Fig. 3

a–g Four-dimensional imaging of tg(kdrl:eGFP) transgenic embryos. a–cptena^−/−ptenb^−/− mutant embryos and d–g wild-type embryos. Imaging was done from 35 hpf onwards following treatment with 5 µM LY294002 from 32 hpf. Still frames from movie S3 (a–c) and movie S4 (d–g). arrowheads: HSPCs. Asterisks: disintegrating HSPCs. Different colors of arrowheads distinguish separate EHT events. Images were taken with ×40 objective and ×1 zoom. Time in hh:mm; DA dorsal aorta, PCV posterior cardinal vein. h, i CHTs of tg(kdrl:mCherry-CAAX/cd41:eGFP) control (n = 10) and LY294002-treated (5 µM, 32–50 hpf) (n = 16) embryos were imaged at 50 hpf. The vasculature is highlighted in red (mCherry) and some GFPlow HSPCs are indicated by arrows. j, k CHTs of tg(cd41:eGFP) control (n = 11) and LY294002-treated (5 µM, 30–60 hpf) (n = 19) embryos were imaged at 4 dpf. Anterior to the left; 2 µm step size. Representative embryos are shown and the number of embryos that showed this pattern/total number of embryos is indicated. The number of GFP^low HSPCs was determined at 50 hpf (l) and 4 dpf (m) and is expressed as average number of cells; error bars indicate standard error of the mean (SEM). Shapiro–Wilk test for normal distribution and Welch’s two-tailed t-test were used for statistical analysis; ***p < 0.001.