What Happened
Reverse transcriptase · Central dogma · Research briefScience, 16 April 2026, record to August 2026
Stanford, the Alex Gao lab · bacterial anti-phage defence
A Way of Making DNA in Which the Protein Is the Blueprint
A team led by Alex Gao at Stanford has described a bacterial defence enzyme, Drt3b, that builds DNA using the arrangement of amino acids in its own active site as the template — not RNA. The finding is peer-reviewed, published in Science and confirmed by a cryo-EM structure of the complex. Drt3b manufactures one fixed, repeating sequence; whether that counts as a challenge to the central dogma of molecular biology is disputed among the biologists who have examined it.
- 2 RTDRT3 pairs a conventional Drt3a with an unconventional Drt3b
- 2.6 Åresolution of the cryo-EM structure of the Drt3a:Drt3b:ncRNA complex
- 6:6:6stoichiometry of that resolved complex
- 103days from the original paper to a confirmed defence mechanism
How Firmly Each Part Stands Confirmed by peer review, and confirmed but not readable at its own source.
| How firm | What | Who says so |
|---|---|---|
| Confirmed | Deng, Lee, Armijo, Wang and Gao describe the Drt3a/Drt3b structure and mechanism, resolved by cryo-EM to 2.6 Å | Science, peer-reviewed and PubMed-indexed (PMID 41990131) |
| Confirmed | DRT3 is cytotoxic in bacterial cells lacking the host recombination complex RecBCD; the phage-encoded RecBCD inhibitor Gam triggers DRT3's abortive-infection response | Cell, peer-reviewed and PubMed-indexed (PMID 42520802) |
| Confirmed | Richard Stone's news article reporting the finding, quoting Gao, Millman, Kranzusch and Bernheim, was published in Science on 16 April 2026 | Corroborated via an EBSCO database record and a Crossref DOI record, both read directly; the article's own page returned a bot-detection wall (HTTP 403) in this session |
| Confirmed, but not verifiable here | The quotations attributed to Gao, Millman, Kranzusch and Bernheim, as Stone's article reports them | Corroborated by independent secondary outlets (ScienceAlert, AOL/Popular Mechanics, CRISPeR FRENZY) that attribute the same words to the same names; Stone's own page could not be read directly in this session |
Timeline
Every Publication With a Date Eight papers and reports, over sixteen weeks
- 16 Apr 2026Science publishes Deng, Lee, Armijo, Wang and Gao's paper on DRT3, and Richard Stone's accompanying news article.
- 21 Apr 2026ScienceAlert and the CRISPeR FRENZY blog each publish independent commentary on Stone's report.
- 24 Apr 2026AOL, syndicating Popular Mechanics, publishes further coverage carrying a critique from Nikolai Slavov of Northeastern University.
- 20 May 2026Synthetic and Systems Biotechnology publishes Yaojun Tong's commentary situating DRT3 among a wider class of defence-associated reverse transcriptases.
- 16 Jun 2026Frontiers in Microbiology publishes Sarfaraz Niazi's review naming the limits on DRT3's programmability.
- 2 Jul 2026Molecular Cell publishes a comment by Yang, Zhang and Yang on the April paper.
- 28 Jul 2026Cell publishes Wang and colleagues' paper reporting the RecBCD/Gam-linked defence mechanism, from a separate group at Columbia and the University of Tokyo.
- 6 Aug 2026Cell publishes Figiel and Nowotny's preview of the Wang paper.
The Argument
Two Readings of the Same Finding Biologists do not agree on what it means.
| Position | Argument |
|---|---|
| A genuine conceptual shift Adi Millman (MIT) | Information appears to flow from the protein's structure to a DNA sequence, which — as Stone reports it — Millman calls a meaningful shift in how the central dogma is understood. |
| A rigid, single-purpose machine Nikolai Slavov (Northeastern) | Drt3b is a highly specialised structural constraint, a “stuttering machine” that produces one repeat — not a general-purpose protein-to-DNA code. |
| No dogma broken Anna Meldolesi and the biologists she cites | The DNA that Drt3b makes is not integrated into the genome, so heritable information flow is unchanged and the dogma as originally stated still holds. |
| What both sides agree on | Drt3b makes one specific repeating sequence. It is not, as things stand, a general protein-to-DNA coding mechanism. |
What Others Add
How Reverse Transcriptase Keeps Being Repurposed Three earlier acts, and Drt3b's own.
- Retroviruses: HIV and others use reverse transcriptase to turn an RNA genome into DNA that integrates into the host chromosome — the best-known use.
- CRISPR: reverse transcriptase is a component of some CRISPR-linked defences, helping bacteria record viral sequences.
- New genes: it also participates in mechanisms by which bacteria generate new genes.
- Drt3b: it extends the line furthest yet, templating DNA on its own protein structure rather than on any nucleic acid at all.
The Defence Question, Now Partly Answered A different group finds a trigger.
Wang and colleagues at Columbia and the University of Tokyo — a separate group from the one that first described DRT3 — report that DRT3 is toxic to bacterial cells lacking RecBCD, the host's own DNA-repair and recombination complex. The phage-encoded protein Gam, which normally works by disabling RecBCD, is what triggers DRT3's abortive-infection response instead. Drt3b itself forms a hexamer, using residues next to its active site as gates that enforce the alternating addition of deoxyadenosine and deoxycytidine. Philip Kranzusch of Harvard, commenting on the original paper before this mechanism was reported, had noted that if a defence role were confirmed, DRT3 would complement other defence polymerase-like proteins that also produce nucleic acid polymers — roughly the shape the Columbia/Tokyo finding takes.
Not the Only One How many bacterial defence systems remain unexamined.
Yaojun Tong's commentary situates DRT3 within a wider, only recently recognised class of defence-associated reverse transcriptases across bacteria — DRT3 is an extreme case on that spectrum, not an isolated one. Aude Bernheim of the Institut Pasteur, discussing the original finding, said the prospect that many still-unannotated bacterial defence systems “encode exotic biochemical functions” like this one is exciting. Alex Gao himself, quoted by Stone, put the core claim plainly: “The protein itself serves as the blueprint for the DNA sequence.”
What Has Not Been Shown As of Niazi's June 2026 review.
- Drt3b has not been engineered to produce any sequence other than its own fixed repeat.
- It has not been validated in a mammalian system.
Conclusion
What to Take Away A confirmed mechanism, a live disagreement, and an early-stage technology.
A confirmed first, a disputed meaning
That a protein's own structure can template a DNA sequence is now a peer-reviewed, independently indexed finding, not merely a claim in a single news report. What it means for the central dogma remains a live disagreement among named biologists — not something this rewrite can settle, and not something it needs to.
The defence answer came from elsewhere
The RecBCD/Gam mechanism answers part of the question the original paper left open — but it comes from a different group, at Columbia and the University of Tokyo, not from the Stanford lab that first described DRT3. Whether it is the whole account of DRT3's defensive role, or one part of it, is not established by the published abstract alone.
Early days for engineering
As of the most recent review, in June 2026, Drt3b makes only its one fixed repeat and has not been shown working in a mammalian system. Whether it becomes a programmable way of writing DNA, in the way CRISPR became an editing tool, is a question the field has raised and not yet answered.