Receptors and Transmission in the Brain-Gut Axis:

Transcription

1 Am J Physiol Gastrointest Liver Physiol 280: G787 G794, Receptors and Transmission in the Brain-Gut Axis: Potential for Novel Therapies I. Receptors on visceral afferents themes ANTHONY J. KIRKUP, ALAN M. BRUNSDEN, AND DAVID GRUNDY Department of Biomedical Science, University of Sheffield, Western Bank, Sheffield S10 2TN, United Kingdom Received 22 February 2001; accepted in final form 22 February 2001 Kirkup, Anthony J., Alan M. Brunsden, and David Grundy. Receptors and Transmission in the Brain-Gut Axis: Potential for Novel Therapies. I. Receptors on visceral afferents. Am J Physiol Gastrointest Liver Physiol 280: G787 G794, Visceral afferents are the information superhighway from the gut to the central nervous system. These sensory nerves express a wide range of membrane receptors that can modulate their sensitivity. In this themes article, we concentrate on those receptors that enhance the excitability of visceral afferent neurons. Some receptors are part of a modality-specific transduction pathway involved in sensory signaling. Others, which are activated by substances derived from multiple cellular sources during ischemia, injury, or inflammation, act in a synergistic fashion to cause acute or chronic sensitization of the afferent nerves to mechanical and chemical stimuli. Such hypersensitivity is the hallmark of conditions such as irritable bowel syndrome. Accordingly, these receptors represent a rational target for drug treatments aimed at attenuating both the inappropriate visceral sensation and the aberrant reflex activity that are the foundation for alterations in bowel function. spinal afferents; sensitization THERE ARE TWO REASONS WHY receptors on the peripheral processes of afferent neurons are a rational target for treating conditions of visceral hyperalgesia like irritable bowel syndrome (IBS). First, these afferent nerve terminals initiate the sensory signals that are transmitted from the gut to the brain and ultimately bring about the perception of visceral events including nausea, satiety, and pain. Second, these same visceral afferents, through local and central reflexes, modulate gastrointestinal motor and secretory activity that ensures that the digestive needs of the individual are met and that, under certain circumstances, may be disturbed, leading to diarrhea or constipation. Therefore, the Holy Address for reprint requests and other correspondence: D. Grundy, Dept. of Biomedical Science, Univ. of Sheffield, Western Bank, Sheffield S10 2TN, United Kingdom. Grail is a single therapeutic target with the potential to attenuate both visceral sensation and aberrant reflex activity and so restore motility and secretion to normal. The properties of visceral afferents have been dealt with extensively in recent themes articles, and we do not propose to cover this ground. Instead, we focus on the excitatory receptors expressed on the afferent terminals in the gut wall and consider the role these receptors play in visceral sensation and how this may relate to the development of novel therapies. In this context, we exclusively consider the receptor mechanisms present in extrinsic afferent nerves of the gastrointestinal tract, which send processes to the central nervous system. The activation pathways of primary afferent neurons or intestinofugal fibers that are a component of the enteric nervous system are not covered because, although they may modulate conscious sensation indirectly through changing the environment around extrinsic afferent nerves, they do not themselves impinge on consciousness directly. The mechanisms involved in visceral perception are generally less well understood than their corresponding sensations in the somatic realm, but there are a number of common features. Similar types of nerve, e.g., small-diameter unmyelinated (C) or thinly myelinated (A ) fibers convey both visceral and somatic painful sensations. However, visceral pain is poorly localized and is often referred to somatic sites because their inputs converge in the dorsal horn of the spinal cord. In addition, the thoracic and abdominal viscera project sensory information to the brain stem via vagal afferent pathways. From what we describe below, the available evidence indicates that these two classes of visceral extrinsic nerves (vagal and spinal) exhibit a number of contrasting properties, which in turn reflect their diverse roles in sensory signaling. RECEPTORS ON SENSORY NEURONS An enormous range of chemical mediators have been implicated in sensory signal transduction in the vis /01 $5.00 Copyright 2001 the American Physiological Society G787

