DEV030924 Supplementary Material

Files in this Data Supplement:

• Supplemental Tables S1-S3 -

• Supplemental Figure S1 -

  Fig. S1. BrdU pulse labeling of cardiac cells. Sagittal sections of a single embryo after pulse labeling with BrdU from 24-48 hpf. The sections were immunohistochemically stained with an α-BrdU antibody (brown) and counter-stained with Hematoxylin and Eosin (blue). Arrows indicate BrdU-positive myocardial cells. Note the high degree of labeling within the embryonic blood cells and endocardial cells. Anterior is to the left, dorsal to the top.

• Supplemental Figure S2 -

  Fig. S2. Double labeling of BrdU and cmlc2. (A,A′,B,B′) Representative sections of BrdU-soaked embryos from 24 hpf until 48 hpf stained for cmlc2 in blue and BrdU in brown. These sections show that in the tissue expressing the cardiac marker cmlc2 very few cells are positive for BrdU (black arrow). The tissue lining the myocardium (the endocardium) is negative for cmlc2 and is highly proliferative, indicated by red arrowheads. Quantitation of the number of BrdU-positive cells within the cmlc2-positive area resulted in a similar number to that previously described in the results section (6±2, mean±s.e.m., n_5).

• Supplemental Figure S3 -

  Fig. S3. eGFP and DsRed protein expression overlaps at 24 hpf and 48 hpf. (A-F) Reconstruction of confocal z-stacks of Tg(cmlc2:eGFP)/Tg(cmlc2:DsRed) embryos at 24 hpf and 48 hpf. Hearts stained immunohistochemically with α-eGFP (A,D) and α-DsRed (B,E) antibodies. C and F represent the overlay. At 24 hpf and at 48 hpf, both eGFP and DsRed protein expression overlaps throughout the heart, with the exception of a few cells. (G-J) The number of cells in Tg(cmlc2:eGFP)/Tg(cmlc2:DsRed) hearts that are eGFP^posDsRed^pos (G,I) or eGFP^posDsRed^neg (H,J; mean±s.e.m., n_5). The eGFP^posDsRed^neg cardiomyocytes are subdivided into those present at the arterial pole and those present at the venous pole (24 hpf, 3±1 and 6±2; 48 hpf, 6±1 and 2±1, respectively). Note that almost all cardiomyocytes are eGFP^posDsRed^pos (24 hpf, 151±12; 48 hpf, 311±10).

• Supplemental Figure S4 -

  Fig. S4. The number of cmlc2-expressing cells increases gradually over time. (A-D) In situ hybridization for cmlc2 in wild-type embryos; dorsal views, anterior to the top. (A) At the 14-somite stage, only a few cells have initiated cmlc2 expression. (B) By the 16-somite stage, more cmlc2-expressing cells are detectable. These cells appear to be located near the medial edge of the lateral plate mesoderm, which is migrating towards the embryonic midline. (C) More cmlc2-expressing cells are evident by the 18-somite stage. As the bilateral cardiomyocyte populations prepare to fuse, cmlc2 expression appears to be spreading laterally within the lateral plate mesoderm. (D) Continued increase in the lateral expansion of cmlc2-expressing cells is apparent at the 20-somite stage, as cardiac fusion begins. The medial-to-lateral spread of cmlc2 expression within the lateral plate mesoderm suggests that cmlc2 expression in ventricular cells precedes cmlc2 expression in atrial cells, as ventricular precursors are located in the medial portion of the lateral plate mesoderm and atrial precursors are located more laterally (Berdougo et al., 2003; Glickman and Yelon, 2002; Schoenebeck et al., 2007)

  Glickman, N. S. and Yelon, D. (2002). Cardiac development in zebrafish; coordination of form and function. Semin. Cell Dev. Biol. 13, 507-513.

  Schoenebeck, J. J., Keegan, B. R. and Yelon, D. (2007). Vessel and blood specification override cardiac potential in anterior mesoderm. Dev. Cell 13, 254-267.

• Supplemental Figure S5 -

  Fig. S5. Photoconversion of Kaede efficiently marks differentiated cardiomyocytes. (A-I) Lateral views of Tg(cmlc2:kaede) embryos, ventricle to the left, demonstrating effectiveness of photoconversion technique. (A-C) Prior to photoconversion, the 32-hpf heart exhibits green Kaede fluorescence (A) but not red Kaede fluorescence (B), resulting in an entirely green heart (C). (D-F) Immediately after exposure to UV light, all of the green Kaede (D) has been converted into red Kaede (E), resulting in an entirely red heart (F). (G-I) Sixteen hours after photoconversion, red Kaede (H) persists in the cells that were expressing the transgene. These cells also generate new green Kaede (G), as do newly differentiating cells that initiated transgene expression after 32 hpf. As a result, cardiomyocytes at the arterial pole fluoresce green, but not red, at 48 hpf (arrow in H). V, ventricle; A, atrium.

