| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular & Cellular Proteomics 5:2167-2174, 2006.
© 2006 by The American Society for Biochemistry and Molecular Biology, Inc.
From the Biological Sciences Division and Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, Washington 99352
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
99% of the bulk mass of the total proteins in human plasma; this leaves perhaps hundreds of thousands of other proteins in only 1% of the protein mass of this biofluid (1). Hence removal of these high abundance proteins in the serum/plasma sample preparation process is becoming increasingly widespread to provide higher sensitivity for achieving broader proteome coverage, particularly of low abundance proteins that are normally present in the concentration range of ng/ml and lower. Regardless of the success of single protein depletion schemes (e.g. removing only albumin or immunoglobulin G), biomarker detection is usually facilitated by removal of as many high abundance proteins as possible. Several antibody-based approaches aimed at simultaneously removing multiple high abundance proteins have been developed to achieve a more comprehensive survey of the human plasma proteome. For example, Pieper et al. (2) reported an approach that is capable of removing 10 high abundance proteins in a single step, and two other recently commercialized systems offer the capability of simultaneously removing six (3) and 12 (4) of the most abundant proteins. These antibody-based depletion systems have been demonstrated to be highly efficient in removing the specifically targeted proteins. However, questions still arise with regard to their reproducibility and selectivity (3–8). In particular there are two primary questions: whether these systems are capable of removing their target proteins in a reproducible fashion and whether there is any loss of non-target proteins along with the high abundance proteins through non-selective binding or physiologically relevant association/interaction. Ideally potential losses of non-target proteins should be minimized during removal of multiple high abundance proteins, but in cases where losses do occur, such losses should be reproducible if the depletion strategy is to be applied for comparative quantitative biomarker discovery studies.
To address these concerns, we evaluated the performance of one widely used immunoaffinity subtraction system, the Beckman Coulter ProteomeLabTM IgY-121 system, for separating high abundance proteins. Our results indicate that high abundance protein separation using the IgY-12 system is highly reproducible, and the low level binding of non-target proteins in this system occurred in a reproducible fashion.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Plasma Sample—
The human blood plasma sample supplied by the Stanford University School of Medicine (Palo Alto, CA) was obtained from a single, healthy volunteer. An initial plasma protein concentration of 65 mg/ml was determined by the Coomassie protein assay (Pierce). To generate a spiked plasma sample, bovine carbonic anhydrase II, rabbit glyceraldehyde-3-phosphate dehydrogenase (G3PDH), and bovine ß-lactoglobulin (Sigma) were added to unprocessed human plasma samples to final concentrations of 40 µg/ml, and equine skeletal muscle myoglobin and chicken ovalbumin (Sigma) were added to final concentrations of 4 µg/ml. Unless otherwise noted, all protein sample processing was performed at 4 °C.
Immunoaffinity Subtraction—
Twelve high abundance plasma proteins (albumin, IgG, 
1-antitrypsin, IgA, IgM, transferrin, haptoglobin, 1-acid glycoprotein, 2-macroglobulin, apolipoprotein A-I, apolipoprotein A-II, and fibrinogen) that constitute 95% of the total protein mass of human plasma were simultaneously separated from other proteins using a prepacked 7 x 52-mm (loading capacity, 25 µl of plasma) IgY-12 affinity LC column (Beckman Coulter, Fullerton, CA; previously known as SepproTM MIXED12 column from GenWay Biotech, San Diego, CA) using an Agilent 1100 series HPLC system (Agilent, Palo Alto, CA). All IgY-12 separations were performed within the manufacturer-instructed column usage and loading capacity. Three buffers (dilution/washing buffer: 10 mM Tris-HCl, 150 mM NaCl, pH 7.4 (TBS); stripping buffer: 100 mM glycine, pH 2.5; neutralization buffer: 100 mM Tris-HCl, pH 8.0) were used in a separation scheme that consisted of sample loading-washing-eluting-neutralization followed by a re-equilibration scheme for a total cycle time of 48 min. The spiked human plasma sample was subjected to five individual IgY-12 separations, and the resulting flow-through fraction and the bound/eluted fraction were collected separately. To generate sufficient samples for subsequent two-dimensional (2D) LC-MS/MS analysis, a non-spiked plasma sample was repeatedly processed eight times, and corresponding flow-through and bound fractions were pooled. All fractions were then individually concentrated in iCON concentrators with 9-kDa molecular mass cutoffs (Pierce) followed by buffer exchange to 50 mM NH4HCO3 in the same unit according to the manufacturer’s instructions. Protein concentration was then measured using the BCA protein assay (Pierce).
