Empagliflozin rescues pro-arrhythmic and Ca2+ homeostatic effects of transverse aortic constriction in intact murine hearts

Abstract

We explored physiological effects of the sodium-glucose co-transporter-2 inhibitor empagliflozin on intact experimentally hypertrophic murine hearts following transverse aortic constriction (TAC). Postoperative drug (2-6 weeks) challenge resulted in reduced late Na⁺ currents, and increased phosphorylated (p-)CaMK-II and Nav1.5 but not total (t)-CaMK-II, and Na⁺/Ca²⁺ exchanger expression, confirming previous cardiomyocyte-level reports. It rescued TAC-induced reductions in echocardiographic ejection fraction and fractional shortening, and diastolic anterior and posterior wall thickening. Dual voltage- and Ca²⁺-optical mapping of Langendorff-perfused hearts demonstrated that empagliflozin rescued TAC-induced increases in action potential durations at 80% recovery (APD₈₀), Ca²⁺ transient peak signals and durations at 80% recovery (CaTD₈₀), times to peak Ca²⁺ (TTP₁₀₀) and Ca²⁺ decay constants (Decay_30-90) during regular 10-Hz stimulation, and Ca²⁺ transient alternans with shortening cycle length. Isoproterenol shortened APD₈₀ in sham-operated and TAC-only hearts, shortening CaTD₈₀ and Decay_30-90 but sparing TTP₁₀₀ and Ca²⁺ transient alternans in all groups. All groups showed similar APD₈₀, and TAC-only hearts showed greater CaTD₈₀, heterogeneities following isoproterenol challenge. Empagliflozin abolished or reduced ventricular tachycardia and premature ventricular contractions and associated re-entrant conduction patterns, in isoproterenol-challenged TAC-operated hearts following successive burst pacing episodes. Empagliflozin thus rescues TAC-induced ventricular hypertrophy and systolic functional, Ca²⁺ homeostatic, and pro-arrhythmogenic changes in intact hearts.

Keywords: Action potential duration; Ca2+ homeostasis; Empagliflozin; Heart failure; Transverse aortic constriction (TAC); Ventricular arrhythmia.

Figure 1

Empagliflozin decreases ATX-II induced I_NaL while sparing Nav1.5 gating in Nav1.5 expressing CHO cells. (a) Na⁺ currents under control conditions and following addition of 10 nM ATX-II and 10 nM ATX-II + 10 μM empagliflozin. (b) Normalized inward Na⁺ current measured at 60 ms and 100 ms (designated Amp-60 ms/100 ms) and from the area under the curve between 0 and 100 ms poststimulation (designated Area-100 ms). N = 11 for each group. One way ANOVA with Holm-Sidak’s tests for post hoc multiple comparisons, ***P < 0.001. (c) Steady-state Nav1.5 activation data (right curves) obtained using 200 ms duration test steps made to test potentials between − 120 and + 60 mV at 10 mV increments. Steady-state inactivation data (left curves) obtained by a double-pulse protocol imposing a series of 200 ms duration conditioning depolarizing pulses to voltages between − 120 and + 60 mV with 10 mV increments followed by a 300 ms duration test pulse to a fixed voltage of − 30 mV. ATX-II, anemonia sulcata toxin II; Empa, empagliflozin; I_NaL, late sodium current.

Figure 2

Effect of empagliflozin on CaMK-II phosphorylation and Nav1.5 and NCX expressions. Western blot determinations of protein expression in sham-operated, TAC-only and TAC + empagliflozin (Empa) groups. (a) Images of total CaMK-II, phosphorylated CaMK-II, Nav1.5 and NCX. (b–d) Semi-quantitative analysis of expression results for t/p-CaMK-II (b), Nav1.5 (c) and NCX (d) with one way ANOVA with Holm-Sidak’s tests for post hoc multiple comparisons. **P < 0.01, ***P < 0.001. t/p-CaMK-II, total/phosphorylated calmodulin-dependent protein kinase II; Nav1.5, sodium channel 1.5; NCX, Na⁺/Ca²⁺ exchanger; GAPDH, glyceraldehyde-3-phosphate dehydrogenase. Original Western blot results shown in supplementary file 2.

