Protease-activated receptor (PAR2) is expressed by nociceptive neurons and activated during inflammation by proteases from mast cells, the intestinal lumen, and the circulation. Agonists of PAR2 cause hyperexcitability of intestinal sensory neurons and hyperalgesia to distensive stimuli by unknown mechanisms. We evaluated the role of the transient receptor potential vanilloid 4 (TRPV4) in PAR2-induced mechanical hyperalgesia of the mouse colon. Colonic sensory neurons, identified by retrograde tracing, expressed immunoreactive TRPV4, PAR2, and calcitonin gene-related peptide and are thus implicated in nociception. To assess nociception, visceromotor responses (VMR) to colorectal distension (CRD) were measured by electromyography of abdominal muscles. In TRPV4+/+ mice, intraluminal PAR2 activating peptide (PAR2-AP) exacerbated VMR to graded CRD from 6–24 h, indicative of mechanical hyperalgesia. PAR2-induced hyperalgesia was not observed in TRPV4−/− mice. PAR2-AP evoked discharge of action potentials from colonic afferent neurons in TRPV4+/+ mice, but not from TRPV4−/− mice. The TRPV4 agonists 5′,6′-epoxyeicosatrienoic acid and 4α-phorbol 12,13-didecanoate stimulated discharge of action potentials in colonic afferent fibers and enhanced current responses recorded from retrogradely labeled colonic dorsal root ganglia neurons, confirming expression of functional TRPV4. PAR2-AP enhanced these responses, indicating sensitization of TRPV4. Thus TRPV4 is expressed by primary spinal afferent neurons innervating the colon. Activation of PAR2 increases currents in these neurons, evokes discharge of action potentials from colonic afferent fibers, and induces mechanical hyperalgesia. These responses require the presence of functional TRPV4. Therefore, TRPV4 is required for PAR2-induced mechanical hyperalgesia and excitation of colonic afferent neurons.
- visceral pain
- protease-activated receptors
- transient receptor potential channels
mechanical sensation in the gastrointestinal tract is required for normal digestion, but defects in mechanical sensation can cause pain. Normally, distension that occurs during feeding, the transit of digesta, and defecation is a nonnoxious mechanical stimulus, and the detection of such mechanical stimuli is required for enteric reflexes and physiological gastrointestinal motility (22). However, allodynia and hyperalgesia to distension are observed in adults and children with colitis and functional bowel disorders (20, 21, 38). Effective therapies for visceral pain are limited by an incomplete understanding of the ion channels that transduce mechanical pain and of the mechanisms by which they are activated or sensitized during disease to cause allodynia and hyperalgesia to mechanical stimuli.
Members of the transient receptor potential (TRP) family of ion channels are important sensors of environmental stimuli and heightened responsiveness of these channels occurs during inflammatory diseases, causing pain (16). TRP vanilloid 1 (TRPV1), a prototypical member of this family, is expressed by a subset of primary sensory neurons that contain substance P (SP) and calcitonin gene-related peptide (CGRP) (12). Protons (pH < 5.5), heat (>42°C), bioactive lipids, and capsaicin directly activate TRPV1 to stimulate SP and CGRP release in peripheral tissues, causing neurogenic inflammation (extravasation of plasma proteins and granulocytes, and hyperemia), and in the dorsal horn of the spinal cord, resulting in activation of nociceptive neurons (23). Inflammatory agents, including proteases, indirectly sensitize TRPV1, resulting in thermal hyperalgesia (5). Although several ion channels respond to mechanical stimuli, TRPV4 has been identified as a channel that mediates somatic mechanosensation and inflammatory mechanical hyperalgesia. TRPV4 is gated by altered tonicity (hypotonic, mildly hypertonic) and by temperature >27°C (26, 35). Identified as the mammalian homologue of the Caenorhabditis elegans gene Osm-9 (36), TRPV4 is expressed by neurosensory structures, including Merkel cells and sensory neurons, where it responds to osmotic and mechanical stimuli. TRPV4 mediates responses to shear stress in endothelial cells (34) and hypotonicity-induced currents in nociceptive neurons (3). The cytochrome P-450 product of arachidonic acid, 5′,6′-epoxyeicosatrienoic acid (5,6-EET), activates TRPV4 and may be an endogenous agonist (56). Hypoosmotic stimuli cause cell swelling and production of arachidonic acid due to the activation of phospholipase A2 (45). A synthetic TRPV4 agonist is the phorbol ester 4α-phorbol 12,13-didecanoate (4αPDD) (55). TRPV4−/− mice show diminished nociceptive responses to noxious pressure (36, 51), and nociceptive responses to altered tonicity are reduced by deletion or knockdown of TRPV4 (2, 3). Inflammatory agents, such as prostaglandins and proteases, also indirectly sensitize TRPV4, resulting in hyperalgesia to somatic mechanical stimuli (1, 3, 25). However, the role of TRPV4 in mechanosensation in the digestive tract, including inflammatory hyperalgesia to mechanical stimuli, is unknown.
