Dynamic patterns of apoptosis in the developing chicken heart


  • Katherine S. Schaefer,

    1. Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, Ohio
    Current affiliation:
    1. Randolph-Macon Woman's College, Department of Biology, 2500 Rivermont Avenue, Lynchburg, VA 24503
    Search for more papers by this author
  • Yong Qiu Doughman,

    1. Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, Ohio
    Search for more papers by this author
  • Steven A. Fisher,

    1. Department of Medicine, Case Western Reserve University School of Medicine, Cleveland, Ohio
    Search for more papers by this author
  • Michiko Watanabe

    Corresponding author
    1. Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, Ohio
    • Department of Pediatrics, Rainbow Babies and Children's Hospital, 11100 Euclid Avenue, Cleveland OH 44106-6011
    Search for more papers by this author


The outflow tract (OFT) is abnormal in many congenital heart defects. One critical mechanism for morphogenesis of this complex structure is apoptosis. Chicken embryos (stages 19–38; ED4–10) stained with a fluorescent supravital lysosomal dye (LysoTracker Red; LTR) revealed the three-dimensional relationship between structural changes and apoptosis. The LTR staining peaked in the OFT myocardium at stages 27–32, consistent with our previous analyses using other apoptosis assays. While LTR stained under both the pulmonary artery and the aorta, it was most prevalent in the subaortic myocardium before its elimination. Furthermore, LTR staining was most abundant in the myocardium under intensely cytokeratin-positive, thick epicardium. These data support the hypothesis that temporally and spatially restricted apoptosis in the OFT myocardium allows the aorta and pulmonary artery to dock at the appropriate angle and level with the proper ventricle. These data also support a relationship between the differentiating epicardium and cardiomyocyte apoptosis. Developmental Dynamics 229:489–499, 2004. © 2004 Wiley-Liss, Inc.


Programmed cell death (PCD) or apoptosis is a widely used developmental mechanism for removal and remodeling of tissue (Jacobson et al., 1997; Vaux and Korsmeyer, 1999). The presence of PCD has been documented previously in developing cardiac tissues by our laboratory (Watanabe et al., 1998) and others (for reviews, see Pexieder, 1975; Fisher et al., 2000; Poelmann et al., 2000; van den Hoff et al., 2000). We correlated apoptosis in the outflow tract (OFT) myocardium with shortening of the chicken OFT myocardium (Watanabe et al., 1998) and found that inhibition of apoptosis in that tissue caused defects in the shortening and the rotation of the OFT (Watanabe et al., 2001). We also determined that epicardial coverage was critical for a normal level of apoptosis in the avian OFT myocardium and that at least one component from the embryonic epicardium, QH-1–positive pro-endothelial cells, had invaded the OFT myocardium and were present as vascular tubes in areas where apoptosis levels were high (Rothenberg et al., 2002). This finding suggested a relationship between epicardial coverage of the OFT and remodeling of that region by apoptosis.

Cell death has been detected in embryos by using a variety of compounds and techniques, including supravital dyes (reviewed in Watanabe et al., 2002; Zakeri and Lockshin, 2002). The advantage of the standard supravital dyes was that they marked cell death in intact hearts. The disadvantage was that the staining could not be retained through fixation or during processing for paraffin sectioning. In this study, we used LysoTracker Red (LTR), a red fluorescent supravital dye that allowed us to identify apoptotic regions in intact hearts and to follow the stain in fixed and immunostained histological sections. LTR labels acidic compartments within the apoptotic cell itself as well as in the acidic compartments of healthy cells that are engulfing apoptotic debris (Zucker et al., 2000). This technique has been used previously to study apoptosis in the embryonic rat and mouse (Zucker et al., 1998; Dunty et al., 2001, 2002; Price et al., 2003). While the staining patterns observed using LTR resembled those detected with other apoptosis assays in cardiac tissue, LTR allowed us to detect regions of cell death at earlier stages than was detected by terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) staining in sections. In addition, this technique revealed intriguing relationships between cell death and morphogenesis that were not previously appreciated in histologic section. The LTR-staining patterns supported a role for apoptosis in the selective removal and sculpting of the OFT myocardium during docking of the roots of the aorta and pulmonary artery with the appropriate ventricular chambers. Costaining with an epicardial marker and LTR staining in whole-mount hearts also revealed a relationship between the differentiating embryonic epicardium and apoptosis in the underlying myocardium.


Ao aorta CNC cardiac neural crest-derived cell IVS interventricular septum LA left atrium LTR LysoTracker Red LV left ventricle OFT outflow tract PA pulmonary artery PCD programmed cell death RA right atrium RV right ventricle TUNEL terminal deoxynucleotidyl transferase mediated dUTP nick end labeling


Comparison of LysoTracker Red and TUNEL Staining

We compared the LysoTracker Red (LTR) staining with TUNEL staining in the same sections to validate the LTR assay as an indicator of apoptosis in the embryonic heart and to identify strengths and limitations of this technique. LTR and TUNEL staining patterns were similar with colocalization of the staining in many regions, including the OFT and even in many particles (Fig. 1). One difference was that the LTR labeled a greater number of particles than did the TUNEL. Another difference was that TUNEL staining was detected on the innermost endocardial cushion tissues of stage 27 and older embryos, but intense LTR staining was absent from these same tissues.

