Non-classical neoantigens from fusion junctions of chimeric RNAs are novel targets for tumor-specific personalized vaccines with low risk of autoimmunity. We present a platform to discover immunogenic neoantigens driving CD8+ T-cell clonotypes from chimeric RNA fusion junctions, with the aim of promoting tumor-reactive T-cell expansion. RNA sequencing data from 1315 patient-derived xenograft (PDX) models harboring 28 different cancer types were analyzed for chimeric RNA. The CD74 [Exon 1-6] | NRG1 [Exon 5-12] fusion was selected based on its established role as an actionable cancer driver and its presence in 4 PDX models with a consistent fusion junction.

We assessed the affinity of 9,10,11mer neopeptides from the CD74-NRG1 fusion to MHC Class I molecules using in silico tools and confirmed these findings in vitro using flow cytometry analyses. Predicted binders were further modeled and ranked by structural features with APEGEN 2.0. Immunogenicity was evaluated via IFN-γ ELISpot assays using HLA-B*07:02-matched PBMCs. Dextramers conjugated with the feature barcode technology were utilized for single-cell 5’ gene expression RNA sequencing and T-cell receptor mapping of antigen-specific activated T cells. This study demonstrates a robust pipeline for identifying immunogenic neoantigens from chimeric RNAs, offering the potential for designing personalized cancer vaccines.