Historically, pain among patients developing inflammatory arthritis has been thought to be mainly attributed to peripheral inflammation. This has notably led to the development of a new class of therapeutics acting on key tissues to decrease inflammation, called disease-modifying antirheumatic drugs (DMARDs), which includes conventional DMARDs, such as methotrexate or anti-TNF monoclonal antibody biologics . Although DMARD therapy indirectly manages pain symptoms by preserving joint tissues and slowing down the disease progression, a large proportion of patients still reports feeling moderate to severe pain . Consequently, more attention has recently been directed towards the treatment of pain itself, with the development of new analgesic agents targeting peripheral nociceptive pathways, such as CNTX-4975, a synthetic form of capsaicin which selectively targets the TRPV1 receptor or Tanezumab, an anti-nerve growth factor (NGF) monoclonal antibody. The promising analgesic results obtained in recent phase III clinical trials with these new treatment options further encourage the development of new pharmacological strategies that successfully target nociceptors .
Among the therapeutic options, inhibition of the CCL2/CCR2 chemokine axis holds great promise for controlling chronic painful arthritis. Indeed, the CCL2/CCR2 signaling has been found to play key roles in peripheral and spinal nociceptive processing, mediating nociceptor sensitization and increase in the synaptic transmission in the spinal dorsal horn [4, 9, 11, 22, 58]. Importantly, standing at the crossroads of the immunobiology and neurobiology, the CCL2/CCR2 chemokine system is also able to trigger peripheral inflammation at the distal site, to promote neuron-glia interaction, and to orchestrate the neuroinflammation response through the recruitment of peripheral T cells and monocytes and/or activation of resident glial cells [12, 59]. Here, we provide significant mechanistic insights into the role of the CCL2/CCR2 signaling within the DRG in the development of peripheral inflammation, nociceptor sensitization and pain hypersensitivity. As schematically represented in Fig. 9, peripheral tissue injury followed by intraplantar injection of CFA, which results in paw oedema and inflammation induces an increase in CCL2/CCR2 and SP expression in ipsilateral DRGs. This upregulation is accompanied by an enhanced excitability of primary nociceptive neurons on days 3 and 10 post-CFA, as revealed by the CCR2-dependent increase in intracellular calcium mobilization following CCL2 stimulation. As shown using the ex vivo superfusion of DRG explants of CFA-treated rats, this is followed by a potassium-evoked calcium-dependent release of CCL2. Finally, the excitation and sensitization of nociceptors following peripheral inflammation drives the anterograde transport of SP at their peripheral nerve terminals as well as paw swelling. Importantly, our results highlight that blockade of the CCL2/CCR2 signaling following repeated i.t. administration of the CCR2 antagonist, INCB3344 reduces the expression of both CCL2 and SP in DRGs of CFA-treated rats, dampens sensory neuron excitability by limiting the intracellular calcium mobilization and subsequently decreases peripheral transport and release of SP at the terminal nerve endings. Then, this pharmacological inhibition of CCR2 significantly reduces the neurogenic inflammation as well as the stimuli-evoked and movement-evoked nociceptive behaviors in CFA-treated rats.
The results of the present study reinforce the previous concept on the neuromodulator/neurotransmitter role of CCL2 [22, 60]. Indeed, our data support the idea that CCL2, synthesized and released by the soma of DRG neurons, directly excites sensory nociceptive neurons by autocrine and/or paracrine processes under peripheral chronic inflammation. Accordingly, previous findings demonstrated in the neuropathic pain model that both CCL2 and CCR2 were upregulated in injured DRGs [13, 61–63] and that application of CCL2 increased the excitability of acutely dissociated small sensory neurons . Likewise, CCL2-positive DRG neurons were found to be increased in inflamed rat tissues [64, 65]. The specific downstream signaling pathways by which CCL2 drives the neuronal hyperexcitability under chronic inflammatory pain conditions yet remain to be uncovered. However, since we demonstrate, as previously reported in neuropathic pain models [61, 66, 67], that following acute (D3) or chronic (D10) CFA treatment, CCL2 elicits a greater calcium elevation in sensory neurons through a CCR2-dependent mechanism compared to naïve DRG neurons, we can hypothesize that functional changes in extracellular