Research Paper 研究论文

Triclocarban triggers osteoarthritis via DNMT1-mediated epigenetic modification and suppression of COL2A in cartilage tissues
三氯卡班通过DNMT1介导的表观遗传修饰和软骨组织中COL2A的抑制引发骨关节炎

2.1. Zebrafish line and zebrafish maintenance
2.1. 斑马鱼线和斑马鱼维护

The Casper zebrafish line was obtained from the China Zebrafish Resource Center (Wuhan, China). The zebrafish were maintained at 28.5 ± 0.5 °C with a pH of approximately 7.4 and a conductivity of approximately 550 μS/cm. Zebrafish exposed to TCC or DMSO were placed in individual tanks with self-circulating water, and the water was replaced daily. All experiments conducted in the current study were approved by the Research Ethics Committee of Jinan University.
Casper斑马鱼品系是从中国斑马鱼资源中心（中国武汉）获得的。斑马鱼保持在28.5±0.5°C，pH值约为7.4，电导率约为550μS/cm。将暴露于TCC或DMSO的斑马鱼放入装有自循环水的单独水箱中，并每天更换水。本研究进行的所有实验均已获得暨南大学研究伦理委员会的批准。

2.2. TCC quantification in synovial fluid of OA patients
2.2. OA患者滑液中的TCC定量

Chemicals and materials: Triclocarban was purchased from Beijing Solarbio Science and Technology Co., Ltd (Beijing, China). Clozapine was purchased from National Institutes for Food and Drug Control (Beijing, China). HPLC-grade acetonitrile, methanol and isopropanol were purchased from TEDIA (USA). Water was prepared using Elga Purelab Felx 3 water purification system (ELGA, High Wycombe, UK). All other chemicals and solvents were of the analytical grade available. Joint fluid was donated by the first Affiliated Hospital of Jinan University.
化学品和材料：三氯卡班购自北京索罗生物科技有限公司（中国北京）。氯氮平购自国家食品药品检定研究所（中国北京）。HPLC级乙腈、甲醇和异丙醇购自TEDIA（美国）。使用Elga Purelab Felx 3净水系统（ELGA，海威科姆，英国）制备水。所有其他化学品和溶剂均为可用分析级。关节液由暨南大学第一附属医院捐赠。

Patient sample preparation: To a 50 μL of joint fluid sample, 20 μL of the IS solution and 250 μL of acetonitrile were added. The samples were vortexed for 30 s and centrifuged for 10 min at 12,000 g. A 100 μL of the supernatant was transferred to an LC vial and diluted with 100 μL of water-acetonitrile (80/20: v/v). A 5 μL of the diluted supernatant was injected and analyzed by LC-MS/MS. All prepared samples were kept in an autosampler at 4 ℃ before injection.
患者样本制备：向 50 μL 关节液样本中加入 20 μL IS 溶液和 250 μL 乙腈。将样品涡旋30秒，并以12,000g离心10分钟。将 100 μL 上清液转移到 LC 小瓶中，并用 100 μL 水乙腈 （80/20： v/v） 稀释。进样 5 μL 稀释上清液，并通过 LC-MS/MS 进行分析。所有制备的样品在进样前保持在4°C的自动进样器中。

LC-MS/MS: LC-20 CE HPLC system (Shimadzu, Kyoto, Japan) was comprising a binary pump, a thermostatted autosampler and a thermostatted column compartment. An Ultimate XB-C18 column (5 µm, 50 ×2.1 mm, Welch, China) maintained at 40 ℃ was used for separation. A gradient program was used with the mobile phase, combining solvent A (methanol: acetonitrile: isopropanol 7:1.5:1.5, 0.2 % formic acid) and solvent B (10 mM ammonium formate in water) at a flow rate of 0.4 mL/min and with a total run time of 3.5 min as follows: 20 % A (0–0.2 min), 20–88 % A (0.21–1.3 min), 88–98 % A (1.31–1.8 min), 98 % A (1.81–2.4 min), 98–20 % A (2.41–2.5 min), 20 % A (3.51–3.5 min).

