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

AMPK activation by AICAR treatment enhances autophagy, leading to phagosomal maturation and antimicrobial response against mycobacterial infection

To identify a potential antimicrobial agent against intracellular Mtb based on autophagy activation, we first examined several AMPK-activating agents that enhanced antibacterial autophagy and antimicrobial responses (data not shown). In our search for autophagy-activating agents through AMPK activation, we found that the AMP-mimetic 5-aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR)^16,17 robustly activated autophagy in murine bone marrow-derived macrophages (BMDMs), as determined by immunoblotting of autophagosomal membrane-associated LC3-II fractions (Fig. S1A). In addition, AICAR treatment upregulated the lipidated form of LC3, but downregulated the autophagy substrate SQSTM1/p62, in uninfected and in Mtb-infected BMDMs in a dose-dependent manner (Fig. 1A). Flow cytometric analysis using anti-LC3B-specific antibody¹¹ showed a significant increase in autophagosome formation in AICAR-treated, Mtb-infected BMDMs (Fig. 1B). It was noted that AICAR-induced LC3B and LC3 puncta formation was greatly increased in Mtb-infected BMDMs, when compared with the uninfected condition (Fig. 1B; Fig. S1B).

We next monitored autophagic flux in macrophages using different assays.¹⁸ The formation of endogenous LC3 punctate structures was significantly abrogated by pretreatment with 3-methyladenine (3-MA), an inhibitor of autophagy (Fig. S1B), and increased by the vacuolar H⁺-ATPase inhibitor, bafilomycin A₁ (Fig. S1B). Additionally, pretreatment of BMDMs with bafilomycin A₁ or chloroquine markedly increased AICAR-induced LC3-II fractions and accumulation of SQSTM1 in macrophages (Fig. S1A). Moreover, the long-lived protein degradation assay showed that the AICAR-induced proteolysis rate of long-lived proteins was increased in BMDMs in a time-dependent manner, and was suppressed by bafilomycin A₁ or chloroquine, both inhibitors of lysosomal acidification (Fig. S1C). We further measured mRFP-GFP-LC3 delivery to lysosomes in AICAR-treated cells. As shown in Figure S1D, both yellow and red puncta were increased in AICAR-treated BMDMs; however, only yellow puncta increased upon pretreatment of the cells with bafilomycin A₁, suggesting that autophagosome maturation into autolysosomes was blocked by this inhibitor. Together, these data indicate that AICAR treatment enhances autophagosome maturation and autophagic flux in BMDMs.

To further evaluate phagosomal maturation by AICAR treatment, we infected BMDMs with ERFP-Mtb, followed by treatment with or without AICAR. Notably, AICAR treatment of Mtb-infected BMDMs resulted in an increased colocalization of the autophagosomal marker LC3 and phagosomes containing labeled Mtb (Fig. 1C and D; Video 1). Ultrastructural analysis of AICAR-treated, Mtb-infected BMDMs revealed bacilli within characteristic double-membrane autophagosomes (Fig. 1E). We next assessed the role of autophagy in antimicrobial responses in macrophages. Treatment of Mtb-infected BMDMs with AICAR inhibited the survival of intracellular bacilli in a dose-dependent manner (Fig. 1G), and this was reversed by treatment with pharmacologic inhibitors of autophagy pathways (Fig. 1G) or by transduction with lentiviral shRNA specific to autophagy-related genes (Atgs) including ATG5, ATG7, and BECN1 (Fig. 1H). These data indicate that AICAR induces autophagy-dependent phagosomal maturation and antimicrobial responses against Mtb in macrophages.

AICAR-induced AMPK activation is required for MTOR inhibition, autophagy activation, and antimicrobial activity in macrophages infected with Mtb

As AICAR induces autophagy and antimicrobial responses, we sought to determine whether AMPK activation by AICAR involves the inhibition of Mtb-mediated MTOR activation. As AMPK activity is regulated by reversible phosphorylation of Thr172 of the catalytic α-subunit of AMPK,^7,19 we first measured AICAR-induced phosphorylation of AMPK at Thr172 and the downstream target, acetyl-CoA carboxylase (ACACA/B) (Fig. 2A). Importantly, Mtb infection of BMDMs also induced a much less, but significant, phosphorylation of AMPK and ACACA/B, than AICAR-treated cells, in a time-dependent manner (Fig. 2B). It was noted that AICAR treatment markedly amplified the AMPK and ACACA/B phosphorylation in Mtb-infected conditions in a dose-dependent manner (Fig. 2C). We previously showed that Mtb infection of human macrophages increases activation of the MTOR-RPS6KB1/S6K1 (ribosomal protein S6kinase, 70 kDa, polypeptide 1) pathway.²⁰ Therefore, we further examined the effects of AICAR on MTOR phosphorylation in BMDMs infected with Mtb. Infection of macrophages with Mtb rapidly increased the phosphorylation of MTOR at Ser-2448 within 15 to 30 min, which then substantially decreased until 8 h, and increased again at later time points (after 16 h) (Fig. 2B). In Mtb-infcted cells at the early (30 min) or late (16 h) time points, AICAR treatment decreased the Mtb-induced phosphorylation of MTOR in a dose-dependent manner (Fig. 2D).

