Recent Sex Chromosome Divergence despite Ancient Dioecy in the Willow Salix viminalis

Abstract

Sex chromosomes can evolve when recombination is halted between a pair of chromosomes, and this can lead to degeneration of the sex-limited chromosome. In the early stages of differentiation sex chromosomes are homomorphic, and even though homomorphic sex chromosomes are very common throughout animals and plants, we know little about the evolutionary forces shaping these types of sex chromosomes. We used DNA- and RNA-Seq data from females and males to explore the sex chromosomes in the female heterogametic willow, Salix viminalis, a species with ancient dioecy but with homomorphic sex chromosomes. We detected no major sex differences in read coverage in the sex determination (SD) region, indicating that the W region has not significantly degenerated. However, single nucleotide polymorphism densities in the SD region are higher in females compared with males, indicating very recent recombination suppression, followed by the accumulation of sex-specific single nucleotide polymorphisms. Interestingly, we identified two female-specific scaffolds that likely represent W-chromosome-specific sequence. We show that genes located in the SD region display a mild excess of male-biased expression in sex-specific tissue, and we use allele-specific gene expression analysis to show that this is the result of masculinization of expression on the Z chromosome rather than degeneration of female-expression on the W chromosome. Together, our results demonstrate that insertion of small DNA fragments and accumulation of sex-biased gene expression can occur before the detectable decay of the sex-limited chromosome.

Keywords: allele-specific expression; masculinization; sex chromosomes; sex-biased gene expression.

Fig. 1

Phylogeny showing the relative divergence times of the evolution of dioecy and the sex chromosomes as well as the lineage split between poplars and willows.

Fig. 2

Genome-wide analysis of patterns of sequence divergence. The innermost circle (A) displays the association (as Bonferroni corrected P-values of a Fisher’s exact test) of genetic SNP markers with the sex of the individual based on a bi-parental population of 271 offspring individuals (Pucholt et al. 2017). The second circle (B) displays the relative Log₂ female:male genome coverage, whereas the outermost circle (C) displays the Log₂ absolute female (red) and male (blue) SNP density. The region on chromosome 15 with genetic markers strongly associated with phenotypic sex is highlighted throughout all circles. The lines represent a moving average over a window size of 25 scaffolds/markers.

Fig. 3

Patterns of sex chromosome differentiation on chromosome 15. (A) Log₂ normalized per base coverage of all individuals that were used in the study (red: females, blue: males). (B) Log₁₀ female:male SNP density. (C) Log₂ female:male FPKM expression values in catkins. (D) Log₂ female:male FPKM expression values in leaves. Shaded areas represent the bootstrap based 95% confidence interval, horizontal dashed lines represent the bootstrap median value whereas solid lines represent a moving average over a window size of 25 scaffolds/genes for the metrics in question. Grey vertical dashed lines represent the border of the SD region as defined by the SNP analysis.

Fig. 4

Heatmap of FPKM values of genes that show sex biased expression in at least one tissue (q < 0.05; log₂ FC > 1) and that are expressed with an average FPKM value of > 1. Samples were clustered using hierarchical cluster analysis (top). Red: female tissue, Blue: male tissue.