(A) Dorsal view of a 10 dpf Tg(plvap:EGFP);Tg(glut1b:mCherry) head immunostained for Claudin-5 (magenta) shows heterogeneous Tg(plvap:EGFP) and Tg(glut1b:mCherry) expression levels in the brain and meningeal vasculature. Strong Tg(plvapb:EGFP) and weak Tg(glut1b:mCherry) expression was observed in the mCP vasculature. (B) Dorsal view of a 10 dpf Tg(kdrl:EGFP) larval head immunostained for Claudin-5 (magenta). (C–C") Magnified images of the top boxed area in (B) show Claudin-5⁺ mesencephalic veins (MsV, yellow arrows). (D–D") Magnified images of the bottom boxed area in (B) show Claudin-5^- mCP vasculature composed of the dorsal longitudinal vein (DLV) and posterior cerebral vein (PCeV). (E–H) Dorsal views of 32 (E), 46 (F), 54 (G), and 72 (H) hpf cranial vasculature visualized by Tg(kdrl:EGFP) expression. The DLV began sprouting before 46 hpf (F) and extended posteriorly toward the myelencephalic choroid plexus (mCP) to connect with the bilateral PCeV around 54 hpf (G) to form the mCP vasculature through 72 hpf (H). (I) Schematic representation of the dorsal view of the cranial vasculature at 72 hpf. Scale bars: 50 µm.

(A) Dorsal view of a 10 dpf Tg(plvap:EGFP);Tg(glut1b:mCherry) head immunostained for Claudin-5 (magenta) shows heterogeneous Tg(plvap:EGFP) and Tg(glut1b:mCherry) expression levels in the brain and meningeal vasculature. Strong Tg(plvapb:EGFP) and weak Tg(glut1b:mCherry) expression was observed in the mCP vasculature. (B) Dorsal view of a 10 dpf Tg(kdrl:EGFP) larval head immunostained for Claudin-5 (magenta). (C–C") Magnified images of the top boxed area in (B) show Claudin-5⁺ mesencephalic veins (MsV, yellow arrows). (D–D") Magnified images of the bottom boxed area in (B) show Claudin-5^- mCP vasculature composed of the dorsal longitudinal vein (DLV) and posterior cerebral vein (PCeV). (E–H) Dorsal views of 32 (E), 46 (F), 54 (G), and 72 (H) hpf cranial vasculature visualized by Tg(kdrl:EGFP) expression. The DLV began sprouting before 46 hpf (F) and extended posteriorly toward the myelencephalic choroid plexus (mCP) to connect with the bilateral PCeV around 54 hpf (G) to form the mCP vasculature through 72 hpf (H). (I) Schematic representation of the dorsal view of the cranial vasculature at 72 hpf. Scale bars: 50 µm.

(A–D) Dorsal views of 102 hpf WT (A), vegfab^-/- (B), vegfc^-/- (C), and vegfd^-/- (D) cranial vasculature visualized by Tg(kdrl:EGFP) expression. A significant fraction of vegfab^-/- and vegfc^-/- larvae lacked the DLV at 102 hpf. (E and F) Percentage of the fish of indicated genotype with and without the DLV at 102 (E) and 154 (F) hpf. A significantly higher fraction of vegfab^-/- and vegfc^-/- fish lacked the DLV at 102 hpf (47% and 40%, respectively) than WT (0%) (n = 13 for WT, n = 19 for vegfab^-/-, n = 25 for vegfc^-/-, and n = 19 for vegfd^-/- fish examined at 102 hpf). At 154 hpf, while 63% of vegfab^-/- fish exhibited the ‘No DLV’ phenotype, only 13% of vegfc^-/- fish showed this phenotype (n = 16 for WT, n = 27 for vegfab^-/-, n = 23 for vegfc^-/-, and n = 26 for vegfd^-/- fish examined at 154 hpf). Only a small fraction of vegfd^-/- fish examined lacked the DLV at 102 and 154 hpf (5% and 0%, respectively). (G–H") Dorsal views of 96 hpf vegfab^+/+ (G–G") and vegfab^-/- (H–H") heads of larvae that carry the Tg(kdrl:EGFP) transgene (G and H) and that were injected with streptavidin-conjugated Qdot 655 nanocrystals (G’ and H’) (merged images shown in G" and H"). In vegfab^+/+ fish (n = 10), the DLV is visualized by both Tg(kdrl:EGFP) expression (G) and blood-circulating Qdot 655 nanocrystals (G’). In contrast, in the vegfab^-/- larvae that lack the Tg(kdrl:EGFP)-positive DLV (H, n = 8), blood-circulating Qdot 655 nanocrystals did not label the DLV (H’), showing the lack of a perfused, functional DLV. Yellow arrows point to the presence and absence of the DLV in vegfab^+/+ and vegfab^-/- larvae, respectively. Scale bars: 50 µm.

