Tải bản đầy đủ - 0 (trang)
10 Detergents, Lipids, and Final Remarks

10 Detergents, Lipids, and Final Remarks

Tải bản đầy đủ - 0trang

90



Chapter 4 PROTEIN EXTRACTION AND PRECIPITATION



6. Cohn EJ, Conant JB. The molecular weights of proteins in

phenol. Proc Natl Acad Sci USA. 1926;12(7):433-438.

7. Ferro M, Seigneurin-Berny D, Rolland N, Chapel A, Salvi D,

Garin J, et al. Organic solvent extraction as a versatile

procedure to identify hydrophobic chloroplast membrane

proteins. Electrophoresis. 2000;21(16):3517-3526.

8. Zellner M, Winkler W, Hayden H, Diestinger M, Eliasen M,

Gesslbauer B, et al. Quantitative validation of different protein

precipitation methods in proteome analysis of blood platelets.

Electrophoresis. 2005;26(12):2481-2489.

9. van Oss CJ. On the mechanism of the cold ethanol

precipitation method of plasma protein fractionation.

J Protein Chem. 1989;8(5):661-668.

10. Sivaraman T, Kumar TK, Jayaraman G, Yu C. The mechanism

of 2,2,2-trichloroacetic acid-induced protein precipitation.

J Protein Chem. 1997;16(4):291-297.

11. Xu Z, Xie Q, Zhou HM. Trichloroacetic acid-induced molten

globule state of aminoacylase from pig kidney. J Protein Chem.

2003;22(7-8):669-675.

12. Bensadoun A, Weinstein D. Assay of proteins in the presence

of interfering materials. Anal Biochem. 1976;70(1):241-250.

13. Rey M, Mrazek H, Pompach P, Novak P, Pelosi L, Brandolin G,

et al. Effective removal of nonionic detergents in protein mass

spectrometry, hydrogen/deuterium exchange, and

proteomics. Anal Chem. 2010;82(12):5107-5116.



5

IMMUNOAFFINITY

DEPLETION OF HIGHABUNDANT PROTEINS

FOR PROTEOMIC SAMPLE

PREPARATION

Pawel Ciborowski

University of Nebraska Medical Center, Omaha, Nebraska



CHAPTER OUTLINE

5.1 Capacity of Immunodepletion Columns and Other

Devices 93

5.2 Reproducibility 95

5.3 Quality Control of Immunodepletion 96

5.4 Albuminome 97

5.5 Summary 104

References 104



The HUPO Plasma Proteome Project (HPPP) 2005

multicenter study reported that mass spectrometry

(MS)-MS data sets from all participating laboratories

identified 15,710 proteins based on the International

Protein Index (IPI). After an integration algorithm

was applied to multiple matches of peptide

sequences, this data set yielded 9504 proteins based

on IPI and identified with one or more peptides. Of

these 9504 proteins, 3020 protein were identified with

two or more peptides and constituted the core data

set.1 In 2003, Anderson and Anderson published

a comprehensive overview of the plasma proteome

showing a 1012 range of protein concentrations in

plasma with hemoglobin and albumin as the most

Proteomic Profiling and Analytical Chemistry. http://dx.doi.org/10.1016/B978-0-444-59378-8.00005-0

Ó 2013 Elsevier B.V. All rights reserved.



91



92



Chapter 5 IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



abundant and interleukins as the least abundant

proteins.2 Taking together these two reports and the

fact that there is no platform able to analyze proteins

(peptides) with such a large range of concentrations,

the need to remove the most abundant proteins

became obvious. This necessity prompted the question of how many and which proteins should be

removed to make the concentration range narrow

enough for proteomic analysis (profiling) to

successfully measure and quantitate all remaining

proteins. It was obvious that plasma must be free of

hemoglobin and should be depleted from albumin.

This led to the development of a number of liquid

chromatography (LC) columns or spin devices to

remove the most abundant proteinsdall based on

affinity or immunoaffinity principles. Aurum Affi-Gel

Blue and Aurum serum protein minikits and columns

(Bio-Rad, Inc.) are used to remove albumin or

albumin and IgG, respectively, by affinity chromatography. These methods, starting with two most

abundant proteinsdalbumin and IgGdas well as

spin and high-capacity LC columns, evolved over

time and now are able to remove many proteins as

well as have increasing capacity. In 2005, Bjorhal and

coauthors performed a systematic and formal

comparison of available devices: Aurum serum

protein minikit from Bio-Rad, ProteoExtract

albumin/IgG removal kit from Merck Biosciences,

multiple affinity removal system from Agilent Technologies, POROS affinity depletion cartridges (antiHSA and protein G) from Applied Biosystems, and

albumin and IgG removal kit from Amersham

Biosciences.3 The authors concluded that a polyclonal antibody-based depletion of the six most

abundant proteins removed up to z87% of the total

protein content in serum, reducing the number of

proteins from 3020 to 3014, with the efficiency of

albumin removal being 99.4%. Subsequently, new

columns were developed to remove 12, 14, 20, and

more proteins and the IgY antibody was employed.

