Transposons passively and actively contribute to evolution of the two-speed genome of a fungal pathogen

Abstract

Genomic plasticity enables adaptation to changing environments, which is especially relevant for pathogens that engage in "arms races" with their hosts. In many pathogens, genes mediating virulence cluster in highly variable, transposon-rich, physically distinct genomic compartments. However, understanding of the evolution of these compartments, and the role of transposons therein, remains limited. Here, we show that transposons are the major driving force for adaptive genome evolution in the fungal plant pathogen Verticillium dahliae We show that highly variable lineage-specific (LS) regions evolved by genomic rearrangements that are mediated by erroneous double-strand repair, often utilizing transposons. We furthermore show that recent genetic duplications are enhanced in LS regions, against an older episode of duplication events. Finally, LS regions are enriched in active transposons, which contribute to local genome plasticity. Thus, we provide evidence for genome shaping by transposons, both in an active and passive manner, which impacts the evolution of pathogen virulence.

Figure 1.

Extensive rearrangements in Verticillium dahliae genomes are mediated by repetitive elements. (A) Syntenic regions, indicated by ribbons, between chromosomes of the two highly similar V. dahliae strains JR2 (chromosomes displayed in white) and VdLs17 (chromosomes displayed in gray) reveal multiple synteny breakpoints caused by inter-chromosomal rearrangements, highlighted by red arrows for the JR2 genome. Red bars on the chromosomes indicate lineage-specific genomic regions (LS) that lack synteny in the other strain. To facilitate visibility, some chromosomes of V. dahliae strain VdLs17 have been reversed and complemented (indicated by asterisks). (B) Detailed view of the genomic regions surrounding selected synteny breakpoints. Rearrangements over short homologous regions such as repetitive elements (black boxes) or genes (colored boxes) resulted in inter-chromosomal rearrangements (translocations). V. dahliae strain VdLs17 genes were inferred by mapping of the V. dahliae strain JR2 genes to the genome assembly of V. dahliae strain VdLs17. Dashed gray lines indicate rearrangement sites. The numbers correspond to rearrangement numbers in A and Table 1.

Figure 2.

Whole-genome alignments of Verticillium dahliae strain JR2 reveals two duplication events. (A) Circos diagram illustrating sequence alignments within V. dahliae strain JR2. Black lines indicate genomic regions with sequence similarity. The inner circle shows LS regions (red lines), the middle circle indicates clusters of LS regions, and the outer circle shows the identity between pairs of secondary alignments. Each cluster of LS region is color coded: LS1 in blue, LS2 in yellow, LS3 in magenta, and LS4 in light blue (see Supplemental Table S2). (B) K_s distribution of paralogs of which both genes are located in the core genome (red) or at least one paralog is located in an LS region (blue). (C) Duplication events are estimated by calculating the K_s value for paralogous gene pairs and displayed in the line graph. Speciation events are estimated by calculating the K_s value for orthologous gene pairs based on genes from V. dahliae strains JR2 and their respective orthologs in the other genomes and displayed in the box plot. Distributions and median divergence times between 1:1:1 orthologous pairs, displayed by box plots, were used to estimate relative speciation events.

Figure 3.

Example of gene losses after segmental duplications within the V. dahliae strain JR2 genome. Example of a segmental duplication between LS regions located on Chromosome 2. Red ribbons indicate regions of homology between the two loci. Blue arrows indicate gene models present only at one of the two loci, whereas green and red arrows indicate common genes and transposable elements, respectively.

Figure 4.

Details of the Ave1 locus in V. dahliae strain JR2. (A) Genome assemblies of race 1 and race 2 V. dahliae strains were aligned to the reference genome assembly of V. dahliae strain JR2. The red arrow indicates the location of the Ave1 gene. (B) Single nucleotide polymorphism (SNP) density (mean number of SNPs per 1 kb) over the Ave1 locus indicates depletion of SNPs in the Ave1 region when compared with neighboring regions. (C) A large genomic region on Chromosome 5 of V. dahliae strain JR2 containing the Ave1 gene is characterized by presence/absence polymorphisms between strains. Lines indicate the corrected average read depth (per 5-kb window, 500-bp slide) of paired-end reads derived from genomic sequencing of 11 V. dahliae strains. Different colors indicate distinct patterns of coverage across the Ave1 locus. Genes (Ave1 is marked in red) and transposable elements/repeats (excluding simple repeats) are indicated.

Figure 5.

Dynamics of transposable elements in the genome of Verticillium dahliae strain JR2. (A) The divergence time of transposable elements identified in the genome of V. dahliae strain JR2 (Faino et al. 2015) was estimated using the Jukes-Cantor distance calculated between repeat copies and their consensus sequence. The distributions of divergence times between transposable elements located in the core genome (red) and in the LS regions (blue) differ. Estimations of speciation events in the evolutionary history of V. dahliae are indicated by triangles based on analyses in C. (B) The distributions of divergence times between expressed/active (log₁₀[RPKM+1] >0) transposable elements (red) and nonexpressed (blue) transposable elements differ. Estimations of speciation events are indicated by triangles. (C) Speciation events are estimated by calculating the Jukes-Cantor distance for orthologous gene pairs based on genes from V. dahliae strains JR2 and their respective orthologs in the other genomes. Distributions and median divergence times between 1:1:1 orthologous pairs, displayed by box plots, were used to estimate relative speciation events.