Transient receptor potential ankyrin 1 (TRPA1) receptor is involved in chronic arthritis: in vivo study using TRPA1-deficient mice

Chronic arthritis is one of the greatest health problems worldwide due to its high prevalence and insufficient therapeutic outcomes [1, 2]. Pain is the key complaint for arthritic patients, but its precise mechanism is poorly understood and its treatment is also a great unresolved problem. The currently used analgesics are often ineffective or their long-term use induces severe adverse effects [3‐5]. Therefore, there is a pressing need to understand the pathomechanisms of arthritis-related pain and identify more effective pharmacological targets for analgesia. Investigation of the regulatory role and activation mechanisms of peptidergic sensory nerves and the complexity of neuro-immune interactions in this condition can be a promising research area [6‐9].

Transient receptor potential ankyrin 1 (TRPA1) is a calcium-permeable non-selective cation channel predominantly expressed on capsaicin-sensitive primary sensory neurons, co-localized with the transient receptor potential vanilloid 1 (TRPV1) receptor in over 90 % of these cells [10, 11]. Besides the nociceptors, functional TRPA1 has also been described on non-neuronal cells, such as keratinocytes [12], fibroblasts [13], synoviocytes [14], macrophages [15, 16], lymphocytes [17], thymocytes [17] and endothelial cells [18] suggesting its complex involvement in inflammatory mechanisms. TRPA1 was characterized originally as a noxious cold- (<17 °C) activated channel [11], although its function in cold sensation is still a matter of debate. Several studies showed that TRPA1 is not required for normal cold sensitivity [19, 20], but others suggest that it acts as a major sensor for noxious cold [20], contributes to cold nociception and cold hypersensitivity after inflammation and nerve injury [21‐23]. TRPA1 is also directly stimulated by intracellular calcium [24] and a broad range of noxious endogenous oxidative products, such as 4-hydroxy-2-nonenal, hydrogen peroxide, hypochloride, hydrogen sulphide, 15-delta prostaglandin J2 [25‐28]. Furthermore, there are several exogenous irritants like mustard oil (allyl isothiocyanate: AITC) [29], cinnamaldehyde [30, 31], allicin [32, 33] and formalin [34] that are known to be potent agonists of TRPA1. Inflammatory mediators, such as bradykinin and serotonin, can sensitize this receptor and increase the responsiveness of the nerve endings [19, 35]. These findings suggest that TRPA1 may be involved in the development and maintenance of arthritic pain, but the precise mechanisms are still unknown.

Few data are available on the involvement of this receptor in inflammation and pain. Previously, some studies showed the role of TRPA1 in nociceptive processes in vivo using selective antagonists of the receptor. The pharmacological blockade of TRPA1 using intraplantar injection of AP-18 24 hours after complete Freund’s adjuvant (CFA) injection and intraplantar, intraperitoneal or intrathecal administration of HC-030031 1, 7 and 28 days after the administration of the adjuvant significantly attenuated mechanical hypersensitivity in mice and cold hypersensitivity in rats [36, 37]. Oral HC-030031 significantly reversed mechanical hypersensitivity in the CFA model of inflammatory pain at the 24-hour time point and the spinal nerve ligation model of neuropathic pain in rats 6 weeks post surgery [38]. HC-030031 given intraperitoneally or orally significantly reduced formalin- and AITC-evoked nocifensive behaviours [34, 38]. Its intraplantar administration prevented and reversed carrageenan-induced mechanical hypersensitivity in rats [10], inhibited AITC- and carrageenan-induced paw oedema in mice at the 3- and 6-hour time point [39]. In the monosodium iodoacetate (MIA)-induced osteoarthritis (OA) model systemic or intra-articular HC-030031 failed to block weight asymmetry and ongoing pain at the 1-hour time point [40], systemic injection of another TRPA1-antagonist, A-967079, reduced the evoked neuronal responses to high-intensity mechanical stimulation, but did not alter their spontaneous firing in osteoarthritic animals [41].