calcium influx and/or CCL2-induced calcium release from internal stores might underlie CCL2-induced neuronal activation. In that sense, it has been shown that inhibition of N-type Ca2+ channels by the omega-conotoxin GVIA channel blocker or treatment with either thapsigargin or ryanodine significantly reduced CCL2-induced intracellular calcium influx and the concomitant release of CGRP by primary sensory neurons . In the same line of thinking, the secretion of CCL2 enhanced the activity of voltage-dependent Ca2+ channels by inducing upregulation of the a2d1 subunit expression in DRG neurons . Other modes of action of CCL2/CCR2 could also drive the neuronal hyperexcitability. Indeed, we and others have demonstrated that CCL2 enhanced sensory neuron excitability by increasing the functional activity of tetrodotoxin-resistant (TTX-R) sodium channel Nav1.8 currents [54, 70]. This effect is CCR2-mediated as treatment with INCB3344 blocked the potentiation of Nav1.8 currents by CCL2 in a concentration-dependent manner [54, 71]. Consistent with this hypothesis, mRNA and protein expressions of Nav1.8 are upregulated in DRG neurons following peripheral inflammation and knockdown of Nav1.8 or use of Nav1.8 selective channel blockers reverse inflammation-induced hyperalgesia [72–76]. Similarly, the increase in CCR2/CCL2 signaling observed following tissue inflammation could cause the peripheral sensitization of DRG nociceptive neurons and drive the hyperalgesic state by upregulating the expression and function of the capsaicin-sensitive TRPV1 ion channel [19, 60, 70]. This idea is supported by the presence of CCL2/CCR2 within TRPV1-expressing sensory neurons [14, 60, 64]. In turn, as demonstrated in in vitro superfusion assay, capsaicin evokes calcium-dependent release of CCL2 [14, 19] and TRPV1 inhibition decreases CCL2-induced hyperalgesia . Collectively, these results demonstrate that the CCL2/CCR2 axis and TRPV1 act in unison to sensitize nociceptors. It is therefore tempting to speculate that part of the analgesic action of the synthetic capsaicin CNTX-4975, which targets TRPV1 can be associated to downregulation of CCL2/CCR2 signaling.
To gain further insights into the mechanisms by which CCR2 activation induces inflammatory hypernociception, we determined whether the increase in intracellular calcium mobilization was translated into a greater CCL2 secretion by DRG explants from CFA-treated rats. As previously observed in naïve and neuropathic animals [14, 19], K+ stimulation inducing neuronal depolarization increases CCL2 release from CFA-exposed DRGs compared to controls. It is widely accepted that extracellular calcium influx and calcium-induced calcium release (CICR) from internal stores play an important role in the release of the pro-nociceptive neuropeptides SP and CGRP from nociceptors [68, 78]. Therefore, since CCL2 stimulates intracellular calcium elevation through both ryanodine-sensitive calcium stores and N-type Ca2+ channels, the release of CCL2 from CFA-exposed DRGs may thus influence the transport and release of these pain-related neuropeptides by presynaptic mechanism. Accordingly, our results demonstrate for the first time that the anterograde transport of SP (but not CGRP) towards the peripheral nerve terminals was inhibited by blocking CCR2 using INCB3344. Interestingly, despite the increase in CCL2 mRNA expression and release by CFA-exposed nociceptors, immunohistochemical analysis of ligated nerves reveals no increased CCL2 immunoreactivity in CFA-treated animals, ruling out the contribution of a DRG-derived CCL2 release toward the periphery, at least at day 10 post-CFA. This is consistent with the demonstration that CCL2 is locally produced at the inflammation site by macrophages/monocytes in CFA inflamed rats and that treatment with INCB3344 dose-dependently inhibits macrophage influx [47, 79]. As previously shown , CCL2 is probably conveyed to the terminals of nociceptors and released at the spinal dorsal horn to modulate the activity of post-synaptic neurons and surrounding glial cells. Although we did not further investigate the cellular mechanisms behind the relation between CCL2 and SP, CCR2 is known to sensitize TRPV1 , whose stimulation induces the release of SP from sensory nerve fibers [80, 81]. Surprisingly, i.t. treatment with INCB3344 does not induce a decrease in the anterograde transport of CCL2 and even increases its accumulation centrally to the sciatic ligature. This seems to indicate that chronic blockade of CCR2 activation leads to increased CCL2 expression, as previously reported in a clinical study in patients with advanced solid tumors treated with an anti-CCL2 human monoclonal antibody .
