A major milestone in the field of cannabinoid pharmacology was the discovery that many effects of Δ9-THC, the main psychoactive constituent of cannabis, are mediated by receptors, two types of which have so far been identified. These are the CB1 and CB2 receptors, which both signal through Gi/o protein (reviewed in Howlett et al., 2002; Pertwee, 2005a, 2006). A second equally momentous discovery was that mammalian tissues can synthesize and release compounds that activate cannabinoid receptors, the first of these ‘endocannabinoids’ to be identified being N-arachidonoylethanolamine (anandamide) and 2-arachidonoylglycerol (Devane et al., 1992; Mechoulam et al., 1995; Sugiura et al., 1995). Together with their receptors, these and other more recently discovered endocannabinoids (reviewed in Pertwee, 2005b) form a part of the ‘endocannabinoid system’.

There is growing evidence that some but not all ligands for CB1 and/or CB2receptors target additional receptors, either established or putative (reviewed in Pertwee, 2004, 2005a). One such receptor that has been attracting particular interest among cannabinoid scientists in recent times is the orphan receptor, GPR55. This interest was initially prompted by patents lodged by GlaxoSmithKline (Brown and Wise, 2001) and AstraZenica (Drmota et al., 2004), which claim that this receptor is activated by several CB1/CB2 receptor ligands, and was further fuelled by two presentations made at the 2005 meeting of the International Cannabinoid Research Society by scientists from each of these pharmaceutical companies (Brown et al., 2005; Sjögren et al., 2005) and by a review article based mainly on these presentations and on the GPR55 patents (Baker et al., 2006). Frustratingly, however, there was a total absence of any peer-reviewed data supporting the claim that GPR55 is a cannabinoid receptor. Happily, this need has now been met by Ryberg et al. (2007) in this issue of the BJP.

In their paper, Ryberg et al. (2007) describe the cloning and protein sequences of human, mouse and rat GPR55 and report that the CB1/CB2 receptor ligand, [3H]CP55940, exhibits specific binding to human GPR55. Furthermore, they present evidence that this receptor can be activated by the established CB1/CB2receptor agonists, Δ9-THC, HU-210, CP55940, anandamide and 2-arachidonoylglycerol, and by the CB1-selective agonist noladin ether. These agonists were all found to share the ability to stimulate [35S]GTPγS binding to GPR55 with EC50 values in the low nanomolar range (Table 1) and Emax values ranging from 73 to 100%. The finding that anandamide and 2-arachidonoylglycerol activated GPR55 raises the possibility that these compounds may serve as endogenous agonists not only for CB1 and CB2receptors but also for GPR55. Also described is the heterogeneous distribution pattern in mouse of GPR55 mRNA, the levels of which were the highest in adrenal tissue, ileum, jejunum, frontal cortex and striatum. Mouse brain contained less GPR55 mRNA than CB1 mRNA in many of the areas investigated, and this difference is most apparent in cerebellum and hippocampus.

Table 1

EC50 values of GPR55 agonists for stimulation of [35S]GTPγS binding to membranes obtained from cultured cells transfected with human GPR55, CB1 or CB2 receptors (from Ryberg et al., 2007)

Several pharmacological differences between GPR55 and CB1 or CB2 receptors are reported by Ryberg et al. (2007) (see Table 1). Particularly notable are their findings that the endocannabinoid, 2-arachidonoylglycerol, displays more than 170 times greater potency as an agonist at GPR55 than at the CB1 or CB2receptor and that Δ9-THC has greater efficacy (though not greater potency) as an agonist for GPR55 (Emax=92%) than for CB1 or CB2 receptors (Emax=61 and 67% respectively). Also noteworthy are their findings that GPR55 couples to Gα13 and not to Gi/o or Gq, that this receptor is activated by AM251 at concentrations at which it behaves as an antagonist/inverse agonist at the CB1receptor (reviewed in Pertwee, 2005a), that CP55940 is 25 times and HU-210 is 130 times less potent at activating GPR55 than at activating CB1 receptors, that CP55940-induced activation of GPR55 is antagonized by cannabidiol with an IC50 (445nm) that is below any concentration at which this plant cannabinoid displaces [3H]CP55940 from CB1 or CB2 receptors (Thomas et al., 2007) that GPR55 is neither activated nor antagonized by the CB1/CB2 agonist WIN55212-2 or by the CB1-selective antagonist/inverse agonist, AM281, and that [3H]WIN55212-2 lacks significant GPR55 affinity. Ryberg et al. (2007)also report that GPR55 is activated by the endogenous ligands, palmitoylethanolamide and oleoylethanolamide, which as expected, were found to lack significant activity as CB1 or CB2 receptor agonists in their experiments. Similarly, virodhamine displayed markedly greater potency as a GPR55 agonist than as a CB1 or CB2 agonist and, in addition, was found to exhibit particularly high efficacy at GPR55 (Emax=160%). They also found GPR55 to be activated in the [35S]GTPγS-binding assay by O-1602 and abnormal cannabidiol. Neither of these compounds display detectable activity as a CB1 or CB2 agonist (Table 1), though both have been reported previously to target the putative abnormal-cannabidiol receptor, as has cannabidiol (reviewed in Pertwee, 2004, 2005a).

The findings reported by Ryberg et al. (2007) raise a number of important questions. First, to what extent are the pharmacological profiles of some CB1 or CB2 receptor agonists and antagonists shaped by their apparent ability to also activate GPR55? This is a question that will be particularly important to answer for AM251, because it is widely used as a research tool with which to block CB1 receptors, for 2-arachidonoylglycerol because of the greater potency this endocannabinoid exhibited at GPR55 than at CB1 or CB2 receptors (Table 1), and for the psychoactive plant cannabinoid, Δ9-THC, because of the greater efficacy it displayed at GPR55 than at CB1 or CB2 receptors (see above). Second, does rimonabant (SR141716A), which is now a licensed medicine (Acomplia, Sanofi-Aventis, Paris, France), a CB1 receptor antagonist/inverse agonist (reviewed in Howlett et al., 2002) and a structural analogue of AM251 and AM281, activate GPR55 (as reported for AM251) or fail to target this receptor (as reported for AM281)? Third, anandamide and 2-arachidonoylglycerol were found to activate GPR55, but to what extent are GPR55-mediated effects modulated by these or other endogenous compounds when they are released in vivo? Also, what are these effects and which endogenous compounds contribute most to any such modulation? Fourth, what physiological and pathological processes are modulated by GPR55, in what cell types and neuronal subpopulations is it most highly expressed, and will the signs of cannabinoid-induced activation of GPR55 be detectable in mice from which CB1 and/or CB2 receptors have been genetically deleted? Above all, what is the role of GPR55 in health and disease? Does this include the regulation of vascular tone and immune-cell migration, as suggested by Ryberg et al. (2007), and will any important clinical applications be discovered for cannabinoids that behave as selective agonists or antagonists for this receptor or for compounds that modulate the tissue levels of any endogenously released GPR55 receptor ligands?

In conclusion, Ryberg et al. (2007) present convincing evidence that GPR55 is targeted by a number of cannabinoids and that, intriguingly, it is activated as potently and with greater efficacy than the CB1 receptor by the main psychoactive constituent of cannabis (Δ9-THC) and much more potently than CB1 or CB2 receptors by the endocannabinoid, 2-arachidonoylglycerol. Now, it will be important to characterize the pharmacology of GPR55 and its ligands more completely through both in vitro and in vivo research, not least so that a conclusive decision can finally be made on whether GPR55 should indeed be accepted as a new member of the cannabinoid receptor clan.