Immunomodulators targeting MARCO expression improve resistance to postinfluenza bacterial pneumonia

Abstract

Downregulation of the alveolar macrophage (AM) receptor with collagenous structure (MARCO) leads to susceptibility to postinfluenza bacterial pneumonia, a major cause of morbidity and mortality. We sought to determine whether immunomodulation of MARCO could improve host defense and resistance to secondary bacterial pneumonia. RNAseq analysis identified a striking increase in MARCO expression between days 9 and 11 after influenza infection and indicated important roles for Akt and Nrf2 in MARCO recovery. In vitro, primary human AM-like monocyte-derived macrophages (AM-MDMs) and THP-1 macrophages were treated with IFNγ to model influenza effects. Activators of Nrf2 (sulforaphane) or Akt (SC79) caused increased MARCO expression and a MARCO-dependent improvement in phagocytosis in IFNγ-treated cells and improved survival in mice with postinfluenza pneumococcal pneumonia. Transcription factor analysis also indicated a role for transcription factor E-box (TFEB) in MARCO recovery. Overexpression of TFEB in THP-1 cells led to marked increases in MARCO. The ability of Akt activation to increase MARCO expression in IFNγ-treated AM-MDMs was abrogated in TFEB-knockdown cells, indicating Akt increases MARCO expression through TFEB. Increasing MARCO expression by targeting Nrf2 signaling or the Akt-TFEB-MARCO pathway are promising strategies to improve bacterial clearance and survival in postinfluenza bacterial pneumonia.

secondary bacterial pneumonia contributes to the high mortality from seasonal and pandemic influenza (23, 80). The most common pathogens identified include Streptococcus pneumoniae, Staphylococcus aureus, and Hemophilus influenzae. Susceptibility to superimposed bacterial infection is highest approximately 4–14 days after influenza onset (60). Multiple mechanisms underlie increased susceptibility (8, 15, 19, 22, 27, 33, 38, 41, 43–45, 49, 50, 65, 66, 73, 79, 81, 82, 85, 91), including influenza virus-induced interferon-γ (IFNγ), which reduces expression of the phagocytic receptor macrophage receptor with collagenous structure (MARCO) on alveolar macrophages (AMs) (73).

MARCO is a highly conserved trimeric class A scavenger receptor (7) that mediates phagocytosis of unopsonized particles and bacteria, thereby playing a key role in innate host defense (5). Upregulation of MARCO expression enhances bacterial binding and phagocytosis and alters cytokine expression (51). Dysregulation of MARCO expression leads to impaired phagocytosis and bacterial clearance, as illustrated by increased lung injury and inflammation in MARCO^−/− mice exposed to pathogens or particulates (5, 11, 76). Therefore, we might anticipate that immunomodulators upregulating MARCO expression would improve innate immune function and enhance bacterial clearance in postinfluenza bacterial pneumonia. Hence, we analyzed the dynamics of MARCO expression after influenza infection to identify novel immunomodulatory strategies. A striking recovery and increase in MARCO were observed between days 9 and 11 after influenza infection. RNAseq transcriptome profiling of AMs on these days showed that transcription factor E-box (TFEB) and nuclear factor erythroid 2-related factor 2 (Nrf2) were among the most overrepresented transcription factors regulating differentially expressed genes during the recovery period of postinfluenza MARCO expression.

Sulforaphane (SFN), an isothiocyanate present in cruciferous vegetables, increases macrophage MARCO expression and antibacterial function through Nrf2 signaling (25, 77). We investigated the effects of SFN on postinfluenza host defenses. Treatment with SFN increased MARCO expression and phagocytosis in IFNγ-treated AM-MDMs and significantly improved bacterial clearance and survival in a mouse model of secondary pneumococcal pneumonia.

TFEB is an evolutionally conserved master gene of lysosomal biogenesis, autophagy, and lysosomal exocytosis (46, 61, 63) and plays an important role in regulating host defenses against pathogens (83) and responses to nutritional stress (64). Transcriptome profiling suggested important roles for Akt and the transcription factor E-box (TFEB) in recovery of MARCO expression. Treatment with the Akt activator SC79 improved MARCO expression and bacterial phagocytosis in IFNγ-treated AM-MDMs. However, the effects of Akt-induced MARCO expression in IFNγ-treated AM-MDMs were blocked after TFEB knockdown with siRNAs. Akt activation also improved survival of mice with secondary bacterial pneumonia. These observations identify a novel role for TFEB in the postinfluenza host response as a mediator of Akt activation-induced MARCO upregulation.

Our data identify novel regulators of AM MARCO expression after influenza and broaden the range of potential immunomodulatory strategies to improve host resistance to postinfluenza secondary pneumonia.

METHODS

Cells.

Human alveolar macrophage (AM)-like monocyte-derived macrophages (AM-MDMs) were prepared as described previously (13, 71). Briefly, human monocytes (New York Biologics, Southampton, NY) were isolated from discarded normal blood by negative selection using RosetteSep (StemCell Technologies, Vancouver, BC, Canada) and matured into AM-like macrophages (1, 88) by 10–11 days of culture in Roswell Park Memorial Institute-1640 (RPMI-1640) medium containing 2 mM l-glutamine, 20 mM HEPES, 1 mM sodium pyruvate, 4,500 mg/l glucose, 10% fetal bovine serum, 100 IU/ml penicillin, 100 µg/ml streptomycin, and 20 ng/ml human granulocyte/macrophage colony-stimulating factor (GM-CSF; Peprotech, Rocky Hill, NJ). B6 cells are an immortalized cell line derived from alveolar macrophages obtained from normal C57Bl6 mice, as detailed in a previous report (89).

Human monocyte cell line THP-1 cells (ATCC TIB-202), were grown in RPMI-1640 medium containing 2 mM l-glutamine, 20 mM HEPES, 1 mM sodium pyruvate, 4,500 mg/l glucose, 1500 mg/l sodium bicarbonate, 10% fetal bovine serum, and 1% of penicillin-streptomycin. THP-1 cells were differentiated into macrophages by growing in media containing 100 nM phorbol 12-myristate 13-acetate (Sigma-Aldrich) and 20 ng/ml GM-CSF for 24 h.

Bacteria.

Green fluorescent protein (GFP)-expressing Staphylococcus aureus strain RN6390 was prepared as described previously (13). Streptococcus pneumoniae serotype 3 (strain ATCC6303) was cultured on 5% sheep blood-supplemented agar plates (VWR International, Radnor, PA). Bacteria were grown overnight and resuspended in sterile phosphate-buffered saline (PBS) before infection. Concentration of the bacterial suspension was determined by measuring the optical density at A₆₀₀ with a spectrophotometer and by colony-forming units (CFUs).

