Developmental pattern of ANF gene expression reveals a strict localization of cardiac chamber formation in chicken.код для вставкиСкачать
THE ANATOMICAL RECORD 266:93–102 (2002) DOI 10.1002/ar.10042 Developmental Pattern of ANF Gene Expression Reveals a Strict Localization of Cardiac Chamber Formation in Chicken ARJAN C. HOUWELING, SEMIR SOMI, MAURICE J.B. VAN DEN HOFF, ANTOON F.M. MOORMAN, AND VINCENT M. CHRISTOFFELS* Experimental and Molecular Cardiology Group, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands ABSTRACT In mouse, atrial natriuretic factor (ANF) gene expression was shown to be a marker for chamber formation within the embryonic heart. To gain insight into the process of chamber formation in the chicken embryonic heart, we analyzed the expression pattern of cANF during development. We found cANF to be specifically expressed in the myocardium of the morphologically distinguishable atrial and ventricular chambers, similar to ANF in mouse. cANF expression was never detected in the myocardium of the atrioventricular canal (AVC), inner curvature, and outflow tract (OFT), which is lined by endocardial cushions. Expression was strictly excluded from the interventricular myocardium and most proximal part of the bundle branches, as identified by the expression of Msx-2, whereas the rest of the bundle branches, trabeculae, and surrounding working myocardium did express cANF. The myocardium that forms de novo within the cushions after looping did not express cANF. At HH9 cANF expression was first observed in a subset of cardiomyocytes, which was localized ventrally in the fused heart tube and laterally in the unfused cardiac sheets. Together, these results show that cANF expression can be used to distinguish differentiated chamber (working) myocardium, including the peripheral ventricular conduction system, from embryonic myocardium. We conclude that differentiation of chamber myocardium takes place already at HH9 at the ventral side of the linear heart tube, possibly preceded by latero-medial signals in the unfused cardiac sheets. Anat Rec 266:93–102, 2002. © 2002 Wiley-Liss, Inc. Key words: cANF; cTnI; Msx-2; cIrx4; heart development; chamber formation The formation of the heart is initiated at the beginning of gastrulation, Hamburger and Hamilton (HH) (Hamburger and Hamilton, 1951) stage 3 in chicken, when cardiac precursor cells migrate from the primitive streak to form two sheets of cardiac mesoderm. The two sheets of cardiac mesoderm migrate medially and fuse in anteroposterior (AP) direction to form the linear heart tube from stage HH9 onward (for a review, see Fishman and Chien, 1997). Fate map experiments have shown that chamber primordia are laid down in linear order in position and time, first and anterior the right ventriclular primordium, last and posterior the atrial primordia (Garcia-Martinez and Schoenwolf, 1993; Markwald et al., 1999; De la Cruz © 2002 WILEY-LISS, INC. and Sanchez-Gomez, 1999). The embryonic myocardium of the cardiac crescent and of the heart tube originating from Grant sponsor: Netherlands Heart Foundation; Grant number: NHS M96.002. *Corresponding author: V.M. Christoffels, Dept. of Anatomy and Embryology, Academic Medical Center, Meibergdreef 15 L2255, 1105 AZ Amsterdam, The Netherlands. Fax: ⫹31-206976177. E-mail: firstname.lastname@example.org Received 7 August 2001; Accepted 10 October 2001 94 HOUWELING ET AL. it is patterned along the AP axis. Dominant pacemaker activity and highest beat rate frequency are always found at the posterior side (van Mierop, 1967; Satin et al., 1988; Kamino et al., 1981). Several genes are expressed in an AP pattern in mouse and chicken, including AMHC1, MLC2V, Tbx5, HRT1, HRT2, and Irx4 (Yutzey et al., 1994; O’ Brien et al., 1993; Bruneau et al., 1999; Nakagawa et al., 1999; Bao et al., 1999; Christoffels et al., 2000a). Furthermore, the embryonic myocardium contains distinct anterior and posterior cardiomyogenic lineages that are precursors of ventricular and atrial cardiomyocytes, respectively (Yutzey et al., 1995). In spite of the early patterning and specification of myocytes within the crescent and tubular heart, several observations indicate that subsequent steps of differentiation of embryonic myocytes take place within the heart tube to generate the (working) myocardium of the chambers. The embryonic myocardium of the entire tubular heart is characterized by slow conduction and peristaltoid csontractions (Patten, 1949; Kamino, 1991; Moorman et al., 1998). Only after looping, in both chicken and rat hearts, the conduction velocity increases in the myocardium of atria and ventricles, resembling the conduction velocity in mature working myocardium. At this time the flanking myocardium of the atrioventricular canal (AVC) and outflow tract (OFT) still displays the phenotype of the linear heart tube (de Jong et al., 1992; Moorman and Lamers, 1994; Moorman et al., 2000). The chamber myocardial cells also show other indications of a more mature phenotype, such as a higher activity of the sarcoplasmic reticulum and development of the sarcomeres (Moorman et al., 1998, 2000). In mouse, the genes for ANF (Zeller et al., 1987), Chisel (Palmer et al., 2001), and Cx40 (Delorme et al., 1997) are not expressed in the linear heart tube. With further