Files
Abstract
This dissertation explores a strategy to enhance antitumor immunity by delivering xenoantigens, non-self antigens readily recognized by the host immune system, directly into the tumor microenvironment, with the goal of locally triggering immune activation and promoting epitope spreading to endogenous tumor antigens. Using the poorly immunogenic B16F10 melanoma model, we tested both protein-based and mRNA lipid nanoparticle-based delivery platforms for local intratumoral xenoantigen delivery, either alone or in combination with αPD1 checkpoint blockade. Protein-based delivery of xenoantigens, including tumor-binding antibody fusion proteins and tumor-binding polymer conjugates, failed to generate significant antitumor responses, even with prior vaccination, likely due to insufficient immunogenicity in the absence of adjuvants. In contrast, intratumoral delivery of mRNA LNPs encoding for membrane-bound xenoantigens generated robust local immune activation, with evidence of increased immune infiltration and activation within tumor-draining lymph nodes. However, repeated intratumoral delivery paradoxically led to a dramatic collapse in intratumoral immune cell populations by the third injection, culminating in impaired antitumor immunity upon secondary tumor rechallenge. Notably, prior vaccination with mRNA did not improve therapeutic outcomes, and in some cases, appeared to hinder the development of durable immune responses. These findings suggest that while local xenoantigen delivery can transiently stimulate immune activation, its efficacy may be limited by the immunosuppressive tumor microenvironment and potential overactivation-induced immune dysfunction. The work has potential implications for the design of mRNA-based cancer vaccines, oncolytic virus therapies, and CAR-T approaches that rely on robust, localized antigen recognition, and highlights the importance of tuning immune activation to achieve productive and lasting antitumor responses.