Emerging patterns of cardiac conduction in the ... - Wiley Online Library

3 downloads 246965 Views 428KB Size Report
Mar 18, 2005 - ventricular activation pattern from immature base-to-apex to mature apex-to-base pattern at the time of ..... Schematic illustration of trabecular structure and ventricular activation in pre- and postseptation ... Photoshop (Adobe).
DEVELOPMENTAL DYNAMICS 233:456 – 465, 2005

RESEARCH ARTICLE

Emerging Patterns of Cardiac Conduction in the Chick Embryo: Waveform Analysis With Photodiode Array-Based Optical Imaging Florence Rothenberg,1 Michiko Watanabe,2* Benjamin Eloff,1 and David Rosenbaum1

Major difficulties investigating the developing cardiac conduction system stem from that the embryonic heart is extremely small (< 2 mm) and cardiac activation is relatively rapid (< 8 msec). The objective of this study was to investigate the electrophysiology of the embryonic chick cardiac conduction system at periseptation stages with a photodiode array-based detection method of optical mapping capable of high spatial and temporal resolution. Previous work indicated that, in chicken embryos, a switch occurs in ventricular activation pattern from immature base-to-apex to mature apex-to-base pattern at the time of ventricular septation. It was our aim to map activation in more detail to identify the active pathway or pathways of atrioventricular conduction at these particular stages. Analysis of preseptated hearts (n ⴝ 10) showed that the latest atrial activation took place just above the site of the earliest ventricular activation at the ventral left ventricular base. Analysis of postseptated hearts (n ⴝ 11) showed apex-to-base conduction consistent with activation through the maturing His–Purkinje system. Evaluation of hearts during septation revealed a gradual transition of ventricular activation patterns rather than an abrupt “switch.” External pacing of preseptated hearts revealed significant slowing of interventricular conduction compared with spontaneous beats (spontaneous, 61.7 cm/sec ⴞ 9 cm/sec vs. paced, 36.5 cm/sec ⴞ 10 cm/sec). The more detailed mapping revealed that, before septation, the pattern of activation of the ventricular myocardium is consistent with direct atrial–ventricular myocardial connections at the left lateral atrioventricular junction; however, functional evidence for a preferential conduction pathway within the ventricles was present before septation. Developmental Dynamics 233:456 – 465, 2005. © 2005 Wiley-Liss, Inc. Key words: His-Purkinje system; electrophysiology Received 11 October 2004; Revised 17 November 2004; Accepted 7 December 2004

INTRODUCTION A specialized cardiac pacemaking and conduction system (Gourdie et al., 2003; Gourdie and Watanabe, 2004) is responsible for coordinating electrical activation of the heart in many species. Normal cardiac conduction produces sequential contraction of the atria and ventricles necessary for ad-

equate forward flow of blood. Clarification of the mechanisms of normal conduction system formation is critical to determine processes that go awry during development, which may result in cardiac arrhythmias postnatally (Brugada et al., 1997; Schott et al., 1998; Clancy et al., 2002; Darbar et al., 2003; Wehrens et al., 2003). An

1

understanding of these developmental mechanisms had been limited by the resolution of electrophysiologic techniques used to investigate electrical activation in small embryonic hearts. Clear action potential waveforms from embryonic cardiomyocytes have been obtained with intracellular electrodes (Lieberman and de Carvalho, 1965;

Heart & Vascular Research Center, MetroHealth Campus, Case Western Reserve University, Cleveland, Ohio Division of Pediatric Cardiology, Department of Pediatrics, Rainbow Babies & Children’s Hospital, Case Western Reserve University, Cleveland, Ohio Grant sponsor: NIH; Grant numbers: 5T32 HL07653; HL38172; HL54807; Grant sponsor: AHA; Grant number: 0365348B. *Correspondence to: Michiko Watanabe, Ph.D., Department of Pediatrics, Case Western Reserve University, School of Medicine, RB&C Hospital, 11100 Euclid Avenue, Cleveland, OH 44106-6011. E-mail: [email protected] 2

DOI 10.1002/dvdy.20338 Published online 18 March 2005 in Wiley InterScience (www.interscience.wiley.com).

© 2005 Wiley-Liss, Inc.

