Contribution of N- and C-terminal channel domains to Kv channel interacting proteins in a mammalian cell line
1 Institut für Neurale Signalverarbeitung, Zentrum für Molekulare Neurobiologie, Martinistrasse 52, 20246 Hamburg, Germany
Abstract
Association of Shal gene-related voltage-gated potassium (Kv4) channels with cytoplasmic Kv channel interacting proteins (KChIPs) influences inactivation gating and surface expression. We investigated both functional and biochemical consequences of mutations in cytoplasmic N and C-terminal Kv4.2 domains to characterize structural determinants for KChIP interaction. We performed a lysine-scanning mutagenesis within the proximal 40 amino acid portion and a structure-based mutagenesis in the tetramerization 1 (T1) domain of Kv4.2. In addition, the cytoplasmic Kv4.2 C-terminus was truncated at various positions. Wild-type and mutant Kv4.2 channels were coexpressed with KChIP2 isoforms in mammalian cell lines. The KChIP2-induced modulation of Kv4.2 currents was studied with whole-cell patch clamp and the binding of KChIP2 isoforms to Kv4.2 channels with coimmunoprecipitation experiments. Our results define one major interaction site for KChIPs, including amino acids in the proximal N-terminus between residues 11 and 23, where binding and functional modulation are essentially equivalent. A further interaction site includes residues in the T1 domain. Notably, C-terminal deletions also had marked effects on KChIP2-dependent gating modulation and KChIP2 binding, revealing a previously unknown involvement of domains within the cytoplasmic Kv4.2 C-terminus in KChIP interaction. Less coincidence of binding and functional modulation indicates a more loose ‘anchoring’ at T1- and C-terminal interaction sites. Our results refine and extend previously proposed structural models for Kv4.2/KChIP complex formation.
, 百拇医药
Introduction
Voltage-gated potassium channels of the Kv4-subfamily represent the molecular substrate of two transient currents of known physiological relevance: First, Kv4 channels mediate the neuronal somatodendritic A-type current (Serodio et al. 1994), which controls regular discharge behaviour as well as dendritic signal integration. Secondly, they mediate the transient outward current (Ito) in cardiomyocytes (Dixon et al. 1996), which participates in the early repolarization phase of the heart action potential.
, 百拇医药
Kv4 channels may coassemble with membrane-bound CD26-related dipeptidyl aminopeptidase-like proteins (Nadal et al. 2003), and with cytoplasmic Kv channel interacting proteins (KChIPs; An et al. 2000), which belong to the neuronal calcium sensor (NCS)-superfamily of EF-hand proteins (Burgoyne & Weiss, 2001). In both brain and heart the biophysical properties and surface expression levels of Kv4 channels may be controlled by these accessory subunits. This paper focuses on the interaction between Kv4 channels and KChIPs, which are integral components of brain Kv4 channel complexes (An et al. 2000; Nadal et al. 2003) and essential for the expression of Ito in rodent heart (Kuo et al. 2001).
, http://www.100md.com
In heterologous expression systems KChIPs have been shown to influence Kv4 channel assembly, phosphorylation status and stability (Shibata et al. 2003; Kunjilwar et al. 2004). KChIPs may enhance surface expression of Kv4 channels through a subcellular redistribution from perinuclear regions to the plasma membrane, which is reflected by an increase in the measured current density (An et al. 2000; Bhring et al. 2001). KChIPs may also modulate Kv4 channel inactivation kinetics (An et al. 2000; Bhring et al. 2001; Beck et al. 2002). This modulation is characterized by a slowing of the initial fast inactivation component of Kv4 channel-mediated currents and by an acceleration of recovery from inactivation.
, 百拇医药
KChIPs bind to the cytoplasmic N-terminus of Kv4 -subunits (An et al. 2000). Certain residues within the proximal 40 amino acid portion preceding the Kv4 T1 domain must be essential for KChIP-mediated gating modulation and KChIP binding, since Kv4.22–40 channels show neither increased current densities nor a modulation of inactivation kinetics in the presence of KChIP2.2 (Bhring et al. 2001). Furthermore, in vitro binding of KChIP2.2 is attenuated if the first 20, but not if the first 10 amino acids of Kv4.2 N-terminal fragments are deleted (Bhring et al. 2001). These results suggested that the proximal Kv4.2 N-terminus harbours a major KChIP interaction site, and that KChIP binding to this site is required for the effects on channel trafficking and gating.
, 百拇医药
It has been shown that effective association of Kv4.3 and KChIP2 does not require the variable and often alternatively spliced KChIP2 N-terminus, but critically depends on regions within the core domain of KChIP2 and its C-terminus (Ren et al. 2003). Crystal structure analyses have indicated that the KChIP core domain possesses a hydrophobic crevice (Scannevin et al. 2004; Zhou et al. 2004). This crevice may represent a binding pocket for hydrophobic residues in the proximal Kv4 N-terminus as indicated by the crystal structure of a KChIP1–Kv4.2 fusion protein, which contains the first 30 amino acids of Kv4.2 (Zhou et al. 2004).
, http://www.100md.com
In addition to the proximal Kv4 N-terminus, a cytoplasmic loop region on the N-terminal side of the T1 domain may play a role in KChIP interaction (Scannevin et al. 2004). However, correlating the effects of KChIP1 coexpression on Kv4.2 inactivation gating and the ability of KChIP1 to bind to Kv4.2 N-terminal fragments in a yeast two-hybrid assay showed some inconsistencies. Certain T1-domain mutations still allowed KChIP1-induced modulation of inactivation while interfering with KChIP1 binding (Scannevin et al. 2004). Such findings suggest that KChIP binding and coassembly with Kv4 -subunits may not be an absolute requirement for the observed KChIP-induced modulation of Kv4 channel gating in coexpression experiments. Alternatively, there may be additional, as yet undetected KChIP interaction sites on Kv4 channels.
, http://www.100md.com
Here we used naturally assembled and fully functional Kv4.2 channel constructs to study structural determinants for KChIP interaction. We focused on three regions of the Kv4.2 -subunit: The proximal Kv4.2 N-terminus, the Kv4.2 T1 domain and the cytoplasmic Kv4.2 C-terminus. The consequences of mutations or deletions within these regions were studied with both electrophysiological and immunocytochemical methods. KChIP-mediated Kv4.2 gating modulation and KChIP-induced changes in functional Kv4.2 surface expression were assessed with whole-cell patch-clamp recordings and direct Kv4.2/KChIP binding with immunoprecipitation (IP) experiments. The data reveal that in all three regions examined critical sites exist, that are important for functional Kv4.2/KChIP interaction.
, http://www.100md.com
Methods
Clones and channel constructs
We tested human KChIP1 (AF199597), KChIP2.1 (AF199598), KChIP2.2 (Bhring et al. 2001) and KChIP2.8 (NM_173194) coexpressed with different constructs of human Kv4.2 (Zhu et al. 1999). We performed a scanning-mutagenesis of the proximal Kv4.2 N-terminus (individual replacement of residues 8–30 by lysine or glutamate; see Figs 2 and 3 and text). Site-directed mutagenesis in the Kv4.2 T1 domain was based on its three-dimensional structure (Nanao et al. 2003; see Fig. 5E examined with the program InsightII (Accelrys). We introduced point mutations at exposed T1-domain sites (mutation of residues 64, 66, 69, 70, 71, 74, 76, 79 and 81 to lysine, glutamate or alanine; see Fig. 5 and text). Finally, we generated the C-terminal deletion mutants Kv 4.2 420, 435, 446, 461, 478, 490 and 580 (numbers indicate the last amino acid residue of the respective construct). All Kv4.2 constructs carried an extracellular haemagglutinin (HA)-tag (not further indicated) between transmembrane segments S1 and S2, which had no influence on channel function or KChIP-mediated gating modulation (not shown). All constructs were subcloned into the pcDNA3 mammalian expression vector (Invitrogen) and verified by sequencing. In the text, individual amino acids with residue numbers are given in three-letter code and individual point mutations in single-letter code. Since all mutations in this study were introduced in Kv4.2 this will not be indicated further.
, http://www.100md.com
Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (n-fold slowing of inactivation; A), the kinetics of recovery from inactivation (n-fold acceleration of recovery from inactivation; B) and the peak current density at +40 mV (n-fold increase in current density; C). Data are shown for Kv4.2 wild-type (wt) and for the individual scanning mutants of the proximal N-terminus. Open bars indicate significant differences, filled bars no significant differences, after coexpression of KChIP2.2 (see also Methods and Table 1). Horizontal broken lines indicate KChIP2.2 effect on wild-type and no effect (n-fold = 1), respectively. Error bars are the standard error.
, http://www.100md.com
A, results obtained with immunoprecipitation (IP) after heterologous coexpression of KChIP2.1 and individual scanning mutants of the proximal Kv4.2 N-terminus in CHO cells. Prior to immunoprecipitation (pre-input) western-blot analysis was performed to quantify the amount of protein expressed. After adjustment of input amount and volume (not shown) KChIP2.1 and HA-tagged Kv4.2 proteins were coimmunoprecipitated and detected with panKChIP (anti-panKChIP) and HA antibodies (anti-HA). Note the impairment of binding for mutations within a stretch of amino acids including Phe11 and Ala23, which shows a discontinuity at Arg13 and Ala16. In the control lane an unspecific band was occasionally detected by the panKChIP antibody. B, structural modelling of the N-terminal residues, mutated and examined in the present study, onto the crystal structure of a KChIP1–4.2N30 fusion protein (Zhou et al. 2004). KChIP1 portion in CPK (EF-hands in bronze colour) and Kv4.2 portion in ribbon presentation. Residues in the flanking regions not involved in KChIP binding as well as Arg13 and Ala16, which reside within the mutation-sensitive domain but exhibit strong coimmunoprecipitation signals, are labelled green. The remaining residues, which are critical for KChIP binding, are labelled red.
, 百拇医药
Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (A), the kinetics of recovery from inactivation (B) and the peak current density measured at +40 mV (C). Data are shown for all tested Kv4.2 T1-domain mutants. D, results obtained with immunoprecipitation (IP) after heterologous coexpression of KChIP2.1 and Kv4.2 T1-domain mutants. Note the apparent lack of binding for L66D, S69D and F74K. E, structural modelling of residues mutated and examined in this study (Asp64, Leu66, Ser69, Ser70, Glu71, Phe74, Glu79 and Gln81) on the Kv4.2 T1-domain crystal structure (Nanao et al. 2003); Zn2+ coordination residues in bronze colour). Only two opposite T1 monomers are shown from a side view. Green residues: mutation does not interfere with KChIP2.1 binding; red residues: mutation leads to a loss of binding as inferred from a lack of coimmunoprecipitation; blue residue: mutation produced non-functional channels. Error bars are the standard error.
, 百拇医药
Heterologous expression
Kv4.2 channels, either alone or in combination with KChIP, were expressed in human embryonic kidney (HEK) 293 and Chinese hamster ovary (CHO) cells as previously described (Bhring et al. 2001). Cells were plated at a density of 3–4 x 104 per 35 mm dish and transiently transfected with cDNA using Lipofectamine (Invitrogen). For the kinetic analyses of inactivation gating, done in HEK 293 cells, we used either 1 μg Kv4.2 cDNA alone or 0.1 μg Kv4.2 in combination with 1 μg KChIP2.2 cDNA per 35 mm dish. For the current density measurements, done in CHO cells, we used 0.1 μg Kv4.2 both in the absence and presence of 1 μg KChIP2.2 cDNA. In all cases 0.5 μg EGFP cDNA (Clontech) was cotransfected, or 1.5 μg of a bicistronic cDNA vector encoding both KChIP2.2 and EGFP was used to identify successfully transfected cells. No functional expression after transient transfection, both in the absence and in the presence of KChIP2.2, was obtained with G18K, S24K, Y76A, Y76V, 420, 446 and 461. These mutants were not used for the analysis of KChIP binding. For immunoprecipitation (IP) experiments CHO cells were plated at a density of 4–5 x 105 per 50 mm dish and transiently transfected with 2 μg Kv4.2 and 2 μg KChIP2 cDNA per dish using FuGENE6 (Roche).
, 百拇医药
Electrophysiological analyses were done with KChIP2.2, because in our hands this variant yields the strongest effects on Kv4.2 functional parameters (B.C. and R.B., unpublished observation). IP experiments were performed mostly with KChIP2.1, because we noticed that KChIP2.2 co-migrated with an unspecific band occasionally detected in control lanes (see Fig. 3A). KChIP2.1 and KChIP2.2 only differ in their alternatively spliced N-termini (Bhring et al. 2001), a region of minor importance for the binding to Kv4 -subunits (Ren et al. 2003).
