Abstract
After more than a decade of electronic cigarette (E-cig) use in the United States, uncertainty persists regarding E-cig use and long-term cardiopulmonary disease risk. As all E-cigs use propylene glycol and vegetable glycerin (PG-VG) and generate abundant saturated aldehydes, mice were exposed by inhalation to PG-VG–derived aerosol, formaldehyde (FA), acetaldehyde (AA), or filtered air. Biomarkers of exposure and cardiopulmonary injury were monitored by mass spectrometry (urine metabolites), radiotelemetry (respiratory reflexes), isometric myography (aorta), and flow cytometry (blood markers). Acute PG-VG exposure significantly affected multiple biomarkers including pulmonary reflex (decreased respiratory rate, −50%), endothelium-dependent relaxation (−61.8 ± 4.2%), decreased WBC (−47 ± 7%), and, increased RBC (+6 ± 1%) and hemoglobin (+4 ± 1%) versus air control group. Notably, FA exposure recapitulated the prominent effects of PG-VG aerosol on pulmonary irritant reflex and endothelial dysfunction, whereas AA exposure did not. To attempt to link PG-VG exposure with FA or AA exposure, urinary formate and acetate levels were measured by GC-MS. Although neither FA nor AA exposure altered excretion of their primary metabolite, formate or acetate, respectively, compared with air-exposed controls, PG-VG aerosol exposure significantly increased post-exposure urinary acetate but not formate. These data suggest that E-cig use may increase cardiopulmonary disease risk independent of the presence of nicotine and/or flavorings. This study indicates that FA levels in tobacco product-derived aerosols should be regulated to levels that do not induce biomarkers of cardiopulmonary harm. There remains a need for reliable biomarkers of exposure to inhaled FA and AA.
NEW & NOTEWORTHY Use of electronic cigarettes (E-cig) induces endothelial dysfunction (ED) in healthy humans, yet the specific constituents in E-cig aerosols that contribute to ED are unknown. Our study implicates formaldehyde that is formed in heating of E-cig solvents (propylene glycol, PG; vegetable glycerin, VG). Exposure to formaldehyde or PG-VG–derived aerosol alone stimulated ED in female mice. As ED was independent of nicotine and flavorants, these data reflect a “universal flaw” of E-cigs that use PG-VG.
Listen to this article’s corresponding podcast at https://ajpheart.podbean.com/e/e-cigarettes-aldehydes-and-endothelial-dysfunction/.
Keywords: aldehydes, cardiovascular disease, electronic cigarettes, endothelium, irritants, tobacco
INTRODUCTION
Cardiovascular disease (CVD) is the leading cause of morbidity and mortality worldwide (1, 2). In 2016, more than one-fourth of the almost 57 million global deaths were attributed to CVD deaths, and ischemic heart disease and stroke have been classified as the two leading causes of death for the last 20 years (2). Although a number of physical, behavioral, and socioeconomic factors have been linked to CVD (3), tobacco smoke has been characterized as the single most significant modifiable risk factor in CVD development (4, 5). Extensive evidence demonstrates that exposures to both mainstream (MCS) (6, 7) and secondhand (SHS) (6, 8) cigarette smoke increase the risk for CVD. Smoking increases the risk of heart attack, coronary artery disease, atherosclerosis, and stroke (9). Smoking is also linked with endothelium dysfunction (9), sine qua non in atherosclerosis. However, the constituents in tobacco smoke that promote CVD and the mechanisms by which these constituents enhance CVD risk remain enigmatic.
After more than a decade of electronic cigarette (E-cig) use in the United States, uncertainty persists regarding E-cig use and long-term CVD risk. Exposure to MCS or SHS significantly increases exposure to formaldehyde (FA) and acetaldehyde (AA) as these are major constituents (10, 11), and the Institute of Medicine ranks FA and AA as two of the most significant toxins in MCS, particularly in relationship to non-cancer disease risk, i.e., CVD (12). Similarly, E-cigs generate a variety of abundant aldehydes including saturated aldehydes (and to a much lesser quantity unsaturated aldehydes) dependent on propylene glycol (PG) and vegetable glycerin (VG) ratio (PG-VG), power of the device and user topography as key factors (13–17). Regardless of differences in E-cig device attributes, two saturated aldehydes, i.e., FA and AA, are generated abundantly (15, 16). Thus, we hypothesized that these aldehydes likely are potential mediators of PG-VG–induced harm.
