Regenerative effects of human chondrocyte sheets in a xenogeneic transplantation model using immune‐deficient rats

Abstract Although cell transplantation has attracted much attention in regenerative medicine, animal models continue to be used in translational research to evaluate safety and efficacy because cell sources and transplantation modalities are so diverse. In the present study, we investigated the regenerative effects of human chondrocyte sheets on articular cartilage in a xenogeneic transplantation model using immune‐deficient rats. Osteochondral defects were created in the knee joints of immune‐deficient rats that were treated as Group A, untreated (without transplantation); Group B, transplantation of a layered chondrocyte sheet containing 5.0 × 105 cells (layered chondrocyte sheet transplantation); Group C, transplantation of a synoviocyte sheet containing 5.0 × 105 cells (synoviocyte sheet transplantation); or Group D, transplantation of both a synoviocyte sheet plus a layered chondrocyte sheet, each containing 5.0 × 105 cells (synoviocyte sheet plus layered chondrocyte sheet transplantation). Histological evaluation demonstrated that Group B showed cartilage regeneration with hyaline cartilage and fibrocartilage. In Groups C and D, the defect was filled with fibrous tissue but no hyaline cartilage. Transplanted cells were detected at 4 and 12 weeks after transplantation, but the number of cells had decreased at 12 weeks. Our results indicate that layered chondrocyte sheet transplantation contributes to articular cartilage regeneration; this model proved useful for evaluating these regenerative effects.

Translational models including rats (Takaku et al., 2014;Takatori et al., 2018), rabbits (Ito et al., 2012;Kaneshiro et al., 2006;Tani et al., 2017), and minipigs (Ebihara et al., 2012) were used previously to confirm the safety and efficacy of chondrocyte sheet transplantation. Next, we conducted a clinical study using autologous transplantation of human chondrocyte sheets to treat patients with knee OA, after which we detected no severe adverse events after more than 3 years of follow-up. In addition, histological analysis of the regenerated cartilage demonstrated that chondrocyte sheet transplantation could lead to regeneration of hyaline cartilage . Recently, we reported the mode of action of chondrocyte sheets in hyaline cartilage regeneration by transcriptomic and proteomic analyses (Toyoda et al., 2019). We are currently conducting our second clinical study using allogeneic transplantation of chondrocyte sheets fabricated from unwanted surgical tissue derived from patients with polydactyly .
To evaluate new human cell sources of chondrocyte sheets, it is preferable to use translational models that can directly evaluate the final product. In other words, xenogeneic orthotopic transplantation models that approximate the in vivo efficacy of clinical transplantation are required. We previously reported the use of an immunosuppressed rabbit xenogeneic transplantation model to directly evaluate human chondrocyte sheets for articular cartilage repair , but the duration for which immunosuppression was effective limited long-term studies. Ito et al. (2012) and Shimizu et al. (2015) reported conflicting results concerning the effectiveness of chondrocyte sheets and synoviocyte (SY) sheets using allogeneic transplantation in rabbits and rats, respectively. In the present study, we investigated the regenerative effects of human chondrocyte sheets and SY sheets for articular cartilage regeneration in a xenogeneic model using immune-deficient rats to establish a translational model that can approximate the efficacy of allogeneic transplantation of human chondrocyte sheets.

| MATERIALS AND METHODS
All procedures using animals in this study were performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 2010)

