The cell surface is where infection begins and where many cancers become treatable. Membrane proteins at this interface act as gateways for viral entry and as points of therapeutic vulnerability, yet many remain structurally uncharacterised. My PhD combined cryo-electron microscopy (cryo-EM), hydrogen-deuterium exchange mass spectrometry (HDX-MS) and computational protein design to define how two disease-associated receptors, neuropilin-1 (NRP1) and LRRC15, engage their molecular partners.
NRP1 is a host factor that promotes SARS-CoV-2 entry, and has been assumed to engage the viral spike protein through a single classical binding site. Using cryo-EM and HDX-MS, we identified non-canonical binding sites between NRP1 and the Omicron spike, revealing an unexpected mode of engagement with potential implications for how the virus enters host cells. We also tested whether LRRC15 is a direct spike receptor, but in vitro reconstitution showed no detectable interaction, suggesting additional cofactors are required for the binding reported in cells.
Beyond its proposed viral role, LRRC15 is an emerging cancer target. It is a structurally uncharacterised membrane protein enriched in cancer-associated fibroblasts and desmoplastic tumours, and the target of the antibody-drug conjugate samrotamab vedotin, now in Phase I trials. We determined the first high-resolution structure of LRRC15, bound to samrotamab at 2.6 Å by cryo-EM, showing that the antibody engages a membrane-proximal epitope and leaves the canonical concave surface fully exposed, a surface we then targeted with computationally designed minibinders of nanomolar affinity.
Together, these studies define how two cell-surface receptors are engaged at the molecular level, revealing a non-canonical mode of spike recognition by NRP1, and providing the first structural framework for therapeutic targeting of LRRC15.