ABSTRACT
One of the hurdles in the discovery of antibiotics is the difficulty of linking antibacterial compounds to their cellular targets. Our laboratory has employed a genome-wide approach of over-expressing essential genes in order to identify cellular targets of antibacterial inhibitors. Our objective in this project was to develop and validate a more sensitive disk diffusion based platform of target identification (Target Identification Platform for Antibacterials version 2; TIPA II) using a collection of cell clones in an Escherichia coli mutant (AS19) host with increased outer membrane permeability. Five known antibiotics/inhibitors and 28 boron heterocycles were tested by TIPA II assay, in conjunction with the original assay TIPA. The TIPA II was more sensitive than TIPA because eight boron heterocycles previously found to be inactive to AG1 cells in TIPA assays exhibited activity to AS19 cells. For 15 boron heterocycles, resistant colonies were observed within the zones of inhibition only on the inducing plates in TIPA II assays. DNA sequencing confirmed that resistant clones harbor plasmids with fabI gene as insert, indicating that these boron heterocycles all target enoyl ACP reductase. Additionally, cell-based assays and dose response curved obtained indicated that for two boron heterocycle inhibitors, the fabI cell clone in AG1 (wild-type) host cells exhibited at least 11 fold more resistance under induced conditions than under non-induced conditions. Moreover, TIPA II also identified cellular targets of known antibacterial inhibitors triclosan, phosphomycin, trimethoprim, diazaborine and thiolactomycin, further validating the utility of the new system.
Subject(s)
Boron Compounds/chemistry , Heterocyclic Compounds/chemistry , Boron Compounds/pharmacology , Drug Delivery Systems/methods , Escherichia coli/drug effects , Heterocyclic Compounds/pharmacologyABSTRACT
2-Acylated 2,3,1-benzodiazaborines can display unusual structures and reactivities. The crystal structure analysis of the boron heterocycle obtained by condensing 2-formylphenylboronic acid and picolinohydrazide reveals it to be an NâB-chelated zwitterionic tetracycle (systematic name: 1-hydroxy-11-oxo-9,10,17λ(5)-triaza-1λ(4)-boratetracyclo[8.7.0.0(2,7).0(12,17)]heptadeca-3,5,7,12,14,16-hexaen-17-ylium-1-uide), C13H10BN3O2, produced by the intramolecular addition of the Lewis basic picolinoyl N atom of 1-hydroxy-2-(pyridin-2-ylcarbonyl)benzo[d][1,2,3]diazaborinine to the boron heterocycle B atom acting as a Lewis acid. Neither of the other two pyridinylcarbonyl isomers (viz. nicotinoyl and isonicotinoyl) are able to adopt such a structure for geometric reasons. A favored yet reversible chelation equilibrium provides an explanation for the slow D2O exchange observed for the OH resonance in the (1)H NMR spectrum, as well as for its unusual upfield chemical shift. Deuterium exchange may take place solely in the minor open (unchelated) species present in solution.
ABSTRACT
A set of 2-acylated 2,3,1-benzodiazaborines and some related boron heterocycles were synthesized, characterized, and tested for antibacterial activity against Escherichia coli and Mycobacterium smegmatis. By high-field solution NMR, the heretofore unknown class of 2-acyl-1-hydroxy-2,3,1-diazaborines has been found to be able to exist in several interconvertable structural forms along a continuum comprised of an open hydrazone a, a monomeric B-hydroxy diazaborine b, and an anhydro dimer c. X-Ray crystallography of one of the anhydro dimers, 17c, revealed it to have an unprecedented structure featuring a double intramolecular OâB chelation. The crystal structure of another compound, 37, showed it to be based on a new pentacyclic B heterocycle framework. Nine compounds were found to possess activities against E. coli, and two others were active against M. smegmatis. The finding that these two contain isoniazid covalently embedded in their structures suggests that they might possibly be acting as prodrugs of this well-known antituberculosis agent in vivo.
