Efficient full-length IgG secretion and sorting from single yeast clones in droplet picoreactors.

The search for new antibodies is a major field of pharmaceutical research that remains lengthy and costly due to the need for successive library screenings. Existing and antibody discovery processes require that libraries are repeatedly subcloned to switch the antibody format or the secretory host, a resource-intensive process. There is an urgent need for an antibody identification platform capable of screening large antibody libraries in their final soluble format. Previous attempts to develop such a platform have struggled to combine large antibody libraries with screening of high specificity, while retaining sufficient library diversity coverage (ability to detect rare events). Here, we describe a new antibody screening platform based on the encapsulation of antibody secreting yeast cells into picoreactor droplets. We developed and optimized a yeast strain capable of growing and secreting full-length human IgGs in picoreactors, and applied a microfluidics-based high-throughput screening approach to sort and recover target-specific antibody-secreting yeasts. Critically, the direct recovery of secretory yeasts allows for downstream screening and antibody characterization, without the need to reformat or subclone the coding sequences. We successfully increased the diversity coverage of sorting the antibody library without compromising sorting specificity by developing a new fluorescence signal processing methodology. By combining this drastically enhanced sorting efficiency with the high-throughput capability of droplet microfluidics, and the rapid growth of , our new platform is capable of screening millions of antibodies per day and enriching for target-specific ones in 4 days. This platform will enable the efficient screening of antibody libraries in a variety of contexts, including primary screening of synthetic libraries, affinity maturation, and identification of multi-specific or cross-reactive antibodies.