Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Dec 7.
Published in final edited form as: J Proteome Res. 2012 Oct 29;11(12):5947–5958. doi: 10.1021/pr300686k

Assessment of Two Immunodepletion Methods: Off-Target Effects and Variations in Immunodepletion Efficiency May Confound Plasma Proteomics

Bhavinkumar B Patel 1,, Carlos A Barrero 1,, Alan Braverman 1,§, Phillip D Kim 1,, Kelly A Jones 1,, Dian Er Chen 1,#, Russell P Bowler 1,**, Salim Merali 1,, Steven G Kelsen 1,§, Anthony T Yeung 1,*,
PMCID: PMC3518753  NIHMSID: NIHMS425311  PMID: 23082855

Abstract

Immunodepletion of abundant plasma proteins increases the depth of proteome penetration by mass spectrometry. However, the nature and extent of immunodepletion and the effect of off-target depletion on the quantitative comparison of the residual proteins have not been critically addressed. We performed mass spectrometry label-free quantitation to determine which proteins were immunodepleted and by how much. Two immunodepletion resins were compared: Qproteome (Qiagen) which removes albumin+immunoglobulins and Seppro IgY14+SuperMix (Sigma-Aldrich) which removes 14 target proteins plus a number of unidentified proteins. Plasma collected by P100 proteomic plasma collection tubes (BD) from 20 human subjects was individually immunodepleted to minimize potential variability, prior to pooling. The abundant proteins were quantified better when using only albumin+immunoglobulins removal (Qproteome) while lower abundance proteins were evaluated better using exhaustive immunodepletion (Seppro IgY14+SuperMix). The latter resin removed at least 155 proteins, 38% of the plasma proteome in protein number and 94% of plasma protein in mass. The depth of immunodepletion likely accounts for the effectiveness of this resin in revealing low abundance proteins. However, the more profound immunodepletion achieved with the IgY14+SuperMix may lead to false-positive fold-changes between comparison groups if the reproducibility and efficiency of the depletion of a given protein is not considered.

Keywords: immunodepletion, Seppro, IgY, Qproteome, iTRAQ, EMMOL normalization, off-target

INTRODUCTION

The large dynamic range of protein concentration in plasma, i.e., nine orders of magnitude1, has necessitated the removal of the higher abundance proteins to visualize the low abundance proteins. Immunodepletion of one or more high abundance proteins using antibody affinity immunoabsorption of specific antigens is one method used to accomplish this purpose. Initial application of immunodepletion to proteomics consisted of the depletion of serum albumin2. Subsequent methods, including the Qproteome (Qiagen) matrix, also remove immunoglobulins. Further generations of immunoaffinity columns, including the popular Multiple Affinity Removal System Columns (Agilent) and the ProteoPrep20 Immunodepletion Column (Sigma-Aldrich) remove several more proteins35.

To accomplish depletion of a greater number of proteins, the SuperMix resin was created by Sigma-Aldrich by immunization of rabbits with the protein fraction that did not bind to an IgY14 column. In principal, the medium abundance, immunogenic proteins, now enriched, in this fraction induced antibodies for the SuperMix column. However, the targets of the SuperMix resin have never been fully identified and the ability of the Seppro IgY14+SuperMix column (Sigma-Aldrich) to allow quantitation of low abundance proteins is unstudied.

When a given protein is targeted for antibody-mediated removal, protein-protein interactions can lead to the removal of multiple proteins. The terms albuminome and depletome have been proposed to describe the proteins that are bound to albumin or other proteins during immunodepletion 69. Estimations of the number of proteins that bind to albumin range from 24 to 67 depending on the methods used 8, 9. A recent report performed an IgY14 column and a SuperMix column in tandem, compared their depleted proteomes and concluded that numerous proteins were immunodepleted by taking the IgY14 exhaustive immunodepletion to the next higher level of exhaustive immunodepletion10. In contrast, our report quantified the immunodepletion efficiencies of individual plasma proteins by a single-stage pre-mixed IgY14+SuperMix column.

Success of immunodepletion approaches has often been measured as the detection of low abundance proteins. However, the nature and extent of immunodepletion and the effect of off-target depletion on the quantitative comparison of the residual proteins have not been critically addressed. We performed mass spectrometry label-free quantitation to determine which proteins were immunodepleted and by how much. Two popular immunodepletion resins were compared: Qproteome (Qiagen) which removes albumin and immunoglobulins and Seppro IgY14+SuperMix (Sigma-Aldrich) which removes 14 target proteins plus a number of unidentified proteins. We found that the IgY14+SuperMix column depleted a substantial portion of the plasma proteome including 155 proteins many of which were previously identified as disease biomarkers. The degree and variability of depletion of such biomarkers can be an issue when comparing their expression levels between groups. Thus exhaustive immunodepletion may lead to false positive or false negative results if both the variability and efficiency of the immunodepletion and the off-target removal of individual proteins are not addressed.

Finally, we combined the data from these two proteomes to yield a continuous quantitative picture of human plasma with 5 logs of dynamic range. The abundant proteins were favored when using only albumin+immunoglobulins removal while lower abundance proteins were evaluated better using exhaustive immunodepletion.

MATERIALS AND METHODS

Study Samples

Plasma samples were collected from 20 subjects using the BD P100 Blood Collection Tubes that contained protease inhibitors, according to manufacturer’s instructions. The subjects studied were Caucasian male ex-smokers enrolled in the NIH COPDGene® project, a large multi-center, genome-wide association study designed to elucidate the genetic basis for Chronic Obstructive Pulmonary Disease (COPD)1118. The use of plasma from subjects in this study was incidental and resulted from a desire to use high quality samples in this analysis.

Study Design

The study design for each of the two immunodepletion methods is shown in Figure 1. The 20 subjects were divided into two study groups of 10 subjects each i.e., Groups 1 and 2. Individual samples taken from disease and control groups were not mixed. To provide technical replicates for the study, iTRAQ 4-plex methodology was used since it allows four replicate experiments to be performed for each immunodepletion method and has the advantage of avoiding LC-MS/MS under-sampling.

Figure 1.

Figure 1

Work flow of an immunodepletion evaluation experiment. The process was done once for the Qproteome method and once for the Seppro IgY14+SuperMix method.

The Depletion of High-Abundance Proteins

Two methods of immunodepletion were used according to the manufacturer’s protocols i.e., the Qproteome spin column approach (Qproteome albumin/IgG Depletion Kit, Qiagen, Carson City, CA) 19 and the Seppro IgY14+SuperMix20 (Seppro Human IgY14 Human SuperMix LC5, catalog number SEP000-KT, Sigma-Aldrich Inc., St. Louis, MO). Researchers using immunodepletion have expressed concern over the possibility of run-to-run variations in the performance of immunodepletion chromatography. One way to prevent such potential variations from negatively impacting this study is to smooth the putative fluctuations by pooling the samples produced by ten chromatography runs. 100 microliter of plasma of each subject was individually immunodepleted by each method. Then aliquots of depleted plasma of 20 subjects were pooled as two groups of 10 each after quality analysis by SDS PAGE and quantitation.

iTRAQ Proteomics

The iTRAQ proteomic analysis procedures used in the current study have been described in detail previously21. From the Qproteome immunodepletion, 100 micrograms of the pooled immunodepleted plasma protein from Group 1 subjects was digested with trypsin and labeled with the iTRAQ 114 label and as a replicate by the iTRAQ 115 label. 100 micrograms of the pooled immunodepleted protein from Group 2 was labeled with the iTRAQ 116 label and as a replicate by the iTRAQ 117 label. These four samples were pooled into an iTRAQ peptide isoelectric focusing fractionation experiment 2123 using the pH range 4.0–4.5 18 cm long region of a 24 cm IPG strip (pH 3.5–4.5 GE Healthcare) separating the peptides into 65 fractions as previously described in detail21. An identical approach was used for the Seppro IgY14+SuperMix immunodepletion flow through fraction. LC-MS/MS on an Applied Biosystems QSTAR XL qTOF mass spectrometer was performed on each IEF fraction as previously described in detail 24 including a second LC-MS/MS run for each sample using an exclusion list generated from the first LC-MS/MS run to obtain better peptide coverage. Peak lists were generated by the Mascot.dll script written by Matrix Science for the QSTAR instrument. Peptide charges +2 and +3, and monoisotopic. Enzyme was specified as trypsin. The peak lists were combined. Data processing of peptides based on 95% confidence for peptide assignment, with ion score >20 and at least one Bold Red by Mascot 2.2 software as described, allowing no trypsin missed cleavage and no methionine oxidation when calculating the emPAI score. Carbamidomethylation (Cys) was set as fixed modification. The precursor ion m/z tolerance was set at 150 ppm; the product ion m/z tolerance at 0.5 Da. SwissProt Database Release 2010_09 containing 519348 sequence entries was used. The false discovery rate was <3.5 % at the peptide level as determined by Mascot using a randomized SwissProt database. To avoid the possibility of missing identification of low abundance proteins, the database was searched again allowing one trypsin missed cleavage and variable methionine oxidation.

