DsbA-L deficiency in T cells promotes diet-induced thermogenesis through suppressing IFN-γ production

Abstract

Adipose tissue-resident T cells have been recognized as a critical regulator of thermogenesis and energy expenditure, yet the underlying mechanisms remain unclear. Here, we show that high-fat diet (HFD) feeding greatly suppresses the expression of disulfide-bond A oxidoreductase-like protein (DsbA-L), a mitochondria-localized chaperone protein, in adipose-resident T cells, which correlates with reduced T cell mitochondrial function. T cell-specific knockout of DsbA-L enhances diet-induced thermogenesis in brown adipose tissue (BAT) and protects mice from HFD-induced obesity, hepatosteatosis, and insulin resistance. Mechanistically, DsbA-L deficiency in T cells reduces IFN-γ production and activates protein kinase A by reducing phosphodiesterase-4D expression, leading to increased BAT thermogenesis. Taken together, our study uncovers a mechanism by which T cells communicate with brown adipocytes to regulate BAT thermogenesis and whole-body energy homeostasis. Our findings highlight a therapeutic potential of targeting T cells for the treatment of over nutrition-induced obesity and its associated metabolic diseases.

Fig. 1. HFD feeding downregulates T cell mitochondrial function in BAT.

a, b Mitochondrial mass of T cells in BAT (a) and iWAT (b) from ND- and HFD-fed mice was analyzed by Mito-Tracker Green staining (n = 4/group). c Gating strategies of mitoSOX-positive and MMP^low T cells in BAT and iWAT. d, e MMP of T cells in BAT (d) and iWAT (e) from ND- and HFD-fed mice was analyzed by Mito-Tracker Green and Mito-Tracker Red staining. (n = 4/group). f, g Mitochondrial ROS of T cells in BAT (f) and iWAT (g) from ND- and HFD-fed mice was analyzed by MitoSOX staining. (n = 4/group). Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test. Source data are provided as a Source Data File.

Fig. 2. DsbA-L is a critical regulator of T cell mitochondrial function.

a DsbA-L expression in CD3⁺ T cells stimulated with different concentrations of CL316,243 for 24 h. b OCR of CD4⁺ (n = 5/group) and CD8⁺ (n = 4–5/group) T cells were measured under basal conditions and in response to indicated drugs. Oligo, oligomycin; Rot, rotenone; Ant, antimycin. c ATP production of activated CD4⁺ (n = 3/group) and CD8⁺ (n = 3/group) T cells isolated from the spleen of DsbA-L^CD4-KO mice and control littermates. d The relative mtDNA content in activated CD4⁺ (n = 4/group) and CD8⁺ (n = 4/group) T cells isolated from the spleen of DsbA-L^CD4-KO mice and control littermates. e Mitochondrial fission and fusion in activated CD4⁺ T cells isolated from the spleen of DsbA-L^CD4-KO mice and control littermates were measured by western blot analyses. f Mitochondrial morphology was evaluated at indicated time points using live-cell confocal microscopy after staining with 100 nM Mito-Tracker Green and 10 μg/ml Hoechst 33342. Scale bars: 5 μm. g Mitochondrial calcium flux was measured in activated CD4⁺ T cells isolated from the spleen of DsbA-L^CD4-KO mice and control littermates. Results are representative of three independent experiments. Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test. Source data are provided as a Source Data File.

Fig. 3. T cell-specific knockout of DsbA-L alleviates diet-induced obesity and insulin resistance.

