Skip to main page content
U.S. flag

An official website of the United States government

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Sep 1;9(9):2336-2353.
doi: 10.1093/gbe/evx179.

Similar Evolutionary Trajectories for Retrotransposon Accumulation in Mammals

Reuben M Buckley  1 , R Daniel Kortschak  1 , Joy M Raison  1 , David L Adelson  1
Affiliations

Similar Evolutionary Trajectories for Retrotransposon Accumulation in Mammals

Reuben M Buckley et al. Genome Biol Evol. .

Abstract

The factors guiding retrotransposon insertion site preference are not well understood. Different types of retrotransposons share common replication machinery and yet occupy distinct genomic domains. Autonomous long interspersed elements accumulate in gene-poor domains and their nonautonomous short interspersed elements accumulate in gene-rich domains. To determine genomic factors that contribute to this discrepancy we analyzed the distribution of retrotransposons within the framework of chromosomal domains and regulatory elements. Using comparative genomics, we identified large-scale conserved patterns of retrotransposon accumulation across several mammalian genomes. Importantly, retrotransposons that were active after our sample-species diverged accumulated in orthologous regions. This suggested a similar evolutionary interaction between retrotransposon activity and conserved genome architecture across our species. In addition, we found that retrotransposons accumulated at regulatory element boundaries in open chromatin, where accumulation of particular retrotransposon types depended on insertion size and local regulatory element density. From our results, we propose a model where density and distribution of genes and regulatory elements canalize retrotransposon accumulation. Through conservation of synteny, gene regulation and nuclear organization, mammalian genomes with dissimilar retrotransposons follow similar evolutionary trajectories.

Keywords: genome architecture; genome evolution; transposable element.

