Mass spectrometry-based proteomics is an approach to studying protein expression, posttranslationalmodifications, and interactions that provides a wide range of possibilities for analyzing protein functions on a global level. Whereas mass spectrometry is the key tool in a proteomic workflow, carefully selected protein extraction and fractionation methods are prerequisite for successful analysis (Figure 1). Proteomic studies require simplification of the protein mixtures into less-complex components that are more amenable to analysis. For this purpose, proteins are usually separated either by 1-dimensional (1D) or 2D gel electrophoresis or by a chromatographic technique. Because measurement of masses of fragmented peptides is more accurate than measuring masses of intact proteins, a typical proteomic experiment involves digestion of protein before mass spectrometric analysis. Mass spectrometry measures masses of peptides and reveals their primary structures.1 Searching these information databases enables identification of proteins.
MASS SPECTROMETRY-THE KEY TOOL FOR PROTEOMICS
Mass spectrometry is a physicochemical analysis technique that determines the mass to charge ratio of ions in a gas phase, which is the basis of various strategies for identification and characterization of proteins in complex mixtures. In proteomic research, 2 major types of spectrometers are used, which differ in the way in which proteins or peptides are ionized for mass determination. In the first, the electrospray ionization, a solvent containing ionized analyte, is pumped through a very small, charged capillary (Figure 2). In this way, the generated aerosol contains droplets of solvent and analyte. Following evaporation of the solvent, the ions move to the mass analyzer of the mass spectrometer. Matrix-assisted laser desorption/ionization is the second type of ionization technique, where a laser beam excites the solid matrix containing the analyte molecules. The ionization reaction takes place in the desorbed matrix-analyte cloud, just above the surface. The ions are then extracted into the mass spectrometer for analysis.
Modern mass spectrometers allow identification of peptides and proteins at subfemtomolar (<10^sup -15^ mol) levels, thus, potentially enabling the study of low-abundant proteins, such as most proteins involved in signal-transduction processes. However, the major challenge in identification of low-abundant proteins is the dynamic range of measurements in mass spectrometers. The instruments, with fewer than 4 orders of magnitude of dynamic range, face up to billion-fold differences in protein abundances in cells. This limitation can be circumvented by a welldesigned strategy of cell, organelle, or protein fractionation. The combination of properly designed extraction and separation techniques, followed by advanced, highaccuracy mass spectrometry, allows identification of thousands of proteins or their posttranslational modifications from minute amounts of tissues, such as clinical samples.
Quantitative Proteomics
Identification of biomarkers or drug targets requires tools for a quantitative comparison of protein levels between analyzed samples. Differences between healthy and diseased tissues or between drug-treated and drug-untreated cells can be measured using different proteomic approaches (Figure 3). In the most widespread and simplest method for quantitation of protein, abundances are measured by densitometry of stained 2D gel, before mass spectrometric identification of the protein. Because identification of gel-resolved proteins is limited to proteins of high or intermediate abundance, alternative mass spectrometer-based methods have been developed. Roughly, they can be classified as methods that are either label-free or that are based on incorporating stable isotopes.2
In the label-free approach, the intensities of mass spectral peaks are typically compared directly between samples, based on their mass and retention times.3,4 Label-free methods are well suited for multiple replicate and batch analysis; however, they require that the sample-preparation and analytical systems be robust and reproducible and that sophisticated data-processing capabilities exist to support the analysis.
The most popular quantification approach uses metabolic labeling of cultured cells, stable-isotope labeling by amino acids in cell culture, a technique in which either the whole medium is labeled with 15N or the medium is supplemented with amino acids containing 2H or 13C isotopes. 5 A major limitation of this approach is that it is limited to biological material cultured in vitro and, therefore, labeling of proteins extracted from tissues has to be achieved in alternative ways. For this purpose, a variety of reagents carrying stable isotopes have been developed that are based on derivatization of the thiol moiety of cysteines or primary amino groups. Thiol-reactive reagents that carry affinity tags, such as isotope-coded affinity tags6 or HysTag,7 offer a several-fold reduction in sample complexity when combined with affinity chromatography. Similarly, up to 2 18O atoms can be incorporated into carboxyl groups of peptides by digestion with trypsin or other endopeptidases in the presence of H2 18O.8 Such approaches are gaining in popularity and are now being used in larger-scale projects.
