Portal de Periódicos da CAPES

Next-Generation Proteomics and Its Application to Clinical Breast Cancer Research

2017; Elsevier BV; Volume: 187; Issue: 10 Linguagem: Inglês

10.1016/j.ajpath.2017.07.003

1525-2191

Mariya Mardamshina, Tamar Geiger,

Molecular Biology Techniques and Applications

Proteomics technology aims to map the protein landscapes of biological samples, and it can be applied to a variety of samples, including cells, tissues, and body fluids. Because the proteins are the main functional molecules in the cells, their levels reflect much more accurately the cellular phenotype and the regulatory processes within them than gene levels, mutations, and even mRNA levels. With the advancement in the technology, it is possible now to obtain comprehensive views of the biological systems and to study large patient cohorts in a streamlined manner. In this review we discuss the technological advancements in mass spectrometry–based proteomics, which allow analysis of breast cancer tissue samples, leading to the first large-scale breast cancer proteomics studies. Furthermore, we discuss the technological developments in blood-based biomarker discovery, which provide the basis for future development of assays for routine clinical use. Although these are only the first steps in implementation of proteomics into the clinic, extensive collaborative work between these worlds will undoubtedly lead to major discoveries and advances in clinical practice. Proteomics technology aims to map the protein landscapes of biological samples, and it can be applied to a variety of samples, including cells, tissues, and body fluids. Because the proteins are the main functional molecules in the cells, their levels reflect much more accurately the cellular phenotype and the regulatory processes within them than gene levels, mutations, and even mRNA levels. With the advancement in the technology, it is possible now to obtain comprehensive views of the biological systems and to study large patient cohorts in a streamlined manner. In this review we discuss the technological advancements in mass spectrometry–based proteomics, which allow analysis of breast cancer tissue samples, leading to the first large-scale breast cancer proteomics studies. Furthermore, we discuss the technological developments in blood-based biomarker discovery, which provide the basis for future development of assays for routine clinical use. Although these are only the first steps in implementation of proteomics into the clinic, extensive collaborative work between these worlds will undoubtedly lead to major discoveries and advances in clinical practice. Omics technologies have revolutionized cancer research, through mapping of somatic mutations, gene copy number variations, and profiling gene expression alterations using genomic technologies, primarily next-generation sequencing. Proteomics, which is the focus of this review, in analogy to genomics, aims to profile the entire protein content of a biological sample, including the protein modifications and interactions.1Aebersold R. Mann M. Mass-spectrometric exploration of proteome structure and function.Nature. 2016; 537: 347-355Crossref PubMed Scopus (1105) Google Scholar The samples may be cells, tissues, or body fluids and, specifically in cancer clinical research, can include tumor tissues, plasma, urine, or proximal body fluids. The mainstream proteomic approach is termed bottom-up or shotgun proteomics and is based on mass spectrometry (MS) technology. In this approach proteins are extracted from the biological samples, followed by their digestion to peptides, chromatographic separation, and MS analysis (Figure 1).2Mann M. Kulak N.A. Nagaraj N. Cox J. The coming age of complete, accurate, and ubiquitous proteomes.Mol Cell. 2013; 49: 583-590Abstract Full Text Full Text PDF PubMed Scopus (285) Google Scholar Typically, proteins are digested with trypsin, which cleaves after every lysine and arginine residues in the protein, resulting in peptides of approximately 6 to 25 amino acids, which are normally unique to the protein. The MS determines the mass-to-charge ratio of each peptide in the complex mixture of hundreds of thousands of peptides and further fragments each peptide to