Experiments were performed under a project license (HKDL2018377) granted by the Medical Ethics Committee of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, in compliance with Chinese national or institutional guidelines for the care and use of animals.

Six four-month-old male Chinese white pigs, weighing approximately 110 kg, were purchased from Shanghai Chuansha Experimental Animal Raising Farm, Shanghai, China. Fifteen eight-week-old male BALB/c-nude mice were provided by the Animal Laboratory, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China.

Cartilage tissues were harvested from adult pig ears and cut into circular cylinders with a diameter of 6 mm. Then, the cartilage sheets were obtained by freeze-sectioning at a thickness of 10 μm, followed by rinsing with 1% (wt/vol) sodium dodecyl sulfate (SDS, Sigma-Aldrich, St. Louis, MO, United States) in phosphate-buffered saline (PBS, HyClone, GE Healthcare, Little Chalfont, United Kingdom) for 1 day to remove cells (Young et al., 2011). Subsequently, the decellularized cartilage was rinsed with deionized water overnight to remove the detergent. For histological analyses, hematoxylin and eosin (HE) and 4,6-diamidino-2-phenylindole (DAPI, Biomol, Plymouth Meeting, United States) staining of the cartilage sheets were performed before and after decellularization. Then, cartilage sheets were lyophilized and milled into a fine powder for preparing hydrogels.

To liquefy the dcECM, the milled powder was resolubilized as described previously (Young et al., 2011; Fu et al., 2019). Briefly, 1 mg/ml porcine pepsin (Sigma-Aldrich, St. Louis, MO, United States) in 0.01 N HCl was used to digest the comminuted dcECM enzymatically under a constant stir rate for 24 h at room temperature. The pepsin digested dcECM stock solutions at 15 mg ECM/mL (dry wt.) were cryopreserved for subsequent experiments. Gelation of the stock solution was induced by neutralizing the pH with one-tenth the digest volume of 0.1 N NaOH, balancing the salt concentration of the pepsin digest with one-ninth the digest volume of 10 × PBS. Then, 1 × PBS was added at 4°C to obtain the desired dcECM concentration. The mixture was placed in a non-humidified incubator at 37°C for 30 min to form 1, 5, and 10 mg/ml dcECM hydrogels spontaneously.

To examine the surface morphology of dcECM hydrogels, SEM (JEOL 6380LV, Japan) was used as described previously (Xue et al., 2013). Briefly, dcECM hydrogels at concentrations of 1, 5, and 10 mg/ml were fixed in cold 2.5% glutaraldehyde for 24 h, rinsed in PBS, dehydrated using graded ethanol, and placed in 100% ethanol for 12 h at 4°C. Finally, the hydrogels were air-dried and sputter-coated with gold before imaging. The morphology of the specimens was photographed using SEM. The diameters of at least 100 fibers of each sample from different SEM images were measured and analyzed using Image J 1.50i software (National Institutes of Health, Bethesda, MD, United States).

The gelation kinetics of the cartilage dcECM hydrogels were evaluated turbidimetrically and compared between groups, as described previously (Young et al., 2011; Wolf et al., 2012). Briefly, 1, 5, and 10 mg/ml dcECM hydrogels were plated in a 96-well plate (100 μl/well) at 4°C. The plate was read using a spectrophotometer (Biotek Devices, Vermont, United States) and the absorbance was measured at 405 nm every 2 min for 1 h. Then the readings were scaled from 0 (at time 0) to 100% (at the maximum absorbance) to determine the normalized absorbance (NA) according to Equation (Hunziker, 1999). A is the absorbance at a given time, A₀ is the initial absorbance, and A_max is the maximum absorbance. The time required to reach 50 and 95% of A_max is defined as t₅₀ and t₉₅, separately, and the gelation rate (S) represents the slope of the linear region of the gelation curve. NA = ( A - A 0 ) / ( A max - A 0 ) (1)

The porcine chondrocytes were isolated and cultured, as described previously (Yan et al., 2009). Briefly, fresh cartilage tissue from one pig ear was cut into 2 × 2 mm² slices and digested with 0.25% trypsin plus 0.02% EDTA (Sigma-Aldrich, St. Louis, MO, United States) at 37°C for 30 min. Then, the slices were digested with 0.1% collagenase II (Sigma-Aldrich, St. Louis, MO, United States) in serum-free Dulbecco’s-modified Eagles medium (DMEM, HyClone, GE Healthcare, Little Chalfont, United Kingdom) at 37°C for an additional 12–16 h. The chondrocytes were counted and seeded in dishes at a cell density of 2 × 10⁴/cm² in DMEM with 10% fetal bovine serum (FBS, HyClone, GE Healthcare, Little Chalfont, United Kingdom). The cells at passage 1 and 2 were used for further experiments. Chondrocytes from different pigs were applied for repetitive experiments.

