Author + information
- Received October 2, 2011
- Revision received January 4, 2012
- Accepted January 10, 2012
- Published online June 12, 2012.
- Maxim Hardziyenka, MD, PhD⁎,†,
- Maria E. Campian, MD⁎,
- Arie O. Verkerk, PhD⁎,
- Sulaiman Surie, MD‡,
- Antoni C.G. van Ginneken, PhD⁎,
- Sara Hakim, BSc§,
- André C. Linnenbank, PhD⁎,
- H.A.C.M. Rianne de Bruin-Bon, BSc∥,
- Leander Beekman, BSc⁎,
- Mart N. van der Plas, MSc‡,
- Carol A. Remme, MD, PhD⁎,
- Toon A.B. van Veen, PhD§,
- Paul Bresser, MD, PhD⁎,‡,
- Jacques M.T. de Bakker, PhD⁎,†,§ and
- Hanno L. Tan, MD, PhD⁎∥,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Hanno L. Tan, Department of Cardiology, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands
Objectives The purpose of this study was to analyze the electrophysiologic remodeling of the atrophic left ventricle (LV) in right ventricular (RV) failure (RVF) after RV pressure overload.
Background The LV in pressure-induced RVF develops dysfunction, reduction in mass, and altered gene expression, due to atrophic remodeling. LV atrophy is associated with electrophysiologic remodeling.
Methods We conducted epicardial mapping in Langendorff-perfused hearts, patch-clamp studies, gene expression studies, and protein level studies of the LV in rats with pressure-induced RVF (monocrotaline [MCT] injection, n = 25; controls with saline injection, n = 18). We also performed epicardial mapping of the LV in patients with RVF after chronic thromboembolic pulmonary hypertension (CTEPH) (RVF, n = 10; no RVF, n = 16).
Results The LV of rats with MCT-induced RVF exhibited electrophysiologic remodeling: longer action potentials (APs) at 90% repolarization and effective refractory periods (ERPs) (60 ± 1 ms vs. 44 ± 1 ms; p < 0.001), and slower longitudinal conduction velocity (62 ± 2 cm/s vs. 70 ± 1 cm/s; p = 0.003). AP/ERP prolongation agreed with reduced Kcnip2 expression, which encodes the repolarizing potassium channel subunit KChIP2 (0.07 ± 0.01 vs. 0.11 ± 0.02; p < 0.05). Conduction slowing was not explained by impaired impulse formation, as AP maximum upstroke velocity, whole-cell sodium current magnitude/properties, and mRNA levels of Scn5a were unaltered. Instead, impulse transmission in RVF was hampered by reduction in cell length (111.6 ± 0.7 μm vs. 122.0 ± 0.4 μm; p = 0.02) and width (21.9 ± 0.2 μm vs. 25.3 ± 0.3 μm; p = 0.002), and impaired cell-to-cell impulse transmission (24% reduction in Connexin-43 levels). The LV of patients with CTEPH with RVF also exhibited ERP prolongation (306 ± 8 ms vs. 268 ± 5 ms; p = 0.001) and conduction slowing (53 ± 3 cm/s vs. 64 ± 3 cm/s; p = 0.005).
Conclusions Pressure-induced RVF is associated with electrophysiologic remodeling of the atrophic LV.
Right ventricular (RV) failure (RVF) secondary to chronic pressure overload is an important determinant of survival in patients with chronic thromboembolic pulmonary hypertension (CTEPH) and other forms of pulmonary arterial hypertension (PAH) (1,2). These patients develop not only RV dysfunction, but also left ventricular (LV) dysfunction (3,4). LV diastolic filling is diminished, because diastolic LV peak filling rate is related to RV ejection fraction (5). We recently reported in CTEPH patients with RVF and in an experimental model of RVF (monocrotaline [MCT]-induced PAH in rats) that this is associated with a reduction in LV mass and an altered gene expression profile of the LV, which is most likely due to atrophic remodeling of the LV (6). Because unloading and/or atrophy of the LV is associated with electrophysiologic remodeling (7), we hypothesized that electrophysiologic remodeling of the LV also occurs in PAH-induced RVF. We therefore studied electrophysiologic properties and gene expression profiles of the LV in rats with MCT-induced RVF. Moreover, we investigated whether CTEPH patients with RVF exhibited similar electrophysiologic changes in the LV.
