Proteomic profiling and analytical chemistry

1
INTRODUCTION
Jerzy Silberring*, † and Pawel Ciborowski‡
*

AGH University of Science and Technology, Krakow, Poland
Centre of Polymer and Carbon Materials, Polish Academy of
Sciences, Zabrze, Poland
z
University of Nebraska Medical Center, Omaha, Nebraska
y

CHAPTER OUTLINE
1.1 Why Analytics Matter? 1
1.2 Expectations: Who and What? 3
1.3 What Is Next and Where Are We Going?

4

1.1 Why Analytics Matter?
The sum of the optimal steps in analytical and
proteomic analysis (process) is not equal to the
optimal process in its entirety! Proteomic studies
are multistep tasks involving a variety of methods,
each governed by its own strengths and limitations. As much as it is a trivial statement, with
which all or at least most of us agree, it has not
been fully appreciated, despite having a profound
impact on the success of laborious, expensive, and,
in many instances, lengthy projects. The concept
of a proteomic study can be depicted in many
ways.

Figure 1.1 shows schematic representation of
a proteomic study. In this model, analytical components/phases are highlighted intentionally because
the same rules of analytical chemistry/biochemistry
apply to discovery, as well as to validation experiments. Experimental design will be governed by a set
of different rules, which does not include instrumentation, but has biology heavily involved. Bioinformatics will be governed by its own set of rules,
Proteomic Profiling and Analytical Chemistry. />Ó 2013 Elsevier B.V. All rights reserved.

1


2

Chapter 1 INTRODUCTION

PROTEOMICS
Experimental
design phase

Analytical phase I

Performance of
instrumentation

Detection level

Your question

Reproducibility
Sample loss

New experiments
New questions
Knowledge

Data analysis
phase

Bioinformatics

Data
Analytical phase II
Validation

Information

Figure 1.1 Schematic representation of a proteomic study.

which are applied to a validation of algorithms.
Nevertheless, looking at constituents of a proteomic
study, it must be realized that the scientist conducting such experiments must grasp the overview of not
only how biological systems work, but also analytical
boundaries for sample preparation, fractionation,
and measurements; tools for database searches;
bioinformatic tools for data analysis; and, of course,
statistics. Because of such complexity, proteomic
studies should be conducted by a team of experts.
Still, a lack of uniformed quality criteria accepted by
all will cause disconnectedness of individual
components in a proteomic study. For instance,
currently there is no consensus on how a “good” twodimensional (2D) gel should look like or how good

(efficient) should, for example, 2D liquid chromatography analysis be. Such criteria exist in analytical/
clinical chemistry, where quality control is an integral part of each analysis on a daily basis, but proteomics falls short in this area. Therefore, this book
attempts to highlight, in a short but comprehensive
manner, the impact of analytical chemistry/
biochemistry basic principles on the final success of
a proteomic experiment. It is hoped that this point of
view will help all, biologists and chemists, in better
understanding of all components of a complex proteomic study.


Chapter 1

1.2 Expectations: Who and What?
If two scientists, a biologist and a chemist, sit at
a table and discuss proteomic methodology, they
likely will emphasize different aspects of the same
study, which, in each viewpoint, is critical for the
outcome. Nevertheless, they very often speak in
technical language that is not fully understood by the
other. This is because chemists are focused on the
sensitivity and accuracy of analytical measurements,
whereas biologists pay attention to explaining biological/pathological effects and are less concerned
with the exact quantitation of analytes. This resembles the famous poem by John G. Saxe “The Blind
Men and the Elephant,” where everyone tries to
identify the part they are touching (i.e., biologist/
chemist) but nobody can get a sense of the whole
system (i.e., proteomic study). Biologists are willing
to accept a high range of responses resulting in high
standard deviations showing or indicating “trends” in
data behavior that support their hypothesis. Chemists, however, expect data to be expressed by

numerical values with high precision, accuracy,
reproducibility, and low standard deviation. Indeed,
as much as the precision of analytical measurements
is important, in many instances such efforts will not
improve the overall output discriminating between
true and false mostly because an exact correlation
between quantitative change and biological effect is
not defined very often. For example, how important
is it to measure a difference between levels of protein
expression above a 10-fold change when the
response of a biological system is already saturated
by the 5-fold change of this protein? A similar question may arise from enzymology, where the most
important factor is enzymatic activitydnot the
protein expression measured by a typical proteomic
approach. If statisticians and bioinformaticians are
brought to the very same table of biologist and
chemist, which happens very often, the discussion
becomes even more complicated. So our question is
what do we see on the other side of the wall of our
office when we look into the office space of our fellow
colleagues with their expertise? We tried to illustrate
this in Figure 1.2. Therefore, it is critical for each of us

INTRODUCTION

3


4

Chapter 1 INTRODUCTION

BIOLOGISTS

Chemists and
Mass Spectrometrists

Biostatisticians and
Bioinformaticians

Figure 1.2 What we see on the other side of the wall of our office when we look into the office
space of our fellow colleagues with their expertise.

to peer outside of our walls that confine us and look
into the world of those who surround us.

