Life uses four DNA letters, but scientists made an eight-letter system work

Every known organism writes genetic information with four DNA letters, but laboratory chemistry is not limited to that alphabet. Researchers at the University of California San Diego have shown that bacterial RNA polymerase can accurately read an eight-letter system. Structural images revealed how the enzyme handled synthetic base pairs alongside natural ones.

Hachimoji DNA Doubles the Genetic Alphabet

Ordinary DNA uses two complementary base pairs assembled from four chemical letters. The expanded system adds four synthetic letters, creating two additional pairs and bringing the total to eight. “Hachimoji” combines Japanese words for eight and letter, a compact name for the enlarged code.

The UC San Diego research account reports that RNA polymerase from Escherichia coli transcribed information containing the synthetic pairs. Accurate transcription is crucial because the enzyme must recognize the template, select the matching RNA component, and advance without losing the sequence.

RNA Polymerase Read Unnatural Base Pairs

RNA polymerase normally copies DNA into RNA at the start of gene expression. The experiment asked whether machinery shaped by evolution around four natural letters could process a larger chemical vocabulary. Biochemical tests showed that it could incorporate the synthetic information with high accuracy under the tested conditions.

The result does not mean that an eight-letter organism has been created. Transcription is one stage in a much larger system that includes DNA replication, repair, RNA processing, translation, metabolism, and reproduction. Demonstrating one compatible enzyme establishes an important step without completing the entire chain.

Cryo-Electron Microscopy Showed the Molecular Fit

High-resolution cryo-electron microscopy captured RNA polymerase while it recognized two synthetic base pairs. The structures showed the enzyme using many of the same chemical and geometric cues that guide recognition of natural pairs. The added letters therefore fit the active machinery without requiring a completely redesigned polymerase.

Structural evidence strengthens the biochemical result because it shows how successful transcription occurred. An output measurement can reveal that copying worked; an image of the molecular complex can identify contacts and conformational changes that explain why the enzyme tolerated the unfamiliar bases.

A Second Pair Worked Without Familiar Hydrogen Bonds

A related study tested another synthetic base pair that lacks the hydrogen bonds normally associated with holding natural base pairs together. RNA polymerase still recognized the pair and completed chemical steps linked with transcription. That result suggests that shape and hydrophobic interactions can also support the enzyme’s decision.

The finding does not make hydrogen bonding irrelevant throughout DNA. It shows that one part of the transcription machinery can accommodate a pair stabilized differently under experimental conditions. Stability during long-term storage and repeated copying remains a separate challenge.

More Letters Could Carry More Chemical Possibilities

An expanded alphabet can encode molecular structures unavailable to four-letter DNA. Earlier work has used synthetic genetic systems to make molecules that recognize liver cancer cells, illustrating how added chemistry could support diagnostic or therapeutic tools. Other possibilities include engineered biological components that perform functions not found in nature.

Those applications require control as well as capacity. Synthetic letters must remain available to the system, avoid unwanted exchange with natural sequences, and be copied accurately over many cycles. Any practical platform would need safeguards against mutation, loss, and unintended behavior.

The Next Test Is a Complete Information Cycle

Transcription answers whether RNA polymerase can read the larger alphabet. A more complete system must also write, preserve, copy, and use that information reliably. Researchers can measure error rates across longer sequences, repeated generations, and different sequence contexts to find where the expanded code begins to fail.

The achievement is therefore precise: a central cellular enzyme processed eight genetic letters using recognition strategies surprisingly similar to those used for natural DNA. The result expands the design space for synthetic biology while leaving the much harder task of building a stable, self-maintaining genetic system ahead.

The extra letters also increase the burden on cellular supply systems. Natural cells manufacture the components needed for ordinary DNA and RNA, while synthetic bases may need to be provided or produced through engineered pathways. A system that reads eight letters accurately for a short experiment still must maintain enough of each component to copy longer sequences without stalling or substituting a natural letter.

Error measurement will be central. A small mistake rate can become important when many synthetic positions are copied across long molecules or repeated generations. Researchers can sequence both the DNA template and RNA output to locate substitutions, skipped bases, and context-dependent failures. Those maps would show whether some neighboring letter combinations are easier for polymerase than others.

Containment is another design requirement. Expanded genetic systems can be built so that synthetic information depends on components unavailable outside the laboratory, creating a chemical barrier to persistence. Whether that barrier remains reliable would need direct testing. The same unusual letters that expand function could also provide a way to keep engineered sequences separate from ordinary biological information.

Translation presents another boundary. An expanded RNA message is valuable only if downstream machinery can use it to build an intended product, or if the RNA itself performs a designed function. Reading more letters is necessary, but not sufficient, for expanded biological output.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview