Dynamic positional fate map of the primary heart-forming region

NIH Public Access Author Manuscript Dev Biol. Author manuscript; available in PMC 2010 August 15. NIH-PA Author Manuscript Published in final edited form as: Dev Biol. 2009 August 15; 332(2): 212–222. doi:10.1016/j.ydbio.2009.05.570. Dynamic Positional Fate Map of the Primary Heart-Forming Region Cheng Cui1,6, Tracey J. Cheuvront1, Rusty D. Lansford2, Ricardo A. Moreno-Rodriguez3, Thomas M. Schultheiss4, and Brenda J. Rongish1,5 1Department of Anatomy and Cell Biology, University of Kansas Medical Center, Kansas City, KS 66160; 2Biology Division, California Institute of Technology, Pasadena, CA 91125; 3Department of Cell Biology and Anatomy, Medical University of South Carolina, Charleston, SC 29425; NIH-PA Author Manuscript 4Department of Anatomy and Cell Biology, Technion-Israel Institute of Technology, Haifa, Israel 31096. Abstract NIH-PA Author Manuscript Here we show the temporal-spatial orchestration of early heart morphogenesis at cellular level resolution, in vivo, and reconcile conflicting positional fate-mapping data regarding the primary heart-forming field(s). We determined the positional-fates of precardiac cells using a precision electroporation approach in combination with wide-field time-lapse microscopy in the quail embryo, a warm-blooded vertebrate (HH Stages 4 through 10). Contrary to previous studies, the results demonstrate the existence of a “continuous” circle-shaped heart field that spans the midline, appearing at HH Stage 4, which then expands to form a wide arc of progenitors at HH stages 5–7. Our time-resolved image data show that a subset of these cardiac progenitor cells do not overlap with the expression of common cardiogenic factors, Nkx-2.5 and Bmp-2, until HH Stage 10, when a tubular heart has formed, calling into question when cardiac fate is specified and by which key factors. Sub-groups and anatomical bands (cohorts) of heart precursor cells dramatically change their relative positions in a process largely driven by endodermal folding and other large-scale tissue deformations. Thus, our novel dynamic positional fate maps resolve the origin of cardiac progenitor cells in amniotes. The data also establish the concept that tissue motion contributes significantly to cellular position fate — i.e., much of the cellular displacement that occurs during assembly of a midline heart tube (HH Stage 9) is NOT due to “migration” (autonomous motility), a commonly held belief. Computational analysis of our time-resolved data lays the foundation for more precise analyses of how cardiac gene regulatory networks correlate with early heart tissue morphogenesis in birds and mammals. © 2009 Elsevier Inc. All rights reserved. 5Corresponding author: Brenda J. Rongish, Anatomy and Cell Biology, WHW 1008, University of Kansas Medical Center, 3901 Rainbow Blvd., Kansas City, KS 66160, USA, Tele: 913/588-1878, Fax: 913/588-2710, E-mail: brongish@kumc.edu. 6Current address for Cheng Cui, Laboratory of Developmental Biology, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD 20892, USA Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Cui et al. Page 2 Keywords NIH-PA Author Manuscript time-lapse; cardiovascular; cardiac development; heart development; imaging; fate-map; cell tracking; avian; embryogenesis; primary heart field Introduction Avian and human embryonic hearts share similar morphologies (De la Cruz et al., 1977; De la Cruz et al., 1983; De La Cruz et al., 1989; Abu-Issa and Kirby, 2008) and many cardiovascular malformations found in the chicken embryo are similar to those found in humans (Nishibatake et al., 1987). However, unlike mammals, the avian embryo is readily amenable to ex ovo culture and imaging. This optical accessibility allows direct observation of the cellular movements comprising heart formation in a warm-blooded experimental system. NIH-PA Author Manuscript During amniote cardiogenesis, epiblast cells ingress through the primitive streak and the newly formed mesodermal cells move cranio-laterally to segregate into two layers: the splanchnic mesoderm and the somatic mesoderm. At Hamburger and Hamilton (HH) Stage 6 (Hamburger and Hamilton, 1951), as the quail anterior endoderm folds, heart progenitor cells in the splanchnic mesoderm translocate to the midline and fuse to form a tubular heart. As the “foregut”, or anterior intestinal portal (AIP), regresses caudally, a primitive heart “trough” is formed; open along its dorsal aspect (Moreno-Rodriguez et al., 2006; Stalsberg and DeHaan, 1969). The anterior and posterior portions of the dorsal roof of the trough continue to close, and the tube elongates bidirectionally (Moreno-Rodriguez et al., 2006) forming a “linear” heart tube between HH stages 9 and 11. The linear heart tube consists of the apical portions of both ventricles. By HH Stage 11 the heart begins to bulge to the embryonic-right. At HH stage 12 the beating heart has formed a loop, and is comprised of a proximal primitive outlet, which connects to the two ventral aortae cephalically, the apical portions of both ventricles, the primitive atria and a primitive inlet (De La Cruz et al., 1989). Beginning with Rawles (Rawles, 1936; Rawles, 1943), researchers have proposed positional fate maps of the heart forming regions by observing the location of heart precursor cells at progressive developmental stages (DeHaan, 1963; Garcia-Martinez and Schoenwolf, 1993; Rosenquist, 1970; Stalsberg and DeHaan, 1969; and reviewed in Abu-Issa et al., 2004; AbuIssa and Kirby, 2007; Yutzey and Kirby, 2002). NIH-PA Author Manuscript Presumptive molecular “markers” have also been used to identify precardiac cells in the primary heart -forming field (Ehrman and Yutzey, 1999; Schultheiss et al., 1995; Yuan and Schoenwolf, 2000). Among these, the homeobox gene, Nkx-2.5 (Evans, 1999; Harvey, 1996; Schultheiss et al., 1995; reviewed in Olson, 2006), Bmp-2 (Schultheiss et al., 1997) and the transcription factor, GATA-4, along with other GATA family members (Jiang et al., 1998; Kostetskii et al., 1999), are commonly acknowledged as markers of cardiogenesis. GATA-4 and Nkx-2.5 are expressed in cardiogenic precursors in a pattern that largely overlaps with the Bmp-2 expression in the endoderm. Nkx-2.5 and Bmp-2 gene expression is observed in a crescent shape pattern at HH stages 5–8 (Schultheiss et al., 1995), with Bmp-2 expressed more caudally than Nkx-2.5 (Schultheiss et al., 1997; Schultheiss et al., 1995). In contrast, GATA-4 is absent at the anterior midline prior to HH8 (Jiang et al., 1998; Kostetskii et al., 1999). Numerous other studies have focused on if, and when, precardiac cells are pre-patterned to occupy a certain positional fate in the tubular heart. These studies generated a wide range of hypotheses regarding the spatial and temporal boundaries of the heart-forming region of amniotes. While valuable, this literature is difficult to reconcile into a unified perspective, and underscores the critical need for unbiased, dynamic cell and tissue positional fate maps. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 3 NIH-PA Author Manuscript The reliability and interpretation of previous heart fate-mapping studies was limited by the existing technical methods of the time. For example, early workers such as DeHaan, Stalsberg and Rosenquist, could not exclusively label mesodermal cells for tracking, fluorescently or otherwise. Use of lipophilic fluorescent dyes, such as DiI, does not permit simultaneously tagging of large numbers of mesodermal cells at distinct locations in the same embryo — which, in turn, precludes computationally mapping their relative anterior-posterior and dorsal-ventral positions at high spatio-temporal resolution (Ehrman and Yutzey, 1999; Hochgreb et al., 2003; Redkar et al., 2001; Xavier-Neto et al., 2001). In short, cellular resolution, “real-time” precardiac cellular displacements were not recorded in previous studies. Using a precision DNA plasmid electroporation protocol, and our well-established computational time-lapse microscopy, we show the trajectories of cardiac precursor cells as they contribute to forming the tubular heart. A major advantage of this approach is that the technology allows one to “run the movies backward”. This important feature permits an observer to identify cells within the definitive linear heart (HH St 10) and to follow these cells, or their progenitors, frame-by-frame, backwards in time to their original positions in the mesoderm. NIH-PA Author Manuscript Here we show dynamic positional fate maps of precardiac cells. In doing so, we demonstrate that a bilateral, continuous primary heart-forming region exists at early stages in the avian embryo. More important, we document relative positional changes of precardiac cells as they exit the primitive streak and form the tubular heart (between HH stage 3+ and HH 10). When these empirical positional fate map data are compared with the expression patterns of Nkx-2.5 and Bmp-2 (HH stages 5 and 10), the results identify a subset of cardiac progenitor cells that do not share overlapping expression patterns with these purported cardiogenic markers — until the tubular heart stage. Importantly, our time-lapse data reveal that the bulk of the pre-cardiac cellular displacements during heart tube formation are driven by long-range tissue convection as opposed to cellular autonomous motility. Materials and Methods Embryo culture and electroporation method NIH-PA Author Manuscript Fertile chicken (Cornish-Cross, Ozark Eggs, Stover, MO) and Japanese quail (Coturnix coturnix japonica, Ozark Eggs, Stover, MO) eggs were incubated for 8–12 hours. The HH stage 3 embryos (Hamburger and Hamilton, 1951) were mounted on filter paper rings and cultured at 37°C in 35 mm Petri dishes coated with agar-albumen substrate (Chapman et al., 2001). Epiblastic cells were electroporated at and near the primitive streak at HH stages 3 to 3+ with DNA plasmids encoding fluorescent proteins, then incubated at 37°C (Cui et al., 2006). Cells in the electroporated regions express fluorescent proteins within 3hr, subsequent time-lapse recording commenced at HH3+ to HH5 and continued until HH12. Image acquisition and data analysis Wide-field time-lapse recording (Czirok et al., 2002) was used to record large-scale tissue displacements, at 1 µm resolution, every 6–8 minutes in bright-field and fluorescent modes. For each field and optical mode, 7–9 images were acquired in multiple focal planes, separated by 10 µm. As a result of over-sampling in the z plane, no feature moves out of focus during the extended (20–30 hours) recording time. Post image-acquisition, the digital files were processed to produce full-scale images registered to correct for X-Y shift during the experiment. Fluorescently-labeled precardiac cells of interest were confirmed to be within the heart tube at HH stage 12 — i.e., after heartbeat and bending to the embryonic-right (See Movie 1). These precardiac tagged cells were traced backward through successive image frames from HH stage 10 to HH Stage 3+ as cells first exit from the primitive streak (Movie 2–Movie 4). Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 4 NIH-PA Author Manuscript Cell positions were marked at sequential time frames with assigned grayscale values, and the images were processed in MATLAB (Mathworks) to calculate the coordinates of the tracked cells (Figure 1A and 1B). Using a custom C++ program, tracked cells from 6 different embryos were combined and mapped to a “standard” embryo at progressive developmental stages. These cells were assigned colors according to their final anatomical positions in the tubular heart at HH stage 10 (Figure 1C) with respect to both the antero-posterior (Figure 2) and to the dorsalventral (Figure 3) axes; See also QuickTime Movie 3–Movie 4. NIH-PA Author Manuscript Using our electroporation approach and H2B-GFP, H2B-RFP or mito-YFP we were able to label a subset of epiblastic cells, some of which gastrulated and formed precardiac mesodermal cells. In other words, the number of cells tracked was limited by the following: 1) not all mesodermal cells (including precardiac cells) are fluorescently labeled following electroporation of epiblastic cells, 2) not all fluorescent cells contribute to the tubular heart. Only fluorescent cells/cell nuclei that came to reside in the heart by HH Stage 10 were followed. Thus, we could only track a subset of cardiac precursor cells. When choosing cells to track from HH 10 back to earlier stages, we chose myocardial cells based on the following three conditions: 1) the tracked cells were located on the outer surface of the heart tube when the heart tube rotated to one side of the embryo at HH Stage 12 (See Movie 1); Movie 2) the myocardial cells moved in clusters (in earlier stages); whereas endothelial/endocardial cells arise as polygonal networks that are drawn into the myocardial mantle by passive tissue deformations; i.e. regression of the anterior intestinal portal (Czirok et al., 2002); and 3) the tracked (myocardial) cells behaved differently when compared to the endocardial and endothelial cells observed in time-lapse movies of Tie1 transgenic quail (unpublished data, Drs. R. Lansford and B. Rongish). All cells tracked in this study contribute to the heart tube by HH stage 10 — therefore the region from which these tracked cells arise is referred to as the “primary heart forming region”, while the electroporated cells residing in this region are referred to as “primary cardiac precursor cells”, or simply “cardiac precursor cells”. References to the primitive streak are as follows: along the anterior-posterior axis — the 100% position is at the rostral/cranial tip of the primitive streak while the 50% position is at the middle of the primitive streak. Plasmid DNA vectors A combination of three DNA plasmids encoding fluorescent proteins were used: H2B-GFP (Cui et al., 2006), Mito-YFP (Clontech, Mountain View, CA) and H2B-RFP (Campbell et al., 2002). NIH-PA Author Manuscript In situ hybridization In situ hybridization was performed as described previously (Henrique et al., 1995). Nkx-2.5 expression patterns were analyzed in chicken and quail embryos, and Bmp-2 expression patterns were analyzed in chicken embryos using the antisense digoxigenin-RNA probes prepared from plasmids containing chicken Nkx-2.5 and Bmp2 cDNAs (Schultheiss et al., 1997; Schultheiss et al., 1995). Whole-mount immunolabeling and histology Embryos were fixed with 3% paraformaldehyde in PBS for 30 minutes and washed with PBS/ Azide solution. For whole-mount immunolabeling using anti-GFP (H2B-GFP and Mito-YFP labels), embryos were incubated in 0.4% Triton for 15 minutes at 4ºC and washed with PBS/ Azide solution. Embryos were incubated in BSA blocking solution for 3 hours at 4ºC followed by anti-GFP antibody (Molecular Probes) at 1:1000 in BSA blocking solution overnight, and Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 5 NIH-PA Author Manuscript subsequently were washed 3 times with PBS/Azide solution. For histology, embryos were dehydrated in a series of ethanol, placed in JB4 infiltration medium (Electron Microscopy Sciences, Hatfield, PA) at 4ºC overnight, and embedded in JB-4 resin. Subsequently, 10µm sections were prepared. Results The primary heart field is contiguous at all stages HH Stages 3+ and 4- precardiac cells, which will later comprise the tubular heart at HH Stage 10, are arranged in an essentially complete circular pattern, while exiting the 50–100% primitive streak (Figure 2 panel A; Movie 3). At HH stage 5 and later stages, the primary heartforming region presents as a continuous arc extending across both sides of the embryo (Figure 2 panels B-D; Movie 3). A small number of cells cranial