- Email: [email protected]
Molecular oncology in the post-genomic era: the challenge of proteomics
24
Review
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
Molecular oncology in the post-genomic era: the challenge of proteomics Simone Mocellin1, Carlo Riccardo Rossi1, Pietro Traldi2, Donato Nitti1 and Mario Lise1 1 2
Department of Oncological and Surgical Sciences, University of Padova, via Giustiniani 2, 35128 Padova, Italy CNR, Institute of Molecular Sciences and Technologies, Corso Stati Uniti, 4– 35127 Padova, Italy
One of the greatest challenges in modern medicine is to dissect the cascade of molecular events that lead to the development and progression of tumors. The genomewide study of proteins embodies the exciting promise of proteomics. Unlike genomics, proteomics aims ambitiously to study not only protein expression profiles but also protein functions, which should provide researchers with a more comprehensive view of the molecular machinery that governs tumor biology. Recent technological advances are making proteomics a user-friendly, high-throughput laboratory tool that is likely to have an unprecedented impact on the pace of discovery in medicine. After completing the sequencing of the human genome, the shift towards studying the proteome seems to be a logical step in the highly challenging task of dissecting the molecular basis of health and disease [1]. The term ‘proteome’ was first coined to describe the set of proteins expressed by a given cell population [2] and should be considered a dynamic entity that continuously changes to shape the myriad of biological phenomena that characterize cell life. The study of the proteome, termed ‘proteomics’, offers a wealth of opportunities, but it also presents enormous technical hurdles as it tries to catalog not only protein expression profiles but also protein functions. During the early years of proteomics and until relatively recently, profiling the expression of proteins primarily relied on the use of 2D electrophoresis (2DE), a rather low-throughput method that was later combined with mass spectrometry (MS) [3,4]. The dynamic nature of protein expression, protein interactions and posttranslational modifications requires that measurements are made as a function of time and cellular state. This type of study requires several data and, thus, the development of both high-throughput technologies for protein investigation and adequate bioinformatics tools for analyzing large data sets [5]. Recent technological advances in MS and array miniaturization are fostering the implementation of high-throughput proteomics in biological research, particularly in the field of oncology. Here, the delineation of altered protein expression, not only at the whole-cell or Corresponding author: Simone Mocellin ([email protected]).
tissue level, but also in subcellular structures, protein complexes and biological fluids, might lead to the development of novel biomarkers for the early detection of disease, the identification of new therapeutic targets and the acceleration of drug development, as well as to the development of more effective strategies to predict and to evaluate therapeutic effect and toxicity [6 – 8]. Although high-throughput proteomics is in its infancy and many advances are needed to overcome its limitations (Table 1), for the first time researchers are working on the ambitious project of studying cell biology by looking at the whole protein repertoire. Here we review the principles and technologies that support proteomics, as well as the main findings that have been so far reported in the field of oncology. Principles and technological platforms Proteomics deals with three fundamental biological issues: protein characterization (including single protein identification and/or quantification and sample protein profiling), protein – protein interactions (how proteins work together to build up metabolic and/or signaling pathways), and posttranslational modifications of proteins (how protein function can be activated, silenced or modulated). Depending on the aim of the research, two main categories of proteomics studies can be identified [9]. First, ‘expression proteomics’ addresses the issue of protein expression levels in a given sample, such as a tumor, normal tissue or body fluid; this approach is particularly useful for protein profiling studies aimed at determining the difference between normal and tumor tissues. The protein profile (or ‘signature’) of cancer should allow investigators to identify tumor markers and, ultimately, disease-specific and/or patient-tailored molecular targets for the development of effective, non-toxic antiblastic drugs. Second, ‘functional proteomics’ includes the analysis of protein – protein interactions, protein – non-protein (e.g. DNA, RNA or lipids) interactions and protein posttranslational modifications, which will help researchers to dissect the intricate network of molecular pathways underlying cellular activities in health and disease. These types of data should greatly improve our understanding of the biological process of tumor development and progression, thus fostering the
http://tmm.trends.com 1471-4914/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.molmed.2003.11.001
Review
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
25
Table 1. Limitations of the two main proteomics technological platforms MS-based proteomics
Protein microarray-based proteomics
Protein identification is limited to known proteins (the mass of proteins under investigation must be compared to those reported in existing protein databases) Protein quantification requires complex procedures to be accurately determined Protein interaction requires sophisticated procedures (e.g. tag-based co-purification)
Antibody-based protein microarrays require thousands (or even hundreds of thousands) of specific antibodies to cover the entire protein Miniaturization can cause problems of steric hindrance among probes (e.g. antibodies) Protein identification by antibodies can be misleading due to antibody cross reaction Protein function of a given biological sample can only be probed with few antibodies at a time
identification of not only single protein-specific therapeutic targets but also pathway-specific targets. From a technical point of view, two main types of technological platform are currently used for highthroughput oncoproteomics research: MS-based proteomics and microarray-based proteomics. Mass spectrometry Several proteomics platforms rely on MS, which is usually associated with other techniques such as electrophoresis or liquid chromatography, for separating the target protein from the sample protein mixture [10,11]. Currently, dedicated MS laboratories can profile thousands of samples and can identify and/or quantify thousands of proteins per day, and rapid advances in sample throughput, sensitivity and accuracy are upcoming [12]. Mass spectrometers are based on a combination of three essential components: (1) the ion source, which produces ions from the sample protein mixture; (2) the mass analyzer, which resolves ionized analytes on the basis of their mass-to-charge (m/z) ratio; and (3) the detector, which detects the ions resolved by the mass analyzer (Figure 1). Two types of MS are used for most proteomics studies: matrix-assisted laser-desorption ionization (MALDI) and electrospray ionization (ESI) MS instruments, which volatize and ionize the proteins or peptides for their further mass analysis (Figure 1). They are very different in design and performance, and each has its own strength and weakness. MALDI is usually coupled to timeof-flight (TOF) analyzers that measure the mass of whole proteins or peptides, whereas ESI has been mainly coupled to different MS instruments, such as ion traps or triple quadrupole instruments, to generate fragment ion (collisioninduced) spectra of selected precursor ions (Figure 1). MALDI-TOF is still widely used to study large proteins by what is known as ‘peptide-mass mapping’. In this method, proteins are identified by matching a list of experimental peptide masses with a calculated list of all of the peptide masses of each entry in, for example, a comprehensive protein database. For an in-depth analysis of individual MS spectral peaks, however, a combination of multiple MS analyzers in a tandem-MS (MS/MS) instrument, such as MALDI-TOF/TOF, ESI-ion trap or ESI-triple quadrupole, is usually adopted to perform a twostage or multistage mass analysis of ions; this approach is ideal for analyzing both biomacromolecules and small molecules because it yields an accurate determination (10 – 100 kDa) of the molecular weight of large proteins and fine structural information directly from the biomolecules (Figure 1). As compared with peptide mass mapping, this http://tmm.trends.com
approach provides information not only about peptide mass but also about peptide sequence, which is scanned against comprehensive protein sequence databases by using one of several different algorithms. Because protein identification relies on matches with sequence databases, however, high-throughput proteomics is currently restricted largely to those species for which comprehensive sequence databases are available. Protein microarray There is considerable benefit to be gained from microarray technology [13,14]. First, in principle thousands of proteins can be spotted on a single slide or similar support, enabling researchers to interrogate simultaneously the function of many different proteins with minimal sample consumption. Second, hundreds or even thousands of copies of an array can be fabricated in parallel, enabling the same proteins to be probed repeatedly with many different molecules under numerous conditions [15]. Protein microarray is finding its way into both expression and functional proteomics [15,16] (Figure 2a). In expression proteomics, protein-detecting microarrays comprise several different affinity reagents, such as antibodies, arrayed at high spatial density on a solid support. Each agent (or ‘probe’) captures its target protein from a complex mixture (e.g. serum or cell lysate), and the captured proteins are subsequently detected and quantified. The three main methods used for antibody microarray are shown in Figure 2b. In contrast to relatively unbiased separation methods such as gel electrophoresis and liquid chromatography, affinity-based approaches enable the investigator to direct the experiment. If the goal of the experiment is to study a particular biological process, only those proteins involved in that process need to be examined. The drawback is that these investigations require both a prior knowledge of the proteins to be studied and appropriate affinity reagents: for instance, the analysis of 30 000–50 000 human genes could require more than 200 000 highly specific antibodies, taking into consideration posttranslational modifications, degradation processes and mutation phenomena. Moreover, steric hindrance, competition and antibody specificity are the primary causes of nonspecific binding or background signals on a multiplexed protein microarray. Protein microarrays also provide a well-controlled, in vitro way in which to study protein function on a genome-wide basis. For this application, sample proteins (or peptides), rather than affinity reagents, are arrayed on a solid support and then probed with compounds of
26
Review
TRENDS in Molecular Medicine
(a)
Protein–peptide liquid mixture
Vol.10 No.1 January 2004
Sampling cone
Nozzle
Spray needle ESI Ions
(b)
Matrix plate
Laser source
Extraction grid
Ions
(c)
Ion source
Mass analyzer
Detector
MALDI
Mass spectra
Intensity
Electrodes
m z –1 Acceleration
TOF Reflector
(d)
TOF-1
Mass selector
TOF-2
Collision chamber TRENDS in Molecular Medicine
Figure 1. Mass spectrometry (MS) technology for proteomics. Two ionization techniques are currently used for biomolecules. (a) Electrospray ionization (ESI) is used to volatilize and to ionize peptides and proteins from liquid samples, and is usually coupled to two mass analyzers (tandem MS or MS/MS) for protein or peptide characterization. (b) For matrix-assisted laser-desorption ionization (MALDI), the analyte is first co-crystallized in a surplus of matrix molecules, which commonly comprise an organic acid on a solid support. A pulsed ultraviolet laser then evaporates the matrix and the analyte into the gas phase. In addition to desorption support, the matrix ionizes the analyte molecules by proton transfer. Single charged ions are mostly generated in this process. (c) The ions are accelerated by an electrical field and then enter the fieldfree mass analyzer, in which they are separated according to their velocity. The ions are turned around in a reflector, which compensates for slight differences in kinetic energy, and then impinge on a detector that amplifies and counts them as they arrive. From the time of flight (TOF) along the mass analyzer, the ratio between the mass and the charge (m z21) of the analyte can be calculated easily: for single charged ions, this value is equivalent to their molecular weight. MALDI-TOF is used for protein profiling purposes and protein identification by the peptide-mass mapping method. Because mass mapping requires an essentially purified target protein, this technique is often used in conjunction with prior protein fractionation using 2D electrophoresis (2DE) (see Figure 3). (d) Scheme of a MALDI-TOF/TOF instrument, shown as an example of tandem MS. The first mass analyzer (TOF-1) acts as a mass filter to separate analytes with the mass of interest from the rest of the ions. For example, the mass of a potential biomarker can be sorted from a mixture of serum proteins. Selected ions are then fragmented in a collision cell. The mass of the resulting fragments is finally analyzed by the second mass analyzer (TOF-2). MALDI-MS/MS enables protein identification, such as the characterization of a tumor marker identified in patient sera by protein profiling, and extends the use of MALDI to the analysis of more complex samples, thereby uncoupling MALDI-MS from 2DE. Other tandem-MS platforms, such as ion-trap MS or quadrupole-TOF MS, rely on similar principles.
