Background: Transient receptor potential (TRP)A1, a member of the TRP family of ion channels, has been proposed to function in diverse sensory processes, including thermosensation and pain. However, TRPA1 has not been directly implicated in stomach mechanosensation, and its contribution to acute visceral pain from this organ is unknown. Here, we investigated the expression of TRPA1 in primary sensory afferents and its involvement in visceral hypersensitivity in rats.
Methods: We examined TRPA1 expression in the dorsal root ganglion (DRG), nodose ganglion (NG), and stomach of rats by using immunohistochemistry. Electromyographic responses to gastric distention (GD) were recorded from the acromiotrapezius muscle in TRPA1 knockdown rats and in control rats.
Results: TRPA1 was predominantly expressed with sensory neuropeptides in DRG and NG neurons, and in nerve fibres in the rat stomach. Gastric distention induced the activation of extracellular signal-regulated protein kinase 1/2 (ERK1/2) in DRG and NG neurons 2 min after stimulation, and most of the phosphorylated-ERK1/2-labelled DRG neurons were TRPA1-positive neurons. Intrathecal injection of TRPA1 antisense attenuated the visceromotor response, and suppressed ERK1/2 activation in the DRG, but not NG, neurons produced by GD. Furthermore, intrathecal and intraperitoneal injections of the TRPA1 inhibitor HC-03003 suppressed the response to noxious GD.
Conclusions: The activation of TRPA1 in DRG neurons by noxious GD may be involved in acute visceral pain. Our findings point to the potential blockade of TRPA1 in primary afferents as a new therapeutic target for the reduction of visceral hypersensitivity.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Functional dyspepsia (FD), defined as pain or discomfort located in the upper abdomen which is altered by food intake, is one of the most common ailments of the bowel (intestines). No structural or biochemical abnormality can be identified in most of those seeking medical attention. Visceral hypersensitivity has been identified in patients with a variety of such functional disorders, and changes in both the properties of primary afferent neurons and abnormal processing of sensory information within the central nervous system likely contribute to the development of this acute visceral pain.1 2 3 The stomach is innervated with afferent vagal fibres that project via the nodose ganglion (NG) to the nucleus of the solitary tract (NTS) in the medulla, and by splanchnic nerves that project via the dorsal root ganglion (DRG) to thoracic spinal cord segments. Several studies show that both vagal and spinal pathways are involved in mediating signals about noxious stimulation of the stomach with mechanonociception carried preferentially through spinal afferents.4 5 6 However, very little is known about the molecules responsible for mechanosensation in the viscera and their contribution to visceral pain.7 8
Multiple mammalian transient receptor potential (TRP) genes have been cloned and classified into six subfamilies that are widely expressed and play diverse roles in thermosensation, mechanosensation, pheromone sensing, and gustation.9 10 Among them, TRPA1 can be activated by a variety of stimuli, including icilin, a chemical that induces a cooling sensation, by noxious cold below 17°C, and by pungent chemicals.11 12 13 14 TRPA1 is expressed in a subset of small-sized DRG and trigeminal ganglia neurons that also express noxious heat-sensing TRPV1, substance P (SP), calcitonin gene-related peptide (CGRP), and is believed to be involved in somatosensory function, pain and neurogenic inflammation.11 15 Recent reports have demonstrated that TRPA1 knock-out mice displayed a deficit in sensing mechanical stimuli, suggesting that TRPA1 may contribute to somatic mechanosensation.16 17 However, TRPA1 has not been directly implicated in stomach mechanosensation, and its contribution to visceral pain from this organ is unknown. In the present study, we examined the localisation of TRPA1 protein in primary afferent neurons and the stomach of the rats, and investigated the roles of TRPA1 in visceral mechanosensation.
Materials and methods
A total of 152 male Sprague–Dawley rats weighing 190–260 g were used. Food, but not water, was withheld for 24 h before surgery.
All surgical procedures were performed with rats under pentobarbital anaesthesia (50 mg/kg, ip). A left lateral incision, 3–4 cm in length, was made to expose the stomach. As described previously,5 latex balloons (diameter, 2.0 cm) were inserted through the fundus into the stomach. The polyethylene tubing for connection of the balloon to the pressure control device was exteriorised at the back of the neck.
