Cooperative Roles of Nitric Oxide-Metabolizing Enzymes To Counteract Nitrosative Stress in Enterohemorrhagic Escherichia coli

Enterohemorrhagic Escherichia coli (EHEC) has at least three enzymes, NorV, Hmp, and Hcp, that act independently to lower the toxicity of nitric oxide (NO), a potent antimicrobial molecule. This study aimed to reveal the cooperative roles of these defensive enzymes in EHEC against nitrosative stress.

ABSTRACT

Enterohemorrhagic Escherichia coli (EHEC) has at least three enzymes, NorV, Hmp, and Hcp, that act independently to lower the toxicity of nitric oxide (NO), a potent antimicrobial molecule. This study aimed to reveal the cooperative roles of these defensive enzymes in EHEC against nitrosative stress. Under anaerobic conditions, combined deletion of all three enzymes significantly increased the NO sensitivity of EHEC determined by the growth at late stationary phase; however, the expression of norV restored the NO resistance of EHEC. On the other hand, the growth of Δhmp mutant EHEC was inhibited after early stationary phase, indicating that NorV and Hmp play a cooperative role in anaerobic growth. Under microaerobic conditions, the growth of Δhmp mutant EHEC was inhibited by NO, indicating that Hmp is the enzyme that protects cells from NO stress under microaerobic conditions. When EHEC cells were exposed to a lower concentration of NO, the NO level in bacterial cells of Δhcp mutant EHEC was higher than those of the other EHEC mutants, suggesting that Hcp is effective at regulating NO levels only at a low concentration. These findings of a low level of NO in bacterial cells with hcp indicate that the NO consumption activity of Hcp was suppressed by Hmp at a low range of NO concentrations. Taken together, these results show that the cooperative effects of NO-metabolizing enzymes are regulated by the range of NO concentrations to which the EHEC cells are exposed.

INTRODUCTION

Enterohemorrhagic Escherichia coli (EHEC) is classified into multiple serotypes, among which O157:H7 is the serotype most commonly linked to epidemic and sporadic disease in humans throughout Japan, North America, and parts of Europe. EHEC infections are a primary cause of hemorrhagic colitis and hemolytic-uremic syndrome (HUS) (1–3). The pathogenesis of EHEC infections is associated with the production of Shiga toxins (Stxs) that are identical to the Shiga toxin produced by Shigella dysenteriae type I (4). Stxs produced by EHEC include Stx1 and Stx2, and the nucleotide sequences of these two Stxs are ∼56% identical (5).

In 2006, a high frequency of HUS was seen following an outbreak of the EHEC O157 TW14359 strain caused by consumption of contaminated spinach (6). The percentage (15%) of TW14359 cases that developed HUS was significantly higher than the percentage (4.1%) of all EHEC O157 cases that developed HUS (7). Comparison of the genome sequences between TW14359 and other EHEC O157 strains that caused lower frequencies of HUS identified the intact nitric oxide (NO) reductase gene norV as a candidate gene for virulence (8). There are two types of norV genes in EHEC O157, an intact norV and a deleted norV gene (9–11). The product of deleted norV is an internal 68-amino-acid deletion in the norV coding sequence and did not exhibit NO reductase activity (9, 10). Moreover, intact norV-type EHEC showed better survival in macrophages than deleted norV-type EHEC (10). An evolutionary analysis revealed that intact norV in EHEC O157 was strictly correlated with subgroup C cluster 1, and deleted norV was correlated with subgroup C clusters 2 and 3, but cluster 2 is a minor group (11). EHEC O157 subgroup C cluster 1 and cluster 2 usually produce only Stx2, while almost all the members of EHEC O157 subgroup C cluster 3 produce both Stx1 and Stx2 (11). Epidemiological studies have shown that EHEC O157 strains that produce only Stx2 are more commonly associated with severe human disease than EHEC O157 strains that produce both Stx1 and Stx2 (12–14). Therefore, the intact norV gene is a direct virulence determinant of EHEC O157, even though, with the exception of serotype O157, no EHEC strains carrying the deleted norV gene have been obtained (11). This fact indicates that reducing NO in bacterial cells is important for the growth and survival of EHEC O157 in the intestinal tract during EHEC infection (11).

