885. VEGF Enhances Bone Formation and Bone

BONE AND CARTILAGE 7 after PPE cDNA gene-gun delivery (ICI 42.8 % decrease). This analgesic effect induced by PPE gene gun treatment was reversed by intravenous naloxone, an opioid antagonist (5mg/kg). Increased enkephalin immunoreactivity in bladder was observed from gene gun treated group at day 4, which was reduced at day 7. Conclusions: The PPE gene can be effectively transferred in the bladder using gene-gun, and increased expression of enkephalin in bladder can suppress nociceptive responses. These results support potential clinical application of PPE gene-gun delivery system for the treatment of bladder pain and other types of visceral pain.

885. VEGF Enhances Bone Formation and Bone Healing Elicited by Transduced MuscleDerived Stem Cells Expressing Human BMP2 Hairong Peng,1,2 Arvydas Usas,1,2 Brian Gearhart,2 Johnny Huard.1,2 1 Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, PA; 2 Growth and Development Laboratory, Children’s Hospital of Pittsburgh, Pittsburgh, PA. Introduction: Bone formation and bone healing require concerted interactions among many different pathways, including osteogenesis and angiogenesis. Previous studies have shown that the angiogenic factor VEGF is important for both normal bone development and BMP4-induced bone formation. More importantly, VEGF can synergistically enhance the bone formation and bone healing elicited by BMP4. The aim of this study was to determine whether VEGF also can enhance the bone formation and bone healing elicited by BMP2, another widely studied BMP. Methods: Muscle-derived stem cells (MDSCs) were transduced with retroviral vectors expressing BMP2 or VEGF. The expression of BMP2 by the retro-BMP2-transduced cells was determined by Western blot and BMP bioassay. The level of VEGF secreted by the VEGF-transduced cells was measured using ELISA. The effect of VEGF on BMP2-induced bone formation was tested by implantation of the transduced cells into the thigh muscles of syngeneic C57BL/ 6J mice. We implanted the same number of cells in all the groups. The effect of VEGF on BMP-induced bone healing was tested by implantation of the cells into critical-sized calvarial defects created in 6J mice. Bone formation and bone healing were monitored radiographically and histologically at designated time points after surgery. Results: Our in vitro studies showed that the BMP2-transduced MDSCs secreted approximately 250 ng BMP2/105 cells/24 h. MDSCs transduced with a retroviral vector expressing VEGF secreted approximately 200 ng of VEGF/105 cells/24h. Neither untransduced cells nor cells transduced with a control vector expressing LacZ secreted detectable levels of either BMP2 or VEGF. Our findings at 2 weeks after implantation and at later time points revealed more bone formation at the sites implanted with both MDSCs expressing BMP2 and MDSCs expressing VEGF than at the sites treated only with MDSCs expressing BMP2. We detected no bone formation at the sites implanted only with MDSCs expressing VEGF. The co-implantation of MDSCs expressing VEGF also enhanced the healing of critical-sized calvarial defects treated with MDSCs expressing BMP2; this improved healing was detectable beginning at 3 weeks after implantation. The transplantation of only MDSCs expressing VEGF did not elicit bone healing. Conclusions: As demonstrated via ex vivo gene therapy mediated by MDSCs, VEGF can enhance the bone formation and bone healing elicited by BMP2. Thus, the synergistic interaction between angiogenic factors (e.g., VEGF) and osteogenic factors is not unique to BMP4-induced bone formation and healing. The use of musclederived stem cell–mediated gene therapy for delivery of BMP2 and VEGF should have great potential for improving bone repair in a variety of orthopaedic conditions. Dr. Huard is the co-founder of CookMyosite. Molecular Therapy Volume 9, Supplement 1, May 2004 Copyright  The American Society of Gene Therapy

886. Determination of Osteoprogenitor-Specific Promoter Activity in Murine Mesenchymal Stem Cells by Recombinant Adeno-Associated Virus Transduction Sanjay Kumar,1 Gandham Mahendra,1 Selvarangan Ponnazhagan.1 1 Pathology, The University of Alabama at Birmingham, Birmingham, AL. The strength and integrity of metabolically active bone depends on the balanced activity of two different cell types, namely osteoclasts, which resorbs the bone and osteoblasts which form new bone. Loss of this homeostasis results in a decrease in bone mass, leading to defects in mineralization of bone, resulting in osteopenic conditions including osteoporosis. Ex vivo gene therapy for osteoporosis is a rational alternative to the existing therapies and the potential of rAAV with long-term expression capability should provide greater benefit in this regard. A potential osteoporosis gene therapy is by using an ex vivo approach with gene modified mesenchymal stem cells (MSC) expressing osteogenic agents, which provide a local supply of precursor cells and a supraphysiological dose of osteoinductive molecules, which may promote bone formation. Towards achieving this goal with rAAV, capable of long term expression, we determined osteoprogenitor-specific expression of rAAV-transduced bone marrow stromal cells in vitro. rAAV vectors containing luciferase gene under the control of CMV, alkaline phosphatase, cbfa-1, collagen type-I, osteopontin and osteocalcin promoters were transduced to mouse MSC in replicates in vitro. Following transduction, cells were either maintained as undifferentiated cultures or induced to undergo early differentiation by the addition of 10 mM β-glycerophosphate, 50 µM ascorbic acid and 100 ng/ml recombinant BMP-2. Luciferase activity was determined at different time points as a measure of promoter activation. Results of these studies indicated that there was no significant difference in CMV activity between undifferentiated and BMP-2 differentiated MSC at different time points. However, luciferase activity from alkaline phosphatase, cbfa-1, collagen typeI, osteopontin and osteocalcin promoters in BMP-2 induced MSC cultures indicated increase in transgene expression over time, whereas no significant increase was observed in undifferentiated cultures. Among the bone-specific promoters tested, expression of collagen type-I and cbfa-1 promoters were more pronounced during early differentiation (48 hours) compared to osteocalcin, alkaline phophatase and osteopontin promoters. Northern blot analysis was done with RNA obtained from BMP-2 differentiated MSC at different time points to confirm BMP-2 induced osteoblastic changes in cultured and differentiated MSC. Results of these studies indicate that by using bone specific promoters that are active during early differentiation of MSC, lineage-restricted expression of therapeutic proteins is possible to enhance osteoblast production for increased bone mass.

