Transforming Growth Factor β1 Suppresses the Type I Interferon Response and Induces Mitochondrial Dysfunction in Alveolar Macrophages

Abstract

Transforming growth factor β1 (TGFβ1) is a pleiotropic cytokine with an established role in fibrosis; however, the immunosuppressive effects of TGFβ1 are less characterized. Elevated levels of TGFβ1 are found in patients with acute and chronic lung diseases, and the underlying disease processes are exacerbated by respiratory viral infections. The alveolar macrophage is the first line of cellular defense against respiratory viral infections and its response to infections is dependent on environmental cues. Using the mouse alveolar macrophage line, MH-S, and human CD14⁺ monocyte-derived macrophages, we examined the effects of TGFβ1 on the Type I interferon (IFN) antiviral response, macrophage polarization, and mitochondrial bioenergetics following a challenge with human respiratory syncytial virus (RSV). Our results showed that TGFβ1 treatment of macrophages decreased the antiviral and proinflammatory response and suppressed basal, maximal, spare mitochondrial respiration, and mitochondrial ATP production. Challenge with RSV following TGFβ1 treatment further exacerbated mitochondrial dysfunction. The TGFβ1 and TGFβ1+RSV treated macrophages had a higher frequency of apoptosis and diminished phagocytic capacity, potentially through mitochondrial stress. Disruption of TGFβ1 signaling or rescue of mitochondrial respiration may be novel therapeutically targetable pathways to improve macrophage function and prevent secondary bacterial infections that complicate viral respiratory infections.

Introduction

TGFβ1 is a cytokine that can promote a pro-inflammatory or an immunosuppressive state depending on the environment, and TGFβ1 is well characterized as playing a role in acute respiratory distress syndrome and pulmonary fibrosis (1-3). Children with severe asthma (4), cystic fibrosis (5-7), and bronchopulmonary dysplasia (8-13) have elevated levels of TGFβ1 in their lung epithelial lining fluid, and these children are susceptible to exacerbation of their underlying lung disease when infected with common respiratory viruses such as respiratory syncytial virus (RSV), rhinovirus, and influenza. While the immunosuppressive properties of TGFβ1 in response to respiratory viruses have been studied (14-17), the antiviral response of the alveolar macrophage exposed to an elevated TGFβ1 environment has not been characterized (18).

Alveolar macrophages are sentinel innate immune cells that patrol the alveolar spaces in the lung and respond to environmental stimuli leading to changes in this cell’s immune phenotype (19, 20). Traditionally, the alveolar macrophage phenotype has been dichotomized into host-defense/pro-inflammatory/glycolytic (M1) as defined by their response to Th1 cytokines and stimuli, interferon gamma (IFNγ) + lipopolysaccharide (LPS) versus an immune regulatory/anti-inflammatory/oxidative phosphorylation (M2) state as defined by their response to Th2 cytokines, interleukin 4 (IL4) or IL13 (21). Although macrophage phenotype has been dichotomized into M1 vs. M2 states, subsets of M2 macrophages exist and polarization to one phenotype is not absolute (19, 22-24). Unlike M1/M2 macrophages which have established cell surface and gene expression markers, macrophages polarized by TGFβ1 or RSV have poorly characterized gene expression, cell surface, and bioenergetic profiles.

Mitochondria sense viral pathogens through the intracellular RNA sensing retinoic acid inducible like receptor-1 (RLR)/mitochondrial antiviral signaling protein (MAVS) pathway (25) that activates both proinflammatory and Type I interferon (IFN) antiviral responses by increasing mitochondrial reactive oxygen species (mROS). mROS generation is tightly regulated by the flux of electrons through the mitochondrial electron transport chain, thus coupling the antiviral response to mitochondrial oxidative phosphorylation. Clearance of airway pathogens by phagocytosis is an energy demanding process that is a main function of the alveolar macrophage, and boosting mitochondrial capacity to generate ATP through oxidative phosphorylation may rescue phagocytosis in an immunosuppressed macrophage. However, the energetic states of macrophages exposed to elevated TGFβ1, or those exposed to respiratory syncytial virus (RSV) have not been studied.

The primary goal of this study was to determine whether treatment of an alveolar macrophage with TGFβ1 would lead to an altered macrophage phenotype that cannot mount a robust antiviral response to an RSV challenge. Specifically, we tested whether pretreatment of alveolar macrophages with TGFβ1 prior to RSV infection would: 1) decrease the Type I IFN response, 2) deactivate the macrophage and prevent its repolarization to fight infection, 3) diminish ATP production via mitochondrial oxidative phosphorylation, 4) increase apoptosis, and 5) decrease phagocytosis. When further challenged with RSV, the anti-viral response was muted including genes associated with the RLR/MAVS pathway and mitochondrial respiration.

Materials and Methods

MH-S Macrophage Cultures

The mouse alveolar macrophage cell line, MH-S, was purchased from the American Type Culture Collection (ATCC, Manassas, Virginia, USA) and cultured in RPMI-1640 medium with 10% fetal bovine serum, 0.05 mM 2-mercaptoethanol, 11.9 mM sodium bicarbonate, penicillin (100 U/mL), streptomycin (100 μg/mL), and gentamicin (40 μg/mL). MH-S were cultured in a humidified incubator at 37 °C, 10% CO₂ to mimic the gas composition in the lung. MH-S were seeded at 0.5 × 10⁶ cells/well and treated with recombinant mouse IFNγ (5 ng/ml), IL4 (20 ng/ml), IL10 (10 ng/ml), or TGFβ1 (5 ng/ml; Peprotech, Rocky Hill, NJ and BioLegend, San Diego, CA) for 24 hr. The following day, the media was replaced as above, with mouse recombinant TGFβ1 (designated MH-S(TGFβ1) at 5 ng/ml), RSV (46 × 10⁶ pfu/mL; MOI 30:1, designated MH-S(RSV)), or the combination of TGFβ1+RSV (designated MH-S(TGFβ1+RSV), and the cells cultured for an additional 24 hr. Media was then harvested, centrifuged at 1500 × G, 4°C for 10 min and the supernatant was saved for enzyme-linked immunosorbent assays. Cells were washed twice with PBS-EDTA, scraped into Eppendorf tubes using a rubber policeman, and centrifuged at 400 ×G, 4°C for 10 min. The cell pellet was resuspended in 1 mL RNALater (Sigma) for future RNA isolation.

RSV and Plaque Assay

RSV clone rA2|19F and Hep-2 (CCL-23) cells were a gift from Martin L. Moore (Emory University). RSV was propagated in Hep-2 cells in MEM medium supplemented with 10% fetal bovine serum and Amphotericin B (2.5 μg/mL) (Sigma). RSV was harvested 5-6 days after inoculation of Hep-2 cells and cells were sonicated on ice (10-20 sec amplitude, 25 cycles, 1 sec on/1 sec off). Debris was cleared by centrifugation 720 × G for 10 min at 4 °C. RSV titer was determined by a plaque assay performed by serially diluting supernatant and infecting Hep-2 cells in 24-well plates for 1 hr while rocking at room temperature. The inoculum was removed and a 0.75% methylcellulose overlay was added to a confluent monolayer of cells and incubated for 6 days at 37 °C in 5% CO₂ humidified incubator. Plaques were visualized by immunostaining as previously described (26). RSV titers are reported as plaque forming units (pfu)/mL of media.