We investigated the role of TRPV4 in mechanosensation and mechanical hyperalgesia in the intestine. To induce mechanical hyperalgesia, we treated animals with agonists of protease-activated receptor 2 (PAR2), a member of a family of four G-protein coupled receptors for serine proteases (43). Proteases that are generated and released during injury and inflammation, including mast cell tryptase (18), trypsins (41), coagulation factors VIIa and Xa (11), and kallikreins (42), cleave and activate PAR2. PAR2 is prominently expressed by neurons of the dorsal root ganglia (DRG) that also express TRPV1, TRPV4, SP, and CGRP (5, 25, 49). PAR2 agonists stimulate release of SP and CGRP from these neurons in peripheral tissues such as the skin and intestine, causing neurogenic inflammation (15, 40, 49). Activation of PAR2 on DRG neurons also causes somatic hyperalgesia to thermal and mechanical stimuli by central mechanisms, by promoting the release of neuropeptides in the dorsal horn of the spinal cord (54). When injected into the colonic lumen of rats, PAR2 agonists increase fos expression in nociceptive spinal neurons and cause hyperalgesia to colorectal distension (17). Proteases released from the intestinal mucosa of patients with irritable bowel syndrome, including trypsins and tryptase, cause visceral hyperalgesia in mice by activating PAR2, suggesting a role for proteases and PAR2 in pain in humans (13). However, the mechanisms by which PAR2 agonists cause hyperalgesia to intestinal distension are unknown.
We recently reported that PAR2 agonists sensitize TRPV4-induced Ca2+ signals, currents, and neuropeptide release from nociceptive neurons, resulting in somatic mechanical hyperalgesia (25). However, mechanisms of visceral pain and hyperalgesia differ from those of somatic nociception. Afferent signals from the colon are carried centrally by two distinct pathways: the lumbar splanchnic nerve (LSN) and the pelvic nerve. The mechanosensory, chemosensory, and electrophysiological properties of these two pathways differ between one another and from somatic nociceptive pathways (9, 24), as do their relative activity at baseline and during inflamed conditions (10, 52). These differences may be attributable, in part, to differential contributions of nociceptive ion channels between the pathways (50). However, the contribution of TRPV4 to PAR2-induced hyperalgesia to colonic distension is unknown.
In the present investigation, we tested the hypothesis that activation of PAR2 sensitizes TRPV4 on colonic afferent neurons, leading to enhanced visceral fiber sensitivity and colonic mechanical hyperalgesia. We examined whether 1) TRPV4 is expressed with PAR2 on spinal afferent nerves innervating the colon; 2) TRPV4 is required for PAR2-induced mechanical hyperalgesia in the colon; 3) TRPV4 is required for PAR2-induced activation of afferent neurons innervating the colon; and 4) PAR2 sensitizes TRPV4-induced action potentials in colonic sensory nerves and TRPV4 currents in colonic DRG neurons.
MATERIALS AND METHODS
Generation of TRPV4+/+ and TRPV4−/− mice in a C57/B6 background has been described (36). Male mice (15–25 g) were used. CD-1 mice (males and females, 15–20 g) and Sprague-Dawley rats (male, 200–300 g) were from Charles River Laboratories (Montreal, QC, Canada and Wilmington, MA). Animals were housed in microisolator cages with free access to food and water, under a 12-h light-dark cycle. The Institutional Animal Care and Use Committees of all institutions approved animal procedures. Animals were killed with pentobarbital sodium (200 mg/kg ip) or CO2 inhalation and bilateral thoracotomy, or by cervical dislocation.
Agonists and antagonists.
PAR2 activating peptide (AP), corresponding to the peptide sequence of the tethered ligand domain of rat/mouse PAR2 (SLIGRL-NH2), and the reverse peptide (RP, LRGILS-NH2), were from CPC Scientific (San Jose, CA). PAR2-AP is a selective agonist of PAR2, and PAR2-RP does not activate this receptor (49). The TRPV4 activators 4αPDD and 5,6-EET and the TRP channel antagonist ruthenium red (RR) were from Sigma Chemical (St. Louis, MO).
Retrograde labeling of colonic neurons.
To identify DRG neurons projecting to the colon for subsequent staining or recording, a retrograde tracer was injected into the colonic wall. Mice were anesthetized with ketamine and xylazine (87.5 and 12.5 mg/kg ip, respectively). The descending colon was exposed through a midline incision and the overlying viscera were kept moist with saline-soaked gauze. Under a dissecting microscope, the dicarbocyanine dye, 1,1-dioctadecyl-3,3,3,3-tetramethlindocarbocyanine methanesulfonate (DiI, 50 mg/ml in methanol; Invitrogen, Eugene, OR) or fast blue (17 mg/ml in 0.9% saline; Cedarlane Laboratories, Hornby, ON, Canada) was injected into in the wall of the colon (6–10 sites, 1 μl per site) using a Hamilton microliter syringe and a 32-gauge needle. To prevent leakage to other organs, the needle was held in place for 30 s after the injection, and any excess dye was carefully removed. The abdominal wall and skin were closed, and animals were allowed to recover. After 7–10 days, animals were killed and DRG (T10–L1, T5 for control) were removed and fixed for immunohistochemistry or dissociated and cultured for electrophysiology.