, 百拇医药
Electrophysiological recordings and data analysis
Cells were continuously superfused at room temperature (20–22°C) with a solution containing (mM): NaCl 135, KCl 5, CaCl2 2, MgCl2 2, Hepes 5, sucrose 10 and phenol red 0.01 mg ml–1; pH 7.4 (NaOH). Recordings were performed in the whole-cell configuration of the patch-clamp technique. Patch-pipettes were pulled from thin-walled borosilicate glass using a DMZ-puller (Zeitz). Pipette-to-bath resistances ranged between 2 and 3 M with internal solution containing (mM): KCl 125, CaCl2 1, MgCl2 1, EGTA 11, Hepes 10, K2ATP 2 and glutathione 2; pH 7.2 (KOH). Currents were recorded using an EPC9 patch-clamp amplifier controlled by PULSE software (HEKA) with maximal series resistance compensation (usually between 80 and 90%). Signals were filtered at 0.2–4 kHz with low-pass Bessel characteristics, amplified as required, and digitized at 50–1000 μs sample intervals. Capacitive current transients were compensated on-line, and the P/n method was applied for leak subtraction. The program PULSEFIT (HEKA) was used to analyse current traces, and the obtained data were further processed using Kaleidagraph (Synergy Software). We analysed the following functional parameters: the extent of fast inactivation was quantified by the current amplitude remaining 30 ms after the activating voltage-step relative to the peak current amplitude (I30/Ipeak; see Figs 1, 4 and 6). The kinetics of recovery from inactivation were measured at –80 mV and fitted with a single-exponential function. As a further electrophysiological parameter we calculated, from the peak current amplitude measured at +40 mV and the whole-cell capacitance, the current density (pA pF–1). Testing more than only one functional parameter helped to eliminate false negatives for the judgement of functional KChIP2.2 interaction, since in many cases the experimentally introduced channel mutation caused extremely low current expression in the absence of KChIP2.2, leading to analysis problems and small sample sizes. Pooled data are presented as mean ± S.E.M., and statistical analyses were performed using GraphPad Prism (GraphPad Software; see text and Table 1). Differences between the mean values obtained in the absence (meana ± S.E.M.a) and presence of KChIP2.2 (meanb ± S.E.M.b) were examined for significance with unpaired Student's t tests (one-tailed P-values and Welch-correction for unequal variances; Table 1). KChIP effects on Kv4.2 current parameters were quantified as the average n-fold change (n = meanb/meana for I30/Ipeak and current density; n = meana/meanb for recovery from inactivation) induced by KChIP2.2 coexpression. Assuming linear and independent error propagation the standard error S.E.M.n of this quotient was calculated to be (see Figs 2, 5 and 7).
, http://www.100md.com
Left panels show currents mediated by Kv4.2 wild-type (wt; A1), W8K (B1) and F11K (C1) channels expressed in HEK 293 cells in the absence or presence of KChIP2.2. Currents were measured in the whole-cell configuration during depolarizing test-pulses from –100 to +40 mV. Inactivation of A-type currents was analysed and plotted (middle panels) as the relative current remaining 30 ms after the beginning of the test pulse (I30/Ipeak, vertical broken lines), in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels show the effects of KChIP2.2 coexpression on the kinetics of recovery from inactivation for wt (A2), W8K (B2) and F11K (C2). Data were plotted on a semilog scale and fitted with single-exponential functions. , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, http://www.100md.com
Left panels, currents mediated by 435 (A1), 478 (B1), 490 (C1) and 580 (D1) in the absence and presence of KChIP2.2. Middle panels, relative amount of inactivation after 30 ms in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels, effect of KChIP2.2 coexpression on the kinetics of recovery from inactivation for 435 (A1), 478 (B1), 490 (C1) and 580 (D1). , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, 百拇医药 Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (A), the kinetics of recovery from inactivation (B) and the peak current density at +40 mV (C). D, results obtained with immunoprecipitation (IP) experiments after heterologous coexpression of KChIP2.1 and Kv4.2 mutants with different C-terminal truncations. Note the strong impairment of KChIP2.1 binding for constructs shorter than 580, with an apparent regain of KChIP binding ability for 435. E, IP was repeated with 580, 490 and 478 after coexpression with KChIP1, KChIP2.2 or KChIP2.8. Note that all tested KChIP variants yielded the same results. Error bars are the standard error.
, 百拇医药
Binding experiments
From CHO cells transfected with Kv4.2 and KChIP cDNA total cell lysates were prepared (lysis buffer: 150 mM NaCl, 50 mM Hepes pH 7.4, 0.5% Triton) in the presence of a protease inhibitor cocktail (Sigma). After pelleting the insoluble matter (10 000 g; 10 min) equal amounts of the supernatant were separated by 4–12% gradient SDS-PAGE and transferred to nitrocellulose membranes. After a blocking reaction (PBS; 0.05% Tween; 5% non-fat milk powder) affinity-purified HA- (1 : 500, 3F10; Roche) or panKChIP-antibody (1 : 600) was incubated overnight. After washing the membrane, horseradish-peroxidase-conjugated goat antibody (1 : 5000; Jackson ImmunoResearch) was incubated for 1 h. Enhanced chemiluminescence reagents were used for signal detection. To quantify the amount of protein expressed, signal intensities were evaluated with the program Quantify One (Bio-Rad).
, http://www.100md.com
Equal amounts of protein were immunoprecipitated by addition of HA- or panKChIP-antibody at a 10 μg ml–1 final concentration overnight. Mixtures were incubated with ProteinG Dynabeads (Dynal Biotech) for 2 h, and immunoprecipitates were washed twice with lysis buffer and eluted with Nupage sample buffer (Invitrogen) under reducing conditions (10 min at 70°C). With the supernatant, western blot analysis was done as described.
Results
, 百拇医药
Mutant scanning analysis of the proximal Kv4.2 N-terminus
In a previous study (Bhring et al. 2001) we have shown that efficient in vitro binding of KChIP2.2 to N-terminal Kv4.2 fragments tolerates a deletion of the first 10 (2–10) but not a deletion of the first 20 amino acids (2–20). Also, in contrast to Kv4.2 wild-type and 2–10, 2–20-mediated current densities were no longer increased in the presence of KChIP2.2 (Bhring et al. 2001). The data suggested that the N-terminal Kv4.2 sequence between amino acid residues 10 and 20 was critical for KChIP2.2 interaction. In order to characterize the N-terminal Kv4.2 interaction domain more precisely we now performed a lysine-scanning mutagenesis of the proximal Kv4.2 N-terminus between residues Trp8 and Pro30 with the positively charged Arginine at position 13 mutated to glutamate (R13E).
, http://www.100md.com
Mutant Kv4.2 channels were first tested for their functionality and then coexpressed with KChIP2.2. G18K and S24K yielded no functional expression in the absence and presence of KChIP2.2. However, the respective glutamate mutants G18E and S24E expressed currents and were used in this scanning analysis. Typically, KChIP2.2 slows the initial onset of inactivation, accelerates recovery from inactivation and increases Kv4.2 current densities. We first analysed these parameters in the absence and presence of KChIP2.2 to assess potential effects of the N-terminal point mutations on KChIP interaction. We analysed fast Kv4.2 inactivation by measuring the ratio I30/Ipeak (see Methods; Fig. 1). For Kv4.2 wild-type channels expressed in the absence of KChIP2.2 I30/Ipeak was 0.27 ± 0.02 (n = 3). In agreement with previous results (Bhring et al. 2001), coexpression of Kv4.2 wild-type with KChIP2.2 increased I30/Ipeak to 0.58 ± 0.04 (n = 3; P = 0.0101) due to a slowing of the initial current inactivation phase (Fig. 1A1), and accelerated recovery from inactivation (rec = 290 ± 43 ms, n = 3 in the absence and 39 ± 3 ms, n = 4 in the presence of KChIP2.2; P = 0.0141; Fig. 1A2). Table 1 summarizes current parameters determined for Kv4.2 channel mutants in the absence and presence of KChIP2.2. Effects of KChIP2.2 coexpression on the Kv4.2 current parameters expressed as n-fold change (see Methods) are shown in Fig. 2.
, 百拇医药
More than 50% of the proximal mutations (L9K, P10K, R13E, A16K, A23K, S24E, G25K, P26K, P28K, M29K and P30K) generated Kv4.2 channels with current parameters affected by KChIP2.2 like wild-type (Fig. 2). Thus, many N-terminal point mutations did not alter the interaction of Kv4.2 channels with KChIP2.2, indicative of local rather than long distance mutational effects. Also W8K appeared to interact with KChIP2.2. Notably, W8K inactivation kinetics were slower than wild-type in the absence of KChIP2.2 (I30/Ipeak = 0.81 ± 0.03, n = 3), however, in the presence of KChIP2.2 both onset of (I30/Ipeak = 0.60 ± 0.04, n = 7; Fig. 1B1) and recovery from inactivation (rec = 50 ± 6 ms, n = 5; Fig. 1B2) were comparable to wild-type coexpressed with KChIP2.2 (Table 1).
, 百拇医药
Only four point mutations (F11K, G18E, P21K and V22K) produced Kv4.2 channels with current parameters that remained unchanged upon KChIP2.2 coexpression (Fig. 2), indicating that these mutations impaired the KChIP2.2-induced modulation of Kv4.2 currents. Coexpression of F11K with KChIP2.2 changed neither onset of (I30/Ipeak 0.70 ± 0.06, n = 6 in the absence and 0.67 ± 0.06, n = 6 in the presence of KChIP2.2; P = 0.2034; Fig. 1C1) nor, significantly, recovery from inactivation (rec = 401 ± 69 ms, n = 5 in the absence and 285 ± 41, n = 4 in the presence of KChIP2.2; P = 0.0993; Fig. 1C2).
, http://www.100md.com
The remaining point mutations (A12K, A14K, A15K, I17K, W19K, M20K and M27K) showed intermediate KChIP sensitivity; i.e. coexpression with KChIP2.2 significantly influenced only one or two of the measured current parameters (Table 1). Although significant, these changes were always smaller than those obtained for wild-type (Fig. 2). Thus, these mutations also affected the KChIP2.2-induced Kv4.2 current modulation.
A likely explanation for our mutational results is that the Kv4.2 N-terminus contains a KChIP2.2-binding domain between residues 11 and 22, and that single point mutations in this binding domain interfere with KChIP binding. We tested this hypothesis by assaying direct KChIP binding to each of the tested Kv4.2 constructs in IP experiments (Fig. 3A). A15K and P28K were not investigated due to their extremely low expression levels. Signal intensities for coimmunoprecipitated Kv4.2 and KChIP2.1 polypeptides were similar to wild-type controls if the point mutation had been introduced N-terminal to Phe11 (W8K, L9K, P10K) or C-terminal to Val23 (A24K and further downstream); i.e. outside the proposed KChIP-binding domain in the Kv4.2 N-terminus (Fig. 3A). By contrast, signal intensities for coimmunoprecipitated Kv4.2 and KChIP2.1 polypeptides were significantly weaker if the point mutations were located within the proposed KChIP-binding domain, except for R13E and A16K (Fig. 3A). These data are in agreement with X-ray structural analysis of a KChIP1–Kv4.2N30 fusion protein (Zhou et al. 2004), in which the side chains of Arg13 and Ala16 do not make contact with the KChIP surface (Fig. 3B; see Discussion).
, 百拇医药
Taken together, our IP and electrophysiology results show a good correlation regarding binding strength and KChIP-induced changes in current properties. This implies that KChIP2.2 modulates Kv4.2 inactivation kinetics and current density through a direct interaction with a binding domain between residues 11 and 23 of the Kv4.2 N-terminus. Thus, our data extend previous KChIP1–Kv4.2N30 X-ray analysis (Zhou et al. 2004), which did not yield atomic coordinates for residues 21–23, leaving the C-terminal border of the proximal KChIP-binding domain undefined.
, http://www.100md.com
Role of exposed amino-acids in the Kv4.2 T1 domain
Ca2+-binding proteins like calmodulin may accommodate in their binding pocket more than one -helix (Schumacher et al. 2001). We hypothesized that KChIPs may interact in a similar fashion with more than one Kv4.2 domain.
In the light of a report that certain regions of the Kv4.2 T1 domain may also be critical for KChIP binding (Scannevin et al. 2004), we examined if the Kv4.2 T1 domain possesses KChIP interaction sites. We used crystal structure information (Nanao et al. 2003) to mutate residues exposed on the cytoplasmic surface of the Kv4.2 T1 domain (Fig. 5E).
, http://www.100md.com
Within the cytoplasmic loop-region of the Kv4.2 T1 domain we mutated Tyr76, Glu79 and Gln81 (Fig. 5E). Neither Y76A nor Y76V yielded functional expression in the absence or presence of KChIP2.2. E79A current parameters including onset of (I30/Ipeak = 0.32 ± 0.03, n = 4 in the absence and 0.47 ± 0.02, n = 6 in the presence of KChIP2.2; p = 0.0044; Fig. 4A1) and recovery from inactivation (rec = 247 ± 14 ms, n = 4 in the absence and 42 ± 2, n = 5 in the presence of KChIP2.2; P = 0.0004; Fig. 4A2) were modulated by coexpression of KChIP2.2 comparable to wild-type parameters, and similar results were obtained for Q81A (Fig. 5A and B, Table 1).
, http://www.100md.com
Left panels, currents mediated by E79A (A1), L66D (B1), S69D (C1) and F74K (D1), in the absence and presence of KChIP2.2. Middle panels, relative amount of inactivation after 30 ms in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels, effect of KChIP2.2 coexpression on the kinetics of recovery from inactivation for E79A (A2), L66D (B2), S69D (C2) and F74K (D2). , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, 百拇医药
Next, we introduced individual lysine residues, in analogy to our terminal scanning analysis, for residues Asp64, Leu66, Ser70, Glu71 and Phe74 (D64K, L66K, S70K, E71K, F74K), which are more laterally located with respect to the cytoplasmic loop region of the Kv4.2 T1 domain. In this region we also mutated some residues to aspartate (L66D, S69D, S70D) in order to detect possible charge-dependent effects on KChIP interaction. Most of these Kv4.2 mutants showed current properties like wild-type in the absence or presence of KChIP2.2 (Table 1). Normal KChIP-induced current modulation is shown for L66D and S69D in Fig. 4B and C, respectively. In the case of F74K and L66K, coexpression of KChIP2.2 slowed the onset of inactivation comparable to wild-type (F74K: I30/Ipeak = 0.36 ± 0.04, n = 3 in the absence and 0.50 ± 0.04, n = 3 in the presence of KChIP2.2; P = 0.0343; L66K: I30/Ipeak = 0.20 ± 0.02, n = 3 in the absence and 0.55 ± 0.05, n = 10 in the presence of KChIP2.2; P < 0.0001; Figs 4D1 and 5A), but only caused a weak or insignificant acceleration of recovery from inactivation (F74K: rec = 254 ± 15 ms, n = 5 and 166 ± 25 ms, n = 6; P = 0.0097; L66K: rec = 232 ± 39 ms, n = 3 and 155 ± 23 ms, n = 9 in the absence and presence of KChIP2.2, respectively; P = 0.0938; Figs 4D2 and 5B).
, 百拇医药
Figure 5C illustrates the relative increases in current density observed for the T1-domain mutants when coexpressed with KChIP2.2. Current density increases were smallest in the case of L66K and S70D (5- and 15-fold, respectively). In contrast to the proximal mutations, the Kv4.2 T1-domain mutations did not completely attenuate KChIP2.2-induced current modulation, but produced more intermediate effects.
Next, we investigated whether the attenuation of KChIP2.2 effects on Kv4.2 T1-domain mutants correlated with an impairment of KChIP binding. Coimmunoprecipitation signal intensities were comparable to wild-type for most T1-domain mutants except L66D, S69D, and F74K. This result was unexpected, since current properties of the latter three mutants were modulated by KChIP2.2 (Fig. 5A–C). Thus, our IP data suggest that T1-domain residues Leu66, Ser69 and Phe74 may be critical for KChIP binding, however, the results of our electrophysiological and binding experiments did not strictly correlate (see Discussion).
, http://www.100md.com
Mutational analysis of the cytoplasmic Kv4.2 C-terminus
Previously, it was shown by mutational deletion that Kv4.1 channel inactivation is mediated by the concerted action of cytoplasmic N- and C-terminal domains (Jerng & Covarrubias, 1997). Since Kv4.2 N-terminal deletion mutants show no functional modulation by KChIP we now examined if C-terminal deletions of varying lengths may similarly influence functional KChIP interaction. The initial onset of inactivation of 435-mediated currents was on average slightly decelerated after coexpression with KChIP2.2 (I30/Ipeak = 0.25 ± 0.03, n = 3 in the absence and 0.38 ± 0.03, n = 9 in the presence of KChIP2.2; P = 0.0111; Figs 6A1 and 7A). However, recovery of 435 from inactivation was not accelerated (rec = 364 ± 29 ms, n = 5 in the absence and 343 ± 18 ms, n = 8 in the presence of KChIP2.2; P = 0.3099; Figs 6A2 and 7B) and 435 current densities were not increased by KChIP2.2 (Fig. 7C). Investigating less severe C-terminal deletions showed that the effects of KChIP2.2 coexpression on Kv4.2 current properties became gradually larger (Fig. 7B and C). In the case of 478, KChIP2.2 seemed to slightly enhance both onset of (I30/Ipeak = 0.32 ± 0.05, n = 7 in the absence and 0.21 ± 0.03, n = 12 in the presence of KChIP2.2; P = 0.0443; Figs 6B1 and 7A) and recovery from inactivation (rec = 256 ± 17, n = 10 in the absence and 165 ± 9 ms, n = 9 in the presence of KChIP2.2; P = 0.0002; Figs 6B2 and 7B). However, also for this mutant no significant increase in current density was observed (Fig. 7C). 490 channels tended to show stronger KChIP2.2 effects (Figs 6C and 7A–C), but only 580 channels showed pronounced KChIP2.2 effects on inactivation kinetics and current density, similar to the ones observed for wild-type (Figs 6D and 7A–C).
, 百拇医药
Next, we examined KChIP2.1 binding to the Kv4.2 C-terminal deletion mutants in IP experiments (Fig. 7D). The results showed that only 580 channels were efficiently coimmunoprecipitated with KChIP2.1 showing signal intensities comparable to wild-type. Neither with HA- nor with panKChIP-antibodies did we obtain signals with the deletion mutants 478 and 490, but the shortest deletion mutant 435 yielded a weak signal when immunoprecipitated with panKChIP-antibodies (Fig. 7D). For Kv4.2 wild-type, 580, 490, and 478 we performed further IP experiments together with KChIP1, KChIP2.2 and KChIP2.8, respectively (Fig. 7E). The outcome of these experiments was similar for all KChIP variants used; i.e. prominent coimmunoprecipitation signals comparableto wild-type were only observed with 580. Thus, effects of C-terminal deletions on binding of KChIPs to Kv4.2, and modulation of Kv4.2 current properties by KChIP2.2 did not strictly correlate, very similar to the observations we made with certain T1-domain mutants.
, 百拇医药
Discussion
In this study we carried out a mutational analysis of the cytoplasmic Kv4.2 N- and C-termini to investigate structural determinants for channel modulation by accessory KChIP2 subunits. For each Kv4.2 mutant the KChIP2-induced modulation of current properties and KChIP2 binding was tested. Thus, only properly assembled and fully functional constructs were analysed. The results suggest that some mutations affect a major Kv4.2/KChIP2-binding interface and therefore directly interfere with KChIP2-induced modulation of Kv4.2 current properties. Other mutations may have a more indirect effect on functional interaction. Previously, the interaction of Shaker-related Kv1 channels with their accessory Kv-subunits has been studied in great detail (Gulbis et al. 2000). Our results demonstrate that the Kv4.2 channel domains involved in KChIP-dependent gating modulation are disitinct from those in Shaker-related Kv1 channels for Kvinteraction. The differences in the mode of interaction for the two types of accessory subunits reflect the distinct inactivation mechanisms and functional roles demonstrated for Kv1/Kvand Kv4/KChIPs channel complexes, respectively.
, 百拇医药
The cytoplasmic Kv4.2 N-terminus harbours multiple KChIP interaction sites
The spacing of N-terminal Kv4.2 amino acids, mutation of which affected KChIP binding and current modulation, suggests that the KChIP-binding interface is -helical. According to our data, residues Phe11, Ala14 and Trp19 of the N-terminal Kv4.2 -helix mainly contribute to this binding interface. This view is consistent with the crystal structure of a KChIP1–Kv4.2N30 fusion protein (Zhou et al. 2004). The structure shows the N-terminal Kv4.2 domain as an -helix embedded in the hydrophobic crevice of KChIP1, and side chains of Kv4.2 amino acid residues Trp8 and Phe11 appear to make direct contact with KChIP1 (Zhou et al. 2004). Possibly the contact between Trp8 and KChIP1 seen in the crystal structure is not important for KChIP binding, because our data show that W8K channels bound KChIP2.1 in a similar way to wild-type. Our data, which were obtained with a drastic amino acid substitution, clearly demarcate Phe11 as the N-terminal border of the KChIP-binding interface in Kv4.2, consistent with our previous finding that deletion of the first 10 amino acids did not affect KChIP interaction with Kv4.2 channels (Bhring et al. 2001). The fact that hydrophobicity was not changed dramatically by a mutation of Phe11 to alanine may explain why it was tolerated in a previous study (Scannevin et al. 2004). In addition, our results reveal Kv4.2 residues relevant for KChIP2 interaction near the C-terminal border of the interface (Pro21, Val22, Ala23 and possibly Met27). In this region the crystal structure of the KChIP1–Kv4.2N30 fusion protein lacks atomic resolution (Zhou et al. 2004).
, http://www.100md.com
It is intriguing that neither R13E, which we have analysed in a previous study (Gebauer et al. 2004), nor proximal lysine-scanning in Kv4.2 had a pronounced effect on inactivation per se. The results corroborate the view that, unlike classical models of Shaker/Kv1 N-type inactivation, electrostatic interactions mediated by charged amino acid residues play a minor role in Kv4.2 N-type inactivation (Gebauer et al. 2004). In this context, our results obtained with W8K deserve further consideration: Obviously, in this mutant the introduction of a positively charged amino acid for a hydrophobic one did have a strong influence on inactivation. In the absence of KChIP2.2, W8K channels inactivated markedly more slowly than wild-type. In the presence of KChIP2.2, the initial slow decay of W8K currents was accelerated to a value similar to the one observed for Kv4.2 wild-type coexpressed with KChIP2.2 (see Fig. 1), which may reflect the previously described ‘streamlining’ of Kv4-mediated A-type currents by KChIPs (Beck et al. 2002). Recovery of W8K channels from inactivation showed the characteristic KChIP-induced acceleration. In experiments where we coexpressed the slowly inactivating N-terminal deletion mutant 2–10 with KChIP2.2 we observed no change in the onset, but normal acceleration of recovery from inactivation (data not shown). Similar findings were obtained previously for the functional interaction between KChIP1 and N-terminally GFP-tagged Kv4.3 channels (Hatano et al. 2002), where the GFP-tag only compromises fast inactivation and its modulation by KChIP but not the KChIP-induced acceleration of recovery from inactivation. The data are difficult to reconcile with a gating model, in which KChIP binds to and immobilizes the proximal Kv4 N-terminus to modulate all aspects of inactivation (Beck et al. 2002).
, 百拇医药
The interaction between Kv4 channels and KChIPs seems to be more complex, as indicated by the observation that additional domains of the cytoplasmic Kv4.2 N-terminus may contribute to an effective KChIP-induced channel modulation (Scannevin et al. 2004). A Kv4.2 N-terminal fragment (amino acids 1–180) with an internal deletion (71–90) showed no interaction with KChIP1 in a yeast two-hybrid assay (Scannevin et al. 2004). The deletion eliminated residues normally located in a cytoplasmic loop region on the N-terminal side of the Kv4.2 T1 domain (Nanao et al. 2003). In Kv1.4 channels this T1-loop region is important for the functional interaction with accessory Kv-subunits (Gulbis et al. 2000). Intersubfamily protein sequence alignments suggest that the critical phenylalanine in Kv1.4, although replaced by a tyrosine in Kv4.1 and Kv4.2, is conserved in Kv4.3. Given the high similarity of the KChIP-induced effects observed with the different wild-type Kv4 subfamily members (Bhring et al. 2001), it seems unlikely that the same T1-loop residues as in Kv1/Kvchannel complexes play a central role in KChIP-mediated Kv4 channel modulation. Nevertheless, in a Kv1.2 N-terminal fragment substitution of the T1-loop sequence by the respective Kv4.2 sequence (amino acids 71–90) conferred the ability to bind KChIP1 in the yeast two-hybrid assay (Scannevin et al. 2004). The conclusion drawn from these results was that the loop region in the Kv4.2 T1 domain contains key residues for KChIP binding (Scannevin et al. 2004). We mutated residues directly within as well as proximal to the T1 loop to investigate potential effects on KChIP-induced Kv4.2 current modulation or KChIP binding. With the exception of Y76A and Y76V, all T1-domain mutants expressed functional channels, excluding long-range mutational effects that may have caused folding and assembly problems. Our results clearly exclude Glu71 and Glu79 as important KChIP interaction sites. For other T1-domain sites, however, the data were not as clear-cut as for our proximal mutagenesis scan (see Figs 2, 3 and 5). T1-domain mutations affected KChIP2.2-induced Kv4.2 current modulation to variable extents, but the effects did not strictly correlate with the IP data. We only obtained a consistent picture for the mutant F74K: The loss of KChIP2-binding ability combined with only a very modest acceleration of recovery from inactivation observed for this mutant confirmed previous observations (Scannevin et al. 2004). Taken together, our data support the view that amino acids in the Kv4.2 T1 domain including the cytoplasmic loop region may be involved in KChIP interaction. The interaction seems to be not strong enough for coimmunoprecipitation and may be indirect.
, 百拇医药
Involvement of C-terminal cytoplasmic Kv4.2 domains in KChIP interaction
It has been shown previously that Kv4 channel inactivation is mediated by the concerted action of cytoplasmic N- and C-terminal domains (Jerng & Covarrubias, 1997). This may indicate that N- and C-terminal Kv4.2 domains interact with each other, as has been reported for the related Kv2.1 channels, where residues close to the cytoplasmic loop of the T1 domain interact with a cytoplasmic C-terminal domain (Ju et al. 2003). These previous findings combined with our experimental results obtained with Kv4.2 T1-domain mutants led us to hypothesize that the cytoplasmic Kv4.2 C-terminus may also play a role in mediating KChIP2 effects on Kv4.2 inactivation gating. Indeed, our experiments showed that deleting the Kv4.2 C-terminus proximal to amino acid residue 580 (490, 478 and 435) may attenuate the KChIP2.2 sensitivity of Kv4.2 currents. The results suggest that domain(s) in the Kv4.2 C-terminus are critical for KChIP2-induced Kv4.2 modulation. Consistent with the electrophysiological data, the IP experiments showed that 490 and 478 mutants did not form a stable complex with KChIPs. Notably, 435 and KChIP2.1 polypeptides were coimmunoprecipitated to some degree, although we did not observe a marked effect of KChIP2.2 on 435 current.
