Marc R. Wilkins and Denis F. Hochstrasser University of Geneva and Central Clinical Chemistry Laboratory, Geneva University Hospital
Introduction
In the last twenty years the field of biochemistry has changed beyond recognition. This is mostly due to the molecular biology revolution, which has provided a great deal of information about biological systems at a nucleic acids level. In fact, there are now over 550,000,000 bases from 827,174 sequence entries in DNA databases (EMBL release 47, June 1996), and the amount of sequence information in these databases doubles almost every twelve months.
However, this wealth of information has revealed comparatively little about how the proteins inside any organism operate individually or synergistically to fulfill biological functions. So, while we know the complete genome sequences of the organisms Haemophilus influenzae, Mycoplasma genitalium, Methanococcus jannaschi, and even the eukaryote Saccharomyces cerevisiaeand will soon know the complete sequence of Escherichia coliwe still do not understand how the simplest living organisms actually work. For example, we do not understand the functions and interrelationships of the 470 proteins of M. genitalium, and more than 30% of S. cerevisiae proteins have no known function.
In some ways this paints a dark picture of biochemistry. If we cannot yet understand how the proteins in life's simplest organisms operate, how can we ever hope to understand the workings of the 60,000 to 70,000 polypeptides thought to be in the human body, which after post-translational modifications may total more than 700,000? The answer to this question is simply that more effort must be placed into the investigation of proteins, or on a larger scale, proteomes.
From Genomes to Proteomes
The primary purpose of this review is to introduce the resources that are available for proteome projects, but first some discussion is warranted about the term "proteome" itself. The "proteome" is simply the PROTEin complement expressed by a genOME or tissue (1). The concept of the proteome is
fundamentally different to that of the genome: while the genome is virtually static and can be well-defined for an organism, the proteome continually changes in response to external and internal events. For example, E. coli will express different proteins and so have a different proteome when cultivated with minimal media instead of complete media. Similarly, during mammalian development cells express different proteins, develop dissimilar but characteristic proteomes, and ultimately differentiate into tissues.
It is also appropriate to articulate the reasons why proteins, as well as nucleic acids, should be studied on a very large scale. Essentially, there is certain information that a genome sequence can provide. This includes protein amino acid sequences and likely initiation and termination codons, although even these can be difficult to predict when intron-exon junctions are not clear or when variable mRNA splicing occurs. mRNA expression research (e.g., cDNA microarrays as reviewed in reference 2) can provide some information about the protein expression levels and tissue distributions.
However, to obtain further information about a protein including subcellular location, turnover rate, post-translational modification, covalent and noncovalent associations, and how all this is affected by different external and internal conditionsit is necessary to study the proteins themselves. Only then can the subtleties be appreciated, like tissue-dependent variable post -translational modification of the same protein (e.g., human serotransferrin in plasma and cerebrospinal fluid) or the processing of a single polypeptide to produce many different products (e.g., the post-translational cleavage of protachykinin beta precursor into three peptide hormone products). Clearly proteome projects, which aim to identify and characterize all proteins expressed by a tissue or organism, are the necessary complement to large scale genome analysis.
Resources Available for the Study of Proteomes
Recent advances in the fields of two-dimensional gel electrophoresis, protein analysis, and computer databases together make proteome analysis possible. Figure 1 shows how these techniques are interfaced in a proteome project. Recent reviews provide detailed discussions of these techniques (1, 3-7). Genome sequencing technologies and projects, which provide the framework underlying proteome projects, are described in a recent "genome" issue of Nature Biotechnology (Vol. 14, No. 10, 1996).
Figure 1: Some resources available for the study of proteomes, and their application. The dotted rectangle highlights databases, which are now interconnected through the World Wide Web and so function more as a single entity. Note that for micro-organisms, the whole cell is often used for two-dimensional separations, but for multicellular organisms it is more useful to analyze one tissue at a time. If needed, samples can be prefractionated before two-dimensional separation.
