1 An indispensable tool for systems biology: mass spectrometry-based proteomics Wing Yung Mass spectrometry is an instrument that can identify and quantify thousands of proteins from complex sample. To date, proteomics, the large-scale analysis of proteins, has combined with mass spectrometry to analyze quantitative protein profiles, protein-protein interactions and post-translational modifications of proteins. By providing diverse, quantitative, high-quality proteomic data, mass spectrometry based proteomics becomes an essential component of systems biology that seeks understanding of biological processes and developing predictive models of biological systems. Keywords: Protein, proteomics, mass spectrometry, systems biology I. INTRODUCTION P is a systematic study of the many and diverse properties of proteins in parallel to provide detailed descriptions of structure, function and control of biological systems in health and disease. In general, proteomics deals with the large-scale determination of gene and cellular function directly at the protein level. It is a particularly rich source of biological information because proteins are involved in almost all biological activities, which contribute greatly to our understanding of biological system. Proteomics, an analysis process, is an essential component of the systems biology approach that seeks to comprehensively describe biological systems through integration of diverse types of data and to allow computational simulations of complex biological system in the future [3]. Proteomics can be divided into three main areas: 1) protein micro-characterization for large-scale identification of proteins and their post-translation modifications; 2) differential display proteomics for comparison of protein levels with potential application in a wide range of diseases; and 3) studies of protein-protein interactions using techniques such as mass spectrometry of the yeast two-hybrid system [1]. The ability of mass spectrometry (MS) to identify ever smaller amounts of protein from increasingly complex mixtures is a primary driving force in proteomics as described in the Nature by Mike Tyers et al [2]. Patterson et al defined mass spectrometry as accurate measurement of charged analytes (ions); in the context of proteomics, analytes are usually peptides or less frequently ions; a mass spectrometer ROTEOMICS measures the mass to charge ratio of charged species under vacuum and comprises. MS-based proteomics is a discipline made possible by the availability of gene and gnome sequence databases and protein ionization methods. MS-based proteomics has established itself as an indispensable technology to interpret the information encoded in genomes. So far protein analysis (primary sequence, post-translational modifications or protein-protein interactions) by MS has been most successful when applied to small sets of proteins isolated in specific functional contexts [3]. In this paper, I will firstly describe what mass spectrometry is and how can it identify and quantify thousands of proteins. I will then review the applications of combined MS and proteomics, to analyze protein profile, protein-protein interactions and post-translation modification. This paper will be concluded with key challenges and perspective of MS based proteomics, especially it acts as an input (proteomic data) and/or a component (proteomic analysis) of systems biology to achieve system level understanding of cells or networks of cells [8]. II. MASS SPECTROMETRY A. Ionization Mass spectrometry (MS) is carried out in the gas state on ionized analytes. By definition, a mass spectrometer consists of ion source, a mass analyzer that measure mass-to-charge ratio of the ionized analytes, and a detector to register the number of ions at each mass-to-charge value. Electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI) are the two common techniques to ionize the proteins or peptides for mass spectrometry. Patterson et al described MALDI as a process by which ion formation is promoted by short laser pulses; the sample is deposited on a sample plate into the source and then embedded in a matrix that promotes ionization; a laser fired at the sample that is co-crystallized with the matrix results in the desorption of the analyte from the sample plated and its ionization [3]. Mann stated in Annual Review Biochem that the precise nature of the ionization process in MALDI is still largely unknown and the signal intensities depend on incorporation of the peptides into crystals, their likelihood of capturing and /or retaining a proton during the desorption process., and a number of other factors including suppressing effects in peptide mixtures. Proteins generally undergo fragmentation to 2 some extent during MALDI, resulting in broad peaks and loss in sensitivity; therefore MALDI is mostly applied to the analysis of peptides [6]. Different from MALDI, ESI takes place in atmosphere and is therefore very gentle without fragmentation of analyte ions in gas phase. The molecules are transferred into the MS with high efficiency for analysis. A wide range of compounds especially protein can be analyzed by ESI. Large ions are typically multiply charged which brings them into the range of mass to charge ratios of typical MS. The distribution of charges gives rise to the typical multiple charge envelopes. These spectra can be simplified by deconvolution, an algorithm that sums up the signal intensity into a single peak at the molecular weight of the analyte [6]. B. Mass Analyzer According to Abersold et al, the mass analyzer is central to the MS technology. In the context of Proteomics, its key parameters are sensitivity, resolution, mass accuracy and the ability to generate information-rich ion mass spectra. There are three basic types of mass analyzer used in mass spectrometryproteomics. These are time of flight (TOF), quadruple and three-dimensional