Optimizing the co-assembly of hydrophobic amino acid-modified hexahistidine and bevacizumab for near-complete inhibition of corneal neovascularization.

Corneal neovascularization (CorNV) is a major cause of blindness, affecting millions globally. Bevacizumab (Beva), an anti-angiogenic antibody, shows promise for treating CorNV due to its specificity and low side effects but is limited by the corneal tissue barrier, requiring invasive delivery methods. This has spurred significant research into modulating nanoparticles (NPs) physicochemical properties for improved corneal penetration. Herein, we proposed an innovative strategy involving the co-assembly of hexahistidine modified with various hydrophobic amino acids (valine (V), tryptophan (W), tyrosine (Y), and isoleucine (I)) with Beva to form XHmA@Beva NPs (VHmA@Beva, WHmA@Beva, YHmA@Beva, and IHmA@Beva), aiming for non-invasive, efficient delivery of Beva to treat CorNV. In vitro, all XHmA@Beva NPs demonstrated the ability to inhibit tubulogenesis and cell migration, with the inhibitory effects ranked as follows: IHmA@Beva > YHmA@Beva > WHmA@Beva ≈ VHmA@Beva. In an alkali-burned rat model of CorNV, VHmA@Beva failed to inhibit neovascularization, weak inhibitory effect for WHmA@Beva. Notably, IHmA@Beva and YHmA@Beva, due to their prolonged ocular retention and improved corneal penetration, demonstrated significant inhibition of neovascularization, with IHmA@Beva achieving near-complete inhibition of CorNV. These findings underscore the crucial role of hydrophobic amino acid residues in peptide NPs assemblies for drug delivery across biological barriers, providing a strategy for noninvasive treatment of CorNV. STATEMENT OF SIGNIFICANCE: Corneal neovascularization (CorNV) is a leading cause of anterior segment blindness, affecting millions of patients worldwide each year. Although anti-VEGF antibody drugs such as bevacizumab (Beva) are highly effective in inhibiting neovascularization, noninvasive delivery remains a major challenge due to multiple barriers on the ocular surface. Current noninvasive delivery strategies achieve only about 90% inhibition, which is insufficient for clinical translation. In this study, we developed a noninvasive delivery approach based on Beva-peptide-zinc ion assemblies, by tuning the hydrophobic amino acids at the peptide termini to produce a series of XHmA@Beva particles. Among them, IHmA@Beva almost completely inhibited CorNV in a rat model of alkali-induced corneal injury, representing the best reported outcome for noninvasive delivery of anti-VEGF antibodies to date. This work not only overcomes the efficiency limitations of noninvasive antibody delivery to the ocular surface but also demonstrates the potential of IHmA@Beva as a safe and effective strategy for clinical translation, offering a therapeutic approach for CorNV.