Dual delivery of rhPDGF-BB and bone marrow mesenchymal stromal cells expressing the BMP2 gene enhance bone formation in a critical-sized defect model

Abstract

Bone tissue healing is a dynamic, orchestrated process that relies on multiple growth factors and cell types. Platelet-derived growth factor-BB (PDGF-BB) is released from platelets at wound sites and induces cellular migration and proliferation necessary for bone regeneration in the early healing process. Bone morphogenetic protein-2 (BMP-2), the most potent osteogenic differentiation inducer, directs new bone formation at the sites of bone defects. This study evaluated a combinatorial treatment protocol of PDGF-BB and BMP-2 on bone healing in a critical-sized defect model. To mimic the bone tissue healing process, a dual delivery approach was designed to deliver the rhPDGF-BB protein transiently during the early healing phase, whereas BMP-2 was supplied by rat bone marrow stromal cells (BMSCs) transfected with an adenoviral vector containing the BMP2 gene (AdBMP2) for prolonged release throughout the healing process. In in vitro experiments, the dual delivery of rhPDGF-BB and BMP2 significantly enhanced cell proliferation. However, the osteogenic differentiation of BMSCs was significantly suppressed even though the amount of BMP-2 secreted by the AdBMP2-transfected BMSCs was not significantly affected by the rhPDGF-BB treatment. In addition, dual delivery inhibited the mRNA expression of BMP receptor type II and Noggin in BMSCs. In in vivo experiments, critical-sized calvarial defects in rats showed enhanced bone regeneration by dual delivery of autologous AdBMP2-transfected BMSCs and rhPDGF-BB in both the amount of new bone formed and the bone mineral density. These enhancements in bone regeneration were greater than those observed in the group treated with AdBMP2-transfected BMSCs alone. In conclusion, the dual delivery of rhPDGF-BB and AdBMP2-transfected BMSCs improved the quality of the regenerated bone, possibly due to the modulation of PDGF-BB on BMP-2-induced osteogenesis.

FIG. 1.

The in vitro effects of rhPDGF-BB on AdBMP2-transfected BMSCs. (A) Effects of rhPDGF-BB treatment for 48 h on the proliferation of AdBMP2-transfected BMSCs (5×10³ cells/well). The cell proliferation of both nontransfected BMSCs and AdBMP2-transfected BMSCs was significantly increased by the rhPDGF-BB treatment at days 3 and 7 (ANOVA, p<0.001 and p<0.001, respectively). (B) The amount of BMP-2 produced by AdBMP2-transfected BMSCs (1×10⁴ cells/well), as determined by ELISA. Although BMP-2 was overexpressed over a 21-day period in AdBMP2-transfected BMSCs, BMP-2 expression was significantly decreased after 21 days compared to the initial levels (ANOVA, p<0.001). Furthermore, the rhPDGF-BB treatment did not significantly influence BMP-2 expression at each interval. (C) Effects of rhPDGF-BB on ALP (1×10⁴ cells/well). The ALP activity was significantly increased by BMP2 gene delivery, which was suppressed by the rhPDGF-BB treatment for 48 h after 7 and 14 days (ANOVA, p<0.001 and p<0.001, respectively). (D) Effects of rhPDGF-BB on the mineralization of AdBMP2-transfected BMSCs (1×10⁴ cells/well). AdBMP2-transfected BMSCs formed mineralization nodules at day 7, which was significantly suppressed by the rhPDGF-BB treatment for 48 h (ANOVA, p=0.001). *Statistically significant difference from the BMSCs and AdBMP-2-trasnfected BMSCs at day 3. ^†Statistically significant difference from the BMSCs and AdBMP-2-trasnfected BMSCs at day 7. ^‡Statistically significant difference from the other groups. ^#Statistically significant difference from the other groups. Values represent mean±SD of three samples. OD, optical density. BMP-2, bone morphogenetic protein-2; BMSC, bone marrow stromal cell; ALP, alkaline phosphatase; ANOVA, analysis of variance; PDGF-BB, platelet-derived growth factor-BB; AdBMP-2, adenoviral vector containing the BMP2 gene. Color images available online at www.liebertpub.com/tea

FIG. 2.

