Kay, C. J., Spang, A., Szöllősi, G. J., Pisani, D., Williams, T. A. & Donoghue, P. C. J. (2026). Dated gene duplications elucidate the evolutionary assembly of eukaryotes. Nature 650(8100): 129–140. doi:10.1038/s41586-025-09808-z.
Key takeaways
From the reading, NOTE-1234.
- A timeline. The archaeal stem (nFECA to LECA) is about twice as long as the alphaproteobacterial one (mFECA to LECA): roughly 1.1 against 0.6 billion years by the authors’ medians.
- A marker for the mitochondrion. Duplications in genes of alphaproteobacterial origin begin within about 200 million years of mFECA, around 2.2 billion years ago, among them parts of the protein import system. The authors take this onset as the start of endosymbiosis. That equation is argued, not tested.
- The host came first. 85% of archaeal-origin duplications predate the marker, including ARP2/3, γ-tubulin, Golgi and ER trafficking and nuclear-targeted paralogues (THEORY-1099). Each date bounds a function from above, and many intervals overlap the marker.
- One partner, not a series. 16 of 19 duplications in non-alphaproteobacterial bacterial genes come after the marker, so the authors argue against serial endosymbiosis. What is dated is duplication, not acquisition.
- Left open. “our results do not allow us to discriminate between the origin of phagocytosis and the timing of mitochondrial endosymbiosis”. Nothing here bears on when the partners became dependent on each other.
Why this is here
A2 cites it for the timing of mitochondrial acquisition, as one of the three programmes in the live disagreement. Bears on QUESTION-277’s point that the innovations of eukaryogenesis need not share one time or one cause. The reading finds it takes a side on the question’s choice between one merger and a sequence of associations: one endosymbiotic partner, after a long phase of host elaboration, with other bacterial genes arriving by transfer.