Dual role of IL6 mediated by mesenchymal stromal cell signaling to joint macrophages in osteoarthritis

Abstract

Background & Aim Osteoarthritis (OA) is a joint disease affecting> 5 million Canadians and patients have limited palliative and joint-replacement surgical options. Stromal cell therapy is emerging as a compelling treatment for OA. Our first-in-Canada Ph1/2 trial with bone marrow mesenchymal stromal cells (BM-MSCs) in OA patients showed significant improvements in patient outcomes. Levels of pro-inflammatory monocytes/macrophages (MΦs) were reduced in the synovial fluid (SF), suggestive of a clinical MSC anti-inflammatory action. Understanding mechanism of action of MSCs in OA will enable effective donor MSC screening, and generation of more efficient engineered MSCs. While IL6 is a known pro-inflammatory mediator in the joint, blocking IL6 has not been so effective in OA suggesting complex roles for IL6. Indeed, MSCs induced homeostatic polarization of macrophages via IL6. This effect is not well studied in an OA joint where high levels of IL6 may have pro-inflammatory roles. We hypothesize that IL-6 mediates anti-inflammatory function of MSCs via IL6 membrane-bound receptor (mIL6R), but not soluble IL6 receptors (sIL6R). Methods, Results & Conclusion We co-cultured blood-derived MΦs with OA-MSCs in the presence of SF from OA patients; or in the presence of IL6 antibody. Polarization of MΦs were assessed by analysis of surface markers. Levels of IL10, IL6 and sIL6R were measured by ELISA. Preliminary analysis of IL6 from MSC from OA patients revealed high IL6 secretion is associated with positive patient-reported outcomes. IL6 antibody abrogated MSCs-induced homeostatic polarization of MΦs, suggesting a key role for IL6 in MSC anti-inflammatory effects. Paradoxically, IL6 levels were increased in SF from late stage OA, whereas sIL6R levels were comparable in early and late OA-SF. Further, the effect of MSCs on MΦs polarization was attenuated in the presence of late OA-SF compared to early OA-SF. This could be due to enhanced IL6 trans-signaling (via sIL6R) in late-OA SF expected to occur at the expense of reduced IL6 classical pathway (via mIL6R). To discriminate between classical vs. trans-signalling, we are using soluble gp130, a non-signaling decoy receptor which complexes IL6-sIL6R without affecting the classical pathway. In conclusion, MSCs exert their anti-inflammatory action in OA, at least partly via polarization of MΦs to homeostatic phenotype mediated by IL6. The ratio of IL6 signaling through trans-vs. classical pathways is an important arbiter of MSC-modulation of joint inflammation.