Distinct Osteogenic Potentials of BMP-2 and FGF-2 in Extramedullary and Medullary Microenvironments

Abstract

Bone morphogenetic protein-2 (BMP-2) and fibroblast growth factor-2 (FGF-2) have been regarded as the major cytokines promoting bone formation, however, several studies have reported unexpected results with failure of bone formation or bone resorption of these growth factors. In this study, BMP-2 and FGF-2 adsorbed into atellocollagen sponges were transplanted into bone defects in the bone marrow-scarce calvaria (extramedullary environment) and bone marrow-abundant femur (medullary environment) for analysis of their in vivo effects not only on osteoblasts, osteoclasts but also on bone marrow cells. The results showed that BMP-2 induced high bone formation in the bone marrow-scarce calvaria, but induced bone resorption in the bone marrow-abundant femurs. On the other hand, FGF-2 showed opposite effects compared to those of BMP-2. Analysis of cellular dynamics revealed numerous osteoblasts and osteoclasts present in the newly-formed bone induced by BMP-2 in calvaria, but none were seen in either control or FGF-2-transplanted groups. On the other hand, in the femur, numerous osteoclasts were observed in the vicinity of the BMP-2 pellet, while a great number of osteoblasts were seen near the FGF-2 pellets or in the control group. Of note, FCM analysis showed that both BMP-2 and FGF-2 administrated in the femur did not significantly affect the hematopoietic cell population, indicating a relatively safe application of the two growth factors. Together, these results indicate that BMP-2 could be suitable for application in extramedullary bone regeneration, whereas FGF-2 could be suitable for application in medullary bone regeneration.

Keywords: BMP-2; FGF-2; bone formation; bone marrow.

Figure 1

BMP-2, but not FGF-2, promotes repair of mouse calvarial defect. Collagen pellets containing BMP-2 (10 µg) or FGF-2 (1 µg, 10 µg) or distilled water (DW, control) were transplanted into mouse calvarial defects. (A) Frontal section (upper panel) and 3D (lower panel) images of micro-CT, 14 days after transplantation. (B) Graph shows the quantitative analysis of the regenerated bone volume (RBV) (n = 3, *** p < 0.001 versus control pellet (#). One-way ANOVA/Tukey). (C) HE-staining of frontal sections at the defect area at 5 days and 14 days after surgery. The results are representative of at least three independent experiments. Lower panels are high magnification images of the squares in the upper images. Two-way arrows, white arrowheads and black arrowheads indicate the bone defects, newly-formed bone and calvarial bone, respectively.

Figure 2

FGF-2, but not BMP-2, promotes repair of mouse femoral defect. Collagen pellets containing BMP-2 (10 µg) or FGF-2 (1 µg, 10 µg) or DW (control) were transplanted into mouse femur defects. (A) Micro-CT sagittal images of the femur (upper panel) and 3D reconstructed images of the trabecular bone (lower panel) of an area ranging from 1 mm above and below the defect. (B) Graph shows the quantitative analysis of BV/TV of an area within 1 mm above and below the defect in the femur. (n = 3–4, * p < 0.05 versus control pellet (#). One-way ANOVA/Tukey). (C) HE-staining of frontal sections at the femur defect area at 5 days and 14 days post-surgery. The results are representative of at least three independent experiments. Lower panels are high magnification images of the squares in the upper images (remained collagen pellet *). Two-way arrows and white arrowheads indicate the bone defects and newly-formed bone, respectively.

Figure 3

Depletion of bone marrow cells inhibits FGF-2-induced bone formation in the marrow cavity. FGF-2-adsorbed collagen pellets were implanted in mouse femur defects in the following conditions: (#1) Intact marrow (normal, control); (#2) Ablated marrow (ablation); (#3) Femur translocation; (#4) Marrow ablation and femur translocation. (A) Sagittal section images of femur defect areas (left panel) and 3D images of the trabecular bone (right panel). (B) Quantitative evaluation of BV/TV of trabecular bone in femur defect area. (n = 4, *** p < 0.001 versus control pellet (#). One-way ANOVA/Tukey).

Figure 4

Effects of BMP-2 and FGF-2 on osteoblast and osteoclast in mouse calvarial and femoral defect. Cross-sectional frozen sections of calvaria (frontal plane, A) and femurs (B) obtained from Col1a1(2.3)-GFP/Trap-tdTomato mice after 5 and 14 days of collagen pellet implantation. GFP-positive osteoblasts and tdTomato-positive osteoclasts are shown in green and red, respectively. Dashed lines indicate the cortical bone (white) and the remained collagen pellet (*, yellow). Nuclei were stained with DAPI (blue). The results are representative of at least three independent experiments.

Figure 5

Effects of BMP-2 and FGF-2 on angiogenesis in mouse calvarial and femoral defect. Cross-sectional frozen sections of calvaria (frontal plane, A) and femurs (B) obtained from wild-type mice after 5 and 14 days of collagen pellet implantation. EMCN-positive endothelial cells are shown in purple. Dashed lines indicate the cortical bone (white) and the remained collagen pellet (*, yellow). Nuclei were stained with DAPI (blue). The results are representative of at least three independent experiments. Collagen pellets containing BMP-2 (10 µg) or FGF-2 (10 µg) or DW (control) were transplanted into mouse femur defects and samples were analyzed by FCM (C) at 5 days and 14 days after transplantation. Endothelial cell: CD31⁺CD45⁻. (n = 3–6, One-way ANOVA/Tukey).

Figure 6

Effects of BMP-2 and FGF-2 on bone marrow cell populations. Collagen pellets containing BMP-2 (10 µg) or FGF-2 (10 µg) or DW (control) were transplanted into mouse femur defects and samples were analyzed by FCM at 5 days (A) and 14 days (B) after transplantation. HSCs: Lin⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻, Erythroid cells: Ter119⁺, Myeloid cells: CD11b⁺Gr-1⁺, B cells: B220⁺, T cells: CD3⁺, Leptin R⁺ stromal cells: CD45⁻Ter119⁻CD31⁻LepR⁺ (n = 4–11, * p < 0.05, ** p < 0.01, *** p < 0.001 versus control pellet (#). One-way ANOVA/Tukey).