887. Bone Formation by Ex Vivo Gene Therapy Using Lentiviral Vector Expressing BMP-2 under the Control of the MLV Promoter Osamu Sugiyama,1 Dong Sung An,2 Seth Gamradt,1 Nancy Q. Liu,1 Irvin S. Y. Chen,2 Jay R. Lieberman.1 1 Orthopaedic Surgery, UCLA David Geffen School of Medicine, Los Angeles, CA; 2UCLA AIDS Institute, UCLA David Geffen School of Medicine, Los Angeles, CA. Objective: Bone morphogenetic protein-2 (BMP-2) is an osteoinductive growth factor that can induce bone formation in vivo. Bone marrow stromal cells (BMSCs) are attractive targets for ex vivo gene therapy, and have been reported to differentiate into osteoblasts in the presence of BMP-2. Although we previously demonstrated that ex vivo gene therapy using an adenovirus S337

BONE AND CARTILAGE expressing the BMP-2 induced bone formation successfully in animal models, the transient protein production may not be of adequate duration to heal large bone defects. Lentivirus is known to infect both dividing and non-dividing cells and induce long-term transgene expression. The objective of this study is to examine the potential of ex vivo gene therapy to enhance bone repair using a lentiviral vector. Methods: Primary rat BMSCs were isolated from the femurs and tibias of 4 to 8 week-old female Lewis rats and maintained in culture. Rat BMSCs were transduced with the lentiviral vector encoding enhanced green fluorescent protein(EGFP), which had either an MLV promoter or a CMV promoter (Lenti-MLV-EGFP or LentiCMV-EGFP). The mean fluorescent intensity (MFI) of EGFP expression was detected in cells at 1, 2, 4, 6 and 8 weeks after transduction by flow cytometry. Next, we constructed the lentiviral vector encoding human BMP-2 cDNA, which had either an MLV promoter or a CMV promoter (Lenti-MLV-BMP-2 or Lenti-CMVBMP-2). To compare in vitro BMP-2 production induced by these vectors, rat BMSCs were transduced and cultured. Cultured medium of cells were harvested at 1, 2, 4, 6 and 8 weeks after transduction, and quantified BMP-2 production during a 24-hour period by enzyme linked immunosorbent assay (ELISA). Finally, 2.5x10 6 rat BMSCs transduced with Lenti-MLV-BMP-2 or Lenti-CMV-BMP2 were impregnated on to a collagen sponge and implanted into a muscle pouch in hindlimbs of severe combined immune deficient (SCID) mice. Animals were sacrificed at 3 weeks after implantation for radiographic and histological examinations. Results: Flow cytometric analysis showed significantly higher MFI in Lenti-MLV-EGFP-transduced cells than in Lenti-CMVEGFP-transduced at each time point. Although the BMP-2 production from Lenti-CMV-BMP-2-transduced cells did decrease over time on ELISA, Lenti-MLV-BMP-2-transduced cells continued to secrete BMP-2 eight weeks after transduction. Ectopic new bone formation was revealed on plain radiographs in the hindlimbs of mice implanted with BMP-2 transduced cells. Increased bone formation was noted with implantation of Lenti-MLV-BMP-2transduced cells in comparison to Lenti-CMV-BMP-2-transduced cells. Histology revealed bony trabeculae and mature osseous components within the muscle pouch of BMP-2-transduced cells. Conclusions:These results demonstrate that lentiviral vectors expressing BMP-2 can induce long term transgene expression in vitro and new bone formation in vivo under the control of the MLV promoter. This may be advantageous when developing tissue engineering strategies to repair large bone defects.