The origin of the primary afferent innervation of the stomach was examined by retrograde tracing using fluorescent fluorogold dye (FG; Fluorochrome, Denver, Colorado, USA). The stomach was exposed by left lateral incision, and 2.5-μl injections of 4% FG were made in the ventral (eight sites) and dorsal (eight sites) walls of the stomach with a Hamilton microsyringe (Hamilton, Reno, Nevada, USA).
Intrathecal delivery of HC-030031 (1 or 10 μg/μl; a kind gift from Hydra Biosciences, Cambridge, Massachusetts, USA), was performed basically as described previously.18 A laminectomy of the L3 vertebrae was performed under adequate anaesthesia with sodium pentobarbital. The dura was cut, and a soft tube (Silascon; Kaneka Medix Company, Osaka, Japan; outer diameter, 0.64 mm) was inserted into the subarachnoid space of the spinal cord at the T9/10 DRG level, and 10 μl of HC-030031 (1 or 10 μg/μl, dissolved in 10% dimethyl sulfoxide) was administered 30 min before gastric distention (GD). HC-030031 (150 mg/kg, dissolved in 0.5% methyl cellulose) was also administered intraperitoneally 60 min before GD. Ten per cent dimethyl sulfoxide or 0.5% methyl cellulose was used as vehicle control.
Antisense knockdown of TRPA1 expression
Antisense oligodeoxynucleotide (AS-ODN; 5′-TCTATGCGGTTATGTTGG-3′), mismatch ODN (MM-ODN; 5′-ACTACTACACTAGACTAC-3′), and fluorescein isothiocyanate (FITC)-labelled ODN directed to TRPA1 have been designed and manufactured by BIOGNOSTIK (Gottingen, Germany).15 Intrathecal delivery of AS-ODN, MM-ODN and FITC-labelled ODN was performed basically as described previously.18 To obtain a sustained drug infusion, a mini-osmotic pump (Alzet type 2003; Durect, Cupertino, California, USA), which operates at a rate of 1 μl/h for a period of 3 days, was filled with AS-ODN (0.5 nmol/μl) or MM-ODN (0.5 nmol/μl) in normal saline. Normal saline was used as the vehicle control. Although it has frequently been questioned whether ODN can reach the DRG in a sufficient concentration by intrathecal delivery, several reports have demonstrated that intrathecal ODN accumulates in DRG cells.19 20
All stimulation procedures were performed on rats that were anaesthetised with urethane (0.8–1.1 g/kg body weight, intraperitoneally). Visceromotor responses were recorded by quantifying the activity on electromyography (EMG) recorded with electrodes implanted in the acromiotrapezius muscle. The EMG before (baseline) and during constant pressure distention of the stomach (60 mm Hg, 30 s) was amplified and filtered (10–5000 Hz; PowerLab/Bio Amplifier; ADInstruments, Castle Hill, New South Wales, Australia), digitised, and integrated using the PowerLab4/25/Chart/ADInstruments data acquisition interface. The raw EMG data were rectified and quantified by calculating the area under the curve. Each distention trial consisted of three segments: a 10 s predistention baseline period, a 30 s distention period (60 mm Hg), and 10 s after termination of GD. Responses to GD are reported as a percentage compared with baseline during the distention period. Rats without stimulation (0 mm Hg) were used as naive controls.
Generation of TRPA1 antisera
To raise antibodies against TRPA1 proteins, rabbits were injected with the peptide RFKKERLEQMHSKWNF coupled via the cysteine to keyhole limpet haemocyanin (KLH).21 The peptide corresponds to a sequence present in the constant region of the rat or mouse proteins (residues 1092–1107 in NP_997491 for rat; residues 1092–1107 in NP_808449 for mouse). The KLH-linked TRPA1 peptide was injected into rabbits using standard procedures for antiserum production from the Peptide Institute (Osaka, Japan). For affinity purification, the peptide was coupled to an affinity column made of aminoalkyl agarose (Bio-Rad, Hercules, California, USA), and the antibody was purified following the manufacturer’s protocols.
