Cell-Based Osteoprotegerin Therapy For Debris-Induced Aseptic Loosening Using A Murine Model

PARTⅠMURINE MODEL OF PROSTHESIS FAILURE CAUSED BY WEAR DEBRISBackgroundTotal joint arthroplasty has become the most successful and effective procedure in the treatment of severe arthritis. Unfortunately, there is relatively high incidence of late failure in cemented and cementless joint replacements although the materials and design have always been improved. However, treatment of the debris associated loosening process is highly problematic, partially due to the difficulty of administering and maintaining effective dosages of therapeutic agents at the site of loosening.Although the precise pathogeneses of aseptic loosening is unclear, cumulative evidence indicates that particulate biomaterial debris generated from the mechanical wear of prosthetic components plays a critical role. Since polymeric and metallic wear particles phagocytosed by macrophages are impervious to enzymatic destruction, these particles frustrate the degenerative function of these cells. Repeated phagocytosis of particles results in activated cells producing proinflammatory cytokines and proteolytic enzyme, which are known to both damage bone and cartilage, and activate more immune cells. Unchecked production of cytokines can lead to persistent stimulation of the inflammatory response, causing chronic inflammation and tissue damage, such as that associated with periprosthetic osteolysis. So it has become an important pathway to solve the aseptic loosening caused by wear debris.However, treatment of the debris-associated loosening is highly problematic, partially due to the difficulty of administering and maintaining effective dosages of therapeutic agents at the loosening joints. When anti-inflammation or anti-osteolysis drugs are administered systemically, their effects at the bone-implant interface rely on vascular perfusion, suggesting that relatively high systemic levels are required to achieve anti-inflammatory activity at the site of loosening process. High systemic drug levels often induce adverse side affects and subsequent poor patient compliance. Gene therapy, though still in its infancy, provides an attractive alternative to overcome this difficulty.To investigate the mechanisms underlying aseptic loosening and explore novel therapeutic approaches, the development of an appropriate animal model is an essential prerequisite. We have developed a murine model of air pouch and obtain some important findings using this model including the effects of OPG gene therapy against wear debris-induced oteolysis. The limitations of the model, such as the short-term characteristics and the absence of a prosthetic implant, prevent it from use to study biomechanical properties of implant stability and long-term effectiveness and safety issues of therapeutic interventions for aseptic loosening. The recent animal models for implant-bone/implant-cement bonding or long-term studies of aseptic loosening pathogenesis typically involve the use of large animals such as dogs, sheep and rabbits. Although these models have the advantages of large bone/joint size and the ability to use a realistic implant, the cost, feasibility and ethical issues prevent broad employment of large animals, especially in the screening studies to explore the therapeutic approaches.In this study, we intend to establish a novel murine model of knee joint replacement by implanting a titanium pin with particles before insertion of the pin, to set up the true model of aseptic loosening. Since LacZ gene has become the most popular target gene to detect vector transduction efficacy, we use LacZ gene to investigate the feasibility of gene therapy in the new murine model.ObjectiveTo investigate the feasibility of establishing murine model of prosthesis failure, to characterize the biomechanical and pathological aspects of loosening process and to detect the feasibility of gene therapy as a potential alternative avenue to control this common complication using the new model.Materials and methods1. Animals 54 female BALB/c mice at eight weeks of age were used as hosts for the model. All mice weighed were divided into three groups:Ti-particle group (18 mice), the control group (18 mice) and LacZ group (18 mice). All mice were quarantined for one week prior to experimentation.2. Establish the Pin-model Under aseptic condition, the tibia plateau was exposed via the left media parapatellar approach and a cavity for the implant was reamed with a 0.7 mm dental drill through the center of the tibia plateau and the cavity was about 5mm in depth. A titanium pin was press-fitted into the canal with 10μl titanium