Both IEF (e.g.: OFFGEL of Agilent) and strong cation exchange HPLC (IEX) chromatography can produce the hyper-fractionation desired to minimize signal contamination in an iTRAQ quantitation experiment. IEF was used. Peptides focus at three tight pH ranges22. The narrow pH range of 4.0–4.5 contains about 1/3 of the total peptides and thus is another fractionation step. Having less peptide in the LC-MS/MS minimizes the amount of contamination of each targeted peptide peak being quantified.

Informatics

To calculate the immunodepletion efficiency of a protein, its concentration in the depleted plasma sample and in the initial plasma sample must be compared. Since our focus is the Seppro IgY14+Super Mix column, the Qproteome-depleted plasma was used as a surrogate for the initial plasma sample. This allowed us to assess the effect of the Seppro IgY14+Super Mix column on the lower abundance proteins that are difficult to detect in crude plasma.

Although the four samples in an iTRAQ 4-plex experiment were intended to be identical in total peptide quantity, we sought to correct for variations in protein load caused by protein assay, trypsin digestion, pipetting, and the labeling-chemistry efficiency in each of the iTRAQ channels. We thus used our recently-reported method of iTRAQ data normalization for hyper-fractionated samples called EMMOL21 (emPAI score × molecular weight normalization). emPAI is the exponentially modified Peptide Abundance Index. EMMOL uses mass spectrometry data to deduce the amounts of each protein in the four iTRAQ channels from the iTRAQ ratios. EMMOL normalization of the total protein quantity in each iTRAQ channel is guided by weighting towards the more abundant proteins where mass spectrometry measurements are more accurate for determining the relative values of total protein. In contrast, an iTRAQ normalization scheme that weighs all iTRAQ-labeled peptides equally regardless of protein abundance will suffer from the low accuracy of the identification and quantitation of low-abundance proteins. The peptides derived from low-abundance proteins are relatively low in intensity, producing iTRAQ reporter ions of low intensities. Moreover, fewer of the different peptides anticipated from the protein sequences will be observed by the mass spectrometer due to ion suppression by the higher abundance proteins and the threshold of sensitivity of the instrument.

Protein Quantitation

Key values included in Tables 14 are the iTRAQ reporter ratios, UniProt ID, Title, Gene Symbol, prot score, prot_mass, prot_matches, prot_cover, prot_pi, and emPAI.

Table 1.

Immunodepletion Efficiencies of the Targets of Qproteome and IgY14+SuperMix column

Immuno-depletion method Title Gene Symbol prot_score prot_mass prot_matches prot_cover emPAI Average mg/mL in depleted plasma % immunodepletion
Qproteome Serum albumin ALB 4726 80107 236 39.2 2.71 1.17 97
Seppro Serum albumin ALB 3083 80107 137 38.4 2.24 0.09 99
Qproteome Ig gamma-1 chain C region IGHG1 8609 40775 409 39.4 2.43 0.58 93
Seppro Ig gamma-1 chain C region IGHG1 464 40775 28 10.3 0.42 0.01 99
Qproteome Ig alpha-1 chain C region IGHA1 6711 40503 264 25.8 1.89 0.42 93
Seppro Ig alpha-1 chain C region IGHA1 936 40503 43 22.4 1.03 0.02 96
Qproteome Ig kappa chain C region IGKC 6260 13070 168 69.8 3.84 0.29 93
Seppro Ig kappa chain C region IGKC 2458 13070 61 53.8 1.86 0.01 96
Qproteome Ig mu chain C region IGHM 5726 53274 209 40.7 2.38 0.72 93
Seppro Ig mu chain C region IGHM 4036 53274 196 44 2.87 0.08 87
Qproteome Immunoglobulin J chain IGJ 304 17049 8 7.5 0.5 0.05 93
Seppro Immunoglobulin J chain IGJ 174 19984 5 11.9 0.7 0.01 80
Qproteome Complement C3 C3 29308 204997 1046 53.1 3.61 4.13
Seppro Complement C3 C3 1321 204997 51 13.1 0.31 0.03 99
Qproteome Serotransferrin TF 18294 87782 961 55.7 8.3 4.08
Seppro Serotransferrin TF 1458 87782 64 27.9 1.1 0.05 98
Qproteome Haptoglobin-related protein HPR 7464 43675 334 30.7 3.4 0.82
Seppro Haptoglobin-related protein HPR 875 43697 36 20.4 1.27 0.03 97
Qproteome Haptoglobin HP 10951 51048 620 47.3 6.74 1.90
Seppro Haptoglobin HP 1523 51048 83 31.3 2.82 0.08 97
Qproteome Apolipoprotein A-I APOA1 24599 34073 1192 76.4 58.75 11.12
Seppro Apolipoprotein A-I APOA1 11090 34073 532 77.5 34.37 0.62 95
Qproteome Apolipoprotein A-II APOA2 3135 12867 334 74 17.91 1.29
Seppro Apolipoprotein A-II APOA2 508 12867 86 74 10.08 0.07 95
Qproteome Alpha-1-acid glycoprotein 1 ORM1 3162 25886 150 48.3 4.19 0.61
Seppro Glial fibrillary acidic protein ORM1 4211 25886 214 46.8 3.52 0.05 92
Qproteome Alpha-1-acid glycoprotein 2 ORM2 1418 25890 95 31.3 2.44 0.35
Seppro Alpha-1-acid glycoprotein 2 ORM2 2546 25890 156 29.9 2.44 0.03 91
Qproteome Alpha-2-macroglobulin A2M 19430 177570 838 46 3.2 3.18
Seppro Alpha-2-macroglobulin A2M 36377 177584 1446 53.4 4.96 0.48 84
Qproteome Fibrinogen gamma chain FGG 4757 57150 302 33.6 3.55 1.12
Seppro Fibrinogen gamma chain FGG 4348 57150 283 48.1 4.85 0.15 84
Qproteome Fibrinogen alpha chain FGA 12367 101996 656 36.5 3.01 1.69
Seppro Fibrinogen alpha chain FGA 13296 101996 690 42.6 5.36 0.30 79
Qproteome Apolipoprotein B-100 APOB 13207 568255 471 25.2 0.91 2.90
Seppro Apolipoprotein B-100 APOB 31350 568096 1259 48.3 3.12 0.97 66

Table 4.

Putative Concentrations of the Composite 412 Protein Plasma Proteome less Albumin and Immunoglobulins

