Lack of p38 activation in T cells increases IL-35 and protects against obesity by promoting thermogenesis

Abstract

Obesity is characterized by low-grade inflammation, energy imbalance and impaired thermogenesis. The role of regulatory T cells (Treg) in inflammation-mediated maladaptive thermogenesis is not well established. Here, we find that the p38 pathway is a key regulator of T cell-mediated adipose tissue (AT) inflammation and browning. Mice with T cells specifically lacking the p38 activators MKK3/6 are protected against diet-induced obesity, leading to an improved metabolic profile, increased browning, and enhanced thermogenesis. We identify IL-35 as a driver of adipocyte thermogenic program through the ATF2/UCP1/FGF21 pathway. IL-35 limits CD8⁺ T cell infiltration and inflammation in AT. Interestingly, we find that IL-35 levels are reduced in visceral fat from obese patients. Mechanistically, we demonstrate that p38 controls the expression of IL-35 in human and mouse Treg cells through mTOR pathway activation. Our findings highlight p38 signaling as a molecular orchestrator of AT T cell accumulation and function.

Keywords: Adipose Tissue; Obesity; T Regulatory Cells; Thermogenesis; p38 Stress Kinases.

Figure 1. p38 MAPK pathway is upregulated in T cells in obese human white adipose tissue.

(A) Dot plot illustrating the expression of the indicated T cell marker genes in the different cell type clusters in human white AT single-cell RNA-seq data from Emont et al (2022). (B) Dot plot showing the expression of the indicated genes by BMI range in human white AT T cell cluster shown in (A). (C) Volcano plot of differentially expressed genes in human white AT T cell cluster in severe obese (BMI 40–50 kg/m²) versus non-obese (BMI 20–30 kg/m²) subjects. The vertical dashed line indicates a log2 fold change cut-off of 0.15. The horizontal dashed line indicates a −log10 adjusted p-value (using Bonferroni correction) cut-off of 1.3 (adj p-value < 0.05). Data information: Differentially expressed genes between BMI ranges were identified with a non-parametric Wilcoxon rank sum test. Obese (BMI 40–50 kg/m²) n = 6 biologically independent patients; Non-obese (BMI 20–30 kg/m²) n = 10 biologically independent patients; DC: dendritic cells; Mac: macrophages; Mono: monocytes; Neu: neutrophils; Mast: mastocytes. Source data are available online for this figure.

Figure 2. MKK3/6 deficiency in T cells protects against HFD-induced obesity.

(A–J) MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 8 weeks (starting at 8–10 weeks old). (A) Body weight evolution in MKK3/6^CD4-KO and CD4-Cre male mice for 8 weeks. Data are presented as the increase above initial weight (left) and absolute weight at the end of the experiment (right). (B) MRI analysis of body and fat mass in CD4-Cre and MKK3/6^CD4-KO mice after 8 weeks of HFD. Representative images are shown on the left. (C) eWAT, sWAT, BAT, and liver mass relative to tibia length. (D) Food intake during 46 h. (E, F) Comparison of energy balance between HFD-fed MKK3/6^CD4-KO and CD4-Cre mice examined in metabolic cages. Hour-by-hour variation in EE (kcal/h) (E); mean EE (Kcal/h) and ANCOVA analysis of EE (kcal/h) (F). (G, H) Comparison of energy balance between HFD-fed MKK3/6^CD4-KO and CD4-Cre mice examined in metabolic cages. Hour-by-hour lean–mass-corrected variation in EE (kcal/h/kg) (G) and mean lean mass-corrected EE (kcal/h/kg) (H). (I) Blood glucose concentration after 8 weeks of HFD in mice fed (left) or fasted overnight (right). (J) Representative haematoxylin–eosin and oil-red O staining of liver sections. Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Analysis by 2-way ANOVA coupled to the Bonferroni post-test (A) or coupled to the Sidak’s multiple comparison post-test (E, G); and by t test or by the Welch test when variances were different (A–C, F, H, I). n = 6–9 biologically independent mice for each group, represented as single dots in the graphs (A–I). Scale bar: 1 cm (B), 100 µm (J). Source data are available online for this figure.

Figure 3. MKK3/6 deficiency in T cells protects against HFD-induced obesity by increasing BAT temperature.

