The animal experiments performed in this investigation conformed to the Guide for Care and Use of Laboratory Animals of Hubei Province, China, and the study protocol was approved by Experimental Animal Ethics Committee of Wuhan University of Science and Technology. Hearts from adult New Zealand white rabbits (1.5–2 kg) of either sex were quickly removed and retrogradely perfused by the Langendorff method, as described previously (Wu, 2005), with Ca²⁺-free Tyrode solution containing the following compounds (in mM): 135 NaCl, 5.4 KCl, 1.0 MgCl₂, 10 glucose, 0.33 NaH₂PO₄, and 10 HEPES, pH 7.4 with NaOH for 5 min. Then, hearts were perfused with Ca²⁺-free Tyrode solution containing collagenase type I (1 g/l) and bovine serum albumin (BSA, 1 g/l) for 30–40 min before being perfused with KB solution for another 5 min. After perfusion, the left ventricle and left atrium were isolated and gently agitated in KB solution. The cardiomyocytes were filtered through a nylon mesh and stored in KB solution containing the following compounds (in mM): 70 KOH, 40 KCl, 20 KH₂PO4, 50 glutamic acid, 20 taurine, 0.5 EGTA, 10 glucose, 10 HEPES, and 3.0 MgSO4, pH 7.4 with KOH. All solutions used in this study were saturated with 95% O₂ and 5% CO₂ and were maintained at 37°C.

For AP recording, quiescent and Ca²⁺-tolerant cardiomyocytes were bathed in standard Tyrode solution. The patch pipette solution contained the following reagents (in mM): 110 K-aspartate, 30 KCl, 5 NaCl, 10 HEPES, 0.1 EGTA, 5 MgATP, 5.0 creatine phosphate, and 0.05 CAMP, pH 7.2 with KOH. When filled with pipette solution, the electrode resistance was in the range of 1.5–2.5 MΩ. APs were induced in current-clamp mode by 1.5-fold diastolic threshold current pulses of 5 ms in duration at different pacing cycle lengths (CLs).

Currents were recorded with a patch-clamp amplifier (EPC9, Heka electronic, Lambrecht, Pfalz, Germany) and were filtered at 2 kHz and digitized at 10 kHz.

The bath solution used for I_NaT recording contained the following compounds (in mM): 30 NaCl, 1.0 CaCl₂, 105 CsCl, 1.0 MgCl₂, 0.05 CdCl, 5.0 HEPES, and 5.0 glucose, pH 7.4 with CsOH, and 1 μM nicardipine was added to the bath solution to block I_CaL. The pipette solution contained the following compounds (in mM): 120 CsCl, 1.0 CaCl₂, 5.0 MgCl₂, 5.0 Na₂ATP, 10 TEA-Cl, 11 EGTA, and 10 HEPES, pH 7.3 with CsOH. I_NaT was determined by 300-ms depolarization pulses from −70 mV to +40 mV in 5-mV increments—using a holding potential (HP) of −90 mV—at 0.5 Hz. For the steady-state inactivation protocols, currents were recorded using 100-ms conditional prepulses from −100 mV to −50 mV in 5 mV increments—using a HP of −90 mV—followed by a 100-ms test pulse at −20 mV and 0.5 Hz.

The bath solution used for I_NaL recording contained the following compounds (in mM): 135 NaCl, 5.4 CsCl, 1.0 MgCl₂, 10 glucose, 0.33 NaH₂PO₄, 0.3 BaCl₂, 10 HEPES, and 1.8 CaCl₂, pH 7.4 with NaOH, and 1 μM nicardipine was added to the bath solution to block I_CaL. The pipette solution used for this experiment was the same as that used for I_NaT recording I_NaL was recorded using a 300-ms depolarization pulse at a HP of −90 mV, followed by pulses with potentials that were increased from −80 mV to +60 mV in 10-mV increments, and was measured at 200 ms in depolarization testing pulse.

