Two billion years ago, around the time atmospheric oxygen levels were rising, one cell engulfed another, and instead of becoming lunch, the ingestee became an Earth-changer and, eventually, a vital part of you: mitochondria.
These microscopic cell inhabitants/engines allowed their host cell to suddenly begin to burn oxygen when digesting their food, an energy source that vastly expanded the amount of energy they could harvest from a given morsel of food. The magic born of this union helped enable nearly all multicellular life on Earth to evolve and get big, complicated, and, in our case, hairy and prone to back problems.
Most multicellular organisms would agree it was a good move. However, there have long been fights among biologists as to whether this aborted lunch was actually the source of mitochondria, and then, when that was all settled, what the identity of the engulfer and engulfee actually were.
Many clues lead us to the former conclusion, not least of which is the fact that mitochondria contain their own DNA separate from your cells' own nucleic DNA, and that mitochondrial DNA looks suspiciously like bacterial DNA. As for the latter, the identity of the engulfer is still hotly contested. But scientists have gotten closer to the identity of that first mitochondrion-to-be.
For years, they have known it was likely an Alphaproteobacterium, and that the mitochondria are closely allied with the Rickettsiales, a group of largely parasitic intracellular bacteria that include organisms that cause typhus and Rocky Mountain Spotted Fever. That would make sense: both mitochondria and intracellular parasites obviously have lots of adaptations that make them good at living inside other cells.
But what is the identity of the extant free-living bacterium that is most closely related to the mitochondria/Rickettsiales? Recently, biologists identified it as some member of the SAR11 group of bacteria including, most notably, Pelagibacter ubique (which still has not been properly named according to the taxonomic code, so my use of italics is not quite right). As its name implies, it may be the most common bacterium on the planet. Under the right conditions, it can make up of half of the living cells in seawater, and likes to hang out in freshwater too. But a recent paper by a team of Swedish scientists in PLoS ONE has found that not only does it seem that mitochondria are *not* most closely related to this group, their closest free-living relatives might be quite obscure indeed.
To gauge the relationships between these bacteria and attempt to identify better the closest free-living mitochondrial relative, scientists looked at the sequences of the respiration-related mitochondrial genes. Scientists have done this before, but they tended to consider groups of bacteria with completely sequenced genomes, which are filled with medically and agriculturally-signficant bacteria, because those are the ones that tend to interest us most financially and get their genomes sequenced. The authors of the study noted that this doesn't cover very well the diversity of bacteria living in sunlit, oxygenated surface ocean water, the kind of environment in which mitochondria were likely to evolve.
So they turned to the Global Ocean Survey, a marine metagenomic sequencing initiative, to see what else was out there. Perhaps to their surprise, they found that the most closely related group in their samples was an obscure little group of bacteria that made up less than 1% of the cell population of ocean surface waters in their Global Ocean Survey samples. They named it Ocean Mitochondrial Affiliate Clade, or OMAC.
To explain the earlier results implicating SAR11 bacteria as mitochondrial relatives, they speculated that members of the SAR11 clade -- including P. ubique -- and members of the mitochondrial Rickettsiales clade are rich in the nucleotide bases adenosine and tyrosine in their DNA (as opposed to the bases guanine and cytosine). Because most other alphaproteobacteria are GC-rich, this may be a convergent feature that lead them to falsely conclude they were related. When they looked at genes that were more essential, more conserved, less AT-rich, and likely more similar to the common ancestors of these bacteria, they found that the mitochondria/Rickettsiales were *not* closely related to SAR11/P. ubique.
The authors note that since we've always thought it was likely that aerobic respiration -- the neat trick that mitochondria (and many other bacteria) do to harvest vastly more energy from their food using oxygen -- evolved in ocean surface waters where oxygen was most likely to filter in first once the atmosphere started filling with oxygen (how that happened is a different story), it makes sense that mitochondria's closest extant free-living relative lives there still. They also noted that after billions of years of climate change and mass extinctions, conditions are likely to have changed for the descendants of the last free-living ancestor of our mitochondria, and thus it's not surprising they may not represent a huge proportion of the oceanic bacterial party.