Characterisation Of Fenofibrate As An Agonist And Negative Allosteric Modulator Of Cannabinoid Receptors

RS Priestley, SPH Alexander, DA Kendall. University of Nottingham, Nottingham, UK

The seven-transmembrane cannabinoid receptors CB1 and CB2 regulate cellular function via coupling to a number of intracellular pathways. They are activated by endogenous lipid mediators, as well as by phytocannabinoids derived from Cannabis sativa, and a diverse range of synthetic agonists. Several cannabinoid receptor ligands are also agonists of the peroxisome proliferator-activated receptors - nuclear receptors which act as lipid sensors, regulating metabolism and cellular differentiation and which are targets for therapeutics including fibrates and thiazolidinediones. Here, we have investigated the PPARα pro-drug fenofibrate for potential activity at the cannabinoid CB1 and CB2 receptors.

Membranes derived from CHO cells overexpressing the cannabinoid receptors were used in radioligand ([3H]-CP 55,940) and [35S]-GTPγS binding assays to determine the compound’s binding and functional properties. In addition CB1-mediated ERK1/2 phosphorylation and β-arrestin recruitment was measured in intact cells, using “In-Cell Western” and PathHunter assays, respectively.

Initial screening revealed fenofibrate as a cannabinoid receptor ligand, exhibiting nanomolar affinity for both cannabinoid receptor subtypes with ~25-fold CB2 selectivity (pK i CB1 = 6.29 ± 0.90, CB2 = 7.68 ± 0.08). The active metabolite fenofibric acid, as well as other PPAR agonists were inactive with regard to CB1/CB2 receptor interaction. In the GTPγS assay, fenofibrate acted as a near full agonist at the CB2 receptor (pEC50 = 7.71 ± 0.04; E max = 90 ± 2% response compared with the full agonist CP 55,940), and a partial agonist at CB1, (maximum response = 59.5 ± 8.1 % CP 55,940 response), though with a decrease in GTPγS binding at higher concentrations, producing an atypical bell-shaped concentration-response curve. This pattern of response was also present in assays measuring β-arrestin recruitment or ERK1/2 phosphorylation. Studies with the CB1-selective antagonist AM 251 showed the high potency positive response to be CB1-mediated, while the low potency negative response was insensitive to the antagonist. A high concentration of fenofibrate (30 µM) was able to significantly increase the dissociation rate constants for the agonist [3H]-CP 55,940 (kslow without fenofibrate = 1.19 ± 0.20, with fenofibrate = 3.84 ± 0.08 x 10-2 min-1; p<0.01; paired t-test), (kfast without fenofibrate = 0.79 ± 0.14, with fenofibrate = 1.50 ± 0.16 min-1; p<0.05; paired t-test), suggesting it to be a negative allosteric modulator at the CB1 receptor. Fenofibrate was also able to modulate the response to CP 55,940 as demonstrated by a reduction in the CP 55,940 maximal response (86.4 ± 1.79 % CP 55,940 response; p<0.01; ANOVA) in the GTPγS assay.

In conclusion, fenofibrate is a somewhat selective CB2 receptor agonist with allosteric modulator properties at the CB1 receptor. Whether its rapid metabolism to fenofibric acid allows the survival of sufficient parent drug to have cannabinoid effects in vivo remains to be determined, but the present findings provide the basis for a new structural class of cannabinoid compounds with potential therapeutic properties.