|Home | About | Journals | Submit | Contact Us | Français|
The therapeutic strategies that have thus far been employed for the treatment of intervertebral disc degeneration (IDD) have focused on relieving the symptoms, while reversal of the degeneration remains an important challenge for the effective treatment of IDD. Growth and differentiation factor-5 (GDF5), of which deficiency leads to early disc degeneration changes, has the potential to increase proliferation of disc cells and expression of extracellular matrix proteins.
To develop a lumbar disc degeneration model in mice and determine the effect of adenoviral GDF5 gene therapy.
Compare the degeneration changes of discs punctured by different size needles to develop a mice lumbar disc degeneration model. Evaluate the effects of in vivo gene therapy for the mice disc degeneration model by an adenoviral vector carrying GDF5 gene.
A lumbar disc degeneration model was developed by needle punctures to the discs in Balb/c mice. Afterwards, a gene therapy treatment to disc degeneration was evaluated. Two of the mice lumbar discs were randomly chosen to be punctured by a 30- gauge needle and then injected with adenovirus that had been engineered to express either the luciferase gene (Ad-Luc) or the GDF5 gene (Ad-GDF5). Animals were analyzed by bioluminescent imaging, radiographic and MRI scanning, then sacrificed at 1-, 2-, 4-, or 8- week post operation and subjected to histological and biochemical assays.
By the detection of T2-weighted MRI scanning and histological study, the degeneration was found in all of the discs punctured by different size needles. But the development of the degeneration in the discs injured by 30-gauge needle was more reliable and moderate compared with other groups. The detection of luciferase activity by bioluminescent imaging revealed that adenovirus survived and the introduced genes were expressed over 6 weeks after injection. There were no T2-weighted MRI signals in either the Ad-Luc or Ad-GDF5 injected mice up to 4 weeks post operation. At 6 and 8 weeks, T2-weighted signals were detected in the Ad-GDF5 group, but none in the Ad-Luc control group. The percent disc height index (%DHI) was significantly decreased (~ 20%) by 1 week following injury in both groups, indicating the development of disc degeneration. At 2 weeks, the %DHI in the mice injected with Ad-GDF5 increased significantly compared with that of the mice injected with Ad-Luc group; the increase was sustained for the rest of experiment period. The disc histology treated with Ad-GDF5 was improved compared with that in control group. Glycosaminoglycan (GAG) levels were significantly decreased in the Ad-Luc injection group since 2 weeks after injury, and the DNA content had diminished by 4 weeks after the operation. In contrast, in the discs injected with Ad-GDF5, there was no decrease in the GAG and DNA levels following injury throughout the 8 weeks treatment period.
Disc degeneration animal model can be developed by using needle puncture to the discs in mice. The adenovirus is an effective vehicle for gene delivery with rapid and prolonged expression of target protein, and resulting improvement in markers of disc degeneration. Ad-GDF5 gene therapy could restore the functions of injured discs and has the potential to be an effective treatment.
In Western societies, intervertebral disc degeneration (IDD) related lower back pain affects between 60% and 80% of the population at some point in their lives, making it one of the most important public health issues today. [1,2] IDD is thought to result from multifatorial causes. [3–7] However, regardless of the etiological factors, the major pathological manifestations of IDD are the progressive declining proteoglycan content, disappearance of nucleus pulposus and collapse of annulus fibrosus. [8,9] Finally, the height of the intervertebral discs is reduced and osteophytes appear at the anterior and lateral edges of the disc. 
The treatments available for IDD include conservative methods and surgical therapies. The traditional surgical discectomy does not reconstitute the functions of disc or allow the spine to regain its characteristic multi-directional flexibility conferred by unique combination of properties of the annulus and nucleus. The challenge of developing a therapeutic method that is able to restore disc functions has therefore been a long-standing aim of researchers.
