A diversified approach to the design of recombinant silk proteins for biomaterials

Silk offers a sustainable alternative to synthetic polymers, yet widespread industrial adoption has not been realised. Many current silk protein engineering efforts have focused on spider dragline silk; a pursuit made difficult by the metabolic burden of expressing high molecular weight, repetitive spidroins. This Perspective argues for a strategic redirection toward the broader sequence space of non-canonical silks for silk protein engineering. By mining the diversity of taxa such as Hymenoptera, Neuroptera, and Trichoptera, we can identify compact or compositionally distinct protein templates that may offer advantages for heterologous expression, stability, or targeted applications. Integrating these sequences with computational protein design, alongside experimental workflows, could expand the range of recombinant silk-like proteins available for biomaterial design. However, translating these proteins into useful materials will require parallel advances in expression, hierarchical assembly control, and scalable aqueous processing. We therefore frame silk protein engineering diversification not as a replacement for processing challenges, but as a route to broaden the molecular starting points available for application-specific biomaterials. Many current silk protein engineering efforts have focused on spider dragline silk; a pursuit made difficult by the metabolic burden of expressing high molecular weight, repetitive spidroins. This Perspective argues for a strategic redirection toward the broader sequence space of non-canonical silks for silk protein engineering.