Male and female βENaC were purchased from The Jackson Laboratory (congenic C57BL/6J background, Stock No. 006438, Bar Harbor, ME). Mice were housed in a specific pathogen-free facility on a 12-hour light/dark cycle with free access to water and a standard chow diet throughout treatment. All animal procedures described were approved by the Institutional Animal Care and Use Committee (IACUC) at Georgia State University. To have enough animals for this study, βENaC mice were bred with wild-type C57BL/6J mice to produce an appropriate number of hemizygous βEnaC for our studies. Before weaning age, animals were genotyped to determine the presence of Scnn1b overexpression. Briefly, tail snips were collected and lysed in Proteinase K overnight at 56°C in preparation for DNA extraction. DNA was extracted by Dneasy Blood and Tissue kit (Qiagen, Foster City, CA) in preparation for polymerase chain reaction (PCR). PCR was performed using primers specific for Scnn1b gene expression as suggested (Forward: CCTCCAAGAGTTCAACTACCG; Reverse: TCTACCAGCTCAGCCACAGTG) (16). To verify the expression of the Scnn1b gene, samples were run on a 4% agarose gel containing ethidium bromide. Mice identified as possessing Scnn1b overexpression were tagged as βENaC mice for future studies.

At 12 weeks of age, mice (n=10 per group) were randomized into either the ENDS exposure group or the control group (sham air). Mice in the ENDS group were exposed to e-cigarette vapor for 1 hour twice daily at a frequency of 5 times per week for a total duration of 10 days as shown in Figure 1A . Mice were placed in a smoking chamber attached to an inExpose Smoking Robot (SCIREQ, Montreal, QC, Canada) with an attached e-cigarette accessory (ECX JoyeTech E-Vic Mini, SCIREQ, Canada) for the administration of vaporized S brand e-cigarette liquid containing nicotine salt at 50 mg/ml or sham air (control). During vapor administration, an optimal vapor/air ratio of 1:6 was obtained based on our previous publication (19). After 10 days of ENDS exposure, animals were euthanized within 24 hours after the last e-cigarette vapor exposure using CO₂. Bronchoalveolar lavage (BAL) fluid and lung tissues were collected for analysis as described below.

After the mice were euthanized, a tracheostomy was performed, and the lungs were immediately flushed with 1 mL ice-cold phosphate-buffered saline (PBS) twice for collection of BAL fluid. Blood was collected via the abdominal aorta, allowed to sit for 30 min to promote coagulation, and centrifuged at 5,000 rpm for 10 min for isolation of serum. Total cells from the BAL were centrifuged and counted using Countess™ II Automated Cell Counter (Invitrogen, Carlsbad, CA) as previously described (16). Differential immune cell count (5,000 cells/slide) was performed on Shandon cytospin slides (Thermo Shandon, Pittsburgh, PA) stained with Diff-Quik (Dade Bering, Newark, DE). The BAL supernatant and serum samples were stored at -80°C until analysis. Expression of inflammatory cytokines, proteases, and growth factors was assessed in BAL fluid via Mouse XL Cytokine Array (R&D Systems, Minneapolis, MN) following the manufacturer’s instructions to determine the presence of a localized inflammatory response within the lung.

In order to determine the effects of ENDS exposure on the development of emphysema, we did a morphometric assessment according to the previous protocol (17, 20). Briefly, the mouse lungs were inflated with 1% low melting-point agarose to 25 cm of fixative pressure. After 48 hours of fixation, lungs were paraffin-embedded, sectioned, and stained with hematoxylin and eosin (H&E). Mean linear intercepts were determined and calculated (21).