Recently, the efficacy of a variety of growth factors for the IDD treatment has been investigated [11–15], which included growth and differentiation factor 5 (GDF5). [12,15] The concept underlying these investigations was that growth factors will improve cells proliferation and extracellular matrix recovery. GDF5, as a member of the BMP family, plays an important role in the growth and development of the skeletal system. In humans, the loss of GDF5 gene causes chondrogenic dysplasias, [16,17] while GDF5 deficient mice showed early disc degeneration changes.  In vitro, incubation with GDF5 can augment the proteoglycan and collagen production by chondrocytes or disc cells. [15,18–20] Furthermore, the injection of GDF5 protein into rabbit discs can promote cell proliferation.  However, the treatment was found to be dose-dependent.  Such dose-dependency increases the challenge of translating such a therapy to larger animals or humans. In an attempt to ensure that a sufficient concentration of GDF5 is maintained, multiple injections of the growth factor have been administered; however, the injury caused by the multiple punctures counteracted the beneficial effects of GDF5 and even worsened the disc degeneration. 
As an alternative therapeutic strategy for GDF5 delivery, gene therapy has the potential to provide a long-term in vivo expression of functional protein. In this study, we evaluated the effect of Ad-GDF5 gene therapy in a novel mice disc degeneration model.
The replication-incompetent adenovirus containing the firefly luciferase reporter gene (Ad-Luc) was a kind gift from Dr. Greg Helm (University of Virginia), and was at a concentration of 2×1011pfu/ml. We generated the replication-incompetent adenovirus that was engineered for expression of the GDF5 gene (Ad-GDF5) using AdEasy System (Stratagene, La Jolla, CA) according to the manufacturer’s instructions.  The final concentration of Ad-GDF5 was determined to be 1.8×1011pfu/ml.
All of the animal work procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Virginia. A total of 56 Balb/c mice (8 weeks old) from Harlan Laboratories (Harlan Laboratories, Indianapolis, IN) were studied, which included 16 mice for the study of the disc degeneration model and 40 mice for the evaluation of gene therapy. Four mice in model study and 10 mice in gene therapy study were sacrificed at each of the time points (1, 2, 4 and 8 weeks after operation).
Prior to the procedure, the animals were put under general anesthesia. Using aseptic techniques, the spine was exposed through an anterior midline transperitoneal approach. After separating the hind peritoneum and psoas major muscles, the L4, L5 and L6 vertebral bodies were identified. For the mice model study, the L4/5, L5/6 and L6/S1 discs were punctured by a 27-, 30-, and 33-gauge needle (Hamilton, Reno, NV), respectively. For gene therapy study, the L5/6 and L6/S1 discs were randomly injected with 5µl of Ad-GDF5 (1.8×1011pfu/ml) or Ad-Luc (2×1011pfu/ml) through a 30-gauge needle. After the operation, the incision was closed in a layered fashion and the animals were allowed to move freely around their cages.
Four of the animals received Ad-Luc injections were subjected to non-invasive, real-time in vivo imaging of bioluminescence, which was assayed on the first and third day after the injection, and then once a week through until the 8th week following the operation. The bioluminescence was imaged using a cryogenically cooled In Vivo Imaging System (IVIS) (Xenogen Corp., Alameda, CA). Using IGOR analysis software (Xenogen Corp.), defined round region of interest (ROI) of 1.5-mm diameter that included the peak photon count was selected at the same position in each image of the same animal. The images obtained from each animal on the first day were used to normalize the images taken at each subsequent time-point.
Four mice in gene therapy study received the radiographic study at each time point. Animals were put under general anesthesia using a 2% isofluorane/oxygen mixture. Data were then acquired using a small animal Digital X-ray scanner. Lateral radiographs were obtained prior to the operation, and then at 1, 2, 4, and 8 weeks after the operation. The heights of vertebral bodies and discs were measured with imaging software (Image J, NIH) and were expressed as Disc Height Index (DHI) using the method developed by Lu et al.  The average measurement of anterior, middle and posterior height of the disc was divided by the average height of the adjacent vertebral body. The change in the DHI was expressed as the percentage DHI (%DHI) and normalized to the DHI that was obtained from the pre-operation radiograph.
Four mice in model study and two mice in gene therapy study were subjected to an MRI scan at each of the different time points through until 8 weeks after operation. Imaging was acquired using a Varian Inova 4.7 Telsa (T) MR (Varian Inc., Palo Alto, CA) scanner. T2-weighted sections in the sagittal plane were obtained as a series of multiple two-dimensional slices.
