Molecular and genetic advances in the regeneration of the intervertebral disc.

BACKGROUND: Owing to the debilitating nature of degenerative disc disease (DDD) and other spine pathologies, significant research has been performed with the goal of healing or regenerating the intervertebral disc (IVD). Structural complexity, coupled with low vascularity and cellularity, make IVD regeneration an extremely challenging task.
METHODS: Tissue engineering-based strategies utilize three components to enhance tissue regeneration; scaffold materials to guide cell growth, biomolecules to enhance cell migration and differentiation, and cells (autologous, or allogeneic) to initiate the process of tissue formation. Significant advances in IVD regeneration have been made utilizing these tissue engineering strategies.
RESULTS: The current literature demonstrates that members of the transforming growth factor beta (TGF-β) superfamily are efficacious in the regeneration of an anabolic response in the IVD and to facilitate chondrogenic differentiation. Gene therapy, though thwarted by safety concerns and the risk of ectopic transfection, has significant potential for a targeted and sustained regenerative response. Stem cells in combination with injectable, biocompatible, and biodegradable scaffolds in the form of hydrogels can differentiate into de novo IVD tissue and facilitate regeneration of the existing matrix. Therapies that address both anabolism and the inherent catabolic state of the IVD using either direct inhibitors or broad-spectrum inhibitors show extensive promise.
CONCLUSION: This review article summarizes the genetic and molecular advances that promise to play an integral role in the development of new strategies to combat DDD and promote healing of injured discs.

INTRODUCTION

The limited regenerative response, biomechanical importance, and structural complexity of the intervertebral disc (IVD) pose significant challenges to both the clinician and researcher in the development of regenerative strategies. Degenerative disc disease (DDD) is the result of adverse loading, dehydration, cellular apoptosis, and an imbalance in tissue anabolism and catabolism. In the setting of DDD, local metabolism favors tissue lysis as manifested by the increased expression of pro-inflammatory cytokines and proteolytic enzymes with a concomitant reduction in matrix anabolism. The presence of catabolic biomolecules that drive this degenerative process not only facilitate tissue degeneration, but they also hinder regenerative efforts.

Rapid advances in our understanding of the biology of the IVD from a genetic and molecular perspective have resulted in several promising methods to facilitate the translation of laboratory-based techniques to clinical disc regeneration. Tissue engineering strategies that employ biomimetic scaffold materials, differentiation-driving bioactive molecules, and multipotent cells to enhance disc regeneration are being developed at a rapid pace. This review compiles and summarizes high-quality basic science studies that employ molecular therapy and tissue engineering-based techniques to aid in regeneration of the IVD. All discussed techniques are summarized in Table 1.

Table 1

Summary table of therapy techniques used in intervertebral disc regeneration

DISC REGENERATION VIA MORPHOGENS AND MITOGENS

Delivery of exogenous proteins: Growth factors and cytokines

The delivery of exogenous proteins in the form of growth factors and/or cytokines is a promising therapy to address DDD. While a wide variety of bioactive proteins and the associated biologic responses they elicit within the IVD have been studied, the principle behind this therapy remains consistent: Altering cellular metabolism and/or phenotype to increase proliferation, extracellular matrix (ECM) synthesis, cell signaling, and/or to downregulate catabolic and inflammatory processes that perpetuate DDD.

Factors with biologic activity: Mitogens, morphogens, and intracellular mediators

Mitogens, morphogens, and intracellular mediators known to be present during neonatal development of the IVD and factors known to induce differentiation of stem cells toward the IVD phenotype have both been identified as candidate therapeutic factors with pleiotrophic biologic activity. Members of the transforming growth factor beta (TGF-β) superfamily, namely TGF-β1 and TGF-β3 in addition to bone morphogenetic proteins (BMPs), as well as insulin-like growth factor-1 (IGF-1), growth and differentiation factors (GDFs), epidermal growth factor (EGF), platelet-derived growth factor (PDGF), basic fibroblastic growth factor (bFGF), and others, have all been studied in the context of IVD regeneration.

Transforming growth factor beta superfamily; morphogens and biologic mediators of intravertebral disc development

Members of the TGF-β superfamily are morphogens and biologic mediators during neonatal development of the IVD.[16] TGF-β is present in lower levels in herniated discs,[27] and its concentration in the IVD has been shown to decrease with age.[4254] Early in vitro studies by Risbud et al. demonstrated that the stimulation of proteoglycan (PG) synthesis in IVDs by transforming growth factor-beta 3[72] (TGF-β3) increases proliferation and ECM macromolecule synthesis via the extracellular signal-regulated kinase/mitogen activated protein (ERK/MAP) signaling pathway.[61] Human annulus fibrosus (AF) cells grown in 3D agarose or alginate systems have shown increased proliferation following treatment with TGF-β1[14] Abbott et al. demonstrated that TGF-β3 in combination with dexamethasone (Dex) induces a decreased expression of the catabolic genes MMP1 (encoding for matrix metalloproteinase-1) and ADAMTS5 (encoding for a disintegrin and metalloproteinase with thrombospondin motifs 5) and increased expression of ECM genes.[1]

