquid phase. This route of exposure has some practical
advantages relative to inhalation studies because the methodol-
ogy is (1) relatively inexpensive, (2) simple to implement, and (3)
easily permits the instillation of a sufficient, quantifiable bolus of
particle suspension to reach target tissues in the respiratory tract.
This third advantage is of considerable importance when testing
particles or materials in the nanoscale-sized regime. Researchers
must employ a method of exposing animals to small amounts of
nanoparticles, most often in suspension or solution because syn-
thesizing and producing new nanoparticle types on a large scale
requires a nontrivial amount of effort.
This chapter describes an intratracheal instillation bioassay
performed in rats that can be used to predict the potential for
inhaled particles to produce lung hazard effects, such as pulmo-
nary fibrosis and sustained inflammation in exposed humans. We
postulate that the mechanisms related to particle-induced pulmo-
nary disease are dependent upon three interdependent general
factors: (1) the tendency of inhaled particles or materials to cause
lung cell injury, measured by general cytotoxicity, (2) the affinity
for inhaled particles to produce ongoing inflammation, and (3)
the reduction of pulmonary macrophage clearance. The occur-
rence of these three factors has been documented individually or
in various combinations in previous reports of chronic lung dis-
ease with fibrosis (5–7). Thus, the major endpoints of this study
are the following: (1) time course and dose–response intensity of
pulmonary inflammation and cytotoxicity, (2) airway and lung
parenchymal cell proliferation, and (3) histopathological evalua-
tion of lung tissue. This method utilizes both dose and time-
course variations in order to determine the range of hazard
potentials as well as particle clearance assessments. The multidis-
ciplinary approach of this bioassay/screen is considered to be an
important component in the accuracy of its predictions of pulmo-
nary toxicity and for the investigation of clearance and inflamma-
tory mechanisms. It is also expected that the results from the
bioassay will provide valuable level-setting information for sub-
chronic and chronic inhalation studies.
Nanoparticle Toxicology: Measurements of Pulmonary Hazard Effects 315
2. Materials
2.1. Intratracheal 1. Animals: Groups of male Crl:CD®(SD)IGS BR rats (Charles

Instillation Exposure River Laboratories, Inc., Raleigh, NC). Five animals/particu-
late-type or control group/dose/time point. The rats should
be approximately 8 weeks old at study start (mean weights
between 240 and 255 g).
2. Particle suspensions in phosphate-buffered saline (PBS) solu-
tion (purchased from Sigma in a packet, Ca, Mg-free pH = 7.4,
instillation volume = 0.5 mL): (a) particle of interest, (b) car-
bonyl iron powder (metallic iron, as a negative control, BASF
Chemical Company (Wayne, NJ) at >99.5% purity), (c) crys-
talline silica (Min-U-Sil 5, a-quartz, as a positive control (5,
8, 9), US Silica Co. (Berkeley Springs, WV) at >99% purity).
3. Halothane anesthetic.
4. Disposable animal feeding needle (20 gauge – 1.5 in.) modi-
fied from a gavage needle.
2.2. Pulmonary Lavage 1. B-Euthanasia-D (sodium pentobarbital, 390 mg/mL)

and Cell Differentials (Intervet/Schering-Plough Animal Health).
2. Scissors.
3. Forceps.
4. Tubing adapter.
5. Suture material.
6. Syringes.
7. PBS, pH = 7.4.
8. Trypan blue filtered using a 0.2-mm Nalgene syringe filter –
25 mm surfactant-free cellulose acetate filter. This dye is fil-
tered once and prepared in a 1:1 ratio with lavagate containing
cells.
9. Hemocytometer.
10. Microscope slides.
11. Diff-Quick staining solutions.
12. Optical microscope.
2.3. Biochemical 1. Olympus AU640 Chemistry Analyzer.

Assays on 2. Olympus® reagents for analyses of bronchoalveolar lavage
Bronchoalveolar (BAL) fluid lactate dehydrogenase (LDH) and alkaline phos-
Lavage Fluid phatase (AlkP).
3. Reagent kit based on Coomassie blue dye binding (QuanTtest,
Quantimetrix, Hawthorne, CA) for measuring lavage fluid
protein.
2.4. Pulmonary Cell 1. 5-Bromo-2¢deoxyuridine (BrdU): 100 mg/kg body weight.

Proliferation Studies 2. PBS solution, pH = 7.4.
and Histopathological
3. Scissors.
Evaluations
4. Forceps.
5. 10% Formalin.
6. Euthanasia-B solution (0.3–0.5 mL depending on body
weight).
7. Anti-BrdU antibody with a 3-amino-9-ethyl carbazole (AEC)
marker. 100 mg/kg body weight.
8. Hematoxylin.
9. 70% Ethanol.
10. Optical microscope.
11. Preservation jars.
12. Tissue monocassettes.
13. Paraffin.
14. Microtome.
15. Microscope slides.
16. H&E staining solutions.
17. Butterfly catheter (Abbott Labs, North Chicago, IL).
3. Methods
3.1. Intratracheal 1. Prepare all particles in 0.5 mL PBS at a concentration of 1 or

Instillation Exposure 5 mg/kg (see Note 1).
(See Ref. 10) 2. Anesthetize the rats with halothane in a jar and watch for
slow respiration. Gauge level of anesthesia by assessing reflexes
(see Note 2).
3. For the BAL studies and lung tissue studies, intratracheally
instill groups of rats (five rats/group/dose/time point) with
single doses of either 1 or 5 mg/kg of the particle type of
interest, benchmark control particles [i.e., crystalline silica
(a-quartz) (positive control) or carbonyl iron particles (nega-
tive control)], or PBS solution (vehicle control) (see Table 1
and Note 3). Use the same rats for the lung cell proliferation
and histopathology studies, but different rats for the BAL
studies.
4. Evaluate the lungs (see Subheadings 3.2–3.5) of sham and
particle-exposed rats at 24 h, 1 week, 1 month, and 3 months
postinstillation (after a recovery period) (see Note 4).
Table 1
Protocol for intratracheal instillation particle bioassay study
Exposure Groups
• PBS (vehicle control)
• Particles (1 and 5 mg/kg)
o α-Quartz particles (positive control)
o Carbonyl iron (negative control)
o Other particles of interest
Instillation
Exposure
Post-instillation evaluation via BAL and lung tissue
24 h 1 wk 1 mo 3 mo
3.2. Pulmonary Lavage 1. Euthanize the rats via intraperitoneal injections of Euthanasia-B
and Cell Differentials at 0.1 mL/300 g of animal.
(See Figs. 1 and 2) 2. Lavage the lungs, trachea, and airways of particle-exposed
and sham rats with PBS solution warmed to 37°C.
3. Carry out this BAL procedure 2–3 times or until 12 mL of
fluid is collected from each animal.
4. Centrifuge lavaged fluids at 700 × g (1,800 rpm) to collect the
cells in a pellet.
5. Remove 2 mL of the supernatant, store it on ice for bio-
chemical analysis (see Subheading 3.3), and decant the
remaining supernatant (see Note 5).
6. Resuspend the cell pellet in 5 mL of cold (4°C) PBS.
7. Add filtered trypan blue to an equal volume of resuspended
lavage fluid and quantify cell numbers and viabilities using a
hemocytometer (see Note 6).
8. Stain cytocentrifuge preparations using Diff-Quick and per-
form cell differential counts using light microscopy with a
minimum of 500 cells per slide. Identify, quantify, and cate-
gorize the cells according to the following cell-types: mac-
rophages, neutrophils, lymphocytes, and eosinophils.
24 hours
100 1 week
1 month 2.0x107
3 months
75
* *
* 1.5x107 *
Total Cells
% PMN s
50
* 1.0x107
* **
25 5.0x106
0 0.0
PBS Carbonyl Crystalline Crystalline PBS Carbonyl Crystalline Crystalline
Iron Silica Silica Iron Silica Silica
(5 mg/kg) (1 mg/kg) (5 mg/kg) (5 mg/kg) (1 mg/kg) (5 mg/kg)
Fig. 1. Left : Pulmonary inflammation in sham and particle-exposed rats as evidenced by % neutrophils (PMN) in BAL
fluids at 24 h, 1 week, 1 month, and 3 months postinstillation. Intratracheal instillation exposures of the carbonyl iron
particles produced a short-term, pulmonary inflammatory response, as evidenced by an increase in the percentages/
numbers of BAL-recovered neutrophils, measured at the 24-h time point. However, the exposures to crystalline silica
(a-quartz) particles (1 and 5 mg/kg) produced sustained pulmonary inflammatory responses, as measured through 3
months postexposure (*p < 0.05 vs. PBS controls). Right : Numbers of cells recovered in BAL fluids from sham and parti-
cle-exposed rats. The numbers of cells recovered by BAL from the lungs of high-dose crystalline silica (a-quartz) exposed
(5 mg/kg) groups were higher than any of the other groups for all postinstillation time points, indicating pulmonary
inflammation. In both figures, all values are given as means ± a standard deviation.
3.3. Biochemical 1. Perform all biochemical assays on BAL fluids at 30°C using a
Assays on semiautomated clinical chemical analyzer.
Bronchoalveolar 2. Measure LDH and AlkP activity using commercially available
Lavage Fluid reagent kits (see Note 7).
(See Fig. 3)
3. Measure lavage fluid protein using a commercially available
reagent kit (see Note 8).
3.4. Pulmonary Cell 1. After the desired recovery period, intraperitoneally inject
Proliferation Studies groups of sham and particle-exposed rats with BrdU (100 mg/
and Histopathological kg body weight in PBS); inject 5 mL/kg body weight.
Evaluations Euthanize the animals 6 h later with an intraperitoneal injec-
(See Fig. 4) tion of Euthansia-B; this is referred to as a “6-h pulse.” BrdU
labels all dividing cells during the 6-h pulse period.
2. Following cessation of spontaneous respiration (within
1–3 min), fix the lungs of the rats either through the vascula-
ture (vascular perfusion) or airway (intratracheal infusion)
(11). In both cases, exsanguinate (bleed out) the animal to
reduce artifacts, expose its trachea, and clamp it with a hemo-
stat to prevent lung collapse. For intratracheal fixation, make
a small incision below the clamp and secure a 19-gauge but-
terfly catheter into the trachea. Connect the catheter to a res-
ervoir (containing a neutral buffered, 10% formalin fixative)
Fig. 2. Cytocentrifuge preparation of lavaged cells recovered from a rat exposed to 5 mg/kg of (a) carbonyl iron particles
1 week postinstillation exposure, (b) carbonyl iron particles 3 months postinstillation exposure, (c) crystalline silica (Min-
U-Sil, a-quartz) particles 1 week postinstillation exposure, and (d) crystalline silica (Min-U-Sil, a-quartz) particles
3 months postinstillation exposure, demonstrating the sustainability of the a-quartz-induced pulmonary inflammatory
responses. Arrows indicate neutrophil recruitment (magnification = 200×).
located 15 cm above the thorax of the animal and infuse the
lungs at 21 cm H2O.
3. After 15 min of fixation, carefully remove the heart and lungs
together and immersion-fix them in 10% formalin.
4. In addition, remove a 1-cm piece of duodenum (which serves
as a positive labeling tissue control) and store it in 10% forma-
lin. The duodenum has a high cell proliferation rate and is
utilized as a positive control tissue.
5. Subsequently, dehydrate the lung lobes, heart, and duode-
num in 70% ethanol and then weigh the lungs (lung weight is
a potential indicator of lung fibrotic responses). Section for
histology.
6. For cell proliferation analyses, embed tissue sections from the
right cranial, caudal, and left lobes of the lung, as well as duo-
denal sections, in paraffin using a tissue monocassette, cut
them using a microtome and mount them on glass slides.
24 hours
1 week
1 month
500
3 months
400
*
LDH (U/L)
300
200
100
150
125
AIKP (U/L)
100
75
50
25
100
75 *
MTP (mg/dL)
50
*
25
0
PBS Carbonyl Crystalline Crystalline
Iron Silica Silica
(5 mg/kg) (1 mg/kg) (5 mg/kg)
Fig. 3. BAL fluid (top) lactate dehydrogenase (LDH), (middle) alkaline phosphatase (AlkP),
and (bottom) micrototal protein (MTP) values for sham and particle-instilled rats at 24 h,
1 week, 1 month, and 3 months postinstillation. Values given are means ± a standard
deviation. Exposures to carbonyl iron did not produce any differences in the LDH, AlkP,
or MTP values when compared to the PBS controls. Further, no significant increases in
BAL fluid AlkP values are measured in any groups at any exposure time. Exposures to 1
or 5 mg/kg crystalline silica particles produced a sustained increase in BAL fluid LDH
and MTP values vs. controls and the 5 mg/kg produced a decrease after week 1 through
the 3-month postexposure period, demonstrating a sustained cytotoxic effect on the
lungs (*p < 0.05 vs. PBS controls).
24 hours
1 week
1 month
15
Lung Weights (g)

3 months
10
Trachieobroncial Cells
1.00 *
% Proliferating
0.75
0.50
0.25
0.00
1.00 *
Parenchymal Cells
% Proliferating
0.75
*
0.50
0.25
0.00
PBS Carbonyl Crystalline Crystalline
Iron Silica Silica
(5 mg/kg) (1 mg/kg) (5 mg/kg)
Fig. 4. Top : Lung weights, middle : tracheobronchial cell proliferation rates (% cells
immunostained for BrdU), and bottom : lung parenchymal cell proliferation rates (% cells
immunostained for BrdU) of sham and particle-instilled rats at 24 h, 1 week, 1 month,
and 3 months postinstillation. Values given are means ± a standard deviation. Lung
weights of rats increased with increasing postinstillation time; however, no difference in
the lung weights was observed between the particle-instilled and sham groups at any
postexposure time point. Significant increases in airway tracheobronchial cell prolifera-
tion indices were measured in high-dose a-quartz exposed rats at 24 h postinstillation,
but these effects were not sustained. Significant increases in lung parenchymal cell
proliferation indices were measured in rats exposed to high-dose a-quartz at 24 h and
1 month postinstillation (*p < 0.05 vs. PBS controls). Exposures to carbonyl iron particle
did not produce any significant differences in cell proliferation indices compared to
vehicle controls.
7. Stain the slides with an anti-BrdU antibody containing a AEC

marker and counter-stain with aqueous hematoxylin.
8. For each treatment group, count immunostained nuclei in
airways (i.e., terminal bronchiolar epithelial cells) or lung
parenchyma (i.e., epithelia, interstitial cells, or macrophages)
by light microscopy at 1,000× magnification (12, 13). Count
a minimum of 1,000 cells/animal in both the terminal bron-
chiolar and alveolar regions of the lung.
9. For histopathological analyses, make sagittal sections of the
left lung with a razor blade.
10. Dissect tissue blocks from upper, middle, and lower regions
of the lung and prepare them for light microscopy (paraffin
embedded, sectioned, and hematoxylin–eosin-stained).
11. Also, evaluate lungs for inflammatory reactions and lung
fibrotic effects (see Note 9).
3.5. Statistical 1. Calculate a one-way analysis of variance (ANOVA) and

Analyses Bartlett’s test for each sampling time (see Note 10).
2. Compare the means of experimental values to their corre-
sponding sham control values for each time point.
Subsequently, normalize the data and represent it as a per-
centage of the sham control values for that experiment.
4. Notes
1. Ensure that the particles are evenly dispersed in the PBS vehicle
otherwise a nonuniform exposure could result. Sonication
can be used to disperse the nanoparticles if necessary.
2. Ensure that loss of animals does not occur during the instilla-
tion procedure due to an overdose of anesthesia.
3. When the experimenter incorporates a material (chemical or
particle) that induces an inflammatory response (positive
control), a material that induces little or no response in the
animal (negative control), and the vehicle control (no parti-
cles), then accurate interpretation of data is more likely to be
achieved.
4. Time-course experiments are needed to determine if the
effects are transient or sustained; transient inflammation can
occur at the 24-h postexposure time point.
5. All refrigerated lavaged samples containing enzymes and pro-
teins are stable for a minimum period of 24 h.
6. Trypan blue is a stain that colors dead cells.
7. LDH is a cytoplasmic enzyme and an indicator of cell injury.

AlkP activity is a measure of Type II alveolar epithelial cell
secretory activity, and increased AlkP activity in BAL fluids is
considered to be an indicator of Type II lung epithelial cell
toxicity.
8. Increased concentrations of BAL fluid micrototal protein
(MTP) generally are consistent with enhanced permeability
of vascular proteins into the alveolar regions, indicating a
breakdown in the integrity of the alveolar–capillary barrier.
Normal variability for protein and LDH values among sham
control samples for each time period average 17%.
9. Progressive lung tissue thickening is a prelude to the develop-
ment of fibrosis.
10. When the F test from ANOVA is significant, the Dunnett’s
test is used to compare means from the control group and
each of the groups exposed either to silica or to carbonyl iron.
Significance is judged at the 5% probability level.
Acknowledgment
This work was supported by DuPont Haskell Global Centers for

Health and Environmental Sciences.
References
1. Beck, B. D., Brain, J. D., and Bohannon, D. E. 5. Lugano, E. M., Dauber, J. H., and Daniele, R. P.
(1982) An in vivo hamster bioassay to assess (1982) Acute experimental silicosis: Lung
the toxicity of particulate for the lungs. Toxicol. morphology, histology, and macrophage
Appl. Pharmacol. 66, 9–29. chemotaxin secretion. Am. J. Pathol. 109,
2. Lindenschmidt, R. C., Driscoll, K. E., Perkins, 27–36.
M. A., Higgins, J. M., Maurer, J. K., and 6. Bowden, D. H. (1987) Macrophages, dust
Belfiore, K. A. (1990) The comparison of a and pulmonary disease. Exp. Lung Res. 12,
fibrogenic and two nonfibrogenic dusts by 89–107.
bronchoalveolar lavage. Toxicol. Appl. 7. Reiser, K. M. and Last, J. A. (1986) Early cel-
Pharmacol. 102, 268–281. lular events in pulmonary fibrosis. Exp. Lung.
3. Driscoll, K. E., Lindenschmidt, R. C., Maurer, Res. 10, 311–355.
J. K., Higgins, J. M., and Ridder, G. (1990) 8. Morgan, A., Moores, S. R., Holmes, A., Evans,
Pulmonary response to silica or titanium diox- J. C., Evans, N. H., and Black, A. (1980) The
ide: Inflammatory cells, alveolar macrophage- effect of quartz, administered by intratracheal
derived cytokines, and histopathology. Am. instillation, on the rat lung. 1. The cellular
J. Respir. Cell Mol. Biol. 2, 381–390. response. Environ. Res. 22, 1–12.
4. Warheit, D. B., Webb, T. R., Reed, K. L., and 9. Bowden, D. H. and Adamson, I. Y. R. (1984)
Sayes, C. M. (2007) Pulmonary toxicity study The role of cell injury and the continuing
in rats with three forms of ultrafine-TiO2 par- inflammatory response in the generation of
ticles: Evidence for differential responses. silicotic pulmonary fibrosis. J. Pathol. 144,
Toxicology 230, 90–104. 149–161.
10. Warheit, D. B., Brock, W. J., Lee, K. P., 12. Warheit, D. B., Carakostas, M. C., Hartsky,
Webb, T. R., and Reed, K. L. (2005) M. A., and Hansen, J. F. (1991) Development
Comparative pulmonary toxicity inhalation of a short-term inhalation bioassay to assess
and instillation and studies with different pulmonary toxicity of inhaled particles:
TiO2 particle formulations: Impact of surface Comparisons of pulmonary responses to car-
treatments on particle toxicity. Toxicol. Sci. bonyl iron and silica. Toxicol. Appl. Pharmacol.
88, 514–524. 107, 350–368.
11. Warheit, D. B., Chang, L. Y., Hill, L. H., 13. Warheit, D. B., Hansen, J. F., Yuen, I. S.,
Hook, G. E. R., Crapo, J. D., and Brody, A. R. Kelly, D. P., Snajdr, S., and Hartsky, M. A.
(1984) Pulmonary macrophage accumulation (1997) Inhalation of high concentrations of
and asbestos-induced lesions at sites of fiber low toxicity dusts in rats results in pulmonary
deposition. Am. Rev. Respir. Dis. 129, and macrophage clearance impairments.
301–310. Toxicol. Appl. Pharmacol. 145, 10–22.
Chapter 21
Nanoparticle Therapeutics: FDA Approval, Clinical Trials,

Regulatory Pathways, and Case Study
Aaron C. Eifler and C. Shad Thaxton
Abstract
The approval of drugs for human use by the US Food and Drug Administration (FDA) through the
Center for Drug Evaluation and Research (CDER) is a time-consuming and expensive process, and
approval rates are low (DiMasi et al., J Health Econ 22:151–185, 2003; Marchetti and Schellens, Br J
Cancer 97:577–581, 2007). In general, the FDA drug approval process can be separated into preclinical,
clinical, and postmarketing phases. At each step from the point of discovery through demonstration of
safety and efficacy in humans, drug candidates are closely scrutinized. Advances in nanotechnology are
being applied in the development of novel therapeutics that may address a number of shortcomings of
conventional small molecule drugs and may facilitate the realization of personalized medicine (Ferrari,
Curr Opin Chem Biol 9:343–346, 2005; Ferrari, Nat Rev Cancer 5:161–171, 2005; Ferrari and
Downing, BioDrugs 19:203–210, 2005). Appealingly, nanoparticle drug candidates often represent
multiplexed formulations (e.g., drug, targeting moiety, and nanoparticle scaffold material). By tailoring
the chemistry and identity of variable nanoparticle constituents, it is possible to achieve targeted delivery,
reduce side effects, and prepare formulations of unstable (e.g., siRNA) and/or highly toxic drugs (Ferrari,
Curr Opin Chem Biol 9:343–346, 2005; Ferrari, Nat Rev Cancer 5:161–171, 2005; Ferrari and
Downing, BioDrugs 19:203–210, 2005). With these benefits arise new challenges in all aspects of regu-
lated drug development and testing.
This chapter distils the drug development and approval process with an emphasis on special consid-
erations for nanotherapeutics. The chapter concludes with a case study focused on a nanoparticle thera-
peutic, CALAA-01, currently in human clinical trials, that embodies many of the potential benefits of
nanoparticle therapeutics (Davis, Mol Pharm 6:659–668, 2009). By choosing CALAA-01, reference is
made to the infancy of the therapeutic nanoparticle field; in 2008, CALAA-01 was the first targeted
siRNA nanoparticle therapeutic administered to humans. Certainly, there will be many more that will
follow the lead of CALAA-01 and each will have its own unique challenges; however, much can be
learned from this drug in the context of nanotherapeutics and the evolving development and approval
process as it applies to them.
Key words: Nanotherapeutics, FDA approval, Regulation, CALAA-01, Drugs
325
326 Eifler and Thaxton
1. The Drug
Approval Process:
An Overview
Drug development and approval by the US Food and Drug
Administration (FDA) can be broadly categorized into three
major phases (see Fig. 1) (1, 2). The preclinical phase includes the
initial discovery of a candidate therapeutic, demonstration of effi-
cacy, and investigation of toxicity. Following discovery, the major
focus of preclinical development is animal testing to obtain evi-
dence supporting drug safety and efficacy, and to determine
appropriate dosing parameters. The data gathered during the pre-
clinical phase is used to support an Investigational New Drug
(IND) application filed with the FDA. Following IND approval,
the candidate drug enters the clinical phase during which human
trials are initiated and are subcategorized into Phases I, II, and
III. If during the clinical phase, the drug is considered safe and
efficacious, the manufacturer files a New Drug Application
(NDA), also with the FDA. Upon NDA approval, the drug can
then be marketed according to specific labeling and put into use
by medical practitioners. As part of the NDA approval, the FDA
may request further studies be done in what is referred to as the
postmarketing phase, or Phase IV. Studies done after NDA approval
are used to further confirm efficacy and/or safety, especially in
groups that may not have been well studied in pre-NDA clinical
trials. The entire FDA approval process is lengthy, labor-intensive,
and stringent. It is estimated that it takes approximately
10–15 years to develop a new medicine at a cost of approximately
$1 billion (3–5, 6). In addition, a number of drugs that are dis-
covered and evaluated in the preclinical stages of drug develop-
ment have an exceedingly high-attrition rate (5).
Fig. 1. The major phases of the FDA drug approval process.

Nanoparticle Therapeutics: FDA Approval, Clinical Trials, Regulatory Pathways 327
1.1. The Preclinical The preclinical phase is the most diverse phase of drug development.
Phase New small molecule drug candidates are discovered through a
number of different means. In some cases, candidates can be pre-
dicted in silico based upon known target and ligand interactions
(e.g., crystal structure data) (7–10). Combinatorial approaches
are also used to produce and screen libraries of small molecule
drugs against known targets (11–13). Currently, high through-
put methodologies are not commonplace in the fabrication and
interrogation of nanoparticle therapeutics. In many cases, nano-
therapeutics are built upon mature, well-characterized nanopar-
ticle platforms whose syntheses and surface and/or internal
modification are well understood and can easily be controlled.
For instance, liposomes, metal nanoparticles (e.g., colloidal gold),
metal oxides (e.g., superparamagnetic iron oxide nanoparticles),
nanoshells (e.g., gold), quantum dots, self-assembling peptides/
proteins, fullerenes, and dendrimers are all mature platform tech-
nologies used to fabricate nanoparticle therapeutics (14).
In general, nanoparticle platforms require some degree of
functionalization(s) to impart desired drug function (15–17).
Nanotherapeutics are being built upon relatively few platform
technologies, but with the potential for near endless platform
modification(s) and chemical tailoring. Nanoparticle drugs often
consist of more than one molecular component coupled to the
nanoparticle surface or contained within (see Fig. 2). Systematic
manipulation of individual, and often multiple, components of a
nanoparticle platform can generate a substantial number of unique
therapeutics. Owing to their high degree of tailorability, nanopar-
ticle therapeutics can be rationally designed based upon precon-
ceived notions of what the “ultimate” nanoparticle therapeutic
Fig. 2. The tailorability of nanoparticles allows for the manipulation of nearly every physi-
cal characteristic. Shown are four characteristics that can be manipulated to impart
desired function, but that can also result in significant changes in in vivo drug behavior.
Adapted by permission from the authors and from Macmillan Publishers Ltd, ref. 13.
Copyright 2007.
represents from the standpoint of adaptability, ease of manufacture,

tailorability, functionality, etc. (16–19). Nanoparticle therapeu-
tics envisioned for human use must have robust methodologies
for synthesis, characterization, quality control, and potential
scale-up. These issues should be kept in mind during planning
and early development phases to facilitate translation.
Realizing the need for robust methods to characterize the
multitude of these nanoparticle entities prior to IND evalua-
tion, the National Cancer Institute (NCI) created the
Nanotechnology Characterization Laboratory (NCL) in 2004
(20). The NCL represents the desire of the NCI to fund and
centralize the development of standardization protocols by
which emerging nanotherapeutics and other nanomaterials can
be preclinically evaluated with regard to efficacy and toxicity
(20). The NCL has identified a number of physicochemical
parameters that must be well controlled and understood in the
early stages of nanotherapeutic development. Importantly, sur-
face chemistry has a significant impact on in vitro and in vivo
nanotherapeutic behavior (14, 20, 21). The systematic manipu-
lation of surface chemistry is an ongoing effort, and a rich area
for future research, in order to more accurately predict the
behavior of nanoparticle therapeutics (20).
In developing a therapeutic for use in humans, significant
in vitro and in vivo testing is required to assess drug safety and
efficacy. According to the FDA, “During a new drug’s early pre-
clinical development, the sponsor’s primary goal is to determine
if the product is reasonably safe for initial use in humans, and if
the compound exhibits pharmacological activity that justifies
commercial development” (2). Addressing drug safety, preclinical
testing focuses specifically on how the drug will interface with
biological systems, and currently takes place on a drug-by-drug
basis. A relatively comprehensive list of in vitro and in vivo testing
currently performed on materials submitted to the NCL serves as
a useful guide (20). Initial studies include data demonstrating
efficacy in a number of cell and animal models appropriate to the
disease under scrutiny. In cell culture, it is imperative to demon-
strate the mechanism of action of the therapeutic and that the
observed effect (e.g., apoptosis) is due to the therapeutically
active component(s) of the nanoparticle drug and not the result
of the nanoparticle platform devoid of drug component (i.e.,
vehicle). In order to get smooth transition from cell culture mod-
els to animal testing, it is important to pair cell culture models
and controls with the same or comparable animal models; often,
testing commences in mice or rat models. For nanoparticle thera-
peutics that contain nucleic acid-based therapeutic moieties (e.g.,
siRNA), testing should also assess potential off-target effects and
any nucleic acid-mediated inflammatory response. Further, when
possible, nucleic acid therapeutics (e.g., siRNAs) should target
conserved sequences among different animal species that will be

tested (18). It should also be kept in mind that most often the
assay used for assessing a given response was not designed for a
nanoparticle therapeutic (14). Thus, it is important to discern
how the physicochemical properties of nanoparticle therapeutics
(e.g., UV–Vis absorbance) impact assay readout and, potentially,
generate false results (14).
Once a promising compound has been discovered, character-
ized, and demonstrated efficacy and safety in the treatment of a
disease in cell culture and in an animal model(s), the ultimate goal
of further preclinical testing is to amass sufficient evidence to sup-
port an IND submission to the FDA. Other data required to sup-
port an IND application includes toxicology, manufacturing
information, and proposed clinical protocols (2). Importantly,
such parameters are often assessed in multiple animal models
(e.g., mouse, dog, nonhuman primate). Based on these data, the
manufacturer must establish a dose range for initial administra-
tion to human subjects to be used in Phase I studies pending
approval of the IND application.
1.2. Investigational Before filing an IND application with the FDA, a manufacturer
New Drug Application may choose to participate in the pre-IND Application Consultation
Program (2). The consultation program is designed to foster early
communication between the FDA and drug developers and to
provide guidance regarding the information and data required for
a successful IND application. Multiple consultation divisions exist
and are organized by therapeutic class and organ system. Currently,
there is no specific division for nanotherapeutics.
The IND application must contain information in three areas:
animal pharmacology and toxicity, manufacturing information,
and clinical protocols and investigator information (2). Animal
pharmacology and toxicity information comes from the extensive
preclinical testing of the drug that occurs after discovery and
allows the FDA to determine whether the product is reasonably
safe for testing in humans. This information also allows investiga-
tors to propose an initial dose to be tested in humans.
Manufacturing information pertains to the composition, and sta-
bility of the drug as well as information on the manufacturer and
the manufacturer’s methods for quality control. The FDA uses
this information to ensure that the company can adequately pro-
duce the drug in sufficient quantity and consistency across batches
of the drug. The final piece of information is twofold; it concerns
the protocols that will be used in the clinical phase of testing and
the investigators who will oversee it. First, detailed protocols are
scrutinized by the FDA to ensure that patients are not exposed to
unnecessary risks and that proper informed consent will be
obtained. This aspect is bolstered with a commitment for involve-
ment by the Institutional Review Board (IRB) to review the study.
Second, information must be provided that demonstrates the

qualifications of clinical investigators (generally physicians) who
will oversee the administration of the compound. It must be
shown that these physicians will be able to fulfill their clinical trial
duties which often go above and beyond those of typical practice.
After submission, the investigator must wait 30 days before initi-
ating any clinical testing while the FDA reviews the application
and ensures that research subjects will not be subjected to unrea-
sonable risk. If approved, the potential drug is able to move into
Phase I clinical testing in humans based on the study protocols
proposed in the IND application.
Further, drugs are required to be the subject of an approved
marketing application by federal law prior to their shipment across
state lines (2). Exemption must be sought from this federal law
for clinical trials designed to be conducted in multiple states.
Technically, a successful IND application becomes the means by
which drug developers are exempted from marketing application
approval, and are able to conduct clinical drug trials across state
lines. More practically, the approval of an IND application signi-
fies the transition from preclinical development to clinical
testing.
1.3. The Clinical Phase Phase I trials represent the first human dose of a drug whose IND
application has just been approved (2). The main goal of Phase I
1.3.1. Phase I
testing is to assess dosing, acute toxicity, and drug excretion in
humans. Typically, the candidate drug is administered to 20–100
healthy volunteers often with dose escalation. Healthy patients
provide a baseline evaluation of initial dosing regimens derived
from animal studies. In cases of severe or life-threatening illnesses,
studies may enroll volunteers with the disease. For nanotherapeu-
tics, as in the case of all therapeutics, testing for therapeutic-spe-
cific side effects identified in preclinical studies are particularly
and formally scrutinized in Phase I trials. On average, Phase I
studies can take from 6 months to one and a half years to com-
plete. An estimated two thirds of Phase I compounds will move
on to Phase II trials (1, 4).
1.3.2. Phase II Phase II studies involve more patients (approximately 100–300)

than Phase I studies and further evaluate the safety and efficacy of
a potential drug in a group of patients who suffer from the disease
being studied (2). Phase II studies also can be placebo-controlled
in order to establish that the drug effectively treats the disease for
which it is intended. To avoid unnecessarily exposing patients to
a potentially harmful compound, the number of patients enrolled
is based on appropriate power calculations, statistically restricting
the study size to one where meaningful data can be expected with
the fewest patient exposures. Phase II trials can take from 6
months to 2 years, however, recent evidence suggests that while
the duration of Phase I and Phase III studies are staying relatively
constant, the duration of Phase II studies has significantly
increased (22). One reason for Phase II study prolongation is the
desire to more thoroughly interrogate drug safety and efficacy in
multiple patient groups, potentially, at multiple sites (22). This
may increase the time that it takes to accrue patients. However,
more data in Phase II provides increasing confidence to either
halt evaluation or move forward to large and expensive Phase III
trials (22).
1.3.3. Phase III Phase III studies are large, randomized, placebo-controlled trials
including typically 1,000–5,000 patient volunteers from hospi-
tals, clinics, and/or physician offices across the country (1, 2).
These studies are used to demonstrate further safety and efficacy
and typically the investigational drug is directly compared to the
gold standard treatment for a given indication, provided that one
exists. Phase III studies can take from 1 to 10 years to complete.
Even at this stage, after extensive testing is already completed,
approximately 10% of medications fail in Phase III trials (1).
Success in Phase III trials is a complex decision which largely
rests upon drug safety and efficacy relative to the gold standard
of care.
1.4. New Drug Since 1938, every new drug has been the subject of an approved
Application NDA before US commercialization (2). The NDA application is
the vehicle through which drug sponsors formally propose that
the FDA approve a new pharmaceutical for sale and marketing.
The data gathered during all prior phases of drug development
become part of the NDA. The documentation required in an
NDA recounts the drug’s entire history, including the drug iden-
tity, the results of animal studies, clinical trial outcomes, how the
drug behaves in the body, and how it is manufactured, processed,
and packaged. After appropriate deliberation, the FDA may
request additional information by written request, issue a tenta-
tive approval under which minor deficiencies need to be corrected
prior to final approval, or NDA approval may be granted and
contain conditions that must be met after initial marketing, such
as Phase IV studies (1, 2). Typically, an NDA takes from 6 months
to 1 year for review (1).
1.5. The Postmarketing Postmarketing studies, also known as Phase IV studies, are under-
Phase taken after the NDA has been approved by the FDA (2). The
impetus for postmarketing studies can come from a number of
places including the FDA, practicing clinicians, or the manufac-
turer. As mentioned previously, the FDA may request a postmar-
keting study to examine its effects on a high-risk population, or a
population that was not well represented in the Phase III trial.
Clinicians may be interested in further study of the drug to assess
its efficacy or side effect profile compared to other treatments.

