Computationally identified peptides immobilized to hydrogel delivery vehicles enable affinity-controlled release of proteins

The administration of therapeutic proteins, such as growth factors, can be accomplished through affinity-based release systems, yet identifying suitable binding partners that achieve high protein loading and desirable release kinetics remains a significant challenge. In this work, we used a computationally guided approach to identify binding peptides for therapeutic proteins, significantly increasing the loading capacity of the functionalized hydrogel delivery platform. Using crystal structure data from the Protein Data Bank and Rosetta-based rational design, we extracted a 10 amino acid peptide segment from the heavy chain of bevacizumab to modulate the release of vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF). Biolayer interferometry confirmed moderate binding affinity to VEGF (K_D 10^-5 M) and higher affinity for PDGF (K_D 10^-7 M). Mutational analysis revealed that aromatic residues were essential for binding. When covalently immobilized to a hyaluronan-methylcellulose hydrogel, the peptide enabled prolonged and sequential release of VEGF and PDGF. Computational modeling of the release kinetics using finite element analysis agreed with experimental data, validating the model and guiding the design of the delivery system. Importantly, the released proteins remained bioactive as demonstrated by cell growth and angiogenesis tube formation assays. This study demonstrates a simple and effective approach to achieve controlled and sequential release of therapeutic growth factors. STATEMENT OF SIGNIFICANCE: Proteins are promising therapeutics for multiple diseases; however, their delivery remains challenging due to the use of generic strategies, which often degrade the proteins. We used computational tools to identify specific binders for two growth factors (VEGF and PDGF), and then demonstrated the utility of these binders to control the release of bioactive proteins for a sustained period from a novel hydrogel. We achieved sequential release of VEGF followed by PDGF, which is important in vessel formation. This new system of identifying protein binding partners is broadly applicable and shown herein to be useful to control the release of two specific proteins. This combination of theoretical and experimental approaches is of broad interest to biomaterial researchers.