The HPLC system was connected to a QTRAP 5500 mass spectrometer (AB Sciex, Singapore) equipped with an electrospray ionization (ESI) source. The Turbo-Ion-Spray interface was operated in positive-ion mode (MRM: triclocarban 314.1/161.1; clozapine 327.2/270.0) with nitrogen as the nebulizing, turbo spray, and curtain gas, with the optimum values set at 40, 40, and 35 psi, respectively. The turbo-gas temperature was set at 500 ℃, and the ESI needle voltage in positive-ion modes was adjusted to 4500 V. Instrument control and data processing was carried out by the Analyst 1.6.2 software. The retention times of triclocarban and clozapine was 2.72 and 2.26 min, respectively.

Calibrators, quality control and internal standards preparation: Stock solutions of triclocarban and clozapine were prepared by dissolving these compounds in methanol at 0.2 and 1 mg/mL, respectively. The stock solution of triclocarban was further diluted with methanol to obtain working standard solutions at several concentrations. The calibration curve was obtained using six calibration standards, i.e., blank / joint fluid samples prepared by the addition of the working solutions to drug-free blank plasma/joint fluid sample, giving final concentrations of 0.5, 1, 2.5, 5, 8, 10 ng/mL for triclocarban. Calibration curves for triclocarban in human plasma/joint fluid were derived from their peak area ratios relative to that of internal standard (IS) using linear regression with 1/x² as a weighting factor. Quality control (QC) samples were prepared at three concentration levels (low, medium and high) for final concentrations of 1.5, and 7.5 ng/mL. The stock solution of clozapine was diluted to 200 ng/mL in methanol for routine use as a IS. All prepared plasma/joint fluid samples and stock solutions were stored at − 80 ℃.

Calibration curves were extracted in duplicate by adding the following volumes to 50 μL of joint fluid: 50 μL of the standard, 20 μL of the IS solution and 200 μL of acetonitrile. The standards were vortexed for 30 s and centrifuged for 10 min at 12,000 g. An aliquot of 100 μL of the supernatant was transferred to LC vial and diluted with 100 μL of water-acetonitrile (80/20: v/v). A 5 μL aliquot of the diluted supernatant was injected and analyzed by LC-MS/MS. The calibration curve for triclocarban was linear over the concentration ranges, with determination coefficients (r²) higher than 0.99.

Cells and cell culture: Human primary chondrocytes were purchased from ICell Bioscience Incorporated (Shanghai, China). Chondrocytes from six different healthy donors were employed in this study. All primary chondrocytes were characterized by col2a-positive immunofluorescent staining (Fig. S7). The cells were maintained in the standard medium provided by the company. Subculture was performed when the cells grew to 85 % confluence at a 1:2 ratio. The cells were incubated at 37 °C with 5 % CO₂.

2.3. Zebrafish imaging

Imaging of the anaesthetized or stained zebrafish was performed under a stereomicroscope (Olympus SZX7, Ming Mei, China). Hyperaemia was quantified by a virtual ruler in the software of the imaging system. Briefly, the virtual ruler was first calibrated by a physical ruler, and the calibrated virtual ruler was employed to detect the length of the whole anal fin (L0) and the length of the hyperemic area (Lh) in a double-blind manner. The hyperaemia rate is presented as 100 % X (Lh/L0).

2.4. Heartbeat rate

The treated zebrafish larva was placed on an agarose gel plate under a stereomicroscope (Olympus SZX7, Ming Mei, China), and the heartbeats were recorded over the course of a minute in a double-blind manner. The heartbeat rate was indicated as times/min.

