Animals used in the present study were treated in accordance with the Guide for the Care and Use of Laboratory Animals regulated by the Administrative Regulation of Laboratory Animals of Hubei Province and all experimental methods were approved by the Animal Research Committee of the First Clinic College of Wuhan University (Wuhan, Hubei, China).

Healthy rabbits of both sexes, weighing 1.5–2.5 kg and ~6–8 weeks old, were anaesthetized with 30 mg/kg sodium pentobarbital intravenously. Following exsanguination, the hearts were rapidly removed and immersed in cold (0–4°C) oxygenated Ca²⁺-free tyrode solution containing 135 mM NaCl, 5.4 mM KCl, 1 mM MgCl₂×6H₂O, 0.33 mM NaH₂PO₄, 10 mM hydroxyethyl piperazineethanesulfonic acid (HEPES) and 10 mM glucose (4°C; pH 7.4). The SA node preparations, bounded by the crista terminalis, the superior and inferior vena cava and the interatrial septum, were carefully dissected out to be pinned in the experimental chamber. The preparations were continuously superfused with modified tyrode solution at a rate of 10 ml/min and a temperature of 37±0.5°C until ~1 h prior to the recordings. The composition of modified tyrode solution was as follows: 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, 0.5 mM MgCl₂×6H₂O, 5 mM HEPES and 10 mM glucose (pH 7.4).

Following 1 h of recovery, the transmembrane potentials were recorded with glass microelectrodes filled with 3 mM KCl (10–15 MΩ) connected to a high input impedance amplifier (Dua 773; World Precision Instruments, Sarasota, FL, USA). The signal was digitalized and collected using specific software (AcqKnowledge 4.1; BIOPAC Systems, Inc., Norfolk, UK). Spontaneous APs from the pacemaker cells were recorded for 20–30 min in control conditions. In this experiment, three concentrations of berberine (0.3, 3 and 30 μM) were added and the spontaneous AP firing rate was measured every 5 min for 1–2 h. In addition, changes in the AP amplitude and duration (APA, APD₅₀ and APD₉₀), the spontaneous firing frequency, the maximal diastolic potential (MDP) and the diastolic depolarization rate (DDR) were determined at the end of the drug exposure period.

Xenopus frogs were anesthetized by cooling on crushed ice for 30–40 min. Ovarian lobes were digested with 1 mg/ml type IA collagenase (Sigma Chemicals, St. Louis, MO, USA) in Ca²⁺-free ND96 solution for 30 min to remove follicle cells. Stage IV and V xenopus oocytes were injected with 30 nl (1 μg/μl) of hHCN4 cRNAs per oocyte using a Nanoject microdispenser (Nanoliter 2000; World Precision Instruments, Sarasota, FL, USA) and then cultured in ND96 solution supplemented with 100 U/ml penicillin, 100 U/ml streptomycin and 2.5 mM pyruvate at 17°C for 2–3 days prior to their use in the voltage clamp experiments. The ND96 solution contained 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 2 mM MgCl₂, 5 mM HEPES, titrated to pH 7.40 with NaOH. The recordings were performed 2–9 days following injection. A standard two-microelectrode voltage-clamp technique was used to record the currents at 21–23°C. The glass microelectrodes were filled with 3 M KCl solution to obtain a resistance of 1–3 MΩ. The oocytes were clamped with a standard two-microelectrode voltage-clamp amplifier (DAGAN CA-1B; Dagan Corporation, Minneapolis, MN, USA) and PLAMP software (Axon Instruments, Foster City, CA, USA). Oocytes injected with hHCN4 cRNA were superfused with ND96 solution at a rate of 2.0 ml/min. The control currents were recorded repeatedly at 1 min intervals, with drug application continuing until the control currents achieved a stable level.

Collagenase type I, zaterbradine, CsCl, HEPES and 4-aminopyridine were purchased from Sigma Chemicals. Pronase E was obtained from Roche (Basel, Switzerland). Berberine hydrochloride was obtained from the Yichang Humanwell Pharmaceutical Co., Ltd (Yichang, Hubei, China) as base powders and dissolved in distilled water. To maintain the drug and ion concentrations constant, the perfusion rate was strictly controlled using the perfusion device BPS-4 (ALA Scientific Instruments, Inc, Westbury, NY, USA) and a constant-flow pump.

All data were stored on a computer hard disk and analyzed off-line using Clampfit 10.0 (Axon Instruments) and Origin 8.0 software (Origin Laboratory, Northampton, MA, USA). The amplitudes of HCN-mediated currents were defined as the time-dependent components (Istep) at the end of the hyperpolarizing pulses or peak tail currents (Itail) at the beginning of the depolarizing pulses. To construct the I–V correlations, the currents were normalized to their own maximum current measured prior to drug treatment and then plotted as a function of the test potential (Vt). The voltage dependency of the HCN current activation was determined by analysis of the Itail measured at depolarizing potentials. All tail current amplitudes from each individual oocyte were normalized to their own Imax, plotted as a function of Vt and fitted with a Boltzmann function: I/Imax=1/[1+exp(Vt-V_1/2)/k)] to determine the values of the half-point (V_1/2) and the slope (k). The time constants for HCN current activation or deactivation (τactivation or τdeactivation) at different Vt were determined using standard exponential curve fitting. Activating or deactivating currents were fitted to a single exponential function: I(t)=Ae^−t/τ+C. The concentration-effect curves were fitted using the Hill equation in the form f=1/[1+(IC₅₀/D)n], where f represents the increase in HCN currents, expressed as a percentage change from the control values, IC₅₀ was the half-maximum inhibitory concentration of berberine, D was the concentration of berberine and n was the Hill coefficient.

