Non‐neuronal TRPA1 encodes mechanical allodynia associated with neurogenic inflammation and partial nerve injury in rats

The pro‐algesic transient receptor potential ankyrin 1 (TRPA1) channel, expressed by a subpopulation of primary sensory neurons, has been implicated in various pain models in mice. However, evidence in rats indicates that TRPA1 conveys nociceptive signals elicited by channel activators, but not those associated with tissue inflammation or nerve injury. Here, in rats, we explored the TRPA1 role in mechanical allodynia associated with stimulation of peptidergic primary sensory neurons (neurogenic inflammation) and moderate (partial sciatic nerve ligation, pSNL) or severe (chronic constriction injury, CCI) sciatic nerve injury.


| INTRODUCTION
The transient receptor potential (TRP) family of channels encompasses several nonselective cation channels expressed in a variety of cells, including a subpopulation of primary sensory neurons, where they encode sensory modalities that span from thermosensation to mechanical and chemical stimuli Story et al., 2003;Talavera et al., 2020). Major attention has been paid to the TRP vanilloid 1 (TRPV1), also known as the capsaicin (hot pepper) receptor, and the TRP ankyrin 1 (TRPA1), also known as the allyl isothiocyanate (AITC, a spicy component of wasabi of wasabi) receptor (Szallasi & Blumberg, 1999;Talavera et al., 2020). TRPV1 and TRPA1 are abundantly expressed in a heterogeneous subpopulation of primary sensory neurons that consists of peptidergic and non-peptidergic C-fibre and Aδ-fibre nociceptors (Bhattacharya et al., 2008). TRPV1 and TRPA1 stimulation results in the release of the neuropeptides, substance P (SP) and calcitonin gene-related peptide (CGRP; Nassini et al., 2014), which elicit plasma protein extravasation and adhesion of leukocytes to postcapillary venules (mediated by SP) and arteriolar vasodilatation (mediated by CGRP), collectively referred to as neurogenic inflammation (Geppetti & Holzer, 1996). Capsaicin, acting on TRPV1, is the prototypical agent that evokes sensory neuropeptide release and neurogenic inflammation (Szallasi & Blumberg, 1999). Capsaicin injection in the mouse and human skin also elicits an acute and transient nociceptive response or pain respectively, and in both species a sustained mechanical hypersensitivity (De Logu et al., 2022;LaMotte et al., 1992) that in mice is mediated by CGRP (De Logu et al., 2022).
In recent years, the role of TRPA1 in sustaining mechanical allodynia has been identified in mouse models of inflammatory, neuropathic, cancer and migraine pain. These include intra-articular injection of monosodium urate (Trevisan et al., 2014), hind limb ischemia and reperfusion (De Logu et al., 2020), partial sciatic nerve ligation (pSNL) (De Logu et al., 2017), alcoholic polyneuropathy , melanoma cells inoculation (De Logu et al., 2021) and CGRP injection into the periorbital skin (De Logu et al., 2022). In addition to the contribution of neuronal TRPA1, which signals agonistinduced acute nociception, a critical role of Schwann cell TRPA1 has been identified in macrophage-dependent (De Logu et al., 2021, 2017) and macrophage-independent  sustained mechanical allodynia.
Pain-like responses produced in rat models of inflammatory and neuropathic conditions have been reported to be reduced by first generation TRPA1 receptor antagonists (Eid et al., 2008;McNamara et al., 2007;Petrus et al., 2007;Wei et al., 2009), which, however, suffered from poor selectivity or suboptimal pharmacokinetics. More recently, TRPA1 deletion in rats by CRISPR technology, while attenuating acute nociception by AITC, failed to reduce mechanical hypersensitivity in models of inflammatory and neuropathic pain, such as those evoked by chronic constriction injury (CCI), the chemotherapeutic agent bortezomib and complete Freund adjuvant (Reese et al., 2020). These findings led to the conclusion that TRPA1 involvement in pathophysiological models of various pain states is confined to mice and cannot be replicated in other rodent species (Reese et al., 2020).
Here, we aimed to answering the general question as to whether the pro-algesic role of TRPA1 is limited to mice or if it can be observed in other rodent species. To this purpose, we explored whether TRPA1 is implicated in sustaining mechanical allodynia associated with neurogenic inflammation and involved in two models of neuropathic pain in rats. We found that while acute nociceptive responses elicited by AITC and capsaicin injection in the rat hind paw were dependent on their respective selective targets (TRPA1 and TRPV1, respectively), mechanical allodynia was exclusively due to CGRP release and the activation of non-neuronal TRPA1, most likely on Schwann cells, that senses, amplifies and sustains the pro-allodynic oxidative stress signal. We also confirmed that in rats the failure of TRPA1 antagonism fails to attenuate allodynia in a severe model of neuropathic pain (CCI), however TRPA1 and oxidative stress were markedly implicated in allodynia in the less severe pSNL model. Thus, TRPA1 seems to have a conserved ability to encode various pain modalities across different mammal species, including rats, in pathophysiological pain models.