a Representative images of mice after a 12-week HFD feeding regimen starting at 8 weeks of age. b Body weight gain of DsbA-L^CD4-KO mice (n = 12) and control littermates (n = 7) during HFD feeding. c Body composition of DsbA-L^CD4-KO mice (n = 12) and control littermates (n = 7) fed an HFD for 12 weeks. d Representative images of fat pads and liver collected from DsbA-L^CD4-KO mice and control littermates fed an HFD for 12 weeks. e Relative tissue weights of HFD-fed DsbA-L^CD4-KO mice (n = 12) and control littermates (n = 7) fed an HFD for 12 weeks. f Representative H&E staining of eWAT, iWAT, BAT, and liver of DsbA-L^CD4-KO mice and control littermates fed an HFD for 12 weeks. Scale bars: 50 μm. g The average diameters of fat cells in DsbA-L^CD4-KO mice and control littermates were analyzed and quantified by the Image J software. h Fasting glucose levels in DsbA-L^CD4-KO mice (n = 4) and control littermates (n = 5) fed an HFD for 12 weeks. i Glucose tolerance test was performed in DsbA-L^CD4-KO mice (n = 12) and control littermates (n = 7) fed an HFD for 12 weeks. j Fasting insulin levels in DsbA-L^CD4-KO mice (n = 4) and control littermates (n = 5) fed an HFD for 12 weeks. k Insulin tolerance test was performed in DsbA-L^CD4-KO mice (n = 14) and control littermates (n = 11) fed an HFD for 12 weeks. Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test. Source data are provided as a Source Data File.

Fig. 4. T cell-specific knockout of DsbA-L increases energy expenditure and BAT thermogenic function.

a UCP1 expression in BAT and iWAT of DsbA-L^CD4-KO mice (n = 7) and control littermates (n = 5) fed an HFD for 12 weeks. b Immunohistochemical staining of UCP1 in BAT of DsbA-L^CD4-KO mice and control littermates fed an HFD for 12 weeks. Scale bars: 50 μm. c Thermogenic gene expression in BAT and iWAT of DsbA-L^CD4-KO mice (n = 7) and control littermates (n = 4) fed an HFD for 12 weeks. d The oxygen consumption of DsbA-L^CD4-KO mice (n = 8) and control littermates (n = 7) fed an HFD for 5 weeks was measured. e The average of oxygen consumption and energy expenditure between DsbA-L^CD4-KO mice and control littermates fed an ND (n = 4/group) or HFD (n = 7–8/group) for 5 weeks. f The basal OCR of BAT (n = 5/group) and iWAT (n = 4/group) collected from DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks. g mRNA levels of genes involved in mitochondrial and fatty acid metabolism in the BAT (n = 4–5/group) and iWAT (n = 3–4/group) of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks. h Western blot analyses of UCP1 and PGC1α in the BAT collected from DsbA-L^CD4-KO mice (n = 6) and control littermates (n = 5) fed an HFD for 5 weeks. i mRNA levels of genes involved in thermogenesis in the BAT (n = 4–5/group) and iWAT (n = 4/group) of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks. j The oxygen consumption of DsbA-L^CD4-KO mice (n = 5) and control littermates (n = 4), which were fed an HFD for 5 weeks and then tested under thermoneutral conditions (30 °C). k The average of oxygen consumption and energy expenditure between DsbA-L^CD4-KO mice (n = 5) and control littermates (n = 4) which were fed an HFD for 5 weeks and then tested under thermoneutral conditions (30 °C). Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test (a (right panel), c, f, g, h (right panel), i or ANCOVA test (e, k)). Source data are provided as a Source Data File.

Fig. 5. DsbA-L deficiency in T cells suppresses IFN-γ production while promoting Treg cell accumulation.

DsbA-L^CD4-KO mice and control littermates were fed an HFD for 5 weeks. a Gating strategies of T cell subsets in adipose tissues. b Representative FACS plots of CD4⁺IFN-γ⁺ Th1 cells (left panel) and quantification of the frequencies and total numbers of CD4⁺IFN-γ⁺ Th1 cells (right panel). (n = 4/group). c Representative FACS plots of CD8⁺IFN-γ⁺ T cells (left panel) and quantification of the frequencies and total numbers of CD8⁺IFN-γ⁺ T cells (right panel). (n = 4/group). d, e Quantification of the frequencies and total numbers of Th2 cells (d) and Treg cells (e). (n = 4/group). f mRNA levels of T cell-associated cytokines in the SVF of BAT. (n = 5/group). g Protein levels of IFN-γ in the BAT and iWAT. (n = 4/group). h Protein levels of IFN-γ in the serum of DsbA-L^CD4-KO mice (n = 10) and control littermates (n = 12). Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test. Source data are provided as a Source Data File.