Human CD14+ Monocyte-derived Macrophage Cultures

Human CD14⁺ peripheral monocytes were purchased frozen and revived according to the supplier’s protocol (Lonza, Alpharetta, GA). Monocytes were seeded at 2 × 10⁶ cells/well into 24-well tissue culture dishes using RMPI-1640 with L-glutamine, 10% heat inactivated fetal bovine serum, penicillin (100 U/mL), streptomycin (100 μg/mL), and gentamicin (40 μg/mL) (Sigma). Human recombinant M-CSF was added to the monocytes according to the CellXVivo Human M2 Macrophage Differentiation Kit (R&D Systems, Minneapolis, MN). Differentiation of human monocytes into macrophages using M-CSF has been described (27). On day 3, half of the media (0.5 mL/well) was replenished with media and human recombinant M-CSF. On day 6, the media was replaced with fresh media as above (without M-CSF) or human recombinant TGFβ1 at 5 ng/mL (BioLegend, San Diego, CA), and the cells were cultured for 24 hr in a humidified incubator at 37 °C, 5% CO₂. The following day, the media was again replaced as above, with human recombinant TGFβ1 at 5 ng/mL, RSV (46 × 10⁶ pfu/mL; MOI 30:1, designated HM(TGFβ1)), or the combination of TGFβ1+RSV (designated HM(TGFβ1+RSV)), and the cells cultured for an additional 24 hr. Media and cells were harvested as described above for MH-S cells.

Gene Expression Analysis

RNA was isolated with the NucleoSpin RNA II kit with on-column genomic DNA digestion according to the manufacturer’s protocol (TaKaRa, Mountain View, CA). RNA was quantified using a NanaDrop Fluorospectrometer (Thermo Scientific). cDNA was synthesized with a High Capacity cDNA Reverse Strand Synthesis kit (Applied Biosystems, Foster City, CA). The cDNA was then pre-amplified by PCR using the TaqMan PreAmp Master Mix and appropriate TaqMan Assays according to the manufacturer’s protocol (Applied Biosystems/Thermo Fisher). Quantitative PCR (qPCR) was performed using TaqMan Gene Expression assays and Master Mix (Applied Biosystems/Thermo Fisher) on a StepOnePlus Real Time PCR System (Thermo Fisher). Mouse housekeeping genes Rpl4, Rsp29, and Oaz1 and human housekeeping genes Rpl30, Rps13, and Rpl27 were used for normalization, respectively (28). TaqMan primer assays are listed in Table S1. Each measurement was performed in duplicate and averaged. The data are reported as the mean ± SEM of six independent experiments.

Flow Cytometry

Cells were pre-incubated with mouse Fc-Block (TruStain fcX, anti-mouse CD16/32 antibody, BioLegend) and Live/Dead Aqua (Thermo Fisher) for 10 min on ice in the dark followed by labeling antibodies, listed in Table S2, for 30 min. Data were acquired on a CytoFLEX flow cytometer (Beckman Coulter, Indianapolis, IN) and analyzed using FlowJo (Tree Star, Ashland, OR). Gates were drawn using fluorescence minus controls. Single MH-S cells were separated from doublet cells by gating on forward scatter area (FSC-A) versus forward scatter height (FSC-H). Single MH-S cells were then selected for being alive by exclusion of the viability dye Live Dead Aqua (Live/Dead Aqua^-). CD45⁺ and F4/80⁺ cells were then gated. This was followed by plotting histograms of the remaining markers: CD86, HLA-DR, CD206, CD71, CD301, Galectin 3, CD204, MARCO, and CD36 (Table S2). For intracellular staining of RSV, cells were fixed and permeabilized using BD Cytofix/Cytoperm kit according to the manufacturer’s protocol (Becton Dickinson), incubated with goat anti-RSVA2 monoclonal antibody (Novus Bio) at a 1:100 dilution in Cytoperm buffer on ice for 2 hours, washed twice using 1 mL of Cytoperm buffer. Cells were then incubated with a 1:200 dilution of the AF488-labeled anti-goat IgG secondary antibody in Cytoperm buffer on ice for 1 hour and washed twice in Cytoperm buffer. Cells were resuspended and stored in 4% paraformaldehyde in PBS (Sigma) until assayed using a Mark^II ImageStream flow cytometer (Amnis). IDEAS software, available for use with the ImageStream cytometer, was used to perform imageflow analysis. The RSV index was calculated as the (percent RSV⁺ cells) × mean fluorescence intensity RSV⁺ cells) and normalized to the average RSV Index from the RSV exposed cells. Four independent experiments were performed, and 15,000 nucleated cells were collected and analyzed for each experiment. The RSV Index was normalized to the RSV-only treated MH-S, and the mean ± SEM is reported.

Cytokine Analysis

Mouse DuoSet ELISA kits for IFNβ, IFNα, IL-6, TNFα and IL12p40 were assayed from media supernatant from 5 × 10⁵ cells per condition in accordance with the supplier’s protocols (R&D Systems) and protein was quantified using a standard curve supplied with the kit. Human DuoSet ELISA kits for IFNβ, TNFα, IL6, IL8, IL10, and IL12p40 were assayed in accordance with the supplier’s protocols (R&D Systems) and protein was quantified using a standard curve supplied with the kit. All measurements were performed in duplicate and averaged. Five independent experiments were performed and the mean ± SEM are reported.

Arginase Activity Assay

Since increased expression of arginase is a marker of a M2 phenotype, arginase activity (1 U = amount of enzyme that catalyzes 1 μmole of L-arginine to urea and ornithine/min at pH 9.5 and 37 °C) was assessed from culture media supernatant from 5 × 10⁵ cells per well according to the supplier’s protocol (Sigma-Aldrich, St. Louis, MO). Activity was quantified relative to a known standard supplied in the kit. All measurements were performed in duplicate and averaged. Six independent experiments were performed and the mean ± SEM are reported.

Metabolic Extracellular Flux Analysis

MH-S were plated at 60,000 cells/well in XF-96-well culture plates (Seahorse Biosciences) and treated as indicated. Extracellular acidification rates (ECAR) and oxygen consumption rates (OCR) were measured in an XFe96 Flux Analyzer (Agilent Seahorse XF Technology). On the morning of the metabolic assays, cell media was aspirated and exchanged for XF Base media (DMEM without bicarbonate (Agilent Seahorse XF Technology) containing 10 μM glucose, 1 mM pyruvate, and 2 mM L-glutamine (Sigma). Media was adjusted to a pH of 7.4 and filter sterilized prior to addition to cells. The cells were then degassed in a humidified 37 °C incubator for at least 1 hour prior to running the assay. The Cell Energy Phenotype assay was performed according the manufacturer’s protocol (Agilent Seahorse XF Technology) using simultaneous injection of oligomycin (2 μM) and FCCP (0.5 μM). The cell energy phenotype test was used to measure the maximal oxygen consumption rate and the maximal extracellular acidification rate was assessed after the simultaneous injection of oligomycin and FCCP. The Mitochondrial Stress Test was performed to calculate oxidative phosphorylation characteristics in response to oligomycin (2 μM), FCCP (0.5 μM), and Rotenone/Antimycin A (0.5 μM) injections according to the manufacturer’s protocol (Agilent Seahorse XF Technology). Each experiment was performed independently 3 times with 10 to 12 replicates per condition. Data from one representative experiment is shown, and mean ± SEM are reported.

