Formaldehyde and the transient receptor potential ankyrin-1 contribute to electronic cigarette aerosol-induced endothelial dysfunction in mice

Abstract

Electronic nicotine delivery systems (ENDS) aerosol exposures can induce endothelial dysfunction (ED) in healthy young humans and animals. Thermal degradation of ENDS solvents, propylene glycol, and vegetable glycerin (PG: VG), generates abundant formaldehyde (FA) and other carbonyls. Because FA can activate the transient receptor potential ankyrin-1 (TRPA1) sensor, we hypothesized that FA in ENDS aerosols provokes TRPA1-mediated changes that include ED and "respiratory braking"-biomarkers of harm. To test this, wild-type (WT) and TRPA1-null mice were exposed by inhalation to either filtered air, PG: VG-derived aerosol, or FA (5 ppm). Short-term exposures to PG: VG and FA-induced ED in female WT but not in female TRPA1-null mice. Moreover, acute exposures to PG: VG and FA stimulated respiratory braking in WT but not in TRPA1-null female mice. Urinary metabolites of FA (ie, N-1,3-thiazolidine-4-carboxylic acid, TCA; N-1,3-thiazolidine-4-carbonyl glycine, TCG) and monoamines were measured by LC-MS/MS. PG: VG and FA exposures significantly increased urinary excretion of both TCA and TCG in both WT and TRPA1-null mice. To confirm that inhaled FA directly contributed to urinary TCA, mice were exposed to isotopic 13C-FA gas (1 ppm, 6 h). 13C-FA exposure significantly increased the urine level of 13C-TCA in the early collection (0 to 3 h) supporting a direct relationship between inhaled FA and TCA. Collectively, these data suggest that ENDS use may increase CVD risk dependent on FA, TRPA1, and catecholamines, yet independently of either nicotine or flavorants. This study supports that levels of FA in ENDS-derived aerosols should be lowered to mitigate CVD risk in people who use ENDS.

Keywords: aldehydes; cardiovascular disease; electronic nicotine delivery systems (ENDS); endothelium; irritants; tobacco.

Graphical Abstract.

Endothelial dysfunction (ED) in aorta ex vivo after short-term exposure of female WT mice to PG: VG-derived aerosol or formaldehyde (FA) was absent in similarly exposed female TRPA1-null mice indicating a role of TRPA1. TRPA1-dependent nervous system-mediated respiratory braking response (lightning bolt in WT mice) was absent in female TRPA1-null mice. FA exposure also induced ED in WT but not in TRPA1-null mice. Irritant compounds such as FA in PG: VG-derived aerosols promote ED through a TRPA1-dependent mechanism. Notably, PG: VG-derived aerosol exposure in TRPA1-null mice induced a compensatory response in the aorta (excessive relaxation with ACh). These findings suggest exposures to E-cig-derived aerosols have complex effects on the vasculature, and thus, long-term E-cig use likely will induce ED and increase CVD risk. Created with BioRender.com.

Fig. 1.

ED of short-term inhalation exposure in female WT mice. Endothelium function was measured in thoracic aorta ex vivo after 4 days of exposure of female WT mice to either filtered air, PG: VG (30:70, 6 h/d, 4 d) or FA (5 ppm). Acetylcholine (ACh) was used to assess concentration- and endothelium-dependent relaxation efficacy (% relaxation; A) and sensitivity (EC₅₀; B) in phenylephrine (PE)-precontracted aortic rings by isometric myography. Changes in the relaxation to an endothelium-independent, nitric oxide donor vasorelaxant, sodium nitroprusside (SNP), were also measured by efficacy (% relaxation; C) and sensitivity (EC₅₀; D) in thoracic aorta of female WT mice. Values = mean ± SE (n = 13–30 mice per group); * P < 0.05 versus air control.

Fig. 2.

ED of short-term inhalation exposure in female TRPA1-null mice. Endothelium function was measured in thoracic aorta ex vivo after 4 days of exposure of female TRPA1-null mice to either filtered air, PG: VG (30:70, 6 h/d, 4 d) or FA (5 ppm, 6 h/d, 4 d). Acetylcholine (ACh) was used to assess concentration- and endothelium-dependent relaxation efficacy (% relaxation; A) and sensitivity (EC₅₀; B) in phenylephrine (PE)-precontracted thoracic aorta rings by isometric myography. Changes in the relaxation to the endothelium-independent, nitric oxide donor vasorelaxant, sodium nitroprusside (SNP) were also measured by efficacy (% relaxation; C) and sensitivity (EC₅₀; D) in thoracic aorta of female TRPA1-null mice. Values = mean ± SE (n = 5–25 mice per group); * P < 0.05 versus air control.

