Role of KATP Channels in the Maintenance of Ventricular Fibrillation in Cardiomyopathic Human HeartsNovelty and Significance
Rationale: Ventricular fibrillation (VF) leads to global ischemia. The modulation of ischemia-dependent pathways may alter the electrophysiological evolution of VF.
Objective: We addressed the hypotheses that there is regional disease-related expression of KATP channels in human cardiomyopathic hearts and that KATP channel blockade promotes spontaneous VF termination by attenuating spatiotemporal dispersion of refractoriness.
Methods and Results: In a human Langendorff model, electric mapping of 6 control and 9 treatment (10 μmol/L glibenclamide) isolated cardiomyopathic hearts was performed. Spontaneous defibrillation was studied and mean VF cycle length was compared regionally at VF onset and after 180 seconds between control and treatment groups. KATP subunit gene expression was compared between LV endocardium versus epicardium in myopathic hearts. Spontaneous VF termination occurred in 1 of 6 control hearts and 7 of 8 glibenclamide-treated hearts (P=0.026). After 180 seconds of ischemia, a transmural dispersion in VF cycle length was observed between epicardium and endocardium (P=0.001), which was attenuated by glibenclamide. There was greater gene expression of all KATP subunit on the endocardium compared with the epicardium (P<0.02). In an ischemic rat heart model, transmural dispersion of refractoriness (ΔERPTransmural=ERPEpicardium−ERPEndocardium) was verified with pacing protocols. ΔERPTransmural in control was 5±2 ms and increased to 36±5 ms with ischemia. This effect was greatly attenuated by glibenclamide (ΔERPTransmural for glibenclamide+ischemia=4.9±4 ms, P=0.019 versus control ischemia).
Conclusions: KATP channel subunit gene expression is heterogeneously altered in the cardiomyopathic human heart. Blockade of KATP channels promotes spontaneous defibrillation in cardiomyopathic human hearts by attenuating the ischemia-dependent spatiotemporal heterogeneity of refractoriness during early VF.
Ventricular fibrillation (VF), once established, rarely self-terminates. The factors that tend to maintain reentry in fibrillating myocardium include increases in spatial dispersion of refractoriness and time-dependent alteration of refractory periods, which may occur in a spatially heterogeneous manner.1–4 VF leads to global ischemia and resultant activation of cardiac KATP channels, which shorten action potential duration and refractoriness.5 If the expression and/or function of KATP channels is heterogeneous, VF-induced ischemic KATP channel activation may also result in increasing spatiotemporal dispersion of refractoriness, providing conditions that are conducive for sustaining VF after its initiation. Thus, KATP channel blockade may prevent shortening of refractoriness and cause attenuation of spatiotemporal dispersion of refractoriness during early VF, thereby forestalling VF perpetuation. In some conditions, such as ischemic cardiomyopathy, it has been suggested that the expression of KATP subunits is altered in infarct border zone.6 We have previously presented preliminary data suggesting that that there is altered expression of KATP channel subunits in human cardiomyopathic hearts compared with normal hearts.7 It is not known if there is spatial heterogeneity and/or altered expression of KATP channels in human cardiomyopathic hearts and if modulation of these channels in cardiomyopathic states alters the maintenance of VF.
We tested the hypotheses that (1) there is differential regional/disease-related expression of KATP channels in human cardiomyopathic hearts, causing spatiotemporal dispersion of refractoriness, and that (2) blockade of KATP channels by glibenclamide promotes spontaneous VF termination by attenuating this spatiotemporal dispersion.
All procedures conformed to the Helsinki Declaration of the World Medical Association. Only essential information is presented in this article in summary form. Please see the Online Data Supplement at http://circres.ahajournals.org for complete methodological details.
Human Langendorff Model
After informed consent, hearts were explanted from 15 patients with dilated cardiomyopathy who underwent cardiac transplantation and were Langendorff-perfused.
Electric mapping of the Langendorff-perfused hearts was performed with an epicardial sock covering the entire epicardial surface and a left ventricular (LV) endocardial balloon (Figure 1B and 1C). VF was recorded in ischemic condition for 20 seconds at the onset and after 180 seconds, in control and treatment groups (10 μmol/L glibenclamide) (Figure 1A). We compared spontaneous defibrillation and effective refractory period (ERP) dispersion between control and treatment groups.
