Every living eukaryote appears to descend from a single ancestral lineage in which an archaeal-related cell formed a permanent partnership with a bacterium that became the mitochondrion. It is the only successful origin of mitochondria known, although neither the merger’s precise mechanism nor the number of failed, vanished attempts can be established.

Every animal, plant, fungus and protist belongs to one surviving branch of cellular history. Near the base of that branch sits a partnership so thoroughly integrated that one member became part of the other.

The host was related to archaea. The partner was a bacterium from the broad alphaproteobacterial line. Over evolutionary time, the bacterium surrendered most of its independence and became the mitochondrion.

The evidence supports one origin for the mitochondria inherited by living eukaryotes. That is not the same as proving the event happened only once. Biology can reconstruct the merger that left descendants. Attempts that failed before producing a durable lineage may have vanished without a readable trace.

Mitochondria still carry a bacterial signature

Mitochondria are usually introduced as the structures that generate much of a cell’s usable chemical energy. Their evolutionary significance is deeper. They contain membranes, genetic systems and molecular machinery whose ancestry points back to bacteria.

Many mitochondria retain a small genome. They divide from pre-existing organelles rather than being assembled from nothing. Their ribosomes and core metabolic genes have bacterial affinities, while phylogenetic analyses place the ancestral organelle somewhere among Alphaproteobacteria.

The organelle’s two surrounding membranes fit an endosymbiotic origin, although membrane history has been extensively remodelled. Its innermost energy-producing membrane preserves a bacterial logic: electron-transfer chains pump protons, and ATP synthase uses the resulting gradient to make adenosine triphosphate.

The precise bacterial sister group remains disputed because almost two billion years of evolution, gene loss and transfer have blurred the signal. A broad Current Biology review by Andrew Roger, Sergio Muñoz-Gómez and Ryoma Kamikawa describes both the secure bacterial origin and the unresolved placement of the mitochondrial ancestor within that bacterial diversity.

The partner became an organelle by giving up autonomy

An endosymbiont is an organism living inside another. An organelle is more deeply integrated. The early bacterial partner lost many genes, transferred others into the host’s nuclear genome and became dependent on proteins made elsewhere in the cell.

Modern mitochondria import hundreds or thousands of proteins made from nuclear genes. These proteins cross organelle membranes through specialised molecular machines. Mitochondrial division is coordinated with the cell, while the organelle retains only a tiny fraction of the genome once carried by its free-living ancestor.

The direction of dependence runs both ways. Most mitochondria cannot survive independently, and eukaryotic cells rely on biochemical systems inherited through the organelle. The original partnership became permanent not merely because the bacterium remained inside, but because separating the two eventually became unworkable.

Energy production is only part of that dependence. Mitochondria contribute to iron-sulfur cluster assembly, carbon metabolism, lipid chemistry and cellular signalling. Some descendants adapted their organelles to oxygen-poor environments, producing hydrogenosomes or highly reduced mitosomes rather than familiar oxygen-respiring mitochondria.

Why the evidence points to one surviving origin

Eukaryotes are extraordinarily varied, but their mitochondrial systems share deep molecular homologies. The organelles use related protein-import machinery, descended from the same bacterial genetic system and occupy corresponding positions in the eukaryotic family tree.

If mitochondria had arisen independently many times, their genomes and integration machinery would be expected to trace separate bacterial origins. Instead, the simplest explanation for the shared architecture is inheritance from one ancient acquisition before the last eukaryotic common ancestor, often shortened to LECA.

LECA was not the cell that first accepted the bacterium. It was the later common ancestor from which all living eukaryotic lineages descend. By LECA’s time, the partnership was already deeply integrated and the cell was surprisingly elaborate.

Reconstructions generally give LECA a nucleus, internal membranes, a cytoskeleton, mitochondria, sophisticated trafficking and sexual or sex-like genetic exchange. That leaves an evolutionary interval between the initial host-bacterium encounter and the fully developed ancestor. The order of changes inside that interval remains one of the central disagreements in eukaryogenesis.