In-vitro study of bone morphogenetic protein-2 for IVD regeneration

BMPs, named after the osteoinductive potential of some of its members, have also been extensively studied in the context of IVD regeneration. Bone morphogenetic protein-2 (BMP-2) and bone morphogenetic protein-7 (BMP-7 or OP-1) are the only human recombinant growth factors approved by the Federal Drug Administration (FDA). Although BMP-2 is an osteoinductive growth factor approved for use in lumbar spine fusion, off-label use is wide-spread, and has been implicated as a potential source of numerous complications.[80] In vitro treatment of IVD cells with BMP-2 upregulates genes for aggrecan, type II collagen, and sex determining region Y-Box9 (Sox9). It increases protein expression of proteoglycans and type II collagen.[3573] without inducing an osteogenic phenotype and/or associated calcification or mineralization.[2443]

In-vivo study of bone morphogenetic protein-2 for IVD regeneration

The regenerative potential of BMPs in the setting of disc regeneration has also been studied with animal models. In an in vivo model of an annular tear, Huang, et al.[20] demonstrated that administration of rhBMP-2 with or without coral, actually caused a greater degree of degenerative changes in the IVD compared with controls. Additionally, the study demonstrated that rather than regenerating the AF with new fibrous tissue, several animals exhibited bony fusions. The study also noted increased inflammation, vascularity, and fibroblastic proliferation in rhBMP-2-treated animals compared with saline-treated animals. Future in vivo studies are necessary to understand the apparent discrepancy between in vitro and in vivo findings.

BMP-7 (OP-1) induces similar in vitro effects as BMP-2 in IVD cells

BMP-7 or OP-1 induces similar in vitro effects as BMP-2 by stimulating proliferation, proteoglycan, and collagen synthesis in IVD cells.[25417190] Furthermore, BMP-7 exhibits antifibrotic, antiapoptotic, antiinflammatory, and proangiogneic properties.[8687] Takegami, et al. demonstrated that primary rabbit AF and nucleus pulposus (NP) cells previously exposed to interleukin-1 beta (IL-1β) recovered their proteoglycan and collagen synthesis after BMP-7 treatment.[41] Im, et al. showed that BMP-7 decreased IL-1β-induced expression of matrix metalloproteinase-13 (MMP-13) in chondrocytes.[21] Masuda, et al. studied the in vivo effect of 10[72] μg/mL BMP-7 injection into rabbit nucleus pulposi following the induction of degeneration using the needle puncture model.[40] The study showed recovered disc height, improved magnetic resonance image (MRI) findings, and higher proteoglycan content in both the AF and NP in the BMP-7-treated animals compared with control (lactose-treated) animals. The same group performed a similar experiment on rabbit discs treated with chondroitinase-ABC to induce degeneration.[22] They observed a recovery of disc height and higher proteoglycan content, but no significant histological improvement.

As very high concentrations of BMP-7 were used in these studies, future experimentation is necessary to determine whether less protein/BMP-7 would achieve comparable positive results. Further research should also explore the variable efficacy of BMP-7 treatment in a model using annular puncture versus a model employing chemical digestion (e.g., chondroitinase-ABC injection). This research should also include an assessment of whether a proteinase inhibitor adds a synergistic effect by counteracting catabolic processes.

In vitro and in vivo studies of BMP-14 (GDF-5): Morbidity and questions regarding impact of dosing

Similar to BMP-2 and -7, bone morphogenetic protein-14 (BMP-14), also known as growth and differentiation factor-5 (GDF-5), or cartilage-derived morphogenetic protein-1 (CDMP-1), was initially investigated as an osteoinductive protein for use in spine fusion. BMP-14 is present during precartilaginous mesenchymal condensation and localized in the cartilaginous cores of long bones during embryogenesis and skeletal development.[4] While its precise role in IVD development, metabolism, degeneration, and regeneration is not fully known, GDF-5 deficiency causes a variety of ECM abnormalities and defects in the IVD.[36]

In vitro, BMP-14 induces cellular activation and proliferation in primary IVD cells.[78] When compared with BMP-2 and BMP-7, BMP-14 appears more effective at promoting collagen synthesis, whereas BMP-2 and BMP-7 stimulate proteoglycan synthesis.[91] In a murine model of disc degeneration, a single 1 μg/mL injection of GDF-5 caused increased disc height compared with saline-injected controls.[77] Multiple 1 μg/mL injections of GDF-5 (once per week for 4 weeks), however, caused inflammatory reactions in the adjacent vertebrae and connective tissue that infiltrated the nucleus. These data indicated that the positive effect of GDF-5 may depend on dosing, and that multiple treatments can cause catabolic and/or inflammatory processes.