The abrupt termination of VF without electric defibrillation during an episode of induced VF (lasting spontaneously >30 seconds) was considered to represent spontaneous defibrillation.
Measurement of ERP During VF
Since the traditional method of ERP measurement cannot be used at multiple sites during VF because of the dynamic nature of VF,8 we used average local fibrillation intervals as a surrogate for ERP.9–14 To further validate the mechanisms found in human hearts, we studied the consequences of ischemia on Langendorff-perfused rat hearts with direct ERP and action potential measurements (see below).
Spatial Dispersion of Refractoriness
Dispersion of refractoriness was calculated between LV epicardium and endocardium (LV transmural dispersion), LV and right ventricular (RV) (interventricular dispersion), anterior and posterior LV (anterio-posterior dispersion) and apex and base (apico-basal dispersion).
Human Heart Samples
Human heart samples were dissected from normal and cardiomyopathic hearts for RNA and protein expression.
RNA Preparation and Quantitative PCR
RNA was isolated from endocardial and epicardial samples for quantitative PCR. β-Actin served as the control.
Langendorff-Perfused Rat Hearts
This protocol was approved by University Health Network Ethics Board. Male Sprague-Dawley rats were anesthetized. Hearts were rapidly isolated and Langendorff-perfused.
The heart was placed on a plastic base in which 4 contact electrodes were inserted (Figure 2A). Two electrodes were used to record LV epicardial signals and 2 were used for pacing LV epicardium. For endocardial mapping and pacing, 4 electrodes were inserted into the LV cavity, with 2 electrodes used for endocardial recording and 2 for endocardial pacing (Figure 2B).
ERP Dispersion Determination Using S1-S2 Protocol
Figure 2C shows the experimental protocol. Hearts were paced at 150 ms (S1) at twice diastolic threshold, and an extrastimulus (S2) was delivered, starting from S1-S2=140 ms, decreasing until capture was lost. ERP dispersion and arrhythmogenesis was measured in control group (no treatment) and after glibenclamide, pinacidil, and lidocaine.
Floating Microelectrode Recording
Recordings were obtained from epicardium and endocardium under control conditions and during ischemia with and without drug treatment.
Data were analyzed using SAS 9.0. Values are shown as mean±SEM. The Fisher exact test and ANOVA were used, and a probability value <0.05 was considered statistically significant.
Fifteen human hearts (6 control and 9 treatment) were studied. Spontaneous VF termination occurred in only 1 of 6 control hearts. In the 9 treatment hearts, spontaneous termination occurred in 2 hearts before glibenclamide administration. Figure 3A shows an example of unipolar electrograms during VF and ischemia. After treatment with glibenclamide (n=9), VF could not be induced in 1 heart and terminated spontaneously in 7 of the remaining 8 hearts (P=0.026 versus control). Figure 3B shows an example of spontaneous termination of VF after the administration of glibenclamide.
KATP Channel Blockade and Spatiotemporal Dispersion of Refractoriness During Fibrillation in Human Langendorff Model
Initially, mean VF cycle length (CL) was compared, averaging VF at onset and 180 seconds, between epicardium and endocardium in the control and glibenclamide treatment groups. Mean VFCL was 248±8 ms and 289±27 ms in control and 397±28 ms and 408±27 ms in the treatment group (P≤0.001 versus control), in LV endocardium and epicardium, respectively. The VFCL was then compared between epicardium and endocardium in onset and early VF (180 seconds). At the onset of VF, the mean VFCL in LV endocardium and epicardium was 230±13 ms and 242±25 ms in the control group and 343±29 ms and 353±32 ms in the treatment group (Figure 4A). The difference between control group and the treatment group was statistically significant (P=0.017), whereas the difference between epicardium and endocardium was not significant. After 180 seconds of VF, as ischemia progressed, a gradient in VFCL developed in the control group, with a mean VFCL of 279±17 ms and 337±36 ms in LV endocardium and epicardium, respectively (P=0.01). In the treatment group, the difference was attenuated, with a mean VFCL of 454±34 ms and 463±27 ms, respectively, in endocardium and epicardium (P=0.6), suggesting that KATP channels play a key role in establishing this gradient. Figure 3C illustrates refractoriness for the LV endocardium and epicardium during the VF sequences in 1 heart.
