In a laboratory in Bremen a culture of microbes grows at 75 degrees, with no oxygen. The microbe is Methanocaldococcus infernus, an archaeon. Archaea are single-celled organisms without a nucleus, forming a domain of life separate from bacteria. This one was isolated from a deep-sea hydrothermal vent. The team grows it with nitrogen gas as its only source of nitrogen.
Nitrogen in the air is two atoms held together by a triple bond. Few microbes can break it, using an enzyme called nitrogenase, and turn it into ammonia. According to the paper this biological fixation supplies half the nitrogen used by living things. Industry does the same job with the Haber-Bosch process, at 200 bar and 400 to 500 degrees. The process accounts for 1 to 2 per cent of global carbon dioxide emissions.
The work is by Nevena Maslać, Mustafa Rasim Törer, Pauline Bolte and Tristan Wagner, of the Max Planck Institute for Marine Microbiology in Bremen. It came out open access in Nature Communications on 8 September 2026. Wagner is also at the Institut de Biologie Structurale in Grenoble. The institute’s press release followed on 15 September, and ScienceDaily ran it on the 19th.

In culture M. infernus fixes nitrogen at up to 91.7 degrees and doubles in two hours. It survives up to 93 degrees but loses viability after 24 hours. To grow on nitrogen it needs molybdenum and a high nitrogen partial pressure, 0.49 atmospheres. It tolerates tungstate, a metal that usually blocks the enzyme, at levels up to a thousand times higher than Azotobacter vinelandii or the archaeon Methanococcus maripaludis can bear.
The team purified the enzyme and heated it in tubes for 25 minutes at different temperatures. The part that binds nitrogen is called NifDK. In M. infernus it unfolds at 92.3 degrees. This is the melting temperature, the point at which the protein comes apart. The same part of the enzyme from Azotobacter vinelandii, the reference bacterium in this field, unfolds at 60.1 degrees. The archaeon’s other enzyme component, NifH, which supplies electrons, holds up to 82.5.

Activity is a different measurement. At room temperature, 18 to 20 degrees, the enzyme makes no measurable ammonia. At 50 degrees it makes 280 nanomoles per minute per milligram of protein. Fifty degrees is the ceiling of the system that regenerates ATP, the molecule that powers the reaction. The authors did not measure activity above it. They assume it follows the Arrhenius law, which links reaction speed to temperature. On that assumption the enzyme would reach 2,240 nanomoles at 80 degrees. That would be the highest activity ever measured for a nitrogenase, against 500 to 900 for the native Azotobacter enzyme. It is an estimate, not a measurement. The authors think the rigidity that stabilises the protein robs it, when cold, of the flexibility it needs to work.
The structure was solved with synchrotron X-rays at the ESRF in Grenoble, down to 1.21 ångström. The authors call it the most simplified nitrogenase known. It mixes traits of the three known families, molybdenum, vanadium and iron-only, without being closer to any one of them. In the active site the team caught the resting state together with a working state never seen before in a molybdenum nitrogenase. The stabilising regions lie mostly on the surface, far from the catalytic centre. The authors say the structure supports the idea that ancient nitrogenases looked more like this one than like the bacterial ones.

The press release says the enzyme only produces ammonia at high temperatures. There are two activity measurements, at room temperature and at 50 degrees. For an organism living near 90 degrees, 50 is low. The ScienceDaily headline says the enzyme survives heat that destroys most proteins. For stability the data hold. For work at those temperatures the measurement is missing.
The work was funded by the Max Planck Society. The authors declare no competing interests. The structures are deposited in the Protein Data Bank, the public archive of protein structures. The paper carries no statement on the use of artificial intelligence.
At 80 degrees, for now, there is only a calculation.

References
Maslać N., Törer M.R., Bolte P., Wagner T., «Molecular basis of N2 fixation in a hyperthermophilic archaeon», Nature Communications, 17, 9602, 8 September 2026, open access under CC BY 4.0. doi: 10.1038/s41467-026-77173-0
Max Planck Institute for Marine Microbiology, press release of 15 September 2026, «Molecular insight into biological nitrogen fixation close to the boiling point», republished by ScienceDaily on 19 September 2026 as «This deep-sea enzyme survives heat that destroys most proteins» (secondary source).