Comparison of the mechanism of antimicrobial action of the gold(I) compound auranofin in Gram-positive and Gram-negative bacteria

Abstract

The antirheumatic gold(I) compound auranofin has been suggested to exhibit bactericidal activity by disrupting the thiol homeostasis through direct inhibition of bacterial thioredoxin reductase (TrxB). While highly effective at killing Gram-positive bacteria, it lacks significant activity against Gram-negative species for reasons that largely remain unclear. Here, we aimed to elucidate the molecular mechanisms underlying the low susceptibility of the Gram-negative model organism Escherichia coli to auranofin when compared to the Gram-positive model organism Bacillus subtilis. A change in the proteome of E. coli exposed to auranofin suggested that the effect of this gold compound is a combination of inactivation of thiol-containing enzymes, upregulation of proteins involved in basal metabolism, and the consequent induction of systemic oxidative stress. Susceptibility tests in E. coli mutants lacking the proteins upregulated upon auranofin treatment suggested that none of them are directly involved in E. colis high tolerance to auranofin. To elucidate factors that could make Gram-negative bacteria less susceptible to auranofin, we tested E. coli cells lacking the efflux pump component TolC. These cells were more sensitive to auranofin treatment than the wild type, but not to an extent that would fully explain the observed difference in susceptibility of Gram-positive and Gram-negative organisms. We thus tested if E. colis thioredoxin reductase (TrxB) is inherently less sensitive to auranofin than TrxB from B. subtilis, which was not the case. E. coli cells lacking components of the thioredoxin-system were also only marginally more susceptible to auranofin. However, E. coli strains lacking the low molecular weight thiol glutathione, but not glutathione reductase, showed a high susceptibility to auranofin. Bacterial cells expressing the genetically encoded redox probe roGFP2 allowed us to observe the oxidation of cellular protein thiols in situ. In line with their susceptibility, the kinetics of probe oxidation and the degree of oxidation promoted by auranofin is significantly higher in Gram-positive bacteria when compared to Gram-negative bacteria. Based on our findings, we hypothesize that auranofin leads to a global disturbance in the cellular thiol redox homeostasis in bacteria, but Gram-negative bacteria are inherently more resistant due to the presence of drug export systems and high cellular concentrations of glutathione.