OF CELL MEMBRANE DISRUPTION

BY ACOUSTIC CAVITATION

A Thesis
Presented to
The Academic Faculty

by

Stephen Andrew Cochran

In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Bioengineering

Georgia Institute of Technology

November 2000
SONOLUMINESCENCE AS AN INDICATOR

OF CELL MEMBRANE DISRUPTION

BY ACOUSTIC CAVITATION

Approved: I

- » \i

Mark }$,. Prausnitz, Chairman

* s.

William D. Hunt

c^.>c^ - ••— fm-S 1

Feng B. Dong

Date Approved \\po/00

 DEDICATION

To the creator, God of the universe, personal, and unchanging,

to my parents and sisters, and

to my wife.
 Ill

ACKNOWLEDGEMENTS

There are many whom I would like to acknowledge and so I will inevitably leave

someone off of this list but let me begin by thanking Hector Guzman, Thomas Lewis,

Robyn Schlicher, Esi Ghartey-Tagoe, Paul Canatella and Steve Woodard for technical

assistance and helpful discussions. I also thank many of the others in lab who helped to

make my stay there an enjoyable one. I would like to especially thank my thesis advisor,

Dr. Mark Prausnitz, who found funding for me and brought me into his lab. He was

always available for quick questions and strongly encouraged me to think independently.

In times when I was confused about which direction to take, he was there to guide me in

which experiments to conduct or which analysis method was better. I really appreciate

all that he did for me. The time that I spent here enriched me not only professionally but

personally as well.

I also thank the following organizations for their support of this work: the

National Science Foundation, the Whitaker Foundation and the National Institutes of

Health (through a sub-contract from EKOS Corporation).

 iv

TABLE OF CONTENTS

DEDICATION ii
ACKNOWLEDGEMENTS Hi
LISTS OF TABLES vi
LISTS OF FIGURES vii
SUMMARY viii

CHAPTER I - INTRODUCTION 1

1.1 Introduction 1
1.2 Objectives 3
1.3 Overview of the Thesis 3

CHAPTER II - BACKGROUND AND THEORY 4

2.1 Introduction and Overview 4

2.2 Acoustic Waves 4
2.3 Cavitation and Sonoluminescence 5
2.4 Drug Delivery using Acoustic Cavitation 7
2.5 Relevance of Work 8

CHAPTER III - MATERIALS AND METHODS 9

3.1 Introduction and Overview 9

3.2 Ultrasound apparatus 11
3.3 Verification of ultrasound apparatus 12
3.4 Preparing Cell Samples 12
3.5 Flow Cytometry Assay 15
3.6 Data Analysis 16
3.7 Analysis of Three Populations 17
3.8 Sonoluminescence Analysis 20

CHAPTER IV - RESULTS AND DISCUSSION 22

4.1 Introduction 22
4.2 Heterogeneous molecular uptake 22
 V

4.3 Effects of ultrasound parameters on bioeffects 24

4.3.1 Effects of intensity on bioeffects 24
4.3.2 Effects of total exposure time on bioeffects 24
4.3.3 Effects of pulse length on bioeffects 28
4.4 Correlation of sonoluminescence with bioeffects 28
4.5 Discussion 33

CHAPTER V - CONCLUSIONS AND RECOMMENDATIONS 36

5.1 Conclusions 36
5.2 Recommendations 37

APPENDICES 39

Appendix A - Figure 6 graphs shown in greater detail 39

Appendix B - Figure 6 graphs shown in greater detail 43

REFERENCES 49
 vi

LISTS OF TABLES

Table 1. Ultrasound transducer specifications 9

 VI1

LISTS OF FIGURES

Figure 1. Diagram of apparatus 9

Figure 2. Results of apparatus validation experiments 13

Figure 3. Chemical structure of the calcein molecule 15

Figure 4. Series of representative histograms 18

Figure 5. Typical light intensity waveforms 21

Figure 6. Cell viability and distribution of cells 25

Figure 7. Data from Figure 6 are re-plotted 26

Figure 8. Uptake and cell viability shown as functions of sonoluminescence... 30

Figure 9. Cell viability correlations with other sonoluminescent parameters... 32

 Vlll

SUMMARY

Ultrasound has been shown to transiently disrupt cell membranes and thereby

load drugs and genes into viable cells. In this study, we quantified the number of calcein

molecules delivered and the loss of viability in prostate cancer cells exposed to 24 kHz

ultrasound over a range of different pulse lengths (1-100 ms), total exposure times (0.1-

10 s) and pressures (1.0-9.8 atm). Bioeffects were observed to increase with pulse length,

total exposure time and pressure. Using this large data set, we established a correlation

between bioeffects and the amount of light produced by sonoluminescence. These results

support a cavitation-based mechanism for bioeffects and suggest a means to control

ultrasound's effects on cells using sonoluminescence-based feedback.

 1

CHAPTER I

INTRODUCTION

1.1 Introduction

One of the critical elements of drug therapy is effective and targeted delivery of

drugs into cells and tissue (Park 1997; Langer 1998). The lipid bilayer cell membrane

poses the primary barrier to transport of molecules into cells (e.g., gene therapy (Rolland

1998)), into tissues (e.g., targeted chemotherapy of tumors (Mir and Orlowski 1999)) and

across tissues (e.g., transdermal drug delivery (Bronaugh and Maibach 1999)). Some

approaches to increase drug transport have used chemical methods, whereby

physicochemical properties of drug molecules are modified to promote their diffusion

across membranes or to facilitate binding to receptors expressed, for example,

preferentially on cancer cells (Silverman 1992). Another approach to enhancing and

targeting drug delivery involves using ultrasound to transiently disrupt cell membranes

(Fechheimer et al. 1987; Mitragotri et al. 1995; Liu et al. 1998). The advantage of this

method is that as a physical, rather than chemical, approach, the enhancement is likely to

be broadly applicable (i.e., can be applied to many cell types and many drugs, without

modifying drug structure) and can be readily targeted, based on the established ability of

ultrasound to be focused almost anywhere in the body (Kremkau 1998).

 A number of studies have shown that ultrasound can disrupt cell membranes in a

reversible manner and thereby load viable cells with molecules of interest. These studies

have demonstrated delivery of small molecules, such as drugs (Saad et al. 1992;

Mitragotri et al. 1995), as well as DNA for transfection (Fechheimer et al. 1987; Bao et

al. 1997; Wyber et al. 1997). This "proof-of-principle" that ultrasound can be used for

drug and gene delivery motivates our study, in which we strive to better elucidate

ultrasound's mechanism of action and develop methods to predict and control

ultrasound's effects on cells.

In this study, we sought to bridge the gap between simply seeing the bioeffects

and being able to control them. Our approach to controlling ultrasound's biological

effects is based on achieving feedback in the form of real-time physical measurements.

