Degenerative disc disease (DDD) is a chronic process that can become clinically manifest in multiple disorders such as idiopathic low back pain, disc herniation, radiculopathy, myelopathy, and spinal stenosis. The limited available technology for the treatment of these and other pathologic and disabling conditions arising from DDD is highly invasive (eg, surgical discectomy and fusion), manifesting a certain degree of complications and unsatisfactory clinical outcomes. Although the precise pathophysiology of DDD remains to be clearly delineated, the progressive decline in aggrecan, the primary proteoglycan of the nucleus pulposus, appears to be a final common pathway. It has been hypothesized that imbalance in the synthesis and catabolism of certain critical extracellular matrix components can be mitigated by the transfer of genes to intervertebral disc cells encoding factors that modulate synthesis and catabolism of these components. The successful in vivo transfer of therapeutic genes to target cells within the intervertebral disc in clinically relevant animal models of DDD is one example of the rapid progress that is being made towards the development of gene therapy approaches for the treatment of DDD. This chapter reviews the ability of gene therapy to alter biologic processes in the degenerated intervertebral disc and outlines the work needed to be done before human clinical trials can be contemplated.
Spinal impairments, in particular low back pain, frequently occur in Western society with a lifetime prevalence of 60–80%. Low back pain is the most common cause of disability in people below the age of 45 years and accounts for 8 million physician visits and 89 million lost work days per year.1 Prevalence of musculoskeletal impairment of the spine in the US is estimated to be greater than 18 million – over one-half of all musculoskeletal impairment. Many of these spinal impairments are the direct or indirect result of degenerative disc disease (DDD). DDD is a chronic process that is characterized in part by progressive loss of proteoglycans and water content in nucleus pulposus and that can clinically manifest in multiple disorders, such as disc herniation, radiculopathy, myelopathy, and spinal stenosis. It is a significant source of patient pain and morbidity, utilizing a large portion of health care resources.2,3,4,5,6 Current treatment options, including both conservative measures such as bed rest, anti-inflammatories, analgesia, and physical therapy as well as surgical measures, target the clinical symptoms of DDD as opposed to addressing the pathologic processes occurring early in the course of degeneration. Conservative approaches can be effective in alleviating symptoms within 2 months in the majority of cases. However, when conservative methods fail, invasive surgical procedures such as discectomy with or without fusion are often considered. While these costly surgical procedures can be highly successful, they hold significant risks for complications, such as nonunion and infection. Moreover, the clinical outcomes can be unpredictable. However, with recent advancements in molecular biology (including recombinant DNA technology, the cloning of genes, and gene transfer technology), it becomes possible to contemplate treating the intervertebral disc itself at the molecular level to prevent or delay the progression of disc degeneration.
The precise etiology and pathophysiology of DDD remains to be clearly delineated. It is known that the matrix of a healthy nucleus pulposus is rich in proteoglycans and type II collagen, whereas the annulus fibrosus is rich in type I collagen.7,8 The proteoglycan content of the disc declines with age, a process that partly reflects decreased synthesis of these macromolecules by the disc cells.9,10,11 Although the reasons for this decline are unknown, reduction in proteoglycan content can have consequences for the disc's ability to resist mechanical loads. Traditionally, disc degeneration has been considered to be largely a biomechanical phenomenon. It is now believed to result from complex interactions of biologic and biomechanical factors.12 The progressive decline in aggrecan, the primary proteoglycan of the nucleus pulposus, is known to be a significant and characteristic factor of the early stages of DDD.13,14,15 At the biochemical level, the diminished proteoglycan content reflects an imbalance in the normal anabolic and catabolic functions of the nucleus pulposus cells, resulting from decreased synthesis, increased catabolism, or a combination of these two processes. With reductions in proteoglycan content of the intervertebral matrix, the nucleus pulposus dehydrates, decreasing disc height and altering its load-bearing capacity.16,17,18 This in turn may directly affect overall spinal biomechanical function by altering the loads experienced by the facet joints, leading to degenerative changes in these structures. Although disc degeneration most probably evolves in response to a complex interplay of multiple biochemical and biomechanical factors,12 the ability to restore proteoglycan content may have therapeutic benefit by increasing disc hydration and potentially improving biomechanics.
The potential to treat DDD as well as other musculoskeletal disorders on a molecular level evolved with the recognition of potentially therapeutic genes,19 and has progressed to the successful clinical use of growth factors to improve spinal fusion rates.20 With regard to intervertebral disc degeneration, the ability to increase proteoglycan synthesis by intervertebral disc cells was first demonstrated by Thompson et al,21 who showed that the exogenous application of recombinant human transforming growth factor (TGF)-β1 to canine disc tissue in culture stimulated in vitro proteoglycan synthesis. These authors suggested that growth factors may be useful for the treatment of disc degeneration. Gruber et al22 also demonstrated that TGF-β1 can induce changes in proliferative and extracellular matrix cellular activities in human disc cells in three-dimensioned culture. Subsequent studies with other growth factors such as insulin-like growth factor 1 (IGF-1) and bone morphogenic protein 2 (BMP-2) have also exhibited the ability to upregulate proteoglycan synthesis activity in disc cells.23 Furthermore, An and co-workers demonstrated that OP-1 stimulated proteoglycan and collagen synthesis in rabbit intervertebral disc cells cultured in alginate beads.24 Although these recombinant growth factors have been shown to stimulate extracellular matrix synthesis in a variety of musculoskeletal tissues, their potential use in the management of DDD (a chronic, progressive disorder) would be limited by their relatively short biologic half-lives that would enable only transient biologic effects after their delivery. The main hypothesis of much of the research detailed in this chapter is that direct transfer of the genes of therapeutic proteins (including growth factors) to intervertebral disc cells in vivo enables the disc cells to manufacture these proteins endogenously, on a continuous basis – thereby enabling long-term regulation of matrix synthesis with the potential to prevent or delay DDD.
Intervertebral disc is a well-encapsulated, avascular organ with a gel-like inner core (nucleus pulposus) consisting of nucleus pulposus cells scattered within an extracellular matrix. It is contained laterally by an outer ring consisting of concentric laminae of collagen fibers running obliquely between adjacent vertebral bodies (annulus fibrosus).
The avascular disc receives the majority of its nutrition via passive diffusion through the cartilaginous endplates. The lack of a direct vascular supply results in low oxygen tension within the disc and causes the cells of the nucleus pulposus to undergo anaerobic metabolism. The ensuing high lactate concentration and subsequent low environmental pH most likely inhibit matrix repair. The matrix of a healthy nucleus pulposus is normally rich in proteoglycans and type II collagen, with a water content of over 85% by volume in juveniles. In adults, proteoglycan content decreases, and type I collagen becomes more prevalent. Water content decreases to approximately 70–75% in adults, and decreases even further with aging and degeneration.25,26 Whether the proteoglycans and collagens are produced by cells within the annular fibrosus, nucleus pulposus, or chondrocytes from the vertebral cartilage endplates remains to be determined, and little is known concerning the phenotype of disc cells and their contribution to the characteristics of the matrix. The cells of the nucleus pulposus may include cells of neural crest origin, namely notochordal cells. In human discs, the notochordal cells typically disappear by 10 years of age, at about the same time morphological signs of degeneration can be seen. Some evidence suggests that these notochordal cells may have a role in the homeostasis of the proteoglycan matrix.27
A variety of biochemical mediators of inflammation and tissue degeneration, including matrix metalloproteinases (MMPs), prostaglandin E2 (PGE2), nitric oxide (NO), and a variety of cytokines – particularly interleukin-1 (IL-1) and interleukin-6 (IL-6) – have been found to be involved in the matrix breakdown of articular cartilage. The current authors demonstrated increased levels of MMPs in degenerated human discs compared to normal, nondegenerated controls. Similarly, the levels of PGE2, NO, and IL-6 were significantly higher.28 These experimental findings suggested that these biochemical agents are involved in the net loss of proteoglycans associated with disc degeneration.
Normally, the swelling pressure resulting from the high concentration of proteoglycans in the nucleus pulposus helps to maintain disc height and contributes to the load-bearing ability of the disc.17,18 Therefore, loss of proteoglycans may directly affect the biomechanical function of the intervertebral disc as well as alter loading of the facets joints and other structures, leading to degenerative changes. There is also clinical evidence that biomechanical alterations may contribute to the earlier onset of DDD. This is most often seen in patients undergoing spinal fusion surgery, which seems to place a higher mechanical ‘burden’ on the discs adjacent to the fusion. Furthermore, mechanical instability of the functional spinal unit, as in the setting of a complete pars interarticularis defect, has been shown clinically to lead to premature DDD. Interestingly, even in discs that have been ‘biomechanically’ stressed, the degeneration process is always associated with net progressive loss of proteoglycans and matrix. Although the exact etiology of disc degeneration is complex and may involve different clinical mechanisms, strategies that result in a net increase in proteoglycans may have therapeutic potential in altering the natural history of disc degeneration. Therefore, gene therapy research in this field has focused on upregulating matrix synthesis by the cells of the intervertebral disc. These strategies could involve increasing the synthesis of certain extracellular matrix components such as proteoglycans, blocking their catabolic breakdown, or combination of increased synthesis and decreased catabolism.
Overview of gene therapy for intervertebral discs disease
Growth factors are small proteins with pleiotrophic effects on cells, including stimulation of cell differentiation and division. TGF-β1 and BMPs are two examples of growth factors that have demonstrated strong potential for altering intervertebral disc biology. As mentioned, Thompson et al studied the in vitro response of canine intervertebral disc tissue to the following growth factors: human recombinant insulin-like growth factor-1(IGF-1), epidermal growth factor (EGF), fibroblast growth factor (FGF), and TGF-β1. The observed responses – particularly in the nucleus and transition zone – suggested to the authors the possibility that disc repair could be modulated by growth factors. The authors concluded that a therapeutic approach to DDD involving enhanced tissue repair by exogenous growth factors could be of great clinical significance.21 Furthermore, apoptosis is an important type of cell death that plays a role in tissue homeostasis. Gruber et al29 demonstrated that a significant reduction in the percentage of apoptotic disc cells could be achieved using IGF-1 and platelet-derived growth factor (PDGF). They suggested that these growth factors may have future therapeutic potential in the treatment of disc degeneration.
However, clinically important issues not addressed in this landmark paper include the delivery mechanism and the length of therapeutic effect of these exogenous growth factors to targeted cells in the intervertebral disc. The normal half-life for most of these growth factors in vivo is on the order of approximately 20 min. Therefore, the therapeutic effect of injecting growth factors directly into the intervertebral disc may be too transient to have a major long lasting effect on a chronic disorder such as DDD, and repeated or continuous injections may not be practical with current technology (nor well-tolerated by patients). An alternative therapeutic approach, which could address these concerns, would be to genetically modify disc cells in situ through gene transfer so that the cells themselves could endogenously manufacture the desired growth factors for a sustained length of time.
Gene transfer involves the delivery of exogenous genes to target cells, resulting in gene expression and subsequent production of proteins. There are two basic strategies for gene delivery. Either vectors with appropriate genes are introduced directly into the body, or the target cells are removed, cultured, genetically altered in vitro, and then reimplanted. The former strategy is known as direct, or in vivo, gene therapy, and the latter as indirect, or ex vivo, gene therapy. The relative merits of these strategies depend on the anatomy and physiology of the target organs, the pathophysiology of the disease, the vector of choice, safety considerations, and other variables. Based on the biology of intervertebral discs and the difficulty in accessing the tissue, the current authors believe that in vivo gene transfer is most suited to intervertebral gene therapy for minimally invasive treatment of early disc degeneration.
Successful gene therapy depends on the efficient transfer of genes to target cells with subsequent sustained expression. With few exceptions, naked DNA is not taken up and expressed by cells. Consequently, vectors are necessary to package and insert genes into cells in such a way that the genetic information can be expressed. Several types of vector systems are available for the delivery of genes to cells in vivo. There are a variety of available viral vectors for this purpose, and the choice of vector is usually depend on the type of the cell that is being transduced as well as the disease that is being treated.
