HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (61) Citing Articles via Google Scholar Google Scholar Articles by Birnbaum, S. G. Articles by Schrader, L. A. Search for Related Content PubMed PubMed Citation Articles by Birnbaum, S. G. Articles by Schrader, L. A.

Structure and Function of Kv4-Family Transient Potassium Channels

Division of Neuroscience, Baylor College of Medicine, Houston, Texas

ABSTRACT

I. INTRODUCTION

A. K+ Channel Overview

II. MOLECULAR STRUCTURE

A. Mechanisms of Voltage-Sensing

B. Mechanisms of Inactivation

C. The T1 Domain and K+ Channel Multimerization

III. SUBCELLULAR LOCALIZATION AND TRAFFICKING

IV. INTERACTING SUBUNITS

A. Kv{beta}

B. K+ Channel Interacting Proteins

C. NCS-1

D. K+ Channel Accessory Protein

E. DPPX

F. Cytoskeletal Proteins

G. Neuron-Specific Modulator: PSD-95

H. Heart-Selective Modulator: MinK-Related Peptide 1

V. REGULATION BY POSTTRANSLATIONAL MODIFICATION

A. Palmitoylation

B. Glycosylation

C. Phosphorylation

VI. CELL BIOLOGY AND REGULATION IN NONNEURONAL SYSTEMS

A. Cardiovascular System

B. Smooth Muscle

C. Lung

D. Other

VII. CELL BIOLOGY AND REGULATION IN NEURONS

A. Distribution of Kv4.x Subunits in the Brain

B. K+ Channels in Neuronal Information Processing

C. Neuromodulatory Effects on A-Type K+ Currents in Hippocampal Pyramidal Neurons

D. Additional Levels of Complexity: Kv4.x Channels as Multimodal Signal Integrators

VIII. PATHOPHYSIOLOGY

A. Epilepsy

B. Alzheimer's Disease

C. Cardiac Pathology

IX. SUMMARY AND FUTURE DIRECTIONS

REFERENCES

	ABSTRACT

Top Next References

	I. INTRODUCTION

Top Previous Next References

 
This review focuses on the shal-type (Kv4.x) primary subunits of K⁺ channels and what we have learned about their function and regulation over the past few years. In our view, this is a timely topic and an emerging area of great physiological importance. A number of important and exciting new papers, primarily concerning studies of the nervous and cardiovascular systems, have been published in this area recently. This has prompted us to bring together an overview of this rapidly progressing area. One emerging theme from the recent literature is that K⁺ channels operate as supramolecular complexes, comprised of pore-forming primary, or , subunits plus a number of associated ancillary subunits and scaffolding proteins. Until recently these proteins, which interact with the primary subunits, were unknown, and their discovery has revised our consideration of shal-type K⁺ channel function. This viewpoint will be one of the unifying themes of this review in which we will specifically discuss proteins that have recently been shown to interact with Kv4.x subunits.

A second theme we will emphasize is the complexity of the mechanisms mediating dynamic regulation of the channel. In particular, the structure-function relationships for posttranslational modification of Kv4.x channels are beginning to be understood in much greater detail. These mechanisms will be discussed in the context of the physiological role of Kv4.x channels in modulating membrane excitability, particularly in the heart and in neurons. Much work has focused on kinase regulation of A-type K⁺ currents in neurons and direct phosphorylation of the Kv4.2 channel subunit. Regulation of Kv4.2 channels and the A-type K⁺ currents that they mediate in hippocampal dendrites will be discussed in particular detail. An especially intriguing aspect of these types of regulation is the possibility that Kv4.2 channels serve as molecular signal integration devices, allowing the cell to integrate a variety of cell-surface signals into a coordinated output in terms of membrane electrical properties.

A. K⁺ Channel Overview

Three groups of K⁺ channels have been characterized based on putative membrane topology of their principal subunits. In all three cases the primary subunits tetramerize to form a single transmembrane pore (reviewed in Ref. 42). The first group, typified by voltage-activated and Ca²⁺-activated K⁺ channels, has six transmembrane domains per -subunit. The second group, typified by the "leak" K⁺ channels, has four transmembrane domains in their -subunits. Finally, the "inward rectifiers" are the simplest structurally and have two transmembrane domains in each -subunit. Each of these three groups comprises a discrete family, based on sequence homology, which is further divided into subfamilies. Thus, based on nucleotide sequence analysis, it is obvious that many different K⁺ channels with diverse kinetics and functions exist. Diversity is also increased by their ability to form functional heterotetrameric structures and to associate with auxillary or -subunits.

The Shaker channel was the first K⁺ channel to be cloned. It is a voltage-dependent channel that was identified in a hyperexcitable Drosophila mutant (99, 147, 153). Other genes from Drosophila have been identified that bear 40% sequence homology with Shaker channels, and thus are included in the Shaker superfamily. These include the Shab, Shaw, and Shal subfamilies (179). The Kv channels are the mammalian gene counterparts for these Drosophila genes and bear 50–75% homology to the Drosophila genes. A systematic nomenclature based on amino acid sequence of the -subunits has been developed that defines the Shaker subfamily as Kv1.x, the Shab subfamily as Kv 2.x, the Shaw subfamily as Kv 3.x, and the Shal subfamily as Kv4.x. All these channels are gated by transmembrane voltage, hence the Kv nomenclature.

The Shal-type family in mammals is comprised of three distinct genes: Kv4.1, Kv4.2, and Kv4.3. The proteins encoded by these genes are highly homologous within the transmembrane regions, with divergent amino and carboxy termini. Kv4.x family channels in general are highly expressed in the brain, heart, and smooth muscles. Heterologous expression of these proteins in expression systems demonstrates that they activate at subthreshold membrane potentials, inactivate rapidly, and recover from inactivation quickly compared with other Kv channels (See Fig. 1A). Therefore, they are termed transient currents. Recent data from transgenic and knockout animals as well as expression of dominant negative constructs indicate that these channels participate in the transient outward A-type K⁺ current characterized in the somatodendritic compartments of neurons, as well as form the Ca²⁺-independent A-type K⁺ current (transient outward current or I_to) in cardiac myocytes. The A-type K⁺ current in these specific tissue systems is discussed in great detail in sections VI and VII. It is important for us to note that Kv4.x are not the only primary subunits that form A-type K⁺ currents; Kv1.x primary subunits also are capable of forming A-type K⁺ currents, but are not discussed in great detail in this review. Finally, molecular approaches such as RNA interference, antisense knockdown, dominant negative constructs, and genetically engineered mice are becoming increasingly utilized to define roles for Kv4.x channels.