ULK1, the upstream kinase of autophagy machinery, has been known to be regulated via differential phosphorylation by AMPK and MTOR.²¹ We further investigated AICAR-induced ULK1 phosphorylation at Ser317 and Ser555 (activating), and at Ser757 (inhibitory)²² in BMDMs. AICAR treatment induced robust ULK1 phosphorylation at Ser317 and Ser555, the levels of which were significantly amplified in BMDMs infected with Mtb (Fig. S2). However, Mtb infection led to a significantly lower induction of ULK1 phosphorylation at Ser317 and Ser555 in BMDMs, whereas it strongly induced ULK1 Ser757 phosphorylation (Fig. S2). In addition, Mtb-induced ULK1 Ser757 phosphorylation was significantly inhibited by AICAR or rapamycin treatment (Fig. S2 and data not shown), confirming the dependence of ULK1 Ser 757 phosphorylation on MTOR activity.

We further investigated whether AICAR-induced autophagy activation and antimicrobial responses were mediated through AMPK pathway activation. We observed that AICAR-induced LC3 puncta formation was significantly inhibited in Mtb-infected BMDMs at 1 h by pretreatment with compound C, a pharmacological inhibitor of the AMPK pathway (Fig. 2E). We then examined whether the AICAR-induced antimicrobial activity was dependent on AMPK activation. Blockade of AMPK activity by compound C significantly decreased AICAR-mediated antimicrobial responses in a dose-dependent manner (Fig. 2F). Additionally, AICAR-induced antimicrobial responses against Mtb were significantly downregulated in cells transduced with shAmpka (Fig. 2G). Pretreatment of BMDMs with compound C considerably increased the intracellular survival of Mtb even without AICAR treatment (Fig. 2H), suggesting that Mtb-induced AMPK activation is required for the basal host defense against Mtb in macrophages. However, Mtb-induced AMPK activation appeared to be transient and insufficient to inhibit MTOR activation during Mtb infection. Thus, our data indicate that treatment of macrophages with an AMPK-activating agent may overcome MTOR activation during Mtb infection, leading to antimicrobial activity.

AICAR treatment increases the induction of mitochondrial oxidative phosphorylation (OXPHOS) genes, ATP, and mass in Mtb-infected macrophages

AICAR treatment increases mitochondrial density and oxidative machinery in various tissues, including white adipose tissue.²³ However, in macrophages, it is not known whether AICAR treatment can control mitochondrial function, density, and morphology, especially during infection. To address the role of AICAR in the activation of mitochondrial function, we first examined the mRNA and protein expressions of macrophage respiratory chain complexes I–V, which were all significantly elevated in Mtb-infected, AICAR-treated BMDMs, compared with untreated or other conditions (Fig. 3A; Fig. S3A). Notably, 10 of these 12 genes (except SDHB and ATP5G3) were significantly upregulated by AICAR treatment in macrophages infected with Mtb, compared with other conditions (Fig. 3A). We next assessed mitochondrial membrane potential (ΔΨm) in BMDMs using a potentiometric fluorescent probe (TMRE) and flow cytometry. The Mtb-infected, AICAR-treated ΔΨm was significantly higher than those of cells under other conditions (Fig. S3B). In addition, ATP levels were significantly increased by AICAR treatment in Mtb-infected cells, compared with those under other conditions (Fig. 3B).

AICAR treatment significantly increased the mitochondrial content of Mtb-infected BMDMs (Fig. 3C) and HeLa cells (data not shown). Moreover, morphologic analysis revealed that AICAR treatment of Mtb-infected BMDMs led to a substantial increase in the proportion of elongated mitochondria, compared with those of untreated or Mtb-infected BMDMs (Fig. 3D). As shown in Figure 3E, ultrastructural imaging showed that few mitochondria were present in resting, untreated BMDMs. However, Mtb infection caused an increase in the number of mitochondria, most of which were swollen. AICAR treatment resulted in an apparent increase in mitochondrial number when compared with untreated controls. Notably, Mtb-infected, AICAR-treated BMDMs showed a greater increases in number in the number of mitochondria together with qualitative improvements (Fig. 3E). Consistent with this, we also observed that mitochondrial DNA content was increased in Mtb-infected BMDMs after AICAR treatment (Fig. S3C). Thus, these data indicate that AICAR treatment enhances expression of the genes involved in mitochondrial OXPHOS and ATP synthesis as well as mitochondrial mass in macrophages infected with Mtb.

