| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,||
,1
* Greenberg Division of Cardiology, Department of Medicine,
Department of Pharmacology, and
Department of Anesthesiology, Weill Cornell Medical College, New York, New York, USA;
Department of Cardiology, St. Mary’s Hospital, Imperial College London, London, UK;
|| Department of Medicine, Division of Cardiology, and
¶ Department of Cell Biology, New York University School of Medicine, New York, New York, USA;
# The Cardiovascular Research Center, Mount Sinai School of Medicine, New York, New York, USA; and
** FoxHollow Technologies, Redwood City, California, USA
1Correspondence: Starr 463, Greenberg Division of Cardiology, Weill Medical College of Cornell University, 1300 York Ave., New York, NY 10065, USA. E-mail: gwa2001{at}med.cornell.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunits, was also diminished, by 
25%. Ventricular MiRP1 protein coimmunoprecipitated with native Kv1.5 and Kv4.2 but not Kv1.4 or Kv4.3. Unexpectedly, kcne2 (–/–) ventricular membrane fractions exhibited 50% less mature Kv1.5 protein than wild type, and disruption of Kv1.5 trafficking to the intercalated discs. Consistent with the reduction in ventricular K+ currents and prolonged ventricular APD, kcne2 deletion lengthened the QTc under sevoflurane anesthesia. Thus, targeted disruption of kcne2 has revealed a novel cardiac partner for MiRP1, a novel role for MiRPs in subunit targeting in vivo, and a role for MiRP1 in murine ventricular repolarization with parallels to that proposed for the human heart.—Roepke, T. K., Kontogeorgis, A., Ovanez, C., Xu, X., Young, J. B., Purtell, K., Goldstein, P. A., Christini, D. J., Peters, N. S., Akar, F. G., Gutstein, D. E., Lerner, D. J., Abbott, G. W. Targeted deletion of kcne2 impairs ventricular repolarization via disruption of IK,slow1 and Ito,f.
Key Words: cardiac arrhythmia • Kv1.5 • Kv4.2 • long QT syndrome • MiRP1 • potassium channel
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunits to alter their gating, conductance, and pharmacology (for a review, see ref. 1
) (Fig. 1A
). The functional effects of MiRPs on different Kv channels are diverse and often profound, and naturally occurring mutations in the KCNE genes that encode MiRPs can cause channel dysfunction and disease. Inherited or sporadic mutations in the human KCNE2 gene, which encodes MiRP1, associate with inherited long QT syndrome (LQTS), and common KCNE2 polymorphisms increase susceptibility to acquired (drug-induced) LQTS, suggested to arise from reduced IKr current due to dysfunction and/or increased sensitivity to drug blocking of ventricular MiRP1-hERG channels (2
, 3)
.
|
For simplicity and to follow convention, MiRP1 is hereafter referred to in this report as KCNE2 (for human protein), KCNE2 (human gene), kcne2 (murine protein), kcne2 (murine gene). KCNE2 has a marked promiscuity of function in vitro, being able to regulate hERG, KCNQ1–3, Kv3.1, Kv3.2, Kv4.2, Kv4.3, and even HCN (pacemaker) cardiac ion channel subunits (1)
. This potential for interaction promiscuity, combined with varying reports of the efficacy of KCNE2 in regulating specific channels, the relatively subtle effects of coassembly in some cases, and reportedly low (or nonexistent) cardiac expression of KCNE2 in some studies compared to others, have led to intense debate regarding its putative roles in cardiac physiology (4
5
6
7
8
9
10
11)
. Here, we generated kcne2 (–/–) mice and conducted a series of experiments on these mice to better understand the role of kcne2 in cardiac physiology in vivo. These experiments demonstrated that kcne2 is required for timely ventricular repolarization and that it regulates native murine ventricular IKslow1 and Ito,f by modulating Kv1.5 and Kv4.2, respectively, in the adult mouse heart.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and were housed and utilized according to the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals and Weill Medical College of Cornell University animal care and use policies. All mice in this study were generated from kcne2 (+/–) x (+/–) crosses. The single coding exon, which we previously designated exon 1 (12)
, is now designated exon 2 due to subsequent identification of an untranslated exon in the 5' untranslated region of the kcne2 gene. For reverse transcriptase-polymerase chain reaction (RT-PCR) from cardiac tissue, RNA was extracted from 3 separate preps per genotype using TriZOL (Invitrogen, Carlsbad, CA, USA), then reverse-transcribed to give cDNA as before (13)
. Primers were hypoxanthine-guanine phosphoribosyltransferase (HPRT): forward, 5'-3' TGGAAAGAATGTCTTGATTGTTGA and reverse, 5'-3' ACTTCGAGAGGTCCTTTTCACC, which gives a 130-bp product; Kv1.5: forward, TTATTCTTATGGCTGACGAGTGC and reverse, AAGGCACCAATAGTACATCCCAG, which gives a 203-bp product. kcne2 (–/–) samples were paired with kcne2 (+/+) samples of similar RNA concentration as assessed by measuring band density of PCR products obtained with specific primers for the reference HPRT transcript; then the paired samples were amplified with Kv1.5-specific primers, run on a 1% agarose gel, and stained with ethidium bromide; optical density was measured using a Fluor-S MultiImager (Bio-Rad, Hercules, CA, USA). Results are expressed as Kv1.5 band density in kcne2 (–/–) ventricles normalized to that in kcne2 (+/+) ventricles.
