Skip Navigation


Cardiovascular Research Advance Access first published online on November 9, 2007
This version [Corrected Proof] published online on December 20, 2007
Cardiovascular Research, doi:10.1093/cvr/cvm060
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
77/3/489    most recent
cvm060v2
cvm060v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Stoll, M.
Right arrow Articles by Decking, U. K.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stoll, M.
Right arrow Articles by Decking, U. K.M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Spatial heterogeneity of myocardial perfusion predicts local potassium channel expression and action potential duration

Marion Stoll, Michael Quentin, Andrej Molojavyi, Volker Thämer and Ulrich K.M. Decking*

Department of Cardiovascular Physiology, Heinrich-Heine-University Düsseldorf, Universitätsstr. 1, Building 22.03,40225 Düsseldorf, Germany

* Corresponding author. Tel: +49 211 81 12651; fax: +49 211 81 12672. E-mail address: decking{at}uni-duesseldorf.de

Time for primary review: 26 days


   Abstract
Top
 Abstract
1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Funding
 References
 
Aims: In the heart, there is not only a transmural gradient of left ventricular perfusion and action potential duration (APD), but also spatial heterogeneity within each myocardial layer, where local blood flow and energy turnover vary more than three-fold between individual regions. We analysed at high spatial resolution whether a corresponding heterogeneity also extends to ion channel gene expression and APD.

Methods and results: In the open-chest beagle dog, left ventricular 300 µL samples of very low or high flow were identified by radioactive microspheres and expression levels determined by quantitative PCR. The distribution of epicardial APD was assessed by mapping local activation repolarization intervals (ARIs) and QT interval (QT). ERG, the potassium channel mediating IKr, and KChIP2, the interacting protein modulating Ito, were increased in Low flow (3.3- and 2.5-fold, P < 0.001 and <0.05, respectively; n = 6 hearts, 30–31 samples each) as compared with High flow areas. This suggested enhanced repolarizing currents in Low flow areas, and in consequence, mathematical model analysis predicted a shorter local APD upon enhanced ERG and IKr. Epicardial mapping revealed a patchy, temporally stable APD pattern (n = 11), a small apico-basal gradient and an APD prolongation induced by the ERG blocker dofetilide predominantly in areas of short basal ARI or QT, respectively (n = 9). In addition, in Short QT areas, ERG expression was three-fold increased (P < 0.05, n = 4).

Conclusion: The spatial pattern of perfusion is matched by the novel patterns of K+ channel expression and APD. Whenever this newly recognized intramural dispersion of APD increases, it may contribute to arrhythmogenesis.

KEYWORDS Ion channels; Gene expression; Action potentials; Coronary circulation; Regional blood flow; Mapping; Microspheres; Computer modelling

Received September 19, 2006; revised November 1, 2007; accepted November 4, 2007


   1. Introduction
Top
 Abstract
 1. Introduction
2. Methods
 3. Results
 4. Discussion
 Funding
 References
 
In the past decade, it is becoming increasingly evident that the left ventricular myocardium cannot be simply regarded as a homogeneous structure, which serves as a single, contractile function. It is rather a complex organ, rich in its cellular diversity, which displays substantial spatial heterogeneity of function, metabolism as well as gene and protein expression.1 In the left ventricular myocardium, a transmural gradient of perfusion and electrophysiological properties, e.g. action potential duration (APD), is well established. Under physiological conditions, subendocardial blood flow exceeds subepicardial perfusion by about 1.25:1, reflecting greater oxygen requirements and systolic wall stress in the deeper myocardial layers.2,3

This transmural flow gradient is paralleled by a gradient of APD, where the longer APD in the subendocardium results in later recovery from depolarization compared with the subepicardium.4 In most species, there is a marked gradient of the transient outward current (Ito) from the subepicardium to the subendocardium. This Ito gradient is based on an expression gradient of the K+ channel subunits Kv4.3 and KChIP2, the Kv channel-interacting protein 2.5 It contributes to the prominent phase 1 repolarization and the shorter action potential in the subepicardium (for review:6–8). In the canine heart, a transmural gradient of the rapid component of the delayed rectifier K+ current (IKr) mediated by the >two-fold enhanced ERG (KCNH2 or LQT2) expression in the subepicardium was recently reported as well.4 In addition to these transmural gradients, there is also a gradient in APD from apex to base,9 consistent with the late activation and early repolarization of the myocardial base.

Beyond these transmural gradients of flow and functional properties, within each myocardial layer there is also a spatial heterogeneity: initial studies, focusing on myocardial blood flow, demonstrated a fractal pattern in various species10,11 that proved to be temporally stable for weeks.12 For example at a myocardial resolution of 200 µL in sheep or baboons, a spatial dispersion of blood flow of 20% was observed, where local perfusion between individual myocardial tissue samples differed more than three-fold. Subsequent studies elucidated the metabolic and molecular basis of this spatial heterogeneity. Low flow areas, i.e. areas receiving less than 60% of the average left ventricular perfusion under physiological conditions, are not hypoxic.13 High flow areas, i.e. areas receiving >140% of the mean blood flow, display enhanced fatty acid uptake14 and energy turnover.15 In fact, the local flow pattern reflects the local energy turnover and demand.13,15 In spite of a similar morphology, Low and High flow areas differ also in their local proteome, which favours fatty acid oxidation in high flow regions.12 Although the above data translate into a coherent spatial pattern of flow, metabolism and gene and protein expression, both the underlying mechanism and functional correlates remain elusive.

