03. The Molecule That Copied Itself

Somewhere in the interval between roughly 4.5 billion years ago, when the Earth-Moon system had only just finished forming, and roughly 4.2 billion years ago, when LUCA's population already carried a working genetic code, something existed that had not yet split its two jobs apart. The 2024 study that dates LUCA calls this interval "surprisingly short." Inside it, one kind of molecule appears to have served as both the message and the machine reading it — a single substance holding heredity and chemistry at once, before life divided that work between DNA, the archive, and protein, the machinery.

That molecule is RNA, and the idea that an entire era of early life ran on RNA alone has a name and a birth date: the "RNA world," coined by Walter Gilbert of Harvard University's Biological Laboratories in a one-paragraph commentary in Nature, volume 319, page 618, in the issue dated 20 February 1986. Gilbert was the first to propose explicitly that the earliest living systems were built entirely from RNA, carrying out both information storage and catalysis — the triggering and speeding-up of chemical reactions — in one molecule, roles later split between DNA and protein. The commentary is short, but by one count from Nature's own metrics, it has since been cited 2,623 times.

Gilbert had a year-old discovery to build his idea on. Two researchers, working independently, had just shown that RNA could act as a genuine biological catalyst — a ribozyme — rather than only a carrier of genetic information. Thomas Cech, at the University of Colorado, Boulder, showed in 1982 that an unprocessed RNA molecule from the ciliate Tetrahymena thermophila could splice itself, cutting and rejoining its own sequence with no protein present anywhere in the reaction. Sidney Altman, at Yale University, built his case in stages: from a 1978 finding that the bacterial enzyme RNase P lost its function when its RNA component was removed and regained it when the RNA was put back, through to a fuller 1983 demonstration that the RNA component alone, unassisted, catalyzes the cutting reaction RNase P is known for. Cech and Altman shared the 1989 Nobel Prize in Chemistry "for their discovery of catalytic properties of RNA." By the time the prize was announced, "close to a hundred" such ribozymes were already known.

Neither discovery is a museum piece. RNase P, which processes transfer RNA, runs today in essentially every bacterium, archaeon, and eukaryote alive. Cech's self-splicing introns are still found in Tetrahymena and in the genes of many other organisms besides.

The single strongest piece of evidence for an RNA-run past, though, sits at the exact center of the machine every living cell still depends on to build proteins at all: the ribosome. In 2000, Nissen, Hansen, Ban, Moore, and Steitz published the atomic-resolution structure of the large ribosomal subunit from the archaeon Haloarcula marismortui in Science, volume 289, pages 920 to 930, and found that the peptidyl transferase center — the exact site where a growing protein chain receives its next amino acid — is built entirely from ribosomal RNA. No protein side-chain atom sits closer than about 18 angstroms to the reaction itself, a distance far too great to take part; both substrate molecules touch only conserved RNA belonging to a section of the ribosome called domain V. Every ribosome alive today, in every domain of life, still runs this same RNA-only reaction chamber. It may be the single most universal fossil the RNA world left behind, because it sits at the very center of biology's most essential machine rather than in some peripheral, vestigial corner of it.

RNA's capacity to copy itself, rather than merely catalyze a single reaction, was demonstrated directly in the laboratory in 2009. Tracey Lincoln and Gerald Joyce, at the Scripps Research Institute, engineered a pair of RNA enzymes that catalyze each other's synthesis from four short building-block molecules, so that each enzyme manufactures its own partner. The result, published in Science, volume 323, pages 1229 to 1232, was self-sustained, exponential amplification with a doubling time of about one hour, continuing indefinitely as long as fresh raw material kept arriving. When several variant versions of the replicators were set to compete for one shared pool of material, new recombinant replicators appeared during the run and came to dominate the population — Darwinian evolution, acting on a purely chemical system, with no protein involved anywhere. The honest limit on the experiment is one its own authors would recognize: whether this particular, carefully engineered pair resembles how a natural, prebiotic replicator could first have arisen on its own, rather than requiring the kind of deliberate, iterative human design that built these two, is a fair question the result cannot answer by itself.

Where the Raw Material Came From

None of this chemistry works without amino acids and nucleotides already lying around to be strung together, and the question of where those came from has its own experimental history, starting in 1953. Stanley Miller, working with Harold Urey at the University of Chicago, sealed a mixture of methane, ammonia, hydrogen, and water vapor inside a flask, a guess at early Earth's atmosphere, and ran an electric spark through it continuously for about a week, standing in for lightning. Nineteen organic compounds turned up afterward, including the amino acids glycine, alpha-alanine, and beta-alanine, built from nothing but those four inorganic starting materials and a spark. The manuscript reached Science in February 1953.

