Jeanne M. Nerbonne

Chapter Outline

Summary

Electrophysiologic studies on isolated mammalian cardiac myocytes have identified the presence of multiple types of voltage-gated K⁺ (Kv) channel currents with distinct time- and voltage-dependent properties that contribute to determining action potential amplitudes, waveforms and durations. A number of non–voltage-gated, inwardly rectifying K⁺ (Kir) channels that contribute to action potential repolarization have also been functionally identified. The various types of Kv and Kir channels are differentially expressed in different cardiac cell types, contributing to cellular and regional differences in action potential waveforms and refractoriness. In the diseased myocardium, K⁺ current remodeling is evident and correlated with changes in action potential waveforms and durations, increased dispersion of repolarization and the development of substrates for reentrant arrhythmias. The cloning of Kv and Kir channel pore-forming (α) and accessory (β) subunits, as well as the subsequently identified two pore domain K⁺ (K2P) channel and small conductance Ca²⁺-dependent K⁺ (SK) channel subunits, has provided insights into the molecular basis of functional myocardial K⁺ channel diversity. In addition, a variety of experimental approaches have been, and continue to be, used to define the molecular determinants of native cardiac K⁺ channels and facilitated efforts focused on exploring the molecular mechanisms controlling the properties and the functional cell surface expression of these channels in the normal and in the diseased myocardium.

In atrial and ventricular myocytes, the action potential upstroke, attributed to inward currents through voltage-gated Na⁺ (Nav) channels, is rapid (Figure 3-1). In nodal tissues, which lack functional Nav channels, the upstroke of the action potential is substantially slower and is dominated by Ca²⁺ influx through voltage-gated Ca²⁺ (Cav) channels. There are also marked regional differences in action potential heights and durations, as well as in the time courses of action potential repolarization (see Figure 3-1). These differences, which affect the normal spread of excitation in the myocardium and influence the dispersion of repolarization in the ventricles, primarily reflect regional differences in the functional expression and the properties of the outward K⁺, as well as the inward (Nav and Cav) currents.¹

Cellular electrophysiologic studies have detailed the properties of the major inward (Nav and Cav) and outward (K⁺) currents that shape the waveforms of atrial and ventricular action potentials (Figure 3-2). In contrast to the cardiac Nav and Cav currents, there are multiple types of myocardial voltage-gated K⁺ (Kv) and non-voltage, inwardly rectifying K⁺ (Kir) channels (Table 3-1), many of which are differentially expressed, contributing to regional variations in myocardial action potential waveforms (see Figure 3-1) and refractoriness.^1–3 In addition, changes in the densities, distributions, and properties of Kv and Kir channels are evident in a variety of myocardial diseases, and these changes affect repolarization, influence propagation and decrease rhythmicity, effects that can produce substrates for the generation of life-threatening arrhythmias.¹ There is, therefore, considerable interest in defining the molecular mechanisms controlling the biophysical properties and the functional cell surface expression of these channels. A large number of Kv and Kir pore-forming α and accessory β (Table 3-2) subunits have been identified,^4,5 and considerable progress has been made in defining the relationships between these subunits and functional myocardial Kv and Kir channels.^1,6,7 Importantly, the studies completed to date have revealed that the molecular correlates of the various types of Kv and Kir channels distinguished electrophysiologically (see Table 3-1) are indeed distinct.¹ In the long term, defining the molecular compositions of myocardial Kv and Kir channels will also facilitate studies aimed at determining the molecular mechanisms controlling the marked regional differences in the expression of these channels in the normal myocardium, as well as the derangements in the expression or functioning of these channels that occur with myocardial disease. In this chapter, the electrophysiologic and molecular diversity of repolarizing myocardial Kv and Kir channels and the molecular determinants of native myocardial K⁺ channels will be reviewed.

Myocardial Voltage-Gated K⁺ Channels: Transient Outward Voltage-Gated K⁺ Channels

Kv currents, activated on membrane depolarization, influence myocardial action potential amplitudes and durations and, in most cells, two broad classes of Kv currents have been distinguished: transient outward K⁺ currents, I_to; and delayed, outwardly rectifying K⁺ currents, I_K (see Table 3-1). The transient currents (I_to) activate rapidly and underlie early (phase 1) repolarization, whereas the delayed rectifiers (I_K) determine the latter phase (phase 3) of action potential repolarization (see Figure 3-2) back to the resting membrane potential. These classifications are broad, however, and there are actually multiple types of transient (I_to) and delayed rectifier (I_K) Kv currents (see Table 3-1) expressed in cardiac cells.^1,3 Electrophysiologic and pharmacologic studies, for example, have clearly demonstrated that there are two types of transient outward K⁺ currents, now referred to as I_to,fast (I_to,f) and I_to,slow (I_to,s), and that these currents are differentially distributed.^6,7 The rapidly activating and inactivating transient outward K⁺ current, I_to,f, is also characterized by rapid recovery from steady-state inactivation, whereas I_to,s recovers slowly from inactivation.^6,7 In addition, I_to,f is readily distinguished from other Kv currents (including I_to,s) using Heteropoda toxin-2 or -3 (see Table 3-1).