Echocardiography
Transthoracic echocardiograms were recorded in 12- to 15-wk-old conscious-sedated (1% isoflurane in 100% oxygen) kcne2 (+/+) and kcne2 (–/–) mice, as described previously with a Sequoia C256 and 15L8 probe (Acuson, Mountain View, CA, USA) (14)
. Left ventricular end-systolic dimension (LVESD), left ventricular end-diastolic dimension (LVEDD), interventricular septal thickness (IVST), and posterior wall thickness (PWT), both in diastole and systole, were measured at the level of the papillary muscles on the short-axis view using 2-dimensional guided M-mode imaging at 3 cardiac cycles in kcne2 (+/+) and kcne2 (–/–) mice. Left ventricular (LV) fractional shortening (FS) was calculated from the M-mode recordings using the equation FS (%) = (LVEDD– LVESD)/LVEDD x 100.
Electron microscopy
Electron microscopy was performed as described previously (12)
. Briefly, LV tissue samples from kcne2 (+/+) and kcne2 (–/–) mice (2 per gender, per genotype) were washed, fixed, stained, and dehydrated, then infiltrated and embedded in Spurr’s resin. Sections were cut, contrasted with lead citrate, and viewed on a JSM 100 CX-II electron microscope (JEOL, Peabody, MA, USA) operated at 80 kV. Images were recorded on Kodak 4489 Electron Image film (Eastman Kodak, Rochester, NY, USA) then digitized on an Epson Expression 1600 Pro scanner (Seiko Epson, Suwa, Japan) at 900 dpi.
Western blot analysis and coimmunoprecipitation (co-IP)
Crude whole-heart or ventricular (as indicated) membrane fractions for Western blot analysis and co-IP were isolated, according to a previously described potassium iodide protocol (15)
, from 15-wk-old kcne2 (+/+) and kcne2 (–/–) mouse hearts. Membrane fractions in a buffer comprising (in mM) 150 NaCl, 50 Tris-HCl (pH 7.4), 20 NaF, 10 NaVO4, 1 phenylmethylsulfonyl fluoride (Fisher Scientific, Pittsburgh, PA, USA), 1% Nonidet P-40 (Pierce, Rockford, IL, USA), 1% CHAPS (Sigma, St. Louis, MO, USA), 1% Triton X-100 (Fisher Scientific), and 0.5% SDS (Sigma) were incubated on a rocking platform for 1 h at 4°C. Debris was removed by centrifugation at 20,000 g for 5 min, and the supernatant was retained. For Western blot analysis, supernatants (40 µg protein/lane) were size-fractionated by SDS-PAGE, transferred onto polyvinylidene difluoride (PVDF) membranes, and probed with polyclonal anti-kcne2 antibody (Sigma or in-house) or antibodies raised against murine Kv1.5 (Alomone, Jerusalem, Israel), Kv4.2 (Chemicon, Temecula, CA, USA), Kv2.1, Kv4.3, or Kv1.4 (Sigma), and horseradish peroxidase (HRP) -conjugated goat anti-rabbit immunoglobulin G (IgG) secondary antibody (Bio-Rad Labs) for visualization with fluorography (ECL Plus; Amersham Biosciences, Piscataway, NJ, USA). Protein concentration was quantified using the Fluor-S MultiImager (Bio-Rad) after normalization for total protein concentration using the Bradford assay. For co-IPs, supernatants (100 µg/µl total protein concentration) were first precleared with Protein A Sepharose beads (Amersham Biosciences), incubated with antibodies raised against subunits, and precipitated with Protein A Sepharose beads. The beads were then washed, and bound proteins were eluted with SDS-PAGE loading buffer for Western blotting as above.
Immunohistochemistry and immunofluorescence
The immunohistochemical detection of Kv1.5 was performed using a Discovery XT processor (Ventana Medical Systems, Tucson, AZ, USA). Hearts from kcne2 (–/–) and kcne2 (+/+) mice were excised post mortem, fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) at 4°C overnight, and transferred into 30% sucrose. Cryosections were blocked for 30 min in 10% normal goat serum, 2% bovine serum albumin (BSA; Sigma) in PBS. Sections were incubated with 1 µg/ml primary antibody (rabbit polyclonal anti-Kv1.5; Alomone) for 3 h followed by 32 min incubation with biotinylated goat anti-rabbit IgG (Vector Labs, Burlingame, CA, USA) in 1:200 dilution. Secondary Antibody Blocker, Blocker D, streptavidin-HRP and 3,3'-diaminobenzidine (DAB) detection kit (Ventana Medical Systems) were used according to the manufacturer’s instructions.
