Diversity and evolution of mariner-like elements in aphid genomes

Abstract

BackgroundAlthough transposons have been identified in almost all organisms, genome-wide information on mariner elements in Aphididae remains unknown. Genomes of Acyrthosiphon pisum, Diuraphis noxia and Myzus persicae belonging to the Macrosiphini tribe, actually available in databases, have been investigated.ResultsA total of 22 lineages were identified. Classification and phylogenetic analysis indicated that they were subdivided into three monophyletic groups, each of them containing at least one putative complete sequence, and several non-autonomous sublineages corresponding to Miniature Inverted-Repeat Transposable Elements (MITE), probably generated by internal deletions. A high proportion of truncated and dead copies was also detected. The three clusters can be defined from their catalytic site: (i) mariner DD34D, including three subgroups of the irritans subfamily (Macrosiphinimar, Batmar-like elements and Dnomar-like elements); (ii) rosa DD41D, found in A. pisum and D. noxia; (iii) a new clade which differs from rosa through long TIRs and thus designated LTIR-like elements. Based on its catalytic domain, this new clade is subdivided into DD40D and DD41D subgroups. Compared to other Tc1/mariner superfamily sequences, rosa DD41D and LTIR DD40-41D seem more related to maT DD37D family.ConclusionOverall, our results reveal three clades belonging to the irritans subfamily, rosa and new LTIR-like elements. Data on structure and specific distribution of these transposable elements in the Macrosiphini tribe contribute to the understanding of their evolutionary history and to that of their hosts.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-017-3856-6) contains supplementary material, which is available to authorized users.

Diversity of <i>mariner</i>-like elements in Orthoptera

“…Cluster VII includes vertumnana and marmoratus subfamilies (Bui et al, 2007(Bui et al, , 2008. Although no previously defined subfamilies were found grouped in cluster I, it includes elements recently identified in the insect order Hemiptera (Bouallègue et al, 2017) and assigned to irritans subfamily (which, however it is not supported by our phylogenetic analysis). Clusters II and IV are comprised of single Orthoptera lineage each (Ort-mar37 and Ortmar45, respectively).…”

“…Therefore, to fully understand the evolution of a transposable element lineage it is important to evaluate its horizontal transfer rate. For 16 lineages identified in the studied Orthoptera genomes, horizontal transfer between dis- (Bouallègue et al, 2017) Apismar1.1 Acyrthosiphon pisum (Bouallègue et al, 2017) Mpmar1.1 Myzus persicae (Bouallègue et al, 2017) Gbmar…”

“…In order to extract the full-length sequence of each copy, the region of each assembled copy was expanded to 1500 bp. The copies were next clustered together into lineages by iterative BLASTn, setting the similarity threshold to 75 % which was previously defined in (Bouallègue et al, 2017), retrieving a majority rule consensus sequence for each lineage during the clustering process. Singlet sequences and consensuses derived from less than 10 sequences were removed from further analysis.…”

“…Corresponding full-length and partial protein sequences of MLE transposases were identified in lineage consensuses using BLASTx search against the original transposase tBLASTn query set. The obtained protein sequences were aligned together with fulllength transposase proteins of known MLEs from the previous work (Bouallègue et al, 2017) using MAFFT v7.312 (Katoh, Standley, 2013). The resulted alignment was used to build a Bayesian phylogeny of MLEs using the MrBayes v3.2.5 (Ronquist et al, 2012).…”

“…The full-length nucleotide consensus sequences of the identified Orthoptera li neages were used in MEGABLAST search against the genomic scaffolds of 259 insect species from different orders available in the NCBI. Cases of potential horizontal transfer between the studied Orthoptera genomes and other insect taxa were considered when MEGABLAST hit showed more than 90 % of identity covering more than 90 % of the query sequences as it was proposed by several authors (Filée et al, 2015;Wallau et al, 2016;Bouallègue et al, 2017).…”