Investigations conformed with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996). Animals were cared for in accordance with institutional guidelines for the care and use of laboratory animals. Twenty-five 7-week-old male Wistar rats received an intraperitoneal injection of MCT (60 mg/kg) to induce PAH and RVF. Eighteen control animals were injected intraperitoneally with the same volume of saline (3 ml/kg). As reported previously (8), MCT injection leads to subsequent disease stages, culminating in RVF. To recognize when a rat reached the RVF stage, we used clinical criteria: body weight loss and ≥1 of the following: dyspnea, hypothermia, cyanosis, and lethargy. Rats were sacrificed when they developed RVF (RVF rats) or at the same time after saline injection (control rats).
Echocardiographic and electrocardiographic analysis
Echocardiographic and electrocardiographic (ECG) (limb leads) recordings were made just before sacrifice. Echocardiographic analysis of the RV and LV was conducted as reported previously (8,9). LV volumes were estimated on the basis of the short-axis area–length method (10). Cardiac output was calculated as LV stroke volume multiplied by heart rate. RR interval and QRS/QT duration were analyzed in ECG lead II. Because the diagnosis of RVF was determined on the basis of clinical, rather than echocardiographic criteria, it was not necessary to perform echocardiography in all animals. Accordingly, we conducted echocardiography in 20 rats.
Hearts were harvested immediately after sacrifice, and perfused according to the Langendorff technique at constant pressure (70 cm water) with Tyrode's solution (37°C) containing: sodium chloride 130 mM, potassium chloride 5.6 mM, calcium chloride 2.2 mM, magnesium chloride 0.55 mM, sodium bicarbonate 24.2 mM, and glucose 11.1 mM (pH 7.4 with 95% oxygen and 5% carbon dioxide) (11). Extracellular electrograms of the LV and RV were recorded with a 208-point multielectrode array, mounted on a micromanipulator. Electrodes were silver wires of 0.1 mm diameter and arranged in a 16 × 13 grid at 0.5-mm interelectrode distances. Recordings were made during stimulation (rectangular pulses of 2-ms duration and twice diastolic threshold) from the center of the grid at a basic cycle length of 200 ms. The effective refractory period (ERP) (the coupling interval of the shortest premature stimulus that failed to achieve local capture) was determined for each ventricle separately as follows. Every eighth stimulus was followed by 1 premature stimulus. Starting at 180 ms, the coupling interval of the premature stimulus was reduced in steps of 10 ms until conduction block. Then, starting from the coupling interval of the last conducted premature stimulus, the coupling interval of the premature stimulus was reduced in steps of 1 ms until ERP. We constructed activation maps from activation times (12), and calculated longitudinal and transversal conduction velocities (11) (Fig. 1). The susceptibility to tachyarrhythmias was tested as reported previously (13).
Myocyte isolation and measurement of cell dimensions
LV and RV myocytes of 7 RVF rats and 4 control rats were isolated enzymatically (14), and the length and width of 50 randomly selected viable rod-shaped myocytes were measured (3-μm resolution).
Action potentials (APs) and whole-cell sodium current (INa) were recorded using standard methods. Voltage control, data acquisition, and analysis were accomplished using custom-made software. INa signals were low-pass filtered (cutoff 5 kHz) and digitized at 20 kHz. APs were filtered and digitized at 10 and 40 kHz, respectively. Cell membrane capacitance (Cm) was estimated as described previously (15). Series resistance was compensated to reduce resistance voltage errors and increase the temporal resolution by 80% in a whole-cell voltage clamp. APs were recorded at 36°C ± 0.2°C with the perforated patch-clamp technique. INa was recorded at 20°C with the whole-cell ruptured patch-clamp technique.
LV and RV tissues were excised from the positions where the multielectrode recordings were made. These tissues were obtained from the same positions in all animals (i.e., the LV and RV free walls). These tissues were fixed in formalin (RVF rats, n = 4 and control rats, n = 4). Seven-micrometer-thick sections were cut parallel to the epicardium, stained with Picro-sirius Red, and examined by light microscopy (10× magnification) for collagen. The amount of collagen in the recording area was determined using the Image-Pro Plus software package (version 5.02, Media Cybernetics, Bethesda, Maryland) after digitizing 6 randomly selected fields per section with a slide scanner.
