Oxygen fills the air today, but for most of Earth’s early history it barely existed. Scientists say the atmosphere did not hold steady oxygen until about 2.33 billion years ago, during the Great Oxidation Event, or GOE. That long wait has puzzled researchers for decades.
Now geobiologists at the Massachusetts Institute of Technology say some microbes may have learned to use oxygen hundreds of millions of years earlier. The work comes from MIT’s Fatima Husain and Gregory Fournier, along with Haitao Shang and Stilianos Louca at the University of Oregon. Their study appears in Palaeogeography, Palaeoclimatology, Palaeoecology.
The team tracked the origin of a crucial oxygen-using enzyme found across most oxygen-breathing life today. Their results point to the Mesoarchean era, between 3.2 and 2.8 billion years ago. That timing lands well before the GOE.
A new twist on an old oxygen mystery
The first big oxygen makers were cyanobacteria, microbes that use sunlight and water to make energy. Oxygen comes out as a byproduct. Scientists place the rise of cyanobacteria around 2.9 billion years ago, long before oxygen stayed in the air.
Maximum-likelihood phylogenetic tree of the heme-copper oxygen reductase family. The tree includes 5360 sequences from all domains of life and shows the recovery of subtype classification grouping, as well as the identification of a group of A2-type within the A-type oxygen reductases. (CREDIT: Palaeogeography, Palaeoclimatology, Palaeoecology)
So why did oxygen take so long to build up? Many scientists argue oxygen reacted with rocks and dissolved chemicals, which removed it fast. The MIT team adds another sink: living things that could “eat” oxygen early on.
“This does dramatically change the story of aerobic respiration,” says study co-author Fatima Husain, a postdoc in MIT’s Department of Earth, Atmospheric and Planetary Sciences (EAPS). “Our study adds to this very recently emerging story that life may have used oxygen much earlier than previously thought. It shows us how incredibly innovative life is at all periods in Earth’s history.”
Husain says the question came straight from the timeline mismatch. “We know that the microorganisms that produce oxygen were around well before the Great Oxidation Event,” Husain says. “So it was natural to ask, was there any life around at that time that could have been capable of using that oxygen for aerobic respiration?”
The enzyme that links life to oxygen
To chase that idea, the researchers focused on heme-copper oxygen reductases. These enzymes power the final step of aerobic respiration, the process that lets cells use oxygen to unlock energy from food.
In simple terms, the enzyme helps turn oxygen into water. That reaction also helps cells build a proton gradient, which then drives ATP production. ATP is the cell’s main energy currency.
The team targeted the enzyme’s “core,” where oxygen chemistry happens. They used a conserved piece called subunit I as a marker. That section holds metal centers that do the work, including a copper ion and heme groups. It also contains six key histidine building blocks that bind those metals. Those features made it easier to spot true members of the enzyme family.
Not all oxygen reductases work the same way. Scientists group many of them into A, B, and C families based on sequence features tied to proton movement. The groups also differ in how tightly they grab oxygen. A-type enzymes tend to have low oxygen affinity, while B-type and C-type tend to have higher affinity.
Multiple sequence alignment of heme-copper oxidase family members. The residues outlined in black and shaded in blue represent the six invariant histidines present in subunit I for all members of the family; the sequence segments outlined in black and shaded in green represent the A1- and A2-type sequence motifs. (CREDIT: Palaeogeography, Palaeoclimatology, Palaeoecology)
That difference matters because early Earth likely offered oxygen in tiny, local bursts, not a global blanket. If low-affinity enzymes can still function at very low oxygen levels, microbes could keep using oxygen without leaving a big atmospheric footprint.
How the team built a “family tree” of respiration
The researchers started with a huge genomic hunt. They gathered 35,984 subunit I sequences from A-, B-, and C-type oxygen reductases, plus related nitric oxide reductases. Nitric oxide reductases reduce nitric oxide, not oxygen, but they look similar enough to complicate deep evolutionary questions.
They aligned sequences, built an initial tree, and then trimmed it to cut repeats while keeping diversity. They removed partial sequences and any that lacked the six invariant histidines. That left 5,360 strong candidates.
To avoid bias in how they rooted the tree, the team used a method called Minimal Ancestor Deviation rooting. In their large tree, that approach placed nitric oxide reductases as an outgroup to oxygen reductases.
