Mice with a deficiency in the Col9a2 gene (Col9a2^-/-) were provided by the Nanjing Biomedical Research Institute of Nanjing University. The DNA from tail biopsies of mice was genotyped using PCR. All mice were housed at a constant room temperature of 20±2˚C and humidity of 50-60% with a 12-h light/dark cycle and free access to water and standard food. A total of 36 male mice at 4 (12-15 g body weight), 8 (20-23 g body weight) and 12 (26-29 g body weight) weeks of age were used for further analysis. At each time-point, the Col9a2^-/- group was compared with the wild-type (WT) control group (n=6 per group). All mice were humanely euthanized via intraperitoneal injection of pentobarbital sodium (150 mg/kg). Animal death was confirmed by observation of cessation of heartbeat and respiration. The present study was approved by the Ethics Committee of Zhejiang Chinese Medical University (Hangzhou, China; approval no. 20190401-10) according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

The lower thoracic and whole lumbar vertebrae of mice were fixed in 4% neutral formaldehyde solution at 4˚C for 48 h, transferred to 70% anhydrous ethanol and then scanned and examined using a high-resolution µCT scanner (Skyscan 1176; Bruker) with the following settings: Isotropic voxel size resolution, 9 µm; energy, 45 kV; current, 500 µA; and integration time, 780 msec. The lower thoracic ribs were included for the identification of L4-L5 IVD localization. Image reconstruction and analysis was performed using NRecon v1.6 (Bruker) and CTAn v1.9 (Bruker), respectively, followed by analysis of the L4-L5 IVD parameters using three-dimensional model visualization software (CTVolx, v3.0; Bruker). The IVD volume was defined by the region of interest (ROI) to cover the entire invisible space between the L4-L5 vertebrae and the cartilage EP volume was defined as covering the visible bone plate close to the vertebrae. A total of 50 consecutive ROI images were used to display the three-dimensional reconstruction of the EP and IVD space. The average of the volume and the height of the IVD were calculated.

The specimens were fixed in neutral formaldehyde solution at a volume fraction of 4% (0.1 mol/l, pH 7.4) for 3 days and then decalcified in 10% EDTA (buffered with pure water, pH 7.4) for 2 weeks. The samples were embedded in paraffin, cut into 3 µm-thick sections and stained with Alcian blue hematoxylin/orange G (ABH/OG) and safranin O and fast green. A total of six different mice from each group at each time-point were selected for the assessment of EP, annulus fibrosus (AF) and nucleus pulposus (NP) using a light microscope (Carl Zeiss AG). The EP score was obtained as previously described and used to evaluate the degeneration of the EP, including structural disorganization, clefts and bony sclerosis (24,25).

IHC was performed using a standard protocol. In brief, 3-µm-thick paraffin-embedded sections were incubated at 60˚C for 4 h prior to deparaffinizing and dehydrating. The sections were then incubated in citrate buffer (0.01 M, pH 6.0; Beijing Solarbio Science & Technology Co., Ltd.) at 60˚C for 4 h, or in pepsinum (OriGene Technologies, Inc.) at 37˚C for 30 min for antigen retrieval. The samples were then treated with endogenous peroxidase blocker (cat. no. PV-6001; OriGene Technologies, Inc.) for 10 min and incubated with 0.3% Triton X-100 for 15 min at room temperature. The sections were then incubated with primary antibodies against Col9a2 (diluted 1:100; cat. no. sc398130; Santa Cruz Biotechnology, Inc.), Col2a1 (diluted 1:1,000 for IHC and 1:50 for IF; cat. no. ab34712; Abcam), Aggrecan (diluted 1:200 for IHC and 1:50 for IF; cat. no. NB100-74350; Novus Biologicals, LLC), Mmp13 (diluted 1:200; cat. no. ab39012; Abcam), ADAM metallopeptidase with thrombospondin type 1 motif 5 (Adamts5; diluted 1:200; cat. no. ab182795; Abcam), Col10a1 (diluted 1:200; cat. no. ab58632; Abcam) and Runx family transcription factor 2 (Runx2; diluted 1:300; cat. no. ab76956; Abcam) overnight at 4˚C. For IHC staining, the sections were incubated with a secondary goat anti-mouse/rabbit antibody (diluted 1:1,000; cat. no. 31234; Invitrogen; Thermo Fisher Scientific, Inc.) for 20 min at room temperature. Positive staining of sections was visualized using diaminobenzidine solution (Invitrogen; Thermo Fisher Scientific, Inc.), while hematoxylin was used for counterstaining. For the IF assay, the slides were incubated with fluorophore-conjugated secondary antibodies at room temperature for 30 min in the dark. The sections were counterstained with DAPI and observed under a fluorescence microscope (Carl Zeiss AG). A total of five images from each section were analyzed using Image-Pro Plus 6.0 (Media Cybernetics, Inc.). The area of Col2a1 or Col10a1-positive staining was calculated, while the Aggrecan-, Mmp13-, Adamts5- or Runx2-positive cells were obtained by counting the number of positively stained cells.

