Author + information
- Received December 24, 2009
- Revision received February 26, 2010
- Accepted March 30, 2010
- Published online August 31, 2010.
- Daehyeok Kim, MD, PhD*,
- Tetsuji Shinohara, MD, PhD*,
- Boyoung Joung, MD, PhD*,
- Mitsunori Maruyama, MD, PhD*,
- Eue-Keun Choi, MD, PhD*,
- Young Keun On, MD, PhD*,
- Seongwook Han, MD, PhD*,
- Michael C. Fishbein, MD†,
- Shien-Fong Lin, PhD* and
- Peng-Sheng Chen, MD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Peng-Sheng Chen, Krannert Institute of Cardiology, 1801 North Capitol Avenue, E475, Indianapolis, Indiana 46202
Objectives The purpose of this study was to test the hypothesis that rhythmic spontaneous sarcoplasmic reticulum calcium (Ca) release (the “Ca clock”) plays an important role in atrioventricular junction (AVJ) automaticity.
Background The AVJ is a primary backup pacemaker to the sinoatrial node. The mechanisms of acceleration of AVJ intrinsic rate during sympathetic stimulation are unclear.
Methods We simultaneously mapped transmembrane potential and intracellular Ca in Langendorff-perfused canine AVJ preparations that did not contain sinoatrial node (n = 10).
Results Baseline AVJ rate was 37.5 ± 4.0 beats/min. The wavefront from leading pacemaker site propagated first through the slow pathway, then the fast pathway and atria. There was no late diastolic Ca elevation (LDCAE) at baseline. Isoproterenol up to 3 μmol/l increased heart rate to 100 ± 6.8 beats/min, concomitant with the appearance of LDCAE that preceded the phase 0 of action potential by 97.3 ± 35.2 ms and preceded the onset of late diastolic depolarization by 23.5 ± 3.5 ms. Caffeine also produced LDCAE and AVJ acceleration. The maximal slope of LDCAE and diastolic depolarization always colocalized with the leading pacemaker sites. Ryanodine markedly slowed the rate of spontaneous AVJ rhythm. Isoproterenol did not induce LDCAE in the presence of ryanodine. The Ifblocker ZD 7288 did not prevent LDCAE or AVJ acceleration induced by isoproterenol (n = 2).
Conclusions Isoproterenol and caffeine induced LDCAE and accelerated intrinsic AVJ rhythm. Consistent colocalization of the maximum LDCAE and the leading pacemaker sites indicates that the Ca clock is important to the intrinsic AVJ rate acceleration during sympathetic stimulation.
The mechanisms of automaticity have traditionally been attributed to the actions of multiple time- and voltage-dependent membrane ionic currents. However, recent studies showed that in addition to these membrane ionic clocks, rhythmic spontaneous sarcoplasmic reticulum (SR) Ca release (the “Ca clock”) can result in rhythmic sodium-calcium exchange current (INCX) activation and sinoatrial node (SAN) depolarization (1,2). We recently confirmed that the membrane ionic clock worked synergistically with the Ca clock to generate sinus rhythm in dogs (3). Whether or not Ca clock contributed to heart rhythm generation in other parts of the heart was unclear. Atrioventricular junction (AVJ) contains specialized conduction tissues, including proximal atrioventricular (AV) bundle, His bundle, and AV node (4). These specialized structures may participate in AVJ automaticity (5). The AVJ is also a common source of cardiac arrhythmia. However, relatively little is known about the mechanisms of automaticity in the AVJ. It has been demonstrated that INCXis present in cells from rabbit atrioventricular node (AVN), and removal of external Na produced a rise of intracellular Ca (Cai) through the reverse mode of INCX(6,7). It is also known that the rate of spontaneous activity of myocytes isolated from rabbit AVN may be decreased by ryanodine and increased by isoproterenol. These changes are accompanied by a decrease and an increase, respectively, in the slope of the preceding Ca ramp (8). These findings suggest that subcellular Caidynamics (the Ca clock) may contribute to the automaticity of the AVJ. However, this hypothesis has not been tested in intact AVJ preparations. We hypothesize that Ca clock is important in AVJ automaticity in intact canine AVJ preparations and that isoproterenol accelerates AVJ rhythm through the increased magnitude of SR Ca release. The purpose of this study is to perform simultaneous transmembrane potential (Vm) and Caimapping to test these hypotheses in a canine model.
