Abstract by Stefania Kalogera

Background: Osteoarthritis (OA) is the most common form of arthritis and a significant cause of disability and pain. Pain is the major debilitating symptom of OA, and the underlying mechanisms that cause OA pain are still at the forefront of medical research. OA is characterized by continuous extracellular matrix (ECM) remodeling of the joint which leads to the release of neo-epitope fragments into the circulation. These epitopes can be used to describe structural changes in OA and potentially pain as they may reflect cartilage, synovium, bone turnover and inflammation. However, OA pain is highly heterogeneous and involves multiple components. The role of proteinase-activated receptor 2 (PAR2) has been identified as a contributor to pathophysiology and pain sensitization in OA. A clear need for more objective measurements of pain is needed to improve the success of therapeutic strategies. Biomarkers of joint tissue turnover as well as markers that reflect inflammatory pathways, such as the activation of PAR2, could be useful tools for pain characterization in OA.

Hypothesis and aims: The main hypothesis of the thesis is that biomarkers of joint tissue remodeling and inflammation can be used to describe structural changes and profile pain in OA patients. The specific aims were 1) to develop and validate a quantitative assay for serological determination of the pro-fragment of PAR2 2) to investigate the relationship between PAR2 activation and synovial ECM remodeling, 3) to investigate the associations between PRO-PAR2 levels and pain outcomes in serum of OA patients, 4) and finally to investigate the associations between PRO-PAR2 and already established biomarkers of ECM remodeling, and clinical features in OA both systemically and locally.

Methods: An immunoassay targeting the PAR2 pro-fragment that is cleaved and released when the PAR2 receptor is activated, was developed and technical parameters were evaluated. PRO-PAR2 levels were tested after in vitro cleavage of the full-length PAR2 protein with matriptase. PRO-PAR2 release was investigated in supernatants from OA synovial explants treated with matriptase. To understand the relationship between PAR2 activation and ECM remodeling matrix metalloproteinase (MMP)-generated collagen type I (C1M) and type III (C3M) degradation fragments’ levels were quantified in the same supernatants. To assess clinical relevance PRO-PAR2 levels were quantified in the serum of 22 healthy individuals, 15 rheumatoid arthritis (RA) and 23 OA patients, and a subset of RA patient treated with tocilizumab. Furthermore, PRO-PAR2 was measured in the Innovative Medicine Initiative – Applied Public-Private Research enabling Osteoarthritis Clinical Headway (IMI-APPROACH) cohort of 297 patients with knee OA, and associations with pain outcomes were investigated with linear regression models. Lastly, PRO-PAR2 and biomarkers of ECM turnover were evaluated in serum and synovial fluid (SF) of 31 OA patients, and their association with Kellgren – Lawrence grade (KL grade), osteophyte area, subchondral bone density, minimum joint space width (minJSW) and pain outcomes from multiple questionnaires were investigated. Biomarkers included degradation fragments of collagen type I (C1M), II (C2M), III (C3M), X (C10C), and aggrecan (ARGS).

Results: PRO-PAR2 was found to be specific for the neo-epitope but not to elongated or truncated forms of the peptide. The assay was technically robust. Neo-epitope specificity was further, confirmed by the release of PRO-PAR2 from full-length PAR2 cleaved in vitro with matriptase. PRO-PAR2 levels were increased in OA synovial explants treated with matriptase alone and together with oncostatin M (OSM)+ tumor necrosis factor-alpha (TNF-α). Matriptase synergistically with oncostatin M (OSM) and TNF-α induced release of collagen type I (C1M) and type III (C3M) degradation fragments in the treated explants compared to controls. In RA patients, PRO-PAR2 levels were elevated compared to healthy controls, and levels were reduced by treatment with tocilizumab. In the IMI-APPROACH cohort, PRO-PAR2 levels were negatively associated with pain outcomes in patients with high inflammation (high MMP-generated fragment of CRP; CRPM). In a subset of 31 OA individuals who had matching SF, there was a correlation of C2M and C3M between serum and SF, while no correlation was observed for C1M, C10C, and ARGS. PRO-PAR2 levels were below the detection limit of the assay in SF. In the serum of the 31 OA individuals from the IMI-APPROACH cohort, C1M was found to be negatively associated with subchondral bone density, while C2M was positively associated with minimum joint space width. C10C levels in SF were negatively associated with minJSW and positively associated with KL grade and osteophyte area. Lastly, C3M in serum was negatively associated with constant pain assessed by the intermittent and constant osteoarthritis pain (ICOAP) questionnaire. None of the other biomarkers showed any association with structural or pain outcomes.

Conclusion: PRO-PAR2 was increased in RA patients and reduced by treatment with anti-IL-6 receptor treatment indicating a pharmacodynamic marker in RA. In OA, PRO-PAR2 was negatively associated with pain outcomes in a subset of patients with high CRPM levels suggesting the presence of other pain mediators or even a protective role of PAR2 in this subgroup of patients. No associations were observed in the full population and in the patients with low inflammation. Some association between ECM turnover markers and structural and pain outcomes in OA serum and SF was observed; however, this was dependent on the individual biomarker. While associations between biomarkers and pain were observed in this PhD thesis, further studies are needed to validate these early exploratory findings.