2 G788 cera (Fig. 1). By and large, the evidence for a role of these substances in visceral sensation has come from studies on the cell bodies of visceral sensory neurons in the dorsal root or nodose ganglia. Electrophysiological, immunocytochemical, and molecular biological techniques revealed the functional expression of receptors to these mediators in these cells. To be firmly convincing for a role in sensory signal generation, such mediators should be effective when applied to the sensory nerve ending in the gut wall, and this has indeed been shown for several of the mediators we describe below. These substances are thought to produce their effects on visceral afferent nerves by three distinct processes: 1) direct activation, which generally involves the opening of ion channels present on the nerve terminals; 2) sensitization, which may occur in the absence of a direct stimulation but which usually results in afferent hyperexcitability to both chemical and mechanical modalities; and 3) alteration of the phenotype of the afferent nerve, for example through alterations in the expression of mediators, channels, and receptors or modulating the activity of these by changing the ligand-binding characteristics or coupling efficiency of other receptors. In addition to a role in the generation or modulation of sensory signals, the receptors and channels expressed may initiate or promote the release of preformed mediators in the afferent nerve endings such as substance P (SP) and calcitonin gene-regulated peptide (CGRP) either locally or from axon collaterals. These efferent functions form the basis of axon reflexes that appear to play an important role in neurogenic inflammation, and this may induce or exacerbate the sensitization and/or altered phenotype of afferent terminals in their vicinity. Any given mediator may recruit one or more of these pathways to produce its effect on visceral sensation, and interference with any of these mechanisms is likely to modulate the gain in visceral sensory pathways in the short and/or long term. By implication, manipulation of these processes therefore represents a valid therapeutic target to ex- Fig. 1. Some of the potential receptor mechanisms underlying activation (depolarization) and sensitization at the terminal of a gastrointestinal sensory afferent. Separate mechanisms underlie activation and sensitization. Some mediators such as serotonin (5-HT) cause activation via 5-HT 3 receptors, whereas others like PGE 2 acting at EP 2 receptors sensitize visceral afferent responses to other stimuli. Still others, for example, adenosine (Adeno), cause both stimulation and sensitization, possibly through distinct receptor mechanisms (4). Bradykinin (BK) has a self-sensitizing action, stimulating discharge through activation of phospholipase C (PLC) and enhancing excitability via prostaglandins (PGs) after activation of phospholipase A 2 (PLA 2). Inflammatory mediators can be released from different cell types (e.g., sympathetic varicosities, mast cells, and blood vessels) present in or around the afferent nerve terminal. 5-HT, ATP, and capsaicin (Cap) can directly activate nonselective cation channels (NSCCs), whereas adenosine, histamine, prostaglandins (not PGE 2), and proteases such as mast cell tryptase (Tryp) and thrombin (Thro) act on G protein-coupled receptors, leading to a Ca 2 -dependent modulation of ion channel activity. Sensitization, however, may be mediated by increased intracellular camp. Adenosine and PGE 2 can generate camp directly through G protein-coupled stimulation of adenylate cyclase (AC). In contrast, histamine (Hist) may act indirectly through the generation of prostaglandins (4). The actions of camp downstream are currently unknown but may involve modulation of ion channels, interaction with other second messengers (e.g., Ca 2 ), or even changes in receptor expression. PARs, protease activated receptors; COX-1, COX-2, cyclooxygenase-1 and -2; AA, arachidonic acid; DAG, diacylglycerol; IP 3, inositol 1,4,5-trisphosphate.

3 G789 ploit in the management of unpleasant sensations and/or hypersensitive states of visceral origin. In the ensuing paragraphs, we discuss the mechanisms by which a number of diverse mediators elicit their effects on visceral afferent sensitivity. Substances that produce a direct stimulation of visceral sensory nerve endings may function in one of two ways. The first function is as a component of a discrete sensory signaling pathway. In this case, the afferent neuron does not respond directly to a stimulus but responds after the release of a mediator from a primary sense cell. A classic example of this type of sensory signal transduction is found in the gustatory afferents innervating the taste buds. These gustatory cells release chemical mediators that act in paracrine fashion on sensory nerve endings within their close proximity. As discussed below, a similar taste mechanism exists in the gut mucosa to monitor luminal composition. The second function is as an indiscriminate modulator of visceral afferent sensitivity. Such substances are usually released under conditions of inflammation, injury, or ischemia from a plethora of cell types, e.g., platelets, leukocytes, lymphocytes, macrophages, mast cells, glia, fibroblasts, blood vessels, muscle cells, and neurons. Each of these specific cells (e.g., mast cells) may release several of these modulating agents, some of which may act directly on the sensory nerve terminal, whereas others may act indirectly after release of other agents from other cells in a series of cascades. This type of activation, which is analogous to an alarm system, is believed to initiate sweeping behavioral and reflex mechanisms that serve to protect the host. There are several examples of humoral agents that function in either or both of the described manners. We