Abstract
INTRODUCTION: Pseudomonas aeruginosa lung infections are difficult to eradicate because of biofilms and rising resistance. MraY (MRAY_PSEAE), an essential membrane enzyme in peptidoglycan biosynthesis, depends on divalent metals (notably Mg(2+)); therefore, metal coordination is critical for reliable structure-based prioritization. METHODS: We built an explicit protein-ligand-metal model of MRAY_PSEAE and extracted pocket constraints from point-cloud geometry and polar/charged features. Candidates were generated through similarity-guided library expansion, combining PubChem 2D similarity retrieval with local rule-based BRICS enumeration. After canonicalization and 2D/property filtering, compounds were clustered by fingerprint similarity and prioritized through chemical-state plausibility, docking, Prime/MM-GBSA rescoring, and ligand strain filtering, followed by interaction analysis, 1-μs MD, and in silico ADMET. RESULTS: Three hits (6675, 2733768, 5311309) were prioritized. In microsecond MD, compound 5311309 showed the most stable pose and key hotspot retention. ADMET indicated a permeability/efflux bottleneck along with liver-kidney and irritation/sensitization alerts, whereas hERG and AMES signals were favorable. In growth assays, compound 5311309 showed an IC50 of approximately 15.1 μM and an MIC of 64 μM. CONCLUSION: Metal-aware pocket modeling plus multi-level screening improved plausibility and ranking for this metal-dependent target, identifying compound 5311309 as a practical lead optimization starting point.