Figure 1.

Comparison of LysoTracker Red (LTR) and terminal deoxynucleotidyl transferase-mediated biotinylated UTP nick end labeling (TUNEL) staining. LTR-stained hearts were embedded and sectioned in paraffin and photographed for LTR. The same sections were immunostained for MF20 and photographed and subsequently stained with the TUNEL technique. A,B: In sagittal paraffin sections of a stage 29 chick heart, the TUNEL-positive staining is dark brown. The area of the proximal outflow tract (oft) enclosed in a box in A is shown at higher magnification in panels B and C. LTR and TUNEL staining patterns were similar. The myocardium, delineated by MF20 immunostaining (green in B,C), contained the highest level of TUNEL (brown in B) and LTR (red in C) staining in the area studied. A brightfield TUNEL image was merged with a fluorescent image in B. C: An LTR-positive (red) image and myocardium staining (MF20 in green) were merged. The arrows indicate three examples of points where LTR and TUNEL staining colocalized within the myocardial layer. a, atrium; v, ventricle; epi, epicardium; endo, endocardium; myo, myocardium. Scale bars = 400 μ in A, 100 μ in C (applies to B,C).

LysoTracker Red Staining of the Stages 19 and 26

At stages of heart looping (stage 19–22) LTR-positive particles were infrequent and widespread with no particularly high concentrations or foci for most regions of the heart (Fig. 2A,B). An exception was the “inner curvature,” where LTR staining was noted as early as stage 19 (Fig. 2C–E). This is the region where the heart tube folds on itself to form an acute angle and where connective tissue bridges (Nahirney et al., 2003) and processes from the proepicardial serosa first contact and begin to cover the heart surface (Hiruma and Hirakow, 1989). In sections through the inner curvature coimmunostained with anti-sarcomeric actin (Sr-1), the LTR particles were found in clusters within the myocardium nearest the epicardium.

Figure 2.

LysoTracker Red (LTR) staining of stage 19–20 hearts. A,B: In intact hearts of stage 19 (A, ventral view; B, view of the right side) up to stage 22, the LTR stained sparse, widely spaced bright red particles distributed over most of the surface of the heart. C–E: An exception to this pattern of LTR staining was detected at the inner curvature, arrow (C, stage 20 ventral view; D,E, stage 19 sagittal section). Region delineated by a white box in D is enlarged in E. The LTR (red) image and myocardium stained image (Sr-1 in green) were superimposed in E. The LTR staining in the inner curvature (arrow in E) is within the myocardial layer with a few LTR particles in the epicardial layer. at, atrium; oft, outflow tract; v, ventricle. Scale bars = 200 μ in A, 200 μ in B (applies to B,C), 100 μ in D,E.

A higher incidence of LTR-positive particles was observed in the heart at stages 22 and 23 on the right dorsal surface of the ventricle near its junction with the proximal OFT (Fig. 3A). Sagittal sections of stage 23 hearts immunostained for sarcomeric actin (Sr-1) revealed that the LTR signal was within the myocardium (Fig. 3B,C) and adjacent to the epicardial interface rather than the endocardial side. At this stage in chicken heart development, the epicardium covers this same area of the heart (Hiruma and Hirakow, 1989).

Figure 3.

A–F: LysoTracker Red (LTR) staining of stage 23 and 24 hearts. LTR-positive particles were concentrated near the base of the outflow tract (oft) on the future right ventricle (at arrows in A and C). Sagittal paraffin sections of a stage 23 chick heart with Sr-1 costaining (green) showed that the LTR signal (red) in this location was in the myocardium (B and C). By stage 24, a high concentration of LTR particles (white arrows) was present on the proximal oft visible on the ventral (D), right (E), and dorsal (F) surfaces. at, atrium; avc, atrioventricular cushions; v, ventricle. Scale bars = 200 μ in A, 100μ in B (applies as 50 μ in C), 200 μ in D (applies to D–F).

Stage 24 embryos had a concentration of LTR staining (Fig. 3D–F) along a region of the right ventricle near the OFT and on the right, dorsal surface at the proximal region of the OFT. This intense LTR staining was localized to the myocardium in sections costained with Sr-1 (data not shown).

