Mutations in the mechanosensory gene PIEZO1 are associated with familial finger interphalangeal joint osteoarthritis

      Purpose: Osteoarthritis (OA) is a common debilitating disease of cartilage degradation for which there is no known cure. We lack a comprehensive understanding of the pathways that can be altered to cause OA. To gain further understanding of the OA disease processes, our goal is to find common pathways and genes that are associated with and have a strong effect on OA. To do this we identify rare gene variants that are associated with familial OA. We use the Utah Population Database (UPDB), a large statewide medical genetics resource, to identify families that have a significant enrichment of OA. We have analyzed the exomes of over 50 families with severe or early-onset OA, focusing on foot, hand, and shoulder OA. Identifying variants in OA families allows us to test the biological significance of these genes and their roles in the mechanisms of OA initiation and progression. Using different animal models, we can also investigate which tissues and cells these genes are functioning in to further understand how they contribute to OA pathogenesis. Importantly, data on these potentially OA-causative genes could aid in designing new, targeted therapeutic interventions for OA.
      Methods: The UPDB allows us to link medical records to large multigenerational pedigrees. We first use medical coding to identify individuals diagnosed with OA in Utah. Affected individuals were mapped to pedigrees, and then we identified families that have a significant enrichment of OA compared to the rest of the population. We recruited both affected and unaffected family members and performed whole-exome sequencing to find genes associated with severe or early-onset OA. We focused specifically on variants that lead to nonsynonymous changes that could alter protein function.
      Whole exome sequencing identifies candidate variants that are predicted to be damaging. Our goal is to quickly identify OA gene variants that disrupt WT gene function. We do this in several ways. First, we use zebrafish and Crispr/Cas9 technology to generate mutant lines in which the ortholog of the potential OA-pathogenic gene of interest is knocked out. We then study the effects of gene loss on zebrafish development and morphology. To determine if there is a role for the gene in OA, we study the effects of gene loss using an injury-induced OA model in zebrafish. When we observe an effect in zebrafish, we generate transgenic mice wherein the gene of interest has been mutated to code for disease variant, and compare the effects of ACL rupture on the mutant versus wildtype mice, to determine effects on OA pathogenesis in mouse models.
      In addition, we use cell culture models to study the biochemical pathways affected by the disease allele variants. The variants we have discovered are involved in pathways for which there are known agonists and inhibitors, and we are using these reagents to further understand the contribution of the disease variants to cell signaling and physiology. Using cell culture, we can generate cell lines that express either wildtype or the OA-associated variants gene of interest. We can then use these cell lines to assess how the mutated protein functions compared to wildtype using appropriate in vitro assays such as electrophysiology and pathway reporter assays.
      Results: We found independent PIEZO1 gene variants in 2 unrelated families with finger interphalangeal joint OA. PIEZO1 encodes a mechanosensitive ion channel that is expressed in chondrocytes and may have a role in cartilage homeostasis. We hypothesized that loss of piezo1 may affect early skeletal and cartilage development. To test this, we generated knockout zebrafish by deleting the 5’ UTR and first 5 exons of piezo1 using Crispr/cas9 technology. However, we found no difference in zebrafish homozygous for the piezo1 deletion compared to wildtype at juvenile stages with respect to skeletal morphology or hematopoietic physiology. We are currently testing the effect of loss of piezo1 in injury-induced OA model on adult zebrafish, and determining if genetic compensation is occurring in piezo1 mutants. The OA-associated PIEZO1 variants we identified are located in separate domains of the protein, indicating that alteration to PIEZO1 function could be impactful to OA pathogenesis. Previous reports indicate that mechanotransduction through PIEZO1 plays roles both in bone formation and in sensing of mechanical stress in cartilage. To determine if these coding variants affect the downstream signals emanating from this mechanosensitive channel, we are currently generating plasmid constructs that contain either wildtype or the 2 PIEZO1 variants. These constructs will be transfected into 293 cells, and this cell culture system will enable us to use electrophysiology assays to test what effect the disease variants have on PIEZO1 biochemical signaling.PIEZO1 is highly conserved across species and data indicate that it may function in normal chondrocyte physiology. To determine if PIEZO1 has a function in chondrocytes, we generated PIEZO1-floxed-Aggrecan-cre mice, wherein PIEZO1 will be specifically deleted in cartilage upon induction with tamoxifen. This will allow us to define the function of PIEZO1 in cartilage more clearly. Using these mice, we are testing the function of PIEZO1 in cartilage in both injury-induced OA (using the ACL-rupture mouse model), as well as ageing-induced OA.
      Conclusions: Our discovery of the PIEZO1 variants in 2 distinct families with enrichment in finger interphalangeal joint OA indicates that mechanotransduction thru this channel may be an important pathway for cartilage function and maintenance. By generating specific biological tools, we can test the role of PIEZO1 and define how its aberrant function could contribute to OA initiation and progression. These data will enable us to determine whether targeting PIEZO1 function would be a viable therapeutic intervention for OA patients.