The Enteric Glia and Its Modulation by the Endocannabinoid System, a New Target for Cannabinoid-Based Nutraceuticals?

2.1. Enteric Neurons

IPANs possess mechano- or chemosensory activity, and besides the straight signal reception, they are able to receive and process the message of the intensity, duration, and pattern of stimuli. These neurons usually form a circumferential internetwork encircling the intestine. Within the group of IPANs, several classes may be listed, for example, according to their localization (myenteric/submucosal plexus) or the direction of signal transduction. Therefore, IPANs can receive, integrate and reinforce signals both locally and across the network (alike interneurons) [25,26].

Interneurons, like IPANs, may be divided into ascending or descending. Furthermore, within the population of interneurons there are several classes that may be distinguished neurochemically and the proportion of interneurons in these classes may differ between the parts of the GI tract, that may reflect the regional diversity in the motor patterns in the intestines [4,27,28].

The last group of neurons in the ENS are motor neurons, which are divided into two subgroups: inhibitory and excitatory. They participate in the control of intestinal motility as they contribute to the contractions and relaxations of the circular and longitudinal smooth muscles in a mechanism dependent on acetylcholine (ACh) (excitatory neurons), or nitric oxide (NO), vasoactive intestinal peptide (VIP) and pituitary adenylate cyclase-activating polypeptide (PACAP) (inhibitory neurons) [2].

3.1. The Endocannabinoid System in the Gastrointestinal Tract

The ECS components are widely expressed in the GI tract. The presence of CB1 receptors, the most predominant in the intestines, was confirmed in the enteric neurons [156], myocytes [157], and epithelial cells [158]. It should be emphasized that intestinal CB1 receptors participate in epithelial regeneration [159] and therefore play a crucial role in the maintenance of the intestinal barrier integrity [158]. Activation of CB1 receptors influences GI motility as it evokes the relaxation of the longitudinal smooth muscles. This action is a combination of the neurogenic effect and direct impact on myocytes [157]. Noteworthy, mRNA expression of CB1 receptors in the mouse colon does not change significantly during inflammatory conditions, i.e., induced by dextran sulfate sodium (DSS) and LPS [160].

CB2 receptors are present in the GI tract, but their expression is lower in comparison to CB1 receptors [14,156]. Similarly to CB1 receptors, LPS- or DSS-induced colitis does not alter the mRNA expression of CB2 receptors, but these receptors do play an important role during intestinal inflammation. In particular, the activation of CB2 receptors with a selective agonist (JWH-133) attenuates chronic colitis in IL-10 deficient mice [161,162]. Furthermore, according to Storr et al. [163], CB2 mRNA is up-regulated in mice with 2,4,6-trinitrobenzene sulfonic acid (TNBS) induced intestinal inflammation.

In animal models of intestinal inflammation (induced by DNBS or TNBS), the level of AEA is increased in the mucosa, but not in the muscular layer of the colon [164]. Furthermore, AEA is up-regulated in the colonic biopsies collected from patients with UC. However, the expression of 2-AG is not altered in both, animal, and human, inflamed colonic samples [164].

According to Grill et al. [160], MAGL mRNA expression decreases in mouse intestines in LPS- and DSS- induced colitis. Furthermore, it was found that the inhibition of enzyme activity increases the level of 2-AG and thus improves TNBS-induced colitis in mice [165]. Interestingly, Wasilewski et al. [166] investigated the impact of inhibition of the activity of enzymes involved in the degradation of ECBs (FAAH, MAGL) on the colonic secretion stimulated with forskolin (cyclic adenosine monophosphate, cAMP-dependent secretagogue), veratridine (voltage-dependent sodium channel activator) or bethanechol (cholinergic receptor agonist, resistant to the action of cholinesterases). The inhibition of FAAH activity abolishes the pro-secretory effect induced by forskolin, which is mediated through CB1/CB2 receptors. On the other side, the anti-secretory action of MAGL inhibitors is reversed by the CB2 receptor antagonist AM-631, but not by AM-251 (CB1 receptor antagonist).

The ECS is also involved in the control of GI peristalsis and transit [14]. In particular, CB1 receptors were pointed out as involved in the inhibitory effect of cannabinoids, both, under control [167,168,169,170,171] and under inflammatory conditions [172]. IBS, or chemotherapy-induced dysmotility are associated with alterations in CB1 and CB2 receptors expression/activity, levels of ECBs or ECB biosynthesis/degradation.