(A–E) MKK3/6^CD4-KO and MKK3/6^f/f mice were fed a high-fat diet (HFD) for 9 weeks. (A) Body weight increased for 9 weeks. (B) MRI analysis of body and fat mass and representative images on the left. (C) Fecal lipid excretion over 5 days. (D) Locomotor activity and rearing over 24 h. (E) Skin temperature surrounding interscapular BAT. Right panels show representative infrared thermal images. Data Information: Data are presented as mean ± SEM, *p < 0.05, ***p < 0.001, ns: not significant. Analysis by 2-way ANOVA coupled to the Sidak’s multiple comparison post-test (A) or t test or by the Welch test when variances were different (B–E). n = 6–8 biologically independent I’mmice for each group, represented as single dots in the graphs (A–E). Scale bar: 1 cm (B). Source data are available online for this figure.

Figure 4. MKK3/6 deficiency in T cells protects against HFD-induced obesity in isothermal housing.

(A–E) MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 8 weeks (starting at 8–10 weeks old) and housed at 30 °C during whole course of HFD. (A) Body weight evolution in CD4-Cre and MKK3/6^CD4-KO male mice for 8 weeks. Data are presented as the increase above initial weight (left) and absolute weight at the end of the experiment (right). (B) MRI analysis of body and fat mass in MKK3/6^CD4-KO and CD4-Cre mice after 8 weeks of HFD. Representative images are shown on the right. (C) eWAT, sWAT, BAT, and liver mass relative to tibia length. (D) Representative haematoxylin–eosin and oil-red O staining of liver, eWAT, and sWAT sections. (E) Skin temperature surrounding interscapular BAT. Right panels show representative infrared thermal images. Data information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001. Exact p-values are shown. Analysis by 2-way ANOVA coupled to the Bonferroni post-test (B) or t test or by the Welch test when variances were different (A–C, E). n = 7–9 biologically independent mice for each group, represented as single dots in the graphs. Scale bar: 1 cm (B), 100 µm for liver and 250 µm for eWAT and sWAT (D). Source data are available online for this figure.

Figure 5. Lack of MKK3/6 in T cells increases BAT thermogenesis and adipose tissue browning.

(A–H) MKK3/6^CD4-KO and control CD4-Cre mice were HFD-fed for 8 weeks. (A) Skin temperature surrounding interscapular BAT. Right panels show representative infrared thermal images. (B) Representative H&E staining of BAT sections. (C) Western blot analysis of UCP1 in BAT. (D) qRT-PCR analysis of thermogenic gene mRNA expression in BAT isolated from control or MKK3/6^CD4-KO mice. mRNA expression was normalized to the expression of β-actin mRNA and presented as fold increase compared to CD4-Cre. (E) qRT-PCR analysis of browning genes mRNA expression from eWAT isolated from control or MKK3/6^CD4-KO mice. mRNA expression was normalized to β-actin mRNA and presented as fold increase compared to CD4-Cre. (F) Representative H&E staining of eWAT sections. (G) qRT-PCR analysis of browning genes mRNA expression from sWAT isolated from control or MKK3/6^CD4-KO mice. mRNA expression was normalized to β-actin mRNA and presented as fold increase compared to CD4-Cre. (H) Representative H&E staining of sWAT sections. Data information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Exact p-values are shown. Analysis by t test or by the Welch test when variances were different. n = 7–9 biologically independent mice for each group, represented as single dots in the graphs. Scale bar: 100 µm (B, F, H). Source data are available online for this figure.

Figure 6. MKK3/6 deficiency in T cells promotes Treg cell accumulation in adipose tissue.

(A–C) MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 8 weeks. (A) FACS quantification and representative dot plots of Treg cells (CD4⁺CD25⁺Foxp3⁺) in SVF and (B) CD8⁺ T cells in SVF. (C) FACS quantification and representative dot plots of myeloid (F4/80⁺CD11b⁺), M1 macrophages (Mϕ) (F4/80⁺CD11b⁺CD11c⁺), M2 Mϕ (F4/80⁺CD11b⁺CD206⁺), and NK (NK1.1⁺CD3^-) cells in SVF. Data Information: Data are presented as mean ± SEM, *p < 0.05, ***p < 0.001. Exact p-values are shown. Analysis by t test or by the Welch test when variances were different. n = 5–7 biologically independent mice for each group, represented as single dots in the graphs (A–C). Source data are available online for this figure.

Figure 7. p38 activation in Treg cells inhibits IL-35 production.