Protein Digestion—
Proteins were denatured and reduced in 50 mM NH4HCO3, 8 M urea, 10 mM dithiothreitol for 1 h at 37 °C followed by alkylation where the proteins were incubated with 40 mM iodoacetamide for 1 h at room temperature in the dark. The protein mixture was then diluted 10-fold using 50 mM NH4HCO3. Sequencing grade, modified porcine trypsin (Promega, Madison, WI) was added at a trypsin:protein ratio of 1:50, and the mixture was incubated at 37 °C overnight. The tryptic digest was loaded on a 1-ml solid-phase extraction (SPE) C18 column (Supelco, Bellefonte, PA) and washed with 4 ml of 0.1% TFA, 5% ACN. Peptides were eluted from the SPE column with 1 ml of 0.1% TFA, 80% ACN and lyophilized. Samples were reconstituted in 25 mM NH4HCO3, and the peptide concentration was measured using the BCA protein assay (Pierce).
Strong Cation Exchange Peptide Fractionation—
A total of 300 µg of tryptic peptides from flow-through and bound fractions of the non-spiked plasma was individually reconstituted with 300 µl of 10 mM ammonium formate (pH 3.0), 25% ACN and fractionated by strong cation exchange chromatography on a Polysulfoethyl A 2.1 x 200-mm column (PolyLC, Columbia, MD) that was preceded by a 2.1 x 10-mm guard column. The separations were performed using an Agilent 1100 series HPLC system at a flow rate of 200 µl/min and mobile phases that consisted of 10 mM ammonium formate (pH 3.0), 25% ACN (A) and 500 mM ammonium formate (pH 6.8), 25% ACN (B). After loading 300 µl of sample onto the column, the mobile phase was maintained at 100% A for 10 min. Peptides were then separated by using a gradient from 0 to 50% B over 40 min followed by a gradient of 50–100% B over 10 min. The mobile phase was then held at 100% B for 10 min. A total of 30 fractions were collected for each separation, and each fraction was dried under vacuum. The fractions were dissolved in 30 µl of 25 mM NH4HCO3, and 10 µl of each fraction were analyzed by capillary LC-MS/MS.
Reversed-phase Capillary LC-MS/MS Analyses—
Peptide samples were analyzed using a custom built high pressure capillary LC system (9) coupled on line to a linear ion trap mass spectrometer (LTQ; ThermoElectron) via an electrospray ionization interface manufactured in-house. The reversed-phase capillary column was prepared by slurry-packing 3-µm Jupiter C18 bonded particles (Phenomenex, Torrence, CA) into a 65-cm-long, 150-µm-inner diameter x 360-µm-outer diameter fused silica capillary (Polymicro Technologies, Phoenix, AZ). The mobile phase consisted of 0.2% acetic acid and 0.05% TFA in water (A) and 0.1% TFA in 90% ACN, 10% water (B). After loading 10 µl of peptides onto the column, the mobile phase was held at 100% A for 20 min. Exponential gradient elution was performed by increasing the mobile phase composition from 0 to 70% B over 150 min. To identify the eluting peptides, the linear ion trap mass spectrometer was operated in a data-dependent MS/MS mode (m/z 400–2000) in which a full MS scan was followed by 10 MS/MS scans. The 10 most intensive precursor ions were dynamically selected in the order of highest intensity to lowest intensity and subjected to collision-induced dissociation using a normalized collision energy setting of 35%. A dynamic exclusion duration of 1 min was used. The temperature of the heated capillary and the ESI voltage were 200 °C and 2.2 kV, respectively.
Data Analysis—
The SEQUEST algorithm (part of the Bioworks software package, Version 3.1 SR1; ThermoFinnigan) was used to search all MS/MS spectra independently against the human International Protein Index (IPI) database (Version 3.05 that consists of 49,161 protein entries; available on line at www.ebi.ac.uk/IPI) supplied with the sequences of the five protein standards and the corresponding reversed human IPI database with no enzyme constraint. Dynamic carboxamidomethylation of cysteine and oxidation of methionine were used during the database search. The reversed human protein database was created as reported previously (10) by reversing the order of the amino acid sequences for each protein. The false discovery rates of peptide identifications were estimated as described previously by dividing the number of unique peptides from the reversed database search by the number of unique peptides from the normal database search (10).