development they are specifically expressed in the morphologically and electrophysiologically distinguishable chambers and not in the myocardium that is lined by endocardial cushions (Christoffels et al., 2000a; Moorman et al., 2000). Expression of these genes therefore discriminates between embryonic and more mature chamber myocardium. Their onset of expression indicates a highly localized initiation of differentiation toward the chamber-myocardial phenotype. This differentiation occurs at the ventral side of the linear heart tube and at the outer curvature of the looped heart (Christoffels et al., 2000a). The chicken heart is formed essentially similar to the mammalian heart. It can be easily manipulated and is very suitable for developmental and functional studies, particularly during early developmental stages, because early cardiac development proceeds at a rate slower than that in rodents. To get an insight into the developmental timing and precise localization of chamber formation in the embryonic chicken heart, we analyzed the expression pattern of the cANF gene by a detailed whole mount and serial section in situ hybridization (ISH) analysis. As a reference, we analyzed the patterns of cardiac troponin I (cTnI), which marks myocardial cells; Msx-2, a homeobox gene related to the Drosophila muscle segment homeobox gene, which marks the proximal ventricular conduction system; and chicken Iroquois related homeobox gene 4 (cIrx4), which marks a segment of the heart tube within which the ventricular chambers will develop. Like its mouse homologue, cANF expression was found to be restricted to the morphological chambers. Expression was not detected in the embryonic myocardium of the AVC, inner curvature, or OFT; in the Msx-2-positive components of the conduction system; or in the cardiomyocytes that form de novo within the cushions. This indicates that cANF expression is indeed marking the myocardium of the differentiating chambers. Its onset at HH9 revealed the formation of ventricular chamber myocardium at the ventral side of the tubular heart and the lateral side of the cardiac sheets before fusion, indicating that chamber formation at the ventral side/outer curvature of the tubular heart may be guided by latero-medial signals in the cardiac crescent. MATERIALS AND METHODS Isolation of Chicken Embryos and Tissue Processing Fertilized chicken eggs were obtained from a local hatchery (Drost BV, Nieuw Loosdrecht, The Netherlands), incubated at 37.5° C in a moist atmosphere and turned automatically every hour for appropriate periods. Embryos were isolated in sterile phosphate-buffered saline (PBS), staged according to the HH classification (Hamburger and Hamilton, 1951) and fixed for 4 hrs to overnight in freshly prepared 4% paraformaldehyde in PBS by rocking at 4° C. For ISH on sections, embryos were dehydrated in a graded ethanol series, embedded in paraplast, sectioned at 15 m, and mounted onto aminoakylsilanecoated slides. For whole-mount ISH, embryos were washed in 0.1% Tween-20 in PBS (PBST), dehydrated in a graded methanol series in PBST, and stored at –20° C in 100% methanol until use. Probe Synthesis cIrx4 probe was kindly provided by Dr. C.L. Cepko (Bao et al., 1999). Msx-2 probe was kindly provided by Dr. P.S. Thomas. Chicken cANF cDNA (Akizuki et al., 1991) was obtained by PCR amplification from a chicken heart cDNA library and cloned into pBluescript SK⫹ (Stratagene, La Jolla, CA). The following primers were designed based on the sequence of cANF cDNA (GenBank accession no. X57702): forward: 5⬘ AACTTCCCCTATTCCCAACGAA 3⬘ (position, nucleotides 30 –51); reverse: 5⬘ AGACAGGAGAGAGGTCCAGCAT 3⬘ (position, nucleotides 706 – 727). cTnI cDNA was obtained by PCR amplification from a chicken heart cDNA library and cloning into pBluescript SK⫹. The following primers were designed based on the Fig. 1. cANF (A and C) expression compared to cTnI (B and E) and Msx-2 (D) expression at HH30 on serial sections. cANF is not expressed in the cTnI-positive myocardium within the cushion mesenchyme (A and B; * in C and E) or the Msx-2-positive conduction system (C and D). ra, right atrium; la, left atrium; rv, right ventricle; lv, left ventricle; ep, epicardium; avb; atrioventricular bundle; ravrb, right atrioventricular ring bundle; me, mesenchyme; ec, endocardium. Scale bar ⫽ 100 m. Fig. 2. cANF (A, C, and D) expression compared to Msx-2 (B and E) and cTnI (F) expression at HH28 on serial sections. No cANF expression is observed in Msx-2-positive areas, and almost no expression is found in the compact myocardium (enlargement of boxed area in A is shown in C). The region boxed in B is shown for ANF, Msx-2, and cTnI in D–F, respectively. Note the very weak Msx-2 staining in the left atrioventricular junction (arrow in B). cTnI staining is less intense in the AVB (arrow in F). Abbreviations are as in legend to Figure 1; tm, trabeculated myocardium; cm, compact myocardium; ep, epicardium. Scale bar ⫽ 100 m. Figure 1. Figure 2. 