NO is a water-soluble reactive nitrogen species. It is rapidly oxidized by molecular oxygen but is more stable under anaerobic conditions. It is potentially toxic in biological systems because of its ability to react with a variety of cellular molecules, especially thiol groups and transition metal centers in proteins (15). NO is an important component of the innate immune system against infection, and it is synthesized at high concentrations (∼10⁻⁷ M to 10⁻⁶ M) by the inducible NO synthases (iNOSs) in phagocytic cells (15–18). On the other hand, lower concentrations of NO (∼10⁻⁹ M to 10⁻⁷ M) are produced by certain bacteria, either as a by-product of nitrite reduction to ammonia or as an intermediate of denitrification (19, 20). In fact, the physiological concentrations of NO within the gut are thought to be within the range of ∼10⁻⁷ M to 10⁻⁶ M (21, 22). Therefore, bacteria can potentially be exposed to various concentrations of NO produced by host cells and endogenously generated NO.

Bacteria have evolved specific sensor proteins to survive against nitrosative stress; these proteins detect the NO molecule and switch on the expression of enzymes that rapidly detoxify NO before it reaches lethal levels. NO sensor proteins are iron-containing transcriptional factors that have evolved roles in the cellular response to NO (23). NO interacts with the heme, iron-sulfur clusters, and iron present in NO sensor proteins to form iron-nitrosyl complexes, which likely leads to protein conformational changes and/or changes in DNA affinity that modulate gene expression in response to the presence of NO. In Escherichia coli, two direct NO-sensing proteins, the transcriptional activator NorR and the transcriptional repressor NsrR, appear to exclusively sense NO (24–28). NorR is a σ⁵⁴-dependent transcriptional activator that activates the transcription of the norVW operon in E. coli in response to NO. NorR senses NO directly through a mononuclear nonheme iron center. On the other hand, NsrR contains an iron-sulfur cluster and regulates the expression of several genes in response to NO. NsrR binds to the promoter region upstream of its target gene and represses transcription. NsrR controls the expressions of the flavohemoglobin Hmp and the hybrid cluster protein Hcp in response to NO in E. coli (26, 27, 29).

NO detoxification pathways appear to be widespread among bacteria to defend against NO. Enteric bacteria have developed multiple mechanisms for protecting themselves from NO. Three main methods for detoxifying NO have been identified in E. coli (30–32). The NorV flavorubredoxin and its redox partner protein NorW NADH:flavorubredoxin oxidoreductase constitute an electron transfer chain, with NADH acting as the primary electron donor (33). NorV and NorW exhibit NADH-dependent NO reductase activity, which is considered to be responsible for the detoxification of NO under anaerobic conditions (33). NorV and NorW are encoded in a bicistronic unit, which is under the regulation of NorR. NorV is composed of three structural domains, the flavodiiron structural core, a flavin mononucleotide (FMN)-binding flavodoxin-like module, and a C-terminal rubredoxin module, while NorW contains a flavin adenine dinucleotide (FAD) moiety (33, 34).

The Hmp flavohemoglobin in E. coli catalyzes both a rapid reaction of heme-bound O₂ with NO to form nitrate under aerobic conditions and a slower reduction of heme-bound NO to form N₂O under anaerobic conditions (35, 36). Hmp is made of an N-terminal globin domain and a C-terminal FAD- and NAD-binding domain (37, 38). The heme in the globin domain is covalently linked to the polypeptide chain through a proximal histidine residue (His85) in addition to the conservation of the typical globin fold, and a phenylalanine residue (Phe43) appears to be necessary to keep the heme in the correct orientation within the pocket (37, 38). In the FAD- and NAD-binding domain of Hmp, the aromatic ring of the conserved residue Phe390 is packed against the isoalloxazine central ring of FAD (37, 38).

