Fig. 2. Hox paralogue group 1 genes: sequence and functional domains. (A) Clustal X (Thompson et al., 1997) alignment of amino acid sequences of zebrafish, mouse and amphioxus Hox paralogue group 1 genes. Conceptual translations of the four zebrafish Hox PG1 genes are compared with mouse Hoxa1, Hoxb1 and Hoxd1 and AmphiHox-1. Identical residues in red, conserved changes in blue. Hexapeptide and homeodomain are overlined in blue and green, respectively; note the unusually long linker region between the hexapeptide and homeodomain regions of hoxc1a. The diagnostic PG1 residues (Sharkey et al., 1997) are indicated with asterisks (below sequence). The 2/7 diagnostic residues not conserved in hoxc1a are indicated with black rather than red asterisks. (B) Schematic of intron/exon structure for the four zebrafish PG1 genes, drawn to scale; hexapeptide and homeodomain are indicated in blue and green, respectively, alternatively spliced exon is indicated in yellow. Based on comparison to genomic sequences (GenBank Accession Numbers AF071243, AF071251, U40995, AF071263). Note that hoxc1a has no intron, as confirmed by PCR on genomic DNA. Numbers indicate intron/exon boundaries with respect to the start of translation; the length of the primary intron is also indicated. The following coding sequences have been placed on the EMBL database: hoxc1a, Accession Number AJ306432; hoxb1b, Accession Number, AJ306433; and hoxa1a, Accession Numbers AJ306430 and AJ306431. (C) Neighbour-joining tree to show the phylogenetic relationships between Hox PG1 genes (based on Clustal X alignment in A; displayed using NJ-Plot) (Perriere and Gouy, 1996), bootstrap values based on 1000 replicates are shown; scale bar refers to branch lengths. The tree suggests that mouse Hoxd1 and zebrafish hoxc1a group together; however, the long branch lengths imply that these genes are more distantly related.