FA and AA also are ubiquitous in a variety of food and drinks, and anthropogenic sources account for the majority of human exposures (18–20). FA is in consumer products, including carpets, manufactured wood products, paints, and preservatives, and both FA and AA are formed as products of combustion from industry, vehicle exhaust, and fires (18, 19, 21, 22). The negative cardiovascular effects of FA and AA exposure have been documented. FA causes hematotoxicity, decreasing both red blood cell (RBC) and white blood cell (WBC) and platelet counts (23–26) as well as the levels of myeloid progenitor cells (23, 26). FA exposure also decreases blood pressure via inhalation in rats (27) and ingestion in humans (28), consistent with FA-induced concentration-dependent vascular relaxation in vitro (29, 30). Occupational exposure to FA is also associated with increased incidence of heart disease (31–33), and a brief (90 min) exposure of young, healthy women to formalin led to acute onset of conduit blood vessel endothelial dysfunction (34). The cardiovascular effects of AA have largely been studied in relationship to aldehyde dehydrogenase (ALDH2), the enzyme responsible for AA metabolism, and specifically ALDH2 mutation that prevents AA metabolism, causing blood levels of AA to increase after ethanol consumption (35). Increased blood levels of AA are linked with AA-hemoglobin adducts (36, 37) and CVD risk (38–40).
Despite the significant exposures to FA and AA from MCS and E-cig–derived aerosols as well as other sources, few studies have addressed the effects of direct inhalation exposures to either FA or AA alone on cardiovascular biomarkers of harm such as endothelial dysfunction. To address this gap in knowledge regarding the potential cardiovascular toxicity of E-cig aerosols and two abundant saturated aldehydes (FA, AA), healthy adult male and female C57BL/6J mice were exposed acutely by inhalation to PG-VG, FA, or AA, and biomarkers of harm and exposure were measured.
MATERIALS AND METHODS
Materials
Reagent-grade chemicals were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise stated. The data sets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Mice and Exposures
Mice.
Male and female C57BL/6J (wild-type, WT) mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were treated according to the “Guiding Principles for the Care and Use of Animals in Research and Teaching” as adopted by the American Physiological Society, and all protocols were approved by the University of Louisville Institutional Animal Care and Use Committee. Before and during exposures, mice were housed under pathogen-free conditions, controlled temperatures, and a 12:12 h light-dark cycle. Mice were maintained on a standard chow diet (Rodent Diet 5010, 4.5% fat by weight, LabDiet, St. Louis, MO).
E-cigarette aerosol exposures.
A software-controlled (FlexiWare) cigarette-smoking robot (SCI-REQ, Montreal, Canada) system was used in the mechanical generation of aerosols from PG-VG mixtures. To control the generation of volatile organic compounds (VOCs) in E-cig aerosols, we used a defined E-cig platform. PG-VG mixture (50:50 or 30:70 ratio, vol/vol) was loaded into a refillable, clear tank atomizer with a fixed coil resistance (Mistic Bridge; ∼3.0 ohm; purchased online) coupled with a rechargeable bluPLUS+ (3.7 V) battery (power output of ∼8 W) (41). The atomizer tank was weighed before and after use to quantify solution consumption (g/puff). A 9-min session was composed of 18 puffs (4 s/puff, 91.1 mL/puff, 2 puffs/min). For respiratory parameters, each exposure had 3 PG-VG sessions evenly spaced over 1 h. For other exposures, 20 sessions were evenly spaced over a 6-h exposure per day for 4 consecutive days. Total suspended particulate (TSP) matter was monitored in real time with an inline infrared Microdust Pro 880 nm (Casella) positioned upstream of the exposure chamber (5 L, SCI-REQ). All whole body exposures were done between 7:00 AM and 2:00 PM in the absence of food or water.
Formaldehyde and acetaldehyde exposures.