interest. With this strategy, new substrates or ligands of known enzymes (or receptors or drugs), as well as diseaserelated posttranslational modifications, can be identified [17]. An advantage of studying protein function in an array format is that the investigator can control the conditions of the experiment. This includes factors such as pH, temperature, ionic strength and the presence or absence http://tmm.trends.com
of cofactors, as well as the modification states of the proteins under investigation. Furthermore, protein microarrays can be used to study the interaction of proteins with non-protein molecules, including nucleic acids, lipids and small organic compounds. An extension of these methodologies is tissue microarray technology (Box 1) – a high-throughput technique
Review
TRENDS in Molecular Medicine
27
Vol.10 No.1 January 2004
(a) Detection system
Sample proteins
Immobilization of recognition molecules
Target-protein binding
Data analysis
(b)
Sandwich assay
Antigen-capture assay
Direct assay
TRENDS in Molecular Medicine
Figure 2. Protein microarrays and strategies used in antibody-based arrays. (a) Principles of protein microarrays. This platform can be compared to a library of spatially addressable and highly miniaturized recognition ligands immobilized on a chip surface that require very low amounts of source material for analysis. After a washing step, only target proteins will be specifically retained on the chip, detected and analyzed for identity. In protein expression microarrays, different types of probe, such as antibodies, antigens, peptides, DNA or RNA aptamers, carbohydrates or chemicals, with high affinity and known specificity are spotted on a microarray support. The levels of different proteins from a sample such as serum, cell lysates or living cells are assessed in parallel. This approach can be applied usefully to differential display experiments – for example, normal versus tumor samples – and to clinical diagnostics. If proteins from two different samples are labeled with two different fluorochromes (e.g. red and green), the absence (gray spots), the relative excess (red and green spots) and the similar representation (yellow spots) of each protein in one sample relative to the other can be determined. In protein function microarrays, individually purified or synthesized proteins are separately spotted on a suitable solid surface. These protein sets are then probed with enzyme- or fluorescently labeled compounds, such as proteins, nucleic acids, drugs or enzymes, to analyze protein activity, binding properties and posttranslational modifications. With this approach, the identification of enzyme substrates, drug targets and biological networks is made possible on a proteome-wide level. (b) Different strategies in antibody-based protein microarrays. In a sandwich assay, capture antibodies are immobilized on the solid support and bound proteins are detected by using a second, labeled detection antibody. In an antigen-capture assay, proteins are chemically labeled before being applied to the array and are captured by immobilized antibodies. In a direct assay, the complex mixture of proteins is itself immobilized on the solid support, and specific target proteins in that mixture are visualized by labeled detection antibodies.
that allows the rapid visualization of one or a few molecular targets (proteins or nucleic acids) in hundreds or thousands of tissue specimens at a time [18]. Box 1. Tissue microarrays Proteomics can rapidly identify in parallel thousands of proteins that are expressed in a given biological sample. Validation and selection of the best targets from those candidate proteins is a challenging task. Analysis of such molecular targets at the cellular level across a panel of normal and diseased tissue specimens, and evaluation of their clinical significance, would provide significant information for cancer diagnosis, chemotherapy monitoring and novel target selection. Tissue microarray (TMA) is a high-throughput technology developed for the rapid visualization of one or a few molecular targets (i.e. proteins or nucleic acids) in hundreds or thousands of tissue specimens at a time [18]. Arrays are generated by robotically spotting small cylinders (, 0.6 mm3) of tissue derived from individual paraffin-embedded specimens onto a slide. Proteins and nucleic acids are usually identified by immunohistochemistry and fluorescence in situ hybridization, respectively. The TMA technical principle is therefore similar to that of the direct assay version of antibody microarrays (see Figure 2), with the fundamental difference that TMA provides investigators with precious information on target distribution at both the cell level (e.g. cytoplasm versus nucleus) and the tissue level (e.g. stromal versus tumor cells). Most applications of TMA have been in the field of cancer research [72]. http://tmm.trends.com
Proteomics applications Protein identification and quantification In MS-based proteomics, no method or instrument exists that is capable of identifying and quantifying the components of a complex protein sample in a simple, single-step operation. Instead, individual components for separating, identifying and quantifying the polypeptides, as well as tools for integrating and analyzing all of the data, must be used in concert. Laser capture microdissection (Box 2) is now a well-established tool for facilitating the enrichment of cells of interest and for overcoming the issue of tissue heterogeneity. Among the multitude of techniques and instruments, however, two main methods can be identified. In the first method, which is still largely used, the proteins in a sample are separated by 2DE and stained, and each observed protein spot is quantified by its staining intensity. Selected spots are excised, digested and analyzed by MS (Figure 3). This rather low-throughput
Review
28
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
approach has several limitations, such as the difficulty of automating the process, the relatively low resolution (for separating proteins with posttranslational modifications and for detecting proteins that are expressed less quantitatively), and poor performance in the presence of proteins with low water solubility (such as membrane and nuclear proteins). The second method combines limited protein purification, which is usually based on liquid chromatography (e.g. multidimensional separation systems), with the more recently developed techniques of automated peptide tandem MS, thereby improving both sensitivity and throughput [19,20] (Figure 3). A recently introduced technology, called surface-enhanced laser-desorption ionization (SELDI), seems to combine the advantages of the two methods, facilitating sample analysis with minimal manipulation [21] (Figure 3). In both MALDI- and ESI-MS, the relationship between the amount of analyte present and the measured signal intensity is complex and incompletely understood. Mass spectrometers are therefore relatively poor quantitative devices. To add a quantitative dimension to peptide tandem-MS experiments, the isotope-coded affinity tag (‘ICAT’) technique has been developed [22]. This method exploits the fact that pairs of chemically identical analytes
Box 2. Laser capture microdissection One of the main limitations of proteomic in vivo studies is the heterogeneity of the primary specimen (for both normal and tumor tissue). As in genomics studies, all of the current proteomics techniques except for tissue microarray (see Box 1) require tissue homogenization and thus do not take into account the cell source that originally contributed to the protein content. On the one hand, because normal and malignant cells can share more than 98% of the protein profile, aberrantly expressed proteins might be masked by the overwhelming population of shared molecules. On the other hand, in vitro propagated tumor cell lines lack the influences of the tumor microenvironment and their protein expression patterns can widely differ from those of malignant cells in vivo. Laser capture microdissection (LCM) is now a well-established tool for facilitating the enrichment of cells of interest and for overcoming the issue of tissue heterogeneity [73]. LCM enables the selection of cells with a precision of 3 –5 mm under direct microscopic vision control. In brief, after the slides are stained, a specific adherence cap containing an ethylene vinyl acetate film is placed over the tissue. The user then moves the slide and, when the cells of interest are located, an infrared laser is fired that melts the film in the target area. Laser-captured cells are removed from the specimen and analyzed by MS. Applications of LCM in oncoproteomics studies have been already reported [54,74].
2DE
IEF electrophoresis
SDS–PAGE
Protein digestion
2D or 3D LC
Protein digestion
Biological sample (cells, tissues, biological fluids)
Lysis buffer
Protein extraction
Multidimensional LC
MS or tandem MS
Protein separation
Microbeads technology
Washing
Elution
Bead Protein
SELDI Washing SELDI chip TRENDS in Molecular Medicine
Figure 3. Phases of mass spectrometry (MS)-based proteomics experiments. After extraction from a biological sample, the protein mixture is usually fractionated to decrease signal suppression by major proteins (e.g. albumin in serum samples) so that less-abundant proteins can be detected, and to resolve spectral peaks made by multiple protein species with a similar molecular weight. Four main protein separation methods are shown. In 2D electrophoresis (2DE), proteins are separated by isoelectrofocusing (the first dimension) according to their isoelectric points. The gel strips are then equilibrated in a sodium dodecyl sulfate (SDS) buffer, transferred and separated according to their molecular weight (the second dimension) by polyacrylamide gel electrophoresis (PAGE). Although MS can profile mixtures of both whole proteins and peptides, for the purposes of protein identification, whole proteins must be first digested with enzymes (i.e. proteases) because the sensitivity of mass measurement for whole proteins is lower than that for peptides. This approach is typically coupled to MALDI-TOF-MS. In multidimensional liquid chromatography, after protein digestion and protease inactivation the protein mixture is fractionated by 2D (strong cation exchange and reversed phase) or 3D (strong cation exchange, avidin and reversed phase) liquid chromatography. This separation procedure is meant to maximize the probability that every protein is represented by at least one peptide in the subsequent tandemMS analysis. Proteins can also be fractionated on functionalized surfaces according to their physical, chemical or biochemical properties. This is the basis of magnetic beads and surface-enhanced laser-desorption ionization (SELDI) technology. The support surface can be chemically functionalized (e.g. rendered hydrophobic, hydrophilic, anionic or cationic, or coordinated to a metal ion) or biochemically functionalized (e.g. with antibodies, nucleic acids or antigens) in order to retain a functional class of protein that can be analyzed by MS or tandem MS. http://tmm.trends.com
Review
TRENDS in Molecular Medicine
