The rearranged mitochondrial genome of Leptopilina boulardi (Hymenoptera: Figitidae), a parasitoid wasp of Drosophila

Abstract The partial mitochondrial genome sequence of Leptopilina boulardi (Hymenoptera: Figitidae) was characterized. Illumina sequencing was used yielding 35,999,679 reads, from which 102,482 were utilized in the assembly. The length of the sequenced region of this partial mitochondrial genome is 15,417 bp, consisting of 13 protein-coding, two rRNA, and 21tRNA genes (the trnaM failed to be sequenced) and a partial A+T-rich region. All protein-coding genes start with ATN codons. Eleven protein-coding genes presented TAA stop codons, whereas ND6 and COII that presented TA, and T nucleotides, respectively. The gene pattern revealed extensive rearrangements compared to the typical pattern generally observed in insects. These rearrangements involve two protein-coding and two ribosomal genes, along with the 16 tRNA genes. This gene order is different from the pattern described for Ibalia leucospoides (Ibaliidae, Cynipoidea), suggesting that this particular gene order can be variable among Cynipoidea superfamily members. A maximum likelihood phylogenetic analysis of the main groups of Apocrita was performed using amino acid sequence of 13 protein-coding genes, showing monophyly for the Cynipoidea superfamily within the Hymenoptera phylogeny.

the nearly complete mitogenome of L. boulardi (Figitidae) is described. The main objective was to use this mitogenome to test the monophyly of the Cynipoidea superfamily and compare the synteny with I. leucospoides to verify whether the GO is a conserved character in Cynipoidea.
Specimens of L. boulardi were collected in Santa Maria,). To collect the wasps, ripe bananas were placed in field sites for 4 days to allow oviposition. The fruits were then maintained in the laboratory until emergence of the flies and their parasitoid wasps (Ortiz et al., 2015). Genomic DNA was isolated from a pool of 20 individuals using the NucleoSpin Tissue XS kit (Macherey-Nagel). The sample was sequenced using a Illumina HiSeq 2000 Next Generation Sequencing (NGS) device through the Fasteris DNA Sequencing Service (Plan-les-Ouates, Switzerland). A single-end approach with a read size of 100 bp was employed. The reads were filtered by quality to eliminate low quality reads. The FASTX-Toolkit (http://hannonlab.cshl.edu/fastx_toolkit/) was implemented within the Galaxy webserver (Giardine et al., 2005;Blankenberg et al., 2010;Goecks et al., 2010) using a quality cut-off value of 20 and a percent of bases that should possess a quality value equal to or higher than the cut-off value of 90 (Ortiz et al., 2015). A total of 35,999,679 reads were selected as of high quality.
The mitogenome was assembled using the MITObim software package (Hahn et al., 2013), with 102,482 reads being used in the assembly, this corresponding to 0.28% of total reads. The COI gene from Lepitopila victoriae (AB583620.1) was used as the seed for the assembly. MITObim uses an in silico baiting approach, which was implemented in the MIRAbait module of the MIRA assembler (v3.4.1.1) (Chevreux et al., 1999). The L. boulardi mtDNA showed a coverage of 655.24 x (on average), and the test for circularity using the software mitoMaker showed it is not circularized. The sequence was deposited in GenBank under the accession number KU665622.
The characterization and annotation of the assembled L. boulardi mitogenome was performed on the MITOS Web Server (Bernt et al., 2013a), using default parameters and UGENE software (Okonechnikov et al., 2012), respectively.