For FA and AA exposures, naïve mice were exposed to either HEPA- and charcoal-filtered air, FA, or AA for either two 9-min sessions in 1 h (respiratory parameters) or for 4 consecutive days (5 ppm, 6 h/day) using an exposure system equipped with a certified permeation tube in a calibrated heating oven (Kin-Tek, LaMarque, TX) as described (42). Levels of AA were monitored real time with an inline photoionization detector (isobutylene calibration), whereas levels of FA were monitored real time with an inline electrochemical sensor (CO calibration) (MultiRAE Pro, RAE Systems, Burlington, VT). Both gases were monitored upstream of the exposure chamber (30 liter, flow 7 lpm). All whole body exposures were done between 7:00 AM and 2:00 PM in the absence of food or water.
Urine Collection and Metabolism
Urine collection.
Prior to exposures, mice were held and a small drop of d-glucose-saccharin solution (wt/wt 3.0%/0.125%; Sigma-Aldrich; St. Louis, MO) was touched to their lips and mouth. After 6 h of air and toxic exposures, mice were placed singly per metabolic cage (Harvard Apparatus, Cambridge, MA) without food but with glucose-saccharin solution in drinking water for urine collection (in graduated cylinders surrounded by 4°C water-jacketed organ baths). Urine was collected from 0 to 3 h post exposure, and in a second overnight collection (3–16 + h, overnight) during which mice were provided glucose-saccharin solution as well as food (43). Urine samples were centrifuged (1,800 g, 5 min; to pellet feces or food) before being decanted and stored at −80°C.
Urine metabolite analysis.
Urinary levels of formate and acetate, the primary metabolites of FA and AA, respectively, were measured by gas chromatography-mass spectrometry (GC-MS) as adapted and modified from previous reports (44, 45). Urine (20 µL) was mixed with sodium phosphate (20 µL; 0.5 M, pH 8.0) containing [13C]formate (2.3 mM) and [13C]acetate (0.23 mM) internal standards, and pentafluorobenzyl bromide (130 µL, 0.1 M). The mixture was vortexed for 1 min and then incubated at 60°C for 15 min, and the resulting reaction products were extracted using hexane (300 µL) before being transferred to glass tubes for GC-MS analysis. Analytes in urine samples were quantified using the peak area ratio based on seven-point standard curves that were run before and after the urine samples. MassHunter software (Agilent) was used for peak integration, calibration, and quantification. Measured formate and acetate sample concentrations were corrected for the natural abundance of 13C-isotopes, and normalized to urinary creatinine (41). Additionally, we estimated the total excreted formate and acetate by multiplying measured concentrations (ng/mL) and total volume (mL) of urine collected at each time point (0–3 h, overnight post exposure).
Systemic Outcomes
Pulmonary (irritant) reflexes.
To measure pulmonary irritant reflexes real time during exposures (i.e., respiratory rate, inspiratory and expiratory time, and amplitude), male mice were implanted with a pressure cannula tunneled into the esophageal serosa via the diaphragm for radiotelemetry recordings (PA-C10; DSI, St. Paul, MN). Mice recovered for 1-wk post surgery before any exposure. Signals were continuously acquired before (baseline), during, and after exposures (1 kHz). Respiratory rate (breaths per minute, bpm), amplitude (mmHg), and expiratory and inspiratory times (s) were calculated from pressure signal waveform (DSI) (46).
Euthanization.
Immediately following the final exposure, mice were euthanized with sodium pentobarbital (≈150 mg/kg ip), ventral thoracotomy, and exsanguination via right ventricle cardiac puncture for blood collection in EDTA-coated (0.2 M) syringes. Organs were removed, weighed, and snap frozen in liquid N2 and kept at −80°C until further analysis.
Complete blood counts.
Complete blood counts (CBC) were measured (20 µL whole blood) with a hematology analyzer calibrated with multispecies hematological reference controls (Hemavet 950FS; Drew Scientific, Inc., Miami Lakes, FL) as described (43).
Plasma biomarkers.
Plasma total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, albumin, total protein, alanine transaminase (ALT), aspartate transaminase (AST), lactate dehydrogenase (LDH), creatine kinase (CK), and creatinine were measured on a Cobas Mira Plus Clinical Chemistry Autoanalyzer (Roche Diagnostics, Indianapolis, IN) as previously described (43).
