Marrow stromal cell-based cyclooxygenase 2 ex vivo gene-transfer strategy surprisingly lacks bone-regeneration effects and suppresses the bone-regeneration action of bone morphogenetic protein 4 in a mouse critical-sized calvarial defect model.

This study evaluated whether the murine leukemia virus (MLV)-based cyclooxygenase-2 (Cox-2) ex vivo gene-transfer strategy promotes healing of calvarial defects and/or synergistically enhances bone morphogenetic protein (BMP) 4-mediated bone regeneration. Gelatin scaffolds impregnated with mouse marrow stromal cells (MSCs) transduced with MLV-expressing BMP4, Cox-2, or a control gene were implanted into mouse calvarial defects. Bone regeneration was assessed by X-ray, dual-energy X-ray absorptiometry, and histology. In vitro, Cox-2 or prostanglandin E(2) enhanced synergistically the osteoblastic differentiation action of BMP4 in mouse MSCs. In vivo, implantation of BMP4-expressing MSCs yielded massive bone regeneration in calvarial defects after 2 weeks, but the Cox-2 strategy surprisingly did not promote bone regeneration even after 4 weeks. Staining for alkaline phosphatase (ALP)-expressing osteoblasts was strong throughout the defect of animals receiving BMP2/4-expressing cells, but defects receiving Cox-2-expressing cells displayed weak ALP staining along the edge of original intact bone, indicating that the Cox-2 strategy lacked bone-regeneration effects. The Cox-2 strategy not only lacked bone-regeneration effects but also suppressed the BMP4-induced bone regeneration. In vitro coculture of Cox-2-expressing MSCs with BMP4-expressing MSCs in gelatin scaffolds reduced BMP4 mRNA transcript levels, suggesting that Cox-2 may promote BMP4 gene silencing in BMP4-expressing cells, which may play a role in the suppressive action of Cox-2 on BMP4-mediated bone formation. In summary, the Cox-2 ex vivo gene-transfer strategy not only lacks bone-regeneration effects but also suppresses the bone-regeneration action of BMP4 in healing of calvarial defects.

Stem cell antigen-1+ cell-based bone morphogenetic protein-4 gene transfer strategy in mice failed to promote endosteal bone formation.

BACKGROUND: This study assessed whether a Sca-1+ cell-based ex vivo gene transfer strategy, which has been shown to promote robust endosteal bone formation with a modified fibroblast growth factor-2 (FGF2) gene, can be extended to use with bone morphogenetic protein (BMP)2/4 hybrid gene. METHODS: Sublethally irradiated recipient mice were transplanted with lentiviral (LV)-BMP2/4-transduced Sca-1+ cells. Bone parameters were monitored by pQCT and microCT. Gene expression was assessed by the real-time reverse transcriptase-polymerase chain reaction. RESULTS: Recipient mice of LV-BMP2/4-transduced Sca-1+ cells yielded high engraftment and increased BMP4 mRNA levels in marrow cells; but exhibited only insignificant increases in serum and bone alkaline phosphatase activity compared to control mice. pQCT and microCT analyses of femurs showed that, with the exception of small changes in trabecular bone mineral density and cortical bone mineral content in LV-BMP2/4 mice, there were no differences in measured bone parameters between mice of the LV-BMP2/4 group and controls. The lack of large endosteal bone formation effects with the BMP4 strategy could not be attributed to ineffective engraftment or expansion of BMP4-expressing Sca-1+ cells, an inability of the transduced cells to secrete active BMP4 proteins, or to use of the LV-based vector. CONCLUSIONS: Sca-1+ cell-based BMP4 ex vivo strategy did not promote robust endosteal bone formation, raising the possibility of intrinsic differences between FGF2- and BMP4-based strategies in their ability to promote endosteal bone formation. It emphasizes the importance of choosing an appropriate bone growth factor gene for delivery by this Sca-1+ cell-based ex vivo systemic gene transfer strategy to promote bone formation.

Modifications of the fibroblast growth factor-2 gene led to a marked enhancement in secretion and stability of the recombinant fibroblast growth factor-2 protein.

Progress in FGF-2 gene therapy has been hampered by the difficulty in achieving therapeutic levels of FGF-2 secretion. This study tested whether the addition of BMP2/4 hybrid secretion signal to the FGF-2 gene and mutation of cys-70 and cys-88 to serine and asparagine, respectively, would increase the stability and secretion of active FGF-2 protein in mammalian cells using MLV-based vectors. Single or double mutations of cys-70 and cys-88 to ser-70 and asp-88, respectively, markedly increased the amounts of FGF-2 protein in conditioned media and cell lysates, which may be due to glycosylation, particularly at the mutated asp-88 residue. Addition of BMP2/4 secretion signal increased FGF-2 secretion, but also suppressed FGF-2 biosynthesis. The combination of BMP2/4 secretion signal and double cys-70 and cys-88 mutations increased the total amount of secreted FGF-2 protein >60-fold. The modifications did not alter its ability to stimulate cell proliferation and Erk1/2 phosphorylation in marrow stromal cells or its ability to bind heparin in vitro, suggesting that the modified FGF-2 protein was functionally as effective as the unmodified FGF-2. An ex vivo application of rat skin fibroblasts (RSF) transduced with the modified FGF-2 vector in a subcutaneous implant model showed that rats with implants containing cells transduced with the modified FGF-2 vector increased serum FGF-2 level >15-fold, increased growth of the implant, and increased vascularization within the implant, compared to rats that received implants containing beta-galactosidase- or wild-type FGF-2-transduced control cells. This modified vector may be useful in FGF-2 gene therapy investigations.