Figure 3

Empagliflozin rescues left ventricular remodeling in TAC-HF mice. (a,b) Representative heart images (mm scale bar) (a) and typical images from M type echocardiography (b) from animals in the sham-operated, TAC-only and TAC + Empagliflozin (Empa) groups after 4 weeks of treatment. (c) blood flow velocity (mm/sec) determinations across the aortic constriction.

Figure 4

Quantifications of empagliflozin rescue of left ventricular remodeling in TAC-HF mice. (a) Heart weight/tibial length (HW/TL) and left ventricular internal diameter ratios. (b,c) Indices for (b) cardiac function: ejection fraction (EF) (i) and fractional shortening (FS) (ii) results; (c) structural remodeling: left ventricular diastolic and systolic anterior (i) and posterior (ii) wall thicknesses. *P < 0.05, **P < 0.01, ***P < 0.001, one way ANOVA with Holm-Sidak’s tests for post hoc multiple comparisons. TAC, transverse aortic constriction; Empa, empagliflozin; HW/TL, heart weight/tibia length; LVID, left ventricular internal diameter; EF, ejection fraction; FS, fractional shortening; LVAW, left ventricular anterior wall; LVPW, left ventricular posterior wall; d, diastolic; s, systolic.

Figure 5

Mapping of electrophysiological and Ca²⁺ homeostatic activity in sham-operated, TAC-only and TAC + empagliflozin hearts. (a) Schematic diagram summarizing configuration of dual optical mapping recording system including mapping areas. In these experiments hearts were studied at a regular stimulation frequency of 10 Hz. (b–e) Action potential (AP) studies: (b) Action potential (AP) traces: AP trace illustrating determination of APD₈₀ (left) and typical AP traces under baseline conditions (middle) and following 1 μM isoproterenol (ISO) challenge (right) from sham-operated, TAC-only and TAC + empagliflozin groups (colour-coded). (c) Typical APD₈₀ maps for sham-operated, TAC-only and TAC + empagliflozin groups obtained under baseline conditions and following 1 μM isoproterenol challenge. (d,e) Summary of APD₈₀ (d), and APD₈₀ heterogeneity results (e). (f–j) Studies of Ca²⁺ signals: (f,g) typical peak CaT maps from sham-operated, TAC-only and TAC + empagliflozin groups before (f) and following isoproterenol challenge (g). (h,i) corresponding Ca²⁺ signal (CaT) traces. (j) peak F/F₀ values in sham-operated, TAC-only and TAC + empagliflozin groups before (left) and following 1 μM isoproterenol challenge (right). N = 5 for sham-operated and TAC-only groups and N = 4 for TAC + empagliflozin group. *P < 0.05, **P < 0.01, ***P < 0.001, two way ANOVA with Holm-Sidak’s tests.

Figure 6

Kinetic features of Ca²⁺ transients in sham-operated, TAC-only and TAC + empagliflozin hearts. (a–d) Ca²⁺ signal (CaT) recovery properties: (a) CaT traces illustrating determination of CaT duration at 80% recovery CaTD_80, and typical baseline CaT records (middle panel) and CaT records obtained following 1 μM isoproterenol (ISO) challenge (right) from sham-operated, TAC-only and TAC + empagliflozin groups. (b) typical CaTD₈₀ maps from sham-operated, TAC-only and TAC + empagliflozin groups. (c,d) Summarized CaTD₈₀ (c) and heterogeneity results (d). N = 5 for sham-operated and TAC-only groups and N = 4 for TAC + Empagliflozin group. Scale bar = 20 ms. (e–g) Further analysis of Ca²⁺ signals into early release and late recovery phases quantified by TTP₁₀₀ and Decay_30–90, respectively. (e) illustrates determination of TTP₁₀₀ and Decay_30–90 from a typical CaT waveform. (f,g) TTP₁₀₀ (f) and Decay_30–90 results (g) for sham-operated, TAC-only and TAC + empagliflozin groups under regular 10 Hz pacing before and following 1 μM isoproterenol (ISO) challenge. N = 5 for sham-operated and TAC-only groups and N = 4 for TAC + empagliflozin group. *P < 0.05, **P < 0.01, ***P < 0.001, two way ANOVA with Holm-Sidak’s tests for post hoc multiple comparisons. TAC, transverse aortic constriction; APD, action potential duration; CaTD, calcium transient duration; Empa, empagliflozin; ISO, isoproterenol; TTP, time to peak.

