γ-Cyclodextrin hydrogel for the sustained release of josamycin for potential ocular application

Abstract

Glaucoma is the leading cause of blindness worldwide. However, its surgical treatment, in particular via trabeculectomy, can be complicated by fibrosis. In current clinical practice, application of the drug, Mitomycin C, prevents or delays fibrosis, but can lead to additional side effects, such as bleb leakage and hypotony. Previous in silico drug screening and in vitro testing has identified the known antibiotic, josamycin, as a possible alternative antifibrotic medication with potentially fewer side effects. However, a suitable ocular delivery mechanism for the hydrophobic drug to the surgical site does not yet exist. Therefore, the focus of this paper is the development of an implantable drug delivery system for sustained delivery of josamycin after glaucoma surgery based on crosslinked γ-cyclodextrin. γ-Cyclodextrin is a commonly used solubilizer which was shown to complex with josamycin, drastically increasing the drug's solubility in aqueous solutions. A simple γ-cyclodextrin crosslinking method produced biocompatible hydrogels well-suited for implantation. The crosslinked γ - cyclodextrin retained the ability to form complexes with josamycin, resulting in a 4-fold higher drug loading efficiency when compared to linear dextran hydrogels, and prolonged drug release over 4 days.

Keywords: drug delivery; fibrosis; glaucoma; hydrogel; josamycin; γ-cyclodextrin.

Graphical abstract

Figure 1.

Characterization of γCD-JM complex formation and JM stability in aqueous solutions. (A) Phase solubility curve of josamycin in aqueous solution with γCD. (B) JM stability in aqueous solutions at 4 °C (blue points), 25 °C (black points) and 37 °C (red points) was measured in γCD (solid lines) and PBS only (dashed lines) solutions.

Figure 2.

Physical characterization of crosslinked γCD hydrogel. (A) Macroscopic images of γCD gels show that the gel is transparent in both the dry and hydrated state. Significant swelling of the hydrogel is evident in the size change after hydration. (B) The SEM image of the interior of the freeze-dried γCD gel shows a consistent porous structure. C) The γ-CD gels were highly swellable with an average swelling ratio of 4.33 ± 0.63. Overlaid points represent individual replicates and are color coded by hydrogel batch

Figure 3.

Josamycin loading improved in cyclodextrin hydrogels. (A) The total mass of JM loaded into one 9 mm γCD gel disk was significantly higher than the total loaded drug in one dextran gel disk. Overlaid points represent individual replicates and are color coded by hydrogel batch. (B) Periodic measurement of the JM concentration in the loading solution during drug loading shows a significant drop in JM concentration for the γCD, but not for the dextran gels. This drop in concentration indicates that JM is interacting with the γCDs in the gel, pulling the JM out of solution and explains the higher total loading. (C) To account for differences in swelling when comparing drug loading, the local concentration of JM was calculated by normalizing the total loaded drug to the final volume of the hydrated gel. This demonstrates that the dextran gel had a local concentration close to the aqueous solubility of JM (intrinsic solubility of JM indicated by dotted line), while the γCD gel had a final concentration 4 times higher. This additionally indicates that JM was forming complexes with the crosslinked γCD and therefore the gel can support JM concentrations much higher than the drug’s intrinsic solubility. * indicates a significant difference with p < 0.05.

Figure 4.

Prolonged drug release from γCD hydrogels. (A) Cumulative in vitro drug release of JM shown as a percent of the total released over the course of the experiment. γCD gels showed significantly prolonged release profiles. (B) The amount of JM released at each time point also clearly demonstrates a prolonged release from the γCD gel in comparison to dextran gels, which do not release a measureable amount of JM past 6 h. The inset graph shows a magnified view of the data from 0.5 to 6 h. * indicates time points where the groups are significantly different (p > 0.05)

Figure 5.

γCD Hydrogels demonstrated good biocompatibility. (A) L929 cells were cultured in eluent medium from γCD gels and showed no loss of viability compared to cells grown in normal medium. (B) γCD gels do not support direct cell attachment and L929 cells seeded directly on the gels remain rounded instead of spreading out when seeding on the surface of the γCD gel (left image). However, cells that were seeded on the plate next to the gels grew up to and under the gel, indicating that direct contact is tolerated (right image). The edge of the gel can be seen running vertically through the middle of the image.