OUP user menu

Temporal patterns of electrical remodeling in canine ventricular hypertrophy: Focus on IKs downregulation and blunted β-adrenergic activation

Milan Stengl, Christian Ramakers, Dirk W. Donker, Ashish Nabar, Andrew V. Rybin, Roel L.H.M.G. Spätjens, Theo van der Nagel, Will K.W.H. Wodzig, Karin R. Sipido, Gudrun Antoons, Antoon F.M. Moorman, Marc A. Vos, Paul G.A. Volders
DOI: http://dx.doi.org/10.1016/j.cardiores.2006.07.015 90-100 First published online: 1 October 2006

Abstract

Objectives: Electrical remodeling in cardiac hypertrophy often involves the downregulation of K+ currents, including β-adrenergic (β-A)-sensitive IKs. Temporal patterns of ion-channel downregulation are poorly resolved. In dogs with complete atrioventricular block (AVB), we examined (1) the time course of molecular alterations underlying IKs downregulation from acute to chronic hypertrophy; and (2) concomitant changing responses of repolarization to β-adrenergic receptor (β-AR) stimulation.

Methods and Results: Serial left-ventricular (LV) biopsies were collected from anesthetized dogs during sinus rhythm (SR; control) and at 3, 7 and 30 days of AVB. KCNQ1 mRNA and protein decreased within 3 days (protein expression 58±10% of control), remaining low thereafter. β1-AR mRNA and protein decreased more gradually to 53±8% at 7 days. In chronic-AVB LV myocytes, IKs-tail density was reduced: 1.4±0.3 pA/pF versus 2.6±0.4 pA/pF in controls. β-A enhancement of IKs was reduced. Isoproterenol shortened action-potential duration in control cells, while causing heterogeneous repolarization responses in chronic AVB. β-A early afterdepolarizations were induced in 4 of 13 chronic-AVB cells, but not in controls. In intact conscious dogs, isoproterenol shortened QTc at SR (by −8±3% from 295 ms), left it unaltered at 3 days AVB (+1±3% from 325 ms) and prolonged QTc at 30 days (+6±3% from 365 ms).

Conclusions: Profound decrease of KCNQ1 occurs within days after AVB induction and is followed by a more gradual decrease of β1-AR expression. Downregulation and blunted β-A activation of IKs contribute to the loss of β-A-induced shortening of ventricular repolarization, favoring proarrhythmia. Provocation testing with isoproterenol identifies repolarization instability based on acquired channelopathy.

Keywords
  • Ion channels
  • Remodeling
  • Autonomic nervous system
  • Membrane potential
  • Ventricular function

1. Introduction

Electrical remodeling in cardiac hypertrophy often comprises the downregulation of sarcolemmal K+ currents, including the delayed-rectifier K+ current IKs [1,2]. K+-current downregulation may act to prolong the action-potential duration and enhance excitation–contraction coupling during chronic overload. However, it can confer exaggerated spatial and temporal heterogeneities of ventricular repolarization, rendering the heart more susceptible to deleterious tachyarrhythmia [3]. Temporal patterns of ion-channel downregulation during chronic overload are still poorly resolved.

In the canine model of chronic complete atrioventricular block (AVB) and ventricular hypertrophy, increased spatial [1,4] and temporal dispersion of repolarization [5] predispose to acquired torsades de pointes and sudden cardiac death [6]. At the cellular level, K+ currents IKs and IKr are reduced [1], whereas sarcoplasmic reticulum Ca2+ release and INaCaX are enhanced, especially at slow heart rates [7]. An increase in subsarcolemmal Na+ concentration underlies the altered Ca2+ handling [8]. Previous work has shown that the reduction of IKs in chronic AVB can be due to a downregulation of KCNQ1- and KCNE1-gene transcription in the basal and midlateral parts of the left-(LV) and right-ventricular wall [9]. In the interventricular septum only KCNQ1 mRNA is decreased [10]. These findings indicate regional heterogeneity of ion-channel remodeling in the overloaded myocardium. Until now the time courses of electromolecular changes have remained unclear, especially in the early phase of overload.

Modern insights support the contention that an intact basal IKs expression plus β-adrenergic receptor (β-AR) stimulation are both essential for IKs to figure in repolarization, which has been demonstrated in the rabbit [11], dog [12,13] and human heart [14]. This suggests that the separate or combined loss of IKs-channel subunits and IKs-relevant β-adrenergic (β-A) signaling molecules in acquired cardiac pathology carries a risk of repolarization instability and arrhythmia. In relation to this, failure of appropriate action-potential shortening to β-AR stimulation, and torsades-like arrhythmias have previously been reported in LV wedge preparations under pharmacological IKs inhibition [15,16].

