Switching on a subtype of the receptor that binds cannabis, the active ingredient in marijuana, can suppress inflammation -- suggesting a new and particularly promising target to treat autoimmune problems such as multiple sclerosis and the damage caused by immune cells after a stroke. But hotly contested evidence for whether or not this cannabinoid receptor is expressed on neurons may limit the potential for pursuing that target in the search for new medicines.
Understanding the receptor's role in damaged neurons and autoimmunity "has pretty important therapeutic implications," said Ken Mackie, a neuroscientist at Indiana University in Bloomington who studies cannabinoid receptors in the brain.
It's commonly accepted that marijuana's "high" stems from cannabis binding to one type of cannabinoid receptor -- called the CB1 receptor -- which is widespread on neurons in the brain. Scientists have tied CB1 receptors to a host of health problems, from depression to obesity to cardiovascular disease, as well as therapeutic benefits such as pain and nausea relief. (A drug called Marinol, a synthetic version of the active ingredient in marijuana, was approved for reducing nausea and boosting appetite in cancer patients in 1985). Although some researchers say cannabis has medicinal properties, FDA approval is a long shot for Schedule 1 compounds such as cannabis -- which the government outlaws on the grounds that they are harmful and have no legitimate medical uses.
A decade or so ago, researchers uncovered a second variant of cannabinoid receptor, called the CB2 receptor, on immune cells and found that they seem to play a role in immune response. Those receptors have been thought to be largely absent from the brain, meaning chemicals that selectively target these sites could sidestep cannabis's psychoactive effects -- and might skirt approval problems. But one group of researchers has shown that CB2 receptors are widespread in neuronal tissue. If true, it could put a wrinkle in hopes for developing a CB2 targeting anti-inflammatory without the unwanted side effects.
Ron Tuma, a physiologist at Temple University in Philadelphia, PA, has been studying the role of CB2 receptors in multiple sclerosis and stroke. CB2 receptors are widely distributed on immune cells, and Tuma's group wanted to see if activating them could quiet inflammation in mice. In June, they published a paper in Microvascular Research showing that selectively targeting CB2 receptors reduces injury and tissue death after ischemic stroke. The effect is "totally dramatic," said his sometime-collaborator Toby Eisenstein, a microbiologist and immunologist at Temple who studies CB2 receptor selective agonists.
Meanwhile, Michelle Glass, a pharmacologist at the University of Auckland in New Zealand, found that CB2 receptor activation can shield neurons, likely by tamping down the immune cells called microglia from activating an inflammatory response. CB2 receptors are especially promising drug targets, she said, because unlike other immuno-suppressive targets, studies suggest CB2 receptors are mostly quiescent at basal levels, reducing the potential for side-effects. Keith Sharkey, an enteric neurobiologist at the University of Calgary in Canada, agreed. "My sense is that there is not very much [CB2] in the brain under baseline conditions," but that the receptor expression gears up during trauma or inflammation, he said.
The hope, researchers say, is that a drug that targets CB2 receptors would be free of the mind-altering effects of their Schedule 1 cousins, since it's generally thought that CB2 receptors are mostly absent from neurons. Eisenstein estimates that "99% of [receptors] are on cells in the immune system." In a 2005 Science paper, Sharkey and colleagues found a low density of CB2 receptors on a few neurons in the brain stem and the cerebellum, as well as on immune cells like microglia in the brain, consistent with the idea that CB2 receptors are mostly absent from brain tissue.
But one researcher's findings throw a wrench in the works. Microbiologist Emmanuel Onaivi at William Paterson University in Wayne, NJ, and his collaborators showed last year that stressed mice treated with a CB2 receptor agonist were more likely to crave alcohol; in a 2006 study, he reported that Japanese men with a variant form of CB2 genes had higher rates of alcoholism and depression. Both studies hint at a role for the receptor in the brain. Onaivi's most controversial finding, however, came from two papers: one published in 2006 in Brain and Research, and the other published in the New York Academy of Sciences in 2008. Using tagged antibodies, his group detected the widespread presence of CB2 receptors on neurons. Taken together, his results suggest that CB2 selective agonists may still have the psychological effects that make CB1-targeting compounds clinically unusable -- removing some of the luster of CB2 receptors as therapeutic targets.
To date, no other labs have been able to reproduce Onaivi's findings. Multiple groups have probed for the messenger RNA or genes that express CB2 receptors on neurons, to no avail, Glass said. "And yet [Onaivi's group is] managing to publish all this immunohistochemistry showing that it's on every brain cell," she said.
Onaivi argues that earlier studies missed CB2 expression in neurons because they used crude Western blot and gene expression probes. Other CB2 researchers disagree. "I don't think Emmanuel has done anything particularly different than anyone else," said Mackie, who studies both CB1 and CB2 receptors. Mackie believes the real discrepancy occurred because Onaivi's 2006 Brain and Research paper didn't include "fastidious" controls. "There've been several commentaries published, laying out what are the necessary controls. Unfortunately some journals don't adhere to them," he said.
For instance, Onaivi has looked for multiple CB2 receptor variants in wild type brains, but has used only a single group of CB2 knockout mice as a control group, Mackie said. In addition, the group used a peroxidase stain that can produce a false positive when testing for the presence of receptors, and didn't run wild type and knockout samples side-by-side, Mackie explained. The longer brain slices are incubated in peroxidase, the likelier it is for them to show some evidence of receptors, regardless of whether they're actually there, he said -- adding that he has found that peroxidase staining appears to show the presence of CB2 receptors even in knockout mice lacking the CB2 gene.
"Some of the antibodies available in particular for the CB2 receptor don't validate out quite as cleanly as one would like," Sharkey added, so they may not have been specific to CB2 receptors, reacting as well with other compounds.
Definitively determining whether CB2 receptors localize to the brain would require running several studies in parallel in both knockout and wild type mouse brain, and leaving the stain on for the same exact amount of time, Mackie said. You'd also want to use a blinded procedure to process and analyze the brain slices -- which Onaivi's group did not do. Without such a rigorous approach, there is a danger of vastly overestimating the receptor's prevalence, Mackie explained. "Sure, there may be some CB2 receptors expressed by these cells, but there's nowhere near the level" Onaivi's group showed with immuno-staining, he said. On top of showing the presence of receptors, studies would also need to demonstrate that neurons express high levels of mRNA for CB2 receptors, and that these same cells also have a lot of CB2 proteins, he said.
But Onaivi thinks he has an alternative explanation for why others have seen little trace of CB2 receptors in neurons when he has found them everywhere. Using genomic analysis, he found "two different isoforms" of CB2 receptors, he said. He presented his preliminary data at the Society for Neuroscience Meeting last November, and published them in June in Genes, Brain and Behavior. "One [isoform] is more expressed in the testes and in the brain, than in the spleen where they were originally discovered," he said. As a result, antibodies known to react with gut CB2 may not be detecting the different form expressed in the brain. "We are asking the question why, for years, people never found CB2 in the brain, and I think we now have the answers," he said.
Mackie concedes that different isoforms of CB2 exist, each with distinct genetic sequences, but he's "doubtful" that could explain the discrepancy. "But I don't have data to say one way or another," he said.