Fig. 6

Figure 6. The stand-alone iLV-to-BV transdifferentiation is dependent on Notch and critical for postinjury survival

(A?C) Transdifferentiation of the stand-alone iLVs (A, arrow) was inhibited by the LEC-specific overexpression of dnMAML-Flag (B, arrowhead). The statistics show the ratios of Dendra2+GFP+ vessels among all the GFP+ vessels (C) (n = 10 larvae; two-tailed unpaired t test; ^???, p < 0.0001).

(D?F) The track iLVs (D, arrows) were induced to undergo LV-to-BV transdifferentiation by the LEC-specific overexpression of NICD-HA (E, arrowhead). The statistics show the ratios of Dendra2+GFP+ vessels among all the GFP+ vessels (F) (n = 10 larvae; two-tailed unpaired t test; ^???, p < 0.0001).

(G?K) In contrast to the wild-type (G), the uninjured dll4 heterozygotes (H) were physiologically normal including brain BVs and BLECs. But after Mtz treatment, the stand-alone iLV-to-BV transdifferentiation (I, arrow) failed to occur in the dll4 heterozygotes (J, arrowhead). The statistics show ratios of Dendra2+GFP+ vessels among all the GFP+ vessels (K) (n = 10 larvae; two-tailed unpaired t test; ^???, p < 0.0001).

(L?T) The dll4 heterozygotic mutation (L?N), the LEC-specific overexpression of dnMAML-Flag (O?Q), and general overexpression of EphrinB2a (R?T), led to defects in the formation of early-regenerated BVs and became lethal. The statistics show the survival rates (N, n = 32; Q, n = 40; T, n = 36). Note that the majority of the larvae with defective formation of early-regenerated BVs died before 6 dpt, while all of them died by 7 dpt.

Scale bar, 50 ?m. Data are represented as mean ± SD. Hets, heterozygotes; HS, heat shock. See also Figures S6 and S7.