Transplantation of mesenchymal stem cells (MSCs) inhibits the progression of disc degeneration in animal models. We know of no study to determine the optimal number of cells to transplant into the degenerated intervertebral disc (IVD). To determine the optimal donor cell number for maximum benefit, we conducted an in vivo study using a canine disc degeneration model. Autologous MSCs were transplanted into degenerative discs at 105, 106, or 107 cells per disc. The MSC-transplanted discs were evaluated for 12 weeks using plain radiography, magnetic resonance imaging, and gross and microscopic evaluation. Preservation of the disc height, annular structure was seen in MSC-transplantation groups compared to the operated control group with no MSC transplantation. Result of the number of remaining transplanted MSCs, the survival rate of NP cells, and apoptosis of NP cells in transplanted discs showed both structural microenvironment and abundant extracellular matrix maintained in 106 MSCs transplanted disc, while less viable cells were detected in 105 MSCs transplanted and more apoptotic cells in 107 MSCs transplanted discs. The results of this study demonstrate that the number of cells transplanted affects the regenerative capability of MSC transplants in experimentally induced degenerating canine discs. It is suggested that maintenance of extracellular matrix by its production from transplanted cells and/or resident cells is important for checking the progression of structural disruption that leads to disc degeneration. Published by Wiley Periodicals, Inc. J Orthop Res 28:1267–1275, 2010
Degenerative spinal disease not only causes lumbar pain in individuals, but also has a major socioeconomic impact from stress on the medical system and loss of productivity from occupational disease.1, 2 While intervertebral disc (IVD) degeneration has been implicated as one of the major causes of degenerative spinal disease, other factors thought to be involved including age, trauma, genetic predisposition, and lifestyle factors, such as obesity, smoking, occupation, and stress.3–5
Because IVD degeneration is an irreversible progressive disease, various treatments are currently being developed to repair and regenerate damaged IVDs. Experimental regenerative medicine techniques for IVD degeneration include the intradiscal injection of cytokines and growth factors,6–10 gene delivery to IVD cells,11–13 creation of artificial IVDs using tissue engineering,14 and cell transplantation.15–18 In vitro and in vivo studies using these techniques have reported increased IVD cell activity, the stimulation of extracellular matrix synthesis, and the suppression of IVD degeneration.
Recent advances in molecular biology and regenerative medicine have drawn an attention to mesenchymal stem cells (MSCs), because these immature cells remain pluripotent, capable of differentiating into bone, cartilage, or fat cells depending on their treatment. Undifferentiated MSCs are also immune tolerant. Autologous MSC transplantation has become an important treatment for various diseases.19, 20 Facilitating clinical applications, MSCs are easily harvested, and extracorporeal isolation and culture are relatively simple, requiring no highly specialized procedures. A number of basic studies have been performed using MSC transplantation into IVDs. Sakai et al.15 performed autologous MSC transplantation using a rabbit IVD degeneration model, showing by radiology, histology, and biochemistry that this procedure suppressed IVD degeneration. They also analyzed the differentiation of transplanted MSCs into nucleus pulposus cells, establishing that transplanted MSCs expressed markers for the resident nucleus pulposus cells, indicating that the differentiation of MSCs is dependent on their environment.16, 17 Hiyama et al.18 examined the immune privilege of IVD tissue, particularly the nucleus pulposus, including the anatomical characteristics separating disc tissue from the host immune system. Using a canine IVD degeneration model, these authors transplanted MSCs and measured changes in the expression of the Fas/Fas-ligand (FasL) system in addition to ascertaining any suppressive effects of the transplant on disc degeneration. In their study, the transplantation of MSCs suppressed IVD degeneration as assessed by radiological, histological, and biochemical findings. Their results also suggest that MSC transplantation could be effective in maintaining the immune privilege of IVDs.