1.3 What Is Next and Where Are We
Going?
Since proteomics moved from qualitative to
quantitative profiling using liquid phase-based
methods of sample fractionation, it fully entered
a domain of analytical chemistry. As much as it is very
beneficial for proteomics due to the wide range of
well-established analytical methods, the complexity of
proteomic profiling creates multiple technical issues.
First, classical analytical chemistry focuses on high
accuracy measurements of single or few compounds
at the same time. It allows adjusting methods of
sample preparation and analytical parameters with
a specific objective(s) scarifying measurements of

other compounds that are contaminants rather than
analytes. Importantly, analytical chemistry exploits
specific characteristics of analyzed compounds, and
this concept fulfills its purpose. In contrast, proteomics attempts measurements of hundreds and thousands of molecules at the same time, which can have


Chapter 1

a wide range of chemical characteristics (e.g., posttranslational modifications of proteins and peptides)
and which have a wide dynamic range of concentrations, such as the circumstance with plasma or serum.
One good example is the use of isobaric tags for
relative and absolute quantitation (iTRAQ) as a means
of quantitation, which is, to some extent, separate
from peptide identification. We have observed nicely
quantitated species that otherwise generated very
poorly fragmented spectra, making confident identification nearly impossible. In Figure 1.1, all three steps
of a proteomic study are equally important. It would
have been a trivial statement if we looked at each step
separately. Caveats arise from connection of these
three steps as a “well-oiled logically working
machine.”
In summation, the main goals of this book are to
highlight points of junction between proteomics and
analytical chemistry and to link experimental design
with analytical measurements, data analysis, and
quality control. It also provides “a list” of points to
consider for those who are planning or entering the
field of proteomics and have minimal experience.

INTRODUCTION

5


2
BIOMOLECULES
Pawel Olszowy,1, 2 Ariel Burns1 and
Pawel Ciborowski1
1
2

University of Nebraska Medical Center, Omaha, Nebraska
Nicolaus Copernicus University, Torun, Poland

CHAPTER OUTLINE
2.1. Major Features and Characteristics of Proteins
and Peptides 7
2.2. Hydrophilicity and Hydrophobicity 8
2.3. Effect of Protein Fragmentation 10
2.4. Effect of Post-translational Modifications 14
2.5. Amino Acid Sequence and Separating Conditions 15
2.6. Cysteine and Methionine; Amino Acids Containing
Sulfur 16
2.7. Protein Identification and Characterization 19
2.8. StructureeFunction Relationship and Its Significance
in Systems Biology Function 19
2.9. Protein Folding and ProteineProtein Interactions 21
2.10. Moonlighting of Proteins 22
References 23

2.1 Major Features and
Characteristics of Proteins
and Peptides
Proteins are very, if not the most, diverse naturally
occurring heteropolymers. This is because they
consist of 20 different monomers (amino acids), vary
in length, and carry multiple modifications. Amino
acids, monomers of proteins, and peptides range in
their properties from hydrophilic (acidic or basic) to
hydrophobic in nature. Therefore, a combination of
amino acid composition and combination (sequence)
will have a big impact on the overall physicochemical
Proteomic Profiling and Analytical Chemistry. />Ó 2013 Elsevier B.V. All rights reserved.

7


8

Chapter 2 BIOMOLECULES

properties of proteins or peptides. Moreover, proteins
may have regions of quite opposite hydrophobic/
hydrophilic properties. For example, membrane
proteins have very hydrophobic transmembrane
domains and hydrophilic extra- and intracellular
domains that have to interact with ligands. This being
said, fractionation of proteins, which is an essential
step in any proteomic profiling experiment, is challenging. One approach to making this step easier is
to fragment proteins into short peptides by various

chemical and enzymatic methods. The resulting pool
of peptides will still form a wide spectrum of molecules, ranging from hydrophobic to hydrophilic;
however, each peptide will have a much more narrow
range of chemical characteristics and will be easier
to be separated as a single, narrow peak in liquid
chromatography. Although true in many instances,
many peptides still carry post-translational modifications influencing overall chemical properties.

2.2 Hydrophilicity and
Hydrophobicity
Amino acids range from hydrophilic or hydrophobic depending on their chemical nature of their
side chains. This feature was used by Jack Kyte and
Russell Doolittle who calculated the hydropathy
index1 based on a measurement of how the R group
(side chain) interacts with water. Their calculations
are based on the free energy of transfer (DG trans) of
the solute amino acid between water and condensed
vapor phase. A negative DG trans indicates a strong
preference for the R group to water, whereas a positive value indicates the opposite.
The hydropathy index is applied to proteins as
well. Starting at the N terminus, individual hydropathy indexes are summed over an arbitrary length of
the protein, usually 7, 9, 11, or 13 residues. Then
a sliding window shifts one amino acid and sums the
hydropathy index of those amino acids and will be
continued until the end of the protein. The hydropathic index versus the position of the amino acid
gives a graphic representation of which parts of the
protein have transmembrane domains.


