One Step Beyond: Going Beyond Genomics With Proteomics And Two-Dimensional Gel Technology

Proteomes and 2D Gel Apparatus Providers Big science has moved on to proteins. With the new brand of science termed proteomics--named by the Australians Marc Wilkins and Keith Williams to mean the "set of PROTEins encoded by the genOME"--the push is on around the globe to produce a complete description of a cell/tissue/organism in terms of the proteins produced. The challenge is all the greater because the expressed protein pattern changes with time and environment--responding to developmental signals, disease, nutrients available, and more. And despite the elegance and speed with which gene expression can be monitored with some of the new array and chip technologies, mRNA levels don't necessarily correlate with that of proteins (L. Anderson and J. Seilhamer, Electrophoresis, 18: 533-7, 1997). Hence, many are turning their attention to where they believe the true action is. At the center of this effort are two-dimensional gels (2-D gels). When some 20 years ago, Patrick O'Farrell described the first 2-D gel system (P. H. O'Farrell, Journal of Biological Chemistry, 250(10):4007-21, 1975), he predicted that with this technology it should be possible to resolve 5,000 protein species on a single gel. He wasn't far off. Today, a gel with thousands of spots is considered routine, with claims in the literature of separations of close to 10,000 species (J. Klose and U. Kobalz, Electrophoresis, 16:1034-59, 1995). Using 2-D gels in their various emanations, the proteomes, or rather partial proteomes from close to 20 organisms and a myriad of tissues--some rather exotic such as newborn ear--have been cataloged and can be displayed The year 1975 was an important year in the science of proteomics--of course back then no one was talking about proteomics; the focus was on improving the resolution of polyacrylamide gel electrophoresis. In that year, several groups published papers describing a method for displaying proteins in two dimensions. The concept was simple yet elegant. After running out proteins in tube gels based on one property, isoelectric point, the gel was turned 90 degrees and run into a slab gel that resolved the species based on a second property, molecular weight. This disarmingly simple trick increased by close to 100-fold the number of species that could be resolved from a cell extract by separating into distinct spots the numerous species that made up what appeared to be a single band when viewed in one dimension. Two-dimensional gels, while simple in concept, are anything but simple to run. From start to finish--from casting gels to interpreting the patterns obtained--this technique poses numerous challenges to those attempting to use it. But since its inception, some refinements in the technology have addressed the problem areas, and a variety of devices and tools have become available that may also advance this field. While with careful attention to detail individual investigators or labs could reproduce 2-D protein patterns, it was rare for disparate groups to get the same pattern, rendering comparisons between laboratories difficult and the creation of shared databases largely pointless. One source of the variability for which a technological solution has been found is instability (drift) of the pH gradient in the first dimension. (The pH gradient is established through carrier ampholytes--amphoteric molecules that when mixed in varying proportions and subjected to an electric field produce gradients covering specific pH ranges.) In the early 1980s, immobilized pH gradients (IPGs), in which the functional groups are covalently bound to the polyacrylamide matrix, were developed by Bengt Bjellqvist at LKB (B. Bjellqvist, et al., Biochemical and Biophysical Methods, 6:317-39, 1982) which in large measure have put this problem to rest. Using this technology, investigators in labs on two continents have run the same samples and report less than 1 mm variation in spot migration patterns (A. Blomberg et al., Electrophoresis 16:1935-45, 1995). The reason 2-D gels were developed, in the words of one of its developers, was to provide an experimental tool for looking at change--changes in protein expression accompanying development, for example, or changes in the forms a protein can take on through various modifications. Only in the last few years, with the advent of DNA chips, has any technique even come close to rivaling 2-D gels for looking at global changes in expression. Nor has any simpler method for looking at protein modification come along. However, using 2-D gels as screening tools is a misuse of the technology, according to O'Farrell, who would rather it be used more in hypothesis-driven research. With the emphasis on the cataloguing and the cross-identification of every protein species, 2-D gels have turned into a workhorse rather than a showcase. The question can be asked, Are 2-D gels up to the task, or have they been pushed to their