| Animals
Sprague-Dawley rats (male, 150 g, Charles River, Milan, Italy, RRID: RGD_734476) were used throughout. Given that, in various previous reports in mice no gender difference was found in the pro-algesic What is already known • The TRPA1 channel has been implicated in various pain models in mice.

What does this study add
• The role of TRPA1 in mechanical allodynia has been investigated in a rat model of neurogenic inflammation and moderate/severe sciatic nerve injury.
What is the clinical significance • TRPA1 implication in mechanical hypersensitivity is a common feature in rodents and may be explored in humans.
role of TRPA1 (De Logu et al., 2020, here, in accordance with the 3Rs guidelines to minimize animal number, we used only male rats. The group size of n = 6 animals for behavioural experiments was determined by sample size estimation using G*Power (Version 3.1.9.6-available from https://gpower.software. informer.com/3.1/) (Faul et al., 2007) to detect size effect in a post hoc test with type 1 and 2 error rates of 5 and 20%, respectively.
Rats were allocated to vehicle or treatment groups using a randomization procedure (http://www.randomizer.org/). Investigators were blinded to treatments, which were revealed only after data collection.  (Lilley et al., 2020). Rats were housed in a temperature-and humidity-controlled vivarium (12-h dark/light cycle, free access to food and water, five animals per cage).
At least 1 h before behavioural experiments, rats were acclimatized to the testing room and behaviour was evaluated between 9:00 am and 5:00 pm. All the procedures were conducted following the current guidelines for laboratory animal care and the ethical guidelines for investigations of experimental pain in conscious animals set by the International Association for the Study of Pain (Kilkenny et al., 2010). Animals were killed with inhaled CO 2 plus 10-50% O 2 and confirmation of death was achieved by a physical method of killing (decapitation) (AVMA Guidelines for the Euthanasia of Animals, 2020).

| Partial ligation of the sciatic nerve
pSNL was performed in rats as previously described (Seltzer et al., 1990). Briefly, rats were anaesthetized with a mixture of ketamine (100 mg kg À1 ) and xylazine (10 mg kg À1 ), placed on a heated surface to maintain body temperature and the right sciatic nerve was exposed at high-thigh level. Under a magnification of 25Â, the dorsum of the nerve was carefully freed from surrounding connective tissues at a site near the trochanter just distal to the point at which the posterior bicep semitendinosus nerve branches off the common sciatic nerve. The nerve was fixed in its place by pinching the epineurium on its dorsal aspect, taking care not to press the nerve against underlying structures. A silicon treated silk suture was inserted into the nerve and tightly ligated so that the dorsal l/3-1/2 of the nerve thickness was trapped in the ligature. The wound was then closed. In sham-operated rats, used as controls, the right sciatic nerve was exposed, but not ligated. Rats were maintained in spontaneous breathing, monitored, and adequately rehydrated until fully recovered from anaesthesia. After surgery rats were maintained one per cage and fed with standard rodent chow (Envigo, Milan, Italy) and water ad libitum.
2.3 | CCI to sciatic nerve CCI to sciatic nerve was performed as previously described (Bennett & Xie, 1988). Briefly, rats were anaesthetized with a mixture of ketamine (100 mg kg À1 ) and xylazine (10 mg kg À1 ) placed on a heated surface to maintain body temperature and the common sciatic nerve was exposed at the level of the middle of the thigh by blunt dissection through the bicep femoris. Proximal to the sciatic trifurcation, about 7 mm of nerve was freed of adhering tissue and four ligatures (5.0 Ethicon chromic catgut) were tied loosely around it with about 1-mm spacing. Great care was taken to tie the ligatures, such that the diameter of the nerve was seen to be just barely constricted. In shamoperated rats, used as controls, the right sciatic nerve was exposed, but not ligated. Rats were maintained in spontaneous breathing monitored, and adequately rehydrated until fully recovered from anaesthesia. After surgery rats were maintained one per cage and fed with a standard rodent chow and water ad libitum. Tween 80 in 0.9% NaCl). Other rats received intraperitoneal (i.p., 10 ml kg À1 ) A-967079 (10, 30 and 100 mg kg À1 ), AMG0902 (AMG, 10, 30 and 100 mg kg À1 ) capsazepine (CPZ, 4 mg kg À1 ) or vehicle (4% DMSO 4% Tween 80 in 0.9% NaCl) before the stimulus or at day 15 after pSNL, CCI or sham surgery.

