Animal studies have led to significant advances in our understanding of pain mechanisms in the intestine that could lead to altered signaling in disorders such as irritable bowel syndrome. However, how these translate to the human afferent nervous system is unclear. Recent studies have demonstrated that it is possible to use a variety of techniques, including electrophysiological recordings, to begin to examine these concepts in humans. This mini-review examines these studies to explore how well animal studies translate to humans suffering from irritable bowel syndrome, highlights some of the advantages and technical limitations of these approaches, and identifies some priorities for future studies using human tissues.
- irritable bowel syndrome
- human afferent nerves
- dorsal root ganglia neurons
- visceral hyperalgesia
abdominal pain is a major cause of morbidity in chronic gastrointestinal (GI) diseases such as irritable bowel syndrome (IBS) but the mechanisms are poorly understood. Visceral hypersensitivity, a common mechanism that underlies pain in IBS, signifies increased sensitivity of nerves in response to stimuli within visceral organs such as the intestine. This results in exaggerated pain signaling due to noxious stimuli (hyperalgesia) or previously nonnoxious stimuli (allodynia). While it can result from either peripheral mechanisms (“bottom up” from the intestine) or central mechanisms (“top down” from the brain), there is growing interest in peripheral mechanisms in humans because of rapidly accumulating evidence that changes in tissue mediators and/or receptors and ion channels cause dysregulated signaling from peripheral nociceptive nerves in IBS (5). Moreover, these peripheral mechanisms provide attractive targets for pain management given their relative accessibility and potential for fewer side effects compared with those in the central nervous system. Despite these advances, most of our data on these mechanisms is based on animal studies and it is unclear whether these translate to humans. This knowledge gap is largely due to the relatively inaccessible human visceral nervous system for mechanistic studies. Recently, however, there have been significant advances and this review highlights emerging data based on these techniques to study peripheral visceral pain signaling in humans. The goal is to highlight new findings in the context of existing animal studies that are relevant to the irritable bowel syndrome, as well as potential controversies and knowledge gaps.
Mechanisms of Sensitization of Afferent Neurons
Animal studies have provided a rapidly growing understanding of pain signaling in the intestine and have been reviewed elsewhere in detail (44). Briefly, peripheral nociceptive (sensory information leading to generation of painful sensations in the brain) signals are relayed by dorsal root ganglion (DRG) neurons. The distal axons of these pseudo-unipolar neurons innervate multiple layers of the intestine and their proximal terminals synapse in the spinal cord (Fig. 1). They have unmyelinated axons whose terminals appear to lack specialized endings. The nociceptive DRG nerves are thought to innervate the serosa and travel into the intestine with arterioles, thereby projecting into the submucosa. They can be activated by distention of the bowel and possible chemical triggers, but there is ongoing controversy around the transduction mechanism including the “adequate stimulus” for these nerves. This is highlighted by our ability to perform large polypectomies on awake patients, where 4–5 cm of tissue can be excised from the mucosa and submucosa, without them experiencing any discomfort.
Animal studies have identified a wide array of tissue mediators that can sensitize nerves and many are thought to be potentially important in IBS (6). These can be generated by immune cells (especially mast cells) and may be triggered by neurohumoral signaling from the central nervous system or from the mediators originating in the intestinal lumen (44). The neuroimmune mediators includes inflammatory/immune (e.g., histamine, proteases, and bradykinin), neurotrophins (e.g., nerve growth factor), and cytokines (e.g., tumor necrosis factor-α) and these are released by a number of immune cells (Fig. 1) (29, 44). Conversely, immune cells can also release inhibitory factors such as opioids that decrease visceral sensitivity (29, 44). Luminal signaling could be mediated from multiple sources, including bacteria-derived cell products, their metabolome, and dietary factors such as short-chain fatty acids or food antigens. Changes in mucosal permeability may provide access to the immune compartment and/or alter mucosal signaling at the basolateral membrane, such as changes in 5-HT released by enterochromaffin cells or release of mucosal proteases (19, 24).
Recently, there is growing evidence that the microbiota can signal directly to nociceptive neurons in the intestine under certain conditions. Toll-like receptors have been identified in DRG neurons innervating the intestine and their activation increases neuronal excitability (37). Pertubation of the mouse microbiome with antibiotics can induce visceral hyperalgesia, an effect reversed by a probiotic (45). Additionally, a recent animal study showed that a specific probiotic reduces intestinal afferent firing mostly by inhibiting the transient receptor potential (TRP) channel TRPV1 (39), but it is unknown if these actions are relevant to human disorders such as IBS.