Fig. 6. IFN-γ inhibits thermogenic gene expression in brown adipocytes.

a BAT explant cultures from DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks were stimulated with a submaximal dose of PMA/ionomycin for 24 h. IFN-γ secreted into the medium was collected and quantified by ELISA. (n = 5/group). b, c Differentiated primary brown adipocytes were left untreated or treated with conditioned medium (CM) of BAT collected in a in the presence or absence of IFN-γ for 24 h. UCP1 expression and thermogenesis markers were tested by western blot (b) and qPCR analyses (c), respectively. (n = 4–5/group). d The mRNA levels of TH and β3AR (Adrb3) in the BAT of DsbA-L^CD4-KO mice (n = 5) and control littermates (n = 4) fed an HFD for 5 weeks. e Differentiated primary brown adipocytes were treated with CL316,243 (1 μM) in the presence or absence of IFN-γ (5 ng/ml) for 24 h. The mRNA levels of thermogenic markers were determined by qPCR analyses. (n = 4/group). f Differentiated primary brown adipocytes were treated with CL316,243 (1 μM) in the presence or absence of IFN-γ (1 or 5 ng/ml) for 24 h. The phosphorylation and protein level of CREB, phosphorylation of PKA substrates, and protein levels of PGC1α, and UCP1 were analyzed by western blot. g cAMP levels accumulated in differentiated BFC adipocytes that were untreated or treated with CL316,243 (1 μM) in the presence or absence of IFN-γ (5 ng/ml) for 24 h. (n = 4/group). h mRNA levels of PDE4 family members in differentiated BFC adipocytes treated as in g. (n = 4/group). i Differentiated BFC adipocytes were left untreated or pre-treated with rolipram (1 μM) for 30 min, followed by stimulated with CL316,243 (1 μM) in the presence or absence of recombinant IFN-γ protein (5 ng/ml) for 24 h. Ucp1 mRNA level was determined by qPCR analyses. (n = 3/group). Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by two-tailed unpaired Student’s t-test (a, d) or ANOVA adjusted for multiple comparisons (c, e, g–i). Source data are provided as a Source Data File.

Fig. 7. Administration of IFN-γ reverses the enhanced BAT thermogenesis in HFD-Fed DsbA-L^CD4-KO mice.

a Schematic for IFN-γ dosing and metabolic analyses IFN-γ (100 μg/kg) was injected intraperitoneally every other day for the last week. b The oxygen consumption of wild-type mice fed an HFD for 5 weeks with (n = 4) or without IFN-γ (n = 5) administration for the last week. c The average oxygen consumption shown in b. d ANCOVA analyses of average VO₂ (ml/h) in c by lean mass were shown. e The basal OCR of BAT collected from DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks with (n = 4–5/group) or without IFN-γ (n = 4/group) treatment for the last week. f Western blot analyses of PGC1α and UCP1 levels in BAT of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks with or without IFN-γ treatment for the last week. (n = 3/group). g mRNA levels of thermogenic genes in BAT of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks with or without IFN-γ treatment for the last week. (n = 4/group). h The mRNA levels of PDE4D in BAT of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks with or without IFN-γ treatment for the last week. (n = 4/group). i cAMP levels accumulated in BAT of DsbA-L^CD4-KO mice and control littermates fed an HFD for 5 weeks with or without IFN-γ treatment for the last week. (n = 3/group). j A proposed model of the mechanism by which T cell DsbA-L regulates energy expenditure in BAT. A diet-induced signal stimulates the sympathetic nerve system and downstream cAMP-PKA activation to promote UCP1 expression in BAT. Meanwhile, HFD feeding leads to decreased DsbA-L expression in BAT-resident T cells. DsbA-L deficiency blocks T cell mitochondrial function and decreases IFN-γ-producing Th1 and CD8⁺ T cells accumulation in BAT. IFN-γ could act on brown adipocytes and inhibit UCP1 expression and cAMP accumulation via promoting PDE4D expression. Decreased IFN-γ production in BAT thus contributes to enhanced diet-induced BAT thermogenic capacity and alleviating over obesity. Data shown are representative of three independent experiments. All data are presented as mean ± SEM. Statistical values p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) are determined by ANOVA adjusted for multiple comparisons (e, f (right panel), g–i) or ANCOVA test (c). Source data are provided as a Source Data File.