The current authors have extensively examined the use of adenoviral vectors for direct, in vivo gene transfer to the intervertebral disc. The results suggest that adenoviral vectors may present an acceptable gene transfer system for the intervertebral disc. Nevertheless, the issue of an immune response to adenoviral exposure is critical, especially in light of the reported death of a patient undergoing a clinical trial with adenovirus. For this reason, the potential use of other vectors must be investigated. The following sections will examine adenovirus, adeno-associated virus (AAV), and nonviral vectors in context of gene therapy for DDD.
Adenoviral vectors are being widely used for gene delivery in vivo and are in clinical trials for cancer. Adenovirus-based vectors have the advantage of infecting nondividing cells efficiently in culture as well as in vivo, a process resulting in a high level of transient gene expression. Adenoviruses are double-stranded linear DNA viruses of approximately 30–35 kb in length. To generate a defective adenovirus for gene transfer application, the E1 gene, important for viral gene expression and replication, can be removed.30 Successful gene therapy depends not only on efficient gene transfer to the targeted cells, but also on the expression of therapeutic transgenes for sufficiently long periods of time. In most gene transfer experiments, a decline in transgene expression with time is observed. To date, this has been one of biggest limitations to effective gene therapy for chronic diseases. The duration of gene expression following adenovirus-mediated gene transfer is limited in certain tissue/organs by immune reactions to viral proteins31 and to foreign proteins encoded by the transgenes.32 In addition, adenoviral vectors cause inflammation when injected into the joint space33,34 and a variety of other organs.35 Owing to the immune response, gene expression for longer than 12 weeks has been difficult to achieve in most musculoskeletal tissues following adenoviral gene transfer to an immunocompetent animal. However, gene expression can occur for longer periods of time in an ‘immune privileged’ site. In studies by the author's group, we have achieved intradiscal marker gene expression (luciferase) in the rabbit lumbar disc for up to 1 year after transduction by adenovirus (Figure 1).36 As the intervertebral disc is well encapsulated and avascular, it seems to act as an ‘immune priviledged’ organ – thereby permitting longer duration of transgene expression than in other musculoskeletal tissues.
An alternative strategy is utilizing vectors that are less immunogenic than the adenovirus. The AAV, a single-stranded DNA virus that requires a helper virus such as adenovirus or herpes simplex virus for replication, is one such vector that has been shown to have little or no immune reaction – thus allowing for more prolonged transgene expression in a variety of tissue. Wild-type, but not recombinant, AAV integrates the therapeutic gene into a specific site on chromosome 19 in a stable, nonpathogenic manner. The main shortcomings of AAV vectors are that they are difficult to produce and are capable of carrying only small amounts of foreign DNA. These issues, as they relate to the intervertebral disc, have not been extensively investigated. Wild-type AAV is nonpathogenic and is not associated with any known disease. The virus infects a wide variety of dividing and nondividing cells, although with varying levels of efficiency. Furthermore, the efficiency of gene transfer in several tissues is sufficient to achieve significant levels of cellular transduction after direct, in vivo delivery. Also, long-term gene expression has been reported in several tissues types. Finally, recombinant AAV vectors express no viral gene products following infection of target cells. The lack of viral gene expression and the prolonged transgene expression are major reasons AAV is being extensively explored as a gene transfer vector in multiple tissue types. For these reasons, the current authors have studied the ability of AAV vectors to transfer and express exogenous genes in the intervertebral disc.37 In this study, AAV vectors were used to transduce human and rabbit nucleus pulposus cells in vitro and in vivo. The in vivo data suggested that transgene expression was low at 2 weeks but increased to a sustained level after 4 weeks. Gene expression could be observed up to 6 weeks. These data suggested that AAV might be a valuable alternative to other viral vectors for the use in gene therapy for DDD. The expression potency and the less immunogenic nature of the AAV make it a potentially attractive alternative to adenoviral gene therapy strategies.
The use of nonviral, plasmid-based systems where no infectious agents are administered to the patient or the patient's cells would eliminate a large portion of the potential risks associated with viral gene transfer. A number of different nonviral vectors for gene transfer have been developed. The most commonly used nonviral vectors are liposomes. These phospholipid vehicles deliver genetic material into a cell by fusing with the cell's phospholipid membrane. Liposome vectors are simple, inexpensive, and safe. Their drawbacks are transient expression of transgene, cytotoxicity at higher concentrations, and low efficiency of transfection. Other nonviral methods of gene delivery include DNA–ligand complexes and the gene gun. Nonviral vectors are nonpathogenic and relatively inexpensive to construct. However, when utilizing nonviral vector systems, the main concern is the transient transgene expression observed following vector delivery. The use of these vectors for gene transfer to joints and other tissues promotes transgene expression for a limited period, usually just days. Obviously, this length of transgene expression would be insufficient for treatment of a chronic disease such as DDD. Thus, the majority of gene therapy studies of the disc currently employ viral vectors.
The notion of using gene transfer for intervertebral disc applications was initially introduced by Wehling et al.38 In an in vitro study, they reported on a retrovirus-mediated transfer of two different exogenous genes to cultured chondrocytic cells from bovine intervertebral end plates: (1) the bacterial β-galactosidase (LacZ) gene, and (2) the cDNA of the human interleukin-1 receptor antagonist (IL-1Ra). β-galactosidase activity was determined by X-gal staining, and IL-1Ra was quantified by ELISA. Transfer of the β-galactosidase gene resulted in transduction of approximately 1% of the cell population. Transfer of the IL-1Ra cDNA resulted in the production of a significant amount of IL-1Ra in 48 h. The authors suggested that the removal of endplate tissue from a degenerating disc (under arthroscopic or X-ray control), followed by transfer of therapeutic genes to the cultured cells, and reinjection of the cells into the disc – thereby turning the host tissues into sites for the synthesis of the drug – could open a new avenue for the treatment of degenerative disease of the spine. Candidate genes were suggested that could conceivably produce proteins that would contribute to the stabilization and/or conservation of spinal structures: IL-1 antagonists, tumor necrosis factor alpha (TNF-α) antagonists, and inhibitors of MMPs.