2.5. Alcian blue and alizarin red double staining

The treated zebrafish were dissected, fixed with paraformaldehyde, and maintained in absolute ethanol. Staining was performed using 1 % Alcian Blue dissolved in 20 % glacial acetic acid and 80 % EtOH and 0.01 % alizarin red dissolved in 1 % KOH according to the standard protocol described in reference [28]. The stained zebrafish were placed on an agarose gel plate under a stereomicroscope (Olympus SZX7, Ming Mei, China). The intraarticular space was quantified by a virtual ruler in the imaging system. Briefly, the virtual ruler was first calibrated by a physical ruler, and the calibrated virtual ruler was employed to detect the shortest distance between the bone of the joint, which was stained with alizarin red in a double blind manner. The intraarticular space is presented in μm.

2.6. Metabolome analysis

Treated zebrafish were sacrificed in ice water, and the articular areas of the anal fin were dissected and subjected to metabolomics analysis. Metabolite extraction was primarily performed according to the following methods. In short, 25 mg tissues were weighed and extracted by directly adding 800 μL of precooled extraction reagent (methanol:acetonitrile:water (2:2:1, v/v/v)), internal standards mix 1 (IS1), and internal standards mix 2 (IS2) for quality control of sample preparation. After homogenization for 5 min using TissueLyser (JXFSTPRP, China), the samples were sonicated for 10 min and incubated at − 20 °C for 1 h. The samples were centrifuged for 15 min at 25,000 rpm at 4 °C, and the supernatant was then transferred for vacuum freeze drying. The metabolites were resuspended in 200 μL of 10 % methanol and sonicated for 10 min at 4 °C after centrifugation for 15 min at 25,000 rpm, and the supernatants were transferred to autosampler vials for LCMS analysis. A quality control (QC) sample was prepared by pooling the same volume of each sample to evaluate the reproducibility of the whole LCMS analysis. LCMS/MS was employed for untargeted metabolomics analysis. Data from both positive and negative ions were collected by a high-resolution mass spectrometer Q Exactive HF (Thermo Fisher Scientific, USA) to improve the coverage. After MS analysis, the data were processed in Compound Discoverer 3.1 (Thermo Fisher Scientific, USA) software. Metabolome bioinformatic analysis was performed using the metabolomics R package metaX developed by BGI (Shenzhen, China). The fold change and Student's t test were used to screen for differential metabolites.

2.7. Transcriptome analysis

Treated zebrafish were sacrificed in ice water, and the articular areas of the anal fin were dissected for mRNA sequencing. Total RNA was extracted using the RNeasy Micro Kit (Cat# 74004, Qiagen). Quantification and qualification were performed using an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, US). The library was constructed following the instructions of the VAHTS Universal V6 RNA-seq Library Prep Kit for Illumina® before being subjected to an Illumina NovaSeq6000 sequencer. The expression abundance is presented as fragments per kilobase million (FPKM). Differential expression analysis was conducted using DESeq2 software [29]. Genes with a parameter of false discovery rate (FDR)≤ 0.05 and absolute fold change≥ 2 were filtered as differentially expressed genes (DEGs).

2.8. GO analysis

DEGs were mapped to the Gene Ontology (GO) database using the Omicshare online software (https://www.omicshare.com/tools/). A hypergeometric test defined significantly enriched GO terms among the DEGs.

2.9. KEGG analysis

DEGs were mapped to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database using Omicshare online software (https://www.omicshare.com/tools/). A hypergeometric test defined significantly enriched KEGG pathways among the DEGs.

2.10. GSEA

All genes with FPKM changes were analysed by Gene Set Enrichment Analysis (GSEA) using Omicshare online software (https://www.omicshare.com/tools/). A hypergeometric test defined significantly enriched gene sets among the DEGs.

2.11. Sirius red staining

Chondrocytes were seeded in 48-well plates at 105 cells/mL. Different concentrations of TCC and its control solvent, DMSO, were added to the cells for 72 h. The treated cells were then fixed with paraformaldehyde and washed with PBS. Picrosirius Red Solution (Abcam, US) was applied for 60 min. The stained cells were then rinsed with an acetic acid solution (0.5 %). The cells were observed and imaged under an inverted microscope (Olympus ix70, Japan). The Sirius red staining of the cells was then quantified according to the gradient densitometry using ImageJ software in a double-blind manner.

