✦ The story
DGRs (Diversity-Generating Retroelements) are perhaps the most counter-intuitive system in the entire bacterial arsenal. Discovered by Jeff Miller at UCLA in 2002, in a phage that infects Bordetella (whooping cough), they were initially mistaken for parasitic retroviruses. They are in fact accelerated-evolution engines. The principle: a target gene (often a surface protein interacting with a variable environment) contains a "variable repeat" sequence (VR). An adjacent cassette contains a template copy (TR, template repeat). During replication, a reverse transcriptase (RT) copies the TR via an RNA intermediate — but introduces targeted errors exclusively on adenines (specific A→N mutagenesis). The result: the target protein evolves at high speed, but only in a specific region, keeping its scaffold intact. It's surgical accelerated evolution.
Discovered 2002
By Liu M., Deora R., Miller J.F. et al. (UCLA) — *Science* 295 (2002)
★ Why we care
For synthetic biology: a programmed DGR could generate in vivo libraries of protein variants optimized by selection — a directed evolution tool integrated into the chromosome. For virology: understanding how some phages use their own DGRs to escape bacterial defenses illuminates the rapid evolution of viral surface proteins. The closest eukaryotic parallel is somatic hypermutation in B lymphocytes — same principle, different biology.
◇ The detail that lands
The Bordetella DGR generates up to 10²⁰ possible variants of the Mtd adhesion protein from a single gene — a number greater than the number of atoms in the human body. It's one of the most powerful targeted diversification engines known, and we're only just beginning to repurpose it for biomolecular engineering: Miller and colleagues have shown that one can hijack a DGR to generate custom libraries of antibody or enzyme variants, without having to do classical random mutagenesis.