Immunofluorescence localization of TRPV4 and PAR2.
Mouse DRG (T10–L1) were immersion fixed in 4% paraformaldehyde in 100 mM PBS pH 7.4, for 2 h. Tissues were washed, incubated in 30% sucrose in PBS overnight at 4°C, and embedded in OCT (BD Biosciences, San Jose, CA). Sections of DRG (10 μm) were cut and mounted on silane-coated slides (Sigma). Tissue sections were washed and incubated in 100 mM PBS, pH 7.4, containing 5% NGS and 0.3% Triton X-100, with the following primary antibodies: rabbit antibody to residues 853–871 in the COOH terminus of rat TRPV4 (Alomone, Jerusalem, Israel, ACC-034, 17 of 19 residues conserved in mouse, 1:500); rabbit antibody to the COOH terminus of mouse TRPV4 (58) (1:500); guinea pig anti-CGRP (RDI, Flanders, NJ, 1:250); rabbit anti-PAR2 (19) (9717, 1:250) (all overnight, 4°C). In controls, the TRPV4 antiserum ACC-034 was preabsorbed by preincubation with the channel fragment used for immunization prior staining (10 μM, 24 h, 4°C). Tissues were washed and incubated with secondary antibodies: goat anti-rabbit IgG conjugated to fluorescein isothiocyanate (Jackson Immuno-Research, West Grove, PA, 1:200) and goat anti-guinea pig IgG conjugated to Alexa 647 (Invitrogen, Carlsbad, CA, 1:1,000) (all 2 h, room temperature). Specimens were mounted in Prolong Gold (Invitrogen).
Specimens were observed with a Zeiss LSM510 Meta confocal microscope and Zeiss Plan Apo Chromat ×20, Fluor Plan Apo ×40, or Apo Chromat ×100 (NA 1.4) objectives. Acquisition parameters were set to minimize bleed-through. Images were processed to adjust contrast and brightness and were colored to represent appropriate fluorophores using Adobe Photoshop 7.0 (Adobe Systems, Mountain View, CA). Images of stained and control slides were collected and processed identically.
Measurement of visceromotor responses to CRD using EMG.
The placement of electrodes for electromyography (EMG) recording and the construction of balloons for colorectal distension (CRD) have been described (31). For electrode placement, mice were anesthetized with ketamine and xylazine (87.5 and 12.5 mg/kg ip, respectively) and received buprenorphine (0.1 mg/kg ip or sc) for preemptive analgesia. Electrodes were made by stripping 3 mm of insulation from the ends of Teflon-coated stainless steel wires (Cooner Wire, Chatsworth, CA), which were sutured into the abdominal musculature just superior to the inguinal ligament. Electrodes were tunneled subcutaneously to the dorsum of the neck, externalized, and secured. The skin was closed and animals allowed to recover for 7 days. During the recovery period, mice were acclimatized to the recording chamber. On the day of testing, mice were briefly anesthetized with isoflurane (1–1.5%), and distending balloons were placed in the rectum, 0.5 cm proximal to the anus, and secured by tape to the tail. Mice were placed in clear polycarbonate tube restrainers and connected to the CRD apparatus and amplifier. Graded distensions were generated by using a compressed helium source with an ultra-low-flow regulator and digital pressure gauge. Mice were challenged at 15, 30, 45, and 60 mmHg distending pressures, with three 10-s trials at each pressure and a 3-min recovery period between each distension. EMG data were amplified and filtered with a 15RXi amplifier and collected into Polyview Recorder (Grass-Telefactor, Braintree, MA). Data were imported into Spike II Software (Cambridge Electronic Design, Cambridge, UK) for analysis. EMG activity was quantified by integrating rectified EMG data over the period of interest and expressed as millivolt-seconds.
PAR2-AP or PAR2-RP (control) were administered to mice immediately after measurement of baseline visceromotor responses (VMR). Under brief isoflurane anesthesia, peptide (100 μl of 1 mg/ml peptide in 80% normal saline, 10% ethanol, 10% Tween-80) was injected by enema into the lumen of the colon 3 cm proximal to the rectum via a PTFE 24 catheter attached to a 1-ml syringe. While anesthetized, mice were held inverted for 5 min. This dose of PAR2-AP causes neurogenic inflammation of the mouse intestine (15, 40). At 6, 12, 24, and 48 h after treatment, VMR to CRD was assessed.
Afferent nerve fiber recordings.