OPTICAL MAPPING OF EMBRYONIC CARDIAC CONDUCTION 457

Arguello et al., 1986, 1988); however, the small size of the embryonic heart limits simultaneous recording to two or at most three sites. Optical mapping using voltage-sensitive dye has made it possible to investigate patterns of cardiac activation and repolarization in mature heart models (reviewed in Efimov et al., 2004) for nearly 30 years (Salama and Morad, 1976). Recent application to the embryonic model has been instructive and has led to further questions. The first studies to use microscopebased optical imaging of embryonic cardiac conduction with a photodiode array detection system were performed with transmitted light in the chick embryo (Hirota et al., 1979). These groundbreaking investigations were limited to the analysis of young, six- to nine-somite transparent embryos (29 –33 hr of development, Hamburger-Hamilton [1951], HH, stage 8 –9) that permitted transmission of light. Investigations using extracellular recordings of older chick hearts (day 4 – 6 of development, HH stages 24 –32) showed that the emergence of the mature ventricular activation sequence in chick embryos coincided with cardiac septation (closure of the interventricular septum with the atrioventricular [AV] junction) and delineation of the cardiac conduction system by transient immunohistochemical markers (Chuck et al., 1997). These investigations of embryonic ventricular activation have been confirmed with voltage-sensitive dyes using CCD detection of the fluorescent signals (Reckova et al., 2003; Hall et al., 2004). Investigations of mouse (Rentschler et al., 2001) and rabbit embryos (Rothenberg et al., 2005), on the other hand, suggest that mammalian hearts may form functioning cardiac conduction pathways well before ventricular septation is complete. The importance of cardiac septation in the patterning of the cardiac conduction system in the chick, then, requires further investigation for several reasons. First, the chick is an important model for investigations of mechanisms of normal and abnormal heart development—it is valuable to know whether the mechanisms deduced in this system are relevant to the mammalian system. Second, the differences ob-

served between species may provide insights to events, signals, or mechanisms critical for the formation of the conduction system. In this study, we used technology that produces highfidelity activation signals from the small embryonic heart that will allow detailed physiologic investigations of the embryonic cardiac conduction system. Photodiode array-based optical mapping of cardiac activation has both high spatial (150 ␮m) and temporal (3.4 kHz) resolution. Whereas CCD detection of optical signals provides excellent spatial resolution, a limitation of previous reports using this method had been the lower sampling frequencies reported (approximately one sample/msec). This outcome can introduce error, as the activation upstroke itself is 1–2 msec in duration. Increased temporal resolution also allows visualization of the activation waveform with improved signal-to-noise ratio. Although sampling frequencies for CCD cameras have improved, data from such devices used in the embryo have not yet been published. A photodiode arraybased system was developed to investigate the emergence of the chick cardiac conduction system. This method produced clear optical activation signals simultaneously from multiple sites over the surface of the developing heart. Using this technology, we tested the hypothesis that activation of the ventricles in the preseptated chick heart occurs through direct AV muscular connections before maturation of the His–Purkinje system (HPS).

RESULTS Validation Orientation of the digital camera image and the photodiode array output were confirmed using a light-emitting diode (LED) placed on the stage of the microscope. The LED was oriented in nine positions throughout the photodiode recording field, and the photodiode array image was aligned with the digital image of the LED. The two pixels with the highest intensity of signal corresponded within half a pixel to the location of the digital image of the LED. The intensity of signal dropped precipitously to that of back-

ground noise within two pixels beyond the LED location (data not shown), indicating that far-field effects of fluorescence in the embryonic hearts were unlikely. Digital images of a reticle placed on the microscope recording chamber were taken at ⫻4 and ⫻10. The images were used to calculate and confirm the magnification of the imaged hearts with respect to the photodiode array image.

Global Activation Sequences Changed After Cardiac Septation We focused our investigation on periseptation stages, as previous results (Chuck et al., 1997) indicated that the pattern of ventricular activation undergoes a switch at this time. Preseptated hearts of stages 24 to 26 showed similar patterns of activation across the ventral surface (Fig. 1). In the ventral view as depicted, atrial activation began on the right (medial) surface of the single atrium and continued across the ventral surface toward the left lateral aspect of the atrium. Crowding of isochrones relative to ventricular activation suggested a slower rate of conduction across the atrium in the preseptated heart (Fig. 1, apparent conduction velocity in this example is 8 cm/sec). After an AV delay, the ventricles of the preseptated hearts activated earliest in the left base just below the site of the last atrial activation. From the left base beneath the AV junction, ventricular activation proceeded rapidly across the surface of the ventricle toward the right-sided outflow tract. Crowding of isochrones over the outflow tract suggested slowing of conduction, consistent with previous reports using extracellular electrode plaques (de Jong et al., 1992) and intracellular electrodes (Arguello et al., 1986). Conduction velocity of ventricular tissue (pixels E through G) was 67 cm/sec (SD, 0.15) as opposed to 39 cm/sec (SD, 0.15) in the outflow tract region (pixels H and I). In lateral views of the preseptated hearts (Fig. 2), atrial activation proceeded in a posterior-to-anterior direction. After the AV delay, the anterior left base activated first and proceeded in an anterior-to-posterior direction. In these preseptation hearts, the

Fig. 1.