Manufacturers may initiate Phase IV studies to assess long-term
effects of the drug or to examine the drug’s effectiveness in addi-
tional indications (23). Although not structured as a strict exper-
iment-based study, postmarketing surveillance of drug
administration continues throughout the lifetime of the drug.
Physicians continue to report untoward side effects of the drug
and manufacturers must submit periodic reports to the FDA
describing any cases of adverse events.
As relatively new entities, nanoparticle therapeutics present
the pharmaceutical industry with a vast array of opportunities.
There is great anticipation surrounding the potential of nano-
therapeutics and nanoparticle platform technologies to overcome
current therapeutic barriers. This chapter heavily focuses on pre-
clinical nanoparticle development due to the more robust litera-
ture available, and the paucity of nanotherapeutics that have
reached the clinical stages of development. Progression of nano-
therapeutics from preclinical to clinical phases of development is
expected to be taken on a case-by-case basis with similar interro-
gation at each phase as would be applied with traditional thera-
peutics. Emphasized throughout this chapter is the need for
robust nanotherapeutic physicochemical and biological character-
ization, and quality control so as to build a solid foundation for
moving through preclinical and into clinical development. At the
time of this writing, there are a number of nanoparticles in pre-
clinical and early stage clinical development (14, 20) demonstrat-
ing that progress is being made. The approval of a new drug is a
laborious and costly process and there is uncertainty due to a
general lack of experience with this class of potential drugs. This
is necessitating partnerships with academia, small start-up compa-
nies, and large pharmaceutical companies with expertise in a par-
ticular nanotherapeutic. Certainly, general knowledge of different
nanoparticles and nanoparticle platforms will disseminate as indi-
vidual materials and candidate nanotherapeutics find their way
into development pipelines. Despite the unique challenges, the
future is bright for the development of nanoparticle drugs, col-
laboration, and discovery. The hope is that the novel properties
displayed by nanoparticle therapeutics will directly lead to
advancements in the successful treatment of human disease.
2. Case Study
This case highlights CALAA-01, the first example of a targeted

nanoparticle used to deliver a siRNA drug. At the time of this
writing, CALAA-01 has progressed through the preclinical stages
of drug development and is now in human clinical trials (18).
Work leading to the first targeted nanoparticle delivery of

siRNA in humans, the drug CALAA-01, was begun by Mark
Davis and colleagues at the California Institute of Technology in
1996 (18). The goal was to develop a multifunctional targeted
cancer therapeutic which would enable the systemic administra-
tion of nucleic acids. Their self-stated design methodology was a
systems approach to a multifunctional colloidal particle – a ratio-
nally derived nanoparticle therapeutic employed before the term
“nanoparticle” was generally used. The initial drug schematic (see
Fig. 3) presents the desired components of an envisioned thera-
peutic with (a) a cyclodextrin-containing polymer (CDP) core
that spontaneously self-assembles with nucleic acids yielding small
colloidal particles less than 100 nm in diameter, (b) a targeting
ligand (R) providing for tumor cell specificity and uptake, and,
(c) the appreciation of endosomal acidification as a mechanism
for particle disassembly and endosomal escape of the therapeutic
nucleic acid (18). This CDP-system was initially envisioned for
use with plasmid DNA; however, as the association of the nucleic
acid with CDP is based upon electrostatic interactions, it was
appreciated that the approach could be a rather general one for
candidate nucleic acid therapeutics. In addition to platform gen-
erality, the platform components were also chosen due to their
amenability for scale-up and manufacture.
CALAA-01 is an embodiment of this initial vision. CALAA-
01 ultimately evolved to include a number of key components
which spontaneously assemble into therapeutic nanoparticles ~70 nm
in diameter (see Fig. 4) (18). In addition to the CDP particle core
DNA
Cell membrane
DNA
pH − 7
R
DNA
DNA
pH < 7
Endosome
Fig. 3. Initial schematic of the delivery system that would become CALAA-01. Reprinted
with permission from author and from ref. 15. Copyright 2009 American Chemical
Society.
Fig. 4. Schematic of the two-vial formulation. The siRNA is contained in one vial, and the
delivery components are contained in the other. Upon mixing, targeted nanoparticles
form via self-assembly. Reprinted with permission from author and from ref. 15.
Copyright 2009 American Chemical Society.
and the siRNA payload, the surface of the formed nanoparticles is

decorated with (a) adamantane-polyethylene glycol (AD-PEG)
for drug stabilization in biological matrices, (b) adamantane-
PEG-transferrin (AD-PEG-Tf) for tumor-specific targeting and
cellular uptake, and (c) imidazole residues to titrate the decrease
in endosomal pH upon cellular uptake and promote the endo-
somal escape of the otherwise sequestered nucleic acid drug.
Adamantane is a small molecule that binds tightly and forms an
inclusion complex with cyclodextrin on the surface of the formed
nanoparticles, thus, displaying PEG and Tf.
Shortly after the conjugate therapeutic was developed, inves-
tigators found that the nanoparticles could be formed by self-
assembly through simultaneous component reconstitution and
mixing (24, 25). This finding gave rise to a unique two-vial for-
mulation strategy (see Fig. 4) which allows for the rapid and
straight forward self-assembly of the nanoparticle delivery system
components (CDP, AD-PEG, AD-PEG-Tf) with siRNA, at the
point of care (25). This formulation provides for siRNA solvation
immediately prior to reconstitution with the nanoparticle delivery
system components. This is an advantage because siRNA is highly
unstable when solvated, but, following the self-assembly process,
is protected from nuclease degradation. In addition, this two-vial
formulation allowed the molecular delivery system components of
CALAA-01 to be separated from siRNA and tested separately for
safety in animal models prior to introduction in humans (26). The
composition of the formed nanoparticle therapeutic following
mixing of the separated components has been well characterized
from the standpoint of size and molecular composition (25).
The initial in vitro and in vivo demonstration of the CDP-
based siRNA delivery system was published in 2005 using a dis-
seminated murine model of Ewing’s sarcoma (27). In these
studies, the siRNA targeted the breakpoint of the EWS–FLI1
fusion gene, which is an oncogenic transcriptional activator, in
TC71 cells positive for EWS–FLI1 and for the transferrin recep-
tor (27). In addition to in vitro inhibition of their targeted gene
product, TC71 cells transfected with firefly luciferase and injected
into NOD/SCID mice served as a model system of metastatic
Ewing’s sarcoma where tumor dissemination and treatment effi-
cacy could be assessed using bioluminescent imaging (27). In this
murine model, investigators administered their targeted, CDP-
based siRNA delivery particle and demonstrated antitumor effects
and target-specific mRNA down regulation (27). Further studies
provided evidence that the CDP-based delivery system does not
illicit an innate immune response (27), and that active targeting
to the transferrin receptor enhances tumor cell uptake (28). In
the case of CALAA-01, siRNA targeting ribonucleotide reductase
subunit 2 (RRM2) was identified (29). In addition to potency,
siRNA targeting RRM2 demonstrates complete sequence homol-
ogy in mouse, rat, monkey, and humans which allowed for a sin-
gle siRNA to be used for conducting preliminary studies in all
animal models. Targeting RRM2, Davis and colleagues confirmed
effective protein knockdown with concomitant reduction in
tumor cell growth potential in a subcutaneous mouse model of
neuroblastoma (30).
Building upon the above experiments and drawing closer to
the initiation of clinical drug testing, Davis and colleagues per-
formed the first study showing that multidosing of siRNA, in the
context of the CDP-based nanoparticles, could be done safely in
a nonhuman primate (26). This study demonstrated that dosing
parameters that were well tolerated were similar to those which
had demonstrated antitumor efficacy in mouse models.
Furthermore, reversible toxicity was observed in the form of mild
renal impairment at high dose; however, extrapolations from
mouse model efficacy studies suggested that the therapeutic dos-
ing window would be large. With these promising animal results,
a solid foundation was provided for moving the CDP-based
siRNA therapeutic platform to the clinic.
In May of 2008, CALAA-01 became the first targeted deliv-
ery of siRNA in humans (18). Details of this study, and others
focused on nanoparticle therapeutics, can be found at http://
www.clinicaltrials.gov. In Phase I studies, patients with solid-
organ tumors refractory to treatment were administered CALAA-
01 to assess drug safety. Patients were administered CALAA-01
by way of intravenous infusion on days 1, 3, 8, and 10 of a 21-day
cycle (18). Importantly, in a recent study by the Davis group, the
authors demonstrate that CALAA-01 effectively targets RRM2
through an RNAi mechanism of action in tumor tissue taken from

patients with melanoma after systemic administration of CALAA-
01 (31). Certainly, the updated results of clinical trials with
CALAA-01 are eagerly anticipated.
Many more nanoparticle therapeutics are on the horizon and
will follow CALAA-01 into the clinical setting. By providing this
brief case-study, certain elements of nanotherapeutic design and
development are highlighted with the hope of providing a con-
crete example for referring back to the, more general, main text
of this chapter. As a pioneering nanoparticle therapeutic, much
can be learned from CALAA-01. For instance, scrupulous detail
to the design, fabrication, characterization, and quality control of
CALAA-01 provided a solid platform for moving the drug for-
ward in preclinical and clinical trials. In addition, the choice of a
species-generic but target-specific siRNA sequence targeting
RRM2 provided direct access to multiple animal models, ulti-
mately, for translation to humans. Finally, testing in a wide variety
of preclinical animal models provided the data necessary to antici-
pate safe dosing parameters in humans, and dictated drug-specific
safety monitoring protocols in humans. Building upon CALAA-
01, it is anticipated that what is now an “infant” field will ulti-
mately provide new therapeutics, based upon advances in
nanotechnology, which will provide for significant improvements
in human health.
Acknowledgments
C. S. T. wishes to thank the Zell Family for their generous sup-

port of his research through the Robert H. Lurie Comprehensive
Cancer Center of Northwestern University, the National Science
Foundation/Nanoscale Science and Engineering Center at
Northwestern University for seed grant funding, the Howard
Hughes Medical Institute for a Physician Scientist Early Career
Award, the National Institutes of Health/NCI for funding
through the Center of Cancer Nanotechnology Excellence, the
International Institute for Nanotechnology at Northwestern
University, the Institute for BioNanotechnology and Medicine
at Northwestern University; and the Northwestern University
Department of Urology, Feinberg School of Medicine.
Furthermore, C. S. T. thanks all of the faculty, staff, postdocs,
Urology residents, graduate students, and research technolo-
gists who dedicate their time and effort to the unwavering
development of an ever expanding array of nanoparticle-based
diagnostics and therapeutics in hopes of improving human
health. A. C. E. wishes to thank the Howard Hughes Medical
Institute for their Research Training Fellowship for Medical
Students and the Northwestern University Feinberg School of

Medicine Departments of Radiology and Urology for their
guidance and generous support.
References
1. Lipsky, M. S. and Sharp, L. K. (2001) From 13. Schnur, D. M. (2008) Recent trends in library
idea to market: the drug approval process. design: ‘rational design’ revisited. Curr. Opin.
J. Am. Board Fam. Pract. 14, 362–367. Drug Discov. Devel. 11, 375–380.
2. http://www.fda.gov. 14. Dobrovolskaia, M. A. and McNeil, S. E. (2007)
3. DiMasi, J. A., Hansen, R. W., and Grabowski, Immunological properties of engineered nano-
H. G. (2003) The price of innovation: new materials. Nat. Nanotechnol. 2, 469–478.
estimates of drug development costs. J. Health 15. Ferrari, M. (2005) Cancer nanotechnology:
Econ. 22, 151–185. opportunities and challenges. Nat. Rev. Cancer
4. DiMasi, J. A. (2002) The value of improving 5, 161–171.
the productivity of the drug development pro- 16. Ferrari, M. (2005) Nanovector therapeutics.
cess: faster times and better decisions. Curr. Opin. Chem. Biol. 9, 343–346.
Pharmacoeconomics 20 Suppl. 3, 1–10. 17. Ferrari, M. and Downing, G. (2005) Medical
5. Kola, I., and Landis, J. (2004) Can the phar- nanotechnology: shortening clinical trials and
maceutical industry reduce attrition rates? regulatory pathways? BioDrugs 19, 203–210.
Nat. Rev. Drug Discov. 3, 711–715. 18. Davis, M. E. (2009) The first targeted delivery
6. Marchetti, S. and Schellens, J. H. (2007) The of siRNA in humans via a self-assembling,
impact of FDA and EMEA guidelines on drug cyclodextrin polymer-based nanoparticle: from
development in relation to Phase 0 trials. Br. concept to clinic. Mol. Pharm. 6, 659–668.
J. Cancer 97, 577–581. 19. Davis, M. E., Pun, S. H., Bellocq, N. C.,
7. Brooks, W. H., McCloskey, D. E., Daniel, K. G., Reineke, T. M., Popielarski, S. R., Mishra, S.,
Ealick, S. E., Secrist, III, J. A., Waud, W. R., et al. (2004) Self-assembling nucleic acid
et al. (2007) In silico chemical library screen- delivery vehicles via linear, water-soluble,
ing and experimental validation of a novel cyclodextrin-containing polymers. Curr. Med.
9-aminoacridine based lead-inhibitor of human Chem. 11, 179–197.
S-adenosylmethionine decarboxylase. J. Chem. 20. http://ncl.cancer.gov/.
Inf. Model. 47, 1897–1905. 21. Dobrovolskaia, M. A., Germolec, D. R., and
8. Lee, P. H., Ayyampalayam, S. N., Carreira, L. Weaver, J. L. (2009) Evaluation of nanoparticle
A., Shalaeva, M., Bhattachar, S., Coselmon, immunotoxicity. Nat. Nanotechnol. 4, 411–414.
R., et al. (2007) In silico prediction of ioniza- 22. Accenture (2007) The Pursuit of High
tion constants of drugs. Mol. Pharm. 4, Performance through Research and
498–512. Development: Understanding Pharamceutical
9. Shacham, S., Marantz, Y., Bar-Haim, S., Kalid, Research and Development Cost Drivers.
O., Warshaviak, D., Avisar, N., et al. (2004) http://www.phrma.org/files/attachments/
PREDICT modeling and in-silico screening AccentureR&D Report-2007.pdf.
for G-protein coupled receptors. Proteins 57, 23. Leonard, E. M. (1994) Quality assurance and
51–86. the drug development process: an FDA per-
10. Sivanesan, D., Rajnarayanan, R. V., Doherty, spective. Qual. Assur. 3, 178–186.
J., and Pattabiraman, N. (2005) In-silico 24. Pun, S. H., Bellocq, N. C., Liu, A., Jensen, G.,
screening using flexible ligand binding pock- Machemer, T., Quijano, E., et al. (2004)
ets: a molecular dynamics-based approach. Cyclodextrin-modified polyethylenimine poly-
J. Comput. Aided Mol. Des. 19, 213–228. mers for gene delivery. Bioconjug. Chem. 15,
11. Aina, O. H., Liu, R., Sutcliffe, J. L., Marik, J., 831–840.
Pan, C. X., and Lam, K. S. (2007) From com- 25. Bartlett, D. W. and Davis, M. E. (2007)
binatorial chemistry to cancer-targeting pep- Physicochemical and biological characteriza-
tides. Mol. Pharm. 4, 631–651. tion of targeted, nucleic acid-containing nano-
12. Batra, S., Srinivasan, T., Rastogi, S. K., and particles. Bioconjug. Chem. 18, 456–468.
Kundu, B. (2002) Identification of enzyme 26. Heidel, J. D., Yu, Z., Liu, J. Y., Rele, S. M.,
inhibitors using combinatorial libraries. Curr. Liang, Y., Zeidan, R. K., et al. (2007)
Med. Chem. 9, 307–319. Administration in non-human primates of
escalating intravenous doses of targeted nano- 29. Heidel, J. D., Liu, J. Y., Yen, Y., Zhou, B.,
particles containing ribonucleotide reductase Heale, B. S., Rossi, J. J., et al. (2007) Potent
subunit M2 siRNA. Proc. Natl Acad. Sci. USA siRNA inhibitors of ribonucleotide reductase
104, 5715–5721. subunit RRM2 reduce cell proliferation in vitro
27. Hu-Lieskovan, S., Heidel, J. D., Bartlett, D. W., and in vivo. Clin. Cancer Res. 13,
Davis, M. E., and Triche, T. J. (2005) 2207–2215.
Sequence-specific knockdown of EWS–FLI1 30. Bartlett, D. W. and Davis, M. E. (2008)
by targeted, nonviral delivery of small interfer- Impact of tumor-specific targeting and dosing
ing RNA inhibits tumor growth in a murine schedule on tumor growth inhibition after
model of metastatic Ewing’s sarcoma. Cancer intravenous administration of siRNA-contain-
Res. 65, 8984–8992. ing nanoparticles. Biotechnol. Bioeng. 99,
28. Bartlett, D. W., Su, H., Hildebrandt, I. J., 975–985.
Weber, W. A., and Davis, M. E. (2007) 31. Davis, M. E., Zuckerman, J. E., Choi, C. H.
Impact of tumor-specific targeting on the J., Seligson, D., Tolcher, A., Alabi, C. A.,
biodistribution and efficacy of siRNA nano- et al. (2010) Evidence of RNAi in humans
particles measured by multimodality in vivo from systemically administered siRNA
imaging. Proc. Natl Acad. Sci. USA 104, via targeted nanoparticles. Nature 464,
15549–15554. 1067–1070.
Chapter 22
Legislating the Laboratory? Promotion and Precaution

in a Nanomaterials Company
Robin Phelps and Erik Fisher
Abstract
Legislation is a form of governance that directs attention and prescribes action. Within the domain of
nanoscience, the US 21st Century Nanotechnology Research and Development Act contains mandates
not only for rapid development for economic competitiveness but also for responsible implementation,
which is required to take place by integrating societal considerations into research and development. This
chapter investigates whether these two mandates tend more to coexist or compete with one another,
both in the purview of nanoscience policy and in the venue of nanoscience practice. This chapter first
reviews macrolevel analysis of the directives contained in the legislation. It then examines, drawing on an
empirical case study, how these directives manifest at the microlevel of a nanoscience research and devel-
opment laboratory.
Key words: Innovation, Integration, Nanotechnology, Precaution, Promotion, Responsible
1. Introduction
On December 3, 2003 the 21st Century Nanotechnology

Research and Development Act (NRDA) was signed into law (1).
This legislation established the National Nanotechnology
Initiative (NNI) as the National Nanotechnology Program (NNP)
and authorized multiyear federal funding for nanotechnology
research and development. Since then, more than US$6.5 billion
of federal funding has been authorized over the 4-year period,
from fiscal year (FY) 2005 to FY 2009, that the legislation has
been in effect.
The genesis of the NNP was a series of informal meetings in
1996 of the federal agencies involved in nanotechnology research.
In 1998, this informal group became a formal Interagency
Working Group (IWG). Over the next year, the IWG issued
339
340 Phelps and Fisher
three reports: Nanostructure Science and Technology (August

1999), Nanotechnology Research Directions (February 2000), and
National Nanotechnology Initiative (July 2000). Combined, these
reports provided a blueprint for the strategic intent of US invest-
ment in nanotechnology research and development. One founda-
tional footing of the blueprint was rapid technological development
and accelerated market deployment intended to keep the USA
competitive in the international arena for economic and other
gains projected to be realized from nanotechnology products.
Another foundational footing was responsible implementation
intended to proactively and adequately address public concern.
Eventually, both of these foundations became incorporated into
the US nanotechnology legislation. Since the policy foundations
of “rapid” and “responsible” nanotechnology research and devel-
opment appear, at least on the surface, to be contradictory (2),
it remains unclear and uncertain whether tensions between the
two foundational footings play themselves out in actual research
and development contexts. In short, do they coexist, and perhaps
even mutually reinforce one another? Or do they remain irrecon-
cilable, competing for focus and attention?
This chapter examines how these two policy foundations
manifest themselves in a US nanotechnology research and devel-
opment laboratory. First, an overview of the tensions as defined
in the various Program Activities of the Act provides context for
the case study. We then provide a brief review of issues and con-
cerns that have been stated and documented in preparation for
the Act’s reauthorization. This is followed by a description of the
case study including the overarching research project it is situated
in, the methodology and methods employed, the initial findings
based on a limited analysis, and a discussion of those findings.
2. Discussion
2.1. US The 21st Century Nanotechnology Research and Development

Nanotechnology Act (NRDA) defines eleven Program Activities which serve as
Legislation guideposts for intent and implementation. A reproduction of
(See Note 1) Section 2(b) of the Act, listing the eleven Program Activities, is
contained in Appendix 1. These Program Activities logically clus-
ter into three groups, which we label here as technical, promo-
tional, and precautionary.
Seven of the Program Activities pertain to the technical
objectives of nanotechnology research. These consist of the
methods and resources for the cultivation of nanotechnology as
an interdisciplinary science. Though some have expressed skepti-
cism about the interdisciplinary nature of nanotechnology
research and development, case studies of biomedical nanotech-
Legislating the Laboratory? Promotion and Precaution in a Nanomaterials Company 341
nology research settings provide a measure of confirmation of

such an interdisciplinarity (3–5).
The four remaining Program Activities address two areas that
policy makers deemed crucial to the public business of nanotech-
nology: promotion of the economic outcomes and precaution
regarding the societal dimensions.
2.1.1. Economic-Promotion Three Program Activities focus nanotechnology research and devel-
Considerations opment efforts on economic considerations that promote meeting
global competitiveness and extensive projected opportunities for
nanotechnology applications. These promotional considerations
include “ensuring…global leadership,” “advancing…productivity
and industrial competitiveness,” and “accelerating” nanotechnol-
ogy deployment. These Activities represent a key policy objective
behind the NNP: US domination in this new competitive global
market. The economic prospect for nanotechnology is projected to
be substantial. Lux Research, an international market research firm,
projected that between 2006 and 2014 global revenues from nan-
otechnology-enabled products will grow from $50 billion to $2.6
trillion and will comprise 15% of projected global manufacturing
output (6). Notably, nearly every industrialized and developing
country has initiated national research programs in nanotechnology
to capture a share of the projected economic and societal benefits.
Global competition for the prospective nanotechnology mar-
ket had reportedly grown over the 5-year period before the
NRDA’s passage. Mihail Roco, chair of the National Science and
Technology Council’s Subcommittee on Nanoscale Science,
Engineering, and Technology reported that at least 30 countries
had created national nanotechnology programs and that interna-
tional nanotechnology funding increased multiple times for a
global investment of approximately US$3 billion (7, 8).
The NRDA specifically requires “accelerating the deployment
and application of nanotechnology research and development in
the private sector, including startup companies.” This language
seeks to position nanotechnology deployment on a well-established
research-to-technology commercialization path within the US
innovation system – a system consisting of academic and federal
lab research, startup companies, venture capital firms, and other
entrepreneurial supporting infrastructure.
Combined, these promotional activities drive a policy focused
on rapid development not only to keep pace with international
competition but also to capture the benefits as well as the perva-
sive impacts of nanotechnology, which have been deemed “cru-
cial” for the country’s future economic health (9).
2.1.2. Societal-Precaution The single remaining Program Activity contained in the NRDA
Considerations stands by itself as much for its content as for its intent. Program
Activity (10) requires “ensuring that ethical, legal, environmental,
and other appropriate societal concerns… are considered during

the development of nanotechnology.” While this requirement for
responsible implementation can be seen as a counterpressure to
that of rapid implementation, it is also possible to regard it as a
notable recognition of the importance of social trust of institu-
tions for commercial success.
The impetus for this legislative directive came from a number
of concerns surrounding nanotechnology that had been expressed
in public and political discourses prior to the legislation. Policy
makers appeared eager to separate concerns they could regard as
credible and convincing from others that could be regarded as
too speculative or fictional. One prominent critical theme during
this time cited potential harm from exposure to nanotechnology
particles, suggesting its potential as the “next asbestos” (10).
Additional concerns encompassed other potential health, safety,
and environmental risks, and they extended to broad ethical and
political questions, including the role of democratic governance
in nanotechnology.
Citing a severe lack of governmental monitoring and regula-
tion of nanotechnology, the nongovernmental organization ETC
Group (11) called for a global, mandatory moratorium on nano-
technology research and product development to allow time for a
closer examination of the potential negative impacts on environ-
mental, health, and safety. The report criticized the “substantial
equivalence” regulatory approach being implemented at the time
for nanoscale materials, a policy that had been used previously to
show the safety of genetically modified organisms (GMOs) with-
out doing full toxicological analysis of GM crops. As applied to
nanotechnology, substantial equivalence presumes that novel
nanoparticles are similar enough to their larger-scale particles that
they do not warrant new toxicology studies. One of the distinc-
tive features of nanomaterials is that they have properties that are
different from those of the analogous bulk material; substantial
equivalence fails to take this feature into account. The suggested
moratorium would remain in effect until scientific communities
could come together to develop and adopt monitoring mecha-
nisms and reporting procedures in a “precautionary principle”
approach to regulatory governance.
Understanding societal implications and addressing societal
concerns about nanotechnology was also a prominent topic within
the US government prior to the legislation enactment. It was a
frequent topic of discourse among the Government agencies
involved directly or indirectly in nanotechnology. Also, the
National Science Foundation (NSF) held a national conference in
2000 and issued a subsequent report in 2001 on the topic of
societal implications of nanotechnology. More than 50 distin-
guished professionals and executives from government and
national laboratories, academic institutions, and the private sector
were among the conference participants and contributors. In

April of 2003, the US Congress held a public hearing on the soci-
etal implications of nanotechnology signaling recognition among
legislators that societal concerns were an important consideration
that needed to be addressed publicly.
The NRDA contains provisions outlining how the sociotech-
nical integration is to be accomplished. General strategies encom-
pass what can be termed both “wide” and “deep” integration,
where “wide” consists of research into societal concerns and dis-
semination thereof, and “deep” consists of feeding research on
societal concerns directly into the NNP including the nanosci-
ence research and development itself. The interdisciplinary socio-
technical integration potentially allows research on societal
considerations to shape the course and outcomes of nanotechnol-
ogy research and development. As such, it envisions a new form
of scientific research in which explicitly “societal” considerations
manifestly influence the design and pursuit of scientific research
and the technology it is meant to enable.
2.1.3. Coexisting In total, the legislation is an acknowledgement that the success of

or Competing Mandates any federal nanotechnology program will not occur solely based
on best efforts to increase the pace of scientific discoveries and of
technology developments. It is a recognition, at least rhetorically,
that a broad range of legitimate societal concerns exist, some of
which could manifest as health and environmental product-related
issues, choice and governance issues, and distribution of benefits
and burdens, to name a few examples. Any of these concerns,
whether “real” or “perceived” (12), could influence public trust,
and hence commercial success. On this view, socially acceptable
outcomes and commercially robust products can be seen to result
from a dual focus on economic and societal considerations of
nanotechnology.
Yet efforts to attempt a dual approach that combines acceler-
ated economic promotion with more deliberative precautionary
methods could manifest as dueling pressures on laboratory
researchers and administrators, who may be confronted with what
appears to be a largely irreconcilable tension between these two
policy objectives. Perhaps the key difference between the two
objectives is in how societal concerns are factored in to nanotech-
nology development. In the traditional economic-promotion
approach to R&D, societal concerns are to be corrected by mech-
anisms that are seen to be external to the laboratory, such as mar-
ket forces and regulation. In contrast, the sociotechnical
integration approach present in the NRDA would be an internal
mechanism that encompasses and intentionally addresses societal
concerns during R&D decisions.
This type of integrated approach represents a small but grow-
ing trend in US federal science and technology policy. Yet none
of the other programs that have attempted to employ it are as

explicit or as high level as the NRDA. The primary previous
attempt was the Ethical, Legal, and Societal Implications (ELSI)
program of the US Human Genome Project (HGP). The HGP
mandate to examine and consider the ethical, legal, and social
implications was thought by the program leaders to be both
visionary and unique (13). The ELSI program, however, has been
criticized for lack of integrative outcomes and in general for fail-
ing to fulfill its mandate (13, 14). The NNI has funded two
Centers for Nanotechnology in Society, one at Arizona State
University (CNS-ASU) and one at the University of California at
Santa Barbara (CNS-UCSB). In particular, the CNS-ASU employs
an integrative approach to research known as “Real-Time
Technology Assessment” (15) and, more recently, has developed
the strategic vision of “Anticipatory Governance” (16).
2.1.4. Nanotechnology Since its authorization, there have been a number of reviews of
Legislation Reauthorization the NRDA program performed – some as specified in the NRDA
legislation, others independently and externally organized. In
2005, the President’s Council of Advisors on Science and
Technology (PCAST), an outside advisory board designated in
NRDA legislation to provide biennial assessments of the NNI to
Congress, acknowledged in its first report that current knowledge
and data to assess the actual risks posed by nanotechnology prod-
ucts were incomplete (17). This point was reiterated by House
Science and Technology Committee Chairman Bart Gordon in a
press release issued after a 2005 committee hearing on the topic
There seems to still be ample unanswered questions in this field, but
what is clear is that commercialization of the technology is outpacing
the development of science-based policies to assess and guard against
adverse environmental, health and safety consequences. The horse is
already out of the gate... Prudence suggests the need for urgency in
having the science of health and environmental implications catch up
to, or even better surpass, the pace of commercialization (18).
Later that same year, the Nanotechnology Environmental and

Health Implications (NEHI) Working Group was established to
provide an infrastructure for coordination with and between
Federal agencies focusing on nanotechnology environmental,
health, and safety research and programs. One year later, a com-
prehensive examination of the NNP was conducted by the
National Research Council of the National Academies of Science
(NAS) per their legislative directive to perform a triennial review.
Their report noted that there was very little published research
addressing the toxicological and environmental effects of engi-
neered nanomaterials and that environmental, health, and safety
issues were of “significant concern to and a topic of serious
discussion by government agencies and commissions, nongovern-
mental organizations (NGOs), the research community, industry,
insurers, the media, and the public” (19). According to the report,
effective solutions required a balancing of promotion with that

of precaution and recommended NEHI facilitate research and
development in a full life-cycle analysis of the precautionary
aspects. In February of 2008, NEHI released its report defining
an environmental, health, and safety research strategy and calling
for the six regulatory agencies in the NNI to work individually
and jointly to implement the strategy (20). A subsequent 2008
National Academy of Science report delivered harsh criticism of the
NEHI plan concluding that there was no strategy in place (21).
Reports, analysis, and testimony from nongovernmental sources
contained similar conclusions and recommendations. A report by
the Project for Emerging Nanotechnologies (22) argued that bet-
ter and more aggressive oversight and new resources were needed
to manage the potentially adverse effects of nanotechnology and
promote its continued development. In its 2007 nanotechnology
policy report, Greenpeace proclaimed that “no regulatory frame-
work has been developed to address the emerging issues” (23).
Richard Denison, a Senior Scientist at NGO Environmental
Defense Fund and former analyst with the US Congress Office
of Technology Assessment leveled a succinct summation of the
criticisms:
NNI and many of its member agencies are talking and writing a great
deal about the need to address nanotechnology’s risks as well as ben-
efits…But there is a continuing disparity between NNI’s words and
actions (24).
Denison reiterated the NRC report call for “a balanced approach

to addressing both the applications and implications of nanotech-
nology [as] the best hope for achieving the responsible introduc-
tion” of nanotechnology products. In his 2008 Senate committee
testimony, Matthew Nordan, President of Lux Research, echoed
this sentiment noting that the current ambiguity and the “glacial
pace” of setting specific regulatory guidelines is becoming a gat-
ing factor for commercialization (25).
On January 15, 2009, the US House of Representatives introduced the
National Nanotechnology Initiative Amendments Act of 2009 (H.R.
544). In February of 2009, the legislation was passed by the House
without amendment and forwarded to the Senate Commerce, Science
and Transportation. The bill reauthorizes and makes incremental chang-
es to several key provisions of the NRDA. One intention of the reau-
thorization bill, as passed in the House, was to better address environ-
mental, health, and safety (EHS) issues associated with nanotechnology
while continuing to encourage promotion of the commercialization of
the technology for economic growth and competitiveness. As stated
by House Science and Technology Committee Chairman Bart Gordon
(26) in conjunction with the passage of the House bill in 2009:
It is important that potential downsides of the technology be ad-
dressed from the beginning in a straightforward and open way, both to
protect the public health and to allay any concerns about the validity of
the results. A thorough, transparent process that ensures the safety of new
products will allow both the business community and the public to benefit
from the development of these new technologies. (Emphasis added).
In summary, much of the reauthorization discourse has been

directed toward developing a national governance framework
with coordination amongst agencies and associated increase in
funding to better address the precautionary aspects contained in
the NRDA mandate to ensure consideration of the ethical, legal,
environmental, and other appropriate societal concerns during
nanotechnology development. The criticisms of the NNI’s
approach to responsible implementation suggest that this empha-
sis has received less attention and may be in direct competition to
the emphasis on rapid implementation. It is also noteworthy that
the reauthorization discourse focused on a more traditional top-
down approach than that outlined in the NRDA’s sociotechnical
integration mandate. The next section describes a research proj-
ect that investigates the possibility and utility of sociotechnical
integration and then turns to a limited analysis of one of the case
studies it has supported.
2.2. Socio-Technical The US legislative mandate for sociotechnical integration during

Integration Research nanotechnology R&D has opened up new opportunities to design
Project and conduct experiments aimed at assessing the possibility and
utility of sociotechnical integration to influence the direction of
R&D. One such undertaking, the Socio-Technical Integration
Research (STIR) project, is a three-year program that is adminis-
tered by the Center for Nanotechnology in Society at Arizona
State University. STIR is funded by the National Science
Foundation (NSF #0849101) with the specific objective to assess
and compare the varying pressures on and capacities for labora-
tory researchers to integrate broader societal considerations into
their work. STIR places ten doctoral students into 20 laboratories
in ten countries on three continents to conduct 20 “laboratory
engagement studies,” a cutting edge form of collaborative par-
ticipant observation (27). The STIR method builds upon ethno-
graphic qualitative research, a methodological paradigm pioneered
by anthropologists and sociologists in the early twentieth century
(28, 29). Ethnography uses extended, primarily participant obser-
vation, to examine the “shared patterns of behavior, beliefs, and
language” of a “culture-sharing group” (29). Traditional labora-
tory studies employed an ethnographic method for studying sci-
ence by examining the internal dynamics of scientific work
through in situ observation “as it happens” and were pioneered
by sociologist of science Bruno Latour (30–32).
The laboratory engagement study also transforms the “reflex-
ive ethnography,” which, in Woolgar’s account, focuses on the
reasoning practices used within the research laboratory to “gen-
erate awareness of reasoning practices as they are deployed in
analysis” (32). Within STIR, the reflexive awareness is not only
applied by the ethnographer to their own thinking about the
phenomena they observe but also accomplished through an
DO
M W
EA
TR
NS
What Whether
U PS
TRE
R&D to to adopt
authorize? R&D
AM
outputs?
How
to implement
R&D?
MID
STREAM
Fig. 1. Stages of science and technology governance (Adapted from STIR: Socio-Technical
Integration Research Project Description, p. 6).
interdisciplinary collaboration between natural and social scientists.