Immunomodulator assays.

AM-MDMs were prepared from primary human monocytes as, described previously (13, 71). AM-MDMs or THP-1 cells were treated with human IFNγ (10, 20, or 50 IU/ml; Peprotech) for 20–24 h to simulate effects of influenza infection and simultaneously incubated with or without R,S-sulforaphane (5 or 10 µM; LKT Laboratories, St. Paul, MN) or with or without SC79 (2, 5, or 8 µg/ml; EMD Millipore, Billerica, MA) before quantitation of MARCO expression or challenge with GFPexpressing S. aureus strain RN6390.

For the in vivo study of immunomodulator effects in postinfluenza bacterial pneumonia, wild-type C57BL/6 male mice were treated intraperitoneally with sulforaphane (20–25 mg·kg⁻¹·mouse⁻¹) or subcutaneously with SC79 (20 mg·kg⁻¹·day⁻¹) on days 6–9 or 10 after influenza infection.

For the in vitro study of the EGCG effect in postinfluenza bacterial pneumonia, AM-MDMs or THP-1 cells were treated with human IFNγ (10, 20, or 50 IU/ml; Peprotech) for 20–24 h to simulate influenza infection and simultaneously incubated with or without EGCG (10 or 50 µM, Sigma-Aldrich) before quantitation of MARCO expression or challenge with GFP-expressing S. aureus strain RN6390 by scanning cytometry.

Bronchoalveolar lavage.

After euthanasia, the mouse trachea was cannulated with a 20-gauge catheter for bronchoalveolar lavage (BAL). For colony-forming unit studies, BAL was performed on day 2 after S. pneumoniae infection. Mouse lungs were lavaged with a total of 4 ml of PBS. Serial dilutions of the lavage fluid were spread on blood agar plates and incubated for 17 h at 37°C before CFU counting. For cytokine/chemokine studies and collection of alveolar macrophages, mouse lungs were first lavaged with 0.5 ml of PBS via cannulation, followed by five lavages with 0.7 ml of PBS. The first 1.2 ml of the total 4-ml lavage fluid was centrifuged at 200 g for 10 min at 4°C. Supernatants from BAL fluid (BALF) collected on days 3, 5, 7, 9, and 11 after influenza infection were used for cytokine/chemokine analysis. Cell pellets were resuspended with the cells in the remaining 2.8-ml lavage for each sample.

Cytokine analysis.

Cytokines/chemokines in mouse BALF after influenza virus strain PR8 infection were measured by using the Mouse Cytokine Array/Chemokine Array 32-plex Panel (Eve Technologies, Calgary, AB, Canada). The 32 cytokines/chemokines include eotaxin, granulocyte colony-stimulating factor, granulocyte macrophage colony-stimulating factor (GM-CSF), IFNγ, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12 (p40), IL-12 (p70), IL-13, IL-15, IL-17A, interferon-γ-induced protein (IP-10), ketatinocyte chemoattractant, leukemia inhibitory factor (LIF), lipopolysaccharide-induced CXC chemokine, monocyte chemotactic protein 1, macrophage colony-stimulating factor, monokine induced by interferon-γ (76), macrophage inflammatory protein (MIP-1α), MIP-1β, MIP-2, regulated on activation, normal T cell expressed and secreted, tumor necrosis factor-α, and vascular endothelial growth factor. The sensitivity of the aforementioned cytokines in the panel ranged from 0.07 to 15.85 pg/ml. Interferon-γ concentrations in BALF were also measured by using the mouse IFNγ ELISA MAX Standard Kit (BioLegend, San Diego, CA) per the manufacturer’s protocol.

RNA sequencing analysis.

RNA sequencing was conducted at the Bauer Center Sequencing Core at Faculty of Arts and Sciences Center for Systems Biology, Harvard University. Ribo-depleted RNA was isolated from total RNA of purified murine alveolar macrophages collected on postinfluenza days 9 and 11. Automated library preparation was performed using the Apollo 324 System (WaferGen Biosystems, Fremont, CA) to generate a library of 100-base pair single-end reads for RNA sequencing. DNA fragments were purified with PCRClean DX Kit from Aline Biosciences (Woburn, MA). Samples were pooled in equal molar amounts and sequenced with the Illumina HiSeq 2500 System (Illumina, San Diego, CA).

For data filtering and mapping of reads, Trim Galore (http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) was used to quality trim raw reads and adapters with a minimum quality score of 30 and a minimum sequence length of 80. FASTX-Toolkit (http://hannonlab.cshl.edu/fastx_toolkit/) was then used to remove the first 10 bases. Remaining reads were mapped to the reference mouse genome from the Ensembl database using the “sensitive” preset parameters in Bowtie 2 (36). A count table of mapped reads was created for each gene, and reads per kilobase of transcript per million mapped reads (RPKM) were calculated from the Bowtie 2 output using the generalized fold change (GFOLD) algorithm (17). The R package of DESeq (3) was used to identify differentially expressed genes and to calculate adjusted P values for comparisons.

Two replicates from each group were analyzed using Qiagen’s Ingenuity Pathway Analysis (Qiagen) and MetaCore (Thomson Reuters) for identification of overrepresented transcription factors and functional analysis of differentially expressed genes. The sequence data are available through the NCBI database using the accession no. PRJNA311622.

For analysis using a Connectivity Map approach (35), we used the Touchstone database available at clue.io/touchstone. The top differentially expressed genes (DEGs) passing a filter of adjusted P value < 0.01 were used, resulting in submission of 476 and 428 up- and downregulated DEGs, respectively, for analysis. The results table was sorted by percentile rank based on their ability to cause a gene expression pattern most similar to the gene signature found by comparing DEGs on day 11 vs. day 9 macrophages.

Quantitative reverse transcription-polymerase chain reaction.

Total RNA was extracted according to the manufacturer’s manual using the Qiagen (Valencia, CA) RNeasy Micro Kit. For each sample of fluorescence-activated cell sorting (FACS)-sorted mouse alveolar macrophages, 150 ng of RNA was reverse transcribed into cDNA using Applied Biosystems High Capacity cDNA Reverse Transcription Kits (Grand Island, NY). For each sample of AM-MDMs or THP-1 cells, 1 µg of RNA was reverse transcribed. The quantitative reverse transcription-PCR (qRT-PCR) was performed on an Applied Biosystems 7300 Real Time PCR System with amplification cycles of 95°C for 10 min, 40 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 10 s.

Immunoblot analysis.