View larger version (19K): [in this window] [in a new window]   FIG. 1. Voltage-dependent properties of neuronal A-type K+ current and Kv4.2-mediated currents in heterologous systems. A: representative A-type K+ current obtained from a hippocampal CA1 pyramidal neuron. Neuron was held at –80 mV and stepped to 0 mV. Decay of the A-type K+ current was fit with a single exponential (red) with time constant () around 28 ms. B: steady-state activation and inactivation curves of A-type K+ currents recorded from cell-attached patches at different locations on CA1 pyramidal dendrites. Note the shift of activation curves between channels located at proximal/soma and distal dendrites. [From Hoffman et al. (78), copyright 1997 Nature.] C: steady-state activation and inactivation curves of Kv4.2 subunits expressed in oocytes.

The kinetic properties of the transient A-type K⁺ current differ greatly depending on the type of cell in which the current is expressed. This can depend on heteromultimerization between primary subunits when a cell expresses more than one Kv4.x subtype primary subunit (e.g., Kv4.2 multimerizes with 4.3 in rat ventricular myocytes; see sect. VI), interaction with various other interacting proteins expressed selectively in various cells (see sect. IV), or posttranslational modification of the channels (see sect. V). We illustrate this diversity in A-type K⁺ currents, even within a single cell, by drawing your attention to Figure 1. This figure illustrates a current-membrane voltage plot from recordings at two different sites on the dendrite of a hippocampal pyramidal neuron. Notice that the activation curve is shifted 10 mV in the hyperpolarized direction for currents that are >100 µm away from the cell soma versus the currents recorded in more proximal dendritic regions. This indicates that the channels underlying the current in the distal dendrite open at a lower membrane potential than those in the proximal dendrites. This likely indicates differing primary or interacting subunits in Kv4.x-encoded channels in proximal dendrites compared with distal dendritic regions in these cells. However, we also know that the voltage dependence of activation of the distal currents is modulated by protein kinases, which is discussed in section V (76, 77, 228), so the different properties within different dendritic regions could be the result of differing posttranslational modification. Regardless of the underlying molecular mechanism, these findings clearly indicate subdomain specificity in the biophysical properties of A-type K⁺ channels.

Further evidence of cell context-dependent A-current properties is provided by considering the biophysical properties of the pore-forming Kv4.x -subunit expressed in heterologous systems versus the properties of native Kv4.x-encoded channel in the cell membrane. This also is illustrated in Figure 1. Consider that the A-type K⁺ current in hippocampal dendrites (Fig. 1B) is composed of, at least in part, Kv4.2 subunits. The activation and inactivation curve recorded from Xenopus oocytes expressing only Kv4.2 is shown in Figure 1C. Kv4.2 inactivation and activation curves in the oocyte expression system are hyperpolarized relative to the hippocampal A-type K⁺ current, particularly the current in the more proximal dendrites. This suggests that less depolarization is necessary to open the channels and fewer channels are inactivated at a given membrane potential in the ooctye versus dendrite. In addition, the Kv4.2 currents in oocytes inactivate slower and recover from inactivation much more slowly than the native current (not shown). Therefore, the Kv4.2 subunits must be modified in the hippocampal pyramidal neuron relative to the isolated -subunit expressed in oocytes, possibly by the effects of interacting subunits as well as by phosphorylation. We discuss specific possible mechanisms for these effects in later sections of the review.

	II. MOLECULAR STRUCTURE

Top Previous Next References

 
Analysis of various K⁺ channel sequences indicates several structural elements that are present in most K⁺ channels (see Fig. 2). These include the amino-terminal cytoplasmic domain, the T1 assembly domain, the six transmembrane -helical domains (S1-S6); the voltage sensor (S4), the pore domain (P-loop), and a carboxy-terminal cytoplasmic domain.

View larger version (23K): [in this window] [in a new window]   FIG. 2. Structure of K+ channels. A: schematic drawing of several structural elements present in most K+ channels. See text for discussion. B: a model of the S5, P-loop, and S6 regions of a K+ channel a: Ribbon representation of the transmembrane section of two opposing S5/P-loop/S6 structures based on the published crystal structure of two KirBac1.1 monomers. The residues forming the selectivity filter (P-loop), displayed as ball-and-stick, interact with four K+, colored green and cyan. The inner -helices, analogous to transmembrane helix S6, are colored purple, whereas the side chains of two hypothetical pore-blocking phenylalanine residues are colored red. The S5 -helix is colored green. [From Kuo et al. (107), copyright 2003 AAAS.] b: Relative positions of the P-loop helices that form the ion selectivity filter are depicted in red in the crystal structure of two K+ channels: Kir-Bac1.1 and KcsA. This view is from the extracellular side of the channel looking directly down the central ion-conduction pathway. Black lines through the center of each pore helix indicate its orientation relative to the center of the cavity.

A. Mechanisms of Voltage-Sensing

A detailed review of the mechanisms of voltage-sensing is beyond the scope of this review, but we will superficially discuss several recent, exciting papers that have sparked a debate in this area. We will review what is known about K⁺ channel voltage-sensing from known crystal structures to date. These data are derived from the bacterial K⁺ channels, whose pore-forming region exhibits homology with known mammalian K⁺ channels including the Kv4.x channels.

The membrane-spanning domain of all voltage-dependent K⁺ channels contains two highly conserved portions, the voltage-sensing portion that surrounds the central pore and the pore domain itself. The pore domain, S5, the P-loop, and S6 all together make up the ion permeation pathway, including the selectivity filter (Fig. 2A). A large body of evidence suggests that the voltage-sensing region of voltage-gated K⁺ channels is the fourth transmembrane -helix, or S4. This region of the protein contains positively charged arginine or lysine residues at essentially every third position and is the only transmembrane domain that is appreciably charged (82). Membrane depolarization causes a movement of the positively charged residues of S4 through the gating canal. This movement of charges through the electric field of the membrane mediates the actual opening of the channel and generates what is referred to as a gating current.

Most data on the pore domains and voltage-sensing mechanisms come from the crystal structures of the bacterial channels, KcsA (52) and MthK (94), which show a high degree of homology with voltage-gated K⁺ channels. Previous studies suggested that the S4 segment lies perpendicular to the plane of the membrane and is shielded by other parts of the channel from direct contact with the lipid environment. Upon membrane depolarization, the S4 segment transfers charged side chains between locations on opposite sides of the membrane, generating a gating current (24). Measurements of the gating charge moving across the membrane suggest that approximately four charges per subunit move entirely across the membrane during the channel's activation by depolarization (5). Mutagenesis studies in which these charges are neutralized reduce the gating charge measured with activation (5, 175). In addition, an acidic residue in S2 was also found to contribute to gating charge of Shaker channels (175). Independent groups have found that the S4 helix responds to voltage with rotational motion (37, 64) using LRET and FRET imaging, respectively. In addition, a similar rotational motion was found using histidine scanning to study changes in exposure of basic residues (190, 191). This rotational motion is likely coupled with translational motion to ultimately give rise to a "helical screw" type of motion (62).