2.12. Real-time quantified PCR

Total RNA was extracted using a RNeasy Micro Kit (Cat# 74004, Qiagen). Reverse transcription was conducted by following the standard protocol of the PrimerScript RT kit (Takara, Japan). The qPCR was performed with a thermal cycling condition of 95 °C (denature) for 15 s and 60 °C (anneal and elongation) using a qPCR mix (MCE, New Jersey, US) on a Step One Real-Time system (ABI, USA). The expression of beta actin (ACTB) was used as a housekeeping gene, and the gene expression is presented as the fold change compared to the control. Three biological repeats were performed for each treatment group.

PCR primers:

hCOL2A1-F: 5’- CCTGGCAAAGATGGTGAGACAG.

hCOL2A1-R: 5’-CCTGGTTTTCCACCTTCACCTG.

hCOL9A1-F: 5’- GGCAGTAGAGGAGAATTAGGACC.

hCOL9A1-R: 5’- GTTCACCGACTACACCCCTG.

hCOL9A3-F: 5’- GTGGATGGTCTGACTGGACG.

hCOL9A3-R: 5’- GGGCAGATACTTGGGCACTG.

hCOL5A2-F: 5’- CCTGGCAAAGATGGTGAGACAG.

hCOL5A2-R: 5’-CCTGGTTTTCCACCTTCACCTG.

hDNMT1-F: 5’- CGAGGAAGTAGAAGCGGTTGG.

hDNMT1-R: 5’- AAGGAGCCCGTGGATGAGG.

hATCB-F: 5’-CACCATTGGCAATGAGCGGTTC.

hATCB-R: 5’-AGGTCTTTGCGGATGTCCACGT.

2.13. Methylation-specific PCR

Zebrafish exposure to TCC was executed in ice water, and the articular areas of the anal fin were dissected and treated with bisulfite. The TCC-exposed chondrocytes were also treated with bisulfite. DNA samples from zebrafish articular areas and chondrocytes were then extracted and purified following the instructions of the Methylation Specific PCR Kit (EM101, Tiangen, China). The methylation-specific and unmethylation-specific primers were designed through the website http://www.urogene.org/cgi-bin/methprimer/methprimer.cgi. The methylation-specific PCR was performed under the same conditions as the real-time quantified PCR section.

Methylation-specific primers:

hCOL2A1-MF: 5’- TGGCCAGGGCCGCAGCAGTGACCAACC.

hCOL2A1-MR: 5’- GACTGAAAGAGCCACAGAGGACGGAG.

hCOL2A1-UF: 5’-TTAGGGTCGTAGTAGTGATTAATCGT.

hCOL2A1-UR: 5’-AACTAAAAAAACCACAAAAAACGAAA.

hATCB-F: 5‘-AGCGATGCGTTCGAGCATCGCUTAGGGAGTATATAGGTTGGGGAAGTT.

hATCB-R: 5‘-AACACACAATAACAAACACAAATTCAC.

2.14. Western blotting

Western blotting was performed to measure the protein level of type II collagen. Briefly, the chondrocytes were seeded in 6 cm cell plates and treated with different concentrations of TCC (0.03–3 μM) for 72 h. The treated cells were then harvested using cell scrapers and lysed using RIPA buffer (HY-K1001, MCE, US). The cell lysate was then subjected to noncontact sonication for 10 min followed by centrifugation at 10000XG for 10 min at 4 °C. The supernatant was collected, and BCA was employed to quantify the total protein, followed by boiling with loading buffer (S3014, Merk, US). The same protein samples were loaded for SDSPAGE. After electrophoresis, the proteins were transferred to a PVDF membrane and blocked with 5 % nonfat milk at room temperature. The blocked membrane was then incubated with COL2A1 primary antibody (ab239007, Abcam, US) at 4 °C overnight. The membrane was washed three times with TBST before incubation with HRP-linked anti-rabbit antibody (ab205718, Abcam, US). Another three rounds of TBST washing were performed before developing the protein bands with ECL reagents (WBULS0100, Merck, US). GAPDH (ab9485, Abcam, US) was used as a loading control.

