A congruent phylogenomic signal places eukaryotes within the Archaea

Tom A Williams,Peter G Foster,Tom M W Nye,Cymon J Cox,T Martin Embley

doi:10.1098/rspb.2012.1795

Abstract

Determining the relationships among the major groups of cellular life is important for understanding the evolution of biological diversity, but is difficult given the enormous time spans involved. In the textbook ‘three domains’ tree based on informational genes, eukaryotes and Archaea share a common ancestor to the exclusion of Bacteria. However, some phylogenetic analyses of the same data have placed eukaryotes within the Archaea, as the nearest relatives of different archaeal lineages. We compared the support for these competing hypotheses using sophisticated phylogenetic methods and an improved sampling of archaeal biodiversity. We also employed both new and existing tests of phylogenetic congruence to explore the level of uncertainty and conflict in the data. Our analyses suggested that much of the observed incongruence is weakly supported or associated with poorly fitting evolutionary models. All of our phylogenetic analyses, whether on small subunit and large subunit ribosomal RNA or concatenated protein-coding genes, recovered a monophyletic group containing eukaryotes and the TACK archaeal superphylum comprising the Thaumarchaeota, Aigarchaeota, Crenarchaeota and Korarchaeota. Hence, while our results provide no support for the iconic three-domain tree of life, they are consistent with an extended eocyte hypothesis whereby vital components of the eukaryotic nuclear lineage originated from within the archaeal radiation.

Highlights

The early evolution of eukaryotes remains a fascinating and poorly understood period in the history of life
These differences have been interpreted in terms of model fit, with the NDRH þ NDCH and CAT models, for example, accounting for properties of the sequence alignment that are poorly anticipated by single-matrix models such as the general time reversible (GTR) model
With an updated sampling of archaeal diversity, we found no support for the three-domains hypothesis either from ribosomal RNA (rRNA) or protein-coding genes under any phylogenetic model

Summary

INTRODUCTION

The early evolution of eukaryotes remains a fascinating and poorly understood period in the history of life. More recent phylogenetic analyses with extended taxonomic samplings have united eukaryotes with a clade comprising the Crenarchaeota plus Thaumarchaeota [5], with the Thaumarchaeota alone [21], or as part of an unresolved eukaryotes plus TACK supergroup [19] All of these topologies are consistent with an eocyte hypothesis broadened in scope to include the newly discovered lineages. A supertree analysis of single-copy protein families found in Bacteria, Archaea including Crenarchaeota, and eukaryotes, was interpreted to reject the eocyte hypothesis in favour of alternative hypotheses whereby eukaryotes emerge from within the Euryarchaeota [22,23] Another analysis [24] has suggested that archaeal genes in eukaryotes derive from an ancient, probably extinct and in any case unknown, archaeal lineage. Trees consistent with a broadly defined eocyte hypothesis are recovered both with standard and with more complex evolutionary models and for all subsets of data

RESULTS

Pyrococcus furiosus Caldivirga maquilingensis

CONCLUSIONS