Hyaline Cartilage Regeneration Using Mixed Human Chondrocytes and Transforming Growth Factor- 1 - Producing Chondrocytes ABSTRACT

Transcription

1 TISSUE ENGINEERING Volume 11, Number 9/10, 2005 Mary Ann Liebert, Inc. Hyaline Cartilage Regeneration Using Mixed Human Chondrocytes and Transforming Growth Factor- 1 - Producing Chondrocytes SUN U. SONG, Ph.D., 1 YOUNG-DEOG CHA, M.D., 1 JEOUNG-UK HAN, M.D., 1 IN-SUK OH, M.D., 2 KYOUNG BAEK CHOI, M.S., 3 YOUNGSUK YI, Ph.D., 3 JONG-PIL HYUN, B.S., 3 HYEON-YOUL LEE, B.S., 3 GUANG FAN CHI, Ph.D., 3 CHAE-LYUL LIM, M.S., 3 J. KELLY GANJEI, B.S., 4 MOON-JONG NOH, Ph.D., 4 SEONG-JIN KIM, Ph.D., 5 DUG KEUN LEE, Ph.D., 4 and KWAN HEE LEE, M.D. 1,3,4 ABSTRACT The purpose of this study was to investigate the efficacy of cartilage regeneration when using a mixture of transforming growth factor- 1 (TGF- 1 )-producing human chondrocytes (hchon-tgf- 1 ) and primary human chondrocytes (hchon) ( mixed cells ), compared with either hchon-tgf- 1 or hchon cells alone. Specifically, mixed cells or hchon cells were first injected intradermally into the backs of immune-deficient nude mice to test the feasibility of cartilage formation in vivo. Both the mixed cells and the hchon-tgf- 1 cells alone induced cartilage formation in nude mice, whereas hchon cells alone did not. To further test the efficacy of the cells in generating cartilage, an artificially induced partial thickness defect of the femoral condyle of a rabbit knee joint was loaded with hchon-tgf- 1 cells with or without mixing additional untransduced hchon cells, and hyaline cartilage regeneration was observed at 4 or 6 weeks. The efficiency of complete filling of the defect and the quality of tissue generated after implanting were evaluated on the basis of a histological grading system modified from O Driscoll et al. (J. Bone Joint Surg. 70A, 595, 1988). Significantly, mixed cells ( ) produced significantly better results than hchon-tgf- 1 ( ) or hchon ( ) cells alone. Histological and immunohistochemical staining of the newly repaired tissues produced after treatment with either mixed cells or hchon-tgf- 1 cells alone showed hyaline cartilage-like characteristics. These results suggest that the implantation of mixed cells may be a clinically efficient method of regenerating hyaline articular cartilage. INTRODUCTION THE FORMATION AND MAINTENANCE of hyaline articular cartilage depend largely on chondrocytes. Chondrocytes are derived from mesenchymal cells, which differentiate during skeletal morphogenesis and development to form chondrocytes. Thus, chondrocytes play an essential role in the upkeep of normal joint function by maintaining the homeostasis of cartilage extracellular matrix (ECM). They serve to increase the volume of the ECM and occupy less than 10% of the total cartilage tissue volume. Chondrocytes are metabolically active and are able to respond 1 Clinical Research Center, College of Medicine, Inha University, Inchon, South Korea. 2 Department of Orthopedic Surgery, College of Medicine, Inha University, Inchon, South Korea. 3 TissueGene Asia, Inchon, South Korea. 4 TissueGene, Gaithersburg, Maryland. 5 Laboratory of Cell Regulation and Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, Maryland. 1516

2 HYALINE CARTILAGE REGENERATION 1517 to a variety of environmental stimuli, such as growth factors, cytokines, matrix molecules, and mechanical loads. Hyaline articular cartilage is avascular in nature, having no nerves or blood supply; therefore, neural impulses cannot provide information to the tissue, and cellular and humoral immune responses are not likely to occur. 1 Hyaline articular cartilage has a limited regenerative capacity and therefore it often even inhibits the healing of traumatic injuries and other types of defects. 2 Several groups have developed strategies to repair or stimulate the regeneration of articular cartilage. These strategies include the transplantation of osteochondral grafts, periosteum, cultured autologous chondrocytes, mesenchymal stem cells, or mixtures of biodegradable polymers containing cytokine proteins Not surprisingly, each procedure has its advantages and disadvantages. We previously reported that the injection of mouse fibroblasts containing the transforming growth factor- 1 (TGF- 1 ) transgene into a partial hyaline articular cartilage defect produced hyaline-like cartilage within 6 weeks of transplantation in the rabbits. 