A new publication in Genome Biology finds that low bone mineral density (BMD) is a major risk factor for osteoporosis. Genome wide association studies have reported 1,100+ BMD loci, but the vast majority reside in non-coding regions genomes thus making it unclear which effector genes they control. This study first asked: which human cell types best reflect the genetics of BMD ? Using heritability analyses across 98 cell types, we found strong enrichment in the bone-forming osteoblasts; however, interesting we observed no such enrichment for the bone resorbing osteoclasts. We then used a cellular model of human fetal osteoblasts and ran a CRISPR interference screen that selectively “dimmer-switched” 89 candidate regulatory elements informed by our in-house 3D genomic data, analyzing the consequences with single-cell RNA sequencing. This linked 23 genes to those elements; four genes — ARID5B, CC2D1B, EIF4G2, and NCOA3 — showed most consistent effects across experimental settings conducted both at CHOP/Penn and the university of Michigan, marking them as promising drivers of BMD biology. Finally, genetic correlation and fine-mapping suggested many BMD signals may act partly through non-bone tissues, not just bone cells. Overall, the study provides a practical blueprint for turning non-coding genetic hits into key target genes for bone health.
Grant Lab Team Publishes Groundbreaking Research in Genome Biology
