Year: 2026
Type: preprint
Abstract: Molecular probes derived from antimicrobial peptides are powerful tools for diagnosing bacterial infectious diseases and guiding antibiotic therapy. However, existing probes rely on a limited repertoire of natural antimicrobial peptides, leaving the vast sequence space of linear peptides largely unexplored. Here, we used explainable artificial intelligence to design AIGP-3, a de novo peptide vector for fluorescent probes. Conjugated to Nile Red, NR-AIGP-3 exhibits high selectivity for Gram-positive bacteria, combining the environmental sensitivity of the fluorophore with the strong membrane binding of the vector and favorable MIC values for live-cell imaging. NR-AIGP-3 outperforms a ubiquicidin-derived probe in both flow cytometry and confocal microscopy, demonstrating its high potential for developing molecular probes for the diagnosis of bacterial infections.
Source: ChemRxiv
Institutions: Centre National de la Recherche Scientifique, Laboratoire d'Innovation Thérapeutique, Université de Strasbourg, Chimie de la Matière Complexe, Laboratoire de bioimagerie et pathologies, Hokkaido University, Institute for Chemical Reaction Design and Discovery, Institut Universitaire de France
Language: English
Subfield: Microbiology
Field: Immunology and Microbiology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7214598302