Formation of the Building Plan of the Human Heart

Abstract

HomeCirculationVol. 123, No. 10Formation of the Building Plan of the Human Heart Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissionsDownload Articles + Supplements ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toSupplemental MaterialFree AccessResearch ArticlePDF/EPUBFormation of the Building Plan of the Human HeartMorphogenesis, Growth, and Differentiation Aleksander Sizarov, MD, Jing Ya, MD, PhD, Bouke A. de Boer, MSc, Wouter H. Lamers, MD, PhD, Vincent M. Christoffels, PhD and Antoon F.M. Moorman, PhD Aleksander SizarovAleksander Sizarov From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author , Jing YaJing Ya From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author , Bouke A. de BoerBouke A. de Boer From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author , Wouter H. LamersWouter H. Lamers From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author , Vincent M. ChristoffelsVincent M. Christoffels From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author and Antoon F.M. MoormanAntoon F.M. Moorman From the Heart Failure Research Center (A.S., B.A.d.B., V.M.C., A.F.M.M.) and Tytgat Institute for Liver and Intestinal Research (J.Y., W.H.L.), Academic Medical Center, Amsterdam, Netherlands; and Department of Histology and Embryology, Shanxi Medical University, Taiyuan, People's Republic of China (J.Y.). Search for more papers by this author Originally published15 Mar 2011https://doi.org/10.1161/CIRCULATIONAHA.110.980607Circulation. 2011;123:1125–1135The intricate process of formation of the heart relies on the spatiotemporal regulation of differentiation and growth of its different parts. Errors in these processes can result in hypoplasia of the chambers and incorrect alignment of the atria and great arteries with the ventricles, with such abnormalities producing the worst forms of congenital heart disease.1 Although recent advances in molecular embryology have greatly improved our understanding of normal cardiac development, our insights into the pathogenesis of human cardiac malformations remain limited.2 With the current lack of data on gene expression in humans, inferences concerning morphogenesis rely on classic studies of human cardiac development or on extrapolation of experimental data from animal studies. Most textbooks on human embryology describe the embryonic heart tube as already possessing a linear array of the definitive components, despite this so-called segmental concept being disproved some time ago by fate-mapping experiments in chickens,3,4 which are now endorsed by recent molecular lineage analyses in mice.5Proliferation studies in chickens revealed that the primary heart tube proliferates minimally and that its growth occurs through addition of differentiating cells from visceral mesoderm at the venous and arterial poles.6 These findings underscore previous lineage studies performed in mice, which have shown that transcription factors Islet1 and Tbx1 play a crucial role in the intricate balance between the cardiac precursor state and differentiation.7,8 Subsequent to its formation, the heart tube elongates and loops, producing its inner and outer curvatures and providing the general building plan for formation of the chambers and conduction system.9 The chambers themselves form by local proliferation and differentiation at the outer curvature.10,11 At the same time, the differentiating chamber myocardium remains flanked by primary myocardium, which is prevented from further differentiation by the transcriptional repressors T-box factors 2 and 3 (TBX2 and TBX3), thus providing the precursors of the cardiac conduction system.9Because it is impossible to perform experiments in the developing human heart and because of the scarcity of gene expression data, the descriptions of its development remained controversial. In this study, we report patterns of proliferation and gene expression in tubular and chamber-forming stages of the human heart, supplementing our account with 3-dimensional (3D) analyses presented in interactive fashion. Our findings show the comparability of the mechanisms governing cardiac development in humans, mice, and chickens.MethodsHuman EmbryosWe used human embryos ranging from Carnegie stages 9 to 16 derived from 2 different collections. Acquisition and preparation of the first group of embryos were performed at the Gynecology Department of Tartu University Hospital, Tartu, Estonia. To extend the number of embryos in this study, we have included a series of previously performed stainings of serial sections of embryos, which were collected from medically induced abortions performed at the Gynecology Department of the First