Hcp is a high-affinity NO reductase that is the major enzyme for reducing NO to N₂O under physiologically relevant conditions (32). Recently, the Stamler group demonstrated that Hcp is an essential enzyme required for S-nitrosylation (39). Hcp contains two types of iron-sulfur clusters unique in biological systems: a [4Fe-4S] cubane cluster and a novel type of hybrid [4Fe-2S-2O] cluster (40). The hcp gene is closely linked to a gene for an NADH-dependent Hcp oxidoreductase, Hcr, which serves as the electron donor for the Hcp in the presence of NADH (40). Hcr contains FAD and a [2Fe-2S] cluster as cofactors (40) and protects Hcp against nitrosylation by NO (32).

In this study, we clarified the cooperative roles of these NO-metabolizing enzymes, NorV, Hmp, and Hcp, in reducing the NO level and protecting against the growth of EHEC bacterial cells in response to nitrosative stress.

RESULTS

Expressions of NO-metabolizing enzyme genes in EHEC in response to NO.

EHEC contains three NO-metabolizing enzymes, NorV, Hmp, and Hcp. To determine the expressions of norV, hmp, and hcp and, thus, the levels of the corresponding enzymes in EHEC following exposure to NO, we constructed EHEC reporter strains using a chromosome-plasmid hybrid bioluminescent reporter system (41). Three single-copy chromosomal norV::luxE fusion, hmp::luxE fusion, and hcp::luxE fusion strains were constructed, and these mutant strains were transformed with the LuxCDAB expression plasmid pluxCDAB3 (41). The levels of luminescence in EHEC reporters depended on the level of luxE expression from each promoter. To determine the effect of the NO-induced expressions of norV, hmp, and hcp on growth conditions, EHEC cells were cultured in LB broth at 37°C for 6 h (early stationary phase) under various growth conditions, and specific luminescence was determined by a luminometer as a measure of the expression of the corresponding gene, followed by normalization by colony counting. Since the concentration of NO in culture following treatment with 200 μM diethylenetriamine NONOate (DETA/NO) under anaerobic conditions was 12 μM (42), the concentration of NO in culture after treatment with 100 μM DETA/NO under anaerobic conditions might be no more than about 10⁻⁶ M, which corresponds to the physiological concentration of NO in the gut (21, 22). In the presence of 100 μM DETA/NO, the expression levels of norV, hmp, and hcp were increased ∼160- to 200-fold, ∼16- to 53-fold, and ∼133- to 163-fold, respectively, under anaerobic growth and aerobic static growth conditions (Fig. 1A to C). However, under aerobic growth conditions with shaking, the expressions of norV and hcp were not increased, although the expression of hmp was increased (Fig. 1A to C). It was previously reported that the hcp promoter is not NO inducible under aerobic conditions (43). Therefore, these results indicated that the culture under aerobic growth with shaking constituted aerobic conditions and that the cultures under both anaerobic growth and aerobic static growth were oxygen limited. To further determine the differences between anaerobic growth and aerobic static growth, we investigated bacterial growth of wild-type, ΔnorV mutant, Δhmp mutant, and Δhcp mutant EHEC strains following treatment with an NO donor. Under aerobic static growth conditions, Δhmp mutant EHEC displayed a reduction of growth 18 h after treatment with 200 μM DETA/NO, although the growth levels of wild-type, ΔnorV mutant, Δhmp mutant, and Δhcp mutant EHEC strains were not inhibited under anaerobic growth conditions (Fig. 1D and E). Since NO detoxification by Hmp occurs most effectively via O₂-dependent NO dioxygenation (36), and NO reductase activity of NorV is exhibited in the absence of an O₂ molecule (44), these results indicate that the culture under aerobic static growth and anaerobic static growth constituted microaerobic and anaerobic conditions, respectively.