allow determination of the amino acid sequence. To add a quantitative dimension to the MS analysis, sample preparation is often combined with labeling with stable isotopes. Metabolic labeling with heavy amino acids, known as the stable isotope labeling with amino acids in cell culture (SILAC) technique, or chemical labeling with isobaric tags techniques, such as isobaric tag for relative and absolute quantitation (iTRAQ) or tandem mass tag (TMT), are extensively used for relative quantification of a large variety of sample types.3Bantscheff 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 Combination with heavy proteins or peptides can further provide the absolute amount of selected proteins of interest. All of these data are then computationally analyzed against protein databases that provide the identity (amino acid sequence) and quantitative information about each peptide and protein.4Nesvizhskii A.I. A survey of computational methods and error rate estimation procedures for peptide and protein identification in shotgun proteomics.J Proteomics. 2010; 73: 2092-2123Crossref PubMed Scopus (380) Google Scholar Dramatic advancements in MS technology in the past decade increased the resolution, mass accuracy, and speed of modern mass spectrometers, thereby allowing higher coverage of the human proteome and increased throughput of the analysis.2Mann M. Kulak N.A. Nagaraj N. Cox J. The coming age of complete, accurate, and ubiquitous proteomes.Mol Cell. 2013; 49: 583-590Abstract Full Text Full Text PDF PubMed Scopus (285) Google Scholar As an example, 8 hours of measurement in the MS were required for the identification of approximately 5000 proteins only 5 years ago, whereas recent studies achieved similar results in 90 minutes.5Richards A.L. Merrill A.E. Coon J.J. Proteome sequencing goes deep.Curr Opin Chem Biol. 2015; 24: 11-17Crossref PubMed Scopus (74) Google Scholar, 6Scheltema R.A. Hauschild J.P. Lange O. Hornburg D. Denisov E. Damoc E. Kuehn A. Makarov A. Mann M. The Q Exactive HF, a Benchtop mass spectrometer with a pre-filter, high-performance quadrupole and an ultra-high-field Orbitrap analyzer.Mol Cell Proteomics. 2014; 13: 3698-3708Crossref PubMed Scopus (231) Google Scholar Further peptide fractionation before the MS runs enables much more comprehensive coverage of the proteome and increases the identifications to >10,000 proteins.7Beck M. Schmidt A. Malmstroem J. Claassen M. Ori A. Szymborska A. Herzog F. Rinner O. Ellenberg J. Aebersold R. The quantitative proteome of a human cell line.Mol Syst Biol. 2011; 7: 549Crossref PubMed Scopus (585) Google Scholar, 8Nagaraj N. Wisniewski J.R. Geiger T. Cox J. Kircher M. Kelso J. Paabo S. Mann M. Deep proteome and transcriptome mapping of a human cancer cell line.Mol Syst Biol. 2011; 7: 548Crossref PubMed Scopus (751) Google Scholar This depth allows identification of lowly expressed proteins and therefore opens new possibilities for application of the technology to clinical samples, aiming to reveal disease mechanisms and novel biomarkers. Beyond the discovery approach, which is based on the genome-scale proteome analysis, MS-based proteomics technology can also be used as a validation approach of specific biomarker candidates and can be applied to routine diagnostics in the clinic (Figure 1). In this targeted MS approach, rather than examining the entire proteome, only predetermined peptides are analyzed with high sensitivity, speed, and quantitative accuracy.9Gillette M.A. Carr S.A. Quantitative analysis of peptides and proteins in biomedicine by targeted mass spectrometry.Nat Methods. 2013; 10: 28-34Crossref PubMed Scopus (359) Google Scholar, 10Picotti P. Aebersold R. Selected reaction monitoring-based proteomics: workflows, potential, pitfalls and future directions.Nat Methods. 2012; 9: 555-566Crossref PubMed Scopus (991) Google Scholar In the present review, we focus on the application of the MS technologies to discovery proteomics in breast cancer research and the potential application of targeted proteomics to the clinic. Alternative approaches, such as matrix-assisted laser desorption/ionization (MALDI) imaging or protein-based approaches, which are primarily based on antibodies such as the human protein atlas antibody-based genome-scale initiative or the reverse-phase protein array approach, have been discussed elsewhere11Kriegsmann J. Kriegsmann M. Casadonte R. MALDI TOF imaging mass spectrometry in clinical pathology: a valuable tool for cancer diagnostics.Int J Oncol. 