Multiplexed isobaric tagging (eg, iTRAQ, Applied Biosystems, Foster City, Calif)9 represents another group of methods that uses a chemical-tagging reagent that allows multiplexing of samples. In single MS mode, the differentially labeled versions of a peptide are indistinguishable. However, in MS2 mode (in which peptides are isolated and fragmented), each tag generates a unique reporter ion. Protein quantitation is then achieved by comparing the intensities of the 4 reporter ions in the MS2 spectra.
Posttranslational Modifications
Proteins are often modified to their functional forms through regulated posttranslational processing, including processing of the polypeptide chain and modification of residues. There are hundreds of types of posttranslational modifications of proteins, such as phosphorylation, glycosylation, and acetylation, to mention a few. In disease, an appearance of, or change in, the abundance of posttranslational modifications can be considered a biomarker. Within a short time, mass spectrometry became the ultimate technology for mapping protein modifications.10 Identification of posttranslational modifications, which usually are present in only a fraction of a protein, requires affinity enrichment. Most advanced are the technologies for mapping phosphorylation sites, which involves affinity enrichment. These technologies can routinely provide information on thousands of phosphorylation sites and their dynamics in a single study.11
PERSPECTIVES AND LIMITATIONS
DNA constitutes the genomic information. Messenger RNA reveals the activities of genes and the primary structures of proteins. Proteomics provides knowledge on the nature of the mature final effectors in cellular processes- the proteins. The application of mass spectrometry-based proteomics to the study of diseases will ultimately lead to identification of biomarkers for the detection, diagnosis, and prognosis of diseases.12 The achieved information will facilitate development of novel medicines.
In the past, enormous effort in biomarker discovery was mostly limited to molecular biology technologies. Studying nucleic acids can reveal primary structures and changes in the abundance of proteins. Proteomics-based approaches can, in addition, enable mapping of protein changes introduced posttranslationally, such as modifications of single residues, or processing of the polypeptide chain. Moreover proteomic studies lead to discoveries of proteins that have not been predicted from the genome or to those for which predictions were ambiguous. Another great opportunity in proteomics is its unprecedented potential for studying protein-protein interactions. Understanding interaction networks can facilitate identification of abnormalities in diseases.
Although the field of mass spectrometry has advanced enormously in recent years, there are significant technical challenges that imply limitations to the routine application of proteomics to clinical research, including limitations in sensitivity and in the dynamic range of mass spectrometers, and difficulties in the availability and processing of clinical samples. Even though the sensitivity of mass spectrometry is usually several orders of magnitude higher than classical biochemical approaches, it does not allow identification of low-abundant proteins from a clinical, biopsy-sized sample. When complex mixtures, such as cell lysates, are analyzed, the routine sensitivity of mass spectrometry is around 1 fmol. Because 1 mol corresponds to 6.02 × 1023 molecules, identification of a protein requires approximately 1 billion molecules, which means that, to identify a protein occurring at 1000 copies per cell, 1 million cells of starting material are required. That is a relatively high amount, which cannot be easily obtained from a biopsy sample or be isolated by laser microdissection from tissue. However, even these minute amounts of clinical material are enough to detect known biomarkers, such as HER-2/neu and estrogen receptor in breast cancer samples, in routine proteomic analysis. Future developments in mass spectrometry and identification methods will probably circumvent some of these limitations and will continue to revolutionize modern biology and to influence clinical research.
© 2008 College of American Pathologists Provided by ProQuest LLC. All Rights Reserved.
Copyright 2008 Archives of Pathology & Laboratory Medicine