13 We hypothesize that regeneration of hyaline cartilage could be accomplished with human TGF- 1 -producing chondrocytes (hchon-tgf- 1 cells) and even further by including extra, untransduced primary chondrocytes (hchon cells) at the defect site. These additional primary chondrocytes, stimulated by TGF- 1 produced from coinjected TGF- 1 -producing chondrocytes, are likely to generate more reparative proteins, such as the various ECM molecules, cytokines, and growth factors, which are required to induce the cartilage healing. In this study, cartilage tissue formation in nude mice and cartilage regeneration in a rabbit partial cartilage defect model were investigated after injection of TGF- 1 - producing human chondrocytes with or without additional primary human chondrocytes. Here we report that a mixture of hchon and hchon-tgf- 1 cells ( mixed cells ) or hchon-tgf- 1 cells alone are able to induce cartilage formation in nude mouse and rabbit cartilage defects, and that the mixture showed a higher histological grading score than did hchon-tgf- 1 cells alone. Histological and immunohistological staining results suggest that the newly formed tissue resulting from treatment with a combination of human TGF- 1 -transduced and untransduced chondrocytes had hyaline cartilage-like characteristics. MATERIALS AND METHODS Plasmid DNAs, recombinant TGF- 1 virus construction, and the generation of hchon-tgf- 1 cells The commercially available retroviral vector pdon- AI (TaKaRa Bio, Otsu, Shiga, Japan) was used to construct the pdon 1 vector, which carries the full-length 1.2-kb human TGF- 1 cdna under the control of the cytomegalovirus (CMV) promoter. The retroviral envelope viral protein-producing vector pvsv-g was obtained from BD Biosciences Clontech (Palo Alto, CA). pdon 1 and pvsv-g plasmid DNAs were cotransfected into GP2-293 cells by FuGENE transfection methods (Roche Diagnostics, Mannheim, Germany). After 48 h of culture, the supernatant was filtered through a m pore size polysulfone filter. The filtered culture supernatant was then diluted 2-fold with medium plus 10% fetal bovine serum (FBS) and used to infect human chondrocytes. Human chondrocytes were obtained from a 1- year-old polydactyl patient. Cells were passaged every 4 5 days. Passage 5 chondrocytes ( ) were seeded in 60-mm culture dishes 18 h before infection with the filtrate plus Polybrene (8 g/ml; Sigma, St. Louis, MO). After 4 h of incubation, the medium was aspirated, saved, and replaced with fresh medium. Infection was repeated 24 h later with the saved viral supernatant. Transduced cells were cultured in Dulbecco s modified Eagle s medium containing 10% FBS and selected with G418 from 48 h after the second infection. Selected colonies were transferred to 24- or 12-well plates and TGF- 1 production was measured with an enzyme-linked immunosorbent assay (ELISA) kit (Quantikine; R&D Systems, Minneapolis, MN). Reverse transcriptase-polymerase chain reaction Total RNA was extracted from cultured cells with the TRIzol reagent (GIBCO; Invitrogen, Carlsbad, CA). Reverse transcriptase-polymerase chain reaction (RT-PCR) was carried out with a SuperScript kit (Invitrogen). The RT and PCR were carried out with the following primers: type I collagen (forward, 5 -TAAAGGGTTCTGGGTC- 3 ; reverse, 5 -CGAACCACATTGGCATCA-3 ) and type II collagen (forward, 5 -CCAAAGGGACAGAAA- GGAGAACCT-3 ; reverse, 5 -CTCTGTGACCTTTGA- CACCAGGAA-3 ) under the following conditions: 95 C for 3 min followed by 95 C for 45 s, 58.5 C for 1 min, and 72 C for 1 min and 40 s, for 30 cycles. The PCR products were analyzed by agarose gel electrophoresis. -Actin was used as an internal control and the primer was designed by BD Biosciences Clontech. Nude mice and subcutaneous injection A mixture of hchon and hchon-tgf- 1, 1:1 ( : ) or 5:1 ( : ), or hchon-tgf- 1 ( ) cells alone in 50 L of serum-containing medium were injected intradermally into the backs of BALB/cAnNCrj-nu nude mice (n 4 per group, 7 to 8 weeks old; Charles River Japan, Yokohama, Japan). Primary untransduced human chondrocytes ( ) were injected as control.