At each time point, five mice in gene therapy study were used for the biochemical assay. Animals were sacrificed and the discs were harvested immediately. After removing the endplate, the discs were digested in 200 µL of sterile papain solution (125 g/mL in 1× PBE, pH 6.5) at 60°C for 24 hours. DNA content was determined by Hoechst 33258 dye assay with calf thymus DNA as a standard. The glycosaminoglycan (GAG) content of the disc was detected by the dimethylmethylene blue (DMMB) colorimetric assay, with chondroitin sulfate as a standard.
Four mice in model study and five mice in gene therapy study were sacrificed at each time point. Lumbar spinal column specimens were harvested and fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) solution for 48 hours. The specimens were then decalcified with 0.25M ethylenediamine tetraacetic acid (EDTA) in PBS for 2 weeks. After being embedded in paraffin, the specimens were sectioned mid-sagittally at 8µm and stained with hematoxylin and eosin staining, or Safranin-O and fast green staining.
An ANOVA test was performed to determine whether there was a significant difference in disc height among the three experimental groups at different time points. We also analyzed variance of the GAG and DNA content within each group at different time points using ANOVA test. Data are presented as means ± S.D. and a p value of less than 0.05 was considered significant.
MRI scans of injured mice lumbar spines are represented by T2-weighted, midsagittal plane images obtained 1 and 2 weeks after needle puncture (Fig. 1). Signal in the nucleus pulposus area disappeared in the 30- and 27-gauge injured discs at 1 week after injury, while the signal intensity of 33-gauge needle injured discs did not show appreciable differences at 1 week post injury. However, all injured discs, including discs injured by 33-gauge needle, lost T2 weighted signals from 2 weeks after injury.
Sections of non-injured and injured discs were stained with Safranin-O at 1, 2, 4 and 8 weeks (Fig. 2). The non-injured discs had distinct nucleus and annulus regions. The nucleus demonstrated a gelatin like structure with scattered cells, while annulus was well organized with its lamellar sheets of collagen. In the 27- and 30-gauge needles injured disc, the nucleus disappearance accompanied with the loss of red staining and the lamellar structure in the annulus since the first week post operation. The reduction of cells and the accumulation of fiber-like tissue aggravated as the degeneration progressing in both groups. In 8 weeks after injury, the 27 gauge needle injured disc almost lost all the red staining and lamellar structure, which was worse than 30-gauge needle injured disc. However, in the 33-gauge needle injured discs, the histological morphology didn’t show obvious changes compared with non-injured disc at the first week. But since 2 weeks post injury, the inner layer of annulus lost the concentric lamellar structure with cracks and fissures appearance. Although the general loss of definition between the annulus and nucleus presented as the evidence of IDD, the nucleus tissue of 33-gauge punctured discs still can be identified by 2 weeks post injury. As the degeneration progressed, the nucleus tissue eventually disappeared at 8 weeks post operation and the space was occupied by disorganized, fibrocartilaginous tissue.
To determine whether mice had been successfully infected with adenovirus competent to express exogenous proteins, we tracked the expression of luciferase in situ in animals transduced with Ad-Luc. Fluorescence was observed on the first day after injection of the Ad-Luc (Fig. 3). In most of the animals it was no longer visible by the 6th week post-operation, but in some mice the expression was sustained through until the mice were sacrificed at 8 weeks. The majority of luciferase activity was confined to the injection site instead of distributed throughout the animal. To determine the expression profile, data were normalized to the fluorescence levels measured on the first day. Quantitative analysis of the peak amplitudes of fluorescence from four mice demonstrated that the activity levels of expressed luciferase steadily declined after the injection of Ad-Luc. At the end of the first week, the fluorescence intensity had decreased to nearly half the peak levels, but after a further 2 weeks the expression was still a quarter of the peak levels, and was still greater than 10% after 28 days.