(A) Experimental time course of the heat-shock treatments. Tg(hsp70l:sflt1), Tg(hsp70l:sflt4), and their control siblings were subjected to a heat-shock at 41, 49, 57, and 65 hpf and analyzed at 72 hpf. (B–G) Dorsal views of the cranial vasculature of 72 hpf Tg(kdrl:EGFP) (B and E), Tg(hsp70l:sflt1);Tg(kdrl:EGFP) (C and F), and Tg(hsp70l:sflt4);Tg(kdrl:EGFP) (D and G) larvae after treatment with (E–G) or without (B–D) multiple heatshocks. (H–K) Percentage of the 72 hpf fish with and without the DLV after the indicated treatment (H and I). Quantification of the DLV lengths of fish subjected to the indicated treatment (J and K). The results involving sFlt1 overexpression are shown in (H and J) and those involving sFlt4 overexpression are shown in (I and K). The results of their control siblings with and without the heatshocks are shared in the graphs for both groups. Heatshock-induced overexpression of sFlt4 led to a significantly increased percentage of the ‘No DLV’ phenotype (33%) compared to the other control groups (I), while DLV lengths were significantly shorter in the fish subjected to heatshock-induced overexpression of sFlt1 compared to the other control groups (J). The number of the animals examined is listed in the graph. Data are means ± SD. Scale bar: 50 µm.

(A) Schematic representation of the dorsal view of WT cranial vasculature, illustrating the mCP vasculature (DLV, PCeV) and the neighboring cranial vessels (MsV, MCeV). (B) Percentage of 96 hpf fish of indicated genotype with and without the DLV. vegfab^-/-;vegfd^-/- larvae displayed a significantly higher penetrance of the ‘No DLV’ phenotype (97%) than vegfab^-/- (69%) and vegfd^-/- (12.5%) larvae, showing that vegfab and vegfd genetically interacts in DLV formation (n = 24 for vegfd^-/-, n = 36 for vegfab^-/-, n = 41 for vegfab^-/-;vegfd^+/-, and n = 29 for vegfab^-/-;vegfd^-/- fish). (C) Quantification of PCeV, MCeV, and MsV formation at 96 hpf. PCeV formation was severely compromised in vegfab^-/-;vegfd^-/- larvae compared to the corresponding single mutants. At this stage, MsV, but not MCeV, formation was mildly affected in vegfab;vegfd double mutants compared to the corresponding single mutants. Data are means ± SD.

(A–C") Dorsal views of 120 hpf WT (A–A"), vegfab^-/- (B–B"), and vegfc^-/-;vegfd^-/- (C–C") heads of larvae that carried the Tg(kdrl:EGFP) (A–C) and Tg(lyve1:DsRed) (A’–C’) transgenes (merged images shown in A"–C"). All of the vegfab^-/- larvae that lacked the DLV (n = 7) formed FGPs in the dorsal surfaces of the brain meninges (B–B"), as observed in WT larvae (n = 13). In contrast, vegfc^-/-;vegfd^-/- larvae (C–C") that exhibited the complete loss of FGPs (n = 9) formed the DLV and PCeV, as observed in WT fish. Yellow arrows point to the DLV and white arrows to the FGPs. Scale bar: 50 µm.

