Notch Signaling Negatively Regulates the Number of Endothelial Cells during Zebrafish Development

(A) Transverse sections of 18 hpf embryos, DMSO- or DAPT-treated, visualized for TOPRO (white), kdrl:GFP (green), and phalloidin (red).

(B) Quantification of endothelial nuclei per focal plane in DMSO-treated (n = 30) or DAPT-treated (n = 44) embryos.

(C) Transverse sections of 18 hpf phenotypic wild-type siblings and Tg(hsp70l:Gal4)^kca4;Tg(UAS:myc-NICD)^kca3 embryos visualized for TOPRO (white), kdrl:GFP (green), and c-Myc as a surrogate measure for NICD expression (red).

(D) Quantification of endothelial nuclei per focal plane in phenotypic wild-type (n = 31) and NICD-overexpressing (n = 28) embryos. Embryos treated with DAPT contained 6.66 (s = 2.42) endothelial nuclei per section, whereas those treated with DMSO had 4.03 (s = 1.27) endothelial nuclei (ANOVA, p < 10^-6). Conversely, embryos overexpressing the NICD contained 3.14 (s = 0.80) endothelial nuclei, in comparison to 4.42 (s = 1.09) nuclei per section in phenotypic wild-type siblings (ANOVA, p < 10^-5), suggesting that Notch signaling negatively regulates endothelial cell number.

Notch Signaling Positively Regulates the Number of Hematopoietic Cells during Zebrafish Development

(A) Whole-mount RNA in situ hybridization showing the expression of gata1 in 18 hpf DMSO-treated or DAPT-treated embryos. Two additional markers, krox20 and otx1, which label the hindbrain and neuroectoderm, respectively, were used as positive controls. Notice the reduced expression of gata1 in DAPT-treated embryos (black arrow).

(B) Transverse sections of 18 hpf DMSO- or DAPT-treated embryos visualized for TOPRO (white), gata1:GFP (green), and phalloidin (red).

(C) Quantification of the hematopoietic nuclei per focal plane in DMSO-treated (n = 27) or DAPT-treated (n = 19) embryos.

(D) Transverse sections of 18 hpf phenotypic wild-type siblings and Tg(hsp70l:Gal4)^kca4;Tg(UAS:myc-NICD)^kca3 embryos visualized for TOPRO (white), gata1:GFP (green), and c-Myc as a surrogate measure for NICD expression (red).

(E) Quantification of the hematopoietic nuclei per focal plane in phenotypic wild-type siblings (n = 28) and NICD-overexpressing embryos (n = 23). DAPT-treated embryos contained 4.89 (s = 2.15) erythrocyte nuclei, in comparison to 8.07 (s = 2.38) nuclei in DMSO-treated embryos (ANOVA, p < 10^-4). Conversely, embryos overexpressing the NICD contained 9.89 (s = 2.53) erythrocyte nuclei, in comparison to 6.72 (s = 1.43) nuclei per section in phenotypic wild-type siblings (ANOVA, p < 10^-6), indicating that Notch signaling positively modulates the number of hematopoietic cells.

Reduced Notch Activity Causes an Increase of Endothelial Cells at the Expense of Hematopoietic Cells

(A) Transverse sections of 18 hpf DMSO- or DAPT-treated embryos, visualized for TOPRO (white), kdrl:mCherry (red), and gata1:GFP (green). DAPT-treated embryos exhibited a concomitant gain of endothelial cells and loss of hematopoietic cells.

(B) Transverse sections of 20 hpf DMSO- or DAPT-treated phenotypic bloodless (bds) mutant embryos, visualized for TOPRO (white), kdrl:GFP (green), and phalloidin (red).

(C) Quantification of the number of endothelial nuclei per focal plane in DMSO-treated bds (n = 22), DAPT-treated bds (n = 20), and DAPT-treated wild-type sibling (n = 24) embryos. DAPT treatment in embryos lacking primitive hematopoietic cells (bds + DAPT) did not result in a change in the number of endothelial cells, suggesting that the ectopic endothelial cells observed in DAPT-treated wild-type embryos might have originated from cells that normally produce hematopoietic cells.

Notch Signaling Functions as a Cell-Fate Switch between the Endothelial and the Hematopoietic Lineages in the Nascent Mesoderm

(A) Experimental scheme of the lineage-tracing experiment. Single-cell-stage embryos were injected with DMNB-caged fluorescein-conjugated dextran. At shield stage, ten cells within the ventral margin, which predominantly gives rise to hematopoietic cells, were uncaged with UV irradiation. Subsequently, embryos were treated with either DMSO or DAPT until 18 hpf.

(B) Transverse sections of 18 hpf embryos treated with DMSO or DAPT after uncaging, visualized for TOPRO (white), kdrl:GFP (green), and uncaged fluorescein (red). White arrowheads point to endothelial cells that contain uncaged fluorescein in DAPT-treated embryos.

