A pond microbe at Oxford breaks the genetic code, reading two stop signals as amino acids

A single-celled organism scooped from an artificial pond in Oxford University Parks reads the stop signals UAA and UAG as two different amino acids: lysine and glutamic acid. The Earlham Institute notes that UAA and UAG usually change meaning together, and the paper describes the ciliate, labeled Oligohymenophorea sp. PL0344, as the first reported genetic code variant of this split kind.

The finding dates to 5 October 2023, when it appeared in PLOS Genetics, and it keeps circulating because the organism turned up by accident. A sequencing test on a handful of cells exposed a rule of molecular biology with an exception living in a pond.

A sequencing test that stumbled on a new code

Dr Jamie McGowan, then a postdoctoral scientist at the Earlham Institute, was working with Professor Thomas Richards’s group at the University of Oxford to test a DNA sequencing pipeline built for very small amounts of material, down to a single cell. The test subject was an unidentified protist from a freshwater pond. According to the Earlham Institute’s release, the genome assembly showed an unexpected change in the genetic code. McGowan called it sheer luck that this particular protist was the one chosen to test the pipeline.

The PLOS Genetics paper records that surface water was sampled in April 2021, that attempts to culture the ciliate failed, and that the team therefore relied on single cells picked by hand and pools of 5 to 50 cells sorted for sequencing.

The genome that came out of those few cells is sizable: 69.7 megabases in 3,671 scaffolds, with 20,141 gene models and a BUSCO completeness score of 94.7 percent, along with a 35,635-base-pair mitochondrial genome. The organism sits in the class Oligohymenophorea, and the authors placed it among 46 ciliate species and 9 outgroup species using 89 conserved proteins. Ciliates are the clade where genetic code variation runs furthest; the Wikipedia survey of the genetic code notes that in some ciliates a stop codon can encode an amino acid when it sits inside a message and end translation only near the 3 prime end.

UAA as lysine, UAG as glutamic acid

In the standard genetic code, UAA, UAG and UGA tell the ribosome to stop building a protein. In PL0344 only UGA keeps that job. The paper reports that about 85 percent of in-frame UAA hits (74 of 87) landed on sites where related proteins carry lysine, and about 89 percent of in-frame UAG hits (56 of 63) landed on conserved glutamic acid sites. Genes containing both codons made up 95.9 percent of the annotated set, so the reassignment is routine rather than an occasional slip.

The authors annotated 320 transfer RNA genes, including 23 putative suppressor tRNAs: 12 resembling lysine tRNAs that could read UAA and 10 resembling glutamic acid tRNAs that could read UAG. That machinery makes sequencing error an unlikely explanation. McGowan said he knew of no other case in which these two stop codons are linked to two different amino acids.

UGA as the lone remaining stop signal

With two of three stop signals repurposed, UGA has to do the stopping alone, and the genome shows signs of it. The paper reports that UGA is significantly overrepresented in the first four in-frame codons after genes, and the Earlham release notes more UGA codons than expected, which may act as a backup against proteins that run on too long. In figures from the paper, 12.3 percent of genes carry at least one UGA within the first six in-frame codons downstream, against 13.6 percent of highly expressed genes and 11.5 percent in the model ciliate Tetrahymena thermophila, so the enrichment is a statistical lean rather than a wall of stop codons. The paper also finds a selenocysteine tRNA suggesting UGA may carry a second meaning in some contexts.

Reassignments of this type are not confined to one pond. A 2024 paper by McGowan, Richards, Neil Hall and David Swarbreck, published in PLOS Genetics on 17 December 2024, inferred at least three independent changes of UAG in phyllopharyngean ciliates, to leucine in three uncultivated species from the TARA Oceans dataset and to glutamine in Hartmannula sinica and Trochilia petrani. In all five of those species UAA stays a stop codon, so none repeats the split seen in the Oxford ciliate. The authors say the reassignments rest on genomic predictions and still need confirmation at the protein level, for instance by mass spectrometry.

The ScienceDaily page that re-surfaced the story on 8 October 2026 does not establish how the split code evolved, and neither does the 2023 paper, which also reports that attempts to establish a stable culture failed, leaving the single-cell data as the only record of this organism. What it leaves open is the route by which a lineage trades two shared stop signals for two distinct amino acids while keeping one stop signal in service.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview