DEVELOPMENTAL DYNAMICS 218:123–135 (2000) The Rostro-Caudal Position of Cardiac Myocytes Affect Their Fate VIDYA PATWARDHAN,1 SONALIS FERNANDEZ,1 MICHAEL MONTGOMERY,1 AND JUDITH LITVIN1,2* 1 Department of Anatomy and Cell Biology, Temple University, Philadelphia, Pennsylvania 2 Fels Institute for Cancer Research and Molecular Biology, Philadelphia, Pennsylvania ABSTRACT During chick embryogenesis, cells destined to form cardiac myocytes are located within the primitive streak at stage 3 in the same relative anterior-posterior distribution as in the prelooped heart. The most rostral cells contribute to the extreme anterior pole of the heart, the bulbus cordis, and the most caudal to the extreme posterior end, the sinoatrial region. After gastrulation, these cells commit to the myocyte lineage and, retaining their relative positions, migrate to the anterior lateral plate. From stages 5 to 10 they diversify into atrial and ventricular myocytes, with the former located posteriorly and the latter, anteriorly. To determine the effect of a change in the rostro-caudal position of these cells on their diversification, anterior lateral plate mesoderm and the underlying endoderm were cut and rotated 180° along the longitudinal axis, at stages 4 – 8. The subsequent diversification of these precursor cells into atrial and ventricular myocytes was examined using lineage-specific markers. Our results showed that altering location along the longitudinal axis through stage 6 changed the normal fate of a precursor cell. The orientation of the overlying ectoderm did not alter normal morphogenesis or determination of fate. Dev Dyn 2000;218:123–135. © 2000 Wiley-Liss, Inc. Key words: diversification; chamber specification; chick embryo; heart formation INTRODUCTION In the early chick gastrula (Hamburger and Hamilton stage 3), cardiac myocyte precursors are present in the middle two-thirds of the primitive streak, just caudal to Hensen’s node (Garcia-Martinez and Schoenwolf, 1993), arranged in the same relative order as in the prelooped heart (Fishman and Chien, 1997). The cells fated to become the bulbus arteriosus are located most anteriorly and the sinus venosus most posteriorly. By stage 4, the cells have gastrulated and migrated into the anterior lateral plate (ALP) mesoderm on both sides of Hensen’s node. Mesodermal cells in the ALP are specified to the cardiac lineage by this stage (Sater and Jacobson, 1989 and 1990; Gonzalez-Sanchez and Bader, 1990). During the next several hours, these cells migrate anteriorly and medially to meet at the midline, form the cardiac crescent by stage 8, and fuse to form a © 2000 WILEY-LISS, INC. median single heart tube by stage 9 –10. The mesodermal cells from the anterior part of the heart tube form the ventricles (DeHaan, 1965). Cardiac myocyte differentiation begins with these cells and subsequently, the posteriorly located cells in the heart-forming region (HFR) differentiate to form the atria (Bisaha and Bader, 1991; Yutzey et al., 1994). The final destination and fate of the cardiac myocytes is reflected by their initial position in the primitive streak prior to gastrulation (Garcia-Martinez and Schoenwolf, 1993; Fishman and Chien, 1997). In the differentiating cardiac myocytes, muscle-specific gene expression precedes diversification into atrial and ventricular myocytes (reviewed in Litvin et al., 1992). During subsequent stages of development, the anterior cells diversify into ventricular myocytes and posterior cells diversify into atrial myocytes. These two subsets of myocytes can be distinguished from each other by virtue of their differences in contractile and conductive properties (DeHaan, 1965) as well as the expression of different myosin isoforms (Bisaha and Bader, 1991; Yutzey et al., 1994; Fishman and Chien 1997). The initial expression pattern of lineage-specific structural genes proceeds along the rostro-caudal axis of the fusing heart tube between stages 7 and 12, and reflects the subsequent diversified phenotype (Bisaha and Bader, 1991; Yutzey et al., 1994; Yutzey and Bader, 1995). Thus, in normal development, differentiation and diversification of cardiac myocytes always occurs in a rostral-to-caudal direction. However, the fate of these cells apparently can be altered during early embryogenesis, as was shown by transplantation of presumptive atrial cells into the anterior ventriculogenic area prior to stage 8 (Satin et al., 1988). These cells then showed a beating frequency characteristic of ventricular cells rather than atrial cells. This change in electrophysiology suggests an underlying plasticity in the diversification of the cardiac precursors. The recent identification of chamber-specific molecular markers like ventricular myosin heavy chain 1 (VMHC1) Grant sponsor: National Institutes of Health; Grant number: HL52052-04. M. Montgomery and J. Litvin contributed equally to this work. *Correspondence to: Judith Litvin, Temple University Medical School, Department of Anatomy and Cell Biology, 3420 N. Broad Street, Philadelphia, PA 19140. E-mail: [email protected] Received 14 July 1999; Accepted 31 January 2000 124 PATWARDHAN ET AL. (Bisaha & Bader 1991) and atrial myosin heavy chain 1 (AMHC1) (Yutzey et al., 1994) presents the opportunity to accurately identify the subsequent fate of transplanted precursor myocytes. We altered the spatial arrangement of the cardiac myocyte precursors along the rostro-caudal axis and, by in situ hybridization, identified them as either ventricular or atrial. The HFR (mesoderm and endoderm) of stages 4 – 8 chick embryos were cut and rotated by 180°, bringing the posterior presumptive atrial cells to the anterior and vice versa. AMHC1 or VMHC1 expression was used as the criterion to identify the fate of these cells in their new locations. Our results showed that altering location along the longitudinal axis at stage 4, 5, and 6 changed the normal fate of a precursor cell. This occurred through a reassignment of fate and not by remigration of the cells from their new location to the original one, as shown by DiI-labeling studies. At