Hospital of Shanxi Medical University, Taiyuan, China. As soon as feasible, the aborted tissues were fixed in 4% paraformaldehyde solution. Embryos were then examined under a stereomicroscope, investigated for gross anomalies, and graded according to the Carnegie criteria. After further fixation overnight, embryos were dehydrated in graded ethanol series, immersed in butanol, and embedded in Paraplast. We included only embryos considered normal. Collection and use of the human embryonic material were approved by the medical ethical committees of the universities of Tartu, Estonia; Amsterdam, Netherlands; and Shanxi Medical University, Taiyuan, China.Immunohistochemistry, In Situ Hybridization, and 3D ReconstructionsImmunofluorescent staining, in situ hybridization, and morphological and quantitative 3D reconstructions were performed on the first group of embryos as detailed in the online-only Data Supplement. Sections of the second group of embryos were processed by indirect immunohistochemistry. For a detailed description of the protocols and specification of the antibodies and riboprobe, see the online-only Data Supplement.Limitations of Our StudyOur study has several limitations owing to the use of human embryos. Fixation could not be entirely standardized. The limited number of available embryos did not permit complete optimization of the staining protocol for some antibodies and probes and prevented assessment of biological variation between specimens at similar developmental stages. Immunohistochemical and in situ hybridization stainings, nonetheless, proved to be reproducible. Because incorporation of bromodeoxyuridine cannot be performed in human embryos, proliferative status was evaluated indirectly by immunohistochemistry for Ki67 antigen and proliferating cell nuclear antigen (PCNA),12,13 the expression of which correlates well with bromodeoxyuridine incorporation.14 The scarcity of young human embryos allowed only the full quantitative reconstruction of Ki67 expression in 1 stage-10 embryo, whereas systematic samples of sections were available in 3 stages to determine the PCNA labeling index of the whole heart. Although these observations are strictly an n=1 observation, a statistical test based on the number of counted nuclei (Z test between proportions) showed that the observed differences in labeling indices exceed the methodological variation. The use of immunohistochemically stained serial sections permits 3D molecular analysis of the structures, but the corollary is that reconstructions have to be corrected for distortions of the surface because of uneven stretching and occasional damage of the sections. Although very fine details could not be reproduced in the 3D models, no spurious structures were gained or important details lost during the correction procedure.ResultsThe reader is encouraged to read the results along with the interactive 3D PDF file in the online-only Data Supplement.Formation and Growth of the Primary Heart TubeFirst we analyzed the formation and growth of the primary heart tube in 1 stage-9, 4 stage-10, and 3 stage-11 embryos, which correspond to ≈20 to 25 days of human development and embryonic days 8.0 through 9.0 in the mouse. Figure 1 shows a 3D overview of the formation and looping of the myocardial tube. In the stage-9 embryo, the youngest of our series, the heart tube is not yet formed. At the cranial end of the embryo, the myocardialized part of the walls of the bilateral intraembryonic coelomic cavities are already fused, with the fusion line4 recognizable by a characteristic groove in the midline of the myocardial trough (Figure 1A and 1A*). The borders of the myocardium are contiguous with the coelomic wall. The 2-cell-thick myocardial layer of the forming heart tube is separated by acellular jelly from the endothelial cells lining 2 narrow lumens. The cardiac jelly extends outside the cardiac trough dorsally, where the ventral wall of the foregut borders the cardiac trough (Figure 1A**). In late-stage-10 embryos, the heart tube is closed, and the tubes show various stages of looping. At this stage, the heart tube contains a single and wider lumen surrounded by jelly, which still extends outside the myocardial tube (Figure 1B, 1B*, and 1B**). In stage-11 embryos, the heart tube is almost completely looped (Figure 1C and 1C*), and the first signs of formation of the atrial and ventricular chambers are seen at the dorsal and ventral sides of the tube, respectively (Figure 1C**). Regionally, the heart tube has expanded and the myocardial layer has thickened, although trabeculae have yet to form. Interiorly, the heart tube is still filled by acellular jelly, which separates the myocardial and endothelial layers. As yet, no