Assembly-Induced Membrane Selectivity of Artificial Model Peptides through Entropy–Enthalpy Competition

Abstract

Peptide design and drug development offer a promising solution for combating serious diseases or infections. In this study, using an AI–human negotiation approach, we have designed a class of minimal model peptides against tuberculosis (TB), among which K7W6 exhibits potent efficacy attributed to its assembly-induced function. Comprising lysine and tryptophan with an amphiphilic α-helical structure, the K7W6 sequence exhibits robust activity against various infectious bacteria causing TB (including clinically isolated and drug-resistant strains) both in vitro and in vivo. Moreover, it synergistically enhances the effectiveness of the first-line antibiotic rifampicin while displaying low potential for inducing drug resistance and minimal toxicity toward mammalian cells. Biophysical experiments and simulations elucidate that K7W6’s exceptional performance can be ascribed to its highly selective and efficient membrane permeabilization activity induced by its distinctive self-assembly behavior. Additionally, these assemblies regulate the interplay between enthalpy and entropy during K7W6–membrane interaction, leading to the peptide’s two-step mechanism of membrane interaction. These findings provide valuable insights into rational design principles for developing advanced peptide-based drugs while uncovering the functional role played by assembly.