, 百拇医药
In the absence of additional data, it is certainly difficult to reconcile apparent inconsistencies between electrophysiological and IP results. One intriguing explanation would be that a palmitoylation-mediated anchoring of KChIPs in the plasma membrane (Takimoto et al. 2002) could lead to functional interaction with Kv4 channels without tightly binding to them. However, the human KChIP2.2 isoform used in this and our previous study (Bhring et al. 2001) lacks the N-terminal double cysteine motif necessary for palmitoylation. Alternatively, due to strong overexpression, the cytoplasmic content of KChIP2.2 may be high enough to promote interaction with Kv4.2 channels based on the law of mass action. In fact, such mass action effects have been reported previously for the promoting effect of KChIP3 on the assembly of Kv4.2 Zn2+-binding mutants (Kunjilwar et al. 2004). We favour the following explanation for our experimental results: mutations, where a loss of binding correlated with a complete loss of KChIP2.2 effects on Kv4.2 current properties, most likely affected directly the Kv4.2/KChIP2.2 interface. This was observed for mutations in the proximal Kv4.2 N-terminus. We propose therefore that the corresponding amino acids are key residues for the interaction of Kv4.2 with KChIP2. Mutations where a complete or partial loss of KChIP2.2 effects on Kv4.2 current properties did not coincide with a loss of KChIP2 binding, are likely to interfere more indirectly with KChIP-induced Kv4.2 current modulation. Finally, some mutations altered, to varying degrees, some aspects of KChIP2-induced Kv4.2 current modulation but at the same time showed a loss of KChIP binding. Possibly, the corresponding Kv4.2/KChIP2 complexes were not stable enough for coimmunoprecipitation. This explanation is consistent with our observation that in many of these cases one of the two antibodies we used still produced a coimmunoprecipitation signal. Thus, we propose that the respective deletions of the cytoplasmic C-terminus, similar to certain point mutations in the T1 domain, exert indirect effects on Kv4.2/KChIP2.2 interaction. T1 and C-terminal sites may represent anchoring domains, helping KChIPs to achieve an optimal orientation for docking to the major binding site between residues 11 and 23 of the Kv4.2 N-terminus. Mutation-induced poor KChIP anchoring may cause defective binding. This disruption may then lead to various degrees of KChIP-mediated channel modulation. Furthermore, different mechanisms of modulation affecting different current parameters may become decoupled. Although tightly bound to the N-terminus, KChIPs may not be anchored and therefore the complex is poorly constrained. As a result, fast inactivation would be slowed but recovery may not be affected (e.g. 435 in Fig. 7). Finally, disrupted KChIP anchoring may also cause incorrect or weaker KChIP binding to the N-terminus (Figs 5D and 7D and E).
, 百拇医药
Implications for the macromolecular structure of Kv4.2 channel complexes
Although there is strong evidence that Kv4.2/KChIP2 channels are octomeric complexes with a 4 : 4 stoichiometry (Kim et al. 2004a), the exact binding mode has remained unsolved. In principle, Kv4/KChIP assembly could be based on different mechanisms leading to the same stoichiometry: Based on the ability of KChIPs to tetramerize under certain conditions (Osawa et al. 2001), they could bind to Kv4 channels as a tetramer, similar to Kv-subunits on the bottom of the T1-tetramer (Gulbis et al. 2000); the binding data obtained by Scannevin et al. (2004) support this view. Alternatively, two KChIP molecules, each bound to a different Kv4 N-terminus, could dimerize, which eventually would result in a dimer-of-dimers organization, as has been shown previously for the binding of the KChIP-related protein calmodulin to Ca2+-activated SK channels (Schumacher et al. 2001); the structural analysis of the KChIP1–Kv4.2N30 fusion protein, which was crystallized as a dimer, supports this view (Zhou et al. 2004).
, 百拇医药
Single-particle electron-microscopy analysis of Kv4.2/KChIP2 complexes (Kim et al. 2004b) proposed that Kv4.2/KChIP2 channels are octomeric complexes, in which KChIPs bind to the channel as four individual protein units, each attached to one of the proximal N-termini and placed in a four-fold rotational symmetry lateral to the T1 domain. Furthermore, the cytoplasmic C-terminus of each Kv4.2 -subunit appears in close neighbourhood both to the T1 domain and bound KChIP (Kim et al. 2004b). The results of our mutational analyses strongly support these features of the proposed structure as outlined in Fig. 8. Native brain Kv4 channel complexes may contain not only cytoplasmic KChIPs but membrane-bound dipeptidyl aminopeptidase-like proteins as additional accessory subunits (Nadal et al. 2003). It remains to be seen how the presence of another accessory subunit may influence the interaction of KChIPs with the binding sites identified on the Kv4.2 -subunit. Eventually, high-resolution structural analyses based on cocrystallized complexes containing all relevant structural determinants will be necessary to elucidate the actual Kv4/KChIP binding mode.
, http://www.100md.com
A simple conceptual model was developed, which combines our experimental results with previous Kv4.2/KChIP2 single-particle electron-microscopy analysis (Kim et al. 2004b; contours of the published structure indicated by broken lines). The cartoon schematically illustrates two opposite Kv4.2 channel -subunits with KChIPs bound. The channel core domain; i.e. voltage-sensor (S1–S4) and pore domain (S5 and S6) are in grey, T1 domain in green, proximal N-terminus in orange and cytoplasmic C-terminus in red. KChIPs (blue) are attached laterally, making contact with both the proximal N-terminus and the T1 domain. The cytoplasmic C-terminus constitutes the lateral columns detected with electron microscopy (Kim et al. 2004b) and may come in close proximity to the T1 domain and the proximal N-terminus as well as bound KChIP.
, http://www.100md.com
In summary, the proximal Kv4.2 N-terminus constitutes the major Kv4.2/KChIP-binding interface. There, mutations directly interfere with KChIP binding and KChIP-mediated channel modulation. Mutations which we classified to indirectly interfere with KChIP binding and modulation were either located laterally on the T1 domain or within the cytoplasmic C-terminus of Kv4.2. The T1 domain and cytoplasmic C-terminus may represent KChIP anchoring sites. The data support a macromolecular structure for Kv4.2/KChIP2 complexes as revealed by single-particle electron-microscopy (Kim et al. 2004b). These findings imply that the Kv4.2 channel domains involved in KChIP-dependent gating modulation are disitinct from those found in Shaker-related Kv1 channels for Kvinteraction (Gulbis et al. 2000). The differences in the mode of interaction for the two types of accessory subunits reflect the distinct inactivation pathways and according functional roles demonstrated for Kv1/Kvand Kv4/KChIP channel complexes.
, 百拇医药
References
An WF, Bowlby MR, Betty M, Cao J, Ling HP, Mendoza G, Hinson JW, Mattsson KI, Strassle BW, Trimmer JS & Rhodes KJ (2000). Modulation of A-type potassium channels by a family of calcium sensors. Nature 403, 553–556.
Bhring R, Dannenberg J, Peters HC, Leicher T, Pongs O & Isbrandt D (2001). Conserved Kv4 N-terminal domain critical for effects of Kv channel-interacting protein 2.2 on channel expression and gating. J Biol Chem 276, 23888–23894.
, 百拇医药
Beck EJ, Bowlby M, An WF, Rhodes KJ & Covarrubias M (2002). Remodelling inactivation gating of Kv4 channels by KChIP1, a small-molecular-weight calcium-binding protein. J Physiol 538, 691–706.
Burgoyne RD & Weiss JL (2001). The neuronal calcium sensor family of Ca2+-binding proteins. Biochem J 353, 1–12.
Dixon JE, Shi W, Wang HS, McDonald CYuH, Wymore RS, Cohen IS & McKinnon D (1996). Role of the Kv4.3 K+ channel in ventricular muscle. A molecular correlate for the transient outward current. Circ Res 79, 659–668.
, 百拇医药
Gebauer M, Isbrandt D, Sauter K, Callsen B, Nolting A, Pongs O & Bhring R (2004). N-type inactivation features of Kv4.2 channel gating. Biophys J 86, 210–223.
Gulbis JM, Zhou M, Mann S & MacKinnon R (2000). Structure of the cytoplasmic beta subunit–T1 assembly of voltage-dependent K+ channels. Science 289, 123–127.
Hatano N, Ohya S & Imaizumi Y (2002). Functional interaction between KChIP1 and GFP-fused Kv4.3L co-expressed in HEK293 cells. Pflugers Arch 444, 80–88.
, 百拇医药
Jerng HH & Covarrubias M (1997). K+ channel inactivation mediated by the concerted action of the cytoplasmic N- and C-terminal domains. Biophys J 72, 163–174.
Ju M, Stevens L, Leadbitter E & Wray D (2003). The Roles of N- and C-terminal determinants in the activation of the Kv2.1 potassium channel. J Biol Chem 278, 12769–12778.
Kim LA, Furst J, Butler MH, Xu S, Grigorieff N & Goldstein SA (2004a). Ito channels are octomeric complexes with four subunits of each Kv4.2 and K+ channel-interacting protein 2. J Biol Chem 279, 5549–5554.
, 百拇医药
Kim LA, Furst J, Gutierrez D, Butler MH, Xu S, Goldstein SA & Grigorieff N (2004b). Three-dimensional structure of I(to); Kv4.2-KChIP2 ion channels by electron microscopy at 21 resolution. Neuron 41, 513–519.
Kunjilwar K, Strang C, DeRubeis D & Pfaffinger PJ (2004). KChIP3 rescues the functional expression of Shal channel tetramerization mutants. J Biol Chem 279, 54542–55451.
Kuo HC, Cheng CF, Clark RB, Lin JJ, Lin JL, Hoshijima M, Nguyen-Tran VT, Gu Y, Ikeda Y, Chu PH, Ross J, Giles WR & Chien KR (2001). A defect in the Kv channel-interacting protein 2 (KChIP2) gene leads to a complete loss of I(to) and confers susceptibility to ventricular tachycardia. Cell 107, 801–813.
, 百拇医药
Nadal MS, Ozaita A, Amarillo Y, de Miera EV, Ma Y, Mo W, Goldberg EM, Misumi Y, Ikehara Y, Neubert TA & Rudy B (2003). The CD26-related dipeptidyl aminopeptidase-like protein DPPX is a critical component of neuronal A-type K+ channels. Neuron 37, 449–461.
Nanao MH, Zhou W, Pfaffinger PJ & Choe S (2003). Determining the basis of channel-tetramerization specificity by X-ray crystallography and a sequence-comparison algorithm: family values (FamVal). Proc Natl Acad Sci U S A 100, 8670–8675.
, 百拇医药
Osawa M, Tong KI, Lilliehook C, Wasco W, Buxbaum JD, Cheng HY, Penninger JM, Ikura M & Ames JB (2001). Calcium-regulated DNA binding and oligomerization of the neuronal calcium-sensing protein, calsenilin/DREAM/KChIP3. J Biol Chem 276, 41005–41013.
Ren X, Shand SH & Takimoto K (2003). Effective association of Kv channel-interacting proteins with Kv4 channel is mediated with their unique core peptide. J Biol Chem 278, 43564–43570.
, 百拇医药
Scannevin RH, Wang K, Jow F, Megules J, Kopsco DC, Edris W, Carroll KC, Lu Q, Xu W, Xu Z, Katz AH, Olland S, Lin L, Taylor M, Stahl M, Malakian K, Somers W, Mosyak L, Bowlby MR, Chanda P & Rhodes KJ (2004). Two N-terminal domains of Kv4 K+ channels regulate binding to and modulation by KChIP1. Neuron 41, 587–598.
Schumacher MA, Rivard AF, Bachinger HP & Adelman JP (2001). Structure of the gating domain of a Ca2+-activated K+ channel complexed with Ca2+/calmodulin. Nature 410, 1120–1124.
, 百拇医药
Serodio P, Kentros C & Rudy B (1994). Identification of molecular components of A-type channels activating at subthreshold potentials. J Neurophysiol 72, 1516–1529.
Shibata R, Misonou H, Campomanes CR, Anderson AE, Schrader LA, Doliveira LC, Carroll KI, Sweatt JD, Rhodes KJ & Trimmer JS (2003). A fundamental role for KChIPs in determining the molecular properties and trafficking of Kv4.2 potassium channels. J Biol Chem 278, 36445–36454.
, 百拇医药
Takimoto K, Yang EK & Conforti L (2002). Palmitoylation of KChIP splicing variants is required for efficient cell surface expression of Kv4.3 channels. J Biol Chem 277, 26904–26911.
Zhou W, Qian Y, Kunjilwar K, Pfaffinger PJ & Choe S (2004). Structural insights into the functional interaction of KChIP1 with Shal-type K+ channels. Neuron 41, 573–586.