Chujo, et al. demonstrated that primary bovine AF and NP cells cultured in alginate beads and treated with rhGDF-5 exhibited significantly increased DNA content, proteoglycan synthesis, and collagen synthesis.[6] Following their in vitro experimentation, injections of rhGDF-5 were assessed in an annular puncture model of disc degeneration in rabbits. Results demonstrated increased disc height, improved scores on MRI grading, and improved histologic scores. Improved healing as measured by disc height, MRI grade, and histologic score was, however, not exhibited with increasing doses.

More work is needed to ascertain effects of lesser known bone morphogenetic proteins family members

BMP-4

The BMP family has numerous members, which have yet to be investigated in terms of their in vitro, or in vivo regenerative potential for disc regeneration. Although BMP-4 is a stimulator of chondrogenesis, little work has been done to study the specific effect of BMP-4 on the IVD.[44] Many studies have assessed the regenerative effect of BMP-4 on articular cartilage,[2937] articular cartilage chondrocytes,[57] and on the differentiating effect on MSCs toward chondroprogenitor cells.[44] Similar to BMP-4, little high-quality work has been done to study the regenerative effect of BMP-13 (also termed GDF-6 or CDMP-2) on DDD.

BMP-13

BMP-13, a member of the BMP family with only 50% homology to BMP-2, is highly evolutionarily conserved across mammals (95% +), which suggests an important biologic function. It has been detected in hypertrophic chondrocytes in ossifying long bone centers, but its exact role in the IVD remains unknown.[38] Li, et al. showed that murine chondrocytes treated with BMP-13 increased total proteoglycan production and aggrecan mRNA in a dose-dependent manner, but that this effect was not synergized by the addition of BMP-2.[34] Future work is necessary to study the efficacy of BMP-13 on IVD regeneration.

Regenerative capacity of mitogenic factors

IGF-1, PDGF, EGF, and bFGF are mitogenic factors of the IVD. While they do not act as chondrogenic factors like the morphogenic molecules previously outlined, they act as “growth factors” by increasing the rate of mitosis. The earliest work with mitogenic molecules in the context of IVD regeneration was performed by Thompson et al. in 1991.[66] This study showed that mitosis is affected by mitogenic factors in canine IVD cells.[72] After Okuda, et al. demonstrated an age-related, IGF-1-dependent decline in proteoglycan synthesis in rat IVDs,[53] they suggested that IGF-1 therapy may be beneficial in reversing age-related degeneration.

Although most authors have used mitogenic factors in combination with morphogens in regenerative therapeutic approaches, some studies have used them in isolation. Walsh, et al.[77] showed that a single injection of IGF-1 into a degenerate murine disc increased cell density and proliferation at 1 week, but this effect was not sustained at later time points. Furthermore, multiple IGF-1 injections caused inflammatory reactions in adjacent vertebrae, indicating a dose-dependent response. When administered in combination with BMP-7, IGF-1 caused a synergistic effect on anabolic BMP-7. This was attributed to signaling by activating the Smad2 pathway and, consequently, causing increased proteoglycan synthesis. IGF-1 and PDGF both possess in vitro antiapoptotic action,[15] but this effect has not been shown in vivo. In vitro, EGF was found to increase activation and proliferation,[72] but bFGF by itself did not elicit a significant proliferative or anabolic effect when injected into a degenerate disc in vivo.[77]

CELL-BASED REGENERATIVE STRATEGIES

Cell-based strategies for IVD regeneration represent an important family of techniques that are rooted in the hypothesis that scaffolds and growth factors alone are insufficient for the regeneration of tissues. Delivery of exogenous cells to injured tissues has been shown to be effective in a variety of clinical scenarios. In the setting of the IVD, these techniques can be classified by the cell types utilized to effect regeneration, including chondrocytes and fibrochondrocytes from the IVD, notochordal cells (NCs), and mesenchymal stem cells (MSCs).

