In Silico Identification and Molecular Characterization of Lentilactobacillus hilgardii Antimicrobial Peptides with Activity Against Carbapenem-Resistant Acinetobacter baumannii

Background/Objectives: Biofilm formation by Acinetobacter baumannii contributes to its persistence in clinical settings and resistance to antibiotic treatment. This study aims to identify and characterize antimicrobials from lactic acid bacteria (LAB) using molecular and in silico approaches that can prevent and disrupt A. baumannii biofilms, assess their antimicrobial host range, and define their synergy with current antibiotics. Methods: Thirteen LAB isolates from the Human Microbiome Project were screened for anti-biofilm activity against A. baumannii. Conditioned media was further tested against six ESKAPE pathogens and three skin commensals. Lentilactobacillus hilgardii was selected for detailed study and antimicrobial peptide (AMP) prediction analysis due to limited toxicity toward commensals. In silico identified peptides were synthesized and tested individually and in combination with sub-MIC doses of an antibiotic. Results: Conditioned media from five LAB species (Lentilactobacillus hilgardii, Lentilactobacillus buchneri, Ligilactobacillus ruminis, Limosilactobacillus fermentum, and Limosilactobacillus antri) significantly inhibited A. baumannii biofilm formation and reduced biomass of mature biofilms. LAB-conditioned media also exhibited broad-spectrum activity against ESKAPE pathogens, though effects on commensals varied. Bioinformatically predicted AMPs from L. hilgardii inhibited planktonic A. baumannii growth but showed no direct biofilm activity even at high doses. However, AMPs were found to synergize with sub-MIC doses of meropenem against mature biofilms leading to decolonization. Conclusions: Our study provides a comprehensive platform for the discovery and characterization of AMPs and supports using commensal bacteria to reduce, prevent, and decolonize biofilms from pathogenic bacteria in community and nosocomial settings.