The bath solution (except nicardipine) used for I_CaL recording was the same as that used for I_NaL recording. The electrode was filled with an internal solution containing the following compounds (in mM): 80 CsCl, 60 CsOH, 40 aspartate acid, 0.65 CaCl₂,5.0 HEPES, 10 EGTA, 5.0 MgATP, and 5.0 Na₂-creatine phosphate, pH 7.2 with CsOH. I_CaL was determined using 300-ms voltage steps with potentials that were increased from −40 mV to +50 mV in 5-mV increments at 0.5 Hz. For the steady-state inactivation protocol, I_CaL was determined using 2,000-ms conditional prepulses with potentials that were increased from −50 mV to 0 mV in 5-mV increments—using a HP of −40 mV—followed by a 300-ms test pulse at 0 mV.

For I_K1 recording, the cells were bathed with Tyrode solution, and 1 μM nicardipine was used to block I_CaL. The internal solution contained the following compounds (in mM): 140 KCl, 1.0 MgCl₂, 5.0 K₂ATP, 10 EGTA, and 5.0 HEPES, pH 7.3 with KOH.

The external solution used to record I_Kr contained the following compounds (in mM): 135 NaCl5.4 KCl, 1.0 MgCl₂, 5.0 glucose, 0.2 CdCl₂, 0.33 NaH₂PO₄, 5.0 HEPES, and 1.0 CaCl₂, pH 7.4 with NaOH, and 30 μM chromanol 293B was used to block I_Ks. The pipette solution contained the following compounds (in mM): 140 KCl, 1.0 MgCl₂, 2.0 Na₂ATP, 10 EGTA, and 5.0 HEPES, pH 7.25 with KOH.

The bath solution used to elicit I_to contained the following compounds (in mM): 140 NaCl, 5.4 KCl, 1.0 MgCl₂, 10 glucose, 0.33 NaH₂PO₄, 5 HEPES, and 1.8 CaCl₂, pH 7.4 with NaOH. The internal solution contained the following compounds (in mM): 110 K-aspartate, 20 KCl, 0.1 GTP, 1.0 MgCl₂, 10 HEPES, 5.0 EGTA, 5.0 MgATP, and 5.0 creatine phosphate, pH 7.2 with KOH. BaCl₂(200 μM), CdCl₂ (200 μM), and atropine (1 μM) were used to block I_K1, I_CaL, and I_KAch, respectively.

The bath solution and pipette solution used to record I_Kur were the same as those used to record I_to, but the pulse protocol was different from that used to record I_to (see the Results Section).

Twenty healthy New Zealand rabbits were randomly divided into two groups (n = 10 for each group): normal saline (NS) and icariin. In the NS group, saline was injected intraperitoneally within half an hour before the experiment. In the icariin group, 3 mg/kg icariin was injected intraperitoneally within half an hour before the experiment. At the beginning of the experiments, both groups of rabbits were anesthetized with xylazine (7.5 mg/kg, i.m.) and ketamine (30 mg/kg, i.v.) through ear vein injection. A standard limb lead II electrocardiogram (ECG) was recorded using the BL-420F data acquisition and analysis system (Chengdu TaiMeng, Sichuan, China) for 120 min following the application of 2 μg/kg/min aconitine, which was injected by a constant velocity pump and used to induce arrhythmias. The onset time and onset dosage of aconitine that induced ventricular premature contraction (VPC), ventricular tachycardia (VT) and ventricular fibrillation (VF) were measured.

Icariin (purity >97%) was obtained from Sigma Aldrich (Saint Louis, MO, USA). Collagenase type I and CsCl were purchased from Gibco (GIBCO TM, Invitrogen Co., Paisley, UK). BSA and HEPES were obtained from Roche (Basel, Switzerland), and the other chemicals were obtained from Sigma Aldrich (Saint Louis, MO, USA). Dimethyl sulfoxide (DMSO) was used to dissolve icariin to obtain a 1 mM stock solution. The final concentration of the DMSO added to the bath solution was less than 0.1%.

