What was the ancestor of everything alive?

In 1977 Carl Woese and George Fox used ribosomal RNA to show that life divides into three primary lines of descent, revealing the archaea as distinct from bacteria. That method (comparing sequences that all organisms share) is still how the tree of life is drawn.

Modern versions of the tree are built from thousands of genomes, many from organisms that have never been grown in a laboratory and are known only from environmental DNA. The tree is far bushier than it looked fifty years ago, and most of it is microbial.

About thirty genes are present in all genomes, most of them for ribosomal proteins. They tell us that LUCA had a ribosome and the genetic code, which the universality of the code had already made plain. They say almost nothing about how or where LUCA lived, because the traits that would answer that are the ones microbes differ in.

One influential reconstruction searched 6.1 million genes from sequenced prokaryotes and found 355 protein families that survive a strict test for inheritance from LUCA rather than later transfer. The picture they produce is specific: an anaerobe, dependent on hydrogen, carbon dioxide and metals, living somewhere hot and without oxygen.

Genes move sideways between microorganisms. This horizontal gene transfer is common, ancient, and it blurs ancestry: a gene present in both bacteria and archaea today may have been inherited from LUCA, or swapped between lineages a billion years later. Of the gene clusters found in both domains, 97 per cent show signs of having crossed between them at some point.

Different filtering choices produce different LUCAs. Allow gene loss freely and LUCA becomes the most versatile organism that ever lived; assume many later gains and it has few genes. A 2024 reconstruction gives it a genome of at least 2.5 megabases and around 2,600 proteins, comparable to a modern prokaryote. The vent-dwelling anaerobe is a serious result, not a consensus one. The 2024 reconstruction agrees that LUCA was an anaerobe that fixed carbon, and does not settle where it lived: at a deep-sea vent, or near the ocean surface.

Molecular clock estimates disagree, and not narrowly. One analysis, dating the root of a tree built from universally conserved genes, puts LUCA at 4.52 to 4.48 billion years ago, just under the ceiling it set at the Moon-forming impact. Another, dating gene duplications that predate LUCA, puts it at 4.33 to 4.09 billion. Both are older than the 3.4-billion-year-old Strelley Pool microfossils, which the first team counts as the oldest conclusive evidence of life, so the date rests on the clocks and on the limits each team sets for them.

LUCA is a floor, not a foundation. The lineages that branched off before it, and any that lived beside it and died out, left no descendants and therefore no genomes to compare. At most they may have passed genes sideways into lineages that survived. The line that led to LUCA leaves a little more, in gene families that had already duplicated before it lived. The rest of that history is beyond the reach of this method.

It may be more accurate to picture a community of entities exchanging genes freely, with no stable species boundaries: a population that behaved less like an ancestor and more like a shared pool. Carl Woese put it that way in 1998, calling the universal ancestor “not a discrete entity” but “a diverse community of cells that survives and evolves as a biological unit”. In that case “the last universal common ancestor” is a convenient abstraction rather than something that could have been picked up and looked at. Most reconstructions since point the other way, to a relatively complex cell, and the 2024 one describes an organism of modern prokaryote grade living as one part of an ecosystem. A 2026 proposal goes further than Woese did, treating LUCA as a process of convergence rather than any organism or collection of organisms.

If the 2016 reconstruction holds, LUCA lived in a hydrothermal setting and used hydrogen as fuel. One model adds that it had not yet built a membrane of its own that could hold an ion gradient, and lived off natural gradients through leaky walls instead, which would explain why bacteria and archaea build their membranes from different chemistry.

Alternatively, LUCA was a stage rather than a thing: a phase in which inheritance was communal and the distinction between organisms had not yet hardened. Modern lineages would then have crystallised out of that pool.