Primary afferent neurons can directly contribute to peripheral inflammation and immune cell recruitment through the release of neuropeptides, such as SP and CGRP [83, 84]. As superfusion experiments does not distinguish between CCL2 release toward the spinal cord or the periphery, we initially thought that CCL2 would be concomitantly released with SP and/or CGRP following CFA intraplantar administration, thus contributing to the neurogenic inflammation process. As indicated above, contrary to SP, CCL2 does not seem to be transported to the peripheral inflammation site. In accordance with these findings, we and others previously reported that CCL2 can be packaged into SP-containing vesicles as well as SP-free vesicles [14, 60]. This might suppose that these two CCL2-containing populations of secretory vesicles can be alternatively released in response to nociceptive signals and then enhanced nociceptor sensitization and pain hypersensitivity. Interestingly, our results also reveal that the nerve ligation reduces peripheral inflammation, indicative of a contribution of neurogenic inflammation to the CFA-induced peripheral oedema. Moreover, the reduction in hind paw volume was similar to INCB3344-treated animals, suggesting that CCR2 activation at the DRG level contributes to peripheral inflammation, probably through the release of SP. Accordingly, administration of NK1 antagonists (i.e. the main SP receptor) reduces the plasma extravasation induced by intra-articular administration of carrageenan [85, 86]. Finally, as joint inflammation directly contributes to joint pain , this peripheral reduction in swelling could contribute to the observed analgesic efficacy of INCB3344.
There is an abundant literature supporting the role of CCL2 and its receptor CCR2 in the regulation of nociceptive transmission, especially for the management of chronic neuropathic pain [11, 12, 58, 59, 87]. Here, we further unveil the therapeutic potential of a CCR2 antagonist to relieve the pain behaviors associated to painful inflammatory conditions. We first demonstrated that CCL2 exerted a pronociceptive action in the inflammatory phase of the formalin test through the exacerbation of the pain-related behaviors induced by a mild injection of formaldehyde. As expected, co-administration of the CCR2 antagonist INCB3344 completely prevented CCL2-induced pain hypersensitivity, thus indicating that CCL2 elicited pain facilitation via a CCR2-dependent mechanism. Similarly, mice overexpressing CCL2 were hypersensitive to chemical-induced nociception , while mice deficient for CCR2 displayed decreased nociception in the inflammatory phase of the formalin test [16, 88]. Accordingly, intracisternal administration of CCL2 also facilitated formalin-induced scratching behavioral responses in the orofacial area . Mechanistically, it is well accepted that the inflammatory phase of the formalin test results from the combination of activation of primary afferent fibers by peripheral inflammatory mediators and functional changes in the dorsal spinal horn, notably through NMDA and NK1 receptor activation, thus leading to central sensitization . Interestingly, it was recently demonstrated that in addition to sensory neuron modulation, CCL2 participates in central sensitization by potentiating the activity of NMDA receptor currents in CCR2-expressing excitatory neurons located in lamina IIo of the spinal dorsal horn under peripheral inflammation [91–93]. Thus, these results further emphasize that CCL2 is anterogradely transported by primary afferent neurons to be released in the spinal dorsal horn in inflamed rats.
Chronic pain in patients with RA or OA leads to important physical distress as well as to the loss of patients’ autonomy and quality of life. Weight bearing activities are the main source of severe pain episodes in people suffering from arthritis [94, 95]. In addition, it is clinically demonstrated that most pharmacological agents relieve pain at rest while being less effective on movement-evoked pain . Here, we demonstrated that repeated i.t. injection of INCB3344 exerted tactile allodynia as previously observed in neuropathic, postoperative and cancer-induced bone pain models using anti-CCL2 antibodies and small molecule antagonists of CCR2 [15, 21, 97–103]. In contrast, inhibition of CCR2 using intraplantar or subcutaneous injection of the antagonist RS504393 was only effective in revering thermal hyperalgesia in CFA inflamed mice . Importantly, this decrease in mechanical hypersensitivity was accompanied by a reduction in the ambulation-evoked pain behaviors in freely-moving CFA-treated rats. Indeed, in line with clinical reports, CFA-induced paw inflammation generated an important load redistribution on the contralateral non-injured limb. Following repeated administration with INCB3344, we observed a partial recovery of the weight borne on the ipsilateral limb. This reversal in pain-induced weight redistribution was of a similar extent than the reversal in stimulus-evoked mechanical allodynia. Interestingly, Longobardi et al. recently showed that systemic blockade of CCR2 by RS504393 improved the weight redistribution in a murine model of injured-induced OA (i.e. destabilization of medial meniscus; DMM) . Likewise, mice invalidated either for CCL2 or CCR2 exhibited less pain-related behaviors post-DMM .