Comparative genomic paleontology across plant kingdom reveals the dynamics of TE-driven genome evolution

Abstract

Long terminal repeat-retrotransposons (LTR-RTs) are the most abundant class of transposable elements (TEs) in plants. They strongly impact the structure, function, and evolution of their host genome, and, in particular, their role in genome size variation has been clearly established. However, the dynamics of the process through which LTR-RTs have differentially shaped plant genomes is still poorly understood because of a lack of comparative studies. Using a new robust and automated family classification procedure, we exhaustively characterized the LTR-RTs in eight plant genomes for which a high-quality sequence is available (i.e., Arabidopsis thaliana, A. lyrata, grapevine, soybean, rice, Brachypodium dystachion, sorghum, and maize). This allowed us to perform a comparative genome-wide study of the retrotranspositional landscape in these eight plant lineages from both monocots and dicots. We show that retrotransposition has recurrently occurred in all plant genomes investigated, regardless their size, and through bursts, rather than a continuous process. Moreover, in each genome, only one or few LTR-RT families have been active in the recent past, and the difference in genome size among the species studied could thus mostly be accounted for by the extent of the latest transpositional burst(s). Following these bursts, LTR-RTs are efficiently eliminated from their host genomes through recombination and deletion, but we show that the removal rate is not lineage specific. These new findings lead us to propose a new model of TE-driven genome evolution in plants.

Keywords: LTR-retrotransposons; comparative genomics; deletion; genome dynamics; plants; solo-LTR; transposable elements; transpositional burst.

Fig. 1.—

(A2 and B2) Phylogenetic tree and (A1 and B1) schematic representation and comparison of different subfamilies of LTR-RTs Uwum from maize (A) and RLC_Gm6 and RLC_Gm18 from soybean (B). The neighbor-joining tree was constructed based on the alignment of the most conserved part of the LTRs sequences. The asterisks indicate the branch with a bootstrap value higher than 90. Color coding: different color indicates different subfamilies of the same LTR-RT family. Gray areas represent conserved region between subfamilies. Scale bar indicates nucleotide sequence divergence. SubF, subfamily; GAG, group-specific antigens; IN, intergrase; RT, reverse transcriptase; kb, kilobase.

Fig. 2.—

Distribution of copy number per LTR-RTs families in eight plant genomes. x axis represent different families and y axis the copy number per family. For each species, a pie chart represents the proportion of all LTR-RT families in the genome. The copy number of the three most repeated families is given in each corresponding pie chart. Only 40 families are represented for each species.

Fig. 3.—

Distribution of estimated insertion ages (in Myr) of LTR-RTs in maize, sorghum, rice, Brachypodium for monocots (A) and soybean, grapevine, Arabidopsis Lyrata, and A. thaliana for dicots (B). Note that for monocots, the number of copies represented is limited to 500 to facilitate interspecies comparison.

Fig. 4.—

Comparison of average percentage of deletion of LTR-RTs between different plant species for four different time range (Myr). Different colors represent different age class. Brackets represent standard deviation of each age class.

Fig. 5.—

Comparative deletion pattern of LTR-RTs families within different plant species (0.5–1 Myr). The horizontal line shows the average percentage of deletion of different families belonging to six plant species (see supplementary data S3, Supplementary Material online).

Fig. 6.—

Distribution of copies length within one LTR-RT family (Gmr19 from soybean genome) (see Materials and Methods).