3 1518 SONG ET AL. Rabbits and intraarticular injection New Zealand White rabbits weighing kg were selected for the animal study. Animals were mature and had a tidemark junction. The knee joint was exposed and a partial cartilage defect (3 6 mm, 1 2 mm deep) was made on the hyaline cartilage layer of the femoral condyle with a surgical knife. Those rabbits that showed bleeding due to bone damage were eliminated from the study. The same clone but different passage numbers (passage TABLE 1. HISTOLOGICAL AND HISTOCHEMICAL GRADING SCALE Score Nature of the predominant tissue Cellular morphology Hyaline cartilage 4 Mostly hyaline cartilage 3 Mostly fibrocartilage 2 Mostly noncartilage 1 Noncartilage only 0 Mason s trichrome staining of the matrix Normal or nearly normal 3 Moderate 2 Slight 1 None 0 Structural characteristics Surface regularity Smooth and intact 3 Superficial horizontal lamination 2 Fissures: 25 to 100% of thickness 1 Severe disruption, including fibrillation 0 Structural integrity Normal 2 Slight disruption, including cysts 1 Severe disruption 0 Thickness 100% of normal adjacent cartilage % of normal cartilage % of normal cartilage % of normal cartilage 0 Bonding to the adjacent cartilage Bonded at both ends of graft 2 Bonded at one end, or partially at both ends 1 Not bonded 0 Freedom from cellular changes of degeneration Hypocellularity Normal cellularity 3 Slight hypocellularity 2 Moderate hypocellularity 1 Severe hypocellularity 0 Chrondrocyte clustering No clusters 2 25% of cells % of cells 0 Total (maximum) 22 FIG. 1. (A) TGF- 1 protein secretion rates of hchon-tgf- 1 cells. The TGF- 1 protein secretion rate was measured by ELISA, and the average secretion rate of hchon-tgf- 1 cells used in this study was about 20 ng/10 5 cells per 24 h. (B) Collagen expression from chondrocytes. hchon cells (lane J; passage 6) do not express type II collagen, whereas hchon-tgf- 1 cells (lanes J- ; passage 5 and 6) do express type II collagen. 6 or 7) of human chondrocytes used in the nude mouse study were used for the rabbit experiments. A mixture of hchon and hchon-tgf- 1, 1:1 ( : ) or 5:1 ( : ), or hchon-tgf- 1 ( ) cells alone in L of medium containing serum were loaded on top of the defect. The cells were left in the defect for min to allow the cells to permeate the wound before suturing. Primary untransduced human chondrocytes ( ) were injected as control. The study was approved by the Institutional Animal Care and Use Subcommittee of Inha University College of Medicine (Inchon, South Korea). Histology and histological evaluation Nude mice or rabbits were killed 4 or 6 weeks after injecting the cells. Staining was performed by published methods. 13 After sacrifice, newly formed tissues or knee samples were examined macroscopically and photographed. Tissues and the distal part of the femur were fixed in formalin for 3 days and then decalcified in nitric acid for an ad-

4 HYALINE CARTILAGE REGENERATION ditional 3 to 4 days. Blocks of specimens were embedded in paraffin and cut into 0.5- to 0.8-mm-thick slices. Sections were stained with hematoxylin eosin, safranin O, toluidine blue, or Masson s trichrome for microscopic examination of the repair tissue. Sections from each animal were randomized to eliminate bias, and then examined and scored independently by three independent researchers (S.U.S., K.B.C., and H.Y.L.). Samples were graded according to a histological scale (refer to Table 1), which is a modification of the method described by O Driscoll et al. 16 Specifically, Masson s trichrome was used instead of toluidine blue for histological staining as it provided a more obvious mechanism to distinguish regenerated cartilage tissue from normal cartilage tissue. The scale is composed of eight categories and assigns a score ranging from 0 to 22 points. Immunohistological staining Sections, prepared as described above, were deparaffinized and hydrated by sequential incubations in xylene and ethanol. After washing in 1 phosphate-buffered saline (PBS) for 2 min, sections were blocked with 3% H 2 O 2 for 10 min. Primary antibody against TGF- 1 (a gift from S.