The SuperMix column developed by GeneWay, Inc.

aimed to immunodeplete the 81 most abundant

proteins.

The 12 most abundant proteins included in Seppro IgY12 (Sigma-Aldrich, Inc.) are albumin, IgG,



Chapter 5



IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



fibrinogen, transferrin, IgA, IgM, apolipoprotein A-I

and -II, haptoglobin, a1-antitrypsin, a1-acid glycoprotein, and a2-macroglobulin. Two additional

proteins included in the Seppro IgY14 column are

apolipoprotein B and complement C3. Specific

removal of these 14 proteins depletes z95% of the

total protein mass from human serum, plasma, or

cerebrospinal fluid (CSF). Twenty-two most abundant proteins constitute z99% of the total mass of

proteins, thus, based on HPPP, leaving at least 2998

proteins in the remaining 1%.4 The SuperMix system

(Sigma-Aldrich, Inc.) was developed by immunizing

chickens with a flow-through fraction of IgY-12 or

IgY-14 column and constructing the column with

affinity-purified IgY antibodies against the flowthrough proteins of IgY-12 or IgY-14. The goal of the

SuperMix system is to remove medium abundant

proteins from plasma/serum/CSF samples that were

already immunodepleted using IgY-12 or IgY-14

columns. It is important to note that because antibodies used to make the SuperMix system are not

fully standardized, the immune response of the

particular chicken being immunized may vary.

Subsequently, variability is introduced by sample

preparation, which may mask differences resulting

from factors such as treatment and disease

development.



5.1 Capacity of Immunodepletion

Columns and Other Devices

The reference range for total protein in human

plasma is 60e85 g/liter. The term “reference range”

in this case is a value used to interpret medical tests

in clinical biochemistry. As much as this broad range

can be considered normal and used in clinical practice, it has significant consequences for constructing

immunodepletion columns or other devices and

their quality control. In most cases, the capacity of

immunodepletion columns is given by the manufacturer in microliters of serum/plasma. For

example, the capacity of the Seppro IgY-14 liquid

chromatography 5 (LC5) column is 100 ml of normal

human serum or plasma. It can only be assumed,



93



94



Chapter 5 IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



based on an average 60- to 85-g/liter range, that this

100 ml equals 6.0e8.5 mg of protein. If a sample from

any given patient has less protein due to an ongoing

disease, the column will not be overloaded; however,

if the patient has a higher level of proteins, for

example, in paraproteinemia, Hodgkin’s lymphoma,

or leukemia, less volume of sample needs to be used

or it is assumed again that the tolerance of the

protein capacity of such a column is broad enough

that more than 8.5 mg of protein can be loaded and

immunodepletion will be based completely on efficiencies provided by the manufacturer. Nevertheless,

there is not a fast and easy method to test whether

the sample was immunodepleted properly. The

situation is even more complicated when we try to

immunodeplete cerebrospinal fluid. CSF contains

10 to 100 times less protein mass than serum/plasma,

which is yet another broad range. Eventually we end

up with a methodology of immunodepletion based

on a wide range of concentrations of total protein.

Therefore, an amount of microliters of serum/

plasma/CSF to be loaded onto the immunodepletion

column needs to be assessed conservatively, protein

concentration is not measured, and only volume is

used as a measure of quantity.

The immunodepletion method of removal of most

abundant proteins has several limitations. Columns

are relatively expensive and after 100 samples cannot

be regenerated anymore. In addition, immunodepletion of many samples is time and labor-intensive.

Spin column devices are less expensive but can be

used for small volumes of plasma/serum/CSF, as well

as a limited number of samples. Therefore, this

approach is not very suitable for high-throughput

studies with volumes of samples exceeding 100 ml. An

alternative technique was proposed by Kovac and coworkers5 and is based in Blue Sepharose 6 Fast Flow

affinity chromatography using an XK 26/40 column

in the AKTA liquid chromatography system to

immunodeplete albumin from 500 ml of human

plasma. The authors reported that based on SDSPAGE analysis, the majority of albumin was removed;

however, the majority of depleted albumin contained

albumin-associated proteins and proteins showing

affinity to Blue Sepharose. This approach presents



Chapter 5



IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



a compromise between price and throughput and

yield of albumin removal; however, analytical

parameters allowing comparisons between experiments were not presented by the authors. SDS-PAGE

separation followed by any type of protein staining

might be good visual measure; however, the pure

analytical value of gel-based densitometry has many

limitations and usually a low level of precision. This

fundamental analytical aspect of immunodepletion

has a profound effect on any type of downstream

quantitation.