After appropriate survival times (the survival time after stimulation in all experiments was 2 min, except in the time-course study), rats were perfused transcardially with 1% paraformaldehyde in 0.1 mol/l phosphate buffer (pH 7.4), followed by 4% paraformaldehyde in 0.1 mol/l phosphate buffer (n = 4 per group). DRG, NG and stomach sections were cut on a cryostat and processed for TRPA1, neurofilament 200 (NF200; Sigma, St Louis, Missouri, USA), tyrosine kinase A (TrkA; Chemicon, Temecula, California, USA), SP (DiaSorin, Stillwater, Minnesota, USA), CGRP (DiaSorin), TRPV1 (Oncogene, San Diego, California, USA), TRPM8 (made by M Tominaga, Okazaki Institute for Integrative Bioscience, Okazaki, Japan), and phosphorylated extracellular signal-regulated protein kinase 1/2 (p-ERK1/2; Cell Signaling Technology, Beverly, Massachusetts, USA) immunohistochemistry according to the procedure used in our previous study.22 The rabbit polyclonal TRPA1 (1:200) and monoclonal NF200 (1:100 000) antibodies were used for single-immunofluorescent staining. For double-immunofluorescent staining, tyramide signal amplification (NEN, Boston, Massachusetts, USA) fluorescence procedures23 were used for rabbit anti-TrkA (1:5000), rabbit anti-SP (1:20 000), rabbit anti-CGRP (1:5000), rabbit anti-TRPV1 (1:5000), rabbit anti-TRPM8 (1:5000), and rabbit anti-p-ERK1/2 (1:5000) staining. The p-ERK1/2 antibody (1:500) was also used for diaminobenzidine tetrahydrochloride staining. The number of TRPA1- or p-ERK1/2-immunoreactive (IR) neurons per section was counted in the DRG and NG.
Differences in values over time of each group were tested using one-way analysis of variance (ANOVA), followed by individual post hoc comparisons (Fisher’s exact test). One-way ANOVA, followed by individual post hoc comparisons (Fisher’s exact test) or pairwise comparisons (t test), were used to assess differences of values between the treatment groups. A difference was accepted as significant if the p value was less than 0.05.
Expression of TRPA1 in visceral afferent pathways
To determine the distribution pattern of the TRPA1 channel protein, we raised a polyclonal antibody against the C-terminal 16-amino acid residues of rat or mouse TRPA1.21 This antibody recognised the predicted band (128 kDa) in the western blot from rat DRG extract, and pretreatment with excess antigenic peptide AD195 completely eliminated the band and immunohistochemical labelling.21 We used this affinity-purified antiserum on tissue sections from the rat thoracolumbar DRG and NG, and found that TRPA1 was expressed by 51.2% (SD 3.1%) and 33.4% (SD 2.3%) of rat DRG and NG neurons, respectively (fig 1A).
FG injected into the stomach labelled cells bilaterally, not only in the thoracolumbar DRG, but also in the NG. We found that the majority of the TRPA1-IR cells were double labelled with FG in both DRG and NG neurons, indicating that TRPA1 is expressed in the primary afferents innervating the stomach (fig 1A). In the DRG and NG, the proportion of identified visceral afferents expressing TRPA1 was larger than that in total ganglion neurons, indicating that TRPA1 expression is more prevalent in gastric sensory neurons (fig 1B).
TRPA1 is predominantly expressed in TRPM8-negative peptidergic neurons
To determine whether the expression of TRPA1 in the thoracolumbar DRG and NG is in a subpopulation of the DRG neurons with myelinated fibres, we performed double labelling for TRPA1 and NF200, because NF200 is a marker for myelinated A-fibres. The results of the co-localisation study with TRPA1 and NF200 in the thoracolumbar DRG and NG are shown in figs 2 and 3, respectively. Only 5.0 (SD 1.9)% and 10.7 (SD 0.7)% of all TRPA1-IR neurons also were labelled with NF200 in the DRG and NG, respectively, indicating that TRPA1 is predominantly expressed in neurons with unmyelinated axons, ie, the C-fibres (table 1). Next, we examined the co-localisation of TRPA1 with TrkA, SP, CGRP and TRPV1. TRPA1 heavily co-localised with TrkA, SP, CGRP and TRPV1 in the thoracolumbar DRG and NG, indicating that these TRPA1-expressing neurons were nerve growth factor-responsive peptidergic neurons (table 1). Furthermore, to determine whether the TRPA1- and TRPM8-positive neurons belonged to the same subset of DRG neurons, co-localisation of TRPA1 with TRPM8 was performed. In the thoracolumbar DRG and NG, TRPM8 was detected in a small subpopulation of TRPA1-labelled neurons (10.6 (SD 2.3)% and 13.8 (SD 5.0)%, respectively). The results of the co-localisation study with TRPA1 with NF200, TrkA, SP, CGRP, TRPV1 and TRPM8 in FG-positive gut-innervating neurons in the thoracolumbar DRG and NG are shown in table 2.