(A–C") Dorsal views of 120 hpf WT (A–A"), vegfab^-/- (B–B"), and vegfc^-/-;vegfd^-/- (C–C") heads of larvae that carried the Tg(kdrl:EGFP) (A–C) and Tg(lyve1:DsRed) (A’–C’) transgenes (merged images shown in A"–C"). All of the vegfab^-/- larvae that lacked the DLV (n = 7) formed FGPs in the dorsal surfaces of the brain meninges (B–B"), as observed in WT larvae (n = 13). In contrast, vegfc^-/-;vegfd^-/- larvae (C–C") that exhibited the complete loss of FGPs (n = 9) formed the DLV and PCeV, as observed in WT fish. Yellow arrows point to the DLV and white arrows to the FGPs. Scale bar: 50 µm.

(A) Predicted domain structure of zebrafish Ccbe1. Ccbe1 consists of a signal peptide (SP), an EGF domain, a calcium-binding EGF domain (Ca-EGF), and two collagen repeat domains (ColA and ColB). Arrows indicate the approximate positions of the protein sequences corresponding to the target sequences of the three designed CRISPR RNA (crRNA). (B) Experimental flows of the microinjection experiments for panels (C and D). Injection cocktails containing Cas9 protein with or without the three ccbe1 crRNA:tracrRNA duplex complexes were injected into one-cell stage embryos produced from Tg(kdrl:EGFP);Tg(lyve1:DsRed) fish crossed with wild-type AB counterparts. Injected progeny were analyzed at 120 hpf for the presence or absence of Tg(lyve1:DsRed)⁺ FGPs in the dorsal meningeal surfaces of the optic tectum. (C) Brightfield and confocal images of 120 hpf Tg(kdrl:EGFP);Tg(lyve1:DsRed) larvae after no injection or Cas9 protein injection with and without the three ccbe1 crRNA:tracrRNA complexes. Although the larvae injected with the ccbe1 crRNA ribonucleoprotein (RNP) complexes exhibited no apparent differences in their gross morphology compared to uninjected and Cas9 injected controls, they completely lacked Tg(lyve1:DsRed)⁺ FGPs (white arrows) in the dorsal meningeal surfaces of the optic tectum. (D) Percentage of 120 hpf larvae of indicated treatment with and without Tg(lyve1:DsRed)⁺ FGPs in the dorsal meningeal surfaces of the optic tectum (n = 20 animals examined for each group). (E) Experimental flows of the microinjection experiments for panel (F and G). (F) Percentage of 54 hpf larvae of indicated treatment with and without the DLV (the number of the animals examined per treatment is listed in the panel). A significant fraction of the embryos injected with the ccbe1 crRNA RNP complexes lacked the DLV at this stage. (G) Quantification of DLV lengths of the fish of indicated treatment that formed the DLV at 54 hpf. (H) Experimental flows of the microinjection experiments for panels (I–K). (I) Dorsal views of 96 hpf vegfab^+/+ and vegfab^-/- larvae that carried Tg(kdrl:EGFP) and Tg(lyve1:DsRed) transgenes after Cas9 protein injection with and without the three ccbe1 crRNA:tracrRNA complexes. While control solution injected vegfab^+/+ and vegfab^-/- larvae formed Tg(lyve1:DsRed)⁺ FGPs in the dorsal surfaces of the brain meninges, those injected with the ccbe1 crRNA RNP complexes completely lacked FGPs regardless of vegfab genotypes. A significant fraction of vegfab^-/- larvae in both injection groups exhibited the ‘No DLV’ phenotype. (J) Percentage of 96 hpf larvae of indicated genotype and treatment with and without the DLV (the number of the animals examined per genotype is listed in the panel). vegfab^-/- larvae injected with the ccbe1 crRNA RNP complexes displayed a markedly increased penetrance of the ‘No DLV’ phenotype. (K) Quantification of PCeV, MCeV, and MsV formation at 96 hpf. vegfab^-/- larvae injected with the Cas9 control solution as well as vegfab^+/+ larvae injected with the ccbe1 crRNA RNP complexes did not exhibit a defect in PCeV formation. However, vegfab^-/- larvae injected with the ccbe1 crRNA RNP complexes exhibited severe defects in PCeV formation. Scale bars: 50 μm in (C) (fluorescence image), (J); 1 mm in (C) (brightfield image).