| Treatment protocols
In different experiments, rats were randomly allocated to the groups receiving perineural (p.n., 10 μl) or intrathecal (i.th., 10 μl) treatment with TRPA1 antisense (AS) or mismatch (MM) oligonucleotide (ODN) (10 nmol), once a day for four consecutive days, or once a day for four consecutive days, starting from day 10 to day 14 after pSNL, CCI, or sham surgery. TRPA1 AS-oligonucleotide sequence was 5 0 -TATCGCTCCACATTGCTAC-3 0 , TRPA1 MM-oligonucleotide sequence was 5 0 -ATTCGCCTCACATTGTCAC-3 0 . Perineural injections were performed by injecting the compound into the region surrounding the sciatic nerve at high thigh level of right hind limbs without skin incision using a microsyringe fitted with a 30-gauge needle.
2.5 | Behavioural assay 2.5.1 | Acute nociception Each rat was lightly restrained in a towel and an intraplantar injection of 20 μl was made to the right hindpaw using a 30-gauge disposable needle attached to a luer-tipped Hamilton syringe. Immediately after the i.pl. injection, rats were placed inside a plexiglass chamber and spontaneous nociception was assessed for 10 min by measuring the time (seconds) that the animal spent licking/lifting the injected paw.

| Paw mechanical allodynia
Paw mechanical allodynia was evaluated by measuring the paw withdrawal threshold using the up-down paradigm (Chaplan et al., 1994;Dixon, 1980). Rats were acclimatized (1 h) in individual clear plexiglass boxes on an elevated wire mesh platform, to allow for access to the plantar surfaces of the hind paws. Von Frey filaments of increasing stiffness (0.4, 0.6, 1.0, 1.4, 2, 4,8,10 and 15 g) were applied to the hind paw plantar surfaces of rats with enough pressure to bend the filament. The absence of a paw being lifted after 5 s led to the use of the next filament with an increased force, whereas a lifted paw indicated a positive response, leading to the use of a subsequently weaker filament. Six measurements were collected for each rat or until four consecutive positive or negative responses occurred. The 50% mechanical withdrawal threshold (expressed in grams) was then calculated. Rat Schwann cells were isolated from sciatic nerve of Sprague-Dawley rats (Tao, 2013). Briefly, sciatic nerve was dissected, the epineurium was removed and nerve explants were divided into 1-mm segments and dissociated enzymatically using collagenase (0.05%) and hyaluronidase (0.1%) in HBSS (2 h, 37 C). Cells were collected by centrifugation (150Âg, 10 min, room temperature) and the pellet was resuspended and cultured in DMEM containing fetal calf serum (10%), L-glutamine (2 mM), penicillin (100 U ml À1 ), streptomycin (100 mg ml À1 ), neuregulin (10 nM) and forskolin (2 μM). Three days later, cytosine-b-D-arabino-furanoside free base (ARA-C, 10 mM) was added to remove fibroblasts. Cells were cultured at 37 C in 5% CO 2 and 95% O 2 . Purity of primary Schwann cells cultured according to the present protocol reaches almost 100%. The culture medium was replaced every 3 days and cells were used after 15 days of culture.

| Cell cultures
For primary culture of rat dorsal root ganglia (DRG) neurons, DRGs (combined cervical, thoracic and lumbar) were bilaterally excised under a dissection microscope and enzymatically digested using 2 mg ml À1 of collagenase type 1A and 1 mg ml À1 of trypsin in 4 ml of HBSS for 35 min at 37 C. Ganglia were disrupted by several passages through a series of syringe needles (23-25G). Rat neurons were then pelleted by centrifugation at 1200Â rpm for 5 min at 4 C and resuspended in DMEM supplemented with 10% heat inactivated horse serum containing 10% heat-inactivated FBS, 100 U ml À1 of penicillin, 0.1 mg ml À1 streptomycin and 2-mM L-glutamine added with 100 ng ml À1 nerve growth factor (NGF) and 2.5 mM ARA-C and maintained at 37 C in 5% CO 2 and 95% O 2 for 2 days before being used.