The present study was designed to examine the time course of molecular alterations underlying IKs downregulation in the canine LV after AVB induction. We quantified the mRNA and protein expression of KCNQ1 and KCNE1 subunits and β-AR subtypes in serial myocardial biopsies, and measured catecholamine levels in blood plasma. Concomitant changes of cellular and in-vivo repolarization were closely examined. We hypothesized that the propensity of repolarization to shorten upon β-AR stimulation would be undermined after AVB, favoring proarrhythmia during maintained overload.

2. Methods

Animal handling was in accordance with the ‘European Directive for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes (86/609/EU)’. The experiments were approved by the Committee for Experiments on Animals of Maastricht University and conformed with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Adult mongrel dogs of either gender were used. Premedication consisted of 10 mg oxycodone HCl, 1 mg acepromazine maleate and 0.5 mg atropine sulphate i.m. Complete anesthesia was induced by thiopental (20 mg/kg i.v.) and maintained by inhalation of halothane (0.5–1.0%) and N2O and O2 (2:1). Ampicilline (1000 mg) was administered before and after the experiment. For additional analgesia, buprenorphine was given (0.015 mg/kg i.m.) after the operation.

2.1. In-vivo studies

AVB was induced by radiofrequency catheter ablation of the His bundle under anesthesia (ndogs=31). In-vivo experiments were performed serially at sinus rhythm (SR) and at 0, 3, 8 and 30 days of AVB, using each dog as its own control. Venous blood samples were collected from these trained animals at conscious state (after minutes of quiescence) for determination of plasma catecholamines (see below). Premedication was subsequently administered to achieve conscious sedation, and a standard 6-lead ECG was registered along with a unipolar recording at the chest position corresponding to V4. Under these conditions, a short-lasting β-A challenge with isoproterenol (1 μg/kg, i.v.; provocation testing) was given to increase heart rate >25% during SR or idioventricular rhythm, followed by relaxation (ndogs=6). The QT interval was measured and corrected for heart rate by Van-de-Water's formula: QTc=QT−0.087(RR−1000) [17]. Beat-to-beat variability of QT interval was quantified as short-term variability of 30 consecutive beats: STV=Σ (QTn−QTn+1)/30×√2 [5].

In 12 animals, a minimally-invasive procedure was used for biopsy sampling under closed-chest conditions, allowing fast recovery of the dogs and permitting repetitive sampling at SR and 3, 7 and 30 days of AVB. Intramural needle biopsies (1×1×10 mm) extending from the epicardium into the LV wall were obtained from the apicolateral region under fluoroscopic guidance with a commercially-available automated biopsy device (14-G, Acecut, TSK Laboratory, Japan). This device was percutaneously inserted into the thorax, i.e., through a small dermal incision near the processus xyphoideus. At least 2 biopsies were obtained per experiment. Tissue quality was excellent in nearly all samples (>98%). Histology confirmed minimal injury to the myocardium at the sampling sites. Biopsies were frozen to −80 °C in isopentane chilled with liquid nitrogen and then stored until molecular analysis. Dogs were monitored by echocardiography after this procedure to exclude cardiac tamponade, although in our experience this closed-chest approach is generally uncomplicated (≥95%) and fatal outcome is rare (<3%). Such complications did not occur in this study.

2.2. Analysis of plasma catecholamines

Blood samples (10 mL) were collected into tubes containing lithium heparin as the anticoagulant and transported to the laboratory on ice within 30 min. Samples were centrifuged at 3000 rpm for 10 min at 4 °C. 100 μL of glutathione (95 g/L) was added to 2 mL plasma before storage at − 20 °C. For analysis, plasma was mixed with an internal standard (3,4-dihydroxybenzylamine-hydrobromide) followed by liquid–liquid extraction with 99% n-heptane and 1-octanol/acetic acid. After evaporation of the organic phase, 100 μL of the acetic acid layer was injected and analyzed by reversed phase-high performance liquid chromatography with electrochemical detection.

2.3. mRNA and protein expression studies

Total RNA was isolated from the biopsies using the RNEasy MINI kit (Qiagen/Westburg, Leusden, The Netherlands) according to the manufacturer's protocol. DNAseI-digestion was performed after elution of the total RNA from the column. First-strand cDNA was synthesized using 1 μg total RNA in combination with an anchored oligo-dT14VN and reverse 18S rRNA primer [18]. Fluorescence-based kinetic real-time PCR was performed using a LightCycler™ system (Roche Diagnostics Nederland, Almere, The Netherlands) in combination with the intercalating fluorescent dye SYBR Green I [10]. Amplicons (KCNQ1, KCNE1, and β1-, β2- and β3-ARs) were quantified using the LinRegPCR method [19] and were expressed relative to the constitutively active 18S rRNA gene. For primers used see Table 1.

