Engineering Saccharomyces cerevisiae secretory system for recombinant protein production: strategies from signal peptides to protease control.

Saccharomyces cerevisiae remains one of the most widely used eukaryotic platforms for recombinant protein production because of its GRAS status, genetic accessibility, and ability to perform post-translational modification. Despite these advantages, secretion of heterologous proteins is frequently limited by inefficient ER folding, trafficking bottlenecks, and proteolytic degradation. This review critically examines both established and emerging strategies aimed at overcoming these barriers, with particular emphasis on signal peptide selection and design, engineering of the early secretory pathway, and control of vacuolar and Golgi-associated proteases. Rather than treating these approaches in isolation, we integrate growing evidence showing that secretion efficiency is governed by coordinated interactions between signal sequences, folding capacity, vesicular transport, and degradation pathways. Recent advances demonstrate that single genetic interventions are rarely sufficient, and that improved secretion outcomes are more consistently achieved through combinatorial strain engineering, fusion-based designs, and pathway-level optimization. Finally, we highlight emerging computational and data-driven tools, including machine learning–guided signal peptide design and predictive models for folding and trafficking bottlenecks, which offer new opportunities for rational host optimization. Together, these insights provide a framework for developing more robust yeast secretion platforms capable of supporting complex recombinant proteins.