Modular mRNA LNP design integrates RNA, lipid, and antigen engineering for protective vaccination

Abstract

mRNA-lipid nanoparticle (LNP) vaccines are programmable, multi-component systems in which immune outcomes emerge from coupled control of nanoparticle chemistry, RNA regulatory architecture, and antigen design. Here we establish an integrated engineering framework that quantitatively maps how ionizable lipid identity, untranslated region (UTR) configuration, and 5' cap structure shape innate activation landscapes and thereby tune the magnitude, cellular distribution, and polarization of adaptive immunity. Benchmarking three ionizable lipids shows that lipid chemistry imprints distinct cytokine and chemokine milieus during dendritic cell-T cell priming that mirror downstream T cell activation phenotypes, identifying lipid structure as a determinant of pathway-selective activation; notably, our first-in-study lipid exhibits benchmark-comparable immunostimulatory profiles, supporting further translational evaluation. Using Crimean-Congo hemorrhagic fever virus (CCHFV) as a model high-consequence pathogen, we show that UTRs act as modular regulatory elements that redirect cytokine outputs, tuning effector versus proliferative programs and expanding helper polarization in an antigen-dependent manner. Cap structure functions primarily as a quantitative gain control, scaling cellular and humoral response magnitude without overriding antigen-defined polarization. Within this optimized platform, antigen architecture defines functional constraints on protection: structural modifications reshape immune hierarchies and antibody quality. Integrating these design axes yields an optimized mRNA-LNP vaccine encoding the CCHFV secreted glycoprotein complex (sGCs) that achieves 90% protection in an immunosuppressed murine lethal-challenge model with minimal clinical signs. Together, these data define generalizable design principles for rational, multi-parameter optimization of mRNA-LNP vaccines.