Electrophysiology of hyperpolarization-activated cyclic nucleotidegated cation channel 2 and hyperpolarization-activated cyclic nucleotide-gated cation channel 4 expressed in HEK293 cells

Abstract

The action potential of pacemaker cells in the sino-atrial node forms the automatic rhythm of the heart. The automatic depolarization in phase 4 is the basis of the automaticity in pacemaker cells. Many currents are included in phase 4, such as calcium current,1 TTX-sensitive sodium current,2 sustained inward current (Isi),3 decay of delayed rectifier potassium current,4 etc. Funny current (If) has long been recognized as important in this phase and is activated at hyperpolarized potentials during cell diastole and in turn activates other currents to form automatic depolarization.5 It is controlled by voltage and cAMP, an intracellular messenger, so the auto no-mic nerve can modulate this current by increasing or decreasing intracellular cAMP. Since a gene family termed hyperpolarization-activated cyclic nucleotidegated cation channel (HCN) was discovered to encode this channel, there are few domestic articles concerning the electrophysiology of the different types of this channel.6 The aim of this article is to describe the electrophysiological properties of human HCN2 and HCN4 channels expressed in HEK293 cells. METHODS Cell culture HEK293 cells were grown in Dulbecco's DMEM medium with penicillin (10 U/ml), streptomycin (100 g/ml), 10% fetal bovine serum and incubated at 37°C in 5% CO2 incubator. Construction of plasmid cDNA encoding human HCN2 or HCN4 (a gift from Professor Kaupp, Germany) were separately cloned and transiently expressed in HEK293 cells as described previously. Briefly, the human HCN2 or HCN4 gene was ligated into a shuttle vector pAdTrack-cytomegalovirus (CMV). Homologous recombination was performed in BJ5183 bacteria by cotransforming a linearized shuttle plasmid and adenovirus backboned vector pAdEasy-1. The positive recombinant adenovirus plasmid was digested with Pac I and transfected into HEK 293 cells for current recording.7 Solutions Currents were recorded 1 to 3 days after transfection. The recording solution contained (mmol/L): NaCl 120, KCl 20, CaCl2 2, MgCl2 1, 4-aminopyridine 2, Hepes 10, Glucose 5; pH was adjusted to 7.35 by NaOH. The pipette solution contained (mmol/L): K-aspartate 130, MgCl2 2, EGTA 11, Na2ATP 5, CaCl2 5, Hepes 10; pH was adjusted to 7.35 by KOH. Electrophysiological recording The whole cell currents were recorded with an Axopatch 200B amplifier (Axon Instruments, USA). Data were collected with software Pclamp 6.0 (Axon Instruments). The pipettes were made from borosillicate capillaries, pulled by a PP-83 puller (Narishiger, Japan) and had resistances of 3 to 8 M.. All experiments were done at room temperature. Data analysis The activation curves of the HCN2 and HCN4 channels were obtained by standard two-step protocols: the cells were held at -50 mV and hyperpolarized from -60 mV to -150 mV, with 10 mV decrements. The difference between the beginning of hyperpolarization and the end was calculated as “hyperpolarization-activated currents”. After normalization to maximal amplitude, currents were plotted to obtain the activation curve. The activation curve can be fitted by the Boltzmann equation: y=1/(1+exp((V-V1/2)/s)), where y is the normalized currents, V is voltage, V1/2 is the half-activation voltage and s is the inverse slope factor. The time constants of current activation were obtained by fitting the hyperpolarization-activated currents with a single exponential curve. Reversal potentials were obtained by linear extrapolation of the tail currents amplitudes recorded from a series of depolarizations (from -60 mV to +20 mV) from -140 mV. The relative permeability of HCN channel to sodium and potassium (PNa/K) was calculated by the Goldman-Hodgkin-Katz equation. The student's unpaired t-test was done by software SPSS11.5. All results are expressed as mean ± standard error (SE). A P<0.05 was considered statistically significant. RESULTS Voltage-dependence of the inward currents HCN2 or HCN4 channel protein was detected in infected cells by immunofluorescence cytochemistry.7 At hyperpolarized potentials there were no inward currents in native HEK293 cells, while If-like time- and voltage-dependent inward currents can be found in HEK293 cells transfected with HCN2 or HCN4. The normalized currents were fitted by the Boltzmann equation, and the V1/2 was (-114.8±3.3) mV and (-125.9±2.9) mV in HCN2 and HCN4 channels respectively (P=0.024). The inverse slope factor was (11.1±1.2) mV and (13.7±1.3) mV, with no statistical significance (P=0.22) (Fig. 1).Fig. 1.: The activation curves of