stages 7 and 8, the fate of these cells could not be altered. Rotation of the overlying ectoderm alone did not alter normal morphogenesis or determination of fate. Diversification fate of atrial and ventricular myocytes is plastic up to stage 6 and set at stage 7. RESULTS Embryos Develop Normally in Culture The culture technique used in our studies supported normal development of chick embryos for up to 3 days (Fig. 1A). After 3 days in culture, both vasculogenesis and angiogenesis were initiated and the heart appeared normal (Fig. 1A and 1C). Immunohistochemical staining with MF20 (antibody directed against sarcomeric myosin heavy chain) showed the expected pattern of differentiation of cardiac myocytes (Fig. 1B). HFR at Stages 4 and 8 To determine the effect of position along the rostrocaudal axis on the fate of precursor cardiac cells, mesoderm and endoderm of the bilateral HFRs of stages 4 – 8 were cut and rotated by 180°. The number of embryos used, and the results of the controls and experimental manipulations are summarized in Table 1. The region of the HFR chosen for manipulation was based on the fate maps of Rawles (1943), Rudnick (1944), Stalsberg and DeHaan (1969), and Erhman and Yutzey (1999). The recent findings of Erhman and Yutzey (1999) define 1: the posterior margin of the HFR at the level of Hensen’s node at stage 5 and at the first condensing somite at stage 8 corresponding to the posterior boundary of nkx2.5 expression; 2: the lateral boundary close to the area pellucida–area opaca interface at stage 5, which is more lateral than described previously (Rawles, 1943); and 3: the medial boundary at the lateral edge of the prospective neural plate. In our experiments, the boundaries of the region cut and rotated at stage 4 are as shown in Fig. 2A. When this fragment was cut, rotated 180° (so that after rotation the prospective ventricular cells are located posterior to the prospective atrial cells), and allowed to develop for 24 hr, the embryo developed normally (Fig. 2B). In addition, the cardiac myocytes expressed myosin heavy chain as detected by staining with MF20 (Fig. 2C) and the heart beat as in a control embryo. To ensure that the HFR contained all of the prospective cardiogenic cells, a fragment of the HFR (mesoderm and endoderm) identical in size and location was removed (Fig. 2D) and the embryo was allowed to develop for 24 hr. The embryo formed head structures, an axis and somites, but a heart was not visualized under light microscopy (Fig. 2E), and myosin heavy chain expression was not detected after staining with MF20 (Fig. 2F). The same experiment was done to determine that the entire HFR was manipulated at stage 8. The HFR at stage 8 includes cells located both lateral to the axis and in the midline, anterior to the anterior intestinal portal. The anterior set of cells is technically difficult to include without severely damaging the embryo. Therefore, we restricted our manipulations to cutting and rotating, or removing as much of the lateral HFR as possible. Thus all manipulations at stage 8 resulted in a centrally located pulsating “heart” that expressed myosin heavy chain (asterisk Fig. 3C and 3F). Removal of a fragment whose posterior margin extended to the first somite (n ⫽ 6; Fig. 3A) resulted, after 24 hr of development, in three small “hearts” (one medial and two lateral; Fig. 3B) all expressing myosin heavy chain as detected by staining with MF20 (Fig. 3C). The lateral structures suggested that the entire lateral HFR was not removed. Therefore, a fragment that extended to the bottom margin of the second somite was removed (Fig. 3D) and in this case, after 24 hr, the two lateral “hearts” were not detected under light microscopy (Fig. 3E) and myosin heavy chain expression was not detected laterally with MF20 (Fig. 3F). Based on these data and mapping studies in our laboratory using DiI injected into subsets of mesodermal cells (manuscript in preparation), we were certain that the manipulations at stages 4 and 8 included the lateral and posterior boundaries of the HFR. Similar studies at stages 5, 6, and 7 were used to ensure that the entire HFR was cut and rotated 180°. Within 60 min of postsurgical incubation, the healing process was established; and after approximately 3 hr complete healing was observed at all stages studied. After 24 hr in culture, in situ hybridization was used to detect mRNA expression of VMHC1 or AMHC1 in the developing heart. Cutting the HFR Mesoderm and Endoderm As controls, the HFR was either not manipulated or cut but not rotated, and the embryos were assessed for morphologic development and expression of VMHC1 or AMHC1. Cutting alone had no effect on chamber specification as determined by VMHC1 and AMHC1 expression at stages 4, 5, 6, 7, and 8. Embryos with cut HFR (not rotated) were identical to unmanipulated control embryos both morphologically (normal beating hearts) and in regards to chamber specification. VMHC1 and AMHC1 mRNA localization are shown for cut control, stage 4 and 8 embryos [Fig. 4A and 4J (VMHC1) and 4C and 4L (AMHC1)]. DIVERSIFICATION OF CARDIAC MYOCYTES 125 Fig. 1. Chicken embryo grown in culture for 72 hr. A: Light micrograph (dissecting microscope) of an embryo cultured for 72 hr. This stage 15 embryo was placed in culture at stage 4 (Hamburger and Hamilton staging). Extensive vasculogenesis and angiogenesis are observed. B: Fluorescent microscopic montage of the same embryo after staining with MF20 (anti-sarcomeric myosin-heavy chain antibody). Normal heart morphology and differentiation of cardiac myocytes are seen. C: Bright field montage of the same embryo as in B. Rotation of the HFR Mesoderm and Endoderm at Stages 4 and 5 The majority of embryos in which the HFRs were cut and rotated at stage 4 developed to stage 10/11 or older at the end of 24 – 48 hr of incubation and were morpho- logically normal and appeared identical to the control embryos. The size, shape, and appearance of the heart was normal: a single median heart with normal looping and beating patterns. As in the control embryos, the beat originated in the posterior part of the heart. 