The genomic architecture and molecular evolution of ant odorant receptors

Abstract

The massive expansions of odorant receptor (OR) genes in ant genomes are notable examples of rapid genome evolution and adaptive gene duplication. However, the molecular mechanisms leading to gene family expansion remain poorly understood, partly because available ant genomes are fragmentary. Here, we present a highly contiguous, chromosome-level assembly of the clonal raider ant genome, revealing the largest known OR repertoire in an insect. While most ant ORs originate via local tandem duplication, we also observe several cases of dispersed duplication followed by tandem duplication in the most rapidly evolving OR clades. We found that areas of unusually high transposable element density (TE islands) were depauperate in ORs in the clonal raider ant, and found no evidence for retrotransposition of ORs. However, OR loci were enriched for transposons relative to the genome as a whole, potentially facilitating tandem duplication by unequal crossing over. We also found that ant OR genes are highly AT-rich compared to other genes. In contrast, in flies, OR genes are dispersed and largely isolated within the genome, and we find that fly ORs are not AT-rich. The genomic architecture and composition of ant ORs thus show convergence with the unrelated vertebrate ORs rather than the related fly ORs. This might be related to the greater gene numbers and/or potential similarities in gene regulation between ants and vertebrates as compared to flies.

Figure 1.

Genome assembly and annotation of the clonal raider ant. (A) Map of contigs within clonal raider ant chromosomes. Alternating gray and black indicate different contigs. Large stretches of uninterrupted gray or black indicate very large contigs, often containing most of a chromosome arm. Putative centromeres were identified based on Hi-C interaction (see Methods). (B) Whole-genome alignment of the second generation draft assembly (Oxley et al. 2014) with the new assembly. Largely continuous, noninverted alignment across the genome indicates little large-scale structural variation between the two assemblies. (C,D) BLAST hit overlap between clonal raider ant gene sets and the Drosophila melanogaster official gene set (v6). The good performance of OGS 4.0.2 in terms of number of genes, and intermediate performance in terms of fraction of genes, indicates high sensitivity and intermediate specificity. (E) Number of odorant receptors found in the annotation based on the previous assembly (OGS 1.8.6) vs. the current annotation (OGS 4.0.2). The large increase in intact ORs and decrease in predicted pseudogenes shows that the previous OR annotations were compromised by poor assembly and misassembly.

Figure 2.

Phylogeny and genomic organization of odorant receptors in four Hymenoptera species. Branches from the five largest ant OR subfamilies are colored by subfamily. OR loci indices are assigned by ordering loci as they appear in the phylogeny from top to bottom. Sets of vertically aligned points represent a single tandem array, while breaks in a line represent transposition events. Loci indices assigned this way are not homologous between species. Loci ages are indicated by color and calculated via phylostratigraphy and synteny analysis (see Methods). The few large, contiguous, and old tandem arrays in the top half of A and C and throughout B indicate tandem duplications as the predominant drivers of gene expansion in these gene clades. The many short species-specific tandem arrays in the bottom half of A and C and throughout D indicate that genomic transpositions followed by tandem duplications drove expansions in these clades.

Figure 3.

Genomic colocalization of transposable elements (TEs), odorant receptor genes, and other genes. (A) Percentage of OR genes and other genes within TE islands. (B) Density of TEs in OR loci and other gene-dense regions (genomic stretches with at least one OR gene [OR loci] or one non-OR gene [gene-dense regions] per 10 kbp) and within the genome as a whole. These data show that TE islands are depauperate in ORs, but OR loci are enriched for TEs relative to other gene-rich loci and the genome as a whole.

Figure 4.

Nucleotide composition of odorant receptors (ORs) vs. other genes in the clonal raider ant and the vinegar fly (Drosophila melanogaster). (A) Frequency distribution of AT content for coding sequence (CDS), introns, downstream, and upstream sequence of OR genes (orange) and other genes (cyan). Y-axes are log-scaled to allow visualization of fly ORs, which make up a small proportion of fly genes. (B) Mean nucleotide composition along the CDS, intronic, and flanking sequences of OR genes or other genes in the clonal raider ant (left) and the vinegar fly (right). These data show that clonal raider ant ORs are AT-rich relative to most other genes, while fly ORs possess typical AT content relative to other genes.