Metformin enhances anti-mycobacterial responses by educating CD8+ T-cell immunometabolic circuits

Abstract

Patients with type 2 diabetes (T2D) have a lower risk of Mycobacterium tuberculosis infection, progression from infection to tuberculosis (TB) disease, TB morality and TB recurrence, when being treated with metformin. However, a detailed mechanistic understanding of these protective effects is lacking. Here, we use mass cytometry to show that metformin treatment expands a population of memory-like antigen-inexperienced CD8⁺CXCR3⁺ T cells in naive mice, and in healthy individuals and patients with T2D. Metformin-educated CD8⁺ T cells have increased (i) mitochondrial mass, oxidative phosphorylation, and fatty acid oxidation; (ii) survival capacity; and (iii) anti-mycobacterial properties. CD8⁺ T cells from Cxcr3^-/- mice do not exhibit this metformin-mediated metabolic programming. In BCG-vaccinated mice and guinea pigs, metformin enhances immunogenicity and protective efficacy against M. tuberculosis challenge. Collectively, these results demonstrate an important function of CD8⁺ T cells in metformin-derived host metabolic-fitness towards M. tuberculosis infection.

Fig. 1. Metformin-educated CD8⁺ T cells protect against M. tuberculosis.

a Experimental strategy and assessed readouts. b, c Bacterial load in the lung of recipient mice 21 days post transfer, from two different experiments. No transfer—Irradiated recipients that did not receive any cells. Metformin-educated—CD4⁺ or CD8⁺ cells from donor mice treated with metformin; Control—cells from untreated mice. n = 5–6 mice/group/experiment. d Flow cytometric analysis of splenic CD8⁺ T cells from mice of experiment 1 b. T_M - memory T cells, CD44⁺CD62L⁺; T_E effector T cells, CD44⁺CD62L⁻; T_N - naive T cells, CD44⁻CD62L⁺. e Frequency and number of CD8⁺ T_E, T_M, and T_N cells in the recipient mice of experiment 1. n = 5 mice/group. f Frequency and number of IFNγ⁺, TNF⁺, or Perforin⁺ CD8⁺ T_M cells from e. n = 5 mice/group. g Frequency and number of IFNγ⁺ and/or Perforin⁺ CD45.1⁺CD8⁺ T_M cells in the recipient mice of experiment 2. n = 5 mice per group. Box-and-whisker plots show the median, 5 and 95 percentiles. All data are analyzed using Kruskal–Wallis test with Dunn’s multiple comparison.

Fig. 2. Metformin treatment expands CD8⁺CXCR3⁺ T_M cell population in spleen and lungs of mice.

a Experimental set-up of in vivo experiments. b Visualized t-SNE map of CD8⁺ T cells from spleen of wild-type (WT) mice with automated cluster gating. t-SNE was performed on live CD45⁺CD90⁺TCRαβ⁺CD4⁻CD8⁺ cells after gating out B cells. c Normalized expression intensities of indicated marker was calculated and overlaid on the t-SNE plot, characterizing all clusters of CD8⁺ T cells in spleen. Means ± SD of four mice per group, one-tailed Mann–Whitney U test. d Manual gating analysis of CD8⁺CXCR3⁺ T_M cells in control and metformin-treated WT mice. e, f Flow cytometric analysis of splenic CD8⁺ T-cell subsets from control and metformin-treated WT mice. Pooled data from three independent experiments. Means ± SD of n = 13–15 mice per group, two-tailed Mann–Whitney U test. g Heat map displaying frequency of various surface markers in lung CD8⁺ T cells of control and metformin-treated WT mice. n = 4. *p ≤ 0.05 by two-tailed paired t test. h Flow cytometric analysis of spleen CD8⁺ T_E, T_M, and T_N cells of control and metformin-treated Cxcr3^−/− mice. Data shown one out of two experiments, mean ± SD of five mice per group. i Left—volcano plot showing DEGs from RNA sequencing data of CD8⁺ T cells from Cxcr3^−/− and WT mice. n = 5 mice/group. Right—Canonical pathway enrichment via Ingenuity pathway analysis (IPA) of 202 DEGs. j Heat map displaying the expression level of 35 FOXO1 targets in data shown in i. NS not significant.

Fig. 3. Metformin promotes CD8⁺ T-cell SRC, mitochondrial mass, survival, and FAO.

a Oxygen consumption rate (OCR) and spare respiratory capacity (SRC) in spleen CD8⁺ T cells from control and metformin-treated WT mice in response to different drugs. n = 4 control mice, n = 5 metformin-treated mice. b OCR in spleen CD8⁺ T cells from control and metformin-treated Cxcr3^−/− mice. n = 6 mice/group. c Mitotracker green staining of CD8⁺ T cells from control and metformin-treated WT mice. MFI - median fluorescence intensity. n = 4 mice/group. d TMRM staining of CD8⁺ T cells from control and metformin-treated WT mice. n = 4 mice/group. e Experimental schematic of adoptive transfer of CD8⁺ T cells into congenic WT mice. f Frequency of CD45.1⁺ cells in the spleen and lymph node of recipient mice. n = 4 mice/group. g OCR in spleen CD8⁺ T cells from control and metformin-treated WT mice in response to different drugs under low glucose (LG) condition in the presence and absence of palmitate. n = 5 mice/group. h, i OCR in spleen CD8⁺ T cells from metformin-treated WT mice in response to different drugs under LG condition in the presence of palmitate. h n = 5 mice/group; i n = 8 mice/group. j Glucose uptake (2-NBDG) in CD8⁺ T_M cells, analyzed by flow cytometry. n = 5 mice/group. Data in a, c, d, f, g, and j are presented as mean ± SD, two-tailed Mann–Whitney U test.