The pattern of LTR staining in these stages, the inner curvature first and the region around the juncture of the OFT and right ventricle second, generally follows the pattern reported for epicardial coverage of the OFT (Ho and Shimada, 1978; Hiruma and Hirakow, 1989). This finding was particularly interesting in light of our previous work showing that epicardial coverage was required over the OFT for the normal level of myocardial apoptosis in that region (Rothenberg et al., 2002). To determine the relationship between the epicardial development and myocardial apoptosis, we costained intact hearts for LTR and for cytokeratin. The antibody to cytokeratin had been shown to mark the epicardial cells in quail embryos from the time that they are within the outermost layer of squamous epithelial cells, when they become mesenchymal cells or tubular structures within the connective tissue of the epicardium, and even when they have invaded the myocardium (Viragh et al., 1993; Vrancken-Peeters et al., 1995). At stage 22, a region of the dorsal right ventricle adjacent to the OFT became intensely keratin-positive (data not shown) and overlapped with the LTR-positive region at that site. By stage 24, the entire heart was covered by cytokeratin-positive epicardium, but the most intense LTR staining again was under those regions of the epicardium where cytokeratin staining was most intense and where the epicardial surface appeared bumpy (Fig. 4). These LTR and intensely cytokeratin-positive regions included the base of the right ventricle and the right proximal OFT. The LTR particles and the intensely keratin-positive cells always appeared in close proximity, but we rarely detected LTR particles within the intensely cytokeratin-positive cells themselves.

Figure 4.

Anti-keratin and LysoTracker Red (LTR) staining of stage 24 hearts. A: Stacked confocal images show that while the entire heart surface is positive for anti-keratin staining (green), the regions of particularly intense anti-keratin staining coincided with regions with the highest density of LTR particles (red particles in A, white particles in B). A is a stacked projection of 2 × 50 individual confocal sections (20-μ-thick z sections, 50 red and 50 green). B: This image is a stacked projection of 50 individual confocal sections (20-μ-thick z sections, 50 red). The LTR-positive regions (B) included the right dorsal surfaces of the right ventricle (rv), the proximal outflow tract (oft), and the distal oft (white arrow). C: The rough quilted texture of the epicardium (white arrow) was best illustrated by standard fluorescence stereomicroscopy of the whole-mount anti-keratin-stained heart. D,E: Sections in the frontal plane through these regions revealed that the anti-keratin staining (green) was most intense on the epicardial epithelium on the right side of the oft (arrows), and the LTR staining was found within the adjacent myocardium. E is a higher magnification of a subregion of D. ra, right atrium; rv, right ventricle; endo, endocardium; myo, myocardium. Scale bars = 0.5 mm in A (applies to A–C), 100 μ in E (applies at 200 μ in D).

The highest levels of LTR staining at stage 25 and 26 were evident in the intact heart within the OFT from its proximal to distal regions (Fig. 5). A heart fixed with blood-engorged epicardial vessels illustrated the location of vessels at the juncture between the OFT and the right ventricle at sites adjacent to high levels of LTR staining (Fig. 5). These vessels are present at stage 26, well before the vessels from the epicardium form their definitive connections to the lumen of the aorta, the coronary arterial ostia (stage 30; Aikawa and Kawano, 1982; Waldo et al., 1990). In contrast to the OFT, the LTR particles on the ventricles and atria were evenly distributed with no concentrations or foci. Sr-1-stained sagittal sections revealed that the LTR staining at these stages was abundant in the endocardial mesenchyme of the OFT ridges with substantial LTR staining within the myocardium at the proximal base of the OFT. Most of the ventricular LTR staining was within the epicardium with only a few particles per section within the apical ventricular myocardium.

Figure 5.

LysoTracker Red (LTR) staining of stage 25 and 26 hearts. A–C: Ventral (A), dorsal (B), and cranial (C) views of the stage 25 LTR-stained embryonic heart. D,E: Right side of the stage 26 heart LTR stained (D) and under brightfield (E). F: The left side of a LTR-stained stage 26 heart. A high concentration of the LTR signal was observed in intact hearts along the length of the outflow tract (oft) at stages 25 and 26. In stage 25, substantial LTR staining (white arrows) was present in the myocardium at the proximal base of the oft (A, ventral view; B, dorsal view with atria removed; C, view of the cranial surface with atria and distal oft removed). C: The proximal oft shown in cross-section with the distal oft removed has intense LTR staining in the myocardium along the proximal right cushion, arrow. D,F: In stage 26 embryos the LTR signal remains high at the proximal base and in the endocardial cushions of the oft. E: In a stage 26 heart, blood vessels were delineated by their blood cell content on the dorsal right side of the right ventricle and oft (black arrow). These vessels appear in the same area, but a stage after the intense LTR staining. G,H: Sagittal paraffin sections of a stage 25-chick heart with Sr-1 costaining (green) showed that the anti-keratin staining (red) was in both the epicardium and myocardium (white arrows). Keratin-positive epicardial cells were present in the mesenchyme over the oft myocardium but not in the myocardium itself (G). In contrast, elongated keratin-positive cells (H, arrows) were detected within the compact zone of the dorsal apical ventricular myocardium. Red and green images were overlaid using Spot software (G, H). at, atrium; avc, atrioventricular cushions; v, ventricle. Scale bars = 500 μ in A (applies to A–C); 500 μ in E (applies to D–F); 100 μ in G (applies at 50 μ in H).