-J. Kim, National Institutes of Health, Bethesda, MD), type I collagen (Santa Cruz Biotechnology, Santa Cruz, CA), or type II collagen (Southern- Biotech, Birmingham, AL) was applied to the sections and incubated for 1 h. Control sections were incubated in 1 PBS without the primary antibody. Sections were then washed and blocked with 5% skim milk in 1 PBS for 20 min before incubation with horseradish peroxidase (HRP)-conjugated secondary antibody. The chromogen reaction was performed with 0.05% diaminobenzidine (DAB) in 1 PBS for 5 min, and sections were subsequently stained with hematoxylin and mounted. RESULTS Construction of hchon-tgf- 1 cells 1519 To generate TGF- 1 -expressing chondrocytes, primary human chondrocytes were infected with recombinant TGF- 1 retroviruses, in which the expression of the TGF- 1 gene was driven by the CMV promoter. Several stable cell lines (hchon-tgf- 1 ) were then generated by G418 selection, as described in Materials and Methods. Levels of active TGF- 1 proteins were determined by FIG. 2. Cartilage formation in nude mice 6 weeks after subcutaneous injection of hchon-tgf- 1 cells with or without the addition of untransduced hchon cells. (A C) Photographs of nude mice 6 weeks after injecting hchon-tgf- 1 without (A) or with hchon cells (B and C). (D F) Photographs of newly formed, skin-on tissues in nude mice. (G I) Photographs of newly formed, skin-off tissues in nude mice. Cartilage-like tissues were formed after injecting hchon-tgf- 1 cells with or without hchon into nude mice (RT), whereas no tissues formed after injecting hchon cells alone (LT). LT, left; RT, right.

5 1520 ELISA, and high producers were selected. The average rate of TGF- 1 expression by the hchon-tgf- 1 cells used in this experiment was about 20 ng/10 5 cells per 24 h (Fig. 1A). Collagen expression was checked by the RT-PCR method. hchon-tgf- 1 cells express type II collagen continuously, whereas hchon cells do not. Type I collagen is expressed in both cells (Fig. 1B). Cartilage formation in nude mice To explore whether cartilage tissue can be formed in vivo by the administration of a mixture of hchon and hchon-tgf- 1 cells or by hchon-tgf- 1 cells alone, SONG ET AL. cells were injected intradermally into the backs of immune-deficient nude mice and cartilage tissue formation was monitored. Normal primary human chondrocytes (hchon) were used as controls. As shown in Fig. 2, cartilage-like tissues were formed in the backs of the nude mice, when either mixed cells or hchon-tgf- 1 cells alone were injected (Fig. 2A C, right). In contrast, hchon cells alone did not induce a cartilage formation (Fig. 2A C, left). These cartilage-like tissues started to form 1 week after injection of the cells and kept growing constantly for up to 7 8 weeks postinjection. In some animals, necrosis of skin cells around the site of injection was observed approximately 4 weeks after treatment. Fig- FIG. 3. Histological examination of newly formed tissue 6 weeks after injecting hchon-tgf- 1 cells with or without hchon cells into nude mice. (A, D, G, and J) Hematoxylin eosin (H&E) staining of newly formed tissues in nude mice. (B, E, H, and K) Toluidine blue (TB) staining of newly formed tissues in nude mice. (C, F, I, and L) Safranin O (S-O) staining of newly formed tissues in nude mice.