(A–C) MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 8 weeks. (A, B) qRT-PCR analysis of mRNA expression in (A) eWAT and (B) SVF isolated from control or MKK3/6^CD4-KO mice. mRNA expression was normalized to b-actin mRNA. (C) FACS quantification and representative dot plots of IL-35⁺ Treg cells in lymph nodes. (D) In vitro Treg cell induction (iTreg). Naive CD4⁺ T cells were isolated from the spleens of CD4-Cre and MKK3/6^CD4-KO mice stimulated for 96 h with plate-bound anti-CD3, soluble anti-CD28 + IL-2 + TGFβ. qRT-PCR analysis of p35 and Ebi3 mRNA in iTregs derived from control or MKK3/6^CD4-KO mice. mRNA expression was normalized to b-actin mRNA. (E) Induction of iTregs from CD4⁺ T cells isolated from healthy human donor buffy coats and stimulated with plate-bound anti-CD3, soluble hIL-2 + hTGFβ for 6 days in the presence or absence of the p38 pan inhibitor BIRB796. qRT-PCR analysis of P35 and EBI3 mRNA in iTregs. mRNA expression was normalized to GAPDH mRNA. (F) FACS analysis of IL-35 MFI in in vitro induced iTreg cells from CD4-Cre and TSC1^CD4-KO mice. (G) Western blot analysis of p-s6 protein S240/244 and p-p38 Thr180/Tyr182 in iTreg cells from CD4-Cre and MKK3/6^CD4-KO mice. Loading control for p-p38 was run on different gel and not presented. (H) FACS analysis of IL-35 MFI in in vitro induced iTreg cells from MKK3/6^CD4-KO mice in the presence or absence of rapamycin for 4 h. Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Exact p-values are shown. Analysis by t test. n = 4–10 biologically independent mice (A–C) or n = 4–9 biologically independent wells (D–H) for each group, represented as single dots in the graphs. Source data are available online for this figure.

Figure 8. Treg-derived IL-35 promotes thermogenesis by increasing ATF-2 phosphorylation and UCP1 and FGF21 levels.

(A) mRNA expression of p35 subunit of IL-35 in human visceral fat isolated from lean and obese patients. (B) MKK3/6^CD4-KO and control CD4-Cre mice were exposed to cold for 4 h. Body and BAT temperature was measured every hour. (C) C57BL6 mice were treated with recombinant IL-35 i.v. (300 ng per mouse) and BAT temperature was measured 4 h later (D) Immortalized brown preadipocytes were differentiated in vitro. Once differentiated, cells were stimulated in the presence or absence of IL-35 (100 ng/ml) for 48 h and UCP1 and FGF21 levels were analyzed by immunoblot. Loading control for UCP1 was run on different gel and not presented. (E) Immortalized brown preadipocytes were differentiated in vitro. Once differentiated, cells were stimulated in the presence or absence of IL-35 (100 ng/ml) for 0–120 min and ATF2 phosphorylation was analyzed by immunoblot. (F) Differentiated adipocytes were stimulated with IL-35 (100 ng/ml) for 48 h in the presence or absence of SB203580 inhibitor (10 μM). The expression of Ucp1 level was measured by qRT-PCR and relativized to b-actin. (G) Primary white preadipocytes were isolated from C57BL6 mice and differentiated in vitro. Once differentiated, cells were stimulated with PBS or with IL-35 (100 ng/ml) for 48 h in the presence or absence of SB203580 inhibitor (10 μM). The expression of principal adipogenic markers (Pparg, Adipoq, Leptin, Perlinipin) level was measured by qRT-PCR and relativized to b-actin. Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001. Exact p-values are shown. Analysis by t test (A, C, D, F), 2-way ANOVA (B), or 1-way ANOVA (G). Lean N = 12 biologically independent patients; Obese N = 52 biologically independent patients (A). n = 5–9 biologically independent mice (B, C) or n = 2–6 biologically independent wells (D–G) for each group, represented as single dots in the graphs. Source data are available online for this figure.

Figure EV1. MKK3/6 deficiency in T cells increases energy expenditure and BAT temperature.

(A) Comparison of energy balance between ND-fed MKK3/6^CD4-KO and CD4-Cre mice examined in a metabolic cage over a 3-day period. Hour-by-hour lean–mass-corrected variation in energy expenditure (EE) (left panel); mean lean–mass-corrected EE (middle panel); and ANCOVA analysis of EE (kcal/h) (right panel). (B) Food and water intake, locomotor activity, and respiratory quotient obtained from metabolic cages. (C, D) Body temperature and skin temperature surrounding interscapular BAT. Right panels show representative infrared thermal images in (C) CD4-Cre and MKK3/6 ^CD4-KO mice and (D) in littermates (MKK3/6^f/f) and MKK3/6^CD4-KO mice fed chow diet. Data information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ns: not significant. Exact p-values are shown. Analysis by 2-way ANOVA coupled to the Sidak’s multiple comparison post-test (A) or by t test or by the Welch test when variances were different (B–D). n = 5–8 biologically independent mice for each group, represented as single dots in the graphs (A–D). Source data are available online for this figure.

