Bacteria rewrite DNA

Bacteria usually make DNA by copying another strand. A Science paper published April 17 reports one bacterial enzyme can build a specific DNA sequence without any DNA or RNA template. (science.org) The work came from Alex Gao’s lab at Stanford University and focused on a bacterial anti-phage system called DRT3. The complex contains two reverse transcriptases, Drt3a and Drt3b, plus a noncoding RNA, and the paper says it makes alternating poly(GT/AC) double-stranded DNA. (science.org) DNA-copying enzymes are usually described in two camps: template-directed enzymes that read an existing nucleic-acid strand, and template-independent enzymes that add low-complexity or random stretches. The Science paper says Drt3b does something different by making a defined complementary poly(AC) strand “in the complete absence of a nucleic acid template.” (science.org) The researchers used cryo-electron microscopy at 2.6-angstrom resolution and reported a 6:6:6 assembly of Drt3a, Drt3b, and the noncoding RNA. In that structure, Drt3a reads an ACACAC sequence in the RNA to make the poly(GT) strand, while Drt3b uses conserved residues in its own active site to enforce alternating bases in the matching strand. (science.org) That places the result in a narrow but important category. Template-free DNA synthesis is not new by itself — a February 26, 2026 Nature Communications paper described “untemplated” activity across several polymerases — but that work dealt with fragments and sequence biases shaped by conditions, not a protein-guided route to a defined repeat pattern. (nature.com) The new study sits inside bacterial virus defense, not ordinary chromosome copying. The Science authors describe defense-associated reverse transcriptases as widespread anti-phage systems and say DRT3 uses an unconventional synthesis pathway as part of that response. (science.org) A second April 2026 preprint points in the same direction in another defense system, DRT7, from Southern University of Science and Technology. That manuscript says an enzyme with reverse-transcriptase and primase-polymerase domains makes protein-templated poly(T) DNA without a complementary nucleic-acid template, then extends it into poly(A)/poly(T)-rich DNA. (biorxiv.org) The caution is that the headline claim is still bounded by the chemistry the paper actually shows. In DRT3, the protein-guided step produces a highly repetitive sequence, not an open-ended rewrite of any arbitrary DNA code, and the broader significance will depend on whether similar mechanisms turn up in other enzymes and organisms. (science.org) For now, the cleanest takeaway is smaller and stranger: at least one bacterial enzyme appears to use part of its own protein structure as the guide for writing DNA. That leaves molecular biologists with a new exception to test, not yet a replacement for the standard rule. (science.org)