The first studies suggesting that n-6 oxylipins could modulate enteric neurotransmission and thus neuronal activity were performed in the myenteric plexus of guinea pig ileum and focused on the release of acethylcholine (Ach). Two independent groups showed that the release of Ach by either protein kinase C, DMPP (a nicotinic agonist) or neuronal activity was decreased in the presence of mepacrine, a PLA₂ inhibitor, and that the opposite was observed when samples of myenteric plexus were treated with either a PLA₂ activator (melittin) or ARA (30, 31). Additional pharmacological experiments allowed to identify the prostaglandins PGE₂, PGI₂ and PGF₂ as critically involved in the chemically or electrically-induced release of Ach by enteric neurons (32–35) (Figure 2). Regarding the LOX pathway, AA861, a selective inhibitor of 5-lipoxygenase, inhibited acetylcholine release from the myenteric plexus induced by electrical field stimulation, an effect that was reversed by metabolites from the 5-lipoxygenase pathway, including 5-HETE, LT C₄, D₄ and E₄ (36) (Figure 2). Besides Ach, it has also been shown that prostanoid signaling, including PGE₂, regulated the cytokine-induced suppression of norepinephrine release in rat myenteric plexus (37). Although scarcer than for the myenteric plexus, some findings suggest that the n-6 oxylipins produced by the submucosal plexus might be also involved in the release of Ach. Interleukin-1β (IL-1β) potentiated the release of Ach induced electrical transmural stimulation in whole guinea pig ileum but not in longitudinal muscle/myenteric plexus (LMMP) only or in mucosal-free preparations and these effects were prevented with of PLA₂, COX or LOX inhibitors (38) (Figure 2). Taken as a whole, these findings suggest that IL-1β stimulates cholinergic enteric neurons via a cascade of ARA metabolites produced in tissues at the mucosal and submucosal level, but not in the myenteric plexus or longitudinal muscles.

Prostaglandins are also capable of directly modulating the electrophysiological properties of enteric neurons. Treatment of myenteric ganglia from the guinea pig colon or ileum with PGE₂ led to prolonged depolarization of enteric neurons with enhanced spike discharge together with an increased neuronal excitability (39–41) (Figure 2). More recently, myenteric neuron activity and contraction have been shown to depend on prostaglandin release (COX-2 pathway) from mucosal mast cells in a model of TNBS-induced colitis (42). Similar findings were observed when submucosal neurons of guinea pig colon were treated with PGD₂ (43). The application of leukotrienes C₄, D₄ and E₄ to myenteric and submucosal neurons of guinea pig resulted in slow depolarization of small intestine neurons, thereby suggesting that derivatives of the LOX pathway are also modulators of enteric neuronal excitability (Figure 2).

Oxylipins are well-known neuroprotectors in the CNS and this might be also the case in the ENS. 15-Deoxy-Delta-12,14-prostaglandin J₂ (15d-PGJ₂), a PGD₂ metabolite produced by EGCs, has been identified as having neuroprotective effects (44) (Figure 2). Pretreatment of primary cultures of ENS with this n-6 oxylipin prevented the neurotoxicity induced by hydrogen peroxide. The involvement of EGCs in this beneficial effect was validated after glial genetic invalidation of the key enzyme involved in the synthesis of 15dPGJ₂, the L-PGDS. The neuroprotective effect of this glial mediator appeared to be due to increased expression in enteric neurons of both glutamate cysteine ligase and intracellular glutathione.

In sharp contrast with the regulation of enterics neuron properties described below, data on the effects of n-6 oxylipins on the EGCs are lacking. The observation showing that a high fat diet, which controls in a major way the GI oxylipin content, increased the number of EGCs in myenteric plexus in a juvenile mouse model (45), but alter submucosal glial density where the expression of the glial marker calcium-binding protein S100β (S100β) was also decreased (46), might suggest that EGC morphology and/or reactivity could be modified via oxylipins. But this is the few we know, and while many studies described the regulation of central glia (astrocytes or microglia) through oxylipin (mostly n-3 oxylipin), the impact of a specific oxylipin on EGC markers and functions remain to be addressed.

The intestinal epithelial barrier (IEB) represents a central actor of intestinal homeostasis. Intestinal epithelial barrier integrity is ensured by a constant renewal thanks to a balance between proliferation, differentiation, migration and elimination of cells, but also by a tight regulation of permeability and repair. The role of EGCs in controlling the properties of the IEB is quite well studied, and our laboratory has highlighted the crucial role of n-6 oxylipins in this regulation. Thus, the 15dPGJ₂ induced differentiation and inhibited the proliferation of epithelial cells (52) (Figure 3). We also demonstrated the involvement of glial 11βPGF_2α in the control of IEB repair via an increase in intestinal epithelial cell spreading (53) (Figure 3). Finally, we have demonstrated that another glial mediator, the 15-HETE, controls IEB paracellular permeability, both in vitro and in vivo (54) (Figure 3). Thus, EGCs represent a source of n-6 oxylipins responsible for IEB strengthening.