Fitmaster (v2x32, HEKA) was used for data analysis, and the figures were plotted by Origin 8.0 (OriginLab Co., MA, USA). All data were expressed as the mean ± SD. Data pertaining to the I_NaT and I_CaL steady-state activation and steady-state inactivation relationships were fitted by the Boltzmann equation, Y = 1/{1+exp (V_m−V_1/2)/k}, where V_m is the membrane potential, V_1/2 is the half-activation and half-inactivation potential, k is the slope factor, and Y is relative conductance (G/G_max, steady-state activation) and relative current (I/I_max, steady-state inactivation). The dose-response relationship curves for the effects of icariin on I_NaT and I_CaL were fitted to the Hill equation, (I_control−I_drug)/I_drug = E_max/[1 + (IC₅₀/C)ⁿ], where I_control and I_drug represent the amplitude of I_NaT and I_CaL obtained in the absence and presence of icariin, respectively, E_max is the maximum inhibition, IC₅₀ is the concentration of icariin at which its half-maximum inhibitory effects are exerted, C is the concentration of icariin, and n is the Hill coefficient. Current density was calculated by dividing the current amplitude by the cell capacitance. The statistical significance of the differences between two groups was determined by Student's t-test, and mean comparisons among multiple groups were performed by one-way analysis of variance (ANOVA) followed by Bonferroni's test. P < 0.05 was considered significant.

APs were consecutively recorded by 5-ms and 1.5-fold threshold current pulses at 1 Hz in the absence and presence of icariin. Icariin attenuated AP amplitude (APA) and the maximum upstroke velocity (V_max), shortened action potential durations (APDs) at 50 and 90% repolarization (APD₅₀ and APD₉₀, respectively) in a concentration-dependent manner in LVMs and LAMs. However, icariin had no significant effects on resting membrane potential (RMP) at concentrations of 5 and 10 μM (Figure 1B; Table 1).

In our study, 5 and 10 μM icariin attenuated the rate-dependence (RD) of the APDs (n = 9 cells/4 rabbits; Figures 1C,D) in LVMs by 10.5 ± 4.3% and 28.5 ± 7.2% at a pacing cycle length (CL) of 500 ms, by 13 ± 4.9% and 32.2 ± 7.4% at a pacing CL of 1,000 ms and by 16.5 ± 4.8% and 34.5 ± 6.4% at a pacing CL of 2,000 ms, respectively.

In the present study, we used 10 nM anemonia toxin II (ATX-II) and a stimulation frequency of 0.25 Hz to elicit early afterdepolarizations (EADs) in LVMs. ATX-II significantly lengthened the APD from 179.78 ± 18.64 ms to 1186.44 ± 93.13 ms and induced EADs in 7 of 10 cells (70%; n = 10 cells/5 rabbits; Figures 2A–C), and 20 μM icariin decreased the APs prolonged by ATX-II from 1186.44 ± 93.13 ms to 360.08 ± 41.95 ms and completely abolished the EADs induced by ATX-II in seven cells. In another group, to elicit delayed afterdepolarizations (DADs) and triggered activities (TAs) in LVMs, we added 1 μM isoproterenol (ISO) to the external solution and the extracellular calcium concentration was elevated to 3.6 mM following a baseline pacing CL of 9,000 ms and on top of that 15 beats with a stimulation frequency of 2.5 Hz. DADs were noted in 6 of 9 cells (3 rabbits; 66.7%), and TAs were noted in 3 of 9 cells (33.3%). Administration of 10 μM icariin significantly suppressed the ISO-induced DADs and completely abolished the ISO-induced TAs (Figure 2D).