Neuronal tracts supplying the colon were dissected as described (9, 37). Briefly, the colon was removed from mice and rats with an attached bundle containing the inferior mesenteric artery, abdominal aorta, inferior mesenteric ganglion, intermesenteric nerve, and LSN and placed in cold modified Krebs solution (bicarbonate buffer, 4°C). The distal colon was opened longitudinally and pinned flat, mucosal side down, in an isolated perfused organ bath (Danz Instruments, Blackwood, South Australia) with the LSN insertions oriented to lie along the edge of the opened preparation. Connective tissue was dissected away from the neurovascular bundle, which was then drawn into a separate paraffin-filled recording chamber. The attached LSN was finely dissected from the aorta and desheathed from the transected nerve stump to expose the nerve fibers; these were divided into fine bundles to be placed onto platinum recording electrodes for single afferent fiber recording. Extraluminal receptive fields on the serosal surface were located by responses to blunt probing with a fine brush and classified as serosal on the basis of their lack of response to low-intensity stretch (≤5 g) and fine mucosal stroking (10 mg von Frey hair), as described (9, 37). PAR2-AP (10–300 μM) was applied into a small steel ring placed over the corresponding receptive field for 2 min, either alone or after a 10-min preincubation with RR (10–100 μM). In some experiments the afferent response to the TRPV4 agonist 5,6-EET (10 μM) was tested before and after PAR2-AP. Signals from the recording electrode were differentially amplified and filtered, and the analog signal was sampled at 20 kHz by using a 1401 data interface (CED, Cambridge, UK). Single units were discriminated with Spike II software (CED).
Electrophysiological recordings from acutely dissociated DRG neurons.
Neurons were acutely dissociated as described (39). Briefly, DRG (T10–13) were removed from mice and incubated in collagenase (1 mg/ml, Worthington, Lakewood, NJ) and dispase (4 mg/ml, Roche, Indianapolis, IN) in HBSS for 10 min at 37°C. Neurons were triturated with a fire-polished Pasteur pipette and incubated again for an additional 5 min at 37°C. Dissociated neurons were suspended in MEM with Earle's salts and HCO3 (Invitrogen) containing 1% penicillin-streptomycin, 2 mM glutamine, and 0.2% (wt/vol) glucose and placed onto Vitrogen-coated coverslips and maintained in a humidified incubator (95% air-5% CO2, 37°C) until electrophysiological studies 4–24 h later. Whole-cell membrane currents of freshly dispersed, fast blue-labeled mouse DRG neurons were recorded with an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA). Only neurons with a capacitance <30 pF were recorded from, since these are considered small nociceptive neurons. The pipette solution contained (in mM) 110 CsCl, 3 MgCl2, 10 EGTA, 10 HEPES, 3 Mg-ATP, 0.6 GTP, pH 7.2 with CsOH. The extracellular solution superfused over the cells during recordings contained (in mM) 100 NaCl, 25 tetraethylammonium chloride (TEA-Cl), 5 KCl, 4 CaCl2, 2 MgCl2, 0.5 4-aminopyridine (4-AP), 10 glucose, 10 HEPES, pH 7.4 with NaOH. Pipette resistance was 2–4 MΩ. Whole-cell currents were recorded and analyzed with Clampfit 10.1 software (Molecular Devices). Data were digitized at 10 kHz and filtered at 5 kHz. Neurons were voltage-clamped at 0 mV to inactivate voltage-gated calcium and sodium channels, and a 150-ms linear ramp protocol (−100 mV to +100 mV) was applied every 13 s. Neurons were exposed to PAR2-AP (100 μM) for 3 min and subsequently exposed to 4αPDD (10 μM). Current amplitude at −80 and +80 mV was normalized to cell capacitance to obtain current densities. Only cells with detectable whole-cell currents responses to TRPV4 agonists were included in the analysis.
Data are presented as means ± SE. The VMR for each CRD trial was quantified as the difference between EMG activity recorded during distension and 10 s immediately prior to distension. Differences between baseline and posttreatment VMR were examined by a repeated-measures ANOVA and Dunnett's post hoc test. The electrophysiological data were analyzed by a Student's two-tailed t-test or paired t-test. The proportion of PAR2-AP-responsive fibers in TRPV4+/+ and TRPV4−/− mice was compared by Fisher's exact test.
TRPV4 and PAR2 are coexpressed by DRG neurons innervating the colon.
TRPV4 is expressed by nociceptive neurons of DRG of rats that also express PAR2 and the proinflammatory and nociceptive neuropeptides SP and CGRP (25). However, it is not known whether TRPV4 is expressed by DRG neurons innervating the colon and whether these neurons also express PAR2 and CGRP. To identify neurons innervating the colon, we injected the retrograde tracer DiI into the wall of the descending colon of mice. DiI was detected in a small proportion of neurons in DRG from the thoracolumbar regions (Fig. 1). Cells from the T5 level did not fluoresce, serving as controls (not shown). To localize TRPV4, we used two different antibodies raised to the COOH terminus of rat and mouse TRPV4. Both antibodies have been reported to specifically detect TRPV4 expressed in transfected cell lines (25, 58). In the present study, both antibodies similarly stained mouse DRG neurons. Immunoreactive TRPV4 was detected in neurons of varying diameter in DRG from mice (Fig. 1A). TRPV4 was detected at the plasma membrane, the cytosol, and the nucleus. Preabsorption of the TRPV4 antibody with the COOH terminal channel fragment used for immunization abolished the signal, confirming specificity (Fig. 1B). Some DiI neurons expressing TRPV4 also contained immunoreactive CGRP, confirming that they are nociceptive neurons (Fig. 1C). Immunoreactive PAR2 was also detected in some DiI neurons (Fig. 1D). It was not possible to simultaneously localize TRPV4 and PAR2 since the antibodies were both raised in the same species (rabbits). However, TRPV4 was detected in 64% (52 of 81) of DiI-positive neurons, and PAR2 was detected in 66% (64 of 82) of DiI-positive neurons. Conversely, 76% (53 of 70) of TRPV4-positive DRG and 76% (73 of 92) of PAR2-positive DRG were also stained with DiI. Examination under high power indicated that immunoreactive TRPV4 and PAR2 were present at the plasma membrane of some DiI neurons (Fig. 1E). Thus in thoracolumbar DRG the majority of neurons expressing either TRPV4 or PAR2 innervate the colon, and a proportion of these neurons express both receptors.