This view has been somewhat challenged by the recent X-ray structure of another bacterial voltage-gated K⁺ channel, KvAP (95, 96), which interestingly contains an intracellular S4 domain. An examination of the voltage-sensing region of this channel suggests that this domain lies in a reclining position at the periphery of the channel, nearly in the plane of the membrane, in the closed state. Upon activation, it is proposed that the S4 domain moves through the hydrophobic lipid environment in a rowing, or paddlelike, motion to a more perpendicular position. The approach taken in these experiments, while innovative, may however introduce problems in interpretation of the data. These studies used detergent solubilization and formation of a complex of the ion channel with a monoclonal Fab fragment, which might have altered the gating mechanism. In this vein, Laine et al. (111) recently published evidence against the paddlelike movement using the MthK channel. Their experiments indicate that the S4 region actually lies in close proximity to the pore domain and that it interacts directly with the pore domain upon membrane depolarization. These results are not consistent with the paddle model of voltage-sensing. Thus the precise structural basis for voltage-sensing is not completely clear at present. Although all current models invoke the S4 domain as a voltage sensor, the details of how this works awaits further investigation.

B. Mechanisms of Inactivation

Channels formed from Kv4.x subunits rapidly inactivate. In general, two potential mechanisms for the inactivation of voltage-gated ion channels are well characterized. Both mechanisms apply to Kv4.x-encoded channels, although they have been more extensively studied in Shaker family channels. One mechanism of inactivation is termed N-type or "ball and chain" type inactivation. This mechanism involves the tethered amino-terminal inactivation domain (ball) of the channel (230) or a similar domain in a -subunit (158), binding to the intracellular entrance of the pore (90). This type of inactivation occurs only when the channel is already open, has fast kinetics, and can be eliminated by removal of the amino terminus (83, 84). A second type of inactivation is known as C-type inactivation and involves the pinching of the pore near the selectivity filter, resulting from a structural change in the four subunits (119, 142, 146). This type of inactivation is typically slower than N-type inactivation and can occur even in the absence of the amino-terminal domain of the channel (84). C-type inactivation is slower in the absence of N-type inactivation, suggesting that the two mechanisms are coupled (21, 84).

However, evidence suggests that Kv4.x family members, and Kv4.2 in particular, may not manifest these inactivation mechanisms in the classical sense. For example, deletion of the first 40 amino acids from the amino terminal of Kv4.2 results in a slowing of the fast and intermediate components of inactivation (15), rather than the complete loss of the fast component observed with amino-terminal deletion of Shaker channels. Kv4.1, on the other hand, loses the fast component of inactivation when the amino terminus is deleted (92, 93). Moreover, a positively charged domain at the carboxy terminus of Kv4.1 (amino acids 420–550) is necessary for rapid inactivation. Therefore, it appears that both the amino and carboxy termini of Kv4.1 are involved in inactivation gating (92, 93). Furthermore, several criteria for C-type inactivation, including interference by external tetraethylammonium (40), and slowing by high external potassium concentrations (21), are not found in Kv4.1- or Kv4.2-encoded channels (15, 17, 92). One key difference between Kv4.2 and Shaker channels that may account for this difference is that Kv4.2 channels predominantly inactivate from the closed state and recover directly, bypassing the open state (15), whereas Shaker channels inactivate only from the open state (45). One report, however, does suggest that Kv4.3 exhibits C-type inactivation (54). Thus subtle structural differences may account for the different attributes of channel inactivation manifest by Kv4.x versus Shaker subfamily channels.

C. The T1 Domain and K⁺ Channel Multimerization

Voltage-gated K⁺ channels only multimerize with members of their own subfamily. For example, mammalian members of the Kv4.x subfamily can multimerize and form functional channels with Kv4.x members from invertebrates, but they will not multimerize with mammalian members of the Kv1–3 subfamilies. The structural feature that mediates this is a highly conserved, cytoplasmic amino-terminal portion of the channel known as the tetramerization domain, or T1 domain (180). The T1 domain consists of 130 amino acids directly preceding the first transmembrane domain, and on its own, can form a stable tetramer in solution. The crystal structure of the T1 domain suggests that the specificity of subfamily associations are based on the polar interfaces between subunits (106). While the primary function of the T1 domain is in channel tetramerization, it also appears to play a role in channel gating. Mutations within this region can produce either rightward or leftward shifts in the activation curve and either a speeding or slowing of inactivation depending on the particular mutation (44).

	III. SUBCELLULAR LOCALIZATION AND TRAFFICKING

Top Previous Next References

 
Specific localization and trafficking mechanisms are relevant to Kv4.x channel function in most cells where they have been studied. This is particularly relevant in cells with a complex morphology and many distinct subcellular compartments, such as neurons. In this section we discuss mechanisms for general channel trafficking and for the localization of Kv4.x channels to specific subcellular compartments. One exciting area of recent research is investigating the role of Kv4.x ancillary subunits in channel expression and trafficking. We will highlight recent findings from these studies in the following section as well.

One factor necessary to understand the role of Kv4.x channels in neuronal function is to understand their specialized distribution within a cell. For example, both Kv4.2 and Kv1.4 channels mediate a transient, A-type K⁺ current; however, Kv4.2 is localized to the soma and dendrites of neurons, while Kv1.4 is concentrated in axons (181). Intrinsic structural features of Kv4.x channels may regulate the subcellular trafficking of these channels. A recent study has shown that this subcellular targeting is mediated at least in part by a 16-amino acid dileucine-containing motif (160) within the pore-forming -subunit. A comparison of all known mammalian and invertebrate Kv4.x channels reveals that 13 of the 16 amino acids in this sequence motif are conserved. Deletion of these amino acids (474–489) from Kv4.2, or merely replacing the two leucines with an alanine and valine, disrupts the polarization normally seen with Kv4.2. Instead of being targeted to the cell body and entire dendrite, distribution is limited to the proximal dendrite and proximal axon. Conversely, insertion of the 16-amino acid dileucine motif into the carboxy terminus of Kv1.3 and Kv1.4 is sufficient to alter their subcellular localization from axons to dendritic processes.

In cardiac ventricular myoctyes, Kv4.2 has been shown to be concentrated in the sarcolemma (plasma membrane, Ref. 19), where the highest level of immunoreactivity is found at the intercalated disk region that connects myocytes together. Using high-resolution techniques, Takeuchi et al. (192) have shown that Kv4.2 localizes to the sarcolemma in atrial myoctyes; however, in ventricular myocytes Kv4.2 is predominant in the t tubules rather than the peripheral membrane. Thus Kv4.2 is selectively targeted to specific subcellular domains in cardiac myocytes as well. However, the mechanism that targets Kv4.x channels in myoctyes has not yet been investigated.