View this table:
Table 1

Primers

TranscriptForward (F)/reverse (R)TAnneal (°C)Amplicon length (bp)NCBI
Accession number
KCNQ1F: GTCGAGTTTGGCAGCTATGCA60274Not available
R: CGCAAAGAAGGAGATGGCAA
KCNE1F: ATGCTGAGTTACATCCGCTC5798Not available
R: TACGCCTTGTCCTTCTCCT
β1-adrenergic receptorF: TACAACGACCCCAAGTGCT55100U73207
R: AGGTACACGAAGGCCATGA
β2-adrenergic receptorF: AATCCCCTTATCTACTGCCG5593U73206
R: ATTCCCATAGGCCTTCAGG
β3-adrenergic receptorF: TTCTGTCCCTGACTCCATCA55106U92468
R: AAGTCCTCCTGACACTGCTGT

Protein expression was analyzed by Western blotting and densitometric quantification. Proteins were extracted from total homogenates derived from unfixed 5-μm-thick cryosections dissolved in equal amounts of lysis buffer containing 62 mmol/L Tris-HCl, 1.25 mmol/L EDTA, 2% NP40, 2.5 mmol/L PMSC, 12.5 μg/mL leupeptin, 12.5% glycerol, 100 μg/mL aprotinin, 2.3% SDS for 30 min on ice and sample buffer containing 62 mmol/L Tris-HCl, 2.3% SDS, 10% glycerol, 5% β-mercaptoethanol, 0.05% bromophenol blue for 4 min at 95 °C. Equal amounts of total protein (30 μg/lane), determined by protein assay, were loaded onto each lane of Tris-HCl precast polyacrylamide gradient gels (4–20%) (BioRad Laboratories, Hercules, CA, USA). After separation, proteins were transferred to PVDF membranes (BioRad Laboratories, Hercules, CA, USA). Transfer and loading was controlled by Ponceau S and Coomassie brilliant blue staining. Immunostaining was performed with primary antibodies directed against KCNQ1 (1:50, Santa Cruz Biotechnology, Santa Cruz, CA, USA), KCNE1 (1:100, Alomone Labs, Jerusalem, Israel), β1-AR (1:250, Calbiochem, Merck Biosciences, Darmstadt, Germany), and β2-AR (1:500, Chemicon, Temecula, CA, USA). Membranes were blocked for at least 15 min in 2% nonfat dry milk (NFDM) in Tween-buffer (TWB), containing 0.13% Tween 20 in phosphate-buffered saline (PBS). Incubation with the primary antibodies was done for 2 h in 1% NFDM or 5% BSA. After 3×5 min washing in TWB, blots were incubated for 1 h using the appropriate peroxidase-conjugated secondary antibodies: swine anti-rabbit (1:5000, Dako Cytomation, Glostrup, Denmark), goat anti-chicken (1:10,000, Jackson Immuno Research, West Grove, PA, USA) and bovine anti-goat (1:250, Santa Cruz Biotechnology, Santa Cruz, CA, USA), washed in TWB (3×5 min) and visualized on X-ray films (Fujifilm, Rotterdam, The Netherlands) using ECL or ECLplus (Amersham Biosciences, Amersham, UK). Densitometric quantification was performed with Quantity One® software using a GS 800 scanner (BioRad Laboratories, Hercules, CA, USA).

2.4. Cellular experiments

Hearts were quickly excised from anesthetized SR (ndogs=20; controls) and AVB (ndogs=22) dogs. For the cell-isolation method see Ref. [1]. A collagenase-digested part of the apicolateral LV wall was dissected and myocytes were isolated from most of the transmural wall except for a rim (≥1.5 mm) of epicardial and endocardial tissue. Only quiescent rod-shaped cells with clear cross-striations were used for the experiments.

Transmembrane potentials were recorded using high-resistance (30–60 MΩ) glass microelectrodes filled with 3 mol/L KCl. Whole-cell currents were measured using patch pipettes with resistances of 1.0–3.0 MΩ when filled with internal solution. Experiments were performed at 37±1 °C. Cell capacitances were 186±6 pF (ncells=54) in control and 192±9 pF (ncells=36) in AVB cells.

The external solution contained (in mmol/L): NaCl 145, KCl 4.0, CaCl2 1.8, MgCl2 1.0, glucose 11, HEPES 10, pH 7.4 with NaOH at 37 °C. In most patch-clamp experiments KCl was omitted from the external solution to increase IKs amplitude and suppress IK1 [1,20]. The patch–pipette solution contained (in mmol/L): potassium aspartate 125, KCl 20, MgCl2 1.0, MgATP 5, HEPES 5, EGTA 10, pH 7.2 with KOH. L-type Ca2+ current (ICaL) was blocked with nifedipine (5 μmol/L). Almokalant was dissolved in the superfusate (2 μmol/L) to block IKr. Isoproterenol was first dissolved in distilled water containing 30-μmol/L ascorbic acid, kept at 5 °C and in the dark until use. HMR1556 ((3R,4S)-(+)-N-[3-hydroxy-2,2-dimethyl-6-(4,4,4-trifluorobutoxy)chroman-4-yl]-N-methylmethanesulfonamide, Aventis Pharma Deutschland GmbH, Sanofi–Aventis, Frankfurt am Main, Germany) was used to block IKs. This compound is highly selective for IKs: at a concentration of 500 nmol/L it blocks the current completely [13], whereas other currents are unaffected [21]. HMR1556 was dissolved in dimethyl sulfoxide as 100 mmol/L stock solution and diluted in the superfusate to achieve a final concentration immediately before experimental application. The final concentrations of dimethyl sulfoxide (maximally 0.1%) in the superfusate had no measurable effects on ion currents and action potentials.