Mesenchymal stem cells

MSCs are multipotent cells that are capable of migration and homing, as well as differentiation toward chondrocytic, osteoblastic, adipocytic, myofibroblastic, and tenofibroblastic lineages.[8] MSCs are found in numerous anatomic locations, including the bone marrow where they associate with stromal cells in a stromal cell-derived factor-1/C-X-C chemokine receptor type 4 (SDF-1/CXCR4)-dependent fashion until mobilized by the inflammatory cascade initiated by tissue injury.[69] They maintain an inherent stemness and are capable of self-renewal by expressing specific self-renewal genes such as Oct-4, Sox-2, and Rex-1.[23] The International Society for Cellular Therapy (ISCT) published criteria by which to identify MSCs, which include positive expression of cluster of differentiation (CD) 73, CD90, and CD105 and negative expression of CD45, CD34, CD14 or CD11b, CD79α, or CD19 and human leukocyte antigen-DR (HLA-DR).[8]

Marrow-derived MSCs differentiate toward a nucleus pulposus-like phenotype

Risbud, et al. demonstrated that marrow-derived MSCs are capable of differentiating toward a NP-like phenotype under the in vitro conditions of hypoxia and exogenous TGF-β1.[60] Marrow-derived MSCs from rats also exhibit the ability to differentiate into NP-like cells when co-cultured with intact IVDs.[82] Using an organ culture model, Le Maitre, et al. showed that intradiscal injection of human, marrow-derived MSCs were not only capable of surviving, but also induced expression of SOX-9, aggrecan and type II collagen indicative of differentiation toward a NP-like phenotype.[31] Richardson et al. found that direct contact between MSCs and NP chondrocytes stimulated the differentiation of the MSCs toward an NP-like lineage, as illustrated by upregulation of SOX-9, type I, II, and VI collagen, as well as aggrecan, versican, and elastin.[59]

In vivo experimentation demonstrates the promise of marrow-derived mesenchymal stem cells

In vivo experimentation has also demonstrated the promise of marrow-derived MSCs. Transplantation of marrow-derived MSCs into rabbit degenerative discs enhanced glycosaminoglycan (GAG) content and gene-level expression of aggrecan and type II collagen. Hiyama et al. transplanted marrow-derived MSCs into previously nucleotomized discs.[19] They demonstrated not only histologic and reverse transcription polymerase chain reaction (RT-PCR) evidence of disc regeneration, but that the transplanted cells could survive 8 weeks posttransplantation.[19]

Multiple niches of mesenchymal stem cells display regenerative properties

There are several other niches of MSCs that display regenerative properties. Utilizing a pellet co-culture of synovium-derived MSCs and NP chondrocytes, Chen, et al. demonstated NP-like differentiation.[5] He and Pei showed that culturing NP chondrocytes in plastic flasks coated with ECM produced by synovium-derived MSCs resulted in increased NP cell proliferation and redifferentiation capacity.[17] Transplanting synovium-derived MSCs immediately after nucleotomy in a rabbit model has also been shown to prevent radiographic, MRI, and histologic evidence of disc degeneration. The transplanted cells suppressed the expression of genes, which encode for MMPs, while also enhancing matrix synthesis.[47] Murrell, et al. transplanted olfactory neurosphere-dervied stem cells into nucleotomized rat discs, and observed increased type II collagen and chondroitin sulfate proteoglycans within the disc tissue.[50] In a canine nucleotomized disc model with transplanted adipose-derived MSCs with a hyaluronic acid carrier, Ganey, et al. demonstrated radiographic, MRI, histologic, and biochemical evidence of disc regeneration.[10]

INTERVERTEBRAL DISC-DERIVED CELLS

Cells derived from intervertebral disc respond may halt degeneration and rejuvenate endogenous cells

Cells derived from the IVD have been studied extensively in terms of their response to soluble factors, variable culture conditions and paracrine signaling. In contrast to MSCs, IVD-derived cell therapy focuses on utilizing cells of a committed phenotype to halt the process of degeneration and, in some instances, rejuvenating endogenous cells that have been damaged by aging, or pathology. In 2003, Watanabe, et al. examined a novel technique of “activating” autologous NP cells via co-culture with AF cells and subsequently reinserting them into the degenerate disc.[79] Degenerative changes in discs that were reinserted with activated NP cells were less severe than in sham-treated discs. However, the results from this study must be interpreted with care, as the model used was a rabbit, which is known to retain NCs in the disc much longer than other animals. Xin, et al. injected unmodified NP cells and NP cells transduced with telomerase reverse transcriptase (hTERT) into explanted canine whole IVDs, which were subsequently transplanted into donor canines.[84] NP cell modification of the allograft discs resulted in significantly less degeneration in the recipients, with hTERT-transduced NP cells offering the greatest protection. Ganey, et al. demonstrated that reimplantation of ex vivo cultures of autologous disc chondroctyes resulted in the production of proteoglycans and a matrix similar to the native disc, while maintaining disc height.[11] The study also highlighted that the disc chondrocytes maintained viability at all examined end points.