AMPK activation leads to the induction of PPARGC1A, which is required for mitochondrial biogenesis and ATP synthesis in Mtb-infected macrophages

Previous studies have demonstrated that AMPK-mediated PPARGC1A induction plays an important role in the coordinate regulation of mitochondrial gene expression and glucose metabolism in a variety of tissues.^24,25 However, little is known about the profiles of PPARGC1A and other genes involved in mitochondrial biogenesis in macrophages. We next investigated whether AICAR treatment of BMDMs leads to the induction of PPARGC1A and genes regulating mitochondrial biogenesis. Both Mtb infection and AICAR treatment induced substantially more PPARGC1A mRNA than observed in untreated, resting BMDMs (Fig. 4A and B). Notably, AICAR treatment of Mtb-infected BMDMs robustly upregulated the mRNAs and proteins involved in mitochondrial biogenesis, such as PPARGC1A, NFE2L (nuclear factor, erythroid 2-like) 1, NFE2L2, TFAM, and CYCS (cytochrome C, somatic) (Fig. 4A and B). In line with the role of AMPK in the upregulation of PPARGC1A,²⁶ we further examined whether AMPK signaling plays a role in the induction of PPARGC1A in macrophages. Blockade of the AMPK pathway caused a substantial decrease in the Ppargc1a mRNA level (Fig. S4A and B), suggesting that AMPK functions downstream of PPARGC1A expression.

Given that AMPK activation upregulated PPARGC1A in Mtb-infected macrophages, we determined whether PPARGC1A is required for the induction of genes involved in mitochondrial biogenesis and ATP synthesis in Mtb-infected macrophages. BMDMs were transduced with shRNA lentiviral vectors that specifically targeted Ppargc1a. We found that the mRNA levels of Nrf1, Nrf2, and Tfam were significantly depressed in BMDMs upon silencing of Ppargc1a (Fig. 4B). In addition, the markedly increased mitochondrial mass and ATP generation in Mtb/AICAR-treated macrophages were significantly decreased by Ppargc1a-knockdown of cells (Fig. 4C and D). Further, the mitochondrial DNA content was considerably decreased in Ppargc1a-knockdown BMDMs (Fig. 4E). Together, these data suggest that AMPK-PPARGC1A is required for the induction of genes involved in mitochondrial biogenesis and the synthesis of ATP in Mtb-infected macrophages.

PPARGC1A contributes to antimicrobial responses against mycobacteria in vitro and in vivo

It has previously been shown that ATP stimulation promoted killing of virulent Mtb and BCG mycobacteria in macrophages in a P2X7-dependent and -independent manner.^27,28 The finding of the critical role of PPARGC1A in mitochondrial biogenesis and ATP synthesis in Mtb-infected macrophages led us to hypothesize that PPARGC1A contributes to AMPK-induced antimicrobial responses. We thus examined whether PPARGC1A is required for antimicrobial responses against Mtb infection. Remarkably, Ppargc1a-knockdown using shRNA specific for Ppargc1a led to a significant increase of intracellular Mtb in BMDMs, when compared with those transduced with nonspecific shRNA (Fig. 5A). In addition, AICAR-mediated antimicrobial responses were dose-dependently attenuated in Ppargc1a-knockdown BMDMs (Fig. 5A). We further questioned whether the colocalization of Mtb phagosomes and autophagosomes are decreased in BMDMs by knockdown of Ppargc1a. As shown in Figure 5B, AICAR-induced upregulation of Mtb phagosomes and autophagosomes was significantly decreased by silencing of Ppargc1a in BMDMs, suggesting an essential role for PPARGC1A in the phagosomal maturation of Mtb. Importantly, we found that the AICAR-induced increase of LC3 autophagosome formation was also markedly abrogated in Mtb-infected BMDMs by knockdown of Ppargc1a (Fig. 5B).

To study the in vivo innate role of PPARGC1A using Drosophila melanogaster, a genetically tractable model host for Mm,²⁹ we examined whether the Drosophila PPARGC1 family ortholog, Spargel, is required for the expression of multiple genes encoding mitochondrial proteins.³⁰ When control and Spargel mutant flies were injected with Mm and monitored survival for 25 d, Spargel mutant flies succumbed rapidly to Mm challenge, compared with control flies (Fig. 5C). Control flies given food containing AICAR showed significantly increased survival after Mm injection (Fig. 5C). A substantially increased, but not complete, protective effect was observed in AICAR-administered Spargel mutant flies after Mm infection (Fig. 5C). Viable bacterial counts in surviving Spargel mutant flies infected with Mm were consistently higher than those in control flies infected with Mm (after 3 d; n = 20 per group) (Fig. 5D). Notably, AICAR administration to Spargel mutant flies did not decrease bacterial loads to the levels observed in control flies (Fig. 5D). Taken together, our data support a critical function of PPARGC1A in antibacterial host defense against mycobacteria in vitro and in vivo.

AMPK-induced PPARGC1A is required for the upregulation of multiple autophagy-related genes via CEBPB

The key functions of the autophagy pathway are mediated by a number of autophagy-related (ATG) genes that were identified in yeast genetic screens, and found in higher eukaryotes.^31,32 Despite the proposed functions of the ATG proteins, which are involved in the essential steps of autophagosome formation and lysosomal degradation,^31,32 the molecular mechanisms by which the induction of Atg genes are modulated are poorly understood. As we observed that PPARGC1A was required for an AICAR-induced increase in LC3-dependent autophagosome formation, we next determined whether PPARGC1A is involved in autophagy induction through activation of Atg genes in macrophages. We first examined whether AICAR treatment results in the mRNA expression of Atg genes including Atg5, Becn1, Atg7, and Lc3b. Treatment of Mtb-infected BMDMs with AICAR induced significant upregulation of Atg5, Becn1, Atg7, and Lc3b mRNA (Fig. 6A). Importantly, PPARGC1A was required for the induction of mRNA expression (Fig. 6A) and promoter activation of Atg5, Becn1, and Atg7 (Fig. 6B) in Mtb-infected BMDMs treated with AICAR.

