Bone formation never occurs without vascular interactions.1–3 This obvious and simple statement of biological fact does not do justice to the physiological and pharmacological implications of the relationship. The vasculature is, of course, the conduit for calcium phosphate and nutrient exchange between bone and the rest of the body; this is relevant not only to the rapid mobilization of skeletal calcium when urgently needed4 but also to the delivery of metabolic substrate to the basic multicellular unit (BMU) for bone-forming osteoblast functions.5 However, the vasculature also provides the sustentacular niche for osteoblast progenitors,6 and is the conduit for egress of bone marrow-derived formed elements from the osteoblast-regulated hematopoietic niche.7 Indeed, Ignarro and colleagues8 have coupled parathyroid hormone (PTH)-mediated bone anabolism with granulocyte colony stimulating factor (G-CSF) treatment as a strategy to mobilize endothelial progenitor cells (ePCs) and enhance ischemic limb recovery in mice. Surprisingly, the inextricable interdependence of vascular physiology, skeletogenesis, bone remodeling, and mineral metabolism has escaped widespread appreciation until very recently. Studies by Clemens et al.9 established that the constitutive elaboration of vascular endothelial growth factor A (VEGF) by osteoblasts increases bone formation via mechanisms that require expansion of skeletal vascularity. They went on to show that pharmacologic augmentation of endogenous skeletal VEGF production accelerates skeletal repair during distraction osteogenesis.10 Exogenous VEGF implants exhibit similar properties in critical bone defects.11 However, the regulation of bone vascular biology and anatomy in response to metabolic, morphogenetic, mechanical, inflammatory, and endocrine cues is very poorly understood. A fundamental understanding of bone–vascular interactions will be necessary to develop new strategies that safely and efficaciously enhance bone formation in the settings of malignancy, fracture nonunion, the open epiphyses of childhood, and the arteriosclerotic vasculopathy of diabetes, uremia, rheumatoid arthritis, and advanced age.
In this issue of the Journal,12 Prisby, Lafage-Proust, and colleagues12 address this important unmet scientific need by examining the regulation of long bone vasculature by PTH, the prototypic osteoanabolic hormone, in growing rats. PTH dosing transiently upregulates VEGF expression in bone followed by subsequent increases in bone formation and bone mass. The skeletal anabolic response to PTH is fully dependent upon intact VEGF signaling; PTH-induced increases in the key histomorphometric indices of osteoblast synthetic function and numbers—and accompanying improvements in trabecular bone mass and architecture—were all inhibited by administration of a VEGF-neutralizing antibody. Anti-VEGF treatment also inhibited PTH induction of Neuropilin1 and Neuropilin2, two key VEGF coreceptors that enhance endothelial sprouting and migration in response to matrix-bound VEGF isoforms liberated with matrix turnover.13 Intriguingly, bone vessel density was not increased by PTH administration;12 in fact, bone vessel density was decreased as assessed by synchrotron radiation microCT analysis, paralleling changes in trabecular osteoclast numbers.12 Of note, this response is remarkably different from the bone anabolic responses elicited by genetically induced, sustained osteoblast VEGF secretion;9 bone anabolic responses arising from unrestrained paracrine VEGF signaling markedly increases vessel density.9 The investigators' painstakingly detailed histomorphometric analyses demonstrated that PTH treatment reduced the distance between capillaries or small arterioles (<29-micron diameter) and the osteoid surface of the BMU.12 However, PTH did not change the spatial relationships for larger vessels (>100-micron diameter). Additionally, no PTH-induced differences were noted in the proximity of vessels of any diameter to bone trabeculae in general, highlighting the specific reorientation of microvessels toward sites of new bone formation.12 Thus, the authors conclude that intermittent PTH exerts bone anabolism via VEGF-dependent mechanisms that do not involve changes in skeletal vascular mass.12 Rather, intermittent PTH dosing regulates the spatial relationships between the anabolic BMU and the microvasculature, localizing capillaries near sites of new bone formation as necessary for bone mass accrual.
