A Small Molecule BH3-mimetic Suppresses Cigarette Smoke-Induced Mucous Expression in Airway Epithelial Cells

Cigarette smoke (CS) exposure is one of the primary risk factors associated with the chronic mucous hypersecretion (CMH). The antiapoptotic protein, Bcl-2 sustains hyperplastic mucous cells, and the airway epithelium of ex-smokers with CMH as well as mice exposed to chronic CS showed increased Bcl-2 expression. Therefore, we investigated whether Bcl-2 plays a role in CS-induced mucous expression. Primary airway epithelial cells (AECs) of murine and human origin were treated with CS extract (CSE), and there was a concentration- and time-dependent increase in secretory mucin (MUC5AC), mucous regulator (SPDEF) and Bcl-2 expression. Using differentiated human AECs cultured on air-liquid interface, EGFR and ERK1/2 pathways were interrogated. Bcl-2 activity was blocked using a small molecule BH3 mimetic ABT-263 that disrupts the Bcl-2 interaction with pro-apoptotic proteins. The ABT-263 treatment resulted in the downregulation of CSE-induced mucus expression and disrupted the EGFR-signaling while inducing the apoptosis and the pro-apoptotic protein, Bik expression. This strategy significantly suppressed the mainstream CS-induced mucous phenotype in a 3-D human airway epithelium model. Therefore, the present study suggests that CS induces Bcl-2 expression to help promote mucous cell survival; and small molecule BH3 mimetics targeting Bcl-2 could be useful in suppressing the CS-induced mucous response.

secretory mucin that is primarily produced by mucous cells in the airway epithelium is MUC5AC, which is induced upon CS exposure and other airway injuries 8,23,24 . In chronic airway diseases such as COPD and asthma, the debilitating mucus or phlegm production is highly associated with increased numbers of mucous cells with increased mucin synthesis and secretion 8 and this pathology is primarily driven by MUC5AC, as shown by a recent study 25 .
In an animal model of chronic CS exposure, we had observed increased expression of Bcl-2 mRNA in mice exposed to CS for 16 weeks with 4-fold higher number of airway epithelial cells (AECs) showing Bcl-2 immunopositivity in CS-exposed mice compared to air-exposed controls 22 . More importantly, bronchial biopsies from ex-smokers with CMH showed significantly increased Bcl-2 levels with 5-fold increased immunopositivity compared to control subjects 20 . Therefore, we investigated the role of Bcl-2 in CS-induced mucous expression using cultured murine and human airway epithelial cells and tested whether targeting Bcl-2 using a small molecule BH3 mimetic compound, ABT-263, could help in modulating CS-induced mucous expression.

CS induces mucus and Bcl-2 levels in a concentration-and time-dependent manner in murine
AECs. CS induces mucus production and mucous cell hyperplasia in airway epithelium 13,16,26,27 , nonetheless, the molecular mechanisms involved in CS-induced mucous expression remain elusive. We analyzed the effect of CS extract (CSE) on primary murine AECs by treating them with 0, 1, 10 and 100 µg/ml of CSE for 24 h. Cells were analyzed for the expression of a secretory mucin, Muc5ac 8,28 ; a master transcriptional regulator of mucous response, Spdef or SAM pointed domain containing ETS transcription factor 29 ; and Bcl-2, a key anti-apoptotic protein that sustains mucous cells 14,15,20,21 . There was a dose-dependent increase in Muc5ac mRNA levels with significant change following 10 and 100 µg/ml CSE exposure (Fig. 1A). A similar change was observed in Spdef mRNA levels (Fig. 1B), however CSE treatment induced Bcl-2 mRNA levels at all tested concentrations (Fig. 1C). Next, we assessed the expression kinetics of these mRNAs over 0, 3, 24, 48 and 72 h following 10 µg/ml CSE treatment. The Muc5ac mRNA levels were highest at 24 h post CSE treatment (Fig. 1D), and Spdef mRNA levels were increased within 3 h of CSE treatment (Fig. 1E). Bcl-2 mRNA levels peaked at 48 h post CSE exposure (Fig. 1F).
