Fig 2: IOX1 regulates Il17a expression through direct association with TET2 on the Il17a promoter. (a) DNA methylation status was assayed using bisulfite sequencing analysis of named CpG sites (#1–12) on the proximal promoter of Il17a in murine total CD4 polarized Th17 cells. A total of 36 clones from four independent experiments were sequenced. The methylation status of CpG site #7 and #8 were statistically different between DMSO and IOX1 treated Th17 cells (Fisher exact test, *P < 0.05). (b) Summary of the frequencies of demethylated CpG sites on the Il17a promoter. (c) Binding of TET2 on the CpG-5 and CpG-6-8 sites of Il17a promoter in response to IOX1 treatment, assayed by ChIP. (N = 3) (d) Binding of IOX1 to the CpG-5 and CpG-6-8 sites of Il17a promoter assayed by Chem-IP. (N = 2) (e) The chemical structure of biotinylated IOX1. (f) Representative FACS analysis of IL-17 expression in response to IOX1 (20 µM) in WT (CD4Cre) and TET2 deficient (CD4CreTet2f/f) Th17 cells. (g) Summary of IOX1 induced IL-17 suppression in WT (CD4Cre) and TET2 deficient (CD4CreTet2f/f) Th17 cells (N = 3). (h) The fold suppression of IL-17+ cell frequencies induced by IOX1 in WT (CD4Cre) and TET2 deficient (CD4CreTet2f/f) Th17 cells (N = 3). (*P < 0.05, **P < 0.01, Mann–Whitney U test). (i) Summary of IOX1 induced suppression of CCL20 expression (in culture supernatants, measured by ELISA) in WT (CD4Cre) (N = 4) and TET2 deficient (CD4CreTet2f/f) (N = 3) Th17 cells. (j) The fold suppression of CCL20 expression induced by IOX1 in WT (CD4Cre) (N = 4) and TET2 deficient (CD4CreTet2f/f) (N = 3) Th17 cells. (*P < 0.05, Mann–Whitney U test).

Fig 3: a-ketoglutarate upregulated the expression of TET2/3 and IFN?-induced STAT1/3 and PD-L1 in melanoma cells.A, B Cell lysates were extracted from B16F10 A and Sk-Mel-28 B cells, and immunoblotting was then performed. C, D Extraction of mRNA and protein from treated melanoma cells, and qRT–PCR C and immunoblotting D were then performed. E, F Extraction of mRNA and protein from treated B16F10 tumor-bearing mouse tissues, and qRT-PCR E and immunoblotting F were then performed. G Detection of the expression of p-STAT1/3 and PD-L1 in B16F10 tumor-bearing mouse tissues by TSA-fluorescent multiplex immunofluorescence staining after the indicated treatments. Multiple experimental data (n = 3) are counted and presented according to the statistical methods, and an asterisk (*) indicates the degree of significant difference.

Fig 4: TET2/3-STAT1/3-CD274 signaling positively correlates with anti-PD1 treatment responsiveness in melanoma patients.A, B Data from the GEO database (GSE 91061) were used to analyze the TET2/3 A and p-STAT1/3 B expression in PD-1 blockade-treated melanoma patients. C, D Data from the GEO database (GSE 91061) were used to analyze the relevance of TETs and STAT-CD274 signaling. The results are expressed as the Spearman rank correlation coefficient. E, F Data from the GEO database (GSE 91061) were analyzed by the ESTIMATE algorithm to characterize the relationships between the PD or PRCR groups and immune infiltration. G, Data from the GEO database (GSE 91061) were analyzed by the ESTIMATE algorithm to characterize the correlations between STAT1 expression and immune scores (G, left panel), estimate scores (G, middle panel) or tumor purity (G, right panel). H–J Data from the GEO database (GSE 91061) were analyzed by the ImmuneCellAI-human method to characterize the immune microenvironment of melanoma tissue. The relative proportions of CD8+ T cells H, cytotoxic CD8+ T cells (I, left panel), CD4+ T cells (I, right panel), and ?d T cells (J, left panel) and immune infiltration scores (J, right panel) in the tumor microenvironment. Data are expressed as the mean (PD n = 19, PRCR n = 9) ± SD. PR partial response, CR complete response, PD progressive disease. Asterisks (*) indicate the degree of significant difference.

Fig 5: Regulation of PD-L1 by the a-KG-TET2/3-p-STAT1/3 axis is required.A Designed primers for hMeDIP analysis and ChIP experiments. B 5-hmC levels of PD-L1 promoters in B16F10 tumor-bearing mouse tissues after the designated treatments. C, D ChIP-qPCR was performed to detect the binding of STAT1 C or STAT3 D to PD-L1 promoter sites in B16F10 cells with or without a-KG treatment. E Immunoprecipitation analysis of TET3 with p-STAT1/3 in B16F10 cells receiving the designated treatments. F P-STAT1/3 protein levels at different time points after IFN? stimulation in B16F10 cells treated with CHX (100 mg/mL) or a-KG (200 µm). G The expression of p-STAT1/3 and PD-L1 protein levels was detected by immunoblotting after knocking down TET2 (left panel) /TET3 (right panel) under the designated treatment. H Schematic diagram of the mechanism. Multiple experimental data (n = 3) are counted and presented according to the statistical methods, and an asterisk (*) indicates the degree of significant difference.