This study protocol was approved by the Institutional Animal Care and Use Committee of Indiana University. Hearts from 10 normal mongrel dogs were excised under general anesthesia and were perfused through the aorta with cardioplegic solution. The proximal right and left circumflex arteries were separately cannulated (9). The AVJ preparations (Fig. 1A)were perfused with Tyrode's solution at 37°C with 95% O2and 5% CO2to maintain a pH of 7.4 through the coronary cannula. The composition of Tyrode's solution was NaCl 125 mmol/l, KCl 4.5 mmol/l, NaH2PO41.8 mmol/l, NaHCO324 mmol/l, MgCl20.5 mmol/l, CaCl21.8 mmol/l, and glucose 5.5 mmol/l. Albumin 100 mg/l was added in deionized water. Contractility was inhibited by 10 to 17 μmol/l of blebbistatin. Pseudo electrocardiogram was recorded with bipolar electrodes in the right atrium. A multipolar electrophysiological catheter was used to record the His bundle electrogram.
We performed simultaneous dual optical mapping of Vmand Caiwhile the hearts were Langendorff perfused (3). After mapping of baseline spontaneous beats, pharmacologic intervention was performed. All 10 hearts were mapped both at baseline and during pharmacologic intervention. Among them, 4 were used for β-adrenergic stimulation with isoproterenol. In 2 hearts, ryanodine (3 μmol/l) was used without and with isoproterenol (1 μmol/l) infusion. In the remaining 4 hearts, we performed the following pharmacologic interventions: caffeine infusion (20 mmol/l given as a bolus in 1 s, n = 2), ZD 7288 (3 μmol/l), followed by isoproterenol (1 μmol/l, n = 2).
The tissues were fixed in formalin, and the AVJ region was sectioned into 5 rectangular blocks as described by Inoue and Becker (10). The tissues were paraffin embedded, sectioned, and stained with hematoxylin and eosin and with Masson's trichrome stain. In addition, we performed immunostaining of hyperpolarization-activated cyclic nucleotide-gated potassium channel 4 (HCN4) with rabbit anti-HCN4 polyclonal antibody (Santa Cruz Biotechnology Inc, Santa Cruz, California).
The Caiand Vmtraces were normalized to their respective peak-to-peak amplitude for comparison of timing and morphology. We assign the maximum amplitude of Caiand Vmduring baseline pacing as 1 arbitrary unit (AU). The amplitude and the slope of the late diastolic Ca elevation (LDCAE) were expressed as AU and AU/s, respectively (3). The same applies to diastolic depolarization (DD) detected on the Vmtracings. To generate an isochronal map, the optical mapping data were spatiotemporally filtered with a 3 × 3 × 3 moving average operation. The activation time at each pixel was determined by threshold crossing, usually set at the mid-point of the optical action potential. The activation isochrones were then constructed by grouping the pixels with the same activation time. Paired ttests were used to compare the means at baseline and during pharmacologic intervention in the same preparation. Analysis of variance with Bonferroni post hoc test was used to compare the slopes of LDCAE and DD at different distances and with different doses of isoproterenol. Data were presented as mean ± standard error of the mean. A p value of ≤0.05 was considered statistically significant.
Anatomy of the AVJ
A picture of the AVJ preparation is shown in Figure 1A. The rectangle in the left panel was enlarged and shown at the right. The optical signals during atrial pacing (600-ms cycle length) at different sites are shown in Figure 1B. The locations of the fast pathway (FP), slow pathway (SP), transitional zone (TZ), and AVN were determined both by its anatomical locations (10) and by the characteristics of their optical signals (9). Specifically, the AVN was identified as sites with slow phase 0 with a notch (arrow) that corresponded temporally with the His potential registered by an electrophysiological catheter in the His area (bottom tracing). The location of the AVN was confirmed microscopically (red arrow, Fig. 1C). The HCN4 staining was positive in the AVN and negative in the surrounding tissues (Fig. 1D).
