Osteoarthritis of the knee and spine is highly prevalent in modern society, yet a disease-modifying pharmacological treatment remains an unmet clinical need. A major challenge for drug development includes selection of appropriate preclinical models that accurately reflect clinical phenotypes of human disease. The aim of this study was to establish an ex vivo explant model of human knee and spine osteoarthritis that enables assessment of osteochondral tissue responses to inflammation and drug treatment. Equal-sized osteochondral fragments from knee and facet joints (both n = 6) were subjected to explant culture for 7 days in the presence of a toll-like receptor 4 (TLR4) agonist and an inhibitor of transforming growth factor-beta (TGF-β) receptor type I signaling. Markers of inflammation, interleukin-6 (IL-6) and monocyte chemoattractant protein-1 (MCP-1), but not bone metabolism (pro-collagen-I) were significantly increased by treatment with TLR4 agonist. Targeting of TGF-β signaling resulted in a strong reduction of pro-collagen-I and significantly decreased IL-6 levels. MCP-1 secretion was increased, revealing a regulatory feedback mechanism between TGF-β and MCP-1 in joint tissues. These findings demonstrate proof-of-concept and feasibility of explant culture of human osteochondral specimens as a preclinical disease model, which might aid in definition and validation of disease-modifying drug targets.

No abstract

Background:Disease heterogeneity, both clinically and molecularly, has been a major hurdle in the development of efficacious disease-modifying osteoarthritis drugs (DMOADs). Biomechanical, inflammatory, osteoporotic and metabolic OA have been proposed as clinically relevant subtypes for stratification of knee OA patients, yet this remains to be included in clinical trial design. Disease heterogeneity does not only occur within, but also between joint types. However, robust data on joint-specific pathomechanisms of OA are still lacking.Objectives:In this study, we performed ex vivo biomarker profiling of human osteochondral tissue of knee and spine OA to identify joint-specific pathomechanisms and DMOAD treatment responses.Methods:Facet joint and tibial plateaus were obtained from patients undergoing lumbar spinal fusion (n=11, mean age 72.8) and total joint arthroplasty (n=8, mean age 73.0) respectively. Osteochondral specimens were cut in equal-sized samples (100-300 mg wet weight) and randomly assigned to treatment groups: control (DMSO), inflammation (1 μg/mL LPS) or inflammation + DMOAD (TGF-betatype I receptor inhibitor,10 μM SB-505124). Explant culture was conducted for one week and biomarkers of bone metabolism (Pro-Col-Ia, SOST, OPG, SPP1), inflammation (MCP-1, IL-6, MMP3, OSM, TIMP1, VEGFA) and cartilage metabolism (ACAN, COMP) were determined by ELISA. Normalized biomarker secretion was analysed using clusteranalyses and ANOVA. Cartilage proteoglycans were assessed by whole mount Alcian blue staining. Expression of Oncostatin-M (OSM) and its receptors OSMR and LIFR in joint tissues was assessed by RT-PCR and immunohistochemistry.Results:Clusteranalyses revealed that LPS stimulation increased IL-6 and MCP-1 secretion by both facet joint (FJ) and knee joint (KJ) tissues. Interestingly, Oncostatin-M (OSM) and its downstream mediators MMP3 and TIMP1 were increased in the majority of FJ, but not KJ specimens. Statistical analyses corroborated increased OSM, MMP3 and TIMP1 levels in a spine-specific fashion (Figure).Whole mount Alcian blue staining revealed heterogeneous effects of LPS treatment on cartilage proteoglycans, which was negatively correlated with OSM (r=-0.54) and TIMP1 levels (r=-0.45) – yet poorly associated with ACAN (r=0.19). Inhibition of TGF-beta type I receptor signalling in osteochondral tissues led to a drastic reduction of Pro-Collagen-Ia and IL-6 secretion in both spine and knee OA specimens. Interestingly, DMOAD treatment significantly reduced OSM, TIMP1 and MMP3 levels in FJ specimens only. Vice versa, KJ tissues revealed a specific upregulation of monocyte chemoattractant protein-1 (MCP-1) and osteopontin (SPP1) upon inhibition of TGF-beta signalling. OSM was exclusively expressed in subchondral bone marrow macrophages. Isolated chondrocytes and osteoblasts expressed both LIFR and OSMR, yet intact cartilage only showed OSMR expression, while OSMR and LIFR was expressed in marrow tissueConclusion:Oncostatin-M expression and signalling was uncovered as specific pathomechanism of spine OA. DMOAD treatment effects suggested interplay of OSM and TGF-beta signalling pathways in facet joint osteoarthritis. Known to be predominantly expressed by macrophages and immune cells, OSM may be an important osteoimmunological mediator of tissue damage and remodelling in spine, but not knee OA. This study also highlights the value of ex vivo human tissue models for OA phenotyping and preclinical evaluation of DMOADs.Disclosure of Interests:None declared