The Peroxisome Proliferator-Activated Receptor α- Agonist Gemfibrozil Promotes Defense Against Mycobacterium abscessus Infections

Peroxisome proliferator-activated receptor α (PPARα) shows promising potential to enhance host defenses against Mycobacterium tuberculosis infection. Herein we evaluated the protective effect of PPARα against nontuberculous mycobacterial (NTM) infections. Using a rapidly growing NTM species, Mycobacterium abscessus (Mabc), we found that the intracellular bacterial load and histopathological damage were increased in PPARα-null mice in vivo. In addition, PPARα deficiency led to excessive production of proinflammatory cytokines and chemokines after infection of the lung and macrophages. Notably, administration of gemfibrozil (GEM), a PPARα activator, significantly reduced the in vivo Mabc load and inflammatory response in mice. Transcription factor EB was required for the antimicrobial response against Mabc infection. Collectively, these results suggest that manipulation of PPARα activation has promising potential as a therapeutic strategy for NTM disease.


Introduction
The incidence of nontuberculous mycobacterial (NTM) infection is increasing worldwide, not only in immunocompromised patients, but also in healthy individuals [1][2][3][4]. However, little is known about the host factors that influence the pathogenesis of NTM infection [1,3,4]. Mycobacterium abscessus (Mabc), differentiation, adherent monocytes were incubated in Roswell Park Memorial Institute 1640 medium (Lonza, 12-702F) containing 10% pooled human serum (Lonza, , 1% L-glutamine, 50 IU/mL penicillin, and 50 mg/mL streptomycin for 1 h at 37 • C, after which the nonadherent cells were removed. Human monocyte-derived macrophages (MDMs) were prepared by culturing peripheral blood monocytes for 4 d in the presence of 4 ng/mL human CSF/macrophage colony-stimulating factor (Sigma-Aldrich, St. Louis, MO, USA, M6518) as described previously [24]. All experimental procedures were approved by the Chungnam National University Institutional Research and Ethics Committee.

Mycobacterial Strains and Culture
Mabc (ATCC 19977), Mabc (CIP 104536T), Mabc-enhanced green fluorescent protein (EGFP), and Mabc-LuxG13 were incubated at 37 • C with shaking in Middlebrook 7H9 broth (BD Biosciences, Franklin Lakes, NJ, USA, 271310) supplemented with 0.2% glycerol, 0.05% Tween-80 (Sigma-Aldrich), and albumin-dextrose catalase. Bacterial cultures were harvested by centrifugation at 500 × g for 20 min, and the pellets were resuspended in bacterial culture medium. Bacterial cultures were divided into 1 mL aliquots and stored at −80 • C. Mabc cultured cells in the logarithmic phase were centrifuged, washed, and resuspended in PBS plus 0.05% Tween 80 (PBST) before homogenizing through a 26-gauge needle and sonicated in a water bath. Bacterial stock cultures were stored at −80 • C. The number of colony-forming units (CFUs) of the inoculum was verified by plating serial dilutions. We used 50 µg/mL kanamycin to culture Mabc-EGFP. pMV306hsp+LuxG13 (plasmid #26161) was introduced into competent Mabc cells by electroporation and plated on 7H10 agar containing 50 µg/mL kanamycin. Kanamycin-resistant colonies were picked and inoculated into 7H9 broth containing kanamycin. Luminescent Mabc was identified using a SpectraMax M3 luminescent plate reader (Molecular Devices, San Jose, CA, USA). Bacteria in mid-logarithmic phase (absorbance 0.4) were used in all experiments. Representative vials were thawed and the CFU enumerated by serially diluting and plating on Middlebrook 7H10 agar (BD Biosciences, 262710).

