A total of 24 six-week-old male BALB/c mice were purchased from Shanghai Jiesijie Experimental Animal Co., Ltd. Mice were raised with a constant temperature of 23±2̊C, 50% humidity, a 12/12 h light/dark cycle and ad libitum access to food and water. The animal protocol in this work was in accordance with guidelines for the care and use of laboratory animals authorized by the Medical Ethics Committee of Minhang Hospital, Shanghai, China. The animal research was approved by the Ethics Committee of Minhang Hospital, Shanghai, China [Medical Ethics Committee (2018) Approval no. 2]. A total of 24 mice were randomly divided into four groups (n=6 per group). Three groups of mice were used as experimental colitis models and one group of mice served as normal controls. An experimental colitis model was established by supplementing BALB/c mice with 3% (w/v) DSS (Sigma-Aldrich; Merck KGaA) in drinking water once a day for 7 days. On the 8th day, the decitabine, sulfasalazine (SASP) positive control and model groups were intraperitoneally administered with decitabine (Merck KGaA; cat. no. 2353-33-5; 0.5 mg/kg), SASP (Merck KGaA; cat. no. 599-79-1; 100 mg/kg) and 1% DMSO (1 μl/20 g), respectively, for 7 consecutive days. Mice were injected twice a day at an interval of 12 h. During the experiment, no mice died unexpectedly. From the day of treating with DSS changes in body weight, external characteristics and clinical symptoms of the mice were monitored every day. The combination of these indicators can reflect the pain and suffering experienced by the animals during the experiment (12). Mice were euthanized with an intraperitoneal injection of 180 mg/kg sodium pentobarbital, and then the heartbeat of the mice was examined. The duration of animal experiments was 2 weeks from the first day of treatment with DSS to euthanasia.

Mice of each group were examined daily for body mass, diarrhea and blood in the stool. Disease activity index (DAI) was calculated as follows: DAI = (weight loss score + fecal trait score + fecal occult blood score)/3) (13).

After the animals were sacrificed on the 8th day, the entire colonic intestine from the anus to the end of the cecum was isolated. Following washing, the extent of inflammation and ulceration was assessed and the length of the colon was recorded. Colon gross morphological damage index (CMDI) score was calculated as reported previously (1).

Distal colon tissue (1 cm) was fixed in 4% paraformaldehyde for 24 h at room temperature, embedded in paraffin, sectioned to a thickness of 8-μm, and stained with H&E for 1 min at room temperature. Histopathological changes were observed under a light microscope (magnification, ×20 or ×40) and scored according to the criteria described by Andújar et al (14). The colon histopathology index (CHPI) was determined as previously described (15).

Since the effect of DSS is more pronounced at the distal end compared with the proximal colon (16), small sections (~1 cm) of excised distal colonic tissue were collected for ELISA and western blot assays. Subsequently, ~10 mg colon tissue of each group was collected and the tissue was homogenized with 1 ml PBS (pH, 6.0; containing 1 μg aprotinin (Shanghai Qcbio Science & Technologies Co. Ltd.; cat. no. 20105ES08) and 1 μg leupeptin pepstatin A (Maokangbio Co. Ltd.; cat. no. 103476-89-7). After homogenate liquid was centrifuged at 16,099 × g for 20 min at 4°C, 500 μl supernatant was collected for ELISA. ELISA kits for IL-17 (Boster Biological Technology; cat. no. EK0431), TGF-β (Beijing Solarbio Science & Technology Co., Ltd.; cat. no. SEKM-0035) and IL-10 (Boster Biological Technology; cat. no. EK0417) were used to measure the cytokine levels, according to the manufacturer's instructions.

