Multiple lines of research began to elucidate the possible mechanism by which cannabinoids modulate intestinal transit. Studies from the 1970s showed that THC reversed cholinergic-mediated contractions within the guinea pig ileum, indicating that cannabinoids might act to reduce peristalsis via an—at the time—undiscovered receptor that inhibits the release of the excitatory neurotransmitter, acetylcholine, from enteric nerves in the gastrointestinal tract.^15–17 Later in vitro work using human and rodent small intestinal tissues supported this hypothesis.^18–21 Croci et al. reported that electrically evoked twitch responses in a human ileum longitudinal smooth muscle preparation were blocked by the general muscarinic acetylcholine receptor (mAChR) inhibitor, atropine, or the neurotoxin, tetrodotoxin (TTX), suggesting that twitch responses were mediated by cholinergic neurons.¹⁹ Importantly, application of the general cannabinoid receptor agonist, (+) WIN 55,212-2 (WIN), dose dependently inhibited twitch responses, and when WIN was applied in combination with atropine or TTX, no additive effects were observed. Furthermore, WIN was found to exert its effects in the ileum through activating CB₁Rs because the selective CB₁R antagonist/inverse agonist, rimonabant—but not the CB₂-selective antagonist, SR144528—blocked WIN-mediated inhibition of twitch responses. These results suggest that CB₁Rs control cholinergic neurotransmission in the human gastrointestinal tract and are the key regulators of contractility (Fig. 1D).

Several other research groups found similar results in the rodent small intestine, suggesting that interactions between endocannabinoid and cholinergic pathways in the gut are conserved across many species and likely serve a similar physiological role in the enteric nervous system to control gastrointestinal contractility.^18,20–22 For example, Coutts and Pertwee reported that WIN and the nonselective cannabinoid receptor agonist, CP-55940, dose dependently inhibited twitch responses to electrical stimulation in the myenteric plexus longitudinal preparation of the guinea pig small intestine, an effect blocked by rimonabant.¹⁸ WIN or CP-55940 had no effect on contractility induced by exogenous application of acetylcholine, but rimonabant alone significantly increased the release of acetylcholine, indicating a prejunctional site of action for CB₁Rs in controlling endogenous acetylcholine release. Studies by Izzo et al. found a similar effect on contractility in the guinea pig ileum circular smooth muscle for WIN and the endocannabinoid, anandamide.²⁰ Both cannabinoid receptor agonists dose dependently inhibited cholinergic- and noncholinergic-mediated contractile responses to electrical field stimulation, and all effects were blocked by rimonabant. Importantly, when rimonabant was given alone, contractile responses were enhanced but failed to affect exogenously applied acetylcholine-induced contractions, further indicating that CB₁Rs control contractility by inhibiting acetylcholine release from enteric neurons. In the mouse, anandamide or the selective CB₁R agonist, arachidonoyl-2′-chloroethylamide (ACEA), reduced spontaneous contractility of ileum longitudinal muscle in a dose-dependent manner.²¹ The actions of anandamide or ACEA were inhibited by rimonabant, but not the selective CB₂R antagonist, AM630, again implicating CB₁Rs in the response. Indeed, CB₁R immunoreactivity was found to be colocalized with choline acetyltransferase in the myenteric plexus of rat and guinea pig, further supporting the hypothesis that CB₁Rs interact with cholinergic neurons to regulate contractility.²³ In addition, monoacylglycerol lipase (MGL), the primary degradative enzyme for 2-arachidonoyl-sn-glycerol (2-AG),²⁴ was found to be heavily expressed throughout the rat small intestine and colocalized with the enteric neural marker, PGP 9.5.²⁵ Blocking the degradation of 2-AG by pharmacologically inhibiting the activity of MGL with URB602 reduced whole-gut transit, an effect completely absent in CB₁ null mice, suggesting that endogenously produced 2-AG acting at CB₁Rs controls gut contractility.

