Delivery of VEGFA in bone marrow stromal cells seeded in copolymer scaffold enhances angiogenesis, but is inadequate for osteogenesis as compared with the dual delivery of VEGFA and BMP2 in a subcutaneous mouse model

Abstract

Background: In bone tissue engineering (BTE), extensive research into vascular endothelial growth factor A (VEGFA)-mediated angiogenesis has yielded inconsistent results. The aim of this study was to investigate the influence on angio- and osteogenesis of adenoviral-mediated delivery of VEGFA alone or in combination with bone morphogenetic protein 2 (BMP2) in bone marrow stromal cells (BMSC) seeded onto a recently developed poly(LLA-co-CL) scaffold.

Methods: Human BMSC were engineered to express VEGFA alone or in combination with BMP2 and seeded onto poly(LLA-co-CL) scaffolds. Changes in angiogenic and osteogenic gene and protein levels were examined by quantitative reverse-transcription polymerase chain reaction (RT-PCR), PCR array, and alkaline phosphatase assay. An in vivo subcutaneous mouse model was used to investigate the effect on angio- and osteogenesis of VEGFA alone or in combination with BMP2, using microcomputed tomography (μCT), histology, immunohistochemistry, and immunofluorescence.

Results: Combined delivery of a lower ratio (1:3) of VEGFA and BMP2 (ad-BMP2 + VEGFA) led to upregulation of osteogenic and angiogenic genes in vitro at 3 and 14 days, compared with mono-delivery of VEGFA (ad-VEGFA) and other controls. In vivo, in a subcutaneous mouse model, both ad-VEGFA and ad-BMP2 + VEGFA scaffold explants exhibited increased angiogenesis at 2 weeks. Enhanced angiogenesis was largely related to the recruitment and differentiation of mouse progenitor cells to the endothelial lineage and, to a lesser extent, to endothelial differentiation of the implanted BMSC. μCT and histological analyses revealed enhanced de novo bone formation only in the ad-BMP2 + VEGFA group, corresponding at the molecular level to the upregulation of genes related to osteogenesis, such as ALPL, RUNX2, and SPP1.

Conclusions: Although BMSC expressing VEGFA alone or in combination with BMP2 significantly induced angiogenesis, VEGFA alone failed to demonstrate osteogenic activity both in vitro and in vivo. These results not only call into question the use of VEGFA alone in bone regeneration, but also highlight the importance in BTE of appropriately formulated combined delivery of VEGFA and BMP2.

Keywords: Angiogenesis; BMP2 and VEGFA; Bone regeneration; Gene delivery; Mesenchymal stem cell; Scaffold.

Fig. 1

VEGFA and BMP2 adenoviral vectors (ad) respectively upregulated the expression of vascular endothelial growth factor A (VEGFA) and bone morphogenetic protein 2 (BMP2) in BMSC. Representative immunofluorescence images of BMSC in monolayer transduced with ad-GFP (a), ad-VEGFA (b), and ad-VEGFA + BMP2 (c–e) adenoviral particles. BMSC (5× 10⁴) were seeded in each scaffold and harvested after 3 and 14 days for mRNA and protein analyses. Significant upregulation of VEGFA mRNA (f) and VEGFA protein (h) levels were found at both time points in both ad-VEGFA and ad-BMP2 + ad-VEGFA BMSC compared with the ad-GFP BMSC. Similarly, significant upregulation of BMP2 mRNA (g) was found in ad-BMP2 + VEGFA BMSC compared with the ad-GFP and ad-VEGFA BMSC at both time points. VEGFA and BMP2 mRNA levels were normalized to GAPDH mRNA level. i ELISA disclosed higher levels of secreted BMP2 in the culture supernatant of ad-BMP2 BMSC compared with the ad-GFP BMSC at both 3 and 14 days. Error bars in (f) and (g) represent SEM of three repeated experiments (n = 3) performed in three technical replicates. ANOVA test with Bonferroni post hoc analysis was used for statistical analysis in (f) and (g) and Student’s t test was performed in (i). Error bars in (i) represent SEM of three repeated experiments (n = 3). ***p < 0.001, **p < 0.01. GFP green fluorescent protein, ns not significant