Mean VFCL was 297±29 ms and 274±22 ms in the control group and 408±28 ms and 496±24 ms in the treatment group for LV epicardium and RV epicardium, respectively. The difference was statistically significant between control and treatment (P<0.001) but not significant between LV and RV epicardium (P=0.69). VFCL was then compared in onset and early VF (180 seconds). At the onset of VF, mean VFCL in LV and RV epicardium was 248±28 ms and 237±16 ms in control group and 353±33 ms and 363±9 ms in the treatment group. The difference between control and treatment groups was statistically significant (P=0.02), whereas the difference between LV epicardium and RV epicardium was not significant (P=0.9). During early VF, a small difference of 32 ms was observed between LV and RV epicardium (LV epicardium=345±37 ms, RV epicardium=313±34 ms), but this failed to achieve statistical significance (P=0.07). In the treatment group, the difference was attenuated, with a mean VFCL of 463±27 ms and 450±26 ms in LV and RV epicardium, respectively (P=0.23).
At the onset of VF, the mean VFCL was 231±30 ms and 239±30 ms in the control group (P=0.9) and 345±26 ms and 350±25 ms in the treatment group (P=0.9) in anterior and posterior LV epicardium, respectively. During early VF (180 seconds), mean VFCL in the control group was 300±33 ms and 333±33 ms in anterior and posterior LV epicardium, respectively. The difference of 33 ms between anterior and posterior LV epicardium after 180 seconds of ischemia failed to achieve statistical significance (P=0.5). After the administration of glibenclamide in the treatment group, this difference was attenuated, with a mean VFCL of 453±28 ms and 455±28 ms in the anterior and poster LV epicardium, respectively (P=0.9).
Mean VFCL at the apex and the base of the heart was 258±30 ms and 287±29 ms in the control group and 373±25 ms and 423±25 ms in the treatment group, respectively. In both regions, the difference was statistically significant between control and treatment groups (P<0.001), but the difference between apex and base was not statistically significant (P=0.15). During onset VF, the mean VFCL was 217±15 ms and 264±23 ms in the control group (P=0.9) and 344±29 ms and 369±36 ms in the treatment group (P=0.2) at the apex and base, respectively. During early VF (180 seconds), VFCL was 306±40 ms at the apex and 347±31 ms at the base in the control group (P=0.8) and 445±36 ms at the apex and 471±25 ms at the base in the treatment group (P=0.9). Though a small difference in VFCL was observed between apex and base, it was consistent in the onset, and early VF and was not affected by glibenclamide.
KATP Channel Subunit Regional mRNA Expression in Cardiomyopathic Human Hearts
We quantified the gene expression of the 4 common subunits that make up the human cardiac KATP channels, Kir6.1, Kir6.2, SUR1, and SUR2A, in epicardium and endocardium in 6 myopathic hearts.
KATP Subunit Expression in LV Endocardium Versus Epicardium
KATP subunit expression was done to evaluate if the LV transmural gradient of refractoriness was related to the differential expression of KATP subunits in epicardium versus endocardium in cardiomyopathic hearts. All expression levels are presented relative to the housekeeping gene and normalized to LV endocardium. Mean mRNA expression levels in LV epicardium were 61±10% (P=0.02), 59±13% (P=0.03), 57±11% (P=0.016), and 72±5% (P=0.004) for Kir6.1, Kir6.2, SUR1, and SUR2A, respectively (Figure 4B and 4C). All subunits were less strongly expressed in LV epicardium compared with LV endocardium.
KATP Channel Expression in LV Epicardium Versus RV Epicardium
We compared KATP gene expression in LV epicardium and RV epicardium. Mean expression levels (relative to housekeeping gene and relative to LV endocardium) were 62±10%, 59±13%, 57±11%, and 72±5% for Kir6.1, Kir6.2, SUR1, and SUR2A, respectively, in the LV epicardium, and 80±38% (P=0.68), 106±47% (P=0.44), 80±32% (P=0.60), and 128±69% (P=0.47) in the RV epicardium.
