On the Fast Track in Functional Proteomics

Graphic: Leza Berardone Researchers in Canada and Denmark are employing mass spectrometry, three-dimensional tissue biology, and supercomputing to blaze a trail in functional proteomics research. In the process, they're putting their company, MDS Proteomics Inc., on the fast track in the latest race to develop new drug targets and eventually better treatments for all kinds of diseases. By using this combination of technologies, MDS Proteomics is accelerating the process of identifying, analyzing, and understanding the myriad structures, interactions, and pathways of proteins. "This integrated approach is critical to developing a new generation of drugs to treat disease," says Matthias Mann, scientific founder of MDS Proteomics and chief proteomics officer. Medical Data Services (MDS) Inc., an international health and life sciences company based in Toronto, launched MDS Proteomics in 2000 following the purchase of Protana, a Danish proteomics company. It is currently expanding its headquarters, and research facilities are under development in Charlottesville, Va., and Boston. MDS Proteomics uses a "one-two-three, punch" approach in its research: The process begins with growing three-dimensional cells in a state-of-the-art optimized suspension culture device. Researchers then use highly sensitive mass spectrometry to analyze the protein activity in the cells. A highly sophisticated bioinformatics system crunches the huge amounts of data generated so that protein 'pictures' emerge in a new, more revealing way. From Gel to BioReactor Although researchers had been investigating gel-purified proteins, this past January MDS Proteomics hired a young cellular biologist from Odense University, Jorge Sans Burns, who brought with him the notion of using three-dimensional tissue biology to grow the cells. Burns had learned about the optimized tissue culture suspension device, now known as the BioReactor, by reading the current literature in cell biology. NASA scientists developed this rotating wall vessel in the mid-1980s to give the space agency's biologists a way to research the genetic impact of microgravity on astronauts. The BioReactor has since been shown to produce three-dimensional tissue.1,2 "The idea struck me that sample handling is a key issue," says Burns, now a senior research scientist at MDS Proteomics lab in Odense. "It's an overworn phrase, but 'garbage in, garbage out.' If we want to have something that is representative of what is going wrong inside a patient, we need to be very careful about how we handle that sample, especially in the context of proteomes, because they're highly context dependent. They'll be changing very rapidly, contrary to the genome, which is more stable." As a result, how the tissue is handled is of "prime importance" in the end result and "whether it means anything in terms of what's going on inside the patient," says Burns, who earned his cellular pathology degree at the University of Bristol, studying under Sir Michael Anthony Epstein of Epstein-Barr virus fame. 3 Burns points out that conventional, gel-purified methods can be problematic, and "they may introduce artifacts" because the cells are trying to grow on a piece of plastic. Beyond that, he adds: "Since we are 3-D organisms with tissues, it seemed to make a lot of sense to approach this from that angle for the simple reason that the BioReactor system of three-dimensional tissue modeling appears to be a way to create the proteins in a more 'relevant' way to people." Upon suggesting the company employ BioReactors to grow the cells, Burns got an almost immediate 'thumbs-up.' MDS Proteomics has since purchased enough BioReactors "to conduct the first stage of experiments and to look at a number of things in parallel." Burns is now looking to collect information in the BioReactors that is "more representative" of how given cells would behave inside the patient rather than cells grown in other, conventional cultures. Although the BioReactor is still relatively new, "a technology that's still to be explored ... its potential still to be realized," Burns notes that the BioReactor is "a valid tool" for getting closer to an in vivo like situation. "If nothing else, it is a cost-effective way of growing large numbers of cells," he says. "That's the least it does." Burns has already found that "the cells are behaving [in a way] they don't in other culture systems, more tissue-like, more as if they were in the human body." Therefore, in addition to allowing the company to scale-up its approach and enabling these researchers to acquire more relevant data on specific diseases, the BioReactors, he suggests, will also help them "tackle some of the issues related to the complexity of tissue organization." Such insights should help open the door to determine the use of certain proteins as diagnostic markers. "To express the cells in a more tissue-like manner is, I feel, one step toward being able to express the targets in the right sort of way," comments Burns. From BioReactor to Mass Spectrometer Once the cells emerge from the BioReactor, they are analyzed in Mann's laboratories, which house a large number of highly customized mass spectrometers. "We are using mass spectrometry to analyze very small amounts of proteins, as well as protein-protein interactions," explains Mann. Mass spectrometry, of course, has long been used for chemical and elemental analyses, but during the last decade, it has been successfully redefined to identify and characterize proteins, and Mann is recognized as one of the leading pioneers of its use in functional proteomics.4 "We are trying to study the function of the proteins and one of those things is to see what a protein interacts with and how it interacts with other proteins, and we are [looking to] do that on a large scale," he elaborates. "We can fuse a cancer gene, for example, to tag and then transvect it, [resulting] in interacting proteins, which we then analyze. So, with mass spectrometry, we have a way of seeing interacting proteins and thereby, we can determine the function of a specific gene," he explains. Given the revelation in February that the human genome consists of fewer than expected genes, one of the major take-home points is that human complexity lies in the proteins genes produce. "To quote my managing director [for Europe], Ole Vorm: 'The human genome data is water on the treadmill of proteomics,'" says Burns. "That there are fewer genes than we had believed puts much more onus on the proteome; there must be significantly more differences based on the proteomes." "There may be several hundred thousand proteins made from these 35,000 genes," Mann adds, reiterating what Francis S. Collins and J. Craig Venter