• Supplemental Figure S6 -

  Fig. S6. Schematic overview of tissue organization at the 23-somite stage. Schematic of the tissue organization seen in a transverse section through the cardiac disk at the 23-somite stage. The top structure is the neural tube around the anterior-posterior level of the mid-hindbrain boundary and the bottom structure is the yolk. The blue circles located directly ventral to the neural tube represent the endoderm. The green circles represent endothelial cells and the orange and yellow circles represent cardiomyocytes. Orange indicates the future ventricle, marked at this stage by vmhc expression and the yellow indicates the future atrium, marked at this stage by amhc expression.

• Supplemental Figure S7 -

  Fig. S7. isl1^K88X mutant embryos are deficient for Isl protein. (A-D) Live images of 48-hpf isl1^K88X sibling and mutant embryos showing no apparent morphological phenotype. (E-J) Single z-scans of confocal images of 23-somite stage isl1^K88X Tg(cmlc2:eGFP) embryos after immunofluorescence staining with α-eGFP (E,H) and α-Isl (F,I) antibodies. (E-G) isl1^K88X sibling embryo; (H-J) isl1^K88X mutant embryo. Arrowheads in the overlay of α-eGFP (green) with α-Isl (red) indicate the eGFP^posIsl^pos cells in the cardiac disk and trigeminal sensory ganglia. In the control (E-G), Isl is expressed in cardiac cells located at lateral positions in the cardiac disk (white arrowhead) and in the trigeminal sensory ganglion (green arrowhead), described in Fig. 3A-C. Note that the nuclear Isl signal in the cardiac disk in H-J is absent. Scale bars: 80 µm.

• Supplemental Figure S8 -

  Fig. S8. isl1 morpholinos reduce Isl1 protein and phenocopy the isl1 mutant in the developmental timing assay. (A,B) Live images of a representative control embryo and an embryo that was injected with the isl1 MO. isl1 morphants present with cardiac edema and loss of motility. (C-K) Single z-scans of confocal images of 23-somite stage Tg(cmlc2:eGFP) embryos after immunofluorescence staining with α-eGFP (C,F,I) and α-Isl (D,G,J) antibodies. (C-E) Control injected embryo; (F-H,I-K) examples of isl1 morphants. Arrowheads in the overlay of α-eGFP (green) with α-Isl (red) indicate the eGFP^posIsl^pos cells in the cardiac disk. In the control (C-E), Isl is expressed in cardiac cells located at lateral positions in the cardiac disk (white arrowhead) and in the trigeminal sensory ganglion (green arrowhead), described in Fig. 3A-C. Note that the nuclear Isl signal in the cardiac disk in F-H and I-K is absent; however, some signal is still present in the trigeminal sensory ganglion in F-H. Scale bars: 80 µm. (L,M) The number of eGFP^posDsRed^pos (L) and eGFP^posDsRed^neg (M) cardiomyocytes/embryo. (M) The eGFP^posDsRed^neg cardiomyocytes are subdivided into those present at the arterial pole and those present at the venous pole. Note the significant reduction of eGFP^posDsRed^neg cells at the venous pole in the isl1 morphant embryos. Bars represent mean±s.e.m. *P<0.05; **P<0.01.

• Supplemental Figure S9 -

  Fig. S9. Cardiac dysfunction has no effect on the addition of cardiomyocytes to both poles of the heart. (A,B). Graphical representation of the number of eGFP^posDsRed^pos (A) and eGFP^posDsRed^neg (B) cardiomyocytes/embryo for control MO-injected embryos (blue bar, n_4) and silent heart (sih/tnnt2) MO-injected embryos (red bar, n_3). (B) The eGFP^posDsRed^neg cardiomyocytes are subdivided into those present at the arterial pole and those present at the venous pole. Bars represent mean±s.e.m. No significant differences in the number of cells quantified in control MO- and sih MO-injected embryos were observed.

• Supplemental Figure S10 -

  Fig. S10. fgf8 MO injection results in reduced Fgf signaling. (A,A′,B,B′) Live images of control embryos (A,A’) or fgf8 MO-injected embryos (B,B′). As previously shown by Draper et al. (Draper et al., 2001), the fgf8 MO used generates a range of phenotypic classes (Draper et al., 2001). In our experiments, injected embryos exhibited phenotypes similar to those described as class 1-3 phenotypes, most typically the class 2 phenotype (Draper et al., 2001). fgf8 MO-injected embryos were selected for the absence of the midbrain-hindbrain boundary, indicated by the black arrowhead. Approximately 40% of the injected embryos presented this phenotype and only these were imaged in the developmental timing assay.

• Movie 1 -

  Movie 1. Recording of the irregular heart beating of an isl1^k88x mutant embryo. Lateral view (anterior to the left) of an isl1^k88x mutant embryo at 72 hpf. Recording was performed without anesthetizing the embryo on a Zeiss Axiophot microscope with a Leica DFC490 camera (7 frames/second).