Immuno-depletion method Title Gene Symbol prot_score prot_mass prot_matches prot_cover emPAI Average mg/mL in depleted plasma
Qproteome Centrosomal protein of 164 kDa CEP164 94 179858 78 0.4 0.02 0.020
Seppro Attractin ATRN 803 172529 48 5.8 0.21 0.020
Qproteome Inner centromere protein INCENP 42 117004 16 2.7 0.03 0.020
Qproteome Tumor necrosis factor receptor superfamily member 8 TNFRSF8 86 69388 10 1 0.05 0.019
Seppro Thrombospondin-1 THBS1 774 141361 23 7.1 0.23 0.017
Qproteome Ras GTPase-activating protein SynGAP SYNGAP1 33 158671 5 0.6 0.02 0.017
Seppro Multiple epidermal growth factor-like domains protein 8 MEGF8 282 320217 10 3.2 0.07 0.012
Seppro Neural cell adhesion molecule L1-like protein CHL1 246 147296 28 2.4 0.13 0.010
Seppro Phosphatidylcholine-sterol acyltransferase LCAT 288 51762 11 5 0.32 0.009
Seppro Pantetheinase VNN1 295 60742 8 11.5 0.27 0.009
Seppro Intercellular adhesion molecule 1 ICAM1 693 62046 17 8.8 0.26 0.009
Seppro GDH/6PGL endoplasmic bifunctional protein H6PD 118 93441 7 5.4 0.17 0.009
Seppro Transforming growth factor-beta-induced protein ig-h3 TGFBI 808 80161 23 9.4 0.2 0.009
Seppro Interleukin-1 receptor accessory protein IL1RAP 269 72170 12 5.8 0.22 0.008
Seppro Receptor-type tyrosine-protein phosphatase eta PTPRJ 421 156271 9 4.6 0.1 0.008
Seppro Nidogen-1 NID1 71 143465 4 2.6 0.11 0.008
Seppro Neogenin NEO1 147 170601 8 3.3 0.09 0.008
Seppro Laminin subunit alpha-2 LAMA2 89 379925 22 2.1 0.04 0.008
Seppro Tenascin-X TNXB 306 488484 13 1.9 0.03 0.008
Seppro 1 Macrophage mannose receptor 1-like protein MRC1L1 100 183170 4 3 0.08 0.008
Seppro Platelet basic protein PPBP 686 16188 22 18.8 0.91 0.008
Seppro Intercellular adhesion molecule 2 ICAM2 345 32877 9 9.5 0.39 0.007
Seppro Cell surface glycoprotein MUC18 MCAM 108 77287 19 2.8 0.15 0.006
Seppro Hypoxia up-regulated protein 1 HYOU1 392 122734 13 6 0.09 0.006
Seppro Vascular endothelial growth factor receptor 3 FLT4 184 155572 7 4.3 0.07 0.006
Seppro Protein S100-A9 S100A9 110 15020 4 24.6 0.59 0.005
Seppro Cystatin-C CST3 379 17170 10 18.5 0.5 0.005
Seppro Coiled-coil domain-containing protein 126 CCDC126 303 17244 8 8.6 0.5 0.005
Seppro Centromere protein F CENPF 36 417533 22 1 0.02 0.005
Seppro Metalloproteinase inhibitor 1 TIMP1 281 25137 6 5.8 0.33 0.004
Seppro C-reactive protein CRP 289 27355 11 11.2 0.3 0.004
Seppro Multimerin-2 MMRN2 195 110792 8 1.8 0.07 0.004
Seppro Ficolin-2 FCN2 147 36885 8 6.7 0.21 0.004
Seppro Noelin OLFM1 177 59540 8 4.7 0.13 0.004
Seppro L-selectin SELL 108 47532 9 7 0.16 0.004
Seppro Endoglin ENG 255 74730 10 5.8 0.1 0.004
Seppro Scavenger receptor cysteine-rich type 1protein M130 CD163 246 136636 9 2.1 0.05 0.004
Seppro Plectin-1 PLEC1 40 569055 21 1.1 0.01 0.003
Seppro Poly [ADP-ribose] polymerase 14 PARP14 54 218539 12 0.8 0.02 0.002
Seppro Rapamycin-insensitive companion of mTOR RICTOR 63 208185 22 1.2 0.02 0.002
Seppro Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit beta isoform PIK3CB 74 135868 25 0.7 0.03 0.002
Seppro Nucleoporin NUP188 homolog NUP188 37 209465 8 0.3 0.02 0.002
Seppro Pleckstrin homology domain-containing family A member 7 PLEKHA7 38 135984 22 1.2 0.03 0.002
Seppro Enhancer of polycomb homolog 1 EPC1 47 102045 69 1.8 0.04 0.002
Seppro Kinesin-like protein KIF21B KIF21B 104 199879 49 1 0.02 0.002
Seppro Atrophin-1 ATN1 40 130632 28 1 0.03 0.002
Seppro MIF4G domain-containing protein MIF4GD 67 27334 18 2.7 0.14 0.002
Seppro Formin-2 FMN2 36 194304 12 0.6 0.02 0.002
Seppro Probable ATP-dependent RNA helicase DDX41 DDX41 44 77826 5 1.1 0.05 0.002
Seppro Uncharacterized protein C2orf77 C2orf77 100 77323 22 1.1 0.05 0.002
Seppro Probable G-protein coupled receptor 25 GPR25 46 39984 5 1.9 0.09 0.002
Seppro Tumor necrosis factor receptor superfamily member 10A TNFRSF10A 37 53964 12 1.5 0.07 0.002
Seppro Interleukin-12 receptor subunit beta-1 IL12RB1 36 77456 12 0.9 0.05 0.002
Seppro Plexin domain-containing protein 2 PLXDC2 153 62998 5 2.1 0.06 0.002
Seppro SPRY domain-containing protein 3 SPRYD3 34 53346 6 1.4 0.07 0.002
Seppro HMG domain-containing protein 3 HMGXB3 62 181488 9 0.5 0.02 0.002
Seppro Calcyclin-binding protein CACYBP 108 31207 43 5.7 0.12 0.002
Seppro UPF0505 protein C16orf62 C16orf62 82 120515 4 2.1 0.03 0.002
Seppro Synaptotagmin-13 SYT13 46 51806 15 2.6 0.07 0.002
Seppro Cholesteryl ester transfer protein CETP 169 59011 4 3 0.06 0.002
Seppro ATP-binding cassette sub-family B member 9 ABCB9 232 88768 17 0.9 0.04 0.002
Seppro E3 ubiquitin-protein ligase TRIM37 TRIM37 70 116552 6 1.7 0.03 0.002
Seppro 26S proteasome non-ATPase regulatory subunit 1 PSMD1 38 117026 8 1 0.03 0.002
Seppro Sodium- and chloride-dependent transporter XTRP3 SLC6A20 37 69839 7 1 0.05 0.002
Seppro Retina-specific copper amine oxidase AOC2 38 86239 6 0.8 0.04 0.002
Seppro DNA ligase 1 LIG1 31 112676 6 1.2 0.03 0.002
Seppro Leucine-rich repeat serine/threonine-protein kinase 2 LRRK2 36 315074 4 0.4 0.01 0.002
Seppro Structural maintenance of chromosomes protein 3 SMC3 35 160154 23 2.2 0.02 0.002
Seppro Cohesin subunit SA-1 STAG1 54 156257 10 1 0.02 0.002
Seppro Centrosomal protein of 135 kDa CEP135 34 149283 17 1 0.02 0.002
Seppro Voltage-dependent N-type calcium channel subunit alpha-1B CACNA1B 56 279251 22 0.6 0.01 0.001
Seppro Nuclear mitotic apparatus protein 1 NUMA1 104 261102 21 0.7 0.01 0.001

EMMOL method of proteome normalization21 was used to calculate the protein concentration of each protein in each pooled plasma sample. EMMOL was demonstrated to be valid by using defined E. coli lysate samples over a 20 fold range of protein quantities and over three logs of dynamic range of protein abundance21.

EMMOL uses the relationship from equation 4 of Ishihama et al.25:

Proteincontent(weight%)=(emPAI×Mr)/(emPAI×Mr)×100

where Mr = molecular weight.

emPAI score is proportional to the fraction of observed peptides/theoretical peptides for a given protein (allowing no methionine oxidation and no missed trypsin cleavage). “emPAI × molecular weight” value is roughly proportional to the abundance of a protein.

Thus the EMMOL data processing workflow involves:

  1. Calculate (emPAI × molecular weight) for each protein as its total in four iTRAQ channels

  2. Normalize each protein to the sum of all proteins for the 4 iTRAQ channels, herein 180 μg protein (the total amount loaded into the 4 channels but any arbitrary total protein value such as the normal plasma protein concentration could have been used instead).

  3. Use the iTRAQ ratios of each protein to calculate the μg of this protein in each iTRAQ channel

  4. Sum the total protein of each iTRAQ channel

  5. Normalize the four channels each to 45 μg of protein

  6. Normalize each protein in each channel to the total protein concentration, determined by protein assay, for the immunodepleted plasma with respect to original plasma volume

  7. Compare the corresponding concentrations from the two immunodepletion methods for each protein

Calculation of the Efficiencies of Immunodepletion of Individual Proteins in Two Immunodepletion Experiments

To calculate the immunodepletion efficiencies for removing albumin and immunoglobulins from plasma by the Qproteome method, the values of serum albumin before immunodepletion was stipulated as 44 mg/mL and immunoglobulins were 24 mg/mL. These are representative values from clinical test reports. After immunodepletion, the Qproteome depleted plasma had a protein concentration of 62.4 mg/mL when equated to the original plasma volume. The IgY14+SuperMix depleted plasma had a protein concentration of 7.4 mg/mL when equated to the original plasma volume. These values were used to normalize the measurements of individual protein quantities obtained from EMMOL calculations. The averaged results of the four iTRAQ channels are shown under the header “Average mg/mL in depleted plasma” in Tables 14. The ratio of this value for each protein from the IgY14+SuperMix column immunodepletion compared with the value from the Qproteome immunodepletion produces the values under the header “% immunodepletion”.