Chapter 2 BIOMOLECULES

It is thermodynamically favorable for water to
minimize the interaction with nonpolar, hydrophobic moieties. This causes nonpolar molecules to
accumulate with each other and form a clathrate
structure. A clathrate structure is a cage-like network
of water surrounding all the hydrophobic interactions of the nonpolar molecules.2 Reversed-phase
chromatography (RPC) is an important tool that uses
hydrophobicity to purify peptides and proteins. RPC
has nonpolar, stationary phases (C-4, C-8, or C-18
consisting of aliphatic chains containing 4, 8, or
18 carbon atoms, respectively) linked covalently to
a silica support and a moderately polar mobile phase.
The ability of the sample to bind to the stationary
phase is proportional to the contact surface area
around the nonpolar stationary phase.
For example, a peptide consisting of only 7 amino
acids has less surface area and hence, less hydrophobic amino acids to come in contact with the
stationary phase than a peptide with 16 amino acids.
These conditions, among others, will determine
which column is best and how good the separation
will be. The KyteeDoolittle analysis1 can be used to
determine the hydrophobicity of peptides and
whether they elute in the beginning or the end of the
chromatography run. Keeping with the 7 amino acid
peptide, every addition of an amino acid will cause
a secondary structure to arise. The secondary structure could hurt the ability to bind to the matrix due to
shielding of the hydrophobic R groups. As the polypeptide chain grows even more, the protein will fold
spontaneously to the most thermodynamically
stable, tertiary structure by confining the most
hydrophobic regions to the interior to minimize the

interaction with water.3 The capacity of a RPC
column to purify a peptide is related to the amount of
surface area that can bind to the nonpolar stationary
phase, as mentioned previously. Because a large
polypeptide or protein has more surface area, a RP
C-18 column is not as efficient in separating.
In peptide sequencing using mass spectrometry
(MS), precursor ions used for consideration have to
be larger than 700 Da (m/z > 700þ1 or 350þ2).
Assuming that the average mass of the amino acid is
z110 Da, peptides to be considered as having

9


10

Chapter 2 BIOMOLECULES

a sequence unique for any given protein have to
consist of six or more amino acids. Peptides of such
length have a limited surface area to interact with the
stationary phase used for separation. The addition of
each amino acid will gradually lead to formation of
a secondary structure(s) and will impact elution time
under any given chromatographic condition. Longer
peptides will tend to bury hydrophobic side chains
and prevent them from interactions with the
stationary phase. As much as this property is
exploited in the separation of peptides of various

lengths, peptides carrying mutations may have quite
different physicochemical properties. The effect will
depend on the amino acid change, its position, and
the overall length of the peptide.

2.3 Effect of Protein Fragmentation
As pointed out earlier, protein fragmentation will
generate a set of peptides. This set of peptides can be
unique if the method used for protein fragmentation
has high enough specificity. This feature was used to
develop protein identification based on peptide
fingerprinting. This method exploits the specificity of
molecular masses of peptide fragments generated by
a specific method. Tables 2.1A and 2.1B show characteristics of peptide sets generated by pepsin and
trypsin digestion of insulin-like growth factor II
(IGF2).
Tables 2.1A and 2.1B show an example of differences of physicochemical properties of peptides
derived from the same IGF2 protein fragmented in
silico by trypsin and pepsin. It is important to note
that trypsin derived set of peptides that are either
acidic with pI points below 4.33 or basic with pI
points above 10.34. If such a digest is fractionated
further based on isoelectric focusing, for example
OFFGEL, we expect them to be on opposite ends of
the fractionation spectrum, and most of these
peptides will be in one or two fractions. In contrast,
complete pepsin digestion will generate only four
peptides suitable for mass spectrometry-based
protein identification due to the length of the
resulting fragments (m/z bigger than 300 for doubly

Table 2.1
A. Composition and properties of fragments from in silico trypsin digestion of IGF2
Fragment
number

Isoelectric
pointa

Hydrophobicityb

Molecular mass
(Da)

Amino acid
residues

Amino acid
sequencec

1
2
3
4
5
6
7
8

4.13
11.01
10.34
10.34
10.34
4.33
4.17
3.67

À3.4
À6.5
À1.1
À4.5
À5.3
4.6
5.8
À4.3

2761
1187
360
174
261
1055
1699
234

24
10
3

1
2
9
16
2

AYRPSETLCGGELVDTLQFVCGDR
GFYFSRPASR
VSR
R
SR
GIVEECCFR
SCDLALLETYCATPAK
SE

B. Composition and properties of fragments from in silico pepsin digestion of IGF2
Fragment
number
1
2
3
4
5
6

Isoelectric
pointa
6.38
5.79
3.67

5.79
3.49
5.79

Hydrophobicityb
À10.6
3.8
À1.8
3.8
0.0
3.8

Molecular mass
(Da)
823
131
364
131
333
131

Amino acid
residues
7
1
4
1
3
1

Amino acid
sequencec
AYRPSET
L
CGGE
L
VDT
L
Continued


Table 2.1 (continued)
B. Composition and properties of fragments from in silico pepsin digestion of IGF2
Fragment
number

Isoelectric
pointa

Hydrophobicityb

Molecular mass
(Da)

Amino acid
residues

Amino acid
sequencec

7
8
9
10
11
12
13
14
15
16
17
18
19

5.79
5.95
5.79
5.77
5.79
10.22
5.79
5.95
5.79
5.79
5.79
5.79
4.33

À3.5
2.8

À2.1
1.5
2.8
À15.0
2.8
À6.3
3.8
1.8
3.8
3.8
À9.9

146
165
606
328
165
2148
165
480
131
89
131
131
1199

1
1
6
2

1
19
1
4
1
1
1
1
11

Q
F
VCGDRG
FY
F
SRPASRVSRRSRGIVEECC
F
RSCD
L
A
L
L
ETYCATPAKSE


C. Composition and properties of fragments from pepsin digestion of IGF24
Fragment
number

Isoelectric

pointa

Hydrophobicityb

Molecular mass
(Da)

