A Deep Learning-Driven Framework Integrating Organoid-Based Functional Validation Identifies Universal Neoantigens...

Area of Science:

• Oncology
• Immunology
• Bioinformatics

Background:

• Glioblastoma (GBM) is an aggressive brain tumor with poor prognosis.
• Neoantigen therapy shows promise for GBM by modulating the tumor microenvironment.
• Current neoantigen prediction methods lack T cell immunogenicity assessment and physiological validation.

Purpose of the Study:

• To develop and validate a novel neoantigen prediction algorithm, TCRscore, integrating T cell receptor recognition features.
• To establish patient-derived glioblastoma organoid models for robust functional validation of predicted neoantigens.
• To enhance the accuracy and reliability of neoantigen identification for GBM immunotherapy.

Main Methods:

• Developed the TCRscore algorithm using public datasets, incorporating human leucocyte antigen binding and T-cell receptor recognition.
• Established 21 patient-derived glioblastoma organoid models from isocitrate dehydrogenase wildtype tumors.
• Validated predicted neoantigens using ELISpot assays, flow cytometry, and in vitro killing assays with organoid-T cell co-cultures.

Main Results:

• TCRscore demonstrated superior performance compared to six existing tools in predicting immunogenic neoepitopes.
• Glioblastoma organoid models accurately recapitulated parental tumor features, serving as effective validation platforms.
• Neoantigen-specific T cells successfully induced targeted killing in glioblastoma organoids, confirming functional immunogenicity.
• Identified a recurrent PIK3R1G376R mutation as a potential shared neoantigen in glioblastoma.

Conclusions:

• The TCRscore algorithm, combined with organoid-based validation, provides a high-fidelity method for GBM neoantigen discovery.
• This approach significantly enhances prediction accuracy, offering a valuable resource for developing effective GBM neoantigen-based immunotherapies.
• The study highlights the potential of targeting shared neoantigens, like those arising from PIK3R1 mutations, for broader therapeutic impact.