TRPA1 co-localises predominantly with the sensory neuropeptides in gastric nerve fibres
We then examined the expression of TRPA1 protein in the rat stomach using immunohistochemistry. In the stomach, nerve fibre-like structures labelled for TRPA1 were detected in the mucosa, around blood vessels in the submucosa, and in the muscle layers (fig 4A,B). Next, the expression of SP- or CGRP-IR nerve fibres was compared with TRPA1 using double-immunofluorescent staining. Both SP and CGRP were present in networks of fibres within mucosal villi in a gastric section, and abundant in submucosa and external muscle layers of rat stomach. We also found that TRPA1 co-localised with SP or CGRP in the mucosa, around blood vessels in the submucosa, and in the external muscle layers. TRPA1 labelling was also observed in SP- and CGRP-positive intramuscular endings, indicating that TRPA1 is expressed with the sensory neuropeptides in gastric nerve fibres of extrinsic and intrinsic origin (fig 4A,B).
Knockdown of TRPA1 expression affects the visceromotor response produced by GD
Noxious mechanical stimulation of the stomach in rats triggers a reproducible visceromotor response that can be quantified using EMG recordings from the acromiotrapezius muscle.5 Therefore, we first confirmed that noxious mechanical stimulation caused a significant increase in EMG activity (fig 5A). The GD of 20 or 40 mm Hg did not cause significant changes in EMG activity recorded from the acromiotrapezius muscle (data not shown), consistent with a previous study.5 24 However, 60 mm Hg of GD led to an increase in visceromotor responses (p<0.05, one-way ANOVA) (fig 5B). To test whether acute visceral pain after noxious mechanical stimulation of the stomach is critically dependent on the presence of TRPA1 in sensory neurons, rats were treated intrathecally with either an AS-ODN targeting TRPA1 or a MM-ODN. We found that 60 mm Hg of GD induced an increase in visceromotor responses that was significantly less in the TRPA1 AS-ODN group than in the MM-ODN group (fig 5A,B). The increase in visceromotor responses in the MM-ODN group was not different from that of the vehicle control rats.
Knockdown of TRPA1 expression attenuates the ERK1/2 activation in DRG neurons produced by GD
ERK1/2 activation in primary afferents has been suggested to be involved in peripheral sensitisation in acute pain conditions.25 26 27 We therefore examined the p-ERK1/2 labelling 2 min after mechanical stimulation (fig 6A). We found few neurons labelled for p-ERK1/2 in the naive control DRG and NG (0 mm Hg), consistent with previous reports.25 27 We also found that the percentage of p-ERK1/2-IR neurons after GD of 60 mm Hg was significantly higher than that of control rats both in the DRG and NG (fig 6B). The increase in p-ERK1/2 labelling was found mainly in small- to medium-diameter neurons (fig 6A, arrows). Furthermore, 17.0 (SD 1.4)% and 14.4 (SD 0.8)% of FG-positive neurons in DRG and NG, respectively, were double-labelled with p-ERK1/2, and ∼62% and ∼49% of p-ERK1/2-IR cells in DRG and NG neurons, respectively, were double-labelled with FG. We found that the pretreatment with AS-ODN, but not MM-ODN, inhibited the noxious mechanical stimulation-induced ERK1/2 activation in the DRG (fig 6A,B), implying that the TRPA1 channel in the DRG is required for ERK1/2 activation after noxious mechanical stimulation. However, there was no effect of AS-ODN on the number of p-ERK1/2-IR cells in the NG.