Previously, we have shown that vitamin D-dependent BECN1 and ATG5 transcriptional activation was mediated through CEBPB activation.³³ In the promoter regions of ATG5, ATG7, and BECN1, putative CEBPB-binding sites were found (Fig. 6C). Thus, we considered whether the AICAR-induced AMPK-PPARGC1A pathway is involved in the activation of CEBPB to induce Atg genes. Treatment of macrophages with AICAR significantly increased the promoter activity and mRNA levels of autophagy genes including ATG5, BECN1, and ATG7, which were abrogated by knockdown of CEBPB (Fig. 6D and data not shown). Silencing of Ppargc1a significantly decreased CEBPB, but not RELA/NF-κB p65, transcription, as measured by luciferase reporter activity (Fig. 6E and data not shown). Furthermore, the silencing of Ppargc1a markedly reduced LC3-II in all BMDM cultures (Fig. 6F), suggesting an important role for PPARGC1A in autophagy activation. Taken together, these findings suggest a role for the AMPK-PPARGC1A pathway in the activation of autophagy through CEBPB-dependent regulation of autophagy genes.

Defective autophagy in macrophages leads to mitochondrial dysfunction and inhibition of host defense against mycobacteria

As our data showed that AICAR treatment resulted in the activation of mitochondrial function in Mtb-infected macrophages, represented by the increased ATP synthesis and mitochondrial contents, we next determined whether there exists a potential link between autophagy and mitochondria in the AMPK-induced activation of host defense using macrophages from Atg7^fl/flLysM-Cre⁺ mice. The ablation of Atg7 expression in macrophages resulted in a significant decrease in LC3-II levels in response to AICAR (Fig. 7A). In addition, a common substrate of autophagic degradation, SQSTM1, was found to accumulate, indicating a lack of autophagic degradation in BMDMs from Atg7^fl/flLysM-Cre⁺ mice (Fig. 7A). Notably, Atg7-deficient macrophages infected with Mtb showed depletion of ATP content in response to AICAR treatment, when compared with their Atg7⁺ littermates (Atg7^fl/flLysM-Cre^-) (Fig. 7B). It is worth noting that BMDMs from Atg7^fl/flLysM-Cre⁺ mice showed a marked loss of ΔΨm, even in untreated conditions (Fig. 7C). Mitochondrial DNA content in response to AICAR treatment was also significantly decreased in Atg7^fl/flLysM-Cre⁺ macrophages infected with Mtb (Fig. 7D). Further, ultrastructural analysis revealed that the mitochondria of Atg7^fl/flLysM-Cre⁺ BMDMs were significantly damaged; i.e., there was destruction of mitochondrial cristae, which were shortened, blurred, and dissolved. In addition, almost all mitochondria showed extreme swelling and vacuolization. The external and internal membranes were destroyed, and numerous mitochondria had ruptured (Fig. 7E; Fig. S5). AICAR treatment did not reverse the damaged mitochondrial morphology of Atg7^fl/flLysM-Cre⁺ BMDMs (Fig. 7E; Fig. S5).

Next, we determined whether AICAR-induced autophagy is required for the enhancement of phagosomal maturation and antimicrobial responses against intracellular mycobacteria. To evaluate phagosomal maturation, we infected Atg7^fl/flLysM-Cre⁺ and their Atg7⁺ littermates with ERFP-Mtb, followed by treatment with or without AICAR. Notably, the increased colocalization of the autophagosomal marker LC3 and bacterial phagosomes in Atg7⁺ macrophages were significantly decreased in Atg7-deficient BMDMs (Fig. 7F). When treated with AICAR, Atg7⁺ macrophages showed significantly increased killing of Mtb or BCG (Mycobacterium bovis BCG), whereas this was almost completely abrogated in Atg7-deficient BMDMs (Fig. 7G and H). In vivo administration of AICAR (500 mg kg-1 body weight) in BCG-infected Atg7^fl/flLysM-Cre⁺ mice³⁴ significantly inhibited in vivo bacterial growth in spleens, whereas this was markedly reduced in their Atg7⁺ littermates (Fig. 7I). Collectively, these data indicate that defective autophagy involves mitochondrial dysfunction and decreased antimicrobial responses to mycobacterial infection.

Reagents

5-Aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR, A9978), 3-methyladenine (3-MA, M9281), 4,6-diamidino-2-phenylindole dihydrochloride (DAPI, D9542), and N′-(7-chloroquinolin-4-yl)-N,N-diethyl-pentane-1,4-diamine (chloroquine, C6628) were purchased from Sigma. Compound C (171260), wortmannin (WM, 681675), bafilomycin A₁ (Baf-A, 196000) were purchased from Calbiochem. Dimethyl sulfoxide (DMSO; Sigma, D8418) was added to the cultures at 0.1% (v/v) as a solvent control.