Construction of a Plasma Proteome from Two Immunodepletion Experiments

A subset of proteins was each quantified in both immunodepletion approaches. However, a given protein may be more abundant in the depleted proteome of one immunodepletion method than the other, resulting in better peptide statistics for identification and better iTRAQ quantitation accuracy. The better quality of these two results where available for each protein was used to construct a composite continuous proteome for further analysis. Moreover, the comparison of the proteomes of the two immunodepletion methods (Tables 14) was used to identify the medium abundance proteins and the non-targeted proteins removed by the Seppro IgY14+SuperMix column.

RESULTS AND DISCUSSION

A comparison of the proteins in the pooled plasma samples immunodepleted by two procedures is shown in Tables 14. Supporting Information for Publication is the full display of all the tables that are partially presented herein plus the original Mascot output data of the Qproteome proteome and the Seppro IgY14+SuperMix proteome. Although the intended quantity of protein in each label reaction was 45 micrograms, EMMOL calculations suggested that the values of the total peptides for each iTRAQ channel during LC-MS/MS varied by as much as 20%. For the Qproteome experiments, iTRAQ 114, 115, 116, and 117 channels were 40.1, 41.7, 47.6, and 50.6 micrograms, respectively. For the Seppro IgY14+SuperMix experiments, iTRAQ 114, 115, 116, and 117 channels were 39.9, 43.9, 46.6, and 49.9 micrograms, respectively. A 20% bias in an iTRAQ channel would tend to complicate accurate determination of small fold-changes among proteins that differ in expression in two proteomes and confound proteomics pathway analysis. Hence, normalization based on the mass spectrometry measurements of the high abundance proteins tends to restore precision for the iTRAQ comparison.

The resulting proteome contained 412 proteins with quantitation, of which 120 were present in both immunodepletion samples

IgY14 Depletion

The Qproteome and the Seppro IgY14+SuperMix methods removed most of the albumin and the immunoglobulins according to SDS PAGE (data not shown) and the latter matrix removed many other proteins not visualized in SDS PAGE. The IgY14 was designed to remove human serum albumin, IgG, fibrinogen, transferrin, IgA, IgM, haptoglobin, alpha2-macroglubulin, alpha1-acid glycoprotein, alpha1-antitrypsin, apo A-I HDL, apo A-II, complement C3, and apo B. Table 1 illustrates their immunodepletion efficiencies.

SuperMix Depletion

A reasonable indication that a protein has been immunodepleted by the SuperMix resin is that its level has decreased by about 80–90 % after passing through the SuperMix resin. This arbitrary cutoff range seemed suitable for this dataset and revealed that the resin removed most of another 76 proteins (Table 2), including: complement factor B, antithrombin-III, inter-alpha-trypsin inhibitor heavy chain H1, H2, and H4; ceruloplasmin, complement C4-A, vitronectin, hemoglobin subunit alpha, beta and delta, plasma protease C1 inhibitor, prothrombin, angiotensinogen, vitamin D-binding protein, histidine-rich glycoprotein, and alpha-1B-glycoprotein. In all, most of the 20 top-abundance plasma proteins were depleted, but not eliminated, by IgY14+SuperMix column.

Table 2.

