A Computationally Optimised Structural Integrity Sequence Enhances Vaccine Stability, Yield and Safety Profile

Recombinant vaccines are a cornerstone of global health, exemplified by the eradication of type 3 wild poliovirus in 2020 due to extensive vaccination campaigns. However, sequence similarity between vaccine antigens and human proteins could theoretically risk autoimmune reactions via molecular mimicry. We present the 'Vaccine Candidate de-Risking and Stabilisation' (VaCRiSta) computational pipeline; integrating sequence similarity searches, epitope prediction, homology modeling, and molecular dynamics simulations to de-risk and structurally stabilize vaccine candidates. As a case study, we applied VaCRiSta to the GCN4 trimerization domain in the context of the mumps F protein. The optimized sequence (GCN4_QM) eliminates 139 7-mer and 50 8-mer human proteome matches, is predicted to enhance stability of the trimeric F protein assembly and doubles protein expression yield (2.2 vs. 1.1 mg/L) for a mumps vaccine candidate. GCN4_QM maintains structural integrity, confirmed by negative-stain electron microscopy, and elicits approximately 3-fold higher neutralizing antibody titers in mice (p < 0.0407). GCN4_QM is a useful structural module for protein engineering and multimeric vaccine design, particularly in replacement of a transmembrane region to ensure the solubility of a trimeric protein. Accordingly, the VaCRiSta approach may support vaccine safety, stability and yield, potentially providing benefits for clinical efficacy and delivery to populations.