Intracellular copper does not catalyze the formation of oxidative DNA damage in Escherichia coli

Abstract

Because copper catalyzes the conversion of H(2)O(2) to hydroxyl radicals in vitro, it has been proposed that oxidative DNA damage may be an important component of copper toxicity. Elimination of the copper export genes, copA, cueO, and cusCFBA, rendered Escherichia coli sensitive to growth inhibition by copper and provided forcing circumstances in which this hypothesis could be tested. When the cells were grown in medium supplemented with copper, the intracellular copper content increased 20-fold. However, the copper-loaded mutants were actually less sensitive to killing by H(2)O(2) than cells grown without copper supplementation. The kinetics of cell death showed that excessive intracellular copper eliminated iron-mediated oxidative killing without contributing a copper-mediated component. Measurements of mutagenesis and quantitative PCR analysis confirmed that copper decreased the rate at which H(2)O(2) damaged DNA. Electron paramagnetic resonance (EPR) spin trapping showed that the copper-dependent H(2)O(2) resistance was not caused by inhibition of the Fenton reaction, for copper-supplemented cells exhibited substantial hydroxyl radical formation. However, copper EPR spectroscopy suggested that the majority of H(2)O(2)-oxidizable copper is located in the periplasm; therefore, most of the copper-mediated hydroxyl radical formation occurs in this compartment and away from the DNA. Indeed, while E. coli responds to H(2)O(2) stress by inducing iron sequestration proteins, H(2)O(2)-stressed cells do not induce proteins that control copper levels. These observations do not explain how copper suppresses iron-mediated damage. However, it is clear that copper does not catalyze significant oxidative DNA damage in vivo; therefore, copper toxicity must occur by a different mechanism.

FIG. 1.

Copper supplements are more toxic during anaerobic growth than during aerobic growth. (A and B) E. coli cultures were grown aerobically in LB, and CuSO₄ was added to a final concentration of 0 mM (□), 1 mM (○), 2 mM (⋄), 3 mM (▵), or 4 mM (▿). (A) Wild-type strain W3110. (B) GR17 (copA cueO cusCFBA mutant). (C and D) Cultures were grown anaerobically in LB, and CuSO₄ was added to a final concentration of 0 mM (▪), 0.25 mM (•), 0.5 mM (♦), or 1 mM (▴). (C) Wild-type strain W3110. (D) GR17 (copA cueO cusCFBA mutant). The data are representative of three independent experiments.

FIG. 2.

Copper protected DNA-repair-deficient mutants from killing by H₂O₂. The LEM1 (recA CopA⁺ CueO⁺ Cus⁺) (squares) and LEM2 (recA copA cueO cusCFBA) (circles) mutants were grown to the early log phase in LB without CuSO₄ (open symbols) or with 2.0 mM CuSO₄ (solid symbols). The cultures were then challenged with 2.5 mM H₂O₂. At each time, aliquots were removed and viability was determined. Recombination-proficient strains (RecA⁺) exhibited >80% survival during the same treatment (data not shown). The data are representative of three independent experiments.

FIG. 3.

Copper suppresses mutagenesis by H₂O₂. GR17 (copA cueO cusCFBA) mutant cells were grown to the early log phase in LB without CuSO₄ (squares) or with 2.0 mM CuSO₄ (circles). Cultures were then diluted to an OD₅₀₀ of 0.025, and either no H₂O₂ (open symbols) or 2.5 mM H₂O₂ (solid symbols) was added. At each time, catalase was added to aliquots to scavenge H₂O₂, and both viability and thyA mutants were quantified. The data are representative of three independent experiments.

FIG. 4.

Kinetic analysis indicates that copper specifically suppresses iron-mediated killing by low doses of H₂O₂. (A and B) λ_vir was exposed in vitro to 2.5 μM FeSO₄ (A) or 4 μM CuSO₄ (B) in the presence of various concentrations of H₂O₂. After a 30-s H₂O₂ challenge for iron-exposed phage and a 5-min challenge for copper-exposed phage, catalase was added to scavenge the H₂O₂, and phage viability was determined by examining plaque formation on W3110. (C and D) Mutants LEM1 (recA) (C) and LEM2 (copA cueO cusCFBA recA) (D) were grown to early log phase in LB containing 2.0 mM CuSO₄ (solid symbols) or no CuSO₄ (open symbols). Subsequently, cells were challenged for 2.5 min with various concentrations of H₂O₂. Catalase was added, and viability was determined. The data are representative of three independent experiments.

