Reconstruction of spinal stability after treatment of congenital spinal deformities, degeneration, acquired instability, and spinal tumors requires bone grafting to facilitate spinal fusion. Clinical statistics in the United States indicate that the majority of the more than 250,000 spinal fusion procedures performed each year require bone grafting to promote spinal fusion, and although exact statistics are not available for China, the actual situation may exceed that of the United States in terms of population percentage. Although autologous bone grafting remains the gold standard for promoting spinal fusion, bone nonunion still occurs in 5-35% of patients with single-segment spinal fusions, and the incidence of bone nonunion is even higher in multisegment spinal fusions. In addition, there are still many intractable problems with autologous bone, allogeneic bone, and bioreplacement material grafts. Therefore, there is an urgent need to further investigate the biological mechanisms of bone regeneration in order to find better ways to promote spinal fusion. The application of local gene therapy can release osteogenic transmitters in specific anatomical regions of the spine to stimulate the formation of new bone and thus maximize spinal fusion, which is a promising alternative to traditional autologous bone grafting. In this paper, we will review the recent research progress on gene therapy for spinal fusion. 1. Prospects of gene therapy for spinal fusion Spinal fusion is different from bone healing of long bone fractures of the extremities: 1) it is not a reconstruction of the original anatomical structure, but a reconstruction of a reasonable mechanical structure of the spine to change the anatomical changes of spinal pathology due to scoliosis, slippage, tumor, infection and trauma. 2) posterior lateral intertransverse fusion of the spine has a unique local microenvironment: the transverse processes are small, and the de-corticalization of the spine is very important for the blood flow to the graft. The blood supply to the graft after corticalization is small, especially in the central part of the graft, which requires revascularization; intertransverse fusion is a large segment of ectopic osteogenesis between the paravertebral muscles, and the mechanism of fusion and the mechanical environment to which it is subjected are different from the orthotopic bone repair of extremity bone grafts. In recent years, bone tissue engineering techniques involving life sciences and engineering have become an important research hotspot in the field of bone repair and reconstruction. In clinical experimental studies of spinal fusion, implantation of up to milligrams of recombinant human bone morphogenetic proteins (rhBMPs) in the posterior lateral intertransverse process is required to effectively promote spinal fusion to overcome the disadvantages of easy loss, short half-life and low efficiency of exogenous protein application. However, the clinical application of supraphysiological doses of rhBMPs is difficult to be further applied in the clinic due to the disadvantages of expensive and uncertain long-term physiological effects. Currently, in the context of a gene therapy research strategy that has been reoriented to focus on the treatment of non-lethal diseases, gene therapy holds the most promise for success in clinical tissue repair (e.g., bone healing) in the field of orthopedics using existing technologies. In the case of challenging clinical spinal fusions, gene therapy is suitable for the treatment of spinal fusions: (1) the bone healing process is highly conserved and well studied, and the efficacy of BMP as an effective osteogenic gene is well established. (2) Unlike the treatment of chronic diseases, bone healing is usually completed within a few weeks, and this process does not require long-term, efficient expression of exogenous genes. By selecting a specific gene introduction vector, the expression of the therapeutic gene can be controlled, thus avoiding excessive new bone production that could affect the function of the body. (3) By local gene therapy, specific seed cells can be selected and the therapeutic gene can be introduced into specific anatomical sites of limb fractures or spine through these genetically modified cells, thus avoiding the complications of systemic application and its technical requirements are relatively simple. (4) The results of preliminary animal experiments are relatively satisfactory. 2. Strategies and methods of gene therapy for spinal fusion 2.1 Adenovirus-mediated gene therapy for spinal fusion Gene introduction systems are the core technology of gene therapy and can be divided into viral and non-viral vector systems, among which adenoviral vectors are fully used in experimental and preclinical studies because of their advantages of efficient transfection, ease of application and high level of gene expression products. Riew et al. used a recombinant BMP2 adenovirus (AdBMP2) to transfect rabbit autologous bone marrow mesenchymal stem cells (BMSC) in vitro and subsequently implanted the cells and collagen vector into the posterior lateral transverse process of L5-6 in rabbits, and found that only one of the five animals produced spinal fusion. The group then increased the time of in vitro transfection of BMSC with AdBMP2 from 1 day to 7 days and found that all experimental animals achieved satisfactory spinal fusion.Riew et al. further used AdBMP2 to transfect porcine BMSC in vitro and then implanted BMP2 gene-modified porcine