Although originally identified in Purkinje fibers, I_to,f is a prominent repolarizing Kv current in atrial and ventricular myocytes, as well as in nodal cells, in most species.¹ There are, however, marked regional differences in I_to,f densities, with the highest densities typically in atrial myocytes. In addition, in mammalian ventricles, I_to,f and I_to,s are differentially distributed. In canine left ventricles (LVs), for example, I_to,f density is fivefold to sixfold higher in epicardial and midmyocardial, than in endocardial, cells.² There are also marked regional differences in I_to,f densities in adult mouse ventricles.^8–10 Specifically, I_to,f density is higher in right than in left ventricular myocytes, and within the LV, I_to,f densities are higher in apex than in base myocytes.^8–10 In the mouse, even greater Kv current heterogeneity is seen in cells isolated from the septum: all ventricular septum cells express I_to,s, and most (approximately 80%) also express I_to,f.⁸ When present, however, I_to,f density is significantly (p < 0.001) lower in septum than in RV or LV cells.⁸ I_to,f and I_to,s are also differentially expressed in ferret LV, and I_to,s is detected only in endocardial LV cells.¹¹ Despite heterogeneities in functional expression, the properties of I_to,f and I_to,s in different cardiac cell types (and species) are remarkably similar, leading to suggestions that the molecular compositions of the underlying (I_to,f and I_to,s) channels are also similar.³

Myocardial Voltage-Gated K⁺ Channels: Delayed Rectifier Voltage-Gated K⁺ Channels

Electrophysiologic and pharmacologic studies have also distinguished multiple types of cardiac delayed rectifier K⁺ currents, I_K (see Table 3-1). In atrial myocytes, for example, the dominant repolarizing K⁺ current is a rapidly activating, non-inactivating K⁺ current, I_Kur (I_K,ultrarapid), which is not detected in ventricular or nodal cells.¹ In most ventricular myocytes, there are two prominent components of delayed rectification, I_Kr (I_K,rapid) and I_Ks (I_K,slow), that are different from I_Kur in terms of time- and voltage-dependent properties.¹ The biophysical properties of I_Kr and I_Ks are distinct: I_Kr activates and inactivates rapidly, displays marked inward rectification, and is selectively blocked by class III anti-arrhythmics, including dofetilide and sotalol.¹² In contrast, I_Ks activates slowly and does not display inward rectification.¹²

Similar to the transient outward K⁺ currents, there are also marked regional differences in the functional expression of I_Ks and I_Kr in mammalian ventricular myocytes.1,2 The density of I_Ks in canine LV, for example, is higher in epicardial and endocardial than in M cells.² There are also regional differences in the functional expression of I_Kr and I_Ks channels in guinea pig LV.¹³ In cells isolated from the LV free wall, for example, the density of I_Kr is higher in subepicardial than in midmyocardial or subendocardial myocytes.¹³ At the base of the LV, in contrast, I_Kr and I_Ks densities are significantly lower in endocardial, than in either midmyocardial or epicardial, LV cells.¹³ Differences in Kv current densities contribute to the variations in action potential waveforms recorded in different regions (i.e., atria and ventricles, right and left ventricles, apex and base of the ventricles) of the heart, as well as in different layers (epicardial, midmyocardial, and endocardial) of the LV and RV walls.^1–3,8–13

In rodent ventricles, there are additional components of I_K with properties distinct from I_Ks and I_Kr (see Table 3-1). In mouse and rat ventricular myocytes, for example, there are novel delayed rectifier Kv currents that have been referred to as I_K, I_Kslow and I_ss (see Table 3-1).^1,3 Mouse ventricular I_K,slow was first identified as a rapidly activating and slowly inactivating K⁺ current with properties distinct from I_to,f, I_to,s and I_ss expressed in the same cells.¹⁴ In addition, I_K,slow was shown to be blocked selectively by micromolar concentrations of 4-aminopyridine (4-AP), which does not affect I_to,f or I_to,s.¹⁴ Subsequent work, however, revealed the presence of two components of mouse ventricular I_K,slow: I_K,slow1, which is blocked by µM 4-AP; and, I_K,slow2, which is blocked selectively by TEA.^15–18 In addition, it has been demonstrated that I_K,slow1 and I_K,slow2 reflect the expression of distinct molecular entities.^15–18 In contrast to the differential distribution of I_to,f and I_to,s, however, I_K,slow1, I_K,slow2, and I_ss appear to be uniformly expressed in mouse atrial and ventricular myocytes.^{8–10,15–19}