2.15. Statistical analysis

Statistical data analysis was conducted using GraphPad Prism 9.0 on Macintosh OS X, and all data are expressed as the mean ± SD. One-way ANOVA and multiway ANOVA were used to perform a significant analysis of multiple groups, and significant differences were considered when the p value was < 0.05.

3.1. TCC triggers osteoarthritis in the zebrafish anal fin

As the first step, we quantified the OA in the joint fluid from articulatio genus of 15 OA patients and 10 non-OA donors. Results demonstrated that the average TCC concentration in OA patients reaches 2 ng/mL, which is significantly higher than that in non-OA donors (0.15 ng/mL) (Fig. S1). The acute toxicity of TCC was then evaluated in zebrafish larvae. Although the morphology of zebrafish larvae was healthy overall, the decrease in the larval survival rate became significant when the TCC concentration was increased to 10 μM (Fig. S2A-B). The heartbeat rate of zebrafish larvae decreased after exposure to 1 μM and 3 μM TCC for 72 h, although they the zebrafish were still alive at those concentrations (Fig. S2C). The adult zebrafish demonstrated more resistance to TCC, as indicated by the unchanged stander length and body weight after long-term exposure to TCC for three weeks (Fig. S2D-E). However, we found that chronic exposure to TCC resulted in hyperaemia around the joints of the anal fin in adult zebrafish in a time- and dose-dependent manner (Fig. 1 and Fig. S3). These results suggested that TCC may injure the joints in adult zebrafish.

1. Download : Download high-res image (730KB)
2. Download : Download full-size image

Fig. 1. TCC induces hyperaemia in the joint of the zebrafish anal fin. (A) Typical images reveal that TCC induces hyperaemia in the reciprocal of the zebrafish anal fin in a time- and dose-dependent manner. The zebrafish were soaked in 0.03–3 μM TCC for different periods, and the same dose of DMSO (0.025 %, V/V) was employed as a control. The black arrows point to the hyperaemia area, and the number in the frame refers to the number of zebrafish as shown in the image/total number of zebrafish involved in this group. Experiments were performed three times with similar results. (B) Quantification of the hyperaemia in the joint of the zebrafish anal fin according to the virtual ruler in the software of the imaging system. L0 represents the length of the whole anal fin, while Lh represents the length of the hyperaemia area. The hyperaemia rate is presented as 100 % X (Lh/L0); (C) Quantification of the time- and dose-dependent hyperaemia in the joints of zebrafish anal fins induced by TCC. * indicates P < 0.05, * * indicates P < 0.01, * ** indicates P < 0.001, * ** * indicates P < 0.0001. The comparison was performed between the TCC-treated group and the DMSO (0.025 %, V/V) control at the same time point.

To further investigate the impact of TCC on the articular area around the anal fin, Alcian blue and alizarin red double staining was performed, and the results indicated that the bone and cartilage tissue were well stated in zebrafish exposed to TCC for three weeks (Fig. 2 A). Interestingly, we found that the intraarticular space narrowed after exposure to TCC (Fig. 2B). We further quantified the intraarticular space in the anal fin. The healthy zebrafish demonstrated an average intraarticular distance of approximately 16 µm. In contrast, the distance diminished to only one-half after exposure to TCC (Fig. 2 C), suggesting that TCC might cause OA in zebrafish anal fins.

1. Download : Download high-res image (354KB)
2. Download : Download full-size image

Fig. 2. TCC reduces the joint space of the zebrafish anal fin. (A) Scheme of Alcian blue and alizarin red staining of zebrafish treated with TCC. Zebrafish were treated with 3 μM TCC or and equal concentration of DMSO (0.025 %, V/V) as a control for three weeks before staining. (B) Typical images of the zebrafish anal fin with Alcian blue and alizarin red staining reveal that TCC reduces the joint space of the zebrafish anal fin. The intraarticular space was quantified by the virtual ruler in the imaging system, which was precalibrated by a physical ruler. (C) Quantification of the joint space of the zebrafish anal fin treated with 3 μM TCC or an equal concentration of DMSO (0.025 %, V/V). * ** * refers to P < 0.0001 and n = 12.