The effects of Ad-GDF5 therapy were determined by assessing changes to the %DHI compared with the pre-operation values (Fig. 4). There were no differences in the %DHI for the control discs, which had been left intact, at any of the time points. However, one week after injection, the %DHIs of discs that had been injected with either Ad-Luc or Ad-GDF5 were remarkably decreased (P<0.05). For the discs that had been injected with Ad-Luc, the %DHI continued to decrease with time. Although %DHI for the discs injected with Ad-GDF5 had initially decreased to a similar extent to those injected with Ad-Luc, the %DHI of these discs recovered by 2 weeks post-operation, and this recovery was maintained for the remaining duration of the experiment, until sacrifice at 8 weeks. Indeed, from the 2-week time point, there was no statistic difference between the %DHIs of the intact and the Ad-GDF5 injected disc, with the difference between them and Ad-Luc injected discs growing more pronounced at each time point (2 and 4 weeks P<0.05; 8 weeks P<0.01).
MRI scanning was performed at 2, 4, 6 and 8 weeks following the operation to determine the tissue water content in the discs. In all mice, the intensity of the T2-weighted signal of both the Ad-Luc and Ad-GDF5 injected groups was clearly decreased at both the 2- and 4-week time points (Fig.5). However, the signal from the discs that had been injected with Ad-GDF5 had begun to recover by the 6-week time point. The trend continued to the 8-week time point, when an even higher signal was measured. On the other hand, there were no visible signs of recovery of the signal intensity in the discs that had been injected with Ad-Luc at any of the time points prior to sacrifice at 8 weeks.
The effect of Ad-GDF5 and Ad-Luc on total DNA, a marker of changes to the overall number of cells in the disc, was determined (Fig. 6). The average DNA content of the disc that had been injected with Ad-Luc had decreased by the 4-week time point, and this was sustained through until the 8-week time point, the differences achieved statistical difference compared with the intact and Ad-GDF5 injection discs (P<0.05). However, the DNA content of the disc that had been injected with Ad-GDF5 did not vary from that of the intact disc at any of the time points.
The GAG content was normalized by the DNA content of the disc and compare among the groups as GAG/DNA ratios. Compared with the intact disc, the GAG/DNA ratio of Ad-Luc injected disc level decreased at the 2- and 8- week post operation (P<0.05) (Fig. 6). In contrast, the GAG/DNA ratio in Ad-GDF5 group increased at the 4-week compared with other groups and had no statistic difference compared with the intact disc group at 2- and 8-week time points.
The disc degeneration was studied histologically by staining with hematoxylin and eosin (Fig. 7). We found that discs injected with Ad-Luc and Ad-GDF5 had lost the majority of the nucleus pulposus by the first week following the operation. In addition, by this time point the chondroid cells had begun to proliferate from the annulus towards the center of the disc, an effect that was more pronounced in the Ad-GDF5 injected discs than the Ad-Luc injected discs. The preponderance of cells growing towards the center of the disc after injection with Ad-GDF5 continued throughout the remaining time points. At the 8-week time point, the disc height of the Ad-Luc groups was obviously lower than Ad-GDF5 groups, which confirmed the result of %DHI measurement of radiographic images. Although the Ad-GDF5 discs did not regenerate the typical appearance of the nucleus pulposus, they were nonetheless much more cellular compared with the Ad-Luc injected discs, which were instead filled with fibrous tissue.
In the present research, we investigated the efficacy of Ad-GDF5 gene therapy in an in vivo mouse model of IDD. Several different animal species have been used for in vivo studies of the biology and mechanical characteristics of IDD, including sheep , pig , dog , rabbit [11,26–28], rat [29,30] and mouse. [12,31] Dawn M. et al. made a comparison of the disc geometry parameters of baboons, sheep, rabbits, rat tail, rat lumbar, mouse tail and mouse lumbar with humans, the mouse lumbar discs were found to be the closest representation of the human lumbar discs geometry. This result revealed the advantage of using mouse lumbar discs to mimic the situation of human lumbar IDDs, which related to the changes of anatomy parameters. We thus developed a lumbar needle puncture model of IDD in mice. Similar to the pathological changes in human lumbar IDD [33,34], the collapse of nucleus pulposus tissue followed by the damages of annulus and endplates appeared in this IDD model. The matrix breakdown resulted in the reduction of water content and the decrease of the space in the disc.
