Electronic versus Combustible Cigarette Effects on Inflammasome Component Release into Human Lung

To the Editor:

Electronic cigarettes (e-cigarettes) include a wide variety of battery-powered devices that produce an aerosol for inhalation by heating a liquid that contains a solvent (usually composed of propylene glycol and/or glycerin), flavoring, and nicotine. The popularity of e-cigarette use has increased in recent years among both smokers and nonsmokers. Thirteen percent of U.S. adults reported e-cigarette use in 2013, up from 1.8% in 2010, but a third of current users were nonsmokers or former tobacco smokers (1). In this regard, the potential for functional effects of e-cigarette use on the first-pass organ, the human lung, remain understudied.

In this context, lung immunosurveillance is partly dependent on the sensing of danger by macrophage inflammasomes. Inflammasomes are multimeric complexes formed in response to various stimuli and are vital to the clearance of pathogens (2). Most inflammasomes activate caspase-1 via an adaptor molecule, an apoptosis-associated speck-like protein containing a caspase activation and recruitment domain (ASC) that promotes pro-IL-1β cleavage to mature IL-1β and induces cell pyroptosis. Furthermore, it has been proposed that extracellular ASC may have additional prion-like activity, as ASC characteristically forms polymers that may cause amyloid-like deposits that promote tissue injury (3). Of note, elevated ASC has also been reported in BAL fluid obtained from humans, and in particular patients with chronic obstructive pulmonary disease (3). We therefore elected to determine whether e-cigarette use may promote ASC release into the lung by comparing BAL fluid ASC levels from normal healthy never-smokers, smokers, and e-cigarette users.

Methods

Forty-three normal subjects composed of smokers, never-smokers, and e-cigarette users (Tables 1 and 2) underwent elective bronchoscopy with 100–140 ml saline BAL fluid after Institutional Review Board (IRB 2015C0088)–approved informed consent, as part of a larger study analyzing the effects of e-cigarettes on lung biomarkers. Smokers were delineated as those who smoked at least 10 cigarettes a day for at least 6 months with no prior e-cigarette use within a year. E-cigarette users were defined as nonsmokers who used at least one nicotine-containing cartridge per day or 1 ml liquid per day for at least 3 months, 12 of whom were former smokers (>5 mo; mean = 8 yr). Never-smokers had smoked fewer than 100 cigarettes in their lifetime, and not for at least 1 year, with no e-cigarette use within that timeframe. Nonsmoking status was confirmed by salivary cotinine (NicAlert; Nymox Pharmaceuticals). All participants were randomly assigned to the right or left lung for bronchoscopy in a 1:1 ratio within each smoking status group (never-smokers, e-cigarette users, and smokers). BAL fluid samples were spun down at 300g at 4°C to separate cells from the supernatant. Cell-free lavage fluid was stored in aliquots and frozen at −80°C until processed. BAL fluid was analyzed by sandwich ELISAs for ASC and caspase-1 combining commercial mouse monoclonal with rabbit polyclonal detection antibodies generated in our laboratory (4) and a bead-based ELISA for IL-1β (Meso Scale Discovery).

Table 1.

Baseline Demographics by Smoking Status

Characteristics	Never-Smokers (n = 12)	E-Cig Users (n = 15)	Smokers (n = 16)
Age, yr, mean (range)	26 (21–30)	27 (21–30)	26 (21–30)
Sex
Female, n (%)	7 (58%)	5 (33%)	4 (25%)
Male, n (%)	5 (42%)	10 (67%)	12 (75%)
Race
White, n (%)	10 (83%)	12 (80%)	14 (88%)
Black or African American, n (%)	1 (8%)	1 (7%)	1 (6%)
Asian, n (%)	1 (8%)	2 (13%)	1 (6%)
Smoking
Former, n (%)	0 (0%)	12(80%)	0 (0%)
Current, n (%)	0 (0%)	0 (0%)	16 (100%)
Never, n (%)	12 (100%)	3 (20%)	0 (0%)
Years of smoking, mean (range)	—	8 (1–15)*	8 (0.6–10)
Cigarettes/d, mean (range)	—	13 (2–20)*	16 (10–20)
Years since last cigarettes, mean (range)	—	2 (0.4–4.1)*	—
Vaping
Years of vaping, mean (range)	—	3 (0.5–4)	—
Puffs/d, mean (range)	—	127 (20–600)	—
E-liquid, ml/d, mean (range)	—	7 (2–20)	—
Nicotine, mg/ml, mean (range)	—	12 (1.5–36)	—

Definition of abbreviation: E-cig = e-cigarette.

^*Prior smoking e-cig users.

Table 2.

Cross-Sectional Analysis of BAL Fluid Cell Counts

BAL Fluid Cell Count	Never-Smokers (n = 11)*	E-Cig Users (n = 13)*	Smokers (n = 16)	P Value†
				Never-Smoker vs. E-Cig User	Never-Smoker vs. Smoker	E-Cig User vs. Smoker
Total cell concentration, ×106/L, median (IQR)	240 (168–267)	306 (212–369)	434 (300–699)	0.41	0.003	0.14
Macrophages, ×106/L, median (IQR)	212 (159–264)	265 (184–321)	411 (289–652)	0.53	0.006	0.06
Lymphocytes, ×106/L, median (IQR)	5 (0–16)	13 (6–42)	10 (1–16)	NA‡	NA‡	NA‡
Neutrophils, ×106/L, median (IQR)	0 (0–6)	3 (2–20)	10 (2–21)	0.14	0.05	0.72

Definition of abbreviations: e-cig = e-cigarette; IQR = interquartile range; NA = not applicable.