3.2. TCC programs an osteoarthritis-prone metabolome

To validate whether TCC triggered OA in zebrafish, the articular area of the anal fin was dissected and subjected to untargeted metabolomics and transcriptomics analysis (Fig. S4A). The metabolites in the articular area significantly changed after TCC exposure in both positive and negative ion modes, as indicated by tables (Tables S2 and S3), volcano diagrams (Fig. S4B and C), and heatmaps (Fig. 3A-B). By using principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA), the distribution of metabolites demonstrated good categorization between TCC and its solvent control group in both positive and negative ion modes (Fig. 3C-F). These differential metabolites were further analysed according to the Kyoto Encyclopedia of Genes and Genomes (KEGG). As indicated in Fig. 3 G, several pathways, such as the citrate cycle, fatty acid biosynthesis pathway, and Foxo1 signalling pathway, were significantly enriched in the joints of the anal fin of zebrafish exposed to TCC. These pathways are also significantly altered in clinical OA patients [30], [31], [32]. We further analysed the metabolites that were significantly changed after exposure to TCC. Table S1 demonstrates the intersection of between the increased and decreased metabolites in the joints exposed to TCC and the significantly changed metabolites in the joint fluids of OA patients according to the previous clinical studies. Most of these metabolites are highly consistent between TCC-exposed zebrafish and clinical OA patients. These results confirm that TCC triggers OA at the molecular level.

1. Download : Download high-res image (718KB)
2. Download : Download full-size image

Fig. 3. TCC causes metabolic dysfunction in the joint of the zebrafish anal fin. (A) Heatmap revealing the differential metabolites collected in positive ion mode; (B) Heatmap showing the differential metabolites collected in negative ion mode; (C) Principal component analysis (PCA) of the differential metabolites collected in positive ion mode; (D) Partial least squares-discriminant analysis (PLS-DA) of the differential metabolites collected in positive ion mode; (E) PCA of the differential metabolites collected in negative ion mode; (F) PLS-DA of the differential metabolites collected in negative ion mode; (G) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment for the differential metabolites collected in positive and negative ion modes.

3.3. TCC suppresses chondrogenesis and extracellular matrix gene expression

Transcriptomics analysis of the articular area of the anal fin was also performed, and the differentially expressed genes (DEGs) are presented in a volcano diagram (Fig. S5A) and heatmap (Fig. 4A). We further performed gene set enrichment analysis (GSEA) of the transcriptome results, and the top 5 upregulated and downregulated gene sets are indicated in Fig. S5B. Interestingly, we found that the extracellular matrix structural constituent and the collagen trimer gene sets, which are critical for cartilage tissue, are the two most enriched gene sets in the downregulated list. We then analysed the related gene sets and found that several extracellular matrix (ECM)- and collagen-related gene sets were overall downregulated after TCC exposure, as presented in Fig. 4B.

1. Download : Download high-res image (649KB)
2. Download : Download full-size image

Fig. 4. TCC reprograms the transcriptome of the zebrafish anal fin joint by reducing the extracellular matrix. (A) Heatmap of the differentially expressed genes treated with TCC. (B) Multiple GSEA curves of subsets of genes encoding the extracellular matrix. (C) Distribution of genes within the GO term of collagen trimer according to GSEA analysis; (D) Heatmap revealing the changes in the expression of extracellular matrix- and inflammation-related genes after exposure to TCC.