Another factor necessary for regulating channel localization is release from the endoplasmic reticulum (ER). Several intrinsic sequences that regulate retention of proteins in the ER or export of the protein from the ER have been identified (120). One important signal for the trafficking of membrane channel proteins in general is the RXR ER retention motif. This has been shown to be necessary for ER retention in both K_ATP channels (231) and the NR1 subunit of N-methyl-D-aspartate (NMDA) receptors (173). Although an RXR ER retention signal has not been specifically identified in Kv4.2, there exists an RKR in its amino terminus, an RYR in an intracellular loop between the second and third transmembrane domains, and an RIR in its carboxy terminus, any of which would serve as a good candidate for ER retention. Mutation of the amino-terminal RKR potential ER retention signal did not alter the ER localization of Kv4.2 in COS cells (183), suggesting that this particular site is not involved in ER retention; however, other potential ER retention sites in Kv4.2 have not been examined. Furthermore, ER export signals, which can accelerate the ER export of proteins, are now being identified (reviewed in Ref. 120). This is another area that has not been closely examined for Kv4.x channels.

Channel trafficking from the ER may also be regulated by interactions with auxiliary proteins as well as potentially by channel phosphorylation (reviewed in Ref. 120). For example, coexpression of a PDZ domain-containing protein, SAP-97, with the Kv1 family of K⁺ channels inhibits ER release of the channel and dramatically reduces their cell surface expression (195). Protein phosphorylation has also been shown to enhance ER export for NMDA receptor subunits (173). As Kv4.x channels can be phosphorylated by several kinases (see sect. V), this is another potential mechanism for regulating channel surface expression. Overall, much more work is required to understand the targeting and distribution of Kv4.x channels; however, these studies suggest that altered release from the ER may regulate surface expression of Kv4.x channels, and ultimately changes in cellular excitability.

A large variety of proteins that interact with Kv4.x channels have now been identified (see Table 1). Many of these auxiliary proteins appear to play a role in the localization of Kv4.x channels within the cell. Furthermore, many of these interacting proteins have also been shown to alter the biophysical properties of Kv4.x channels. The effects of these interacting proteins on Kv4.x channel kinetics and cellular distribution are discussed in section IV.

	IV. INTERACTING SUBUNITS

Top Previous Next References

A. Kv

Functional diversity of K⁺ channels can be augmented through the association of the pore-forming -subunits with auxiliary subunits. We now know of many auxiliary subunits that interact with -subunits and contribute to the regulation of biophysical properties and expression levels of K⁺ channels. The auxiliary subunits discovered first and characterized in most detail are the members of the Kv auxiliary subunit family.

Kv subunits lack putative transmembrane domains and potential glycosylation sites or leader sequences, suggesting that they are cytoplasmic proteins (174). Three Kv genes have been identified: Kv1, Kv2, and Kv3 (73, 113). Several functional consequences of -subunits on Kv channels have been described. One effect is to increase the inactivation rate of Kv channels (73). In a striking example, Kv subunits can convert normally noninactivating delayed rectifier channels to a rapidly inactivating channel (113, 158). Kv1 and Kv2 have also been shown to modulate voltage dependence of Kv channels in heterologous expression systems (56). Another function of -subunits is to serve as chaperone proteins that promote and/or stabilize cell surface expression of K⁺ channel -subunits in general (57, 116, 135, 136, 182, 199), including Kv4.3 channels (224). Another proposed role for Kv subunits is as redox sensors, as they structurally are similar to oxidoreductase enzymes (67) and may confer redox sensitivity to channel function through a bound NADPH cofactor.

The effects of -subunits specifically on Kv4.x channel function is still under investigation. While Kv1.2 does not affect Kv4.2 channel gating kinetics, it does confer sensitivity to redox modulation and hypoxia (149). Kv4.3 interacts with all three known -subunits (1, 2, and 3, see Refs. 48, 224). Kv1 and -2 cause an increase in Kv4.3 channel current density, with no effect on channel gating (224), while 3 shifts steady-state inactivation and slows recovery from inactivation. In addition, Kv1 and -2 subunits were found to localize in many of the same brain structures as Kv4.x-family -subunits, particularly the hippocampus (159), suggesting they are an integral component of neuronal A-type K⁺ currents.

B. K⁺ Channel Interacting Proteins

Although Kv4.x channels are the best candidates for mediating A-type K⁺ currents in neurons, Kv4.2 homomers expressed in nonneuronal systems manifest activation curves that are somewhat shifted in the depolarizing direction, and their recovery from inactivation is slower than what is observed for native A-type K⁺ currents. Interestingly, in early studies Serodio et al. (176) found that rat brain mRNA coexpressed with Kv4.2 in oocytes could restore many of the properties seen in native A-type K⁺ currents, suggesting that some cofactor was responsible for this change. A good candidate for this cofactor came to light in the form of a family of auxiliary proteins known as K⁺ channel interacting proteins, or KChIPs (11). Four KChIPs have been identified to date. KChIPs 1–3 were originally identified in a yeast-two-hybrid assay using the amino terminus of Kv4.3 as bait (11). KChIP4 was identified later as a binding partner of presenilin 2 (131). The KChIPs are members of the recoverin-neuronal calcium sensor superfamily that are Ca²⁺ binding proteins. KChIPs share sequence similarity with neuronal calcium sensor 1 (NCS-1) or frequenin (discussed below), a calcium-binding protein that is involved in the regulation of transmitter release in Drosophila (154). Moreover, KChIP3 is identical to calsenilin (33) and DREAM, a Ca²⁺-regulated transcriptional repressor (35). The various KChIPs have a variable amino-terminal region, but a conserved carboxy-terminal region that contains four EF-hand-like calcium-binding motifs. The function of these Ca²⁺ binding domains in K⁺ channel regulation is unknown; however, a Ca²⁺ dependence of transient K⁺ currents has been reported (28, 59, 188).

The KChIPs colocalize and coimmunoprecipitate with brain Kv4.x subunits through binding to the amino terminus of Kv4.x -subunits and are thus considered integral components of the native Kv4.x channel complexes (11, 16). KChIP2 has also been shown to be an integral subunit in the ventricular wall of the heart where its expression parallels the gradient in transient outward current I_to (47, 148, 161, 162).

KChIPs 1, 2, and 3 all have similar effects on the physiological properties of Kv4.x channels, these effects having been mostly studied in Kv4.2 channels specifically. The physiological effects of KChIPs include an increase in the density of Kv4.2 currents, a hyperpolarizing shift in the activation curve, an increase in the rate of recovery from inactivation, and a slowing of the time constant of inactivation. The variable amino-terminal region of the KChIPs is unnecessary for this modulation to occur. However, modulation fails to occur when mutations are introduced into the EF-hand region in the conserved carboxy terminus of KChIPs (11).