Figure 4—source data 1.

(A–D") Magnified views of 10 dpf WT (A–A"), vegfab^-/- (B–B"), vegfab^-/-;vegfc^-/- (C–C"), and vegfab^-/-;vegfd^-/- (D–D") larvae carrying the Tg(kdrl:EGFP) transgene that were immunostained for GFP and Claudin-5 (magenta). In 10 WT fish, Claudin-5 immunoreactivity is prominent in mCP epithelial cells and weakly present in the tissues located ventrally to the DLV and PCeV but is barely detectable in the DLV and PCeV (A–A’). No obvious differences were observed in this low level of Claudin-5 expression in these vessels between WT and any of the vegfab^-/-, vegfab^-/-;vegfc^-/-, and vegfab^-/-;vegfd^-/- larvae that formed these vessels (B–D"). Yellow arrows point to the DLV and white arrows to the PCeV. The number of the animals examined was: n = 65 for WT, n = 50 for vegfab^-/-, n = 26 for vegfab^-/-;vegfc^-/-, and, n = 24 for vegfab^-/-;vegfd^-/- fish. Scale bar: 50 μm.

(A–D") Magnified views of 10 dpf WT (A–A"), vegfab^-/- (B–B"), vegfab^-/-;vegfc^-/- (C–C"), and vegfab^-/-;vegfd^-/- (D–D") larvae carrying the Tg(kdrl:EGFP) transgene that were immunostained for GFP and Claudin-5 (magenta). In 10 WT fish, Claudin-5 immunoreactivity is prominent in mCP epithelial cells and weakly present in the tissues located ventrally to the DLV and PCeV but is barely detectable in the DLV and PCeV (A–A’). No obvious differences were observed in this low level of Claudin-5 expression in these vessels between WT and any of the vegfab^-/-, vegfab^-/-;vegfc^-/-, and vegfab^-/-;vegfd^-/- larvae that formed these vessels (B–D"). Yellow arrows point to the DLV and white arrows to the PCeV. The number of the animals examined was: n = 65 for WT, n = 50 for vegfab^-/-, n = 26 for vegfab^-/-;vegfc^-/-, and, n = 24 for vegfab^-/-;vegfd^-/- fish. Scale bar: 50 μm.

(A–F’) Endogenous mRNA transcripts of vegfab (A and A’), vegfc (C and C’), and vegfd (E and E’) at the indicated embryonic stages were detected by in situ hybridization. Lateral views of the EGFP expression patterns driven by the Tg(UAS:EGFP) embryo carrying the TgBAC(vegfab:gal4ff) (B and B’), TgBAC(vegfc:gal4ff) (D and D’), or TgBAC(vegfd:gal4ff) (F and F’) transgene. The EGFP expression patterns driven by each of the BAC transgenic lines closely matched those of the corresponding endogenous mRNA. (G–J) Dorsal views of 48 hpf Tg(kdrl:ras-mCherry) embryos with no additional gene (G) or carrying the TgBAC(vegfab:EGFP) (H), TgBAC(vegc:EGFP) (I), or TgBAC(vegd:EGFP) (J) transgene. No detectable auto-fluorescence was observed in the control embryo (G) under these microscopy settings. TgBAC(vegfab:EGFP), TgBAC(vegfc:EGFP), and TgBAC(vegfd:EGFP) expression was observed in dorsal meningeal surfaces and distinct cell types. (K and L) Lateral (K) and dorsal (L) views of 72 hpf TgBAC(vegfab:EGFP);Tg(kdrl:ras-mCherry) larvae, showing the TgBAC(vegfab:EGFP) expression in the DMJ and mCP. TgBAC(vegfab:EGFP) expression was not observed in the DLV. Scale bars: 200 μm in (F); 100 μm in (F’); 50 μm in (J), (K), (L).

