The AMPK-PPARGC1A pathway is required for antimicrobial host defense through activation of autophagy

Abstract

AMP-activated protein kinase (AMPK) is a crucial energy sensor and plays a key role in integration of cellular functions to maintain homeostasis. Despite this, it is largely unknown whether targeting the AMPK pathway can be used as a therapeutic strategy for infectious diseases. Herein, we show that AMPK activation robustly induces antibacterial autophagy, which contributes to antimicrobial defense against Mycobacterium tuberculosis (Mtb). AMPK activation led to inhibition of Mtb-induced phosphorylation of the mechanistic target of rapamycin (MTOR) in macrophages. In addition, AMPK activation increased the genes involved in oxidative phosphorylation, mitochondrial ATP production, and biogenesis in Mtb-infected macrophages. Notably, peroxisome proliferator-activated receptor-gamma, coactivator 1α (PPARGC1A) was required for AMPK-mediated antimicrobial activity, as well as enhancement of mitochondrial function and biogenesis, in macrophages. Further, the AMPK-PPARGC1A pathway was involved in the upregulation of multiple autophagy-related genes via CCAAT/enhancer binding protein (C/EBP), β (CEBPB). PPARGC1A knockdown inhibited the AMPK-mediated induction of autophagy and impaired the fusion of phagosomes with MAP1LC3B (LC3B) autophagosomes in Mtb-infected macrophages. The link between autophagy, mitochondrial function, and antimicrobial activity was further demonstrated by studying LysMCre-mediated knockout of atg7, demonstrating mitochondrial ultrastructural defects and dysfunction, as well as blockade of antimicrobial activity against mycobacteria. Collectively, our results identify the AMPK-PPARGC1A axis as contributing to autophagy activation leading to an antimicrobial response, as a novel host defense mechanism.

Keywords: AICAR; AMP-activated protein kinase; MTOR; Mycobacterium tuberculosis; PPARGC1A; autophagy.

Figure 1. AICAR-induced autophagy activation leads to phagosomal maturation and antimicrobial responses against mycobacterial infection. (A and B) BMDMs were infected with Mtb (MOI = 5) for 4 h, incubated with or without AICAR (0.1, 0.5, 1 mM for A, 0.5 mM for B) for 16 h. (A) Cells were harvested, followed by immunoblotting (IB) with anti-LC3, anti-SQSTM1, and anti-ACTIN. (B) Cells were fixed and immunolabeled with the anti-LC3B antibody, followed by staining with Alexa 488-conjugated goat anti-rabbit IgG. The expression of LC3B was analyzed by flow cytometry. (Left) A representative histograms of 7 independent replicates. (Right) Average mean fluorescence intensities (MFIs). (C and D) BMDMs were infected with ERFP-Mtb (MOI = 10) for 4 h, incubated with or without AICAR for 24 h, then fixed and stained with αLC3 (Alexa 488) to detect autophagosomes and DAPI. (C) Confocal z-stack images were obtained and reconstituted to 3 dimensions using the Imaris software for Mtb-infected/AICAR-treated conditions. Insets, enlargement of outlined areas. Scale bars: 5 μm. See also the corresponding Video 1. (D) Quantitative data of colocalization analyses (%). Colocalization analyses between the autophagosomal marker LC3 (green) and the bacterial phagosome containing ERFP-Mtb (red). See also the corresponding Video 1. (E) BMDMs were infected with Mtb, treated with AICAR, and visualized by transmission electron microscopy (TEM). Scale bar: 1 µm. Magnified view of corresponding boxed areas in the right panels (scale bar: 200 nm). (Left) Representative TEM images from 3 independent experiments. (Right) Quantification of 200 internalized mycobacteria per experimental condition (mean ± SD). (F–H) Intracellular survival of Mtb assessed by CFU assay. (F) BMDMs were infected with Mtb for 4 h, and then incubated with AICAR for 3 d. (G) BMDMs were infected with Mtb, followed by treatment with 3-MA (10 μM; 2 h), wortmannin (100 nM), chloroquine (10 μM), or bafilomycin A₁ (100 nM), and then incubated with AICAR (0.5 mM) for 3 d. (H) THP-1 cells were transduced with lentivirus expressing nonspecific shRNA (sh-NS) or shRNA specific for BECN1 (sh-BECN1), ATG5 (sh-ATG5), ATG7 (sh-ATG7) with polybrene (8 μg/mL). After 3 d, THP-1 cells were infected with Mtb, treated with AICAR, and then lysed to determine intracellular bacterial loads. **P < 0.01 and ***P < 0.001 (2-tailed Student t test), are indicated. All data above represent the means ± SD from 3 experiments. SC, solvent control. CFU, colony-forming units.

