MI2026 – Yoonjoo Choi

Title:
Allosteric deimmunisation of a Salmonella phosphatase enhances catalytic function

Abstract:
Genetically engineered Salmonella strains are promising vectors for cancer therapy, but their clinical use is limited by host immune responses. A key immunogenic antigen, the phosphatase PhoN, is metabolically essential for bacterial survival in the vitamin B6-depleted tumour microenvironment and thus cannot be simply deleted. While conventional deimmunisation targets surface residues, we present a strategy of allosteric deimmunisation by mutating non-surface, buried residues within the dominant h2 helix T-cell epitope. We hypothesised that the epitope is allosterically coupled to the distal active site gate, allowing for the simultaneous modulation of immunogenicity and enzymatic function. Using a computational pipeline, we designed deimmunised variants and used molecular dynamics (MD) simulations to predict their functional effects. The simulations revealed that mutations in the h2 helix allosterically controlled the gate’s flexibility. The enzymatic activities of the deimmunized variants were correlated with the MD predictions: two of the five designed variants exhibited 2~3-fold increases in catalytic function. This work demonstrates that non-surface epitopes can be rationally engineered to not only ablate immunogenicity but also to allosterically enhance protein function.

Biography:
Professor Yoonjoo Choi is a leading expert in computational protein engineering, with a particular focus on immunoinformatics, deimmunisation, and therapeutic protein design. Trained in theoretical physics at King’s College London and earning a DPhil in structural bioinformatics from the University of Oxford, he brings a rare multidisciplinary background to biomedical research. He currently holds a professorship at Chonnam National University Medical School in South Korea, where his research bridges computational biology and therapeutic translation.

Professor Choi has developed pioneering methods to identify and redesign immunogenic regions in therapeutic proteins, with a strong emphasis on structure-guided approaches. His work has contributed significantly to the rational engineering of proteins for reduced immunogenicity while preserving or enhancing biological function, including bacterial enzymes, antibodies, and tumour neoantigens. With over a decade of experience spanning institutions in the UK, USA, and Korea, he has co-led numerous national and international research programmes in immunotherapy, vaccine adjuvants, and AI-driven epitope prediction. His research continues to influence the fields of structural vaccinology and immunoengineering, making him a sought-after collaborator at the interface of computation and medicine.