Escherichia coli cytochrome c peroxidase is a respiratory oxidase that enables the use of hydrogen peroxide as a terminal electron acceptor

Abstract

Microbial cytochrome c peroxidases (Ccp) have been studied for 75 years, but their physiological roles are unclear. Ccps are located in the periplasms of bacteria and the mitochondrial intermembrane spaces of fungi. In this study, Ccp is demonstrated to be a significant degrader of hydrogen peroxide in anoxic Escherichia coli Intriguingly, ccp transcription requires both the presence of H₂O₂ and the absence of O₂ Experiments show that Ccp lacks enough activity to shield the cytoplasm from exogenous H₂O₂ However, it receives electrons from the quinone pool, and its flux rate approximates flow to other anaerobic electron acceptors. Indeed, Ccp enabled E. coli to grow on a nonfermentable carbon source when H₂O₂ was supplied. Salmonella behaved similarly. This role rationalizes ccp repression in oxic environments. We speculate that micromolar H₂O₂ is created both biologically and abiotically at natural oxic/anoxic interfaces. The OxyR response appears to exploit this H₂O₂ as a terminal oxidant while simultaneously defending the cell against its toxicity.

Keywords: OxyR; anaerobic respiration; oxidative stress.

Fig. 1.

Under anoxic conditions, a new H₂O₂-degrading activity appears. Wild-type (MG1655) and Hpx⁻ (LC106) cells were grown and assayed aerobically or anaerobically. Rates of H₂O₂ scavenging were measured as described. Error bars in this and subsequent figures represent SEM of three independent experiments. Asterisks represent statistical significance [*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; not significant (ns), P > 0.05].

Fig. S1.

H₂O₂ scavenging in Hpx⁻ strains is not due to reactions with secreted metabolites or intracellular Fenton chemistry. The rate of H₂O₂ scavenging was measured with Hpx⁻ cells, with the filtrate of anaerobically grown Hpx⁻ (LC106) cells, or with cells treated with 1 mM dipyridyl (a cell-permeable iron chelator). ND, the rate was below the detection limit (0.1 μM H₂O₂/min-OD). ns, not significant compared with Hpx⁻.

Fig. 2.

Ccp scavenges H₂O₂ in anoxic Hpx⁻ cells. The Hpx⁻ parent strain and mutant derivatives were grown anaerobically, and the rate of H₂O₂ scavenging was measured. Cross-hatched bars: The cytochrome c maturation machinery is required for Ccp activity. Spotted bar, far right: the Hpx⁻ Δccp mutant was genetically complemented using pACYC184-ccp under its own promoter. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05). The single mutants other than Hpx⁻ Δccp and Hpx⁻ Δccm were not significantly different from Hpx⁻. None of the double mutants were significantly different from Hpx⁻ Δccp. Strains used: Hpx⁻ (LC106), Hpx⁻ Δbcp (MK150), Hpx⁻ Δtpx (MK154), Hpx⁻ ΔbtuE (MK158), Hpx⁻ Δccp (MK146), Hpx⁻ Δcyd Δcyo (SSK53), Hpx⁻ ΔappBC (MK180), Hpx⁻ ΔosmC (MK208), Hpx⁻ Δccp Δbcp (MK172), Hpx⁻ Δccp Δtpx (MK174), Hpx⁻ Δccp ΔbtuE (MK176), Hpx⁻ Δccp ΔappBC (MK182), Hpx⁻ Δccp Δcyd (MK164), Hpx⁻ Δccp Δcyo (MK166), Hpx⁻ Δccp Δcyd Δcyo (MK170), Hpx⁻ Δccp ΔosmC (MK210), Hpx⁻ Δccm (MK198), Hpx⁻ Δccp Δccm (MK418), and Hpx⁻ Δccp pAcyc184-ccp (MK430).

Fig. 3.

(A) Ccp has a low effective K_m for H₂O₂. Rates of anoxic H₂O₂ scavenging were measured with Hpx⁻ (LC106) and Hpx⁻ Δccp (MK146) cells. Triplicate measurements determined K_m(app) to be 5.2 ± 0.6 μM. A single trial is shown here. (B) Respiratory quinones are required for Ccp function. The Hpx⁻ strain (LC106) and its derivatives lacking ubiquinone (LC148); menaquinone (SSK6); both (LC160); ccp (MK146); or ubiquinone, menaquinone, and ccp (MK184) were grown anaerobically in LB medium, and the rates of H₂O₂ scavenging were measured. Asterisks represent statistical significance compared with the Hpx⁻ strain (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. S2.

