Clonality, inbreeding, and hybridization in two extremotolerant black yeasts

Abstract

Background: The great diversity of lifestyles and survival strategies observed in fungi is reflected in the many ways in which they reproduce and recombine. Although a complete absence of recombination is rare, it has been reported for some species, among them 2 extremotolerant black yeasts from Dothideomycetes: Hortaea werneckii and Aureobasidium melanogenum. Therefore, the presence of diploid strains in these species cannot be explained as the product of conventional sexual reproduction.

Results: Genome sequencing revealed that the ratio of diploid to haploid strains in both H. werneckii and A. melanogenum is about 2:1. Linkage disequilibrium between pairs of polymorphic loci and a high degree of concordance between the phylogenies of different genomic regions confirmed that both species are clonal. Heterozygosity of diploid strains is high, with several hybridizing genome pairs reaching the intergenomic distances typically seen between different fungal species. The origin of diploid strains collected worldwide can be traced to a handful of hybridization events that produced diploids, which were stable over long periods of time and distributed over large geographic areas.

Conclusions: Our results, based on the genomes of over 100 strains of 2 black yeasts, show that although they are clonal, they occasionally form stable and highly heterozygous diploid intraspecific hybrids. The mechanism of these apparently rare hybridization events, which are not followed by meiosis or haploidization, remains unknown. Both extremotolerant yeasts, H. werneckii and even more so A. melanogenum, a close relative of the intensely recombining and biotechnologically relevant Aureobasidium pullulans, provide an attractive model for studying the role of clonality and ploidy in extremotolerant fungi.

Keywords: Aureobasidium melanogenum; Hortaea werneckii; extremotolerance; halophilic fungus; halotolerance; hybridization; population genomics.

Figure 1:

Single-nucleotide polymorphism (SNP) diversity of Hortaea werneckii (A, B) and Aureobasidium melanogenum (C, D). Names of diploid strains are written in bold. (A, C) Phylogenetic networks reconstructed with a Neighbor-Net algorithm from a dissimilarity distance matrix calculated from SNP data. (B, D) Principal component analysis of SNPs. The genomes are represented by circles. The average divergence between groups of haploid genomes (dashed lines) is expressed as millions of SNPs (numbers next to dashed lines).

Figure 2:

Linkage disequilibrium (LD) decay in Hortaea werneckii (A) and Aureobasidium melanogenum (B). Squared correlation coefficient (r²) was calculated for all pairs of nonsingleton biallelic loci within the distance of 10 kbp or less and plotted as a function of the distance between the loci (blue line). The maximum observed value and its half value are marked with red horizontal dashed lines. A generalized additive model curve was fitted to the data (black line).

Figure 3:

Phylogenies of 50 longest alignable genomic regions of Hortaea werneckii (A, C) and Aureobasidium melanogenum (B, D). The alignable regions were extracted from the genomes and aligned with SibeliaZ, optimized with Gblocks, manually inspected, and used for phylogeny reconstruction with IQ-TREE and standard model selection. (A, B) Overlay of 50 phylogenies for each species. Numbers on leaf nodes represent genomes, and different sequences from the same genomes (for genomes with ploidy >1) are distinguished with letters added to the genome numbers. Vertical lines mark major clusters and the proportion of trees that supported them. (C, D) Consensus supernetworks calculated from 50 phylogenies for each species in SplitsTree. Names of diploid (and tetraploid) strains are written in bold, and tetraploid strain is additionally marked with an asterisk.

Figure 4:

Hypothesis of the genome evolution and hybridization in Hortaea werneckii (A) and Aureobasidium melanogenum (B). The hypothesis is based on the majority consensus phylogeny of 50 longest alignable regions per species. Each colored line in the central tree represents a haploid genome. The distances between the nodes of the tree correspond to the distances in an ultrametric majority consensus phylogeny. Haploid genomes are represented by a single colored line in the outermost edge of the tree, diploid genomes are represented by a double colored line, and the only tetraploid genome is represented by 4 colored lines. Around the tree, colored symbols mark the continent (inner circle) and habitat (outer circle) from which the strains have been originally isolated. Black lines and numbers in the outermost circle mark the genome/strain groups presumably originating from the same hybridization event.

Figure 5:

Aneuploid regions in Hortaea werneckii (A) and Aureobasidium melanogenum (B) genomes. Per-nucleotide sequencing depth of regions corresponding to the 50 and 35 longest contigs of H. werneckii and A. melanogenum was converted into proportion of the median sequencing depth of each individual genome. Circles represent an average of this depth in 30-kbp windows. The central horizontal line marks the median sequencing depth of the genome. Upper and lower horizontal lines mark the expected depth for haploid and triploid regions in an otherwise diploid genome. Genomes with at least 1 putatively aneuploid region are plotted in color. Other genomes are plotted in light gray. Colors of strain names in the legend mark haploid (blue) and diploid (red) genomes.