Papex

Open problem · Biology · Open

Origin of eukaryotes

How did archaea and bacteria combine to give rise to the first complex cells with a nucleus?

Summary

Eukaryotic cells are widely thought to have arisen from a partnership between an archaeon and one or more bacteria. What drove the merger, and the order in which mitochondria, internal membranes and the nucleus appeared, are not known.

Why it is open

Eukaryotic cells differ sharply from those of archaea and bacteria: they have a nucleus, an internal membrane system, an elaborate cytoskeleton and mitochondria, which evolved from an alphaproteobacterium [13]. No eukaryotes are known that lack mitochondria primarily rather than through later loss, so the last eukaryotic common ancestor (LECA) probably already had them [13]. The order of the changes preceding LECA remains difficult to recover from present-day organisms [5].

The problem is several questions: which archaea the host came from, which bacteria besides the ancestor of mitochondria contributed genes, what drew the partners together, and whether the mitochondrion arrived early, into a simple host, or late, into a host that was already complex, the scenarios that Alexandros Pittis and Toni Gabaldón group as mito-early and mito-late [1, 5, 15]. Asgard genomes support an archaeal ancestry, but the branching position and effects of gene transfer are disputed [3, 9, 11, 12].

Dates alone cannot settle the mitochondrial question: Tara Mahendrarajah and colleagues note that a dated tree "does not distinguish between hypotheses" of early or late acquisition, because these concern the order of features that do not correspond to nodes in the tree [10]. The sources point to what would be needed: a placement of eukaryotes among archaea that is stable to the choice of genomes and markers [9, 11, 12], a way of ordering gene acquisitions that is not confounded by differences in evolutionary rate [6, 13], and the biology of living relatives of the partners [7].

What is known

The archaeal host (2015–2020). In 2015 Anja Spang, Thijs Ettema and colleagues described Lokiarchaeota from genomes assembled from environmental DNA: a candidate archaeal phylum that grouped with eukaryotes in phylogenomic analyses and whose genomes encode an expanded set of eukaryotic signature proteins, suggestive of membrane remodelling [3]; these are proteins otherwise characteristic of eukaryotes [12]. In 2020 Hiroyuki Imachi and colleagues reported, after a decade of work, the first isolate of an Asgard archaeon, related to Lokiarchaeota [7, 11]: Candidatus Prometheoarchaeum syntrophicum, an anaerobic, extremely slow-growing cell about 550 nanometres across that degrades amino acids and depends on partner microorganisms to consume the hydrogen and formate it releases, with long branching protrusions and no visible organelle-like structures [7].

Where eukaryotes branch (2023–2026). In 2023 Laura Eme and colleagues placed eukaryotes "with high confidence" inside the Asgard archaea, as sister to the Hodarchaeales, an order within the class Heimdallarchaeia [9]. In a correction of February 2026 they reported that, after removing three partly redundant markers, the placement held when two treatments against analytical artefacts were combined [9]. In 2025 Jiawei Zhang and colleagues added 223 new Asgard genomes and concluded that eukaryotes branched before the diversification of all sampled Heimdallarchaeia; they attribute the difference probably to Njordarchaeales genomes, which they find to be chimeras of Asgard sequences and sequences from TACK archaea, Asgard's sister group [11]. In a 2025 review and reanalysis, Patrick Forterre found eukaryotes next to Hodarchaeales in only about 10% of Eme and colleagues' single-gene trees, and argued that the Asgard signal reflects long co-evolution with gene transfer in both directions between proto-eukaryotes and Asgard archaea [12].

Where the genes came from (2026). In January 2026 Victor Tobiasson, Eugene Koonin and colleagues traced core LECA genes with constrained trees and found dominant Asgard contributions to most eukaryotic systems; the alphaproteobacterial contribution was limited, mainly to energy transformation and iron–sulfur cluster assembly, and other bacterial ancestries were scattered without consistent trends [13]. In June 2026 Moisès Bernabeu, Gabaldón and colleagues reported "multiple waves" of gene transfer from diverse bacteria, with Myxococcota and Planctomycetota as consistent donors beyond the two main partners, some transfers probably preceding the mitochondrion, and large DNA viruses of the Nucleocytoviricota mediating most of the at least 4.5% of acquisitions that came through viruses [15].

Dates (2023–2025). In 2023 Mahendrarajah and colleagues' dated tree of life put the divergence of eukaryotes from their closest Asgard relatives at 2.67–2.19 billion years ago, from Alphaproteobacteria at 2.58–2.12 billion years ago, and LECA at 1.93–1.84 billion years ago [10]. Zhang and colleagues dated the Heimdallarchaeia crown group to 3.12–2.26 billion years ago [11]; this is a different node from the eukaryote–Asgard split [10, 11]. The last universal common ancestor is a much earlier ancestor [10].

