Autophagy, TGF-β, and SMAD-2/3 Signaling Regulates Interferon-β Response in Respiratory Syncytial Virus Infected Macrophages

Human respiratory syncytial virus (RSV) is a lung tropic virus causing severe airway diseases including bronchiolitis and pneumonia among infants, children, and immuno-compromised individuals. RSV triggers transforming growth factor-β (TGF-β) production from lung epithelial cells and TGF-β facilitates RSV infection of these cells. However, it is still unknown whether RSV infected myeloid cells like macrophages produce TGF-β and the role of TGF-β if any during RSV infection of these cells. Our study revealed that RSV infected macrophages produce TGF-β and as a consequence these cells activate TGF-β dependent SMAD-2/3 signaling pathway. Further mechanistic studies illustrated a role of autophagy in triggering TGF-β production from RSV infected macrophages. In an effort to elucidate the role of TGF-β and SMAD-2/3 signaling during RSV infection, we surprisingly unfolded the requirement of TGF-β—SMAD2/3 signaling in conferring optimal innate immune antiviral response during RSV infection of macrophages. Type-I interferon (e.g., interferon-β or IFN-β) is a critical host factor regulating innate immune antiviral response during RSV infection. Our study revealed that loss of TGF-β—SMAD2/3 signaling pathway in RSV infected macrophages led to diminished expression and production of IFN-β. Inhibiting autophagy in RSV infected macrophages also resulted in reduced production of IFN-β. Thus, our studies have unfolded the requirement of autophagy—TGF-β—SMAD2/3 signaling network for optimal innate immune antiviral response during RSV infection of macrophages.


INTRODUCTION
Human respiratory syncytial virus (RSV) causes severe lung diseases bronchiolitis and pneumonia among high risk individuals (e.g., infants, children, immuno-compromised individuals; Hall, 2001;Falsey et al., 2005). Innate immune antiviral response mediated by type-I interferon (e.g., interferon-β or IFN-β) is critical for combating virus infection (Uematsu and Akira, 2007;Wilkins and Gale, 2010;Newton and Dixit, 2012). IFN-β plays a pivotal role in host defense against RSV infection and is a major player driving virus clearance from the respiratory tract. RSV infected cells utilize various mechanisms including activation of pattern recognition receptors (PRRs), to trigger IFN-β release during infection (Sabbah et al., 2009;Tsai et al., 2015). In the current study we have unfolded a previously unknown mechanism regulating IFN-β production during RSV infection.
In the current study we have characterized the role of TGF-β and SMAD-2/3 signaling during RSV infection of macrophages. RSV triggered TGF-β production from macrophages, which resulted in activation of SMAD-2/3 signaling pathway.

