Abstract Osteoclast formation depends on the cytokine receptor activator of NFB ligand (RANKL), whose actions are inhibited by the decoy receptor osteoprotegerin (OPG). Osteocytes are an important source of RANKL and we have recently shown that osteoblasts, but not osteocytes, are an essential source of the OPG that suppresses resorption of cancellous bone. In contrast, osteoblasts and osteocytes provide only a portion of the OPG protecting cortical bone. Thus, the identity of the cells providing the OPG that protects large regions of cortical bone remains unclear, but may involve osteoblast progenitors and vascular endothelial cells, both of which express OPG. Previous studies suggest that a major function of beta-catenin is to promote OPG expression in osteocytes. Our finding that osteoblasts, but not osteocytes, are a major source of OPG necessitates a reevaluation of the role of beta-catenin in osteocytes. Previous beta-catenin loss-of-function studies have been hampered by the lack of Cre driver strains that can distinguish between osteoblasts and osteocytes and by the dramatic bone loss caused by beta-catenin deletion in late-stage osteoblastic cells. That osteoblasts are an important source of OPG also suggests that the rapid increase in resorption and rebound bone loss following discontinuation of anti-RANKL (denosumab) therapy may be due, in part, to the absence of osteoblasts, and thus OPG, during the period following discontinuation. Based on these findings, we propose the hypotheses that osteoblast progenitors or vascular endothelial cells are important sources of the OPG protecting cortical bone and that the beta-catenin pathway in osteocytes contributes to bone remodeling independent of its control of OPG expression. We also propose that the lack of osteoblasts, and thus OPG, contributes to the rebound resorption following discontinuation of denosumab. Aim 1 will identify cellular sources of OPG that control cortical bone resorption by deleting a conditional OPG allele in mesenchymal progenitors using Prx1-Cre mice, and vascular endothelial cells using Tek-Cre mice, and compare the effects on cortical bone to those seen in OPG- null mice. Aim 2 will determine the role of beta-catenin in osteocytes by deletion of beta-catenin using Sost-Cre mice, which delete target genes in osteocytes but not osteoblasts. Aim 3 will determine whether the rebound resorption caused by discontinuation of denosumab results in part from the profound lack of osteoblasts producing OPG using a novel humanized RANKL mouse line treated with denosumab. These mice will be used to map the cellular and molecular conditions associated with discontinuation of denosumab and to determine whether promotion of osteoblast formation via anti-sclerostin antibody administration can restore OPG levels and ameliorate rebound resorption.