Notochordal cells represent a promising cell population for regeneration

NCs represent a promising cell population for regenerative approaches for the IVD. NCs are involved in the formation and development of the IVDs in early life. Early work by Aguiar, et al. demonstrated that coculture of NCs with NP cells resulted in significant increases in proteoglycan content in both a juxtacrine and paracrine fashion.[2] Conditioned media from NCs was shown to be more effective than traditional chondrogenic media (containing Dex and TGF-β3) in terms of inducing NP-like differentiation in MSCs.[28] Erwin, et al. found that NP cells cultured with NC conditioned media displayed reduced apoptosis via inhibition of caspase-3/7 and -9.[9] The NP cells also showed reduced expression of matrix catabolic factors, including MMP-3, while showing upregulation of genes associated with matrix anabolism, including aggrecan, collagen type II, and tissue inhibitor of MMP-1.

ADVANCES IN GENE THERAPY AND DELIVERY

Gene therapy aims to alter the genome of a cell population to elicit a desired molecular effect – generally the increased or decreased production of a gene product to increase/decrease the production of a protein of interest. While the direct administration of recombinant proteins has been shown to be successful, the short half-life of a protein, unknown protein–protein interactions, localized and undesirable thermodynamic phenomena, and cost are all considered disadvantages of this type of therapy.

Altering genotype of cell population, transcription/translation sustains protein synthesis

By altering the genotype of a cell population of interest, gene transcription and translation can be “reprogrammed” to invoke sustained protein synthesis. Diffusion properties of most delivery vehicles containing proteins (e.g., collagen sponge containing BMP-2) cause a burst release followed by a rapid decrease in protein concentration. This burst release, often times over 90% of the total protein in the first 24 hours, can cause protein concentrations to exceed the upper therapeutic threshold, rendering the therapy inefficient. Gene therapy induces a sustained, biomimetic production and release of a protein and can be tailored to adjust the magnitude of total release.

Direct in vivo gene therapy: Direct administration of infecting agent(s) to host

Direct, or in vivo, gene therapy involves the direct administration of the infecting agent into the host.[68] While this modality is less invasive and less costly, the efficiency and location of infection is highly variable, and the desired effect may be difficult to achieve with a single injection. Some authors have demonstrated active infection of rabbit disc cells for up to 2 years, but infection rate and duration depend on cell type and environment. Indirect, or ex vivo, gene therapy necessitates the isolation of host cells, subsequent in vitro infection of these cells, followed by surgical reimplantation of gene-modified cells. While this therapy is costly and requires multiple surgical procedures, the efficiency of infection is high and ectopic infection in undesirable anatomic locations can be avoided. However, cell viability can be lower, and cells may need to be reimplanted in a bioconductive scaffold to ensure cellular incorporation. Cells do not uptake DNA molecules by themselves, and, as such, delivery vehicles must be utilized to transfer the gene of interest into the cell to cause efficient infection. Recombinant adenoviral vectors, adeno-associated virus (AAV), or nonviral vectors can be utilized to deliver a gene of interest.[68] The multiplicity of infection (MOI) describes the ratio of infecting agents (i.e., the vector containing the gene) to host agents (i.e., the cell) and is optimally minimized to reduce cost and avoid ectopic infection.

Safety concerns and positive results with gene therapy

There are significant safety concerns regarding the clinical in vivo use of gene therapy. Major hurdles to the “safe” in vivo use of gene therapy include ectopic transfection, patient-specific dose responses, immune reactions to viral proteins, and unknown side effects. Following the death of a seemingly healthy patient in a clinical trial of gene therapy for liver disease in 1999,[39] extensive concern triggered research studies assessing the potential side effects of gene therapy. While such studies related to IVD regeneration are limited, some authors have drawn conclusions regarding the safety of gene therapy in the spine. In an animal model, Wallach, et al. performed intradural injections of adenoviral TGF-β1 to study how the spinal cord would react to ectopic adenoviral therapy.[75] They found no adverse clinical symptoms or histologic abnormalities in animals that received control (saline injections), 10[61] plaque-forming units (PFU) of Ad-TGF-β1, 10[1] PFU of Ad-BMP-2, or injections of rhTGF-β1. However, animals that received supra-therapeutic doses of Ad-TGF-β1 developed bilateral lower extremity paralysis on postoperative day 10. Other evidence confirming the potential risks of supra-therapeutic gene therapy in the spine included the documentation of: Dural fibrosis, inflammation and fibroblastic infiltration of dural tissue.[75]

Numerous other studies, however, document the safety and efficacy of in vivo gene therapy. Wehling, et al. were the first to employ gene therapy to address IVD degeneration.[81] In their study, primary bovine chondrocytes isolated from vertebral end plates were infected with retroviruses containing bacterial β-galactosidase (LacZ) and the complementary DNA (cDNA) of human interleukin-1 receptor antagonist (IL-1Ra), the direct inhibitor of IL-1 receptors. They demonstrated that sustained release of a therapeutic protein was feasible in IVD cells; X-gal staining indicated only 1% of the total cell population was infected with the LacZ gene, and 24 ng/mL/10[61] cells of IL-1Ra was secreted in 48 hours.

