The hyperpolarization-activated current (I) is a nonselective inward current activated on hyperpolarization during the diastolic depolarization period of action potential in heart sino-atrial node and Purkinje system, and plays an important role in the generation and regulation of pacemaker activity in both primary and secondary pacemaker cells1, 2, 3. The biophysical characteristics of I comprise time-dependent activation kinetics, a linear current-voltage (I-V) relationship and voltage-dependent block upon extracellular application of Cs+. It has been well documented that both sympathetic and parasympathetic control of heart rate involves modulation of I 4. In adult mammalian heart, β-adrenergic agonist isoproterenol (ISO) stimulates I by shifting the activation curve to more positive potentials1, whereas the vagal neurotransmitter acetycholine (ACh) inhibits I by shifting the activation curve to more negative potentials5. This control is exerted through modulation of adenylate cyclase (AC) and cAMP6. Recently, a multigenic family of pacemaker channels has been cloned, the name of which is hyperpolarization-activated and cyclic nucleotide-gated (HCN) channel genes. The I channel of heart may be encoded by HCN4 or HCN27,8.

Though predominantly expressed in the conductive system of adult heart, If is also present in ventricular myocytes of immature heart since it has been recorded in chick embryonic ventricle9 and neonatal rat ventricle10. It is well known that hypertrophy induces a re-expression of genes encoding fetal proteins11, therefore the currents exist during the fetal life may occur in pathological heart. In fact, If is present in ventricular myocytes of hypertensive rat12 as well as recently in human failing heart13. The presence of If in ventricular myocytes may have relevance with arrhythmogenesis in hypertrophy and heart failure28. Murine embryonic stem (ES) cells are being used by our collaborator to study cardiomyocyte development14. The developmental changes in If channel are among the major interests. The ES cells differentiation in vitro should be similar to the corresponding process happening in vivo. We studied the functional expression and β-adrenergic regulation of the current If in the heart of embryonic mouse instead of ES cell derived cardiomyocytes. Compared with the data resulted from ES cells15, the present study exposed us to the fact how If channel progresses during heart development in vivo. We found that If is highly expressed but can not be stimulated by ISO in early stage cardiomyocytes. ISO stimulates If at late stage of development, but the expression of the current decreases greatly compared with that of early stage. We further show that If is modulated by phosphorylation via protein kinase A (PKA) both in early and late stage cardiomyocytes.


Preparation of embryonic cardiomyocytes

Embryos were removed from pregnant female mice. Through our study 10.5 d, 13.5 d and 16.5 d after coitus were considered as EDS, IDS and LDS. After hearts were dissected from embryos, atrial and ventricular tissues were separated under a dissecting microscope and incubated in an Eppendorf tube with enzyme-containing solution (1 mg/ml collagenase B, Roche Molecular Biochemicals, Mannheimm, Germany) for 30 min at 37±0.3°C Isolated cells were cultured on sterile, gelatin-coated glass cover slips in Dulbecco' modified Eagle' medium (DMEM, Gibco) containing 20% fetal bovine serum for 18-24 h before current recording.


The cells were placed in a temperature-controlled (37±0.3oC recording chamber mounted on the stage of an inverted microscope (Zeiss, Germany) and superfused with the normal Tyrode' solution containing (mM): NaCl 140, NaOH 2.3, KCl 5.4, CaCl2 1.8, MgCl2 1, Hepes 10, Glucose 10, pH 7.4 (adjusted with NaOH). When If was recorded, the normal Tyrode' solution was modified by adding (mM): CdCl2 0.5, BaCl2 1, 4-aminopyridine 2. Extracellular application of drugs was performed by superfusing cells with Tyrode' solution containing the drugs.

