Copy Number Analysis of the Murine Platelet Proteome Spanning the Complete Abundance Range
2014; Elsevier BV; Volume: 13; Issue: 12 Linguagem: Inglês
10.1074/mcp.m114.038513
ISSN1535-9484
AutoresMarlis Zeiler, Markus Moser, Matthias Mann,
Tópico(s)Peptidase Inhibition and Analysis
ResumoKnowledge of the identity and quantity of expressed proteins of a cell type is a prerequisite for a complete understanding of its molecular functions. Mass-spectrometry-based proteomics has allowed the identification of the entire protein complement of yeast and the close-to-complete set of proteins expressed in mammalian cell lines. Using recent technological advances, we here characterized the proteome of murine platelets, key actors in mediating hemostasis and thrombosis. We accurately measured the absolute protein concentrations of 13 platelet proteins using SILAC-protein epitope signature tags and used them as reference points to estimate the copy numbers of all proteins of the platelet proteome. To distinguish contaminants such as plasma or erythrocyte proteins from true platelet proteins, we monitored protein abundance profiles across multiple purification steps. In total, we absolutely quantified 4,400 platelet proteins, with estimated copy numbers ranging from less than 10 to about a million per cell. Stoichiometries derived from our data correspond well with previous studies. Our study provides a close-to-complete reference map of platelet proteins that will be useful to the community, for instance, for interpreting mouse models of human platelet diseases. Knowledge of the identity and quantity of expressed proteins of a cell type is a prerequisite for a complete understanding of its molecular functions. Mass-spectrometry-based proteomics has allowed the identification of the entire protein complement of yeast and the close-to-complete set of proteins expressed in mammalian cell lines. Using recent technological advances, we here characterized the proteome of murine platelets, key actors in mediating hemostasis and thrombosis. We accurately measured the absolute protein concentrations of 13 platelet proteins using SILAC-protein epitope signature tags and used them as reference points to estimate the copy numbers of all proteins of the platelet proteome. To distinguish contaminants such as plasma or erythrocyte proteins from true platelet proteins, we monitored protein abundance profiles across multiple purification steps. In total, we absolutely quantified 4,400 platelet proteins, with estimated copy numbers ranging from less than 10 to about a million per cell. Stoichiometries derived from our data correspond well with previous studies. Our study provides a close-to-complete reference map of platelet proteins that will be useful to the community, for instance, for interpreting mouse models of human platelet diseases. Platelets are cells derived from the cytoplasm of megakaryocytes, which are found in bone marrow and constantly produce and release platelets into the blood. In the blood they circulate and survey the integrity of the vasculature. Upon injury of the endothelium, platelets prevent hemorrhages and uncontrolled blood loss by sealing the vascular lesions. Platelets' ability to form aggregates is important for their hemostatic function; however, pathological platelet activation—for example, during rupture of an atherosclerotic plaque—may reduce blood supply to the heart or brain during vascular occlusion and thereby induce cardiac infarction or stroke. It is therefore important to understand the molecular processes that control platelet activation and aggregation and to develop new therapeutic strategies to block critical platelet proteins involved in these processes (1Michelson A.D. Antiplatelet therapies for the treatment of cardiovascular disease.Nat. Rev. Drug Discov. 