In a later study, Le Maitre, et al. also utilized IL-1Ra gene therapy in attempts to inhibit molecular processes of DDD in vitro.[32] Degenerate human IVD cells were infected with adenoviral vectors containing the IL-1Ra gene (Ad-IL-1Ra) and showed sustained release of IL-1Ra. Infected cells were demonstrated to be IL-1-resistant, as measured by IL-1-mediated MMP-3 and MMP-13 gene expression. The same group later studied IL-1Ra gene therapy on explants of degenerate human IVDs and found that this approach inhibits IL-mediated MMP production and, consequently, matrix degradation.[33]

In early work, Nishida, et al. delivered adenoviral TGF-β1 into lumbar IVDs of rabbits.[52] Culturing isolated cells after 1 postoperative week, revealed significantly more in vitro TGF-β1. Immunohistochemical staining of whole IVDs also showed significantly more tissue staining positive for TGF-β1 than control tissue. Recently, Leckie, et al. treated rabbits with adeno-associated virus sero-type 2 (AAV2) vectors carrying either BMP-2 or TIMP-1 following annulotomy. Serial quantitative T2 mapping, histologic analysis, and biomechanical testing were performed to compare healthy controls, defect-only, and gene therapy-treated animals. In addition, the study quantified circulating concentrations of C-telopeptide of collagen type II (CTX-II), a serum biomarker previously shown to correlate to DDD.[125870] AAV2-TIMP2-treated animals exhibited a significant difference in MRI index and quantitative T2 values compared with defect-only animals at 6 weeks. Biomechanical testing demonstrated that AAV2-BMP2 and AAV2-TIMP2-treated groups had mechanical properties that were more representative of healthy animals, although these findings were not indicated to be significant. Qualitative histologic analysis showed preservation of NP area, architecture, and cellularity in treated groups compared with defect-only animals. Significantly lower concentrations of CTX-II were detected in treated animals at 12 weeks compared with defect-only animals.

Adenoviral vectors expressing numerous bone morphogenetic proteins

Zhang, et al. undertook a study to assess the effects of adenoviral vectors expressing numerous BMPs and the transcription factor Sox9 on ECM production of bovine NP cells.[91] Proteoglycan production, collagen production, and cellular proliferation were measured 6 days following infection. BMP-2, -3, -4, -5, -7, -8, -10, -13, -15-infected and Sox9-infected cells all exhibited significantly more proteoglycan accumulation compared with controls. BMP-2, -4, -5, -7, -8, -10, -14, -15-infected and Sox9-infected cells exhibited significantly more collagen accumulation compared with controls. Only BMP-2 and -8 increased cellular proliferation. One of the study's most significant findings was that BMP-2 and -7-transduced cells exhibited similar proteoglycan production as cells treated with 100 ng/mL rhBMP-7 for one day.

Sox9 gene delivery in human IVD cells

Sox9 gene delivery has also been studied in human IVD cells. Paul et al. demonstrated that Sox9 gene delivery increased Type II collagen production in primary cells from degenerated IVDs; this finding was duplicated utilizing intradiscal injections of Ad-Sox9 in rabbits.[56]

Primary human IVD cells respond favorably to adenoviral infection with growth factors

Primary human IVD cells have also been shown to respond favorably to adenoviral infections with TGF-β1, IGF-1, and BMP-2 in another study.[48] The study demonstrated increased proteoglycan production due to treatment with each single adenovirus. However, and more interestingly, they demonstrated the largest proteoglycan production due to infection with the triple combination adenovirus containing all three genes. This suggested that gene therapy with both morphogens and mitogens combined could exhibit synergistic effects.

Treatment of IVD cells with lim mineralization protein-1 causes increased production of proteoglycans both in vitro and in vivo

Yoon et al. demonstrated that adenoviral treatment of IVD cells with LIM Mineralization Protein-1 (LMP-1), an intracellular regulator known to induce the production of BMPs in osteoblasts,[46] causes increased production of proteoglycans both in vitro and in vivo.[85] The study showed that rat disc cells respond to Ad-LMP-1 in a dose-dependent manner. In addition, this process was BMP-dependent, as BMP inhibition with noggin diminished the production of sGAGs. In vivo, rabbit IVDs were treated with an optimized MOI of Ad-LMP-1, and qPCR of harvested NP at 3 weeks demonstrated upregulation of aggrecan, BMP-2, and BMP-7 mRNA.

MULTI-FACETED APPROACHES TO DISC REGENERATION

The three vertices of the tissue engineering triangle include scaffolds, cells, and growth factors. This section of the review will focus on summarizing in vivo studies that describe strategies that are dependent on two or more of the vertices to effect IVD regeneration.