Mycobacterial Infection
BMDMs were infected for 2 h with Mabc, Mabc-EGFP, or Mabc-LuxG13, washed three times with PBS to remove extracellular bacteria, and incubated in fresh medium for the indicated periods at 37 • C. For mouse infection in vivo, frozen Mabc cells were thawed and centrifuged, and the pellet was resuspended in PBST. Mice were administered with Mabc either intravenously (iv; 1 × 10 7 CFU/mouse) or intranasally (in; 1 × 10 6 CFU/mouse). To assess the bacterial burden, mice were euthanized at the indicated times after Mabc infection, and the lung, spleen, and liver were removed and homogenized using a tissue homogenizer (OMNI TH) in PBST. To assess bacterial viability in BMDMs, infected cells were lysed in distilled water to release intracellular bacteria. Thereafter, the serially diluted homogenates of Mabc-infected tissues and cells were inoculated onto 7H10 agar, and colonies were counted 3-4 days later. To quantify intracellular Mabc-LuxG13 based on luciferase activity, the luminescence signal of Mabc-LuxG13 was measured for 10 s using a luminescence microplate reader, and bioluminescence was expressed as relative light units [25].

GEM Treatment
GEM was dissolved in dimethyl sulfoxide (DMSO) followed by dilution in 0.1% methyl cellulose (Sigma-Aldrich, M7027) and was used at 15 mg/kg body weight. Mice were treated with vehicle (0.1% methyl cellulose) or GEM once daily via gavage from days 3 to 7 after infection.

Immunopathology of Mycobacterium-Infected Mice
Lungs were harvested from mice infected with Mabc for the indicated times, fixed in 10% formalin, and embedded in paraffin wax. For histopathology, lung paraffin sections (4 µm) were cut and stained with hematoxylin and eosin (H&E). H&E-stained images were scanned using the Aperio Digital Pathology Slide Scanner (Leica, Wetzlar, Germany) and imaged using the ScanScope®CS System (Leica). The histopathological severity score was graded by scanning multiple random fields in six sections per mouse. For immunohistochemical (IHC) staining, lung paraffin sections (4 µm) were cut and immunostained with antibodies specific for myeloperoxidase (MPO; Abcam, Cambridge, MA, USA, ab9535) and LyG6 (Bio X Cell, West Lebanon, NH, USA, BP0075-1). Immunohistochemically stained lung tissue slides were examined using a confocal laser scanning microscope.

RNA Extraction, Semiquantitative Real-Time PCR, and Quantitative Real-Time PCR
Total RNA was isolated from homogenized tissues or cells using TRIzol (Thermo Fisher Scientific, Waltham, MA, USA, 15596-026), following the manufacturer's instructions. After RNA quantitation, cDNA was synthesized by reverse transcription using the reverse transcriptase premix (Elpis Biotech, Daejeon, Korea, EBT-1515). Semiquantitative PCR was performed using the Prime Taq Premix (GeNet Bio, Nonsan, Korea, G-3000). Quantitative real-time PCR (qRT-PCR) was performed on the Rotor-Gene Q 2plex system (Qiagen, Hilden, Germany) using SYBR Green master mix (Qiagen, 204074). Gene expression was relatively quantitated using the 2 −∆∆Ct method, with normalization to β-actin expression, and expressed as the relative fold change. The primer sequences used are listed in Supplementary Table S1.

Immunofluorescence
After the appropriate treatment, cells on coverslips were washed three times with PBS, fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.25% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) for 10 min, and incubated with the appropriate primary antibody overnight at 4 • C. Cells were washed with PBS to remove excess primary antibody and incubated with the secondary antibody for 2 h at room temperature. Nuclei were stained with DAPI for 2 min. After mounting, fluorescence images were acquired and analyzed using the TCS SP8 confocal microscope (Leica, Wetzlar, Germany) and MetaMorph Advanced Imaging acquisition software version 7.8 (Molecular Devices, San Jose, CA, USA).