HCN2 (n=10) and HCN4 channels (n=11).Activation kinetics of channels The hyperpolarization-activated channels can be perfectly fitted by single exponential function. As Fig. 2 shows, the activation becomes faster at more hyperpolarized potentials. The activation time constants at -110 mV were 0.99±0.21 seconds and 8.47±2.85 seconds respectively in HCN2 and HCN4 channels (P<0.001).Fig. 2.: The activation kinetics of HCN2 (n=19) and HCN4 channels (n=10).Ion selectivity The cells were held at -140 mV, then clamped to -60 mV until to +20 mV, the reversal potentials of HCN2 and HCN4 channels were -20.1 mV and -23.0 mV with 20 mmol/L potassium in the external solution. And the corresponding PNa/K was 0.40 and 0.34 respectively (Fig. 3).Fig. 3.: The relationship between test potentials and the tail current amplitudes of HCN2 (n=13) and HCN4 (n=6).Response to external Cs+ When 2 mmol/L Cs+ was added, the inward currents of HCN2 and HCN4 channels obviously decreased, with little change in the deactivated outward currents. DISCUSSION The “funny” current (If) is important in the pacemaker cells of sino-atrial node; it causes the slow diastolic depolarization of the action potential that determines heart rate and rhythm. In the 1990's a gene family termed HCN was cloned and investigators have tried to identify the molecular constituents of native pacemaker channels. Four types of HCN channels have been found; HCN1-HCN4.6 So far, studies confirm that there are three types of HCN isoforms present in cardiac tissues; HCN1, HCN2 and HCN4.8 The HCN3 isoform only exists in brain tissues. But recently there is a report that transcripts of HCN3 were also detected by reverse transcriptase PCR in heart ventricle.9 HCN4 channels are essential for the proper generation of pacemaker potentials in the fetal emerging sino-atrial node.10 The rabbit SA node contains predominantly HCN4 channels whereas HCN2 channels are dominant in rabbit ventricle. Both HCN2 and HCN4 channels respond best to intracellular cAMP and can be influenced greatly by the autonomic nerve, while HCN1 channels are moderately modulated by cAMP and less influenced by the autonomic nerve. So we studied the electrophysiology of HCN2 and HCN4 channels. The activation potentials of both HCN2 and HCN4 channels were more negative than that of native If channels. The reason is still unknown. Recent evidence suggests that the activation curve is context-dependent. When expressed in ventricular myocytes, the activation curves of HCN2 and HCN4 are shifted 16-19 mV more positively than that of the curves of the same types expressed in HEK293 cells.11 A reasonable explanation is that unknown intracellular factors are present in native cells but not in HEK293 cells that inhibit the voltagesensitive protein in channel gating. As our results showed, HCN2 channels were activated at more positive potentials than HCN4 channels and activated dramatically faster. The C-terminus of HCN channels inhibits the gating of the channel by an unclear mechanism and is important in channel kinetics. A cyclic nucleotide binding domain (CNBD) at the C-terminus corresponds with the sensitivity to cAMP of HCN channels. After cAMP binding to CNBD, the inhibition of the C-terminus is released, resulting in less energy needed for channel activation, thus the activation curve shifts to more positive potentials and the channel activates faster. The response to cAMP is obvious in the HCN2 and HCN4 channels. In the absence of cAMP, the activation curve in HCN1 channels is more positive than in the HCN2 and HCN4 channels; the basic inhibition by the C-terminus is weak, thus the modulation of HCN1 channels by cAMP is moderate.12 The V1/2 of HCN2 and HCN4 channels are more negative than that of other studies, we do not know the exact reason. Many factors can affect the channel activation. The length of the C-terminus influences its inhibition to channel gating; the shorter the C-terminus, the less the basic inhibition. The C-terminus of the HCN4 channel contains more amino acids than any other HCN channel, and it is activated at more hyperpolarized potentials.13 If the C-terminus of HCN4 is substituted by that of HCN1, the channel is activated dramatically faster and loses sensitivity to cAMP. The S3-S4 linker and S4-S5 linker also take part in the activation kinetics of HCN channels.14,15 Given the fact that the activation kinetics of native f-channels is between that of HCN2 and HCN4, one would expect that different isoforms of HCN may co-assemble to form the native f-channel. There is controversy as to whether native f-channels are formed by the same kind of isoforms. More work is still needed to be done to elucidate the f-channels.