(A–F’) Endogenous mRNA transcripts of vegfab (A and A’), vegfc (C and C’), and vegfd (E and E’) at the indicated embryonic stages were detected by in situ hybridization. Lateral views of the EGFP expression patterns driven by the Tg(UAS:EGFP) embryo carrying the TgBAC(vegfab:gal4ff) (B and B’), TgBAC(vegfc:gal4ff) (D and D’), or TgBAC(vegfd:gal4ff) (F and F’) transgene. The EGFP expression patterns driven by each of the BAC transgenic lines closely matched those of the corresponding endogenous mRNA. (G–J) Dorsal views of 48 hpf Tg(kdrl:ras-mCherry) embryos with no additional gene (G) or carrying the TgBAC(vegfab:EGFP) (H), TgBAC(vegc:EGFP) (I), or TgBAC(vegd:EGFP) (J) transgene. No detectable auto-fluorescence was observed in the control embryo (G) under these microscopy settings. TgBAC(vegfab:EGFP), TgBAC(vegfc:EGFP), and TgBAC(vegfd:EGFP) expression was observed in dorsal meningeal surfaces and distinct cell types. (K and L) Lateral (K) and dorsal (L) views of 72 hpf TgBAC(vegfab:EGFP);Tg(kdrl:ras-mCherry) larvae, showing the TgBAC(vegfab:EGFP) expression in the DMJ and mCP. TgBAC(vegfab:EGFP) expression was not observed in the DLV. Scale bars: 200 μm in (F); 100 μm in (F’); 50 μm in (J), (K), (L).

(A–C") Dorsal views of 54 hpf vegfab^+/+ (A–A") and vegfab^-/- (B–C") heads of embryos that carried the TgBAC(vegd:EGFP) and Tg(kdrl:ras-mCherry) transgenes. vegfab^-/- embryos that exhibited the presence (B–B"), or absence (C–C"), of the DLV were examined. Overall TgBAC(vegd:EGFP) expression patterns and signal intensity did not apparently differ between vegfab^+/+ (A’, n = 14) and vegfab^-/- embryos regardless of the presence (B’, n = 7), or absence (C’, n = 4), of the DLV. (D–F") Dorsal views of 48 hpf vegfab^+/+ (D–D") and vegfab^-/- (E–F") heads of embryos that carried the TgBAC(vegc:EGFP) and Tg(kdrl:ras-mCherry) transgenes. vegfab^-/- embryos that exhibited the presence (E–E"), or absence (F–F"), of the DLV were examined. Overall TgBAC(vegc:EGFP) expression patterns and signal intensity did not apparently differ between vegfab^+/+ (D’, n = 3) and vegfab^-/- embryos regardless of the presence (E’, n = 2), or absence (F’, n = 3), of the DLV. Scale bars: 50 μm.