2010; 9: 154-169Crossref PubMed Scopus (307) Google Scholar). A deeper, quantitative understanding of the platelet proteome will facilitate the identification of new drug targets and therefore the development of novel anti-platelet therapies. With a diameter of only 0.5 to 1 μm in mice and 2 to 5 μm in humans, platelets are the smallest blood cell type and have a very short life span of 3 to 4 days (7 to 10 days in humans). Platelets lack a nucleus, and therefore there is no transcription that could replenish their residual megakaryocyte-derived mRNA. As a consequence, their mRNA levels are very low. Nevertheless, platelets translate mRNA into protein upon activation; however, whether this is important for platelet function is not clear (2Weyrich A.S. Schwertz H. Kraiss L.W. Zimmerman G.A. Protein synthesis by platelets: historical and new perspectives.J. Thrombosis Haemostasis. 2009; 7: 241-246Crossref PubMed Scopus (225) Google Scholar). The low mRNA levels make transcriptomics challenging because even a minimal contamination of the platelet sample by nucleated cells could make a substantial contribution to the measured transcriptome. In addition, functional interpretation of the measured transcript levels is complicated by the fact that they may reflect the parental megakaryocyte transcriptome rather than platelet-specific processes (3McRedmond J.P. Park S.D. Reilly D.F. Coppinger J.A. Maguire P.B. Shields D.C. Fitzgerald D.J. Integration of proteomics and genomics in platelets: a profile of platelet proteins and platelet-specific genes.Mol. Cell. Proteomics. 2004; 3: 133-144Abstract Full Text Full Text PDF PubMed Scopus (257) Google Scholar). Despite these difficulties, several studies have measured mouse and human platelet transcriptomes, leading to the identification of ∼6,500 and 9,500 transcripts, respectively (4Rowley J.W. Oler A.J. Tolley N.D. Hunter B.N. Low E.N. Nix D.A. Yost C.C. Zimmerman G.A. Weyrich A.S. Genome-wide RNA-seq analysis of human and mouse platelet transcriptomes.Blood. 2011; 118: e101-e111Crossref PubMed Scopus (411) Google Scholar). In contrast to transcriptomics, proteomics approaches are intrinsically better suited for studying the cellular functions of platelets, because proteins are the biochemical functional units. Furthermore, they are the drug targets in antithrombotic or antiplatelet therapy (1Michelson A.D. Antiplatelet therapies for the treatment of cardiovascular disease.Nat. Rev. Drug Discov. 2010; 9: 154-169Crossref PubMed Scopus (307) Google Scholar). Historically, studies of the platelet proteome have utilized two-dimensional gel electrophoresis and typically have quantified up to several dozens of proteins (5Zufferey A. Fontana P. Reny J.L. Nolli S. Sanchez J.C. Platelet proteomics.Mass Spectrom. Rev. 2012; 31: 331-351Crossref PubMed Scopus (36) Google Scholar). This approach has now been superseded by mass-spectrometry-based proteomics with much higher resolution and mass accuracy. The high peptide sequencing speed of modern instrumentation, combined with other technological advances, enables the mapping of close-to-complete proteomes with high confidence, despite the broad dynamic range of protein quantities expressed (6Altelaar A.F. Munoz J. Heck A.J. Next-generation proteomics: towards an integrative view of proteome dynamics.Nat. Rev. Genet. 2013; 14: 35-48Crossref PubMed Scopus (521) Google Scholar, 7Bantscheff M. Lemeer S. Savitski M.M. Kuster B. Quantitative mass spectrometry in proteomics: critical review update from 2007 to the present.Anal. Bioanal. Chem. 2012; 404: 939-965Crossref PubMed Scopus (581) Google Scholar). Recently, Burkhart et al. employed modern mass spectrometric instrumentation to confidently identify the deepest proteome to date of about 4,000 human platelet proteins (8Burkhart J.M. Vaudel M. Gambaryan S. Radau S. Walter U. Martens L. Geiger J. Sickmann A. Zahedi R.P. The first comprehensive and quantitative analysis of human platelet protein composition allows the comparative analysis of structural and functional pathways.Blood. 