Minimum Inhibitory Concentration (MIC) Determination Using Resazurin Microtiter Assay (REMA)
The MICs of the compounds were determined as described previously [26]. Briefly, 100 µL of cation-adjusted Mueller-Hinton (CAMH) and 7H9 broth supplemented with 10% ADC was added to every well of a 96-well microtiter plate. A two-fold serial dilutions were prepared in 96-well clear microplates to obtain concentration ranges from 200 µM to 0.39 nM. Log-phase Mabc strain CIP 104536T cultures were added to each wells to an OD600 of 0.05 and the plates were incubated at 37 • C for 3 days prior to addition of resazurin (0.025% [wt/vol]). After overnight incubation, fluorescence was measured by excitation at 530 nm and emission at 590 nm using a SpectraMax®M3 Multi-Mode Microplate Reader (Molecular Devices, San Jose, CA, USA). The MIC was calculated using GraphPad Prism 6 software (GraphPad Software, Inc., La Jolla, CA, USA).

Statistical Analysis
Each experiments was carried out three times. Statistical tests were performed using GraphPad Prism version 5 by Student's t-test and one-way ANOVA. Data are presented as means ± standard deviation (SD).

Ppara -/-Mice Exhibited an Upregulated Proinflammatory but Defective Antimicrobial Response During Mabc Infection Compared with Ppara +/+ Mice
To examine whether PPARα is involved in the defense against Mabc infection, we infected Ppara +/+ and Ppara -/mice with Mabc (ATCC 19977; iv, 1 × 10 7 CFU/mouse) for 7 days and examined the mycobacterial burden in the lung, liver, and spleen. As shown in Figure 1A, the in vivo bacterial load was significantly increased in the lung, liver, and spleen of Ppara -/mice compared with Ppara +/+ mice at 7 days post-infection (dpi). To confirm that PPARα is required for host defense, we infected mice with Mabc (in, 1 × 10 6 CFU/mouse). The number of bacteria was greater in the lung of Ppara -/mice than Ppara +/+ mice ( Figure 1B).
We next examined the histopathological features on the lungs of Ppara +/+ and Ppara -/mice infected with Mabc. As shown in Figure 1C, the number of granuloma-like inflammatory lesions was significantly increased in Ppara -/mice compared with Ppara +/+ mice. Neutrophil infiltration in the lungs may play a protective or pathological role during Mabc infection [27,28]. Thus, we examined neutrophil infiltration in the lungs of Ppara +/+ and Ppara -/mice.

PPARα Deficiency Increased the Inflammatory Response During Mabc Infection
PPARα is a regulator of the inflammatory response. We next examined whether PPARα is involved in Mabc-induced inflammation in vivo. The mRNA levels of proinflammatory cytokines and chemokines were markedly upregulated in the lung of Ppara -/compared with Ppara +/+ mice Cells 2020, 9, 648 7 of 17 ( Figure 2A). Notably, the mRNA levels of Tnf, Il6, Il1b, Cxcl10, and Cxcl5 were significantly increased and that of Il10 decreased in lung tissue from Ppara -/mice, compared with Ppara +/+ mice, during Mabc infection (Figure 2A).

PPARα Activation Enhances Macrophage Antimicrobial Responses Againt Mabc Infection
We next investigated the effects of PPARα agonists on Mabc infection in macrophages. Although PPARα agonist GEM [29] did not exhibit any direct antimicrobial effect against Mabc (Supplementary Figure S2), it substantially inhibited the intracellular growth of Mabc in BMDMs ( Figure 3A and 3B) and human MDMs ( Figure 3C). Notably, GEM (100 μM) treatment of BMDMs and human MDMs resulted in significant inhibition of the intracellular growth of Mabc (Figures We next examined the mRNA and protein levels of these cytokines and chemokines in Ppara +/+ and Ppara -/-BMDMs after Mabc infection. As shown in Figure 2B, the mRNA levels of Tnf, Il6, and Il1b were significantly increased in BMDMs from Ppara -/mice, compared with Ppara +/+ mice, at 6 h after Mabc infection. In addition, the TNF and IL-6 protein levels were markedly upregulated in Ppara -/-BMDMs compared with Ppara +/+ BMDMs, at 6 and 18 h after Mabc infection ( Figure 2C). Furthermore, the expression of TNF, IL-6, and IL-1β mRNA and protein levels were significantly increased in Ppara -/-BMDMs, compared with Ppara +/+ BMDMs, after Mabc infection ( Figure 2D,E; protein and mRNA, respectively). Together, these data suggest that proinflammatory responses are markedly upregulated in the lungs and macrophages from Ppara -/mice, compared with Ppara +/+ mice, during Mabc infection.

