A big circuit model.
The field's identity crisis did not dampen the enthusiasm of the more than 185 participants at Metabolic Profiling: Applications to Toxicology and Risk Reduction, a conference held 14-15 May 2003 at the NIEHS in Research Triangle Park, North Carolina. The meeting brought together molecular biologists, analytical chemists, toxicologists, clinicians, nutritional scientists, and computational biologists from government, academia, and industry to assess the current state of the science in the emerging area of metabolic profiling, and to define its potential applications to the health sciences. It was organized by the NIEHS, the National Institute of Diabetes and Digestive and Kidney Diseases, the NIH Office of Rare Diseases, the U.S. Food and Drug Administration, Paradigm Genetics, and Waters Corporation.
In metabonomics, biofluids and tissues are analyzed to determine the metabolites present, both in homeostasis and when the organism has been affected by factors such as environmental exposures. "It is both a systems biology and a dynamic approach," said speaker Hector Keun, a postdoctoral researcher at Imperial College of the University of London, "as metabonomics analysis can provide a description of the integrated physiological behavior of an entire living system across time."
Metabolic profiling is a hugely complex undertaking, generating huge amounts of data to be analyzed and mined for nuggets of significant information about metabolic pathways and networks, novel biomarkers, and how metabolites interact not only with genes and proteins, but also with environmental, nutritional, and lifestyle factors.
The integration of this dizzying array of variables holds the potential to tell researchers a great deal about human health and disease etiology, with translational rewards already emerging in diagnostics and drug targeting, development and safety screening. "Metabolism is phenotype," said Steve Watkins, president of Lipomics Technologies of West Sacramento, California. "It integrates all the factors you'd want to know about, from nutrition to environment to genetics; it's really the only way to assess how you're doing as an individual. If you want to assess your current state of health, you have to [study] the metabolome."
Lipomics Technologies has developed a quantitative assay that measures several hundred lipid metabolites from biosampies. The assay is used to study the role of lipid metabolism in disease and develop diagnostic profiles of drug safety and efficacy. A Research Triangle Park company, Metabolon, has identified a metabolic signature for amyotrophic lateral sclerosis. This profile holds great promise for identifying disease-related biochemical and signaling events, diagnostic markers, and potential therapies.
Metabolic profiling technology is booming in both the analytical and computational realms, but the metabolome is only one element of the entire picture. "If you want to learn about pathways," said Trey Ideker, Pfizer Computational Biology Fellow at the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts, "you're going to have to characterize metabolites, genes, and proteins, and for each we have interactions and levels. That's six different things we're going to have to integrate together in a big circuit model, or blueprint, of the cell, and of every cell type, and of every tissue type." The key to meeting this huge challenge, he said, is data mining.
Ideker described work being conducted by his group in developing a computational "scaffold" approach to modeling complex cellular interactions. The modeled scaffolds are mined to reveal a hierarchy of signaling, regulatory, and metabolic pathways. Keun told participants of a system to analyze metabolic profiling data being constructed by the Consortium on Metabonomic Toxicology, a project of scientists at six pharmaceutical companies and Imperial College. The consortium's prototype has already shown promise in elucidating the nature of the relationships between traditional toxicological end points and metabolic descriptors, helping to validate the role of metabonomics data in predictive and mechanistic toxicology.
Scientists expressed excitement at the potential contributions offered by metabolic profiling to toxicology, toxicogenomics, and risk assessment, reduction, and prevention. "There is some incredible technology that can be used to assess risk, link exposure with disease etiology, and reduce the uncertainty of risk in the population," said William Suk, director of the NIEHS Center for Risk and Integrated Sciences. "This is just the beginning."
Kenneth Ramos, chairman of the Department of Biochemistry and Molecular Biology at the University of Louisville Health Sciences Center and toxicogenomics editor for EHP, said, "Metabolic profiling can be a much more effective way of communicating risk and of having an impact on risk reduction strategies in the future, because a metabolite is something people relate to and have a better grasp on than genes. And of course, metabolites ultimately being a reflection of the genomic network, it's probably going to be quite significant."
Participants agreed that perhaps the most consequential future direction for the field--where it will ultimately yield its most profound benefits--lies in the integration of metabolic profiling data with genomic and proteomic data. "One of the most important take-home messages from this meeting," said Ramos, "is the recognition that metabonomics is a systems biology-integrated approach that's going to give us meaningful answers to vitally important questions about ecological and human health."