International Cartilage Repair Society

Transcription

1 OsteoArthritis and Cartilage (2004) 12, 834e842 ª 2004 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. doi: /j.joca One day exposure to FGF-2 was sufficient for the regenerative repair of full-thickness defects of articular cartilage in rabbits 1 H. Chuma M.D.y, H. Mizuta M.D., Ph.D.y, S. Kudo M.D., Ph.D.y, K. Takagi M.D., Ph.D.y and Y. Hiraki Ph.D.z y Department of Orthopaedic and Neuro-Musculoskeletal Surgery, Faculty of Medical and Pharmaceutical Sciences, Kumamoto University, Kumamoto , Japan z Department of Cellular Differentiation, Institute for Frontier Medical Sciences, Kyoto University, Kyoto , Japan Summary International Cartilage Repair Society Objectives: Administration of fibroblast growth factor (FGF)-2 for 2 weeks induces a successful cartilaginous repair response in 5-mm fullthickness articular cartilage defects in rabbits. The purpose of this study was to investigate the effects of a short time exposure to FGF-2 on the repair of the defects. Methods: Five-mm-diameter cylindrical defects, which do not repair spontaneously, were created in the femoral trochlea of the rabbit knees. The defects were administered sterile saline or FGF-2 (150 pg/h) via an osmotic pump for the initial 1 day, 3 days, or 2 weeks, and we assessed the FGF-2 action on the proliferation and migration of mesenchymal cells in the reparative tissue. Using a total of 126 rabbits, we performed three sets of experiments. We also studied the effect of FGF-2 on migration of marrow-derived mesenchymal cells in vitro. Results: FGF-2 treatment for 1 day or 3 days induced the sequential chondrogenic repair responses that led to successful cartilaginous resurfacing of defects within 8 weeks as well as the 2-week treatment did. We confirmed by a radioisotope study that FGF-2 injected was rapidly eliminated from the defects (a residual ratio of 50% within 30 min). The effect of FGF-2 on cultured marrow-derived cells suggested that FGF-2 facilitated the mobilization and migration of replicating mesenchymal cells from bone marrow. Conclusions: Only 1 day exposure to FGF-2 is sufficient for induction of the chondrogenic repair response in 5-mm-diameter full-thickness defects of articular cartilage in rabbits. FGF-2 stimulated the recruitment of mesenchymal cells into the defects, which was a limiting step for the induction of cartilage. ª 2004 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. Key words: Articular cartilage, Chondrogenesis, FGF-2, Cell migration. Introduction Articular cartilage has a very limited capacity for repair. Partial-thickness defects that are limited to the layer of articular cartilage do not respond significantly to any repair process 1,2. Full-thickness defects penetrating the subchondral bone undergo repair processes that result in the generation of either fibrous or fibrocartilaginous tissue or, to a very limited extent, hyaline cartilage. Bone marrow contains pluripotent mesenchymal progenitor cells, which can differentiate into multiple differentiated cell-types such as chondrocytes, osteocytes, adipocytes, and fibroblasts 3. These marrow-derived mesenchymal cells have an essential role in the repair of the full-thickness defects of articular cartilage 4. The lineage progression of mesenchymal cells in 1 This work was partly supported by grants from the Ministry of Health, Labor and Welfare, and Grant-in Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. Address correspondence and reprint requests to: Dr Yuji Hiraki, Department of Cellular Differentiation, Institute for Frontier Medical Sciences, Kyoto University, 53 Shogoin-Kawahara-cho, Sakyo-ku, Kyoto , Japan. Tel/Fax: ; hiraki@ frontier.kyoto-u.ac.jp Received 25 November 2003; revision accepted 14 July the defect cavities largely depends on the size of the defects 4e6. In a mature rabbit model of small cylindrical defects (%3 mm in diameter) created in the femoral trochlea, marrow-derived mesenchymal cells spontaneously undergo chondrogenic differentiation within 2 weeks after the creation of the defects to regenerate surfacing articular cartilage concomitantly with the repair of subchondral bone 4. By contrast, chondrogenesis does not occur in larger defects (R5 mm in diameter), and the surface of the defects was covered only with fibrous tissue. The analyses of skeletal development in embryos and chondrogenic cell line in culture have shown that a variety of signaling molecules are implicated as a critical regulatory factor for a sequence of events during chondrogenesis 7e10. Based on these findings, attempts have been made to promote the repair of full-thickness defects by the use of such bioactive molecules 11e13. Fibroblast growth factor-2 (FGF-2) is known to be a potent mitogen for a wide variety of cell-types derived from mesoderm and neuro-ectoderm in vitro 14. Previous studies have also suggested that FGF signaling participates in the support of the proliferation of limb bud mesenchymal cells 15,16. We previously demonstrated the importance of FGF-2 released from surrounding bone matrix to the defect cavities for the cartilaginous repair: the local administration of neutralizing monoclonal 834

