Lipid-coated solid and mesoporous silica nanoparticles (LC-SiNPs and LC-MSNs) were developed as modular^platform for the delivery of Toll-like receptor (TLR) agonists. In this study, the incorporation of the amphiphilic TLR4 agonist monophosphoryl lipid A (MPLA) into lipid bilayers of the newly developed platform was systematically optimized. Both core particle types -solid and mesoporous- provided a stable surface structure to enhance membrane stability. By varying cholesterol (15–45%) and anionic lipid (DPPG) (10–30%) content, we
identified key composition parameters that influence TLR4 activation, membrane fluidity and particle-cell interactions. Solid-core MPLA-SiNPs showed enhanced immunostimulatory activity compared to unformulated MPLA when formulated with cholesterol levels increased to 45% and moderate DPPG fractions of 20%, while
mesoporous MPLA-MSNs required higher DPPG content of 30% for comparable activation. To further refine performance, Bayesian optimization (BO) was applied, leading to a significant improvement in MPLA-SiNPs. The BO-optimized MPLA-SiNPs achieved an EC50 value of 87 ng/mL, outperforming classical formulation strategies.
This optimized EC50 was 60 ng/mL lower than for the particles optimized using the classical One-Factor-At-a- Time approach and 280 ng/mL lower compared to unformulated MPLA. Macrophages as antigen-presenting cells carrying TLR4 receptors efficiently internalized and processed the particles. These findings underscore the importance of rational formulation design and demonstrate the potential of lipid-coated silica nanoparticles as modular vaccine carriers. This versatile and modular platform can in future be exploited for delivery of other
TLR agonists and additionally be equipped with antigens.