Thus, it is also a methodological practice of introducing ethno-
graphic observations and findings into the laboratory research
context itself – both for verification and so as to stimulate mutual
learning and reflection by both parties to the sociotechnical
collaboration (27).
STIR studies take place over a 4-month period and utilize the
novel methodological approach of midstream modulation. As
shown in Fig. 1, within laboratories research and development
decisions are conceptually situated in the midstream of the sci-
ence and technology governance process, occurring between
upstream policy decisions and downstream regulatory and market
activities. Midstream agents, including those who make basic
research decisions, thus perform the functional role of imple-
menting authorized research agendas. Research developments,
which are measured by mapping the evolution of research deci-
sions over time, are theorized to be modulated or incrementally
shaped by a variety of institutional, social, and cognitive factors.
Modulation at the midstream is posited to occur in three succes-
sive stages: de facto modulation, the factors that influence deci-
sions; reflexive modulation, laboratory practitioners’ awareness of
these factors and of their own roles within larger social systems;
and deliberative modulation, in which scientists consciously form
decisions that are tempered by a reflexive awareness of these
factors. Thus, midstream modulation provides a mechanism for
evaluating and adjusting research decisions during the research
process and constitutes a bottom-up approach for shaping
research and development directions in light of relevant societal
considerations – what has been termed “governance from
within” (33). This unique application of reflexive ethnography

to science and engineering research decisions serves to interact
with the content of research decisions, thereby in theory lending
visibility to both the promotional and precautionary influences
on research decisions.
STIR engages in midstream modulation through interdisci-
plinary collaboration between social and natural scientists. Despite
longstanding calls for such sociotechnical integration and collab-
orations (14), there have been very few laboratory engagement
studies conducted using this approach. The NRDA mandate
affords a renewed call for and recognition of the need for socio-
technical integration in nanotechnology development. Various
types of relationships between social and natural science have
recently been initiated in emerging technology research programs
including nanotechnology, genetic engineering, and synthetic
biology. Often the failed genetically modified crop debate is used
as an example of the need and justification for including social
scientists in these primarily public-funded research programs. In
these relationships, Calvert and Martin (34) suggest two different
roles for social scientists: the social scientist can perform the role
of either a “contributor” or a “collaborator.” A contributor is one
who contributes to (at times as a representative of the “public”),
facilitates the discussions of, and studies the ethical, legal, and
social implications of research. In contrast, a collaborator is one
who is involved with the research and interacts with the research-
ers in ways that can potentially shape the research agenda and
influence the research direction. The collaborator role STIR sci-
entist has shown some positive indications of success in initial
laboratory engagements (27, 35). The next section offers findings
drawn from one such engagement study.
2.3. STIR Case Study: The site for one STIR case study was a company, Rocky Mountain
Rocky Mountain Nanomaterials (see Note 2), producing novel nanomaterials using
Nanomaterials a patented application technology. Nanomaterials can be consid-
ered a nanotechnology sector with numerous applications across
the spectrum from biomedical, energy, and various technology and
industrial markets. A report by market analyst firm Lux Research
identifies nanomaterials at the beginning of the nanotechnology
value chain (36). Thus, the nanomaterials sector represents a major
portion of the economic potential for nanotechnology and is
therefore posited to exhibit a number of influences for economic
promotion. In addition, according to the Nanotechnology
Industries Association (NIA), a UK-funded organization formed
in 2005 to establish a framework for the safe, sustainable, and
socially supportive development of nanotechnology, a complex
and convoluted mixture of regional, national, trade, industry, and
international voluntary and regulatory governance initiatives for
nanomaterials exists (37). These disparate governance initiatives
Fig. 2. Six dimensions of risk for Nanomaterials (see ref. 37).
need to address six dimensions of risk identified for nanomaterials

(see Fig. 2). The nanomaterials sector is thus also posited as likely
to exhibit evidence of influences for precaution, hence making it a
viable source for examining the nature of interactions between the
NRDA’s promotional and precautionary mandates.
Rocky Mountain Nanomaterials is a university spin-out, a
company which emanated from research conducted at a univer-
sity in Colorado, which utilizes a patented process to create novel
nanomaterials. University spin-outs have been shown to be impor-
tant contributors in the emergence of nanomaterial applications
(38). The company has been in business since 2002 and has a
staff of four PhD scientists, one operations manager, and two
founding executives. In addition, the laboratory has strategic

partnerships with departments at the university where two addi-
tional founders of the company work.
This laboratory engagement study was conducted from April
through August of 2009. The length of the study was predeter-
mined and was consistent across the STIR laboratory sites. The
primary empirical data collection methods were participant obser-
vation and semistructured and unstructured interviews. The
researcher met individually each week with two scientists and par-
ticipated in the weekly laboratory project review meetings. One
scientist (“C4”) received his PhD in Chemical Engineering from
a university in the US Rocky Mountain region; the other scientist
(“M1”) received his PhD in Electro Chemistry from the same
university. Interview responses and observations were recorded in
a field notebook, and many of the individual interviews, with the
scientist’s permission, were digitally recorded. Documents,
obtained with permission from the laboratory, and content from
archival research form the remainder of the empirical data
sources.
Interviews were guided by a protocol developed during the
STIR pilot study (27). The model (see Fig. 3) consists of four
distinct conceptual components intended to describe research
decisions as well as to capture and make visible – to both the
investigating STIR scientist and to the participating laboratory
scientists – the de facto influences of societal considerations dur-
ing research activities. The model was often utilized during the
study to initiate the semistructured interviews with the scientists.
The remainder of this chapter presents the results of a limited
study of a subset of the data generated in the Rocky Mountain
Fig. 3. Decision protocol components (Adapted from STIR: Socio-Technical Integration

Research Project Description, p. 7).
Nanomaterials case study. Rather than attempting to investigate

the possibility and utility of sociotechnical integration, this par-
ticular study sought to develop broad classification categories
which could characterize and relate the “nontechnical” or “societal”
influence on technical research decisions.
2.3.1. Findings: Decision A preliminary analysis of a subset of the data – drawn from inter-
Influences views, lab meetings, and informal conversations – was conducted
using Conceptually Clustered Matrix display format (39). The
data were placed into two major categories of influence: external
and internal. Internal influences originate from the people and
the policies within the company and indicate cultural norms that
can guide decisions and behaviors. External influences originate
from outside the company and indicate the institutional context
of the innovation system, which can also guide decisions and
behaviors. Within each category, the type of influence was catego-
rized as either “technical” or “societal.” From this grouping of
empirical data, four distinct societal influences on the laboratory
decisions emerged: economic, intellectual property, university
relations, and environment, health, and safety (EHS). Of these,
economic considerations had the greatest number of instances
and dominated the external societal influences; however, it
occurred only in a few instances in the case of internal societal
influences. In contrast, university relations considerations were
a much stronger internal societal influence but only occurred in a
small number of instances as an external societal influence. That
university relations were stronger internally is to be expected
given the fact that the laboratory is a university spin-out and
maintains ongoing ties to the university for research. Similarly,
intellectual property was mentioned more often as an internal
rather than as an external societal consideration. This may be due
to the fact that intellectual property serves as a competitive advan-
tage and as a barrier to entry into the market for others, thus
having a significant potential economic impact.
EHS was the only consideration mentioned by all partici-
pants, and there was a near balance in the number of EHS consid-
erations between internal and external societal influences. For
example, a new opportunity required the use of hydrazine, an
inorganic chemical compound. The researchers were aware of the
potential negative and positive external societal considerations of
hydrazine given its use in a range of applications from rocket fuel
to pharmaceuticals to automotive airbags. They were aware that
the US Occupational Safety and Health Administration (OSHA)
was looking at toxicology, an internal EHS consideration. During
the discussions of the opportunity, the question came up about
the safe handling of hydrazine, an example of an internal EHS
consideration. One of the participants agreed to make contact
with the largest producer of hydrazine to find out the standard
safe handling practices. In another example, one of the participants

was looking at their current use of carbon in a nanomaterials
application. He expressed an external EHS consideration and
concern that in this specific application their existed the possibility
of overheating potentially causing the nanomaterial to catch on
fire because carbon, which was selected for this application because
of its high conductivity, is combustible at high temperatures.
In addition to these four categories of societal influences, the
conceptually clustered matrix also yielded a broader societal influ-
ence theme, which is here termed “green-nano.” This theme
appeared in both economic as well as EHS considerations. A major-
ity of the external economic considerations were focused on so
called “green energy,” energy developed from renewable sources.
This focus can more than likely be associated with the current
Colorado and national priorities of becoming a “green economy”
by both reducing the country’s dependence upon foreign oil and
reducing the carbon output during the production of energy
from fossil fuels. It must be noted however that the only instance
when “green-nano” was used in this latter sense was in the case of
“carbon avoidance” as an internal EHS consideration. For the
most part when “green-nano” was used as an economic consider-
ation it was in reference to funding that could be obtained from
government, venture capital, and/or strategic customers by pur-
suing green energy opportunities. (This finding is confirmed by
one researcher’s statement, “We can raise money with green” (see
Note 3) and by another participant’s statement that a carbon
monetization mechanism called carbon credits could be a “cash
cow” (see Note 4).) Notably, “green-nano” was employed as an
EHS consideration both in pursuit of a green energy funding
opportunity for the lab and in critically questioning the same
opportunity. (The former inference is based on the potential
funding source’s emphasis on no carbon byproducts, while the
latter is derived from one participant’s statement, “How green is
it when it uses nasty precursors?” (see Note 5)).
2.3.2. Analysis of Findings Analysis of the data subset produced four categories of “societal”
(See Note 6) influence on laboratory decisions in a nanomaterials laboratory:
economic, intellectual property, university relations, and EHS
considerations. Of these, both intellectual property and university
relations emerged more in relation to economic justification (in
cases of competitive differentiation and outsourcing partnership).
Accordingly, these two categories fall primarily under economic
promotion and appear less frequently under societal precaution.
The analysis did find evidence of societal influences present
in research decisions; however, economic considerations by far
outweighed any other societal consideration. Thus, within the
scope of the nanomaterials sector in which the Rocky Mountain
Nanomaterials company operates, promotion far outweighs

precaution.
This limited analysis did discover some product-related societal
influences concerning EHS; however, broader issues including
those of power, choice, and distribution were usually not men-
tioned by research participants without prompting by the STIR
scientist. It is intuitive that promotion considerations would be
strongly influential given the fact that this laboratory is a startup
company that uses a combination of market signals, customer
opportunities, and government sources of funding to derive com-
pany growth and survival. It is also not unreasonable to anticipate
that there would be a deficit of precautionary influences given the
lack of clear downstream regulatory mechanisms governing risk,
as previously detailed, and given a relatively unprecedented legis-
lative directive for sociotechnical integration that in effect requires
significant changes to the institutional settings governing nano-
scale research and development. Such changes in the norms, val-
ues, and rules that shape organizational behavior regarding
economic promotion and societal consideration would need to
occur not only at the microlevel of laboratories but also at the
mesolevel of institutional environments that constrain or encour-
age innovation (40–42). Institutions and organizations constitute
primary elements of an established innovation system (43).
Though innovation systems are evolutionary in many ways, they
can also be slow to change and adapt in others (41, 42, 44). New
technologies produce pressures on the institutions and organiza-
tions within a sector to change or adapt in response to new con-
cerns such as the precautionary and promotional considerations
of US legislation. Institutional and organizational response to
these pressures is distinct within a sector and is based in part on
the transformative capacity of the technology, whether it is endog-
enous or exogenous to the sector, and on the sectoral adaptabil-
ity, the supportive or disruptive effects of the technology on the
sector (45). The overall commercial success or failure of a techno-
logical regime can be a function in part of in what ways an innova-
tion system retains old characteristics and in what ways it remains
flexible and open to adaptation.
A thorough examination of the source and nature of the
mesolevel institutions that shape the economic-promotion and
societal-precaution influences would be a next reasonable step
toward creating insight and understanding for policy and prac-
tice. Review of the initiatives of the US agencies involved in nano-
materials risk governance – the Federal Drug Administration, the
Environmental Protection Agency, and the National Institute for
Occupational Safety and Health – may provide insights into ways
the nanomaterials sector may be responding to precautionary
issues. A report from a UK pilot study (46) into how public policy
might encourage the dual objective of promotion with precaution

approach specifically for nanomaterial research and development
could provide additional insight.
3. Notes
1. The material in this section draws heavily upon material in

ref. 2.
2. The company’s name has been changed for confidentiality
purposes.
3. Participant interview, May 6, 2009, Laboratory site.
4. Participant interview, May 6, 2009, Laboratory site.
5. Project Review meeting, May 6, 2009, Laboratory site.
6. The limited analysis applied to the subset of data from the
Rocky Mountain Nanomaterials case study did not indicate
how best to characterize the relation between the two policy
goals of promotion and precaution. Whether these two goals
coexist or compete remains a broad question that requires more
nuanced analysis. While the statements “we can raise money
with green” would seem to indicate coexistence, the statement
“how green is it?” implies competition. Similarly, this limited
analysis did not seek to provide insight into the capacities of
laboratories to engage in sociotechnical integration attempts or
into how such capacities might be enhanced. We note that
throughout this case study, broader societal dimensions of
research decisions that were evident to the STIR scientist were
not always indicated by the laboratory scientists. This was taken
not to be due to intentional efforts by laboratory participants
to ignore or negate these dimensions; rather, such dimensions
simply did not appear to be in the de facto cognitive frame of
decision alternatives of the company scientists. Once societal
considerations were brought to the attention of a lab scientist
by the STIR researcher through the use of the decision proto-
col, however, opportunities arose to discuss these dimensions
further. Over time, these discussions extended from single deci-
sions to more encompassing ones related to industry, market,
and society. Other forms of analysis of the data set, including
more narrative-based accounts, are therefore likely to produce
more penetrating insights into the possibility and utility of
sociotechnical integration mandated by the US legislation and
investigated by projects like STIR. It is also worth noting that,
in other settings, nanoscientists have been documented to be
more concerned about some nanotechnology-associated risks
than are members of the public (47).
Acknowledgments
The authors gratefully acknowledge the participants from the

laboratory of this case study whose willingness to mutually share,
discuss, and explore the content and context of laboratory deci-
sions greatly increased the value of the experience. This material
is based upon work supported by the US National Science
Foundation under Grant No. #0849101.
Appendix 1.
Program Activities
of the National
Nanotechnology 1. Developing a fundamental understanding of matter that
Program Laid Out enables control and manipulation at the nanoscale.
in Section 2(b) of 2. Providing grants to individual investigators and interdisci-
the 21st Century plinary teams of investigators.
Nanotechnology 3. Establishing a network of advanced technology user facilities
Research and and centers.
Development Act 4. Establishing, on a merit-reviewed and competitive basis, inter-
disciplinary nanotechnology research centers, which shall
(a) Interact and collaborate to foster the exchange of techni-
cal information and best practices.
(b) Involve academic institutions or national laboratories and
other partners, which may include States and industry.
(c) Make use of existing expertise in nanotechnology in their
regions and nationally.
(d) Make use of ongoing research and development at the
micrometer scale to support their work in nanotechnology.
(e) To the greatest extent possible be established in geo-
graphically diverse locations, encourage the participation
of Historically Black Colleges and Universities that are
part B institutions as defined in section 322(2) of the
Higher Education Act of 1965 (20 U.S.C. 1061(2)) and
minority institutions [as defined in section 365(3) of that
Act (2 U.S.C. 067k(3))], and include institutions located
in States participating in the Experimental Program to
Stimulate Competitive Research (EPSCoR).
5. Ensuring US global leadership in the development and appli-
cation of nanotechnology.
6. Advancing the US productivity and industrial competitiveness
through stable, consistent, and coordinated investments in long-
term scientific and engineering research in nanotechnology.
7. Accelerating the deployment and application of nanotechnology

research and development in the private sector, including
startup companies.
8. Encouraging interdisciplinary research, and ensuring that
processes for solicitation and evaluation of proposals under
the program encourage interdisciplinary projects and
collaborations.
9. Providing effective education and training for researchers and
professionals skilled in the interdisciplinary perspectives nec-
essary for nanotechnology so that a true interdisciplinary
research culture for nanoscale science, engineering, and tech-
nology can emerge.
10. Ensuring that ethical, legal, environmental, and other appro-
priate societal concerns, including the potential use of nano-
technology in enhancing human intelligence and in developing
artificial intelligence which exceeds human capacity, are con-
sidered during the development of nanotechnology by
(a) Establishing a research program to identify ethical, legal,
environmental, and other appropriate societal concerns
related to nanotechnology, and ensuring that the results
of such research are widely disseminated.
(b) Requiring that interdisciplinary nanotechnology research
centers established under paragraph (4) include activities
that address societal, ethical, and environmental concerns.
(c) Insofar as possible, integrating research on societal, ethi-
cal, and environmental concerns with nanotechnology
research and development, and ensuring that advances in
nanotechnology bring about improvements in quality of
life for all Americans.
(d) Providing, through the National Nanotechnology
Coordination Office established in section 3, for public
input and outreach to be integrated into the Program by
the convening of regular and ongoing public discussions,
through mechanisms such as citizens’ panels, consensus
conferences, and educational events, as appropriate.
11. Encouraging research on nanotechnology advances that
utilize existing processes and technologies.
References
1. US Congress (2003) 21st Century anotechnology research and development.
n
Nanotechnology Research and Development Sci. Public Policy 33, 5–16.
Act of 2003, Public Law no. 108–153. 3. Meyer, M. and Persson, O. (1998)
2. Fisher, E. and Mahajan, R. L. (2006) Nanotechnology-interdisciplinarity, patterns
Contradictory intent? US federal legislation of collaboration and differences in applica-
on integrating societal concerns into tion. Scientometrics 42, 195–205.
4. Rafols, I. and Meyer, M. (2007) How 18. US House of Representatives: Committee

cross-disciplinary is bionanotechnology? on Science and Technology (2005) Research
Explorations in the specialty of molecular is key to safety when tackling unknown
motors. Scientometrics 70, 633–650. aspects of nanotechnology. Press release 17
5. Rafols, I. and Meyer, M. (2010) Diversity and November.
network coherence as indicators of interdisci- 19. National Research Council; Division on
plinarity: case studies in bionanoscience. Engineering and Physical Sciences; Committee
Scientometrics 82, 263–287. to Review the National Nanotechnology
6. Lux Research (2007) The nanotech report. 5th Initiative (2006) A matter of size: triennial
Ed., v. 1, New York, NY. review of the national nanotechnology initiative.
7. Roco, M. C. (2003) Government nanotech- p. 74, National Academics Press, Washington,
nology funding: an international outlook. DC.
National Science Foundation. 30 June. 20. NEHI – Nanotechnology Environmental
http://www.nsf.gov. Health Implications Working Group (2008)
8. Roco, M. C. (2003) Broader societal issues of National nanotechnology initiative strategy for
nanotechnology. J. Nanopart. Res. 5, nanotechnology-related environmental, health,
181–189. and safety research. National Nanotechnology
Coordination Office, Arlington, VA.
9. Hines, C. (2003) Nanotechnology: small mir-
acles foreseen; some lawyers say trend will 21. National Research Council; Division on
rival semiconductors and the Internet. New Engineering and Physical Sciences; Committee
York Law J. 13, 5 November. to Review the National Nanotechnology
Initiative (2008) Review of federal strategy for
10. Mnyusiwalla, A., Daar, A. S., and Singer, P. A. nanotechnology-related environmental health
(2003) “Mind the gap”: science and ethics in and safety research. National Academies Press,
nanotechnology. Nanotechnology 14, 9–13. Washington, DC.
11. ETC Group (2003) Nanotech and the pre- 22. Davies, J. C. (2006) Managing the effects of
cautionary prince. Available at http://www. nanotechnology. Project on Emerging
etcgroup.org. Nanotechnologies. http://www.nanotech-
12. Norden, M. (2008) Congressional testimony, project.org.
Hearing before the Committee on Commerce, 23. Greenpeace (2007) Nanotechnology: policy
Science, and Transportation. US Senate. 24 and position paper. http://www.greenpeace.
April. org.
13. McCain, L. (2002) Informing technology 24. Denison, R. (2007) Congressional testimony,
policy decisions: the US Human Genome Hearing before the Committee on Science
Project’s ethical, legal, and social implications and Technology, US House of Representatives,
programs as a critical case. Technol. Soc. 24, 110th Congress. 31 October.
111–132.
25. Nordan, M. (2008) Congressional testimony,
14. Fisher, E. (2005) Lessons learned from the Hearing before the Committee on Commerce,
ethical, legal and social implications program Science, and Transportation, US Senate,
(ELSI): planning societal implications research 110th Congress. 24 April.
for the national nanotechnology program.
Technol. Soc. 27, 321–328. 26. US House of Representatives: Committee on
Science and Technology (2009) Bill intro-
15. Guston, D. H. and Sarewitz, D. (2002) Real- duced to ensure safety of nanotechnology,
time technology assessment. Technol. Soc. 24, transparency of research. Press release 15
93–109. January.
16. Barben, D., Fisher, E., Selin C., and Guston, 27. Fisher, E. (2007) Ethnographic invention:
D.H. (2008) Anticipatory Governance of probing the capacity of laboratory decisions.
Nanotechnology: Foresight, Engagement, NanoEthics 1, 155–165.
and Integration. in: The new handbook of sci-
ence and technology studies (Hackett, E. J., 28. Sanday, P. R. (1983) The Ethnographic
Amsterdamska, O., Lynch, M. E., and Paradigms. in: Qualitative methodology (Van
Wajcman, J., eds.) MIT, Cambridge, MA. Maanen, J., ed.) Sage, Newbury Park, CA.
17. President’s Council of Advisors on Science 29. Cresswell, J. W. (2007) Qualitative inquiry
and Technology (2005) The national nano- and research design: choosing among five differ-
technology initiative at five years: assessment ent approaches. Sage, Thousand Oaks, CA.
and recommendations of the National 30. Godfrey-Smith, P. (2003) Theory and reality:
Nanotechnology Advisory Panel. May. an introduction to the philosophy of science. The
Available at http://www.nano.gov. University of Chicago Press, Chicago, IL.
31. Latour, B. and Woolgar, S. (1979) Laboratory 41. Nelson, R. R. (1994) The co-evolution of
life: the social construction of scientific facts. technology, industrial structure, and support-
Sage, Beverly Hills, CA. ing institutions. Ind. Corp. Change 3,
32. Woolgar, S. (1982) Laboratory studies: a 417–419.
comment on the state of the art. Soc. Stud. Sci. 42. Nelson, R. R. (1995) Co-evolution of indus-
12, 481–498. try structure, technology, and supporting
33. Fisher, E., Mahajan, R. L., and Mitcham, C. institutions, and the making of comparative
(2006) Midstream modulation of technology: advantage. Int. J. Econ. Bus. 2, 171–184.
governance from within. Bull. Sci. Technol. 43. Edquist, C. (1997) Systems of innovation:
Soc. 26, 485–496. technologies, institutions, and organizations.
34. Calvert, J. and Martin, P. (2009) The role of Cambridge University Press, Cambridge,
social scientists in synthetic biology. EMBO UK.
Rep. 10, 201–204. 44. Nelson, R. R. and Winter, S. G. (1982) An
35. Schuurbiers, D. and Fisher, E. (2009) Lab- evolutionary theory of economic change.
scale Intervention. EMBO Rep. 10, 424–427. Harvard University Press, Cambridge, MA.
36. Lux Research (2004) Sizing nanotechnology’s 45. Dolata, U. (2009) Technological innova-
value chain. New York, NY tions and sectoral change. Transformative
37. NIA – Nanotechnology Industries Association capacity, adaptability, patterns of change: an
(2007) Risk classification made easy. Presented analytical framework. Res. Policy 38,
at NanoEXPO 2007, May 16, in Nottingham, 1066–1076.
England. 46. Nightingale, P., Morgan, M., Rafols, I., and
38. Libaers, D., Meyer, M., and Geuna, A. (2006) van Zwanenberg, P. (2008) Nanomaterials
The role of university spinout companies in an innovation systems: their structure, dynamics
emerging technology: the case of nanotech- and regulation. Report for the Royal Commission
nology. J. Technol. Transf. 31, 443–450. on Environmental Pollution. Science Policy
39. Miles, M. B. and Huberman, A. M. (1994) Research Unit, Brighton, UK.
Qualitative data analysis. Sage, Thousand 47. Scheufele, D. A., Corley, E. A., Dunwoody,
Oaks, CA. S., Shih, T.-J., Hillback, E. D., and Guston,
40. Hollingsworth, J. R. (2000) Doing institu- D. H. (2007) Nanotechnology scientists
tional analysis: implications for the study of worry about some risks more than the general
innovations. Rev. Int. Polit. Econ. 7, 595–644. public. Nat. Nanotechnol. 2, 732–734.
Chapter 23
Navigating the Patent Landscapes for Nanotechnology:

English Gardens or Tangled Grounds?
Douglas J. Sylvester and Diana M. Bowman
Abstract
The patent landscape, like a garden, can tell you much about its designers and users: their motivations,
biases, and general interests. While both patent landscapes and gardens may appear to the casual observer
as refined and ordered, an in-depth exploration of the terrain is likely to reveal unforeseen challenges
including, for example, alien species, thickets, and trolls. As this chapter illustrates, patent landscapes are
dynamic and have been forced to continually evolve in response to technological innovation. While
emerging technologies such as biotechnology and information communication technology have chal-
lenged the traditional patent landscape, the overarching framework and design have largely remained
intact. But will this always be the case? The aim of this chapter is to highlight how nanotechnology is
challenging the existing structures and underlying foundation of the patent landscape and the implica-
tions thereof for the technology, industry, and public more generally. The chapter concludes by asking
the question whether the current patent landscape will be able to withstand the ubiquitous nature of the
technology, or whether nanotechnology will be a catalyst for governments and policy makers for over-
hauling the current landscape design.
Key words: Intellectual property, TRIPS, Patent thickets, Patent pools, Trolls, Technology innovation
1. Introduction
One is tempted to think of the patent landscape as a refined

English garden. Views of gently rolling lawns spotted by outcrop-
pings of majestic trees, a few revival buildings, and inundated by
hundreds of floral and shrub varieties might leave the casual
observer with the view that it is entirely organic and naturalistic.
For those who look closer, however, one sees the tenders’ efforts.
The lack of straight lines, walls, or delineated beds masks the
perfect visual delineation of the form – a form evolving over
decades (if not centuries) and one that is largely in balance.
359
360 Sylvester and Bowman
Organic contours hide the hundreds of small decisions that are

continually being made to retain the appropriate balance. But
those decisions are nevertheless made. Although such gardens
may give the impression that they arise “just so,” they hide the
enormous and complicated efforts of their tenders to organize,
weed, and design them.
Since the 1970s, patent gardens have been under continued
attack, and their carefree style seems under threat as alien species
invade these once tranquil spaces. The dual shocks of software (1)
and much more importantly biotechnology (2) wreaked havoc on
these once tranquil and seemingly unchanging spaces. Well-
tended beds, ancient perennials, and majestic arbors were threat-
ened by rapidly growing and unanticipated thickets and brambles
(3–6). Rolling meadows were quickly dotted with pitfalls, and
once-languid pools, (7) now choked with unforeseen infestations,
threatened to become unsightly and unrecoverable bogs and
quagmires (8). Worst of all, these gardens (and especially their
beautiful marble bridges) were invaded by trolls (9, 10)!
In these dark days, the garden’s tenders created new tools,
brought in help from abroad (as discussed in Subheading 2.3)
and, although forced to make certain concessions to these alien
species (see, generally, the Uruguay Round Agreements Act of
1994), finally succeeded by the end of the millennia to return
much of the garden to its former apparent placidity. Sure some
ancient varietals were replaced by new and, for a time, foreign
blossoms. A few hedges, long put up to keep out unwanted visi-
tors, or at least to make the entry difficult, were, if not fully
removed, trimmed a tad. Finally, it appears that furtive efforts
were made to gather pitchforks and torches to drive off those
pesky trolls – although some are obviously lurking underneath
some of the murkier bridges (10). In the end, the threat, although
not entirely gone, seemed largely under control and much of
what we had always expected in our patent garden remained
familiar and friendly.
However, new threats are looming on the edges of our serene
plot. Tendrils of invasive species can be seen sprouting all over the
garden, and destructive vines threaten the integrity of the garden’s
walls. In short, nanotechnology brings with it the potential to upset
not only some aspects of the patent garden, but may also force a
complete rethinking of its function and form (see Note 1).
Biotechnology thickets may have grown over some beloved blooms,
but nanotechnology’s brambles threaten to take down the entire
field (6, 8). Those seeking refuge in the cooling waters of the gar-
den’s pools, now clear after a decade of invasion, may find them
once again choked to a vivid green. And, horror, the trolls appear
to be breeding again! Our patent garden, so perfect in its form to
handle the challenges of past patent revolutions, seems particularly
unable to handle what nanotechnology may be bringing.
Navigating the Patent Landscapes for Nanotechnology 361
Leaving aside (thankfully we can only suppose) the metaphor

of the garden, the increasing pace, complexity, and importance of
technological revolutions has put real pressure on the patent sys-
tem. Revered doctrines, designed for pre- and early industrial inno-
vations, seem quaint (if not dangerous) for these times (6). Patent
institutions, organized around silos of knowledge and focused on
local inventorship, may not be able to stand in the face of massively
complex and global innovations. Finally, the pace of patenting,
both in terms of process and conceptual foundations, seems dan-
gerously ill-suited to technological advances that have the ability to
challenge existing national and international legal frameworks,
including those relating to patents, in the blink of an eye. In numer-
ous other publications, we have examined the many challenges
nanotechnology poses to traditional regulatory structures related
to environmental and human health and safety (see, for example,
(11–15)). In this article, we examine the challenges that nanotech-
nology has and will pose for patent frameworks and institutions.
We already know that biotechnology radically shifted the pat-
ent landscape both in terms of patenting practices of researchers
(16), and the scope and breadth of patentable subject matter
(17–19). In just a few short years, biotechnology and rapidly
advancing pharmaceutical patenting forced a fundamental rethink-
ing of what was and should be patentable (19), as well as substan-
tial hand-wringing about why we allow patenting of socially
beneficial inventions at all (5). Although biotechnology, software,
and pharmaceutical patenting may have spurred a substantial
“rethinking [of] intellectual property rights” (19), much of the
prior system remains in place. Patents are still, largely, national
affairs (20) with massive bureaucratic costs that view patents as
arising from traditional scientific disciplines (21, 22). In addition,
patents are still issued without great oversight and on increasingly
early stage technologies (23). These were, similarly, issues in the
biotechnology and pharmaceutical revolutions of the 1980s – but
the system was able to accommodate these issues without funda-
mentally altering the process and purpose of patenting.
Returning (apologies) to our metaphor – it may be time to
decide whether the garden, as it has been known for centuries,
should be dozed. The garden, whether we are conscious of it or
not, picks winners and losers. It gives preference to those that
bloom first – crowding out latecomers and preventing variations.
Hedges and walls, meant to keep out unwelcome visitors and
maintain tranquility, reduce hybridization, and competition and,
arguably, reduce overall social utility. Finally, those darn trolls
really need to be run out!
In this chapter, we explore the challenges that nanotechnol-
ogy may pose for the traditional patent landscape. In particular, we
wish to address more of the garden metaphors (although,
mercifully, these are not our creations) that speak of potential
thickets, bogs, brambles, quagmires, pitfalls, pools, and other

problems that new technologies may pose for our little patent
garden. To this end, this chapter proceeds in several sections.
In Subheading 2.1, we set out some of the patent basics that
have arisen over the course of the past few centuries. In so doing,
our analysis focuses mainly on the USA, but we do include some
important comparative and multinational aspects of the current
landscape. In Subheading 2.2, we discuss some of the unique
aspects of nanotechnology’s development, current regulation,
and potential that many have predicted will pose real problems
for traditional patent systems. Further, we provide an overview
of research into whether some of those predictions have started
to come true. Subheading 2.3 sets out some of the remaining
challenges that nanotechnology may pose for patent frame-
works. Finally, in Subheading 2.4, we set forth some (admit-
tedly tentative) thoughts about how we may avoid some of the
potential problems with nanotechnology and how the patent
system may need to reform itself to better accommodate
technological invasions that will inevitably occur again.
2. Discussion
2.1. Patent Basics For centuries, the patent system has been predicated on a few
unchanging (and nearly universal) precepts. First, patents are
national in scope (and often favor national inventors) (24, 25).
Second, patents reward innovation by granting rights to inven-
tors (24). Third, patents only apply to inventions and not discov-
eries (24). And, fourth, patents are ultimately intended to benefit
society by encouraging technological innovation and must, there-
fore, seek the balance between encouraging invention and ensur-
ing social benefits through access and use (20, 25). To achieve
these various principles, patent systems around the world have
created both institutions and doctrinal frameworks dedicated to
ensuring their fulfillment.
These precepts were built up during periods of relatively low
patenting and invention. They made sense in that era, but their
relevance today is increasingly coming under question. First, the
national scope of patents produces real inefficiencies in that inven-
tors must seek approval in each nation, greatly increasing the cost
on both inventors and societies to manage that system. In addi-
tion, in an era of increasing global research and patenting, it
makes little sense to continue to favor one citizen inventor over
another merely because of accident of birth.
As a result, there have been some tentative efforts to stream-
line this process. First, the United States Patent Office (USPTO)
and Japanese Patent Office (along with many others) have begun
to take some of the administrative burden off of inventors by
allowing for streamlined patenting through individual treaty

arrangements, or application of the Patent Cooperation Treaty. A
discussion of how these procedures have streamlined the process
is outside the scope of this chapter, but the real point is that it has
not been wildly successful. It is still an extraordinarily complex
and expensive process to engage in global patenting and calls for
reform only continue to grow (26, 27).
A secondary, and closely related, framework for achieving
patent’s precepts is through doctrinal limitations on patentability.
At their base, these rules seek to ensure that only those patents
that add something to the world of knowledge in a given area are
patentable (24). The doctrines of novelty, nonobviousness, and
utility are all, in some respects, attempts to ensure that only those
patents worthy of intellectual property protection are granted a
bundle of legal rights (20, 28). Yet, given patent law’s overall goal
of providing social goods through appropriate incentivized inno-
vation, a fundamental practical principle of nearly every patent
system has been to “grant the patent and let the market figure it
out!” In low-innovation periods, this approach makes perfect
sense (23). In addition, it may even have net benefits in periods
of explosive innovation in technological applications. Where it is
deeply problematic is in times – and you guessed that the nano-
technology revolution may be one of those times – of immense
patenting of basic research and fundamental research tools (6).
Again, this is an issue we take up later in the chapter.
One area of genuine progress in patent law, at least in terms
of overall efficiency and systemic fairness, has been in the area of
doctrinal harmonization. In particular, the World Trade
Organization’s (WTO) Trade-Related Intellectual Property Rights
(TRIPS) Agreement has provided an opportunity for tremendous
progress in the harmonization of basic standards of patentability.
Pursuant to Article 27(1) of the TRIPS Agreement,
patents shall be available for any invention, whether products or proc-
esses, in all fields of technology provided they are new, involve an
inventive step and are capable of industrial application.
While conditions for the granting of a patent and nature of the

exclusive rights will vary between jurisdictions, Mandel (28) notes
that these requirements, and therefore the bundle of rights granted,
are “largely harmonized throughout the world.” Unfortunately,
as hinted above, substantive or doctrinal harmonization has not
been accompanied with procedural efficiencies (25).
Despite these efforts at substantive harmonization, it is
important to note that not all subject matter may be the subject
of a patent grant and that, at its outer limits, the question of what
is patentable is still an open discussion. As highlighted by Article
27 of the TRIPS Agreement, patents may be granted for “any
inventions,” but the Agreement does not define what constitutes
an “invention.” This area of potential disunity has been largely

avoided as most nations have adopted permissive definitions of
patentability (29), thereby ensuring continuation and global
propagation of the “patent now, sue later” mentality noted above.
To be clear, we do not view this practice as necessarily problem-
atic. Indeed, as we note later in this chapter, an equally thorny
problem is making patent grants too difficult and/or expensive to
obtain – because they create disincentives to innovate, rob inven-
tors of the fruits of their labor, or unduly delay beneficial applica-
tions. Nobody said tending the patent garden was easy.
As a result of the efforts of the WTO and, indeed, the USA
through numerous bilateral agreements (30), there is wide con-
sensus on what is not patentable as well as buy-in for the general
principle that most things should be. According to Mandel (28),
laws of nature, natural phenomena, abstract ideas, aesthetic creations,
and information and data per se generally are not patent eligible. Almost
everything else is.
Thus in most jurisdictions, “discoveries” or “products of

nature” fall outside the traditional scope of patentable subject
matter. Despite these similarities, there are some differences.
Eisenberg (2), for example, has observed a “shifting landscape
of discovery in genetics and genomics research” that presents
moral and conceptual difficulties about what is or should be
patentable. In particular, biotechnology (and, to a much lesser
extent, software) forced national patent systems to reevaluate
the scope of what they consider to be patentable subject matter.
For example, in 1976, the Australian Patent Office adopted a
fairly liberal approach to the patenting of living subject matter
when it held that,
living organisms were determined to be patentable provided they were
not in a naturally occurring state and they had improved or altered
useful properties, and not merely changed morphological characteristics
which had no effect on the working of the organism (31).
While the USA has adopted a similarly liberal approach, the posi-
tion of these two jurisdictions may be contrasted to that of, for
example, the EU. Pursuant to the European Parliament and
Council Directive 98/44/EC of 6 July 1998 on the legal protection
of biotechnological inventions, the European Union places specific
limitations on the patenting of, for example, plants and animal
varieties (see, for example, Article 4).
Nanotechnology, and other emerging technologies such as
synthetic biology, would appear to have the potential to further
impact this landscape. Where biotechnology created fundamental
challenges for many on the moral nature of what is patentable
(32, 33), nanotechnology seems less controversial although the
potential for ethical challenges remains. As a result, the real chal-
lenges for patenting of nanotechnology will flow from doctrinal
conceptions of novelty and nonobviousness, and the fact that

nanotechnology does not fall neatly into any one silo.
2.2. Nanotechnology Nanotechnology is a field of technological effort that holds tre-