2 Osteoarthritis and Cartilage Vol. 12, No antibody against FGF-2 (bfm-1) interrupted the chondrogenic repair in the 3-mm-diameter small defects 6. Administration of FGF-2 for 2 weeks in the 5-mm-diameter large defects successfully stimulated accumulation of the proliferating cell nuclear antigen (PCNA)-positive proliferating mesenchymal cells to the same extent as in the 3-mmdiameter smaller defects at 1e2 weeks after creation of defects, and subsequently induced regeneration of articular cartilage and the subchondral bone within 8 weeks 17. Several reports documented the effects of a short time treatment to FGF-2 on chondrogenesis in vivo 18,19. Shida et al. demonstrated that single injection of FGF-2 into the young rat knee joint promoted the expansion of chondroprogenitor cell population in the femoral condylar ridge 18.In fracture-healing model, Kawaguchi et al. found a marked increase in the cartilage formation in the callus at 1 week after fracture, although more than 90% of FGF-2 disappeared from the injected site within 3 days 19. Thus we hypothesized that FGF-2 may act on the early stages of chondrogenic repair process and enable the full-thickness articular cartilage defects to initiate cartilaginous repair by administration of the factor for a shorter period within 2 weeks. Here we investigated the effects of the exposure to FGF-2 (150 ng/ml) for 1 day or 3 days on the repair response of 5-mm-diameter full-thickness defects. Clearance of FGF-2 administered into the defects was also assessed by the radioisotope study. In addition, using a radioisotope study, we assessed and examined the effect of FGF-2 on migration of marrow-derived mesenchymal cells in vitro. Materials and methods ANIMALS AND EXPERIMENTAL PROCEDURES Adolescent (15e16 weeks) male Japanese white rabbits, weighing 3.0e3.4 kg, were used in this study. All animals were housed individually, and animal care and experimental procedures were conducted in accordance with institutional guidelines ( ac.jp/card/japanese/kisoku/kisoku.html). Forty-two rabbits were anesthetized with intravenous administration of sodium pentobarbital (30 mg/kg body weight). The right knee joint was approached by means of a medial parapatellar incision under sterile conditions. The patella was dislocated laterally to expose the articular surface on the femoral trochlea. Full-thickness cylindrical defects (5 mm in diameter, 4 mm in depth) were created in the center of the femoral trochlea with a hand drill equipped with a 5-mmdiameter drill-bit, as described previously 6. After creation of the defects, the animals were fitted with an osmotic pump (Model 2002; Alzet, Palo Alto, CA, USA) connected with silastic medical grade tubing (0.75 mm ID/ 1.45 mm OD); a tubing about 5 mm long was introduced into the articular capsule. The pump was placed subcutaneously in the posterolateral region of the thigh. Twenty-one rabbits were fitted with an osmotic pump, which had been previously filled with 200 ml of sterile phosphate buffered saline (PBS) containing human recombinant FGF-2 (300 ng/ml; Takeda Chemical Industries, Osaka, Japan). The pump has a normal pumping rate of 0.5 ml/h over a 2-week period. Another group of twenty-one control animals received PBS alone with the same osmotic pump system. The articular capsule and skin were closed independently with 4-0 nylon sutures. All animals were allowed to walk freely without any splintage. HISTOCHEMICAL ANALYSES The animals treated with FGF-2 were randomly allocated into three groups: administration for 1 day, 3 days, or 2 weeks (n Z 7 each). Control animals were also randomly divided into three groups: administration of PBS for 1 day, 3 days, or 2 weeks (n Z 7 each). Depending on the experimental protocols, the animals were re-anesthetized on the final day of FGF-2 administration, and the osmotic pumps were removed without reopening the joint capsule. In previous studies 6,20, we confirmed that a tip of the tubing connected with an osmotic pump remains in the articular cavity of the knee joint during the experiment. All animals were sacrificed with an overdose of sodium pentobarbital at 8 weeks after creation of the defects for the subsequent histochemical analyses. This time point was chosen to evaluate the repair of the defects, because the previous study revealed that 3-mm-diameter defects, which heal spontaneously, and 5-mm-diameter defects with FGF- 2 treatment for 2 weeks regenerated the epiphyseal structure by 8 weeks 6. The distal portion of each femur was removed, and the harvested tissues were then fixed in 4% paraformaldehyde at room temperature for 1 h, decalcified with 10% ethylenediamine tetraacetic acid (EDTA) for 3 weeks, and embedded in paraffin. Sections (5 mm thick) were cut in the transverse plane, and stained with hematoxylin and eosin or with safranin-o. For semi-quantitative analysis of the reparative tissue, the sections were examined in a blinded manner by two observers not informed on the group assignment, and were scored according to the histological grading scale of Pineda et al., with some modifications (Table I) 21. The scale used here is inversely correlated to the original one so that better recovery of articular structure has a higher score; the score ranges from 0 (worst) to 14 (best) 6. Three sets of experiments were performed by using a total of 126 rabbits, and reproducibly yielded similar results. Table I Scoring system for the histological appearance of full-thickness defects of articular cartilagey Characteristics Score Filling of defects (%) Reconstitution of osteochondral junction Yes 2 Almost 1 Non-close 0 Matrix staining Normal 4 Reduced staining 3 Significant staining 2 Faint staining 1 No staining 0 Cell morphology Normal 4 Mostly hyaline and fibrocartilage 3 Mostly fibrocartilage 2 Some fibrocartilage, but mostly non-chondrocytic cells 1 Non-chondrocytic cell only 0 Perfect score 14 ymodified from Pineda et al. 21.

