Poster Presentation 50th Lorne Proteins Conference 2025

TFold for accurate peptide-HLA structure prediction (#220)

Janesha Maddumage 1 2 , Stephanie Gras 3 4 5
  1. Infection and Immunity Program, La Trobe Institute for Molecular Science (LIMS), La Trobe University , Bundoora, Victoria 3086 , Australia
  2. Department of Biochemistry and Chemistry, School of Agriculture, Biomedicine and Environment (SABE), La Trobe University, Bundoora, Victoria 3086, Australia
  3. Department of Biochemistry and Chemistry, School of Agriculture, Biomedicine and Environment (SABE), La Trobe University, Bundoora, Victoria 3086, Australia
  4. Infection and Immunity Program, La Trobe Institute for Molecular Science (LIMS), La Trobe University , Bundoora, Victoria 3086, Australia
  5. Department of Biochemistry and Molecular Biology, Biomedicine Discovery Institute, Monash University, Clayton, Victoria 3800, Australia

The Human Leukocyte Antigen (HLA) system is integral to the adaptive immune response, facilitating the presentation of antigenic peptides to T cells. Highly polymorphic HLA molecules, with > 28,000 alleles, present both endogenous and exogenous peptides, allowing for robust immune surveillance. The binding of peptides to HLA involves complex hydrogen and hydrophobic interactions, with structural HLA pockets that bind to peptide anchor residues, determining peptide binding affinity (Nguyen et al, 2021). These binding mechanisms have been structurally characterised using X-ray crystallography, and yet their accurate prediction remains challenging. Understanding this molecular mechanism at the atomic level is crucial, as the specificity conferred by peptide-HLA binding enables T cell activation and protection against infection and cancer.

 

Recent advancements in sophisticated structure prediction tools, such as AlphaFold2, have revolutionised the field of structural biology by providing reliable models of protein structures.

Our research utilises the TFold pipeline, specifically designed for predicting the three-dimensional structures of peptide-HLA with higher accuracy than AlphaFold2 (Mikhaylov et al., 2024). By using TFold’s peptide binding scores, we assess the suitability and feasibility of peptides recognised by specific HLA allomorphs. These binding scores will be compared with experimentally obtained stability data and alignement with experimentally obtained structures to identify peptides with high potential of HLA binding.

 

This integrated approach enhances our understanding of the molecular mechanisms underlying peptide presentation and immune recognition while facilitating the identification of promising vaccine candidates. These findings aim to significantly advance the fields of immunology and vaccine development by elucidating the intricacies of peptide-HLA interactions.

  1. References Mikhaylov V, Brambley CA, Keller GLJ, Arbuiso AG, Weiss LI, Baker BM, Levine AJ (2024) Accurate modeling of peptide-MHC structures with AlphaFold. Structure 32: 228-241.e224 Nguyen AT, Szeto C, Gras S (2021) The pockets guide to HLA class I molecules. Biochemical Society Transactions 49: 2319-2331