126 PATWARDHAN ET AL. TABLE 1. Heart Development in Stages 4 to 8 Embryos, 24 Hours After Manipulation of the Mesoderm/Endoderm or Ectoderm Hamburger & Hamilton stage 4 4 4 4 5 5 6 6 7 7 Operation performed Mesoderm and endoderm cut controls Mesoderm and endoderm cut and rotated Ectoderm cut controls Ectoderm cut and rotated Mesoderm and endoderm cut controls Mesoderm and endoderm cut and rotated Mesoderm and endoderm cut controls Mesoderm and endoderm cut and rotated Mesoderm and endoderm cut controls Mesoderm and endoderm cut and rotated Total operated and survived at 24 hr 25 36 4 6 6 10 9 5 5 8 8 8 Mesoderm and endoderm cut controls Mesoderm and endoderm cut and rotated 11 22 8 8 Ectoderm cut controls Ectoderm cut and rotated 7 15 Phenotype Normal hearta Normal heart Normal heart Normal heart Normal heart Normal heart Normal heart 4 “hearts” Normal heart (4) attached and 1 median “heart” Normal heart (4) attached & 1 median “heart” Normal heart Normal heart Number with the given phenotype 22 30 4 4 6 9 8 5 4 7 9 22 4 11 a Normal heart, morphology as described by Rawles (1943), Olson and Srivastava (1996), Fishman and Chien (1997), and DeHaan (1965). At both stages 4 and 5 (data shown for stage 4), in situ hybridization using the VMHC1 probe localized the ventricular myocytes to the anterior part of the heart (Fig. 4B). The same was true for the unoperated (data not shown) and cut but not rotated controls (Fig. 4A). Atrial myocytes were located (staining with AMHC1) in the posterior part of the heart in the operated (Fig. 4D) as well as control embryos (Fig. 4C). Rotation of the HFR at stages 4 and 5 had no effect on heart morphogenesis, chamber formation, or overall development of the embryo. Rotation of the HFR Mesoderm and Endoderm at Stage 6 Rotation of the HFR at stage 6 was technically more difficult to perform than at the other stages. The mesodermal cells were very adherent to both the ectoderm and endoderm at this stage. Therefore, to cut and accurately rotate all of the mesodermal cells with the endoderm was not easy, with many embryos being damaged in the process, so much so, that they did not develop normally and were not considered in this study. Of the five embryos that were successfully cut and rotated and stained for expression of VMHC1 and AMHC1 (Table 1), one is represented in Fig. 4H and 4I. To ensure that the entire HFR was cut and rotated, the HFR was removed (Fig. 4E), the embryo was allowed to develop for an additional 24 hours (Fig. 4F), and then stained with MF20 to assess for myosin heavy chain expression (Fig.4G). When the entire HFR was cut and rotated 180°, four pulsating groups of tissues were formed. The anterior structures expressed VMHC1 (Fig. 4H arrows) and the posterior structures expressed AMHC1 (Fig.4I arrowheads). Rotation of the HFR Mesoderm and Endoderm at Stage 8 Rotation of the HFR at stage 8 resulted in condensation of the heart-forming cells into four groups of “hearts” on either side of the axis (Fig. 4K and 4M). The two anterior “hearts” (black arrows) were often attached to the two posterior “hearts” (black arrowheads). The anterior groups of beating cells expressed AMHC1 (Fig. 4M), whereas the posterior groups expressed VMHC1 (Fig. 4K). A medially located group of cells was also detected. In all of these cases, the medially located group of cells expressed VMHC1 (Fig. 4K). This expected result is explained by the fact that the anterior-most centrally located cardiac progenitor cells (anterior to the anterior intestinal portal) were not included in the cut-androtated HFR at stage 8. They remained along the medial axis and in all cases expressed VMHC1 (Fig. 4K white arrowhead). A similar group of medially located cells was seen but did not express AMHC1 (Fig. 4M, white arrowhead). Rotation of the HFR Ectoderm The HFR ectoderm alone was rotated to determine whether it was the source of the diversification signal. After rotating the ectoderm overlying the HFR at stages 4 and 8, the embryos developed normally and had normal, single, looped, beating heart tubes located at the midline. Ventricular and atrial chamber specification was normal (detected by VMHC1 and AMHC1 expression) (Fig. 5B and 5D, respectively) as in the control embryos (Fig. 5A and 5C, respectively). DiI Labeling of Rotated Cells To determine the migratory patterns of cells in the HFR after cutting and rotation at stages 4 – 8, we labeled either the new anterior or the new posterior edge with DiI in different embryos. Visualization of the DiI fluorescence after 16 –24 hr of incubation identified the destination of these labeled cells. When the anterior cells on the left side of the embryo (ventral side up) of the rotated HFR at stage 4 (data DIVERSIFICATION OF CARDIAC MYOCYTES 127 Fig. 2. The heart-forming region (HFR) at stage 4. Ventral view of the same embryo at different stages of development is seen in A–C and another embryo in D–F. A: Stage 4 embryo immediately following cutting of the mesoderm and endoderm of the HFR. Arrowheads point to the cut HFR. These pieces correspond to the regions fated to become the heart (Rawles, 1943; Rosenquist and DeHaan, 1966). The mesoderm and endoderm in this HFR was rotated 180° bilaterally and the embryo was cultured for 24 hr. B: Bright field image of the same embryo as in A at approximately stage 12, 24 hr postsurgery. Normal morphology was observed. Hd, head; Ht, heart; S, somites. C: Fluorescent image of the same embryo as in B, reacted with anti-myosin-heavy chain antibody, MF20. Normal cardiomyocyte differentiation occurred as assessed by expression of myosin. D: A stage 4 embryo with the HFR endoderm and mesoderm removed. Arrowheads point to the removed HFR. E: Bright field image of the same embryo as in D 24 hr postsurgery. Although head and somites are visible, no heart structures were observed. F: Fluorescent image of the same embryo as in E, reacted with anti-myosin-heavy chain antibody, MF20. Cardiomyocyte differentiation was not detected. not shown), or stage 8 (Fig. 6A) were marked, the labeling was confined to the anterior portion (encircled by black dots) of the anteriorly located “heart” (Fig. 6B). The more posterior part of the heart was only stained when posterior cells were labeled at stages 4 – 8 (data not shown). The embryo in Fig. 6A was at stage 8 at the time of cut and rotation of the HFR. Twenty hours after surgery (Fig. 6B), four “hearts” were detected. In addition, a midline heart was detected as described earlier (white arrowhead). 