Paleopolyploidy in the Brassicales: analyses of the Cleome transcriptome elucidate the history of genome duplications in Arabidopsis and other Brassicales

Abstract

The analysis of the Arabidopsis genome revealed evidence of three ancient polyploidy events in the evolution of the Brassicaceae, but the exact phylogenetic placement of these events is still not resolved. The most recent event is called the At-alpha (alpha) or 3R, the intermediate event is referred to as the At-beta (beta) or 2R, and the oldest is the At-gamma (gamma) or 1R. It has recently been established that At-gamma is shared with other Rosids, including papaya (Carica), poplar (Populus), and grape (Vitis), whereas data to date suggest that At-alpha is Brassicaceae specific. To address more precisely when the At-alpha and At-beta events occurred and which plant lineages share these paleopolyploidizations, we sequenced and analyzed over 4,700 normalized expressed sequence tag sequences from the Cleomaceae, the sister family to the Brassicaceae. Analysis of these Cleome data with homologous sequences from other Rosid genomes (Arabidopsis, Carica, Gossypium, Populus, and Vitis) yielded three major findings: 1) confirmation of a Cleome-specific paleopolyploidization (Cs-alpha) that is independent of the Brassicaceae At-alpha paleopolyploidization; 2) Cleome and Arabidopsis share the At-beta duplication, which is lacking from papaya within the Brassicales; and 3) rates of molecular evolution are faster for the herbaceous annual taxa Arabidopsis and Cleome than the other predominantly woody perennial Rosid lineages. These findings contribute to our understanding of the dynamics of genome duplication and evolution within one of the most comprehensively surveyed clades of plants, the Rosids, and clarify the complex history of the At-alpha, At-beta, and At-gamma duplications of Arabidopsis.

Keywords: Arabidopsis; Brassicales; Cleome; polyploidy; transcriptome.

FIG. 1.—

Histograms of Brassicales gene duplication ages with mixture model fits. (A) A. thaliana, Brassicaceae. (B) C. papaya, Caricaceae. (C) C. spinosa, Cleomaceae. Plots of normal distributions were fitted from mixture model analyses; the orange plot represents At-α, the blue plot represents At-β, green plot represents At-γ, and the purple plot represents Cs-α.

FIG. 2.—

Phylogeny of Brassicales taxa and related Rosid outgroups displaying inferred paleopolyploidizations. Branch lengths are mean K_s values from 270 nuclear ortholog sets (supplementary table S1, Supplementary Material online). Colored dots indicate inferred paleopolyploidizations placed in relation to lineage divergence based on our rate corrections. Gray dots represent paleopolyploidizations inferred in previous analyses (Blanc and Wolfe 2004; Sterk et al. 2005).

FIG. 3.—

Phylogeny of the General Regulatory Factor (GRF or 14-3-3) gene family with nodes highlighted that support inferred paleopolyploidy events. Sequences from key taxa used to infer polypolyploidizations are labeled: Brassica (red), Arabidopsis (orange), and Cleome (pink). Nodes supporting particular paleopolyploidy events (i.e., clades containing paralogous loci) are labeled with colored circles: Br-α in red (Brassica genome triplication), At-α in orange (Brassicaceae genome duplication), Cs-α in pink (Cleome genome triplication), At-β in blue (core-Brassicales genome duplication), and the potential placement of the At-γ is shown in green (Rosid or Angiosperm genome triplication). For the analysis, EST sequences from several species (sequences labeled by genus or common names) as well as from the following genomic sequences: Os (Oryza sativa), Sb (Sorghum bicolor), Mt (Medicago truncatula), Vv (V. vinifera), Pt (P. trichocarpa), Cp (C. papaya), and At (A. thaliana). The eight members of the Arabidopsis gene family used in the analysis are additionally labeled by their gene family names: GRF1–GRF8.