Capsaicin, the active ingredient of hot peppers, selectively stimulates and then desensitizes a subpopulation of C-fiber polymodal nociceptive and Aδ sensory nerve fibers that are classified as capsaicin-sensitive afferents (1, 2). Capsaicin acts via transient receptor potential vanilloid 1 (TRPV1), which was cloned in 1997 and was formerly termed vanilloid receptor 1 (3). TRPV1 is a nonselective cation channel, which can be activated by noxious heat, protons, and vanilloids such as capsaicin, as well as a range of putative endogenous mediators (4, 5). The acute stimulation of sensory nerves by capsaicin is associated with acute pain and neurogenic inflammation (1, 2). Thermal nociceptive thresholds, observed after exposure to a hot plate or inflammatory insult in the paw, have been shown to be blunted in TRPV1-knockout mice, indicating the potential for TRPV1 in mediating inflammatory hyperalgesia (6, 7). The stimulation of TRPV1 leads to release of neuropeptides, including substance P and calcitonin gene-related peptide (CGRP) from sensory nerves, (2, 8, 9). Neuropeptides have the potential to contribute to inflammatory disease as the “neurogenic component” via a variety of mechanisms (10).
Levels of neuropeptides have been shown to be increased in samples of synovial fluid from patients with rheumatoid arthritis (11), and the ligation of sensory nerves in experimental joint inflammation is associated with a beneficial effect (12), providing evidence for involvement of sensory nerve activation in arthritis. Indeed, continued topical treatment with capsaicin has been shown to be beneficial in patients with hand osteoarthritis (13), and capsaicin creams are marketed for the alleviation of pain in diseases including both osteoarthritis and rheumatoid arthritis (14). There is evidence from rodent studies that TRPV1 activation can influence pathophysiologic events in arthritis. For example, capsaicin pretreatment depletes substance P, and this has been associated with attenuation of the severity of arthritis (15–18). However, it is difficult to utilize studies involving capsaicin to pinpoint the influence of endogenous activation of TRPV1 in vivo on joint inflammation, due to the capsaicin-stimulated events leading to both sensory nerve depletion and nonspecific desensitization. In addition, pharmacologic antagonists developed to date are, to our knowledge, unsuitable for studies of long-term responses (19). Thus, mice lacking TRPV1 present a distinct experimental advantage, since the influence of TRPV1 can be investigated in the absence of exogenously administered vanilloids.
In the present study, TRPV1−/− mice were used in protocols involving recently developed techniques designed to enable correlation of knee swelling and inflammatory edema (20). Unilateral adjuvant-induced arthritis was elicited by intraarticular (IA) injection of Freund's complete adjuvant (CFA) into the knee joint. The objective was to investigate the sensory neurogenic involvement through use of methods developed to facilitate the analysis of microvascular events in addition to the measurement of nociceptive behavior. The results provide evidence that the deletion of TRPV1 led to reduced acute (18 hours) and chronic knee swelling. Acute knee swelling was associated with TRPV1-sensitive hyperpermeability in the knee joint. Knee swelling was also reduced long term in TRPV1−/− mice, and reduced nociceptive responses were observed. The techniques were also used to evaluate the role of TRPV1 in the local generation of, and acute effects of, tumor necrosis factor α (TNFα). Our findings provide evidence for a key role of TRPV1 in mediating ongoing, in addition to acute, joint inflammation.
- Top of page
- MATERIALS AND METHODS
In this study we demonstrated a pivotal role of TRPV1 in mediating vascular changes that occur in both acute and chronic joint inflammation. Results from studies of acute effects (at 18 hours) showed that knee swelling is reduced in TRPV1−/− mice compared with WT mice. Furthermore, this decreased swelling is directly related to attenuated plasma extravasation, suggesting that TRPV1 participates in influencing hyperpermeability for at least 18 hours after induction of joint inflammation. Despite this, TNFα levels and polymorphonuclear leukocyte levels were found to be similar in WT and TRPV1−/− mice at this time point, providing evidence that TRPV1 acts as a modulatory, rather than a primary, mediator. Studies with TNFα administered IA over 4 hours, by which time TNFα-induced synovial leukocyte accumulation is high, revealed a link between knee swelling and the presence of TRPV1, since TNFα-induced knee swelling was significantly greater in WT compared with TRPV1−/− mice.
Importantly, when the chronic effects of CFA-induced joint inflammation were observed over a 3-week period, knee swelling in TRPV1−/− mice showed an altered profile compared with that in WT mice. The responses in WT mice remained significantly increased for 1 week, while those in the TRPV1−/− mice were lower throughout, with a nonsignificant increase observed at 1 week. These results implicate TRPV1 in maintaining the inflammatory response. Furthermore, measurement of thermal nociceptive thresholds revealed a stable level in CFA- and vehicle-treated limbs except for a significantly decreased nociceptive threshold at 24 hours and at 1 week in WT, but not TRPV1−/−, mice. Taken together these results indicate an essential role of TRPV1 in the ongoing as well as the acute pathophysiologic events in joint inflammation.
