Molecular genetic and epigenetic analysis of osteoarthritis

Abstract

Osteoarthritis (OA) is a common degenerative disease of synovial joints that principally affects older individuals. The genetic architecture of OA is highly complex, with heterogeneous pathological pathways converging on a common end phenotype characterised by the painful loss of articular cartilage. In this study, I investigated both genetic and epigenetic aspects of the OA disease process. The genetic investigation focused on the functional analysis of a female hip OA association signal that was discovered in the arcOGEN study, marked by the single nucleotide polymorphism (SNP) rs4836732. The signal surpasses the genome-wide significance threshold, with p = 6.11x10-10. The signal encompasses only two other SNPs that have a high degree of linkage disequilibrium (LD) with rs4836732. Functional studies using luciferase reporter assays, electrophoretic mobility shift assays (EMSAs) and a range of transformed cell lines did not identify any differential allelic activity between SNP alleles, but did identify differential EMSA banding patterns for the C allele of rs4836732 when compared to the T allele, suggestive of differential protein complex binding dependent on the allele present at rs4836732. Subsequent investigations focused on three genes in the immediate vicinity of the association signal: ASTN2, PAPPA and TRIM32. All three genes were expressed in cartilage, synovium and fat pad from OA patients, though none displayed differential allelic expression correlating with rs4836732 genotype in these tissues. This lack of positive evidence in end-stage disease tissue may be indicative of the association signal mediating risk during joint development or growth. Of the three genes, PAPPA was deemed the most likely candidate to have a functional affect in joint development. Knock-down of the gene in mesenchymal stem cells revealed that PAPPA is required for osteogenic and chondrogenic differentiation, providing support ii for further investigation of this gene. My second line of investigation focused on OA epigenetics, specifically on subgroups of hip and knee OA patients who displayed altered cartilage methylation of inflammatory genes when analysed by high-density CpG methylation array. As part of this study, I developed pyrosequencing assays to determine the inflammatory subgroup of new patient cartilage samples, negating the need for a high-density array. I also further interrogated the inflammatory status of the patients, analysing serum markers of systemic inflammation and gathering biometric data to ascertain the primary cause of inflammation, with a focus on the metabolic status of patients in relation to body fat. Interestingly, no systemic differences in inflammation, circulating leptin or BMI were observed. This suggests the elevated inflammatory status may be local to the synovial joint, as opposed to being secondary to increased systemic inflammation. Overall, the two investigation routes that I have pursued further emphasise the heterogeneous nature of OA. They also highlight the importance of patient stratification for OA research and provide platforms to improve our comprehension of this debilitating disease.