Fig 2: Intestinal epithelial cell-derived or exogenous CS is protective against 2.5% DSS-induced colitis.a RT-qPCR analysis of Sult2b1 mRNA in different organs from Sult2b1f/f and Sult2b1?IEC mice (n = 4 mice/group). b Quantification of CS in serum, cecal contents, and colonic tissues from Sult2b1f/f and Sult2b1?IEC mice using LC–MS (n = 6 mice/group, two-tailed Student’s t test). c Schematic diagram of treatment and body weight reduction rate of Sult2b1f/f and Sult2b1?IEC mice during DSS challenge (n = 5 mice/group), two-way ANOVA with Tukey’s multiple comparisons test, day 3 *P = 0.0139; day 4 ***P < 0.0001; day 5 ***P < 0.0001; day 6 ***P < 0.0001. d, e Faecal occult blood (FOB) of Sult2b1f/f and Sult2b1?IEC mice on day 3, and colon length/body weight of these mice on day 6 of DSS challenge (n = 5 mice/group, two-tailed Student’s t test). f Schematic diagram of treatment and body weight reduction rate of Sult2b1f/f, Sult2b1?IEC, and CS-fed Sult2b1?IEC mice during DSS challenge (n = 5, 5, 4 mice/group, respectively, two-way ANOVA with Tukey’s multiple comparisons test, day 4 (KO vs. KO + CS) ***P = 0.0003; day 5 (KO vs. KO + CS) ***P < 0.0001; day 6 (KO vs. KO + CS) ***P < 0.0001. g, h FOB and colon length/body weight of Sult2b1f/f, Sult2b1?IEC, and CS-fed Sult2b1?IEC mice during DSS challenge (n = 5, 5, 4 mice/group, respectively, one-way ANOVA with Sidak’s multiple comparison test). i, j: Hematoxylin and eosin- and EdU-stained colonic sections excised from the same parts (2 cm away from the anus) of normal Sult2b1f/f, Sult2b1?IEC, 6-day DSS-challenged Sult2b1f/f, Sult2b1?IEC, and CS-fed Sult2b1?IEC mice (n = 5, 5, 5, 5, 4 mice/group, respectively, one-way ANOVA with Sidak’s multiple comparison test). Data are shown as the mean ± SEM. Source data are provided as a Source Data file.

Fig 3: Promotion of cholesterol biosynthesis by CS was dependent on SREBP2.a Immunofluorescent staining of SREBP2 in HT-29 and SULT2B1-KO cells in the presence or absence of CS (50 µM), where nuclei were counterstained with DAPI (n = 4 fields/group, one-way ANOVA with Sidak’s multiple comparison test). b Western blotting analysis of mature/precursor SREBP2 in 293T cells transiently transfected pCDH-CMV-HA-SREBF2 /pCDNA3-SCAP and treated with DMSO, 25 µM CS, or 50 µM CS for 6 h, 12 h or 18 h. The ratio of mature/precursor SREBP-2 was summarized. (n = 3 biological replicates, one-way ANOVA with Sidak’s multiple comparison test). c Western blotting analysis of SREBP2 in SULT2B1-KO + sh-CHERRY and SULT2B1-KO + sh-SREBF2 cells. d RT-qPCR analysis of SREBF2 in SULT2B1-KO + sh-CHERRY and SULT2B1-KO + sh-SREBF2 HT-29 cells (n = 2 independent culture wells, two-tailed Student’s t-test). e RT-qPCR analysis of cholesterol synthesis genes in SULT2B1-KO + shCherry and SULT2B1-KO + shSREBF2 cells treated with DMSO or CS (50 µM). The values were expressed as CS/DMSO (n = 2 independent culture wells, two-tailed Student’s t-test). f Filipin III staining in SULT2B1-KO cells, SULT2B1-KO + shSREBF2 cells, and SULT2B1-KO + shSREBF2 + CS (50 µM) cells (n = 4 fields/group, one-way ANOVA with Sidak’s multiple comparison test). g Cell viability of SULT2B1-KO and SULT2B1-KO + shSREBF2 cells treated with DMSO (served as a control) or 25 µM CS (n = 5 biologically independent samples, one-way ANOVA with Sidak’s multiple comparison test). Data are reprehensive of at least three independent experiments. Data are shown as the mean ± SEM. Source data are provided as a Source Data file.

