Figures and data
Mtb rewires the metabolic network of monocyte-derived DCs (Mo-DCs).
Mo-DCs were stimulated with viable or irradiated Mtb (iMtb) at two MOI (1 or 2 Mtb per DC) for 24 h. Glycolysis was measured as: (A) Lactate release in culture supernatants; (B) Glucose uptake measured in culture supernatants; (C) Relative expression of HIF-1α mRNA normalized to EeF1A1 control gene. (D) Representative histograms of the mean fluorescence intensity (MFI) of HIF-1α as measured by flow cytometry. Quantification shown in graph to the right. (E) Relative expression of lactate dehydrogenase A (LDHA) mRNA normalized to EeF1A1 control gene. (F) FACS plots show the percentage of Glut1+ cells with and without iMtb stimulation or infected with viable Mtb in a representative experiment. Quantification of Glut1+ cells plotted below. (G) MFI of Mitospy probe as a measurement of mitochondrial mass for Mo-DCs treated (or not) with iMtb (upper panel) or infected with viable Mtb (lower panel). The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown. (H) Representative electron microscopy micrographs of control and iMtb-stimulated DCs showing mitochondria colored in cyan (left panels) and quantified morphometric analysis (right panels). Statistical significance was assessed in (A-E) using 2-way ANOVA followed by Tukey’s multiple comparisons test (∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001), and in (F-H) using paired t test (∗p < 0.05) for iMtb versus controls. All values are expressed as means ± SEM.
Mtb skews DC metabolism toward glycolysis.
Mo-DCs were stimulated with irradiated Mtb (iMtb) or infected with Mtb expressing Red Fluorescent Protein (Mtb-RFP, panel C). (A) Representative histograms showing the translation level after puromycin (Puro) incorporation and staining with a monoclonal anti-Puro (anti-Puro MFI) in response to inhibitor treatment (C, Control; DG, 2-Deoxy-D-Glucose; Oligomycin, O; or combination treatment, DG+O). The bar plots show the values of the anti-Puro MFI from 6 donors. Arrows and numbers inside boxes denote the differences between the MFI of puro in the different treatments that are used to calculate the glucose dependence (1) and fatty acids and amino acids oxidation (FAO & AAO) capacity (4); and the mitochondrial dependency (2) and glycolytic capacity (3). (B) Relative contributions of glycolytic and FAO & AAO capacities and glucose and mitochondrial dependences to overall DC metabolism analyzed with SCENITH. (C) DCs were infected with Mtb-RFP for 24 h, thereafter the metabolic profile was evaluated by SCENITH. Representative plots showing the gating strategy to distinguish the populations within Mtb-infected cultures, which includes RFP+ (Mtb-infected DCs) and RFP- (bystander DCs) cells. Representative histograms showing the translation level after Puro incorporation are shown for uninfected, Mtb-infected and bystander DCs. The bar plots show the values of the anti-Puro MFI from 4 donors. Right panels show the relative contributions of glycolytic and FAO & AAO capacities and glucose and mitochondrial dependences to DC metabolism. (D) Kinetic profile of proton efflux rate (PER; lower panel) and oxygen consumption rate (OCR; upper panel) measurements in control and iMtb-stimulated DCs in response to inhibitor treatments (Oligomycin, O; ROT/AA, Rotenone/Antimycin A), obtained using an Agilent Seahorse XFe24 Analyzer. PER and OCR measurements were normalized to the area covered by cells. (E) ATP production rate from mitochondrial oxidative phosphorylation (MitoATP) and glycolysis (glycoATP). MitoATP production rate and glycoATP production rate were calculated from OCR and ECAR measurements in control and iMtb-stimulated DCs. (F) Percentages of MitoATP and GlycoATP relative to overall ATP production. Statistics in (B, E-F) are from paired t test (∗p < 0.05; ∗∗p < 0.01) for iMtb versus controls. Statistics in (C) are 2-way ANOVA followed by Tukey’s multiple comparisons test (∗p < 0.05) as depicted by lines. The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown.
Mtb triggers glycolysis through TLR2 ligation in Mo-DCs.
Mo-DCs were stimulated with irradiated Mtb (iMtb) in the presence of neutralizing antibodies against either TLR2 (aTLR2), TLR4 (aTLR4), or their respective isotype controls. (A) Lactate release as measured in supernatant. (B) Glucose uptake as measured in supernatant. (C) Mean fluorescence intensity (MFI) of HIF-1α as measured by flow cytometry. (D) Kinetic profile of proton efflux rate (PER) and oxygen consumption rate (OCR) measurements (left panels). Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production (right panel). (E-F) Mo-DCs were stimulated with Pam3Cys or Mtb peptidoglycan (PTG) at the indicated concentrations. (E) Lactate release as measured in supernatant. (F) Glucose uptake as measured in supernatant. Statistics in (A-B, E-F) are 2-way ANOVA followed by Tukey’s multiple comparisons test (∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001). Statistics in (C-D) are from paired t test (∗p < 0.05) for iMtb versus controls. The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown.
HIF-1α is required for DC maturation upon iMtb stimulation but not for CD4+ T lymphocyte polarization.