2.5. Statistics

The data are expressed as mean±SEM. Data comparisons were made with the Student's t test for unpaired or paired data groups. Serial changes (mRNA, protein expressions; QT intervals) were tested using linear-regression analysis and repeated-measures ANOVA. Differences were considered statistically significant if P<0.05.

3. Results

3.1. Temporal patterns of mRNA and protein expression of IKs subunits and β-adrenergic receptors during emerging hypertrophy

The rate of hypertrophic growth was maximal within 1 week of AVB induction, when echographic LV mass increased by 2.2±0.4 g/day [22]. At chronic AVB, heart weights (upon sacrifice) were >30% heavier than SR controls (306±17 g versus 233±11 g, respectively; P<0.05). When corrected for body weight, this difference remained significant: 10.7±0.3 g/kg versus 8.7±0.5 g/kg, respectively (P<0.05), confirming the presence of hypertrophy.

KCNQ1 mRNA decreased to 58±13% of control (P<0.05) already after 3 days of AVB and further down to 42±9% at 30 days. KCNQ1 protein paralleled this decline: 58±10% of control after 3 days, 53±10% at 30 days (P<0.05; Fig. 1). Expression levels of KCNQ1 at chronic AVB match with previous findings [10]. KCNE1 mRNA and protein did not change significantly after AVB at any of the time points (Fig. 1).

Fig. 1

Temporal patterns of mRNA and protein expression of IKs subunits and β-ARs after AVB. Left upper panels, KCNQ1. Right upper panels, KCNE1. Left lower panels, β1-AR. Right lower panels, β2-AR. Serial myocardial biopsies were obtained at SR and 3, 7 and 30 days of AVB in 12 dogs. *, P<0.05 versus SR.

β1-AR mRNA decreased more gradually to 73±5% of control (P<0.05) at 7 days and 63±8% at 30 days. β1-AR protein showed a similar significant decline, as illustrated in Fig. 1. β2-AR mRNA and protein remained unaltered (Fig. 1). Levels of β3-AR mRNA were too low to be detected and quantified reliably, and were not examined by Western blotting.

For control purposes, we collected serial biopsies (7, 14 and 49 days apart) in 3 dogs with sinus rhythm. Western-blot analysis revealed no significant differences in protein expression of KCNQ1 and β1-AR, indicating that the observed changes after AVB (in the other dogs) were not contaminated by the biopsy sampling itself.

3.2. Biophysical properties of downregulated IKs

Biophysical properties of IKs in myocytes from dogs with chronic AVB (ncells=15; ndogs=7) were compared to controls (ncells=19; ndogs=12). Consistent with protein downregulation, IKs density was significantly smaller in chronic AVB: 1.4±0.3 pA/pF versus 2.6±0.4 pA/pF in controls (P<0.05; tail currents at −25 mV after 3-s steps to 50 mV; Fig. 2A). The kinetics of both activation and deactivation of IKs were not different between chronic-AVB and control myocytes (Fig. 2B, Table 2). Voltage dependence of IKs activation was also similar in chronic-AVB and control cells (Fig. 2C). The sensitivity of IKs to pharmacological block by HMR1556 remained unchanged in chronic-AVB myocytes (ncells=6) with an IC50 of 81 nmol/L versus 78 nmol/L in controls (ncells=6; Fig. 2D).

Fig. 2

Downregulation of IKs in chronic AVB cells. (A) Chronic AVB (CAVB) led to a decrease of IKs density. (B) Kinetics of both activation and deactivation of IKs were not different between AVB and control myocytes. Upper panels, arrows indicate IKs traces at depolarizations to 40 mV (SR versus CAVB), which were used for comparison of kinetics after artificial equalization of amplitude (in left panel below; see also Table 2). Vertical bar, 500 pA. Horizontal bar, 1 s. Inset, voltage-clamp protocol. (C) Voltage dependence of IKs activation was similar in CAVB (filled circles) and control cells (open squares). (D) Sensitivity of IKs to pharmacological block by HMR1556 remained unchanged in CAVB cells (filled circles). *, P<.05.

View this table:
Table 2

Kinetics of IKs activation and deactivation

Control, halftimes (ms)Chronic AVB, halftimes (ms)
ActivationDeactivationActivationDeactivation
Baseline511±63128±13509±36121±9
Iso (100 nmol/L)412±53*185±15*429±36*171±20*
  • Activation, depolarizing step from −50 mV to +20 mV. Deactivation, repolarization to −25 mV. Numbers: ncells=10, ndogs=8 for chronic AVB; ncells=10, ndogs=6 for control. Iso, isoproterenol. *P<0.05 versus baseline.