Our group reported the successful adenovirus-mediated transfer of exogenous marker genes to intervertebral disc cells of skeletally mature female New Zealand White (NZW) rabbits in vitro and in vivo in 1998.39 Adenovirus was selected as the vector of choice for several reasons. First, the adenoviral vector is very efficient in transferring foreign genes to quiescent, nondividing, highly differentiated cells. Second, the intervertebral disc, being relatively encapsulated and avascular, seemed to offer an ideal environment for maintaining a high concentration of directly injected viral vectors – thus increasing the likelihood of transduction. Third, the well-encapsulated and avascular environment of the nucleus pulposus also appeared to have the potential to limit the entry of immunocompetent cells – thereby preventing immune reactivity and permitting prolonged transgene expression. For the in vivo portion of this study, the anterolateral aspects of lumbar intervertebral discs of 15 NZW rabbits were surgically exposed, and 20 μl of Ad/lacZ in saline solution (6 × 106 plaque-forming units (PFU)) was injected directly into the nucleus pulposus of three lumbar discs per rabbit. An equal volume of saline-only solution was injected into control discs. The rabbits were killed at 1, 3, 6, and 12 weeks after injection. Expression of the transferred lacZ gene was determined by X-Gal staining.
Direct, in vivo injection of Ad/lacZ into the nucleus pulposus resulted in the transduction of considerable numbers of cells, with marker gene expression persisting in vivo at an apparently undiminished level for at least 12 weeks – with no evidence of immune reactions (Figure 2). These results supported the hypothesis that the disc is an immune-privileged organ. Follow-up studies revealed evidence of continued foreign gene expression, even at 1 year.
The immune response
Notably, the rabbits in the studies involving adenovirus-mediated transgene expression showed no signs of systemic illness in response to the adenoviral vector or the resulting transgene expression. In addition, no histological changes suggesting a cellular immune response were observed.
Further studies suggested that the adenoviral vector might be suitable for delivery of therapeutic genes to the disc to modulate its biologic activity.36 For intradiscal gene therapy to be considered feasible for DDD, sufficient levels and duration of transgene expression must be achieved in the long term, resulting in demonstrable therapeutic biologic effects with minimal associated side effects. Adenovirus carrying luciferase marker genes was injected into lumbar discs of NZW rabbits. Careful immunology studies were performed measuring the titer of neutralizing antibodies to the virus in three different study groups. In the first study group, the disc was injected with the adenovirus luciferase construct (Figure 3a). In the second group, subcutaneous injection of adenovirus was performed simultaneous to the intradiscal injection (Figure 3b). In the third group, the adenovirus was injected subcutaneously 2 weeks prior to intradiscal injection, thereby deliberately immunizing the rabbits (Figure 3c). As shown in Figure 3, all rabbits from the three groups exhibited significant amounts of intradiscal transgene expression at 42 days after intradiscal injection. There was no correlation between the neutralizing antibody titer in the peripheral circulation and the intradiscal transgene expression at 6 weeks post intradiscal injection. Variable amounts of neutralizing antibodies to adenovirus were detected in the peripheral circulation in all the three groups, but with no demonstrable effects on intradiscal transgene expression – suggesting that the disc limits access of antibodies and immune-competent cells to the transduced cells.
Gene transfer of therapeutic growth factors
Encouraged by the results with marker proteins, the successful in vivo transduction of the intervertebral disc with a therapeutic gene, human TGF-β1, was soon accomplished.40 In this study, the current authors performed an in vivo study, again using the rabbit model, to determine the feasibility of adenovirus-mediated transfer of a potentially therapeutic gene to the intervertebral disc. TGF-β1 was selected because it has a wide range of potentially therapeutic as well as quantifiable biologic effects, and was previously found by Thompson et al21 to increase proteoglycan synthesis in cultured canine disc tissues. The anterior aspects of the middle three lumbar intervertebral discs of 11 New Zealand white rabbits were surgically exposed and directly injected with adenovirus containing human TGF-β1 cDNA. The supraadjacent disc served as an intact control for each rabbit. The rabbits were killed 1 week later. The expression of transferred genes was determined by immunohistochemical staining and ELISA, and proteoglycan synthesis was assessed by measurement of sulfate incorporation using PD-10 column assay. In vivo injection of Ad/CMV-TGF-β1 into the nucleus pulposus was found to result in an approximately 30-fold increase in active TGF-β1 synthesis and a five-fold increase in total TGF-β1 production in discs injected with the adenoviral-growth factor construct (Figure 4). Biologic modulation was also documented by a 100% increase in proteoglycan synthesis (Figure 5). Assays for TGF-β1 production and proteoglycan synthesis were also performed with discs that had been injected with viral control group (adenovirus luciferase marker gene). These viral control discs demonstrated no increase in production of TGF-β1 or proteoglycan, indicating that the increases in the TGF-β1 experimental group were a result of transgene expression and not a nonspecific response to the adenoviral vector. This observation of a significant increase in proteoglycan synthesis secondary to gene transfer strongly suggested that gene transfer has the potential to alter the time course of DDD – a spinal disorder characterized in part by loss of proteoglycans in the nucleus pulposus.
This study established adenovirus-mediated gene transfer as a technique for favorably modifying the biologic activity of intervertebral disc cells of rabbit in vivo. Before contemplating extending this approach to the treatment of human disease, it is necessary to demonstrate that human disc cells are indeed susceptible to adenovirus-mediated transduction. Accordingly, the objective of the following studies was to test the efficacy of the adenovirus-mediated gene transfer technique for transferring exogenous genes to human intervertebral disc cells in vitro, and to measure the effects of therapeutic gene transfer on matrix synthesis by human intervertebral disc cells.