Since the transmural gradient of myocardial perfusion was paralleled by a subendo-/ subepicardial gradient of APD based on differential expression of potassium channels, we wondered whether the spatial heterogeneity of flow, metabolism and gene expression observed within each myocardial layer extended also to ion channel expression and APD. Specifically, we compared Low, Mean and High flow samples and tested the hypothesis that in Low flow areas the expression of voltage-gated potassium channels mediating repolarizing outward currents (Ito, IKs, IKr, and IK1) is increased, resulting in a shortening of APD, which might explain the previously observed decreased energy demand. Candidate genes included Kv4.3 (KCND3), KChIP 2 (KCNIP2), KvLQT1 (KCNQ1), minK (KCNE1), ERG (KCNH2), and Kir2.1 (KCNN4), as well as Nav1.5 (SCN5A) and Cav1.2 (CACNA1C). To assess the local APD, epicardial mapping enabled the determination of the local QT and the local activation repolarization interval (ARI), the established measure of the local APD.16


   2. Methods
Top
 Abstract
 1. Introduction
 2. Methods
3. Results
 4. Discussion
 Funding
 References
 
The investigation was approved by the local government, carried out in a total of 17 open-chest, ventilated beagle dogs (14.6 ± 3.4 kg) and conforms 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). The animals were sedated by thiopental sodium, anaesthetized by piritramide and midazolam and instrumented as described previously.15

2. 1 Flow measurement and sample selection
To determine local myocardial perfusion and identify Low and High flow areas, radioactive microspheres (15 µm, 57Co or 153Gd, Perkin-Elmer) were injected under basal conditions into the left atrium while withdrawing an arterial reference organ.12 At the end of the experiment, the anterior free wall was excised, freeze-clamped, lyophyillized, and cut into ~170 samples (60 ± 22 mg dry weight) dividing the tissue into columns, rows, and at least three transmural layers. The deposition of radioactive microspheres was measured by {gamma}-scintigraphy for flow determinations.12 This enabled the identification of left ventricular samples receiving ≤60%, 100 ± 5% or ≥140% of overall average flow, i.e. Low, Mean, and High flow samples in all transmural layers.

2.2 Experimental protocols and procedures
In a first experimental series (n = 6), Low, Mean, and High flow samples were obtained for the analysis of ion channel gene expression.

Having established significant differences in potassium channel expression, in the second series (n = 11) the spatial distribution of APD and the role of potassium channel expression was assessed by epicardial mapping. In most experiments of this series (n = 9), the effect of the ERG-blocker dofetilide (0.1 mg/kg) on ARI, the local QT interval (QT) and blood flow was assessed. In a subset of experiments (n = 5), the cardiac cycle length was varied by electrical stimulation in the range from 400 to 550 ms before and after dofetilide to assess its effect on QT and ARI independent of its effect on heart rate.

To identify areas of short and long QT under basal conditions, in four of these hearts epicardial mapping of the anterior free wall was repeated five times in 15 min intervals prior to pacing and dofetilide and the time-averaged QT mean for each electrode determined. Selected areas were marked with an intramyocardial injection of 25 µL trypan blue at the end of the experiment. After withdrawal of the electrode sock, 3 mg subepicardial biopsies were obtained using a 16G biopsy needle (Temno Evolution, Allegiance) with a 4 mm biopsy range. This tissue was immediately stored in RNAlater (Qiagen) for further preparation. Thereafter, the left ventricular free wall was excised and treated as described above.

2.3 Epicardial mapping of local surface electrograms
A nylon mesh sock equipped with 128 electrodes (J. Davis, Utah) equally spaced at 6–8 mm distance on the epicardial surface was employed. Following pericardiotomy, the sock was gently pulled across the complete left and right ventricular surface. The junction of left anterior descending and circumflex artery served as anatomical landmark for a consistent placement. Once a steady-state was attained, unipolar electrograms, measured against a common chest reference electrode, were acquired at a sampling rate of 1 kHz by a 16-channel A/D-converter (ADInstruments). The electrodes were divided into eight groups and within 3 min all traces were obtained while heart rate and blood pressure remained stable. The traces were manually analysed for RR, QT and also ARI as a measure of local APD.16 While QT was defined as the period between the initial deflection of the activation complex (QRS correlate) and the end of the T wave, while ARI is the delay between the moment of minimum dV/dt during the activation complex and the moment of maximum dV/dt during the T wave16 (see also Figure 3A). The latter is independent of the polarity of the T wave.17 QT and ARI were corrected for heart rate by Van de Water's formula [e.g. QTc = QT – 0.087(RR – 1000)].18 Preliminary experiments had indicated that for each individual electrode, the variability of the measured QTc and ARIc, respectively, was only about ±3 and ±2% and independent of the local time-averaged QTc or ARIc, which differed by more than 60 and 40 ms, respectively. As a result, an individual electrode displaying e.g. a short QTc in one measurement showed a relatively short QTc in all registrations throughout the experiment. To further enhance the reliability, for local biopsies and selective flow determinations, areas of Short and Long QTc were identified on the basis of average values from five consecutive measurements.