Fig. 2. Lateral view of a representative stage 25 embryonic heart. A: Digital image of heart in B. B: Isochrone maps have been placed over the image of the heart. The legends are next to the maps for the atrium and ventricle. White, early activation; black, late activation.

OPTICAL MAPPING OF EMBRYONIC CARDIAC CONDUCTION 459

apex-to-base pattern of conduction was never seen. Embryonic ventricles imaged after cardiac septation (stage 30) demonstrated a different sequence of activation similar to that of the mature heart (Fig. 3). The site of earliest ventricular activation occurred either slightly to the right or left of the apex rather than the base. Activation then spread rapidly from apex to base. The last region activated was over the outflow tract, which at later stages becomes repositioned toward the center of the ventral surface of the heart. Detection of cardiac activation with a photodiode array at high sampling rates permitted visualization of distinct activation signals (Fig. 1). Signals from photodiode array pixels over atrial sites had slow onset (Fig. 1B, waveforms A–D) relative to ventricular sites, and rapid repolarization. Activation signals from ventricular sites (Fig. 1B, waveforms E–I), on the other hand, had rapid activation and prolonged repolarization. Activation signals were also obtained from photodiode array pixels situated over the AV junction myocardium (Fig. 1e). Activation times could not be accurately ascribed to these sites, as the first derivative of the activation onset was ambiguous. Upstrokes are broad, suggestive of slower conduction. A series of embryos in intermediate stages of development had neither the typical “apex-to-base” nor “base-toapex” global activation patterns seen in pre- and postseptation embryos (n ⫽ 5, stages 26 –29). In all cases, regions along the left lateral ventricle activated earliest (Fig. 4). Activation of these ventricles proceeded rapidly

in a sweeping, left-to-right manner instead of a longitudinal orientation.

Effects of Pacing Pacing of the embryonic heart was performed both to validate the detection method and to investigate physiologic parameters of the nascent conduction substrates. Conduction in 10 embryonic hearts was assayed with voltage-sensitive dyes before and during external pacing. The pacing electrode was placed on the surface of the heart and, therefore, activating the ventricular wall from the epicardial side. There was a significant decrease in the conduction velocity with pacing (Fig. 5). The average conduction velocity of the spontaneous beat (sinus rhythm) was 62 cm/sec compared with the paced beat that was 37 cm/sec. The difference in conduction velocity was significant whether the paced beat was in the same direction of the spontaneous beat or in the opposite direction of the spontaneous beat. Analysis of atrioventricular conduction times showed no difference before or after ventricular septation (Fig. 6).

DISCUSSION Activation waveforms were assayed from multiple sites in the periseptation stages of chick heart development (stages 25–32), including the atrium and AV junction. These results were consistent with and expanded previous findings using extracellular recordings at two sites (Chuck et al., 1997): that the base-to-apex activation pattern was present before ventricular septation was replaced by the

apex-to-base activation pattern after septation, and a transitional pattern of simultaneous base and apex activation was observed at stages of development in between. Detailed mapping of multiple ventricular sites uncovered a distinct focus of early excitation in the left ventral base in all preseptation embryos (n ⫽ 16). Significantly shorter conduction times during sinus rhythm (compared with paced rhythms) support the presence of a preferential conduction pathway within the ventricle. Lastly, distinct atrial activation waveforms throughout the atrium were demonstrated. Previous extracellular recordings (Chuck et al., 1997) indicated that, just before ventricular septation, impulses from the atrial myocardium may have two avenues to reach the ventricular myocardium: nonspecialized atrioventricular myocardial connections and the developing common (His) bundle. Before septation, there are atrioventricular myocardial connections (Lieberman and de Carvalho, 1965) all around the circumference of the AV junction that are capable of conducting impulses. A novel finding from the present study indicated that, despite the many potential sites where the impulse could travel across the AV junction, the first site of ventricular activation was consistently at a discrete site at the left base as assayed in the frontal and lateral surfaces of the ventricles. This ventricular site is directly below the last atrial site of activation, consistent with conduction across a ventrolateral atrioventricular myocardial connection in a region recently shown to have fibers consistent with the anatomy of conductive cells and active conducting