Inhibitors of both COX isoforms increased the contractions of human longitudinal muscles induced by electrical stimulation and are atropine-sensitive. It has been suggested that cholinergic neurons are thought to be primarily modulated by COX-1 activity, whereas COX-2 acts primarily at the muscular level to decrease muscarinic responses (50). Another study on rat duodenum also showed that inhibition of COX isoforms with indomethacin increased motility, probably via enteric nerve excitation (57). While the expression of COX1 and COX2 enzymes in neurons of myenteric ganglia of mouse or human colon, their role in the neuromuscular function is unclear. The inhibition of COX does not have the same effects on neuromuscular functions (peristalsis, spontaneous or Ach-induced activity, electrically-induced contractions, etc.…) depending on the species, but also according to the intestinal region (stomach, jejunum, distal colon, etc.) or the studied muscle layer (circular/longitudinal) (58). In addition, as enteric neurons can be the target but also the source of COX-pathway metabolites, and as many studies used COX inhibition or deletion without neuronal specification, it is difficult to conclude about the specific neuronal contribution (58) (Figure 3).

The LOX pathway has also been involved in motility regulation as a recent work demonstrate how the 12-HETE, can remodel the ENS and inhibit duodenal contraction (59).

The prostaglandins PGD₂, E₂, I₂ and F_2a as well as LTC₄ increased short-circuit current, which represents chloride secretion and activated submucosal enteric neurons. The secretory response was reduced by treatment with tetrodotoxin or atropine, suggesting an involvement of cholinergic submucosal neurons in this regulation (60–62). Nevertheless, the prostaglandin regulation of secretion through ENS could be more complex, being accelerated with high concentrations of PGE₂ (63) or sometimes presenting anti-secretory effect for rat colonic preparations treated with PGD₂ (64, 65).

Prostanoids and leukotrienes are mediator of communication between macrophages or mast cells and ENS, but the ENS to immune cell communication through n-6 oxylipins has not been described yet.

One study demonstrated that in a co-culture of mast cells with rat submucosal neurons, inhibition of 5-LOX, resulted in a reduction of the number of responding neurons (66). In the same way, in small intestine of guinea pig sensitized to cow’s milk and exposed to beta-lactoglobulin, was observed an increase in neuronal excitability, both suppressed by treatment with a COX or 5-LOX inhibitors (67). Finally, one study suggests the involvement of macrophage-like cells found in rat myenteric plexuses in the effect of IL1β and IL-6 on the suppression of prostanoid-mediated norepinephrine release (37).

The role of oxylipins in inflammatory bowel processes is multifaceted and unclear. Many studies associate n-6 with proinflammatory effects, and thereby an increased risk of developing IBD (8, 69–71). Conversely, many derivatives of ARA could also play an anti-inflammatory and pro-resolutive role (72), and treatment with COX inhibitors led to an exacerbation of the activity of the disease or induce relapse in IBD patients (73, 74). More recently, two eicosanoid synthesis pathway enzymes (PTGIS and PGD₂ synthase) were identified among genes dysregulated in fibrotic Crohn’s disease (75). In the same manner, the concentration of a cluster of 10 eicosanoids (PGF2_a, 15-HETE, TxB₂, 11β PGF2_α, PGE₃, 15dPGJ₂, 6kPGF_1α, PGE₂, PGD₂ and 8 isoPGA₂) were decreased in biospie supernatant form UC or CD patients, when biopsies came from a non-inflamed areas (76). These studies suggest that, while upregulated by acute inflammation, the defects in expression of synthesis enzyme and/or the low production of n-6 oxylipins are associated with chronic inflammation.

These changes of n-6 oxylipins production could be related to alterations of EGCs and enteric neurons observed in IBD patients. Indeed, in addition to neuronal structural alterations (77), increase in neuronal degeneration (78–81), and changes in neurochemical coding (82–85), neurons of the myenteric plexuses from active IBD patients expressed high level of COX2 that could contribute to the dysmotility observed in patients (86). Indeed, a decrease in the number of myenteric neurons per ganglion was observed after the induction colitis with TNBS in guinea pigs and especially an increase in COX2 expression and release of PGE₂, TXA₂ and LTB₄ in the neuromuscular layer, was associated with a decrease in colonic propulsion and an increase in the excitability of myenteric neurons (87). In addition to neurons, EGCs from IBD patients present structural alterations (80), changes in glial marker expression (77, 88, 89) and a decrease in the production of four glial lipid mediators: the 15-HETE, 18-HEPE, 15dPGJ₂ and 11βPGF_2α (53) (Figure 3). The defect in 15dPGJ₂ and 11βPGF_2α production was explained by a decrease in L-PGDS and AKR1C3 synthesis enzyme expression by EGCs from CD patients (53). The functional consequences of the decrease of 15-HETE or 11βPGF_2α productions by CD EGC have been identified as a loss of control of the IEB permeability or healing, respectively (53, 54) (Figure 3). Similarly, the decrease of 15dPGJ₂ could be linked to a loss of neuroprotection from oxidative stress (44) and a defect in the proliferation and differentiation of intestinal epithelial cells (52) (Figure 3).