2.4 RNA preparation, reverse transcription, and quantitative real-time PCR analysis
Total RNA was isolated from 31 Low, 53 Mean, and 30 High flow samples and in another set of experiments from biopsy tissue using the RNeasy Fibrous Tissue Mini Kit and the RNeasy Micro Kit, respectively (Qiagen). First-strand cDNA was synthesized using Superscript II RT (Invitrogen) in the presence of 1 µg of total RNA and oligo (dT)12–18 primers.

Gene expression of Kv4.3, KChIP2, KvLQT1, minK, ERG, Kir2.1, Nav1.5, and Cav1.2 was determined by quantitative real-time PCR. For gene amplification, cDNA was amplified using SYBR Green PCR Core Reagents (Applied Biosystems) and primers as listed in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1 Primers used in quantitative real-time PCR experiments to determine mRNA expression of eight cardiac ion channel genes

 
The comparative CT method was used for relative quantification, relating the number of cycles of the target genes to ß-actin at CT 0.5 and at CT 1.0. Because cycle efficiency was 100% in the range studied, relative amount was determined by using the formula 2–{Delta}CT where {Delta}CT describes the difference between ß-actin and the six genes.

2.5 Model analysis of ion channel expression and APD
To evaluate the relevance of different ion channel expression levels observed in Low, Mean, and High flow areas for APD, a mathematical model of canine ventricular cardiomyocyte electrophysiology19 was employed. This model had been translated to run within the Cellular Open Resource (COR) environment.20 Consistent with experimental observations, all simulations were carried out at a cycle length of 550 ms and only steady-state data were used for further analysis. The standard parameter set of the model was taken to present Mean flow areas. To simulate Low and High flow areas, the maximal channel conductances were modulated in proportion to the expression levels observed in Low and High flow areas (Table 2). Since Ito is dependent both on KChIP2 and Kv4.3, the average expression of these two genes was employed, and the same was true for IKs, which is dependent on KvLQT1 and minK. Blocking ERG by dofetilide was simulated by a 90% decrease of maximal ERG conductance.


View this table:
[in this window]
[in a new window]

  		Table 2 Differential mRNA expression of selected ion channels in Low and High flow (LF, HF) areas relative to Mean flow (MF) areas

 
2.6 Statistics
All data are given as mean ± standard deviation except where indicated. Regional differences within the heart between Low and High flow samples were compared using Student's t-test for paired data, taking separate samples from the same flow or QT domain as replicate measurements. One-way ANOVA followed by Bonferonni's correction for multiple comparisons was applied where appropriate. Results were considered significant at P < 0.05.


   3. Results
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
4. Discussion
 Funding
 References
 
In the 17 anaesthetized dogs studied, basal heart rate, mean aortic pressure, and myocardial blood flow were 82 ± 25 bpm, 79 ± 11 mmHg, and 0.91 ± 0.42 mL/min/g, the subendo-/subepicardial perfusion ratio being 1.21 ± 0.12.

When analysed at the spatial resolution of 300 µL (sample mass 60 mg dry weight), there was substantial flow heterogeneity within each myocardial layer. The coefficients of variation (CV) for subepicardial, midmyo- and subendocardial flow ranged from 0.29 to 0.36, the overall CV being 0.32 ± 0.08. In each dog, about one-tenth of samples received less than 60% or more than 140% of the overall average flow and were designated Low and High flow areas, respectively. The flow difference between Low and High flow areas exceeded the small transmural gradient by far.

3.1 Ion channel expression in Low and High flow areas
When comparing Low, Mean, and High flow samples from all transmural layers, a significant inverse correlation between ERG, but also KChIP2 expression in the individual samples and their flow could be observed (Figure 1A). While ERG mediates IKr, KChIP2 is the Kv channel-interacting protein modulating Kv4.3 and thus Ito. In principle, both enhanced IKr and Ito will contribute to a shorter APD in Low flow areas (see below). ERG showed a 3.3-fold and KChIP2 a 2.5-fold higher expression in Low flow as compared with High flow areas (n = 6, Table 2). Also Kv4.3 mediating Ito displayed a tendency towards enhanced expression in Low flow areas (Table 2). Enhanced ERG expression was qualitatively also seen on the protein level (data not shown). As shown in Figure 1B, enhanced ERG and KChIP2 expression levels in Low flow areas were present in two out of three transmural layers; however, neither ERG nor KChIP2 differed significantly in their expression between subendo-, midmyo-, and subepicardium in any of the flow domains (Figure 1B). Thus, the correlation seen was indeed related to flow.