3 836 H. Chuma et al.: Short-term exposure to FGF-2 on cartilage repair ESTIMATION OF CELL DENSITY AND CELL PROLIFERATION IN THE REPARATIVE TISSUE We assessed the proliferative capacity of the mesenchymal cells that had migrated and filled the defects by immunostaining the PCNA, in addition to the packing cell density of the reparative tissues. Full-thickness defects were created for 36 rabbits, which were fitted with an osmotic pump as described above. Eighteen animals treated with FGF-2 were divided into three groups: administration of FGF-2 for 1 day, 3 days, or 1 week (n Z 6 each). Eighteen control animals treated with PBS were also divided into three groups: administration of PBS for 1 day, 3 days, or 1 week (n Z 6 each). Implanted osmotic pumps were removed at the end of FGF-2 administration, as described above. All animals were sacrificed at 1 week after creation of the defects for the subsequent histological and immunohistochemical analyses. The distal portion of each femur was removed, and the harvest tissues were fixed in 4% paraformaldehyde for 1 h, decalcified with 10% EDTA for 3 weeks, and embedded in paraffin. Sections (5 mm thick) were cut in the transverse plane, and stained with hematoxylin and eosin. For the immunostaining, the sections were deparaffinized, and hydrated. Endogenous peroxidase activity was blocked by 0.5% hydrogen peroxide in methanol and washed in 0.1% bovine serum albumin (BSA) in Tris-buffered saline (TBS). Nonspecific staining was reduced by incubation with normal horse serum. The sections were incubated with a monoclonal antibody against PCNA (dilution 1:400; Dakopatts, Copenhagen, Denmark) at room temperature for 1 h. Antibody binding was visualized by using a Vectastain avidinebiotineperoxidase complex (ABC) kit (Vector Laboratories, Burlingame, CA) in combination with diaminobenzidine (DAB) solution according to the manufacturer s instructions. The sections were then counterstained with methyl green. For the detection of PCNA, the growth plate of Japanese white rabbits was used as a positive control. Sections were incubated with horse serum instead of specific primary antibodies and stained as a negative control. The packing cell density of the reparative tissue in the defects was examined by the methods of Gallay et al. 22 using the sections stained with hematoxylin and eosin. A section of the defect was divided into 20 sampling square areas (1 mm! 1 mm). We selected 15 areas other than the five deeper peripheral ones where bone formation was evident. Two fields were selected at a magnification of!200 in each of sampling areas, and all nuclei except those of endothelial cells in capillaries and inflammatory cells were counted. To calculate the cell density in the defect, the mean number of cell nuclei was divided by the sampling area. Three sections from each paraffin block were used to calculate the average cell density for one animal. The mean and SD of thus obtained average cell densities was calculated for five animals in each experimental group. The PCNA-positive cells in the reparative tissue were counted according to the method of Aizawa et al. 23 and expressed by the percentage of the total number of cells in the defect. Sampling square areas (1 mm! 1 mm) were determined similarly as described above. Five fields were selected randomly at a magnification of!400 in each sampling area in the section. Because three sections from each paraffin block were used, the positive and negative undifferentiated cells were counted in a total of 225 fields for one animal. The rate of occurrence of PCNA-positive cells in each reparative tissue was expressed as the mean number of positive cells divided by the total number of cells for six animals in each experimental group. Sections were examined in a blinded manner by two evaluators not informed on the group assignment. Two sets of experiments were performed by using a total of 72 rabbits, and yielded similar results. ASSESSMENT OF CLEARANCE OF FGF-2 ADMINISTERED INTO FULL-THICKNESS DEFECTS Clearance of FGF-2 administered into the full-thickness articular cartilage defects was monitored using a radioisotope study 12,24, I-labeled FGF-2 was purchased from NEN Life Science Products, Inc. (Boston, MA, USA), and the specific activity was 88.4 mci/mg. For 12 rabbits, 5-mmdiameter full-thickness articular cartilage defects were created as above. The capsule and skin were repaired with 4-0 nylon sutures independently. On the day following the creation of the defects, three rabbits were re-anesthetized to expose the defect of the right knee again, and a total of 5ngof 125 I-labeled FGF-2 suspended in 5 ml of PBS, whose radioactivity was estimated to be mci, was injected into the blood clot filling the defect. After 30, 60, or 90 min, one rabbit at a time was sacrificed, and blood clots were immediately collected by wiping off with the cotton. The radioactivity included in the cotton was counted with an automatic gamma counter (ARC-380; Aloka, Tokyo, Japan). Just after sacrificing the animals, the same amount of 125 I-labeled FGF-2 was injected into the blood clot filling the defect of the left knee, and radioactivity in the clot was counted immediately in the same manner. All measurements of radioactivity were completed within 2 min after wiping off the blood clots with the cotton. In