Phagocytosis Assay

Following culture in the above mentioned experimental conditions, 100 μL of pHRodo Staph aureus Red beads (3:1 beads:cell ratio; Thermo Fisher) were added to each well of cultured MH-S and human CD14⁺ monocyte-derived macrophages and incubated for two hours in a humidified, 37 °C, 5% (Human) or 10% (MH-S) CO₂ incubator. Media was aspirated, cells were washed twice with 2 ml PBS, and then scraped with a rubber policeman into Eppendorf tubes. After centrifugation at 400 ×G, 4°C for 10 min, the cell pellet was resuspended in 4% paraformaldehyde in PBS and stored at 4 °C. Cells were analyzed using the ImageStream flow cytometer, and the data were analyzed using the Internalization Wizard in IDEAS software. Four independent experiments were performed using MH-S with 15,000 cells analyzed per sample. For the studies of macrophages derived from human CD14⁺ peripheral monocytes, two human donors were used with each experimental condition performed in triplicate and 10,000 – 20,000 cells analyzed per sample. The phagocytic index was calculated as the (percent pHRodo⁺ cells) × mean fluorescence intensity R1 gate internalized pHRodo⁺ cells) and normalized to the average phagocytic index from the untreated cells.

Mitochondrial Reactive Oxygen Species Assay

Following experimental treatment, cells were washed twice with PBS, and 1 mL of PBS containing Ca, Mg and 500 nM MitoTracker Red CM-H₂X ROS (Thermo Fisher) was added to each well of a 12-well plate. Cells were incubated for 30 min at 37 °C and 5% CO₂, then washed with PBS, and harvested using a rubber policeman into a 96-well plate for immediate analysis of 50,000 cells per experiment on a CytoFlex flow cytometer. Six independent experiments were performed. A mitochondrial ROS gate was set using an unstained control. Data were analyzed using Flow Jo software. Mean fluorescence intensity ± SEM are reported.

Apoptosis Assay

Following experimental treatments, cells were washed twice with PBS and harvested by scraping with a rubber policeman into flow cytometry tubes. After centrifugation at 400 ×G at 4 °C for 10 min, cells were resuspended in 1 mL of 4% paraformaldehyde in PBS for 15 min at 37 °C and 5% CO₂. Cells were washed twice in PBS-EDTA and permeabilized by dropwise addition of ice-cold methanol as the sample was vortexed. Cells were incubated in methanol at 4 °C for 30 min and then washed twice with PBS-EDTA. Cells were resuspended in residual buffer and blocked for 10 min at RT with 2.5 μl of mouse FC Block for 10 min. Then, 2 μl of R-phycoerythrin (PE)-cleaved caspase (Thermo Fisher) was added to the cells and allowed to incubate for 1 hour in the dark at room temperature. At the end of the incubation, cells were washed twice with PBS-EDTA and immediately run on a FACSCalibur flow cytometer. Six independent experiments were performed. A cleaved Caspase 3 positive gate was set using an unstained control. Data were analyzed using Flow Jo software. Mean percent in the cleaved-Caspase 3⁺ gate ± SEM are reported. Cells were also analyzed using the ImageStream flow cytometer, and the data were analyzed using the Apoptosis Wizard in IDEAS software.

Statistics

Data were analyzed with GraphPad Prism 7 for Windows using one-way ANOVA with post-hoc Tukey’s test for multiple comparisons or Student t-test for bivariable comparisons. A p-value ≤ 0.05 was considered significant.

Results

TGFβ1 Suppresses the Type I Interferon Antiviral and Proinflammatory Response to RSV in Cultured Mouse Alveolar Macrophages

We examined the effects of TGFβ1 on the antiviral and proinflammatory responses triggered by RSV exposure. Exposure of MH-S to RSV versus control resulted in a significant upregulation of a pro-inflammatory response including increased gene expression of Type I IFN and pro-inflammatory cytokines: IFNβ1, RSV: 2.96 vs. Control: 1.0, p < 0.0001), IL1b (RSV: 2.58 vs. Control: 1.01, p < 0 .0001), IL6 (RSV: 1.77 vs. Control: 1.0, p <0.0001), IL12b (RSV: 1.5 vs. Control: 1.0, p < 0.0001), nitric oxide synthase 2 (NOS2), RSV1.67 vs. Control: 1.01, p < 0.0001), TNFα (RSV: 1.72 vs. Control: 1.0, p = 0.0004, and IL18 (RSV: 1.59 vs. Control: 1.06, p = 0.0052 (Figure 1A). RSV alone upregulated cell surface protein expression of the M1 surface markers, CD86 (MFI RSV: 41645 vs. MFI Control: 30051, p = 0.0098) and HLA-DR (MFI RSV: 5520 vs. MFI Control: 1851, p < 0.0001). TGFβ1 treatment decreased the surface expression of CD86 (MFI TGFβ1: 20306 vs. MFI Control: 30051, p = 0.0384), but not HLA-DR (MFI TGFβ1: 1982 vs. MFI Control: 1851, p = 0.9547) at baseline (Figure 1B). Similar to the gene expression profiles, secretion or release of IFNβ (RSV: 24.25 pg/ml vs. Control: 2.54 pg/ml, p < 0.0001), IL6 (RSV: 11.05 pg/ml vs. Control: Not detected (ND), p < 0.0001), and TNFα (RSV: 10.71 pg/ml vs. Control: 5.47 pg/ml, p < 0.0001) into the culture media was increased in MH-S(RSV) cells (Figure 1C).

Figure 1.