TRPV4 is required for PAR2-induced visceral mechanical hyperalgesia.
Mice lacking functional TRPV4 show diminished pain responses to noxious somatic mechanical stimuli (51). Distension of the colon and rectum of mice induces VMR that is indicative of pain (31). To examine the role of TRPV4 in colonic mechanical sensation in mice, we compared VMR to CRD in TRPV4+/+ and TRPV4−/− mice at baseline. In both groups, graded CRD (15, 30, 45, 60 mmHg) caused a significant (P < 0.001) linear increase in EMG activity, indicative of VMR. The baseline VMR to all distending pressures was similar in TRPV4+/+ and TRPV4−/− mice (Figs. 2A and 3 for 60 mmHg). Thus functional TRPV4 is not required for perception of these stimuli in mice under basal conditions.
Intraplantar injection of PAR2 agonists cause allodynia and hyperalgesia to somatic mechanical stimuli (54). These responses are not observed in mice lacking functional TRPV4 and thus depend on sensitization of this channel (25). Intracolonic administration of PAR2 agonists also enhances the VMR to CRD in rats, indicative of mechanical hyperalgesia (17). To examine the role of TRPV4 in PAR2-induced visceral pain, we compared VMR to CRD in TRPV4+/+ and TRPV4−/− mice at 6, 24, and 48 h after intracolonic administration of PAR2-AP or PAR2-RP. In TRPV4+/+ mice, intracolonic injection of PAR2-AP significantly enhanced the VMR to the higher distending pressures (45 and 60 mmHg) measured at 6 and 24 h after injection, compared with baseline responses (Figs. 2B and 3A). Responses to 60 mmHg at 6 and 24 h after PAR2-AP were significantly increased by twofold over baseline (P < 0.05, P < 0.01, respectively, Fig. 2B). Comparisons of the total VMR response to all distending pressures confirmed that there was a 2.5-fold increase in response at 6 and 24 h after PAR2-AP compared with baseline responses (Fig. 4). At 48 h after PAR2-AP, responses were not different from those measured at baseline (not shown). In control experiments, intracolonic PAR2-RP, which does not activate PAR2, had no effect on the magnitude of the VMR to CRD at any distending pressure or time point in TRPV4+/+ mice (Figs. 2C and 4). Thus intracolonic PAR2-AP causes a robust and sustained increase in the VMR to CRD in TRPV4+/+ mice, indicative of mechanical hyperalgesia.
In marked contrast to the results obtained from TRPV4+/+ mice, PAR2-AP did not affect VMR to CRD in TRPV4−/− mice. In TRPV4−/− mice, the VMR to CRD at 15, 30, 45, or 60 mmHg measured at 6, 24, or 48 h after intracolonic PAR2-AP was not different from the baseline values (Figs. 2D and 3B). Comparisons of the total VMR response to all distending pressures confirmed the lack of sensitization in TRPV4−/− mice (Fig. 4). Thus TRPV4 mediates both acute and delayed hyperalgesia induced by PAR2 activation.
A TRP channel antagonist inhibits mechanosensitivity of serosal fibers through TRPV4.
Splanchnic serosal afferent fibers detect transient, high-intensity mechanical stimuli, such as those occurring during noxious balloon distension (9). Colonic inflammation (52, 53) and bradykinin (10) sensitize responses of these fibers to mechanical stimuli, indicating a role for this pathway in processing inflammatory hyperalgesia. To determine whether TRPV4 contributes to the mechanosensitivity of splanchnic serosal fibers, we examined responses to mechanical probing in the presence of RR, which antagonizes TRP channels, including TRPV4 (55). In TRPV4+/+ mice, focal compression of identified receptive fields with a calibrated von Frey hair (2,000 mg)-stimulated action potential discharge in splanchnic afferent fibers, and RR inhibited this response in a concentration-dependent manner, which was significant (P < 0.001) at the highest concentration tested (100 μM, Fig. 5A). In TRPV4−/− mice, mechanically stimulated discharge of action potentials was reduced by ∼50% compared with TRPV4+/+ mice, and RR had no effect on this response (Fig. 5B). Thus TRPV4 contributes to the mechanosensitivity of splanchnic serosal afferent fibers in mice, and the effect of RR on these fibers is to inhibit TRPV4 activity.
