Brown adipose TRX2 deficiency activates mtDNA-NLRP3 to impair thermogenesis and protect against diet-induced insulin resistance

Figure 1. BAT-specific TRX2 deficiency protects mice from diet-induced hepatic steatosis and hypertriglyceridemia. (A) Growth trend of Trx2BATKO and WT mice under NCD (n = 6) or HFD (n = 10). (B) Ratio of adipose depots to body weight from 16-week-old Trx2BATKO (n = 3) and WT (n = 4) mice under HFD. (C) Representative histologic images of iBAT from HFD-fed mice. Arrowheads denote large unilocular LDs. BAT adipocytes with unilocular LD diameter of 20 μm or more were quantified. (D) Representative histologic images of adipose tissues and liver from Trx2BATKO and WT mice under HFD. Arrows denote multilocular LDs in WAT, while arrowheads denote LDs in the liver. Adipocytes with multilocular LDs (%) are quantified. n = 6. (E–H). Lipid profile, including plasma TG (E), LDL-C (F), TC (G), and HDL-C (H) levels from 16-week-old WT and Trx2BATKO mice under HFD (n = 6). (I) Representative BODIPY staining images of liver from mice under HFD. White box denotes magnified areas. Arrowheads denote LDs in the liver. (J) TG contents of liver from HFD-fed mice (n = 3). (K) Expression of lipid metabolic–related genes in livers from HFD-treated mice (n = 3). Quantitative data are presented as mean ± SEM. *P <…

Figure 2. BAT-specific TRX2 deficiency protects mice from diet-induced insulin resistance and hypertriglyceridemia. (A) Fasting blood glucose levels of mice under NCD or HFD (n = 6) and quantification by incremental AUC. (B) Fasting plasma insulin levels of HFD-fed mice (n = 4). (C) GTT of HFD-treated mice (n = 6) and quantification. (D) Insulin tolerance test of HFD-treated mice (n = 6). Quantitative data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. Significance was assessed by 2-way ANOVA with Bonferroni’s post hoc tests (A, C, and D) and 2-tailed Student’s t test (B).

Figure 3. TRX2 loss enhances lipid uptake in BAT. (A) Intestine lipid absorption capacity of 16-week-old WT and Trx2BATKO mice by detecting plasma radioactivity under oral gavage of [3H]-labeled triolein along with poloxamer 407 injection (n = 5). CPM, counts per minute. (B) Hepatic VLDL production detection by measuring plasma TG levels from WT and Trx2BATKO mice under overnight fasting and treated with LPL inhibitor poloxamer 407 to inhibit catabolism (n = 4). (C) Oral lipid tolerance test performed on mice to determine circulating TG clearance (n = 5). (D–E) Exogenous lipid uptake in tissues determined by detecting radioactivity after oral gavage of [3H]-labeled triolein of WT and Trx2BATKO mice (n = 3). Plasma radioactivity of WT and Trx2BATKO mice are shown in E. Quantitative data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. Significance was assessed by 2-way ANOVA with Bonferroni’s post hoc tests (A–C) and 2-tailed Student’s t test (D and E).

Figure 4. TRX2 deficiency in BAT impairs thermogenesis. (A) Intrarectal temperature of 16-week-old WT and Trx2BATKO mice during acute cold intolerance test in the absence of food (n = 6). (B–F) WT and Trx2BATKO mice after 3 days of cold exposure. (B) Intrarectal temperature (n = 6). (C) Food intake during cold exposure. (D) Angptl4 mRNA expression in iBAT (n = 3). (E) Western blots of indicated proteins in iBAT. Relative protein levels are presented as fold changes by taking WT as 1.0. n = 2. (F) Histologic iBAT images. Arrowheads denote large LDs. (G and H) Ex vivo lipolysis assay detecting FFA levels (G) and FFA release efficiency (H) of isolated iBAT treated with saline or ISO (n = 3). (I) In vivo lipolysis assay detecting plasma NEFA (n = 6). (J) PPARα and PPARγ protein expression in iBAT. Relative protein levels are presented by taking saline as 1.0. n = 2 mice. (K) Relative mRNA expression of Ppara (n = 4) and its target genes (n = 3) in iBAT. (L) Ex vivo FA oxidation of iBAT (n = 6). Quantitative data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. Two-way ANOVA with Bonferroni’s post hoc tests (A–C), 1-way ANOVA followed by Tukey’s post hoc test (G and I), and…

