Unraveling mechanisms of protein encapsulation and release in coacervates via molecular dynamics and machine learning

Abstract

Coacervates play a pivotal role in protein-based drug delivery research, yet their drug encapsulation and release mechanisms remain poorly understood. Here, we utilized the Martini model to investigate bovine serum albumin (BSA) protein encapsulation and release within polylysine/polyglutamate (PLys/PGlu) coacervates. Our findings emphasize the importance of ingredient addition sequence in coacervate formation and encapsulation rates, attributed to preference contact between oppositely charged proteins and poly(amino acid)s. Notably, coacervates composed of β-sheet poly(amino acid)s demonstrate greater BSA encapsulation efficiency due to their reduced entropy and flexibility. Furthermore, we examined the pH responsiveness of coacervates, shedding light on the dissolution process driven by Coulomb forces. By leveraging machine learning algorithms to analyze simulation results, our research advances the understanding of coacervate-based drug delivery systems, with the ultimate goal of optimizing therapeutic outcomes.