Remodeling of cortical bone allografts mediated by adherent rAAV-RANKL and VEGF gene therapy

Abstract

Structural allograft healing is limited because of a lack of vascularization and remodeling. To study this we developed a mouse model that recapitulates the clinical aspects of live autograft and processed allograft healing. Gene expression analyses showed that there is a substantial decrease in the genes encoding RANKL and VEGF during allograft healing. Loss-of-function studies showed that both factors are required for autograft healing. To determine whether addition of these signals could stimulate allograft vascularization and remodeling, we developed a new approach in which rAAV can be freeze-dried onto the cortical surface without losing infectivity. We show that combination rAAV-RANKL- and rAAV-VEGF-coated allografts show marked remodeling and vascularization, which leads to a new bone collar around the graft. In conclusion, we find that RANKL and VEGF are necessary and sufficient for efficient autograft remodeling and can be transferred using rAAV to revitalize structural allografts.

Figure 1

The mouse femoral allograft model. Mice received a femoral autograft or allograft, and were killed at 3 weeks (a,c), 4 weeks (b,d,f,h) or 2 weeks (e,g). Representative radiographs from an autografted (a,b) and an allografted (c,d) mouse are illustrated at 3 and 4 weeks after fracture. The arrows indicate the presence of callus on the cortical surface of the autograft at 3 weeks (a), which is remodeled by 4 weeks (b), and is completely absent in the allograft (c,d). Hematoxylin, eosin, orange G and acian blue–stained sections show the endochondral bone formation at the graft-host junctions (arrow heads) of both auto and allografts at 2 weeks (e,g), which is remodeled to form a bony union at 4 weeks (f,h). Of note is the periosteal intramembranous bone formation (*), which only occurs in autografts (e), producing a new cortical bone collar with bone marrow at four weeks (f). In contrast, allografts are encased by fibrous tissue (#), heal through creeping callus (g), and are dependent on dead cortical bone for structural integrity after remodeling (h).

Figure 1

The mouse femoral allograft model. Mice received a femoral autograft or allograft, and were killed at 3 weeks (a,c), 4 weeks (b,d,f,h) or 2 weeks (e,g). Representative radiographs from an autografted (a,b) and an allografted (c,d) mouse are illustrated at 3 and 4 weeks after fracture. The arrows indicate the presence of callus on the cortical surface of the autograft at 3 weeks (a), which is remodeled by 4 weeks (b), and is completely absent in the allograft (c,d). Hematoxylin, eosin, orange G and acian blue–stained sections show the endochondral bone formation at the graft-host junctions (arrow heads) of both auto and allografts at 2 weeks (e,g), which is remodeled to form a bony union at 4 weeks (f,h). Of note is the periosteal intramembranous bone formation (*), which only occurs in autografts (e), producing a new cortical bone collar with bone marrow at four weeks (f). In contrast, allografts are encased by fibrous tissue (#), heal through creeping callus (g), and are dependent on dead cortical bone for structural integrity after remodeling (h).

Figure 2

Altered Tnfsf11 and Vegfa gene expression during allograft healing. Total RNA was extracted from femoral autografts and allografts at the indicated time and processed for real time RT-PCR. The data are presented as the fold induction ± s.d., compared to the day 0 control, after standardization with the internal β-actin control. *P < 0.05 for autograft versus allograft at the same time point.

Figure 3

Systemic and local loss of either RANKL or VEGF results in defective autograft healing. Mice received untreated autografts followed by injections of control IgG (a), RANK:Fc (b), or anti-VEGF (c) therapy, and were killed four weeks later. Representative hematoxylin and eosin–stained sections from these mice show a reduction in the amount of new bone formation around the autographs (arrow heads) and persistence of cartilage (blue in b,c). Representative radiographs from mice that received autografts transduced with rAAV-β-gal (e) or a combination of rAAV-OPG and rAAV-sFlt1 (f), 2 weeks after fracture are shown. Histomorphometry of the area of new bone formation on the autografts (d,g). *P < 0.05 compared to the IgG or rAAV-β-gal controls.

Figure 4

Transduction efficiency of rAAV-β-gal following freeze-drying onto allografts and implants in vitro and in vivo. 5 × 10⁷ transducing units of rAAV-β-gal was lyophilized onto mouse femoral allografts (a) or stainless steel pins (b). The transduction efficiency was determined in vitro by incubating the coated pins on top of a monolayer of confluent 293 human embryonic kidney cells for 72 h. Photographs of the X-gal-stained cells distal (c) and proximal (d) to the coated pin, as well as an uncoated control pin (e) are shown. The transduction efficiency of the coated pins was also quantified after the indicated storage time at −80 °C. RLU, relative light units. (f). As a control, 5 × 10⁷ transducing units of rAAV-β-gal in 50 μl PBS was directly placed on a monolayer of 293 cells. The β-galactosidase activity in the cultures was determined using the Galacto-Light system. No significant differences were observed. The efficiency of in vivo transduction 14 d after transplantation is shown at ×10 (g) and ×40 (h) magnification, where the blue staining indicates transduction of the fibroblasts (f) between the allograft (a) and the muscle (m).

Figure 5

Revitalization of processed allografts via rAAV mediated-RANKL and VEGF gene transfer. Allografts containing 5 × 10⁷ particles of rAAV-β-gal, rAAV-RANKL, rAAV-VEGF or a combination of rAAV-RANKL and rAAV-VEGF were transplanted into mice and evaluated 28 d after surgery. In vivo VEGF expression was analyzed in sera taken from the combined coated allografts at the indicated time after surgery (a). VEGF levels in uncoated allografts were consistently >50 pg/ml throughout the time course. Representative histology from the medial segment of the lateral cortex of a rAAV-β-gal (b) and rAAV-RANKL + rAAV-VEGF (c) coated allograft on day 28. Of note is the considerable amount of new bone on the rAAV-RANKL + rAAV-VEGF–coated allograft highlighted by a reversal line (arrows) and its similarity in cortical thickness to the rAAV-β-gal coated allograft. The new bone that formed on the allografts was quantified by histomorphometry (d) and the data are presented as the area of new bone formation on the graft ± s.d. (*P < 0.05 versus β-gal control).

Figure 6

rAAV-mediated gene transfer of RANKL and VEGF induces cortical bone resorption, vascularization and remodeling in processed allografts in vivo. Representative TRAP-stained histologic sections from mice in the combination group (a–d). An example of a rAAV-VEGF + rAAV-RANKL–coated allograft in which remodeled bone extends the entire length of the graft (arrow heads) is shown (a). The novel histologic features of the combination group were characterized by osteoclastic resorption of the necrotic bone (black arrows in b,d,g), osteoblastic new bone formation in the resorption lacunae (white arrows in c,d) and osteoclastic remodeling of the new woven bone (yellow arrows in d). Hematoxylin and eosin–stained sections of allografts from the combination group revealed asymmetric reversal lines (dashed line in e and shown at higher magnification without the lines in g, and black arrow in c) between dead bone and newly formed live bone, new blood vessel formation inside the marrow cavities (* in e,f), and active tunneling resorption (arrows in f) in the necrotic bone. In contrast, none of the other groups showed these features and were all characterized by a fibrotic tissue (f) that covered the periosteal surface and necrotic tissue (n) that filled the marrow cavity (h).