TGFβ1 decreases the Type I interferon (IFN) and proinflammatory response to respiratory syncytial virus (RSV) in cultured mouse alveolar macrophages (MH-S). A) Gene expression profiles of M1 markers from six independent experiments performed in duplicate for each sample and reported as mean ± SEM. B) Cell surface expression of M1 markers CD86 and HLA-DR assessed by flow cytometry from six independent experiments and reported as mean fluorescence intensity ± SEM. C) Protein expression of M1 markers measured by enzyme-linked immunosorbent assay (ELISA) from cell culture media from five independent experiments performed in duplicate for each sample and reported as mean ± SEM. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

In the absence of RSV, there were no significant changes in the majority of pro-inflammatory (M1) markers in MH-S(TGFβ1) cells when compared to control (Figure 1A). However, IFNβ1 (TGFβ1+RSV: 1.96 vs. RSV: 2.96, p = 0.0324) and IL12b (TGFβ1+RSV: 1.29 vs. RSV: 1.5, p = 0.0001) gene expression were significantly decreased in the MH-S(TGFβ1+RSV) group when compared to MH-S(RSV) and slightly increased in NOS2 (TGFβ1+RSV: 186 vs. RSV: 1.67, p = 0.0373), but there were no changes in gene expression of IL1b (TGFβ1+RSV: 2.54 vs. RSV: 2.58, p = 0.9977), IL6 (TGFβ1+RSV: 1.69 vs. RSV: 1.77, p = 0.1218), TNFα (TGFβ1+RSV: 1.35 vs. RSV: 1.72, p = 0.0765), or IL18 (TGFβ1+RSV: 1.06 vs. RSV: 1.24, p = 0.1260), (Figure 1A). Both CD86 (MFI TGFβ1+RSV: 3505 vs. MFI RSV: 11,860, p = 0.0079) and HLA-DR (MFI TGFβ1+RSV: 4360 vs. MFI RSV: 5520, p = 0.0010) surface expression were also significantly decreased in response to TGFβ1 when comparing MH-S (RSV) to MH-S (RSV+TGFβ1) (Figure 1B). Treatment with TGFβ1 alone attenuated the secretion of TNFα (TGFβ1: 3.57 pg/ml vs. Control: 5.47 pg/ml, p = 0.9929), but not IFNβ1 (TGFβ1: 2.94 pg/ml vs. Control: 2.54 pg/ml, p = 0.0171). IL6 was not detected in MH-S (TGFβ1) or MH-S cells (Figure 1C). However, TGFβ1 attenuated RSV-associated increases in the secretion of IFNβ1 (TGFβ1+RSV: 17.15 pg/ml vs. RSV: 24.25 pg/ml, p < 0.0001), IL6 (TGFβ1+RSV: 4.27 pg/ml vs. RSV: 11.05 pg/ml, p = 0.0091), and TNFα (TGFβ1+RSV: 3.08 pg/ml vs. RSV: 10.71 pg/ml, p < 0.0001) (Figure 1C).

To further assess activation of the antiviral response genes, we assessed gene expression of the retinoic acid-inducible gene-I (RIG-1) like receptor (RLR) RNA-sensing pathway and interferon-inducible antiviral response genes. When compared to controls, RSV upregulated gene expression of RIG-1 (RSV: 1.59 vs. Control: 1.06, p = 0.0116), Mx1 (RSV: 1.64 vs. Control: 1.01, p < 0.0001), 2′-5′-oligoadenylate synthetase 2 (Oas2) (RSV: 1.20 vs. Control: 1.01, p = 0.0359), and suppressor of cytokine signaling (SOCS1) (RSV: 1.21 vs. Control: 1.0, p = 0.0048) (Figure 2A). There were no differences in gene expression of MAVS (RSV: 0.99 vs. Control: 1.01, p = 0.9857) or Pias1 (RSV: 1.08 vs. Control: 1.12, p = 0.9845) in MH-S(RSV) compared to controls (Figure 2A). When compared to the control, TGFβ1 decreased MAVS (TGFβ1: 0.80 vs. Control: 1.01, p = 0.0002), but did not change RIG-1 (TGFβ1: 0.88 vs. Control: 1.06, p = 0.6448), Oas2 (FC 0.84, p = 0.0920), Mx1 (TGFβ1: 0.98 vs. Control: 1.01, p = 0.9636), SOCS1 (TGFβ1: 0.90 vs. Control: 1.0, p = 0.2831), or Pias1 (TGFβ1: 0.94 vs. Control: 1.12, p = 0.4072). When cells were treated with TGFβ1+RSV, gene expression of MAVS (TGFβ1+RSV: 0.83 vs. RSV: 0.99, p = 0.0038) and Mx1 (TGFβ1+RSV: 1.45 vs. RSV: 1.64, p = 0.0320) were decreased when compared to RSV alone (Figure 2A). In contrast, MH-S(TGFβ1+RSV) did not change RIG-I (TGFβ1+RSV: 1.40 vs. RSV: 1.59, p = 0.6051), Oas2 (TGFβ1+RSV: 1.09 vs. RSV: 1.20, p = 0.3435), Pias1 (TGFβ1+RSV: 0.98 vs. RSV: 1.08, p = 0.8139), or SOCS1 (TGFβ1+RSV: 1.15 vs. RSV: 1.21, p = 0.6899) compared with MH-S(RSV) (Figure 2A).

Figure 2.

TGFβ1 decreases the antiviral response to respiratory syncytial virus (RSV) in cultured mouse alveolar macrophages (MH-S) without affecting the uptake of RSV virus into MH-S. A) Gene expression profiles of RLR-MAVS antiviral and Type I interferon signaling genes from six independent experiments performed in duplicate for each sample and reported as mean ± SEM. B) RSV uptake as measured by RSV index and reported as mean ± SEM. C) Plaque assay of media supernatant and reported as mean ± SEM. Results were analyzed with a Student’s t-test from six independent experiments performed in duplicate for each sample. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

To ensure that a difference in the uptake of RSV was not responsible for the differential antiviral and proinflammatory response of the MH-S following TGFβ1 treatment, we stained cells for internalized RSV and performed image flow cytometry which demonstrated no difference between groups (normalized mean ± SEM, TGFβ1+RSV: 103.3 ± 7.374 vs. RSV: 100 ± 13.5, p = 0.8368) (Figure 2B). In addition, we performed a plaque assay on both the supernatant and cell lysates from MH-S cells. Figure 2C shows that both MH-S(TGFβ1+RSV) and MH-S(RSV) cleared greater than 99% of the RSV particles and that there were no differences in the number of plaques from the media from cells with or without TGFβ1 treatments (Mean ± SEM, TGFβ1+RSV: 223,889 ± 15,949 vs. RSV: 234,444 ± 24,641, p = 0.7266). Cell lysates grew fewer than two RSV plaques in either condition suggesting that RSV infection was not propagated in the MH-S cells (data not shown).

TGFβ1 Suppresses Anti-Inflammatory Response to RSV in Cultured Mouse Alveolar Macrophages

