Simple high-throughput encoding of deep mutational scanning libraries by oligo-based Golden Gate assembly

Abstract

Control over mutational library diversity is an essential consideration when engineering proteins, but is often fraught with trade-offs between diversity, specificity, and affordability. Contemporary library assembly approaches often incorporate oligonucleotide pool synthesis to achieve affordable, precise mutagenesis; however, these oligos are often reliant on complex designs to facilitate downstream PCR and/or restriction digests. Direct hybridisation of oligo pools is an overlooked strategy to simplify mutagenesis, especially when paired with a type IIS restriction cloning approach. We validate this approach by designing, hybridising, and deep sequencing single and dual substitution CDR region parts derived from nanobody GA10. Assembly of these parts into a full-length nanobody CDS facilitated the phage display of variant libraries for affinity maturation against its cyclic peptide target. Variants identified through enrichment analysis were expressed in isolation and yielded improved affinities by more than 100-fold. Recent advances in machine learning have successfully inferred improved variants outside of screened library space, but require controlled, multi-mutant libraries. The library assembly approach outlined in this research is well-suited for such approaches.