Cindy Boer

126 | Chapter 3.1 another study showed that an MGP protein complex is excreted by healthy articular chondrocytes, but not by OA-affected chondrocytes,[34] although we only assessed MGP expression and not MGP protein complex excretion. Although the loci with ASE are known to be enriched for eQTLs, [35] we were unable to detect an association between the MGP genotype and MGP RNA expression levels in cartilage. This could have been due to our modest sample size (knee joint, n=25and or hip joint, n=22) in combination with large heterogeneity of the tissue. No- tably, the available cartilage samples originated from different joint sites (knee and hip) and different disease stage (preserved versus affected) and had large age range of the individuals. Also, it is known that ASE is a more powerful technique than classical eQTL analysis to identify functional SNPs influencing expression of genes.[28] While the ex- tent of imbalance could be considered relatively modest, an increasing number of OA associated SNP alleles appear to mark ASE by comparable amount.[19,36–38] From a more biological perspective, one could consider a prolonged, although slight, deviation from homeostasis due to modest ASE of cartilage relevant genes to be of substantial in- fluence over time. This latter hypothesis could contain the molecular basis for increased risk towards developing OA among the ageing population. Additionally, we observed that the rs1800801 alleles also affected expression of MGP in subchondral bone sam- ples. This could imply that, in parallel to an effect in cartilage, the presumed disturbed cartilage homeostasis is further affected by the underlying bone, further enabling the view that OA is a pathology of the entire joint. Our findings may give an explanation for the known vitamin K association with OA: MGP-mediated calcification inhibition is dependent on γ-carboxylation by vitamin K.[39] It has been shown that low vitamin K intake is correlated with OA.[40] Thus, vitamin K intake may be a potential therapeutic treatment in OA. Recently, a first ran- domised control trial testing the effects of vitamin K on OA was published, which re- ported no overall effect of vitamin K on hand OA.[41] Despite the low power of the trial, there was a significant beneficial effect on joint space narrowing (cartilage degrada- tion) among those individuals that were vitamin K deficient at the start of the trial.[41] Thus, an adequately powered study of vitamin K may be justified based on the found MGP association. Furthermore, genetic predisposition for hand OA was not taken into account in the trial. Perhaps, genetic predisposition for hand OA ( MGP -risk variants) in combination with insufficient vitamin K intake might potentiate cartilage calcification and subsequent risk for developing hand OA. Therefore, future OA trails, therapeutic and preventive treatments might benefit from taking a personalised medicine approach since genetically supported drug targets double the success rate of therapeutics in clin- ical development.[31]