# Hotspot Interactions between Two Fab Molecules in Molecular Dynamics Simulations Improve Predictive Models of Aggregation Kinetics

## TL;DR

This paper uses molecular dynamics simulations to identify key interaction sites between antibody fragments, improving models for predicting protein aggregation.

## Contribution

The study introduces a novel approach combining MD simulations and hotspot analysis to enhance aggregation kinetics predictions.

## Key findings

- Specific residues were identified as persistent hotspots through contact map and PCA analysis.
- Including solvent accessibility of hotspots and APRs improved aggregation models across 49 conditions.
- Interfragment interactions significantly influence conformational dynamics compared to single Fab simulations.

## Abstract

Protein–protein interactions (PPIs) are fundamental
to numerous
biological processes, and the identification of interaction hotspots
is essential for understanding the mechanisms of protein aggregation
and informing protein engineering efforts. Although various algorithms
have been developed to predict hotspot regions for protein–protein
interactions, little research has focused on understanding the relative
roles of these largely protein surface-based interactions and the
interactions between cross-β sheet-forming aggregation-prone
regions (APRs) that are largely buried within proteins. This study
uses all-atom molecular dynamics (MD) simulations to investigate the
interactions between two Fab antibody fragments, focusing on the identification
and characterization of the key interaction sites. Through frequency
contact map analysis and principal component analysis, we identified
specific residues that consistently formed stable contacts, distinguishing
them from transient random interactions. Our findings revealed that
while numerous contact points occurred throughout the simulations,
relatively few sites acted as persistent hotspots based on the duration
of their contacts. Comparison to single Fab simulations highlighted
the influence of interfragment interactions on conformational dynamics.
Inclusion of solvent accessibility for two surface hotspots, alongside
one predicted APR, significantly improved models for predicting aggregation
kinetics over 49 formulation conditions. The molecular-level insights
gained will be important for guiding protein engineering strategies
aimed at modulating these interactions to enhance product stability
and retain therapeutic efficacy.