Origin of metabolism: from geochemistry to biochemistry

Origin of metabolism is one of the central problems in the history of life: how could a network of hundreds of linked reactions emerge before the enzymes that drive those reactions today existed? The new study proposes a specific sequence. Early biochemical reactions may have relied partly on inorganic catalysts in the environment, while enzymes and cofactors gradually replaced them and made metabolism increasingly self-sufficient.

At the center of the reconstruction is LUCA, the last universal common ancestor of modern cellular life. LUCA was not the first life. It represents a much later evolutionary stage at which the genetic code, ribosomes, and protein synthesis were already operating. The authors asked a narrower question: how complete was LUCA's own enzyme-catalyzed metabolism?

What modern genomes retain from ancient metabolism

The researchers analyzed 953 genomes—552 bacterial and 401 archaeal—and matched protein sequences and structures to reactions in core metabolism. Their updated network contained 424 reactions that convert simple inputs, including H2, CO2, NH3, H2S, and inorganic phosphate, into amino acids, nucleotides, and cofactors. Enzyme families could be assigned to 361 of those reactions.

Of the reconstructed protein families, 166 were traced to LUCA. Another 89 were inferred to have arisen on the lineage leading to the last bacterial common ancestor, and 38 on the lineage leading to the last archaeal common ancestor. A further 37 were too sparsely distributed for confident lineage assignment.

The implication is that LUCA already possessed a substantial metabolic network, but its enzymatic metabolism was incomplete. Nucleotide synthesis appeared comparatively complete, while amino-acid synthesis, cofactor synthesis, and parts of central carbon metabolism still required later innovations.

The study also identified a more direct signature of independent post-LUCA assembly: five reactions for which bacteria and archaea use structurally unrelated enzymes. When the same chemical task is carried out by proteins with different evolutionary origins, that pattern is consistent with independent invention after the two lineages diverged.

Before enzymes, metals may have done part of the work

An incomplete enzymatic network raises an obvious question: what carried out the missing chemistry? The authors connect this stage to serpentinizing hydrothermal systems, where water reacts with rock to generate hydrogen, alkaline conditions, and metal catalysts.

For 37 reactions in the analyzed set—about 10%—the same metabolic reaction has already been reproduced experimentally under hydrothermal conditions with metal catalysts. For many others, the relevant reaction type or a related reaction sequence has been demonstrated. In total, the authors estimate that 46% of core metabolic reactions and their geochemical analogs now have experimental examples of proceeding without enzymes and cofactors in aqueous, transition-metal-catalyzed conditions.

Phosphorylation is a particularly important problem. Modern cells constantly form and break phosphate bonds, most visibly in ATP, yet generating such bonds in water is chemically difficult. The researchers therefore tested phosphite, a reduced form of phosphorus that occurs in serpentinized rocks.

In the laboratory, nickel converted phosphite to phosphate at roughly 75% yield at 100°C. Nickel also promoted a small amount of AMP phosphorylation to ADP. Palladium on carbon was more effective: at pH 9 and 50°C, AMP was converted to ADP at about 8% yield after 72 hours, while serine was converted to phosphoserine at 42% yield after 18 hours. These experiments establish that aqueous phosphorylation can occur without an enzyme and without ATP under the tested conditions.

From geochemistry to an autonomous metabolic network

The authors use these observations to propose a gradual replacement model. At an early stage, native metals and other environmental components could have supplied some catalytic functions. Cofactors—small helper molecules that transfer electrons or chemical groups—then emerged, followed by a progressively larger set of enzymes. Together, these biological catalysts made reactions more selective and allowed metabolism to become less dependent on particular mineral surfaces.

One striking argument comes from the order in which cofactors appear in the reconstructed network. All 17 cofactors examined are required in at least some reactions that occur earlier in the network than the reactions that synthesize the cofactors themselves. The authors interpret this as evidence that simpler precursors must initially have performed those functions, with metals and hydrothermal chemistry providing plausible candidates.

This is not a laboratory recreation of the origin of life, and it does not prove that life necessarily began at a hydrothermal vent. The paper combines evolutionary reconstruction from modern genomes with laboratory chemistry, so its conclusions depend on both inference and experimental analogy. Archaeal enzymes remain less well characterized and less densely sampled than bacterial ones. The catalog of nonenzymatic reactions is incomplete, and the authors explicitly note that metals are unlikely to replace every missing enzyme. The phosphite concentrations used experimentally were also higher than those measured in modern serpentinized rocks, while palladium is rare in Earth's crust.

The paper's central conclusion is therefore more specific than claims that life “originated twice.” The authors instead argue for chemical continuity from environmental reactions to LUCA, followed by independent completion of enzymatic metabolism along the bacterial and archaeal lineages. In this model, the boundary between geochemistry and biochemistry was not a single event but a long evolutionary transition.