| Abstrakt: |
Colistin, an antibiotic, has become a last-resort therapy for serious infections caused by Antimicrobial Resistance (AMR) diseases during the last decade. The positively charged colistin coupled to the negatively charged lipid A can rupture the outer cell membrane of Gram-negative bacteria. However, the presence of a mobile colistin resistance gene (mcr gene) in Enterobacteriaceae has resulted in colistin resistance. MCR function transfers phosphoethanolamine (PEA) of phosphatidylethanolamine (PE) to lipid A, neutralizing its negative charge and preventing the binding of positively charged colistin. Currently, mcr isoforms varied from mcr-1 to mcr-10 have been discovered in environmental and clinical isolates, but only the three-dimensional structures of the catalytic portion of two MCR proteins, MCR-1 and MCR-2, were crystallized. Full-length MCR protein structures may be necessary for understanding MCR function and developing inhibitors; therefore, the structures of MCR-1 to 10 proteins were predicted by novel accurate protein prediction utilizing Deep Learning (RoseTTAFold). Based on multiple-sequence alignment and superposition on all MCR protein structures, there are six conserved residues at the active site, HIS¹, HIS2, HIS3, ASP, GLU, and THR. Tunnel analysis was utilized to determine the possible routes for substrate PE entering into MCR proteins. Among the four substrate-binding paths to the MCR active site (tunnels 1-4), PE preferentially binds at the active site via tunnel 1. This discovery not only anticipates PE as a substrate-binding to MCR protein, but it might also be beneficial for guiding MCR inhibitors. [ABSTRACT FROM AUTHOR] |
