A protein-building instruction can be correct while the machinery occasionally reads it in the wrong grouping. Understanding that distinction turns a dramatic-sounding molecular finding into a practical design problem: how can a sequence be written so the intended reading is more reliable?
- The new work examines laboratory translation, not clinical rates of illness or adverse events.
- Sequence context matters: the same modified base does not produce the same reading behaviour everywhere.
- The detailed proposed mechanism was dissected in a bacterial translation system and still needs testing in human translation.
- Research published
- Nature Communications · 28 September 2026
- Cell experiment
- HEK293T fluorescent reporters
- Other measurements
- Reconstituted translation; single-molecule FRET
- Focus
- +1 frameshifting and sequence context
Original publication: 28 Sep 2026 · The date above refers to this brief.
What it means to move the reading frame
RNA instructions are read in groups of three bases called codons. A +1 frameshift moves the grouping forward by one base, so the same sequence can lead to a different chain of amino acids. That is a change in how an RNA message is translated; it is not a change to someone’s DNA.
For an everyday analogy, put separators into a continuous string of letters, then move every separator by one place. The letters remain, but the groups change. The analogy explains the grouping problem; it does not reproduce the chemistry or suggest that every message is read incorrectly.
Source 1 ↗What the researchers actually measured
Poulis and colleagues used paired fluorescent reporters in HEK293T cells, reconstituted translation reactions and single-molecule FRET microscopy. The reporters turned a change in reading frame into a measurable fluorescence signal. The modified base, N1-methylpseudouridine (m1Ψ), increased +1 frameshifting in particular UUUC sequence contexts.
For the tested BNT162b2-derived reporter sequences, reported estimates ranged from 0.82 ± 0.02% to 1.7 ± 0.3% with unmodified RNA, and from 1.0 ± 0.2% to 2.7 ± 0.5% with modified RNA. These are ranges across different reporter constructs, with mean ± standard error at the endpoints. They are not pooled estimates for entire vaccines or percentages of patients harmed.
Source 1 ↗Why the next letter can change the result
The paper links the reading shift to pauses in the ribosome—the molecular machine that builds proteins—and to the pairing between a codon and a transfer RNA. Replacing susceptible UUUC sequences with UUCC or UUUU prevented the modification-dependent effect in the tested designs. This makes the finding relevant to sequence optimization rather than a claim that all modified RNA behaves identically.
A central qualification sits in the discussion: the detailed mechanism was demonstrated in a bacterial translation system and remains to be established in human translation. Cell reporter observations and a biochemical explanation are complementary evidence, but they should not be compressed into a claim that every mechanistic step has already been demonstrated in human tissue.
Source 1 ↗What is new after the earlier frameshifting paper?
Mulroney and colleagues reported related frameshifting work online in December 2023. That study included immune recognition of products made in the shifted frame and showed that sequence changes could reduce their production. Its report did not identify adverse human outcomes caused by those products. This is historical context, not an extra patient cohort in the new study.
Our interpretation is that the 2026 paper sharpens the mechanistic and sequence-design questions. Recognizing an unexpected peptide, explaining how it is made and establishing a health consequence are different claims. The first two can guide design and quality control without automatically proving the third.
Source 2 ↗Source 1 ↗The useful next step is a stronger design check
For follow-up research, we would compare original and redesigned sequences in relevant human cell types, assess the intended protein and unintended products with complementary assays, and check whether changes preserve useful expression. The tests should cover a range of sequence contexts rather than just one convenient reporter.
This is our proposed evaluation plan, not a completed result. The longer-term opportunity is more precise engineering of therapeutic RNA. Any claim about improved clinical safety or effectiveness would require evidence at that level; a fluorescence ratio alone cannot supply it. The paper is useful because it gives designers a testable problem and candidate sequence rules.
Three levels of evidence in an RNA-design story
| Measurement | What it can address | What it cannot supply alone |
|---|---|---|
| Cell fluorescence reporter | Reading behaviour in a designed construct | A whole-product clinical risk estimate |
| Reconstituted translation and FRET | A mechanism under controlled molecular conditions | Proof that every step is identical in human tissue |
| Clinical outcome study | Benefits and harms in a defined human use | The detailed molecular explanation without additional experiments |
Our comparison of evidence levels. Laboratory estimates in the text refer to reporter constructs, not clinical outcomes.
Your questions, answered
Does frameshifting mean that RNA changes DNA?
No. The term describes a shift in the grouping used while translating an RNA message into a protein. It does not mean the underlying DNA sequence was edited.
Are the percentages a risk estimate for vaccinated people?
No. They describe fluorescence-derived estimates for selected laboratory reporter sequences. The experiment, denominator and outcome differ from a clinical study.
Why can changing the RNA wording preserve the intended protein?
More than one codon can specify the same amino acid. This gives designers options to change parts of the nucleotide sequence while preserving the intended protein sequence. The performance of a particular redesign still needs experimental confirmation.
What is single-molecule FRET adding here?
It follows distance-sensitive fluorescent signals from labelled molecular components, helping investigators examine conformational states and timing. It adds mechanistic detail under an experimental preparation; it is not a patient test.
What should I take away from this study?
The most useful lesson is that molecular design can be tested at several levels. Keep the reporter result, proposed mechanism, sequence redesign and any future clinical claim distinct.
Limits of this interpretation
- Reporter constructs and a cultured cell line do not reproduce every tissue, delivery condition or complete therapeutic product.
- The detailed bacterial-system mechanism has not yet been demonstrated in human translation.
- Raw data, supplemental results and the source-data workbook were not independently assessed; no clinical outcome estimate is offered.
Sources & transparency
- Poulis, Robecchi, Kurochkina et al. (2026): Recoding by N1-methylpseudouridine guides rational mRNA vaccine design
Publisher-indexed abstract, selected results, discussion, methods and figure captions checked. Direct complete HTML retrieval failed; supplements, raw fluorescence data and source-data workbook were not independently assessed. CC BY 4.0 verified in the publisher rights statement; newly written explanation and original comparison table, no publisher figures reproduced. · Accessed 30 Sep 2026
DOI: 10.1038/s41467-026-77796-3 - Mulroney, Pöyry, Yam-Puc et al. (2023 online; 2024 issue): N1-methylpseudouridylation of mRNA causes +1 ribosomal frameshifting
Publisher-indexed abstract, selected conclusions and bibliographic dates checked for historical context. No complete methods or source-data assessment. Separate earlier study, not additional participants in the 2026 laboratory work. · Accessed 30 Sep 2026
DOI: 10.1038/s41586-023-06800-3 - Creative Commons Attribution 4.0 — licence for the Poulis et al. and Jhamvar et al. papers
Licence linked in those two publisher rights statements. Our articles use new wording, explanations and reading aids, with attribution to the original investigators. · Accessed 30 Sep 2026
Prepared and source-checked with AI on 30 September 2026. Press-news Team is our collective publication byline, not a claim of medical credentials or human review. No human editorial or clinical review has taken place. We did not conduct the experiments or reanalyse the raw data. We distinguish reported findings from our own explanations and proposed follow-up questions. Access limits are listed with each source. This article explains basic research and research tools; it does not evaluate an individual’s treatment. The photograph is illustrative.
Source check: AI source check — primary research, dates and selected results
Clinical review: Not applicable to this educational guide
Suggest a correction