Fig 2: TRPM2 regulated HG-induced ROS production and NADPH oxidase activation, which linked TXNIP to NLRP3 inflammasome activation in human monocytic cell lines.(A) ELISA for IL-1ß secretion from the supernatants of treated cells. (B,C) The ROS production was measured by CM-F2DCFDA staining. (D) The level of protein carbonyl content was measured by protein carbonyl content assay. (E) The cellular NADPH level was measured by NADP/NADPH assay, and NADPH oxidase activity was normalized to total cellular protein levels. (A,B) U937 cells were stimulated with low glucose (LG; 5.5 mM glucose) or high glucose (HG; 30 mM glucose for 24, 48, 72 h) (A) with pre-treatment of N-acetyl-L-cysteine (NAC; 25 mM) or diphenyleneiodonium chloride (DPI; 10 µM), or (B) in the presence of GAPDH- (GA si), p47 phox- or TRPM2-siRNA (n = 6–7). (C–E) The cells were pre-treated with 3-aminobenzamide (3-AB; 5 mM), 3,4-dihydro-5-[4-(1-piperidinyl)butoxy]-1(2H)-isoquinolinone (DPQ; 100 µM), DPI (10 µM), GAPDH- or TRPM2-siRNA under LG, mannitol (Ma; 30 mM) or HG conditions in (C) THP-1 cells or (D,E) U937 cells (n = 5–6). (F) Representative immunoblots for pro-IL-1ß, IL-1ß p17, pro-caspase-1, cleaved caspase-1 (p20), and ß-actin in the presence of GADPH- or TXNIP-siRNA under HG in U937 cells (n = 5). (G) Immunofluorescence images showing the location of the TXNIP and NLRP3 in fixed cells by using confocal microscopy, in the presence of GAPDH- or TRPM2-siRNA, or EGTA-AM (5 mM) under HG in U937 cells (n = 4). The percentage of TXNIP co-localization with NLRP3 inflammasome was calculated as the average volume of the overlapping areas. Data were shown as mean ± S.E.M. (A) *P < 0.05 and ***P < 0.001 vs. HG. (B) **P < 0.01 vs. HG + GAPDH-siRNA. (C–E) *P < 0.05 and ***P < 0.001 vs. LG; #P < 0.05 and ##P < 0.01 vs. HG. (F) **P < 0.01 vs. LG + GAPDH-siRNA; ##P < 0.01 vs. HG + GAPDH-siRNA.

Fig 3: TRPM2 knockdwon markedly reduced p47 phox phosphorylation, TXNIP expression and Trx activity under HG in U937 monocytes.(A) Representative immunoblots and graphs for protein expressions of p47 phox, p22phox, gp91pox, TXNIP or GAPDH in the presence of BAPTA-AM (2.5, 5, 10 µM), or EGTA (0.5, 1, 5 mM) under low glucose (LG; 5.5 mM glucose) or high glucose (HG; 30 mM glucose) (n = 4–5). (B) Representative immunoblots and graphs for protein expressions of p47 phox, TXNIP or GAPDH or ß-actin in the presence of GAPDH- or TRPM2-siRNA under HG (n = 4). (C) Quantitative PCR was performed on TXNIP mRNA in GAPDH- (GA si) or TRPM2-siRNA-treated cells under LG or HG, and it was normalized to LG + GAPDH-siRNA (n = 4). (D) TRX activity was determined by the insulin disulfide reduction assay, and was normalized to LG + GAPDH-siRNA. The cells were stimulated with HG for 24, 48, 72 h in the presence of GAPDH- or TRPM2-siRNA (n = 4). Data were shown as mean ± S.E.M. (C,D) *P < 0.05 and ***P < 0.001 vs. LG + GAPDH-siRNA; #P < 0.05 and ###P < 0.001 vs. HG + GADPH-siRNA.

Fig 4: Ca2+ influx via TRPM2 was required for HG-induced NLRP3 inflammasome activation in U937 cells.(A) Fluorescence images showing fluo-4-loaded cells taken at indicated timepoints, and representative flow cytometric plot and graph showing relative median fluorescence intensity (MFI) of intracellular Ca2+ concentration by Fluo-4 staining under low glucose (LG; 5.5 mM glucose;) or high glucose (HG; 30 mM glucose) (n = 5). (B) Relative changes in [Ca2+]i, evoked by H2 O2 (1 mM) over the time course. The cells were treated in the presence of 2-aminoethoxydiphenyl borate (2-APB; 100 µM), or with pre-treatment of 3-aminobenzamide (3-AB; 5 mM) or 3,4-dihydro-5-[4-(1-piperidinyl)butoxy]-1(2H)-isoquinolinone (DPQ; 100 µM), or GAPDH- (GA si) or TRPM2-siRNAs (TRPM2 si) under HG (n = 4). (C) Representative immunoblots for pro-IL-1ß, IL-1ß p17, pro-caspase-1, cleaved caspase-1 (p20), and GAPDH in the presence of EGTA-AM (0.5, 1, 5 mM) under HG (n = 4). (D) Immunofluorescence images showing the location of caspase-1 and ASC in fixed cells using confocal microscopy in the presence of GAPDH- or TRPM2-siRNAs, or EGTA-AM (5 mM) under HG. The percentage of co-localization of caspase-1 with ASC was calculated as the average volume of the overlapping areas (n = 4). Data were shown as mean ± S.E.M. (A) *P < 0.05 vs. LG. (B) *P < 0.05 and **P < 0.01 vs. HG; ##P < 0.01 vs. HG + GA si. (D) **P < 0.01 vs. LG + GA si; ##P < 0.01 vs. HG + GA si.

Fig 5: TRPM2 was required for HG-induced caspase-1-dependent IL-1ß secretion in U937 monocytes.(A,C) ELISA for IL-1ß or IL-18 secretion from the supernatants of treated cells. (A) U937 cells were stimulated with low glucose (LG; 5.5 mM glucose), mannitol (Ma; 30 mM), high glucose (HG; 10, 20, 30 mM glucose) for 48 h (n = 6–7). (B,D) Representative immunoblots for pro-IL-1ß, IL-1ß p17, pro-caspase-1, cleaved caspase-1 (p20), GAPDH and ß-actin in the presence of (B) GAPDH-, NLRP3-, ASC- or caspase-1-siRNA, or (D) GAPDH- or TRPM2-siRNAs under HG. (C) U937 or THP-1 cells were stimulated with HG in the presence of GAPDH- or TRPM2-siRNAs, or 3-AB (5 mM) or DPQ (100 µM) (n = 4). Data shown were shown as mean ± S.E.M. (A,C) *P < 0.05, **P < 0.01 and ***P < 0.001 vs. LG; #P < 0.05 and ##P < 0.01 vs. HG.