Western blots and immunohistochemistry
Total cellular protein was isolated from the LV and RV free walls from the positions where the multielectrode recordings were made (RVF rats, n = 5 and control rats, n = 5). Standard SDS-PAGE and Western blot analyses were conducted using antibodies directed against the Connexin-43 protein (Cx-43, Transduction). To compare the expression levels of Cx-43 between RVF and control hearts, the pooled samples were separated on gel (25 μg total cellular protein per sample) and blotted onto nitrocellulose. The optical densities of the obtained signals were quantified (ImageQuant software, Molecular Dynamics, Sunnyvale, California) and corrected for possible deviations in the amount of transferred total protein. For this, reversible Ponceau-S staining of the blots was performed, after which signals were digitized by scanning at 400 dpi. Cryosections (10-μm thickness) were sliced perpendicular to the epicardial surface from biopsies rapidly frozen in liquid nitrogen. Thus, each section encompassed the entire ventricular wall. Immunohistochemistry was performed as described previously (13).
Quantitative real-time polymerase chain reaction
RNA was isolated from LV samples of 4 RVF and 5 control rats from the positions where the multielectrode recordings were made. mRNA expression levels of the main subunits of the cardiac sodium channel (Scn5a), the L-type calcium channel (Cacna1c), and the various potassium channels (Kcnip2, Kcnd2, Kcnd3, Kcna4, Kcnq1, Kcne1) were quantified using the LightCycler system for real-time polymerase chain reaction (RT-PCR) (Roche Applied Science, Basel, Switzerland). Quantitative RT-PCR data were analyzed with the LinRegPCR program (16). All samples were processed in triplicate and expression levels were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH).
We studied 26 CTEPH patients who were included in a previous study from our group (17). For the present study, we divided them into 2 groups: patients with RVF (n = 10) and patients without RVF (n = 16). RVF was defined as tricuspid annulus plane systolic excursion (TAPSE) <16 mm (18). All subjects gave written informed consent. Investigations were approved by the local institutional review board.
Echocardiographic data of the whole cohort were presented in a previous study (17). In this study, we compared the echocardiographic data between RVF and non-RVF patients.
All patients underwent in situ epicardial mapping of the LV lateral wall during native circulation at pulmonary endarterectomy (body core temperature 36.0°C to 36.5°C) as described previously (17). Using a multielectrode grid (144 or 126 electrodes, 1 × 1- or 0.4 × 0.4-mm interelectrode distance, respectively), we determined LV ERP by programmed electrical stimulation from the center of the grid, and calculated longitudinal and transversal conduction velocities from ellipsoid activation patterns.
Statistical analysis was performed using SPSS software (version 16.0.0, SPSS, Inc., Chicago, Illinois). All data are presented as mean ± SEM. Groups were compared using unpaired 2-tailed Student t tests, Mann-Whitney U nonparametric test, or chi-square test when appropriate. A p value ≤0.05 was considered statistically significant.
Echocardiographic and autopsy findings
RVF occurred at 26 ± 2 (range, 22 to 30) days after MCT injection. Table 1 shows echocardiographic and morphometric variables of control and RVF rats. Compared with control rats, RVF rats had a significantly thicker RV wall, a dilated RV with reduced contractility (TAPSE), and an abnormal echocardiographic pattern of RV diastolic filling (lower tricuspid E/A ratio). Diastolic filling of the LV was also altered (reduced mitral E/A ratio and LV end-diastolic volume). This was associated with significantly diminished LV stroke volume and cardiac output. These echocardiographic data were in agreement with autopsy findings; compared with control rats, RVF rats had significantly reduced body, liver, and LV weights, and increased lung and RV weights. Moreover, RVF rats had pleural effusion, whereas control rats did not.
Electrocardiographic and epicardial mapping
RVF rats (n = 6) had significantly longer QRS and QT durations than control rats (n = 4, 34 ± 1 ms vs. 28 ± 1 ms; p = 0.004 and 120 ± 10 ms vs. 64 ± 4 ms; p = 0.003, respectively). RR interval was not significantly different between RVF and control rats (196 ± 14 ms vs.154 ± 16 ms; p = 0.1). LV ERP was significantly longer in RVF rats than in control rats (Fig. 2A), whereas LV longitudinal conduction velocity was significantly reduced (Fig. 2B). In the RV, ERP was also longer in RVF rats than in control rats (Fig. 2A). In contrast to the reduction in longitudinal conduction velocity in the LV of RVF rats, longitudinal and transversal conduction velocities in the RV were higher in RVF rats than in control rats (Figs. 2B and 2C). Tachyarrhythmias were noninducible during programmed electrical stimulation, neither in RVF rats, nor in control rats.
