Osteoporosis—a disease of low bone mass and a predisposition to fracture—remains a pressing public health burden. Osteoporotic fractures occur in 1 in 3 women and 1 in 5 men over the age of 50, at a cost of more than $22 billion annually in the US. To prevent osteoporotic fractures, we need additional strategies to increase bone formation and reverse the disease. A potent strategy to increase bone formation is mechanical loading. However, with aging, bone becomes less responsive to the same stimulus. Because we do not fully understand the agerelated decrease in bone mechanoresponsiveness, we have been unable to leverage the involved pathways for maximal clinical benefit. My overall goal was to better understand the aged loading response and to test rescue strategies to restore the mechanoresponsiveness of the aged skeleton, which could lead to novel strategies to treat osteoporosis.

In aim one, we used systems biology approaches to more fully understand the aged loading response. Specifically, we compared the loading responses of young-adult (5-month) and old (22-month) mice using (1) transcriptomics and (2) proteomics to better define the differences that drive the diminished bone formation. Transcriptomics and proteomics—the large-scale study of RNA messages (transcripts) and proteins, respectively—are two gold standards for profiling tissues in vivo, but these techniques had not been employed to fully explore the aged loading response. At the transcript level compared to young-adult mice, old mice had less activation following loading at each time point, as measured by the number of differentially expressed genes (DEGs) and the fold-changes of the DEGs. Old mice engaged fewer pathways and gene ontology (GO) processes, showing less activation of processes related to proliferation and differentiation. In bones of young-adult mice, we observed prominent Wnt signaling, extracellular matrix (ECM), and neuronal responses, which were diminished with aging. Additionally, we identified several targets that may be effective in restoring the mechanoresponsiveness of aged bone, including nerve growth factor (NGF), Notum, prostaglandin signaling, Nell-1, and the AP-1 family. To facilitate the proteomics experiments, we successfully developed the first method to profile the proteome from mouse cortical bone. We performed a fundamental comparison of proteomics and transcriptomics on paired limbs from the same animal and provided the first estimate for the correlation between the proteome and transcriptome in bone, which was in line with other tissues. At the protein level compared to young-adult mice at baseline, old mice displayed reduced TGF-beta signaling and Wnt signaling. With loading, proteomics revealed only modest changes, perhaps due to the lower protein turnover of bone.

In aim two, we set out to use our increased understanding of the aged loading response to test rescue strategies for restoring the mechanoresponsiveness in aged bone. While it was critical to define how the loading response changed with aging, the translational impact of these discoveries will be realized only if they inform ways to rescue the loading response in old mice or lead to new osteoporosis treatment candidates. Specifically, we tested the extent to which (1) exogenous NGF delivery to enhance Wnt signaling, (2) EP-1 antagonism to modulate prostaglandin signaling, and (3) inhibition of the senescence-associated secretory phenotype (SASP) with a Jak inhibitor could restore the loading response in old mice. Contrary to our hypotheses, none of these rescue approaches enhanced the loading response in aged mice, underscoring the complexity of restoring this failure of biology with aging. More recently, this aim has led to ongoing studies to restore Wnt signaling more directly with a Wnt agonist. These strategies show promise for increasing skeletal mass at baseline and will be tested in the context of loading in aged mice in the future.

In aim three, we explored glucose metabolism as a potential link between the diminished Wnt signaling activation and diminished bone formation in aged mice following loading. Wnt signaling is known to regulate cellular metabolism in bone-forming osteoblasts, and changes in cellular metabolism, specifically toward aerobic glycolysis, may facilitate osteoblastic differentiation. Given that both Wnt signaling and the bone formation response are reduced with aging, we hypothesized that alterations in osteoblast glucose metabolism mediated this association. We specifically used (1) a pharmacological inhibitor of aerobic glycolysis (dichloroacetate) and (2) a genetic loss-of-function of the primary glucose transporter in the osteoblast lineage (Glut1) to investigate if impaired glucose metabolism phenocopied the aged loading response. These studies did not reveal strong effects of impaired glucose metabolism on the bone formation response to mechanical loading but cannot definitively rule out the role of glucose metabolism in these processes.

Altogether, this work advances our understanding of how aged bone responds to mechanical loading at both the transcript and protein levels. It experimentally tested a number of targets to show that they did not effectively restore the loading response in aged mice. It also showed that impairing glucose metabolism does not phenocopy the aged loading response. This work has laid the foundation for new studies of bone tissue using proteomics and has suggested additional targets for restoring the aged loading response that have yet to be tested—including Wnt signaling directly.