Using a similar approach, Liu et al. (1998) correlated observed bioeffects (red blood cell

disruption) with measured features of the acoustic frequency spectrum known to be

associated with cavitation, i.e., subharmonic frequency (f/2). This correlation could be

used to determine the effects a given exposure of ultrasound had on cells based only on a

relatively simple measurement of the acoustic spectrum. Using a related approach,

Wyber et al. (1997) correlated ultrasound's bioeffects (cell viability) with an iodine-based

measurement of cavitation. Although assay of this technique is more time intensive and

unlikely to be used in vivo, it further validates the idea of using physical measures of

cavitation as correlates for ultrasound's bioeffects. In this study, we seek to correlate

observed bioeffects (molecular uptake and cell viability) with another physical

phenomenon associated with cavitation: sonoluminescence.

 3

1.2 Objectives

The objectives of this study were threefold:

1) study the effects of different ultrasound pressure, total exposure time,

and pulse length on molecular uptake and cell viability,

2) establish a relationship between ultrasound's effect on cell viability and

molecular uptake with a known acoustic cavitation phenomenon,

sonoluminescence and

3) identify regions of high molecular uptake and high viability.

1.3 Overview of the Thesis

This chapter provides an introduction to the thesis. Chapter 2 provides the

background and theories behind acoustic cavitation as well as the relevance of this study.

In chapter 3, the materials and methods that were used in this study are described in

detail. Results and discussion then follow in chapter 4. Lastly, in chapter 5, conclusions

and recommendations for future studies are provided.

 CHAPTER II

BACKGROUND AND THEORY

2.1 Introduction and overview

The cell membrane poses a difficult barrier across which to deliver drugs into

cells or tissues for targeted drug therapy. There is a need for new and innovative

technology that is able to transiently disrupt the cell membrane in a controllable fashion.

Ultrasound-induced cavitation has been shown to disrupt cell membranes and provides an

attractive solution if its effects can be controlled. It is believed that there exist methods

to control these effects by monitoring physical measures of acoustic cavitation. In this

study, we seek to find a correlation between sonoluminescence, which is closely

associated with acoustic cavitation, and molecular uptake and cell viability. In order to

build a foundation for this study, some background must be given on acoustics, acoustic

cavitation and sonoluminescence.

2.2 Acoustic Waves

Acoustic waves are mechanical energy waves that create changes in the density of

the medium as they travel through it. These waves are considered longitudinal waves

because particles are displaced parallel or anti-parallel to the plane of motion. Over the

cycle of an acoustic wave there exists a compression stage, where the particles are
 5

pressed together, and a rarefaction stage, where the particles are spread apart. A more

detailed explanation of the physical phenomenon that occurs with the passage of an

acoustic wave can be found, for example, in Pierce (1989).

Acoustic waves can be classified according to the frequency at which these

compressions and rarefactions occur. In this study, the waves oscillate at a frequency of

24 kHz and are therefore considered to be ultrasonic, i.e. above the normal threshold of

human hearing, 20 kHz (Pierce 1989). There are other ways to describe acoustic waves

and some of the terms used to characterize them are: frequency, pulse length, incident

pressure, total exposure time, and duty cycle.

Terms Defined

• Frequency - number of cycles in a given time period.

• Pulse length - duration of one pulse of acoustic energy.

• Total exposure time - the pulse length multiplied by the total number of pulses

during an exposure.

• Incident pressure - the peak pressure of an acoustic wave in the absence of

acoustic cavitation.

• Duty cycle - the ratio of the time that the transducer generates acoustic waves to

the total duration of the exposure.

2.3 Cavitation and Sonoluminescence

Cavitation involves the creation of gas bubbles in a liquid medium and is a

phenomenon that occurs in many places (Munson et al., 1994). For example, people who

design propellers for boats have to understand how the propeller will cut through the
 6

water and if it will induce regions of cavitation. In other applications, people who design

artificial heart valves have to consider the pressures their valve generates in blood flow

and if it will induce cavitation as well. The term, cavitation, has been used to describe

the creation as well as the dynamics of bubbles. Cavitation occurs when liquid that was

at high pressure enters a region of low pressure, resulting in the "tearing" apart of the

liquid and formation of a cavity or bubble. When this bubble reenters a region of high

pressure, it can result in the violent collapse of the bubble. This violent phenomenon can

lead to pitting of metal propellers and heart valves.

Cavitation that is acoustically driven has been named acoustic cavitation.

Acoustic cavitation harnesses the mechanical energy inherent in acoustic waves to

generate, grow, and subsequently collapse bubbles in liquid media. The size of the

bubbles generated in liquid media is inversely related to the driving frequency. One of

the interesting aspects of acoustic cavitation is that acoustic waves can be focused and

thereby target areas where cavitation activity would be useful. Focusing could play an

important role in situations where only a small region of the body should be affected

while leaving other tissues unharmed.

Acoustic cavitation is believed to be the mechanism by which cell membranes are

ruptured by ultrasound. The violent bubble collapse that occurs during cavitation creates

extremely high local temperatures (thousands of degrees Kelvin) and pressures (hundreds

to thousands of atmospheres), high-velocity jets of fluid, and free radicals (Leighton

1994, Apfel 1997). Light can also be generated as a result of the extremely high

temperatures and pressures, a phenomenon called sonoluminescence. We believe that

these phenomena directly or indirectly lead to cell membrane disruption. It is important

to point out that the resonant bubbles created in water at 24 kHz have been shown to be

-150 microns (Leighton, 1994), which is approximately 10 times the size of the cells we

utilized. During our experiments, we measured changes in molecular uptake and cell

viability induced by a broad range of ultrasound conditions. We also measured the

amount of sonoluminescence light generated. Our hypothesis is that since cavitation is

believed to cause the observed bioeffects, then a measure of cavitation - i.e.,

sonoluminescence - should correlate with these bioeffects over the many ultrasound

conditions examined.

2.4 Drug Delivery Using Acoustic Cavitation

Given the violent nature of acoustic cavitation, it can be deduced that there are

several ways that acoustic cavitation could reversibly or irreversibly damage cell

membranes. A cell in close proximity to one of these violent collapses could be damaged

in a variety of ways. Some cells might avoid damage, others might be reversibly

damaged and still others might be fatally damaged. As will be explained in the Results

section, this range of damage is often what is seen among cells exposed to ultrasound.

In this study, we do not attempt to identify which of the physical events caused by

cavitation leads to the bioeffects that we see. Instead, we are attempting to correlate the

observed bioeffects with quantitative measurements of acoustic cavitation. With a

quantitative measurement of acoustic cavitation, its effects can be monitored and thereby

controlled independent of acoustic parameters. In this study, we monitored two effects of

acoustic cavitation; its ability to kill cells and its ability to deliver a model drug, calcein,
to living cells. Depending on the acoustic conditions utilized, these occurrences are

observed in varying proportions.

