Altered gastrointestinal (GI) function contributes to the debilitating symptoms of inflammatory bowel diseases (IBD). Nerve circuits contained within the gut wall and outside of the gut play important roles in modulating motility, mucosal fluid transport, and blood flow. The structure and function of these neuronal populations change during IBD. Superimposed on this plasticity is a diminished responsiveness of effector cells — smooth muscle cells, enterocytes, and vascular endothelial cells — to neurotransmitters. The net result is a breakdown in the precisely orchestrated coordination of motility, fluid secretion, and GI blood flow required for health. In this review, we consider how inflammation-induced changes to the effector innervation of these tissues, and changes to the tissues themselves, contribute to defective GI function in models of IBD. We also explore the evidence that reversing neuronal plasticity is sufficient to normalize function during IBD.
functio laesa, the disturbance of normal function, is one of the cardinal signs of inflammation and along with pain contributes to the majority of symptoms of inflammatory disease. During gastrointestinal (GI) inflammation the control of smooth muscle cells, enterocytes, and vascular endothelium is dysregulated leading to altered motility, blood flow, and water balance. Research over the past decade has elegantly delineated the cellular mechanisms of plasticity in enteroendocrine (EE) cells and neural pathways (7, 26, 41, 42) in the inflamed bowel and proposed how these changes may lead to altered function. However, at the same time as inflammation is altering the innervation of the gut, the effector cells of the muscularis externae, mucosa, and vasculature are also changed in ways that often diminish their responsiveness to released neurotransmitter. Therefore, it is reasonable to question whether reversing changes to the innervation of the inflamed bowel would normalize function given that the effector tissues themselves would presumably remain dysfunctional. In this review, we describe enteric neural pathways and how the effector components are changed in models of inflammatory bowel disease (IBD) and explore the evidence that reversing neuronal plasticity can restore normal gut function.
Neural Pathways to GI Effector Tissues
Effector cells in the gut are in the muscularis externae, mucosa, and vasculature. Each tissue is innervated from two or more populations of neurons. The muscularis externae is composed of longitudinal and circular smooth muscle layers that are each innervated by separate populations of inhibitory and excitatory motor neurons from the myenteric plexus. Similarly, the enterocytes and goblet cells of the mucosa are innervated by two or more types of secretomotor neurons from the submucosal plexus. The submucosal vasculature is innervated by enteric and extrinsic afferent vasodilator neurons and sympathetic vasoconstrictor nerves (59).
The function of these effector tissues is regulated by nerve circuits that are activated via chemical or mechanical stimulation of the GI mucosa (3) or stretch of the intestinal wall (31, 55). The mucosal EE cells (Fig. 1) respond to sensory stimuli such as short-chain fatty acids, glucose, or mechanical deformation of the villi (3) and release a variety of gut hormones such as PYY, GLP-1, and CCK (32). The most common are the serotonin (5-HT)-releasing enterochromaffin cells; they also seem to respond to the broadest range of stimuli (3). Activation of luminal and basolateral 5-HT receptors can have important effects on propulsive motility, visceral sensation, and immune system activation (18, 20, 26, 56).
Release of gut hormones from the EE cells activates enteric sensory neurons [intrinsic primary afferent neurons (IPANs); Fig. 1] which respond with a burst of action potentials and/or a slow excitatory postsynaptic potential associated with action potential firing. The information is processed by the IPANs (4) and passed on to interneurons and motor neurons via a burst of fast excitatory postsynaptic potentials (EPSPs) mediated mainly by nicotinic receptors or a slow excitatory postsynaptic potential mediated mainly by neurokinin receptors (16). Activity in these pathways modulates motility, enterocyte Cl− secretion, and blood flow (Fig. 1).
Neuromodulation of the muscularis externae.