6 HYALINE CARTILAGE REGENERATION ure 2G I shows the newly formed tissues in nude mice with the skin layer removed. We performed an additional experiment in which the cell number was fixed at but included other ratios of cell types; results indicated a similar effect (data not shown). Histological and immunohistochemical staining of newly formed cartilage tissue in nude mice To examine whether the newly formed tissue in nude mice had cartilage-like characteristics, histological analyses were performed by hematoxylin eosin (H&E), toluidine blue (TB), and safranin O (S-O) staining. Cartilage specific TB and S-O staining of tissues showed that the newly formed tissues in nude mice administered mixed cells or hchon-tgf- 1 cells alone were similar, in terms of staining depth and color, to normal cartilage tissue (Fig. 3E, F, H, I, K, and L). To determine whether injected hchon-tgf- 1 cells continued to produce TGF- 1 and whether they induced the expression of type II collagen protein, we performed immunohistochemical staining with antibodies against TGF- 1 and human type II collagen. The results obtained showed the strong expression of TGF- 1 and human type II collagen in the newly formed tissues in nude mice (Fig. 4D, E, G, H, J, and K). Higher TGF- 1 expression was FIG. 4. Immunohistological examination of newly formed tissues 6 weeks after the subcutaneous injection of hchon-tgf- 1 cells with or without hchon into nude mice. (A, D, G, and J) TGF- 1 staining of newly formed tissues in nude mice. (B, E, H, and K) Type II collagen staining of newly formed tissues in nude mice. (C, F, I, and L) Secondary antibody control staining of newly formed tissues in nude mice.

7 1522 observed in the specimen injected with a 1:1 ratio of mixed cells than with the 5:1 ratio, because of the greater density of hchon-tgf- 1 cells in the 1:1 ratio. These results suggest that TGF- 1 -expressing chondrocytes are able to express TGF- 1 transgene and to induce type II collagen expression in vivo. Type I collagen expression was observed at similar levels in all samples, in accordance with RT-PCR results in Fig. 1B (data not shown). Regeneration of hyaline cartilage with a mixture of hchon and hchon-tgf- 1 cells, or with hchon-tgf- 1 cells alone, in rabbits To determine whether mixed cells or hchon-tgf- 1 cells alone were able to stimulate the repair of a hyaline cartilage defect, cells were implanted into an artificially created partial defect in the femoral condyle of a rabbit knee joint as previously described. 13 Cartilage defects treated with a mixture of cells were found to have filled in with regenerated tissues by 6 weeks postinjection (Fig. 5C H). SONG ET AL. In contrast, in most cases, hchon cells alone did not fill the defect with repair tissue (Fig. 5A and B). These animal experiments were evaluated on the basis of a modified histological grading scale, previously described by O Driscoll et al. 16 (Table 1). The grading results summarized in Table 2 indicate that mixed cells at a ratio of 5:1 maximally stimulated hyaline cartilage tissue repair. The grading scores for mixed cells at ratios of 5:1 and 1:1 were and , respectively. These values are greater than that with hchon-tgf- 1 cells alone ( ). The grading score of hchon cells alone still shows a modest level ( ), indicating that there is some spontaneous cartilage regeneration in this rabbit model. Immunohistochemical and histological staining analyses of reparative cartilage tissue in rabbits To investigate the expression of TGF- 1 and type II collagen after the implantation of mixed cells, regenerated tissues were analyzed by immunohistochemical and FIG. 5. Regeneration of hyaline articular cartilage after injury through administration of hchon-tgf- 1 cells with or without hchon cells in rabbits. (A, C, E, and G) Photographs of the femoral condyle of rabbit knee 6 weeks after injecting hchon cells alone (A), hchon-tgf- 1 cells alone (C), a mixture of hchon and hchon-tgf- 1 cells (1:1 ratio) (E), or a mixture of hchon and hchon-tgf- 1 cells (5:1 ratio) (G). (B, D, F, and H) Overview of defects, stained with Masson s trichrome, from knee joints injected with hchon cells alone (B), hchon-tgf- 1 cells alone (D), a mixture of hchon and hchon-tgf- 1 cells (1:1 ratio) (F), or a mixture of hchon and hchon-tgf- 1 cells (5:1 ratio) (H).