Figure 1
View larger version (35K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1 Correlation between ERG respectively KChIP2 expression and flow, demonstrating enhanced expression of KChIP2 and ERG in Low as compared with High flow areas (LF, HF) independent of the transmural layer. (A) Data ({Delta}CT) from all layers, normalized for the expression in Mean flow areas (MF). (B) Expression in LF, MF, and HF samples from the subepicardial (Subepi), midmyocardial (Mid), and subendocardial (Subendo) layer, normalized for midmyocardial MF samples. (**P < 0.01, and ***P < 0.001 vs. HF, +P < 0.05 LF vs. HF, n = 6 hearts).

 
There was no flow-related difference in mRNA expression of KvLQT1 and minK, mediating, respectively, modulating IKs, in Kir2.1, the inward rectifier K+-channel, nor in Nav1.5, the voltage-gated Na+-channel or Cav1.2, the {alpha}-subunit of the voltage-gated Ca2+-channel (Table 2).

When comparing subepi- and subendocardial Mean flow samples, there was no difference in Kv4.3, KvLQT1, minK, ERG, and Cav1.2, the respective ratios ranging from 0.75 to 1.37. Only KChIP2 tended to display a 2.86-fold greater expression in the subepicardium that however did not reach the level of statistical significance.

3.2 Model analysis of ion channel expression and action potential duration
To predict the effects of enhanced potassium channel expression on APD in Low flow areas, a mathematical model of the isolated canine ventricular myocyte was employed.19 At a cycle length of 550 ms chosen according to experimental data, the model calculated the AP and the APD at 90% of repolarization (APD90) (Figure 2A). Increasing selectively the maximal conductance for IKr according to the enhanced ERG expression in Low flow areas resulted in a >10 ms shortening of the predicted APD90. Reducing the IKr conductance consistent with the lower ERG expression in High flow areas prolonged APD90 to a smaller extent (Figure 2B, IKr). Similarly, enhancing the maximal conductance of Ito according to the expression data shortened APD90 while reducing Ito had hardly any effect (Figure 2B, Ito). The non-significant differential expression of KvLQT1 and minK governing IKs had no discernible effect on APD90 (data not shown). When modulating both the maximal conductance for Ito, IKr, and IKs taking all the expression data from Table 2 for Low and High flow areas into account, the effect on the predicted action potentials was much greater than upon modulation of a single channel conductance only (Figures 2A and B, IKr+IKs+Ito). The model also predicted that blocking ERG by 90% (dofetilide) would result in an increase in APD90 predominantly in areas characterized by enhanced ERG expression, i.e. in Low flow areas (Figure 2B, IKr+IKs+Ito+dofetilide).


Figure 2
View larger version (21K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2 Model analysis predicted a shorter action potential in Low flow regions. (A) Model results integrating the relative levels of ion channel expression in LF, MF, and HF areas (Table 2). (B) The effects of differential regulation of only IKr or Ito on APD90 were smaller than the combined effect of simultaneous modulation of IKr+Ito+IKs. In the combined model, dofetilide's effects were simulated by decreasing maximal IKr conductance by 90%.

 
3.3 Epicardial mapping to assess spatial differences in local electrograms
We next explored whether the predicted heterogeneity of APD (Figure 2A) translated into a spatial heterogeneity in the intact beating heart. To this end, 128 local cardiac electrograms (Figure 3A) were obtained per heart, enabling the reliable detection of local QT intervals and ARIs. Figure 3B demonstrates a representative map of the anterior cardiac QT distribution. Consistent with the local differences in potassium channel expression, a spatial pattern of epicardial QT intervals was observed, with differences between the shortest and longest left ventricular QTc of 51 ± 15 ms. There was also a similar pattern of local ARI, the average difference between shortest and longest ARIc being 55 ± 17 ms. These patterns were temporally stable for at least 60 min (data not shown). Qualitatively similar maps were obtained in 11 experiments, the average QTc in a total of 1315 traces being 320 ± 48 ms and the average ARIc being 247 ± 31 ms. Within a given heart, the average coefficient of variation was 3.7 and 5.6%, respectively. There was no difference between the left and right ventricular traces in terms of mean QTc (320 ± 48 vs. 317 ± 48 ms) and ARIc (249 ± 32 vs. 237 ± 28 ms); LV and RV traces also did not differ in variability within each ventricle (QTc CV 3.7 vs. 3.5%, and ARIc CV 5 vs. 4.1%). As expected, QTc was about 7 ms and ARIc 14 ms shorter in the basal compared with the apical region (P < 0.05 each).


Figure 3
View larger version (21K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3 Spatial heterogeneity of local unipolar electrograms as observed by epicardial mapping. (A) Unipolar electrogram from a single electrode site. (B) Representative anterior surface map of local QT durations (grey to black colour code: 287 to 349 ms). LV and RV: left and right ventricular surface; LAD: left anterior descending artery.

