Molecular studies on the autoantibody-mediated bone destruction in rheumatoid arthritis

Abstract

Autoantibody-positive rheumatoid arthritis (RA), also called seropositive RA, is characterised by the presence of anti citrullinated proteins antibodies (ACPA) that can be detected in blood several years before disease onset. Joint inflammation, pain and bone destruction are major features of the disease. Classically bone destruction and pain have been considered to be late events in the disease development, resulting from long lasting and uncontrolled inflammation. However, both bone loss and pain have been reported in both seropositive individuals not yet having the disease and those seropositive individuals just being diagnosed with seropositive RA. Taken together these findings suggest that other factors than uncontrolled joint inflammation might drive the process of bone loss and pain in RA. We hypothesized that antibodies themselves and specifically ACPA might play a direct role in mediating bone loss and pain and aimed to investigate the role of ACPAs in bone metabolism and pain-like behaviour in mice.

To investigate bone metabolism, we focused on studies of osteoclasts (OC), cells responsible for bone loss in vitro (by assessing formation and maturation of OC in cell cultures and estimating their capacity to degrade bone matrix in vitro) and in vivo (by micro-CT analysis of the bone density). We have demonstrated that polyclonal and monoclonal ACPAs isolated from the blood and synovial fluid of RA patients enhance the number of OC and of their bone resorptive capacity in vitro. This effect was accompanied by a significant increase of IL-8 levels in OC supernatants and abolished by neutralizing anti IL-8 antibodies. Further, ACPA injected in mice were shown to bind to CD68-positive OC precursors in bone marrow in the vicinity of the joints and to promote trabecular bone loss, which was also reversed by blocking the mice homologues of IL-8, CXCL1/2 by using reparixin. In parallel to bone destruction ACPA also induced pain-like behaviour in mice, that similar to bone loss was also abolished by CXCL1/2 blocking. Taken together these findings suggest that ACPA promote bone loss by inducing IL-8 that in turn can further amplify the bone loss process and induce pain-behaviour.

As ACPA, but no other immunoglobulins (non-ACPA immunoglobulins from RA patients, non-ACPA Ig from healthy individuals) were able to promote bone loss, we investigated the role of citrullination in ACPA-mediated osteoclastogenesis. We demonstrated that citrullination by peptidyl arginine deiminases (PAD) enzymes is essential for the physiological development and maturation of OCs but no other cells (such as synovial fibroblasts). This finding might explain the ACPA preference for OCs. Further we showed that ACPA bind to targets expressed on the surface of OCs. Blocking the citrullination machinery by PAD enzyme inhibitors significantly abrogates ACPA binding to OCs and ACPA-mediated osteoclastogenesis.

OCs could develop from different cell precursors and inflammatory conditions, such as joint inflammation in RA, promote the transdifferentitation of immature DC (iDC) into OCs. In order to investigate if ACPA might also play a role in this proces, we analyzed the capacity of polyclonal and monoclonal ACPA to promote the in vitro transdifferentation of iDC to OC. We showed that despite a clearly distinct protein profile as compared to classical macrophage OC precursors, iDC are able to develop into remarkably similar OCs. Plasticity towards OC differentiation correlated with PAD activity and protein citrullination expression levels in iDC cultures. Citrullinated actin and vimentin were present in iDCs and iDC-derived OCs and both proteins were deposited on the cell surface, co-localising with ACPAs binding to the cells. ACPAs enhanced OC differentiation from both monocyte-derived iDCs and from circulating CD1c+ DCs. Blocking either PAD activity or ACPA-induced IL-8 secretion completely abolished the stimulatory effects of citrulline-targeting antibodies on DC-OC transdifferentiation. We further explored the mechanisms involved in the plasticity of iDC and their capacity to develop into OCs showing that cell culture densities and lactate concentrations are essential mediators. DCs originating from dense cultures developed in the presence of high lactic acid doses, have high PAD activity and increased efficiency to convert into OC and erode bone. In contrast, DCs from sparse cultures have low PAD activity with decreased OC potential.

In conclusion, the current thesis describes novel mechanisms by which RA-associated antibodies target OC to induce bone loss and pain. Our studies provide insights into the mechanisms by which systemic autoimmunity might target the joints and suggest potential novel ways to prevent this.