Horsetails Are Ancient Polyploids: Evidence from Equisetum giganteum

Abstract

Horsetails represent an enigmatic clade within the land plants. Despite consisting only of one genus (Equisetum) that contains 15 species, they are thought to represent the oldest extant genus within the vascular plants dating back possibly as far as the Triassic. Horsetails have retained several ancient features and are also characterized by a particularly high chromosome count (n = 108). Whole-genome duplications (WGDs) have been uncovered in many angiosperm clades and have been associated with the success of angiosperms, both in terms of species richness and biomass dominance, but remain understudied in nonangiosperm clades. Here, we report unambiguous evidence of an ancient WGD in the fern lineage, based on sequencing and de novo assembly of an expressed gene catalog (transcriptome) from the giant horsetail (Equisetum giganteum). We demonstrate that horsetails underwent an independent paleopolyploidy during the Late Cretaceous prior to the diversification of the genus but did not experience any recent polyploidizations that could account for their high chromosome number. We also discuss the specific retention of genes following the WGD and how this may be linked to their long-term survival.

Figure 1.

The K_S-Based Age Distribution of the E. giganteum Transcriptome Provides Support for a Paleopolyploidy Event. (A) The full K_S-based age distribution of the paranome is presented. The x axis shows the synonymous distance until a K_S cutoff of 5 in bins of 0.1, while the y axis shows the number of retained duplicated paralogous gene pairs. Note that the y axis is broken because the first bin contains a very high number of recent duplicates. (B) A subset of (A) is shown with K_S values < 2 in bins of 0.04, containing the K_S values that were used for mixture modeling (excluding those with a K_S < 0.1). Note that the y axis has a different scale compared with (A) because of the different bin size. The component of the Gaussian mixture model as identified by EMMIX (McLachlan et al., 1999), which corresponds to a significant WGD feature based on SiZer analysis (Chaudhuri and Marron, 1999), is plotted on the age distribution in red, while other components are colored in black. An ancient WGD is identified with its peak centered around a K_S of 0.81 (see Table 1). (C) SiZer output. The transition from the blue to the red color at a K_S of 0.81 indicates a significant bump in the distribution.

Figure 2.

Genes Generated by the Paleopolyploidy Are Enriched for GO Terms with Roles in Plant Regulation and Development. The bar plot shows the number of genes on the x axis (note the logarithmic scale) and the GO-slim categories that were found to be overrepresented in homoeologs on the y axis. Many GO-slim categories are related to plant regulation and development, indicating that such genes have preferentially been retained after the WGD. See also Supplemental Data Set 1.

Figure 3.

Absolute Age Distribution for the Dated Peak-Based Duplicates Representing the Paleopolyploidy in E. giganteum. The solid black line represents the kernel density estimate of the dated peak-based duplicates, while the vertical dashed black line represents its peak, used as the WGD age estimate. Gray lines represent the density estimates for the 1000 bootstrap replicates, and the vertical black dotted lines represent the corresponding 90% confidence intervals for the WGD age estimate. The original raw distribution of dated peak-based duplicates is also indicated by open circles. The mode used as an estimate for the consensus WGD age is found at 92.42 mya with lower and upper 90% confidence interval boundaries at 75.16 and 112.53 mya, respectively.

Figure 4.

Pruned Tree Topology and K_S Age Distributions for the Green Land Plants. The left panel presents the topology of a few representative key plant species: C. arietinum (chickpea), Arabidopsis (thale cress), Solanum lycopersicum (tomato), L. sativa (lettuce), B. distachyon (purple false brome), O. sativa (rice), S. bicolor (sorghum), M. acuminata (banana), P. equestris (orchid), E. giganteum (giant horsetail), and P. patens (moss). Genome duplications are indicated by colored boxes. Ages for the WGDs are taken from Vanneste et al. (2014a). Red boxes indicate ages of WGDs obtained through absolute dating. Black boxes indicate ages of WGDs obtained from literature. The orange box represents the WGD for E. giganteum, as inferred in this study, while the triangle denotes the divergence of the different horsetail crown group species in the Early Cenozoic (Des Marais et al., 2003). The right panel presents K_S-based age distributions for all species included in the tree, with the exception of the distribution for E. giganteum, which is presented in more detail on Figure 1. Age distributions were constructed as described in Methods, while sequence data for these species were taken from Vanneste et al. (2014a). The x axis always shows the synonymous distance until a K_S cutoff of 3 in bins of 0.1, while the y axis shows the number of retained duplicated paralogous gene pairs. The red dotted lines indicate K_S boundaries of 0.6 and 1.1 on all individual age distributions.