IKs showed a decreased responsiveness to isoproterenol in chronic-AVB myocytes. SC50 values were 63 nmol/L for AVB and 30 nmol/L for control cells. Maximal stimulation at 300 nmol/L isoproterenol enhanced IKs 4.0±0.7 fold in AVB versus 7.9±0.7 fold at control (P<0.05; Fig. 3). The kinetics of IKs were affected similarly by β-AR stimulation in AVB and control cells (Table 2). For voltage steps from −50 mV to 20 mV, halftimes for activation decreased by approximately 100 ms, whereas halftimes for deactivation increased by about 50 ms.

Fig. 3

Decreased responsiveness of IKs to β-A stimulation in chronic-AVB (CAVB) cells. Upper panel, representative examples of IKs at baseline and when increased by β-A stimulation (isoproterenol, 100 nmol/L) in a control (SR) and a CAVB cell. Relative data shown are based on IKs-tail amplitudes. Vertical bar, 5 pA/pF. Horizontal bar, 2 s. Lower panel, dose dependence of IKs stimulation by isoproterenol at control and CAVB. Inset, voltage-clamp protocol. *, P<0.05.

3.3. Consequences of IKs downregulation for repolarization response to isoproterenol

Isoproterenol (100 nmol/L) shortened action-potential duration (APD95) significantly in control cells (ncells=6; ndogs=3; Fig. 4A). This was mainly due to the stimulation of IKs, because block of IKs by HMR1556 (500 nmol/L) in the presence of isoproterenol reversed it [13]. Heterogeneous responses of repolarization were found in chronic AVB: APD95 remained unchanged in 7, decreased in 4 and increased in 2 myocytes (ndogs=10; Fig. 4B). Despite the limited ability of chronic-AVB cells to shorten APD in response to isoproterenol (100 nmol/L), IKs still contributed to some degree under these conditions, as evident from the (extra) APD prolongation observed after application of HMR1566 (Fig. 4A and B). In the absence of isoproterenol, HMR1556 did not change APD (ncells=5), in line with recent findings in normal dog [13] and human [14].

Fig. 4

Effects of β-A stimulation on the action potential. Action potentials/APD95 at baseline (bas), in the presence of isoproterenol (iso; 100 nmol/L) and after addition of HMR1556 (500 nmol/L) in control (A) and chronic-AVB (B) myocytes. Pacing cycle length, 1000 ms. Upper panels, representative examples. Lower panels, APD95 values (mean±SEM) for the different conditions. Vertical bar denotes 50 mV. Horizontal bar, 100 ms. *, P<0.05 versus baseline. #, P<0.05 versus iso 100.

Early afterdepolarizations (EADs) were induced in 4 of the 13 chronic-AVB cells with isoproterenol (100 nmol/L) alone, and in 6 of 13 myocytes during isoproterenol plus HMR1556 (Fig. 5). EADs developed at short (500 ms) and long (1000 and 2000 ms) pacing cycle lengths. They were not observed at baseline. In control myocytes, no EADs were observed under any of these conditions (Fig. 5).

Fig. 5

EADs in the presence of β-A stimulation in chronic-AVB (CAVB) cells. Incidence of EADs at baseline, in the presence of isoproterenol (iso; 100 nmol/L) and after addition of HMR1556 (500 nmol/L) in control (SR) and CAVB cells. Right panels, representative action potentials with EADs in the presence of isoproterenol at pacing cycle lengths (CL) of 2000 ms and 500 ms.

A cAMP-activated Cl current (ICl) could also contribute to the action-potential shortening in response to β-A stimulation. In dog ventricle however, such a current is reportedly absent [23], whereas it was also shown that anthracene increased the incidence of EADs by isoproterenol suggesting the presence of cAMP-activated ICl [24]. We performed additional experiments and measured currents during wash-in of 3 μmol/L isoproterenol in conditions that should reveal this ICl (K+- and Na+-free solutions with symmetrical [Cl]; current measured at −80 mV). Current density at baseline was −0.32±0.06 pA/pF in control (ncells=7) and −0.30±0.05 pA/pF in chronic-AVB myocytes (ncells=7, P=NS). In the presence of isoproterenol (3 μmol/L) the current was −0.36±0.06 pA/pF and −0.35±0.04 pA/pF, respectively (P=NS). In 4 cells, this current was not sensitive to the Cl channel blocker niflumic acid (100 μmol/L). These data suggest that if present, the cAMP-dependent ICl is very small and does not contribute to differences in repolarization between control and chronic AVB.

3.4. QT responses to isoproterenol in conscious AVB dogs

To investigate whether repolarization responses to β-AR stimulation were also altered in vivo, we subjected 6 dogs to short-lasting challenges with isoproterenol (1 μg/kg i.v.; >25% increase of heart rate) at SR control and during idioventricular rhythm at 3, 8 and 30 days of AVB.