AM-MDMs were washed with cold phosphate-buffered saline and lysed with protease and phosphatase inhibitors in 1% NP-40 lysis buffer for 30 min on ice. Cell lysates were resolved on NuPAGE Novex 4–12% Bis-Tris Protein Gels (Life Technologies, Grand Island, NY), and transferred to nitrocellulose membranes. Antibodies used for immunoblot analysis included from Cell Signaling Technology were as follows: phosphor-Akt (Ser⁴⁷³, rabbit mAb no. 4060), Akt (rabbit mAb no. 4685), and β-actin; poluclonal rabbit anti-MARCO, sc-68913, was obtained from Santa Cruz Biotechnology. Immunoreactive bands were detected by chemiluminescence. Densitometric analysis of the immunoblots was performed using ImageJ (National Institutes of Health).

Stable transduction of THP-1 cells.

MISSION Lentiviral Transduction Particles with human TFEB shRNA (Sigma-Aldrich, St. Louis, MO) or human TFEB (BC032448.1) ORF cDNA lentiviral particles (GeneCopoeia, Rockville, MD) were added to 1 × 10⁶ human THP-1 cells at a multiplicity of infection of 0.3 after cells were incubated in 5 µg/ml hexadimethrine bromide for 10 min. Cells were then spun at 1,000 g for 90 min at 30°C and incubated at 37°C for 5 h before fresh RPMI-1640 medium was added. Following overnight incubation at 37°C, cells were washed and grown in culture media containing 5 µg/ml puromycin.

Transfection of human monocyte-derived macrophages.

AM-MDMs were transfected with ON-TARGETplus Human TFEB siRNA or ON-TARGETplus Non-targeting Pool (GE Healthcare Dharmacon, Lafayette, CO) using PolyJet DNA In Vitro Transfection Reagent (SignaGen Laboratories, Rockville, MD) according to the manufacturer’s protocol.

Mouse model of postinfluenza bacterial pneumonia.

Eight- to nine-week-old MARCO^−/− (4) or wild-type control male mice on a C57BL/6 background (The Jackson Laboratory, Bar Harbor, ME) were used in the in vivo model of postinfluenza bacterial pneumonia. Animals were cared for in accordance with the Institutional Animal Care and Use Committee guidelines. Experimental procedures on animals were conducted after approval by the Harvard Medical Area Standing Committee on Animals. For the postinfluenza pneumonia model, mice were intranasally inoculated with 1 plaque-forming unit of influenza virus strain A/Puerto Rico/8/34 (PR8; Virasource, Raleigh-Durham, NC) per mouse in 25 μl of PBS after anesthesia with ketamine (200 mg/kg) and xylazine (20 mg/kg). On day 7 after influenza infection, mice were intranasally infected with 300–500 CFUs of S. pneumoniae serotype 3 in 25 μl of PBS after anesthesia. Body weight and morbidity were monitored after infections.

Scanning cytometry analysis of MARCO expression and bacterial uptake.

For analysis of treatment effects on MARCO expression, AM-MDMs were incubated with mAbs PLK-1 (5) or IgG3 isotype control at 4°C after Fc receptor blockade. Cells were then washed and stained with Alexa Fluor 594 F(ab′)2 fragment of goat anti-mouse IgG (Invitrogen, Grand Island, NY). Dead cells were stained with the Invitrogen LIVE/DEAD Fixable Green Dead Cell Stain Kit per the manufacturer’s instructions. After fixation with 4% formaldehyde, cells were stained with Hoechst nuclear stain and CellMask blue stain (Invitrogen).

For analysis of treatment effects on AM-MDM uptake of GFP-labeled bacteria, extracellular bacteria were stained with Alexa Fluor 594-labeled monoclonal antibody to S. aureus (ABIN4356218; Novus Biologicals) or lysed with lysostaphin (20 μg/ml, 30 min). For MARCO inhibition experiments, cells were incubated in 200 μg/ml of polyinosinic acid [poly(I)], a class A scavenger receptor inhibitor, at 37°C for 30 min before bacterial infection. Alternatively, GFP-S. aureus suspensions were incubated with or without MARCO blocking peptides (Santa Cruz Biotechnology, Dallas, TX) for 30 min before infection. Dead cells were stained with the Invitrogen LIVE/DEAD Fixable Red Dead Cell Stain Kit. After fixation with 4% formaldehyde, cells were stained with Hoechst nuclear stain and CellMask blue stain (Invitrogen).

Scanning cytometry BD Pathway 855 High-Content Bioimager (BD Biosciences, San Jose, CA) was used to acquire confocal fluorescence images of MARCO surface labeling and binding and uptake of bacteria, and image data were analyzed using MATLAB image analysis software (13).

Fluorescence-activated cell sorting.

Cells collected from BAL on days 3, 5, 7, 9, and 11 after influenza infection of C57BL/6 mice were stained with fluorescein isothiocyanate (FITC)-conjugated anti-MARCO antibodies (AbD Serotec, Raleigh, NC), Alexa Fluor 647-conjugated F4/80 antibodies (BioLegend), PE-Cy7-conjugated CD11c+ antibodies (BioLegend), and their corresponding isotypes after Fc Receptor blockade. F4/80+CD11c+ cells were sorted using the BD FACSAria II cell sorter (BD Biosciences, San Jose, CA) for RNA extraction.

Chromatin immunoprecipitation.

The SimpleChIP (chromatin immunoprecipitation) Enzymatic Chromatin IP Kit (Magnetic Beads) no. 9003 from Cell Signaling Technology (Danvers, MA) was used for chromatin immunoprecipitation according to the manufacturer’s protocol. Briefly, AM-MDMs were formalin-fixed, treated with micrococcal nuclease, and sonicated. Chromatin from the AM-MDMs was immunoprecipitated with anti-TFEB antibody and its corresponding isotype (Cell Signaling Technology). Immunoprecipitated nuclear DNA was extracted and analyzed by PCR, with primers targeting the human MARCO upstream promoter area. The MARCO primers used for the ChIP assay were as follows: forward, 5′-CCCTGGTAGAGATGCCTGAC-3′; reverse, 5′-GAAGCCTTGCTCTGAACCAC-3′.

Statistical analysis.

Statistical analyses were performed using GraphPad Prism 6 (GraphPad Software, San Diego, CA). Results are expressed as means ± SD. Student’s t-test was performed to compare treatment effects between two groups and one-way analysis of variance (ANOVA) with Bonferroni adjustment for comparison between multiple groups. Mortality of mice was analyzed by the Kaplan-Meier and log rank methods.