Fig. 1. Global activation sequence and optical signals from the ventral surface of a representative stage 25 embryonic chick heart. A: Isochrone maps constructed from activation signals detected from the epicardial surface of the atrium and ventricles. The atrial map was constructed from the same beat as the ventricular map. The atrial activation map is represented in green, isochrones are 2 msec. Light green represents early signals; dark green represents late signals. Activation was earliest over the medial common atrium and proceeded to the left lateral surface of the atrium. A total of 30 msec was required for the activation to traverse the entire atrium (see the legend to the right of the map). Ventricular activation occurred after an atrioventricular (AV) delay of 220 msec, began in the left base and proceeded toward the right (of all preseptated hearts investigated, this heart had a representative global activation sequence, but a very prolonged AV delay). Isochrones are 1 msec apart. Note that only 9 msec was required to traverse the entire ventricular surface. B: Same heart as in A. Optical signals as recorded from individual pixels are demonstrated to the right of the image. A–I: Note that atrial upstrokes (A–D) occurred over a longer period of time than ventricular upstrokes (E–I). Furthermore, a spike can be seen in the atrial signals that correspond to the ventricular upstroke. At: common atrium, which appears on the left of the embryonic heart; V, ventricle; oft, outflow tract; AVJ, atrioventricular junction; ecg, electrocardiogram. C: Atrial upstrokes superimposed to show progression of activation across the surface of the atrium (normalized for differences in amplitude). D: Ventricular upstrokes superimposed to show progression of ventricular activation across the surface of the ventricle (normalized for differences in amplitude). Atrial activation occurred over a longer period of time than ventricular activation, consistent with data in A. e: Optical signals recorded from photodiode array pixels over the AVJ. The first signal was recorded from a location beneath the asterisk, and each signal proceeds as shown by the arrow. Note the slow upstroke compared with ventricular activation signals (B, letters E through I), indistinct activation time, and blurring of atrial with ventricular signals.

Fig. 3. Postseptation activation sequence. Global activation across the ventral surface of a representative stage 31 heart. Isochrone map of activation sequence superimposed over a digital image. White, early activation; black, late activation; isochrones ⫽ 1 msec. After septation, the earliest activation of the ventricle occurred to the right of the ventricular apex and proceeded toward the base. A total of 8 msec was required for complete activation of the ventricle. Inset shows a digital image of the same heart to indicate landmark structures: OFT, outflow tract; LA, left atrium; RA, right atrium; LV, left ventricle; RV, right ventricle.

Fig. 4.

OPTICAL MAPPING OF EMBRYONIC CARDIAC CONDUCTION 461

Fig. 5. Conduction velocity of paced and spontaneous beats in the same embryonic hearts (n ⫽ 10). There is a significant reduction in conduction velocity for paced vs. spontaneous beats; the asterisk indicates a significant P value (0.00003). Pacing, therefore, resulted in a reduced conduction velocity compared with spontaneous beats (sinus rhythm).

pathways (Thompson et al., 2003; Sedmera et al., 2004). This pattern confirms that there is a preferential conduction route from the sinoatrial node through the atrium to the left base of the ventricles that overrides the use of other potential atrioventricular myocardial connections before septation. Another novel finding is the ventricular pattern of activation and conduction at the transitional stages (stages 26 –29) between the time when the base-to-apex pattern disappears and the apex-to-base becomes dominant. The transitional stages exhibited a simultaneous activation of the left lateral side at several sites. This pattern of activation would promote the ejection of blood from the left to the right ventricles through the ventricular foramen that is still open at these stages and supports a gradual change in activation pattern from base-to-apex followed by left-to-right and finally apexto-base rather than an abrupt change from base-to-apex to apex-to-base.

Is Septation an Important Landmark for Development of the Conduction System? The hypothesis that the changing ventricular activation sequence is related to ventricular septation has not been

Fig. 4. Representative example of a stage 29 embryonic heart with a “transitional” ventricular global activation sequence. These embryos have a more diffuse activation of the left lateral ventricular surface with a sweeping pattern of electrical activation toward the right ventricle. Raw activation signals are shown (unfiltered, no averaging). OFT, outflow tract; LA, left atrium; RA, right atrium; ecg, electrocardiogram.

supported in experimental studies. Reckova et al. (2003) showed that conotruncal banding causes precocious maturation of the conduction system as demonstrated by early appearance of “apex-to-base” pattern of conduction; however, histology of these preparations clearly showed a ventriculoseptal defect suggesting failure of ventricular septation. In the present work, lack of a significant change in the AV conduction time before or after septation and the transitional left-to-right pattern of activation argue against a dramatic switch from an immature pathway to the mature pathway induced by septation but rather for a gradual transition that may be explained in part by trabecular morphology.