2.5 Relevance of Work

A critical problem in pharmaceutical research is the desire to deliver molecules

into cells more efficiently. Acoustic cavitation is the mechanism that we propose to use

to transiently disrupt cell membranes and thereby accomplish this goal. Ultrasound-

induced cavitation is especially attractive because of its ability to target local regions in

the body. For example, tumors could be targeted and made more permeable to

chemotherapeutic agents utilizing ultrasound. In this study, we sought to increase the

utility of ultrasound-induced cavitation by developing a method to control the effects

caused by acoustic cavitation. To that end, we monitored one of the physical measures of

acoustic cavitation, sonoluminescence. By recording the sonoluminescent light output as

we monitored bioeffects and by finding a relationship between the two, we achieved a

feedback mechanism with which to monitor and control molecular uptake and cell

viability.
 CHAPTER III

MATERIALS AND METHODS

3.1 Introduction and Overview

To determine the effects of acoustic cavitation on molecular uptake and cell

viability, and to correlate those effects with sonoluminescence, cavitation was induced in

cell suspensions using a cylindrical piezoelectric transducer (Table 1) and

sonoluminescent light output was measured using a photo-multiplier tube (PMT) (Figure

1). The exposure chamber was similar to that described by Liu et al. (1998) with the

addition of a lens and PMT (R5600U Series, Hamamatsu, Bridgewater, NJ) attached to

the bottom of the chamber. The exposure chamber was filled with 200 mL deionized

water that was degassed for 3 hr at -650 mmHg gage pressure using a vacuum chamber

(Nalgene, Rochester, NY) and pump (KNF Neuberger, Trenton, NJ). Degassing the

water bath removes nucleation bubble sites and therefore limits the majority of cavitation,

and hence sonoluminescent light output, to within the cell sample.

Table 1. Ultrasound transducer specifications

Materials Lead Zirconate Titanate

Outer Diameter 2.0 inches
Wall Thickness 0.125 inches
Length 1.0 inches
Piezoelectric Constant -107 x 10 , 2 m/V
 10

Cell Sample
l
Transducer Mr~ Input Function
Generator
& Amplifier

PMT

Figure 1. Diagram of apparatus, which applies ultrasound to cell suspensions and

simultaneously measures sonoluminescence light production. A cell suspension

contained in a plastic sample chamber is placed within a water bath containing a

cylindrical transducer that is controlled by a function generator and amplifier to produce

ultrasound. At the base of the exposure chamber is a photo-multiplier tube (PMT) which

is used to measure light produced by sonoluminescence.

3.2 Ultrasound Apparatus

The ultrasound generation system consisted of a function generator (DS345 SRI,

Stanford Research Systems, Sunnyvale, CA), ultrasonic amplifier (Macro-Tech 2400,

Crown Audio, Elkhart, IN), matching transformer (MT-56R, Krohn-Hite, Avon, MA),

and transducer (24kHz, Channel Industries, Santa Barbara, CA). This system controlled

the frequency, duty cycle, incident pressure, and pulse length. The total exposure time

was controlled manually by turning the system on and off.

The data acquisition system was composed of a digital oscilloscope (54603B,

Hewlett Packard, Palo Alto, CA) equipped with a HPIB interface module (54650A)

allowing the voltage applied to the transducer and the voltage produced by the PMT to be

downloaded to a PC and stored for analysis. The voltage applied to the transducer was

measured and converted to an incident pressure using the method of Liu et al (1998).

This approach determines the voltage-pressure correlation under conditions where there

is no cavitation and then linearly extrapolates this correlation out to higher pressures.

This approximation will be an overestimate of the actual pressure but provides a better

measure of the total acoustic input. According to this analysis, the following correlation

was established: P = 0.0083 V, where P is incident pressure (atm) and V is voltage (V)

applied to the transducer. To be sure this correlation was developed in the absence of any

cavitation, the acoustic spectrum was monitored using a FFT spectrum analyzer (SR 670,

Stanford Research Systems) and incident pressure measurements were only used if there

was no peak present at the subharmonic frequency (f/2).

3.3 Verification of Ultrasound Apparatus

In order to verify that the light we gathered during the ultrasound exposure was

coming from the cell sample and not from the water bath, we performed a series of

experiments. The first set of experiments performed included setting up the apparatus

exactly like we would for any other experiment, the only difference being that the water

was not degassed and there was no cell sample. We then ran the apparatus through the

different ultrasound conditions and recorded the light output, as shown in Figure 2.

The second set of experiments was similar to the first except this time we

degassed the water bath thereby reducing the number of cavitation nucleation sites (again

no cell sample used). Again we tested the range of different ultrasound conditions and

from these two sets of experiments, we concluded that the light generated was indeed

coming from the cell sample and not from the water bath itself.

3.4 Preparing Cell Samples

DU 145 prostate cancer cells (American Type Culture Collection, Rockville, MD) were

grown and harvested using protocols described below.

Cell Harvest Protocol

Harvesting a 150 cm2 flask (approximately 12 million cells):

1) Growth media over the cell monolayer was removed using a pipette.
2+ 2+
2) Monolayer was washed with 15 mL PBS (Cellgro) without Ca / Mg for 1-2

minutes and then removed using a pipette.

 13

6.E-02

0 4 6 8 10
Pressure (atm)

Figure 2. Results of apparatus validation experiments showing the ability to distinguish

the light output from the sample and the light output from the water bath. The integrated

light output from a sample was five times or more the light output generated from the

degassed water bath without a sample.

3) 5 mL of trypsin were then added to the flask and the flask's cap was replaced and

tightened down. The flask was gently shaken to ensure full coverage of the monolayer by

the trypsin. The flask was then placed in the incubator (Model 3110, Forma Scientific,

37°C) for 2 minutes.

4) The flask was then removed from the incubator. The trypsin was known to have

successfully dislodged the cells when the monolayer could be seen to slide off the flask

wall. The flask was tapped gently to ensure that all cells were suspended in the trypsin.

30 mL of full growth media were added to the cell suspension to deactivate the trypsin.

After harvest, the cells were spun down (1000 x g, Model BS-15R, Beckman

centrifuge) and resuspended in growth media. The cell concentration was determined

using a cell counter (Coulter Multisizer II) and then diluted to a final concentration of

750,000 cells/mL. The sample chambers (Sedi-Pet transfer pipette, 241, SAMCO, San

Fernando, CA; stems cut to 2 cm in length) each contained 1.2 mL of a well-mixed

suspension of 750,000 cells/mL in a 10 |LiM solution of calcein (623 Da, r = 0.6 nm,

Molecular Probes, Eugene, OR). Calcein (Figure 3) was used as a model drug because it

is of similar size to many drugs in common use and is unable to readily penetrate the cell

membrane. The suspension was introduced into the chambers using a 22-gauge needle

and 3 mL syringe (Becton Dickinson, Franklin Lakes, NJ). A meniscus was left at the

top of the pipette stem and a 1/16" stainless steel rod was inserted 2 mm into the stem.