Based on these studies and others, MSC transplantation into degenerated IVDs offers a promising new interventional technique in animal models.21, 22 Although the efficacy of MSC transplantation has been documented, some questions remain unanswered. For example, the optimal cell number for achieving IVD regeneration from MSC transplantation have not been elucidated. Using a canine model of disc degeneration, the present study examined the dosage of autologous MSCs to transplant, investigated the inhibitory effects of the MSC transplant on disc degeneration, and determined the viability of transplanted MSCs.
MATERIALS AND METHODS
Animal experiments were carried out with IACUC approval. A total of 30 beagle dogs, 12- to 18-month old, weighing approximately 10 kg, (Nosan Beagle; Nosan Corporation, Kanagawa, Japan) were divided equally into five groups of six. The N group was the unoperated control group. Three groups with induced disc degeneration were given subsequent MSC transplants of 105, 106, or 107 cells per disc. The D group of operated controls had induction of disc degeneration without MSC transplants. Lateral radiographs and magnetic resonance imaging (MRI) were obtained for every animal before experimental use to confirm the absence of abnormalities. At 12 weeks after the first operation, final radiological and MRI assessments were obtained and all dogs were euthanized by a lethal dose of 120 mg/kg sodium pentobarbital (Abbott Laboratories, Abbott park, IL). The L3/4, L4/5, and L5/6 discs were isolated with the vertebral bodies attached.
Disc Degeneration Model
Four weeks before the initial MSC transplantation, disc degeneration was induced in L3/4, L4/5, and L5/6 IVDs by NP aspiration according to the method of Hiyama et al.18 Briefly, an 18-gauge needle was inserted at the center of the disc through the AF into the NP. To induce disc degeneration, the NP was aspirated using a 10-ml syringe, as previously described. The mean mass of the nucleus pulposus aspirated from each disc was 13.4 ± 4.7 mg.
Bone Marrow Collection, Analyses, and Transplantation of MSCs
Autologous MSCs were obtained from the iliac crest from each animal receiving autologous transplants according to the method of Hiyama et al.18 To measure survival rate of MSCs transplanted into the NP cavity, the MSCs were infected with AcGFP1 (Takara, Japan), a retrovirus vector expressing the green fluorescent protein (GFP) gene (conditioning studies were performed, data not shown). Vector incorporation was more than 90% (data not shown) using FACS analysis. Following the second passage in culture, adherent autologous MSCs were enzymatically released from monolayer culture and the designated cell number was transplanted percutaneously into IVDs in which degeneration had been induced (L3/4, L4/5, and L5/6) 4 weeks before. Transplants were done in the three lumbar discs in a randomized sequence using a discogram needle (25-gauge) guided by fluoroscopic imaging.
Lateral radiographs were taken under general inhalation anesthesia in all groups at 0, 4, 8, and 12 weeks after induction of degeneration. A fluoroscopic imaging intensifier (70 kV, 10 mA, distance 100 cm) was used. Vertebral body heights and disc heights were measured using image J software (Free soft, Image Processing and Analysis, http://rsb.info.nih.gov/ijl). The data were then transferred to the Excel software program (Microsoft Excel, 2003) and the disc height index (DHI) was calculated, using the method of Masuda et al.25 Changes in the DHI were expressed as %DHI and normalized to the measured preoperative IVD height by the following equation [DHI = (post-operative DHI/preoperative DHI) × 100].
MRI images were also taken to evaluate signal changes in T2-weighted images at 0, 4, 8, and 12 weeks after the first operation in all groups. All MRIs were obtained using a spine coil (1.5T, Gyroscan, ACS-NT, Powertrak6000, Philips, Amsterdam, the Netherlands) under anesthesia. T2-weighted sections in the sagittal plane were obtained using a fast spin echo sequence with time-to-repetition (TR) of 4,000 ms and time-to-echo (TE) of 150 ms; interslice gap of 0.3 mm; matrix at 512 × 512; the field-of-view (FOV) was 200 mm × 200 mm; the number of excitations was 4 and TSE echo spacing was at 18.8. At 12 weeks after the induction of degeneration, the signal intensity of the T2-weighted image of each disc was evaluated using the Pfirmann classification.26, 27
Gross Anatomical Findings
The spinal segments from five dogs from each group (n = 25 IVDs) were fixed in 10% formalin neutral buffer solution (Wako, Osaka, Japan) and decalcified in Plank–Rychlo solution (Decalcifying Solution A; Wako, Tokyo, Japan). Specimens were cut longitudinally through the center of the disc for macroscopic evaluation.