Langendorff-Perfused Rat Model
ERP Dispersion During KATP Modulation
Dispersion in refractoriness (ΔERP) was defined as the difference in the refractoriness between epicardium and endocardium (Figure 2D). Twenty-eight hearts were studied with a traditional S1-S2 protocol. Mean ΔERP was 5±2 ms and 36±5 ms during baseline (without ischemia) and ischemia (P=0.02 versus baseline), respectively. After 10 μmol/L glibenclamide (n=13), mean ΔERP was 0.1±3.0 ms and 4.9±4.0 ms during baseline (no ischemia) and ischemia with glibenclamide (P=0.9 versus glibenclamide alone), respectively. Hence, glibenclamide significantly attenuated the ΔERP during acute ischemia (P=0.019). In 7 hearts, ERP in control conditions was followed by pinacidil. Mean ΔERP dispersion in the presence of pinacidil was 2.7±2.0 ms in the baseline condition and 28±13 ms during acute ischemia. There was no statistically significant difference in ΔERP after pinacidil, whether in the baseline condition or acute ischemia.
Arrhythmogenesis During KATP Modulation
During control conditions, cardiac arrhythmia occurred in 15 of 28 hearts in the baseline condition and 19 of 28 hearts during acute ischemia. The difference was not statistically significant (P=0.4). In 13 hearts, measurements in control conditions were followed by 10 μmol/L glibenclamide. After glibenclamide, cardiac arrhythmias occurred in 10 of 13 hearts in the baseline condition (no ischemia) and 4 of 13 hearts during acute ischemia (P=0.047). When compared with control, glibenclamide significantly suppressed the incidence of cardiac arrhythmias during acute ischemia (4/13 versus 19/28, P=0.043). In 7 hearts, measurements during control conditions were followed by 10 μmol/L pinacidil. After pinacidil, cardiac arrhythmias occurred in 5 of 7 hearts in the baseline condition (no ischemia) and 7 of 7 hearts during acute ischemia (P=0.5). When compared with control, there was no difference in the incidence of cardiac arrhythmias after pinacidil during acute ischemia (P=0.2).
Action Potential Changes During KATP Modulation
To study the effect of KATP modulation on action potential duration, membrane potentials were measured under control conditions and after KATP modulation (Figure 2E). Action potential durations at 90% of repolarization (APD90) were analyzed. In the baseline control condition (no ischemia, no drug), mean APD90 was 56±3 and 55±2 ms in endocardium and epicardium, respectively (n=17, P=NS). Glibenclamide did not cause any significant changes in APD in the baseline condition (n=7, APD90 epicardium: 58±5 ms, endocardium: 53±3 ms; P=NS). Pinacidil resulted in APD shortening, mainly in epicardium (n=5; mean APD90 epicardium: 41±8 ms; pinacidil versus baseline; P=0.049).
After 3 minutes of ischemia, APD90 was significantly shorter on epicardium compared with endocardium (epicardium: 55±6 ms; endocardium: 82±5 ms; P=0.02). After the administration of glibenclamide, the difference in APD90 between the epicardium and endocardium was attenuated and was not statistically significant during acute ischemia (APD90 at 3 minutes of acute ischemia; endocardium: 83±5 ms, epicardium: 83±12 ms; P=NS).
Lidocaine Effects on ERP and APD
In 8 hearts, control ERP dispersion measurements were followed by administration of 10 μmol/L lidocaine. ERP dispersion measurements were repeated in the presence of lidocaine under baseline conditions as well as during ischemia. The mean ERP dispersion was 24±6 ms in the baseline condition (no ischemia). During ischemia, the mean ERP dispersion was 91±21 ms. The difference of dispersion between baseline and acute ischemia after lidocaine was statically significant (P=0.006). When compared with control, lidocaine resulted in increased transmural gradient of ERP during ischemia (P=0.02). After lidocaine, arrhythmias resulted in 50% hearts (4/8), not statistically significant from control (P=0.2). Lidocaine had no significant effect on APD90 either in epicardium or endocardium (n=5; mean APD90 epicardium: 50±3 ms; endocardium: 57±2 ms). Similarly, lidocaine had no significant effect on APD in acute ischemia (eg, mean APD90 after 3 minutes of ischemia: epicardium 43±13 ms, endocardium 70±32 ms; P=NS versus no-drug control).
