Accelerator Startup, GPC-Rx, Uses Computers to Make Drugs Without Side Effects

11/21/08Follow @xconomy

Accelerator’s newest startup could, if it wanted to, make a version of marijuana that wouldn’t give people the munchies. Seattle-based GPC-Rx isn’t going to do that, but it is using sophisticated 3-D computer modeling to help researchers design other drugs to specifically do their main job, without the usual side effects.

This company has its roots in the lab of William Goddard, the noted Caltech chemist. It’s the second time Accelerator has backed one of his inventions (the first was Seattle-based Allozyne). Accelerator isn’t disclosing the amount of money going into the new company, but it’s on the “upper end” of its usual investments, says president Carl Weissman. That means it’s in the neighborhood of $4 million to $5 million.

This idea is one that only a biotech geek can really love. Goddard’s crew has set up proprietary computer models that drug hunters can use to accurately predict how a snake-like 3-dimensional protein structure will look before, during, and after it’s been hit with a drug. It focuses on a particular type of target called G-protein coupled receptors (GPCRs), which are blocked by common drugs like antihistamines, antidepressants, and, yes, marijuana. Many of these treatments aren’t very good at distinguishing between subtle differences in the appearance of these targets, which is why they cause side effects. It’s been obvious to pharmaceutical companies for years that if they could get a clearer picture of how the targets appear in 3-D, then they could design better drugs.

“With a lock and key, even subtle differences can make a big difference,” says Bard Geesaman, a biotech entrepreneur recruited by Weissman to lead GPC-Rx as chief scientist. “If you’re a pharmaceutical company, you can make a master key, it unlocks a lot of things, but you get side effects.”

Plenty of money is at stake in these subtle differences. Sanofi-Aventis’ weight-loss drug rimonabant (Acomplia) failed to win FDA approval last year because it had a rare side effect of being linked to suicidal thinking. Since drugs that control weight loss, depression, and allergies represent some of the biggest pharmaceutical markets on the planet, there’s no shortage of competition to come up with a better way to hit these targets, Weissman says. “Every major pharmaceutical company has a program looking at GPCR’s,” he says. (My Boston colleague Ryan McBride also recently profiled the Cambridge, MA-based startup Ascent Therapeutics, which is developing a different class of molecules it calls “pepducins” to hit GPC targets inside cells.)

Geesaman, 41, sounds like a rare combination of scientist and businessman for a task like this. He has a doctorate in systems and computational neurobiology from MIT, and a medical degree in internal medicine from Harvard Medical School. He did a stint as a venture partner at MPM Capital scouting healthcare investments, co-founded an MPM-backed pharma company in Japan, and most recently was executive director of life sciences for the X Prize Foundation in Santa Monica, CA. He got to know Weissman in 2001 when they worked together at Cambridge, MA-based Elixir Pharmaceuticals.

So far, GPC-Rx has no practical way to double-check the accuracy of its computer models with certainty through an imaging technique, Weissman says. Instead, GPC will test its ideas by using the computer’s readouts, developing small molecule compounds to specifically block the targets as they appear, and run experiments in petri dishes to see if they really block the chosen targets exclusively, or whether they still are likely to cause side effects by hitting proteins that belong to the same family.

This is all still a long way from clinical trials, so it will be years before researchers will truly know the answer of whether this will be able to yield a big payoff. If they’re right, we may see antihistamines that won’t make you drowsy, or maybe a new breed of weight-loss drugs without any mood-altering side effect. I can only speculate how long it might be before someone comes up with munchie-free-marijuana.

By posting a comment, you agree to our terms and conditions.