In our experiment only spontaneously beating single cardiomyocyte was selected using the whole-cell configuration of the patch-clamp technique16. The cells were held in voltage-clamp or current-clamp mode using an Axopatch 200-A amplifier (Axon Instruments, CA, USA). In voltage-clamp experiments, I was elicited by a hyperpolari n 10 mV increments (holding potential −35 mV). For analysis of the I-V relationship, after a 2000 ms hyperpolarizing voltage step to −115 mV, depolarizing voltage steps (1500 ms) from −110 to +50 mV in 10 mV increments were used to elicit tail current.

Patch pipettes prepared from glass capillary tubes (Liuhe Laboratory Apparatus Factory, Nanjing, China) by means of a two-stage vertical puller (David Kopf Instruments) had a resistance of 2-5 MΩ when filled with the pipette solution (mM): NaCl 10, potassium aspartate 130, Na2ATP 2, EGTA 1, MgCl2 2, Na2GTP 0.1, Hepes 10, pH 7.2 (adjusted with KOH). Cell membrane capacitance (Cm) was determined on-line using the ISO2 (MFK, Frankfurt, FRG) software program. Data were acquired at a sampling rate of 10 kHz, filtered at 2kHz, stored on hard disk and analyzed off-line using the ISO2 analysis software package.

Data analysis

The amplitude of I was measured as the difference between the instantaneous current at the beginning of the hyperpolarizing pulse and the steady-state current at the end of hyperpolarization17. Currents were normalized to membrane capacitance to calculate current densities. The steady-state activation curve was constructed from the amplitude of time-dependent inward current during hyperpolarizing steps (−40 mV to −140 mV). Specific conductance of If was determined for each cell according to the equation g=I/(Vm−Vrev), where g is the conductance calculated at the membrane potential Vm, I is the current amplitude, and Vrev is calculated from the analysis of tail currents. The values were normalized to the maximal current conductance (gmax) and fitted with the Boltzmann equation: g/gmax=1/{1+exp[(V1/2−Vm)/S]}, where Vm is the membrane voltage, V1/2 is the voltage at half-maximal activation, and S is a slope factor at Vm=V1/2. For I-V relation, the amplitude and reversal potential of currents were analyzed from the tail current measured between peak current (15 ms after depolarization to omit possible interference of sodium or capacitance currents18) and maintained current at the end of the clamp pulse. Data are presented as mean±standard error of the mean (S.E.M) when appropriate. Statistical analysis was performed using Student's paired or unpaired t tests and values of P < 0.05 were considered significant.


The following chemicals were all purchased from Sigma: protein kinase A inhibitor (PKI), ISO, 8-Br-cAMP, forskolin, Hepes, CsCl, and 4-aminopyridine. PKI was dissolved in pipette solution and stored frozen at −20oC. The PKI aliquots were thawed immediately prior to use and diluted with pipette solution to the desired concentration. After the whole cell configuration was constructed, PKI was dialyzed into the cell through the pipette. 8-Br-cAMP was stored and used in a dark room. The substances for cell culture were purchased from Gibco. And the other chemicals, if not stated, were all from Chinese companies.


Electrophysiological properties of I

We used the EDS ventricular myocytes to identify the electrophysiological characteristics of If. A serial of hyperpolarizing steps in 10 mV increments from −40 mV to −140 mV (holding potential −35mV) elicited time-dependent and voltage-dependent inward currents. The current traces are showed in Fig 1A, and the amplitudes of If were measured in four EDS ventricular myocytes. Specific conductance of If was normalized and fitted with the Boltzmann equation, yielding an activation threshold around −40 mV, a half maximal activation voltage (V1/2) of −76±2.24 mV (n=4) and a slope factor (S) of −16±1.89 mV−1 (n=4) (Fig 1B). Fig 1C and D show the current traces for analyzing I-V relationship of If before and after extracellular application of Cs+ (2 mM). Compared with the control condition (Fig 1C), the inward currents were almost completely blocked by 2 mM CsCl (Fig 1D). Fig 1E shows the I-V relation of I in absence and presence of extracellular Cs+. Under the control condition, the I-V curve is relatively linear with a reversal potential of −32.6±1.42 mV (n=5). In the presence of Cs+, at potentials negative to the reversal potential of If, the I-V curve demonstrates a voltage dependent inhibition of If by Cs+. But when positive to the reversal potential, the amplitude of If evoked by depolarizing voltage steps is unaltered. Our results demonstrate that the electrophysiological properties of If are similar to those reported in adult mammalian heart3,19,20,