96 HOUWELING ET AL. sequence of chicken cTnI cDNA (GenBank accession no. M73703 (Hastings et al., 1991)): forward: 5⬘ GAGCAAAGCGGGAGTTGGA 3⬘ (position, nucleotides 32–50); reverse: 5⬘ CCACAACGCTGCCCTTAAAG 3⬘ (position, nucleotides 600 – 619). Digoxigenine-labeled antisense mRNA complementary to cANF, Msx-2, chicken cTnI, cIrx4 (Bao et al., 1999), rat cTnI (Ausoni et al., 1991), and rat ANF (Zeller et al., 1987) were produced by in vitro transcription according to the manufacturer’s instructions (Roche, Mannheim, Germany). Whole-Mount ISH and ISH on Sections ISHs were performed as described recently for mouse and human tissue (Moorman et al., 2001). Probe concentration was 1 g/ml in hybridization mix. Individual sections were surrounded with a hydrophobic barrier using an ImmEdge pen (Vector Laboratories, Burlingame, CA) to confine hybridization mix to the sections during prehybridization and hybridization. Identical hybridization conditions (hybridization at 70° C, in hybridization mix containing 50% formamide and 5⫻ SSC) were used for all probes. After overnight color reaction, the sections were dehydrated in a graded ethanol series followed by 3 ⫻ 5 minutes xylene and embedded in Entallan. Images of sections were taken using a digital Nikon Coolpix 950 camera coupled to a Zeiss Axiophot microscope. Flatfield correction was performed with a user-written macro in PMISS200. Images of whole-mount embryos were taken on a thin layer of 1% agarose using a Nikon Coolpix 950 camera and a Wild M7 S microscope. RESULTS To gain insight into the process of chamber formation in the chicken embryonic heart, we analyzed the developmental expression pattern of cANF on serial sections. The results are presented in retrograde order to first demonstrate which cardiac structures were found to express cANF before defining the stage and localization of the onset of expression. Chicken cTnI was used as a marker for cardiomyocytes and cIrx4 was used as a marker identifying the AVC, ventricles, and proximal OFT. To mark the changing position of the interventricular ring that contributes to part of the conduction system, Msx-2 expression was assessed. cTnI Is a Marker for Cardiomyocytes in Chicken We first assessed whether chicken cTnI is a suitable marker for all myocardium by whole-mount ISH and ISH on sections at all developmental stages analyzed (HH8 – HH30). Expression was detectable in all myocardium of the heart at embryonic and fetal stages, including the myocardium developing in the dorsal mesocardium, atrioventricular cushions, and truncal ridges (Figs. 1B, E; 2F; 3C, F; and 4F). No cTnI expression was detected in cushion mesenchyme (e.g., Fig. 1B, E; 2F; and 3C, F), endocardium (Fig. 1E), or epicardium (Fig. 1B). This underlines the myocardial specificity of the probe and the resolving power of the ISH method used. cTnI gene expression was weaker in the Msx-2 positive myocardium of the conduction system (Figs. 1B, D, E and 2E, F), which is in line with the notion that in the conduction system the contractile apparatus does not develop extensively. cTnI was found to be expressed selectively in the cardiac mesoderm from HH8 onward (Fig. 5A) and in the entire tubular heart from HH10 onward (not shown). ISH on sections revealed strong expression from HH9 in the cardiac sheets (Fig. 6B, D, F). Aside from some background that was sometimes observed in ISH experiments at earlier stages (e.g., Fig. 6D), other tissues did not stain with this probe (compare Figs. 5A and 6D, F). No signal was observed in control experiments using a sense probe. As was shown previously for rat and Xenopus (Ausoni et al., 1991; Drysdale et al., 1994), cTnI appears to mark all cardiomyocytes in chicken from early stages of cardiogenesis. Expression Patterns in the Fetal Heart (Figs. 1 and 2) To reveal whether cANF marks the myocardium of the chambers of the heart, we first studied its expression pattern in the fetal heart. At stages HH28 and HH30, when the heart has obtained its adult configuration, cANF expression was observed in the atria and in the entire trabeculated component of the ventricle, which includes the bundle branches and subendocardial peripheral conduction system. Expression was very low in the compact ventricular myocardium and absent from the atrial septum, AVC, OFT, mesenchyme of the cushions, endocardium, and epicardium. Msx-2 marks the myocardium surrounding the interventricular foramen that gives rise to the atrioventricular bundle (AVB) (His bundle), right atrioventricular ring bundle (RAVRB), retroaortic root branch (RAORB), and proximal part of the bundle branches (Figs. 1D and 2B, E), as was previously shown (Chan-Thomas et al., 1993; Moorman et al., 1998). cANF expression was excluded specifically from these Msx-2-positive myocytes (Figs. 1A, C, D and 2A, B, D, E). The part of the ventricular septum on top of the Msx-2-expressing AVB that originates by muscularization of the mesenchymal cushion after HH27– HH28 (van den Hoff and Moorman, unpublished observations) did not express cANF either (Fig. 1C–E). cIrx4 mRNA was detected in the AVC in both ventricles and in all cTnI-expressing myocardium of the ventricular septum, including the Msx-2-expressing myocytes and myocardium formed by muscularization (not shown). Fig. 3. cANF (A and D) expression compared to cIrx4 (B and E) and cTnI (C and F) expression at HH24, and ANF (G) compared to cTnI (H) in mouse at E10.5 on serial sections. cTnI-expressing myocytes within the mesenchyme of the dorsal mesocardium do not express cANF (* in A and B). Note cIrx4 expression in the AVC that borders the cANF expression