Guidelines To Predict Binding Poses of Antibody-Integrin Complexes.

Area of Science:

• Structural Biology
• Computational Chemistry
• Biopharmaceutical Development

Background:

• Integrins are critical cell adhesion receptors involved in numerous signaling pathways and disease pathologies.
• Therapeutic antibodies targeting integrins are increasingly important, necessitating accurate prediction of their binding interactions.
• Current methods for determining integrin-antibody complex structures require robust computational strategies.

Purpose of the Study:

• To evaluate protein-protein docking programs for generating accurate binding poses of integrin-antibody complexes.
• To develop and validate a reliable strategy for identifying the native binding pose and avoiding false positives.
• To provide insights for the rational design of novel anti-integrin antibody therapeutics.

Main Methods:

• Utilized protein-protein docking simulations using ClusPro and GRAMM-X to generate potential binding poses.
• Applied a novel energy-based rule to select the most likely native pose from docking outputs.
• Employed heated molecular dynamics simulations as a secondary validation method for ambiguous cases.

Main Results:

• ClusPro and GRAMM-X demonstrated superior performance in generating the native pose of integrin-antibody complexes.
• A specific energy difference threshold (first model >5% more negative than the second) successfully identified native poses in most cases.
• Heated molecular dynamics confirmed native poses with RMSD values below 0.5 nm when the energy rule was inconclusive.

Conclusions:

• The proposed methodology, combining docking with an energy-based rule and molecular dynamics, is efficient for predicting integrin-antibody binding poses.
• This approach helps avoid misranking incorrect poses, enhancing confidence in computational models.
• The validated methods support the rational design and development of effective anti-integrin biopharmaceuticals.