To determine the biocompatibility of dcECM hydrogels in vitro, we chose collagen I derived from rat tail (Col-I, BD Biosciences, San Jose, CA, United States) as the control, which gelatinized stably under the concentration of 1 mg/ml as reported (Fu et al., 2019). Chondrocytes were seeded in 24-well plates, coated with collagen I or 10 mg/ml dcECM hydrogels at a density of 2 × 10⁴ cells/well, and viable chondrocytes were imaged and quantified using the Cell Counting Kit-8 (CCK-8, Dojindo, Kumamoto, Japan) Assay. According to the manufacturer’s instructions, at 1, 3, 5, 7, 9, and 11 days after seeding, the cells were washed with PBS and incubated with 10% CCK-8 in DMEM for 2 h. Then, the absorbance of each well was measured at 450 nm using a microplate reader (ELX800, BioTek, Vermont, United States). The cell proliferation assay was presented by the mean optical density (OD) value from six wells, and experiments were repeated three times by using chondrocytes from different pigs.

To determine the biocompatibility and chondrogenesis supporting abilities of dcECM hydrogels in vivo, a subcutaneous transplantation model in nude mice was constructed, as described previously (Sawkins et al., 2013). Mice were separately kept in colony room with a 12-h light/dark cycle at 25 °C for 7 days before initiating experiments. Briefly, 350 μl collagen I (Col-I, 1 mg/ml) or the dcECM hydrogel at 1, 5, or 10 mg/ml was mixed with 50 μl chondrocytes (5 × 10⁶ cells in 50 μl PBS), respectively, to prepare Col-I grafts or dcECM grafts. Subsequently, this mixture was injected subcutaneously into the dorsal region of mice via 25 G needles at different points, the injection dose of the mixture at each injection point was 400 μl. Ten mice were randomly selected and each mouse received four plugs, which were from the different groups above. At 4 or 8 weeks after transplantation, five mice were randomly selected and sacrificed for harvesting the implants. In addition, five mice separately received one plug from the 10 mg/ml dcECM group, which would be harvested at 12 weeks after transplantation. The implants were used for gross evaluation and further analyses (n = 5/each time point). The weights and volumes were recorded and compared as reported previously (Xue et al., 2013).

Grafts were digested in papain solution (Sigma-Aldrich, St. Louis, MO, United States ) and assayed for soluble, triple-helical collagen content via the Sircol Collagen Assay (Biocolor Ltd., Carrickfergus, United Kingdom). To determine the GAG content, 1,9-dimethylmethylene blue dye solution (Sigma-Aldrich, St. Louis, MO, United States) was used according to the instruction of the kit. A pepsin buffer solution was used as the negative control and subtracted from the signal, as described previously (Young et al., 2011).

The implants harvested at investigated time points and normal cartilage tissue of pig ears were collected for further histological staining as described previously (Liu et al., 2008). Briefly, samples were fixed in 4% paraformaldehyde, paraffin-embedded, and sliced into 5-μm-thick sections. HE staining was performed to analyze the structure of implants. Toluidine blue and Safranin O staining were used to evaluate GAG deposition in the engineered cartilage tissues. Collagen II expression of pig chondrocytes was specifically detected by a mouse anti-human collagen II antibody (1:200; Abcam, Cambridge, MA, United States) and horseradish peroxidase-conjugated anti-mouse secondary antibody (1:50; Dako, Denmark). The sections were developed using diaminobenzidine tetrahydrochloride (DAB, Dako, Denmark), and images were acquired by microscope (Olympus BX51, Japan) and at least six representative fields from each sample were examined.

Data were expressed as the mean ± SD. The statistical analysis was performed with one-way analysis of variance test (ANOVA) for comparisons across multiple groups, followed by post hoc analysis Tukey test in GraphPad Prism 8.0 (GraphPad, San Diego, CA, United States). p < 0.05 was considered statistically significant.

The diameter of cartilage cylinders was 6 mm (Figure 1A) and the 10 µm thick cartilage sheets were sectioned and rinsed (Figures 1B,C). Following the decellularization protocol, the ECM fine powder was made (Figure 1D) and solubilized with pepsin to liquefy the cartilage matrix. This soluble matrix displayed properties similar to those of purified collagen gels, facilitating it to be a viscous liquid at 4°C and polymerizing after incubation at 37°C (Figure 1E). The histological analyses of sheets revealed that the cartilage ECM remained intact after decellularization and almost all nuclei were absent by DAPI staining (Figure 1F).

Hydrogels were successfully prepared from dcECM scaffolds at the concentrations of 1, 5, and 10 mg/ml. Qualitatively, the hydrogels of higher ECM concentrations (5 and 10 mg/ml) were more viscous than the 1 mg/ml (Figures 2A–C). The SEM images of the hydrogel surface showed randomly oriented fibrillar structures with interconnecting pores in dcECM hydrogels at different concentrations (Figures 2D–F). The dcECM hydrogels at 10 mg/ml contained the thickest fibrils and the highest fibril density as compared to the hydrogels at lower concentrations (Figures 2D–F). The fiber diameter increased non-linearly from 75.5 ± 13.5 nm at 1 mg/ml to 101.1 ± 22.0 nm at 10 mg/ml (Figures 2G–J).