To establish the cellular basis of ERP prolongation and reduction in longitudinal conduction velocity, we conducted electrophysiologic studies in isolated LV and RV myocytes. Figure 3A shows representative APs from a LV and RV myocyte of a control rat and a RVF rat. LV and RV myocytes of RVF rats had significantly longer APs at 90% repolarization at all pacing rates tested (Figs. 3B and 3C). In contrast, no significant differences in maximal upstroke velocity were observed between RVF and control rats. Similarly, no differences in resting membrane potential or AP amplitude were found (data not shown). In further support of the notion that conduction slowing was not explained by changes in ionic currents, patch-clamp studies of INa revealed that mean INa densities and voltage dependence of INa steady-state activation/inactivation in LV myocytes were not different between RVF and control rats (Table 2,Fig. 4). In RV myocytes, V1/2 of activation was ∼9 mV more negative in RVF than in control rats (p < 0.05), whereas voltage-dependent inactivation was not different (Table 2, Fig. 4).
Morphometric and tissue properties, protein levels, and mRNA expression patterns
Given that changes in conduction velocity may occur on the basis of altered cell dimensions (reduced conduction velocity at smaller myocyte size ) and/or changes in cell-to-cell impulse transmission, we next studied cell dimensions, protein levels of Cx-43, and tissue properties. LV myocytes were shorter and narrower in RVF rats than in control rats (111.6 ± 0.7 μm vs. 122.0 ± 0.4 μm ; p = 0.02, and 21.9 ± 0.2 μm vs. 25.3 ± 0.3 μm; p = 0.002, respectively) (Fig. 3D), although their membrane capacitances were similar (30 myocytes of RVF rats, 144 ± 6 pF vs. 35 myocytes of control rats, 143 ± 7 pF; p = 0.9). Compared with control rats, RV myocytes of RVF rats were wider (27.7 ± 0.5 μm vs. 23.1 ± 0.3 μm; p = 0.015) and had larger membrane capacitance (166 ± 9 pF [26 myocytes] vs. 132 ± 6 pF [46 myocytes], p = 0.002), whereas their lengths were not different (108.4 ± 0.9 μm vs. 110.9 ± 0.5 μm; p = 0.9) (Fig. 3D). Figure 5A shows double labeling of Cx-43 and alpha-actinin. Cx-43 expression was homogeneous and aligned along the intercalated disc, and, to a lesser extent, along the lateral myocyte borders. Cx-43 expression in the LV was less dense in RVF rats than in control rats, but remained homogeneous. In the RV, Cx-43 expression was denser in RVF rats than in control rats. Quantification of Cx-43 protein expression by Western blotting indicated a 24% decrease in the LV and a 30% increase in the RV of RVF rats (signals were corrected for differences in the total amount of transferred protein) (Fig. 5B). The amount of interstitial collagen deposition was increased in the RV, but not the LV, of RVF rats compared with controls (Fig. 5C).
Finally, we conducted quantitative PCR analysis in LV myocytes of the major ion channels that constitute the rat LV AP. This analysis revealed that mRNA expression levels of Kcnip2 (which encodes KChIP2, an auxiliary subunit of the transient outward potassium current, Ito) were significantly reduced in RVF rats, in accordance with the observed AP/ERP prolongation. In contrast, expression of other potassium, sodium, and calcium channel subunits remained unchanged (Fig. 5D).
Patient characteristics and echocardiographic analysis
Patient characteristics are summarized in Table 3. All patients used oral anticoagulants for at least 3 months, 7 used the dual endothelin receptor antagonist bosentan, and none used antiarrhythmic drugs (including calcium channel blockers and beta-blockers). There were no significant differences in age and/or gender between patients without RVF (n = 16) and with RVF (n = 10) (56 ± 4 years vs. 61 ± 3 years; p = 0.3; and 7 women/9 men vs. 6 women/4 men; p = 0.7). No patient had coronary artery disease at coronary angiography. Patients with RVF had, on average, longer PQ and QTc durations, higher mean pulmonary arterial pressure, right atrial pressure, and pulmonary capillary wedge pressure, as well as larger total pulmonary resistance. In contrast, they had smaller RV stroke volume and cardiac output. Echocardiographic findings are summarized in Table 4. Patients with RVF had larger RV end-diastolic diameter with lower indexes of early diastolic relaxation (E′) and contractility (TAPSE). They also had significantly lower LV end-diastolic diameter and volume, along with lower early LV diastolic relaxation velocity and LV ejection fraction, and more prominent leftward shift of the interventricular septum (LV early diastolic eccentricity index).