Coexpression of KChIP1, 2, or 3 together with Kv4.2 increases the peak current density 10-fold, suggesting that the KChIPs may promote surface expression, or stabilize the expression of Kv4.x channel proteins at the surface (11). A recent study has shown that COS1 cells transfected with Kv4.2 alone exhibit a perinuclear pattern of immunofluorescent staining, which is typical for ER retained proteins (183), and no detectable surface staining (Fig. 3). Furthermore, double-labeling of these cells for Kv4.2 and the ER protein calnexin revealed an overlapping staining pattern, while double-labeling with a Golgi apparatus marker revealed no overlapping staining. Coexpression of Kv4.2 with KChIP1, 2, or 3 dramatically altered the subcellular distribution of Kv4.2 channels (Fig. 3). Kv4.2 staining in the cotransfected cells exhibited significant cell-surface staining, and the remaining intracellular Kv4.2 colocalized with the Golgi apparatus marker rather than the ER marker. In contrast, KChIP4 does not promote surface expression of Kv4.2, and in fact, KChIP4 may competitively antagonize binding of the other KChIPs to Kv4.2 (183). Furthermore, unlike the other KChIPs, KChIP4 eliminates fast inactivation of Kv4.x-encoded currents (79).

View larger version (44K): [in this window] [in a new window]   FIG. 3. KChIP coexpression leads to changes in the subcellular localization of Kv4.2 expressed in COS-1 cells. A–E: COS-1 cells were transfected with EGFP-Kv4.2 alone (A) or with EGFP-Kv4.2 and KChIP1 (B), KChIP2 (C), KChIP3 (D), and KChIP4a (E) at 2:1 cDNA ratios and stained for cell surface Kv4.2 to determine the extent of Kv4.2 surface expression in transfected cells, which were detected by EGFP signals. Left panels: external anti-Kv4.2 antibody staining performed in the absence of detergent permeabilization. Right panels: EGFP signals show the distribution of total Kv4.2. F and G: detergent-permeabilized cells expressing EGFP-Kv4.2 alone (F) or EGFP-Kv4.2 and KChIP2 (G) were stained with anti-calnexin antibody (as an ER marker; F) or L. culinaris agglutinin (LCA; as a Golgi marker; G) (right panels). The localization of Kv4.2 is shown as EGFP signal (left panels). Scale bar, 10 µm. [From Shibata et al. (183), copyright 2003 The American Society for Biochemistry and Molecular Biology.]

Overall, the ability of KChIPs to restore more native-like properties to Kv4.x channels and their colocalization with Kv4.x channels in several systems suggests that KChIPs may be an important and integral part of the composition of A-type K⁺ channels in cells.

C. NCS-1

Interestingly, NCS-1 (also known as frequenin), a member of the EF-hand family of Ca²⁺ sensing proteins which includes KChIPs, is also expressed in mouse brain and coimmunoprecipitates with Kv4.2 and Kv4.3 (69, 138). In heterologous systems, NCS-1 coexpression with Kv4.x -subunits increases the current density and slows the rate of inactivation of the Kv4.x current. In contrast to KChIPs, however, NCS-1 does not affect the voltage dependence of inactivation or rate of recovery from inactivation of the channel. The effects of NCS-1 on current density and rate of inactivation are Ca²⁺ dependent, as they are blocked by a membrane-permeable Ca²⁺ chelator. These results are not due to a nonselective effect of Ca²⁺ binding proteins on Kv4.x channels, as other closely related members of the Ca²⁺ binding family (VILIP1, hippocalcin, neurocalcin, and calmodulin) have no significant effect on the amplitude or time course of Kv4.2-encoded current (138).

In addition to altering the biophysical properties of Kv4.x channels, NCS-1 can also alter the cellular distribution of Kv4.x channels. In COS cells transfected with Kv4.2 cDNA, Kv4.2 proteins were observed predominantly in perinuclear regions (138). However, cotransfection of Kv4.2 and NCS-1 increased the Kv4.2 immunoreactivity in the outer margins of the cells. Similarly, Guo et al. (70) reported an increase in surface expression of Kv4.3 channels coexpressed with NCS-1 in HEK-93 cells, and a decrease in surface expression of the channel when NCS-1 levels were reduced with an antisense oligonucleotide. In accordance with the enhanced surface expression of the Kv4.x channels, both groups reported an increase in A-type K⁺ current when the Kv4.x channel was coexpressed with NCS-1. These data suggest that in addition to regulating the rate of current inactivation for Kv4.2 and Kv4.3, NCS-1 can also regulate A-type K⁺ current by enhancing the surface expression of these channels.

D. K⁺ Channel Accessory Protein

K⁺ channel accessory protein (KChAP) was first cloned from a rat cDNA library and is not yet described in other species (211). KChAP is a member of the protein inhibitor of activated signal transducer STAT3 gene family. KChAPs are known to interact with several different families of Kv channels, including Kv1.3, 2.1, 2.2, and 4.3 (110, 109). Coexpression of KChAPs with these Kv channels increases current expression, without an effect on current kinetics. This chaperone activity of KChAPs may be mediated indirectly through an interaction with Kv1.2 as expression of Kv1.2 inhibits the chaperone effects of KChAP on Kv4.3 (110). The effects of KChAP on Kv4.1 or 4.2 are currently unknown.

E. DPPX

Although association of KChIPs with Kv4.2 or Kv4.3 in heterologous expression systems can restore most of the properties of native A-type K⁺ currents, one shortcoming is that they slow the kinetics of inactivation, compared with the fast kinetics of inactivation of native A-type K⁺ currents. In recent studies Bernardo Rudy's group (133) discovered that coexpression of high-molecular-weight mRNA isolated from rat cerebellum with Kv4.2 in oocytes resulted in an acceleration of the kinetics of inactivation. This effect was not eliminated by using KChIP antisense oligonucleotides (133), suggesting the existence of another Kv4.x auxiliary protein that the authors labeled K⁺ channel accelerating factor, or KAF. Later, Rudy's group found that KAF is a transmembrane protein called DPPX (134).

DPPX had no previously known function, although it is related to the dipeptidyl aminopeptidase CD-26, which has a role in cell adhesion. Coexpression of DPPX and Kv4.2 had similar effects to coexpression of KChIPs and Kv4.2: an increase in surface expression, increased speed of recovery from inactivation, and a shift in inactivation voltage dependence. However, the time constant of inactivation of Kv4.2 was decreased with coexpression of DPPX, making the current more similar to native currents (134). DPPX had similar effects when coexpressed with Kv4.3, but not with other fast-inactivating K⁺ channels such as Kv1.4.

DPPX localizes to the somatodendritic compartments of neurons that are known to contain Kv4.2, such as pyramidal neurons of the hippocampus and striatal neurons, as well as to regions that contain Kv4.3, such as the Purkinje cells of the cerebellum. DPPX also localizes to areas known to contain both Kv4.2 and Kv4.3, such as granule cells of the cerebellum and dentate gyrus (134).

