Designing a Novel di-epitope Diphtheria Vaccine: A Rational Structural Immunoinformatics Approach

The design of an epitope-based vaccine against diphtheria toxin (DTx) originated from the idea that many strong binder epitopes may be structurally located in the depth of DTx. Subsequently, many ineffective antibodies may be produced by the presentation of those epitopes to T and B lymphocytes. The other critical issue is the population coverage of a vaccine that has been neglected in traditional vaccines.

Given the issues above, our study aimed to design a peptide-based diphtheria vaccine, considering the issues of unwanted epitopes and population coverage.

The frequencies of pre-determined HLA alleles were listed. A country in which almost all HLA alleles had been determined in almost all geographical distribution was selected. The epitopes within the sequence of diphtheria toxin were predicted by the NetMHCIIPan server based on the selected HLA alleles. Strong binder epitopes on the surface of diphtheria toxin were selected by structural epitope mapping. The epitopes, which cover almost all the human population for each of the HLA alleles in the candidate country, were then selected as epitopebased vaccines.

At first, 793 strong binder epitopes were predicted, of which 82 were surface epitopes. Nine surface epitopes whose amino acids had extruding side chains were selected. Finally, 2 epitopes had the most population coverage and were suggested as a di-epitope diphtheria vaccine. The population coverage of the di-epitope vaccine in France and the world was 100 and 99.24 %, respectively. HLA-DP had the most roles in epitope presentation.

Our results indicated that 97.75% of unwanted antibodies (791 epitopes) have been reduced. Achieving two immunodominant surface epitopes confirmed our rational filtration strategy for sequential reduction of unwanted epitopes. Our novel insight may pave a new way to designing novel peptide-based vaccines to avoid producing non-specific antibodies.