Discs L3/4, L4/ 5, and L5/6 were excised from the lumbar spine of 10 dogs from each group (n = 50 IVDs). Each IVD was fixed in 10% formalin, decalcified, and embedded in paraffin. The paraffin blocks were stained with hematoxylin and eosin or Safranin-O, and evaluated by histology and immunostaining. The hematoxylin and eosin stain evaluated degenerative changes in annulus fibrosus using the disc degeneration grading system of Nishimura and Mochida.28 Two observers familiar with human and animal IVD specimens and blinded to this study evaluated the sections. The intraobserver error was very small. The kappa value for grading scale was 0.90, showing an excellent agreement. For immunohistochemistry, primary mouse monoclonal antibody (collagen type II, Daiichi Fine Chemical Co., Toyama, Japan) was used with standard protocol.
Apoptotic Cell Counts
The DeadEnd™ Colorimetric TUNEL System (Promega, Madison, WI) was used according to the manufacturer's instructions to detect apoptotic cells in the NP. End labeling was performed on formalin-fixed sections from lumbar spine of five dogs from MSC-transplantation groups. Positive control sections were pretreated with DNase I (Promega). The terminal deoxynucleotidyl transferase (TdT) was replaced by PBS for the negative controls. Two pathologists counted the cells in the nucleus pulposus area in 10 random high power fields (HPFs; magnification ×400). The percent of TUNEL positive cells of the total nucleus pulposus cells was calculated.
Viability of Nucleus Pulposus Cells After Transplantation
The viability of cells recovered from nucleus pulposus samples of five dogs from each group (n = 25 IVDs) were analyzed at 12 weeks after induction of IVD degeneration. Specimens were weighed, then digested in Dulbecco's modified Eagle medium (DMEM; Gibco, Invitrogen Corporation, Carlsbad, CA) containing 0.4% (w/v) actinase E (Kaken Pharmaceutical Inc., Tokyo, Japan) for 1 h, followed by 2 h in DMEM containing 0.016% (w/v) bacterial collagenase P (Roche Diagnostics GmbH, Mannheim, Germany). The digested tissue was passed through a 70 µm pore size cell filter (Becton Dickinson Labware Co. Ltd, Franklin Lakes, NJ). Cells were washed, filtered (Falcon cell strainer, 100 µm), seeded in 96-well microtiter plates (100 µl, 20,000 cells/well, four-wells/disc), and incubated for 20 min to allow settling. Then, 50 µl of ethidium homodimer-1 (EthD-1, 2 µmol/L)/calcein AM (1.6 µmol/L) solution (Live/Dead® cell viability kit; Invitrogen, Basel, Switzerland) were added to each well and incubated for 20 min. Methanol-treated cells (100 µl/150 µl cell suspension) were used as negative controls. Cell fluorescence was observed using an inverse microscope (Leica, DM IL; filters: I3 S 450–490 nm and N2.1 S 515–560 nm). High-resolution digital photos of three visual fields/well were taken using both filters at 25×. Image analysis software (Quantity One®; Bio-Rad, Hercules, CA) was used to quantify cell numbers.
