Does endo-symbiosis explain the origin of the nucleus?

To the editor

Horiike et al.¹ give an excellent bioinformatic analysis showing relationships between yeast genes that function in the nucleus and archaeal genes, and between yeast genes that function in the cytoplasm and bacterial genes. However, their conclusion that the nucleus originated as an archaeal endosymbiont fails to explain the following features of the nucleus: the structure of the nuclear envelope; the nuclear pore complex; linear chromosomes; absence of phagocytic bacteria; the preservation of RNA-world relics in eukaryotes, and reduction of these in prokaryotes. Furthermore, their explanation contradicts the general trend of gene loss reported in parasitic, endosymbiotic and organellar genomes².

Clear parallels exist between bacterial, mitochondrial, hydrogenosomal and chloroplast membranes. No such parallel exists for the nuclear envelope where the inner and outer membranes are continuous. Likewise, the nuclear pore complex bears no resemblance to prokaryotic transmembrane pores. Hence, unlike for other organelles, ultrastructure does not favour endosymbiotic origins³.

The nucleus contains linear chromosomes with telomeres, which have not been found in archaea and arguably predate circular chromosomes. Forterre's thermoreduction hypothesis⁴, that prokaryotes arose through reductive evolution at high temperature, argues for circularization being derived; circular DNA is more thermostable than linear. Maintenance of telomeres by telomerase probably originated in the RNA world, before modern cells⁵; telomerase has an RNA core and is highly conserved among eukaryotes. Using RNA relics to root the tree of life argues that some eukaryote nuclear traits are ancestral, having been lost through reductive evolution in prokaryotes⁵; thermoreduction explains this pattern because RNA is thermolabile^4,5. If some eukaryote nuclear traits predate archaeal traits, these cannot be explained by an archaeal endosymbiont.

The conclusion of Horiike and colleagues¹ requires that the endosymbiont gained genes from its host, which is counter to known examples of endosymbiosis (including eukaryotic organelles) and intracellular parasitism, where the unifying feature is gene loss. Intracellular existence makes primary synthetic pathways redundant². Furthermore, the yeast cytoplasmic–bacterial gene relationship described¹ can be explained by Muller's ratchet — the irreversible accumulation of mutations in small asexual populations. Relocation of organellar genes to the nucleus results in escape of the effects of the ratchet^2,3 but extensive transfer from host to endosymbiont would place genes under greater mutational pressure.

Neither reductive evolution nor endosymbiosis explains nuclear origins. The former, however, explains RNA-world relics and linear chromosomes in eukaryotes, is consistent with Horiike and colleagues' results¹ and argues against an archaeal origin for the nucleus.