domain in the atria (arrows in A and B). Arrows in G and H indicate the AVC and part of the right atrium, which do not express ANF. Note the weak cTnI expression in the right ventricle (* in H) in the region where ANF expression is also observed (* in G). Abbreviations are as in previous legends; avc; atrioventricular canal; oft, outflow tract; ift, inflow tract. Scale bar ⫽ 100 m. Fig. 4. cANF (A, C, and E) expression compared to cIrx4 (B and D) and cTnI (F) expression at HH21 and HH22, and ANF (G) compared to cTnI (H) in mouse at E9.5 on serial sections. Note the sharp border between cANF and cIrx4 expression domains in the AVC (arrows in A and B). No cANF is expressed in the AVC, inner curvature, or OFT at HH22 (E and F). Note the restricted ANF expression near the dorsal mesocardium, compared to cTnI (G and H). Abbreviations are as in previous legends; oc, outer curvature; ic, inner curvature; dm, dorsal mesocardium; v, ventricle. Scale bar ⫽ 100 m. Figure 3. Figure 4. 98 HOUWELING ET AL. In summary, cANF expression was found only in the atrial appendages, in the trabeculated myocardium of the ventricles, and in the bundle branches of the fetal heart. cANF Expression During Chamber Formation (Figs. 3 and 4) To investigate whether cANF is selectively expressed in myocardium of the forming chambers of the looped heart, we examined hearts from stage HH16 –HH24. At these stages the heart tube has looped and has become S-shaped (De la Cruz and Sanchez-Gomez, 1999; Männer, 2000). cANF expression at stages 16 –24 was detected in the inflow tract (IFT), embryonic atria, and trabecular component of the ventricle (shown for HH21 and HH24 in Figs. 3A and 4A). Expression was not detected in the AVC, inner curvature, or OFT (shown for HH21, HH22, and HH24 in Figs. 3A, D and 4A, C, E), i.e., in the smoothwalled myocardium that is lined by cushions. cIrx4 was found to be expressed in the AVC, ventricles, and proximal OFT, forming a “segment” of expression within the tubular heart (Figs. 3B, E and 4B, D). Serial sections revealed that the trabeculated, cANF-expressing ventricular myocardium forms only a restricted domain at the outer curvature within the cIrx4-expressing segment. The border of ventricular cANF expression in the trabeculae coincides with the border of the atrioventricular and OFT cushions (Fig. 4E). At the junction between atria and AVC, the expression domains of cANF and cIrx4 were found to be mutually exclusive, their border coinciding with the border of the atrioventricular cushions (Fig. 4A, B). cANF expression was absent from the cTnI-expressing newly formed myocardium within the dorsal mesocardium and the atrium septum (Fig. 3A, C). At comparable stages of mouse heart development, after looping (E9.5 and E10.5), strong ANF expression was detected at the outer curvature at the level of the left ventricle, and weaker expression in the outer curvature of the right ventricle and in the embryonic atria. No expression was observed in the IFT, AVC, inner curvature, or OFT (Figs. 3G and 4G). The mouse cTnI gene was expressed in the entire myocardium specifically, although at lower levels in the right ventricle and OFT (Figs. 3H and 4H). In summary, at stages of chamber formation and expansion, expression of cANF was observed only in the trabeculated ventricular myocardium and in the atrial myocardium of the forming appendages, similar to the pattern of ANF in mouse. In contrast to ANF in mouse, cANF was found to be expressed in the IFT as well. Onset of Chamber Formation in the Early Embryonic Heart (Figs. 5 and 6) The expression data indicated that cANF is a marker for chamber myocardium differentiated from the embryonic myocardium. To investigate the process of differentiation, we next investigated the onset and localization of cANF in the cardiac crescent and subsequently in the tubular heart. The stage of onset of cANF expression was defined by whole-mount ISH. Expression of cANF was not detected at HH7 and HH9 – (Fig. 5B). At this stage the cardiac sheets start to fuse in a cranial to caudal direction to give rise to the heart tube. Weak expression became detectable first at HH9⫹ at the ventral side of the linear heart tube restricted at both the cranial and caudal sides to the prospective ventricular region. The timing and lo- calization of cANF expression appeared to be similar to those of ANF expression in mouse. In mouse the linear heart tube did not express ANF, but ANF expression was detected at the ventral side of the future ventricular portion soon after looping of the ventricular part of the heart tube became evident (Fig. 5D, E). With ISH on sections, which appeared to be more sensitive than the wholemount ISH method, we observed cANF expression at HH9 at the ventral side of the cTnI-positive myocardium in the region where the heart tube has been formed (Fig. 6A, C, E). More caudal, in the region of fusion, restricted cANF expression was detected laterally in the cTnI-expressing cardiac sheets (Fig. 6C, D). Caudal to this region, cANF expression was detectable in a few myocytes of the unfused sheets. When fusion progresses caudally, this lateral myocardium will become positioned at the ventral side (de Jong et al., 1990) (Fig. 7) and later form the (trabeculated portion of the) ventricles (De la Cruz and