Oral Presentation 31st Annual Lorne Proteomics Symposium 2026

Actionable T Cell Targets in Paediatric Diffuse Midline Glioma: From H3K27M Neoantigen to Dark Antigens (133458)

Tima TS Shamekhi 1 , Terry C.C Lim Kam Lim 1 2 , Gabriel Goncalves 1 2 , Nivedhitha Selvakumar 1 , Anh Doan 3 , Kylie Loh 3 , Joshua Ooi 3 4 , Bijun Zeng 5 6 , Roberta Mazzieri 5 6 , Riccardo RD Dolcetti 6 7 , Ron RF Firestein 1 , Pouya PF Faridi 1 2 8
  1. Centre for Cancer Research , Hudson Institute of Medical Research, Clayton, Victoria, Australia
  2. Monash Proteomics and Metabolomics Facility, Department of Biochemistry and Molecular Biology, Monash Biomedicine Discovery Institute, Monash University, Clayton, Victoria, Australia
  3. School of Clinical Sciences, Department of Medicine, Monash University, Clayton, Victoria, Australia, Centre for Inflammatory Diseases, Monash University Department of Medicine, Monash Medical Centre, Clayton, Victoria 3168, Australia, Clayton, VIC, Australia
  4. School of Clinical Sciences, Department of Medicine, Monash University, Clayton, Victoria, Australia, Regulatory T-cell Therapies Lab School of Clinical Sciences at Monash Health Monash University., Clayton, VIC, Australia
  5. Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
  6. Department of Microbiology and Immunology, The University of Melbourne, Parkville, Victoria 3010, Australia, Australia
  7. Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
  8. School of Clinical Sciences, Department of Medicine, Monash University, Clayton, Victoria, Australia

Diffuse midline glioma (DMG) remains among the most lethal paediatric malignancies, characterised by the H3K27M mutation and profound resistance to current therapies. Despite remarkable progress in cancer immunotherapy, the antigenic determinants capable of eliciting effective T-cell responses in DMG remain undefined. Here, we delineate the first comprehensive map of H3K27M-derived and H3K27M-induced antigens, spanning canonical and noncanonical sources, and demonstrate their therapeutic tractability across multiple HLA contexts.

Using isogenic H3K27M mutant–wild-type cell line pairs, we resolved how this onco-histone remodels the HLA class I ligandome. Through an integrated multi-omics and functional pipeline combining deep immunopeptidomics, predictive modelling, and T-cell functional assays we identified and validated five naturally presented, immunogenic H3.3K27M neoepitopes restricted by three distinct HLA supertypes. These neoantigens were confirmed in patient tumours by PRM-targeted mass spectrometry and elicited potent cytotoxicity when targeted by cloned H3K27M-specific TCRs in co-culture assays.

To chart the broader antigenic landscape, we profiled 22 patient-derived DMG lines and 17 primary tumours, integrating fractionated data-dependent and data-independent acquisition immunopeptidomics. By filtering tumour ligandomes against benign brain and reference HLA datasets and using a custom proteogenomic database incorporating Ribo-seq-defined noncanonical ORFs, we uncovered a rich layer of cryptic antigens arising from H3K27M-driven chromatin dysregulation. Remarkably, 18% of these tumour-exclusive ligands were shared across samples, indicating recurrent, non-mutational vulnerabilities exploitable by “off-the-shelf” immunotherapies. Functional testing confirmed that selected dark-proteome-derived peptides robustly activated T cells in an HLA-restricted manner.

Together, this work defines the first integrated antigenic atlas of DMG, encompassing both mutation-derived and epigenetically induced antigens. It establishes a direct mechanistic link between H3K27M-mediated chromatin remodelling and tumour-specific antigen presentation providing an actionable foundation for next-generation TCR-T and vaccine therapies in a cancer long deemed immunologically silent.