In vivo performance of cell-seeded scaffolds

Scaffolds represent a multi-functional tool in tissue engineering. The scaffold structure provides mechanical support to resist biomechanical loading, while guiding new tissue growth by providing compositional and structural cues to cells. Sato, et al. seeded allogenic AF cells on an atelocollagen scaffold, with a honeycomb morphology.[66] Cells remained viable throughout the in vitro culture period, and were also able to prevent disc space collapse when transplanted into the nuclear space of the rabbit IVDs. Sakai, et al. also used an atelocollagen scaffold, but transplanted marrow-derived MSCs, rather than AF cells into the discs of rabbits.[65] The study demonstrated radiographic, MRI, and histologic evidence that MSC-seeded ateolocollagen scaffolds were able to restore disc height, promote proteoglycan synthesis, and preserve normal disc morphology. Huang, et al. utilized a tri-copolymer consisting of collagen type II, hyaluronan, and chondroitin-6-sulfate to enhance NP tissue regeneration following nucleotomy in a rabbit model.[20] Disc regeneration was markedly improved when the tri-copolymer constructs were cultured with allogenic NP cells for one week prior to implantation. Using a canine model, Ruan, et al. investigated the effects of NP cell-seeded poly(lactic-co-glycolic acid) (PLGA) scaffolds on nucleotomized discs.[63] They found that the segmental stability, and disc height were preserved in the animals receiving the cell-seeded scaffolds, while the scaffold alone was insufficient to promote these “regenerative” changes.

HYDROGEL-BASED SCAFFOLDS: A PROMISING METHOD FOR NUCLEUS REGENERATION

Hydrogel-based scaffolds also represent a promising method for nucleus regeneration. These polymers can be synthesized to be responsive to environmental stimuli, and to promote cell attachment. The gel-like consistency also mimics the gelatinous NP. Henriksson et al. showed that human MSC's were more effective in contributing to IVD tissue regeneration when suspended in a peptidic hydrogel (PuraMatrix, BD Biosciences), versus cell culture media.[18] Cells transplanted into degenerate porcine IVDs with the peptide-based hydrogel exhibited greater in vivo viability, and enhanced production of type II collagen, aggrecan, and type I collagen.

In vivo growth factor delivery from scaffolds

In addition to providing a synthetic microenvironment to facilitate cell delivery to the IVD, scaffolds can also be implemented as growth factor delivery systems. Most often, the release kinetics of the growth factor are dependent on the degradation profile of the carrier system. Incorporation of growth factors within polymeric carriers delays proteolysis and also facilitates the sustained and localized release of growth factors, or drugs. Zhang, et al. utilized a polyethylene glycol - poly(lactic-co-glycolic acid) - polyethylene (PEG-PLGA-PEG) thermosensitive gel to deliver simvastatin intradiscally in a rat disc stab injury model.[89] They demonstrated that the slow release of simvastatin from the thermosensitive and resorbable PEG-PLGA-PEG gel “regenerated” the degenerative disc, and “slowed” the degenerative process. Nagae, et al. described using gelatin hydrogel microspheres for the sustained, intradiscal delivery of platelet rich-plasma, an autologous depot of growth factors and cytokines.[51] Implanted gelatin microspheres could not be discerned at 8 weeks postimplantation, but increased immunohistochemical evidence of proteoglycan production was observed. In 2009, the group used similar gelatin microsphere technology to again deliver platelet rich plasma in a degenerative rabbit disc model, but studied the effects in greater detail.[67] They demonstrated a lower number of apoptotic cells in the nucleus, and higher mRNA levels of proteoglycans and collagen type II.

ENHANCING DISC REGENERATION BY SLOWING THE DEGENERATIVE PROCESS

Catabolism and anabolism are clearly not in balance in the setting of DDD. Promising therapeutic approaches for IVD regeneration, therefore, include downregulating the expression or directly inhibiting inflammatory and/or catabolic moieties are. In vitro, Kuroki, et al. demonstrated that canine chondrocytes cultured in 3D produced decreased concentrations of MMP-1 and MMP-3 when treated with TIMP-1 and TIMP-2. However, IL-1β-associated matrix degradation was not significantly thwarted, indicating the need to address pathways of degradation not inhibited by TIMP-1 and TIMP-2.[30]

Proteinase inhibitors utilized in gene therapy-based approaches

Proteinase inhibitors have also been employed in gene therapy-based approaches. Wallach et al. delivered adenovirus TIMP-1 to human primary cells isolated from degenerate IVDs.[76] They demonstrated the in vitro benefit of downregulating catabolic processes to allow anabolism to predominate when they showed that gene delivery of TIMP-1 increased proteoglycan production in 3D pellet cultures.