limit? Can further refinements be made to the technology to enable realization of the goal that the protein jockeys have set for themselves? What other new technologies are needed, and are there any in the offing? Running 2-D gels successfully requires a panoply of research tools. Starting at the beginning, sample preparation/protein solubilization is the first in a long series of nontrivial tasks, and in fact, one key to the success of 2-D gels and the proteome may lie in the way samples are prepared and prefractionated before reaching the gels. The first dimensional isoelectric focusing gel (IEF) or IPG, is the next hurdle, and here the challenge is getting reproducible gradients. As mentioned above, IPGs have made this task easier, but researchers who have accumulated both experience and data with carrier ampholytes may be reluctant to part with them. And in experienced hands carrier ampholytes can give reproducible results (Lopez and Patton, Electrophoresis, 18:338-43, 1997) without some of the pitfalls of immobilines--difficulties at the second dimension interface, or the tendency for proteins to precipitate at their isoelectric points when immobilines are being used preparatively. On the other hand, immobilines have a large loading capacity--several fold more than an IEF gel--good news for those wishing to do downstream analyses, such as sequencing or mass determination. Options for staining/detecting protein spots continue to mount, as new fluorescent dyes are added to coomassie, silver, zinc, and of course the old standby, radioactivity. Each has its advantages; nothing can surpass radioactivity for sensitivity (How can you do better than a dpm, after all?), although silver will detect nanogram quantities of protein. For quantitation, colloidal Coomassie blue demonstrates good stoichiometry, though it is limited in sensitivity. Imaging and image analysis follow next. While most standard imaging systems--storage phosphor screen or fluoro imagers, flatbed document scanners, or even old-fashioned densitometers--will do the job of collecting images of 2-D gels, specialized software has been developed for analyzing the complex patterns of spots. Packages are available that will go from spot detection to database searches and, in one case, even to web page construction for posting your 2-D gel pattern on the internet. And as seems to be true for all biological computing, PC based systems are slowly supplanting the UNIX systems, though both are in common use. Cataloging the protein readout of an organism is actually not a novel concept, only the word "proteome" is new. Since the late 70s, several research groups have been working toward this goal: Leigh and Norman Anderson developed the idea of a human protein index, using immunological means to identify proteins, and Frederick Neidhardt and his group have worked on the E. coli proteome, as it now would be called, having created an inventory of proteins displayed on 2-D gels by using reference 2-D gels with known proteins radiolabeled in total cell extracts to make id's on individual spots. The tools being applied to the proteome, however, have evolved since the 1980s, as other analyses are now being applied to 2-D gel profiles--in particular, mass spectrometry (MS), both matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF) and electrospray ionization tandem quadrupole (ESI/MS/MS) techniques, along with amino acid composition. Proteins are first subjected to controlled proteolysis, either in the gel or after transfer to a blotting membrane, and the resulting peptides analyzed. The properties of the fragment collection from an individual spot (size and amino acid composition) can then be compared with the same properties of known proteins, or even putative proteins from gene data banks, to provide often unequivocal identification of individual proteins. Downstream analyses of peptides solve some of the thorny problems of 2-D gels. Immobilines notwithstanding, vagaries of 2-D gels persist--spots have been known to move around a bit, making a positive identification based on position alone uncertain, especially for small spots. However, the application of MS to protein identification relieves the pressure for getting reproducible gels, since the gels in this situation are being used merely to isolate species one from the other, not to identify the proteins. However, there is a limit to how many spots can be analyzed this way, and only with high-throughput systems (and maybe not even there) could more than a few hundred species be analyzed. Two-dimensional gels fail in several areas: the ability to detect and analyze low-abundance proteins and the ability to separate proteins at the extremes of both isoelectric point (pI) and molecular weight. Furthermore, using denaturing conditions--the second dimension typically is SDS-PAGE--destroys activity, eliminating any possibility of ascribing functionality to spots. Some new devices or approaches are being developed to address these