Because there was a modest (~2-fold) increase in the mRNA levels, we analyzed the protein expression in murine AECs at 48 h post 10 µg/ml CSE treatment. There was increased expression of Muc5ac, Spdef and Bcl-2 in a subset of murine AECs following 10 µg/ml CSE treatment ( Fig. 1G and H). CSE treatment resulted in more than 10-fold higher number of Muc5ac-and Spdef-immunopositive cells compared to non-treated controls, along with 5-fold higher Bcl-2-immunopositive murine AECs (Fig. 1H). Therefore, these results suggest that CSE exposure results in a CSE concentration-and time-dependent increase in Muc5ac, Spdef and Bcl-2 levels in murine AECs.

Concomitant induction of MUC5AC and Bcl-2 by CSE in Human AECs.
To characterize the CSE response in human cells, monolayers of human AECs grown in submerged cultured conditions were treated with 0, 1, 10 and 100 µg/ml of CSE for 48 h. There was a dose-dependent increase in MUC5AC ( Fig. 2A) and SPDEF (Fig. 2B) mRNA levels, however, CSE treatment induced Bcl-2 mRNA levels at 1 and 10 µg/ml CSE concentration (Fig. 2C). Bcl-2 protein levels examined by western bot analysis showed >4-fold increase following 24 and 48 h of 10 µg/ml CSE treatment ( Fig. 2D and E). Specifically, following in-situ immunostaining, all of the MUC5AC-positive cells showed higher Bcl-2 expression at 48 h post 10 µg/ml CSE treatment (Fig. 2F). There were 5-fold higher Bcl-2 + and 6-fold higher MUC5AC + human AECs following CSE treatment (Fig. 2G). Thus, like murine AECs, CSE exposure resulted in a concomitant increase in both MUC5AC and Bcl-2 expression in human AECs.

CSE engages EGFR pathways to induce MUC5AC and Bcl-2 expression. To mimic the in-vivo
physiology of AECs, we next analyzed the effect of CSE on air-liquid interface differentiated human AECs. At 48 h post 10 µg/ml CSE treatment, the MUC5AC (Fig. 3A), SPDEF (Fig. 3B) and Bcl-2 ( Fig. 3C) mRNA levels were induced by CSE exposure. At protein levels too, there was increased MUC5AC expression (Fig. 3D) with a 4-fold higher number of MUC5AC + cells at 48 h post 10 µg/ml CSE treatment compared to controls (Fig. 3E). We next investigated the pathways implicated in CS-induced mucus expression by analyzing EGFR (Epithelial Growth Factor Receptor) and ERK1/2 (Extracellular-signal Regulating Kinase 1/2) pathways 23,[30][31][32] . The immunoblot analyses showed increased ERK1/2 phosphorylation at 24 and 48 h post CSE treatment whereas EGFR phosphorylation was observed only at 48 h post CSE treatment ( Fig. 3F and G). EGFR mRNA levels were also an increased at 48 h post CSE treatment (Fig. 3H). These data thus suggest that CSE engages EGFR and ERK1/2 pathways to induce mucus and Bcl-2 expression.
Small molecule BH3 mimetic, ABT-263, suppresses CSE-induced mucous expression by inducing apoptosis. Bcl-2 levels regulate epithelial cell survival in IL-13-and allergen-induced airway inflammation 14 and in the present study, CSE treatment induced both MUC5AC and Bcl-2 expression. Therefore, we reasoned that blocking Bcl-2 might be helpful in restricting the CS-induced mucous response. We had tested various chemical and molecular approaches to inhibit Bcl-2 and found that ABT-263, a small molecule BH3-domain mimetic compound efficiently blocks Bcl-2 activity with fewer off-target effects 14 . Therefore, we tested the effect of ABT-263 on air-liquid interface differentiated human AECs. Pretreatment with ABT-263 (1uM) significantly suppressed the CSE-induced MUC5AC (Fig. 4A), SPDEF (Fig. 4B) and EGFR (Fig. 4C) mRNA as observed in CSE + ABT treatment group with no discernible effect in ABT-only treated cells. ABT-263 pretreatment suppressed the CSE-induced EGFR pathway by 2-fold in CSE + ABT condition compared to CSE only group and there were no changes in ERK1/2 pathway ( Fig. 4D and E).