Immunodepletion Efficiencies of the 76 Putative Targets of Seppro SuperMix Resin

Immuno-depletion method Title Gene Symbol prot_score prot_mass prot_matches prot_cover emPAI Average mg/mL in depleted plasma % immunodepletion
Qproteome Complement factor B CFB 2265 94629 110 21.6 1.32 0.698
Seppro Complement factor B CFB 102 94629 2 1.8 0.04 0.002 100
Qproteome Inter-alpha-trypsin inhibitor heavy chain H4 ITIH4 2298 109541 95 24.5 0.94 0.576
Seppro Inter-alpha-trypsin inhibitor heavy chain H4 ITIH4 51 109573 3 1.8 0.07 0.004 100
Qproteome Inter-alpha-trypsin inhibitor heavy chain H2 ITIH2 1671 116337 55 15 0.55 0.357
Seppro Inter-alpha-trypsin inhibitor heavy chain H2 ITIH2 198 116364 5 2.3 0.03 0.002 100
Qproteome Complement C4-A C4A 4364 205487 181 24.2 1.03 1.178
Seppro Complement C4-A C4A 168 205487 8 2.5 0.07 0.008 100
Qproteome Antithrombin-III SERPINC1 5520 58501 167 46.3 2.43 0.794
Seppro Antithrombin-III SERPINC1 160 58501 4 5 0.13 0.004 100
Qproteome Ceruloplasmin CP 3498 132638 124 25.1 0.93 0.687
Seppro Ceruloplasmin CP 74 132638 2 2.1 0.06 0.004 99
Qproteome Vitronectin VTN 1661 58096 47 20.7 0.86 0.281
Seppro Vitronectin VTN 210 58096 4 3.1 0.06 0.002 99
Qproteome Hemoglobin subunit alpha HBA1 945 17034 37 35.2 2.43 0.231
Seppro Hemoglobin subunit alpha HBA1 55 17034 2 8.5 0.23 0.002 99
Qproteome Plasma protease C1 inhibitor SERPING1 2382 59671 105 23.8 1.33 0.444
Seppro Plasma protease C1 inhibitor SERPING1 124 59671 3 5.4 0.13 0.004 99
Qproteome Vitamin D-binding protein GC 1757 61010 83 23.2 1.03 0.351
Seppro Vitamin D-binding protein GC 112 61010 3 6.3 0.19 0.006 99
Qproteome Prothrombin F2 1017 75798 28 12.5 0.4 0.169
Seppro Prothrombin F2 49 75798 2 1.6 0.05 0.002 99
Qproteome Angiotensinogen AGT 977 56576 42 13.2 0.47 0.149
Seppro Angiotensinogen AGT 278 56576 9 2.7 0.07 0.002 99
Qproteome Inter-alpha-trypsin inhibitor heavy chain H1 ITIH1 2049 108699 55 13.4 0.4 0.242
Seppro Inter-alpha-trypsin inhibitor heavy chain H1 ITIH1 179 108699 5 4.2 0.07 0.004 99
Qproteome Hemoglobin subunit beta HBB 2335 17832 71 67.3 7.62 0.761
Seppro Hemoglobin subunit beta HBB 629 17832 15 44.9 1.66 0.016 98
Qproteome Histidine-rich glycoprotein HRG 945 63825 36 17.9 0.57 0.203
Seppro Histidine-rich glycoprotein HRG 60 63825 4 5 0.12 0.004 98
Qproteome Alpha-1B-glycoprotein A1BG 249 56538 24 9.5 0.21 0.066
Seppro Alpha-1B-glycoprotein A1BG 38 56538 3 2 0.07 0.002 98
Qproteome Hemoglobin subunit delta HBD 1079 17889 42 50.3 2.94 0.294
Seppro Hemoglobin subunit delta HBD 87 17889 8 32 0.8 0.008 98
Qproteome Afamin AFM 1139 78456 40 20.7 0.59 0.259
Seppro Afamin AFM 775 78456 31 8.2 0.2 0.008 97
Qproteome Complement component C7 C7 200 103711 5 3 0.07 0.039
Seppro Complement component C7 C7 219 103711 5 1.8 0.04 0.002 96
Qproteome Alpha-1-antitrypsin SERPINA1 16719 51922 799 64.6 8.86 2.550
Seppro Alpha-1-antitrypsin SERPINA1 4261 51922 220 53.1 4.66 0.129 96
Qproteome C4b-binding protein alpha chain C4BPA 770 74086 23 8.9 0.41 0.169
Seppro C4b-binding protein alpha chain C4BPA 395 74086 9 3.9 0.16 0.006 96
Qproteome C4b-binding protein beta chain C4BPB 77 32046 4 4 0.25 0.045
Seppro C4b-binding protein beta chain C4BPB 51 32046 2 4 0.12 0.002 96
Qproteome Dermcidin DCD 56 12976 2 10 0.3 0.022
Seppro Dermcidin DCD 77 12976 2 10 0.3 0.002 95
Qproteome Kinesin-like protein KIF21B KIF21B 77 199879 30 1 0.04 0.045
Seppro Kinesin-like protein KIF21B KIF21B 104 199879 49 1 0.02 0.002 95
Qproteome Apolipoprotein C-III APOC3 1489 11854 56 55.6 6.52 0.437
Seppro Apolipoprotein C-III APOC3 1199 11854 49 53.5 3.23 0.021 95
Qproteome Complement C1q subcomponent subunit B C1QB 238 28976 7 5.5 0.28 0.045
Seppro Complement C1q subcomponent subunit B C1QB 276 29238 10 5.5 0.13 0.002 95
Qproteome Heparin cofactor 2 SERPIND1 917 62105 29 18.4 0.5 0.174
Seppro Heparin cofactor 2 SERPIND1 292 62105 10 9.6 0.26 0.009 94
Qproteome Ficolin-2 FCN2 152 36885 7 9.6 0.34 0.070
Seppro Ficolin-2 FCN2 147 36885 8 6.7 0.21 0.004 94
Qproteome Interleukin-12 receptor beta-1 chain IL12RB1 34 77456 6 2.3 0.05 0.022
Seppro Interleukin-12 receptor subunit beta-1 IL12RB1 36 77456 12 0.9 0.05 0.002 94
Qproteome Serum amyloid A-4 protein SAA4 355 16007 12 26.2 0.92 0.082
Seppro Serum amyloid A-4 protein SAA4 250 16004 7 8.5 0.55 0.005 94
Qproteome Phospholipid transfer protein PLTP 182 57526 5 8.1 0.29 0.093
Seppro Phospholipid transfer protein PLTP 381 57526 17 5.3 0.21 0.007 94
Qproteome Complement factor I CFI 306 74412 12 9.3 0.21 0.088
Seppro Complement factor I CFI 154 74412 14 4.6 0.16 0.006 93
Qproteome Tripartite motif-containing protein 37 TRIM37 46 116552 2 0.7 0.03 0.020
Seppro E3 ubiquitin-protein ligase TRIM37 TRIM37 70 116552 6 1.7 0.03 0.002 93
Qproteome Complement component C9 C9 759 70091 29 8.4 0.36 0.140
Seppro Complement component C9 C9 337 70091 16 8.9 0.29 0.011 93
Qproteome R3H domain-containing protein 1 R3HDM1 63 129519 2 0.6 0.03 0.022
Seppro R3H domain-containing protein 1 R3HDM1 54 129519 2 0.6 0.03 0.002 93
Qproteome Complement C1s subcomponent C1S 794 83650 29 14.2 0.54 0.251
Seppro Complement C1s subcomponent C1S 766 83650 39 13.7 0.48 0.022 91
Qproteome Apolipoprotein C-I APOC1 627 10767 46 32.5 3.85 0.232
Seppro Apolipoprotein C-I APOC1 373 10767 23 32.5 3.85 0.022 91
Qproteome Alpha-1-antichymotrypsin SERPINA3 4424 51682 158 38.3 1.65 0.475
Seppro Alpha-1-antichymotrypsin SERPINA3 2975 51682 114 40.4 2.27 0.062 90
Qproteome Apolipoprotein C-II APOC2 247 12285 13 45.5 2.05 0.142
Seppro Apolipoprotein C-II APOC2 313 12285 12 57.4 2.05 0.014 90
Qproteome UPF0505 protein C16orf62 C16orf62 32 120515 3 2.1 0.03 0.020
Seppro UPF0505 protein C16orf62 C16orf62 82 120515 4 2.1 0.03 0.002 90
Qproteome Obscurin OBSCN 217 928336 71 0.3 0.01 0.051
Seppro Obscurin OBSCN 87 928336 34 0.6 0.01 0.005 90
Qproteome Apolipoprotein D APOD 147 23276 9 10.1 0.35 0.045
Seppro Apolipoprotein D APOD 128 23276 5 10.1 0.35 0.004 90
Qproteome Complement C1q subcomponent subunit C C1QC 155 27859 3 7.3 0.14 0.022
Seppro Complement C1q subcomponent subunit C C1QC 174 27859 3 7.3 0.14 0.002 90
Qproteome A disintegrin and metalloproteinase with thrombospondin motifs 12 ADAMTS12 37 196369 8 0.9 0.02 0.022
Seppro A disintegrin and metalloproteinase with thrombospondin motifs 12 ADAMTS12 33 196355 6 0.9 0.02 0.002 89
Qproteome Phosphatidylinositol-4,5- bisphosphate 3-kinase catalytic subunit beta isoform PIK3CB 61 135868 26 1.2 0.03 0.023
Seppro Phosphatidylinositol-4,5- bisphosphate 3-kinase catalytic subunit beta isoform PIK3CB 74 135868 25 0.7 0.03 0.002 89
Qproteome Uncharacterized protein C2orf77 C2orf77 47 77323 3 1.1 0.05 0.022
Seppro Uncharacterized protein C2orf77 C2orf77 100 77323 22 1.1 0.05 0.002 89
Qproteome Vacuolar fusion protein MON1homolog A MON1A 52 64884 8 1.4 0.06 0.022
Seppro Vacuolar fusion protein MON1homolog A MON1A 62 64884 9 1.4 0.06 0.002 89
Qproteome T-complex protein 1 subunit zeta-2 CCT6B 127 65072 61 1.5 0.06 0.021
Seppro T-complex protein 1 subunit zeta-2 CCT6B 42 65072 23 2.8 0.06 0.002 88
Qproteome ATP-binding cassette sub-family B member 9 ABCB9 126 88768 7 0.9 0.04 0.020
Seppro ATP-binding cassette sub-family B member 9 ABCB9 232 88768 17 0.9 0.04 0.002 88
Qproteome Keratin, type I cytoskeletal 9 KRT9 221 66211 13 7.7 0.24 0.087
Seppro Keratin, type I cytoskeletal 9 KRT9 138 66146 13 5.8 0.24 0.008 88
Qproteome Apolipoprotein L1 APOL1 573 47751 16 9.5 0.46 0.123
Seppro Apolipoprotein L1 APOL1 517 47751 25 11.6 0.57 0.015 87
Qproteome Protein AMBP AMBP 934 42624 36 25 0.52 0.125
Seppro Protein AMBP AMBP 1308 42624 37 25 0.66 0.015 87
Qproteome Fibronectin FN1 1774 277418 49 7.4 0.17 0.261
Seppro Fibronectin FN1 1648 277437 44 6.7 0.19 0.029 86
Qproteome Hyaluronan-binding protein 2 HABP2 95 70504 2 1.6 0.11 0.043
Seppro Hyaluronan-binding protein 2 HABP2 349 70504 12 3.8 0.17 0.006 86
Qproteome Complement-activating component of Ra-reactive factor MASP1 161 86325 8 2.7 0.09 0.043
Seppro Mannan-binding lectin serine protease 1 MASP1 485 86325 20 2.7 0.13 0.006 86
Qproteome Isocitrate dehydrogenase [NAD]subunit alpha, mitochondrial IDH3A 131 43625 12 2.7 0.09 0.022
Seppro Isocitrate dehydrogenase [NAD]subunit alpha, mitochondrial IDH3A 52 43625 3 2.7 0.18 0.004 84
Qproteome Keratin, type I cytoskeletal 10 KRT10 728 63161 21 14.4 0.41 0.144
Seppro Keratin, type I cytoskeletal 10 KRT10 840 62478 29 17 0.68 0.023 84
Qproteome Apolipoprotein E APOE 1982 38263 45 41.3 2.71 0.579
Seppro Apolipoprotein E APOE 2852 38263 100 46.1 5.5 0.113 83
Qproteome Gelsolin GSN 695 92672 16 11.8 0.22 0.114
Seppro Gelsolin GSN 808 92672 24 13.9 0.37 0.019 83
Qproteome CD5 antigen-like CD5L 88 42052 2 4 0.09 0.022
Seppro CD5 antigen-like CD5L 110 42052 4 6.1 0.19 0.004 82
Qproteome Keratin, type II cytoskeletal 2epidermal KRT2 359 71298 12 7.5 0.23 0.092
Seppro Keratin, type II cytoskeletal 2epidermal KRT2 362 70866 18 11.7 0.43 0.016 80
Qproteome Hypoxia up-regulated protein 1 HYOU1 99 122734 2 1.9 0.03 0.021
Seppro Hypoxia up-regulated protein 1 HYOU1 392 122734 13 6 0.09 0.006 80
Qproteome Corticosteroid-binding globulin SERPINA6 259 47877 9 7.2 0.25 0.067
Seppro Corticosteroid-binding globulin SERPINA6 1344 47877 42 12.8 0.57 0.015 79
Qproteome Inter-alpha-trypsin inhibitor heavy chain H3 ITIH3 262 108718 7 4 0.14 0.085
Seppro Inter-alpha-trypsin inhibitor heavy chain H3 ITIH3 513 108718 16 12 0.31 0.018 79
Prostaglandin-H2 D-isomerase PTGDS 139 22829 3 8.9 0.17 0.022
Seppro Prostaglandin-H2 D-isomerase PTGDS 248 22829 15 12.6 0.36 0.004 79
Qproteome L-lactate dehydrogenase B chain LDHB 55 40791 2 7.5 0.19 0.043
Seppro L-lactate dehydrogenase B chain LDHB 445 40791 11 12.3 0.42 0.009 79
Qproteome Keratin, type II cytoskeletal 1 KRT1 380 70184 23 13.5 0.51 0.197
Seppro Keratin, type II cytoskeletal 1 KRT1 846 70349 40 19.9 1.16 0.043 78
Qproteome Selenoprotein P SEPP1 163 48016 8 8.1 0.25 0.067
Seppro Selenoprotein P SEPP1 536 48016 25 16 0.57 0.015 77
Qproteome Cystatin-C CST3 125 17170 2 11 0.23 0.022
Seppro Cystatin-C CST3 379 17170 10 18.5 0.5 0.005 77
Qproteome Transthyretin TTR 2397 17288 141 57.1 4.03 0.390
Seppro Transthyretin TTR 2970 17288 191 73.5 8.21 0.078 76
Qproteome Cartilage oligomeric matrix protein COMP 131 88457 2 2.6 0.04 0.020
Seppro Cartilage oligomeric matrix protein COMP 482 88457 9 5.8 0.13 0.006 76
Qproteome Coagulation factor XIII A chain F13A1 195 89348 8 6.1 0.13 0.065
Seppro Coagulation factor XIII A chain F13A1 697 89348 22 10.5 0.28 0.014 76
Qproteome Plasma serine protease inhibitor SERPINA5 90 49389 4 5.9 0.16 0.044
Seppro Plasma serine protease inhibitor SERPINA5 221 49389 10 11.1 0.34 0.009 75
Qproteome Monocyte differentiation antigen CD14 CD14 162 42119 4 7.2 0.19 0.045
Seppro Monocyte differentiation antigen CD14 CD14 581 42119 21 14.7 0.53 0.012 75
Qproteome Coagulation factor V F5 249 271099 9 2.1 0.06 0.091
Seppro Coagulation factor V F5 809 271099 33 5.5 0.16 0.023 74
Qproteome Serum paraoxonase/arylesterase 1 PON1 1071 42921 25 18 0.29 0.070
Seppro Serum paraoxonase/arylesterase 1 PON1 1409 42921 59 27 0.95 0.022 74

