E. coli RNA Polymerase Successfully Transcribes Synthetic Eight-Letter DNA Alphabet

For decades, a central assumption in molecular biology held that the chemistry of life was strictly constrained by the four natural DNA bases—adenine (A), thymine (T), cytosine (C), and guanine (G)—and the specific hydrogen bonds that hold them together. The opposing view, championed by synthetic biologists, argued that the genetic code is merely a chemical scaffold. If synthetic molecules could mimic the geometry of natural base pairs, they hypothesized, the cellular machinery would process them just the same, regardless of their chemical makeup.[4]

Now, high-resolution structural evidence has settled the debate. In a pair of studies published in late summer 2026, researchers at the University of California San Diego demonstrated that Escherichia coli RNA polymerase—the enzyme responsible for reading DNA and synthesizing RNA—can accurately transcribe an expanded, eight-letter genetic alphabet.[1][2]

The expanded system, known as Hachimoji DNA (from the Japanese words for "eight" and "letters"), doubles the natural genetic code by introducing four synthetic nucleotides: P, Z, B, and S.[1]

In a study published September 2 in Nature Communications, the research team, led by Dong Wang, used cryo-electron microscopy to capture the bacterial enzyme in the act of transcribing the synthetic base pairs.[2]

The structural snapshots, resolved to between 2.42 and 2.75 angstroms, revealed that RNA polymerase recognizes the synthetic letters using the exact same biochemical and structural signals it uses for natural DNA. The enzyme's active site adopts a catalytically competent configuration, seamlessly incorporating the artificial bases into the growing RNA strand.[2]

"By expanding the genetic code, we could create new molecules that have never been seen before and explore new ways of making proteins as therapeutics," Wang stated.[1]

The flexibility of the transcription machinery extends even further than geometric mimicry. In a companion study published August 12 in the Proceedings of the National Academy of Sciences, the same team showed that RNA polymerase can process a hydrophobic unnatural base pair—known as Ds:Pa—that completely lacks the hydrogen bonds normally required to hold DNA strands together.[3]

The PNAS study captured the enzyme at a 3.20-angstrom resolution, showing that the hydrophobic Ds:Pa pair forms an edge-to-edge alignment. The enzyme's "trigger loop"—a critical structural element for catalysis—closes fully around the synthetic pair, proving that hydrogen bonding is not a strict prerequisite for transcription.[3]

However, the process is not entirely symmetrical. The kinetic data revealed that the enzyme incorporates the synthetic Ds nucleotide much more efficiently than its partner, Pa. Similarly, the Nature Communications study noted that the synthetic P:Z base pair is processed at a transcription speed roughly two times slower than a natural G:C pair.[2][3]

Despite these kinetic speed limits, the overall fidelity remains high. The findings provide the first structural proof that a parallel, alternative genetic system can be fully supported by natural cellular enzymes.[1][4]

This structural foundation moves synthetic biology closer to practical applications. By proving that living cells can process an eight-letter code, researchers can now design synthetic DNA templates that instruct cells to manufacture novel amino acids, complex diagnostics, and targeted cancer therapeutics that the natural four-letter alphabet could never encode.[1][4]