A Robot Is Rewriting the Genetic Code

The Discovery

Researchers have built an automated platform that prototypes alternate genetic codes without first rewriting a living genome. Called AGENTEX, the system combines engineered ribosomes, synthetic transfer RNAs, cell-free extracts, sequencing and laboratory robotics.

The peer-reviewed Nature study reports that the platform can assemble pools of engineered tRNAs, test how well they are charged with amino acids and measure whether a proposed code successfully produces polypeptides. It turns a laborious biological redesign into a repeatable experimental cycle.

What Makes It Strange

The standard genetic code uses 64 three-letter codons to specify amino acids, starts and stops. Because several codons usually mean the same thing, the team designed compressed codes using one codon for each of the 20 standard amino acids, plus a start and a stop. That leaves 14 codons open for reassignment to non-standard building blocks.

• Engineered ribosomes accept the platform's altered tRNAs while avoiding the natural translation system.
• The workflow robotically synthesizes and tests whole pools rather than changing one codon at a time.
• The experiments demonstrated minimal codes and incorporation of non-standard amino acids.
• The cell-free components are designed not to function as a complete system in organisms outside the experiment.

Biology's dictionary just became something a robot can edit and test.

The Implications

Expanded codes could produce proteins and sequence-defined materials using chemical building blocks unavailable to ordinary biology. Compressed codes are also being investigated for virus resistance and reduced transfer of engineered genes, although AGENTEX itself is a prototyping platform rather than a finished organism with those properties.

This is not the creation of a new form of life. The work demonstrates controlled translation in cell-free extracts and supplies a faster route for evaluating designs before the much harder task of genome engineering.

What Happens Next

The platform gives researchers a way to compare many code designs, learn which tRNA arrangements work and feed those measurements into another design round. The near-universal code is still running life on Earth. For the first time, radically different drafts can be auditioned at robotic speed.