 
3.4 ERG channel blockade, local APD prolongation and perfusion
To test the model prediction of a preferential APD prolongation by ERG blockade in areas of high ERG expression (Figure 2B, dofetilide), the effects of dofetilide on the local QTc and ARIc were measured in the heart in vivo. Dofetilide increased mean QTc from 311 by 34 ms (+10 ± 6%) and mean ARIc from 243 by 31 ms (+12 ± 7%). The QT prolongation was significantly greater in areas of short basal QTc and smaller in areas of long basal QTc (Figure 4A). In consequence, when correlating the increase in local QTc to its basal value in each heart (n = 9), the average slope was –0.68 ± 0.17 at an average r2 of 0.32. When allocating the basal values of each heart to nine percentile classes and calculating the dofetilide-induced increase, there was a significantly greater increase in the two lowest and a much smaller increase in the two highest QTc classes compared with the overall mean (Figure 4A). In fact, in three hearts in areas of long basal QTc dofetilide induced hardly any rise in QTc (≤3%). When relating the dofetilide-induced rise in local ARIc to basal ARIc, the data were qualitatively the same: dofetilide caused a significantly greater ARI prolongation in the shortest basal ARIc class than in three out of four long ARI classes (Figure 4B).


Figure 4
View larger version (19K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4 Greater APD prolongation by dofetilide in areas of short basal APD as assessed by QTc and ARIc. (A) Basal QTc values from nine hearts were grouped in percentile classes. The average increase for each class induced by dofetilide was determined. **P < 0.01 vs. mean {Delta}QTc. (B) Using the same approach, the increase in ARIc was assessed. **P < 0.01 and ***P < 0.001 vs. ARIc class shown.

 
To test whether a shorter APD in low flow areas might impair the local Ca2+-transient and energy demand and thus directly contribute to the lower flow, in most of the experiments (n = 7) the effects of APD prolongation by dofetilide on local flow were determined. While dofetilide increased both QTc and ARIc substantially in Short QTc, respectively, Short ARIc and thus presumably Low flow areas (Figure 4), there was only a modest elevation of myocardial perfusion in Low flow regions. When correlating flow before and after application of dofetilide, both the positive offset and the slope being <1 reflected a preferential rise in perfusion in Low flow areas (P < 0.001, Figure 5), while the spatial heterogeneity of perfusion was not affected by dofetilide.


Figure 5
View larger version (28K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5 Modest, but significant effect of dofetilide on local flow distribution. The slope (0.89) and offset (10.8) of the linear regression (r2 = 0.704, P < 0.001) indicate a preferential increase in low flow areas (n = 7 hearts, 1084 samples).

 
3.5 ERG expression in Short and Long QTc areas
To directly demonstrate enhanced ERG and KChIP2 expression in short QTc regions, a total of 20 Short and Long local QTc (<97 vs. >103% of the mean) areas were identified by averaging of repeated QTc measurements, which reduced the local SEM to 1.3%. Analysis of biopsies from Short QTc areas indicated a modest increase in KChIP2 and revealed a three-fold higher ERG expression (Figure 6), consistent with a substantial role of IKr in APD shortening. When relating local ERG expression to ARIc in those Short and Long QTc areas, there was a significant negative correlation, the slope being –9,8, i.e. a doubling of ERG expression upon a 10% shorter basal ARIc.


Figure 6
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 6 Enhanced KChIP2 and ERG mRNA expression in short QTc areas (*P < 0.05 short vs. long QTc, n = 4; 9–11 samples each).

 

   4. Discussion
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
Funding
 References
 
While across the left ventricular myocardial wall, transmural gradients have long been established, the current study adds a novel dimension to the spatial heterogeneity in the heart. At high spatial resolution, there is not only a fractal patchwork of perfusion within each myocardial layer,11 but there is also substantial variability in potassium channel expression, notably in ERG and KChIP2, which mediate, respectively, govern IKr and Ito. In fact, Low flow areas, receiving <60% of the average left ventricular blood flow under physiological conditions, display an up to 3.5-fold higher expression of ERG and KChIP2 resulting locally in a shorter action potential. This may contribute to a reduced energy demand in Low flow regions.13 The spatial dispersion of APD seen on the epicardial surface could well increase under pathophysiological conditions such as myocardial hypertrophy and subsequently form the arrhythmogenic substrate e.g. of drug-induced torsades de points.21

4.1 Spatial heterogeneity of ion channel expression and function
Previous low resolution studies focused on transmural gradients, e.g. the expression gradient of the K+ channel subunits Kv4.3 and KChIP2 governing Ito.5,6 In the present study, a three-fold higher KChIP2 expression in the subepicardium might have been significant when restricting our subendo- and subepicardial samples to a small rim. A prominent subepicardial ERG expression had been reported in the ferret and the canine heart.4,22 However, in the present and other canine studies there was no transmural difference in IKr23 and ERG expression (Figure 1B). Another traditional focus was the apico-basal gradient. In previous9 and in the present study, a longer apical APD was seen, suggesting enhanced inward or decreased outward currents. However, there was no apico-basal ERG gradient in the canine and human heart24 and no evidence for decreased apical Ito or IKs.