RESULTS

MARCO mRNA expression of sorted alveolar macrophages after influenza virus infection.

Lung cells from influenza-infected mice were collected by lavage on days 3, 5, 7, 9, and 11 after infection. Lavaged F4/80+CD11c+ cells were sorted as alveolar macrophages (6, 34) (Fig. 1A), and Marco expression was measured by qRT-PCR. Corresponding to the increased IFNγ levels on postinfluenza days 7 and 9 (Fig. 1B), Marco expression was downregulated on days 7 (0.46-fold, P < 0.0001; Fig. 1C) and 9 (0.51-fold, P < 0.0001). However, there was a surprising 4.86-fold increase (P < 0.0001) in Marco expression between days 9 and 11. Moreover, Marco expression was upregulated 2.48-fold (P < 0.0001) on day 11 compared with uninfected controls.

Fig. 1.

Alveolar macrophage (AM) receptor with collagenous structure (MARCO) expression is spontaneously restored and increased in the later phase of influenza virus infection. A: bronchoalveolar lavage samples were collected after influenza infection, and F4/80+CD11c+ cells (alveolar macrophages) were sorted and subjected to RNA extraction and quantitative (q)RT-PCR for evaluation of MARCO expression. B: ELISA of bronchoalveolar lavage fluids collected from influenza-infected mice showed that IFNγ levels increased significantly on days 7 and 9 compared with uninfected controls; n = 3–5/group. C: qRT-PCR of sorted alveolar macrophages showed that MARCO expression was downregulated on days 7 (P < 0.0001) and 9 (P < 0.0001) postinfluenza virus infection but upregulated on day 11 (P < 0.0001) compared with uninfected controls; n = 3–9/group. Data are shown as fold change after normalization to the housekeeping gene TATA box-binding protein (TBP) expression. P values in in B and C are as follows: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ●, ■, ▲,▼, ◆, and ○ correspond to samples collected from groups defined in the x-axes.

To explore the basis for these changes in Marco expression, levels of large panels of cytokines were measured in lavage fluids using a multiplexed array assay. The results are summarized in Fig. 2 and identify several cytokines whose expression declines from day 7 onward. Consistent with prior studies (73), IFNγ exhibited the greatest increase on day 7 compared with day 5 (261-fold) and also when day 7 was compared with day 9 or day 11 (524- and 66-fold, respectively). A subset of the cytokines assayed showed a similar pattern (e.g., IL-4, 10, 12p70, fold change; data not shown), but no single cytokine showed a unique change in magnitude or pattern that might easily explain (by itself) the marked increase in Marco expression between days 9 and 11.

Fig. 2.

Cytokine multiplex assay of bronchoalveolar lavage fluids from influenza-infected C57BL/6 mice. Day 0: uninfected control; days 3, 5, 7, 9, and 11: postinfluenza. There were 3 mice in each group, represented by individual symbols (■, ▲,▼, ◆, and ○); bars show mean values.GM-CSF, granulocyte/macrophage colony-stimulating factor; G-CSF, granulocyte colony-stimulating factor; KC, ketatinocyte chemoattractant; LIF, leukemia inhibitory factor; MCP-1, monocyte chemotactic protein-1; M-CSF, macrophage colony-stimulating factor; MIG, monokine induced by interferon-g; MIP-1α, macrophage inflammatory protein-1α; MIP-2, macrophage inflammatory protein 2; RANTES, regulated on activation, normal T cell expressed and secreted.

RNA sequencing analysis of sorted alveolar macrophages on days 9 and 11 after influenza infection.

To explore potential mechanisms for upregulation of Marco expression more broadly, RNAseq analysis was performed on F4/80+CD11c+ alveolar macrophages purified by flow sorting from mice on postinfluenza days 9 and 11; 1,537 differentially expressed genes (DEGs) were identified between day 9 and day 11 samples, using criteria of adjusted P < 0.01; (Supplemental Table S1; Supplemental Material for this article is available online at the AJP-Lung Cellular and Molecular Physiology website). Notably, these included Marco, v-Akt murine thymoma viral oncogene homolog 1 (Akt1 or Akt), mitogen-activated protein kinase 3 (Mapk3), and another Nrf2-regulated gene, heme oxygenase 1 (Hmox1). Enrichment analysis using MetaCore showed that the candidate genes were associated with process networks relevant to the resolution phase of influenza (Supplemental Table S2) and also that NRF2 and TFEB are among the most overrepresented transcription factors (based on binding sites in genes differentially expressed in AMs between day 9 and day 11; Supplemental Table S3). Finally, analysis of the DEGs using the CLUE portal to query the Connectivity Map Touchstone data set (clue.io/touchstone) identified agents that cause gene expression changes that correlate strongly with the gene signature linked to upregulated MARCO on day11. Six of the top 10 compounds (of the 2,837 compounds evaluated) activate NRF2 [by rank: celastrol (no. 1) (16), topoisomerase inhibitors SN38 and topotecan (nos. 5 and 6, respectively) (54), mycophenolate (7) (75), triptolide (no. 8) (90), and trichostatin (no. 9) (9)]. We further investigated immunomodulation of both of these targets, focusing on sulforaphane activation of NRF2 (47) and Akt-mediated activation of TFEB.

Sulforaphane increases MARCO expression of IFNγ-treated human macrophages.

To model the effects of influenza in vitro, we treated primary AM-MDMs with IFNγ for 24 h, which resulted in downregulation of MARCO expression (0.54 ± 0.20-fold change vs. control, P < 0.001; Fig. 3, A and B). SFN treatment upregulated MARCO expression and diminished the IFNγ-mediated reduction in MARCO expression (0.92 ± 0.25 vs. 0.54 ± 0.20-fold change, P < 0.01). Cell death following IFNγ treatment was also significantly reduced in SFN-treated groups (SFN = 10 µM) compared with untreated controls (1.3 ± 1.5 vs. 4.5 ± 2.0%, P < 0.01, Fig. 3, A and C). Similarly, treatment of differentiated human THP-1 macrophages with IFNγ (5–50 IU/ml) resulted in a dose-dependent downregulation of MARCO expression and increased cell death, which were reversed by treatment with SFN (Fig. 4, A and B).

Fig. 3.

