Obesity and Gut–Brain Communication: The Cholinergic-Endocannabinoid Link

Obesity is a complex disease characterized by disruptions in energy homeostasis. Energy intake is governed by many factors, including bidirectional communication between the gastrointestinal tract and the brain via the vagus nerve (Moura-Assis et al., 2021). The gastrointestinal tract provides the brain with information about food intake, such as the macronutrients and volume of food consumed, and thus signals to the brain when to terminate a meal (Moura-Assis et al., 2021). Conversely, the brain regulates digestion and absorption in the gut via vagal efferents (Browning and Travagli, 2011), whose cell bodies are located in the dorsal motor nucleus of the vagus (DMV) in the brainstem. The activity of these neurons is partially governed by central melanocortin-4 receptor (MC4R) signaling, which plays a pivotal role in food intake and energy homeostasis (Yeo et al., 2021). Previous work has demonstrated significant perturbations in gut–brain signaling in obesity, particularly in vagal afferents (de Lartigue, 2016). The extent to which vagal efferent signaling contributes to obesity, however, is understudied.

One way vagal efferents may influence energy intake is by stimulating the production of the endocannabinoid 2-arachidonylglycerol (2-AG). 2-AG is the main ligand of the cannabinoid receptor CB1R, which is highly expressed in the gastrointestinal tract and the vagus, as well as in the CNS. Numerous reports have shown that endocannabinoids influence energy homeostasis, food consumption, and food preference, in part by activating the hypothalamic orexigenic system, which promotes eating. Importantly, there is a positive correlation between plasma endocannabinoid levels and indicators of obesity, such as body mass index and abdominal adiposity (Côté et al., 2007), and levels of key endocannabinoids are elevated in the gut epithelium in obesity (DiPatrizio, 2021). Previous work has shown that vagal CB1Rs regulate gastrointestinal motility, although not body weight (Vianna et al., 2012).

Vagal efferent control of the stomach is mediated exclusively through the release of acetylcholine, and numerous reports have shown that activation of muscarinic acetylcholine receptors (mAChRs) drives the production of endocannabinoids (DiPatrizio, 2021). 2-AG is synthesized via hydrolysis of 1,stearoyl,2-arachidonyl-sn-glycerol (SAG, 18:0, 20:4) by diacylglycerol lipase α/β (DAGLα/β; DiPatrizio, 2021), and in the gut, this is hypothesized to be driven by vagal cholinergic signaling during fasting (DiPatrizio, 2021). Although the system remains locally homeostatic through degradation of 2-AG by monoacylglycerol lipase (MAGL), which in turn is inhibited by 2-linoleoyl glycerol (2-LG, 18:2) and 2-oleolglycerol glycerol (2-OG; Murataeva et al., 2016), this homeostasis might be disrupted during obesity.

The epidemic of obesity is thought to stem partly from overconsumption of highly palatable foods. It is therefore possible that vagal efferent stimulation of the gut following consumption of highly palatable food results in the production of endocannabinoids that promote additional eating. To test this hypothesis, Wood et al. examined the interplay between cholinergic and endocannabinoid signaling in mice exhibiting diet-induced obesity (DIO) after being fed a high-fat, high-sucrose diet that mimics the modern western-world diet (Wood et al., 2024).

Immunohistological staining for the activity-induced protein cFos showed significantly greater activation of cholinergic neurons in the DMV of DIO mice than in the DMV of control mice. In addition, levels of MAGs (2-AG, 2-DG, and 2-OG) in the gut epithelium were significantly higher in DIO mice than control mice. To determine whether the increase in MAG levels resulted from increased cholinergic transmission, the authors injected a selective M3 mAChR antagonist. This significantly reduced the levels of all three MAGs. Similarly, treatment with a selective M1 mAChR antagonist reduced MAG levels, although only the reduction in 2-OG was statistically significant. Finally, injection of the nonselective mAChR antagonist atropine reduced levels of all three MAGs. These results suggest a direct link between the cholinergic system and the production of endocannabinoids in DIO mice, wherein activation of M3 and possibly M1 mAChR leads to increases in 2-AG in the gut epithelium.

Next the authors investigated whether increased endocannabinoids in the gut resulted from an increase in levels of SAG, the precursor of 2-AG, and/or the activity of DAGL, which hydrolyzes SAG into MAGs. Indeed, SAG formation and DAGL activity were higher in DIO mice than in mice fed standard diet. This increase could be reversed by mAChR antagonism, further demonstrating that the production of gut MAGs is under cholinergic control. In contrast, MAGL activity was unaffected by drug treatment. This suggests the cholinergic system drives the production of endocannabinoids without increasing their breakdown, thereby leading to elevated levels in the DIO model.