Collagen is a crucial ECM component of cartilage tissue. Therefore, the collagen trimer set genes are indicated in a lattice diagram based on the GSEA results (Fig. 4C). Almost all collagen coding genes were inhibited by TCC exposure. Type II collagen is the primary ECM in cartilage tissue. Moreover, the reduced cartilage tissue-related ECM genes, including col2a1a, and elevated inflammation-related genes are indicated in Fig. 4D. The col2a1a gene, the major gene that codes Type II collagen, is highlighted in the lattice diagram, and it also demonstrated a marked decline after TCC exposure. The progesterone−mediated oocyte maturation-related genes were enriched among the upregulated pathways. We also found that TCC might bind to the progesterone receptor (PGR) at its ligand-binding domain (Fig. S6). These results suggested that TCC triggers a decrease in cartilage tissue at the transcriptional level.

Primary human chondrocytes for six non-OA donors were introduced to investigate whether TCC suppresses the expression of the type II collagen coding gene conserved in humans. The primary human chondrocytes were characterized by immunofluorescence for type II collagen (Fig. S7). As shown in Fig. 5 A and B, TCC exposure caused a dose-dependent collagen decline, as indicated by the gradually weakened Sirius red staining. Expression of the human COL2A gene demonstrated a dose-dependent decrease at both the mRNA and protein levels after exposure to TCC (Fig. 5 C and D). These results suggested that the collagen formation inhibitory effect of TCC is conserved between zebrafish and human beings.

1. Download : Download high-res image (683KB)
2. Download : Download full-size image

Fig. 5. TCC suppresses the expression of the type II collagen coding gene via DNMT1. (A) Different concentrations of TCC reduced type II collagen in human chondrocytes, as indicated by Sirius red staining. The experiments were performed with 3 biological replicates. An equal concentration of DMSO (0.025 %, V/V) was used as control. Similar results were obtained in human chondrocytes from 6 independent donors (3 female donors and 3 male donors). (B) Quantification of Sirius red staining in human chondrocytes exposed to different concentrations of TCC or an equal concentration of DMSO (0.025 %, V/V). Columns with the same letter indicate P > 0.05, while columns with different letters indicate P < 0.05. (C) TCC reduces COL2A1 expression in human chondrocytes. qPCR was performed with 3 biological replicates and 2 technical replicates. Similar results were obtained in human chondrocytes from 6 independent donors. An equal concentration of DMSO (0.025 %, V/V) was used as control. Columns with the same letter indicate P > 0.05, while columns with different letters indicate P < 0.05. (D) TCC reduces type II collagen in human chondrocytes. Western blotting was performed with 3 biological replicates, and similar results were obtained in human chondrocytes from 6 independent donors. GAPDH was employed as a loading control. (E) TCC stimulates DNMT1 expression in human chondrocytes. qPCR was performed with 3 biological replicates and 2 technical replicates. Similar results were obtained in human chondrocytes from 6 independent donors. Columns with the same letter indicate P > 0.05, while columns with different letters indicate P < 0.05; (F) GO enrichment of the differentially expressed genes in the zebrafish anal fin joint treated with TCC; (G) Distribution of genes within the GO term of negative regulation of gene expression according to GSEA; (H) Distribution of genes within the GO term of DNA metabolic processes according to GSEA.

3.4. TCC stimulates DNMT1 expression and methylation of the type II collagen coding gene

To further investigate the potential upstream regulator of type II collagen coding gene expression in response to TCC, we predicted the potential transcription factors of COL2A1; however, none of the possible transcription factors changed after TCC exposure (Fig. S3B). We then reanalyzed the transcriptome results and found that the DNA methylation genes had significantly different gene ontology (GO) analysis results (Fig. 5 F). Moreover, the GSEA lattice diagram also demonstrated that DNA metabolism and negative regulation of gene expression subsets were significantly elevated after TCC exposure. Interestingly, DNA methyltransferase 1 (DNMT1), highlighted in Fig. 5 G and H, belongs to both gene subsets. We also quantified the expression in human chondrocytes, and the results demonstrated that the DNMT1 transcript gradually increased when the TCC concentration increased (Fig. 5E), indicating that DNMT1 expression stimulated by TCC is conserved between zebrafish and human beings.