(A–A") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(ntla:gal4);Tg(UAS:Kaede);Tg(sox10:mRFP) embryo, showing intermingled distribution of Kaede⁺ and mRFP⁺ meningeal cells (n = 7 embryos; 180 Tg(ntla:gal4);Tg(UAS:Kaede)⁺ cells examined; 1 cell co-localized with Tg(sox10:mRFP)⁺ cell). Yellow arrows point to some of the Kaede⁺ meningeal cells that were clearly not mRFP⁺ even on the maximum projection images. (B–B") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(sox10:EGFP);TgBAC(vegfab:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 12 embryos; 80 TgBAC(vegfab:NTR)⁺ cells examined; 2 cells co-localized with Tg(sox10:EGFP)⁺ cells). Yellow arrows point to some of the NTR-mCherry⁺ meningeal cells that were clearly not EGFP⁺ even on the maximum projection images. (C–C") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(sox10:EGFP);TgBAC(vegfc:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 4 embryos; 66 TgBAC(vegfc:NTR)⁺ cells examined; 0 cells co-localized with Tg(sox10:EGFP)⁺ cells). Yellow arrows point to some of the NTR-mCherry⁺ meningeal cells that were clearly not EGFP⁺ even on the maximum projection images. (D–D") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(sox10:EGFP);TgBAC(vegfd:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 4 embryos; 66 TgBAC(vegfd:NTR)⁺ cells examined; 2 cells co-localized with Tg(sox10:EGFP)⁺ cells). Yellow arrows point to some of the NTR-mCherry⁺ meningeal cells that were clearly not EGFP⁺ even on the maximum projection images. Scale bar: 50 μm.

(A–A") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(mpeg1.1:Dendra2);TgBAC(vegfab:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 16 embryos; 89 Tg(mpeg1.1:Dendra2)⁺ cells examined; 0 cells co-localized with TgBAC(vegfab:NTR)⁺ cells). (B–B") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(mpeg1.1:Dendra2);TgBAC(vegfc:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 17 embryos; 95 Tg(mpeg1.1:Dendra2)⁺ cells examined; 0 cells co-localized with TgBAC(vegfc:NTR)⁺ cells). (C–C") Confocal maximum projection images of the dorsal views of a 48 hpf Tg(mpeg1.1:Dendra2);TgBAC(vegfd:NTR) embryo, showing non-overlapping distribution of EGFP⁺ and NTR-mCherry⁺ meningeal cells (n = 8 embryos; 39 Tg(mpeg1.1:Dendra2)⁺ cells examined; 0 cells co-localized with TgBAC(vegfd:NTR)⁺ cells). Scale bar: 50 μm.

(A) Schematic of the ntla:EGFP construct used for injection experiments (B–E"). (B–E") Dorsal views of 48 hpf Tg(ntla:NTR) (B–C") and TgBAC(vegfc:NTR) (D–E") heads of embryos that were injected with the ntla:EGFP construct at the one-cell stage. Magnified images of the boxed area in (B) and (D) are shown in (C–C") and (E–E"), respectively. Most of the EGFP⁺ cells (C and E) were co-localized with NTR-mCherry⁺ meningeal cells (C’ and E’) in Tg(ntla:NTR) (C–C") and TgBAC(vegfc:NTR) (E–E") embryos. Yellow arrows point to the co-localized cells (n = 5 embryos, 71 co-localized cells observed for Tg(ntla:NTR); n = 3 embryos, 10 co-localized cells observed for TgBAC(vegfc:NTR)). (F–H) Dorsal magnified views of 48 hpf Tg(ntla:Gal4-VP16);Tg(UAS:EGFP-CAAX), abbreviated Tg(ntla:EGFP), TgBAC(vegfd:EGFP), and TgBAC(vegfab:EGFP) embryonic heads near the DMJ in the dorsal surfaces of the brain meninges. Close similarity in the morphology of the EGFP⁺ meningeal cells marked by each of the lines was observed. Scale bars: 50 μm in (B–E"); 25 μm in (H).

