HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 49, 45497-45500, December 7, 2001

This Article Full Text (PDF) All Versions of this Article: 276/49/45497    most recent R100014200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Griffin, T. J. Articles by Aebersold, R. Search for Related Content PubMed PubMed Citation Articles by Griffin, T. J. Articles by Aebersold, R. Social Bookmarking                      What's this?

MINIREVIEW
Advances in Proteome Analysis by Mass Spectrometry^*

Timothy J. Griffin and Ruedi Aebersold

From the Institute for Systems Biology, Seattle, Washington 98105

	INTRODUCTION

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

The interpretation of the information contained in the genomic sequence of a species with respect to the structure, function, and control of biological processes is a technical and conceptual challenge for current research methods. Systematic and quantitative analysis of gene expression is emerging as a valuable tool to diagnostically distinguish between cell types (1-5) and to differentiate between states (metabolic, activation, pathological) of a particular cell type (5-7). More elaborate strategies such as the combination of systematic, quantitative gene expression analysis with targeted, hypothesis-guided perturbations of cells are being explored for the comprehensive mechanistic analysis of cellular pathways and processes (5, 7-10).

Measuring gene expression at the protein level is potentially more informative than the corresponding measurement at the mRNA level. Proteins, the major catalysts of biological function, contain several dimensions of information that collectively indicate the actual rather than the potential functional state as indicated by mRNA analysis. These include the abundance, state of modification, subcellular location, and three-dimensional structure of proteins and their association with each other and/or with biomolecules of different types. Mass spectrometry has become the analytical technology of choice for many of the aspects of proteome analyses that are reflected by the covalent structure of proteins.

This review describes recent, innovative advances in mass spectrometry-based proteome analysis that collectively have converged into a generic new approach to proteomics. The specific advances include high throughput protein identification by multidimensional chromatography, automated tandem mass spectrometry and sequence data base searching, accurate quantification by the application of stable isotope dilution theory to protein analysis, and the targeted isolation of selected analytes by the use of highly selective chemistries. Selected applications of these methods along with speculations about future prospects and directions of proteomics research are also included.

	The Emergence of Proteomics: the First Generation Technology

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

The development of methods to separate complex protein mixtures at high resolution by two-dimensional gel electrophoresis (2DE)¹ into reproducible patterns (11, 12) presented the opportunity to diagnose quantitative and qualitative differences in the protein composition of two or more cell or tissue samples long before gene array techniques to measure gene expression were conceived. Unfortunately, 2DE by itself was an essentially descriptive technique and, without the availability of reliable tools for the identification of the separated protein species, of limited utility as a molecular biology research tool.

This changed in the early 1990s when two revolutionary techniques, matrix-assisted laser desorption ionization (MALDI) time-of-flight (TOF) mass spectrometry (MS) (13, 14) and electrospray ionization (ESI) (15, 16) MS and tandem mass spectrometry (MS/MS) replaced the slower and less sensitive chemical degradation methods (17, 18) as the methods of choice for the identification of proteins separated by 2DE (19-24). Typically, these methods involved excision of gel bands of interest, in-gel digestion of the proteins contained in the band using the enzyme trypsin (25), and finally mass spectrometric analysis of the peptides produced. Protein identification was accomplished using peptide-mass fingerprinting by MALDI-TOF MS, as initially described by Henzel et al. (26) and independently by others (reviewed in Ref. 23), nano-ESI tandem mass spectrometry (MS/MS) (27, 28), or reverse-phase (RP) microcapillary liquid chromatography (µLC) ESI MS/MS (29-34) using automated, data-dependent scanning and dynamic exclusion of peptide ions already analyzed in the same experiment (29, 35-37). The tandem mass spectra produced by the collision-induced dissociation (CID) (38, 39) of selected peptides are searched against theoretical tandem mass spectra of peptide sequences contained in a data base using a variety of search algorithms (40), the pioneering program being Sequest developed by Eng et al. (41). The specific techniques and instruments have been reviewed in detail and are not discussed further here (42, 43).

