Impact of Osteoblastic Hypoxia Signaling in Bone and Beyond: Skeletal regulation of systemic glucose homeostasis

Although the skeleton is equipped with a well-organized and dens vascular network, gradients of oxygenation exist due to its specific architecture and high cellularity in the marrow. To face these challenging conditions, cells within the bone environment are well-adapted and respond to low oxygen by activating a specialized transcriptional program, mediated by hypoxia-inducible factors (HIFs), to reduce their need for oxygen on the one hand and stimulate the supply of oxygen on the other. In the bone field, the hypoxia signaling pathway has been extensively studied and found to play a central role in bone formation and skeletal homeostasis. In fact, HIF and its downstream target vascular endothelial growth factor (VEGF) represent the first recognized couplers of osteogenesis and angiogenesis and fulfil a bone-anabolic role. Additionally, recent work for the first time allocated a role for HIF-induced upregulation of glycolysis in the development of a high bone mass phenotype in genetically modified mice. In addition to its classically recognized functions, the skeleton has also become increasingly recognized as an important regulator of whole-body energy metabolism and glucose homeostasis, with osteocalcin and insulin representing the prime known endocrine mediators of this interplay. Genetic evidence has indicated that osteoblasts may additionally influence global energy metabolism through mechanisms independent from these known effectors, although their nature remains elusive. In this thesis, we studied the effects of the hypoxia signaling pathway on bone formation and homeostasis at the tissue level, but also on the cellular level by investigating osteoblast biology and cellular energy metabolism. We also addressed the implications of altered hypoxia signaling in bone on processes extending beyond skeletal tissues. For these studies, we generated a model of increased HIF signaling in osteolineage cells, by osteoprogenitor-specific deletion of the Von Hippel Lindau (Vhl) gene, a negative regulator of the hypoxia pathway, using the Osx-Cre:GFP driver strain. First, we investigated the bone phenotype of these Vhl conditional knockout (cKO) mice in-depth, as described in Chapter 3 of the thesis. We found that mice in which the hypoxia signaling pathway was constitutively activated in osteoblast lineage cells showed a marked increase in bone mass with hypervascularization and disruption of the bone marrow environment. However, despite the high bone mass, adult Vhl cKO mice showed greatly decreased bone formation and mineralization rates. This was associated with impaired terminal differentiation of osteoblasts and an expansion of the pool of relatively immature osteolineage cells, which were likely responsible for the deposition of the rather disorganized, woven bone matrix. Despite normal osteoclast numbers, the presence of cartilaginous remnants extending into the diaphysis and reduced serum levels of collagen degradation products relative to bone volume were suggestive of a reduced bone turnover status in Vhl cKO mice. In the next Chapter 4, we aimed to shed light on the cell-intrinsic effects of increased HIF signaling in osteolineage cells, with a special focus on cellular energy metabolism. Both in vitro and in vivo, Vhl-deficient osteoblasts showed increased glucose consumption and glycolysis, associated with upregulated glucose transporters and glycolytic enzymes, and a reduced oxygen consumption rate. This effects resembles the Warburg effect that typifies cancer cells. The pharmacologic agent dichloroacetate (DCA), a glycolysis inhibitor, corrected the bio-energetic switch towards glycolysis and increased glucose consumption of Vhl-deficient cells in vitro. In search of the mechanisms causing the high bone mass phenotype in the Vhl mutant mice, we next generated a postnatally (PN)-induced Vhl cKO model and treated these mice with DCA in vivo. Our results showed that the high bone density observed in PN-Vhl cKO mice was not reverted upon DCA administration, suggesting that the altered osteogenic metabolism was not the prime cause of the high bone mass phenotype in our model of skeletal Vhl-deletion. While handling Vhl cKO mice we noticed their lean appearance. We therefore wanted to gain understanding of how increased hypoxia signaling in the skeleton could affect body composition and energy homeostasis. As such, in Chapter 5 we focused on the characterization of the systemic metabolic phenotype of Vhl cKO mice. Vhl cKO mice showed consistently reduced blood glucose levels and increased glucose tolerance from the age of 6 weeks onward. Hypoglycemia in the mutant mice was associated with lower body weights and reduced energy stores, including decreased peripheral fat accumulation and glycogen content in the liver, despite normal food intake and even reduced physical activity. The hypoglycemia could not be explained through abnormal osteocalcin or insulin signaling. Instead, glycemia levels inversely correlated with high overall uptake of glucose by the skeleton. Considering the highly glycolytic nature of Vhl-deficient osteolineage cells, we hypothesized that the low glycemia levels in the mutant mice were the result of a continual drain of glucose towards the skeleton. Hence, we investigated the metabolic phenotype of vehicle- versus DCA-treated PN-Vhl cKO mice. PN-Vhl cKO mice recapitulated the key systemic features of constitutive Vhl-deletion, including increased glucose tolerance and reduced glycemia, but without the potentially confounding or complicating aspects of lipodystrophy and reduced BW and body size, thereby representing a hypomorph model of osteoblastic Vhl-deletion. Intriguingly, when treated with DCA, the development of the metabolic phenotype was prevented in the PN-Vhl cKO mice, suggesting that cellular glucose utilization in bone may be a major determinant of systemic glucose homeostasis. Altogether, in this thesis we revealed that Vhl-deletion in osteoprogenitors leads to high bone mass, despite reduced terminal osteoblast differentiation and reduced bone formation and turnover at adult age. Additionally, we provide genetic and pharmacologic evidence supporting the notion that local glucose utilization in the skeleton contributes to systemic glucose clearance and homeostasis, a novel concept that may possibly have wide-ranging clinical implications with regard to bone and metabolic disorders.status: publishe