Empagliflozin stabilizes electrical activity in failing heart cells

Empagliflozin stabilized the electrical activity of ventricular cardiomyocytes isolated from two animal models of non-ischemic heart failure with reduced ejection fraction, according to a new experimental study. Three hours of pretreatment shortened overlong action potentials, reduced beat-to-beat variability, and lowered susceptibility to two hallmarks of pathological electrical activity: alternation between long and short action-potential durations and delayed afterdepolarizations. The same concentration had no significant effect when applied acutely for only three minutes, suggesting that the drug’s cellular action requires time-dependent processes such as altered signaling or post-translational regulation rather than immediate ion-channel blockade.

The researchers examined heart cells from male mice with severe heart failure induced by transverse aortic constriction and male rabbits with chronic failure produced by combined pressure and volume overload. Healthy controls were included for both species. The mouse model produced an ejection fraction below 25%, while rabbit hearts with left ventricular end-systolic dimension above 1.4 cm and ejection fraction below 35% were classified as failing. Cells were randomly assigned to vehicle, 1 micromolar empagliflozin, or 1 micromolar cariporide, a selective inhibitor of the sodium-hydrogen exchanger NHE-1. Patch-clamp recordings were conducted at 37 degrees Celsius, and electrophysiological outcomes were compared using hierarchical analysis of variance.

Failing cardiomyocytes had prolonged action-potential duration at 90% repolarization, greater short-term variability, and, in the rabbit model, a more depolarized resting membrane potential. Mouse cells also showed a lower maximal action-potential upstroke velocity. Empagliflozin pretreatment attenuated these abnormalities in both models: it reduced action-potential prolongation and variability, improved resting potential in rabbit cells, and increased upstroke velocity in failing mouse cells. Three-hour treatment produced no significant electrophysiological changes in healthy controls, while three-minute exposure had no measurable effect in either healthy or failing cells. The delay between empagliflozin and its observed effects was a central feature of the results.

The drug also shifted the pacing threshold at which action-potential alternans appeared, a beat-to-beat oscillation linked with unstable repolarization. In failing mouse cells, alternans appeared at rates of at least 7 hertz; at 10 hertz it occurred in all 15 untreated failing cells but in only one of 10 cells after empagliflozin pretreatment. In rabbits, the alternans threshold was 3 hertz in failing cells and 5 hertz in controls. Pretreatment increased the threshold in failing rabbit cells and reduced the magnitude of alternans to control levels. These observations were made in isolated cells and therefore do not establish that the treatment prevents arrhythmias in a whole heart or in a patient.

Delayed afterdepolarizations, spontaneous depolarizations that can trigger abnormal beats, were also reduced. In failing mouse cells, empagliflozin lowered their incidence and prevented spontaneous action potentials. In failing rabbit cells, they occurred in 10 of 15 cells without pretreatment, compared with 2 of 12 after pretreatment, a statistically significant difference. Acute empagliflozin exposure did not reduce alternans or afterdepolarizations. Cariporide did not reproduce the broad effects of empagliflozin: it did not improve action-potential duration or alternans in either model and only modestly reduced afterdepolarizations in mouse cells. This comparison does not rule out NHE-1 as a contributor, because the experiments used normoxic conditions and an internal pH of 7.2, where NHE-1 activity is relatively low.

Strong buffering of intracellular calcium with 10 millimolar EGTA reproduced much of the electrical stabilization. It reduced action-potential prolongation and variability, prevented alternans in both failing models, and abolished delayed afterdepolarizations. The results identify disturbed intracellular calcium handling as a major component of the cellular phenotype, although they do not prove that every effect of empagliflozin is calcium-dependent. Calcium and sodium handling are closely coupled in failing hearts: increased late sodium current can promote intracellular sodium accumulation, alter calcium exchange, activate calcium/calmodulin-dependent protein kinase II, and reinforce abnormalities in calcium release. The study therefore supports a mechanistic link between improved sodium and calcium regulation and the observed electrical effects, without establishing a single molecular target.

Direct current measurements identified two ion currents consistent with the action-potential changes. Late sodium current was small in healthy cells but markedly increased in failing mouse and rabbit cardiomyocytes, by 126% and 93%, respectively, relative to healthy controls. Three-hour pretreatment with empagliflozin reversed this increase in both species. The inward rectifier potassium current, which supports resting membrane potential and repolarization, was reduced in failing cells. Empagliflozin increased its density in both models, including the physiologically relevant outward current in mouse cells, where it rose by 46%, compared with a 15% increase in inward current. As with the action-potential results, empagliflozin did not significantly change these currents in healthy control cells.

The time dependence is compatible with several non-acute mechanisms discussed by the authors. Empagliflozin may suppress post-translational changes that enhance late sodium current in heart failure, potentially involving altered CaMKII-dependent phosphorylation. It might also act as a pharmacological chaperone, changing the trafficking or functional membrane expression of cardiac sodium channels. Previous studies have reported both direct inhibition of Nav1.5, generally at concentrations above the 1 micromolar used here, and increased functional expression of peak sodium current after longer exposure. The present experiments did not directly distinguish these mechanisms or test whether changes in late sodium and inward rectifier currents caused the action-potential improvements.

Several boundaries constrain the findings. The study used only male animals and deliberately excluded diabetes and ischemic disease, isolating a glucose-independent effect in chronic non-ischemic failure; whether responses differ by sex or in ischemic heart failure remains unknown. The work also examined isolated cells rather than intact hearts, animals, or people. Although empagliflozin improves clinical outcomes in heart failure, these experiments do not show that the electrophysiological changes observed here explain its clinical benefits or predict reduced arrhythmias in patients. The next steps would include tissue and organ-level studies using whole-heart optical mapping and in-vivo arrhythmia testing, alongside experiments designed to identify the relevant molecular targets and determine whether the effects are specific to empagliflozin or shared by other SGLT2 inhibitors.

Subject of Research: Cellular electrophysiological effects of empagliflozin in non-ischemic HFrEF cardiomyocytes

Article Title: Complex antiarrhythmic effects of empagliflozin in non-ischemic HFrEF cardiomyocytes

Article References: Hegyi, B., & Bers, D. M. (2026). Complex antiarrhythmic effects of empagliflozin in non-ischemic HFrEF cardiomyocytes. Frontiers in Pharmacology, 17, Article 1955044. https://doi.org/10.3389/fphar.2026.1955044

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1955044

Keywords: empagliflozin, heart failure, HFrEF, cardiac arrhythmias, action potential, ion channels, intracellular calcium, late sodium current, IK1, SGLT2 inhibitors, cardiomyocytes, Complex