particles solution injection before the insertion of the pin in Ti-particle group and 20μl particles solution injection for every month. No particles but PBS was used for the control group.Rinse the wound with PBS containing penicillin and streptomycin, and close the wound by sutures in layers.The mice were sacrificed at 3,7 and 12 weeks for biomechanical and histological evaluation.3. Construct retroviral vectors and gene transduction Adeno-associated virus coding for the LacZ gene (AAV-LacZ) was used to examine the feasibility of in vivo gene transfer using this model. At 3 weeks after pin-implantation surgery,50μl of sterile culture medium containing 5×108 cfu of AAV-LacZ was injected into each prosthetic joint of the twelve mice and the virus-free medium was injected as control. LacZ transgene expression in the prosthetic joint was detected using X-gal staining at 7 and 12 weeks after surgery.4. Micro-Computerized Tomography (Micro CT) Scans All mice were scanned immediately following surgery using Micro CT system to confirm the proper position of pin implantation. Following acquisition and reconstruction, the image data were collected at 3,7 and 12 weeks post operation to calculate the bone mineral densities (BMD) of the tibia bone surrounding the titanium pin. 5. Pull-out test At 3,7 and 12 weeks after surgery, the mouse limb with the implant intact was removed by disarticulating the knee joint following sacrifice. All soft tissue around the prosthetic joint was carefully removed to expose the implanted pin surface and proximal tibia. The distal tibia was cemented into a custom-designed jig using dental cement. The head of the pin was secured between the razor blades of the holding jig. After the fixture was attached to the loading cell, the MTS actuator pulled the implant out of the bone at a rate of 2.0 mm/minute under displacement control. Actuator positions and loading data were recorded.6. Histological and Immunohistochemical (IHC) Analyses Following the pull-out test, the peri-implant proximal tibiae were formalin-fixed and decalcified using 12% EDTA before paraffin-embedding. Sections were stained with hematoxylin & eosin to examine new bone formation or bone erosion around the prosthetic pin and The thickness of the periprosthetic membrane was measured using Image-Pro software. Modified Masson staining was performed to examine bone collagen changes. Immunohistochemical stains were carried out to evaluate pro-inflammatory cytokines and mediators (IL-1, TNFαand CD68) of osteoclastogenesis in periprosthetic tissues. X-gal staining was employed to trace LacZ gene transduced cells.7. Statistical Analyses Statistical analysis between different groups was performed by Student T-test, or the ANOVA test; with the S chafer formula for post hoc multiple comparisons, using the SPSS software package. A p value of less than 0.05 was considered as significant difference. Data are expressed as mean±standard deviation.Results1. Surgical outcomes The mice tolerated the surgical procedure well and ambulated with the implanted limb within 3 days after surgery. The wound healed in 7 days and the suture released with no signs of scratching or inflammation. There was no influence on daily activity between the three groups. Ti-particles were examined near the prosthesis and knee with irregular distribution after knee dissection. There were no obvious structural differences between stable and particle-stimulated implantation groups by naked eyes.2. Micro CT scans Micro CT scans indicated that the implants were well fixed with no obvious migration up to 12 weeks after surgery in the control group. However, titanium particle injection induced marked periprosthetic bone resorption illustrates typical debris-associated aseptic loosening on CT imaging and a significant BMD loss was observed in 7 weeks (p<0.05) and 12 weeks (p<0.01).3. Implant Stability Tested by Pull-out Test The head of the pin was secured between the razor blades of the holding jig which was sufficiently strong to hold the titanium pin to the extraction instrumentation during pullout tests reaching at least 100 N forces. For the control group, the average peak interfacial shear strength against pulling was 24.12±3.3 N at 12 weeks after surgery and there was no statistical difference between 3,7, and 12 weeks after surgery. However, the introduction of titanium particles significantly decreased the implant stability at 7 weeks after surgery (p<0.05) and only 4.32±0.55 N of pulling force was required at 12 weeks (p<0.01).4. Histological and Immunohistochemical Assessment Histology showed a stable condition and irregular new bone formation in the control