In vitro characterization of HCN channel kinetics and frequency dependence in myocytes predicts biological pacemaker functionality

Abstract

The pacemaker current, mediated by hyperpolarization-activated cyclic nucleotide-gated (HCN) channels, contributes to the initiation and regulation of cardiac rhythm. Previous experiments creating HCN-based biological pacemakers in vivo found that an engineered HCN2/HCN1 chimeric channel (HCN212) resulted in significantly faster rates than HCN2, interrupted by 1-5 s pauses. To elucidate the mechanisms underlying the differences in HCN212 and HCN2 in vivo functionality as biological pacemakers, we studied newborn rat ventricular myocytes over-expressing either HCN2 or HCN212 channels. The HCN2- and HCN212-over-expressing myocytes manifest similar voltage dependence, current density and sensitivity to saturating cAMP concentrations, but HCN212 has faster activation/deactivation kinetics. Compared with HCN2, myocytes expressing HCN212 exhibit a faster spontaneous rate and greater incidence of irregular rhythms (i.e. periods of rapid spontaneous rate followed by pauses). To explore these rhythm differences further, we imposed consecutive pacing and found that activation kinetics of the two channels are slower at faster pacing frequencies. As a result, time-dependent HCN current flowing during diastole decreases for both constructs during a train of stimuli at a rapid frequency, with the effect more pronounced for HCN2. In addition, the slower deactivation kinetics of HCN2 contributes to more pronounced instantaneous current at a slower frequency. As a result of the frequency dependence of both instantaneous and time-dependent current, HCN2 exhibits more robust negative feedback than HCN212, contributing to the maintenance of a stable pacing rhythm. These results illustrate the benefit of screening HCN constructs in spontaneously active myocyte cultures and may provide the basis for future optimization of HCN-based biological pacemakers.

Figure 1. Expression of HCN2 and HCN212 in newborn rat ventricular myocytes

A, representative whole-cell current tracings of myocytes expressing HCN2 (top) and HCN212 (bottom) channels. Currents were evoked by stepping from a holding potential of −25 mV to different hyperpolarizing voltage steps ranging from −25 to −125 mV in 10 mV increments. B, the mean I –V relationship of HCN2 (n = 8, filled circles) and HCN212 (n = 12, open circles). The HCN current (I_HCN) is defined as the time-dependent component at the end of the 3 s test pulse. C, representative instantaneous current I_{ins, HCN2}, I_ins,HCN212 and I_ins,GFP recorded at a single hyperpolarizing step (−120 mV) from a holding potential of −30 mV. The inset shows how the instantaneous component was measured; it is defined as the difference between the zero current level and the initial current level where the time-dependent current component begins, and includes leak current. D, the instantaneous current density in response to voltage steps to −120 mV in myocytes over-expressing HCN2 (n = 15), HCN212 (n = 13), GFP (n = 17). *P < 0.05 versus GFP.

Figure 2. Steady-state properties of HCN2 and HCN212 in newborn rat ventricular myocytes

A, the mean activation–voltage relation of HCN2 and HCN212 in the absence and presence of cAMP. Mean fractional activation curves of HCN2 (squares) and HCN212 (circles) obtained in the absence (black symbols) and in the presence (grey symbols) of 10 μmol l⁻¹ cAMP in the pipette solution. For illustrative purposes, the mean data were fitted to the Boltzmann equation for experiments in the absence (continuous black lines) and in the presence (dashed grey lines) of cAMP. Inset: mean values without (black) and with (grey) cAMP. B, voltage dependence of activation (filled symbols) and deactivation (open symbols) time constants of HCN2 (squares) and HCN212 (circles). Mean activation values were obtained from 10 cells for both HCN2 and HCN212; mean deactivation time constant values were obtained from 7 and 6 cells for HCN2 and HCN212, respectively. The curves are the fits to the equation . *Voltages differ by multiple comparison (P < 0.05).

