Computer Software

Daniel D. Clark *
Virtual Protein Purification: A Simple Exercise Daniel J. Edwards
to Introduce pH as a Parameter that Effects
Ion Exchange Chromatographyw s

From the Department of Chemistry and Biochemistry, California State

University-Chico, Chico, California 95929-0210

Abstract
This article describes a simple exercise using a free, easy-to- majors at California State University-Chico. Given the versatil-
use, established online program. The exercise helps to rein- ity of the program, this work is also a model for instructors that
force protein purification concepts and introduces undergradu- wish to develop their own exercise to help teach other protein
ates to pH as a parameter that affects anion-exchange purification techniques. VC 2017 by The International Union of

chromatography. The exercise was tested with biochemistry Biochemistry and Molecular Biology, 46(1):91–97, 2018.

Keywords: Protein purification; ion exchange; computer software;

web-assisted learning

Introduction purification. While actual lab time may never permit it, a
virtual lab experience can easily facilitate the exploration
Protein purification is an important topic in the undergrad-
of chromatographic parameters in a rapid manner. Herein
uate biochemistry curriculum. In a lecture setting, a discus- we present a simple virtual protein purification exercise
sion of ion exchange, size-exclusion, and affinity-based that utilizes a free, established online program [1]. The pri-
purification techniques is common. In the laboratory, the mary objectives of the exercise were to (1) reinforce previ-
implementation of one or more of these techniques is also ously learned protein purification concepts (specifically ion
common. However, despite this coverage, undergraduates exchange), and (2) introduce students to pH as a parameter
typically do not have an opportunity to explore key chro- that effects anion-exchange chromatography. A secondary
matographic parameters, for a particular media (or resin) objective was to have students discover the value of in-
type, that can affect a protein separation. To afford stu- vestigating both “strong” (Q) and “weak” (DEAE) anion-
dents this kind of experience, not available from one itera- exchange media as part of a protein purification. We
tion of a supplied protocol, would enhance their conceptual sought to accomplish these objectives through six protein
understanding and practical ability to execute a protein purification simulations that would generate realistic FPLC
chromatograms for students to think about, discuss, and
interpret.
Volume 46, Number 1, January/February 2018, Pages 91–97
Abbreviations: FPLC, fast protein liquid chromatography; DEAE-cellu-
lose, diethylaminoethyl-cellulose; CM-cellulose, carboxymethyl-cellu- Background
lose; Q-Sepharose, trademarked (GE Healthcare) resin; Q refers to
quaternary amine; S-Sepharose, SP-Sepharose, trademarked (GE Proteins are purified using techniques that separate them
Healthcare) resin; S refers to sulfopropyl; LC-MS/MS, liquid according to differences in charge, size, hydrophobicity,
chromatography-tandem mass spectrometry and molecular recognition (affinity). Ion exchange chroma-
*To whom correspondence should be addressed. Department of tography separates proteins based on differences in net
Chemistry and Biochemistry, California State University-Chico, 400 surface charge, which is highly pH dependent [2]. The iso-
West First Street, Chico, CA 95929-0210. Tel.: (530)-898-5251.
electric point (pI) of a protein is the pH at which it carries
E-mail: ddclark@csuchico.edu
w
s Additional Supporting Information may be found in the online no net charge. Protein titration curves indicate that at pH
version of this article. values less than the pI, a protein will carry an increasing
Received 10 April 2017; Revised 13 June 2017; net positive charge, while at pH values greater than the pI,
Accepted 18 July 2017 a protein will carry an increasing net negative charge [2].
DOI 10.1002/bmb.21082 Ion exchange chromatography includes anion and cation
Published online 6 August 2017 in Wiley Online Library exchangers. Anion exchangers employ a positively charged
(wileyonlinelibrary.com) resin interacting with an exchangeable anion, while cation