Inwardly Rectifying Myocardial K⁺ Channels Also Contribute to Repolarization

In addition to the Kv currents, Kir currents, specifically I_K1 and the ATP-dependent K⁺ current, I_KATP (Table 3-1), contribute to shaping myocardial action potential waveforms.^20–22 Similar to the Kv currents, the densities of the Kir currents vary in different regions of the heart (e.g., atria, ventricles, conducting tissue).^1,3 In contrast to the Kv currents, however, myocardial Kir current densities are similar in myocytes in different regions of the ventricles.¹ In mammalian atrial and ventricular myocytes, I_K1 plays a role in establishing resting membrane potentials and plateau potentials, and contributes to phase 3 repolarization (see Figure 3-2). The fact that the conductances of I_K1 channels are high at negative membrane potentials underlies the contribution of I_K1 to atrial and ventricular resting membrane potentials.²⁰ Although the voltage-dependent properties of I_K1 channels are such that conductance is low at potentials positive to approximately −40 mV, these channels nevertheless contribute outward K⁺ currents during the plateau phase of the action potential in ventricular cells,²⁰ as well as during phase 3 repolarization (see Figure 3-2), because the driving force on K⁺ is high at depolarized membrane potentials.

Myocardial ATP-dependent K⁺ channels are weak, inwardly rectifying channels that are inhibited by (elevated) intracellular ATP and activated by nucleotide diphosphates.²¹ In ventricular myocytes, activation of I_KATP channels during periods of hypoxia and ischemia results in action potential shortening, suggesting that these channels provide a link between cellular metabolism and membrane potential.^21,22 The opening of I_KATP channels appears to contribute to the cardioprotection resulting from ischemic preconditioning.²² In contrast with some ventricular Kv channels, I_KATP channels appear to be distributed uniformly in the right and left ventricles and through the thickness of the right or left ventricular walls. Interestingly, I_KATP channels are expressed at much higher densities than other sarcolemmal myocardial K⁺ channels,²¹ suggesting that action potentials will be shortened markedly when only a few I_KATP channels are activated.

Pore-Forming (α) Subunits of Myocardial Voltage-Gated K⁺ Channels

Kv channel pore-forming (α) subunits are six transmembrane-spanning domain proteins (Figure 3-3) with a region between the fifth and sixth transmembrane domains that contributes to the K⁺-selective pore.⁴ The positively charged fourth transmembrane domain in the Kv α-subunits (see Figure 3-3) is homologous to the corresponding regions in Nav and Cav channel α-subunits, placing them in the S4 superfamily of voltage-gated channels.^3,4 In contrast to Nav and Cav channels, in which only a single α-subunit is required to form a channel, functional Kv channels comprise four α-subunits (see Figure 3-3). Similar to the diversity of functional myocardial Kv channels (see Table 3-1), however, multiple Kv α-subunits have been identified. These subunits comprise several homologous Kv α-subunit subfamilies, Kv1.x, Kv2.x, Kv3.x, Kv4.x, Kv10.x, Kv11.x, and many members of these Kv α-subunit subfamilies are expressed in the mammalian heart.¹ In addition, further functional Kv channel diversity could arise through alternative splicing of transcripts and through the formation of heteromultimeric channels between two or more Kv α-subunit proteins in the same Kv α-subunit subfamily.^1,3,4

Additional subfamilies of Kv α-subunits were revealed with the cloning of the human eag-related (HERG) gene, KCNH2, subsequently identified as the locus of mutations underlying one form of familial long QT-syndrome, LQT2, and KCNQ1 (KvLQT1), the locus of mutations in another inherited long QT syndrome, LQT1.²³ Heterologous expression of KCNH2 (ERG1) reveals inwardly rectifying Kv currents²³ with properties similar to cardiac I_Kr (see Table 3-1). Although there are several ERG (KCNH) subfamily members, only KCNH2 (which encodes ERG1) appears to be expressed in the myocardium.¹ Heterologous expression of KCNQ1 (KvLQT1) alone reveals rapidly activating and non-inactivating Kv currents, whereas co-expression with the Kv accessory subunit, minK (see Table 3-2), produces slowly activating Kv currents that resemble the slow component of cardiac-delayed rectification, I_Ks.^1,3,23