We have demonstrated that KATP channel blockade during VF promotes spontaneous defibrillation of cardiomyopathic human hearts by attenuating the ischemia-dependent spatiotemporal dispersion of refractoriness during early VF. The molecular basis probably involves heterogeneous and altered expression of KATP channel subunits in cardiomyopathic hearts. After 180 seconds of VF, there is heterogeneity in refractoriness which is attenuated by blockade of KATP channels. Taken together, our findings indicate that glibenclamide promotes spontaneous defibrillation of VF in the setting of global ischemia by the attenuation of spatiotemporal heterogeneity and increasing refractoriness during VF in cardiomyopathic hearts.
Spontaneous defibrillation (termination of VF without high-voltage shock) is an uncommon event that has occasionally been reported in certain animal species15 and in humans,16–18 but rarely in a cardiomyopathic state. The development of therapeutic strategies that would promote spontaneous defibrillation has significant clinical implications for prevention of sudden cardiac death in cardiomyopathic patients at the highest risk of VF. It is well described that shortening of the APD plays an important role in maintaining cardiac arrhythmias.19,20 There are various reports that glibenclamide can either prevent21 or partially reverse22 the APD shortening in hypoxia and ischemia21,22 in both isolated22 and in situ guinea pig myocardium.21 Our observations in the current study suggest that during VF, glibenclamide prevents APD shortening caused by acute hypoxia, ischemia, and KATP-activation, leading to attenuation of spatiotemporal dispersion of refractoriness, thus causing spontaneous termination. A gradient in local activation rate/refractoriness existed between the LV endocardium and epicardium, probably due to differential KATP channel gene expression. Glibenclamide, through KATP channel blockade, resulted in a loss of this gradient. These observations suggest that spatiotemporal action potential characteristics within the LV contribute to the maintenance of VF in cardiomyopathic human hearts and are in part controlled by regionally differential consequences of KATP activation, the effects of which are attenuated by KATP channel blockade.
A dominant-frequency gradient has been reported during VF in various animal23–25 and human26 models. We hypothesized that differential function of KATP channels might contribute to this gradient. The fact that blockade of KATP channels by glibenclamide abolished this gradient supports this hypothesis. We further tested the hypothesis by studying the differential expression of KATP channel subunits in LV epicardium and endocardium in cardiomyopathic hearts. The mRNA expression levels of KATP channel subunits were significantly higher in LV endocardium compared with LV epicardium. Miyoshi et al have reported that blockade of cardiac KATP channels by glibenclamide in dogs suppresses the extracellular K+ rise in epicardium, whereas the extracellular K+ level in endocardium remains unaffected.27 We attribute this difference to species specific differential functioning of KATP channels in the heart. It should also be noted that the hearts that we studied were cardiomyopathic hearts explanted from patients undergoing cardiac transplantation.
KATP Modulation in Experimental VF
In an in vivo ischemic porcine model, the KATP channel opener pinacidil significantly shortened the ERP from 162±16 to 130±28 ms and increased the peak frequency of the LV power spectrum during VF from 9.3±0.6 to 10.5±1.0 Hz.28 Our data are consistent with the concept proposed regarding KATP modulation during VF and also with the observed relation between refractory period and cycle length. Because measuring dynamic ERP changes during VF in human hearts was impossible, we used frequency indices as indicators of ERP change and conducted confirmatory experiments in an animal model to measure dispersion of refractoriness (ΔERP) with traditional S1-S2 methods. In the absence of ischemia, there was no difference in ΔERP between control and glibenclamide groups (5±2 versus 0.1±3.0 ms), but during ischemia there was a large, statistically significant difference (36±5 ms versus 4.9±4.0 ms). These results implicate KATP activation and consequent increased ERP heterogeneity in ischemic conditions that could potentially serve as a substrate for sustaining VF. The fact that ventricular arrhythmias were ameliorated by glibenclamide supports this concept.