128 PATWARDHAN ET AL. Fig. 3. The heart-forming region (HFR) at stage 8. Ventral views of the same embryo at different stages of development are shown in A–C and another embryo in D–F. A: The mesoderm/endoderm HFR at stage 8, marked by arrowheads was removed. The * marks the location of the midline HFR above the anterior intestinal portal that was not removed. The posterior boundary of the HFR was at the level of the first somite. B: Bright field image of the embryo in A, 24 hr postsurgery. The embryo was at approximately stage 14. Head and somites are visible. Small “hearts” (arrows) are seen on either side of the axis. C: Fluorescent image of the embryo in B reacted with MF20. The lateral “hearts” (arrow) and centrally located “heart” (*) expressed myosin heavy chain suggesting that not all of the HFR was removed. D: The mesoderm/endoderm HFR at stage 8, marked by arrowheads was removed. The posterior boundary of the HFR was at the level of the second somite. E: Bright field image of the embryo in D, 24 hr postsurgery. The embryo was at approximately stage 14. Head and somites are visible. No laterally located “hearts” were observed. A centrally located “heart” (*) was observed. F: Fluorescent image of the embryo in E. Only a centrally located “heart” expressed myosin (*). These cells arose from the HFR most anteriorally located (*), above the anterior intestinal portal that was not removed. DIVERSIFICATION OF CARDIAC MYOCYTES Figure 4. 129 130 PATWARDHAN ET AL. Fig. 5. Rotation of ectoderm in the heart-forming region (HFR) does not alter normal cardiac morphogenesis and chamber specification. A: Stage 8 embryo in which the HFR ectoderm was cut but not rotated and hybridized with VMHC1 probe, localizing the ventricular myocytes to the anterior part of the heart. B: Stage 8 embryo in which HFR ectoderm was cut and rotated and probed with VMHC1. Staining is confined to the anterior portion of the heart. C: Stage 8 cut control embryo, hybridized with AMHC1 probe, localizing the atrial myocytes to the posterior part of the heart. D: Stage 8 embryo in which the HFR ectoderm was cut and rotated and probed with AMHC1. Staining is confined to the posterior portion of the heart. DISCUSSION the plasticity of the cardiac myocyte precursor fate over time and the influence of position along the rostrocaudal axis on this fate. We rotated the HFR by 180°, bringing the posteriorly located atrial precursors to an anterior position normally occupied by ventricular precursor cells, and bringing the anteriorly located ventricular cells to the posterior position. We ensured that entire HFR was included within the manipulated fragment. This was done by removing the HFR, and staining for myosin heavy chain expression. The assumption was that if Cardiac myocytes, once committed, continue to diversify into various sublineages such as atrial, ventricular, and conduction system myocytes. The factors influencing this diversification are not well understood. The spatial order of the cardiac precursors within the primitive streak prior to gastrulation reflects their ultimate fate and destination. However, this influence appears to be purely positional at this stage because the cells are neither specified nor determined (Garcia-Martinez and Schoenwolf, 1993). Our study examines, in vivo, Fig. 4. (overleaf.) Effects of position on diversification of cardiac precursor cells. Whole mount in situ hybridization of chick embryos using digoxygenin labeled VMHC1 or AMHC1 antisense riboprobe. A–D: Change in position of endodermal and mesodermal cells in the heartforming region (HFR) at stages 4 and 5 did not alter normal heart morphogenesis or chamber orientation. A: Stage 4 cut control (HFR mesoderm and endoderm cut but not rotated) embryo 24 hr postsurgery, hybridized with VMHC1 probe, localizing the ventricular myocytes to the anterior part of the heart. B: Stage 4 embryo in which mesoderm and endoderm of the HFR was cut and rotated, 24 hr postsurgery, probed with VMHC1. Staining is confined to the anterior portion of the heart. C: Stage 4 cut control embryo 48 hr postsurgery, hybridized with AMHC1 probe, localizing the atrial myocytes to the posterior part of the heart. D: Stage 4 embryo in which mesoderm and endoderm of the HFR was cut and rotated, 24 hr postsurgery, probed with AMHC1. Staining is confined to the posterior portion of the heart. E–I: At stage 6 cut and rotation of the HFR results in normal chamber orientation and multiple heart formation. E–G: The same embryo observed at different stages of development. E: Stage 6 embryo with the HFR marked by arrowheads was removed. F: Bright field image of the embryo 24 hr postsurgery at approximately stage 13. G: Fluorescent image of the embryo in F reacted with MF20. All of the HFR was removed since no myosin expression was observed. The HFR of a stage 6 embryo was cut and rotated 180° and VMHC1 mRNA localization 24 hr postsurgery is shown in H, and AMHC1 expression in I. Arrows point to anterior “hearts” and arrowheads point to posterior “hearts.” J–M: At stage 8, cut and rotation of the HFR resulted in reversed chamber orientation and multiple “heart” formation. J: Stage 8 cut control embryo 24 hr postsurgery, hybridized with VMHC1 probe, localizing the ventricular myocytes to the anterior part of the heart. K: Stage 8 embryo 24 hr postsurgery, in which mesoderm and endoderm of the HFR was cut and rotated, probed with VMHC1. Staining is exclusive to the posterior “hearts” (arrowheads). The centrally located “heart” from unmanipulated cells at the anterior intestinal portal expressed VMHC1 (white arrowhead). L: Stage 8 cut control embryo 