Double immunofluorescence detection of zonula-occludens 1 (ZO-1) and Kv1.5 was performed using a Discovery XT processor (Ventana Medical Systems). Preceding incubation with 1 µg/ml mouse monoclonal anti-ZO-1 antibody (Zymed, San Francisco, CA, USA), tissue sections were blocked for 30 min in mouse IgG blocking reagent (Vector Labs). Primary antibody incubation was for 3 h, followed by 60 min incubation with 1:200 biotinylated anti-mouse IgG (Vector Labs). Mouse IgG1 (5 µg/ml) was used as an isotype negative control. Detection was performed with Streptavidin-HRP D (Ventana Medical Systems) followed by incubation with Tyramide-Alexa Fluor 488 (Invitrogen). Prior to incubation with rabbit polyclonal Kv1.5 antibody (Alomone; 2.5 µg/ml), tissue sections were blocked for 30 min in 10% normal goat serum, 2% BSA in PBS. Primary antibody incubation was for 3 h, followed by 32 min incubation with biotinylated goat anti-rabbit IgG (Vector Labs; 1:200 dilution). Detection was performed with streptavidin-HRP D (Ventana Medical Systems), followed by incubation with Tyramide-Alexa Fluor 568 (Invitrogen). Stained slides were viewed with a Zeiss Axiovert 200 wide-field microscope (Carl Zeiss, Thornwood, NY, USA), and pictures were acquired using MetaMorph 7.1 software (Molecular Devices, Sunnyvale, CA, USA).
Electrocardiography
Electrocardiographic studies were performed, as described previously (16
, 17)
. Briefly, leads were placed on all four limbs of anesthetized mice, and a baseline electrocardiogram was recorded. Intervals were calculated from leads I, II, and III, and corrected QT interval (QTc) was calculated according to the Mitchell formula, QTc = QT/(RR/100)1/2, where QT is the interval between the start of the Q wave and the end of the T wave, and RR is the interval between the onset of one QRS complex to the onset of the next QRS complex, measured in seconds (18
, 19)
.
Optical action potential (AP) mapping
A state-of-the-art charge-coupled device (CCD) -based optical mapping system, capable of simultaneously measuring high-fidelity fluorescent signals from 6400 recording sites across the explanted intact murine heart (see details below) with high spatial and temporal resolutions, was used. AP was measured, as described previously, using the voltage-sensitive indicator dye di-4-ANEPPS (Molecular Probes, Eugene, OR, USA). A high-intensity filtered excitation light (515±5 nm; Omega Optical, Brattleboro, VT, USA) was obtained from a 180-W quartz tungsten halogen lamp light source (Oriel, Stratford, CT, USA) and directed onto the heart in an epifluorescence manner via custom-built optics designed to maximize the incident excitation light efficiency and homogeneity. Anatomical features of the preparation for each measurement were digitized and stored by the main 80 x 80 pixel CCD detector.
Emitted fluorescent light was collected using a custom-built macroscopic imaging system consisting of a high numerical aperture photographic lens (25 mm, f/0.95, Nikon, Melville, NY, USA). Fluorescent light exiting the detector lens was filtered [>610 nm for AP measurements] and focused onto the CCD detector. Data acquisition software was used to precisely control the timing and duration of recordings in order to minimize unnecessary photobleaching effects and display data for online inspection by the investigators.
Adult kcne2 (+/+) and (–/–) mice were anesthetized with pentobarbital sodium (0.1 ml), and their hearts were rapidly excised and Langendorff-perfused with Krebs solution, containing (in mM) 118 NaCl, 29 NaHCO3, 1.0 MgSO4, 4.8 KCl, 10 dextrose, and 3.4 CaCl2 (pH 7.40; 35°C; 95% O2:5% CO2). Perfusion pressure was maintained at 60–70 mmHg by regulating coronary perfusion flow with a digital dual-head roller pump. Hearts were stained by direct coronary perfusion for 20 min with di-4-ANEPPS, as described previously.
Preparations were placed in a tissue bath on a custom-built frame and gently pressed against a glass imaging window by a stabilizing piston. To avoid surface cooling, the preparations were immersed in the coronary effluent, which was maintained at a constant temperature (36±1°C) equal to the perfusion temperature by a heat-exchanger assembly. Cardiac rhythm was monitored for changes in heart rate or loss of pacing capture via silver electrodes placed in proximity to the mouse heart. Electrocardiograph signals were filtered, amplified, and displayed in real time on a computer monitor using the MP-100 amplifier system and the Acknowledge software package (Biopac Inc, Goleta, CA, USA). Cardiac rhythm, perfusion pressure, and flow were continuously monitored during each experiment.
Data were analyzed offline in a masked fashion, with the use of custom-developed software coded using the Interactive Data Language (IDL) platform. In all experiments, average action potential duration (APD) times (APD 50, 75, and 90) were measured by performing spatial averaging across the array. Both depolarization and repolarization times (at 50, 75, and 90% of the AP amplitude) were measured using validated computer algorithms. Depolarization time was defined as the point of maximum positive derivative in the AP upstroke (dF/dtmax). Repolarization time was measured at 50, 75, and 90% of the AP amplitude. APD was defined as the difference between repolarization time (at 50, 75, and 90%) relative to the depolarization time.
