The Application of New Software Tools to Quantitative Protein Profiling Via Isotope-coded Affinity Tag (ICAT) and Tandem Mass Spectrometry

• DOI: 10.1074/mcp.m300041-mcp200
• Publication Date: 2003-07-01T00:14:23Z
• Authors: Priska D. von Haller; Eugene Yi; Samuel Donohoe; Kelly Vaughn; Andrew Keller; Alexey I. Nesvizh...; Jimmy Eng; Xiao-jun Li; David R. Goodlett; Ruedi Aebersold; Julian D. Watts
• Keywords: Database search engine; Isobaric labeling; Tandem mass tag

Proteomic approaches to biological research that will prove the most useful and productive require robust, sensitive, reproducible technologies for both qualitative quantitative analysis of complex protein mixtures. Here we applied isotope-coded affinity tag (ICAT) approach profiling, in this case proteins copurified with lipid raft plasma membrane domains isolated from control stimulated Jurkat human T cells. With ICAT approach, cysteine residues two related isolates were covalently labeled isotopically normal heavy versions same reagent, respectively. Following proteolytic cleavage combined proteins, peptides fractionated by multidimensional chromatography subsequently analyzed via automated tandem mass spectrometry. Individual spectrometry spectra searched against a sequence database, variety recently developed, publicly available software applications used sort, filter, analyze, compare results repetitions experiment. In particular, robust statistical modeling algorithms assign measures confidence peptide sequences which they likely derived, identified database searches. We show applying such tools identification cell raft-associated able estimate accuracy identifications made. These also allow determination false positive rate as function user-defined data filtering parameters, thus giving user significant over information about final output large-scale proteomic experiments. ability probabilities all identifications, need manual verification is substantially reduced, making rapid evaluation large datasets possible. Finally, repeating experiment, relating general reproducibility validity analyses was obtained. A main objective proteomics systematic quantification expressed cell, or contained within compartment other complex. The common date has been combination separation, commonly high-resolution two-dimensional polyacrylamide gel electrophoresis (2DE) 1The abbreviations are: 2DE, electrophoresis; EM, expectation maximization; IADIFF, INTERACT differential; ICAT, tag; LC, liquid chromatography; μLC-MS/MS, microcapillary-liquid spectrometry; MIF, macrophage inhibitory factor; MS, MS/MS, pcomp, computed probability given assignment correct.; Pcomp, TCR, receptor. (MS/MS). For accomplished individual spot excision, in-gel-digestion, MS/MS. When desired, relative achieved visualizing differences 2DE patterns samples silver staining radiolabeling (1Boucherie H. Sagliocco F. Joubert R. Maillet I. Labarre J. Perrot M. Two-dimensional Saccharomyces cerevisiae.Electrophoresis. 1996; 17: 1683-1699Crossref PubMed Scopus (70) Google Scholar, 2Gygi S.P. Rochon Y. Franza B.R. Aebersold Correlation between mRNA abundance yeast.Mol. Cell. Biol. 1999; 19: 1720-1730Crossref (3101) 3Link A.J. Hays L.G. Carmack E.B. Yates 3rd, J.R. Identifying major proteome components Haemophilus influenzae type-strain NCTC 8143.Electrophoresis. 1997; 18: 1314-1334Crossref (161) 4Garrels J.I. McLaughlin C.S. Warner Futcher B. Latter G.I. Kobayashi Schwender Volpe T. Anderson D.S. Mesquita-Fuentes Payne W.E. Proteome studies cerevisiae: Identification characterization abundant proteins.Electrophoresis. 1347-1360Crossref (115) 5Shevchenko A. Jensen O.N. Podtelejnikov A.V. Wilm Vorm O. Mortensen P. Boucherie Mann Linking genome spectrometry: Large-scale yeast dimensional gels.Proc. Natl. Acad. Sci. U. S. 93: 14440-14445Crossref (1275) 6Bardel Louwagie Jaquinod Jourdain Luche Rabilloud Macherel D. Garin Bourguignon survey plant mitochondrial relation development.Proteomics. 