Accurate phenotype-to-genotype mapping of high-diversity yeast libraries by heat-shock-electroporation (HEEL).

Abstract

High-throughput DNA transformation techniques are invaluable when generating high-diversity mutant libraries, a cornerstone of successful protein engineering. However, transformation efficiencies have a direct correlation with the probability of introducing multiple DNA molecules into each cell, although reliable library screenings require cells that contain a single unique genotype. Thus, transformation methods that yield a high multiplicity of transformations are unsuitable for high-diversity library screenings. Here, we describe an innovative yeast library transformation method that is both simple and highly efficient. Our dual heat-shock and electroporation approach (HEEL) creates high-quality DNA libraries by increasing the fraction of mono-transformed yeast cells from 20% to over 70% of all transformed cells, thus allowing for near-perfect phenotype-to-genotype associations. HEEL also allows more than 10⁷ yeast cells per reaction to be transformed with a circular plasmid molecule, which corresponds to an almost 100-fold improvement compared with current yeast transformation methods. To further refine our library screening approach, we integrated an automated yeast genotyping workflow with a dual-barcode design that employs both a single nucleotide polymorphism and a high-diversity region. This design allows for robust identification and quantification of unique genotypes within a heterogeneous population using standard Sanger sequencing. Our findings demonstrate that the longstanding trade-off between the size and quality of transformed yeast libraries can be overcome. By employing the HEEL method, large DNA libraries can be transformed into yeast with high-efficiency, while maintaining high library quality, essential for successful mutant screenings. This advancement holds significant promise for the fields of molecular biology and protein engineering.IMPORTANCEWith the recent expansion of artificial intelligence in the field of synthetic biology, there has never been a greater need for high-quality data and reliable measurements of phenotype-to-genotype relationships. However, one major obstacle to creating accurate computer-based models is the current abundance of low-quality phenotypic measurements originating from numerous high-throughput but low-resolution assays. Rather than increasing the quantity of measurements, new studies should aim to generate as accurate measurements as possible. The HEEL methodology presented here aims to address this issue by minimizing the problem of multi-plasmid uptake during high-throughput yeast DNA transformations, which leads to the creation of heterogeneous cellular genotypes. HEEL should enable highly accurate phenotype-to-genotype measurements going forward, which could be used to construct better computer-based models.