The movements of the GI smooth muscle layers that underlie motility reflexes for mixing, and for pendular and propagating contractions reflect complex interactions between myocytes, interstitial cells and the enteric nerve circuits that innervate them. The major excitatory neuromuscular transmitters are acetylcholine (ACh) and substance P whereas nitric oxide, purines, and vasoactive intestinal peptide (VIP)/pituitary adenylate cyclase activating peptide (PACAP) are the major inhibitory neurotransmitters (Fig. 1) (17). The influence of inhibitory and excitatory junction potentials (IJPs and EJPs) on smooth muscle excitability is superimposed on pacemaker activity generated in interstitial cells of Cajal, which is conducted into the smooth muscle layers (27). Recent work from Sanders and colleagues (54) has defined how enteric motor neurons modulate the syncytium of smooth muscle cells, intramuscular interstitial cells, and fibroblast-like cells. Innervation of the smooth muscle cells, interstitial cells, and platelet-derived growth factor α+ cells (the SIP syncytium) by enteric motor neurons can modulate smooth muscle action potential discharge in response to slow waves in phasic muscles and cause contraction or relaxation in tonic muscles.
Effects of colitis.
Models of colitis in guinea pigs and rodents result in a loss of coordinated contractile patterns in inflamed regions associated with enteric neuronal loss, axon sprouting, and an increase in the excitability of enteric nerve circuits (41, 46). In guinea pigs, where electrophysiological experiments have documented the changes in enteric neural excitability and synaptic transmission during trinitrobenzene sulfonic (TNBS)-induced colitis and ileitis, propulsive motility in the colon is impaired (7, 41). This has been compared with a form of enteric attention deficit disorder: hyperexcitability of enteric nerve circuits leads to the loss of the spatiotemporal coordination of smooth muscle movements required for normal propulsion of luminal contents (41). Although detailed discussion of enteric glial cells is beyond the scope of the present review, recent findings have highlighted a role for them in modulating motility (43). Since enteric glia are altered during colitis (22), inflammation-induced glial plasticity may contribute to altered motility during colitis. In addition, models of colitis also cause impairments of nitrergic (21), purinergic (53), and cholinergic (12) neurotransmission (Fig. 2). Furthermore, damage to interstitial cells of Cajal (30), changes to smooth muscle responsiveness to neurotransmitters (28), and altered activity of myocyte ion channels (1) during colitis impair the ability of effector cells to respond to the nervous system. Therefore, it is possible that altered motility in models of colitis is the result of neuronal plasticity, impaired SIP function, or a combination of both.
Experiments in guinea pigs with TNBS colitis indicate that drugs that reverse enteric neuronal plasticity restore colonic propulsive motility in animals with colitis. Hyperpolarization-activated cation (HCN) channels are found on many functional types of enteric neuron but are especially prominent on the IPANs (61). TNBS colitis leads to hyperexcitability of IPANs due, in part, to an increase in activity of HCN channels (34). Two different drugs that block HCN channels in IPANs are able to restore in vitro propulsion of colonic contents when applied to colons from guinea pigs with colitis (25). The authors importantly demonstrated that neither drug has any direct effect on the SIP syncytium, suggesting that the restoration of motility is only due to neuronal effects. Other experiments have examined whether drugs that mimic the effect of colitis on the excitability of IPANs can recapitulate the effects of colitis on motility. Supporting the idea that IPAN dysfunction is at the crux of the motility disturbances during colitis, pharmacologically blocking the intermediate conductance Ca2+-activated K+ channel on IPANs causes a similar disruption of motility to that seen during TNBS colitis (15, 51).
Complementary evidence supporting the importance of neuronal changes in motility defects during colitis comes from a study addressing the role of pannexin-1 hemichannels in myenteric neuron loss (Fig. 2) during colitis (9, 21). Gulbransen and colleagues (21) found that probenecid in vivo prevented the ~20% loss of myenteric neurons that normally occurs in a mouse model of DNBS-induced colitis. Furthermore, they demonstrated that this restored the inhibitory and excitatory innervation of the circular smooth muscle. While probenecid may act on other transporters in addition to pannexin-1 hemichannels, and may also impact nonglial cells, this study suggests that preventing neuronal loss during colitis may prevent the impairment of motility regulation in colitis. In human IBD, it is unlikely that prevention of neuronal loss will be possible, so reversal of neuronal loss may have value in restoring normal motility. In this regard, regenerative medicine strategies to implant stem cell-derived enteric neurons in the areas with reduced numbers of neurons may be of benefit (5, 10), as may therapies that augment the regenerative capacity of the enteric nervous system (ENS) (35, 57).