Background mendous promise as well as potential peril (34, 35). Despite the
concerns of many groups and governments, the regulatory
response to nanotechnology has been, largely, one of research
and development. While nanotechnology-specific safety or envi-
ronmental regulation has been slow to develop (36), with the EU
only recently adopting the first national or supranational legisla-
tive instrument containing nano-specific provisions (see Note 2),
one area of legal action has not – the willingness of governments
to fund research and development of nanotechnology. In the
USA, for example, two (of the four) goals of the National
Nanotechnology Initiative (NNI) are to “[a]dvance a world-class
nanotechnology research and development program” and “[f]
oster the transfer of new technologies into products for commer-
cial and public benefit” (37). To that end, the US federal govern-
ment has ponied up more than $US1.6 billion in 2010 alone to
foster basic and applied research into the technology (38). Other
countries have been equally quick to promote their nascent nano-
technology efforts with government funding (39, 40).
One consequence of this approach is that nanotechnology may
be the most multijurisdictional and multinational technology to have
emerged so far (41). The result of this is that nanotechnology patents
are likely to have numerous inventors from more than one institu-
tion and, just as likely, from more than one country (42, 43).
For example, the atomic force microscope (AFM), one of the
most basic research tools necessary to do almost any work in nan-
otechnology (44), was patented in 1988 (see below for a discus-
sion of how patenting of basic research tools may be a problem)
and awarded to IBM and, in particular, its Swiss research center
(see Note 3). The initial patented invention was not multina-
tional, however, a recent survey of patents arising out the original
AFM patent shows that more than 3,000 patents now relate to
(either in improvements, modifications, or processes) the original
AFM patent (8). In writing this chapter, the authors conducted a
rather unscientific survey of just a few dozen of those comple-
mentary patents and discovered related filings from Japan,
Germany, the USA, China, Canada, France, and many others.
Among these patents, a select few showed researchers from differ-
ent jurisdictions collaborating on inventions (see Note 4).
In an even more unscientific study, we leafed through a few
of the more than 2000 currently pending nano-patent applications
and found numerous multinational collaborations on inventions
including inventors from the USA collaborating with inventors
from (1) India, (2) Great Britain, (3) the Netherlands, (4) Poland,
(5) Belgium, and (6) Japan. A takeaway from this small sample is
that the level of cross-country collaborations is growing. Perhaps

more important, as collisions between patent holders will inevita-
bly grow, so will calls for patent pools (see Subheading 2.4 below
for a broader discussion of these) that will require inventors from
numerous jurisdictions to collaborate on and share potential
inventions. As inventors and inventions will become increasingly
multinational, so too the pressure on national patent systems to
increase efficiency and harmonize doctrines that discriminate
against foreign inventors will also increase.
Massive government funding often means direct funding to
universities for early stage research. In this way, nanotechnology
is nothing new under the sun – governments have a long history
of funding basic research into potential applied sciences. What is
new, however, is that the outgrowth of this funding has come in
an era of hyperpatenting on all points of the research curve.
Indeed, this is one area that may separate nanotechnology from
all other prior technological revolutions – every aspect of this
technology may be patented (6).
All other technological innovations, from biotechnology to
software, initially developed during a time when those who received
federal funding (universities mainly) were unable to effectively com-
mercialize or patent inventions (6). Software, arising in the 1960s
and 1970s, not only arose in an era where basic research could not
often be patented in the USA as a result of its federal funding, but
also arose in a culture of publication over patenting. Biotechnology,
although eventually reaching the hyperpatenting stage, was also
developed in an era and culture that dampened much of the urge to
patent. The only two historical parallels to nanotechnology are the
patent wars of the radio and the airplane (6, 45, 46). In each of
those cases, market-based attempts to override patent thickets failed.
Government action was required in each case to break the patent
logjam and allow research to transform to public goods.
Nanotechnology is global not only in its funding and com-
mercial reach, but also in its ability to patent not only the applica-
tions of the technology but also the basic research tools necessary
to conduct the research. The ability of inventors to patent basic
research tools is not, doctrinally, novel (47). What is new is, as
already noted, the willingness and desire of universities, the origin
of most basic research in industrialized countries, to patent such
tools and, more important, the erosion of traditional experimen-
tal use exceptions (48). Indeed, there are very few basic nano-
technology building blocks and research tools that are not,
already, patented (6, 49, 50). There has been much hand-wringing
about this fact – and real concern that nanotechnology’s real
potential will be swallowed up in a morass of litigation (51). This
is an issue we explore in more detail in Subheadings 2.3 and 2.4.
Another characteristic, as described by Maynard (52), is
its multidisciplinarity “which crosses established boundaries of
scientific inquiry and agency jurisdiction.” Based on the complexity

and multidisciplinarity of most nanotechnology patents, many
have been concerned that the patent offices of various govern-
ments will prove unable to handle, at least not appropriately,
requests to patent these developments (53). Most concerning is,
apparently, the fact that poor review at patent offices (6, 54) will
result in overpatenting of inventions that do not meet minimum
requirements of patentability in all areas (55). As one author put
it, concerns over workload and expertise are prevalent at patent
offices, but these concerns are “especially true [for nanotechnol-
ogy] since the PTO’s current staff lack the cutting edge knowl-
edge to completely understand these inventions. This problem is
exasperated [sic] by the fact that the PTO still faces an enormous
backlog of nanotechnology patent applications” ((51); see also
(56, 57)). As Tegart (51) noted, “Inventors need fast and unbu-
reaucratic help to realize an idea with importance for the future.”
In the sections to come, we explore in more detail the problems
hinted by Harris and attempt to assess the likelihood that these
problems will increase.
Finally, as evidenced by their financial commitments, the
potential of nanotechnology is obviously not lost on govern-
ments. Much of this potential is in public health and, in particu-
lar, in cancer research and other disease areas (34, 58). Although
there is real concern that the commercial viability of nanotechnol-
ogy applications may be hindered by patenting practices, a larger
concern is that legal wrangling will slow the health benefits of
nanotechnology. Slowing down economic growth may be a cost
more easily borne by nations with strong patenting regimes – but
as debates over pharmaceuticals have made clear, issues involving
health applications of nanotechnology have the potential to
rework patent systems in much more substantial ways.
2.3. Mapping Given the general background of nanotechnology patenting dis-

the Current cussed above, the question that many have asked is whether the
Nanotechnology potential disaster many feared is starting to, or perhaps has, come
Patent Landscape to pass. The short answer is that it looks like things are not exactly
going according to plan at the patent offices.
With Lux Research (50) having stated that “corporations,
start-ups, and labs depend on patents to protect their nanotech
innovation – and turn them into cash,” the importance of securing
patents for nanotechnology-based inventions – including both
product and processes – is arguably best highlighted by reference
to the increasing levels of patent activity within key patent offices
such as the USPTO and the European Patent Office (EPO).
Recent efforts to report on this activity, including performance
data relating to jurisdictions, institutions, and individuals, have
included work by, for example, Marinova and McAleer (59),
Huang et al. (60–62), Bawa (63), Koppikar et al. (64), Heinze (65),
Lux Research (50, 53), and Chen and Roco (66). All have shown
a deluge of nanotechnology patents.
In one of the first published studies examining longitudinal
patent activity for nanoscale science and engineering activities
within the USPTO, Huang et al. (61) reported rapid growth in
patenting activity over the period 1976–2002. By using a “full
text” keyword-based approach (see Note 5), and subsequent fil-
tering process, the authors found that the USPTO processed
approximately 8,600 “nano-based” patents over this period; pat-
enting activity was found to be steep after 1997 and 2001. These
periods of growth in patent activity occurred, as was observed by
the authors, around periods of program growth and other institu-
tional activities within, for example, the USA. It is perhaps unsur-
prising then that the authors (61) found that “the [nanoscale
science and engineering] patents grew significantly faster than the
USPTO database as a whole, especially beginning with 1997.”
Other key findings reported by Huang et al. (61) included
the diversity of countries and institutions involved in the patent-
ing activity (albeit still dominated by the USA), and the strength
of patenting activity within particular technological fields includ-
ing chemicals, catalysts, and pharmaceuticals. The observed
growth in patenting activity would appear to highlight the impor-
tance of patent law, and the protections therefore afforded to pat-
entees under the legal framework; this is despite the costs
associated with securing patent protection for an invention (see
Note 6).
In a more recent study of patenting activity within the
USPTO, Lux Research (50) similarly reported a “ramp-up” in
the number of patents being issued by the national patent office,
with steep growth continuing in the post-2003 period. According
to their analysis,
the number of nanotech patents issued ha[d] risen steadily from a base
of 125 in 1985 to 4,995 today …Nanotech patents far outpace other
areas of innovation, with a compound growth rate of 20% versus just
2% for patents overall (50).
Their analysis supported the findings of Huang et al. (61), with

Lux Research noting that patentees were more likely to be from
the USA than any other jurisdiction but with a growing percent-
age from other jurisdictions. In addition, patents were likely to be
assigned to a patentee in the private sector than any other sector
(for example, university, government and/or research organiza-
tion). Along with significant growth in patent activity, the authors
found that the average number of claims within each patent had
also increased. In their words:
inventors are authoring more sophisticated patents that cover more
nuanced variations of the same theme in a single filing. The average
nanotech patent issued in 2005 has 23.5 claims, compared to only
15.8 in 1985 (50).
As discussed below, this observed trend has significant implications

for not only patent examiners who must be able to deal with the
complexities associated with the applications, but also patent
growth more generally.
In addition to overall patenting activity within the USPTO,
Lux Research (50) also looked at patenting activity (applications
and grants) for eight specific nanomaterials, each of which has the
ability to be utilized across five different applications areas. They
included the following: carbon nanotubes, metal nanoparticles,
ceramic nanoparticles, dendrimers, quantum dots, fullerenes, and
nanowires. The purpose of the report was to provide an in-depth
analysis of patent density for each of these platform materials, to
determine the breadth of the patent claims, and to identify areas of
potential entanglement; vulnerability to potential challenges (con-
flict) and market potential of patents were also considered (50).
Based on this examination, the authors found that significant
growth occurred in relation to patenting activity for all eight
materials; ceramic nanoparticles and carbon nanotubes, which
have broad applications across numerous fields, were found to
have experienced particularly steep growth over the time period
examined, resulting in high patent density (50). This, they sug-
gested, had the potential to create an unfavorable patent environ-
ment for inventors/patentees in relation to, for example, carbon
nanotubes within the electronics field and ceramic nanoparticles
within personal health-care and cosmetics applications. However,
the authors (50) went on to suggest that there may still be hidden
opportunities in relation to these two materials, and that given
their potential breadth as structural materials, “it [was] likely that
these nanomaterials will emerge as battles worth fighting by
2008.” Varying trends in patent filing and density, as well as
future potential based on market opportunities, were observed
for the other six materials (50).
The continued increase in patenting activity in key national
patent offices suggests that industry, research organizations, uni-
versities, and other key bodies remain positive about the market
opportunities and associated economic benefits for nanotechnology-
based inventions. This is despite the costs associated with techno-
logical innovation and the increasingly vocal debates occurring
within jurisdictions over potential, yet unquantified, risks associ-
ated with some facets of the technology. Yet, as the ETC Group
(49) has sought to remind us, the successful granting of a pat-
ent, albeit for a nanotechnology-based invention or any other
type, is not enough in itself to ensure the commercial success of
that invention. Success or failure, as witnessed in the EU in rela-
tion to, for example, genetically modified foods is dependent on
a far broader range of criteria, including consumer acceptance of
the invention and/or technology (12, 67, 68).
While much of the literature relating to intellectual property
rights and nanotechnology has canvassed the patent landscape
and paints a detailed picture thereof, there is an increasing body

of work that has focused on the potential challenges and barriers
that the technology and its inventors may face in the coming
years. Concerns have been expressed, for example, in relation to
the breadth of patent claims for platform or structural materials,
patent thickets, overlapping patent claims, and the institutional
capacity of patent offices to assess nanotechnology-based applica-
tions. Many of these issues are not in themselves unique to nano-
technology; rather, as highlighted below, they are common to
other emerging technologies, including synthetic biology and
reflect many of the experiences of prior technologies.
However, nearly all current research into patenting activity
for nanotechnology paints a gloomy picture of the patent land-
scape. National patent offices are overwhelmed, with the time for
nanotechnology patents taking far longer than other technologies
(50). Indeed, one study revealed that nanotechnology applica-
tions, the average time for a nanotechnology-based patent appli-
cation to be processed and granted by the USPTO had increased
from 33 months in 1985 to 47 months in 2006 (50). Whether
the delays are caused by the sheer number of applications in rela-
tion to an understaffed patent office or if it is because of the
immense complexity of patent applications is not yet clear (see
Note 7). According to Halluin and Westin (69),
because few individuals have an in-depth and complete knowledge of
nanotechnology, patent examiners may not have the tools necessary to
understand the complexities of the field. In fact, both the European
and US Patent Offices have admitted that they do not fully understand
nanotechnology.
Nevertheless, these delays have real consequences for the com-

mercialization and development of beneficial nanotechnology
applications, especially within the field of nanomedicine. Delays
and costs at patenting obviously delay research and dissemination
of key technologies and, just as obviously, deeply affect pricing.
This challenge is not however unique to nanotechnology,
with other sophisticated technologies likely to also challenge the
capacity of patent offices.
In addition to these economic inefficiencies, the studies we
have discussed above, as well as countless others, have shown that
feared patent thickets, brambles, quagmires, and other natural-
disaster-themed descriptions have apparently taken over the land-
scape (70–73). In basic research tools, fundamental materials,
building-block structures, and numerous other crucially impor-
tant aspects of nanotechnology, vast numbers of overlapping and
broadly written patents, held by varied institutions and competi-
tors, have already been issued. We have already seen, in the USA,
a series of patent infringement lawsuits filed among competitors.
Although no study has yet compared the level of infringement
suit activity compared to prior technologies, there are many reasons
to believe that nanotechnology’s future may be threatened, or at

least made less bright, by these looming controversies.
2.4. Moral and Ethical It is arguably not surprising, when considered against the backdrop
Implications of of the patenting of human genes debate and associated concerns
Nanotechnology over the breadth of patents being granted on human genes, that this
Patenting issue has also become a topic of debate in regards to the patenting
of nanotechnology. This has been particularly the case regarding
platform or structural materials, such as the eight considered by Lux
Research (50) in their report. The concern here, as articulated by
the ETC Group, is primarily in relation to the issues of concentrated
ownership and therefore control over patents, and the subsequent
implication of this in terms of economics, innovation, and access –
especially for developing economics (49). Of course, such concerns
are not new, nor unique to nanotechnology. However, it would
appear that nanotechnology does create additional challenges here;
the ETC Group (49) has suggested that, for example,
breathtakingly broad nanotech patents are being granted that span mul-
tiple industrial sectors and include sweeping claims on entire classes of
the Periodic Table.
They go on to suggest that (49),

it’s not just the opportunity to patent the most basic enabling tools,
but the ability to patent the nanomaterials themselves, the products
they are used in and the methods of making them.
The ETC Group’s concern is that the dense concentration of

patents, which are held by a small number of patent holders, com-
bined with the ability to patent basic nanoscale materials, “could
mean monopolizing the basic elements that make life possible”
(49). The implications of this, it was argued, would be profound
for developing countries with the ETC Group (49) stating that,
researchers in the global South are likely to find that participation in
the proprietary “nanotech revolution” is highly restricted by patent
tollbooths, obliging them to pay royalties and licensing fees to gain
access.
Given their concern over patent concentration and breadth of

claims being made, the nongovernmental organization went on
to highlight the emergence of so-called “patent thickets” within
the nanotechnology patent landscape. This refers to, as explained
by Shapiro (3), “an overlapping set of patent rights requiring that
those seeking to commercialize new technology obtain licenses
from multiple patentees.” Patent thickets therefore have the abil-
ity to hinder technological innovation and therefore the commer-
cialization of new technologies.
In their examination of patenting activity for five nanomateri-
als within the USPTO and the EPO (carbon nanotubes, inor-
ganic nanostructures, quantum dots, dendrimers, and the
Scanning Probe Microscopes), the ETC Group was able to paint
a picture of the emerging patent thicket for some nanomaterials;

this was illustrated primarily by reference to the number of pat-
ents relating to each of the applications currently held by different
institutions, and the so-called patent density for each applications.
By way of example, the ETC Group reported that the USPTO
had issued 227 patents for carbon nanotubes between 1999 and
2004. While it found that the patents for the material were held
by a number of different parties, across a range of different sec-
tors, it nevertheless came to the conclusion that a patent thicket
for the material had already occurred and that,
a swarm of existing patents, whose claims are often broad, overlapping
and conflicting, means that researchers hoping to develop new tech-
nology based on carbon nanotubes must first negotiate licenses from
multiple patent owners (49).
The ETC Group is not the only commentator to have voiced its
concern about the potential implications of overlapping patent
claims and the emergence of “patent thickets” or “nano-thickets,”
with a number of commentators having expressed concern over
the potential creation, and the implications thereof, for nanotech-
nology (45, 50, 53, 70–74).
Having observed the problems associated with patent thickets
in other areas of technological innovation, including biotechnol-
ogy and information and communication technologies, Clarkson
and DeKorte (70) noted that patent thickets have the potential to
give rise to a range of issues, including the unintentional infringe-
ment of patents and the subsequently liability created as a conse-
quence of the said infringement, the problem of anticommons,
the creation of barriers to entry, and the need for licensing. While
the issues are not unique to any one area, they (70) noted that,
the nanotechnology patent space experiences an even greater level of
these problems because it is much more complicated than other tech-
nology areas.
In order to determine the extent of the growing patent thickets

for nanotechnology, Clarkson and DeKorte (70) undertook a
mapping exercise of patent space and density within the USPTO.
The authors then used network analytic techniques as a way to
“visualize” the growth at three time points – 2000, 2002, and
2004. By plotting individual patents and then references between
patents, the authors demonstrated not only the growth in nano-
technology patents within the USPTO during that time period,
but also the increasing interconnectivity – or network – between
the patents. This visualization process enabled the authors to map
potential patent thickets.
The increasing crowding of the nanotechnology patent
space, and therefore emerging patent thicket in some areas, was
found to support the findings of Lux Research (50, 53). Having
determined that patent thickets were not just theoretical for
nanotechnology, but were already occurring, the authors then

took a somewhat pragmatic approach to the issue; if thickets
cannot be avoided, what strategies can be employed to protect
patents while also promoting innovation?
Traditionally cross-licensing arrangements – which Shapiro
(3) has eloquently defined as “an agreement between two
companies that grants each the right to practice the other’s patents
– are one way in which this may be achieved.” However, while
such arrangements are relatively straightforward when involving
only two parties, Clarkson and DeKorte (70) note a number of
limitations, including high transaction costs, which can make such
arrangements prohibitive when more than two parties are involved
in the contractual negotiations. In light of these limitations,
Clarkson and DeKorte (70) proposed a second alternative for
avoiding validity challenges and potential patent litigation: “pat-
ent pools.” These contractual undertakings are, as summarized by
Clark et al. (75) “an agreement between two or more patent own-
ers to license one or more of their patents to one another or third
parties.” Such arrangements have been an important tool for pro-
viding parties with access to proprietary information for over a
century. It has been suggested that the ability for parties to readily
access patent information through such pool arrangements pro-
mote innovation within areas that may have otherwise become the
subject of patent blocking and legal challenges – and at a lower
transactional cost than cross-licensing (3, 70).
But establishing and relying on patent pools as a mechanism
to access patent information would appear to be only one poten-
tial approach to addressing the challenges presented by the thick-
ets. Another, arguably somewhat more radical, approach would
be to promote an “open source” approach. The open source
“movement” has been widely adopted in relation to software
development (76) and to varying degrees within the field of med-
icine and drug development (77, 78). The movement’s potential
application to the nanotechnology patents landscape has there-
fore raised some level of discussion among commentators. One of
the earliest contributions was from Bruns (79), who looked at the
applicability of the open source movement to molecular nano-
technology. In his view, “open source approaches might offer
advantages for faster, more reliable and more accessible research
and development” (79). He advocated the adoption of an open
source approach to the technology where public money had been
used to generate the intellectual property in question. Bruns (79)
argument was that such an approach would not only encourage
innovation but also assist in diffusing the technology, and its asso-
ciated benefits, to developing economies.
While the open source movement has continued to gain trac-
tion within, for example, the field of software development, “there
is not yet an “open source nanotechnology” movement” (80).
This may in part be explained by the fact that software develop-

ment is process based, where developers of nanotechnology are at
this time largely focused on product generation. Moreover, with
open source software, the primary ‘cost’ is the programmer’s time,
which they give freely to further develop and refine the code lines.
The same cannot be said with the development of nanotechnology,
which requires not only human resources but also infrastructure
and consumables. Prisco (81) and Peterson (82) have however
begun to further explore open source for nanotechnology. As
such, we would suggest that the nanotechnology patent landscape
is likely to significantly evolve over the short to medium term,
with the open source movement just one way in which individuals
and organizations attempt to circumnavigate the emerging patent
thickets and promote technology innovation.
As any individual with a green thumb will know, tending to a
garden – albeit a refined English garden or a small herb garden –
requires constant care and attention. Any such garden is dynamic
by its very nature and will evolve over time. A constant state of
vigilance is needed to ward off pests and other challenges, and the
more proactive, educated, and vigilant the gardener is, the better
the outcome will be.
As this chapter has sought to highlight, there are many simi-
larities between the needs and challenges of a garden and that of
a patent landscape. As with our garden, the patent landscape has
evolved and been refined over centuries in response to new spe-
cies and the introduction of new technologies. Sometimes, the
landscape has been better prepared to handle the attacks than
others. Nanotechnology is the latest species to place a strain on
the fundamental features of the landscape, pushing up against his-
torical walls and threatening traditionally well-tended fields. This
is due to a number of factors: its multidisciplinary character, its
trans-jurisdictional nature, the ability for inventors to apply for
and be granted patents not only to products but also to the basic
building blocks, and claims which related to a diverse number of
areas and/or applications. It is also in part due to the immense
public and private sector interests in nanotechnology, and the
rush to secure legal rights over their inventions.
But the question is: will nanotechnology be permitted to devas-
tate that which has taken centuries to build up? Or will the garden-
ers – primarily national governments in this instance – see the
emergence and growth of nanotechnology as an opportunity to
reconsider the borders and features of the current landscape and
revamp it accordingly? This would of course involve significant time
and energy, but with other equally complex and multifaceted tech-
nologies already in the research and development pipeline, it would
appear that policy makers need to “stop and smell the roses” to
ensure that the economic and social benefits of the technology are
released. Perhaps it is time to modernize the landscape to meet the
needs of the current climate – a more global approach to patenting

is one obvious option – and provide the gardeners with the tools
that they need to do their job in a timely and efficient manner.
But those who use and enjoy the garden must also take some
responsibility for its future to ensure that the benefits of the tech-
nology are realized. Rather than, for example, relying on costly
and time-consuming litigation, beneficiaries of the patent system
should be encouraged to explore arrangements such as
cross-licensing and patent pools early on, or where appropriate,
be encouraged to look to open source approaches. Governments
can also play a role here by, for example, creating a framework
which encourages and/or rewards these approaches.
3. Notes
1. In this chapter, we do not provide background or definitional

sections on what we consider to be nanotechnology. For our
views on this subject, see, for example, (11, 12, 83).
2. In November 2009, the European Parliament and Council
adopted the final text for the Cosmetic Regulation (84).
3. See US Patent Number: 4 724 318.
4. Cites available upon request.
5. As noted by Huang et al., there were “seven basic keywords
with several variations” (61).
6. For a more recent longitudinal study of nanotechnology-pat-
enting activity within the USPTO, the EPO, and the Japan
Patent Office (JPO). See also Chen and Roco (66).
7. Koppikar et al. (64) have suggested that the latter may be
largely attributed to the interdisciplinary nature of the appli-
cations and the “many and diverse applications [are] often
associated with a single nanotechnology invention.”
References
1. Graham, S. and Mowery, D. C. (2003) Standard Setting. Innov. Pol. Econ. 1,