| Materials
If with A-967079 (i.p. or i.pl.) did not affect the nocifensive behaviour but did prevent HPMA induced by capsaicin (Figure 1c,d).
To understand the mechanism underlying TRPA1-dependent HPMA, antagonists were given after the administration of the stimulus. Post-treatment with capsazepine did not affect HPMA elicited by both capsaicin and AITC (Figure 1e), whereas post-treatment with A-967079 markedly attenuated both responses (Figure 1f). These data suggest that spontaneous nocifensor behaviour elicited by TRPV1 and TRPA1 activators are mediated by the activation of the respective channel in the sensory nerve terminal. However, whatever the initial stimulus targeting the peptidergic nerve terminal, the prolonged HPMA is due to a common pathway implicating the TRPA1 channel.

| Cellular and molecular mediators of neurogenic inflammation associated allodynia
AS-oligonucleotide, which does not interfere with neuronal RNA, did not affect acute nocifensor behaviour, but attenuated HPMA evoked by AITC (Figure 3g). Treatment (p.n.) with TRPA1 AS-oligonucleotide did not affect acute nocifensor behaviour evoked by capsaicin, but reduced HPMA elicited by either capsaicin or CGRP (Figure 3h,i).
The expression of TRPA1 mRNA in rat Schwann cells was confirmed by RNAscope in rat sciatic nerve tissue and primary culture of rat Schwann cells by co-expression of TRPA1 mRNA with staining for the Schwann cell marker, S100 (Figure 4a,b). Primary rat Schwann cells were harvested and grown in culture and their identity was verified by RT-qPCR with the S100 gene expression (Figure 4c). TRPA1 mRNA expression was also confirmed in cultured rat Schwann cells and DRG neurons (Figure 4c). The identity of DRG neurons was verified by advillin (AVIL) gene expression (Figure 4c).

Exposure of cultured primary rat Schwann cells and DRG neurons
to AITC elicited a rapidly developing calcium response (Figure 4d,e) that was inhibited in the presence of A-967079 and in calcium-free medium. Exposure to CGRP elicited a delayed calcium response in rat Schwann cells (but not in DRG neurons) that was attenuated by olcegepant, A-967079 and calcium-free medium (Figure 4f To further verify the TRPA1 protein expression in rat Schwann cells, a photochromic selective TRPA1 ligand (A1CA) was used (Qiao et al., 2020). A1CA exhibits no fluorescence in free style due to the free rotation of rotors. However, its interaction with TRPA1 inhibits the rotation and enhances the fluorescence (Qiao et al., 2020).
rTRPA1-HEK293T transfected cells, HEK293T cells and cultured primary rat Schwann cells were exposed to A1CA (10 μM) for three minutes and then visualized with a fluorescent microscope.

| TRPA1 role in chronic constriction injury (CCI) and partial sciatic nerve ligation (pSNL)
Rats undergoing CCI or pSNL developed, at day 15 after surgery, a robust HPMA ipsilateral to the lesion that was not observed in sham rats. Systemic (i.p.) administration of two different TRPA1 antagonists, AMG-0902 and A-967079, while not affecting HPMA