Zhu X-R, Wulf A, Schwarz M, Isbrandt D & Pongs O (1999). Characterization of human Kv4.2 mediating a rapidly-inactivating transient voltage-sensitive K+ current. Receptors Channels 6, 387–400., http://www.100md.com(Britta Callsen, Dirk Isbr)
Abstract
Association of Shal gene-related voltage-gated potassium (Kv4) channels with cytoplasmic Kv channel interacting proteins (KChIPs) influences inactivation gating and surface expression. We investigated both functional and biochemical consequences of mutations in cytoplasmic N and C-terminal Kv4.2 domains to characterize structural determinants for KChIP interaction. We performed a lysine-scanning mutagenesis within the proximal 40 amino acid portion and a structure-based mutagenesis in the tetramerization 1 (T1) domain of Kv4.2. In addition, the cytoplasmic Kv4.2 C-terminus was truncated at various positions. Wild-type and mutant Kv4.2 channels were coexpressed with KChIP2 isoforms in mammalian cell lines. The KChIP2-induced modulation of Kv4.2 currents was studied with whole-cell patch clamp and the binding of KChIP2 isoforms to Kv4.2 channels with coimmunoprecipitation experiments. Our results define one major interaction site for KChIPs, including amino acids in the proximal N-terminus between residues 11 and 23, where binding and functional modulation are essentially equivalent. A further interaction site includes residues in the T1 domain. Notably, C-terminal deletions also had marked effects on KChIP2-dependent gating modulation and KChIP2 binding, revealing a previously unknown involvement of domains within the cytoplasmic Kv4.2 C-terminus in KChIP interaction. Less coincidence of binding and functional modulation indicates a more loose ‘anchoring’ at T1- and C-terminal interaction sites. Our results refine and extend previously proposed structural models for Kv4.2/KChIP complex formation.
, 百拇医药
Introduction
Voltage-gated potassium channels of the Kv4-subfamily represent the molecular substrate of two transient currents of known physiological relevance: First, Kv4 channels mediate the neuronal somatodendritic A-type current (Serodio et al. 1994), which controls regular discharge behaviour as well as dendritic signal integration. Secondly, they mediate the transient outward current (Ito) in cardiomyocytes (Dixon et al. 1996), which participates in the early repolarization phase of the heart action potential.
, 百拇医药
Kv4 channels may coassemble with membrane-bound CD26-related dipeptidyl aminopeptidase-like proteins (Nadal et al. 2003), and with cytoplasmic Kv channel interacting proteins (KChIPs; An et al. 2000), which belong to the neuronal calcium sensor (NCS)-superfamily of EF-hand proteins (Burgoyne & Weiss, 2001). In both brain and heart the biophysical properties and surface expression levels of Kv4 channels may be controlled by these accessory subunits. This paper focuses on the interaction between Kv4 channels and KChIPs, which are integral components of brain Kv4 channel complexes (An et al. 2000; Nadal et al. 2003) and essential for the expression of Ito in rodent heart (Kuo et al. 2001).
, http://www.100md.com
In heterologous expression systems KChIPs have been shown to influence Kv4 channel assembly, phosphorylation status and stability (Shibata et al. 2003; Kunjilwar et al. 2004). KChIPs may enhance surface expression of Kv4 channels through a subcellular redistribution from perinuclear regions to the plasma membrane, which is reflected by an increase in the measured current density (An et al. 2000; Bhring et al. 2001). KChIPs may also modulate Kv4 channel inactivation kinetics (An et al. 2000; Bhring et al. 2001; Beck et al. 2002). This modulation is characterized by a slowing of the initial fast inactivation component of Kv4 channel-mediated currents and by an acceleration of recovery from inactivation.
, 百拇医药
KChIPs bind to the cytoplasmic N-terminus of Kv4 -subunits (An et al. 2000). Certain residues within the proximal 40 amino acid portion preceding the Kv4 T1 domain must be essential for KChIP-mediated gating modulation and KChIP binding, since Kv4.22–40 channels show neither increased current densities nor a modulation of inactivation kinetics in the presence of KChIP2.2 (Bhring et al. 2001). Furthermore, in vitro binding of KChIP2.2 is attenuated if the first 20, but not if the first 10 amino acids of Kv4.2 N-terminal fragments are deleted (Bhring et al. 2001). These results suggested that the proximal Kv4.2 N-terminus harbours a major KChIP interaction site, and that KChIP binding to this site is required for the effects on channel trafficking and gating.
, 百拇医药
It has been shown that effective association of Kv4.3 and KChIP2 does not require the variable and often alternatively spliced KChIP2 N-terminus, but critically depends on regions within the core domain of KChIP2 and its C-terminus (Ren et al. 2003). Crystal structure analyses have indicated that the KChIP core domain possesses a hydrophobic crevice (Scannevin et al. 2004; Zhou et al. 2004). This crevice may represent a binding pocket for hydrophobic residues in the proximal Kv4 N-terminus as indicated by the crystal structure of a KChIP1–Kv4.2 fusion protein, which contains the first 30 amino acids of Kv4.2 (Zhou et al. 2004).
, http://www.100md.com
In addition to the proximal Kv4 N-terminus, a cytoplasmic loop region on the N-terminal side of the T1 domain may play a role in KChIP interaction (Scannevin et al. 2004). However, correlating the effects of KChIP1 coexpression on Kv4.2 inactivation gating and the ability of KChIP1 to bind to Kv4.2 N-terminal fragments in a yeast two-hybrid assay showed some inconsistencies. Certain T1-domain mutations still allowed KChIP1-induced modulation of inactivation while interfering with KChIP1 binding (Scannevin et al. 2004). Such findings suggest that KChIP binding and coassembly with Kv4 -subunits may not be an absolute requirement for the observed KChIP-induced modulation of Kv4 channel gating in coexpression experiments. Alternatively, there may be additional, as yet undetected KChIP interaction sites on Kv4 channels.
, http://www.100md.com
Here we used naturally assembled and fully functional Kv4.2 channel constructs to study structural determinants for KChIP interaction. We focused on three regions of the Kv4.2 -subunit: The proximal Kv4.2 N-terminus, the Kv4.2 T1 domain and the cytoplasmic Kv4.2 C-terminus. The consequences of mutations or deletions within these regions were studied with both electrophysiological and immunocytochemical methods. KChIP-mediated Kv4.2 gating modulation and KChIP-induced changes in functional Kv4.2 surface expression were assessed with whole-cell patch-clamp recordings and direct Kv4.2/KChIP binding with immunoprecipitation (IP) experiments. The data reveal that in all three regions examined critical sites exist, that are important for functional Kv4.2/KChIP interaction.
, http://www.100md.com
Methods
Clones and channel constructs
We tested human KChIP1 (AF199597), KChIP2.1 (AF199598), KChIP2.2 (Bhring et al. 2001) and KChIP2.8 (NM_173194) coexpressed with different constructs of human Kv4.2 (Zhu et al. 1999). We performed a scanning-mutagenesis of the proximal Kv4.2 N-terminus (individual replacement of residues 8–30 by lysine or glutamate; see Figs 2 and 3 and text). Site-directed mutagenesis in the Kv4.2 T1 domain was based on its three-dimensional structure (Nanao et al. 2003; see Fig. 5E examined with the program InsightII (Accelrys). We introduced point mutations at exposed T1-domain sites (mutation of residues 64, 66, 69, 70, 71, 74, 76, 79 and 81 to lysine, glutamate or alanine; see Fig. 5 and text). Finally, we generated the C-terminal deletion mutants Kv 4.2 420, 435, 446, 461, 478, 490 and 580 (numbers indicate the last amino acid residue of the respective construct). All Kv4.2 constructs carried an extracellular haemagglutinin (HA)-tag (not further indicated) between transmembrane segments S1 and S2, which had no influence on channel function or KChIP-mediated gating modulation (not shown). All constructs were subcloned into the pcDNA3 mammalian expression vector (Invitrogen) and verified by sequencing. In the text, individual amino acids with residue numbers are given in three-letter code and individual point mutations in single-letter code. Since all mutations in this study were introduced in Kv4.2 this will not be indicated further.
, http://www.100md.com
Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (n-fold slowing of inactivation; A), the kinetics of recovery from inactivation (n-fold acceleration of recovery from inactivation; B) and the peak current density at +40 mV (n-fold increase in current density; C). Data are shown for Kv4.2 wild-type (wt) and for the individual scanning mutants of the proximal N-terminus. Open bars indicate significant differences, filled bars no significant differences, after coexpression of KChIP2.2 (see also Methods and Table 1). Horizontal broken lines indicate KChIP2.2 effect on wild-type and no effect (n-fold = 1), respectively. Error bars are the standard error.
, http://www.100md.com
A, results obtained with immunoprecipitation (IP) after heterologous coexpression of KChIP2.1 and individual scanning mutants of the proximal Kv4.2 N-terminus in CHO cells. Prior to immunoprecipitation (pre-input) western-blot analysis was performed to quantify the amount of protein expressed. After adjustment of input amount and volume (not shown) KChIP2.1 and HA-tagged Kv4.2 proteins were coimmunoprecipitated and detected with panKChIP (anti-panKChIP) and HA antibodies (anti-HA). Note the impairment of binding for mutations within a stretch of amino acids including Phe11 and Ala23, which shows a discontinuity at Arg13 and Ala16. In the control lane an unspecific band was occasionally detected by the panKChIP antibody. B, structural modelling of the N-terminal residues, mutated and examined in the present study, onto the crystal structure of a KChIP1–4.2N30 fusion protein (Zhou et al. 2004). KChIP1 portion in CPK (EF-hands in bronze colour) and Kv4.2 portion in ribbon presentation. Residues in the flanking regions not involved in KChIP binding as well as Arg13 and Ala16, which reside within the mutation-sensitive domain but exhibit strong coimmunoprecipitation signals, are labelled green. The remaining residues, which are critical for KChIP binding, are labelled red.
, 百拇医药
Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (A), the kinetics of recovery from inactivation (B) and the peak current density measured at +40 mV (C). Data are shown for all tested Kv4.2 T1-domain mutants. D, results obtained with immunoprecipitation (IP) after heterologous coexpression of KChIP2.1 and Kv4.2 T1-domain mutants. Note the apparent lack of binding for L66D, S69D and F74K. E, structural modelling of residues mutated and examined in this study (Asp64, Leu66, Ser69, Ser70, Glu71, Phe74, Glu79 and Gln81) on the Kv4.2 T1-domain crystal structure (Nanao et al. 2003); Zn2+ coordination residues in bronze colour). Only two opposite T1 monomers are shown from a side view. Green residues: mutation does not interfere with KChIP2.1 binding; red residues: mutation leads to a loss of binding as inferred from a lack of coimmunoprecipitation; blue residue: mutation produced non-functional channels. Error bars are the standard error.
, 百拇医药
Heterologous expression
Kv4.2 channels, either alone or in combination with KChIP, were expressed in human embryonic kidney (HEK) 293 and Chinese hamster ovary (CHO) cells as previously described (Bhring et al. 2001). Cells were plated at a density of 3–4 x 104 per 35 mm dish and transiently transfected with cDNA using Lipofectamine (Invitrogen). For the kinetic analyses of inactivation gating, done in HEK 293 cells, we used either 1 μg Kv4.2 cDNA alone or 0.1 μg Kv4.2 in combination with 1 μg KChIP2.2 cDNA per 35 mm dish. For the current density measurements, done in CHO cells, we used 0.1 μg Kv4.2 both in the absence and presence of 1 μg KChIP2.2 cDNA. In all cases 0.5 μg EGFP cDNA (Clontech) was cotransfected, or 1.5 μg of a bicistronic cDNA vector encoding both KChIP2.2 and EGFP was used to identify successfully transfected cells. No functional expression after transient transfection, both in the absence and in the presence of KChIP2.2, was obtained with G18K, S24K, Y76A, Y76V, 420, 446 and 461. These mutants were not used for the analysis of KChIP binding. For immunoprecipitation (IP) experiments CHO cells were plated at a density of 4–5 x 105 per 50 mm dish and transiently transfected with 2 μg Kv4.2 and 2 μg KChIP2 cDNA per dish using FuGENE6 (Roche).
, 百拇医药
Electrophysiological analyses were done with KChIP2.2, because in our hands this variant yields the strongest effects on Kv4.2 functional parameters (B.C. and R.B., unpublished observation). IP experiments were performed mostly with KChIP2.1, because we noticed that KChIP2.2 co-migrated with an unspecific band occasionally detected in control lanes (see Fig. 3A). KChIP2.1 and KChIP2.2 only differ in their alternatively spliced N-termini (Bhring et al. 2001), a region of minor importance for the binding to Kv4 -subunits (Ren et al. 2003).