epicardial layer covers the outer surface of the myocardial heart tube.Download figureDownload PowerPointFigure 1. Formation and looping of the human embryonic heart. At stage 9 (A through A**), the heart tube (ht) is just forming. At stage 10 (B through B**), the heart tube begins to loop, and at the right side of the venous pole, the forming atrium is already discernable. At stage 11 (C through C**), the heart tube is elongated and looped, and it is possible to distinguish the morphological landmarks of the primordia of the future cardiac chambers, as well as the inner and outer curvatures (white and black dotted lines in C**). At shown stages, myocardium at the venous and arterial poles is directly contiguous with the coelomic wall (black arrows). aip indicates anterior intestinal portal; ao, (dorsal) aorta; avc, atrioventricular canal; cc, coelomic cavity; fg, foregut; OFT, outflow tract; PV, primitive ventricle; and v, vein. The scale bars apply to the dorsal and ventral views.To provide insight into the precursor, differentiation, and proliferative status of the primary heart tube and adjacent mesoderm, we stained sections with ISL1, NKX2-5, troponin-I, and the proliferation markers PCNA and Ki67 (Figures 2 and 3). In the stage-9 embryo, the domains of the NKX2-5 and ISL1 expression in the forming heart, adjacent mesoderm, and foregut are largely overlapping. Weak ISL1 expression in the myocardial trough (Figure 3A and 3B) indicates the recent differentiation of the ISL1-positive precursors toward cardiomyocytes. At stages 10 to 11, the nondifferentiated (ie, troponin I–negative) mesodermal cells at the venous and arterial poles, as well as cells of the dorsal coelomic wall and ruptured mesocardium, coexpress NKX2-5 and ISL1, which suggests that these cells are the progenitors to be added to the heart tube. In contrast, most of the myocardial heart tube is negative for ISL1, with the exception of the venous and arterial poles, where strong myocardial expression of ISL1 is present (Figure 2). All myocardial cells express NKX2-5, indicating that the sinus venosus, characterized by absence of this factor,15 is not yet formed at these stages. The level of expression of the proliferation markers PCNA and Ki67 in the myocardial heart tube at stage 10 was much lower than in surrounding tissues (Figures 2 and 3). The coelomic wall contiguous with the myocardium at the venous and arterial poles highly coexpresses proliferation and progenitor cell markers, the expression of which tapered off in the heart.Download figureDownload PowerPointFigure 2. Growth of the primary heart tube (ht) at the venous and arterial poles. Sections of embryos at stages 10 through 12 were incubated with antibodies as indicated; no proliferation data are available for the stage-9 embryo. Note the sharp border of myocardial (troponin I–positive) to nonmyocardial coelomic wall at both poles of the heart tube. At these borders, the expression of ISL1 (A, C, I, K, M, O), proliferating cell nuclear antigen (B, D, J, L, N, P), and Ki67 (E, G) is almost overlapping, being high in the nonmyocardial cells and decreasing along the myocardial wall of the inflow and outflow tracts (OFT) (arrows). The expression patterns of Ki67 and proliferating cell nuclear antigen are similar (B, E, D, G), and the number of phosphohistone H3 (PH3)–expressing cells is very low, even in highly proliferating regions (F, H). Bars=100 μm. ao indicates (dorsal) aorta; as, aortic sac; cc, coelomic cavity; fg, foregut; LV, left ventricle; RV, right ventricle; st, septum transversum; and SV, sinus venosus.Download figureDownload PowerPointFigure 3. Formation and growth of the primary heart tube (ht) during stages 9 through 11. Transverse sections, which correspond to the cross sections of the 3-dimensional (3D) models shown in Figure 1, were incubated with antibodies as indicated. In the stage-9 embryo (A through C), the domains of expression of NKX2-5 and ISL1 in the forming heart and foregut (fg) are largely overlapping, although expression of ISL1 in the myocardium is very weak. Compared with the expression of troponin I, the expression of NKX2-5 extends into the adjacent coelomic wall. At the next stages, most of the cardiomyocytes, marked by the expression of NKX2-5 (D, J), do not express ISL1 (E and K), irrespective of their proliferative status. Importantly, the linear myocardial tube contains very few cells positive for Ki67 or proliferating cell nuclear antigen (F and G). The quantitative reconstruction of Ki67 expression in a late-stage-10 embryo (H and I) shows a striking discrepancy of the percentage of the Ki67-positive cells in the myocardium compared with the coelomic wall, indicating that the vast majority of