Electrocardiographic analysis and epicardial mapping
Patients with RVF had significantly longer PQ and QTc duration than patients without RVF (Table 3). LV ERP in patients with RVF was significantly longer than in patients without RVF (Figs. 6A and 6B), and longitudinal conduction velocity was significantly reduced (Figs. 6C and 6D).
The atrophic LV of rats with MCT-induced RVF exhibited evidence of electrophysiologic remodeling (i.e., longer AP/ERP and slower conduction velocity). AP/ERP prolongation was consistent with reduced expression of Kcnip2. Conduction slowing was not explained by impaired impulse formation, as AP maximum upstroke velocity, INa magnitude/properties, and mRNA levels of Scn5a were unaltered. Instead, impulse transmission was hampered, both by reduction in cell size and by impaired cell-to-cell impulse transmission (reduced Cx-43 levels). AP/ERP prolongation and conduction slowing of the LV were also found in CTEPH patients with RVF.
LV function in RVF due to chronic RV pressure overload
Clinical (3,5,19,20) and experimental (21) studies demonstrated that diastolic filling of the LV is impaired in chronic PAH. LV filling rate relates to RV ejection fraction (5). Accordingly, we found that LV early diastolic relaxation velocity and LV end-diastolic volume were reduced in patients with RVF. Similarly, LV diastolic filling was impaired in RVF rats. Aside from abnormal LV diastolic filling, LV systolic function (LV stroke volume and ejection fraction) was affected in CTEPH patients with RVF and in RVF rats. These findings agree with previous studies that showed diminished LV contractility in patients with chronic RV pressure overload (4,20). Moreover, in a previous study (6), we found that LV mass was reduced in CTEPH patients with chronic RV failure, but that LV mass normalized after pulmonary endarterectomy. This finding suggests that reduced LV systolic function in the setting of RV failure may be due to a secondary adaptation with abnormal LV underfilling and mechanical unloading.
Possible mechanisms of LV electrophysiologic remodeling during RVF
The electrophysiologic changes in the LV during RVF that we observed both in patients and rats (AP/ERP prolongation and conduction slowing) were similar to changes occurring in the LV during LV failure (22,23). Moreover, a recent study showed that (sub)epicardial LV monophasic APs are prolonged in MCT rats with RV hypertrophy (24). Various neurohumoral and mechanical factors may affect LV electrophysiology in RVF after chronic PAH (3,25–27). In the present study, Kcnip2 mRNA expression levels were reduced. This was predicted to result in Ito reduction, consistent with previous reports that linked hypertrophy and/or failure to reduced KChIP2 expression and Ito density (24,28–30). Although not measured in this study, elevated angiotensin II levels may cause both Ito reduction (26,31) and reduction in Cx-43 expression (32). Similarly, endothelin-1 may play a role, as it caused AP prolongation in an experimental study (33), whereas its plasma levels were augmented in CTEPH patients (25) and MCT-treated rats (34). Alternatively, underfilling may impair LV diastolic function in patients with CTEPH (3) and MCT-injected rats with RV hypertrophy and failure (21). Mechanical unloading reduces cell dimensions (35,36), regardless of enhanced neurohumoral stimulation (35), and results in AP prolongation (7). In our study, length and width of LV myocytes of RVF rats were smaller than those in control rats, and LV weight was reduced, similar to previous studies (27,37,38). Reduced cell dimensions may in part explain the reduction in longitudinal conduction velocity (11,39).
Differences in electrophysiologic remodeling between the left and right ventricles
We found that changes in AP/ERP during RVF were qualitatively similar in the RV and LV, whereas changes in conduction velocities were divergent between the RV and LV. Moreover, AP/ERP prolongation was more prominent in the RV than in the LV. The resulting dispersion in electrophysiologic properties between the RV and LV could have predisposed to arrhythmias by generating a substrate for reentry. However, we failed to induce arrhythmias in RVF rats or control rats. This may be, in part, due to prolongation of the excitation wave (the mathematical product of conduction velocity and ERP) (40) found in the present study (data not shown), along with the preservation of the homogeneous distribution of Cx-43.