KATP Subunit Expression
In rodent ventricular cardiomyocytes, Kir6.2/SUR2 make up the dominant KATP channel expression. There is, however, heterogeneity within rodent hearts, with atrial cardiomyocytes expressing Kir6.2/SUR1.29 Although the classic understanding is that the KATP channel in cardiomyocytes is composed mainly of Kir6.2 and SUR2A,30–32 more recent findings suggest that Kir6.1 and SUR1 subunits may also play significant roles.33 Our study shows that in cardiomyopathic human left ventricles, gene expression of Kir6.1, Kir6.2, SUR1, and SUR2A is heterogeneous with greater preponderance on the endocardium. Kir6 subunit upregulation may contribute to the substantial time-dependent increases in ERP-heterogeneity during VF in cardiomyopathic hearts. Tavares et al showed similar Kir6.1-upregulation in a rodent cardiomyopathic model.6 Hence, the relative expression of KATP channels may be different not only in different tissues but also in different species and in different disease states. Establishing effective clinical treatments with KATP blockade requires further evaluation of the mechanism and physiological significance of differential KATP subunit expression/function with heart disease. The present study provides a basis for such further work by identifying changes in KATP subunit gene expression in cardiomyopathic human hearts.
A difference in refractoriness was observed between cardiomyopathic LV epicardium and endocardium that was attenuated by glibenclamide. In our data set, it was accompanied by increased gene expression of all KATP subunits in LV endocardium compared with epicardium. There are other potential mechanisms of transmural KATP function gradients in addition to transmural protein expression gradients in underlying subunits. There could be another, unidentified subunit associated with Kir6/SUR subunits that importantly modifies its function and shows differential endocardial/epicardial expression. Alternatively, differences in myocardial energetics and oxygen demand could produce discrepant KATP activation in LV endocardium versus epicardium. Lee et al have reported that glibenclamide inhibits Na+-K+ pump and Ca2+ channels in guinea pig hearts.34 These effects were observed at much higher concentrations (≥100 μmol/L) than used in our study. Still, the possibility of involvement of off-target actions on other ion channels cannot be ruled out. The differential activation rate seen in our study could have another potential explanation. Conduction block of wave fronts should be considered as an explanation in addition to local refractoriness changes. Morley et al previously showed in a single-cell preparation that metabolic inhibition has a significant influence on intercellular membrane resistance and may promote conduction block.35 However, in our rat heart study, we paced locally in a tissue preparation and were able to establish that indeed refractoriness was locally altered.
The only effective therapeutic strategy by which VF can be terminated once it has occurred is electric defibrillation. For a defibrillation shock to succeed, it must extinguish existing VF activations throughout a critical mass of myocardium, as well as not initiate new fibrillatory wave fronts that propagate into regions containing excitable gaps. Therefore, the present clinical strategy involves the use of implantable defibrillator devices in cardiomyopathic patients. Various chronic pharmacological strategies have been tested for the prevention or termination of VF with limited success, primarily due to proarrhythmic effects.36–40 The strategy that we tested presents the intriguing possibility of extinguishing VF activations throughout the myocardium by prolonging refractoriness and not initiating new fibrillatory wave fronts due to voltage gradients set up by high-voltage shocks. In addition, by virtue of having therapeutic electrophysiological properties mainly during ischemia, this strategy has the potential of maximal effect during the global ischemia that accompanies VF. Recently, a therapeutic strategy using spatially focalized intervention to modulate Na+ channel expression in the border zone of infarcted hearts has been proposed.41 Our findings suggest a role for modulation of global ischemia-dependent pathways in a nonelectrical focal temporal defibrillation strategy by blockade of ischemia-activated KATP channels in diseased human hearts.