Accessory/Auxiliary Subunits of Myocardial Voltage-Gated K⁺ Channels

In addition to the Kv α-subunits, a number of Kv channel accessory (Kv β) subunits have also been identified (see Table 3-2). The first of these subunits was KCNE1, which encodes a small (130-aa) protein (minK) with a single transmembrane spanning domain.²⁴ It appears that minK coassembles with KvLQT1 to form functional cardiac I_Ks channels.^23,24 Additional minK homologs, MiRP1 (KCNE2), MiRP2 (KCNE3), and MiRP3 (KCNE4) have also been identified (see Table 3-2), and it has been suggested that MiRP1 (KCNE2) functions as an accessory subunit coassembling with ERG1 to generate cardiac I_Kr.²⁵ It has also been reported that the MiRP subunits interact with multiple Kv α-subunit subfamilies and modify channel properties.²⁴ MiRP2, for example, coassembles with Kv3.4 in mammalian skeletal muscle,²⁶ and MiRP1 coassembles with Kv4.x α-subunits when coexpressed in heterologous cells.²⁷ These observations suggest that the MiRP (KCNE) accessory subunits can assemble with a variety of Kv α-subunits and contribute to the formation of multiple types of myocardial Kv channels. Direct experimental support for this hypothesis, however, has not been provided, and the roles of the various KCNE subunits in the generation of functional cardiac Kv^26,27 and other non-Kv²⁸ channels remain to be defined.

Another type of Kv channel accessory subunit was revealed with the isolation of low molecular weight (approximately 45 kD) cytosolic (Kvβ) subunits from brain.⁵ Three homologous Kv β-subunits, Kvβ1, Kvβ2, and Kvβ3 (see Table 3-2), as well as alternatively spliced transcripts, have been identified, and both Kvβ1 and Kvβ2 are expressed in heart.¹ Previous studies have shown that Kvβ subunits interact with the intracellular domains of Kv1 α-subunits, and functional studies demonstrated that Kvβ subunits affect the biophysical properties and increase the cell surface expression of heterologously expressed Kv1 α-subunit–encoded Kv currents.⁵ Because Kv α- and β-subunits appear to coassemble in the endoplasmic reticulum,²⁹ the increase in functional channel expression suggests that the Kv β-subunits affect channel assembly, processing or stability or, alternatively, function as chaperone proteins. Although Kvβ1 and Kvβ2 have been shown to associate with Kv4 subunits in the mouse myocardium and the targeted deletion of Kvβ1 results in the attenuation of Kv4-encoded I_to,f in mouse ventricular myocytes,³⁰ the roles of Kvβ subunits in the generation of myocardial Kv channels in other species have not been explored.

A yeast two-hybrid screen, using the N terminus of Kv4.2 as the bait, led to the identification of accessory Kv channel interacting proteins, (KChIPs) (see Table 3-2) in brain.³¹ Of the four KChIPs identified,^31,32 only KChIP2 is expressed in heart,^31,33 although there are several KChIP2 splice variants.^34–36 Interestingly, the KChIPs belong to the recovering family of neuronal Ca²⁺-sensing (NCS) proteins,³⁷ and contain four EF-hand domains.³¹ When coexpressed with Kv4 α-subunits, each of the KChIPs increases K⁺ current densities, slows inactivation, speeds recovery from inactivation, and shifts the voltage dependence of current activation.³¹ In contrast, KChIP coexpression does not affect the properties or the densities of Kv1.4- or Kv2.1-encoded K⁺ currents, consistent with the suggestion that the modulatory effects of the KChIPs are specific for Kv4 α-subunit–encoded Kv channels.³¹ More recent studies, however, suggest that the KChIPs modulate the functional cell surface expression of Kv1.5-encoded Kv channels,³⁸ as well as myocardial Cav channels.³⁹ In addition, although KChIP binding to Kv4 α-subunits is not Ca²⁺ dependent, mutations in EF hand domains 2, 3, and 4 eliminate the modulatory effects of KChIP1 on Kv4- and Kv1.5-encoded Kv currents,^31,38 suggesting a role for voltage-dependent Ca²⁺ entry and intracellular Ca²⁺ levels in the regulation of functional cardiac (Kv4-encoded) I_to,f channels, as has been demonstrated for neuronal Kv4-encoded channels.⁴⁰

The transmembrane diaminopeptidyl transferase-like protein 6 (DPP6 or DPPX) has also been suggested as an accessory subunit of cardiac⁴¹ and neuronal⁴² Kv4-encoded channels (see Table 3-2). Lacking enzymatic activity, DPP6 increases the cell surface expression of Kv4 α-subunits, shifts the voltage dependences of activation and inactivation of heterologously expressed Kv4 currents to more negative potentials, and accelerates the rates of current activation, inactivation, and recovery.^41,42 Interestingly, heterologous coexpression of DPP6 with Kv4.3 and KChIP2 produces Kv currents that closely resemble native cardiac I_to,f.⁴¹ Another member of the DPP-like subfamily of proteins, DPP10, has been demonstrated to associate with Kv4.2 and KChIP3 in the rat brain and to have regulatory effects similar to DPP6 on heterologously expressed Kv4 (with and without KChIP) channels.^43,44 In addition, DPP10 has been shown to be expressed in normal and failing human ventricles.⁴⁵