Cytochrome bo and bd oxidases require millimolar doses of H₂O₂ for effective turnover. The rates of NADH oxidation by the bo and bd oxidases were measured using inverted membrane vesicles under anoxic conditions. The vesicles were prepared from Δcyd (AL454) or Δcyo (JEM1513) strains. The figure shows a representative curve from three independent experiments. The cytochrome bd-1 oxidase also exhibits some catalase activity, but this action similarly requires millimolar levels of H₂O₂ for rapid turnover (82).

Fig. S3.

Nap and Nrf are not required for H₂O₂ scavenging. The rate of H₂O₂ scavenging was measured using anoxic Hpx⁻ Δnrf (MK218) and Hpx⁻ Δnap (MK212) cells. The slight decrement in the Δnap mutant may be due to a polar effect upon downstream cytochrome c biosynthetic genes. ns, not significant compared with Hpx⁻.

Fig. 4.

Strong ccp expression requires both the absence of oxygen and the presence of H₂O₂. (A) The rates of H₂O₂ scavenging were measured in Hpx⁻ cells grown in oxic or anoxic conditions. Where indicated, cells were incubated with 40 μM H₂O₂ for 1 h before assay. (B) β-galactosidase activity from the transcriptional ccp′–lacZ⁺ reporter fusion was measured in Hpx⁻ (MK250), Hpx⁻ Δfnr (MK274), and Hpx⁻ ΔoxyR (MK278) mutants grown in oxic or anoxic media. Where indicated, cells were incubated with 40 μM H₂O₂ for 1 h before harvesting. (C) β-galactosidase activity from the fusion was measured in wild-type cells containing either a pACYC184-oxyR2 plasmid (expressing a constitutively active form of OxyR) (MK346) or empty vector (MK344), grown anaerobically. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. S4.

ccp expression requires both the absence of oxygen and the presence of H₂O₂. β-galactosidase activity from the transcriptional ccp′–lacZ⁺ reporter fusion was measured in WT (MK246) cells grown in oxic or anoxic media. Where indicated, cells were incubated with 40 μM H₂O₂ for 1 h before harvesting. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. S5.

Under anoxic conditions, cytoplasmic enzymes remain the primary scavengers of H₂O₂. (A) β-galactosidase activity of the transcriptional ahpCF′–lacZ⁺ reporter fusion was measured in Hpx⁻ (MK188) cells grown anaerobically or aerobically. (B) The rate of H₂O₂ scavenging was measured in anoxic wild-type (MG1655) cells, the Hpx⁻ (LC106) mutant, and the Δccp (MK416) mutant. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. 5.

Ccp cannot protect the cytoplasm from exogenous H₂O₂. (A) Wild-type (MG1655), Δccp (MK416), Hpx⁻ (LC106), and Hpx⁻ Δccp (MK146) mutant cells were grown to OD₆₀₀ ∼0.1 and spread on a plate. Disks soaked in H₂O₂ were put on the plate, and the diameter of zone of inhibition was measured after 24 h. (B) Data from A are shown as bar graphs. (C) Modeling (SI Materials and Methods) shows that H₂O₂ exchange between the external environment and the periplasm is too fast for Ccp to significantly diminish the periplasmic H₂O₂ level. Therefore, Ccp has minimal effect on H₂O₂ entry into the cytoplasm. Steady-state fluxes (in %) are calculated relative to the rate of H₂O₂ entry into the periplasm. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. S6.

A Δccp mutant is not more sensitive to millimolar doses of H₂O₂ than is the wild-type strain. Wild-type (MG1655) and Δccp mutant (MK416) cells were grown to OD₆₀₀ ∼0.3 and challenged with 2.5 mM H₂O₂. Percent survival was calculated based on cfus. In contrast to an earlier report (28), we did not observe any difference in the sensitivities of these strains.

Fig. 6.