Intellectual property protection in the U.S. 119–150.
software industry. In: Patents in the Knowledge- 4. Burk, D. L. and Lemley, M. A. (2003) Policy
Based Economy. Board on Science, Technology Levers in Patent Law. Va. Law Rev. 89,
and Economic Policy (STEP) (Cohen, W. and 1575–1696.
Merrill, D., eds.) The National Academies, 5. Heller, M. A. and Eisenberg, R. S. (1998)
Washington, DC. 219–258. Can Patents Deter Innovation? The
2. Eisenberg, R. S. (2002) How Can You Patent Anticommons in Biomedical Research. Science
Genes? Am. J. Bioeth. 2, 3–11. 280, 698–701.
3. Shapiro, C. (2000) Navigating the Patent 6. Lemley, M. (2005) Patenting Nanotechnology.
Thicket: Cross Licenses, Patent Pools, and Stanford Law Rev. 58, 601–630.
7. Bessen, J. Patent Thickets: Strategic Patenting 21. Eisenberg, R. S. (1989) Patents and the
of Complex Technologies. http://www.resear- Progress of Science: Exclusive Rights and
choninnovation.org/thicket.pdf. Accessed 26 Experimental Use. Univ. Chic. Law Rev. 56,
December 2009. 1017–1086.
8. D’Silva, J. (2009) Pools, Thickets and Open 22. Rai, A. K. (1999) Regulating Scientific
Source Nanotechnology. European intellec- Research: Intellectual Property Rights and the
tual property review, 31(6), 300–306. Norms of Science. Northwest Univ. Law Rev.
9. Ratingen, J. (2006) Slaying the Troll: 94, 77–152.
Litigation as an Effective Strategy Against 23. Masur, J. S. (2008) Process as Purpose: Costly
Patent Threats. Santa Clara Comput. High Screens, Value Asymmetries, and Examination
Technol. Law J. 23, 159–210. at the Patent Office. http://ssrn.com/
10. Magliocca, G. N. (2007) Blackberries and abstract=1105184. Accessed 26 December
Barnyards: Patent Trolls and the Perils of 2009.
Innovation. Notre Dame Law Rev. 82, 24. United States Patent and Trademark Office
1809–1838. (2005) General Information Concerning
11. Abbott, K. W., Sylvester, D. J., and Marchant, Patents. http://www.uspto.gov/patents/
G. E. (2010) Transnational regulation of basics/index.html#patent. Accessed 15
nanotechnology: Reality or romanticism? In: December 2009.
International Handbook on Regulating 25. Webber, P. M. (2003) Protecting Your
Nanotechnologies (Hodge, G. A., Bowman, D. Inventions: the Patent System. Nat. Rev.
M., and Maynard, A. D., eds.) Edward Elgar, Drug Discov. 2, 823–830.
Cheltenham. 525–544. 26. Maskus, K. E. (2000) Intellectual Property
12. Sylvester, D. J., Abbott, K. W., and Marchant, Rights in the Global Economy. Institute for
G. E. (2009) Not Again! Public Perception, International Economics, Washington, DC.
Regulation, and Nanotechnology. Regul. Gov. 27. Grossman, G. M. and Lai, E. C. (2004)
3, 165–185. International Protection of Intellectual
13. Marchant, G. E. and Sylvester, D. J. (2006) Property. Am. Econ. Rev. 94, 1635–1653.
Transnational Models for Regulation of 28. Mandel, G. (2010) Regulating nanotechnol-
Nanotechnology. J. Law Med. Ethics 34, ogy through Intellectual Property Rights. In:
714–725. International Handbook on Regulating
14. Bowman, D. M. and van Calster, G. (2007) Nanotechnologies (Hodge, G. A., Bowman, D.
Does REACH Go Too Far? Nat. Nanotechnol. M., and Maynard, A. D., eds.) Edward Elgar,
1, 525–526. Cheltenham. 388–405.
15. Bowman, D. M. and Hodge, G. A. (2009) 29. Caulfield, T., Gold, E. R., and Cho, M. K.
Counting on Codes: An Examination of (2000) Patenting Human Genetic Material:
Transnational Codes as a Regulatory Refocusing the Debate. Nat. Rev. Genet. 1,
Governance Mechanism for Nanotechnologies. 27–231.
Regul. Gov. 3, 145–164. 30. Abbott, F. M. (2006) Intellectual Property
16. Rai, A. K. and Eisenberg, R. S. (2003) Bayh- Provisions of Bilateral and Regional Trade
Dole Reform and the Progress of Biomedicine. Agreements in Light of U.S. Federal Law.
Law Contemp. Probl. 66, 289–314. http://www.unctad.org/en/docs/iteipc
17. Caulfield, T., Cook-Deegan, R. M., Kieff, F. 20064_en.pdf. Accessed 13 December 2009.
S. and Walsh, J. P. (2006) Evidence and 31. Australian Government (2008) Patentable
anecdotes: an analysis of human gene patent- Subject Matter – Issues Paper. Advisory Council
ing controversies. Nat. Biotechnol. 24, on Intellectual Property, Canberra.
1091–1094. 32. Drahos, P. (1999) Biotechnology Patents,
18. Klein, R. D. (2007) Gene patents and genetic Markets and Morality. Eur. Intellect. Prop.
testing in the United States. Nat. Biotechnol. Rev. 21, 441–449.
25, 989–990. 33. Bagley, M. A. (2003) Patent First, Ask
19. Andrews, L. B. (2002) Genes and patent pol- Questions Later: Morality and Biotechnology
icy: rethinking intellectual property rights. in Patent Law. William Mary Law Rev. 45,
Nat. Rev. Genet. 3, 803–808. 469.
20. Dinwoodie, G. B., Hennessey, W. O., and 34. Royal Society and Royal Academy of
Perlmutter, S. (2001) International Engineering (2004) Nanoscience and
Intellectual Property Law and Policy. Nanotechnologies: Opportunities and
LexisNexis, Newark. Uncertainties. RS-RAE, London.
35. Maynard, A. D. (2007) Nanotechnology: The 50. Lux Research (2006) Nanotech Battles Worth
Next Big Thing, or Much Ado about Nothing? Fighting. Lux Research, New York.
Ann. Occup. Hyg. 51, 1–12. 51. Harris, D. L., Hermann, K., Bawa, R.,
36. Lux Research (2009) Nanotech’s Evolving Cleveland, J. T., and O’Neill, S. (2004)
Environmental, Health, and Safety Landscape: Strategies for Resolving Patent Disputes Over
The Regulations Are Coming. Lux Research, Nanoparticle Drug Delivery Systems.
New York. Nanotechnol. Law Bus. 1, 372–390.
37. National Nanotechnology Initiative (undated) 52. Maynard, A. D. (2006) Nanotechnology: A
About the NNI – Home. http://www.nano. Research Strategy for Addressing Risk. Project
gov/html/about/home_about.html. on Emerging Nanotechnologies, Washington,
Accessed 15 December 2009. DC.
38. National Nanotechnology Initiative (2009) 53. Lux Research (2005) The Nanotech Intellectual
Funding. http://www.nano.gov/html/about/ Property Landscape. Lux Research, New York.
funding.html. Accessed 15 December 2009. 54. Thomas, J. R. (2001) Collusion and Collective
39. Roco, M. C. (2005) International perspec- Action in the Patent System: A Proposal for Patent
tives on government nanotechnology funding Bounties. Univ. Ill. Law Rev. 1, 305–353.
in 2005. J. Nanopart. Res. 7, 707–712 55. Tegart, G. (2004) Nanotechnology: the
40. European Commission (2005) Nanosciences Technology for the Twenty-First Century.
and Nanotechnologies: An Action Plan for Foresight 6, 364–370.
Europe 2005–2009. European Parliament, 56. Zekos, G. I. (2006). Nanotechnology and
Brussels. Biotechnology Patents. Int. J. Law Inf.
41. Hullmann, A. and Meyer, M. (2003). Technol. 14, 310–369.
Publications and Patents in Nanotechnology: 57. Bowman, D. (2007) Patently Obvious:
an Overview of Previous Studies and the State Intellectual Property Rights and
of the Art. Scientometrics 58, 507–527. Nanotechnology. Technol. Soc. 29, 307–315.
42. Zucker, L. G. and Darby, M. R. (2005). Socio- 58. Ferrari, M. (2005). Cancer nanotechnology:
economic Impact of Nanoscale Science: Initial opportunities and challenges. Nat. Rev.
Results and Nanobank (working paper Cancer 5, 161–171.
11181). NBER available at http://www.nber. 59. Marinova, D. and McAleer, M. (2003)
org/papers/w11181. Nanotechnology Strength Indicators:
43. Zucker, L. G., Darby, M., Furner, J., Lieu, R., International Rankings Based on US Patents.
and Ma, H. (2007). Minerva Unbound: Nanotechnology 14, R1–R7.
Knowledge Stocks, Knowledge Flows, and 60. Huang, Z., Hu, R., and Pray, C. (2003)
New Knowledge Production. Res. Policy 36, Longitudinal Patent Analysis for Nanoscale
850–863. Science and Engineering: Country, Institution
44. Binnig, G., Quate, C. F., and Gerber, C and Technology Field. J. Nanopart. Res. 5,
(1986) Atomic Force Microscope. Phys. Rev. 333–363.
Lett. 56, 930–934. 61. Huang, Z., Chen, H., Chen, Z. K., and Roco,
45. Sabety, T. (2004) Nanotechnology Innovation M. C. (2004) International Nanotechnology
and the Patent Thicket: Which IP Policies Development in 2003: Country, Institution,
Promote Growth? Albaby Law J. Sci. Technol. and Technology Field Analysis Based on
15, 477–516. USPTO Patent Database. J. Nanopart. Res. 6,
46. Johnson, H. A. (2004) Wright Patent Wars 325–354.
and Early American Aviation. J. Air Law 69, 62. Huang, Z., Chen, H., Li, X., and Roco, M. C.
21–64 (2006) Connecting NSF Funding to Patent
47. Mueller, J. M. (2001) No Dilettante Affair: Innovation in Nanotechnology (2001–2004).
Rethinking the Experimental Use Exception J. Nanopart. Res. 8, 859–879.
to Patent Infringement for Biomedical 63. Bawa, R. (2004) Nanotechnology Patenting
Research Tools. Wash. Law Rev. 76, 1–66. in the US. Nanotechnol. Law Bus. 1, 31–51.
48. Sylvester, D. J., Menkhus, E., and Granville, K. J. 64. Koppikar, V., Maebius, S. B., and Rutt, J. S.
(2005) Innovation Law Handbook. Available at (2004) Current Trends in Nanotech Patents:
SSRN: http://ssrn.com/abstract=999451. A View From Inside the Patent Office.
Accessed 26 December 2009. Nanotechnol. Law Bus. 1, 24-30.
49. ETC Group (2005) Nanotech’s “Second 65. Heinze, T. (2004) Nanoscience and
Nature” Patents: Implications for the Global Nanotechnology in Europe: Analysis of
South. ETC Group, Ottawa. Publications and Patent Applications including
Comparisons with the United States. 76. Lakhani, K. R. and von Hippel, E. (2003)
Nanotechnol. Law Bus. 1, 1–19. How Open Source Software Works: “Free”
66. Chen, H. and Roco, M. C. (2008) Mapping User-to-User Assistance. Res. Policy 32,
Nanotechnology Innovations and Knowledge. 923–943.
Springer, New York. 77. Economist (2004) An Open-Source Shot in
67. Bauer, M. W. and Gaskell, G. (eds.) (2002) the Arm? Economist, 10 June, 17.
Biotechnology: The Making of a Global 78. Munos, B. (2006). Can Open-Source R&D
Controversy. Cambridge University Press, Reinvigorate Drug Research? Nat. Rev. Drug
London. Discov. 5, 723–729.
68. Jasanoff, S. (2005) Designs on Nature: 79. Bruns, B. (2001). Open Sourcing
Science and Democracy in Europe and the Nanotechnology Research and Development:
United States. Princeton University Press, Issues and Opportunities. Nanotechnology 12,
Princeton. 198–210.
69. Halluin, A. P. and Westin, L. P. (2004) 80. Kelty, C., Lounsbury, M., Yavuz, C. T., and
Nanotechnology: The Importance of Colvin, V. L. (undated) Towards Open Source
Intellectual Property Rights in an Emerging Nanotechnology: Arsenic Removal and
Technology. Howrey Simon Arnold & White, Alternative Models of Technology Transfer.
Washington, DC. http://opensourcenano.net/images/GRC-
70. Clarkson, G. and DeKorte, D. (2006) The Poster2.pdf. Accessed 15 April 2009.
Problem of Patent Thickets in Convergent 81. Prisco, G. (2006) Globalization and Open
Technologies. Ann. NY Acad. Sci. 1093, Source Nano Economy. KurzweilAI.net.
180–200. http://www.kurzweilai.net/meme/frame.
71. Lee, A. (2006) Examining the Viability of html?main=/articles/art0659.html. Accessed
Patent Pools for the Growing Nanotechnology 12 April 2009.
Patent Thicket. Nanotechnol. Law Bus. 3, 82. Peterson, C. L. (2008) Citizen-Controlled
317–328. Sensing: Using Open Source & Nano
72. Harris, D. L. (2008) Carbon Nanotube Patent technology to Reduce Surveillance & Head
Thickets. In: Nanotechnology & Society: Off Iraq-Style Wars. http://www.open-
Current and Emerging Ethical Issues (Allhoff, sourcesensing.org/proposal.pdf. Accessed 12
F. and Lin, P., eds.) pp. 163–186. Springer, April 2009.
New York. 83. Hodge, G. A., Bowman, D. M., and Ludlow,
73. Miller, J., Serrato, R., Represas-Cardenas, K. (2007) Introduction: big question for
J. M., and Kundahl, G. (2005) The Handbook small technologies. In: New Global Regulatory
of Nanotechnology: Business, Policy, and Frontiers in Regulation: The Age of
Intellectual Property Law. Wiley, New York. Nanotechnology (Hodge, G. A., Bowman, D.
74. Bastani, B. and Fernandez, D. (2004) M., and Ludlow, K., eds.) pp. 3–26. Edward
Intellectual Property Rights in Nanotechnology. Elgar, Cheltenham.
Fernandez & Associates, LLP, Menlo Park. 84. EurActiv.com (2009) Germany opposed to
75. Clark, J., Piccolo, J., Stanton, B., and Tyson, ‘nano’ label for cosmetics, 24 November.
K. (2000) Patent Pools: A Solution to the http://www.euractiv.com/en/enterprise-jobs/
Problem of Access in Biotechnology Patents? germany-opposed-nano-label-cosmetics/arti-
USPTO, Washington, DC. cle-187583. Accessed 15 December 2009.
Chapter 24
Scientific Entrepreneurship in the Materials

and Life Science Industries
Jose Amado Dinglasan, Darren J. Anderson, and Keith Thomas
Abstract
Scientists constantly generate great ideas in the laboratory and, as most of us were meant to believe, we
should publish or perish. After all, what use is a great scientific idea if it is not shared with the rest of the
scientific community? What some scientists forget is that a good idea can be worth something – some-
times it can be worth a lot (of money)! What do you do if you believe that your idea has some commercial
potential? How do you turn this idea into a business? This chapter gives the aspiring scientific entrepre-
neur some (hopefully) valuable advice on topics like choosing the right people for your management
team, determining inventorship of the technology and ownership shares in the new company, protecting
your intellectual property, and others; finally, it describes some of the various pitfalls you may encounter
when commercializing an early stage technology and instructions on how to avoid them.
Key words: Entrepreneurship, Technology transfer, Nanomaterials, Nanotechnology, Start-up,

Venture capital
1. Introduction
Why become an entrepreneur? Certainly, there are risks associated

with becoming an entrepreneur. The odds of failure are high, and
entrepreneurs typically face an emotional “roller-coaster” of good
news and very bad news. This can be exacerbated by the fact that
entrepreneurs take a very personal interest in their company. It is
easy to get so personally invested that it can be difficult to keep an
even emotional keel.
So why become an entrepreneur? When people speak about
becoming an entrepreneur, they often lead with the bad side of
entrepreneurship, much like we have done here. The simple truth
is not everyone should become an entrepreneur. Entrepreneurship
requires significant risk tolerance, emotional balance, and a keen
379
380 Dinglasan, Anderson, and Thomas
sense of vision, but for those who are well suited and who are
willing to take the risks, it is an incredibly fulfilling career option
for scientists.
We believe that entrepreneurs are, more than anything else,
motivated by impact. Impact can be in the purely monetary sense –
few other scientific careers offer the opportunity for a PhD to be
a multimillionaire 5 years after graduation. More importantly,
however, impact also has a huge nonfinancial dimension. In a
company that is commercializing a new technology, one can have
an opportunity to help address major world issues, to directly
impact the success or failure of your venture, to build a team of
employees from ground zero, and to influence the culture of a
new organization. In addition to all of this, if you are a recent
graduate from a doctoral, master’s, or undergraduate program, it
is an opportunity to create an exciting role for yourself in a grow-
ing company that suits your unique skills.
To introduce ourselves and to give you a little perspective on
where we are coming from, two of the authors of this chapter
(Anderson and Dinglasan) are scientific entrepreneurs and
cofounders of Vive Nano. The third author (Thomas) is a serial
entrepreneur who has been Vive Nano’s President and CEO since
the very early stages of the company. Vive Nano is a spin-off com-
pany from the University of Toronto that was founded in 2006.
At the time of writing (2010), we have 16 employees, a pilot
manufacturing facility in Toronto, and a range of issued patents
and pending patent applications. Vive Nano is focused on using
its proprietary process to provide simple small solutions to big
issues. Specifically, our development activities are targeted in two
areas with major global impact: new crop protection formulations
with improved efficacy and decreased environmental impact, and
improved nanoscale heterogeneous catalysts.
In this chapter, we provide an overview of our experiences in
founding Vive Nano, including its history to date, some high-
level but useful advice to aspiring entrepreneurs from academic
institutions on topics like choosing the right people for your
management team, determining inventorship of your technology
and ownership shares in your new company, protecting your
intellectual property, and others; finally, we give descriptions of
some of the various pitfalls you may encounter when commer-
cializing an early stage technology and instructions on how to
avoid them.
This chapter is primarily aimed at graduate students and
postdoctoral fellows that are thinking of starting a company based
on research that they have performed at a university. The infor-
mation may also be useful to professors, postdoctoral fellows or
research associates in government laboratories, or researchers in
private industry. There are many different types of companies that
can be founded by researchers in these circumstances, including
Scientific Entrepreneurship in the Materials 381
service businesses and companies that start small and grow slowly,
but provide sufficient profits for the founding team to live well.
We are focused in this chapter on companies that are targeting
large problems in large markets that will (typically) require financ-
ing from external resources.
Of course, there is no one-size-fits-all solution to starting a
company, and circumstances may vary greatly. One thing that an
entrepreneur learns very early on is to take all advice with a grain
of salt and come to their own conclusions. We suggest you do the
same with this chapter.
2. Discussion
The idea of starting a company began when a group of us (including

Anderson and Dinglasan) took part in a noncredit course offered
by the University of Toronto called “Entrepreneurship 101.”
Most of us were at turning points in our careers and were trying
to decide what to do after graduate school or a postdoctoral fel-
lowship. This full-year seminar course was geared toward scientists/
academic personnel who wanted to explore the idea of starting a
business based on the technologies or ideas that they had worked
on in the laboratory. Each of the course’s lecturers was practicing
in industry and had advanced scientific training. Lecture topics
included intellectual property (IP), market research, project man-
agement, licensing, and others. This course gave us some insight
into how to bring something that has commercial potential out of
the laboratory and make it into a business.
At the end of the course, we were expected to make a busi-
ness pitch based on a technology that had been developed as part
of our research. The process of making the pitch made us think
about the impact the technology that we were trying to commer-
cialize could have on certain markets/sectors, the advantages
(and disadvantages) of our technology over existing technologies,
the different products we could potentially produce, and a possi-
ble plan for the business. Most importantly, it also taught us just
how much we did not know about starting a company. Although
we realized that a lot of additional groundwork would be needed
to even start a sustainable business, we also convinced ourselves
that the technology that we had developed was exciting enough
to form the foundation of a new company.
We were also fortunate enough to belong to a research group
that had prior start-up experience. Professor M. Cynthia Goh, in
whose laboratory we were working, had previously spun out Axela
Biosensors, a diagnostics company that was based on a technol-
ogy developed in her laboratory. This company was about 5 years
into its existence at that point, and had about 20 employees,
and we had seen it grow from an idea in the laboratory to a

well-financed, product-based company. This means that we knew
it was possible to start a company from an idea developed in the
laboratory and that we had access to people who could help advise
us about the risks, opportunities, and surprises common to early
stage companies. We hope to provide much of the same advice to
the readers of this chapter.
2.1. The Idea As scientists, we create and generate knowledge in the research
laboratory. What do we do if we think this knowledge is poten-
tially commercializable? For a start-up company, this knowledge
is one of the two critical early stage assets of the company (the
other is people, outlined in further detail in Subheading 2.2).
This knowledge can take many forms, and if it can be owned by
the company, it is known as intellectual property (IP). Intellectual
property could be a new technology that it is possible to get
patented – and thereby prevent anyone else from using – or
simply “know-how,” things that you know that are necessary for
the knowledge to be applied, but may not be patentable. For
most start-up companies, patentable IP is the most important
type of IP and will need to be protected using patents at a very
early stage.
Because of the importance of IP, it is important to clearly
define at the very beginning who owns the IP and how it will
be commercialized. Assuming you developed an invention or
technology while working at a university, depending on the IP
policies set up at that particular university, the university may
own part, if not all, of your invention. This is also true in many
government laboratories. Depending on the country in which
you live, each university will have its own version of a technology
transfer office that is responsible for handling ownership of
IP developed within university premises. They can usually assist
the inventors of the technology with commercializing the
technology, if desired.
The first step once an invention or technology has been
developed is to let the university know about the invention. This
is normally done by filing an invention disclosure to the univer-
sity’s technology transfer office. The disclosure documents the
circumstances under which the invention was created as well as
provides the university with the information necessary to evalu-
ate inventorship, patentability, and obligations to research spon-
sors outside the university. Depending on the policies at your
institution, you and the other inventors may now have the option
to commercialize the technology or invention yourself, or you
may need to work with the technology transfer office to do so.
Usually, one of the first activities that is required is to file a
patent on the invention that will help you prove that you own the
invention or technology. Unless a technology has been protected
in this way, it is often difficult to attract the financing that is

necessary to build the business. It is only possible to get a patent
on an invention if it has not yet been disclosed to the public.
Public disclosure includes publishing a scientific paper or presenting
a poster or a talk at a conference, so it is critical to make sure that
you have filed any relevant patents before you give even a single
presentation about your technology. Your technology transfer
office can assist you with understanding this issue in more detail.
A discussion of drafting patents is beyond the scope of this
chapter, but we do suggest that you recruit legal counsel (a patent
agent or a lawyer) to assist you with this process. While expensive,
IP is a critical asset for the company, and a mistake made at this
stage could cause dramatic problems down the road.
It also is important to determine the inventors of the invention
at an early stage. The rules of determining inventorship of an
invention are quite different from the rules of determining author-
ship of a scientific paper. Inventors are the people who were
responsible for the creative breakthrough that led to the inven-
tion. You, along with the colleagues that worked with you, must
come into agreement about who contributed to the original idea
and overall intellectual input to the invention in question. People
that merely work under the direction of another are not consid-
ered coinventors unless they have made inventive or creative con-
tributions to a concept that led to the invention. For you to be
considered an inventor, you should be able to point to a specific
idea that you contributed without which the invention could not
exist. It is essential to settle any inventorship issues at this early
stage because certain problems can arise later on if inventorship
is not correctly attributed. Further, outstanding inventorship
issues can leave your future patent open to challenges from other
parties that may wish to invalidate it. From a practical point of
view, it is just good sense to think about these issues early, while
there is no money involved. Things become quite different once
the invention is actually worth something. If necessary, your
patent agent or patent lawyer can assist you with determining
the inventors of an invention.
2.2. The Team Now that you have a technology or an invention that you have
decided to commercialize, it is time to start recruiting the people
that will make your company a success. Before bringing on exter-
nal people, we recommend you turn your attention to the found-
ing group. In our experience, the start-up company has a better
chance of success if there are multiple founders. This guarantees
different points of view when the critical early stage decisions for
the company are made and also increases the odds that people on
the founding team have complementary skills. However, having
multiple founders does complicate issues, for instance, in deter-
mining how much of the new company each founder will own.
For the sake of argument, let us say there are 5 individuals

that decide they want to start a company together. Each of the
individuals brings different contributions and skills to the table;
some are inventors of the technology, some have business savvy
and can help with initial financing, and some can provide mentor-
ship or training to the others. In addition, usually (though not
always) the university takes some stake in the start-up company by
providing the background facilities necessary for the invention to
be developed. In this scenario, each individual obviously deserves
partial ownership of the company that is to be created based on
the invention, but how do we now assign how much of the start-
up company each individual should own? Do you divide the com-
pany equally? There is really no perfect way to do this. It is all
really quite arbitrary at this point, but what is essential is that
everyone comes into agreement on the amount of initial shares
assigned to them. Each individual should negotiate with the
group for his/her fair share. It is important to make sure that, in
exchange for the shares in the company, the individuals involved
have contributed to the company in a significant way. Ownership
should not be handed out unless it has truly been earned. The
goal is that everyone is comfortable that the company’s initial
ownership is fair and that everyone has been treated equitably.
This is much easier to do now, when the company (and therefore
the shares) is worth very little. Unfair ownership distribution or
unearned ownership can cause major problems down the road,
when the shares are actually worth something.
It is helpful not only to consider the amount of time each
individual has already contributed to the company up to the point
where the shares are being divided, but also the potential amount
of time he or she will be spending with the company after the
shares have been assigned. Those who are sticking around should
be given some incentive (i.e., someone who is willing to stay on
for a year or more should be allocated more shares than someone
who is not going to be with the company next year). You could
also set up a scheme to reward individuals for making significant
contributions such as hitting certain milestones, developing new
products, or finding funding, just to name a few examples.
Regardless of the exact details of what you work out, it is essential
to reward individuals with enough incentive to make them feel
that what they have done and will do in the future is valuable to
the company. We have found that employee stock option grants
based on performance are one of the best ways to do this.
Once you raise any external financing, it is usually important
to set up a shareholder’s agreement that will govern share owner-
ship, voting rights, and many other important issues that define
how the company is run. At this time, the ownership of the com-
pany among the founding team is locked in, and some of the
flexibility that existed initially disappears.
The founding team (and investors, depending on their interest)

will need to recruit other people who can help the new company
succeed. In addition to IP, the people involved in the new
company are its greatest asset. In order to meet potential
new hires, the most critical piece of advice that we can provide is
to network. Just as scientific entrepreneurship is not for everyone,
joining a new, growing company is not for the faint of heart! This
means that people who are well suited to your new company can
be difficult to find – you will need to search hard for the best
people. One way to expand your network is by joining profes-
sional organizations and trade associations, as well as by
attending scientific conferences and trade shows. Entrepreneurship
programs run by your university or other nearby institutions are
an excellent place to meet people with a keen interest in entrepre-
neurship. Usually, even if the speakers and attendees are not
available to consider joining your team, they can connect you
with good potential hires.
As one of the founding members of a technology-based com-
pany, you should realize that as smart as you are on the technical
side of things, you probably do not know much about business.
This will be particularly true if you are founding the company
straight out of graduate school or a postdoctoral fellowship,
though this point can apply just as much to professors! You will
need other people – people with different backgrounds and exper-
tise – to help create a successful company. The following is a brief
overview of the people you will need in your team to build a suc-
cessful start-up.
2.2.1. Management Team You should preferably try to find people with previous start-up
experience, like serial entrepreneurs who have gone through the
ups and downs of starting their own companies and have previous
experience or familiarity with the markets or industry you are
focused on. These people will be directing and mentoring the
people in the company (including you), so their vision and the
way they interact with people is important. If you are lucky, you
may be able to recruit experienced entrepreneurs who have been
successful in past companies and may be willing to invest in your
company, along with providing experienced management skills.
2.2.2. Scientific Team Remember, your company will need people with a wide range of
technical skills in addition to those of the members of the founding
team. This means that you are not necessarily going to want to
hire people with the same background as yourself, but will instead
want to look for people with complementary skills. The specifics
of what you need will, of course, depend on the nature of the
technology you are commercializing, as well as the markets that
you are targeting. Keep in mind that you have (likely) not worked
in the industry that will use your products, so recruiting someone
from that industry would be useful. Also, if you are a scientist,

you may need engineering resources to scale-up your product or
assist with other types of product development. Note that the
number of people you hire on the management and scientific
teams will obviously depend on your resources, and the type of
people you will hire as employees #1 and #2 will likely be very
different from employee #30.
2.2.3. Service Providers In addition to the people inside your company who work for you,
you will also need assistance from a wide range of other people as
you and your new team establishes and grows the company. These
people will include lawyers, accountants, bank account managers,
Web site designers, and many others. Particularly for the more
critical service providers, it is important to make sure that they
have experience working with new companies in your industry.
After all, you are unlikely to be experienced in these areas, and
you want service providers that can keep you from making com-
mon early stage mistakes. These run the gamut from things as
simple as missing payroll tax payments to accidentally publishing
a new technology before filing for a patent – invalidating any
chance you have to get a patent on that technology. Ideally, your
service providers will also be able to provide other assistance,
including introductions to financing sources, potential hires, and
potential customers. The chances that they will be able to assist in
this way are greatly increased if they have past experience working
with companies such as yours.
As a new company, you are likely going to have limited
resources that you can spend on instruments, so you will also
need to find facilities and people that you can use for analytical
services. Universities can be ideal for this purpose as their facilities
can be less expensive to use than those in industry; it is also typi-
cally less expensive to use university facilities rather than purchas-
ing the instrumentation yourself. Further, you can often access
government grants to set up collaborations with universities.
However, it is again important to make sure that the specific
group you are working with at the university has experience
working with companies like yours, as you are likely to be moving
at a very fast pace and will need fast, effective turnaround of
analytical services.
2.3. The Product Products and sales are the most important reason for your
and Sales company’s existence. If no one wants to use your product, you
have a product with no value. Ergo, you have a company with no
value.
Your technology foundation will need to be made into prod-
ucts that are critical to the market, robust, and economically viable.
If your technology is none of these, but it is interesting to you
and your staff, it will keep you interested in your work, but
ultimately it should stay as a university research project, and not

be the foundation for a company. In addition, if you get too
excited about the technology and forget about what your
customers will want, you might as well stay in academia.
Product development and sales in start-ups are an interlinked
and iterative process, largely driven by the need to fund growth.
Typically, you will start by creating an “alpha” product internally
and then fix problems in it to create a “beta” product that is then
trialed with initial customers. This product is then modified again
to create something that is ready for a broader market launch.
The materials that we currently sell to researchers are always being
improved – as we receive feedback; we add characterization details,
tighten specifications, add new products, and decrease costs.
This continuous improvement will lead to the best possible
product – and it is critical to start that process as early as possible.
It is important to get your initial trial customers to help with
product development, as they can provide: technical guidance,
market access, customer validation that you are focused on the
right problem, and reference customers, for other sales and
importantly for financing your next development. In addition,
these initial customers act as validation to investors to give you
additional money, which you can then use to build the next
product iteration.
It is also important to target the right size of initial customer.
Everyone – you, your investors, and your family – will think it is
fantastic if you land a large company as a customer. True, it will
help you impress investors and enable them to dream of what
would happen if your product is successful. But these “marquee”
clients will be hard negotiators and the deals may not be that
lucrative. We found that small-to-medium-sized companies were
the most fertile ground for trial customers. These customers can
make decisions more quickly, act as true partners as they are closer
to your size and, in some cases, may still remember what it was
like to start a new business. Though you should still work to sell
to them, large company buyers may steer away from small
company product development projects – unless the technology
has great potential.
You may require additional staff to help build your sales.
Always remember, though, that your first salesperson is yourself.
You are inexpensive, know the technology, and should be able to
synthesize customer feedback to develop and improve your
product. But once you can no longer carry the burden alone, you
should add a business development person, rather than sales-
person. The distinction here is important. Our view was that a
salesperson is very good at using his or her contacts to sell large
quantities of a well-defined product when pointed in the right
direction. A person specializing in business development, on
the other hand, suits an earlier stage of product development.
These are people who can be trusted advisors for target customers,
walk the halls to determine what their needs are, and then inform
product development staff. In addition, a salesperson is largely
commission driven, but an advisor typically is compensated with
a higher base and lower commissions.
2.4. The Cash There is no one who will give you money for free. You need to
demonstrate value before any investor – besides your family –
investor will think that there is a value in what you do. You will
need to iteratively build your products and customers to demon-
strate value to investors.
It is important that you constantly try to get money – and not
worry about giving away too much when you get it. Some people
say that you should think of your shares as if they could be worth
$100 each in the future, but they will not get there unless you
have money. Many start-ups go out of business simply because
they run out of money, not due to a bad idea or product. One
thing that you can count on is that things always take longer, and
cost more, than you expect.
There are a range of sources for financing your venture,
including:
●● Friends and family – Fairly evident from the title, these are
people who are close to you that will often invest because
they trust in you. It is your responsibility to ensure that this
trust is not misplaced and that they are suitable candidates for
risky investments. It can be very uncomfortable at events with
friends if you just lost their money. Frequently, arm’s-length
investors want to ensure that friends and family are invested,
as it is a test of your resolve to make the company work. Your
desire to not disappoint your friends and family then is a big
driver to ensure your success.
●● Angel investors – These are typically individuals or groups of
individuals who have from ten thousand to several hundred
thousand dollars to invest in early stage companies. They are
the bridge to later financing and will typically invest under
terms that give them special preferred rights.
●● Government grant and regional development programs –
These are typically grant or loan programs that we have found
to be very helpful in providing growth capital without giving
away large amounts of the company. They will vary from
country to country, but most governments have various
programs in place.
●● Customers – This is the ideal source of financing as it provides
market validation and reference customers for other sales
while building your company and not diluting your owner-
ship stake. Customers can sometimes creatively finance capital
expenditure and may be eventual buyers of your company.
●● Venture capital – This is institutional money that invests in

high-growth companies. This can be very useful, and most
people look at this as the “holy grail” because they hear stories
of venture-backed companies like Google that have done
fantastically well. But it is important to remember that the
chances of getting venture funding are very low, and typically
only 10% of those companies are really successful. In addi-
tion, only certain types of companies will be attractive to these
investors, and you will largely be signing control of your
company over to someone else.
In our case, we built Vive Nano largely through angel investors,
customers, and grants.
As a scientist, you should not rely on your ability to learn
about and raise financing on your own. It is critical to add financing
capability at an early stage. While our company’s initial manage-
ment team was all technical, the next stage of the management
team’s development was in hiring individuals that could provide
our company with the ability to finance itself. Many companies
fail because they have outlasted their cash, so this was a critical
development. So crucial, in fact, that at one point, 60% of our
management team had experience in raising financing.
You should realize that the financing process is rarely fast. We
have seen a typical “fast financing” take 6 months and a longer
financing took 18 months. You will need to budget for this when
considering your cash needs.
Lastly, once you get the money, manage it very carefully. One
of us (Keith Thomas) used to be a banker and would quickly take
loans away from companies that were “building temples to them-
selves.” If you are focusing more on working in – or worse –
building a lavish workspace, you have taken your mind off getting
and keeping customers.
The Chinese philosopher Lao-tzu said that, “A journey of a
thousand miles begins with a single step.” The science you have
created in the university laboratory is the first step but only the
first step.
You may believe that you have invented the next world
wonder. You might be right. But you will definitely need other
people’s help to turn your science into something that will have
any benefit. Business people will tell you that it is a long road to
turn your science into replicable technology, find a profitable
application, build it into a saleable product, and package it for
sale. This is all true. The entrepreneurial journey entails a lot of
risks – risks of failure and risks of success. You just have to have
it in your guts to take the risk. The rewards can be great, and
taking these risks is necessary if you and your technology are
going to change the world.
Good luck in your journey!
3. Notes
1. Consciously build a company culture. Culture can make or

break your company. At our company, we focused on building
a culture that is smart (smart in how we execute our work),
open (open to new ideas and new challenges), and respon-
sible (responsible in the application of what we do for the
greater good).
2. Select employees who are a good fit for your company’s
culture. Our company required different resources at each
stage of its development, but the cultural DNA of the people
at all stages was common. Our people are aggressive, critical
thinkers with the ability to network. To help foster a culture
of critical thinking, we have a rule that at least two sets of eyes
look at everything before it goes outside the company. This
rule helps ensure consistent messaging, but more importantly
makes sure that we test every document, research, or product
before it is presented to clients.
3. Do not insist on doing it all by yourself and ensuring you get
credit for it. It requires a range of partners to support the
company. You should be thankful for anyone who helps build
your company and share the credit with them. We were – and
still are! – thankful for the university, government, and corpo-
rate partners who pushed us in the beginning to do better
every day.
4. Do not only hire your friends. It is good to work with people
with whom you are compatible and potentially even with
your friends, but you may need to discipline or fire staff at
some point and this is hard to do to close friends.
5. Do not hire the wrong start-up manager. Every company
needs what is sometimes called “gray hair” or “adult supervi-
sion.” Adding this person is critical and where we have seen a
number of companies go wrong. Your company will require
the experience that comes from having “done things” before
to speed growth and avoid pitfalls, especially in cash and
people management. There are a number of former executives
from large companies who are looking to work in small
companies, but it is important to not accept the first “gray
hair” you see. A good start-up manager has experience in
building start-ups and can provide access to their own money
or network to get things going. In addition, that person will
be your chief salesperson – to employees, customers, investors,
funding agencies – until you hire a salesperson or business
development manager later. If they cannot sell, they are not
the right person.
6. Do not refuse to “let people go.” Not everyone will be able

to grow with the company, and this leads to hard decisions.
It is often better for the company and for the individual to let
that individual go after the job has surpassed their capa-
bilities. Even retaining an employee in a reduced role and
hiring someone to be their boss can lead to unnecessary
tension.
7. Remember that there is a difference between how business-
people and academics think. We have come to the under-
standing that this is one of the primary reasons why
commercialization of university science is so tough. You have
two partners working together, each with differing world-
views. The academic wants academic freedom – to research
what they want and to enable their discovery to be widely
known through publishing or conferences– while the busi-
nessperson wants the academic to perform dedicated
research and to keep the discovery quiet to enable patenting.
There are ways to ensure both goals are met, but they require
understanding from both sides.
Acknowledgments
D. J. A. and J. A. D. thank Professor M. Cynthia Goh, who has

been an excellent mentor for them both. The authors also thank
their coworkers at Vive Nano, who keep the company growing.
Chapter 25
Applying the Marketing Mix (5 Ps) to Bionanotechnology

