HUMAN JOINT-ON-A-CHIP: MODEL DEVELOPMENT AND APPLICATION TO STUDY PATIENT-SPECIFIC CROSSTALK

Synovial inflammation is accepted as a contributing factor to osteoarthritis (OA) in at least a subset of patients. The interaction between the cartilage and synovium is of particular interest, but is yet to be fully elucidated. Here, we report on the development of a vascularised synovium-on-a-chip to study synovial inflammation and immune cell recruitment. We also report on the development of a cartilage-synovium-on-a-chip, to study patient specific crosstalk. Human fibroblast-like synoviocytes (hFLS) and human articular chondrocytes (hAC) were commercially-sourced, or freshly isolated from tissue collected from OA patients, as required per configuration. Preliminary, in vitro experiments revealed that conditioned media from hFLS influenced the inflammatory response and expression of matrix turnover genes, in hAC, and revealed between-donor variation in these responses. Subsequently, two organ-chip models were developed using the Chip-S1 (Emulate Inc.; Figure 1). The vascularised synovium-on-a-chip consisted of an hFLS monolayer in the top channel, and a human umbilical vein endothelial cell (HUVEC)-seeded bottom channel, mimicking vasculature. The cartilage-synovium-on-a-chip similarly consisted of an hFLS monolayer in the top channel, and the bottom channel was seeded with hAC suspended in hydrogel. Both chip configurations were established and cultured for a maturation period (6 days and 10 days respectively) before uniaxial cyclic tensile strain was applied (2 h at 12% strain, 0.2 Hz) for 4 days. In both chip configurations, hFLS exhibited characteristic morphology, cytoskeletal architecture and matrix protein deposition. In the vascularised synovium chip, HUVECs grew on all inner surfaces of the bottom channel forming a tubule-like structure. Synovial inflammation was initiated through the addition of IL-1 β into the synovium channel resulting in the increased secretion of inflammatory and catabolic mediators. Enhanced expression of the inflammatory marker, ICAM-1, was observed in HUVECs, accompanied by increased attachment of circulating monocytes. In the cartilage-synovium chip, hAC were well-distributed throughout the bottom channel and demonstrated a characteristic rounded morphology with cortical actin. The expression of hyaluronic acid synthase-1, 2 and 3 genes were increased in the hFLS layer following the application of strain. Our findings highlight the importance of synovial inflammation, cartilage-synovium crosstalk and donor-specific responses in unpicking OA pathogenesis, and in establishing stratified and personalised therapies. We have developed two organ-on-a-chip models that are currently being utilised to study patient-specific interactions between the cartilage and synovium to unpick pathological crosstalk in OA.