Basis for the Change in Ventricular Activation Sequence Before optical mapping techniques, it was not clear how the “switch” from base-to-apex to apex-to-base activation of the ventricles occurred, as there were only a few points that could be resolved at one time (de Jong et al., 1992; Chuck et al., 1997). The ability to assess the activation of multiple sites across the surface of the developing heart showed that the most dramatic change took place across the surface of the left ventricle (Fig. 7). In the embryonic left ventricle, there is a 180-degree shift in the direction of activation with the completion of ventricular septation. Activation of the right ventricle, however, only shifts approximately 45 degrees after ventricular septation. This finding is consistent with changes that take place in the orientation and

structure of ventricular trabeculations in the chick ventricles (Ben-Shachar et al., 1985; Sedmera et al., 2000). Several authors have shown that the endocardial surface of trabeculae activates earlier than the adjacent compact myocardium (de Jong et al., 1992; Reckova et al., 2003), supporting the idea that internal myoarchitecture may dictate the epicardial activation patterns seen in this investigation. Anatomic studies of the trabeculations of the embryonic chick heart (Sedmera et al., 2000) and the adult Xenopus and zebrafish have shown beautifully that the orientation of the trabeculations is consistent with the direction of epicardial electrical activation (Sedmera et al., 2003, 2004). Work in the developing chick, mouse, and human show that significant remodeling of the cardiac myoarchitecture takes place throughout embryonic development (Ben-Shachar et al., 1985; Sedmera et al., 2000; Meilhac et al., 2003). This remodeling may influence the epicardial pattern of ventricular activation as shown in this and other investigations (Chuck et al., 1997; Reckova et al., 2003). Figure 7 schematically demonstrates how the pattern of trabeculations changes from the preseptation stages of embryonic development to the four-chambered structure and the reorientation of activation in the left ventricle that ensues. Before septation, trabeculations are plate-like structures that sweep from the inner curvature to the outer curvature, including the ventricular apex, where they are distributed like pages in an open book. The plate-like structures appear to be in contact with the inner

462 ROTHENBERG ET AL.

Fig. 6. Time from end of P-wave to earliest ventricular depolarization on activation waveform. The atrioventricular delay is not significantly different before or after septation. The average time from the end of atrial excitation (P-wave) to the onset of ventricular excitation (QRS) before vs. after septation was 114.7 vs. 95.3 msec; SE 68 and 18.7, respectively; P ⫽ 0.6; n ⫽ 4 in each group. NS, not significant.

surface of the ventricular myocardium from base to apex. Rapidly conducting pathways running along the endocardial surface of these plate-like structures may permit activation of the epicardial surface in the base-to-apex pattern observed. In contrast, in the adult, the trabeculations are rod-like and span the ventricular lumen in a longitudinal orientation. They do not appear to be in contact with the entire inner surface of the ventricle, unlike the preseptation plate-like trabeculae, except at the base and apex. This remodeling of the trabeculae and shift in trabecular realignment from transverse to longitudinal may produce the apex-to-base pattern of epicardial activation observed. It will be necessary to correlate activation from the internal structures throughout development as well as global activation in a three-dimensional manner that includes the luminal surface to accurately determine the functional status of the entire early developing cardiac conduction system.

Clarity of Local Waveforms With Photodiode ArrayBased Detection Methods Photodiode array-based detection of electrical activation of the periseptation chick heart revealed clear waveforms remarkably similar to action potentials observed with intracellular recordings of developing chick cardiac conduction cells (Arguello et al., 1986, 1988). Similarities with intracellular recordings include a shorter repolarization period in atrial tissues compared with ventricular tissues, the presence of a ventricular plateau during repolarization, slower activation upstrokes in cells within the AV canal,

and rapid activation of ventricular tissues. Furthermore, this is the first demonstration of cardiomyocyte activation from simultaneously sampled sites within the AV canal. Pixels overlying these cardiomyocytes image action potentials from both atrial and ventricular cardiomyocytes; features from both can be seen (Fig. 1e). The upstrokes from AV canal cells, however, are more broad and lower in amplitude than atrial cells, a feature characteristic of AV canal cells as observed with intracellular recording in similarly staged chick embryos (Arguello et al., 1986) and extracellular plaque electrodes (de Jong et al., 1992). Intracellular electrode recording has suggested that this reduced upstroke may be due to the presence of calcium and slow-sodium ionic currents (Arguello et al., 1986). High resolution optical imaging has been used to investigate the role of individual ionic currents in mature cardiac preparations (Akar et al., 2002; reviewed in Jalife et al., 2003). Based on the quality of the signals obtained by our method, it might be possible to use this method to investigate the individual contributions of emerging ionic currents in the developing embryonic heart, as has been shown with mature cardiac preparations.

