Transcriptomic Profiling of Primary Osteoblast from a panel of Collaborative Cross mice

Genome-wide association studies (GWASs) for bone mineral density (BMD), one of the most significant predictors of osteoporotic fracture, have identified over 1100 independent associations; however, few of the causal genes have been identified. Recently, the “omnigenic” model of the genetic architecture of complex traits proposed two general categories of causal genes, core and peripheral. Core genes play a direct role in regulating traits; thus, their identification is key to revealing critical regulators and potential therapeutic targets. Here, we identified a co-expression module enriched for genes exhibiting properties consistent with core genes for BMD by analyzing GWAS data through the lens of a cell-type and timepoint-specific gene co-expression network for mineralizing osteoblasts. We identified multiple co-expression modules enriched for genes implicated by BMD GWAS and prioritized modules based on their enrichment for genes with core-like properties. Only one module, the purple module, was enriched for genes correlated with in vitro mineralization (r = 0.49; FDR = 0.012), with known roles in skeletal development (P <2.2 x 10-16), that when perturbed produce a bone phenotype in mice (OR = 4.1; P = 2.14 x 10-9), and are monogenic bone disease genes in humans (OR=21.3; P=6.94 x 10-9). Furthermore, the purple module contained genes from two distinct transcriptional profiles with regards to osteoblast differentiation, one of which, termed the late differentiation cluster (LDC), was more highly enriched for genes with core-like properties. Within the LDC, we found that the most highly connected genes were more likely to overlap a BMD GWAS association and associations that contained LDC genes overlapped enhancers and promoters in osteoblasts. Finally, we identified four LDC genes (B4GALNT3, CADM1, DOCK9, and GPR133) with colocalizing expression quantitative trait loci (eQTL) and altered BMD in mouse knockouts. Our network-based approach identified a “core” module for BMD and has provided a resource for expanding our understanding of the genetics of bone mass. Overall design: 42 strains were sequenced with some replicates within strain, and sample 129 was sequenced twice, as a technical control.