Application to human intervertebral disc cells
In vitro studies with cultured human nucleus pulposus cells yielded similar promising results. Successful transduction of the lacZ marker gene delivered via adenoviral vectors was achieved with human cells from degenerated discs.41 Similar experiments with retroviral delivery of marker genes resulted in a smaller percentage of transduction,42 perhaps due to the minimal mitotic activity of the intervertebral disc cells. The response of human cells from degenerated discs to adenoviral-mediated delivery of TGF-β1 was assessed in the following manner.43
In the first study, the authors harvested lumbar and cervical disc tissue from 15 patients (age range: 18–67 years) during surgical disc procedures performed for disc herniation, stenosis, and idiopathic scoliosis. Disc degeneration grade – by analysis of preoperative MRI – ranged from 1 to 4 (no degeneration to moderately severe).44 The disc cells were isolated; cultured in monolayer, and treated with five different doses of adenovirus carrying the LacZ gene. A minimum dose of 150 MOI Ad/CMV-lacZ was found to be sufficient to achieve transduction of approximately 100% of disc cells – regardless of patient age, sex, surgical indication, disc level, and disc degeneration grade. The finding that cells from degenerated discs were no less susceptible to adenovirus-mediated gene transfer than those from nondegenerated discs is encouraging, considering that direct, in vivo adenovirus-mediated gene transfer to a human disc (whether degenerated or nondegenerated) would necessitate transduction of the native intervertebral disc cells. The determination of optimal dosage of adenovirus to achieve efficient transduction of human intervertebral disc cells in situ is therefore simplified by the fact that disc cell transducibility by adenovirus is relatively unaffected by patient age, sex, surgical indication, disc level, and degeneration grade (within the ranges encountered in this study). As the rate-limiting step for successful gene therapy is the ability to transfer genes efficiently, the achievement of efficient gene transfer to human intervertebral disc cells in this study was an important and necessary step in the development of gene transfer strategies for the management of human disc disorders.
Having established the efficacy of marker gene transfer to human intervertebral disc cells, the next step was to transfer potentially therapeutic genes to human disc cells to elucidate the biologic effects of therapeutic gene transfer. In the second study, cervical and lumbar intervertebral disc tissue from the nucleus pulposus region was obtained from 22 patients during surgical disc procedures. The cells were isolated and cultured in monolayer. At full confluence, the cultures were organized into four groups and treated with (1) saline, (2) 2 ng/ml of exogenous TGF-β1, (3)Ad/CMV-luciferase, or (4) Ad/CMV-hTGF-β1. The viral constructs were administered at doses of 75 and 150 MOI. At 2 days after treatment, the concentration of TGF-β1 in the cell culture supernatant of each group was measured by ELISA. Then after 2 days of being cultured in fresh Neuman–Tytell serum-less medium containing either 35S-sulfate or 3H-proline, newly synthesized proteoglycan and collagen/noncollagen was analyzed using established techniques. The measurement of TGF-β1 concentration showed that cultures treated with exogenous TGF-β1 (2 ng/ml initially) exhibited a decrease in TGF-β1 concentration by day 2 (0.56±0.09 ng/ml). In contrast, cultures treated with Ad/CMV-TGF-β1 (MOI=75) exhibited an increase in TGF-β1 concentration from less than 0.05 ng/ml at the time of treatment to 3.12±0.28 ng/ml by day 2. The measurements of proteoglycan synthesis showed that after 2 days of incubation, cultures treated with exogenous TGF-β1 failed to show increase in proteoglycan synthesis, whereas the cultures transduced by Ad/TGF-β1 exhibited an increase of approximately 200% in proteoglycan synthesis (P<0.05) over that of the other groups. These results point to certain advantages of gene transfer over treatments based on delivery of exogenous growth factors, which include superior bioavailability of endogenous TGF-β1 produced by transduced disc cells and a possible upregulation of TGF-β1 receptors. For collagen and noncollagen synthesis, however, exogenous TGF-β1 (2 ng/ml) and the adenoviral construct encoding TGF-β1 (MOI of 150) exhibited similar upregulatory potential, with both treatments resulting in approximately a 350% increase in collagen synthesis (P<0.05) and 250% increase in noncollagenous protein synthesis (P<0.05) over that of control cultures.
Interestingly, the viral dose required to increase proteoglycan synthesis was significantly less than that required for 100% transduction of the cells, perhaps highlighting the ability of a transduced cell to influence the biologic activity of nongenetically altered neighboring cells. The concept that successfully transduced cells exert a paracrine-like effect on their nontransduced neighboring cells implies that significant alteration in protein synthesis can be achieved with a small number of transduced cells. A better understanding of this paracrine effect may enable the use of decreased viral loads to achieve a therapeutic effect, thereby minimizing potential viral toxicity. These experiments were also performed with a viral control, which further established that the increase in biologic activity was the result of the delivered genetic material and not of the adenoviral vector.
Subsequent in vitro studies with other promising growth factors such as BMP-2 and IGF-1 documented the potential of adenoviral delivery of these factors to increase proteoglycan synthesis in a viral dose-dependent manner. Considering the potential adverse effects of viral vectors, studies have been undertaken to develop strategies to minimize viral loads while maintaining the same biologic effects. Experiments with combination gene therapy involving TGF-β1, IGF-1, and BMP-2 suggested that these growth factors are synergistic in amplifying matrix synthesis.45 In this study, human intervertebral disc cells were obtained during surgical procedures and nucleus pulposus cells were isolated by enzymatic digestion. Upon full confluence of the cells, individual cultures were transduced with Ad/TGF-β1, or Ad/IGF-1, or Ad/BMP-2, or with a combination of two viral constructs (Ad/TGF-β1+Ad/IGF-1; Ad/TGF-β1+Ad/BMP-2; Ad/IGF-1+Ad/BMP-2) or with all three viral constructs [Ad/TGF-β1+Ad/IGF-1+Ad/BMP-2]. Total viral concentration applied to each culture was adjusted to 3.75 × 107, MOI=75. Newly synthesized proteoglycan was assessed by 35S radioactive sulfate incorporation using chromatography and scintillation counter. A significant increase in proteoglycan synthesis over controls was obtained for cultures treated with single agents (295% for Ad/TGF-β1, 180% for Ad/IGF-1, 190% for Ad/BMP-2). A synergistic or additive effect was observed for double combinations of therapeutic genes (398% for Ad/TGF-β1+Ad/IGF-1, 348% for Ad/TGF-β1+Ad/BMP-2, and 322% for Ad/IGF-1+Ad/BMP-2). Cultures that received all three therapeutic genes demonstrated a 471% increase in newly synthesized proteoglycans (Figure 6). The results of this study demonstrated that combination or ‘cocktail’ gene therapy offers a promising method for maximizing matrix regeneration for low-dose viral mixtures.
