Fig. 5
Traces of inversions and high sequence diversification within a pannier intron in ladybird beetles. a Sequence comparison of the genomic region surrounding the pannier locus. 700 kb genomic sequences surrounding the pannier locus were extracted from the genome assembly of each allele in H. axyridis (hC, hA, h) and C. septempunctata (C. sep). Strain names are given in parentheses. Arrows indicate genes predicted by the exonerate program (Orange, pannier; Blue, GATA transcription factor genes paralogous to pannier; Green, other genes). Gene names are listed at the top. Vertical or diagonal bars connecting adjacent genomic structures indicate BLAST72 hit blocks (bluish, forward hit; reddish, reverse hit) in the comparison between the two adjacent genomic scaffolds. The colour code for colouring the bars is at the bottom. The exon−intron structure and the first intron (*) of pannier (1A isoform) is depicted on the top of the panel. The upper half (first intron) of the pannier locus is diversified (whitish) between different h alleles in H. axyridis, and shows traces of inversions (crossed reddish bars). Several intronic sequences are conserved between H. axyridis and C. septempunctata (bars located in the upper half of pannier in C. sep). The black arrow indicates the region specifically expanded in H. axyridis. b Overview of the size of the upper noncoding regions (the first intron + the upper intergenic region) at the pannier (pnr) locus in holometabolous insects. The topology of the phylogenetic tree of surveyed insects is adapted from ref. 107 (Coleoptera), and TIMETREE108 (Diptera, Hymenoptera, Lepidoptera). The sizes of the first introns are given in parenthesis if cDNA information was available. H. axyridis has the largest noncoding sequence at the pannier locus. In some species, synteny of the three paralogous GATA genes was broken up by translocation (*) or insertion (**). c, d ML phylogenetic trees constructed with nucleotide sequences of pannier (pnr) coding region (c), and those of the conserved region in the first intron (d). The trees were drawn to scale with branch lengths measured in the number of substitutions per site. Bootstrap values were calculated from 1000 resampling of the alignment data. Bars, 0.01 substitutions/site