To investigate whether ERK1/2 activation in DRG neurons produced by noxious mechanical stimulation is mediated through TRPA1, we performed double staining for p-ERK1/2 and TRPA1 (fig 6C). The majority of p-ERK1/2-labelled neurons 2 min after GD at 60 mm Hg also expressed TRPA1 in DRG neurons. We found that 12.5 (SD 1.1)% and 19.3 (SD 3.7)% of TRPA1-positive neurons in DRG and NG, respectively, were double-labelled with p-ERK1/2, and that 79.7 (SD 1.1)% and 69.7 (SD 5.2)% of p-ERK1/2-IR cells in DRG and NG neurons, respectively, were double-labelled with TRPA1. We then confirmed that the level of TRPA1 protein in the DRG of the AS-ODN-treated rats was significantly lower than that in the MM-ODN-treated rats, whereas there was no difference in TRPA1 expression in the NG neurons that innervate the stomach (fig 7).
Effect of HC-030031 on visceromotor responses to noxious GD
To further confirm the functional roles of TRPA1 in DRG neurons, we investigated whether inhibition of TRPA1 activation modifies the response to noxious mechanical stimulation. HC-030031 is a newly identified potent and selective TRPA1 inhibitor. HC-030031 blocked calcium response and inward currents elicited by allyl isothiocyanate (AITC) in hTRPA1-expressing cells in a concentration-dependent manner and greatly attenuated inflammatory- and neuropathy-induced pain in vivo.28 The TRPA1 inhibitor HC-030031 was delivered intrathecally 30 min before stimulation via a catheter whose tip was positioned close to the T9/10 DRG to target TRPA1 activity in the DRG. We found that intrathecal administration of HC-030031 dose-dependently reversed the response to 60 mm Hg of GD, compared with vehicle-treated rats (fig 8A). Furthermore, we found that a single intraperitoneal administration of HC-030031 also inhibited the 60 mm Hg of GD-induced increase in visceromotor responses (fig 8B).
Effect of afferent denervation on visceromotor responses to noxious GD
To determine whether spinal or vagal pathways mediate the aversive response to noxious mechanical stimulation of the stomach, we selectively dissected the bilateral splanchnic or subdiaphragmatic vagus nerves. Consistent with previous results,6 splanchnic nerve resection significantly affected the response to GD at 60 mm Hg, whereas results after subdiaphragmatic vagotomy did not differ from those in sham-operated rats (fig 8C).
Two thermosensive ion channels, TRPV1 and TRPV4, have been implicated previously in visceral mechanosensation. Deletion of TRPV1 or TRPV4 impairs afferent fibre transmission of mechanical stimuli in the colon.29 30 However, new mechanotransduction candidates are emerging, such as TRPA1, but its contribution to visceral mechanosensation is unclear. Although previous studies have shown TRPA1 is localised in a subpopulation of DRG neurons,11 15 we now show that TRPA1 is associated with neurons that innervate a particular target. Studies using knockout mice demonstrated that TRPA1 is not essential for hair-cell transduction but contributes to the transduction of mechanical stimuli in nociceptor sensory neurons.16 In the present study, we found that intrathecal TRPA1 AS-ODN significantly affected the visceromotor response to noxious GD. Our functional evidence was correlated with intense enrichment of TRPA1 expression in splanchnic gastric afferents traced to their cell bodies in the thoracolumbar DRG. This was evident both in terms of the proportions of neurons expressing TRPA1 and the relative quantities of TRPA1 in the gastric population of DRG neurons. Further, we also found that TRPA1-IR nerve fibres in the rat stomach were expressed in the mucosa, around blood vessels in the submucosa, and in the external muscle layers. TRPA1 labelling was also observed in intramuscular endings, indicating that TRPA1 is expressed in gastric nerve fibres of extrinsic and intrinsic origin.
There may be additional roles for TRPA1 in visceral afferents, such as in detecting changes in osmolality, as shown for this channel in other systems.31 There may also be alternative roles for visceral afferents expressing TRPA1, such as in mediating local vasodilatation, which is suggested by their expression of the vasodilator SP and CGRP and their proximity to blood vessels. Indeed, we found that TRPA1-expressing neurons expressed TrkA, SP and CGRP. In gastric nerve fibres, TRPA1 was also expressed with the sensory neuropeptides.