RSV Triggers TGF-β Production and Activates SMAD-2/3 Signaling in Macrophages
RSV infection results in TGF-β production from non-myeloid cells like lung epithelial cells (McCann and Imani, 2007;Gibbs et al., 2009;Mgbemena et al., 2011;Bakre et al., 2015). However, it is still unknown whether TGF-β is released from RSV infected myeloid cells like macrophages. Therefore, we investigated TGFβ production from RSV infected macrophages. For these studies, we infected primary mouse bone marrow derived macrophages (BMDMs) and mouse macrophage RAW 264.7 cell-line with RSV. At 2, 4, and 8 h post-infection, medium supernatant was collected to analyze active TGF-β levels by ELISA. RSV infection triggered TGF-β release from macrophages, since we detected TGF-β in the medium supernatant of RSV infected BMDMs and RAW 264.7 cells (Figures 1A,B). TGF-β production was independent of cell toxicity or cell death since we failed to detect LDH (LDH cytotoxicity assay), apoptosis and necrosis during the 2-8 h post-RSV infection (data not shown).
RSV induced autophagy in macrophages was deduced from Western blotting with LC3 antibody (Figures 2A,B). In order to assess autophagy's role, we treated cells with autophagy inhibitor 3 MA. Autophagy plays an important role in TGF-β production during RSV infection, since 3 MA treatment abrogated TGFβ release from RSV infected BMDMs ( Figure 2C). A similar result was observed following treatment of RSV infected RAW 264.7 cells with 3 MA (data not shown). Autophagy was also required for activation of SMAD-2/3 signaling during RSV infection. Western blot analysis revealed drastic reduction in phospho-SMAD2 levels following treatment of RSV infected BMDMs with 3 MA (Figure 2D). A similar result was observed in 3 MA treated RAW 264.7 cells infected with RSV (data not shown).
To further validate our results we generated macrophages deficient in beclin-1, a critical cellular factor required for autophagy induction. Beclin-1 silencing by siRNA led to loss of beclin-1 protein expression in BMDMs ( Figure 3A) and RAW 264.7 ( Figure 3B) cells. In accordance with our studies with 3 MA, we observed loss of phospho-SMAD2 protein following RSV infection of beclin-1 deficient BMDMs ( Figure 3C) and RAW 264.7 ( Figure 3D) macrophages. Thus, we have identified autophagy as one of the major cellular events triggering TGF-β production and SMAD-2/3 signaling pathway activation during RSV infection.  Figure 2A) was quantified and the ratio of LC3-II/LC3-I was plotted to denote autophagy induction during RSV infection. (C) BMDMs were pre-treated with autophagy inhibitor (3 MA; 5 mM) for 2 h and infected with RSV (1MOI) in presence of DMSO (vehicle control) or 3 MA. Following infection, TGF-β levels in the medium supernatant was measured by ELISA. (D) BMDMs were pre-treated with 3 MA (5 mM) for 2 h and infected with RSV (1MOI) in presence of DMSO or 3 MA. Cell lysates were subjected to western blot analysis with phospho-SMAD2 (p-SMAD2), SMAD2, and β-actin antibodies. The ELISA value (C) represents the mean ± standard deviation. *p ≤ 0.05 using a Student's t-test.
RSV infection of macrophages is abortive (Segovia et al., 2012). Cellular entry and replication of RSV has been documented in macrophages. However, infectious RSV progeny virus is not released (virus budding) from macrophages. Although macrophages do not support RSV budding, RSV infected macrophages produce high level of IFN-β. It is now becoming apparent that type I interferon (IFN-α/β) released from RSV infected myeloid cells like macrophages are critical for airway host defense against RSV. Interestingly, TGF-β and SMAD-2/3 signaling can positively regulate type-I interferon production and response (Qing et al., 2004). In light of the role of TGF-β and SMAD-2/3 signaling in modulating IFNβ response, we next investigated whether TGF-β-SMAD2/3 pathway regulates IFN-β production during RSV infection. For these studies we utilized the TGF-β inhibitor SB-431542, that blocks extracellular TGF-β activity by inhibiting TGF-β binding to TGF-β receptor. In addition, we utilized SMAD-2 deficient macrophages to study the role of SMAD-2/3 signaling during IFN-β response. TGF-β was required for optimal IFNβ response during RSV infection since significant reduction in IFN-β production was observed following treatment of RSV infected BMDMs ( Figure 4A) and RAW 264.7 ( Figure 4B) cells with SB-431542. SB-431542 treatment also drastically abrogated IFN-β expression in RSV infected macrophages ( Figure 4C). These results were further validated by using SMAD-2 deficient macrophages. We used siRNA to silence SMAD-2 expression in RAW 264.7 cells. Reduced SMAD-2 protein expression was noted in cells transfected with SMAD-2 siRNA ( Figure 4D). RSV infection of SMAD-2 deficient cells resulted in significant loss of IFN-β production ( Figure 4E). These results have highlighted a mechanism of IFN-β production during RSV infection. Our study suggests a role of TGF-β induced SMAD-2/3 signaling pathway in promoting activation of antiviral response by virtue of facilitating production of IFN-β from RSV infected macrophages. These studies have also illustrated a biological role of TGF-β and SMAD-2/3 signaling in positively regulating antiviral response in macrophages during RSV infection.
The role of TGF-β and SMAD-2/3 signaling in positively regulating IFN-β production was validated by investigating RSV infection in macrophages. IFN-β mediated antiviral response restricts RSV replication and therefore, loss of IFN-β production results in enhanced viral replication. Since RSV does not productively infect macrophages, we did not perform plaque assay analysis with medium supernatant of RSV infected macrophages. Instead, we analyzed RSV replication by evaluating expression of RSV nucleocapsid (N) protein mRNA. Indeed diminished IFN-β production from TGF-β signaling inhibited cells led to enhanced RSV replication since elevated expression of RSV N mRNA was observed in cells silenced for SMAD2 expression (Figures 5A,B). Similar enhanced expression of RSV fusion (F) protein was noted following Western blot analysis (with RSV F protein antibody) of RSV infected cell lysate derived from macrophages treated with TGF-β inhibitor SB-431542 (data not shown). A temporal relationship of TGF-β and IFNβ production during RSV infection was also evident. Although we detected TGF-β production from RSV infected macrophages at 2 h post-infection (Figures 1A,B), we failed to detect IFNβ at 2 h post-infection ( Figure 5C). However, IFN-β could be detected at 4 h post-RSV infection ( Figure 5C). This result demonstrated TGF-β production preceding IFN-β production during RSV infection and therefore this temporal event suggested TGF-β mediated regulation of IFN-β production.