Fig 4: SULT2B1 and cholesterol sulfate (CS) levels are increased in ulcerative colitis (UC) mucosa.a Normalized FPKM (fragments per kilobase of transcript per million mapped reads) of SULT2B1 in inflamed rectum mucosa of 26 patients with UC and 22 healthy controls based on RNA-seq data in GSE 111889. (Two-tailed Student’s t test). b SULT2B1 protein expression in the uninflamed and inflamed colonic mucosa from the same patients with UC using immunohistochemistry (n = 9, two-tailed, Wilcoxon matched-pairs signed rank test). c Quantification of CS in colonic tissues extracted by colonoscopy from healthy controls (n = 19) and patients with UC (n = 27) using LC–MS. CS concentration was calculated from the linear regression equation plotted from the standard curve and normalized with wet weight of colonic tissues. (Two-tailed Student’s t test). d Western blotting analyses were performed for SULT2B1 protein in the colon of mice challenged with the following: 3-day 2.5% DSS; 3-day 2.5% DSS + 3-day H2O; 3-day 2.5% DSS + 6-day H2O; 3-day 2.5% DSS + 9-day H2O; (Con: control mice; D3: 3-day 2.5% DSS; D3 + H3: 3-day 2.5% DSS + 3-day H2O; D3 + H6: 3-day 2.5% DSS + 6-day H2O; D3 + H9: 3-day 2.5% DSS + 9-day H2O) (n = 3, 2, 2, 2, 3 mice/group, respectively, one-way ANOVA with Sidak’s multiple comparison test). e Western blotting analyses were performed for SULT2B1 protein in the colon of mice challenged with the following: 5-day 0.5% DSS; 10-day 0.5% DSS; 15-day 0.5% DSS; (D5: 5-day 0.5% DSS; D10: 10-day 0.5% DSS; D15: 15-day 0.5% DSS) (n = 3 mice/group, one-way ANOVA with Sidak’s multiple comparison test). f RT-qPCR analysis of Sult2b1 mRNA in distal, media, and proximal colonic tissues with 2-, 4-, and 6-day 2.5% DSS challenge, respectively. (n = 6, 3, 3, 3 mice/group, respectively, one-way ANOVA with Sidak’s multiple comparison test). g Quantification of CS in cecal contents and colonic tissues from mice with 6-day 2.5% DSS challenge and control group using LC–MS. CS concentration was calculated from the linear regression equation plotted from the standard curve and normalized with wet weight of cecal contents or colonic tissues. (n = 6 mice/group, two-tailed Student’s t test). h Quantification of CS in colonic tissues from mice with 2-, 4-, and 6-day 2.5% DSS challenge using DESI-MSI. Cholesterol sulfate ion is at m/z 465.3044, the signal variation was normalized by the total ion current (TIC, range from m/z 400 to 500) pixel-by-pixel. The approximate lateral resolution is 200 µm (n = 5 mice/group, 2 fields/mouse, one-way ANOVA with Sidak’s multiple comparison test). Data from in vitro assays are reprehensive of at least three independent experiments. Data are shown as the mean ± SEM. Source data are provided as a Source Data file.

Fig 5: CS is required for cholesterol biosynthesis in colonic epithelial cells.a The heatmap shows the relative levels of cholesterol metabolism-related genes in inflamed colonic tissues from 6-day 2.5% DSS-treated Sult2b1f/f and Sult2b1?IEC mice. (Fold change = 2, P < 0.05) (n = 4 mice/group). b Gene Set Enrichment Analysis shows the enrichment of the cholesterol synthesis gene signature in the downregulated genes (Sult2b1?IEC vs. Sult2b1f/f mice). c Quantification of free cholesterol concentration in colonic tissues from Sult2b1f/f and Sult2b1?IEC mice in the presence or absence of 6-day 2.5% DSS challenge using the Amplex® Red Cholesterol Assay Kit (n = 5 mice/group, one-way ANOVA with Sidak’s multiple comparison test). d Filipin III staining in HT-29 and SULT2B1-KO cells. The boxed areas are shown in the lower panels. (n = 3 fields/group, two-tailed Student’s t test). e CCK-8 assay shows the proliferation of HT-29 and SULT2B1-KO HT-29 cells in the presence or absence of cholesterol (50 µM) dissolved in ß-cyclodextrin (CHOL: cholesterol; ß-CYC: ß-cyclodextrin). Two-way ANOVA with Tukey’s multiple comparisons test, 48 h (HT-29+ß-CYC vs. SULT2B1-KO + ß-CYC, *P = 0.0209; SULT2B1-KO + ß-CYC vs. SULT2B1-KO + CHOL, *P = 0.0127), 72 h (HT-29+ß-CYC vs. SULT2B1-KO + ß-CYC, **P = 0.0012; SULT2B1-KO + ß-CYC vs. SULT2B1-KO + CHOL, *P = 0.0122). f Normalized FPKM of the cholesterol biosynthesis genes in inflamed rectum tissues of 26 patients with UC and 22 healthy controls based on RNA-seq data in GSE 111889. (Two-tailed Student’s t test). g Quantification of total and free cholesterol in colonic tissues extracted by colonoscopy from healthy controls (n = 19) and patients with UC (n = 27) using the Amplex® Red Cholesterol Assay Kit. (Two-tailed Student’s t-test). h Pearson’s correlation analysis was performed between CS and total cholesterol levels, or CS and free cholesterol levels in colonic tissues from 27 patients with UC. (Two-tailed P value). Data from in vitro assays are reprehensive of at least three independent experiments. Data are shown as the mean ± SEM. Source data are provided as a Source Data file.