Ultimately, mediators modulate ion channels on nociceptive nerves causing neuronal sensitization. As shown in Fig. 1, activation of G-protein coupled receptors [GPCRs; e.g., proteases acting on protease-activated receptor (PARs) or histamine on H1 receptors], ligand-gated ion channels (e.g., 5-HT3 receptors and P2X purinoreceptors), voltage-gated Ca2+ channels, or mechanosensitive cationic channels (especially TRPV1, TRPV4, and TRPA1 channels) depolarizes nociceptive nerve terminals producing generator potentials that initiate action potentials when the membrane potential threshold is reached. Signaling from many GPCRs can also modulate properties of these channels leading to greater depolarizing currents for a given stimulus, thereby amplifying action potential discharge. In addition, activation of GPCRs can increase voltage-gated Na+ channel conductance and/or decrease K+ channel conductance involved in action potential electrogenesis, leading to a decrease in action potential threshold and/or greater action potential number and firing rate (9). These changes reflect a balance of excitatory and inhibitory mediators and can result from conformational changes in ion channels (e.g., phosphorylation), or transcriptional and/or posttranscriptional modifications.
The Study of Human Tissue
Until recently most of our understanding of altered pain signaling in IBS patients has been derived from rectal barostat measurements (distension of a balloon within the rectum to different pressures with patient reporting of sensations ranging from “just sensible” to “maximum pain,” and often coupled with brain imaging). This approach has served to demonstrate that altered sensory signaling such as visceral hyperalgesia is a common finding in IBS patients (40) but is hampered by an inability to perform mechanistic studies, especially in the intestine. Electrophysiological studies in animal models, such as patch-clamp recording, Ca2+ imaging, and extracellular afferent nerve recording, have yielded detailed information regarding the cellular properties of mechanoreceptors that underlie these distention-induced responses. Fortunately, many of these techniques have now been adapted for study in humans, as shown in Table 1 and Fig. 2. These human studies have served to support postulated mechanisms but also identify new mechanisms or heighten the relevance for some in the pathophysiology of exaggerated pain signaling in IBS.
Immunohistochemical studies have examined the innervation of the human colon as well as changes in nerve fiber density, ion channel, or receptor expression in GI disorders. A pan-neuronal marker revealed innervation in all layers of the gut (i.e., from mucosa to mesentery) (34). Immunoreactivity for calcitonin related gene-related peptide (CGRP) and MRGPRD, both markers for nociceptors based on animal studies, revealed CGRP fibers associated with large blood vessels in the mesentery and MRGPRD fibers associated with small blood vessels within the mesentery and submucosa (34). These CGRP-positive fibers associated with mesenteric vessels are likely nociceptors in humans as well (12), but this is yet to be confirmed.
Mediators that sensitize nociceptive nerves.
Tissue supernatants derived from colonoscopic biopsies obtained from IBS patients have enabled identification of key tissue mediators and the study of their effects on the properties of nociceptive DRG neurons (see Ref. 36 for detailed review). Two groups [Barabara et al. (7) and Cenac et al. (17)] were the first groups to use this technique to demonstrate that acute application of IBS colonic biopsy supernatants excited mesenteric afferent nerves and increased intracellular calcium in DRG neurons from animals, compared with supernatants from healthy controls. These supernatants have also been shown to sensitize these neurons. For example, action potential firing evoked by a mechanical stimulus in high-threshold afferent nerves in the mouse distal colon is increased following acute application of the IBS supernatant (4). In DRG neurons, overnight incubation with biopsy supernatant from IBS with predominant diarrhea (IBS-D) patients decreased the rheobase (current required to generate a single action potential) thus demonstrating an increase in the intrinsic excitability in these neurons by IBS supernatant (42). This effect in DRG neurons was not seen in protease-activated receptor 2 (PAR2) knockout mice demonstrating a role for proteases in this sensitization (42). Other studies using fecal supernatants from IBS patients with predominant constipation (IBS-C) also suggest proteases could sensitize nerves but via a PAR2-independent mechanism (3).