Mice and cell culture

The mice used in individual experiments were age- (6 to 8 wk) and sex-matched. Wild-type C57BL/6 mice were purchased from SAMTAKO BIO (Gyeonggi-do, Korea). Atg7-floxed mice (Atg7^fl/fl, C57BL/6 background) mice were kindly provided by Dr Masaaki Komatsu (Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan). To generate transgenic myeloid lineage cell-specific Atg7-deficient mice (Atg7^fl/flLysM-Cre⁺) using the Cre/loxP recombination system, Atg7^fl/fl mice were crossed with B6.129P2-Lyz2^tm1(cre)Ifo/J mice (The Jackson Laboratory) expressing Cre recombinase under the control of a Lyzs promoter. F1 progeny from Atg7^fl/flLysM-Cre⁺ × Atg7^fl/fl crosses were genotyped for the presence (LysM-Cre⁺) or absence (LysM-Cre^-) of the LysM-Cre allele. Cre-mediated recombination was confirmed by PCR using genomic DNA from isolated from peritoneal macrophages and primers flanking the floxed region, as described previously.⁵⁸ Bone marrow was harvested from these mice and differentiated into macrophages. Absence of expression of Atg7 in these macrophages was confirmed by immunoblot analysis using an ATG7-specific antibody (Cell Signaling, 2631) (Fig. 7A). Mice were maintained under specific pathogen-free conditions. All animal-related procedures were reviewed and approved by the Institutional Animal Care and Use Committee, Chungnam National University School of Medicine (Daejeon, Korea). All animal-related procedures were reviewed and approved by the Institutional Animal Care and Use Committee, Chungnam National University School of Medicine (Daejeon, Korea).

Mice were maintained under specific-pathogen-free conditions. Primary bone marrow-derived macrophages (BMDMs) were isolated and differentiated by growth for 5 to 7 d in medium containing macrophage colony-stimulating factor (25 ng/mL; R&D Systems, 416-ML) as described previously.¹¹ The culture medium was Dulbecco’s modified Eagle’s medium (DMEM; Lonza, 12-604F), containing 4 mM glutamine and 10% FBS (Lonza, BW14-503E).

The mouse macrophage cell line RAW264.7 (ATCC TIB-71; American Type Culture Collection) was maintained in DMEM containing 10% FBS, sodium pyruvate, nonessential amino acids, penicillin G (100 IU/ml), and streptomycin (100 µg/ml). Human monocytic THP-1 (ATCC TIB-202) cells were grown in RPMI 1640/glutamax (Lonza, 12-702F) supplemented with 10% FBS, and then were treated with 20 nM PMA (Sigma-Aldrich, P8139) for 24 h to induce their differentiation into macrophage-like cells, followed by washing 3 times with phosphate-buffered saline (PBS). Transient transfections were performed using Lipofectamine 2000 (Invitrogen, 11668) according to the manufacturer’s instructions.

Mycobacterial strains

Mycobacterium tuberculosis H37Rv (Mtb) was kindly provided by R L Friedmann (University of Arizona, Tucson, AZ USA). Mycobacterium bovis BCG was obtained from the Korean Institute of Tuberculosis (Osong, Korea). Mycobacteria were grown in Middlebrook 7H9 (Difco, 271310) medium supplemented with 0.5% glycerol, 0.5% Tween-80 (Sigma, P1754), and 10% oleic acid, albumin, dextrose, and catalase (OADC; BD Biosciences, 212240). The enhanced Mtb-red fluorescent protein (ERFP) was described previously.¹¹ The Mtb-ERFP strain was cultivated in 7H9-OADC plus kanamycin (50 μg/ml; Sigma, 60615).

For Drosophila infection, the M. marinum strain Aronson (ATCC 927, fish isolate) was cultured in 7H9 broth with OADC and 0.2% Tween 80, at 30 °C in the dark without agitation, as described previously.¹¹

For all assays, mid-log-phase bacteria (absorbance, 0.4 to 0.7) were used. Batch cultures were aliquoted and stored at -80 °C. Representative vials were thawed and enumerated for viable colony-forming units (CFU) on Middlebrook 7H10 agar (Difco, 262710). The effective concentration of LPS was < 50 pg/ml in those experiments with a bacterium-to-cell ratio of 10:1. Single-cell suspensions of mycobacteria were prepared and used to infect cells as described previously.⁵⁹

Mycobacterial infection and colony-forming unit assays

To ensure the reliable quantification of intracellular bacteria, CFU assays were utilized. Cells were infected for 2 to 4 h with Mtb, Mtb-ERFP, or M. bovis BCG at different MOIs. Next, cells were washed 3 times with PBS, and fresh medium containing amikacin (Sigma; 200 μg/ml, A0368000) was added. After various periods of incubation, cells were lysed with 0.3% saponin (Sigma, 47036) to release the intracellular bacteria, and the lysates of infected cells were then resuspended vigorously, transferred to screw-capped tubes, and sonicated in a preheated 37 °C water bath sonicator (Elma) for 5 min. Aliquots of the sonicates were diluted 10-fold in 7H9 medium. Four dilutions of each sample were plated separately on 7H10 agar plates and incubated at 37 °C with 5% CO₂ for 14 to 21 d.