Off-target Protein Removal

Many abundant proteins in serum double as carrier proteins with high affinity binding but this phenomenon has only been studied for a few abundant proteins. The carrier role of serum albumin for small molecules26 and proteins 9 is well known. Another example is alpha 2 macroglobulin, the most abundant globulin in the blood that is not an immunoglobulin, is an effective protease inhibitor, blocks fibrinolysis and is a known carrier of prostate-specific antigen PSA27, 28. However, it also binds to numerous growth factors and cytokines, such as platelet-derived growth factor, basic fibroblast growth factor, TGF-β, insulin, and IL-1β. Driven by affinity removal by the high-abundance carrier proteins, these off-target depleted proteins shown in Table 3 were not seen in the highly enriched Seppro IgY14+SuperMix fraction in spite of their relatively ease of identification, with many peptides each, in the Qproteome fraction. Table 3 illustrates 65 proteins of this class. Mass spectrometry under-sampling is unlikely to explain their absence in the Seppro IgY14+SuperMix fraction; and we designate these as off-target removals. Other explanations are plausible, including the possibility that some of these proteins were exceptionally immunogenic in production of SuperMix matrix. In all, about 155 plasma proteins, 38% of the plasma proteome in protein number in this report and 94% of plasma protein in mass, were removed from the plasma by the Seppro IgY14+SuperMix resin.

Table 3.

65 Proteins Putatively Absorbed by Protein-protein Interactions on Seppro IgY14+SuperMix Column

Immuno-depletion method Title Gene Symbol prot_score prot_mass prot_matches prot_cover emPAI Average mg/mL in depleted plasma
Qproteome Hemopexin HPX 3086 55555 219 32.3 2.02 0.628
Qproteome Phosphatidylinositol-4-phosphate 3-kinase C2domain-containing gamma polypeptide PIK3C2G 106 182657 102 1.3 0.04 0.041
Qproteome Centrosomal protein of 164 kDa CEP164 94 179858 78 0.4 0.02 0.020
Qproteome Kinesin-like protein KIF16B KIF16B 41 169204 77 1.5 0.02 0.019
Qproteome Alpha-2-HS-glycoprotein AHSG 557 42548 76 19.1 0.97 0.232
Qproteome Dynein heavy chain 17, axonemal DNAH17 214 563266 74 0.2 0.01 0.031
Qproteome Complement factor H CFH 834 155353 57 11.2 0.26 0.225
Qproteome Elongation factor Ts, mitochondrial TSFM 143 39169 52 2.2 0.1 0.022
Qproteome Protein FAM171B FAM171B 61 100849 49 1.4 0.04 0.022
Qproteome Kininogen-1 KNG1 1267 80489 48 20.3 0.72 0.326
Qproteome WD repeat-containing protein 19 WDR19 47 166391 38 0.4 0.02 0.018
Qproteome Plasminogen PLG 521 100452 34 5.1 0.2 0.112
Qproteome Beta-2-glycoprotein 1 APOH 282 44051 27 7.8 0.28 0.069
Qproteome Serum amyloid P-component APCS 780 27358 22 14.3 0.68 0.104
Qproteome Transmembrane protein 201 TMEM201 42 76326 20 1.2 0.05 0.021
Qproteome Complement C5 C5 504 207189 16 3.7 0.11 0.127
Qproteome Complement component C6 C6 663 118022 16 6.7 0.13 0.085
Qproteome Dynein heavy chain 8, axonemal DNAH8 59 565682 16 0.3 0.01 0.032
Qproteome Negative elongation factor B COBRA1 87 71759 16 2.2 0.05 0.020
Qproteome Inner centromere protein INCENP 42 117004 16 2.7 0.03 0.020
Qproteome Formin-like protein 1 FMNL1 32 132291 12 0.6 0.03 0.023
Qproteome Vacuolar protein-sorting-associated protein 36 VPS36 40 48397 12 1.6 0.08 0.022
Qproteome Actin, cytoplasmic 1 ACTB 474 44934 11 15.5 0.49 0.125
Qproteome Pericentrin PCNT 48 411049 11 0.5 0.01 0.023
Qproteome Complement component C8 beta chain C8B 462 73758 10 7.1 0.22 0.090
Qproteome Tumor necrosis factor receptor superfamily member 8 TNFRSF8 86 69388 10 1 0.05 0.019
Qproteome Staphylococcal nuclease domain-containing protein 1 SND1 45 111120 9 2.3 0.07 0.044
Qproteome SET domain-containing protein 3 SETD3 37 73898 9 0.8 0.05 0.021
Qproteome CAP-Gly domain-containing linker protein 1 CLIP1 29 183824 9 1 0.02 0.021
Qproteome Retinol-binding protein 4 RBP4 349 25067 8 19.4 0.53 0.075
Qproteome Alpha-2-antiplasmin SERPINF2 396 57755 8 6.5 0.21 0.068
Qproteome Heparanase HPSE 46 67043 8 2.4 0.06 0.023
Qproteome Actin-binding LIM protein 1 ABLIM1 32 97150 7 2.6 0.08 0.043
Qproteome Trichoplein keratin filament-binding protein TCHP 69 67056 7 2.8 0.06 0.023
Qproteome Uncharacterized protein FLJ44048 - 59 205325 7 1.1 0.02 0.023
Qproteome ADAMTS-like protein 1 ADAMTSL1 29 64754 7 0.4 0.06 0.022
Qproteome P protein OCA2 46 96527 7 0.7 0.04 0.022
Qproteome Kappa-actin FKSG30 65 45645 6 5.5 0.17 0.044
Qproteome Probable E3 ubiquitin-protein ligase HECTD3 HECTD3 35 104187 6 1.4 0.04 0.023
Qproteome Cyclin-dependent kinase-like 1 CDKL1 39 46856 6 2 0.08 0.020
Qproteome Golgi resident protein GCP60 ACBD3 40 65308 6 1.3 0.06 0.020
Qproteome ANKRD26-like family C member 1A A26C1A 161 135563 5 2.6 0.08 0.061
Qproteome Plasma kallikrein KLKB1 97 79485 5 1.7 0.05 0.022
Qproteome Zinc finger protein 37A ZNF37A 47 74724 5 1.4 0.05 0.021
Qproteome 26S proteasome non-ATPase regulatory subunit 5 PSMD5 40 59730 5 1.4 0.06 0.020
Qproteome Ras GTPase-activating protein SynGAP SYNGAP1 33 158671 5 0.6 0.02 0.017
Qproteome N-acetylmuramoyl-L-alanine amidase PGLYRP2 161 64621 4 8 0.18 0.065
Qproteome MAP7 domain-containing protein 1 MAP7D1 42 102617 4 1.5 0.04 0.022
Qproteome Ladybird homeobox corepressor 1-like protein CORL2 44 111363 4 0.8 0.03 0.019
Qproteome Eukaryotic translation initiation factor 3 subunit J EIF3J 29 34779 4 3.1 0.11 0.021
Qproteome Syntaxin-1B STX1B 41 37343 4 2.1 0.1 0.021
Qproteome UDP-N-acetylglucosamine--peptide N-acetylglucosaminyltransferase 110 kDa subunit OGT 47 125886 4 0.6 0.03 0.021
Qproteome Envoplakin EVPL 38 252963 4 0.4 0.01 0.014
Qproteome Complement factor H-related protein 1 CFHR1 115 41216 3 6.7 0.19 0.043
Qproteome DENN domain-containing protein 4C DENND4C 50 204270 3 0.7 0.02 0.023
Qproteome Disks large homolog 5 DLG5 36 233891 3 0.7 0.02 0.026
Qproteome Putative myosin-XVB MYO15B 47 176166 3 0.5 0.02 0.020
Qproteome 1-phosphatidylinositol-4,5-bisphosphate phosphodiesterase gamma-1 PLCG1 42 160269 3 0.5 0.02 0.018
Qproteome Lebercilin-like protein LCA5L 49 88817 2 1.3 0.04 0.020
Qproteome FACT complex subunit SPT16 SUPT16H 28 135972 2 0.7 0.03 0.023
Qproteome Complement component C8 gamma chain C8G 120 23299 2 7.4 0.16 0.021
Qproteome Complement component C8 alpha chain C8A 75 71587 2 1.4 0.05 0.020
Qproteome Structural maintenance of chromosomes protein 1B SMC1B 32 165103 2 0.7 0.02 0.018
Qproteome Platelet-activating factor acetylhydrolase PLA2G7 44 54825 2 2.3 0.07 0.021
Qproteome Tubulin polyglutamylase TTLL5 TTLL5 41 155257 2 0.5 0.02 0.017