The following sections illustrate the concept of large-scale protein analysis, where the goal is to separate the proteins of a tissue or an organism and then identify, analyze, and characterize them at a rate of hundreds of samples per week. Several dedicated "Proteome Analysis Facilities" employing these strategies have been established already, for example, at Odense University (Denmark), Macquarie University (Australia), Proteome Inc. (United States), Large-Scale Biology Corporation (United States), and Oxford GlycoSciences (United Kingdom).
Separating Complex Protein Mixtures by Two-Dimensional Gel Electrophoresis In proteome projects, one of the primary goals is to separate and visualize as many proteins from a sample as possible, thus allowing them to be catalogued by computer and studied by analytical techniques. Two-dimensional polyacryl -amide gel electrophoresis (2-D PAGE), as first described by Klose (8) and O'Farrell (9), is used to separate proteins by isoelectric focusing in a first dimension and then by their apparent molecular weight in an SDS-PAGE second dimension. Current methods can resolve as many as 3,000 proteins on a 16 x 18 cm gel, so in principle it is possible to resolve the entire proteome of a simple organism like M. genitalium or H. influenzae on a single two-dimensional gel. 2-D PAGE is one of the most efficient and powerful methods for purifying proteins in small quantities.
Immobilized pH gradients (IPG) (10,11) can now be used for the pH range 3 to 12 (Görg et al., unpublished results) and have become the method of choice for the isoelectric focusing. IPG gels do not suffer from cathodic drift and focus proteins to
equilibrium, thus providing very high reproducibility. Furthermore, IPG gels are commercially available, so the reproducibility among different laboratories needed to establish and standardize two-dimensional gel databases is now possible. For example, when S. cerevisiae proteins were separated by 2-D PAGE at laboratories in three countries, the correlation coefficient for pattern reproducibility of 470 proteins was 0.9994 (12). When very high resolution is required, the pH range of IPG gels can be narrowed, for example, so one pH unit spans 16 cm (13), allowing us to "zoom in" on a pH range of interest. An important feature for using IPG gels in the first-dimension is their ability to accommodate the high sample loads needed for micropreparative 2-D PAGE. When loaded by in-gel rehydration, 15 mg of protein can be separated on narrow or wide pH range gels (14), and it is anticipated that more than 90 mg could be loaded onto IPG gels that are wider and thicker than current designs. Current methods for sample loading produce hundreds of protein spots on a single gel, with the quantities of each protein ranging from high nanogram to low microgram amounts. The study of very low-abundance proteins by 2-D PAGE is still challenging, even after loading milligram quantities of samples. However, prefractionation either by subcellular compartmentalization (15) or by narrow-range micropreparative 2-D PAGE (16) often helps.
Protein Identification and Characterization Strategies Identifying proteins in two-dimensional gels is the first step towards understanding what proteins are expressed in an organism or tissue and to what degree. Traditionally, protein identification in two-dimensional gels has been difficult. The amounts present on two-dimensional gels have been limiting, and protein databases contained information for only a small proportion of proteins for any organism. Nevertheless, protein identification was achieved with methods such as immunoblotting, amino-terminal and internal peptide sequence analysis, co-migration of unknown proteins with known proteins, or overexpression of homologous genes (see reference 1). With the rapid advance of genome sequencing projects and with improvements in micropreparative two-dimensional gel technology, these two major obstacles of protein identification have been overcome. Now the challenge is to identify the best methods to rapidly and economically screen the hundreds to thousands of proteins in a two-dimensional gel, for protein identification and for evaluation of post-translational modifications.
