The three stop codons are UAA, UAG, and UGA, and their job is to signal the end of a protein-building message. Unlike the 61 other three-letter codes in the genetic code, which each call for a specific amino acid, these three do not code for any amino acid at all. Instead, they tell the ribosome, the cell’s protein-assembly machine, to release the finished protein and move on. The way cells recognize and act on these signals involves dedicated proteins, surrounding sequence context, and a surprisingly nuanced set of rules that organisms have occasionally bent or rewritten over the course of evolution.
How Stop Codons End Translation
When a ribosome slides along a messenger RNA and encounters one of the three stop codons in its reading slot (called the A site), no matching transfer RNA arrives carrying an amino acid. Instead, a protein called a release factor drops in. In bacteria, two release factors handle the job: RF1 recognizes UAA and UAG, while RF2 recognizes UAA and UGA. In eukaryotes (everything from yeast to humans), a single release factor called eRF1 recognizes all three stop codons.
What makes the recognition selective is not so much how fast the release factor lands on the ribosome, but how long it stays. Studies of bacterial RF1 show that it arrives at the ribosome at roughly the same speed regardless of whether a stop codon or an ordinary sense codon is sitting in the A site. The difference is in departure: RF1 falls off sense codons up to a thousand times faster than it falls off stop codons, meaning the release factor lingers only when the right signal is present.1PubMed Central. Kinetics of stop codon recognition by release factor 1 That lingering is what triggers the next step.
Once firmly seated on a stop codon, the release factor reaches into the part of the ribosome where the growing protein chain is still tethered to its last transfer RNA. A short stretch of amino acids in the release factor, known by the shorthand GGQ, catalyzes the chemical reaction that snips the protein free from the transfer RNA.2PubMed. A model for how ribosomal release factors induce peptidyl-tRNA cleavage in termination of protein synthesis This GGQ sequence is found in release factors across all domains of life, from bacteria to humans, pointing to an ancient and universal mechanism.3PubMed Central. Mechanism of release factor-mediated peptidyl-tRNA hydrolysis on the ribosome
Not All Three Are Used Equally
Although UAA, UAG, and UGA all serve as stop signals, organisms show strong preferences for one over the others. In simpler eukaryotes like yeast, UAA is heavily favored. In complex organisms like mammals, UGA dominates. UAG is the least-used stop codon across virtually all eukaryotes studied.4PubMed. Relationships among stop codon usage bias, its context, isochores, and gene expression level in various eukaryotes This pattern tracks with the overall composition of an organism’s genome: species with higher GC content in their DNA tend to use UGA more, while AT-rich genomes lean toward UAA.
Why would it matter which stop codon a gene uses if all three do the same thing? Part of the answer is efficiency. Not all stop codons halt the ribosome with equal reliability, and different organisms face different selective pressures around how cleanly translation should end. A gene that needs to be expressed at very high levels, where even a tiny fraction of ribosome “leaking” past the stop codon would be costly, may benefit from the stop codon that works most reliably in that organism’s cellular environment.
The Neighborhood Around a Stop Codon Matters
A stop codon does not operate in isolation. The letters immediately surrounding it, both upstream and downstream, influence how efficiently the ribosome actually stops. Work in yeast showed that the six nucleotides after the stop codon are a key factor in how much readthrough (the ribosome accidentally continuing past the stop signal) occurs.5PubMed Central. Impact of the six nucleotides downstream of the stop codon on translation termination The single nucleotide right after the stop codon (position +4) has the largest individual effect. For UAA and UGA in yeast, a G at that position gave the most efficient termination, while a C gave the least.6PubMed. The efficiency of translation termination is determined by a synergistic interplay between upstream and downstream sequences in Saccharomyces cerevisiae
In mammalian cells, a systematic test of all 192 possible combinations of stop codon plus three downstream nucleotides found that termination fails in fewer than 0.2% of translation events for most contexts. The exception was UGA followed by a C at position +4, where readthrough could climb to around 3% in some combinations.7Nucleic Acids Research. Eukaryotic translational termination efficiency is influenced by the 3′ nucleotides within the ribosomal mRNA channel That might sound small, but when a cell is churning out thousands of copies of a protein, even a few percent readthrough means a meaningful number of extended, potentially dysfunctional products. Evolution has responded by making the most commonly used stop codon-plus-context combinations in natural genes the very ones that terminate most efficiently.