	Apparent Limitations of the 2DE-MS Approach: an Outline of a Second Generation Technology

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

It has become apparent that the 2DE-MS method as most frequently practiced has significant, inherent limitations. First, the combination of limited sample capacity and limited detection sensitivity of 2DE restricts the detection of low abundance proteins. If total yeast cell lysates are separated and detected by silver staining, proteins present at less than 1000 copies per cell are not detected (44). As proteins expressed at low abundance may make up a large portion of a given proteome (44), it is apparent that the proteins detected by 2DE do not give a true representation of all the expressed proteins. Second, despite substantial recent advances (45, 46), the separation of transmembrane proteins by 2DE remains challenging. Third, a substantial fraction of spots contain more than one protein and/or differentially modified or processed forms of a protein that migrate to different positions in the gel, thus complicating quantification (44). Fourth, the method is based on the sequential identification of individually processed protein spots, which limits its throughput. Fifth, the method is inherently labor-intensive and requires a high skill level, which limits the potential for full automation. Collectively, these limitations indicate the need for the development of improved or alternative technologies if routine proteome analysis is to become a reality. To address these limitations, incremental improvements of the 2DE-MS approach have been made that include sample prefractionation prior to 2DE (47), the use of fluorescent protein dyes with enhanced detection sensitivity (48), and the use of gels with expanded separation range (zoom gels) (46, 49) to improve the detection sensitivity of low abundance proteins, the search for new detergent systems to maximize solubility of membrane proteins (45), and the development of robotic and software systems to increase the level of automation of the process (42, 50).

Concurrently, an alternative technique has been emerging that has the potential to systematically identify and quantify all the proteins in a cell or tissue type. It is based on three principles. The first is rapid protein identification by automated tandem mass spectrometry and sequence data base searching. Essentially the same methods developed for the identification of gel-separated proteins are applied to identify the components of unseparated protein mixtures. The second is the determination of the ratio of abundance (relative quantification) for proteins present in different protein samples by stable isotope dilution. Stable isotope dilution theory (51) states that the relative signal intensity in a mass spectrometer of two analytes that are chemically identical but of different stable isotope composition (and thus distinguishable in a mass analyzer) are a true representation of the relative abundance of the two analytes in the sample. The third principle is the targeted isolation of selected peptide analytes from complex peptide mixtures via specific chemical reactions. Collectively, the three components permit the relative measurement of abundance and the identification of the components of very complex protein mixtures rapidly and with a high degree of automation, without the need to separate protein mixtures prior to analysis. Variations of this technology also have the potential to systematically and quantitatively determine properties of proteins that reflect their functional state. These include the phosphorylation state and the activity of some classes of enzymes. The evolution and early applications of this second generation proteomics technology are described below.

	The Emergence and Initial Applications of a Second Generation Proteomics Technology

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

The challenge facing any comprehensive proteomics approach is one of separating and simplifying very complex mixtures of proteins in which individual components differ in abundance by six or more orders of magnitude, while retaining enough information to allow for comprehensive characterization of expressed proteins. To this end, the combination of selective labeling of proteins with stable isotope-containing affinity reagents and multidimensional liquid chromatography in conjunction with automated, data-dependent tandem mass spectrometry and sequence data base searching has proven effective and has been shown to overcome at least some of the critical limitations of the 2DE-MS-based approach to proteomics.