When the effects of icariin on I_NaT reached a steady state (3 min), the next concentration of the drug could be added to the external recording solution. Icariin (1, 5, 10, and 20 μM) reduced I_NaT in a dose-dependent manner in LVMs and LAMs. Figures 3A,B show the representative recordings for I_NaT in LVMs and LAMs, respectively, and Figure 3C shows the corresponding current-voltage relationships in LVMs and LAMs. The IC₅₀ values for I_NaT in LVMs and LAMs were 11.83 ± 0.92 μM (n = 10 cells/6 rabbits) and 12.28 ± 0.29 μM (n = 8 cells/4 rabbits; p > 0.05 LAMs vs. LVMs; Figure 3D), respectively. Figures 3E,G show typical current recordings, which were generated according to the steady-state inactivation protocol, in LVMs and LAMs. In the absence and presence of 20 μM icariin, the V_1/2 values of the steady-state inactivation curves in LVMs were −85.47 ± 1.36 mV and −91.45 ± 1.48 mV (n = 8 cells/5 rabbits; p < 0.01 vs. control), respectively, with corresponding k-values of 8.48 ± 1.05 and 8.28 ± 0.76 (n = 8 cells/5 rabbits; p > 0.05 vs. control). Administration of 20 μM icariin shifted the V_1/2 value of the steady-state inactivation curve in LAMs from −76.1 ± 1.52 mV to −82.28 ± 0.96 mV (n = 6 cells/3 rabbits; p < 0.01 vs. control), with k-values of 8.29 ± 1.64 and 8.72 ± 0.81 (n = 6 cells/3 rabbits; p > 0.05 vs. control). These results indicate that icariin induced a leftward (negative potential) shift of the steady-state inactivation curve of I_NaT in LVMs and LAMs (Figures 3F,H). However, it had no significant effects on the activation process in LVMs and LAMs (Figures 3F,H).

To identify I_NaL, we recorded current before and after the application of 4 μM TTX using 300-ms depolarization pulses with potentials ranging from a HP of −90 mV to a potential of −20 mV. TTX (4 μM) had no significant effects on I_NaT but decreased the amplitude of I_NaL from −0.39 ± 0.004 pA/pF to 0.023 pA/pF (n = 6 cells/3 rabbits; p < 0.01 vs. control), indicating that the TTX-sensitive current was I_NaL. Administration of 10 nM ATX-II significantly enhanced I_NaL, an effect that was reversed by administration of 1, 10, 20, and 40 μM icariin (n = 6 cells/4 rabbits; Figures 4A,C). The percentage inhibitions by 1, 10, 20, and 40 μM icariin of ATX-II augmented I_NaL were 7.8 ± 1%, 29 ± 6.4%, 43.68 ± 5.6%, and 61.4 ± 5.7%. Figure 4B shows the representative current recordings of I_NaL at −20 mV that are shown in Figure 4A.

To elicit I_CaL, we clamped LVMs at −40 mV and then depolarized the cells to +5 mV for 300 ms at 0.2 Hz. As shown in Figure 5A, the I_CaL run-down phenomenon lasted for approximately 5 min after membrane rupture in the control condition and then reached a steady state for 15 min (n = 5 cells/2 rabbits). I_CaL decreased by 8.5% during the this 5-min period. We performed a series of experiments on I_CaL during the stabilization period. To investigate the efficiency of the effects of icariin on I_CaL in LVMs, we recorded the current sequentially. As shown in Figure 5B, 10 μM icariin was added to the bath solution after the first (1st) current curve (control). I_CaL decreased rapidly between the tenth (10th) current curve (45 s after perfusion with icariin) and the thirteenth (13th) current curve (60 s after perfusion with icariin) and then decreased gradually until it reached a steady state (the twenty-seventh current curve). Icariin was washed out after 27th current curve (130 s after perfusion with icariin). I_CaL increased rapidly between the 27th current curve and the thirtieth (30th) current curve and then increased gradually until it reached its maximum value (82%) at the fifty-fifth current curve (270 s after perfusion with icariin). The summary data are shown in Figure 5C (n = 10 cells/4 rabbits). The above results indicate that icariin rapidly and reversibly inhibited I_CaL in LVMs.

