Mechanism of H2S-mediated protection against oxidative stress in Escherichia coli

Abstract

Endogenous hydrogen sulfide (H₂S) renders bacteria highly resistant to oxidative stress, but its mechanism remains poorly understood. Here, we report that 3-mercaptopyruvate sulfurtransferase (3MST) is the major source of endogenous H₂S in Escherichia coli Cellular resistance to H₂O₂ strongly depends on the activity of mstA, a gene that encodes 3MST. Deletion of the ferric uptake regulator (Fur) renders ∆mstA cells hypersensitive to H₂O₂ Conversely, induction of chromosomal mstA from a strong pLtetO-1 promoter (P _tet -mstA) renders ∆fur cells fully resistant to H₂O₂ Furthermore, the endogenous level of H₂S is reduced in ∆fur or ∆sodA ∆sodB cells but restored after the addition of an iron chelator dipyridyl. Using a highly sensitive reporter of the global response to DNA damage (SOS) and the TUNEL assay, we show that 3MST-derived H₂S protects chromosomal DNA from oxidative damage. We also show that the induction of the CysB regulon in response to oxidative stress depends on 3MST, whereas the CysB-regulated l-cystine transporter, TcyP, plays the principle role in the 3MST-mediated generation of H₂S. These findings led us to propose a model to explain the interplay between l-cysteine metabolism, H₂S production, and oxidative stress, in which 3MST protects E. coli against oxidative stress via l-cysteine utilization and H₂S-mediated sequestration of free iron necessary for the genotoxic Fenton reaction.

Keywords: antibiotics; cysteine; hydrogen sulfide; oxidative stress; sulfur metabolism.

Fig. 1.

Quantitation of H₂S production by WT, 3MST-deficient (∆mstA), and 3MST-overproducing (P_tet-mstA) E. coli. (A) Representative Pb(Ac)₂-soaked paper strips show a PbS brown stain as a result of the reaction with H₂S. Strips were affixed to the inner wall of a culture tube above the level of the liquid culture of WT or mutant bacteria for 18 h. Numbers show the change in H₂S production relative to WT cells. The values are means from three independent experiments with a margin error of less than 10%. (B) Representative fluorescence images of H₂S production by live WT and mutant E. coli cells treated with the TICT-based fluorescent H₂S probe (15). (Magnification: 100×.) (C) Fluorescence intensities of WT and mutant E. coli cells in Luria–Bertani broth or Luria–Bertani broth plus 2 mM H₂O₂ treated with fluorescent H₂S probe as detected by Cytation 3 (BioTek Instruments Inc.). Values are means ± SD (n = 3). RFU, relative fluorescence unit. *P < 0.05 (Student’s t test; equal variance).

Fig. 2.

3MST-derived H₂S protects E. coli from H₂O₂ toxicity. (A) Representative survival curves show the effect of H₂S deficiency (∆mstA) or overproduction (P_tet-mstA) on H₂O₂-mediated killing. (B) An fur mutation promotes H₂O₂ cytotoxicity in WT and ∆mstA cells but not in P_tet-mstA cells. The percentage of surviving cells was determined by counting cfu and is shown as the mean ± SD from three experiments. (C) Relative change in H₂O₂ sensitivity of WT, ∆mstA, and P_tet-mstA cells in response to Fur deficiency (∆fur). Values are means ± SD from three experiments.

Fig. 3.

3MST-derived H₂S protects genomic DNA from the damaging Fenton reaction. (A) 3MST-derived H₂S sequesters intracellular iron. Representative Pb(Ac)₂-soaked paper strips show the decrease in the amount of H₂S generated in P_tet-mstA cells in response to deletion of fur or sodA sodB genes. Such deletions cause a drastic increase in the intracellular free iron content (21). Addition of 200 µM 2,2′-dipyridyl (dp), an iron chelator, restores H₂S to its original level in each case. The values (percentages) are means from three independent experiments with a margin error of less than 10%. (B) 3MST-derived H₂S renders cells less susceptible to DNA damage as evidenced by the higher H₂O₂ concentration necessary to induce the SOS response in P_tet-mstA cells. The SOS response was monitored by bioluminescence of the lux biosensor (pColD′::lux) in Δfur, Δfur ΔmstA, Δfur P_tet-mstA, and WT cells in the presence of different concentrations of H₂O₂. J/J_k indicates the induction factor in percentage compared with the maximal intensity of bioluminescence of samples in the presence of H₂O₂. Values are means ± SD from three experiments. (C) 3MST-derived H₂S renders cells less susceptible to H₂O₂-induced DNA breaks as detected by TUNEL. The graph shows the percentage of gated propidium iodide cells that are TUNEL-positive as detected by fluorescence intensity greater than that of untreated cells. Statistical evaluation (one-way ANOVA and Tukey’s post hoc test) was performed to evaluate differences in the cell population.

Fig. S1.

The relative expression of a chromosomal copy of mstA from its native promoter or Ptet-mstA in exponentially grown WT and mutant cells as measured by qRT-PCR. Values are means ± SD from three independent experiments.