| Fabrication of LC sheets and SY sheets
Articular cartilage and synovial tissues were collected from the unwanted postoperative tissues of six patients (age 71-78 years; average age 74 years) who underwent total knee arthroplasty at Tokai University Hospital. Articular cartilage and synovial tissues were enzymatically digested separately to isolate chondrocytes and synoviocytes, respectively. Layered chondrocyte sheets (LC sheets) and SY sheets were made according to the methods described by Kokubo et al. (2016) and Takahashi et al. (2018), as summarized below. Briefly, articular cartilage or synovial tissues were minced in a petri dish using scissors, and then digested in Dulbecco's modified Eagle's medium-Nutrient Mixture F-12 (DMEM/F12; Gibco, Waltham, MA, USA) containing 5 mg/ml collagenase type I (Worthington, NJ, USA), 20% fetal bovine serum (FBS; Ausgenex, Molendinar, Australia), and 1% antibiotic-antimycotic solution (AA; Gibco). The cell suspension was passed through a 100-μm cell strainer (Becton Dickinson and Company [BD], Franklin Lakes, NJ) and washed in 1 × Dulbecco's phosphate-buffered saline (PBS; Gibco). Primary chondrocytes were suspended in Cellbanker 1 cryopreservation medium (Zenoaq, Fukushima, Japan) and cryopreserved at −80 C. Synoviocytes were cultured to Passage 1 and preserved in the same manner as the chondrocytes. The frozen chondrocytes and synoviocytes were thawed, and the chondrocytes were seeded on temperature-responsive culture inserts (UpCell ® 4.2 cm 2 , CellSeed, Tokyo, Japan) at 50,000 cells/cm 2 , whereas the synoviocytes were seeded on six-well companion plates (Corning, Corning, NY, USA) at 10,000 cells/cm 2 , followed by coculturing. The culture medium was initially composed of DMEM/F12, 20% FBS, and 1% AA. The medium was changed every 3 or 4 days thereafter, with the addition of 100 μg/ml ascorbic acid (Wako Pure Chemical Industries, Osaka, Japan). After coculturing for 2 weeks, three chondrocyte sheets were layered using a polyvinylidene difluoride membrane and were cultured for another 7 days.
The synoviocytes were also collected in sheet forms and were cultured for another 7 days. To count the cells, fabricated LC sheets and SY sheets were digested enzymatically, and trypan blue exclusion assays were performed. LC sheets and SY sheets were cut appropriately just prior to transplantation so that they contained approximately 5.0 × 10 5 cells.

| RNA extraction and reverse transcription-PCR
LC sheets and SY sheets were incubated with TRIzol (Thermo Fisher Scientific) for RNA extraction. Total RNA was extracted using the RNeasy Mini Kit (Qiagen Inc., Valencia, CA, USA), according to the manufacturer's instructions. cDNA was synthesized using the QuantiTect Reverse Transcription Kit (Qiagen Inc., Hilden, Germany) at 42 C for 15 min and 95 C for 3 min. Table 1 lists the oligonucleotide primers used. Polymerase chain reaction (PCR) was performed using SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) and the cDNA template (2.5 ng) in a total reaction volume of 25 μl using a 7300 Real-time PCR system (Applied Biosystems) at 95 C for 15 s, 60 C for 60 s, for 35-45 cycles. The results were analyzed using Smart Cycler II software (Applied Biosystems). The internal control glyceraldehyde-3-phosphate dehydrogenase was used as a reference gene, and the comparative 2 -ΔΔCt method was used for analysis.

| Measurement of humoral factors
A random selection of LC sheets and SY sheets were further cultured for 72 h in 3 ml of DMEM/F12, supplemented with 1% FBS and 1% AA incubated at 37 C under appropriate cell culture condition.
Supernatants were collected and centrifuged at 15,000 g for 10 min to remove cell debris. The concentrations of transforming growth factor-β-1 (TGF-β1; R&D Systems), and melanoma inhibitory activity (MIA; Roche, Basel, Switzerland) were measured using enzyme-linked immunosorbent assay (ELISA) kits. The signal detected for blank medium containing 1% FBS was subtracted to adjust for the protein content of the FBS. Measurements were repeated at least twice for each donor, and averages are reported. The concentrations of TGF-β1 or MIA were normalized by the number of cells (1.0 × 10 6 cells).

| Transplantation of LC sheets and SY sheets
Forty-eight 12-week-old rats (F334/NJcl-rnu/rnu; Clea Japan, Tokyo, Japan), weighing approximately 270 g, were used in the transplantation experiments. The animals were housed two animals per cage and were given daily standard chow and access to water ad libitum. LC sheets and SY sheets fabricated from each of the donors were allocated equally to each transplantation group. The rats were anesthetized using isoflurane, N 2 O, and O 2 . An osteochondral defect was created using the methods described by Itokazu et al. (2016). Briefly, a medial parapatellar incision was made on the right knee; the patella was moved laterally and an osteochondral defect (diameter 2 mm; depth 1 mm) was created on the patellar groove of the femur using a biopsy punch (Kai Industries, Seki, Japan) and hand drill. Marrow bleeding during creation of the osteochondral defect was confirmed.
Next, LC sheets and SY sheets were transplanted as: Group A, untreated (without transplantation); Group B, LC sheet containing 5.0 × 10 5 cells; Group C, SY sheet containing 5.0 × 10 5 cells, and Group D: both SY and LC sheets, each containing 5.0 × 10 5 cells. For Group D, the SY sheet was transplanted first and then covered with T A B L E 1 Primers used for reverse-transcriptase-polymerase chain reaction
the LC sheet. The patella was anatomically repositioned and the quadriceps femoris muscle, tendon, and skin were sutured. After the surgery, all the rats were returned to their cages without splinting or immobilization.