Amino acid
residues

Amino acid
sequencec

1
2-7
2A-7

6.38
3.93
3.93

À6.8
0.0
3.9

936
1436
1650

8

13
15

2:7:
2A:7:

2B-7

3.93

3.8

1549

14

2B:7:

3
3A
4-10

3.49
3.49
6.09

0.3
3.8
À2.2

575
446
1867

5
4
17

4:10:

4-10A

4.50

À6.4

2097

19

4:10A:

4B-10

6.09

À2.1

2030

18

4B:10:

12.78
10.34
3.67
5.79
3.67

À18.9
À3.1
5.2
9.4
À0.4

1619
664
359
315
361

14
6
3
3
3

5
5A

6
8
9
a

Average pI calculated from a computer program based on Kozlowski 2007e2011 ( />Calculated according to Kyte and Dolittle.1
Single-letter code for amino acids used.

b
c

AYRPSETL
CGGELECCFRSCD
TLCGGEL
ECCFRSCD
LCGGEL
ECCFRSCD
VDTLQ
VDTL
FVCGDRGF
YCATPAKSE
FVCGDRGF
ETYCATPAKSE
FVCGDRGFY
YCATPAKSE
FSRPASRVSRRSRG
FSRPAS
IVE
LAL
LET

14

Chapter 2 BIOMOLECULES

charged species). However, two of these peptides will
be located in the middle of the OFFGEL spectrum
when a 3.0 to 10.0 immobilized pH gradient strip is
used. Trypsin digestion will generate four peptides
with m/z bigger than 300 for doubly charged species,
which will be fragmented for MS/MS identification
and, if tagged, also for quantitation. Table 2.1C shows
an example of observed peptides generated by pepsin
digestion of IGF2. In this case, due to miscleavages,
pepsin digestion generated eight such peptides.
Considering the contribution of each peptide to a high
confidence of identification and quantitation, in this
particular case the pepsin digest will have an analytical advantage over the tryptic digest. Another issue is
that the selection of peptides for the multiple reaction
monitoring (MRM) type of experiment will be reduced
when IGF2 is fragmented using trypsin than pepsin.
Therefore, depending on the focus of the proteomic
experiment, the choice of proteolytic enzyme or other
means of peptide fragmentation may have an effect on
the accuracy of quantitation. It also needs to be noted
that in silico digestion using tools such as a PeptideCutter ( are of
great help but do not reflect the real effect of proteolytic digestions.

2.4 Effect of Post-translational

Modifications
Chemical modifications have an impact on the
overall chemical property of proteins and peptides.
Single site acetylation of 50-kDa or larger proteins
may not be detectable by many analytical methods
because an increase of hydrophobicity of acetylated
versus nonacetylated forms can be negligible if the
protein is by itself quite hydrophobic in nature. The
situation changes significantly when such a protein is
digested enzymatically for MudPIT proteomic
profiling. In this situation, acetylation may be located
on a relatively short peptide, for example, 8 to 10
amino acids, and have a profound impact on the
overall hydrophobicity of this molecule, leading to
a shift in elution time in reversed-phase liquid
chromatography (RP-LC).


Chapter 2

Physicochemical properties of proteins and
peptides are complicated further when multiple
residues on one protein or a longer peptide are
modified and, in extreme cases, modifications are
heterogeneous.

2.5 Amino Acid Sequence and
Separating Conditions
There is no “one-size-fits-all” solution in protein
and peptides analyses and chromatographic separation of peptides. It will also depend on how we match

structure and properties of peptides of interest with
characteristics of separation media. This is a very
important and quite often neglected issue when all
details of a proteomic profiling experiment(s) need to
be considered.
In most instances, very standard supplies, prepacked columns or bulk resins, are used, such as an
RP-LC column, without analyzing what type of resin is
used in any given product. It is more evident now than
10 years ago that the success of proteomic profiling
requires narrowing the scope of investigation to
improve sensitivity and specificity. One example is
immunodepletion of the most abundant proteins
from plasma/serum or cerebrospinal fluid samples to
reduce the dynamic range of protein concentrations,
thus reducing the dynamic range of the concentration
of peptides if such a sample is first digested enzymatically and then subjected to first dimension
separation. The example provided in Figure 2.1 shows
selectivity comparison between different silica-based
media at pH 2.0 and pH 6.5. In this case, a mixture of
closely related angiotensin peptides was used as the
sample. Peptides 1, 2, and 3 are different in one amino
acid and their sequences are as follow: 1RVYVHPI,
2
RVYIHPI, and 3RVYVHPF, respectively. While the
third peptide has a distinct value of mean hydrophobicity (0.08) compared to the first and second peptides
of 0.32 and 0.37, respectively, peptides 2 and 3 can be
coeluted (Figure 2.1a) or eluted separately
(Figure 2.1h). However, if peptides 1, 2, and 3 are to be
eluted separately, peptides 5 and 6 and peptides 7 and
8 are coeluted (Figure 2.1). This pattern would depend

BIOMOLECULES

15


16

Chapter 2 BIOMOLECULES

µRPC C2/C18

Sephasil Peptide C18

Sephasil Peptide C18

7+8
5+6

pH 2

1

4
2

0

7+8

pH 6.5

pH 2
3
2
1

65

6
43 8
5
12
7

4

3

25

0

25

0

25

Time (min.)

