Functional Roles of a Ca²⁺-Activated K⁺ Channel (SK2) in Atrioventricular Nodes

Abstract

Since the first description of the anatomical atrioventricular nodes (AVN), a large number of studies have provided insights into the heterogeneity of the structure as well as a repertoire of ion channel proteins which govern this complex conduction pathway between the atria and ventricles. These studies have revealed the intricate organization of multiple nodal and nodal-like myocytes contributing to the unique electrophysiology of the AVN in health and diseases. On the other hand, information regarding contribution of specific ion channels to the function of the AVN remains incomplete. We reason that the identification of AVN-specific ion channels may provide a more direct and rationale design of therapeutic target in the control of AVN conduction in atrial flutter/fibrillation, one of the most common arrhythmias seen clinically. In this study, we took advantage of two genetically altered mouse models with over-expression or null mutation of one of the small conductance Ca²⁺-activated K⁺ channel isoform, SK2 channel and demonstrated robust phenotypes of AVN dysfunction in these experimental models. Over-expression of SK2 channels results in the shortening of the spontaneous action potentials (APs) of the AVN cells and an increase in the firing frequency. On the other hand, ablation of the SK2 channel results in the opposite effects on the spontaneous APs of the AVN. Furthermore, we directly documented the expression of SK2 channel in mouse AVN using multiple techniques. The new insights may have important implications in providing novel drug targets for the modification of AVN conduction in the treatment of atrial arrhythmias.

INTRODUCTION

The atrioventricular node (AVN) is a highly specialized pacemaking tissue located at the junction of the right atrium and ventricle. Indeed, it is the only electrical connection between the atria and ventricles and provides the critical delay between atrial and ventricular contraction to allow for proper atrial emptying prior to the start of the ventricular contraction. Pharmacological slowing of impulses across AVN is widely used clinically in atrial flutter/fibrillation to ensure physiological ventricular responses in these conditions. Previous studies have identified the roles of several distinct ion channels in the AVN function^1-5 and recent work have begun to assemble an array of ion channel genes in the pacemaking tissues.⁶ On the other hand, information regarding contribution of specific ion channels to the function of the AVN remains incomplete. We reason that the identification of AVN-specific ion channels may provide a more direct and rationale design of therapeutic target in the control of AVN conduction in atrial flutter/fibrillation, one of the most common arrhythmias seen clinically.

Specifically, we have recently identified several isoforms of Ca²⁺-activated K⁺ channels (K_Ca) in human and mouse cardiac myocytes which we have shown to be critical in sculpting the duration of the cardiac action potential (AP).^7,8 K_Ca channels are highly expressed in atrial compared to ventricular tissues. Moreover, not only does the current play important functional roles in mouse atrial myocytes, it also contributes significantly to the repolarization process in human atria.^7,8 K_Ca channels are present in a wide variety of cells, where they integrate changes in intracellular Ca²⁺ concentration [Ca²⁺_i] with changes in K⁺ conductance and membrane potential.^9,10 K_Ca channels can be divided into three main subfamilies: the large-conductance Ca²⁺- and voltage-activated K⁺ channels (BK), the intermediate-conductance Ca²⁺-activated K⁺ channels (IK, K_Ca3) and the small conductance Ca²⁺-activated K⁺ channels (SK or K_Ca2).^9-13 SK channels are encoded by at least three distinct genes, namely KCNN1 (SK1), KCNN2 (SK2), KCNN3 (SK3).^9,10,13

Here, we directly document the robust expression of SK2 channel in mouse AVN. Moreover, using genetically altered mouse models with over-expression or null mutation of one of the K_Ca isoforms, SK2 channel, we demonstrate significant changes in the AVN function. The new insights into the functional roles of SK2 channel in AVN may have important implications in providing novel drug targets for the modification of AVN conduction in the treatment of atrial arrhythmias.

MATERIALS AND METHODS

A detailed Materials and Methods is described in the Online Data Supplement.