Autophagy Regulates IFN-β Production from RSV Infected Macrophages
Autophagy triggered TGF-β-SMAD2/3 signaling, which was involved in IFN-β production during RSV infection of macrophages. These results suggested a role of autophagy in IFN-β production from infected macrophages. Interestingly, autophagy induction was required for optimal IFN-β production from RSV infected dendritic cells (DCs) and mouse airway (Morris et al., 2011;Reed et al., 2013;Owczarczyk et al., 2015). However, autophagy induction has not been studied yet in RSV infected macrophages. RSV infection led to autophagy induction in primary BMDMs (Figures 2A,B) and macrophage cell-line RAW 264.7 cells (data not shown). Role of autophagy was next evaluated by analyzing IFN-β production from macrophages treated with autophagy inhibitor 3 MA. Autophagy induction   Figure 5A) was quantified and the ratio of N/GAPDH was plotted to denote enhanced expression of RSV N in SMAD2 siRNA transfected cells. (C) RAW 264.7 were infected with RSV (1MOI) for 0 h (mock), 2, 4, and 8 h. Following infection, the medium supernatant was collected to assess IFN-β production by ELISA analysis. The ELISA value represents the mean ± standard deviation. *p and **p ≤ 0.05 using a Student's t-test.
in macrophages during RSV infection was required for IFN-β production since 3 MA treatment drastically reduced IFN-β release from RSV infected BMDMs ( Figure 6A) and RAW 264.7 ( Figure 6B) macrophages. Concomitantly, RSV failed to induce IFN-β expression in 3 MA treated macrophages ( Figure 6C). This result was further validated by using autophagy deficient RAW 264.7 cells that are silenced for beclin-1 expression ( Figure 3B). RSV infection of beclin-1 deficient macrophages led to drastic reduction in IFN-β expression ( Figure 6D). Loss of IFN-β production from autophagy inhibited cells led to enhanced RSV replication since elevated RSV F protein expression was observed in cells treated with autophagy inhibitor 3 MA (Figures 6E,F). These results have demonstrated a critical role of autophagy in triggering antiviral response in RSV infected macrophages by promoting IFN-β release.

