Antigen Processing and Presentation- Elliott Group
- Elliott Group
My lab has worked on Major histocompatibility complex class I (MHC I) biology for over 35years. MHC I molecules protect us from infections and from cancer by binding to peptides and presenting them to receptors on the surface of cytotoxic T cells or Natural Killer cells. Recognition of peptide-MHC complexes by these receptors activate the cytotoxic cells, leading to the elimination of the cells on which the peptide-MHC complexes are expressed.
We combine experimental biology with mathematical modelling, computational chemistry and in vivo immunology to investigate how mechanisms at the atomic-level control antigenic peptide selection by MHC I, and how this impacts on immune responses to cancer. For example we are exploring how the abundance of specific peptides at the cell surface of antigen presenting cells or tumour cells influences their immunogenicity and immunodominance; and how T cell avidity for those peptides guides their downstream function and effectiveness in fighting cancer.
We are interested in how intracellular chaperones including the molecules tapasin and calreticulin regulate the qualitative and quantitative selection of peptides inside the endoplasmic reticulum for presentation at the cell surface. We have also shown how polymorphic variations in the enzymes that generate peptides might influence the development of non-communicable diseases, including cancer
We were the first to show that short peptides were an integral component of MHC I and that particular peptides stabilised the MHC I structure and induced a conformational change. This led to the novel concept that MHC I protein dynamics is important for its ability to select only the most stably binding peptides– a feature that we have incorporated into mechanistic models that accurately simulate the biology of antigen presentation.
With the CRUK Oxford Centre COMBATcancer consortium https://www.cancer.ox.ac.uk/research/projects/combatcancer , we are building spatial maps of tumour antigen presentation to understand how tumour associated peptide presentation can shape the tumour immune microenvironment, in order to design ways of generating more aggressive and effective anti-cancer immunity.
We have combined these experimentally validated mechanistic models with ML/AI approaches to develop new “hybrid intelligence” models for predicting immunogenicity, and to develop new ways of designing vaccines that take into account the impact of antigen processing and intracellular peptide selection on immunogenicity. We combine these in silico approaches with experimental antigen discovery approaches using mass-spectrometry based immunopeptidomics (with the Adamopoulou lab). Mechanism-based models are also opening up new avenues for simulation and we are working on models that simulate immune surveillance in precancers and even virtual clinical trials for vaccines designed to intercept the development of cancer from precancer states.
We belong to the GO-PRECISE alliance (https://www.ox.ac.uk/news/2025-01-27-oxford-and-gsk-launch-50million-immuno-prevention-programme-advance-novel-cancer) and are actively involved in developing interception vaccines for lung, colon and breast pre-cancers – with others on the horizon.