These previous studies could not assess the spatial pattern of ERG and KChIP2 expression now revealed, since they either integrated across large tissue volumes or randomly picked few tissue samples, most probably from Mean flow areas. Only the present high-resolution approach, i.e. systematically dividing subendo-, mid-, and subepicardium of the left ventricular wall and selecting 300 µL samples according to local perfusion made it possible to detect the several-fold differences in ERG and KChIP2 expression between Low and High flow samples. Model analysis indicated substantial consequences for the local APD, consistent with the spatial heterogeneity of the local ARI and QT demonstrated by epicardial mapping (Figures 3–4). The specific ERG blocker dofetilide predominantly prolonged ARI and QT in short basal ARI and QT areas, respectively (Figure 4), which can be attributed to an enhanced local ERG expression. ARI is an established index of the APD at the site of the electrode.16,17 However, the close correlation of QT prolongation and basal QT compared with the looser dependence of ARI prolongation on basal ARI (cf. Figure 4A and B) and the predictive power of basal QT for the level of ERG expression (Figure 6) may indicate that local QT is also a measure of APD in the vicinity of the electrode.

The spatial heterogeneity of ERG and KChIP2 is consistent with the non-uniform, patchy ERG distribution seen in the ferret heart.22 Since flow-related differences in ERG and KChIP2 were present also in the subendocardium (Figure 1), a contribution of M cells to these differences is highly unlikely.23

4.2 Spatial heterogeneity in the heart and its potential basis
Although the myocardium is still viewed by many as a very homogeneous tissue, recent research has firmly established spatial patterns of gene and protein expression,12 flow,10,11 metabolism,25 energy turnover,15 and demand13 even within the left ventricular free wall, revealing a rather complex and nevertheless coherent structure. The current study adds potassium channel gene expression and APD as an additional dimension to this already impressive list.

On the transmural level, under physiological conditions, there is a higher subendocardial blood flow2 associated with enhanced energy turnover. The transmural flow gradient is paralleled by a gradient of APD induced by transmural differences in K+ channel expression (Kv4.3 and KChIP2) and Ito.5 To what extent this APD gradient contributes to gradients of the Ca2+ transient and energy demand remains to be seen.

When studying the myocardium at higher resolution, within each myocardial layer a substantial spatial heterogeneity is revealed (for review:1). Local blood flow differs more than three-fold between Low and High flow areas. The fractal pattern of myocardial perfusion correlates with local metabolism, gene, and protein expression, and the current study extends these observations to local potassium channel expression and APD. The spatial heterogeneity of perfusion reflects the local energy turnover and demand.13,15 Due to the inherently higher energy demand, High flow areas are more likely to suffer from stop-flow ischaemia and reperfusion.26 It is most likely that the quantitative differences between local myocardial tissue areas will increase at even higher spatial resolution, and therefore may represent only the ‘tip of the iceberg’. However, current technology for flow measurements27 does not allow to enhance the spatial resolution while still correlating precisely local perfusion and gene or protein expression in one tissue sample.

Surprisingly, the fundamental basis of this spatial heterogeneity is not well understood. In Low flow areas in the present study, the shortened APD was expected to induce a lower energy demand. However, since ERG blockade increased APD (Figure 4), but only slightly enhanced local perfusion in Low flow (and thus short APD) regions (Figure 5), the observed differences in ERG and APD can only explain a minor fraction of the variability of perfusion. To what extent the local differences in KChIP2 expression and Ito govern local Ca2+ transients, energy demand and flow remains to be seen. A further important factor may be local differences in myocardial fiber strain and work,28 which consecutively result in local variations of flow, gene, and protein expression.12 However, the relevant mechano-sensitive feedback transduction mechanisms have not been elucidated so far. Also the regulatory mechanisms in the ventricular expression of ion channels are incompletely understood and the transcription factors governing transmural or intramural ion channel expression are only beginning to emerge. In the mouse heart, e.g. a transmural expression gradient of a transcription factor governs the gradient for Kv4.2 and Ito.29 However, since Kv4.2 is not expressed in larger species, this finding cannot explain the heterogeneity seen in the present study.

