Broadly sampled multigene trees of eukaryotes

Abstract

Background: Our understanding of the eukaryotic tree of life and the tremendous diversity of microbial eukaryotes is in flux as additional genes and diverse taxa are sampled for molecular analyses. Despite instability in many analyses, there is an increasing trend to classify eukaryotic diversity into six major supergroups: the 'Amoebozoa', 'Chromalveolata', 'Excavata', 'Opisthokonta', 'Plantae', and 'Rhizaria'. Previous molecular analyses have often suffered from either a broad taxon sampling using only single-gene data or have used multigene data with a limited sample of taxa. This study has two major aims: (1) to place taxa represented by 72 sequences, 61 of which have not been characterized previously, onto a well-sampled multigene genealogy, and (2) to evaluate the support for the six putative supergroups using two taxon-rich data sets and a variety of phylogenetic approaches.

Results: The inferred trees reveal strong support for many clades that also have defining ultrastructural or molecular characters. In contrast, we find limited to no support for most of the putative supergroups as only the 'Opisthokonta' receive strong support in our analyses. The supergroup 'Amoebozoa' has only moderate support, whereas the 'Chromalveolata', 'Excavata', 'Plantae', and 'Rhizaria' receive very limited or no support.

Conclusion: Our analytical approach substantiates the power of increased taxon sampling in placing diverse eukaryotic lineages within well-supported clades. At the same time, this study indicates that the six supergroup hypothesis of higher-level eukaryotic classification is likely premature. The use of a taxon-rich data set with 105 lineages, which still includes only a small fraction of the diversity of microbial eukaryotes, fails to resolve deeper phylogenetic relationships and reveals no support for four of the six proposed supergroups. Our analyses provide a point of departure for future taxon- and gene-rich analyses of the eukaryotic tree of life, which will be critical for resolving their phylogenetic interrelationships.

Figure 1

Overall evaluation of support for nodes in 105 and 92 taxon sets. Support for nodes using different taxon datasets, algorithms and evolutionary models. The top portion lists taxa with clear ultrastructural identities [1] whereas taxa below the line represent putative eukaryotic supergroups. As described in the text, two datasets were examined (105 and 92 taxa). R:n = RAxML analysis of 4 genes as nucleotides, with third codon positions removed; B:n = Bayesian analysis of 4 genes as nucleotides, with third codon positions removed; B:S-A = Bayesian analysis of the SSU-rDNA gene and the remaining three genes as amino acids; B:A = Bayesian analysis of three genes as amino acids; PhyML = Likelihood analysis under PHYML of three genes as amino acids. The 'Rhizaria' are indicated by grey boxes for the 92 taxon analysis as there were insufficient taxa to address this hypothesis; instead, we report support for the nested group 'Cercozoa.'

Figure 2

Eukaryotic phylogeny based on 105 taxon set. (A) Phylogeny of the major eukaryotic groups inferred from a Bayesian analysis of the combined SSU-rDNA and amino acid sequences of actin, alpha-tubulin, and beta-tubulin from 105 taxa. This is the 50% majority rule consensus tree with average branch lengths that were calculated from the Bayesian posterior tree distribution. Results of a RAxML bootstrap analysis are shown above the branches. Node thickness indicates >95% Bayesian posterior probability support, as indicated in the key. The branch lengths are proportional to the number of substitutions per site. The six eukaryotic supergroups are shown in different colors (see key). (B) Schematic phylogeny summarizing the results of Hackett et al. 2007. The thickest branches define clades that received ≥ 90% maximum likelihood bootstrap support, whereas the relatively thinner branches provide 80 – 89% bootstrap support. Stars indicate taxa that were characterized in this study (see additional file 1 for details).

Figure 3

Eukaryotic phylogeny based on 92 taxon set. Phylogeny of the major eukaryotic groups inferred from a Bayesian analysis of the combined SSU-rDNA and amino acid sequences of actin, alpha-tubulin, and beta-tubulin from 92 taxa. Other notes as in Figure 2.