Scientists Found an Enzyme Unchanged for 2 Billion Years

Every time geochemists crack open a billion-year-old rock and read the nitrogen locked inside, they are trusting one enormous assumption: that the enzyme responsible for fixing nitrogen has always left the same chemical mark. Now, for the first time, that assumption has been tested directly. A team led by Betül Kaçar at the University of Wisconsin-Madison reconstructed and experimentally characterised a library of synthetic ancestral nitrogenase genes spanning over two billion years of evolutionary history — and the resurrected nitrogenases recapitulate canonical N-isotope biosignatures with striking consistency across the entire span.

The Billion-Year Assumption Buried in Every Rock Sample

Nitrogen is indispensable to life — every amino acid, every strand of DNA contains it. When organisms fix atmospheric N₂ into usable ammonia, the enzyme nitrogenase leaves a subtle isotopic trace in the biomass it produces. That trace, measured as δ¹⁵N, gets buried in sediment and preserved for billions of years. Archean rocks dating to 3.2 billion years ago carry δ¹⁵N values between +0.7‰ and −2.8‰, a range that matches modern Mo-nitrogenase almost exactly. The entire interpretive framework of ancient nitrogen geochemistry rests on the presumption that nitrogenase has always behaved this way — yet no one had ever tested whether molecular evolution could have shifted those values.

Building Ancient Enzymes From Scratch

Previous work on ancestral proteins modified existing genes. Kaçar's team took a different approach: they built fully synthetic ancestral DNA from first principles, using phylogenetic inference to reconstruct nitrogenase sequences at four key nodes along the oxygen-tolerant Group-I lineage. The resulting variants — Anc1 through Anc4, spanning roughly 700 million to 2.3 billion years ago — were individually inserted into a nitrogenase-deficient strain of Azotobacter vinelandii, a workhorse diazotrophic bacterium. Every molecule of fixed nitrogen in those engineered cells came solely from the synthetic ancient enzyme, giving the team a clean, direct read of each ancestor's isotopic behaviour.

Why Do Resurrected Nitrogenases Recapitulate Canonical N-Isotope Biosignatures Over Two Billion Years?

The answer turns out to be deep molecular conservation. All four ancestral strains produced ε¹⁵N values between −0.9‰ and −2.9‰ — a range that overlaps entirely with the modern wild-type and sits squarely within the canonical Mo-nitrogenase window recorded in ancient rocks. There was no directional trend: older enzymes did not produce more or less fractionation than younger ones. Anc3, the second-oldest variant at roughly 1.5–2 billion years old, showed the broadest individual spread but still did not escape the modern range. Alternative nitrogenase isozymes — vanadium-based and iron-only variants — produce far more negative values, between −6‰ and −8‰. The new data confirm those enzymes were not dominant before the Great Oxidation Event at ~2.45 billion years ago, exactly as the rock record implies.

What the Stable Signature Means for Earth Science and Beyond

For geochemists, the study removes a long-standing interpretive uncertainty. Sedimentary nitrogen isotope records from the Mesoproterozoic, from ocean anoxic events in the Devonian, Jurassic, and Cretaceous — all now stand on firmer ground. Researchers can use δ¹⁵N values in those rocks to argue for Mo-nitrogenase activity without worrying that molecular evolution has quietly shifted the goalposts. For astrobiologists, the implications extend further. A biosignature that holds constant across two billion years of Earth's evolutionary noise is exactly the kind of signal worth searching for on exoplanets with nitrogen-rich atmospheres. Roger Buick of the University of Washington, a co-author of the study, has long championed nitrogen isotopes as a proxy for early life; this work directly vindicates that approach.

The Open Questions That Will Drive the Next Decade

The study has real boundaries. All four reconstructed ancestors fall after the Great Oxidation Event, leaving the pre-GOE Archean — the very interval where the oldest nitrogen biosignatures sit — untested. The team's next goal is pushing reconstructions back past 2.5 billion years, into anaerobic ancestors that predated atmospheric oxygen. There are also open questions about how ocean chemistry, temperature, and microbial ecology might modulate fractionation in ways the lab cannot fully replicate. What this work establishes clearly is a platform: synthetic ancestral enzymes expressed in living cells are now a viable tool for ground-truthing geochemical proxies, and that changes what deep-time biology can claim to know.

• Biosignatures are molecularly anchored. — The ε¹⁵N signature of Mo-nitrogenase has not drifted in over two billion years, meaning ancient rock chemistry reflects real biology, not evolutionary drift.
• Synthetic biology reads deep time. — Fully synthetic ancestral DNA expressed in living bacteria is now a proven method for testing geochemical assumptions that could never be verified any other way.
• The pre-GOE gap remains open. — Reconstructions covering the Archean before 2.5 billion years ago are the critical next step; those enzymes lived in an oxygen-free world and may tell a different story.

"We reconstructed and experimentally characterised a library of synthetic ancestral nitrogenase genes spanning over 2 billion years — establishing a molecular platform for probing ancient biosignatures from first principles." — Rucker, Kaçar et al., Nature Communications, 2026.