Applying computational protein design to engineer affibodies for affinity-controlled delivery of vascular endothelial growth factor and platelet-derived growth factor

Abstract

Effective angiogenesis requires the coordinated presentation of vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF), which play competing roles in stimulating vascular outgrowth and stabilization. Dysregulation in VEGF and PDGF secretion are implicated in abnormal vascular structures, poor vessel stability, and inadequate tissue repair. Current biomaterial delivery vehicles for these proteins have a limited ability to precisely control the kinetics of protein release, preventing systematic exploration of their temporal effects. Here, we combined yeast surface display and computational protein design to engineer eight VEGF-specific and PDGF-specific protein binders called affibodies with a broad range of affinities (dissociation constants = 2.23-9260 nM) for affinity-controlled protein release. Starting from one VEGF- and one PDGF-specific affibody discovered from yeast surface display, computational modeling was used to select disruptive mutations to create VEGF-specific affibodies with lower affinities for VEGF and expand the affinity range of PDGF-specific affibodies. In both cases, the specificity of the affibodies for their target protein was either maintained or enhanced. Soluble protein-specific affibodies modulated protein bioactivity as evidenced by changes in VEGF-induced endothelial cell proliferation and luminescent output of a PDGF-responsive fibroblast cell line. Affibody-conjugated polyethylene glycol maleimide (PEG-mal) hydrogels sustained VEGF and PDGF release compared to hydrogels without affibodies, enabling tunability of protein release over 7 days. VEGF and PDGF released from affibody-conjugated hydrogels exhibited higher bioactivity compared to proteins released from hydrogels without affibodies, highlighting the potential of these engineered affinity interactions to prolong protein bioactivity. This work underscores the power of computational protein design to enhance biomaterial functionality, creating a platform for tunable protein delivery to support angiogenesis and tissue repair.