Thalidomide increases endogenous ceramide during the development of zebrafish embryos, and exogenous C2-ceramide mimics thalidomide-induced antiangiogenic action. (A,B) Zebrafish embryos were treated at 75% epiboly stage (8 hpf) with the indicated concentrations of thalidomide, and harvested at various times (A) or 18 hours after treatment (B). Lipids were extracted, and ceramide content was measured by the DGK assay as described in “Materials and methods.” (A) ○ indicates control; ♦, 400 μM thalidomide; and •, 800 μM thalidomide. The results were obtained from 3 independent experiments. The error bars indicate 1 SD. *The difference in ceramide content between the embryos treated with thalidomide and the control embryo is statistically significant at P < .01 by ANOVA test. (C-F) Lateral view of fixed embryo shows that 10 μM C2-ceramide–treated embryos (D) are shorter in length than the control (C), as indicated by white arrows. Hematoxylin and eosin staining for the transverse section of C2-ceramide–treated embryos shows that the lumens of the dorsal artery (marked “A” in panel E) and posterior cardinal vein (marked “V” in panel E), which are evident in the control embryo (E), are lost in the C2-ceramide–treated embryo (F). (G-L) Expressions of neuropilin-1 and FLK-1 mRNA at the region corresponding to dorsal artery and vein are not detected after treatment with C2-ceramide as well as thalidomide (indicated by boxes; H,K). The defect of FLK-1 expression is more evident by the higher magnification (J,L). The results are representative of each experiment. noto indicates notochord; A, dorsal artery; V, posterior cardinal vein; hpf, hours after fertilization. Original magnifications of panels are as follows: C, D, G, I, H, and K, ×40; E and F, ×150; and J and L, ×63. Image acquisition was performed as described for Figure 1.