, 百拇医药
Electrophysiological recordings and data analysis
Cells were continuously superfused at room temperature (20–22°C) with a solution containing (mM): NaCl 135, KCl 5, CaCl2 2, MgCl2 2, Hepes 5, sucrose 10 and phenol red 0.01 mg ml–1; pH 7.4 (NaOH). Recordings were performed in the whole-cell configuration of the patch-clamp technique. Patch-pipettes were pulled from thin-walled borosilicate glass using a DMZ-puller (Zeitz). Pipette-to-bath resistances ranged between 2 and 3 M with internal solution containing (mM): KCl 125, CaCl2 1, MgCl2 1, EGTA 11, Hepes 10, K2ATP 2 and glutathione 2; pH 7.2 (KOH). Currents were recorded using an EPC9 patch-clamp amplifier controlled by PULSE software (HEKA) with maximal series resistance compensation (usually between 80 and 90%). Signals were filtered at 0.2–4 kHz with low-pass Bessel characteristics, amplified as required, and digitized at 50–1000 μs sample intervals. Capacitive current transients were compensated on-line, and the P/n method was applied for leak subtraction. The program PULSEFIT (HEKA) was used to analyse current traces, and the obtained data were further processed using Kaleidagraph (Synergy Software). We analysed the following functional parameters: the extent of fast inactivation was quantified by the current amplitude remaining 30 ms after the activating voltage-step relative to the peak current amplitude (I30/Ipeak; see Figs 1, 4 and 6). The kinetics of recovery from inactivation were measured at –80 mV and fitted with a single-exponential function. As a further electrophysiological parameter we calculated, from the peak current amplitude measured at +40 mV and the whole-cell capacitance, the current density (pA pF–1). Testing more than only one functional parameter helped to eliminate false negatives for the judgement of functional KChIP2.2 interaction, since in many cases the experimentally introduced channel mutation caused extremely low current expression in the absence of KChIP2.2, leading to analysis problems and small sample sizes. Pooled data are presented as mean ± S.E.M., and statistical analyses were performed using GraphPad Prism (GraphPad Software; see text and Table 1). Differences between the mean values obtained in the absence (meana ± S.E.M.a) and presence of KChIP2.2 (meanb ± S.E.M.b) were examined for significance with unpaired Student's t tests (one-tailed P-values and Welch-correction for unequal variances; Table 1). KChIP effects on Kv4.2 current parameters were quantified as the average n-fold change (n = meanb/meana for I30/Ipeak and current density; n = meana/meanb for recovery from inactivation) induced by KChIP2.2 coexpression. Assuming linear and independent error propagation the standard error S.E.M.n of this quotient was calculated to be (see Figs 2, 5 and 7).
, http://www.100md.com
Left panels show currents mediated by Kv4.2 wild-type (wt; A1), W8K (B1) and F11K (C1) channels expressed in HEK 293 cells in the absence or presence of KChIP2.2. Currents were measured in the whole-cell configuration during depolarizing test-pulses from –100 to +40 mV. Inactivation of A-type currents was analysed and plotted (middle panels) as the relative current remaining 30 ms after the beginning of the test pulse (I30/Ipeak, vertical broken lines), in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels show the effects of KChIP2.2 coexpression on the kinetics of recovery from inactivation for wt (A2), W8K (B2) and F11K (C2). Data were plotted on a semilog scale and fitted with single-exponential functions. , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, http://www.100md.com
Left panels, currents mediated by 435 (A1), 478 (B1), 490 (C1) and 580 (D1) in the absence and presence of KChIP2.2. Middle panels, relative amount of inactivation after 30 ms in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels, effect of KChIP2.2 coexpression on the kinetics of recovery from inactivation for 435 (A1), 478 (B1), 490 (C1) and 580 (D1). , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, 百拇医药 Effects of KChIP2.2 coexpression on the initial Kv4.2 current inactivation (A), the kinetics of recovery from inactivation (B) and the peak current density at +40 mV (C). D, results obtained with immunoprecipitation (IP) experiments after heterologous coexpression of KChIP2.1 and Kv4.2 mutants with different C-terminal truncations. Note the strong impairment of KChIP2.1 binding for constructs shorter than 580, with an apparent regain of KChIP binding ability for 435. E, IP was repeated with 580, 490 and 478 after coexpression with KChIP1, KChIP2.2 or KChIP2.8. Note that all tested KChIP variants yielded the same results. Error bars are the standard error.
, 百拇医药
Binding experiments
From CHO cells transfected with Kv4.2 and KChIP cDNA total cell lysates were prepared (lysis buffer: 150 mM NaCl, 50 mM Hepes pH 7.4, 0.5% Triton) in the presence of a protease inhibitor cocktail (Sigma). After pelleting the insoluble matter (10 000 g; 10 min) equal amounts of the supernatant were separated by 4–12% gradient SDS-PAGE and transferred to nitrocellulose membranes. After a blocking reaction (PBS; 0.05% Tween; 5% non-fat milk powder) affinity-purified HA- (1 : 500, 3F10; Roche) or panKChIP-antibody (1 : 600) was incubated overnight. After washing the membrane, horseradish-peroxidase-conjugated goat antibody (1 : 5000; Jackson ImmunoResearch) was incubated for 1 h. Enhanced chemiluminescence reagents were used for signal detection. To quantify the amount of protein expressed, signal intensities were evaluated with the program Quantify One (Bio-Rad).
, http://www.100md.com
Equal amounts of protein were immunoprecipitated by addition of HA- or panKChIP-antibody at a 10 μg ml–1 final concentration overnight. Mixtures were incubated with ProteinG Dynabeads (Dynal Biotech) for 2 h, and immunoprecipitates were washed twice with lysis buffer and eluted with Nupage sample buffer (Invitrogen) under reducing conditions (10 min at 70°C). With the supernatant, western blot analysis was done as described.
Results
, 百拇医药
Mutant scanning analysis of the proximal Kv4.2 N-terminus
In a previous study (Bhring et al. 2001) we have shown that efficient in vitro binding of KChIP2.2 to N-terminal Kv4.2 fragments tolerates a deletion of the first 10 (2–10) but not a deletion of the first 20 amino acids (2–20). Also, in contrast to Kv4.2 wild-type and 2–10, 2–20-mediated current densities were no longer increased in the presence of KChIP2.2 (Bhring et al. 2001). The data suggested that the N-terminal Kv4.2 sequence between amino acid residues 10 and 20 was critical for KChIP2.2 interaction. In order to characterize the N-terminal Kv4.2 interaction domain more precisely we now performed a lysine-scanning mutagenesis of the proximal Kv4.2 N-terminus between residues Trp8 and Pro30 with the positively charged Arginine at position 13 mutated to glutamate (R13E).
, http://www.100md.com
Mutant Kv4.2 channels were first tested for their functionality and then coexpressed with KChIP2.2. G18K and S24K yielded no functional expression in the absence and presence of KChIP2.2. However, the respective glutamate mutants G18E and S24E expressed currents and were used in this scanning analysis. Typically, KChIP2.2 slows the initial onset of inactivation, accelerates recovery from inactivation and increases Kv4.2 current densities. We first analysed these parameters in the absence and presence of KChIP2.2 to assess potential effects of the N-terminal point mutations on KChIP interaction. We analysed fast Kv4.2 inactivation by measuring the ratio I30/Ipeak (see Methods; Fig. 1). For Kv4.2 wild-type channels expressed in the absence of KChIP2.2 I30/Ipeak was 0.27 ± 0.02 (n = 3). In agreement with previous results (Bhring et al. 2001), coexpression of Kv4.2 wild-type with KChIP2.2 increased I30/Ipeak to 0.58 ± 0.04 (n = 3; P = 0.0101) due to a slowing of the initial current inactivation phase (Fig. 1A1), and accelerated recovery from inactivation (rec = 290 ± 43 ms, n = 3 in the absence and 39 ± 3 ms, n = 4 in the presence of KChIP2.2; P = 0.0141; Fig. 1A2). Table 1 summarizes current parameters determined for Kv4.2 channel mutants in the absence and presence of KChIP2.2. Effects of KChIP2.2 coexpression on the Kv4.2 current parameters expressed as n-fold change (see Methods) are shown in Fig. 2.
, 百拇医药
More than 50% of the proximal mutations (L9K, P10K, R13E, A16K, A23K, S24E, G25K, P26K, P28K, M29K and P30K) generated Kv4.2 channels with current parameters affected by KChIP2.2 like wild-type (Fig. 2). Thus, many N-terminal point mutations did not alter the interaction of Kv4.2 channels with KChIP2.2, indicative of local rather than long distance mutational effects. Also W8K appeared to interact with KChIP2.2. Notably, W8K inactivation kinetics were slower than wild-type in the absence of KChIP2.2 (I30/Ipeak = 0.81 ± 0.03, n = 3), however, in the presence of KChIP2.2 both onset of (I30/Ipeak = 0.60 ± 0.04, n = 7; Fig. 1B1) and recovery from inactivation (rec = 50 ± 6 ms, n = 5; Fig. 1B2) were comparable to wild-type coexpressed with KChIP2.2 (Table 1).
, 百拇医药
Only four point mutations (F11K, G18E, P21K and V22K) produced Kv4.2 channels with current parameters that remained unchanged upon KChIP2.2 coexpression (Fig. 2), indicating that these mutations impaired the KChIP2.2-induced modulation of Kv4.2 currents. Coexpression of F11K with KChIP2.2 changed neither onset of (I30/Ipeak 0.70 ± 0.06, n = 6 in the absence and 0.67 ± 0.06, n = 6 in the presence of KChIP2.2; P = 0.2034; Fig. 1C1) nor, significantly, recovery from inactivation (rec = 401 ± 69 ms, n = 5 in the absence and 285 ± 41, n = 4 in the presence of KChIP2.2; P = 0.0993; Fig. 1C2).
, http://www.100md.com
The remaining point mutations (A12K, A14K, A15K, I17K, W19K, M20K and M27K) showed intermediate KChIP sensitivity; i.e. coexpression with KChIP2.2 significantly influenced only one or two of the measured current parameters (Table 1). Although significant, these changes were always smaller than those obtained for wild-type (Fig. 2). Thus, these mutations also affected the KChIP2.2-induced Kv4.2 current modulation.
A likely explanation for our mutational results is that the Kv4.2 N-terminus contains a KChIP2.2-binding domain between residues 11 and 22, and that single point mutations in this binding domain interfere with KChIP binding. We tested this hypothesis by assaying direct KChIP binding to each of the tested Kv4.2 constructs in IP experiments (Fig. 3A). A15K and P28K were not investigated due to their extremely low expression levels. Signal intensities for coimmunoprecipitated Kv4.2 and KChIP2.1 polypeptides were similar to wild-type controls if the point mutation had been introduced N-terminal to Phe11 (W8K, L9K, P10K) or C-terminal to Val23 (A24K and further downstream); i.e. outside the proposed KChIP-binding domain in the Kv4.2 N-terminus (Fig. 3A). By contrast, signal intensities for coimmunoprecipitated Kv4.2 and KChIP2.1 polypeptides were significantly weaker if the point mutations were located within the proposed KChIP-binding domain, except for R13E and A16K (Fig. 3A). These data are in agreement with X-ray structural analysis of a KChIP1–Kv4.2N30 fusion protein (Zhou et al. 2004), in which the side chains of Arg13 and Ala16 do not make contact with the KChIP surface (Fig. 3B; see Discussion).
, 百拇医药
Taken together, our IP and electrophysiology results show a good correlation regarding binding strength and KChIP-induced changes in current properties. This implies that KChIP2.2 modulates Kv4.2 inactivation kinetics and current density through a direct interaction with a binding domain between residues 11 and 23 of the Kv4.2 N-terminus. Thus, our data extend previous KChIP1–Kv4.2N30 X-ray analysis (Zhou et al. 2004), which did not yield atomic coordinates for residues 21–23, leaving the C-terminal border of the proximal KChIP-binding domain undefined.
, http://www.100md.com
Role of exposed amino-acids in the Kv4.2 T1 domain
Ca2+-binding proteins like calmodulin may accommodate in their binding pocket more than one -helix (Schumacher et al. 2001). We hypothesized that KChIPs may interact in a similar fashion with more than one Kv4.2 domain.
In the light of a report that certain regions of the Kv4.2 T1 domain may also be critical for KChIP binding (Scannevin et al. 2004), we examined if the Kv4.2 T1 domain possesses KChIP interaction sites. We used crystal structure information (Nanao et al. 2003) to mutate residues exposed on the cytoplasmic surface of the Kv4.2 T1 domain (Fig. 5E).
, http://www.100md.com
Within the cytoplasmic loop-region of the Kv4.2 T1 domain we mutated Tyr76, Glu79 and Gln81 (Fig. 5E). Neither Y76A nor Y76V yielded functional expression in the absence or presence of KChIP2.2. E79A current parameters including onset of (I30/Ipeak = 0.32 ± 0.03, n = 4 in the absence and 0.47 ± 0.02, n = 6 in the presence of KChIP2.2; p = 0.0044; Fig. 4A1) and recovery from inactivation (rec = 247 ± 14 ms, n = 4 in the absence and 42 ± 2, n = 5 in the presence of KChIP2.2; P = 0.0004; Fig. 4A2) were modulated by coexpression of KChIP2.2 comparable to wild-type parameters, and similar results were obtained for Q81A (Fig. 5A and B, Table 1).
, http://www.100md.com
Left panels, currents mediated by E79A (A1), L66D (B1), S69D (C1) and F74K (D1), in the absence and presence of KChIP2.2. Middle panels, relative amount of inactivation after 30 ms in the absence (filled bars) and presence of KChIP2.2 (open bars). Right panels, effect of KChIP2.2 coexpression on the kinetics of recovery from inactivation for E79A (A2), L66D (B2), S69D (C2) and F74K (D2). , without KChIP2.2; , in the presence of KChIP2.2. Error bars are the standard error.