FIG 1.

Effect of growth conditions on gene expressions of NO-metabolizing enzymes and bacterial growth in EHEC treated with an NO donor. EHEC cells grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) with or without DETA NONOate (DETA/NO) and grown at 37°C for 6 h for gene expression or 18 h for bacterial growth under anaerobic growth, aerobic static growth, or aerobic growth with shaking at 220 rpm. Relative light units (RLU) and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. Data are the means ± standard deviations of values from three experiments. (A) norV reporter, 100 μM DETA/NO; (B) hmp reporter, 100 μM DETA/NO; (C) hcp reporter, 100 μM DETA/NO; (D) anaerobic growth, 200 μM DETA/NO; (E) aerobic static growth, 200 μM DETA/NO.

EHEC enters the gastrointestinal tract during infection. While the lumen of the intestinal tract is anaerobic, there is a zone of lower oxygenation adjacent to the mucosal surface that is generated by diffusion from the microvillus capillary network (45). Moreover, oxygen levels modulate pathogen virulence during infection (46). Therefore, we analyzed the roles of NO-metabolizing enzymes in EHEC under anaerobic and microaerobic conditions. EHEC reporters were cultured in LB broth at 37°C for 6 h and 18 h (late stationary phase) under anaerobic and microaerobic conditions. In this study, nitrite was used as a source of NO (40, 41). In the presence of 3 mM nitrite, the expression levels of norV, hmp, and hcp were increased ∼7- to 11-fold, ∼10- to 12-fold, and ∼180- to 290-fold under anaerobic conditions, respectively (Fig. 2A to C). Under microaerobic conditions, the expression levels of norV, hmp, and hcp in the presence of 3 mM nitrite were increased ∼4- to 5-fold, ∼10- to 13-fold, and ∼300- to 400-fold, respectively (Fig. 2E to G). The expression of espA in the presence of 3 mM nitrite was not increased under anaerobic and microaerobic conditions (Fig. 2D and H). The product of espA is one of the type III secreted proteins in EHEC (47). These results indicated that among these three genes, induction of hcp is most sensitive to NO under anaerobic and microaerobic conditions.

FIG 2.

Effect of nitrite on gene expressions of NO-metabolizing enzymes in EHEC. EHEC reporters for norV, hmp, or hcp grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) with or without 3 mM nitrite and grown statically at 37°C for 6 h or 18 h under anaerobic and microaerobic conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. The results are expressed relative to the values obtained for EHEC reporters without nitrite treatment after 6 h. Data are the means ± standard deviations of values from three experiments. An asterisk indicates a significant difference (P < 0.01). (A) Anaerobic, norV; (B) anaerobic, hmp; (C) anaerobic, hcp; (D) anaerobic, espA; (E) microaerobic, norV; (F) microaerobic, hmp; (G) microaerobic, hcp; (H) microaerobic, espA.

To further determine the expressions of norV, hmp, and hcp in EHEC following exposure to NO, we treated the EHEC reporter strains with various concentrations of the NO donor DETA/NO for 18 h. At DETA/NO concentrations as low as 0.1 μM, the expression level of hcp was significantly higher than those of norV and hmp under both anaerobic and microaerobic conditions, indicating that the expression of hcp has a lower threshold for a response to NO than the expressions of norV and hmp (Fig. 3A and B). The expression of espA following treatment with 100 μM DETA/NO was not increased under anaerobic or microaerobic conditions (Fig. 3A and B). Under anaerobic and microaerobic conditions, the expression levels of norV, hmp, and hcp following treatment with 10 μM DETA/NO were increased ∼7- to 8-fold, ∼5- to 6-fold, and ∼140- to 200-fold, respectively (Fig. 3A and B). These results were similar to those in the presence of 3 mM nitrite.

FIG 3.

