In April 2021, Jamie McGowan was running a routine test. A computational biologist at the Earlham Institute in Norwich, England, McGowan had fed a single-celled organism’s genetic data into a sequencing pipeline to check that the software was working. The cell had been scooped from a freshwater pond in Oxford University Parks, a leafy green space where students walk between lectures and ducks paddle past Victorian-era trees. Nobody expected the sample to rewrite a foundational principle of molecular biology.
But the software flagged something that should not have been there. Two of the three “stop” signals in the organism’s genetic code were not stopping anything. Instead of telling the cell’s protein-building machinery to halt, the codons UAA and UAG were coding for amino acids, the building blocks of proteins. Across nearly every known bacterium, plant, animal, and fungus on Earth, those signals mean the same thing: stop here. In this pond ciliate, they meant keep going.
The findings, published in PLOS Genetics, confirmed that the organism, cataloged as Oligohymenophorea sp. PL0344, has reassigned UAA to code for the amino acid lysine and UAG to code for glutamic acid. Only UGA, the third canonical stop codon, still functions as a termination signal. As of June 2026, no other known organism has been documented reassigning two stop codons simultaneously to two different amino acids in this way.
Why stop codons matter
A codon is a three-letter sequence of nucleotide bases along a strand of messenger RNA. Each codon specifies either one of the 20 standard amino acids or a signal to stop translating. The National Human Genome Research Institute defines this system in its glossary entry on the genetic code, and for decades, biology textbooks have described it as universal: the same 64 codons carry the same meanings in organisms from E. coli to elephants.
That universality has always come with fine print. Mitochondria, the energy-producing compartments inside cells, use slightly altered codes. A handful of microbes have been caught reassigning a single stop codon. But the core table of 64 codons has held remarkably steady across billions of years of evolution, which is why PL0344’s double reassignment stands out. It is not a minor footnote. It is a wholesale revision of two of the three punctuation marks that every other known genome treats as non-negotiable.
How the discovery was confirmed
McGowan’s team performed both genome and transcriptome sequencing on PL0344. When they examined the transcriptome, the collection of all RNA messages the cell was actively reading, they found UAA and UAG embedded within long, intact coding sequences across thousands of genes. If those codons were still acting as stops, the predicted proteins would have been truncated fragments. Instead, the proteins aligned with known functional domains from related organisms, a pattern consistent only with the codons being read as amino acids.
To rule out contamination, software errors, or misassembly, the researchers analyzed codon-usage patterns across the full dataset. The signal was consistent: UAA appeared in positions where lysine would be expected, and UAG appeared where glutamic acid fit. The raw sequencing data and genome assemblies have been deposited in the European Nucleotide Archive under project accession PRJEB58266, making them freely available for any lab to download and independently verify.
That open-data step matters. Extraordinary claims in genomics demand reproducibility, and the public archive gives other research groups a clear path to check the work using their own computational pipelines.
What scientists still do not know
The PLOS Genetics paper establishes what PL0344’s code does but not yet how it does it. No one has identified the specific molecular machinery that makes the reassignment work. In standard cells, proteins called release factors recognize stop codons and trigger the ribosome to let go of the finished protein. Transfer RNAs, or tRNAs, are the adapter molecules that carry amino acids to the ribosome and match them to sense codons. For PL0344 to read UAA as lysine and UAG as glutamic acid, it must have evolved novel tRNAs, altered release factors, or both.
A review in the Annual Review of Biochemistry outlines the general mechanisms behind stop-codon reassignment, including tRNA evolution, release-factor modification, and context-dependent decoding. But mapping those players in PL0344 specifically will require biochemical experiments that have not yet been published.
Ciliates, the broader group to which PL0344 belongs, are already recognized as hotspots for genetic-code variation. Research documented through Oxford University’s digital repository has cataloged multiple independent reassignments of the UAG stop codon across phyllopharyngean ciliates, showing that even closely related lineages can settle on different decoding schemes. Some species in the genus Condylostoma show context-dependent behavior in which stop codons can act as either termination signals or sense codons depending on their position in a gene. PL0344 fits within this broader pattern of flexibility, but its clean, genome-wide reassignment of two codons to two specific amino acids appears to be unique.
Whether the dual reassignment provides any survival advantage is also unresolved. One hypothesis is that repurposing stop codons lets PL0344 produce longer or structurally distinct proteins that help it cope with the fluctuating conditions of a freshwater pond: shifting nutrient levels, temperature swings, microbial competition. Another possibility is that the change is largely neutral, a product of genetic drift in a lineage whose unusual genome architecture made the switch tolerable rather than beneficial. Comparative proteomics between PL0344 and ciliates that retain the standard code could test these ideas, but no such studies have appeared.
There is also the question of error tolerance. Using two former stop codons as amino-acid signals raises the risk that random mutations could eliminate the organism’s remaining genuine stop signals, potentially producing runaway protein chains that waste energy or form toxic aggregates. Whether PL0344 has evolved compensatory safeguards, such as stronger sequence signals around UGA stops, enhanced protein quality-control pathways, or RNA-level markers that flag true gene endings, remains unknown.
What it means for synthetic biology
For researchers who engineer organisms for a living, PL0344 is a natural proof of concept. Synthetic biologists have spent years manipulating stop codons in lab-built organisms, introducing artificial tRNAs and tweaking release factors to make cells incorporate non-standard amino acids into proteins. That work has produced promising results in drug development, biomaterials, and diagnostics, but engineered systems often struggle with efficiency and fidelity. The reassigned codons can “leak,” causing the ribosome to stop when it should keep going, or keep going when it should stop.
PL0344 shows that a living cell can maintain a double reassignment stably across its entire genome, across thousands of genes, without apparent collapse. If researchers can characterize the molecular toolkit that makes this possible, the specific tRNAs, their anticodon modifications, the configuration of release factors, those components could inform the design of more reliable engineered systems. The practical applications range from producing therapeutic proteins with precisely placed non-standard amino acids to building biosensors with expanded chemical capabilities.
The next concrete milestone for the field is a biochemical dissection of PL0344’s translation machinery. Until that work is done, the organism’s code remains a sequencing-level observation, robust and reproducible, but not yet mechanistically explained.
A pond that rewrote the textbook
PL0344 was not collected on an expedition to an extreme environment or isolated from a deep-sea vent. It came from a park pond in the middle of a university campus, a body of water that thousands of people walk past every week without a second thought. That ordinariness is part of the point. If a single cell from an unremarkable English pond can overturn one of biology’s most conserved rules, the diversity of translation systems in nature is almost certainly wider than current databases reflect.
Microbial surveys have historically focused on organisms that can be cultured in a lab, which excludes the vast majority of single-celled life. Advances in single-cell genomics, the same technology that flagged PL0344’s anomalous code, are now making it possible to sequence organisms that refuse to grow in a petri dish. As those tools become cheaper and more widely deployed, biologists may find that the “universal” genetic code is less a law and more a strong default, one that life has been quietly editing in ponds, soils, and oceans for millions of years.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