to Hensen’s node at HH4- form the anterior-most and ventral-most portions of the heart at HH Stage 10 — of this small group, 8 cells (4 cells in two embryos) were tracked on both sides of the embryonic vertebral axis (See Movie 4 and Movie 5). NIH-PA Author Manuscript Tagged cells remaining within the larger cohort were tracked only on the left side (68 cells in four embryos), and were observed to move laterally and cranially (see cells in white circle, Figure 2 A and Movie 3). Cells were tracked on one side of the specimens, as it is widely accepted that the spatial pattern of the primary heart-forming region is bilaterally symmetrical. Assuming there is a mirror-image cohort of cells on the embryonic-right side — the primary heart field at Stage 3+ and 4 conforms to a virtually “closed” circular pattern that segues into a continuous arc at Stage 5, when cardiac precursor cells cease exiting the primitive streak. To verify if there are bona fide mesodermal cells residing in the anterior-most cardiac precursor field, longitudinal plastic sections (10 µm) were cut from the anterior midline (solid white line in Figure 5) of labeled HH Stages 5 (panel A), 6 (panel B) and 7 (panel C) embryos. Sections were whole mount labeled using anti-GFP antibody to identify the H2B-GFP and H2B-YFP labeled cells. The sections show fluorescent cells within both the primitive endoderm and a discontinuous layer of mesoderm. The observed variability in fluorescence between cells in the same germ layer and/or between cells in the endoderm versus the mesoderm may be due to the number of cell divisions that have occurred since labeling of the parent epiblastic cell. The cells are visible using both the DIC and fluorescence optics. The solid white line represents the positions of longitudinal section images and the dashed white lines and asterisks represent the levels of the anterior-most cardiac precursor cells observed at each stage. NIH-PA Author Manuscript It is interesting that at the level of the anterior-most cardiac precursor cells mapped in this study (marked by white dashed lines and asterisks in Figure 5 and Figure 6), mesodermal cells are present at the midline at all stages examined (white arrowheads in Figure 5). In contrast, mesodermal cells found more anterior to the levels occupied by pre-cardiac cells at HH stages 5 and 6 (arrowheads shown in boxed areas in Figure 5 panels A and B), were not observed in the lip of the headfold by HH stage 7+ (boxed area in Figure 5C). The (GFP) longitudinal sections of electroporated embryos show that in the lip of the headfold at HH stage 7+ (boxed area, Figure 5C), fluorescently labeled cells are found in the endoderm, but not in the ectoderm. These cells overlap with a region of Nkx-2.5-expressing cells (Figure 6C). Cardiac progenitor locomotion and distribution at progressively later developmental stages HH stages 3–4—At stages 3–4, the cardiac precursor cells intermix with each other during ingression and as they move away from the 50–100% primitive streak position. There is no easily discerned motion pattern that characterizes cellular intermingling along the ventraldorsal axis (as indicated by the colors of the labeled cells; see Figure 3 Panel A). However, a Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 6 NIH-PA Author Manuscript general anterior-posterior (A-P) pattern was discerned (Figure 2 Panel A). Most red-colored cells (in Figure 2A) have previously ingressed through the streak. Ultimately all the red-colored cells establish an anterior position-fate in the HH10 heart. Cells assigned with the colors green and yellow are intermixed, but generally take positions anterior to blue cells. Pink cells, the sinoatrial precursor cells, are still ingressing through the primitive streak at HH3+. By HH stage 4− (Figure 2A), all cardiac precursor cells that originated from the 75–100% primitive streak sector have exited the primitive streak and departed the midline. Meanwhile other cardiac precursor cells continue to ingress at the 50–75% sector of the primitive streak. HH stage 5—All cardiac precursor cells leave the primitive streak by HH Stage 5 and the heart-forming region exists as an arc (Figure 2 panel B). The HH5 cardiac progenitors continue to shift their positions relative to nearby labeled cells at high frequencies when compared to later stages; meanwhile the overall A-P and V-D patterns become more discernable. Simultaneously, the anterior-most cardiac precursor cells gradually begin to move in a collective, non-random, fashion. Thus, cardiac precursor cells gradually acquire relatively stable anterior-posterior and ventral-dorsal position-fates — which correspond to their definitive position in the HH stage 10 heart tube. NIH-PA Author Manuscript HH stage 7—Compared to earlier stages, HH 7 cardiac precursor cells and surrounding mesodermal cells rarely change relative positions and begin to show a spatial regularity that bridges several cell-diameters — further stabilizing their positional fate pattern along the antero-posterior and ventral-dorsal axes. This spatial constancy will continue to be manifested in the HH stage 10 heart tube. With respect to the A-P axis (Figure 2 panel C), the anterior cardiac precursor cohort, in red, expands to occupy tissue from the cranial-most position to a position approximately 35 degrees along a radian originating at Hensen’s node (i.e., between the red lines in panel C of Figure 2). Further, these red-colored (presumptive anterior) cells are found in the medial-most positions compared to the other labeled cells. The cardiac precursor cells in green extend across a sector 25 degrees to 80 degrees (i.e., between the green lines), and are positioned between red- and blue-labeled cells. The cardiac precursor cells denoted in blue contribute to the posterior-most positions at HH 10 and extend from about 35 degrees to 90 degrees (i.e., between the blue lines). NIH-PA Author Manuscript With respect to the D-V axis (panel C in Figure 3) — the cardiac precursor cells denoted in red contribute to the ventral-most positions in the HH 10 heart tube and extend from the cranialmost positions to about the 50 degree sector (i.e., between the red lines). With respect to the mediolateral axis, these same red-colored (presumptive ventral) cells tend to be observed later at the medial-most positions. The cardiac precursor cells denoted in green contribute to the lateral side positions in an HH 10 heart tube and are observed in a D-V sector 35 degrees to 80 degrees (i.e., between the green lines). The green-colored cells are positioned between redand blue-labeled cells. The cardiac precursor cells denoted in blue, which come to reside in the dorsal-most positions in the HH 10 heart tube, are present in a 70 to 90 degree sector (i.e., between the blue lines). HH stage 7+—After HH7, the epithelialized mesodermal cell layer (splanchnic mesoderm) and the endoderm-fold separate from the somatic mesoderm/ectoderm layer as the AIP regresses and the headfold progresses caudally. The resulting cardiac splanchnopleura rollsup at an angle oblique to the embryonic midline (panel D in Figure 2). As a consequence of the rolling-up, the cells in the outer positions “flip” and assume more posterior and inner positions, while the cells observed in the intermediate positions assume more anterior and outer Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 7 NIH-PA Author Manuscript positions (See Movie 3). Meanwhile, the rolling-up mechanism proceeds posteriorly — thus anterior cardiac precursor cells “exchange” or flip positions at earlier stages compared to cardiac progenitors at more posterior locations. HH stages 8 through 10—By HH8, our movies confirm that cardiac precursor tissues are engaged in the bilateral fusion necessary to form a single tubular heart (Panels 2E and 2F and Movie 2–Movie 4). The cardiac cells/tissue