When the effects of icariin on I_CaL reached a steady state (2.5 min), the next concentration of the drug could be added to the bath solution. Figure 5D shows the representative I_CaL recordings in LVMs after sequential treatments of 0.1, 1, 5, 10 μM icariin and 1 μM nicardipine. Icariin decreased I_CaL in a concentration-dependent manner in LVMs, with an IC₅₀ of 4.78 ± 0.89 μM (n = 8 cells/4 rabbits; Figure 5G). Nicardipine (1 μM) almost completely inhibited I_CaL in LVMs in the presence of 10 μM icariin, indicating that I_CaL was the nicardipine-sensitive current. Figure 5E shows the representative I_CaL recordings in LAMs after sequential treatments of 1, 5, 10, and 20 μM icariin. Icariin reduced I_CaL in a dose-dependent manner in LAMs, with an IC₅₀ of 13.43 ± 2.73 μM (n = 9 cells/5 rabbits; p < 0.01 vs. LVMs; Figure 5G). Figure 5F shows the I_CaL current-voltage relationships in LVMs (left, n = 13 cells/6 rabbits) and LAMs (right, n = 9 cells/5 rabbits). Icariin shifted the I_CaL steady-state inactivation curves to the left in LVMs and LAMs (Figures 5I,K). The V_1/2 values before and after 10 μM icariin administration were shifted from −25.7 ± 1.01 mV and −29.96 ± 0.85 mV to −28.87 ± 2.18 mV (n = 10 cells/4 rabbits; p < 0.01 vs. control; Figures 5H,I) and −33.94 ± 1.33 mV (n = 10 cells/6 rabbits; p < 0.01 vs. control; Figures 5J,K), and the k-values were shifted from 7.16 ± 1.08 and 5.86 ± 0.82 to 7.52 ± 2.31 (p > 0.05 vs. control) and 7.4 ± 1.14 (p < 0.01 vs. control) in LVMs and LAMs, respectively. However, the drug has no significant effects on the activation process in these cells (Figures 5I,K).

To elicit I_K1 in LVMs, we clamped the cells at −40 mV (to inactivate their sodium channels) and depolarized them from −120 mV to +50 mV in 5-mV increments for 400 ms at 0.5 Hz. As shown in Figures 6A,B, icariin (10 and 40 μM) had no effect on I_K1 (n = 18 cells/8 rabbits). I_Kr in LVMs was elicited using a 3-s depolarization pulse whose potential was increased from a HP of −40 mV to a potential of 50 mV in 10-mV increments before returning to a potential of −40 mV for 5 s. Only the I_Kr tail-current (I_Kr−tail) was measured. Icariin (10 and 40 μM) had no significant effects on I_Kr−tail (n = 12 cells/5 rabbits; Figures 6C,D). I_to in LAMs was elicited by 400 depolarization voltage steps with potentials that were increased from −80 mV to +50 mV in 10-mV increments, followed by a conditional test in which −40 mV was administered for 100 ms to block sodium currents. Forty micrometer icariin had no significant effect on I_toin LAMs (n = 15 cells/6 rabbits; Figures 6E,F). I_Kur in LAMs was elicited by an 80-ms prepulse whose potential was increased from a HP of −50 mV to a potential of 30 mV (to inactivate I_to), followed by 140-ms test pulses with potentials that were increased from −40 mV to +60 mV in 10-mV increments—using a HP of −50 mV—after a 50-ms interval before returning to −30 mV. Figure 6G shows the I_Kur current-voltage relationship in LAMs in the absence and presence of icariin (20 and 40 μM). Icariin had no significant effect on I_Kur (n = 1 5 cells/5 rabbits).

In the NS group, VPC, VT, and VF were observed in all 10 rabbits. In the icariin group, VPC, VT and VF occurred in 9, 4 and 1 of 10 rabbits, respectively. Compared with the NS group, icariin application prior to aconitine administration increased the onset time (Figures 7A,B,D) and onset dosage (Figure 7C). The administration of icariin attenuated the incidence of arrhythmias induced by aconitine (Figure 7E) and rabbit mortality (Figure 7F).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.