To explore the properties of dcECM hydrogels, the gelation kinetics of the hydrogels at 1, 5, and 10 mg/ml were evaluated. The no change turbidity of 1 mg/ml indicated this concentration dcECM hydrogels could not cross-link well (Figure 3A). The turbidimetric gelation kinetics for 5 and 10 mg/ml showed sigmoidal shapes (Figures 3A,B). The lag phase for dcECM hydrogels at 5 mg/ml (6.153 ± 0.324) has no statistical significance with that at 10 mg/ml (7.876 ± 1.051) (Figure 3C). In addition, the times to reach 50 and 95% gelation, as well as the gelation rate (S) are also similar for dcECM hydrogels at 5 mg/ml and at 10 mg/ml (Figures 3D–F). The results indicated that the turbidimetric gelation kinetics for dcECM hydrogels at 5 mg/ml and 10 mg/ml are both good.

To evaluate the effects of dcECM on the viability of cells, chondrocytes were cultured on uncoated plates (TCP), Col-I-coated plates and 10 mg/ml dcECM hydrogel-coated plates (dcECM). The images of chondrocytes after 7 days of culture indicated that dcECM hydrogel coating could support the adhesion and proliferation of chondrocytes (Figures 4A–C). Furthermore, the proliferation of chondrocytes at day 1, 3, 5, 7, 9, and 11 after seeding on different coating plates was compared and shown in Figure 4D. The viability of chondrocytes on dcECM-coated plates was similar to those grown on uncoated TCP plates and Col-I-coated plates from day 3 to day 11, indicating dcECM hydrogels could support cells well.

To further explore the biological characteristics of dcECM hydrogels, chondrocytes mixed with Col-I or dcECM hydrogels at 1, 5, and 10 mg/ml were transplanted subcutaneously in nude mice. During the experimental period, we monitored the physical state and behavior of mice, and found that implants had no obvious adverse effects on mice. The representative macrography of grafts after 4 weeks post-grafting indicated that all concentrations of dcECM hydrogels had the same good gelation as collagen I in Figures 5A–D. The wet weights, volumes, collagen, and GAG contents of grafts increased gradually with increasing concentration of hydrogels (Figures 5I–L). Grafts of the 10 mg/ml group weighed similar to those of the Col-I group (Figure 5I) and the volumes of 10 mg/ml grafts were significantly larger than those of 1 mg/ml (Figure 5J) (p < 0.05). The collagen ratios of 10 mg/ml grafts were significantly higher as compared to those of the Col-I group (Figure 5K) (p < 0.01). However, the GAG assay did not show a marked difference between Col-I and dcECM hydrogel groups (Figure 5L). Furthermore, HE images indicated more viable chondrocytes and thicker tissues in the 10 mg/ml grafts in comparison to the other concentrations (Figures 5E–H). With increasing concentrations, the rate of positive Toluidine-blue and Safranin O staining tissue improved (Figures 6A–H); also collagen II immunohistochemistry indicated collagen deposition in the dcECM groups (Figures 6I–L). Taken together, the results put forth that dcECM hydrogels supported the viability of chondrocytes and the formation of lacunae in vivo, which induced the subcutaneous chondrogenesis.

To further explore the longer performances of dcECM hydrogels in vivo, grafts were analyzed at 8 weeks after transplantation (n = 5/each time point). Macroscopic appearances of grafts in Figures 7A–D showed that 10 mg/ml dcECM hydrogel grafts seemed bigger and more compact when compared with the 1 or 5 mg/ml group. Similarly, with the increasing concentrations of dcECM hydrogels, the wet weights, volumes and ratios of collagen and GAG contents of implants presented a rising trend via the 8-week culture in vivo (Figures 7I–L). The weights and volumes of 1 mg/ml grafts were significantly lower than those of Col-I grafts (p < 0.05), while the sizes of 10 mg/ml implants were similar to those of the Col-I group (Figures 7I,J). After 8 weeks, 10 mg/ml dcECM hydrogels could retain maximal collagen and GAG among the groups (Figures 7K,L). Further analyses of HE images showed that the dcECM hydrogels supported the survival of chondrocytes as well as collagen I (Figures 7E–H). Also, Toluidine-blue and Safranin O staining displayed the biological properties of dcECM hydrogels that improved the formation of cartilage-like tissues (Figures 8A–H). The expression of collagen II in dcECM hydrogel grafts was similar to that of grafts with collagen I (Figures 8I–L). Therefore, dcECM hydrogels could be deemed to perform adequately to maintain the long-term survival and function of chondrocytes in vivo.

Moreover, dcECM hydrogel grafts after 12 weeks of in vivo culture were compared to the original normal cartilage tissue of ears via immunohistochemistry analyses (Figure 9). HE staining showed similar numbers of chondrocytes and morphology of tissue between 10 mg/ml dcECM hydrogel grafts and source ear cartilage (Figures 9A,B). Toluidine blue, Safranin O, and collagen II staining showed chondrocytes in dcECM hydrogel were as rich as those in source tissue and apparent lacunae were also observed in the engineered cartilage tissue (Figures 9C–H). Therefore, the subcutaneous cartilage-like tissue formed by dcECM hydrogels was similar to the normal ear cartilage (Figures 9C–H).