Fig. 3

Asymmetric evolution of FoxG genes. (a) The maximum likelihood tree of the FoxG genes using 243 aligned sites of amino acid sequences with the JTT + I + G4 amino acid substitution model. The ultrafast bootstrap approximation values of 75 or more are indicated at the nodes. FoxG2 and FoxG1 are categorized in the elusive and non-elusive genes defined in this study, respectively. The absence of FoxG2 orthologs in mammals and birds may be attributable to information loss or secondary gene loss (see Discussion). Whole-mount in situ hybridization for FoxG1 (b) and FoxG2 (f) genes using 3 and 4 dpo embryos of the Madagascar ground gecko, respectively. Expression signals of FoxG1 (c–e) and FoxG2 (g) by sections of in situ hybridization using 2 dpo embryos. FoxG1 is expressed in the otocyst, vestibulocochlear ganglion (shown by arrowheads in c), retina (shown by arrowheads in d), and cerebrum (shown by arrowheads in e), whereas FoxG2 is specifically expressed in the vestibulocochlear ganglion (shown by an arrow in g). (h) The levels of the sections for the gecko embryo are indicated by dashed lines. Whole-mount in situ hybridization for foxg1a (i, the FoxG1 ortholog), foxg1c (m, the FoxG2 ortholog), and foxg1d (o) and foxg1b (t) (the FoxG3 orthologs) using 2 dpf zebrafish embryos. Expression signals of foxg1a (j–l), foxg1c (n), and foxg1d (p–s) by sections of in situ hybridization using the 2 dpf embryos. Any expression signals of foxg1b were not observed in the ganglion. (u) The levels of the sections for the zebrafish embryos are indicated by dashed lines. Scale bars in b–f, 1 mm, and those in i, j, m, o, and t, 100 μm. k, l, n, and p–s have the same scale as j