The L4-L5 segments of the spinal cords were removed from 12-week-old mice and the soft tissues were excised. Total RNA was extracted using the Qiagen RNeasy Mini kit (Qiagen GmbH) and then reverse transcribed into cDNA using the RevertAid First Strand cDNA Synthesis kit (Invitrogen; Thermo Fisher Scientific, Inc.) according to manufacturers' protocols. As per the manufacturer's instructions, qPCR was performed for target genes using SYBR Premix Ex Taq™ II (Takara Biotechnology Co., Ltd.) with a QuantStudio™ 7 Flex Real-Time PCR System (Thermo Fisher Scientific, Inc.). The thermocycling conditions were: Pre-denaturation at 94˚C for 5 min; followed by 40 cycles of denaturation at 94˚C for 30 sec, annealing at 60˚C for 30 sec and extension at 72˚C for 30 sec. The primer sequences are presented in Table I and the relative mRNA expression was measured using the 2^-ΔΔCq method (26).

Statistical analysis was performed using SPSS software (version 25.0; IBM Corp). Values are expressed as the mean ± the standard error of the mean and analyzed using a one-way ANOVA followed by Bonferroni's post-hoc test as appropriate. P<0.05 was considered to indicate a statistically significant difference.

µCT scanning was implemented to qualitatively estimate the lumbar spine of the Col9a2^-/- and WT mice at 12 weeks of age. The cavities within the EPs were significantly increased in the CT three-dimensional images of the L4-L5 IVD (Fig. 1A), as well as in the median sagittal images of the CT scan (Fig. 1B). Although no statistically significant differences were observed, the IVD volume and height of Col9a2^-/- mice were reduced compared with those of WT mice (Fig. 1C and D), as determined through the analysis of the three-dimensional reconstruction images of the L4-L5 IVD. Thus, the CT data suggested that although there were no significant differences in the L4-L5 IVD volume and height between the Col9a2^-/- and WT mice, the ossification of the EPs was slightly increased and this is a phenotype of early-stage IVDD (27).

Histochemistry and histomorphometry were performed to further observe the phenotypes of IVDD induced by Col9a2 gene knockout. ABH/OG staining revealed that there were no obvious histological differences between the two groups at 4 and 8 weeks, except that the 8-week-old Col9a2^-/- mice exhibited a slight calcification of the EPs. The EPs of the Col9a2^-/- mice began to undergo significant degeneration at 12 weeks of age compared with the WT mice, as indicated by the apparent calcification and ossification formation of the EPs (Fig. 2A). Furthermore, oteoglycan loss was also present in the NP of Col9a2^-/- mice, as evidenced by less ABH/OG and safranin O and fast green staining compared with the WT mice (Figs. 3 and 4). Furthermore, a large number of clefts was observed in the AF of the Col9a2^-/- mice with collagen dislocation and cytopenia, and even AF rupture (Figs. 3 and 4B). The EP score, a histological evaluation of EP degeneration, was used to assess pathological changes, such as the degree of osteosclerosis, structural disorders and neovascularization. Of note, it was indicated that the EP score was significantly increased in the Col9a2^-/- mice compared with the WT mice at 12 weeks, confirming EP degeneration (Fig. 2B). The results of the histochemical and histomorphometric analysis indicated that the 12-week-old Col9a2^-/- mice exhibited certain other manifestations of IVDD, such as AF rupture and proteoglycan reduction, in addition to distinct EP degeneration, relative to WT mice.

Further experiments were performed to determine whether the protein levels of IVD-related proteins in Col9a2^-/- mice were also altered. IF was used to locate the expression of Col9a2 and verify the efficiency of gene knockout (Fig. 5A). Col9a2 was extensively expressed in the IVDs of WT mice, while its expression was low in the Col9a2^-/- mice (Fig. 5B-D). IHC and IF analysis were then performed to assess the expression levels of Col2a1, Aggrecan, Mmp13, Adamts5, Col10a1 and Runx2 in the IVDs. The expression levels of Col2a1 and Aggrecan, which indicate ECM deposition, were markedly decreased in the IVDs of Col9a2^-/- mice as compared with those in WT mice (Figs. 6 and 7). Furthermore, chondrocyte matrix damage, as assessed by staining for Mmp13 and Adamts5 in the EPs and AF of Col9a2^-/- mice, increased significantly compared with that in the WT mice (Fig. 8). Furthermore, chondrocyte hypertrophy, as evaluated by the number of Col10a1- and Runx2-positive cells, was uniformly distributed in the EPs and AF of WT mice, whereas in the EPs and outer layer of AF of Col9a2^-/- mice, positively stained cells were observed at sites where the chondrocytes merged together; however, positive cells were not present in the larger EP cavities and the expression levels of Col10a1 and Runx2 were markedly increased in the IVDs of Col9a2^-/- mice compared with those in WT mice (Fig. 9). These results suggested that the absence of the Col9a2 gene decreased the expression levels of ECM deposition-associated proteins and increased that of matrix degradation- and chondrocyte hypertrophy-related proteins in IVD cartilage tissues, which exacerbated the occurrence and development of IVDD.

RT-qPCR was performed to further determine the mRNA expression levels of ECM deposition-, matrix degradation- and chondrocyte hypertrophy-related genes in the IVD tissue of Col9a2^-/- mice. As expected, the data demonstrated that the mRNA expression trends of Col2a1, Aggrecan, Mmp13, Adamts5, Col10a1 and Runx2 were similar to the results of protein expression described above: Knockout of the Col9a2 gene diminished the mRNA expression levels of Col2a1 and Aggrecan (Fig. 10A and B), while the mRNA expression levels of Mmp13, Adamts5, Col10a1 and Runx2 were increased (Fig. 10C-F). These results suggested that deletion of the Col9a2 gene suppressed ECM synthesis and accelerated matrix degradation and chondrocyte hypertrophy in the IVD tissue.