SK2 Null Mutant Mice (SK2 +/Δ mice) and Transgenic Mice Over-Expressing SK2 Channel (SK2 +/T)

SK2 null mutant mice were generated as previously described.¹⁴ Transgenic mice over-expressing SK2 channels were developed via insertion of a tetracycline-regulatory cassette into the SK2 locus as previously described.^15,16 Both SK2 null and transgenic mouse lines were backcrossed more than seven generations onto the C57Bl/6J background. Compared with wild-type (WT) littermate mice, in the absence of doxycycline, the SK2 protein and SK2 mRNA are over-expressed in heterozygotes (SK2 +/T) ~10 and 4.5 folds as shown by Western blot and quantitative PCR, respectively. Since the transgenic animals were generated by site-specific insertion of a tetracycline-regulatory cassette into the 5' untranslated region of the SK2 gene, the SK2 channels could be over-expressed without interfering with the normal profile of SK2 expression.

Atrioventricular Node Recordings

AVN preparation was prepared as previously described.^1,6 AVN region was recognized by its anatomic landmarks (Figure 2A). Spontaneous APs were recorded from isolated AV nodal preparations using microelectrode techniques with 3 M KCl microelectrodes at 33°C as previously described.¹⁷

Figure 2.

Spontaneous APs recorded from intact AVNs from WT, SK2+/T and SK2+/Δ mice. (A) Photomicrograph of endocardial view of typical AVN preparation from mouse hearts. RA, right atrium; IVC, inferior vena cava; RV, right ventricle; CS, coronary sinus. (B) Representative examples of spontaneous APs from intact AVNs showing an increase in the AVN firing frequency in SK2 +/T mice and a decrease in the firing frequency in SK2 +/Δ mice. (C) Summary data from the three groups of animals for DDR (mV/s), CL (ms), MDP (mV), APA (mV), V_max (V/s), APD₅₀ (ms), and APD₈₀ (ms). *P<0.05.

Single AVN cells were isolated from WT and mutant mice as previously described with some modification.^1,17-19 Whole-cell Ca²⁺-activated K⁺ current (I_K,Ca) was recorded from single AVN cells at room temperature using patch-clamp techniques as previously described.^7,20

Electrocardiographic (ECG) Recordings

ECG recordings were obtained at 33°C using Bioamplifier (BMA 831, CWE, Incorporated, Ardmore, PA) as previously described.¹⁷

Immunofluorescence Confocal Microscopy and Immunohistochemistry

Immunofluorescence labeling was performed as described previously.⁷ Immunohistochemistry and antibodies used are described in the Online Data Supplement.

RESULTS

SK2 +/T and SK2 +/Δ mice show evidence of sinoatrial node (SAN) and AVN dysfunction

We documented robust phenotypes of SAN and AVN dysfunction in the SK2 +/T and SK2 +/Δ mice. Figure 1A shows ECG recordings from SK2 +/T and SK2 +/Δ mice compared to WT animals illustrating significant sinus bradycardia with prolongation of the PR intervals in SK2 +/Δ mice compared to WT animals. In contrast, SK2 +/T mice show significant shortening of the RR and PR intervals. Figure 1B further illustrates examples of ECG recordings in SK2 Δ/Δ mice showing complete AV block with AV dissociation. Summary data for PR intervals are shown in Figure 1C (n=10 from each group, *p<0.05). These differences in the RR and PR intervals may represent intrinsic abnormalities in the pacemaking activities in SA and AV nodal cells and/or the His-Purkinje system, respectively in genetically targeted mice. However, alternative possibilities include altered autonomic input into the SAN and AVN resulting from the over-expression or deletion of the SK2 channel. To rule out this latter possibility, we recorded ECG in the WT and mutant mice using combined intra-peritoneal injection of atropine (1mg/kg) and propanolol (20 mg/kg), concentration previously shown to abolish autonomic control of the heart.^21,22 Results are presented in Figure 1D. Administration of atropine and propanolol resulted in the prolongation of the PR intervals in the SK2 +/T and SK2 +/Δ mice as well as WT animals. More importantly, significant abnormalities in the AVN in SK2 +/T and SK2 +/Δ animals remain after treatment with atropine and propanolol compared to the WT controls, suggesting that the defects observed are intrinsic to the pacemaking tissues. On the other hand, the prolongation of the PR interval may be related to either AV node and/or the His-Purkinje system. Moreover, it is well known that the specific conductance-mediated influences on spontaneous pacemaker activity are highly rate-dependent. In order to further assess the effects of SK2 channels on AV nodes at constant rates, we performed in vivo electrophysiologic studies as described in Online Data Supplement. Figure S1 in Online Data Supplement shows the prolongation of the AV node refractory period in SK2 +/Δ and SK2 Δ/Δ compared to wild-type littermates consistent with the prolongation of the PR interval as described above.