Characterization of AVJ rhythm
Baseline AVJ rhythm had a rate of 37.5 ± 4.0 beats/min. Figure 2Ashows the mapped region. Among them, sites a, b, and c corresponded to the anatomical location of the AVN. Figure 2B shows the isochronal activation map of a single beat during the typical AVJ rhythm, with the earliest activation in the leading pacemaker site as time 0. In the all cases at baseline, the AVJ rhythm began near AVN and slowly propagated toward SP region inferior to AVJ (Fig. 2B, top), and then rapidly propagates toward the other parts of the atrium (Fig. 2B, bottom). In this example, the former portion of the propagation took 130 ms and the latter portion only 15 ms (from 130 to 145 ms). There was an obvious conduction delay between AVN region and the rest of the RA preparation. Figure 2C shows the optical signals recorded at different sites shown in Figure 2B. There was diastolic depolarization (arrows) at the leading pacemaking site near AVN. These diastolic depolarizations occurred without preceding LDCAE on the Caitracing. Panel 2D shows these optical signals in greater detail. The upstroke slope of optical Vmand Caifluorescence was shallow in the AVN and the slow pathway (sites a through d). There was very slow propagation between sites c and f. The propagation from f to i was fast and was associated with a steep slope in the phase 0. The delay between phase 0 of site c and phase 0 of site f in all preparations averaged 102 ± 25.5 ms.
Effects of pharmacologic interventions
Isoproterenol increased the rate of AVJ rhythm in a dose-dependent fashion from 0.01 to 3.0 μmol/l (Fig. 3A).In all preparations studied, the heart rate increased by a maximum of 167.2% (from 37.5 ± 4.0 beats/min to 100.2 ± 6.8 beats/min) during isoproterenol infusion. The slope of DD progressively decreased as the distance between the recording site and the leading pacemaker site progressively increased (Fig. 3B). We also noted that LDCAE appeared at the leading pacemaker sites in all 6 preparations during isoproterenol infusion (Fig. 3C, arrow). LDCAE at the leading pacemaker site preceded the phase 0 action potential upstroke by 97.3 ± 35.2 ms and preceded the onset of late DD by 23.5 ± 3.5 ms (Fig. 3D). Similar to DD, the slope of LDCAE progressively decreased as the distance from pacing site increased (Fig. 3B). In this (Fig. 3C) and an additional 3 preparations, the same site served as the leading pacemaker both at baseline and during isoproterenol infusion. In 4 preparations, however, the leading pacemaker sites shifted during isoproterenol infusion. At the leading pacemaker sites, the slopes of LDCAE are 0 AU/s, 1.47 ± 0.16 AU/s, 2.04 ± 0.18 AU/s, 2.93 ± 0.18 AU/s, 3.38 ± 0.27 AU/s, and 4.00 ± 0.21 AU/s with 0, 0.01, 0.03, 0.1, 0.3, and 1.0 μmol/l of isoproterenol, respectively (p < 0.001). The slopes of DD are 0 AU/s, 0.55 ± 0.15 AU/s, 0.95 ± 0.16 AU/s, 1.40 ± 0.15 AU/s, 1.93 ± 0.10 AU/s, and 2.28 ± 0.10 AU/s with 0, 0.01 l, 0.03, 0.1, 0.3, and 1.0 μmol/l of isoproterenol, respectively (p < 0.001). Post hoc tests showed that there were significant differences among all groups, with p values ranging from 0.000 to 0.001 for LDCAE and from 0.000 to 0.0038 for DD.
Figure 4Asummarizes the responses to isoproterenol of all preparations studied. Arrows in Figure 4A show the original leading pacemaker site (solid black dots) and the leading pacemaking sites during isoproterenol infusion (open black dots). The filled red dots indicate sites where there were no shifts. Figure 4B shows the isochronal activation maps and the corresponding Caimaps of preparation, labeled “3” in Figure 4A. The baseline activation originated in the AVN region (site a, blue color). The activation then shifted downward with 0.01 μmol/l isoproterenol (site b). Further increase of dose to 0.1 μmol/l resulted in a leftward shift to site c. Figure 4C shows that the sites of LDCAE (arrows) shifted along with the leading pacemaker site. In all 4 preparations with shifting leading pacemaker sites, the LCDAE always colocalized with the leading pacemaker site.