Figure 2.1 Separation of various peptides: 1. Val4-lle7-AT III (RVYVHPI), 2. Ile7-AT III (RVYIHPI),
3. Val4-AT III (RVYVHPF), 4. Sar1-Leuß-AT II (Sar-RVYIHPL) (Sar=sarcosine, N-methylglycine),
5. AT III (RVYIHPF), 6. AT II (DRVYIHPF), 7. des-Asp1-AT I (RVYIHPLFHL), 8. AT I (DRVYIHPFHL)
using different silica based media and/or pH. A mixture of closely related angiotensin peptides
was used as sample. (Based on “Reverse Phase Chromatography, Principles and Methods”,
Brochure by Amersham Biosciences 18-1134-16, Edition AA, 1999).

on medium type and pH of mobile phase used. Such
a dual factor effect on peptide separation can be
exploited with great benefits if the project is more
focused in addressing much more specific questions
than full-range unbiased proteomic profiling.
Similarly, in gel-based proteomic profiling, application of different conditions, such as percentage
of the gel and buffer system used, may favor separation in specific regions of molecular weight. Application of various conditions is much broader for
one-dimensional gel electrophoresis than in twodimensional gel electrophoresis.

2.6 Cysteine and Methionine; Amino
Acids Containing Sulfur
Cysteine and methionine are two amino acids that
contain sulfur. While methionine is an essential amino
acid, cysteine is synthesized from methionine, thus it
is nonessential. Cysteine is classified as a polar, noncharged amino acid, whereas the side chain of
methionine is quite hydrophobic. The chemical
linkage of sulfur in methionine is thiol ether. The
hydropathy index of methionine and cysteine


Chapter 2

according to the Kyte and Doolittle scale1 is positive
and equals 1.9 and 2.5, respectively. Unlike cysteine,
the sulfur of methionine is not highly nucleophilic,
although it will react with some electrophilic centers.
It is generally not a participant in the covalent
chemistry that occurs in the active centers of enzymes.
The thiolate anion, which is formed after the ionization of cysteine in basic solutions, does not change the
character of this amino acid. Therefore, it is very
uncommon to find cysteine on the surface of a protein
even after ionization. The sulfur of methionine, as
with that of cysteine, is prone to oxidation; therefore,
oxidated methionine is added to database searches of
tandem mass spectra. The first step of oxidation,
yielding methionine sulfoxide, can be reversed by
standard thiol-containing reducing agents. The
second step yields methionine sulfone and is effectively irreversible. When oxidized, cysteine residues
can form disulfide bonds, strengthening protein
tertiary and quaternary structures. Additionally, many
metal-containing proteins use cysteines to hold their
metals in place, as the sulfhydryl side chain is a strong
metal binder. There are a few reasons why sulfur
atoms in amino acids do not affect the position of
those amino acids in proteins. One of the most
important aspects is the strong ability to create
disulfide bonds in comparison with the creation of the
weakest, noncovalent hydrogen bond with water.
However, the weakest ability to attract electrons (in
comparison to oxygen) results in a lack of hydrogen
bonds using sulfur atoms.
Cysteine stabilizes the tridimensional structure of

proteins, which is critical for extracellular proteins that
may be exposed to harsh conditions. Because proteins
containing multiple disulfide bridges are more resistant to thermal denaturation, they may maintain their
biological activity at more extreme conditions.
The existence of a disulfide bridge inside a protein
(intramolecular) and/or between different proteins
(intermolecular) makes it necessary to break those
bonds before proteomic analysis. The standard
approach is a two-step procedure that is almost
always applied to prepare proteins samples for proteomic analysis. In the first step, proteins are reduced
using dithiothreitol (C4H10O2S2) or mercaptoethanol,

BIOMOLECULES

17


18

Chapter 2 BIOMOLECULES

although the latter agent is now used rather seldom.
In this step, disulfide bridges break, yielding free
sulfhydryl groups. In the following second step, free
sulfhydryl groups are alkylated to prevent reoxidation
and formation of bridges.
The chemical feature of cysteinedisotopic-coded
affinity tag (ICAT)dhas also been utilized in a gelfree MS-based technique for quantitative comparisons of up to two samples. This approach uses
a chemical reagent consisting of a thiol-reactive
group (labeling cysteines), linker and acid cleavable

biotin moiety (for affinity-based purification) as
presented in Figure 2.2.5,6 Quantification can be
performed using either carbon or hydrogen isotope
labeling. In case of carbon labeling, isotopic linkers
contain nine carbon isotopes 13C (heavy tag) and
nine carbon isotopes 12C (light tag). ICAT reagents
using labeled hydrogen atoms contain eight
hydrogen isotopes 2H (heavy tag) and eight hydrogen
isotopes 1H (light tag). Application of incorporated
13
C rather than 2H allows increase accuracy and
precision of quantification based on mass spectrometry using both electrospray ionization and
matrix-assisted laser desorption ionization techniques. A limitation of the ICAT technique is possible
quantification of only cysteine-containing proteins.
The biological importance of sulfur-containing
amino acids is multifold. Methionine is necessary for
the synthesis of proteins. It forms S-adenosyl-Lmethionine, which serves at a methyl donor in reactions, prevents fatty liver through transmethylation
and choline formation, and can lower toxic acetaldehyde levels in humans after alcohol ingestion. It also
plays an important role in preserving the structure of
O
HN