the preliminary experiments, we injected the same amount of 125 I-labeled FGF-2 into the cotton and the blood clot filling the defect, and the blood clots were immediately collected by wiping off with the cotton. The comparison of the radioactivity included in each cotton demonstrated that the recovery of the radioactivity was 91.3 G 4.8%. The residual radioactivity in each animal was calculated from the ratio of radioactivity in the clot of the right knee divided by that of the left knee. Four sets of experiments were performed by using a total of 12 rabbits within a day. The residual radioactivity in each time point was expressed as the mean G SD for four animals in each experimental group. MIGRATORY RESPONSE OF MARROW-DERIVED MESENCHYMAL CELLS TO FGF-2 The effect of FGF-2 on migration of marrow-derived mesenchymal cells was assessed in vitro. Rabbit marrowderived mesenchymal cell cultures were established using modified previously described methods 26,27. A rabbit was sacrificed with an overdose of sodium pentobarbital, and tibias and femurs were excised aseptically. Their proximal and distal ends were cut off, and the marrow from the femoral and tibial midshafts was flushed out into serum-free Ham s F-12 medium (Biofluids Biosource International, Rockville, MD, USA). A single cell suspension was prepared by sequentially drawing the marrow sample into a syringe three times through needles of decreasing size ( gauge 18, 21, 23). The dissociated cells were washed twice with PBS, pelleted by centrifugation for 5 min, and then resuspended in Ham s F-12 medium containing 10% fetal bovine serum (FBS; Cosmo Bio Co., Ltd., Tokyo,

4 Osteoarthritis and Cartilage Vol. 12, No. 10 Japan) and antibiotics ( penicillin G, 100 U/ml; streptomycin, 100 mg/ml; Celox Co., Hopkins, MN, USA). The bone marrow cells were seeded at 5.0! 10 6 nucleated cells/ml in a 90-mm plastic dish and cultured at 37(C in a humidified atmosphere of 5% CO 2. After 5 days, non-adherent cells were washed off. From this time onward the medium was changed every other day. When cells had grown to confluence, they were harvested by trypsinization, collected by centrifugation, suspended in serum-free Ham s F-12 medium containing 0.1% BSA, and used for subsequent experiments. 837 Migratory response of marrow-derived mesenchymal cells to a gradient of FGF-2 was measured in multiwell chemotaxis chambers with polycarbonate nucleopore filters of 8 mm pore size (Kurabo Industries Ltd., Osaka, Japan) using a modified checkerboard format. FGF-2 was added at a concentration of 0e50 ng/ml in serum-free Ham s F-12 medium containing 0.1% BSA to the bottom well of the chamber. The upper chamber was then placed on the plate, and a suspension of marrow-derived mesenchymal cells in serum-free Ham s F-12 medium containing 0.1% BSA supplemented with the concentration of 0e10 ng/ml of Fig. 1. Regeneration of 5-mm full-thickness defects treated with PBS alone for 1 day (A), 3 days (B) or 2 weeks (C), or with FGF-2 (150 pg/h) for 1 day (D), 3 days (E), or 2 weeks (F). Panel G is higher-power photomicrograph of the framed area in panel D. All sections were stained with safranin-o. Bar, 1 mm for AeF; 100 mm for G.

5 838 H. Chuma et al.: Short-term exposure to FGF-2 on cartilage repair FGF-2 was added in each upper well at 1! 10 5 cells/well. After incubation of the chamber for 5 h at 37(C in5%co 2, the upper surface of the filter of each well was rubbed with a cotton swab to remove the cells which had attached but not migrated. The filter was fixed in 100% methanol for 5 min, stained with Giemsa, and then mounted upside down on glass slides. For each well, the number of migrated cells was counted in 16 randomly selected fields under light microscopy at a magnification of!400. Seven sets of experiments were performed independently by using seven rabbits, and the data were expressed as the mean G SD in each experiment. STATISTICAL ANALYSES Statistical analyses on the histological score, the rate of occurrence of PCNA-positive cells, and data of migratory response were performed using KruskaleWallis test and ManneWhitney s U-test. A P value of less than 0.05 was considered significant. Results EFFECT OF FGF-2 ON REPAIR RESPONSE Total score FGF-2 PBS alone 1 Day 3 Days 2 Weeks Treatment period Fig. 2. Histological score for the regeneration of full-thickness defect at 8 weeks after creation of the defects according to the histological grading scale (Table I). The defects were treated with PBS alone or FGF-2 (150 pg/h) for 1 day, 3 days or 2 weeks. Values are means G SD (n Z 7). Asterisks indicate results that were significantly different between the two groups (Manne Whitney s U-test; P! 0.05). The histological appearance of the reparative tissues at 8 weeks after creation of 5-mm-diameter full-thickness defects is shown in Fig. 1. In the control groups treated with PBS alone, the defects were covered only with fibrous tissue, although the subchondral bone was almost fully reconstituted. The cells surfacing the defects were mostly non-chondrocytic and the matrix was not stained by safranin-o. By contrast, in the FGF-2-treated groups, welldeveloped cartilage stained by safranin-o covered the defects with the reconstruction level of the subchondral bone up to the original boneearticular cartilage junction, regardless of whether FGF-2 