DPPX also strongly alters the cellular localization of Kv4.x channels (134). Coexpression of Kv4.2 and DPPX in CHO cells results in a 20-fold increase in surface expression of the channel as measured by a surface protein biotinylation assay. This dramatic effect was due to a 2-fold increase in total protein levels as well as a 10-fold increase in the surface-exposed channel protein. The increase in surface expression of Kv4.2 protein correlated with a 25-fold increase in A-type K⁺ current when Kv4.2 was coexpressed with DPPX in Xenopus oocytes. The striking alteration in cellular localization of Kv4.2 channels as well as the ability to restore a more native A-type K⁺ current suggests that DPPX is also an integral molecular component of Kv4 channels in cells where they are coexpressed.

F. Cytoskeletal Proteins

A search for proteins responsible for Kv4.2 localization to somatodendritic compartments of neurons (181) led to the discovery that Kv4.2 interacts with filamin, a member of the -actinin/spectrin/dystrophin family of actin-binding proteins. Petrecca et al. (151) demonstrated an association between the carboxy terminal of Kv4.2 and filamin in neurons. These investigators also showed that Kv4.2 and filamin colocalize in cerebellum and cultured hippocampal neurons. Furthermore, Kv4.2 is expressed in a punctate pattern in neuronal dendrites that colocalizes with the synaptic marker synaptophysin. Transfection of Kv4.2 into a heterologous cell line lacking filamin resulted in a uniform expression pattern of the channel, while similar expression of Kv4.2 in a cell line that contains filamin resulted in Kv4.2 colocalizing with filamin at the roots of filopods (151). Transfection of a mutant form of Kv4.2 in which the filamin-binding region had been altered resulted in a uniform expression pattern similar to that observed in the cells without filamin. Although total Kv4.2 protein expression was the same, the whole cell Kv4.2 current density from Kv4.2 transfected cells was two- to threefold greater in the filamin-expressing cells than in cells without filamin. This difference was due to a higher density of Kv4.2 channels in the surface membrane rather than any changes in single-channel conductance. These data suggest that filamin may function as a scaffold protein that enhances Kv4.2 channel expression on the surface membrane. Furthermore, an interaction with filamin may target Kv4.2 to synapses, although further research is needed to confirm this.

Interestingly, the critical region of Kv4.2 interaction with filamin appears to be the proline-rich regions of 601–604 and is identical to a sequence in Kv4.3 that also binds to filamin (151). These Kv4.2 residues critical for filamin association are in the region of an ERK phosphorylation site (T602) discussed in section VI. We speculate that ERK phosphorylation at this site could interfere with the Kv4.2/filamin association and result in altered Kv4.2 localization.

Another protein family that may play a role in Kv4.x localization and distribution is the integrins. Interaction with these cell adhesion molecules may underlie the restricted cellular distribution of the Kv4.x-interacting protein filamin (31). A variety of studies indicate that filamin and integrin interact and are components of the neuromuscular junction, where -integrin plays a role in the signaling events that lead to agrin-induced clustering of ACh receptors (127).

A possible interaction of Kv4.2 with integrins is supported by a recent study that showed that the extracellular matrix protein vitronectin affects the expression of Kv4.2 in hippocampal explants (205). Vitronectin is an extracellular ligand for transmembrane integrin proteins. In developing hippocampal pyramidal neurons, vitronectin exposure for 3–4 days was found to increase the peak current amplitude of A-type K⁺ current. Vitronectin treatment had no significant effect on the voltage dependence of activation or inactivation, or the kinetics of inactivation or recovery from inactivation of the A-type K⁺ current. Immunocytochemical analysis revealed that vitronectin increased the immunoreactivity of Kv4.2 proteins, but not Kv1.4 proteins, suggesting that the changes in A-type K⁺ current are mediated by Kv4.2. Furthermore, vitronectin had no effect on the amplitude of the sustained components of K⁺ currents in these hippocampal pyramidal neurons, indicating a specific effect on A-type K⁺ currents. These data are indicative of a functional interaction of Kv4.2 with integrins, at least in neurons. While this prospect is enticing and promising, more research is necessary to confirm such an interaction and determine its molecular basis.

G. Neuron-Specific Modulator: PSD-95

Another synapse-associated protein found to interact with Kv4.2 is the scaffolding protein PSD-95 (216). Using a surface biotinylation assay, Wong et al. (216) showed that coexpression of Kv4.2 with PSD-95 in CHO cells enhanced the surface expression of Kv4.2 2-fold. Deletion of the last four amino acids of Kv4.2's sequence (VSAL), which constitute a putative PDZ-binding domain, or using a palmitoylation-deficient PSD-95 mutant which blocks Kv1.4 channel clustering (85) blocked this increase in surface expression of Kv4.2 without affecting total channel levels (216). These results were confirmed using a fluorescently tagged Kv4.2 protein to monitor channel protein distribution. In cells transfected with Kv4.2 alone, the majority of the fluorescence was located in an internal reticular network with some fluorescence at the outer cell margins. Coexpression of Kv4.2 and PSD-95 resulted in an increase in channel expression on the cell surface as well as the formation of clusters of Kv4.2 channels. Thus PSD-95 could be another protein responsible for recruiting Kv4.x channels to the cell surface.

H. Heart-Selective Modulator: MinK-Related Peptide 1

MinK-related peptide 1 (MiRP1) is a protein linked to maintaining electrical stability in the heart. Initial functional data suggested that MiRP1 associates with and modulates the function of HERG, which contributes to the fast component of the delayed rectifier current of heart (2). More recently KvLQT1, the slower component of the delayed rectifier (196), as well as Kv3.4 and KCNQ4 (1, 171) were found to interact with MiRP1, suggesting it is a rather promiscuous binding protein. MiRP1 has been found to affect Kv4.2 currents expressed in Xenopus oocytes, slowing the rates of activation and inactivation and shifting the voltage dependence of activation in the depolarizing direction (232). These data strongly suggest that MiRP1 may contribute to regulating I_to in the heart, although this has not been tested directly. MiRP1 has been undetected in the brain other than in the pituitary (217), so the relevance of MinK to neuronal A-type K⁺ current remains to be seen.

	V. REGULATION BY POSTTRANSLATIONAL MODIFICATION

Top Previous Next References

 
An enormous variety of posttranslational modification of proteins has been described, including glycosylation, methylation, acetylation, ubiquitation, attachment of fatty acids, covalent attachment of coenzymes, and phosphorylation. Posttranslational modification of proteins can play an important role in the proper folding, assembly, and trafficking of many membrane-associated proteins. The role of several common forms of posttranslational modification of channels will be discussed in this section: palmitoylation, glycosylation, and phosphorylation, specifically in the context of regulating Kv4.x channels.