Isolation of adult mouse ventricular myocytes
Mouse ventricular myocytes were isolated as described previously (20)
. Briefly, adult (3- to 4-mo-old) kcne2 (+/+) and kcne2 (–/–) mice were euthanized with CO2, and the hearts were rapidly excised and mounted by aortic cannulation onto a Langendorff apparatus. Subsequently, hearts were perfused retrogradely with warmed (37°C), oxygenated, calcium-free HEPES-buffer containing (in mM) NaCl 137, KCl 5.4, MgSO4 1.2, NaH2PO4 15, glucose 20, HEPES 10, and L-glutamine 2 (pH 7.3, Buffer A). After 5 min, 0.8 mg/ml Type II collagenase (Worthington, Lakewood, NJ) and 10 µM CaCl2 (Buffer B) were added to the solution, and hearts were perfused for 15–20 min, with the temperature of the perfusate and the hearts maintained at 37°C. Following the perfusion, the atria and free right ventricular wall were removed using a fine scalpel and scissors. The left ventricle was cut open, and the ventricular septum and the top 0.3 mm of tissue at the apex of the left ventricle were removed. These tissue pieces were transferred to a falcon tube containing prewarmed Buffer A supplemented with 1.25 mg/ml taurine, 5 mg/ml BSA, and 150 µM CaCl2 (Buffer C). After mechanical dispersion by gentle shaking at 37°C, cells were collected by sedimentation. Isolated myocytes were resuspended in -modified essential medium (-MEM; Invitrogen) supplemented with 10% FBS and antibiotics (penicillin, 100 U/ml; streptomycin, 100 µg/ml), plated on laminin-coated 60 mm culture dishes, and kept in a 5% CO2:95% air incubator at 37°C. Approximately 45 min after plating, the medium was replaced with fresh -MEM to wash off damaged and nonadhering cells. Rod-shaped, striated myocytes of healthy appearance with no spontaneous contractions were subjected to electrophysiological recording within the first 8 h following isolation.
Cellular electrophysiology
Chinese hamster ovary (CHO) cells were cultured and transfected as described previously (21)
with cytomegalovirus-based expression vectors containing cDNAs for murine kcne2 (or blank plasmid control) (2 µg) and murine Kv4.2 (1 µg) 24 h before voltage-clamp experiments. Whole-cell patch-clamp recordings using CHO cells were performed at room temperature using an IX50 inverted microscope equipped with epifluorescence optics for green fluorescent protein detection (Olympus, Center Valley, PA, USA), a Multiclamp 700A Amplifier, a Digidata 1300 Analog/Digital converter and PC with pClamp9 software (Molecular Devices). Bath solution was (in mM) 135 NaCl, 5 KCl, 1.2 MgCl2, 5 HEPES, 2.5 CaCl2, and 10 D-glucose (pH 7.4). Pipettes were of 3–5 M
resistance when filled with intracellular solution containing (in mM) 10 NaCl, 117 KCl, 2 MgCl2, 11 HEPES, 11 EGTA, and 1 CaCl2 (pH 7.2). Cells were stepped from a holding potential of –80 mV to test potentials from –60 to +50 mV in 10-mV increments. Current amplitudes, measured in individual cells, were normalized to cell size (whole-cell membrane capacitance). Data were analyzed using pClamp9.1 software (Molecular Devices), and statistical analysis (ANOVA) was performed using Origin 6.1 (Microcal, Northampton, MA, USA) software.
Similar procedures were used for ventricular myocyte analyses with the following exceptions: the bath solution contained (in mM) 136 NaCl, 4 KCl, 1 CaCl2, 2 MgCl2, 5 CoCl2, 10 HEPES, 10 glucose, and 0.02 tetrodotoxin (TTX) (pH 7.4). The recording pipette solution contained (in mM) 135 KCl, 10 EGTA, 10 HEPES, and 5 glucose (pH 7.2). Outward K+ currents were evoked during 4.5-s voltage steps to test potentials between –60 and +50 mV in 10-mV increments from a holding potential of –70 mV after a 20-ms prepulse to –20 mV. Leak currents were always <100 pA and were not corrected. For inhibition of K+ currents with 4-aminopyridine (4-AP; ICN Biomedicals, Irvine, CA, USA), heteropodatoxin 2 (HpTx2; Alomone) or tetraethylammonium (TEA; Sigma) stock solutions were prepared in bath solution, and applied directly to the bath after "baseline" recordings and allowed to equilibrate for 2–3 min before "drug" recordings. The amplitudes of Ito,f, Ito,s, IKslow, and Iss were determined by fitting the decay phases of the K+ outward currents to the sum of three exponentials, as described previously (22)
. When necessary for fitting, Iss was fixed to the steady-state current value at the end of the test pulse. Correlation coefficients were determined to assess the quality of fits; only fits with correlation coefficients >0.970 were used in this study.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
24
25)
. Previously, we reported that kcne2 (–/–) mice exhibit striking gastric morphological and functional abnormalities, including achlorhydria, gastric hyperplasia, and abnormal parietal cell architecture (12)
. Here, we first established that kcne2 protein is expressed in kcne2 (+/+) but not kcne2 (–/–) hearts by Western blot analysis (Fig. 1B
). Next, echocardiographic analysis of hearts from adult (12–15 wk) kcne2 (–/–) mice bred from kcne2 (+/–) x (+/–) crosses revealed no evidence of LV hypertrophy, chamber dilation, or contractile dysfunction, and no differences in these parameters when compared to kcne2 (+/+) littermates (Fig. 1C
; Supplemental Table 1). Electron micrographs from kcne2 (–/–) mice revealed no structural abnormalities at the cellular level (Fig. 1D
). Similarly, kcne2 disruption did not significantly alter heart size, heart-to-body weight ratio, histological presentation, or serum electrolyte concentrations in 12- to 15-wk-old mice (data not shown).
