Antibody activation of sensory neurons : exploring novel pain mechanisms in rheumatoid arthritis

Chronic pain is a worldwide major problem that presents several challenges due to lack of treatment efficacy and/or side effects associated with long-term usage of analgesics. Autoimmune diseases such as rheumatoid arthritis (RA) are often characterized by pain components, which generally poorly respond to drug treatment. In fact, RA patients suffer from persistent pain even if the active disease and inflammation is under medical control or in remission. Moreover, pain appears years before the onset of the active disease. This indicates that RA pain components might underlie additional unknown mechanisms rather than only the classical view of pain strictly correlating with inflammation. Of note, recent studies show that RA autoantibodies are present in RA patients up to 10 years before the onset of inflammation and most of the available treatment options in the clinics do not affect antibody titers. Therefore, the aim of this thesis is to investigate possible autoantibody actions that could represent the missing link explaining pain in RA in the pre- and post-inflammatory phases of the disease.

In study I, we explored the role of RA-relevant autoantibodies in directly activating sensory neurons. Injection of anti-collagen type II (CII) antibodies (Abs) promoted pain-like behavior in mice in the absence of any visual, histological or molecular inflammation. This pain-like behavior was not dependent on complement activation or destabilization of cartilage structure. Instead, our data suggested a direct activation of CII-immune complexes (ICs) on sensory neurons via the activation of Fc gamma receptors (FcγRs). Indeed, we found expression of FcγRI and FcγRIIb proteins on peripheral neuronal terminals in mouse skin. In addition, CII-IC in vitro stimulation of cultured dorsal root ganglia (DRGs) neuronal cells promoted release of a calcitonin gene related peptide (CGRP), intracellular increase of calcium levels and membrane depolarization. Interestingly, CGRP release was prevented in cultures from FcRγ chain deficient mice (lacking activating FcγRI, III and IV, but still expressing inhibitory FcγRIIb). Accordingly, injection of anti-CII Abs failed to induce pain-like behavior in FcRγ chain deficient mice or when the Ab-FcγR interaction was altered. Instead, mice expressing activating FcγRs only on non-hematopoietic cells (including neurons), but not on hematopoietic cells, displayed similar pain thresholds to wild type mice when injected with anti-CII Abs. Altogether our data suggested a novel RA-associated pain mechanism of direct interaction between Abs and FcγRI present on sensory neurons that is independent of inflammatory functions of pathological Abs. Finally, we showed that human DRG neurons also express the activating FcγRIIIA making our data translational to clinics, possibly explaining pain in RA patients before the onset of the disease or even when it is under medical control or in remission.

In study II, we investigated pain-associated pathological actions of human anticitrullinated proteins antibodies (ACPA) purified from RA-patients. Injection of human ACPA, but not non-ACPA or IgGs from healthy individuals, promoted pain-like behavior in mice in the absence of visual, histological and molecular inflammation. Furthermore, ACPA did not induce significant increase of intracellular calcium levels or membrane depolarization in cultured DRG neurons, suggesting that ACPA do not exert their nociceptive functions through a direct action of their Fab region on sensory neurons. However, ACPA bound to osteoclasts, inducing the release of the mouse interleukin-8 analogue CXCL1, which subsequentially sensitized neurons. In fact, a CXCL1 receptor antagonist or an osteoclasts inhibitor prevented ACPAinduced pain-like behavior. In conclusion, we provided evidence of novel nociceptive actions of human ACPA, offering new targets in IL-8 and osteoclasts for the pain treatment of the ACPA-positive subgroup of RA patients.

In study III, we characterized B35, Neuro-2a (N2a) and F11 neuroblastoma cell lines, trying to find an alternative method to primary DRG cultures from rodents for pain-related in vitro experiments. We compared the cell lines subjected to two differentiation media to promote the acquisition of more neuronal-like features on parameters such as morphology, proliferation, metabolic activity, expression of neuronal markers and functional activity. While B35 showed the highest neuronallike morphological features, N2a the highest neuronal markers expression and F11 the highest neuronal excitability in functional assays, all the cell lines compared to primary DRG cultures only to some extent. Therefore, our findings indicated that neuroblastoma cell lines should be carefully selected by researchers for studying neuronal processes, as they do not represent a complete substitute of primary DRG cultures.

In summary, this thesis addresses the crucial need of better understanding the underlying pain mechanisms in RA and provides novel insights that could potentially benefit the clinical therapeutic strategies, opening new avenues for the development of innovative pain-relief drugs.

List of scientific papers

I. Bersellini Farinotti A*, Wigerblad G*, Nascimento D*, Bas DB, Morado Urbina C, Nandakumar KS, Sandor K, Xu B, Abdelmoaty S, Hunt MA, Ängeby Möller K, Baharpoor A, Sinclair J, Jardemark K, Lanner JT, Khmaladze I, Borm LE, Zhang L, Wermeling F, Cragg MS, Lengqvist J, Chabot-Doré AJ, Diatchenko L, Belfer I, Collin M, Kultima K, Heyman B, Jimenez-Andrade JM, Codeluppi S, Holmdahl R**, Svensson CI**. Cartilage-binding antibodies induce pain through immune complex-mediated activation of neurons. Journal of Experimental Medicine. (2019), 216 (8):1904-1924. *,**Contributed equally.
https://doi.org/10.1084/jem.20181657

II. Wigerblad G, Bas DB, Fernandes-Cerqueira C, Krishnamurthy A, Nandakumar KS, Rogoz K, Kato J, Sandor K, Su J, Jimenez-Andrade JM, Finn A, Bersellini Farinotti A, Amara K, Lundberg K, Holmdahl R, Jakobsson PJ, Malmström V, Catrina AI, Klareskog L, Svensson CI. Autoantibodies to citrullinated proteins induce joint pain independent of inflammation via a chemokine-dependent mechanism. Annals of the Rheumatic Diseases. (2016), 75 (4):730-738.
https://doi.org/10.1136/annrheumdis-2015-208094

III. Bersellini Farinotti A, Abdelmoaty S, Kurtovic Z, Krishnan S, Delaney A, Codeluppi S, Emami Khoonsari P, Rogoz K, Kultima K, Svensson CI. Comparing the characteristics of neuronal cell lines with primary DRG neurons in culture and the effect of serum starvation/differentiation. [Manuscript]