Sanchez-Gomez, 1999). Using ISH on mouse sections of early tubular heart stages, ANF was never detected outside the future ventricular region of the heart tube. In summary, cANF expression in chicken was found to be initiated not only at the ventral side of the fused heart tube, as in mouse, but also laterally in the unfused cardiac sheets. Expression is therefore initiated at a relatively earlier stage of heart development than in mouse. DISCUSSION cANF Is a Marker Specific for Chamber Myocardium We previously found that in mouse expression of ANF and a few other genes (Chisel, SERCA2A, Cx40, Cx43, and Irx5) are suitable markers to discriminate between embryonic and chamber myocardium (Christoffels et al., 2000a). In the linear heart tube and in the myocardium of the AVC, inner curvature, and OFT that flanks the morphological chambers, expression of these genes was distinctly lower or absent. To gain insight into the developmental timing and precise localization of chamber formation in the embryonic chicken heart, we performed a detailed whole-mount and section ISH study using the presumed homologue of mammalian ANF, cANF (Akizuki et al., 1991), and some reference markers, cTnI, Msx-2, and cIrx4, as probes. From HH16 onward, expression of cANF was found to be restricted to the atria and the trabeculated portion of the ventricles at the outer curvature. Expression was absent from the cushion-lined AVC, inner curvature, and OFT, which play an important role in alignment of the chambers and septation (Mjaatvedt et al., 1999). The restriction of expression to the morphologically identifiable chambers indicates that the myocardium of the chambers is phenotypically distinct from that of the cushion-lined compartments. This observation is supported by electrophysiological measurements, showing that the atrial and ventricular chambers obtained fastconducting properties during development (from HH23 onward). The flanking AVC and OFT retained the slowconducting properties, as found in the original tubular heart of the younger chicken embryo (Arguello et al., 1988; de Jong et al., 1992; Moorman et al., 1998). The electrical properties of the distinct cardiac compartments have a matching contractile and molecular phenotype (Moorman et al., 1998, 2000). Therefore, cANF is specifically ex- LOCALIZATION OF CARDIAC CHAMBER FORMATION Fig. 5. Onset of cTnI (A) and cANF (B and C) expression in chicken and of ANF (D and E) in mouse detected by whole-mount ISH. cTnI expression is detectable in the cardiogenic mesoderm from HH9 – onward. cANF expression was first observed at HH9⫹ in the embryonic Fig. 6. cANF (A, C, and E) expression in the early tubular heart and in the cardiac mesoderm in an HH9 embryo from cranial (A) to caudal (E) compared to cTnI (B, D, and F) expression. Note the restriction of cANF expression to the ventral side of the heart tube, compared to cTnI (arrows in A and B indicate borders of cANF- and cTnI-expressing regions, respectively). Caudal of fusion cANF expression (arrows in C and E) is observed only in the lateral part of the cTnI-positive cardiac sheets (borders of expression indicated by arrows in D and F). Scale bar ⫽ 100 m. pressed in chamber myocardium that is more mature than the neighboring embryonic myocardium. Not all chamber myocardium continues to express cANF. As was observed for ANF in mouse, transmural expression was observed in the embryonic ventricle (HH9 –HH10). Soon thereafter the expanding compact myocardial outer layer of the ventricles gradually looses cANF expression. This is in line with observations of De la Cruz and Sanchez-Gomez (1999), who found that tagged embryonic ventricular wall myocardium maps to the trabeculated myocardium of the formed ventricles. Therefore, cANF expression does not mark chamber myocardium continuously throughout development. 99 ventricle of chicken (arrow in C) and in the tubular heart of mouse shortly after the initiation of looping (E8.5). cm, cardiac mesoderm; ev, embryonic ventricle; aip, anterior intestinal portal. Fig. 7. A highly schematic representation of the morphogenetic events that transform the flat cardiac mesodermal sheets (pink) into the heart tube during development (adapted from de Jong et al., 1990). Four imaginary sections (A–D) illustrate the process of folding of the embryo, by which the lateral borders of the separate cardiac sheets (green boxes) become apposed in the ventral mesocardium after a movement of 180°. By this process of folding the intestine is also formed. Note the transition from a latero-medial axis in the cardiac mesodermal sheets (A) to a ventro-dorsal axis in the tubular heart (D). As the process of folding and fusion of the cardiac sheets proceeds in a cranial to caudal direction during development, section A is both developmentally earlier than and caudal to section D, in which the folding process is complete. Abbreviations are as in previous legends; cc, coelomic cavity; ht, heart tube; vm, ventral mesocardium; Lat, lateral; M, medial; D, dorsal; V, ventral; L, left; R, right. The spatial and developmental expression pattern of cANF is similar to that of ANF in mouse, indicating that cANF is the homologue of ANF in mammals. However, in contrast to mouse, cANF expression in chicken is retained in the right ventricle during