Molecular Determinants of Native Myocardial Transient Outward Voltage-Gated K⁺ Channels

Considerable experimental evidence has accumulated demonstrating a critical role for Kv α-subunits of the Kv4 subfamily in the generation of cardiac I_to,f channels. In rat and mouse ventricular myocytes exposed to antisense oligodeoxynucleotides targeted against Kv4.2 or Kv4.3, I_to,f density is reduced by approximately 50%.46,47 Rat ventricular I_to,f density is also reduced in cells exposed to an adenoviral construct encoding a truncated Kv4.2 subunit (Kv4.2ST) that functions as a dominant negative.⁴⁸ In addition, I_to,f is eliminated in ventricular and in atrial myocytes isolated from transgenic mice expressing a pore mutant of Kv4.2, Kv4.2W362F (Kv4.2DN), which also functions as a dominant negative.^49,50 Although biochemical and electrophysiologic studies suggested that Kv4.2 and Kv4.3 are associated in adult mouse ventricles and that functional mouse ventricular I_to,f channels are heteromeric,⁴⁷ targeted deletion of Kv4.2 eliminates mouse ventricular I_to,f,⁵¹ whereas elimination of Kv4.3 has no effect.⁵² In mouse ventricles, therefore, Kv4.2 is the critical α-subunit required for the generation of functional I_to,f channels.⁵¹ Given the similarities in the properties of I_to,f (see Table 3-1), it seems reasonable to suggest that Kv4 α-subunits also underlie I_to,f in other species. In canine and in human myocardium, however, the candidate subunit is Kv4.3 because Kv4.2 is barely detectable.⁵³ Although two splice variants of Kv4.3 have been identified,⁵⁴ the expression levels of the two predicted Kv4.3 proteins and the functional roles of these variants in the generation of functional cardiac I_to,f channels have been determined.

It has also been demonstrated that the Kv channel accessory subunit KChIP2 coimmunoprecipitates with Kv4 α-subunits from adult mouse ventricles, which is consistent with a role for this subunit in the generation of Kv4-encoded mouse ventricular I_to,f channels.⁴⁷ In ferret and canine hearts, a gradient in KChIP2 message expression is observed through the thickness of the ventricular wall,^35,36 leading to suggestions that the differential expression of KChIP2 underlies the epicardial–endocardial differences in I_to,f densities. Subsequent studies revealed that the patterns of expression of the KChIP2 message, the KChIP2 protein, and I_to,f densities in canine ventricles are indeed similar,⁵⁵ which is consistent with an important role for KChIP2 in determining functional canine and human ventricular I_to,f densities. In contrast, in rat and mouse ventricles, there is little or no gradient in KChIP2 message or protein (51,56) expression, and it appears that regional differences in Kv4.2 expression underlie the heterogeneities in I_to,f densities in rodent ventricles.^56,57 Molecular insights into the regulation of the observed regional variations in the expression of the Kv4.2 transcripts were provided with the demonstrations that the expression levels of the transcription factors, Irx5 and NFAT, are positively and negatively correlated, respectively, with the differences in Kv4.2 expression and I_to,f densities.^58,59 Interestingly, approximately 25 transcription factors were subsequently shown to be differentially expressed in the ventricles,⁶⁰ although the functional import of these findings remains to be determined.

A role for the KChIP2 splice variant, KChIP2b, in determining regional differences in ferret ventricular I_to,f densities has also been proposed.³⁵ The KChIP-related (NCS family) protein, NCS-1, appears to be uniformly expressed in adult mouse ventricles and, in addition, coimmunoprecipitates with Kv4 α-subunits.⁶¹ The protein expression pattern of this putative Kv4 channel accessory subunit has not been examined in canine or human hearts, and a role for NCS-1 in controlling functional myocardial I_to,f densities has not been determined directly. Although accumulating evidence suggests that cardiac I_to,f channels function in macromolecular complexes, comprising Kv4 α-subunits and multiple cytosolic and transmembrane accessory subunits (Figure 3-4), the molecular compositions of native myocardial I_to,f channels have not been determined. In addition, the functional roles of additional Kv accessory subunits, including members of the KCNE27⁶² and DPPX41⁴⁵ subfamilies, in the generation of native myocardial I_to,f channels and in determining regional differences in myocardial I_to,f densities remain to be defined.