Recently, a variety of techniques for protein identification have been defined (Figure 2), and the results obtained with these techniques have been spectacular. Proteins have been identified with a 75% success rate by matching their amino acid composition, estimated isoelectric point, molecular weight, and the species of origin against protein databases (17). This technique was shown to be useful with as little as 250 ng of protein, making it applicable
for many proteins prepared by 2-D PAGE. An even more sensitive approach is peptide mass fingerprinting: here low nanogram quantities of proteins are enzymatically cleaved, and the peptide masses are determined by mass spectrometry and are used for database searches (see reviews in references 1, 5, 7). Protein identification by amino acid composition and peptide mass fingerprinting are completely or partially automated, and so twenty or more samples can be analyzed each day with a single instrument. The time needed for sample preparation and analysis is so short that throughput is often limited by data interpretation.
Among the more impressive recent developments is tandem mass spectrometry to sequence gel-separated proteins. Wilm et al. were able to sequence up to 16 residues of trypsin-digested peptides, starting with only 5 ng of a silver-stained protein from a gel; by analyzing several peptides, they could assign all 73 residues of one protein (18) . Similarly, Yates et al. have developed approaches to automatically interpret and match raw tandem mass spectrometry data against nucleotide sequence databases, simplifying and streamlining the process of sequence determination and protein identification (19). Although these last two approaches are not yet commonly used, they illustrate the exciting potential of current protein analytical techniques.
Development of high-throughput identification methods has spawned a shift in identification strategy. Formerly, it was common practice to identify proteins by using a single, usually slow, and expensive technique to obtain one piece of high -fidelity information. Identification by extensive Edman degradation is one example, where proteins would be sequenced for 12 to 20 residues and only one or two proteins could be analyzed each day.
Now proteome researchers use methods that generate data quickly, economically, and easily: often the pieces of data by themselves are not sufficient for identification but when used in combination, they can identity proteins with high confidence. Notable in this regard are the combinations of peptide mass fingerprinting and Edman sequence data (20), amino acid composition and peptide mass fingerprinting data (21), amino acid composition and Edman sequence tag data (22), and peptide mass fingerprinting and mass spectrometry sequence data (23, 24). These strategies often also use estimated molecular weight, isoelectric point, and species of origin to help reduce the number of candidate proteins in databases. An hierarchical approach (Table 1) has been proposed for large-scale protein identification: inexpensive and rapid technologies are used first, followed by methods that require more time, resources, and labor. In this manner, proteins already included in databases are quickly identified and do not consume resources, and novel proteins become the focus of further research efforts.
After proteins in two-dimensional gels are identified, studies can be initiated to evaluate post-translational modifications. In proteome projects, large-scale approaches can be used to define sets of proteins with specific modifications. For example, two -dimensional electrophoresis of samples labeled with 32P or [3H]-mannose identifies proteins containing phosphate groups or mannose-glycans (25). Similar experiments with [3H]-farnesyl- or [3H]-geranylgeranyl-pyrophosphate (26) should identify farnesylated or geranylgeranylated polypeptides.
Figure 2: Strategies for identifying and characterizing proteins separated by 2-D PAGE (modified from reference 41). To generate peptides, proteins can be excised directly from gels, but it is advantageous to blot to membranes for other techniques, which can sometimes be performed sequentially on the same sample (e.g., Edman tag sequencing and amino acid analysis). Abbreviations: liquid chromatography (LC), mass spectrometry (MS), matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS).
Antibodies specific for phosphothreonine, phosphoserine, and phosphotyrosine can be applied to a blot of an entire two -dimensional gel to locate phosphorylated proteins (27), and lectin binding studies can reveal glycoproteins (28). These techniques provide information about the presence or absence of a specific protein modification, but to evaluate the level and sites of modifications, it is necessary to use chromatographic or mass spectrometric techniques. High performance chromatography techniques have been described for characterizing phosphoamino acids and monosaccharides in single spots from two-dimensional gels (29, 30), and glyco- and phospho-amino acids have been localized in proteins by mass spectrometry or Edman degradation techniques (5, 31). In contrast to protein identification, rapid methods for identifying post-translational modifications are not well established, partly because many types of modifications rem
Table 1: Hierarchy of analysis tools for mass screening of 2-D gel-separated proteins (modified from Wasinger et al., 1995). Rapid, inexpensive, and automated techniques that give easily interpreted data are used as a first step in protein identification, followed by slower, more expensive, analysis intensive techniques. The order of analysis is expected to change with advances in technology particularly in mass spectrometry. These analysis techniques are reviewed in references 1, 4, 5, and 7.
