How might cannabinoids influence sexual behavior?

M and hashish are widespread drugs of abuse that contain D9-tetrahydrocannabinol (THC), a bioactive ingredient best known for its psychotropic effects. Remarkably, THC also produces multiple nonpsychotropic effects: for example, analgesia, hypotension, modulation of bronchospasm, and reduction of inflammation (1–6). That THC also influences sexual behavior was clearly demonstrated for the first time in the early 1980s; yet the precise molecular mechanism of this effect has remained unsolved. In this issue of PNAS, Mani et al. (7) revisit these seminal experiments. They identify the molecular target by which THC affects sexual behavior, and unveil a remarkable operative cross talk mechanism between THC and the progesterone and dopamine signaling pathways, which were already known to play a central role in reproductive behavior. The chemical synthesis of THC was first described in 1964 (8). In the following years, researchers used synthetic THC to study its diverse biological effects. Whether these effects occurred through receptors or simply by changing plasma membrane fluidity was unknown. One early hypothesis, which has now been abandoned, was that THC could directly bind to the estrogen receptor, thus competing for the sequence of events initiated by estrogen (9, 10). Although controversial, this hypothesis had the advantage of stimulating research to ascertain a link between cannabinoids and sexual behavior. An important step forward in understanding the biological effects of cannabinoids was made in 1988 when Howlett and colleagues (11) used a high-affinity radioactive cannabinoid ligand, [3H]CP-55,940, to demonstrate the existence of a specific cannabinoid binding site in cell membranes of rat brain. This discovery was shortly followed by the molecular cloning and sequencing of the cannabinoid CB1 and CB2 receptors (12, 13). Both receptors are seven transmembrane Gi/ocoupled receptors that display distinct patterns of tissue expression. CB1 receptors are abundantly expressed in the CNS and to a lesser extent in the periphery, whereas CB2 receptors seem to be exclusively expressed by immune cells (14). It is well known that activation of CB1 receptors modulates the release of various neurotransmitters, which could account for the effects produced by THC on higher cognitive functions (15, 16). Engagement of CB2 receptors expressed by circulating macrophages reduces their immune response, which might account for the antiinflammatory effect produced by THC intake (6, 17, 18). An influence of THC on reproductive behavior has been suspected for at least 30 years. Studies carried out in the 1960s reported that chronic oral administration of marijuana resin somehow ‘‘significantly reduces fertility’’ in rats (19). THC can influence reproductive behavior by acting at multiple levels. In males, THC suppresses spermatogenesis, reduces the weight of reproductive organs, decreases the concentration of circulating hormones (such as testosterone) in plasma, and affects some components of sexual behavior. In females, THC prolongs the estrous cycle and decreases the proestrous surge of luteinizing hormone inhibiting ovulation. On the other hand, if THC facilitates sexual behavior: Where does it act? Mani et al. (7) address these questions by using ovariectomized rats and quantifying lordosis quotient, one of the well characterized components of sexual receptivity. The inhibitory effect of ovariectomy on lordosis is quite dramatic; however, complete receptive behavior can be restored by exogenous hormone administration, such as estradiol benzoate, an estrogen receptor agonist. Ovariectomized animals treated with high doses of THC alone fail to show lordosis, but relatively low doses of THC significantly increase the lordosis primed by estradiol benzoate. By using antagonists against either CB1 or CB2 receptors, namely SR141617A and SR14528, Mani et al. (7) demonstrate that the effect produced by THC on sexual behavior occurs through engagement of CB1 receptors. This finding is consistent with the fact that CB1 receptors are expressed in the hypothalamus, in particular at the level of the ventromedial nucleus (20, 21). Future studies using CB1 knockout mice could unambiguously demonstrate the involvement of CB1 receptors in the THC-induced lordosis (22, 23). At this point, however, it was necessary to address the noninvolvement of CB2 receptors, as at least one study suggested their presence in the CNS (24). Indeed, Mani et al. show that SR14528 does not antagonize the effect produced by THC on lordosis. A bell-shaped curve response is often observed with in vivo administration of THC. In this study, Mani et al. (7) give an interesting rationale for this phenomenon observed on lordosis quotient. At high concentrations, such as 400 ng injected i.c.v., THC reduces motor activity. Thus, the smaller lordosis quotient observed at high concentrations of THC might be attributed to the fact that the females are less mobile, which is obviously important for this behavior. It has been shown that the estradiol benzoate-induced lordosis is also increased by progesterone and dopamine (25). The temporal pattern of hormone levels, as well as the behavioral receptivity in intact or ovariectomized estradiol benzoateand progesteroneprimed rats is consistent with an important role for both progesterone and estrogen in the control of sexual receptivity. Interestingly, a molecular mechanism has been proposed for this cross talk between progesterone and estrogen, in which progestin receptors could directly interact with estrogen receptors to activate MAP kinase (26). Dopamine is also a crucial part of this cross talk mechanism (25), acting through D1B (also known as dopamine D5 receptors) (27, 28). Mani et al. (7) push this idea one step further and explore the possibility that CB1 receptors