Thus, the levels of 2-AG in plasma were higher in IBS-D patients, while oleoylethanolamide (OEA) and palmitoylethanolamide (PEA) were lower in comparison to healthy participants [173]. On the contrary, OEA concentration in plasma was higher in IBS-C patients as compared to controls. The AEA levels were similar in all IBS subtypes. Interestingly, the ECB turnover may be a key factor in the pathophysiology of IBS as the mRNA expression of FAAH in the colonic biopsies was significantly lower in IBS-C patients in comparison to healthy controls and the IBS-D group [173]. These results suggest that the impaired process of ECB degradation may lead to sustained activation of CB receptors and be the cause of prolonged intestinal transit. Furthermore, the symptoms of IBS have been related to the genetic polymorphism of genes encoding CB1 receptors and FAAH [174,175]. For example, a higher number of AAT triplets in the CNR1 gene (10 or more triplets) was noted in IBS patients in comparison to healthy controls. Moreover, this polymorphism was reported in patients with higher scores on the abdominal pain/discomfort symptoms scale [174]. The polymorphism in CNR1 rs806378 (CC vs. CT/TT) correlated with the efficacy of colonic transit: patients with TT variant were characterized with the fastest colonic transit [175]. The polymorphism in genes encoding FAAH may be related to IBS: conversion of 385C to A leads to decreased expression of FAAH. The CA/AA polymorphism was more frequent in IBS-D and IBS-M patients. There was a strong association between the FAAH CA/AA and accelerated colonic transit in IBS-D patients [176].

With regards to chemotherapy-induced dysmotility, it was found that the non-selective synthetic cannabinoid agonist WIN 55,212-2 (WIN), at a low dose devoid of central effects, partially decreased diarrhea associated with 5-FU treatment in rats, despite not being able to improve gut inflammation [177], suggesting a direct effect on the myenteric plexus. In contrast, animals treated with vincristine displayed paralytic ileus that improved when treated previously with the CB1-selective antagonist AM251, suggesting that vincristine-induced GI motor reduction is associated with an increase in ECS activity, involving CB1 receptors [178]. This is consistent with findings of ECS activation in other paralytic ileus conditions, like that produced by LPS administration [179]. Finally, WIN was not able to prevent repeated cisplatin-induced pica or gastric dysmotility (indirect markers of nausea/vomiting in non-vomiting species [180,181]) in the rat [182,183]. These results are conflicting with the known empiric use of cannabinoids to prevent chemotherapy-induced nausea and vomiting but might be related to similar mechanisms occurring during the so-called cannabis hyperemesis syndrome, suffered by heavy cannabinoid consumers [184], at least those genetically susceptible [20].

Finally, a role for ECS has been demonstrated in abdominal pain [185]. Thus, in early preclinical studies using colorectal distension (CRD), both CB1 and CB2 receptors elicited analgesic effects under basal conditions and during inflammation-induced hyperalgesia [186,187]. Interestingly, an endogenous cannabinoid tone able to activate CB1 receptors was found in response to noxious CRD [188]. Moreover, TRPV1 channels were early demonstrated to be involved in the development of chemically induced visceral hypersensitivity to CRD [189] and acute mechanical colonic hyperalgesia without prior chemical challenge [190]. Furthermore, compared with control animals, AEA levels and TRPV1 expression increased whereas CB1 receptor expression decreased in DRG from rats submitted to water avoidance stress for 10 consecutive days (a paradigm well known to produce visceral hypersensitivity to CRD), and these changes were prevented by prior administration of the non-selective cannabinoid agonist WIN or the TRPV1 antagonist capsazepine [191]. More recently, it has been proposed that modulation of both FAAH and MAGL by dual inhibitors (i.e., increasing both AEA and 2-AG) might be a good strategy to control visceral pain [192]. Importantly, changes in the ECS in some cerebral areas, at least those affecting CB1 and TRPV1, may contribute to visceral hypersensitivity development at the supraspinal level [193].

Whatever the mechanism, the preferred cannabinoid-based strategies to counteract visceral pain and other altered functions of the GI tract are those not capable of exerting psychotropic effects, including peripherally-acting CB1 agonists [167], CB2 agonists (which regulate inflammation but lack psychotropic actions), inhibitors of ECB degradation or other compounds like the phytocannabinoid cannabidiol (CBD), whose amazing number of molecular targets and lack of psychoactivity constitute an attractive alternative in this field [18]. Many other compounds in hemp (the variety of Cannabis sativa with lower concentrations of the main psychoactive compound of marijuana, Δ⁹-tetrahydrocannabinol), including other phytocannabinoids, terpens, alkaloids, steroids, flavonoids, and lignans are being studied for their potential use as nutraceuticals in the context of the GI ailments [194].

Likewise, other cannabinoid-like compounds may produce beneficial effects in those GI conditions [18]. However, as shown in the following section, thus far only a few studies have addressed the effects on the EGCs of phytocannabinoids and other cannabinoid-like molecules with potential nutraceutical use.