Recent claims. The metabolism of the Asgard ancestor of eukaryotes is contested. Eme and colleagues inferred that the lineage leading to eukaryotes adapted to moderate temperatures and acquired the genetic potential for a heterotrophic lifestyle [9]. Zhang and colleagues inferred an anaerobic, hydrogen-dependent acetogen (an organism that makes acetate) and took this as support for the hydrogen hypothesis [11]. In February 2026 Kathryn Appler, Brett Baker and colleagues, analysing 404 Asgard genomes including 136 new Heimdallarchaeia, reported genes typical of an aerobic lifestyle, among them complex IV of the respiratory chain, and proposed that hydrogen production and aerobic respiration may both have been present in the Asgard–eukaryote ancestor [14]. Forterre, discussing the 2024 preprint of that study and citing a 2025 preprint that reports the first cultures of Hodarchaeales, writes that these are anaerobes that thrive in low-oxygen conditions only alongside aerobic organisms, so their oxygen-related proteins may serve detoxification [12].

Main approaches

Metabolic partnership. The merger began as an exchange of metabolites. In William Martin and Miklós Müller's 1998 hydrogen hypothesis, a strictly hydrogen-dependent archaeal host depended on a bacterium that released hydrogen as waste [1]. In Imachi and colleagues' entangle–engulf–endogenize model, an amino-acid-degrading host interacts with partners, including an oxygen-scavenging future mitochondrion that its protrusions entangle and engulf [7]. Strongest evidence: Forterre’s 2025 review describes the first four cultured Asgard archaea as living with partner microorganisms, mostly methanogens [12], and Zhang and colleagues reconstruct a hydrogen-dependent ancestor [11]. Main difficulty: reconstructions of the ancestor's metabolism disagree [9, 11, 14], and the first isolate releases hydrogen rather than consuming it, unlike the host of the hydrogen hypothesis [1, 7].

Mitochondria first. Nick Lane and Martin argued in 2010 that mitochondria supplied the energy needed for eukaryotic genomic complexity [2]. Strongest evidence: the absence of eukaryotes that primarily lack mitochondria, which Tobiasson and colleagues say lends credence to scenarios in which the endosymbiosis triggered the reorganisation of the cell [13]. Main difficulty: Michael Lynch and Georgi Marinov found in 2015 that a gene's cost, relative to a cell's lifetime energy budget, falls with cell volume in bacteria and eukaryotes alike, and concluded that the mitochondrion "was not a prerequisite for genome-size expansion" [4].

Mitochondria late. In 2016 Pittis and Gabaldón used scaled phylogenetic branch lengths to argue for late mitochondrial acquisition by a host with mixed ancestry [5, 6, 13]. Strongest evidence: Tobiasson and colleagues largely recovered the same pattern in a larger dataset, and their gene origins fit a host with eukaryotic features before the mitochondrion [13]. Main difficulty: Martin and colleagues argued in 2017 that the analysis overfitted a mixture model and rested on an untested molecular clock [6]; Tobiasson and colleagues found that branch lengths for respiratory complexes, presumably acquired together, vary greatly, so they are "determined primarily by factors other than the gene acquisition time" [13]. Bernabeu and colleagues' ordering of donors uses the same kind of measure [15].

Complexity in steps. Julian Vosseberg and colleagues used ancient gene duplications to order events: duplications roughly doubled the proto-eukaryotic gene set, families involved in the cytoskeleton and membrane trafficking duplicated early, and most others expanded after the mitochondrion arrived [8]. Strongest evidence: the inferred duplication order accommodates both earlier host complexity and later mitochondrial effects [8]. Main difficulty: the timing again comes from phylogenetic distances [8], open to the same objection [6, 13].

Open sub-questions

  1. Hodarchaeales or deeper. Does the Hodarchaeales sister grouping survive when Njordarchaeales genomes are removed or re-binned and marker sets are swapped [9, 11]?
  2. Descent or transfer. With taxon sampling balanced between archaea and eukaryotes, in what fraction of single-gene trees do eukaryotes group with Heimdallarchaeia, against the 10% Forterre found for Hodarchaeales [12]?
  3. Oxygen in the ancestor. Do the complex IV genes of Heimdallarchaeia support respiration, or only detoxification, when tested in culture [12, 14]?
  4. A better clock. Can any relative-timing method recover the simultaneous acquisition of respiratory complex subunits from branch lengths [6, 13]?
  5. Bacterial-type lipids. Which donors supplied eukaryotes' bacterial-type membrane lipid pathways, whose ancestry Tobiasson and colleagues found unusually mixed [13]? See Lipid divide.
  6. Viruses as carriers. Did virus-mediated transfers into the LECA lineage come from the same donors as direct transfers, and before or after the mitochondrion [15]? See Origin of viruses.