2012; 120: e73-e82Crossref PubMed Scopus (506) Google Scholar). Based on the tendency of the shotgun proteomics workflow to identify peptides from more abundant proteins more frequently (spectral counting), the authors were able to derive a quantitative measure of the majority of the identified proteome. These values were then scaled to copies per cell through a literature review of absolute copy number measurements from diverse sources, such as quantitative Western blotting. We have recently developed a method for absolute protein quantification in which we produce isotope-labeled recombinant protein fragments (PrESTs) 1The abbreviations used are:PrESTprotein epitope signature tagABPalbumin binding proteinGPglycoproteiniBAQintensity-based absolute quantificationSILACstable isotope labeling by amino acids in cell culture.1The abbreviations used are:PrESTprotein epitope signature tagABPalbumin binding proteinGPglycoproteiniBAQintensity-based absolute quantificationSILACstable isotope labeling by amino acids in cell culture. in Escherichia coli and combine them with stable isotope labeling in cell culture (SILAC) (9Ong S.E. Blagoev B. Kratchmarova I. Kristensen D.B. Steen H. Pandey A. Mann M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics.Mol. Cell. Proteomics. 2002; 1: 376-386Abstract Full Text Full Text PDF PubMed Scopus (4569) Google Scholar, 10Zeiler M. Straube W.L. Lundberg E. Uhlen M. Mann M. A protein epitope signature tag (PrEST) library allows SILAC-based absolute quantification and multiplexed determination of protein copy numbers in cell lines.Mol. Cell. Proteomics. 2012; 11 (O111.009613)Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). In this SILAC-PrEST method, protein fragments are expressed as fusion proteins with albumin binding protein (ABP) as a solubility tag. Upon purification, their absolute concentrations are determined in relation to a common sample of ultra-pure ABP whose concentration has previously been measured via amino acid analysis. The heavy PrESTs with known concentrations are then spiked into cell lysates, and the SILAC ratios of several peptides enable calculation of the cellular concentrations and copy numbers of their endogenous protein counterparts. Using this approach, we were able to quantify HeLa cell proteins with copy numbers ranging from thousands to several millions per cell (10Zeiler M. Straube W.L. Lundberg E. Uhlen M. Mann M. A protein epitope signature tag (PrEST) library allows SILAC-based absolute quantification and multiplexed determination of protein copy numbers in cell lines.Mol. Cell. Proteomics. 2012; 11 (O111.009613)Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). protein epitope signature tag albumin binding protein glycoprotein intensity-based absolute quantification stable isotope labeling by amino acids in cell culture. protein epitope signature tag albumin binding protein glycoprotein intensity-based absolute quantification stable isotope labeling by amino acids in cell culture. In this study we set out to analyze the murine platelet proteome via the high-resolution, quantitative methods developed in our laboratory (11Wisniewski J.R. Zougman A. Nagaraj N. Mann M. Universal sample preparation method for proteome analysis.Nat. Methods. 2009; 6: 359-362Crossref PubMed Scopus (5043) Google Scholar, 12Wisniewski J.R. Zougman A. Mann M. Combination of FASP and StageTip-based fractionation allows in-depth analysis of the hippocampal membrane proteome.J. Proteome Res. 2009; 8: 5674-5678Crossref PubMed Scopus (437) Google Scholar). Using a quadrupole Orbitrap mass spectrometer (13Michalski A. Damoc E. Hauschild J.P. Lange O. Wieghaus A. Makarov A. Nagaraj N. Cox J. Mann M. Horning S. Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer.Mol. Cell. Proteomics. 2011; 10 (M111.011015)Abstract Full Text Full Text PDF PubMed Scopus (626) Google Scholar), we obtained label-free quantification values for more than 4,400 proteins. These values were converted to copy numbers per platelet using PrESTs of 13 proteins for calibration. Furthermore, to distinguish true platelet proteins from contaminants experimentally, we followed their decreasing intensity profile through successive stages of purification. Our accurate and quantitative picture of a mammalian platelet proteome shows that it is much larger than might have been expected from its specialized functions. Mice (strain C57BL/6) were bled under anesthesia from the retro-orbital plexus, and ∼1 ml of blood was collected using heparin (20 U/ml in TBS) as the anticoagulation reagent. Blood was then centrifuged at 100 × g for 7 min to obtain platelet-rich plasma, which we termed the "crude fraction." The platelet-rich plasma was centrifuged at 700 × g to concentrate the platelets in the top layer (termed the "purified fraction"). This