Biochemistry and Molecular Biology Education 91

 Biochemistry and
Molecular Biology Education

exchangers employ a negatively charged resin interacting completed without instructor help in about an hour. The
with an exchangeable cation. A “strong” anion exchanger, exercise as provided to students (see Supporting Informa-
such as Q-Sepharose is positively charged across a broad pH tion, pages 2–4) includes Part 1, a concept review, and Part
range, whereas a “weak” anion-exchanger, such as DEAE- 2, the virtual ion exchange chromatography, which uses
cellulose will lose its charge above pH 9 [3]. Therefore, col- the online program. In Part 2 students use the three-
umn equilibration pH is an important parameter in ion component “Easy3” protein mixture to execute six virtual
exchange chromatography as it can affect protein charge ion exchange separations. The simulations (A–F) vary the
and, depending on the resin, the charge on the column. Both ion exchange media, equilibration pH, and the salt gradient
of these affect protein adsorption and separation. as follows: (A) Q-Sepharose, pH 6.0, 0–0.5 M salt, (B) Q-
Sepharose, pH 7.0, 0–0.5 M salt, (C) Q-Sepharose, pH 8.0,
0–0.5 M salt, (D) Q-Sepharose, pH 10.0, 0–0.5 M salt, (E) Q-
Program Description Sepharose, pH 10.0, 0–1.0 M salt, (F) DEAE-Cellulose, pH
Protein Purification is a free online program developed by 10.0, 0–0.5 M salt. In each simulation, students use the
the late (June 2016) Dr. Andrew G. Booth at the University “assay enzyme activity” feature to locate fractions that con-
of Leeds (http://www.agbooth.com/pp_ajax/) [1]. The pro- tain Protein 1. Students draw the results as a FPLC chro-
gram offers four protein mixtures of varied complexity to matogram on the appropriate panel (A–F) of a six-panel fig-
explore: Easy3 (three proteins), Example (six proteins), ure in the exercise (see Supporting Information, page 3).
Default (20 proteins), and Complex (60 proteins). Users can Their chromatograms include an A280 trace, the linear salt
select any of the proteins within a mixture to purify. The gradient, and the location of Protein 1. Additionally stu-
separation techniques available are ammonium sulfate dents use the linear salt gradient to estimate the [salt]
fractionation, heat treatment, gel filtration, ion exchange (within 25 m M) required for elution of Protein 1. Figure 1
chromatography, hydrophobic interaction chromatography, shows the six screen shots (A–F) that students would gener-
and affinity chromatography. Within each of these techni- ate using the program and the annotations (*) that students
ques, users can alter various parameters. In the case of ion would use to indicate the location of Protein 1, and the esti-
exchange chromatography, anion-exchange media (Q- mated [salt] required to elute Protein 1. Table 1 presents
Sepharose and DEAE-cellulose) or cation-exchange media the seven questions that follow use of the program, our
(S-Sepharose and CM-cellulose) are available. With either, rationale for their inclusion, and an evaluation of student
users define the method of elution as either a “salt performance on each question.
gradient” or “pH gradient”. With salt gradient selected,
users can define the pH of the equilibration buffer (2.0–
11.0) and the start and end salt concentrations within a 0 Implementation
to 3.0 M range. When a column-based purification tech- At California State University-Chico, we tested this exercise
nique is chosen, a FPLC chromatogram appears that dis- in our seven hours per week, biochemistry lab (designed
plays absorbance at 280 nm (A280) on the primary y-axis, for biochemistry majors) during the Spring 2017 semester.
fraction number on the x-axis, and [salt] on the secondary All of the students had previously learned about protein
y-axis (see Fig. 1). Although not employed in our exercise, purification, including ion exchange chromatography, in
the program also features virtual 1D- and 2D-gels (with lecture. In the schedule of the course, the exercise pre-
coomassie blue and immunoblot staining available) that ceded two major projects (1) a scripted three-week purifi-
enable users to analyze fractions from a column. The pro- cation and characterization of a recombinant fluorescent
gram also contains a help feature that includes six tutorial protein, and (2) an unscripted seven-week student designed
exercises, but practice with these is not required for the project aimed at the purification, identification (via LC-MS/
exercise described in this work. A literature search indi- MS), and characterization of a dehydrogenase enzyme from
cates that two publications using the program are avail- its native source. The exercise was presented during a
able, and both are from the author of the program [1, 4]. three hour lab as the prelude to an open-ended “dry lab”
More recently, a web-based applet that simulates an ion protein purification experience (described below) using the
exchange purification of over produced proteins from E.coli same program. Each student completed the exercise indi-
was described [5]. However, we decided to use the program vidually without instructor help. Immediately afterwards,
devised by Dr. Andrew Booth for its ease of use and versa- each student’s attitudes toward the exercise were assessed
tility that extends to techniques beyond ion exchange chro- by way of an anonymous survey (Table 2).
matography [1, 4]. Following assessment, each student was assigned a
protein (1–60) at random from the “complex mixture” and
had about two hours to optimize a purification strategy.
Exercise Description Importantly, at this stage, students could explore the vari-
In an effort to keep the exercise simple, we designed it ety of techniques the program offers, use virtual gels, and
with a focus on anion-exchange chromatography. It can be pool fractions. To increase the difficulty, and emphasize the