An important aspect of sensory nerve biology is that, soon after the onset of inflammation, the different classes of primary sensory neurons and their dorsal root ganglion (DRG) cell bodies exhibit phenotypic changes. Findings in some rodent studies have suggested that these include up-regulation of TRPV1. For example, an increase in TRPV1 protein (but not messenger RNA) in DRG neurons at 2 days after induction of CFA inflammation, which resulted in an increase in TRPV1 levels in peripheral nociceptor terminals, has been reported (26). The notion of TRPV1 up-regulation during inflammation is further supported by findings in studies of CFA-induced inflammation (27, 28). In comparison, a negligible increase in the number of TRPV1-positive L1–L5 ganglia receptors was found in a model of CFA-induced monarthritis in rat knees examined at 3 days (acute response) and 21 days (chronic response). The variation in these results highlights the importance of analyzing functional, in addition to histochemical and molecular, components. Altered activation of the receptor leading to a functional response does not appear to be necessarily associated with up-regulation of TRPV1.
In our study, bilateral hyperalgesia was seen at 24 hours postinjection in the WT mice, but this had disappeared by 1 week. There is evidence that dorsal root reflexes occur bilaterally in fine articular afferents in CFA-induced arthritis in the rat (24). Dorsal root reflexes are action potentials that originate in afferent fibers at the spinal cord entry point and propagate to the periphery (24). In addition, bilateral up-regulation of the expression of neurokinin 1 and bradykinin 2 receptors during acute inflammation in the DRG neurons of rats with antigen-induced arthritis has been demonstrated (29). Both the central nervous system and local inflammatory processes have been shown to be involved in bilateral inflammation (30). As discussed above, phenotypic changes occur in sensory neurons during inflammation. It may be that activity from the central nervous system influences phenotypic changes in the DRG neurons in the periphery, thus leading to bilateral hyperalgesia. Our results observed in WT but not TRPV1−/− mice suggest an essential involvement of TRPV1 in this process. However, TRPV1 protein was not seen to be increased in the contralateral paw at 48 hours after CFA injection in the rat (28).
The finding of similar levels of TNFα in synovial fluid extracts from WT and TRPV1−/− mice led us to examine the possibility that TNFα may influence TRPV1 activity due to an action of TNFα upstream from TRPV1. Expression of TNF receptor subtypes occurs on cultured neonatal DRG cells (31, 32), and TNFα can influence neuropathic pain via a p38 MAPK–dependent mechanism (33). Furthermore, TNFα and TNF receptor 1 are suggested to play a prominent role in priming sensory nerve endings in mouse peripheral airways, leading to potentiation of a sensory nerve–mediated tracheal hyperpermeability response (34). The finding that TNFα-induced knee swelling is greater in WT mice compared with TRPV1−/− mice is consistent with the concept that TNFα can prime TRPV1 sensory nerve–mediated swelling in the knee joint, via as-yet-uncharacterized mechanisms.
The most common hypothesis is that the inflammatory events following TRPV1 activation are mediated by release of neuropeptides, principally substance P (acting via its vasoactive neurokinin 1 receptor) and CGRP, which are up-regulated in clinical and experimental joint inflammation (29, 35–38). Substance P stimulates acute plasma protein extravasation when injected into joints (39, 40), although its possible role in more severe and chronic phases of disease is unclear (40, 41). However, we have recently shown that there was no difference in responses in the CFA-induced model of joint inflammation, used in the present study, in WT versus neurokinin 1–knockout mice after 18 hours (20). Thus, one cannot assume a pivotal involvement of tachykinins or the major vasoactive tachykinin neurokinin 1 receptor in this model of arthritis. Results of clinical studies with neurokinin 1 antagonists do not suggest that they have a beneficial effect, at least in osteoarthritis (42), although there has been some debate regarding the potency of neurokinin 1 antagonists used in clinical trials to date (43).
The role of CGRP is less well understood. It is a potent vasodilator that can potentiate joint plasma extravasation (44), and it has been suggested to influence the inflammatory process in the joint (45–48) and contribute to hyperalgesia (49). A CGRP antagonist has recently been shown to have a favorable effect in the treatment of pain associated with migraine (50), and it is possible that CGRP is responsible for mediating some of the proinflammatory effects observed after TRPV1 activation. However, sensory nerve–mediated events are more complex in that activated sensory nerves also release antiinflammatory neuropeptides, in particular, somatostatin (51), in and around the joint. It should be noted that if the sum effect of TRPV1 activation is antiinflammatory, one would not expect the deletion of TRPV1 to act in the beneficial manner observed in the present study.
In summary, we propose that TRPV1 is involved in mediating joint swelling at both the acute and chronic stages and is essential in the development of a bilateral decreased threshold of thermal nociception at 24 hours, which persists for 1 week on the arthritic ipsilateral side. We conclude that TRPV1 does not influence the acute generation of the prominent inflammatory cytokine TNFα. However, this cytokine may be involved in stimulating or potentiating TRPV1-mediated responses. These results are consistent with the possibility that TRPV1 antagonists may have therapeutic potential for both acute and chronic inflammatory joint disease. Indeed, it has been recently shown that a novel TRPV1 antagonist, AMG 9810, reverses CFA-induced thermal and mechanical hyperalgesia in the rat paw (52), illustrating the potential of TRPV1 antagonists in the treatment of inflammatory disease.