Cytokeratin-Positive Cells Were Within the OFT Epicardium but Not in the Myocardium During Initiation of Myocardial Apoptosis

To identify which steps in the development of the embryonic epicardium correlated best with the presence of LTR-positive staining within the myocardium, LTR-stained hearts were sectioned and subsequently immunostained with anti-cytokeratin at stages 25 and 27. Cytokeratin-positive cells were detected in the myocardium of the dorsal and ventral regions of the ventricle (Fig. 5H). In contrast, cytokeratin-positive cells were not detected within the OFT myocardium in the region of LTR-positive cardiomyocytes at stage 25, even though they were abundant in the connective tissue of the subepicardium in these regions (Fig. 5G). This finding indicated that epicardial cells had undergone mesenchymal transformation and were populating the matrix of the epicardium but had not yet invaded the myocardium at the time when apoptosis was occurring in the OFT myocardium.

LysoTracker Red Stained Foci in the OFT of stage 27–35

At stages 27 through 29, LTR staining was abundant in the OFT with a concentrated region of LTR staining (Figs. 6, 7) at a proximal site and a distal site. The proximal focus of LTR staining was wedge shaped with the widest region on the ventral and right side at the base of the OFT (st. 27–29; Figs. 6, 7). When compared with the location of the great vessels connected to the OFT, the LTR staining was greater in the myocardium on the right side under the aorta (Fig. 7A–D). This finding was confirmed in transverse sections of stage 28 hearts immunostained with Sr-1 where LTR staining was present at higher levels and throughout more sections on the aortic side of the OFT myocardium (Fig. 7I,J) as well as in TUNEL-stained sections (Fig. 7H).

Figure 6.

LysoTracker Red (LTR) staining in foci at stage 27. Two LTR foci (white arrows) were visible in the OFT region of stage 27 intact hearts, one proximal and another distal (A–D). A–C: The proximal LTR band has a wedge-shaped pattern when viewed from the left side (A) and covers the base of the OFT when viewed from the ventral or right side (B,C). D,E: Sagittal paraffin sections of a stage 27 chick heart that was anti-cytokeratin stained as a whole-mount and stained with Sr-1 after sectioning (both in green) showed that the LTR signal (red, arrows) in the proximal focus was within the myocardium, while the distal focus lay above the myocardial cuff and within the epicardium and mesenchyme. E is a higher magnification of the distal LTR focus (distal arrow in D). la, left atrium; ra, right atrium; lv, left ventricle; rv, right ventricle; oft, outflow tract; epi, epicardium; myo, myocardium; v, ventricle. Scale bars = 1 mm in A (applies to A–C), 500 μ in D, 50 μ in E.

Figure 7.

LysoTracker Red (LTR) staining in foci at stage 28. A,B: Two LTR foci (white arrows) continued to be visible in the outflow tract (oft) on the left side view of stage 28 intact hearts, one proximal and another distal. Brightfield images are shown in A and C to allow comparison of the position of the blood-filled great vessels within the oft and the LTR foci. The proximal LTR band has a wedge-shaped pattern when viewed from the left side (B). C,D: In the ventral view, the proximal focus covers the area on the right side under the aorta (Ao; arrow) compared with the left side under the pulmonary artery (PA). E–G: Stage 28 whole-mount LTR (red) and MF20 (green) costained hearts (ventral view of distal outflow tract (oft). The distal LTR focus (white arrows) was above the myocardial cuff (arrowheads). H–J: Transverse sections of the stage 28 proximal oft (H) assayed by terminal deoxynucleotidyl transferase-mediated biotinylated UTP nick end labeling (TUNEL; brown staining) or costained (I,J) with LTR (red) and Sr-1(green). A higher frequency of TUNEL/LTR particles were within the myocardium on the right side of the transverse sections, corresponding to the subaortic region of the OFT. J is a higher magnification of the right side of the oft shown in I. at, atrium; la, left atrium; lv, left ventricle; ltr, LTR-stained red fluorescent image; mf20, mf20-stained green fluorescent image; rv, right ventricle; sub Ao, sub aortic; sub PA, sub pulmonic. Scale bars = 500 μ in A (applies to A–D), 200 μ in G (applies to E–G), 200 μ in H (applies to H,I), 100 μ in J.