There is also evidence that mediators from peripheral blood mononuclear cells (PBMC) of IBS patients can increase DRG excitability, based on studies of mechanosensitive responses (27). However, the TNF-α antibody inhibitor infliximab did not fully abolish this increased mechanosensitivity, suggesting multiple cytokines may be involved (27). Studies of human PBMCs have also revealed that mediators can decrease sensitization of nerves in the gut. In these studies, PBMC supernatants from healthy subjects inhibited afferent firing to colon wall stretch in a mouse model of visceral hypersensitivity (28). A similar effect was seen with IBS-C PBMC supernatant but not IBS-D supernatant; this effect was blocked with a μ-opioid receptor antagonist. Another inhibitory mediator identified in human tissues is extracellular cyclic guanosine monophosphate (cGMP). Exogenous application of cGMP reduces excitability of human DRG neurons (16), similar to that observed in animal studies (15). Extracellular cGMP can be increased in the colonic epithelium by the drug linaclotide acting on apical guanylase cyclase receptors on enterocytes (10) and is currently used to treat IBS-C. Drugs acting on guanylate cyclase C (GC-C) decrease pain in IBS patients, and it is postulated this occurs by inhibitory effects on nociceptive DRG neurons (10).
Human studies support the evidence from animal studies that TRP channels are key cellular pathways underlying pain signaling from the gut and that they exhibit considerable plasticity under a number of conditions (11). In humans, altered TRPV1 expression on nerve fibers in the distal gut has been reported, although the results have been inconsistent. For example, Akbar et al. (2) found an increase in TRPV1 immunoreactive fibers in biopsies of IBS patients compared with controls and this correlated with degree of abdominal pain. Similarly, increased TPRV1 expression was observed in patients with quiescent inflammatory bowel disease (IBD) with IBS symptoms compared with patients with quiescent IBD without IBS symptoms as well as control subjects (1). In contrast, van Wanrooij et al. (43) found no difference in TRPV1 expression in biopsies from IBS patients compared with control subjects. Nonetheless, IBS patients with visceral hypersensitivity evoked by rectal balloon distension had a significantly greater pain response to rectal application of capsaicin compared with both normosensitive IBS patients and control subjects; pain perception to capsaicin application was significantly associated with abdominal pain symptom scores (43). Additionally, it has been demonstrated that TRPV4 may sensitize afferents innervating the human colon as application of a TRPV4 antagonist decreased mechanosensitivity (33). Together, this suggests that TRP channels are important in nociceptive signaling in the human colon and their signaling may be exaggerated in IBS.
These TRP channels may be sensitized by signaling from proteases and other tissue mediators acting directly on the neuron or through indirect pathways. For example, proteases released by IBS patients can sensitize TRPV4 (18) through indirect pathways. Exposing IBS supernatants to mouse DRG neurons caused an increase in the TRPV4 agonist 5,6-epoxyeicosatrienoic acid (5,6-EET); this was inhibited in PAR2 knockout mice suggesting that proteases induce an increase of 5,6-EET within nociceptive neurons that could then act on TRPV4 (18). Reports of a histamine-TRPV1 interaction is an example of a direct pathway. In these studies, incubation of biopsy supernatants from IBS patients with isolated murine DRG neurons increased the intracellular calcium response to the TRPV1 agonist capsaicin and this potentiation was blocked by a H1 receptor antagonist (46). Taken together, TRP channels are sensitized by multiple mediators obtained from human tissue, ultimately resulting in exaggerated pain signaling in IBS.
Sensitization of voltage-gated ion channels by IBS tissue mediators will also increase excitability of nociceptive neurons. IBS-D supernatant decreased the rheobase and increased action potential discharge (reflecting changes in voltage gated Na/K channels underlying action potential electrogenesis) in mouse DRG neurons, an effect that was blocked in PAR2 knockout mice (42). Voltage-gated calcium channels appear to be an additional target of these tissues mediators (14).
Signaling by the microbiome.
Manipulation of the microbiome with either antibiotics (35) or probiotics (23) has had some success in IBS treatment suggesting a potential role for the microbiome in symptom generation. However, a “signature” microbiota specific to IBS has not been identified (8) despite numerous reports of an altered microbiome in IBS patients.
Although human studies are limited, transplanting fecal micriobiota from IBS patients into germ-free animals has provided important data. For example, when fecal microbiota from IBS patients was transferred to germ-free rats, it caused visceral hypersensitivity compared with rats receiving microbiota from healthy controls or conventional rats (20). Increased hydrogen gas production and dissolved sulfides were seen in IBS microbiota rats suggesting bacterial metabolites may be important signaling molecules underlying visceral sensitivity (20). Furthermore, the human microbiome is capable of producing proteases (32) and histamine (41) suggesting multiple factors produced by the microbiome could also sensitize sensory nerves within the gut.