92 Virtual Protein Purification

 Screen shots of FPLC chromatograms using the program. Each simulation used the “Easy3” mixture. Asterisks indicate
FIG 1 the location of Protein 1 with the estimated [salt] required for elution given. Ion-exchange media, equilibration pH, and
the salt gradient varied as follows:(A) Q-Sepharose, pH 6.0, 0–0.5 M salt, (B) Q-Sepharose, pH 7.0, 0–0.5 M salt, (C) Q-
Sepharose, pH 8.0, 0–0.5 M salt, (D) Q-Sepharose, pH 10.0, 0–0.5 M salt, (E) Q-Sepharose, pH 10.0, 0–1.0 M salt, (F) DEAE-
Cellulose, pH 10.0, 0–0.5 M salt. Panels lettered according to the exercise (see Supporting Information, page 3). [Color
figure can be viewed at wileyonlinelibrary.com]

other available techniques, affinity chromatography was these restrictions was to align the experience with their
not allowed. As analytical tools to guide their purifications, seven-week student designed purification later in the
only the “assay enzyme activity” and 1D-gel (coomassie semester. Optimization focused on employing a logical
blue) features were permitted. The greater purpose for sequence of purification steps (for example, size-exclusion

Clark and Edwards 93

 Biochemistry and
Molecular Biology Education

Exercise questions, rationale for inclusion, and student performance on each question
TABLE 1

Percentage of
effective student
Questions Rationale for inclusion explanations (%)a

1. Examine the chromatograms for Students are required to think about two 67
Q-Sepharose run at pH 6 and 7 (Panels situations where Protein 1 adsorption did
A & B). Where did Protein 1 elute in both not occur. They will need to consider the
cases? Explain the basis for the charge on the resin and the charge on
observation. Protein 1 at each pH value.
2. Examine the chromatograms for Students discover that an increase in the 83
Q-Sepharose run at pH 6, 7, and 8 (Panels equilibration pH led to a stronger Protein
A–C). What happened to the [salt] 1-column interaction and this necessitated
required for the elution of Protein 1 as the a higher [salt] for elution. Their
equilibration pH increased? Explain the explanation will require consideration of
basis for the observation. how a higher pH might have strengthened
the interaction.
3. Examine the chromatogram for Students are required to consider a situation 67
Q-Sepharose run at pH 10, specifically the where Protein 1 elution did not occur and
one that employed a 0–0.5 M NaCl think about why this happened.
gradient (Panel D). At what [salt] did
Protein 1 elute? Explain the basis for the
observation.
4. Examine the chromatogram for Students discover that a strong Protein 50
Q-Sepharose run at pH 10, specifically the 1-column interaction can be overcome by
one that employed a 0–1 M NaCl gradient increasing the final [salt] of the linear
(Panel E). At what [salt] did Protein 1 gradient.
elute? How does this result compare to the
0–0.5 M NaCl gradient (Panel D)? Explain
the basis for any observed differences.
5. Compare the chromatograms for The comparison illustrates differences 50
Q-Sepharose (Panel D) and DEAE-cellulose between strong (Q-Sepharose) and weak
(Panel F) run at pH 10. How do they (DEAE-cellulose) anion exchangers. An
compare? Explain the basis for any effective answer will require students to
observed differences. consider the structure of the ion exchange
groups on each resin.
6. Examine the chromatogram for Students will discover that Protein 1 does 67
DEAE-cellulose run a pH of 10 (Panel F). not adsorb to DEAE-cellulose at pH 10, but
How might the results have changed if the does at pH 8. An explanation will require
equilibration pH was lowered to pH 8? Try students to think about the structure of the
it with the program and explain the basis DEAE-cellulose ion exchange group, how
for the observed result. its charge is affected by pH, and how this
relates to Protein 1 adsorption.
7. Use the chromatograms to estimate the This question requires student reflection on 100
pI of Protein 1. Explain how you did this. all of the chromatograms, which further
reinforces pH as a parameter that effects
anion-exchange chromatography.

a
Percentages calculated using the student exercises that were available for analysis (n 5 6).

