Structural modeling of antibody variant epitope specificity with complementary experimental and computational techniques

Published: 06 Mar 2025, Last Modified: 26 Apr 2025GEMEveryoneRevisionsBibTeXCC BY 4.0
Track: Biology: datasets and/or experimental results
Nature Biotechnology: Yes
Keywords: antibody, antibody design, binding, HDX, cryo-EM, SPR, molecular docking, molecular modeling, molecular dynamics, structural modeling, structural features, computational structural biology
TL;DR: Our study describes the structural rules of antibody-antigen binding, showing how antibodies with different sequences follow shared antigen-binding principles based on geometry, surface, and biochemical properties.
Abstract: Antibodies are key therapeutic biomolecules yet the principles underlying diverse paratopes binding to the same epitope remain unexplained. An insufficient understanding of the structural rules of antibody-antigen binding, due to a lack of experimentally resolved structures, leads to the current inability to characterize antibody variants in silico. Here we explore a rule-based antibody design strategy that relies on a thorough understanding of epitope-paratope interactions, in contrast to generative design based on millions of trials and errors. We identified the epitope of six affinity-verified complexes between HER2 and Trastuzumab paratope variants using cryo-EM and position-resolved HDX-MS. Computational analysis of modeled structural conformational ensembles replicates and expands experimental results and highlights the importance of flexibility in understanding high and low-affinity binders. Structural parameters calculated based on geometry, surface, and biochemical properties were able to stratify antibodies by affinity. Overall, our study describes the structural interfaces of the paratope variants, showing how antibodies with diverse sequences share similar binding rules.
Anonymization: This submission has been anonymized for double-blind review via the removal of identifying information such as names, affiliations, and identifying URLs.
Presenter: ~Eva_Smorodina1
Format: Yes, the presenting author will attend in person if this work is accepted to the workshop.
Funding: Yes, the presenting author of this submission falls under ICLR’s funding aims, and funding would significantly impact their ability to attend the workshop in person.
Submission Number: 25
Loading