A. Palmitoylation

To our knowledge, most forms of posttranslational modification have not been well studied for Kv4.x channels. However, a few studies have examined the role of posttranslational modification of subunits that interact with Kv4.x channels (see sect. IV). There is substantial evidence that attachment of palmitate, a long-chain fatty acid, to KChIPs is required for the plasma membrane localization of both ancillary subunits and the associated Kv4.x channel. For example, two "long" isoforms of KChIP2 have been identified which contain potential palmitoylation sites. These two isoforms show enhanced surface membrane expression compared with the shorter isoform when expressed in the absence of Kv4.3 (194). Furthermore, the longer two isoforms increased the surface membrane expression of Kv4.3 channels and produced larger increases in Kv4.3 current density compared with the shorter KChIP2 isoform. Mutation of the palmitoylation sites on the longer KChIP2 isoform reduced both the plasma membrane localization of Kv4.3 channels as well as the enhanced current observed with wild-type KChIP2. These data suggest that palmitoylation of KChIP is an important mechanism for enhancing cell surface localization of Kv4.x channels.

Palmitoylation also appears to be important for trafficking of PSD-95 to postsynaptic densities and for the formation of ion channel clusters (55, 198). The palmitoylation of PSD-95 is also required for PSD-95-mediated surface expression of Kv4.2. Furthermore, clustering of the K⁺ channel Kv1.4 at postsynaptic densities requires palmitoylation of PSD-95. These data suggest that a similar mechanism may be required to cluster Kv4.x channels at the cell membrane in postsynaptic densities. Obviously, there is much more work required to understand the role of palmitoylation in regulating Kv4.x channel function, cellular localization, and stability/degradation.

B. Glycosylation

Glycosylation of proteins has been shown to promote proper protein folding in the ER as well as to alter protein transport and targeting (for review, see Ref. 74). In general, glycosylation of K⁺ channels is not required for surface expression, but does appear to increase surface expression of the channel by decreasing channel turnover and increasing channel stability. For example, blocking glycosylation of Shaker-type K⁺ channels has been shown to dramatically decrease the stability and cell surface expression of the channel but not effect the folding (104) and assembly of functional channels, or their transport to the cell surface (168). Similar results have been shown for the HERG K⁺ channel (65).

Concerning Kv4.x-family channels specifically, none of the channels contains a consensus sequence for attachment of an N-linked oligosaccharide (N-X-S/T) on their extracellular surface. While there is no known consensus sequence for O-linked glycosylation, we are not aware of any studies that have determined that any of the Kv4.x channels are directly glycosylated. However, treatment of dissociated ventricular myocytes with a neuraminidase to remove sialic acid, a negatively charged sugar residue, decreases I_to (which is produced by Kv4.2 and Kv4.3 channels, see Ref. 202). Accordingly, an increase in the duration of action potentials was also observed in these cells. Removal of sialic acid did not block the channels from reaching the cell surface, nor did it block the formation of functional channels, but it did alter the voltage dependence of activation in addition to reducing I_to amplitude. These results were mimicked by expression of Kv4.3 in a sialylation-deficient heterologous cell line. The reduction of I_to suggests that removal of sialic acid reduces the number of K⁺ channels at the plasma surface, possibly through a faster channel turnover rate that leads to a decrease in Kv4.3 channel stability. While it is feasible that sialic acid residues on Kv4.3 directly modulate the channel, it is more likely that the effects observed in this study were due to deglycosylation of a Kv4.x interacting subunit (see sect. IV).

C. Phosphorylation

In the last few years there has been a great deal of evidence that A-type K⁺ currents in both neuronal and cardiac cells can be regulated by phosphorylation. Application of a phorbol ester, which activates protein kinase C (PKC), suppresses I_to in ventricular myocytes (13, 137). Similarly, recording from pyramidal cell dendrites in the hippocampus, Hoffman and Johnston (76) showed that activation of either PKA or PKC decreased the probability of A-type K⁺ channel opening. PKA and PKC activation also increased the amplitude of back-propagating action potentials in distal dendrites, consistent with a decrease in K⁺ channel current (see sect. VII). It has subsequently been shown that PKA and PKC modulation of the A-type K⁺ current in hippocampal neurons is mediated through the ERK/mitogen-activated protein kinase (MAPK) pathway (228). Finally, activation of PKC suppressed Kv4.2 or Kv4.3 current in Xenopus oocytes (137).

Kv4.2 contains a consensus sequence for phosphorylation by protein tyrosine kinase (PTK), and it has previously been shown that Kv1.2 channels can be inhibited by PKC activation of PTK phosphorylation (115). Thus Nakamura et al. (137) mutated the consensus sequence in Kv4.2 for phosphorylation by PTK to determine if this site mediated the PKC inhibition of the A-type K⁺ current. Modulation of this Kv4.2 mutant was similar to the wild-type; therefore, PKC inhibits the Kv4.2 currents independent of direct PTK phosphorylation. Furthermore, Kv4.3, which is also inhibited by PKC, does not contain a consensus sequence for PTK phosphorylation.

Finally, it is worth noting that splice variants may potentially exhibit differential sensitivity to phosphorylation. For example, two splice variants of human Kv4.3 have been identified; the longer splice variant contains 19 amino acids with a consensus PKC phosphorylation site inserted into the carboxy terminus after the last transmembrane spanning region (105). PKC activation reduced the peak current in the longer splice variant but had no effect on the shorter splice variant when expressed in a heterologous expression system (152). Mutation of the PKC consensus site abolished the sensitivity of the long splice variant of Kv4.3 to PKC activation, suggesting that direct phosphorylation of Kv4.3 in the alternatively spliced domain modulated the K⁺ current.

As discussed above, phosphorylation of Kv4.2 by PKA has been shown to modulate the K⁺ current in hippocampal neurons (76). Surprisingly, PKA has no effect on K⁺ current when Kv4.2 is expressed alone in Xenopus oocytes (169). Further studies revealed that coexpression of KChIP3 with Kv4.2 was sufficient to rescue the PKA regulation of Kv4.2 (see Fig. 4). Using mutations of the PKA phosphorylation sites in Kv4.2, Schrader et al. (169) found that both PKA phosphorylation of Kv4.2 at serine-552 in addition to an interaction between Kv4.2 and the ancillary subunit KChIP3 were required for PKA regulation of the A-type K⁺ current in oocytes. This finding leads to a completely unexpected conclusion—direct phosphorylation of the Kv4.2 -subunit by PKA is not sufficient to modulate channel function; the effect of phosphorylation requires the presence of the KChIP ancillary subunit.

View larger version (13K): [in this window] [in a new window]   FIG. 4. KChIP3 coexpression is required for modulation of the current by PKA. Activation curve of current obtained from oocytes transfected with either Kv4.2 alone (A) or Kv4.2 + KChIP3 (B) is shown. The current was evoked from depolarizing pulses in control (squares) and 15 min after commencement of forskolin application (triangles). Forskolin does not significantly alter recovery from inactivation (not shown). [Adapted from Schrader et al. (169).]