5.2 Reproducibility

The reproducibility of each step in a multistep

proteomic profiling experiment is critical and is

associated with a variability of all parameters.

However, we need to keep in mind that there are two

major sources of variability in proteomics studies:

technical and biological. The impact of technical

variability has decreased in the last decade due to the

development of standardized protocols (kits),

robotics, and autosamplers, as well as an overall

improvement in the quality of instrumentation,

supplies, and reagents. In chromatographic resin

(packing) used to make columns or devices for

immunodepletion, the antibodies are oriented on the

surface of solid beads and cross-linked chemically via

the Fc region; as a result, the Fab interacting regions

are exposed. Covalent cross-linking also prevents

leaching; nevertheless, such resins need to be used

with great care and protected with sodium azide if

not used daily. Based on our experience,6e10 washing

the column with sodium azide every 15 to 20 cycles is

a good practice and extends the lifetime of a column.

At the same time, not much, if anything, can really

be done to reduce biological variability, which

increases with an increased complexity of organisms

and/or biological systems under investigation. One

way to offset biological variability in biomedical

research is the use of transformed cell lines, established viral or bacterial strains, and in-bred animals

based on the assumption that if two laboratories use

the same bacterial strain grown under the same



95



96



Chapter 5 IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



conditions, variability from this part of experiment

will be minimized and results of manipulation of

such systems can be compared meaningfully. This,

however, does not apply to humans and nonhuman

primates. Thus the assembly of cohorts of human

subjects participating in clinical studies can only be

based on a set of predetermined clinical evaluations,

which usually do not cover all variability. Furthermore, use of predetermined conditions for the

collection of biological samples (i.e., blood collected

at same time of day after 12 hours of fasting) cannot

result in the same concentration of highly abundant

proteins and percentage of albumin. If these subjects

come back for a second visit, the variable levels of

highly abundant proteins present in the biological

sample will affect the amount of proteins removed by

immunodepletion during sample processing. This

needs to be kept in mind while planning and

executing proteomic experiments, particularly those

using body fluids.



5.3 Quality Control of

Immunodepletion

In addition to comparison of LC and one-dimensional gel electrophoresis (1DE) profiles, there is no

good method to monitor the quality of immunodepletion. Moreover, 1DE can indicate whether there is

any residual of depleted proteins present in the flowthrough fraction, if the gel staining method is sensitive

enough. For quality control, we cut a band from 1DE

gel where we would expect albumin (the most abundant protein) to be, perform in-gel digestion and

analyze such sample using nano-LC-MS/MS. This

method is sensitive enough to show a slowly

increasing number of albumin peptides after immunodepletion of 120 plasma samples using a column

that was recommended by the manufacturer to

perform according to specification to up to 100

samples. Comparison of LC profiles as shown in

Figure 5.1 is a crude assessment and can be indicative

only of loss of more than 50% of column performance.

A more thorough examination of bound rather than

flow-through protein was performed by Bjorhall and



Chapter 5



IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



97



Figure 5.1 Reproducibility of immunodepletion of plasma samples using Seppro IgY-14 column.

(A) Representative profile of immunodepletion of one of first 1e10th plasma samples.

(B) Representative profile of immunodepletion of one of last 90e100th plasma samples.



co-workers,3 which is summarized in Table 5.1. The

authors also used 1DE and 2DE to examine the efficiency of immunodepletion.



5.4 Albuminome

Immunodepletion is based on the interaction of

a proteinaceous antigen with immunoglobulin,

which is a protein as well. Although it is a highly

specific interaction, it still remains to be a proteine

protein interaction that might be nonspecific to some

extent. It needs to be noted that this interaction is

sensitive to harsh conditions of 8 M urea, SDS, or

guanidine hydrochloride (Gu-HCl), and a limited

concentration of selected detergents, as well as

ambient temperatures, can be used. However, under

these conditions, interactions of many proteins may

occur with those being immunodepleted, thus the

elimination of most abundant proteins may lead to



Table 5.1 Protein identification of bound protein fractions



from a multiple affinity removal column, separated

on a 1D SDS-PAGE gel

No



Protein ID



Protein name



Mw (kDa)



Mascot MS

scorea)



Sequence

coverage (%)



Mascot MS/MS

scoreb)



1



TRFE_HUMAN



77.0



172



22



121



2

3



ALBU_HUMAN

GC1_HUMAN

A1AT_HUMAN

HPT_HUMAN

KAC_HUMAN

HPT_HUMAN

KAC_HUMAN

HPT_HUMAN



Serotransferrin precursor

(transferrin)

Serum albumin precursor

lg g-1 chain C region

a-1-Antitrypsin precursor

Haptoglobin precursor

lg k-chain C region

Haptoglobin precursor

lg k-chain C region

Haptoglobin precursor

(a1-chain)



