RNA strands separated, held apart by chemistry, replicating without proteins
In the long search for how life first learned to copy itself, researchers at Cambridge's MRC Laboratory have cleared a formidable obstacle: they demonstrated that RNA can replicate exponentially, strand by strand, under conditions plausible for the early Earth — no proteins required. The key was an unlikely pairing of primitive three-letter building blocks and the freeze-thaw rhythms of ancient geothermal pools, which together solved the 'strand separation problem' that had stalled origin-of-life science for decades. This is not a complete account of how life began, but it is a rare moment when a specific, testable mechanism emerges from one of nature's deepest mysteries.
- For decades, RNA replication stalled at a molecular trap: copied strands bind so tightly to their templates that without protein machinery to pry them apart, the process cannot repeat — a chicken-and-egg problem at the very origin of life.
- The Cambridge team broke the deadlock by abandoning modern biochemistry entirely, using three-letter RNA building blocks that exist in no living organism today, paired with cycles of heat, acid, and freezing that mimic ancient geothermal springs.
- In the thin liquid veins between ice crystals, trinucleotides concentrated around separated RNA strands, holding them apart long enough for replication to proceed — then thawing released the system to begin again, producing exponential copying across multiple rounds.
- A striking side effect emerged: random RNA sequences replicated under these conditions drifted toward what researchers identify as hypothesized primordial codons, hinting that the chemistry of replication itself may have helped sculpt the genetic code.
- The result solves one bottleneck but leaves the larger puzzle intact — longer sequences, self-replicating ribozymes, and the full transition to evolvable life remain open frontiers, and the trinucleotides used have no known counterpart in nature.
In May 2025, a paper in Nature Chemistry described something the field had long sought but never fully demonstrated: RNA copying itself repeatedly, exponentially, under conditions that could have existed on the early Earth — with no protein machinery involved. James Attwater and Philipp Holliger at Cambridge's MRC Laboratory of Molecular Biology, working with UCL colleagues, had cracked what researchers call the strand separation problem.
The obstacle is simple to state and brutal to solve. When RNA copies itself, the original and its complement bind into a double helix so stable it becomes a molecular trap. Modern cells deploy specialized proteins to unzip these helices constantly, but on the early Earth, before proteins existed, nothing could do that work. Earlier experiments had shown RNA could be copied once, or that fragments of the process could function in isolation — but a complete, repeating cycle had never been demonstrated without biological help.
The solution came from an unexpected direction. Instead of single RNA letters, the team used trinucleotide building blocks — three-letter units found in no living organism today. They dissolved these in a solution with RNA strands, used heat and acid to separate the double helices, then froze the mixture. In the thin liquid channels between ice crystals, trinucleotides concentrated and coated the separated strands, holding them apart like chemical spacers. Replication proceeded in those frozen veins. When the ice thawed, the cycle began again — driven by alternating temperature and pH shifts — and RNA replicated exponentially across multiple rounds, on both defined and random sequences.
The proposed natural setting is a geothermal freshwater pool where underground heat meets a cold surface, generating daily freeze-thaw cycles. Saltwater fails — it disrupts the freezing process. Evaporation fails — RNA degrades at the temperatures required. The chemistry is specific enough to be testable, and specific enough to be plausible.
What the result does not do is equally important to state. It does not explain how life began, nor does it confirm the RNA world hypothesis. It addresses one bottleneck — strand separation — with a physical, chemistry-based solution. Attwater himself noted that the Last Universal Common Ancestor was already a complex entity with vast evolutionary history behind it; the period this research probes predates any fossil or molecular evidence we can directly examine.
One detail in the findings has drawn particular attention: random RNA sequences replicated under these conditions drifted toward what researchers identify as hypothesized primordial codons — the earliest precursors to the genetic code. If that observation holds up under independent scrutiny, it would suggest that the mechanics of replication itself may have shaped the code, rather than the code arising purely through selection.
The gap between a controlled laboratory cycle and a self-sustaining, evolvable system remains real. Whether the trinucleotide-freeze-thaw mechanism can work with longer RNA sequences, and eventually with a ribozyme replicating under the same conditions, is the field's next test. For now, origin-of-life research has a specific mechanism to work with — and conditions precise enough to push further.
In May 2025, a paper appeared in Nature Chemistry describing what may be the first time scientists have watched RNA copy itself repeatedly under conditions that could have existed on the early Earth, without any help from the protein machinery that does this work in living cells today. The researchers—James Attwater and Philipp Holliger at Cambridge's MRC Laboratory of Molecular Biology, working with colleagues at UCL—had solved a problem that has frustrated the field for decades: how RNA strands could separate from their copies long enough for those copies to become templates for the next round of replication.
The obstacle they tackled is deceptively simple to state and brutally difficult to overcome. When an RNA molecule copies itself, it creates a complementary strand, and the two bind together into a double helix so stable and so quick to form that it acts like a molecular trap. In modern cells, specialized proteins unzip these helices constantly, but on the early Earth, before proteins existed, there was nothing to do that work. Earlier experiments had shown that RNA could be copied once, or that pieces of the replication process could work in isolation, but no one had demonstrated a complete cycle—replication happening over and over, exponentially, without biological machinery to assist it.
