Every living cell on Earth descends from a single population of ancestors. Biologists call it LUCA, the last universal common ancestor, and the new question about it is not whether it existed, but how complete it was. A team led by researchers at Heinrich Heine University Düsseldorf argues that LUCA carried a working genetic code and ribosome but a half-built metabolism, and that the two great branches of life, bacteria and archaea, finished the job separately. The study, published August 5 in Science Advances, is open access.
The claim rests on a painstaking reconstruction of the metabolic core that all cells share. The researchers mapped the reactions that make up core biosynthesis, updated from earlier work to a list of 424 reactions, and asked which enzyme families were present in LUCA and which appeared later. Using 953 sequenced prokaryotic genomes, 552 bacterial and 401 archaeal, they clustered similar enzymes by sequence and structure and traced each cluster’s distribution across both domains of life. An enzyme family was assigned to LUCA if it was retained in a majority of genomes across at least four major taxonomic groups in both bacteria and archaea. The statistical test for the bacterial-archaeal split returned a p-value of 0.002.
Their answer: 166 enzyme families, about 46 percent of the reactions they could map, trace back to LUCA. A further 89 arose on the lineage leading to modern bacteria, and 38 on the lineage leading to archaea, with 37 too sparsely distributed to assign. The ribosome tells the same story from a different angle. LUCA’s ribosome, the protein factory present in all life, is reconstructed as having 33 proteins; the archaeal lineage added 29, the bacterial lineage added 21. In five cases, including the enzyme alanine dehydrogenase and riboflavin synthase, the two lineages appear to have invented structurally unrelated enzymes that catalyze the same reaction, independently. The picture is of an ancestor with the machinery of heredity in place but with its biochemistry still under construction, completed separately in the two domains.
The paper’s abstract states the claim in careful terms: enzymatic metabolism in the universal common ancestor was incomplete, undergoing final assembly independently in the lineages leading to bacteria and archaea. Notably, the paper does not say LUCA was not alive, and it explicitly credits LUCA with a functional ribosome, genetic code and translation. The attention-grabbing summary that appeared in the press release, one genetic code, two origins of life, is the interpretive framing of senior author William Martin, not a phrase that appears in the paper itself. Whether the independent completion of metabolism deserves the label “two origins” is now the subject of a definitional debate, since the genetic code, the ribosome and translation all trace to a single origin.
The study also offers a chemical story for how the unfinished metabolism could have run before enzymes existed. About 46 percent of core metabolic reactions have geochemical analogs, and 37 of 361 reactions have been experimentally demonstrated under conditions resembling a hydrothermal vent, using plain metal catalysts: iron, cobalt, nickel and palladium. The most striking demonstration involves phosphate. Phosphite, a phosphorus compound produced by serpentinization, the same water-rock reaction that powers vent chemistry, can donate phosphate to biological molecules. In experiments, nickel metal converted phosphite to phosphate and hydrogen gas at roughly 75 percent yield overnight at 100 degrees Celsius, and a palladium catalyst with phosphite converted the energy currency AMP into ADP at about 8 percent yield within 72 hours at 50 degrees Celsius. Serine was phosphorylated to phosphoserine at 42 percent yield in 18 hours. Nickel alone, the authors note, replaces about ten enzymes and ten cofactors of the acetyl-CoA pathway, the ancient carbon-fixation route. The picture is one of metals doing the chemistry first, with enzymes gradually taking over.
That scenario is plausible but contested in its details. Betül Kaçar of the University of Wisconsin-Madison, who studies early life and was not involved in the work, told Scientific American that the idea of the environment supplying catalysis is very plausible, but cautioned that the last common ancestor is not the same thing as the origin of life, and that how chemistry and geology passed the baton to biology remains unknown. Competing origin scenarios, including cyanosulfidic protometabolism and hot springs, are cited and argued against in the paper. The authors also acknowledge limits: archaeal genomes are less completely annotated than bacterial ones, which could bias the counts; there is no direct geological evidence from the Hadean eon; palladium is rare in Earth’s crust; and the phosphite concentrations used in the lab exceed those measured in serpentinized rocks today.
What the study does not do is close the case. It reconstructs individual reactions, not a continuous network running on metals alone, and the authors say the decisive experiment, assembling a connected metabolism without enzymes, is ongoing. If it succeeds, the picture of life’s beginning would shift from a single origin to something stranger: one code, one ribosome, and two separate completions of the chemistry needed to be a free-living cell.
Sources:
- Mrnjavac, N., Hoffmann, N.K., Schlikker, M.L. et al. “Intermediate stages in the origin of metabolism at a phosphorylating hydrothermal vent.” Science Advances, August 5, 2026. DOI: 10.1126/sciadv.aef3128.
- Heinrich Heine University Düsseldorf press release, “Early Evolution of Life: Two Origins of Life,” August 5, 2026.
- Pappas, S. “Life on Earth May Have Emerged Not Once but Twice.” Scientific American, August 7, 2026.
- ScienceDaily, “Evidence for two origins of life on Earth,” August 11, 2026. https://www.sciencedaily.com/releases/2026/08/260811011134.htm