8 HYALINE CARTILAGE REGENERATION 1523 TABLE 2. COMPARISON OF HISTOLOGICAL GRADING SCORES SIX WEEKS AFTER INJECTION a hchon hchon- hchon hchonhchon hchon-tgf- 1 TGF- 1 (1 1) TGF- 1 (5 1) b Number of animals Nature of the predominant tissue Cellular morphology Masson s trichrome staining of the matrix Structural characteristics Surface regularity Structural integrity Thickness Bonding to the adjacent cartilage Freedom from cellular changes of degeneration Hypocellularity Chrondrocyte clustering Total Abbreviations: hchon, human chondrocytes; hchon-tgf- 1, transforming growth factor- 1 -producing human chondrocytes. a Values represent averages SD for each category, as assessed by three different investigators. b Total score for hchon hchon-tgf- 1 (5 1) was significantly different (p 0.05) from those for hchon or hchon-tgf- 1 alone. histological staining 4 weeks postinjection in the abovedescribed rabbit model (n 5 each). Immunohistochemical staining of tissue sections with TGF- 1 and type II collagen antibodies showed that the proteins were expressed at detectable levels in the regenerated tissue (Fig. 6A and B). These results indicated that injected cells were still present and functioning 4 weeks postinjection and that the newly formed tissue was in the process of forming cartilage. Newly formed tissues were stained with H&E, Masson s trichrome, safranin O, or toluidine blue and representative results at 4 weeks postinjection are presented in Fig. 7. Masson s trichrome and safranin O staining suggested that extracellular matrix proteins, such as colla- gens, were present (Fig. 7B and C). On the other hand, toluidine blue staining indicated that carbohydrates were not fully saturated in the reparative tissue (Fig. 7D). DISCUSSION This study shows that mixed cells are capable of inducing hyaline-like articular cartilage more efficiently than are hchon-tgf- 1 or hchon cells alone in vivo. The nature of the newly formed tissue in the rabbit knee defect after implanting mixed cells was similar to that of normal hyaline cartilage tissue. These results suggest the clinical viability of implanting TGF- 1 -producing human FIG. 6. Representative results of immunohistological staining of regenerated cartilage 4 weeks after administration of a mixture of hchon and hchon-tgf- 1 cells. Shown are regenerated tissues stained with TGF- 1 (A), type II collagen (B), and control secondary antibody (C) against goat immunoglobulin protein. The expression of TGF- 1 protein was detected in cells mostly on the surface of cartilage and in some cells in the regenerated cartilage (RC). The expression of type II collagen was detected in both regenerated cartilage (RC) and normal cartilage (NC) tissues. The arrows point to the stained areas. Insets: Higher magnification views.

9 1524 SONG ET AL. FIG. 7. Representative results of histological staining of the regenerated cartilage formed 4 weeks after the subcutaneous injection of a mixture of hchon and hchon-tgf- 1 cells. (A D) Regenerated articular cartilage was stained with hematoxylin - eosin (A), Masson s trichrome (B), safranin O (C), or toluidine blue (D). The regenerated hyaline-like articular cartilage formed after administering a mixture of hchon and hchon-tgf- 1 cells shows normal hyaline cartilage characteristics. chondrocytes in combination with untransduced primary human chondrocytes for hyaline articular cartilage regeneration. We have previously shown that hyaline cartilage tissue can be regenerated by a cell-mediated gene therapy approach using TGF- 1 -producing NIH 3T3 mouse embryonic fibroblasts. 