What Happens After the Protein Is Released
Cutting the protein loose is not the end of the story. After termination, the ribosome is still sitting on the mRNA, clamped around a now-empty transfer RNA. Before it can be used again, the ribosome needs to be taken apart and its components recycled.
In bacteria, a protein aptly named ribosome recycling factor (RRF) teams up with EF-G, a factor normally involved in moving transfer RNAs through the ribosome during elongation. RRF enters the A site, mimicking a transfer RNA, and EF-G drives what is essentially a translocation step that ejects the leftover tRNA and mRNA. The 70S ribosome is then split into its two subunits by an initiation factor so the cycle can begin again.8PubMed Central. Post-termination complex disassembly by ribosome recycling factor, a functional tRNA mimic
Eukaryotic cells use a different toolkit. A protein called ABCE1, a member of the ABC family of energy-consuming enzymes, pries apart the ribosome’s two subunits after termination. ABCE1 can burn ATP, GTP, and other nucleotide fuels, and it specifically requires the release factor eRF1 to be present on the ribosome before it will act.9PubMed Central. The role of ABCE1 in eukaryotic posttermination ribosomal recycling This requirement acts as a checkpoint, ensuring that recycling only happens after legitimate termination has occurred.
When Stop Codons Appear in the Wrong Place
Mutations can create a stop codon in the middle of a gene where none belongs. These premature stop codons (often called nonsense mutations) would produce a truncated, usually nonfunctional protein if the cell simply obeyed the signal. To guard against this, eukaryotic cells run a surveillance process called nonsense-mediated mRNA decay, or NMD.
NMD works by exploiting a difference in context between a legitimate stop codon and a premature one. Legitimate stop codons sit near the end of the mRNA, close to a stretch of sequence and associated proteins that signal “this is really the end.” Premature stop codons are typically far upstream of those markers. Surveillance proteins, particularly one called UPF1, accumulate downstream of the premature stop codon and slow down the termination process. That delay marks the mRNA for destruction: enzymes strip away protective proteins and expose the mRNA to degradation machinery.10BMB Reports. Nonsense-mediated mRNA decay, a simplified view of a complex mechanism
NMD is medically significant because nonsense mutations cause or contribute to hundreds of genetic diseases, including cystic fibrosis, Duchenne muscular dystrophy, and certain cancers. In many of these diseases, the mutant protein would actually retain some function if it could be completed, but NMD destroys the mRNA before the cell gets a chance to make even the truncated version.
Drugs That Force the Ribosome to Read Through Stop Codons
The medical relevance of premature stop codons has fueled interest in drugs that trick the ribosome into ignoring them. Aminoglycoside antibiotics like gentamicin can promote low-level readthrough of stop codons, and researchers have been searching for compounds that boost this effect without the kidney and ear toxicity that aminoglycosides carry.
One approach combines aminoglycosides with small-molecule enhancers. A compound called CDX5-1, when paired with the aminoglycoside G418, increased production of full-length protein from premature stop codons of all three types, with UGA being the most susceptible to readthrough, followed by UAG and then UAA.11Nucleic Acids Research. Novel small molecules potentiate premature termination codon readthrough by aminoglycosides That ordering matters for drug development: the stop codon type in a patient’s specific mutation partly determines how responsive it might be to readthrough therapy.
Another promising molecule, 2,6-diaminopurine (DAP), specifically promotes readthrough of UGA premature stop codons. In a mouse model of cystic fibrosis caused by a UGA nonsense mutation in the CFTR gene, DAP corrected the defect and restored protein function. It even worked when administered to pregnant mice, crossing to the offspring through breastfeeding.12PubMed Central. Use of 2,6-diaminopurine as a potent suppressor of UGA premature stop codons in cystic fibrosis These results are still in animal models, but they illustrate how understanding the biology of stop codons opens doors to treating diseases once considered untreatable.