The heart rate dropped from 128±12 bpm at SR to 41±4 bpm at 3 days AVB, remaining low thereafter (Fig. 6). At baseline during conscious sedation, QT time (chest lead V4) measured 253±10 ms at SR. It increased thereafter from 371±12 ms at 3 days to 414±7 ms at 30 days AVB (P<0.05 versus 3 days), indicating electrical remodeling [1].

Fig. 6

QT responses to isoproterenol in conscious AVB dogs. (A) Heart rates (HR; filled squares) and corresponding QT times (open circles) during infusion of isoproterenol (1 μg/kg, i.v.) in a representative dog during SR (upper panel), and at 3 days (middle panel) and 30 days of AVB (lower panel). The grey zone indicates the period of isoproterenol infusion. Vertical lines indicate the time points of 25% heart-rate increase (inserted for comparability). (B) Heart-rate responses to isoproterenol. White bars indicate baseline conditions during SR and after 3, 8, 30 days of AVB. Hatched bars, 25% heart-rate increase above baseline. Black bars, peak responses. (C) QTc responses to isoproterenol at baseline (white bars) and during isoproterenol infusion (hatched bars, at 25% heart-rate increase above baseline; black bars, during peak responses of heart rate). (D) Beat-to-beat variability of QT (quantified as short-term variability; STV) at baseline (open squares) and during isoproterenol (filled circles) in individual dogs (connected by lines) during SR and after 3 and 30 days of AVB. Same symbols at the side indicate mean±SEM. *, P<0.05 for isoproterenol versus baseline.

Next we investigated the actions of isoproterenol (1 μg/kg i.v.). Both in SR and idioventricular rhythm a strong chronotropic effect was noted, as expected: heart rates increased from 128±12 bpm to 201±9 bpm (SR) and from 41±4 bpm to 58±5 bpm (3 days AVB), respectively. These chronotropic responses remained proportionally similar at the later stages of AVB (Fig. 6B). QT-interval measurements were performed at 2 time points during each experiment: (1) when heart rate increased 25% above the baseline value (for comparability between experiments; hatched bars Fig. 6B and C); and (2) during peak chronotropic response (black bars Fig. 6B and C). A profound shortening of repolarization accompanied the peak heart-rate increase at SR: QT decreased from 253±10 to 211±5 ms (QT-17%; P<0.05). At 3 days AVB, QT shortened from 371±12 ms to 347±6 ms (heart rate +25%; QT-6%; P<0.05) to 324±4 ms (peak response; QT-13%; P<0.05). At 30 days AVB, however, QT changes during isoproterenol became much less pronounced despite similar increases in heart rate: QT altered from 414±7 ms to 409±11 ms (heart rate +25%; QT-1%; P=NS) to 392±10 ms (peak response; QT-5%; P<0.05). Individual QT times were often longer than their baseline values and they varied considerably on a beat-to-beat basis, indicating in-vivo repolarization instability (Fig. 6A and D). The latter was not observed at SR or acute AVB. When corrected for heart rate, QTc was shortened by isoproterenol during SR, unchanged at 3 and 8 days AVB, and overtly prolonged in chronic AVB (Fig. 6C). Beat-to-beat variability of QT (quantified as STV) decreased from 2.5±0.4 ms to 1.6±0.3 ms (P<0.05) in response to isoproterenol during SR. At 3 days AVB QT STV remained unchanged (4.0±0.4 ms to 4.8±0.7 ms; P=NS), whereas at 30 days it actually increased from 5.4±0.6 ms to 7.4±0.7 ms (P<0.05) during β-A stimulation (Fig. 6D). Ventricular tachycardia was not observed, supposedly because local myocardial β-A stimulation was not intense enough.

3.5. Plasma levels of catecholamines

To assess the influence of natural catecholamines on repolarization, plasma levels of dopamine, epinephrine and norepinephrine were measured in venous blood samples at all time points, including the recovery phase after induction of AVB (0 days). Measured levels of dopamine ranged from ‘not detectable’ to 0.32 nmol/L under control conditions (median 0.12 nmol/L) and did not alter during AVB. Likewise, epinephrine concentrations did not change significantly (Fig. 7; open squares). Norepinephrine, however, (control concentration 0.87±0.23 nmol/L) increased significantly upon induction of AVB (0 days) and peaked at 2.50±0.56 nmol/L around 8 days AVB. In the chronic phase of AVB thereafter (>8 days), norepinephrine returned to control levels (Fig. 7; filled diamonds). We speculate that the early rise of this catecholamine is a response mechanism to overcome incomplete circulatory compensation in the first 2 weeks of AVB [22].

Fig. 7

Plasma levels of natural catecholamines during AVB. Concentrations of epinephrine and norepinephrine in venous blood plasma at SR and 0, 3, 8 and 30 days of AVB. *, P<0.05 versus SR.