issues. One avenue is prefractionation of proteins through the isolation of subcellular fractions, for example, which can improve the detection of low abundance proteins. The Australian company Gradipore is developing a flow-through system for rapid separation of mixtures of protein with respect to size and pI. Its GradiFLOW technology uses a series of acrylamide membranes that fractionate macromolecules with a range of sizes (104-106 daltons). When used in conjunction with an electric field, the membranes can effect pI separations of as little as 0.2 pH units, a kind of in-solution 2-D separation. The virtue of this instrument is its ability to process in a matter of 30 minutes to an hour large amounts of material in a native state, which can then be loaded onto a 2-D gel to reveal low-abundance molecules. Analyses of proteins on microchips are now possible as well. Ciphergen, a Palo Alto, California-based company, offers ProteinChip™ arrays that employ a technology called SELDI™ (surface-enhanced laser desorption/ionization). Proteins are captured on the surface of a 2 mm chip bearing any of a variety of affinity surfaces--antibodies, receptors, ligands, nucleic acids, carbohydrates, or chromatographic surfaces--after which the protein composition can be mapped and compared with other proteins (Retentate Mapping™) or evaluated with respect to a variety of protein- protein interactions. A variety of specialized tools are available that have been and are continuing to be developed for proteome research. A selection of these follows. Screen shot from Amersham Pharmacia Biotech's ImageMaster software Amersham Pharmacia Biotech, inventors of the Immobiline™ IPG technology, have taken this concept one step further and now offer, in addition to solutions of Immobiline buffers for preparing gels, Immobiline DryStrip preformed IPG gel strips in a variety of lengths, pH ranges, and gradient shapes. The company also offers a complete system for the first dimension, IPGphor™, which is a single unit for rehydrating, applying sample, and running up to 12 IPG strips at up to 8,000 V. Amersham Pharmacia Biotech supports this technology with a vast array of equipment for the running of 2-D gels in both vertical and horizontal formats: gel apparatus from mini to its high-throughput system, DALT, with which as many as 23 gels can be poured and 10 run simultaneously. The latest addition to its suite of products is the Image-Master™ 2-D software. Bio-Rad has developed the ProteomeWorks™, an integrated set of standardized materials including prepared solutions, gels, and equipment to produce reproducible data with advantages in process efficiency and throughput. ProteomeWorks system includes kits, reagents, and equipment for sample preparation and quantification, a new line of IPG strips and a 10,000 V first dimension cell designed specifically for the proteome process, precast SDS-PAGE gels and the PROTEAN® II xi Multi-Cell, colloidal coomassie stain, silver stain kits, premixed buffers, blotting apparatus, reagents, and PVDF membranes; standards for protein electrophoresis. In addition, Bio-Rad offers a full range of imaging equipment options (densitometry, CCD camera, laser scanning, and storage phosphor imaging), imaging software, and an automated protein spot excision system which provides hands-free cutting of protein spots from gels or membranes for mass spec and other identification processes Screen shot of Genomic Solutions' 2-D Analyzer The ESA electrophoresis group and Genomic Solutions have collaborated in the development of a series of products that allow the separation of protein mixtures with high resolution 2-D gels, and provide the follow-on products to automatically analyze gels and identify individual proteins within the gel. Genomic Solutions also offers complete image acquisition and analysis packages such as the Bio Image Visage™ System and the complete line of Phoretix 2-D Analysis Software packages. Soon to be introduced is a digestion robot (ProGest™) for preparing samples for mass spec, which will be joined later with a spot excision robot to further automate the process of picking spots from gels, along with a new version of software to automate the analysis process. Also offered by Genomic Solutions is a complete 2-D Gel Running Service including sample preparation, gel running, image analysis and comparison, spot excision and extraction, Western blotting, and mass profiling, on a fee for service basis (no royalties). Large Scale Biology Corporation study Large Scale Biology Corporation builds systematic proteome databases for the discovery and development of improved drugs, diagnostics and agricultural products. Major efforts include the Molecular Anatomy and Pathology (MAP™) database, describing the protein composition of human cells and tissues and how these change in disease, and the Molecular Effects of Drugs (MED™) database, characterizing the effects of drugs and other xenobiotics on protein expression. In addition, LSB