The spleens of the mice were cut, filtered through a 200-μm nylon mesh, centrifuged at 251.55 × g for 5 min at 4°C, and then the supernatant was removed. NH₄Cl (2 ml; 0.83%) was added to lyse the red blood cells. After 2 min, 15 ml PBS was added to terminate the lysis. The cell density was adjusted to 1×10⁶ cells/ml in PBS and the cells were blocked with 2.5% BSA at room temperature for 1 h. Subsequently, the cells were incubated with FITC-rat anti-mouse CD4 (cat. no RM4-5; eBioscience; Thermo Fisher Scientific, Inc.) in 2.5% BSA for 1 h at room temperature in the dark and washed twice with PBS. Cells were then fixed with fixation reagent (cat. no. 00-5523-00; eBioscience; Thermo Fisher Scientific, Inc.) for 15 min at room temperature and permeabilized for intracellular staining. Subsequently, 1 ml of a 3:1 permeabilization reagent (cat. no. 00-5523-00; eBioscience; Thermo Fisher Scientific, Inc.) was added, and the cells were incubated at 4°C for 1 h in the dark, followed by the addition of cell permeabilization buffer (2 ml; cat. no. 39487s; Cell Signaling Technology, Inc.), centrifugation twice, and the resuspension of cells to a volume of 100 μl. PE-rat anti-mouse Foxp3 (cat. no. 72-5775-40; eBioscience; Thermo Fisher Scientific, Inc.) in 2.5% BSA was added and incubated for 1 h at 4°C in the dark. Cell phenotypes were detected using a flow cytometer and the data were analyzed using FlowJo software (FlowJo LLC; version 7.6.1).

Sections were blocked in 5% BSA for 1 h at room temperature and were stained with anti-mouse antibodies against zonular occludens-1 (1:1,000; ZO-1; Novus Biologicals, Ltd.; cat. no. NBP1-85047) and occludin (1:200; Abcam; cat. no. ab216327) in 2.5% BSA for overnight at 4°C, and then with the secondary antibody goat anti-rabbit IgG (H+L) with a FITC fluorescent label (1:1,000; GeneCopoeia; cat. no. L147A) or Cy-3 tag (1:1,000; Boster Biological Technology; cat. no. BA1032) in 2.5% BSA for 2 h at room temperature. Sections were examined under a fluorescence microscope (magnification, ×20). Quantitative analysis was conducted using Image J software (National Institutes of Health; version 1.46).

ZO-1 and occludin anti-mouse antibodies were diluted with PBS at a ratio of 1:100; and PBS was used as a negative control for overnight at 4°C. Subsequently, eight fields of view were randomly observed under a fluorescence microscope (magnification, ×40), and the percentage of positive cells and the intensity of staining was scored. The percentage of stained cells in the field of view was scored as follows: <5%, 0 point; 5-25%, 1 point; 26-50%, 2 points; 51-75%, 3 points; and >75%, 4 points. The staining intensity was scored as follows: No staining, 0 point; light yellow staining, 1 point; brown staining, 2 points; and tan staining, 3 points. The two scores were then multiplied to calculate the immunoreactivity score.

For western blotting, ~100 mg distal colonic tissue was weighed and cut into pieces. The tissue was then grinded with liquid nitrogen and added to 1 ml ice-cold PBS in a round bottom flask. The tissue was homogenized at a speed of 90.56 × g and transferred to a 1.5 ml EP tube. The homogenate liquid was washed twice with PBS containing PMSF (Beijing Solarbio Science & Technology, Co., Ltd.) and then centrifuged at 16,099 × g for 5 min at 4°C. The supernatant was removed and 500 μl protein lysis buffer (Beyotime Institute of Biotechnology; cat. no. P0013) was added. After the tissue was lysed on ice for 2 h, it was centrifuged at 16,099 × g for 30 min at 4°C. The supernatant was collected and mixed with 5X loading buffer (Beyotime Institute of Biotechnology) and then the protein was denatured at 100̊C for 10 min. Protein concentration was detected using a BCA kit (Beyotime Institute of Biotechnology) and 30 μg protein was separated on a 10% SDS-PAGE gel before being transferred to PVDF membranes (Thermo Fisher Scientific, Inc.). Membranes were blocked with 4% BSA-TBST for 90 min and probed overnight at 4°C with primary antibodies in 5% BSA against the following ZO-1 (1:2,500; Novus Biologicals, Ltd.; cat. no. NBP1-85047), occludin (1:1,000; Abcam; cat. no ab216327), ERK1/2 (1:1,000; Abways Technology; cat. no. CY1066), phosphorylated (p)-Erk1/2 (1:1,000; Abways Technology; cat. no. CY6190), JNK (1:1,000; Abways Technology; cat. no. AB3296), p-JNK (1:1,000; Abways Technology; cat. no. CY6315), p38 (1:1,000; Abways Technology; cat. no. AB3374) and p-p38 (1:1,000; Abways Technology; cat. no. CY6390). GAPDH (1:1,000; Santa Cruz Biotechnology, Inc.; cat. no. sc-166574) was used as a loading control. Then, the membranes were incubated with an anti-rabbit HRP secondary antibody (1:1,000; Jackson ImmunoResearch Laboratories, Inc.; cat. no. 111-005-003) for 2 h at room temperature and bands were visualized using super ECL detection reagent (Shanghai Yeasen Biotechnology, Co., Ltd.; cat. no. 36208ES60). Quantitative analysis was conducted using Image J software (National Institutes of Health; version 1.46), and the ratio of phosphorylated protein/total protein was evaluated.