The kinetic and pharmacologic properties of the slow transient outward K⁺ currents, I_to,s, in ventricular myocytes are different from I_to,f (see Table 3-1), which is interpreted as suggesting that the molecular correlates of ventricular I_to,s and I_to,f channels are also distinct. Direct experimental support for this hypothesis was provided in electrophysiologic experiments on myocytes isolated from mice with a targeted deletion of the Kv1.4 gene, Kv1.4^–/–,⁶³ demonstrating that I_to,s is undetectable in septum cells.⁶⁴ The properties and the densities of I_to,f, I_K,slow1, I_K,slow2, and I_ss in Kv1.4^–/– left and right ventricular (and in atrial) myocytes, however, are indistinguishable from those measured in wild type cells.^64,65 Interestingly, upregulation of I_to,s is evident in the ventricles of Kv4.2DN-expressing mice in which I_to,f is eliminated.⁶⁵ Given the similarities in the time- and voltage-dependent properties of the slow transient outward K⁺ currents in other species with mouse I_to,s (1), it seems reasonable to suggest that Kv1.4 likely also encodes I_to,s in ferret, rabbit, human, and canine ventricular myocytes.

Molecular Determinants of Native Myocardial Delayed Rectifier Voltage-Gated K⁺ Channels

As noted earlier, ERG1 is the locus of mutations in LQT2, and heterologous expression of ERG1 reveals voltage-gated, inwardly rectifying K⁺-selective channels that are similar to cardiac I_Kr.²³ Alternatively processed forms of ERG1, with unique N- and C-termini, have also been identified in mouse and human hearts and suggested to have roles in the generation of native cardiac I_Kr channels.^66–68 It has also been suggested that functional cardiac I_Kr channels are multimeric, comprising ERG1 and minK, and biochemical studies have demonstrated coimmunoprecipitation of ERG1 and minK from equine ventricles.⁶⁹ It is not clear whether ERG1 and minK or other members of the KCNE family are also found in association in other species.

Although heterologous expression of KCNQ1, the locus of mutations in LQT1, reveals rapidly activating, non-inactivating Kv currents, coexpression with minK produces slowly activating Kv currents similar to cardiac I_Ks.²³ These observations, together with biochemical data demonstrating that heterologously expressed KvLQT1 and minK associate, were interpreted as suggesting that minK coassembles with KvLQT1 to form functional cardiac I_Ks channels.^1,23 Direct biochemical evidence for the in situ coassembly of KvLQT1 and minK was recently provided in studies of equine ventricles.⁶⁹ Similar data for human ventricular I_Ks are yet to be provided. Interestingly, it was reported that KvLQT1 modulates the distribution and properties of ERG1-encoded channels, an observation interpreted as suggesting that cardiac I_Ks and I_Kr channels are regulated directly through Kv α-subunit–Kv α-subunit interactions.⁷⁰ The molecular mechanisms controlling the cell surface expression of functional I_Ks channels, the regional differences in functional ventricular I_Ks densities, and the interactions between cardiac I_Ks and I_Kr (and possibly other cardiac channels), however, remain to be determined.

Similar to the transient outward Kv currents, molecular genetic methods, in combination with biochemistry and electrophysiology primarily in mice, have provided molecular insights into the basis of functional delayed rectifier Kv channel diversity in the murine myocardium. A role for Kv1 α-subunits in the generation of mouse ventricular I_K,slow, for example, was revealed with the demonstration that I_K,slow is selectively attenuated in ventricular myocytes isolated from transgenic mice expressing a truncated, dominant negative Kv1 α-subunit, Kv1.1DN.¹⁴ It was subsequently shown, however, that I_K,slow is also reduced in ventricular myocytes expressing a dominant negative Kv 2.1 mutant, Kv2.1DN,¹⁵ revealing that there are two molecularly distinct components of mouse ventricular I_K,slow: I_K,slow1 that is sensitive to micromolar concentrations of 4-AP and encoded by Kv1 α-subunits, and I_K,slow2 that is sensitive to TEA and encoded by Kv2 α-subunits.¹⁵ Subsequent studies revealed that I_K,slow1 is eliminated in ventricular myocytes isolated from mice harboring the targeted disruption of the KCNA5 (Kv1.5) locus, revealing that Kv1.5 encodes I_K,slow1.¹⁶ These findings, together with the previous studies completed on cells from Kv1.4^–/– animals⁶³ in which I_to,s is eliminated,⁶⁴ reveal that, in contrast to the Kv4 α-subunits, Kv4.2, and Kv4.3,⁴⁷ the Kv1 α-subunits, Kv1.4 and Kv1.5, do not associate in adult mouse ventricles in situ. Rather, functional Kv1 α-subunit–encoded Kv channels in mouse ventricular myocytes are homomeric, composed of Kv1.4 α-subunits (I_to,s)⁶³ or Kv1.5 α-subunits (I_K,slow1).¹⁶ The roles of Kv accessory subunits in the generation of functional I_to,s, I_K,slow1 and I_K,slow2 channels and the molecular mechanisms controlling the differential expression of these channels remain to be determined.