autologous BMSC cells into the thoracic intervertebral disc through a minimally invasive thoracoscopic technique, and after 6 weeks It was confirmed that satisfactory anterior fusion was obtained between all six thoracic vertebral bodies in the experimental group. Wang et al. compared the efficacy of BMP2-modified murine BMSC cells and autologous bone in the fusion of posterior lateral transverse processes in a single segment of the murine spine and found that the local gene therapy approach produced a large amount of nearly normal small sorghum-like bone between the transverse processes, whereas the histology of the autologous bone fusion samples showed thin, lace-like small sorghum bone. Hidaka et al. used AdBMP7 transfection of bone marrow cells cultured for 4 weeks in vitro and grown onto an allogeneic osteoconductive scaffold to construct artificial bone material, which was subsequently transplanted into the posterior lateral spine of nude mice. 8-week postoperative imaging and mechanical data showed a 70% and 80% spinal fusion rate, respectively. Helm’s group attempted a simpler method to promote spinal fusion by injecting an amount of AdBMP2 or AdBMP9 percutaneously directly into the lumbosacral paraspinal muscles of nude mice, and subsequent CT imaging and histology confirmed satisfactory spinal fusion in the animals. 2.2 Experimental study of adenovirus-mediated LMP-1 mineralizing protein (LMP-1) gene therapy for spinal fusion The practical application of BMP-like secretory osteoinductive factors for spinal fusion in higher animals has several drawbacks: 1) supraphysiological doses of BMP protein up to milligram level are required to effectively promote spinal fusion and it is difficult to maintain effective physiological concentrations of BMP protein in vivo; 2) BMP 2) BMP as a secretory osteoinductive protein is only one member of a complex osteogenic pathway with multi-factor interactions, and therefore BMP alone cannot fully mimic the molecular mechanisms inherent in spinal fusion. 3) The physiological side effects of high doses of BMP applied in vivo need to be further investigated. Based on this, Boden et al. successfully cloned a new intracellular signaling molecule, LMP-1 (LIM minialize protein), using differential PCR display technique, which is expressed several hours before osteoblast differentiation and induces the expression of various BMP molecules such as BMP2, 4, 6, 7 and other BMPs in target cells transfected with LMP-1 gene. growth factors. Therefore, as an intracellular signaling molecule, only a small amount of cellular expression is required to exert cascade amplification of osteogenic effects and improve osteogenic efficiency. The authors suggest that LMP-1 as an early osteogenic intracellular signaling molecule can overcome the deficiencies of BMP-like secretory factors and thus can mimic the process of osteogenesis in its physiological form. Boden et al. used liposomes as a gene transfer vector and found that artificial bone constructed from the composite of rat bone marrow cells transfected with LMP-1 gene and decalcified bone matrix could effectively promote the fusion of thoracic and lumbar spine in nude rats. Although the transfection efficiency of liposomes used as gene transfer vectors in this study was only 1% in vitro, the experiment still yielded good osteogenic results. Subsequently, they found that adenovirus with recombinant LMP-1 (AdLMP-1) required only a very low number of infection complexes (MOI=O.25 pfu/cell) for transfection of osteogenic precursor cells in vitro, which is several orders of magnitude lower compared to other gene therapies using adenovirus as gene transfer vector. In rabbit spinal fusion studies, the group demonstrated that AdLMP-1 (MOI=0.25-0.4pfu/cell) was effective in inducing fusion between the posterior lateral transverse processes of the rabbit lumbar spine after in vitro transfection of rabbit bone marrow or peripheral blood cells and implantation in decalcified bone matrix or collagen-ceramic complexes. Boden et al. further explored the feasibility of clinical application of the above experimental results, considering that adenovirus-neutralizing antibodies in the body of previously infected adenovirus patients may affect the clinical efficacy of AdLMP-1 when they are treated. They replicated a rabbit model of pre-infected adenovirus after pre-injection of rabbits with Ad5 adenovirus, and 4 or 16 weeks later, in transplantation of autologous LMP-1 gene-modified peripheral blood cells to promote rabbit spinal fusion assays, by increasing the dose of virus infection or the number of infected cells (MOI = 10 pfu/cell, 1 × 107 cells), and by increasing the (30 min) to overcome humoral immune rejection and achieve a satisfactory spinal fusion effect. 2.3 Study of naked plasmid-mediated BMP7 gene therapy for spinal fusion In order to avoid the possible side effects such as toxicity and immune reaction caused by the in vivo application of viral vectors, some scholars have tried to apply non-viral gene transfer vectors for spinal fusion. The complex was implanted into the posterior lateral intertransverse muscle of L5-6 in SD rats, and histological examination 2 and 4 weeks after surgery revealed that the above method was successful in inducing intramuscular ectopic osteogenesis, but did not produce spinal fusion. The authors suggest that it may be that the total amount of BMP7 protein produced by the constructed naked plasmid and collagen complex in the animal muscle was not sufficient to successfully induce spinal fusion. Further experiments by the authors will promote the production of BMP7 protein by improving the structure of the naked plasmid and optimizing the vector, or by intramuscular electrotransfer methods, in the hope that this non-viral gene therapy method with few side effects will be effective in promoting spinal fusion. 