the cardiomyocytes at this stage are quiescent. Note the local increase in proliferative activity of the myocardial tube of this particular embryo (arrowheads in H). L shows the section through the looped heart of the stage-11 embryo, where the inner curvature (dotted line) is negative for proliferating cell nuclear antigen, whereas the walls of the primitive ventricle (PV) and atrium show reinitiation of proliferation (green arrowheads). Note that cells making up the foregut and neuroectoderm (ne) are strongly positive for proliferating cell nuclear antigen and Ki67. Yellow arrows in L point to the developing epicardium. Bars=100 μm. ao indicates (dorsal) aorta; cc, coelomic cavity; and st, septum transversum.To quantify and visualize the proliferation of the primary heart tube and its adjacent mesoderm, a quantitative reconstruction of Ki67 expression in a single embryo of late stage 10 was made (Figure 3H and 3I). From this reconstruction, one can appreciate the significant contrast in proliferative status of the myocardium and adjacent mesoderm; on average, only 33% of the cardiomyocytes versus 84% of the cells within the coelomic wall are positive for Ki67. At the end of stage 10, approximately half of the cardiomyocytes within the outer layer of the forming outer curvature of the cardiac tube are positive for Ki67 and PCNA (Figure 3H). At stage 11, expression of proliferation markers is confined to the myocardium of the forming ventricular and atrial chambers (arrowheads in Figure 3L). The PCNA labeling index displays an increase in overall proliferation rate from stage 10 to 12 (Figure 4). Ki67- and PCNA-stained sections of other embryos confirm these quantitative observations. Because the peak level of PCNA is reached around the early S-phase of the cell cycle,14 the observed PCNA labeling is lower than the average Ki67 labeling index. Phosphohistone H3-[Ser10] expression16 showed a low number of M-phase nuclei, although perhaps slightly more at the venous pole than at the arterial pole (Figure 2F and 2H).Download figureDownload PowerPointFigure 4. Quantitative assessment of the overall increase of the proliferating cell nuclear antigen (PCNA) labeling index of the myocardial heart tube at stages 10 through 12 (1 embryo per stage; the error bars indicate the standard error due to binomial sampling).Growth and Differentiation of the Cardiac ChambersNext we analyzed the 3D morphology, proliferation, and expression of molecular markers in chamber-forming hearts of 8 embryos of stage 12 to 13, 10 embryos of stage 14 to 15, and 6 embryos of stage 16, which corresponds to ≈26 to 38 days of human development and embryonic day 9.5 through embryonic day 12.5 in the mouse. The inner and outer curvatures of the chamber-forming heart display significant differences in the patterns of proliferation and gene expression. At the outer or greater curvature, atrial and ventricular chambers differentiate and balloon out from the cardiac tube dorsally and ventrally, respectively. The inner curvature is the point of continuity between the primary myocardium of the floor of the common atrium, the atrioventricular canal, and the proximal outflow tract (Figure 5A through 5C). Overt signs of differentiation toward chamber myocardium were first observed at stage 12 at the outer curvature of the completely looped heart tube. Now tiny trabeculae, covered by a markedly thinner layer of acellular jelly than at the smooth-walled inner curvature, are formed in the left ventricular wall (Figure 5A*). From stage 14, when septation has started, the expansion of the chambers progresses significantly (Figure 5B), and ventricular trabeculae become extensive (Figure 5B*). Subsequent to the local reinitiation of proliferation of the myocardial walls of the future atrial and ventricular chambers in the looping heart at late stage 10 (Figure 3H and 3L), the proportion of proliferating cardiomyocytes at stage 12 increases substantially (Figure 4). Later on, atrial and ventricular chamber myocardium of stages 12 through 16 show high PCNA expression (Figure 6), although proliferation in the ventricular trabeculae declines by stage 16 (Figure 6C*). Similar to the previous stages, the expression of phosphohistone H3 was observed only in a few cells throughout the sections (not shown).Download figureDownload PowerPointFigure 5. Chamber formation in the heart of embryos during stages 12 through 16. The expansion of the chamber cavities occurs through ballooning of the walls out of the outer curvature of the looped heart tube (A, B, C). Atrial chamber formation is marked by high expression of atrial natriuretic factor (ANF) (A*, B*, C*), and formation of the ventricular chambers is characterized by a gradual