We gave caffeine as a 2-ml bolus (20 mmol/l) directly into the left circumflex artery in 2 preparations. Figure 5Ashows the direction of impulse propagation (arrows) along the SP and FP. Figures 5B and 5C show Caiand Vmtracings along the SP and FP, respectively, during caffeine infusion. Arrows point to LDCAE at the leading pacemaker (LP) site. Figure 5D shows actual Vmand Cairecorded at baseline and during caffeine infusion at LP, FP, and SP. The slopes of LDCAE (3.25 ± 0.14 AU/s) and DD (2.41 ± 0.05 AU/s) during caffeine infusion are significantly steeper than LDCAE and DD at baseline (0 AU/s, p = 0.020; and 0.27 ± 0.16 AU/s, p = 0.022, respectively). LDCAE (arrows) appeared in the LP site, and the mean AVJ rate increased by 108% (from 37.5 ± 4.0 beats/min to 78.3 ± 3.2 beats/min). In these 2 preparations, the LP sites did not shift locations during caffeine infusion. Similar to that seen during isoproterenol infusion, the sites with maximum slopes of LDCAE and DD always colocalized with the LP sites. The slopes decreased progressively as the recording sites moved away from the LP site (Fig. 5E).
Figure 6shows the effects of ryanodine in 1 preparation. Ryanodine 3 μmol/l markedly slowed the rate of spontaneous AVJ rhythm by 83.2%, from 37.5 ± 4.0 to 6.3 ± 6.2 beats/min (Fig. 6A). Isoproterenol infusion after ryanodine (n = 2) increased the heart rate only to 16.4 ± 6.3 beats/min, which was 56.2% less than the baseline rate before ryanodine (Fig. 6A). Figure 6B shows that ryanodine reduced the AVJ rate in a preparation receiving isoproterenol infusion. Isoproterenol did not induce LDCAE in the presence of ryanodine, and the heart rate increase was not associated with LDCAE (Fig. 6C).
The Ifblocker ZD 7288 (3 μmol/l) decreased basal AVJ rate by 35% (n = 2). The ZD 7288 did not prevent 1.0 μmol/l of isoproterenol from increasing AVJ rate by 158% (p = 0.011, compared with basal rate), accompanied by the appearance of LDCAE with LP shifted to the upper site in the AVN. There was apparent LDCAE at the LP site (Fig. 7B).
The present study shows that isoproterenol and caffeine-induced AVJ rhythm acceleration was accompanied by increased LDCAE. The isoproterenol effects were suppressed by ryanodine, but not by ZD 7288. There was colocalization of the maximum LDCAE with the LP sites in the AVJ. These findings suggest that spontaneous SR Ca release plays an important role in the mechanisms of AVJ rate acceleration during isoproterenol or caffeine infusion. A functioning Ca clock in the AVJ is important to the generation of heart rhythm when SAN is impaired or not present.
Mechanisms of AVJ rhythm
Whether or not the studies of SAN can be directly applied to the mechanisms of AVJ rhythm acceleration during isoproterenol infusion is unclear. Because the AVN and posterior extension express HCN4 (channel responsible for If), it is plausible to hypothesize that Ifis responsible for the AVJ automaticity (11). The present study shows that in addition to Ifand the membrane voltage clock, the Ca clock is also important in AVJ automaticity. Unlike SAN, the anatomic location of the AVJ pacemaker in rabbit heart is stable during autonomic modulation (12). Consistent with that finding, the LP sites in a majority of preparations in our study do not move during isoproterenol infusion. However, in 4 preparations, the LP sites do move with isoproterenol. More importantly, simultaneous Caiand Vmmapping showed that LDCAE preceded the onset of late DD and that the LP site always colocalized with the maximum slope of the LDCAE. Ryanodine was more effective than ZD 7288 in preventing heart rate acceleration induced by isoproterenol. Ridley et al. (8) isolated myocytes from the AVN of rabbit hearts and performed optical mapping of the Caitransients in those myocytes. They found that there are spontaneous Caitransients in the AVN cells. These transients ramped up the amplitude before the onset of the upstroke of the Cai. The morphology of the Cairamp in that study was the same as the LDCAE at the LP sites in the present study. The authors also reported that ryanodine inhibited the spontaneous Caitransients and reduced the rates of spontaneous activation. Our study extended their observations to the intact AVJ. The availability of simultaneous Vmsignals confirmed that these Cairamps described by Ridley et al. (8) in fact occurred before the onset of the phase 0 of the action potential. Taken together, these data indicate that SR Ca clock plays an important role in AVJ automaticity.