To investigate whether the reversal of endocannabinoid production by mAChR antagonism translates to functional effects, the food intake of mice fed a western diet was measured. Single-dose intraperitoneal injection of a nonselective mAChR antagonist, a selective M3 mAChR antagonist, or a CB1R antagonist reduced caloric intake in DIO mice but not in control mice; however, the M1 mAChR agonist had no detectable effect. Overall, these results indicate that upregulated cholinergic signaling, likely through M3 mAChR, and subsequent increased endocannabinoids acting on CB1Rs, contribute to increased eating in this model of western-DIO.

To verify that CB1R in the intestinal epithelium is required for the suppression of eating produced by mAChR and CB1R antagonists in DIO mice, mice in which CB1Rs were knocked out selectively in the intestine (intCB1-/- mice) were exposed to the western diet. Results showed that intCB1-/- mice gained less body mass compared with controls. Furthermore, intCB1-/- mice were no more affected by injection of the CB1R antagonist or nonselective mAChR antagonist than by vehicle injection, whereas intCB1+/+ mice injected with vehicle showed a higher caloric intake than those injected with either antagonist. This supports the hypothesis that endocannabinoids act via gut epithelial CB1Rs to stimulate western diet intake in DIO mice.

In summary, Wood et al. have uncovered a pathway by which endocannabinoid signaling drives overeating in diet-induced obesity. Specifically, a high-fat, high-sugar western diet is associated with increased activation of vagal cholinergic motor neurons. This increase in cholinergic signaling activates M3 mAChR in the gut, leading to elevated endocannabinoid production. These endocannabinoids, in turn, act on intestinal CB1Rs to drive an increase in feeding.

Wood et al. observed greater activation of cholinergic neurons in the DMV of DIO mice than in controls. They note that further investigation is required to confirm that the activated neurons are stomach-innervating vagal efferents, as this region houses cell bodies from vagal efferents innervating numerous other organs. This could be achieved through viral tracing. In addition, it will be important for future work to identify the source of the increased activation. Previous work has demonstrated that DMV neurons expressing MC4R activate vagal efferents that project to the gut, but activation of these receptors does not appear to affect feeding; instead, these neurons influence energy expenditure and glucose homeostasis (Yeo et al., 2021). Therefore, it would be informative to investigate activation of feeding-related neurons in brain regions known to project to DMV, such as POMC neurons in the hypothalamic areas.

Wood et al. focused most of their study on 2-AG, even though levels of 2-OG and 2-DG were also increased in DIO mice and were reduced by cholinergic drug treatment. Notably, levels of 2-OG were 10-fold higher than those of 2-AG in DIO mice. Moreover, this bioactive lipid is usually considered part of the 2-AG entourage and is speculated to be a functional antagonist of the CB1R (Murataeva et al., 2016). Future research should therefore investigate the role of 2-OG and 2-DG in the gut generally, as well as in relation to DIO.

Additional work should also elucidate the specific role of gut cholinergic and endocannabinoid systems in feeding in both normal and DIO conditions. In comparison to the findings of gut CB1R action in this study, recent data has suggested that vagal afferent CB1Rs do not regulate body weight homeostasis (Vianna et al., 2012); hence it will be important to determine how the effects of blocking CB1R in the gut in this current study influence feeding and appetite, which are controlled centrally, and if so, whether this effect is mediated via vagal afferents. Wood et al. hypothesize that the hypophagic effect of CB1R and M3 mAChR antagonists occurs via a cholecystokinin (CCK)-mediated mechanism, as endocannabinoids have been demonstrated to inhibit CCK release from I cells, a subtype of hormone-releasing cells that line the gut epithelium (DiPatrizio, 2021). CCK acts as a satiety signal, regulating short-term food intake, known to act via receptors in the stomach, vagal afferents, and the CNS. Further investigation into how intestinal CB1R signaling affects feeding should be a key next step for this research.

The results of the study by Wood et al. may ultimately lead to improved or new clinical treatments for obesity, although it is important to consider the nonspecificity and off-target effects of many of the pharmaceutical approaches used in this study. In particular, peripheral mAChR antagonists affect many other organ systems, such as cardiac function. Furthermore, given that vagal efferents signal to the gut only via acetylcholine, it is likely that by blocking mAChRs, particularly M3, digestion and stomach contractility will also be affected (Tanahashi et al., 2021), with other possible effects on energy homeostasis. Finally, the authors found that a muscarinic antagonist, both alone and in combination with a CB1R antagonist, led to a decrease in ambulatory movement, which could indicate not only malaise, but also anxiety or reduced food-seeking behavior. Overall, any potential off-target effects should be thoroughly examined if the results of this study are translated into clinical treatments. Nevertheless, the work by Wood et al. highlights a mechanism that may contribute to dysregulated food intake in obesity and opens up further questions and potential targets for the treatment of obesity.