M. bovis BCG infection and bacterial growth in vivo

M. bovis BCG were thawed immediately before injection and diluted with saline to a final volume of 100 μL per mouse. Mice were held down in a restrainer, and their tails were mildly heated with a heating lamp to achieve vasodilatation. M. bovis BCG (1 × 10⁷ CFU/mouse) were injected through the tail vein slowly with a 30-gauge needle. After injection, mild pressure was applied at the spot of injection until no bleeding was achieved, to prevent the backflow of virus solution. Mice were monitored daily for 30 d. For measurement of the bacterial burden in the spleen, mice were sacrificed at 4 d after AICAR (500 mg/kg, i.p.) injection for 3 d, organs were homogenized in PBST, and serial dilutions of the homogenates were plated on 7H10 agar plates, with colonies counted 14 d later.

M. marinum infection and bacterial growth in vivo

M. marinum was cultured, and flies were maintained as described previously.¹¹ The Ppargc1 homolog Spargel mutant (P{SUPor-P}CG9809^KG08646ry⁵⁰⁶) and wild-type control (ry⁵⁰⁶) flies were obtained from the Bloomington Stock Center (Indiana University). For infection, 3- to 5-d-old Ppargc1 mutant or control³⁰ male flies were collected and infected as previously described.¹¹ After infection, flies were transferred to fresh standard cornmeal medium or AICAR medium (1 mM) and maintained at 25 °C and 60% humidity. Flies were transferred to fresh vials at least once per wk. For the AICAR effect of the M. marinum infection, Ppargc1 mutant or control files were infected with GFP-M. marinum (CFU = 500) by direct injection, and then were incubated with or without AICAR-containing medium. These files were incubated at 25 °C and 80% humidity for 3 d. GFP-expressing M. marinum in flies were observed using an Olympus IX81 confocal microscope with fluorescence optics. Determination of bacterial growth within infected flies was performed as described previously.¹¹ These files were infected with M. marinum with or without AICAR-containing medium for 3 d. The homogenized M. marinum-infected flies were serially diluted and plated onto 7H10 agar. After 14 d, calculation of the colony numbers on plates was performed. The experiments were repeated 3 times.

Lentiviral shRNA production and transduction

Lentiviral shRNA produced by transient transfection using packaging plasmids (psPAX2, pMD2.VSV-G, purchased from Addgene) and the target shRNA plasmid DNA (BECN1; RHS4531-EG8678, ATG5; RHS4531-EG9474, ATG7; RHS4531-EG10533, and Ppargc1a (RMM4532-NM_008904) were purchased from Open Biosystems; AMPKA (sc-29674-SH) was purchased from Santa Cruz Biotechnology) were cotransfected into 293T cells using Lipofectamine 2000 (Invitrogen, 11668) as described previously.³³ Virus-containing medium was collected at 72 h post-transfection and concentrated by ultracentrifugation. Lentiviral vector titration was determined using 293T cells, and the generated lentiviruses were transduced into BMDMs, THP-1, or Raw264.7 cells as described previously.³³ For lentivirus infection, cells (5 × 10⁵ /ml) in medium containing 10% FBS were seeded in 24-well plates, and then infected with lentiviral vectors (MOI = 10) in the presence of polybrene (Sigma; 8 μg/ml, H9268). On the following day, the medium was replaced, the cells were incubated for 2 d, and they were then analyzed for target gene knockdown. A parallel experiment using a GFP-encoding lentivirus (the pGIPZ lentiviral vector; Open Biosystems, RMM4431) indicated that 80% of cells were successfully transduced by the virus.

Immunoblotting

Immunoblotting was performed as described previously.³³ For immunoblotting, polypeptides were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to a PVDF membrane (Bio-Rad). Specific antibodies to LC3 (L8918) and SQSTM1/p62 (P0067) were purchased from Sigma. Antibodies against phospho-ACACA/B (Acetyl-CoA carboxylase) (3661), phospho-AMPKA (2535), AMPKA (2532), phospho-MTOR (2971), phospho-ULK1 Ser317 (12753), Ser555 (5869), Ser757 (6888), MTOR (2983), CYCS/Cyt C (4272), SDHA (5839), and ATG7 (2631) were purchased from Cell Signaling. Anti-TFAM (sc-28200), anti-NFE2L1/NRF1 (sc-28379), anti-NDUFA9 (sc-58392), and anti-ACTIN (sc-1616) were purchased from Santa Cruz Biotechnology. Anti-PPARGC1A (ab54481), anti-COX4l1/COX IV (ab16056), and anti-UQCRC2 (ab14745) were purchased from Abcam (Cambridge). Anti-ATP5A1 (459240) was purchased from Invitrogen. These proteins were visualized by enhanced chemiluminescence (ECL; Millipore) and exposure to film (FUJI medical X-ray film). For autophagy analysis, cells were treated for 30 min on ice, lysed with 1% Triton X-100, and then subjected to immunoblotting analysis with an antibody against LC3 or SQSTM1/p62.

