High-throughput screening and Bayesian machine learning for copper-dependent inhibitors of Staphylococcus aureus

Area of Science:

• Medicinal Chemistry
• Drug Discovery
• Antimicrobial Research

Background:

• Copper-dependent inhibitors (CDIs) represent a promising source of novel antibacterial agents.
• Previous research indicated the frequent presence of staphylococcal CDIs in small chemical libraries.

Purpose of the Study:

• To conduct a large-scale industrial screen for novel anti-staphylococcal copper-dependent inhibitors (CDIs).
• To identify and characterize chemical motifs associated with CDI activity.
• To develop and validate a computational model for predicting CDI activity.

Main Methods:

• A large-scale screen of 40,771 compounds was performed in parallel in standard and copper-supplemented media.
• Sphere-exclusion clustering was used to analyze the chemical structures of identified hits.
• Structure-activity relationship (SAR) analysis was conducted on pyridinyl thieno-pyrimidines.
• Bayesian classification models were built using Discovery Studio and Assay Central.
• In silico evaluation of FDA-approved drugs was performed using the developed model.

Main Results:

• 483 compounds exhibited confirmed copper-dependent IC50 values under 50 μM.
• Sulfur-containing motifs, including benzimidazole-2-thiones and pyridinyl thieno-pyrimidines, dominated the identified hits.
• SAR analysis yielded improved CDI candidates with potential ligand/ion coordination activity.
• A validated computational model was developed for predicting copper-dependent activity.
• Two anti-helminths, albendazole and thiabendazole, were identified in silico as potential CDIs.

Conclusions:

• Large-scale screening effectively identified novel staphylococcal CDIs, primarily featuring sulfur-containing structures.
• The developed computational model accurately predicts copper-dependent activity and can be applied to existing drug libraries.
• Albendazole and thiabendazole show potential as copper-dependent antibacterial agents, warranting further investigation.