Flow Cytometry
Circulating angiogenic cells (CACs).
Blood CACs were identified by flow cytometry and quantified as events double positive for Flk-1 (endothelial marker; Vegfr2 homolog) and Sca-1 (hematopoietic stem cell marker) as previously described (47).
Platelet-leukocyte aggregates (PLAs).
PLAs were identified by flow cytometry and quantified as events double positive for CD41 (platelets) and CD45 (leukocytes) as previously described (43) with slight modifications. Briefly, aliquots of whole blood were diluted (1:4) with HEPES-Tyrode solution before fixation [paraformaldehyde, Fc 1.6%, room temperature (RT)]. Red blood cells were lysed (MilliQ water), and the sample was centrifuged (400 g, 5 min, RT). The sample pellet was incubated with 1% Fc Block (5 µL; 10 min) before staining for 30 min with FITC-labeled anti-CD41 and APC-labeled anti-CD45 or isotype-matched negative controls (FITC-IgG1; APC-IgG2bκ). Stained cells were washed with HEPES-Tyrode solution containing 1% BSA, centrifuged at 400 g for 5 min, and resuspended in HEPES-Tyrode solution (250 µL). A BD LSR Flow Cytometer (BD Biosciences, San Jose, CA) was used to analyze stained cells; a minimum of 20,000 events was collected for each sample.
Endothelium Dysfunction
Immediately after the final exposure (4th day), male and female mice were euthanized and the thoracic aorta was isolated and placed into cold (4°C) physiological salt solution (PSS). Thoracic aorta rings (3–4 mm) were carefully cleaned and prepared for isometric myography as described (42).
Physiological salt solutions (PSS).
PSS for aorta was (in mM) NaCl, 118; KCl, 4.7; CaCl2, 2.5; KH2PO4, 1.2; MgSO4, 1.2; NaHCO3, 25; and, glucose, 5.5; pH 7.4. High K+ PSS (100 mM) was substituted with equimolar K+ for Na+ (29).
Aorta reactivity.
Organ baths contained PSS bubbled with 95% O2-5% CO2 at 37°C. Briefly, after 10 min without tension, aortic segments were equilibrated to ≈1 g of loading tension over 1 h. All segments were stimulated with high K+ to test for viability, washed three times with PSS over 30 min, and re-equilibrated to 1 g of resting tension. The segments were then stimulated again with high K+, followed by three bath changes and a re-equilibration to resting tension. To test whether exposures altered vascular reactivity, the following responses were measured: 1) contractions induced by high K+; 2) concentration-dependent contractions of phenylephrine (PE; 0.1 nM to 10 µM); 3) concentration-dependent relaxations of acetylcholine (ACh; 0.1 nM to 10 µM) in PE-precontracted segments; and 4) concentration-dependent relaxations of sodium nitroprusside [SNP; 0.01 nM to 10 µM; nitric oxide (NO) donor] in PE-precontracted segments after addition of NG-nitro-l-arginine methyl ester (l-NAME) (100 µM) and addition of ACh (10 µM). Measures of efficacy (Emax: contractions normalized to aortic length and percentage relaxation of PE contraction) (48) and sensitivity (EC50, effective concentration producing 50% response, i.e., cumulative concentration responses normalized to 100% with interpolation of EC50) were calculated (49). To assess specific alteration in endothelial nitric oxide synthase (eNOS) function, the effect of l-NAME on PE-induced tension was calculated as a “PE contraction ratio”: PE tensionpost-L-NAME/PE tensionpre-L-NAME.
Statistics
Data are presented as means ± SE. For two group comparisons, rank sum tests with Bonferroni’s post test or paired (or one-way repeated measures ANOVA) or unpaired t tests as appropriate were applied. For multiple group comparisons, one-way ANOVA with Bonferroni’s post test or when variation indicated Kruskal-Wallis ANOVA on ranks with Dunn’s post test were used (SigmaPlot, version 12.5; Systat Software, Inc., San Jose, CA). Statistical significance was set at P < 0.05.