was administered for 1 day, 3 days, or 2 weeks. Spherical-shaped chondrocytes were arranged in vertical columnar distribution in the repair cartilage layer of the defects. Figure 2 shows the semiquantitative histological evaluation of the reparative tissues in each group according to the grading scale of Pineda et al. with some modifications 21. The scores of FGF-2-treated groups were significantly higher than those of control groups. Interestingly, the FGF-2 administration for 1 day or 3 days induced a remarkable chondrogenic repair response to the same extent as the administration for 2 weeks 6 (Fig. 2). The results indicate that the 1 day exposure to FGF-2 was sufficient for the induction of a regenerative repair response in large defects. EXPRESSION OF PCNA IN THE REPARATIVE TISSUE We assessed the proliferative capacity of the mesenchymal cells in the reparative tissues at 1 week after creation of the defects by immunostaining using the PCNA. We first stained the sections of the reparative tissues with hematoxylin and eosin. Then, semi-serial sections were stained with PCNA antibody, and examined at high magnification (Fig. 3). The histological appearance of the reparative tissues at 1 week in the FGF-2-treated defects was not significantly different from that of the control defects treated with PBS alone, regardless of the duration of FGF-2 administration. Undifferentiated spindle-shaped mesenchymal cells proliferated at the periphery of the defect, and the center of the defects was filled with blood clots. Osteoblastic differentiation took place near the surface of subchondral bone in the depth and edge of the defects (Fig. 3). In the control defects, immunoreactivity to PCNA was barely detectable in mesenchymal cells in the reparative tissue, except in the marginal area [Fig. 3(DeF)]. By contrast, most of the undifferentiated cells were positively stained with PCNA antibody in the FGF-2-treated defects [Fig. 3(JeL)] 28. There was no significant difference in the PCNA immunoreactivity among the groups treated with FGF-2. Table II summarizes the packing cell density and the rate of occurrence of PCNA-positive cells in the reparative tissues. The rate of occurrence of PCNA-positive cells was significantly higher in all FGF-2-treated defects (50% or more) than those of control defects at 1 week after creation of the defects 28. The packing cell density at 1 week was also significantly higher in all FGF-2-treated defects than that in the control defects. No significant difference was seen in the packing cell density among the groups treated with FGF-2. CLEARANCE OF 125 I-FGF-2 FROM THE DEFECTS Clearance of FGF-2 administered in the defects was assessed using a radioisotope study using 125 I-labeled FGF-2 24, as described in Materials and Methods section. The residual radioactivity in each time point is shown in Fig. 4. It rapidly decreased in a time-dependent manner. Radioactivity found 30 min after injection into the defect was approximately 40% of total, and only 20% of total activity remaining at 90 min after injection. MIGRATORY RESPONSE OF MARROW-DERIVED MESENCHY- MAL CELLS TO FGF-2 The migratory responses of rabbit marrow-derived mesenchymal cells to FGF-2 are shown in Fig. 5. Chemotactic migration was significantly stimulated by FGF-2 in a concentration-dependent manner from 0.1 ng/ml to 10 ng/ml. Maximum response was observed at 10 ng/ml of FGF-2, where migrated cell number was 9.6-fold higher than that in the absence of FGF-2. At a concentration of 50 ng/ml of

6 Osteoarthritis and Cartilage Vol. 12, No Fig. 3. Transverse sections of the reparative 5-mm-diameter defects treated with PBS alone for 1 day (A and D), 3 days, (B and E) or 1 week (C and F), and treated with FGF-2 (150 pg/h) for 1 day (G and J), 3 days, (H and K) or 1 week (I and L). All animals were killed at 1 week after creation of the defects. Hematoxylin and eosin staining (AeC, GeI) and PCNA immunostaining (DeF, JeL) were carried out. Panels D, E, F, J, K and L represent the PCNA-positive cells in the areas denoted by the framed areas in panels A, B, C, G, H and I, respectively, at a higher magnification. The sections were counterstained with methyl green. Bar, 1 mm for AeC and GeI; 125 mm for DeF and JeL. FGF-2, there was a decline in the migratory response, although a significant stimulation of chemotactic migration was still evident. We then performed a checkerboard analysis to distinguish chemotactic migration observed in response to a positive concentration gradient of FGF-2 from the random migration observed in response to FGF-2 (chemokinetic migration). As shown in Table III, the chemotactic action of FGF-2 was evident, but only minimum or no chemokinetic migration was observed. Discussion We previously reported that administration of FGF-2 for initial 2 weeks in the 5-mm-diameter full-thickness defects of articular cartilage, which is large enough not to repair spontaneously, resulted in the successful regeneration of articular cartilage and the subchondral bone within 8 weeks 6. In the present study, we investigated the effects of the short time exposure to FGF-2 ( for 1 day or 3 days) on the repair of the 5-mm-diameter defects. The results indicated that a short-term administration of FGF-2 was sufficient for the induction of a regenerative repair response in large defects. The FGF-2 administration for 1 day yielded the regenerative repair response in large defects to the same degree as the administration for 2 weeks. In the repair of full-thickness articular cartilage defects, undifferentiated cells migrating from the underlying bone marrow into the defect cavities are thought to be a primary source of repair cells 4, although synovium has been suggested as the other progenitor pool in several studies 2,29,30. In previous studies 6,17,20, we have shown a sequential reparative response after creation of fullthickness defects: the defect cavities are first filled with blood clots. Undifferentiated cells appear at the periphery of blood clots at 2e3 days after creation of defects, and then