Recent studies by the Sharkey group (Bashashati et al.²⁶) found a similar role for the biosynthetic enzyme for 2-AG, diacylglycerol lipase (DGL),²⁷ in the control of intestinal transit and constipation. In these studies, CB₁^−/−, wild-type controls, and a mouse strain with a constipated phenotype, C3H/HeJ, were used. Immunoreactivity for DGLα in the ileum and colon myenteric plexus colocalized mostly with the vesicular acetylcholine transporter in nerve processes and punctate terminals surrounding enteric nerves. The highest expression of DGL was found in the stomach and colon compared to levels in the duodenum and ileum. Inhibiting the degradation of 2-AG with the MGL inhibitor, JZL-184, decreased whole-gut transit in wild-type controls, but not in CB₁^−/− mice, and enhanced the inhibitory effects of systemically administered 2-AG on transit. Scopolamine, an anticholinergic that inhibits mAChRs, and loperamide, a general opioid receptor agonist, were used to reduce contractility in vitro in the mouse ileum and colon during electrical field stimulation, and the effects of inhibiting 2-AG biosynthesis by DGL with orlistat or OMDM-188 were evaluated. DGL inhibition reversed the inhibitory effects on contractility induced by scopolamine or loperamide. Similar results were found for orlistat or OMDM-188 on whole-gut transit when systemically administered to mice. Importantly, normalization of scopolamine or loperamide effects on transit was found in wild-type mice, but not in CB₁^−/− mice, indicating that DGL likely exerts its actions on intestinal transit through the biosynthesis of 2-AG, which acts at local CB₁Rs. C3H/HeJ mice had reduced fecal output when monitored for more than 1 h, and DGL inhibition with OMDM-188 increased output. Together, the results suggest that contractility and fecal output in the mouse ileum and colon may fall under the regulation of 2-AG signaling at local CB₁Rs.

In contrast to the above work describing a role for CB₁Rs—but not for CB₂Rs—in the human and rodent small intestines on cholinergic-mediated contractility, pharmacological evidence suggests that both, CB₁Rs and CB₂Rs, control cholinergic neurotransmission in the mouse stomach.²⁸ Both, atropine and TTX, abolished intraluminal pressure changes following electrical field stimulation to the mouse stomach, implicating cholinergic neurotransmission in the response. Interestingly, the effects of electrical field stimulation on intraluminal pressure were inhibited by all test compounds, including WIN, anandamide, and ACEA, as well as the selective CB₂R agonists, JWH015 or JWH133. The evidence suggests that cholinergic-mediated gastric functions can be regulated through cannabinoid activity at both CB₁Rs and CB₂Rs differentially across specific organs in the gut (i.e., stomach vs. small intestine).

Boesmans et al. reported that CB₁Rs might also control mitochondrial transport in enteric nerves.²⁹ In vitro experiments using cultured guinea pig myenteric neurons revealed that spontaneous activity was increased in the presence of rimonabant and another selective CB₁R antagonist, AM251, but inhibited when CB₁Rs were activated with anandamide or methanandamide, a stable analogue of anandamide. Activity was blunted in the presence of URB957 and AA-5HT, both inhibitors of anandamide degradation by FAAH, highlighting the presence and role for endogenously synthesized CB receptor ligands in mediating enteric activity. Importantly, mitochondrial transport among enteric neurons was enhanced by the CB₁R antagonist and decreased in the presence of the CB₁R agonists, suggesting that CB₁Rs might control energy metabolism in these cells. Further studies, however, will be necessary to confirm the precise physiological relevance of CB₁Rs in mitochondrial function in the gastrointestinal tract.

Important studies also suggest that the endocannabinoid system in the gut is altered after consumption of high-fat diets, which is proposed to influence gut motility.^5,30 Mice were fed a high-fat diet for 8 weeks, and small intestinal levels of anandamide and 2-AG were quantified, as well as intestinal transit.⁵ Interestingly, levels of anandamide in the small intestine were decreased, while levels of 2-AG increased, and these effects were met with increases in intestinal transit. Given the well-established evidence of a decrease in gastrointestinal motility following CB₁R activation,¹ that levels of 2-AG are on average 10–20 times higher than those for anandamide in the same tissue, and that 2-AG acts as a full agonist at CB₁ compared to the partial agonist, anandamide,³¹ the results were surprising. The authors concluded that decreases in small intestinal content of anandamide led to the observed increases in transit.⁵ Endocannabinoids in these studies were extracted from whole homogenates of intestine, including mucosal and serosal layers, and thus make it difficult to ascertain whether specific areas within the intestine contain varying levels of the endocannabinoids, which could have distinct functional consequences on intestinal transit. An in-depth analysis of the subcellular localization for the endocannabinoid system as well as studies investigating the contribution of modified endocannabinoid signaling in obesity for gastrointestinal function are warranted.