kcne2 disruption prolongs ventricular APD
Optical mapping studies were performed to quantify ventricular APD in isolated, perfused, intact hearts from kcne2 (+/+) and (–/–) mice. Measurements were performed in a blinded manner to genotype in three separate batches that included age- and sex-matched animals delivered on the same day and using identical conditions for both genotypes. Representative AP recordings are shown in Fig. 2A
, indicating considerable AP prolongation in kcne2 (–/–) compared to kcne2 (+/+) animals. Quantification is shown for the three batches, with each point representing the mean of 6400 recording sites across the ventricle; kcne2 (–/–) hearts consistently demonstrated increased APD (Fig. 2B
), as early and late phases of repolarization were delayed (50, 75, and 90%). These data demonstrate that kcne2 (–/–) hearts have compromised ventricular repolarization, suggesting loss of function of ventricular K+ currents, which was next tested directly using cellular electrophysiological measurements.
|
kcne2 disruption reduces ventricular Ito,f and IK,slow currents but not Ito,s or Iss
Mean whole-cell membrane capacitances were 135.9 ± 4.5 and 137.0 ± 4.7 pF, and mean access resistances were 12.3 ± 0.8 and 12.1 ± 1.0 M, in kcne2 (+/+) and kcne2 (–/–) LV myocytes, respectively (n=59–71). Figure 3
shows representative currents from whole-cell voltage-clamp recordings of LV myocytes isolated from the apex and septum of kcne2 (+/+) and kcne2 (–/–) mice using 4.5-s depolarizing steps from a holding potential of –70 mV to potentials between –60 and +50 mV. We detected four kinetically distinct voltage-gated potassium currents in ventricular myocytes, as previously reported (26)
, from both kcne2 (+/+) and kcne2 (–/–) mice: a fast-activating and fast-inactivating component (Ito,f) present in myocytes from the apex and some from the septum; a more slowly inactivating transient outward current (Ito,s) present in septal myocytes lacking Ito,f; and two currents present in all myocytes tested—an even more slowly inactivating current (IK,slow) and a noninactivating, steady-state current (Iss). Disruption of kcne2 caused statistically significant reductions in mean peak current density (at +40 mV) in LV apex myocytes (22%) and in septal myocytes expressing Ito,f (16%), with 21 and 15% reductions, respectively, in Ito,f density in these two cell types (Fig. 3
; Table 1
). Furthermore, in cells expressing Ito,f, the IK,slow current density was reduced by 32% (apex) and 30% (septum), respectively. Interestingly, in the subset of septal myocytes expressing Ito,s but not Ito,f, the density of IK,slow was unaffected by kcne2 deletion. There were no significant differences in Ito,f or IK,slow gating kinetics due to kcne2 disruption, as derived from fitting of whole-cell currents, even in cells where the density of these currents changed with kcne2 deletion (Table 1)
. Neither the current density nor the gating kinetics of the other currents detected (Ito,s and Iss) were affected by kcne2 disruption (Fig. 3E, F
; Table 1
).
|
|
kcne2 disruption reduces HpTx2-sensitive Ito,f and 4-AP-sensitive IK,slow currents
Functional effects of kcne2 disruption were further dissected pharmacologically. HpTx2 is a relatively specific blocker of Kv4.2 and Kv4.3 currents, which generate Ito,f in murine (Kv4.2 and perhaps Kv4.3) and human (Kv4.3) heart (27
28
29)
. Here, in myocytes expressing Ito,f, kcne2 disruption reduced the mean (300 nM) HpTx2-sensitive current density at +50 mV membrane potential by 26% (from 12.8±0.5 to 9.5±0.7 pA/pF, n=7–8; P<0.05) in the septum and by 21% (from 19.9±0.9 to 15.7±1.4 pA/pF, n=7–8; P<0.05) in the apex (Fig. 4
, top). Thus, currents generated by Kv4.2 and/or Kv4.3 were diminished by kcne2 disruption.
|
Murine ventricular IK,slow has two pharmacologically distinct components: a 50 µM 4-AP-sensitive current generated by Kv1.5 (IK,slow1); and a 25 mM TEA-sensitive current generated by Kv2.1 (IK,slow2) (26
, 30
31
32)
. Here, kcne2 disruption reduced the 4-AP-sensitive IK,slow current (IK,slow1) by 47% (at +50 mV), from 11.1 ± 0.7 to 5.9 ± 0.5 pA/pF (n=10–11; P<0.001; Fig. 4
, middle). In contrast, the TEA-sensitive-component (IK,slow2) was unchanged, with current densities at +50 mV of 7.2 ± 0.8 pA/pF for kcne2 (+/+) vs. 7.2 ± 0.6 pA/pF for kcne2 (–/–) (n=10–13; P>0.05; Fig. 4
, bottom). Thus, currents generated by Kv1.5, not previously thought to interact with kcne2, were also diminished by kcne2 disruption, whereas those generated by Kv2 subunits were unaffected. There were no significant genotype-specific differences in activation or inactivation kinetics, or voltage dependence, observed for any of the three pharmacologically isolated currents studied (data not shown). It should be noted that native murine Ito,f is also sensitive to 4-AP, albeit weakly, 20% inhibition being previously reported with 50 µM 4-AP (33)
. Thus, the 4-AP-sensitive current shown here (Fig. 4
, middle) probably also contains a relatively minor Ito,f component.