(C) Quantification of the number of endothelial nuclei with uncaged fluorescein per focal plane in embryos treated with DMSO (n = 18) or DAPT (n = 13). Embryos treated with DAPT contained 1.61 (s = 1.01) fluorescein-positive endothelial cells, in comparison to 0.16 (s = 0.37) in DMSO-treated embryos.

(D) Our proposed model: Notch signaling negatively regulates endothelial cell number by limiting the mesodermal cells that can differentiate into endothelial progenitors. Attenuated Notch activity causes mesodermal cells with hematopoietic potential to differentiate into endothelial progenitors, thus increasing the number of endothelial cells.

DAPT Treatment Leads to an Increased Number of Endothelial Cells at the Fifth Somite Level
(A) Transverse sections of 18 hpf embryos, DMSO- or DAPT-treated, visualized for TOPRO (white), kdrl:GFP (green), and phalloidin (red). (B) Quantification of endothelial nuclei per focal plane in DMSO-treated (n=30) or DAPT-treated (n=35) embryos. Embryos treated with DAPT contained 5.74 (s=1.37) endothelial nuclei, while those treated with DMSO had 3.93 (s=1.57) nuclei per section (ANOVA, p<10^-5), indicating that the effect of Notch signaling on the number of endothelial cells is not limited to specific regions within the embryos.

mib^ta52b Mutant Embryos Display an Increased Number of Endothelial Cells
Transverse sections of 18 hpf embryos, phenotypic wild-type (A) or mib^ta52b (B), visualized for TOPRO (white), kdrl:GFP (green), and phalloidin (red). (C) The number of endothelial nuclei per focal plane was quantified. Homozygous mib^ta52b mutant embryos exhibited an increased number of endothelial cells.

Multiple notch and delta Genes Are Expressed in the Zebrafish Gastrula
Expression of known notch (notch1a, notch1b, notch2, and notch3) and delta (deltaa, deltab, deltac, and deltad) genes at 6 hpf were analyzed by RT-PCR (A). Whole mount RNA in situ hybridization showing the expression of known notch and delta genes in 6 hpf embryos (B).

A Distinct Role for Notch Signaling in the Formation of Endothelial Progenitors
Transverse sections of embryos treated with DMSO (A-F) or DAPT (G-L), visualized for kdrl:GFP ( and K) (green) and EphrinB2 (B, D, F, H, J and L) (red). Embryos were treated from 10 to 18 hpf (A-B, and G-H), from 10 to 14 hpf (C-D, and I-K), or from 14 to 18 hpf (E-F, and K-L). White dotted lines highlight the cell boundaries, and white arrowheads point to the expression of EphrinB2 within endothelial cells. DAPT treatment from 10 to 14 hpf leads to an increase in endothelial precursors, but does not affect EphrinB2 expression. Treatment from 14 to 18 hpf leads to a loss of EphrinB2 expression, but does not affect endothelial cell number.

Cell Proliferation Is Unaffected in Embryos with Reduced Notch Signaling
Transverse sections (A-F, H-K, and M-P) or lateral whole mount view (G and L) of 18 hpf embryos treated with DMSO (A-C and G-K) or DAPT (D-F and L-P), with two-hour pulse of 10mM BrdU starting at 12 hpf (A-F) or stained with anti-phospho-Histone H3 (G-P), visualized for kdrl:GFP (A, D, G, K, L, and K) (green), BrdU (B and E) or TOPRO (I ,K, N, and P) (white), and anti-phospho-Histone H3 (G, J, L, and O) (red). Inhibition of Notch signaling did not increase the number of proliferating endothelial cells.

Notch Signaling Does Not Affect the Number of Other Nonaxial Mesoderm-Derived Lineages
(A) Transverse sections of 18 hpf embryos, DMSO- or DAPT-treated, visualized for TOPRO (white), pax2a:GFP (green), and phalloidin (red). (B) Quantification of pronephric duct cell nuclei per focal plane in DMSO-treated (n=13) or DAPT-treated (n=16) embryos. (C) Transverse sections of 18 hpf embryos, DMSO- or DAPT-treated, visualized for TOPRO (white), S59 antibody (green), and phalloidin (red). (D) Quantification of slow muscle fiber cell nuclei per focal plane in DMSO-treated (n=20) or DAPT-treated (n=20) embryos. Embryos treated with DAPT contained 12.50 (s=1.63) pronephric duct cell nuclei and 29.20 (s=2.84) slow muscle fiber cell nuclei per section, which were comparable to DMSO-treated embryos (12.23 (s=2.05) pronephric duct cell nuclei and 29.70 (s=2.77) slow muscle fiber cell nuclei).

Gross Developmental Process Is Unaffected in Embryos with Modulated Notch Signaling
Bright-field micrographs of 18 hpf DMSO- (A and E) or DAPT-treated embryos (B and F), and phenotypic wild-type sibling (C and G) and NICD over-expressing embryos (D and H). Both the morphology (A-D) and the expression of myoD (E-G) in DAPT-treated or NICD overexpressing embryos appear to be comparable to DMSO-treated or phenotypic wild-type sibling embryos, suggesting that modulation of Notch does not affect the gross developmental process.