The Antiarrhythmic Potential of Caffeine

Introduction: Caffeine is consumed regularly by large swathes of the population. The pharmacological effects of caffeine remain controversial to this day, especially with regards to its ability to cause or prevent arrhythmias. To further analyze its arrhythmogenic effects, our aim was to study the direct effects of caffeine on cardiac K⁺-currents which are important for cardiac repolarization and thus have been implicated in the pathogenesis of many arrhythmias, including atrial fibrillation (AF). Special emphasis is placed on the two-pore (K2P-) potassium-channel family, whose function in the heart is incompletely understood at present. 

Goals/Purpose: The aim of this study is to elucidate the acute effects of caffeine on cardiac action potential (AP) formation.

Methods: Cardiac ion-channels were heterologously expressed in Xenopus leavis oocytes. 1-3 days after cRNA injection two-electrode voltage-clamp measurements were taken. Whole cell patch-clamp recordings of isolated human atrial single cardiomyocytes obtained from human atrial tissue samples were used to assess the effects of caffeine on the atrial AP.

Results: Application of 20 mM caffeine caused significant inhibition of the TREK-1 (94.0±1% inhibition; n=4; p<0.0001), TREK-2 (61.9±2.6% inhibition; n=3; p=0.0019), TASK-1 (53.3±2.6% inhibition; n=7; p<0.0001), hERG (80.8±3.2% inhibition; n=5; p<0.0001), K_v2.1 (50.0±2.8% inhibition; n=3; p<0.0001), K_v4.3 (34.6±5.2% inhibition; n=3; p=0.0111) and K_ir2.1 (30.3±5.9% inhibition; n=5; p=0.0102) currents in TEVC experiments. IC₅₀ values were obtained for TASK-1 (14.2 mM; n=3-5) and TREK-1 (621.4 µM, n=3-4). Application of 20mM theophylline, a well-known caffeine metabolite, led to significant inhibition of TASK-1 (69.2±8.3% inhibition; n=4; p<0.0001), TREK-1 (95.0 ± 4% inhibition; n=3; p<0.0001) and hERG (65.9±3.4% inhibition; n=3; p<0.0001) currents with an IC₅₀ value for TREK-1 of 644.6µM (n=3). Patch-clamp recordings of isolated human atrial cardiomyocytes revealed a reduction of potassium current densities (0.698 ΔρA/μF inhibition; p=0.017; 7/4 (n/N)), as well as a prolongation of the atrial action potential duration at 30%, 50% and 90% repolarization (APD₃₀: 100 µM caffeine: +14,16 ms vs. baseline, p=0.0192, APD₅₀: 100 µM caffeine: +24.87 ms vs baseline, p=0.0118, APD₉₀: 100 µM caffeine: +36.45 ms vs baseline, p=0.0077, 9/2 (n/N)).

Conclusion: Caffeine robustly inhibits K⁺-channels. We observed the strongest effects for the TREK1 channel at concentrations which may be reached in plasma by oral uptake of 250 mg of caffeine or more (approx. 3 cups of coffee). Consistent with these findings, in whole cell patch clamp recordings of outward K⁺-currents on isolated human cardiomyocytes we recorded a drop in the current density upon caffeine application, as well as a prolongation of the AP duration. This data suggests that caffeine exerts a class III antiarrhythmic effect on atrial cardiomyocytes. Further studies will be required to rule out confounding factors (e.g. caffeine mediated effects on [Ca²⁺]_i).