| Histological evaluation of repaired cartilage
The animals were euthanized by administration of high-dose anesthesia at 4 or 12 weeks after transplantation. The operated knee was opened, and the distal portion of the femur was excised and fixed for 1 week in 20% formalin (Wako Pure Chemical Industries).

| Statistical analysis
The results are presented as mean ± standard deviation, and p < 0.05 was deemed significant. Statistical analysis was performed using SPSS v23.0 software (IBM Corp., Armonk, NY, USA). The reverse transcription PCR and ELISA results were analyzed using the Mann-Whitney U test. Analysis of variance was used to analyze the weight distribution ratios and ICRS scores, and the Tukey honest significant difference (HSD) method was used for post hoc tests. Compared with SY sheets, LC sheets secreted higher concentrations of TGF-β1 (LC sheets 1.6 ± 0.2 ng/1.0 × 10 6 cells; SY sheets 0.40 ± 0.02 ng/1.0 × 10 6 cells) and MIA (LC sheets 16.5 ± 2.1 ng/1.0 × 10 6 cells; SY sheets were below the detection limit).

| Pain evaluation
The weight distribution ratio was used as a measure of pain and was tested daily for 28 days after transplantation (Figure 2a). Group A showed minimal improvement from Day 1 to Day 28: 30.9 ± 6.8% to 41.9 ± 5.0%. Groups B, C, and D all showed improvement from Day 1 to Day 28: 30.1 ± 5.3% to 47.9 ± 3.9% in Group B, 31.8 ± 4.5% to 45.3 ± 4.0% in Group C, and 32.2 ± 6.4% to 46.3 ± 3.6% in Group D. The weight distribution ratios for Day 28 were significantly higher for Groups B and D than for Group A, as shown in Figure 2b (p < 0.05).

| Microscopic analysis of repaired tissue
No signs of infection or rejection were observed in any group during the experiment. Safranin O staining was performed to evaluate the repaired tissue, as shown in Figure 3. Four weeks after transplantation, the defects in Groups B, C, and D were filled with repaired tissue.
In Group A, the defects were not filled with a chondral layer, but with either bone-like tissue or fibrous tissue that did not stain for Safranin O or COL II staining but did stain for COL I. In Group B, the repaired tissue showed Safranin O and COL II staining and no COL I staining, good integration with the surrounding cartilage, a high defect-filling rate, and a smooth superficial layer, although the subchondral bone formation was inadequate. In Groups C and D, the repaired tissue tissue, a high defect-filling rate, a smooth superficial layer, and a columnar arrangement of the chondrocytes. There was moderate formation of subchondral bone. In Groups C and D, the repaired tissue had not changed significantly from that which was present 4 weeks after transplantation. A modified version of the ICRS grading system was used to evaluate cartilage repair (Table 2). Figure 4 and Table 3 show the ICRS grading system results 4 and 12 weeks after transplantation. At 4 weeks after transplantation, the scores for Groups A, B, C, and D were 24.6 ± 4.1, 32.3 ± 5.4, 24.9 ± 3.2, and 25.2 ± 5.9, respectively. The score for Group B was significantly higher than that for Group A. At 12 weeks after transplantation, the scores for Groups A, B, C, and D were 22.8 ± 4.6, 33.5 ± 4.4, 23.3 ± 2.1, and 25.5 ± 2.1, respectively. The score for Group B was significantly higher than those for Groups A, C, and D.