PAR2 activates serosal afferent fibers by a TRPV4-dependent mechanism.
PAR2-AP increases the discharge rate of afferent fibers from the small intestine of rats (33) and also evokes spontaneous depolarization and prolonged hyperexcitability of DRG neurons in rats (4, 5, 32) and enteric neurons in guinea pigs (46). To determine whether activation of PAR2 increases the activity of colonic afferent neurons and to further investigate the contribution of TRPV4 to PAR2-induced hyperalgesia in the colon, we recorded action potentials from colonic afferent fibers of mice during administration of PAR2-AP. In TRPV4+/+ mice, PAR2-AP (300 μM) induced a robust increase in the frequency of action potential firing in 40% of studied fibers (13 of 32), which was sustained for the 2-min period of application (Fig. 6, A and B). In those fibers that were responsive to PAR2-AP, the maximal rate of action potentials was approximately sevenfold greater after administration of agonist. This response was concentration dependent (10–300 μM PAR2-AP, with 30 μM as threshold, not shown). Fibers identified as unresponsive had similar discharge rates before and during PAR2-AP administration. In marked contrast, none of the fibers (0 of 19) from TRPV4−/− mice were responsive to PAR2-AP (Fig. 6, A and B). However, fibers from TRPV4−/− mice still respond to blunt probing, indicating that TRPV4−/− mice retained their ability to detect mechanical stimuli (Fig. 6C). These fibers also had normal electrical activation thresholds (not shown). Thus TRPV4 is required for the excitatory effects of PAR2-AP on colonic afferent fibers.
We also determined whether activation of PAR2 can sensitize subsequent responses to the TRPV4 agonist, 5,6-EET. In afferents from TRPV4+/+ mice that responded to PAR2-AP, the action potential response evoked by 5,6-EET was markedly enhanced compared with 5,6-EET responses in afferents that were unresponsive to PAR2-AP (P < 0.05, Fig. 6D). However, PAR2-AP had no effect on the mechanosensory responses to blunt probing in either PAR2-AP-responsive or -unresponsive fibers (Fig. 7). Thus PAR2 does not appear to directly increase mechanosensitivity in the same fibers that it activates, at least over a short time frame of a few minutes.
To determine whether these effects of PAR2 agonist are conserved in another species, we examined the effects of PAR2-AP on activity of colonic afferent fibers innervating the rat colon. Consistent with our results in mice, PAR2-AP significantly increased the peak action potential rate in a rat serosal afferent preparation (Fig. 8). RR, which inhibited the effects of PAR2-AP in mice by antagonizing TRPV4, abolished the action of PAR2-AP on activity of colonic afferent fibers in rats. These results suggest that PAR2 activation increases activity of colonic serosal afferent neurons in rats by a TRPV4-mediated mechanism.
PAR2 sensitizes TRPV4 currents in colonic DRG neurons.
PAR2 agonists can sensitize TRPV4-induced Ca2+ signals and currents in TRPV4-transfected HEK cells and in DRG neurons (25). However, somatic and visceral DRG neurons differ in their electrophysiological properties and responses to inflammatory mediators (24). Whereas activation of PAR2 has been previously shown to increase the excitability of colonic DRG (32), it is not known whether activation of PAR2 sensitizes TRPV4 in DRG neurons innervating the colon. To investigate whether colonic neurons express functional TRPV4 and to determine whether PAR2 agonists sensitize TRPV4 currents in these neurons, we recorded currents from fast blue-labeled colonic DRG neurons from mice during voltage ramps. Exposure of neurons to the TRPV4 agonist 4αPDD (10 μM) enhanced current responses compared with untreated neurons (Fig. 9, A and C). The mean increase in current density produced by 4αPDD was −7.11 pA/pF (±1.81 pA/pF) at −80 mV and 7.01 pA/pF (±1.83 pA/pF) at +80 mV (n = 7).
Of the colonic neurons examined, 55% responded to 4αPDD (7 of 13). Preincubation of neurons with PAR2-AP (100 μM, 3 min) enhanced the responses to 4αPDD (Fig. 9, B and C). In neurons preincubated with PAR2-AP, the activation by 4αPDD was increased to −25.91 pA/pF (±10.34 pA/pF) at −80 mV and 42.50 pA/pF (±11.17 pA/pF) at +80 mV (n = 11). No significant change was produced by application of PAR2-AP alone (not shown). Thus DRG neurons innervating the colon respond to the TRPV4 agonist 4αPDD, providing functional evidence for expression of this channel. PAR2-AP enhances these responses, indicating that PAR2 can sensitize TRPV4 in these neurons. This response to TRPV4 activation by PAR2-AP was similar to the sensitization of colonic serosal afferent endings.
