Decreased bioavailable estrogen levels are a major cause of bone loss in postmenopausal women, but sex hormones are important regulators of bone mass in both sexes. Estrogen signaling in bone occurs mainly through estrogen receptors ERα and ERβ. ERα in particular is important in regulating bone mass and bone’s response to mechanical loading, but its particular role in each bone cell type and its cross-talk with BMP signaling are not well studied.

Osteoblast-specific ERα knockout (pOC-ERαKO) and littermate control (LC) were bred by crossing Osteocalcin-Cre and ERα^fl/fl mice. The effects of removing ERα in osteoblasts and osteocytes on bone mass, bone strength, and bone’s response to mechanical loading were studied in 10-week-old animals. In general, cancellous and cortical bone mass were both reduced in pOC-ERαKO female mice, while bone mass was increased in pOC-ERαKO male mice compared to their sex-matched LC, measured by microCT in the proximal and midshaft tibia, femur, and L5 vertebra. These bone mass changes correlated with decreased vertebral compressive strength in female knockout mice and increased femoral bending strength in male knockout mice. After two weeks of in vivo tibial compression, female pOC-ERαKO mice showed a greater increase in bone mass in the proximal tibia, where baseline bone mass was decreased, and at the tibial midshaft, where baseline bone mass was similar to LC. Male pOC-ERαKO mice exhibited a normal response to mechanical loading

Next, 10-week-old female pOC-ERαKO and LC mice were administered either RAP-661, a BMPR1a inhibitor, or placebo, and all mice were subjected to daily in vivo tibial compression for two weeks. RAP-661 markedly increased bone mass in the L5 vertebra and cancellous tibial metaphysis of both genotypes, but not at the femoral midshaft, tibial midshaft, or tibial metaphyseal cortex. In the vertebra, the drug-induced increase in bone mass was less in pOC-ERαKO mice than LC. Animals treated with RAP-661 responded less to mechanical loading in the tibial metaphysis than placebo animals, but similarly at the tibial midshaft. This is the first evidence to indicate that BMPR1a may mediate bone’s response to mechanical loading and interact with ERα in osteoblasts in vivo.