Underexpression and overexpression of magp1 result in vascular defects in zebrafish embryos. Bright-field images of an uninjected Tg(fli1/EGFP) embryo (A) and an magp1 MO–injected embryo (B) at 30 hours after fertilization. magp1 MO–injected embryos display normal morphology, except for the expanded tail fin (arrowheads). Compared with the uninjected Tg(fli1/EGFP) embryo (C, E, G), magp1 MO–injected Tg(fli1/EGFP) embryos exhibit distended vessels in the brain and the eye (D, arrowheads; 47% ± 16%; N = 60), axial vessels with irregular lumens (F, arrows; 70% ± 3%; N = 36), and dilated vascular channel in the venous plexus (H, arrows) with only a few branches of venous microvessels (H, arrowheads; 96% ±;N = 84). DA indicates dorsal aorta; CV, cardinal vein. To assess functionality of the circulation, microangiography was conducted on Tg(fli1/EGFP) embryos (I, M, uninjected embryos; K, O, magp1 MO–injected embryos) at 48 hours after fertilization using TRITC dextran. Compared with uninjected Tg(fli-1/EGFP) embryos (J, N), magp1 MO–injected Tg(fli1/EGFP) embryos (L, P) display lumenized vasculature. However, pooling of TRITC dextran in dilated regions of the blood vessels is observed in magp1 MO–injected Tg(fli1/EGFP) embryos (L, P, arrowheads). The corresponding region of dilated vessels for panel P is shown in panel O as indicated by arrowheads. (Q) Summary of 3 independent experiments showing alleviation of 3 ng magp1 MO–induced vascular defects with the introduction of Magp1 protein through injection of 10 pg magp1 expression construct (compare columns 2 and 1; P = .01). N indicates the number of embryos scored. In contrast, coinjection of 10 pg control expression construct, FRM-DsRED, does not affect the frequency of the magp1 MO–induced vascular defects (column 3). Error bars indicate SEM. (R) Overexpression of magp1 generates a range of vascular defects, including the dilation of caudal vein (arrowheads). Original magnifications: (A-D, G-H, I-P, R) 5 ×; (E-F) 10 ×.