The mouse lungs were inflated with 1% low melting-point agarose to 25 cm of fixative pressure and then fixed with 4% neutral buffered formalin for 48 hours. Later, these tissue samples were embedded in paraffin and sectioned into 4 µm sections using a rotary microtome. For immunohistochemical analysis, tissue sections were dewaxed with xylene and rehydrated with graded concentrations of ethanol, followed by antigen retrieval in 10 mM citric acid solution (pH = 6) for 20 min using a pressure cooker. Endogenous peroxidase activity was blocked by 3% hydrogen peroxide. After blocking with 5% Bovine serum albumin (BSA) for 30 min, the slides were incubated with the primary antibodies against the following proteins at 4°C overnight: α smooth muscle actin (α-SMA) 1:500 (Ab5694; Abcam); Von Willebrand Factor (vWF) 1:300 (A008229; Agilent); CD71 (Ab84036, Abcam), proSP-C (AB3786, Sigma), HOPX (11419-AP, Proteintech). After washing, the sections were incubated with the appropriate HRP polymers and developed with 3–3′ diaminobenzidine solution (DAB substrate kit; Vector Laboratories). After counterstaining with hematoxylin, the slides were dehydrated and mounted with a mounting medium. For immunofluorescent staining, after incubated with primary antibodies, the sections were incubated with fluorescent secondary antibody Goat anti-Mouse IgG 488 (A11001, Thermo Scientific), Goat anti-Mouse IgG 594 (A11005, Thermo Scientific), Goat anti-Rabbit IgG 488 (A11008, Thermo Scientific), Goat anti-Rabbit IgG 594 (A11012, Thermo Scientific) for 1h at room temperature. Then the sections were mounted with ProLong Diamond Antifade Mounting with DAPI (P36962, Thermo Scientific).

For Hematoxylin and Eosin (H&E) staining, the paraffin slides were baked at 60°C for 2 hours, then dewaxed with xylene and rehydrated with graded ethanol solutions. Next, the slides were incubated with Mayer’s Hematoxylin for 10 min and washed under tap water for 10 min. Slides were then immersed in Eosin for 30 sec and washed under tap water, after which, the slides were dehydrated and mounted with a mounting medium.

Periodic-acid Schiff (PAS; Sigma Aldrich, Saint Louis, MO) staining was used to evaluate mucus accumulation and immune cell number. Masson’s Trichrome (Sigma, St. Louis, MO) and Picro-Sirius Red (Abcam, Cambridge, MA) staining were used to assess collagen deposition in lung tissue. All stainings were performed following the manufacturer’s instructions. All images were taken at 20x magnification using Keyence Fluorescent Microscope (Keyence, Itasca, IL).

Human airway epithelial cells (16HBE) were a gift from Dr. Pierre Massion (Vanderbilt University) and maintained in DMEM media. 16HBE cells were grown in 1% penicillin/streptomycin with 10% fetal bovine serum. Human Aortic Endothelial Cells (HAEC) were obtained from Invitrogen (Invitrogen, Carlsbad, CA) and cultured in endothelial cell growth medium supplemented with low serum (2% V/V FBS) (PromoCell; Heidelberg, Germany), and both cells were maintained in a cell culture incubator at 37°C and 5%p CO₂. 16HBE cells were plated at 30,000 cells/well in 6-well plates. Then cells were treated with ENDS liquid (50 mg/ml) or ENDS liquid without nicotine for 48 hours and harvested for analysis of apoptosis by flow cytometry or western blot. Similarly, HAEC cells were cultured in 6-well plates at a density of 2.0 × 10⁵ cells/well, starved overnight, and treated with 50 mg/ml ENDS liquid with nicotine or ENDS liquid without nicotine for 24 hours or 1 mM H₂O₂ for 12 hours (positive control) and then subjected to Annexin V/PI apoptosis analysis by flow cytometry.

The cells were detached with accutase solution and washed with PBS and later binding buffer, followed by resuspension in binding buffer at 10⁶ cells/ml. The cell suspensions were incubated with 5 μL of FITC-conjugated Annexin V (V13242, Thermo) for 15 min at room temperature without exposure to light, followed by washing and resuspension in binding buffer. Propidium Iodide (Sigma Aldrich, Saint Louis, MO) was added followed by incubation for 5 min. The cells were acquired using LSRFortessa flow cytometer (BD Biosciences, Durham, NC), and the data was analyzed by FlowJo software (BD, Durham, NC).

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay was performed using VasoTACS In Situ Apoptosis Detection Kit (4826–30-K, Trevigen, Gaithersburg, MD) following the manufacturer’s instructions. In brief, after rehydration, tissue slides were incubated with Proteinase K Solution for 20 min at room temperature, followed by quenching for 5 min. 50 μL of Labeling Reaction Mix were added to the tissues and the slides were incubated at 37°C for 60 min, after which the reaction was terminated by adding the stop solution. The slides were covered with Strep-HRP Solution for another 10 min and developed using TACS Blue Label solution. After counterstaining by Red Counterstain C, slides were dehydrated and mounted with the mounting medium.