initially consider certain mediators that are primarily believed to function in dedicated sensory signaling roles before expanding on those agents that act on visceral afferent nerves in a more promiscuous fashion. SENSORY SIGNAL TRANSDUCTION Before activation of extrinsic afferent nerves, specific stimuli arising within the lumen of the gastrointestinal tract may activate specialized cells present in the mucosa. Examples of these cells, which effectively act as principal sensory transducers, are enterochromaffin (EC) cells, which release 5-hydroxytryptamine (5-HT). These epithelial cells are strategically positioned in the intestinal mucosa to taste luminal contents and release their mediators across the basolateral membrane to generate action potentials in the afferent nerve endings within the lamina propria. Stimulus intensity is encoded in the amount of mediator release and represents the balance between the mechanisms causing release and the uptake mechanisms that limit the site and duration of activation. In the case of 5-HT liberated from EC cells, this agent is thought to act directly on vagal extrinsic afferent nerves in the mucosa through activation of ionotropic 5-HT 3 receptors expressed on the nerve terminal (17). The physiological stimuli for the release of 5-HT from EC cells may be, in certain instances, mechanical, suggesting a role for this process in mechanotransduction (13). However, a large body of data implicates this mechanism in the detection of bacterial enterotoxins, e.g., cholera toxin. These toxins trigger release of 5-HT from EC cells to bring about an orchestrated response to dilute and subsequently eliminate the pathogenic material from the body and preclude further consumption of the potentially harmful material. However, certain drugs used in cancer chemotherapy are thought to inappropriately trigger these mechanisms, thus provoking debilitating nausea and emesis in patients. These symptoms are responsive to selective 5-HT 3 receptor antagonists, such as ondansetron and granisetron, which strongly supports the role for 5-HT in nausea and vomiting pathways from the gastrointestinal tract. However, other mechanisms may be involved, especially in the delayed emetic response to chemo- and radiotherapy (1). The mediator(s) responsible for emesis that is resistant to 5-HT 3 receptor antagonists is not known, and this supports the view that there is the potential for plasticity in the emetic pathway. From a therapeutic perspective, elucidating the identity(ies) of the substance(s) that account for this delayed response may assist in the development of effective antiemetic agents for patients who are or become refractory to the presently available therapy. Electrophysiological evidence suggests that 5-HT acts on a distinct population of vagal mucosal afferent nerves (16). As such, this represents an example of a high-fidelity, modality-specific signal transduction pathway. This mechanism more than likely functions in the detection of moment-to-moment changes in luminal composition and operates, in the main, below the level of consciousness. In contrast to such a specific signaling pathway, it is apparent that a battery of mediators, some of which are described below, may influence visceral sensation in a less ordered manner. VISCERAL HYPERSENSITIVITY Vagal and spinal afferent fibers each respond to mechanical stimulation such as distension and contraction. Vagal afferents encode events within the physiological range. In contrast, spinal afferents respond over a wide dynamic range extending from physiological to pathophysiological levels of distension (12). These spinal endings can contribute to signaling visceral pain through some intensity code that recognizes extreme levels of distension or contraction. Other spinal afferents, however, respond only to noxious levels of distension, and an extreme example of these is the high-threshold mechanoreceptors that fail to respond under normal circumstances. These are the so-called sleeping or silent nociceptors that can be awakened under conditions of injury or inflammation (12). The latter illustrate the fact that mechanosensitivity is not fixed either in terms of threshold for activation or gain in the stimulus-response relationship, and as such the threshold can be reduced and the gain increased under certain situations. A number of proinflammatory me-

4 G790 diators have been implicated in this sensitization process (see Fig. 1), and examples of some of the key agents in this phenomenon are detailed below. Bradykinin. Bradykinin (BK) is a nonapeptide that is generated from plasma during tissue damage and inflammation. This peptide mediates its effects via two G protein-coupled receptors, B 1 and B 2, with the latter being constitutive and the former induced by some cytokines and nerve growth factor (NGF) (see Ref. 5). In vitro studies in uninflamed preparations have shown that BK powerfully activates mesenteric spinal afferents with serosal terminals through an action on B 2 receptors and that BK-induced release of prostaglandins (see below) contributes to the overall magnitude of the response (4, 31). These findings corroborate whole animal studies showing that B 2 receptor antagonists attenuate visceral