Proteins made Visible by Immunodepletion by IgY14+SuperMix

276 proteins were visible only in the Seppro IgY14+SuperMix sample (Supporting Information for Publication and partially in Table 4). Most of these proteins are of low abundance in the plasma.

Continuous Plasma Proteome is best Constructed from Two Immunodepletion Approaches

In general, proteins depleted by the Seppro IgY14+SuperMix column were analyzed with better peptide coverage in the Qproteome method. The opposite is true for proteins enriched by the IgY14+SuperMix column. Using EMMOL normalization to calculate the approximate concentrations of each protein in the plasma and the availability of two immunodepletion results, we combined the two proteomes to form a continuous protein concentration range of 5 logs. For each protein that can be quantified in the proteome of each immunodepletion protocol, the value with higher confidence level, because of higher peptide coverage (emPAI), was selected. A portion of this proteome is presented in Table 4 and in whole in Supporting Information for Publication.

Why is Assessment of the On-target and Off-target Removed Proteins Important to Blood Biomarker Studies?

Immunodepletion can be conveniently performed once on a pool of plasma of cases and then once on a pool of the plasma of the controls. That approach is expedient but variation in the efficiencies of the immunodepletion process from run to run between cases and controls can increase the potential for false discovery. The performance of immunodepletion may vary in different laboratories depending on the composition of the immunodepletion matrix used, the chosen ratio of plasma volume to immunodepletion column capacity, and the age of the column. Our comparison reported the overall immunodepletion efficiency of the combined resin for each protein in the proteome. Thus a useful database that indicates which proteins are at risk for false positives is provided. Based on our quantitation of the immunodepletion efficiency of each individual protein, we hypothesize that those proteins which have the highest percentage immunodepletion are at the highest risk of being identified as false-positives in differential proteomic studies. For example, 98% removal of a protein in cases versus 96% removal in controls may suggest one fold change incorrectly in the residual proteins when the two groups were equal in the original plasma. However, the design of our study does not allow us to identify individual at-risk proteins as false positives or false negatives. Some of the removed off-target proteins can be relatively low abundance proteins. Thus variation in their detection or quantification in comparator groups may cause them to be misinterpreted as disease biomarkers. On the other hand, we noticed that proteins that have been depleted by less than 50% in the Seppro IgY14+SuperMix immunodepletion method still report protein fold-changes comparable to the Qproteome method that did not deplete those proteins.

In summary, the current study illustrates that caution needs to be exercised in experiments involving immunodepletion. Nonetheless, after proper accounting for off-target protein loss, immunodepletion proteomics by two methods can provide accurate comparison of both high-abundance and low-abundance proteins. One method should account for the highly-abundant proteins and a separate method should more exhaustively deplete the less-abundant proteins.

Supplementary Material

Supporting Information for Publication

Acknowledgments

We thank Brian Searle for critical reading of this manuscript and Dr. Karen Kawarta, Sigma-Aldrich, for performing the immunodepletion procedures.

This work was partially supported by the National Institutes of Health (NIH) grant RC2 HL101713, the National Cancer Institute, Work Assignment #16 of N01-CN-43309, the Driskill Foundation, Institutional Core Grant P30CA06927, Tobacco Settlement Funds from the Commonwealth of Pennsylvania, the Pew Charitable Trust, and the Kresge Foundation.

The abbreviations used are

COPD

Chronic Obstructive Pulmonary Disease

iTRAQ

Isobaric Tag for Relative and Absolute Quantitation

IEF

Isoelectric Focusing

IPG

Immobilized Polyacrylamide Gel

EMMOL

emPAI-Molecular Weight method of proteome normalization

Footnotes

Conflicts of interests

Dr. Dian Er Chen who performed the immunodepletion steps was employed by Sigma-Aldrich that sells the IgY14+SuperMix column. Other authors declare no competing interest and no support by any company in connection with this work or future related work.

This article contains Supporting Information for Publication.