The sequenced length of the genome was 15,417 bp, containing 13 protein-coding, two rRNA, and 21 tRNA genes, as well as an A+T-rich region having 316 bp and 84.2% of A+T content ( Figure 1). We believe that a small portion of the genome failed to assemble because it was not possible to identify the trnM sequence. The mtDNA of L. boulardi is AT rich, totaling 80.3% A+T content. This high A+T composition is typical of other hymenopterans, with values ranging from 82.4% to 87.2% (Wei et al., 2010). In the Hymenoptera species here studied, these values range 612 Oliveira et al. Figure 1 -Summary of L. boulardi mitochondrial genome content and organization. ND1-6 and 4L refer to NADH dehydrogenase subunits 1-6 and 4L, COI-III refers to cytochrome c oxidase subunits 1-3, ATP6 and ATP8 refer to ATPase subunits 6 and 8, and Cyt b refers to cytochrome b; rrn refers to ribosomal RNA genes. Letters are the respective tRNAs genes. Op refers the missing region in the assemblage. Arrows indicate gene direction. from 77.7 to 87.4% (Supplementary Table S1). The plus strand comprised nine protein coding and two RNA genes, whereas the minus strand encompassed four protein coding genes and no RNA genes. All protein-coding genes initiated with an ATN codon, except for COI that started with the tetranucleotide, TTAG. Furthermore, among the protein-coding genes 11 had typical stop codons, TAA and TAG, while the ND6 and COII genes showed an incomplete stop codon T. The synteny observed in the L. boulardi mitogenome differs from that found in PanGo (Figure 2). Two rearrangements involved protein-coding genes, Nad1 is here positioned between Nad3 and Nad5, Nad2 suffered an inversion, changing from the plus strand to the minus strand, and the rRNA genes rrnL and rrnS also underwent an inversion. An extensive change of positions was involved in the tRNAs genes. A total of 16 tRNA genes changed their positions compared with PanGo (trnL2, trnD, trnG, trnA, trnS1, trnN, trnE, trnF, trnH, trnS2, trnL1,trnV, trnI, trnQ, trnW, trnC) ( Figure S1). Other five tRNA genes maintained synteny with respect to PanGo (trnK, trnR, trnT, trnP, trnY). As coverage is almost constant throughout the assembled genome, this indicates that the gene order is real. Looking at the full set of mitochondrial genes, an ample array of rearrangement was observed in L. boulardi when compared with the PanGo pattern, that is considered the ancestral gene order for insects. Figure 2 also shows a synteny comparison of the Ibalia leucospoides mitochondrial genome with PanGo. The Ibalia leucospoides mitogenome showed extensive rearrangements involving 15 tRNA genes, as well as seven protein-coding genes compared to ancestral PanGo. Fur-thermore, it contains two extra trnM genes (Mao et al., 2015). I. leucospoides and L. boulardi belong to the same superfamily, Cynipoidea, but are within different specific families. The first species is included in the Ibaliidae family and the second is a member of the Figitidae family. As can be observed in Figure 2, the rearrangements present in both genomes diverge significantly, suggesting that gene order is not conserved within the Cynipoidea and, thus, can be an important molecular marker for phylogenetic studies of this group, similar to other hymenopteran taxa (Oliveira et al., 2008;Wei et al., 2010;Mao et al., 2015).
The phylogenomic analysis positioned L. boulardi as the sister species of I. leucospoides (Figure 3). These species belong to the same superfamily, Cynipoidea, but belong to different families (Figitidae and Ibalidae respectively), and thus formed a monophyletic clade within the Apocrita phylogeny. The branches observed in the obtained tree are, in general, in accordance with the study of the evolutionary relationship among the Hymenoptera groups conducted by Sharkey et al. (2012). The major difference observed in the phylogenetic tree obtained here and that described by those authors is related to the group, Proctotrupomorpha, here represented by Cynipoidea, Proctotrupoidea, and Chalcidoidea. In the data of Sharkey et al. In Ibalia leucospoides, the mitogenome has three large rearrangements, with a repositioning of seven protein-coding and two rRNA genes. Minor rearrangements also occcurred within the 15 tRNA genes (Mao et al., 2015). In L. boulardi, there we found the presence of three large rearrangements that changed the position of two protein-coding and two rRNA genes. However, the rearrangements differ among the species.
(2012), Cynipoidea clustered with Proctotrupoidea, and this being the sister group to Chalcidoidea. In the results obtained here, Cynipoidea clustered with Chalcidoidea, and this grouping with Proctotrupoidea. The other cluster retrieved in the phylogeny consists of Aculeata (Apoidea, Sphecoidea and Vespoidea), having as sister group the Evanioidea. This clade clustered with Trigonaloidea. The relationship seen for this clade is, thus, similar to that reported by Sharkey et al. (2012). Possible reasons for the differences observed between our results and those obtained by Sharkey et al. (2012) may be related to differences in the characters used for phylogenetic analysis. Sharkey et al. (2012) used 392 morphological characters and sequence data for four loci (18S, 28S, COI and EF-1a), while we used 13 mitochondrial genes.
In summary, we consider three main contributions of this study: i) the phylogenomic analyses showed that L. boulardi and I. leucospoides form a clade representative of the Cynipoidea superfamily within the Hymenoptera phylogeny, strengthening, as observed in other studies, the monophyly of the Cynipoidea superfamily; ii) the divergent gene order observed in L. boulardi and I. leucospoides suggests that this character is not conserved in Cynipoidea; and iii) the description of the L. boulardi partial mitochondrial genome should be relevant for future phylogenomics studies in Hymenoptera, as well being useful to future population genetics studies on this species.