Usually, DNMT1 maintains DNA methylation and thus suppresses the expression of methylated genes [30]. We then scanned human COL2A1 (Fig. 6 A) and zebrafish col2a1a (Fig. 6B) genes. There are GC-rich sequences in both of their 5’ UTR regions. Bisulfite-based qPCR was employed to evaluate the methylation of the GC-rich sequence of human COL2A1 and zebrafish col2a1a genes. As indicated in Fig. 6 C, the solvent control demonstrated a higher unmethylated level in the GC-rich sequence of the COL2A1 gene than that in the TCC-exposed chondrocytes using unmethylated primers (UP). When using methylated primers (MPs), TCC-exposed chondrocytes demonstrated higher methylation levels in the GC-rich sequence of the COL2A1 gene than solvent control cells, indicating that TCC stimulated DNA methylation in the GC-rich sequence of the COL2A1 gene in human chondrocytes. The methylation of the zebrafish col2a1a gene was also quantified in the articular tissue of zebrafish anal fins, and the results were similar to those of human chondrocytes (Fig. 6D). These results collectively demonstrated that the methylation of the GC-rich sequence in COL2A1/col2a1a genes stimulated by TCC is conserved between zebrafish and human beings.

1. Download : Download high-res image (596KB)
2. Download : Download full-size image

Fig. 6. TCC reduces type II collagen expression via DNMT1-dependent DNA methylation. (A) The CpG island in the human COL2A1 gene. The arrows refer to the primer pairs. (B) The CpG island in the zebrafish col2a1a gene. The arrows refer to the primer pairs; (C) Quantification of CpG island methylation in the human COL2A1 gene treated with TCC or an equal concentration of DMSO (0.025 %, V/V).; (D) Quantification of CpG island methylation in the zebrafish col2a1a gene treated with TCC or an equal concentration of DMSO (0.025 %, V/V).; (E) Quantification of CpG island methylation in the human COL2A1 gene exposed to TCC with or without the presence of DC_517 or CM-272. An equal concentration of DMSO (0.025 %, V/V) was used as control; (F) Quantification of COL2A1 expression in human chondrocytes exposed to TCC with or without the presence of DC_517 or CM-272. An equal concentration of DMSO (0.025 %, V/V) was used as control. UP refers to unmethylation primers; MP refers to methylation primers. qPCR was performed with 3 biological replicates and 2 technical replicates. Similar results were obtained in human chondrocytes from 6 independent donors. * indicates P < 0.05, * ** indicates P < 0.001, and * ** * indicates P < 0.0001.

3.5. TCC suppresses COL2A1 via DNMT1-dependent DNA methylation

To confirm the role of DNMT1 in the suppression of type II collagen expression by TCC, two DNMT1 inhibitors, DC_517 and CM-272, were introduced to pretreat human chondrocytes before exposure to TCC. As shown in Fig. 6E, the methylation of the COL2A1 gene was also quantified, and the bisulfite-based qPCR results demonstrated that both DC_517 and CM-272 abolished the methylation of COL2A1 induced by TCC. Moreover, TCC significantly suppressed COL2A1 expression in a dose-dependent manner, while the inhibitory effect of TCC became insignificant when the chondrocytes were pretreated with either DC_517 or CM-272 (Fig. 6 F). Quantification was also performed for COL9A1, COL9A3 and COL5A2, which encode other core proteins of type II collagen. Fig. S8 demonstrated that TCC decreased the expression of COL9A1, COL9A3 and COL5A2, which is consistent with the expression of zebrafish col9a1a, col9a3 and col5a2b (Fig. 4D). Two DNMT1 inhibitors, DC-517 and CMC-272, significantly reversed the decreased expression COL9A1, COL9A3 and COL5A2 induced by TCC, indicating the TCC supressed the expression of COL9A1, COL9A3 and COL5A2 in a DNMT1-dependent manner. Sirius red staining of human chondrocytes demonstrated that inhibition of DNMT1 by either DC_517 or CM-272 abolished the suppressive effect of TCC on collagen formation (Fig. 6 G). These results indicated that TCC suppresses COL2A1 via DNMT1-dependent DNA methylation.