36 hr postsurgery, hybridized with AMHC1 probe, localizing the atrial myocytes to the posterior part of the heart. M: Stage 8 embryo 24 hr postsurgery, in which mesoderm and endoderm of the HFR was cut and rotated, probed with AMHC1. Staining is confined to the anterior groups of “hearts” (arrows). Although a centrally located “heart” was present (white arrowhead), it did not express AMHC1. Arrows indicate the multiple anterior hearts expressing AMHC1, and black arrowheads indicate posterior hearts expressing VMHC1. DIVERSIFICATION OF CARDIAC MYOCYTES Fig. 6. Cells maintain their relative positions in the rotated mesoderm and endoderm of the heart-forming region (HFR). DiI was visualized using fluorescence microscopy. Top: Embryo in which a few cells on the left side (ventral-side up) at the anterior end of the cut and rotated mesoderm and endoderm of the HFR were labeled with DiI at stage 8. The bright field image of the stage 8 embryo at 0 hours post-cut and rotate was superimposed on the fluorescent image. Arrow points to the 131 subset of DiI-labeled cells. Bottom: The same embryo, 24 hr postsurgery has developed approximately 16 pairs of somites. The bright field image of the embryo was superimposed on the fluorescent image. Two attached “hearts” were observed on each side of the axis. The anterior heart on the left side (encircled by black dots) contains DiI-labeled cells. The posterior left heart is encircled by green dots. The centrally located “heart” is marked by a white arrowhead. 132 PATWARDHAN ET AL. any precursor cardiac cells were not included in the HFR fragment that was removed, they would have differentiated into cardiac myocytes and would be detected by staining with an anti-myosin heavy chain antibody (MF20). The lineage markers AMHC1 and VMHC1, expressed by cells in the HFR, were used to identify the developmental fate of cardiac myocytes in their new positions. Rotation of the HFR at stages 4 and 5 had no effect on chamber specification, as seen by detection of lineage markers, or cardiac morphogenesis after 24 hr of postsurgical development. Cells from the posterior margin of the HFR would normally express AMHC1 whereas those from the anterior margin would express VMHC1 at this time in development. When both groups of cells were rotated to new anterior or posterior positions, they changed their fate as determined by expression of myosin mRNA. The cells that were located anteriorly always expressed VMHC1 and the cells that were located posteriorly always expressed AMHC1. An alternative explanation of our results at stage 4 –5 is that, because differentiation occurs in a rostralto-caudal direction, the rotated new anterior cells are merely expressing VMHC1 prior to their expression of AMHC1 (Bisaha and Bader, 1991). To elaborate, during normal development, all cardiac myocytes initially express VMHC1. It is only later in development (stage 9 and later) that the posterior cells begin to express AMHC1. However, the rotated new posterior cells express AMHC1 (this never occurs during normal development), which is due to respecification. At stage 6, the majority of the cut and rotated HFR embryos showed cardia bifida with two to four groups of “hearts.” VMHC1 was expressed in the anteriorly located cells and AMHC1 was expressed in the posteriorly located cells. At stage 6, with a change in the rostro/caudal position of cardiogenic cells, a change in cell fate was observed. Second, unlike the results obtained at stages 4 and 5, cardia bifida or multiple “hearts” (up to four) was observed. Stage 6 is when precardiac cells begin their movement anteriorly towards the midline. Between stages 5 and 8 many changes in the extracellular matrix components and surface molecules on precardiac mesodermal cells allow this migration to occur (Linask and Lash, 1986). Also, the stage 6 time-point is pivotal in the process of differentiation (Gonzales-Sanchez and Bader, 1990; Montgomery et al, 1994). These authors showed that at stage 4 most cardiac progenitor cells were sensitive to the deleterious effects of 12-0-tetradecanoylphorbol 13-acetate (TPA) or Bromodeoxyuridine (BrdU) (which inhibited differentiation), whereas at stage 8 precardiac cells were refractory. At stage 6, precardiac cells in the caudal HFR were sensitive and remained undifferentiated, whereas the more rostral cells were refractory to the effects of TPA and BrdU. These findings suggest that stage 6 represents a timepoint in cardiac development when events crucial to normal morphologic heart development and differenti- ation of cardiac myocytes occurs. Along these lines, the posterior cells sensitive to BrdU (possibly because they have not yet undergone a crucial step in the differentiation process) at stage 6 include the atrial progenitor cells. When these cells are relocated to their new position, at this crucial point in the process of differentiation, they may take on a ventricular fate only because they had not already completely differentiated into atrial cells at the time of cut and rotation. Note that differentiation proceeds in a rostral-to-caudal direction. Therefore, ventricular cells differentiate before atrial cells. Because the original anterior cells (with a ventricular fate), when relocated to their new posterior location, express AMHC1, they were respecified. The data we obtained also suggest that the surrounding milieu plays an important role in regulation of this process. The region of the heart comprising the ventricle is larger than that comprising the atria. Yet, when the anterior cells (prospective ventricular cells) were placed in a posterior position and expressed AMHC1, the AMHC1-positive area (Fig. 4I) was smaller than the VMHC1-positive area (Fig. 4H). This suggests that when the anterior cells were placed in the posterior region, only a small percentage of them expressed AMHC1, possibly because only a select region around the posterior margin of the HFR supports atrial differentiation and AMHC1 expression. The cardia bifida or multiple heart formation may be explained by interference of the migratory