Every living cell on Earth runs on the same division of labour: DNA stores the instructions, proteins do the work, and RNA carries messages between them. That arrangement is also a chicken-and-egg problem. Proteins are needed to copy DNA, and DNA is needed to make proteins. Something had to come first.
In the late 1980s, researchers found an answer of sorts. Some RNA molecules can catalyse chemical reactions — not just carry information. They were named ribozymes, after the protein catalysts called enzymes. A single molecule that both stores sequence information and speeds up chemistry could, in principle, do the jobs of DNA and proteins at once. The physicist-turned-biologist Walter Gilbert gave the idea its name: the RNA world.
The picture is easy to sketch. Nucleotides form and join into chains in some prebiotic pool. By chance one chain folds into a ribozyme that helps other RNAs assemble — perhaps even itself. Copying is noisy, so copies sometimes differ from their parent. Mutation plus selection is Darwinian evolution, and evolution can improve the replicators. From there, the story goes, some RNAs learned to catalyse metabolism, and the road to the first protocell comes into view.
Where the story breaks
The trouble is that the chemistry has not cooperated easily. In 2019, Gerald Joyce of the Salk Institute and colleagues made synthetic ribozymes that could build other RNAs, then scrambled their sequences to let them evolve in the lab. Mutations introduced at each round tended to destroy the catalytic ability; very few variants in each generation could still construct RNA at all. Self-sustaining RNA evolution, the experiments suggested, is hard.
There was a size problem too. Joyce's ribozymes were large — several hundred nucleotides — while experiments had only managed spontaneous assembly of fragments around 10 to 15 nucleotides long. Getting from randomly joined stubs to a molecule big enough to replicate and evolve looked like a leap no primordial pond would make.
QT45 changes the arithmetic
Recent work from Philipp Holliger's group at the MRC Laboratory of Molecular Biology in Cambridge has narrowed that gap. His team ran artificial evolution on much smaller RNAs — 20 to 40 nucleotides, with small length changes allowed each cycle — and found a 45-nucleotide molecule they called QT45, for "quite tiny." It can seize another RNA strand and copy it.
"In principle it can copy any sequence," Holliger said. And a 45-nucleotide molecule arising spontaneously, he argued, "does not seem an unreasonable ask." The team even suggests the activity might survive splitting QT45 into two fragments, since the parts should still be able to do their jobs — which matters, because short fragments assemble far more readily than long ones.
The hit rate is the more striking number. The researchers found two other small RNAs with similar activity in a starting pool of roughly 10 trillion different molecules. That pool represents about a hundred-billionth of all possible sequences of that length — so finding functional replicators that often implies this kind of chemistry is not a one-in-the-universe accident.
Holliger's group has also shown a plausible energy source: rounds of RNA replication that produce exponential growth can be driven by cycling acidity combined with repeated freezing and thawing. Icy poles on the young Earth are one candidate setting — a habitat that would have to remain life-friendly long enough for a spark to catch.
The objection that will not go away
The strongest criticism is not about copying at all. It is about whether copying is enough.
Life is not merely a self-replicating carrier of genetic information. It needs metabolism — a way to harvest energy from the environment and spend it on building useful molecules. An RNA world, as described, is a chemical system that reproduces. Whether it would be alive is a separate question, and possibly a definitional one. However life started, it probably needed more than a string of nucleotides.
There is also the question of place and time. Nobody knows whether early Earth offered a niche for an RNA world, geographically or temporally — a pond, a pore, an ice seam that stayed suitable for long enough. And the transition from replicating RNA to something with a membrane, a metabolism and heredity is itself unexplained.
What would settle it
Three kinds of evidence would move the argument. A ribozyme that copies a sequence substantially longer than itself, without falling apart under mutation. A demonstration that the same small replicators work in a plausible prebiotic mixture rather than a designed pool. And a laboratory system that couples replication to a simple metabolism — showing that a self-copying chemistry can start paying its own energy bills.
Until then, the RNA world remains the best-motivated hypothesis in origin-of-life research and an incomplete one. The chemistry is no longer the obstacle it looked like in 2019. The biology — what it would take for chemistry to become a living system — still is.



