Scientists Show a Key Enzyme Can Read an Eight-Letter Genetic Alphabet

All known life on Earth relies on the same four-letter genetic alphabet. Researchers at the University of California San Diego have now shown that RNA polymerase, the enzyme that reads DNA and produces RNA in the first step of gene expression, can accurately read and transcribe a synthetic eight-letter version of DNA called "Hachimoji." The finding suggests cells can use their existing molecular machinery to handle synthetic genetic information, a step toward the synthetic-biology goal of expanding the language of DNA beyond the four letters found in nature.

To see how the enzyme handles this expanded alphabet, the researchers combined biochemical experiments with high-resolution cryo-electron microscopy, a technique that can reveal structures smaller than the width of a single atom. They captured detailed structural views of RNA polymerase from Escherichia coli (E. coli) bacteria as it recognised and incorporated two synthetic base pairs not found in nature. The images showed the enzyme identifies these synthetic letters using many of the same biochemical and structural signals it relies on to recognise natural base pairs, helping explain why it can accurately copy information written in the expanded alphabet.

In a related study published in the Proceedings of the National Academy of Sciences (PNAS), the same team found that RNA polymerase can also recognise another pair of synthetic base pairs, even though these lack the hydrogen bonds that normally help hold DNA base pairs together. Earlier research has already used expanded genetic alphabets to create synthetic DNA molecules capable of recognising liver cancer cells; the new structural findings provide a foundation for further technologies built around expanded genetic codes, including new diagnostic tools, therapeutics and engineered biological systems.

The Nature Communications study was led by Dong Wang, a professor at UC San Diego's Skaggs School of Pharmacy and Pharmaceutical Sciences, and published on September 2, 2026. The related PNAS study was published on August 12, 2026, also led by Wang.