Sulforaphane (SFN) significantly improves MARCO expression and bacterial phagocytosis and reduces cell death in IFNγ-treated human monocyte-derived macrophages. A: representative images of scanning cytometry analysis of AM-like monocyte-derived macrophages (AM-MDMs) after treatment with IFNγ and/or SFN. Red, MARCO on the cell surface; green, dead cells. B: quantitation of results of scanning cytometry analysis of AM-MDMs treated with or without IFNγ and SFN. Fluorescence index (FI) = %positive × mean fluorescence intensity of all cells (MFI); no. of donors = 6. Data are represented as means ± SD. C: SFN treatment reduced cell death in IFNγ-treated AM-MDMs; no. of donors = 6. Data are represented as means ± SD. D: representative images of scanning cytometry analysis of AM-MDMs after treatment with IFNγ and/or SFN, followed by GFP-S. aureus infection. Extracellular bacteria were labeled with red fluorescent dye. E: individual donor results showing decreased phagocytosis in response to IFNγ, improvement by SFN, and blockade of SFN effect with the scavenger receptor inhibitor poly(I); no. of donors = 6. Colors indicate different donors. F: individual donor results showing decreased phagocytosis in response to IFNγ, improvement by SFN, and blockade of SFN effect with the MARCO peptide for competition studies; no. of donors = 6. Colors indicate different donors. P values in B, C, E, and F are as follows: *P < 0.05, **P < 0.01, and ***P < 0.001. NS, not significant.

Fig. 4.

Sulforaphane (SFN) significantly improves MARCO expression and reduces cell death in IFNγ-treated THP-1 macrophages. A: scanning cytometry analysis showed that IFNγ reduced MARCO expression on THP-1 cells in a dose-dependent manner, but simultaneous treatment with SFN upregulated and/or restored MARCO expression in IFNγ-treated THP-1 cells. Fluorescence index (FI) of MARCO expression = %THP-1 cells expressing MARCO × MFI (mean fluorescence intensity of all cells). Data represent means of 4 replicates from 2 independent experiments. B: SFN treatment also reduced cell death in the IFNγ-treated THP-1 cells.

Sulforaphane improves bacterial phagocytosis in IFNγ-treated AM-MDMs through MARCO upregulation.

AM-MDMs were challenged with S. aureus following ∼24 h of IFNγ pretreatment. Phagocytosis of fluorescent GFP-labeled S. aureus by AM-MDMs was impaired by IFNγ treatment (P < 0.05; Fig. 3, D–F) but significantly improved when treated simultaneously with SFN (P < 0.01). Cell death following IFNγ and S. aureus infection was also reduced by concurrent SFN treatment (data not shown).

To further delineate the causal link between sulforaphane treatment, MARCO expression, and bacterial phagocytosis, we treated AM-MDMs with either the broad class A scavenger receptor blocker polyinosinic acid [poly(I)] or a specific MARCO-blocking peptide before S. aureus infection. The rescue effect of SFN on bacterial phagocytosis after IFNγ treatment was significantly inhibited by both poly(I) (5.19 bacteria/cell, 95% CI 2.62 to 7.76, vs. 3.25, 95% CI 0.47 to 6.04, P < 0.01; Fig. 3E) and the specific MARCO-blocking peptide (5.21, 95% CI 4.39 to 6.02, vs. 3.04, 95% CI 1.75 to 4.33, P < 0.05; Fig. 3F). Taken together with the previous findings, these results indicate that SFN improves bacterial phagocytosis in IFNγ-treated macrophages through upregulation of the class A scavenger receptor MARCO. In contrast to uniform results seen with the THP-1 cell line, considerable donor-to-donor variability was seen with primary AM-MDM cells. However, AM-MDMs from all donors showed a similar pattern, with IFNγ causing a reduction in bacterial phagocytosis that was reversed by SFN and MARCO blockade resulting in inhibition of phagocytosis that was not abrogated by SFN (Fig. 3, E and F).

We also tested the effects of another Nrf2 activator, epigallocatechin gallate (EGCG) (24, 86), on MARCO expression and bacterial phagocytosis in AM-MDMs from two donors. EGCG also reversed declines in MARCO expression and S. aureus phagocytosis caused by IFNγ (Fig. 5).

Fig. 5.

Epigallocatechin gallate (EGCG) increases MARCO expression and bacterial phagocytosis in IFNγ-treated cells. A: the Nrf2 activator EGCG improved MARCO expression in IFNγ-treated AM-MDMs. AM-MDMs were incubated with or without IFNγ (20 or 50 IU/ml) and EGCG (50 µM). Data shown are mean fold changes in MFI (mean fluorescence intensity) in cells from 2 donors. B: EGCG improved phagocytosis of GFP-S. aureus in IFNγ-treated macrophages. Fluorescence index (FI) of phagocytosis = %cells with phagocytosed bacteria × no. of bacteria/cell.

Sulforaphane improves bacterial clearance and host survival in postinfluenza bacterial pneumonia.

We investigated the effects of SFN using the in vivo model of postinfluenza bacterial pneumonia summarized schematically in Fig. 6A. To study the effects of SFN on bacterial clearance, mice were subcutaneously injected with SFN or vehicle on postinfluenza days 6–9 (i.e., days −1 to 2 post-S. pneumoniae infection). SFN treatment significantly improved clearance of S. pneumoniae, which was measured as bacterial colony-forming units (CFU) in bronchoalveolar lavage fluid, compared with untreated controls (log₁₀ CFU = 6.09 ± 0.60 vs. 7.57 ± 0.20, P = 0.036; Fig. 6B). To study the effects of SFN on host survival in postinfluenza bacterial pneumonia, mice were injected with SFN or vehicle on postinfluenza days 6–9 or 10 (i.e., days −1 to 2 or 3 post-S. pneumoniae infection; Fig. 6A). SFN-treated mice exhibited lower body weight loss and faster weight recovery after postinfluenza bacterial infection compared with controls (data not shown). Notably, SFN improved survival in wild-type (53.3 vs. 5.9%, P < 0.0001; Fig. 6C) but not Marco^−/− mice (Fig. 6D), indicating that Marco expression was required for the beneficial effect of SFN. Median survival times for SFN-treated and vehicle-treated wild-type mice were >12 and 4.7 ± 0.2 days, respectively.

Fig. 6.

Sulforaphane (SFN) improves clearance of S. pneumoniae and host survival in wild-type C57BL/6 mice with postinfluenza bacterial pneumonia. A: summary of secondary pneumonia model starting infection with influenza virus strain A/Puerto Rico/8/34 (PR8) on day 0 (d0), followed by S. pneumoniae infection on day 7. Mice received daily treatments with either vehicle or SFN from postinfluenza days 6 to 9 (d6, d7, d8, and d9) or 10 (d10). B: treatment with SFN significantly improved bacterial clearance, measured as colony-forming units (CFU) in bronchoalveolar lavage fluids on day 9 in postinfluenza pneumococcal pneumonia (P = 0.036). C and D: survival studies of wild-type C57BL/6 mice (C) and MARCO knockout (D) mice with postinfluenza bacterial pneumonia. SFN significantly improved host survival in wild-type mice (P < 0.0001), but not that of MARCO knockouts. No. of vehicle-treated wild-type mice = 34, no. of SFN-treated wild-type mice = 30, no. of vehicle-treated MARCO knockouts = 20, and no. of SFN-treated MARCO knockouts = 20.