CONCLUSIONS Before septation, the activation pattern of the embryonic heart is consistent with activation of the embryonic ventricle occurring through lateral myocardial connections between the atrium and ventricle rather than through a centrally localized developing HPS. Evidence for a preferential

conduction pathway within the ventricle, however, is present before septation. The dynamic and gradual change in the patterns of epicardial ventricular activation in the periseptation period suggested that remodeling of the ventricular trabeculations may play a role in this electrophysiological transition. Photodiode array-based detection of voltage-sensitive fluorescent markers of cardiac activation is a powerful method for investigation of conduction system transitions in embryonic hearts. The quality of the waveforms collected by this method suggests that this technique may be used in the future to determine transmembrane currents that emerge during conduction system development.

EXPERIMENTAL PROCEDURES Preparation Embryonic hearts from the White Leghorn chicken (Gallus gallus, Squire Valleevue Farm, OH) at stages of development before, during, and after cardiac septation (days 4 –7, or HH stages 19 to stage 32) were isolated from the embryo after cervical dislocation. The hearts were superfused in warmed oxygenated Tyrode’s solution in a temperature controlled recording–perfusion chamber (Warner Instrument Corp., Model RC-22C, 37°C– 39.5°C). A total of 98 embryos were studied. Electrocardiograms were measured before and after di-4ANEPPS application with silver wire electrodes that were placed in close proximity to the beating heart (Fig. 8). The electrodes were connected to a Gould preamplifier (Cleveland, OH), and signals were displayed on a Gould DataSYS 740 Digital Storage Oscillo-

OPTICAL MAPPING OF EMBRYONIC CARDIAC CONDUCTION 463

Fig. 7. Schematic illustration of trabecular structure and ventricular activation in pre- and postseptation chick embryos (modified from Ben-Shachar et al., 1985 and Sedmera et al., 2000). Curved lines represent trabecular structure; large arrows represent the direction of conduction across the epicardial surface. The number “1” represents latest atrial activation; “2” represents earliest ventricular activation. In preseptation embryos (⬍ stage 25), trabeculae are plate-like structures that sweep from the dorsal inner curvature to the ventral greater curvature with atrioventricular conduction occurring through the junctional myocardium. The small arrows suggest that each plate-like trabeculum activates the ventricle sequentially, producing a base-to-apex direction. Postseptation trabeculae have lost their sheet-like appearance and are cylindrical structures that traverse primarily in a longitudinal (apex-to-base) orientation with respect to the ventricles. IVS, interventricular septum; RV, right ventricle.

Fig. 8. Typical electrocardiograms from one stage 25 embryo in saline (green tracing) and after a 10-min bath in di-4ANEPPS (orange tracing) and schematic diagram of the optical recording system used in these experiments. Electrocardiogram tracings were recorded and printed on the same scale. Administration of di-4ANEPPS did not prolong conduction through the atrioventricular (AV) node. The photodiode array, digital camera, and eyepiece objectives were optically aligned. See the Experimental Procedures section for system details. CPU, central processing unit

464 ROTHENBERG ET AL.

scope. Signals were amplified either ⫻4,000 or ⫻10,000, and filtered (30 Hz high pass, 30 Hz low pass). Di-4ANEPPS (20 ␮l in oxygenated Tyrode’s solution, pH 7.4) was superfused for 10 min to ensure adequate membrane binding of the dye. Heart rate, QRS duration, and AV delay were measured before and after application of the voltage-sensitive dye, and there were no significant detrimental effects on cardiac conduction (heart rate [HR] saline, 118 msec ⫾ 38 bpm vs. HR di-4-ANEPPS, 132 msec ⫾ 40 bpm [P ⫽ 0.22]; QRS duration saline, 8.4 msec ⫾ 3.3 vs. QRS duration di-4-ANEPPS, 8.8 ⫾ 3.4 msec [P ⫽ 0.16]; AV delay saline, 86.7 msec ⫾ 23.3 msec vs. AV delay di-4ANEPPS, 96.9 msec ⫾ 24.1 msec [P ⫽ 0.27], n ⫽ 19, data includes embryos both before and after septation).

Measurements of AV Delay A subset of hearts (in which P-waves could be clearly delineated, n ⫽ 8) was analyzed by measuring the time from the end of the P-wave to the beginning of the ventricular spike.