Statistical analysis was performed using GraphPad Prism 5.0 software (GraphPad Software, Inc.). The data are presented as the mean ± standard deviation. Significant differences for between multiple groups were analyzed by one-way ANOVA followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference. Each experiment was repeated a minimum of three times.

Decitabine, also known as 5-aza-2′-deoxycytidine, is a natural adenosine analog of 2′-deoxycytidine (17), and its structure is presented in Fig. 1A. Apart from the normal control group, all other BALB/c mice were given 3% DSS for 7 consecutive days to establish an experimental colitis model. On the 8th day, decitabine (0.5 mg/kg), SASP (100 mg/kg) or placebo (1% DMSO) were intraperitoneally injected for a further 7 consecutive days. The recommended clinical dose of decitabine is 15 mg/m² (18). The Meeh-Rubner formula (19) calculates the body surface area of mice. The body surface area of 20 g mice is ~0.67 dm² (20). Thus, the mice were intraperitoneally injected with decitabine at a dose of 5 mg/kg, which was only used in the preliminary experiments. A total of one week after the first treatment with decitabine, two mice in the decitabine group died; therefore, the administration concentration was reduced by ten times to 0.5 mg/kg (Fig. 1B).

The DAI was used to evaluate the therapeutic effect of decitabine on DSS-induced IBD. No mice died during the modeling process. After 24 h of modeling, which was after the animals were treated with DSS, the animals developed symptoms such as loose stools, severe perianal contamination, reduced food intake and weight loss. During the whole administration period, the DAI of the normal control group was 0. Following the last day of DSS administration, the mice in the model group slightly recovered over the following three days but then their condition began to decline on the fourth day; they ate and drank less, the body weight declined and they had loose stools with mucus and pus. In the SASP group, the general condition of the mice gradually improved, and the DAI score gradually decreased. After 7 days of treatment with SASP, the symptoms of IBD were relieved and the mean DAI score was only 0.73 points. On day 4 of treatment with decitabine, food intake increased and the extent of loose stools was reduced. Compared with the model group, the body weight was increased in the decitabine group. After 7 days of treatment with decitabine, the mean average DAI score reached 0.8. (Fig. 1C).

From the results of colon length measurement, the colon length of the model group was shorter than that of the normal control group. The shortening of the colon length in mice was relieved in the SASP and decitabine groups. The general morphology of the colon and the CMDI score demonstrated that the normal control group had a smooth colon surface and a clear texture. In the DSS alone group, the surface of the intestine was uneven, and the phenomenon of inflammatory congestion and edema. In the model group, severe inflammatory hyperemia and edema were observed in the colon and the CMDI score was significantly higher compared the normal control group (P<0.01). Local hyperemia and edema were observed in the decitabine intervention group, but mainly concentrated in the distal colon, and the CMDI score was significantly lower compared with in the model group (P<0.05; Fig. 1D).

Histopathological observation and the CHPI score demonstrated that there was no obvious inflammatory cell infiltration in the normal control group, and the intact mucosal epithelium was well arranged. In the model group, the colonic mucosa was exfoliated, the gland structure was severely damaged, the gland was disordered and a large number of lymphocytes and neutrophils were infiltrated, and the CHPI score was significantly higher compared with the normal group (P<0.01). The infiltration of inflammatory cells, hyperplasia and edema in was alleviated in the SASP group, and the CHPI score was significantly reduced compared with the model group (P<0.001). In the decitabine group, the intestinal mucosa of the mice was intact, inflammatory cell infiltration was low and the CHPI score was significantly decreased compared with the model group (P<0.001; Fig. 1E).

