A new tree-based methodological framework to infer the evolutionary history of Mesopolyploid lineages: An application to the Brassiceae tribe (Brassicaceae)

Whole genome duplication events are notably widespread in plants and this poses particular challenges for phylogenetic inference in allopolyploid lineages, i.e. lineages that result from the merging of two or more diverged genomes after interspecific hybridization. The nuclear genomes resulting from allopolyploidization contain homologous gene copies from different evolutionary origins called homoeologs, whose orthologs must be sorted out in order to reconstruct the evolutionary history of polyploid clades. In this study, we propose a methodological approach to resolve the phylogeny of allopolyploid clades focusing on mesopolyploid genomes, which experienced some level of genome reshuffling and gene fractionation across their subgenomes. To illustrate our methodological framework, we applied it to a clade belonging to the model Brassicaceae plant family, the Brassiceae tribe, that experienced a mesohexaploidy event. The dataset analysed consists of both publically available genomic sequences and new transcriptomic data according to taxa. The present methodology requires a well-annotated reference genome, for which the identification of the parental subgenome fragments has been performed (e.g. Brassica rapa and Brassica oleracea). Focusing on fully retained genes (i.e., genes for which all homoeologous gene copies inherited from the parental lineages are still present in the reference genome), the method constructs multilabelled gene trees that allow subsequent assignment of each gene copy to its diploid parental lineage. Once the orthologous copies are identified, genes from the same parental origin are concatenated and tree-building methods are used to reconstruct the species tree. This method allows resolving the phylogenetic relationships (i) among extant species within a mesopolyploid clade, (ii) among the parental lineages of a mesopolyploid lineage, and (iii) between the parental lineages and closely related extant species. We report here the first well-resolved nuclear-based phylogeny of the Brassiceae tribe.