PPARα Activation Enhances Macrophage Antimicrobial Responses Againt Mabc Infection
We next investigated the effects of PPARα agonists on Mabc infection in macrophages. Although PPARα agonist GEM [29] did not exhibit any direct antimicrobial effect against Mabc (Supplementary Figure S2), it substantially inhibited the intracellular growth of Mabc in BMDMs ( Figure 3A,B) and human MDMs ( Figure 3C). Notably, GEM (100 µM) treatment of BMDMs and human MDMs resulted in significant inhibition of the intracellular growth of Mabc ( Figure 3A-C). As a positive control, treatment of Mabc-infected BMDMs with clarithromycin resulted in a marked inhibitory effect on intracellular growth of Mabc ( Figure 3B). However, it is noted that clarithromycin is a drug with pleiotropic activity [30]. In addition, the combined treatment of BMDMs with GEM and clarithromycin did not show any significant additional/synergistic effect upon anti-Mabc response (data not shown). Moreover, GEM-mediated antimicrobial activity was mediated through PPARα, since the GEM-induced suppressive effect on intracellular growth of Mabc in Ppara +/+ BMDMs was significantly abrogated in Ppara -/-BMDMs ( Figure 3D). We also showed that application of the PPARα antagonist GW6471 [31] increased intracellular bacterial growth in a dose-dependent manner ( Figure 3A). These data suggest that GEM treatment increased the antimicrobial activity against intracellular Mabc through PPARα.

PPARα Activation Regulates the Inflammatory Response to Mabc Infection
We next examined regulation of the inflammatory response by GEM during Mabc infection. As shown in Figure 4A, GEM treatment (100 µM) of BMDMs dramatically suppressed the expression of a variety of inflammatory cytokines and chemokines (Il6, Il12p40, Il1b, Cxcl10, and Cxcl5) at 18 h after infection. The Il10 mRNA level was substantially decreased in BMDMs at 18 h but significantly increased at 6 h after Mabc infection ( Figure 4B). In addition, we examined whether GEM treatment inhibited Mabc-induced inflammatory responses in macrophages through PPARα. To examine this, we infected Ppara +/+ and Ppara -/-BMDMs with Mabc (MOI of 5), and treated with GEM (100 µM) for 18 h prior to measure the mRNA levels of Il6 and Il1b. As shown in Figure 4C, we found that the inhibitory effects of GEM on Mabc-mediated Il6 and Ilb expression in Ppara +/+ BMDMs were significantly abrogated in Ppara -/-BMDMs. Therefore, GEM treatment suppressed Mabc-induced inflammatory response depending on PPARα.

PPARα Activation Upregulates Antimicrobial Responses and Downregulates Pathological Inflammation during Mabc Infection
To confirm and extend the in vitro data, we next assessed the function of PPARα activation by GEM in the defense against Mabc infection in vivo. C57BL/6 mice were infected with Mabc iv, and GEM (15 g/kg body weight) was administered orally. As shown in Figure 5A and 5B, the numbers of viable bacteria in the lung, liver, and spleen were significantly reduced in GEM-treated mice compared with control mice (Figure 5A,B, iv and in, respectively). Consistent with the in vitro