, 百拇医药
Next, we introduced individual lysine residues, in analogy to our terminal scanning analysis, for residues Asp64, Leu66, Ser70, Glu71 and Phe74 (D64K, L66K, S70K, E71K, F74K), which are more laterally located with respect to the cytoplasmic loop region of the Kv4.2 T1 domain. In this region we also mutated some residues to aspartate (L66D, S69D, S70D) in order to detect possible charge-dependent effects on KChIP interaction. Most of these Kv4.2 mutants showed current properties like wild-type in the absence or presence of KChIP2.2 (Table 1). Normal KChIP-induced current modulation is shown for L66D and S69D in Fig. 4B and C, respectively. In the case of F74K and L66K, coexpression of KChIP2.2 slowed the onset of inactivation comparable to wild-type (F74K: I30/Ipeak = 0.36 ± 0.04, n = 3 in the absence and 0.50 ± 0.04, n = 3 in the presence of KChIP2.2; P = 0.0343; L66K: I30/Ipeak = 0.20 ± 0.02, n = 3 in the absence and 0.55 ± 0.05, n = 10 in the presence of KChIP2.2; P < 0.0001; Figs 4D1 and 5A), but only caused a weak or insignificant acceleration of recovery from inactivation (F74K: rec = 254 ± 15 ms, n = 5 and 166 ± 25 ms, n = 6; P = 0.0097; L66K: rec = 232 ± 39 ms, n = 3 and 155 ± 23 ms, n = 9 in the absence and presence of KChIP2.2, respectively; P = 0.0938; Figs 4D2 and 5B).
, 百拇医药
Figure 5C illustrates the relative increases in current density observed for the T1-domain mutants when coexpressed with KChIP2.2. Current density increases were smallest in the case of L66K and S70D (5- and 15-fold, respectively). In contrast to the proximal mutations, the Kv4.2 T1-domain mutations did not completely attenuate KChIP2.2-induced current modulation, but produced more intermediate effects.
Next, we investigated whether the attenuation of KChIP2.2 effects on Kv4.2 T1-domain mutants correlated with an impairment of KChIP binding. Coimmunoprecipitation signal intensities were comparable to wild-type for most T1-domain mutants except L66D, S69D, and F74K. This result was unexpected, since current properties of the latter three mutants were modulated by KChIP2.2 (Fig. 5A–C). Thus, our IP data suggest that T1-domain residues Leu66, Ser69 and Phe74 may be critical for KChIP binding, however, the results of our electrophysiological and binding experiments did not strictly correlate (see Discussion).
, http://www.100md.com
Mutational analysis of the cytoplasmic Kv4.2 C-terminus
Previously, it was shown by mutational deletion that Kv4.1 channel inactivation is mediated by the concerted action of cytoplasmic N- and C-terminal domains (Jerng & Covarrubias, 1997). Since Kv4.2 N-terminal deletion mutants show no functional modulation by KChIP we now examined if C-terminal deletions of varying lengths may similarly influence functional KChIP interaction. The initial onset of inactivation of 435-mediated currents was on average slightly decelerated after coexpression with KChIP2.2 (I30/Ipeak = 0.25 ± 0.03, n = 3 in the absence and 0.38 ± 0.03, n = 9 in the presence of KChIP2.2; P = 0.0111; Figs 6A1 and 7A). However, recovery of 435 from inactivation was not accelerated (rec = 364 ± 29 ms, n = 5 in the absence and 343 ± 18 ms, n = 8 in the presence of KChIP2.2; P = 0.3099; Figs 6A2 and 7B) and 435 current densities were not increased by KChIP2.2 (Fig. 7C). Investigating less severe C-terminal deletions showed that the effects of KChIP2.2 coexpression on Kv4.2 current properties became gradually larger (Fig. 7B and C). In the case of 478, KChIP2.2 seemed to slightly enhance both onset of (I30/Ipeak = 0.32 ± 0.05, n = 7 in the absence and 0.21 ± 0.03, n = 12 in the presence of KChIP2.2; P = 0.0443; Figs 6B1 and 7A) and recovery from inactivation (rec = 256 ± 17, n = 10 in the absence and 165 ± 9 ms, n = 9 in the presence of KChIP2.2; P = 0.0002; Figs 6B2 and 7B). However, also for this mutant no significant increase in current density was observed (Fig. 7C). 490 channels tended to show stronger KChIP2.2 effects (Figs 6C and 7A–C), but only 580 channels showed pronounced KChIP2.2 effects on inactivation kinetics and current density, similar to the ones observed for wild-type (Figs 6D and 7A–C).
, 百拇医药
Next, we examined KChIP2.1 binding to the Kv4.2 C-terminal deletion mutants in IP experiments (Fig. 7D). The results showed that only 580 channels were efficiently coimmunoprecipitated with KChIP2.1 showing signal intensities comparable to wild-type. Neither with HA- nor with panKChIP-antibodies did we obtain signals with the deletion mutants 478 and 490, but the shortest deletion mutant 435 yielded a weak signal when immunoprecipitated with panKChIP-antibodies (Fig. 7D). For Kv4.2 wild-type, 580, 490, and 478 we performed further IP experiments together with KChIP1, KChIP2.2 and KChIP2.8, respectively (Fig. 7E). The outcome of these experiments was similar for all KChIP variants used; i.e. prominent coimmunoprecipitation signals comparableto wild-type were only observed with 580. Thus, effects of C-terminal deletions on binding of KChIPs to Kv4.2, and modulation of Kv4.2 current properties by KChIP2.2 did not strictly correlate, very similar to the observations we made with certain T1-domain mutants.
, 百拇医药
Discussion
In this study we carried out a mutational analysis of the cytoplasmic Kv4.2 N- and C-termini to investigate structural determinants for channel modulation by accessory KChIP2 subunits. For each Kv4.2 mutant the KChIP2-induced modulation of current properties and KChIP2 binding was tested. Thus, only properly assembled and fully functional constructs were analysed. The results suggest that some mutations affect a major Kv4.2/KChIP2-binding interface and therefore directly interfere with KChIP2-induced modulation of Kv4.2 current properties. Other mutations may have a more indirect effect on functional interaction. Previously, the interaction of Shaker-related Kv1 channels with their accessory Kv-subunits has been studied in great detail (Gulbis et al. 2000). Our results demonstrate that the Kv4.2 channel domains involved in KChIP-dependent gating modulation are disitinct from those in Shaker-related Kv1 channels for Kvinteraction. The differences in the mode of interaction for the two types of accessory subunits reflect the distinct inactivation mechanisms and functional roles demonstrated for Kv1/Kvand Kv4/KChIPs channel complexes, respectively.
, 百拇医药
The cytoplasmic Kv4.2 N-terminus harbours multiple KChIP interaction sites
The spacing of N-terminal Kv4.2 amino acids, mutation of which affected KChIP binding and current modulation, suggests that the KChIP-binding interface is -helical. According to our data, residues Phe11, Ala14 and Trp19 of the N-terminal Kv4.2 -helix mainly contribute to this binding interface. This view is consistent with the crystal structure of a KChIP1–Kv4.2N30 fusion protein (Zhou et al. 2004). The structure shows the N-terminal Kv4.2 domain as an -helix embedded in the hydrophobic crevice of KChIP1, and side chains of Kv4.2 amino acid residues Trp8 and Phe11 appear to make direct contact with KChIP1 (Zhou et al. 2004). Possibly the contact between Trp8 and KChIP1 seen in the crystal structure is not important for KChIP binding, because our data show that W8K channels bound KChIP2.1 in a similar way to wild-type. Our data, which were obtained with a drastic amino acid substitution, clearly demarcate Phe11 as the N-terminal border of the KChIP-binding interface in Kv4.2, consistent with our previous finding that deletion of the first 10 amino acids did not affect KChIP interaction with Kv4.2 channels (Bhring et al. 2001). The fact that hydrophobicity was not changed dramatically by a mutation of Phe11 to alanine may explain why it was tolerated in a previous study (Scannevin et al. 2004). In addition, our results reveal Kv4.2 residues relevant for KChIP2 interaction near the C-terminal border of the interface (Pro21, Val22, Ala23 and possibly Met27). In this region the crystal structure of the KChIP1–Kv4.2N30 fusion protein lacks atomic resolution (Zhou et al. 2004).
, http://www.100md.com
It is intriguing that neither R13E, which we have analysed in a previous study (Gebauer et al. 2004), nor proximal lysine-scanning in Kv4.2 had a pronounced effect on inactivation per se. The results corroborate the view that, unlike classical models of Shaker/Kv1 N-type inactivation, electrostatic interactions mediated by charged amino acid residues play a minor role in Kv4.2 N-type inactivation (Gebauer et al. 2004). In this context, our results obtained with W8K deserve further consideration: Obviously, in this mutant the introduction of a positively charged amino acid for a hydrophobic one did have a strong influence on inactivation. In the absence of KChIP2.2, W8K channels inactivated markedly more slowly than wild-type. In the presence of KChIP2.2, the initial slow decay of W8K currents was accelerated to a value similar to the one observed for Kv4.2 wild-type coexpressed with KChIP2.2 (see Fig. 1), which may reflect the previously described ‘streamlining’ of Kv4-mediated A-type currents by KChIPs (Beck et al. 2002). Recovery of W8K channels from inactivation showed the characteristic KChIP-induced acceleration. In experiments where we coexpressed the slowly inactivating N-terminal deletion mutant 2–10 with KChIP2.2 we observed no change in the onset, but normal acceleration of recovery from inactivation (data not shown). Similar findings were obtained previously for the functional interaction between KChIP1 and N-terminally GFP-tagged Kv4.3 channels (Hatano et al. 2002), where the GFP-tag only compromises fast inactivation and its modulation by KChIP but not the KChIP-induced acceleration of recovery from inactivation. The data are difficult to reconcile with a gating model, in which KChIP binds to and immobilizes the proximal Kv4 N-terminus to modulate all aspects of inactivation (Beck et al. 2002).
, 百拇医药
The interaction between Kv4 channels and KChIPs seems to be more complex, as indicated by the observation that additional domains of the cytoplasmic Kv4.2 N-terminus may contribute to an effective KChIP-induced channel modulation (Scannevin et al. 2004). A Kv4.2 N-terminal fragment (amino acids 1–180) with an internal deletion (71–90) showed no interaction with KChIP1 in a yeast two-hybrid assay (Scannevin et al. 2004). The deletion eliminated residues normally located in a cytoplasmic loop region on the N-terminal side of the Kv4.2 T1 domain (Nanao et al. 2003). In Kv1.4 channels this T1-loop region is important for the functional interaction with accessory Kv-subunits (Gulbis et al. 2000). Intersubfamily protein sequence alignments suggest that the critical phenylalanine in Kv1.4, although replaced by a tyrosine in Kv4.1 and Kv4.2, is conserved in Kv4.3. Given the high similarity of the KChIP-induced effects observed with the different wild-type Kv4 subfamily members (Bhring et al. 2001), it seems unlikely that the same T1-loop residues as in Kv1/Kvchannel complexes play a central role in KChIP-mediated Kv4 channel modulation. Nevertheless, in a Kv1.2 N-terminal fragment substitution of the T1-loop sequence by the respective Kv4.2 sequence (amino acids 71–90) conferred the ability to bind KChIP1 in the yeast two-hybrid assay (Scannevin et al. 2004). The conclusion drawn from these results was that the loop region in the Kv4.2 T1 domain contains key residues for KChIP binding (Scannevin et al. 2004). We mutated residues directly within as well as proximal to the T1 loop to investigate potential effects on KChIP-induced Kv4.2 current modulation or KChIP binding. With the exception of Y76A and Y76V, all T1-domain mutants expressed functional channels, excluding long-range mutational effects that may have caused folding and assembly problems. Our results clearly exclude Glu71 and Glu79 as important KChIP interaction sites. For other T1-domain sites, however, the data were not as clear-cut as for our proximal mutagenesis scan (see Figs 2, 3 and 5). T1-domain mutations affected KChIP2.2-induced Kv4.2 current modulation to variable extents, but the effects did not strictly correlate with the IP data. We only obtained a consistent picture for the mutant F74K: The loss of KChIP2-binding ability combined with only a very modest acceleration of recovery from inactivation observed for this mutant confirmed previous observations (Scannevin et al. 2004). Taken together, our data support the view that amino acids in the Kv4.2 T1 domain including the cytoplasmic loop region may be involved in KChIP interaction. The interaction seems to be not strong enough for coimmunoprecipitation and may be indirect.
, 百拇医药
Involvement of C-terminal cytoplasmic Kv4.2 domains in KChIP interaction
It has been shown previously that Kv4 channel inactivation is mediated by the concerted action of cytoplasmic N- and C-terminal domains (Jerng & Covarrubias, 1997). This may indicate that N- and C-terminal Kv4.2 domains interact with each other, as has been reported for the related Kv2.1 channels, where residues close to the cytoplasmic loop of the T1 domain interact with a cytoplasmic C-terminal domain (Ju et al. 2003). These previous findings combined with our experimental results obtained with Kv4.2 T1-domain mutants led us to hypothesize that the cytoplasmic Kv4.2 C-terminus may also play a role in mediating KChIP2 effects on Kv4.2 inactivation gating. Indeed, our experiments showed that deleting the Kv4.2 C-terminus proximal to amino acid residue 580 (490, 478 and 435) may attenuate the KChIP2.2 sensitivity of Kv4.2 currents. The results suggest that domain(s) in the Kv4.2 C-terminus are critical for KChIP2-induced Kv4.2 modulation. Consistent with the electrophysiological data, the IP experiments showed that 490 and 478 mutants did not form a stable complex with KChIPs. Notably, 435 and KChIP2.1 polypeptides were coimmunoprecipitated to some degree, although we did not observe a marked effect of KChIP2.2 on 435 current.
