Genome-wide association study of extreme high bone mass: Contribution of common genetic variation to extreme BMD phenotypes and potential novel BMD-associated genes

Background: Generalised high bone mass (HBM), associated with features of a mild skeletal dysplasia, has a prevalence of 0.18% in a UK DXA-scanned adult population We hypothesized that the genetic component of extreme HBM includes contributions from common variants of small effect and rarer variants of large effect, both enriched in an extreme phenotype cohort. Methods: We performed a genome-wide association study (GWAS) of adults with either extreme high or low BMD Adults included individuals with unexplained extreme HBM (n = 240) from the UK with BMD Z-scores >=+ 3.2, high BMD females from the Anglo-Australasian Osteoporosis Genetics Consortium (AOGC) (n = 1055) with Z-scores +1.5 to +4.0 and low BMD females also part of AOGC (n = 900), with Z scores -1.5 to - 4.0. Following imputation, we tested association between 6,379,332 SNPs and total hip and lumbar spine BMD Z-scores. For potential target genes, we assessed expression in human osteoblasts and murine osteocytes. Results: We observed significant enrichment for associations with established BMD associated loci, particularly those known to regulate endochondral ossification and Wnt signalling, suggesting that part of the genetic contribution to unexplained HBM is polygenic Further, we identified associations exceeding genome-wide significance between BMD and four loci: two established BMD-associated loci (5ql4.3 containing MEF2C and lp36.12 containing WNT4) and two novel loci: 5pl3.3 containing NPR3 (rs9292469; minor allele frequency [MAF] = 0.33%) associated with lumbar spine BMD and llpl5.2 containing SPON1 (rs2697825, MAF = 0.17%) associated with total hip BMD. Mouse models with mutations m either Npr3 or Spon1 have been reported, both have altered skeletal phenotypes, providing in vivo validation that these genes are physiologically important in bone. NRP3 regulates endochondral ossification and skeletal growth, whilst SPON1 modulates TGF-beta regulated BMP-driven osteoblast differentiation Rs9292469 (downstream of NPR3) also showed some evidence for as sociation with forearm BMD in the independent GEFOS sample (n = 32,965) We found Sponl was highly expressed in murine osteocytes from the tibiae, femora, humeri and calvana, whereas Npr3 expression was more variable. Conclusion: We report the most extreme-truncate GWAS of BMD performed to date. Our findings, suggest potentially new anabolic bone regulatory pathways that warrant further study.