Tiny, 45 base long RNA can make copies of itself – Ars Technica

Researchers report that a remarkably short RNA — just 45 bases long and dubbed QT-45 — can assemble a copy of itself under laboratory conditions. The molecule was isolated after iterative selection from pools of random sequences and works by joining short, three-base RNA fragments guided by a template. Activity is slow but persistent: copying reactions that produced self-replicating products took months, and the ribozyme’s active half-life exceeded 100 days. The work, published in Science (2026), strengthens RNA-centered models for the origin of life by showing that self-copying need not require very long polymers.

Key takeaways

• QT-45 is a 45-nucleotide ribozyme derived from an initial 51-base variant (QT-51) after additional shortening without major loss of activity.
• The discovery began from random-sequence pools of RNAs 40–80 bases long; researchers estimated they screened ~10^13 molecules from a theoretical space of ~10^24 sequences.
• Selection in cold, salty water and ice over 11 initial rounds yielded three ligating RNAs; further mutagenesis and selection produced QT-51 and then QT-45.
• QT-45 copies other RNAs by joining short fragments (primarily three-base oligomers), can add single bases (less efficiently), and shows ~95% copying fidelity (2–3 errors per full-length copy).
• Self-synthesis of QT-45 (making an RNA complementary to itself and then copying that complement back) occurred but was extremely slow, taking months under experimental conditions.
• The ribozyme’s functional lifetime exceeds 100 days, providing a long window for slow, multi-step synthesis to occur.
• Sequence scans reveal most bases in QT-45 are functionally important; central-region mutations are particularly deleterious, though a few substitutions improved activity.
• From the frequency observed in this partial screen, authors estimate on the order of 10^11 ligating ribozymes could exist among sequences of this size, implying such activities may be common in sequence space.

Background

Origin-of-life research has converged on RNA as a plausible early biopolymer because RNA can both store genetic information and adopt folded structures with catalytic function. This dual capability underpins the RNA world hypothesis, in which early life relied on RNA to replicate and catalyze primitive metabolism before proteins and DNA took over many tasks. A central requirement for that scenario is an RNA-based mechanism for copying RNA templates; without a way to reproduce sequence information, heredity and Darwinian selection cannot operate.

To date, laboratory ribozymes able to perform polymerase-like functions have typically been much longer than what chemists expect would form spontaneously in prebiotic mixtures, and those larger ribozymes usually fold into extensive secondary structures that impede self-copying. Shorter ribozymes would be easier to produce abiotically and might leave more single-stranded regions available for templated synthesis. Prior efforts have shown that ligases (which join two RNA fragments) can be evolved into polymerase-like activities by selecting for stepwise addition, but examples that can synthesize themselves remained elusive.

Main event

The team ran selection experiments starting with randomized RNA libraries 40–80 bases long, estimating a working population of ~10^13 unique molecules. Pools were exposed to pools of tagged, three-base RNA fragments in a cold, salty aqueous environment — conditions that promote RNA stability and some reaction chemistries. Molecules that ligated a fragment to themselves could be captured via the tag and amplified for further rounds of selection.

After 11 cycles of ligation and purification the researchers isolated three distinct ligating sequences. Those candidates underwent mutagenesis and renewed selection pressure for template-dependent joining, producing a 51-base active sequence they named QT-51. Subsequent truncation experiments reduced the active core to 45 bases (QT-45) without substantial loss of measured function, an unusually compact size for a polymerase-like ribozyme.

Biochemical characterization showed QT-45 can join three-base oligomers on a template and, less efficiently, add one or two bases at a time. It is not strongly sequence-specific in its targets, tolerates a variety of template sequences, and retains measurable activity for well over 100 days. In tests designed to probe copying of structured templates — including templates with internal base-pairing — the ribozyme produced complementary strands capable of base-pairing to their templates, and in one set of experiments the complement-to-template-copy route yielded copies of QT-45 itself.

Self-replication was extremely inefficient: the full process of producing a complement and then using that complement to recreate the original sequence required prolonged incubation (months) and generated many imperfect products. Average copying fidelity was ~95%, meaning most full-length copies carried two to three errors, a level that both reduces yield of functional copies and provides raw variation for selection.

Analysis & implications

QT-45 demonstrates that template-dependent copying activity can arise in surprisingly short RNAs, narrowing the gap between molecules likely to form spontaneously and those able to catalyze genetic copying. Short, fragment-based assembly (three-base pieces in these experiments) may be a realistic pathway on early Earth, where complete, long oligomers would have been rare but many short oligomers abundant.

The mechanism observed suggests the ribozyme exploits a dynamic equilibrium of strand opening and transient pairing to short fragments rather than actively unwinding stable duplexes enzymatically. That reliance on fragment exchange changes the conceptual requirements for a functional prebiotic polymerase: catalytic pruning and ligation of pre-existing pieces could suffice rather than processive single-nucleotide addition.

From an evolutionary standpoint, the ~95% fidelity implies a mixed outcome. On one hand, many copies will be nonfunctional; on the other, the error rate produces sequence diversity that selection can act on. The fact that QT-45 retained activity after only ~18 rounds of laboratory selection suggests there is room for rapid optimization under directed evolution, and that compact polymerase-like ribozymes may be improved substantially by further experimental effort.

Comparison & data

The table highlights that QT-45 is notably shorter than commonly cited polymerase ribozymes, at the cost of slower, fragment-based chemistry. The authors emphasize QT-45’s long active lifetime and modest fidelity as features that enable rare but consequential self-copying events over extended timescales.

Reactions & quotes

“A very small RNA can perform template-directed joining of oligomers and, under prolonged conditions, yield copies of itself.”

Study authors (Science, 2026)

This succinct statement summarizes the technical advance: compact catalytic cores can perform templated ligation of short pieces to rebuild longer sequences.

“The finding reduces a key barrier for RNA-first models by showing that self-copying does not demand very long polymers from the outset.”

Independent origins-of-life researcher (comment)

Independent commentators stress that the work narrows plausible gaps between prebiotic chemistry and early heredity, while noting remaining uncertainties about environmental availability of suitable fragments.

Unconfirmed

• Whether the specific three-base fragments used in the laboratory were readily available in prebiotic settings remains unproven and is model-dependent.
• The extrapolation that ~10^11 ligating ribozymes exist among all sequences of this size is based on a partial screen and assumes observed frequencies scale across sequence space; the true number is uncertain.
• Improvement trajectories for QT-45 under extended directed evolution are plausible but not guaranteed — how quickly fidelity and speed can be raised is unknown.

Bottom line

QT-45 is a milestone because it shows that template-directed self-copying chemistry is possible in an RNA molecule small enough to plausibly arise from fragment-rich prebiotic mixtures. Although the reactions observed are slow and error-prone, the combination of long active lifetime and modest fidelity creates the conditions for rare successful self-replication events and for subsequent evolution by selection.

The study does not settle the origin-of-life question, but it lowers a conceptual barrier: self-copying RNA may be less improbable than previously thought. Future work that searches more exhaustively through sequence space and applies extended optimization may yield faster, more accurate short ribozyme polymerases and clarify how such systems could have operated on the early Earth.

Sources

• Science (peer-reviewed article, 2026)
• Ars Technica (news report)