Sustained expression of exogenous proteins for the treatment of chronic diseases, such as IDD, is essential to the success of the therapeutic strategy. In a study of transgene expression mediated by an adenoviral vector, the target gene expression in rabbit discs can be found even more than one year post injection. Nishida and Kang et al. reported adenoviral vector carrying TGF-²1 gene can increase proteoglycan synthesis after injected in rabbit discs for one week.  In this study, the result of Ad-Luc injection demonstrated that the expression of a target protein after infection of living cells in the disc can be maintained at an assayable level for at least 5 weeks. This results in an extension of the functional exposure of host cells to exogenous proteins, resulting in a long-term effect after a single injection into the disc.
GDF5 is important for the development of skeletal system, especially for the cartilage-like tissue. In our previous study, GDF5 deficient mice showed conspicuous depauperate and degenerated discs.  Therefore, it is attractive for using GDF5 in the therapeutic strategy of IDD. We also found that the GDF5 protein can enhance cells proliferation and increase the gene expression for extracellular matrix proteins such as type II collagen and aggrecan in in vitro cultured disc cells.  Moreover, GDF5 protein suppressed the expression of matrix metalloproteinase-3 (MMP-3), part of the mechanism underlying the pathological degradation of the extracellular matrix in disc. [37,38] These observations have been confirmed by other research groups.[12,15,20]
In this present study, the cells proliferation of disc influenced by the injection of Ad-GDF5 or Ad-Luc was revealed by DNA content assay and histological study. There was no statistical difference of DNA content at all of the time points between Ad-GDF5 injected and intact discs. We considered that the injection of Ad-GDF5 enhanced the cells proliferation, which compensated the loss of the cell number after the needle puncture induced IDD. Besides the cells proliferation, another effect of GDF5 promotion is the increase of proteoglycan in the extracelluar matrix of discs. There was no reduction of GAG/DNA ratio at any of the time points in the discs that had been injected with Ad-GDF5 compared with intact discs. Moreover, at 4-week time point, the ratio of GAG/DNA in Ad-GDF5 treatment group even higher than that in the intact disc. This indicates that Ad-GDF5 injection not only improved the cells proliferation of the chondriod cells but also promoted proteoglycan production. A previous report described increases in the proteoglycan content of the disc after injection of osteoprotein-1 (OP-1), which were present at the 2-week time point, but not at later time points.  It is possible that the stimulating effect of the injected growth factor on the secretion of matrix components was lost after 2 week post injection. However, we found that the expression of GAG content was maintained at the same level throughout the experiment in the Ad-GDF5 group. We conjecture that sustained expression of GDF5 would maintain the continuous secretion of matrix components for long periods of time. The promotion of proteoglycan production after the injection of Ad-GDF5 was also confirmed by MRI studies. [40,41] In Ad-GDF5 treatment group, the T2 signals begin to reappear at the 6-week time point and were more obvious at 8 weeks. We believe this may reflect the appearance of more chondroid cells in the discs injected with Ad-GDF5, which correlated with greater secretion of proteoglycan.
The changes to the number of cells, the expression of matrix components and the water content can all be interpreted in terms of the degeneration and recovery of the disc. These transformations also reflect changes of the height of discs, while the narrowing of disc is considered to be one of the important characteristic features of IDD.[42,43] The height of the discs injected with Ad-GDF5 had recovered and the effect was maintained throughout the rest of the experiment.
Our research reports the curative effects of gene therapy for the treatment of IDD in a novel mice model by adenoviral mediation of human GDF5 gene. The results of this study not only confirmed the long-term expression of the target protein in the disc in vivo, but also highlighted the physiological improvements that occurred to the disc. Our research also confirmed this IDD model developed in mice will be suitable for the study of the physiopathologic changes and the effect of therapeutic intervention.
In support of their research for or preparation of this manuscript, this project was supported by AO Research Fund of the AO Foundation and the National Institute of Health. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.