Tregs are important for the homeostasis of the immune system (21). CD4⁺ CD25⁺ Tregs are a subset of Tregs that maintain immune tolerance through direct contact with effector T cells and secretion of TGF-β, IL-10 and other cytokines (22). Compared with the normal control group, the level of cytokine IL-17 in the colon tissue of the model group was significantly increased, while the levels of cytokines IL-10 and TGF-β were significantly decreased (P<0.001). Compared with the model group, the level of IL-17 in the SASP group was significantly lower (P<0.05) and the IL-10 level was significantly increased (P<0.01). Compared with the model group, the TGF-β levels were significantly increased (P<0.001). There was a significant increase of IL-10 in the decitabine group compared with the model group (P<0.01; Fig. 2A).

In the present study, Foxp3 expression in the spleen of the model group was significantly decreased compared with the normal control group (P<0.01), while treatments with SASP or decitabine resulted in significant increases in Foxp3 expression (P<0.01 and P<0.001, respectively; Fig. 2B).

DSS can promote edema and inflammation, and significantly increases the weight of mice spleens (P<0.001) (23). SASP reduced the weight of the mice spleens compared with the model group (P<0.001). In addition. decitabine reduced the weight of the mice spleen, but it was not statistically significant compared with the model group (Fig. 3A).

Compared with the normal control group, DSS significantly reduced the proportion of CD4⁺ Foxp3⁺ T cells in CD4⁺ T cells in the spleen of mice (P<0.01). Compared with the model group, the proportion of CD4⁺ Foxp3⁺ T cells in CD4⁺ T cells was significantly increased following 7 days of intraperitoneal injection of decitabine (P<0.001; Fig. 3B). This indicates that intraperitoneal injection of decitabine can promote the re-expression of Foxp3 in naive T cells of mice spleen, and increase the proportion of CD4⁺ Foxp3⁺ T cells in CD4⁺ T cells, suggesting that the use of methyltransferase inhibitors in vivo can play a role in demethylation and amplify Tregs.

After the mice were sacrificed, the fluorescence staining of ZO-1 and occludin in the colonic mucosa of the normal control group was continuously distributed at the junction of the colon cells, and the edges were smooth and the fluorescence scope was wide. The mice in the model group demonstrated both a decreased fluorescence intensity and discontinuity distribution of tight junction proteins in the cell apical membrane. Compared with the model group, the tight junction proteins in SASP group and decitabine group were continuously distributed in the apical membrane of intestinal epithelial cells, the edges were smooth, and the fluorescence area and intensity were increased (Fig. 4A and B). The immunohistochemistry demonstrated that the intestinal epithelial structure of the normal control group was intact, while in the model group the tight junction structure and microvilli were destroyed; the gap between cells was widened and the cells were vacuolated. As observed by immunofluorescence, the expression of ZO-1 was significantly lower in the model group compared with the normal control group, and significantly higher in the SASP (P<0.001) and decitabine groups (P<0.001) compared with the model group (Fig. 4A). As observed by immunofluorescence, the expression of Occludin was significantly lower in the model group compared with the normal control group, and significantly higher in the SASP (P<0.001) and decitabine groups (P<0.01) compared with the model group (Fig. 4B). However, differences in occludin expression among all groups were not statistically significant (Fig. 5A). The levels of ZO-1 and occludin in colon tissue of each group were detected by western blotting. The results demonstrated that the levels of ZO-1 and occludin in the model group were significantly lower compared with those in the normal control group (P<0.001), indicating impaired colon barrier function. Compared with the model group, the expression levels of ZO-1 and occludin were significantly higher in the SASP and decitabine groups (P<0.001; Fig. 5B).

The expression levels of ERK1/2, JNK and p38 are presented in Fig. 6A. MAPK signaling is an important signaling system that utilizes a step-by-step phosphorylation process to amplify signals into the nucleus, regulate the activity of transcription factors and the expression of corresponding genes, and then cause cellular responses (24). It was determined that decitabine not only affected the expression of ERK1/2, JNK and P38 in the MAPK pathway, but also p-ERK1/2, p-JNK and p-p38. The phosphorylation of ERK1/2, JNK and p38 in each group is presented in Fig. 6B. Compared with the normal control group, the phosphorylation of ERK1/2, JNK and p38 protein in the model group were significantly increased (P<0.001), and significantly decreased after SASP (P<0.05) and decitabine (P<0.001) intervention. Phosphorylation of ERK1/2, JNK and p38 proteins represents the activation state of MAPK (25), and decitabine inhibits phosphorylation of ERK1/2, JNK and p38 in colon tissues, indicating that it can exert anti-inflammatory effects by inhibiting the activation of the MAPK signaling.