PTH Stimulates Osteoblastic VEGF-A Production and Remodels Bone Marrow Micro-Endothelial Structures

Abstract 722

In their niches within the bone marrow (BM) microenvironment, hematopoietic stem cells (HSCs) interact with a number of other cellular and molecular factors that can affect their regulation. We previously demonstrated that activation of osteoblasts (OBs) by intermittent Parathyroid Hormone (PTH) stimulation or by expression of a constitutively active PTH receptor (Col1caPTH1R mice, hereafter referred to as TG) expands HSCs with long term repopulating ability. PTH treatment has also been found to improve survival after BM injuries that severely disrupt the hematopoietic system as well as the BM vasculature; however the mechanism for this effect is unknown. Osteoblastic cells respond directly to PTH administration, whereas HSCs lack expression of the PTH1R, indicating that PTH-mediated HSC expansion occurs through the BM microenvironment. Both endosteal osteoblastic cells and the vasculature constitute HSC niches within the BM. Intermittent PTH exerts an anabolic effect on bone, expanding trabecular bone and bone-lining stromal cells. New vessel formation is required for adult bone remodeling. To test if PTH also increases the BM vasculature, we treated mice with an intensive regimen of systemic PTH that expanded hematopoietic stem and progenitor cells (HSPCs) and analyzed hindlimb histology as well as BM cell immunophenotype. In the tibial and femoral metaphyseal region, PTH increased microvessels (167 ± 18.1 vs 348 ± 39.3 microvessels/hindlimb, n=5 mice per group, p=0.0030) and vascular area ((mm²): 0.135 ± 0.0194 vs 0.281 ± 0.00951, n=5 mice per group, p=0.0001) measured by histomorphometry. PTH also increased the frequency of PECAM+ BM endothelial cells (0.0935 ± 0.0173 vs 0.196 ± 0.0172, n=4–5 mice per group, p=0.0043) measured flow cytometrically. To determine if PTH increases osteoblast-derived angiogenic signals, mouse calvarial MC3T3 cells were treated with PTH(1-34) at various stages of osteoblastic maturation. Differentiation of MC3T3 cells did not affect baseline Vegfa expression. PTH strongly induced total Vegfa expression in day 7 cells at two hours, an effect that peaked at six hours (3.83 ± 1.76 vs 50.7 ± 9.72 fold change above baseline, n=3, p=0.0090) and was sustained 24 hours after treatment. The magnitude of the effect increased throughout osteoblastic differentiation. In vitro PTH also increased levels of secreted VEGF-A protein at corresponding time points (VEGF-A(pg/mL): 43.4 ± 2.27 vs 464 ± 24.2, n=3, p<0.0001). Similarly, PTH induced Vegfa expression two hours post-stimulation in a second osteoblastic cell line, derived from rat osteosarcoma (1.133 ± 0.2963 vs 8.500 ± 1.320 fold change above baseline, n=3, p=0.0055). Vegfa pre-mRNA comprises eight exons that undergo alternative splicing to form variants encoding several VEGF-A isoforms. These isoforms differ functionally primarily due to their varying heparan-binding affinity. Specifically, VEGF189 remains mostly cell and matrix-associated due to its interactions with heparan sulfate proteoglycans in extracellular matrix and on cell surfaces. Because PTH stimulation dramatically increases extracellular matrix and bone volume, expression of the splice variant encoding matrix-associated VEGF189 was assayed in PTH-stimulated MC3T3 cells. PTH strongly increased expression of the VEGF189-encoding variant six hours after treatment of maturing MC3T3 cells (1.70 ± 1.100 vs 113 ± 25.1, n=3, p=0.0416), suggesting that PTH may favor expression of a more highly localized VEGF-A isoform. Analysis of BM plasma revealed that soluble VEGF-A levels were significantly decreased both in TG mice (VEGF-A(pg/mL): 74.2 ± 7.05 vs 44.0 ± 3.67, n=3-5 mice per group, p=0.0054) and in mice treated systemically with PTH (VEGF-A(pg/mL): 105 ± 6.62 vs 78.6 ± 4.95, n=9 mice per group, two independent experiments, p=0.0054), while serum VEGF-A levels were unchanged. Since PTH strongly stimulates osteoblastic expression of matrix-bound VEGF-A but less soluble VEGF-A is detected in the BM during PTH stimulation in vivo, we speculated that PTH-dependent proangiogenic signals in the BM microenvironment may be highly localized via modulation of Vegfa pre-mRNA splicing in osteoblastic cells. Because the blood perfusion status of BM niches profoundly affects HSPC behavior, these findings may suggest a mechanism by which PTH establishes highly localized niches to instruct the fate of resident HSCs.