4. Discussion

The present study demonstrates that a profound decrease of KCNQ1 mRNA and protein occurs within days after induction of AVB in the apicolateral region of the canine LV. It is followed by a more gradual decrease of β1-ARs. The kinetics of downregulated IKs and its responsiveness to the selective blocker HMR1556 remain unaltered in AVB myocytes. However, β-AR stimulation of IKs is decreased. At the level of the cellular action potential these molecular and biophysical characteristics translate to the loss of β-A-induced shortening of repolarization, favoring the occurrence of EADs. In conscious dogs, QT and QTc responses to β-A challenges confirm the IKs-related loss of ventricular repolarization shortening in chronic AVB, when beat-to-beat QT instability is also eminent.

4.1. Downregulation of IKs and β-adrenergic regulation

Downregulation of IKs at the chronic stage of AVB has been reported previously in the dog [1,9] and rabbit [25]. Decreases in IKs were also found in other animal models of cardiac hypertrophy and/or failure [2,26,27], and in humans [28]. In the present study we closely analyzed the biophysical properties of downregulated IKs in chronic-AVB myocytes. The unaltered kinetics and responsiveness to HMR1556 indicated that reductions in channel numbers, rather than conformational changes, underlie the diminution of this current. Another aspect of IKs downregulation was the finding, that, in contrast to the reduction of KCNQ1, KCNE1 did not change. We observed a similar differential response in the canine interventricular septum after chronic AVB [10]. However, reductions of both KCNQ1 and KCNE1 mRNA and protein were found in the basal and midlateral parts of the LV and right ventricle [9]. In the latter study, mRNA expression was quantified using a combination of Northern blot and multiplex competitive real-time PCR, whereas in the recent septum study [10] and the present one fluorescence-based kinetic real-time PCR was employed. The latter is a more accurate method. To exclude a methodological reason for the apparent discrepancies in KCNE1 results, we redetermined KCNE1 mRNA in tissue from the basal and midlateral LV of normal and chronic-AVB dogs (as in Ref. [9]) using the fluorescence-based kinetic real-time PCR (as in the present study), and found a similar decrease of expression as described previously [9]. Based on these data, we consider regional heterogeneity of IKs-subunit remodeling in the LV, even among α- and β-channel subunits, the most likely explanation for the different results in the three studies.

The question arises why the decrease of KCNQ1 did not change the kinetics of IKs. Wang et al. [29] examined the influence of various ratios of KCNQ1: KCNE1 on the current kinetics of the assembled IKs channel. Ratios of 2:1 and 1:1 resulted in similar current characteristics, and both closely resembled the activation kinetics and voltage dependence of native IKs [29]. Thus, a unilateral decrease of KCNQ1 of ∼50% (as in AVB) would not be expected to affect the kinetics of native IKs significantly, in line with the present findings.

Stimulation of IKs by β1-ARs depends predominantly (if not exclusively) on protein-kinase-A stimulation of KCNQ1 subunits [30]. Modulation via β2-ARs has also been suggested, mainly by the finding that cardiac IKs density is increased in transgenic mice expressing fusion proteins of IKs subunits and human β2-ARs [31]. β2-A signaling could modulate the function of these IKs channels even in the absence of exogenous β-A agonists [31]. β3-A regulation of IKs has also been reported, although with different results in different models: β3-AR stimulation enhanced IKs in Xenopus oocytes expressing human KvLQT1/MinK channels [32], but inhibited the current in guinea-pig ventricular myocytes [33]. From the three receptor subtypes (β1, β2 and β3) examined in the present study, only β1 expression showed a clear and sustained decrease compatible with the decreased β-A responsiveness of IKs in AVB cells. We cannot exclude that, apart from the decreases of KCNQ1 and β1-AR, other elements of the IKs-macromolecular signaling complex [34] or β-A cascade are also involved in the downregulation of IKs.

The effect of β-AR downregulation is probably not the same for all ion channels. We found that the maximal response of ICaL during depolarizing steps to 0 mV was the same in control and chronic AVB (manuscript under revision). Differences in β-A responsiveness between ICaL and IKs could be explained by different coupling of these channels to their receptors and/or signaling cascades. Isoproterenol stimulation of L-type Ca2+ channels results mainly from a local activation of the signaling cascade [35]. At the present time there is no such data available for IKs channels.

4.2. Temporal patterns of remodeling at the molecular and in-vivo level: do they match?

Using a minimally-invasive procedure under closed-chest conditions, we obtained LV myocardial biopsies in a serial manner. Thus, follow-up of the transcriptional expression of IKs-channel subunits and other IKs-relevant proteins (including β-AR subtypes) was possible using each dog as its own control, thereby minimizing problems of biological variability. Such problems are often encountered in group comparisons when different animals are used. To the best of our knowledge, this is the first study to describe serial molecular-expression patterns in LV myocardial tissue from individual dogs. We believe that this approach holds great promise for future studies of myocardial remodeling in chronic pathological conditions.

