Antibody- and osteoclast-dependent mechanisms of chronic pain

Abstract

Chronic pain is a frequent condition that affects approximately 20% of worldwide population with a higher prevalence among women. Due to the lack of optimal treatment, chronic pain leads to decreased life quality as well as significant physical and psychological complications. Consequently, it has become a major socio-economical problem, creating huge expenses in the form of medical care, sick leave and loss of productivity. Recent findings in the fields of immunology and neuroscience have underlined the important effector functions of autoantibodies in the development of several chronic pain conditions. Although pain in autoimmune diseases has been historically associated with inflammation due to the antibody-induced activation of inflammatory pathways, pain in these conditions often occurs before inflammation development or is the only symptom of the disease. Increasing amount of preclinical evidence suggests that autoantibodies engage multiple mechanisms, which give rise to hypersensitivity independent of inflammation. Thus, the aim of this thesis was to explore the mechanisms behind the pronociceptive properties of autoantibodies isolated from rheumatoid arthritis (RA) and fibromyalgia patients.

In Study I we have used a combination of two RA-associated monoclonal antibodies (mAbs) isolated from B cells of RA patients, to study the connection between increased bone erosion and pain in RA. We found that mice injected with 1103:01B02/1325:01B09 mAbs developed long-lasting mechanical hypersensitivity and bone erosion in the absence of overt joint inflammation. Furthermore, we found that pain-like behavior induced by 1103:01B02/1325:01B09 is associated with a moderate increase in the expression of few proinflammatory factors in the joints and is resistant to treatment with conventional non-steroidal anti-inflammatory drugs (NSAIDs), suggesting it does not depend on classical inflammatory processes. Instead, we demonstrated that inhibiting osteoclast activity and acid-sensing ion channel 3 (ASIC3) signaling prevented the development of 1103:01B02/1325:01B09-induced hypersensitivity. We have also established that secretory phospholipase A2 (sPLA2) and lysophosphatidylcholine 16:0 (LPC 16:0) are critical mediators of B02/B09-induced mechanical hypersensitivity, since treatment with sPLA2 inhibitor reversed pain-like behavior and bone erosion induced by 1103:01B02/1325:01B09 mAbs. Collectively, these findings provide a novel link between bone erosion and pain, in a state of subclinical inflammation, and progress our knowledge about the mechanisms of bone-related pain in RA.

In Study II we have used a monoclonal anti-citrullinated protein antibody (ACPA) previously reported to stimulate osteoclasts in vitro, but not tested in vivo. We found that intravenous (i.v.) injection of 1325:04C03 IgG into mice induced pain-like behavior in the absence of visible signs of joint inflammation and in a fashion that is insensitive to conventional analgesics like NSAIDs or gabapentin. Instead, 1325:04C03-induced mechanical hypersensitivity was attenuated by nerve growth factor (NGF)-neutralizing antibody and a CXCR1/2 antagonist. Although 1325:04C03 only mildly stimulated osteoclast activity and did not lead to bone erosion in vivo, treatment with osteoclast inhibitor zoledronate partially reversed 1325:04C03-induced mechanical hypersensitivity, indicating a contribution of the bone compartment to pain-like behavior. Furthermore, increased expression of Ngf and neurotrophin 3 (Ntf3) in the ankle joints, as well as increased expression of several pronociceptive factors in the dorsal root ganglia (DRG), was prevented by osteoclast inhibition, pointing towards a relationship between altered bone metabolism and NGF-driven nociception. In conclusion, our data point to a concomitant role of NGF and osteoclast-derived ligands in mediating ACPA-induced pain-like behavior.

In Study III we have focused on the pronociceptive properties of 1325:01B09 mAb, an antimodified protein antibody (AMPA) with reactivity towards several modified peptides. Intravenous injection of 1325:01B09 into mice induced mechanical and thermal hypersensitivity without any visual, histological or transcriptional signs of inflammation in the joint and was not alleviated by NSAID treatment. Instead, we found that 1325:01B09 caused a global increase in expression of several inflammatory-, macrophage-, satellite-glia cells (SGC)- and nociceptor-related factors in lumbar DRGs. Using transgenic mice that lack activating Fc-gamma receptors (FcgRs) we found that FcgRs are critical for the development of intravenous and intra-articular 1325:01B09-induced mechanical hypersensitivity and partially drive transcriptional changes in the DRGs. Finally, we have observed that 1325:01B09 binds SGCs in vitro, and in synergy with LPS, stimulates cells to release CXCL1. Overall, our findings point to the possibility that 1325:01B09 forms immune complexes and induces mechanical hypersensitivity through stimulation of FcgRs. Further studies are warranted to delineate if 1325:01B09 acts also locally in the DRGs by binding to SGC epitopes or FcgRI located on DRG macrophages.

In Study IV we have provided evidence for the pronociceptive properties of IgG isolated from serum of patients with fibromyalgia syndrome (FMS). We have shown that passive transfer of FMS IgG, but not IgG from healthy controls (HC) or IgG-depleted FMS serum into mice, gave rise to pressure, mechanical and thermal hypersensitivity in mice, accompanied by decreased muscle strength and diminished locomotor activity. Moreover, injection of FMS IgG resulted in enhanced nociceptor responsiveness to mechanical and cold stimulation in the skin as well as loss of intraepidermal innervation. While FMS IgG did not directly activate sensory neurons, it bound to SGCs in vitro and in vivo and stimulated their activity measured by immunohistochemistry and gene expression. Furthermore, FMS IgG did not accumulate in mouse spinal cord or brain. Instead, FMS IgG bound to human DRGs and colocalized with both SGCs and sensory neurons. To conclude, these findings demonstrate that passive transfer of IgG from FMS patients into mice recapitulates key features of the disease. While in-depth studies are required to understand the cellular and molecular mechanisms of FMS IgG-induced hypersensitivity, we suggest that therapies that reduce IgG titres or decrease IgG binding might be successful in treating symptoms of FMS.

In summary, this thesis underlines the urgency of in-depth understanding of mechanisms behind autoantibody-induced pain in diseases like rheumatoid arthritis or fibromyalgia syndrome. Studies described here provide evidence that autoreactive antibodies engage several inflammation-independent mechanisms to induce pain and provide novel insights that can aid in designing better pain therapies in the future. Furthermore, only by undertaking translational studies like the ones described above we will be able to understand the complex mechanisms that drive pain in RA or FMS.