Overflow of solution was allowed to ensure no visible air bubbles were trapped in the

sample.
 o o
II II
(HOCCH2)2NCH2 CH2N(CH2COH)2

Figure 3. Chemical structure of the calcein molecule (courtesy of Molecular Probes).

The sample was positioned in the cylindrical water bath exposure chamber so that

the center of the sample was in the radial and axial center of the transducer. Pulsed

ultrasound at 24 kHz and 10% duty cycle was applied to the cell sample using

combinations of 3 pulse lengths (1-100 ms), 3 total exposure times (0.1-10 s), and 10

different pressures (1.0-9.8 atm). After ultrasound application, the cells remained at

room temperature for approximately 5 min to allow cell recovery. Cell samples were

then washed 3 times by centrifugation (735 x g, 3 min, Eppendorf 5415 C, Brinkmann

Instruments, Westbury, NY), resuspended in Dulbecco's phosphate-buffered saline

(Cellgro, Herndon, VA), and kept on ice until analysis the same day.

3.5 Flow Cytometry Assay

The samples were assayed using flow cytometry to quantify cell viability and the

number of calcein molecules taken up by each cell. To assess viability, we added to each

cell sample 5 uL of 10 mg/mL propidium iodide (Molecular Probes) to stain non-viable

cells. We also added 105/mL green fluorescent latex microspheres (d = 2.5 um,

Molecular Probes, 488/515, lot #5881-2), which were used as an internal volumetric

standard to determine the cell concentration after ultrasound application, as described by

Prausnitz et al. (1994).

A FACScan flow cytometer using Cellquest software (BD Immunocytometry

Systems, San Jose, CA) was utilized to measure green calcein fluorescence associated

with molecular uptake and red propidium iodide fluorescence associated with non-viable

cells. At least 10,000 viable cells were collected from each sample. The samples were

excited using a 488 nm argon laser. The fluorescence of propidium iodide (675/20 nm

bandpass filter for emission collection) was used to distinguish between viable and non-

viable cells. Viable cells were analyzed to determine their mean fluorescence intensity of

calcein (530/30 nm bandpass filter). Fluorescence intensity was then converted to an

average number of calcein molecules per cell using calibration beads (Quantum 25, Flow

Cytometry Standards Corporation, Fishers, IN, d = 7.2 um, lot #B00136) with known

fluorescence intensity (Prausnitz et al. 1994).

3.6 Data Analysis

As described previously (Guzman et al. submitted), the fluorescence intensity of a

population of cells can be represented by a histogram (Figure 4). These histograms from

flow cytometry data were initially analyzed with WinMDI 2.8 (Windows Multiple

Documents Interface, Joseph Trotter, Scripps Research Institute, La Jolla, CA) to

determine the fraction of cells that were viable and, among the viable cells, the number of

molecules per cell. Then, molecular uptake data were imported to a Microsoft Excel
(version 9.0) spreadsheet using Ldata (Robinson and Kelly, 1998), a numerical

conversion formula; reformatted; and finally imported to the statistical software package,

MiX 3.1 (Ichthus Data Systems, Hamilton, Ontario, Canada) where histograms of

molecular uptake were fitted as three normal distributions.

3.7 Analysis of Three Populations

As described in the Results section, histograms were best described as complex

distributions containing up to three sub-populations. In order to determine the relative

percentages of cells in the three sub-populations, we first had to decide how many

populations were present. As shown in Figure 4 A, a single population sometimes

adequately describes histograms. In other cases, two sub-populations are evident: one at

the same position as the single population and another over a broad distribution at higher

levels of fluorescence (Figure 4 B). Finally, three sub-populations are sometimes needed

to describe the data, which include the previously mentioned two distributions, with a

third population at even higher fluorescence (Figure 4 C-E).

Although one could visually identify when multiple sub-populations were present,

we wanted a uniform, quantitative method to determine the number of sub-populations

present. First, we classified control (no ultrasound) samples as having a single

population and identified the value below, which the fluorescence of 97% of cells fell.

This value was considered the maximum fluorescence of control (i.e., the "cut-off).

Excluding the brightest 3% of "events" in flow cytometry is often done to discard

artifacts (Canatella et al., submitted).

Number of Cells Number of Cells Number of Cells Number o
r
Figure 4. Series of representative histograms showing calcein uptake by viable cells at

increasing ultrasound pressures. (A) A control population of cells shows weak

background fluorescence. (B) Cells exposed to ultrasound at 1.9 atm show a large sub-

population with essentially no effect and smaller group of cells with higher fluorescence,

indicating uptake of calcein. (C-E) Ultrasound applied at increasing pressure (2.9, 3.9

and 4.9 atm) causes increasingly more cells to take up calcein molecules. Because these

highly non-Gaussian distributions cannot be characterized by a single average, we

separated each population into nominal uptake (NUP), low uptake (LUP) and high uptake

(HUP) sub-populations by fitting three Gaussian distributions, as illustrated in (C)

(Guzman et al. submitted). Ultrasound exposures were performed for 2 sec total

exposure time and 10 ms pulse length.

 For samples exposed to ultrasound, we determined the fraction of cells above the

"cut-off fluorescence. If this value was less than 5%, then that sample was classified as

one population (i.e., no different from control); if it was greater than 5% but less than

15% then it was classified as two populations (unless we could see from the uptake

distribution that it was clearly three populations); and lastly if cells with fluorescence

above the "cut-off represented 15% or more of the cells then we classified that sample

as having three populations. The selection of 5% and 15% criteria is somewhat arbitrary.

In the absence of a more rigorous way to determine the number of sub-populations, we

selected these criteria because they classified the samples in a manner consistent with our

visual classification. Once we had classified all of the sample histograms, then we could

properly analyze them with the MiX 3.1 software, which cannot determine how many

sub-populations are present but will find the distribution among a known number of sub-

populations.

3.8 Sonoluminescence Analysis

Light output gathered by the digital oscilloscope (Hewlett Packard, 54603B) was

downloaded to a computer (Ginstar Computer, Norcross, GA) for analysis. Light output

generated during all ultrasound pulses applied to a given cell sample were combined to

yield an average light-vs.-time trace for each sample. Although the light data had

different shapes, in Figure 5 we have displayed some representative waveforms.

 14 -l

Q. 12
1
A
E
o 10 -I
>
E
> Peak
«
c
•2
c

0.2 0.4 0.6 0.8 1 1.2 1.4

T i m e (m s)

Figure 5. Typical light intensity waveforms showing that for the same conditions the

light output waveform varied. We sought to determine if ultrasound's effects on cells

could be correlated with integrated light output (i.e., total area under the curve), light

output above a threshold (i.e., area under the curve above a threshold level) or peak light

output (i.e., light intensity at its peak value).