RESULTS
Pulmonary Irritant Responses
Because PG-VG aerosols contain high levels of reactive aldehydes that can stimulate pulmonary irritant reflexes (e.g., respiratory braking, cough), we tested whether PG-VG aerosol was irritating to mice. We used radiotelemetry to monitor real-time changes in respiratory parameters before, during, and after exposures as previously performed during acrolein exposures in mice (46).
Propylene glycol-vegetable glycerin.
Within the first puff of PG-VG, respiratory parameters were significantly altered and remained altered until the end of the 9-min vaping session (18 puffs) after which parameters returned toward baseline (Fig. 1). Moreover, similar irritant responses (temporally, directionally, and magnitude) were observed in each successive vaping session. The ∼50% decrease in respiratory rate (Fig. 1A) with reciprocal increases in both expiratory (strongly) (Fig. 1B) and inspiratory (minimally) (Fig. 1C) times were hallmark changes of an “irritant” (or nocifensive) response. Amplitude (a measure of thoracic effort) also was lessened slightly by PG-VG aerosol perhaps reflecting shallower as well as slower breathing (Fig. 1D).
Figure 1.
Effects of PG-VG–derived aerosol on pulmonary function in mice. Male C57BL/6J mice instrumented with pressure transmitters to detect changes in pleural cavity pressure were exposed to filtered air or three 9-min sessions (18 puffs/session) of PG-VG–derived aerosol. Pressure waveforms were analyzed for respiratory rate (breaths per min, bpm; A), expiratory (B), and inspiratory (C) times (s), and, peak amplitude (mmHg, a measure of respiratory effort; D). Onset of exposure to PG-VG–derived aerosol provoked rapid and robust changes in respiratory rate (A), expiratory (B), and inspiratory (C) times, and, more modest changes in peak amplitude (D). Values are 1-min means ± SE (n = 4 mice per group) (*P < 0.05) vs. air control; 0.10 > #P > 0.05 vs. air control. Magenta line represents real-time total suspended particulate matter (TSP; g/m3) measured upstream of whole body exposure chamber (5-l) using a Microdust Pro 880 nm. ○, air; ●, PG-VG. PG, propylene glycol. VG, vegetable glycine.
Formaldehyde.
To test if the PG-VG–provoked response was potentially due to FA, an abundant aldehyde generated from PG-VG, mice were exposed to FA (5 ppm). Although initial exposure to 5 ppm FA had minimal effect on respiratory parameters, there was a rapid noticeable alteration in respiratory parameters during a second FA challenge when FA levels briefly increased to almost 10 ppm (Fig. 2). These changes included both robust depression of respiratory rate (Fig. 2A) and increased expiratory time (Fig. 2B) with less effect on inspiratory time (Fig. 2C) and amplitude (Fig. 2D). Thus, FA at levels between 5 and 10 ppm stimulated an irritant response that mimicked, in part, the robust response to PG-VG aerosol.
Figure 2.
Effects of formaldehyde (FA) on pulmonary function in mice. Male C57BL/6J mice instrumented with pressure transmitters to detect changes in pleural cavity pressure were exposed to filtered air or two 9-min sessions of FA (5-10 ppm). Pressure waveforms were analyzed for respiratory rate (breaths per min, bpm; A), expiratory (B), and inspiratory (C) times (s), and, peak amplitude (mmHg, a measure of respiratory effort; D). Onset of exposure to FA had modest effect at 5 ppm but when >5 ppm provoked rapid and robust changes in respiratory rate (A) and expiratory time (B). Less noticeable changes occurred in inspiratory (C) time and peak amplitude (D). Values are 1-min means ± SE (n = 2 mice per group); 0.10 > #P > 0.05 vs. air control. Blue line represents real time monitoring of FA level (ppm) measured upstream of whole body exposure chamber (30-L) with an inline MultiRAE Pro. ○, air; ●, FA.
Acetaldehyde.
As AA is also an abundant aldehyde generated in PG-VG aerosol, we tested for effects of AA. In contrast to “respiratory braking” effects observed with both PG-VG aerosol and FA, AA exposure (5–10 ppm) had no obvious effects on any respiratory parameter (Fig. 3). These data indicate AA did not trigger an irritant response at any level reached under these exposure conditions.