When we align the patterns of KCNQ1 and β1-A downregulation with the altering QT responses in vivo and the changes of plasma norepinephrine, an interesting picture emerges. Considering that IKs downregulation is a major determinant of the AVB-induced repolarization prolongation [1], the fast decrease of KCNQ1 (at 3 days; Fig. 1) apparently contrasts with the much slower response of QT and QTc prolongation by remodeling (after >8 days of AVB; Fig. 6). However, the contribution of IKs to repolarization depends critically on β-AR stimulation [13] and therefore this system must be taken into account when explaining the temporal differences. We find it likely that the early rise of norepinephrine (Fig. 7) ‘neutralizes’ the loss of KCNQ1, effectively maintaining a pseudonormal IKs in the first week of AVB. In the period thereafter, the waning of norepinephrine levels and the additional downregulation of β1-ARs (Fig. 1) gradually uncovers the QT prolongation. In a similar time-dependent manner, the loss of β-A-induced shortening of ventricular repolarization could be explained.

The lack of an effect of isoproterenol on the QTc interval at 3 and 7 days AVB (Fig. 6C) could arise from the influence of high levels of endogenous norepinephrine in the ventricles at this stage. This could have obscured the effect of the exogenous β-A agonist.

Different mechanisms control the chronotropic responses to isoproterenol during SR and AVB, i.e., via stimulation of the sinus node and via stimulation of distal His-bundle cells, Purkinje cells or even ventricular working myocardium, respectively. The sensitivity of these pacemaking tissues to isoproterenol may not be the same in the two conditions.

4.3. Consequences of IKs downregulation

In control myocytes, β-A activation shortened the APD consistently, which indicated that the stimulation of outward currents (mainly IKs) prevailed over the stimulation of inward currents (mainly ICaL). In chronic-AVB cells, heterogeneous APD responses were observed, likely as the result of IKs downregulation in the presence of an unchanged ICaL and enhanced INaCaX [7]. The reduced IKs still contributed to repolarization during β-AR stimulation, as evident from the (extra) APD prolongation after administration of HMR1566 (Fig. 4). Under these conditions (isoproterenol plus HMR1566) the frequent generation of EADs in chronic-AVB cells, and their absence in control myocytes (Fig. 5), strongly suggested that other ionic changes (e.g., enhanced Na+–Ca2+ exchange [7]) also contributed to the proarrhythmia. Taken together our results indicate that the already labile ventricular repolarization in chronic AVB is adversely challenged during episodes of intense β-A activation. The generation of β-A EADs would exaggerate this repolarization lability and also trigger abnormal impulses.

Approximately 10% of the dogs with chronic AVB die suddenly. In some of these, sudden death occurred during excitement, as witnessed by animal technicians. In other animals, ambulatory electrocardiographic recordings revealed polymorphic ventricular tachycardia as the mode of death [6]. Tachycardia was immediately preceded by an accelerated idioventricular rhythm and short-coupled ventricular beats from different foci [6]. The data suggest a significant contribution of the sympathetic nervous system to the initiation of these arrhythmic events. Given the findings of KCNQ1 downregulation and β-A proarrhythmia in this dog model, there is an interesting similarity with the human congenital long-QT 1 syndrome. Sympathetic triggers of torsades de pointes, as during exercise and emotion, have long been recognized in long-QT-1 patients [36]. In latent carriers of KCNQ1 mutations, intravenous epinephrine can unmask the abnormal repolarization [37], much like our findings with isoproterenol in dogs with AVB. To our knowledge, the present study is the first to describe provocation testing using isoproterenol in acquired cardiac channelopathy.

Acknowledgments

This study was supported by The Netherlands Organization for Health and Development, The Hague (ZonMw 906-02-068 to P.G.A.V.), The Netherlands Heart Foundation, The Hague (NHS 98.042 to C.R.), the ‘Stichting Hartsvrienden Rescar’, Maastricht, The Netherlands (to P.G.A.V.), and the United States National Heart, Lung, and Blood Institute (USPHS-NHLBI grant HL-67101 to A.V.R.). Monique M.J. de Jong, Departments of Cardiology and CardioThoracic Surgery, Maastricht, and Jet D.M. Beekman, Department of Medical Physiology, University Medical Center Utrecht, The Netherlands provided excellent technical assistance. Michel E.F.H. Pereboom, Department of Clinical Chemistry, Academic Hospital Maastricht, performed the catecholamine analyses. The primers for canine KCNQ1 and KCNE1 genes were kindly provided by Robert Dumaine, PhD, Department of Physiology and Biophysics, Faculty of Medicine and Health Sciences, University of Sherbrooke, QC, Canada. The authors are grateful to Jos G. Maessen, MD, PhD, Department of CardioThoracic Surgery, Cardiovascular Research Institute Maastricht and Academic Hospital Maastricht for his role in establishing the serial-biopsy technique.

Footnotes

  • Time for primary review 12 days

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
View Abstract