NH

O

X
NH

S

X

X

X

NH

O

O

X

X
Biotin

O
X

O

Linker

X
Thiol specific
reactive group

Figure 2.2 Schematic representation of ICAT reagent. X could be either hydrogen (light reagent)

or deuterium (heavy reagent). Eight 2H atoms could be used, making an 8-Da difference in a singly
charged or a 4-Da difference in a doubly charged fragment.


Chapter 2

cell membranes 7 and it has an important function for
some reactions involved in protein and DNA
synthesis.8 Cysteine is found in b-keratin, an important
component of skin, hair, and nails. A huge number of
disulfide bonds causing keratin can be very hard, such
as in nails or teeth, or flexible, such as in hair. The
smallest number of disulfide bonds creates soft keratin
in skin. The human body uses cysteine to produce the
antioxidant glutathione, as well as the amino acid
taurine. The body can also convert cysteine into
glucose for a source of energy. Cysteine also plays a role
in the communication between immune system cells.

2.7 Protein Identification and
Characterization
High confidence protein identification and indepth characterization in one proteomic experiment
is the most favorable goal. Although new tools have
been developed during the last decade, the inherent
properties of proteins and peptides create limitations
of how much information can be obtained. For
example, using one enzyme for protein fragmentation
will generate peptides that can be too short or too
long. For a protein with high confidence, two or three
peptides are usually sufficient; however, it may not be

enough for characterization and/or identification of
specific regions of a protein. One example is histones.
These proteins contain multiple lysine residues in one
string and can be highly methylated and/or acetylated. It is an analytically challenging task to identify
the exact position of methylation or acetylation.
Therefore, protein characterization usually requires
more than one analytical approach, which will require
more biological material not always abundantly
available.

2.8 StructureeFunction Relationship
and Its Significance in Systems
Biology Function
The major goal of proteomic profiling experiments is to get an insight into how the complex

BIOMOLECULES

19


20

Chapter 2 BIOMOLECULES

biological system works; therefore, the most desirable outcome is new functional information. When
proteomics was born in the mid-1990s, everybody
was fascinated with the ability to identify (catalog)
tens, hundreds, and then thousands of proteins
in one analytical experiment. This did not last long,
as we realized that answers are in relative quantitation rather than the presence or absence of

a particular protein. At this point we hit yet another
wall, which was post-translational modifications,
which increased the complexity of proteomic
experiments by at least two orders of magnitude.
New experimental approaches have been proposed
and collectively great progress has been made in
accumulating huge amounts of data. Although
significant steps in the biological interpretation of
such massive data have been made, our knowledge
about how biological systems function is growing at
a disproportionally low rate. Two hurdles in progress here are correlation of protein structure and
function and protein localization and function. The
latter phenomenon is also called protein moonlighting. This brings us to question what a protein
structure represents in defining its biological function and further on how the structure of a protein
defines its physiological function.
What if we assume that similar sequences of
proteins represent similar functions, whereas
different sequences are responsible for different
functions? We will certainly find many examples to
support such assumptions. Let us consider transmembrane domains of receptors that are hydrophobic and have a helical structure to be
accommodated by a hydrophobic environment of
a lipid bilayer. Furthermore, integrins a 1, 2, and 4
have single-pass transmembrane helical domains
that all play the same function: anchoring these
proteins into the cell membrane. They are all close to
the C-terminal end of the polypeptide chain;
however, all of them have a different primary structure (Figure 2.3).
As we know, integrins are responsible for transmitting signals related to numerous functions and
are part of a/b heterodimers.

Chapter 2

BIOMOLECULES

21

Integrin alpha 1
1131 ISKDGLPGRVPLWVILLSAFAGLLLLMLLILALWKIGFFKRPLKKKMEK-COOH 1179
Integrin alpha 2
1121 IMKPDEKAEVPTGVIIGSIIAGILLLLALVAILWKLGFFKRKYEKMTKNPDEIDETTELSSCOOH 1181
Integrin alpha 4
971 RPKRYFTIVIISSSLLLGLIVLLLISYVMWKAGFFKRQYKSILQEENRRDSWSYINSKSNDDCOOH 1132

Figure 2.3 Amino acid sequences of transmembrane domains of integrins a 1, 2, and 4.

