Identification of a potent and highly efficacious, yet slowly desensitizing CB1 cannabinoid receptor agonist

Abstract

1 The relationship of agonist efficacy to the rate of G protein-coupled receptor signaling desensitization is controversial. 2 Expressing inwardly rectifying potassium channels (GIRKs) in Xenopus oocytes, we have devised a signaling assay that clearly identifies CB1 cannabinoid receptor agonists with low intrinsic efficacy. 3 In this assay, the synthetic CB1 agonists, AM411, AM782, AM1902, AM2233 and WIN55,212-2 and the endogenous cannabinoid, 2-arachidonoyl ester, were full agonists. 4 The synthetic CB1 agonist AM356 (methanandamide), the endogenous cannabinoids, anandamide and 2-arachidonoyl ether, and the phytocannabinoid, Delta9THC, were partial agonists. 5 The rate of desensitization of CB1 was independent of agonist efficacy. WIN55,212-2, AM782, AM1902, AM2233, and 2-arachidonoyl glycerol ester all desensitized quickly, with desensitization rates varying from 14% min(-1) to 10% min(-1). AM356, AM411, anandamide, and Delta9THC all desensitized considerably slower, at a rate of 5% min(-1). 6 Despite high potency and efficacy, AM411 desensitized as slowly as anandamide and Delta9THC. 7 CB1 agonist efficacy and rate of desensitization are not necessarily related.

Figure 1

Calculation of agonist efficacy and rate of desensitization. (a) Oocytes injected with 0.025 ng oocyte⁻¹ of CB1 and 0.02 ng oocyte⁻¹ of GIRK channel were perfused with high potassium buffer. CB1 agonist was applied after GIRK current stabilized. The amplitude of response to agonist (I_agonist) was compared to the amplitude of response to 1 μM WIN55,212-2 (I_1μMWIN), giving ‘% Maximal WIN Response' (b) Oocytes were injected with 0.05 ng oocyte⁻¹ CB1 cRNA, 0.02 ng oocyte⁻¹ GIRK channel, 0.625 ng oocyte⁻¹ GRK3, and 4 ng oocyte⁻¹ bovine Arrestin3. After GIRK current stabilized, oocytes were perfused with 1 μM WIN55,212-2 for 5 min followed by application of 1 μM SR141716. Desensitization was defined as the difference between the current at the beginning and end of the trace, normalized to the initial current. The rate was defined as the percent desensitized per minute.

Figure 2

Decreasing the amount of CB1 cRNA injected reveals partial agonism of anandamide and AM356. (a) Oocytes were injected with 0.02 ng oocyte⁻¹ GIRK channel and either 2.0 or 0.15 ng oocyte⁻¹ CB1 cRNA. WIN55,212-2, anandamide, and AM356 were applied to oocytes in concentrations ranging from 1 nM to 10 μM and results are expressed as ‘% Maximal WIN Response.' In oocytes expressing 2.0 ng oocyte⁻¹ CB1 cRNA, all agonists fully activated the current by 1 μM. However, in oocytes expressing 0.15 ng oocyte⁻¹ CB1 cRNA, 1 μM anandamide and AM356 only activated the current to 25–30% of the WIN55,212-2 response. Thus, 0.15 ng oocyte⁻¹ of CB1 cRNA unmasked the partial agonism of AM356 and anandamide. (b) WIN55,212-2, a full agonist, and AM356 and anandamide, both lower efficacy agonists, elicited the same GIRK current at 1 μM in oocytes injected with 2.0 ng oocyte⁻¹ of CB1 cRNA. However, in oocytes injected with 0.15 ng oocyte⁻¹ of CB1 cRNA, 1 μM AM356 and anandamide only elicited 25–30% of the WIN-activated currents. The number of oocytes tested for each condition is indicated in parentheses.

Figure 3

Dose–response curves for endogenous, plant-derived and synthetic agonists. (a) Oocytes were perfused with AM1902, AM782, AM2233, AM411, and WIN55,212-2 at the indicated concentrations. Despite widely varying EC50's, all compounds showed full agonism. (b) Oocytes were perfused with anandamide, AM356, 2-AG (ester), 2-AG ether, Δ⁹-THC, and WIN55,212-2 at the indicated concentrations. 2-AG and WIN55,212-2 were fully efficacious at 10 μM, whereas, 2-AG ether, Δ⁹-THC, anandamide, and AM356 elicited only 40–60% of the Maximal WIN Response at 10 μM (N=11–20 per concentration).

Figure 4

Relationship between CB1 affinity and EC50 for GIRK current activation. EC50's (determined from the data in Figure 3) and K_i (from same sources as Table 2) are well correlated, except for 2-AG (affinity lower than predicted by EC50) and AM2233 (EC50 lower than predicted by affinity).

Figure 5

Representative traces showing desensitization with selected agonists. Oocytes were injected with 0.05 ng oocyte⁻¹ CB1 cRNA, 0.02 ng oocyte⁻¹ GIRK channel, 0.625 ng oocyte⁻¹ GRK3, and 4 ng oocyte⁻¹ Arrestin3. Oocytes were perfused with high potassium buffer. After GIRK currents stabilized, agonists (WIN55,212-2, AM411, AM356, and 2-AG ether, as labeled) were applied at the indicated concentrations for 5 min, followed by application of 1 μM SR141716.

Figure 6

Agonist desensitization rates. (a) Desensitization rates of fully efficacious agonists varied greatly. Desensitization rates were determined as previously described in ‘Methods.' Using a one-way ANOVA nonparametric test, statistical analysis of the desensitization rates of the highly efficacious agonists (all at 1 μM) were compared to 1 μM WIN55,212-2 (P<0.001 (^**), P<0.05 (^*)). (b) Less efficacious agonists desensitized more slowly than WIN55,212-2. To evaluate whether or not desensitization rates correlate with agonist efficacy, the desensitization rates of agonists and WIN55,212-2 at equally efficacious concentrations were determined. The desensitization rate of 100 nM WIN55,212-2 was not quite statistically different from 1 μM WIN 55,212-2 with P>0.08 (^). 10 μM Δ⁹-THC, 10 μM AM356, and 1 μM AM411 all desensitized at significantly slower rates compared to 100 nM WIN55,212-2 (P<0.001(^**)), however, the desensitization rate of 10 μM 2-AG ether was not statistically different from 100 nM WIN55,212-2 (P>0.17).