Gene similarity networks provide tools for understanding eukaryote origins and evolution

Abstract

The complexity and depth of the relationships between the three domains of life challenge the reliability of phylogenetic methods, encouraging the use of alternative analytical tools. We reconstructed a gene similarity network comprising the proteomes of 14 eukaryotes, 104 prokaryotes, 2,389 viruses and 1,044 plasmids. This network contains multiple signatures of the chimerical origin of Eukaryotes as a fusion of an archaebacterium and a eubacterium that could not have been observed using phylogenetic trees. A number of connected components (gene sets with stronger similarities than expected by chance) contain pairs of eukaryotic sequences exhibiting no direct detectable similarity. Instead, many eukaryotic sequences were indirectly connected through a "eukaryote-archaebacterium-eubacterium-eukaryote" similarity path. Furthermore, eukaryotic genes highly connected to prokaryotic genes from one domain tend not to be connected to genes from the other prokaryotic domain. Genes of archaebacterial and eubacterial ancestry tend to perform different functions and to act at different subcellular compartments, but in such an intertwined way that suggests an early rather than late integration of both gene repertoires. The archaebacterial repertoire has a similar size in all eukaryotic genomes whereas the number of eubacterium-derived genes is much more variable, suggesting a higher plasticity of this gene repertoire. Consequently, highly reduced eukaryotic genomes contain more genes of archaebacterial than eubacterial affinity. Connected components with prokaryotic and eukaryotic genes tend to include viral and plasmid genes, compatible with a role of gene mobility in the origin of Eukaryotes. Our analyses highlight the power of network approaches to study deep evolutionary events.

Fig. 1.

Selection of connected components containing a eukaryote–archaebacterium–eubacterium–eukaryote path. Eubacterial genes are represented in red, archaebacterial genes in blue, eukaryotic genes in green, plasmid genes in purple, and virus genes in black. Nodes were automatically distributed within each connected component using the edge-weighted spring-embedded visualization algorithm. This algorithm tends to place highly connected nodes and their neighbors close together. For a visualization of all CCs with an E-A-B-E topology, see SI Appendix, Fig. S1.

Fig. 2.

Connected component with eukaryotic genes likely contributed by the archaebacterial and eubacterial ancestors (A), and the likely evolutionary history of the gene family (B). In this connected component, eukaryotic genes contributed by one domain do not exhibit detectable similarity to eukaryotic genes contributed by the other domain. The shortest paths linking eukaryotic genes of eubacterial and archaebacterial affinity involve an archaebacterial and a eubacterial sequence (resulting in a eukaryote–archaebacteria–eubacteria–eukaryote path). Eubacterial genes are represented in red, archaebacterial genes in blue, eukaryotic genes in green, and plasmid genes in purple.

Fig. 3.

Correlation between the number of genes of each eukaryotic genome and the archaebacterial-to-eubacterial gene ratio.

Fig. P1.

(A) Significant similarities between eukaryotic (green), archaebacterial (blue) and eubacterial (red) sequences belonging to the same gene family. (B) Relationship between genome size and the archaebacterial-to-eubacterial gene ratio for 14 eukaryotic genomes: Bigelowiella natans, Hemiselmis andersenii, Encephalitozoon intestinalis, Plasmodium knowlesi, Saccharomyces cerevisiae, Giardia lamblia, Entamoeba histolytica, Chlorella variabilis, Naegleria gruberi, Phytophtora infestans, Trypanosoma cruzi, Homo sapiens, Tetrahymena thermophila, and Arabidopsis thaliana. Nodes in the network represent sequences, whereas links represent significant similarities. The connected component exhibits a remarkable Eukaryote-Archaebacteria-Eubacteria-Eukaryote structure, with some eukaryotic genes exhibiting similarity to eubacterial homologs and other eukaryotic genes exhibiting similarity to archaebacterial homologs. This topology is in agreement with a chimerical origin of Eukaryotes. Notice that a tree encompassing all the sequences in the gene family cannot be easily constructed, because not all sequences are significantly similar to each other at the specified thresholds.