Not too long ago, a drug called rimonabant seemed to be just such a magic bullet. Rimonabant blocked the receptors in the body that are sensitive to cannabinoids like THC. There are two known varieties of cannabinoid receptors, by the way, CB1 and CB2, though they stand among other multi-purpose receptors that are sensitive to a number of things including certain cannabinoids. The problem with rimonabant was that it was something of a pharmacological sledgehammer, blocking CB1 receptors throughout the brain and body. The brain relies of cannabinoid signaling for a number of things including mood, and perhaps not surprisingly rimonabant was associated with severe depression in a number of patients. For this reason, it was pulled in most markets.
The trail needn’t go dead just because of this setback, however. Numerous studies have suggested that blocking peripheral (non-brain) CB1 receptors is sufficient to reduce food intake and abrogate many of the metabolic disturbances of obesity. Since it was blocking CB1 in the brain that was causing all of the psychological problems, perhaps the solution is to develop a drug that only works in the periphery. And today’s paper concerns just such a drug!
The drug in question is a novel CB1 antagonist, AM6545. This drug blocks the CB1 receptors, but for whatever reason it has limited ability to penetrate the blood brain barrier and thus affect the central nervous system. The first major order of business, having first characterized the drug and its pharmacodynamics, is to see whether it has the same pitfalls of other, similarly acting drugs. Since cannabinoid antagonists sometimes induce nausea (agonists, by the way, are great for alleviating the nausea induced by chemotherapy). Now rats cannot vomit (they are thus very cautious about trying new food), so the best behavioral surrogate appears to be conditioned gaping and conditioned taste aversion. In conditioned gaping, rats are essentially taught to pair the taste of some innocuous substance (saccharin) with the physiological effects of some possibly nocuous substance (the antagonists, and sometimes LiCl as a positive control). Should any of the substances actually induce nausea, the rat will “gape” in response to the saccharin; I am assuming gaping resembles gagging. AM6545 did not induce conditioned gaping, meaning that it probably does not induce nausea. The higher dose of the drug did, on the other hand, reduce food intake. Since that same dose did not induce nausea, we can probably rule that out as a cause.
Interestingly, if the 10th cranial nerve (the vagus) is cut, the effect of AM6545 on food intake remains. This is interesting, because the vagus nerve is one of the main connection points between the brain and viscera, aside from the manifold hormonal and somatosensory signals that the two exchange. Often times sectioning the vagus nerve destroys the effect of a drug, and if this happens, it’s a safe bet that some portion of that drug’s effect relied on neural conduction through the vagus rather than something blood borne (though it’s never so simple).
Things get weird at the end when the authors tried the drug on CB1, CB2, and CB1&2 knockout mice. These mice are missing those receptors, so presumably if the drug still affects food intake in them, then its actions on the cannabinoid receptors cannot be the entire story. The drug worked to reduce food intake on CB1 knockouts, CB2 knockouts, but “not” on dual knockouts. The effect in the dual knockouts was just barely non-significant (p=0.08, the cutoff is supposed to be p=0.05). The authors say that it’s straight up non-significant, but I don’t really buy this interpretation. There’s probably a good chance that AM6545 does act somewhere else to regulate food intake, but it’s not at all clear where that might be.
The drug needs some follow up investigation, preferably by other labs who don’t have patent interests in the mix. Importantly, the nausea related endpoints here are difficult to interpret. An ideal follow-up might include measure of anhedonia and perhaps even sickness behavior and “depression” (such as it is in rodents) to rule out some of the problems we’ve seen with other CB1 antagonists. Still, drugs that play nice and keep out of the CNS are always preferable (provided that they don’t need CNS access to work), so there could be some potential here.
Cluny, N., Vemuri, V., Chambers, A., Limebeer, C., Bedard, H., Wood, J., Lutz, B., Zimmer, A., Parker, L., Makriyannis, A., & Sharkey, K. (2010). A novel peripherally restricted cannabinoid receptor antagonist, AM6545, reduces food intake and body weight, but does not cause malaise, in rodents British Journal of Pharmacology, 161 (3), 629-642 DOI: 10.1111/j.1476-5381.2010.00908.x