Background Many osteoarthritis (OA) patients show synovial activation, which is thought to be involved in joint destruction. Previously, we found increased expression of both the alarmins S100A8/A9 and members of the Wnt signaling pathway in the joints during experimental OA. Active roles in the development of the OA pathology are attributed to both groups of proteins.
Objectives To determine whether an interrelationship between S100A8/A9 and canonical Wnt signaling exists.
Methods Gene expression of S100A8/A9 and Wnt members was measured in the destabilization of the medial meniscus (DMM) and collagenase-induced OA (CIOA) experimental OA models. Selected Wnts were overexpressed and expression levels of S100A8/A9 were measured. Activation of canonical Wnt signaling was determined after intra-articular injection of S100A8 into naïve mouse knee joints and after induction of CIOA in S100A9-deficient mice. Expression of Wnts was tested in macrophages and fibroblasts after stimulation with S100A8. To determine if the effects of S100A8 injections run via activation of Wnt signaling, canonical Wnt signaling was inhibited in vivo by feeding inducible DKK-1 Tg mice a diet supplemented with doxycycline to induce overexpression of the DKK-1 transgene.
Results qPCR analysis showed increased expression of the alarmins S100A8/A9 and several members of the canonical Wnt signaling pathway in the CIOA model at all time points measured. In the DMM model, the expression of S100A8/A9 and Wnt16 and WISP1 was mainly increased at day 28 after induction. Kinetics of S100 and Wnt expression were comparable, as was observed in both models. This gave rise to the question if an interrelationship existed between these factors. Therefore, we overexpressed Wnt8a and Wnt16, two canonical Wnts, with the use of adenoviral vectors. However, this did not result in increased expression of S100A8 and S100A9, both on RNA and protein level. In contrast, we found that injection of S100A8 increased the expression of Wnt16 in the synovium and accumulation of β-catenin, a hallmark of canonical Wnt signaling, in both cartilage and synovium. In addition, the downstream mediator of canonical Wnt signaling WISP1, was increased. Furthermore, we found reduced β-catenin accumulation in the cartilage and synovium after induction of the CIOA model in S100A9 deficient mice. However, differences in inflammation as the result of divergent S100 levels could possibly explain the differences in canonical Wnt signaling. Therefore, we stimulated both murine and human macrophages and fibroblasts with S100A8 in vitro and found increased expression of Wnt16 and WISP1 in macrophages, but not in fibroblasts. Finally, we determined if activation of canonical Wnt signaling was required for the effects of S100A8. Therefore, we injected S100A8 into knee joints of DKK-1 Tg mice that were fed a standard diet or a diet supplemented with doxycycline, which led to overexpression of DKK-1. Injection of S100A8 significantly induced the expression of MMP3, IL-6, MIP1α and KC. However, inhibition of canonical Wnt signaling significantly reduced the upregulation of MMP3, IL-6 and MIP1α but not KC.
Conclusions S100 proteins increase Wnt signaling. In addition, part of the S100 effects in the synovium is dependent on activation of Wnt signaling.
Disclosure of Interest None declared