conducts numerous targeted proteomics research projects with partners in the pharmaceutical, diagnostic and biotechnology industries. The Company continues to make significant investments in the development of proteomics technology, building on its longstanding record of innovation in this area Oxford GlycoSciences, developers of technology for each step in the process from preparing samples to analysis, has introduced a novel approach to 2-D separation in the use of matrices stabilized by bonding to glass. Proteome, Inc. is a provider of proteome databases that integrate proteomic and genomic data with protein information in the research literature. The company, founded as a 2-D gel technology developer and service provider, focuses now on building the databases and software tools needed for interpretation of proteomic and functional genomic data. SITES: Australian Proteome Analysis Facility www.proteome.org.au Swiss Institute of Bioinformatics www.expasy.ch Danish Center for Human Genome http://biobase.dk/ COMMERCIAL SITES: Large Scale Biology Corporation www.lsbc.com Phoretix www.phoretix.com Proteome Inc. www.proteome.com A particularly helpful and detailed review of the technologies used in proteome research can be found at: http://www.proteome.usyd.edu.au/proteome_review.html (Reprinted from Ian Humphery-Smith, Stuart J. Cordwell and Walter P. Blackstock, Electophoresis 18(8): 1217-1242, 1997.) Alpha Sean- Densitometer designed for 2-D proteomics from Alpha Innotech Alpha Innotech, developers of CCD detection systems, has introduced a large format (12.2x17.2) densitometer to scan in large 2-D gels. In addition to the hardware, a 2-D software package is also available. The 2-D AlphaMatch analysis software package is a comprehensive package offering automatic spot detection, measurement and matching. After analysis the 2-D database package provides the user with unlimited archiving and searching capabilities. Flicker is a public domain software program for comparing images from different Internet sources on your web browser. Melanie II 2-D (available from Bio-Rad) was developed by Denis Hochstrasser and the MELANIE group in Geneva, Switzerland. Melanie II 2-D Analysis Software automates the identification, quantitation, and comparison of 2-D samples in various formats--IPG gels, silver-stained gels, amino-black PVDF membranes, phosphorescence plates, and autoradiograms. It can be used for both single and high volume 2-D comparison studies. PDQuest is offered by Bio-Rad for the analysis and databasing of 2-D gels. With PDQuest, you can analyze up to 100 gels at a time in a matched set, compare 100 match sets to each other, and form databases of up to 1,000 gels with web based querying. The landmarking ability is intuitive, allowing alignment of all gels quickly and properly the first time. A matching algorithm ensures that spots are correctly matched, with little or no editing. Once the gels are analyzed, you can perform a wide variety of queries on the gels to discover subtle differences between samples, which can be applied, in gel-to-gel comparisons, as well as protocol-to-protocol experiments. PDQuest is available for a free unlimited 30-day trial at www.discover.bio-rad.com. Phoretix is concentrating its efforts on the data analysis side of proteomics with its Phoretix 2--D, a software system for analyzing 2-D gels. In addition to the usual constellation of functions--spot detection, editing, measurement and spot matching between gels--Phoretix 2-D can be partnered with a 2-D gel database package designed for project work. Phoretix is in the final stages of developing a totally new version which addresses the problems of variations between gels, the difficulties in accurate quantitation and provides for creation of web pages from 2-D experiments. A test version of the new software is currently available to those scientists that become part of its test program. Scanalytics : GELLAB II+ is a Windows-based software package that automates the analysis of 2-D gels. The analytical package, which includes automatic spot detection, spot normalization to compensate for distortion during the run, and spot matching overlay, is combined with a dedicated relational database based on the program of Peter Lemkin of NCI. Segmentation analysis of 2-D protein gel done with Scanalytics GELLAB II+ software. Spots are identified and assigned numbers. Color indicates intensity of spot, with blue being least intense, black most intense. Proteomics is a field with both enthusiasts and detractors. The detractors argue that not much has changed in protein analysis by 2-D gels in twenty years with only modest improvements in technology. In addition, they say that no new knowledge or understanding has been gained from all the effort and money that has been put into this endeavor--or at least the gains in knowledge are out of proportion with the effort expended. On the other hand, the enthusiasts, while acknowledging that proteomics isn't easy to do and 