Figure 2. AICAR-induced AMPK activation is required for MTOR inhibition, autophagy activation, and antimicrobial activity in Mtb-infected BMDMs. (A–D) AMPK activation and MTOR inhibition by AICAR in BMDMs. BMDMs were treated with AICAR (0.5 mM, for A) or Mtb (MOI = 5, for B) for the times indicated. BMDMs were infected with Mtb for 4 h, incubated with or without AICAR (0.1, 0.5, 1 mM) for 1 h (C and D), followed by IB to detect phosphorylated forms of AMPKA, ACACA/B, and ACTIN (A–C) or phosphorylated and total forms of MTOR (B and D). (E) BMDMs pretreated with AMPK inhibitor compound C (Comp C, 10 μM for 1 h) were infected with Mtb (MOI = 10) for 4 h, and then incubated with AICAR for 24 h. The cells were fixed and labeled with Alexa 488-conjugated anti-LC3 and DAPI. Quantitative analysis of LC3+ puncta per cell. (F–H) Intracellular survival of Mtb assessed by CFU assay. (F and H) BMDMs were infected with Mtb, followed by treatment with Comp C (1 h), and then incubated with (F; AICAR, 0.5 mM) or without AICAR (H) for 3 d. (G) BMDMs were transduced with lentivirus expressing nonspecific shRNA (sh-Ns) or shRNA specific for AMPKa (sh-Ampka) with polybrene (8 μg/mL). After 3 d, BMDMs were infected with Mtb, treated with AICAR, and then lysed to determine intracellular bacterial loads. (Top) Semiquantitative RT-PCR was used to determine the efficiency of lentiviral transduction. All data represent the means ± SD from 3 experiments. *P < 0.05, **P < 0.01, and ***P < 0.001 (2-tailed Student t test), compared with solvent control (F and H) or AICAR-treated condition (G). SC, solvent control. CFU, colony-forming units. ns, not significant.

Figure 3. AICAR treatment increases the induction of mitochondrial OXPHOS gene expression, ATP synthesis, and mitochondrial mass in Mtb-infected macrophages. (A and B) AICAR induces upregulation of OXPHOS genes. BMDMs were infected with Mtb (MOI = 5) for 4 h, incubated with AICAR (0.5 mM) for 6 h (A) or the times indicated (B), and subjected to quantitative real-time PCR of OXPHOS genes (A) or Intracellular ATP was quantified by a fluorometric method (B). (C–E) BMDMs were infected with Mtb for 4 h, incubated with AICAR for 24 h. (C) Mitotracker fluorescence signals assessed by a flow cytometric analysis. (Left) Representative histograms from 7 independent replicates. (Right) Bar graph indicates the mitochondrial mass MFIs. (D) Mitochondrial marker TOMM20 signals were assessed using a confocal microscope (60×). Magnified view of corresponding boxed areas in the right panels. Scale bar: 5 µm. (E) Transmission electron microscopy (TEM). (Top) Representative TEM images from 3 independent experiments are shown. (Bottom) Bar graph indicates the mitochondrial number. Scale bar: 1 µm. Data shown are from one representative of at least 3 independent experiments (means ± SD of triplicate [A–C right, and E bottom] samples). *P < 0.05, **P < 0.01, and ***P < 0.001 (2-tailed Student t test), compared with AICAR-treated condition (A, C, and E). U, untreated control.