Observation of the LTR-stained and MF20-immunostained intact stage 27 and 28 hearts revealed that the distal LTR-positive focus was just above the distal rim of the myocardial tissue (Figs. 6D,E, 7E–G). This LTR focus followed a ridge of tissue that protruded on the outer surface just distal to the myocardial cuff and the developing semilunar valves. It appeared as a band at stage 27 (Fig. 6B) and became more compact and triangular-shaped at stage 28 (Fig. 7D,E). Sections of stage 27 OFT revealed that the abundant LTR staining was located in two regions with the proximal foci in the myocardial portion of the OFT and the distal foci just distal to the myocardial rim of the OFT (Fig. 6 D,E) within the epicardium and the mesenchyme.

By stage 30, a concentration of LTR staining on the right OFT was still present just below the aorta (Fig. 8A,B). The staining was located in the epicardium and myocardium in Sr-1–immunostained transverse sections of stage 30–31 hearts (Fig. 8G,H). This region of tissue disappeared in later stages (≥ st. 32), and the most intense LTR staining shifted to the epicardial area between the aorta (Ao) and pulmonary artery (PA) and finally to the subpulmonic region (st. 31–32; Fig. 8C–F). At stage 32, the LTR signal was under the PA and the area under the Ao was indented or concave. The Ao was directly connected to the ventricle by this time and appeared to be in its final location with respect to surrounding tissues.

Figure 8.

A–F: Frontal whole-mount LysoTracker Red (LTR) staining of stage 30–32 hearts. B, D, F, and H are higher magnifications of the images in A, C, E, and G, respectively. The outflow tract region just below the aorta (Ao) and pulmonary artery (PA) had a high concentration of LTR particles. This LTR staining concentration is located on the Ao side of the outflow tract at stage 30 (A, arrowhead), then shifts from the Ao side to the PA side (C,E, arrowheads). G,H: Transverse sections of stage 31 LTR (red) -stained hearts costained with Sr-1 (green) show that the majority of the LTR staining at this stage is within the epicardial tissue (G, arrowheads) with less signal in the myocardium (H, arrows). ra, right atrium; la, left atrium; v, ventricle; I, infundibulum; B,B′, brachiocephalic arteries. Scale bars = 100 μ in B (applies to B,D,F and applies at 200 μ in A,C,E), 200 μ in G (applies at 100 μ in H).

The LTR staining of the OFT gradually subsided from that stage onward. By stage 35, the LTR staining was restricted to the great vessels with the most intense staining at the branch point of vessels. The LTR staining also delineated smaller vessels that ran diagonally over the surface of the pulmonary artery. The patterns of LTR staining observed in whole-mount embryos hearts from stage 19 to 32 are diagrammed (Fig. 9).

Figure 9.

Summary of the changing patterns of LysoTracker Red (LTR) staining in the intact embryonic heart. The dots represent sites of most intense LTR staining. at, atrium; ra, right atrium; la, left atrium; oft, outflow tract; v, ventricle; ao, aorta; pa, pulmonary artery; rv, right ventricle; lv, left ventricle. The hearts are not drawn to scale relative to each other.


The study of outflow tract morphogenesis is complicated by rapid appearances and disappearances of cardiac structures simultaneous with their relocation in three-dimensions relative to the body wall and other cardiac structures. One dramatic change is the reduction in the overall length of the OFT coincident with high levels of cell death in the avian OFT myocardium and endocardium at stages of OFT morphogenesis (stages 24–32; Pexieder, 1975; Hurle et al., 1978; Hurle and Ojeda, 1979; Watanabe et al., 1998; Cheng et al., 2002; Rothenberg et al., 2002). Our analysis showed that (1) the easily obtained LTR assay results are consistent with previous findings using other more time-consuming apoptosis detection techniques, (2) cell death in the OFT myocardium begins earlier than previously reported with TUNEL, (3) apoptosis is in the OFT myocardium at higher levels under the aorta during specific stages, and (4) epicardial thickening and differentiation was noted over OFT regions at the time of myocardial apoptosis.

Utility of the LTR Technique

The LTR assay was useful to rapidly locate regions of apoptosis within the intact developing chicken heart. The pattern of LTR staining in whole-mounted hearts and in histologic sections was similar to that detected by other apoptosis assay methods including TUNEL, confirming that LTR can be used to assay for apoptosis in the chicken embryo heart. The properties of LTR give it an added advantage over TUNEL in studying developing cardiac structures. The heart undergoes three-dimensional contortions during its morphogenesis, making it difficult to follow any particular heart structure in sections alone. Being able to follow the LTR staining within the three-dimensional structure of the heart allowed us to see relationships not previously noticed: the preferentially high levels of apoptosis in the subaortic myocardium and the relationship of the cytokeratin-positive epicardial cells to apoptosis.

