Bone mineral density (BMD) is a heritable risk factor for osteoporosis and bone fracture, both of which constitute common and serious health concerns in the elderly. A new study published in Nature Genetics uses a genome-wide association study (GWAS) for estimated BMD (eBMD) to provide an atlas of associated genetic loci and lays the foundation for further analyses of a set of targetable genes for prevention of fracture.
The researchers utilized the UK Biobank cohort, which provides strong statistical power for association testing thanks to the large sample size; eBMD in this cohort was measured by quantitative ultrasound of the heel bone.
The first GWAS in >400,000 individuals identified 515 loci (301 of which were novel) associated with eBMD. Although the individual effects on eBMD are small, together these loci explain ~20% of the variation in bone density in the general population.
In a second GWAS, the researchers analysed fracture risk itself and identified 13 loci in UK Biobank participants, classified into cases (n = 53,184) and controls (n = 373,611) using hospital records. All 13 loci were replicated in a cohort of research participants from the personal genetics company 23andMe (n = 731,819) and all 13 loci were also associated with eBMD with genome-wide significance in the initial GWAS.
To shorten the list to potential causal genes, the authors applied a number of statistical methods to obtain a set of fine-mapped single-nucleotide polymorphisms (SNPs). These SNPs were overlapped with ATAC-sequencing data on chromatin accessibility in an osteosarcoma cell line and ENCODE tracks of DNase I hypersensitivity sites in human primary osteoblasts. Then, in conjunction with knowledge of genomic features of genes with known function in bone biology, a set of target genes was generated for functional analysis.
A high-throughput screen of knockout mice from the Origins of Bone and Cartilage Disease study showed that loss of any given selected target gene (n = 126) more often resulted in an abnormal skeletal phenotype than when a gene from an unselected set of genes (n = 526) was knocked out.
Taking the combined evidence from the eBMD GWAS, fracture-risk GWAS, statistical fine-mapping and mouse skeletal phenotype into account, DAAM2 (which encodes a protein involved in Wnt/β-catenin signalling) was selected as an example target gene for validation.
“Daam2-knockout mice had high cortical porosity and reduced bone strength”
The researchers showed that CRISPR–Cas9-mediated disruption of DAAM2 reduced mineralization in human osteoblast cells in vitro. Furthermore, Daam2-knockout mice had high cortical porosity and reduced bone strength, validating this gene as a potential target for prevention or treatment of osteoporosis.
“We expect that this work will be used to highlight further targets for drug development,” says Brent Richards, one of the corresponding authors “Furthermore, this work might help us identify people with genetically increased risk of osteoporosis”.
Morris, J. A. et al. An atlas of genetic influences on osteoporosis in humans and mice. Nat. Genet. https://doi.org/10.1038/s41588-018-0302-x (2018)
About this article
Cite this article
Trenkmann, M. GWAS cracks fracture risk. Nat Rev Rheumatol 15, 126 (2019). https://doi.org/10.1038/s41584-019-0173-2