Order Identification Technique
1 peptide mass fingerprinting
2 amino acid composition
3 combined amino acid composition and peptide mass fingerprinting
4 amino-terminal Edman "sequence tag" and amino acid composition
5 mass spectrometry sequence tag (MS/MS or post -source decay)
6 mass spectrometry sequencing (MS/MS)
7 extensive amino-terminal Edman sequencing
8 internal peptide Edman sequencing
9 carboxyl-terminal sequencing (chemical or enzymatic)
ain poorly understood. However, it is expected that screening proteins for many types of modifications will become a routine part of proteome projects in the near future.
Reference Maps and Databases A major challenge in proteome projects is storing and organizing huge volumes of protein sequence, structure, and annotation data, integrating these with two-dimensional gel images, and linking protein and nucleic acid sequence databases. Ideally, databases carrying all this information should be user-friendly, globally accessible, and designed to allow seamless navigation between different sites.
Two-dimensional reference maps of tissues or organisms are becoming a central tool for organizing and understanding proteome data (3). Reference maps are made by scanning stained two-dimensional gels and then analyzing the digitized gel using imaging software such as Quest or Melanie (25, 32). Protein spot positions and boundaries are detected, spot intensities are calculated for quantitation, reference numbers are assigned to each spot, and isoelectric points and apparent molecular masses are calculated. Any protein spot can be annotated with information, including its identity, up- or down-regulation in response to various conditions, and association with disease states. With the improvements in analytical techniques, many reference maps are now highly detailed, with some showing several hundred identifications (33, 34). These detailed maps define the proteins present in a particular sample and their levels of expression.
The number of reference maps available on the World Wide Web is increasing (e.g., Figure 3). A list of currently available maps can be found at the WORLD-2DPAGE site, at the URL address:
http://expasy.hcuge.ch/ch2d/2d-index.html
Figure 3: A reference map of E. coli proteins, as viewed from the SWISS-2DPAGE database. The URL address for this specific page is: http://expasy.hcuge.ch/cgi-bin/map2/def?ECOLI. A two -dimensional reference map is shown, and all identified proteins are highlighted with a cross. Selecting the cross takes the user to a database text entry for that protein, which in turn is linked to other database entries for that protein.
Hypertext links have helped integrate reference maps with protein and other databases. An example of an integrated "proteome database" is the group including SWISS-2DPAGE, SWISS -PROT, PROSITE, and SWISS-3DIMAGE, accessible through the ExPASy World Wide Web server (35) at the URL address:
This site can also be accessed from some gel image analysis programs (32). In the SWISS-2DPAGE database, there are E. coli, S. cerevisiae, and twelve human tissue two-dimensional reference maps: all identified proteins are highlighted with a red cross (36). To access text information about an identified protein, the user simply selects a spot on the gel image (Figure 3). A hypertext link can then be used to obtain the full SWISS-PROT entry for that protein, displaying protein sequence, domain structure, information on known post-translational processing and modifications, and references. From SWISS-PROT, the user can select a link to SWISS-3DIMAGE to see the three-dimensional structure of the protein, if it is known, or to submit the sequence to the SWISS-MODEL three-dimensional modeling tool (37). Also, from SWISS-PROT, the user can select links to pertinent information from DNA sequence databases (EMBL/Genbank), chromosomal and genomic maps (GDBGenome Database), bibliographic references and abstracts (Medline), and databases on the association of human proteins with diseases (OMIM Online Mendelian Inheritance in Man).
Applications of Proteome Analysis
The applications of proteome projects are, of course, numerous, and it is impossible to provide a comprehensive list here. However, in a broad manner, we can discuss how these technologies can be applied to certain biological issues.