FIG. 5.

Copper exposure confers H₂O₂ resistance before growth slows. Mutant LEM63 (copA cueO cusCFBA recA) was grown to early log phase in LB. CuSO₄ (2.0 mM) (•) or double-distilled H₂O (○) was added. At subsequent times the OD₅₀₀ was determined (A), and an aliquot was removed and challenged with 2.5 mM H₂O₂ for 5 min (B). Catalase was then added, and viability was determined by plating. The data are representative of three independent experiments.

FIG. 6.

Copper prevented the formation of oxidative DNA lesions in vivo. GR17 (copA cueO cusCFBA) mutant cells were grown to the early log phase in LB with or without 2.0 mM CuSO₄. Cells were diluted to an OD₅₀₀ of 0.025 and incubated for 5 min in the presence of 3.3 mM KCN. H₂O₂ (2.5 mM) was then added. After 5 min catalase was added. (A) The total genomic DNA was isolated, and qPCR was performed using equal amounts of template DNA. The qPCR products were stained with ethidium bromide and scanned. Damage to the template DNA reduced the yield of the PCR product, and Poisson analysis allowed us to calculate the number of H₂O₂-induced lesions per genome that blocked the PCR polymerase. (B) After H₂O₂ challenge, cell viability was determined by plating. Solid bars, cells treated with 2.5 mM H₂O₂; open bars, untreated cells. The data are the means of three independent experiments. Error bars represent standard deviations.

FIG. 7.

Copper treatment induces both cell catalases. Strains GR17 (copA cueO cusCFBA; expressing KatG and KatE), LEM25 (copA cueO cusCFBA katG; expressing KatE), LEM37 (copA cueO cusCFBA katE; expressing KatG), and LEM29 (copA cueO cusCFBA katE katG; lacking both catalases) were grown in LB without CuSO₄ (open bars) or with 2.0 mM CuSO₄ (solid bars). Extracts were prepared, and their abilities to scavenge 10 mM H₂O₂ were determined at room temperature. The H₂O₂-scavenging ability of LEM29 (copA cueO cusCFBA katE katG) was ≤0.005 unit of absorbance (Abs)/min/mg protein and therefore could not be determined. The data are the means of three independent experiments. Error bars represent standard deviations.

FIG. 8.

Copper did not change hydroxyl radical generation inside cells. LEM29 (copA cueO cusCFBA katE katG) mutant cells were grown in LB without CuSO₄ (a) or with 2.0 mM CuSO₄ (b and c). For panel c, 1 mM dipyridyl was added to cells 5 min prior to exposure to 1 mM H₂O₂. Intracellular hydroxyl radicals were trapped and detected by EPR spectroscopy as described in Materials and Methods. No signal was present if H₂O₂ was not added (data not shown). The data are representative of three independent experiments.

FIG. 9.

H₂O₂-oxidizable copper is located in the periplasm. Cells were grown in LB with or without 2.0 mM CuSO₄. Cells were concentrated as described in Materials and Methods, and 10 mM H₂O₂ was added immediately before freezing. (A) EPR spectra of LEM29 (copA cueO cusCFBA katE katG) cells grown with and without 2.0 mM CuSO₄. No signal was evident if H₂O₂ was not added (data not shown). (B) The relative signal intensities of LEM27 (katE katG), LEM120 (copA katE katG), and LEM29 (copA cueO cusCFBA katE katG) mutant cells grown with (solid bars) and without (open bars) 2.0 mM CuSO₄ and exposed to H₂O₂ were calculated by dividing the EPR signal intensity by the OD₆₀₀ of the cell paste. The data are representative of three independent experiments.

FIG. 10.

H₂O₂ does not induce copper efflux genes. WOII260E (copA-lacZ) (A), WOII260B (cueO-lacZ) (B), and WOIII1A (cusC-lacZ) (C) cells were grown to an OD₅₀₀ of 0.1 in LB and then challenged with 60 μM H₂O₂ (solid bars) or 2.0 mM CuSO₄ (shaded bars). Control cells were not challenged (open bars). Cells were grown for 60 min at 37°C and then harvested. The data are the means of three independent experiments. Error bars represent standard deviations.