Measurement of the NO responsiveness of NO-metabolizing enzyme genes in EHEC. EHEC reporters for norV, hmp, or hcp grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) containing various concentrations of DETA/NO and grown statically at 37°C for 18 h under anaerobic (A) and microaerobic (B) conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. The results are expressed relative to the values obtained for EHEC reporters without DETA/NO treatment. Data are the means ± standard deviations of values from three experiments. An asterisk indicates a significant difference (P < 0.01).

Role of NO-responsive transcription factors in expressions of NO-metabolizing enzyme genes and growth protection in EHEC in response to NO.

NorR is a σ⁵⁴-dependent transcriptional activator that regulates the expression of norV in E. coli (48). In addition, there is direct evidence for NsrR binding at sites upstream of hmp and hcp, and these are known to be NsrR targets (29, 43, 49). Fnr-binding sites have also been predicted in the promoter regions of hmp and hcp (31). Thus, to determine whether the expressions of these NO-metabolizing enzyme genes depend on the response of NorR, NsrR, and/or Fnr to NO in EHEC, we constructed ΔnorR mutant, ΔnsrR mutant, and Δfnr mutant EHEC reporters and determined their specific luminescences under anaerobic and microaerobic conditions. As expected, the expression of norV in the ΔnorR mutant EHEC reporter was not increased in response to NO under either anaerobic or microaerobic conditions after 18 h (Fig. 4A and D). However, the ΔnorR mutant norV reporter could not be complemented by a plasmid clone containing the norR gene (data not shown). This was consistent with a previous report (50). These results indicate that the transcriptional activator NorR in EHEC positively regulates the expression of norV in response to NO under anaerobic and microaerobic conditions. In addition, the expressions of norV in the ΔnsrR mutant and Δfnr mutant EHEC reporters were slightly and largely increased in response to NO under anaerobic and microaerobic conditions after 18 h, respectively (Fig. 4A and D). The ΔnsrR mutant and Δfnr mutant norV reporters could be complemented by plasmid clones containing the nsrR gene and the fnr gene, respectively (see Fig. S1 in the supplemental material).

FIG 4.

Measurement of the NO responsiveness of NO-metabolizing enzyme genes in NO-sensing transcription factor gene-deficient EHEC. EHEC reporters for norV, hmp, or hcp grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) containing various concentrations of DETA/NO and grown statically at 37°C for 18 h under anaerobic and microaerobic conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. The results are expressed relative to the values obtained for the wild-type EHEC reporter without DETA/NO treatment. Data are the means ± standard deviations of values from three experiments. (A) Anaerobic, norV; (B) anaerobic, hmp; (C) anaerobic, hcp; (D) microaerobic, norV; (E) microaerobic, hmp; (F) microaerobic, hcp.

The expressions of hmp in the wild-type, ΔnorR mutant, and Δfnr mutant EHEC reporters were increased in response to NO under anaerobic and microaerobic conditions after 18 h (Fig. 4B and E). However, the expression level of hmp in the ΔnsrR mutant EHEC reporter was constantly higher than the hmp expression levels in the other strains in both the presence and absence of NO donor treatment and under both anaerobic and aerobic conditions (Fig. 4B and E). The ΔnsrR mutant hmp reporter could be complemented by a plasmid clone containing the nsrR gene (Fig. S1). These facts indicated that the transcriptional repressor NsrR in EHEC repressed the expression of hmp and was inactivated by exposure to NO, resulting in the derepression of hmp.

The expressions of hcp in the wild-type and ΔnorR mutant EHEC reporters were increased in response to NO under anaerobic and microaerobic conditions after 18 h (Fig. 4C and F). However, the expression level of hcp in the ΔnsrR mutant EHEC reporter was constantly higher than those in the other strains either with or without NO donor treatment and under both anaerobic and microaerobic conditions. In contrast, the expression of hcp in the Δfnr mutant EHEC reporter remained at the basal level either with or without NO donor treatment and under both anaerobic and microaerobic conditions (Fig. 4C and F). The ΔnsrR mutant and Δfnr mutant hcp reporters could be complemented by plasmid clones containing the nsrR gene and the fnr gene, respectively (Fig. S1). These results indicated that the transcriptional repressor NsrR in EHEC repressed the expression of hcp and was inactivated by exposure to NO, resulting in the derepression of hcp, and that Fnr was essential for the expression of hcp under anaerobic and microaerobic conditions.