Figure 1.

Surface ECG recordings from WT, SK2+/T and SK2+/Δ mice. (A) Examples of ECG recordings in WT, SK2+/T and SK2+/Δ mice. (B) Examples of ECG recordings in SK2 Δ/Δ mice showing complete AV block with AV dissociation. (C) Summary data for PR intervals in WT, SK2+/T and SK2+/Δ mice in control conditions. There were also significant differences in the RR interval among the three groups of animals (135.0±7.8, 109.7±6.1 and 153.8±6.1 ms for WT, SK2 +/T and SK2 +/Δ, respectively. (D) Summary data for PR intervals after intraperitoneal injection of atropine and propranolol to abolish autonomic control of the heart (n=6, *p<0.05). There were no significant differences in the percent increase in PR intervals after autonomic blockade in WT, SK2+/T and SK2+/Δ mice (Lower Panel in D).

To further document whether there are significant alteration within the AVN among the different genetically-altered mouse models, we evaluated the spontaneous AP characteristics from isolated AV node preparations.

AVNs isolated from SK2 +/T mice show an increase in the rate of firing while the opposite findings were observed in SK2 +/Δ mice

Spontaneous APs were recorded from isolated intact AVN preparation using microelectrode techniques at 33-34°C. Figure 2A shows a photomicrograph of an isolated mouse AVN illustrating the important landmarks used in the identification of the AVN region as previously described.¹ Representative spontaneous APs recorded from the regions within the AVN are shown in Panel B comparing WT, SK2 +/T and SK2 +/Δ. Specifically, APs recorded from within the AVN can be identified by the presence of the slow diastolic depolarization and a very slow upstroke of phase 0. SK2 +/T mice show a significant increase in the spontaneous activities of the AVNs compared to age-matched WT controls (Figure 2B&C). In contrast, SK2+/Δ mice show a significant decrease in the firing frequency of the AVNs compared to the age-matched WT controls. Indeed, data obtained from isolated AVN preparations are consistent with the findings from in vivo measurement as presented in Figure 1 further supporting the notion that the abnormalities observed are intrinsic to the AVN. Moreover, application of apamin (500 pM) resulted in a significant decrease in the frequency of firing consistent with data obtained from the SK2 +/Δ mice (see Figure S2 in Online Data Supplement).

Figure 2C shows summary data of cycle-length (CL) in ms, the maximum diastolic potential (MDP), AP amplitude (APA), maximum upstroke velocity (V_max), rate of diastolic depolarization (DDR), AP duration at 50 and 80% repolarization (APD₅₀, APD₈₀). Detailed analysis of the spontaneous AP reveals significant changes in the CL, APD as well as DDR in the SK2 +/T and SK2 +/Δ compared to the age-matched WT control. Over-expression of the SK2 channel in SK2 +/T mice results in a significant shortening of the APD₈₀ compared to WT, while APD₅₀ and APD₈₀ were significantly prolonged in SK2 +/Δ mice compared to WT mice. Moreover, SK2 +/T mice show a significant increase in the DDR and a corresponding decrease in the CL. The opposite effects were observed in the SK2 knock-out mice. There was also a decrease in V_max in the SK2 +/Δ mice.