Michael S. Tomczyk
Abstract
This chapter, based on concepts developed for my book, NanoInnovation (Tomczyk, Nanoinnovation:
What Every Manager Needs to Know, 2011), is one of the first attempts to evaluate nanotechnology in
the context of the “marketing mix” – a conceptual challenge given that nanotechnology is not one product
or even a set of products, but rather a technology that is incorporated in an expanding list exceeding a
1,000 products – encompassing materials, structures, processes, and devices. My purpose is to use this
context to identify some of the critical issues and factors that will influence development of “nanotech-
nology markets” at this very early stage in the evolution of nanotechnology, and more specifically,
bionanotechnology. As technological innovations continue to promote the market growth for nanotech-
nology, especially in the field of medicine and healthcare, sensemaking frameworks are needed to help
decision makers keep pace with these evolving markets. One of the best frameworks is the “marketing
mix” which has been used for decades to identify the controllable factors that decision makers can
influence through marketing strategies. With so many game-changing innovations poised to move from
nanotech research to commercialization, marketing issues are becoming increasingly important to
decision makers in science/academia, business/venture development, and government/policymaking.
Key words: Bionanotechnology, Bioscience, Biotechnology, Innovation, Marketing, Nanobio

technology, Nanoinnovation, Nanotechnology
1. Introduction
The promise of nanotechnology, especially in medicine and

healthcare, is profound.
Each day, we hear how cancer researchers are using nano-
particles to tag, target, and destroy tumors. We read about
nano-enabled labs-on-a-chip and portable devices that will afford-
ably identify dozens of diseases, and bring sophisticated diagnostics
to remote corners of the world. Science journals describe how
nanoparticles are delivering drug molecules to diseased cells
without harming healthy tissue, and delivering therapies inside
393
394 Tomczyk
cells – activated by a pulse of light, a magnetic or electrical charge,

or infrared heat. Researchers are striving to create a test that can
read an entire human genome in less than an hour for $50 or
$100. These are only a few examples.
Marketing nanotechnology is not like marketing a traditional
product or service (1). Nanotechnology involves nanoscale mate-
rials, processes, structures, and devices, typically 1–100 nm in
size. In context, a strand of DNA is about 2.5 nm wide. The
diameter of the West Nile virus is about 50 nm (2).
The products that represent the nanotechnology market
cover an enormous spectrum of applications and industries. They
include imaging systems such as scanning tunneling microscopes
that help medical researchers observe, understand, and manipu-
late biological processes to effect novel diagnostics and therapies.
They include material forms that we have only known about for a
couple of decades, such as carbon nanotubes, quantum dots, and
nanowires, and variations such as nanoshells that can be used to
deliver chemotherapy molecules. A key capability of nanotech-
nology is the capability to manipulate collections of atoms and
molecules that exhibit quantum effects at the scale that make
elements such as gold – inert in bulk form – surprisingly reactive
at the nanoscale. These capabilities and phenomena are fueling
emerging fields, such as “theranostics,” which combine diagnostics
and therapy.
Semiconductors are undoubtedly one of the most familiar
nanotechnology products – nanoscale circuits have been present
in computers and cell phones for decades. Intel has developed
circuits as small as 22 nm, enabling the placement of 2.9 billion
circuits on a chip the size of a small fingernail (3).
As exciting as these innovations and possibilities might be,
they also introduce a variety of marketing challenges. Moving
what works in a laboratory to what works in a treatment regimen
can take years or decades to achieve – while patients and practitioners
read the headlines and assume that a treatment or cure is imminent.
For example, delivering a nanoparticle to one type of cell in a
Petri dish is different from targeting a given amount of nanopar-
ticles to a specific location in the human body, when the body
includes more than 200 different types of cells.
Beyond the research challenges, nanotechnology faces a
variety of image challenges associated with the fuzzy nature of
health, safety, and environmental issues, many of which are unre-
solved. For example, until recently there were no globally accepted
standards for nanomaterials or engineered nanoparticles. If you
buy carbon nanotubes from two suppliers, the specifications can
vary considerably. Studies linking nanoparticles to disease or
environmental contamination are still being validated – which
raises yet another marketing issue: should manufacturers advertise
“nano-inside” or “nano-free” on their labels, or ignore nano-
branding altogether?
Applying the Marketing Mix (5 Ps) to Bionanotechnology 395
To put these issues in context, I present some sensemaking

frameworks and personal perspectives using a classic marketing
framework known as the “marketing mix.”
2. The
Nanotechnology
Marketing Mix
In traditional marketing, companies use a framework called the
“5 Ps,” based on factors that decision makers can control and
incorporate into their business strategy. Traditionally, the 5 Ps
include the following: Product, Price, Place, Promotion, and
People. A version proposed by Dev and Schulz in 2005 (4) builds
on this framework with an expanded next-gen mix called SIVA,
an acronym that stands for: Solution, Information, Value, and
Access – where Product becomes Solution, Price becomes Value,
Place becomes Access, and Promotion becomes Information.
Although Dev/Schultz did not include the fifth P (People)
in their framework, I am including People and adding the
dimension of “Community,” which expands “SIVA” to “SIVAC.”
Communities of interest have become an important element in
global marketing. In the case of bionanotechnology, communities
of interest include scientists and research groups as well as patients
and practitioners.
Each of the five parameters in the marketing mix poses some
thought-provoking issues and challenges for decision makers.
In Table 1, I have applied this expanded marketing mix to
bionanotechnology and included some general examples for each
parameter (1).
2.1. P1/Product The first wave of products/solutions that emerged from nano-
(Solution): Nanotech technology included tools such as imaging systems, materials such
Definitions, Maps, as carbon nanotubes, and nano-sized catalysts, which are more
and Killer Applications efficient than comparable “bulk”-size catalysts. Nanotechnology
products range from raw materials such as silver nanoparticles
used in antimicrobial coatings (for example, in medical environ-
ments) to titanium dioxide nanoparticles used in sunscreens.
Probably, one of the best-known nanostructures is the carbon
nanotube (CNT), a tubular form of carbon molecule that is
typically 1 or 2 nm in diameter. CNTs come in several varieties
including single-walled and multiwalled nanotubes – most of
these are used in applications that are embedded in industrial
processes and hidden from view.
2.1.1. The “Products” In nanotechnology, the “product” we are marketing is not always
of Nanotechnology tangible; this makes the SIVAC term “solution” more appropriate.
“Bionano” involves biological structures and processes as well as
engineered materials and devices. These solutions range from
nano-sized drugs, to therapies that use nanoparticles to deliver
396 Tomczyk
Table 1
Applying the marketing mix to nanotechnology
Examples (5 Ps) 5 Ps SIVAC Examples (SIVAC)

(a) Labs-on-a-chip P1 Solution (a) $1 disposable disease test
(b) Nanoshells for drug delivery Product
(b) A “silver bullet” cure for cancer
(c) Nano-sized drugs for
bioavailability (c) Longer lasting chemotherapy
drugs with minimal side effects
Nanoparticles and nanoshells are P2 Information How cancer patients can enroll in
being used to treat cancer; Promotion clinical trials using nanotherapies;
the first patients have been where clinical trials are underway
“cured” (experimentally)
$1,000 personal genome (not P3 Value Affordable personal gene profile to
currently available, but an Price identify susceptibility to disease
industry goal) (especially inherited conditions)
Medical research centers and P4 Access Clinical trials, experimental research
research hospitals Placement programs; foreign countries with
these
Patients with diseases (such as P5 Community (a) High-risk patients who need to
cancers) that are most responsive People have their cancers detected as
to nano-enabled (a) diagnostics soon as possible
and (b) therapies (b) Patients who need nanotherapy
to minimize chemotherapy side
effects (these communities can
share life-saving info online,
worldwide)
drugs or destroy diseased cells, to digestible computer chips that

can be embedded in a pill and monitored wirelessly.
While there are only a few examples of “nanodrugs” on the
market today, dozens of nanodrugs and nano-enabled therapies
are currently in clinical trials in the USA and overseas. Nanosizing
a drug can make it more soluble (and thus more bioavailable),
allow it to remain intact until it reaches the desired organ, or
provide a “time-release” feature that keeps it working longer.
One of the early successful uses of bionanotechnology was the
ability to deliver water insoluble drugs to tumors. Early examples
include nano-sized drugs coated with a liposome, most notably
Doxil (a first generation drug approved in 1995) and Abraxane.
Abraxane is a nano-sized formulation of the cancer drug
paclitaxel that uses nanoparticles made from the human protein
albumin to treat metastatic breast cancer. In 2009, Abraxane was
approved for the treatment of pancreatic cancer. Abraxane
minimizes toxic side effects and delivers a 50% higher dose in
30 min than standard paclitaxel, which originally needed to be
administered over several hours (5).
In bionanotechnology, sometimes we are marketing a

discovery – such as how to unwind a DNA molecule so that it
can be maneuvered through a channel on a biochip. We may be
marketing a phenomenon such as increased solubility that results
when a drug is nano-sized, or a new capability that a researcher
has demonstrated in a laboratory showing, for example, how
wrapping a drug in a lipid makes it more bioavailable. The
“product” may be a revelation that explains why a disease resists
or responds to a particular treatment. While not always of
immediate commercial value, such innovations do create value in
the marketplace, through shared research, patents, product
pipelines, and commercial applications.
Finding a comprehensive definition that encompasses imaging
systems, nanomaterials, bionanostructures and processes, devices,
and more is an important part of the marketing puzzle.
2.1.2. Defining the A core challenge in marketing nanotechnology (including

Nanotechnology “Product” “bionano”) is getting the definitions right. The problem is that
nanotechnology is not one discrete thing. It is a set of ideas,
concepts, and metrics that describe not only what exists and can
be observed and manipulated at the nanoscale, but also what is
possible in the future as we increase our ability to engineer atoms
and molecules.
Unfortunately, there is still no ubiquitous consensus definition
for nanotechnology, per se. Nanotech means different things to
different people and organizations. However, this is not as serious
a consideration as we might imagine, since the definitions used
are each tailored to the needs of the (mostly scientific) organiza-
tions that are doing the defining. The nuances are subtle, but
important. Here are a few examples:
The National Nanotechnology Initiative (NNI) defines nano-
technology as: “the understanding and control of matter at
dimensions between approximately 1 and 100 nm, where unique
phenomena enable novel applications.” The National Institutes
of Health and the US patent office also use the NNI definition.
The National Science Foundation uses one of the longest and
most comprehensive definitions for nanotechnology: “Research
and technology development at the atomic, molecular or macro-
molecular levels, in the length scale of approximately 1–100 nm
range, to provide a fundamental understanding of phenomena
and materials at the nanoscale and to create and use structures,
devices and systems that have novel properties and functions
because of their small and/or intermediate size. The novel and
differentiating properties and functions are developed at a critical
length scale of matter typically under 100 nm. Nanotechnology
research and development includes manipulation under control
of the nanoscale structures and their integration into larger
material components, systems, and architectures. Within these
larger scale assemblies, the control and construction of their
398 Tomczyk
structures and components remains at the nanometer scale.

In some particular cases, the critical length scale for novel properties
and phenomena may be under 1 nm (e.g., manipulation of atoms
at ~0.1 nm) or be larger than 100 nm (e.g., nanoparticle reinforced
polymers have the unique feature at ~200–300 nm as a function
of the local bridges or bonds between the nanoparticles and the
polymer) (6).”
Most definitions of nanotechnology reference the nanoscale –
from 1 to 100 nm. However, if we use this scale we exclude atoms,
which are under 1 nm in diameter – and manipulation of atoms is
an important function of nanotechnology. The European
Commission addresses this by using the term “of the order of
100 nm” which includes some flexibility at both ends of the
scale (7). Many bioscience applications – stem-cell scaffolds and
clusters, use of nanoreagents in diagnostic tests, drug delivery
systems, and convergent applications that combine semicon-
ductors with diagnostics – may be greater than this scale; but
most of these applications do involve nanoparticles, nanoscale
structures, and processes.
Further, there is no consensus definition for bionanotech-
nology, which is often characterized as the intersection or conver-
gence of biology and nanotechnology. This is also the juncture
where novel applications for medicine are being developed.
Nanotechnology structures and processes – atoms and molecules
and their interactions, essentially – form the foundations of
biology and chemistry. Many biological processes, such as self-
replication of biological structures, are ingrained in Nature.
Medical researchers are drawing inspiration from biology.
Examples include self-replicating and self-healing systems, nano-
scale structures that can be used as stem-cell scaffolds, and various
types of biomarkers and drug delivery mechanisms. The term
“biomimetics” describes the field of replicating natural structures
and processes, sometimes called “biomimetic nanotechnology.”
2.1.3. Mapping The accompanying technology map (see Fig. 1) shows the com-
Bionanotechnology plexity of bionanotechnology, with particular focus on human
health care (1). This map is not intended to be comprehensive,
but rather representative of the range of functions, applications,
and innovations encompassed by bionanotechnology. This is a
dynamic map, which is constantly changing. It shows the myriad
domains encompassed by bionanotechnology, including some
emerging sectors that are being created by the convergence of
different streams of research such as theranostics (the conver-
gence of diagnostics and therapeutics).
While these are technology sectors and not market segments,
they offer a starting point for understanding the promising
potential of bionanotechnology to address a wide variety of
markets in the key areas of instruments and tools, diagnostics, and
Probes
Imaging
Labs Systems Novel Nanoskins,
on a Scaffolds,
Sensors Drug
Matrices
Chip Delivery
Biochips Nanoscopic Instruments & Tools
Systems
Gene
Diagnostics Theranostics Therapies Therapy
Biomarkers DNA
Magic
Genetic Nano
Nano Particles Bullets
Tests Drugs
Disease
Prevention
Antimicrobial
Materials, Filters, Membranes
Fig. 1. A bionanotechnology technology map. Copyright © 2011 by Michael S. Tomczyk. All rights reserved.
therapies – including “killer applications” that are the bionano market

incarnation of these emerging technologies (see Subheading 2.1.4).
Each of these circles represents not just commercial products,
but also a set of very important and promising solutions across a
broad spectrum of possibilities.
2.1.4. Killer Applications In any market, the hottest products are those that address killer
applications. A killer application (or “killer app”) includes a use of
technology that provides a core value, such as a computer operating
system, which is where the term originated. In popular jargon, a
killer app can also include an application that becomes so popular
and widely diffused that it establishes itself as indispensable.
Many of the most important applications are under develop-
ment – and in this sense, it can be argued that bionanotechnology
is lagging the “nanotech revolution” in contrast to semicon-
ductors, which are leading the charge. This is due in part to the
long development and testing cycle for biomedical research in
general, as well as the need to overcome technical issues that
range from scaling up laboratory results to providing efficient
methods for microencapsulation, genetic manipulation, etc.
In medicine, there are several holy grails that bionanotech-
nology is striving toward, the foremost being the quest to find a
novel treatment or cure for cancer – one that destroys cancerous
cells and tumors without damaging healthy cells and tissues. Many
novel therapies that use nanoparticles are being researched and
tested, including those which can be activated or triggered by
electricity, heat, magnetism, and light. One experimental therapy
involves using hollow gold nanoshells to deliver a drug to a
specific disease site. Another approach involves concentrating
gold nanoparticles in tumors and using near infrared radiation to
heat the particles to kill the tumors (leaving the surrounding
400 Tomczyk
healthy tissue undamaged). Yet another involves tagging a drug

with a semiconductor device similar to an RFID chip – made out
of digestible materials – that tracks and transmits the location of
the drug, or simply confirms that it was swallowed (an important
consideration given that an alarming percentage of patients delay
or stop taking their life-saving medications). In the area of diag-
nostics, researchers are investigating novel methods for detecting
cancer cells before they form tumors.
Other “killer applications” that are being intensively pursued
by research teams include: delivering disease-fighting molecules
to specific cells, tissues, and organs that are difficult to treat with
conventional therapies, delivering genetic materials using nano-
carriers, providing controlled release of biological chemicals such
as toxic chemotherapeutic agents, reducing the size (and cost) of
reagents used in diagnostic tests, reducing the cost and effi-
cacy of labs-on-a-chip, automating and cost-reducing personal
genome sequencing, and using nanocapsules to target and deliver
drugs to disease sites. Nanotechnology researchers are striving to
revolutionize diagnostic testing by moving the field toward
point-of-care testing. Scientists are engineering portable diag-
nostic systems that can be used in a doctor’s office instead of
having to be sent to a specialized laboratory. Portable diagnostic
devices can also be used to quickly and inexpensively identify
diseases in rural villages, and in remote regions of the world.
These are just a few examples.
Entrepreneurs, companies, and venture capitalists may view
these as “products” but they are much more than that. Most of
these medical applications represent solutions that will lower
diagnostic costs and greatly improve the delivery and effective-
ness of health care. As these solutions begin to move from labo-
ratories to commercial availability, there will be a need for a clear
marketing message to position these innovations and show the
relative benefits in relation to existing tried-and-true solutions
such as traditional drugs, surgery, medical devices, and other
modalities. It is not enough to simply announce that a new nano-
therapy now exists. Practitioners, hospitals, and clinics will need
to be shown that the “switching costs” justify adopting a nano-
technology solution, no matter how effective it might be. For
example, why use a nano-sized version of a drug instead of the
meta-sized version? One reason is that the nano-sized drug may
be more soluble and bioavailable. Such distinctions are important
and need to be included in product descriptions, to differentiate
novel applications from existing alternatives.
2.2. P2/Price (Value): In marketing, “price” includes value throughout the supply chain.
Trillions in Revenues, At the macro level, we can ask: what is the expected value of the
Thousands in Cost nanotechnology market? The “value” of nanotechnology is
Savings significant, whether we calculate revenues represented by the
entire nanotech sector, or cost/price savings for individual products.
Lux Research has forecasted that the US nanotechnology market

will reach $2.5 trillion by 2015, after adjusting for the effects of
the recession. However, as the Lux report observes, this estimate
includes the value of finished goods such as expensive automo-
biles that incorporate nanomaterials, which is not a true indica-
tion of the value of the nanomaterials themselves. In 2000, the
National Science Foundation predicted that the nanotechnology
market will reach $1 trillion by 2015, with nanomedicine repre-
senting about $180 billion. Since then, other industry watchers
including Cientifica and Lux Research (8, 9) have endorsed this
prediction, with the caveat that these trillion dollar predictions
include the value of nano-enabled products, which is much greater
than the value of nanomaterials themselves. Lux Research has
observed that the nanotechnology “market” is actually a nano-
technology “value chain” that ranges from nanomaterials to
nanointermediaries to nano-enabled products. By any measure,
we are talking about a technology-driven, trillion dollar industry.
One of the benefits of nanotechnology is the ability to reduce
the cost (and price) of expensive products and services. If we drill
down to the price of individual nanotechnology solutions, we can
easily gain a sense of the value proposition that lies ahead, even if
only a few nanoinnovations achieve success.
For example, taking advantage of the larger surface area-
to-volume ratio of nanoparticles enables the use of smaller
numbers of particles in expensive diagnostic tests. This lowers the
cost and can make the tests faster.
Many innovations (most notably nanotech “labs-on-a-chip”)
will lead to compact, disposable, “instant” diagnostic tests for a
variety of diseases. The cost target for these tests is $1 per test.
The combination of labs-on-a-chip with portable readers offers
the potential to deliver quick, convenient tests to people in cities
as well as in rural areas. The low cost can make diagnostics
available in developing countries that otherwise could not afford
this technology. Portable diagnostics may also make home tests
more readily available, which could eventually change the paradigm
for medical diagnostics in general.
Personal genomics is another emerging market that will benefit
from lower costs enabled by nanotechnology. The first individual
human genome sequenced in 2007 cost about $60 million. The
second genome to be sequenced was that of DNA pioneer James
Watson – 454 Life Sciences did Dr. Watson’s profile in 2008,
which took about 2 months and cost under $1 million. Currently,
personal genome tests cost up to $200,000 depending on the level
of detail that is required (10); with six billion base pairs of DNA in
an individual genome, the processing power, time, and costs
remain high. Several gene testing services offer targeted tests at
lower costs, checking for specific traits and diseases, including
testing for “point mutations.” These profiles are less expensive –
$400–$3,000 – however, they are not complete. They do offer a
402 Tomczyk
way to identify specific genes in at-risk populations where

hereditary conditions may exist. The current objective, expressed
by companies such as IBM and Invitrogen, is to lower the cost
of a reasonably detailed personal genome analysis to $1,000.
Several research teams are working to unwind and analyze DNA
molecules – most notably the Philadelphia-based Princeton
University spin-out BioNanomatrix, which has announced that it
will beta-test its single-molecule DNA analyzer in 2010 (11). This
line of research could result in targeted DNA tests at a price as low
as $100–$200.
Of course, price and value are relative, when you consider
that a bionanotech solution may be an early stage therapy or
medical choice that can save a life – or thousands of lives. And in
an era when it is imperative to reduce the cost of health care,
nanotechnology innovations could play an important role in
delivering these life-saving solutions at an affordable cost.
If bionanotechnology delivers extremely cost-effective solu-
tions, it may be easier to market these products/technologies; on
the other hand, it could be more difficult to market these solu-
tions, if the profits are too low to recover research investments, or
if the solutions (such as $1 laboratory tests) are commoditized.
There are also “switching costs” involved in transitioning
from an existing technology to a bionanotechnology solution. It
is conceivable to envision a time when there will be a dozen
different ways to deliver drugs and treat diseases that use nano-
technology and that are radically different than current proce-
dures. Whether these are heated nanoparticles, or metal shells, or
lipid envelopes, or boxes made from DNA, or magnetic materials
such as iron oxide, or other innovations remains to be seen. The
marketing question is: when these solutions do become available,
will practitioners switch from surgery and drugs to nanotherapies,
and if so, how soon?
2.3. P3/Promotion Two decades ago, medical solutions were provided by practitioners
(Information): Nano- through hospitals and clinics, but patients were often adminis-
inside or Nano-free? tered solutions without fully understanding the exact purpose of
a surgical technique, why a particular drug was prescribed, alter-
native treatments, or salient details concerning their medical
condition. Today in the era of ubiquitous communication, the
same information available to medical researchers and practi-
tioners is available online to the patients who are the “customers”
of bionanotechnology. Unfortunately, some aspects of bionano-
technology, including the safety of many nanomaterials, are still
largely unknown, and these unknowns can color public percep-
tion and acceptance.
Strategically, one of the decisions companies need to make is
whether to promote the use of nanotechnology in their products,
or not. This is a “nano-inside” or “nano-free” decision that could
have a positive or negative impact. This marketing decision depends
on many variables, from the value of a solution (e.g., a medical

cure versus a cosmetic cream), the risk–reward tradeoffs, and
perhaps most important, whether the public accepts and trusts
nanotechnology in general or not.
2.3.1. Public Acceptance So far, it appears that nanotechnology has enjoyed a free ride in
terms of public acceptance. The term “nano” has not encoun-
tered the level of resistance that plagued genetically modified
organisms (GMO) in the 1990s. This could be due to low public
awareness, or to the fact that there has not been a public relations
disaster or events to trigger a wave of public concern. For
example, genetically modified foods were being introduced in
Europe soon after the “mad cow disease” scare, which made the
public especially sensitive about scientists and researchers tampering
with their food supply.
Many nanoinnovations are marketed in the media as if they
are already available, although most are years or decades away –
these nanotech “breakthroughs” are trumpeted on the Internet
and even in the most reputable science publications.
In November 2009, I did a quick search on Google for
“nanotechnology breakthroughs” and generated 7.1 million hits.
Does this represent marketing hype, or hope, or glimpses of
realities to come?
Let us say that some of these seven million search items are
redundant. Even at a rate of ten redundancies for each item, this
equates to 700,000 breakthrough items online. So let us assume
these search items span at least a decade, so divide the total by 10
and this yields 70,000 “breakthrough” items per year.
Given the amount of buzz around nanotechnology and any-
thing that carries the name “nano” (including such products as
cars (the Tata Nano) and media players (the iPod Nano) that may
not actually include nanotechnology), promotional awareness of
“nano” would seem to be fairly well established. Americans in
particular should have a strong awareness of nanotechnology,
given the amount of activity on the Internet – and the fact that
there are more than 1,000 consumer products that use nanotech-
nology, according to the Project on Emerging Nanotechnologies
at the Woodrow Wilson International Center for Scholars (http://
www.nanotechproject.org).
However, while we are supposedly being flooded with nano
news, several sources report that the public at large, including
college graduates, is largely uninformed when it comes to nano-
technology. According to one 2009 study, 68% of survey respon-
dents reported having heard “just a little or nothing” about
nanotechnology. Other studies have confirmed that public aware-
ness of nanotechnology is low, including Andrew Maynard, chief
science advisor for the Project on Emerging Technologies (12).
So, how do we reconcile this enormous number of nanotech-
nology breakthrough search items with survey results that show
404 Tomczyk
that a majority of Americans (and probably people in most

countries) are largely unaware of nanotechnology? Is there a
communication disconnect? Are only a minority of knowledge-
able insiders reading these thousands of announcements, articles,
news items, and blogs?
One interpretation of these results is that the plethora of
nanotechnology news is only reaching an elite audience or that
nanotechnology simply has not resonated with the public. This
can also be explained by the fact that most nanotechnology
products to-date have been hidden in industrial processes, such as
coatings and catalysts, and are not designated on consumer
products including foods that contain nanoparticles.
In nanobiotechnology, only a few dozen drugs have been
approved, and the 100 or so in clinical trials are not widely publi-
cized. Most companies do not advertise the fact that they have
nanotechnologies in their products. You will not find a “nano-
technology” label on most sunscreens, for example. However,
this is changing. In November 2009, the European Union
announced a regulation requiring all cosmetics products to
include the word “nano” in brackets after any ingredient that is
less than 100 nm (13). This labeling requirement applies to all
cosmetics marketed in the EU but could easily be extended to
other product categories, from foods to medicines. This regula-
tion could also provide a framework for new labeling standards in
other countries including the USA.
Opponents of the new regulation have argued that this
designation could be viewed by the public as a warning. The issue
is, how will “nano-labeling” affect the marketing of consumer
products, foods, and medicines? It depends on whether the
pro-nano or antinano forces win the tug-of-war for public
perception. Whether “truth in nano” laws and labels are beneficial,
detrimental, or carry no effect remains to be seen.
Regardless of whether “nano” is perceived as a threat or a
benefit, it is easy to envision a day very soon when nano-sized
drugs, nanoscale therapies, and in vivo diagnostics or theranostics
that involve nanomaterials will require doctors, nurses, and
patients to sign consent forms. Tighter controls could be trig-
gered by insurance companies providing risk protection, or by a
specific “nano incident.” A few nano incidents have already
occurred, although they did not receive widespread attention.
A major incident could, of course, damage public opinion and
lead to laws and regulations that could restrict the use of nano-
technology, including biotechnology therapies.
2.3.2. The Implications There have been a few notable examples of PR brushfires
of a Nanoincident involving “nano” although they did not turn into firestorms. In
2006, an aerosol product called Magic Nano, a household glass
and ceramic tile sealant sold in an aerosol can, was blamed for
over 90 customer reports of respiratory distress in Germany.
The product was promptly recalled. However, it was not

confirmed that the aerosol particles were in fact nano-sized.
A side issue that this case raised is that a product such as a solution
in an aerosol can may not be nano in its present form, but it may
become nano when you push the button on the top of the can.
Another example of negative publicity involves the use of
silver nanoparticles used in Samsung washing machines, refrigera-
tors, and other appliances under the trademarked brand, “Silver
Nano™” also marketed as “SilverCare™” and as the “Silver Nano
Health System™.” Samsung claims that this silver nanotech-
nology sterilized against over 650 types of bacteria by releasing
up to 400 billion silver ions into fabrics to create a sterilizing
protection up to 30 days after washing (14). It is feared that these
particles will find their way into waterways and kill fish and other
aquatic life – which is a real concern given that ionic silver can be
toxic to aquatic life. While metallic silver is not water soluble, if
you remove one electron from an atom of metallic silver you
create silver ions which are not only water soluble, but toxic at
certain levels. Samsung’s Silver technology includes a patented
release mechanism for silver ions that claims to kill 99% of bacteria
in cold water. However, the wastewater from washing machines
eventually enters streams and groundwater, which is potentially
troubling since ionic silver can be toxic to friendly bacteria as well
as larger aquatic species. In 2007 the US Environmental Protection
Agency (EPA) determined that the silver ions in Samsung’s
washers were subject to the Federal Insecticide, Fungicide, and
Rodenticide Act (FIFRA) and needed to be regulated as a
pesticide (15).
Time will tell if nanotechnology dodges these and other
negative PR bullets. A very real concern is that one significant
“toxic nano” event could taint the entire field of nanotechnology.
Labeling the entire area of science and technology across indus-
tries with the descriptive term “nanotechnology” runs the risk
that if one incident occurs in any one sector, this could be extrap-
olated to the entire field of “nano.”
The most significant potential hazard – and potential image
problem – associated with nanotechnology has been the comparison
of carbon nanotubes to asbestos fibers. Carbon nanotubes have a
thread-like structure that is similar to asbestos, and studies have
demonstrated an “asbestos-like pathogenicity” in mice (16). The
question is that is this a problem for humans, and if so, is it
limited to nanotubes of a certain length – for example, longer
nanotubes seem to be more hazardous than shorter tubes. The
jury is still out on this issue, yet it points to the need to address
the handling and disposal of nanotubes and the specifications for
CNTs used in products. On the other hand, it can be argued that
nanomaterials have been used for decades in many industries. In
bionanotechnology, clinical trials are in place to determine safety,
and at the other end of the life cycle, there are procedures for
406 Tomczyk
handling and processing “biohazard” materials so that nanotech

waste residues should be able to be properly disposed of.
There are some other important marketing issues evoked by
these examples. First, labeling the entire field of nanotech-
nology under one umbrella term runs the risk that one negative
incident could taint the entire field. I believe that the term
“nanoscale technology” is more appropriate and that “nano-
technology” or “nanotech” – while effective as a buzzword and
currently friendly in the marketplace – groups the entire sector
across all industries under one label. If one fatal incident occurs
that is associated with a certain type of nanoparticle or nano-
material, it is possible that headlines could create a backlash
against nanotechnology in general.
Another aspect that companies in particular need to be aware
of is the possibility that some of their products such as paints and
coatings may contain nanoparticles and if so, they could wake up
one morning to discover that their existing products – which may
be a 100-years-old – are suddenly reclassified as “nano” and
forced to meet stringent and costly requirements for packaging,
handling, labeling, and insurance.
How companies and institutions deal with these image issues
will determine whether nanotechnology, including bionanotech-
nology, will be accepted as a customer-friendly technology like
microelectronics, or as a “customer-beware” technology such as
genetically modified organisms (GMO).
Informing the public about the benefits as well as the potential
(including unknown) risks of nanotechnology – while guarding
against misleading the public and encouraging fear-mongering –
is a major marketing challenge that scientists as well as marketers
need to address proactively.
2.4. P4/Place (Access): The “place” of nanotechnology is not a traditional retail outlet
Diffusing such as a pharmacy, or even the Internet. Most nanotechnology
Nanotechnology products such as carbon nanotubes are only available from a few
reliable sources worldwide. Most bionanotech solutions are being
researched in multimillion dollar laboratories in large corpora-
tions, research hospitals, and government-sponsored programs.
Aside from nano-sized drugs, the most promising emerging
innovations, the really radical solutions, are still being developed
in research laboratories or are only available in early stage clinical
trials. “Placing” these solutions in the health-care market requires
the solution to run a gauntlet of animal tests and clinical trials
before it is commercialized, and even then a new therapy needs to
compete with existing accepted therapies that range from surgical
procedures to drugs.
One marketing question we can ask is: how will bionanotech-
nology solutions be diffused, once there are more commercial
applications available? Manufacturers need expensive imaging
systems – or nanoimaging services – to produce nanoparticles,
fabricate nanoscale biochips and other devices, and perform quality

control on products that use nanotechnology. In medicine, nano-
imaging is needed to diagnose the impact of nanoscale therapies
and to monitor the amount of nanoparticles that linger in the
body. The cost of these tools can be a limiting factor to the devel-
opment and diffusion of nanotechnology products and solutions.
While the public may not worry about this issue, it is possible that
some solutions may take longer to reach the market, or to become
diffused, because companies developing nanosolutions may not
be able to afford the tools. The high cost of imaging systems also
may limit the ability of schools to train the next generation of
nanotechnology scientists.
2.4.1. The Role of Imaging Imaging systems play a critical role in the adoption and availability
in Bionano Diffusion of bionanotechnology solutions. Scanning tunneling microscopes
and other “nanoscopes” cost as much as $50,000–$500,000
depending on the accessories such as probes and sensors that are
included. The resource Web site, Nanotechnology Now, lists
about 200 companies under “Nanotechnology Tool Makers and
Service Providers (17).” If there is one parameter that could
expedite the diffusion of nanotechnology research worldwide, it
is the availability of lower cost imaging systems.
Until the tools of nanotechnology become more affordable
and accessible, access to “bionano” will be limited to research and
educational laboratories. This does not mean that we should
despair. Many analogous examples exist of innovations that
required expensive infrastructures to create the market, including
MRI imaging systems, medical implants such as pacemakers and
stents, and high-definition television and cell phone technologies.
We can learn from these examples as bionano therapies, diagnostic/
theranostic tests, and medical devices become available.
2.4.2. The Impact of Health Wherever a nanotechnology solution is developed, tested, manu-
and Safety Regulations factured, delivered, and/or used, there are important safety and
environmental considerations that need to be considered. Safety
issues in particular can impact not just market acceptance, but
commercial viability and access. Regulations by government
agencies have the power to impact public access in many ways,
from policies that restrict government funding for research
(e.g., stem cells) to regulations that require additional tests and
studies as part of a clinical trial.
While access to nanotechnology in general is currently not
regulated in most markets, globally – in most countries, nano-
technology falls under existing health and safety regulations. In
the USA, the FDA – which does not regulate “technology” per
se – has been studying how to regulate the “combination
products” represented by nanobiotechnology (drugs, medical
devices, and biological products) (18).
408 Tomczyk
The European Union enacted the REACH program

(Registration, Evaluation, Authorisation, and Registration of
Chemicals) in 2007. REACH requires companies to manage the
risks associated with chemicals (including nanomaterials) and to
provide safety information on process chemicals, mixtures of
chemicals, chemicals used to prepare or manufacture products,
and even chemicals that might be released from products.
Manufacturers and importers are required to gather information
on the properties of their chemicals, so they may be handled
safely, and to register the information in a central database run by
the newly formed European Chemicals Agency (19).
Most industrial organizations are still wrestling with nano-
tech standards. For example, while nanoscale features have been
used in semiconductors and electronic devices for more than a
decade, organizations such as IEEE, ISO, and IEC are still a few
years away from establishing standards for the use of nanomate-
rials in electronics and other sectors.
We have not seen a heated public debate over the safety of
bionanotechnology because there are so few solutions available at
this early stage in the development of this emerging market.
Bionano is lagging the consumer market in terms of applications,
which provides some breathing room – however, this “market
lag” suggests a scenario where we could see bionano being
subjected to the policies and regulations developed to address
“nonbio” applications of nanotechnology.
For example, if carbon nanotubes or “silver nanoparticles”
are shown to be hazardous to human health or to the environ-
ment, it is reasonable to expect that bionano applications in
hospitals, pharmacies, and diagnostic laboratories will face more
stringent controls than those that currently exist. On balance,
reasonable standards of safety and efficacy should prevail, and
these are precepts that already exist.
For the foreseeable future, it is conceivable that we will not
be able to advertise or describe many applications involving nano-
technologies as “safe” or “healthy” or “environmentally friendly”
with any degree of certainty until more research is available. In
the meantime, the ability of marketing professionals to guarantee
that there will only be positive effects and no negatives or liabili-
ties will remain in limbo until generally accepted standards, rules,
and policies are put into place. These policies will range from safe
handling during processing of nanomaterials to administration
and safety in the delivery of bionano therapies and to disposal of
devices and materials that contain nanomaterials. In this area, the
bionano community will benefit from existing procedures that
already exist for the handling of biohazardous materials.
2.5. P5/People As a “marketplace,” the nanobiotechnology market acts more like