PubMed Disclaimer

Figures

<sc>Fig</sc>. 1.—
Fig. 1.—
Overview of humanizing retrotransposon distributions. (a) Genomes are segmented and filtered according to a minimum mapping fraction threshold, removing poorly represented segments from both species. The black X shows which segments were not able to reach the minimum mapping fraction threshold. (b) Fragments of query species’ genome segments are matched to their corresponding human genome segments using genome alignments. (c) Query species genomes are humanized following equation (1). (d) Pairwise genomic correlations are measured between each humanized retrotransposon family and each human retrotransposon family. (e) The effect of humanizing on retrotransposon density distributions is measured by performing a Kolmogorov–Smirnov test between the humanized query retrotransposon density distribution and the filtered query retrotransposon density distribution. (f) The effect of filtering on retrotransposon density distributions is measured by performing a Kolmogorov–Smirnov test between the segmented human retrotransposon density distribution and the filtered human retrotransposon density distribution. (g) The pairwise correlation analysis results and the P values from the Kolmogorov–Smirnov tests are integrated into heatmaps (fig. 4 and supplementary figs. S18–S22, Supplementary Material online) that compare the genomic relationships of retrotransposons between species.
<sc>Fig</sc>. 2.—
Fig. 2.—
Similar genomic distributions of retrotransposons across mammals. Principal Component 1 and Principal Component 2 of nonhuman and nonmouse genome retrotransposon content, each vector loading has been colored according to the retrotransposon group it represents. Principal components have been renamed according to the retrotransposon group whose variance they principally account for.
<sc>Fig</sc>. 3.—
Fig. 3.—
Genomic distributions of lineage-specific retrotransposons associate with distinct genomic environments. (a) PCA of human and mouse retrotransposon content and mean genome replication timing in human HUVEC cells and mouse EpiSC-5 cells. (b) Retrotransposon density per nonoverlapping 50 kb intervals from a pooled set of ERD boundaries across all 16 human cell lines. Black dashed lines indicate 2 standard deviations from the mean (solid horizontal black line). Red line indicates mean replication timing across all samples.
<sc>Fig</sc>. 4.—
Fig. 4.—
Genome-wide spatial correlations of humanized retrotransposon families. Heatmap colors represent Pearson’s correlation coefficient for genomic distributions between humanized (a) dog and human retrotransposon families, and humanized (b) horse and human retrotransposon families. Values at the top left of each heatmap reflect the proportion of each genome analyzed after filtering at a 10% minimum mapping fraction threshold (fig. 1a). Dog and horse P values represent the effect of humanizing on filtered nonhuman retrotransposon density distributions (fig. 1e). Human P values represent the effect of filtering on the human retrotransposon density distributions (fig. 1f).
<sc>Fig</sc>. 5.—
Fig. 5.—
Retrotransposon accumulation patterns are conserved across mammals. (a) Top 10% of genome segments based on retrotransposon density of new SINEs and new L1s. (b) Top 10% of genome segments based on retrotransposon density of ancient elements and old L1s. In both a and b, segments for nonhuman genomes were ranked according to their humanized values. Large ERDs (>2 Mb) from HUVEC cells are marked in orange.
<sc>Fig</sc>. 6.—
Fig. 6.—
Retrotransposon accumulation occurs in open chromatin near regulatory regions. (a) The activity of DNase1 clusters in cERDs and cLRDs. DNase1 clusters were identified by merging DNase1 hypersensitive sites across 15 tissues. Their activity levels were measured by the number of DNase1 hypersensitive sites overlapping each DNase1 cluster. (b) Retrotransposon density of nonexonic regions and DNase1 clusters in cERDs and cLRDs. (c) Observed minus expected retrotransposon density at the boundary of DNase1 clusters corrected for interval size bias (see Methods). Expected retrotransposon density was calculated as each group's nonexonic total retrotransposon density across cERDs and cLRDs. A confidence interval of 3 standard deviations from expected retrotransposon density was also calculated, however, the level of variation was negligible.
<sc>Fig</sc>. 7.—
Fig. 7.—
Retrotransposon insertion size is inversely proportional to local regulatory element density. (a) Observed to expected ratio of retrotransposon position coverage depth measured from consensus 30 end. Expected retrotransposon position coverage depth was calculated as total retrotransposon coverage over consensus element length. We used 6 kb as the consensus new L1 length and 300 bp as the consensus Alu length. (b) New L1 and Alu position density ratio (cERDs: cLRDs). (c) Alu and (d) new L1 observed over expected retrotransposon insertion rates at DNase1 cluster boundaries in cERDs and cLRDs. Insertion rates were measured by prevalence of 30 ends and expected levels were calculated as the per Mb insertion rate across cERDs and cLRDs.
<sc>Fig</sc>. 8.—
Fig. 8.—
Retrotransposon accumulation within intergenic and intronic regions correlates with the distribution of DNase1 clusters. Density of DNase1 clusters and retrotransposons at each position upstream and downstream of genes and exons in (a) intergenic and (b) intronic regions. For DNase1 clusters, dotted lines represent exon overlapping clusters and solid lines represent clusters that do not overlap exons. For retrotransposons, solid lines represent the uncorrected retrotransposon density at exon and gene boundaries. Bar plots show expected retrotransposon density across cERDs and cLRDs. Highlighted regions outline DRBZs, regions extending from the gene or exon boundary to the point where retrotransposon levels begin to increase.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ashida H, Asai K, Hamada M.. 2012. Shape-based alignment of genomic landscapes in multi-scale resolution. Nucleic Acids Res. 40(14): 6435–6448. - PMC - PubMed
    1. Baillie JK, et al.2011. Somatic retrotransposition alters the genetic landscape of the human brain. Nature 479(7374): 534–537. - PMC - PubMed
    1. Bao W, Kojima KK, Kohany O.. 2015. Repbase update, a database of repetitive elements in eukaryotic genomes. Mobile DNA 6: 11. - PMC - PubMed
    1. Bourque G, et al.2008. Evolution of the mammalian transcription factor binding repertoire via transposable elements. Genome Res. 18(11): 1752–1762. - PMC - PubMed
    1. Capilla L, et al.2016. Mammalian comparative genomics reveals genetic and epigenetic features associated with genome reshuffling in rodentia. Genome Biol Evol. 8(12): 3703–3717. - PMC - PubMed

LinkOut - more resources