Bacteria, Archaea and eukaryotes: domains and phylogenomic evidence

Cell organisation and ancestry are different comparisons

Bacteria and Archaea are major lineages of cellular life with many similarities in organisation but substantial molecular differences. Eukaryotic cells have a characteristic organisation involving a nucleus and other membrane-bound compartments, with modifications and exceptions in particular cells and lineages. These categories contain far more diversity than one representative drawing can show.

The term prokaryote is useful when describing cells without a membrane-bound nucleus. It does not by itself demonstrate a clade consisting of every such organism and excluding every eukaryote. As in the earlier lessons, a shared organisational state needs an ancestry hypothesis before it becomes a claim about a complete group of descendants.

In 1990, Carl Woese, Otto Kandler and Mark Wheelis proposed the domain rank and the names Bacteria, Archaea and Eucarya, drawing on molecular and cellular comparisons. This was a particular historically documented classification proposal. Its existence is distinct from the question of exactly how every lineage branches.

Ribosomal RNA and other conserved molecular components allow comparisons across very different cells. Phylogenomics extends such questions across many genes or genomic datasets. The difficult work includes sampling the lineages, identifying comparable copies, analysing alignment quality and checking whether evolutionary models fit the data.

A nuclear lineage can branch within another named category

Many phylogenomic analyses support an archaeal ancestry for the eukaryotic nuclear lineage, with a close relationship to Asgard archaea. In a tree with eukaryotes nested within the archaeal radiation, an ancestry-defined clade descending from that archaeal ancestor includes eukaryotes. A classification that lists three familiar categories need not assign them three equally independent first branches.

A 2023 study by Laura Eme and colleagues analysed expanded Asgard sampling and supported a nested placement. Its 2026 correction removed three partly redundant markers from a central dataset and repeated the analyses. The authors reported that the overall trends persisted, while support varied and the precise sister-group result depended on stated treatments. This illustrates why a paper’s dataset, methods and corrections belong with its claim.

A separate 2025 analysis also supported ancestry within Asgard but favoured a different finer placement. Our schematic therefore collapses broad branches and marks an Asgard-related connection without naming one living species as the eukaryotic ancestor or asserting a permanently settled closest order. Relationships inferred from sampled sequences are not recovered ancestral cells. More taxon sampling and model evaluation can refine the placement.

The sketch is conditional: it illustrates the nested ancestry interpretation, not an analysis locating the universal root or resolving all archaeal branches. It omits viruses and many finer relationships. A diagram designed for one question should not acquire unsupported authority for every deep-history question.

Eukaryotic cells combine inherited histories

Mitochondrial ancestry provides another important connection. Molecular and cellular evidence supports origin from a bacterial endosymbiont: a bacterium living in association within another lineage became an integrated organelle. Genes and functions subsequently changed, including transfer of some organelle-associated genes to the nuclear genome.

An organelle history is therefore not identical to the host lineage’s branching history. Nor does bacterial ancestry imply that every protein in a present mitochondrion came from that one bacterial source. Cells and their genomes are modified mosaics, not intact ancient ancestors preserved inside modern organisms.

Many eukaryotes have mitochondria or modified related organelles, and documented secondary loss cautions against defining every living eukaryote as a cell that must currently contain one. A 2016 study of Monocercomonoides reported evidence of complete organelle loss in a lineage descended from mitochondrial ancestors. That bounded example is different from claiming eukaryotic life originally lacked mitochondrial ancestry.

The network sketch distinguishes an archaeal-related host-lineage connection from a bacterial contribution to mitochondrial origin. Its crossing connection represents a historical integration, not an ordinary parent-to-offspring species split. It does not calculate the merger’s date, describe the complete host physiology or settle the order of all early cellular innovations.

Worked example

Check your understanding

1. Does the shared absence of a nucleus prove that Bacteria and non-eukaryotic Archaea form a clade excluding eukaryotes?

   Show reasoning

   No. The shared organisational state is not sufficient. In the supplied nested history, the relevant archaeal ancestor also has eukaryotic descendants, so excluding them omits part of its descent. Use the term prokaryote with its descriptive scope.

2. Why does the Asgard study’s correction matter even if its broad conclusion persisted?

   Show reasoning

   The evidence units and analyses changed: partly redundant markers were removed, results were recalculated and statistical support varied. A dated, corrected account preserves the basis and limits of the claim rather than treating the original number or a precise placement as immutable.

3. Does a eukaryote lacking a present mitochondrial organelle prove that its ancestry never involved one?

   Show reasoning

   No. A lineage can lose an ancestral feature. The Monocercomonoides research interpreted its evidence as secondary loss in an otherwise eukaryotic lineage. Present absence and historical absence are different claims.

What to carry forward

Broad cellular names organise diversity; molecular comparisons and organelle histories explain ancestry with explicit scope and dated evidence.