, 百拇医药
In the absence of additional data, it is certainly difficult to reconcile apparent inconsistencies between electrophysiological and IP results. One intriguing explanation would be that a palmitoylation-mediated anchoring of KChIPs in the plasma membrane (Takimoto et al. 2002) could lead to functional interaction with Kv4 channels without tightly binding to them. However, the human KChIP2.2 isoform used in this and our previous study (Bhring et al. 2001) lacks the N-terminal double cysteine motif necessary for palmitoylation. Alternatively, due to strong overexpression, the cytoplasmic content of KChIP2.2 may be high enough to promote interaction with Kv4.2 channels based on the law of mass action. In fact, such mass action effects have been reported previously for the promoting effect of KChIP3 on the assembly of Kv4.2 Zn2+-binding mutants (Kunjilwar et al. 2004). We favour the following explanation for our experimental results: mutations, where a loss of binding correlated with a complete loss of KChIP2.2 effects on Kv4.2 current properties, most likely affected directly the Kv4.2/KChIP2.2 interface. This was observed for mutations in the proximal Kv4.2 N-terminus. We propose therefore that the corresponding amino acids are key residues for the interaction of Kv4.2 with KChIP2. Mutations where a complete or partial loss of KChIP2.2 effects on Kv4.2 current properties did not coincide with a loss of KChIP2 binding, are likely to interfere more indirectly with KChIP-induced Kv4.2 current modulation. Finally, some mutations altered, to varying degrees, some aspects of KChIP2-induced Kv4.2 current modulation but at the same time showed a loss of KChIP binding. Possibly, the corresponding Kv4.2/KChIP2 complexes were not stable enough for coimmunoprecipitation. This explanation is consistent with our observation that in many of these cases one of the two antibodies we used still produced a coimmunoprecipitation signal. Thus, we propose that the respective deletions of the cytoplasmic C-terminus, similar to certain point mutations in the T1 domain, exert indirect effects on Kv4.2/KChIP2.2 interaction. T1 and C-terminal sites may represent anchoring domains, helping KChIPs to achieve an optimal orientation for docking to the major binding site between residues 11 and 23 of the Kv4.2 N-terminus. Mutation-induced poor KChIP anchoring may cause defective binding. This disruption may then lead to various degrees of KChIP-mediated channel modulation. Furthermore, different mechanisms of modulation affecting different current parameters may become decoupled. Although tightly bound to the N-terminus, KChIPs may not be anchored and therefore the complex is poorly constrained. As a result, fast inactivation would be slowed but recovery may not be affected (e.g. 435 in Fig. 7). Finally, disrupted KChIP anchoring may also cause incorrect or weaker KChIP binding to the N-terminus (Figs 5D and 7D and E).
, 百拇医药
Implications for the macromolecular structure of Kv4.2 channel complexes
Although there is strong evidence that Kv4.2/KChIP2 channels are octomeric complexes with a 4 : 4 stoichiometry (Kim et al. 2004a), the exact binding mode has remained unsolved. In principle, Kv4/KChIP assembly could be based on different mechanisms leading to the same stoichiometry: Based on the ability of KChIPs to tetramerize under certain conditions (Osawa et al. 2001), they could bind to Kv4 channels as a tetramer, similar to Kv-subunits on the bottom of the T1-tetramer (Gulbis et al. 2000); the binding data obtained by Scannevin et al. (2004) support this view. Alternatively, two KChIP molecules, each bound to a different Kv4 N-terminus, could dimerize, which eventually would result in a dimer-of-dimers organization, as has been shown previously for the binding of the KChIP-related protein calmodulin to Ca2+-activated SK channels (Schumacher et al. 2001); the structural analysis of the KChIP1–Kv4.2N30 fusion protein, which was crystallized as a dimer, supports this view (Zhou et al. 2004).
, 百拇医药
Single-particle electron-microscopy analysis of Kv4.2/KChIP2 complexes (Kim et al. 2004b) proposed that Kv4.2/KChIP2 channels are octomeric complexes, in which KChIPs bind to the channel as four individual protein units, each attached to one of the proximal N-termini and placed in a four-fold rotational symmetry lateral to the T1 domain. Furthermore, the cytoplasmic C-terminus of each Kv4.2 -subunit appears in close neighbourhood both to the T1 domain and bound KChIP (Kim et al. 2004b). The results of our mutational analyses strongly support these features of the proposed structure as outlined in Fig. 8. Native brain Kv4 channel complexes may contain not only cytoplasmic KChIPs but membrane-bound dipeptidyl aminopeptidase-like proteins as additional accessory subunits (Nadal et al. 2003). It remains to be seen how the presence of another accessory subunit may influence the interaction of KChIPs with the binding sites identified on the Kv4.2 -subunit. Eventually, high-resolution structural analyses based on cocrystallized complexes containing all relevant structural determinants will be necessary to elucidate the actual Kv4/KChIP binding mode.
, http://www.100md.com
A simple conceptual model was developed, which combines our experimental results with previous Kv4.2/KChIP2 single-particle electron-microscopy analysis (Kim et al. 2004b; contours of the published structure indicated by broken lines). The cartoon schematically illustrates two opposite Kv4.2 channel -subunits with KChIPs bound. The channel core domain; i.e. voltage-sensor (S1–S4) and pore domain (S5 and S6) are in grey, T1 domain in green, proximal N-terminus in orange and cytoplasmic C-terminus in red. KChIPs (blue) are attached laterally, making contact with both the proximal N-terminus and the T1 domain. The cytoplasmic C-terminus constitutes the lateral columns detected with electron microscopy (Kim et al. 2004b) and may come in close proximity to the T1 domain and the proximal N-terminus as well as bound KChIP.
, http://www.100md.com
In summary, the proximal Kv4.2 N-terminus constitutes the major Kv4.2/KChIP-binding interface. There, mutations directly interfere with KChIP binding and KChIP-mediated channel modulation. Mutations which we classified to indirectly interfere with KChIP binding and modulation were either located laterally on the T1 domain or within the cytoplasmic C-terminus of Kv4.2. The T1 domain and cytoplasmic C-terminus may represent KChIP anchoring sites. The data support a macromolecular structure for Kv4.2/KChIP2 complexes as revealed by single-particle electron-microscopy (Kim et al. 2004b). These findings imply that the Kv4.2 channel domains involved in KChIP-dependent gating modulation are disitinct from those found in Shaker-related Kv1 channels for Kvinteraction (Gulbis et al. 2000). The differences in the mode of interaction for the two types of accessory subunits reflect the distinct inactivation pathways and according functional roles demonstrated for Kv1/Kvand Kv4/KChIP channel complexes.
, 百拇医药
References
An WF, Bowlby MR, Betty M, Cao J, Ling HP, Mendoza G, Hinson JW, Mattsson KI, Strassle BW, Trimmer JS & Rhodes KJ (2000). Modulation of A-type potassium channels by a family of calcium sensors. Nature 403, 553–556.
Bhring R, Dannenberg J, Peters HC, Leicher T, Pongs O & Isbrandt D (2001). Conserved Kv4 N-terminal domain critical for effects of Kv channel-interacting protein 2.2 on channel expression and gating. J Biol Chem 276, 23888–23894.
, 百拇医药
Beck EJ, Bowlby M, An WF, Rhodes KJ & Covarrubias M (2002). Remodelling inactivation gating of Kv4 channels by KChIP1, a small-molecular-weight calcium-binding protein. J Physiol 538, 691–706.
Burgoyne RD & Weiss JL (2001). The neuronal calcium sensor family of Ca2+-binding proteins. Biochem J 353, 1–12.
Dixon JE, Shi W, Wang HS, McDonald CYuH, Wymore RS, Cohen IS & McKinnon D (1996). Role of the Kv4.3 K+ channel in ventricular muscle. A molecular correlate for the transient outward current. Circ Res 79, 659–668.
, 百拇医药
Gebauer M, Isbrandt D, Sauter K, Callsen B, Nolting A, Pongs O & Bhring R (2004). N-type inactivation features of Kv4.2 channel gating. Biophys J 86, 210–223.
Gulbis JM, Zhou M, Mann S & MacKinnon R (2000). Structure of the cytoplasmic beta subunit–T1 assembly of voltage-dependent K+ channels. Science 289, 123–127.
Hatano N, Ohya S & Imaizumi Y (2002). Functional interaction between KChIP1 and GFP-fused Kv4.3L co-expressed in HEK293 cells. Pflugers Arch 444, 80–88.
, 百拇医药
Jerng HH & Covarrubias M (1997). K+ channel inactivation mediated by the concerted action of the cytoplasmic N- and C-terminal domains. Biophys J 72, 163–174.
Ju M, Stevens L, Leadbitter E & Wray D (2003). The Roles of N- and C-terminal determinants in the activation of the Kv2.1 potassium channel. J Biol Chem 278, 12769–12778.
Kim LA, Furst J, Butler MH, Xu S, Grigorieff N & Goldstein SA (2004a). Ito channels are octomeric complexes with four subunits of each Kv4.2 and K+ channel-interacting protein 2. J Biol Chem 279, 5549–5554.
, 百拇医药
Kim LA, Furst J, Gutierrez D, Butler MH, Xu S, Goldstein SA & Grigorieff N (2004b). Three-dimensional structure of I(to); Kv4.2-KChIP2 ion channels by electron microscopy at 21 resolution. Neuron 41, 513–519.
Kunjilwar K, Strang C, DeRubeis D & Pfaffinger PJ (2004). KChIP3 rescues the functional expression of Shal channel tetramerization mutants. J Biol Chem 279, 54542–55451.
Kuo HC, Cheng CF, Clark RB, Lin JJ, Lin JL, Hoshijima M, Nguyen-Tran VT, Gu Y, Ikeda Y, Chu PH, Ross J, Giles WR & Chien KR (2001). A defect in the Kv channel-interacting protein 2 (KChIP2) gene leads to a complete loss of I(to) and confers susceptibility to ventricular tachycardia. Cell 107, 801–813.
, 百拇医药
Nadal MS, Ozaita A, Amarillo Y, de Miera EV, Ma Y, Mo W, Goldberg EM, Misumi Y, Ikehara Y, Neubert TA & Rudy B (2003). The CD26-related dipeptidyl aminopeptidase-like protein DPPX is a critical component of neuronal A-type K+ channels. Neuron 37, 449–461.
Nanao MH, Zhou W, Pfaffinger PJ & Choe S (2003). Determining the basis of channel-tetramerization specificity by X-ray crystallography and a sequence-comparison algorithm: family values (FamVal). Proc Natl Acad Sci U S A 100, 8670–8675.
, 百拇医药
Osawa M, Tong KI, Lilliehook C, Wasco W, Buxbaum JD, Cheng HY, Penninger JM, Ikura M & Ames JB (2001). Calcium-regulated DNA binding and oligomerization of the neuronal calcium-sensing protein, calsenilin/DREAM/KChIP3. J Biol Chem 276, 41005–41013.
Ren X, Shand SH & Takimoto K (2003). Effective association of Kv channel-interacting proteins with Kv4 channel is mediated with their unique core peptide. J Biol Chem 278, 43564–43570.
, 百拇医药
Scannevin RH, Wang K, Jow F, Megules J, Kopsco DC, Edris W, Carroll KC, Lu Q, Xu W, Xu Z, Katz AH, Olland S, Lin L, Taylor M, Stahl M, Malakian K, Somers W, Mosyak L, Bowlby MR, Chanda P & Rhodes KJ (2004). Two N-terminal domains of Kv4 K+ channels regulate binding to and modulation by KChIP1. Neuron 41, 587–598.
Schumacher MA, Rivard AF, Bachinger HP & Adelman JP (2001). Structure of the gating domain of a Ca2+-activated K+ channel complexed with Ca2+/calmodulin. Nature 410, 1120–1124.
, 百拇医药
Serodio P, Kentros C & Rudy B (1994). Identification of molecular components of A-type channels activating at subthreshold potentials. J Neurophysiol 72, 1516–1529.
Shibata R, Misonou H, Campomanes CR, Anderson AE, Schrader LA, Doliveira LC, Carroll KI, Sweatt JD, Rhodes KJ & Trimmer JS (2003). A fundamental role for KChIPs in determining the molecular properties and trafficking of Kv4.2 potassium channels. J Biol Chem 278, 36445–36454.
, 百拇医药
Takimoto K, Yang EK & Conforti L (2002). Palmitoylation of KChIP splicing variants is required for efficient cell surface expression of Kv4.3 channels. J Biol Chem 277, 26904–26911.
Zhou W, Qian Y, Kunjilwar K, Pfaffinger PJ & Choe S (2004). Structural insights into the functional interaction of KChIP1 with Shal-type K+ channels. Neuron 41, 573–586.
Zhu X-R, Wulf A, Schwarz M, Isbrandt D & Pongs O (1999). Characterization of human Kv4.2 mediating a rapidly-inactivating transient voltage-sensitive K+ current. Receptors Channels 6, 387–400., http://www.100md.com(Britta Callsen, Dirk Isbr)