4.3 Possible implications and outlook
The current study indicates a spatial heterogeneity of APD at least in the epicardium, induced by a newly recognized spatial pattern of ion channel expression clearly present in each transmural layer. Although in the past only transmural and apico-basal expression gradients were studied, we have now demonstrated a local pattern of ERG and KChIP2 expression that is related to local flow and present in both the subendo-, midmyo-, and subepicardial layer. This pattern results in a dispersion of ventricular APD and contributes to the pattern of ventricular repolarization as well.30 Enhanced spatial heterogeneity may be of considerable importance under pathophysiological conditions, e.g. during heart failure. Proarrhythmic properties of drugs inducing QT prolongation are frequently studied in the AV-block dog.31 Since in this model dofetilide is well known to induce torsades de pointes, it is tempting to speculate that the spatial heterogeneity of flow, ion channel expression, and APD is increased, possibly resulting in a critical level of dispersion of repolarization. This could also be true in heart failure or myocardial hypertrophy, creating the substrate for functional reentry responsible for the development of life-threatening arrhythmias.32


   Funding
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Funding
References
 
This study was supported by the Deutsche Forschungsgemeinschaft (SFB-612-A4 and De 487/7-1). Dofetilide was provided by Pfizer (Karlsruhe, Germany) free of charge.


   Acknowledgements
 
We are grateful to Jürgen Schrader, MD, for continuous advice and support and to Penny Noble and Alan Garny, Oxford, for providing the CellML translation of the Fox model and the COR environment. We appreciated helpful discussions with Dieter Hafner, Department of Pharmacology, Düsseldorf and skillful technical assistance by Gerald Große, Artyom Verner, and Daniela Haubs.

Conflict of interest: none declared.


   References
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Funding
 References
 

