The ubiquity and ancestry of insect doublesex

Abstract

The doublesex (dsx) gene functions as a molecular switch at the base of the insect sex determination cascade, and triggers male or female somatic sexual differentiation in Drosophila. Having been reported from only seven current insect orders, the exact phylogenetic distribution of dsx within the largest Arthropod sub-phylum, the Hexapoda, is unknown. To understand the evolution of this integral gene relative to other arthropods, we tested for the presence of dsx within public EST and genome sequencing projects representative of all 32 hexapod orders. We find the dsx gene to be ubiquitous, with putative orthologs recovered from 30 orders. Additionally, we recovered both alternatively spliced and putative paralogous dsx transcripts from several orders of hexapods, including basal lineages, indicating the likely presence of these characteristics in the hexapod common ancestor. Of note, other arthropods such as chelicerates and crustaceans express two dsx genes, both of which are shown to lack alternative splicing. Furthermore, we discovered a large degree of length heterogeneity in the common region of dsx coding sequences within and among orders, possibly resulting from lineage-specific selective pressures inherent to each taxon. Our work serves as a valuable resource for understanding the evolution of sex determination in insects.

Figure 1. Summary of evidence for hexapod doublesex recovered in this study.

Orders for which we identified a high-confidence dsx EST encoding both OD1 and OD2 domains (see Fig. S1) are marked ‘LINKED’; those for which the domain was found on a singleton EST contig are marked “YES” (see Figs S2, S3 and S4), and those for which we recovered only a Type-B OD1 domain (i.e Type-A absent, see Fig. S3) are annotated as such. Phylogeny as per Misof et al. (2014).

Figure 2. Alternative splicing of doublesex transcripts from Anurida maritime (Collembola) and Lepismachilis y-signata (Archaeognatha).

The consensus EST contig from each species (top, bold text), with short-read data mapped below illustrates the diverging transcript isoforms. The common 5′ OD2 sequence is boxed in each species, while the diverging 3′ ends are in red and orange text.

Figure 3. Distance in nucleotides (grey) between the OD1 (red) and OD2 domains (blue), for each high-confidence doublesex transcript reported in this study.

Total EST length (which may include untranslated regions) is to the right of each bar.