For single-cell organisms with known genome sequences such as H. influenzae, M. genitalium, M. jannaschi, E. coli, and S. cerevisiae proteome projects can provide reference maps where all proteins are identified, with immediate and considerable details about protein expression. From there, cells can be treated in ways that help define protein function, metabolic pathways, and involvement in pathogenesis. For example, to better define yeast protein functions and their involvement in pathways, genes on chromosome III are systematically deleted one at a time, and then global protein expression is evaluated with two-dimensional reference maps (Stephen Fey et al., unpublished results). As expected, deletion of a protein involved in a metabolic pathway can affect expression of other proteins in that pathway and proteins with similar functions in other compensating, redundant pathways. In a second example, E. coli proteins are evaluated for their response to changes in growth conditions to learn how these responses might be regulated (38). This project has already examined global changes due to nitrogen, phosphate, and glucose starvation; treatment with substances such as cadmium chloride, hydrogen peroxide, 2,4-dinitrophenol, and nalidixic acid; and growth at different temperatures. Over 1,000 proteins have been classified into sets of co-stimulated proteins (stimulons) and sets of co-regulated proteins (regulons). Not surprisingly, many organizations would like to test the effects of drugs using this approach.
With eukaryotes, subcellular fractionation before two -dimensional separation helps define the location of proteins. Similarly, fractionation of native protein complexes by methods such as immunoprecipitation (39) followed by SDS-PAGE can help define covalent and non-covalent associations of proteins, aiding the construction of protein linkage maps.
Proteome analysis of multicellular organisms also presents a myriad of possibilities. The application of two-dimensional reference maps to the study of development presents an opportunity to examine the regulation of gene and protein expression on a large scale. In mature organisms, protein expression differences between tissue types can be defined by comparing reference maps, highlighting proteins that are tissue-specific (e.g., reference 36). This analysis also defines the common "housekeeping" proteins, which account for about 80% of proteins in any tissue. This high level of common proteins means that much of the knowledge gained by studying one tissue can be extrapolated to other tissues in the same organism. An important application of proteome analysis is the comparison of human tissue in normal and diseased states. In the study of cancer, a number of marker proteins have been defined using this approach (e.g., reference 40), and some may be useful for early disease diagnosis.
Finally, it is important to note that proteome analysis presents unique opportunities for the study of organisms that are not well defined at the molecular level. This is because many proteins from poorly defined organisms (e.g., Candida albicans ) can be identified by comparison to well-defined systems (e.g., S. cerevisiae). Rapid methods for these comparisons have been described (21). In this manner, proteins for many species can be identified from reference maps without cloning and sequencing the corresponding genes. Proteins that cannot be easily identified across species boundaries then become the focus of detailed characterization and DNA sequencing efforts.
A Final Word
This review has presented an overview of proteome analysis, but it has not covered work from the many fields that intersect those described here. For example, three-dimensional structural analysis has not been discussed here, even though it is central to understanding protein function and will become more important as structural and sequence databases expand. Also, important conceptual advances in the field of cell biology have not been described here. However, the potential of large-scale protein analysis has been illustrated, along with several possible applications of this technology.
To conclude, we want to emphasize that numerous challenges will remain for understanding even the simplest biological systems, even after their genomes have been sequenced and proteins characterized. For example, when we have finally defined the individual functions of the 470 proteins in M. genitalium or the 6,000 proteins in S. cerevisiae, the challenge will become to understand the complexities of how these proteins interrelate and work in a coordinated, cohesive manner to achieve metabolic and reproductive functions in optimal and adverse conditions. There remains much work to be done!
Acknowledgments
We thank Keith Williams, Amos Bairoch, Andrew Gooley, Manuel Peitsch, Ron Appel, Steve Pennington, and Jean-Charles Sanchez for many proteome discussions. We acknowledge support of the Swiss National Fund for Scientific Research (grant 31 -33658.92), and MRW acknowledges financial support of the Helmet Horton Foundation.