pain, at least in acute models in which inflammation is present (see Ref. 5). In contrast, it seems that in chronic inflammation models, the role of the inducible B 1 receptor in visceral nociception mechanisms becomes more dominant (5). The wealth of evidence clearly indicates a role for BK in the generation of visceral pain in the acute and chronic phases of inflammation, and, therefore, antagonists of BK receptors could be useful therapeutically to treat visceral hypersensitivity in inflammatory conditions. Prostaglandins and leukotrienes. Products of arachidonic acid oxygenation are a major contributor to hyperalgesia in the somatic realm, and recent evidence suggests that they may play a similar role in visceral sensory transmission. This group of mediators comprises the prostaglandins (PGs) and leukotrienes (LKs), which are synthesized from the precursor arachidonic acid by cyclooxygenase (COX) and lipoxygenase enzymes, respectively. PGE 2 acts through multiple EP receptors (see Ref. 14). In the gastrointestinal tract, EP 1 receptors appear to play a major role in direct activation of mucosal mesenteric afferents, but EP 2 receptors may also play a sensitizing role (14, 31). Critical to this latter function in visceral nerves may be the activation of adenylate cyclase and elevation of intracellular camp, because the membrane-permeant camp analog dibutyryl camp mimics the sensitization process (Ref. 4; see Fig. 1). Such mechanisms may underlie the enhanced responsiveness of visceral afferent neurons to chemical and mechanical stimuli in inflammatory conditions and may be involved in wakening the so-called silent nociceptors after an inflammatory insult. Two isoforms of the COX enzyme have been characterized, and these are termed COX-1 and COX-2 (22). COX-1 is constitutive and may be involved in controlling baseline visceral afferent sensitivity because, in native tissue, naproxen significantly reduced the magnitude of the response to BK (31). However, during inflammatory conditions such as colitis, upregulation of the inducible COX-2 occurs, leading to augmented PG synthesis, and this enzyme may therefore be important in the genesis of persistent pain in this syndrome (39). Interleukin (IL)-1 and tumor necrosis factor (TNF)- may underlie this increased expression of COX-2 (41), which is consistent with the view that PGs contribute to the illness behavior and somatic and visceral hyperalgesia associated with elevated levels of these cytokines. PGs are derived from virtually every type of tissue, in particular sympathetic nerve terminals and immunocompetent cells, and these sources may be important in the maintenance of the inflammatory state. Very little is known about the role of lipoxygenase products such as LKB 4 in visceral nociception. However, lipoxygenase metabolites have recently been shown to activate vanilloid receptors, suggesting that they may contribute to visceral pain in inflammatory conditions (see below). 5-HT. 5-HT is discussed above in relation to EC cell mechanisms of luminal tasting. However, this humoral agent may also be released from a variety of other sources including platelets, mast cells (5), endothelial cells, and serotonergic neurons (13) in the gastrointestinal tract. mrna for several 5-HT receptors is present in dorsal root ganglion cells (8), and this suggests that multiple potential signaling mechanisms are activated by this mediator. Although it remains to be unequivocally demonstrated in the gastrointestinal tract, the presence of 5-HT 3 receptors on visceral spinal afferent terminals (11) may explain the ability of some 5-HT 3 receptor antagonists to provide pain relief in certain IBS patients. However, the clearly defined functional role of these receptors in the peristaltic reflex (see Ref. 37) may explain the marked constipating side effects of such agents, which may ultimately limit their widespread clinical usefulness in the treatment of visceral hyperalgesia. Other 5-HT receptors such as the 5-HT 4 receptor have generated a recent upsurge in interest regarding a putative role of these receptors in modulating visceral sensation. 5-HT 4 receptors are generally coupled to adenylate cyclase (15), which would be expected to promote an increase in neuronal excitability, but a partial 5-HT 4 agonist, counterintuitively, attenuates visceral spinal afferent transmission (38). Although the prokinetic effects of these agents, which would provide symptomatic relief in constipation-predominant IBS, are well documented (see Ref. 37), further studies are necessary to characterize the nature of this apparent inhibitory effect on visceral spinal afferent nerves. ATP. There is good evidence that ATP, which is released from damaged tissues, participates in visceral afferent signaling. This molecule exerts its physiological effects through the interaction with two major classes of receptor (34). One group, which comprises ligand-gated cation channels, is termed P2X receptors and is currently classified into seven subtypes. The other group, which consists of G protein-coupled receptors, is termed P2Y receptors, and seven subtypes of these receptors have been functionally characterized in mammalian systems. To date, however, convincing evidence only exists for the presence of functional P2X receptors in visceral neurons. Electrophysiological and immunocytochemical studies have demonstrated their expression in cells derived from nodose (30) and dorsal

