Cindy Boer

174 | Chapter 4.1 For six SNVs (three new and three known), a single variant could be postulated as causal with >95% PP ( Supplementary Table 7 ). The known signals were success- fully fine mapped to the same variants as previously reported (rs4252548, a missense variant in IL11 ; rs13107325, a missense variant in SLC398A ; and rs75621460, an in- tronic variant in TGFB1 ). The new fine-mapped variants were rs72979233 (total knee replacement) residing in an intron of the polymerase (DNA-directed), delta 3, accessory subunit ( POLD3 ) gene; rs74852393 (total hip replacement) situated in the intron of talin 2 ( TLN2 ); and rs201194999 (osteoarthritis at any site) which is located between RP11-807H7.2 and RNU2-40P . We identified four lead missense SNVs directly affecting protein coding sequence that have the highest prior of being causal than the other vari- ants in the credible set (rs28929474 in SERPINA1 , rs3740129 in CHST3 , rs143083812 in SMO and rs2276749 in VGLL4 ( Supplementary Table 7 ). Additionally, 5 lead SNVs were in high LD with coding variants with moderate to severe consequences (R 2 ≥0.8) in the 95% PP set ( Supplementary Table 7 ) ( Table 3 ). The majority of complex disease SNVs is thought to exert their risk by affecting regulation of a target gene in a tissue and cell-specific context[25, 26]. We examined if variants in the fine mapped 95% credible sets ( Supplementary Table 7 ) were en- riched for active gene regulatory histone marks in 127 cell tissues[26]. Osteoarthritis affects multiple tissues within the joint, most notably the cartilage and bone, but there is also evidence for involvement of the synovium, and possibly the muscles and tendons of the joint [27](PMID. We found significant enrichment (p-value<1.3x10 -08 ) of osteo- arthritis-associated SNVs in active gene regulatory histone marks in known osteoar- thritis tissues, such as cartilage (mesenchymal stem cell derived chondrocytes), bone (primary osteoblast) and muscle (muscle, heart) ( Figure 3 ). In line with this we also find enrichment with tissues from the same developmental lineage (mesoderm, mesen- chymal stem cells, adipose-derived mesenchymal stem cells). Interestingly, we also find enrichment in epithelial tissues (i.e., fibroblasts). The epithelial cell type enrichment may be due to the extracellular matrix structure of skin, which like cartilage, muscle and tendons, is also for a large part made out of glycans and collagens [28, 29]. Enrichment in gene regulatory elements was also seen with neurosphere, endocrine (ovary, placen- ta and adrenal gland) and non-endocrine organs (lung, stomach and spleen). Functional genomics in primary disease tissue In order to identify putative effector genes for each of the lead SNVs, we examined if any osteoarthritis signals colocalized with the expression of genes residing within ~1Mb of the osteoarthritis-associated lead variant, using quantitative trait loci (eQTLs) across the 48 tissues in the Genotype-Tissue Expression (GTEx) project [30]. We found colocal-