Roughly two billion years ago, one single-celled organism ended up living permanently inside another. The descendants of that arrangement now power nearly every plant and animal on Earth.
One of the most consequential events in the history of life on Earth may have begun as something like a failed meal. Somewhere between 1.5 and 2 billion years ago, one single-celled organism engulfed another. The intended meal survived. The two cells settled into an arrangement that proved more useful than digestion, and the bacterium that almost became food became, over the next several hundred million years, the mitochondrion: the energy-producing organelle that now sits inside nearly every cell of nearly every plant, animal, fungus, and protist on Earth.
This is the endosymbiotic theory of mitochondrial origin. It is one of the most thoroughly supported propositions in evolutionary biology. The specific details, including exactly when the merger happened and exactly which bacterial lineage was involved, are still actively debated. The general account is not.
What the evidence looks like
Mitochondria carry their own DNA, separate from the DNA in the cell’s nucleus. In humans, this mitochondrial genome is small, around 16,500 base pairs encoding 37 genes. It is organised as a single circular molecule, structurally closer to a bacterial chromosome than to anything in the rest of the eukaryotic cell. The mitochondrion replicates this DNA on its own schedule, partitions copies between daughter mitochondria, and uses its own machinery, its own ribosomes, its own transfer RNAs, to translate the genes its DNA still carries.
Mitochondria also divide by splitting in two, the way bacteria reproduce. They are bounded by two membranes rather than one. The inner membrane, in chemical composition, more closely resembles a bacterial cell membrane than a eukaryotic one. Their ribosomes are bacterial in size and structure. That bacterial resemblance is one reason some antibiotics that target bacterial ribosomes, including tetracycline and chloramphenicol, can also interfere with mitochondrial protein synthesis.
A 2015 phylogenomic study in Nature’s Scientific Reports sums up the consensus position: mitochondria evolved only once, from bacteria living within their host cells, probably around two billion years ago, and the bacterial ancestor sat within the alphaproteobacteria, a diverse group of modern free-living and parasitic bacteria. Beyond that, the picture gets contested.
What is still being argued about
Pinpointing the exact alphaproteobacterial lineage closest to the mitochondrial ancestor is, on the same Scientific Reports paper, “highly debated.” Some analyses place mitochondria close to the order Rickettsiales, which today includes obligate intracellular parasites such as Rickettsia. A 2018 PNAS study argued that mitochondria branched off from a proteobacterial lineage before the divergence of all currently sampled alphaproteobacteria, which would mean the closest living relative may not be in any group we have yet sequenced.
The timing is also a range, not a single number. Most current estimates fall between roughly 1.5 and 2 billion years ago. The earliest unambiguous fossils of eukaryotic cells, which by definition would already contain mitochondria, are around 1.6 billion years old, so the merger has to have happened before that. The exact upper bound is open.