Succinic Acids as Potent Inhibitors of Plasmid-borne IMP-1 Metallo-β-lactamase

Jeffrey H Toney,Gail G Hammond,Paula M.D Fitzgerald,Nandini Sharma,James M Balkovec,Gregory P Rouen,Steven H Olson,Milton L Hammond,Mark L Greenlee,Ying-Duo Gao

doi:10.1074/jbc.m104742200

Abstract

IMP-1 metallo-beta-lactamase (class B) is a plasmid-borne zinc metalloenzyme that efficiently hydrolyzes beta-lactam antibiotics, including carbapenems, rendering them ineffective. Because IMP-1 has been found in several clinically important carbapenem-resistant pathogens, there is a need for inhibitors of this enzyme that could protect broad spectrum antibiotics such as imipenem from hydrolysis and thus extend their utility. We have identified a series of 2,3-(S,S)-disubstituted succinic acids that are potent inhibitors of IMP-1. Determination of high resolution crystal structures and molecular modeling of succinic acid inhibitor complexes with IMP-1 has allowed an understanding of the potency, stereochemistry, and structure-activity relationships of these inhibitors.

Highlights

Carbapenems such as imipenem (Scheme 1) have proven useful for the treatment of a variety of Gram-negative and Gram-positive infections (1, 2)
We have identified a series of 2,3-(S,S)-disubstituted succinic acids that are potent inhibitors of IMP-1
The IMP-1 gene encoding an MBL has been identified on a plasmid and in Japan has transferred among clinical isolates such as Pseudomonas aeruginosa (11, 12), Klebsiella pneumoniae, Serratia marcescens, and other members of the Enterobacteriaceae (13, 14)

Summary

EXPERIMENTAL PROCEDURES

The IMP-1 metallo-␤-lactamase lacking the N-terminal 18 hydrophobic amino acids was expressed and purified as described (32). Molecular Modeling—Compound 8 was docked manually into the enzyme active site using the x-ray structure of the complex between IMP-1 and compound 1. The structures were docked into the active site overlapping with the structure of 1 as it is shown in the crystal structure, and the IMP-1-inhibitor complexes were energy-minimized using the same procedure as described for compound 8. For compounds 3 (R,R) and 4 (R,S) and the two stereoisomers of compound 2 (S,S), 100 conformations were generated, respectively, using the method described in Feuston et al (48) and were docked into the IMP-1 active site. 100 conformations were generated using the method described by Feuston et al (48) and docked into the IMP-1 active site using the x-ray structure of the complex between IMP-1 and compound 1. The complexes between imipenem and the IMP-1 enzyme were energy-minimized using the same procedure as described for compound 8

RESULTS

DISCUSSION

H H PhCH2 PhCH2 HHHHHHH