7 840 H. Chuma et al.: Short-term exposure to FGF-2 on cartilage repair Table II Cell density and occurrence of the PCNA-positive undifferentiated cells in the reparative tissue at 1 week after creation of defects Treatment period Cell density (cells! 10 ÿ3 /mm) PCNA-positive cells (% of total) FGF-2 PBS alone FGF-2 PBS alone 1 day 1.22 G G G G days 1.12 G G G G week 1.21 G G G G 4.3 Values are means G SD for six animals. All the nuclei were counted in a total of 90 fields using three sections for each animal at a magnification of!200, as described in Materials and Methods. PCNA-positive and negative undifferentiated cells were also counted in a total of 225 fields using three sections for each animal at a magnification of!400. they migrate toward the center of the clot. By 1 week, the defect cavities are mostly filled with the mesenchymal cells 28. With immunohistochemical analysis using CD68, CD31 and alkaline phosphatase to assess the presence of macrophages, vascular endothelial cells, and osteoblasts, we detected no positive staining of these antigens in the reparative tissue of the defect cavities 28. These indicated that the reparative tissue is primarily composed of undifferentiated fibroblastic cells. Active expansion and maintenance of progenitor cell populations is thought to be an important prerequisite for initiation of chondrogenesis 13,31e33. We previously demonstrated that the capacity of reparative tissue to form cartilage in full-thickness articular cartilage defects was well correlated with the occurrence in the defects of PCNApositive undifferentiated cell populations 17 : the rate of occurrence of PCNA-positive cells in the reparative tissue at 1 week was about 60% in 3-mm-diameter defects which are small enough to regenerate articular cartilage spontaneously, while it was significantly lower in the 5-mmdiameter defects (about 15%) 17. Supplementation of FGF-2 for 1 week increased the rate of PCNA-positive cells in the 5-mm-diameter defects to the level of the 3-mm-diameter Clearance ratio (%) Time (min) Fig. 4. Time course of the 125 I-labeled FGF-2 clearance ratio. Values are mean G SD (n Z 4). Number of migrated cells (cells/field) Concentration of FGF-2 (ng/ml) defects. The present study revealed that the rate of occurrence of PCNA-positive cells was as high in the 5-mm-diameter defects treated with FGF-2 only for 1 day as in the defects treated with FGF-2 for 1 week. PCNA is a 36 kda acidic nuclear protein which is essential for DNA synthesis 34e36. This protein is synthesized from the late G 1 phase to the S phase of the cell cycle 35,37, which means that not all proliferating cells are stained by PCNA antibodies. The rate of occurrence of PCNA-positive cells is smaller than actual proportion of the proliferative component. Although the proportion of the duration of each phase in the cell cycle varies among the cell-types 38e40, high ratio of occurrence of PCNA-positive cells of greater than 50% in the defects treated with FGF-2 for 1 day may indicate that almost all repair cells in the defect cavities are in a proliferative state. Several investigators reported the rapid clearance of FGFs from the applied sites in vivo 19,24. FGF-2 injected into adult mouse knee joints remained for only several hours after injection 18. In agreement, when FGF-2 was injected into the blood clot filling the defect in our model, more than half of FGF-2 injected was cleared from the defect cavity within 30 min (Fig. 4). The half-life of PCNA is reported to be 8e60 h depending on the type of cells 41. Therefore, when the rapid clearance of FGF-2 is taken into consideration, it is unlikely that FGF-2 treatment for 1 day directly promotes a high proliferative activity of the marrow-derived mesenchymal cells, as indexed by a high rate of occurrence of PCNA-positive cells. The critical limiting step for the induction of cartilaginous repair is speculated to be migration and recruitment of marrow-derived mesenchymal cells into defect cavities, which is one of the earliest cellular events in the repair of full-thickness articular cartilage defects. In addition to the mitogenic effects, FGF-2 is known Fig. 5. Chemotactic response of marrow-derived mesenchymal cells to FGF-2. Marrow-derived mesenchymal cells placed in the upper well of the chamber after loading the lower well with FGF-2 at 0.1e50 ng/ml and FGF-free medium (0 ng/ml) as a negative control. Values are mean G SD for seven experiments. Significantly different from 0 ng/ml, significantly different from 0.1 ng/ml, and significantly different from 1 ng/ml (Manne Whitney s U-test; P! 0.05).