UGA Does Double Duty for Selenocysteine
One of the most remarkable exceptions to the “stop means stop” rule involves UGA and the amino acid selenocysteine, sometimes called the 21st amino acid. In certain genes, UGA does not trigger termination at all. Instead, it codes for selenocysteine, which contains the element selenium and is essential for enzymes that protect cells from oxidative damage.
The cell distinguishes between “stop here” and “insert selenocysteine” using a special signal in the mRNA itself. A hairpin-shaped structure called SECIS, located in the untranslated region downstream of the coding sequence, recruits a binding protein (SBP2) and a specialized elongation factor (eEFSec) that together deliver selenocysteine-loaded transfer RNA to the ribosome.13PubMed Central. Structure of the mammalian ribosome as it decodes the selenocysteine UGA codon Without the SECIS element, the same UGA codon is read as a stop signal. The SECIS element can override UGA codons placed at various positions in an mRNA, giving the system considerable flexibility.14PubMed Central. Functional characterization of the eukaryotic SECIS elements which direct selenocysteine insertion at UGA codons Modifications on the selenocysteine-specific transfer RNA also play a role in making this recoding work reliably.15PubMed. Understanding the role of tRNA modifications in UGA recoding as selenocysteine in eukaryotes
There are about 25 known selenoprotein genes in the human genome, and they participate in thyroid hormone metabolism, antioxidant defense, and other processes. Selenocysteine insertion is found across all three domains of life, making it an ancient and conserved exception to standard stop codon decoding.
UAG and the 22nd Amino Acid
UGA is not the only stop codon that sometimes encodes an amino acid. In certain archaea (single-celled organisms that thrive in extreme environments) and some bacteria, the UAG codon, normally a stop signal, is read as pyrrolysine, sometimes called the 22nd genetically encoded amino acid. Pyrrolysine is essential for enzymes involved in methane production from methylamines, a metabolically important process in these organisms.
Like selenocysteine, pyrrolysine insertion depends on a dedicated transfer RNA and its own specialized enzyme that attaches pyrrolysine to that tRNA. Expressing just these two genes in a different organism is enough to add pyrrolysine to its genetic code.16PubMed. The direct genetic encoding of pyrrolysine In the model methanogen Methanosarcina acetivorans, the UAG codon carries dual meaning: it functions as both a stop codon and a pyrrolysine codon, with the organism tuning expression of the pyrrolysine-insertion machinery in response to environmental cues like substrate availability.17PubMed Central. Methanogenic archaea encoding Pyrrolysine maintain ambiguous amber codon usage This ambiguity, a single codon simultaneously meaning “stop” and “insert this amino acid,” challenges the textbook picture of the genetic code as a fixed, unambiguous lookup table.
Viruses That Exploit Leaky Stop Codons
Some viruses have evolved to depend on stop codon readthrough as a regulatory strategy. Murine leukemia virus (MuLV), a retrovirus, places a UAG stop codon between two essential genes, gag and pol. Most of the time the ribosome stops at that codon, producing the Gag structural protein. But a small fraction of the time the ribosome reads through the UAG, producing a longer Gag-Pol fusion protein that includes the viral replicase enzymes.18PubMed Central. Modulation of stop codon read-through efficiency and its effect on the replication of murine leukemia virus The ratio of Gag to Gag-Pol is critical for the virus to assemble properly, and the leakiness of the stop codon is what sets that ratio.
This strategy is not limited to retroviruses. Several plant viruses and other RNA viruses use stop codon readthrough or related mechanisms like ribosomal frameshifting to regulate expression of their replication enzymes. The downstream sequences and RNA structures flanking the stop codon influence how often readthrough occurs, giving the virus fine-grained control over protein ratios.19PubMed Central. RNA Cis-Elements Involved in Animal Virus Stop Codon Readthrough: Stop Codon Context and Downstream RNA Structures Because readthrough is essential for these viruses, it also represents a potential antiviral target: drugs that make termination more efficient at the viral stop codon could cripple replication without affecting normal host protein synthesis.