Fig. 4. Metformin reprograms gene expression of CD8⁺ T_M cells.

a Schematic depicting sorting of CD8⁺ T_M cells for RNA sequencing and identification of DEGs in metformin-educated CD8⁺ T_M cells. b Molecular function enrichment via Ingenuity pathway analysis (IPA) of DEGs in a. c Heat map displaying the expression level of 86 genes related to cell death and survival. d Flow cytometric analysis of caspase-1 in splenic CD8⁺ T cells from Mtb-infected mice treated or not with metformin. n = 5 mice/group. e Experimental set-up for the transfer of CD8⁺ T_M or T_E cells into TCRβδ^−/− mice followed by Mtb or BCG infection. Three experiments were performed. f, g Flow cytometric analysis of the frequency of CD45.1⁺ lung f and splenic g cells in recipient mice. n = 4 control group, n = 6 metformin group. h Bacterial load in the lung of Mtb-infected TCRβδ^−/− recipient mice. i, j Bacterial load in the lung of BCG-infected TCRβδ^−/− recipient mice. Data of mean ± SD of 4–5 mice per group, Kruskal–Wallis test with Dunn’s multiple correction. Data in d, f, g, and h is presented as mean ± SD, two-tailed Mann–Whitney U test.

Fig. 5. Metformin promotes BCG vaccine immunogenicity and efficacy.

a PPD-specific response of spleen cells, analyzed by flow cytometry. Data on IFNγ producing CD8⁺ T, T_M, and T_E cells. n = 5 mice/group. Box-and-whisker plot show the median, 5 and 95 percentiles, one-way ANOVA with Turkey’s multiple comparison test. b, c Mitotracker green b TMRM c staining of splenic CD8⁺ T cells from control and metformin-treated BCG-vaccinated WT mice. b n = 4 mice/group; c n = 5 mice/group. means ± SD, Mann–Whitney U test. d Changes in gene expression in CD8⁺ T_M cells from BCG-vaccinated mice treated or not with metformin. IPA analysis of 607 DEGs. n = 5 mice/group. e Schematic of Mtb infection of metformin-treated BCG-vaccinated mice. Lung bacillary load at 30 days post infection (p.i) is shown. n = 6 mice/group. f Schematic of Mtb infection of metformin-treated unvaccinated and BCG-vaccinated guinea pigs. g Lung bacillary load in unvaccinated and BCG-vaccinated Mtb-infected guinea pigs treated or not with metformin. n = 6–7 animals/group. h Light micrographs of hematoxylin and eosin (H&E) staining of lung sections from animals in g. Magnification for main image ×40 (scale bars, 50 μm) and insert ×100. i Morphometric analysis of lung sections shown in h indicating the mean lesion lung involvement from each guinea pig. n = 6–7 animals/group. Data in a–c, e is representative of two independent experiments. Guinea pig data in g and i is from one experiment. Data in e, g, and i, means ± SD, Kruskal–Wallis test with Dunn’s multiple correction.

Fig. 6. Metformin treatment expands CD8⁺CXCR3⁺ T_M cell population in humans.

a Strategy of analyzing PBMCs from metformin-treated healthy individuals. b t-SNE map of human peripheral CD8⁺ T cells with manually cluster gating for T_E, T_M, and T_N cells. c Normalized expression intensities of indicated markers were calculated and overlaid on the t-SNE plot, characterizing all cluster of human CD8⁺ T cells. d Manual gating identified differential expression of CD127 and CXCR3, distinguishing two CD8⁺ T_M subpopulations, T_M1 and T_M2. e Frequency of CD8⁺CD127⁺CXCR3⁺ T_M cells before and after metformin treatment. n = 8 subjects, analysis by Wilcoxon matched-pairs signed rank test. f Overlay plot showing expression of BCL2 in CD8⁺CD127⁺CXCR3^- and CD8⁺CD127⁺CXCR3⁺ T_M cells. BCL2 expression in CD8⁺ T_N cells is also indicated. Data shown from one representative donor. g Flow cytometric analysis of frequency of peripheral CD8⁺CXCR3⁺ T_M cells in T2D patients receiving either metformin or other diabetic monotherapy, n = 18 other monotherapy group, n = 42 metformin group. Box-and-whisker plot show the median, 5 and 95 percentiles. Analysis by two-tailed t test with Welch’s correction.