Large-scale transcript variants dictate neoepitopes for cancer immunotherapy

The success of cancer therapeutic vaccines hinges on precise and efficient discovery of tumor neoepitopes. However, conventional approaches for neoepitope identification face the challenge of building a repertoire containing sufficient immunogenic epitopes. Leveraging the advancements in full-length ribosome nascent-chain complex-bound mRNAs sequencing (FL-RNC seq) technology coupled with cutting-edge artificial intelligence (AI)-based predictive models. Our study developed a comprehensive workflow that facilitates the accurate and expansive discovery of the neoepitope landscape, with a particular emphasis on those epitopes that remained elusive or were overlooked owing to the constraints inherent in short read sequencing methods, specifically large-scale transcript variants (LSTVs). We located 22 MHC-restricted epitopes from LSTVs in an MC38 mouse model. Subsequently, a vaccine encoding these neoepitopes was synthesized employing the mRNA-LNP methodology and was evaluated both as a standalone therapy and in conjunction with anti-PD-1 immunotherapy. The outcomes of this study underscored that our elaborately engineered vaccine not only curbed tumor progression but also induced robust and specific T-cell mediated immunity, in addition to modulating the tumor micro-environment, underscoring the multifaceted potentials of LSTVs-derived neoepitope vaccine in shaping tumor immunotherapy strategies. Therefore, we introduce an approach that significantly expands the repertoire of neoepitope sources, offering a more universal methodology for discovering personalized cancer vaccines applicable to a broader array of tumor indications