5 G791 root (7) ganglia, and recent evidence indicates the existence of functional P2X receptors on the peripheral endings of mesenteric afferent nerve bundles (23). These nerve trunks contain a mixed population of vagal, spinal, and intestinofugal afferent fibers, which terminate in different levels of the bowel wall (mucosa, muscle, and serosa), so whether ATP is involved in conveying luminal or noxious signals or mediating peripheral reflexes is not currently clear. The functional ATP-gated cation channels in native sensory neurons were initially thought to consist, on the one hand, of either homomeric P2X 3 subunits in the case of spinal neurons (7) or, on the other, heteromultimers of P2X 2 and P2X 3 receptors in the case of vagal fibers (30). However, the situation is probably more complex, because functional channels with diverse properties can be formed from heterologous expression of individual types or combinations of different types of P2X receptors (34). Despite this potential complexity, studies with P2X 3 receptor knockout mice have indicated possible roles of this subtype in sensory signal transduction in the visceral realm (9). Notably, these null mutants display hyporeflexia of the urinary bladder (9). Because ATP is released from the urothelium in response to stretch, it is possible that the loss of P2X 3 receptors, which are normally present on fibers innervating the bladder, may impair the functioning of the micturition reflex (see Ref. 9). This raises the possibility that a similar process underlies a component of mechanotransduction in the gastrointestinal tract, because mechanical stress has been shown to stimulate the release of ATP from several types of cell that are present in the bowel wall (see Ref. 35). Indeed, ATP can be liberated from cells by stimuli as diverse as changes in cell volume, hypoxia, extracellular acidosis, inflammation, lipopolysaccharide exposure, and membrane receptor stimulation (see Ref. 35 for references). Clearly, there is a potential for this nucleotide to be involved in the transduction of nonnoxious and/or noxious stimuli arising in the viscera. It is tempting to speculate that the physiological and/or pathophysiological signaling roles of ATP may be determined not just by anatomic location of the afferent nerve endings in the gut wall but also by the differing expression at their peripheral terminals of homomeric and/or heteromeric P2X receptors. The studies with P2X 3 receptor knockouts indicated certain modalities in which this subtype of receptor is involved in encoding. In view of the paucity of selective agents, investigations in knockouts of the other P2X receptors found in sensory ganglia (P2X 2, P2X 5, and P2X 6 ) should permit the appreciation of the roles of these subtypes in sensory signal transduction in the gastrointestinal tract in both normal and dysfunctional states. This in turn should provide important information as to which of the P2X receptors should be specifically targeted to alleviate aberrant and/or noxious sensory signal transduction, without perturbing normal visceral sensation. Adenosine. The enzymatic breakdown product of ATP, adenosine, may also function as a sensory signaling molecule in the viscera. Four types of adenosine receptors, termed A 1, A 2A, A 2B, and A 3, have been identified (34). These receptors are metabotropic and, depending on the tissue, they either inhibit or stimulate adenylate cyclase or activate phospholipase C. Functional adenosine receptors are expressed in nodose (29) and dorsal root (34) ganglion cells, and these generally exhibit inhibitory effects. In contrast, with recordings from mesenteric afferent nerve bundles innervating the jejunum it has been shown that adenosine receptor activation elicits excitation of their peripheral endings both in vitro (4) and in vivo (26). This effect appears to involve direct and indirect mechanisms on the sensory nerve via the stimulation of A 1 and A 2B receptors (4, 26). The direct excitation of the nerve terminal may involve a sensitization to substances present in the extracellular fluid surrounding the nerve ending. In contrast, a component of the indirect excitation produced by adenosine receptor activation is probably secondary to contraction of intestinal smooth muscle. In addition, this excitatory response involves the stimulation of the visceral afferents by mediators released from mast cells (3), of which histamine, acting via H 1 receptors, has been shown to be a powerful activator (27). In vitro evidence indicates that some of the effects of adenosine involve an action on mesenteric afferent fibers located in the serosa (4). This suggests that these fibers are spinal in origin. However, additional effects of this mediator on vagal and intestinofugal populations of sensory nerves cannot be discounted as yet. Nevertheless, these observations suggest that adenosine may activate and/or sensitize gastrointestinal nociceptive visceral fibers, and this may occur in conditions of inflammation or ischemia, where extracellular concentrations of