Engineering Peptide Hydrogels to Extend the Release of Protein and Small-Molecule Drugs through Dynamic Covalent Bonding

Introduction:: Pharmaceutical drugs have demonstrated their ability to increase longevity and quality of life. As a result, more than half of the world’s population takes at least one drug every day. However, conventional drug formulations have several key limitations, including the need for frequent administration, which undermines their real-world utility. Self-assembling peptide hydrogels represent one class of injectable controlled-release systems with the ability to sustain drug release from an all-natural and protein drug-compatible depot. Unfortunately, the extension of release, and thus extension of efficacy, is limited when governed by diffusion alone (Figure 1A). To overcome this challenge, we have developed modifications for peptide hydrogels that enable dynamic covalent crosslinking between diols and boronic acids (Figure 1B-C) to extend the duration of in vitro and in vivo drug release for far longer than electrostatic and hydrophobic interactions can. In total, we synthesized and evaluated three different peptide functionalization strategies (catechol, Cat; nitrocatechol, nitroCat; or salicylhydroxamic acid, SHA), two base hydrogels (K₂ and E₂), and three drugs, including two small-molecule boronic acid-containing drugs (Bortezomib, GSK656) and one model IgG antibody functionalized with boronic acid.


Materials and Methods:: Multi-domain peptide hydrogels were synthesized to include catechol (Cat), nitrocatechol (nitroCat), or salicylhydroxamic acid (SHA) groups, all of which can form dynamic covalent bonds with boronic acid. Proper peptide synthesis was confirmed using mass spectrometry and purified using high-performance liquid chromatography. Hydrogels with or without drug were then studied using circular dichroism and rheology to observe changes in hydrogel properties. In vitro release experiments using FITC-phenylboronic acid were performed in phosphate-buffered saline at 37 °C, measuring fluorescence of the supernatant frequently. Naïve BALB/c mice were then injected subcutaneously with bortezomib- of GSK656-loaded K₂ hydrogels with the nitroCat or SHA modifications. Blood from treated animals was collected periodically using a blood card, drug was then extracted from the card using methanol, and the resulting drug concentrations in serum were analyzed using liquid chromatography-mass spectroscopy. A subset of mice were sacrificed early to determine local tissue concentrations, tissue was collected, sectioned, and then imaged using mass spectrometry imaging. Finally, fluorescently labeled IgG was injected into SKH1-elite mice, who are naturally hairless, with or without hydrogel-forming K₂ and E₂ peptides. Mice were then imaged non-invasively using an IVIS and antibody release was determined by observing the fluorescent signal at the injection site decrease over time.


Results, Conclusions, and Discussions:: Modifying K₂ and E₂ peptides to enable dynamic covalent bonding with boronic acid-containing drugs did not meaningfully affect beta sheet formation, mechanical properties, shear thinning, or shear recovery. K₂ hydrogels modified with SHA or nitroCat greatly extended in vitro release of bortezomib compared to unmodified K₂ hydrogels or those modified with Cat (Figure 1D); therefore, the Cat modification was deprioritized. A murine pharmacokinetic study showed that bortezomib concentrations in blood were similar between the soluble injection and K₂ injection, as expected, whereas the SHA and nitroCat modifications could reduce the peak concentration after dosing as much as 8-fold compared to a bolus injection (Figure 1E) or be used to maintain drug levels for up to 7-fold longer with the same peak drug concentration as the soluble dose. The tissue surrounding the SHA-K₂ hydrogel depot also showed substantial bortezomib surrounding the injection site after one week whereas very little drug was present in the soluble injection group at that time (Figure 1F). We then used this drug to release GSK656, a boronic acid-containing drug being explored for tuberculosis. In vivo pharmacokinetic studies showed that SHA-K₂ extended the duration over which GSK656 remained above its effective dose 50 (ED50) from 2 days to 21 days (Figure 1G). Because protein therapeutics—and antibody drugs in particular—are among the most popular new drugs owing to their potency and specificity, we then explored the ability to release macromolecules that do not innately possess boronic acid. We modified IgG with varying numbers of boronic acid by labeling amines on the protein. Using an in vivo imaging system (IVIS) and antibodies further labeled with a fluorophore, we assessed the release of antibodies from the depot and found that SHA-K₂ from approximately 1 day to approximately 12 days (Figure 1H). Lastly, to avoid the inflammation that has been associated with K₂, we modified E₂, a less inflammatory peptide that assembles via the same mechanism, and observed an even greater slowing of release (Figure 1I). These results provide strong evidence for dynamic covalent bonding as a tool for extending drug release from peptide hydrogels.


Acknowledgements (Optional): :

References (Optional): :