Fig 1.
Karyotype and genome of D. busckii.
A. Karyotype of D. busckii. The ancestral karyotype of Drosophila species consists of six chromosomal arms termed ‘Muller’s elements’. Muller-A element is the ancestral sex chromosome shared by all Drosophila species, and Muller-F is the dot chromosome. In D. busckii, the dot chromosome pair has fused to the ancestral X and Y chromosome and became the neo-X and neo-Y chromosome. B. Coverage and heterozygosity patterns of the D. busckii genome. For each chromosome of D. busckii (named after its homologous chromosome in D. melanogaster), we show mapped read coverage in male (blue) and female (red), and single nucleotide polymorphism (SNP) density (sites/kb) within 5kb non-overlapping windows along the chromosome. chrX shows a reduction of male coverage because the ancestral Y chromosome is completely degenerated in male D. busckii. The neo-sex (dot) chromosome shows similar coverage to autosomes, indicating that most neo-Y reads can still be mapped to the neo-X. The increase of male SNP density on the neo-sex chromosome indicates sequence divergence between the neo-X and neo-Y alleles.
Fig 2.
We used 6189 orthologous gene pairs from 9 Diptera species and constructed a phylogenomic tree. Although the bootstrap value at the ancestral node of D. busckii and D. albomicans is low, D. busckii is grouped with high confidence within the Drosophila subgroup instead of as a sister group to all Drosophila species, as previously hypothesized [16,30].
Fig 3.
Functional degeneration of neo-Y genes.
A. Composition of neo-Y linked genes. We show numbers of putative functional genes (‘Intact’), genes with premature stop codons (‘PTC’) and/or frameshift (‘Shift’) mutations on the neo-Y. B. Boxplots of gene expression level on each chromosome. We divide neo-sex linked genes according to the functional status of the neo-Y genes: functional (func) neo-Y genes, and their diploid (dpd) neo-X homologs; non-functional (psd) neo-Y genes and their hemizygous (hmz) neo-X homologs. The former group of neo-sex linked genes shows a higher expression level than the latter. C. Allelic expression bias of neo-sex linked genes in male adults. Shown are the log ratios of neo-X expression vs. neo-Y expression along the neo-sex chromosome, with putatively functional neo-Y genes in red and pseudogenes in green. We also plot the loess smooth lines separately for the two categories of genes, in order to show the local variation of the log ratio along the chromosome position. Any genes above 0 have higher neo-X expression relative to the neo-Y. D. Sex-bias expression of neo-sex linked genes. We show the expression difference between sexes for neo-sex linked genes, with neo-X/Y gene expression level combined in male, and only neo-X gene expression in female. E. Correlation between relative neo-sex allelic expression vs. sex-biased expression and relative neo-X expression. Shown are the ratios of neo-X vs. neo-Y expression level for neo-sex linked genes, vs. their expression ratio between sexes (in blue), and the ratio of neo-X expression in male vs. that in female (in orange), as well as their linear regression lines. F. Density plot of the ratios of male neo-sex alleles (neo-X in orange, neo-Y in blue) vs. female expression levels. Assuming an equal expression level between sexes, we expect the distribution of relative neo-X alleles’ expression to be around half of the female expression level (dashed line).
Fig 4.
Heterochromatin evolution on the D. busckii neo-Y.
Shown is the normalized log2 enrichment level of H3K9me2 (A-C) or H3K9me3 (D-F) over genes on different chromosomes. A. Enrichment level of H3K9me2 at silent neo-Y linked genes (in blue) is significantly higher than that of the neo-X (in orange, Wilcoxon test significance level, P<0.001:***), chrX (red) and autosomes (green). B-C. ‘Metagene’ profiles for H3K9me2 enrichment. Metagene profiles scale all genes of the same chromosome into the same number of bins for calculating average enrichment frequency along the gene body (Methods and Materials). We divide genes into actively transcribed (B.) and silent (C.) genes based on the gene expression levels of neo-Y alleles. We also include the up- and down- stream 1.5kb flanking regions. D. Enrichment level of H3K9me3. E-F. Metagene profiles for H3K9me3 enrichment at active (E.) and silent (F.) genes.
Fig 5.
Dosage compensation in male D. busckii.
A. Immunostaining of male and female D. busckii polytene chromosomes with MSL-2 antibody. The neo-X / X chromosome is marked with an arrow, and the male X shows binding of MSL-2 protein. B. Comparison of normalized log2 enrichment level of H4K16ac across genes on different chromosomes. Enrichment level of H4K16ac on X-linked genes (red) is significantly higher (Wilcoxon test, P<0.001) than on any other chromosomes, while neo-sex linked genes show a significantly lower enrichment level than autosomes (green), and there is no significant difference between the neo-X (orange) and neo-Y (blue) alleles. C. Enrichment level of H4K16ac is strongly correlated between orthologous genes of D. melanogaster and D. busckii. Genes are color-coded according to chromosomal location. D-E. Metagene profiles of H4K16ac over active (D.) and silent (E.) genes. For neo-sex genes, we defined the expression status by the expression level of neo-X alleles. Note that H4K16ac is significantly more enriched at active X-linked genes, and shows a characteristic 3’ binding bias.