94 Virtual Protein Purification

 Postexercise assessment of student attitudes
TABLE 2

Student response percentages (%)a

Question SA A N D SD

1. This exercise increased my understanding of ion 37.5 37.5 25 0 0

exchange chromatography as part of a protein
purification.
2. This exercise increased my understanding of how 50 37.5 12.5 0 0
equilibration pH can affect protein binding to and
elution from, an anion-exchange column.
3. This exercise helped me to understand how 50 50 0 0 0
protein binding and elution can vary between
“Q” and “DEAE” anion-exchange columns.
4. Although this exercise focused on anion-exchange 37.5 50 0 12.5 0
chromatography, I feel it gave me the foundation
to understand how equilibration pH can affect
protein binding to and elution from, a
cation-exchange column.
5. This exercise should be part of future 37.5 37.5 25 0 0
iterations of the course.

a
Strongly agree (SA), agree (A), neither agree nor disagree (N), disagree (D), strongly disagree (SD). Percentages calculated using data
from all students in the course (n 5 8).

is not an appropriate first step) and a reduction in the total indicated by percentage in Table 1, were those that were
number of steps, but was not concerned with yield or cost well articulated and on trajectory with the high quality
(in person-hours), which the program tracks. Students answers provided in Supporting Information Table S1. As
were required to illustrate their optimized strategy with a shown in Table 1, at least 50% of the available student
detailed flow diagram that included (if applicable) the sepa- explanations were effective for any given question. These
ration technique, type of chromatography resin, equilibra- percentages would likely have been higher if instructor–stu-
tion pH and type of gradient used, which fractions were dent and student–student interactions were permitted during
pooled and the salt concentration range that the pool the exercise. While such interactions are synergistic and
encompassed. Additionally, students completed a purifica- would normally be desirable, for the purpose of this work,
tion table and determined the pI and molecular weight of we chose to forbid them to control influences on student per-
their purified protein using a 2-D gel. In fact, a 2D-gel (coo- formance data (Table 1) and student attitude data (Table 2).
massie blue or immunoblot) was only authorized on the It is noteworthy that 100% of the available student explana-
purified protein obtained via their optimized strategy—for tions were effective for Question 7, as it required reflection
the express purpose of pI and molecular weight determina- on all of their chromatograms to estimate the pI (Table 1).
tion. Therefore, as it would be for a new protein (whose As it was beyond the focus of the present work, the
sequence was unknown), pI information was not directly purification of a protein from the sixty component “complex
available to guide the development of their purification. mix” was not targeted for assessment. Despite this, we
wish to offer some insight into how students performed. All
students successfully completed their optimized purification
Evaluation of Student Performance in under two hours with a couple of students finishing in
The evaluation of student performance on each of the exer- about an hour. The task challenged students, but by deem-
cise questions is provided in Table 1. Provided as Supporting phasizing yield and cost, it was readily achievable. Each
Information (Table S1, pages 5–6), are examples of high student performed many iterations before arriving at an
quality and low quality student answers, to each of the seven optimized protocol. The techniques that students often
exercise questions in Table 1. Effective student explanations, employed were ammonium sulfate fractionation, heat

Clark and Edwards 95

 Biochemistry and
Molecular Biology Education

Postexercise assessment of student learning

TABLE 3

Question:
A certain protein has a pI of 7.5. At a pH of 7.0 this protein Percentage of students
would be expected to: selecting an answer (%)a

(a) Adsorb to an anion exchange column 25

(b) Adsorb to a cation exchange column 75
(CORRECT RESPONSE)
(c) Adsorb to a size exclusion column 0
(d) Adsorb to a hydrophobic interaction column 0
(e) Cannot be predicted without more information 0

a
Percentages calculated using data from all students in the course (n 5 8).

