Deep mutational scanning-guided design of a high-affinity helix–loop–helix peptide targeting G-CSF receptor

Abstract

At present, mid-sized binding peptides have emerged as a new class of drug modalities. We have de novo designed a helix–loop–helix (HLH) peptide (MW: ∼4,500), constructed phage-displayed libraries, and screened the libraries against a variety of disease-related proteins to successfully obtain molecular-targeting HLH peptides. The next essential step in developing HLH peptides into therapeutics involves affinity engineering to optimize binding affinity and specificity. Here, we demonstrate deep mutational scanning to improve binding affinity over 1000-fold for an HLH peptide (P8-2KA; K_D = 380 nM) targeting granulocyte colony-stimulation factor receptor (G-CSFR). Site-saturation mutagenesis on the two helices was performed to produce a phage-displayed library that was screened against G-CSFR. The DNA sequences of mutants from the unselected and selected phage libraries were analyzed with next-generation sequencing. The enrichment ratio of each mutant was calculated from the sequencing data to identify beneficial mutations for G-CSFR binding. Grafting of the five beneficial mutations on P8-2KA dramatically increased the binding affinity (K_D = 16 nM), while cyclization of the HLH peptide with an intramolecular disulfide bond further increased binding affinity for G-CSFR (K_D = 0.18 nM). The combined strategy of phage-displayed library selection and deep mutational scanning-guided design generated high-affinity HLH peptides, emphasizing the potential of molecular-targeting HLH peptides as a new drug modality that serves as an alternative to antibodies.