Together, these findings suggest that, despite impairments to neuromuscular transmission, interstitial cells, and myocytes during colitis, restoration of normal excitability patterns in the ENS or preventing the loss of enteric neurons during colitis is capable of restoring normal motility patterns.
Secretion of Cl− from enterocytes, which is accompanied by the paracellular movement of Na+ and water, is tonically regulated by the autonomic nervous system. Enteric secretomotor neurons have cell bodies primarily in submucosal ganglia and are innervated by IPANs and myenteric interneurons (Fig. 1). Secretomotor neurons evoke mucosal Cl− secretion via release of ACh or VIP, which leads to activation of Ca2+ activated Cl− channels or CFTR, respectively. This neurogenic secretion compensates for normal colonic water and salt absorption; disruption of this balance leads to diarrhea, which is a common symptom of IBD. Sympathetic postganglionic neurons inhibit electrolyte secretion from enterocytes indirectly, by inhibiting enteric secretomotor circuits (38) (Fig. 2).
Effects of colitis.
IBD models have provided insight into the impact of intestinal inflammation on electrolyte homeostasis. Both the TNBS and DSS models of chemically induced colitis, and the IL-10 knockout mouse model of IBD, lead to impaired secretagogue-induced Cl− secretion from enterocytes (2, 39) (Fig. 2). This occurs despite axonal sprouting in the mucosa, hyperexcitability of IPANs, enhanced excitatory synaptic drive to secretomotor neurons and a reduction in the release of sympathetic neurotransmitters during colitis (36, 45, 52). This impairment in Cl− secretion appears to be due to the actions of prostaglandin (PG) D2, downstream of COX-2 induction, as PGD2 recapitulated the effect of colitis on enterocyte function, and in vitro administration of COX-2 inhibitors restored normal colonic Cl− secretion (62). Enteric glia appear to also play an important role in inflammation-associated epithelial dysfunction. Inhibition of enteric glial cells in vivo with fluoroacetate restored epithelial secretory and, importantly, barrier function in mouse models of IBD (39). MacEachern and colleagues (39) further demonstrated a role of inducible nitric oxide synthase (iNOS), presumably from enteric glia, in colitis-associated epithelial dysfunction. Inhibition of iNOS but not nNOS in vitro restored neurogenic Cl− secretion.
Taken together these data suggest that normalization of neuronal function may not be sufficient, on its own, to restore secretomotor function. However, this has not yet been directly tested. What seems clear is that enteric glial production of toxic concentrations of NO may contribute to the loss of epithelial responsiveness to neural secretagogues in models of colitis. Thus inhibition of glial iNOS activity may prevent defective water and electrolyte secretion during colitis.
The blood supply of the GI mucosa is tightly regulated to meet the metabolic demands of the epithelium and to efficiently transport digested nutrients. As with the motility pathways, mucosal stimuli initiate intrinsic and extrinsic neural reflexes that modulate blood flow into mucosal capillaries by effects on submucosal arterioles (Fig. 1) (59). Sympathetic vasoconstrictor neurons provide noradrenergic and purinergic vasoconstrictor innervation of vascular smooth muscle whereas enteric neurons and extrinsic afferent neurons provide vasodilator innervation of vascular endothelial cells Enteric vasodilator neurons are important components of submucosal reflexes that coordinate mucosal blood flow with fluid secretion (16).
Effects of colitis.