Organisms That Have Reassigned Stop Codons Entirely
Selenocysteine and pyrrolysine represent context-dependent exceptions, but some organisms have gone further and permanently rewritten what a stop codon means. In ciliates, a group of single-celled eukaryotes, multiple lineages have independently reassigned one or more stop codons to code for amino acids. Some ciliate genetic codes are “ambiguous,” meaning a codon like UGA serves as both a stop signal and a tryptophan codon depending on context. This same UGA-to-tryptophan reassignment has arisen independently in trypanosomatids and other distantly related organisms.20eLife. Evolution of nuclear genetic codes in ciliates and the role of eRF1 and tRNA in stop codon reassignment
In mitochondrial genomes, stop codon reassignment is even more common. Many animal mitochondria use UGA to code for tryptophan rather than as a termination signal.21PubMed Central. The mechanisms of codon reassignments in mitochondrial genetic codes Bacteria have gotten in on the act too. In members of the family Eggerthellaceae (phylum Actinomycetota), genomic analysis of organisms recovered from the guts of horses, primates, and other mammals revealed UGA-to-tryptophan reassignment, accompanied by loss of the release factor RF2 that normally recognizes UGA and the appearance of a tRNA capable of reading UGA as tryptophan.22PubMed Central. Stop codon reassignment to tryptophan in members of the bacterial phylum Actinomycetota That combination, losing the termination factor and gaining a competing tRNA, is the recipe for a permanent code change.
What Happens When a Stop Codon Is Missing
If stop codons are so important, what happens when an mRNA simply does not have one? This can occur through errors in processing, premature cleavage of the mRNA, or mutations that delete the stop codon. The result is a ribosome that translates all the way to the physical end of the mRNA and stalls there, unable to either terminate or continue.
Cells have evolved rescue pathways to deal with these stuck ribosomes. In bacteria, the primary rescue system is trans-translation, which uses a hybrid molecule called tmRNA that acts partly as a transfer RNA and partly as a messenger RNA. It enters the stalled ribosome, adds a short peptide tag to the incomplete protein (marking it for destruction), and provides a built-in stop codon so the ribosome can terminate normally. A backup system involves alternative ribosome-rescue factor A (ArfA), which slots into the empty mRNA channel of the stalled ribosome and recruits RF2 to trigger peptide release even though no stop codon is present.23PubMed Central. Translational termination without a stop codon The ribosome quality control pathway provides yet another layer of backup, supporting cell survival even when the primary rescue factors are absent.24PubMed Central. The ribosome-associated quality control pathway supports survival in the absence of non-stop ribosome rescue factors
In eukaryotes, a parallel process called nonstop decay detects ribosomes that have reached the end of an mRNA without encountering a stop codon. Specialized factors recruit degradation machinery to chew up the problematic mRNA, while other quality control proteins extract the stalled ribosome and dispose of the incomplete protein.25PubMed Central. Degradation of mRNAs that lack a stop codon: a decade of nonstop progress These rescue systems underscore how central stop codons are to normal cellular function: when they are absent, cells have multiple independent safety nets to prevent the damage that stuck ribosomes can cause.
How Cryo-EM Has Revealed the Machinery in Action
Much of the detailed understanding of how release factors recognize stop codons and trigger protein release has come from cryo-electron microscopy, a technique that flash-freezes molecular complexes and images them with electron beams. By capturing ribosomes at different stages of the termination process, researchers have been able to watch the release factor cycle unfold nearly frame by frame.
Structures of bacterial ribosomes caught in the act with RF2 have been resolved at resolutions fine enough to see individual amino acid side chains, revealing how the release factor rearranges itself as it moves from initial binding to the catalytically active state.26eLife. Extensive ribosome and RF2 rearrangements during translation termination On the eukaryotic side, structures of the mammalian ribosome bound to eRF1 and eRF3 have captured the initial moment of factor attachment, before the energy-releasing step that commits the complex to termination.27PubMed Central. Cryo-EM structure of the mammalian eukaryotic release factor eRF1-eRF3-associated termination complex These snapshots have answered long-standing questions about which parts of the release factor touch the stop codon, which parts contact the ribosome’s catalytic center, and how the ribosome itself reshapes during the process. Termination turns out to be far more dynamic than early models suggested, with both the ribosome and the release factor undergoing large-scale structural shifts rather than simply locking together like puzzle pieces.