pathway of the cardiogenic cells by the cut and rotation manipulation. Similar findings of multiple “hearts” were seen at stages 7 and 8. Although we do not know the exact reason for the disruption of heart tube organization at stages 6 – 8, it is a common finding in experiments involving alterations in the endoderm at these stages (DeHaan, 1963; Linask and Lash, 1988; Satin et al., 1988; Osmond et al., 1991). One possible explanation could be a disturbance of a fibronectin concentration gradient between the endoderm and mesoderm as reported by Linask and Lash (1986). They reported that this gradient is established at stage 5⫹/6 along the rostro-caudal axis, with the highest concentration at the rostral end of the embryo. The cardiac precursors migrate on this gradient via a haptotactic mechanism in a caudal-to-rostral direction to meet at the head fold region in the midline at stage 8 (Linask and Lash, 1986). The HFR rotation at stage 4 would have no effect on cardiac morphogenesis because the rotation and wound healing occurred well before the gradient began to form. After stage 5⫹, any disturbance of the gradient would disrupt the morphogenetic migration of the cardiac precursors, resulting in their aggregation at multiple sites along their migratory route and the formation of multiple “hearts.” There is some controversy over the existence of a fibronectin gradient, though this may be due to a difference in fibronectin expression between different species [chicken (Linask and Lash, 1986) versus rat (Suzuki et al., 1995)]. It is clear from previous studies that disturbances of the endoderm always result in formation of multiple DIVERSIFICATION OF CARDIAC MYOCYTES “hearts” (tubular structures of beating tissue) (DeHaan, 1963; Satin et al., 1988; Linask and Lash, 1988). In these studies, as well as ours, at stages ⬎5, the mechanical or chemical disruption of either epithelialization of mesoderm (Linask et al., 1997) or of extracellular matrix may well be the basic mechanism underlying the formation of multiple hearts. In any case, the interpretation of our findings, concerning the respecification of atrial and ventricular myocytes would not be altered regardless of the underlying mechanisms (discussed above) that provide guidance for the directional movement of heart primordia during cardiac morphogenesis. At stages 7 and 8, four to five groups of independently beating cells were seen. Following the HFR rotation, ventricular myocytes were always found posterior to atrial myocytes suggesting that by stage 7– 8 the fate of these cells has been determined and was unaffected when their position along the rostro-caudal axis was changed. A centrally located pulsating group of cells was always detected that expressed VMHC1. A group of cells in the region of the anterior intestinal portal remain undisturbed after the cut and rotation manipulation of the HFR. These cells remained in their central position and express VMHC1, as expected. Our results show that between stages 4 and 6, the choice between an atrial and ventricular fate, for a precursor cell, is ultimately dependent on its position along the rostro-caudal axis. The anteriorly located cells always become ventricular, and the posteriorly located cells become atrial, irrespective of their original position and fate. The precursor cells lose this plasticity at the stage of cardiac crescent formation (stage 8). Our findings explain at the molecular level the transplantation experiments (stages 5–7) of Satin et al. (1988). In these experiments, atrial tissue that was transplanted into the ventriculogenic region exhibited the slower frequency of beating characteristic of ventricular myocytes. We suggest that at a molecular level the alteration of beat frequency seen in DeHaan’s experiments is based on a respecification of cell fate, which lead to a different developmental program. This in turn suggests that the transplanted atrial progenitor cells altered their beat frequency because they became ventricular myocytes. A previous study (Yutzey et al., 1995) has shown that the undifferentiated cardiogenic tissue from stage 4 and older chick embryos can diversify into distinct atrial and ventricular lineages when grown in vitro in a defined minimal medium. This study also showed that the diversification potential of these cells could be changed through retinoic acid treatment only at early stages of development. In that study, the precursor cells were isolated in vitro from any new interactions with other cells or extracellular influences, allowing them to undergo diversification as dictated by their original program. In our in vivo study, the cardiac progenitors were placed in a new environment (rotated 180° along the anterior-posterior axis), where they were subjected to new environmental signals that reset 133 their diversification program. Yutzey et al. (1995) also showed that while exposure to retinoic acid was capable of respecifying cells from the HFR of early stage embryos, cells from embryos older than stage 8 were refractory to the retinoic acid-mediated fate change. These observations are completely consistent with our present study that shows that the cells of the HFR at stages 4 – 6 are plastic and can be respecified, whereas at stage 8 they are no longer plastic and cannot be respecified. We propose that the maintenance of the relative positions of ventricular myocytes, with respect to atrial myocytes, after rotating the HFR in the present study is through a process of respecification of the diversified fate of the committed cells. This respecification may be mediated via factors intrinsic or extrinsic to the HFR. When the environmental signals do not match the intrinsic fate decision, as in the case of cut and rotated endoderm-mesoderm HFR at stages 4 – 6, information from the surrounding environment dominates and the cell fates are respecified according to these environmental conditions. This could represent a backup system of diversification that ties the intrinsic cell fate to the external environment, which in this case is position along the