The tree of life describes a tripartite cellular world

Abstract

The canonical view of a 3-domain (3D) tree of life was recently challenged by the discovery of Asgardarchaeota encoding eukaryote signature proteins (ESPs), which were treated as missing links of a 2-domain (2D) tree. Here we revisit the debate. We discuss methodological limitations of building trees with alignment-dependent approaches, which often fail to satisfactorily address the problem of ''gaps.'' In addition, most phylogenies are reconstructed unrooted, neglecting the power of direct rooting methods.Alignment-free methodologies lift most difficulties but require employing realistic evolutionary models. We argue that the discoveries of Asgards and ESPs, by themselves, do not rule out the 3D tree, which is strongly supported by comparative and evolutionary genomic analyses and vast genomic and biochemical superkingdom distinctions. Given uncertainties of retrodiction and interpretation difficulties, we conclude that the 3D view has not been falsified but instead has been strengthened by genomic analyses. In turn, the objections to the 2D model have not been lifted. The debate remains open.

The expanding Asgard archaea and their elusive relationships with Eukarya

“…In the case of actin, the proto‐eukaryotic HGT hypothesis is strongly supported by phylogenetic analyses showing that the various forms of actins present in Asgards branch between the various clades of eukaryotic actin paralogs (cytoplasmic actin and ARPs) that were already present in LECA 43,44 . Also, corroborating this hypothesis, Nasir et al 7 concluded from a comparative analysis of conserved folds that ESPs are of relatively recent origin and could correspond to late HGTs between Archaea and Eukarya.…”

“…The discovery of Asgards thus did not reduce the formidable divide existing between the archaeal and eukaryotic phenotypes. The transformation of an archaeon into a eukaryote still involves many transitions that have never been observed in nature 7,[60][61][62] . As pointed out by Nasir et al 7 , all known organisms forming intimate endosymbiotic or ectosymbiotic associations have maintained their domain identity.…”

“…The transformation of an archaeon into a eukaryote still involves many transitions that have never been observed in nature 7,[60][61][62] . As pointed out by Nasir et al 7 , all known organisms forming intimate endosymbiotic or ectosymbiotic associations have maintained their domain identity. They noticed that "even mitochondria are still recognizable as highlyreduced Bacteria despite billions of years of coexistence inside the eukaryotic cell.…”

“…Among them, multiple sequence alignments are critical, and despite undeniable improvements in their methods, the risk of misalignment increases along with the size of the data set. Misalignments of universal proteins due to domain inversion/substitution or mismatch have been detected by Nasir et al 7 in 42% of the universal markers used in the first Asgard paper. Since some of these markers are also present in the data sets of Liu et al and of Xie et al (reference 22 and personal observation), it will be important to check the new alignments for possible mistakes.…”

“…The discovery of Asgard archaea has revitalized an intense debate about the universal tree of life (uTol) topology 1 , 2 . Proponents of the two primary domain (2D) uTol propose a scenario in which Eukarya emerged within Archaea, with Eukarya being a subgroup of the Asgard superphylum 3 , 4 , whereas others still support a uTol in which all three domains (3D) are monophyletic, with Asgards nested within Archaea 5 , 6 , 7 (Figure 1 ).…”