2-D gels are far from perfect, feel that this is the best approach for studying the expression of large numbers of proteins. New methodologies are in the works (see call out box), but are still years away from approaching the utility of 2-D gels. But there's no denying that there is excitement in the field of proteomics--with a host of companies, big and small, trying to fill in the gaps in present technologies, and an ever-increasing collection of databases. While 2-D gel patterns abound in the literature and on the web, the jury is still out on what the pay-off will be. The web is a rich resource for information on 2-D gels and proteomics. Several nationally funded groups devoted to proteomics have web sites with information, databases, and links to others. Technology reviews for 2-D gel electrophoresis and for proteomics in general also are available on the web. Amersham Pharmacia Biotech has produced a handbook of 2-D electrophoresis using immobilized pH gradients--principles and methods--that is available for researchers in this field, and a number of monographs have been published with collected works. For more information about products mentioned in this article please circle Reader Service #73. Author Laura De Francesco can be contacted at ldefrancesco@the-scientist.com. She acknowledges the help in the preparation of this manuscript of Patrick O'Farrell (Department of Biochemistry and Biophysics, UCSF), Jim Garrels (Proteome Inc.), and Brad Walsh (Australian Proteome Analysis Facility) AUSTRALIA by Brad Walsh, Facility Manager, Australian Proteome Analysis Facility The Australian Proteome Analysis Facility (APAF) was funded by the Australian federal government in 1996 and has been at the forefront of proteome technology since that time. Indeed, the term "proteome" was invented by Dr. Marc Wilkins and the founding director of the facility, Professor Keith Williams. Once the facility was established, it was clear that to support this new and exciting technology would require a commitment to the development of automation and robotics and integrated systems for proteomics. In fact, in the recent past the Macquarie University Centre for Analytical Biotechnology (MUCAB), from which APAF grew, was active in automating the postseparation analysis aspect of proteomics. Added to this was the facility researchers' expertise in studying posttranslational modification, including glycosylation and phosphorylation, with techniques developed in house. Thus as APAF began its life, it received further funding in the form of a state government grant to enable it to develop technologies in use today, including the following: Machines such as the Glycosite™, a protein sequencer specially modified to detect glyco-amino acids, as well as high-throughput amino acid composition using F-moc chemistry and a robot that was capable of both adding the derivatizing agents to the amino acids and injecting them onto the HPLC, thus allowing amino acid composition to be determined rapidly, with a turn-around time of about 40 minutes per sample. Tag sequencing, in which a protein sequencer is modified to include a carousel. On this carousel, one could place up to 16 protein spots, each separated by 2-D gels, and simply do a short N-terminal tag on the molecule. It has been found that an N-terminal tag of four to five amino acids is sufficient to identify a protein and certainly, coupled with the information of species, pI and molecular weight, is more than adequate to identify known proteins from databases. Since then, APAF has been very active in developing second-generation proteomic: improved electrophoresis and blotting apparatus, in conjunction with Bio-Rad Laboratories, as well as new solubilization techniques to allow one to run better 2-D gels; new blotting techniques to allow one to blot proteins better to PVDF membranes for subsequent identification; and further developments on the Glycosite, C-terminal sequencing and in automating the process of cutting spots from gels and blots (in conjunction with Advance Rapid Robotic Manufacturers). Of course, automation of the peptide mass fingerprinting process has been a priority, as well as development of bioinformatic software that is able to take peaks from MALDI mass spectrometers and translate them into identified proteins with little human intervention. A close association with Geneva University Hospital and the Swiss Institute of Bioinformatics has seen development of proteomics tools on ExPASy (http://expasy.hcuge.ch). The wave of proteomics is building to a crest, and the excitement of this field can be felt as you walk around the APAF laboratories. Discovery projects have been initiated in the areas of mammalian biology and medicine, microbiology, and agriculture, and these samples are beginning to yield up their secrets. Proteomes and 2D Gel Apparatus Providers

One Step Beyond: Going Beyond Genomics With Proteomics And Two-Dimensional Gel Technology

Interested in reading more?

Become a Member of