Oxygen and nitrate block both Ccp activity and synthesis. (A) Hpx⁻ (LC106) cells were grown anaerobically, and the rate of H₂O₂ scavenging was measured in absence or presence of oxygen. (B) Hpx⁻ (LC106) cells were both grown and assayed for H₂O₂ scavenging either without an electron acceptor or in the presence of fumarate, nitrate, or oxygen. (C) Expression of the transcriptional ccp′–lacZ⁺ reporter fusion in Hpx⁻ (MK250) cells grown either without an electron acceptor or in the presence of fumarate, nitrate, or oxygen. Cells were incubated with 40 μM H₂O₂ for 1 h before harvesting. Asterisks represent statistical significance compared with anoxic assay in A and no terminal acceptor in B and C (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. S7.

NarL, NarP, and ArcA do not mediate the inhibition of ccp transcription by nitrate. β-galactosidase activity of the transcriptional ccp′–lacZ⁺ reporter fusion was measured in Hpx⁻ (MK250), Hpx⁻ ΔnarL (MK404), Hpx⁻ ΔnarP (MK412), and Hpx⁻ ΔarcA (MK408) strains cultured with or without nitrate. If indicated, cells were incubated with 40 μM H₂O₂ for 1 h before harvesting. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. 7.

Rate of respiration of H₂O₂ is comparable to that of fumarate. Comparison of respiration rates by wild-type cells with different electron acceptors in glycerol medium. The grid pattern extension of the H₂O₂ bar represents calculated V_max, because unlike the other acceptors, H₂O₂ was provided at a subsaturating (10 μM) concentration. See Materials and Methods for details. The rate of respiration by fumarate was calculated in SI Materials and Methods.

Fig. 8.

Ccp allows respiratory growth using H₂O₂ as the final electron acceptor. (A) Wild-type cells were grown in anoxic glycerol medium without any electron acceptor (○) or in the presence of 40 mM nitrate (▲), 25 mM fumarate (▪), or 5 μM H₂O₂ (●). Viable cells (cfus) were determined at different time points. Residual growth in the absence of H₂O₂ appears to be due to trace oxygen in the anaerobic chamber (Fig. S8). (B) WT and Δccp (MK416) strains were grown in anoxic glycerol medium in the presence or absence of 5 μM H₂O₂. (C) Δccp mutants were complemented with the ccp gene under its own promoter in a plasmid. Strains with pACYC184-ccp (MK436) and the empty vector (MK432) were grown in presence or absence of 5 μM H₂O₂. (D) The initial (t = 0) and the final (14-h) time points from three biological replicates of A were used to calculate the growth ratio for each growth condition. (E) The initial (t = 0) and the final (14-h) time points from three biological replicates of B were used to calculate the growth ratio for each growth condition. (F) The initial (t = 0) and the final (14-h) time points from three biological replicates of C were used to calculate the growth ratio for each growth condition. Asterisks represent statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. 9.

E. coli B and S. typhimurium LT2 can grow using H₂O₂ as the final electron acceptor. (A) Wild-type E. coli B cells were grown in anoxic glycerol medium without any electron acceptor (○) or in the presence of 5 μM H₂O₂ (●). Viable cells (cfus) were determined at different time points. (B) The initial (t = 0) and the final (14-h) time points from three biological replicates of A were used to calculate the growth ratio for each growth condition. (C) Wild-type S. typhimurium LT2 was grown in anoxic glycerol medium without any electron acceptor (○) or in the presence of 5 μM H₂O₂ (●). Viable cells (cfus) were determined at different time points. (D) The initial (t = 0) and the final (14-h) time points from three biological replicates of A were used to calculate the growth ratio for each growth condition. Asterisks represent statistical significance compared with growth with no terminal acceptor (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, P > 0.05).

Fig. 10.

Proposed model for Ccp expression and function. H₂O₂ rapidly enters the periplasm through porins but penetrates the inner membrane at a lower rate. Activation of OxyR stimulates synthesis of catalase and NADH peroxidase, which keep the cytoplasmic H₂O₂ concentration low. In anoxic conditions, FNR is also activated, and together it and OxyR induce ccp. Ccp enters the periplasm and allows the respiratory chain to reduce H₂O₂ to H₂O. In isolated cells, Ccp enables anaerobic respiration, but it does not degrade H₂O₂ quickly enough to reduce the periplasmic or cytoplasmic H₂O₂ concentrations.

Fig. S8.

Trace oxygen is the probable source of slight residual anaerobic growth in the absence of added electron acceptors. Growth occurred only at very low cell densities (○); at higher cell densities that would quickly scavenge oxygen, no increase in biomass was perceptible (□). One sample culture is shown from two replicates.