(A–C) Dorsal views of 54 hpf vegfc^+/+ (A), vegfc^um18/+ (B), and vegfc^um18/um18 (C) cranial vasculature visualized by Tg(kdrl:EGFP) expression. vegfc^um18/um18 fish did not exhibit a defect in DLV formation (C) in contrast to the ‘No DLV’ phenotype found in vegfc^{hu6410/hu6410} fish (Figure 1N). (D) Percentage of 54 hpf fish of indicated genotype with and without the DLV (n = 22 for vegfc^+/+, n = 52 for vegfc^um18/+, and n = 21 for vegfc^um18/um18 fish). (E) Time course of the cell ablation experiments for panels (F–M). (F) Brightfield images of 72 hpf Tg(kdrl:EGFP), Tg(sox10:NTR), TgBAC(vegfab:NTR), and TgBAC(vegfd:NTR) larvae after treatment with DMSO or 10 mM Metronidazole (MTZ) from 34 to 72 hpf. (G–L) Dorsal views of 72 hpf Tg(kdrl:EGFP);TgBAC(vegfab:NTR) (G and H), Tg(kdrl:EGFP);TgBAC(vegfd:NTR) (I and J), and Tg(kdrl:EGFP);Tg(sox10:NTR) (K and L) cranial vasculature after treatment with DMSO (G, I, K) or MTZ (H, J, L) from 34 to 72 hpf. Efficient ablation of NTR-mCherry⁺ cells was observed after treatment with MTZ (H, J, L) compared to their respective DMSO-treated controls (G, I, K). MTZ-treated TgBAC(vegfab:NTR) and TgBAC(vegfd:NTR), but not Tg(sox10:NTR), larvae lacked the DLV (arrows). (M) Percentage of the indicated 72 hpf transgenic fish with and without the DLV after treatment with DMSO or MTZ (the number of the animals examined per each treatment group is listed in the graph). MTZ-treated TgBAC(vegfab:NTR) and TgBAC(vegfd:NTR) larvae exhibited a drastic increase in the ‘No DLV’ phenotype when compared to their corresponding DMSO treatment groups and the MTZ-treated Tg(kdrl:EGFP) larval group. In contrast, MTZ-treated Tg(kdrl:EGFP) and Tg(sox10:NTR) animals showed no significant differences compared to their corresponding DMSO treatment groups. Scale bars: 1 mm in (F); 50 μm in (C and L).

(A–B’) Lateral views of endogenous vegfaa mRNA expression patterns detected by in situ hybridization at 24 hpf (A and A’). Lateral views of the EGFP expression patterns driven by the Tg(UAS:EGFP) embryo carrying the TgBAC(vegfaa:gal4ff) transgene (B and B’). The EGFP expression domains closely matched the endogenous vegfaa mRNA expression patterns. (C, C’ and D–D") Lateral (C and C’) and dorsal (D–D") views of 48 hpf TgBAC(vegfaa:EGFP);Tg(kdrl:ras-mCherry) embryos, showing the TgBAC(vegfaa:EGFP) expression in the cells located ventrally to the DLV. TgBAC(vegfaa:EGFP) expression was not observed in the extending DLV or in the dorsal surfaces of the brain meninges. (E) Time course of the cell ablation experiments for panels (F–J). (F and G) Brightfield images of 72 hpf TgBAC(vegfaa:NTR) larvae after treatment with DMSO (F) or 10 mM MTZ (G) from 34 to 72 hpf. (H and I) Dorsal views of 72 hpf TgBAC(vegfaa:NTR) cranial vasculature visualized by Tg(kdrl:EGFP) expression after treatment with DMSO (H) or MTZ (I) from 34 to 72 hpf. Efficient ablation of NTR-mCherry⁺ cells was observed after treatment with MTZ (I) compared to the DMSO-treated control (H). MTZ-treated TgBAC(vegfaa:NTR) larvae formed the DLV (yellow arrows) and PCeV. (J) Percentage of the 72 hpf transgenic fish with and without the DLV after treatment with DMSO or MTZ. No significant difference was observed between MTZ-treated (n = 60) and DMSO-treated (n = 30) groups. Scale bars: 1 mm in (G); 200 μm in (B); 100 μm in (B’); 50 μm in (C), (C’), (D’ and I).