2015; 46: 893-906Crossref PubMed Scopus (114) Google Scholar, 12Akbani R. Ng P.K. Werner H.M. Shahmoradgoli M. Zhang F. Ju Z. Liu W. Yang J.Y. Yoshihara K. Li J. Ling S. Seviour E.G. Ram P.T. Minna J.D. Diao L. Tong P. Heymach J.V. Hill S.M. Dondelinger F. Stadler N. Byers L.A. Meric-Bernstam F. Weinstein J.N. Broom B.M. Verhaak R.G. Liang H. Mukherjee S. Lu Y. Mills G.B. A pan-cancer proteomic perspective on The Cancer Genome Atlas.Nat Commun. 2014; 5: 3887Crossref PubMed Scopus (350) Google Scholar, 13Uhlen M. Fagerberg L. Hallstrom B.M. Lindskog C. Oksvold P. Mardinoglu A. et al.Proteomics. Tissue-based map of the human proteome.Science. 2015; 347: 1260419Crossref PubMed Scopus (7243) Google Scholar and are beyond the scope of this review. Genomic analysis of breast cancer clinical samples has been applied to thousands of tumor samples as a part of the Cancer Genome Atlas14Cancer Genome Atlas NetworkComprehensive molecular portraits of human breast tumours.Nature. 2012; 490: 61-70Crossref PubMed Scopus (8301) Google Scholar and additional international initiatives.15Nik-Zainal S. Davies H. Staaf J. Ramakrishna M. Glodzik D. Zou X. et al.Landscape of somatic mutations in 560 breast cancer whole-genome sequences.Nature. 2016; 534: 47-54Crossref PubMed Scopus (1257) Google Scholar, 16Curtis C. Shah S.P. Chin S.F. Turashvili G. Rueda O.M. Dunning M.J. Speed D. Lynch A.G. Samarajiwa S. Yuan Y. Graf S. Ha G. Haffari G. Bashashati A. Russell R. McKinney S. Langerod A. Green A. Provenzano E. Wishart G. Pinder S. Watson P. Markowetz F. Murphy L. Ellis I. Purushotham A. Borresen-Dale A.L. Brenton J.D. Tavare S. Caldas C. Aparicio S. METABRIC GroupThe genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups.Nature. 2012; 486: 346-352Crossref PubMed Scopus (3696) Google Scholar These consortia, aiming to classify breast cancer subtypes and to associate the profiles with cancer prognosis, profiled the gene expression patterns, somatic mutations, copy number variations (CNVs), and epigenetic profiles. Despite this breadth of molecular information, the protein level has been, to a large extent, neglected. To understand the potential contribution of proteomics to the molecular characterization of cancer, multiple studies performed detailed comparisons of proteomic profiles and the genomics and transcriptomics of the same systems. Comparing cancer proteomes with CNV data showed low correlation of approximately 0.2, highlighting the relatively minor contribution of these genomic alterations to the levels of their end products, the proteins.17Geiger T. Cox J. Mann M. Proteomic changes resulting from gene copy number variations in cancer cells.PLoS Genet. 2010; 6: e1001090Crossref PubMed Scopus (102) Google Scholar Comparative analysis of breast cancer tissue analysis showed that only 30% of the chromosomal aberrations are translated to protein abundance changes.18Mertins P. Mani D.R. Ruggles K.V. Gillette M.A. Clauser K.R. Wang P. et al.Proteogenomics connects somatic mutations to signalling in breast cancer.Nature. 2016; 534: 55-62Crossref PubMed Scopus (978) Google Scholar These studies showed that only some of the amplifications and deletions, presumably the cancer drivers, have a functional output, whereas most are further controlled by other gene expression regulatory mechanisms of transcription, translation, and mRNA and protein stability. Much higher correlation is observed between the transcriptome and the proteome, because in many systems transcription is the major determinant of protein amount.19Aviner R. Shenoy A. Elroy-Stein O. Geiger T. Uncovering hidden layers of cell cycle regulation through integrative multi-omic analysis.PLoS Genet. 2015; 11: e1005554Crossref PubMed Scopus (33) Google Scholar, 20Jovanovic M. Rooney M.S. Mertins P. Przybylski D. Chevrier N. Satija R. Rodriguez E.H. Fields A.P. Schwartz S. Raychowdhury R. Mumbach M.R. Eisenhaure T. Rabani M. Gennert D. Lu D. Delorey T. Weissman J.S. Carr S.A. Hacohen N. Regev A. Immunogenetics. Dynamic profiling of the protein life cycle in response to pathogens.Science. 