The study of human tissues have heightened interest in changes to nerve fiber density in IBS patients (2, 22, 48) although this has not been observed in all studies (43). Nerve growth factor (NGF) immunoreactivity (22) and brain-derived neurotrophin factor (BDNF) concentration (48) are reported to be elevated in IBS patients, in keeping with the reported changes in nerve fiber density. Additionally, Dothel et al. (22) demonstrated an increase in both tyrosine kinase receptor A (TRKA), the preferred receptor for NGF, and mucosal growth-associated protein 43 (GAP43), which is linked to neurite elongation, in IBS patients. Taken together, this suggests there is increased nerve fiber density, likely a combination of extrinsic nociceptive nerves and intrinsic enteric nerves, at least in some IBS patients likely as a result of increased expression of nerve growth factors.
The Importance of Patient Phenotyping
Studies of specific mediators in subsets of IBS patients highlight the importance of studying human tissues. For example, a small study suggests increasing age resulted in altered sensitivity of human afferents (47). Afferents were less spontaneously active and sensitivity to bradykinin was reduced. This was contrasted by the finding that mast cells were in closer association with nerve fibers with increasing age suggesting the potential for greater neuroimmune interactions (47). The clinical phenotype of the patient may also be important. For example, some studies report sensitization of DRG neurons by supernatants from IBS-D but not IBS-C. In addition, whether an IBS patient has visceral hypersensitivity (determined by barostat measurements) (13) may predict how murine DRG neurons respond to biopsy supernatants from IBS patients. Finally, within a given patient subtype (i.e., IBD-C patients), there may also be differences in responses. For example, preliminary studies of the expression of GC-C in IBS patients appear to vary, potentially underlying the variable response to drugs (25). Together these studies highlight the importance of patient phenotyping given the heterogeneity of IBS symptoms (e.g., IBS-D vs IBS-C), but deeper phenotyping is likely needed to fully understand the full range of mechanisms leading to symptom expression (e.g., IBS-C patients with visceral hypersensitivity vs normosensitivity determined by barostat or IBS-D patients identified by their exaggerated symptom response to a diet challenge rich in poorly digested carbohydrates; Refs. 13, 31).
Limitations of the Study of Human Tissues
While IBS supernatant studies have yielded important translational discoveries, there are several important caveats. First, how long biopsies are processed to generate supernatants could affect the proportions of mediators. Second, it is important to recognize that the exposure time of the supernatants with isolated neurons (36) or intestine used in extracellular afferent nerve recordings can dictate which cellular mechanisms underlie neural activity (i.e., minutes can activate nerves; minutes vs. hours to examine sensitization of nerves). For human tissue harvest from the operating theater (either intestine or DRG neurons), some signaling mechanisms will be more sensitive to ischemic time than others (e.g., G protein-coupled receptor mediated greater than sensitivity than ligand gated). This may be due to intraoperative time to complete the resection once tissue blood supply is clamped (greater in a laparascopic than an open colonic resection), time required for the release of tissue from the pathologist (in proper transport medium), and ultimately when the experiments are eventually initiated by the investigator. Colonic resections or human DRG neurons are more readily obtained from patients who do not have IBS (e.g., resection for colorectal cancer in patient without GI symptoms). This may result in a different expression of receptors for important mediators previously identified in IBS tissues thus making interpretation of the results with respect to IBS challenging (21). Finally, it is not possible to determine if human DRG neurons used in experiments innervate the GI tract vs. other visceral or somatic organs.
Recent studies using human tissues have advanced our understanding of important factors involved in sensitizing nociceptive nerves in patients with IBS and attention to several areas could accelerate this potential. Protocols for accessing and processing human tissue (e.g., biopsy supernatants, surgical resections) need to be standardized across laboratories to ensure that data are comparable. The development of cell lines from human intestinal tissues and DRG neurons will help to address concerns about the viability of tissue samples. The studies of human tissues should continue to be driven by both translational (use of discoveries from animal studies) and “reverse translational” studies (mechanisms implied by clinical studies). Finally, the early use of human tissues in the drug discovery pathway is needed to prioritize targets with the greatest therapeutic potential and translational relevance.
No conflicts of interest, financial or otherwise, are declared by the authors.
D.E.R. and S.J.V. prepared figures; D.E.R. and S.J.V. drafted manuscript; D.E.R. and S.J.V. edited and revised manuscript; D.E.R. and S.J.V. approved final version of manuscript.
We thank Dr. Kaede Takami for assistance with the preparation of the figures in this manuscript.
- Copyright © 2017 the American Physiological Society