group. The introduction of Ti-particles provoked the formation of periprosthetic inflammatory soft membrane and the thickness was higher than the control group in 7 weeks (p<0.01)and 12 weeks (p<0.01). The periprosthetic bones stimulated with titanium particles exhibited much fainter blue color staining using Modified Trichrome staining. Using a computerized image analysis system, the IOD of Trichrome staining in debris-induced bone resorption group averaged 47.66±5.21%loss at 6 weeks (p<0.01) and 52.29±6.72% loss at 12 weeks (p<0.01) compared with the control.The time study of pin-implantation with monthly titanium particle injections to mimic debris-associated loosening process indicated a continuous inflammatory cellular infiltration and periprosthetic membrane formation, starting at 7 weeks following surgery. Immunohistochemical assessment using anti-mouse cytokine antibodies revealed a profound accumulation of TNFa and IL-1 expressing cells in the particle-stimulated sections. CD68+macrophages were also present in marked aggregations in particle-stimulated periprosthetic membranes.5. In vivo LacZ gene transfer detection X-gal staining revealed that a direct single injection of AAV-LacZ into the implanted joint resulted in strong transgene expression, indicated by strong blue or green coloration in the synovial membranes and periprosthetic tissue. tissue. In contrast, the joints receiving virus-free medium injections remained negative using this staining.Conclusions1. We have successfully established a murine Pin-model of aseptic loosening and represented how the wear debris induced prosthetic failure.2. The time study of pin-implantation with monthly titanium particle injections to mimic debris-associated loosening process indicated a continuous inflammatory cellular infiltration and periprosthetic membrane formation.3. The successful in vivo LacZ gene transfer showed the feasibility of gene therapy in the model. PARTⅡCELL-BASED OSTEOPROTEGERIN GENE THERAPY FOR DEBRIS-INDUCED ASEPTIC LOOSENING USING A MURINE MODELBackgroundTotal joint arthroplasty is a highly successful procedure in the treatment of end stage of arthritis. While failure due to infection and operative error has become relatively rare, osteolysis-associated aseptic loosening (AL) has become more common and important. Cumulative evidence indicates that particulate biomaterial debris generated from the mechanical wear of prosthetic components plays a critical role among other factors. The wear debris was engulfed by macrophage and other cells, resulting in cellular activation and release of pro-inflammatory mediators and cytokines such as interleukin-1 (IL-1β), tumor necrosis factor (TNFα), and IL-6. These cytokines in turn provoke cellular proliferation, promote osteoclastogenesis, and stimulate mature osteoclasts to absorb the adjacent bone. This process impacts bone remodeling around the implant and leads to osteolysis and aseptic loosening.The OPG/RANKL/RANK pathway plays a key role in bone metabolism and osteolysis. RANKL (receptor activator of nuclear factor NF-κB ligand) is mainly expressed by macrophages, osteoblasts, marrow stromal cells, and lymphocytes. RANKL binds to RANK, a receptor on the cell surface of osteoclasts and osteoclast precursors, resulting in proliferation, differentiation and maturation of osteoclasts, which can subsequently promote local osteolysis.RANKL (receptor activator of nuclear factor NF-κB ligand) is mainly expressed by macrophages, osteoblasts, marrow stromal cells, and lymphocytes. RANKL binds to RANK, a receptor on the cell surface of osteoclasts and osteoclast precursors, resulting in proliferation, differentiation and maturation of osteoclasts, which can subsequently promote local osteolysis. In contrast, osteoprotegerin (OPG), a secreted protein with homology to members of the TNF superfamily, is considered to be a natural decoy receptor that profoundly modifies the effects of RANKL by inhibiting RANKL/RANK interaction, so it can prevent osteoclasts from differentiation and maturation.Based on the anti-osteolytic nature of OPG, it appears to be a potential therapeutic agent to treat debris-associated periprosthetic bone resorption and aseptic loosening. However, the delivery of OPG to chronic periprosthetic osteolytic sites remains problematic. Current systemic anti-inflammation or anti-osteolysis therapies have several weaknesses including high dose requirements with modest efficiency, systemic side effects and the tendency for poor patient compliance. Gene therapy, though still in its infancy, provides an attractive alternative to treat aseptic loosening since the transfer of genes may facilitate the continuous