Role of Akt activation on MARCO expression and bacterial phagocytosis in postinfluenza bacterial pneumonia.

Akt and MAPK3 were among the top five genes identified as enriched network object analysis from the Metacore analysis of the RNAseq results [in order, with the number (n) of networks involved in parentheses: ERK1/2 (n = 33), MAPK3 (n = 30), PI3KIA (n = 29), AKT (n = 27), and NF-κB (n = 24)]. Both were also previously found to be altered by influenza (18). qRT-PCR showed that Akt and Mapk3 mRNA expression levels in murine alveolar macrophages gradually declined after influenza infection (Figs. 7, A and B) but were increased on postinfluenza day 11 compared with the reduced levels from day 5 to day 9. The nadirs of Akt and Mapk3 expression occurred on days 7 and 9, respectively. These patterns of Akt and Mapk3 expression were concordant with that of MARCO expression and inversely corresponded to IFNγ levels in BALF.

Fig. 7.

Influenza infection and decreased expression of Akt and MAPK3. Numbers in the x-axis indicate days after PR8 influenza virus infection in wild-type adult male C57BL/6 mice. A: quantitative RT-PCR of sorted alveolar macrophages showed that Akt expression was significantly downregulated on days 5–9 postinfluenza virus infection, but there was no significant difference on day 11 compared with uninfected controls. There were 3–5 mice in each group. Data are represented as means ± SD. B: quantitative RT-PCR of sorted alveolar macrophages showed that MAPK3 expression was significantly downregulated on days 5–9 postinfluenza virus infection. The level of MAPK3 expression increased significantly on day 11 from day 9, although it was still lower than controls. There were 3–5 mice in each group. Data are represented as means ± SD. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Both Akt inhibitor perifosine (10–50 µM; Fig. 8A) (26, 74) and MAPK/ERK inhibitors U-0216 and PD-98059 (data not shown) (2, 53) reduced MARCO expression on AM-MDMs, supporting a role for Akt and MAPK signaling pathways in MARCO expression in human macrophages.

Fig. 8.

Akt signaling is involved in postinfluenza MARCO expression. A: MARCO expression of AM-MDMs in vitro (measured by scanning cytometry) is reduced by the Akt inhibitor perifosine; no. of donors = 3. Data are represented as means ± SD. B: AM-MDMs were treated overnight with IFNγ (20 IU/ml) and/or SC79 (0, 2, 5, and 8 µg/ml). Western blot analysis showed that IFNγ treatment decreased levels of MARCO. The Akt activator SC79 increased MARCO expression in a dose-dependent manner in the presence of IFNγ. C: representative scanning cytometry images showed that SC79 improved MARCO expression in IFNγ-treated AM-MDMs. AM-MDMs were treated simultaneously with 20 IU/ml of IFNγ and 8 µg/ml of SC79 for 20–24 h. Red, MARCO proteins on the cell surface; green, dead cells. D: quantitation results of scanning cytometry analysis of AM-MDMs with or without IFNγ or SC79. Results from scanning cytometry analysis showed that interferon-γ (IFNγ) treatment of human monocyte-derived macrophages reduced MARCO expression in a dose-dependent manner (data with 10 IU/ml IFNγ not shown). Fluorescence index (FI) of MARCO = %positive × mean fluorescence intensity of each cell; no. of donors = 6. Data are represented as means ± SD. P values in A and D are as follows: **P < 0.01 and ****P < 0.0001.

Western blot analysis of AM-MDMs showed that IFNγ treatment decreased both p-Akt and MARCO. Moreover, treatment with the Akt activator SC79 (30) increased MARCO levels in the presence of IFNγ (Fig. 8B). Scanning cytometry also showed that the SC79 increased MARCO expression in IFNγ-treated AM-MDMs (P < 0.001, n = 6; Fig. 8, C and D). Treatment with MAPK/ERK activators failed to change MARCO expression (data not shown). Because the available MAPK activators did not increase MARCO expression in vitro, we chose to focus on Akt. SC79 treatment increased phagocytosis of GFP-S. aureus by IFNγ-treated AM-MDMs (Fig. 9, A and B). Finally, the rescue effect of SC79 in increasing bacterial phagocytosis was significantly inhibited by the class A scavenger receptor poly(I) (Fig. 9C).

Fig. 9.

The Akt activator SC79 improves AM-MDM phagocytosis of S. aureus. A: average no. of phagocytosed GFP-S. aureus per cell after treatment with IFNγ ± SC79. SC79 treatment increased the average no. of phagocytosed GFP-S. aureus per cell in IFNγ-treated AM-MDMs; no. of donors = 6. Colors indicate different donors. B: %cells infected with GFP-S. aureus after treatment with IFNγ and/or SC79. SC79 treatment increased the proportion of cells with phagocytosed bacteria in IFNγ-treated AM-MDMs. No. of donors = 6. Colors indicate different donors. C: the ability of SC79 to increase bacterial phagocytosis after IFNγ treatment was significantly inhibited by the class A scavenger receptor inhibitor poly(I); no. of donors = 4. Colors indicate different donors. P values in A–C are as follows: *P < 0.05 and **P < 0.01. D: summary of secondary pneumonia model starting infection with influenza virus (PR8) on day 0, followed by S. pneumoniae infection on day 7. Mice received daily treatments with either vehicle or SC79 from postinfluenza days 6 to 9 or 10. Data are a combination of 4 independent experiments. Thirteen mice were treated with subcutaneous injection of SC79 in the dose of 20 mg/kg from postinfluenza days 6 to 9. Fifteen mice received subcutaneous injection of SC79 in the dose of 20 mg/kg from postinfluenza day 6 to day 10. There was no significant difference in host survival between 4-day and 5-day SC79 treatments. E: survival studies of wild-type C57BL/6 mice with postinfluenza bacterial pneumonia. No. of vehicle-treated mice = 25; no. of SC79-treated mice = 28.