LTR staining of the intact heart was more sensitive in indicating regions of apoptosis than the TUNEL staining in sections. LTR indicated apoptosis starting at stages 22–23 at the base of the OFT. In contrast, TUNEL did not allow detection of apoptosis until stages 25–26 (Watanabe et al., 1998; Cheng et al., 2002), a few stages later than detected by LTR. The higher sensitivity of the LTR analysis is probably due to the property of LTR in accumulating in acidic lysosomal compartments whether they contain condensed DNA or not. The TUNEL assay is limited to labeling only the subset of compartments with condensed DNA fragments. Another reason the LTR was more sensitive was that it was easier to see accumulations of LTR particles through the entire thickness of myocardium rather than the few TUNEL particles in 7- to 10-μm sections.

LTR also allowed detection of apoptosis in the endocardium as well as in the myocardium and epicardium, although the conditions of staining would have to be adjusted for deeper LTR penetration for later stages. Comparison of LTR staining patterns between the embryonic hearts incubated in LTR while intact in the embryo and after explantation of the beating heart into an LTR solution showed no difference in whole-mount staining patterns (data not shown). The sections of the hearts stained by incubation of the intact heart with LTR did not stain the endocardium in the older stages (>stage 30), even though there was TUNEL staining in the endocardium. Our interpretation is that the 15-min incubation in LTR used in this study was optimal for the analysis of myocardial and epicardial structures but was not adequate for LTR uptake and analysis of endocardial structures at these later stages. Longer LTR incubation times, higher temperatures, or staining by perfusion may be preferable for LTR staining of the endocardium and other luminal structures.

In our previous study, we found TUNEL-positive foci that were within the distal and proximal OFT myocardium at stage 31. In this study, we also found two foci in the OFT using the LTR staining at an earlier stage (27–29). The proximal focus of LTR staining was similar in position to what we had identified with TUNEL staining, but the distal LTR focus was distinctly different being above the distal rim of myocardium. Sections revealed the LTR-positive staining in mesenchyme at the point where the secondary heart-forming field adds myocardium to the distal OFT (Mjaatvedt et al., 2001; Waldo et al., 2001). Thus, the LTR staining pattern did differ from that detected by TUNEL in this detail. A complete analysis of apoptosis patterns may require the use of two or more assays for any particular tissue.

Preferentially High Levels of Apoptosis in the Subaortic Myocardium

Our LTR staining patterns are consistent with a higher frequency of apoptosis in the myocardium under the aortic root compared with that under the pulmonary artery root at stage 28 just as the aorta lowers itself behind the pulmonary artery onto the left ventricular chamber. This finding suggests that the disappearance of the subaortic myocardium during OFT maturation may result from preferentially high levels of cell death at this site with little or no compensating cardiomyocyte proliferation. At later stages, LTR staining was also intense under the aorta, but this time within the epicardium. The epicardial layer may also remodel by apoptosis as do the inner myocardial and endocardial layers during OFT morphogenesis. Differential removal of myocardium under the Ao could also explain why the aortic valve is at a lower cranial–caudal position than the valve of the PA, which sits on a muscular cone of myocardium. Differential apoptosis could also partially account for the rotation of the Ao with respect to the PA when the root of the Ao moves posterior to the PA as the valve of the Ao is tilted and lowered onto the left ventricle.

Relationship of the Development of Epicardium and the Appearance of Apoptosis Within the Underlying OFT Myocardium

In our previous study, delay of epicardial coverage by the implantation of a barrier in ovo caused a reduction in the level of cell death in the chicken OFT myocardium as determined by TUNEL staining (Rothenberg et al., 2002). One component from the epicardium, endothelial cells, was detected within the same region of OFT myocardium where there was a high level of cardiomyocyte cell death. These findings suggested a relationship between the embryonic epicardium and myocardial apoptosis. In the current study, the pattern of the progressive onset of myocardial apoptosis over the intact developing heart followed a few stages behind the epicardial coverage of the OFT and in immediately adjacent tissues. The epicardium begins to cover the middorsal surface of the chicken heart at stage 17 and completely encloses the myocardial surface by stage 27 in a predictable pattern (Ho and Shimada, 1978; Hiruma and Hirakow, 1989; Manner, 1992, 1993; Manner et al., 2001). We first detected substantial LTR staining in the chicken embryo heart at the dorsal inner curvature at stage 19, the site where the a connective tissue bridge from the proepicardial serosa (Nahirney et al., 2003) and villi of the proepicardial serosa first contact the wall of the myocardium a few stages earlier, around stage 17 (Manner, 1992). Intense LTR staining was next detected on the dorsal surface of the future right ventricle near the base of the OFT in intact stage 22–23 hearts. The epicardium reaches this surface of the right ventricle and wraps itself around the proximal OFT by stage 21 in chick (Ho and Shimada, 1978; Hiruma and Hirakow, 1989; Manner et al., 2001). Therefore, by the time the apoptosis is detected by LTR staining in the proximal OFT at stage 23, the epicardium has covered that region and has had time to develop and thicken.