Figure 5. TRX2 BAT ablation suppresses FA oxidation. (A and B) Gene expression in iBAT from mice after 3 days of cold exposure (n = 3). (C–I) Mice treated with CL136,243 (1 mg/kg) for 3 hours. (C) Antioxidant protein levels in iBAT from WT mice by taking saline as 1.0 (n = 2). (D) Intrarectal temperature (n = 6). (E) Histologic images of iBAT. Arrowheads denote large LDs, with quantification in the right panel (n = 6). (F and G) Plasma NEFA and glycerol levels (n = 4). (H) Immunoblots of proteins in iBAT (each lane from 3 mice). (I) Relative mRNA expression of genes in iBAT (n = 3). Quantitative data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. Two-tailed Student’s t test (A and B), and 1-way ANOVA followed by Tukey’s post hoc test (D, F, G, and I). Scale bar: 100 μm (E).

Figure 6. TRX2 loss in BAT contributes minimally to whole-body EE. (A–D) Mice were fed or fasted for 24 hours from 7 am. (A) Representative histologic images of iBAT from 16-week-old WT and Trx2BATKO mice. Three sections from each mouse (n = 3 mice). (B) Cd36 expression in iBAT (n = 3). (C) Gene expression of FA oxidation–related genes in iBAT (n = 3 mice). (D) Intrarectal temperature (n = 6). (E–I). Mice were subjected to metabolic cage evaluation. Mice were individually placed in metabolic cages (CLAMs) and allowed to acclimatize for 48 hours before readings were taken. Mice were fasted for 24 hours and refed for 48 hours (n = 5 for WT, n = 4 for KO in fasted, n = 3 for KO in refed). EE (E), oxygen consumption (VO2) (F), food intake (G), physical activity counts (H), and RER (I) were measured. Analyses were performed in male mice. Quantitative data are presented as mean ± SEM. *P < 0.05; ***P < 0.001. Significance was assessed by 2-tailed Student’s t test. Scale bar: 100 μm (A).

Figure 7. TRX2 loss induces excessive mtROS and cytosolic mtDNA release in BAT. (A and B) mtROS were detected by MitoSOX in iBAT. Data are presented as arbitrary fluorescence unit (AFU). n = 5. (C) mtROS were detected by MitoSOX (red) and mitoPY1 (green) in freshly isolated iBAT. Data are presented as relative fluorescence unit (RFU) by taking WT as 1.0. n = 6. (D) Western blot analysis of redox proteins in BAT. Relative protein levels and PRX3 dimer/monomer ratios are presented as fold changes by taking WT as 1.0. n = 2. (E–H) Mitochondrial structures of iBAT by EM. (E) Representative EM images. White boxes denote magnified areas, and arrowheads indicated mitochondria. Quantification of mitochondrial numbers (F), damaged mitochondria percentage (G), and mitochondria with opening outer membrane (H). Ten fields were randomly chosen for each group (n = 3). (I) mtDNA copy number of iBAT (n = 3). Tert was used as a nuclei DNA control. (J and K) mtROS were detected by MitoTracker (green) and MitoSOX (red) in primary brown adipocytes. Arrows indicate MitoSOX+ cells. Quantification of RFU is shown in K (n = 3). (L and M) Cytosolic DNA detected by costaining for TOM20 (red) and double-stranded DNA…

Figure 8. TRX2 deficiency triggers cGAS/STING pathway and NLRP3 inflammasome activation in BAT. (A) Proinflammatory cytokine mRNA expression in iBAT from 16-week-old WT and Trx2BATKO mice (n = 3). (B) NLRP3 inflammasome-related gene expression of iBAT (n = 4). (C) IL-1β levels of iBAT as detected by ELISA (n = 3). (D) Western blots of cGAS/cGAMP/STING pathway proteins in iBAT. Relative protein levels are presented as fold changes by taking WT as 1.0. n = 2. (E and F) Immunostaining of NF-κB p65 (green) in primary brown adipocytes with DAPI counterstaining. Boxes denote magnified areas, and arrowheads denote nuclear p65 or IRF3. Nuclear p65 translocation (% p65+ nuclei) was quantified. (G and H) Immunostaining of IRF3 (green) in primary brown adipocytes with DAPI counterstaining. Nuclear IRF3 translocation (% IRF3+ nuclei) was quantified. (I) Western blots of inflammasome-related proteins in isolated mature brown adipocytes from iBAT. Relative protein levels are presented as fold changes by taking WT as 1.0. n = 2. (J and K) Inflammasome activation in primary brown adipocytes detected by ASC (red) and NLRP3 (green) colocalization. White arrows indicate ASC+ specks. (K) Quantification of the…