Our results show that TRPV4 is required for PAR2-induced mechanical hyperalgesia and activation of afferent neurons in the colon. We observed that immunoreactive TRPV4 is expressed by the majority of primary spinal afferent neurons innervating the colon and that these neurons express PAR2 and respond to agonists of TRPV4 (5,6-EET, 4αPDD) and PAR2 (PAR2-AP), confirming functional coexpression. Activation of PAR2 sensitizes responses to colorectal distension and evokes action potentials in some colonic serosal afferent fibers. Deletion or antagonism of TRPV4 completely abolishes these effects. PAR2 also sensitizes TRPV4 currents in isolated colonic DRG neurons. We have thus described a mechanism whereby proteases that are produced during gastrointestinal injury or inflammation may cleave PAR2 on visceral afferent neurons and sensitize TRPV4, causing neuronal hypersensitivity and consequent hyperalgesia to mechanical stimulation. Together, these results confirm our recent report that PAR2 sensitizes TRPV4 in cell lines and in nociceptive neurons to promote somatic hyperalgesia to mechanical stimulation (25).
Role of TRPV4 in PAR2-induced visceral hyperalgesia.
We have recently reported that PAR2 sensitizes TRPV4 to cause somatic mechanical hyperalgesia (25). However, the role of TRPV4 in mediating the effects of PAR2 in the gastrointestinal tract has not been defined. The observation that TRPV4 is coexpressed with CGRP and PAR2 in some colonic sensory nerves supports the hypothesis that TRPV4 participates in nociception in the colon and that its activity may be regulated by PAR2. We observed that PAR2-AP induces a robust hyperalgesia to CRD up to 24 h after treatment. This finding is consistent with previous observations in rats that intracolonic administration of PAR2 agonists cause hyperalgesia to CRD by an unknown mechanism (17). In contrast, mice lacking functional TRPV4 did not show any increased VMR to CRD at any time point following PAR2-AP treatment. Thus TRPV4 is essential for the hyperalgesic effects of PAR2 in the colon. As expected, there was no significant hyperalgesic effect observed after treatment with PAR2-RP. The specificity of these agonists has been previously well demonstrated (5, 46).
Given its putative role as a mechanoreceptor, it is noteworthy that mice lacking functional TRPV4 responded similarly to 15–60 mmHg CRD as TRPV4+/+ mice in the basal state. These results suggest that other molecular transducers, such as purinergic P2X receptors (57) and acid-sensing ion channels (30, 44), play primary roles in sensing these mechanical stimuli under basal conditions. Our observation that TRPV4 is required for PAR2-induced hyperalgesia suggests that TRPV4 may be particularly important in high-threshold or silent afferents that are activated under inflamed conditions. It will be of interest to study the responses of TRPV4−/− mice under alternative models of inflammation or sensitization, to determine whether they are more generally resistant to visceral hyperalgesia.
The complete absence of PAR2-AP effect in TRPV4−/− mice is even more striking given previous observations that PAR2 sensitizes TRPV1 (5). TRPV1 is widely expressed in mouse colonic DRG neurons (47) from the splanchnic and pelvic innervations (7). Both jejunal (48) and colonic (30) afferents from mice lacking TRPV1 have lower responses to mechanical distension. TRPV1 has been implicated in chronic colonic hyperalgesia induced by zymosan, a noninflammatory insult (29). However, the contributions of TRPV1 and TRPV4 to PAR2-induced hyperalgesia may depend on the expression of these channels in specific neurons. Only 61% of splanchnic serosal afferents respond to capsaicin (7), whereas 40% respond to PAR2-AP (Fig. 6), with a proportion of serosal afferents that respond to PAR2-AP and 5,6-EET being unresponsive to capsaicin (S. M. Brierley, unpublished observation).
Role of TRPV4 in PAR2-induced sensitization of colonic neurons.
Our results demonstrate that PAR2-AP evokes action potential firing in a large population of serosal afferents from the LSN. Activation of PAR2 also increases the excitability of colonic DRG (32) and enteric neurons (46) and increases the discharge rate of small intestinal afferents fibers (33). PAR2-induced discharge of small intestinal afferents depends on both direct excitation of afferent neurons and indirect excitation that is secondary to increased motility of the intestine (33), and it is likely that the same direct and indirect mechanisms contribute to PAR2-induced excitation of colonic afferent neurons. Consistent with our in vivo data, fibers from TRPV4−/− mice were completely unresponsive to PAR2-AP. This represents a robust and novel manifestation of the functional coupling between the excitatory effect of PAR2 activation and TRPV4.
Serosal fibers from TRPV4−/− mice had diminished responses to high-intensity mechanical stimuli, and RR inhibited the mechanosensitivity of fibers from TRPV4+/+ but not TRPV4−/− mice. Together, these observations suggest that mechanosensitive TRPV4 is expressed by serosal afferents of the LSN. This observation contrasts with the finding that TRPV4+/+ and TRPV4−/− mice have almost identical responses to CRD under baseline conditions. However, serosal fibers respond only to relatively high-intensity mechanical stimuli and, on the basis of their previously defined response profiles, these fibers are unlikely to be activated by physiological levels of distension (9). Our data are consistent with a model in which TRPV4 plays a role primarily in mechanotransduction in high-threshold afferents but makes a minor contribution to transduction in low-threshold afferents that would be activated by CRD under basal conditions. Data supporting this model were recently presented (8). When activated by PAR2, these high-threshold fibers evoke action potential discharge, which we suggest contributes to visceral hyperalgesia by converging with signals from low-threshold afferents in the dorsal horn of the spinal cord, similar to the way in which bile-sensitive afferents sensitize spinal responses to distension (6).