The cells of the OFT epicardium enter the epicardial matrix by epicardial–mesenchymal transformation by stages 21–23 (Perez-Pomares et al., 1997, 1998). We found that keratin-positive cells from the epicardium had covered the OFT myocardium and were within the matrix of the epicardium, but none had entered the OFT myocardium by the time LTR-positive apoptotic regions were detected there at stages 23–25. If cells from the epicardium are involved in initiating cardiomyocyte cell death, they do not appear to require direct contact with cardiomyocytes. The epicardial cells may be initiating cardiomyocyte cell death by (1) secreting pro-apoptosis factors and (2) inducing hypoxia. Changes in the cellular and extracellular composition of the differentiating epicardium and its thickening might impede access to oxygenation from the epicardial side contributing to hypoxia within the underlying OFT myocardium. Hypoxia may serve to initiate cell death and vascular development. Another possibility is that epicardial cell differentiation and cardiomyocyte apoptosis are parallel events that are initiated by the same or different stimuli.

At present, there is no evidence that epicardial cells are secreting pro-apoptotic factors; however, there is evidence that growth factor signaling between the myocardium and the epicardium does occur (Morabito et al., 2001), comparable although different in the specifics to what occurs between the myocardium and endocardium (for reviews, see Eisenberg and Markwald, 1995; Nakajima et al., 2000). FGFs and TGFβs expressed in the myocardium have been implicated in regulating epithelial–mesenchymal transformation within the epicardium (Morabito et al., 2001). Such signals may also be involved in initiating or regulating apoptosis within the OFT myocardium.

Why is the OFT myocardium susceptible to apoptosis during epicardial coverage while the ventricular myocardium, which also becomes covered, is not? The explanation may lie in the combination of unique features of the OFT. The OFT cardiomyocytes are intrinsically different from the ventricular cardiomyocytes in gene expression, which may be related to their different sites and timing of origin. All three tissue layers of the OFT, the myocardium, epicardium, and endocardium, arise late from the mesenchyme of the secondary heart-forming field (Mjaatvedt et al., 2001; Kelly et al., 2001; Waldo et al., 2001), while much of the ventricle is derived at an earlier stage from the primary heart-forming field (for reviews, see Kelly and Buckingham, 2002; Yutzey and Kirby, 2002). The OFT myocardial tissue is not trabeculated and is lined by a thick endocardium on the inside and a thick epicardium on the outside. The epicardial coverage may start sooner and thicken more rapidly around the base of the OFT compared with other regions of the ventricles. Under these conditions, the oxygen diffusion into the OFT myocardium may be less than into the trabeculated ventricular myocardium.

Our preliminary results suggest that the chicken OFT myocardium is particularly low in oxygen tension compared with the myocardium of other heart regions at stages of OFT morphogenesis (Sugishita Y, Fisher SA, and Watanabe M, in press). In quail, vasculature was detected at stage 26–27 by using the antibody marker QH-1 for quail endothelial cells within the myocardium at the dorsal atrioventricular junction (Viragh et al., 1993). Furthermore, it has been reported that one of the first places to detect vasculature in chicken development is the bulboventricular sulcus (Rychter and Ostadal, 1971), the junction between the ventricle and the OFT where intense staining for cell death is detected by TUNEL (Watanabe et al., 1998; Cheng et al., 2002) or LTR (this study) within the myocardium. We also saw evidence for substantial growth and development of the vasculature within the epicardium beginning around the base of the OFT as early as stages 25–26 (Fig. 5E). The presence of both dying cardiomyocytes and a concentration of prevascular and vascular cells would be consistent with the tissue hypoxia simultaneously initiating cardiomyocyte cell death and stimulating the development and entry of endothelial precursors into the OFT myocardium.


Preparation of Chicken Embryos

Fertile White Leghorn chicken eggs were obtained from Case Western Reserve University's Squire Valleevue Farm and incubated in humidified air at 38°C. The embryos of stages 19–35 (Hamburger and Hamilton, 1951) were handled in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health and the institutional IACUC guidelines.

LTR Staining

Stock solutions of 1 mM LTR (Molecular Probes, Eugene, OR) in dimethyl sulfoxide (DMSO) were diluted in Hanks' balanced salt solution (HBSS) or Dulbecco's phosphate buffered saline (PBS) to yield a final concentration of 2.5 μM. The embryos were dissected to expose the heart and incubated in the LTR solution for 15 min at room temperature (22°C) while rocking and washed in PBS before fixation. Preliminary experiments were carried out by using 1, 2.5, or 5 μM LTR for 5-, 10-, or 15-min incubations. Staining was observed with all concentrations with the best resolution and least amount of background interference achieved by using 2.5 μM LTR for 15-min incubations.

The embryos were fixed in neutral, buffered formalin solution (10%, Ted Pella, Inc., Redding, CA) for 1 hr at room temperature while rocking, washed extensively with PBS, and stored at 4°C before observation and further processing.