Figure 1
figure 1

Electrophysiological properties of If A, a serial of hyperpolarizing steps from −40 mV to −140 mV elicited time-dependent and voltage-dependent inward currents. B, the steady-state activation characteristics of 4 experiments. C and D, current traces for analyzing I-V relationship of If before and after extracellular application of Cs+ (2 mM). E, the I-V relationship of If in absence and presence of extracellular Cs+ (• ,control; , in the presence of 2 mM CsCl).

The decreases in expression of I f during development

To detect the functional expression of If in the heart at different stage of development, we recorded and calculated the densityof If and the percentage of cells expressing it. The heart of IDS and LDS is big enough for us to separate atrium from ventricle, but that of EDS is so small that only ventricles were collected. Fig 2A shows the percentage of atrial and ventricular myocytes expressing If at different stage of development. If was detected in a large percentage (85.7%, n=21) of EDS ventricular cells, but fewer (71.6%, n=29) in IDS and fewest (44.4%, n=27) in LDS. Moreover, the percentage of atrial cells expressing If also declines from 77% (n=8) in IDS to 48.3% (n=25) in LDS. Fig 2B illustrates decrease in the densities of If during development. The If densities were 7.24±1.28 pA pF−1 (n=18) and 6.48±0.99 pA pF−1 (n=21) in EDS and IDS ventricular myocytes, respectively. LDS ventricular cells display a significantly lower current density of 3.76±1.04 pA pF−1 (n=12, P < 0.01), as compared with EDS and LDS ventricular cells. Current density of atrial cells were also decreased significantly from 8.86±2.34 pA pF−1 (n=6) in IDS to 4.62±1.06 pA pF−1 (n=12, P < 0.01in LDS. Since If may play an important role in determining the diastolic depolarization phase of action potential (AP), a typical recording of AP from spontaneously contracting ventricular myocytes was performed in the current-clmap mode. Fig 2C illustrates that in the AP of an EDS ventricular myocyte, there is a slow repolarization after the upstroke and a rapid diastolic depolarization. In a LDS ventricular myocyte (Fig 2D), the diastolic phase of AP is flat, but that in the EDS ventricular cell is fairly steep.

Figure 2
figure 2

The decreases in expression of If during development A, the percentage of atrial and ventricular myocytes expressing If. B, decreases in the current densities (If/Cm, where Cm is the membrane capacitance) of If during development. *Statistically significant difference between LDS and EDS or IDS at the P < 0.05 level was performed using Student's unpaired t test. If was evoked at −110 mV, lasting for 2000 ms. C, the spontaneous action potentials (AP) recorded from an EDS ventricular myocyte. D, the action potentials of a LDS ventricular myocyte.

Isoproterenol stimulates I f in LDS but not EDS cardiomyocytes

It is well know that If increases in response to ISO in adult cardiomyocytes1,21. We testedβadrenergic regulation of If at −110 mV. Our results show that ISO stimulated If in LDS but not EDS cardiomyocytes. In Fig 3A and B, current traces (left panel) and corresponding time course (right panel) of typical experiments are displayed. Shortly after extracellular application of ISO (2μM), If increased in a LDS ventricular cell (Fig 3B). Fig 3C (right panel) shows that ISO increased the averaged current density significantly from 2.66±0.93 pA pF−1 to 4.18±0.88 pA pF−1 (n=5, P < 0.05) in LDS ventricular cells. The stimulation of If by ISO in LDS atrial cells was also recorded. If densities in control condition and in the presence of ISO were 4.03±1.82 pA pF−1 and 6.34±1.85 pA pF−1 (n=6, P < 0.05), respectively. But in EDS ventricular cells If did not change apparently (Fig 3A). Fig 3C (left panel) shows that If densities of EDS ventricular cells before and after ISO application were 4.01±0.06 pA pF−1 and 4.05±0.10pA pF−1 (n=6).