Cells were lysed with lysis buffer (RIPA buffer from Thermo Scientific, 89900, 1X Halt protease inhibitor cocktail from Thermo Scientific, 78430, and 1X Halt phosphatase inhibitor cocktail from Thermo Scientific, 78426). Mouse tissues were mixed with lysis buffer (300ul of lysis buffer per 5mg piece of tissue) and homogenated with a homogenizer. Cell debris was pelleted for 10 min at 12,000 x g at 4C. After protein concentration quantification by BCA assay (Thermo Scientific 23225), an equal amount protein of each sample was separated using SDS-PAGE and transferred to a nitrocellulose membrane. Membranes were treated with 5% BSA in PBS-T (PBS with 0.1% Tween 20) blocking buffer for at least 1 h at room temperature, then incubated with primary antibody diluted in blocking buffer overnight at 4C. Anti-ACSL4 antibody (Abcam, ab155282, 1:5000 dilution); anti-GAPDH antibody (CST #2118, 1:1000 dilution); Anti-CD71 (Abcam, Ab84036, 1:1000 dilution), anti-Ferritin antibody (Abcam, Ab75973, 1:1000 dilution). Following three 10 min washes in PBS-T, the membrane was incubated with HRP conjugated secondary antibodies, goat anti-rabbit (Thermo Scientific, 65–6120, 1:5000 dilution), or goat anti-mouse IgG antibody (Thermo Scientific, 62–6520, 1:5000 dilution) in blocking buffer for 1 hour at room temperature. Following three 5 min washes in PBS-T, membranes were incubated with Chemiluminescent Substrate (Thermo Scientific, 34580), and bands were visualized using Autoradiography Film (Denville Scientific, E3012).

Since the peroxidation of polyunsaturated fatty acids (PUFAA) has emerged as a key driver of oxidative damage to cellular membranes leading to ferroptosis (22), we explored the effects of ENDS exposure on peroxidation of PUFA in vitro. 16HBE cells were treated with 50ng/ml of ENDS liquid for 24hs or 1uM of RSL3 (positive control) for 6 hours. After the treatments, cells were washed three times for 10 minutes in PBS, and then fixed in 4% paraformaldehyde for 5 minutes. Next, cells were incubated in 0.1mg/ml BODIPY 581/591 (Invitrogen C10445) in DMSO for 30 minutes in the dark, followed by three times of washes. Cells were then mounted with Prolong gold antifade reagent with 4’,6-diamidino-2-phenylindole (DAPI). Images were taken by Keyence Fluorescent Microscope (Keyence, Itasca, IL).

Lipidomic analysis on serum samples was performed at The Harvard Center for Mass Spectrometry according to our previous publication (23). Briefly, 200 µl of serum samples from mice with either ENDS exposure or normal control was mixed with lung tissue was homogenized with 2 mL of pre-cold methanol (MX0486–1, Sigma), then mixed with 4 mL of chloroform (A452–1, Sigma). The mixtures were combined with 1.2 mL of HPLC grade water (WX0001–1, Sigma), and centrifuged at 3000 rpm for 10 min at 4°C. The bottom layer (chloroform phase) was collected for lipidomics analysis using LC-MS on an Orbitrap Exactive (Thermo Scientific) in line with an Ultimate 3000 LC (Thermo Scientific). Each sample was analyzed in positive and negative modes, in the top 5 automatic data-dependent MS/MS modes. Flow rate was set to 100 μL/min for 5 min with 0% mobile phase B, then, switched to 400 μL/min for 50 min, with a linear gradient of mobile phase B from 20% to 100%. The column was then washed at 500 μL/min for 8 min at 100% mobile phase B followed by re-equilibrated at 0% mobile phase B, 500 μL/min for 7 min. For positive mode runs, buffers for mobile phase A contain 5 mM ammonium formate, 0.1% formic acid, and 5% methanol in water, and, while buffers for mobile phase B are made up of 5 mM ammonium formate, 0.1% formic acid, 5% water, and 35% methanol in Isopropanol. For negative runs, buffers consisted of 0.03% ammonium hydroxide and 5% methanol in water for mobile phase A, and, for mobile phase B, 0.03% ammonium hydroxide, 5% water, and 35% methanol in isopropanol. Each of the spectra for each lipid identified with Lipidsearch software (version 4.1.16, Mitsui Knowledge Industry, University of Tokyo) was manually examined for the presence of the head group characteristic fragment, and for the side chain’s fragment if present. Intensity signals are represented as the area of the parent ion. Integrations and peak quality were curated manually before exporting and analyzing the data in Microsoft Excel. Each ID was based on the fragments found in the literature and compiled in the Lipidsearch database.

