W18 - Modulating Antigen Processing through Metal-Organic Framework Vaccines to Bias Adaptive Immunity

Introduction: : Protein-based vaccines function by delivering protein antigen that can be processed into various peptide fragments by antigen-presenting cells (APCs) to stimulate and raise specific adaptive immune responses. However, free native protein antigens are inherently unstable—they are susceptible to degradation both in circulation and within endolysosomal compartments of APCs, where they are rapidly digested by proteases. While genetic engineering or chemical modification increases protein stability, this method is inherently protein specific, requires research for each newly identified antigen, and can lead to critical modifications of immunogenic amino acids and loss of function. To address this, biomaterial-based strategies such as lipid and polymeric nanoparticles have emerged to protect and deliver protein antigens more efficiently. Despite improved delivery, current nanoparticle systems still face challenges in controlling intracellular antigen release and processing rate due to their own instability, incompatible solvent conditions for loading proteins without employing genetic modifications to the protein, and a lack of generalizable strategies across various proteins to stabilize them in conditions of cellular stress. Importantly, the rate of antigen release within APCs is increasingly recognized as a key factor in determining immune outcome—especially the balance of CD4+ versus CD8+ T cell activation, and TH1 versus TH2 polarization. However, most studies have focused on release of short antigenic peptides due to the limitations of current vaccine systems. Moreover, few have enabled tunable delivery of native, unmodified protein which does require genetic or chemical modification. In this research, we employ a modular metal-organic framework (MOF) platform to overcome these limitations.

Materials and

Methods: : NU-1000 or NU-1003 MOF nanoparticles were synthesized at 100-200 nm sizes following previously published protocols and post-synthetically loaded with native ovalbumin or receptor binding domain (RBD) protein antigen. To encapsulate the proteins into the MOFs, 0.25 mg of MOF was resuspended in Milli-Q water and sonicated before being incubated in a 100 µM protein solution for 24 h at 25 °C. After loading, samples were centrifuged and the supernatant was saved for quantification. Loaded MOFs were used to investigate how pore size affects antigen release rate, antigen processing in vitro, and antigen-specific T cell responses ex vivo using flow cytometry and ELISAs. Long-term immune responses were investigated in vivo using C57BL/6 wild-type mice following multiple subcutaneous immunizations and collection of serum via retroorbital blood draw. Analyses measuring adaptive cellular memory and humoral antibody buildup were performed with flow cytometry and ELISAs.

Results, Conclusions, and Discussions:: Using two biocompatible zirconium-based nano-sized MOFs in the NU-100x series with differing pore sizes—NU-1000 (3.3 nm) and NU-1003 (4.7 nm)—we investigate how solely modulating pore structure through size alters protein-MOF interactions, antigen release kinetics, and downstream immune processing. These MOFs allow for post-synthetic loading of large amounts of native protein antigen (e.g., ~1.4 mg/mg of ovalbumin) without requiring harsh solvents during loading or modifications to the protein. We observed that altering MOF pore size changed the electrostatic interactions and thus retention of protein antigens, producing distinct antigen release profiles. These differences had a significant impact on antigen processing and MHC-I presentation in bone marrow-derived dendritic cells (BMDCs), but more excitingly, propagated distinct biases in raised antigen-specific CD8+ and CD4+ T cells. Ova loaded into NU-1003 (Ova@NU-1003) induced a 3-fold higher CD8+:CD4+ T cell proliferation ratio and promoted a TH1-skewed cytokine profile (producing 2.2-fold more TH1:TH2 cytokines compared to ovalbumin-loaded NU-1000 (Ova@NU-1000). While both MOFs induced ex vivo responses to antigen-specific stimuli, for NU-1000 delivery in vivo there were stronger antigen-specific IgG responses, dependent on CD4+ T cells. Ova@NU-1000 induced ~9-fold greater antigen-specific IgG antibodies and a long-term ~5.9-fold elevation in the IgG1:IgG2a ratio compared to Ova@NU-1003. This indicates the strong humoral raised immunity and bias towards TH2. Importantly, these trends were observed when employing a clinically relevant antigen, the SARS-CoV-2 receptor-binding domain (RBD). Unmodified native RBD was loaded into the MOFs at a near 100% encapsulation efficiency. RBD-loaded NU-1000 induced ~60.5-fold greater IgG1:IgG2a ratio compared to RBD@NU-1003. Overall, this research highlights how vaccine-mediated delivery of native non-modified protein antigens can be rationally tuned via MOF architecture to direct immune polarization. Our findings offer a broadly applicable strategy for designing vaccines with tailored immune outcomes through control over antigen release kinetics.

Acknowledgements and/or References (Optional)::