The apparent exceptions turned out to be descendants

For years, several anaerobic eukaryotes were described as primitively amitochondriate. Their lack of visible mitochondria seemed to suggest that some living eukaryotic branches separated before the merger.

Closer molecular study instead found reduced mitochondrial organelles in these groups. Hydrogenosomes can produce hydrogen under oxygen-poor conditions. Mitosomes have lost most familiar mitochondrial functions. Both may lack an organelle genome, yet their proteins and machinery reveal mitochondrial ancestry.

In 2016, researchers reported a genuine eukaryote without any detectable mitochondrial organelle, the gut microbe Monocercomonoides. The team found no mitochondrial genome or import system. Instead, the organism had acquired a bacterial sulfur-mobilisation system that performs essential iron-sulfur chemistry in the cytosol.

Monocercomonoides does not represent a pre-mitochondrial survivor. It sits within a eukaryotic lineage whose ancestors possessed mitochondria, so its absence is a secondary loss. The cell could discard the organelle only after another imported system replaced a function that mitochondria normally make indispensable.

Even the strongest apparent exception therefore carries a family history shaped by the original acquisition. Losing an inherited organelle is different from belonging to a lineage that never had one.

The host side of the merger looks archaeal

Genes used to copy, read and express genetic information connect eukaryotes most closely with archaea. Many metabolic genes have bacterial histories. A eukaryotic cell is not a simple half-and-half mixture, but its genome preserves a deep combination of ancestries.

The discovery of Asgard archaea sharpened the host connection. A 2017 Nature analysis assembled genomes from environmental samples and found that these microbes encode versions of proteins associated with eukaryotic cell shape, membranes and trafficking. Phylogenies placed eukaryotes within or close to this archaeal radiation, depending on the genes and methods used.

Asgard archaea are not living ancestors frozen in time. They have evolved for the same immense interval as every modern organism. Nor do eukaryotic-looking proteins prove that an Asgard cell could engulf bacteria. They are relatives that preserve clues about the lineage from which the host side may have emerged.

The finding nevertheless changed the shape of the problem. It made a separate, already eukaryote-like host less necessary. Instead, much of eukaryotic complexity may have developed from within an archaeal lineage while the bacterial partner was being acquired and integrated.

Engulfment is possible, but not established

The familiar textbook picture shows a large proto-eukaryotic cell swallowing a bacterium. This phagocytosis-first model is intuitive, and modern eukaryotes routinely engulf other cells. It also creates an ordering question. Did the host already possess the membranes, cytoskeleton and energy supply required for phagocytosis, or did mitochondrial integration help make that complexity possible?

Other models begin with metabolic exchange between separate cells. In the hydrogen hypothesis proposed by William Martin and Miklós Müller, a hydrogen-dependent archaeal host forms a partnership with a bacterium able to release hydrogen and carbon dioxide. Metabolic dependence brings the cells together before internalisation.

Not every version requires hydrogen, and researchers disagree over the ancestral metabolisms. Oxygen, sulfur compounds and methane-producing partners appear in competing models. What these syntrophic accounts share is the idea that a useful exchange preceded enclosure.

A third family of ideas has the host growing protrusions around an external partner. Membranes gradually surround the bacterium until it resides inside the expanding cell. These models differ over geometry and metabolism, but all must explain the same endpoint: stable inheritance, gene transfer, protein import and mutual dependence.

A cultured Asgard archaeon supplied a living clue

In 2020, after more than a decade of cultivation, Hiroyuki Imachi and colleagues described Prometheoarchaeum syntrophicum. The Asgard archaeon grows extremely slowly, has a small cell body, depends on metabolic partners and produces long branching protrusions.

The team proposed an “entangle-engulf-endogenize” model. In their scenario, protrusions entangled partner cells, membranes eventually enclosed them, and rising oxygen helped favour the bacterial relationship. The observation made a protrusion-based route biologically imaginable without giving the archaeon modern phagocytosis.

It remains a hypothesis. Prometheoarchaeum is not the ancestral host, and its behaviour cannot show what another cell did almost two billion years ago. Laboratory growth conditions can also select a particular partnership from a wider natural repertoire.