Next, we studied the capacity of TGFβ1 to mount an anti-inflammatory (M2) response. We found that TGFβ1 decreased gene expression of key M2 genes including the mannose receptor (Mrc1, CD206) (TGFβ1: 0.54 vs. Control: 1.03, p < 0.0001) and Clec10a (CD301) (TGFβ1: 0.73 vs. Control: 1.00, p < 0.0001), IL10 (TGFβ1: 0.80 vs. Control: 1.02, p = 0.0011) and TGFβ1 (TGFβ1: 0.71 vs. Control: 0.98, p < 0.0049) when compared to control cells (Figure 3A). Gene expression of the M(IL10) marker, Bcl3, was not different in response to TGFβ1 compared to control cells (TGFβ1: 0.97 vs. Control: 1.00, p = 0.8891) (Figure 3A). Surprisingly, gene expression of the key M2 marker Arginine 1 (Arg1) was increased in MH-S (TGFβ1) cells compared to control cells (TGFβ1: 1.38 vs. Control: 1.01, p < 0.0001) (Figure 3A). M(RSV) cells had no change in Mrc1 (RSV: 0.87 vs. Control: 1.03, p = 0.2371), Clec10a (RSV: 0.92 vs. Control: 1.00, p = 0.0574), IL10 (RSV: 1.00 vs. Control: 1.02, p = 0.9760), TGFβ1 (RSV: 0.97 vs. Control: 0.98, p = 0.9989), and an increase in Arg1 (RSV: 1.18 vs. Control: 1.01, p = 0.0063) and Bcl3 (RSV: 1.25 vs. Control: 1.00, p < 0.0001) compared to control cells (Figure 3A). Compared to MH-S(RSV) cells, MH-S(TGFβ1+RSV) cells had decreased gene expression of Mrc1 (TGFβ1+RSV: 0.55 vs. RSV: 0.87, p = 0.0044), Clec10a (TGFβ1+RSV: 0.75 vs. RSV: 0.92, p < 0.0001), IL10 (TGFβ1+RSV: 0.86 vs. RSV: 1.00, p = 0.0430), and TGFβ1 (TGFβ1+RSV: 0.75 vs. RSV: 0.97, p = 0.0240). The TGFβ1-induced increase in Arg1 was potentiated in response to RSV (TGFβ1+RSV: 2.13 vs. RSV: 1.18, p < 0.0001). Bcl3 gene expression was increased with RSV (RSV: 1.25 vs. Control: 1.00, p < 0.0001), and MH-S(TGFβ1+RSV) did not change Bcl3 gene expression compared to MH-S(RSV) (TGFβ1+RSV: 1.20 vs. RSV: 1.25, p = 0.6368) (Figure 3A). Based on flow cytometry, cell surface expression of the M2 markers CD206 (MFI TGFβ1: 3938 vs. MFI Control: 6059, p = 0.0021), CD71 (MFI TGFβ1: 29,167 vs. MFI Control: 39,797, p = 0.0002), and CD301 (MFI TGFβ1: 202 vs. MFI Control: 409.8, p = 0.0107) were decreased in response to TGFβ1 when compared to control (Figure 3B). MH-S(RSV) cells had a mixed cell surface marker expression with no change in CD206 (MFI RSV: 5625 vs. MFI Control: 6059, p = 0.8216), decreased CD71 (MFI RSV: 30,791 vs. MFI Control: 39,797, p = 0.0012), and increased CD301 (MFI RSV: 739.2 vs. MFI Control: 409.8, p = 0.0001) compared to controls. There was a consistent decrease in surface expression for CD206 (MFI TGFβ1+RSV: 4236 vs. MFI RSV: 5625, p = 0.0523), CD71 (MFI TGFβ1+RSV: 23,998 vs. MFI RSV: 30,791, p = 0.0140) and CD301 (MFI TGFβ1+RSV: 346.3 vs. MFI RSV: 739.2, p < 0.0001) in MH-S(TGFβ1+RSV) compared to MH-S(RSV) cells. Despite an increase in Arg1 gene expression in RSV (RSV: 1.18 vs. Control: 1.01, p = 0.0063), TGFb1 (TGFβ1: 1.38 vs. Control: 1.01, p < 0.0001), and TGFB1+RSV (TGFβ1+RSV: 2.13 vs. Control: 1.01, p < 0.0001) cells, there was no difference in Arg1 activity amongst the groups (Mean Control: 2.85 vs. RSV: 2.89, p = 0.9998, Control: 2.85 vs. TGFβ1: 2.53, p = 0.8958; Control: 2.85 vs. TGFβ1+RSV: 3.14, p = 0.9197, RSV: 2.89 vs. TGFβ1+ RSV: 3.14 p = 0.9463) (Figure 3C).

Figure 3.

TGFβ1 decreases anti-inflammatory (M2) polarization markers. A) Gene expression profiles of M2 markers from six independent experiments performed in duplicate for each sample and reported as mean ± SEM. B) Cell surface expression of M2 markers CD206, CD71, and CD301 assessed by flow cytometry from six independent experiments and reported as mean fluorescence intensity ± SEM. C) Arginase 1 (Arg1) activity assay from five independent experiments performed in duplicate for each sample and reported as mean ± SEM. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

TGFβ1 and RSV Treatment Blunts Mitochondrial Oxidative Respiration in Cultured Mouse Alveolar Macrophages

Mitochondria are key organelles involved in energy metabolism and antiviral sensing. We next examined the effects on energy metabolism and ATP production when RSV exposure was superimposed on TGFβ1-related mitochondrial stress. When compared to control, RSV decreased the mitochondrial respiration rate, a measure of oxidative phosphorylation, of MH-S as shown in the cell energy phenotype test shown in Figure 4A (left graph). Similar depression was observed in the TGFβ1+RSV relative to control but not to RSV alone. Thus, the stressed OCR metabolic potential was decreased for MH-S(TGFβ1) (Mean TGFβ1: 148.7 vs. Control: 166.6, p = 0.0055), MH-S(RSV) (Mean RSV: 107.6 vs. Control: 166.6, p < 0.0001), and MH-S(TGFβ1+RSV) (Mean TGFβ1+RSV: 111.1 vs. Control: 166.6, p < 0.0001) when compared with control cells (Figure 4A, middle graph). When the maximum extracellular acidification rate (ECAR), a measure of glycolysis, was determined, all experimental groups increased in response to mitochondrial stress treatment (Figure 4A, left graph) but the stressed ECAR were not different (Mean RSV: 152.5 vs. Control: 157.5, p = 0.7987; Mean TGFβ1: 149.9 vs. Control: 157.5, p = 0.5179; Mean TGFβ1+RSV: 160.4 vs, Control: 157.5, p = 0.9569; Mean TGFβ1+RSV: 160.4 vs. RSV: 152.5, p = 0.4535) (Figure 4A, middle graph). Compared to control cells, all treatments resulted in a lower basal OCR/ECAR ratio (4.01, control; 1.76, RSV; 1.99, TGFβ1; 1.56, TGFβ1+RSV) (Figure 4A, right graph).

Figure 4.