In the case of only two groups, two-tailed t-tests were performed to assess the differences. Prior to analysis, we confirmed that the data met the assumptions of normality and homogeneity of variances using appropriate tests (Shapiro-Wilk test for normality and Levene’s test for homogeneity of variances). All analyses were carried out using Prism7.0 software (GraphPad Software Inc., San Diego, CA). Values are represented as mean ± SEM. Levels of significance are designated as p < 0.05 unless otherwise indicated.

The β ENaC mice have been used as a mouse model for induction of COPD (24–26). The βENaC mice, after exposure to cigarette smoke, display key pulmonary abnormalities of COPD, including inflammation, emphysema, and destruction of alveolar tissue, thus are used to study the molecular pathogenesis of COPD (25). One of the features of emphysema is the enlargement of alveolar spaces associated with the destruction of alveolar walls (27). To determine whether ENDS exposure can induce COPD, the β ENaC mice were given whole-body ENDS exposure as indicated ( Figure 1A ). At the end of the treatment, the mice were sacrificed, and lungs were collected for further pathological analysis. We observed that ENDS-exposed mice displayed impaired lung structure integrity with alveolar enlargement, compared with control mice ( Figure 1B ). We further quantified the air space size and found that the free distance between gas exchange surfaces was significantly increased in the lungs of ENDS-treated mice, indicating that ENDS caused lung emphysema in βENaC mice ( Figures 1B, C ). As mucus accumulation is another feature of COPD, we then assessed mucus content using PAS staining. The results showed that ENDS exposure increased the deposition of mucus in airways compared with the control group ( Figure 1D ). Altogether, our results suggest ENDS exposure causes the pathogenesis of COPD in β ENaC mice.

COPD is characterized by lung inflammation, which is a critical factor leading to progressive and irreversible airflow obstruction (28). Therefore, we next determined the presence of immune cells in BAL from mice with or without ENDS exposure using the Diff-Quik staining. Clearly, ENDS exposure increased the total number of immune cells including macrophages and neutrophils in BAL ( Figures 2A, B ). We then performed a cytokine array assay to measure cytokine production. Elevated levels of multiple cytokines including CCL2, IL-4, IL-13, IL-10, M-CSF, and TNF- α were found in the lungs of ENDS-exposed mice ( Figures 2C–H ). Interestingly, individual cytokines such as CCL-2, IL-4, IL-10, and IL-13 have been indicated to contribute to macrophage infiltration, M2 macrophage polarization, and pulmonary fibrosis (29–31). Thus, these results indicate ENDS exposure leads to inflammation response in the lung by enhancing immune cell number as well as cytokines productions in the βENaC mice.

Lung fibrosis with excessive deposition of collagen in airways was identified in patients with COPD (32, 33). It has been shown that lung fibrosis promotes airway obstruction thus exacerbating COPD (34). Therefore, we examined whether ENDS affected lung fibrosis during COPD progression. We first assessed the deposition of collagen by Picro-Sirius red staining ( Figure 3A ) and Masson’s Trichrome staining ( Figure 3C ) using lung tissues from mice with or without exposure to ENDS. The results showed that exposure to ENDS elevated total collagen deposition in βENaC mice within the airway and lung parenchyma compared with the control ( Figures 3A–D ). Myofibroblasts, the major player in collagen synthesis, is a well-known hallmark of fibrosis. We therefore examined the existence of myofibroblasts by staining α-SMA, a widely used myofibroblast maker. The results demonstrated that ENDS exposure increased the α-SMA⁺ myofibroblast population ( Figures 3E, F ). Thus, these data indicate that exposure to ENDS enhances collagen deposition and the number of collagen-producing myofibroblasts, revealing that ENDS exposure promotes fibrosis in the lung.