Criteria that would yield an overall confidence of over 95% for peptide identification at the unique peptide level were established for filtering raw peptide identifications. For example, with delta correlation (
Cn) value of 
0.1, the following cross-correlation score (Xcorr) cutoffs were used: for the 1+ charge state, Xcorr 1.5 for fully tryptic peptides and Xcorr 3.0 for partially tryptic peptides; for the 2+ charge state, Xcorr 2.7 for fully tryptic peptides and Xcorr 3.7 for partially tryptic peptides; and for the 3+ charge state, Xcorr 3.3 for fully tryptic peptides and Xcorr 4.5 for partially tryptic peptides. Non-tryptic peptides were not considered. Two additional Cn cutoff values of 0.05 and 0.16 were applied to reduce false negative identifications while maintaining a 95% level of confidence for peptide assignments. For Cn 0.05, the minimum acceptable Xcorr value was raised to achieve a comparable percentage of false positive rate identifications, and similarly for Cn 0.16, the minimum acceptable Xcorr value was reduced. In an attempt to remove redundant protein entries in the reported results, ProteinProphet software was used as a clustering tool to group similar or related protein entries into a "protein group" (11). All unique peptides that passed the filtering criteria were assigned an identical peptide probability score of 1 and entered into the software program (solely for clustering analysis) to generate a final list of nonredundant proteins or protein groups. One protein identification was randomly selected to represent each corresponding protein group that contains member database entries.
|
RESULTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In the LC-MS/MS analyses of complex peptide mixtures, the selection of ions for MS/MS is not totally randomized but rather dependent on the width of the chromatographic peaks or the concentration of peptides delivered to the mass spectrometer. Thus, data acquisition can be biased against the low abundance species, and peptide ions from more abundant proteins are selected more frequently for MS/MS (12). The number of spectra acquired for a given protein (spectral count) was shown to accurately reflect its relative abundance with a linear correlation over 2 orders of magnitude of linear dynamic range (12). It has also been demonstrated that spectral counting and MS-derived peak areas strongly correlate for determining quantitative changes in protein expression (13, 14). In this study, raw peptide identifications were filtered using a set of stringent criteria providing >95% confidence at the unique peptide level, and the spectral counts for each protein were then calculated by summing the numbers of observations of all filtered peptides derived from the protein for comparison of relative protein abundances in the five replicate analyses.
An average of 54 and 122 proteins were identified from the IgY-12-bound fraction and flow-through fraction, respectively (only proteins with two or more unique peptide identifications from all the replicate analyses are considered; the ambiguous immunoglobulin identifications were removed). The numbers of proteins identified in the five replicate analyses are fairly consistent with a coefficient of variance of less than 4% for both fractions. Fig. 1 displays linear correlations of the spectral counts between pairs of analyses among the five replicates of the flow-through fractions. Note that a correlation coefficient (R2) of greater than 0.98 was obtained from each comparison, and an averaged coefficient of variance of spectral count of 17% was obtained for proteins identified from flow-through fractions in the five replicate analyses. Similar results were obtained from the replicate analyses of the bound fractions (data not shown), demonstrating the excellent reproducibility of high abundance protein separation using the IgY-12 system.
|
|
|
A total of 38 bound non-target proteins were identified in the replicate LC-MS/MS analyses of the five individually prepared bound and flow-through fractions of the IgY-12 column. To compare the extent of nonspecific binding in an individual high abundance protein separation, a distribution (D)-value was calculated by dividing the spectral counts for each protein from the bound fraction by the sum of spectral counts for the corresponding protein from the bound fraction and from the flow-through fraction. The D-value serves as an indicator for evaluating the reproducibility of protein partition between the bound and flow-through fractions. A larger D-value suggests relatively significant nonspecific binding. The D-values of each protein in the replicate analyses are listed in Table III; the numbers of unique peptides for specific proteins identified from the bound fraction and from the flow-through fraction are also provided. Details of these peptide and protein identifications in the replicate LC-MS/MS analyses can be found in Supplemental Table 1. A number of proteins bound significantly to the IgY-12 column, such as zinc-2-glycoprotein, apolipoproteins (C-I, C-III, D, E, and L1), CD5 antigen-like protein, serum amyloid proteins A2 and A4, transthyretin, hemoglobin, serum paraoxonase/arylesterase 1, and leucine-rich 2-glycoprotein (Table III). One important finding from this set of analyses is that, although many non-target proteins were observed to bind to the IgY-12 systems at different levels, the fraction of binding for any specific non-target protein to the column appeared to be very reproducible (Table III). Binding of one of the five spiked protein standards (rabbit G3PDH) was also observed (Table II), again in a reproducible manner.