PPARα Activation Upregulates Antimicrobial Responses and Downregulates Pathological Inflammation during Mabc Infection
To confirm and extend the in vitro data, we next assessed the function of PPARα activation by GEM in the defense against Mabc infection in vivo. C57BL/6 mice were infected with Mabc iv, and GEM (15 g/kg body weight) was administered orally. As shown in Figure 5A,B, the numbers of viable bacteria in the lung, liver, and spleen were significantly reduced in GEM-treated mice compared with control mice (Figure 5A,B, iv and in, respectively). Consistent with the in vitro results (Figure 4A), the mRNA levels of proinflammatory cytokines (Tnf and Il1b) were significantly decreased in the lung of GEM-treated mice compared with control mice ( Figure 5C). In addition, the Il10 mRNA level was significantly upregulated in the lung tissue of GEM-treated mice compared with control mice ( Figure 5C).

PPARα is Required for Nuclear Translocation of TFEB in Macrophages during Mabc Infection
The smooth variant of Mabc restricts phagosomal acidification and induces less autophagy compared with the rough variant [33]. We thus evaluated LC3-I/II conversion and p62 expression in BMDMs after Mabc infection. As shown in Figure 6A, Mabc infection increased the LC3-II and p62 levels, suggesting blocking of autophagic flux. In addition, phagosomal acidification, i.e., colocalization of Mabc with the lysosomal marker LAMP1, was markedly reduced in Ppara -/-BMDMs compared with Ppara +/+ at 4 h (Figures 6B and 6C). At 24 h after Mabc infection, the slight A rough variant of Mabc enhanced type I interferon (IFN) production and cell death [32]. We thus examined the expression of Ifna5 in the lung tissues of Ppara +/+ compared with Ppara -/-mice. Interestingly, the Ifna5 mRNA level was markedly upregulated in the lung of Mabc-infected Ppara -/compared with Ppara +/+ mice ( Figure 5C). Moreover, the number of granulomatous lesions was considerably decreased by GEM ( Figure 5D).

PPARα is Required for Nuclear Translocation of TFEB in Macrophages during Mabc Infection
The smooth variant of Mabc restricts phagosomal acidification and induces less autophagy compared with the rough variant [33]. We thus evaluated LC3-I/II conversion and p62 expression in BMDMs after Mabc infection. As shown in Figure 6A, Mabc infection increased the LC3-II and p62 levels, suggesting blocking of autophagic flux. In addition, phagosomal acidification, i.e., colocalization of Mabc with the lysosomal marker LAMP1, was markedly reduced in Ppara -/-BMDMs compared with Ppara +/+ at 4 h ( Figure 6B,C). At 24 h after Mabc infection, the slight intraphagosomal acidification disappeared in Ppara +/+ BMDMs such that there was no difference between Ppara +/+ and Ppara -/-BMDMs ( Figure 6B,C). We next examined whether PPARα activation by GEM induces nuclear translocation of TFEB in BMDMs. GEM treatment of BMDMs led to a significant increase in TFEB nuclear translocation in Mabc-infected BMDMs ( Figure 6E). We further examined whether other PPARα agonist regulated nuclear translocation of TFEB. As shown in Supplementary Figure S3, treatment of BMDMs with GW7647 did not significantly upregulate the nuclear translocation of TFEB in BMDMs during Mabc infection, when compared with untreated controls. Furthermore, the GEM-mediated antimicrobial response in BMDMs was significantly attenuated by knockdown of Tfeb using shRNA targeting Tfeb, compared with a nonspecific shRNA ( Figure 6F). Together, these data suggest that PPARα activation by GEM upregulates phagosomal maturation and the antimicrobial response to Mabc by inducing nuclear translocation of TFEB.  TFEB is an essential transcription factor for genes involved in autophagy and lysosomal biogenesis [34]. TFEB activation is required for IFNγ-or PPARα-induced autophagy and restriction of the intracellular growth of M. tuberculosis [20,35]. However, the nuclear translocation of TFEB has not been investigated in the context of Mabc infection. We compared the intracellular translocation of TFEB in BMDMs after Mabc infection between Ppara +/+ and Ppara -/-BMDMs. As shown in Figure 6D, we found that TFEB nuclear translocation was significantly increased in Ppara +/+ BMDMs, but markedly decreased in Ppara -/-BMDMs, after 30 min of Mabc infection.
We next examined whether PPARα activation by GEM induces nuclear translocation of TFEB in BMDMs. GEM treatment of BMDMs led to a significant increase in TFEB nuclear translocation in Mabc-infected BMDMs ( Figure 6E). We further examined whether other PPARα agonist regulated nuclear translocation of TFEB. As shown in Supplementary Figure S3, treatment of BMDMs with GW7647 did not significantly upregulate the nuclear translocation of TFEB in BMDMs during Mabc infection, when compared with untreated controls. Furthermore, the GEM-mediated antimicrobial response in BMDMs was significantly attenuated by knockdown of Tfeb using shRNA targeting Tfeb, compared with a nonspecific shRNA ( Figure 6F). Together, these data suggest that PPARα activation by GEM upregulates phagosomal maturation and the antimicrobial response to Mabc by inducing nuclear translocation of TFEB.