said at the recent American Association for the Advancement of Science annual meeting in San Francisco. "Because the genes can make different proteins ..., the proteins will be much more diverse than the genes we will be studying; therefore, the numbers will be several-fold higher," Mann continues. "It's not like you measure the sequence once and you're done. You have to measure in all kinds of environments, so you may end up with orders of magnitude more data than the Human Genome Project." Computational Steps Consequently, MDS Proteomics is generating considerably more data than either the public or private human genome projects. Uncovering the multitudes of protein interaction requires substantial bioinformatics and supercomputing capabilities. To address that issue, the company formed a strategic alliance with IBM, which recently launched its Life Science Initiative, and is providing the firm with its primary computer infrastructure in Odense and at the company's corporate headquarters in Toronto. "We now have very large computers that allow us to search this huge amount of data that we're generating," says Mann. Actually, the systems in Odense and Toronto together form one of the largest Linux clusters in the world. The company may be young, but its scientists, notes Burns, "are some of the most cited mass spectrometry and proteomics scientists and have published some 60 papers in Cell, Science, and Nature, among other journals." With remarkable speed, MDS Proteomics has become, in fact, one of the largest functional proteomics ventures--now employing some 225 people (60 scientists), according to Mann--deep into conducting functional proteomics research. Although the company's first functional proteomics facility was established in Odense, MDS is rapidly expanding. In addition to a Toronto facility established in conjunction with Anthony Pawson's laboratory at the Samuel Lunenfeld Research Institute at the Mount Sinai Hospital--Pawson is a fellow of the Royal Societies of London and Canada and is also an MDS Proteomics founder--the company just opened a 50,000-square-foot state-of-the-art laboratory in its headquarters for analyzing protein interaction pathways, producing data for its proprietary BIND database to select and validate targets for its partners. In the soon-to-open Boston division, scientists will focus on high-throughout screening, structural biology and drug design, medicinal chemistry, and protein identification, while in Virginia, they will specialize in the analysis of complex mixtures of proteins to identify novel membrane bound receptor proteins as targets fro therapeutic drugs. Mass spectrometry has provided "the big change" over the last decade, says the German-born and -raised Mann, who earned his chemical engineering degree from Yale University and also has a background in mathematics, physics, and biotechnology. "In terms of general biology, we now have the ability to take apart the cell and look into the different compartments, really look in and know what proteins are there, what they look like, whether they look differently in one compartment than they do in another, what their phosphorylation stages are, and whether they're turned on and when they are turned off."5 "This [research] just was not possible previously, because we didn't have the throughput and the sensitivities, and so it had to be studied indirectly before," he continues. "The promise of proteomics is to study the proteins directly at the cellular level so they're not only in the genetic screen or only in the computer, but [where] you can study them directly. Now, for the first time, because of the advances in the analysis methods, we can do that. Mass spectrometry allows for analyses of hundreds of proteins very sensitively." The simplified notion of one gene, one protein, one function is disappearing fast, Mann says. "We are learning that a protein can have quite a different function depending on where it's located in the cell." The implications of that knowledge may well be enlightening beyond the imagination. Both Mann and Burns are reluctant to specify exactly what diseases they are currently researching, but Mann does offer this: "We do have quite a focused medical interest in this for various forms of cancer, among other things." "The proteome paves the path for the genome," says Burns. "If one can find the right targets, then one would be in a better position to find the right treatments, and we're on that road now. We're in a phase now where we have to address more complex questions, and having tools like we've never had before helps us enormously with the process of identifying proteins and then understanding more about how the human genome functions as it does, in states of health and disease." If the BioReactor lives up to its promises, as these researchers expect it will, three-dimensional biology will aid in rendering "a truer answer--meaning that the changes you see, for example, in cancer vs. noncancer--would hopefully be more reflective of the tissue stage rather than being a culture effect," says Mann. "Mass spectrometry is quite proven now," he adds.6 "We have already generated a lot of data that we have put through the supercomputers, and hopefully with the BioReactors we can do a lot more." MDS Proteomics is venturing that use of the BioReactor, combined with the capability of mass spectrometry analyses and the supercomputing/bioinformatics capacity of the company to sort through and review the endless streams of data will, in turn, offer up a sum result that is far greater than its technological parts. This approach will fill "one of the critical holes," Mann concludes, on the road to the next generation of drug targets and advanced treatments for various diseases. A.J.S. Rayl (ajsrayl@loop.com) is a contributing editor for The Scientist. References 1. T.J. Goodwin et al., "Three-dimensional culture of a mixed mullerian tumor of the ovary: expression of in vivo characteristics," In Vitro Cellular and Developmental Biology, 33:366-74, 1997. 2. T.J. Goodwin et al., "Coculture of normal human small intestine cells in a rotating-wall vessel culture system," Proceedings of the Society for Experimental Biology and Medicine, 202, 1993. 3. M.A. Epstein et al., "Virus particles in cultured lymphoblasts from Burkitts lymphoma," Lancet, 1:702, 1964. 4. J.B. Fenn et al., "Electrospray ionization for mass-spectrometry of large biomolecules," Science, 246:64-71, 1989. 5. A. Pandey, M. Mann, "Proteomics to study genes and genomes," Nature, 405: 837-46, 2000. 6. W. Blackstock, M. Mann, editors, "Proteomics: A Trends Guide," Trends In Biological Sciences, Elsevier, Summer 2000.

On the Fast Track in Functional Proteomics

Interested in reading more?

Become a Member of