treatment (only for those with proteins stable at 508C or equilibration pH can affect protein binding to and elution
greater), hydrophobic interaction, and size-exclusion. In from, a cation-exchange column (Table 2).
terms of basic strategy, ion exchange, followed by hydro- The postexercise assessment of student learning is
phobic interaction, then gel filtration, proved successful in shown in Table 3. The question was part of a comprehen-
many cases. As mentioned previously, students were not sive final examination given three and a half months after
allowed to use a 2D-gel to find the pI of their target pro- students completed the exercise. The results indicated that
tein, which would have helped guide their purifications. the majority of students (75%) correctly predicted that a
Interestingly, a few students estimated pI indirectly, as they protein with a pI of 7.5 would adsorb to a cation exchange
did with the exercise (see Table 1, Question 7), by examin- column at pH 7.0 (Table 3). To answer this question, stu-
ing their target protein’s pH-dependent behavior on an ion dents would have had to consider the net charge on the
exchange column. These students subsequently used this protein at this pH, and the charge on the chromatography
information to assist the design of specific ion exchange column—both of which were fundamental concepts empha-
steps. Given that many aspects of this open-ended purifica- sized in the exercise. This result aligned with the postexer-
tion were new to students, instructor guidance and interac- cise assessment of student attitude, where the majority of
tion appeared to be essential for success. Yet, such guid- students (87.5%) felt that the exercise gave them the foun-
ance was expected, and it provided an opportunity for dation to understand how equilibration pH can affect pro-
students to learn in lieu of their wet-lab purifications later tein binding to and elution from, a cation-exchange column
in the semester. (see Table 2, Question 4).

Assessment Summary
The postexercise assessment of student attitudes via an This article describes a unique exercise using a free online
anonymous survey, which employed a five-point Likert program. The exercise focuses on anion-exchange chroma-
scale, is shown in Table 2. The sample size (n 5 8) reflected tography and utilizes a simple protein mixture. However,
the small number of students that take this course at CSU the versatility of the program affords the creation of similar
Chico; for example, total enrollment in Fall 2016 and exercises with other chromatographic techniques and more
Spring 2017 semesters were seven and eight students, complex protein samples. Therefore, our work also serves
respectively. Student responses reflected a favorable view as a model for instructors to develop their own custom
of the exercise (Table 2). Question 4 was included to gauge exercise to augment their teaching of protein purification.
student confidence with what they had learned and whe- Such an exercise can be implemented in a lecture course,
ther they thought they could extend their knowledge to as a pre-lab assignment, or as an in-lab prelude to a wet-
cation-exchange chromatography (Table 2). The data for lab or dry-lab (as we have done). We do not advocate for a
Question 4 suggested that most students felt the exercise virtual experience to replace an actual protein purification.
had given them the foundation to understand how Instead, we believe this program significantly enhances any

96 Virtual Protein Purification

wet-lab experience by giving students the opportunity to servlet/catalog/en/GELifeSciences-us/service-and-support/handbooks/ (accessed
March 6, 2017).
explore chromatographic parameters to a degree not possi-
[3] Scopes, R. K. (1994) Protein Purification: Principles and Practice, 3rd ed.,
ble within the time constraints of a normal lab. Springer-Verlag, New York.
[4] Phornphisutthimas, S. Panijpan, B. Wood, E. J., and Booth, A. G. (2007)
Improving Thai students’ understanding of concepts in protein purifica-
References tion by using Thai and English versions of a simulation program. Bio-
[1] Booth, A. G. (1986) Simulation of protein separation techniques on a per- chem. Mol. Biol. Educ. 35, 316–321.
sonal computer. Biochem. Soc. Trans. 14, 908–909. [5] Usher, K. C., and Barrette-Ng, I. H. (2012) Web-based applet is a learning tool
[2] GE Healthcare Life Sciences. Handbook: Ion Exchange Chromatography— that simulates ion exchange chromatography purification of overexpressed
Principles and Methods. http://www.gelifesciences.com/webapp/wcs/stores/ proteins from Escherichia coli cell lysate. J. Chem. Educ. 89, 555–556.

Clark and Edwards 97