Item

Abstract

De novo protein design has evolved into a powerful tool capable of providing a new generation of
potential therapeutics. We have recently designed a novel protein (Boskar4) to act as a granulocyte-
colony stimulating factor receptor (G-CSFR) agonist. Unlike the native receptor ligand (G-CSF), Boskar4
is small, highly stable, and can be produced with high yields in Escherichia coli. While Boskar4 showed
nanomolar activity in cell-based assays and in differentiating neutrophils ex vivo, and in animal models,
it was still less active than recombinant human G-CSF (rhG-CSF). This could be in part attributed to the
lower affinity of Boskar4 to the G-CSFR compared to the native ligand. We therefore developed a
pipeline, with Boskar4 as an initial candidate, that is able to quickly and efficiently affinity-maturate de
novo designed proteins. In essence, a bacterial display is used in combination with computationally-
designed libraries, utilizing a new design software Damietta. A set of 19 better binding monomeric
Boskar4 variants were identified with up to 100-fold higher affinity. Nine of them were further
investigated as tandem fusions in a cell-based assay, from which the most active variant nearly reached
the activity of rhG-CSF itself. Additionally, high affinity variants of Boskar4 monomers can be used as
competitive antagonists as well as the bases for new protein design applications like Novokines; novel-function cytokines.