Evaluation of the Survival of Transplanted MSCs
Formalin-fixed sections from lumbar spine of five dogs from the GFP-positive MSC-transplantation group (n = 15 IVDs) were analyzed at 12 weeks after induction of IVD degeneration. At 12 weeks post-degeneration induction were embedded in paraffin blocks and sectioned at about 2 µm. Standard procedures were applied for fluorescent immunohistochemistry with primary antibody (Anti-Green Fluorescent Protein; 1:50; Medical&Biological Laboratories, Nagoya, Japan). Alexa 488 (1:100; Molecular Probes, Eugene, OR) was used as a second antibody. Tissues were mounted on slides using VECTASHIELD Mounting medium with 4′,6′-diamino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA). Two pathologists counted the cells in the nucleus pulposus area in 10 random HPFs (magnification ×400). The percent of GFP-positive MSCs of the total nucleus pulposus cells was calculated.
The significance of differences among means of data on radiograph measurements, TUNEL staining for apoptotic cell counts, Live/Dead® for viability of nucleus pulposus cells after transplantation, and the survival rate of the GFP-positive transplanted MSCs was performed using a repeated measure ANOVA and Fisher's PLSD post hoc test. Statistical significance was accepted at p < 0.05. The Kruskal–Wallis test and Mann–Whitney U-test were used to analyze the nonparametric data from MRI and histology grading. The Statview program was used for the statistical analyses. Error bars were set to represent 1 SD unit.
Radiographic and MRI Findings
Radiographs showed significant (p < 0.01) narrowing of the disc space at 4 weeks after induction of IVD degeneration in the operated control D group and the MSCs transplant groups. The mean %DHI in the D group continued to decrease until 8 weeks post-induction, then plateaued. The mean %DHI of the N group was nearly 100% throughout the study. In contrast, at 4, 8, and 12 weeks after MSC transplantation the mean %DHI of the MSC-transplanted disc groups was higher than the pre-transplant index. At 4, 8, and 12 weeks after the first operation, the mean %DHI increased significantly (p < 0.01) in the transplant groups compared to the D group. No significant differences were seen among the three MSC-transplant groups (Fig. 1). At 12 weeks post-induction, the results of MRI (T2-weighted) signal intensity measurement of discs from the D group was less than all other groups. The MSC-transplant groups showed a significant increase in the disc signal intensity. The use of Pfirrmann's classification revealed a significant delay in the progression of disc degeneration in the transplant groups, suggesting an increase in water content in MSC-transplanted discs.27 The signal intensity level of the N group was constant throughout the study period. Again, no significant differences were detected among the three MSC-transplant groups (Fig. 2).
The gross appearance of the dissected spines at 12 weeks post-induction of degeneration showed more apparent disc space narrowing and connective tissue invasion of the nucleus cavity in the D group than the other groups. The 106 and 107 MSC-transplant groups appeared similar to the N group, but the discs from the 105 cell transplant group appeared more similar to the connective tissue invasion seen in the D group. Overall, the disc structure was maintained better in the 106 and 107 cell transplant discs (Fig. 3).
The histological analysis also showed noteworthy regenerative effects of the MSC transplants. Hematoxylin and eosin staining of N group discs showed a relatively well preserved oval-shaped nucleus with no collapse of inner and outer annular structures. The 106 and 107 cell groups also showed a relatively well preserved inner annulus structure similar to the N group, but discs from the 105 cell group had less preservation of the annular structure (Fig. 4a,c). Safranin-O staining of NPs from the 106 and 107 groups showed relatively dark staining of nucleus tissue, similar to the N group, while discs from the 105 and D groups showed lighter staining of the nucleus and inner annulus (Fig. 4b). The D group stained poorly for collagen type II compared with the other groups (Fig. 5).
Detection of Apoptosis in MSC-Transplanted Discs
The results of TUNEL staining revealed that N group discs had low numbers of apoptotic cells, while D group discs showed significantly higher numbers (p < 0.01) of dying and dead cells compared to the other groups. Among the three MSC-transplant groups, the ratio of apoptotic cells was significantly lower for the 106 group than the 105 and 107 groups. No significant difference existed between 105 and 107 groups (p = 0.25; Fig. 6a). There was not a specific region for TUNEL positive cells. It stained scattered in the NP cavity (Fig. 6b).