trapping field (ion trap, Fourier transform ion cyclotron) analyzers. The analyzers can be stand-alone and put together in tandem to take advantage of the strengths of each [4]. C. MALDI Time-of-Flight MS Aebersold stated that because of its simplicity, excellent mass accuracy, high resolution and sensitivity, MALDI-TOF (figure 1) which is known as peptide-mass mapping or peptidemass fingerprinting is the first approach. In this approach, the mass spectrum of the eluted peptide mixture is acquired, which results in a peptide-mass fingerprint of the protein being studied. Because mass mapping requires a purified target protein, this approach is commonly used with prior protein fractionation using two-dimensional gel electrophoresis (2DE) [4]. This mass spectrum is obtained by MALDI, which results in a TOF distribution of the peptides comprising the mixture. The reason for using peptides rather than proteins is that gelseparate proteins are difficult to elute and to analyze by mass spectrometry, and the molecular weight of proteins is not sufficient database identification. In contrast, peptides are easily eluted from gels and even a small set of peptides from a protein provides sufficient information for identification [1]. Mann identified the fact that some of the peptide ions decay because of the energy departed in the desorption process has been used to obtain structural information in a technique termed post-source decay. It can provide important structural information but does not serve as general peptide sequencing method because it is not sufficiently sensitive and simple to control [6]. Figure 1 Schematic of MALDI process (A) & MALDI-TOF instrument (B) The mass-to-charge ratio is related to time, it takes an ion to reach the detector the lighter ions arrive first. D. ESI Quadrupole MS The quadrupole is a mass filter which consists of four rods to which an oscillating electric field is applied and which lets only a certain mass pass through. To date most peptide sequencing experiments is performed on triple quadrupole that consist of three sections: two mass-separating quadrupole sections separated by a central quadrupole (or a higher multiple) section whose function is to contain the ions during fragmentation. Quadrupole MS are capable of unit mass resolution and mass accuracy of 0.1-1Da and excel at quantitative measurements. The triple quadrupole can be programmed for a variety of different scan modes in addition to the isolation of the peptide followed by obtaining a mass spectrum of the fragments (collision-induced (CID) spectra) [6]. Figure 2 Schematic of a quadrupole TOF instrument. Ions are separated in Q1 and dissociated in q2. They enter the TOF analyze through a grid and are pulsed into the reflector and onto the detector, where they are recorded. In recent years the triple quadrupole has begun to be complemented by the quadrupole time-of-flight (TOF) instrument (figure 2) in which the third quadrupole section is replaced by a TOF analyzer. Its main advantages are that it 3 provides the high accuracy for mass and high resolution, resulting in unambiguous determination of charge state and very high specificity in the database searches. In another method, the peptides are ionized by ESI directly from the liquid phase. The peptide ions are spayed into a tandem mass spectrometer which has the ability to resolve peptides in the mixture and isolate one species at a time. The main advantage of it over MALDI fingerprinting is that sequence information derived from several peptides is much more specific for the identification of a protein than a list of peptide masses. The fragmentation data can not only be used to search protein sequence databases but also nucleotide database such as expressed sequence tag databases and even raw genomic sequence databases [1]. E. Liquid Chromatography and Tandem MS Liquid chromatography (LC) coupled to tandem mass spectrometry called LC-MS/MS, is a power technique for the analysis of proteins and peptides. Complicated mixtures containing hundreds of proteins vary by orders of magnitude, LC-MS/MS can be used alone or in combination of with 1DE or 2DE or other protein purification techniques [6]. Before LC-MS/MS can be used, there are two technical issues need to be addressed. 1) The relationship between the amount of analyte present and measured signal intensity is complex and incompletely understood. MS is therefore a poor quantitative device; and 2) the amount of data collected by the method is huge and its analysis daunting. The question of what constitutes an identified protein in a LC-MS/MS experiment has been difficult to answer. It is therefore important that computer programs that use robust and transparent statistical principles to estimate accurate probabilities indicating the likelihood for the presence of a peptide or protein in the sample are further developed and widely tested and applied [4]. Mann echoed that if it is the approach of using proteolytic digestion of mixtures of proteins and LC-MS/MS analysis of the peptides generated, it is no longer possible to identify proteins based on peptide mass profiling because peptides are scrambled and multiplexed. Instead, tandem mass spectral data must be generated and interpreted. He saw software control and automation accelerate the process of acquiring the MS/MS data and hundreds of MS/MS spectra can be generated in a single run [6]. Data analysis tools are essential in these high capacity experiments. III. MS-BASED PROTEOMICS MS-based proteomics has established itself as an indispensable technology to interpret the information encoded in genomics. So far protein analysis (primary sequence, posttranslational modification (PTM) or protein-protein interactions) by MS has been most successful when applied to small sets of protein isolated in specific functional contexts. The systematic analysis of the much larger number of proteins