kcne2 physically interacts with Kv1.5 and Kv4.2 but not Kv1.4 or Kv4.3 in murine ventricles
The data from Figs. 3
and 4
suggest that kcne2 could interact with Kv4.2, Kv4.3 and/or Kv1.5 in murine ventricles, although gene knockout can also cause electrical remodeling in the heart by altering expression levels of other channel subunits (34)
, and this was another potential mechanism for the effects of kcne2 disruption on Ito,f and IK,slow1. These possibilities were tested using protein biochemistry and RT-PCR (Fig. 5
). First, while mean membrane expression levels of Kv1.4 (which may also generate transient outward currents in murine ventricles; ref. 35
), Kv2.1, Kv4.2, and Kv4.3 protein were unaffected by kcne2 disruption, expression of the 80-kDa band of Kv1.5 was reduced by 46 ± 2%. A lower molecular mass Kv1.5 band thought to correspond to immature, nonglycosylated Kv1.5 (36)
was unaltered by kcne2 disruption (Fig. 5A, B
). The reduction of mature Kv1.5 protein correlated well with the 47% reduction in IK,slow1 current (Fig. 4)
.
|
Co-IP from crude kcne2 (+/+) ventricular membrane preparations using anti-Kv4.2 or anti-Kv1.5 antibodies pulled down kcne2, whereas antibodies raised against Kv4.3 or Kv1.4 did not (Fig. 5C
). This suggested that kcne2 modulates Ito,f in murine ventricles by direct physical interaction with Kv4.2, but not Kv4.3 or Kv1.4. Furthermore, together with the subunit expression data, the results suggested that in kcne2 (+/+) ventricles, kcne2 interacts directly with Kv1.5 to augment IK,slow1 current, perhaps by promoting maturation or surface expression. This hypothesis was further supported by semiquantitative RT-PCR, which failed to show a difference in Kv1.5 mRNA levels between kcne2 (+/+) and (–/–) ventricles (Fig. 5D
).
kcne2 is required for Kv1.5 localization at the intercalated discs in murine ventricular myocytes
Immunohistochemical labeling of murine ventricular tissue using anti-Kv1.5 antibody strongly supported, in conjunction with the preceding data, a role for kcne2 in Kv1.5 surface trafficking. Specifically, in kcne2 (+/+) ventricular myocytes, Kv1.5 was localized prominently in the intercalated discs, as previously reported (37
, 38)
; in contrast, in kcne2 (–/–) ventricular myocytes Kv1.5 appeared to be absent from the intercalated discs (Fig. 6A
). This result was supported by immunofluorescence colabeling of Kv1.5 and ZO-1, a marker for the intercalated discs (39)
. Kv1.5 colocalized with ZO-1 in kcne2 (+/+) but not kcne2 (–/–) myocytes (Fig. 6B
).
|
kcne2 disruption predisposes to QT prolongation with sevoflurane anesthesia
Some human KCNE2 variants are not associated with pathological effects at baseline, but precipitate rhythm disturbances, including prolonged QT interval in the presence of select pharmacological agents or proarrhythmic factors such as hypokalemia, a phenomenon known as acquired long QT syndrome (aLQTS) (3)
. Here, kcne2 (–/–) mice showed an increased propensity for QT interval prolongation when anesthetized with sevoflurane, a volatile anesthetic previously shown to inhibit Kv4 channels by speeding their inactivation (40)
. With 3% sevoflurane (1.5x MAC, minimum alveolar concentration), QTc was not significantly prolonged in kcne2 (+/+) mice compared to values obtained using 1% isoflurane, a volatile anesthetic that does not prolong QTc in mice (41)
. In contrast, in kcne2 (–/–) mice, QTc in the presence of sevoflurane was significantly longer (by 15%) than that in kcne2 (+/+) mice, and also than that in kcne2 (–/–) mice in the presence of isoflurane. Heterozygous kcne2 (+/–) mice had a mean QTc with sevoflurane anesthesia intermediate between those of the two homozygous genotypes (Table 2
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunits that repolarize the ventricular myocardium, relatively little is known about the specific roles of potassium channel ancillary subunits in cardiac repolarization and how they modulate Kv subunits in vivo. The transient outward current, generated by Kv4.2, Kv4.3 (Ito,f), and Kv1.4 (Ito,s) subunits, is considered an important repolarizing force in murine ventricles (20)
. KCNE2 has been found in vitro to modulate several cardiac-expressed Kv subunits, including hERG, KCNQ1 (42)
, pacemaker (HCN) channels (43
, 44)
, and also Kv4 subfamily members 2 and 3 (10
, 45
, 46)
. Previously reported results from Xenopus laevis oocyte coexpression studies showed that human KCNE2 augments Kv4.2 current and slows its time-to-peak and inactivation (45)
. This augmentation is consistent with the reduction in Ito,f observed here in murine ventricular myocytes of kcne2 (–/–) mice. Our data strongly suggest that in adult murine myocytes, kcne2 modulates Ito,f, the HpTx2-sensitive component of Ito, by interaction with Kv4.2. To confirm a functional interaction between murine kcne2 and Kv4.2 subunits, these were coexpressed here in CHO cells, revealing Kv4.2 current augmentation by murine kcne2 (Supplemental Fig. 1) consistent with the previous results using human subunits. The data argue against native interaction with Kv4.3, although it is important to note that the relatively stringent co-IP conditions we employed (see Materials and Methods) may disrupt some native subunit-subunit interactions. Similarly, a kcne2-Kv1.4 interaction in the membrane fraction of murine ventricular myocytes is unlikely given the lack of coassembly and lack of effects of kcne2 deletion on Ito,s observed here.