Mitochondrial oxidative respiration is compromised in MH-S(TGFβ1) and MH-S(RSV). A) Cell energy phenotype test demonstrating oxygen consumption metabolic potential of MH-S(TGFβ1), MH-S(RSV), and MH-S(TGFβ1+RSV) upon simultaneous addition of oligomycin (Complex V inhibitor) and FCCP (uncoupler). B) Mitochondrial stress test demonstrating C) basal OCR, D) maximal OCR, E) spare respiratory capacity, F) ATP Production, and G) Proton leak. Results shown are from one representative experiment from three independent experiments and reported as mean ± SEM. Within an experiment, each condition was performed with 10-12 replicates. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

To gain further insight into the mechanism of decreased mitochondrial oxygen consumption, we explored the mitochondria response to RSV, TGFβ1, and the combination of TGFβ1+RSV using the Mitochondrial Stress Test (Figure 4B). RSV and TGFβ1 each decreased the basal OCR (Mean RSV: 162 vs. Control: 282.5, p < 0.0001; Mean TGFβ1: 207 vs. Control: 282.5, p < 0.0001) (Figure 4C) and maximal OCR (Mean RSV: 279.1 vs. Control: 538.8, p < 0.0001; Mean TGFβ1: 390 vs. Control: 538.8, p < 0.0001) (Figure 4D), spare respiratory capacity (SRC) (Mean RSV: 117.1 vs. Control: 256.4, p < 0.0001; Mean TGFβ1: 182.8 vs. Control: 256.4, p = 0.0002) (Figure 4E), and ATP production when compared to control cells (Mean RSV: 140.8 vs. Control: 257.1, p < 0.0001; Mean TGFβ1: 184.6 vs. Control: 257.1, p < 0.0001) (Figure 4F). Furthermore, treatment with TGFβ1 further potentiated the RSV-induced decreases in the basal (Mean TGFβ1+RSV: 115 vs. RSV: 162, p = 0.0029) and maximal OCR (Mean TGFβ1+RSV: 207.6 vs. RSV: 279.1, p = 0.0265) and ATP production (Mean TGFβ1+RSV: 92.75 vs. RSV: 140.8, p = 0.0010) but not SRC (Mean TGFβ1+RSV: 92.39 vs. RSV: 117.1, p = 0.3775) (compare Figure 4C – 4F). There was no difference in proton leak amongst the groups (Mean RSV: 21.29 vs. Control: 25.32, p = 0.0847; Mean TGFβ1: 22.64 vs. Control: 25.32, p = 03767; Mean TGFβ1+RSV: 22.5 vs, Control: 25.32, p = 0.3321; Mean TGFβ1+RSV: 22.5 vs. RSV: 21.29, p = 0.8680) (Figure 4G).

TGFβ1 Suppresses the Antiviral and Proinflammatory Response to RSV in Human CD14+ Monocyte-derived Macrophages

To test the relevance of our findings in humans, we repeated these experiments with human CD14+ monocyte-derived macrophages (MoDMs) in culture. Similar to MH-S, there was no effect of TGFβ1 on the M1 and anti-viral response markers in human MoDMs (Figure 5A). In response to RSV, there was increased expression of the Type I IFN (IFNβ1, RSV: 14.76 vs. Control: 1.00, p < 0.001; IFNα1, RSV: 2.42 vs. Control: 1.00, p < 0.0001), anti-viral response genes (MAVS, RSV: 1.67 vs. Control: 1.00, p < 0.0001; Mx1, RSV: 3.54 vs. Control: 1.00, p = 0.0286), and pro-inflammatory (CD86, RSV: 3.96 vs. Control: 1.00, p < 0.0001; CCR7, RSV: 14.68 vs. Control: 1.00, p < 0.0001) (Figure 5A). When HM(TGFβ1+RSV) cells were compared with HM(RSV), IFNβ1 gene expression was decreased (TGFβ1+RSV: 7.77 vs. RSV: 14.76, p = 0.0088) but there were no differences in the other M1 and anti-viral markers (Figure 5A). However, HM(TGFβ1+RSV) did decrease secretion of IL12p40 (TGFβ1+RSV: 3.044 pg/ml vs. RSV: 14.38 pg/ml, p = 0.0028) and IL10 (TGFβ1+RSV: 33.86 pg/ml vs. RSV: 83.22 pg/ml, p = 0.0007) when compared to RSV alone but had no effect on IFNβ1 (TGFβ1+RSV: 92.43 pg/ml vs. RSV: 99.08 pg/ml, p = 0.5086), IL6 (TGFβ1+RSV: 793.1 pg/ml vs. RSV: 909.9 pg/ml, p = 0.9139) or IL8 (TGFβ1+RSV: 3836 pg/ml vs. RSV: 4149 pg/ml, p = 0.9422) secretion (Figure 5B). Surprisingly, TNFα expression was increased in HM(TGFβ1+RSV) compared to HM(RSV) (TGFβ1+RSV: 78.12 pg/ml vs. RSV: 42.34 pg/ml, p = 0.0015) (Figure 5B). The human anti-inflammatory (M2) markers, FCER2 (CD23) (TGFβ1+RSV: 0.79, p = 0.003 and TGFβ1: 0.54, p < 0.0001 vs. Control: 1.01 and vs. RSV: 1.17, p < 0.0001 for TGFβ1+RSV and p = 0.0011 for TGFβ1), Mrc1 (CD206) (TGFβ1+RSV: 0.32, p < 0.0001 and TGFβ1: 0.63, p < 0.0001 vs. Control: 1.02 and vs. RSV: 0.37, p = 0.8477 for TGFβ1+RSV and p = 0.0022 for TGFβ1), and CD200R1 (TGFβ1+RSV: 0.68, p < 0.0001 and TGFβ1: 0.68, p < 0.0001 vs. Control: 1.00 and vs. RSV: 0.92, p < 0.0001 for TGFβ1+RSV and p < 0.0001 for TGFβ1) were decreased in HM(TGFβ1) and HM(TGFβ1+RSV) compared to both control and HM(RSV) cells (Figure 5C), similar to that seen in MH-S. There was a decrease in CCL24 gene expression in HM(RSV) (RSV: 0.80 vs. Control: 1.01, p = 0.0182); however, there was no differential gene expression in the remainder of the groups for CCL24 and amongst any of the groups for CCL22 (data not shown).

Figure 5.

TGFβ1 decreases the Type I interferon (IFN) and proinflammatory response to respiratory syncytial virus (RSV) in CD14+ human monocyte-derived macrophages (HmMoDM). A) Gene expression profiles of M1 markers from six independent experiments performed in duplicate for each sample and reported as mean ± SEM. B) Protein expression of M1 markers measured by enzyme-linked immunosorbent assay (ELISA) from cell culture media from five independent experiments performed in duplicate for each sample and reported as mean ± SEM. C) Gene expression profiles of M2 markers from six independent experiments performed in duplicate for each sample and reported as mean ± SEM. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

TGFβ1 Decreases Phagocytic Capacity of Mouse Alveolar Macrophages and TGFβ1+RSV Human Monocyte-derived Macrophages