Assessment of Viability of Nucleus Pulposus Cells after MSC Transplantation
The dual-fluorescence Live/Dead® cell viability assay was used to label viable nucleus pulposus cells 16 weeks after induction of degeneration in the D and MSC-transplant groups. Live cells were stained green with Calcein-AM and dead cells were stained red with Ethidium homodimer-1. The N group had more live cells, while the D group showed significantly (p < 0.01) less than the other groups. Among the three MSC-transplant groups, the number of live cells in the 106 and 107 groups was significantly (p < 0.01) greater than the 105 group. The 107 group had smaller number of live cells than the 106 group, although the difference was not statistically significant (p = 0.20; Fig. 7).
Survival of Transplanted MSCs in the Nucleus Pulposus
GFP-positive MSCs were detected using an FITC filter. GFP-positive MSCs were seen primarily in the central region of the nucleus pulposus. Only cells that stained with DAPI to show a clearly visible nucleus were counted. These cells were consecutively counted in 10 randomly selected visual fields of nucleus pulposus area and averaged after costaining with GFP to detect the labeled transplanted MSCs. Percentages of GFP-positive cells for the 106 cell group was 62.8 ± 12.4, and the 107 group 75.2 ± 16.1%, both were significantly (p < 0.01) higher than the 15.3 ± 9.2% found in the 105 group (Fig. 8a). There was no specific region for its distribution and they were scattered in NP cavity (Fig. 8b).
In order to evaluate the effect of cell doses in MSCs transplantation for experimental disc degeneration, changes in disc height, as assessed by radiographic %DHI, signal, and in relative signal intensity as classified by Pfirrman et al. of the nucleus pulposus on MRI T2-weighted imaging were determined. We found that the increase of %DHI in the MSC-transplant groups compared to the D group decreased beginning at 4 weeks after transplantation. On T2-weighted imaging, MSCs transplanted discs showed less degeneration in Pfirrman classification. Diagnostic imaging results showed that the progression of IVD degeneration was inhibited in all three MSC-transplanted groups, with no significant differences among the three groups. Evaluation of the histology of the experimental discs indicated that disc degeneration was limited in the MSC-transplant groups. The annulus fibrosus structures were maintained at near normal status in the MSC 106 and 107 groups, but not the 105 group. In the NP of the MSC 106 and 107 groups, proteoglycan, type II collagen, and other extracellular matrix components were maintained, indicating that IVD degeneration was suppressed. There were no significant differences in imaging results between 105 versus 106 and 107 groups in MSC-transplanted group discs despite the fact that there were significant differences in histological analysis. This may be resulting from the fact that disc height and T2-weighted signal on MRI would not correlate directly to histological disc degeneration in short observation time. In the previous literature, Ho et al.22 suggested that differences between imaging and histological findings are probably due to differences in sensitivity between diagnostic imaging and histological analysis; however, differences in follow-up time intervals may also contribute. Further studies utilizing longer follow-up times are needed to resolve the differences between imaging and histological findings.
The three parameters used to assess post-transplant cell activity in discs were the number of remaining transplanted MSCs, the survival rate of NP cells, and apoptosis of NP cells in transplanted discs. For the MSC 105 group, the numbers of transplanted MSCs remaining and viable cells in the nucleus pulposus tissue were low. An insufficient cell number may not be capable of reconstructing the matrix and microenvironment of degenerating discs, resulting in increased death of both resident and transplanted cells. As seen in the MSC 107 group, when too many cells were transplanted into the limited space of the IVD, an imbalance between the relatively slow rate of diffusion of nutrients and cell number may have occurred, inducing apoptosis of both transplanted MSCs and resident cells; this needs to be verified in longer-term studies. These findings from the transplantation of 105, 106, and 107 MSCs suggest that quantity of the transplanted cells are important to optimize the inhibition of progression of degeneration and to regenerate damaged IVDs and that there most likely is a minimal number that affects its therapeutic effect. However, there is still much more parameters to discuss such as evaluation of adequate nutrition for the number of MSCs transplanted which may also be important for maintaining the microenvironment of the disc or differentiation status of the transplanted MSCs. Inducing MSCs towards NP in vitro may most likely affect the therapeutic outcome as well.
