Enhancing yeast folding capacity by genome editing unlocks quantitative and qualitative improvements in antibody surface display
Abstract Background Yeast surface display is a widely used platform for antibody affinity maturation; however, constraints in the yeast folding and disulfide bond formation machinery can limit correct antibody expression and bias selection outcomes, favoring variants that satisfy display constraints. This limitation might be of particular relevance when selections are based on biophysical features beyond affinity, such as aggregation, polyreactivity, or thermal stability. To overcome these constraints, we engineered the yeast display system by overexpressing key folding chaperones, yeast BiP and human protein disulfide isomerase (PDI), either through co-expression from the antibody display plasmid or via genomic editing, individually and in combination. Results As a proof of concept, surface display of adalimumab was significantly increased upon chaperone co-expression, with the highest improvement observed in strains with genomic integration of PDI, yielding a 2.5-fold increase in display levels. These findings were consistently reproduced across four additional antibodies using three edited strains expressing BiP, PDI, or both. To assess folding quality directly at the cell surface, we implemented two novel complementary staining strategies: with maleimide-Pacific Blue to detect unpaired thiols as a result of incomplete disulfide bond formation, and with Bis-ANS to quantify exposed hydrophobic regions. Both assays revealed a substancial reduction in free thiols and surface hydrophobicity (as only the proper hydrophobic residues are exposed) in the edited strains, consistent improved disulfide bond formation and overall folding quality relative to the parental strain. Accordingly, the PDI-edited yeast strain showed the best overall performance, improving both display quantity and quality across a panel of ten therapeutic antibodies. Enhanced display translated into improved antigen binding without altering the polyreactivity profiles of several candidates, therefore retaining native biophysical characteristics. Conclusions The enhancement of the yeast folding machinery, particularly through genomic integration of PDI, substantially improves both the quantity and quality of antibody surface display. This optimized yeast display platform enables more faithful translation of antibody biophysical features, supporting its application in antibody workflows, including selections based on biophysical properties.