Although TRPA1-deficient mice are also valuable tools to investigate the role of this receptor in vivo, only few arthritis studies have been performed with these. Two articles reported that TRPA1 knockout (KO) mice did not develop acute pain, thermal or mechanical hypersensitivity after intraplantar injection of bradykinin or AITC, but displayed reduced sensitivity to intense cold and punctate mechanical stimulation [19, 23]. In contrast, one study showed that TRPA1 KO mice exhibited mechanical hypersensitivity 24 hours following CFA injection [36]. In a chronic model of inflammatory pain induced by intra-articular CFA, in TRPA1 KO mice mechanical hypersensitivity developed only 24 hours after CFA administration, but it was significantly smaller than in the wildtypes during the whole 4-week period [42]. TRPA1-deficient mice displayed attenuated carrageenan- and AITC-induced acute inflammatory paw oedema at the 3- and 6-hour time point [39]. The age-dependent role for TRPA1 in pain behaviour occurring in the adjuvant-induced arthritis model has very recently been revealed: old (>18 months old) TRPA1 KO mice developed significantly lower mechanical hypersensitivity as compared to their wildtypes throughout the 8-week experimental period, while the young (3–6 months old) TRPA1-deficient ones revealed that only during the first 2 weeks [43].

There are only few data regarding the role of TRPA1 in inflammation and pain, there are no knockout studies providing evidence for its precise role in chronic inflammatory pain processes in various time points, except for one article [42]. Therefore, in the present study, we aimed to analyze its involvement in chronic arthritis of different mechanisms and related nociception in comparison with acute models using TRPA1-deficient mice.

These results showed that chronic arthritis/osteoarthritis and related pain behaviours are mediated by the TRPA1 receptor activation. Adjuvant-induced oedema and inflammatory mechanical hypersensitivity, as well as MIA-evoked degenerative mechanical hypersensitivity with a potential neuropathic component and reduced weight bearing are diminished in TRPA1 KO mice. Furthermore, we presented the first evidence that TRPA1 is involved in the early neutrophil activation and late plasma extravasation in the CFA arthritis model. In contrast to some data indicating that TRPA1 mediates carrageenan-induced acute mechanical hypersensitivity in rats [10] and paw oedema in mice [39], our results clearly demonstrated that this receptor does not have a pivotal role in acute inflammation and hypersensitivity either in the knee joint or in the paw (Table 1).

The CFA-, MIA- and carrageenan-induced models are widely used and well-characterized rodent models of acute and chronic inflammation [56‐60]. CFA is heat-killed Mycobacterium tuberculosis, muramyl dipeptide involved in the cell wall causes a Th1- and cytokine-driven joint inflammation in rodents, which can mimic the main pathologic features of human rheumatoid arthritis [61‐64]. It can also be used to induce acute (1–3 days) and chronic inflammation (3 weeks). In our chronic model TRPA1 KO mice showed significantly decreased mechanical hypersensitivity from days 2 to 7, attenuated oedema from the second day during the total 3-week period, and reduced arthritic changes in the tibiotarsal joint at the 10-day time point as compared to their WT counterparts. Furthermore, we provided the first in vivo bioluminescence and fluorescence imaging data that TRPA1 activation mediates early neutrophil activation (on day 2) and late plasma protein extravasation (on day 7) in chronic arthritis. The regulatory role of TRPA1 receptors in chronic arthritic pain is supported by the finding that TRPA1 deletion hampered ongoing nociception in CFA-induced monoarthritis [42]. However, this study found no difference in knee joint swelling or histology, and concluded that the hyperalgesic function of TRPA1 was dissociated from joint swelling and inflammation [42]. These differences can be explained by the distinct features of these models (intra-articular vs. intraplantar/tail root administration, localized monoarthritis in the knee joint vs. small joints with systemic symptoms, different kinetics) and different investigational techniques (digital micrometer for the knee vs. plethysmometer for the paw). The knockout studies were confirmed by experiments with TRPA1-selective antagonists. Intraperitoneal, intraplantar or intrathecal injection of HC-030031 significantly reduced the long-lasting mechanical hypersensitivity at the 1-, 7- and 28-day time points [37]. The co-expression and interaction of TRPA1 and TRPV1 in a subpopulation of peptidergic, afferent Aδ and C fibres is well known [11, 29, 65, 66]. We previously showed decreased mechanical hypersensitivity, paw swelling and histopathological changes in TRPV1 KO mice in the same CFA model [44]. These results suggest that both TRPA1 and TRPV1 have relevant regulatory roles in long-term inflammation and hypersensitivity. Thermal hypersensitivity did not develop in our CFA model, the noxious heat threshold did not significantly change in response to the inflammatory reaction. Data showing no difference in thermosensitivity of TRPA1 KO and WT mice are in agreement with previous reports suggesting that TRPA1 is not a heat sensor [19, 23, 38]. Meanwhile, cold sensitivity increased in all groups independently of the inflammation suggesting the induction of cold hypersensitivity by the repeated measurements. Therefore, this investigational technique is not suitable for testing nociception in this model. However, pharmacological blockade of TRPA1 abolished CFA-evoked nocifensive behaviour to cold stimulus determined by a distinct methodology (1 % tetrafluoroethane spray on the hindpaw) in mice [37]. Therefore, the in vivo functional relevance of the noxious cold activation of TRPA1 in the processing of different pain modalities still remains an open question.