Gene expression patterns of osteogenic markers in AdBMP2-transfected BMSCs following rhPDGF-BB treatment. The rhPDGF-BB treatment significantly suppressed the mRNA expression of (A) Runx2, (B) Dlx5, (C) Alp, and (D) Ibsp induced by the AdBMP2-transfected BMSCs, especially at 24 h. The mRNA expression of each osteogenic factor was normalized against Gapdh. *Statistically significant difference between AdBMP2-transfected BMSCs and AdBMP2-transfected BMSCs/rhPDGF-BB (ANOVA, p<0.05). ^#Statistically significant difference between nontransfected BMSCs and AdBMP2-transfected BMSCs (ANOVA, p<0.05). ^†Statistically significant difference between nontransfected BMSCs and AdBMP2-transfected BMSCs/rhPDGF-BB (ANOVA, p<0.05). Cell, nontransfected BMSCs; cells/B2, AdBMP2-transfected BMSCs; cells/B2P, AdBMP2-transfected BMSCs/rhPDGF-BB. Values represent mean±SD of three samples.

FIG. 3.

Gene expression patterns of BMP-2 related factors Bmpr1a (A), Bmpr2 (B), Noggin (C), and Gremlin (D) in AdBMP2-transfected BMSCs following the rhPDGF-BB treatment. The mRNA expression of both BMPRII and Noggin in the AdBMP2-transfected BMSCs was significantly decreased by the rhPDGF-BB treatment. The mRNA expression of each factor was normalized against Gapdh. *Statistically significant difference between AdBMP2-transfected BMSCs and AdBMP2-transfected BMSCs/rhPDGF-BB (ANOVA, p<0.05). ^#Statistically significant difference between nontransfected BMSCs and AdBMP2-transfected BMSCs (ANOVA, p<0.05). ^†Statistically significant difference between nontransfected BMSCs and AdBMP2-transfected BMSCs/rhPDGF-BB (ANOVA, p<0.05). Cell, nontransfected BMSCs; cells/B2, AdBMP2-transfected BMSCs; cells/B2P, AdBMP2-transfected BMSCs/rhPDGF-BB. Bmpr1a, BMP receptor type Ia; Bmpr2, BMP receptor type II. Values represent mean±SD of three samples.

FIG. 4.

In vivo histology and histomorphometric analysis. (A) Representative histology at 2 and 4 weeks postsurgery (decalcified; H&E stain; original magnification 2×). Both the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups showed extensive new bone formation at 2 and 4 weeks. In the BMSC/BMP2/rhPDGF-BB group, bone bridging that covered the central region of the defect area was observed. Arrow: margin of defect. (B) New bone area (mm²) at the defect site. Both the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups showed significantly higher amounts of new bone area at 2 weeks after surgery (ANOVA, p=0.003). However, there was no significant difference between the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups. At 4 weeks, the BMSC/BMP2/rhPDGF-BB group had significantly higher levels of new bone area compared to the other groups (ANOVA, p<0.001). (C) Defect closure (%) within the defects. Both the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups produced a significantly higher percentage of defect closure at 2 weeks (ANOVA, p<0.001). However, there was no significant difference between the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups. At 4 weeks, the defect closure of the BMSC/BMP2/rhPDGF-BB group was significantly higher than the other groups (ANOVA, p<0.001). Defect closure was measured as the ratio of the area of newly formed bone divided by the area of the whole defect. *Statistically different from the BMSCs and rhPDGF-BB groups. ^†Statistically different from the other groups. Values represent mean±SD of six samples.

FIG. 5.

Micro-CT analysis and quantitative measurements of calvarial defects at 2 and 4 weeks. Both the BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups showed extensive new bone formation, in which the new bone had low radiopacity at 2 weeks. The new bone at 4 weeks was more mature and had enhanced radiopacity compared to the bone at 2 weeks. Both BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups showed significantly higher levels of mineralized BV percentage to TV of defect at 2 and 4 weeks (ANOVA, p=0.026 and p=0.001, respectively). However, there was no significant difference between BMSC/BMP2 and BMSC/BMP2/rhPDGF-BB groups. In terms of BMD, there was no statistical significance at 2 weeks. At 4 weeks, the BMSC/BMP2 group showed significantly less BMD of newly formed bone than other groups, including BMSC/BMP2/rhPDGF-BB groups (ANOVA, p=0.009). *Statistically different from the BMSCs and rhPDGF-BB groups. ^†Statistically different from the other groups. Micro-CT images: left, coronal section view; middle, transverse view; right, 3D reconstruction of newly formed bone in the defect site; arrow, margin of defect; BV/TV, the percentage of mineralized bone volume of the defect tissue volume; BMD, bone mineral density. Values represent mean±SD of six samples. micro-CT, microcomputed tomography.