	Selective Protein Labeling and Automated Tandem Mass Spectrometric Analysis

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

The labeling of proteins at specific sites in a complex mixture followed by proteolysis and selective purification of the labeled peptide fragments has proven to be an effective method for the analysis of unseparated protein samples. The nucleophilic thiol group contained in the side chain of reduced cysteine residues is a commonly targeted site for the modification and labeling of proteins and peptides (52, 53). The frequency of cysteine residues in protein sequences makes it an attractive amino acid to target for the reduction of the complexity of peptide mixtures (~10% of all possible tryptic peptides in the yeast Saccharomyces cerevisiae contain a cysteine (54)). However, as ~92% of the total proteins in the S. cerevisiae genome contain at least one cysteine (54), selection and identification of only the cysteine-containing peptides still enables the comprehensive identification of expressed proteins, although the complexity of the sample is significantly decreased. Reduction of the complexity of peptide mixtures prior to mass spectrometric analysis is advantageous for several reasons. First, the selection of a subset of the peptides generated by proteolysis of a protein mixture greatly increases the representation of the selected peptides in the sample that is loaded onto the µLC column. The reduction of the sample complexity achieved by selective tagging is therefore essential for detecting and identifying low abundance proteins. Second, as low intensity, co-eluting peptides are often missed during automated MS/MS analysis due to the relatively long duty cycle of the instrumental routine, simplifying the mixture allows for the identification of a larger proportion of the available peptides. Third, the presence of the relatively rare amino acid cysteine that is indicated by the specific reaction between the alkylating group and the thiol side chain provides a significant constraint for sequence data base searching.

The effective use of selective labeling of cysteine residues for the simplification of the peptide samples generated by proteolysis of protein mixtures prior to mass spectrometric analysis has been demonstrated by Spahr et al. (55). The authors labeled cysteine side chains using a cleavable, biotinylated reagent, and these peptides were then affinity-purified using immobilized avidin and identified by LC-ESI MS/MS. This experiment was part of a study in which a total of 108 soluble intermembrane mitochondria proteins from mouse liver samples were identified. Selective labeling and capture of cysteine-containing peptides is also the basis of the isotope-coded affinity tag (ICAT) approach that was recently developed in our laboratory (56). In this approach the proteins present in two samples (e.g. the proteins expressed by a cell under two different physiological conditions) are labeled separately on the side chains of their reduced cysteine residues using one of two isotopically different but chemically identical sulfhydryl-reactive ICAT reagents (Fig. 1A). One of the ICAT reagents is an isotopically normal reagent containing hydrogen atoms on the carbon backbone (referred to as the d₀ reagent), and the other is an isotopically heavy (d₈) reagent, where the hydrogen atoms have been replaced with deuterium atoms. As the pairs of peptide labeled with the d₀ and d₈ versions of the ICAT reagent are chemically identical, according to stable isotope dilution theory (51) they serve as mutual internal standards for accurate protein quantification. The relative quantity of each protein present in the two biological samples is therefore determined by measuring the relative signal intensities of pairs of isotopically labeled, concurrently eluting peptides using an initial mass spectral scan. The identification of the proteins is accomplished by switching the instrument to MS/MS mode in which it selects peptides for CID.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Global, quantitative mass spectrometric analysis of protein expression. A, the structure of the ICAT reagent. B, mass spectrometric analysis using selective protein labeling with the ICAT reagent and multidimensional chromatography. Equal amounts of total protein are isolated from cells existing in two different biological states and labeled with the d0 or d8 versions of the ICAT reagent. The proteins are mixed, enzymatically digested, separated by multidimensional chromatography, and analyzed by MS. Relative quantification of protein expression between the two states is accomplished by comparison of peak intensities of the isotopically different peptides, and identification is accomplished by selecting these peptides for MS/MS and subsequent sequence data base searching with the generated CID spectra.

Alternative methodologies for quantitative protein analysis using selective peptide labeling and MS have also been developed recently (57, 58). Munchbach et al. (57) labeled the N termini of peptides derived from 2DE-separated proteins using a stable isotope-containing reagent to profile the effects of carbon source restriction on protein expression in Escherichia coli. CID of these N-terminal labeled peptides generates unique mass signatures that facilitate de novo peptide sequencing by MS/MS analysis (i.e. sequence determination from interpretation of raw MS/MS spectra without the need for data base searching). A similar approach involves the labeling of carboxylic acid residues on peptides with isotopically normal or heavy methanol, which allows for both relative quantification of protein expression and de novo peptide sequencing.² Another alternative approach to quantitative protein analysis employs metabolic incorporation of stable isotope-containing amino acids into differentially expressed proteins isolated from cell cultures grown on either normal media or media enriched or depleted in stable isotope-containing amino elements (59, 60).

