Fig. S3

Mutation of sites within gnathostome Hox2 enhancers and their influence on tissue-specific activity in transient transgenic zebrafish and lamprey embryos. a, Alignments of the NC3 region of the wild-type mouse (m) Hoxa2 NC enhancer with two variants, ΔNC3_1 and ΔNC3_2, showing the 15bp portions deleted in each variant. The positions of characterised hindbrain (purple) and NC (green) cis-elements are shown above the alignments. b, Dorsal views of transient transgenic zebrafish embryos with GFP expression mediated by the wild-type Hoxa2(m) enhancer and the two ΔNC3 variants. The left otic vesicle of each embryo is circled, with GFP expression in rhombomeres (r) and neural crest (nc) annotated. Embryos are at approximately 30 hours post-fertilisation. c, Alignment of a portion of the wild-type zebrafish (zf) hoxb2a NC enhancer with a variant in which the Pbx-Hox site has been mutated (ΔPbx-Hox). d, st24 transient transgenic lamprey embryos injected with the wild-type hoxb2a(zf) (lateral view) and mutated hoxb2a(zf)ΔPbx-Hox (dorsal view) enhancers. hoxb2a(zf) mediates expression in the hindbrain and neural crest, posterior to PA1 (pharyngeal arches are numbered), while this activity is lost in hoxb2a(zf)ΔPbx-Hox. GFP-expressing embryos shown in b and d are representative of the expression potential of the reporter construct in each case, as inferred from screening many (typically more than 100) injected embryos. The injection statistics for Hoxa2(m) and mutated variants in transient transgenic zebrafish embryos are provided in Supplementary Table 3, while those for the hoxb2a(zf) constructs in lamprey embryos are given in Supplementary Table 2.