Whole-cell I_K,Ca recorded from isolated AVN cells

In order to document that there are differences in the expression of the SK2 channels in SK2 +/T and SK2 +/Δ mice, we directly assessed the SK2 current density measured as apamin-sensitive I_K,Ca in single isolated AVN cells from transgenic and knock-out animals compared to age-matched WT controls. Shown in Figure 3A are examples of the whole-cell current density elicited from a holding potential of −55 mV to various voltage-steps as shown at baseline and after application of apamin (500 pM). Apamin-sensitive currents were obtained using digital subtraction and are shown to the right. Summary data in Panel B show a significant increase in the apamin-sensitive current density in AVN cells isolated from SK2 +/T and a significant decrease in the current density in the SK2 +/Δ mice compared to WT littermates.

Figure 3.

Whole-cell I_K,Ca density elicited using a holding potential of -55 mV in single AVN cells. I_K,Ca current density was obtained using the difference current before and after application of apamin (500 pmol/L) normalized to the cell capacitance. (A) Representative examples of whole-cell I_K,Ca recorded from single isolated AVN cells from WT, SK2 +/T and SK2 +/Δ mice. (B) Summary data for the current density-voltage relations of the apamin-sensitive current in AVN cells from WT (11.7±1.3 pF), SK2 +/T (10.6±2.0 pF) and SK2 +/Δ mice (13.9±1.6 pF). *p<0.05, n=6-9 cells for each group from 3 set of animals.

Immunohistochemistry and Immunofluorence Confocal Microscopy

We performed immunofluorescence confocal laser scanning microscopy using isolated single AVN cells to document the expression of the SK2 channel in AVN as shown in Figure 4. Figure 4A shows bright-field images of single isolated AVN cells (a&b) compared to atrial (c) and ventricular myocytes (d). In Figure 4B and D, Ca_v3.1 and NF 160 were further used as specific AVN cell markers. It has previously been demonstrated that neurofilament (NF)-160 is a marker of the pacemaker and conduction system.^23-25 Ca_v3.1 was only expressed in AVN cells (Figure 4B, panel a) and not atrial or ventricular myocytes (panels c&d) similar to previously published literature.⁵ Figure 4C&D illustrate positive SK2 staining in AVN (Panels a) as well as atrial (Panels d) and ventricular myocytes (Panels e). In contrast, NF 160 was only expressed in AVN (Figure 4D, Panel a) further documenting the specific cell types used in our study. To further document that the specificity of the anti-SK2 antibody used in our study, we performed control experiments as shown in Figure 4C, D Panels b&c. Anti-SK2 antibody was pre-incubated with antigenic peptide which eliminated the positive staining (Panels b). Panels c shows lack of staining with secondary antibodies only. Finally, experiments were repeated using homozygous null mutant mice (SK2 Δ/Δ) showing lack of staining in the atrial myocytes (Figure 4C, Panel f).

Figure 4.

Subcellular distribution of SK2 channel in mouse AVN cells. (A) Photomicrographs of single isolated AVN cells (a&b). Atrial and ventricular myocytes are shown for comparison in the right panels (c&d, respectively). (B) Photomicrographs of confocal laser scanning microscopy of AVN cells (a) showing positive immunostaining with anti-Ca_v3.1 antibody followed by anti-rabbit IgG-FITC conjugated secondary antibody (green), (b) treatment with secondary antibody only as negative control and (c&d) absence of Ca_v3.1 staining in atrial and ventricular myocytes, respectively. The corresponding differential interference contrast (DIC) images are shown in the right panels. (C) Photomicrographs of confocal laser scanning microscopy of immunostaining of AVN cells, atrial and ventricular myocytes with anti-SK2 and anti-α-actinin2 antibodies: (a) double staining with anti-SK2 (green) and anti-α-actinin2 antibodies (red) in an AVN cell, (b) preincubation of the anti-SK2 antibody with antigenic peptide, (c) treatment with secondary antibodies only (anti-rabbit IgG-FITC conjugated and anti-mouse IgG-Texas Red conjugated antibodies) as additional negative control, (d&e) double staining with anti-SK2 (green) and anti-α-actinin2 antibodies (red) in single atrial and ventricular myocytes, respectively. (f) double staining with anti-SK2 and anti-α-actinin2 antibodies in atrial myocytes from SK2 Δ/Δ mice were used as a negative control showing lack of positive staining in SK2 Δ/Δ for SK2 channel. Merged images are shown in the right panels. (D) Photomicrographs of confocal laser scanning microscopy as in C except that anti-neurofilament 160 (NF 160) antibody (red) was used instead of anti-α-actinin2 antibody in AVN cells, atrial and ventricular myocytes. NF 160 was used as an AVN marker. Scale bars =10 μm.