2015; 347: 1259038Crossref PubMed Scopus (295) Google Scholar Cell line studies have shown correlations of 0.4 to 0.6 between mRNA levels and proteins.8Nagaraj N. Wisniewski J.R. Geiger T. Cox J. Kircher M. Kelso J. Paabo S. Mann M. Deep proteome and transcriptome mapping of a human cancer cell line.Mol Syst Biol. 2011; 7: 548Crossref PubMed Scopus (751) Google Scholar, 21Schwanhausser 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 Although mRNA levels are often used as proxies of protein levels, this limited correlation shows the dramatic differences between these two layers, which reflect the potential contribution and the importance of protein analysis. Analyzing the protein level has the potential to integrate all of the upstream genetic alterations, epigenetic regulations, and environmental effects that are fundamental regulators of the cancer phenotype. If the technology permitted complete coverage of all full-length proteins, these analyses would identify all expressed mutated proteins, the functional CNVs, and output of all regulatory mechanisms of genes and proteins. However, despite the improved proteomic coverage, it does not yet enable identification of most of the mutated proteins.18Mertins P. Mani D.R. Ruggles K.V. Gillette M.A. Clauser K.R. Wang P. et al.Proteogenomics connects somatic mutations to signalling in breast cancer.Nature. 2016; 534: 55-62Crossref PubMed Scopus (978) Google Scholar Furthermore, to identify mutations by standard proteomic approaches, one would have to have a database, which includes all potentially mutated protein sequences. To obtain such a comprehensive database it is possible to integrate the sequencing information of all somatic mutations obtained in the multitude of genomic databases, such as Catalogue of Somatic Mutations in Cancer and The Cancer Genome Atlas (TCGA). With the growing number of sequenced cancer genomes, several databases, for example CanProVar22Li J. Duncan D.T. Zhang B. CanProVar: a human cancer proteome variation database.Hum Mutat. 2010; 31: 219-228Crossref PubMed Scopus (52) Google Scholar, 23Zhang M. Wang B. Xu J. Wang X. Xie L. Zhang B. Li Y. Li J. CanProVar 2.0: an updated database of human cancer proteome variation.J Proteome Res. 2017; 16: 421-432Crossref PubMed Scopus (24) Google Scholar and XMann,24Yang X. Lazar I.M. XMAn: a Homo sapiens mutated-peptide database for the MS analysis of cancerous cell states.J Proteome Res. 2014; 13: 5486-5495Crossref PubMed Scopus (18) Google Scholar integrate these data into proteomic databases that include nonsynonymous mutations, including germline and somatic ones. Alternatively, sample-specific databases can be generated directly from genomics of the same samples, when these are analyzed with both technologies.18Mertins P. Mani D.R. Ruggles K.V. Gillette M.A. Clauser K.R. Wang P. et al.Proteogenomics connects somatic mutations to signalling in breast cancer.Nature. 2016; 534: 55-62Crossref PubMed Scopus (978) Google Scholar, 25Lawrence R.T. Perez E.M. Hernandez D. Miller C.P. Haas K.M. Irie H.Y. Lee S.I. Blau C.A. Villen J. The proteomic landscape of triple-negative breast cancer.Cell Rep. 2015; 11: 630-644Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar Beyond the identification of mutated proteins, this integrated genomic-proteomic analysis can associate between specific mutations and the proteomic profiles. Specifically in breast cancer, given the moderate mutational load and the high variation in mutational profiles between patients, integration of the genomics with proteomics is capable of revealing the convergence of distinct mutations to a limited number of pathways, which can be identified by proteomics. Moreover, because mutations often occur in regulatory proteins, proteomics can reveal the output of these perturbations and can identify the functional drivers of tumorigenesis. With the use of this approach, Lawrence et al25Lawrence R.T. Perez E.M. Hernandez D. Miller C.P. Haas K.M. Irie H.Y. Lee S.I. Blau C.A. Villen J. The proteomic landscape of triple-negative breast cancer.Cell Rep. 2015; 11: 630-644Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar analyzed