with different stable-isotope compositions can be differentiated in a mass spectrometer owing to their mass difference, and that the ratio of signal intensities for such analyte pairs accurately indicates the abundance ratio for the two analytes. This technique relies, however, on the availability of isotopes in the particular protein to be quantified. In microarray-based proteomics, different strategies, such as MS analysis or western blot assay, can be applied to protein identification after target binding and detection (Figure 2a). For antibody microarrays, immunoglobulin specificity theoretically guarantees target identity; however, antigen cross-reactivity might require further investigation for confirmation. Absolute quantification of proteins can be achieved by means of colorimetric methods, such as in a standard enzyme-linked immunosorbent assay, whereas relative quantification (e.g. normal versus tumor samples) can be obtained by using fluorescence-based probe labeling, such as in two-color cDNA microarrays [23] (Figure 2a). Protein interactions Three main techniques are used to study protein interactions: protein microarrays (see above), biochemical copurification and the yeast two-hybrid system. In co-purification, the protein itself is used as an affinity reagent to isolate its binding partners [16,24,25]. Compared with two-hybrid and microarray-based approaches, this strategy has the advantages that the fully processed and modified protein can serve as the ‘bait’, the interactions take place in the native environment and cellular location, and multicomponent complexes can be isolated and analyzed in a single operation. Ideally, endogenous proteins can act as the ‘bait’ if there is an antibody or other reagent that allows specific isolation of the protein with its bound partners. Unfortunately, there are currently no comprehensive antibody collections and many available antibodies do not immunoprecipitate well or lack sufficient specificity. A more generic strategy is to ‘tag’ the proteins of interest, for instance, with a sequence that is readily recognized by an antibody specific for the tag. This facilitates immunoprecipitation of the target-protein complex and its subsequent MS-based analysis. The yeast two-hybrid assay [26] provides a genomewide genetic approach to the identification and analysis of protein – protein interactions [16]. It relies on a property of several eukaryotic transcription factors: that is, the possession of both a site-specific DNA-binding domain and a transcription activation domain that recruits the transcriptional machinery. In this assay, two types of hybrid protein are generated: so-called ‘bait’ proteins, which are fused with the DNA-binding domain; and ‘prey’ proteins, which are fused with the activation domain. Only when partner proteins interact with each other is the activity of the transcription factor reconstituted and the reporter gene expressed. Protein posttranslational modifications Mass spectrometric methods to determine the type and site of posttranslational modifications on single, purified proteins have been refined over the past two decades. http://tmm.trends.com
Vol.10 No.1 January 2004
29
Recently, attempts have been made to define modifications on a proteome-wide scale [27]. Given the difficulty of identifying all modifications even in a single protein, at present scanning for proteome-wide modifications is not comprehensive. One of the current strategies is essentially an extension of the approach used to analyze protein mixtures. Instead of searching the database for only nonmodified peptides, the database search algorithm is instructed also to match potentially modified peptides. To avoid a ‘combinatorial explosion’ resulting from the need to consider all possible modifications for all peptides in the database, the experiment is usually divided into two stages: the identification of a set of proteins on the basis of non-modified peptides, followed by searching this set of proteins for modified peptides [28]. A more function-oriented strategy focuses on the search for one type of modification on all of the proteins present in a sample. Such techniques are usually based on some form of affinity selection (e.g. antibodies specific to phosphotyrosine) that is specific for the modification of interest (e.g. phosphorylation), which is used to purify the ‘subproteome’ (e.g. the phosphoproteome) bearing this modification [29]. Using this strategy, Paweletz et al. [17] have analyzed longitudinally the state of pro-survival checkpoint proteins at the microscopic transition stage from patient-matched, histologically normal prostate epithelium to prostate intraepithelial neoplasia and subsequently to invasive prostate cancer. Results in oncoproteomics Besides providing new insights into cancer pathogenesis [17,30 – 34], proteomics could have an unprecedented impact on some vital areas of cancer patient management, such as (1) the early detection of disease, by using proteomic patterns of body fluid samples [21,35]; (2) cancer diagnosis and/or prognosis, based on proteomic signatures of tumor samples as a complement to histopathological evaluation [36 – 38]; (3) the development of new diseaseand/or patient-specific therapeutic strategies after the identification of differential display between normal tissue and tumor tissue [39,40]; and (4) a rational modulation of therapy according to changes in protein profiles associated with drug resistance [41 – 43]. Although in its infancy, oncoproteomics has already allowed investigators to make interesting observations that demonstrate the potential of this novel approach in cancer translational research, as briefly summarized below. Cancer diagnosis and prognosis Several studies have already shown that protein fingerprints can reproducibly distinguish between normal and tumor samples [34,38,44]. Moreover, protein profiling facilitates the differentiation between samples (including cytological material [45]) obtained from histologically different types of tumor [36,46,47]. Refining diagnosis at the molecular level would obviously greatly improve the classification of tumors, ultimately ameliorating our ability to predict clinical outcome and to identify individuals at a higher risk of disease recurrence. Although studies linking protein expression patterns with patient prognosis are still in their infancy, some
30
Review
TRENDS in Molecular Medicine
reports support the potential of proteomics in this field [36,37,48,49]. Proteomics technology can also accelerate the diagnostic process. For instance, the use of CD antigen expression for the immunophenotyping of leukemias and lymphomas is currently constrained by limitations of flow cytometry, which restricts the assessment of CD antigens to no more than three per assay. Belov et al. [46] have successfully applied antibody microarray to the simple and rapid determination of tens of CD antigens, and potentially more, in a single analysis. So far the use of biomarkers in the clinical setting has been limited to one or a few serum proteins (e.g. PSA, CA-125, CA-19.9 or CEA). In contrast to this ‘univariate’ approach, proteomics facilitates the simultaneous analysis of multiple biomarkers, thereby increasing the sensitivity and specificity of the assay and increasing the possibility of diagnosis at an earlier stage, even before the onset of symptoms. The potential utility of chip-based proteomics in the early detection of cancer has been demonstrated by Petricoin et al. [50,51], who have identified a serum proteomic pattern that accurately distinguishes individuals with cancer from control subjects. Adopting a similar philosophy, Valerio et al. [52] have applied proteomic analysis of serum samples to the differential diagnosis of tumors from tumor-like diseases, such as pancreatic carcinoma and chronic pancreatitis. In our own study, sera from individuals with melanoma and healthy volunteers have been analyzed by MALDI-MS [48]. Several protein species with a molecular weight of less than 30 000 Da, which are completely absent in control subjects, have been found in the sera of individuals with melanoma. Moreover, the presence and abundance of proteins with molecular weights of 2500– 3500 Da show significant variations according to the clinical stage of disease, indicating the potential of protein profiling not only in early diagnosis but also in disease prognosis. On the basis of these encouraging results, proteomics technology is being employed by some national institutions [53,54] for the discovery of cancer biomarkers and for the identification of high-risk subjects. Antiblastic drug development Several proteomics studies have been carried out to identify novel molecularly targeted antiblastic drugs [3,55]. Hanash et al. [56] have found that the levels of OP-18, a regulator of microtubule dynamics, are significantly higher during leukemic blastic crisis than in normal resting or proliferating lymphocytes. Because OP-18 is necessary for maintaining the transformed phenotype, this protein is being investigated as a potential target for antiblastic drugs [57]. Hu et al. [58] have discovered that heat shock protein 70 (HSP70), a stress-inducible antiapoptotic protein, is overexpressed in ovarian cancer cells, and that its inhibition enhances the effect of chemotherapeutic agents. This finding suggests that HSP70 is involved in tumor chemoresistance; thus, neutralization of this protein could open new avenues for the treatment of cancers that have acquired resistance to drugs acting through the apoptotic pathway. By using this proteomics approach, other authors have reported similar results for different types of cancer [30,59 –61]. http://tmm.trends.com
Vol.10 No.1 January 2004
Acquired drug resistance is one of the main causes of therapeutic failure in individuals with cancer [62]. However, surprisingly few studies have been specifically designed to investigate the mechanisms underlying poor treatment response in vivo, as compared with the number of clinical trials investigating treatment effects. Pharmacoproteomics offers the possibility of dissecting in vivo the phenomenon of drug resistance during chemotherapy at a genome-wide level [42]. The implementation of such biological studies in the design of clinical trials is advocated [62]. Tumor immunology The discovery of tumor-associated antigens (TAAs) has renewed the interest for anticancer vaccines [63]. By studying the differential display between tumor and normal tissues, proteomics can identify potential new targets for immunotherapy [64]. A more specific strategy exploits proteomics technology to detect the targets of TAA-specific antibodies that are frequently found in the serum of individuals with cancer. In this case, antibodycontaining sera can be used to probe 2D blots of protein extracts from an individual’s tumor or from tumor-derived cell lines to directly identify antigenic TAAs. Serological proteomics has been adopted for this purpose in several animal models of tumor [65– 70]. Remarkably, this approach also facilitates the detection of other types of antigen, such as glycoproteins; this aspect is particularly important because some TAAs are tumor-specific simply because of their altered glycosylation state [71]. Concluding remarks Recent technological advances in proteomics could revolutionize the way of doing research in medicine by providing researchers with a formidable high-throughput laboratory tool with which to study protein expression profiles and functions in health and disease. Although in its infancy, oncoproteomics holds the promise of dissecting the biology of cancer, thereby improving cancer management from early detection to the development of diseasespecific drugs and patient-tailored therapeutic strategies. This unique opportunity to analyze the dynamics of the proteome on a genome-wide scale is based on different biochemistry techniques such as MS and protein arrays – the principles of which have been described here, together with the main findings reported in the field of oncology. The integration of proteomic and genomic data sets through powerful bioinformatics will yield a comprehensive database of protein properties that will serve as an invaluable tool for researchers to build and to test scientific hypotheses aimed at dissecting the process of tumorigenesis and, ultimately, at unveiling the Achilles heel of tumors. Acknowledgements Apologies are made to those authors whose work on oncoproteomics has not been cited owing to space limitations.