Immunofluorescence and confocal microscopic analysis

Endogenous LC3 punctate staining and immunofluorescence analysis were performed as described previously.¹¹ After the appropriate treatment, cells were washed twice with PBS, fixed with 4% paraformaldehyde in PBS for 10 min, permeabilized with 0.25% Triton X-100 in PBS for 10 min, and incubated with anti-LC3 (1:400; MBL International, PM036) for 2 h at room temperature. Cells were washed to remove excess primary antibody, and incubated with the appropriate fluorescently labeled secondary antibodies for 1 h at room temperature. Nuclei were stained by incubation with DAPI for 5 min. After mounting, fluorescence images were acquired using a confocal laser-scanning microscope (LSM 710; Zeiss, Thomwood, NY USA). Fluorescence intensity was measured using the ImageJ or Adobe Photoshop CS4 software. Endogenous LC3 punctate dots were enumerated using the ImageJ software. Each condition was assayed in triplicate, and at least 200 cells per well were counted.

For colocalization analysis, the co-distribution of the LC3 autophagosome and bacterial phagosome were quantified and validated statistically by Pearson coefficient, as specified by the ZEN 2009 software (version 5.5 SP1; Zeiss). Images of dynamic cell interactions were recorded as vertical z-stacks (3-dimensional images). Imaris 7.1.1 (Bitplane), Ultraview 5.5 (PerkinElmer), and Adobe Photoshop 7 (Adobe Systems) were used for image processing.

Flow cytometry

Cells were analyzed by flow cytometry for LC3B and mitochondria mass using a FACSCanto II flow cytometer as indicated by the manufacturer (Becton Dickinson, San Jose, CA USA). After 2 washes with PBS, cells were fixed in 4% paraformaldehyde for 10 min at 37 °C or permeabilized with 0.25% Triton X-100 in PBS for 10 min. Cells were stained with primary antibodies for 1 to 2 h at 4 °C (1:100) and the secondary antibodies for 1 h on ice. After 2 washes with PBS, cells were fixed in 4% paraformaldehyde and assayed immediately. Flow cytometry data were collected using 30,000 to 100,000 cells and were analyzed using the FlowJo software (Tree Star, Ashland, OR USA).

Transmission electron microscopy

For transmission electron microscopy analysis, BMDMs were washed with PBS, fixed with 4% paraformaldehyde and 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 1 h, post-fixed in 1% osmium tetroxide and 0.5% potassium ferricyanide in cacodylate buffer for 1 h, embedded in resin, and cured at 80 °C for 24 h. Ultrathin sections (70 to 80 nm) were cut using an Sorvall RMC MT 6000 MT6000-XL Ultramicrotome (Microtome Service Company, NY USA), stained with uranyl acetate and lead citrate, and examined using a Tecnai G2 Spirit Twin transmission electron microscope (FEI Co., Hillsoboro, OR USA) and a JEM ARM 1300S high-voltage electron microscope (JEOL, Tokyo, Japan).

Measurement of ATP production

ATP levels were measured by the Luciferin/Luciferase method using the ATP Bioluminescence assay kit (PerkinElmer, 6016943) according to the manufacturer's instructions. Briefly, ~2 × 10⁵ BMDMs were infected with Mtb or were treated with AICAR, then washed with PBS. At 24 h later, they were again washed with PBS. ATP liberation buffer was then added to the cell cultures. Luciferin and Luciferase were mixed together (50 μl, final volume) in a separate vial to which 50 μl of lysate (liberated ATP) from the cell cultures was then added. Luminescence was analyzed after a 30 s delay on a FluoroScan luminometer (Labsystems, Helsinki, Finland). An ATP standard curve prepared at the same time according to the manufacturer’s instructions was used to calculate the concentration of ATP in each sample.

Mitochondrial DNA quantification

To quantify mtDNA copy number, we measured the mitochondrial (mt) to nuclear (n) DNA ratio as described previously.⁶⁰ Pyruvate kinase (Pklr) was used as a marker for nDNA and NADH dehydrogenase subunit 1 (mt-Nd1) for mtDNA. Real-time PCR reactions were performed according to the manufacturer’s instructions (QuantiFast SYBR green PCR master mix; Qiagen, 204052), and thermal cycling was performed in a Rotor gene 6000 instrument (Qiagen, Hilden, Germany). The mtDNA content was normalized to the nucleic DNA content. The primer pairs used for PCR are listed in Table. S1.

Mitochondrial membrane potential measurements

The mitochondrial membrane potential (ΔΨm) of intact cells was measured as described previously,⁶¹ with modifications. Cells were washed with PBS and trypsinized. The protein concentration of cells was adjusted to 0.2 mg/ml in DMEM without phenol red (Life Technologies-Invitrogen), FBS, and antibiotics. TMRE (tetramethylrhodamine, ethyl ester; 200 nM, Molecular Probes-Invitrogen, T669) was added to the cell suspension. Cells were incubated at 37 °C for 30 min in the dark. ΔΨm was measured by flow cytometry, and data were analyzed using the FlowJo software. TMRE fluorescence was measured using the FL2 channel (582 nm).