2.4 Study of non-viral BMP6 gene therapy mediated by nuclear transfection method to promote spinal fusion Conventional calcium phosphate precipitation, liposome and electroporation methods mediated by non-viral gene transfection are less efficient and more toxic to cells, making it difficult to be used in clinical applications. To overcome the technical bottleneck of low efficiency of non-viral gene transfection, a new non-viral gene transfection method, nucleofection, which combines electroporation and liposome technologies, has been recently introduced by Amaxa Biotech (www.amaxa.com), Germany. Sheyn et al [18] isolated and cultured porcine adipose-derived mesenchymal stem cells and transferred BMP6 plasmids into the cells by nucleofection in vitro to test the transfection and expression efficiency of exogenous genes in vitro. The authors then injected these transgenic cells into the paravertebral muscles of nude mice. It was shown that the nuclear transfection method could effectively mediate the transfection of BMP6 gene and also induce spinal fusion in nude mice with little side effects. 2.5 Study of commercial BMPs for spinal fusion In various animal models of spinal fusion, BMPs produced by recombinant genetic techniques achieved satisfactory fusion results, laying the foundation for their transition to clinical application. Satisfactory spinal fusion was achieved in all patients. Clinically, up to milligram levels of commercial BMPs are required to effectively promote spinal fusion, and the high cost limits their clinical application. 3. Safety studies of gene therapy for spinal fusion Although the results of animal experiments on gene therapy for spinal fusion are satisfactory, many problems need to be solved in the clinical application, such as the safety of viral vectors and the controllability of therapeutic gene expression, the optimization of seed cells, the selection of scaffold vectors, and the problem of multi-gene combination therapy. Among these, the safety of gene therapy is the first issue to be considered. 3.1 Safety of viral vectors and controllability of therapeutic gene expression: By removing all viral genes from viral vectors while retaining only the cis-acting elements necessary for their replication and packaging is the direction of viral vector development. This strategy maximizes vector capacity and reduces the immunotoxicity induced by viral protein expression. Early gene expression products in the adenovirus genome are the underlying cause of cellular and humoral responses at the site of adenovirus application in vivo, and this immune rejection can lead to a decrease in the level of therapeutic gene expression. Although the injection of genetically modified seed cells into the body can avoid immune rejection by direct exposure of the body to the virus, studies have shown that in some cases, the adenovirus early gene products expressed by adenovirus-infected cells may escape from the cells and cause immune rejection. Therefore, the production of visceral-free adenoviruses with most of the viral genes deleted is a trend for future applications of adenoviral vectors, which have not yet been used in bone healing research. The adeno-associated virus (AAV) vector was constructed by deleting all components of the viral genome while retaining only the long terminal repeat sequences at both ends, thus making the vector less immunogenic for in vivo application. intramuscular ectopic osteogenesis and no immune inflammatory response at the injection site. In contrast to the AAV vector, in immunocompetent rats, a dose of immunosuppression 24 hours prior to intramuscular injection of AdBMP2 was required to temporarily suppress the body’s immune response before significant osteogenesis could be observed. AAVBMP2/4 vector is expected to achieve satisfactory spinal fusion in immunocompetent organisms using a minimally invasive technique to inject the vector into the posterior lateral intertransverse muscles of the spinal segments requiring fusion. Efficient expression of the therapeutic gene is not the only requirement in gene therapy, as the unregulated state of the gene entering the body can have disastrous consequences for the organism, and the best approach is to be able to control the spatial and temporal expression of the gene in an orderly manner. Although the results of animal studies on gene therapy for spinal fusion are promising, for future clinical applications of gene therapy using stem cells as seed cells, it is particularly important to regulate the expression of therapeutic genes by rigorously regulating the timing, duration, and yield of exogenous gene expression, since bone healing is usually completed within a few weeks and this process does not require long-term, efficient expression of exogenous genes such as BMPs. Excessive new bone production after prolonged and efficient expression of BMP can cause unnecessary compression of nerve roots and spinal cord in the vicinity of spinal fusion; moreover, BMP has multiple physiological effects and the potential systemic side effects of excessive BMP concentrations beyond the therapeutic window should not be underestimated. With the successful