A Survey of Transposon Landscapes in the Putative Ancient Asexual Ostracod Darwinula stevensoni

Abstract

How asexual reproduction shapes transposable element (TE) content and diversity in eukaryotic genomes remains debated. We performed an initial survey of TE load and diversity in the putative ancient asexual ostracod Darwinula stevensoni. We examined long contiguous stretches of DNA in clones from a genomic fosmid library, totaling about 2.5 Mb, and supplemented these data with results on TE abundance and diversity from an Illumina draft genome. In contrast to other TE studies in putatively ancient asexuals, which revealed relatively low TE content, we found that at least 19% of the fosmid dataset and 26% of the genome assembly corresponded to known transposons. We observed a high diversity of transposon families, including LINE, gypsy, PLE, mariner/Tc, hAT, CMC, Sola2, Ginger, Merlin, Harbinger, MITEs and helitrons, with the prevalence of DNA transposons. The predominantly low levels of sequence diversity indicate that many TEs are or have recently been active. In the fosmid data, no correlation was found between telomeric repeats and non-LTR retrotransposons, which are present near telomeres in other taxa. Most TEs in the fosmid data were located outside of introns and almost none were found in exons. We also report an N-terminal Myb/SANT-like DNA-binding domain in site-specific R4/Dong non-LTR retrotransposons. Although initial results on transposable loads need to be verified with high quality draft genomes, this study provides important first insights into TE dynamics in putative ancient asexual ostracods.

Keywords: DNA transposons; asexuality; crustaceans; retrotransposons; transposable elements.

Figure 1

Darwinula stevensoni. (A) A sample of multiple individuals of Darwinula stevensoni. Taken by Jeroen Vendericks. (B) Lateral view of the carapace of an individual D. stevensoni. This picture was taken with the polychromatic polarization microscope [40] with a 4× objective lens and a DP73 camera. The total length of the animals is around 800 μm. In the left corner, an embryo in the brooding pouch is visible.

Figure 2

Box plots of intron and exon abundance. The frequency of introns and exons were calculated as % of total fosmid lengths. Boxes contain the interquartile range from the 25th to the 75th percentile, the horizontal line indicates the median, and vertical lines indicate minimum and maximum distributions of the data. Outliers are shown by dots.

Figure 3

Landscape divergence plots (left hand side) and genome occupancy by known transposable elements (TEs) (right hand side) in D. stevensoni. Divergences were calculated as Kimura substitution levels with adjusted CpG. Genome fraction of TEs was calculated after merging RepeatMasker and Censor outputs. (A, top) TEs in fosmid DNA sequence data. Genome fraction of TEs was calculated after merging RepeatMasker and Censor outputs. (B, bottom) TEs in the preliminary draft genome assembly [45]. The plot was constructed using the REPET library obtained with the Illumina assembly. The pie chart shows genome occupancy for TE categories occupying more than 1% of the genome. Concerning less frequent TEs in the draft genome, we found 0.2% of Penelope, 0.2% of Dong-R4 and 0.8% Jockey elements, 1.1% of gypsy and 0.6% of other LTR elements.

Figure 3

Landscape divergence plots (left hand side) and genome occupancy by known transposable elements (TEs) (right hand side) in D. stevensoni. Divergences were calculated as Kimura substitution levels with adjusted CpG. Genome fraction of TEs was calculated after merging RepeatMasker and Censor outputs. (A, top) TEs in fosmid DNA sequence data. Genome fraction of TEs was calculated after merging RepeatMasker and Censor outputs. (B, bottom) TEs in the preliminary draft genome assembly [45]. The plot was constructed using the REPET library obtained with the Illumina assembly. The pie chart shows genome occupancy for TE categories occupying more than 1% of the genome. Concerning less frequent TEs in the draft genome, we found 0.2% of Penelope, 0.2% of Dong-R4 and 0.8% Jockey elements, 1.1% of gypsy and 0.6% of other LTR elements.

Figure 4

Box plots of TE abundance and overlap of TEs with exons and introns in fosmid data. TEs were identified with Censor from translated fosmid DNA sequences. All frequencies were calculated as % of total fosmid lengths. Boxes contain the interquartile range from the 25th to the 75th percentile, the horizontal line indicates the median, and vertical lines indicate minimum and maximum distributions of the data. Outliers are shown by dots.

Figure 5

Sequence features of representative fosmid contigs. Features of contigs (marked outside in yellow) are plotted on tracks and their identities indicated as colored boxes shown in the figure legend. Shades of red inside the circles represent % identity from BLAST searches (dark color 100%, intermediate 90–99%, and light color 80–89%), indicating recent transposition (see also Table S4). Large numbers on the outside indicate contigs; smaller numbers, the position in kb. Translated TEs were identified with Censor; transcripts, exons and introns were predicted with Augustus; all features visualized with Circos. For further details on contigs, see Table S3A. (A) Contigs Ds_ctg224, Ds_ctg295, Ds_ctg101, Ds_ctg183 and Ds_ctg179 containing mariner-1 DNA TEs. (B) Contigs Ds_ctg 45, Ds_ctg 82, Ds_ctg 89, Ds_ctg 282, Ds_ctg 333 and Ds_ctg 336 with various TEs including DNA (mariner), LINE (PLE and CR1), and LTR (gypsy). (C,D) Contigs Ds_ctg22, Ds_ctg31, Ds_ctg37, Ds_ctg79, Ds_ctg83, Ds_ctg100 and Ds_ctg128 and Ds_ctg133, Ds_ctg135, Ds_ctg155, Ds_ctg247, Ds_ctg257, Ds_ctg299 and Ds_ctg315, respectively, displaying abundant mariner-2 TEs with high sequence similarities. (E,F) Contigs containing LINE-like TEs, including contigs Ds_ctg9, Ds_ctg94, Ds_ctg178, Ds_ctg185 and Ds_ctg221 containing RTE (E) and contigs Ds_ctg37, Ds_ctg49, Ds_ctg74, Ds_ctg83, Ds_ctg100, Ds_ctg141, Ds_ctg247 and Ds_ctg339 containing CR1 (F) TEs.

Figure 6

Regions of homology in R4/Dong (A) and Gypsy (B) retrotransposons of D. stevensoni. Multiple sequence alignments were visualized with Jalview [59]. (A) The alignment includes Dong1 and Dong2 (D. stevensoni) and related elements from cnidarians (Exaiptasia diaphana, Acropora millepora, Acropora digitata), mollusks (Biomphalaria glabrata), insects (Bombyx mori, Anopheles gambiae, Spodoptera frugiperda, Pediculus humanus corporis, Sitophilus oryzae), fish (Oryzias latipes, Takifugu rubripes) and nematodes (Ancylostoma caninum, Oesophagostomum dentatum, Teladorsagia circumcincta) with the corresponding accession numbers, followed by a representative selection of Myb-like domains from the PF00249 seed alignment (25 out of 147). (B) Structure-based alignment of D. stevensoni Gypsy1 C-terminus (amino acids 1322-1479 out of 1486) and the central helix domain of the spike (sigma) proteins from reoviruses and coronaviruses identified by HHpred, with the corresponding PDB accession numbers.