Targeting peptide-MHC complexes with designed T cell receptors and antibodies

Area of Science:

• Immunology
• Structural Biology
• Computational Biology

Background:

• Class I Major Histocompatibility Complexes (MHCs) present intracellular peptides to T cells for immune surveillance.
• T cell receptor (TCR) recognition of peptide-MHC (pMHC) complexes is crucial for identifying pathogens and tumors.
• Designing therapeutic receptors targeting specific pMHCs is challenging due to self-antigen constraints.

Purpose of the Study:

• To introduce ADAPT, a structure-based deep learning framework for designing T cell receptors (TCRs) and antibodies.
• To enable the design of novel receptors targeting specific peptide-MHC (pMHC) complexes.
• To explore therapeutic applications in cancer immunotherapy and autoimmune diseases.

Main Methods:

• Developed a deep learning framework (ADAPT) integrating structural information for antigen-receptor design.
• Designed and characterized TCRs and antibodies against a diverse panel of pMHC targets.
• Utilized cryogenic electron microscopy (cryo-EM) to validate the atomic-level accuracy of designed antibodies.

Main Results:

• Demonstrated the capability of ADAPT to design TCRs and antibodies with high specificity for pMHC targets.
• Cryo-EM structures confirmed atomic-level accuracy of designed antibodies at the pMHC recognition interface.
• Validated the robustness and potential of the structure-based design approach.

Conclusions:

• ADAPT provides a powerful platform for designing therapeutic receptors targeting pMHC complexes.
• Designed receptors hold promise for advancing cancer immunotherapy and treating autoimmune conditions.
• This work enhances predictive understanding of TCR specificity, impacting basic immunology and diagnostics.