DISCUSSION
Based on our study, we postulate a scheme (and a model) of IFNβ production from RSV infected macrophages (Figure 7). RSV infected macrophages will induce autophagy, which will trigger TGF-β release. Cell surface TGF-β receptor is then activated by extracellular TGF-β via paracrine/autocrine action. This will culminate in activation (i.e., phosphorylation of SMAD-2/3) of SMAD-2/3 pathway and subsequent expression/production of IFN-β. Thus, our studies have illustrated an important role of autophagy-TGF-β-SMAD2/3 pathway in launching an antiviral response (i.e., IFN-β production) in RSV infected macrophages.
TGF-β is a pleotropic cytokine regulating a plethora of biological functions and activities, including immune surveillance, pro-, and anti-inflammatory response, and cellcycle regulation (Travis and Sheppard, 2014;Heldin and Moustakas, 2016). Respiratory RNA viruses like influenza A virus (IAV), RSV and rhinovirus triggers TGF-β expression  Figure 6E) was quantified and the ratio of F protein/actin was plotted to denote enhanced expression of RSV F protein in 3 MA treated cells. The ELISA values (A,B) represent the mean ± standard deviation. *p ≤ 0.05 using a Student's t-test.
FIGURE 7 | A model depicting the role of autophagy-TGF-β-SMAD2/3 signaling network in positively regulating IFN-β production during RSV infection of macrophages. RSV infection triggers autophagy in macrophages and this event promotes TGF-β production. Cell surface TGF-β receptor (TGFR) is then activated by extracellular TGF-β via paracrine/autocrine action. Interaction of extracellular TGF-β with TGFR will result in activation of SMAD-2/3 signaling comprising of phosphorylation of SMAD-2/3. Activation (i.e., phosphorylation of SMAD-2/3) of SMAD-2/3 will result in IFN-β expression and production. and release from non-myeloid cells like lung epithelial cells and fibroblasts (Thomas et al., 2009;Bedke et al., 2012;Li et al., 2015). Surprisingly, we and others have shown a presumptive detrimental role of TGF-β in host defense against respiratory RNA viruses. TGF-β facilitated RSV (McCann and Imani, 2007;Gibbs et al., 2009;Mgbemena et al., 2011) and rhinovirus (Thomas et al., 2009;Bedke et al., 2012) infection of lung epithelial cells and airway fibroblasts. Accordingly, TGF-β diminished antiviral response by reducing type-I IFN production from rhinovirus infected fibroblasts and epithelial cells (Thomas et al., 2009;Bedke et al., 2012). An interesting aspect of TGF-β is its cell-type dependent response. Particularly, TGF-β activity (and response) may profoundly differ in myeloid vs. nonmyeloid cells. In that context, there is only one study showing a respiratory virus (RSV) up-regulating TGF-β in a myeloid cell (cord blood DCs) and studies with DC-T cell co-culture (RSV infected DC co-cultured with T-cells) revealed a role of TGF-β in modulating in vitro T-cell response (Thornburg et al., 2010). Due to limited studies with myeloid cells, particularly with no studies being performed with macrophages, we investigated whether-(a) RSV triggers TGF-β release from macrophages; and (b) TGF-β produced from RSV infected macrophages plays any functional role in regulating innate immune response. Our studies have demonstrated that-(a) TGF-β is released from RSV infected macrophages; and (b) TGF-β-SMAD2/3 signaling is required for optimal IFN-β production during RSV infection. SMAD-2/3 pathway represents the major TGF-β signaling cascade responsible for transmitting intracellular response originating on the cell surface following interaction of TGFβ with type-II TGF-β receptor (Heldin and Moustakas, 2016). So far no studies have focused on the SMAD-2/3 pathway during respiratory virus infection. It is unknown whether-(a) respiratory viruses like RSV activates SMAD-2/3 pathway; and (b) SMAD-2/3 pathway play any role in regulating virus infection and innate immune response. Our study revealed -(a) activation of SMAD-2/3 pathway in RSV infected macrophages; and (b) a role of SMAD-2/3 pathway in triggering IFN-β production during RSV infection and thus, "positively" regulating innate antiviral response.
Autophagy constitutes a critical cellular process that maintains cellular homeostasis for normal biological and physiological functions (Shibutani et al., 2015). RSV induces autophagy in dendritic cells (DCs) and in the respiratory tract of infected mice (Morris et al., 2011;Reed et al., 2013;Owczarczyk et al., 2015). Autophagy is required for efficient IFN-β response during RSV infection, since lack of autophagy reduces IFN-β release from RSV infected DCs. Although RSV induced autophagy in DCs, autophagy induction in RSV infected macrophages has not been reported yet. We now show autophagy induction in RSV infected macrophages. Furthermore, our studies have illustrated a functional role of autophagy in positively regulating antiviral response (IFN-β production) in macrophages infected with RSV. Given the fact that TLR3 induces autophagy in macrophages (Delgado et al., 2008) and TLR3 is activated by RSV in these cells , we postulate that autophagy induction in RSV infected macrophages occur due to RSV mediated activation of TLR3 pathway. A relationship between autophagy and TLR3 is evident from strikingly similar lung cytokine profile and pathogenesis severity in RSV infected autophagy deficient (i.e., beclin-1 deficient mice) and TLR3 knockout mice (Reed et al., 2013). In addition to TLR3, it is plausible that ssRNA genome of RSV can induce autophagy following TLR7 activation. TLR7 activation results in autophagy induction (Sanjuan et al., 2007;Delgado et al., 2008) and thus, during early infection ssRNA originating from RSV can induce autophagy via TLR7.
In summary, our studies have resulted in identification of autophagy-TGF-β-SMAD2/3 signaling network in macrophages as a critical cellular event required for optimal antiviral response (IFN-β production) following RSV infection.
Animal studies were performed according to housing and care of laboratory animals guidelines established by National Institutes for Health (NIH). All animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Washington State University.

