Antiarrhythmic potential of the dual SGLT1/2 inhibitor sotagliflozin – may SGLT1 be the key?

Introduction
Sodium-glucose co-transporter inhibitors (SGLTi) have been shown to be cardioprotective in large clinical trials in both heart failure with reduced and preserved ejection fraction. Treatment with the sodium-glucose co-transporter-2 inhibitor (SGLT2i) dapagliflozin was also shown to reduce the incidence of atrial arrhythmias in a subanalysis of a clinical trial, although the antiarrhythmic mechanisms underlying this observation are not yet fully understood. In contrast to SGLT2, which has virtually no cardiac expression, SGLT1 mRNA and protein have been detected in myocardial tissue - yet its functional role remains unclear. The SGLTi sotagliflozin has a potent inhibitory effect on SGLT1, making it an interesting candidate for investigating the antiarrhythmic mechanisms of SGLTi. The aim of this study is to advance our understanding of the antiarrhythmic potential of SGLTi by characterising the effects of sotagliflozin on cardiac ion channels and analysing the RNA expression of atrial SLC5A1 (the gene encoding SGLT1).

Methods
Cardiac ion channels were heterologously expressed in Xenopus laevis-oocytes. Two-electrode voltage clamp (TEVC) technique was used to identify the effects of sotagliflozin on these channels. The interaction of sotagliflozin with Na_V1.5 was further characterized using different stimulation frequencies, alanine-mutagenesis, in silico docking simulation, and a comparison with the effects of ajmaline and lidocaine. mRNA levels of SLC5A1 were analyzed in human atrial tissue samples using qPCR.

Results
The effect of sotagliflozin on a group of 10 cardiac ion channels was investigated using TEVC. Administration of sotagliflozin at 100 µmol/L lead to a Na_V1.5 inhibition of 22.3 ± 2.1 % (P < 0.0001; n = 9). This inhibition was comparable with the effects of class I antiarrhythmic lidocaine on Na_V1.5 (lidocaine 100 µmol/L ; 31.0 ± 2.4 %; P < 0.0001; n = 7), whereas ajmaline at 100 µmol/L showed a significantly stronger inhibitory effect (65.3 ± 2.5 %; P = 0.0015; n = 3). A reduced affinity of sotagliflozin to the alanine pore mutant Na_V1.5-F1760 suggests a direct binding mechanism of sotagliflozin in close proximity to the Na_V1.5 channel pore. This finding is consistent with our in silico docking simulations suggesting a possible binding site at this amino acid based on a published cryo-EM structural model of Na_V1.5. Our analysis further showed a frequency-dependent inhibition of heterologously expressed Na_V1.5 channels by sotagliflozin. To investigate whether cardiac SGLT1 might play a role in AF, expression analyses were performed in atrial tissue samples obtained from patients with either sinus rhythm (SR), paroxysmal/persistent AF (pAF) or chronic AF (cAF). Here, our qPCR analysis showed a significant downregulation of SLC5A1 mRNA levels in atrial tissue samples from cAF patients compared to SR controls (49.1 %; P = 0.0012; n = 12).

Conclusion
The results of our study suggest a direct Na_V1.5 channel inhibitory effect of sotagliflozin, which appears to be mediated by a direct binding mechanism. Further translational studies are warranted to evaluate the effect of SGLTi such as sotagliflozin on cardiac electrophysiology. A significant downregulation of SLC5A1 mRNA levels in cAF samples suggests a not yet understood role of SGLT1 in the pathophysiology of atrial fibrillation.