Proteomic profiling of osteoarthritis. A computational approach to biomarker discovery

Abstract: Understanding the molecular mechanisms of osteoarthritis (OA) is critical for early diagnosis and effective treatment. OA is a leading cause of disability and poses an increasing burden on healthcare systems, particularly with an aging global population. Despite the potential of proteomics to elucidate the complex biology underlying OA, its application remains limited due to challenges including gaps in computational tools and insights into the early stages of the disease. This thesis addresses these limitations through approaches that combines inventive computational strategies with biological exploration. The thesis presents a comprehensive computational framework that aims to facilitate proteomics analyses. In paper I, we developed ProteoMill, a user-friendly, web-based platform designed to make proteomics data analysis and biological interpretation available to a broader scientific community. In paper II, we introduced, proteasy, a specialized computational tool aimed at identifying proteolytic events. Using this tool, we performed a peptidomic analysis to identify key proteolytic enzymes involved in the degradation of proteins that contribute to OA progression in human synovial fluid (SF). This study presented a broad array of differentially abundant endogenously cleaved peptides, and their potential cleaving actor. We demonstrated that the proteolytic activity of the predicted proteases extends beyond the extracellular matrix (ECM) of the surrounding tissues, and can also affect factors such as chylomicron assembly, potentially leading to hampered homeostasis. In paper III, we established a human meniscus ex vivo model, that enabled us to perform controlled studies on cytokine-mediated effects on meniscal tissues. Our analyses highlight an increase in catabolic processes in response to some of the cytokine treatments while IL1ß had a limited catabolic effect. Finally, in paper IV we utilized the SOMAscan assay, an aptamer-based proteomics platform that is capable of measuring 7,000 proteins. This allowed us to get an unprecedented look into early-stage OA. Gaussian Graphical Models (GGMs) were further utilized to elucidate complex protein interactions, revealing new insights into disrupted joint homeostasis in OA. Through this, we identified novel proteins and sub-networks implicated in the early stages of the disease.