Fig 2: GC treatment induces a reduction in ANG expression in vascular-associated osteoclasts.a Three-week-old BALB/c mice were treated with MPS at 10 mg/m2/day or vehicle by daily intraperitoneal injection for 2 weeks. Bone tissue extracts of femoral metaphysis were prepared. Relative levels of 53 mouse angiogenesis-related proteins in the tissue extracts were simultaneously assessed using the Proteome Profiler Mouse Angiogenesis Array Kit. The factors differentially expressed were marked with red squares. b and c Immunofluorescence staining of femoral metaphysis sections of four-week-old (Young) and four-month-old (Adult) C57BL/6J mice were performed using antibody against ANG. Representative images are shown in (b). Red: ANG; Blue: DAPI. Quantified numbers of ANG+ cells in primary spongiosa per mm2 tissue area (N. ANG+ cells/Ar) are shown in (c). d–k Three-week-old C57BL/6J mice were treated with MPS at 10 mg/m2/day or vehicle by daily intraperitoneal injection for 2 weeks. Immunofluorescence staining of femur metaphysis sections were performed using antibodies against ANG. Representative images are shown in (d). Red: ANG; Blue: DAPI. Quantified numbers of ANG+ cells in primary spongiosa and diaphyseal bone marrow per mm2 tissue area (N. ANG+ cells/Ar) are shown in (e) and (f), respectively. Double immunofluorescence staining of femoral metaphysis sections was performed using antibodies against ANG and RANK or ANG and TRAP. Representative images are shown in (g) and (j), respectively. Red: ANG; Green: RANK or TRAP; Blue: DAPI. Boxed areas are shown at a higher magnification in corresponding panels to the right. Quantified numbers of RANK-double and ANG-double positive cells (N. RANK+ANG+ cells/Ar) and TRAP- and ANG-double positive cells (N. TRAP+ANG+ cells/Ar) are shown in (h) and (k), respectively. Percentages of ANG-expressing cells in the RANK+ and TRAP+ cell populations (%RANK+ cells expressing ANG and %TRAP+ cells expressing ANG) are shown in (i) and (l), respectively. m Four-color immunofluorescence staining of femur metaphysis sections was performed using antibodies against ANG (red), RANK (green), and Emcn (cyan). DAPI stains nuclei. Note: cells in yellow (ANG-expressing osteoclasts) are closely associated with the cells in cyan (Emcn+ vascular endothelial cells). Arrows indicate RANK+ANG+ double-positive cells reside closely to Emcn+ vascular endothelial cells. n Schematic diagram showing ANG-expressing osteoclasts that reside closely to type-H vessels in primary spongiosa in the absence or presence of MPS treatment. GP growth plate. n = 4–8 mice. Data are represented as mean ± s.e.m. **p < 0.01, ns not significant as determined by two-tailed Student’s t-tests.

Fig 3: ANG secreted from metaphyseal osteoclasts mediates osteoclast-vascular crosstalk.a Schematic diagram showing the procedure of the ex vivo experiments. Metaphyseal bone of distal femur was isolated from 3-week-old C57BL/6J mice and cultured in serum-free DMEM for 1 day to adjust to the ex vivo environment. The explant was then cultured for 3 days with different treatments. Fresh medium was replaced after washing away the medium. Conditioned medium (CM) was collected after another 3-day culture. b The concentrations of ANG in the CMs were measured by ELISA. c Different CMs were individually added in the HUVEC culture, and the proliferation of the cells was measured by MTT assay. d and e Different CMs were individually added in the HUVEC culture planted on Matrigel. Tube formation assay images are shown in (d). Quantitative analysis of cumulative tubule length is shown in (e). PTH parathyroid hormone, ALN alendronate. Experiments were performed in triplicate and were repeated three times. Data are represented as mean ± s.e.m. **p < 0.01 as determined by one-way ANOVA with post hoc Tukey test.

Fig 4: GC Treatment suppresses osteoclast formation but does not directly regulate ANG gene expression in osteoclasts.a Schematic diagram showing the procedure of the in vitro experiments. Bone marrow mononuclear cells/macrophages were isolated from 3-week-old C57BL/6J mice and cultured with M-CSF and RANKL for 5 days to induce the formation of mature osteoclasts. The cells were treated with vehicle or MPS with or without RU486 for the entire 5 days (i) or only the last day (ii). The cells were subjected to TRAP staining or lysed to measure the mRNA expression level of ANG. TRAP staining of the cells is shown in (b). Quantified number of TRAP+ spreading multinucleated cells/well is shown in (c). The levels of ANG expression were analyzed by qRT-PCR (d). Experiments were performed in triplicate and were repeated three times. Data are represented as mean ± s.e.m. **p < 0.01, ns, not significant as determined by one-way ANOVA with post hoc Tukey test.

Fig 5: The ANG/PLXNB2 axis is essential to maintain the proliferative activity of vascular endothelial cells and protect them from senescence.a–d Three-week-old C57BL/6J mice were treated with MPS at 10 mg/m2/day or vehicle by daily intraperitoneal injection for 2 weeks. Double immunofluorescence staining of femoral sections using antibodies against PLXNB2 (green) and Emcn (red) or Osx (red) (a). Simultaneous fluorescence immunostaining of Emcn (red) and FISH analysis of 47S rRNA transcription (green) in bone tissue sections were shown in (b). DAPI stains nuclei blue. Boxed areas are shown at a higher magnification in corresponding panels to the right. Quantitative analysis of percentage of Emcn+ blood vessels in (c) and non-blood vessel cells (d) that show positive signal of 47S rRNA transcription in primary spongiosa. Blood vessels with more than 20 green dots are considered as positive. Non-blood vessel cell with more than five green dots in nucleus are considered positive. e Immunoblot of PLXNB2 in HUVECs transfected with two individual PLXNB2 siRNAs (siPLXNB2-1, -2) or scrambled control siRNA (siCTRL). f–i HUVECs were transfected with siCTRL or siPLXNB2 for 36 h and then stimulated with 200 ng/mL rhANG or vehicle for another 12 h. Cells were fixed and immunofluorescence staining was performed using antibody against ANG (red) in (f). Arrow heads indicate nuclear translocation of ANG. DAPI stains nuclei blue. Immunoblot was performed using antibodies against p-AKT, AKT, p-ERK, or ERK in (g). Click-iT RNA-imaging assays were performed, and confocal images of the cells pulsed with EU for 2 and 6 h were obtained (green: Alexa 488-EU) in (h). Cells were subjected to SA-ßGal staining with representative images in (i). j HUVECs were treated transfected with siCTRL or siPLXNB2 for 48 h. Immunoblot was performed using individual antibodies against LaminB1, Ki67, p21, and GAPDH. k, l HUVEC was seeded on Matrigel with control IgG, 26-2F or mAb17. Images of tubule formation is shown in (k). Quantitative analysis of cumulative tubule length is shown in (l). **p < 0.01 as determined by one-way ANOVA with post hoc Tukey test. GP growth plate. n = 4-6 mice, Data are represented as mean ± s.e.m. **p < 0.01, ns not significant as determined by two-tailed Student’s t-tests.