release of therapeutic proteins. We have reported that in vivo gene transfer of OPG effectively blocked osteoclastogenesis and reversed periprosthetic bone resorption in a mouse model of knee prosthesis failure.To better control the efficacy of the transgene expression in comparison with the direct in vivo gene transfer, the current study evaluated the potential therapeutic effects of a cell-based OPG gene therapy for aseptic loosening and osteolysis.Objective1. To investigate the in vitro transduction efficiency of adeno-associated virus encoding OPG to fibroblast-like synoviocytes (FLS).2. To evaluate the feasibility and effect of cell-based OPG gene therapy for aseptic loosening using the new model.3. To explore the mechanism of cell-base gene therapy.Materials and methods1. Fibroblast-like synoviocytes (FLS) Sacrifice 10 BALB/c mice by cervical dislocation, sterilize the lower limbs with alcohol and expose the knee joints. After careful removal of the skin and muscle, the tissue of the knee joints was minced, incubated with 1 mg/ml of collagenase in serum-free RPMI 1640 for 2h at 37℃, filtered through nylon mesh, and washed extensively. Cells were cultured in RPMI 1640 supplemented with 10% FBS and 1% penicillin/streptomycin in a humidified 5% CO2 atmosphere. After overnight culture, non-adherent cells were removed, and adherent cells were cultured. The cells were passaged by replating at a 1:5 dilution when the cultures reached confluence.2. Gene transfer into cells The adeno-associated virus coding for OPG or LacZ gene was used for in vitro gene transfer into the FLS.Cells were co-cultured with 107 particles/ml titer of AAV-OPG-EGFP or AAV-LacZ at 30-40% confluence, and a second dose of viral vectors at titer of 107 was added into culture 6 hours later. Transduction efficacy was determined by fluorescent microscopy (for the emission of GFP) and X-gal staining (for LacZ transduction).3. Establishment of the Pin-model Sixty BALB/c mice were divided into 5 groups: AAV-OPG, FLS-AAV-OPG, AAV-LacZ, FLS-AAV-LacZ and PBS control.All mice were quarantined for one week prior to experimentation. Under aseptic condition, the tibia plateau was exposed via the left media parapatellar approach and a cavity for the implant was reamed with a 0.7 mm dental drill through the center of the tibia plateau and the cavity was about 5mm in depth. A titanium pin was press-fitted into the canal with 10μl titanium particles solution injection before the insertion of the pin and 20μl particles solution injection for every month. Three weeks after surgery,50μl medium containing AAV-OPG or AAV-LacZ (106IU/ml),50μl medium containing FLS-AAV-OPG or FLS-AAV-LacZ (106/ml) and 50μl virus-free PBS was injected into the knee joint individually. The mice were sacrificed at 4 weeks after gene modification for biomechanical, histological and molecular evaluation.4. Pull-out test At 4 weeks after gene therapy, the mouse limb with the implant intact was removed by disarticulating the knee joint following sacrifice. All soft tissue around the prosthetic joint was carefully removed to expose the implanted pin surface and proximal tibia. The distal tibia was cemented into a custom-designed jig using dental cement. The head of the pin was secured between the razor blades of the holding jig. After the fixture was attached to the loading cell, the MTS actuator pulled the implant out of the bone at a rate of 2.0 mm/minute under displacement control. Actuator positions and loading data were recorded.5. Histological and Immunohistochemical (IHC) Analysis Following the pull-out test, the peri-implant proximal tibiae were formalin-fixed and decalcified using 12% EDTA before paraffin-embedding. Sections were stained with hematoxylin & eosin to examine new bone formation or bone erosion around the prosthetic pin and The thickness of the periprosthetic membrane was measured using Image-Pro software. Modified Masson staining was performed to examine bone collagen changes. Immunohistochemical stains were carried out to evaluate pro-inflammatory cytokines and mediators (IL-1, TNFαand CD68) of osteoclastogenesis in periprosthetic tissues. X-gal staining was employed to trace LacZ gene modification.6. Tartrate-resistant Acid Phosphatase (TRAP) Stain for Osteoclasts Following deparaffinization and rehydration, the sections were permeated in a microwave oven for 30 seconds. Fix and stain the slides with TRAP commercial kit. TRAP+cells in the periprosthetic membrane were detected by the presence of dark purple staining granules in the cytoplasm.7. OPG transgene and product detection Conventional reverse transcription polymerase chain reaction (RT-PCR) was performed to