Computational study of amphiphilic protein nano-bundle interactions

Abstract

Hydrophobins are natural amphiphilic proteins that contain hydrophobic and hydrophilic regions, which enable their assembly at interfaces. Their unique interfacial properties are attractive, and they can be modified to modulate their structure and function for a variety of applications, including emulsification. Among them, HFBI, a class II hydrophobin, has shown an ability to form self-assembled structures at air-water surfaces. Using this property, a HFBI-based nano-bundle (HFBI-HexCC) has recently been designed. This protein is engineered into a nano-scale material while still maintaining a hydrophobic patch for amphiphilic properties and disulfide bridges for structural stability. In this study, we computationally investigated the protein-protein interactions (PPIs) between HFBI-HexCC, using a deep learning model DDMut-PPI. A total of 380 mutations of HFBI-HexCC residues were studied to identify a set of recommended mutations that could potentially modulate protein-protein interactions between HFBI-HexCC, thereby improving its amphiphilic properties for application in emulsions.