Optical Mapping A high-resolution optical mapping system was used to record cardiac activation from 128 sites simultaneously in the embryonic chick heart. A 128pixel photodiode array (Centronics) and digital camera (Nikon Coolpix 950) were mounted onto an upright microscope (Nikon Eclipse E400; see Fig. 1). Alignment of the digital and photodiode array images was extensively validated (see below). An Oriel halogen light source was used (68831 Radiometric Power Supply). A total of 250 W was required to provide an adequate signal without saturation. Excitation light was filtered with a Chroma (Rockingham, VT) D480/30x filter, and emission light was filtered with a Chroma D605/55m filter; dichroic beam splitting was performed with the Chroma 505 DCLP beam splitter. The sampling frequency for this system was 3.4 kHz. The digital image was acquired with a ⫻4 objective; the photodiode array signal was acquired at ⫻10; images were analyzed and aligned using Photoshop (Adobe). Image superimpo-

sition was tested using graticule images at ⫻4 and ⫻10 and then scaled appropriately. At ⫻10, each pixel is 150 ␮m ⫻ 150 ␮m, detecting fluorescent signal from approximately 20 –30 embryonic cardiomyocytes. The electrocardiogram signal from the oscilloscope and the photodiode array signal were synchronized and directed to a Compaq computer designed for analysis. Analysis of the amplified signal was performed with Matlab-based software (Daqworks, 2000) designed in this laboratory. This software was used for signal averaging, conduction velocity calculation, activation map production, and action potential duration calculation. Conduction velocity was determined by analysis of vectors assigned to the isochrone conduction maps using the method developed by Bayly et al. (1998). Embryonic hearts were paced using an A-M Systems Stimulator-Isolator Model 2200 (Bioscience Tools, San Diego, CA). The amplitude and duration of the stimulation pulse was slowly increased until capture was achieved. The cycle length was slightly faster than the spontaneous heart rate, or cycle lengths of 380-275 for spontaneous heart rates between 160 and 200. Attempts at motion reduction were not made as cardiac activation occurred well before onset of muscular contraction.

Statistical Analysis Results are expressed as means ⫾ SD. Significance was determined by single-factor analysis of variance analysis and assigned as P ⬍ 0.05.

ACKNOWLEDGMENTS The authors thank Drs. Steven Poelzing, Igor Efimov, and Rodolfe Katra for helpful discussions. F.R., M.W., and D.R. received funding from the NIH.

REFERENCES Akar FG, Yan GX, Antzelevitch C, Rosenbaum DS. 2002. Unique topographical distribution of M cells underlies reentrant mechanism of torsade de pointes in the long-QT syndrome. Circulation 105: 1247. Arguello C, Alanis J, Pantoja O, Valenzuela B. 1986. Electrophysiological and ul-

trastructural study of the atrioventricular canal during the development of the chick embryo. J Mol Cell Cardiol 18:499 – 510. Arguello C, Alanis J, Valenzuela B. 1988. The early development of the atrioventricular node and bundle of His in the embryonic chick heart. An electrophysiological and morphological study. Development 102:623–637. Bayly PV, KenKnight BH, Rogers JM, Hillsley RE, Ideker RE, Smith WM. 1998. Estimation of conduction velocity vector fields from epicardial mapping data. IEEE Trans Biomed Eng 45:563–571. Ben-Shachar G, Arcilla RA, Lucas RV, Manasek FJ. 1985. Ventricular trabeculations in the chick embryo heart and their contributions to ventricular and muscular septal development. Circ Res 57:759 – 766. Brugada R, Tapscott T, Czernuszewicz G, Marian A, Iglesias A, Mon L, Brugada J, Girona J, Domingo A, Bachinski L, Roberts R. 1997. Identification of a genetic locus for familial atrial fibrillation. NEJM 337:350. Chuck ET, Freeman DM, Watanabe M, Rosenbaum DS. 1997. Changing activation sequence in the embryonic chick heart: implications for the development of the His-Purkinje system. Circ Res 81: 470 –476. Clancy RM, Askanase AD, Kapur RP, Chiopelas E, Azar N, Miranda-Carus ME, Buyon JP. 2002. Transdifferentiation of cardiac fibroblasts, a fetal factor in anti-SSA/Ro-SSB/La antibody-mediated congenital heart block. J Immunol 169:2156 –2163. Darbar D, Herron K, Ballew J, Jahangir A, Gersh B, Shen W, Hammill S, Packer D, Olson T. 2003. Familial atrial fibrillation is a genetically heterogeneous disorder. JACC 41:2185–2192. de Jong F, Opthof T, Wilde AAM, Janse MJ, Charles R, Lamers WH, Moorman AFM. 1992. Persisting Zones of Slow Impulse Conduction in developing chicken hearts. Circ Res 71:240 –250. Efimov IR, Nikolski V, Salama G. 2004. Optical imaging of the heart. Circ Res 95:21–33. Gourdie RG and Watanabe M. 2004. Excited by the cardiac pacemaking and conduction system. Anat Rec 280A:933–934. Gourdie RG, Harris BS, Bond J, Justus C, Hewett KW, O’Brien TX, Thompson RP, Sedmera D. 2003. Development of the cardiac pacemaking and conduction system. Birth Defects Res Part C Embryo Today 69:46 –57. Hall C, Hurtado R, Hewett K, Shulimovich M, Poma C, Reckova M, Justus C, Pennisi D, Tobita K, Sedmera D, Gourdie R, Mikawa T. 2004. Hemodynamic-dependent patterning of endothelin converting enzyme 1 expression and differentiation of impulse-conducting Purkinje fibers in the embryonic heart. Development 131: 581–592. Hamburger V, Hamilton H. 1951. A series of normal stages in the development of