2002; 2: 880-898Crossref (129) Scholar). This method proven quite successful cataloguing numbers samples. However, highly repetitive, labor intensive, difficult automate. addition, it necessarily selects only can be resolved missing many larger smaller addition lower solubility, proteins. Also, due sample loading limitations generally (4Garrels 7Gygi Corthals G.L. Zhang Evaluation electrophoresis-based technology.Proc. 2000; 97: 9390-9395Crossref (1176) Scholar), abundance, regulatory rarely detected when mixtures are analyzed. typically resolves different posttranslationally modified forms Given high degree post-translational modifications occurring on eukaryotic organisms, great difficulties obtaining accurate separate into multiple spots, well co-migrate spot, during 2DE. because vivo activities regulated modification, readily resolve differentially allows use monitor changes known "active" "inactive" developed technology instead based differential isotopic tagging (8Gygi Rist Gerber S.A. Turecek Gelb M.H. Quantitative using tags.Nat. Biotechnol. 994-999Crossref (4246) 9Han D.K. Eng Zhou profiling differentiation-induced microsomal tags spectrometry.Nat. 2001; 946-951Crossref (819) 10Smolka M.B. electrophoresis, labeling, spectrometry.Mol. Proteomics. 1: 19-29Abstract Full Text PDF (108) 11Flory M.R. Griffin T.J. Martin Advances stable isotope tags.Trends 20: S23-S29Abstract Scholar) summarized schematically Fig. 1. reagents consist three functional elements: thiol-reactive group selective labeling reduced Cys residues, biotin isolation peptides, linker synthesized either an ("light") "heavy" form (utilizing 2H 13C) incorporation tags. typical disulfide bridges under denaturing conditions, free sulfhydryl groups compared respectively "light" reagent. then combined, proteolyzed trypsin, resulting separated any number optional fractionation steps, including removal untagged (i.e. not containing residue) avidin-affinity chromatography. Peptide/protein made MS/MS fractions, followed searching observed spectra. ratio signal intensities unfragmented yields abundances peptide, hence original have investigation role detergent-resistant microdomains receptor (TCR) signaling line, Jurkat. sought evaluate reproducibility, performance, reliability comparing paper, present in-depth technical discussions each dataset, comparisons various datasets. application new, automated, greatly improved assigning scores matched. While our performed analyzing what would normally challenging overlap along observations processing, raised caveats should kept mind performing interpreting data. Furthermore, removed much identifications. experiments illustrated how nature facilitate timely processing datasets, currently time-consuming frequently process. offer some standardization interpretation of, turn, enable researchers perform interpret their consistently, those these allow, first time, transparent comparison laboratories. total 5 × 108 exponentially growing cells resuspended at ∼2 107/ml RPMI 1640 medium supplemented 10% fetal calf serum, split equal aliquots, chilled ice 15 min. Cells simultaneously treated anti-TCR (OKT3) anti-CD28 monoclonal antibodies, cross-linked secondary antibody 2 min simulate costimulation, essentially according standard laboratory protocols (12Heller Goodlett D.R. Watts J.D. comprehensive T-cell antigen composition microcapillary chromatography-tandem spectrometry.Electrophoresis. 21: 2180-2195Crossref (0) 13Heller CD28 stimulation regulates its association N-ethylmaleimide-sensitive fusion involved vesicle sorting.Proteomics. 70-78Crossref 14Watts Sanghera J.S. Pelech S.L. Phosphorylation serine 59 p56lck activated cells.J. Chem. 1993; 268: 23275-23282Abstract Detergent resistant membranes (rafts) purified described elsewhere (15). lysed 25 mm Tris, pH 7.5, 150 NaCl, 10 β-glycerophosphate, EDTA, 1 Na3VO4, phenylmethanesulfonyl fluoride, μg/ml soybean trypsin inhibitor, leupeptin, aprotinin, 0.1% Triton X-100, dounce homogenized (10 strokes), mixed volume 80% sucrose MNE buffer (25 2-morpholinoethanesulfonic acid, 6.5). Rafts density step gradient ultracentrifugation (16–18 h, 200,000 g, 4 °C). low-density raft-containing fraction further diluted buffer, rafts pelleted centrifugation (5 pellet dissolved 50 8, 6 m urea, 0.05% SDS. manufacturer's protocol (ICAT Kit Protein Labeling; Applied Biosystems, Foster City, CA), optimized conditions result (16Smolka Purkayastha Optimization tag-labeling procedure analysis.Anal. Biochem. 