(Community): Creating a community – actually, a network of several communities – than
a Global Nanomarket a traditional consumer market. On the technology side are scientists
and research teams who are networked through scientific

conferences, webinars, online publications – and the companies
and institutions that provide commercial solutions. On the
“customer” side are patients and practitioners who can receive
information about nanotechnology discoveries, as well as how
they relate to orphan diseases, medical therapies, charitable
services, and experimental research programs. Policymakers,
legislators, and regulators determine how solutions are provided
and regulated.
Influencing these communities are networks of media experts,
pundits, bloggers, and commentators. Nanoscale drugs and
therapies already on the market as well as those in clinical trials
and research laboratories are actively discussed and tracked by
advocacy groups and patient networks.
The key to reaching this vast networked market with bio-
nanotechnology solutions will be to demonstrate how a novel
bionano therapy is more efficacious and cost-effective than existing
treatments. This is especially true for therapies that are not drugs
or surgery but something entirely different that does not fall
neatly into either category.
The “market demand” for bionanotech solutions is strong.
Worldwide deaths from cancer alone are expected to increase
from 7.4 million in 1974 to 12 million people by 2030, according
to the World Health Organization. Bionanotechnologists are
working on a wide variety of novel therapies for cancer. In most
countries including the USA, people are living longer. They need
therapies that can provide better quality of life in their 70s, 80s,
and 90s. People in remote villages need nano-enabled portable
devices and test kits to help detect and treat diseases.
Innovations enabled by bionanotechnology offer the promise
to shrink the size and cost of biochips, to identify, track, and treat
disease at the cellular level, and to target diseases in organs that
were previously inaccessible. Nanotechnology scaffolds are already
enabling stem cells to grow into blood vessels and bladders; other
solutions on the horizon that hold promise include inexpensive
DNA tests, artificial synapses, and much more.
As they become available, some of these solutions will be
easier to communicate than others. A surgeon may one day tell a
patient: “We can use this drug to kill your tumors but it may kill
some healthy tissue as well, you will lose your hair, and you could
die from the chemicals…OR…we can inject nanoparticles into
your system that will migrate to the tumor cells, then we’ll use
radiation or magnetic resonance to heat those particles so they kill
your tumor with less destruction of healthy cells…and you can go
home 2 h after the procedure.” Of course, communicating the
implications of a DNA test result, or a genetic defect revealed by
a nano-enabled labs-on-a-chip, will be more complicated and
involve more sophisticated messaging.
410 Tomczyk
Perhaps the best people to explain bionanotechnology to the

beneficiaries of these innovations are the patients themselves. In
an era when health care reform is a vigorous public issue, large
communities of interest are already paying increased attention to
the promise of bionanotechnology. Patients and practitioners
maintain vigorous communities of interest that share life-saving,
life-enhancing medical information through blogs and forums.
As bionanotechnology solutions move from laboratories to
hospitals and clinics, “viral marketing” (an ironic term) will
communicate which of these solutions are most efficacious,
affordable, and available.
3. Summary
At this early stage in the evolution of nanobiotechnology,

scientists need to remain aware of the impact that the marketing
mix can have on the funding, development, adoption, perception,
and ultimate success of the solutions they are developing. In many
areas, what happens outside of the laboratory will play a vital role
in what happens to bionanotechnology solutions in the market-
place. Many of the marketing considerations mentioned in this
discussion will impact the translation of bionanotechnology
solutions from laboratories to hospitals and clinics.
References
1. Tomczyk, M. S. (2011) Nanoinnovation: 9. Lux Research (2009) Nanomaterials State of
What Every Manager Needs to Know. Wiley- the Market Q1 2009: Cleantech’s Dollar
VCH, Weinheim, Germany. Investments, Penny Returns.
2. Boutin, C. (October 9, 2003) Purdue team 10. McGowan, K. (June 22, 2009) Your Genome,
solves structure of West Nile virus. Purdue News. Now Available for a (Relative) Discount.
3. Intel Press Release (September 22, 2009) Discover.
Moore’s Law Marches on at Intel. 11. Karow, J. (October 20, 2009) BioNanomatrix
4. Dev, C. S. and Schulz, D. E. (2005) In the Plans Beta-Launch of Single-Molecule DNA
mix: a customer-focused approach can bring Analyzer for Q2 of 2010. Genomeweb.
the current marketing mix into the 21st century. 12. Hart Research Associates (September 22,
Mark. Manage. 14, 18–24. 2009) Nanotechnology, Synthetic Biology, &
5. McGee, P. (2006) Delivering on nano’s promise. Public Opinion. Commissioned by the Project
Drug Discov. Dev. 9, 12–18. on Emerging Technologies, Woodrow Wilson
6. National Science Foundation (December International Center for Scholars.
2009) Nanotechnology definition NSF web- 13. Stafford, N. (2010) New nano rule for EU
site, http://www.nsf.gov/crssprgm/nano/ cosmetics. Chem. World 7, 6.
reports/omb_nifty50.jsp 14. Samsung (March 29, 2005) Samsung Silver
7. Williams, D. (May/June 2008) Defining Nano Health System™ Gives Free Play to Its
Nanotechnology. Medical Device Technology. ‘Silver’ Magic. Samsung Press Release.
8. Cientifica (April 2007) Halfway to the Trillion 15. Bolles, D. (December 3, 2007) The Irony of
Dollar Market? A Critical Review of the Nanotechnology’s Promise. Water & Waste
Diffusion of Nanotechnologies. water News.
16. Poland, C. A., Duffin, R., Kinloch, I., http://www.nanotech-now.com/tool-makers.

Maynard, A., Wallace, W. A. H., Seaton, A., htm
et al. (2008) Carbon nanotubes introduced 18. US Food & Drug Administration (November 7,
into the abdominal cavity of mice show asbestos- 2009) FDA Regulation of Nanotechnology Products:
like pathogenicity in a pilot study. Nat. FDA and NanoPharmaceuticals. http://www.
Nanotechnol. 3, 423–428. nanopharmaceuticals.org/FDA.html
17. Nanotechnology Now (June 27, 2009) Nano 19. Bergeson, L. L. (May 23, 2007) REACH and
technology Tool Makers & Service Providers. Nano. Nanotechnology Law Blog.
Chapter 26
Managing the “Known Unknowns”: Theranostic Cancer

Nanomedicine and Informed Consent
Fabrice Jotterand and Archie A. Alexander
Abstract
The potential clinical applications and the economic benefits of theranostics represent a tremendous incentive
to push research and development forward. However, we should also carefully examine the possible downsides.
In this chapter, we address the issue of how theranostics might challenge our current concept of informed
consent, especially the disclosure of information concerning diagnosis and treatment options to human
subjects. We argue that our lack of data concerning long-term effects and risks of nanoparticles on human
health and the environment could undermine the process when it comes to weighing the risks against the
benefits. Our lack of an agreed upon framework for risk management in nanomedicine may require us to
adopt an “upstream” approach that emphasizes communication and transparency among all relevant
stakeholders to help them make informed choices that enable safety or progress.
Key words: Informed consent, Nanomedicine, Cancer, Theranostics, Ethics, Policy, Risk management
1. Introduction
Nanotechnology enables its users to control matter and exploit

novel phenomena and properties at the nanoscale. Advances in
nanotechnology provide an opportunity to develop innovative
interventions in various areas of medicine such as cancer treat-
ment. The heterogeneous nature of cancer tumors and the need
to target specific cells constitute major challenges nanomedicine
could overcome contrary to conventional drug therapies. Indeed,
some nanoscale particles or devices smaller than 50 nm can pen-
etrate cells while nanodevices smaller than 20 nm have the poten-
tial to move through the blood stream. These type of capabilities
may allow for better selectivity of drugs toward cancer cells, which
reduces toxicity, increases efficacy of chemotherapy, and leads to
better dosing of medications (1). Theranostics takes advantage of
413
414 Jotterand and Alexander
novel nanotechnology platforms by incorporating therapy, imaging

diagnostics, and cell targeting into one system with multifunctional
capabilities. Theranostic nanomedicine exploits the multifunc-
tional capabilities of nanoscale devices and the molecular knowl-
edge of the human body to diagnose, treat, and possibly, prevent
diseases (2).
Current research and development in theranostic cancer
nanomedicine focus on combining three major areas of nano-
medicine to engineer innovative multifunctional devices that are
capable of (1) targeted drug delivery, (2) diagnostic imaging, and
(3) genetic testing. This combination approach may offer clini-
cians an opportunity to diagnosis and assess disease, deliver a tar-
geted therapy, and monitor therapeutic responses simultaneously
(3, 4). Some believe that theranostics will become the future stan-
dard of care (5), because it enables clinicians to tailor a specific
treatment regimen to the biomarkers expressed by a single patient.
This may prove beneficial in cancer treatment, where high levels
of variations in molecular markers or extreme heterogeneity of
the disease occur (6).
Researchers at UT Southwestern Medical Center and the
University of Texas at Dallas are working on the development of
multifunctional nanomedicine platforms for cancer diagnosis and
therapy. They utilize a “bottom-up” strategy to create highly
integrated architectures with multiple functions for tumor target-
ing, imaging ultra-sensitivity, and controlled drug delivery. New
molecular targets of cancer are exploited to achieve greater imag-
ing specificity of molecular probes and higher therapeutic indices
of drugs. Other researchers focus on theranostic agents that
merge both diagnostic and therapeutic capabilities in one plat-
form. There are currently few viable approaches to theranostics,
and of those available, most combine existing fluorescence or
MRI imaging approaches with traditional chemotherapeutic
drugs in the same molecular complex. For therapy, single wall
nanotubes (SWNTs) efficiently convert absorbed near infrared
(NIR) light into heat, and the thermal ablation of model tumor
cells has been demonstrated. Not only is the therapeutic potential
for nanomedicines promising, but also analysts predict a bright
economic future for this market.
Market analysts believe that theranostic applications are
worthwhile because they see nanoscale structures as a transforma-
tive technology that will revolutionize our health care system (7).
Theranostic companies (imaging source) are partnering with
pharmaceutical companies to merge their diagnostic and thera-
peutic products into combinations (dual-use products) that may
improve our clinical outcomes. The upside for these partnerships
may be greater than the more traditional standalone pharmaceu-
tical firms, because their dual-use products may lead to reductions
in our health care costs. Some financial analysts predict that the
Managing the “Known Unknowns”: Theranostic Cancer Nanomedicine 415
global market for theranostics may be in the billions of dollars by

2014 (8), up to $40 billion (9). In fact, several major pharmaceu-
tical companies are seeking mergers with diagnostic imaging
companies to take advantage of these opportunities within the
US market.
While the potential clinical applications and the economic
benefits of theranostics represent a tremendous incentive to push
research and development forward, we should also carefully
examine possible downsides. In this chapter, we address the issue
of how theranostics might challenge our current concept of
informed consent, especially the disclosure of information con-
cerning diagnosis and treatment options to human subjects. We
argue that our lack of data concerning long-term effects and risks
of nanoparticles on human health and the environment could
undermine the process when it comes to weighing the risks against
the benefits. Our lack of an agreed upon nanospecific framework
for risk management in nanomedicine may require us to adopt an
“upstream” approach that emphasizes communication and trans-
parency among all relevant stakeholders to help them make
informed choices that enable safety or progress or possibly under-
mine them, depending on how the public and private sector han-
dle its content and delivery.
2. Implications
of Theranostics
in the Clinical
Setting The development of new therapeutic applications of nanomateri-
als in cancer research may offer cancer victims new hope but they
may also create ethical, legal, and policy challenges for its stake-
holders and their institutions. One major reason for concern lies
with the ability of “multicomponent nanomedicine with modular
designs” to personalize therapy. Personalization of therapy may
lead to “adaptive targeting” where a course of therapy is altered
in real-time as a response to the adaptive resistance of cancers (3).
Hence, the possibility of “adaptive targeting” raises concern for
the validity of the informed consent process that must account for
a real-time change of the therapeutic options (in this case dosage)
without a formal consent procedure to account for each new
change or alteration. Not only will “adaptive targeting” raise con-
cern for the validity of informed consent, but it also may reshape
our concepts of the patient–physician relationship (10). The use
of nanomaterials in combination with diagnostic/therapeutic
application may produce an “automation of medical expertise,”
where treatment may change without any exchange of informa-
tion between patients and their physicians, unless, of course, there
is a feedback mechanism for real-time assessment (11). But even
then, it is unclear whether patients will have sufficient time to
determine the best course of action especially if therapy involves a

higher level of toxicity requiring them to reassess their risks and
benefits. In other words, the distinction between the process of
communicating a diagnosis and the individual’s assessment of
treatment options may become blurred. Patients may be forced
to make therapeutic decisions without knowing all of relevant
information necessary to make an informed decision. In an ideal
world we may believe that researchers and clinicians foster the
best interests of human subjects and patients alike, and thus, a
change in therapy options would be immaterial. However, because
of the complexity of our medico-legal system, the consenting
process has transitioned from an ethical principle to a legal doc-
trine with legal consequences. Ultimately, theranostics not only
poses new ethical challenges to the process of informed consent,
but likewise it also creates challenges to existing legal and institu-
tional policies.
Any ethical, legal, or policy challenges raised by the transfor-
mative nature of theranostic nanomedicine, especially those
impacting the informed consent process, may be further compli-
cated by our lack of risk data and analysis (12–15). Even so, fore-
casters predict more, not less, research and development with
nanoscale materials and structures, especially those utilized in
nanomedicines and health care (7–9). To protect these develop-
ing markets and reap the health benefits related to nanomedi-
cines, all stakeholders must assume responsibility for insuring the
benefits of these emerging technologies outweigh their risks or
hazards. Given that the general public remains mostly ignorant of
nanotechnology, clinical research, and development of nanomed-
icines may continue unabated until our first health disaster or
nanomedicine scare goes public (16). Once this happens, stake-
holders may not be able to tackle risk issues after the fact or con-
trol public perceptions to preserve their use or markets.
Some commentators warn that the end result for all nano-
technologies will mirror those Europe experienced for genetically
modified organisms (GMOs) or nuclear power and other emerg-
ing technologies. A mistrustful public may simply shun all nano-
technologies to stifle further research and development. Reality is
we had our first nanoscare and -recall in 1999 when the German
manufacturer (Kleinmann GmbH; Sonnenbuehl, Germany)
recalled its aerosol dirt-repellant spray called Magic-Nano after it
reportedly sickened nearly 100 of its users (17). Fortunately for
Magic-Nano and its company, authorities discovered that the
“nano” was not responsible for the adverse events, because it lost
its “nanocharacteristics” during manufacturing. Nevertheless,
authorities simply had no choice but to hunker down and do
damage control while nongovernmental organizations (NGOs)
called for more toxicology studies and regulations. Although the
company survived its scare, many NGOs and interested stakeholders
remain worried about their risks and some express concern for
further research and dissemination of nanotechnology into the
market (17). Perhaps, the better approach is to focus on achiev-
ing more transparency for informed decision-making through the
informed consent (13–16).
3. Informed
Consent
During the last few decades, advances in biomedical research have
provided better diagnostic tools and increased the number of
therapeutic options. Theranostics harvest the benefits in both of
these areas. The hope is to increase objectivity in diagnosis to cre-
ate algorithms for a well-defined understanding of the condition
of the patient and optimization of treatment. However, treatment
cannot rest on objective information alone. It requires subjective data
gained from the perspective of the patient and the subjective
judgments of the physician (18). Indeed, the subjective dimen-
sion stems from the necessity to let the patient decide a particular
treatment option and assess what is morally acceptable for that
individual in relation to potential risks and benefits. In other
words, even though technology will provide tremendous accu-
racy in diagnosis and patient-specific treatments (e.g., pharma-
cogenomics and proteomics), it is highly unlikely that the
biomedical sciences will ever eliminate the role of subjective data
in patient care. Because medicine deals with human beings who
are masters of their own bodies and destinies, informed consent
constitutes an ethical, and by extension, a legal and policy
imperative.
The doctrine of informed consent fosters the autonomy of
individuals and protects them from abuse. It allows individuals to
decide what should be done with their bodies, where no one
should be compelled to act against their will (19). It also provides
guidance and boundaries within the patient–physician relation-
ship to encourage communication and trust. The process of
informed consent facilitates the meaningful exchange of informa-
tion between the physician or researcher and the patient seeking
treatment or subject participating in research, respectively (20).
These discussions should be more than mere recitals of checklists
of information whether they pertain to a choice of treatment or a
decision to participate in a research protocol (21). Both processes
have oral and written components, but their scope or information
content may vary, depending on whether the consent process
focuses on treatment or clinical research. If it is treatment, then
discussions generally include the diagnosis, nature and purpose of
treatment, material risks and outcomes commonly known or
expected for the particular patient, disclosure of all feasible
alternatives and risks, prognosis, if recommendations declined,

prognosis, if recommendations accepted, and disclosure of con-
flicts of interest. Conversely, the clinical research consent process
may involve identifying the study as research and its purpose,
describing the reasonably foreseeable risks and benefits, disclos-
ing appropriate alternatives, handling of confidentiality and
research-related injuries, supplying contact information, and
obtaining a statement of voluntariness (20). And unlike the pro-
cess for treatment, this may require an entirely new written con-
sent form, which must be reviewed and approved by an Institutional
Review Board (IRB) before it ever reaches a participant. Success
of the process in both treatment and clinical research requires
participants to address all components and meaningfully exchange
information to respect ethical norms, or it fails, subjecting parties
to potential legal liability. Not only is the processes of informed
consent essential for good care, but also its successful completion
keeps all parties out of our judicial system and avoids wasting time
and judicial resources litigating matters that could have been
avoided by respecting ethical norms.
Based on a previous analysis (22), we will reiterate our accep-
tance of the seven element model of informed consent developed
by Beauchamp and Childress (23) for the analysis of nanotech-
nology. Some scholars have promoted a three-element model
(decision-making capacity, voluntariness, and informed under-
standing) (24) while usually the following five elements are rec-
ognized: (1) competence, (2) disclosure, (3) understanding, (4)
voluntariness, and (5) consent (23). However, the complexity of
clinical trials warrants a more refined approach as proposed by
Beauchamp and Childress who recognize seven distinctive ele-
ments of informed consent, organized in three main groups:
1. Threshold elements (preconditions): (a) Competence (to under-
stand and decide) – the ability to make a rational decision; (b)
voluntariness (in deciding) – absence of coercion.
2. Information elements: (a) Disclosure (of material informa-
tion) – the patient/research subject must be fully informed;
(b) recommendation (of a plan) – the patient/research sub-
ject must be provided specific recommendations concerning
his/her medical condition and/or the research procedure;
(c) understanding (of 2a and 2b) – the patient/research sub-
ject must be able to process information and understand it.
3. Consent elements: (a) Decision (in favor of a plan) – the patient/
research subject’s ability to make a choice; (b) authorization
(of the chosen plan) – the patient/research subject must con-
sent to the treatment and/or experimental procedure.
The use of theranostics nanomedicine for cancer therapy calls
into question the validity of informed consent from an ethical
standpoint. As already stated, individuals might need to decide

about, and consent to, therapeutic options before a diagnosis is
provided. In order to avoid compromising the ethical standards
of informed consent, it is paramount for us to determine whether
we should rethink the consent process when the potential risks or
side effects are unknown. Specifically, there are concerns about
the disclosure of information (2a) and recommendation (2b) with
regards to the condition of the individual. How we adequately
inform an individual about his or her medical condition and the
potential risks of a treatment option is even more challenging in
cancer therapy because adaptive resistance of cancer cells leads to
“adaptive targeting” and real-time alterations of therapy (3). How
individuals will make decisions about, and consent to, potential
treatment options will depend on the type of information pro-
vided to them and a set of personal considerations when con-
fronted with two or more potential treatment outcomes
(risk–benefit analysis). What does or should the individual under-
stand about the risks associated with theranostics? What
mechanism(s) should a clinician or clinical researcher choose to
frame the discussion of scope of informed consent? Should there
be a centralizing theme if one component is therapeutic while
another component is nontherapeutic? To address these ques-
tions adequately requires a better understanding of the potential
risks and effects associated with the use of nanoparticles and nan-
odevices in therapy and diagnosis. Unfortunately, at this point, a
definitive answer is not possible due to our lack of data concern-
ing long-term effects.
4. Theranostics
and the “Known
Unknowns”
No one disputes our need for more information on the risks and
benefits of nanoscale materials including those utilized in thera-
nostics. Unfortunately, we are (1) dealing with materials that have
novel properties which we may not fully comprehend, (2) lacking
the necessary and sufficient toxicology studies we need to define
its hazards, exposures, and life cycles, (3) moving products from
our laboratories to our markets at a pace that exceeds the ability
of our existing ethical norms and regulatory frameworks to adapt
and respond, (4) lacking long-term studies concerning the major-
ity of the nanomaterials we produce, and (5) hoping our benefits
from nanomaterials, especially those in medicine and cancer ther-
apy, will outweigh our risks (12–15). The resulting risk issues for
the nanoscale materials in nanomedicines may qualify as what for-
mer Secretary of Defense Donald Rumsfeld calls the “known
unknowns” where the existence of possible outcomes is recog-
nized, but their actual occurrence is uncertain or unknown (13).
Not only will our failure to address the “known unknowns” of its
risks impact current research and development of new products
sooner rather than later, but also it may dissuade future partici-
pants from participating in clinical trials. Potential participants
may choose not to participate because clinical researchers may
not have the information necessary to communicate and explain
potential risks. Unfortunately, the negative experiences with a
commercial product such as Magic-Nano or those at Hospital of
Pennsylvania (“HUP”) following the death of one of its gene-
therapy participants may foretell of things to come if we do not
begin assessing and managing risks now (25).
In 1999, investigators at HUP were conducting a gene-therapy
trial when one of their participants experienced a tragic death.
After a thorough investigation by the FDA, HUP ceased all clini-
cal trials with similar agents and followed by some rather trou-
bling revelations about their clinical trials process with this agent
(25). Officials at HUP admitted under intense pressure from the
local press that the parents of the deceased did not know about
the risks of the vector in research animals. Worse still, the parents
did not know that some participants experienced flu-like symp-
toms and their son did not actually qualify for the protocol. Once
this information came to light, the parents claimed they would
have never agreed to enroll their son and participate. The bitter
lessons learned by everyone were communication and full disclo-
sure of the risks or hazards are essential. But in the case of all
nanoscale materials, including those used in nanomedicines, these
lessons may be lost because much of the information physicians,
patients, and participants will need for a meaningful exchange
during informed consent remains unknown (15).
Unfortunately, knowing the risks and balancing them against
their potential benefits may be easier said than actually done.
Clearly, most stakeholders from the basic scientists to policymak-
ers recognize that we lack information on the health, safety, envi-
ronmental, ethical, legal, and social issues associated with
nanomaterials (26–30). The problem is our “known unknowns”
related to risk identification, assessment, management, and com-
munication, or risk governance, are daunting at best (13, 31, 32).
Notwithstanding our “known unknowns” about their risks, we
also have local, national, and international regulatory gaps. Almost
every nation is researching and developing nanotechnologies
without drafting nanospecific laws or regulations because they
believe their existing regulatory schemes are adequate (26–29,
31, 32). In fact, most nations, especially the developing ones, are
shying away from agreeing to any formalized, transnational
approach to the regulation of nanomaterials because they fear fall-
ing further behind more powerful nations in the race to corner
the market on nanoproducts including those in nanomedicine.
Simply put, developing nations do not want to miss out on the
economic benefits of this technology. To regulate or not to regulate

may ultimately depend on what we learn about our “known
unknowns” on risk and how we assess, manage, or govern them.
5. Defining Risk,
Assessment,
and Management
The crucial question then is “what is risk?” Briefly stated risk rep-
resents the likelihood of an adverse event occurring that produces
consequences (31, 33). To adequately manage risk and reduce
adverse events, individuals must identify, assess, manage, and
communicate risk information. In short, individuals must know
the “known unknowns” related to risk, and the existence of a
framework helps to manage them. In a review of multiple risk
management frameworks over the past decade, Jardine and her
co-workers found that agencies responsible for dealing with risk
may do so using a variety of approaches, but all of them generally
require: (1) risk assessment (describing and estimating adverse
outcomes using hazard identification, dose–response assessments,
exposure assessments, and risk characterization), (2) risk manage-
ment (using a process that identifies, evaluates, selects, and imple-
ments a set of actions based on science and designed to
cost-effectively reduce risk while recognizing the social, cultural,
ethical, political, and legal contexts), and (3) risk communication
(facilitating interactive exchanges of information between stake-
holders and essential to effective management) (33). They also
recognized seven key elements or principles essential to compre-
hensive risk management programs dealing with human health,
ecological, and occupational risks (33). These seven key elements
include (1) problem formulation, (2) stakeholders (“interested
and affected parties”), (3) risk communication among stakehold-
ers, (4) quantitative risk assessment components, (5) iteration
and evaluation, (6) informed decision-making among stakehold-
ers, and (7) flexibility. Of these elements, the problem formula-
tion stage may be the most critical element because stakeholders
who fail to identify the right problem usually spend their time,
manpower, and capital solving it to arrive at wrong or unworkable
solutions. More importantly, they found a failure to incorporate
any of these elements will likely generate conflicts and problems
among stakeholders that make risk management more, not less
problematic (33). Jardine and her co-workers also noted these
seven principles should be supported by ten ethical principles in
risk management decision-making: (1) beneficence, (2) fairness,
(3) equity, (4) utility (adequate risk management), (5) honesty,
(6) caution when uncertain (“better safe than sorry”), (7) respect
of autonomy, (8) repeatability, (9) realization that risk cannot be
eliminated (“life is not risk free”), and (10) Golden Rule. Several of
the ethical principles supporting sound risk decision-making may

also serve as the underpinnings for informed consent (15, 25).
Although elements, principles, and practices may prove useful
with more traditional problems, nanotechnologies may raise new
challenges with their “known unknowns” (13, 14, 31, 34, 35).
Given some commentators already question our ability to identify
the risks or hazards associated with all nanomaterials (13–16, 34),
then it also seems stakeholders will likely target the wrong prob-
lem without a sound problem formulation stage. Our need for
discussions about risk management may be rising as reports of
potential toxicities for nanomaterials continually surface in the
academic literature and lay press (25, 31, 32, 35). Depending on
the particle and the animal model, investigators are identifying
particles that may affect our lungs (36), renal cells (37), and other
organ systems (10). Although most of these studies and reports
focus on animal models and cell cultures, a recent report from
China raises the specter for human involvement (38). Apparently,
a group of Chinese workers developed symptoms and pulmonary
findings similar to other pneumoconiosis such as asbestosis fol-
lowing their exposure to nanoparticles over several months (38).
Unlike previous reports in animal models, the materials incrimi-
nated in this report here were not carbon nanotubes (CNTs), but
nanoparticles. Because it represents a single report and likely has
other confounds, it may generate more questions than it answers.
It does, however, emphasize our need to know more about our
“known unknowns.”
Some commentators believe the inadequacies of our tradi-
tional risk management principles may further undermine our
public trust and confidence when it comes to managing the real
or imagined risks of nanomaterials (39–43). Marchant and his
colleagues consistently question whether our current risk man-
agement principles and programs are up to the task of managing
nanotechnology (39). They point out that identifying and quan-
tifying the health, safety, and environmental risks may prove dif-
ficult at best, because we lack both knowledge and experience
with its risk. They believe traditional risk management principles
such as (1) acceptable risk (utilizing risk assessment to identify
and reduce risks to acceptable levels), (2) cost–benefit analysis
(weighing the costs and benefits of proposed risk management
options), (3) best available technology (reducing risks to lowest
level technologically or economically possible), and (4) the pre-
cautionary principles (applying a “safe is better than sorry”
approach to control risk) are unworkable for nanotechnologies.
For example, the acceptable risk approach may not be applicable
because it favors risk reduction over benefit attainment, while the
cost–benefit analysis suffers from the uncertainties about the
“known unknowns” of nanotechnology. In the case of the best
available technology approach, both the risks and benefits of
nanotechnology may be simply ignored and this could lead to

policymakers to make arbitrary decisions. As for the precaution-
ary principle, it appears to be the least favored approach by several
commentators including Marchant for a variety of reasons includ-
ing its multiplicity of versions, low level of risk required to invoke
it, and ability to stop work and hinder progress (32, 39). More
importantly, all of these risk management principles suffer from
the lack of risk information that the public and its policymakers
require to effectively manage risk. To better manage the “known
unknowns,” some commentators and NGOs suggest we adopt
risk assessment and management schemes specifically designed
for nanomaterials (32, 39, 44–46). Currently, there exist various
attempts to construct risk management frameworks, some of
which we outline in the next section.
6. Exploring Risk
Management
Frameworks
One regulatory framework for nanotechnology developed by
Bowman and Hodge covers six regulatory frontiers by creating an
“enforcement pyramid” that will allow regulators to use a variety
of enforcement options (45, 46). Their “enforcement pyramid”
addresses major legal areas related to nanotechnology that include
intellectual property, privacy, product safety, occupational health
and safety, international law, and environmental law. They believe
such an approach will allow multiple stakeholders to utilize cur-
rent regulations to help manage nanomaterials and their risks.
Even so, some commentators question whether this framework
may be too static and likely inapplicable to nations with poorly
developed legal systems (38). One alternative to the enforcement
pyramid is the Nanorisk Framework which is also a nanotechnology-
based paradigm developed by Environment Defense (ED) and
DuPont to regulate a specific industry (44). This framework uses
a stepwise approach that focuses on risk identification, empha-
sizes safety through the life cycle of a given nanomaterial, fosters
transparency, and tracks success of risk management schemes.
Although it focuses on an industry, this scheme may be applicable
to nanomedicines, because it focuses on safety, stakeholder par-
ticipation, and monitoring that fosters trust.
Unlike these frameworks, the International Risk Governance
Council (IRGC) envisions a framework that provides both gover-
nance and risk governance of nanomaterials. Risk governance
incorporates the totality of circumstances, institutions, processes,
and stakeholders related to these activities that lead to risk-related
decisions or actions within the context of the risk situation (32).
Here, governance differs from the traditional concepts of legislative
control or government that mandates behavior through “command
and control measures.” Governance is generally preferable to

