Background: A new 32-valent M protein-based Strep A vaccine was designed using a three-tiered structure-based algorithm to identify M peptides most likely to generate antibodies cross-reactive with non-vaccine M types to improve overall predicted vaccine coverage. The aim of the current study was to assess predicted vs. observed M peptide cross-reactivity after immunization of rabbits with the new 32-valent vaccine.
Methods: Immune sera obtained from 3 immunized rabbits were screened for antibody levels against a total of 127 50aa N-terminal M peptides. Direct ELISA and ELISA inhibition assays were performed to assess levels of vaccine-specific and cross-reactive antibodies in relation to the structure-based predictions used to design the vaccine.
Results: The immune sera contained significant levels of antibodies (>8-fold increase above pre-immune) against 31/32 vaccine M peptides. Importantly, the immune sera also contained significant levels of antibodies (>8-fold) against 28/37 (76%) non-vaccine peptides predicted to cross-react with at least 1 vaccine peptide. Of the 28 non-vaccine M peptides that showed cross-reactivity with the vaccine antisera, ELISA inhibition assays indicated that 18 peptides shared epitopes with the predicted cross-reactive vaccine peptide, suggesting that some cross-reactive antibodies were elicited by other vaccine peptides. Of the 58 remaining non-vaccine peptides that were not predicted to cross-react with vaccine antisera, 8 (14%) resulted in cross-reactive antibody binding.
Conclusion: Structure-based algorithms may be used to design broadly cross-reactive multivalent M peptide vaccines.