Radical views of the tree of life.

The tripartite division of cellular organisms into three domains – the Bacteria, the Archaea and the Eukarya – has become so ingrained that most of us have all but forgotten (or did not even know) that the ancestry and relationships among these ancient lineages are far from settled. The most widely accepted view, which emerged primarily from analyses of anciently duplicated genes (Gogarten et al., 1989; Iwabe et al., 1989), places Bacteria as the basal lineage and distinct from the common ancestor of archaea and eukaryotes. Accordingly, these studies frame the Archaea, despite their appellation, as archaic but not primordial, and the three domains are each viewed as separate lineages, with archaea and eukaryotes as sister taxa. Early on, Lake (1988) developed an algorithm that positioned a clade of thermophilic, sulfur-metabolizing archaea (the Crenarchaeaota or ‘eocytes’) as more closely related to eukaryotes than to other archaea, and the archael Holobacteria as more closely related to Eubacteria. This upended the view that Bacteria, Archaea and Eukarya were separate divisions; however, subsequent analyses of smalland large-subunit rRNA sequences in more conventional phylogenetic frameworks upheld the Archaea as a cohesive group (Gouy and Li, 1989). And although there has since been some tantalizing evidence that eukaryotes emerged from within the Archaea, particularly from analyses of elongation factor Tu sequences (Rivera and Lake, 1992; Baldauf et al., 1996), the ‘eocyte’ hypothesis was all but forgotten . . . until last year. The incentive to re-examine the deepest branches in the Tree of Life was no doubt impelled by the current number and diversity of genome sequences. The addition of genes and taxa has helped resolve some of the thorniest phylogenetic issues and, in this case, can help establish the degree of support for the 3-domain tree, the eocyte tree, or any of the other hypotheses that have been posited over the years (Embley and Martin, 2006; Poole and Penny, 2007). To examine the origin of the eukaryotes, Pisani et al. (2007) applied a supertree approach to each of three large overlapping data sets containing either 0, 8 or 21 eukaryotic genomes, in order to monitor the effects of taxon sampling on the resulting branching orders. Supertrees are phylogenies that are assembled by merging sets of more limited trees, each based on a gene that is not necessarily present in all taxa. Their results resolved bacteria and archaea as separate groups, but favour a chimeric origin of eukaryotes, as the majority of eukaryotic genes with prokaryotic homologues derive from multiple symbioses. These findings help explain how different genes originated and why particular genes show distinct phylogenetic histories, but they do not ascertain whether the cellular lineage leading to eukaryotes – i.e. the lineage that suffered these varied symbioses – preceded, or arose from within, the Archaea. In a parallel approach, Yutin et al. (2008) sifted through nearly 1000 sets of putative orthologues present in all three domains and identified more than a hundred genes that provided sufficient signal to test the relationship of eukaryotes to the two major phyla within the Archaea. Although several individual genes showed closer alliance to the Crenarchaeota, their analyses, for the most part, placed eukaryotes outside the extant archaeal diversity, lending support to the 3-domain tree. Numerous factors, aside from the presence of genes with different evolutionary histories, can complicate the accurate reconstruction of deep evolutionary relationships. Individual genes, or entire genomes, display different evolutionary rates due to selection or to the underlying pattern of mutations. For example, divergent organisms with AT-rich genomes often encode proteins that are enriched in amino acids specified by AT-rich codons, such as leucine, isoleucine and lysine, and such proteins are prone to converge instead of diverge with time, thereby distorting their actual relationships. This is rarely a problem when analysing closely related organisms, which tend to have similar rates and patterns of substitutions, but one might expect its effects to increase with organismal diversity and therefore, to be most pronounced when encompassing the diversity of all life forms. *For correspondence. E-mail [email protected]; Tel. 520626-8355; Fax 520-621-3709. Environmental Microbiology (2009) doi:10.1111/j.1462-2920.2009.01895.x