this substance are likely to be elevated (see Ref. 4 for references). However, it should be borne in mind that extracellular levels of adenosine increase subsequent to release of ATP because of the latter s rapid degradation by ubiquitous ectonucleotidases. In this respect, the effect of exogenous ATP on mesenteric afferent nerve activity comprised a delayed component that was sensitive to an adenosine receptor antagonist (25). Thus the effects of the nucleotide on visceral afferent nerves and, ultimately, the qualities of the sensation generated may involve synergism between ATP and adenosine receptors and/or may involve the stimulation of separate populations of fibers. However, whether functional P2X and adenosine receptors are coexpressed on the peripheral endings of mesenteric afferent fibers remains unresolved. Vanilloid receptors. Vanilloid compounds such as capsaicin have been known for a number of years to stimulate both somatic and visceral afferent nerves (18). The molecular basis for the excitatory effects of these compounds on sensory fibers is the activation of nonselective cation channels that have recently been cloned (6) and are known as vanilloid receptor-1 (VR-1). Functional capsaicin receptors have been demonstrated on cells obtained from vagal and dorsal root

6 G792 ganglia (18), and the presence of mrna for VR-1 has been confirmed in these neurons (32). Consistent with these observations, capsaicin sensitivity has been shown in vagal mucosal and muscle fibers (2) and spinal serosal afferent neurons (31) innervating the gastrointestinal tract. The identity of the natural ligand for this receptor is still not known, and this has hindered progress in understanding the physiological and/or pathophysiological roles of this receptor in visceral sensation. Nevertheless, protein kinase C and ligands such as BK, anandamide, and lipoxygenase products each increase VR-1 receptor channel activity (19, 33). Heat and protons also augment VR-1 receptor activity (6, 42). In the case of the latter, this mechanism may explain or at least contribute to its excitatory effects on visceral afferent nerve activity (42). In addition, protons augment the response of P2X 2 receptors to ATP (34). Because the extracellular fluid in inflamed tissues has a lower ph than normal, the effect of protons on these receptors may be relevant to visceral pain in inflammatory conditions of the gastrointestinal tract. In view of the fact that multiple signaling mechanisms affect the activity of VR-1, there may not be a single endogenous ligand for this channel. Instead, it may represent a general effector mechanism that is recruited by mediators under certain conditions to stimulate or enhance the activation of sensory neurons. Development of a more effective antagonist than capsazepine is clearly essential if we are to fully appreciate the role of VR-1 in visceral pain perception and perhaps provide some insight into its role in vagal afferents, which may not necessarily be involved in nociception in the bowel (12). Proteinase-activated receptors. An emerging class of receptors that is provoking an explosion of interest is the proteinase-activated receptors (PARs) (10). The natural ligands of these receptors, which belong to the superfamily of G protein-coupled receptors, are serine proteases such as thrombin, mast cell tryptase, and trypsin (10). PARs exhibit a novel mechanism of receptor activation and are stimulated by a tethered ligand of six amino acid residues located on the NH 2 -terminal domain, which is exposed after an end portion of the receptor is cleaved by the protease (10). Four types of PARs, denoted PAR1, PAR2, PAR3 (10), and PAR4 (21), have been identified, although PAR3 is probably a cofactor in the activation of PAR4. PAR1 and PAR4 are sensitive to thrombin, whereas mast cell tryptase activates PAR2 and trypsin stimulates both PAR2 and PAR4 (28). Transduction pathways to which the PARs are coupled include the stimulation of phospholipase C and mitogen-activated protein (MAP)-3 kinase mitogenic cascades, the latter of which may have particular important functional significance with respect to longterm effects on visceral afferent signaling (10). PAR2 are expressed in dorsal root ganglion cells (40), and PAR1 has been recently described in this tissue (44). Preliminary evidence suggests that functional PAR2 is also present on the peripheral terminals of mesenteric afferent nerves (24). Because tryptase-containing mucosal mast cells reside in close proximity to afferent terminals in the lamina propria (43), these receptors, together with those for 5-HT and histamine, may participate in neuroimmune signaling from the gastrointestinal tract during allergic reactions. Furthermore, after a breach of the mucosal barrier, for example, by acid erosion, it is possible that trypsin may gain access to the submucosal space and activate the terminals of PAR2-expressing nerves in the vicinity, eventually leading to the orchestration of mucosal defense mechanisms. However, neither the level of the terminations nor the identity of the population of mesenteric afferent nerves functionally expressing PAR2 is known. Until such information is to hand, one can still only speculate on their function. In