Molecular Determinants of Myocardial Kir and K2P Channels

In cardiac and other cells, inwardly rectifying K⁺ (I_K1) channels are encoded by a large and diverse subfamily of inward rectifier K⁺ (Kir) channel pore-forming α-subunit genes,⁴ each of which encodes a protein with two transmembrane domains that assemble as tetramers to form K⁺ selective pores (see Figure 3-3). Based on the properties of the currents produced in heterologous expression systems, Kir2 α-subunits were long thought to encode the strongly inwardly rectifying cardiac I_K1 channels,²⁰ and several members of the Kir2 subfamily are expressed in the myocardium.⁷¹ Direct insights into the roles of Kir2 α-subunits in the generation of cardiac I_K1 channels were provided in studies completed on myocytes isolated from mice lacking KCNJ2 (Kir2.1^–/–) or KCNJ12 (Kir2.2^–/–).⁷² Although Kir2.1^–/– mice have cleft palate and die shortly after birth, precluding electrophysiologic studies on myocytes from adult animals, voltage-clamp recordings from isolated newborn Kir2.1^–/– ventricular myocytes revealed that I_K1 is absent, whereas a quantitative reduction in I_K1 was observed in adult Kir2.2^–/– ventricular myocytes,⁷² suggesting that both Kir2.1 and Kir2.2 contribute to mouse ventricular I_K1 channels and that functional cardiac I_K1 channels are heteromeric. The quantitative differences between the effects of the deletion of KCNJ2 and KCNJ12 further suggest that Kv2.1 (KCNJ2) is the critical subunit underlying (mouse) I_K1 channels.⁷²

Mutations in KCNJ2 have been linked to congenital long QT (Andersen-Tawil syndrome or LQT7) and short QT syndromes,73,74 and increasing expression of Kir2.1 in the mouse heart, which results in the upregulation of I_K1, is proarrhythmic.^75,76 Previous studies have identified regional differences in myocardial I_K1 expression and properties in adult mouse heart,^77,78 and it has been suggested that these differences reflect the variable subunit compositions of the channels, as well as differences in polyamine concentrations.⁷⁹ Studies focused on testing this hypothesis directly and on defining the molecular mechanisms controlling regional differences in the expression and functioning of native I_K1 channels in species other than mice are also clearly warranted.

In the mammalian heart, I_KATP channels appear not to have a prominent role in action potential repolarization, but are thought to be important in myocardial ischemia and preconditioning.21,22 In heterologous systems, I_KATP channels can be reconstituted by coexpression of Kir6.x subunits with one or more ATP-binding cassette proteins that encode sulfonylurea receptors, SURx.⁸⁰ Pharmacologic and mRNA expression studies suggested that cardiac sarcolemmal I_KATP channels are encoded by Kir6.2 and SUR2A, and the essential role of Kir6.2 was demonstrated directly in experiments showing that I_KATP channels are undetectable in ventricular myocytes isolated from Kir6.2^–/– animals.⁸¹ In addition, cardiac I_KATP channel activity is reduced in SUR2^–/– myocytes⁸² and unaffected in SUR1^-/- myocytes,⁸³ suggesting an important role for SUR2. Interestingly, the properties of the residual I_KATP channels in SUR2^–/– myocytes are similar to those produced in heterologous cells on coexpression of Kir6.2 and SUR1,⁸² suggesting that SUR1 also contributes to the generation of functional cardiac I_KATP channels by coassembling with Kir6.2 α-subunits alone or with SUR2A. Biochemical and molecular genetic strategies will need to be combined to define the molecular components of native cardiac I_KATP channels directly.

Although action potential waveforms in Kir6.2^–/– and wild type ventricular myocytes are indistinguishable, the action potential shortening observed in wild type cells during ischemia or metabolic blockade is abolished in the Kir6.2^–/– cells,⁸¹ consistent with the hypothesis that cardiac I_KATP channels have a role under pathophysiologic conditions, particularly those involving metabolic stress. Interestingly, action potential durations are largely unaffected in transgenic animals expressing mutant I_KATP channels with markedly (fortyfold) reduced ATP sensitivity, suggesting that there are additional inhibitory mechanisms that regulate cardiac I_KATP channel activity in vivo.⁸⁴ Similar to the strong inwardly rectifying myocardial Kir currents, I_K1, further study is needed to provide molecular insights into the mechanisms controlling regional differences in the functional expression of myocardial I_KATP channels.