Because ABT-263 is a known apoptotic inducer, therefore, we examined the extent of apoptosis in our experimental setting by flow cytometric analysis of the Annexin V and propidium iodide stained cells. There were 4-fold higher apoptotic cells in human AECs pretreated with ABT-263 after CSE exposure (CSE + ABT) compared to CSE-or ABT-alone treated controls (Fig. 4F). Our previous studies have shown that Bik is a key proapoptotic protein in AECs that induces cell death 14,16 and therefore, we analyzed the effect of ABT-263 on Bik levels. Compared to CSE treated cells, the Bik mRNA levels were upregulated in CSE + ABT treated cells (Fig. 4G). Similarly, in differentiated murine AECs (Fig. 4H), ABT-263 significantly suppressed the CSE-induced expression of Muc5ac, Egfr and FoxA3, another mucous-regulating transcription factor 33 . In addition, there were higher Bik mRNA levels in CSE + ABT treated murine AECs (Fig. 4H).

ABT-263 completely blocks CS-induced mucous in 3-D human airway tissue.
In order to assess the efficacy of ABT-263 in blocking CS-induced mucous, we analyzed the 3-D EpiAirway tissue culture model (MatTek Corp, Ashland, MA). The 3-D human epithelial airway tissues were exposed to mainstream CS using a SCIREQ smoke machine (Montreal, QC, Canada) for three consecutive days and a group of tissues were treated with 1 µM ABT-263 treatment 2 h before exposures. Smoke exposures caused around 5-fold increase in both MUC5AC (Fig. 5A) and SPDEF (Fig. 5B) mRNA expression which was completely blocked by ABT-263 pretreatment. There was a significant increase in MUC5AC + cell population (40% cells) following smoke exposure ( Fig. 5C and D) that was suppressed in ABT-263 pretreated tissues (15% cells) with no discernible changes in

ABT-263 induces Bik expression and apoptosis in CSE-induced mucous cells. To determine
whether CSE-induced mucous cells were undergoing apoptosis, human AECs pretreated with ABT-263 and exposed to CSE for 48 h were analyzed by TUNEL (Terminal deoxynucleotidyl transferase dUTP nick end labeling) assay and were co-stained for MUC5AC (Fig. 5F). There were >10-fold higher number of TUNEL-positive cells in CSE + ABT treated cells with majority of MUC5AC + cells showing TUNEL positivity as represented by MUC5AC + TUNEL + cell counts (Fig. 5G). Next, we analyzed whether Bik expression is associated with the apoptotic AECs by co-staining the CSE + ABT and CSE-treated cells for Bik and activated or cleaved caspase 3 (CC3), a late-stage apoptosis indicator (Fig. 5H). There were higher number of cells expressing Bik (Bik + ) in CSE + ABT treatment condition with most of Bik + cells were undergoing apoptosis as detected by the CC3 immunostaining and are represented as Bik + CC3 + cell counts (Fig. 5I). These data suggest that ABT-263 treatment of CSE exposed cells suppresses mucous expression by upregulating Bik levels and augmenting apoptosis.

Discussion
In this study, we report that CS exposure induces MUC5AC mucin and SPDEF, the mucous master regulator with a concomitant induction in Bcl-2 levels in both human and murine airway epithelial cells. The treatment with a small molecule BH3 mimetic compound, ABT-263, attenuated the CS-induced mucus expression in both human and murine cells. Furthermore, we observed that ABT-263 treatment attenuated the CS-engaged EGFR signaling to help upregulate apoptosis and the proapoptotic Bik levels in order to suppress the CS-induced mucus expression.