Figure EV2. MKK3/6 deficiency in T cells protects females against HFD-induced obesity.

(A–D) Female MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 9 weeks (starting at 8–10 weeks old). (A) Body weight evolution in MKK3/6^CD4-KO and CD4-Cre female mice for 9 weeks. Data are presented as the increase above initial weight (left) and absolute weight at the end of the experiment (right). (B) MRI analysis of body and fat mass in MKK3/6^CD4-KO and CD4-Cre mice after 8 weeks of HFD. Representative images are shown on the right. (C) eWAT, sWAT, BAT, and liver mass relative to tibia length. (D) Skin temperature surrounding interscapular BAT. Right panels show representative infrared thermal images. Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Analysis by 2-way ANOVA coupled to the Bonferroni post-test (A) or t test or by the Welch test when variances were different (A–D). n = 8–10 biologically independent mice for each group, represented as single dots in the graphs (A–D). Source data are available online for this figure.

Figure EV3. Lack of MKK3/6 in T improves adipose tissue metabolic homeostasis.

(A, B) MKK3/6^CD4-KO and control CD4-Cre mice were fed an HFD for 8 weeks. qRT-PCR analysis of adipogenic, lipogenic, β-oxidation, and glycolytic genes mRNA expression from (A) eWAT and (B) sWAT isolated from control CD4-Cre or MKK3/6^CD4-KO mice. mRNA expression was normalized to the amount of b-actin mRNA. Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Exact p values are shown. Analysis by t test or Welch’s test when variances were different. n = 8-9 biologically independent mice for each group, represented as single dots in the graphs. Source data are available online for this figure.

Figure EV4. MKK3/6 deletion in T cells increases Treg cell population in blood and lymph nodes.

(A–C) MKK3/6^CD4-KO and CD4-Cre mice were fed a high-fat diet (HFD) for 8 weeks. FACS quantification and representative dot plots of CD4⁺, CD8⁺ and Treg cells (CD4⁺CD25⁺Foxp3⁺) in spleen (A), blood (B), and lymph nodes (C). Data Information: Data are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. Analysis by t test or Welch’s test when variances were different. n = 7 biologically independent mice for each group, represented as single dots in the graphs. Source data are available online for this figure.

Figure EV5. p38 MAPK pathway is upregulated in Treg cells in obese human adipose tissue.

(A–G) The analysis was performed using human white adipose tissue single-cell RNA-seq data from Emont et al (Emont et al, 2022). (A) Dot plot of the expression of the indicated regulatory T cell (Treg) marker genes in the different cell type clusters. (B) Dot plot of the expression of the indicated genes by BMI range in human white adipose tissue Treg cluster shown in (A). (C) Violin plot showing the level of MAPK14 gene expression by BMI range in the human white adipose tissue Treg dataset. (D–E) Volcano plots of differentially expressed genes in human white adipose tissue Treg subcluster in obese (BMI 30–40 kg/m²) versus non-obese (BMI 20–30 kg/m²) subjects (D) and in severe obese (BMI 40–50 kg/m²) versus non-obese (BMI 20–30 kg/m²) subjects (E). (F) Violin plot showing the level of MAP2K3 gene expression by BMI range in the human white adipose tissue Treg dataset. (G) Volcano plots of differentially expressed genes in human white adipose tissue Treg subcluster in class 1 and 2 obesity (BMI 30–40 kg/m²) versus non-obese (BMI 20–30 kg/m²) subjects. The vertical dashed lines in (D, E, G) indicate a log2 fold change cut-off of 0.25. The horizontal dashed lines in (D, E, G) indicate a −log10 p-value cut-off of 1.3 (p-value < 0.05). Data information: Differentially expressed genes between BMI ranges were identified with a non-parametric Wilcoxon rank sum test. Obese (BMI 30–40 kg/m²) N = 3 biologically independent patients; Severe obese (BMI 40–50 kg/m²) N = 6 biologically independent patients; non-obese (BMI 20–30 kg/m²) N = 5 biologically independent patients. DC: dendritic cells; Mac: macrophages; Mono: monocytes; Neu: neutrophiles; Mast: mastocytes.