To further examine the roles of NorR, NsrR, and Fnr in EHEC under NO stress conditions, we investigated growth inhibitions of the wild-type, ΔnorR mutant, ΔnsrR mutant, and Δfnr mutant EHEC strains following treatment with an NO donor. Under anaerobic conditions, ΔnorR mutant EHEC displayed a reduction of growth 18 h after treatment with 400 μM DETA/NO, although the growth levels of the wild-type, ΔnsrR mutant, and Δfnr mutant EHEC strains were not inhibited (Fig. S2). Under microaerobic conditions, the growth levels of the wild-type, ΔnorR mutant, ΔnsrR mutant, and Δfnr mutant EHEC strains were not inhibited by 400 μM DETA/NO after 18 h (Fig. S2). These results suggested that NorR was essential for anaerobic growth at a very high concentration of NO (400 μM DETA/NO).

Role of NO-metabolizing enzymes in protecting the growth of EHEC in response to NO.

To examine the role of NO-metabolizing enzymes in EHEC under NO stress conditions, we constructed NO-metabolizing enzyme gene-deficient mutant strains and investigated growth inhibition following treatment with an NO donor under anaerobic and microaerobic conditions. The ΔnorV mutant, Δhmp mutant, and Δhmp mutant strains could be complemented by plasmid clones containing the norV gene, the hmp gene, and the hcp gene, respectively (Fig. S3). Under anaerobic conditions, the growth of Δhmp Δhcp mutant EHEC 18 h after treatment with 100 μM DETA/NO was not inhibited in comparison to the growth of ΔnorV Δhmp Δhcp mutant EHEC (Fig. 5A), indicating that NorV plays an important role in anaerobic growth in the presence of a high concentration of NO. However, although norV was expressed at 6 h (Fig. 2A), the growth levels of the Δhmp mutant and Δhmp Δhcp mutant EHEC strains were inhibited 6 h after treatment with 100 μM DETA/NO under anaerobic conditions (Fig. 5B). These results indicated that NorV and Hmp play a cooperative role for anaerobic growth under a high concentration of NO.

FIG 5.

Effect of NO-metabolizing enzymes on bacterial growth (A to D) and NO levels (E to H) in bacterial cells treated with a high concentration of an NO donor. EHEC NO reporters grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) containing 100 μM or 200 μM DETA/NO and grown statically at 37°C for 6 h or 18 h under anaerobic and aerobic conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. Data are the means ± standard deviations of values from three experiments. An asterisk indicates a significant difference (P < 0.01). (A and E) Anaerobic, 100 μM DETA/NO, 18 h; (B and F) anaerobic, 100 μM DETA/NO, 6 h; (C and G) microaerobic, 200 μM DETA/NO, 18 h; (D and H) microaerobic, 200 μM DETA/NO, 6 h. Closed bars, shaded bars, hatched bars, and open bars indicate wild-type strains, single mutant strains, double mutant strains, and triple mutant strains, respectively.

Under microaerobic conditions, Δhmp mutant EHEC displayed a reduction of microaerobic growth in response to 200 μM DETA/NO after 6 h and 18 h (Fig. 5C and D). The growth of ΔnorV Δhcp mutant EHEC was not inhibited by 200 μM DETA/NO after 6 h and 18 h (Fig. 5C and D). These results indicated that Hmp was important for protection of bacterial growth under microaerobic conditions in the presence of a high concentration of NO.

Role of NO-metabolizing enzymes in the regulation of NO toxicity in EHEC bacterial cells.