Moreover, adenoviral delivery of tissue inhibitor of metalloproteinase 1 (TIMP-1) demonstrated the same ability to increase proteoglycan synthesis.46 TIMP-1 is an endogenous inhibitor of MMPs, enzymes capable of degrading the extracellular matrix of the intervertebral disc. Significant increases in proteoglycan synthesis compared to controls were obtained for all doses of Ad/TIMP-1 administered to cultured human nucleus pulposus cells. The increases were 143, 204, 193, 477%, for MOIs of 50, 75, 100, 150, respectively. This finding established a second gene therapy strategy to modify the disrupted balance of synthesis and catabolism occurring in the degenerated intervertebral disc: inhibition of matrix degradation with ensuing net increases in proteoglycan content. It remains to be determined if combination gene therapy with both an anabolic growth factor and a catabolic inhibitor such as TIMP-1 will have a synergistic effect.
As mentioned, the biochemistry of disc degeneration is poorly understood but is believed to involve perhaps hundreds of cytokines and inflammatory mediators in its pathophysiology. A variety of genes might be expressed during the course of disc degeneration that may be in part detrimental to the preservation of matrix (eg catabolic cytokines) but also partly a repair response by the cells to the damaged environment (eg anabolic cytokines and growth factors). To determine the efficacy of vector-mediated gene transfer in preventing or delaying the progression of disc degeneration and to continue its progress toward successful human clinical trials, it is critical to test rigorously the proposed gene therapy strategies in animal models of disc degeneration that closely simulate the human condition. A number of models have been proposed in the literature. Disc degeneration occurs spontaneously in some species, such as the nonchondrodystrophic beagle and the sand rat.47 Other species require artificial interventions to bring about degenerative changes within a reasonable time frame. The annular stab model of degeneration in the NZW rabbit has been well described in the literature by Lipson and Muir.48 In previous in-house studies with this model, the current authors found that the 3 mm incision of the anterior annulus allowed escape of nuclear material from the disc. There was concern that the degenerative changes induced by the full thickness 3 mm annular incision were too abrupt, in contrast to the gradual changes that occur in the human condition. For these reasons, the current authors modified this technique to produce a puncture injury using a 16-gauge hypodermic needle.
Extensive MRI and histological data have shown that the needle stab model produces gradual and consistent degenerative changes that closely parallel the human condition. MRI analysis of 18 rabbits demonstrated that the stabbed discs exhibited a progressive decrease in ‘MRI Index’ (the product of nucleus pulposus area and signal intensity, from T2-weighted mid-sagittal plane images) starting at 3 weeks post stab and continuing through 24 weeks – with no evidence of spontaneous recovery or reversal of MRI changes. (Figure 7). Radiographs revealed disc space narrowing, disc wedging, and osteophyte formation in stabbed discs as compared with healthy discs. Histological examinations of the stabbed discs revealed cracks and clefts within the nucleus as well as delamination and infolding of the annulus. In addition, clusters of notochordal cells were readily apparent in healthy discs, but were sparse in discs that had been stabbed. Stabbing the anterolateral annulus fibrosus of rabbit lumbar discs with a 16-gauge hypodermic needle results in a number of slowly progressive and reproducible MRI, X-ray, and histological changes over 24 weeks that show a similarity to changes seen in human intervertebral disc degeneration (Figure 8). This model would appear to be suitable for studying pathophysiology of DDD and testing efficacy of novel treatments of DDD.49
The potential of gene therapy to alter the biologic processes occurring in the degenerated intervertebral disc has been clearly established, yet significant work remains before human clinical trials can be considered. The next step in this development will be to assess the feasibility of transducing the cells of degenerated rabbit discs with marker and therapeutic genes. Preliminary results are promising (Figure 9). It is the opinion of the current authors that vector-mediated transfer of the genes of anabolic proteins such as TGF-β1, BMP-2, and IGF-1 to the degenerative disc will stimulate upregulation in proteoglycan synthesis. As endogenous biochemical mediators of matrix degeneration (MMPs) are likely to be present in the model, we believe that the discs will also respond to treatment by vector-mediated delivery of anticatabolic genes such as TIMP. A combination of genes producing both anabolic effects and anticatabolic effects may be the most desirable in preventing or delaying degeneration of the intervertebral disc.
Finally, the basic science of the effects of growth factors and catabolic inhibitors in the biologic processes and mechanical functioning of the spine needs to be clearly elucidated. More biochemical studies are warranted to delineate the relationship between viral concentration, transgene expression, and protein synthesis. Despite the hurdles that remain, the potential of gene therapy to alter the course of intervertebral disc degeneration holds much clinical promise, and will continue to stimulate future investigations.
Given the complexity of the clinical entity known as degenerative disc disease, including the multiple pathways that can lead to degenerative disc disease, it is unlikely to expect that one vector will ultimately serve all clinical needs. There clearly are trade-offs between the efficiency of gene transfer, duration of transgene expression, and immunogenecity that are characteristic of each vector, to name just a few variables. In addition, these variables are influenced by the characteristics of the target tissues/organs (especially in various stages of disease.) Ideally, the spine surgeon of the future may have a variety of vectors to choose from for application to specific clinical problems of the disc. Treatment of certain types of rapidly developing DDD might require a rapid uptake of virus and vigorous, early transgene expression of relatively short duration. More indolent types of DDD may benefit from slower release of proteins over a longer period of time. These would likely require different vectors. Further studies will help to elucidate how different vectors perform in vivo in different situations, and will therefore help to further refine the principles of vector selection that we have started to develop for applications to the disc.