“command and control” government initiatives when dealing
with emerging technologies such as nanotechnology. The IRGC
framework acknowledges the inherent “riskiness” of existing and
future nanotechnologies and looks for a flexible process to lead
stakeholders toward prudent risk decisions. Renn and Rocco in
their paper discussing the IRGC framework on risk governance
see the uncertainty or ambiguity of risks of nanomaterials as
becoming greater or more problematic as nanostructures in later
generations increase in their sophistication. To help all stakehold-
ers cope with these eventualities, the IRGC proposes a risk man-
agement framework specifically designed for all nanotechnologies,
including those utilized in nanomedicine.
The IRGC framework categorizes risk-related knowledge as
(1) simple (clear causes and effects), (2) complex (multiple causal
events and effects), (3) uncertain (knowledge lacking), and (4)
ambiguous (variability of norms and interpretations), depending
on the generation and complexity and the generation of the nano-
structure (32). Perhaps, preassessment is the most critical phase
because problems are identified and framed into Frame 1 for pas-
sive, first-generation nanomaterials or Frame II for increasingly
complex nanomaterials in generations two through four. Once
risk-related knowledge is accrued then the risk process moves
from preassessment (risk assessment and risk concern) to risk tol-
erability assessment (risk characterization and evaluation) to risk
management and its risk decision-making and implementation
phases. The goal of this complex nanospecific framework is to
allow stakeholders to identify problems sooner rather than later
while addressing their potential social, legal, and policy impacts.
Each step of the process incorporates stakeholders and utilizes
communication and transparency which are the same key ele-
ments and principles identified by Jardine and her co-workers.
Although this framework is nanospecific, it is highly complex
unlike the framework suggested by Marchant and his co-workers
or other commentators (39).
Compared to IRGC risk management framework, Marchant
and his co-workers have crafted a less complex framework that
incorporates many of the aspects of the aforementioned frame-
works as well as concepts from Ayres and Braithwaite to build and
to create a dynamic regulatory pyramid (39). Rather than viewing
persuasion, soft law, self-regulation, and command and control
regulations as static events, they extend them through time. Their
approach also focuses on self-regulation and stakeholder gener-
ated norms and policies. Each stage of their framework looks to
involve stakeholders who are touched by nanotechnologies. Their
process begins with information gathering, assessment, and
information dissemination (immediate) stage followed by a
stakeholder-based self-regulation and norms (short term) stage
that moves to an enforced self-regulation (long term) stage. Each

level incorporates more supervision while also increasing trans-
parency and participation among stakeholders. The key to their
approach is its incrementalism that allows nanotechnology to
advance while keeping vigilant to risks concerns. It also allows
regulators to use a tit-for-tat regulatory strategy to get compli-
ance without unduly hindering progress. Such a framework may
allow all nanotechnologies to advance while allowing everyone to
reap its benefits.
Because most, if not all, of these risk management frameworks
and principles depend on our “known unknowns” about nano-
material risks, we will always have uncertainties and ambiguities
that will need defining. In fact, many “risky” theranostic nano-
medicine products reside within either the first (invasive and non-
invasive diagnostics and quantum dots) or second generation
(targeted drugs) of nanomaterials. They have “known unknowns”
of risk. Even with formalized risk governance frameworks, stake-
holders will continually be forced to weigh and balance the
“known unknowns” of nanomaterial risks against any real or
imagined benefit. As noted, our “known unknowns” may become
more uncertain and ambiguous as nanotechnologies become
more complex and diverse in their uses.
If Marchant and other commentators are correct about the
deficiencies in our existing risk management principles (39), then
anyone using them for risk calculus may be misled. The lack of
risk information on nanomaterials will likely impact IRB members
who must identify, access, and manage risks related to theranostic
cancer nanomedicines in clinical trials (15, 25). In fact, several
commentators have already questioned the approval process the
“nanoparticle albumin-bound” or nab-technology (Abraxis
Biosciences) anticancer agent Abraxane for market and clinical
use (47). Although many hail the approval process for this agent
as a triumph for abbreviated drug trials (48), others question the
validity of the process since the toxicology on the long-term
effects of the agent are lacking (49). Of the nearly 81,478 clinical
trials identified by the U.S. National Institute of Health (NIH)
on their web site only 1,023 or less than 2% of all trials specify
nanoparticles in their protocols and over 60% of them utilize the
drug Abraxane or nab-paclitaxel (50). The question remains
whether the use of risk–benefit data from “old” or preexisting
macromolecular counterparts is appropriate for nanomaterials
with novel properties (27, 51). Reality is there may be long-term
effects that are not accounted for by more traditional risk–benefit
models for macromolecular forms. More importantly, the compa-
nies and regulatory agencies responsible for assessing risk and
safety may not know the proper benchmarks or safety thresholds
for comparison with these novel nanomaterials (47).
If our traditional risk management principles and benchmarks

are inapplicable to nanoparticles because of their “known
unknowns,” then IRB members may have to rethink their use of
existing IRB risk management strategies such as the component
analysis (52) or the “net risks test” to evaluate the risks and ben-
efits of protocols with nanoparticles (53). Certainly, both meth-
ods require some knowledge of the risks and benefits of the
therapies and procedures (“interventions”) to make their com-
parisons meaningful. And although the former looks at clinical
equipoise to guide therapeutic research while the latter does
not, both must compare proposed research interventions with
existing therapies and procedures. If both assessment schemes
rely on knowing the risks and benefits of nanomaterials, then
both assessment schemes could be flawed by the “known
unknowns” and our lack of knowledge of toxicological studies
on their long-term effects (25, 47). This knowledge gap may be
compounded by IRB members who may choose to fill the gaps
by drawing on their preexisting knowledge and assumptions
about the macromolecular counterparts to evaluate risks, or
benefits, or both (15, 51). It may be the IRB equivalent of the
computer axiom that “garbage in equals garbage out,” where
calculations based on bad information simply produce more
misinformation. Not only may IRB members find their handling
risk-related matters for granting protocols problematic, but they
may also face challenges with crafting informed consent docu-
ments that convey risk information meaningfully to potential
candidates for clinical trials (15, 25). In this case, IRB members
may be unable to help investigators draft informed consent doc-
uments that reasonably convey the risks and benefits to partici-
pants or second (research team members) and third parties
(public).
Worse still, some allege that both clinical trial participants,
and ultimately patients, may be misled by researchers into partici-
pating in these protocols, because participants may be assuming
they are participating in a “bioequivalency trial” when they are
actually not. This situation may be further complicated by the
need for companies to keep their proprietary information on their
nanoparticle-based formulations confidential. In some cases, as in
the case of drugs such as Vioxx, companies and researchers may
not reveal information on safety concerns until it is too late (47, 54).
If this occurs, then everyone, including participants and patients,
may be receiving information that is either deficient or inaccurate
which could open the door to ethical and legal challenges to
informed consent (25, 47). If this lack of information becomes
the basis for a claim of deception, then everyone may find them-
selves at risk for a legal challenge to the process of informed con-
sent (20, 55).
7. Managing Our
“Known
Unknowns”:
Transparency and The current lack of an adequate framework for risk management
Communication requires careful communication on the part of the scientific and
medical community with various stakeholders to insure a suc-
cessful development of theranostics nanomedicine (56). The
economic stakes and the potential therapeutic benefits demand
a strategic plan that will allow the adequate assessment, manage-
ment, and communication of risks. In this chapter, we favored a
risk management framework that emphasizes self-regulation,
responsibility, transparency, and the involvement of all relevant
stakeholders. This bottom-up approach does not guarantee a
unanimous adherence to the aforementioned principles but it
allows a rapid flow of information between stakeholders, the
involvement of all relevant parties and the continuation of
research and development despite our “known unknowns.”
Until we get an agreed upon risk management framework, this
approach minimizes the potential of stigmatizing the technol-
ogy (32, 39).
That being said, the “known unknowns” related to nano-
technology may make it almost impossible for stakeholders to
avoid the public reaching premature opinions on its risks and
benefits (47). The propensity of the public to make risk–benefit
decisions based on mental shortcuts or “heuristics” makes nano-
technology vulnerable to interpretational errors (39). This means
the public may be exposed to an untoward event related to nano-
technology and form an opinion based on their initial perception
of the risks. Once formed, it may be difficult or impossible to
change. Reality is imagery and social reinforcement may give cer-
tain groups, such as NGOs or the media, real leverage in framing
public perceptions of nanotechnology. For these reasons, all
stakeholders must openly discuss the “known unknowns” related
to nanomaterials and engage the public to build public confidence
and trust before rather than after some adverse event taints the
debate.
Maybe the best strategy to avoid public uproar and misinfor-
mation is for all stakeholders to begin engaging in discussions
about the “known unknowns” before some adverse event occurs.
We need to encourage “upstream ethical and policy reflections”
at various level of discourse that includes the public, the scientific,
and medical community as well as various political and economic
stakeholders. Ultimately we need to develop risk management
frameworks that will help us identify the appropriate problems,
engage parties, foster discussions, and develop management
schemes to minimize risks and maximize our benefits.
References
1. Abel, J. “Nanomedicine and Cancer” Available 16. Bawa, R. and Johnson, S. (2007) The ethical
online: http://www.chem.usu.edu/~tapaskar/ Dimensions of nanomedicine. Med. Clin. N.
Joe-Seminar-Nano-Medicine.pdf. Ann. 91, 881–887.
2. European Science Foundation Nanomedicine 17. Wolinski, H. (2006) Nanoregulation. EMBO
Report (2005) Available online: http://www. Rep. 7, 858–861.
esf.org/publication/214/Nanomedicine.pdf. 18. Katz, J. (2009) Informed consent – Must it
3. Sumer, B. and Gao, J. (2008) Theranostic nano- remain a fairy tale? in: Ethical Issues in Modern
medicine for cancer. Nanomedicine 3, 137-140. Medicine (7th edition) (Steinbock, B. Arras, J.
4. Muthu, M. and Singh, S. (2009) Targeted D., and London, A. J., eds.) McGraw Hill,
nanomedicine: effective treatment modalities New York.
for cancer, AIDS and brain disorders. 19. Furrow, B. R., Greaney, T. L., Johnson, S. H.,
Nanomedicine 4, 105–118. Jost, T. S., and Schwartz, R. L. (2000) Health
5. Tassinari, O., Caiazzo R. J., Ehrlich J. R., and Law (2nd edition). Hornbook series, West
Liu, B.C-S. (2008) Identifying autoantigens Group, West Publishing Co.
as theranostic targets: antigen arrays and 20. Grim, D. A. (2007) Informed consent for all!
immunoproteonics approaches. Curr. Opin. No exceptions. New Mexico L. Rev.37, 39–83.
Mol. Ther. 10, 107–115. 21. Etchells, E. (1999) Informed consent in surgi-
6. Del Vecchio, S., Zannetti, A., Fonti, R., Pace, cal trials. World J. Surg. 23, 1215–1220.
L., and Salvatore, M. (2007) Nuclear imaging 22. Jotterand, F., McClintock, S. M., Alexander,
in cancer theranostics. Q. J. Nucl. Med. Mol. A. A., and Husain, M. M. (2010) Ethics and
Imag. 51, 152–163. informed consent of Vagus Nerve Stimulation
7. Ramachandran, G. (June 26, 2009) (VNS) for patients with Treatment-Resistant
Theranostics: an evolving field catering to the Depression (TRD). NeuroEthics 3, 13–22.
unmet needs of the medical world. http:// 23. Beauchamp, T. L. and Childress, J. F. (2009).
www.frost.com/prod/servlet/market-insight- Principles of Biomedical Ethics (6th edition).
print.pag?docid=170968864. Accessed 28 Oxford University Press, Oxford.
October 2009.
24. Brock, D. (1987). Informed consent. in: Health
8. Visogain. (May 29, 2009) In vitro diagnostics: Care Ethics (Van De Veer, D. and Regan T.,
market analysis 2009-2024. http://www. eds.) Temple University Press, Philadelphia.
prlog.org/10246394-in-vitro-diagnostics-
market-analysis-20092024.pdf. Accessed 28 25. Virdi, J. (2008) Bridging the knowledge gap:
October 2009. examining potential limits in nanomedicine.
Spon. Gen. 2, 25–44.
9. Blair, E. D. (2008) Assessing the value-adding
impact of diagnostic-type tests on drug devel- 26. Breggin, L., Falkner, R., Jaspers, N., Pendergrass,
opment and marketing. Mol. Diag. Ther. 12, J., and Porter, R. (2009) Securing the promise
331–337. of nanotechnologies. Chatham House, London,
England: Regulating Nanotechnologies in the
10. Jotterand, F. (2007) Nanomedicine: how it EU and US. http://www.lse.ac.uk/internation-
could reshape clinical practice. Nanomedicine alRelations/centresandunits/regulatingnano-
2, 401–405. technologies/nanopdfs/REPORT.pdf. Assessed
11. 3rd Horizon Scanning Workshop (2007) 29 October 2009.
Theranostics – Ethical, legal and social aspects. 27. Scientific Committee on Emerging and Newly
Muenster, 6th-8th June 2007. Identified Health Risks (SCENIHR) (2009)
12. Girod, J. and Klien, A. R. (2009) Predictive Risk Assessment of Products of Nanotechnologies.
health technologies: personalized medicine http://ec.europa.eu/health/ph_risk/commit-
and toxic exposures. Hous. J. Health L. Policy tees04_scenihr_o_023.pdf. Accessed 28 October
9, 163–178. 2009.
13. Wilson, R. F. (2006) Symposium article: 28. Nanotechnology: a report of the U.S. Food
Nanotechnology: the challenge of regulating and Drug Administration. Rockville, Maryland:
known unknowns. J. Law Med. Ethics 34, Nanotechnology Task Force (2007). http://
704–713. www.fda.gov/downloads/ScienceResearch/
14. Rakhlin, M. (2008) Regulating nanotechnology: SpecialTopics/Nanotechnology/ucm110856.
a private-public insurance solution. Duke Law pdf. Accessed 10 October 2009.
Tech. Rev. 2008, 2–45. 29. Pelley, J. and Saner, M. (2009) International
15. Resnick, B. R. and Tinkle, S. S. (2007) Ethical approaches to the regulatory governance of
issues in clinical trials involving nanomedicine. nanotechnology. Ottawa, Ontario: Regulatory
Contemp. Clin. Trials. 28, 433–441. Governance Initiative. http://www.carleton.ca/
regulation/publications/Nanotechnology_ 43. Kuzma, J., Paradise, J., Ramachandran, G.,

Regulation_Paper_April2009.pdf. Assessed 20 Kim, J.-A., Kokotovich, A., and Wolk, S. M.
September 2009. (2008) An integrated approach to oversight
30. Greenpeace: policy and position paper on assessment for emerging technologies. Risk
Nanotechnology (2002) http://www.green- Anal. 28, 1197–1219.
peace.org/raw/content/denmark/press/ 44. Nano Risk Framework, Washington D.C. and
apporter-or-dokumenter/nanotechnology-pol- Wilmington, Delaware: Environmental
icy-positi.pdf. Accessed on 1 October 2009. Defense – DuPont Nano Partnership (2007)
31. Tyshenko, M. G. and Krewski, D. (2008) A risk http://www.edf.org/documents/6496_
management framework for the regulation of Nano%20Risk%20Framework.pdf. Accessed
nanomaterials. Int. J. Nanotechnol. 5, 143–160. on 20 October 2009.
32. Renn, O. and Roco, M.C. (2006) 45. Bowman, D. M. and Hodge, G. A., (2007) A
Nanotechnology and the need for risk gover- small matter of regulation: an international
nance. J. Nanopart. Res. 8, 153–191. review of nanotechnology regulation.
33. Jardine, C. G., Hrudey, S. E., Shortreed, Columbia Sci. Tech. L. Rev. 8, 1–36.
J. H., Craig, L., Krewski, D., Furgal, C., et al. 46. Bowman, D. M. and Hodge, G. A. (2006)
(2003) Risk management frameworks for Nanotechnology: mapping the wild regulatory
human health and environmental risks. J. frontier. Futures 38, 1060–1073.
Toxicol. Environ. Health, Part B 6, 570–718. 47. Vines, T. and Faunce, T. (2009) Assessing the
34. Maynard, A. D. (2006) Nanotechnology: safety and cost-effectiveness of early nanod-
assessing the risks. Nanotoday 1, 22–32. rugs. J. Law Med. 16, 822–45.
35. Patra, D., Ejnavarzala, H. and Basu, P. K. 48. Till, M. C., Simkin, M. M. and Maebius, S. (2005)
(2009) Nanoscience and nanotechnology: Nanotech meets the FDA: a success story about
ethical, legal, social and environmental issues. the nanoparticle drugs approved by the FDA.
Curr. Sci. 96, 651–657. Nanotechnology Law & Bus. 2.2, 163–167.
36. Orberdorster, G., Oberdorster, E., and 49. Faunce, T. A. (2008) Toxicological and public
Oberdorster, J. (2005) Nanotechnology: an good considerations for the regulation of
emerging discipline evolving from studies of nanomaterial-containing medical products.
ultrafine particles. Environ. Health Perspect. Expert Opin. Drug Saf. 7, 103–106.
113, 623–839. 50. U.S. National Institutes of Health, Nanoparticle
37. Azou, B. L., Jorly, J., On, D., Sellier, E., Studies. [Clinical Trials.gov] http://www.clini-
Moisan, F., Fleury-Feith, J., et al. (2008) In caltrails.gov/ct2/results?term=nano. Accessed
vitro effects of nanoparticles on renal cells. on Nov. 10, 2009.
Part. Fibre Tox. 5, 22. 51. Beasley, J. C., Kramer, V. L., and Priest, S. H.
38. Song, Y., Li, X. and Du, X. (2009) Exposure (2008) Expert opinion on nanotechnology:
to nanoparticles is related to pleural effusion, risks, benefits, and regulation. J. Nanopart.
pulmonary fibrosis and granuloma. Eur. Respir. Res. 10, 549-558.
J. 34, 559–567. 52. Weijer, C. and Miller, P. B. (2004) When are
39. Marchant, G. E., Sylvester, D. J., and Abbott, research risks reasonable in relation to antici-
K. W. (2008) Risk management principles for pated benefits? Nat. Med. 10, 570–573.
nanotechnology. Nanoethics 2, 43–60. 53. Wendler, D. and Miller, F. G. (2004) Assessing
40. Linkov, I., Steevens, J., Adlakha-Hutcheon, research risks systematically: the net risks test.
G., Bennett, E., Chappell, M., Colvin, V., J. Med. Ethics 33, 481-486.
et al. (2009) Emerging methods and tools for 54. Avom, J. (2006) Dangerous deception – hid-
environment risk assessment, decision-making, ing the evidence of adverse drug effects. NEJM
and policy for nanomaterials: summary of 355, 2170–2171.
NATO Advance Research Workshop. J. 55. King, S. K. and Moulton, B. W. (2006)
Nanopart. Res. 11, 513–527. Rethinking informed consent: the case for
41. Linkov, I., Satterstrom, F. K., Steevens, J., shared medical decision making. Am. J. Law
Ferguson, E., and Pleus, R. C. (2007) Multi- Med. 32, 429–493.
criteria analysis environmental risk assessment for 56. Plows, A. and Reinsborough, M. (2008)
nanomaterials. J. Nanopart. Res. 9, 543–554. Nanobiotechnology and Ethics: Converging
42. Farber, D. and Lakhtakia, A. (2009) Scenario Civil Society Discourses. in: Emerging Conceptual,
planning and nanotechnologies futures. Eur. Ethical and Policy Issues in Bionanotechnology
J. Phys. 30, S3-S15. (Jotterand, F., ed.) Dordrecht, Springer.
Index
A C
Actuation................................... 63, 124, 127, 132–135, 166 CALAA–01........................................................... 332–336
Affinity interaction Cancer
biotin-streptavidin.................................................... 5–6 diagnostic.......................................................... 400, 414
metal........................................................................... 97 therapy.............................7, 64, 153, 158, 172, 227, 333,
Angel investor........................................................ 388, 389 396, 409, 413–415, 418–419
Animal model Cantilever................................................119–120, 125–136
hamster...................................... 262, 266, 267, 276, 277 Capsid..................................................7, 207–211, 213–216
mouse..................................................52, 224, 329, 335 Carbon nanotube
rat................................................ 263, 273, 319, 328, 335 multi-walled..............................................223, 301, 395
Antibiotic.........................................................53, 187–189, single-walled......................................223, 226, 300, 395
191, 204, 251–252 Cell
Antibody................... 4–6, 21, 36, 39–41, 43, 44, 55, 60, 64, behavior............................................................ 247, 261
66, 67, 69, 71–73, 87, 89, 90, 96, 131–132, capture........................................................ 73, 141–149
134–136, 142–144, 148, 159–160, 172, 187–205, culture........................................... 38, 53, 55, 56, 68–70,
216, 226, 263, 269, 271–272, 274, 277 72–74, 143, 144, 165, 167, 219, 249, 251,
Assay 252, 254, 255, 263, 264, 269–270, 272, 273,
alkaline phosphatase (AlkP)......................315, 318, 320 299–311, 328, 329, 422
lactate dehydrogenase (LDH)......................... 300–302, death....................................................67, 152, 170, 188
304–311, 315, 320 lysis............................................ 67–70, 72–73, 304, 308
MTT..................165, 219, 300–301, 303–305, 309–311 neural progenitor...................................................... 271
Assembly Schwann........................................................... 263, 273
bottom-up........................................................ 121–124 staining..................................... 144–149, 165, 247, 255,
self.................................. 5, 8, 34, 97–100, 102–106, 108, 257, 278, 317, 322
132, 214, 244, 245, 261, 262, 269, 327, 334 stem...................................................271, 398, 407, 409
top-down.......................................................... 120–121 viability................................. 72–73, 219, 247, 248, 255,
271, 302–305, 309, 310, 317
B
Central nervous system (CNS) regeneration.......... 259, 277
Bacteria..................................... 1–2, 90, 130, 135–136, 165, Charge-transfer complex............................................ 63, 71
167, 187–205, 405 Chemical vapor deposition (CVD)........................... 3, 112,
Bacteriophage......................................4, 187–205, 214–215 120, 124, 230
BAL. See Bronchoalveolar lavage Circulating tumor cells (CTC)............................... 141–149
Biobarcode.................................................................. 17–30 Click chemistry...............................................37, 40–41, 45
Biodetection........................................................... 6–7, 131 Commercialization................................. 331, 341, 344, 345,
Biomarker............................................ 33–47, 398, 399, 414 366, 370, 371, 380–383, 385, 391, 406
Biomedical nanotechnology...................................1–10, 95, Conjugation...............................4, 21, 28, 42, 45–46, 55, 61,
340–341, 348 69, 95–108, 189, 191–193, 195–198, 202, 204, 211,
Biophysics..............................................................20, 33, 51 216–218, 220, 231
Biosensing.............................................. 36, 77–92, 96, 111, Coupling
119–137, 299, 381 EDC-NHS........................................................... 81, 82
Bronchoalveolar lavage (BAL)................315–318, 320, 323 sulfo-SMCC............................ 37, 39–40, 211, 217, 218
DOI 10.1007/978-1-61779-052-2, © Springer Science+Business Media, LLC 2011
431
432 Index

Crosslinking............................... 4–5, 38–39, 41, 43, 45–47, I

78, 84–85, 96, 101, 162, 163, 166, 170, 171, 192,
233–234, 237 Imaging
CTC. See Circulating tumor cells confocal.................................52–54, 68, 70, 73, 74, 171,
CVD. See Chemical vapor deposition 181, 255, 271, 274
magnetic resonance (MRI)..................... 34, 38, 45, 162,
D 163, 166, 167, 172, 226, 228, 407, 414
micro CT.......................................................... 234–235
1D. See One-dimensional
Immobilization............................... 34, 55, 77–92, 131–134,
Dendrimer..........................................................79, 80, 170,
136, 142, 144–145
327, 369, 371–372
Immunochemistry...................................143–144, 146–148
Diagnostic magnetic resonance (DMR)..................... 33–47
Implant....................................... 7, 120, 168, 180–182, 228,
Dielectrophoresis technique........................................... 116
244, 245, 259–261, 273, 277–278, 407
DMR. See Diagnostic magnetic resonance
Informed consent............................................329, 413–427
DNA...............................................................4, 5, 7, 18–20,
Infusion.............................................. 4, 210, 211, 215–216,
22–29, 34, 36, 64–66, 74, 95–108, 131,
218, 318, 335
132, 134, 135, 165, 210, 212, 214, 215,
Innovation.............................................. 181, 341, 351, 353,
224, 226, 227, 283–295, 333, 394, 397,
361–363, 366–369, 371–374, 394, 397, 398,
399, 401–402, 409
400–402, 406, 407, 409, 410
Drug delivery.............................77, 154, 156, 158, 159, 161,
Institutional review board (IRB)............................144, 147,
167, 173, 179, 180, 188, 192, 200, 207–220, 226,
329, 418, 425, 426
299, 396, 398, 399, 414
Intellectual property (IP)............................ 9, 351, 352, 361,
E 363, 369–370, 373, 380–382, 423
Interaction
ECM. See Extracellular matrix charge-charge....................................284, 285, 289, 295
Electrospinning.............................. 3, 8, 121–123, 244–250, nanoparticle-membrane............... 4, 5, 66, 164, 283–295
252, 256–257 International risk governance council (IRGC)....... 423, 424
Encapsulation............................... 4, 77, 158–159, 164, 166, Intratracheal instillation......................................... 314–318
167, 169–172, 213, 259, 399 Investigational New Drug (IND) Application.............. 326,
Entrepreneurship.....................................341, 379–391, 400 329–330
Ethics..................................9, 341–342, 344, 346, 348, 356, In vivo.................................... 7, 8, 17, 20, 38, 51, 52, 64, 95,
364, 371–375, 415–422, 426, 427 152, 156, 164, 166–170, 172, 173, 189, 200–201,
Extracellular matrix (ECM)............................244, 245, 261 209–210, 224, 227–229, 231, 244, 247, 327, 328,
335, 404
F IP. See Intellectual property
Fabrication IRB. See Institutional review board
micro.........................................................120–121, 124 IRGC. See International risk governance council
nano...................................................120–121, 179–180
FDA approval
L
clinical phase.............................................326, 330–331 Lab-on-a-chip................................. 393, 396, 399–401, 409
post-marketing phase................................326, 331–332 Legislation........................................... 9, 339–356, 365, 423
pre-clinical phase.............................................. 326–329 Lipid membranes................................................... 283–295
Liscensing................................................371–373, 375, 381
G Lithography
Genetically modified organisms (GMO)...............342, 403, electron beam (e-beam).............................120–123, 250
406, 416 photo-....................................................................... 120
GMO. See Genetically modified organisms
M
H
Magnetic relaxation switching (MRSw).................... 34–36,
High-pressure CO (HiPco).....................225, 230, 232, 236 38, 42–45, 47
HiPco. See High-pressure CO Marketing mix........................................................ 393–410
Histopathology...................................................9, 166, 314, MD simulation. See Molecular dynamics simulation
316, 318–322 Mechanical properties............................ 224, 228, 229, 236,
Hydrogel..........................................................166, 179–184 244, 245, 247, 260, 299
433
Index

MEMS. See Microelectromechanical system NEMS. See Nanoelectromechanical system
Mercury intrusion porosimetry....................................... 235 NGO. See Non-governmental organization
Microelectromechanical system (MEMS)............. 119–120, NMR. See Nuclear magnetic resonance
124–126 NNI. See National Nanotechnology Initiative
Microfluidic device............................................45, 133, 142 Non-governmental organization (NGO)...............342, 344,
Microparticle 345, 371, 416–417, 423, 427
magnetic (MMP).................... 20–22, 24–26, 28, 29, 36 Nuclear magnetic resonance (NMR)....................34, 35, 38,
polymeric.......................................................... 179–184 43–46, 195
Microscopy
atomic force (AFM)......................... 2, 66, 72, 119, 120, O
136, 171, 365 Oligonucleotide.............................4, 5, 7, 17–30, 38, 39, 43,
electron................... 2, 3, 70, 73, 113, 120, 143, 146, 234 78, 99, 103, 107, 210, 214, 215, 285
fluorescence....................46, 54, 147, 148, 184, 263, 269 One-dimensional (1D)............................... 4, 111, 123, 127,
X-ray fluorescence (XFM).......................................... 66 129, 224, 226, 228
Microsphere........................................................ 77–91, 179 Optical penetration depth...................................... 153–156
Molecular dynamics (MD) simulation................... 220, 285
MRI. See Imaging P
MRSw. See Magnetic relaxation switching
Multi-component............................................4, 5, 8, 9, 415 Patent
Multi-functional..................................... 5, 9, 158, 161, 162, pools..........................................................366, 373, 375
166, 333, 413–414 thickets......................................................366, 370–374
Multiplexing......................................................... 17–30, 96 PDT. See Photodynamic therapy
Multi-valent....................................... 47, 55, 60, 61, 63–64, Phage display................................... 188, 189, 192, 200–202
82, 85, 87–88, 95–108 Photocatalysis............................................................. 67, 69
Photodynamic therapy (PDT).......................... 64, 151–173
N Photosensitizer.............................. 8, 64, 151–159, 161–172
PNA.....................................................................................74
Nanocomposite...................... 8, 64, 224, 229, 233, 235–237 Policy................................ 340–345, 347, 351, 353–354, 374,
Nanodevices........................................................4, 413, 419 382, 407–409, 415–417, 420, 423, 424, 427
Nanoelectromechanical system (NEMS)........... 3, 119–137 Polymer.....................................7, 8, 18, 39, 77–92, 96, 122,
Nanoelectronics.............................................................. 223 123, 159–163, 167, 169–172, 179–184,
Nanomaterial...............................2–9, 18, 19, 64, 72, 95–97, 219, 228–234, 236, 237, 244–251,
123, 207–220, 224, 228, 229, 284, 300, 328, 254–257, 260, 300, 398
339–356, 369, 371–372, 394, 397, 401, 402, Prodrug....................................................156, 157, 187–205
404–406, 408, 415, 419, 420, 422–427 Protein.......................2, 17, 34, 52, 65, 78, 96, 129, 154, 183,
Nanomedicine............. 63–64, 187–205, 370, 401, 413–427 188, 207, 224, 244, 263, 284, 300, 315, 327, 396
Nanometer.......................................... 66, 78, 120, 121, 127, 5 Ps......................................................................... 393–410
188, 245, 397–398 Pulmonary
Nanoparticle bioassay..................................................................... 313
core-shell.................................................................... 39 hazards.......................................................... 9, 313–323
gold........................ 17–30, 214, 284–286, 288, 290–293 toxicity.............................................................. 313, 314
iron oxide............................... 38–39, 161–162, 172, 327
photon-upconverting................................................ 172 Q
plant virus (PVN)..........................7, 207–212, 215–219
titanium dioxide.........................................7, 63–74, 395 Q-factor............................125, 129–131, 133, 134, 136, 137
Nanoscale..................................1, 2, 5, 7, 64–66, 73–74, 95, Quantum dot........................ 5, 6, 51–61, 95–108, 161–162,
111, 134, 142, 181–182, 245, 314, 341, 342, 353, 213–214, 327, 369, 371–372, 394, 425
355, 356, 368, 371, 380, 394, 397, 398, 404,
R
406–409, 413, 414, 416, 419, 420
Nanostructure........................ 4, 57, 123, 124, 339–340, 424 RADA4, 261–264, 266, 267, 269–273, 276
Nanotechnology............................... 1, 19, 63, 95, 217, 259, RCNMV. See Red clover necrotic mosaic virus
328, 339, 359, 393, 413 Reactive oxygen species (ROS)...............................7, 64, 67,
Nanowire..................................... 6, 111–116, 121, 122, 124 72–73, 152, 153, 168–169
National Nanotechnology Initiative (NNI)........... 339–340, Red clover necrotic mosaic virus
344–346, 365, 397 (RCNMV)..........................................207–218, 220
434 Index

Rheometry...................................................................... 236 T
Risk management............................................415, 421–427
RNA..........................................7, 18, 34, 36, 207, 209, 210, Targeting..........5, 18, 33, 52, 64, 77, 96, 112, 128, 151, 179,
212–215, 227, 283, 328–329, 332–336 187, 209, 227, 246, 259, 283, 314, 327, 380, 393
ROS. See Reactive oxygen species Technology transfer................................................ 382, 383
Templation......................................................................... 3
S Theranostics.....................394, 398, 399, 404, 413–420, 427
Tissue
SBCG. See Shape-based coarse-grained model
engineering........................................228–229, 243–257
Scaffold
soft.............................................................243–257, 260
biodegradable................................. 8, 229, 244, 259–278
Toxicology...............................8, 9, 313–323, 329, 342, 351,
nanofiber...................................... 8, 244–247, 249, 252,
416, 419, 425
253, 255, 261, 262
Trade-related intellectual property rights (TRIPS)
porous................................... 8, 224, 228–229, 232–235,
agreement................................................... 363–364
244–245, 252, 259–278
Trafficking
Scanometric detection.....................................23, 26, 27, 29
protein.................................................................. 60, 61
Semiconductor.................................. 51, 64, 65, 67, 95–108,
single particle.............................................................. 58
111, 394, 398–400, 408
Trauma........................................................................... 259
SERS. See Surface enhanced Raman spectroscopy
TRIPS. See Trade-related intellectual property rights
Shape-based coarse-grained (SBCG)
agreement
model...................................................284–289, 295
Silicon nanopillar (SiNP) array...................................... 142 U
SiNP. See Silicon nanopillar array
SIVA. See Solution, information, value, and access United States food and drug administration (FDA)...... 8, 9,
SIVAC.................................................................... 395, 396 170, 173, 325–336, 407, 420
Socio-technical integration research (STIR) United States 21st Century Nanotechnology Research and
project.................................. 346–348, 350, 353, 354 Development Act (NRDA)....................... 339–341,
Sol fraction..................................................................... 234 343–346, 348, 349, 355–356
Solution, information, value, and access (SIVA)............. 395
V
Start-up company.............332, 341, 353, 354, 382–384, 390
STIR. See Socio-technical integration research project Venture capital.........................................341, 352, 389, 400
Surface enhanced Raman spectroscopy (SERS)............. 227 Virus........................................1–2, 7, 36, 67, 125, 134–135,
Surgery.................................... 267, 270, 276, 400, 402, 409 162, 188, 203, 207–220, 394

Biomedical Nanotechnology 2011

Uploaded by

Copyright:

Available Formats

Biomedical Nanotechnology 2011

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Biomedical Nanotechnology 2011

Uploaded by

Copyright:

Available Formats

Methods in Molecular Biology™

For other titles published in this series, go to

Methods and Protocols

ISSN 1064-3745 e-ISSN 1940-6029

Library of Congress Control Number: 2011922747

© Springer Science+Business Media, LLC 2011

Printed on acid-free paper

Humana Press is part of Springer Science+Business Media (www.springer.com)

Argonne, IL Sarah J. Hurst

Part I Using Nanomaterials in Sensing, Imaging, and Therapeutics

13 Antibacterial Application of Engineered Bacteriophage

Part II Translating Nanoscience and Technology

19 Cytotoxic Assessment of Carbon Nanotube Interaction

Fabrice Jotterand • Division of Ethics and Health Policy, Department of Clinical

Tijana Rajh • Center for Nanoscale Materials, Argonne National Laboratory,

Key words: Nanoparticles, Nanodevices, Biomedical nanotechnology, Biodetection, Nanothera-

Nanoscience and technology is a field that focuses on developing

Fig. 1. Length scales.

diameter (e.g., cells, viruses, and bacteria), and as a consequence

(20–23), and branched (24–26) structures (see Fig. 2). Techniques

namely, sensing, imaging, and therapeutics. Finally, it gives a

nanomaterial is a particle with an iron-oxide core and a cross-linked

3.1. Sensing Most of the sensing and imaging applications of nanomaterials in

highly specific binding interaction and the quantum dot label is

detection of conventional techniques [e.g., enzyme-linked

Three examples of nanostructured scaffolds are presented in this

Fig. 4. General schematic of a multicomponent nanoparticle showing targeting moieties,

the components of the degradation of these complex materials

The continued growth of the field of nanoscience and

Using Nanomaterials in Sensing, Imaging, and Therapeutics

Multiplexed Detection of Oligonucleotides

Key words: Oligonucleotide, Multiplexing, Gold nanoparticle, Detection, Sensing, Biomolecule,

The dramatic development of genomics and proteomics in the

is growing significantly since the progress of the current research is

1.1. Oligonucleotide An oligonucleotide is a short nucleic acid, or polymer of nucle-

the duplex pair, can be controlled by changing the salt concentration

1.2. Gold Colloidal gold nanoparticles (AuNPs) have been investigated by

1.3. General Properties An oligonucleotide–gold nanoparticle conjugate (Oligo-AuNP) is

chemical recognition abilities of these structures are derived from

1.4. Biobarcoded Biobarcode assays are characterized by their ultrahigh sensitivity

and the AuNPs generates hundreds of biobarcode strands, lead-

2.1. Gold Nanoparticle 1. Gold (III) chloride trihydrate.

2.2. Oligo-AuNP 1. Thiol-oligonucleotide of a given sequence (HPLC-purified)

2.3. MMP Preparation 1. Amino-functionalized MMPs (2.8 mm in diameter; Invitrogen,

5. Verigene Reader System (Nanosphere, Northbrook, IL).

2.5. Scanometric 1. N-Hydroxysuccinimide (NHS) ester-activated Codelink

3.2. Oligo-AuNP 1. Dissolve the lyophilized thiol-modified oligonucleotides

10. Place the oligo-functionalized MMPs on a magnet. Remove

3.4. Multiplexed 1. Prepare Oligo-AuNPs by conjugating oligonucleotides with

10. Collect the released barcode oligonucleotides from the mixture

3.5. Scanometric 1. Dissolve amine-terminated capture DNA sequence (5¢ H2N-

1. Four exemplary sequences (Au1–Au4) for the AuNP probes

Fig. 2. Scanometric detection of the biobarcode-oligonucleotides released from the

is protected in the oxidized form (disulfide), which needs to

NaCl concentration is 0.2 M. Four thiolated sequences

20. The hybridization temperature is determined empirically not

Molecular Detection of Biomarkers and Cells Using

Key words: Magnetic nanoparticles, Molecular biosensing, Diagnostic magnetic resonance,

Magnetic nanoparticles (MNP) offer unique advantages for

In addition, biological samples exhibit virtually no magnetic

Argonne, IL Sarah J. Hurst

Part II Translating Nanoscience and Technology

1.1. Oligonucleotide An oligonucleotide is a short nucleic acid, or polymer of nucle-

2.3. MMP Preparation 1. Amino-functionalized MMPs (2.8 mm in diameter; Invitrogen,

egatively affect the reaction. Incubate the reaction mixture

2.1. Reagents 1. Biotinylated small molecule or antibody against an extracel-