297: 25-31Crossref (170) short, following reduction cysteines (d0-ICAT) (d8-ICAT) samples, pooled ≤1 M ≤ 0.01% SDS proteolysis, excess (Promega, Madison, WI). cation exchange 4.6 200 Polysulfoethyl column μm particles, 300 Å pore size; Poly Columbia, MD) flow 800 μl/min. Peptides eluted 0–25% B 30 min, 25–100% 20 (buffer A: K2HPO4, 25% CH3CN, 3.0; B: 600 KCl, 3.0). elution profile (Fig. 8A) determined fractions Forty-three (fractions 10–52) individually processed avidin cartridges (Applied Biosystems) Biosystems), isolate Cys-containing peptides. Both eluate flow-through retained. To increase concentration (μLC-MS/MS) analysis, eluates pairs (except 52), 22 μLC-MS/MS. Because higher concentrations, Three sets generated subsequent μLC-MS/MS analysis: elutes mostly ICAT-labeled peptides) iterations experiment unlabeled iteration resultant subsets termed 1, 2, Flow-through Fifty 100% loaded autosampler sequentially data-dependent (17Yi E.C. Marelli Lee Purvine S.O. Aitchison Approaching complete peroxisome gas-phase fractionation.Electrophoresis. 23: 3205-3216Crossref (172) Injections cm 100 capillary packed in-house (Magic C18; Michrom BioResources, Auburn, CA). linear 10–40% ∼200–300 nl/min 0.4% acetic 0.005% heptafluorobutyric acid H2O; acetonitrile). HP1100 solvent delivery system (Hewlett Packard, Palo Alto, CA) precolumn splitting. An LCQ-DEKA ion-trap spectrometer (ThermoFinnigan, San Jose, built micro-spray device analyses. Peptide fragmentation collision-induced dissociation carried out fashion dynamic-exclusion option, recorded. uninterpreted finally submitted suite search results. process, below, more extensively "Results Discussion." Automated SEQUEST™ (18Eng McCormack A.L. correlate spectral amino database.J. Am. Soc. Mass Spectrom. 1994; 5: 976-989Crossref (5198) identify matches recorded spectrum. Uninterpreted locally maintained (version dated 9/8/2002) contaminants porcine (used proteolysis) bovine serum albumin (a component culture medium) additionally included. parameters set follows: static modification d0-ICAT-labeled +442.22, +8 d8-ICAT-labeled Cys; +16 oxidized Met; tolerance ± 3 Da; no enzyme specified. same, but without Cys. ThermoFinnigan. files automatically PeptideProphet™ (19Keller Nesvizhskii A.I. Kolker E. Empirical model search.Anal. 74: 5383-5392Crossref (3733) computation correct (pcomp). outputs displayed (9Han tool web/intranet-based display, sorting range user-definable parameters. restrict pcomp cut-offs, functions determine "single hit" entries one above predetermined threshold) filtered version tool, (IADIFF) side-by-side files. allowed entries) corresponded. generates Excel spreadsheet and/or sorted dataset distribution publication purposes. 1) ProteinProphet™. ProteinProphet™ utilizes list respective minimal (database explain compute (Pcomp) indeed sample(s) (20Nesvizhskii Keller identifying spectrometry.Anal. 2003; (in press)Crossref (3443) together (potentially) match entry). It deals indistinguishable grouping them "protein." occurs (mRNAs) fragments represented entries. Highly homologous gene families dealt formation "protein groups," again single match. discussed detail below web-based exported sorting, distribution, More PeptideProphet™, ProteinProphet™, found Proteomics pages www.systemsbiology.org/. upon request open source. experimental strategy employed study Briefly, (15Zhang W. Trible R.P. Samelson L.E. LAT palmitoylation: essential microdomain targeting tyrosine phosphorylation activation.Immunity. 1998; 9: 239-246Abstract (745) few variations. Cell cross-linking TCR coreceptor Proteins copurifying conventional detergent insolubility X-100) °C, 21Brown D.A. Rose J.K. Sorting GPI-anchored glycolipid-enriched subdomains transport apical surface.Cell. 1992; 68: 533-544Abstract (2568) reagent differed 8 units referred d0- d8-ICAT reagents, Samples chromatography, enrich (ICAT-labeled (unlabeled retained analyses, "Experimental Procedures." repeated second time assessment approach. From above, forward Cys-containing, experiments, respectively; non-Cys-containing dataset. All inferred tools, respectively, Currently, generate way quality assigned date, "correct" "threshold approach," where basis alone, everything threshold being discarded. derived. Typically, visual inspection verify quality, "correctness" peptide/protein particularly close preset threshold, cases hits," whereby identification. process variable. user/laboratory own opinion suitable minimum score set. problem compounded fact laboratories engines, unique scoring system, types spectrometers, producing characteristics. fact,