Immunohistochemistry of Histologic Section Through Mouse AVN

Expression of SK2 channel protein in the mouse AVN was further evaluated using immunostaining of cardiac section through AVN. Figure 5A,A’ show photomicrographs of histologic sections of AVN and working myocardium from WT mice at low and high magnification using hematoxylin and eosin. Masson's trichrome (Panels B,B’) was used to stain fibrous tissue to view the AVN demonstrating an oval of compact node cells next to the central fibrous body and the ventricular septum. Expression of SK2 protein in the AVN and working myocardium in WT animals were documented as perixidase-positive staining (Panels C,C’). Absence of SK2 protein expression in the AVN and working myocardium in SK2 Δ/Δ mice is shown in Panels D,D’. Treatment with secondary antibody only was used as negative control (Panels E,E’).

Figure 5.

Expression of SK2 channel protein in AVN. Photomicrographs of histologic sections of AVN and working myocardium from WT and SK2 Δ/Δ mice at low (A,B,C,D,E) and high magnification (A′,B′,C′,D′,E′). (A,A′) Sections stained with hematoxylin and eosin. (B,B′) Sections stained with Masson's trichrome. (C,C′) Expression of SK2 protein (peroxidase positive) in the AVN and working myocardiumin in WT animals. (D,D′) Absence of SK2 protein expression in the AVN and working myocardium in SK2 Δ/Δ mice. (E,E′) Treatment with secondary antibody only as negative control. AVN, atrioventricular node; Arrows indicate AVN/HB region; RA, right atrium; VS, ventricular septum. Scale Bars=50 μm.

DISCUSSION

The two gene-targeted mouse models used in our study with over-expression and knockout of SK2 channel offer us the unique opportunities to directly test the functional roles of SK2 channel in the whole animals, isolated AVN preparations as well as single isolated AVN cells. Moreover, the two mouse models were backcrossed onto the same genetic background allowing for the direct comparison between these two models. Indeed, we documented robust phenotypes of SAN and AVN abnormalities in the two mouse models consistent with a critical role of SK2 channel in the pacemaking activities of both SAN and AVN. Here, we have focused our efforts on the AVN and have documented the existence of SK2 channel in AVN cells using multiple techniques. Over-expression of SK2 channels results in the the shortening of the APs of the AVN cells associated with an increase in the firing frequency. On the other hand, SK2 channel knockout results in the opposite effects on the spontaneous APs of the AVN. Moreover, SK2 +/T mice show a significant increase in the DDR and a corresponding decrease in the CL. The opposite effects were observed in the SK2 knock-out mice. There was also a decrease in V_max in the SK2 +/Δ mice and the finding may result from changes in the Ca²⁺ current. Indeed, we have previously documented that SK2 channels are functionally coupled to L-type Ca²⁺ channels.²⁶ However, detailed analyses of the Ca²⁺ current in these mouse models are beyond the scope of the present study.

AVN Electrophysiology

Since the first description of the anatomical AVN by Tawara in 1906,²⁷ a large number of studies have provided insights into the heterogeneity of the structure as well as a repertoire of ion channel proteins which govern this complex conduction pathway between the atria and ventricles^1-5,28-32. These studies have revealed the intricate organization of multiple nodal and nodal-like myocytes contributing to the unique electrophysiology of the AVN in health and diseases including slow conduction and dual pathways of the AVN. Indeed, reentrant tachycardia within the AVN represents the most common type of paroxysmal supraventricular tachycardia in patient population. In addition, previous studies have documented several distinct ion channel proteins which contribute to the pacemaking nature of the AVN as well as distribution of the specific connexin within the AVN structure.^1-6,30,31 For example, we and others, have previously documented the important roles of Ca_v1.3, in addition to Ca_v1.2 isoform of the L-type Ca²⁺ channels in pacemaking tissues.^4,17,22 In addition, disruption of the gene encoding for Ca_v3.1 abolishes T-type Ca²⁺ current in SAN and AVN in mice. On the other hand, no T-type Ca²⁺ current was detected in the atria of the wild-type or knock-out animals.⁵