Fig. S2.

H₂S protects cells from H₂O₂-induced DNA damage. TUNEL analysis of H₂O₂-induced DNA damage. A–E are histograms of three biological repeat measurements of TUNEL-positive events analyzed on a FACSCalibur for each strain. Samples were treated with 5 mM H₂O₂ for 30 min at room temperature before fixation and labeling followed by analysis. The x axis (FL1-H FITC) represents the relative FITC fluorescence with respect to the untreated population, and the y axis is the number of events after gating for the propidium iodide-stained population.

Fig. S3.

TnaA has no effect on (A) H₂S production or (B) H₂O₂-mediated bacterial killing. The values (percentages) are means from three independent experiments with a margin error of less than 10%. Representative survival curves show the effect of TnaA deficiency (∆tnaA) or overexpression (P_tet-tnaA) on H₂O₂-mediated killing in a ∆fur background. The percentage of surviving cells was determined by counting cfu and is shown as the mean ± SD from three experiments.

Fig. 4.

Functional interaction between 3MST and CysB regulon. (A) The relative expression of CysB-regulated genes in exponentially grown ΔmstA and P_tet-mstA cells was measured by qRT-PCR. The relative expression (y axis) represents the fold change of each mRNA level compared with that of the untreated cells. Values are means ± SD from four experiments. (B) Induction of CysB-regulated gene expression by H₂O₂ (2 mM, 20 min) in exponentially grown WT and ∆mstA cells as detected by qRT-PCR. The relative expression (y axis) represents the fold change of each mRNA level after treatment of the cells with H₂O₂ compared with that of the untreated cells (dashed line). Values are means ± SD from four experiments.

Fig. S4.

Interplay between H₂S generation, serine acetyltransferase activity, and cysB gene expression. The activity of serine acetyltransferase encoded by the cysE gene is subject to feedback inhibition by l-cysteine. The product of the acetyltransferase reaction is OAS, which spontaneously isomerizes to NAS. NAS acts as an inducer of the master transcriptional regulator CysB. CysB protein binds immediately upstream of the −35 region of positively regulated promoters, where in the presence of inducer, it facilitates formation of a transcription initiation complex. CysB also autoregulates its own synthesis by binding to its own promoter as a repressor. NAS stimulates CysB protein binding to sites involved in positive regulation and inhibits binding to the negatively regulated cysB promoter (9). The right-facing arrow beneath cysB indicates the direction of transcription. P, promoter.

Fig. S5.

Relative expression of CysB-regulated genes in ΔtnaA and P_tet-tnaA cells. The expression was measured by qRT-PCR in exponentially growing cells. Values are means ± SD from three independent experiments.

Fig. 5.

Interdependence between 3MST activity and l-cysteine/cystine import. Constitutive expression of the TcyP transporter suppresses the negative effect of cysB deletion on H₂S production in (A) P_tet-mstA cells or (B) ∆mstA and WT cells as detected by the Pb(Ac)₂ assay. Representative panels show mean values (percentages) from three independent experiments with a margin error of less than 10%. (C) A model of H₂S-mediated defense against oxidative stress in E. coli. A fraction of exogenous H₂O₂ reacts with l-cysteine in the periplasm to form l-cystine and H₂O. This reaction leads to a decrease in the intracellular content of l-cysteine with a subsequent relief of autoregulation of cysB and activation of CysB-dependent genes, including tcyP, which is responsible for transport of l-cystine into the cell. Overflow of cystine/cysteine flux results in increased mstA-dependent generation of H₂S, which sequesters free iron to prevent the Fenton reaction and formation of damaging hydroxyl radicals.

Fig. S6.

Deletion of cysB abolishes the protective effect of H₂S against H₂O₂. Overnight cultures of Δfur (squares), Δfur ΔmstA (triangles), and Δfur P_tet-mstA (circles) in (A) cysB+ and (B) ΔcysB backgrounds were inoculated in Luria–Bertani broth liquid medium and grown to OD ∼ 0.2 followed by addition of 2 mM H₂O₂ (black) or water (white). Cells were grown in triplicate at 37 °C with aeration using a Bioscreen C automated growth analysis system. The curves represent the averaged values from three parallel experiments with a margin of error of less than 5%.

Fig. S7.

Induction of tcyP gene expression by H₂O₂ in exponentially grown WT, ∆mstA, and P_tet-mstA cells as measured by qRT-PCR. The relative expression (y axis) represents the fold change of each mRNA level after treatment of the cells with 2 mM H₂O₂ for 20 min compared with that of the untreated cells (dotted line). Values are means ± SD from three independent experiments.

Fig. S8.

Constitutive expression of the TcyJ transporter does not suppress the negative effect of cysB deletion on H₂S generation in (A) P_tet-mstA or (B) WT, ΔmstA, or ΔcysB cells. Pb(Ac)₂-soaked paper strips show a PbS brown stain as a result of reaction with H₂S. Strips were affixed to the inner wall of a culture tube above the level of the liquid culture of bacteria for 18 h. The values (percentages) are means from three experiments with a margin error of less than 10%.