69.3

36.1

46.7

45.2

11.6

45.2

11.6

45.2



219

e

108

133

e

54



31



336

89

37

60

36

37



4

5

6

7



27

e

11

e



58



No proteins corresponding to unspecific binding were identified. When several precursor ions were used for MS/MS analysis, the total scores are given. A score was considered

significant if either MS or MS/MS searches resulted in a score above the 95th percentile of significance

a)

Protein scores greater than 65 are significant (p < 0.05)

b)

Individual ion scores above 34 indicate identity or extensive homology (p < 0.05)

Source: Bjorhall and colleagues.3



Chapter 5



IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



partial removal of other proteins affecting quantitation. Therefore, we can conclude that more proteins

depleted intentionally, the more other proteins will

unintentionally be removed as well. For qualitative

purposes it may have a lower impact because some

pool of noninteracting proteins will remain in the

flow-through fraction. However, for quantitative

measurements, even small portions of protein being

removed unintentionally may have a decisive effect

on differences in expression whether positive or

negative.

To test how many proteins that were immunodepleted from the human plasma sample, we took 50

mg of protein from the eluted fraction, fragmented by

trypsin digestion, fractionated the resulting peptides

using a 24-well OFFGEL (based on pI), and analyzed

each fraction further using RP-nano-LC¼MS/MS.

Spectra were searched against the UniRef90 database

with Proteome Discovered (Sequest algorithm). We

identified 96 proteins represented by at least one

unique medium and one unique high confidence

peptide, which are summarized in Table 5.2. Because

plasma from HIV-1-infected individuals was used, we

identified such proteins as gp160, gp120, and Pol.

Interestingly, many of proteins listed in Table 5.2 are

putative, and their records in UniProt/TrEMBL still

have the status “unreviewed.”

Because albumin is known from its interactions

with many molecules, including proteins, and is the

most abundant protein in plasma, it is assumed

correctly that the flow-through immunodepletion

fraction containing the most abundant proteins also

contains other coimmunodepleted proteins. This

fraction is called “albuminome,” although it also

refers to proteins co-immunodepleted due to interactions with other abundant proteins. Therefore,

albuminome was a subject of several systematic

studies that provided some insights into the

composition of co-removed subproteomes.5,12,13 Our

considerations here are more focused on quantitative

than qualitative effects of immunodepletion and

other methods. From our laboratory practice we

conclude that the capacity and performance of IgYbased columns are much higher than recommended

by manufacturers. This is because of quite a large



99



100



Chapter 5



IMMUNOAFFINITY DEPLETION OF HIGH-ABUNDANT PROTEINS



Table 5.2 Human plasma proteins



coimmunodepleted with IgY-14 column11

Accession



Molecular

mass (kDa)



Calculated pI



Protein name

Rheumatoid factor D5 light chain

(Fragment)

Similar to Piccolo protein

(Aczonin)

Geminin coiled-coil domaincontaining protein 1

Titin

Fibronectin fragments or splice

variant C (Fragment)

cDNA FLJ78387

cDNA FLJ60561, highly similar

to complement C4-B

Soluble VEGFR3 variant 3

Envelope glycoprotein gp160

or gp120 or (Fragment)

Pol

Lysyl-bradykinin

Uncharacterized protein

Uncharacterized protein

Jumonji domain containing 1A,

isoform CRA_b

Putative uncharacterized protein

TFIIH basal transcription factor

complex subunit, putative

Dynamin-associated protein,

putative

Paramyosin, putative

Putative uncharacterized protein

Putative uncharacterized protein

FYVE-containing protein, putative

Putative uncharacterized protein

Putative uncharacterized protein

Putative uncharacterized protein

Putative uncharacterized protein

Low-density lipoprotein receptor,

putative

Putative uncharacterized protein



A0N5G5



12.8



8.97



A4D1A8



410.9



5.40



A6NCL1



37.9



6.20



A6NKB1

A6YID4



3711.3

57.0



6.52

8.63



A8K008

B4DIE5



51.6

83.8



8.16

9.13



B5A928

B5ANL2



24.5

94.9



7.80

8.43



B6RP19

C9JEX1

C9JLB1

C9JU00

D6W5M4



10.7

43.8

15.6

14.0

78.6



7.12

6.43

4.70

7.20

8.29



E0VB03

E0VCF0



188.4

63.9



9.41

7.24



E0VEP4



175.4



5.86



E0VG64

E0VHH7

E0VMF5

E0VMQ5

E0VNC2

E0VPE6

E0VQH5

E0VVV4

E0VVX3



69.6

150.9

163.7

123.8

373.3

172.6

47.2

67.6

510.2



6.44

8.25

7.33

8.10

5.39

6.38

5.72

9.32

5.44



E0VZE1



68.8



8.92



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

10 Detergents, Lipids, and Final Remarks

Tải bản đầy đủ ngay(0 tr)

×