794. Retroviral-Based Gene Therapy with Cyclo-Oxygenase-2 Promotes Bony Union and Accelerates Fracture Healing in the Rat

Abstract

Fracture repair of endochondral bones involves an ordered series of processes, including the proliferation of fibrous tissues, their replacement by cartilage, and the conversion of cartilage to bony tissue that ultimately bridges the fracture. While recent gene therapy studies with osteogenic genes, e.g., BMPs, yielded encouraging results in that they augmented bone production at the fracture site, none of the published strategies to date have accelerated bony union of the fracture gap, which is the hallmark of successful fracture repair. In this study, we evaluated the efficacy of cyclo-oxygenase-2 (Cox-2) gene therapy in promoting bony union in a rat femur fracture model. The Cox-2 gene was chosen because Cox-2 promotes inflammation, angiogenesis, bone formation and bone resorption; all of these processes are essential for fracture repair. However, previous gene therapy with Cox-2 has been limited by the short half-life of Cox-2 mRNA, a result of the destabilizing ARE domains of the 3|[prime]| untranslated region (3|[prime]|UTR). Thus, we modified the human Cox-2 transgene by deleting its 3|[prime]|UTR and by including an optimized Kozak sequence to increase the stability of Cox-2 mRNA and to enhance Cox-2 protein translation, respectively. A murine leukemia virus (MLV)-based retroviral vector was used to target the expression of the Cox-2 transgene or a |[beta]|-galactosidase control transgene to the cells of the periosteum induced to proliferate by the injury. The retroviral vector was injected directly into the periosteum at the fracture site in a single percutaneous injection at one day post-fracture. Healing was evaluated by X-ray analysis and bone histomorphometry. Real-time RT-PCR analysis of human Cox-2 gene expression in Cox-2 and control fracture tissues at 4, 7, 14 and 21 days healing (N=4 each) established that human Cox-2 transgene expression significantly increased total Cox-2 gene expression at each time point, and greater than 10-fold at 21 days (p<0.03). Endogenous rat Cox-2 gene expression was unaffected by the treatment. X-ray and histology analysis revealed that, unlike BMP therapy, Cox-2 gene therapy produced a symmetric callus around the entire circumference of the fracture without supra-periosteal bone formation. More importantly, 7 of 8 MLV-Cox-2 treated fractures yielded bony union of the fracture callus at 21 days, while only 1 of 6 MLV-|[beta]|-galactosidase treated fractures showed evidence of bony union. This healing occurred well before the 4 to 5 weeks normally required for bony union in this fracture model. Histomorphometric measurements demonstrated that bony union coincided with a reduction in the cartilage area from 3.7% to 0.6% (p<0.005) within the fracture site at 21 days healing, indicating that Cox-2 gene therapy accelerated the replacement of the fracture callus cartilage with bone to generate earlier bony union of the fracture. Sinusoid-like structures in Cox-2 treated fractures suggested that Cox-2 therapy promoted angiogenesis. In summary, we have demonstrated for the first time that local expression of a Cox-2 transgene from a retroviral vector accelerated bony union in a normally healing fracture model.