Additionally, the genes must be expressed at the appropriate levels and at the required times. Excessive growth factor production via gene therapy could be harmful. Regulated expression is being developed so that genes can be turned on or off as needed.50,51,52 Moreover, recent genetics studies of disc degeneration suggest that genes encoding aggrecan and type IX collagen may be good candidate genes for transfer for modification in strategies to delay or prevent disc degeneration in susceptible populations.53,54 Recent studies have demonstrated Sox 9 to be an essential transcription factor for type II collagen synthesis as well as chondrogenesis.55,56 Finally, high throughput technologies, such as microarray, are elucidating thousands of genes that may be good candidates for regeneration of intervertebral disc disease.
Clearly, there are many pertinent issues concerning the timing and control of growth factor delivery to be addressed before gene therapy can be considered for clinical use of humans for the treatment of intervertebral disc disease. Needless to say, clinical trials must establish patient safety before clinical efficacy is sought. In future, a series of vectors encoding multiple crucial genes maintaining extracellular matrix homeostasis regulated by physiologic signals might be used to enhance degenerative intervertebral disc.
Conrad DA et al. Cost of low back pain problems: an economic analysis. In: Weinstein JN, Gordon SL (eds). Low Back Pain: a Clinical and Scientific Overview. AAOS publishers: Rosemont, IL, 1996.
Anderson J . Back pain and occupation. In: Jayson MIV (ed). The Lumbar Spine and Back Pain. Churchill Livingstone: London, 1987, pp 2–36.
Anderson JA . Epidemiological aspects of back pain. J Soc Occup Med 1986; 36: 90–94.
Borenstein D . Epidemiology, etiology, diagnostic evaluation, and treatment of low back pain. Curr Opin Rheumatol 1992; 4: 226–232.
Kraemer J . Natural course and prognosis of intervertebral disc diseases. International Society for the Study of the Lumbar Spine Seattle, Washington, June 1994. Spine 1995; 20: 635–639.
Waddell G . Low back pain: a twentieth century health care enigma. Spine 1996; 21: 2820–2825.
Conventry MB, Ghormley RK, Kernohan JW . The intervertebral disc: its microscopic anatomy and pathology: Part 1. Anatomy, development, and pathology. J Bone Joint Surg 1945; 24: 105–112.
Conventry MB, Ghormley RK, Kernohan JW . The intervertebral disc: its microscopic anatomy and pathology: Part 2. Anatomy, development, and pathology. Changes in the intervertebral disc concomitant with age. J Bone Joint Surg 1945; 27: 233–247.
Adams P, Muir H . Qualitative changes with age of proteoglycans of human lumbar disc. Ann Rheum Dis 1976; 35: 289–296.
Gower WE, Pegrini V . Age-related variations in proteinpolysaccharides from human nucleus pulposus, annulus fibrosus, and costal cartilage. J Bone Joint Surg 1969; 51A: 1154–1162.
Lipson SJ, Muir H . Experimental intervertebral disc degeneration: morphologic and proteoglycan changes over time. Arthritis Rheum 1981; 24: 12–21.
Garfin SR . The intervertebral disc disease – does it exist?. In: Weinstein JN, Wiesel SW (eds). The Lumbar Spine. W.B. Saunders: Philadelphia, 1990, pp 369–380.
Adler JH, Schoenbaum M, Silberberg R . Early onset of disk degeneration and spondylosis in sand rats (Psammomys obesus). Vet Pathol 1983; 20: 13–22.
Buckwalter JA . Aging and degeneration of the human intervertebral disc. Spine 1995; 20: 1307–1314.
Pearce RH, Grimmer BJ, Adams ME . Degeneration and the chemical composition of the human lumbar intervertebral disc. J Orthop Res 1987; 5: 198–205.
Butler D et al. Discs degenerate before facets. Spine 1990; 15: 111–113.
Urban JP, McMullin JF . Swelling pressure of the intervertebral disc: influence of proteoglycan and collagen contents. Biorheology 1985; 22: 145–157.
Urban JP, McMullin JF . Swelling pressure of the lumbar intervertebral discs: influence of age, spinal level, composition, and degeneration. Spine 1988; 13: 179–187.
Evans CH, Robbins PD . Possible orthopaedic applications of gene therapy. J Bone Joint Surg (Am) 1995; 77: 1103–1114.
Boden SD, Zdeblick TA, Sandhu HS, Heim SE . The use of rhBMP-2 in interbody fusion cages. Definitive evidence of osteoinduction in humans: a preliminary report. Spine 2000; 25: 376–381.
Thompson JP, Oegema Jr TR, Bradford DS . Stimulation of mature canine intervertebral disc by growth factors. Spine 1991; 16: 253–260.
Gruber HE et al. Human intervertebral disc cells form the annulus: three-dimensional culture in agarose or alginate and responsiveness to TGF-beta1. Exp Cell Res 1997; 235: 13–21.
Osada R et al. Autocrine/paracrine mechanism of insulin-like growth factor-1 secretion, and the effect of insulin-like growth factors-1 on proteoglycan synthesis in bovine intervertebral discs. J Orthop Res 1996; 14: 690–699.
Takegami K et al. Osteogenic protein-1 enhances matrix replenishment by intervertebral disc cells previously exposed to interleukin-1. Spine 2002; 27: 1318–1325.
Hallen A . Hexosamine and ester suphate content of the human nucleus pulposus at different ages. Acta Chem Scand 1958; 12: 1869–1872.
Hallen A . The collagen and ground substance of the human nucleus pulposus at different ages. Acta Chem Scand 1962; 16: 705–709.
Aguiar DJ, Johnson SL, Oemega TR . Notochordal cells interact with nucleus pulposus cells: regulation of proteoglycan synthesis. Exp Cell Res 1999; 246: 129–137.
