Buxus and Tetracentron genomes help resolve eudicot genome history

Abstract

Ancient whole-genome duplications (WGDs) characterize many large angiosperm lineages, including angiosperms themselves. Prominently, the core eudicot lineage accommodates 70% of all angiosperms and shares ancestral hexaploidy, termed gamma. Gamma arose via two WGDs that occurred early in eudicot history; however, the relative timing of these is unclear, largely due to the lack of high-quality genomes among early-diverging eudicots. Here, we provide complete genomes for Buxus sinica (Buxales) and Tetracentron sinense (Trochodendrales), representing the lineages most closely related to core eudicots. We show that Buxus and Tetracentron are both characterized by independent WGDs, resolve relationships among early-diverging eudicots and their respective genomes, and use the RACCROCHE pipeline to reconstruct ancestral genome structure at three key phylogenetic nodes of eudicot diversification. Our reconstructions indicate genome structure remained relatively stable during early eudicot diversification, and reject hypotheses of gamma arising via inter-lineage hybridization between ancestral eudicot lineages, involving, instead, only stem lineage core eudicot ancestors.

Fig. 1. Habit and genome assembly features of Buxus and Tetracentron.

Flowering branch of Buxus sinica (a) courtesy of PiPi; and leafy shoot of Tetracentron sinense (b) courtesy of Daderot. Hi-C contact heatmaps, intragenomic synteny with syntenic blocks colored according to the Ks scale, and Circos plots for Buxus (c) and Tetracentron (d). Concentric tracks in the Circos plots, from innermost outwards, show gene, Copia, and Gypsy retrotransposon densities per 1 Mb, and chromosomes, while ribbons connect inter-chromosomal syntenic regions. Source data underlying Fig. 1c, d are provided as a Source data file.

Fig. 2. Phylogenetic relations of Buxus and Tetracentron.

a Phylograms depicting the coalescent solution of individual Maximum Likelihood (ML) gene trees (left) and partitioned ML analysis of a supermatrix of nucleotide sequence alignments (right). Node labels indicate quartet (coalescence) and bootstrap (supermatrix) support values, and orange stars highlight the positions of Buxales and Trochodendrales in the two trees. b Cloudogram of 763 SCN gene trees illustrating discordance surrounding the deep branches of eudicot phylogeny. The most frequent trees are blue, the next most frequent red, the third most frequent green, and the rest are dark green. c Ks (left) and Similarity (right) distributions showing peaks (arrows) that stem from WGD (top) and speciation events (bottom), respectively. Source data underlying Fig. 2b, c are provided as a Source data file.

Fig. 3. Synteny and phylogenomics of eudicot subgenomes.

a Macrosyntenic alignments of early-diverged eudicots against Vitis with tracking of genomic positions by color-coded syntenic blocks representing the seven ancestral eudicot chromosomes. b Coalescence-based phylogenies of syntelogs derived from duplication events affecting the seven ancestral eudicot chromosomes. Green, red, blue, purple, yellow, aqua, and brown tracks highlight positions of ancestral chromosomes 1 through 7, respectively. Branch labels are posterior probabilities. c Schematic reconstruction of ancient eudicot WGD history. Differently color-filled circles label putative independent duplication events and stars highlight the two gamma WGDs in which the third genome is donated to the initial tetraploid (green star) from an extinct lineage to form the hexaploid (yellow star). Source data underlying Fig. 3a are provided as a Source data file.

Fig. 4. Ancestral eudicot genomes.

a Schematic phylogeny of the eudicot clade depicting extant and ancestral genomes (nodes 1–3) examined here. Images Credits: Vitis, Bob Nichols; Amaranthus, Patrick Alexander; Buxus, PiPi; Tetracentron, Daderot; Nelumbo, Engin Akyurt; Aquilegia, Ejohnsonboulder; all available in the Public Domain via Wikimedia Commons. b Protogene content of ancestral genomes. c Heatmaps of conserved synteny supporting the delimitation of seven protochromosomes in each ancestral genome (left panel), and Vitis chromosomes painted according to the protochromosomes of the three diploid ancestral genomes (right panel). Source data underlying Fig. 4c are provided as a Source data file.