Dual affibody-conjugated hydrogels tune the co-delivery and activity of immunomodulatory and osteogenic proteins

Introduction:: Bone regeneration requires the well-coordinated presentation of many functionally different proteins. However, clinical treatments deliver the potent osteogenic protein bone morphogenetic protein-2 (BMP-2) alone within a scaffold. Recent studies have demonstrated that combining BMP-2 with other proteins improves bone healing outcomes. In this study, we assess the combination of BMP-2 with the cytokine interleukin-4 (IL-4), which has been shown to polarize M0 macrophages into M2 (pro-therapeutic) phenotypes while also inhibiting osteoclast activity. Traditional delivery vehicles provide limited control over individual protein release kinetics, and thus cannot accurately mimic the healing cascade. Alternatively, affinity-based delivery vehicles present a promising approach to independently control delivery rates of multiple proteins, dictated by the binding strength between the delivery vehicle and each specific protein. The goal of this study was to sequentially co-deliver IL-4 and BMP-2 to target different stages of bone healing, using hydrogels with multiple affinity interactions specific to each protein.

Materials and Methods:: A yeast library displaying protein binders called affibodies was subjected to magnetic- and fluorescence-activated cell sorting to identify affibodies binding specifically to BMP-2 or IL-4. Affibody sequences specific to BMP-2 or IL-4 were modified with C-terminal hexahistidine tag and cysteine and were transformed into BL21 E.coli for soluble affibody expression, collection, and purification. Size, secondary structure, and dissociation constant of the binding interactions of soluble affibodies were characterized using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), circular dichroism, and biolayer interferometry, respectively. Affibody structure and binding to target proteins were predicted using the computational tools AlphaFold-2, Rosetta, and ZDock. To assess the impact of affibody binding on BMP-2’s osteogenic activity, we cultured C2C12 mouse skeletal myoblasts (20,000 cells/well) for 72 hours with 20 nM BMP-2 and/or 20 nM BMP-2-specific affibody. A colorimetric alkaline phosphatase (ALP) assay was used to quantify early osteogenic differentiation of C2C12s. The effects of affibody-binding on IL-4’s immunomodulatory activity was assessed by culturing THP-1 monocytes (10,000 cells/well) with phorbol 12-myristate 13-acetate for 48 hr to mature them into M0 macrophages, followed by 72 hr incubation with 20 nM IL-4 and/or 2 µM IL-4-specific affibody. Enzyme-linked immunosorbent assays (ELISA) for tumor necrosis factor-α (TNF-α) were used to quantify M1 macrophage polarization levels. Polyethylene glycol-maleimide (PEG-Mal) hydrogels containing high- or low-affinity BMP-2 and/or high- or low-affinity IL-4 affibodies were synthesized using 4-arm PEG-Mal, affibodies, and dithiothreitol, and loaded with 50 ng BMP-2 and 50 ng IL-4. Protein encapsulation and release over 1 week was quantified using protein-specific ELISAs.


Results, Conclusions, and Discussions:: Eleven BMP-2-specific and three IL-4-specific affibodies were identified using yeast surface display. Two affibodies with distinctly different dissociation constants for their target proteins were selected for each protein, and will be referred to as either high-affinity or low-affinity binders. SDS-PAGE confirmed the affibodies were approximately 7.4 kDa per their theoretical size, and circular dichroism spectra indicated the affibodies were alpha-helical as expected. Computational predictions suggested that the high-affinity BMP-2 affibody binds to the “wrist” epitope of BMP-2, while the low-affinity affibody binds to the “knuckle” epitope, which are mammalian BMP receptor binding sites, suggesting that affibodies may inhibit BMP-2’s osteogenic activity. C2C12 ALP activity was significantly reduced in BMP-2+affibody groups compared to BMP-2-only controls, supporting the binding location predictions. IL-4 affibodies were predicted to bind at different locations on IL-4, with only the high-affinity affibody predicted to bind at an IL-4 receptor binding site. Preliminary results indicated that TNF-α levels of THP-1s were increased in the IL-4+affibody groups compared to the IL-4 only control, suggesting that the affibodies decreased IL-4-mediated M2 macrophage polarization. However, affibodies alone increased TNF-α secretion, suggesting that dosage of IL-4-specific affibodies need to be refined. Conjugation of BMP-2- or IL-4-specific affibodies into PEG-Mal hydrogels increased encapsulation efficiency of their respective proteins. Hydrogels containing high-affinity BMP-2 affibodies slowed the release of BMP-2 compared to hydrogels without BMP-2-specific affibodies. A similar trend was observed with IL-4-specific affibodies. Combining IL-4- and BMP-2-specific affibodies into a single hydrogel also decreased release of each respective protein, similar to the use of IL-4 or BMP-2 affibodies alone, suggesting that protein-affibody binding interactions were independent.


Future work will involve studying the effect of affibodies on time-dependent macrophage polarization using THP-1 cells in vitro. We also plan to test the ability of this system to effectively deliver proteins to enhance bone repair in vivo in a critically-sized, femoral defect in a rat. Ultimately, affibody-containing hydrogels may be a promising approach to deliver multiple proteins with precise temporal control, leading to improved healing outcomes in severe bone injuries.


Acknowledgements (Optional): :

References (Optional): :