Transposable element accumulation drives size differences among polymorphic Y Chromosomes in Drosophila

Abstract

Y Chromosomes of many species are gene poor and show low levels of nucleotide variation, yet they often display high amounts of structural diversity. Dobzhansky cataloged several morphologically distinct Y Chromosomes in Drosophila pseudoobscura that differ in size and shape, but the molecular causes of their large size differences are unclear. Here we use cytogenetics and long-read sequencing to study the sequence content of polymorphic Y Chromosomes in D. pseudoobscura We show that Y Chromosomes differ almost twofold in size, ranging from 30 to 60 Mb. Most of this size difference is caused by a handful of active transposable elements (TEs) that have recently expanded on the largest Y Chromosome, with different elements being responsible for Y expansion on differently sized D. pseudoobscura Y's. We show that Y Chromosomes differ in their heterochromatin enrichment and expression of Y-enriched TEs, and also influence expression of dozens of autosomal and X-linked genes. The same helitron element that showed the most drastic amplification on the largest Y in D. pseudoobscura independently amplified on a polymorphic large Y Chromosome in Drosophila affinis, suggesting that some TEs are inherently more prone to become deregulated on Y Chromosomes.

Figure 1.

D. pseudoobscura male karyotype and Y Chromosome size variation. (A) D. pseudoobscura male karyotype. Shown are Muller elements: Muller elements A and D form the X Chromosome in D. pseudoobscura. (B) Measurements of Y Chromosome arms (ratio of long arm compared with short; y-axis) from chromosome spreads for each Y-replacement line (x-axis). Colors indicate the Y Chromosomes chosen for further investigation: blue, Y_L; yellow, Y_M; green, Y_S; dark blue, lines not further characterized. Dots correspond to outlier data that fall outside of Q1–1.5 × IQR or Q3 + 1.5 × IQR. (C) Diploid genome size estimations (y-axis) of Y-replacement line males (x-axis) from flow cytometry. The rightmost sample in red is the diploid genome size of females from the reference genome strain (MV25), the strain that was used to generate Y-replacement lines though backcrossing. The numbers show the inferred genome size of three males that were used for more detailed downstream analysis (and referred to as Y_S, Y_M, and Y_L). (D) Heterochromatin estimates in three selected Y-replacement line males. Top shows staining thoracic cells with propidium iodide, and bottom shows staining whole-brain nuclei with DAPI. Leftmost red boxplot is the sequenced/backcross female for comparison.

Figure 2.

De novo assembly of three differently sized Y Chromosomes (Y_S, Y_M, Y_L). (A) Male and female coverage tracks of Y Chromosome assemblies with Log₂(female/male) coverage and repeat landscape shown beneath. One tick mark corresponds to 5 Mb, and collapsed contigs were adjusted in the plots. Landscapes of the most abundant repeats are shown separately in Supplemental Figure S8, A through C. (B, top to bottom) Karyotypes of Y_S, Y_M, and Y_L males with the arrowhead denoting the Y and the arrow denoting the X.

Figure 3.

Repeat abundance and age suggest recent TE mobilization contributes to size differences among polymorphic Y's. (A) TE abundance in Y_M and Y_L relative to Y_S from paired-end mappings to the TE library. TEs for which the absolute difference is >200 kb are shown. For details, see Methods section “Repeat content estimation by reference library.” (B) TE abundance in Y_M and Y_L chromosome assemblies relative to the Y_S assembly. TEs for which the absolute difference is >200 kb are shown. For details, see Methods section “Repeat content estimation by de novo assemblies.” (C) Divergence in TE copies from whole-genome Illumina sequencing across the Y's as calculated by dnaPipeTE. (D) Divergence in TE copies from the Y_L assembly based on the top 50 most abundant TEs (relative to Y_S), other TEs, and rolling-circle transposons as calculated by RepeatMasker. Percentages of TEs sum to 100% for each category.

Figure 4.

Decreased H3K9me3 enrichment on Y_L. (A) Heatmap with corresponding boxplot showing H3K9me3 enrichment at TEs by chromosome (autosomes/X vs. Y Chromosomes). Significance values calculated: (*) <0.05, (**) <0.01, (***) <1 × 10⁻⁵, Wilcoxon test. (B) Scatterplot of H3K9me3 enrichment at TEs by chromosome group for Y_S and Y_L (P < 0.05, two-sample t-test). The gray line indicates similar enrichment levels for both chromosome groups.

Figure 5.

Differential transposon and gene regulation on the Y Chromosome. (A) Differential TE expression between Y_S and Y_L males irrespective of age. Data represent the mean of four replicates with standard error bars (50% higher or lower expression, P < 0.05, Wald test). (B) Differential TE expression between Y_S and Y_L plotted against their differential TE abundance from Illumina mappings. (C) Lifespan curves of Y_S and Y_L males and backcross females. (D) Differential gene expression between Y_S and Y_L irrespective of age. Data represent the mean of four replicates with standard error bars (50% higher or lower expression, P < 0.05, Wald test).

Figure 6.

Y Chromosome expansions in D. affinis by the same TEs as in D. pseudoobscura. (A) Diploid genome size estimations from D. affinis Y-replacement lines. Rightmost boxplot (red) is the backcross female strain. (B) TE abundance in Y_{M, aff} and Y_{L, aff} relative to Y_{S, aff} from paired-end mappings to the TE library, made in a similar manner as in Figure 3B. TEs for which the absolute difference is >50 kb are shown. (C) Divergence in TE copies from whole-genome Illumina sequencing across the Y's estimated by dnaPipeTE, made in a similar manner as in Figure 3C. (D) Phylogenetic trees of Helitron-1_DPe (left) and Polinton-1_DPe (right) found in D. pseudoobscura (green), D. affinis (orange), and D. pseudoobscura Y_L (teal). Repbase sequences are highlighted in blue.