Thus, while most of the studies are focused on an active role of n-6 oxylipins in the deleterious effect of intestinal inflammation, we have to reconsider their necessity in the resolution of inflammation (90, 91) and the possible involvement of their defects in the ENS as a contributing factor to the pathological mechanisms of IBD.

Other studies suggest a detrimental role of n-6 oxylipins that target the ENS in several infectious diseases, especially during parasitic infections.

The Trypanosoma cruzi infection leads to myenteric neuron cell death mainly in the esophagus and colon and results in megaesophagus and megacolon, respectively (92). A treatment with low-dose aspirin during the acute phase of infection, known to inhibit COX, reduced parasitemia and protects myenteric neurons from cell death and plastic changes, notably their hypertrophy. This protection of cholinergic neurons was also associated with a recovered transit time by the aspirin treatment. In addition, treatment with aspirin decreases substance P and increases vasoactive intestinal peptide (VIP) productions, and could therefore help to regulate inflammation in these animals.

The infection with the parasite Cryptosporidium in piglet, induces intestinal villous damages associated with malabsorption and reduced sodium chloride (NaCl) absorption due to prostaglandin released from the inflamed tissue. The prostaglandins alter NaCl transport in this infection primarily by stimulating cholinergic interneurons that innervate VIPergic and cholinergic motor nerves (93). The ENS may be a potential target for pharmacological control of the acute diarrhea in this infection.

The involvement of n-6 oxylipins has also been suggested in many other pathologies associated with intestinal disorders, particularly via their role on neuronal excitation or their inhibitory role on intestinal contractility.

Patients with irritable bowel syndrome (IBS) are characterized by abdominal cramps, diarrhea or constipation. Treatments with the PGD₂ antagonist BW A868C inhibited the longitudinal muscle contractions potentiated by supernatant of biopsies from IBS patient (94). In addition to this effect on motility, lipidomic studies have demonstrated that n-6 oxylipins can participate in visceral sensitivity. Quantification of the PUFA metabolites produced in biopsies from control and IBS patients have shown that 5,6-EET and PGE₂ production are increased in diarrhea-predominant IBS patients and 5-oxoeicosatetraenoic acid (5-oxoETE) in IBS with predominant constipation (95, 96). These productions are correlated with pain and bloating and these oxylipins can activate afferent sensory neurons but the contribution of ENS in these processes is not described.

In the same way, about half of the patients with slow-transit constipation exhibits an elongated colon and outlet obstruction. One study showed increased levels of COX2 in muscularis and mucosa of partially obstructed mice, associated with colon elongation, slowing of transit, and loss of mucosal reflexes. The expression of COX2 was also found in subpopulations of nNos myenteric neurons, and PGE₂ application completely abolished spontaneous inhibitory junction potentials and depolarize the circular muscle, suggesting that it was removing tonic inhibition. Removal of the obstruction or treatment with a selective COX2 inhibitor restored control parameters as spontaneous inhibitory junction potentials and mucosal reflexes (97). This study suggested that an increase in COX2 expression led to a hyperexcitable colon, largely due to suppression of inhibitory nerve pathways by prostaglandins.

In the case of food allergies, products of activated mast cells, including prostanoids and leukotrienes, are also known to stimulate nerve endings, to modulate neuronal excitability and to increase synaptic transmission, resulting in phenotypically and functionally altered submucosal secretomotor neurons (98).

In the context of postoperative ileus, a study showed, in the rat, that a bowel manipulation increased the expression of COX2, associated with an increased PGE₂ release in the immune cells but also in subpopulation of myenteric neurons. This associated with a decrease in contractile smooth muscle activity (99). COX2 metabolites such as PGE₂ produced by the ENS appear to be involved in enteric dysmotility, and administration of specific inhibitors could help to prevent the transit delay induced by bowel manipulation (58). The PGE₂ produced by the ENS is not only involved in smooth muscle control, but also in epithelial proliferation, and a study nicely shown how tumor cells can induce PGE₂ production by EGCs to activate colon cancer stem cells and stimulate tumorigenesis (100).

Finally, some ARA derivatives could also have an emetogenic role, and one study investigated the role of leukotrienes in this effect. Injections of different leukotrienes, such as LTC₄ and D₄ induced an increase in the frequency and percentage of least shrews receiving vomiting. Following LTC₄-induced vomiting, they showed increased labeling of nuclei for Fos in ENS, suggesting that leukotrienes modulate emetic activity in part via enteric nervous components activation (101).