	Multidimensional Separation Strategies

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

Along with selective labeling and purification of proteins, promising approaches that employ multiple, orthogonal liquid chromatography steps in conjunction with automated ESI tandem mass spectrometric analysis have been developed recently as an alternative strategy to 2DE for analyzing complex protein mixtures (61-63). Most notably, Link et al. (63) described an integrated system that employed a biphasic two-dimensional µLC column packed with strong cation exchange (SCX) and RP materials, enabling sequential separation of peptides first by electrostatic charge and then by hydrophobicity, followed by online MS/MS analysis. This method has proven effective for the comprehensive analysis of protein complexes (63), and more recently it has been applied to the identification of nearly 1500 proteins, including low abundance proteins, from a whole-cell yeast lysate (64). Gygi et al. (44) have used a similar approach that employs SCX high performance liquid chromatography (HPLC) in conjunction with off-line RP-µLC-ESI MS/MS and have shown that this method enables the analysis of low abundance proteins in S. cerevisiae that 2DE-based approaches are not sensitive enough to detect.

We have also coupled this multidimensional LC methodology with the ICAT strategy described above to form an integrated approach to the quantitative analysis of protein expression (Fig. 1B). We have applied this strategy to the investigation of changes in the protein expression profile after metabolic perturbation in yeast cells (56) to detect differentially induced changes in the membrane protein composition in UL-60 cells (65), as well as to detect differences in protein profiles isolated from simulated and non-stimulated androgen-dependent human prostate cell lines.³ Additionally, by correlating protein expression data with cDNA array data measuring the corresponding mRNA expression levels, we now have the tools necessary for the global analysis of gene expression on a system-wide level. In an initial study we have applied these two complementary gene expression profiling approaches to the systematic analysis of metabolic pathways in yeast (10).

	Future Prospects

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

It can be anticipated that the traditional 2DE-MS-based approach as well as alternative, second generation proteomics technologies will continue to rapidly evolve and diversify over the next few years. These advances will be accelerated by exciting hardware and software developments related to mass spectrometry. Described below is anticipated progress related to second generation proteomics approaches.

Development of Additional Labeling Chemistries-- The combination of group-specific chemistries, stable isotope dilution, and automated MS/MS can also be used to quantitatively and systematically assess properties of proteomes other than their composition. For example, we have recently developed a method for the quantitative analysis of protein phosphorylation on a proteome-wide scale (66). It involves the specific labeling of phosphate groups on peptides, enabling the isolation of these phosphopeptides for subsequent mass spectrometric analysis using a solid-phase purification strategy. This method also includes a labeling of the carboxylic acid groups of the peptides with stable isotope tags, thus facilitating the quantitative analysis of phosphorylated proteins by RP-µLC-ESI MS/MS in a manner similar to the ICAT strategy. This approach applied to the proteins contained in a total yeast cell lysate identified in a single RP-µLC-ESI MS/MS experiment numerous phosphorylation sites on 13 phosphoproteins, many of which were not previously known to be phosphorylated proteins. The development of chemistries (67-69) that select proteins based on their state of activity rather than simply on the presence of specific amino acid residues represents an exciting opportunity to analyze proteomes functionally (70). This is achieved by the synthesis of chemical reagents that selectively bind to the active site of specific enzymes in an activity-dependent manner. Liu et al. (67) have developed a chemistry that specifically targets the active site of catalytically active serine hydrolases, and Greenbaum et al. (69) have developed an analogous chemistry for the selective labeling of catalytically active cysteine proteases. Similar chemistries, coupled with stable isotope tagging, could enable the quantitative, proteome-wide characterization of the function and activity of selected classes of proteins by MS.