Statins act as intracellular inhibitors of MMP production and enhance BMP-2 mRNA expression

Statins are a class of drugs approved for lowering cholesterol, but their activity as intracellular inhibitors of MMP production make them an attractive therapeutic option in orthopedics. By inhibiting the enzyme 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMG-CoA reductase), MMP transcription is inhibited by downregulating the RhoA/ROCK pathway necessary for MMP production.[64] Barter, et al. showed that statins can reverse the production of MMP-1 and MMP-13 induced by interleukin-1 and oncostatin-M (OSM) in articular chondrocytes and articular cartilage tissue.[3] Simvastatin also specifically inhibits MMP-9 production by saphenous vein smooth muscle cells.[74]

Interestingly, statins have also been shown to increase BMP-2 mRNA in osteoblasts in vitro and cause mature bone formation in vivo.[49] Zhang, et al. has studied the effect of simvastatin both in vitro[88] and in vivo[89] in the context of IVD regeneration. In their in vitro work, they showed that simvastatin upregulated BMP-2 mRNA expression in rat IVD cells and consequently caused increased expression of type II collagen mRNA and proteoglycan. The expression of type II collagen and proteoglycan were dependent upon the expression of BMP-2, as shown by the inhibition of BMP-2 by 500 ng/mL noggin, which reduced the expression of collagen and proteoglycans. Furthermore, the study demonstrated that simvastatin directly facilitates the BMP-2 pathway by inhibiting HMG-CoA reductase; this process can be reversed by the addition of melavonate.[88] In vivo, the same group demonstrated the clinical applicability of the direct administration of simvastatin. They showed that the intradiscal injection of a synthetic hydrogel loaded with 5 mg/mL simvastatin following annular puncture in rats significantly upregulated BMP-2 and type II collagen mRNA while also reversing degenerative changes attributed to the annular puncture.[89]

Lactoferricin causes increased PG accumulation and downregulated catabolic processes

In a recent study, Kim and coworkers treated bovine NP cells with lactoferricin, an amphiphatic heparin-binding glycoprotein known for its antiviral, antioxidant, antioncologic, and analgesic properties,[13] and measured the expression of MMPs, ADAMTSs, TIMPs,[26] and ECM macromolecules. Results show that lactoferricin treatment caused increased PG accumulation and downregulated catabolic processes by increasing the expression of several TIMPs in a dose-dependent manner via the ERK and p38 protein kinase pathways.

TNFα-stimulated gene product TSG-6 with inter-α-inhibitor downregulate MMP activation

TNFα-stimulated gene product (TSG-6) is a 35 kD glycoprotein known to be involved in the inflammatory cascade. It has been implicated as an early marker of osteoarthritis in mice,[45] and shown to have antiinflammatory effects in vivo.[83] When TSG-6 forms a complex with inter-alpha-inhibitor (IαI), it is hypothesized to downregulate MMP activation, therefore thwarting the catabolic processes active in disease states such as osteoarthritis or DDD.[45] Roberts, et al. collected 58 herniated discs from 54 patients, and found TSG-6 and its binding protein, IαI, to be present in nearly all discs.[62] While the precise function of TSG-6, IαI, and their interaction is not fully known, future research is merited to study how these molecules can be exploited in anticatabolic treatment approaches.

Other pharmacologics being studied for their impact on IVD regeneration and/or discogenic back pain

Other pharmacologics currently utilized in other diseases have been studied in the context of IVD regeneration and/or discogenic back pain. Etanercept is an approved TNF-α inhibitor used primarily for the treatment of rheumatoid arthritis. Direct injections of TNF-α inhibitor into perispinal tissue constitute a potential therapeutic approach for treating chronic discogenic pain.[7] The oral administration of CPA-926 an esculetin prodrug, has shown efficacy in delaying the onset of disc height loss following annular puncture in rabbits.[55] The precise molecular mechanisms of these results are, however, not fully outlined and need to be studied in greater detail.

SUMMARY

Many basic science studies, utilizing molecular therapies and tissue-engineering-based strategies, are attempting to regenerate IVDs. Their aim is to treat/reverse DDD, and thereby, devise methods for treating low back pain. Although members of the TGF-β superfamily appear to be the most promising molecules to address the thwarted anabolic processes in DDD, future research should continue to address the catabolic and inflammatory environment of the IVD. The evolution of gene therapy, namely a more extensive knowledge of ectopic transfection, optimizing transfection efficiency, and developing strategies for more localized delivery will play an important role in the future development of IVD regeneration. The continual development of cheap, easily manufactured, biocompatible, and biodegradable scaffolds able to be injected or implanted in the IVD should also be at the forefront of IVD regeneration research. Last, the rapid, intraoperative isolation of MSCs, autologous disc cells, or cartilage progenitor cells to provide cheap and biocompatible cell sources for reimplantation in combination with scaffolds and/or signaling molecules will likely be the most important area of future engineering-based research therapies.