thickening of the compact layer and by appearance of extensive trabeculae, which are positive for ANF and connexin 43 (A**, B**, C**). The inner curvature, atrioventricular canal (avc), and outflow tract (OFT) are devoid of these markers. Bars=200 μm. 3D indicates 3-dimensional; fg, foregut; LA, left atrium; lsh, left sinus horn; LV, left ventricle; RA, right atrium; RV, right ventricle; and SV, sinus venosus.Download figureDownload PowerPointFigure 6. Growth of the chamber-forming and septating heart analyzed by proliferating cell nuclear antigen (PCNA) expression. PCNA expression in transverse sections of embryos at stages 12 through 16 is shown at the atrial (A, B, C) and ventricular (A*, B*, C*) level. The walls of the ballooning atrial and ventricular chambers are highly positive for PCNA, indicating their intensive growth, whereas the walls of atrioventricular canal (avc), inner curvature (asterisk in A*), and outflow tract (OFT) are virtually negative for this proliferation marker. In addition, the ventricular trabeculae at stage 16 are devoid of PCNA (C*). Another interesting observation is the absence of proliferation in the vestibular spine (B). Bars=200 μm. LA indicates left atrium; lsh, left sinus horn; LV, left ventricle; RA, right atrium; RV, right ventricle; and SV, sinus venosus.Along with the morphological changes marking the formation of the chambers, there are several molecular markers, such as atrial natriuretic factor (ANF) and fast-conducting connexins 40 and 43, which become expressed in the developing working myocardium in the mouse.10 In the early chamber-forming heart of the stage-12 embryo, the atrial wall and ventricular myocardium expressed ANF and connexin 40 (Figures 5A* and 7E). At later stages, the expression of ANF and connexin 40, although remaining very strong in the atrial myocardium, became confined to the trabecular myocardium in the ventricles, as reported previously for the human fetal heart17,18 (Figures 5 and 7). Because the antibody staining for connexin 43 was unsuccessful, we investigated the expression of this gene at the mRNA level by in situ hybridization. The expression of connexin 43 was observed in the ventricular trabeculae and was absent from the atrial and ventricular walls at least until stage 16 (Figure 5), as reported previously.18 In contrast, the nonworking myocardium of the atrioventricular canal and outflow tract was devoid of ANF, connexin 40 and 43, and PCNA expression (Figures 5 to 7). Remarkably, the mesenchyme of the vestibular spine, a structure supposed to play a major role in completing atrial septation,19 does not proliferate, whereas the contiguous extracardiac mesenchyme displays high PCNA expression (Figure 6C).Download figureDownload PowerPointFigure 7. Molecular analysis of the non–chamber-forming parts of the heart. At stage 12 (A through C), the myocardium of the inner curvature (dotted line in E) and atrioventricular canal (avc) expresses TBX3 but not connexin 40 (arrowheads in D and E). In contrast, the walls of the developing chambers express connexin 40 and no TBX3. In the stage-14 embryonic heart, the chambers have been expanded substantially. As in the previous stages, the connexin 40–negative regions of venous sinus and atrioventricular canal flanking the chamber myocardium express TBX3 (G through K), marking the regions where the pacemaker and conduction system develops. When assessed by routine histological staining (hematoxylin-azophloxin [H&A]), the cells of the atrioventricular canal and ventricular wall look similar (F and L). Bars=200 μm. LA indicates left atrium; LV, left ventricle; OFT, outflow tract; pas, primary atrial septum; pos positive; pr.f, primary foramen; RA, right atrium; RV, right ventricle; rvv, right venous valve; and SV, sinus venosus.Expression of TBX 2 and TBX3 in the Non–Chamber-Forming MyocardiumAs in mice and chickens,6,11,20 the myocardium of the atrioventricular canal and outflow tract flanking the ballooning atrial and ventricular chambers does not proliferate, as assessed by PCNA expression (Figure 6), and is devoid of chamber markers (Figures 5 and 7), underscoring the hypothesis of the persistence of primary myocardium at these specific locations.21 Expression of the transcriptional repressor and conduction system marker TBX322 was first observed in the floor of the common atrium and in the atrioventricular canal myocardium at stage 12 (Figure 7D), when chamber formation is well under way. Subsequent to the initiation of septation at stage 14, TBX3 expression in the atrioventricular canal becomes asymmetrical, being stronger at the right side (Figure 7J). Weak expression of TBX3 was also observed in parts of the myocardial systemic venous sinus and