Pain in rheumatoid arthritis : bone and neuroinflammation-associated mechanisms

Abstract

Pain is often the primary reason patients with rheumatoid arthritis (RA) seek medical care. Despite effective disease control with currently available disease modifying antirheumatic drugs (DMARDs), there are still hurdles to overcome as a significant proportion of patients still report continuous pain. This suggests that the relationship between joint inflammation and pain severity is not linear. Currently few effective pain treatments for RA are available, which leads to individual and societal burden. Understanding the regulation of chronic pain in RA is thus vital to identify new drug targets and improve therapeutical strategies. This thesis explores alternative mechanisms of pain in RA with a specific focus on bone and neuroinflammation-associated mechanisms.

In Paper I, the contribution of osteoclasts to pain mechanisms was characterized in the collagen antibody-induced arthritis (CAIA) model. As previously reported, CAIA induces transient joint inflammation and persistent mechanical hypersensitivity that outlasts active inflammatory state. Herein, local bone erosion was detected in the calcaneus during both inflammatory and late phases of the CAIA model. Interestingly, while osteoclast activity was prominently increased during the inflammatory phase, pain-like behavior was reversed by two different osteoclast inhibitors in the late phase. In order to understand the contribution of osteoclast activity in nociceptive mechanisms, bone vascularization and innervation were examined. Both vascular and nerve densities were increased in the calcaneus during inflammation, but surprisingly remained elevated in the late phase despite resolution of joint inflammation. Notably, the CAIA-induced changes in bone, vascular and nerve densities in the late phase were attenuated by osteoclast-blocking agents correlating with suppression of osteoclast-derived angiogenic and neurogenic factors, such as netrin-1. Blockade of netrin-1 activity reversed CAIA-induced hypersensitivity in the late phase. Collectively, these findings suggest that the pronociceptive role of osteoclasts is not entirely dependent of their resorbing actions and that osteoclast inhibitors are effective in alleviating pain during the refractive phase of RA.

Janus kinase/signal transducers and activators of transcription (JAK/STAT) inhibitors represent a new class of DMARDs. In Paper II, the antinociceptive effects of the JAK1/2 inhibitor baricitinib on the CAIA model as well as the underlying mechanisms were determined. In this study, baricitinib produced reversal effects on CAIA-induced pain-like behavior, which was more pronounced in the late phase of the model. Importantly, the antinociceptive properties of baricitinib in the CAIA model do not completely covary with its anti-inflammatory effects. Intriguingly, no sign of JAK/STAT activation was detected in the dorsal root ganglia (DRGs) or spinal cords of CAIA-subjected mice, thus prompting other signaling pathways targeted by baricitinib to be explored. The effect of baricitinib on AAK1 activity was examined as this pathway was recently identified to be an additional target of baricitinib. mRNA levels of Aak1 and its downstream target Ap2m1 as well as phosphorylation and total protein of AP2M1 were upregulated in DRGs from the late phase of the CAIA model. Baricitinib treatment was able to normalize phosphorylation and total protein levels of AP2M1. Taken together, our data suggest that baricitinib may exert its antinociceptive effects through modulation of AAK1 rather than JAK/STAT signaling in the phase of refractive arthritis-induced pain.

Paper III and IV delineated the role of peripheral and spinal high mobility group box 1 (HMGB1) in arthritis-induced pain and if the pronociceptive actions of HMGB1 is sex-dependent. In Paper III, blocking the activity of HMGB1 in the periphery was shown to alleviate CAIA-induced pain-like behavior in male but not female mice. Interestingly, local injection of Toll-like receptor (TLR)4-activating disulfide HMGB1 induced mechanical hypersensitivity in both sexes, but was associated with more pronounced contribution of immune cells in males compared to females. CAIA induction has been associated with activation of microglia in the spinal cord. In Paper IV, disrupting microglial activity was shown to prevent development of disulfide HMGB1-induced pain-like behavior in male but not female mice. To further explore sex-specific differences, global spinal protein expression was examined using liquid chromatography-mass spectrometry. Surprisingly, several antinociceptive and anti-inflammatory proteins were elevated in only male mice that received the microglial inhibitor minocycline, some of which modulate protein activation cascade that converges on proteinase-activated receptor (PAR)2. Targeting the identified proteins individually, however, did not produce robust antinociceptive effects as minocycline. Overall, these studies demonstrate the important aspects of sex and cellular location in the contribution of peripheral and spinal HMGB1 and TLR4 to arthritis-induced pain.

In summary, this thesis has described three additional mechanisms of RA-induced pain. The findings suggest the involvement of osteoclasts, AAK1/AP2M1 and HMGB1 in mediating CAIA-induced hypersensitivity, particularly in the refractive state of the model. In addition, this work has highlighted the importance of mapping sex dimorphism and the prospective that pain relief is achieved differently in different sexes. Although more research are warranted in order to decipher the exact mechanisms that drive and maintain chronic pain in RA, this thesis has provided interesting mechanistic insights with respect to the bone environment and neuroinflammatory factors.