KATP channel blockade may have a potential role in both the primary and secondary prevention of VF arrest. We used 10 μmol/L glibenclamide, which is a higher concentration than used in reperfusion injury studies. If used clinically, this could cause hypoglycemia in normoglycemic patients. However, the use of cardioselective KATP channel blockers such as HMR188342–44 could obviate this problem in developing this treatment paradigm. In addition, cardiac KATP channel blockers could prove to be useful adjuncts when defibrillation fails or proves difficult after cardiac surgery or for resistant VF in the cardiopulmonary resuscitation setting.
This study was performed in a Langendorff-perfused human heart model. It is not possible with this model to study the effect of interplay of different neural and autonomic mechanisms on VF that can occur in vivo; hence, care should be maintained when extrapolating the results of this study to the clinical setting. We studied the effect of KATP channel modulation in acute ischemia in isolated cardiomyopathic human hearts. Because the relative mRNA expression of KATP subunits is also different in normal and cardiomyopathic hearts, the effect of modulation of KATP channels in these scenarios needs further investigation.
It has been reported that blockade of KATP channels might counteract ischemic preconditioning and may aggravate acute ischemic injury. The emphasis of our study was on the role of KATP channel blockade in VF during acute ischemia, and we did not study the effect of KATP channel modulation on ischemic injury and preconditioning. It may, however, be possible to retain an antiarrhythmic effect of KATP channel blockade without adverse consequences for cardioprotection by selectively targeting sarcolemmal KATP channels, which mediate action potential changes while leaving intact mitochondrial KATP channels, which are central to preconditioning.
In our human Langendorff model, we used VFCL as a surrogate for ERP to obtain an ERP index simultaneously in multiple cardiac regions and to study the rapidly developing temporal evolution of refractoriness. Though VFCL may overestimate absolute refractoriness, previous studies suggest that there is considerable agreement between pacing-based determination of refractoriness and CL in fibrillation.45,46 Moreover, we confirmed our principal findings with direct ERP and APD measurements in an isolated rat heart model.
In rats, transmural dispersion of ERP and APD increased with ischemia in a glibenclamide-suppressible way, as in humans; however, we observed greater ischemic ERP shortening in epicardium versus endocardium in rats, different from humans. This may be due to a species difference, because in 2 human hearts in which we were able to study monophasic action potentials during ischemia, endocardial APD decreased more than epicardial with ischemia, and the difference was attenuated with 10 μmol/L glibenclamide (Online Data Supplement), fully consistent with our VFCL results.
Western blots for protein expression of KATP channel subunits were attempted. Antibodies available for KATP subunits were imperfect. We were unable to obtain data for SUR2A, and the staining characteristics of the available SUR1 and SUR2B antibodies were suboptimal. In addition, molecular masses of bands detected by the SUR antibodies in human tissue were different from those in heterologous systems. Finally, GAPDH signals in our SUR2B control samples were degraded, and we therefore had to use Ponceau staining as a loading control. We have provided complete original blots in the Online Data Supplement (Online Figures I through Figure IV), and our results must be interpreted in the light of these limitations.
In this study, we have demonstrated that in cardiomyopathic human hearts, heterogeneous regulation of KATP subunit gene expression is associated with increased spatiotemporal dispersion of refractoriness during VF. Glibenclamide promotes spontaneous VF termination by attenuating spatiotemporal dispersion of refractoriness that develops with ischemia as a consequence of VF. These findings suggest a role for modulation of global ischemia-dependent pathways in nonelectrical defibrillation strategies for VF.
This research was funded by grants from CIHR (MOP777687 to Dr Nanthakumar; MOP68929 to Dr Nattel). We thank Dr W. A. Coetzee for his expert advice and for kindly providing cell lysates from HEK cells stably expressing KATP subunits.
In August 2011, the average time from submission to first decision for all original research papers submitted to Circulation Research was 16 days.
Non-standard Abbreviations and Acronyms
- action potential duration
- cycle length
- effective refractory period
- ATP-sensitive potassium channels
- left ventricular
- right ventricular
- ventricular fibrillation
- Received September 20, 2010.
- Revision received September 25, 2011.
- Accepted September 27, 2010.
- © 2011 American Heart Association, Inc.