To study the effects of SC79 on in vivo survival from secondary pneumonia, mice were treated with SC79 (20 mg/kg) or vehicle from days 6 to 9 or 10 (i.e., day −1 to day 2 or 3 post-S. pneumoniae infection) (Fig. 9D). SC79-treated mice had a significantly higher survival rate compared with vehicle-treated mice (25 vs. 4%, P = 0.001; Fig. 9E). The median survival times for SC79-treated and vehicle-treated mice were 6.8 ± 1.1 and 4.9 ± 0.3 days, respectively.

TFEB is an overrepresented transcription factor regulating susceptibility to postinfluenza bacterial pneumonia.

Because TFEB, the most overrepresented transcription factor in our transcriptome data, was also recently identified as an important regulator of innate immunity in C. elegans and in murine macrophages (83), we investigated the role of TFEB in our postinfluenza model. TFEB is known to translocate to the nucleus and bind to E-box sequences (5′-CANNTG-3′) (69). Examination of the MARCO promoter region (−2,000 to +200) showed multiple E-box sequences and thus potential binding sites for TFEB (Fig. 10). To investigate whether TFEB binds to the MARCO promoter, AM-MDMs were subjected to chromatin immunoprecipitation using anti-TFEB antibody. PCR with primers targeting the upstream sequence (−824 to −622) of the human MARCO gene showed enhanced detection consistent with binding of TFEB to the MARCO promoter area (Fig. 11A).

Fig. 10.

Multiple E-box sequences are identified in the MARCO promoter area. A: logo representation (WebLogo 3) of the E-box sequences (CANNTG) within 2,000 base pairs (bp) upstream of the transcription start site of human MARCO gene. B: the CANNTG sites at −2,000 to +200 bp of the transcription start site of the human MARCO gene. Sequences in blue indicate primers used in the chromatin immunoprecipitation (ChIP) assay.

Fig. 11.

transcription factor E-box (TFEB) modulation changes MARCO expression levels. A: ChIP assay of AM-MDMs (± treatment of SC79) showed direct binding of TFEB to the upstream promoter area of human MARCO. B: relative gene expression of TFEB and MARCO in TFEB knockdown THP-1 cells compared with the control transduced cells. Data are shown as fold change after normalization to the housekeeping gene TBP expression. Bars represent the mean, and error bars represent the standard deviation of triplicates from 1 representative experiment of 2. C: relative gene expression of TFEB and MARCO in TFEB-overexpressing THP-1 cells compared with control. Data are shown as fold change after normalization to the housekeeping gene TBP expression. Bars represent the mean, and error bars represent the standard deviation of triplicates from 1 representative experiment of 3. D: representative scanning cytometry images of AM-MDMs with or without IFNγ and/or SC79. AM-MDMs were transfected with human TFEB siRNA or nontargeting pool. E: quantitation results of scanning cytometry analysis of TFEB AM-MDM knockdowns. Akt activator-induced MARCO expression in IFNγ-treated macrophages is blocked in TFEB AM-MDM knockdowns. Fluorescence index (FI) = %positive × mean fluorescence intensity of all cells (MFI); no. of donors = 5. Colors indicate different donors. P values in B, C, and E are as follows: *P < 0.05, **P < 0.01, and ***P < 0.001. TFEB kd, TFEB knockdown; NT, nontarget.

A possible association between MARCO, Akt, and TFEB is suggested by the findings that increased mTORC1 activity downregulates Akt activity and MARCO expression (78, 87) and sequesters TFEB in the cytoplasm through phosphorylation (59). We used RNA interference to knock down TFEB in THP-1 macrophages and AM-MDMs before treatment with the Akt activator SC79 (30). THP-1 cells stably transduced with lentivirus-delivered shRNA against TFEB showed an almost fivefold reduction in MARCO mRNA expression (P = 0.002; Fig. 11B) and a 70% reduction in Hmox1 (data not shown) compared with cells transduced with nontargeting shRNA. Mouse B6 cells (derived from alveolar macrophages) transduced with shRNA against TFEB also showed reduced Marco mRNA expression compared with nontargeting shRNA-transduced controls (data not shown). In contrast, after lentiviral transduction, TFEB-overexpressing THP-1 cells showed a >300-fold increase in Marco expression (P < 0.0001; Fig. 11C) and a fivefold increase in Hmox1 expression (data not shown). To test the effect of TFEB levels on the interaction of Akt and MARCO expression, we treated TFEB knockdown cells with the Akt activator SC79. SC79-induced MARCO expression in IFNγ-treated macrophages was abrogated in primary AM-MDMs transfected with TFEB siRNA (P < 0.01, n = 5; Fig. 11, D and E) and in THP-1 knockdowns (data not shown), suggesting a link between Akt signaling, TFEB nuclear translocation, and MARCO expression (Fig. 12).

Fig. 12.

TFEB mediates Akt activation-induced MARCO expression. A: a schematic illustration depicting postulated activities between Akt, TFEB, and MARCO. Influenza infection stimulates T cell production of IFNγ, which in turn may induce activation of mammalian target of rapamycin complex 1 (mTORC1) and ensuing Akt inhibition. It is also possible that IFNγ inhibits Akt activity through mTORC1-independent pathways. Activated mTORC1 sequesters TFEB in the cytoplasm and also downregulates MARCO expression through Akt inhibition or other downstream effectors, such as c/EBPβ inhibition (87). Akt inhibition results in MARCO downregulation, leading to impaired bacterial phagocytosis and increased susceptibility to bacterial infection. B: treatment with SC79 activates Akt signaling and upregulates MARCO in the presence of IFNγ through the effects of TFEB.

DISCUSSION

Alveolar macrophages are critical for maintaining lung homeostasis and promoting survival following influenza infection (12, 20, 62), including defense against secondary pneumonia. We investigated the role of the AM scavenger receptor MARCO in these defenses to identify potential immunomodulators that could be used to reduce the risk of postinfluenza bacterial pneumonia. Initially, elevated IFNγ levels following influenza infection corresponded to synchronous decreases in AM MARCO expression. However, as IFNγ levels returned to normal on postinfluenza day 11, there was a striking increase in MARCO expression. Transcriptome profiling of purified lung macrophages indicated important roles for NRF2 and Akt/TFEB-related signaling in the unexpectedly high rebound level of MARCO. High levels of IFNγ after influenza virus infection correlated with decreased MARCO expression in vivo and caused impaired phagocytosis in AM-MDMs in vitro. Activating the Nrf2 and Akt signaling pathways with sulforaphane and SC79, respectively, increased MARCO expression and improved bacterial phagocytosis in the presence of IFNγ. In an in vivo model of postinfluenza bacterial pneumonia, both SC79 and sulforaphane improved host survival. The sulforaphane-induced improvement of survival was observed in wild-type but not in MARCO knockout mice, suggesting that MARCO is an indispensable mediator for Nrf2 effects in postinfluenza bacterial pneumonia.

Some limitations of our study merit discussion. Although previous studies have shown that GM-CSF-matured human macrophages (AM-MDMs) used in our in vitro studies resemble primary AMs by numerous criteria (13, 71), and IFNγ treatment of AM-MDMs reproduces the postinfluenza downregulation of lung macrophage MARCO expression and impaired bacterial phagocytosis seen in vivo, it is possible that responses of true primary lung macrophages may differ. In addition, although our mouse model of postinfluenza is similar to that used by many other investigators (20, 28, 41, 67, 68, 72), results may be dependent upon influenza dose as well as the dose and type of bacteria used for secondary infection (29, 38, 45, 70). In our secondary model we did not measure viral load on day 7, the day of secondary bacterial infection. In studies using much higher inocula of virus (designed to cause primary influenza pneumonia), we observed a marked decrease in viral load by day 6 (21). Here, with a much lower inoculum that causes minimal inflammation and weight loss, it is likely that the virus is cleared by day 7. However, whether there remains some residual virus has not been directly addressed in our data. We focused on IFNγ-mediated effects, but our multiplex cytokine results could be interpreted as showing that changes in the levels of several other cytokines (individually or in combinations) may contribute to the striking increase in MARCO levels from days 9 to 11. Another issue that merits mention is our use of S. aureus for in vitro studies, whereas the in vivo studies use S. pneumoniae. Both pathogens are common agents in secondary postinfluenza pneumonia, and both are engaged by MARCO and other scavenger receptors. However, in vivo secondary pneumonia models using pneumococcus may be more relevant since they require much lower inocula (e.g., 500 CFU), whereas inocula of 10⁸ are commonly required for postinfluenza studies with S. aureus. Moreover, using the GFP-S. aureus in vitro allowed the use of fluorescent assays reported, a strategy not possible for pneumococcus given the lack of available fluorescence-tagged S. pneumoniae. Future studies with both in vitro and in vivo analyses using the same organism could add to the understanding of any pathogen-specific aspects. Finally, there is a rich literature that identifies multiple mediators as causal for postinfluenza secondary pneumonia (8, 15, 22, 31–33, 38, 40, 42, 44, 52, 55–58, 72, 79, 82, 85), including type I interferons (37). The proposed multiple mechanisms are not necessarily mutually exclusive and may offer alternative targets for immunomodulator interventions.

Notably, we found that the MARCO rescue effect of Akt activation is dependent upon the transcription factor TFEB, a lysosome and autophagy master regulator. These results are congruent with previous studies in which PI3K inhibitors reduced scavenger receptor-mediated phagocytosis (71). The class A scavenger receptor inhibitor poly(I) blocked the SC79-mediated increase in bacterial phagocytosis, indicating a link between Akt signaling and scavenger receptors. It has previously been shown that IFNγ produced by influenza infection induces hyperactivated mTORC1 (87), which in turn inhibits Akt activity (10) and reduces both MARCO expression and unopsonized phagocytosis (87). It is also possible that IFNγ inhibits Akt activity through undiscovered mTORC1-independent pathways. Taken together, these data suggested that IFNγ production after influenza is associated with attenuated Akt signaling and MARCO expression in alveolar macrophages. The rebound phenomenon of MARCO expression on postinfluenza day 11 may be due to a feedback-based increase of Akt activity from the previously inhibited state. It is worth noting that the magnitude of changes in Akt expression (Fig. 7A) is relatively small. Moreover, phosphorylation state rather than expression level often regulates Akt activity. Future studies are warranted to further delineate these potentially important details.

Notably, both TFEB and Nrf2 are among the overrepresented transcription factors regulating differentially expressed genes during the recovery period of postinfluenza MARCO expression (Supplemental Table S3), indicating that both are essential in restoration of homeostasis and/or host defense against postinfluenza complications. There may be some interplay between these two transcription factors since 1) as we have shown here, TFEB mediates Akt signaling in MARCO upregulation; 2) activation of either insulin receptor (84) or epidermal growth factor receptor tyrosine kinase (48) induces Nrf2 activation through phosphatidylinositol 3-kinase (PI3K)/Akt signaling; 3) PI3K/Akt signaling pathway may be activated by sulforaphane, leading to increased expression of Nrf2-regulated genes (14, 39); and 4) THP-1 cells stably transduced with shRNA against TFEB showed reduced expression of Hmox1, an Nrf2-regulated gene, compared with the nontargeting shRNA control, whereas TFEB-overexpressing THP-1 cells showed increased Hmox1 expression when compared with the control. We speculate that the ability of sulforaphane to also reverse IFNγ-mediated cell death (Figs. 3C and 4A) may also reflect activation of prosurvival Akt signaling, but our data do not directly test this postulate. Taken together, these data suggest that TFEB, Nrf2, and MARCO expression are correlated with each other and all associated with Akt signaling.

The mechanism(s) by which MARCO expression is inhibited by IFNγ is unclear. Previous data showed that tyrosine kinases, protein kinase C, PI3K signaling, and the MAPK pathway are involved in macrophage scavenger receptor-mediated phagocytosis (71). Interactions between Nrf2 and PI3K Akt signaling are known (48, 84). These data indicate that Nrf2 signaling and regulation of MARCO expression are coupled to both the MAPK and PI3K/Akt signaling pathways, but the specific pathways for IFNγ effects need further characterization.

Figure 12 summarizes our working hypotheses regarding the role of TFEB in mediating Akt-dependent signaling to upregulate MARCO, the major opsonin-independent phagocytic receptor on alveolar macrophages, in the postinfluenza alveolar milieu. Our data also suggest that expression of the scavenger receptor MARCO may be linked to additional TFEB-regulated autophagy and/or lysosomal genes. Our results show that, in addition to Nrf2, Akt signaling and TFEB may be effective therapeutic targets in postinfluenza bacterial pneumonia.

GRANTS

This research was supported by National Institutes of Health Grants ES-00002 and HL-115778.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.W., G.M.D., and A.S.B. performed experiments; M.W., J.G.G., G.M.D., A.S.B., R.K.T., S.B., and L.K. analyzed data; M.W., J.G.G., G.M.D., R.K.T., S.B., and L.K. interpreted results of experiments; M.W., G.M.D., and L.K. prepared figures; M.W. and L.K. drafted manuscript; M.W., R.K.T., S.B., and L.K. edited and revised manuscript; M.W., J.G.G., G.M.D., A.S.B., R.K.T., S.B., and L.K. approved final version of manuscript; L.K. conceived and designed research.