triple-negative breast cancer cell lines and showed that multiple mutations are associated with similar proteomic alteration, thereby reducing cancer heterogeneity on the protein level. These common altered proteins can then be targeted with the same drugs despite their distinct genetic profiles. The Clinical Proteomic Tumor Analysis Consortium performed proteomic and phosphoproteomic analysis of 77 breast cancer tissue samples from the TCGA and could therefore integrate the proteomic data with the genomic data of the same samples (detailed below). With the use of the sample-specific databases, they showed the association of mutations with changes in protein levels. The phosphoproteomic analysis of these tissues further identified signaling pathways that change in association with specific mutations.18Mertins P. Mani D.R. Ruggles K.V. Gillette M.A. Clauser K.R. Wang P. et al.Proteogenomics connects somatic mutations to signalling in breast cancer.Nature. 2016; 534: 55-62Crossref PubMed Scopus (978) Google Scholar For example, they associated between phosphorylation of RPS6KA5 and EIF2AK4 with mutations in PIK3CA, and phosphorylation of MASTL and EEF2K were associated with TP53 mutations, independent of the cancer subtype. Beyond the examination of full proteomes, the technology also enables analysis of subproteomes, such as specific cell organelles and protein interactions. One of the recent clinically attractive applications is in the field of immune-oncology. Immunopeptidomics is an emerging field in proteomics, in which major histocompatibility complex molecules are isolated, followed by MS analysis of the bound peptides that are presented to the immune system. This approach reveals all of the antigens expressed by the cancer cells and can serve as a platform for neo-antigen identification toward individualized cancer vaccine development.26Bassani-Sternberg M. Barnea E. Beer I. Avivi I. Katz T. Admon A. Soluble plasma HLA peptidome as a potential source for cancer biomarkers.Proc Natl Acad Sci U S A. 2010; 107: 18769-18776Crossref PubMed Scopus (101) Google Scholar, 27Bassani-Sternberg M. Braunlein E. Klar R. Engleitner T. Sinitcyn P. Audehm S. Straub M. Weber J. Slotta-Huspenina J. Specht K. Martignoni M.E. Werner A. Hein R. H Busch D. Peschel C. Rad R. Cox J. Mann M. Krackhardt A.M. Direct identification of clinically relevant neoepitopes presented on native human melanoma tissue by mass spectrometry.Nat Commun. 2016; 7: 13404Crossref PubMed Scopus (398) Google Scholar, 28Bassani-Sternberg M. Coukos G. Mass spectrometry-based antigen discovery for cancer immunotherapy.Curr Opin Immunol. 2016; 41: 9-17Crossref PubMed Scopus (92) Google Scholar, 29Hickman H.D. Yewdell J.W. Mining the plasma immunopeptidome for cancer peptides as biomarkers and beyond.Proc Natl Acad Sci U S A. 2010; 107: 18747-18748Crossref PubMed Scopus (11) Google Scholar, 30Kalaora S. Barnea E. Merhavi-Shoham E. Qutob N. Teer J.K. Shimony N. Schachter J. Rosenberg S.A. Besser M.J. Admon A. Samuels Y. Use of HLA peptidomics and whole exome sequencing to identify human immunogenic neo-antigens.Oncotarget. 2016; 7: 5110-5117Crossref PubMed Scopus (109) Google Scholar Altogether, analysis of the proteomic level can be integrated with genomics and can complement it by revealing the functional outputs of genomics, integrating with environmental cues that affect the phenotype. Proteomic analysis of cell lines has become a routine procedure that enables coverage of near complete proteomes. Analysis of tumor tissue samples, however, requires various adaptations to reach similar depth. The main challenge is associated with the heterogeneous composition of the tissue, which includes multiple cell types and extracellular matrix (ECM) proteins. Highly abundant ECM proteins may mask the more lowly expressed cellular proteins and may lead to reduced proteome coverage. In addition, analysis of a mixed cell population, which includes cancer cells, together with immune cells, fibroblasts, adipocytes, and endothelial cells, may average out substantial differences associated with each of these populations. To obtain accurate, clinically relevant data, it is therefore essential to dissect the tissue and to extract the specific cells of interest. Tissue macrodissection or laser capture microdissection, isolates the cancer cells from the rest of the tissue, thereby allowing more accurate cancer cell analysis. Microdissected tissue analysis has clear biological advantages and showed higher peptide identification rates because of the lower complexity of the tissue.31De Marchi T. Braakman R.B. Stingl C. van Duijn M.M. Smid M. Foekens J.A. Luider T.M. Martens J.W. Umar A. The advantage of laser-capture microdissection over whole tissue analysis in proteomic profiling studies.Proteomics. 