Acute myeloid leukemia (AML) is a leading blood cancer subtype that can be caused by 27 gene mutations. Previous studies have explored potential vaccine and drug treatments against AML, but many were proven immunologically insignificant. Here, we targeted this issue and applied various clinical filters to improve immune response. KIT is an oncogenic gene that can cause AML when mutated and is predicted to be a promising vaccine target because of its immunogenic responses when activated. We designed a multi-epitope vaccine targeting mutations in the KIT oncogene using CD8+ and CD4+ epitopes. We selected the most viable vaccine epitopes based on thresholds for percentile rank, immunogenicity, antigenicity, half-life, toxicity, IFNγ release, allergenicity, and stability. The efficacy of data was observed through world and regional population coverage of our vaccine design. Then, we obtained epitopes for optimized population coverage from PCOptim-CD, a modified version of our original Java-based program code PCOptim. Using 24 mutations on the KIT gene, 12 CD8+ epitopes and 21 CD4+ epitopes were obtained. The CD8+ dataset had a 98.55% world population coverage, while the CD4+ dataset had a 65.14% world population coverage. There were five CD4+ epitopes that overlapped with the top CD8+ epitopes. Strong binding to murine MHC molecules was found in four CD8+ and six CD4+ epitopes, demonstrating the feasibility of our results in preclinical murine vaccine trials. We then created three-dimensional (3D) models to visualize epitope-MHC complexes and TCR interactions. The final candidate is a non-toxic and non-allergenic multi-epitope vaccine against KIT mutations that cause AML. Further research would involve murine trials of the vaccine candidates on tumor cells causing AML.
Keywords: KIT oncogene; MHC I and MHC II molecules; TCR binding; acute myeloid leukemia (AML); artificial neural networks; epitopes; epitope–MHC complexes; immunoinformatics; murine MHC molecules; vaccine design.