A new endothelial target for cannabinoids.

The article by Offertaler and colleagues in this issue contains evidence pointing to a novel vascular site of action of cannabinoids. The precise vascular actions of endogenous cannabinoids have been surprising complex and controversial (see Hillard, 2000; Ralevic et al., 2002; Randall et al., 2002). In 1995, Ellis et al. demonstrated that anandamide caused cerebrovascular vasodilatation in the rat via the release of prostanoids. The following year, Randall et al. (1996) demonstrated that anandamide was a relaxant in the rat mesenteric vasculature but that this did not involve cyclooxygenase products. Since these initial reports, a variety of mechanisms has been proposed for the vascular actions of anandamide. These have included the release of nitric oxide (Deutsch et al., 1997), activation of potassium channels (Randall et al., 1996, 1997), conversion to metabolites of arachidonic acid (Pratt et al., 1998), the activation of gap junctions and presumed mediation via the endothelium-derived hyperpolarizing factor (EDHF) (Chaytor et al., 1999), blockade of calcium channels (Gebremedhin et al., 1999), and the involvement of prostanoids (Fleming et al., 1999). Perhaps most intriguing was the 1999 proposal that anandamide was in fact an endogenous vanilloid that acted via sensory nerves to release transmitters such as calcitonin gene-related peptide, which mediated the vasorelaxation (Zygmunt et al., 1999). However, almost without exception each proposal or demonstration has been accompanied by subsequent reports that did not support the initial findings, and this may reflect both species and methodological differences between studies. Two particular areas of uncertainty were the involvement of the endothelium and the role of currently recognized cannabinoid CB receptors. Much of the latter stems from the uncertainty surrounding the selectivity of the “classic” cannabinoid CB1 receptor antagonist SR141716A. Having said that, the strongest evidence for the involvement of cannabinoid CB1 receptors came from the absence of cardiovascular responses to anandamide in CB1 knockout mice (Ledent et al., 1999). In 1999, Wagner and colleagues (Jarai et al., 1999) proposed that anandamide acted, in part, via an endothelial anandamide receptor in rat mesenteric vessels. This was based on the observation that relaxation to anandamide was partly sensitive to both removal of the endothelium and SR141716A, but when the endothelium was removed, the sensitivity to the antagonist was lost. This led to the proposal that anandamide acted on a cannabinoid receptor that was sensitive to SR141716A but was not the CB1 receptor and was termed the “anandamide receptor”. An additional observation was that the exogenous cannabinoid, -tetrahydrocannabinol, did not cause vasorelaxation. Subsequent work by that group, demonstrated that the endothelial cannabinoid receptor was also activated by the neurobehaviorally inactive “abnormal cannabidiol” (abn-cbd), which caused vaorelaxation (Jarai et al., 1999). One possibility to arise from the identification of the SR141716A-sensitve endothelium-dependent component is that anandamide acts in part via EDHF and that SR141716A is acting via inhibition of EDHF activity (e.g., through blockade of myoendothelial gap junctions) (Chaytor et al., 1999). However, Mukhopadhyay et al. (2002) demonstrated that the endothelium-dependent component was G-protein-coupled and mediated via nitric oxide whereas the endothelium-independent component was due to activation of vanilloid receptors, at least in rabbit aortic rings. In the present issue, Kunos and colleagues (Offertaler et al., 2003) provide further characterization of the “endothelial anandamide” receptor. Specifically they report that a novel cannabidiol analogue, O-1918, opposes the relaxant effects of anandamide and abn-cbd, the hypotensive effects of abn-cbd and the phosphorylation of p42/44 MAP kinase induced by abn-cbd in endothelial cells. These actions of O-1918 are independent of CB1 and CB2 receptors, and this led the authors to conclude that O-1918 was a selective antagonist of the endothelial anandamide receptor. The authors point out that it is unlikely that O-1918 is acting at a different recognition site on CB1 or CB2 receptors because previous studies have shown that abn-cbd causes relaxation in vessels from