Fig. 2

Adenoviral (ad) vector-mediated combined delivery of bone morphogenetic protein 2 (BMP2) and vascular endothelial growth factor A (VEGFA) led to upregulation of osteogenic and angiogenic molecules in vitro. ad-GFP, ad-VEGFA, or ad-BMP2 + VEGFA BMSC (5 × 10⁴) were seeded in each scaffold and harvested at 3 and 14 days for custom polymerase chain reaction (PCR) array and TaqMan-based qRT-PCR. a, b Unsupervised hierarchical clustering demonstrated two separate clusters consisting of replicates of ad-BMP2 + VEGFA, and replicates of ad-GFP and ad-VEGFA BMSC at both time points. c ANOVA disclosed that ALPL, BMP7, BMP6, FLT1, and PGF mRNA levels were found to be significantly upregulated at 3 days in ad-BMP2 + VEGFA BMSC. d More osteogenic markers were significantly induced at 14 days in ad-BMP2 + VEGFA BMSC compared with the ad-GFP and ad-VEGFA BMSC groups. Error bars represent SEM of three biological replicates (n = 3) performed in three technical replicates. ANOVA test with Bonferroni post hoc analysis was applied in c and d. e Representative images demonstrating higher alkaline phosphatase activity in ad-BMP2 + VEGFA BMSC in monolayer culture as compared with the ad-GFP and ad-VEGFA BMSC at 3 and 14 days. Experiments were repeated at least three times. ***p < 0.001, **p < 0.01, *P < 0.05. GFP green fluorescent protein, ns not significant

Fig. 3

Adenoviral (ad) vector-mediated expression of vascular endothelial growth factor A (VEGFA) alone and in combination with bone morphogenetic protein 2 (BMP2) was associated with enhanced angiogenesis in scaffold explants. a Representative images at 2 weeks revealed more reddish scaffold explants with multiple capillary/vessel like structures (black arrows) radiating from the periphery in ad-VEGFA and ad-BMP2 + VEGFA groups compared with the other control explants. b–e Frozen sections of scaffold explants at 2 weeks were stained with anti-CD31 targeting mouse CD31 protein and examined for capillary-like structures. Few CD31-positive structures were found in untransduced and ad-GFP scaffold explants (b and c, white arrows). However, in ad-VEGFA and ad-BMP2 + VEGFA explants, many CD31-positive capillary-like structures (white arrows) were seen both at the periphery (not shown in the figure) as well as within the scaffolds (d, e). Red arrowheads indicate remnants of the scaffold material. f Quantification of CD31-positive area in the scaffold explants at 2 weeks demonstrated significantly more CD31-positive structures in ad-VEGFA and ad-BMP2 + VEGFA (ad-B + V) scaffolds than in the untransduced and ad-GFP explants. ANOVA test with Bonferroni post hoc analysis was used for statistical analysis in (e). Error bars represent SEM. ***p < 0.001; **p < 0.01; *P < 0.05. GFP green fluorescent protein, ns not significant

Fig. 4

Combined delivery of bone morphogenetic protein 2 (BMP2) and vascular endothelial growth factor A (VEGFA) induced ectopic bone formation in scaffold explants in a subcutaneous NOD/SCID mouse model. The ability of VEGFA alone or in combination with BMP2 to induce ectopic bone formation was investigated in NOD/SCID mice by subcutaneously implanting scaffolds seeded with 5 × 10⁵ BMSC from different experimental groups. Scaffold explants at 8 weeks were examined by μCT to evaluate ectopic bone formation. Negligible amounts of radiopaque bone-like structures were detected in a untransduced, b ad-GFP, and c ad-VEGFA BMSC. In contrast, numerous bony-like radiopaque structures were found at the periphery as well as within the scaffolds of ad-BMP2 + VEGFA explants (d, cross sectional view in e). f Quantification of the radiopaque bone-like structures in the scaffold explants revealed significant amounts of bone formation in ad-BMP2 + VEGFA scaffolds. g–n Representative images of H&E stained FFPE sections of scaffold explants from different experimental groups at 2 and 8 weeks. g–l Formation of bony structures was not detected in untransduced, ad-GFP, or ad-VEGFA groups at either 2 or 8 weeks. m In the ad-BMP2 + VEGFA BMSC group, formation of bony structures (black arrows) could be seen as early as 2 weeks, mostly at the periphery of the scaffold explants. n Extensive bone formation (green arrows) with bony trabeculae extending inside the scaffolds was found at 8 weeks in ad-BMP2 + VEGFA BMSC. The structure consisted of numerous osteocyte-like cells at both 2 (black arrowheads, inset in m) and 8 weeks (green arrowheads, inset in n). ANOVA with Bonferroni post hoc analysis was carried out for statistical analysis in (f). Error bars represent SEM. **p < 0.01. ad adenoviral, GFP green fluorescent protein

Fig. 5

Bone morphogenetic protein 2 (BMP2) and vascular endothelial growth factor A (VEGFA) co-expression-mediated bone formation is associated with upregulation of osteogenic and angiogenic markers in the in vivo scaffold explants. Custom PCR array was used to examine differentially expressed osteogenesis- and angiogenesis-related genes in scaffold explants at 2 weeks. ANOVA revealed upregulation of a number of osteogenesis- and angiogenesis-related genes such as a ALPL, b RUNX2, c SPP1 (osteopontin), d ANGPT1, and e PGF in ad-BMP2 + VEGFA explants compared with the ad-GFP and ad-VEGFA explants. ANOVA test with Bonferroni post hoc analysis was undertaken for statistical analysis. **p < 0.01, **p < 0.05. GFP green fluorescent protein, ns not significant