(316b) N-Terminal Engineering Improves the Quality of Recombinant Proteins Secreted from Yeast

There is a global need to increase the capacity for manufacturing recombinant proteins for food, industrial enzymes, and therapeutics. The cost and complexity of protein manufacturing can be reduced by secretion of the recombinant product into the extracellular medium during fermentation. Secreting cells can be retained in the fermentation to enable continuous upstream processes, and downstream purification of the product from culture supernatant is preferable to cell lysates. Manufacturing cell lines, however, may secrete recombinant products with unwanted modifications or variations in primary, secondary, or tertiary structure. Product-related variants are especially problematic for therapeutic proteins because they may impact the function or safety of a medical intervention. Specifically, product molecules with N-terminal truncation or extension, proteolytic cleavage, or N- or O-linked glycosylation may be non-functional, immunogenic, prone to aggregation, or unstable in formulation. These product-related variants pose challenges in downstream purification because they may not be removed by affinity chromatography and typically require additional unit operations like ion exchange or hydrophobic interaction chromatography. Development and manufacturing costs can be reduced, therefore, by design of upstream processes that produce homogenous, properly modified recombinant proteins.

The yeast Komagataella phaffii (Pichia pastoris) is a common alternative host with high potential for low-cost manufacturing of therapeutic proteins like vaccine antigens and monoclonal antibodies. K. phaffii has a productive, highly developed secretory pathway, in addition to fast growth to high cell densities, and simple genetic manipulation. In this host, secretion of a recombinant protein requires genetic attachment of a signal peptide to facilitate translocation into the endoplasmic reticulum, post-translational modification, and ultimately packing into secretory vesicles. The most commonly used signal peptide in recombinant K. phaffii is the α-mating factor signal peptide (αSP) from Saccharomyces cerevisiae, because it typically yields the highest secreted titer of the protein of interest. Processing of the αSP, however, may depend on the recombinant protein, and improper cleavage of the αSP can result in product-related variants.

Here, we elucidate the impact of signal peptide processing on the quality and secreted titer of several recombinant vaccine antigens. We observed that an aggregated product-related variant of the SARS-CoV-2 receptor binding domain (RBD) is due to N-terminal extension from incomplete processing of the αSP. We eliminated this product variant and increased secreted titers by steric extension of the N-terminus of the RBD by a functional peptide or addition of one or more N-terminal amino acid residues. We also demonstrated that this strategy improves the quality of three other subunit antigens from a trivalent vaccine for rotavirus. Finally, we applied N-terminal engineering to a monoclonal antibody and observed the same improvement in product quality.

These results together suggest that processing of the signal peptide is a critical determinant of the quality of secreted recombinant proteins. To date, wide adoption of alternative manufacturing hosts like K. phaffii has been limited by product-related variants of unique proteins like subunit vaccine antigens, and by improper processing and modification of complex intravenously delivered products like monoclonal antibodies. The framework presented here will enable reliable secretion of a wide range of recombinant proteins with a single signal peptide, which should accelerate the development of new, low-cost manufacturing processes.