Generative AI designs functional thiolation domains for reprogramming non-ribosomal peptide synthetases.

Large language models and generative protein design promise to accelerate biotechnology, but it remains unclear whether they can engineer dynamic megasynth(et)ases whose activity depends on transient, context-specific domain interfaces. Non-ribosomal peptide synthetases (NRPSs) exemplify this challenge and produce many clinically used therapeutics. Here we integrate pretrained generative models (ESM3, ProteinMPNN and EvoDiff) with design-build-test-learn cycles and data-guided prioritization to generate 76 de novo thiolation (T) domains. We build and test 578 recombinant NRPS variants in vivo spanning minimal, full-length, and hybrid assembly lines. AI-designed T-domains support product formation across architectures, enable catalytically active hybrids at recombined junctions, and increase yields by up to ~3-fold relative to NRPSs carrying the native T-domain. A representative design shows improved soluble expression, refolding, and a 12 °C higher melting temperature, while molecular dynamics simulations indicate preserved global stability but reshaped, state-dependent interdomain contact networks. Together, these results establish generative design as an effective route to context-conditioned engineering and reprogramming of biosynthetic assembly lines.