Structure-based targeting of the lipid A-modifying enzyme PmrC to contrast colistin resistance in Acinetobacter baumannii.

Introduction: Antimicrobial-resistant pathogens are an ongoing threat to human and animal health. According to the World Health Organization (WHO), colistin is considered the last resort antibiotic against human infections due to multidrug-resistant Gram-negative organisms-including Acinetobacter baumanni, a priority-1 pathogen. Despite colistin being considered a last resort antibiotic, transferable bacterial resistance to this drug has been reported in humans and animals. This makes addressing colistin resistance a critical priority in public health efforts. The large PetN transferase membrane protein PmrC is responsible for colistin resistance due to its catalysed modification of lipid A of the external membrane. Despite its importance, this potential drug target was never characterised at a molecular level.

Methods: The recombinant production of large membrane proteins in their native forms is a bottleneck in modern molecular biology. In this study, we recombinantly produced PmrC and biophysically characterised it in solution. We employed in silico approaches, including virtual screening and molecular modelling, to identify PmrC ligands. The binding of these ligands to PmrC was measured using Microscale Thermophoresis (MST). The best ligand was tested for its ability to hamper colistin resistance in Acinetobacter baumannii clinical isolates. Finally, we checked that the identified compound was not cytotoxic at the used concentrations by haemolysis assays.

Results: We successfully produced PmrC PetN transferase membrane protein in high yields and showed that PmrC is a stable α-β protein, with melting temperature T_m = 60°C. Based on the PmrC structural model, we identified a promising druggable cavity. Therefore, we used a structure-based virtual screening to identify potential inhibitors. A small molecule, here denominated as s-Phen, was proved to bind PmrC with μM affinity. Microbiological assays confirmed that the s-Phen can drastically reduce colistin minimum inhibitory concentration (MIC) in two A. baumannii-resistant isolates and that it is not cytotoxic. Importantly, PmrC binding pocket to s-Phen is highly conserved in all homologues of PmrC, regardless of the location of genes encoding for them and of their operons.

Discussion: Our study provides a molecular characterisation of PmrC and demonstrates the importance of PmrC as a drug target and the strong potential of PmrC binding molecules to act as colistin adjuvants, operating as synergistic tools to combat multiresistant nosocomial pathogens.