Figure 7

Stimulus protocols to assess ventricular arrhythmogenicity. Inset: overall pulsing scheme. (a) Successive episodes of 30 identical and consecutive S1 stimuli were applied with PCL decreasing from 100 to 50 ms with a 10 ms decrement. If no VT occurred during this procedure then (b) a burst pacing protocol consisting of 50 identical and consecutive stimuli separated by a 20 ms interval was applied. Otherwise, isoproterenol (ISO) was applied followed by similar stimulus protocols (c,d). ISO, isoproterenol; VT, ventricular tachycardia; PCL, pacing cycle length. Scale bar = 0.1 s.

Figure 8

Empagliflozin blunts Ca²⁺ signal alternans induced by TAC. Analysis of CaT alternans studied in Langendorff perfused hearts analysis subject to the S1S1 protocol examining sequences containing more than 10 consecutive alternans. (a) Representative CaT alternans maps and traces obtained at PCLs of 70 ms and 100 ms from the sham-operated, TAC-only and TAC + empagliflozin groups before and following isoproterenol (ISO) challenge. (b) Plots of large and small CaT alternans and definition of CaT alternans ratio. (c) Quantitative plots of CaT alternans ratio at different PCLs for the sham-operated, TAC-only and TAC + empagliflozin groups before (i) and following (ii) isoproterenol challenge. For each group, N = 5 and *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 v.s sham-operated group, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ^####P < 0.0001 v.s TAC-only group, repeated measure ANOVA with Holm-Sidak’s tests for CaT alternans ratio amongst groups at different PCLs. TAC, transverse aortic constriction; Empa, empagliflozin; ISO, isoproterenol; CaT, calcium transient; PCL, pacing length; BL, baseline.

Figure 9

Empagliflozin rescues ventricular arrhythmogenicity following TAC. When intact Langendorff-perfused hearts were subject to S1S1 stimuli delivered at 50 Hz, no VT episodes were induced (data not shown). (a) After recovery, hearts were next challenged with 1 μM isoproterenol (ISO) and again subject to 50 Hz stimulation. Typical ECGs from sham-operated, TAC-only and TAC + empagliflozin (Empa) groups. Scale bar = 0.1s. PVC, premature ventricular contraction. (b) Incidences of VT in the three experimental groups giving total numbers of hearts studied, and the numbers of hearts showing VT in each group. (c) Scoring of occurrences of ventricular arrhythmogenicity in sham-operated, TAC-only and TAC + empagliflozin groups.

Figure 10

Conduction patterns during a ventricular tachycardic (VT) episode in a heart following TAC induced HF. (a–c) Simultaneous ECG (a), membrane potential, V_m (b) and Ca²⁺ recordings (c) close to the onset of VT obtained for phase mapping analysis for conduction made over the region of interest shown in the inset (d) giving representative maps of macro-reentry at different time points in the TAC-only group (e). No VT episodes were observed in sham-operated (f) and TAC + empagliflozin (g) groups which gave rectilinear conduction patterns running from the stimulus point at the apex to the base. TAC, transverse aortic constriction; Empa, empagliflozin; V_m, membrane potential.