EHEC strains that expressed NO-metabolizing enzymes were previously reported to produce a lower level of NO than the corresponding mutant EHEC strains (10, 42). To confirm this, here we determined the NO levels in bacterial cells treated with an NO donor under anaerobic and aerobic conditions. As an NO reporter system, we used the NO reporter plasmid pRPL3, whose luxCDABE expression is dependent on NorR activation of the promoter of norV in response to NO in bacterial cells (10). Under anaerobic conditions, the specific luminescence (relative light units [RLU]/CFU) of the wild-type EHEC NO reporter was much lower than that of the ΔnorV Δhmp Δhcp mutant EHEC NO reporter 6 h and 18 h following treatment with 100 μM DETA/NO, suggesting that NO-metabolizing enzymes in EHEC detoxify NO to yield a low concentration of NO in bacterial cells (Fig. 5E and F). Among the EHEC strains expressing only a single NO-metabolizing enzyme gene, the specific luminescence of the Δhmp Δhcp mutant EHEC NO reporter was lower than those of the ΔnorV Δhcp mutant and ΔnorV Δhmp mutant EHEC NO reporters 18 h after treatment with 100 μM DETA/NO under anaerobic conditions, suggesting that NorV in EHEC was important for NO detoxification at 18 h under anaerobic conditions at a high concentration of NO (Fig. 5E). However, the specific luminescences of the Δhmp mutant and Δhmp Δhcp mutant EHEC NO reporters at 6 h were higher than those 18 h after treatment with 100 μM DETA/NO, suggesting that NorV in EHEC could not detoxify NO in bacterial cells under anaerobic conditions at 6 h in the presence of a high concentration of NO (Fig. 5E and F). This result is consistent with that for growth inhibition following treatment with the same concentration of DETA/NO. On the other hand, under microaerobic conditions, the specific luminescence of the ΔnorV Δhcp mutant EHEC NO reporter was lower than those of the Δhmp Δhcp mutant and ΔnorV Δhmp mutant EHEC NO reporters 6 h and 18 h after treatment with 200 μM DETA/NO, suggesting that Hmp in EHEC is important for NO detoxification under microaerobic conditions at high concentrations of NO (Fig. 5G and H).

Because much lower concentrations of NO were required to induce hcp expression than to induce the expressions of norV and hmp (Fig. 3A and B), we next examined the role of NO-metabolizing enzymes at lower concentrations of NO under anaerobic and microaerobic conditions. Under both anaerobic and microaerobic conditions, the growth of ΔnorV Δhcp mutant EHEC was similar to that of ΔnorV Δhmp Δhcp mutant EHEC and lower than those of the Δhmp Δhcp mutant and ΔnorV Δhmp mutant EHEC strains following treatment with 10 μM DETA/NO or in the presence of 3 mM nitrite (Fig. 6A to D), indicating that NorV and Hcp protect EHEC from growth inhibition at low concentrations of NO under anaerobic and microaerobic conditions.

FIG 6.

Effect of NO-metabolizing enzymes on bacterial growth (A to D) and NO levels (E to H) in bacterial cells treated with a low concentration of an NO donor or in the presence of nitrite. EHEC NO reporters grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) containing 10 μM DETA/NO or 3 mM nitrite and grown statically at 37°C for 18 h under anaerobic and microaerobic conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. Data are the means ± standard deviations of values from three experiments. An asterisk indicates a significant difference (P < 0.01). (A and E) Anaerobic, 10 μM DETA/NO; (B and F) anaerobic, 3 mM nitrite; (C and G) microaerobic, 10 μM DETA/NO; (D and H) microaerobic, 3 mM nitrite. Closed bars, shaded bars, hatched bars, and open bars indicate wild-type strains, single mutant strains, double mutant strains, and triple mutant strains, respectively.