Fig. 3.

Aortic contractile toxicity of short-term inhalation exposure in female WT and female TRPA1-null mice. Thoracic aortic function was measured ex vivo following 4 days of exposure of female WT (A, B, C) and female TRPA1-null (D, E, F) mice to either filtered air, PG: VG (360 puffs/d, 6 h/d, 4 d) or FA (5 ppm, 6 h/d, 4 d). The efficacy (mN/mm; A, D) and sensitivity (EC₅₀; B, E) of concentration-dependent, phenylephrine (PE)-induced contractions in mid-thoracic aortic rings were measured by isometric myography. The PE Contraction Ratio (PECR: a summed measure of both the contribution of NOS activity and the maximal PE-induced contraction) was also measured in all groups (C, F). Values = mean ± SE (n = 5–30 WT mice per group and n = 3–25 TRPA1-null mice); * P < 0.05 versus air control; ^#  P < 0.05 versus air control group.

Fig. 4.

Effects of exposures on pulmonary function in female mice. Female WT and TRPA1-null mice were instrumented with pressure transmitters to detect changes in pleural cavity pressure, and after 1 week of recovery, mice were exposed to filtered air (15 min) punctuated by either 3–9 min sessions of PG: VG-derived aerosols (2 puffs/min; A, B) or 2–9-min sessions of FA (5 ppm and 10 ppm, respectively; C, D). Pressure waveforms were analyzed for respiratory rate (breaths per min, bpm; A, C) or expiratory time (s; B, D). Onset of exposures in WT mice but not in TRPA1-null mice was followed by rapid, robust onset of effects whereas upon cessation there was a rapid reversal of changes back to baseline. Values = 1 min mean ± SE (n = 3 mice per group); * P < 0.05 exposed mice versus baseline level in genotype-matched mice. Solid line represents real-time monitoring of either total suspended particle matter (TSP, g/m³) or FA level (ppm), respectively, as measured upstream of whole-body exposure chambers via inline infrared detector (Casella) or electrochemical detector (MultiRAE Pro), respectively.

Fig. 5.

Urinary excretion of FA metabolites following a 6 h exposure of female WT and TRPA1-null mice to either air, PG: VG-derived aerosol, or FA. Within the first 3 h post-exposure, the FA metabolite (thiazolidine-4-carboxylic acid, TCA) was significantly increased in the urine of female WT and TRPA1-null mice exposed to both PG: VG-derived aerosol (6 h) and FA (5 ppm, 6 h) compared with respective filtered air exposure (6 h) group (A). In the late post-exposure collection (3–18 h after exposure), TCA levels in the urine of female WT and TRPA1-null mice exposed to PG: VG-derived aerosol remained elevated whereas levels after FA exposure had returned toward baseline levels present in respective filtered air group (B). Within the first 3 h post-exposure, the thiazolidine carbonyl glycine (TCG) metabolite of FA was significantly increased in the urine of female WT and TRPA1-null mice exposed to PG: VG-derived aerosol (6 h) but only in TRPA1-null mice exposed to FA (5 ppm, 6 h) compared with respective filtered air exposure (6 h) group (C). In the late post-exposure collection (3 to 18 h after exposure), TCG levels in the urine of female WT and TRPA1-null mice exposed to PG: VG-derived aerosol and to FA had returned toward baseline levels present in the respective filtered air group (D). Within the first 3 h post-exposure to ¹³C-FA (1 ppm; red outlined diamonds), the ¹³C-FA metabolite (¹³C-TCA) was significantly increased in the urine of female WT mice compared with the level of ¹³C-4-TCA in respective filtered air group (E). In the late post-exposure collection (3 to 18 h after ¹³C-FA exposure), the ¹³C-TCA level in the urine of female WT mice was elevated above baseline level (<limit of detection, LOD) in the filtered air group (F). Values = mean ± SE (n = 3–4 mice per group); * P < 0.05 versus genotype- and time-matched air control; ^# 0.05<P < 0.10 versus genotype- and time-matched air control.