One cultured relative supplies plausibility, not a recording of eukaryogenesis. The merger’s physical mechanics remain unresolved because no intermediate lineage is known to have survived.

The merger may have paid for cellular complexity

One influential argument holds that mitochondria transformed the energy available per gene. A prokaryotic cell generates energy across its outer membrane, tying respiratory area to cell geometry. By placing respiratory membranes inside many organelles, a larger cell could expand energy production without relying only on its external surface.

Nick Lane and William Martin developed this case in a 2010 Nature analysis. They argued that mitochondrial bioenergetics supported much larger genomes and energetically expensive internal machinery, helping explain the gulf between typical prokaryotic and eukaryotic complexity.

The scale of that causal role remains debated. Prokaryotes can be structurally sophisticated, and estimates of the eukaryotic energy advantage depend on how cells and membranes are compared. Energy alone does not explain nuclei, sex, dynamic cytoskeletons or multicellularity.

Mitochondria may have enabled evolutionary expansion without dictating every innovation that followed. A power supply creates possibilities; it does not specify the structures a lineage will build.

From symbiont genes to a new kind of cell

The integration process was not a single swallow followed immediately by a modern cell. Genes moved from the symbiont to the host genome across many generations. Mutations removed redundant pathways. New targeting signals sent nuclear-encoded proteins back into the organelle, and transport complexes evolved to recognise them.

Conflict would have accompanied cooperation. Host cells that failed to control symbiont division could lose the partner or be overrun. Symbionts that replicated at the host’s expense might spread within a cell while harming the lineage. Stable inheritance required the interests of formerly separate organisms to become aligned.

The remaining mitochondrial genome shows that integration never became simple absorption. Different eukaryotic groups retained different genes, rearranged them and changed their genetic codes. Yet those varied genomes still descend from the same bacterial source.

This prolonged transition is why “formed a partnership” is more accurate than imagining one clean moment of transformation. The initial association and the finished organelle were separated by an evolutionary process whose intermediate stages are mostly gone.

The family tree cannot count vanished experiments

Modern genomes are records of survival. They can show that all sampled eukaryotic mitochondria converge on one ancestral organelle. They cannot reveal every temporary symbiosis between ancient cells.

A bacterium may have entered a host and killed it. A partnership may have lasted thousands of generations before ending. Another lineage may have integrated a different bacterium but disappeared before leaving descendants that survive today. Without descendants, distinctive transferred genes or an identifiable fossil, such attempts are outside the reach of phylogeny.

Fossils offer little help at that resolution. Ancient microfossils can preserve size, shape, walls or traces of division, but rarely the membranes and molecular relationships required to distinguish a temporary endosymbiosis from a permanent organelle.

The same boundary appears deeper in life’s history. SpaceDaily’s account of the last universal common ancestor noted that LUCA was not necessarily the first organism or the only lineage alive. It was the common ancestor whose descendants survived. The mitochondrial ancestor is known through the same selective archive.

“Only successful origin known” therefore has a precise meaning. It counts origins represented among living descendants. It cannot convert absence of surviving evidence into proof that every other ancient attempt failed to begin.

One known origin is not a probability

Mitochondria have one known origin because evidence from living eukaryotes converges on one acquisition. No second mitochondrial lineage survives for comparison, and no organism preserves an independent route to an equivalent organelle.

That singularity matters in astrobiology. If complex cells depend on a partnership that is extremely difficult to establish, microbial life could be common while eukaryote-like complexity remains unusual. If the partnership was attempted often and one lineage merely happened to survive Earth’s later history, the inference changes.

Earth offers no independent biosphere and no replay. One surviving event cannot reveal the probability of the event, just as one lottery winner cannot reveal how many tickets were discarded unless the full draw is known.

The careful conclusion is narrower. Animals, plants, fungi and the immense diversity of microbial eukaryotes inherited a common cellular merger. Its descendants transformed Earth’s biosphere. The mechanism that began it, and the number of partnerships evolution erased, remain beyond the surviving record.