Kang JD et al. Herniated lumbar intervertebral discs spontaneously produce matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2. Spine 1996; 21: 271–277.
Gruber HE, Norton HJ, Hanley Jr EN . Anti-apoptotic effects of IGF-1 and PDGF on human intervertebral disc cells in vitro. Spine 2000; 25: 2153–2157.
Robbins PD, Ghivizzani SC . Viral vectors for gene therapy. Pharmacol Ther 1998; 80: 35–47.
Yang Y et al. Cellular immunity to viral antigens limits E-1 deleted adenoviruses for gene therapy. Proc Natl Acad Sci USA 1994; 91: 4407–4411.
Tripathy SK et al. Immune responses to transgene-encoded proteins limit the stability of gene expression after injection of replication-defective adenovirus vectors. Nat Med 1996; 2: 545–550.
Nita I et al. Direct gene delivery to synovium. An evaluation of potential vectors in vitro and in vivo. Arthritis Rheum 1996; 39: 820–828.
Sawchuk SJ et al. Anti-T cell receptor monoclonal antibody prolongs transgene expression following adenovirus-mediated in vivo gene transfer to mouse synovium. Hum Gene Ther 1996; 7: 499–506.
McCoy R et al. Expression of human interleukin-1 receptor antagonist in mouse lungs using a recombinant adenovirus: effects on vector-induced inflammation. Gene Therapy 1995; 2: 437–442.
Nishida K et al. Immune-Privilege of the Intervertebral Disc: Long-Term Transgene Expression Following Direct Adenovirus-Mediated Gene Transfer. International Society for the Study of Lumbar Spine: Adelaide, Australia, 2000, (Abstr. 69).
Lattermann C et al. Successful AAV-Vector Mediated Transgene Expression in the Intervertebral Disc in Preimmunized Rabbits. Orthopaedic Research Society: Dallas, 2002, (Abstr. 0116).
Wehling P et al. Transfer of genes to chondrocytic cells of the lumbar spine. Proposal for a treatment strategy of spinal disorders by local gene therapy. Spine 1997; 22: 1092–1097.
Nishida K et al. Adenovirus-mediated gene transfer to nucleus pulposus cells. Implications for the treatment of intervertebral disc degeneration. Spine 1998; 23: 2437–2442.
Nishida K et al. Modulation of the biologic activity of the rabbit intervertebral disc by gene therapy: an in vivo study of adenovirus-mediated transfer of the human transforming growth factor beta 1 encoding gene. Spine 1999; 24: 2419–2425.
Moon SH et al. Human intervertebral disc cells are genetically modifiable by adenovirus-mediated gene transfer. Spine 2000; 25: 2573–2579.
Reinke J, Evans C, Robbins PD . Transfer of therapeutic genes to human chondrocytes-like cells of lumbar disc prolapse. Annual Meeting of International Society for the Study of the Lumbar Spine, Singapore 1997; 56:(Abstract).
Moon SH et al. Proteoglycan Synthesis in Human Intervertebral Disc Cells Cultured in Alginate Beads; Exogenous TGF-Beta1 vs. Adenovirus-Mediated Gene Transfer of TGF Beta1 cDNA. Orthopaedic Research Society: Orland, Florida, 2000, (Abstr. 1061).
Eyre D et al. Intervertebral disc: Part B. Basic science perspective. In New Perspectives in Low Back Pain. American Academy of Orthopaedic Surgeons: Park Ridge, IL, 1989, pp 147–207.
Moon S et al. Biologic Response of Human Intervertebral Disc Cell to Gene Therapy Cocktail. Orthopaedic Research Society: San Francisco, 2001, (Abstr. 0295).
Wallach CJ et al. Gene transfer of the catabolic inhibitor TIMP-1 increases measured proteoglycans in cells from degenerated human intervertebral discs. Spine 2003; 28: 2331–2337.
Silberberg R, Aufdermaur M, Adler JH . Degeneration of the intervertebral disks and spondylosis in aging sand rats. Arch Pathol Lab Med 1979; 103: 231–235.
Lipson SJ, Muir H . 1980 Volvo award in basic science. Proteoglycans in experimental intervertebral disc degeneration. Spine 1981; 6: 194–210.
Sobajima S et al. A slowly progressive and reproducible animal model of intervertebral disc degeneration characterized by MRI, X-Ray, and histology. (in press, Spine 2003).
Ye X, Rivera VM, Zoltick P . Regulated delivery of therapeutic proteins after in vivo somatic cell gene transfer. Science 1999; 283: 88–91.
Ralph GS, Bienemann A, Harding TC . Targeting of tetracycline-regulatable transgene expression specifically to neuronal and glial cell populations using adenoviral vectors. Neuroreport 2000; 11: 2051–2055.
Sclementi CR, Baba EJ, Calos MP . An extrachromosomal tetracycline regulatable system for mammalian cells. Nucleic Acids Res 2000; 28: E80.
Annunen S et al. An allele of COL9A2 associated with intervertebral disc disease. Science 1999; 285: 409–412.
Kawaguchi Y et al. Association between an aggrecan gene polymorphism and lumbar disc degeneration. Spine 1999; 24: 2456–2460.
Huang W et al. Phosphorylation of SOX by cyclic AMP-dependent protein kinase A enhances SOX9's ability to transactivate a Col2a1 chondrocyte-specific enhancer. Mol Cell Biol 2000; 20: 4149–4158.
Baer AE et al. Collagen gene expression and mechanical properties of intervertebral disc cell-alginate cultures. J Orthop Res 2001; 19: 2–10.
About this article
Cite this article
Sobajima, S., Kim, J., Gilbertson, L. et al. Gene therapy for degenerative disc disease. Gene Ther 11, 390–401 (2004). https://doi.org/10.1038/sj.gt.3302200
- intervertebral disc
- growth factors
- viral and nonviral vectors
International Journal of Molecular Sciences (2021)
Scientific Reports (2020)
Advances in Clinical Medicine (2020)
Current Pharmaceutical Biotechnology (2020)
Journal of Ethnopharmacology (2020)