We here also demonstrate that TRPV4 and PAR2 are present on colonic DRG neurons and that activation of PAR2 sensitizes TRPV4 currents. We did not investigate the mechanisms of sensitization specifically in colonic DRGs, although we have recently reported that PAR2 sensitizes TRPV4 in nociceptive neurons by mechanisms involving phospholipase Cβ and protein kinases A, C, and D (25). We have also recently demonstrated in colonic DRG neurons that PAR2 induces a sustained excitability to depolarizing stimuli by suppressing K+ currents, including delayed-rectifier IK currents (32). Given this alternate mechanism by which PAR2 might cause neuronal sensitization, the apparently complete dependence on TRPV4 for the effects of PAR2 is all the more striking.
The lack of effect of PAR2-AP in TRPV4−/− mice is unlikely to be due to diminished expression of functional of PAR2 in these animals, as PAR2 was detected in similar abundance in colonic DRGs of TRPV4−/− mice (N. W. Bunnett, unpublished observation). In addition, RR completely abolished PAR2-AP-induced activation of colonic afferents in rats. Although RR is a nonselective TRP channel blocker, its ability to inhibit afferent mechanosensory function was lost in TRPV4−/− mice, implicating a major action of RR on TRPV4 in these colonic fibers. This lack of effect further suggests that TRPV4 is the only RR-sensitive mechanoreceptor in these fibers. Taken together, these findings provide further support for a central role of TRPV4 in mediating the effects of PAR2.
Physiological role of PAR2 and TRPV4 in visceral hyperalgesia.
PAR2 may induce hyperexcitability of nociceptive neurons and cause visceral hyperalgesia by multiple mechanisms, including sensitization of TRPV1 (5), suppression of K+ currents including delayed-rectifier IK currents (32), and release of histamine from colonic mast cells (28). Our data suggest that expression of TRPV4 is required for PAR2-induced colonic neuronal hyperexcitability and hyperalgesia. Many aspects of the present study confirm our previous report that PAR2 agonists sensitize TRPV4 to cause somatic mechanical hyperalgesia in mice (25). In the present study, we did not examine the molecular mechanisms by which PAR2 sensitizes TRPV4 in colonic nociceptive neurons. However, in cell lines coexpressing PAR2 and TRPV4, antagonism of phospholipase Cβ and protein kinases A, C, and D inhibited PAR2-induced sensitization of TRPV4. It remains to be determined whether these mechanisms mediate sensitization of TRPV4 in colonic nociceptors.
The relatively immediate effect of PAR2-AP on neuronal activation contrasts with the observation that that PAR2-AP induces visceral hyperalgesia for up to 24 h, indicating that additional mechanisms may perpetuate PAR2-induced hyperalgesia in vivo. One possibility is that alterations in colonic permeability contribute to this delayed and sustained effect. However, PAR2 agonists directly activate enterocytes to promote paracellular permeability (14), suggesting that this a TRPV4-independent effect. Mast cells both express PAR2 and release the agonist tryptase (28). Thus mast cell activation may prolong the presence of PAR2 agonists (27), thereby perpetuating neuronal activation. Moreover, PAR2-induced hyperalgesia may in large part be mediated by neuropeptide release in the dorsal horn of the spinal cord (54), and the prolonged effects of PAR2-AP administration in vivo may reflect central sensitization of spinal neurons.
Absence of TRPV4 did not affect behavioral responses to CRD, suggesting that TRPV4 makes relatively little contribution to mechanosensation in low-threshold fibers. However, TRPV4 deletion prevented PAR2-induced activation of colonic neurons and colonic hyperalgesia to mechanical stimulation. Our results are in broad agreement with the emerging role of TRPV4 in somatic hyperalgesia, where deletion or inhibition of TRPV4 abolishes PAR2-induced hyperalgesia to mechanical and osmotic stimuli, but does not affect touch sensation in the absence of inflammation (1, 25). TRPV4−/− mice do have deficits in sensation of high-intensity noxious pressure (51), and a similar deficit may be present for noxious colonic distending pressures greater than 60 mmHg. However, higher distending pressures can cause colonic damage in mice (29). Whether TRPV4 plays a general role in inflammatory hyperalgesia in the colon remains to be determined. Given the prominent role of TRPV4 in mediating colonic mechanical hyperalgesia, antagonists of this channel may be of value in treating the pain of gastrointestinal disorders that involve elevated production or release of proteases.
This research was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-54840 and DK-43207 (N. W. Bunnett) and DK-07762 and the Hartwell Foundation (W. Sipe); National Health and Medical Research Council of Australia (L. A. Blackshaw); AstraZeneca (N. W. Bunnett, L. A. Blackshaw); Canadian Institute of Health Research (S. Vanner); and Crohn's and Colitis Foundation of Canada (S. Vanner).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society