| DISCUSSION
A large series of evidence supports the role of TRPA1 in models of inflammatory, neuropathic, cancer and migraine pain in mice (De Logu et al., 2021, 2017Trevisan et al., 2014). However, recent findings obtained in rats with genetic deletion of the TRPA1 channel by the CRISPR technology showed that, while the agonist-induced TRPA1-mediated acute nocifensor behaviour was attenuated, mechanical allodynia produced by neuropathic and inflammatory pain models was unaffected (Reese et al., 2020). From these findings, the hypothesis was advanced that the pro-algesic role of TRPA1 is confined to mice and it cannot be replicated in another rodent species and possibly in other species (Reese et al., 2020). Here, based on recent findings obtained in mice (De Logu et al., 2022), we explored in rats the role of TRPA1 in mechanical allodynia associated with neurogenic inflammation or neuropathic pain models.
F I G U R E 5 TRPA1 activation mediates hind paw mechanical allodynia (HPMA) in partial sciatic nerve ligation (pSNL), but not in chronic constriction injury (CCI) model. HPMA in rats 15 days after (a) CCI, (b) pSNL or sham procedure treated with intraperitoneal (i.p.) AMG0902 (AMG, 30 and 100 mg kg À1 ), A-967079 (A96, 10, 30 and 100 mg kg À1 ) or vehicle (veh). HPMA in rats 15 days after (c) CCI, (d) pSNL or sham procedure treated with capsazepine (CPZ, 4 mg kg À1 , i.p.) or veh. HPMA in rats 15 days after (e) CCI, (f) pSNL or sham procedure treated (once a day for four consecutive days starting from day 10 to day 14 after surgery) with perineural (p.n., 10 μl) TRPA1 antisense (AS) or mismatch (MM) oligonucleotide (ODN) (10 nmol). (g) Cumulative data for Trpa1 and S100 mRNA in rat sciatic nerve tissue 15 days after pSNL, CCI or sham procedure (n = 6 rats per group  (Bessac et al., 2008), promote a feed-forward pathway that sustains mechanical allodynia over 2-3 h. As AITC-evoked HPMA was attenuated by the same pharmacological interventions and with the same timing that were shown to inhibit capsaicin-evoked HPMA, an identical common pathway is proposed to sustain mechanical allodynia associated with neurogenic inflammation, independently from the stimulus that triggers the activation of peptidergic nociceptors and the ensuing neuropeptide release (Figure 6). attenuated in a dose-dependent manner by the two TRPA1 antagonists. However, in the two neuropathic pain models, CGRP release does not apparently play any role as HPMA was unaffected by olcegepant. It is known that neuropathic pain caused by peripheral nerve lesion (Wallerian degeneration) is due to haematogenic macrophage accumulation at the site of the injury (De Logu et al., 2017;Van Steenwinckel et al., 2015). Oxidative stress generated by invading macrophages mediates mechanical allodynia in both models, although its contribution seems higher in the pSNL model, as the ROS scavenger, N-tert-butyl-α-phenylnitrone, appeared superior in reducing HPMA in the pSNL model than in the CCI model. The unique redoxsensitivity of TRPA1 (Hinman et al., 2006;Macpherson et al., 2007) and the higher oxidative burden in the pSNL might be the reason for the channel involvement in this model. Additional explanations may be proposed. There is evidence of remarkable differences in the nerve lesions in the footpad skin produced in rats by CCI and pSNL (Ma & Bisby, 2000). Whereas in CCI PGP-9.5+ve, nerve fibres dramatically decreased within two weeks, in the pSNL model the decrease was partial and underwent a time-dependent recovery (Ma & Bisby, 2000).
It is possible to hypothesize that the less severe pSNL lesion preserves the integrity of the unit composed by the nerve fibre and the surrounding Schwann cells that contribute to mechanical allodynia. In the more severe CCI lesion, the structural loss of peripheral sensory axons excludes their contribution to generating pain signals.
A TRPA1 antagonist failed to reduce pain symptoms in patients with chronic painful diabetic neuropathy (Jain et al., 2022). However, it did produce a statistically significant attenuation in a subgroup of patients with preserved sensory nerve function and therefore with a less severe neuropathy (Jain et al., 2022). These findings show some similarity with the present rat data, where HPMA associated with a less severe nerve lesion was TRPA1-dependent. The lack of effect of a CGRP antagonist in either the CCI or pSNL model underlines the unique role of CGRP in migraine (Edvinsson et al., 2018;Nassini et al., 2014) and not in other pain conditions, as indicated by the failure of an anti-CGRP monoclonal antibody in reducing pain in patients with osteoarthrosis (Jin et al., 2018). It is possible that in Wallerian degeneration the oxidative burden required for targeting Schwann cell TRPA1 is provided by the massive ROS generation produced by invading macrophages, while under these circumstances the contribution of CGRP to the overall oxidative burden is negligible.
Primary hyperalgesia, a pain response due to sensitization of peripheral nociceptors, has been reported in the cutaneous area of inflammation following the application of a variety of stimuli, including capsaicin (LaMotte et al., 1992). The proposal by Sir Thomas Lewis (Lewis, 1936) that a chemical substance, released from collateral branches by the antidromic invasion of propagated action potentials originating from the injured nerve terminal, causes the flare (inflammation) and increases the sensitivity of other fibres responsible for pain can be applied to the present findings in rats. This mechanism, recently reported in mice (De Logu et al., 2022) and confirmed here in rats, might be a common feature that should be explored also in humans. In conclusion, we have reported that the contribution of TRPA1 to mechanical allodynia appears to be present in models of neurogenic inflammation and models of neuropathic pain characterized by moderate nerve injury, although in this latter case the contribution of CGRP is absent.