888. A BMP2/7 Fusion Gene Enhances Osteoblastic Differentiation in Mesenchymal Cells Wei Zhu,1 Oheneba Boachie-Adjei,1 Bernard Rawlins,1 Ronald G. Crystal,2 Chisa Hidaka.1 1 Hospital for Specical Surgery, New York, NY; 2Weill Medical College of Cornell University, New York, NY. While recombinant bone morphogenic proteins (rBMPs) are homodimers, several studies have suggested that heterodimeric BMPs resulting from the co-expression of two different BMP genes might be more potent. Previously, we have shown that co-transfection of target cells with adenovirus (Ad) vectors encoding BMP2 or BMP7 results in enhanced osteoblastic differentiation of mesenchymal cells in vitro and bone formation in vivo when compared to either vector alone. In the current study, we constructed a novel BMP2/7 fusion gene encoding both BMP2 and BMP7 fused by a linker, and show that transfection with this gene results in the expression of biologically active BMP2/7 heterodimer. A BMP2/BMP7 fusion gene was constructed so that the stop codon of BMP2 cDNA and signal peptide of BMP7 cDNA were replaced by a (Gly4Ser)4 linker. This fusion gene was cloned into a S338

plasmid with expression driven by a CMV promoter (pSCMVBMP2/7), so that a single transcript encoding BMP2, the linker, and BMP7 would be expressed. To detect the presence of BMP2/7 heterodimer, pSCMV-BMP2/ 7 and a control plasmid encoding the marker gene of green florescent protein (pCMV-GFP) were transfected into A549 cells using polyfection. After immunoprecipitation with anti-BMP2 antibody, supernatants of pSCMV-BMP2/7 transfected cells, but not controls, were positive for BMP7 as measured by enzyme-linked immunosorbant assay (ELISA), indicating the presence of BMP2/7 heterodimer. Results were similar when anti-BMP7 immuoprecipitation was followed by BMP2 ELISA. To assess BMP2/7 bio-activity, C2C12 myoblastic cells were stimulated with A549 supernatant of pSCMV-BMP2/7-transfected cells at dilutions ranging from 1:80 to 1:1, and markers of osteoblastic differentiation including osteocalcin (OCN) and alkaline phosphatase (ALP) were evaluated. As controls, C2C12 cells were also stimulated with commercially available recombinant human (rh) BMP2 or rhBMP7 at 1000 ng/ml, pCMV-GFP transfected A549 supernatant at 1:1 dilution and untransfected A549 supernatant. After 5 days, supernatant of pSCMV-BMP2/7-transfected cells increased OCN expression and ALP activity in C2C12 cells at a dose-dependent manner. The level of OCN expression induced by a 1:5 dilution of supernatant of pSCMV-BMP2/7-transfected cells (containing approximately 6 ng/ml BMP2/7) was comparable to that achieved by 1000 ng/ml of rhBMP2 or rhBMP7. Moreover, ALP activity induced by supernatant of pSCMV-BMP2/7-transfected cells at this dose was 2- fold higher than that achieved by rhBMP2 or rhBMP7 at 1000 ng/ml. As expected, no OCN or ALP level was detected in control groups stimulated with supernatants of cells transfected by pCMV-GFP or of untransfected cells. Our study shows that a BMP2/7 fusion gene construct can result in the production of a biologically active BMP2/7 heterodimer. This study may potentially expand the future clinical application of BMP gene therapy.

889. Relationship between the Amount of Bone Formation and Protein Expression in Two Rat Stains Using Direct Adenoviral Vector Injection and Ex Vivo Method Jin Zhong Li,1 Hongwei Li,1 El Noh,1 Lindsey Jacobson,1 AnHsiang Liu,1 Dwight Saulle,1 Gerald R. Hankins,1 Gregory A. Helm.1 1 Neurosurgery, University of Virginia, Charlottesville, VA. Although adenoviral vectors have a very broad range and high efficiency of cell transduction both in vitro and in vivo, the uses of adenoviral vectors have been significantly limited in immunocompetent animals. We tested whether the amount of protein expression was the major factor determining the foreign gene functions. The recombinant adenoviral vectors of human bone morphogentic protein 4 and luciferase were used to answer this question. Direct viral injection and ex vivo methods were chosen in immunodefective athymic nude (AN) and immunocompetent SD rats. On Day 0, the primary rat bone marrow cells were infected with either ADhBMP4 or ADLUC at 107 PFU/flask or PBS. At the same time, five AN and SD rats were given a direct injection of either ADhBMP4 or ADLUC (107 PFU/100 µl) into the right thigh. On Day 1, both control and virally infected cells were injected into the left thigh with 2 ×106 cells/100 µl/flask. The rats underwent imaging performed with a cooled CCD before and after an intraperitoneal luciferin injection on Days 1 and 7. On Day 1, the densities of luminescent signals (×107 photons/sec) were the following: 1.83 ± 0.56 for SD rats with direct ADLUC injection; 3.73 ± 1.15 for SD rats with ex vivo ADLUC injection; 5.01 ± 1.91 for AN rats with direct ADLUC injection; and 11.9 ± 3.35 for AN Molecular Therapy Volume 9, Supplement 1, May 2004

Copyright  The American Society of Gene Therapy