Viral Infection of Cells
BMDMs and RAW264.7 cells were infected with purified RSV at one multiplicity of infection (MOI) in serum free antibiotic free OPTI-MEM medium (Gibco). Virus adsorption was performed for 1.5 h at 37 • C. Following adsorption, cells were washed twice with DPBS (Dulbecco's phosphate-buffered saline) (Gibco) and the infection was continued in presence or absence of serum containing medium. To examine TGF-β secretion and SMAD-2 phosphorylation during RSV infection, BMDMs, and RAW264.7 cells were cultured in serum free medium.
In some experiments, cells were pre-treated with 3methyladenine (3 MA; Sigma Aldrich) or TGF-β inhibitor (SB-431542; InVivogen). After pre-treatment, cells were infected with RSV and infection was continued in presence or absence of respective inhibitors. Dimethyl sulfoxide (DMSO) served as a vehicle control for SB-431542 and 3 MA.

siRNA Transfection
Mouse beclin-1 and mouse SMAD-2 were silenced in BMDMs or RAW 264.7 cells by using Lipofectamine 2000 for 24-48 h. Control siRNA, mouse beclin-1 siRNA and mouse SMAD-2 siRNA were purchased from Santa Cruz Biotechnology. Cells were transfected with 60 pmol of siRNAs.

ELISA Assay
Medium supernatants collected from RAW 264.7 cells and BMDMs were analyzed for active TGF-β (ebioscience) and IFN-β (PBL Assay Science) levels by cytokine-specific ELISA kit.

Western Blotting
BMDMs and RAW264.7 cells were lysed using 1%-Triton X-100 in PBS (pH 7.4), EDTA-free protease inhibitor cocktail (Roche Diagnostics) and 10 mM of sodium pyrophosphate (Sigma) in PBS. Cell lysates were subjected to 10-15% SDS-PAGE. Separated proteins were transferred onto 0.2 µm nitrocellulose membrane (GE Health care) and blotted with specific antibodies. SMAD-2, phospho-SMAD2 and LC-3 antibodies were purchased from Cell signaling. Beclin-1 antibody was purchased from Abcam. RSV fusion (F) protein antibody was obtained from FDA (Dr. Judy Beeler). β-actin antibody was purchased from Bethyl Laboratories. For some experiments, protein bands from Western blot were quantified using Image Lab Software (Bio-Rad).

AUTHOR CONTRIBUTIONS
SP, NS, and SB designed the experiments; SP and NS performed the experiments; SP, NS, and SB analyzed the data; SP and SB wrote the paper.