development (Figs. 1A, 2A, and 4G). Furthermore, the IFT of the chicken heart expressed cANF, whereas the expression at the intake of the heart in mouse is restricted to the future auricles. Therefore, in mouse ANF is expressed in a more restricted region of the heart than cANF is in chicken. cANF Is Not Expressed in Newly Formed Myocardium In both mouse and chicken, ANF expression was not observed in the myocardium developing within the 100 HOUWELING ET AL. cushion mesenchyme, which expresses cTnI (Figs. 1A–C, E and 3A, C) and will contribute to the septa and valves. As this myocardium forms by a process of migration and recruitment relatively late during development (Mjaatvedt et al., 1999; van den Hoff et al., 1999; van den Hoff and Moorman, unpublished observations), it is likely to still have an embryonic myocardial phenotype before it differentiates toward working myocardium. The absence of cANF expression in this myocardium, together with the selective expression of cANF in the chambers, indicates that cANF is a marker for differentiated chamber myocardium. Peripheral Ventricular Conduction System Differs in Origin and Phenotype From the Proximal Conduction System, as Revealed by Expression of cANF and Msx-2 At HH28 –HH30, cANF transcripts were found in all of the trabeculated myocardium of the ventricles and the bundle branches, which together comprise the developing peripheral subendocardial ventricular conduction system (Vassall-Adams, 1982; Moorman et al., 1998). The proximal conduction system is formed by the AVB, the retro-aortic root branch, the RAVRB, and the most proximal part of the left and right bundle branches (Wessels et al., 1992; Moorman et al., 1998), which in chicken specifically express the homeobox gene Msx-2 (Chan-Thomas et al., 1993; this study). cANF expression was strictly absent from the Msx-2-positive cells (Fig. 2D, E). These mutually exclusive patterns indicate that the proximal and peripheral conduction systems are strictly separated from each other at HH28. The myocardium within which the proximal components of the conduction system are developing (AVC, top of the septum) never at any given stage in development did express cANF. In contrast, the myocardium that gives rise to the to bundle branches and the subendocardial peripheral ventricular conduction system (trabeculated component of the ventricle, myocardium flanking the interventricular septum (Vassall-Adams, 1982; Moorman et al., 1998)) was found to express cANF at all stages of development. These observations strongly indicate that the proximal and peripheral components of the conduction system are derived from two phenotypically distinct pools of myocytes. These findings are in line with the observations of Cheng et al. (1999), who found that the ventricular conduction system is largely derived from multipotent cardiomyocytes of the tubular and chambered heart— cardiomyocytes that give rise to either the proximal or the peripheral ventricular conduction system. The distinction between the phenotypes of the two parts of the conduction system, i.e., cANF-negative “embryonic” vs. cANF-positive “chamber-type” myocardium, is consistent with electrophysiological, ultrastructural, and molecular findings. These indicate that the cells of the proximal conduction system, the node, the AVB, and the lower rim under the atria just above the fibrous annulus (former AVC) still have a nodal-like morphology and low abundance of gap-junction proteins reminiscent to the embryonic myocardium, in contrast to the bundle branches and trabeculae that have a high density of gap-junctions (reviewed in Moorman et al., 1998; Lo, 2000). cIrx4-Expressing Segment vs. cANF-Expressing Ventricular Chamber cIrx4 is the first homeobox transcription factor shown to be expressed in an exceptionally restricted pattern in the developing heart (Bao et al., 1999; Bruneau et al., 2000; Christoffels et al., 2000b), where it is involved in compartment-specific MHC gene regulation (Bao et al., 1999). Our analysis shows that cIrx4, in addition to the ventricular chamber, is expressed in the AVC, inner curvature, and proximal OFT. This cIrx4-expressing segment of the tubular heart will not entirely become ventricular chamber. The trabeculated ventricular (cANF-expressing) myocardium will be formed at the ventral side and outer curvature within this segment. The AVC will become incorporated into the atria (Wessels et al., 1996). Recent studies of Irx4 function (Bao et al., 1999; Bruneau et al., 2001; Wang et al., 2001) indicated that Irx4 acts as a repressor of transcription of several genes in the ventricles. These findings and the expression pattern of cIrx4 together suggest that cIrx4 is involved in preventing the atrial chamber phenotype from invading the AVC and ventricles, rather than specifying the ventricular chamber. cANF Expression Reveals Early Pattern Formation and Differentiation in the Fusing Cardiac Mesodermal Sheets As cANF discriminates between embryonic and chamber myocardium, the onset of cANF expression possibly indicates the timing and localization of the transition from the embryonic toward the chamber myocardial phenotype. The pattern of cANF expression in chicken appeared to resemble the expression pattern of ANF in mouse, but, surprisingly, we found the onset