PDGFRα+ cells gave arise to ECs of osteogenic vessels during bone remodeling
To gain an insight into the origin of osteogenic vessels, we took advantage of lineage tracing method to delineate the progeny produced by PDGFRα+ cells by crossing PDGFRα-creERT2 mice with YFP reporter mice to generate tamoxifen-inducible PDGFRα-creERT2;YFP mice. PDGFRα-creERT2;YFP mice received tamoxifen at two-month-old for 3 days, 14 days, 21 days and 28 days. Immunofluorescent analysis revealed that most PDGFRα+/YFP+ cells reside near Tie-2+PECAM-1+ vessels as perivascular cells in metaphysis three days after tamoxifen injection (Fig. 1a). The YFP+Tie-2+PECAM-1+ cells appeared 14 days post injection and the cell numbers were increased afterward (Fig. 1b, c). Osteoprogenitors, highly expressing Sp7, are committed to the osteoblast lineage to contribute to de novo bone formation. We observed a large number of Sp7+ osteoprogenitors in the bone marrow cavity in the metaphysis. However, few YFP+ cells were overlaid with Sp7 (< 8%), suggesting PDGFRα lineage cells mainly participated in osteogenic vessel formation, instead of osteoblast differentiation (Fig. 1d, e).
To further elaborate the contribution of PDGFRα+ cells to osteogenic vessels, we generated a mouse model that expressed diphtheria toxin receptor (iDTR) under PDGFRα-creERT2 promoter to conditionally kill PDGFRα linage cells upon tamoxifen and diphtheria toxin (DT) injection. We found that the number of PDGFRα+ cells were significantly reduced after four weeks’ daily administration related to vehicle treatment (Fig. 1f). Moreover, Sp7+ cell number were also dramatically reduced (Fig. 1f, g). Interestingly, the number of Tie-2+PECAM-1+ vessels also decreased remarkably (Fig. 1h, i). Altogether, these results suggested that the PDGFRα+ cells mainly contribute to Tie-2+PECAM-1+ vessels.
Bone marrow PDGFRα+ cells associate with Tie-2+PECAM-1+ vessels
To assess the meaningful relationship between PDGFRα + cells and osteogenic vessels during aging, we analyzed the femur of 1-, 3-, 6- and 12-month-old wild type mice. In femur metaphysis, abundant PDGFRα+ cells were seen at the metaphysis sites in puberty (1-month-old) mice and more than those in young adults (3-month-old mice). The cell number was significantly decreased with aging and PDGFRα+ cells were nearly absent in aged (12-month-old) mice (Fig. 2a, b). These changes were accompanied by remarkable reduction of Tie-2+PECAM-1+ vessels, with more being seen in 1-month-old mice and 3-month-old mice. The 12-month-old mice were devoid of any Tie-2+PECAM-1+ vessels, leaving PECAM-1−Tie-2+ vessels present at the femur metaphysis (Fig. 2c, d). Almost all Tie-2+PECAM-1+ ECs (> 90%) are Tie-2+ and PECAM-1+ cells in metaphysis in 1 and 3-month-old mice while the percentage dropped to nearly zero in 12-month-old mice (Fig. 2c, e), suggesting Tie-2+PECAM-1+ vessels are related to active bone remodeling. Taking together, simultaneous appearance and loss of PDGFRα + cells and Tie-2+PECAM-1+ vessels suggest a close association between PDGFRα + cells and Tie-2+PECAM-1+ vessels.
Blockage of HIF1α signaling in the PDGFRα + cells reduced osteogenic vessels
It has been previously demonstrated that HIF1α generated signals are required for angiogenesis, which is necessary for initial specification of bone-forming osteoblasts. The knockout of HIF1α from osteoclasts or osteocytes in mice showed a reduction in bone loss8,9. However, the relationship between HIF1α and BMPCs has not been revealed. In this study, we injected Roxadustat (50 mg/kg, intraperitoneal injection)24, an HIF-α prolyl hydroxylase inhibitor to block HIF signaling in 2-month-old mice. We found that the majority of PDGFRα+ cells were co-localized with Tie-2+ ECs and PECAM-1+ ECs in the metaphysis in vehicle (5%DMSO / 45%PEG300 / 50% ddH2O) injected mice (Fig. 3a, b). The numbers of PDGFRα+ cells, Tie-2+ ECs and PECAM-1+ were significantly decreased in metaphysis area in Roxadustat treated mice relative to their age-matched vehicle treated littermates (Fig. 3a - e). Furthermore, the Tie-2+PECAM-1+ vessels were significantly reduced (Fig. 3f, g).
