Clot-Buster Built From Scratch

After a heart attack or stroke, patients are often given a drug called heparin to prevent blood clots. Now scientists have assembled a synthetic abridged form of heparin that, in animals, appears to work better than the real thing.

A complex chain of sugar molecules extracted from pig intestine, heparin binds to antithrombin, heightening the protein's ability to mop up the enzyme thrombin, which normally kick starts coagulation. But heparin also attaches to a substance called PF4, which in certain patients can trigger new clots.

In today's Nature, scientists from the pharmaceutical company Sanofi in Toulouse, France, report that they have stitched together a heparin substitute. Maurice Petitou started with a sugar chain and tacked on the exact five sugars that heparin uses to grapple onto antithrombin. On the other end, he stuck six negatively charged sugars designed to attract thrombin molecules. With a molecular tether in between six uncharged sugars, the synthetic compound is only 17 sugars long, compared to heparin's hundreds.

That small size has a big payoff. The runt, nicknamed compound 12, has no electrically charged stretch long enough to bind PF4, says Petitou. In test tube experiments, he found that compound 12 did not cause platelet coagulation in the blood of patients who suffered the PF4 side-effect. It also appears to work better than heparin. Petitou treated lab rats with either compound 12 or standard heparin, then tried to induce blood clots. The new drug turned out to be up to 10 times more powerful than heparin at preventing new clots. The researchers are now planning clinical trials.

Other heparin experts are elated. Pierre Sinay, an organic chemist from the Ecole Normale Supérieure in Paris, points out that a synthetic drug will remove the theoretical risk of pathogens stowing away in heparin preparations from pigs and causing human disease. "This is a very nice and important piece of work," Sinay says. He thinks that industrial synthesis of the new compound should be feasible.