13 Furthermore, we hypothesized that the TGF- 1 produced by the injected TGF- 1 -transduced fibroblasts would stimulate chondrocytes and chondrocyte precursors residing in the defect area to induce cartilage regeneration. If this hypothesis is true, cartilage regeneration by this method may be dependent on the number of surrounding chondrocytes or chondrocyte precursors. Therefore, in the present study, we tested whether cartilage regeneration by TGF- 1 -producing chondrocytes was enhanced by coinjection of additional primary chondrocytes into the defect site. These additional primary chondrocytes would generate various ECM molecules, cytokines, and growth factors, required for cartilage healing, after being stimulated by the TGF- 1 produced from the transduced chondrocytes. It is likely that TGF- 1 -producing chondrocytes themselves can also produce the factors necessary for cartilage healing after being stimulated by the autocrine action of TGF- 1. We believe that the results, taken together, indicate that the reparative systems of cartilage tissue are fully activated in response to a mixture of hchon to hchon-tgf- 1 cells. We also investigated how the ratio of primary chondrocytes to TGF- 1 -producing chondrocytes affects cartilage regeneration. Histological grading results indicate that a 5:1 ratio of hchon to hchon-tgf- 1 cells produced more enhanced cartilage regeneration than did a 1:1 ratio or hchon-tgf- 1 cells alone. From a clinical safety point of view, using this proposed mixed cell approach provides a mechanism to optimally activate the regenerative process, using the lowest possible amount of recombinant exogenous growth factors. The mixed cell approach demonstrated in this article also provides a viable alternative to using stem cells. Current research involving stem cells inherently involves using a mixed cell approach, as typically they are not highly purified from their source. As a result, any therapeutic use of such stem cells will have to overcome the hurdle of cellular differentiation at the population level. Thus, we believe the mixed cell approach discussed here overcomes these and other issues, by selecting the specific clonal populations used through limiting dilution and phenotype selection methods. Significantly, this also overcomes the plethora of ethical issues associated with the use of stem cells. In addition, untransduced hchon cells serve as additional bulk material for filling the defect site, and provide additional target cells for TGF- 1 expressed from infected cells due to the autocrine and paracrine modes of action of TGF- 1.

10 HYALINE CARTILAGE REGENERATION Our previous approach to the issue of tissue regeneration was also xenogeneic, involving the transplantation of mouse fibroblasts into a rabbit knee joint; no significant immune reaction was observed in the study. 13 This is attributed to the avascular nature of the cartilage and the immunosuppressive activity of TGF- 1 protein. H&E staining of surrounding tissues of the rabbit knee into which TGF- 1 -producing cells were injected in the present study also showed no severe inflammation response (data not shown). The molecular mechanism that controls cartilage regeneration during cell-mediated gene therapy still remains to be resolved. However, these data suggest that TGF- 1 is involved in the proliferation and differentiation of chondrocytes, and it is possible that TGF- 1 secreted by the injected chondrocytes stimulates both primary chondrocytes and TGF- 1 -producing chondrocytes via paracrine or autocrine modes, respectively, thus initiating a redifferentiation of the dedifferentiated state of the chondrocytes It is also likely that TGF- 1 may activate other molecules involved in chondrocyte differentiation and healing. The strength of the regenerated cartilage to withstand the mechanical forces within the joint is the focus of further and ongoing investigation. ACKNOWLEDGMENTS These studies were supported by a grant from the Korea Health 21 R&D Project, the Ministry of Health and Welfare, Republic of Korea (02-PJ2-PG4-PT ). REFERENCES 1. Buckwalter, J.A., Einhorn, T.A., and Simon, S.R. Orthopaedic Basic Science: Biology and Biomechanics of the Musculoskeletal System, 2nd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons. 2000, pp Bentley, G., and Greer, R.B. Homotransplantations of isolated epiphyseal and articular cartilage chondrocytes into joint surfaces of rabbits. Nature 230, 385, Brittberg, M., Lindahl, A., and Nilsson, A. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N. Engl. J. Med. 331, 889, Brittberg, M. Autologous chondrocyte transplantation. Clin. Orthop. 