detect OPG gene expression at the local delivery site and the remote organs/tissues. The PCR product was electrophoresed on 1% agarose gels to confirm the amplification of OPG gene. The transgene product, human OPG protein levels in FLS-AAV-OPG culture medium and variant prosthetic joint tissue were determined by a quantitative sandwich enzyme immunoassay technique (ELIS A) utilizing capture and detection monoclonal antibody pairs against different epitopes of OPG8. Statistical Analysis Statistical analysis between different groups was performed by Student T-test, or the ANOVA test; with the Schafer formula for post hoc multiple comparisons, using the SPSS software package. A p value of less than 0.05 was considered as significant difference. Data are expressed as mean±standard deviation. Results1. Cell culture Cells were fusiform and polygon shaped, of unequal size. The cell shape and proliferation ability of the transfected FLS were the same as untransfected cells. No acute cytotoxicity in cell culture was noticed following viral infection.2. Gene transduction efficiency Fluorescent microscopy was performed to identify green fluorescent protein, expressed by the co-transduced EGFP gene from AAV-OPG-eGFP and the transduction efficiency of AAV-OPG in FLS was 93.1±2.2%. LacZ gene transduced cells showed dark blue by X-gal stain and the transduction efficiency of AAV-LacZ in FLS was 92.6±2.6%.3. Pull-out test The distal tibia was cemented into a custom-designed jig using dental cement. The head of the pin was secured between the razor blades of the holding jig which was sufficiently strong to hold the titanium pin to the extraction instrumentation during pull-out tests reaching at least 100 N forces. The average peak interfacial shear strength against pulling was 10.34±2.05 N in AAV-LacZ group, 8.14±1.23 N in FLS-AAV-LacZ group and 7.32±1.35 N in PBS group and there was no statistical difference between the three groups. However, the ex vivo and in vivo OPG gene therapies significantly increased the implant stability, requiring 21.56±2.44 and 18.19±2.10 N of pulling force respectively, to dissociate the pin implant from the bone. There was no statistical difference between the two OPG gene-modified groups.There was statistical difference among the OPG gene-modified groups and other three groups.4. Histological analysis Histological evaluation revealed that marked bone-implant-interface fibrous membranes were present in both FLS-AAV-LacZ and non-treated control groups. However, the interface membranes in sections from the FLS-AAV-OPG group were dramatically thinner. Both in vivo and ex vivo OPG gene transfer significantly blocked the periprosthetic membrane formation compared with other three groups (p<0.01).While the periprosthetic bone tissues from AAV-LacZ, FLS-AAV-LacZ and virus-free groups generally exhibited weak staining of Trichrome blue coloration, OPG-gene modification using in vivo or ex vivo delivery markedly enhanced the staining as seen previously. Quantified intensity analysis by computerized image analysis system confirmed the statistical significance (p<0.01).A profound accumulation of CD68+macrophages, and TNFαand IL-1 expressing cells were observed at the interface between the pin and surrounding bone in sections from both FLS-AAV-LacZ and virus-free non-treated groups. In contrast, significantly less positive cells were present on tissue sections from OPG gene modified groups.5. TRAP+cells The TRAP staining was conducted to identify osteoclast-like ceils in the model. There was also a marked reduction of TRAP+cells which showed dark purple color in the sections from OPG gene modified animals in comparison with LacZ-treated and virus-free groups (p<0.01).6. OPG gene and protein detection No positive PCR products were detected, except the periprosthetic tissues where the ex vivo gene transfer was conducted, where results were similar to in vivo OPG treatment. ELISA revealed OPG protein production in FLS-AAV-OPG culture medium was 4.20ng/72h/106cells, where OPG levels in peri-prosthetic tissues from ex vivo OPG-treated groups were 2.60ng/mg total protein at sacrifice,comparable with levels observed during in vivo OPG treatment.Conclusions1. Cell-based OPG can be successfully expressed in the Pin-model and the transgene product can reach the therapeutic concentration.2. Cell-based OPG gene transfer appears to be a feasible and effective candidate to treat or prevent the wear debris-induced osteolysis and aseptic loosening.3. OPG gene expression can increase the concentration of OPG in the tissue around prosthesis, block the activation of osteoclasts and reduce the expression of inflammatory cytokines.