Global Air Pollutant Phenanthrene and Arrhythmic Outcomes in a Mouse Model

Background: The three-ringed polycyclic aromatic hydrocarbon (PAH) phenanthrene (Phe) has been implicated in the cardiotoxicity of petroleum-based pollution in aquatic systems, where it disrupts the contractile and electrical function of the fish heart. Phe is also found adsorbed to particulate matter and in the gas phase of air pollution, but to date, no studies have investigated the impact of Phe on mammalian cardiac function. Objectives: Our objectives were to determine the arrhythmogenic potential of acute Phe exposure on mammalian cardiac function and define the underlying mechanisms to provide insight into the toxicity risk to humans. Methods: Ex vivo Langendorff-perfused mouse hearts were used to test the arrhythmogenic potential of Phe on myocardial function, and voltage- and current-clamp recordings were used to define underlying cellular mechanisms in isolated cardiomyocytes. Results: Mouse hearts exposed to ∼8μM Phe for 15-min exhibited a significantly slower heart rate (p=0.0006, N=10 hearts), a prolonged PR interval (p=0.036, N=8 hearts), and a slower conduction velocity (p=0.0143, N=7 hearts). Whole-cell recordings from isolated cardiomyocytes revealed action potential (AP) duration prolongation (at 80% repolarization; p=0.0408, n=9 cells) and inhibition of key murine repolarizing currents—transient outward potassium current (Ito) and ultrarapid potassium current (IKur)—following Phe exposure. A significant reduction in AP upstroke velocity (p=0.0445, n=9 cells) and inhibition of the fast sodium current (INa; p=0.001, n=8 cells) and calcium current (ICa; p=0.0001) were also observed, explaining the slowed conduction velocity in intact hearts. Finally, acute exposure to ∼8μM Phe significantly increased susceptibility to arrhythmias (p=0.0455, N=9 hearts). Discussion: To the best of our knowledge, this is the first evidence of direct inhibitory effects of Phe on mammalian cardiac electrical activity at both the whole-heart and cell levels. This electrical dysfunction manifested as an increase in arrhythmia susceptibility due to impairment of both conduction and repolarization. Similar effects in humans could have serious health consequences, warranting greater regulatory attention and toxicological investigation into this ubiquitous PAH pollutant generated from fossil-fuel combustion. https://doi.org/10.1289/EHP12775


Table of Contents
Table S1.Estimated IC50, n(H) and maximal block.
Table S2.Model output of effect of Phenathrene exposure on the human ventricular action potential.

Figure S1. (A)
A representative surface ECG from an ex vivo mouse heart showing the wave forms and intervals measured.Note the absence of a clear J-wave showing early repolarization.These were often undetectable due to interference with the QRS which is common for ex vivo mouse heart preparations and thus not quantified in this study.Lower panel shows the impact of acute phenanthrene (Phe) exposure on uncorrected QT interval (B), QRS duration (C) and Pwave duration (D).Average parameters calculated from isolated mouse (C57/BJ, Male, 10weeks) hearts at baseline and after 15-minute exposure to either control solution (circle, black), 2.1 µM Phe (square, red) or ~8 µM Phe (triangle, dark-red) and finally following a 15-minute wash-out period.Bars are mean+/-SEM, symbols are individual hearts (N=5-7 hearts).Two-way ANOVA mixed-effects analysis with Tukey's multiple comparison test found no significant differences between groups.Numeric values are provided in Excel Table S5.

Figure S3. Effect of phenanthrene on the inactivation kinetics of outward potassium currents in isolated murine ventricular myocytes.
Differences in time constants of inactivation of Ito (A, n=6-8 cells, N=5 animals) and IKur (B, n=6-8 cells, N=6 animals) under control conditions and after exposure to different concentrations of phenanthrene (Phe).*Significant from control, ***=p<0.001,RM ANOVA.Numeric values are as follows: for Ito tau in ms was 16.4±1.3,13.9±1.8,13.7±1.5,9.0±1.1, and 6.99±0.82;and for IKur tau in ms was 79.8±5.4,58.4 ±5.9, 42.8±4.3,14.4±2.9, and 37.2±2.9,under control and with 1, 3, 10, and 30 µM Phe, respectively.2) where 0 concentration is set to 0 block and maximal inhibition by any concentration is set to 100%.I Kr was calculated from the hERG1a/1b data from 1 and used in the place of I Kur to model impact of phenanthrene (Phe) on human AP (see Figure 7).  2 showing the impact of phenanthrene (Phe) exposure on the human ventricular action potential.Note the significant prolongation of APD predicted at levels found in human serum (i.e.<3 µM 3 ).Output AP waveforms are provided in Figure 7 showing the impact of phenanthrene (Phe) exposure on the human ventricular action potential simulated at 1Hz.Note that APD 90 was prolonged at all concentrations tested.

Figure S2 .
Figure S2.Monophasic action potential duration at 50% repolarization, corrected for peakpeak interval.MAP duration at 50% repolarisation, corrected for peak-peak interval (cMAPD50) at baseline and after exposure to either control solution (A) or ~8 µM phenanthrene (Phe) (B) N=6 hearts.Each point represents an individual heart, with lines showing the change after exposure.Paired Student's T-tests found no significant differences between groups (ns).Numeric values are provided in Excel Tabel S6.

Figure S1 :
Figure S1: (A)A representative surface ECG from an ex vivo mouse heart showing the wave forms and intervals measured.Note the absence of a clear J-wave showing early repolarization.These were often undetectable due to interference with the QRS which is common for ex vivo mouse heart preparations4 and thus not quantified in this study.Lower panel shows the impact of acute Phe exposure on uncorrected QT interval (B), QRS duration (C) and P-wave duration (D).Average parameters calculated from isolated mouse (C57/BJ, Male, 10-weeks) hearts at baseline and after 15-minute exposure to either control solution (circle, black), 2.1µM phenanthrene (Phe) (square, red) or ~8 µM Phe (triangle, dark-red) and finally following a 15-minute wash-out period.Bars are mean+/-SEM, symbols are individual hearts (N=5-7 hearts).Two-way ANOVA mixed-effects analysis with Tukey's multiple comparison test found no significant differences between groups.Numeric values are provided in Excel TableS5.

Figure S2 :
Figure S2: Monophasic action potential (MAP) duration at 50% repolarization, corrected for peak-peak interval.MAP duration at 50% repolarisation, corrected for peak-peak interval (cMAPD50) at baseline and after exposure to either control solution (i) or ~8uM Phe (ii) for N=6 hearts.Each point represents an individual heart, with lines showing the change after exposure.Paired T-test showed no significant differences between groups (ns).

Figure S2 :
Figure S2: Monophasic action potential duration at 50% repolarization, corrected for peak-peak interval.MAP duration at 50% repolarisation, corrected for peak-peak interval (cMAPD50) at baseline and after exposure to either control solution (A) or ~8uM phenanthrene (Phe) (B) N=6 hearts.Each point represents an individual heart, with lines showing the change after exposure.Paired Student's T-tests found no significant differences between groups (ns).Numeric values are provided in Excel Tabel S6.

SUPPLEMENTAL DATA FOR Yaar et al. 2023Table S1 .
Estimated IC 50 , n(H) and maximal blockEstimations of IC 50 , n(H) and maximal block based on fractional block calculated from murine I to , I Ca and I Na densities (Table

Table S2 .
Model output of effect of Phenathrene exposure on the human ventricular action potential