CS exposure has been shown to drive chronic mucus production 6,11,28,[34][35][36] and alters the cell fate by affecting the cell proliferation and the cell death pathways 13,16,26,27 . During the lung development, a network of transcription factors including thyroid transcription factor-1 (TTF1), forkhead box protein A2 (FOXA2) and A3 (FOXA3), and SPDEF, drive the fate of respiratory airway epithelial cells 29,37,38 . This cybernetic transcriptional network regulates the proliferation, differentiation and function of airway epithelial cells including mucous, ciliary and basal cells. Among these transcription factors, SPDEF is the important regulator of the growth and differentiation of mucous cells and mucus production 29 and its levels are upregulated in airway mucous cells of patients with chronic airway diseases 29 and in animal models of allergen exposure 17 . In our studies using CSE exposure model, both murine and human airway epithelial cells showed a dose-dependent response in SPDEF levels, one of the first reports showing the direct effect of CS on this mucous master regulator. It is notable that human airway epithelial cells were more sensitive to CSE exposure than murine counterparts because 1 µg/ml CSE, the lowest dose tested, significantly induced both MUC5AC and SPDEF levels in humans with no discernable changes in murine airway epithelial cells. Moreover, even at the higher CSE dose there was only modest increase in murine Muc5ac and Spdef expression compared to human AECs. This could be due to the inherent differences between human and murine airway epithelial physiology and the mucous regulatory network as discussed in detail elsewhere 8,39 . In our previous studies, we observed that murine AECs inherently show moderate mucous response due to a mucous-limiting p53 genotype that negatively affects Bcl-2 mRNA half-life and SPDEF transcriptional activity 15 . Nonetheless, Bcl-2 seems to be important for mucous expression in both human and murine airway epithelial cells because ABT-263 treatment blocked the mucous expression in cells of both origin. With respect to our studies, the absence of SPDEF in a mice model was shown to attenuate the allergen-induced mucous cell development and mucus production 29,40 . Similarly, knockdown of SPDEF with small interfering RNA (siRNA) in the human bronchial epithelial cell line was found to significantly reduce the expression of IL-13-induced MUC5AC expression 41 . Recently, in order to achieve a long-lasting mucus inhibition Song et al. have successfully used epigenetic silencing of SPDEF to downregulating mucous expression in human lung epithelial cells 42 . These observations together with our data suggest that SPDEF could be a potential therapeutic target for regulating CS-induced mucus hypersecretion and future studies will help determine the feasibility of this proposition.
Cigarette smoke components engage several signaling events that lead to increased differentiation and proliferation of mucous cells leading to chronic mucus production 43,44 and here, we observed that EGFR might be involved in CSE-induced mucin expression. In addition, CS induced inflammatory mediators secreted from airway epithelial cells might be providing a feed-forward to the EGFR-mediated mucin expression in an autocrine or paracrine manner because various EGFR ligands are present in AECs in an inactive form 45 . Furthermore, CS stimulates reactive oxygen species or ROS production that induce TGFα and activation of TNFα-converting enzyme (TACE)-mediated mucin expression 30,46,47 . Similarly, dual oxidase-1 (DUOX-1) is also expressed by airway epithelium that could activate TACE to engage EGFR-mediated mucin expression 48 . More importantly, EGFR pathways has been identified as key proinflammatory mediators that aide in epithelial and mucous cell hyperplasia 23,[49][50][51][52][53] and are associated with the survival of airway epithelial cells in response to a viral challenge 49,54 . We observed an increased phosphorylation of EGFR following CSE exposure that could be implicated in the subsequent induction of both Bcl-2 and MUC5AC. These findings are consistent with other studies, which report TGFα-induced activation of EGFR mediates Bcl-2 induction 52,53 . Moreover, the external stimuli including exposure to CS, LPS, ozone, or allergen that upregulate Bcl-2 expression, also activate EGFR in airway epithelial cells 47,54 . Recently, we reported that when airway epithelial cells were pre-incubated with the specific EGFR and ERK1/2 inhibitors there was a significant suppression of the allergic inflammation induced expression of MUC5AC and Bcl-2 14 . However, in the present study we failed to see any significant changes in ERK1/2 signaling following ABT treatment although CSE exposure caused increased ERK1/2 phosphorylation. There are other ERK-independent pathways engaged by CS and CS-induced ROS to help induce MUC5AC including JNK and AP-1 signal transducers 31 . Further studies are required to delineate the CSE-engaged EGFR signaling mediator(s) that result in mucus expression specifically the one blocked by BH-3 mimetic compound treatment.
Collectively, the present study suggests that Bcl-2 and MUC5AC may be regulated by identical or overlapping pathways implicating EGFR as one of the upstream mediators. Besides epithelial cell proliferation and survival signaling, EGFR mediates MUC5AC synthesis following exposure to CS-, LPS-, or microbial infection 30,55 . Therefore, EGFR activation might be one the primary mechanism responsible for the induction of mucous cell differentiation 52,53 following CS exposure to help sustain hyperplastic mucous cells. However, therapeutic approaches targeting EGFR receptors face considerable safety and efficacy challenges in clinical trials because COPD subjects poorly tolerated EGFR inhibitor, BIBW 2948, and more so, there was no discernible reduction in mucin production 56 . Therefore, downstream mediators of EGFR signaling may offer a more effective target for developing novel mucolytic therapeutics.