Fifty-five years later, the same source material yielded more. In 2008, a team led by Jeffrey Bada rediscovered sealed sample vials from an experiment Miller himself had run in the 1950s but never published, using a modified "volcanic apparatus" built to simulate a water-vapor-rich volcanic eruption laced with lightning. Reanalyzed with instruments Miller never had access to — modern high-performance liquid chromatography and mass spectrometry — those decades-old vials gave up 22 different amino acids and 5 amines, several never before identified in any Miller-type experiment, a notably richer haul than the original published setup produced. Published in Science on 17 October 2008, the reanalysis argues that lightning inside volcanic eruption plumes, a genuine feature of the young, volcanically active Earth, may have been an especially productive corner of the planet for assembling the raw material of life. The chemistry itself is not disputed. What is disputed is how closely Miller's assumed atmosphere, strongly reducing and rich in methane and ammonia, matched the real early atmosphere, which most current atmospheric science now judges was probably closer to a neutral mix rich in carbon dioxide and nitrogen instead.

A wholly different kind of evidence for the same chemistry arrives from outside the planet. The Murchison meteorite fell in Australia in 1969, and it carries a small but real excess of one mirror-image form over the other in the amino acid isovaline, a compound essentially absent from terrestrial biology, which is exactly why it can be trusted as an uncontaminated tracer of extraterrestrial chemistry. Measured excesses of the left-handed form range from none at all up to between 18.5 and 20.5 percent, depending on the specific stone and analytical method used — the largest such excess reported in any meteorite to date. Pizzarello's 2003 measurements, published in Geochimica et Cosmochimica Acta, found a range of 0 to 15.2 percent; a 2009 study in PNAS by Glavin and Dworkin measured 18.5 percent, plus or minus 2.6, specifically in Murchison, and up to 20.5 percent across a wider survey of carbonaceous chondrites, a class of stony meteorite rich in carbon compounds.

Murchison's parent body dates to the early Solar System, roughly 4.5 to 4.6 billion years ago, the standard formation age for meteorites of its class — meaning this chiral bias existed in Solar System rock before Earth's own homochirality was ever locked in. That timing is consistent with, though it does not prove, an extraterrestrial contribution to Earth's original tilt toward one hand. Glavin and Dworkin's own analysis adds an honest complication rather than resolving one: an excess as large as roughly 20.5 percent looks too big to explain by the leading proposed mechanism, irradiation by circularly polarized light, alone, and they propose aqueous alteration — chemical reworking by liquid water moving through the meteorite's parent body — as an additional or alternative process. The gap between mechanism and observation has not closed.

Between a spark that can write amino acids out of raw gas, a meteorite carrying a chiral bias older than Earth itself, and a laboratory demonstration that RNA alone can copy itself, compete, and evolve, the outline of heredity's first moment comes into view: a molecule that copies itself, imperfectly, inside a pool already stocked with its own raw material. Imperfect copying, repeated, is the entire mechanical root of heredity — one generation resembling, but never perfectly repeating, the one before it.

Something that could copy itself, and occasionally copy itself wrong, already had everything it needed to start accumulating history rather than just chemistry. What it did not yet have was permanence: a lock strong enough to hold the arrangement it had found. That lock was still ahead.

Overturned if: the chemistry preceding heredity is shown to have required peptides or a broader metabolic network working alongside RNA from the start, rather than a genuine stage in which RNA alone carried out both message and machine.

Sources

• 1986 · Gilbert, "Origin of life: The RNA world" — Nature 319, 618
• 2000 · Nissen, Hansen, Ban, Moore & Steitz, "The Structural Basis of Ribosome Activity in Peptide Bond Synthesis" — Science 289(5481):920
• 2009 · Lincoln & Joyce, "Self-Sustained Replication of an RNA Enzyme" — Science 323(5918):1229-1232
• 1953 · Miller-Urey spark-discharge experiment, NASA historical summary — NASA NTRS technical report
• 2008 · Johnson et al., "The Miller Volcanic Spark Discharge Experiment" — Science 322(5900):404
• 2009 · Glavin & Dworkin, "Enrichment of the amino acid L-isovaline by aqueous alteration on CI and CM meteorite parent bodies" — PNAS 106(14)