Two sources of endogenous hydrogen peroxide in Escherichia coli

Abstract

Mechanisms of hydrogen peroxide generation in Escherichia coli were investigated using a strain lacking scavenging enzymes. Surprisingly, the deletion of many abundant flavoenzymes that are known to autoxidize in vitro did not substantially lessen overall H(2)O(2) formation. However, H(2)O(2) production diminished by 25-30% when NadB turnover was eliminated. The flavin-dependent desaturating dehydrogenase, NadB uses fumarate as an electron acceptor in anaerobic cells. Experiments showed that aerobic NadB turnover depends upon its oxidation by molecular oxygen, with H(2)O(2) as a product. This reaction appears to be mechanistically adventitious. In contrast, most desaturating dehydrogenases are associated with the respiratory chain and deliver electrons to fumarate anaerobically or oxygen aerobically without the formation of toxic by-products. Presumably, NadB can persist as an H(2)O(2)-generating enzyme because its flux is limited. The anaerobic respiratory enzyme fumarate reductase uses a flavoprotein subunit that is homologous to NadB and accordingly forms substantial H(2)O(2) upon aeration. This tendency is substantially suppressed by cytochrome oxidase. Thus cytochrome d oxidase, which is prevalent among anaerobes, may diminish intracellular H(2)O(2) formation by the anaerobic respiratory chain, whenever these organisms encounter oxygen. These two examples reveal biochemical and physiological arrangements through which evolution has minimized the rate of intracellular oxidant formation.

Figure 1

Suppression of H₂O₂ formation by β-alanine occurs through catabolism rather than pantothenate synthesis. A. In vivo H₂O₂ production by the catalase/peroxidase-deficient parent strain (LC106, triangles) and a panC mutant (SSK62, diamonds). Open symbols: 25 mM β-alanine was included in the medium. B. Rates of H₂O₂ formation by the parent LC106 strain and its puuE (SSK96) and pncB (SSK102) derivatives.

Figure 2

Pathways of NAD⁺ biosynthesis. Top: NadB-dependent formation of NAD⁺ from aspartate. Bottom: Proposed conversion of β-alanine supplements to NAD⁺. The requirement for PuuE and PncB was demonstrated; the involvement of NadA is speculative (see text).

Figure 3

H₂O₂ formation is suppressed by nicotinamide precursors and depends on NadB. A. H₂O₂ formation was measured for the catalase/peroxidase-deficient parent strain LC106 grown aerobically in the presence of 0.6 mM of the indicated supplements. B. H₂O₂ formation by LC106 (catalase/peroxidase mutant) and its nadB (SSK84), nadA (JI426), and nadB nadA (SSK91) derivatives. All strains were supplemented with 1.6 μM nicotinic acid.

Figure 4

Molecular oxygen is a poorer electron acceptor for purified NadB than is fumarate. A. Rate of reduction of oxygen to H₂O₂ by NadB in the absence and presence of fumarate. The dissolved oxygen concentration was 260 μM. B. NadB turnover rate, measured as oxaloacetate (OAA) production, in buffer saturated with air (260 μM oxygen), with 100% oxygen (1.1 mM), or with fumarate in the absence of oxygen.

Figure 5

Molecular oxygen and fumarate compete as NadB substrates in vivo. A. H₂O₂ production was measured for the catalase/peroxidase-deficient parent strain LC106 and its nadB derivative (SSK84) in the presence and absence of 50 mM fumarate supplements. B. Rates of H₂O₂ production as a function of oxygen concentration for catalase/peroxidase deficient LC106 parent strain (diamonds) and its nadB derivative (SSK84, squares). By subtraction, the NadB-dependent component of LC106 H₂O₂ formation was derived (triangles).

Figure 6

Fumarate reductase generates substantial H₂O₂ when E. coli is transferred from anaerobic to aerobic medium. Strains were the catalase/peroxidase-deficient parent strain LC106, its Δfrd derivative (LC126), a fumarate-reductase overproducer (LC106 +pFrd), and a vector control strain (LC106).

Figure 7

The aerobic respiratory system suppresses H₂O₂ formation by reduced fumarate reductase. A. Inverted membrane vesicles (10 μg/ml) were prepared from anaerobic cells that contain fumarate reductase. The rates of H₂O₂ formation were determined during back-reduction of fumarate reductase with succinate, using a catalase/peroxidase-deficient sdhABCD strain with an intact aerobic respiratory system (SSK90) and ones that lacked ubiquinone (ubi, SSK124) or cytochrome d and o oxidases (SSK53). B. H₂O₂ formation by growing cells upon shift from anaerobic to aerobic medium. The full bar indicates the rate of H₂O₂ formation by strains synthesizing fumarate reductase. The black part of the bar indicates the rate of H₂O₂ by matched frd-deficient strains; by subtraction, the gray bar indicates the fumarate-reductase-dependent rate of H₂O₂ formation. The ubi mutation increased frd-dependent H₂O₂ production three-fold, both with normal and overexpressing strains.

Figure 8

Model: Association with the respiratory chain avoids H₂O₂ production by the desaturating dehydrogenases. A. Under anaerobic conditions, both soluble NadB and membrane-bound dihydroorotate dehydrogenase (DHOD) deliver electrons to fumarate. B. Upon aeration, fumarate levels fall, and NadB transfers electrons directly to oxygen, generating H₂O₂. Electrons from dihydroorotate dehydrogenase are delivered to oxygen through cytochrome oxidase, without H₂O₂ formation. (Abbreviations: DHO, dihydroorotate; OR, orotate; Q, the quinone pool; Frd, fumarate reductase; fum, fumarate; suc, succinate; asp, asparate; isuc, iminosuccinate.)

Figure 9

Model: Upon aeration, the aerobic respiratory chain pulls electrons away from fumarate reductase and thereby suppresses H₂O₂ formation. (Abbreviations: Cyd, cytochrome d oxidase; Frd, fumarate reductase; Ndh, NADH dehydrogenase; MQ, menaquinone; UQ and UQH₂, oxidized and reduced ubiquinone; fum, fumarate; succ, succinate.)