Figure 3
figure 3

Isoproterenol stimulates If in LDS but not EDS ventricular myocytes. A, the effect of ISO (2μM) on If in an EDS ventricular myocyte. B, the effect of ISO (2μM) on If in a LDS ventricular myocyte. (left panel: the current traces of If; right panel: time course of the experiments, voltage protocol identical to Fig 2). 1, control; 2, ISO (2 μM). C, the difference in current densities between the absence and presence of ISO was resulted from the test of EDS and LDS ventricular myocytes. *The difference from the test of LDS ventricular myocytes was statistically significant (P < 0.05, paired t test).

Effects of forskolin and cAMP on I in EDS cardiomyocytes

Because ISO did not stimulate I in EDS ventricular cardiomyocytes, we tested the effects of forskolin (a direct activator of AC) and 8-Br-cAMP (a membrane-permeable analogue of cAMP) on If at early stage of development. The current traces of If (Fig 4A) as well as the time course of experiments (Fig 4B) demonstrate that both forskolin and 8-Br-cAMP increase the amplitude of the current. A summary of these data are showed in Fig 4C, where If densities under control conditions and after application of forskolin (3μM) were 5.54±1.33 pA pF−1 and 7.04±2.35 pA pF−1 (n=4), respectively. Similarly, after extracellular application of 8-Br-cAMP (400μM), If densities were enhanced from 4.64±1.77 pA pF−1 to 5.85±1.35 pA pF−1 (n=4). Altogether, these experiments suggest that in the βadrenergic signaling cascade system, AC and cAMP function fairly well at early stage development.

Figure 4
figure 4

Effects of forskolin and 8-Br-cAMP on If in EDS ventricular myocytes. Both forskolin μM) and 8-Br-cAMP (400 μM) can increase the size of If (voltage protocol identical to Fig 2) in EDS ventricular myocytes, 1, control; 2, application of drugs. A, the current traces of If. B, time course of the experiments. C, the If densities were increased apparently after application of forskolin or 8-Br-cAMP.

I f is stimulated by phosphorylation via PKA in EDS and LDS cardiomyocytes

It is well known that cAMP is the second messenger of βadrenergic signaling cascade system. The former results show that cAMP increase the amplitude of If both in LDS and EDS cardiomyocytes. We further studied whether the effect of cAMP on If was mediated through a direct cAMP binding and/or an indirect phosphorylation via cAMP-dependent PKA at these two developmental stages. In order to differentiate between these two regulatory pathways, we test the effects of ISO and 8-Br-cAMP on If again after 15 min of cell dialysis with PKI, a highly selective peptide inhibitor of PKA22. Fig 5A shows that the stimulation of ISO (2μM) on If was greatly depressed by intracellular perfusion with PKI (1 mg/ml) in a LDS ventricular myocyte. If density under control conditions is 2.30±1.02 pA pF−1 (n=3). After 15 min of intracellular dialysis with PKI, the current density in presence of ISO (2μM) was 2.31±1.14 pA pF−1 (n=3). Moreover, ISO had no effect on If under this condition in LDS atrial myocytes (n=3, data not shown). Fig 5B demonstrates that in an EDS ventricular myocyte, 8-Br-cAMP (400μM) did not stimulate If after intracellular perfusion with PKI (1mg/ml). The If densities under control conditions and in presence of 8-Br-cAMP after dialysis with PKI were 3.82±0.28 pA pF−1 and 3.99±0.22 pA pF−1 (n=4), respectively. As it is noted before, in EDS ventricular myocytes, ISO cannot stimulate If, but cAMP can increase the amplitude of the current. Here, the block of PKI on the effect of cAMP in EDS ventricular myocytes was observed.

