Abstract
BACKGROUND: Yarrowia lipolytica has emerged as a well-established platform for producing a wide range of biomolecules, including recombinant proteins (rProteins). Its robust metabolism and resistance to various environmental stressors make it particularly well-suited as a microbial cell factory. However, additional physiological modifications are still required to fully meet industrial demands. Over years of strain development, Y. lipolytica has been engineered to carry auxotrophic markers, streamline the secretory pathway via deletion of native secretory proteins, prevent filamentation, and enable inducible gene expression systems. RESULTS: In this study, we continued the fine-tuning of Y. lipolytica as a platform for rProtein synthesis, building on previous work. Specifically, we: (i) introduced a third auxotrophy to facilitate more complex genetic engineering strategies, (ii) removed bacterial vector elements (including antibiotic resistance genes) from previous constructs, and (iii) carried out extensive deletions of extracellular proteases and a peroxidase gene. The newly constructed chassis strains, JMY9438 and JMY9451/9452, both bear triple auxotrophies. The latter strain additionally lacks proteolytic activity due to the deletion of five protease genes. We evaluated the rProtein production efficiency of these strains harboring one, two or three integrated copies of the target gene. rProtein expression levels increased with copy number up to two; however, no further improvement was observed with three copies. Notably, the strain with protease deletions and a single gene copy showed the highest rProtein production per cell, while the strain retaining proteases but harboring two copies yielded the highest absolute rProtein levels. CONCLUSIONS: We present a new generation of Y. lipolytica chassis strains specifically optimized for recombinant protein production. Our results demonstrate that extensive protease deletions can provide a high-performance genetic background, enabling high-level rProtein production without relying on multi-copy expression strategies.