References 1 Tyers, M. and Mann, M. (2003) From genomics to proteomics. Nature 422, 193 – 197
Review
TRENDS in Molecular Medicine
2 Wilkins, M.R. et al. (1996) Progress with proteome projects: why all proteins expressed by a genome should be identified and how to do it. Biotechnol. Genet. Eng. Rev. 13, 19 – 50 3 Hanash, S. (2003) Disease proteomics. Nature 422, 226– 232 4 Peng, J. and Gygi, S.P. (2001) Proteomics: the move to mixtures. J. Mass Spectrom. 36, 1083 – 1091 5 Boguski, M.S. and McIntosh, M.W. (2003) Biomedical informatics for proteomics. Nature 422, 233 – 237 6 Petricoin, E.F. et al. (2002) Clinical proteomics: translating benchside promise into bedside reality. Nat. Rev. Drug Discov. 1, 683 – 695 7 Simpson, R.J. and Dorow, D.S. (2001) Cancer proteomics: from signaling networks to tumor markers. Trends Biotechnol. 19 (Suppl. 10), S40 – S48 8 Martin, D.B. and Nelson, P.S. (2001) From genomics to proteomics: techniques and applications in cancer research. Trends Cell Biol. 11, S60– S65 9 Pandey, A. and Mann, M. (2000) Proteomics to study genes and genomes. Nature 405, 837 – 846 10 Aebersold, R. and Mann, M. (2003) Mass spectrometry-based proteomics. Nature 422, 198 – 207 11 Lin, D. et al. (2003) Large-scale protein identification using mass spectrometry. Biochim. Biophys. Acta 1646, 1 – 10 12 Smith, R.D. (2002) Trends in mass spectrometry instrumentation for proteomics. Trends Biotechnol. 20 (12), S3 – S7 13 Schweitzer, B. and Kingsmore, S.F. (2002) Measuring proteins on microarrays. Curr. Opin. Biotechnol. 13, 14 – 19 14 Templin, M.F. et al. (2002) Protein microarray technology. Trends Biotechnol. 20, 160– 166 15 MacBeath, G. (2002) Protein microarrays and proteomics. Nat. Genet. 32 (Suppl.), 526 – 532 16 Phizicky, E. et al. (2003) Protein analysis on a proteomic scale. Nature 422, 208 – 215 17 Paweletz, C.P. et al. (2001) Reverse phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front. Oncogene 20, 1981– 1989 18 Kononen, J. et al. (1998) Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat. Med. 4, 844– 847 19 Washburn, M.P. et al. (2001) Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 19, 242 – 247 20 Steen, H. and Pandey, A. (2002) Proteomics goes quantitative: measuring protein abundance. Trends Biotechnol. 20, 361 – 364 21 Issaq, H.J. et al. (2002) The SELDI-TOF MS approach to proteomics: protein profiling and biomarker identification. Biochem. Biophys. Res. Commun. 292, 587 – 592 22 Gygi, S.P. et al. (1999) Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17, 994 – 999 23 Wang, E. et al. (2002) Prospective molecular profiling of melanoma metastases suggests classifiers of immune responsiveness. Cancer Res. 62, 3581 – 3586 24 Ho, Y. et al. (2002) Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415, 180– 183 25 Gavin, A.C. et al. (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141 – 147 26 Fields, S. and Song, O. (1989) A novel genetic system to detect protein – protein interactions. Nature 340, 245 – 246 27 Mann, M. and Jensen, O.N. (2003) Proteomic analysis of posttranslational modifications. Nat. Biotechnol. 21, 255 – 261 28 MacCoss, M.J. et al. (2002) Shotgun identification of protein modifications from protein complexes and lens tissue. Proc. Natl. Acad. Sci. U. S. A. 99, 7900 – 7905 29 Kalume, D.E. et al. (2003) Tackling the phosphoproteome: tools and strategies. Curr. Opin. Chem. Biol. 7, 64 – 69 30 Stierum, R. et al. (2003) Proteome analysis reveals novel proteins associated with proliferation and differentiation of the colorectal cancer cell line Caco-2. Biochim. Biophys. Acta 1650, 73 – 91 31 Pucci-Minafra, I. et al. (2002) Proteomic patterns of cultured breast cancer cells and epithelial mammary cells. Ann. N. Y. Acad. Sci. 963, 122 – 139 32 Jessani, N. et al. (2002) Enzyme activity profiles of the secreted and membrane proteome that depict cancer cell invasiveness. Proc. Natl. Acad. Sci. U. S. A. 99, 10335 – 10340 http://tmm.trends.com
Vol.10 No.1 January 2004
31
33 Wulfkuhle, J.D. et al. (2002) Proteomics of human breast ductal carcinoma in situ. Cancer Res. 62, 6740 – 6749 34 Adam, P.J. et al. (2003) Comprehensive proteomic analysis of breast cancer cell membranes reveals unique proteins with potential roles in clinical cancer. J. Biol. Chem. 278, 6482 – 6489 35 Wulfkuhle, J.D. et al. (2003) Proteomic applications for the early detection of cancer. Nat. Rev. Cancer 3, 267 – 275 36 Yanagisawa, K. et al. (2003) Proteomic patterns of tumour subsets in non-small-cell lung cancer. Lancet 362, 433 – 439 37 Steinert, R. et al. (2002) The role of proteomics in the diagnosis and outcome prediction in colorectal cancer. Technol. Cancer Res. Treat. 1, 297– 304 38 Celis, J.E. et al. (1999) Proteomics and immunohistochemistry define some of the steps involved in the squamous differentiation of the bladder transitional epithelium: a novel strategy for identifying metaplastic lesions. Cancer Res. 59, 3003 – 3009 39 Srinivas, P.R. et al. (2002) Proteomics for cancer biomarker discovery. Clin. Chem. 48, 1160 – 1169 40 Liotta, L.A. et al. (2001) Clinical proteomics: personalized molecular medicine. J. Am. Med. Assoc. 286, 2211 – 2214 41 Verrills, N.M. and Kavallaris, M. (2003) Drug resistance mechanisms in cancer cells: a proteomics perspective. Curr. Opin. Mol. Ther. 5, 258– 265 42 Poland, J. et al. (2002) Study of therapy resistance in cancer cells with functional proteome analysis. Clin. Chem. Lab. Med. 40, 221– 234 43 Hutter, G. and Sinha, P. (2001) Proteomics for studying cancer cells and the development of chemoresistance. Proteomics 1, 1233 – 1248 44 Dwek, M.V. and Alaiya, A.A. (2003) Proteome analysis enables separate clustering of normal breast, benign breast and breast cancer tissues. Br. J. Cancer 89, 305 – 307 45 Fetsch, P.A. et al. (2002) Proteomic evaluation of archival cytologic material using SELDI affinity mass spectrometry: potential for diagnostic applications. Am. J. Clin. Pathol. 118, 870 – 876 46 Belov, L. et al. (2001) Immunophenotyping of leukemias using a cluster of differentiation antibody microarray. Cancer Res. 61, 4483– 4489 47 Conrads, T.P. et al. (2003) Cancer diagnosis using proteomic patterns. Expert Rev. Mol. Diagn. 3, 411 – 420 48 Ferrari, L. et al. (2000) Protein profiles in sera of patients with malignant cutaneous melanoma. Rapid Commun. Mass Spectrom. 14, 1149– 1154 49 Chakravarthy, B. and Pietenpol, J.A. (2003) Combined modality management of breast cancer: development of predictive markers through proteomics. Semin. Oncol. 30 (Suppl. 9), 23 – 36 50 Petricoin, E.F. et al. (2002) Use of proteomic patterns in serum to identify ovarian cancer. Lancet 359, 572 – 577 51 Petricoin, E.F. et al. (2002) Serum proteomic patterns for detection of prostate cancer. J. Natl. Cancer Inst. 94, 1576– 1578 52 Valerio, A. et al. (2001) Serum protein profiles of patients with pancreatic cancer and chronic pancreatitis: searching for a diagnostic protein pattern. Rapid Commun. Mass Spectrom. 15, 2420 – 2425 53 Celis, J.E. and Gromov, P. (2003) Proteomics in translational cancer research: toward an integrated approach. Cancer Cell 3, 9 – 15 54 Verma, M. et al. (2001) Proteomic approaches within the NCI early detection research network for the discovery and identification of cancer biomarkers. Ann. N. Y. Acad. Sci. 945, 103 – 115 55 Krieg, R.C. et al. (2002) Clinical proteomics for cancer biomarker discovery and therapeutic targeting. Technol. Cancer Res. Treat. 1, 263– 272 56 Hanash, S.M. et al. (2002) Identification of novel targets for cancer therapy using expression proteomics. Leukemia 16, 478– 485 57 Iancu, C. et al. (2000) Taxol and anti-stathmin therapy: a synergistic combination that targets the mitotic spindle. Cancer Res. 60, 3537– 3541 58 Hu, W. et al. (2002) Increased expression of heat shock protein 70 in adherent ovarian cancer and mesothelioma following treatment with manumycin, a farnesyl transferase inhibitor. Anticancer Res. 22, 665– 672 59 Jones, M.B. et al. (2002) Proteomic analysis and identification of new biomarkers and therapeutic targets for invasive ovarian cancer. Proteomics 2, 76 – 84 60 Chen, S.T. et al. (2002) Proteomics reveals protein profile changes in doxorubicin – treated MCF-7 human breast cancer cells. Cancer Lett. 181, 95 – 107
Review
32
TRENDS in Molecular Medicine
61 Hondermarck, H. et al. (2001) Proteomics of breast cancer for marker discovery and signal pathway profiling. Proteomics 1, 1216 – 1232 62 Lonning, P.E. (2003) Study of suboptimum treatment response: lessons from breast cancer. Lancet Oncol. 4, 177– 185 63 Mocellin, S. et al. Dissecting tumor responsiveness to immunotherapy: the experience of peptide-based melanoma vaccines. Biochim Biophys Acta Rev Cancer (in press) 64 Le Naour, F. (2001) Contribution of proteomics to tumor immunology. Proteomics 1, 1295– 1302 65 Le Naour, F. et al. (2001) Proteomics-based identification of RS/DJ-1 as a novel circulating tumor antigen in breast cancer. Clin. Cancer Res. 7, 3328 – 3335 66 Brichory, F. et al. (2001) Proteomics-based identification of protein gene product 9.5 as a tumor antigen that induces a humoral immune response in lung cancer. Cancer Res. 61, 7908– 7912 67 Kobayashi, R. (2001) A proteomics approach to find a new breast cancer-specific antigenic marker. Clin. Cancer Res. 7, 3325 – 3327
Vol.10 No.1 January 2004
68 Prasannan, L. et al. (2000) Identification of b-tubulin isoforms as tumor antigens in neuroblastoma. Clin. Cancer Res. 6, 3949– 3956 69 Klade, C.S. (2002) Proteomics approaches towards antigen discovery and vaccine development. Curr. Opin. Mol. Ther. 4, 216 – 223 70 Klade, C.S. et al. (2002) Identification and characterization of 9D7, a novel human protein overexpressed in renal cell carcinoma. Int. J. Cancer 97, 217 – 224 71 Brockhausen, I. (1999) Pathways of O-glycan biosynthesis in cancer cells. Biochim. Biophys. Acta 1473, 67 – 95 72 Kallioniemi, O.P. et al. (2001) Tissue microarray technology for highthroughput molecular profiling of cancer. Hum. Mol. Genet. 10, 657– 662 73 Bonner, R.F. et al. (1997) Laser capture microdissection: molecular analysis of tissue. Science 278, 1481 – 1483 74 Paweletz, C.P. et al. (2001) New technologies for biomarker analysis of prostate cancer progression: Laser capture microdissection and tissue proteomics. Urology 57 (Suppl. 1), 160– 163
News & Features on BioMedNet Start your day with BioMedNet’s own daily science news, features, research update articles and special reports. Every two weeks, enjoy BioMedNet Magazine, which contains free articles from Trends, Current Opinion, Cell and Current Biology. Plus, subscribe to Conference Reporter to get daily reports direct from major life science meetings. http://news.bmn.com Here is what you will find in News & Features: Today’s News Daily news and features for life scientists. Sign up to receive weekly email alerts at http://news.bmn.com/alerts Special Report Special in-depth report on events of current importance in the world of the life sciences. Research Update Brief commentary on the latest hot papers from across the Life Sciences, written by laboratory researchers chosen by the editors of the Trends and Current Opinions journals, and a panel of key experts in their fields. Sign up to receive Research Update email alerts on your chosen subject at http://update.bmn.com/alerts BioMedNet Magazine BioMedNet Magazine offers free articles from Trends, Current Opinion, Cell and BioMedNet News, with a focus on issues of general scientific interest. From the latest book reviews to the most current Special Report, BioMedNet Magazine features Opinions, Forum pieces, Conference Reporter, Historical Perspectives, Science and Society pieces and much more in an easily accessible format. It also provides exciting reviews, news and features, and primary research. BioMedNet Magazine is published every 2 weeks. Sign up to receive weekly email alerts at http://news.bmn.com/alerts Conference Reporter BioMedNet’s expert science journalists cover dozens of sessions at major conferences, providing a quick but comprehensive report of what you might have missed. Far more informative than an ordinary conference overview, Conference Reporter’s easy-to-read summaries are updated daily throughout the meeting. Sign up to receive email alerts at http://news.bmn.com/alerts
http://tmm.trends.com
Review
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
Molecular oncology in the post-genomic era: the challenge of proteomics Simone Mocellin1, Carlo Riccardo Rossi1, Pietro Traldi2, Donato Nitti1 and Mario Lise1 1 2
Department of Oncological and Surgical Sciences, University of Padova, via Giustiniani 2, 35128 Padova, Italy CNR, Institute of Molecular Sciences and Technologies, Corso Stati Uniti, 4– 35127 Padova, Italy
One of the greatest challenges in modern medicine is to dissect the cascade of molecular events that lead to the development and progression of tumors. The genomewide study of proteins embodies the exciting promise of proteomics. Unlike genomics, proteomics aims ambitiously to study not only protein expression profiles but also protein functions, which should provide researchers with a more comprehensive view of the molecular machinery that governs tumor biology. Recent technological advances are making proteomics a user-friendly, high-throughput laboratory tool that is likely to have an unprecedented impact on the pace of discovery in medicine. After completing the sequencing of the human genome, the shift towards studying the proteome seems to be a logical step in the highly challenging task of dissecting the molecular basis of health and disease [1]. The term ‘proteome’ was first coined to describe the set of proteins expressed by a given cell population [2] and should be considered a dynamic entity that continuously changes to shape the myriad of biological phenomena that characterize cell life. The study of the proteome, termed ‘proteomics’, offers a wealth of opportunities, but it also presents enormous technical hurdles as it tries to catalog not only protein expression profiles but also protein functions. During the early years of proteomics and until relatively recently, profiling the expression of proteins primarily relied on the use of 2D electrophoresis (2DE), a rather low-throughput method that was later combined with mass spectrometry (MS) [3,4]. The dynamic nature of protein expression, protein interactions and posttranslational modifications requires that measurements are made as a function of time and cellular state. This type of study requires several data and, thus, the development of both high-throughput technologies for protein investigation and adequate bioinformatics tools for analyzing large data sets [5]. Recent technological advances in MS and array miniaturization are fostering the implementation of high-throughput proteomics in biological research, particularly in the field of oncology. Here, the delineation of altered protein expression, not only at the whole-cell or Corresponding author: Simone Mocellin ([email protected]).