The MAPK family includes ERK1/2, p38 MAPK, c-Jun N-terminal kinase/stress-activated protein kinase, and ERK5.32 Activity-dependent activation of ERK1/2 has been reported in the central nervous system, especially in the hippocampus.33 We examined the phosphorylation of ERK1/2 in the DRG and NG 2 min after stimulation, because activation of ERK1/2 in primary afferents is involved in acute nociceptive processing by a nontranscriptional mechanism, and examination of p-ERK1/2 is very useful as an indicator of the activated DRG neurons after noxious GD.24 We found that noxious GD induced activation of ERK1/2, predominantly in TRPA1-expressing neurons, which was prevented by TRPA1 AS-ODN. This suggests that primary afferent activation, through the TRPA1 by mechanical stimulation, may produce action potentials, which in turn results in the phosphorylation of ERK1/2 in DRG neurons. Alternatively, it is possible that ERK1/2 in the DRG may activate TRPA1 itself by phosphorylation.
The phosphorylation of MAPK may not be an accurate reflection of activity, and indeed, further studies are needed to establish the direct relationship between the phosphorylation of ERK1/2 in the DRG and the electrophysiological activity in individual neurons. Indeed, we found p-ERK1/2-labelled neurons that were not retrogradely labelled. However, the characteristics of p-ERK1/2 labelling in DRG neurons after noxious GD suggests that the p-ERK1/2 labelling is, at least in part, correlated with the activation state of the primary afferent neurons through TRPA1.
In the present study, we found that intrathecal TRPA1 AS-ODN significantly attenuated the activation of ERK1/2 in DRG, but not NG, neurons produced by noxious GD. Indeed, intrathecal injection of TRPA1 AS-ODN significantly decreased the expression of TRPA1 in the DRG, but not NG, neurons. Furthermore, intrathecal administration of the TRPA1 inhibitor HC-030031 reversed the response to noxious GD. Therefore, noxious mechanical stimuli might be carried through spinal splanchnic, but not vagal, afferents. Indeed, we confirmed that splanchnic nerve resection significantly affected the visceromotor response to noxious GD (fig 8C). An unexpected finding in the present study was that after noxious GD, an intensity-dependent ERK1/2 activation occurred in NG neurons. Interestingly, Traub et al34 reported significant elevation of c-Fos in the NTS after GD. However, consistent with previous reports,4 5 6 vagotomy did not significantly affect the response to noxious GD. Therefore, these results suggest that the NG and NTS neurons may not be involved in the perception of noxious signals caused by mechanical stimulation of the stomach, although the vagal pathways are activated after stimulation. Intriguingly, a recent report has shown that cold sensitivity is present in a large proportion of NG neurons and that TRPA1 appears to be a principal receptor molecule for cold transduction in NG neurons.35
We also found that intraperitoneal administration of the TRPA1 inhibitor HC-030031 also inhibited noxious GD-induced increase in visceromotor responses. However, there is a possibility that the effects of TRPA1 AS-ODN or HC-030031 on acute visceral pain could be mediated by the inhibition of TRPA1 activation in the spinal cord, but not in the DRG. In fact, it has been reported that TRPA1 is located at central terminals of primary afferent fibres and its activation modulates synaptic transmission in the spinal dorsal horn.36 This central modulation of sensory signals may be associated with physiological and pathological pain sensations. Because vagal and spinal afferents from TRPV1 knockout mice display altered mechanosensitivity,30 37 we believe that blockade of both TRPA1 and TRPV1 channels in primary afferents simultaneously may provide a more effective means to reduce visceral pain.
The gift of HC-030031 from Hydra Biosciences is gratefully acknowledged. We thank Dr Y Dai at Hyogo University of Health Sciences for valuable suggestions regarding the TRPA1 antibody.
Funding This work was supported, in part, by Grants-in-Aid for Scientific Research, and the Open Research Center grant, Hyogo College of Medicine, both from the Japanese Ministry of Education, Science, and Culture. This work was also supported by a grant from the Japan Health Sciences Foundation.
Competing interests None.
Provenance and Peer review Not commissioned; externally peer reviewed.
Ethics approval All animal experimental procedures were approved by the Hyogo College of Medicine Committee on Animal Research and were performed in accordance with the National Institutes of Health guidelines on animal care.