8 Osteoarthritis and Cartilage Vol. 12, No. 10 Table III Chemotaxis of bone marrow-derived mesenchymal cells to FGF-2 FGF-2 (ng/ml) Number of cells migrated into lower wells Upper well Lower well 0 (ng/ml) 10 (ng/ml) 0 (ng/ml) 8.7 G G (ng/ml) 83.6 G G 11.7 Cells that migrated into lower wells through the nucleopore filters were counted, as described in Materials and Methods. Values are means G SD for seven experiments. Asterisks indicate results that were significantly different from the [upper well 0 ng/ml vs lower well 10 ng/ml] group (ManneWhitney U-test; P! 0.05). to stimulate migration of different cell-types, including keratocytes, vascular endothelial cells, neural crest cells, and embryonic limb bud cells 42e45. However, only limited information is available regarding the migratory effects of FGF-2 on marrow-derived mesenchymal cell. Therefore, we examined the effect of FGF-2 on the migration of rabbit marrow-derived mesenchymal cells (Table III and Fig. 5). Isolation and cultures of marrow-derived mesenchymal cells were carried out using minor modifications of the reported methods 26,27. The osteochondrogenic potential of these cells was confirmed by diffusion chamber assay in nude mice 26. The in vitro migration assay showed that FGF- 2 stimulated the directional chemotactic migration of marrow-derived mesenchymal cells in a concentrationdependent manner, producing a peak response at 10 ng/ml. Thus the present in vitro results support our in vivo observations that mobilization and recruitment of mesenchymal cells is the primary site of FGF-2 action. In the present study, we demonstrated that the FGF-2 administration for 1 day yielded the successful chondrogenic repair response to the same extent as the administration for 3 days or 2 weeks. Histological analysis revealed that FGF-2 administration for 1 day was sufficient to regenerate the epiphyseal structure, including resurfacing articular cartilage and the subchondral bone up to the original bone-articular junction. Previous immunohistochemical studies using specific antibodies against type II collagen and glycosaminoglycans indicated that regenerated tissue contains abundant cartilaginous matrix components in 5-mm-diameter defect with FGF-2 treatment for 2 weeks 28. Further investigations are needed to clarify whether the duration of FGF-2 administration may affect the mechanical properties of the regenerated cartilage or not. Acknowledgments Recombinant human FGF-2 was a generous gift from Dr T. Kurokawa (Takeda Chemical Industries, Osaka, Japan). References 1. Kim HK, Moran ME, Salter RB. The potential for regeneration of articular cartilage in defects created by chondral shaving and subchondral abrasion. An experimental investigation in rabbits. J Bone Joint Surg Am 1991;73:1301e Hunziker EB, Rosenberg LC. Repair of partial-thickness defects in articular cartilage: cell recruitment from the synovial membrane. J Bone Joint Surg Am 1996;78: 721e Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999; 284:143e7. 4. Shapiro F, Koide S, Glimcher MJ. Cell origin and differentiation in the repair of full-thickness defects of articular cartilage. J Bone Joint Surg Am 1993;75: 532e Convery F, Akeson W, Koewn G. The repair of osteochondral defects: an experimental study in horses. Clin Orthop 1972;82:253e Otsuka Y, Mizuta H, Takagi K, Iyama K, Yoshitake Y, Nishikawa K, et al. Requirement of fibroblast growth factor signaling for regeneration of epiphyseal morphology in rabbit full-thickness defects of articular cartilage. Dev Growth Differ 1997;39:143e Shukunami C, Ohta Y, Sakuda M, Hiraki Y. Sequential progression of the differentiation program by bone morphogenetic protein-2 in chondrogenic cell line ATDC5. Exp Cell Res 1998;241:1e Shimizu A, Tada K, Shukunami C, Hiraki Y, Kurokawa T, Magane N, et al. A novel alternatively spliced fibroblast growth factor receptor 3 isoform lacking the acid box domain is expressed during chondrogenic differentiation of ATDC5 cells. J Biol Chem 2001;276: 11031e Kulyk WM, Rodgers B, Greer K, Kosher RA. Promotion of embryonic chick limb cartilage differentiation by transforming growth factor-b. Dev Biol 1989;135: 424e Haas AR, Tuan RS. Chondrogenic differentiation of murine C3H10T1/2 multipotential mesenchymal cells: II. Stimulation by bone morphogenetic protein-2 requires modulation of N-cadherin expression and function. Differentiation 1999;64:77e Sellers RS, Peluso D, Morris EA. The effect of recombinant human bone morphogenetic protein-2 (rhbmp-2) on the healing of full-thickness defects of articular cartilage. J Bone Joint Surg Am 1997;79: 1452e Sellers RS, Zhang R, Glasson SS, Kim HD, Peluso D, D Augusta DA, et al. Repair of articular cartilage defects one year after treatment with recombinant human bone morphogenetic protein-2 (rhbmp-2). J Bone Joint Surg Am 2000;82:151e Caplan AI, Elyaderani M, Mochizuki Y, Wakitani S, Goldberg VM. Principles of cartilage repair and regeneration. Clin Orthop 1997;254e Gospodarowicz D, Neufeld G, Schweigerer L. Molecular and biological characterization of fibroblast growth factor, an angiogenic factor which also controls the proliferation and differentiation of mesoderm and neuroectoderm derived cells. Cell Differ 1986;19: 1e Niswander L, Tickle C, Vogel A, Booth I, Martin GR. FGF-4 replaces the apical ectodermal ridge and directs out growth and patterning of the limb. Cell 1993;75:579e Niswander L. Interplay between the molecular signals that control vertebrate limb development. Int J Dev Biol 2002;46:877e Hiraki Y, Shukunami C, Iyama K, Mizuta H. Differentiation of chondrogenic precursor cells during the