|
2-glycoprotein (12 unique peptides), and CD5 antigen-like protein (11 unique peptides). A significant portion of proteins were observed to be common in both the bound and flow-through fractions based on both one-dimensional and 2D LC-MS/MS, suggesting that many non-target proteins bind to the columns to a certain extent either directly or indirectly. Meanwhile although the 12 target proteins were removed with high efficiency as reported previously (4) (Supplemental Table 5), none of them appeared to be removed at 100% efficiency when 2D LC-MS/MS was used for the analysis (Supplemental Table 2), presumably due to their high original concentration and the increased sampling depth of the flow-through fraction. |
DISCUSSION |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
99% of its protein mass) in a single step recently became commercially available (Sigma ProteoPrep® 20 plasma immunodepletion kit), and LC columns that remove 50–100 proteins are anticipated in the near future. The reproducibility and selectivity for such systems, i.e. the risk of losing potential proteins of interest along with the target proteins, is critical for these systems to be broadly applied for quantitative candidate biomarker discovery or verification studies. In this study, the widely used IgY-12 immunodepletion system was evaluated based on replicate depletion experiments and high resolution LC-MS/MS and 2D LC-MS/MS analyses. The spectral counting approach used here has been demonstrated previously to be applicable to comparative proteomic analysis (12–14). The high abundance protein separations were determined to be very reproducible based on the analyses of replicate flow-through fractions using LC-MS/MS and spectral counts (Fig. 1); similar results were obtained for the bound fractions. High reproducibility was also observed based on the direct measurements of protein recoveries at key sample processing steps (Table II) where an average of 5.5% of proteins were recovered from the flow-through fraction, consistent with the 95% protein mass removal estimated from known concentrations of the 12 target proteins in human plasma and the 95–99.5% depletion efficiency of the IgY-12 system determined in previous studies using SDS-PAGE and ELISA (4). The averaged recovery of flow-through proteins is similar to what was determined in a recent large scale plasma proteome profiling study (5.6%) utilizing the same IgY-12 system (15). Moreover high abundance protein separation on the IgY-12 column did not significantly deteriorate within the manufacturer-defined usage and sample load (data not shown). These results demonstrated that the performance of the IgY-12 column is highly consistent, and the sample processing procedures used in this study are robust and reproducible, thus suitable for comparative proteomic studies for candidate biomarker discovery. Recently a study reported that reproducible and high efficiency depletion can be obtained from the MARS column as determined by using ELISA (8). This complementary study demonstrates yet another example that the selective depletion of abundant plasma proteins by immunoaffinity chromatography is a valid approach for the enrichment of potential biomarkers sought by proteomics-based methodologies.
Estimating the percentage of protein partition between the bound and flow-through fractions is challenging due to the drastic component differences between the two fractions and nearly 20-fold difference in protein quantities as a result of the immunoaffinity separation itself. The calculated D-value does not reflect the true percentage of protein partition in the bound and flow-through fractions but still serves as a good indicator of the reproducibility of the nonspecific binding: the alternation of spectral counts for both the bound and flow-through fractions in the replicate analyses should be consistent if the high abundance protein separation process is reproducible. However, when the spectral counts from both fractions are small, the calculated D-values will be less reproducible due to the undersampling nature of tandem mass spectrometry measurements.
A relatively large number of non-target proteins that bind to the IgY-12 system have regions of amino acid sequence homologous to target proteins, e.g. apolipoproteins (homologous to apolipoproteins A-I and A-II), and these epitopes may be bound directly to the polyclonal antibodies. Furthermore many abundant proteins may bind to the column through nonspecific interactions with the bead surfaces; other proteins are most likely binding to one or more of the affinity column-bound target proteins (that act as "bait") through mechanisms of specific and physiologically relevant protein-protein interactions under certain conditions, i.e. the effect of "interactome" (16, 17). The proteins identified from both the bound and flow-through fractions may provide interesting information related to specific protein-protein interactions (Table III and Supplemental Table 2). For example, hemoglobin and complement C3 are reported to bind to haptoglobin and IgG, respectively (18). Zinc-2-glyoprotein, a protein that binds 100% to IgY-12 column (Table III and Supplemental Table 2), may interact with the targeted apolipoproteins and/or interact directly with the Fc region of IgG because it is a lipid-mobilizing factor similar to the Fc receptor and belongs to the Class I major histocompatibility complex immunoglobulin superfamily (19, 20). Transthyretin (prealbumin, 20–40 µg/ml), another relatively abundant protein, also binds to the IgY-12 column (Table III and Supplemental Table 2) possibly due to protein-protein interactions with the immunoaffinity-bound target proteins because both transthyretin and albumin are major transport proteins for the hydrophobic thyroid hormones. Alternatively a contaminating antigen used for antibody generation in the IgY-12 system may be present (4) because transthyretin binds to completion to the IgY-12 column but only at very low levels to the MARS column (data not shown). One of the five spiked protein standards, rabbit G3PDH (spiked at 40 µg/ml level), also reproducibly binds to the IgY-12 system (Table II) in the five replicate LC-MS/MS analyses. However, a survey of the non-spiked plasma sample using 2D LC-MS/MS revealed that human G3PDH does not bind to the IgY-12 system at all at its normal plasma concentration. Six G3PDH unique peptides were identified only from the flow-through fraction, and among them three peptides are unique to human G3PDH (Supplemental Table 4), although the protein sequences of rabbit and human G3PDH have 94% in common. These results suggest that the binding of G3PDH to the IgY-12 system may be concentration-dependent; however, the exact mechanism remains unclear.