rostro-caudal axis. Any mistakes of positioning during gastrulation or the subsequent migration of the cells would be rectified by respecification using external environmental cues. What is the source of the diversification signal? We were unable to examine the effects of cutting and rotating the HFR endoderm alone because of difficulties in separating endodermal and mesodermal layers consistently and accurately. However, our thinking is as follows: If the signal for respecification were in the endoderm, we would have expected that after rotation the myocyte precursors would diversify into the same lineage as they would have if not rotated. In other words, after rotation, the atrial precursors would still have become atrial myocytes and ventricular precursor cells, ventricular myocytes. Since their fate was respecified at stages 4 – 6, the signal for diversification must originate from somewhere other than the HFR endoderm. This is consistent with the finding that rotation of the early endoderm at stage 3⫹ has no effect on chamber specification or cardiac morphogenesis (Inagaki et al., 1993). We thought that the signal might originate from the ectodermal layer. Therefore, we examined what would happen if only the ectoderm of the HFR was cut and rotated, leaving the mesodermendoderm in the normal orientation. A normal-looking embryo with normal heart morphology, chamber specification, and beating frequency resulted from these manipulations. This indicated that the orientation of the HFR ectoderm had no effect on cardiogenesis at the stages studied and suggests that the signal for respecification is not in the HFR ectoderm. Our experimental data suggest that it is unlikely that the signal is from any of the layers of the HFR itself, since rotation of the HFR ectoderm alone had no effect and rotation of the HFR mesoderm-endoderm resulted in respecification 134 PATWARDHAN ET AL. (stages 4 – 6). If the signal(s) were present in either the endoderm or the mesoderm of the HFR, one would have expected the cells to develop normally regardless of anterior-posterior orientation. We believe that the most likely source for the signal is therefore extrinsic to the HFR. EXPERIMENTAL PROCEDURES Embryo Culture and Microsurgical Manipulations Fertilized white leghorn chicken eggs from Truslow farms (Chestertown, MD) were incubated at 37°C in a humidified egg incubator until the required stages of development. Embryos were dissected out of the eggs using sterile filter paper rings and grown ventral-side up on a 50% albumin– 0.3% agarose matrix in 35-mm petri dishes (G.C. Schoenwolf, personal communication). Embryos between stages 4 and 8 (Hamburger and Hamilton staging, 1951) were used. Indirect Immunofluorescence Staining Embryos were cultured as described, washed in phosphate buffered saline with 0.1% Tween-20 (PBT) and fixed in ice-cold methanol. Embryos were blocked in 10% normal goat serum (NGS)/5% bovine serum albumin (BSA)/PBT, incubated with MF20 (anti-sarcomeric myosin heavy chain antibody; Bader et al., 1982) in 5% NGS/2.5% BSA/PBT, washed extensively in PBT, reacted with goat-anti-mouse conjugated to rhodamine (secondary antibody), and viewed under a fluorescence microscope. All incubations and washes were at 4°C, overnight. Cutting and Rotating the Cardiogenic Mesoderm and Endoderm The cardiogenic areas were identified based on published fate maps (Rawles, 1943; DeHaan, 1965; Rosenquist, 1970; Ehrman and Yutzey, 1999). Under a dissecting microscope, the embryo was placed ventral-side up and, using glass needles, a rectangular incision was made in the ALP through the endoderm and mesoderm, leaving the underlying ectoderm undisturbed. This cut piece of tissue was rotated 180° along the rostro-caudal axis. The entire cardiogenic area [based on the fate maps and studies of Rawles (1943), DeHaan (1965), Yutzey et al. (1995), and Ehrman and Yutzey (1999)] was included in the cut rectangle. To ensure that all of the HFR was included in the cut and rotated fragment, embryos in which the HFR was removed were cultured for 24 hr and stained with MF20 for expression of sarcomeric myosin heavy chain. In addition, two other types of controls were used: embryos in which the cardiogenic areas were cut but not rotated, and those that were not manipulated. After incubating for 24 – 48 hr at 37°C in 5% CO2 and 95% humidity, experimental and control embryos were photographed and fixed at 4°C in 4% paraformaldehyde/1 ⫻ phosphate buffered saline (PBS) overnight, with gentle agitation. The embryos were washed in 1 ⫻ PBS and dehydrated in a graded series of dilutions of ethanol: 1 ⫻ PBS/0.1% TritonX-100 (PBTX) and stored at ⫺20°C in 70% ethanol:PBT until used in situ hybridization. In Situ Hybridization In situ hybridization was carried out as described by Wilkinson (1992) with minor modifications. Embryos were digested with 10 g/ml proteinase K for 30 min at room temperature, incubated in 2% BSA for 20 min, and fixed in 4% paraformaldehyde/0.2% gluteraldehyde for 20 min on ice prior to prehybridization. These solutions were made in PBTX. Hybridization was carried out in 50% formamide at 65°C overnight with 2 g/ml digoxygenin (DIG)-labeled probe. Probes were used multiple times and stored at ⫺70°C between uses. Hybridized probes were detected using anti-DIG antibody at 1:2,000 dilution, and the substrate reaction was accomplished using the Genius (Boehringer Mannheim) detection system. After optimum staining (in this case for not more than 15 min) embryos were rinsed in PBTX and photographed. Synthesis of VMHC1 and AMHC1 Probes Drs. D. Bader and K. Yutzey provided the pVMHC1 and pAMHC1 clones, respectively. VMHC1 was used as a marker for ventricular myocytes and AMHC1 for atrial myocytes. Digoxigenin-UTP labeled antisense RNA probes were prepared using the Genius 4 RNA labeling system (Boehringer Mannheim). Probes were quantitated by transfer of DIG-incorporated RNA from an agarose gel to NYTRAN (Schleicher & Schuell) and the labeled RNA was detected using the Genius detection system. Cutting and Rotating Ectoderm Overlying the HFR Embryos