References
1. Wilkins, M.R., Sanchez, J.-C., Gooley, A.A., Appel, R.D., Humphery-Smith, I., Hochstrasser, D.F. and Williams, K.L. (1995) Biotechnol. Gene. Eng. Rev. 13, 19-50.
2. Schena, M. (1996) Bioessays 18, 427-31.
3. Wilkins, M.R., Sanchez, J.-C., Williams, K.L. and Hochstrasser, D.F. (1996) Electrophoresis 17, 830-838.
4. Patterson, S.D. (1994) Anal. Biochem. 221, 1-15.
5. Patterson, S.D. and Aebersold, R. (1995) Electrophoresis 16, 1791-1814.
6. Anderson, N.G. and Anderson, N.L. (1996) Electrophoresis 17, 443-453.
7. Mann, M. and Wilm, M. (1995) Trends. Biochem. Sci. 20, 219-224
8. Klose, J. (1975) Human Genetik 26, 231-243.
9. O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021.
10. Bjellqvist, B., Ek, K., Righetti, P.G., Gianazza, E., Görg, A., Westermeier, R. and Postel, W. (1982) J. Biochem. Biophys. Meth. 6, 317-339.
11. Görg, A., Postel, W. and Gunther, S. (1988) Electrophoresis 9, 531-546.
12. Blomberg, A., Blomberg, L., Norbeck, J., Fey, S.J., Larsen, P.M.,Larsen, M., Roepstorff, P., Degand, H., Boutry, M., Posch, A., and Gorg, A. (1995) Electrophoresis 16, 1935-1945.
13. Bjellqvist, B., Sanchez, J.-C., Pasquali, C., Ravier, F., Paquet, N., Frutiger, S., Hughes, G.J. and Hochstrasser, D.F. (1993) Electrophoresis 14, 1375-1378.
14. Rabilloud, T., Valette, C. and Lawrence, J.J. (1994) Electrophoresis 15, 1552-1558.
15. Harrington, M.G., Coffman, J.A., Calzone, F.J., Hood, L.E., Britten, R.J. and Davidson, E.H. (1992) Proc. Natl. Acad. Sci. USA 89, 6252-6256.
16. Hochstrasser, A.C., James, R.W., Pometta, D. and Hochstrasser, D.F. (1991) Appl. Theor. Electrophor. 1, 333 -337.
17. Wilkins, M.R., Pasquali, C., Appel, R.D., Ou, K., Golaz, O., Sanchez, J-C., Yan, J.X., Gooley, A.A., Hughes, G., Humphery-Smith, I., Williams, K.L. and Hochstrasser, D.F. (1996) Bio/Technology 14, 61-65.
18. Wilm, M., Shevchenko, A., Houthaeve, T., Breit, S., Schweigerer, L., Fotsis, T. and Mann, M. (1996) Nature 379, 466-469.
19. Yates, J.R. 3rd, Eng, J.K. and McCormack, A.L. (1995) Anal. Chem. 67, 3202-3210.
20. Patterson, S.D., Thomas, D. and Bradshaw, R.A. (1996) Electrophoresis 17, 877-891.
21. Cordwell, S., Wilkins, M.R., Cerpa-Poljak, A., Gooley, A.A., Duncan, M., Williams, K.L. and Humphery-Smith, I. (1995) Electrophoresis 16, 438-443.
22. Wilkins, M.R., Ou, K., Appel, R.D., Sanchez, J.-C., Yan, J.X., Golaz, O., Farnsworth, V., Cartier, P., Hochstrasser,
D.F., Williams, K.L. and Gooley, A.A. (1996) Biochem. Biophys. Res. Commun. 221, 609-613.