Figure 4. AMPK-induced PPARGC1A is required for mitochondrial biogenesis and ATP synthesis in Mtb-infected BMDMs. (A) AICAR induces upregulation of OXPHOS proteins. BMDMs were infected with Mtb (MOI = 5), incubated with AICAR (0.5 mM) for the indicated times, and subjected to IB analysis with antibodies to PPARGC1A, NFE2L1, TFAM, CYCS, and ACTIN. (B) BMDMs were transduced with lentivirus expressing nonspecific shRNA (sh-Ns) or shRNA specific for PPARGC1A (sh-Ppargc1a) with polybrene (8 μg/mL). After 3 d, BMDMs were infected with Mtb (MOI = 5) for 4 h, incubated with AICAR (0.5 mM) for 6 h, and subjected to quantitative real-time PCR to determine mRNA expression of Ppargc1a, Nfe2l1, Nfe2l2, and Tfam. (C–E) The experimental conditions were as outlined in (B). BMDMs were infected with Mtb for 4 h, incubated with AICAR for 16 h. (C) MitoTracker fluorescence signals were assessed by FACS analysis. (Top) Representative histograms of 7 independent replicates. (Bottom) Bar graph indicating the mitochondrial mass as MFIs. (D) Cellular ATP was quantified fluorometrically using an ATP bioluminiscent assay kit. (E) mtDNA content in BMDMs measured by quantitative real-time PCR. The mtDNA content was normalized to nuclear DNA. Data shown are from one representative of at least 3 independent experiments (means ± SD from triplicate (B, C bottom, D, E) samples). *P < 0.05 and **P < 0.01 (2-tailed Student t test), compared with sh-Ns (B–E). U, untreated control. ns, no significant.

Figure 5. PPARGC1A is required for antimicrobial responses against mycobacteria in vitro and in vivo. (A) At 3 d post-transduction with lentivirus-sh-NS or lentivirus-sh-Ppargc1a, BMDMs were infected with Mtb (MOI = 1 or 10) for 4 h and then incubated with AICAR and then lysed to determine intracellular bacterial loads. (Right) Semiquantitative RT-PCR analyses show the efficiencies of lentiviral transduction. (B) (Left) BMDMs were infected with ERFP-Mtb (MOI = 10) for 4 h, incubated with or without AICAR (0.5 mM) for 24 h, then fixed and stained with anti-LC3 (Alexa 488) to detect autophagosomes and DAPI. Scale bars: 5 μm. (Right) Quantitative data of colocalization analyses (%). Colocalization analyses between the autophagosomal marker LC3 (green) and the bacterial phagosome containing ERFP-Mtb (red). (C) Control and Ppargc1 mutant flies were infected with M. marinum (CFU = 500) and then treated with or without AICAR (1 mM). Dead flies were counted at 1-d intervals. The error bars indicate 95% confidence intervals. Log-rank analysis of the survival curves indicated that each condition (n = 120) was significantly different (***P < 0.001). (D) After 3 d, each group of flies was harvested, homogenized, and quantified by CFU assay. The experiments were repeated 3 times. Data shown are from one representative of at least 3 independent experiments (means ± SD from triplicate (A, B right, D) samples). ***P < 0.001 (2-tailed Student t test), compared with sh-Ns (B). SC, solvent control. CFU, colony-forming units. ns, not significant.