kcne2 modulates IK,slow1 in murine ventricles
Analogous to the effects of kcne2 deletion described herein, in a previous study of mice, in which Kv1.5 was replaced by Kv1.1 (relatively 4-AP-insensitive), the mice exhibited decreased density of ventricular IK,slow1 current (32)
. Kv1.5 modulation by kcne2 has not previously been reported; here, we demonstrate their functional and physical interaction in vivo. The data also suggest that kcne2 promotes surface expression, and/or targeting of Kv1.5 to specific myocyte compartments, a property not previously attributed to KCNE subunits. However, MinK (KCNE1) was previously shown to increase hERG current by an unknown mechanism, postulated to involve increasing the active fraction of hERG channels at the cell surface (47)
. In kcne2 (–/–) mice, RT-PCR did not show differences in Kv1.5 mRNA compared to kcne2 (+/+) mice, but the mature form of Kv1.5 was expressed relatively less in kcne2 (–/–) ventricular membrane fractions. This was not accompanied by a relative increase (or decrease) in the immature form of Kv1.5 in kcne2 (–/–) ventricular membrane fractions, suggesting either that misprocessed Kv1.5 is degraded in kcne2 (–/–) mice or that it is trafficked to a subcellular compartment not present in the crude membrane fraction enriched using our protocol.
Previous studies demonstrate that KCNE subunits influence trafficking of subunits to the cell surface in vivo in Caenorhabditis elegans (48
, 49)
. The surprising finding here that trafficking of Kv1.5 to the intercalated discs, where it is normally enriched, was strongly inhibited by kcne2 deletion suggests that KCNE subunits may act to chaperone their subunit partners to specific subcellular compartments—a novel and exciting property for these versatile ancillary subunits. The cytoplasmic KChAP, KChIP, and Kvβ subunits are also known to promote surface expression of their -subunit partners (50
51
52)
, indicating that -subunit chaperoning may be a common role for all of these ancillary subunits. Further studies are required to investigate the mechanism or mechanisms by which kcne2 augments Kv1.5 current in vivo and to elucidate whether this functional interaction can be recapitulated in other systems. It was interesting that in myocytes lacking Ito,f, the density of IK,slow was not changed by kcne2 disruption (Table 1)
. Although not directly tested here, one possible explanation for this might be that kcne2 is not expressed in this specific myocyte subpopulation, but there could be other regulatory processes or ancillary subunits determining in which cells particular interactions occur.
Other subunits interact with Kv1.5 and Kv4.2 in vivo
Kv1.5 and Kv4.2 can each interact with a variety of other ancillary subunits, including KChIPs, KChAPs, and Kvβ, and may do so in the heart (53
, 54)
. Kv4.2 has also been suggested to form complexes with Kv4.3 subunits in murine ventricular myocytes. Previously reported in vitro effects of KCNE2 on Kv4.3, thought to be the primary correlate of human ventricular Ito,f, are to slow time-to-peak and inactivation, although there is disagreement over the effects of KCNE2 on Kv4.3 current density, possibly due to cell type. Studies using human embryonic kidney (HEK) 293 cells showed augmentation of Kv4.3 current by KCNE2 (55)
compared to an inhibitory effect with KCNE2 in COS-7 and CHO cells (10
, 46)
.
Recently, Kv4.2 gene disruption was shown to completely eliminate Ito,f in murine ventricles, despite the observation that Kv4.3 protein expression was detectable and unaffected, suggesting that either Kv4.3 does not contribute to Ito,f, or that it cannot form Ito,f channels in the absence of Kv4.2 (although homomeric Kv4.3 channels are functional in vitro) (34)
. Here, we could not detect kcne2-Kv4.3 complexes in ventricular membrane fractions but could detect kcne2-Kv4.2 complexes; this argues against the existence of heteromeric kcne2-Kv4.2-Kv4.3 channels in vivo, as these would be expected to yield kcne2-Kv4.3 co-IPs, but does not rule out a Kv4.2-Kv4.3 interaction without kcne2 in the complex in vivo. Both these Kv4 subunits are also thought to coassemble in vivo with KChIP2 and Kvβ1 (50
, 55
, 56)
, cytoplasmic ancillary subunits that promote subunit surface expression and (in the case of KChIP2) slow inactivation. Further studies will reveal whether kcne2-Kv4.2-KChIP2 complexes, or complexes containing Kv1.5, kcne2, and Kvβ or other cytoplasmic subunits exist in vivo; functional studies in vitro suggest they are possible (55)
.
Lack of detectable subunit remodeling in kcne2 (–/–) murine ventricles
Transgenic and gene-knockout mouse models are a powerful tool for researchers to identify molecular targets for K+ currents in the heart and to study gene function in vivo, but interpretation of effects can be stymied by structural and electrical remodeling. Mice expressing a dominant-negative truncated form of Kv4.2 exhibit cardiac hypertrophy, in contrast to those with complete elimination of the Kv4.2 gene, but similar to those overexpressing an LQT2-associated HERG mutation (25
, 34
, 57
, 58)
. We previously reported that mice lacking kcne2 show remodeling in the stomach wall, with up-regulation of KCNQ1—the subunit partner of kcne2 in parietal cells—and gastric hyperplasia due to the concomitant loss of gastric acid secretion (12)
. Here, we observed no cardiac structural abnormalities, or alterations in expression levels of subunits that could be indicative of remodeling. This, coupled with the functional effects and pharmacological and co-IP studies described herein lead us to conclude that kcne2 modulates Ito,f and IK,slow1 by direct physical interaction with Kv4.2 and Kv1.5, respectively, in adult murine ventricles, and not by remodeling.