Mitochondrial mass determination and visualization of mitochondria in living cells

Mitochondrial mass was measured using fluorescence levels upon staining with MitoTracker Green FM (Invitrogen, M7514) at 100 nM at 37 °C for 30 min. Subsequently, cells were washed once in PBS and promptly used for flow cytometry. The percentage and mean fluorescence intensity level of the mass were calculated for each sample.

To visualize the mitochondria, mitochondrial marker TOMM20 at 100 nM was used, adding the dye with DMEM media at 37 °C for 30 min. Next, the cover glasses were washed twice with PBS and directly acquired with a confocal laser-scanning microscope (LSM 710; Zeiss). The images were evaluated as best-focus intensity projections.

RNA extraction and quantitative RT/real-time PCR

For semiquantitative RT-PCR, total RNA was extracted from cells using TRIzol (Invitrogen, 15596018), and cDNA synthesis was performed using an iScript cDNA Synthesis Kit (Bio-Rad, 18080200) as described previously.³³ Using iQ SYBR Green Supermix (Bio-Rad, 170-8880), we performed quantitative analysis of mRNA expression using a Veriti 96-well Thermal Cycler (Applied Biosystems). PCR amplification was performed in a total volume of 30 μl containing 2 μl of cDNA and 10 pM primers (Invitrogen). The PCR amplification protocol consisted of an initial incubation step for 10 min at 95 °C, followed by 35 cycles of denaturation at 95 °C for 30 s, annealing for 30 s at 55 °C, and extension at 72 °C for 30 s. PCR products were resolved on 1.5% agarose gels and were stained with ethidium bromide. The relative quantitative real-time mRNA expression was determined using the SYBR green PCR master mix (Qiagen, 204072) and Rotor gene 6000 (Qiagen). Expression levels of genes were calculated by the 2^∆∆Ct method using Gapdh as an internal control gene. Data were analyzed using an unpaired, 2-tailed Student t test, and the results were expressed as relative fold changes. The primer pairs used for PCR are listed in Table. S1.

Luciferase assays

Raw264.7 or THP-1 cells were transiently transfected with a luciferase gene under the control of the pGL3 promoter-containing NFKB1 sites, or the CEBPB transcription recognition sequences were used. We also used the pRL vector expressing a Renilla luciferase gene (Promega, E2241) for constitutive protein expression, and the dual luciferase reporter assay system (Promega, E1910) substrate.

For luciferase reporter gene assays, human ATG5, BECN1, and ATG7 promoter constructs were used. The human ATG5 and BECN1 promoter constructs were described previously.³³ For the human ATG7 promoter construct, a 4463-bp fragment upstream of the predicted transcription start site of the human ATG7 gene was generated by PCR amplification from human genomic DNA using the sense primer 5′-GCACGCGTAG TTTGAGAGGG CAACGG CAGGAGAG-3′ and antisense primer (ATG7-RP) 5′-GGAAGATCTG CAGGTGTGGA GAGC CAGGAGATGA GA-3′. The 5′-promoter deletion construct (2122-bp fragment) was generated using the antisense primer AGT7-RP and the sense primer 5′-GCACGCGTTG AGGTGAAGGT GGGAGGATCA CTTGAGTC-3′. These fragments were digested with Mlu I and Bgl II and cloned upstream of the luciferase gene in the pGL3-Basic reporter vector (Promega). Searches for putative DNA binding sites for transcription factors in the human ATG7 5′-flanking region sequences were conducted using TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html) and MOTIF Search (http://motif.genome.jp).

A schematic representation of the generation of human ATG5, BECN1, and ATG7 promoter-reporter constructs is shown in Figure 6C. The reporter gene assay was performed as described previously.³³ The luciferase reporter vector controlled by a synthetic promoter containing direct repeats of the transcription recognition sequences for the CEBPB was obtained from Stratagene. Raw264.7 or THP-1 cells were plated in 6-well culture plates and transfected with a firefly reporter vector (4.0 μg) and pRL vector expressing the Renilla luciferase gene (0.1 μg), together with the indicated expression plasmids (4.0 μg) using the Lipofectamine 2000 (Invitrogen) reagent according to the manufacturer’s instructions. After 2 d of transfection, cells were lysed in Passive Reporter Lysis Buffer (Promega, E1910), and luciferase activity was measured. Firefly luciferase activity was normalized to Renilla luciferase activity. Each transfection was performed in triplicate, and 3 independent experiments were conducted.

Statistical analyses

For statistical analysis, data obtained from independent experiments (means ± standard deviation [SD]) were analyzed using a 2-tailed Student t test. Differences were deemed to be significant at a P value less than 0.05. For Drosophila survival, data were graphed and analyzed using the GraphPad Prism software (GraphPad Software, Inc., San Diego, CA USA). The 95% confidence intervals were computed as follows: 1.96× standard error in each direction. Statistical significance was evaluated using the log-rank (Mantel–Cox) test.