References

  • 1.Anderson NL, Polanski M, Pieper R, Gatlin T, Tirumalai RS, Conrads TP, Veenstra TD, Adkins JN, Pounds JG, Fagan R, Lobley A. The human plasma proteome: a nonredundant list developed by combination of four separate sources. Mol Cell Proteomics. 2004;3(4):311–26. doi: 10.1074/mcp.M300127-MCP200. [DOI] [PubMed] [Google Scholar]
  • 2.Burgess-Cassler A, Johansen JJ, Kendrick N. Immunodepletion of albumin from human serum samples. Clinica chimica acta; international journal of clinical chemistry. 1989;183(3):359–65. doi: 10.1016/0009-8981(89)90371-9. [DOI] [PubMed] [Google Scholar]
  • 3.Beer LA, Tang HY, Sriswasdi S, Barnhart KT, Speicher DW. Systematic discovery of ectopic pregnancy serum biomarkers using 3-D protein profiling coupled with label-free quantitation. Journal of proteome research. 2011;10(3):1126–38. doi: 10.1021/pr1008866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Maccarrone G, Milfay D, Birg I, Rosenhagen M, Holsboer F, Grimm R, Bailey J, Zolotarjova N, Turck CW. Mining the human cerebrospinal fluid proteome by immunodepletion and shotgun mass spectrometry. Electrophoresis. 2004;25(14):2402–12. doi: 10.1002/elps.200305909. [DOI] [PubMed] [Google Scholar]
  • 5.Shen Z, Want EJ, Chen W, Keating W, Nussbaumer W, Moore R, Gentle TM, Siuzdak G. Sepsis plasma protein profiling with immunodepletion, three-dimensional liquid chromatography tandem mass spectrometry, and spectrum counting. Journal of proteome research. 2006;5(11):3154–60. doi: 10.1021/pr060327k. [DOI] [PubMed] [Google Scholar]
  • 6.Koutroukides TA, Guest PC, Leweke FM, Bailey DM, Rahmoune H, Bahn S, Martins-de-Souza D. Characterization of the human serum depletome by label-free shotgun proteomics. Journal of separation science. 2011;34(13):1621–6. doi: 10.1002/jssc.201100060. [DOI] [PubMed] [Google Scholar]
  • 7.Smith MP, Wood SL, Zougman A, Ho JT, Peng J, Jackson D, Cairns DA, Lewington AJ, Selby PJ, Banks RE. A systematic analysis of the effects of increasing degrees of serum immunodepletion in terms of depth of coverage and other key aspects in top-down and bottom-up proteomic analyses. Proteomics. 2011;11(11):2222–35. doi: 10.1002/pmic.201100005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Scumaci D, Gaspari M, Saccomanno M, Argiro G, Quaresima B, Faniello CM, Ricci P, Costanzo F, Cuda G. Assessment of an ad hoc procedure for isolation and characterization of human albuminome. Analytical biochemistry. 2011;418(1):161–3. doi: 10.1016/j.ab.2011.06.032. [DOI] [PubMed] [Google Scholar]
  • 9.Gundry RL, Fu Q, Jelinek CA, Van Eyk JE, Cotter RJ. Investigation of an albumin-enriched fraction of human serum and its albuminome. Proteomics Clinical applications. 2007;1(1):73–88. doi: 10.1002/prca.200600276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Juhasz P, Lynch M, Sethuraman M, Campbell J, Hines W, Paniagua M, Song L, Kulkarni M, Adourian A, Guo Y, Li X, Martin S, Gordon N. Semi-targeted plasma proteomics discovery workflow utilizing two-stage protein depletion and off-line LC-MALDI MS/MS. Journal of proteome research. 2011;10(1):34–45. doi: 10.1021/pr100659e. [DOI] [PubMed] [Google Scholar]
  • 11.Foreman MG, Zhang L, Murphy J, Hansel NN, Make B, Hokanson JE, Washko G, Regan EA, Crapo JD, Silverman EK, DeMeo DL. Early-onset chronic obstructive pulmonary disease is associated with female sex, maternal factors, and African American race in the COPDGene Study. American journal of respiratory and critical care medicine. 2011;184(4):414–20. doi: 10.1164/rccm.201011-1928OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Han MK, Kazerooni EA, Lynch DA, Liu LX, Murray S, Curtis JL, Criner GJ, Kim V, Bowler RP, Hanania NA, Anzueto AR, Make BJ, Hokanson JE, Crapo JD, Silverman EK, Martinez FJ, Washko GR. Chronic obstructive pulmonary disease exacerbations in the COPDGene study: associated radiologic phenotypes. Radiology. 2011;261(1):274–82. doi: 10.1148/radiol.11110173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Kim DK, Jacobson FL, Washko GR, Casaburi R, Make BJ, Crapo JD, Silverman EK, Hersh CP. Clinical and radiographic correlates of hypoxemia and oxygen therapy in the COPDGene study. Respiratory medicine. 2011;105(8):1211–21. doi: 10.1016/j.rmed.2011.02.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kim V, Han MK, Vance GB, Make BJ, Newell JD, Hokanson JE, Hersh CP, Stinson D, Silverman EK, Criner GJ. The chronic bronchitic phenotype of COPD: an analysis of the COPDGene Study. Chest. 2011;140(3):626–33. doi: 10.1378/chest.10-2948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Rambod M, Porszasz J, Make BJ, Crapo JD, Casaburi R. Six-minute walk distance predictors, including CT scan measures, in the COPDGene cohort. Chest. 2012;141(4):867–75. doi: 10.1378/chest.11-0870. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Regan EA, Hokanson JE, Murphy JR, Make B, Lynch DA, Beaty TH, Curran-Everett D, Silverman EK, Crapo JD. Genetic epidemiology of COPD (COPDGene) study design. COPD. 2010;7(1):32–43. doi: 10.3109/15412550903499522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wan ES, Hokanson JE, Murphy JR, Regan EA, Make BJ, Lynch DA, Crapo JD, Silverman EK. Clinical and radiographic predictors of GOLD-unclassified smokers in the COPDGene study. American journal of respiratory and critical care medicine. 2011;184(1):57–63. doi: 10.1164/rccm.201101-0021OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Washko GR, Lynch DA, Matsuoka S, Ross JC, Umeoka S, Diaz A, Sciurba FC, Hunninghake GM, San Jose Estepar R, Silverman EK, Rosas IO, Hatabu H. Identification of early interstitial lung disease in smokers from the COPDGene Study. Academic radiology. 2010;17(1):48–53. doi: 10.1016/j.acra.2009.07.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Geoui T, Urlaub H, Plessmann U, Porschewski P. Extraction of proteins from formalin-fixed, paraffin-embedded tissue using the Qproteome extraction technique and preparation of tryptic peptides for liquid chromatography/mass spectrometry analysis. In: Ausubel Frederick M, et al., editors. Current protocols in molecular biology. Unit 10. Chapter 10. 2010. pp. 27pp. 1–12. [DOI] [PubMed] [Google Scholar]
  • 20.Qian WJ, Kaleta DT, Petritis BO, Jiang H, Liu T, Zhang X, Mottaz HM, Varnum SM, Camp DG, 2nd, Huang L, Fang X, Zhang WW, Smith RD. Enhanced detection of low abundance human plasma proteins using a tandem IgY12-SuperMix immunoaffinity separation strategy. Molecular & cellular proteomics: MCP. 2008;7(10):1963–73. doi: 10.1074/mcp.M800008-MCP200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kim PD, Patel BB, Yeung AT. Isobaric Labeling and Data Normalization without Requiring Protein Quantitation. Journal of biomolecular techniques: JBT. 2012;23(1):11–23. doi: 10.7171/jbt.12-2301-002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Lengqvist J, Uhlen K, Lehtio J. iTRAQ compatibility of peptide immobilized pH gradient isoelectric focusing. Proteomics. 2007;7(11):1746–52. doi: 10.1002/pmic.200600782. [DOI] [PubMed] [Google Scholar]
  • 23.Eriksson H, Lengqvist J, Hedlund J, Uhlen K, Orre LM, Bjellqvist B, Persson B, Lehtio J, Jakobsson PJ. Quantitative membrane proteomics applying narrow range peptide isoelectric focusing for studies of small cell lung cancer resistance mechanisms. Proteomics. 2008;8(15):3008–18. doi: 10.1002/pmic.200800174. [DOI] [PubMed] [Google Scholar]
  • 24.Ke E, Patel BB, Liu T, Li XM, Haluszka O, Hoffman JP, Ehya H, Young NA, Watson JC, Weinberg DS, Nguyen MT, Cohen SJ, Meropol NJ, Litwin S, Tokar JL, Yeung AT. Proteomic analyses of pancreatic cyst fluids. Pancreas. 2009;38(2):e33–42. doi: 10.1097/MPA.0b013e318193a08f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ishihama Y, Oda Y, Tabata T, Sato T, Nagasu T, Rappsilber J, Mann M. Exponentially modified protein abundance index (emPAI) for estimation of absolute protein amount in proteomics by the number of sequenced peptides per protein. Mol Cell Proteomics. 2005;4(9):1265–72. doi: 10.1074/mcp.M500061-MCP200. [DOI] [PubMed] [Google Scholar]
  • 26.Fanali G, di Masi A, Trezza V, Marino M, Fasano M, Ascenzi P. Human serum albumin: from bench to bedside. Molecular aspects of medicine. 2012;33(3):209–90. doi: 10.1016/j.mam.2011.12.002. [DOI] [PubMed] [Google Scholar]
  • 27.Lin VK, Wang SY, Boetticher NC, Vazquez DV, Saboorian H, McConnell JD, Roehrborn CG. Alpha(2) macroglobulin, a PSA binding protein, is expressed in human prostate stroma. The Prostate. 2005;63(3):299–308. doi: 10.1002/pros.20183. [DOI] [PubMed] [Google Scholar]
  • 28.Zhang WM, Finne P, Leinonen J, Salo J, Stenman UH. Determination of prostate-specific antigen complexed to alpha(2)-macroglobulin in serum increases the specificity of free to total PSA for prostate cancer. Urology. 2000;56(2):267–72. doi: 10.1016/s0090-4295(00)00609-9. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information for Publication

RESOURCES