Delta 9-tetrahydrocannabinol and cannabinol activate capsaicin-sensitive sensory nerves via a CB1 and CB2 cannabinoid receptor-independent mechanism

Abstract

Although Delta(9)-tetrahydrocannabinol (THC) produces analgesia, its effects on nociceptive primary afferents are unknown. These neurons participate not only in pain signaling but also in the local response to tissue injury. Here, we show that THC and cannabinol induce a CB(1)/CB(2) cannabinoid receptor-independent release of calcitonin gene-related peptide from capsaicin-sensitive perivascular sensory nerves. Other psychotropic cannabinoids cannot mimic this action. The vanilloid receptor antagonist ruthenium red abolishes the responses to THC and cannabinol. However, the effect of THC on sensory nerves is intact in vanilloid receptor subtype 1 gene knock-out mice. The THC response depends on extracellular calcium but does not involve known voltage-operated calcium channels, glutamate receptors, or protein kinases A and C. These results may indicate the presence of a novel cannabinoid receptor/ion channel in the pain pathway.

Fig. 1.

The naturally occurring cannabinoids THC and cannabinol evoke sensory nerve-mediated relaxation of rat hepatic and mesenteric arterial segments contracted with phenylephrine (PhE). The concentration-dependent relaxations induced by THC (○) in hepatic (n = 5) (A) and mesenteric (n = 6) (B) arteries, and those induced by cannabinol (○) in hepatic arteries (n = 7) (C) are abolished in arterial segments pretreated with the sensory neurotoxin capsaicin (10 μm; ⋄;n = 5 and 4 for THC and cannabinol, respectively). The CGRP receptor antagonist 8-37 CGRP (3 μm; ▪) also prevents relaxations induced by THC (n = 5 and 6 for hepatic and mesenteric arteries, respectively) and cannabinol (n = 4). B, As shown by thetrace, THC also relaxes the mesenteric artery without endothelium. The dotted line shows the basal tension level before addition of PhE. Data are expressed as mean ± SEM.

Fig. 2.

The vasodilator action of THC and cannabinol is not mimicked by C11 hydroxy and carboxy derivatives of THC. In humans, THC is metabolized to 11-OH-Δ⁹-THC and Δ⁹-THC-11-oic acid (Burstein; 1999), both of which fail to relax phenylephrine (PhE)-contracted rat hepatic arteries (n = 3). The dashed line shows the basal tension level before addition of PhE. The structures of the potent CB₁ and CB₂ receptor agonists HU-210 and CP 55,940 are also shown; these agonists are synthetic derivatives of THC without an intact C11 methyl group. None of these compounds cause sensory nerve-mediated relaxation in the rat hepatic artery (Zygmunt et al., 1999).

Fig. 3.

Effects of CB₁ and vanilloid receptor antagonists on sensory nerve-mediated relaxation induced by THC, cannabinol, and anandamide in rat hepatic and mesenteric arteries.A, The THC-induced vasorelaxation in hepatic arteries (○; n = 8) is not inhibited by the CB₁ receptor antagonists SR141716A (300 nm; ▪; n = 5) and AM251 (30 nm; ●;n = 4). Vasorelaxations evoked by THC (○;n = 10) (B) and cannabinol (○; n = 7) (C) are not inhibited by the competitive vanilloid receptor antagonist capsazepine (3 μm; ●; n = 8 and 4 for THC and cannabinol, respectively) but are abolished by the noncompetitive vanilloid receptor antagonist ruthenium red (1 μm; ▪;n = 8 and 4 for THC and cannabinol, respectively).D, In mesenteric arteries, THC-induced relaxations (○;n = 6; same as in Fig. 1B) are also unaffected by SR141716A (300 nm; ●;n = 4) and capsazepine (3 μm; ▪;n = 4) and are inhibited by 1 μmruthenium red (▴; n = 5). E, Anandamide-induced vasorelaxations in the absence (○;n = 5) and presence (●; n = 5) of 3 μm capsazepine or 1 μm ruthenium red (▪; n = 4). F, THC (10 μm) releases CGRP from rat hepatic arteries in the absence (n = 6; p < 0.001) but not in the presence (n = 6) of 1 μmruthenium red compared with basal CGRP release (n = 5). Data are expressed as mean ± SEM.

Fig. 4.

THC elicits sensory nerve-mediated relaxation in mouse isolated mesenteric arteries via a vanilloid receptor-independent mechanism. A, Capsaicin (●), anandamide (▾), and THC (○) evoke concentration-dependent relaxations of mesenteric arterial segments from wild-type mice contracted with phenylephrine (PhE; n = 4). Traces, all from separate arterial segments, show that THC and anandamide (AEA) fail to relax arteries pretreated with capsaicin (10 μm; top traces) or in the presence of 8-37 CGRP (3 μm; bottom traces) (n = 3–4). This lack of effect of THC and anandamide is not attributable to the inability of arteries to respond to vasodilators, because CGRP and SIN-1 (a nitric oxide donor) cause complete relaxations. B, THC induces relaxations of the same magnitude in arteries from VR1 gene knock-out mice (VR1^−/−; n = 5) and their control littermates (VR1^+/+; n = 7). AEA (n = 6) and capsaicin (CAP; n = 4) are equally as effective as THC at relaxing arteries from VR1^+/+mice, but they produce only minor relaxations in arteries from VR1^−/− mice (n = 6 and 7 for AEA and CAP, respectively). As shown by the traces, THC also relaxes arteries that do not respond to either CAP or AEA, indicating that sensory nerves are functional in VR1^−/− mice (E_max = 72 ± 3%;n = 13). Data are expressed as mean ± SEM. The dashed lines in traces show the basal tension level before addition of PhE.

Fig. 5.

THC-induced release of CGRP from sensory nerves in rat mesenteric arteries is dependent on calcium influx.A, THC (10 μm; n = 5) and caffeine (10 mm; n = 5) release CGRP from rat mesenteric arteries in PSS (p< 0.001 compared with basal CGRP release; n = 4). When calcium in the PSS is replaced by 10 μm EGTA, caffeine (n = 4) but not THC (n= 5) still evokes a release of CGRP (p < 0.001 compared with basal CGRP release; n = 4).B, THC (10 μm; n = 5) and caffeine (10 mm; n = 5) also release CGRP in Tris-buffer solution (p < 0.001 compared with basal CGRP release; n = 5). In the presence of 1 mm lanthanum, caffeine (n = 5) but not THC (n = 5) is able to release CGRP (p < 0.001 compared with basal CGRP release; n = 5). Data are expressed as mean ± SEM.

Fig. 6.

The effect of THC on perivascular sensory nerves does not involve protein kinases A and C. A, THC (10 μm) evokes CGRP release in rat hepatic arteries in both the absence and presence of the nonselective protein kinase inhibitor staurosporine (3 μm; p < 0.001 compared with basal CGRP release; n = 6).B, The protein kinase C activator PDBu releases CGRP from rat hepatic arteries in the absence (p< 0.01; n = 6) but not in the presence of 100 nm staurosporine (n = 5) compared with basal CGRP release (n = 6). C, PDBu elicits concentration-dependent relaxations in rat hepatic arteries contracted with phenylephrine. However, PDBu cannot relax arteries pretreated with 10 μm capsaicin (●;n = 5) or in the presence of 3 μm8-37 CGRP (▪; n = 5). D, PDBu-induced vasorelaxations are also prevented by 1 μmruthenium red (▪; n = 5) and partially inhibited by 3 μm capsazepine (●; n= 6). For clarity, the same controls (○) are shown inC and D (n = 7). Data are expressed as mean ± SEM.