 22

CHAPTER IV

RESULTS AND DISCUSSION

4.1 Introduction

This study examined the effects of pressure, pulse length, and total exposure time

on calcein uptake and cell viability, and sought to correlate those effects with

sonoluminescence. Prostate cancer cells were exposed to 24 kHz ultrasound and then

analyzed using flow cytometry to determine the number of calcein molecules taken up

into the cytosol and the loss of cell viability.

4.2 Heterogeneous molecular uptake

The number of calcein molecules taken up by cells showed significant

heterogeneity. Figure 4 displayed a series of uptake histograms in which acoustic

pressure was varied, while pulse length and total exposure time were held constant. Each

histogram shows the distribution of molecular uptake among a population of 10,000

viable cells. The control sample (Figure 4 A) shows a single peak with weak background

fluorescence probably due to autofluorescence, membrane-bound calcein, and optical and

electrical noise. Histograms of sonicated samples (Figures 4B-E) show broad

distributions of fluorescence, which indicates heterogeneous amounts of calcein uptake

within single populations of cells. Although the distribution shape depended on pressure,
 23

pulse length, and total exposure time (see below), the existence of a low-fluorescence

peak, a high-fluorescence peak and a broad valley in between was seen in most samples

(e.g., Figure 4).

As described previously (Guzman et al. submitted), to facilitate analysis we have

divided each uptake histogram into three sub-populations: nominal uptake population

(NUP), which corresponds to the low-fluorescence peak interpreted as indistinguishable

from controls; low uptake population (LUP), which corresponds to the broad valley

between the peaks interpreted as cells with varying amounts of uptake; and high uptake

population (HUP), which corresponds to the high-fluorescence peak interpreted as cells

with the maximum possible uptake (i.e., achieved equilibrium with extracellular

environment).

Although the relative sizes of these sub-populations depended strongly on

ultrasound conditions, the number of molecules taken up within each sub-population

remained the same order of magnitude. Graphically, this means that the heights of the

peaks changed, but their positions did not vary much. For all of the ultrasound conditions

examined in this study, the average number of molecules taken up by NUP cells was 3.53

x 104 ± 4.08 x 104, LUP cells was 2.55 x 106 ± 4.74 x 105, HUP cells was 8.70 x 106 ±

8.77 x 105 (average ± standard error of the mean). Based on a cell volume of 2xl0"9 ml

per cell (data not shown), the average calcein concentration within HUP cells was 7 JLIM.

Because calcein was supplied extracellularly at 10 uM, this supports the interpretation

that HUP cell uptake approaches equilibrium.

 24

4.3 Effects of acoustic parameters on molecular uptake and cell viability

The effects of pressure, pulse length, and total exposure time on molecular uptake

and cell viability are shown in Figure 6. The overall height of each bar represents cell

viability, while the three stripes within each bar represents the fraction of cells in NUP

(gray), LUP (white) and HUP (black). Each graph shows the effect of pressure, while

each row corresponds to a different pulse length and each column to a different total

exposure time.

4.3.1 Effects of pressure on bioeffects

These graphs show that as pressure increases, cell viability generally decreases.

Similarly, as pressure increases, NUP cells also decrease in number. However, the

fraction of LUP and HUP cells (i.e., cells with significant levels of uptake) does not

change much as a function of pressure, above an apparent threshold of approximately 1.0

atm. Except when cell viability is extremely low, the sum of LUP and HUP cells in each

graph is relatively constant.

4.3.2 Effects of total exposure time on bioeffects

Figure 7 replots the data shown in Figure 6 to better display the effects of total

exposure time and pulse length. In each graph, pressure is held constant and the effects

of three different pulse lengths are examined at three different total exposure times.

Scanning across each graph shows that as total exposure time increases, cell viability

generally decreases.
 100 ms TET, 1 ms pulse 2 s TET, 1 ms pulse 10 s TET, 1 ms pulse
Itl
rh rh h rfirh rh
100^ m

rh r h rh rh
100 \
n [*i
i+i 75 75-1
ih "I r*i r+i rh
r 50
i
H hi rn H \\ H
(+1
rh rh ^]
rh

r*n

WW W W Hn r L-J
S 50-1 25J 25 Hh
FHW m m -••••••• J
Q0 1.0 1.9 29 39 4.9 59 69 7.9 88 9.8
0-
j QO 1.0 1.9 29 09 4.9 59 69 7.9 88 98
i i
QO 1.0 1.9 29 39 4.9 59 69 7.9 88 98
P r e s s u r e (atm) Pressure (atm) Pressure (atm)

100 ms TET, 10 ms pulse 2 s TET, 10 ms pulse 10 s TET, 10 ms pulse

rh M [fl n rh
w

8"
rtwill
II
a 25-1
0 w
00 1.0 1.9 29 39 45 59 69 7.9 88 98 00 1.0 1.9 29 39 4.9 59 69 7.9 88 98 00 1.0 1.9 29 09 4.9 59 69 7.9 88 9.8
P r e s s u r e (atm) Pressure (atm) P r e s s u r e (atm)

100 ms TET, 100 ms pulse 2 s TET, 100 ms pulse 10 s TET, 100 ms pulse
100- 100-

r-J-l [ 75-
Hhf"
3*1 -h rh rh rh
I"]
<u
a 25J
l+l fir h[h t\rh
50-

25-
rh
o-l
fcJJJJ
QO 1.0 1.9 29
w aid
1 1— U-j
39 4.9 59 69 7.9 88 98
JJfi -1 0-
QO
aCw
1.0 1.9 29 39 4.9 59 69 7.9 88 98 QO 1.0
HM-M-T-
1.9 29 39 4.9 59 69 7.9 88 9.8
P r e s s u r e (atm) Pressure (atm) P r e s s u r e (atm)

Figure 6. Cell viability and distribution of cells among NUP (gray), LUP (white) and

HUP (black) sub-populations following exposure to ultrasound as a function of pressure,

total exposure time and pulse length. The height of each bar represents total cell

viability, whereas the stripes show the distribution of cells among the three sub-

populations. Each column shows results at the same total exposure time (TET) and each

row is at the same pulse length. Data represent averages of at least three replicates with

standard errors of the mean shown. (Graphs can be seen in greater detail in Appendix)
 Percent Cells Percent Cells Percent Cells Percent Cells
N3 tn ^i o
3 tn o tn o
I 1 i i i
i

1+ + £ 3^
^ -»
+ +
I
^
1

1 + ± £
10

+ 3 S 3 5

*°
o
3 "
T-
13 1 m o
H 3° ' E 33
~T + + FT" -
+ +
s
M

WFJ} •D3 w _
H °
m + 13-
to+

T
to

100
* + 3' + -

E
3'
- ft +t HLJ
- --rn
J

zn ro _»
CO _,

m
H
i|
3CD m
H°
i|&
CO
03

3
O
«•

Ej
co

3 f"
°

I ED
3ZJ
Percent Cells Percent Cells Percent Cells Percent Cells
r>o tn ro tn r>o tn -J
> tn o tn
• i i

3- -- p EI 13- •+ +
I :E 3-

1
100 ms
^Z3 = r+ + 1 \ +

10
3 o
+
3 S
w 1 T
? + 8 +
3 "
H^ m o
-i °
3 - '+ + "| 8
3 " + +
33 33 - » + +
{
--H CO
bo
_,
o bo o m-
03 » D) »
3 m 3 rn
 27

Figure 7. Data from Figure 6 are re-plotted to better show dependence on total exposure

time and pulse length. Each graph provides results at a different pressure. Within each

graph, three total exposure times (TET) are considered, each at three different pulse

lengths. (Graphs can be seen in greater detail in Appendix)

4.3.3 Effects of pulse length on bioeffects

Looking within each triplet of bars indicates that pulses of 1 ms duration often

cause weaker effects than those of 10 ms or 100 ms duration.

