Identification of small molecule inhibitors for the Brachyspira pilosicoli glutamate racemase (Bp-MurI) enzyme using a computational and experimental approach

Brachyspira pilosicoli (Bp) causes intestinal spirochaetosis in poultry and pigs and more rarely in humans. In this pilot study, we used a computational approach to identify potential small molecule inhibitors for the Bp glutamate racemase (Bp-MurI), an essential enzyme for cell wall biosynthesis, and then tested the Hit compounds against B. pilosicoli in a broth microdilution assay. Protein modelling and validation generated a reliable three-dimensional structure for Bp-MurI and development of a structure-based pharmacophore model led to the virtual screening of > 51,000 drug-like compounds from the Maybridge High Throughput Library. After filtering and molecular docking, three top Hit compounds (Hit 1, HTS02114; Hit 3, AW00718; Hit 4, HTS04172) were selected based on their binding affinities and interactions with the active site residues. Molecular Dynamics Simulations assessed the stability and dynamics of the Bp-MurI-inhibitor complexes. Hit 4 showed no biocidal activity against B. pilosicoli (MIC>1mM), Hit 1 had weak activity (MIC50 of 0.5 mM), and Hit 3 was the most effective (MIC50 and MIC100 values of 0.25 and 0.5 mM, respectively). Tiamulin, the currently used antibiotic, was highly effective (MIC50 and MIC100 values of 0.015 µM and 2µM, respectively). None of the compounds were cytotoxic to an epithelial cell line in culture using a resazurin assay, and SwissADME favourably assessed their physicochemical, pharmacokinetic and drug-likeness properties in silico. In conclusion, a computational modelling approach coupled with high throughput screening in silico is a useful process for more in-depth drug discovery against this pathogen.