23. Mann, M. and Wilm, M. (1994) Anal. Chem., 66, 4390-4399.
24. Jensen, O.N., Vorm, O. and Mann, M. (1996) Electrophoresis 17, 938-944.
25. Garrels, J.I. and Franza, B.R. (1989) J. Biol. Chem. 264, 5283-5298.
26. Danesi, R., McLellan, C.A. and Myers, C.E. (1995) Biochem. Biophys. Res. Commun. 206, 637-643.
27. Coghlan, M.P., Pillay, T.S., Tavare, J.M. and Siddle, K. (1994) Biochem. J. 303, 893-899.
28. Gravel, P., Golaz, O., Walzer, C., Hochstrasser, D.F., Turler, H. and Balant, L.P. (1994) Anal. Biochem. 221, 66-71.
29. Packer, N.H., Wilkins, M.R., Golaz, O., Lawson, M., Gooley, A.A., Hochstrasser, D.F., Redmond, J.W. and Williams, K.L. (1996) Bio/Technology 14, 66-70.
30. Yan, J.X., Packer, N.H., Tonella, L., Ou, K., Wilkins, M.R., Sanchez, J.-C., Gooley, A.A., Hochstrasser, D.F. and Williams, K.L. (1997) J. Chromatogr. A, In Press.
31. Pisano, A., Redmond, J.W., Williams, K.L. and Gooley, A.A. (1993) Glycobiology 3, 429-435.
32. Wilkins, M.R., Sanchez, J.-C., Bairoch, A., Hochstrasser, D.F. and Appel, R.D. (1996) Trends Biochem. Sci., in press.
33. Celis, J.E., Rasmussen, H.H., Gromov, P., Olsen, E., Madsen, P., Leffers, B., Honoré, B., Dejgaard, K., Vorum, H., Kristensen, B., Ostergaard, M., Haunso, A., Jensen, N.A., Celis, B., Basse, B., Lauridsen, J.B., Ratz, G.P., Andersen, A.H., Walbum, E., Kjaergaard, I., Andersen, I., Puype, M., Van Damme, J. and Vanderkerckhove, J. (1995) Electro -phoresis 16, 2177-2240.
34. Golaz, O., Hughes, G.J., Frutiger, S., Paquet, N., Bairoch, A., Pasquali, C., Sanchez, J.-C., Tissot, J.D., Appel, R.D., Walzer, C., Balant, L. and Hochstrasser, D.F. (1993) Electrophoresis 14, 1223-1231.
35. Appel, R. D., Bairoch, A. and Hochstrasser, D. F. (1994) Trends Biochem. Sci. 19, 258-260.
36. Sanchez, J.-C., Appel, R.D., Golaz, O., Pasquali, C., Ravier, F., Bairoch, A. and Hochstrasser, D.F. (1995) Electrophoresis 16, 1131-1151.
37. Peitsch, M.C., Wells, T.N., Stampf, D.R. and Sussman, J.L. (1995) Trends Biochem. Sci. 20, 82-84.
38. VanBogelen, R.A., Abshire, K.Z., Pertsemlidis, A., Clark, R.L. and Neidhardt, F.C. (1996). "Gene-Protein Database of Escherichia coli K-12: Edition 6." In Escherichia coli and Salmonella typhimurium Cellular and Molecular Biology, Second Edition. (F.C. Neidhardt, R.I. Curtiss, C.A. Gross, J.L. Ingraham, and M. Riley, eds.). ASM Press, Washington D.C. pp. 2067-2117.
39. Hynes, G., Sutton, C.W. and Willison, K.R. (1996) FASEB 10, 137-147.
40. Grover, P.K. and Resnick (1995) Prostate 26, 12-18.
41. Williams, K.L., Gooley, A.A. and Packer, N.H. (1996) Today's Life Science 8 (6), 16-21.
Marc Wilkins may be contacted at the Central Clinical Chemistry Laboratory, Geneva University Hospital, 24 Rue Micheli-du-Crest, 1211 Geneva 14; E-mail: marc.wilkins@dim.hcuge.ch.
Return to the
The ABRF Home
Page