Figure 5
figure 5

If is stimulated by phosphorylation via PKA in EDS and LDS ventricular myocytes. A, the stimulation of ISO (2μM) on If current was greatly depressed by intracellular perfusion with PKI (1 mg/ml) in a LDS ventricular myocyte. B, in an EDS ventricular myocyte, 8-Br-cAMP (400μM) can't stimulate If any more under intracellular perfusion with PKI (1 mg/ml). Left panel, the current traces of If. Right panel, time course of the experiments. 1, PKI; 2, PKI and ISO or 8-Br-cAMP. Voltage protocol is identical to Fig 2.


The present results demonstrated that both I densities and the percentage of cells expressing it are high in the working cells during early stage development of murine embryo. Since If is essential for cell's pacemaker activity2,3,4,, the presence of which might imply that there is automatic rhythmicity in these cells. In other word, these cells have not differentiated from pacemaker cells to mature working cells. This point can also be deduced from the AP we recorded from the spontaneously beating ventricular myocytes. The ventricular myocyte of early stage has an AP profile similar to that of adult sino-atrial node cells, characterized by the slow repolarization after the upstroke and the rapid diastolic depolarization. At late stage of development, If densities and the percentage of cells expressing it are all significantly decreased both in atrial and ventricular myocytes; the AP profile of ventricular myocyte is more like the mature ventricle'. We conclude that If exists in most of the early stage cardiomyocytes, which have not differentiated completely. At late stage of development, when atrial and ventricular myocytes develop into mature working cells, which have no automatic rhythmicity, If will be greatly downregulated in these cells. But increased current densities were observed by our collaborator during cardiomyogenesis15. Maybe, only the cells of heart conduction system had been collected with the procedure they used, so there is some differences from atrial and ventricular myocytes. If is re-expressed in the adult ventricular myocytes isolated from heart failure and hypotrophy11,13,23, therefore the automatic rhythmicity may reappear in these ventricular myocytes and the re-expression of If may have relevance with arrhythmogenesis in these heart diseases17,24. Therefore, our research on If in the embryonic mouse heart is also useful to study pathophysiological phenomena of heart failure and hypertrophy.

It has been documented that f-channel, regulated by the autonomous system, is important for the regulation of heart rhythmicity. In the adult mammalian heart, the binding of βadrenergic agonist to β-adrenergic receptor (βAR) is coupled to an intracellular cascade by the stimulatory G protein, Gs. Agonist occupancy activates AC to increase intracellular concentrations of cAMP, which stimulates f-channels directly in sino-atrial node25,26. But in Purkinje fibres, instead of the direct interaction between cAMP and f-channels, the phosphorylation of the channels is involved27,28. Our study demonstrates that ISO stimulates If in LDS but not EDS cardiomyocytes of murine embryo. We conclude that at early stage of development, the β-adrenergic regulation of If is not mature; the β-adrenergic signaling cascade is not able to work well until the late stage of development. We further find that foskolin and 8-Br-cAMP are able to increase the amplitude of If in EDS ventricular cells, indicating that AC and cAMP function fairly well. The lack of response to ISO of If in EDS cardiomyocytes could be due to the coupling deficiency precedes AC: either expressions of β-AR and the Gs-protein or their coupling mechanism.

Altogether, there is a cAMP-dependent regulation of If both in EDS and LDS embryonic cardiomyocytes. In order to detect whether the effect of cAMP on If was mediated through a direct cAMP binding or an indirect phosphorylation via PKA, we used PKI to test the response of If to 8-Br-cAMP in EDS ventricular myocytes. The result is that 8-Br-cAMP did not increase the amplitude of If after intracellular perfusion with PKI. Likewise, at late stage of development, the stimulation of ISO on If was abolished upon intracellular perfusion with PKI. We conclude that at early and late stage of development, the regulation of If channels appears to be mediated by phosphorylation via PKA rather than by a direct cAMP binding. This point is identical to our collaborator's finding resulted from ES cell derived cardiomyocytes.