Blood flow to the mucosa is impaired during IBD, due to angiogenesis and loss of vasoregulation (11, 23, 24). This is thought to contribute to pathogenic hypoxia that accompanies IBD (58). In tissues from IBD patients, endothelium-dependent vasodilation, which is required for the effects of ACh and neuropeptides on blood flow, is markedly reduced (24). In contrast, the vasodilator effects of nitric oxide donors are not affected, suggesting the defect lies in the ability of the endothelium to produce vasodilator substances, rather than an inability of smooth muscle cells to respond to endothelial-derived vasodilators (24). Experiments in the DSS model of colitis in mice found a similar defect in endothelium-dependent vasodilation due to free radical-induced damage to endothelial cells (44). Taken together these findings suggest that, despite increased activity in vasodilatory enteric and extrinsic afferent pathways (46), released neurotransmitters are unable to evoke release of vasodilator substances from endothelial cells in the inflamed gut (Fig. 2).
Sympathetic vasoconstrictor regulation of submucosal arterioles is also altered in models of IBD. The release of norepinephrine from sympathetic axons within the gut is reduced in mouse and rat models of colitis (29, 45). This appears to be due to a combination of effects on presynaptic inhibitory α2-adrenoceptors (6) and inhibition of N-type voltage-gated Ca2+ channels (45). However, the pathophysiological consequences of these changes are as yet unclear because norepinephrine release was measured from whole tissues in response to electrical field stimulation (EFS) or KCl-induced depolarization. Whole tissue concentrations of norepinephrine may not reflect its concentration at individual neuroeffector junctions due to the intermittent nature of norepinephrine release from individual varicosities (8). Furthermore, the stimulation parameters used are likely to have obscured the impact of sympathetic visceromotor neuronal hyperexcitability (14, 33) on norepinephrine release.
In vitro experiments in preparations of submucosal arterioles from the colons of mice treated with TNBS have identified a selective loss of the ability of purines to cause vasoconstrictions during colitis (37). In contrast, noradrenergic vasoconstrictions are not affected (37, 48). The selective defect in purinergic vasoconstriction appears to be due to upregulation of an enzyme that degrades extracellular purines (48). Accordingly, when hydrolysis-resistant purines are applied to vessels they do cause vasoconstriction. Therefore, the inhibition of purine hydrolysis may restore normal vasoconstrictor regulation as the ability of vascular smooth muscle cells to contract to purines remains intact during colitis.
In summary, colitis reduces the responsiveness of the vasculature to released transmitters through free radical-induced damage to vascular endothelium, inhibition of norepinephrine release, and enhanced catabolism of purine neurotransmitters. Future studies should address whether reversing neuroplasticity in vasomotor pathways can overcome these changes and restore vascular function during colitis.
The neurons and glial cells of the enteric nervous system and its extrinsic connections regulate important GI functions that are impaired during colitis. This review has considered the evidence from animal models of IBD that impaired motility, fluid secretion, and blood flow are primarily due to effects of inflammation on the effector cells or their innervation. These processes can be seen summarized in Fig. 2. Based on the evidence to date, it appears that reversal of neuronal plasticity may be able to restore motility and prevention of glial changes may restore mucosal function in models of IBD. In contrast, vascular function seems less likely to be restored by reversing neuronal or glial plasticity. However, studies have yet to be published that definitively test whether reversing neuroplasticity in secretomotor or vasomotor neurons can prevent or overcome secretory of vascular effector cell defects to normalize function. Another important caveat to these studies in animal models is that of the translatability of the mechanisms identified to IBD patients. Future studies that bridge this gap could be very valuable. Although beyond the scope of the present review, there is increasingly strong evidence that the nervous system can modulate mucosal barrier function (50) and the severity of GI inflammation (19, 40, 47, 60). This suggests that, in addition to restoring normal function, neuromodulatory drugs or electroceutical interventions (13) may reduce the severity of inflammation in IBD and thereby reduce symptoms.
Research in the Lomax laboratory on IBD-related neuroplasticity is funded by Crohn’s and Colitis Canada.
No conflicts of interest, financial or otherwise, are declared by the author(s).
A.E.L. drafted manuscript; A.E.L., S.P., and P.P.B. edited and revised manuscript; A.E.L., S.P., and P.P.B. approved final version of manuscript; P.P.B. prepared figures.
1 This article was invited as part of a series highlighting AJPGI-sponsored researchers at Experimental Biology 2016: San Diego, CA.
- Copyright © 2017 the American Physiological Society