  1. Decking UKM. Spatial heterogeneity in the heart: recent insights and open questions. News Physiol Sci (2002) 17:246–250.[Abstract/Free Full Text]
  2. Duncker DJ, Bache RJ. Inhibition of nitric oxide production aggravates myocardial hypoperfusion during exercise in the presence of a coronary artery stenosis. Circ Res (1994) 74:629–640.[Abstract/Free Full Text]
  3. Duncker DJ, Ishibashi Y, Bache RJ. Effect of treadmill exercise on transmural distribution of blood flow in hypertrophied left ventricle. Am J Physiol Heart Circ Physiol (1998) 275:H1274–H1282.[Abstract/Free Full Text]
  4. Obreztchikova MN, Patberg KW, Plotnikov AN, Ozgen N, Shlapakova IN, Rybin AV, et al. I(Kr) contributes to the altered ventricular repolarization that determines long-term cardiac memory. Cardiovasc Res (2006) 71:88–96.[Abstract/Free Full Text]
  5. Zicha S, Xiao L, Stafford S, Cha TJ, Han W, Varro A, et al. Transmural expression of transient outward potassium current subunits in normal and failing canine and human hearts. J Physiol (2004) 561:735–748.[Abstract/Free Full Text]
  6. Nerbonne JM, Kass RS. Molecular physiology of cardiac repolarization. Physiol Rev (2005) 85:1205–1253.[Abstract/Free Full Text]
  7. Patel SP, Campbell DL. Transient outward potassium current, ‘Ito’, phenotypes in the mammalian left ventricle: underlying molecular, cellular and biophysical mechanisms. J Physiol (2005) 569:7–39.[Abstract/Free Full Text]
  8. Schram G, Pourrier M, Melnyk P, Nattel S. Differential distribution of cardiac ion channel expression as a basis for regional specialization in electrical function. Circ Res (2002) 90:939–950.[Abstract/Free Full Text]
  9. Iwata H, Kodama I, Suzuki R, Kamiya K, Toyama J. Effects of long-term oral administration of amiodarone on the ventricular repolarization of rabbit hearts. Jpn Circ J (1996) 60:662–672.[CrossRef][Medline]
  10. King RB, Bassingthwaighte JB, Hales JRS, Rowell LB. Stability of heterogeneity of myocardial blood flow in normal awake baboons. Circ Res (1985) 57:285–295.[Abstract/Free Full Text]
  11. Bassingthwaighte JB, King RB, Roger SA. Fractal nature of regional myocardial blood flow heterogeneity. Circ Res (1989) 65:578–590.[Abstract/Free Full Text]
  12. Laussmann T, Janosi RA, Fingas CD, Schlieper GR, Schlack W, Schrader J, et al. Myocardial proteome analysis reveals reduced NOS inhibition and enhanced glycolytic capacity in areas of low local blood flow. FASEB J (2002) 16:628–630.[Free Full Text]
  13. Loncar R, Flesche CW, Deussen A. Coronary reserve of high- and low-flow regions in the dog heart left ventricle. Circulation (1998) 98:262–270.[Abstract/Free Full Text]
  14. Caldwell JH, Martin GV, Raymond GM, Bassingthwaighte JB. Regional myocardial flow and capillary permeability-surface area products are nearly proportional. Am J Physiol (1994) 267:H654–H666.[Web of Science][Medline]
  15. Decking UKM, Skwirba S, Zimmermann MF, Preckel B, Thämer V, Deussen A, et al. Spatial heterogeneity of energy turnover in the heart. Pflügers Arch (2001) 441:663–673.[CrossRef][Web of Science][Medline]
  16. Haws CW, Lux RL. Correlation between in vivo transmembrane action potential durations and activation-recovery intervals from electrograms. Effects of interventions that alter repolarization time. Circulation (1990) 81:281–288.[Abstract/Free Full Text]
  17. Coronel R, de Bakker JM, Wilms-Schopman FJ, Opthof T, Linnenbank AC, Belterman CN, et al. Monophasic action potentials and activation recovery intervals as measures of ventricular action potential duration: experimental evidence to resolve some controversies. Heart Rhythm (2006) 3:1043–1050.[CrossRef][Web of Science][Medline]
  18. Van de Water A, Verheyen J, Xhonneux R, Reneman RS. An improved method to correct the QT interval of the electrocardiogram for changes in heart rate. J Pharmacol Methods (1989) 22:207–217.[CrossRef][Web of Science][Medline]
  19. Fox JJ, McHarg JL, Gilmour RF Jr. Ionic mechanism of electrical alternans. Am J Physiol Heart Circ Physiol (2002) 282:H516–H530.[Abstract/Free Full Text]
  20. Garny A, Kohl P, Noble D. Cellular Open Resource (COR): a public CellML based environment for modelling biological function. Int J Bif Chaos (2003) 13:3579–3590.[CrossRef][Web of Science]
  21. Belardinelli L, Antzelevitch C, Vos MA. Assessing predictors of drug-induced torsade de pointes. Trends Pharmacol Sci (2003) 24:619–625.[CrossRef][Medline]
  22. Brahmajothi MV, Morales MJ, Reimer KA, Strauss HC. Regional localization of ERG, the channel protein responsible for the rapid component of the delayed rectifier, K+ current in the ferret heart. Circ Res (1997) 81:128–135.[Abstract/Free Full Text]
  23. Liu DW, Antzelevitch C. Characteristics of the delayed rectifier current (IKr and IKs) in canine ventricular, midmyocardial, and endocardial myocytes. A weaker IKs contributes to the longer action potential of the M cell. Circ Res (1995) 76:351–365.[Abstract/Free Full Text]
  24. Szentadrassy N, Banyasz T, Biro T, Szabo G, Toth BI, Magyar J, et al. Apico-basal inhomogeneity in distribution of ion channels in canine and human ventricular myocardium. Cardiovasc Res (2005) 65:851–860.[Abstract/Free Full Text]
  25. Deussen A. Local myocardial glucose uptake is proportional to, but not dependent on blood flow. Pflügers Arch (1997) 433:488–496.[CrossRef][Web of Science][Medline]
  26. Ghaleh B, Shen Y-T, Vatner SF. Spatial heterogeneity of myocardial blood flow presages salvage versus necrosis with coronary artery reperfusion in conscious baboons. Circulation (1996) 94:2210–2215.[Abstract/Free Full Text]
  27. Decking UKM, Pai VM, Bennett E, Taylor JL, Fingas CD, Zanger K, et al. High-resolution imaging reveals a limit in the spatial resolution of blood flow measurements by microspheres. Am J Physiol Heart Circ Physiol (2004) 287:H1132–H1140.[Abstract/Free Full Text]
  28. Bassingthwaighte JB, Beard DA, Li Z. The mechanical and metabolic basis of myocardial blood flow heterogeneity. Basic Res Cardiol (2001) 96:582–594.[CrossRef][Web of Science][Medline]
  29. Costantini DL, Arruda EP, Agarwal P, Kim KH, Zhu Y, Zhu W, et al. The homeodomain transcription factor Irx5 establishes the mouse cardiac ventricular repolarization gradient. Cell (2005) 123:347–358.[CrossRef][Web of Science][Medline]
  30. Coronel R, Opthof T, Plotnikov AN, Wilms-Schopman FJ, Shlapakova IN, Danilo P Jr, et al. Long-term cardiac memory in canine heart is associated with the evolution of a transmural repolarization gradient. Cardiovasc Res (2007) 74:416–425.[Abstract/Free Full Text]
  31. Carlsson L. In vitro and in vivo models for testing arrhythmogenesis in drugs. J Intern Med (2006) 259:70–80.[CrossRef][Web of Science][Medline]
  32. Di Diego JM, Belardinelli L, Antzelevitch C. Cisapride-induced transmural dispersion of repolarization and torsade de pointes in the canine left ventricular wedge preparation during epicardial stimulation. Circulation (2003) 108:1027–1033.[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Circ Cardiovasc ImagingHome page
R. T. George, A. Arbab-Zadeh, J. M. Miller, K. Kitagawa, H.-J. Chang, D. A. Bluemke, L. Becker, O. Yousuf, J. Texter, A. C. Lardo, et al.
Adenosine Stress 64- and 256-Row Detector Computed Tomography Angiography and Perfusion Imaging: A Pilot Study Evaluating the Transmural Extent of Perfusion Abnormalities to Predict Atherosclerosis Causing Myocardial Ischemia
Circ Cardiovasc Imaging, May 1, 2009; 2(3): 174 - 182.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
77/3/489    most recent
cvm060v2
cvm060v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Stoll, M.
Right arrow Articles by Decking, U. K.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stoll, M.
Right arrow Articles by Decking, U. K.M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?