Integrating molecular dynamics simulations to enable rational assembly of immune signals for immunotherapy

Despite the importance of biophysical cues in tuning the immune response, the connections between these cues and immunological outcomes are poorly understood in the context of immunotherapies. To study these connections, our lab designed therapeutic complexes that are self-assembled from peptide antigens modified with cationic amino acid residues and anionic, nucleic acid-based modulatory cues. We utilized the self-assembly platform as a tool to understand how tuning the biophysical properties of immune signals impacts molecular interactions during self-assembly. Here, we implemented molecular dynamics simulations as a tool to study how molecular interactions between cationic peptides and anionic modulatory cues change as a function of peptide design. Using temperature replica exchange molecular dynamics, we compare molecular contacts - including hydrogen bonding and salt bridges - across a library of peptide sequences that are mistakenly attacked during autoimmune disease. We show that peptides with higher cationic charge and peptides anchored with arginine residues form more electrostatic interactions during self-assembly than peptides with lower cationic charge and peptides anchored with lysine residues, respectively. Surface plasmon resonance studies revealed that in addition to the type of anchored amino acid residue, the distribution of charge across the peptide also impacts the binding affinity of self-assembled immune cues. In vitro primary cell studies using these same antigen designs revealed signaling that was likewise sensitive to the total charge, charge distribution, and type of anchored amino acid residues within the therapeutic complexes. Taken together, these insights help intuit how to modify biophysical cues to self-assemble a range of peptide antigens for distinct disease targets. This granular understanding of nanomaterial-immune interactions contributes to more rational immunotherapy design.