Histologic Analysis

A subset of the embryos were processed by standard procedures for paraffin embedding and sectioning. Serial sections (7-μm thick) were collected on Superfrost Plus slides (Fisher Scientific, Pittsburgh, PA). For certain procedures, embryos were processed for frozen sectioning. Briefly, they were incubated in sucrose solutions for cryoprotection and frozen in TBS tissue freezing medium (Fisher Scientific). Frozen sections (10–20 μm) were also collected on Superfrost Plus slides.

Paraffin and frozen sections were observed and photographed before further treatment to document the LTR staining pattern because subsequent procedures reduced the LTR signal. To delineate the myocardium, specific sections were stained with MF20, a mouse monoclonal antibody specific to myosin heavy chain, or Sr-1, an antibody to sarcomeric actin (DAKO M0874, Carpinteria, CA). The MF20 developed by Dr. D. Fishman was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. The secondary antibody for Sr-1 was Alexa fluor 488–conjugated goat α-mouse IgM (Molecular Probes).

Select LTR-stained and formalin-fixed intact embryos were stained for MF20 (van den Hoff et al., 1999). Embryos with their hearts exposed were incubated at 4°C overnight in 1:4 DMSO:methanol solution, rehydrated in a graded alcohol series of 95%, 70%, and 50% methanol for 1 hr each at room temperature, incubated in PBS with 10% normal goat serum (NGS) for 1 hr to block nonspecific staining, incubated with a 1:200 dilution of MF20 overnight at 4°C, washed three times for 30 min in PBS with 10% NGS, incubated for 3 hr at room temperature with a goat anti-mouse IgG antibody labeled with Alexa Fluor 488 (Molecular Probes), washed with PBS, and observed with a fluorescence stereomicroscope or by confocal laser scanning microscopy.

For keratin staining of intact hearts, embryos were processed by using a modification of a previously described protocol (Vrancken-Peeters et al., 1995). LTR-stained embryos with hearts exposed were fixed 1 hr in 10% formaldehyde solution, blocked in 10% NGS and 0.05% Tween-20 in HBSS for 1 hr, incubated overnight at 4°C in 1:250-fold dilution of polyclonal anti-keratin antibody (DAKO Z0622) in blocking buffer, washed in PBS then 3 hr in 1:200 goat anti-rabbit conjugated to Alexa fluor 488 (Molecular Probes), and observed and photographed under a conventional fluorescence stereomicroscope or by using a laser scanning confocal microscope. For sections, stained embryos were fixed in formaldehyde solution and processed for paraffin embedding and sectioning. Alternate sections were collected on separate slides and prepared for immunostaining or observed after removal of paraffin by fluorescence microscopy without further staining.

In selected paraffin sections, the presence of apoptosis was confirmed by using the TUNEL technique. Deparaffinized sections were treated with 2% hydrogen peroxide to quench endogenous peroxidase and stained with the ApopTag peroxidase kit (Intergen, Purchase, NY) using the TUNEL technique (Gavrieli et al., 1992) to identify concentrations of DNA fragments characteristic of apoptosis. Briefly, sections were incubated with proteinase K (20 μg/ml, for 15 min at 25°C), exposed to the TdT (terminal deoxynucleotidyl transferase) enzyme containing reaction buffer to add dUTP-digoxigenin to the 3′-OH end of DNA fragments and incubated with anti-digoxigenin conjugated to peroxidase.

Sections processed with the TUNEL technique were rinsed with PBS and incubated at room temperature for 1 hr in MF20 or Sr-1. The MF20 was diluted with 10% NGS in PBS to a concentration of 1:500. The sections were washed in PBS and incubated in goat anti-mouse IgG Alexa Fluor, 1:250, for 1 hr at room temperature, rinsed, and cover-slipped in Gelvatol solution (Airvol 523 polyvinyl alcohol, Air Products and Chemicals, Inc., Allentown, PA) or glycerol with 2% n-propyl gallate.

Whole-mount embryos were observed by using a Leica MZFLIII stereomicroscope or a laser scanning confocal microscope (Zeiss LSM 410; Axiovert 100; Zeiss 5× Plan-Neofluor; WD 13.5 mm; excitation lines 488 nm for green and 568 nm for red). For the projected image in Figure 4, an overlay was made of two projected images by using the red and green filters. Each projected image was the combination of 50 slices of 66 microns per slice. The embryo sections were observed by using a Leica DMLB microscope. Digital images were obtained by using a Spot RT digital camera and software v3.1 (Diagnostic Instruments, Inc., Sterling Heights, MI). Images were imported and brightness adjusted and, when necessary, overlaid by using Adobe Photoshop 6.0 or 7.0.1. All observations reported were observed in tissues from at least three different embryos.


The authors thank Erica Sieverding, Carrie Jobe, Yaling Qiu, Ankita Patel, Florence Rothenberg, MaryAnn Pendergast, and David Lawrence for technical advice and/or assistance at critical phases of the study.