2016; 16: 1474-1485Crossref PubMed Scopus (29) Google Scholar However, analysis of minute tissue amounts requires substantial optimization of protein extraction and digestion.32Braakman R.B. Tilanus-Linthorst M.M. Liu N.Q. Stingl C. Dekker L.J. Luider T.M. Martens J.W. Foekens J.A. Umar A. Optimized nLC-MS workflow for laser capture microdissected breast cancer tissue.J Proteomics. 2012; 75: 2844-2854Crossref PubMed Scopus (39) Google Scholar An additional challenge in the analysis of clinical samples is the ability to use archived formalin-fixed, paraffin-embedded tissue. Reversal of the formalin crosslinks is routine in antigen retrieval procedure in immunohistochemistry methods. In a similar manner, sample boiling in high detergent concentrations denatures the proteins in the samples and allows exposure of the entire protein to the proteolytic enzymes and results in high digestion efficiency. Comparison of fresh-frozen tissue to formalin-fixed, paraffin-embedded tissue digestion showed that fixation does not reduce the number of identified proteins and does not induce specific crosslinking-related modifications.33Sprung Jr., R.W. Brock J.W. Tanksley J.P. Li M. Washington M.K. Slebos R.J. Liebler D.C. Equivalence of protein inventories obtained from formalin-fixed paraffin-embedded and frozen tissue in multidimensional liquid chromatography-tandem mass spectrometry shotgun proteomic analysis.Mol Cell Proteomics. 2009; 8: 1988-1998Crossref PubMed Scopus (163) Google Scholar, 34Ostasiewicz P. Zielinska D.F. Mann M. Wisniewski J.R. Proteome, phosphoproteome, and N-glycoproteome are quantitatively preserved in formalin-fixed paraffin-embedded tissue and analyzable by high-resolution mass spectrometry.J Proteome Res. 2010; 9: 3688-3700Crossref PubMed Scopus (192) Google Scholar Therefore, for proteomic analysis there is no actual limitation in the types of analyzed tissues. Because of the inability to amplify proteins (as opposed to nucleic acids), sample amounts are often limiting the depth of the identified proteome. To overcome this challenge, several approaches have been developed aiming to minimize sample handling and sample loss, by using a single-reactor for the entire sample preparation procedure. The in StageTip sample-processing method uses pipette tips with various filter for protein digestion, separation, and purification.35Kulak N.A. Pichler G. Paron I. Nagaraj N. Mann M. Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells.Nat Methods. 2014; 11: 319-324Crossref PubMed Scopus (991) Google Scholar The solid phase-enhanced sample preparation platform uses paramagnetic beads to perform the protein digestion and purification in a single tube. This method demonstrated improved efficiency and reduced sample loss.36Hughes C.S. Foehr S. Garfield D.A. Furlong E.E. Steinmetz L.M. Krijgsveld J. Ultrasensitive proteome analysis using paramagnetic bead technology.Mol Syst Biol. 2014; 10: 757Crossref PubMed Scopus (513) Google Scholar The solid phase-enhanced sample preparation–clinical tissue proteomics platform can be easily adapted to high-throughput format for various clinical samples.37Hughes C.S. McConechy M.K. Cochrane D.R. Nazeran T. Karnezis A.N. Huntsman D.G. Morin G.B. Quantitative profiling of single formalin fixed tumour sections: proteomics for translational research.Sci Rep. 2016; 6: 34949Crossref PubMed Scopus (73) Google Scholar Dozens of proteomic studies analyzed clinical breast tumor samples; however, most suffer from low proteome coverage or cohort size. Rezaul et al38Rezaul K. Thumar J.K. Lundgren D.H. Eng J.K. Claffey K.P. Wilson L. Han D.K. Differential protein expression profiles in estrogen receptor-positive and -negative breast cancer tissues using label-free quantitative proteomics.Genes Cancer. 