(A-C) Mo-DCs were stimulated with irradiated Mtb (iMtb) in the presence or absence of the HIF-1α inhibitor PX-478 (PX). (A) Metabolic flux analysis showing quantification of mitochondrial ATP production and glycolytic ATP production, as in Figure 2C. (B) Mean fluorescence intensity (MFI) of CD83, CD86 and PD-L1 as measured by flow cytometry. (C) TNF-α and IL-10 production by Mo-DCs measured by ELISA. (D-E) Monocytes from PPD+ healthy donors were differentiated towards DCs, challenged or not with iMtb in the presence or absence of PX for 24 h, washed, and co-cultured with autologous CD4+ T cells for 5 days. (D) Extracellular secretion of IFN-γ and IL-17 as measured by ELISA. (E) Absolute abundance of Th1, Th17, Th2 and Th1/Th17 CD4+ T cells after coculture with DCs. When indicated lymphocytes without DCs were cultured (Ly). Statistical significance based on 2-way ANOVA followed by Tukey’s multiple comparison test (∗p < 0.05; ∗∗p < 0.01). The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown.
HIF1α-mediated-glycolysis is required to trigger migratory activity in iMtb-stimulated DCs.
Mo-DCs were treated (or not) with HIF-1a inhibitor PX-478 (PX) or glycolysis inhibitor oxamate (OX) and stimulated with iMtb for 24 h. (A) Lactate release as measured in supernatants in DCs stimulated or not with iMtb in the presence of OX. (B) Percentage of migrated cells towards CCL21 relative to the number of initial cells per condition. (C) Three-dimensional amoeboid migration of DCs through a collagen matrix after 24 h. Cells within the matrix were fixed and stained with DAPI. Images of the membrane of each insert were taken and the percentage of cells per field were counted. The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown. (D) Representative schematic of the experimental setup for in vivo migration assays. (E) Percentages of migrating BMDCs (CFSE-labeled among CD11c+) recovered from inguinal lymph nodes. Statistical significance assessed by (A-B) ANOVA followed by Dunnett’s multiple comparisons test (∗p < 0.05; ∗∗p < 0.01); (C) Nested ANOVA followed by Dunnett’s multiple comparisons test (∗p < 0.05; ∗∗p < 0.01); (E) ANOVA followed by Holm-Sidak’s multiple comparisons test (∗p < 0.05).
Stabilization of HIF-1α promotes migration of tolerogenic DCs and Mo-DCs from TB patients.
Tolerogenic Mo-DCs were generated by dexamethasone (Dx) treatment and were stimulated (or not) with iMtb in the presence or absence of HIF-1α activator DMOG. (A) Lactate release and glucose uptake as measured in supernatant. (B) Mean fluorescence intensity (MFI) of HIF-1α. Representative histograms and quantification are shown. (C) Three-dimensional amoeboid migration of DCs through a collagen matrix. After 24 h of migration, images of stacks within the matrix were taken every 30 µm. Percentage of migrating cells were defined as cells in the stacks within the matrix relative to total number of cells. (D) Chemotactic activity towards CCL21 in vitro. (E-H) Mo-DCs were generated from healthy subjects (HS) or TB patients, and DCs were stimulated (or not) with iMtb. (E) Chemotaxis index towards CCL21 (relative to unstimulated DCs). (F) Lactate production ratio relative to unstimulated DCs. (G) Glycolytic capacity assessed by SCENITH. (H) Chemotactic activity towards CCL21 of Mo-DCs from TB patients stimulated with iMtb and treated or not with DMOG. Statistical significance assessed by (A-D) 2-way ANOVA followed by Tukey’s multiple comparisons test (∗p < 0.05; ∗∗p < 0.01); (E-G) Unpaired T test (∗p < 0.05); (H) Paired T test (∗p < 0.05). The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown.
CD16+ monocytes from TB patients show increased glycolytic capacity.
(A) Monocytes from TB patients or healthy subjects (HS) were isolated and cultured with IL-4 and GM-CSF for 24 h. Accumulation of lactate in culture supernatants were measured at 1 and 24 h of differentiation. (B) Glycolytic capacity measured by SCENITH of monocyte subsets as defined by their CD14 and CD16 expression from HS and TB patients. (C) Correlation analysis between the baseline glycolytic capacity and the evolution time of TB symptoms for each monocyte subset (CD14+CD16-, CD14+CD16+ and CD14dimCD16+, n = 14). Linear regression lines are shown. Spearman’s rank test. The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown. (D) BubbleMap analysis, a high-throughput extension of GSEA, on the pairwise comparisons of monocytes from healthy patients (HS) or donors with latent TB (LTB) vs patients with active TB (TB), for each monocyte subset (CD14+CD16-, CD14+CD16+ and CD14dimCD16+). The gene sets shown come from the Hallmark (H.) collection of the Molecular Signature Database (MSigDB). Statistical significance was assessed by (A) Paired T test for 0 vs. 24h (∗p < 0.05) and 2-way ANOVA for HS vs. TB at each time (∗∗p < 0.01); (B) unpaired T test (∗p < 0.05; ∗∗p < 0.01).
HIF1-α activation in CD16+ monocytes from TB patients leads to DCs with poor migration capacity.
(A) Ex-vivo determination of HIF-1α expression by monocytes from healthy subjects (HS) or TB patients (TB) for each monocyte subset (CD14+CD16-, CD14+CD16+ and CD14dimCD16+). (B-C) Monocytes from HS were treated with DMOG during the first 24 h of differentiation with IL-4/GM-CSF (earlyDMOG) and removed afterwards. On day 6 of differentiation, cells were stimulated (or not) with iMtb. (B) Monocyte lactate release after 24 h of DMOG addition. (C) Chemotactic activity towards CCL21 of DCs. Statistical significance was assessed by (B) paired T test (∗∗p < 0.01); (C) 2-way ANOVA followed by Tukey’s multiple comparisons test (∗p < 0.05). The data are represented as scatter plots with each circle representing a single individual, means ± SEM are shown.
