Genetic Modifications and Other Tools for the Enhancement and Optimization of Recombinant Protein Manufacturing in Komagataella phaffii

Abstract

The global demand for high-quality recombinant proteins is rapidly growing. Meeting this demand while sustaining supply chains and improving worldwide accessibility requires scalable, cost-efficient production platforms. Komagataella phaffii (K. phaffii) is a promising host for addressing these challenges and has been used to produce a wide range of recombinant proteins. However, for complex molecules such as monoclonal antibodies (mAbs), space–time yields in K. phaffii remain below those of traditional hosts. This thesis presents several studies that narrowed this productivity gap by enhancing the volumetric productivity of K. phaffii through host-cell engineering and novel reactor operating strategies. First, we explored how disrupting non-essential genes affected the production of recombinant proteins in K. phaffii. Using CRISPR-Cas9, we characterized the essentiality of every gene in the K. phaffii genome. We used these results in conjunction with biological hypotheses to identify targets for genetic engineering. We disrupted 29 genes associated with secretion pathways and observed significant improvements in mAb production (up to 9x) for 19 of those strains. The best performing strains were evaluated across multiple scales, including fed-batch fermentation. We also assessed the effectiveness of combining the most beneficial genetic engineering changes. We next investigated whether transcriptomic data could be used in conjunction with essentiality to identify promising targets for genetic engineering. We found no correlation between gene expression level and the effectiveness of gene disruptions; however, our screen did uncover five genetic edits that significantly improved mAb production in K. phaffii. Beyond genetic engineering of the host organism, genomic integration significantly impacts volumetric productivity – specifically, the number of copies of a heterologous transgene encoding a recombinant protein significantly influences yield and genetic stability. Current methods for determining the copy number of recombinant transgenes have limited quantitative resolution for large constructs or constructs with high copy numbers. We developed a pipeline that uses long-read sequencing data to determine transgene copy number for large, complex, and repetitive integrants. This pipeline can be used to characterize existing strains, evaluate the impact of copy number on productivity, support optimization of process conditions, inform strain engineering decisions, and minimize confounding variables in comparative studies. We also explored novel reactor operating conditions for engineered strains of K. phaffii. We developed a stepped perfusion fermentation method that allowed for testing of up to 11 different operating conditions within a single reactor run. We used this method to identify promising process parameters for the commercial production of NIST mAb in perfusion fermentation. We also developed a novel feeding strategy for fed-batch fermentation of trastuzumab-producing strains. Closing the productivity gap between K. phaffii and traditional manufacturing hosts could unlock considerable economic and operational advantages for agile and flexible manufacturing. This thesis introduces specific genetic engineering changes and operational fermentation modes that significantly improved the productivity of several proteins of interest in K. phaffii. The strategies and tools developed and presented here provide a foundation for continued engineering of this organism to achieve target productivities and can be broadly applied to enhance the production of recombinant proteins across diverse microbial systems.