expressed in a cell, an explicit goal of proteomics is now also rapidly advancing. A. Protein profiling MS-based proteomics has interfaced well with 1) the generation of protein-protein linkage map; 2) the use of protein identification technology to annotate and correct genomics DNA sequences and 3) the use of quantitative methods to analyze protein expression profiles as a function of cellular state as an aid to infer cellular function. Stable-isotope dilution and LC-MS/MS are used to accurately detect changes in quantitative protein profiles and to infer biological function from the observed patterns [4]. The most valuable information on the system being studied is obtained from those proteins that are expressed differentially in a matrix of proteins of unchanged expression, therefore proteomic technologies detecting difference in protein profiles need to be quantitative. Two peptides of identical chemical structure that differ in mass because they differ in isotopic composition are expected according to stable isotope dilution theory, to generate identical specific signals in as mass spectrometer [3]. Gygi et al described an approach based on a class of new chemical reagents termed isotope-code affinity tags (ICATs) and tandem mass spectrometry (LC MS/MS and sequence database searching). The reagent exists in two forms heavy (d8) and light (d0). This approach is based on two principles. First, a short sequence of contiguous amino acids from a protein contains sufficient information to identify that unique protein. Second, pairs of peptides tagged with the light and heavy reagents are chemically identical and serve as ideal mutual internal standards for accurate quantification. In MS, the ratio between the intensities of the lower and upper mass components of these pairs of peaks provide an accurate measure of the relative abundance of the peptides (and hence the proteins) in the original cell pools because the MS intensity response to a given peptide is independent of the isotopic composition of the ICAT reagents [5]. B. Analysis of post-translation modification Proteins are converted to their mature form through a complicated sequence of post-translational protein processing events. Many of the post translational modifications (PTM) are regulatory and reversible. Pandey et al. stated that one of the unique features of proteomics studies is the ability to analyze the PTM of proteins. Phosphorylation, glycosylation and sulphation as well as many other modifications are extremely important for protein function as they can determine activity, stability, localization and turnover. MS is the proteomic method of choice to determine protein modifications and the task is more difficult than protein identification [1]. Mann et al said that the MS methods have been refined to using peptide mapping with different enzymes to “cover” as much of the protein sequence as possible. Protein modifications are then determined by via computer-aided interpretation. For analysis of some types of PTMs, specific MS techniques have been developed that scan the peptides derived from a protein for the presence of a particular 4 modification. But, the analysis of regulatory modification remains a challenge due to the size and ionizability of peptides bearing the modifications and their fragmentation behavior in the MS [4]. Given the difficulties of identifying all modifications in a single protein, at present, scanning for proteome-wide modification is not comprehensive. Another approach is instead of searching the database only for non-modified peptides, the databases algorithm is instructed to also match potentially modified peptides. In summary the experiment is divided into identification of a set of protein based of nomodified peptides followed by searching only these proteins for modified peptides [4]. Overall, many challenges remain in the large-scale mapping of PTMs, but it is clear that MS-based proteomics can make a unique contribution in this area, for example systematic quantitative measurements of PTMs by stable-isotope labeling. C. Analysis of protein interactions Most proteins exert their function by way of protein-protein interactions and enzymes are often held in tightly controlled regions of the cell by such interactions. The question is to what protein does it bind? MS with ICAT strategy of Gygi et al [5] – fully processed and modified protein can serve as the bait that the interactions take place in the native environment and cellular location and that multicomponent complex can be isolated and analyzed in a single operation. However, because many biologically interactions are of low affinity, transient and generally dependent on the specific cellular environment in which they occur. MS based methods in a straightforward affinity experiment only detect a subset of the protein interactions that actually occur. Bioinformatics methods, correlation of MS data with those obtained by other methods or iterative MS measurements possibly in conjunction with chemical cross-linking can often help to further explain direct interactions and overall topology of multi-protein complexes [4]. Mann et al described epitope-tagging strategy, an example of functional proteomics, used in identification of interacting proteins. The cDNA of interest is first cloned into a vector that provides an epitope tag. This is followed by transfection of the tagged “bait” into the cell of interest. The cells are then lysed and the lysates purified by affinity purification using an antibody against epitope. Proteins bound specifically to the bait protein are eluted by competitive elutin using a peptide that encodes the epitope. The proteins are then resolved by gel electrophoresis followed by mass spectrometric identification. Mann continued to describe a modification to this approach which is two epitope tags. This may provide decreased binding to nonspecific proteins as