Animal models for studies of cell transplantation in IVD degeneration have employed rabbits, rats, and larger animals, such as beagles.23, 24, 29 Several techniques have been developed to produce disc degeneration, but an ideal animal model for the human condition has not been established.30–32 It has been well known that notochord-like cells are present in the nucleus pulposus of rats and young rabbits, whereas only chondrocyte-like cells constitute the nucleus pulposus in humans. Small animals may therefore be unsuitable for use as models of the human disease. We use beagles as an alternative animal model because we believe the discs of this chondrodystrophic breed more closely approximate the anatomy of human discs. In addition, beagle IVD degeneration and herniation are likely to occur spontaneously, and beagle nucleus pulposus cells are chondrocyte-like. Beagles are sometimes being used to study IVD degeneration since their disc cells share similarities with human discs.33–37. Degeneration is produced by nucleotomy in the model used in the present study, and thus may have differences from naturally occurring IVD degeneration in humans. Although this may be considered a limitation of this model, no animal model that completely mimics human IVD degeneration exists at this time. The nucleotomy model may well have usefulness for studying disc degeneration progression following disc surgery.
In the human NP, the average total cell density of the disc is about 4–6 × 103 cells/mm3.38, 39 Human discs are rich in extracellular matrix and possess immune privilege. For clinical application, the optimal number of MSCs transplanted to restore these characteristics in damaged IVDs must be determined. In a supplementary experiment, we investigated 20 IVD specimens from 10 beagles, 10 of which were normal IVDs taken from beagles sacrificed at a very similar age and from the same disc level of animals used in the study and 10 of which same procedure was performed as the initial study for induction of disc degeneration. Results demonstrated that average volume of NP tissue in single IVDs of beagles used in the study was 107.18 ± 1.03 mg (wet weight) and the average viable cell number was 1.03 ± 0.19 × 106 cells. For the IVDs, where induction of disc degeneration was performed, average volume of NP tissue at the time of injection was 87.38 ± 4.23 mg (wet weight) and the average viable cell number was 1.36 ± 0.15 × 105 cells. These data indicates that there is not truly a large difference in number of viable cells among the IVDs used in the study. This also justifies the rational of injecting equal numbers of MSCs in the discs. Furthermore, a base line of 106 viable cells may be the adequate number to maintain disc homeostasis, which also is understandable in interpreting the result of initial study where injection of 105 cells was not fully sufficient. Previous animal model studies of cell transplantation to IVDs have used 106 cells/disc for transplants,15–18 therefore, we transplanted 105, 106, or 107 cells per disc in the present study, establishing three transplant groups with the number of cells transplanted increasing at 10- to 100-fold. Although actual clinical situation for translation of this procedure will not be clarified without clinical trials, we believe that grade 2 or 3 in Pfirrman classification, where disc degeneration is mild-to-moderate with less nutritional problem may be one candidate.40
In the present study, the effect of the number of MSCs transplanted into a degenerative disc was investigated using a canine disc degeneration model. From the perspective of clinical application, the optimal number of transplanted cells may also be affected by the severity of recipient disc degeneration.22 However, in the current limited model, the transplantation of 106 MSCs, when compared to 105 or 107, produced the best maintenance of the structure of IVDs and best inhibited IVD degeneration. Because results from a study such as ours may vary in other animal models and in humans, care must be taken in the clinical application of our findings.
The authors would like to thank Tomoko Nakai and Sachie Tanaka and Teaching and Research Support Center at Tokai University School of Medicine for their technical assistance. This work was supported in part by a Grant-in-Aid for Scientific Research and a Grant of The Science Frontier Program from the Ministry of Education, Culture, Sports, Science and Technology of Japan D.S. and J.M., a grant from 2008 Tokai University School of Medicine Research Aid and a grant from AO Spine International to D.S.