Why is this manuscript so significant? Firstly, the ingenious combination of barium sulfate perfusion with Goldner trichrome staining and digital image analysis of tibial plastic sections is a technical advance that should not go unheralded. As such, the authors' approach transforms and elevates our expectations for state-of-the-art histomorphometric phenotyping of bone anabolic responses. This unparalleled, quantitatively robust analysis of bone vascular anatomy and its regulation by PTH has provided truly novel insights into how anabolic signals control and coordinate bone mass accrual via VEGF signaling in the bone–vascular axis.14 As a transient osteoblast VEGF “secretagogue,” the endocrine actions of circulating PTH pulses12 may now be seen to help maintain BMU–microvascular proximity via VEGF paracrine signals that sustain osteoblast synthetic functions. However, although pharmacological proof-of-concept is in hand,12 the impact of endogenous endocrine PTH activity on osteoblast VEGF expression and BMU–bone microvascular alignment has yet to be examined. Nevertheless, this new biology is important to consider in the clinic—particularly so in the setting of end-stage renal disease where uremia-induced PTH resistance, aggressive parathyroidectomy, and/or calcitriol dosing can interact to iatrogenically induce a devastating low turnover mineral and bone disorder.15 Moreover, because inhibitors of VEGF signaling are increasingly utilized for treatment of malignancy,16 significant potential exists for inducing low turnover bone disease in this setting as well via perturbation of the bone–vascular axis relationship. Of note, aminobisphosphonates potently inhibit angiogenesis and endothelial cell (EC) migration;17 thus, alterations in the bone–vascular axis may explain in part the blunting of trabecular bone anabolic responses to PTH treatment when aminobisphosphonates are coadministered.18
Secondly, Prisby et al.12 highlight that it is the proximity of the BMU to the microvasculature—not the mass of the skeletal microvasculature—that is critical in bone anabolism. To be sure, perfusion- and diffusion-dependent nutrient supplies must be rate-limiting contributors.19 Indeed, VEGF is a “vascular permeability factor” that enhances trans-endothelial fluid flow, organizes conduit function, and promotes bone canalicular flow.20 However, it has become increasingly evident that the vasculature itself provides paracrine and juxtacrine cues that regulate osteoblast function.21, 22 For example, BMP2 is expressed by microvascular ECs23 and is upregulated by VEGF, hypoxia,23 and mechanical loading.24 Pro-osteogenic Wnt7 family members are also elaborated by ECs and vascular sinusoids in bone.25 In addition, ephrin B2, an important stimulus for anabolic EphB4 signaling in osteoblasts,26 is expressed by ECs27 and enhances EC motility.28 VEGF upregulates ephrin B2 expression during EC differentiation, and subsequent VEGF-induced EC migration to the BMU may provide coordinated juxtacrine and paracrine signals necessary for bone anabolism.
Finally, and most importantly, when taken against the backdrop of the recent studies of Maes et al.,29 the data of Prisby et al.12 demonstrate that the pharmacokinetic–pharmacodynamic (PK-PD) relationship for angiogenic signals in the bone–vascular axis is of vital interest to skeletal pharmacology. Constitutive VEGF expression by osteoblasts not only increases bone mass and skeletal microvascular density, but also increases osteoclast numbers and induces bone marrow fibrosis.29 All three responses—bone mass accrual, skeletal microvascular density, and marrow fibrosis—arise via constitutive downstream activation of skeletal β-catenin.29 The transient kinetics of VEGF induction by intermittent PTH dosing may go far to explain the differences in bone histology of (a) VEGF-dependent bone anabolism arising from intermittent PTH administration12 versus (b) unrestrained osteoblast VEGF expression.9 This notion is supported by the data of Jilka and colleagues30 and Schipani et al.31 Continuous rather than intermittent PTH signaling increases marrow vessel density, osteoblast, and osteoclast number—and responses were abrogated by inhibition of RANKL-mediated osteoclast recruitment.31 Importantly, the recruitment of osteoclasts indirectly enhances bone angiogenesis and fibrosis.32 Of note, an underappreciated aspect of the histological response to intermittent PTH is the complex time course with respect to osteoclast recruitment vs. trabecular and endocortical resorption—modestly increased early on but reduced with time with intermittent dosing in rats,33, 34 mice,35 and humans.36 PTH-induced reductions in osteoclast density were once again noted by Prisby et al.32 in their pulsatile paradigm at 30 days;12 this PK-PD relationship for PTH-regulated trabecular and endocortical osteoclast activity may mechanistically contribute to the “uncoupling” of osteoanabolism and osteoangiogenesis observed with intermittent dosing but not with continuous PTH administration.30 Of note, in the macrovasculature, viz, aortic vascular smooth muscle—cell-autonomous PTH/PTHrP receptor (PTH1R) activation directly inhibits profibrotic and procalcific β-catenin signaling.37 This macrovascular response is immediately germane to the biology of arteriosclerotic calcification and conduit vessel fibrosis37—but, in principle, might also partially mitigate fibrotic responses in marrow microvasculature. The expression and consequences of PTH1R activation in the microvasculature have not yet been studied in sufficient detail. Nevertheless, PTHrP, a selective PTH1R agonist, induces endothelium-dependent arterial vasorelaxation38, 39 and directly inhibits EC migration and proliferation.40 Thus, VEGF-dependent skeletal anabolism arising from approximation of the bone microvasculature to the BMU after PTH administration may reflect the net actions of (1) PTH-induced osteoblast VEGF secretion and augmented nutrient exchange, (2) microvascular EC-derived osteoanabolic signals, and (3) cell-autonomous, PTH1R-modulated EC responses.
As noted above, recruitment of osteoclasts enhances angiogenic responses and marrow vascularity but also engenders marrow fibrosis. Should inhibition of RANKL signaling reduce osteoclast numbers and activity without altering EC migration and proliferation, various combinations of PTH1R agonists, RANKL antagonists, and VEGF ligands may afford finely tuned bone anabolic responses. This might be particularly useful when crafting personalized bone health management strategies along the low turnover—high turnover—marrow fibrosis continuum that confronts physicians and their patients with end-stage renal disease and/or fracture nonunion.41 Without a doubt, the improved understanding of the bone–vascular axis and its regulation will shape future clinical strategies that address these and other unmet needs in musculoskeletal health and healthcare.