367(Suppl.), 147, Buckwalter, J.A., and Mankin, H.J. Articular cartilage: Tissue design and chondrocyte matrix interactions. Instr. Course Lect. 47, 477, Chen, F.S., Frenkel, S.R., and DiCesare, P.E. Chondrocyte transplantation and experimental treatment options for articular cartilage defects. Am. J. Orthop. 26,, 396, Chesterman, P.J., and Smith, A.U. Homotransplantation of articular cartilage and isolated chondrocytes: An experimental study in rabbits. J. Bone Joint Surg. 50A, 184, Gillogly, S.D., Voight, M., and Blackburn, T. Treatment of 1525 articular cartilage defects of the knee with autologous chondrocyte implantation. J. Orthop. Sports Phys. Ther. 28, 241, Grande, D.A., Singh, I.J., and Pugh, J. Healing of experimentally produced lesions in articular cartilage following chondrocyte transplantation. Anat. Rec. 218, 142, Hunziker, E.B., and Rosenberg, L.C. Repair of partialthickness defects in articular cartilage: Cell recruitment from the synovial membrane. J. Bone Joint Surg. 78A, 721, Johnston, B., and Yoo, J.U. Autologous mesenchymal progenitor cells articular cartilage repair. Clin. Orthop. 367(Suppl.), 156, Kimura, T., Yasui, N., Ohsawa, S., and Ono, K. Chondrocytes embedded in collagen gels maintain cartilage phenotype during long-term cultures. Clin. Orthop. 186, 231, Lee, K.H., Song, S.U., Hwang, T.S., Yi, Y., Oh, I.S., Lee, J.Y., Choi, K.B., Choi, M.S., and Kim, S.-J. Regeneration of hyaline cartilage by cell-mediated gene therapy using transforming growth factor 1 -producing fibroblasts. Hum. Gene Ther. 12, 1805, Mankin, H.J. The reaction of articular cartilage to injury and osteoarthritis. N. Engl. J. Med. 291, 1285, Wakitani, S., Goto, T., Young, R.G., Mansour, J.M., Goldberg, V.M., and Caplan, A.I. Repair of large full-thickness articular cartilage defects with allograft articular chondrocytes embedded in a collagen gel. Tissue Eng. 4, 429, O Driscoll, S.W., Keeley, F.W., and Salter, R.B. Durability of regenerated articular cartilage produced by free autogenous periosteal grafts in major full-thickness defects in joint surfaces under the influence of continuous passive motion. J. Bone Joint Surg. 70A, 595, Yang, X., Chen, L., Xu, X., Li, C., Huang, C., and Deng, C.X. TGF- /Smad3 signals repress chondrocyte hypertrophic differentiation and are required for maintaining articular cartilage. J. Cell Biol. 153, 35, Wildemann, B., Schmidmaier, G., Ordel, S., Stange, R., Hass, N.P., and Raschke, M. Cell proliferation and differentiation during fracture healing are influenced by locally applied IGF-1 and TGF- 1 : Comparison of two proliferation markers, PCNA and BrdU. J. Biomed. Mater. Res. 65B, 150, Morales, T.I., and Roberts, A.B. Transforming growth factor regulates the metabolism of proteoglycans in bovine cartilage organ cultures. J. Biol. Chem. 263, 12828, Redini, F., Daireaux, M., Mauviel, A., Galera, P., Loyau, G., and Pujol, J.P. Characterization of proteoglycans synthesized by rabbit articular chondrocytes in response to transforming growth factor- (TGF- ). Biochim. Biophys. Acta 1093, 196, Iwasaki, M., Nakata, K., Nakahara, H., Nakase, T., Kimura, T., Kimata, K., Caplan, A.I., and Ono, K. Transforming growth factor- 1 stimulates chondrogenesis and inhibits osteogenesis in high density culture of periosteum-derived cells. Endocrinology 132, 1603, Izumi, T., Scully, S.P., Heydemann, A., and Bolander, M.E. Transforming growth factor 1 stimulates type II collagen expression in cultured periosteum-derived cells. J. Bone Miner. Res. 7, 115, 1992.

11 O Driscoll, S.W., Recklies, A.D., and Poole, A.R. Chondrogenesis in periosteal explants: An organ culture model for in vitro study. J. Bone Joint Surg. Am. 76, 1042, Goldberg, A.J., Lee, D., Bader D.L., and Bentley, G. Autologous chondrocyte implantation. Culture in a TGF- containing medium enhances the re-expression of a chondrocytic phenotype in passaged human chondrocytes pellet culture. J. Bone Joint Surg. Br. 87, 128, SONG ET AL. Address reprint requests to: Dug Keun Lee, Ph.D. or Kwan Hee Lee, M.D. TissueGene 209 Perry Parkway, Suite 13 Gaithersburg, MD dklee@tissuegene.com or ortholee@tissuegene.com