Since we have demonstrated that chemical inhibition of HIF1α using Roxadustat in vivo decreased the numbers of osteogenic vessels, we further investigated whether PDGFRα + cells were the targets for HIF1α inhibition in osteogenic vessel formation. We generated an inducible knockout mouse model with HIF1α deficient in the PDGFRα+ cells (HIF1α−/−) by crossing floxed HIF1α mice with PDGFRα-creERT2 mice. In this mouse model, PDGFRα+ cells no longer respond to HIF1α signaling stimulations. We found that the numbers of PDGFRα+ cells and Tie-2+PECAM-1+ ECs in the metaphysis were significantly lower in 3-month-old HIF1α−/− mice than their wild type littermates, vehicle treated PDGFRα-creERT2;HIF1αfl/fl (HIF1αfl/fl) mice (Fig. 3h – k). The trabecular bone volume / tissue volume and numbers were also significantly reduced in 3-month-old HIF1α−/− mice relative to HIF1αfl/fl mice (Fig. 3l, m). Moreover, bone growth in HIF1α−/− mice were remarkably slower related to their wild type littermates visualized by calcein double labeling (Fig. 3n). Accordingly, the dynamic parameters such as mineral apposition rate and bone formation rate were significantly dropped in HIF1α−/− mice (Fig. 3o). Altogether, these results showed that PDGF receptor deficient PDGFRα+ cells fail to differentiate to osteogenic vessels and sequential bone formation, further demonstrating the causality between PDGFRα+ cells and Tie-2+PECAM-1+ vessels regulated by PDGF-BB.
PTH stimulates PDGFRα+ cells commitment to osteogenic vessels.
PTH is the only FDA-approved anabolic agent for osteoporosis. Increasing evidence suggests that PTH regulates bone formation via effects on angiogenesis due to coupled bone and capillaries formation found during development, fracture healing and remodeling25–27. To further validate the role of PDGFRα+ cells in osteogenic vessel formation in vivo, we altered bone marrow environment by administrating PTH intermittently into 3-month-old wild type and PDGFRα-creERT2;YFP mice. PTH dramatically stimulated Tie-2+PECAM-1+ vessel formation after injection with the number of Tie-2+PECAM-1+ cells being increased three folds by 4 weeks’ administration (Fig. 4a, b). Similarly, the number of both PDGFRα+ cells, which attached to Tie-2+ vessels in metaphysis, and PDGFRα lineage cells were also increased by PTH injection, the majority of which constituted to Tie-2+PECAM-1+ vessels, were also elevated (Fig. 4c - f). We also injected PTH in aged PDGFRα-creERT2;YFP mice and found that PTH administration increased Tie-2+PECAM-1+ cell number along with enhancement of bone formation (Fig. 4g, h). Taking together, the results showed PDGFRα+ cells play an important role in bone formation and further validated that PDGFRα+ cells differentiate osteogenic Tie-2+PECAM-1+ vessels.
PTH increased Tie-2+PECAM-1+ vessel formation via HIF1α signaling in BMPC.
To examine the mechanism by which PTH increased Tie-2+PECAM-1+ vessel formation, we injected PTH or PBS along with tamoxifen or vehicle in 3-month-old PDGFRα-creERT2;HIF1α mice to evaluate the effect of HIF1α signaling in PDGFRα+ cells in an active bone remodeling environment. We demonstrated that deprived of HIF1α in PDGFRα+ cells significantly suppressed bone formation after PTH injection in HIF1α−/− mice, which, by contrast, was seen in PTH/veh injected PDGFRα-creERT2;HIF1α (HIF1αfl/fl) mice (Fig. 5a and 5b). Consistent with previous findings (Fig. 4a, b), PTH stimulated Tie-2+PECAM-1+ vessels formation in HIF1αfl/fl mice, which was not noted in HIF1α−/− mice (Fig. 5c, d). Moreover, although the number of osteogenic vessels was slightly increased with PTH treatment vs. PBS treatment, it was significantly decreased in PTH treated HIF1α−/− mice than HIF1αfl/fl mice (Fig. 5c, d). Together, these results demonstrate PDGFRα + cells give rise to Tie-2+PECAM-1+ vessels for angiogenesis coupled with bone formation via HIF1α signaling.