Sophie J. Gilbert ^1* Ben J. Egan ¹ Cleo S. Bonnet ¹ Ryan Jones ¹ Sam Evans ² Deborah Mason ¹

• ¹ School of Biosciences, College of Biomedical and Life Sciences, Cardiff University, Cardiff, United Kingdom
• ² School of Engineering, Cardiff University, Cardiff, United Kingdom

INTRODUCTION: Changes to bone physiology play a central role in the development of osteoarthritis with the mechanosensing osteocyte releasing factors that drive disease progression. This study developed a humanised in vitro model to detect osteocyte responses to either interleukin-6, a driver of degeneration and bone remodelling in animal and human joint injury, or mechanical loading, to mimic osteoarthritis stimuli in joints. METHODS: Human MSC cells (Y201) were differentiated in 3-dimensional type I collagen gels in osteogenic media and osteocyte phenotype assessed by RTqPCR and immunostaining. Gels were subjected to a single pathophysiological load or stimulated with interleukin-6 with unloaded or unstimulated cells as controls. RNA was extracted 1-hour post-load and assessed by RNAseq. Markers of pain, bone remodelling, and inflammation were quantified by RT-qPCR and ELISA.RESULTS: Y201 cells embedded within 3D collagen gels assumed dendritic morphology and expressed mature osteocytes markers. Mechanical loading of the osteocyte model regulated 7564 genes (Padj p<0.05, 3026 down, 4538 up). 93% of the osteocyte transcriptome signature was expressed in the model with 38% of these genes mechanically regulated. Mechanically loaded osteocytes regulated 26% of gene ontology pathways linked to OA pain, 40% reflecting bone remodelling and 27% representing inflammation. Load regulated genes associated with osteopetrosis, osteoporosis and osteoarthritis. 42% of effector genes in a genome-wide association study metaanalysis were mechanically regulated by osteocytes with 10 genes representing potential druggable targets. Interleukin-6 stimulation of osteocytes at concentrations reported in human synovial fluids from patients with OA or following knee injury, regulated similar readouts to mechanical loading including markers of pain, bone remodelling, and inflammation. DISCUSSION: We have developed a reproducible model of human osteocyte like cells that express >90% of the genes in the osteocyte transcriptome signature. Mechanical loading and inflammatory stimulation regulated genes and proteins implicated in osteoarthritis symptoms of pain as well as inflammation and degeneration underlying disease progression. Nearly half of the genes classified as 'effectors' in GWAS were mechanically regulated in this model. This model will be useful in identifying new mechanisms underlying bone and joint pathologies and testing drugs targeting those mechanisms.