Presence of K_Ca Channels in the Heart

Even though the diversity of the Ca²⁺-independent voltage-gated K⁺ channels in cardiac myocytes has been well documented, there has been some uncertainty with regards to the existence of K_Ca channels in the hearts. In general, to ensure stable recording conditions, whole-cell currents are recorded using relatively high concentration of Ca²⁺ chelators, which would mask the Ca²⁺-activated currents. Nonetheless, the presence of I_K,Ca has previously been suggested in rabbit myocytes.³³ On the other hand, one previous study has argued that apamin at a high concentration (50 nmole/L) may have an irreversible effect on the delayed rectifier K⁺ current (I_K,s) in guinea pig ventricular myocytes.³⁴ To resolve some of these uncertainty, we have previously used a combination of techniques to demonstrate the presence of several isoforms of SK channels in human and mouse cardiac myocytes.^7,8 Moreover, we have documented the important functional roles of the SK channels in the heart as well as the more prominent expression in atria compared to ventricular tissues.^7,8 However, the functional roles of the SK channels in pacemaking tissues have not been investigated to date.

SK channels have been shown to play an important role in setting the tonic firing frequency of neurons.³⁵ Their activation causes membrane hyperpolarization, which inhibits cell firing and limits the frequency of repetitive APs. The increase in intracellular Ca²⁺ evoked by AP firing decays slowly, allowing SK channel activation to generate a long lasting hyperpolarization, termed the slow afterhyperpolarization. This spike-frequency adaptation protects the cell from the deleterious effects of continuous tetanic activity. In contrast to the hyperpolarization effects of SK channels in neuron, in atrial myocytes, the current contributes markedly toward the late phase of the cardiac repolarization.^7,8 Nonetheless, contrary to the atrial myocytes, the maximum diastolic potential in AV nodes is ~−60 mV and a significant hyperpolarization may be expected. Therefore, the observed action potential shortening with no associated changes in the maximum diastolic potentials in the SK2 +/T model may not be due entirely to modulation of a Ca²⁺-activated K⁺ current. Additional studies are required to further assess possible changes in other ionic currents which may contribute to the above findings.

Pharmacology of SK Channels

The bee venom peptide toxin, apamin, is an 18-amino acid peptide with two internal disulfide bridges.³⁶ Two residues, an aspartic acid and an asparagine, that reside on opposite sides of the deep pore have been shown to be essential for apamin sensitivity. No other class of K⁺ channels is blocked by this drug, and among the cloned K⁺ channels, the residues that endow sensitivity are present at those positions only in SK2 and SK3 channels. These data suggest that apamin is a very specific blocker for SK channels and that the SK channels may represent the sole class of apamin receptors in the body.³⁷ In our voltage-clamp experiments, a relatively low dose of apamin (500 pM) was used to ensure the specificity of the toxin in our study.

Physiological Significance and Implication for Human Diseases

Here, we demonstrate that over-expression of SK2 channel in the heart increases the frequency of firing in pacemaking tissues. Therefore, in cardiovascular system, SK2 channels may serve a distinct role to potentiate the increase in AVN conduction during exercise or heightened sympathetic responses under normal physiologic conditions. On the other hand, an increase in intracellular Ca²⁺ under certain pathologic conditions may produce profound changes in AVN conduction. For example, during atrial fibrillation, the rapid depolarization may increase intracellular Ca²⁺ and potentiate the I_K,Ca and AVN conduction. Hence, SK2 channel may represent an attractive target to modulate atrial conduction during atrial arrhythmias.

Limitations

Even though the gene-targeted mouse models offer unique advantage for our study to directly probe the functional roles of SK2 channel in AVN, we do recognize the fact that there are well documented differences between small animal models and human.^38,39 Additional studies will need to be carried out using large animal models.