Phagocytosis is a main function of a macrophage and is energy intensive. Given the decreased ATP production with MH-S(TGFβ1) and MH-S(TGFβ1+RSV), we hypothesized that TGFβ1 treatment would decrease the phagocytic capacity of the macrophage. Phagocytosis was measured using pHRodo Staph aureus fluorescently-tagged beads that fluoresce once engulfed and acidified in the phagosome. TGFβ1 treated MH-S were 13.5% less efficient at phagocytosis than controls (p = 0.0095) while TGFβ1+RSV treated MH-S were 20% less efficient at phagocytosis than RSV exposed cells (p = 0.0004) (Figure 6A). In human monocyte-derived macrophages, TGFβ1 did not attenuate phagocytosis when compared to control cells (p = 0.9989) (Figure 6B). Unlike the MH-S cells, RSV exposure decreased phagocytic capacity of human monocyte-derived macrophages by 20% when compared to controls (p = 0.0232) (Figure 6B). However, the combined exposure of human monocyte-derived macrophages TGFβ1+RSV (Normalized mean phagocytic index: 76%) resulted in impaired phagocytosis similar to RSV (Normalized mean phagocytic index: 80%) alone (p = 0.9349) (Figure 6B). To further explore the mechanism resulting in decreased phagocytic capacity following TGFβ1+RSV exposure in MH-S cells, we measured the cell surface expression of several scavenger receptors involved in phagocytosis, CD204, CD36, Galectin 3 and MARCO. CD204 cell surface expression was increased with RSV exposure compared to control (MFI RSV: 61,176 vs. MFI Control: 30,681, p < 0.0001) and decreased in response to TGFβ1 in combination with RSV compared to RSV alone (MFI TGFβ1+RSV: 34,316 vs. MFI RSV: 61,176, p < 0.0001). CD36 was increased in response to RSV (MFI RSV: 32,411 vs. MFI Control: 26,142, p = 0.0006) and TGFβ1 (MFI TGFβ1: 31,844 vs. MFI Control: 26,142, p = 0.0016), but there was no difference between TGFβ1+RSV (MFI: 29,915) compared with RSV (MFI: 32,411, p = 0.2546). There was no difference in response between Galectin 3 (MFI TGFβ1+RSV: 23,282 vs. MFI RSV: 36,693, p = 0.1667) and MARCO (MFI TGFβ1+RSV: 169 vs. MFI RSV: 152, p = 0.6440) when comparing TGFβ1+RSV to RSV treated MH-S cells (Figure 7).

Figure 6.

Phagocytosis of fluorescently-labeled pHRodo Staph aureus particles was analyzed using image flow cytometry. A) The gating strategy and representative flow plots and images are shown for MH-S cells for each condition. B) Phagocytosis of MH-S cells are shown. Results are averaged from four independent experiments with 15,000 cells analyzed per condition in each experiment and are reported as mean ± SEM. Representative images of MH-S cells are shown above the respective bars in the graph. C) Phagocytosis capacity of human CD14+ monocyte-derived macrophages (HmMoDM) are shown for each condition. Two human donors were used to differentiate HmMoDM in triplicate from each condition. Results are averaged from six independent experiments with 10,000 cells analyzed per condition in each experiment and are reported as mean ± SEM.

Figure 7.

Cell surface expression of the scavenger receptors A) CD204, B) CD36, C) Galectin 3, and D) MARCO were measured by flow cytometry from six independent experiments and reported as mean fluorescence intensity ± SEM. *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

TGFβ1 Increases Apoptosis of Mouse Alveolar Macrophages

Mitochondrial-dependent apoptosis can be triggered by elevated levels of mROS as a consequence of mitochondrial antiviral signaling that promotes loss of the mitochondrial membrane potential. TGFβ1 alone did not increase mROS (MFI TGFβ1: 46,464 vs. MFI Control: 45,466, p = 0.9979) but MH-S(RSV) (MFI RSV: 78,092 vs. Control, p < 0.0001, vs. TGFβ1, p < 0.0001) and MH-S(TGFβ1+RSV) cells (MFI TGFβ1+RSV: 71,600 vs. Control, p = 0.0007, vs. TGFβ1, p = 0.0011) had increased mROS levels when compared to control or MH-S(TGFβ1) cells (Figure 8A). There were similar levels of mROS between MH-S(RSV) and MH-S(TGFβ1+RSV) cells (MFI TGFβ1+RSV: 71,600 vs. MFI RSV: 78,092, p = 0.6548) (Figure 8A). Both TGFβ1 and mROS are known inducers of apoptosis, and we explored the possibility that the decreased phagocytosis could also be explained by increased MH-S cell apoptosis. Despite similar increases in mROS levels in RSV and TGFβ1+RSV cells, the addition of TGFβ1 (% apoptotic: 21.08%, vs. control, p < 0.0001; vs. RSV, p < 0.0001) or TGFβ1+RSV (% apoptotic: 16.58%, vs. control, p = 0.0025; vs. RSV, p = 0.0438) resulted in an increase in cleaved caspase 3 levels, an apoptosis marker, compared to control (% apoptotic: 10.84%) or MH-S(RSV) cells (% apoptotic: 12.64%), indicating that TGFβ1 increased apoptosis regardless of RSV exposure (Figure 8B). Image flow analysis confirmed apoptotic nuclear morphology (Figure 8C). Therefore, decreased phagocytic capacity in MH-S(TGFβ1) and MH-S(TGFβ1+RSV) cells may be due to TGFβ1-induced macrophage apoptosis.

Figure 8.

Mitochondrial reactive oxygen species (mROS) production and apoptosis were analyzed by flow cytometry. A) MitoTracker Red CMH₂X ROS is a fluorescent probe that measures mROS as assessed by flow cytometry from six independent experiments and reported as mean fluorescence intensity ± SEM. A representative plot of flow cytometry data is shown for one experiment. B) Cleaved caspase 3 (Cl-Cas3) is a marker of apoptosis as assessed by flow cytometry from six independent experiments. A representative plot of flow cytometry data is shown for one experiment and reported as mean fluorescence intensity ± SEM. C) Apoptotic nuclear morphology was confirmed by image flow cytometry analysis. Cells in the apoptosis negative gate did not stain positive for cleaved Caspase 3, and this the corresponding image is not shown in channel 3 (Ch03). *p ≤ 0.05 vs. control, ^Op ≤ 0.05 vs. RSV, ^#p ≤ 0.05 vs. TGFβ1.

Discussion

In this study, we found that TGFβ1 suppressed the antiviral and pro-inflammatory response to RSV, potentially by disrupting mitochondrial bioenergetics. As the sentinel innate immune cell responsible for activation of the innate and adaptive immune responses to pathogens, the alveolar macrophage is a key instigator of host defense through pathogen phagocytosis. However, the immune phenotype of the alveolar macrophage can change depending on the cytokine environment it occupies. As expected, RSV induced a similar response as IFNγ+LPS treatment of macrophage (27) in addition to a strong Type I IFN antiviral response.

Exposure to TGFβ1 has been shown to disable the proinflammatory response of macrophages stimulated with the gram-negative endotoxin LPS (29). Similarly, we showed that TGFβ1 treatment of mouse alveolar macrophages and human monocyte-derived macrophages diminished the Type I IFN antiviral and pro-inflammatory response to RSV. These results are consistent with previous reports that TGFβ1 acts as an immunosuppressant of proinflammatory cytokines in macrophages 12-16 hours following treatment with TGFβ1 and that down-regulation occurs at the level of translation, rather than transcription (29, 30).

We observed that TGFβ1 elicits a unique macrophage phenotype distinct from the M(IL10) phenotype (27). In previous studies, a longer duration of preincubation and maximal suppression of LPS-triggered TNFα release occurred with TGFβ1 exposure compared to IL10 (29). Our results indicate that TGFβ1 suppresses the expression markers associated with alternatively activated macrophages polarized by IL4 (Figure 3B), suggesting that macrophages exposed to TGFβ1 did not adopt a tissue-healing or wound-repair phenotype typically promoted by IL4 or IL10 macrophages (27) but a mixed phenotype.