well as to improve the recovery of the protein complex [6]. Pandey et al believed that functional proteomics will provide a wealth of protein-protein interaction data, which will probably be its most important and immediate impact on biological science. Because proteins are one step closer to function than are genes, these studies frequently lead directly to biological discoveries or hypothesis [1]. IV. CHALLENGES & EXPECTATIONS As proteins are involved in essentially all biological functions and clinical conditions, MS and proteomics will have an even greater impact on biology and medicine that it has had so far. MS-based proteomics interfaces particularly well with cell biological studies for studying specific protein functions. The success is built on the proven potential of MS techniques to rapidly identify almost any protein, to analyze that protein for the presence of post-translational modifications, to determine how and with what other biomolecules that proteins interact [4]. This shift in focus from the analysis of selected isolated proteins to proteome-wide analyze but poses yet unmet challenges as following. Proteomic MS, different from protein MS, often collect large amounts of data in the absence of hypotheses concerning specific proteins or activities. Successful proteomics experiments need to be designed in such a way that they can take advantage of the power of statistics for interpretation. To achieve this goal, carefully controlled repeat studies and the generation of models describing the source, magnitude and distribution of error will be essential. MS-based proteomics result in large amounts of data. Data collection at a volume and quality that is consistent with the use of statistical methods is a significant limitation of proteomics today. In a typical LCMS/MS, approx. 1000 CID spectra can be acquired per hour. It would take a long time to analyze complete proteomes. High-throughput collection of consistently high quality data remains a challenge in proteomics. The analysis and interpretation of the enormous volumes of proteomic data is also a challenge. In order to compare data from different labs, proteomic tools in the area of data analysis, data storage, visualization and communication are critical. The publication of the large data sets generated by proteomics experiments and the information contained therein poses significant challenge [4]. I agreed with Patterson et al those three challenges needs to be addressed in order for proteomics to have a substantial impact on the systems biology model. The first challenge is the enormous complexity of the proteome. The second challenge is the limited throughput of today’s proteomic platforms: iterative, systematic measurements on differentially perturbed systems demand a sample that is not matched by current proteomic platform. The third challenge is lack of a general technique for the absolute quantification of proteins, eliminating the need of a reference sample [3]. V. CONCLUSION & PERSPECTIVE MS-based proteomics is an essential component to systems biology research because proteins are rich in information that has turned out to be extremely valuable for the description of biological processes. This include protein abundances, linkage maps to other proteins or other types of biomolecules including DNA and lipids, activities., modification states, 5 subcellular location and more. Unfortunately, with the exception of quantitative protein profiles and protein-protein interactions (discussed above), none of these properties can currently be measured systematically, quantitatively and with high throughput [7]. According to Patterson et al in short term, MS based proteomics also can be expected to provide partial data sets of sufficient quality, density and information content to provide the basis for generating sophisticated models of biological processes that will be able to simulate system properties such as adaptation and robustness, which may not be appear form the analysis of isolated elements of a system [3]. In long term, researchers expected that combining different genomic and proteomic results obtained from some biological system will substantially increase understanding of complex biological processes. For instance, mRNA expression profiles and protein expression profiles seem to be largely complementary and therefore contribute to more refined description of the system that each observation by itself is unable to provide [6]. Idekar et al stated the ultimate goal of systems biology is the integration of data represent and simulate the physiology of cell. More specifically, systems biology based on diverse and high quality proteomic data will define functional biological modules, reveal previously unrecognized connections biochemical processes and modules, and generate new hypothesis that can be tested by the targeted generation of more proteomic data [8]. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] A. Pandey, et al “Proteomics to study genes and genomes” Nature, Vol. 405, 15 June 2000, pp. 837-846 M. Tyers, et al “ From genomics to proteomics” Nature, Vol. 422, 13 March 2003, pp. 193-196 S. Patterson, et al “Proteomics: the first decade and beyond” Nature Genetics supplement, Vol. 33, March 2003, pp. 311-321 R. Aebersold, et al “Mass spectrometry-based proteomics” Nature, Vol. 422, 13 March 2003, pp 198-207 S. Gygi, et al “Quantitative analysis of complex protein mixtures using isotope-coded affinity tags” Nature Biotechnology, Vol. 17, October 1999, pp 994-999 M. Mann, et al “Analysis of protein and proteomes by mass spectrometry” Annu Rev. Biochem. Vol. 70, 2001, pp. 437-473 T Ideker, et al “Integrated Genomics and proteomics analysis of a systematically perturbed metabolic network” Science, Vol. 292, 2001, pp 929-934 T Ideker, et al “A new approach of decoding life: system biology” Annu. Rev. Genomics Human Genes, Vol. 2, 2001, pp 343-372