addition, whether PAR4 has a role to play in gastrointestinal afferent signaling awaits exploration. NGF. NGF, acting through the trka receptor located on spinal sensory neurons that also express SP and CGRP, is one mediator that has the potential to exert a long-term effect on the activity of visceral sensory neurons (see Ref. 5). NGF levels are rapidly increased during inflammatory responses, and sequestration of endogenous NGF leads to a decrease in hyperalgesia in experimental models (20). NGF, together with other trophic factors including neurotrophin-3 (NT3), brainderived neurotrophic factor (BDNF, which acts through trkb receptors), and glia-derived neurotrophic factor (GDNF) evoke phenotypic changes in nociceptive neurons and may also be involved in the regeneration of sensory neurons after nerve injury (36). Intriguingly, nodose neurons express trkb but not trka receptors (32), suggesting that different trophic influences may regulate the sensitivity of vagal and spinal visceral afferent nerves, but this still remains to be demonstrated. POTENTIAL FOR NOVEL THERAPIES Considerable recent progress has been made in our understanding of the receptor mechanisms that augment the excitability of visceral afferent nerves. This new information obviously has important implications for the development of novel treatments for conditions in which visceral hypersensitivity is a key feature, such as IBS. However, there are undoubtedly other potentially crucial mediators awaiting discovery, and this and other present deficiencies in our knowledge preclude a full appreciation of the complex interactions between mediators that occur at the level of the visceral afferent terminal. Despite this, it is not unreasonable to propose that an ideal therapeutic agent for the treatment of visceral hyperalgesia would be one that would not interfere with normal signal transduction but serve to prevent the release of, or reverse the sensitizing effect of, mediators released under pathophysiological conditions. In this respect, sensitizing agents act in a highly synergistic fashion and are generally excitatory, although there are other mediators such as somatostatin, -opioids, and GABA that do

7 G793 have an inhibitory influence. The concept of an inflammatory soup has been developed to incorporate the complex interactions that can occur between many different mediators and may help to explain the success of a variety of agents with some efficacy in IBS. Equally, there may be a degree of redundancy among the mediators such that the contribution of one is compensated for by plasticity in the mechanisms that control sensitivity. However, as mentioned above, unless the underlying cause of the visceral hypersensitivity in IBS is unraveled, any real possibility of developing an all-encompassing treatment is some way off. The authors acknowledge the financial support of the BBSRC (UK), the DFG (Germany), and GlaxoWellcome Research and Development Ltd. (UK). REFERENCES 1. Andrews PL, Naylor RJ, and Joss RA. Neuropharmacology of emesis and its relevance to anti-emetic therapy. Consensus and controversies. Support Care Cancer 6: , Blackshaw LA, Page AJ, and Partosoedarso ER. Acute effects of capsaicin on gastrointestinal vagal afferents. Neuroscience 96: , Brasacchio MF, Kirkup AJ, Wagner S, Haupt W, Kreis ME, and Grundy D. Mast cell involvement in mesenteric afferent sensitivity to adenosine: 2B or not 2B? (Abstract). Gastroenterology 116: A593, Brunsden AM and Grundy D. Sensitization of visceral afferents to bradykinin in rat jejunum in vitro. J Physiol (Lond) 521: , Bueno L, Fioramonti J, and Garcia-Villar R. Pathobiology of visceral pain: molecular mechanisms and therapeutic implications. III. Visceral afferent pathways: a source of new therapeutic targets for abdominal pain. Am J Physiol Gastrointest Liver Physiol 278: G670 G676, Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, and Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389: , Chen CC, Akopian AN, Sivilotti L, Colquhoun D, Burnstock G, and Wood JN. A P2X purinoceptor expressed by a subset of sensory neurons. Nature 377: , Chen JJ, Vasko MR, Wu X, Staeva TP, Baez M, Zgombick JM, and Nelson DL. Multiple subtypes of serotonin receptors are expressed in rat sensory neurons in culture. J Pharmacol Exp Ther 287: , Cockayne DA, Hamilton SG, Zhu QM, Dunn PM, Zhong Y, Novakovic S, Malmberg AB, Cain G, Berson A, Kassotakis L, Hedley L, Lachnit WG, Burnstock G, McMahon SB, and Ford AP. Urinary bladder hyporeflexia and reduced pain-related behaviour in P2X 3-deficient mice. Nature 407: , Dery O, Corvera CU, Steinhoff M, and Bunnett NW. Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. Am J Physiol Cell Physiol 274: C1429 C1452, Fu LW and Longhurst JC. Role of 5-HT3 receptors in activation of abdominal sympathetic C fibre afferents during ischaemia in cats. J Physiol (Lond) 509: , Gebhart GF. Pathobiology of visceral pain: molecular mechanisms and therapeutic implications. IV. Visceral afferent contributions to the pathobiology of visceral pain. Am J Physiol Gastrointest Liver Physiol 278: G834 G838, Gershon MD. Roles played by 5-hydroxytryptamine in the physiology of the bowel. Aliment Pharmacol Ther 2: 15 30, Haupt W, Jiang W, Kreis ME, and Grundy D. Prostaglandin EP receptor subtypes have distinctive effects on jejunal afferent sensitivity in the rat. Gastroenterology 119: , Hegde SS and Eglen RM. Peripheral 5-HT 4 receptors. FASEB J 10: , Hillsley K and Grundy D. Serotonin and cholecystokinin activate different populations of rat mesenteric vagal afferents. Neurosci Lett 255: 63 66, Hillsley K, Kirkup AJ, and Grundy D. Direct and indirect actions of 5-hydroxytryptamine on the discharge of mesenteric afferent fibres innervating the rat jejunum. J Physiol (Lond) 506: , Holzer P. Capsaicin: cellular targets, mechanisms of action, and selectivity for thin sensory neurons. Pharmacol Rev 43: , Hwang SW, Cho H, Kwak J, Lee SY, Kang CJ, Jung J, Cho S, Min KH, Suh YG, Kim D, and Oh U. Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci USA 97: , Jaggar SI, Scott HC, and Rice AS. Inflammation of the rat urinary bladder is associated with a referred thermal hyperalgesia which is nerve growth factor dependent. Br J Anaesth 83: , Kahn ML, Zheng YW, Huang W, Bigornia V, Zeng D, Moff S, Farese RV Jr, Tam C, and Coughlin SR. A dual thrombin receptor system for platelet activation. Nature 394: , Kargman S, Charleson S, Cartwright M, Frank J, Riendeau D, Mancini J, Evans J, and O Neill G. Characterization of prostaglandin G/H synthase 1 and 2 in rat, dog, monkey, and human gastrointestinal tracts. Gastroenterology 111: , Kirkup AJ, Booth CE, Chessell IP, Humphrey PP, and Grundy D. Excitatory effect of P2X receptor activation on mesenteric afferent nerves in the anaesthetised rat. J Physiol (Lond) 520: , Kirkup AJ, Bunnett NW, and Grundy D. A proteinase-activated receptor subtype 2 (PAR-2) agonist stimulates mesenteric afferent nerve discharge (Abstract). Gastroenterology 118: A173, Kirkup AJ, Chessell IP, Humphrey PPA, and Grundy D. Dual effect of ATP on mesenteric afferent nerve discharge in the anaesthetized rat (Abstract). J Physiol (Lond) 520: 67P, Kirkup AJ, Eastwood C, Grundy D, Chessell IP, and Humphrey PP. Characterization of adenosine receptors evoking excitation of mesenteric afferents in the rat. Br J Pharmacol 125: , Kreis ME, Haupt W, Kirkup AJ, and Grundy D. Histamine sensitivity of mesenteric afferent nerves in the rat jejunum. Am J Physiol Gastrointest Liver Physiol 275: G675 G680, Lan RS, Stewart GA, and Henry PJ. Modulation of airway smooth muscle tone by protease activated receptor-1, -2, -3 and -4 in trachea isolated from influenza A virus-infected mice. Br J Pharmacol 129: 63 70, Lawrence AJ, Krstew E, and Jarrott B. Complex interactions between nitric oxide and adenosine receptors in the rat isolated nodose ganglion. Eur J Pharmacol 328: 83 88, Lewis C, Neidhart S, Holy C, North RA, Buell G, and Surprenant A. Coexpression of P2X 2 and P2X 3 receptor subunits can account for ATP-gated currents in sensory neurons. Nature 377: , Maubach KA and Grundy D. The role of prostaglandins in the bradykinin-induced activation of serosal afferents of the rat jejunum in vitro. J Physiol (Lond) 515: , Michael GJ and Priestley JV. Differential expression of the mrna for the vanilloid receptor subtype 1 in cells of the adult rat dorsal root and nodose ganglia and its downregulation by axotomy. J Neurosci 19: , Premkumar LS and Ahern GP. Induction of vanilloid receptor channel activity by protein kinase C. Nature 408: , Ralevic V and Burnstock G. Receptors for purines and pyrimidines. Pharmacol Rev 50: , Roman RM and Fitz JG. Emerging roles of purinergic signaling in gastrointestinal epithelial secretion and hepatobiliary function. Gastroenterology 116: , Romero MI, Rangappa N, Li L, Lightfoot E, Garry MG, and Smith GM. Extensive sprouting of sensory afferents and hyper-

8 G794 algesia induced by conditional expression of nerve growth factor in the adult spinal cord. J Neurosci 20: , Sanger GJ. 5-Hydroxytryptamine and functional bowel disorders. Neurogastroenterol Motil 8: , Schikowski A, Mathis C, Thewissen M, Ross H-G, Pak MA, and Enck P. Dose-dependent modulation of rectal afferent sensitivity by a 5-HT 4 receptor agonist (Abstract). Gastroenterology 116: A643, Singer II, Kawka DW, Schloemann S, Tessner T, Riehl T, and Stenson WF. Cyclooxygenase 2 is induced in colonic epithelial cells in inflammatory bowel disease. Gastroenterology 115: , Steinhoff M, Vergnolle N, Young SH, Tognetto M, Amadesi S, Ennes HS, Trevisani M, Hollenberg MD, Wallace JL, Caughey GH, Mitchell SE, Williams LM, Geppetti P, Mayer EA, and Bunnett NW. Agonists of proteinase-activated receptor 2 induce inflammation by a neurogenic mechanism. Nat Med 6: , Takahashi S, Shigeta J, Inoue H, Tanabe T, and Okabe S. Localization of cyclooxygenase-2 and regulation of its mrna expression in gastric ulcers in rats. Am J Physiol Gastrointest Liver Physiol 275: G1137 G1145, Tominaga M, Caterina MJ, Malmberg AB, Rosen TA, Gilbert H, Skinner K, Raumann BE, Basbaum AI, and Julius D. The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron 21: , Williams RM, Berthoud HR, and Stead RH. Vagal afferent nerve fibres contact mast cells in rat small intestinal mucosa. Neuroimmunomodulation 4: , Young SH, Ennes HS, Steinhoff M, Mayer EA, and Bunnett NW. Rat dorsal root ganglion (DRG) neurons express proteinase activated receptors (PARS) 1 and 2 (Abstract). Gastroenterology 118: A381, 2000.