of expression in both the tubular heart and the unfused sheets at HH9. The onset of expression in chicken is earlier than the onset of expression in mouse, which first expresses ANF at E8.5 in the already looping heart tube. This implies that differentiation and chamber specification occur at an earlier stage in chicken than in mouse. However, the apparent difference in the onset of differentiation might also be accounted for by temporal differences in morphogenetic events. If fusion of the two cardiac mesodermal sheets occurs at a relatively earlier stage of development in mouse than in chicken, one would expect onset of ANF expression in mouse in a developmentally more advanced stage. cANF expression was initiated cranially at the ventral side of the fused tubular heart and just caudally to the point of fusion in the lateral region of the cardiac sheets, as identified by cTnI expression. The lateral portions of the cardiac sheets become the ventral side of the tubular heart (Fig. 7A–D) (de Jong et al., 1990) when at HH10 –HH11 fusion has progressed more caudally. Indeed, at later stages most staining of cANF was observed along the ventral side of the tubular heart (not shown). The ventral side of the tube becomes the outer curvature during looping from which the ventricular chambers expand (De la Cruz and Sanchez-Gomez, 1999; Christoffels et al., 2000a). The formation of ventricular chamber myocardium at the ventral side of the heart tube requires ventro-dorsal patterning information in the tubular heart that was indeed found for Hand1 expression in mouse (Christoffels et al., 2000a; Harvey et al., 1999). The expression of cANF in the lateral region of the cardiac sheets, caudally to the point LOCALIZATION OF CARDIAC CHAMBER FORMATION of fusion, possibly provides the first molecular clue that latero-medial patterning in the unfused sheets precedes the ventro-dorsal patterning in the fused tube. ACKNOWLEDGMENTS We thank Dr. Penny Thomas, Imperial College, London, for the Msx-2 probe; Anita Buffing, Corrie de Gier-de Vries, and Piet de Boer for their contributions to the data presented; and Dr. Gertien Smits for critically reading the manuscript. LITERATURE CITED Akizuki N, Kangawa K, Minamino N, Matsuo H. 1991. Cloning and sequence analysis of complementary DNA encoding a precursor for chicken natriuretic peptide. FEBS Lett 280:357–362. 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. Ausoni S, de Nardi C, Moretti P, Gorza L, Schiaffino S. 1991. Developmental expression of rat cardiac troponin I mRNA. Development 112:1041–1051. Bao ZZ, Bruneau BG, Seidman JG, Seidman CE, Cepko CL. 1999. Regulation of chamber-specific gene expression in the developing heart by Irx4. Science 283:1161–1164. Bruneau BG, Logan M, Davis N, Levi T, Tabin CJ, Seidman JG, Seidman CE. 1999. Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev Biol 211:100 –108. Bruneau BG, Bao ZZ, Tanaka M, Schott J-J, Izumo S, Cepko CL, Seidman JG, Seidman CE. 2000. Cardiac expression of the ventricle-specific homeobox gene Irx4 is modulated by Nkx2-5 and dHand. Dev Biol 217:266 –277. Bruneau BG, Bao ZZ, Fatkin D, Xavier-Neto J, Georgakopoulos D, Maguire CT, Berul CI, Kass DA, Kuroski-de Bold ML, de Bold AJ, Conner DA, Rosenthal N, Cepko CL, Seidman CE, Seidman JG. 2001. Cardiomyopathy in Irx4-deficient mice is preceded by abnormal ventricular gene expression. Mol Cell Biol 21:1730 –1736. Chan-Thomas PS, Thompson RP, Robert B, Yacoub MH, Barton PJR. 1993. Expression of homeobox genes Msx-1 (Hox-7) and Msx-2 (Hox-8) during cardiac development in the chick. Dev Dyn 197:203– 216. Cheng G, Litchenberg WH, Cole GJ, Mikawa T, Thompson RP, Gourdie RG. 1999. Development of the cardiac conduction system involves recruitment within a multipotent cardiomyogenic lineage. Development 126:5041–5049. ChristoffelsVM, Habets PEMH, Franco D, Campione M, de Jong F, Lamers WH, Bao ZZ, Palmer S, Biben C, Harvey RP, Moorman AFM. 2000a. Chamber formation and morphogenesis in the developing mammalian heart. Dev Biol 223:266 –278. ChristoffelsVM, Keijser AGM, Houweling AC, Clout DEW, Moorman AFM. 2000b. Patterning the embryonic heart: identification of five mouse Iroquois homeobox genes in the developing heart. Dev Biol 224:263–274. de Jong F, Geerts WJC, Lamers WH, Los JA, Moorman AFM. 1990. Isomyosin expression pattern during formation of the tubular chicken heart: a 3D immunohistochemical analysis. Anat Rec 226: 213–227. 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. De la Cruz MV, Sanchez-Gomez C. 1999. Straight tube heart. Primitive cardiac cavities vs. primitive cardiac segments. In: De la Cruz M, Markwald RR, editors. Living morphogenesis of the heart. Boston: Birkhäuser. Chapter 3, p 85–98. Delorme B, Dahl E, Jarry-Guichard T, Briand JP, Willecke K, Gros D, Théveniau-Ruissy M. 1997. Expression pattern of connexin gene products at the early developmental stages of the mouse cardiovascular system. Circ Res 81:423– 437. Drysdale TA, Tonissen KF, Patterson KD, Crawford MJ, Krieg PA. 1994. Cardiac troponin I is a heart-specific marker in the Xenopus 101 embryo expression during abnormal heart morphogenesis. Dev Biol 165:432– 441. Fishman MC, Chien KR. 1997. Fashioning the vertebrate heart: earliest embryonic decisions. Development 124:2099 –2117. Garcia-MartinezV, Schoenwolf