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. Martin, W. & Müller, M. (1998). The hydrogen hypothesis for the first eukaryote. Nature 392(6671), 37–41. 10.1038/32096 · Checked 15 Oct 2026
  2. Lane, N. & Martin, W. (2010). The energetics of genome complexity. Nature 467(7318), 929–934. 10.1038/nature09486 · Checked 15 Oct 2026
  3. Spang, A., Saw, J. H., Jørgensen, S. L., Zaremba-Niedzwiedzka, K., Martijn, J., Lind, A. E. et al. (2015). Complex archaea that bridge the gap between prokaryotes and eukaryotes. Nature 521(7551), 173–179. 10.1038/nature14447 · Checked 15 Oct 2026
  4. Lynch, M. & Marinov, G. K. (2015). The bioenergetic costs of a gene. Proceedings of the National Academy of Sciences 112(51), 15690–15695. 10.1073/pnas.1514974112 · Checked 15 Oct 2026
  5. Pittis, A. A. & Gabaldón, T. (2016). Late acquisition of mitochondria by a host with chimaeric prokaryotic ancestry. Nature 531(7592), 101–104. 10.1038/nature16941 · Checked 15 Oct 2026
  6. Martin, W. F., Roettger, M., Ku, C., Garg, S. G., Nelson-Sathi, S. & Landan, G. (2017). Late mitochondrial origin is an artifact. Genome Biology and Evolution 9(2), 373–379. 10.1093/gbe/evx027 · Checked 15 Oct 2026
  7. Imachi, H., Nobu, M. K., Nakahara, N., Morono, Y., Ogawara, M., Takaki, Y. et al. (2020). Isolation of an archaeon at the prokaryote–eukaryote interface. Nature 577(7791), 519–525. 10.1038/s41586-019-1916-6 · Checked 15 Oct 2026
  8. Vosseberg, J., van Hooff, J. J. E., Marcet-Houben, M., van Vlimmeren, A., van Wijk, L. M., Gabaldón, T. & Snel, B. (2021). Timing the origin of eukaryotic cellular complexity with ancient duplications. Nature Ecology & Evolution 5(1), 92–100 (published online 2020). 10.1038/s41559-020-01320-z · Checked 15 Oct 2026
  9. Eme, L., Tamarit, D., Caceres, E. F., Stairs, C. W., De Anda, V., Schön, M. E. et al. (2023). Inference and reconstruction of the heimdallarchaeial ancestry of eukaryotes. Nature 618(7967), 992–999; author correction 651(8104), E1 (2026), DOI 10.1038/s41586-026-10115-4. 10.1038/s41586-023-06186-2 · Checked 15 Oct 2026
  10. Mahendrarajah, T. A., Moody, E. R. R., Schrempf, D., Szánthó, L. L., Dombrowski, N., Davín, A. A. et al. (2023). ATP synthase evolution on a cross-braced dated tree of life. Nature Communications 14, 7456. 10.1038/s41467-023-42924-w · Checked 15 Oct 2026
  11. Zhang, J., Feng, X., Li, M., Liu, Y., Liu, M., Hou, L.-J. & Dong, H.-P. (2025). Deep origin of eukaryotes outside Heimdallarchaeia within Asgardarchaeota. Nature 642(8069), 990–998. 10.1038/s41586-025-08955-7 · Checked 15 Oct 2026
  12. Forterre, P. (2025). Extensive lateral gene transfer between proto-eukaryotes and Heimdallarchaeia suggests their close association during eukaryogenesis. mLife 4(4), 345–362. 10.1002/mlf2.70030 · Checked 15 Oct 2026
  13. Tobiasson, V., Luo, J., Wolf, Y. I. & Koonin, E. V. (2026). Dominant contribution of Asgard archaea to eukaryogenesis. Nature 650(8100), 141–149. 10.1038/s41586-025-09960-6 · Checked 15 Oct 2026
  14. Appler, K. E., Lingford, J. P., Gong, X., Panagiotou, K., Leão, P., Langwig, M. V., Greening, C., Ettema, T. J. G., De Anda, V. & Baker, B. J. (2026). Oxygen metabolism in descendants of the archaeal-eukaryotic ancestor. Nature 652(8109), 405–415. 10.1038/s41586-026-10128-z · Checked 15 Oct 2026
  15. Bernabeu, M., Manzano-Morales, S., Marcet-Houben, M. & Gabaldón, T. (2026). Gene ancestries reveal diverse microbial associations during eukaryogenesis. Nature 656(8127), 399–405. 10.1038/s41586-026-10639-9 · Checked 15 Oct 2026

Page reviewed 15 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in biology. Checked 08 Oct 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Life sciences