Bacteria rewrite DNA rules
- Scientists reported bacteria that appear to rewire standard DNA machinery in a fundamentally new way. (x.com) - The team described the mechanism as changing how genetic information is processed, challenging textbook assumptions. (x.com) - The finding arrives as AI tools are also speeding biological discovery, intensifying interest in new molecular systems. ( )
Every biology textbook says DNA-copying enzymes need a nucleic-acid template. A paper in *Science* reports a bacterial enzyme complex that builds a defined DNA strand without one. (science.org) DNA usually stores information in four chemical “letters,” and polymerases normally copy that information by matching each new letter to an existing DNA or RNA strand. The new study says a defense-associated reverse transcriptase system called DRT3 instead makes alternating double-stranded poly(GT/AC) DNA in bacteria that use it against phages, the viruses that infect bacteria. (science.org) The work was led by Pujuan Deng, Hyunbin Lee, Carlo Armijo, Haoqing Wang and Alex Gao at Stanford University and published in *Science* in April 2026. The authors report that DRT3 is built from two reverse transcriptases, Drt3a and Drt3b, plus a noncoding RNA, assembled into a 6:6:6 complex seen by cryo-electron microscopy at 2.6-angstrom resolution. (science.org) One half of the system still follows a familiar rule: Drt3a uses an ACACAC sequence in the noncoding RNA as a template to make the poly(GT) strand. The surprise is Drt3b, which the authors say makes the matching poly(AC) strand “in the complete absence of a nucleic acid template.” (science.org) That means the information for one DNA product is being dictated by the protein machine itself rather than by a prewritten DNA or RNA guide. The paper describes this as a “protein-templated mechanism for sequence-specific DNA synthesis,” expanding the known ways cells can write nucleic acids. (science.org) This is not the first sign that bacteria can bend the usual flow of genetic information. In May 2024, *Nature* reported on bacterial defense systems that “scramble the standard workflow of life,” including systems that generate new genes as part of antiviral defense. (nature.com) One of those earlier systems, DRT2, was also published in *Science* and showed bacteria making a new gene during phage defense from a noncoding RNA template rather than from a conventional gene already sitting in the genome. That work suggested bacterial immune systems were a rich source of unfamiliar DNA chemistry before the DRT3 paper arrived. (science.org; nature.com) Stanford researchers reported another rule-bending example in September 2024, when a *Nature* study found some bacteria can flip a segment inside a single gene and make different protein outcomes from the same stretch of DNA. That finding challenged the simpler textbook idea that one gene maps cleanly to one protein. (stanford.edu) The immediate biological role of DRT3 is still narrower than “rewriting life”: the paper studies an anti-phage defense system, not ordinary chromosome replication in all cells. But the authors place it alongside other odd bacterial polymerases such as AbiK, DRT9 and DRT2, arguing that bacterial immunity has become a catalog of alternative nucleic-acid writing mechanisms. (science.org) That helps explain why these systems are drawing so much attention now. CRISPR also began as a bacterial antiviral defense system before becoming a lab tool, and Stanford and other groups are increasingly pairing large biological datasets with artificial intelligence methods to sift for useful molecular machinery faster. (science.org; cell.com) For now, the cleanest takeaway is narrower and stranger: in at least one bacterial defense pathway, a protein appears to specify a DNA sequence directly. That leaves molecular biologists with one more exception to the old rule that sequence-specific DNA synthesis always needs a nucleic-acid blueprint. (science.org)