tissue level, but also in subcellular structures, protein complexes and biological fluids, might lead to the development of novel biomarkers for the early detection of disease, the identification of new therapeutic targets and the acceleration of drug development, as well as to the development of more effective strategies to predict and to evaluate therapeutic effect and toxicity [6 – 8]. Although high-throughput proteomics is in its infancy and many advances are needed to overcome its limitations (Table 1), for the first time researchers are working on the ambitious project of studying cell biology by looking at the whole protein repertoire. Here we review the principles and technologies that support proteomics, as well as the main findings that have been so far reported in the field of oncology. Principles and technological platforms Proteomics deals with three fundamental biological issues: protein characterization (including single protein identification and/or quantification and sample protein profiling), protein – protein interactions (how proteins work together to build up metabolic and/or signaling pathways), and posttranslational modifications of proteins (how protein function can be activated, silenced or modulated). Depending on the aim of the research, two main categories of proteomics studies can be identified [9]. First, ‘expression proteomics’ addresses the issue of protein expression levels in a given sample, such as a tumor, normal tissue or body fluid; this approach is particularly useful for protein profiling studies aimed at determining the difference between normal and tumor tissues. The protein profile (or ‘signature’) of cancer should allow investigators to identify tumor markers and, ultimately, disease-specific and/or patient-tailored molecular targets for the development of effective, non-toxic antiblastic drugs. Second, ‘functional proteomics’ includes the analysis of protein – protein interactions, protein – non-protein (e.g. DNA, RNA or lipids) interactions and protein posttranslational modifications, which will help researchers to dissect the intricate network of molecular pathways underlying cellular activities in health and disease. These types of data should greatly improve our understanding of the biological process of tumor development and progression, thus fostering the
http://tmm.trends.com 1471-4914/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.molmed.2003.11.001
Review
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
25
Table 1. Limitations of the two main proteomics technological platforms MS-based proteomics
Protein microarray-based proteomics
Protein identification is limited to known proteins (the mass of proteins under investigation must be compared to those reported in existing protein databases) Protein quantification requires complex procedures to be accurately determined Protein interaction requires sophisticated procedures (e.g. tag-based co-purification)
Antibody-based protein microarrays require thousands (or even hundreds of thousands) of specific antibodies to cover the entire protein Miniaturization can cause problems of steric hindrance among probes (e.g. antibodies) Protein identification by antibodies can be misleading due to antibody cross reaction Protein function of a given biological sample can only be probed with few antibodies at a time
identification of not only single protein-specific therapeutic targets but also pathway-specific targets. From a technical point of view, two main types of technological platform are currently used for highthroughput oncoproteomics research: MS-based proteomics and microarray-based proteomics. Mass spectrometry Several proteomics platforms rely on MS, which is usually associated with other techniques such as electrophoresis or liquid chromatography, for separating the target protein from the sample protein mixture [10,11]. Currently, dedicated MS laboratories can profile thousands of samples and can identify and/or quantify thousands of proteins per day, and rapid advances in sample throughput, sensitivity and accuracy are upcoming [12]. Mass spectrometers are based on a combination of three essential components: (1) the ion source, which produces ions from the sample protein mixture; (2) the mass analyzer, which resolves ionized analytes on the basis of their mass-to-charge (m/z) ratio; and (3) the detector, which detects the ions resolved by the mass analyzer (Figure 1). Two types of MS are used for most proteomics studies: matrix-assisted laser-desorption ionization (MALDI) and electrospray ionization (ESI) MS instruments, which volatize and ionize the proteins or peptides for their further mass analysis (Figure 1). They are very different in design and performance, and each has its own strength and weakness. MALDI is usually coupled to timeof-flight (TOF) analyzers that measure the mass of whole proteins or peptides, whereas ESI has been mainly coupled to different MS instruments, such as ion traps or triple quadrupole instruments, to generate fragment ion (collisioninduced) spectra of selected precursor ions (Figure 1). MALDI-TOF is still widely used to study large proteins by what is known as ‘peptide-mass mapping’. In this method, proteins are identified by matching a list of experimental peptide masses with a calculated list of all of the peptide masses of each entry in, for example, a comprehensive protein database. For an in-depth analysis of individual MS spectral peaks, however, a combination of multiple MS analyzers in a tandem-MS (MS/MS) instrument, such as MALDI-TOF/TOF, ESI-ion trap or ESI-triple quadrupole, is usually adopted to perform a twostage or multistage mass analysis of ions; this approach is ideal for analyzing both biomacromolecules and small molecules because it yields an accurate determination (10 – 100 kDa) of the molecular weight of large proteins and fine structural information directly from the biomolecules (Figure 1). As compared with peptide mass mapping, this http://tmm.trends.com
approach provides information not only about peptide mass but also about peptide sequence, which is scanned against comprehensive protein sequence databases by using one of several different algorithms. Because protein identification relies on matches with sequence databases, however, high-throughput proteomics is currently restricted largely to those species for which comprehensive sequence databases are available. Protein microarray There is considerable benefit to be gained from microarray technology [13,14]. First, in principle thousands of proteins can be spotted on a single slide or similar support, enabling researchers to interrogate simultaneously the function of many different proteins with minimal sample consumption. Second, hundreds or even thousands of copies of an array can be fabricated in parallel, enabling the same proteins to be probed repeatedly with many different molecules under numerous conditions [15]. Protein microarray is finding its way into both expression and functional proteomics [15,16] (Figure 2a). In expression proteomics, protein-detecting microarrays comprise several different affinity reagents, such as antibodies, arrayed at high spatial density on a solid support. Each agent (or ‘probe’) captures its target protein from a complex mixture (e.g. serum or cell lysate), and the captured proteins are subsequently detected and quantified. The three main methods used for antibody microarray are shown in Figure 2b. In contrast to relatively unbiased separation methods such as gel electrophoresis and liquid chromatography, affinity-based approaches enable the investigator to direct the experiment. If the goal of the experiment is to study a particular biological process, only those proteins involved in that process need to be examined. The drawback is that these investigations require both a prior knowledge of the proteins to be studied and appropriate affinity reagents: for instance, the analysis of 30 000–50 000 human genes could require more than 200 000 highly specific antibodies, taking into consideration posttranslational modifications, degradation processes and mutation phenomena. Moreover, steric hindrance, competition and antibody specificity are the primary causes of nonspecific binding or background signals on a multiplexed protein microarray. Protein microarrays also provide a well-controlled, in vitro way in which to study protein function on a genome-wide basis. For this application, sample proteins (or peptides), rather than affinity reagents, are arrayed on a solid support and then probed with compounds of
26
Review
TRENDS in Molecular Medicine
(a)
Protein–peptide liquid mixture
Vol.10 No.1 January 2004
Sampling cone
Nozzle
Spray needle ESI Ions
(b)
Matrix plate
Laser source
Extraction grid
Ions
(c)
Ion source
Mass analyzer
Detector
MALDI
Mass spectra
Intensity
Electrodes
m z –1 Acceleration
TOF Reflector
(d)
TOF-1
Mass selector
TOF-2
Collision chamber TRENDS in Molecular Medicine
Figure 1. Mass spectrometry (MS) technology for proteomics. Two ionization techniques are currently used for biomolecules. (a) Electrospray ionization (ESI) is used to volatilize and to ionize peptides and proteins from liquid samples, and is usually coupled to two mass analyzers (tandem MS or MS/MS) for protein or peptide characterization. (b) For matrix-assisted laser-desorption ionization (MALDI), the analyte is first co-crystallized in a surplus of matrix molecules, which commonly comprise an organic acid on a solid support. A pulsed ultraviolet laser then evaporates the matrix and the analyte into the gas phase. In addition to desorption support, the matrix ionizes the analyte molecules by proton transfer. Single charged ions are mostly generated in this process. (c) The ions are accelerated by an electrical field and then enter the fieldfree mass analyzer, in which they are separated according to their velocity. The ions are turned around in a reflector, which compensates for slight differences in kinetic energy, and then impinge on a detector that amplifies and counts them as they arrive. From the time of flight (TOF) along the mass analyzer, the ratio between the mass and the charge (m z21) of the analyte can be calculated easily: for single charged ions, this value is equivalent to their molecular weight. MALDI-TOF is used for protein profiling purposes and protein identification by the peptide-mass mapping method. Because mass mapping requires an essentially purified target protein, this technique is often used in conjunction with prior protein fractionation using 2D electrophoresis (2DE) (see Figure 3). (d) Scheme of a MALDI-TOF/TOF instrument, shown as an example of tandem MS. The first mass analyzer (TOF-1) acts as a mass filter to separate analytes with the mass of interest from the rest of the ions. For example, the mass of a potential biomarker can be sorted from a mixture of serum proteins. Selected ions are then fragmented in a collision cell. The mass of the resulting fragments is finally analyzed by the second mass analyzer (TOF-2). MALDI-MS/MS enables protein identification, such as the characterization of a tumor marker identified in patient sera by protein profiling, and extends the use of MALDI to the analysis of more complex samples, thereby uncoupling MALDI-MS from 2DE. Other tandem-MS platforms, such as ion-trap MS or quadrupole-TOF MS, rely on similar principles.