GC. 1993. Primitive streak origin of the cardiovascular system in avian embryos. Dev Biol 159:706 –719. Hamburger V, Hamilton JL. 1951. A series of normal stages in the development of the chick embryo. J Morphol 88:49 –92. Harvey RP, Biben C, Elliott DA. 1999. Transcriptional control and pattern formation in developing vertebrate heart: studies on NK-2 class homeodomain factors. In: Harvey RP, Rosenthal N, editors. Heart development. Boston: Academic Press. p 111–129. Hastings KE, Koppe RI, Marmor E, Bader D, Shimada Y, Toyota N. 1991. Structure and developmental expression of troponin I isoforms. cDNA clone analysis of avian cardiac troponin I mRNA. J Biol Chem 266:19659 –19665. Kamino K. 1991. Optical approaches to ontogeny of electrical activity and related functional organization during early heart development. Physiol Rev 71:53–91. Kamino K, Hirota A, Fujii S. 1981. Localization of pacemaking activity in early embryonic heart monitored using voltage-sensitive dye. Nature 290:595–597. Lo CW. 2000. Role of gap junctions in cardiac conduction and development. Circ Res 87:346 –348. Männer J. 2000. Cardiac looping in the chick embryo: a morphological review with special reference to terminological and biomechanical aspects of the looping process. Anat Rec 259:248 –262. Markwald RR, Trusk T, Moreno-Rodriguez RA. 1999. Formation and septation of the tubular heart: integrating the dynamics of morphology with emerging molecular concepts. In: De la Cruz MV, Markwald RR, editors. Living morphogenesis of the heart. Boston: Birkhäuser. p 43– 84. Mjaatvedt CH, Yamamura H, Wessels A, Ramdell A, Turner D, Markwald RR. 1999. Mechanisms of segmentation, septation, and remodeling of the tubular heart: endocardial cushion fate and cardiac looping. In: Harvey RP, Rosenthal N, editors. Heart Development. San Diego: Academic Press. p 195–208. Moorman AFM, Lamers WH. 1994. Molecular anatomy of the developing heart. Trends Cardiovasc Med 4:257–264. Moorman AFM, de Jong F, Denyn MMFJ, Lamers WH. 1998. Development of the cardiac conduction system. Circ Res 82:629 – 644. Moorman AFM, Schumacher CA, de Boer PAJ, Hagoort J, Bezstarosti K, van den Hoff MJB, Wagenaar GTM, Lamers JMJ, Wuytack F, Christoffels VM, Fiolet JWT. 2000. Presence of functional sarcoplasmic reticulum in the developing heart and its confinement to chamber myocardium. Dev Biol 223:279 –290. Moorman AFM, Houweling AC, de Boer PAJ, Christoffels VM. 2001. Sensitive nonradioactive detection of mRNA in tissue sections: novel application of the whole-mount in situ hybridization protocol. J Histochem Cytochem 49:1– 8. Nakagawa M, Richardson JA, Olson EN, Srivastava D. 1999. HRT1, HRT2, and HRT3: a new subclass of bHLH transcription factors marking specific cardiac, somitic, and pharyngeal arch segments. Dev Biol 216:72– 84. O’Brien TX, Lee KJ, Chien KR. 1993. Positional specification of ventricular myosin light chain 2 expression in the primitive murine heart tube. Proc Natl Acad Sci USA 90:5157–5161. Palmer S, Groves N, Schindeler A, Yeoh T, Biben C, Wang CC, Sparrow D, Barnett L, Jenkins NA, Copeland NG, Koentgen F, Mohun T, Harvey RP. 2001. The small muscle-specific protein Csl modifies cell shape and promotes myocyte fusion in an insulinlike growth factor 1-dependent manner. J Cell Biol 153:985–997. Patten BM. 1949. Initiation and early changes in the character of the heartbeat in vertebrate embryos. Physiol Rev 29:31– 47. Satin J, Fujii S, de Haan RL. 1988. Development of cardiac heartbeat in early chick embryos is regulated by regional cues. Dev Biol 129:103–113. van den Hoff MJB, Moorman AFM, Ruijter JM, Lamers WH, Bennington RW, Markwald RR, Wessels A. 1999. Myocardialization of the cardiac outflow tract. Dev Biol 212:477– 490. 102 HOUWELING ET AL. van Mierop LHS. 1967. Localization of pacemaker in chick embryo heart at the time of initiation of heartbeat. Am J Physiol 212:407– 415. Vassall-Adams PR. 1982. The development of the atrioventricular bundle and its branches in the avian heart. J Anat 134:169 –183. Wang GF, Nikovits Jr W, Zheng-Zheng B, Stockdale FE. 2001. Irx4 forms an inhibitory complex with the vitamin D and retinoic X receptors to regulate cardiac chamber-specific slow MyHC3 expression. J Biol Chem 276:28835–28841. Wessels A, Vermeulen JLM, Verbeek FJ, Virágh SZ, Kálmán F, Lamers WH, Moorman AFM. 1992. Spatial distribution of “tissue-specific” antigens in the developing human heart and skeletal muscle: III. An immunohistochemical analysis of the distribution of the neural tissue antigen G1N2 in the embryonic heart; implications for the development of the atrioventricular conduction system. Anat Rec 232:97–111. Wessels A, Markman MWM, Vermeulen JLM, Anderson RH, Moorman AFM, Lamers WH. 1996. The development of the atrioventricular junction in the human heart. Circ Res 78:110 –117. Yutzey KE, Rhee JT, Bader D. 1994. Expression of the atrial-specific myosin heavy chain AMHC1 and the establishment of anteroposterior polarity in the developing chicken heart. Development 120: 871– 883. Yutzey KE, Gannon M, Bader D. 1995. Diversification of cardiomyogenic cell lineages in vitro. Dev Biol 170:531–541. Zeller R, Bloch KD, Williams BS, Arceci RJ, Seidman CE. 1987. Localized expression of the atrial natriuretic factor gene during cardiac embryogenesis. Gene Dev 1:693– 698.