One such target could be Bcl-2, which is critical for airway inflammation induced mucous cell hyperplasia 14,20,21 . Bcl-2 is a founding member of a family of proteins that maintain cellular homeostasis by regulating programmed cell death pathways like apoptosis and autophagy 18,57 . Bcl-2 protects cells against a wide range of cell death stimuli by stabilizing the mitochondrial and endoplasmic reticulum (ER) membrane, and preventing permeabilization and release of death mediators 58,59 . Bcl-2 is present in the outer mitochondrial and ER membranes, and inactivates BH3-domain consisting pro-apoptotic members of the Bcl-2 family 58,59 . Here, using a BH3 mimetic compound, ABT-263, which blocks Bcl-2 activity by competing for the BH-3 binding domain with proapoptotic proteins resulting in increased apoptosis; we were able to suppress the CSE-induced mucus expression. Moreover, among the CSE-exposed differentiated airway epithelial cells, there was increased apoptosis in ABT-263-treated ones compared to non-treated controls accompanied by increased Bik levels and reduced EGFR phosphorylation. Bik is the known proapoptotic protein in airway epithelium because Bik-deficient mice fail to resolve allergen-induced mucous expression 60 . In airway epithelium of patients with chronic airway diseases, there is significant reduction in Bik mRNA compared to controls 16 . Therefore, ratio of Bik and Bcl-2 levels could be crucial in determining the epithelial and mucous cell fate because blocking Bcl-2 with ABT-263 results in Bik-mediated cell death.
The present study posits several limitations that merit further discussion. Firstly, the CS extract is primarily consisted of the solution phase extract and do not replicate the actual gaseous and particulate exposure conditions. In addition, there are wide variations in the CS extraction methods. In our studies, we employed both organic and aqueous extraction method to prepare the CS extract and observed that 10 µg/ml concentration was optimal to observe mucous response without altering cell growth. This concentration when extrapolated was equivalent to a 100 µg/mm 3 of actual CS exposure; the levels comparable to those active smokers get exposed to. Secondly, in this study, we have only observed the effect of ABT-263 in cell culture model and to establish any clinical relevance of our findings, the preclinical animal model testing is required. However, like humans where CS-mediated effects take decades, the mouse models also require longer duration of CS exposure to observe any phenotypic changes in airway responses. From our previous time-course in-vivo studies in a mouse model of CS exposure 20 , we had observed that only after 16 weeks of exposure there was an observable mucous response. These longer exposure conditions might need a continuous sustained release of ABT-263 to avoid daily treatment and added irritation to animals. We are currently developing a sustained-release regimen of ABT-263 to help conduct these studies. Nonetheless, our in-vitro data using 3-D human airway tissues and differentiated AECs do provide a strong support for the utility of ABT-263 in blocking CS-induced mucous conditions. With the current success of ABT-263 in the treatment of human cancer and other chronic diseases 61,62 , it will be safer to presume that these BH3 mimetics might find the utility in regulating mucous pathologies associated with chronic airway diseases.

Methods
Cell Culture. Murine AECs were isolated from mouse trachea and cultured on plastic plate or Transwell membranes (Corning, New York, NY) as described previously 63 . All experimental procedures were carried out in accordance with FIU institutional guidelines and regulations. Briefly, tracheas from C57Bl/6 (Jackson Laboratory) mice were excised, connective tissue was cleared, and tracheas were cut open lengthwise. Cleaned tracheas were incubated in Pronase solution (DMEM, 1.4 mg/mL Pronase, and 0.1 mg/mL DNase) overnight at 4 °C to dissociate AECs from BL. Enzymatic activity was stopped with 10% FBS (Invitrogen, Carlsbad, CA), and cells were collected by gently rocking trachea in DMEM/Ham H12 media (Invitrogen), followed by centrifugation at 400 g for 10 minutes at 4 °C. Cell clumps were dissociated by using 5 mL of declumping solution (DMEM and 2 mmol/L ScIeNtIfIc RepoRts | (2018) 8:13796 | DOI:10.1038/s41598-018-32114-w EDTA) and plated at 100,000 cells/well on collagen-coated plates. Primary human AECs were kindly provided by Dr. Scott Randell at the Marsico Lung Institute/Cystic Fibrosis Research Center at the University of North Carolina, Chapel Hill, USA. Lung tissues were procured under protocol #03-1396 approved by the University of North Carolina at Chapel Hill Biomedical Institutional Review Board; informed consents were obtained from all subjects and AECs were procured as previously described 64 . The AECs were maintained in bronchial epithelial growth medium (BEGM, Lonza, Walkersville, MD). For air-liquid interface culture, AECs were seeded on Transwell membranes and differentiated for 14 days. Following treatments, the membrane quarters were used for RNA and protein isolation and were fixed in 4% paraformaldehyde for immunostaining. All the methods were performed in accordance with the institutional guidelines and regulations approved by FIU.