4.4 Correlation of sonoluminescence with bioeffects

Although the data presented in Figures 6 and 7 characterize the effects of

ultrasound parameters on cells, we wanted to determine if there was a single measurable

parameter that would correlate with uptake and viability for all of the ultrasound

conditions tested. The parameter we selected was sonoluminescence, or the light

generated by cavitation. Our hypothesis was that since cavitation is believed to cause the

observed effects on cells, then a measure of cavitation - i.e., sonoluminescence - might

correlate with these bioeffects over the broad range of ultrasound conditions used in this

study.

To test this hypothesis, we have replotted the uptake and viability data not as a

function of ultrasound parameters, as shown in Figures 6 and 7, but instead as a function

of total integrated light output produced by sonoluminescence as measured using a photo-

multiplier tube mounted at the base of the ultrasound exposure chamber. Figure 8 shows

that molecular uptake, viability and the fraction of cells in NUP, LUP, and HUP all

correlate with sonoluminescence. The many graphs shown in Figures 6 and 7 can be

reduced to the set of single curves shown in Figure 8. As mentioned above, molecular

uptake in each of the sub-populations shows scatter, but is generally of the same order of

magnitude within each sub-population (Figure 8 A). Figures 8 B - 8 E indicate an

apparent threshold at an integrated light output near 10"2 Vxsec, above which, cell
viability drops steeply. However, at lower levels of integrated light output, viability

remains high and a significant fraction of cells (e.g., 20 - 30 %) are in the LUP or HUP

sub-populations.

Effects on cells are seen to correlate with total integrated light output, which is

defined as the average integrated light output during a pulse (e.g., the integral of one of

the curves shown in Figure 5) multiplied by the total number of pulses applied during the

given exposure. We also looked for correlations with other features of the measured

sonoluminescence, but none correlated as well as integrated light output. For example,

we tried to correlate bioeffects with the peak level of light output (see Figure 5), since the

creation of bioeffects may be a catastrophic event that only requires achieving some peak

level of cavitation, even for just an instant. However, the data were extremely scattered

using this approach (Figure 9 A), suggesting that this interpretation was incorrect.

Similarly, correlation of peak light output multiplied by the number of pulses (i.e., the

number of times the peak was achieved) also yielded a poor correlation (Figure 9 B).

Another approach was to correlate bioeffects with sonoluminescence only above a

threshold, since weak cavitation that created low levels of sonoluminescence might not

produce bioeffects. However, correlation of bioeffects with light output integrated above

a number of different thresholds (see Figure 5) yielded progressively better correlations

as the threshold was lowered (Figure 9 C), but the best correlation remained when the

threshold was zero (i.e., total integrated light output shown in Figure 8). This suggests

that any cavitation capable of generating sonoluminescence was sufficiently "strong" to

induce effects on cells.

 30

1.E*08

• S 1.E-+07
0
o Q. 100- 0
o U c&
•Si
fl) 1.E+06- Z
iHUP

Cells in
jD OUUP
1.E+05-

0
O ONUP

.E 1.E+04- 0 < ^
<1>
_o c
a> ,
1- °«L *
Q 1.E+03
0
a>
*«w
• •<* 0w°
a.
i.E-tce 0.1- .. -1 —1 1 1

1.E-06 1.E05 1.E-04 1.E«3 1.E02 1.&01 LE4OO 1.E-06 1.E05 1.&C4 1.E03 1.E-02 1.E01 1.&00
I n t e g r a t e d Light Output (Vsec) Integrated LlghtOutput(Vsec)

B D
100 100-

£
ra
10
3*
>
* • • t

0.1 0.1
1.&06 1.E-05 1.E-04 1.E-03 ^.BC2 1.E01 1.E+00 1.E06 1.E05 1.E04 ^.BC3 1.E-02 1.E-01 1.E+00
I n t e g r a t e d Light Output (Vsec) I n t e g r a t e d Light Output (Vsec)

Q. 100
X
A
C k * * l.
W 10
<D
o * Atffft

0.1
1.E06 1.E-C5 1.E04 1.E03 1.E<32 1.E01 1.E+O0
I n t e g r a t e d Light Output (Vsec)

Figure 8. (Caption on next page)

Figure 8. Uptake and cell viability shown as functions of sonoluminescence light output.

(A) Calcein molecules per cell in each of the three sub-populations are relatively

independent of light output. (B) Cell viability decreases sharply above a threshold level

of light output (e.g., 1 x 10"2 Vxsec). (C - E) Percent cells in NUP, LUP and HUP is

relatively constant and then decreases sharply above a threshold level of light output

(e.g., 1 x 10"2 Vxsec). All of the data from Figures 6 and 7 collected over a broad range

of ultrasound conditions are re-plotted here and shown to correlate well with

sonoluminescence light output.

 A
100 H
1 . • *'i***
• J*. • » • • « • « *
/ • • . • «•.
n
(Q

T> 1 H
O
i' < 1 *
0.1 • •• ,
1.E-04 1.E-03 1.E-02 1.E-01
Peak L i g h t O u t p u t (V)

B
100 H

•• •*$ # • • % . •
f ,0- ..•• •*»
!Q
TO
•:«
>
% x** * •
o
0.1
V''W;l
• • *
1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02
Peak L i g h t O u t p u t * # of P u l s e s (V)

c
100-
• V • • • • > ;
• • • •*• ••,• •
£ io- • • /•
_•
(0 • • .* . 7 . ,
Cel IV

0.1 i ^__^^_ T
% •• .'"<*i
1
v
1
•
1
•
1 1
1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00
Integrated Light Above 10 mV Threshold (V*sec)

Figure 9. Cell viability correlations with other sonoluminescent parameters show a more

scattered relationship. (A) Correlation with peak light output, (B) Correlation with peak

light output times the number of pulses, (C) Correlation with integrated light output

above a threshold of 10 mV.

 33

4.5 Discussion

The most significant observation from this study is the overarching correlation

between cavitation-induced sonoluminescence and ultrasonic bioeffects. We believe that

this is the first time such a correlation has been demonstrated. In a related study, Liu et

al. (1998) showed that disruption of red blood cells correlated with another measure of

cavitation - the subharmonic pressure measured at one-half the applied frequency.