Figure 3. Simulated whole-cell current tracing and simulated spontaneous action potentials in the Kurata sinoatrial node (SAN) model with original Kurata I_f, HCN2 expression and HCN212 expression

A, currents were evoked by stepping from a holding potential of −25 mV to different hyperpolarizing voltage steps ranging from −25 to −125 mV in 10 mV increments. B, spontaneous action potentials were simulated with each of the pacemaker current configurations – the SAN model with unchanged I_f (Kurata et al. 2002) is shown on the left, HCN2 in the middle, and HCN212 on the right.

Figure 4. Effects of HCN2 and HCN212 channel expression on spontaneous activity of newborn rat ventricular myocytes

A, mean spontaneous beat rate of myocytes expressing GFP, HCN2 and HCN212. *P < 0.05 versus GFP; **P < 0.05 versus HCN2 and GFP. B, representative examples of spontaneous action potentials with bursting behaviour recorded from an HCN2-expressing (top) and HCN212-expressing culture (bottom). At the far right, the portion of each trace indicated by the box is displayed on an expanded time scale. C, the incidence of bursting behaviour in HCN-expressing monolayers. *P < 0.05 versus HCN212.

Figure 5. Prepulse modulation of HCN channels

A, representative HCN2 currents elicited by protocol 3. The inset shows the entire protocol, which includes an activating prepulse to −80 mV followed by a step to +20 mV with increasing interpulse interval and an additional activating step to −80 mV; the portion of the protocol not displayed in the current traces is represented by dashed lines. The double arrow labelled I_ins represents the constant portion, or minimal, I_ins. The double arrow labelled ΔI_ins indicates the enhancement of the instantaneous current observed at the shortest interpulse interval, reflecting the largest measured ΔI_ins. I_HCN is the time-dependent steady-state current, and its magnitude is labelled for the same current trace. B, dependence of ΔI_ins,HCN2 (n = 3, squares) and ΔI_ins,HCN212 (n = 3, circles) on the interpulse interval. ΔI_ins was normalized to the amplitude of the time-dependent steady-state component of HCN current (ΔI_ins/I_HCN). Data were fitted to a single exponential function; the 150 ms time point is highlighted by the vertical dashed line. C, representive HCN2 currents elicited by protocol 4. The inset shows the full protocol, with the portion omitted from the current traces represented by dashed lines. Prepulses (−80 mV) were evoked for different prepulse lengths (150 ms, 1000 ms, 1850 ms and 2700 ms) and the interpulse interval kept at 150 ms. D, activation time constant of HCN2 (n = 5, squares) and HCN212 (n = 4, circles) during the second −80 mV step. *Statistically significant (P < 0.05) difference from other prepulse durations in same HCN group by multiple comparison.

Figure 6. Frequency-dependent modulation of HCN currents in myocytes

A, representative HCN currents elicited by repetitive steps from −80 mV to +20 mV for 150 ms at 0.5 Hz pacing rates. B, overlapping of the 20th HCN2 current (top) and HCN212 current (bottom) at 0.5 (black) and 3.3 Hz (grey) pacing rates. C, the changes of total I_ins in HCN2 (n = 8) and HCN212 (n = 6) at 0.5 (white) and 3.3 Hz (black). *P < 0.05 versus 3.3 Hz by paired test; #P < 0.05 versus HCN212. D, activation time constant of HCN2 and HCN212 during the last −80 mV step at 0.5 (white) and 3.3 Hz (black) pacing rates. *P < 0.05 versus 3.3 Hz by paired test. E, the mean beat-to-beat changes of the time-dependent current of HCN2 (squares) and HCN212 (circles) at the 150 ms time point (I_{HCN,150 ms}) during the 20-pulse train at 0.5 Hz (open symbols) and 3.3 Hz (filled symbols). *P < 0.05 versus HCN212 at 3.3 Hz by multiple comparison.