OMD is a small proteoglycan involved in bone and dental matrix mineralization but also in ectopic mineralization of other tissues, such as in arteries 32–36, suggesting it could be involved in cartilage mineralization and degradation during aging and OA.
We have previously demonstrated that osteoblasts located in the sclerotic area of OA subchondral bone produced less OMD than neighboring osteoblasts coming from non-sclerotic area 8. Interestingly, OMD levels were also lower in the serum of OA patients. To study the impact of Omd expression on bone remodeling, skeletal development and architecture, we followed mice deficient for Omd and mice overexpressing Omd during 16 months.
While the presence of OMD in bone has been previously reported 31,37, we showed for the first time that OMD is localized in mineralized tissues of the murine knee joints and is identified in calcified cartilage and tidemark. Interestingly, we observed that the calcified cartilage layer was thinner in the medial tibial compartment but thicker in the lateral tibial compartment of KO mice than in other genotypes indicating that OMD plays a key role in cartilage mineralization. The consequences of calcified cartilage thickness on cartilage degradation in OA remain controversial. One study showed that the calcified cartilage was thinning with OA, resulting in the reduction of the cartilage elastic modulus 38. However, other studies showed that the calcified cartilage thickness increased with the progression of OA 39,40. The presence of the more severe cartilage lesions in the medial tibial plateau of aging KO mice, where the calcified cartilage was thinner, supports the hypothesis that a thinner layer of cartilage calcified is a factor promoting cartilage degradation. Of course, this theory needs to be confirmed in other models. In the DMM OA model, there was no significant difference in cartilage damage between genotypes. This finding contrasted with the observation performed in the aging KO mice in which the cartilage lesions severity were higher in KO mice than in other genotype. This observation can be explained by the higher severity of the lesions in DMM model reflecting more of a late stage of OA. We can anticipate a floor effect in DMM-induced OA model because the cartilage lesions were too severe.
At the bone level, 8 and 16-month-old KO mice had more trabecular and cortical BV/TV than the WT while, inversely, UP mice had a reduced ratio. This finding highlights that Omd plays a key role in bone remodeling. More precisely, keeping the homeostatic expression of Omd helps preserve its volume and structure. Omd overexpression not only reduced BV/TV but also increased the structure model index, which is an indicator of the altered shape of trabeculae, and higher bone porosity. Over time, aging was aggravating those observations. This indicates that, when overexpressed, Omd may turn to cause detrimental effects on skeletal tissues. In the KO, the global bone morphology was affected. Their tibia was narrower and their tibial crest longer, this morphological change may affect muscle insertion and, by so, the muscle to bone relationship. Further, KO mice were more prone to spontaneously develop subchondral bone sclerosis, as indicated by higher BV/TV, like in sclerotic subchondral bone in OA.
We also observed sclerosis of the subchondral bone following the DMM procedure, in all the genotypes. Yet, the subchondral bone of the medial tibia of UP mice was thinner than the KO and WT mice suggesting that Omd could prevent subchondral bone sclerosis in OA. These data suggest that Omd is deeply involved in subchondral bone sclerosis in OA, a key feature of OA involved in cartilage degradation. Therefore, we can hypothesize that the impact of Omd on cartilage degradation could be secondary to its effect on bone..
The gait analysis with the Catwalk XT identified different motor patterns between the genotypes. The gait pattern of KO mice, including the print area, the swing, and the intensity of the contact of paws toward the glass platform, were modified. More precisely, for their hind paws, KO mice had a lower print area, a shorter swing, and a higher intensity of the contact of the paw. This may result in pain, discomfort, or of mechanical disorders associated with joint damage or with skeletal tissue abnormalities. It is important to highlight that a decreased hind print area is considered the best predictor for spontaneous OA 41. On the other hand, the different gait patterns of the KO could be the result of OA development leading to abnormal loading on affected limbs.
To corroborate our findings from the mice model, we studied mutant adult zebrafish which did not express omd. In zebrafish, we found cartilage lesions in the articular cartilage in the jaw joint. Again, this suggests that omd prevents spntaneous cartilage lesions during aging and that a decrease of omd production by osteoblasts and hypertrophic chondrocytes could be deleterious for cartilage. Altogether, these findings support that a loss of OMD contributes to OA development. We have then investigated by which mechanism of action OMD could regulate bone and cartilage metabolism. Our transcriptomic data revealed that OMD is unlikely to perform its function on osteoblasts through direct gene expression regulation as very few genes were modified and with a low magnitude. It remains noteworthy to specify that among the regulated genes, IBSP was downregulated by OMD. IBSP overexpression by hypertrophic chondrocytes is associated with OA 42. Consequently, OMD could control cartilage calcification in OA by downregulating IBSP production.
Direct binding of OMD to key bone regulatory factors is another possible mechanism of action. SLRPs are known to bind cytokines, growth factors, and ligands like RANKL 9,24. Herein, we showed that OMD is not only enhancing the differentiation of osteoblasts 31, but is also able to bind directly to RANKL and block its biological activity on osteoclasts. The mutant zebrafish model confirmed the role of omd in osteoclastogenesis. The number of cathepsin K positive osteoclasts increased in the regenerating caudal fin of the mutants. Furthermore, observations of elasmoid scales, which share similar transcriptomic profile with the mammalian skeleton including genes related to human diseases 43, highlighted higher TRAP staining and more circular scales in zebrafish lacking omd. As osteoclasts are inducing OMD expression in mature osteoblasts 44, OMD presents a negative feedback activity on them. Furthermore, sulfated GAGs are known to inhibit the differentiation of osteoclasts and the sulfation level of OMD is higher during the ECM mineralization process 45,46. Those observations are crucial since we know that bone remodeling plays a key role in the bone-driven OA phenotype and that an elevated number of osteoclasts are found in sclerotic subchondral bone 47. Our findings present OMD as a novel regulator of the bone remodeling process able to prevent subchondral bone sclerosis in pathological conditions like OA.
In conclusion, alterations of the OMD expression modify the bone and cartilage metabolism and structure. OMD helps to preserve bone and cartilage integrity and a local decrease of its production leads to the development of OA mainly by increasing subchondral bone sclerosis and thinning the calcified cartilage while its overexpression alleviate the subchondral bone sclerosis. OMD is able to directly bind to RANKL and inhibit osteoclastogenesis to regulate bone remodeling and limit subchondral bone sclerosis. Our previous and current researches, making use of both in vitro and in vivo experiments either with human, mouse or zebrafish models, build a strong and compelling body of evidence of OMD being a key factor in OA associated with subchondral bone sclerosis.