Advances in Mass Spectrometric Instrumentation-- It can be anticipated that parameters critical to the performance of mass spectrometers, including detection sensitivity, sample throughput, mass resolution, and mass accuracy, will continue to improve, although not necessarily all on the same instrument. The recent introduction of a mass spectrometer that combines a MALDI source with a selection quadrupole (Q), a collision cell (q), and a TOF fragment ion analyzer (MALDI QqTOF) (71, 72) offers the previously unavailable ability to combine the sensitivity, amenability to automation, and mass accuracy of MALDI-TOF MS with the capabilities of MS/MS. We have shown the MALDI QqTOF instrument to be effective for the analysis of ICAT-labeled proteins (73) with the main advantage of the approach being that those proteins showing significant differential expression between the two biological conditions can be selectively identified by MS/MS analysis, whereas those peaks showing little or no differences in expression can be omitted. The option to selectively analyze only the differentially expressed proteins could lead to a dramatic increase in throughput for proteome-wide protein expression profiling studies. Another recent advance in instrumentation is the development of a MALDI TOF-TOF mass spectrometer (74). This instrument combines the same advantages of the MALDI QqTOF instrument with the added potential for extremely fast analysis times and thus the potential to significantly increase sample throughput, making the identification of thousands of proteins per day possible. Another type of instrument that holds great potential for proteomic studies is the Fourier Transform (FT) mass spectrometer, which gives significantly increased sensitivity, resolution, and mass accuracy relative to other mass spectrometers (60, 75, 76). It has been demonstrated in yeast that with added constraints, the mass accuracy obtained by FT-MS analysis of peptides is sufficient to uniquely identify the peptides (and thus the proteins from which they are derived) from a data base without the need for tandem mass spectrometric analysis (77, 78). This allows for the detection of low abundance proteins that are many times missed when sequencing peptides by standard MS/MS methods, and thus this approach has great promise for improved sensitivity and increased throughput in proteome-wide analyses. Collectively, these advances are expected to dramatically change the performance of proteomic technology, with respect to sensitivity, level of automation, sample throughput, and accuracy. Furthermore, it can be anticipated that the ability to measure additional properties of proteins will move proteomics ever closer to the comprehensive analysis of biological function.

	ACKNOWLEDGEMENT

We thank David Goodlett for his helpful comments on this manuscript.

	FOOTNOTES

^* This minireview will be reprinted in the 2001 Minireview Compendium, which will be available in December, 2001. This work was supported in part by a grant from the Merck Genome Research Institute (MGRI) and NCI, National Institutes of Health Grant 1R33CA84698.

Funded by a National Institutes of Health postdoctoral genome training grant fellowship. To whom correspondence should be addressed: Institute for Systems Biology, 4225 Roosevelt Way N. Suite 200, Seattle, WA 98105. Tel.: 206-732-1359; Fax: 206-732-1299; E-mail: tgriffin@systemsbiology.org.

Published, JBC Papers in Press, October 3, 2001, DOI 10.1074/jbc.R100014200

² D. Goodlett, personal communication.

³ M. E. Wright and R. Aebersold, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: 2DE, two-dimensional electrophoresis; RP-µLC, reverse-phase microcapillary liquid chromatography; MS, mass spectrometry; MS/MS, tandem mass spectrometry; ESI, electrospray ionization; MALDI, matrix-assisted laser desorption/ionization; TOF, time-of-flight; CID, collision-induced dissociation; ICAT, isotope-coded affinity tag; SCX, strong cation exchange; HPLC, high performance liquid chromatography.

