Beyond the Whole-Genome Duplication: Phylogenetic Evidence for an Ancient Interspecies Hybridization in the Baker's Yeast Lineage

Abstract

Whole-genome duplications have shaped the genomes of several vertebrate, plant, and fungal lineages. Earlier studies have focused on establishing when these events occurred and on elucidating their functional and evolutionary consequences, but we still lack sufficient understanding of how genome duplications first originated. We used phylogenomics to study the ancient genome duplication occurred in the yeast Saccharomyces cerevisiae lineage and found compelling evidence for the existence of a contemporaneous interspecies hybridization. We propose that the genome doubling was a direct consequence of this hybridization and that it served to provide stability to the recently formed allopolyploid. This scenario provides a mechanism for the origin of this ancient duplication and the lineage that originated from it and brings a new perspective to the interpretation of the origin and consequences of whole-genome duplications.

Fig 1. Evidence of a duplication peak pre-dating the WGD.

(A) Evolutionary relationships of the analysed species. The tree was built using a maximum likelihood approach on a concatenated alignment of 516 widespread orthologs. All branches had maximal bootstrap support (100%). The WGD and the pre-KLE (Kluyveromyces, Lachancea, and Eremothecium) branch are marked with coloured circles. Branches in the lineage leading from S. cerevisiae to the root are numbered from more ancestral (n1) to more recent (n8). (B) Duplication densities (duplications per gene per branch) calculated for each annotated branch, either using the entire set of gene trees (green dots) or only the ohnologs (yellow dots). (C) Sequence divergence between yeast sequences belonging to two populations: duplication mapped at the WGD branch (blue) and duplication mapped at the pre-KLE branch (red). Graphs represent frequencies of normalized blast scores, Kimura distances, and estimated divergence age, respectively. Normalized blast score is the result of dividing the blast score obtained when aligning the seed yeast protein to the ohnolog pair by the blast score obtained from aligning the seed yeast protein to itself. The Kimura distance between the two sequences was calculated using protdist as implemented in the phylip package after aligning the two sequences. PL-R8s [14] was used to assess the divergence times in individual trees that contained two ohnologous genes. Data on which this figure is based are provided in S1 Data.

Fig 2. Assessment of hybridization parental lineages.

(A) Schematic example of how the pre-KLE node is found in the common ancestor of the two parents, whereas the tetraploid was formed afterwards. (B) Schematic example of duplication inference at the pre-KLE position from a gene tree with genes coming from two parentals. (C) Top: maximum likelihood species tree representing the evolution of S. pastorianus. The tree was obtained using the same approach as the tree in Fig 1: 215 alignments from genes present in single copy in S. pastorianus and with orthologs in all the species considered were concatenated and analysed using maximum likelihood. Bootstrap support was maximal (100%) in all branches. The red dot represents the branch where the duplication peak can be found. Bottom: Graph representing the duplication density (duplications per gene per branch) found at three different branches in the species tree. (D) Schematic representation of the inferred positions of the putative parents, related to the main fungal groups considered. The most likely position of the two parents is marked in a black, dashed line, while a second possible position is marked in a grey, dashed line. Data on which this figure is based are provided in S1 Data.

Fig 3. Example of a phylogenetic tree with a topology that supports the hybridization scenario.

Example of a phylogenetic tree in which two copies were retained after the formation of the tetraploid. One copy shows topology A, while the second copy shows topology C. The putative duplication event is indicated in red. Support for the topology is indicated as aLRT values.

Fig 4. Topological analysis of polyploids.

(A) Phylogenetic representation of the three possible topologies regarding the placement of the post-WGD and the two parental sequences (ZT and KLE). The pie chart on the left represents the average percentage of trees found in all the S. cerevisiae reduced phylomes that supported each topology. The average was calculated from the results of the different reduced phylomes (see S9 Fig). The pie chart on the right represents the same pie chart but only using those trees within the reduced phylomes that contain S. cerevisiae proteins that have a conserved ohnolog. The average was calculated from the results of the different reduced phylomes considering only trees in which the S. cerevisiae sequence has a conserved ohnolog (see S10 Fig). Numbers below the pie charts indicate the average number of trees that passed the filters and the percentage it represents when compared to the total. (B) Same pie charts as in A but for two genomes that underwent a WGD. (C) Same pie charts as in A but for two genomes that underwent a hybridization. (D) Same pie charts as in A but for two genomes that have not been duplicated. (E) Same pie chart as in A but for the simulated phylome. Data on which this figure is based are provided in S1 Data.

Fig 5. Genome rearrangements of syntenic blocks.

Average number of genome rearrangements as calculated by MGR (Multiple Genome Rearrangements) [43] for each syntenic block inferred from Gordon et al. [11]. Orange dots represent the number of rearrangements between the S. cerevisiae block and the orthologs found in the ZT genomes, while the green dots show the same value for the comparison between the S. cerevisiae genome and the KLE genomes. Data on which this figure is based are provided in S1 Data.

Fig 6. Hybridization scenarios: Schematic representation of possible scenarios of WGD following interspecies hybridization.

Homeologous chromosomes for the two hybridizing species are coloured in yellow. Bands on the chromosomes represent genes. Pairs of genes of the same colour are paralogs. Hypothesis A shows the fusion of two diploids and the formation of the allotetraploid. Hypothesis B shows the mating of two haploids and the posterior WGD that leads to the formation of the allotetraploid. The upper right cell contains two pairs of meiotic homologous chromosomes (1a–1b and 1aʹ–1bʹ) and shows the different events that have affected the hybrid during diploidization. The cell at the bottom right represents the current yeast genome as a homozygous diploid.