Figure 6. The AMPK-PPARGC1A pathway is required for upregulation of multiple autophagy-related genes in macrophages via CEBPB. (A) BMDMs were transduced with lentiviruses expressing nonspecific shRNA (sh-NS) or specific shRNA for PPARGC1A (sh-Ppargc1a). After 3 d, BMDMs were infected with Mtb (MOI = 5) for 4 h and then incubated with AICAR (0.5 mM) for 6 h, followed by RT-PCR analyses of Atg5, Becn1, Lc3b, Ppargc1a, and Gapdh. (B) Raw264.7 cells were transduced with lentiviruses expressing sh-NS or sh-Ppargc1a, followed by transfection with a pGL3 basic, pGL3-ATG5-1P, or pGL3-ATG5-3P (Top), a pGL3 basic, pGL3-BECN1-1P, or pGL3- BECN1-3P (Middle) or a pGL3 basic, pGL3-ATG7-1P, or pGL3-ATG7-2P (Bottom) luciferase reporter construct. After 48 h, cells were infected with Mtb (MOI = 5) for 4 h, incubated with AICAR (0.5 mM) for 6 h and then a luciferase assay was performed. The luciferase activity was measured and normalized to Renilla luciferase enzyme activities. (C) Schematic representation of the human ATG5 (Top), BECN1 (Middle), and ATG7 (Bottom) promoter-reporter construct generation. Potential CEBPB binding sites located in the 5′-flanking region of genes ATG5, BECN1, and ATG7 are shown as numbered boxes. Numbers in parentheses indicate the number of residues matching the consensus. n = A, T, G, or C; K = G or T; *, antisense orientation. (D and E) The experimental conditions were as outlined in (B). (D) THP-1 cells were cotransfected with a scrambled control siRNA (si-NS) or si-CEBPB, together with a pGL3 basic, pGL3-ATG5-1P, or pGL3-ATG5-3P (Top), a pGL3 basic, pGL3-BECN1-1P, or pGL3-BECN1-3P (Middle) or a pGL3 basic, pGL3-ATG7-1P, or pGL3-ATG7-2P (Bottom) luciferase reporter construct (E) Raw264.7 cells were transduced with lentiviruses expressing sh-NS or sh-Ppargc1a, followed by transfection with a pGL3 basic or pGL3-CEBPB luciferase reporter construct. The promoter activities were determined by luciferase assay. (F and G) The experimental conditions were as outlined in (A). BMDMs were infected with Mtb (MOI = 5) for 4 h, incubated with AICAR (0.5 mM) for 16 h. (F) Cells were harvested followed by IB with anti-LC3, anti-PPARGC1A, and anti-ACTIN. Data shown are from 1 representative of at least 3 independent experiments (means ± SD of triplicate (B, D, and E) samples). *P < 0.05, **P < 0.01, and ***P < 0.001 (2-tailed Student t test), compared with sh-NS (B, D, and E). U, untreated control. ns, not significant.

Figure 7. Defective autophagy in macrophages leads to mitochondrial dysfunction and increased intracellular growth of mycobacteria. (A to F) Atg7^fl/fl LysM-Cre+ and Atg7^fl/fl LysM-Cre- BMDMs were infected with Mtb (MOI = 5) for 4 h and then treated with AICAR (0.5 mM) for 24 h (A–C, E and F) or 6 h (D). (A) Cells were harvested and followed by IB with anti-LC3, anti-SQSTM1, anti-ATG7, and anti-ACTIN. (B) Cells were infected with Mtb, incubated with AICAR, and then cellular ATP was quantified fluorometrically using an ATP bioluminescent assay kit. (C) Mitochondrial membrane potential (ΔΨm) was measured using a potentiometric fluorescent probe (TMRE), followed by flow cytometric analysis. (D) mtDNA content in BMDMs measured by quantitative real-time PCR. The mtDNA content was normalized to nuclear DNA. (E) Transmission electron microscopy (TEM) analysis for Atg7^fl/fl LysM-Cre+ BMDMs infected with Mtb (MOI = 10), followed by treatment with AICAR. Scale bar: 1 µm. (F) Quantitative data of colocalization analyses (%). Colocalization analyses between the autophagosomal marker LC3 (green) and the bacterial phagosome containing ERFP-Mtb (red). (G and H) Intracellular survival of Mtb (G) or BCG (H) assessed by CFU assay. Atg7^fl/fl LysM-Cre+ and Atg7^fl/fl LysM-Cre- BMDMs were infected with Mtb (G) or BCG (H) for 4 h, incubated with AICAR (0.5 mM), and then lysed to determine intracellular bacterial loads. (I) The mice were infected with BCG (1 × 10⁷ CFU/mouse, i.v.) for 3 wks and then AICAR (500 mg/kg, i.p.) treated for 3 consecutive days. The mice were sacrificed 4 wks after BCG infection and bacterial loads of infected mice (n = 5 per group) in spleen were determined by CFU assay. Data shown are from one representative of at least 3 independent experiments (means ± SD of triplicate (B, D, F–I) samples). *P < 0.05, **P < 0.01, and ***P < 0.001 (2-tailed Student t test), compared with Atg7^fl/fl LysM-Cre- (B and D) or AICAR-treated condition (G and H). U, untreated control. SC, solvent control. CFU, colony-forming units. ns, not significant.