were dissected out of the eggs and cultured on albumin agar plates as above. The blastoderm was gently detached from the vitelline membrane and embryos were placed dorsal-side up. Using clearly visible cardiogenic mesodermal areas as landmarks, ectoderm of the HFR was carefully cut, leaving the underlying mesoderm and endoderm undisturbed. The cut rectangular piece of ectoderm was rotated 180° and the cuts were allowed to heal for about 60 min at 37°C before flipping the embryos to ventral-side up. Further incubation and processing for in situ hybridization was as before. DiI Labeling of Cells Embryos were dissected and cultured as described. After cutting through the mesoderm and endoderm of the HFR, these tissues were rotated. A subset of cells either at the anterior or posterior end of the rotated piece were labeled with DiI (Molecular Probes). A stock solution of 5mg/ml DiI in dimethly sulfoxide was prepared and stored at ⫺20°C. A working solution of 10 g/ml DiI in PBS was freshly diluted prior to use. Marking was done using glass capillary needles drawn using a Narishige needle puller (Narishige, Japan). DiI was placed on a subset of cells using a Narishige pres- DIVERSIFICATION OF CARDIAC MYOCYTES sure microinjection apparatus. This procedure was carried out at stages 4 and 8. After labeling, bright field and fluorescent images (using a 4⫻ objective) were obtained to mark the point of DiI labeling at 0 hours, the embryos were incubated for 16 –24 hr at 37°C, and bright field and fluorescent images were recorded to determine the subsequent location of the labeled cells. The bright field and fluorescent images at 0 hours were superimposed using Adobe Photoshop 5.5 thereby showing the exact position of the labeled cells. The destination of these cells was determined 16 –24 hours later by superimposing the bright field and fluorescent images obtained. ACKNOWLEDGMENTS We thank Dr. Abhay Redkar for his comments and valuable discussions. This work was supported by a grant from the National Institutes of Health, HL52052 to J.L. REFERENCES Bader D, Masaki T, Fischman DA. 1982. Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J Cell Biol 95:763–770. Bisaha JG, Bader D. 1991. Identification and characterization of a ventricular specific myosin heavy chain VMHC1: expression in differentiating cardiac and skeletal muscle. Dev Biol 148:355– 364. DeHaan RL. 1963. Organization of the cardiogenic plate in the early chick embryo. Acta Embryol Morphol Exp 6:26 –38. DeHaan RL. 1965. Morphogenesis of the vertebrate heart. In DeHaan RL, Ursprung H, editors. Organogenesis. New York: Holt, Rinehart & Winston, Inc. p 377– 420. Erhman LA, Yutzey KE. 1999. Lack of regulation in the heart forming region of avian embryos. Dev Biol 207:163–175. Fishman MC, Chien KR. 1997. Fashioning the vertebrate heart: earliest embryonic decisions. Development 124:2099 –2117. Garcia-Martinez V, Schoenwolf GC. 1993. Primitive streak origin of the cardiovascular system in avian embryos. Dev Biol 159: 706 –719. Gonzalez-Sanchez A, Bader D. 1990. In vitro analysis of cardiac progenitor cell differentiation. Dev Biol 139:197–209. Hamburger V, Hamilton HL. 1951. A series of normal stages in the development of the chick embryo. J Morphol 88:49 –92. Inagaki T, Garcia-Martinez V, Schoenwolf GC. 1993. Regulative ability of the prospective cardiogenic and vasculogenic areas of the primitive streak during avian gastrulation. Dev Dyn 197: 57– 68. 135 Linask KK, Lash JM. 1986. Precardiac cell migration: fibronectin localization at mesoderm-endoderm interface during directional movement. Dev Biol 114:87–101. Linask KK, Lash JM. 1988. A role for fibronectin in the migration of avian precardiac cells. II. Rotation of the heart forming region during different stages and its effects. Dev Biol 129:324 –329. Linask KK, Knudsen KA, Gui YH. 1997. N-Cadherin-catenin interaction: necessary component of cardiac cell compartmentalization during early vertebrate heart development. Dev Biol 185:148 –164. Litvin J, Montgomery M, Gonzalez-Sanchez A, Bisaha JG, Bader D. 1992. Commitment and differentiation of cardiac myocytes. Trends Cardiovasc Med 2:27–32. Montgomery M, Litvin J, Gonzalez-Sanchez A, Bader D. 1994. Staging of commitment and differentiation of avian cardiac myocytes. Dev Biol 164:63–72. Olson EN, Srivastava D. 1996. Molecular pathways controlling heart development. Science 272: 671– 676. Osmond MK, Butler AJ, Voon FCT, Bellairs R. 1991. The effects of retinoic acid on heart formation in the early chick embryo. Development 113:1405–1417. Rawles M. 1943. The heart forming areas of the early chick blastoderm. Physiol Zool 16:22– 42. Rosenquist GC. 1970. Location and movements of cardiogenic cells in the chick embryo: the heart-forming portion of the primitive streak. Dev Biol 22:461– 475. Rudnick D. 1944. Early history and mechanics of the chick blastoderm. A review. Quart Rev Biol 19:187–212. Sater AK, Jacobson AG. 1989. The specification of heart mesoderm occurs during gastrulation in Xenopus laevis. Development 105: 821– 830. Sater AK, Jacobson AG. 1990. The restriction of the heart morphogenetic field in Xenopus laevis. Dev Biol 140:328 –336. Satin J, Fuji S, DeHaan RL. 1988. Development of cardiac beat rate in early chick embryo is regulated by regional cues. Dev Biol 129:103– 113. Stalsberg H, DeHaan RL. 1969. The precardiac areas and formation of the tubular heart in the chick embryo. Dev Biol 19:128 –159. Suzuki HR, Solursh M, Baldwin HS. 1995. Relationship between fibronectin expression during gastrulation and heart formation in the rat embryo. Dev Dyn 204:259 –277. Wilkinson DG. 1992. Whole mount in situ hybridization of vertebrate embryos. In Wilkinson DG, editor. In situ hybridization: a practical approach. Oxford, UK: Oxford University Press. p 75– 83. Yutzey KE, Bader D. 1995. Diversification of cardiomyogenic cell lineages during early heart development. Circ Res 77:216 –219. Yutzey KE, Gannon M, Bader D. 1995. Diversification of cardiomyogenic cell lineage in vitro. Dev Biol 170:531–541. 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.