	REFERENCES

TOP INTRODUCTION The Emergence of Proteomics:... Apparent Limitations of the... The Emergence and Initial... Selective Protein Labeling and... Multidimensional Separation... Future Prospects REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				K. M. Waters, J. G. Pounds, and B. D. Thrall Data merging for integrated microarray and proteomic analysis Brief Funct Genomic Proteomic, December 1, 2006; 5(4): 261 - 272. [Abstract] [Full Text] [PDF]

				J. A. Stead, J. N. Keen, and K. J. McDowall The Identification of Nucleic Acid-interacting Proteins Using a Simple Proteomics-based Approach That Directly Incorporates the Electrophoretic Mobility Shift Assay Mol. Cell. Proteomics, September 1, 2006; 5(9): 1697 - 1702. [Abstract] [Full Text] [PDF]

				L. M. Schnapp, S. Donohoe, J. Chen, D. A. Sunde, P. M. Kelly, J. Ruzinski, T. Martin, and D. R. Goodlett Mining the Acute Respiratory Distress Syndrome Proteome: Identification of the Insulin-Like Growth Factor (IGF)/IGF-Binding Protein-3 Pathway in Acute Lung Injury Am. J. Pathol., July 1, 2006; 169(1): 86 - 95. [Abstract] [Full Text] [PDF]

				J. Hirsch, K. C. Hansen, A. L. Burlingame, and M. A. Matthay Proteomics: current techniques and potential applications to lung disease Am J Physiol Lung Cell Mol Physiol, July 1, 2004; 287(1): L1 - L23. [Abstract] [Full Text] [PDF]

				B. Habermann, J. Oegema, S. Sunyaev, and A. Shevchenko The Power and the Limitations of Cross-Species Protein Identification by Mass Spectrometry-driven Sequence Similarity Searches Mol. Cell. Proteomics, March 1, 2004; 3(3): 238 - 249. [Abstract] [Full Text] [PDF]

				P. M. Lonosky, X. Zhang, V. G. Honavar, D. L. Dobbs, A. Fu, and S. R. Rodermel A Proteomic Analysis of Maize Chloroplast Biogenesis Plant Physiology, February 1, 2004; 134(2): 560 - 574. [Abstract] [Full Text] [PDF]

				D. B. Martin, D. R. Gifford, M. E. Wright, A. Keller, E. Yi, D. R. Goodlett, R. Aebersold, and P. S. Nelson Quantitative Proteomic Analysis of Proteins Released by Neoplastic Prostate Epithelium Cancer Res., January 1, 2004; 64(1): 347 - 355. [Abstract] [Full Text] [PDF]

				L. V. LEAK, E. F. PETRICOIN III, M. JONES, C. P. PAWELETZ, A. M. ARDEKANI, V. A. FUSARO, S. ROSS, and L. A. LIOTTA Proteomic Technologies to Study Diseases of the Lymphatic Vascular System Ann. N.Y. Acad. Sci., December 1, 2002; 979(1): 211 - 228. [Abstract] [Full Text] [PDF]

				G. C. Adam, E. J. Sorensen, and B. F. Cravatt Chemical Strategies for Functional Proteomics Mol. Cell. Proteomics, October 1, 2002; 1(10): 781 - 790. [Abstract] [Full Text] [PDF]

				K. Sasaki, K. Sato, Y. Akiyama, K. Yanagihara, M. Oka, and K. Yamaguchi Peptidomics-based Approach Reveals the Secretion of the 29-Residue COOH-Terminal Fragment of the Putative Tumor Suppressor Protein DMBT1 from Pancreatic Adenocarcinoma Cell Lines Cancer Res., September 1, 2002; 62(17): 4894 - 4898. [Abstract] [Full Text] [PDF]

N. Jessani, Y. Liu, M. Humphrey, and B. F. Cravatt Enzyme activity profiles of the secreted and membrane proteome that depict cancer cell invasiveness PNAS, August 6, 2002; 99(16): 10335 - 10340. [Abstract] [Full Text] [PDF]