that rolled-up to the ventral side of the heart tube, at HH 10, continue to descend towards a more posterior position. Meanwhile, the cardiac precursor cells that maintained their tissue positions and contribute to the dorsal side of the heart tube continue to ascend towards a more anterior position. By HH stage 10, the cardiac precursor tissues have formed a quasi-cylindrical structure. Stages after HH 10—Individual cardiac cells were not tracked after HH stage 10 due to technical limitations imposed by the initiation of the heartbeat. However, inspection of the time-lapse recordings shows that the cardiac cells continue to stream into and form presumptive atrial tissue (accretions at the posterior end of the heart tube) as well as future sinus venosus tissue (caudal to the heart tube). Similarly, outlet tissue is recruited from the cephalic mesoderm. Thus, the heart tube lengthens in both cranial and caudal directions. NIH-PA Author Manuscript Comparison of the dynamically-defined primary heart-forming region with the expression patterns of Nkx-2.5 and Bmp-2 in fixed specimens In situ hybridization results revealed no discernable differences between the expression of Nkx2.5 in chicken and quail embryos. The expression patterns of Nkx-2.5 (quail, Figure 2; chicken, Figure 3) and Bmp-2 (chicken) form a continuous arc extending on both sides of the embryo at HH stages 5–7 and overlap to a great extent with the nascent heart tube at HH stage 8 and thereafter — however, significant discrepancies exist between the expression patterns of Bmp-2/Nkx-2.5 and the location of the primary heart field we dynamically mapped. For example: 1) At the posterior end of the heart-forming region, a subset (labeled with black arrows) of the cardiac cells that will contribute to the presumptive atrium (labeled with the color pink) do not overlap with Nkx-2.5 positive regions until HH stage 9. (See Figure 2 and Figure 3, and Movie 3). When compared to Nkx-2.5 expression, Bmp-2 is expressed more posteriorly, in the region encompassing many of the pink-colored progenitors. However, neither Bmp-2 nor Nkx-2.5 expression overlaps with the lateral-most pink cells (black arrows in Figure 2, panels D, E and F). NIH-PA Author Manuscript 2) Between HH stages 5 and 9, both Bmp-2 and Nkx-2.5 expression patterns overlap with cells colonizing the presumptive ventricle (manifested at HH10), except the cells at the anterior tip of the cardiac crescent. These latter anterior progenitors originated from the Hensen’s node and nearby streak region at HH stage 3+ (white arrows in Figure 2). Between HH stages 5 and 7, the time-lapse data suggest that anterior tip cardiac progenitors do not overlap with the Nkx-2.5 expressing cells. Figure 6 shows longitudinal cross-sectional analysis from wholemount in situ hybridizations. The dashed white line and asterisks mark the level of the anterior most cells that contribute to the tubular heart. In each case these cells are found outside the region of Nkx-2.5-expressing cells at the midline. Prior to HH stage 7+, the Bmp-2 expression pattern is similar to that of Nkx-2.5, except for a small patch of Bmp-2 positive cells just posterior to the endoderm-fold near the midline (panels 2B and 2C and Figure 4 Panels A – C'). The time-lapse data suggest that the small region of Bmp-2 expression overlaps with the anterior-most cardiac precursors at HH Stages 5 and 7 (Figure 2B and 2C; Figure 4 Panels B- C'); however, when the anterior-most cardiac precursor cells incorporate into the headfold at HH8 and 9, the cells reside in a Nkx-2.5-negative and Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 8 NIH-PA Author Manuscript Bmp-2-negative region (Figure 2E and 2F). Eventually, at HH10, when these anterior-most cardiac precursor cells incorporate into the presumptive ventricles, their positions coincide with Nkx-2.5- and Bmp-2-positive regions (data not shown). 3) Some lateral cardiac progenitors are not found in regions overlapping with Nkx-2.5 or Bmp-2 expression at stages earlier than HH stage 10. Note the cells indicated by white arrowheads in Figure 2, Panels E and F. Discussion The heart-forming primordium is a coherent and contiguous circular band of progenitors at stages earlier than previously appreciated We designed a precision electroporation approach (Cui et al., 2006) and wide-field time-lapse imaging technology (Czirok et al., 2002). The instrumentation and software enabled us to determine, in compressed “real-time”, both precardiac cellular trajectories, and the relative position changes of these cells during gastrulation, axis formation and early cardiogenesis. Electroporation of early stage embryos allowed expression of markers by HH Stage 3+. This approach was advantageous in that precardiac cells could be targeted with greater efficiency at earlier stages. NIH-PA Author Manuscript One important observation arising from the recordings, using these technological advances, is that the anterior-most cardiac precursor cells we observe originate immediately adjacent to the midline axis — thus at very early stages (HH Stage 4) a “continuous” primary heart-forming region is present. Earlier work by Rosenquist and DeHaan (Rosenquist and DeHaan, 1966) indicated that at a stage slightly later than our work — HH Stage 5 — bilateral heart forming fields were “bridged” by cells that ingressed through the streak at HH 4 and migrated in a direct route cranially, similar to the cells we refer to as the “anterior-most cardiac precursor cells”. This group referred to these “bridge” cells as prospective endocardial and myocardial cells of the conus. In 1969, Stalsberg and DeHaan (Stalsberg and DeHaan, 1969) reported a few prospective endocardial cells and noncardiogenic mesoderm were present at the midline. NIH-PA Author Manuscript Thus, since the 1960’s, controversy has existed regarding when the heart forming fields form a single unit that spans the midline. Redkar and colleagues have suggested the heart primordia fuse between HH Stage 7 and 8 (Redkar et al., 2001). However it is more generally “accepted” that the chicken primary heart fields remain bilateral and physically separate until HH Stage 9 (Stalsberg and DeHaan, 1969; Reviewed in Abu-Issa and Kirby, 2007), at which time fusion events result in formation of a tubular heart. Recent work by Abu-Issa and Kirby (Abu-Issa and Kirby, 2008) suggests the initial fusion occurs even later in the chick, at HH Stage 10. There appears to be species-specific differences pertaining to when the heart fuses; for example, Baldwin and colleagues showed that a single crescent-shaped population of precardiac mesodermal cells exists in the presomite rat embryo (Baldwin et al., 1991). Although our electroporation technique does not label every cell in a given region, the recordings convincingly show that the labeled cells represent a continuous cohort of cells destined to comprise the midline “linear” heart in quail embryos. Indeed, virtually all the known cardiac “segments”, present at HH Stage 10, consist of cells that map back to a well-defined circular band of cells within HH Stage 3+/4 gastrulae. Moreover, we are able to track the position-fate of the anterior-most cardiac progenitors in time-lapse — both “forward” and “backward”. Based on the present data, the anterior-most cells we tracked move in concert with the rotating heart tube after HH10. These cells appeared to be mostly myocardial in nature and to occupy the outer surface layer of the heart tube at HH12. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 9 NIH-PA Author Manuscript In addition to the recordings, we show labeled cells in a discontinuous layer of mesoderm and in the adjacent endoderm in longitudinal sections prepared through the anterior-most cardiac region of embryos at HH Stage 5, 6, and 7+. Our results clearly show that mesodermal cells exist in HH stage 5, 6, and 7+ embryos, at the level of the anterior-most precardiac cells we dynamically mapped. Thus, HH 10 cardiac cells can be traced back to progenitors that reside at the midline of HH 5, 6, and 7+ quail embryos. Interestingly, at HH stage 5 and 6, “transient” mesodermal cells are also observed at positions cranial to those of the precardiac precursors. At HH stage 7+, as the headfold regresses, the region cranial to the levels of the precardiac precursors moves posteriorly, forming the lip of the headfold (boxed area in Panel 5C). At HH stage 7+, we could not identify mesodermal cells in the lip of the headfold. Instead, the data show that fluorescently-labeled cells are found only in the endoderm of this region, suggesting that the cranial-most midline mesodermal cells incorporate into the endoderm layer between HH stage 5 and HH stage 7. NIH-PA Author Manuscript This notion of plasticity between the endoderm and mesoderm at these stages is supported by that of Seifert and colleagues, who observed closely associated endodermal and mesodermal cell layers in longitudinal sections of avian embryos slightly later, at HH Stage 8. These workers speculated that the adjacent endoderm contributes to mesenchyme formation at the prechordal plate (Seifert et al., 1993). It is important to note that the precardiac cells we track in time-lapse pass through the anterior-most cardiac precursor region discussed by Seifert et al. Other work suggests that endoderm can arise from cells that ingress, but migrate and mix with cells in the middle germ layer already predestined to form mesoderm and endoderm — up to HH Stage 4 (Kimura et al., 2006). The exact cellular architecture of this mesodermal/endodermal interface will have to be resolved by high-resolution lineage fate mapping — work that lies beyond the scope of this study. The rolling-up movement of cardiac progenitor tissue, between HH7 and HH8, dramatically alters cellular/tissue positional-fate NIH-PA Author Manuscript Much debate has centered on if, and when, cranial-caudal patterning occurs during heart tube formation. It has been suggested by some groups that rostrocaudal patterning of the primary cardiac precursor cells occurs within the primitive streak, and that this rostrocaudal alignment is retained as these cells move to the midline to form the heart tube (Fishman and Chien, 1997; Garcia-Martinez and Schoenwolf, 1993). Other groups have suggested such an event may occur later, at HH Stage 5 (DeHaan, 1963; Rosenquist and DeHaan, 1966; and reviewed in Abu-Issa and Kirby, 2007). More recently, the idea of a “continuous” rostrocaudal patterning within the primitive streak has been contradicted (Ehrman and Yutzey, 1999; Redkar et al., 2001). Our data, like those of Ehrman and Yutzey and Redkar and colleagues, also contradict the idea of continuous rostrocaudal patterning which first occurs within the primitive streak. Furthermore, our data show how cells, which originate from similar A-P levels along the primitive streak, change their relative positions when the cardiac splanchnopleura tissues rollup during HH7 and HH8. Specifically, the general movement pattern of mesodermal cells between HH stages 3+ and 7 is that the cells exiting from more anterior (or cranial) positions in the primitive streak eventually reside at more medial locations in the mesoderm; while the cells exiting from more posterior (or caudal) positions in the primitive streak colonize more lateral locations in the mesoderm in later stages. The bilateral tissues will eventually return to the midline and fuse (Yang et al., 2002). The tissue displacement or rolling-up movement occurs first anteriorly and then propagates posteriorly to positions near the atrio-sinus venosus junction after HH10. As the result of the tissue rolling up, the cells that reside at anterior and lateral positions in the heart-forming region are convected to the posterior heart tube at HH10 (the cells denoted with blue color in panel C of Figure 2 and the blue zone on the left side of the embryo in Figure 7). Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 10 NIH-PA Author Manuscript These data indicate that cells at a given medial-lateral level — which corresponds to a given axial level in the primitive streak (A-P axis) — contribute to very different A-P and V-D positions in the HH10 heart tube, as has been suggested previously (Ehrman and Yutzey, 1999; Stalsberg and DeHaan, 1969). The positional fate map reported here agrees substantially with a previous cardiac fate map that showed most cardiac precursor cells arise from 50–70% of the primitive streak at HH3/3 + (Garcia-Martinez and Schoenwolf, 1993). However, since we tracked the cranial cardiac precursor cells back to locations very close to the tip of the primitive streak (Hensen’s node), as shown in panel A in Figure 2 and Figure 4A (See Movie 4 and Movie 5) — we concluded that there are cardiac precursor cells originating from the tip (100%) of the primitive streak, or immediately nearby (±10µm) that contribute to the anterior-ventral side of the heart tube by HH10. The spatial separation of cells occurs during the HH5 to HH7 interval NIH-PA Author Manuscript The time-lapse data show that neighboring mesodermal cells mix relatively frequently at HH4 and HH5, when compared to later stages. At HH6, neighboring mesodermal cells change their relative positions less frequently. This regional decrease in mixing begins at anterior positions, and then propagates posteriorward. We assume the loss of mixing with neighboring cells is due to the segregation and epithelialization of mesoderm into two layers: splanchnic mesoderm and somatic mesoderm. During this time, it is possible to discern the spatial segregation of cardiac precursor cells into coherent tissue bands, as denoted with different colors encoding A-P and D-V relative positional information. These spatial data agree most closely with the detailed fate map at HH5 prepared by Stalsberg and DeHaan (Stalsberg and DeHaan, 1969). We observed that cells clustered within 3–4 cell-diameters at HH stage 7 in some cardiogenic regions are later separated by substantial distances at HH stage 10 — shifting from a 30µm to a 300µm spread. With our technology, it is possible to discern such fine-tuned spatial separation of cells with distinct positional fates (10µm-10,000µm spatial scale). Previous studies using DiI and coordinating grids (Hochgreb et al., 2003; Redkar et al., 2001) were unable to provide sufficient spatiotemporal resolution to resolve the cellular positional fates we document above. For example, Redkar and co-workers incorrectly concluded that there is a significant mixing of cells in the cardiac field up to HH stage 8 (Redkar et al., 2001). This we now know was due to lower spatial resolution data, the absence of time-lapse imaging (temporal resolution) and the fact that cardiac precursor cells “roll-up” and shift their relative positions dramatically between HH stage 7 and HH stage 10. Moreover, the diI labeling techniques used by Ehrman and Yutzey (Ehrman and Yutzey 1999) as well as Redkar and colleagues (Redkar et al., 2001) could have labeled endodermal, as well as mesodermal cells. NIH-PA Author Manuscript Hochgreb et al (Hochgreb et al., 2003) provided a more detailed heart fate map, but also overestimated the degree of anterior and posterior cellular mixing at HH stage 8 compared to HH stage 7 (Hochgreb et al., 2003). With the advantage of new technology, our data show that spatial separation of cardiogenic cells slows significantly and is “fixed” by HH stage 7; thus, supporting the finding that experimental manipulation of cardiac A-P fates is only possible up to approximately HH stage 7 (Hochgreb et al., 2003). However, although cell spatial separation (cell position with respect to neighboring cells) is fixed by HH stage 7, endodermal (tissuelevel) folding results in profound cellular positional changes (i.e. flipping, rotating) between HH 7–10. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 11 Cell-autonomous motility versus tissue motion; i.e., mesenchymal motility versus epithelialization NIH-PA Author Manuscript We operationally define the term “tissue motion” or “tissue convection” to indicate instances where/when cells are displaced in a manner that maintains their positional relationships — thus, entire cohorts of precardiac cells move with the same trajectories. Our data strongly suggest that after stage 6, presumptive myocardial cells begin to lose their cell autonomous motility as coherent cardiac tissues form. Further, between HH stages 7–12, tissue convection is the predominant mechanism by which cells are moved to their definitive positions in the heart. Our findings are similar, but not identical, to those of earlier groups (Rosenquist and DeHaan, 1966; Stalsberg and DeHaan, 1969), who previously reported that the epimyocardial layer of the splanchnic mesoderm moves as a sheet beginning at an earlier stage – stage 5 and this continues until stage 12. NIH-PA Author Manuscript It is important to point out that in certain spatiotemporal positions — such as the anterior-most cells (Movie 4 and Movie 5) — cells still exhibit significant autonomous cellular motility, well past the “linear” heart stage (HH10). Studies in zebrafish have also identified two distinct phases of cell movements during heart tube assembly, including an initial coherent medial movement of the cardiomyocytes. This tissue-like movement pattern is followed by an angling of a subset of cardiomyocytes, located in the periphery, toward the endocardial precursors (Holtzman et al., 2007). Further, this group showed the latter cardiomyocyte behavior pattern is directed by the endocardium. Our unpublished time-lapse data confirm that tissue motion occurs during heart morphogenesis, since extracellular matrix fibrils are displaced into the forming tubular heart (Rongish et al., unpublished data; similar in concept to studies by Zamir et al., 2006; 2008). Similarly, time-lapse recordings show that fluorescent endocardial cells also move into the forming tubular heart as a cohesive tissue (unpublished data, Drs. R. Lansford (Caltech) and Rongish). Nkx-2.5 and Bmp-2 expression in the primary heart-forming region NIH-PA Author Manuscript The expression patterns of Nkx-2.5 and Bmp-2 (upstream of both Nkx-2.5 and Gata-4) do not completely overlap with the heart-forming regions mapped by others (Rosenquist and DeHaan, 1966; Ehrman and Yutzey, 1999), nor with the heart-forming region described here. Although Ehrman and Yutzey believed the heart forming region was defined medially, laterally, and posteriorly by Nkx-2.5 gene expression, our data (in both quail and chicken embryos) show significant discrepancies exist between cardiac gene expression and the heart-forming region. Specifically, the anterior-most, and some lateral cardiac precursor cells do not appear to overlap with Nkx-2.5- and Bmp-2-positive regions until HH10-HH12 (white arrows and arrowheads in Figure 2). Further, the data show that cardiac precursor cells continue to ingress through and exit the primitive streak at HH4 — an observation that is distinct from the Bmp-2 expression pattern (Schultheiss et al., 1997). Unlike observations by Ehrman and Yutzey, our dynamic fate mapping indicates the heart forming region extends posterior to Hensen’s node at HH 4–5, and also extends posterior to the first formed somite at HH 7–8. Like Redkar et al., 2001; we show the posterior border of the heart forming region (including sinoatrial precursors) at HH 8 to be at approximately the level of the 4th somite (Redkar et al., 2001). Moreover, most of the posterior progenitors (labeled with the color pink), which contribute to part of the future atrium (at HH12), do not overlap with the Nkx-2.5- nor Bmp-2- positive region until Stage 9. These discrepancies imply that neither Nkx-2.5 nor Bmp-2 precisely defines the entire primary heart-forming region (Eisenberg, 2002; Redkar et al., 2001; and reviewed in Abu-Issa and Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 12 NIH-PA Author Manuscript Kirby, 2007). Nevertheless, these two cardiac markers remain reasonable candidates for identifying most cells that contribute to the linear heart at HH Stage 10. It remains to be determined whether cardiac markers such as Isl1 (Sun et al., 2006; Abu-Issa et al., 2004; Laugwitz et al., 2008) or GATA4 (Hochgreb et al., 2003) can identify all primary cardiac precursors in quail. Further, recent work indicates Fgf10 is expressed by a subset of cardiac cells in the anterior second heart field in the mouse (Galli et al., 2008). Markers such as these need to be studied in avian embryos to determine the onset and location of their expression. In summary, a combination of real-time expression of putative cardiac markers, cell nuclei labeling and tracking, and time-lapse imaging is needed to address when cardiac lineage fate is specified with respect to position fate. In other words, position fate and lineage fate are distinct cellular properties. True 3D mapping of primary heart morphogenesis is indicated NIH-PA Author Manuscript Additional, more extensive studies are likely to result in a more complete positional fate map of precardiac precursors. For example, future studies may use cardiac-specific promoter-driven plasmids to increase cell labeling in the tubular heart. Regardless, there is and will be a need for increased spatial resolution to image the heart as it increases in thickness; driving a subsequent need for three dimensional cell tracking. The fluorescently-labeled cardiac precursor cells that contribute to the dorsal side of the heart tube by HH10 are difficult to track using our current technology. It is likely that if the positional fate of additional dorsal cells were mapped, the primary heart-forming region would be larger than depicted by our current study. We speculate that the blue zone on the right side of the embryo in Figure 7 (depicting the progenitors contributing to the dorsal side of the heart) would extend further anteriorly; and the red zone on the left side (depicting the progenitors contributing to the anterior heart) would extend further posteriorly. Currently, image deconvolution and three-dimensional cell tracking methods are being developed that will allow more accurate positional-fate mapping during early morphogenesis. Collectively, the experimental methods used in this study resolved cellular displacements at 5–7 µm spatial resolution across a wide length scale (>3 orders of magnitude). The technology allows the movements of multiple cells at widely-separated regions of the embryo to be followed, simultaneously. It is tempting to speculate on the existence of a single heart forming field, as is currently being debated (Abu-Issa et al., 2004; Moorman et al., 2007; Abu-Issa and Kirby, 2008). However, further dynamic imaging studies, including 3D tracking analysis, will be required to place such a conclusion on a firm basis. Supplementary Material NIH-PA Author Manuscript Refer to Web version on PubMed Central for supplementary material. Acknowledgements We thank Dr. Charles Little (University of Kansas Medical Center) for helpful and provocative discussions, and Drs. Olivier Pourquié and Bertrand Bénazéraf (Stowers Institute for Medical Research, Kansas City, MO) for assistance with in-situ hybridization analysis. This study was supported by an American Heart Association fellowship (0620105Z, CC), an award from the G. Harold and Leila Y. Mathers Charitable Foundation (BJR), an award from the University of Kansas Medical Center Research Institute (BJR) and a NIH award HL085694 (BJR). References Abu-Issa R, Kirby ML. Heart Field: From Mesoderm to Heart Tube. Annu Rev Cell Dev Biol. 2007 Abu-Issa R, Kirby