interest. With this strategy, new substrates or ligands of known enzymes (or receptors or drugs), as well as diseaserelated posttranslational modifications, can be identified [17]. An advantage of studying protein function in an array format is that the investigator can control the conditions of the experiment. This includes factors such as pH, temperature, ionic strength and the presence or absence http://tmm.trends.com
of cofactors, as well as the modification states of the proteins under investigation. Furthermore, protein microarrays can be used to study the interaction of proteins with non-protein molecules, including nucleic acids, lipids and small organic compounds. An extension of these methodologies is tissue microarray technology (Box 1) – a high-throughput technique
Review
TRENDS in Molecular Medicine
27
Vol.10 No.1 January 2004
(a) Detection system
Sample proteins
Immobilization of recognition molecules
Target-protein binding
Data analysis
(b)
Sandwich assay
Antigen-capture assay
Direct assay
TRENDS in Molecular Medicine
Figure 2. Protein microarrays and strategies used in antibody-based arrays. (a) Principles of protein microarrays. This platform can be compared to a library of spatially addressable and highly miniaturized recognition ligands immobilized on a chip surface that require very low amounts of source material for analysis. After a washing step, only target proteins will be specifically retained on the chip, detected and analyzed for identity. In protein expression microarrays, different types of probe, such as antibodies, antigens, peptides, DNA or RNA aptamers, carbohydrates or chemicals, with high affinity and known specificity are spotted on a microarray support. The levels of different proteins from a sample such as serum, cell lysates or living cells are assessed in parallel. This approach can be applied usefully to differential display experiments – for example, normal versus tumor samples – and to clinical diagnostics. If proteins from two different samples are labeled with two different fluorochromes (e.g. red and green), the absence (gray spots), the relative excess (red and green spots) and the similar representation (yellow spots) of each protein in one sample relative to the other can be determined. In protein function microarrays, individually purified or synthesized proteins are separately spotted on a suitable solid surface. These protein sets are then probed with enzyme- or fluorescently labeled compounds, such as proteins, nucleic acids, drugs or enzymes, to analyze protein activity, binding properties and posttranslational modifications. With this approach, the identification of enzyme substrates, drug targets and biological networks is made possible on a proteome-wide level. (b) Different strategies in antibody-based protein microarrays. In a sandwich assay, capture antibodies are immobilized on the solid support and bound proteins are detected by using a second, labeled detection antibody. In an antigen-capture assay, proteins are chemically labeled before being applied to the array and are captured by immobilized antibodies. In a direct assay, the complex mixture of proteins is itself immobilized on the solid support, and specific target proteins in that mixture are visualized by labeled detection antibodies.
that allows the rapid visualization of one or a few molecular targets (proteins or nucleic acids) in hundreds or thousands of tissue specimens at a time [18]. Box 1. Tissue microarrays Proteomics can rapidly identify in parallel thousands of proteins that are expressed in a given biological sample. Validation and selection of the best targets from those candidate proteins is a challenging task. Analysis of such molecular targets at the cellular level across a panel of normal and diseased tissue specimens, and evaluation of their clinical significance, would provide significant information for cancer diagnosis, chemotherapy monitoring and novel target selection. Tissue microarray (TMA) is a high-throughput technology developed for the rapid visualization of one or a few molecular targets (i.e. proteins or nucleic acids) in hundreds or thousands of tissue specimens at a time [18]. Arrays are generated by robotically spotting small cylinders (, 0.6 mm3) of tissue derived from individual paraffin-embedded specimens onto a slide. Proteins and nucleic acids are usually identified by immunohistochemistry and fluorescence in situ hybridization, respectively. The TMA technical principle is therefore similar to that of the direct assay version of antibody microarrays (see Figure 2), with the fundamental difference that TMA provides investigators with precious information on target distribution at both the cell level (e.g. cytoplasm versus nucleus) and the tissue level (e.g. stromal versus tumor cells). Most applications of TMA have been in the field of cancer research [72]. http://tmm.trends.com
Proteomics applications Protein identification and quantification In MS-based proteomics, no method or instrument exists that is capable of identifying and quantifying the components of a complex protein sample in a simple, single-step operation. Instead, individual components for separating, identifying and quantifying the polypeptides, as well as tools for integrating and analyzing all of the data, must be used in concert. Laser capture microdissection (Box 2) is now a well-established tool for facilitating the enrichment of cells of interest and for overcoming the issue of tissue heterogeneity. Among the multitude of techniques and instruments, however, two main methods can be identified. In the first method, which is still largely used, the proteins in a sample are separated by 2DE and stained, and each observed protein spot is quantified by its staining intensity. Selected spots are excised, digested and analyzed by MS (Figure 3). This rather low-throughput
Review
28
TRENDS in Molecular Medicine
Vol.10 No.1 January 2004
approach has several limitations, such as the difficulty of automating the process, the relatively low resolution (for separating proteins with posttranslational modifications and for detecting proteins that are expressed less quantitatively), and poor performance in the presence of proteins with low water solubility (such as membrane and nuclear proteins). The second method combines limited protein purification, which is usually based on liquid chromatography (e.g. multidimensional separation systems), with the more recently developed techniques of automated peptide tandem MS, thereby improving both sensitivity and throughput [19,20] (Figure 3). A recently introduced technology, called surface-enhanced laser-desorption ionization (SELDI), seems to combine the advantages of the two methods, facilitating sample analysis with minimal manipulation [21] (Figure 3). In both MALDI- and ESI-MS, the relationship between the amount of analyte present and the measured signal intensity is complex and incompletely understood. Mass spectrometers are therefore relatively poor quantitative devices. To add a quantitative dimension to peptide tandem-MS experiments, the isotope-coded affinity tag (‘ICAT’) technique has been developed [22]. This method exploits the fact that pairs of chemically identical analytes
Box 2. Laser capture microdissection One of the main limitations of proteomic in vivo studies is the heterogeneity of the primary specimen (for both normal and tumor tissue). As in genomics studies, all of the current proteomics techniques except for tissue microarray (see Box 1) require tissue homogenization and thus do not take into account the cell source that originally contributed to the protein content. On the one hand, because normal and malignant cells can share more than 98% of the protein profile, aberrantly expressed proteins might be masked by the overwhelming population of shared molecules. On the other hand, in vitro propagated tumor cell lines lack the influences of the tumor microenvironment and their protein expression patterns can widely differ from those of malignant cells in vivo. Laser capture microdissection (LCM) is now a well-established tool for facilitating the enrichment of cells of interest and for overcoming the issue of tissue heterogeneity [73]. LCM enables the selection of cells with a precision of 3 –5 mm under direct microscopic vision control. In brief, after the slides are stained, a specific adherence cap containing an ethylene vinyl acetate film is placed over the tissue. The user then moves the slide and, when the cells of interest are located, an infrared laser is fired that melts the film in the target area. Laser-captured cells are removed from the specimen and analyzed by MS. Applications of LCM in oncoproteomics studies have been already reported [54,74].
2DE
IEF electrophoresis
SDS–PAGE
Protein digestion
2D or 3D LC
Protein digestion
Biological sample (cells, tissues, biological fluids)
Lysis buffer
Protein extraction
Multidimensional LC
MS or tandem MS
Protein separation
Microbeads technology
Washing
Elution
Bead Protein
SELDI Washing SELDI chip TRENDS in Molecular Medicine
Figure 3. Phases of mass spectrometry (MS)-based proteomics experiments. After extraction from a biological sample, the protein mixture is usually fractionated to decrease signal suppression by major proteins (e.g. albumin in serum samples) so that less-abundant proteins can be detected, and to resolve spectral peaks made by multiple protein species with a similar molecular weight. Four main protein separation methods are shown. In 2D electrophoresis (2DE), proteins are separated by isoelectrofocusing (the first dimension) according to their isoelectric points. The gel strips are then equilibrated in a sodium dodecyl sulfate (SDS) buffer, transferred and separated according to their molecular weight (the second dimension) by polyacrylamide gel electrophoresis (PAGE). Although MS can profile mixtures of both whole proteins and peptides, for the purposes of protein identification, whole proteins must be first digested with enzymes (i.e. proteases) because the sensitivity of mass measurement for whole proteins is lower than that for peptides. This approach is typically coupled to MALDI-TOF-MS. In multidimensional liquid chromatography, after protein digestion and protease inactivation the protein mixture is fractionated by 2D (strong cation exchange and reversed phase) or 3D (strong cation exchange, avidin and reversed phase) liquid chromatography. This separation procedure is meant to maximize the probability that every protein is represented by at least one peptide in the subsequent tandemMS analysis. Proteins can also be fractionated on functionalized surfaces according to their physical, chemical or biochemical properties. This is the basis of magnetic beads and surface-enhanced laser-desorption ionization (SELDI) technology. The support surface can be chemically functionalized (e.g. rendered hydrophobic, hydrophilic, anionic or cationic, or coordinated to a metal ion) or biochemically functionalized (e.g. with antibodies, nucleic acids or antigens) in order to retain a functional class of protein that can be analyzed by MS or tandem MS. http://tmm.trends.com
Review
TRENDS in Molecular Medicine