The solution came from an unexpected direction: using building blocks made of three RNA letters instead of one. These trinucleotides do not exist in any living organism today, but the team reasoned that the earliest chemistry on Earth was probably far simpler and messier than anything that has survived to the present. They dissolved these trinucleotides in a solution containing RNA strands, then subjected the mixture to heat and acid to separate the double helices. After neutralizing the solution, they froze it. In the thin liquid channels that form between ice crystals, the trinucleotides concentrated and coated the separated RNA strands, holding them apart like a chemical spacer. Replication proceeded in those frozen veins. When the ice thawed, the cycle could begin again, driven by alternating temperature and pH shifts. Over multiple rounds, the RNA replicated exponentially, with both strands of the original duplex being copied, and the system working on random RNA sequences as well as defined ones.
The proposed setting for this chemistry is a geothermal freshwater pool or spring where underground heat meets the cold surface, creating daily freeze-thaw cycles. Saltwater does not work—salt disrupts the freezing process and prevents the trinucleotides from concentrating. Evaporation, another possible concentrating mechanism, fails because RNA degrades at the high temperatures required. The chemistry is specific enough to test, and specific enough to be plausible.
It is important to be clear about what this result does and does not do. It does not explain how life began. It does not prove that the RNA world hypothesis is correct, or that RNA was definitely the first self-replicating molecule on Earth. Other recent papers have tackled different pieces of the puzzle: one group demonstrated RNA polymerase ribozymes with much higher copying accuracy; another showed a small ribozyme that could synthesize both itself and its complement, though not yet in the same reaction vessel. This paper addresses a single, specific bottleneck—the strand separation problem—and provides a physical, chemistry-based solution that requires no biological infrastructure.
Attwater himself noted in the press release that the Last Universal Common Ancestor, the earliest organism we can infer from all living things today, was already "a pretty complex entity" with vast evolutionary history behind it. The RNA world hypothesis describes a period that predates any fossil or molecular evidence we can directly examine. The trinucleotides used in the experiment do not occur in nature, which the team acknowledges. But the logic is sound: the first replicators were probably nothing like modern life, and their chemistry was probably simpler and stranger than anything preserved in any cell.
The field's current understanding is that life did not emerge from RNA alone. Nucleotides, amino acids, lipids, and simple metabolic chemistry probably all appeared and began interacting in a prebiotic environment we cannot directly observe. Other groups at Cambridge and UCL are working on how these building blocks could have assembled from even simpler precursors. This paper is one step in a much larger puzzle, and it is incomplete.
What comes next is whether the trinucleotide-freeze-thaw mechanism can work with longer RNA sequences, and eventually with the ribozyme itself replicating under the same conditions. The gap between a replication cycle that works on short sequences in a controlled lab and a self-sustaining system capable of evolution is still real. The paper also hints at something else worth watching: the observation that random RNA sequences, when replicated in this system, drifted toward what researchers call hypothesized primordial codons—the earliest precursors to the genetic code. If that holds up under scrutiny from other laboratories, it would suggest that the chemistry of replication itself may have shaped the code, rather than the code emerging purely through selection. For now, the field has a mechanism to work with and conditions specific enough to test further.
Citações Notáveis
Life is separated from pure chemistry by information, a molecular memory encoded in the genetic material that is transmitted from one generation to the next.— Philipp Holliger, MRC Laboratory of Molecular Biology
The Last Universal Common Ancestor is a pretty complex entity with a great deal of evolutionary history hidden behind it.— James Attwater, MRC Laboratory of Molecular Biology
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter that they used trinucleotides instead of the single-letter building blocks we see in life today?
Because trinucleotides don't stick to RNA strands the way single nucleotides do. They act as a spacer, keeping the strands separated during freezing. It's a chemical solution to a physical problem—the early Earth probably used whatever chemistry was available, not what we see now.
So this is saying RNA could have copied itself without any help from proteins?
Exactly. No enzymes, no machinery. Just chemistry responding to heat, cold, and pH changes that would have happened naturally in a geothermal pool. That's what makes it plausible for the early Earth.
Does this mean we've solved the origin-of-life question?
No. This solves one bottleneck. But you still need to explain where the RNA came from in the first place, how it acquired the ability to catalyze reactions, how it interacted with other molecules like lipids and amino acids. The authors are careful about this—they know what they've shown and what they haven't.
What's the codon drift observation about?
When they replicated random RNA sequences in this system, the sequences that survived tended to drift toward what researchers think the earliest genetic codes looked like. It suggests the chemistry itself may have imposed a bias on what the code became, rather than the code being purely a product of selection.
Is this result already accepted, or will other labs need to verify it?
Other labs will definitely need to reproduce it and push it further. That's how science works. But the paper gives them a specific mechanism to test and specific conditions to work with. That's valuable whether or not every detail holds up.