Mild denaturing conditions (e.g. addition of 5–20% acetonitrile) may help to disrupt the binding of some proteins, particularly low molecular weight proteins, to the carrier proteins, thus decreasing the possibility of co-depleting non-target proteins of interest (21, 22). Although more stringent washing conditions may assist with the dissociation of certain tightly bound proteins, it is compromised by the need to maximize binding of target proteins to the depletion systems to achieve high efficiency immunodepletion. It is likely unavoidable that some of the non-target proteins will bind to the immunoaffinity subtraction systems without compromising the optimized depletion conditions. A practical solution to this scenario would be to characterize both the bound and flow-through fractions, i.e. use these systems as a method for "partitioning" instead of "depletion" so that virtually no protein of interest would be lost along the high abundance protein separation process. This approach would also create an opportunity for investigating the subinteractomes in the blood using individual antibody resin, e.g. albuminome, haptoglobinome, apolipoproteinome, etc. (4). An important observation made in this study is that the partitioning of any specific protein appeared to be very reproducible across replicate analyses regardless of whether some non-target proteins bound at varying levels, indicating the overall high reproducibility of the IgY-12 system as well as the sample processing procedures used in this study.
In summary, the present study demonstrated that immunoaffinity subtraction systems, exemplified by the widely used IgY-12 system, can reproducibly and efficiently remove targeted high abundance proteins. Binding of non-target proteins to the system does occur but in a reproducible fashion under controlled conditions. These findings suggest that these immunoaffinity systems can be readily integrated into an analytical strategy for broader and deeper quantitative proteomic analysis without complicating the overall accuracy of the measurements. As a result, these systems should find broad application in candidate disease biomarker discovery or verification using clinically obtainable biofluids as samples.
|
FOOTNOTES |
|---|
Published, MCP Papers in Press, July 19, 2006, DOI 10.1074/mcp.T600039-MCP200
1 The abbreviations used are: IgY, immunoglobulin yolk; 2D, two-dimensional; G3PDH, glyceraldehyde-3-phosphate dehydrogenase; SPE, solid-phase extraction; Cn, delta correlation; Xcorr, cross-correlation score; MARS, multiple affinity removal system; D, distribution. 
* Portions of this research were supported by NIGMS, National Institutes of Health, Large Scale Collaborative Research Grant U54 GM-62119-02 and National Institutes of Health National Center for Research Resources Grant RR18522. Work was performed in the Environmental Molecular Science Laboratory, a United States Department of Energy (DOE) national scientific user facility located on the campus of Pacific Northwest National Laboratory (PNNL) in Richland, WA. PNNL is a multiprogram national laboratory operated by Battelle Memorial Institute for the DOE under Contract DE-AC05-76RLO-1830. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
S The on-line version of this article (available at http://www.mcponline.org) contains supplemental material.
To whom correspondence should be addressed: Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, P. O. Box 999, MSIN: K8-98, Richland, WA 99352. E-mail: rds{at}pnl.gov
|
REFERENCES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2-glycoprotein, a soluble fat-depleting factor related to major histocompatibility complex proteins.
J. Biol. Chem.
276, 35008
–350132-glycoprotein, a lipid mobilizing factor, is expressed in adipocytes and is up-regulated in mice with cancer cachexia.
Proc. Natl. Acad. Sci. U. S. A.
101, 2500
–2505
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Journal of Biological Chemistry |
| Journal of Lipid Research | ASBMB Today |