Examination of the amino acid sequence for Kv4.2 and Kv4.3 reveals several consensus sequences for phosphorylation by PKA, PKC, ERK/MAPK, and calmodulin kinase II (CaMKII). Thus far, direct biochemical studies of Kv4.x channel phosphorylation have only been published for the ERK and PKA sites of Kv4.2. Two sites on Kv4.2 that are phosphorylated in vitro and in vivo by PKA have been identified (12) as well as three ERK/MAPK phosphorylation sites (3). However, the only functional study of the specific Kv4.2 phosphorylation sites that has been published concerns the serine-552 site, which mediates PKA modulation of Kv4.2 voltage dependence of activation when associated with KChIP (169). In addition, mutation of the other PKA sites on the amino terminal had no effect on channel kinetics.

As part of the biochemical studies of PKA and ERK phosphorylation of Kv4.2, three antibodies have been developed which selectively recognize Kv4.2 when it is phosphorylated at different sites: 1) when it is phosphorylated by PKA at the amino-terminal site (threonine-38), 2) when it is phosphorylated by PKA at the carboxy-terminal site (serine-552, see Ref. 12), and 3) when it is triply phosphorylated at all three ERK sites (threonine-602, threonine-607, and serine-616; see Ref. 3). With the use of these three antibodies, the distribution of phosphorylated Kv4.2 was examined in the mouse brain. Immunoreactivity for all three antibodies was found in many widespread regions including the cerebral cortex, hippocampus, thalamus, cerebellum, striatum, and amygdala (204). While immunoreactivity with each antibody was found in the same structures, the distribution of staining varied between the antibodies (Fig. 5). Of particular interest is the pattern of staining in the hippocampus. For example, the stratum lacunosum moleculare, which receives inputs from the entorhinal cortex via the perforant pathway, displays relatively little ERK-phosphorylated Kv4.2 or PKA carboxy-terminal-phosphorylated Kv4.2. However, this same layer is stained by the antibody that recognizes Kv4.2 when it is phosphorylated by PKA at the amino terminus. Similarly, of the three antibodies tested, the soma of CA3 neurons are primarily recognized by the ERK triply phosphorylated Kv4.2 antibody, and the mossy fiber inputs to CA3 are primarily recognized by the carboxy-terminal PKA-phosphorylated Kv4.2 (see Table 2 for a summary of immunoreactivity in hippocampal subregions).

View larger version (88K): [in this window] [in a new window]   FIG. 5. Immunohistochemistry of the hippocampus using antibodies recognizing phosphorylated Kv4.2. A: staining of ERK triply phosphorylated Kv4.2 at x100. B: carboxy-terminal PKA-phosphorylated Kv4.2 at x100. C: amino-terminal PKA-phosphorylated Kv4.2 at x100. Strong immunoreactivity in the CA1 stratum oriens (so) and stratum radiatum (sr) can be seen in A and B with a dearth of immunoreactivity in these same layers in C. However, there is a relative dearth of immunoreactivity in the stratum lacunosum moleculare (slm) in A and B, whereas immunoreactivity is strong in the slm in C. The immunoreactivity in the molecular layer of the dentate gyrus (ml-dg) is increased in B compared with A and C. sp, Stratum pyramidali. [From Varga et al. (204), copyright Cold Spring Harbor Laboratory Press.]

	VI. CELL BIOLOGY AND REGULATION IN NONNEURONAL SYSTEMS

Top Previous Next References

 
A. Cardiovascular System

The A-type K⁺ current is a critical regulator of the excitable cells of the heart. Various K⁺ currents determine the amplitude and duration of action potentials in the myocardium. Figure 6 shows a cardiac action potential, which is much longer in duration than a neuronal action potential, and the corresponding K⁺ currents that participate in repolarization of the action potential. The cardiac action potentials are initiated by Na⁺ channels that depolarize the membrane and activate voltage-gated Ca²⁺ and K⁺ channels. The K⁺ currents characterized in the heart are highly variable, depending on the species and the specific region of the heart. This subject is quite complicated and has been discussed with great sophistication previously (140); therefore, we will not cover this issue here. For the purposes of this review, we will concentrate on the I_to in the heart, which, similarly to A-type K⁺ current, activates and inactivates rapidly. A great amount of variability across species exists in I_to as well as K⁺ channel subunits that contribute to I_to, an indication of the different requirements for cardiac function in mammals of different sizes. In rodents I_to appears to be the major repolarizing current, whereas in higher mammals I_to is responsible only for the rapid repolarization phase (166). I_to has now been divided into two distinct transient outward K⁺ currents, I_to,f and I_to,s which are differentially distributed in the myocardium. These currents are differentiated based on their rate of inactivation and recovery from inactivation (29, 207, 221). I_to,s appears to be formed by Kv1.4 channels (150, 201, 212). I_to,f has been characterized in ventricular myocytes as well as atrial cells from various species (140). While there is substantial variability across species, considerable evidence suggests that the Kv4.x family of K⁺ channel proteins are the primary subunits that underlie I_to,f in most species (50).

View larger version (20K): [in this window] [in a new window]   FIG. 6. Cardiac action potential and currents that underlie repolarization. A: example of a ventricular action potential. B and C: K+ currents responsible for repolarization. The initial repolarization is the result of the rapidly activating transient outward current (Ito), the ultrarapid delayed rectifier (IKUR), and the leak currents. The rapid and slow (IKs) delayed rectifiers as well as IKur and leak currents underlie the plateau phase. [From Tristani-Firouzi et al. (199a). Copyright 2001, with permission from Excerpta Medica.]

In contrast, Kv4.2 appears to be the only primary subunit that underlies I_to,f in atrial myocytes of rodents. Thus I_to,f is selectively eliminated in atrial myocytes by Kv4.2 antisense oligonucleotides (27), but not anti-Kv4.3 oligonucleotides. In addition, a transgenic mouse expressing a Kv4.2 dominant negative shows no I_to,f in atrial myocytes (220, 221). Taken together, these data suggest that Kv4.2 and Kv4.3 heteromultimerize to form the I_to,f of the rodent ventricle, but only Kv4.2 participates in I_to,f of rodent atrium.

Different K⁺ channel subunits appear to be responsible for the I_to of higher mammalian heart versus rodent heart. For example, Kv4.3 appears to be solely responsible as the pore-forming subunit for I_to,f of canine and human heart (23, 51). The functional diversity of I_to,f of canine and human heart is attributed to differential expression of interacting proteins. Thus a gradient of KChIP2 gene expression that parallels the gradient in I_to expression has been seen in both canine and ferret ventricle (148, 161