Caspase-1, originally described as interleukin converting enzyme, is an intracellular enzyme that plays a role in the maturation of two important mediators of inflammation, IL-1β and IL-18 [14–16]. Recently, it was suggested that caspase-1 also metabolises another inflammatory mediator, IL-33; however, contrary to IL-1β and IL-18, it seems that caspase-1 action upon IL-33 produces an inactive form of this cytokine . In the present study, we demonstrated that caspase-1 plays a crucial role in the cascade of events involved in the genesis of inflammatory hypernociception. Indeed, caspase-1 null mice present a reduction in carrageenin-induced hypernociception, and a selective inhibitor of caspase-1 presents an antinociceptive effect in this mice model. Our present results corroborate a previous finding of Elford et al. (1995) that showed that the treatment of rats with a selective inhibitor of caspase-1 reduced mechanical hypernociception in a model of yeast-induced inflammation . Extending these observations, it was recently shown that caspase-1 mediates the induction of a complex regional pain syndrome that developed after tibia fracture in rats .
It is well known that the release of cytokines in the inflammatory site is involved in the induction of inflammatory hypernociception. In addition, it seems that these pro-nociceptive cytokines are released sequentially . In this cascade, TNFα and chemokines, such as CXCL1/KC, are released earlier . Investigating whether caspase-1 might account for the release of these two cytokines, we detected that the levels of TNFα and CXCL1/KC in carrageenin-inflamed mice paws are similar in WT and casp1-/- mice, suggesting that the caspase-1 pro-nociceptive role is downstream of TNFα and CXCL1/KC. Our results also refute that the pro-nociceptive role of caspase-1 in carrageenin inflammation is dependent on its capacity to process IL-18. This conclusion is based in the fact that IL-18 null mice present similar hypernociception after carrageenin paw injection compared with WT mice. Nonetheless, the involvement of caspase-1 in complex regional pain syndrome triggered by tibia fracture seems to be dependent on IL-18 processing . Moreover, IL-18 also plays a role in the genesis of hypernociception in a model of immune inflammation induced by ovalbumin in previously immunised mice, and in this model caspase-1 might be responsible for IL-18 processing . We also recently showed that besides cytokines, the recruitment of neutrophils to inflammatory sites mediates carrageenin-induced hypernociception . However, it is also seems that the caspase-1 pro-nociceptive role is not related to neutrophils, because neutrophil accumulation in the mice paws of casp1-/- is similar to WT mice.
TNFα and CXCL-1/KC hypernociceptive roles in carrageenin inflammation were demonstrated to be, at least in part, dependent on the production of IL-1β . This fact, together with our present observation that casp1-/- mice present reduced hypernociception induced by TNF-α and CXCL1/KC but not by IL-1β and PGE2, is strongly suggestive that caspase-1 could be mediating inflammatory hypernociception through its product, mature IL-1β. Indeed, we detected that while the mRNA for pro-IL-1β increased at the same levels in casp1-/- and WT mice after carrageenin injection, there was a reduction in the mature form of IL-1β in casp-1-/- mice. This suggestion was supported by the observation that IL-1ra treatment reduced carrageenin-induced hypernociception. Corroborating the idea that IL-1β processing is the mechanism by which caspase-1 mediates inflammatory hypernociception, it was observed that IL-1ra treatment did not alter the neutrophil migration induced by carrageenin similarly observed in casp1-/- mice. It is somehow striking because IL-1β hypernociceptive effect is dependent on neutrophil migration . At this moment, we have the hypothesis that that in carrageenin-induced inflammation neutrophils are recruited mainly by TNFα and CXCR2 ligands (CXCL1/KC in mice or CINC-1 in rats) and they are activated in the site of inflammation by IL-1β that in turn induced the expression of COX-2 and the production of PGE2.
Although casp1-/- mice present a reduction in the expression of the mature form of IL-1β during carrageenin-induced paw inflammation, these mice still present the residual production of active IL-1β, suggesting that alternative mechanisms might trigger the induction of mature IL-1β in this model. For instance, there is evidence that pro-IL-1β can be cleaved by other proteases in addition to caspase-1, including elastase, proteinase 3 matrix metalloprotease 9 (MMP9), which are produced by neutrophils recruited to sites of tissue damage [20–23]. In this context, we have recently shown that MMP9 mediates hypernociception that developed during antigen-induced arthritis, a model in which IL-1β also plays a role .
Besides peripheral participation in the development of inflammatory hypernociception, we could not disregard the fact that caspase-1 might be involved by playing a central role in the response. In fact, activation of caspase-1 in the spinal cord has been associated with an increase of central IL-1β production that promotes COX-2 dependent inflammatory hypernociception . In the periphery, the pro-nociceptive effect of IL-1β was also mediated by cyclooxygenase-derived PGE2 because its effect was inhibited by treatment with indomethacin [6, 25]. Therefore, the observation that COX-2 expression and PGE2 production are reduced in casp1-/- mice challenged with carrageenin corroborates the importance of IL-1β in these processes. Contrary to these results, it was shown that although spinal glial-derived IL-1β is fundamental for the development of neuropathic pain after peripheral nerve injury, caspase-1 is not involved in this process . It was clearly demonstrated that, in this model, MMP-9 and MMP-2 are the enzymes responsible for IL-1β maturation playing a central role in neuropathic pain induction .
One question that emerges from these results is how caspase-1 is activated during carrageenin-induced paw inflammation. Regarding the mechanisms that trigger caspase-1 activation in the inflammatory process, there is now a large body of recent evidence showing that they are dependent on the assembly of cytosolic multiprotein complexes known as inflammasomes [26, 27]. Inflammasomes are formed by self-oligomerising scaffold proteins belonging to the NOD-like receptor family. There are, at least, four different inflammasomes: NALP1/NLRP1, NALP3/NLRP3, IPAF/NLRC4 and the HIN-200 family member, AIM2 . These molecules self-oligomerise after stimuli recognition and form high-molecular weight complexes that trigger caspase-1 autoactivation. Therefore, an interesting question is: which inflammasome is activated during carrageenin inflammation? In our knowledge, there is no study that addresses this issue. However, one can predict the involvement of the NALP3 inflammasome. This suggestion is based on the following indirect evidence: a) NALP3-containing inflammasomes that activate caspase-1 generally depend on stimulation of P2X7 ; b) P2X7 mediates carrageenin-induced inflammatory hypernociception [29, 30]; and c) P2X7 mediation of inflammatory hypernociception depends on stimulation of IL-1β because the antinociceptive effect of a selective P2X7 receptor antagonist is lost in IL-1 knockout mice . Collectively, these data suggest that NALP3 is the inflammasome that triggers caspase-1 activation in the mediation of carrageenin-induced inflammatory hypernociception, yet other inflammasomes could be also involved. For instance, the NALP1 inflammasome is required for caspase-1 activation and mediates the complex regional pain syndrome that developed after tibia fracture in rats . Therefore, additional studies using, for example, NALP3 null mice are required to solve this question.
In summary, the present study presents evidence that caspase-1 is involved in the genesis of inflammatory hypernociception. The participation of caspase-1 in this inflammatory symptom was not associated with the production of pro-inflammatory cytokines (TNFα, CXCL1/KC or IL-18) and recruitment of neutrophils, but was associated with the maturation of IL-1β in the inflammatory site. Caspase-1 seems to also be involved in the induction of COX-2 and consequently in the production of the directly-acting hypernociceptive mediator, PGE2. Together, these results added new information about the physiopathology of inflammatory pain. In conclusion, it is plausible to suggest that caspase-1 constitutes a real target to control inflammatory pain.