2010; 1: 251-271Crossref PubMed Scopus (35) Google Scholar studied protein expression profiles associated with estrogen receptor (ER) status of breast cancer and obtained a signature of 236 differentially expressed proteins between ER-positive and -negative tumors. Among those, they found Iprin-α1, Fascin, death-associated protein 5, and β-arrestin 1 as potential biomarkers of the ER-negative subgroup. However, the cohort consisted of only six patients and an average of 1000 identified proteins per sample. Similarly, Gamez-Pozo et al39Gamez-Pozo A. Ferrer N.I. Ciruelos E. Lopez-Vacas R. Martinez F.G. Espinosa E. Vara J.A. Shotgun proteomics of archival triple-negative breast cancer samples.Proteomics Clin Appl. 2013; 7: 283-291Crossref PubMed Scopus (23) Google Scholar reported identification of >1600 protein groups in a cohort of five triple-negative breast cancer (TNBC) samples. With the use of a set of 18 matched normal breast epithelial samples and ER-positive malignant breast epithelial samples, Cha et al40Cha S. Imielinski M.B. Rejtar T. Richardson E.A. Thakur D. Sgroi D.C. Karger B.L. In situ proteomic analysis of human breast cancer epithelial cells using laser capture microdissection: annotation by protein set enrichment analysis and gene ontology.Mol Cell Proteomics. 2010; 9: 2529-2544Crossref PubMed Scopus (61) Google Scholar obtained a molecular signature that corresponds to the transition of normal epithelial tissue to highly invasive malignant one and identified 298 significantly changing proteins that were more abundant in the tumor samples. These were enriched for biological processes such as focal adhesion and lipid metabolism. With the improvement of MS technologies and sample preparation protocols, the size of cohorts and the quality of proteomic data significantly improved. Liu et al41Liu N.Q. Stingl C. Look M.P. Smid M. Braakman R.B. De Marchi T. Sieuwerts A.M. Span P.N. Sweep F.C. Linderholm B.K. Mangia A. Paradiso A. Dirix L.Y. Van Laere S.J. Luider T.M. Martens J.W. Foekens J.A. Umar A. Comparative proteome analysis revealing an 11-protein signature for aggressive triple-negative breast cancer.J Natl Cancer Inst. 2014; 106: djt376Crossref PubMed Scopus (43) Google Scholar analyzed a cohort of 126 TNBC breast cancer samples using laser capture microdissection–liquid chromatography–MS/MS approach. The total protein coverage obtained was >3500 proteins, and they identified an 11-protein signature for TNBC with 10 proteins that were up-regulated (CMPK1, AIFM1, FTH1, EML4, GANAB, CTNNA1, AP1G1, STX12, AP1M1, and CAPZB)), and one was down-regulated (methylenetetrahydrofolate dehydrogenase 1) in good-prognosis patients. The signature presented high predictive value of patient prognosis with area under the curve of 0.83 of a receiver operating characteristics curve. With the use of the same techniques De Marchi et al42De Marchi T. Liu N.Q. Stingl C. Timmermans M.A. Smid M. Look M.P. Tjoa M. Braakman R.B. Opdam M. Linn S.C. Sweep F.C. Span P.N. Kliffen M. Luider T.M. Foekens J.A. Martens J.W. Umar A. 4-protein signature predicting tamoxifen treatment outcome in recurrent breast cancer.Mol Oncol. 2016; 10: 24-39Crossref PubMed Scopus (26) Google Scholar obtained a four-protein signature (programmed cell death protein 4, cingulin, ovarian carcinoma immunoreactive antigen domain-containing protein 1, and Ras GTPase-activating protein-binding protein 2), which predicts tamoxifen-susceptibility in recurrent breast cancer. The cohort consisted of 112 ER-positive tumor samples with total coverage of 4000 proteins. In our laboratory, we profiled ER-positive breast cancer progression, by comparing matched normal noncancerous tissue and primary tumors with and without lymph node involvement. The cohort was composed of four groups with a total of 88 samples, and the analysis reached the depth of >9000 quantified proteins. This work demonstrated metabolic remodeling represented by up-regulation of oxidative phosphorylation processes and down-regulation of key glycolytic proteins such as glyceraldehyde-3 phosphate dehydrogenase, fructose-bisphosphate aldolase A, hexokinase-2, and L-lactate dehydrogenase A and B chains