Combined, these and other observations suggest that cavitation is the mechanism by

which ultrasound disrupts cells.

The observed correlations also suggest possible applications for drug and gene

delivery, where sonoluminescence or subharmonic pressure might be used to provide

feedback about the cavitation and its associated bioeffects generated by a given

ultrasound exposure. For example, a tissue in the body or solution of cells in the lab

could be exposed to ultrasound while simultaneously monitoring sonoluminescent output.

Once the total integrated light output reached a desired level (e.g., 10" V-sec for the

system used here), sonication could be stopped. In this way, a constant dose of cavitation

(as measured by sonoluminescence), rather than a constant dose of ultrasound, can be

applied and thereby potentially achieve more reproducible bioeffects.

A notable observation from this study is that there was a broad distribution of

cells with different amounts of uptake characterized by two peaks and a broad valley in

between. This type of distribution has been observed before using a similar protocol, but

involving ultrasound applied at 500 kHz (Guzman et al. submitted), which leads us to

believe that this heterogeneity in uptake may be a general feature of ultrasound-disrupted

cells. It is not currently clear whether heterogeneity is due to cell-based differences in

their response to ultrasound or physically-based differences due to non-uniformities in

ultrasound-induced cavitation. In either scenario, one can imagine some cells being

damaged to the point of cell death, some that are extensively disrupted but nevertheless

survive (HUP), some with fewer disruptions (LUP), and some with essentially no

disruptions at all (NUP). More work is needed to provide a fuller mechanistic

interpretation of the observed heterogeneity and its implications for drug and gene

delivery applications.

This study also presented data on the dependence of uptake and viability on

ultrasound parameters. Generally, greater pressures and longer total exposure times

yielded greater effects on the cells, which is similar to previous observations (Liu et al.

1998; Bao et al. 1997; Miller et al. 1996; Miller et al. 1999). This result is consistent

with a cavitation-based mechanism, since cavitation generally increases in intensity with

increasing pressure and increases in duration with longer exposure times (Leighton

1994).

We also observed how pulse length influenced bioeffects: 1 ms pulses had

significantly less effect than 10 or 100 ms pulses. We hypothesize that 1 ms pulses,

which generate only 24 acoustic cycles, may not be sufficiently long to nucleate, grow

and collapse cavitation bubbles powerful enough to disrupt cell membranes. However,

10 ms pulses, which produce 240 cycles per pulse, may be long enough to generate

sufficient inertial cavitation. Others have similarly observed a dependence on pulse

length for disruption of red blood cells (Ciaravino et al. 1981; Kober et al. 1989; Liu et al.
1998). In contrast, Guzman et al. (submitted) did not observe a dependence on pulse

length for prostate cancer cells exposed to 500 kHz ultrasound for pulse lengths between

20 JJLS (10 cycles) and 60 ms (30,000 cycles). This apparent inconsistency may be

explained by the presence of stabilized gas bubbles (i.e., Optison) in the work of Guzman

et al., which served as nucleation sites and thereby facilitated more rapid onset of

cavitation (Flynn and Church 1988).

 36

CHAPTER V

CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

The main obstacle to delivering therapeutic drugs to cells is the cell membrane.

In this study, we sought to overcome that barrier by using mechanical energy in the form

of ultrasonic waves. The primary impetus of this study was to evaluate the effects of

ultrasound parameters on cell viability and molecular uptake and to establish a

relationship between acoustic cavitation and these effects. The acoustic cavitation

phenomenon that we sought a relationship with is sonoluminescence, which is the burst

of light that is given off when bubbles in solution collapse violently.

To meet our first objective of quantifying the effects of ultrasound parameters, we

made the following observations:

1) As pressure increased, cell viability generally decreased above a threshold

pressure. Also, at the same time that cell viability was decreasing, molecular

uptake was often increasing, shown by the increasing percentages of cells with

calcein uptake.

2) As total exposure time increased, we again observed increasing bioeffects, namely

a decrease in cell viability and an increase in cell viability.

 3) As pulse length increased from 1 ms to 10 ms, there was a notable difference in

bioeffects at the higher total exposure times. There was not such a notable

difference between 10 and 100 ms pulse lengths.

Regarding our second objective, we sought to establish a relationship between

acoustic cavitation and bioeffects (cell viability and molecular uptake), and therefore we

measured the amount of light generated during ultrasound exposures. We utilized several

correlation techniques but the best relationship materialized when the integrated light

output was multiplied by the number of pulses during an exposure, a measurement we

called the total integrated light output. From this correlation, we accomplished the

objective of establishing a relationship between bioeffects and a measurable gauge of

cavitation, sonoluminescence.

The last objective was to identify regions of sonoluminescent light output where

there was both high molecular uptake and high cell viability. This objective was

accomplished by treating the nominal uptake population, low uptake population, and the

high uptake population as a single population using a weighted average. With this

representation we were able to locate a region where there was both high molecular

uptake and high cell viability.

5.2 Recommendations

An interesting study might include observing not only the amount of light during

ultrasound exposure but the location of the light generation as well. This might isolate

whether the light is generated near cells and thereby further determine that acoustic

cavitation is responsible for the bioeffects that are seen. This could be accomplished
through a camera device mounted on the bottom of the device. Another study might

include disseminating the spectrum of light generated during routine ultrasound exposure.

Perhaps the water bath and the cell sample emit spectrum that are different enough to be

distinguished.

Further studies could also be done to see if the bioeffects that we observe would

correlate with any other acoustic cavitation phenomenon, such as the generation of free

radicals. This would be important in determining which measure of cavitation is

primarily responsible for the bioeffects that we see. A free radical scavenger that

fluoresces could be used to detect the presence of hydroxide radicals and through the

analysis used in this thesis or a similar one perhaps a relationship would materialize.

One shortcoming of the device is that only the average light output during a total

exposure can be observed instead of gathering the light output singly from each pulse.

With a different data acquisition device, one might be able to gather each pulse and

therefore get an even more accurate picture of the bioeffects.

