Genome Size Evolution Mediated by Gypsy Retrotransposons in Brassicaceae

Abstract

The dynamic activity of transposable elements (TEs) contributes to the vast diversity of genome size and architecture among plants. Here, we examined the genomic distribution and transposition activity of long terminal repeat retrotransposons (LTR-RTs) in Arabidopsis thaliana (Ath) and three of its relatives, Arabidopsis lyrata (Aly), Eutrema salsugineum (Esa), and Schrenkiella parvula (Spa), in Brassicaceae. Our analyses revealed the distinct evolutionary dynamics of Gypsyretrotransposons, which reflects the different patterns of genome size changes of the four species over the past million years. The rate of Gypsy transposition in Aly is approximately five times more rapid than that of Ath and Esa, suggesting an expanding Aly genome. Gypsy insertions in Esa are strictly confined to pericentromeric heterochromatin and associated with dramatic centromere expansion. In contrast, Gypsy insertions in Spa have been largely suppressed over the last million years, likely as a result of a combination of an inherent molecular mechanism of preferential DNA removal and purifying selection at Gypsy elements. Additionally, species-specific clades of Gypsy elements shaped the distinct genome architectures of Aly and Esa.

Keywords: Brassicaceae; Comparative genomics; Copia retrotransposon; Genome size evolution; Gypsy retrotransposon.

Figure 1

Genomic synteny between Aly and Esa The annotations for the 13 tracks showing different genomic elements are: chromosomes of Aly and Esa represented by the outmost tracks colored in red and blue, respectively (A), hotspots of Gypsy element (B), Copia element (C), unclassified Gypsy element (D), g1-g7 families successively (E–K), gene density (L), and repeat density (M).

Figure 2

Different activity of LTR-RTs in the four genomes A. Proportions of repetitive sequences in the four genomes. B. Number of full-length Copia and Gypsy elements in the four genomes. C. Distribution of insertion times of Copia and Gypsy elements in the four species. D. Transposition rates of LTR-RTs in the four species, defined as the ratio of the net increase of LTR-RTs every 0.1 MYs relative to the total LTR-RTs over a 10-MY time-scale. E. Accumulation rates of LTR-RTs in increments of 0.1 MYs over 10 MYs. LTR-RT, long terminal repeat retrotransposon; MY, million year; MYA, million years ago.

Figure 3

Genomic distribution of LTR-RTs in the four genomes A. Distributions of Gypsy and Copia elements on chromosome 1 in the four species. The lengths of the vertical lines indicate the insertion ages: the longer the line, the older the insertion time. B. Distributions of Gypsy-to-gene and Copia-to-gene distances in the four species. C. Relationships of insertion times and insertion distances of the Gypsy or Copia elements to the nearest genes in the four species. TE-to-gene distances were log₁₀ transformed. D. Linear regression analysis of insertion times and insertion distances of Gypsy and Copia elements. The TE-to-gene distanced were log₁₀ transformed.

Figure 4

Phylogenetic classification of active LTR-RTs in the four species A. Numbers of “active”, “inactive” and “fossil” LTR-RTs in the four genomes. B. Percentages of “active”, “inactive” and “fossil” LTR-RTs in the four genomes. C. and D. Neighbor-joining trees of the 20 clades of active Copia elements (C) and the 7 clades of active Gypsy elements (D). Leaf nodes in different colors denote the species origin of individual elements. Colored branches indicate the family identity. Major branch nodes at or near the most recent common ancestors of the 20 clades of Copia elements and the 7 clades of Gypsy elements are labelled with a black-filled circle if the associated bootstrap values exceed 80%.

Figure 5

Estimation of the age of the LTR-RT clades in the four species A. Violin plots representing the age of the 27 clades of LTR-RTs in the four species. Only clades containing more than 30 members are shown. Numbers in the brackets above each violin box indicate the numbers of LTR-RT elements. B. Relationship between the mean and standard deviation (SD) values of the age of an LTR-RT clade. The solid fitting curves were generated by a LOESS model, with the grey area showing the 95% confidence interval.

Figure 6

Gypsy elements carrying chromodomains in the four species A. Phylogenetic tree of chromodomains across species. Colored nodes represent the chromodomains identified in the four Brassicaceae species. Black nodes with labels indicate previously reported sequences by Weber et al. . The two shaded areas indicate that the sequences were from either Tekay (green) or Reina (orange) clade. B. Multiple sequence alignment of the chromodomains of Gypsy elements. C. Chromodomain motif representation. The upper logo indicates the prevalence of amino acids at specific positions of known chromodomain motifs from other plant species. The lower logo shows the composition of chromodomain motifs identified in the four studied species.