Schematic representation, protein sequence alignment and maximum likelihood (ML) nucleotidic phylogeny of endothelin and sarafotoxin sequences. The novel toxin gene identified by Whittington and colleagues (named edn1_O.ananitus) is identified by an orange star. For each endothelin (ET) group, we used two mammalian sequences (Homo sapiens and Mus musculus), one archosaurian sequence (Gallus gallus), one amphibian sequence (Xenopus laevis) and one actinopterygian sequence (Danio rerio). Only one copy of the fish-specific-duplicated edn-2 and edn-3 gene products was used in the alignment and in the phylogenetic analysis for D. rerio. Sequences from the fish-specific edn4 group have not been included. (a) The ET-like structure is represented at the top of the figure; the SRTX-like structure is represented at the bottom. The 'X' used in the alignment for Sarafotoxin sequences means that no homologous regions are identified. Three different sarafotoxin isoforms isolated from Atractaspis microlepidota and encoded on the same precursor are used for multiple alignment. The dots used in the alignment for edn2 of the Ornithorhynchus ananitus sequence indicate missing data. The threshold for identity/similarity shading is 80%. (b) ML was performed on the ET/SRTX region of 108 bp using the generalized time-reversible plus invariant plus gamma (GTR+I+G) model. Topology is rooted with sarafotoxin sequences. Bootstrap values greater than 90% are indicated. The identified edn-1, edn-2 and edn-3 platypus sequences are indicated in red.