with different stable-isotope compositions can be differentiated in a mass spectrometer owing to their mass difference, and that the ratio of signal intensities for such analyte pairs accurately indicates the abundance ratio for the two analytes. This technique relies, however, on the availability of isotopes in the particular protein to be quantified. In microarray-based proteomics, different strategies, such as MS analysis or western blot assay, can be applied to protein identification after target binding and detection (Figure 2a). For antibody microarrays, immunoglobulin specificity theoretically guarantees target identity; however, antigen cross-reactivity might require further investigation for confirmation. Absolute quantification of proteins can be achieved by means of colorimetric methods, such as in a standard enzyme-linked immunosorbent assay, whereas relative quantification (e.g. normal versus tumor samples) can be obtained by using fluorescence-based probe labeling, such as in two-color cDNA microarrays [23] (Figure 2a). Protein interactions Three main techniques are used to study protein interactions: protein microarrays (see above), biochemical copurification and the yeast two-hybrid system. In co-purification, the protein itself is used as an affinity reagent to isolate its binding partners [16,24,25]. Compared with two-hybrid and microarray-based approaches, this strategy has the advantages that the fully processed and modified protein can serve as the ‘bait’, the interactions take place in the native environment and cellular location, and multicomponent complexes can be isolated and analyzed in a single operation. Ideally, endogenous proteins can act as the ‘bait’ if there is an antibody or other reagent that allows specific isolation of the protein with its bound partners. Unfortunately, there are currently no comprehensive antibody collections and many available antibodies do not immunoprecipitate well or lack sufficient specificity. A more generic strategy is to ‘tag’ the proteins of interest, for instance, with a sequence that is readily recognized by an antibody specific for the tag. This facilitates immunoprecipitation of the target-protein complex and its subsequent MS-based analysis. The yeast two-hybrid assay [26] provides a genomewide genetic approach to the identification and analysis of protein – protein interactions [16]. It relies on a property of several eukaryotic transcription factors: that is, the possession of both a site-specific DNA-binding domain and a transcription activation domain that recruits the transcriptional machinery. In this assay, two types of hybrid protein are generated: so-called ‘bait’ proteins, which are fused with the DNA-binding domain; and ‘prey’ proteins, which are fused with the activation domain. Only when partner proteins interact with each other is the activity of the transcription factor reconstituted and the reporter gene expressed. Protein posttranslational modifications Mass spectrometric methods to determine the type and site of posttranslational modifications on single, purified proteins have been refined over the past two decades. http://tmm.trends.com
Vol.10 No.1 January 2004
29
Recently, attempts have been made to define modifications on a proteome-wide scale [27]. Given the difficulty of identifying all modifications even in a single protein, at present scanning for proteome-wide modifications is not comprehensive. One of the current strategies is essentially an extension of the approach used to analyze protein mixtures. Instead of searching the database for only nonmodified peptides, the database search algorithm is instructed also to match potentially modified peptides. To avoid a ‘combinatorial explosion’ resulting from the need to consider all possible modifications for all peptides in the database, the experiment is usually divided into two stages: the identification of a set of proteins on the basis of non-modified peptides, followed by searching this set of proteins for modified peptides [28]. A more function-oriented strategy focuses on the search for one type of modification on all of the proteins present in a sample. Such techniques are usually based on some form of affinity selection (e.g. antibodies specific to phosphotyrosine) that is specific for the modification of interest (e.g. phosphorylation), which is used to purify the ‘subproteome’ (e.g. the phosphoproteome) bearing this modification [29]. Using this strategy, Paweletz et al. [17] have analyzed longitudinally the state of pro-survival checkpoint proteins at the microscopic transition stage from patient-matched, histologically normal prostate epithelium to prostate intraepithelial neoplasia and subsequently to invasive prostate cancer. Results in oncoproteomics Besides providing new insights into cancer pathogenesis [17,30 – 34], proteomics could have an unprecedented impact on some vital areas of cancer patient management, such as (1) the early detection of disease, by using proteomic patterns of body fluid samples [21,35]; (2) cancer diagnosis and/or prognosis, based on proteomic signatures of tumor samples as a complement to histopathological evaluation [36 – 38]; (3) the development of new diseaseand/or patient-specific therapeutic strategies after the identification of differential display between normal tissue and tumor tissue [39,40]; and (4) a rational modulation of therapy according to changes in protein profiles associated with drug resistance [41 – 43]. Although in its infancy, oncoproteomics has already allowed investigators to make interesting observations that demonstrate the potential of this novel approach in cancer translational research, as briefly summarized below. Cancer diagnosis and prognosis Several studies have already shown that protein fingerprints can reproducibly distinguish between normal and tumor samples [34,38,44]. Moreover, protein profiling facilitates the differentiation between samples (including cytological material [45]) obtained from histologically different types of tumor [36,46,47]. Refining diagnosis at the molecular level would obviously greatly improve the classification of tumors, ultimately ameliorating our ability to predict clinical outcome and to identify individuals at a higher risk of disease recurrence. Although studies linking protein expression patterns with patient prognosis are still in their infancy, some
30
Review
TRENDS in Molecular Medicine
reports support the potential of proteomics in this field [36,37,48,49]. Proteomics technology can also accelerate the diagnostic process. For instance, the use of CD antigen expression for the immunophenotyping of leukemias and lymphomas is currently constrained by limitations of flow cytometry, which restricts the assessment of CD antigens to no more than three per assay. Belov et al. [46] have successfully applied antibody microarray to the simple and rapid determination of tens of CD antigens, and potentially more, in a single analysis. So far the use of biomarkers in the clinical setting has been limited to one or a few serum proteins (e.g. PSA, CA-125, CA-19.9 or CEA). In contrast to this ‘univariate’ approach, proteomics facilitates the simultaneous analysis of multiple biomarkers, thereby increasing the sensitivity and specificity of the assay and increasing the possibility of diagnosis at an earlier stage, even before the onset of symptoms. The potential utility of chip-based proteomics in the early detection of cancer has been demonstrated by Petricoin et al. [50,51], who have identified a serum proteomic pattern that accurately distinguishes individuals with cancer from control subjects. Adopting a similar philosophy, Valerio et al. [52] have applied proteomic analysis of serum samples to the differential diagnosis of tumors from tumor-like diseases, such as pancreatic carcinoma and chronic pancreatitis. In our own study, sera from individuals with melanoma and healthy volunteers have been analyzed by MALDI-MS [48]. Several protein species with a molecular weight of less than 30 000 Da, which are completely absent in control subjects, have been found in the sera of individuals with melanoma. Moreover, the presence and abundance of proteins with molecular weights of 2500– 3500 Da show significant variations according to the clinical stage of disease, indicating the potential of protein profiling not only in early diagnosis but also in disease prognosis. On the basis of these encouraging results, proteomics technology is being employed by some national institutions [53,54] for the discovery of cancer biomarkers and for the identification of high-risk subjects. Antiblastic drug development Several proteomics studies have been carried out to identify novel molecularly targeted antiblastic drugs [3,55]. Hanash et al. [56] have found that the levels of OP-18, a regulator of microtubule dynamics, are significantly higher during leukemic blastic crisis than in normal resting or proliferating lymphocytes. Because OP-18 is necessary for maintaining the transformed phenotype, this protein is being investigated as a potential target for antiblastic drugs [57]. Hu et al. [58] have discovered that heat shock protein 70 (HSP70), a stress-inducible antiapoptotic protein, is overexpressed in ovarian cancer cells, and that its inhibition enhances the effect of chemotherapeutic agents. This finding suggests that HSP70 is involved in tumor chemoresistance; thus, neutralization of this protein could open new avenues for the treatment of cancers that have acquired resistance to drugs acting through the apoptotic pathway. By using this proteomics approach, other authors have reported similar results for different types of cancer [30,59 –61]. http://tmm.trends.com
Vol.10 No.1 January 2004
Acquired drug resistance is one of the main causes of therapeutic failure in individuals with cancer [62]. However, surprisingly few studies have been specifically designed to investigate the mechanisms underlying poor treatment response in vivo, as compared with the number of clinical trials investigating treatment effects. Pharmacoproteomics offers the possibility of dissecting in vivo the phenomenon of drug resistance during chemotherapy at a genome-wide level [42]. The implementation of such biological studies in the design of clinical trials is advocated [62]. Tumor immunology The discovery of tumor-associated antigens (TAAs) has renewed the interest for anticancer vaccines [63]. By studying the differential display between tumor and normal tissues, proteomics can identify potential new targets for immunotherapy [64]. A more specific strategy exploits proteomics technology to detect the targets of TAA-specific antibodies that are frequently found in the serum of individuals with cancer. In this case, antibodycontaining sera can be used to probe 2D blots of protein extracts from an individual’s tumor or from tumor-derived cell lines to directly identify antigenic TAAs. Serological proteomics has been adopted for this purpose in several animal models of tumor [65– 70]. Remarkably, this approach also facilitates the detection of other types of antigen, such as glycoproteins; this aspect is particularly important because some TAAs are tumor-specific simply because of their altered glycosylation state [71]. Concluding remarks Recent technological advances in proteomics could revolutionize the way of doing research in medicine by providing researchers with a formidable high-throughput laboratory tool with which to study protein expression profiles and functions in health and disease. Although in its infancy, oncoproteomics holds the promise of dissecting the biology of cancer, thereby improving cancer management from early detection to the development of diseasespecific drugs and patient-tailored therapeutic strategies. This unique opportunity to analyze the dynamics of the proteome on a genome-wide scale is based on different biochemistry techniques such as MS and protein arrays – the principles of which have been described here, together with the main findings reported in the field of oncology. The integration of proteomic and genomic data sets through powerful bioinformatics will yield a comprehensive database of protein properties that will serve as an invaluable tool for researchers to build and to test scientific hypotheses aimed at dissecting the process of tumorigenesis and, ultimately, at unveiling the Achilles heel of tumors. Acknowledgements Apologies are made to those authors whose work on oncoproteomics has not been cited owing to space limitations.