^*Excluded data because of contamination with blood, which altered the total cell count.

^†Kruskal-Wallis followed by all-pairs using Steel-Dwass test to control for the multiple comparisons.

^‡Pairwise not included because of the Kruskal-Wallis test not being statistically significant.

Results

BAL samples were processed for total and differential cell counts. Smokers had the highest total cell counts, which were statistically significant compared with never-smokers (P = 0.003; Tables 1 and 2). Smokers had the highest total macrophage counts compared with never-smokers and e-cigarette users, but were only statistically significant compared with never-smokers (P = 0.006). Regarding BAL fluid inflammasome components, ASC and caspase-1 were readily detected in unconcentrated fluid, and each correlated with macrophage concentrations (r² = 0.52 [P < 0.0001] and r² = 0.55 [P < 0.0001], respectively).

ASC in BAL fluid was higher in smokers and e-cigarette users when compared with never-smokers (P = 0.0005 and P = 0.04, respectively), but e-cigarette user ASC levels were not statistically different from smokers (Figure 1). Caspase-1 and IL-1β levels largely paralleled the ASC data. Caspase-1 levels were higher in smokers than in e-cigarette users and nonsmokers (P = 0.04 and P = 0.0003, respectively), but e-cigarette user caspase-1 was not different from nonsmokers. Finally, IL-1β levels in BAL fluid were highest in smokers (median, 4.6 pg/ml; interquartile range [IQR], 2.7–7.3) and statistically significant compared with both e-cigarette users and never-smokers (median, 1.0 pg/ml [IQR, 0.8–1.6; P = 0.0005], and median, 1.0 pg/ml [IQR, 0.7–1.2; P = 0.001], respectively).

Figure 1.

ASC and caspase-1 levels in BAL fluid by smoking status. Cell-free BAL fluids from never-smokers (n = 12), e-cigarette users (n = 15), and smokers (n = 16) were analyzed by ELISA for ASC and caspase-1. Box plots show medians with interquartile range (IQR), and whiskers show minimum and maximum values. ASC was significantly higher in smokers than never-smokers (median, 37 ng/ml [IQR, 21–64 ng/ml] vs. 11 ng/ml [IQR, 9–15 ng/ml], respectively). E-cigarette user ASC levels (median, 22 ng/ml [IQR, 14–35 ng/ml]) were intermediate to never-smokers and smokers but only statistically significant compared with never-smokers. Caspase-1 levels were significantly lower in never-smokers compared with smokers (median, 12 pg/ml [IQR, 6–14 pg/ml] vs. 42 pg/ml [IQR, 27–68 pg/ml], respectively). The median caspase-1 level was 16 pg/ml [IQR, 9–35 pg/ml] for e-cigarette users, which was statistically different from that of smokers. Kruskal-Wallis was applied comparing inflammasome components among the three smoking status groups. The Steel-Dwass test was used for multiple pairwise comparisons. *P < 0.05, **P < 0.01, and NS = no significant difference. ASC = apoptosis-associated speck-like protein containing a caspase activation and recruitment domain; e-cig = electronic cigarette.

Discussion

This study is the first cross-sectional study of inflammasome protein release in human BAL fluid to compare various smoking categories that include exclusive e-cigarette use; most of the participants were former smokers. Consistent with previous studies, smokers had higher inflammatory total cell counts and macrophage counts compared with never-smokers (5). It is noteworthy that e-cigarette macrophage counts were intermediate between smokers and never-smokers. These findings are consistent with murine (6) and human studies (7) that suggest e-cigarette use can affect innate immunity.

In the context of inflammasome activation, the ASC levels in BAL fluids of all subjects were higher than expected. When normalized to estimated epithelial lining fluid (i.e., about 100× diluted in BAL [8]), smoker ASC approximated 4 μg/ml epithelial lining fluid (i.e., about 1,000 times higher than what we have observed for plasma concentrations; unpublished observation). The significance of this lung-centric elevation of ASC remains to be determined, but it has been linked to “prionoid” functions of oligomeric ASC (9, 10). Importantly, ASC levels did reflect smoking status. E-cigarette users had ASC levels intermediate to those of never-smokers and smokers. Both caspase-1 and IL-1β levels mimicked the ASC levels in the three groups, suggesting that the ASC levels in part reflect inflammasome activation, although there was no notable difference in IL-1β levels between never-smokers and e-cigarette users. Taken as a whole, these data suggest that e-cigarette use contributes less to lung inflammation than does combustible tobacco use. However, in this cross-sectional study, e-cigarette use did provide an inflammatory signal above that of never-smokers. Because we cannot exclude residual influences from prior combustible tobacco use in our e-cigarette cohort, longitudinal studies of e-cigarette use in never-smokers are still needed to better characterize the inflammatory effects of e-cigarette use. Furthermore, future work will need to address the relative effect of nicotine versus other components, including carriers and flavorings, which were not addressed in this study.