May 12, 2006 | How’s your biochemistry expertise? Probably not so great, suggests John Ryals, CEO of six-year-old Metabolon. Pursuit of molecular biology in the biopharmaceutical industry has drained talent (and dollars) from biochemistry, he believes, while the resulting proliferation of genomic and proteomic research hasn’t yet paid off.

Metabolomics — if you do it well — is often a better approach, says Ryals, and Metabolon is betting big on that idea. In March, the company was awarded a U.S. patent (#7,005,255) for “Methods for drug discovery, disease treatment, and diagnosis using metabolomics” based on work by co-founder and Duke University researcher Rima Kaddurah-Daouk.

The patent, first filed in January 2003, claims a method for identifying small molecules relevant to amyotrophic lateral sclerosis. That’s just the first of a volley of claims to broaden the patent, says Ryals. He joined the company in 2002, already an experienced fund-raiser having led Paradigm Genetics to an IPO as CEO, and having held research positions at Novartis and Ciba-Geigy.

Metabolon’s basic premise is that it can identify biomarkers based on small molecules using traditional mass spec, some proprietary biochemical techniques, sophisticated homegrown software, and strength in biochemistry.

“The group here has been involved in the technology since the beginning of metabolomics, back in 1996 and 1997. We have a vision of not just looking at ions but actually getting at molecules present in a sample. We identify every molecule, give it a name, and understand how it relates in biochemical pathways. So we spend a lot of time on software and software engineering. We don’t do a lot to the mass spec itself,” he says.

Kaddurah-Daouk’s credentials are impressive. Trained at Johns Hopkins, Massachusetts General Hospital-Harvard Medical School, and MIT, she has written extensively on “energy impairment disease,” is president of the Metabolomics Society, and maintains a research affiliation with the Dana-Farber Cancer Institute.

Currently, the company headcount is roughly 30 with two-thirds in research, including six in software engineering and two statisticians. Much of the technology used is plain vanilla. Metabolon chose Thermo mass spec machines, but any good machine would have sufficed. A Linux multiprocessor server runs analysis software. A large Dell server runs the Oracle database and LIMS.

Biopharma Wishes
Customers have the usual wish list. “They’ll want biomarkers for a drug. They’ll want to understand if there are off target effects. Can we find biomarkers for a disease that will allow them to get a better stratification of patients? We can probably flip a project like that in two to three months,” says Ryals, who quickly adds that he believes the technology can be applied in several industries.

Metabolon has published with Bristol-Myers Squibb on AIDS work (HIV Protease Inhibitors and the Hepatic Metabolome), and Ryals says the company has worked with eight of the top 10 biopharma — not bad considering the platform has only been in “commercial mode” for about a year and a half.

CSO Michael Milburn is a recent hire (four months ago). He’d been senior VP of research and corporate development at Sirtris Pharmaceuticals, working on metabolic disease and neurodegeneration projects, and had stints at Plexxikon, Structural Genomix, and GlaxoSmithKline.

“The technology has gotten sufficiently robust to the point where we are working with so many companies in these various diseases that having someone onboard with a little bit more pharma/biotech experience made sense,” he says.

Milburn emphasizes the need to stay ahead of the biochemistry curve: “One way we’ve done that is to go out and exclusively sign up, on a consulting basis, some of whom we believe to be the most clinically relevant biochemists and whole body physiology biochemists. That’s an area we see as a strong strategic component of the company moving forward.”

Given FDA’s growing vigor in pressing biopharmaceutical companies to better understand what their drugs do, the outlook for Metabolon seems potentially strong.