References 1 Tyers, M. and Mann, M. (2003) From genomics to proteomics. Nature 422, 193 – 197
Review
TRENDS in Molecular Medicine
2 Wilkins, M.R. et al. (1996) Progress with proteome projects: why all proteins expressed by a genome should be identified and how to do it. Biotechnol. Genet. Eng. Rev. 13, 19 – 50 3 Hanash, S. (2003) Disease proteomics. Nature 422, 226– 232 4 Peng, J. and Gygi, S.P. (2001) Proteomics: the move to mixtures. J. Mass Spectrom. 36, 1083 – 1091 5 Boguski, M.S. and McIntosh, M.W. (2003) Biomedical informatics for proteomics. Nature 422, 233 – 237 6 Petricoin, E.F. et al. (2002) Clinical proteomics: translating benchside promise into bedside reality. Nat. Rev. Drug Discov. 1, 683 – 695 7 Simpson, R.J. and Dorow, D.S. (2001) Cancer proteomics: from signaling networks to tumor markers. Trends Biotechnol. 19 (Suppl. 10), S40 – S48 8 Martin, D.B. and Nelson, P.S. (2001) From genomics to proteomics: techniques and applications in cancer research. Trends Cell Biol. 11, S60– S65 9 Pandey, A. and Mann, M. (2000) Proteomics to study genes and genomes. Nature 405, 837 – 846 10 Aebersold, R. and Mann, M. (2003) Mass spectrometry-based proteomics. Nature 422, 198 – 207 11 Lin, D. et al. (2003) Large-scale protein identification using mass spectrometry. Biochim. Biophys. Acta 1646, 1 – 10 12 Smith, R.D. (2002) Trends in mass spectrometry instrumentation for proteomics. Trends Biotechnol. 20 (12), S3 – S7 13 Schweitzer, B. and Kingsmore, S.F. (2002) Measuring proteins on microarrays. Curr. Opin. Biotechnol. 13, 14 – 19 14 Templin, M.F. et al. (2002) Protein microarray technology. Trends Biotechnol. 20, 160– 166 15 MacBeath, G. (2002) Protein microarrays and proteomics. Nat. Genet. 32 (Suppl.), 526 – 532 16 Phizicky, E. et al. (2003) Protein analysis on a proteomic scale. Nature 422, 208 – 215 17 Paweletz, C.P. et al. (2001) Reverse phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front. Oncogene 20, 1981– 1989 18 Kononen, J. et al. (1998) Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat. Med. 4, 844– 847 19 Washburn, M.P. et al. (2001) Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 19, 242 – 247 20 Steen, H. and Pandey, A. (2002) Proteomics goes quantitative: measuring protein abundance. Trends Biotechnol. 20, 361 – 364 21 Issaq, H.J. et al. (2002) The SELDI-TOF MS approach to proteomics: protein profiling and biomarker identification. Biochem. Biophys. Res. Commun. 292, 587 – 592 22 Gygi, S.P. et al. (1999) Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17, 994 – 999 23 Wang, E. et al. (2002) Prospective molecular profiling of melanoma metastases suggests classifiers of immune responsiveness. Cancer Res. 62, 3581 – 3586 24 Ho, Y. et al. (2002) Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415, 180– 183 25 Gavin, A.C. et al. (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141 – 147 26 Fields, S. and Song, O. (1989) A novel genetic system to detect protein – protein interactions. Nature 340, 245 – 246 27 Mann, M. and Jensen, O.N. (2003) Proteomic analysis of posttranslational modifications. Nat. Biotechnol. 21, 255 – 261 28 MacCoss, M.J. et al. (2002) Shotgun identification of protein modifications from protein complexes and lens tissue. Proc. Natl. Acad. Sci. U. S. A. 99, 7900 – 7905 29 Kalume, D.E. et al. (2003) Tackling the phosphoproteome: tools and strategies. Curr. Opin. Chem. Biol. 7, 64 – 69 30 Stierum, R. et al. (2003) Proteome analysis reveals novel proteins associated with proliferation and differentiation of the colorectal cancer cell line Caco-2. Biochim. Biophys. Acta 1650, 73 – 91 31 Pucci-Minafra, I. et al. (2002) Proteomic patterns of cultured breast cancer cells and epithelial mammary cells. Ann. N. Y. Acad. Sci. 963, 122 – 139 32 Jessani, N. et al. (2002) Enzyme activity profiles of the secreted and membrane proteome that depict cancer cell invasiveness. Proc. Natl. Acad. Sci. U. S. A. 99, 10335 – 10340 http://tmm.trends.com
Vol.10 No.1 January 2004
31
33 Wulfkuhle, J.D. et al. (2002) Proteomics of human breast ductal carcinoma in situ. Cancer Res. 62, 6740 – 6749 34 Adam, P.J. et al. (2003) Comprehensive proteomic analysis of breast cancer cell membranes reveals unique proteins with potential roles in clinical cancer. J. Biol. Chem. 278, 6482 – 6489 35 Wulfkuhle, J.D. et al. (2003) Proteomic applications for the early detection of cancer. Nat. Rev. Cancer 3, 267 – 275 36 Yanagisawa, K. et al. (2003) Proteomic patterns of tumour subsets in non-small-cell lung cancer. Lancet 362, 433 – 439 37 Steinert, R. et al. (2002) The role of proteomics in the diagnosis and outcome prediction in colorectal cancer. Technol. Cancer Res. Treat. 1, 297– 304 38 Celis, J.E. et al. (1999) Proteomics and immunohistochemistry define some of the steps involved in the squamous differentiation of the bladder transitional epithelium: a novel strategy for identifying metaplastic lesions. Cancer Res. 59, 3003 – 3009 39 Srinivas, P.R. et al. (2002) Proteomics for cancer biomarker discovery. Clin. Chem. 48, 1160 – 1169 40 Liotta, L.A. et al. (2001) Clinical proteomics: personalized molecular medicine. J. Am. Med. Assoc. 286, 2211 – 2214 41 Verrills, N.M. and Kavallaris, M. (2003) Drug resistance mechanisms in cancer cells: a proteomics perspective. Curr. Opin. Mol. Ther. 5, 258– 265 42 Poland, J. et al. (2002) Study of therapy resistance in cancer cells with functional proteome analysis. Clin. Chem. Lab. Med. 40, 221– 234 43 Hutter, G. and Sinha, P. (2001) Proteomics for studying cancer cells and the development of chemoresistance. Proteomics 1, 1233 – 1248 44 Dwek, M.V. and Alaiya, A.A. (2003) Proteome analysis enables separate clustering of normal breast, benign breast and breast cancer tissues. Br. J. Cancer 89, 305 – 307 45 Fetsch, P.A. et al. (2002) Proteomic evaluation of archival cytologic material using SELDI affinity mass spectrometry: potential for diagnostic applications. Am. J. Clin. Pathol. 118, 870 – 876 46 Belov, L. et al. (2001) Immunophenotyping of leukemias using a cluster of differentiation antibody microarray. Cancer Res. 61, 4483– 4489 47 Conrads, T.P. et al. (2003) Cancer diagnosis using proteomic patterns. Expert Rev. Mol. Diagn. 3, 411 – 420 48 Ferrari, L. et al. (2000) Protein profiles in sera of patients with malignant cutaneous melanoma. Rapid Commun. Mass Spectrom. 14, 1149– 1154 49 Chakravarthy, B. and Pietenpol, J.A. (2003) Combined modality management of breast cancer: development of predictive markers through proteomics. Semin. Oncol. 30 (Suppl. 9), 23 – 36 50 Petricoin, E.F. et al. (2002) Use of proteomic patterns in serum to identify ovarian cancer. Lancet 359, 572 – 577 51 Petricoin, E.F. et al. (2002) Serum proteomic patterns for detection of prostate cancer. J. Natl. Cancer Inst. 94, 1576– 1578 52 Valerio, A. et al. (2001) Serum protein profiles of patients with pancreatic cancer and chronic pancreatitis: searching for a diagnostic protein pattern. Rapid Commun. Mass Spectrom. 15, 2420 – 2425 53 Celis, J.E. and Gromov, P. (2003) Proteomics in translational cancer research: toward an integrated approach. Cancer Cell 3, 9 – 15 54 Verma, M. et al. (2001) Proteomic approaches within the NCI early detection research network for the discovery and identification of cancer biomarkers. Ann. N. Y. Acad. Sci. 945, 103 – 115 55 Krieg, R.C. et al. (2002) Clinical proteomics for cancer biomarker discovery and therapeutic targeting. Technol. Cancer Res. Treat. 1, 263– 272 56 Hanash, S.M. et al. (2002) Identification of novel targets for cancer therapy using expression proteomics. Leukemia 16, 478– 485 57 Iancu, C. et al. (2000) Taxol and anti-stathmin therapy: a synergistic combination that targets the mitotic spindle. Cancer Res. 60, 3537– 3541 58 Hu, W. et al. (2002) Increased expression of heat shock protein 70 in adherent ovarian cancer and mesothelioma following treatment with manumycin, a farnesyl transferase inhibitor. Anticancer Res. 22, 665– 672 59 Jones, M.B. et al. (2002) Proteomic analysis and identification of new biomarkers and therapeutic targets for invasive ovarian cancer. Proteomics 2, 76 – 84 60 Chen, S.T. et al. (2002) Proteomics reveals protein profile changes in doxorubicin – treated MCF-7 human breast cancer cells. Cancer Lett. 181, 95 – 107
Review
32
TRENDS in Molecular Medicine
61 Hondermarck, H. et al. (2001) Proteomics of breast cancer for marker discovery and signal pathway profiling. Proteomics 1, 1216 – 1232 62 Lonning, P.E. (2003) Study of suboptimum treatment response: lessons from breast cancer. Lancet Oncol. 4, 177– 185 63 Mocellin, S. et al. Dissecting tumor responsiveness to immunotherapy: the experience of peptide-based melanoma vaccines. Biochim Biophys Acta Rev Cancer (in press) 64 Le Naour, F. (2001) Contribution of proteomics to tumor immunology. Proteomics 1, 1295– 1302 65 Le Naour, F. et al. (2001) Proteomics-based identification of RS/DJ-1 as a novel circulating tumor antigen in breast cancer. Clin. Cancer Res. 7, 3328 – 3335 66 Brichory, F. et al. (2001) Proteomics-based identification of protein gene product 9.5 as a tumor antigen that induces a humoral immune response in lung cancer. Cancer Res. 61, 7908– 7912 67 Kobayashi, R. (2001) A proteomics approach to find a new breast cancer-specific antigenic marker. Clin. Cancer Res. 7, 3325 – 3327
Vol.10 No.1 January 2004
68 Prasannan, L. et al. (2000) Identification of b-tubulin isoforms as tumor antigens in neuroblastoma. Clin. Cancer Res. 6, 3949– 3956 69 Klade, C.S. (2002) Proteomics approaches towards antigen discovery and vaccine development. Curr. Opin. Mol. Ther. 4, 216 – 223 70 Klade, C.S. et al. (2002) Identification and characterization of 9D7, a novel human protein overexpressed in renal cell carcinoma. Int. J. Cancer 97, 217 – 224 71 Brockhausen, I. (1999) Pathways of O-glycan biosynthesis in cancer cells. Biochim. Biophys. Acta 1473, 67 – 95 72 Kallioniemi, O.P. et al. (2001) Tissue microarray technology for highthroughput molecular profiling of cancer. Hum. Mol. Genet. 10, 657– 662 73 Bonner, R.F. et al. (1997) Laser capture microdissection: molecular analysis of tissue. Science 278, 1481 – 1483 74 Paweletz, C.P. et al. (2001) New technologies for biomarker analysis of prostate cancer progression: Laser capture microdissection and tissue proteomics. Urology 57 (Suppl. 1), 160– 163
News & Features on BioMedNet Start your day with BioMedNet’s own daily science news, features, research update articles and special reports. Every two weeks, enjoy BioMedNet Magazine, which contains free articles from Trends, Current Opinion, Cell and Current Biology. Plus, subscribe to Conference Reporter to get daily reports direct from major life science meetings. http://news.bmn.com Here is what you will find in News & Features: Today’s News Daily news and features for life scientists. Sign up to receive weekly email alerts at http://news.bmn.com/alerts Special Report Special in-depth report on events of current importance in the world of the life sciences. Research Update Brief commentary on the latest hot papers from across the Life Sciences, written by laboratory researchers chosen by the editors of the Trends and Current Opinions journals, and a panel of key experts in their fields. Sign up to receive Research Update email alerts on your chosen subject at http://update.bmn.com/alerts BioMedNet Magazine BioMedNet Magazine offers free articles from Trends, Current Opinion, Cell and BioMedNet News, with a focus on issues of general scientific interest. From the latest book reviews to the most current Special Report, BioMedNet Magazine features Opinions, Forum pieces, Conference Reporter, Historical Perspectives, Science and Society pieces and much more in an easily accessible format. It also provides exciting reviews, news and features, and primary research. BioMedNet Magazine is published every 2 weeks. Sign up to receive weekly email alerts at http://news.bmn.com/alerts Conference Reporter BioMedNet’s expert science journalists cover dozens of sessions at major conferences, providing a quick but comprehensive report of what you might have missed. Far more informative than an ordinary conference overview, Conference Reporter’s easy-to-read summaries are updated daily throughout the meeting. Sign up to receive email alerts at http://news.bmn.com/alerts
http://tmm.trends.com