Researchers at Columbia and Harvard have successfully engineered an E. coli strain that operates without the amino acid isoleucine. By rethinking the composition of the ribosome, the team reduced the organism's standard genetic code from 20 amino acids to 19. This study marks a significant step in understanding how early life forms might have functioned with limited building blocks.
The project relied on advanced AI tools to predict how proteins could fold correctly despite the removal of a core component. While isoleucine is common in nature, the team used software to redesign structural elements within ribosomal proteins. This required an iterative process of testing and computational modeling to ensure the small subunit of the ribosome remained functional. The resulting cells grew at about 60 percent the rate of unmodified E. coli but proved capable of sustaining life over hundreds of generations.
This experiment highlights the gap between current AI capabilities and our biological understanding. While the AI successfully identified amino acid replacements that human researchers might have ignored, the models could not explain the logic behind their design choices. Scientists remain unsure why the modified cells grow more slowly, though potential causes include reduced ribosomal accuracy or slower catalytic speeds.
Looking ahead, the team plans to investigate whether this approach can scale to an entire genome. Although the immediate practical application remains unclear, the work provides a framework for testing the limits of biological systems. It opens new questions about the evolution of the genetic code and what constitutes the bare minimum for cellular function.