Discussion
Mabc, a rapidly growing NTM, causes serious pulmonary infections in both immunocompetent and immunocompromised individuals [1,36,37]. Within the Mabc complex, Mabc (sensu stricto) is a clinically important species because it is the most drug-resistant NTM and is associated with a high rate of treatment failure [37,38]. Understanding host-pathogen interactions and the underlying mechanisms is vital for the development of new therapeutics against NTM infection [1,39,40]. Recent studies on the roles of the innate and acquired immune system in NTM infection have suggested that multiple pathways/factors contribute to the pathogenesis of Mabc infection and host protective immunity [1,40]. In this study, we found that activation of PPARα induced an antimicrobial response to Mabc infection in vitro and in vivo.
PPARα is a key factor involved in host defense in macrophages against M. tuberculosis and M. bovis BCG [20]. These findings raised the question of whether PPARα promotes innate host defense against NTM infections and, therefore, whether PPARα-activating agents have potential as drug targets for Mabc infection in vivo. We addressed this issue using in vivo and in vitro models of Mabc infection and found that BMDMs from Ppara -/mice exhibited increased bacterial growth compared with those from Ppara +/+ mice. Similar to the findings of previous studies of M. tuberculosis and M. bovis BCG [20], we found that PPARα deficiency increased the levels of proinflammatory cytokines and chemokines in the lung and macrophages during Mabc infection. These findings are in partial agreement with a recent report of a protective role of PPARα activation against lipopolysaccharide-induced acute lung injury in mice [11].
In this study, we found that neutrophil infiltration was significantly increased in Ppara -/-BMDMs and mice after infection. Therefore, the inflammatory response presumably mediated by neutrophil infiltration may play a double-edged role in Mabc infection. Neutrophils have crucial functions in host defense, i.e., bacterial uptake and mycobactericidal activity, against the rough and smooth morphotypes by generating reactive oxygen species and extracellular traps [41]. In addition, impaired TNF signaling and defective neutrophil trafficking result in formation of aberrant granulomas, massive bacterial growth, and death of zebrafish larvae [42]. In line with the clinical finding that anti-TNF therapy exacerbates Mabc infection [43,44], appropriate induction of TNF and other inflammatory cytokines as well as neutrophil infiltration may play a fundamental role in host defense against Mabc infection. However, in human neutrophils, Mabc induces the expression of genes encoding chemokines and proinflammatory cytokines and shows augmented biofilm formation, which promote survival of the bacteria by exploiting the neutrophil-rich environment [45]. During pulmonary NTM infection, recruited neutrophils release elastase and metalloproteinase, damaging the airway epithelium and resulting in formation of microabscesses [40]. Therefore, an excessive inflammatory response due to infiltrated neutrophils may induce an exaggerated immunopathological reaction in vivo.
Our results reveal that Cxcl10 and Cxcl5 mRNA levels were significantly modulated by PPARα during Mabc infection in vivo. In the lung of Ppara -/mice, Cxcl5 and Cxcl10 expression levels were significantly increased. The chemokine CXCL5 is involved in the regulation of neutrophil influx and pulmonary neutrophil inflammation during M. tuberculosis infection [46][47][48]. In addition, the Cxcl5 mRNA level was significantly increased in macrophages and epithelial cells in Sirt3-deficient lung during M. tuberculosis infection [47]. Neutrophil depletion by targeting Ly6G reportedly enhances host defense against M. tuberculosis infection in mice [46][47][48]. In addition, whether CXCL10 is increased in Mabc patients or Mabc infection models is unclear. M1 macrophages secrete proinflammatory cytokines and chemokines, including TNF-α, IL-6, IL-12, and CXCL10 [49]. The production of these cytokines and CXCL10 was markedly upregulated in the lung of Ppara -/compared with Ppara +/+ mice. The CXCL10 level was found to be increased in the plasma of untreated and treatment-refractory patients with Mycobacterium avium complex (MAC)-induced lung diseases, suggesting a role of CXCL10 as a marker of NTM disease [50]. Furthermore, the serum CXCL10 level was increased in patients with the fibrocavitary form of M. avium complex lung disease, compared with healthy controls, suggesting a role of CXCL10 in the pathogenesis of NTM diseases [51]. Together, our data suggest that CXCL10 and CXCL5 contribute to the pathogenesis of Mabc lung infection by recruiting and inducing over-activation of M1 macrophages and neutrophils, respectively.
Having established a relationship between PPARα and TFEB, we next determined whether the PPARα-TFEB axis is required for host defense against Mabc infection. The S variant of Mabc was reported to induce autophagy [33], and we found that Mabc infection of wild-type BMDMs blocked autophagic flux and induced slight phagosomal acidification in Ppara +/+ BMDMs. Moreover, the nuclear translocation of TFEB was slightly, but significantly, increased in Ppara +/+ but not Ppara -/-BMDMs. Because TFEB is a master regulator of, and essential transcription factor for, lysosomal biogenesis and control of xenophagy during M. tuberculosis infection [20,52], the regulatory mechanisms of TFEB activation are critical for the antimycobacterial response. Nuclear receptor subfamily 1, group D, member 1, an adopted orphan nuclear receptor, plays an important role in the regulation of TFEB transcription [53]. Although we did not examine the effect of TFEB on the inflammatory response during Mabc infection, we reported previously that TFEB silencing enhanced the inflammatory response of macrophages infected with M. tuberculosis [20]. Therefore, while the host autophagy response differs between M. tuberculosis and Mabc, the functional significance of PPARα-TFEB in the antimicrobial response is identical.
We found that GEM was effective against Mabc infection in vivo using mouse model. A prior study reported the antimycobacterial activity of GEM in bacteriological medium and in macrophages [54]. Importantly, we found that GEM did not exert a direct antimycobacterial effect against Mabc. However, GEM treatment showed a significant antimicrobial effect against Mabc infection in vitro and in vivo, depending on PPARα. For in vivo treatment, GEM was administered to mice at a dose of 15 mg/kg of body weight/day via gavage, as previously described [55]. In adult humans, GEM is prescribed at a dose of 600-1200 mg/day, therefore dose used in our study (15 mg/kg) is human equivalent dose to mice [55]. Mechanistically, we found that GEM treatment specifically upregulated TFEB nuclear translocation in BMDMs during Mabc infection. Moreover, the GEM-induced antimicrobial response was significantly attenuated in TFEB-knockdown BMDMs. These data suggest the PPARα agonist GEM as a promising target for development of host-directed therapy against Mabc infection, since long term use of antibiotics may cause multidrug resistance in NTM infection by blocking lysosomal acidification and autophagic degradation [56]. Thus, GEM has two mechanisms of action: induction of an antimycobacterial effect and activation of the PPARα-TFEB axis. Further investigation of the effect of GEM on host defense against multidrug-resistant Mabc strains is needed.