The nucleolar proteome and the (endosymbiotic) origin of the nucleus.

Sir, Nucleoli are essential regions of the eukaryotic nuclei where the ribosomes are assembled. The human nucleolar proteome has been studied in the last two years, leading to the identification of several hundreds of proteins. In a recent Bioessays’ article, Staub et al. characterise a large set of conserved domains in these nucleolar proteins. The search of these domains in other organisms revealed a complex pattern. Many exist in archaea, arguing for an archaeal origin of the nucleolar ‘‘core’’ machinery. A substantial bacterial contribution is also identified, while several domains are exclusively eukaryotic. These results support a chimeric origin of the nucleolus, with an ancestral core of archaeal origin supplemented with certain bacterial proteins and several proteins invented during the eukaryotic evolution. This mixed heritage is compatible with the diverse hypotheses that propose the amalgamation of archaea and bacteria to give rise to the first eukaryotes. However, these hypotheses differ in the mechanism explaining the merging of the archaeal and bacterial partners. Some propose an unspecified fusion or the engulfment of an archaeon by a bacterium. The most elaborate invoke symbioses between archaea and bacteria. Another fundamental difference concerns the origin of the nucleus: is it a remnant of endosymbiotic archaea or just an idiosyncratic eukaryotic invention? Staub et al. try to answer this question using the nucleolar proteome information. They conclude that the ribosomes and the components involved in the maturation and assembly of their constituents are of archaeal origin, which excludes a bacterial endosymbiotic origin of the nucleus. However, all modern hypotheses for the origin of the nucleus involve symbiotic or endosymbiotic archaea, not endosymbiotic bacteria. Hence, despite their claim, Staub et al.’s results are fully compatible with these hypotheses. They are reluctant to accept an origin of the nucleus from an endosymbiotic archaeal ancestor, but they support their view with arguments other than the nucleolar protein domain analysis. Are their arguments sufficiently sound? The first concerns the co-existence of three different genomes and protein synthesis machineries in early eukaryotes and the problems eventually derived from the interferences between them. However, this problem is not exclusive for the hypotheses proposing an endosymbiotic origin of the nucleus. Indeed, the study of the nucleolar proteome supports a chimeric origin of eukaryotes involving at least an archaeon andabacterium, that is, twogenomesand twodifferent protein synthesis machineries. This is not necessarily problematic, as revealed by the evolutionary history of eukaryotes. All contemporary eukaryotes have or have hadmitochondria with their owngenomeandprotein synthesismachinery.Moreover, photosynthetic eukaryotes also contain chloroplasts, derived from cyanobacterial endosymbionts with a genome and protein synthesis machinery. In the most complex cases, cryptophyte and chlorarachniophyte algae, chloroplasts derive from endosymbiotic eukaryotic algae, which possess a reduced but functional nucleus (the ‘‘nucleomorph’’). In these algae, four different genomes and protein synthesis machineries co-exist: nuclear-, mitochondrial-, chloroplastand nucleomorph-encoded. This extreme complexity (CavalierSmith called these organisms ‘‘meta-algae’’) eloquently demonstrates that the co-existence of different genomes and protein synthesismachineries is possible and canhave agreat evolutionary success. The second argument evoked by Staub et al. is that the available theories do not explain the energetic advantages for anendosymbiotic origin of the nucleus.Several hypothesesdo not discuss any selective advantage (energetic or other) for the establishment of the endosymbiosis. However, we have already advanced a model, the ‘‘syntrophy hypothesis’’, starting with the establishment of a metabolic symbiosis (‘‘syntrophy’’ meaning ‘‘eating together’’) sustained by interspecies hydrogen transfer between facultative sulfatereducing proteobacteria and methanogenic archaea, a frequent symbiosis in anoxic environments. For us, the ancestral nucleus (derived from the methanogen) was originally a metabolic compartment, which lost methanogenesis during its evolution. The hypothesis preferred by Staub et al. to explain eukaryogenesis, the ‘‘hydrogen hypothesis’’, is also based on a hydrogen-transfer syntrophy between