development of safe, non-toxic, trigger-activated vectors that can induce rigorous regulation of gene expression at the transcriptional level, a powerful tool has been provided to bring gene therapy into clinical application. The new generation of tetracycline-regulated gene expression system (Tet-on/off) effectively solves the problems of toxic effects of inducers, baseline escape and low level expression, and is one of the most commonly used gene expression regulation systems for future clinical applications. gazit et al. transfected the plasmid ptTATop-BMP2 (Tet-off), a tetracycline-regulated BMP2 expression C3H10T1/2 cells and screened cell clones stably expressing BMP2, and used the cells and collagen sponge composite to repair bone defects of the radius in mice. It was the first time that tetracycline derivative doxycycline (DOX) could regulate the expression of BMP2 in mice and effectively repair radial bone defects. The group recently injected 2×106 of these cells directly into the L2-L6 paravertebral muscles of C3H/HeN mice and added DOX to the drinking water of the animals at 3, 7 and 30 days after surgery to terminate the expression of BMP2 and thus determine the minimum window of BMP2 expression required for spinal fusion. The results showed that the animals in the experimental group were able to form new bone in the posterior lateral spine after 4 weeks, and more mature laminae could be formed after 8 weeks to achieve strong spinal fusion. More importantly, it was demonstrated that a minimum time window of 7 days for BMP2 expression in this spinal fusion model was sufficient to induce new bone formation to promote spinal fusion, thus providing the first useful exploration of the relationship between spatial and temporal expression of BMP2 and new bone formation. In addition, the experiment also showed no local or systemic toxic effects in the vicinity of spinal fusion. Gazit et al. further constructed a tetracycline-positive (Tet-on) AAVBMP2 vector, which was injected directly into mouse cranial defects and paravertebral muscles to effectively repair cranial defects and promote spinal fusion. The authors will further investigate the dose-to-osteogenic response curve between DOX and new bone production in order to maximize the natural osteogenic pattern and precisely regulate bone formation. 3.2 Safety study of commercial BMPs for spinal fusion Clinically, laminectomy decompression often requires simultaneous posterior lateral intertransverse, anterior interbody, or posterior interbody bone graft fusion. Martin et al. applied BMP2 to 17 lumbar monkeys after laminectomy to perform posterior-lateral fusion of the lumbar spine. The results showed only reactive bone formation at the laminectomy site without excessive new bone production to compress the spinal cord. In a subsequent experiment, Boden et al. implanted BMP2 with a different vehicle in 24 lumbar laminectomized monkeys with no osteogenesis in the exposed dura after laminectomy. The results of this study confirm that BMPs are safe and effective in promoting spinal fusion. 4. Principles of animal experiments for gene therapy for spinal fusion Gene therapy for spinal fusion must be strictly regulated in animal experiments before entering clinical trials, and animal experiments must follow the principle of lower to higher animals. Because of the differences in the physiological and mechanical anatomical environments of animals and humans, data obtained in animal studies must be extrapolated to human clinical applications with caution; gene therapy protocols that are successful in animal fracture healing and spinal fusion studies may not necessarily be successful in human clinical trials. Conversely, however, protocols that are unsuccessful in lower animal studies will necessarily have difficulty succeeding in the more complex human setting. Therefore, gene therapy for bone healing or spinal fusion that is successful in lower animals needs to be successful in non-human primate experiments before it can enter clinical trials and eventually be transitioned to human clinical applications. At present, experimental studies on gene therapy for spinal fusion are still limited to small animal experiments, and most of them are single posterior lateral intertransverse fusion mode of the spine. 5. Conclusion A large amount of preclinical research data clearly shows that local gene therapy technology can be applied to human disease treatment, but for patients who need to receive bone repair treatment, autologous bone grafting is still the conventional clinical treatment method. In general, gene therapy techniques are not necessary for all patients. However, for refractory bone repair, this technology offers a new therapeutic avenue. Furthermore, when applying gene therapy strategies to treat the aforementioned diseases, investigators must take into account that these new molecular therapies are capable of being used to improve the quality of life of patients, and not to treat lethal diseases. Therefore, the safety of such treatments must first be considered. The more desirable adenoviral vectors used in animal studies will have to be further refined for future clinical applications to ensure a significant degree of biosafety in their application. With further refinement of gene therapy techniques that are safe, efficient, and enable precise regulation of therapeutic genes, gene therapy techniques are expected to replace conventional clinical autologous bone grafts as a routine means of promoting spinal fusion.