9 842 H. Chuma et al.: Short-term exposure to FGF-2 on cartilage repair regeneration of articular cartilage. Osteoarthritis Cartilage 2001;9:S102e Shida J, Jingushi S, Izumi T, Iwaki A, Sugioka Y. Basic fibroblast growth factor stimulates articular cartilage enlargement in young rats in vivo. J Orthop Res 1996; 14:265e Kawaguchi H, Kurokawa T, Hanada K, Hiyama Y, Tamura M, Ogata E, et al. Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 1994;135:774e Kudo S, Mizuta H, Otsuka Y, Takagi K, Hiraki Y. Inhibition of chondrogenesis by parathyroid hormone in vivo during repair of full-thickness defects of articular cartilage. J Bone Miner Res 2000;15: 253e Pineda S, Pollack A, Stevenson S, Goldberg V, Caplan A. A semiquantitative scale for histologic grading of articular cartilage repair. Acta Anat Basel 1992;143: 335e Gallay SH, Miura Y, Commisso CN, Fitzsimmons JS, O Driscoll SW. Relationship of donor site to chondrogenic potential of periosteum in vitro. J Orthop Res 1994;12:515e Aizawa T, Kokubun S, Tanaka Y. Apoptosis and proliferation of growth plate chondrocytes in rabbits. J Bone Joint Surg Br 1997;79:483e Jingushi S, Heydemann SK, Kana SK, Macey LR, Bolander ME. Acidic fibroblast growth factor (afgf) injection stimulates cartilage enlargement and inhibits cartilage gene expression in rat fracture healing. J Orthop Res 1990;8:364e Kawai K, Suzuki S, Tabata Y, Ikada Y, Nishimura Y. Accelerated tissue regeneration through incorporation of basic fibroblast growth factor-impregnated gelatin microspheres into artificial dermis. Biomaterials 2000; 21:489e Ashton BA, Allen TD, Howlett CR, Eaglesom CC, Hattori A, Owen M. Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo. Clin Orthop 1980;294e Wakitani S, Goto T, Pineda SJ, Young R, Mansour JM, Caplan AI, et al. Mesenchymal cell-based repair of large, full-thickness defects of articular cartilage. J Bone Joint Surg Am 1994;76:579e Mizuta H, Kudo S, Nakamura E, Otsuka Y, Takagi K, Hiraki Y. Active proliferation of mesenchymal cells prior to the chondrogenic repair response in rabbit fullthickness defects of articular cartilage. Osteoarthritis Cartilage 2004;12:586e Hunziker EB, Driesang IM, Morris EA. Chondrogenesis in cartilage repair is induced by members of the transforming growth factor-beta superfamily. Clin Orthop 2001;S171e Hunziker EB. Growth-factor-induced healing of partialthickness defects in adult articular cartilage. Osteoarthritis Cartilage 2001;9:22e Osdoby P, Caplan AI. Osteogenesis in cultures of limb mesenchymal cells. Dev Biol 1979;73:84e Osdoby P, Caplan AI. A scanning electron microscopic investigation of in vitro osteogenesis. Calcif Tissue Int 1980;30:43e George M, Chepenik KP, Schneiderman MH. Proliferation of cells undergoing chondrogenesis in vitro. Differentiation 1983;24:245e Tan CK, Castillo C, So AG, Downey KM. An auxiliary protein for DNA polymerase-delta from fetal calf thymus. J Biol Chem 1986;261:12310e Bravo R, Frank R, Blundell PA, Macdonald-Bravo H. Cyclin/PCNA is the auxiliary protein of DNA polymerase-delta. Nature 1987;326:515e Prelich G, Tan CK, Kostura M, Mathews MB, So AG, Downey KM, et al. Functional identity of proliferating cell nuclear antigen and a DNA polymerase-delta auxiliary protein. Nature 1987;326:517e Bravo R, Celis JE. A search for differential polypeptide synthesis throughout the cell cycle of HeLa cells. J Cell Biol 1980;84:795e Steel GG. Growth Kinetics of Tumors. Oxford: Clarendon Press, Wright NA. Cell proliferation in health and disease. In: Anthony PP, MacSween RM, Eds. Recent Advances in Histopathology. Edinburgh: Churchill Livingstone 1984;17e Wright NA, Alison M. The Biology of Epithelial Cell Populations. Oxford: Clarendon Press, Gannon JV, Nebreda A, Goodger NM, Morgan PR, Hunt T. A measure of the mitotic index: studies of the abundance and half-life of p34cdc2 in cultured cells and normal and neoplastic tissues. Genes Cells 1998; 3:17e Andresen JL, Ehlers N. Chemotaxis of human keratocytes is increased by platelet-derived growth factor- BB, epidermal growth factor, transforming growth factor-alpha, acidic fibroblast growth factor, insulinlike growth factor-i, and transforming growth factorbeta. Curr Eye Res 1998;17:79e Kanda S, Lerner EC, Tsuda S, Shono T, Kanetake H, Smithgall TE. The nonreceptor protein-tyrosine kinase c-fes is involved in fibroblast growth factor-2-induced chemotaxis of murine brain capillary endothelial cells. J Biol Chem 2000;275:10105e Kubota Y, Ito K. Chemotactic migration of mesencephalic neural crest cells in the mouse. Dev Dyn 2000; 217:170e Webb SE, Lee KK, Tang MK, Ede DA. Fibroblast growth factors 2 and 4 stimulate migration of mouse embryonic limb myogenic cells. Dev Dyn 1997;209:206e16.