| DISCUSSION
Appropriate translational models that can directly evaluate the final outcomes are necessary to confirm the safety and efficacy of human cell-derived products. In the present study, LC sheets, similar to those used in our previous clinical study, and SY sheets fabricated from synovium-derived cells, were evaluated in vitro and in vivo.
LC sheets and SY sheets were both negative for Safranin O staining and positive for fibronectin and COL I immunostaining.
Compared with the chondrocyte sheets fabricated from the surgical waste material of younger patients (Kokubo et al., 2016), the LC sheets used in this study were fabricated from older patients, which may explain why these LC sheets did not stain for Safranin O or COL II. Gene expression analysis revealed that compared with SY sheets, LC sheets had significantly higher expression of COL2A1 and SOX9 suggesting that they had better chondrogenic properties. MMP13 expression was significantly higher in LC sheets than in SY sheets. In previous study, the LC sheets restrain the catabolic factors matrix metalloproteinase 3, MMP13, and ADAMTS5 in allogenic transplantation on rabbit model (Kaneshiro et al., 2006). Also, MMP13 expression did not differ significantly between LC sheets and layered SY sheets, and that ADAMTS4 and 5 were significantly higher in the layered SY sheets than the LC sheets in allogenic transplantation on rat model (Shimizu et al., 2015). The difference is that the cells for the current study were collected from older patients with knee OA. Furthermore, collagen, type X, alpha 1 expression in this study was higher in the LC sheets than in SY sheets. Thus, it is possible that LC sheets may be particularly affected by OA.
ELISA analysis revealed that LC sheets produced significantly higher levels of TGF-β1 and MIA, which have, respectively, been suggested to make important contributions to cartilage repair and to be a marker for the cartilage phenotype. For each group, n = 6. At 4 weeks after transplantation, the scores for Groups A, B, C, and D were 24.6 ± 4.1, 32.3 ± 5.4, 24.9 ± 3.2, and 25.2 ± 5.9, respectively. ICRS scores in Group B were significantly higher than those in Group A (p < 0.05). At 12 weeks after transplantation, the scores for Groups A, B, C, and D were 22.8 ± 4.6, 33.5 ± 4.4, 23.3 ± 2.1, and 25.5 ± 2.1, respectively. ICRS scores in Group B were significantly higher than those in Groups A, C, and D (p < 0.05). The results are presented as mean ± standard deviation. LC: layered chondrocyte; SY: synoviocyte; ICRS: International Cartilage Regeneration and Joint Preservation Society adhesiveness and maintenance of the cartilage phenotype (Mitani et al., 2009). In the previous study, we confirmed that LC sheets secrete MIA, TGF-β1, and prostaglandin E2, which are anabolic factors involved in cartilage repair (Hamahashi et al., 2015).
Cartilage repair was not observed in Groups C and D at 4 or 12 weeks after transplantation, and the osteochondral defects were filled with fibrous tissue. Ito et al. (2012) reported that, in rabbits, better cartilage regeneration was achieved by transplanting a combination of LC sheets and SY sheets than by LC sheets or SY sheets alone. Shimizu et al. (2015) reported that in rats, the transplantation of SY sheets as a monolayer or a triple layer did not promote hyaline cartilage repair. Our results show that SY sheets, alone or in combination with LC sheets at a ratio of 1:1, did not promote hyaline cartilage repair. These differences may relate to the differing characteristics of SY sheets from different animal species (rabbits, rats, or humans) and the fact that, in our study, the cells were collected from older patients with knee OA. Note: Group A, untreated; Group B, LC sheet containing 5.0 × 10 5 cells alone; Group C, SY sheet containing 5.0 × 10 5 cells alone; and Group D, SY sheet plus LC sheet, each containing 5.0 × 10 5 cells. LC: layered chondrocyte; SY: synoviocyte. For each group, n = 6. ICRS: International Cartilage Regeneration and Joint Preservation Society; Ti: tissue morphology; Matx: matrix staining; Stru: structural integrity; Clus: cluster formation; Tide: tidemark opening; Bform: bone formation; SurfH: histological appraisal of surface architecture; FilH: histological appraisal of the degree of defect filling; Latl: lateral integration of defect-filling tissue; Basl: basal integration of defect-filling tissue; InfH: histological signs of inflammation. The total score (Hgtot) range is 11-45. reported immune rejection at 12 weeks after transplantation, when the effects of the immunosuppression wore off. In the present study, we used F334/NJcl-rnu/rnu rats for xenogeneic transplantation experiments. These rats have an immunological deficit in T-cell function and are not likely to reject transplanted tissues; therefore, replacement of transplanted cells by host-derived cells secondary to chondrogenesis and time is believed to be the reason for the decreased number of cells detected.
There were two limitations to this study. First, 12 weeks posttransplantation is a short period for evaluating cartilage repair.
Therefore, evaluation over a longer term is warranted. Second, this study was conducted on small animals. The rat model is low cost and suitable for the selection of cell sources; however, validation of the efficacy of transplantation is required in larger animals that more closely resemble humans.
Evaluation of human chondrocyte sheets using immune-deficient rats was possible for up to 12 weeks after transplantation. The in vitro evaluation revealed that human chondrocyte sheets with a cartilagespecific phenotype contributed to cartilage regeneration in vivo. We plan to verify the efficacy of various cell sources for chondrocyte sheets using this translational model.

ACKNOWLEDGMENTS
We are grateful to the Support Center for Medical Research and