In June 2006 pharmaceutical giant Sanofi-Aventis began selling a new weight-loss drug called rimonabant in Europe. Rimonabant worked in part by reducing appetite, and the company claimed it could also treat addiction, harmful cholesterol, and diabetes. Lab tests even suggested the drug produced healthier sperm. But within six months, the company had received more than 900 reports of nausea, depression, and other side effects.
By the following summer, the U.S. Food and Drug Administration had rejected rimonabant, noting that relative to a placebo, patients taking it were twice as likely to contemplate, plan, or attempt suicide. The European Medicines Agency soon asked Sanofi-Aventis to address the safety concerns, and on December 5, 2008, the company pulled the drug off the European market.
Rimonabant was a spectacular flop, and yet its lure today is stronger than ever. Researchers worldwide are pursuing novel drugs aimed at the exact same target: the endocannabinoid system, an elaborate network of receptors and proteins that operate within the brain, heart, gut, liver, and throughout the central nervous system. For drug designers, the system’s powerful role in regulating cravings, mood, pain, and memory makes it a tantalizing target. The challenge now is finding sharper, more refined ways to manipulate it without causing the sort of debilitating side effects that derailed rimonabant. “The system is very, very widespread and very effective at a variety of levels,” says neuroscientist Keith Sharkey, who studies the role of endocannabinoids in the gut at the Hotchkiss Brain Institute at the University of Calgary. “It seems to be very important in the body, which is a concern when you develop drugs for it because you will get a range of effects.”
Zheng-Xiong Xi, a pharmacologist at the National Institute on Drug Abuse in Baltimore, explains that the main receptor in the endocannabinoid system, CB1, interferes with brain levels of dopamine, a chemical associated with reward-seeking behavior, pleasure, and motivation. Activating CB1 jump-starts a chain reaction that culminates in an excess of dopamine floating around between neurons. “The dopamine produces a good feeling, a rush, euphoric effects,” Xi says.
A few years ago, Xi was studying this phenomenon in mice, hoping to find a pill to treat addiction to dopamine-boosting drugs such as cocaine. Scientists suspected that rimonabant, which decreases CB1 activity, dampens appetite by decreasing dopamine levels and taking the rush out of eating. Xi was looking for a compound that would have the same effect on cocaine users. Without the high, he theorized, cocaine might lose its appeal.
In studying the endocannabinoid system, Xi tested THC, the active compound in marijuana. THC is thought to produce euphoric feelings by increasing CB1 activity and causing dopamine levels to rise. Instead he saw the opposite effect. “We found that at higher doses it produced a decrease,” Xi says. “So how did this happen?”
When Xi tried THC on mice lacking CB1 receptors, he found the same response: Dopamine dropped. Could THC be acting on the other receptor in the endocannabinoid system, CB2? It was an odd question, since CB2 receptors were not thought to reside in the brain. “For years people didn’t believe that they really existed there,” Sharkey says. But when Xi tested THC in mice without the CB2 receptor, it had no effect at all. CB2 was clearly involved.
Testing the idea further, Xi then switched to JWH-133, a compound designed to latch onto CB2 receptors and decrease dopamine levels. If CB2 receptors were truly absent in the brain, the drug should have no effect on dopamine levels there. Instead, Xi found, they were plunging. In cocaine-addicted mice with unmodified endocannabinoid systems, JWH-133 dramatically reduced the number of times the mice would press a lever for more cocaine. “The drug-taking behavior or drug intake is tremendously decreased,” Xi says. Meanwhile, when he and his team tested JWH-133 on mice without CB2 receptors, the rodents kept going for the cocaine, and their dopamine levels were higher.
This effect on dopamine may explain why rimonabant was dangerous: It blocked the body’s ability to produce a natural high. But Xi’s work, published in Nature Neuroscience, suggests that targeting a different receptor, CB2, makes all the difference. While early tests in rats hinted that rimonabant might have depressive effects, Xi’s group found no evidence of malaise in their mice.