ML. Patterning of the heart field in the chick. Dev. Biol 2008;319:223–233. [PubMed: 18513714] Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 13 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Abu-Issa R, Waldo K, Kirby ML. Heart fields: one, two or more? Dev Biol 2004;272:281–285. [PubMed: 15282148] Baldwin HS, Jensen KL, Solursh M. Myogenic cytodifferentiation of the precardiac mesoderm in the rat. Differentiation 1991;47:163–172. [PubMed: 1959714] Campbell RE, Tour O, Palmer AE, Steinbach PA, Baird GS, Zacharias DA, Tsien RY. A monomeric red fluorescent protein. Proc Natl Acad Sci U S A 2002;99:7877–7882. [PubMed: 12060735] Chapman SC, Collignon J, Schoenwolf GC, Lumsden A. Improved method for chick whole-embryo culture using a filter paper carrier. Dev Dyn 2001;220:284–289. [PubMed: 11241836] Cui C, Lansford R, Filla MB, Little CD, Cheuvront TJ, Rongish BJ. Electroporation and EGFP labeling of gastrulating quail embryos. Dev Dyn 2006;235:2802–2810. [PubMed: 16894628] CzirokARuppPARongishBJLittleCDMulti-field 3D scanning light microscopy of early embryogenesis. ; J Microsc2002206209217 [PubMed: 12067365] De la Cruz MV, Gimenez-Ribotta M, Saravalli O, Cayre R. The contribution of the inferior endocardial cushion of the atrioventricular canal to cardiac septation and to the development of the atrioventricular valves: study in the chick embryo. Am J Anat 1983;166:63–72. [PubMed: 6837479] De la Cruz MV, Sanchez Gomez C, Arteaga MM, Arguello C. Experimental study of the development of the truncus and the conus in the chick embryo. J Anat 1977;123:661–686. [PubMed: 885781] De La Cruz MV, Sanchez-Gomez C, Palomino MA. The primitive cardiac regions in the straight tube heart (Stage 9) and their anatomical expression in the mature heart: An experimental study in the chick embryo. J Anat 1989;165:121–131. [PubMed: 17103606] DeHaan RL. Migration patterns of the precardiac mesoderm in the early chick embryo. Exp. Cell Res 1963;29:511–560. Ehrman LA, Yutzey KE. Lack of regulation in the heart forming region of avian embryos. Dev Biol 1999;207:163–175. [PubMed: 10049572] Eisenberg LM. Belief vs. scientific observation: the curious story of the precardiac mesoderm. Anat Rec 2002;266:194–197. [PubMed: 11920381] Evans SM. Vertebrate tinman homologues and cardiac differentiation. Semin Cell Dev Biol 1999;10:73– 83. [PubMed: 10355031] Fishman MC, Chien KR. Fashioning the vertebrate heart: earliest embryonic decisions. Development 1997;124:2099–2117. [PubMed: 9187138] Galli D, Dominguez JN, Zaffran S, Munk A, Brown NA, Buckingham ME. Atrial myocardium derives from the posterior region of the second heart field, which acquires left-right identity as Pitx2c is expressed. Development 2008;135:1157–1167. [PubMed: 18272591] Garcia-Martinez V, Schoenwolf GC. Primitive-streak origin of the cardiovascular system in avian embryos. Dev Biol 1993;159:706–719. [PubMed: 8405690] Hamburger V, Hamilton HL. A series of normal stages in the development of the chick embryo. J Morphol 1951;88:49–92. HarveyRPNK-2 homeobox genes and heart development. ; Dev Biol1996178203216 [PubMed: 8812123] Henrique D, Adam J, Myat A, Chitnis A, Lewis J, Ish-Horowicz D. Expression of a Delta homologue in prospective neurons in the chick. Nature 1995;375:787–790. [PubMed: 7596411] Hochgreb T, Linhares VL, Menezes DC, Sampaio AC, Yan CYI, Cardoso WV, Rosenthal N, XavierNeto J. A caudorostral wave of RALDH2 conveys anteroposterior information to the cardiac field. Development 2003;130:5363–5374. [PubMed: 13129847] Holtzman NG, Schoenebeck JJ, Tsai HJ, Yelon D. Endocardium is necessary for cardiomyocyte movement during heart tube assembly. Development 2007;134:2379–2386. [PubMed: 17537802] Jiang Y, et al. Common role for each of the cGATA-4/5/6 genes in the regulation of cardiac morphogenesis. Dev Genet 1998;22:263–277. [PubMed: 9621433] Kimura W, Yasugi S, Stern CD, Fukuda K. Fate and plasticity of the endoderm in the early chick embryo. Dev Biol 2006;289:283–295. [PubMed: 16337933] Kostetskii I, Jiang Y, Kostetskaia E, Yuan S, Evans T, Zile M. Retinoid signaling required for normal heart development regulates GATA-4 in a pathway distinct from cardiomyocyte differentiation. Dev Biol 1999;206:206–218. [PubMed: 9986733] Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 14 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Laugwitz K-L, Moretti A, Caron L, Nakano A, Chien KR. Islet1 cardiovascular progenitors: a single source for heart lineages. Development 2008;135:193–205. [PubMed: 18156162] Moorman AF, Christoffels VM, Anderson RH, van den Hoff MJ. The heart-forming fields: one or multiple? Philos Trans R Soc Lond B Biol Sci 2007;362:1257–1265. [PubMed: 17581808] Moreno-Rodriguez RA, Krug EL, Reyes L, Villavicencio L, Mjaatvedt CH, Markwald RR. Bidirectional fusion of the heart-forming fields in the developing chick embryo. Dev Dyn 2006;235:191–202. [PubMed: 16252277] Nishibatake M, Kirby ML, Van Mierop LH. Pathogenesis of persistent truncus arteriosus and dextroposed aorta in the chick embryo after neural crest ablation. Circulation 1987;75:255–264. [PubMed: 3791607] Olson EN. Gene regulatory networks in the evolution and development of the heart. Science 2006;313:1922–1927. [PubMed: 17008524] Rawles ME. A study in the localization of organ-forming areas in the chick blastoderm of the headprocess stage. J. Exp. Zool 1936;72:271–315. Rawles ME. The heart-forming areas of the early chick blastoderm. Physiol. Zool 1943;16:22–42. Redkar A, Montgomery M, Litvin J. Fate map of early avian cardiac progenitor cells. Development 2001;128:2269–2279. [PubMed: 11493546] Rosenquist GC. Location and movements of cardiogenic cells in the chick embryo: the heart-forming portion of the primitive streak. Dev Biol 1970;22:461–475. [PubMed: 5423312] Rosenquist GC, DeHaan RL. Migration of precardiac cells in the chick embryo: a radioautographic study. Contrib. Embryol. Carnegie Inst. Wash 1966;38:111–123. Schultheiss TM, Burch JB, Lassar AB. A role for bone morphogenetic proteins in the induction of cardiac myogenesis. Genes Dev 1997;11:451–462. [PubMed: 9042859] Schultheiss TM, Xydas S, Lassar AB. Induction of avian cardiac myogenesis by anterior endoderm. Development 1995;121:4203–4214. [PubMed: 8575320] Seifert R, Jacob M, Jacob HJ. The avian prechordal head region: a morphological study. J Anat 1993;183 (Pt 1):75–89. [PubMed: 8270478] Stalsberg H, DeHaan RL. The precardiac areas and formation of the tubular heart in the chick embryo. Dev Biol 1969;19:128–159. [PubMed: 4180350] Xavier-Neto J, Rosenthal N, Silva FA, Matos TG, Hochgreb T, Linhares VL. Retinoid signaling and cardiac anteroposterior segmentation. Genesis 2001;31:97–104. [PubMed: 11747199] Yang X, Dormann D, Munsterberg AE, Weijer CJ. Cell movement patterns during gastrulation in the chick are controlled by positive and negative chemotaxis mediated by FGF4 and FGF8. Dev Cell 2002;3:425–437. [PubMed: 12361604] Yuan S, Schoenwolf GC. Islet-1 marks the early heart rudiments and is asymmetrically expressed during early rotation of the foregut in the chick embryo. Anat Rec 2000;260:204–207. [PubMed: 10993956] Yutzey KE, Kirby ML. Wherefore heart thou? Embryonic origins of cardiogenic mesoderm. Dev Dyn 2002;223:307–320. [PubMed: 11891982] Zamir EA, Czirok A, Cui C, Little CD, Rongish BJ. Mesodermal cell displacements during avian gastrulation are due to both individual cell-autonomous and convective tissue movements. Proc Natl Acad Sci USA 2006;103(52):19806–19811. [PubMed: 17179040]Epub 2006 Dec 18 Zamir E, Rongish B, Little C. The ECM Moves during Primitive Streak Formation – Computation of ECM Versus Cellular Motion. MS accepted, PLoS Biol 2008;6(10):e247. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 15 NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. NIH-PA Author Manuscript Colors were assigned to tracked heart precursor cells in quail embryos to denote their positions at HH stage 10 along both the antero-posterior (A-P) and the ventral-dorsal (V-D) axes. Fluorescent tracked precardiac cells were selected from a population of electroporated epiblastic cells that gastrulated, formed splanchnic mesodermal cells, and came to reside on the surface of the tubular heart. A) A fluorescent image of an HH stage 10 heart with tracked cells labeled with grayscale (intensity) values. To simplify, we assume the heart tube is a cylinder with a radius R: defined as the distance from the midline reference points (arrowheads) to lateral reference points (arrows) at successive A-P levels; and that all tracked cells are on the cylinder’s surface. The radius R represents the longest distance between a lateral point (white arrow) and the corresponding midline point (white arrowhead) at the same A-P level. The radius measurement can vary at different A-P levels. B) The distance (x) between a tracked cell and the midline position of the heart tube at the same A-P level is used to calculate the tracked cell’s position on the cylinder surface (c) using the equation shown. C) The tracked cell is then assigned one color according to its position on the cylinder’s surface to indicate its relative position along the V-D axis. The color red represents the ventral-most positions; the color blue represents the dorsal-most positions and the color green represents the lateral-most positions. Along the A-P axis, the tracked cells are assigned a separate color according to their positions relative to two reference points: A, the anterior-most point and B, the posterior-most point of the heart tube (see panel 1A asterisks). The color red represents the anterior-most positions; the color blue represents the posterior-most positions and the color green represents the middle positions. The cells that reside caudal to the heart (future omphalomesenteric veins) by HH stage 10 (cells in the white circle in panel A) are assigned the color pink (See Figure 2 and Movie 3). Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 16 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Sequential time-lapse frames of “tracked” primary cardiac precursor cells compiled from 6 quail embryos between HH stages 4 through 9. The tracked cells are pseudo-colored and superimposed upon a corresponding bright-field image (Panels A-F). The color code denotes a cell’s final position in the HH10 heart tube with respect to the A-P axis, as explained in Figure 1. The cranial-most cells were tracked on both sides of the embryonic vertebral axis (see white arrows); while cells originating inside the region depicted by the white oval (Panel A) were tracked only on the embryonic left side. The colored radian lines emanating from Hensen’s node (HN) denote the spatial boundaries of the color-coded cells at each time point. See Movie 3 for the corresponding time-lapse data. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 17 NIH-PA Author Manuscript The columns of smaller image panels depict a series of whole-mounted embryos subjected to RNA in-situ hybridizations using Bmp-2 or Nkx-2.5 probes. In situ hybridizations are shown in quail embryos for Nkx-2.5, and in chicken embryos for Bmp-2. Cells colored pink will reside caudal to the heart (future omphalomesenteric veins) by HH stage 10 and will eventually contribute to the atrium. Black arrows indicate a subset of pink cells that do not overlap with Nkx-2.5 expression. Scale bar: 200 µm. NIH-PA Author Manuscript NIH-PA Author Manuscript Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 18 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Sequential time points show cellular motion analyzed with respect to the dorsoventral, D-V, axis. The panels are arranged similar to Figure 2, in which primary cardiac precursor cells from 6 quail embryos were tracked between HH stage 4 and 9 (A-F). The color code denotes a cell’s final position in the HH10 heart tube with respect to the D-V axis, as explained in Figure 1. The cranial-most cells were tracked on both sides of the embryonic axis (as indicated in Figure 2), while only left side cells were tracked for the D-V analysis (Movie 3). The colored radian lines emanating from Hensen’s node (HN) denote the spatial boundaries of the color-coded cells at each time point. RNA in situ hybridization panels are shown to the right of each Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 19 respective time-lapse frame. The corresponding in situ hybridizations for Nkx2.5 are shown in chicken embryos. Bmp-2 images are duplicates of those seen in Figure 2. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 20 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Anterior-most cardiac precursor cells overlap with Bmp-2 expression patterns at HH5 and HH7. The image in panel A depicts an HH4- embryo showing only cells fated to assume cranialmost heart positions at subsequent stages (See Movie 4 and Movie 5). Panel B and C are duplicates of Figure 2 HH5 and HH7, with the black boxes showing a region expressing Bmp-2. Panels B' and C' show the boxed regions at higher magnification. Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 21 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Images of whole-mount embryos (WM-BF and WM-GFP) and their longitudinal-sections (DIC and GFP) at HH5 (A), HH6 (B) and HH7+ (C). Quail embryos were electroporated with mito-YFP at HH 3 and whole-mount (WM) immunolabeled with anti-GFP antibody to enhance the fluorescence signal before sectioning. The longitudinal-sections are marked as solid white lines in WM-GFP images. The A-P levels of cardiac precursor cells that reside at the cranialmost positions, as suggested in our time-lapse data, are marked with black asterisks in WMBF images and with white dashed lines and asterisks in images of longitudinal-sections. White arrowheads mark the mesodermal cells. The embryos in whole-mount images are viewed from their ventral side. Double-headed arrows with letters ‘V’ and ‘D’ in DIC images of longitudinalDev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 22 NIH-PA Author Manuscript sections indicate the ventral and dorsal directions of the sections. The white boxes indicate the corresponding regions of the forming headfold at each time shown. The boxed region in Panel C is referred to in the text as the lip of the headfold. White scale bars: 100 µm. HN = Hensen’s node; PS = primitive streak NIH-PA Author Manuscript NIH-PA Author Manuscript Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 23 NIH-PA Author Manuscript Figure 6. NIH-PA Author Manuscript Whole-mount embryo RNA in-situ hybridization for Nkx-2.5 and longitudinal-section images of avian embryos at HH5+ (A; chick), HH6+ (B; quail) and HH7+ (C; quail). The black lines in whole-mounted (WM) in-situ images indicate the site of longitudinal-sections. White dashed lines and asterisks indicate the A-P levels of cardiac precursor cells that reside at the anteriormost positions as suggested in our time-lapse data. The whole-mounted embryos are viewed from their ventral side. Double headed arrows with letters ‘V’ and ‘D’ in longitudinal-section images indicate the ventral and dorsal directions of the sections. White scale bars: 100 µm. NIH-PA Author Manuscript Dev Biol. Author manuscript; available in PMC 2010 August 15. Cui et al. Page 24 NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 7. NIH-PA Author Manuscript Schematic representation of the spatial separation of pseudo-colored (based on cell position at HH 10) cells at HH7 along the A-P (viewer’s left side; hatched) and the V-D (viewer’s right side) axes of the embryo. The color pink represents the region containing cells that will contribute predominantly to the atria by HH12. These data represent cell tracking in 6 embryos. (See also Movie 1–Movie 4). Dev Biol. Author manuscript; available in PMC 2010 August 15.