Injectable Granular Hydrogels Enable Avidity-Controlled Biotherapeutic Delivery

Abstract

Protein therapeutics represent a rapidly growing class of pharmaceutical agents that hold great promise for the treatment of various diseases such as cancer and autoimmune dysfunction. Conventional systemic delivery approaches, however, result in off-target drug exposure and a short therapeutic half-life, highlighting the need for more localized and controlled delivery. We have developed an affinity-based protein delivery system that uses guest-host complexation between β-cyclodextrin (CD, host) and adamantane (Ad, guest) to enable sustained localized biomolecule presentation. Hydrogels were formed by the copolymerization of methacrylated CD and methacrylated dextran. Extrusion fragmentation of bulk hydrogels yielded shear-thinning and self-healing granular hydrogels (particle diameter = 32.4 ± 16.4 μm) suitable for minimally invasive delivery and with a high host capacity for the retention of guest-modified proteins. Bovine serum albumin (BSA) was controllably conjugated to Ad via EDC chemistry without affecting the affinity of the Ad moiety for CD (K_D = 12.0 ± 1.81 μM; isothermal titration calorimetry). The avidity of Ad-BSA conjugates was directly tunable through the number of guest groups attached, resulting in a fourfold increase in the complex half-life (t_1/2 = 5.07 ± 1.23 h, surface plasmon resonance) that enabled a fivefold reduction in protein release at 28 days. Furthermore, we demonstrated that the conjugation of Ad to immunomodulatory cytokines (IL-4, IL-10, and IFNγ) did not detrimentally affect cytokine bioactivity and enabled their sustained release. Our strategy of avidity-controlled delivery of protein-based therapeutics is a promising approach for the sustained local presentation of protein therapeutics and can be applied to numerous biomedical applications.

Keywords: avidity; bioconjugation; cyclodextrin; hydrogel; sustained release.

Figure 1

Synthesis and characterization of DexMA and MeCD. (A) Schematic of methacrylated cyclodextrin (MeCD, top, purple) and methacrylated dextran (DexMA, bottom, blue) synthesis through esterification with GMA. (B) DS and yield dependence on the molar feed ratio of GMA to glucose repeat units. (C) Hydrogels are formed through copolymerization of MeCD and DexMA. Representative oscillatory time sweeps (1 Hz, 1.0% strain) of photopolymerization (10 mW/cm², 5 min as indicated by the shaded area) showing shear moduli (G′) of MeCD alone (D), DexMA alone (E), and their combination (F).

Figure 2

Granular hydrogel formation and characterization. (A) Schematic representation of microgel fabrication by EF. (B,C) Particle diameter throughout the extrusion process of 5%_w/v DexMA gels (mean ± SD; n = 200 particles), quantified using fluorescence microscopy images (C, scale bar = 100 μm). Representative particles are outlined (yellow) for clarity. (D) Final particle diameter of 5%_w/v DexMA and 5%_w/v DexMA +10%_w/v MeCD microgels (mean ± SD; n = 200; ns = not significant). (E) Continuous flow experiments showing the shear stress and viscosity of 5%_w/v DexMA +10%_w/v MeCD granular hydrogels. (F) Cyclic deformation at low (0.5%) and high (500%) strain (1.0 Hz) of 5%_w/v DexMA + 10%_w/v MeCD hydrogels; G′ (storage modulus, dark purple, circle), G″ (loss modulus, light purple, circle). (G) Representative images of granular hydrogel injection (30G needle, 1 mL syringe; scale bar = 5 mm).

Figure 3

Chemical modification of BSA. (A) Schematic of BSA modification with FAM and/or Ad–PEG–amine via EDC-catalyzed amidation. (B) FAM–BSA excitation and emission scans (λ_max ex/em = 480/525 nm, left). Dependence of the FAM-per-BSA modification ratio on the molar feed ratio of FAM to BSA (right). Differences between all reaction conditions were highly significant, ****P < 0.0001. (C) The affinity-based thermodynamic dissociation constant of individual GH interactions (K_D, bottom left) and extent of Ad–BSA modification (N, bottom right), determined by ITC (see Figures S5 and S6). (D) Avidity-controlled Ad–BSA complex half-life (bottom left) and dissociation rate constant (bottom right), determined by SPR. Data represent mean ± SD; n = 3; ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ANOVA, Tukey HSD.

Figure 4

Model biomolecule release. (A) Schematic of biomolecule retention within the granular host hydrogels as a result of avidity-based interaction. (B) Cumulative release of Ad-FAM-BSA (0–10 equiv Ad); n = 4. (C) Controlled release of multiple components from the same hydrogel, including BSA-AlexaFluor555 (0 equiv Ad), Ad–BSA–fluorescein (2.5 equiv Ad), and Ad–BSA-Pacific Blue (5 equiv Ad); n = 6. Data represent the mean ± SD; **P < 0.01, ***P < 0.001, ****P < 0.0001; RM one-way ANOVA, Tukey HSD.

Figure 5

Cytokine functionality and release profile. Heat maps of gene expression levels in bone marrow-derived macrophages (BMDMs) after treatment with modified and unmodified cytokines (A) or final hydrogels at the end of 14 day release (B). Data represent mean ± SD; n ≥ 1; ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; two-way ANOVA, Fisher LSD. Cytokine release profiles of unmodified and Ad-modified (5 equiv) IL-10 (C), IFNγ (D), and IL-4 (E) from DexMA + MeCD hydrogels into media. Data represent the mean ± SD; n ≥ 3; ****P < 0.0001; RM two-way ANOVA.