In addition to Kv and Kir channel α-subunits that assemble as tetramers, a novel type of K⁺ α-subunit with four transmembrane spanning regions and two pore domains (see Figure 3-3) was identified with the cloning of TWIK-1.⁸⁵ Both pore domains contribute to the formation of the K⁺ selective pore, and TWIK-1 subunits assemble as dimers.⁸⁶ A large number of two-pore domain K⁺ (K2P) channel α-subunit genes have been identified, and several are expressed in the mammalian myocardium (Table 3-3). Heterologous expression of K2P subunits reveals currents with distinct biophysical properties and sensitivities to several potential intracellular and extracellular modulators, including anesthetics, change in pH, and fatty acids.⁸⁵

The multiplicity of K2P α-subunits (see Table 3-3), the widespread distribution of expressed subunits and the findings that the properties of the channels encoded by K2P subunits are regulated by a variety of physiologically (and pathophysiologically) relevant stimuli suggest that K2P channels likely subserve a variety of important functions. As with other cell types, the physiologic roles of these subunits and channels in the myocardium are just beginning to be explored. Both TREK-1 and TASK-1 are detected in heart, and heterologous expression of these subunits gives rise to instantaneous, non-inactivating K⁺ currents that display little or no voltage dependence.⁸⁵ These properties suggest that K2P subunits might contribute to background or “leak” K⁺ channels (i.e., channels with properties similar to the steady-state, non-inactivating K⁺ current [I_ss] that is expressed or characterized in rodent myocytes).^8,9,85–87 In recent studies completed on rat atrial and ventricular myocytes, roles for TASK1^88–90 and TREK1^86,87 subunits in the generation of I_ss have been suggested. Clearly, experiments focused on testing these hypotheses are needed to provide clear insights into the molecular basis of I_ss and to allow further study focused on defining the mechanisms controlling the physiologic and the pathophysiologic regulation of these (I_ss) channels.

Conclusions

Cellular electrophysiologic studies have distinguished multiple types of voltage-gated inward and outward currents that contribute to action potential repolarization in mammalian cardiac cells (see Table 3-1). The outward (K⁺) currents are more numerous and more diverse than the inward (Na⁺, Ca²⁺) currents, and most cardiac cells express a repertoire of voltage-gated and inwardly rectifying K⁺ channels (see Table 3-1). In addition, several of these K⁺ channels are expressed differentially in different myocardial cell types and through the thickness of the ventricular walls. Molecular cloning led to the identification of multiple voltage-gated (Kv), non–voltage-gated, inwardly rectifying (Kir), and weakly rectifying, non-inactivating (K2P) K⁺ channel pore-forming α-subunits and a number of channel accessory (β) subunits (see Table 3-2) thought to contribute to the formation of the various cardiac K⁺ currents that have been distinguished electrophysiologically (see Table 3-1). In recent years, considerable progress has been made in identifying the pore-forming Kv and Kir α-subunits contributing to the formation of most of the K⁺ channels expressed in mammalian cardiac myocytes (see Figure 3-2). In addition, biochemical studies have provided some insights into the molecular mechanisms underlying the observed heterogeneities in the expression of myocardial Kv and Kir currents. For cardiac I_to,f, regional differences in current densities are correlated with differences in Kv4.2 protein expression in adult mouse ventricles,⁵⁷ whereas variable expression of the I_to,f channel accessory protein, KChIP2, has been suggested as underlying the transmural gradient in I_to,f densities in canine and human ventricles.⁵⁵ For cardiac I_K1 channels, in contrast, recent studies suggest that differences in Kir channel α-subunit composition or differences in the concentrations of intracellular polyamines appear to have roles in regulating the functional diversity of these channels.⁷⁸

Accumulating evidence suggests that native myocardial Kv channels likely function in macromolecular protein complexes (see Figure 3-4), comprising pore-forming α-subunits and multiple cytosolic and transmembrane accessory subunits (see Table 3-2). The molecular compositions of native myocardial Kv channels, however, have not been determined. In contrast to the progress made in defining the Kv (and the Kir) α-subunits encoding native myocardial Kv (and Kir) channels or currents, however, much less is known about the functional roles of the putative accessory subunits (see Table 3-2) of these channels, as well as about the functioning of the various K2P subunits (see Table 3-3). An important focus of future work will likely be on defining the roles of the various Kvβ and the K2P α-subunits in the generation of functional myocardial K⁺ channels and on defining the molecular mechanisms controlling the properties and the cell surface expression of myocardial K⁺ channels encoded by the various Kv α and β, Kir α and β, and K2P α-subunits. In addition, as numerous studies have documented changes in functional K⁺ channel expression in a variety of myocardial disease states, changes that could reflect modifications in channel properties, as well as alterations in the molecular compositions of the channels or the processing of the underlying channel subunits, it seems clear that a major focus of future research will be on defining these mechanisms in detail. There are numerous possible mechanisms, including transcriptional, translational, and posttranslational, that likely are important in regulating the functional expression and the properties of myocardial K⁺ channels in the normal, as well as in the damaged or diseased, myocardium. Defining these mechanisms will have an important effect on the field, leading to new insights into the mechanisms that regulate myocardial K⁺ channel expression and functioning and to the development of novel therapeutic strategies to prevent or reverse the remodeling of these channels associated with systemic or myocardial disease.

References