The data are presented as the mean ± standard deviation. The Student’s t-test was used for statistical analysis of the paired observations, and an analysis of variance was performed to test the difference among the groups. P<0.05 was considered to indicate a statistically significant difference.

To examine the effects of berberine on the spontaneous APs in rabbit SAN tissues, transmembrane potentials were recorded by a standard microelectrode technique. Berberine application had depressant effects on spontaneous activity, as demonstrated in Fig. 2. The effects of berberine (3 or 30 μM) on the AP parameters in the normal SA node pacemaker cells are shown in Table I. Compared with the control group, berberine (3 or 30 μM) significantly decreased the DDR and rate of pacemaker firing (RPF), and the changes in the RPF induced by berberine paralleled those in the DDR. Meanwhile, the amplitude of the AP and the maximal diastolic potential decreased as a result of berberine treatment. The above effects occurred following 5–10 min of berberine superfusion and reached their peak within 15–20 min.

Xenopus oocytes were utilized as a heterologous expression system and the actions of berberine on the expression of hHCN4 were analyzed. For the voltage-clamp recordings, the hHCN4 current was elicited by hyperpolarization pulses of 4,000 ms from a holding potential of −60 to −150 mV in 10 mV decrements at 0.1 Hz and then clamped back to 10 mV for 2,000 ms (Fig. 2). Thereafter, selective and non-selective f-channel blockers, zatebradine and CsCl, were utilized to confirm the biophysical properties of the HCN channel. The HCN4 (n=4) currents were readily and completely blocked by 5 mM CsCl. By contrast, 10 μM zatebradine markedly inhibited the hHCN4 currents by 82.1±8.4% (n=4; Fig. 3).

In the present study, it was identified that hHCN4 currents were inhibited by berberine in a concentration-dependent manner at the investigated test potential of −110 mV (Fig. 4A). Superfusion of berberine (1 to 300 μM) reduced the normalized Istep at voltages ranging from −100 to −140 mV, with the most pronounced effects observed at the more hyperpolarized voltages (n=8; P<0.05). Fig. 4 also depicts the correlation between the decreased fraction of the hHCN4 current and the concentration of berberine at −120 mV, with an IC₅₀ value of 32.3±2.1 μM and a Hill coefficient of 1.5±0.2 (n=8; Fig. 4B).

In addition to the inhibitory effect of berberine on current amplitude in HCN4 channels demonstrated above, berberine also modulated HCN4 channel current kinetics. The representative traces of the hHCN4 current and the expanded traces of the outward tail current prior to (control) and following 3 or 30 μM berberine treatment are illustrated in Fig. 5A, and the amplitudes of the measured tail currents were normalized to the peak value, plotted as a function of test voltage, and fitted with a Boltzmann function to obtain the isochronal voltage dependence of HCN4-channel activation. Superfusion of berberine (3 or 30 μM) reduced the normalized Itail at voltages ranging from −90 to −140 mV, with more pronounced effects at the more hyperpolarized voltages (n=8; P<0.05; Fig. 5B). Furthermore, the average value for V_1/2 was −102.7±1.9 mV under the control conditions and −93.8±1.7 mV (n=8; P<0.05) or −80.1±2.4 mV (n=8; P<0.05) following the addition of berberine (3 or 30 μM). The slope factor (k) of the activation curve was decreased from 19.0±1.9 to 14.7±1.8 mV and 12.5±2.3 mV in the presence of berberine at 3 and 30 μM, respectively.

In accordance with the more negative Vt caused by berberine treatment, the activation of HCN4 channels was significantly easier. In the presence of berberine, τactivation was significantly increased in the potential channel from −140 mV to −90 mV, and with Vt becoming more negative, this change was more evident. The values of τactivation were markedly increased by berberine at 3 and 30 μM, from 504.6±39.8 ms (n=8) to 588.4±21.7 ms (3 μM, n=8, P<0.05) and 1176.4±57.3 ms (30 μM, n=8, P<0.05), respectively, at a Vt of −110 mV. The time constant for the current deactivation at 10 mV corresponding to a test potential of −120 mV was 504.6±39.8 ms (n=8) in the control and 588.4±21.7 ms (3 μM, n=8, P>0.05) and 1176.4±57.3 ms (30 μM, n=8, P<0.05) in the presence of berberine at 3 and 30 μM, respectively (Fig. 6).

To study the use-dependent blockage of the hHCN4 current by berberine, a standard activation/deactivation protocol of 2,000 ms pulse from a holding potential of −30 mV to −100 mV followed by a 1,000 ms depolarizing pulse to 0 mV at a rate of 0.25 Hz was used. The HCN currents, measured at the end of −100 mV, were plotted as a function of the pulse number (Fig. 7). During the train pulse stimulation there was no decline in the amplitude of the hHCN4 currents in the absence of berberine. In the absence of berberine, hHCN4 currents generated by 30 pulse burst stimulations were essentially identical. Further studies demonstrated that hHCN4 currents appeared to exhibit a gradual decline following superfusion with berberine (1–300 μM), until the amplitude of HCN currents reached a steady state. Furthermore, compared with the control conditions, superfusion with 30 μM berberine may have no marked blocking effect on the amplitude of hHCN4 currents at the first pulse stimulation, which may be considered as the control pulse. Following berberine application at the holding potential, a significant decline in the amplitudes of the hHCN4 currents occurred. From the results, it was identified that the higher concentration of the drug (30 μM) had faster kinetics of blockage than the lower concentration (3 μM).