procedure was repeated once to obtain the "highly purified fraction." Eventually the platelet pellet was resuspended in 1 ml of Tyrode's buffer containing PGl2 and apyrase, after which it underwent a centrifugation step (Fig. 1A) ("ultra-purified fraction"). For protein correlation profiling, because greater sample amounts are required, we mixed blood from different mice and then performed the extensive purification described above. For each sample 30 × 106 platelets were resuspended in lysis buffer (2% SDS, 100 mm Tris, pH 7.5, 100 mm DTT), boiled at 95 °C, and further processed using the filter-aided sample preparation method (11Wisniewski J.R. Zougman A. Nagaraj N. Mann M. Universal sample preparation method for proteome analysis.Nat. Methods. 2009; 6: 359-362Crossref PubMed Scopus (5043) Google Scholar). In brief, SDS was exchanged to urea on a 30-kDa filter. Peptides were eluted after digestion with trypsin and subjected to a StageTip-based (14Rappsilber J. Mann M. Ishihama Y. Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips.Nat. Protoc. 2007; 2: 1896-1906Crossref PubMed Scopus (2570) Google Scholar) strong anion exchange fractionation (12Wisniewski J.R. Zougman A. Mann M. Combination of FASP and StageTip-based fractionation allows in-depth analysis of the hippocampal membrane proteome.J. Proteome Res. 2009; 8: 5674-5678Crossref PubMed Scopus (437) Google Scholar). Platelet counts were determined using a Hemavet950 analyzer (Drew Scientific, Waterbury, CT). For absolute protein quantification using the SILAC-PrEST method (10Zeiler M. Straube W.L. Lundberg E. Uhlen M. Mann M. A protein epitope signature tag (PrEST) library allows SILAC-based absolute quantification and multiplexed determination of protein copy numbers in cell lines.Mol. Cell. Proteomics. 2012; 11 (O111.009613)Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar), the synthetic genes of the protein standards were fused to ABP. The murine PrESTs were designed to be optimal for mass spectrometric analysis; specifically, we selected unique regions having many tryptic peptides also allowing us to distinguish isoforms. SILAC standards were produced using an auxotrophic E. coli strain (15Matic I. Jaffray E.G. Oxenham S.K. Groves M.J. Barratt C.L. Tauro S. Stanley-Wall N.R. Hay R.T. Absolute SILAC-compatible expression strain allows Sumo-2 copy number determination in clinical samples.J. Proteome Res. 2011; 10: 4869-4875Crossref PubMed Scopus (34) Google Scholar) in the presence of heavy arginine (13C615N4) and heavy lysine (13C615N2). The recombinant His-tag-containing proteins were purified via nickel-nitrilotriacetic acid columns and quantified with aliquots of a light ABP preparation on which amino acid analysis had been performed. Next, the 13 protein standards were mixed and spiked into the lysed platelets at approximately endogenous concentration. The sample was further processed using the filter-aided sample preparation method (11Wisniewski J.R. Zougman A. Nagaraj N. Mann M. Universal sample preparation method for proteome analysis.Nat. Methods. 2009; 6: 359-362Crossref PubMed Scopus (5043) Google Scholar). The samples were eluted from the stage tip, resuspended in buffer A (2% acetonitrile, 0.1% trifluoroacetic acid), and loaded onto a fresh 50-cm C18 column packed with Reprosil-Pur 1.9-μm resin (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany). The samples were separated on an ultra-high-performance LC system using a 180-min gradient ranging from 5% to 30% buffer B (80% acetonitrile, 0.1% trifluoroacetic acid) at a constant flow rate of 250 nl/min and injected via a nanoelectrospray ion source into the mass spectrometer. We used a quadrupole Orbitrap mass spectrometer (Q Exactive (13Michalski A. Damoc E. Hauschild J.P. Lange O. Wieghaus A. Makarov A. Nagaraj N. Cox J. Mann M. Horning S. Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer.Mol. Cell. Proteomics. 2011; 10 (M111.011015)Abstract Full Text Full Text PDF PubMed Scopus (626) Google Scholar), Thermo Fisher Scientific) in a data-dependent fashion, acquiring a full scan (300–1750 m/z; 70,000 resolution at m/z 200; target value, 3e6 ions; maximum fill time, 20 ms) and up to 10 subsequent MS/MS scans (17,500 resolution; target value, 1e5 ions; maximum fill time, 120 ms) using higher energy collision fragmentation for peptide identification (13Michalski A. Damoc E. Hauschild J.P. Lange O. Wieghaus A. Makarov A. Nagaraj N. Cox J. Mann M. Horning S. Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer.Mol. Cell. Proteomics. 2011; 10 (M111.011015)Abstract Full Text Full Text PDF PubMed Scopus (626) Google Scholar). Xcalibur software (Thermo Fisher Scientific) was used to acquire data. The raw data were analyzed using MaxQuant version 1.4.1.4 (16Cox J. Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification.Nat. Biotechnol. 2008; 26: 1367-1372Crossref PubMed Scopus (9154) Google Scholar) with the integrated search engine Andromeda (17Cox J. Neuhauser N. Michalski A. Scheltema R.A. Olsen J.V. Mann M. Andromeda: a peptide search engine integrated into the MaxQuant environment.J. Proteome Res. 2011; 10: 1794-1805Crossref PubMed Scopus (3450) Google Scholar). For peptide identification, the fragmentation spectra were searched against the UniProt mouse database (downloaded in May 2013) containing 50,829 entries to which 247 common contaminants were added. A "software lock mass" was used to recalibrate and improve the mass accuracy of the precursor masses (18Cox J. Michalski A. Mann M. Software lock mass by two-dimensional minimization of peptide mass errors.J. Am. Soc. Mass Spectrom. 2011; 22: 1373-1380Crossref PubMed Scopus (108) Google Scholar). During the main search, the maximum allowed initial mass deviation of the precursor ions was set at 4.5 ppm, and the maximum mass deviation of the fragmentation ions was set at 20 ppm. Cysteine carbamidomethylation was set as a fixed modification, and N-terminal acetylation and methionine oxidation were allowed as variable modifications. The "Enzyme" parameter was set as trypsin, for which N-terminal cleavage to proline and two miscleavages were allowed. A minimum of seven amino acids were required for valid peptide identification. In addition to the standard peptide search, the "second peptide" identification and the "match between run" option were enabled in Andromeda. For statistical evaluation of the data, a posterior error probability and a false discovery rate cutoff (determined via target-decoy searching) of 0.01 were used for peptides and proteins. For SILAC quantification, the MaxQuant standard settings were applied requiring at least two ratio counts between SILAC peptide pairs. Bioinformatics analysis was performed with the Perseus tool (version 1.4.1.4) available with the MaxQuant environment. For clustering and subsequent identification of contaminating proteins from other cell types and plasma, we used label-free intensities (19Cox J. Hein M.Y. Luber C.A. Paron I. Nagaraj N. Mann M. Accurate Proteome-wide Label-free Quantification by Delayed Normalization and Maximal Peptide Ratio Extraction, Termed MaxLFQ.Mol. Cell. Proteomics. 2014; 13: 2513-2526Abstract Full Text Full Text PDF PubMed Scopus (2688) Google Scholar). In this case we replaced missing values using data imputation, which is based on the assumption that missing values are caused by the detection limit of the MS measurement (20Deeb S.J. D'Souza R.C. Cox J. Schmidt-Supprian M. Mann M. Super-SILAC allows classification of diffuse large B-cell lymphoma subtypes by their protein expression profiles.Mol. Cell. Proteomics. 2012; 11: 77-89Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). First we determined the Gaussian distribution of the logarithmized data, and next we used a normal distribution with adjusted mean and standard deviation in order to simulate signals of low abundant proteins. We chose parameters (width = 0.3, downshift = 1.8) such that the distribution of the imputed values was placed at the lower end of the distribution of measured values. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (www.proteomexchange.org) via the PRIDE partner repository (21Vizcaino J.A. Cote R.G. Csordas A. Dianes J.A. Fabregat A. Foster J.M. Griss J. Alpi E. Birim M. Contell J. O'Kelly G. Schoenegger A. Ovelleiro D. Perez-Riverol Y. Reisinger F. Rios D. Wang R. Hermjakob H. The PRoteomics IDEntifications (PRIDE) database and associated tools: status in 2013.Nucleic Acids Res. 2013; 41: D1063-D1069Crossref PubMed Scopus (1595) Google Scholar) with the dataset identifier PXD000747. Platelets are highly abundant in blood and can be prepared in high yields through differential centrifugation, but proteins from lysed erythrocytes and plasma are inevitably present in platelet fractions. To distinguish true platelet proteins from such contaminants, we adapted the protein correlation profiling approach (22Foster L.J. de Hoog C.L. Zhang Y. Zhang Y. Xie X. Mootha V.K. Mann M. A mammalian organelle map by protein correlation profiling.Cell. 2006; 125: 187-199Abstract Full Text Full Text PDF PubMed Scopus (468) Google Scholar, 23Andersen J.S. Wilkinson C.J. Mayor T. Mortensen P. Nigg E.A. Mann M. Proteomic characterization of the human centrosome by protein correlation profiling.Nature. 2003; 426: 570-574Crossref PubMed Scopus (1051) Google Scholar). Originally, protein correlation profiling was developed to define organelle proteomes by quantifying the distribution of various organelle marker proteins across different subcellular fractions and subsequently matching proteins with profiles similar to the marker. Here, we instead followed protein abundance profiles across purification fractions to identify the contamination profile. Mouse platelets were separated from erythrocytes and plasma via multiple centrifugation and washing steps (Fig. 1A). Aliquots were taken at each step (crude, purified, highly purified, and ultra-purified fractions; see "Experimental Procedures") for protein correlation profiling. All samples were measured on a Q Exactive mass spectrometer after filter-aided sample preparation and strong anion exchange fractionation (12Wisniewski J.R. Zougman A. Mann M. Combination of FASP and StageTip-based fractionation allows in-depth analysis of the hippocampal membrane proteome.J. Proteome Res. 2009; 8: 5674-5678Crossref PubMed Scopus (437) Google Scholar). To ensure meaningful profiles, we required at least three quantification values either from three different mice and the highly pure platelet preparation or from three unpurified fractions. After this stringent filtering, 4,585 protein groups remained. To identify proteins with similar abundance profiles across different purification steps, we first performed unsupervised hierarchical clustering (Fig. 1B) of the label-free intensities (Fig. 1C). From the hierarchical clustering, we observed two major branches of the dendrogram grouped based on the different levels of purity—all the ultra-purified mouse replicates grouped together, separate from the cruder samples (Fig. 1B). Intriguingly, groups of proteins at each abundance level had similar profiles over the different samples (six clusters in total in six abundance ranges). Through cluster analysis and by comparing protein profiles to known plasma proteins, we identified 191 contaminant proteins. Because there was a gradual decline in protein intensities of the contaminations between the highly purified and ultra-purified fractions, we additionally perform a Welch's t test and detected 191 significant outliers at a false discovery rate of 0.05 (Fig. 2A). Interestingly, these outliers covered 90% of the 191 proteins identified via cluster analysis (18 unique proteins each), together amounting to 209 contaminants, of which 55% (115) were "secreted" according to the UniProt keyword annotation (Fig. 2B). These included apolipoproteins, serine protease inhibitors, antibodies, and complement factors, confirming that plasma was the main source of contaminant proteins. However, contaminants also included known erythrocyte markers such as erythrocytic spectrin, erythrocyte membrane protein band 4.2, and carbonic anhydrase 1 and 2, indicating that erythrocytes were another source of contamination (supplemental Table S1). However, we note that protein correlation profiling does not solve the problem of contaminating proteins completely. Because of platelets' canalicular system and "sponge-like" surface, as well as their uptake of plasma through vesicles (24Berger G. Masse J.M. Cramer E.M. Alpha-granule membrane mirrors the platelet plasma membrane and contains the glycoproteins Ib, IX, and V.Blood. 