First, LV myocardial biopsies were not taken from CTEPH patients to assess cell and/or tissue morphology and study cellular electrophysiology. Instead, in-depth analysis of electrophysiologic properties of the LV was conducted in the MCT rat model of RVF. Second, we assessed refractoriness and conduction velocities in animals during Langendorff perfusion. Therefore, we cannot exclude that acute changes in LV loading conditions influenced the electrophysiologic properties of the LV due to mechanoelectrical feedback (41). However, we also found ERP prolongation and LV longitudinal conduction slowing in CTEPH patients with RVF whose native circulation was maintained during recording. Moreover, ERP prolongation and LV conduction slowing during Langendorff perfusion, and AP prolongation in isolated myocytes support the notion that the observed electrophysiologic remodeling is mainly explained by impact of chronic RV pressure overload. This finding is in agreement with a previous study (41) in a normal canine heart, in which only acute LV volume loading, but not an acute increase in LV afterload changed AP duration. Similarly, acute LV volume loading resulted in conduction slowing in isolated rabbit hearts (41). Another study limitation was that epicardial mapping in CTEPH patients was performed under general anesthesia when the pericardium was open. Although a previous study showed that pericardium-mediated interventricular interaction in patients with CTEPH is negligible because of pericardial adaptation to slowly progressive pulmonary hypertension (42), intrapericardial pressures may well have been increased in the present study, contributing to LV underfilling. If so, opening of the pericardium might have led to changes in interventricular interaction that affected the electrical properties of the LV due to mechanoelectrical feedback.
Electrophysiologic remodeling occurs in the atrophic LV of CTEPH patients with RVF and in the MCT rat model of PAH-induced RVF. This is associated with abnormal LV diastolic filling.
The authors gratefully thank Dr. Antonius Baartscheer and Cees A. Schumacher for their supervision during the cardiac myocyte isolation.
For an expanded Methods section, please see the online version of this article.
Dr. Tan was supported by the Royal Netherlands Academy of Arts and Sciences (KNAW) and the Netherlands Organization for Scientific Research (NWO, grant ZonMW Vici 918.86.616). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- action potential
- brain-type natriuretic peptide
- chronic thromboembolic pulmonary hypertension
- effective refractory period
- whole-cell sodium current
- transient outward potassium current
- left ventricle
- pulmonary arterial hypertension
- real-time polymerase chain reaction
- right ventricle
- right ventricular failure
- tricuspid annulus plane systolic excursion
- Received October 2, 2011.
- Revision received January 4, 2012.
- Accepted January 10, 2012.
- American College of Cardiology Foundation
- Gurudevan S.V.,
- Malouf P.J.,
- Auger W.R.,
- et al.
- Hardziyenka M.,
- Campian M.E.,
- Reesink H.J.,
- et al.
- Morgan E.E.,
- Faulx M.D.,
- McElfresh T.A.,
- et al.
- Reichek N.,
- Helak J.,
- Plappert T.,
- Sutton M.S.,
- Weber K.T.
- Wiegerinck R.F.,
- Verkerk A.O.,
- Belterman C.N.,
- et al.
- Ruijter J.M.,
- Ramakers C.,
- Hoogaars W.M.,
- et al.
- Hardziyenka M.,
- Campian M.E.,
- Bouma B.J.,
- et al.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Louie E.K.,
- Rich S.,
- Brundage B.H.
- Akar F.G.,
- Rosenbaum D.S.
- Akar F.G.,
- Spragg D.D.,
- Tunin R.S.,
- Kass D.A.,
- Tomaselli G.F.
- Benoist D.,
- Stones R.,
- Drinkhill M.,
- Bernus O.,
- White E.
- Radicke S.,
- Cotella D.,
- Graf E.M.,
- et al.
- Yu H.,
- Gao J.,
- Wang H.,
- et al.
- Fischer R.,
- Dechend R.,
- Gapelyuk A.,
- et al.
- Sakai S.,
- Miyauchi T.,
- Sakurai T.,
- et al.
- Miyauchi T.,
- Yorikane R.,
- Sakai S.,
- et al.
- Lisy O.,
- Redfield M.M.,
- Jovanovic S.,
- et al.
- McGowan B.S.,
- Scott C.B.,
- Mu A.,
- McCormick R.J.,
- Thomas D.P.,
- Margulies K.B.
- Handoko M.L.,
- de Man F.S.,
- Happe C.M.,
- et al.
- Werchan P.M.,
- Summer W.R.,
- Gerdes A.M.,
- McDonough K.H.
- Spach M.S.,
- Heidlage J.F.,
- Dolber P.C.,
- Barr R.C.
- Rensma P.L.,
- Allessie M.A.,
- Lammers W.J.,
- Bonke F.I.,
- Schalij M.J.
- Hansen D.E.
- Blanchard D.G.,
- Dittrich H.C.