Fig. 5. Fgf3 signaling from the diencephalon, but not the telencephalon, is sufficient for adenohypophyseal development. (A,C-H) Embryos at 32 hpf, lateral views (A,C,F,G), or frontal views (D,F,H), on head regions. (A) Embryo stained for fgf3 transcripts. Note the fissure of the eye vesicle as a reference point for the anterior border of the infundibular fgf3 expression domain. (B) Cartoon showing the transplantation sites at shield stages. Consistent with results obtained in fate-mapping experiments (Woo and Fraser, 1995; Varga et al., 1999; Mathieu et al., 2002), telencephalic chimeras as shown in G,H were obtained by transplanting cells from/to the animal pole, diencephalic chimeras as shown in C-F by transplanting cells from/to dorsal regions anterior/animal of the shield. Cells of the adenohypophysis (ad) are supposed to derive from region indicated by arrowhead. (C-H) Chimeras, in-situ hybridized for lim3 transcripts in blue, and with transplanted wild-type cells in brown. Out-of-focus bilateral lim3 domains in D,F,H represent hindbrain motoneurons (Glasgow et al., 1997). (C,D) Wild-type recipients with wild-type donor cells in the diencephalon, displaying normal adenohypophyseal lim3 expression (indicated by arrows). (E,F) fgf3 mutant recipients with infundibular wild-type cells (indicated by arrowheads) and adjacent rescued adenohypophyseal lim3 expression (indicated by arrows). Inset in E shows second rescued embryo. (G,H) fgf3 mutant recipient with many wild-type cells in the telencephalon, still lacking adenohypophyseal lim3 expression.