Immunostaining and Fluorescent Imaging Analysis. The murine and human AECs grown on
Labtek-II slides (ThermoFisher Inc.) were fixed in 4% paraformaldehyde and washed in 0.05% v Brij-35 in PBS (pH 7.4) and immunostaining was performed as described previously 14 . Briefly, the cells were blocked using a solution containing 3% BSA, 1% Gelatin and 1% normal donkey serum with 0.1% Triton X-100 and 0.1% Saponin and were stained with antibodies to MUC5AC (Millipore Inc., Burlington, MA), Spdef (Santa Cruz Biotech, Dallas, TX), Bcl-2 (Santa Cruz Biotech, Dallas, TX), Bik (Abcam, Cambridge, MA) and cleaved caspase 3 (Cell Signaling Tech., Danvers, MA) or isotype controls. The immunolabelled cells were detected using respective secondary antibodies conjugated fluorescent dyes (Jackson ImmunoResearch Lab Inc., West Grove, PA) and mounted with 4′,6-diamidino-2-phenylindole (DAPI) containing Fluormount-G TM (SouthernBiotech, Birmingham, AL) for nuclear staining. Immunofluorescent images were captured using BZX700 Microscopy system (Keyence Corp., Japan) and analyzed using NIH Image J software. Quantitative Real-Time RT-PCR. Total RNA was isolated from the snap-frozen cells using RNAeasy kit (Qiagen, Germantown, MD) as per manufacturer's instruction. RNA concentration was determined using the Synergy HTX Multi-Mode reader (BioTek, Winooski, VT) and cDNA were synthesized using iScript advanced cDNA kit (BioRad, Hercules, CA). The primer/probe sets for MUC5AC, SPDEF, FOXA3, and Bcl-2 were obtained either from BioRad (Hercules, CA) or Qiagen (Germantown, MD) and cDNA amplified by q-PCR using the iTaq SYBR-green Master Mix (BioRad, Hercules, CA) in the Agilent Stratagene Mx3000P Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA). Relative quantities were calculated by normalizing averaged C T values to GAPDH or β-Actin to obtain ΔC T , and the fold-change (ΔΔC T ) over the controls were determined as described previously 15 . Western Blot Analysis. Cell extracts were prepared using RIPA buffer (20 mM Tris, pH 7.4, 137 mM NaCl, 1% NP-40, 0.25% Deoxycholate, 0.1% SDS, 1 mM EDTA and 1% protease inhibitor cocktail). Protein concentration was determined by BCA kit (Pierce; Rockford, IL) and 50 µg protein was analyzed by western blotting as described previously 15 . Antibodies to Bcl-2, EGFR, p-EGFR, ERK1/2 and p-ERK1/2 were all from Cell signaling technologies (Danvers, MA) and β-actin antibody was from Sigma co. (St. Louis, MO). Proteins were detected using ECL and visualized by chemiluminescence (Perkin Elmer, Waltham, MA) using the BioRad Chemidoc Imaging system (Hercules, CA). The full-length immunoblots are available online as supplemental information.
ScIeNtIfIc RepoRts | (2018) 8:13796 | DOI:10.1038/s41598-018-32114-w Statistical Analysis. Grouped results were expressed as means ± SEM. Data were analyzed using GraphPad Prism Software (GraphPad Software, Inc., San Diego, CA). Grouped results were analyzed using two-way analysis of variance. When significant main effects were detected (P < 0.05), Fishers least significant difference test was used to determine differences between groups.