KCNE2 in human and murine cardiac physiology and pathophysiology
We previously presented evidence that KCNE2 contributes to human ventricular repolarization by formation of IKr complexes with the hERG potassium channel subunit, the major repolarizing force in human ventricular myocardium (2)
. Sporadic or inherited mutations in KCNE2 that impair the function of human KCNE2-hERG channels are associated with inherited LQTS (2
, 59
, 60)
. aLQTS is more prevalent than the purely inherited form and requires environmental factors such as hypokalemia or drug interaction for pathophysiological consequences (61)
. aLQTS is thought to be commonly caused by pharmacological inhibition of the IKr current because hERG channels are particularly susceptible to drug blocking (62)
. Some cases of aLQTS appear to also require a genetic component, and mutations or even common polymorphisms in KCNE2 are enriched in cohorts of aLQTS patients compared to asymptomatic control subjects (2
, 3)
. Some KCNE2 sequence variants increase sensitivity of KCNE2-hERG channels to blocking by QT-prolonging drugs; other KCNE2 variants impair KCNE2-hERG channel function at baseline, and these effects are thought to superimpose on normally tolerated levels of pharmacological inhibition of IKr to prolong the QT interval (3
, 63)
.
Here, kcne2 deletion reduced specific murine repolarizing currents and prolonged ventricular action potential durations, but a known Kv4-inhibiting (40)
and QT-prolonging (64)
anesthetic, sevoflurane, was needed to bring out kcne2-specific changes in the QTc. Human patients with KCNE2 variants associated with inherited LQTS have thus far all been reported to be heterozygous, with one wild-type and one mutant allele, whereas even homozygous kcne2 null mice showed a normal QTc in the absence of sevoflurane. The most obvious differences between the two circumstances are that KCNE2 probably regulates a different set of subunits (including hERG and perhaps KCNQ1) in human ventricles; furthermore, the macroscopic manifestations of impaired cellular repolarization are somewhat different in murine vs. human ventricles (65)
. However, it is also important to acknowledge the difference between a point mutation and loss of a gene product. Even with ancillary subunits, point mutations can exert dominant-negative effects. In contrast, absence of one copy of an ancillary subunit gene manifests as haploinsufficiency; with absence of both copies of an ancillary subunit gene, the channel subunits, which function as homomeric channels, are still present, albeit with altered properties (including altered cellular distribution, as demonstrated here). Interestingly, wild-type kcne2 reduces outward hERG and KCNQ1 currents in vitro; therefore, deletion of kcne2 would be predicted to increase these currents in cases in which these subunits form complexes in vivo, something we intend to test next using neonatal murine ventricular myocytes, which unlike adult murine myocytes, express mERG and KCNQ1 to a significant extent.
The coassembly with and functional regulation of Kv1.5 by kcne2 in murine ventricles is potentially significant for the human heart. Kv1.5 is not thought to be expressed in human ventricles but is expressed in human atria (66)
, where KCNE2 is also expressed (67)
. The possibility arises, then, that KCNE2 regulates Kv1.5 in human atria. This demands further study because Kv1.5 is currently a major target candidate for atrial fibrillation drugs, which block Kv1.5 to prolong atrial refractoriness without risk of proarrhythmia due to ventricular APD prolongation (68
, 69)
. Interestingly, human KCNE2 gene variants have been tentatively associated with atrial fibrillation, although the emphasis in these cases has been focused on functional effects on KCNE2-KCNQ1 channels, which are hypothesized to occur in the human atrium (67)
.
In conclusion, this genetic examination of the specific cardiac roles of kcne2 shows that it regulates murine ventricular repolarization, as we suggested it does in human heart, albeit via hERG in human ventricles compared to Kv1.5 and Kv4.2 in mice. The possibility that KCNE2 also regulates Kv4 channels in human ventricles, and Kv1.5 channels in human atria, and that KCNE2 mutations cause disease via disruption of these channels, should be considered.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 22, 2008. Accepted for publication June 5, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Jin, E. R. Chemaly, A. Lee, C. Kho, L. Hadri, R. J. Hajjar, and F. G. Akar Mechanoelectrical remodeling and arrhythmias during progression of hypertrophy FASEB J, February 1, 2010; 24(2): 451 - 463. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Sole, M. Roura-Ferrer, M. Perez-Verdaguer, A. Oliveras, M. Calvo, J. M. Fernandez-Fernandez, and A. Felipe KCNE4 suppresses Kv1.3 currents by modulating trafficking, surface expression and channel gating J. Cell Sci., October 15, 2009; 122(20): 3738 - 3748. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Maffie and B. Rudy Weighing the evidence for a ternary protein complex mediating A-type K+ currents in neurons J. Physiol., December 1, 2008; 586(23): 5609 - 5623. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