Optical signals collected from the canine 3-dimensional tissues represent a weighted average of the transmembrane action potentials throughout the entire canine atrial wall (13). It is possible that the little foot in front of the optically recorded action potential or Caitransients is not DD or LDCAE, but rather a strongly filtered optical recording from the deeper structures. However, as shown in Figure 4C, the site of Caielevation moves from one site to another during isoproterenol infusion, and it always colocalizes with the earliest site of activation on the isochronal map. Similarly, the DD in LP site occurs during isoproterenol infusion but not at baseline (Fig. 3C). These findings rule out a fixed deeper structure as a source of these isoproterenol-induced deflections.
Because of the complex origin of the impulse near the AVJ, it is possible that activation propagated from the SAN might contaminate the signals of the AVJ (14,15). Therefore, we trimmed away the sinus node in the present study, allowing the impulse originated from AVJ itself to propagate toward the surrounding atrial myocardium. The late DD and the first action potential upstroke observed in the AVJ region therefore represent signals initiated in the AVJ. The second upstroke in the optical signal represents the signals of a different layer (the nonpacemaking atrial myocardium) that was excited by the AVJ pacemaker. We propose that late DD and LDCAE are not affected by the activations of nonpacemaking cells in this study.
The authors thank Jian Tan, Yanhua Zhang, and Lei Lin for their assistance.
Dr. Kim is currently at the Department of Internal Medicine, Inha University Hospital, Incheon, South Korea. Supported by the National Institutes of Health/National Heart, Lung, and Blood Institutegrants P01 HL78931, R01 HL78932, and 71140; a Korean Ministry of Information and Communicationthrough research and develop support project (Dr. Joung); a Nihon Kohden/St. Jude Medical electrophysiology fellowship(Dr. Maruyama); a Korea Research FoundationGrant (KRF-2008-357-E00028)funded by the Korean Government(Dr. Choi); a Piansky Family Endowment(Dr. Fishbein); an American Heart AssociationEstablished Investigator Award (Dr. Lin); and a Medtronic-Zipes Endowment(Dr. Chen). Medtronic Inc., St. Jude Medical, and Cryocath Inc. donated research equipment. All other authors report that they have no relationships to disclose.
- Abbreviations and Acronyms
- arbitrary unit
- atrioventricular junction
- atrioventricular node
- intracellular calcium
- diastolic depolarization
- fast pathway
- hyperpolarization-activated cyclic nucleotide-gated potassium channel 4
- membrane ionic current
- late diastolic calcium elevation
- leading pacemaker
- sinoatrial node
- slow pathway
- sarcoplasmic reticulum
- transitional zone
- transmembrane potential
- Received December 24, 2009.
- Revision received February 26, 2010.
- Accepted March 30, 2010.
- American College of Cardiology Foundation
- Lakatta E.G.,
- Maltsev V.A.,
- Vinogradova T.M.
- Maltsev V.A.,
- Lakatta E.G.
- Joung B.,
- Tang L.,
- Maruyama M.,
- et al.
- Racker D.K.,
- Kadish A.H.
- Racker D.K.
- Hancox J.C.,
- Levi A.J.,
- Brooksby P.
- Wu J.,
- Olgin J.,
- Miller J.M.,
- Zipes D.P.
- Inoue S.,
- Becker A.E.
- Fedorov V.V.,
- Schuessler R.B.,
- Hemphill M.,
- et al.
- Efimov I.R.,
- Mazgalev T.N.
- Efimov I.R.,
- Fedorov V.V.,
- Joung B.,
- Lin S.F.