UC San Diego researchers show a natural enzyme can read a synthetic eight-letter genetic code

Researchers at the University of California, San Diego have shown that one of biology's most fundamental enzymes can accurately read genetic instructions written in an alphabet that does not exist anywhere in nature. In a study published Sept. 2 in Nature Communications, the team reported that RNA polymerase, the molecular machine that copies DNA into RNA in every living cell, can transcribe a synthetic eight-letter genetic code with the same fidelity it applies to the four letters used by all known life.

The work, led by Dong Wang, a professor at UC San Diego's Skaggs School of Pharmacy and Pharmaceutical Sciences, used high-resolution cryo-electron microscopy to capture E. coli RNA polymerase in the act of copying DNA built from two synthetic base pairs layered on top of the familiar four: adenine-thymine and guanine-cytosine. A companion study from the same lab, published Aug. 12 in the Proceedings of the National Academy of Sciences, found the enzyme could recognize a separate synthetic pair that lacks the hydrogen bonds normally thought necessary to hold DNA's rungs together. Together, the two papers show the cell's ordinary transcription machinery treating manufactured genetic letters almost exactly as it treats natural ones.

Reading letters that do not exist in nature

All terrestrial life stores genetic information in four chemical letters, commonly abbreviated A, T, G and C, arranged in two complementary pairs. The synthetic system used in the new studies, known as "Hachimoji" DNA, adds four more: P and B, which mimic the shape of natural purines, and Z and S, which mimic natural pyrimidines. P pairs only with Z, and B only with S, doubling the number of distinct letters available to encode information without altering DNA's basic double-helix architecture.

Wang's team wanted to know whether RNA polymerase, the enzyme that reads a gene and produces the RNA copy a cell uses to build proteins, would stumble over letters it never evolved to recognize. The cryo-EM structures showed it does not. According to the university's research summary, the enzyme recognizes the synthetic base pairs using much of the same biochemical and structural signaling it relies on for natural ones, despite the artificial letters having no evolutionary history inside the cell.

A quest that began with a search for alien life

The Hachimoji system did not originate at UC San Diego. It was first built in 2019 by a team led by chemist Steven Benner of the Foundation for Applied Molecular Evolution, working with the University of Texas at Austin, Indiana University and DNA Software, in research described in the journal Science. That project was funded in part by NASA, which was interested less in engineering earthly organisms than in the question of what alternative chemistries life elsewhere in the universe might use to store information; the agency's own account at the time framed the synthetic molecule as a tool for broadening the search for extraterrestrial life beyond DNA's familiar four letters. Benner is a co-author on both of the new UC San Diego papers, along with Shuichi Hoshika of his foundation and Dmitry Lyumkis of the Salk Institute for Biological Studies, whose cryo-EM expertise contributed to the structural work.

Wang's lab has been probing how the cell's own enzymes handle these synthetic letters for several years. In an earlier, related study in Nature Communications in December 2023, the group reported that RNA polymerase could not distinguish an earlier generation of synthetic base pairs from natural ones because the artificial letters mimicked natural Watson-Crick geometry closely enough to "slip in under the radar," as Wang put it at the time.

"Considering how diverse life on Earth is with just four nucleotides, the possibilities are enticing," Wang said in 2023, describing the broader premise behind his lab's work on expanding the genetic alphabet.

Who stands to use it, and what is still missing

No living organism carrying an eight-letter genome exists, and the new studies do not create one. The experiments were confined to purified E. coli RNA polymerase acting on synthetic DNA templates in a lab setting, and the researchers have not tested whether the same fidelity holds in a living cell or in human enzymes. What the structures establish, in the researchers' own framing, is a molecular explanation for why synthetic genetic information can move through a natural cell's machinery at all — a mechanistic foundation rather than a working application.

That foundation still matters to a specific set of researchers: those building diagnostic and therapeutic molecules from expanded genetic alphabets. Versions of the Hachimoji system's chemistry have already been used to evolve synthetic DNA "aptamers" — short DNA strands that bind to a target the way an antibody does — capable of latching onto liver, lung and breast cancer cells in laboratory studies cataloged in the National Institutes of Health's PubMed Central archive. Showing that RNA polymerase can transcribe these same synthetic letters strengthens the case that cells could eventually be engineered to manufacture such molecules directly, rather than requiring them to be synthesized letter by letter in a lab.

What happens next

Wang's group has signaled it intends to keep testing how far into the cell's machinery synthetic genetic information can travel — beyond transcription into translation, where a ribosome would need to turn an eight-letter RNA message into a protein containing amino acids beyond nature's standard twenty. Independent commentary on the two papers, including a summary from the Sense About Science blog, has stressed that the results remain confined to bacterial enzymes and purified components, with no timeline yet for engineered cells or clinical applications. Other outlets covering the papers, including ScienceDaily, likewise described the finding as a structural proof of principle rather than a finished technology. For now, the practical payoff Wang's team points to is narrower and further off: a molecular blueprint that other labs can use to design synthetic genetic systems the cell's own machinery will not reject.