- Baker LC,
- London B,
- Choi BR,
- Koren G,
- Salama G
- el-Sherif N,
- Gough WB,
- Restivo M
- Wit AL,
- Allessie MA,
- Bonke FI,
- Lammers W,
- Smeets J,
- Fenoglio JJ Jr.
- Downar E,
- Janse MJ,
- Durrer D
- Farid TA,
- Nair K,
- Bingolis O,
- Masse S,
- Sivagangabalan G,
- Umapathy K,
- Asta J,
- Varro A,
- Nattel S,
- Demolombe S,
- Nanthakumar K
- Toal SC,
- Farid TA,
- Selvaraj R,
- Chauhan VS,
- Masse S,
- Ivanov J,
- Harris L,
- Downar E,
- Franz MR,
- Nanthakumar K
- Misier AR,
- Opthof T,
- van Hemel NM,
- Vermeulen JT,
- de Bakker JM,
- Defauw JJ,
- van Capelle FJ,
- Janse MJ
- Lammers W,
- Allessie M,
- Rensma P,
- Schalij M
- Opthof T,
- Misier AR,
- Coronel R,
- Vermeulen JT,
- Verberne HJ,
- Frank RG,
- Moulijn AC,
- van Capelle FJ,
- Janse MJ
- Opthof T,
- Dekker L,
- Coronel R,
- Vermeulen J,
- Van Capelle F,
- Janse M
- Gavrilescu S,
- Cotoi S
- Chen CC,
- Lin YC,
- Chen SA,
- Luk HN,
- Ding PY,
- Chang MS,
- Chiang CE
- Masse S,
- Farid T,
- Dorian P,
- Umapathy K,
- Nair K,
- Asta J,
- Ross H,
- Rao V,
- Sevaptsidis E,
- Nanthakumar K
- Babenko AP,
- Gonzalez G,
- Aguilar-Bryan L,
- Bryan J
- Zhang H,
- Flagg TP,
- Nichols CG
- Lee SY,
- Lee CO
- Morley GE,
- Anumonwo JM,
- Delmar M
- Selzer A,
- Wray HW
- Lau DH,
- Clausen C,
- Sosunov EA,
- Shlapakova IN,
- Anyukhovsky EP,
- Danilo P Jr.,
- Rosen TS,
- Kelly C,
- Duffy HS,
- Szabolcs MJ,
- Chen M,
- Robinson RB,
- Lu J,
- Kumari S,
- Cohen IS,
- Rosen MR
- Gogelein H,
- Hartung J,
- Englert HC,
- Scholkens BA
- Billman GE,
- Englert HC,
- Scholkens BA
- Duytschaever M,
- Mast F,
- Killian M,
- Blaauw Y,
- Wijffels M,
- Allessie M
Novelty and Significance
What Is Known?
Ventricular fibrillation (VF) is the most common cause of sudden cardiac death.
High-voltage electric shock is the most common procedure to terminate VF.
What New Information Does This Article Contribute?
During VF in myopathic human hearts, there is spatio-temporal heterogeneity in refractoriness across the left ventricular myocardium.
Blockade of KATP channels by glibenclamide attenuates spatio-temporal heterogeneity in refractoriness, causing spontaneous termination of VF.
VF is the most common cause of sudden cardiac death. The only effective therapeutic strategy by which VF can be terminated once it has occurred is electric defibrillation. In this study, we demonstrate for the first time that in cardiomyopathic hearts, gene expression of KATP channel subunits is altered. This is accompanied by difference in refractoriness in myopathic hearts between left ventricular epicardium and endocardium. Blocking KATP channels by glibenclamide decreases this transmural difference in refractoriness and increases spontaneous termination of VF. These findings raise the intriguing possibility of promoting spontaneous defibrillation of VF using intrinsic metabolic mechanisms. In addition, by virtue of having therapeutic electrophysiological properties mainly during ischemia, this strategy has the potential of maximal effect during the global ischemia that accompanies VF. Therefore, KATP channel blockade may have a potential role in both the primary and secondary prevention of VF arrest. In addition, cardiac KATP channel blockers could be useful adjuncts when defibrillation fails or proves difficult after cardiac surgery or for treating resistant VF in the setting of cardiopulmonary resuscitation.