2.9 Protein Folding and
ProteineProtein Interactions
Proteins fold to reach their conformation associated with function. The process of protein folding is
not fully understood; however, we know that most
proteins are folded during or right after synthesis.
Many proteins, although folded properly, need
further processing and help from chaperons to reach
their final functional structure. Many proteins are
maintained unfolded by chaperons as otherwise they
could not be transported outside of the cell. For
example, Escherichia coli developed a specialized Sectranslocase system for post-translational translocation
of proteins.9,10 This system is a complex of the ATPdriven motor protein SecA and the SecYEG protein
functioning as a membrane-embedded translocation
channel. One of the features of this system is that only

unfolded proteins can be translocated, thus they have to
be kept in a translocation-competent state. SecB holdase, which is an export-dedicated molecular chaperon,
prevents proteins to be translocated from folding and
aggregating. Summarizing, if we extract all proteins
from a cell, denaturate, fragment using, for example,
trypsin, and quantitate based on resulting peptides, we
are unable to conclude whether the protein was
unfolded and complexed with a chaperon and will
contribute to the active pool outside of the cell or was
folded and never destined to be exported. Even if we


22

Chapter 2 BIOMOLECULES

measure the stoichiometric ratio of the chaperon to
protein, we do not have evidence of their function and
quantitation gives us limited information.
Another example is the presence of flexible regions
of proteins, which may lead to conformational changes
upon self-interactions forming homopolymers or
upon interactions with other proteins.
Proteineprotein interaction may be mediated by
an induced-folding mechanism. This mechanism has
been proposed for disabling the intrinsic antiviral
cellular defense mechanism by HIV-1 Vif protein.11
Vif neutralizes two components of a human antiviral
defense mechanism, APOBEC3G and APOBEC3F, by
engaging them with the cellular protein complex of

EloB, EloC, Cul5, and Rbx2 to promote degradation
via the proteasomal pathway. In this example,
participation of Vif in such a complex determines one
of its many functions.

2.10 Moonlighting of Proteins
Protein moonlighting is a phenomenon acquired
during the evolutionary process when a single
protein performs more than one function, which is
also associated with specific localization for a specific
function. This phenomenon was described for the
first time by Joram Piatigorsky and Graeme Wistow in
the late 1980s12 but gained more attention after
nailing down this term by Constance Jeffery in
1999.13 The first proteins shown to moonlight were
crystalline and other enzymes14; later proteins, such
as receptors, ion channels, chaperons, or structural
proteins,15 expanded this list.
Due to the lack of a systematic experimental
approach, moonlighting properties of proteins have
been found as a result of other studies that did not
directly target the dual functionality of the proteins of
interest. Nevertheless, the number of such a class of
proteins is increasing rapidly, showing that moonlighting proteins appear to be abundant in all kingdoms of life.16 We may speculate that the list of such
proteins is not complete and that future studies will
add more to the list. Moonlighting phenomenon may
also contribute to various diseases. Therefore, while


Chapter 2

interpreting results of proteomic studies, particularly
when the objective of such studies is to connect
changes in expression levels with a function(s) having
a biological effect, protein moonlighting needs to be
consider.
If a protein binds other molecules, whether small
molecules, carbohydrates, or other proteins, it may
acquire a new function that can also be associated with
a different localization. It has to be determined whether
or not such a property falls under the moonlighting
phenomenon. It becomes more complicated when the
pool of relatively abundant extracellular proteins
circulating in body fluid is considered. Proteins circulating as complexes with antibodies may not be quantitated properly using an ELISA assay, and MRM-based
quantitation after proteolytic fragmentation may give
different concentrations. Very often extracellular
proteins are considered a homogeneous population of
molecules; in fact, they may represent an array of
functionally different subsets. It is also possible that
only one subset might be relevant as a biomarker,
whether diagnostic or reflecting molecular mechanisms underlying a pathological state.

References
1. Kyte J, Doolittle RF. A simple method for displaying the
hydropathic character of a protein. J Mol Biol.
1982;157(1):105-132. Epub 1982/05/05.
2. Biswas KM, DeVido DR, Dorsey JG. Evaluation of methods for
measuring amino acid hydrophobicities and interactions. J
Chromatogr A. 2003;1000(1-2):637-655. Epub 2003/07/25.
3. Cserhati T, Szogyi M. Role of hydrophobic and hydrophilic

forces in peptide-protein interaction: New advances. Peptides.
1995;16(1):165-173. Epub 1995/01/01.
4. Rickard EC, Strohl MM, Nielsen RG. Correlation of
electrophoretic mobilities from capillary electrophoresis with
physicochemical properties of proteins and peptides. Anal
Biochem. 1991;197(1):197-207. Epub 1991/08/15.
5. Dunkley TP, Dupree P, Watson RB, Lilley KS. The use of
isotope-coded affinity tags (ICAT) to study organelle
proteomes in Arabidopsis thaliana. Biochem Soc Trans.
2004;32(Pt3):520-523. Epub 2004/05/26.
6. Yi EC, Li XJ, Cooke K, Lee H, Raught B, Page A, et al. Increased
quantitative proteome coverage with (13)C/(12)C-based, acidcleavable isotope-coded affinity tag reagent and modified data
acquisition scheme. Proteomics. 2005;5(2):380-387. Epub
2005/01/14.