Intra-articular injection of MIA inhibits a key enzyme activity of glycolysis (glyceraldehye-3-phosphate dehydrogenase) in chondrocytes leading to progressive loss of these cells. The destruction of the articular cartilage, transient synovial inflammation and pain behaviour are similar to the human OA [67‐71]. MIA results in decreased weight bearing on the injured limb, movement-evoked pain and hypersensitivity [72, 73]. Previous data showed that chronic joint pain originates from the periphery by the sensitization of primary afferent nerves [74‐77]. It is well-known, that TRPA1 receptors are highly expressed on these fibres, but whether these receptors have a role in OA was investigated only with a selective receptor antagonist [40, 41]. Since systemic or intra-articular HC-030031 failed to block pain-related behaviours 1 hour post injection [40], and the blockade of TRPA1 did not reduce the MIA-induced spontaneous firing of sensory neurons [41], it could be concluded that MIA-induced ongoing pain is independent of TRPA1. Our present results with knockout mice support this concept regarding the knee joint swelling and histopathological alterations. However, we found that TRPA1 contributes to MIA-evoked decreased weight bearing and tactile hypersensitivity between days 3 and 10 of the experimental period.

Intraplantar or intra-articular injection of carrageenan is an appropriate acute technique for testing anti-inflammatory drugs [56]. It involves both neurogenic and non-neurogenic mechanisms characterized by prostaglandin production, cyclooxygenase (COX)-2 upregulation, formation of reactive nitrogen and oxygen species, as well as cytokines and other inflammatory mediators [78‐82]. In the present study we showed that TRPA1 does not have a role in acute carrageenan-induced paw oedema, mechanical and thermal hypersensitivity. In contrast, previous reports demonstrated that the selective TRPA1-antagonist HC-030031 and/or genetic deletion of TRPA1 inhibited the development and maintenance of carrageenan-induced inflammatory hypersensitivity in rats [10], and paw oedema in mice at the 3- and 6-hour time points [39], respectively. Distinct species, measuring time points, volumes, concentrations, and investigational techniques might be possible explanations for the differences. Similarly to what we found in TRPA1 KO mice, we previously described no difference in the carrageenan model in TRPV1-deficient mice [46].

Similarly to the carrageenan model, acute intra-articular CFA-evoked mechanical hypersensitivity and swelling were not altered by the deletion of the TRPA1 receptor over a 24-hour period, which is in agreement with previous reports [36, 42].

The distinct functional outcomes between the roles of the TRPA1 receptor in our four models can be explained by the wide distribution of TRPA1 on sensory nerves and non-neuronal cells, such as keratinocytes [12], fibroblasts [13], synoviocytes [14], macrophages [15, 16], lymphocytes [17], thymocytes [17] and endothelial cells [18]. Keratinocytes stimulated by TRPA1 agonist have been shown to increase the expression and release of pro-inflammatory interleukins [12], which can activate or sensitize sensory nerve endings [83]. The exposure of thymocytes to cinnamaldehyde accelerated T cell differentiation [17], which has a crucial role in the pathomechanism of CFA-induced arthritis. Macrophages, the other key players of CFA-evoked joint inflammation, also express TRPA1 mediating anti-inflammatory effects [15, 16]. Furthermore, there is a broad range of endogenous TRPA1 agonists produced locally during inflammatory processes that might differently modulate the receptor on the sensory nerves and non-neural structures. This would consequently trigger and/or inhibit the inflammatory cascades.

Our findings demonstrate an important regulatory role of the TRPA1 receptor in chronic arthritis/osteoarthritis and related pain behaviours in the mouse. Therefore, it might be a promising target for novel analgesic/anti-inflammatory drugs.