Dating the tree required even more trimming. The researchers downsampled again, using a 97 percent identity filter and manual checks to keep key cyanobacterial and eukaryotic branches. The final dating dataset held 386 sequences across 423 aligned amino acid sites.
“The hardest part of this work was that we had too much data,” Fournier says. “This enzyme is just everywhere and is present in most modern living organism. So we had to sample and filter the data down to a dataset that was representative of the diversity of modern life and also small enough to do computation with, which is not trivial.”
Maximum-likelihood trees for the heme-copper oxygen reductases. Taxa and collapsed clades are labeled according to NCBI taxonomy annotations at the order level when possible; for taxa without order-level classification, the next highest level of classification was used. (CREDIT: Palaeogeography, Palaeoclimatology, Palaeoecology) What the dates say about early oxygen use
With their curated dataset, the team ran molecular clock analyses across many model choices. They tested different priors and relaxed-clock approaches, then looked for the most consistent outcomes.
Across models, the broad message stayed stable: major heme-copper oxygen reductase lineages likely emerged before the GOE. Depending on the tree structure, key ancestors landed in the range of roughly 3.4 to 3.6 billion years ago. For A-type oxygen reductases, estimates clustered around 3.19 to 3.21 billion years ago.
The team also tested what happened when they removed two deeply branching archaeal sequences near the base of the A-type group. Those sequences pushed ages older. Without them, the mean age for the remaining A-type enzymes dropped to about 2.86 to 2.90 billion years ago. Even then, the timing still stayed pre-GOE.
The cyanobacterial branches offered another clue. Cyanobacteria make oxygen, so their oxygen-use tools hold special meaning. The tree showed a deep duplication among cyanobacteria with A-type enzymes. The researchers argue this pattern supports aerobic respiration as an early trait in these microbes, likely present in stem cyanobacteria.
For pre-duplication cyanobacterial A-type oxygen reductases, the mean ages landed near 2.36 to 2.40 billion years ago. That timing sits close to the GOE, around 2.4 to 2.3 billion years ago, and supports diversification during the transition to a more oxygenated world.
Microbes that “ate” oxygen before it filled the sky
The enzyme dates do not mean Earth had modern oxygen levels early on. Instead, they support a world with local “whiffs” of oxygen, especially near cyanobacterial mats.
Geochemical studies have reported signs consistent with short-lived oxygen presence before the GOE, including signals tied to oxidative weathering and certain metal patterns in ancient rocks. The study also notes fossil-like bubble structures linked to cyanobacterial oxygen production in very old microbial mats.
The MIT-led team argues that early oxygen consumers could help explain the long delay. Cyanobacteria may have released oxygen, but nearby microbes could have consumed it quickly. That would keep oxygen local and scarce, while also slowing any atmospheric rise.
The study also points to evidence that low-affinity A-type oxygen reductases can operate at extremely low oxygen concentrations. In lab work cited in the paper, organisms expressed these enzymes at oxygen concentrations as low as 1 nanomolar O2 per liter. Under some conditions, oxygen use never exceeded oxygen supply, keeping oxygen levels very low.
Put together, the picture looks less like a switch and more like a long tug-of-war. Rocks pulled oxygen down, but biology may have pulled, too. Life may have learned to exploit oxygen early, then helped delay oxygen’s global takeover.
“Considered all together, MIT research has filled in the gaps in our knowledge of how Earth’s oxygenation proceeded,” Husain says. “The puzzle pieces are fitting together and really underscore how life was able to diversify and live in this new, oxygenated world.”
Practical Implications of the Research
This work gives researchers a clearer timeline for when oxygen use may have started, even before oxygen became common in the air. That helps scientists interpret ancient rock signals that hint at brief oxygen spikes, since biology could have influenced those patterns.
The findings may also reshape how researchers search for life beyond Earth. If microbes can evolve oxygen use early and survive on tiny oxygen traces, scientists may need to rethink what counts as a strong “oxygen signature” on another planet.
Finally, the study highlights how quickly life can adapt to new energy sources. That insight can guide future research on early metabolism, microbial evolution, and the tight links between biology and planetary chemistry.
Related Stories