 39

APPENDIX A

FIGURE 6 GRAPHS SHOWN

IN GREATER DETAIL
 40

100 ms TET, 1 ms pulse

100 I 1

_w
75
"53
O
*-*
c 50
o
o
1_
<D 25
a

0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.8
Pressure (atm)

100 ms TET, 10 ms pulse

100 -

_w 75 -
o
O rh H-i
50H
i
o —+— -h
k- 25-
a>
Q_
T T-
0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.8
Pressure (atm)

100 ms TET, 100 ms pulse

100-

tf)
^z 75
o
o
+••

c 50
ao>
i_
a)
a. k!6
+ -I-

0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.8
Pressure (atm)
 41

2 s TET, 1 ms pulse

100

v>
a> 75
O
c 50
Q>
O
u
d>
25 -
a

0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.
Pressure (atm)

2 s TET, 10 ms pulse

100
w
~0 75
O
4-«
50
8
i_
0)
°" 25

r *r
0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.8
Pressure (atm)

2 s TET, 100 ms pulse

100
<n
"ai 75
O
*->
c
a) 50
u
a)
l_

a. 25

0.0 1.0 1.9 2.9 3.9 4.9 5.9 6.9 7.9 8.8 9.8
Pressure (atm)
Percent Cells Percent Cells rcen
Ol \3 ai •-J o cn
o o cn o o o
o i

0.0

0.0
1
-

1.0

1.0
1
n II H

1.9
1.9
O

1
cn
H •o H n

2.9
2.9
1

m
H (D
V) 00
^
3.9
^

TET, 1
cn Lu
o
1

O
o c ^ n

4.9
4.9

3 SUI
in 9
1

T3
T3 ^^ rn
C fi) 1
5.9

CO
uls
P *
c/o
1

G) a>
3 CO
P r e s s u r e (atm)

9
6.9
1

B
-•

7.9
7.9
1

8.8
8.8
1

9.8
9.8
 APPENDIX B

FIGURE 7 GRAPHS SHOWN

IN GREATER DETAIL
 1.0 atrr 1
ioo A r*i ph
</> T r^
ph
r+r &
0) 75 -J 1
O rt-
4-
c
O 50 H
O
i_
<D
Q- «r-
25 J

nJ
U -r i i
—s—
i i 1
i i i
m
i i i
+ H
1 10 looms 1 10 100 s 1 10 100 ms
11X 1 mj rEi 2 sTET 10 sTET

1.9 atm
100 H

j2
"5 75
O
+-•
(D 50
o
Q-
25 -\

-r" T
m s m
1 10 100 ms 1 10 100 1 10 100 S
lOOmsTET 2sTET 10STET
 45

2.9 atm
100

J2
O 75
O
* • *

§ 50
k.
o
Q- 25

m s m
1 10 100 ms 1 10 100 1 10 100 s
lOOmsTET 2sTET 10sTET

3.9 atm
100 H

_w
O 75
O
+J

Q> 50
u
l_
Q
Q.
25

m s m s 1 10 100 ms
1 10 100 1 10 100
lOOmsTET 2sTET 10STET
 46

4.9 atm

1 10 100 ms 1 10 100 ms 1 10 100 ms

lOOmsTET 2sTET 10sTET

5.9 atm

1 10 100 ms 10 100 ms 1 10 100 ms

lOOmsTET 2sTET 10sTET
 47

6.9 atm

m s
1 10 100 ms 1 10 100 1 10 100 ™S
lOOmsTET 2sTET 10sTET

7.9 atm

1 10 100ms 1 10 100 m s 1 10 100 m s

lOOmsTET 2sTET 10sTET

 8.8 atm

m ms
1 10 100 s 10 100 1 10 100 ms
lOOmsTET 2sTET 10sTET

9.8 atm
100 A

O 75 A
O
§ 50
O
Q.
25 A

1 10
100msTET
100 ms 1 10
2sTET
100 ms a
1 10
10sTET
100 ms
 REFERENCES

Apfel RE. Sonic effervescence: A tutorial on acoustic cavitation. J Acoust Soc Am

1997;101:1227-1237.

Bao S, Thrall BD, and Miller DL. Transfection of a reporter plasmid into cultured cells
by sonoporation in vitro. Ultrasound Med Biol 1997;23:953-959.

Bronaugh RL and Maibach HI. Percutaneous Absorption: Drugs-Cosmetics-

Mechanisms-Methodology. New York: Marcel Dekker, 1999.

Ciaravino V, Flynn HG, Miller MW. Pulsed enhancement of acoustic cavitation: a

postulated model. Ultrasound Med Biol 1981;7:159-166.

Fechheimer M, Boylan JF, Parker S, et al. Transfection of mammalian cells with plasmid
DNA by scrape loading and sonication loading. Proc Natl Acad Sci 1987;84:8463-8467.

Flynn HG and Church CC. Transient pulsations of small gas bubbles in water. J Acoust
Soc Am 1988;84:1863-1876.

Guzman HR, Nguyen DX, Khan S, and Prausnitz MR. Ultrasound-mediated disruption of
cell membranes II: Heterogeneous effects on cells. J Acoust Soc Am (submitted).

Kober LO, Ellwart JW, Brettel H. Effect of the pulse length of ultrasound on cell
membrane damage in vitro. J Acoust Soc Am 1989;86:6-7.

Kremkau FW. Diagnostic Ultrasound: Principles and Instruments. Philadelphia: W.B.

Saunders, 1998.

Langer R. Drug delivery and targeting. Nature 1998;392:5-10.

Leighton TG. The Acoustic Bubble. London: Academic Press, 1994.

Liu J, Lewis TN, and Prausnitz MR. Non-invasive assessment and control of ultrasound-
mediated membrane permeabilization. Pharm Res 1998;15:918-924.

Miller DL, Bao S, and Morris JE. Sonoporation of cultured cells in the rotating tube
exposure tube. Ultrasound Med Biol 1999;25:143-149.
Miller MW, Miller DL, and Brayman AA. A review of in vitro bioeffects of inertial
ultrasonic cavitation from a mechanistic perspective. Ultrasound Med Biol 1996;
22:1131-1154.

Mir LM and Orlowski S. Mechanisms of electrochemotherapy. Adv Drug Deliv Rev

1999;35:107-118.

Mitragotri S, Blankschtein D, and Langer R. Ultrasound-mediated transdermal protein

delivery. Science 1995;269:850-853.

Munson BR, Young DF, and Okiishi TH. Fundamentals of Fluid Mechanics. John Wiley
& Sons, Inc. New York. 1994.

Park K. Controlled Drug Delivery: Challenges and Strategies. Washington, DC,

American Chemical Society. 1997.

Pierce AD. Acoustics: An Introduction to Its Physical Principles and Applications. The
Acoustical Society of America, 1989.

Prausnitz MR, Milano CD, Gimm R, Langer R, and Weaver JC. Quantitative study of
molecular transport due to electroporation: Uptake of bovine serum albumin by
erythrocyte ghosts. Biophys J 1994;66:1522-1530.

Rolland AP. From genes to gene medicines: Recent advances in nonviral gene delivery.
Crit Rev Ther Drug Carrier Syst 1998;15:143-198.

Saad AH, Hahn GM. Ultrasound-enhanced effects of Adriamycin against murine tumors.
Ultrasound Med Biol 1992;8:715-723.

Silverman RB. The Organic Chemistry of Drug Design and Drug Action. New York:
Academic Press, 1992.

Wyber JA, Andrews J, and D'Emanuele A. The use of sonication for the efficient
delivery of plasmid DNA into cells. Pharm Res 1997;14:750-756.