3.2. EGCs and the ECS

Cannabinoid compounds, due to their antioxidant and anti-inflammatory activity, could help to modulate the inflammatory processes in which EGCs intervene. However, not much is known about the specific interaction between the ECS and EGCs, probably because these cells do not seem to express the typical cannabinoid CB1 and CB2 receptors. In fact, a study using mouse colonic myenteric plexus found no co-localization of the CB1 receptor with specific markers for these cells, S100-β or GFAP [195]. Other immunostainings did not prove the presence of CB1 or CB2 receptors in EGCs in the GI tract (pylorus, duodenum, ileum, and colon) of cats [196]. In a study carried out in dogs, the CB1 and CB2 receptors were not seen either in the EGCs from the myenteric plexus, although weak-to-moderate immunoreactivity was detected in both neurons and glial cells from a few submucosal ganglia [197]. However, in a recent study carried out in cryosections of the distal ileum of horses, EGCs showed immunoreactivity for the CB1 receptor and PPARα [198]. Thus, the expression of cannabinoid receptors by the EGCs may be species-dependent. Not much is known about the expression and role of CB1 and CB2 receptors in human EGCs. Despite the apparent lack of expression of CB receptors in the ENS in most species evaluated, CB agonists and antagonists might exert indirect actions on the EGCs. For example, it was shown that LPS increased the expression of c-fos in rat ileal EGCs and enteric neurons, and this was attenuated by CB2 agonists. Since enteric neurons do express CB2 receptors under inflammatory conditions, LPS activation of EGCs could be secondary to CB2-mediated neuronal activation [156].

TRPV1 is a non-selective cation channel activated by exogenous plant-derived vanilloid compounds as well as by endocannabinoids (namely, anandamide). In a study carried out by Yamamoto et al. [199], TRPV1-immunoreactive signals were detected in EGCs of the myenteric plexus of wild-type mice but not in TRPV1 knockout mice. Altered expression of GFAP at early postnatal time points in knock-out mice suggested that TRPV1 could be involved in enteric glia maturation. Furthermore, the addition of a TRPV1 antagonist to EGC cultures from wild-type mice myenteric plexus, decreased the expression ratio of GFAP to S100-β [199]. Thus, anandamide (and exogenous vanilloid compounds, like capsaicin), acting through TRPV1, may be a regulator of EGCs development and function, a hypothesis that still needs to be proved.

In contrast, EGCs express PPARα [196,197], a nuclear hormone receptor to which transcription-related ligands bind, and whose activation may induce anti-inflammatory and antinociceptive effects [58]. Since 2002, there has been accumulating evidence of the interaction of endocannabinoid compounds with PPARα, which may also be activated by other compounds similar to cannabinoids, phytocannabinoids or synthetic cannabinoids [200]. Since this receptor is expressed in EGCs, some investigations have been carried out to analyze the modulation of the pro-inflammatory activity of these cells by cannabinoid/cannabinoid-like compounds able to activate PPARα.

The AEA-like molecules OEA and PEA have an affinity for different receptors, including nuclear receptors such as PPARα, channels such as TRPV1, and membrane receptors such as GPR119 (OEA) and GPR55 (PEA). Of the two, PEA is the only one whose effects on EGCs have been reported [58]. PEA exerts dose-dependent anti-inflammatory effects on mouse models of UC and in human biopsies. This is due to its ability to reduce EGC activation, with the involvement of a cascade of PPARα activation, decreased expression of S100-β and TLR4, reduced expression of pro-inflammatory proteins such as iNOS, COX-2, and TNF-α, decreased NO production, reduced myeloperoxidase activity in mucosal neutrophils and reduced macrophage infiltration [58].

Interestingly, in a rat model of HIV-1-associated diarrhea induced by administration of HIV-1 tat, PEA reduced diarrhea through PPARα and consequent blockade of TLR4/NFκB activation, of colonic submucosal EGCs activation and overexpression of S100-β and iNOS [201].

Alzheimer’s disease (AD) is one of the most common neurodegenerative disorders characterized by functional digestive disturbances, including infrequent bowel movements, constipation, and defecatory disorder. The possible benefits of PEA administration in counteracting enteric inflammation have been investigated in an AD mouse model (SAMP8; Senescence-Accelerated Mouse-prone 8) [202]. In addition, the effects of PEA on EGCs were tested in cultured cells treated with LPS and β-amyloid 1–42 (Aβ) to mimic AD conditions. In this study, SAMP8 mice showed an increase in the expression of S100-β in colonic tissues, suggesting the presence of reactive gliotic processes. PEA administration induced a reduction of enteroglial-derived S100-β protein expression indicating that PEA is able to blunt the gliotic reaction. In cultures, EGCs treatment with PEA counteracted the increment of S100-β, TLR-4, NF-κB p65, and IL-1β release induced by LPS and Aβ. These results suggest that the anti-inflammatory effect of PEA prevents the enteric glial hyperactivation and counteracts the onset and progression of the colonic inflammatory condition by selectively targeting the S100-β/TLR-4 axis on ECGs, resulting in an inhibition of the NF-κB p65 pathway and cytokines release [202].

To our knowledge, no report has been published evaluating the involvement of other components of the endocannabinoid system (degradation enzymes FAAH and MAGL, GPR55, GPR119, …) in the physiopathology of EGCs.