Figure 9. TRX2 loss–driven NLRP3 inflammasome activation relies on mtDNA release in BAT. Primary brown adipocytes were differentiated and cultured in the absence or presence of indicated inhibitors or siRNAs for 4 days. (A–C) Effects of CsA on mtDNA release and activation of the cGAS/STING and NLRP3 inflammasome pathways. WT and KO primary brown adipocytes were treated with CsA or vehicle (DMSO). (A) Immunostaining for cytosolic dsDNA (with mitochondrial marker Tom20), nuclear IRF3, or ASC/NLRP3 specks. Boxes denote magnified areas, and arrows denote cytosolic mtDNAs, nuclear IRF3, and ASC+ specks. (B) Nuclear IRF3 translocation (% IRF3+ nuclei) and percentages of ASC speck+ cells were quantified. n = 10 random fields. (C) Western blots for the cGAS/STING and NLRP3 inflammasome pathways. Relative protein levels are presented as fold changes by taking WT as 1.0. (D–F) Effects of MCC950 on mtDNA release and activation of the cGAS/STING and NLRP3 inflammasome pathways. WT and KO primary brown adipocytes were treated with MCC950 or vehicle (DMSO). (D) Western blots for the cGAS/STING and NLRP3 inflammasome pathways. Relative protein levels are presented as fold changes by taking WT as 1.0. (E)…

Figure 10. NLRP3 inflammasome inhibitor MCC950 ameliorates the impaired thermogenic phenotype of Trx2BATKO mice. Eight-week-old WT and Trx2BATKO mice under NCD were treated with MCC950 (10 mg/kg) or an equal volume of saline by intraperitoneal injection every other day for 8 weeks (n = 6). (A) mRNA expression of inflammasome-related genes in iBAT (n = 3). Relative mRNA levels are presented as fold changes by taking saline-treated WT as 1.0. (B) Western blots of NLRP3 inflammasome-related proteins in iBAT. Protein levels are presented as fold changes by taking saline-treated WT as 1.0. n = 3. (C) Mature IL-1β levels in iBAT were measured by ELISA (n = 3). (D) Protein levels of the cGAS/STING pathway in iBAT were detected by Western blotting. Protein levels are presented as fold changes by taking saline-treated WT as 1.0. n = 3. (E) mRNA expression of proinflammatory cytokines in iBAT (n = 3). Relative mRNA levels are presented as fold changes by taking saline-treated WT as 1.0. (F) Western blots of PPARα in iBAT. n = 3. (G) mRNA expression of lipid metabolism genes in iBAT (n = 3). Normalized mRNA levels (versus β-actin mRNA) are presented. (H) Intrarectal temperature during acute cold…

Figure 11. Effects of MCC950 on diet-induced glucose metabolism and insulin sensitivity. Eight-week-old WT and Trx2BATKO mice were fed with HFD in the presence of MCC950 (10 mg/kg) or an equal volume of saline by intraperitoneal injection every other day for 8 weeks. (A) Body weight gain (n = 4). (B) Fasting blood glucose levels and incremental AUC (n = 4). (C) Blood glucose levels and incremental AUC in GTT (n = 4). (D) Blood glucose levels and incremental AUC in insulin tolerance test (n = 4). (E–H) Lipid profiles, including LDL-C (E), HDL-C (F), TC (G), and TG (H) (n = 4). Significance was assessed by 2-way ANOVA with Bonferroni’s post hoc tests (A–D) and 1-way ANOVA followed by Tukey’s post hoc test (E–H).

Figure 12. MCC950 treatment reverses the metabolic benefits of Trx2BATKO mice under nutrient overload. (A) Representative histologic images of iBAT, ingWAT, and liver tissues from 16-week-old HFD-fed WT and Trx2BATKO mice. Arrowheads denote LDs in BAT and liver, whereas arrows denote ingWAT with multilocular LDs. (B) BAT adipocytes with unilocular LD diameter of 20 μm or more and proportion of WAT with multilocular LDs are quantified. n = 6. (C) TG contents of liver from HFD-fed mice with or without MCC950 (n = 3). Quantitative data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. Significance was assessed by 1-way ANOVA followed by Tukey’s post hoc test (B and C). Scale bar: 100 μm (A). (D) A schematic diagram summarizing our findings. TRX2 deficiency in BAT protects from insulin resistance and impairs thermogenesis through the mtDNA/NLRP3 inflammasome pathways. TRX2 ablation in brown adipocytes triggers ROS production and mtDNA release, which induce NLRP3-mediated aberrant innate inflammation in BAT. Consequently, enhanced lipid uptake, with suppressed mitochondrial FA oxidation in BAT, results in the “whitening” of BAT, protecting mice from insulin resistance, but…