Next, we try to elucidate how ENDS could promote COPD in the lung. Cell death is reported to be a major contributor to the pathogenesis of COPD and there is increased evidence showing enhanced levels of cell death of alveolar epithelial and endothelial cells in the lungs of COPD patients (35, 36). Therefore, we examined total cell death levels in lung tissues using TUNEL staining. The results demonstrated an elevated TUNEL staining of endothelial cells and airway epithelial cells in ENDS-exposed mice compared with control mice, respectively ( Figures 4A, B , 5A, B ). We observed that the number of endothelial cells, stained by von Willebrand Factor (vWF), a widely used biomarker for vascular endothelial cells, was significantly diminished in ENDS-exposed mice compared with the control mice ( Figures 4C, D ). To confirm this data in vitro, we employed the human aortic endothelial cells (HAEC) and tested the effects of ENDS on endothelial cells in vitro. As shown in Figure 4E , the cell viability was significantly decreased in the nicotine-containing ENDS liquids treated group (57%, Figure 4E bottom right) compared with the control group (82%, Figure 4E top left). Interestingly, ENDS liquid without nicotine also induced remarkable cell death (75%, Figure 4E ). On the other hand, ENDS liquid with nicotine and without nicotine significantly led to cell death compared with the control group in airway epithelial cells (16HBE) ( Figures 5C, D ).

We noticed that the level of cell death detected by TUNEL staining (either apoptosis or necrosis) does not fully match with lung tissue destruction by ENDS as shown in histological staining. We, therefore, questioned whether there are other formats of cell death mechanisms contributing to tissue mass loss during the progression of ENDS-induced COPD. Besides apoptosis, ferroptosis has drawn intensive attention recently as a critical player in COPD pathogenesis and growing evidence has shown involvement of epithelial cell ferroptosis in COPD pathogenesis (37, 38). Ferroptosis is also reported to induce the release of pro-inflammatory cytokines, leading to COPD-related airway remodeling and emphysema (37), which matches our observation of the elevated level of pro-inflammatory phenotype in ENDS-exposed mice. So we speculated ENDS exposure might induce ferroptosis. We performed a mass-spectrometry-based lipidomics analysis using serum samples collected from mice with or without ENDS exposure. The result demonstrated that relative abundances of more than 50 lipid species were altered after ENDS exposure ( Figure 6A ). We identified increases in the level of 9 species of TAGs and 9 major phospholipid classes (with the total number of detected molecular species of 34 distributed between the following major classes: 22 species; phosphatidic acid (PA); lysophosphatidylcholine (LPA) 7 species; phosphatidylethanolamine (PE), 2 species; phosphatidylinositol (PI) 1 species; phosphatidylserine (PS); 1 species, phosphatidylcholine (PC); 1 species) ( Figure 6A ). Furthermore, we measured the expression of biomarkers of ferroptosis using mouse tissues and found ENDS increased expression of ferroptosis-specific markers including Transferrin Receptor/CD71 (39) and ACSL4 (40) and reduced the expression of GPX4 ( Figure 6B ), the loss of which is reported required for ferroptosis (41). Furthermore, IHC staining of CD71 showed ENDS-treated mice have elevated levels of CD71 in the lungs ( Figure 6C , Supplementary Figure 1A ). Then we treated the human bronchial epithelial cell line (16HBE) with or without ENDS (RSL3 as a ferroptosis positive control) and measured ferroptosis by determining the amount of lipid peroxides in cellular membranes using the BODIPY-C11 probe. The data showed that ENDS-treated 16HBE has increased BODIPY-positive ferroptosis cells ( Figure 6D ). The western blot results further supported that ENDS increased the expression level of CD71 and ACSL4, confirming ENDS could induce ferroptosis in bronchial epithelial cells ( Figure 6E ). In order to verify the impact of ENDS on ferroptosis in epithelial cells, we performed immunofluorescent staining with ferroptosis biomarker, CD71. We found ferroptotic type I alveolar epithelial cells (CD71⁺HOPX⁺) and ferroptotic type II alveolar epithelial cells (CD71⁺proSP-C⁺) are both elevated after ENDS exposure ( Figures 7A, B , Supplementary Figures 1B, C ), suggesting that ENDS induced ferroptosis in epithelial cells, which is partially responsible for alveolar degradation and destruction.