the primary atrial septum. Expression of TBX2, which also represses chamber formation,23 is similar to that of TBX3, although with some differences. Unlike TBX3, TBX2 is expressed in the outflow tract. In the atrioventricular canal, the expression of TBX2 observed at stage 12 has gradually disappeared at stage 16 (Figure 8). As in the mouse,20 expression of TBX2 in the atrioventricular canal tapers off into the left ventricular direction (arrowheads in Figure 8A). The expression pattern of connexin 40 (Figure 7E and 7K) is complementary to that of TBX3, except for the atrial septum and vestibular spine, indicating conservation of the repressive action of TBX3 in mice22 and humans. The general histology and cellular morphology of the myocardium in the atrioventricular canal and ventricular chamber wall at stages 12 through 14 (Figure 7F and 7L) are similar despite their distinct molecular signature.Download figureDownload PowerPointFigure 8. Expression of TBX2 in the chamber-forming and septating heart. For description, see text. Bars=200 μm. avc, atrioventricular canal; LA, left atrium; LV, left ventricle; OFT, outflow tract; pas, primary atrial septum; pulm.v.o., pulmonary venous orifice; RA, right atrium; and SV, sinus venosus.DiscussionThe mechanisms of development of the 4-chambered heart from a primary linear heart tube have remained a fertile topic for research and debate for more than a century. Classic studies, which were to date the best original works describing early human heart development,24–26 relied on the evaluation of histologically stained serial sections and 3D reconstructions. Lacking the availability of molecular markers and genetic lineage tools, classic embryologists working on human heart development had to draw their conclusions on morphogenetic mechanisms underlying growth and differentiation only on the basis of changes in visible anatomic landmarks and cellular morphology. Constrictions visible on the external surface of the early heart tube were interpreted as the borders of segments representing the compartments of the septating heart.26–28 However, fate-mapping and lineage studies in chickens and mice have demonstrated that the linear heart tube contains little more than the precursors for the definitive left ventricle.3–7 In humans, molecular data allowing comparison with animal studies on cardiogenesis are scarce. We now report an analysis of morphogenesis, growth, and differentiation of the human embryonic heart at key stages of its development, allowing such a comparison. The significance of our observations is 2-fold. First, our observations support a biphasic model of early cardiac morphogenesis, with the myocardial tube initially recruited from the highly proliferative adjacent coelomic wall and then, after a phase of quiescence, the chambers forming by local reinitiation of proliferation and differentiation. Second, they show that the mechanisms underlying the formation, growth, and differentiation of the primary heart tube and the chambers as seen in mice and chickens are highly conserved in humans.The Early Embryonic Heart Grows Through Proliferation and Differentiation of Cardiac Precursor CellsThe primary linear heart tube is formed through myocardialization and subsequent fusion of parts of the bilateral walls of the intraembryonic coelom.4 Although in 1933 Patten and Kramer29 had demonstrated experimentally that the early embryonic chicken heart does not contain all of its definitive parts, embryologists studying human development continued to assume that the primary myocardial heart tube elongated by growth of its initially present segments.27,28 Later, Sissman30 demonstrated that the early myocardial heart tube in chicken embryos had very low proliferative activity, and, more recently, lineage studies in chickens and mice provided unequivocal evidence of addition of cardiomyocytes from the second heart field to both poles of the heart tube.5,6 The transcription factors Islet1 and Tbx1 were shown to be crucial for the maintenance of cells in the dorsal coelomic wall in the precursor state before their differentiation into cardiomyocytes and addition to the growing heart tube.7,8Our current observations that ISL1 and the proliferation markers PCNA and Ki67 are coexpressed in the cells of the coelomic wall contiguous to the myocardium at the venous and arterial poles suggest a similar role of ISL1 in the growth of the heart tube in human and mouse. The myocardium of the forming heart tube in the youngest embryo of our series still expressed ISL1, supporting a role of ISL1 as cardiac progenitor cell regulator within both the primary and secondary heart fields.31 The proliferating myocardium of the expanding chambers was nega