GoKawiilResearchers design synthetic cysteine proteases from scratch using AI
Scientists reported designing new enzymes capable of cleaving peptide bonds using an activated cysteine nucleophile, achieving success in 13 of 69 designs. The engineered proteases, whose structures differ substantially from natural proteases, accelerated the cutting reaction by up to 30 million times over background rates, and six crystal structures closely matched computational design models. The enzymes also functioned effectively when tested inside mammalian cells.
GoKawiil's interpretation of the reporting above, not reported fact.
The ability to design enzymes that break down proteins at specific sequences from scratch, rather than modifying existing natural proteases, suggests a new level of control over protein engineering that could expand synthetic biology toolkits. The researchers suggest the remaining performance gap compared with nature's most efficient proteases stems from imperfect active-site arrangement, pointing to a clear direction for future refinement. Demonstrated activity in living mammalian cells indicates potential paths toward therapeutic or biotechnological applications, though translation to practical use would require further validation.
- New cysteine protease enzymes were designed computationally with a 19% experimental success rate
- Some designs achieved rate enhancements up to 3×10^7 over uncatalyzed reactions, with structures closely matching predictions
- The designed enzymes functioned inside mammalian cells, suggesting applicability for synthetic biology
Source: nature.com — Choi, 2026-10-14
Published there as: “De novo design of cysteine proteases”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.