The dampening of the Type I IFN response to RSV is not a result of TGFβ1 suppressing the ability of MH-S to phagocytose RSV (Figure 2B – C), and a lower RSV burden is not responsible for the diminished antiviral and proinflammatory response in MH-S(TGFβ1+RSV) cells compared with MH-S(RSV) cells. TGFβ1 has also been shown to inhibit IFNγ-induced expression of the MHC Class II gene (HLA-DR) in the monocytic cell line U937 (31). We also showed that TGFβ1 decreased HLA-DR cell surface expression in MH-S cells. In contrast to our results, Tsunawaki and colleagues did not see a decrease in phagocytosis of starch granules by mouse peritoneal macrophages when exposed to TGFβ1 (32); however, this discrepancy may be due to experimental differences as we challenged macrophages with fluorescently-tagged Staph aureus endotoxin-coated beads and peritoneal macrophages may behave differently than cultured mouse alveolar macrophages or human monocyte-derived macrophages.

There are several possible explanations for why phagocytosis was impaired in TGFβ1 exposed macrophages. First, phagocytosis is an energy intensive process, and we demonstrated using Seahorse metabolic bioenergetic studies that TGFβ1 and TGFβ1+RSV treated macrophages have impaired oxidative phosphorylation and ATP production. Mitochondrial integrity is integral for cell function, and cells initiate apoptosis when their mitochondria are compromised. We found that TGFβ1 induced apoptosis in alveolar macrophages and represents a second mechanism by which TGFβ1 may have compromised phagocytic function. This response contrasts with IFNγ+LPS and IL4 polarized macrophages, neither of which stimulate apoptosis despite diminishing oxidative phosphorylation in IFNγ+LPS treated macrophages (27). A third possibility is that the TGFβ1 polarized macrophage alters scavenger receptor expression affecting the macrophage’s ability to recognize and phagocytose bacteria. In our study, only the M(IL4)/M2 marker, CD204, was increased in response to RSV and decreased in response in TGFβ1 correlating to the differences seen in phagocytic capacity between MH-S(RSV) and MH-S(TGFβ1+RSV) cells. Other scavenger receptors such as galectin 3, CD36, and MARCO increased expression in response to RSV but were not affected by TGFβ1 treatment prior to RSV. While phagocytosis of RSV was not inhibited due to TGFβ1 treatment, the combination of TGFβ1+RSV resulted in the inability of the macrophage to respond to a bacterial toxin phagocytic challenge. It is likely that mitochondrial dysfunction and subsequent apoptosis induced by TGFβ1 is responsible for diminished phagocytosis to a bacterial toxin challenge following RSV exposure.

Involvement of mitochondrial metabolism is integral in the plasticity of macrophages to change from an M(IL4) or M(IL10) phenotype to an M(IFNγ+LPS) phenotype (27). Conversely, an M(IFNγ+LPS) polarized macrophage was not readily able to convert to an M2 phenotype when exposed to either IL4 or IL10 (27). Herein we showed that TGFβ1 treated macrophages, with subsequent RSV exposure, had diminished oxidative respiratory capacity resulting in a loss of mitochondrial ATP production compared to either TGFβ1 or RSV treatment alone. Furthermore, the TGFβ1+RSV macrophages behaved metabolically like an IFNγ+LPS stimulated macrophage. Repolarization of the deactivated macrophage, similar to the M(IFNγ+LPS) macrophage, may not be possible with modification of the pulmonary cytokine environment. Rather, recruitment of proinflammatory monocytes, instead of repolarization of the deactivated resident alveolar macrophages, may be necessary for host defense against secondary bacterial infections.

While the present study describes the effects of TGFβ1 on macrophage response to respiratory viral infections, others have examined the effects of TGFβ1 on primary human bronchial epithelial cells and fibroblasts. For example, TGFβ1-treated human bronchial epithelial cells had increased RSV replication and an exaggerated TNFα response (15). This result was paralleled in TGFβ1-treated fibroblasts whereby rhinovirus replication was increased, and TGFβ1 was shown to blunt the Type I IFN response via the IRF3 pathway in bronchial epithelial cells from asthmatic patients (16). Bedke et al showed that the TGFβ-family member, TGFβ2, was increased in the supernatant from cultured human bronchial epithelial cells. This increase in TGFβ2 was responsible for the increased rhinovirus replication from these cells due to inhibition of the Type I IFN response (14). However, others have shown the reverse interaction in epithelial cells where RLR-activated IRF3 prevented association of Smad3 with the TGFβ1 receptor (TGFβR) (33). Without Smad3 association with TGFβR, formation of functional Smad transcriptional complexes is inhibited thus preventing TGFβ1-induced gene transcription of immunosuppressive proteins during an active viral infection. By comparison, our results showed that TGFβ1 prevents IRF3-gene activation by decreased IFNβ gene and protein expression. Inhibition of the TGFβ1 signaling pathway in the alveolar macrophage may be a target to prevent immunosuppression during the antiviral response.

The addition of physiologic (1 to 10 nM) and pharmacologic (100 nM) concentrations of dexamethasone to differentiating human monocytes increases the localization of transforming growth factor β receptor II (TGFβRII) to the cell surface of mature macrophages in a dose- and time-dependent manner (34). Increased surface expression of TGFβRII amplifies TGFβ1 signaling via the TGFβRII/SMAD pathway, and we predict that activation of this signaling pathway with corticosteroids would lead to increased repression of the Type I IFN and pro-inflammatory cytokine response. Physiologic levels of glucocorticoids are present in fetal calf serum that we used in the monocyte to macrophage differentiation media, and we did not see any difference in surface expression of TGFβRII after supplementation with a pharmacologic dose of dexamethasone added to the culture media (data not shown). While the addition of dexamethasone may amplify the inhibition of the Type I IFN response in the presence of TGFβ1+RSV, we did not add pharmacologic doses of glucocorticoids during monocyte differentiation to measure the effects on the TGFβ1/TGFβRII signaling pathway because the clinical use of corticosteroids to treat viral bronchiolitis does not alter the course of illness (35-37).

Overall, the present study demonstrates that TGFβ1 deactivation of macrophages dampens the Type I IFN and proinflammatory response to RSV. Both TGFβ1 and RSV blunt oxidative phosphorylation, while TGFβ1 induces macrophage apoptosis, thereby preventing macrophages from efficient phagocytosis. Future studies investigating the mechanism by which mitochondrial energetics influence reprogramming of deactivated macrophages may lead to improved host defense against secondary bacterial pathogens and opportunistic infections.

Abbreviations

IFNγ

interferon gamma

MAVS

mitochondrial antiviral signaling protein

LPS

lipopolysaccharide

IL

interleukin

TGFβ1

transforming growth factor beta-1

TNF

tumor necrosis factor

TLR3

toll-like receptor 3

RSV

respiratory syncytial virus

RLR

retinoic acid inducible like receptor-1