BIOMOLECULES

23


24

Chapter 2 BIOMOLECULES

7. Vara E, Arias-Diaz J, Villa N, Hernandez J, Garcia C, Ortiz P,
et al. Beneficial effect of S-adenosylmethionine during both
cold storage and cryopreservation of isolated hepatocytes.
Cryobiology. 1995;32(5):422-427. Epub 1995/10/01.
8. Ahmed HH, El-Aziem SH, Abdel-Wahhab MA. Potential role of
cysteine and methionine in the protection against hormonal

imbalance and mutagenicity induced by furazolidone in
female rats. Toxicology. 2008;243(1-2):31-42. Epub 2007/10/30.
9. Bechtluft P, Kedrov A, Slotboom DJ, Nouwen N, Tans SJ,
Driessen AJ. Tight hydrophobic contacts with the SecB
chaperone prevent folding of substrate proteins. Biochemistry.
2010;49(11):2380-2388.
10. Driessen AJ, Nouwen N. Protein translocation across the
bacterial cytoplasmic membrane. Annu Rev Biochem.
2008;77:643-667.
11. Bergeron JR, Huthoff H, Veselkov DA, Beavil RL, Simpson PJ,
Matthews SJ, et al. The SOCS-box of HIV-1 Vif interacts with
ElonginBC by induced-folding to recruit its Cul5-containing
ubiquitin ligase complex. PLoS Pathogen. 2010;6(6):
e1000925.
12. Wistow GJ, Piatigorsky J. Lens crystallins: The evolution and
expression of proteins for a highly specialized tissue. Annu Rev
Biochem. 1988;57:479-504. Epub 1988/01/01.
13. Jeffery CJ. Moonlighting proteins. Trends Biochem Sci.
1999;24(1):8-11. Epub 1999/03/24.
14. Chen JW, Dodia C, Feinstein SI, Jain MK, Fisher AB. 1-Cys
peroxiredoxin, a bifunctional enzyme with glutathione
peroxidase and phospholipase A2 activities. J Biol Chem.
2000;275(37):28421-28427. Epub 2000/07/14.
15. Kourmouli N, Dialynas G, Petraki C, Pyrpasopoulou A,
Singh PB, Georgatos SD, et al. Binding of heterochromatin
protein 1 to the nuclear envelope is regulated by a soluble
form of tubulin. J Biol Chem. 2001;276(16):13007-13014. Epub
2001/03/30.
16. Huberts DH, van der Klei IJ. Moonlighting proteins: An
intriguing mode of multitasking. Biochim Biophys Acta.

2010;1803(4):520-525. Epub 2010/02/11.


3
FUNDAMENTAL
STRATEGIES OF PROTEIN
AND PEPTIDE SAMPLE
PREPARATION
Anna Bodzon-Kułakowska,* Anna Drabik,*
Przemyslaw Mielczarek,* Filip Sucharski,*
Marek Smoluch,* Piotr Suder* and
Jerzy Silberring*, †
*

AGH University of Science and Technology, Krakow, Poland
Centre of Polymer and Carbon Materials, Polish Academy of
Sciences, Zabrze, Poland

y

CHAPTER OUTLINE
3.1. GENERAL STRATEGIES FOR PROTEOMIC SAMPLE
FRACTIONATION 27
3.1.1 Introduction 27
3.1.2 Inhibition of Protease Activity 28
3.1.3 Homogenization 29
3.1.4 Cells as Source of Biological Material for
Proteomics 30
3.1.5 Subcellular Compartments: Organellar
Proteomics 34

3.1.6 Crude Protein Extract: What Is the Next Step? 36
3.1.7 Fractionation Based on Size-Exclusion Filters 38
3.1.8 Chromatographic Methods of Protein
Fractionation 39
3.1.9 Peptide Purification 41
3.1.10 Summary 43
Acknowledgments 44
References 44

Proteomic Profiling and Analytical Chemistry. />Ó 2013 Elsevier B.V. All rights reserved.

25


26

Chapter 3 STRATEGIES OF PROTEIN AND PEPTIDE SAMPLE PREPARATION

3.2. CAPILLARY COLUMNS FOR PROTEOMIC
ANALYSES 46
3.2.1 Introduction 46
3.2.2 Conventional Capillary Columns 47
3.2.3 Monolithic Columns 48
3.2.3.1 Silica-Based Monolithic Columns 49
3.2.3.2 Organic-Based Monolithic Columns 51
3.2.3.3 Methacrylate-Based Monolithic Columns
3.2.3.4 Styrene-Based Monolithic Columns 52
3.2.4 Summary and Conclusions 54
References 56
3.3.

3.3.1
3.3.2
3.3.3
3.3.4
3.3.5
3.3.6
3.3.7

52

ION-EXCHANGE CHROMATOGRAPHY 58
Historical Perspective 58
Principle of Ion-Exchange Chromatography 58
Common Types of IEC Stationary Phases 60
Choice of Ion Exchanger (Cation or Anion?) 63
Choice of Strong or Weak Ion Exchanger 64
Buffers in IEC 65
Ion-Exchange Chromatography in Proteomic
Studies 66
References 68

3.4. PROTEIN AND PEPTIDE SEPARATION BASED ON
ISOELECTRIC POINT 69
3.4.1 Principles of Isoelectric Focusing (IEF) 69
3.4.2 Sample Preparation Prior to IEF 72
3.4.3 Isoelectric Focusing in Liquid State 73
3.4.4 Immobilized pH gradient IEF 74
3.4.5 Capillary IEF (CIEF) 75
3.4.6 Isoelectric Focusing in Living Organisms 76
3.4.6 Summary 76

References 77