1996; 87: 1385-1395Crossref PubMed Google Scholar), some of these proteins remain bound to or are even taken up by platelets, making them difficult to identify as contaminants. To determine a reference map of the platelet proteome, we performed a second quantitative analysis. In these experiments we performed the same purification strategy but exclusively measured quantitative data on the ultra-purified platelets from three different mice. The different mice provided the range of variation. The final mean copy number values were determined as a combination of these values and the ultra-purified fraction of the combined blood sample to obtain the best possible accuracy. To evaluate and visualize the quality of our measurements, we plotted label-free protein abundance values of different replicates against one another (Fig. 3). This resulted in a Pearson correlation coefficient of r = 0.98, indicating excellent performance of our workflow and of MaxQuant's label-free algorithm, as well as consistency among the inbred mice. Plotting protein abundances of the ultra-purified versus the crude fractions highlighted contaminants previously identified through protein correlation profiling and Welch's t test as outliers from the trend line (depicted in red in Fig. 3). Once we eliminated the contaminant proteins, a final platelet proteome of 4,376 protein groups was obtained (supplemental Table S2). Next, we wished to determine the copy number of each platelet protein. To this end, we first selected 13 platelet proteins (Table I) that covered a wide abundance range of the platelet proteome. Absolute protein quantification was achieved using the recently developed SILAC-PrEST quantification method (10Zeiler M. Straube W.L. Lundberg E. Uhlen M. Mann M. A protein epitope signature tag (PrEST) library allows SILAC-based absolute quantification and multiplexed determination of protein copy numbers in cell lines.Mol. Cell. Proteomics. 2012; 11 (O111.009613)Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). Briefly, from each of the 13 proteins we selected ∼150 amino acids of unique sequence containing multiple tryptic peptides. We recombinantly expressed these PrESTs with a purification and a solubility tag and quantified these standards in a SILAC experiment (see "Experimental Procedures"). Next, we mixed the PrESTs of known concentration in appropriate ratios to obtain a master mix (10Zeiler M. Straube W.L. Lundberg E. Uhlen M. Mann M. A protein epitope signature tag (PrEST) library allows SILAC-based absolute quantification and multiplexed determination of protein copy numbers in cell lines.Mol. Cell. Proteomics. 2012; 11 (O111.009613)Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). This heavy mix was combined with the platelet lysate allowing multiplexed concentration determination of the corresponding endogenous proteins via their SILAC ratios (Fig. 4B). To quantify all other platelet proteins, intensities were normalized and scaled by the number of theoretically observable peptides through the intensity-based absolute quantification (iBAQ) algorithm incorporated in the MaxQuant software, using the PrESTs as the iBAQ standards (25Schwanhausser B. Busse D. Li N. Dittmar G. Schuchhardt J. Wolf J. Chen W. Selbach M. Global quantification of mammalian gene expression control.Nature. 2011; 473: 337-342Crossref PubMed Scopus (4059) Google Scholar). The copy numbers reported are the mean of three different ultra-purified mice and the ultra-purified fraction from the protein correlation profiling. The coefficient of variation (as a percentage) for each protein copy number was calculated from the four measurements and is reported in supplemental Table S2. The overall median coefficient of variation of the all copy numbers between the different mouse replicates was 20.6%.Table ISelected platelet proteins for reference pointsGene nameProtein nameCopy numberCoefficient of variation (%)Akt1RAC-α serine/threonine-protein kinase50035.6Akt2RAC-β serine/threonine-protein kinase70040.9Akt3RAC-γ serine/threonine-protein kinase1,600109.0Fermt3Fermitin family homolog 3286,7003.5FynTyrosine-protein kinase Fyn3,40026.6Gp6Platelet glycoprotein VI19,70033.9Itgb3Integrin β-3136,3004.8Itgb2Integrin β-27,3007.9P2rx1P2X purinoceptor 15,50024.7PrkcaProtein kinase
Referência(s)