OPTICAL MAPPING OF EMBRYONIC CARDIAC CONDUCTION 465

the chick embryo. Journal of Morphology 88:49 –92. Hirota A, Fujii S, Kamino K. 1979. Optical monitoring of spontaneous electrical activity of 8-somite embryonic chick heart. Japanese Journal of Physiology 29:635. Jalife J, Anumonwo JM, Berenfeld O. 2003. Toward an understanding of the molecular mechanisms of ventricular fibrillation. J Int Card Electrophys 9:119. Lieberman MP, de Carvalho A. 1965. The electrophysiological organization of the embryonic chick heart. The Journal of General Physiology 49:351–379. Meilhac S, Kelly R, Rocancourt D, EloyTrinquet S, Nicolas J, Buckingham M. 2003. A retrospective clonal analysis of the myocardium reveals two phases of clonal growth in the developing mouse heart. Development 130:3877–3889. Reckova M, Rosengarten C, deAlmeida A, Stanley CF, Wessels A, Gourdie RG, Thompson RP, Sedmera D. 2003. Hemodynamics is a key epigenetic factor in development of the cardiac conduction system. Circ Res 93:77–85. Rentschler S, Vaidya D, Tamaddon HS, Degenhardt K, Sassoon D, Morley GE,

Jalife J, Fishman GI. 2001. Visualization and functional characterization of the developing murine cardiac conduction system. Development 128:1785–1792. Rothenberg F, Nikolski V, Watanabe M, Efimov IR. 2005. Electrophysiology and anatomy of embryonic rabbit hearts before and after septation. Am J Physiol Heart Circ Physiol:00770.02004. Salama G, Morad M. 1976. Merocyanine 540 as an optical probe of transmembrane electrical activity in the heart. Science 191:485–487. Schott JJ, Benson DW, Basson CT, Pease W, Silberbach GM, Moak JP, Maron BJ, Seidman CE, Seidman JG. 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science (Washington DC) 281:108. Sedmera D, Pexieder T, Vuillemin M, Thompson RP, Anderson RH. 2000. Developmental patterning of the myocardium. Anat Rec 258:319 –337. Sedmera D, Reckova M, deAlmeida A, Sedmerova M, Biermann M, Volejnik J, Sarre A, Raddatz E, McCarthy RA, Gourdie RG, Thompson RP. 2003. Functional and morphological evidence for a

ventricular conduction system in zebrafish and Xenopus hearts. Am J Physiol Heart and Circ Phys 284:H1152– H1160. Sedmera D, Reckova M, Bigelow MR, DeAleida A, Stanley CP, Mikawa T, Gourdie RG, Thompson RP. 2004. Developmental transitions in electrical activation patterns in chick embryonic heart. Anat Rec 280A:1001–1009. Thompson RP, Reckova M, DeAlmeida A, Bigelow M, Stanley CP, Spruill JB, Trusk T, Sedmera D. 2003. The oldest, toughest cells in the heart. In: Chadwick DJ, Goode J, editors. Development of the cardiac conduction system. New York: John Wiley and Sons. p 157–176. Wehrens XH, Lehnart SE, Huang F, Vest JA, Reiken SR, Mohler PJ, Sun J, Guatimosim S, Song LS, Rosemblit N, Armiento JM, Napolitano C, Memmi M, Priori SG, Lederer WJ, Marks AR. 2003. FKBP12.6 deficiency and defective calcium release channel (ryanodine receptor) function linked to exercise-induced sudden cardiac death. Cell 113:829.