To further examine the roles of NorV and Hcp in EHEC at low concentrations of NO, we investigated the NO levels in bacterial cells at low concentrations of NO under anaerobic and microaerobic conditions. The specific luminescence of the ΔnorV Δhmp mutant EHEC NO reporter was lower than those of the Δhmp Δhcp mutant and ΔnorV Δhcp mutant EHEC NO reporters treated with 10 μM DETA/NO or in the presence of 3 mM nitrite under anaerobic and microaerobic conditions (Fig. 6E to H). These results indicate that Hcp in EHEC is the major enzyme responsible for the reduction of NO in the presence of a low concentration of NO under anaerobic and microaerobic conditions.

Hmp suppresses the NO reductase activity of Hcp at a low concentration of NO.

Unexpectedly, the specific luminescences of the Δhmp mutant and ΔnorV Δhmp mutant EHEC NO reporters were lower than that of the wild-type EHEC NO reporter treated with 10 μM DETA/NO or in the presence of 3 mM nitrite under anaerobic and microaerobic conditions (Fig. 6E to H). The levels of specific luminescence of the Δhmp mutant and ΔnorV Δhmp mutant EHEC NO reporters following treatment with 10 μM DETA/NO or in the presence of 3 mM nitrite were similar to those of the wild-type EHEC NO reporter without NO donor or nitrite treatment (data not shown). However, the level of specific luminescence of the ΔnorV mutant EHEC NO reporter was similar to that of the wild-type EHEC NO reporter (Fig. 6E to H). This result indicated that the NO reductase activity of Hcp in EHEC was suppressed by Hmp at a low concentration of NO under anaerobic and microaerobic conditions. Thus, to further examine the effect of Hmp on NO reductase activity of Hcp at a low concentration of NO, we constructed deleted and point-mutated hmp-expressing EHEC strains (Fig. 7). The aromatic side chain of Phe43 in the globin domain binds nearly parallel to the heme plane, and Hmp is stabilized by a stacking interaction of the isoalloxazine central ring of FAD with Phe390 in the FAD- and NAD-binding domain (37, 38). These deleted and point-mutated hmp-expressing Δhmp ΔnorV Δhcp mutant EHEC strains display lower growth and higher NO levels than ΔnorV Δhcp mutant EHEC following treatment with 200 μM DETA/NO under microaerobic conditions, indicating that mutated Hmp did not exhibit NO consumption activity (Fig. 8A and F). The specific luminescences of the deleted-hmp-expressing Δhmp ΔnorV mutant and point-mutated-hmp-expressing Δhmp ΔnorV mutant EHEC NO reporters were lower than that of the ΔnorV mutant EHEC NO reporter (Fig. 8B and G). These results indicated that the NO reductase activity of Hcp in EHEC was suppressed by the NO consumption activity of Hmp in the presence of a low concentration of NO, although the NO consumption activity of Hmp in ΔnorV Δhcp mutant EHEC could not be detected following treatment with 10 μM DETA/NO under anaerobic and microaerobic conditions (Fig. 6E and G).

FIG 7.

Structure of intact and mutant Hmp. Hmp is made of an N-terminal globin domain and a C-terminal FAD- and NAD-binding domain.

FIG 8.

Effect of Hmp on bacterial growth (A to E) and NO levels in bacterial cells (F to J) following treatment with an NO donor. EHEC NO reporters grown overnight were diluted 1:100 with LB medium–10 mM HEPES (pH 7.0) containing 10 μM or 200 μM DETA/NO and grown statically at 37°C for 18 h under anaerobic and microaerobic conditions. RLU and the number of bacteria were measured by a luminometer and bacterial plate counts (CFU), respectively. Data are the means ± standard deviations of values from three experiments. An asterisk indicates a significant difference (P < 0.01). (A and F) Microaerobic, 200 μM DETA/NO; (B, D, G, and I) anaerobic, 10 μM DETA/NO; (C, E, H, and J) microaerobic, 10 μM DETA/NO.

Hcp may accept electrons from NADH via Hcr (40). Thus, to determine whether Hcr played a role in the suppression of NO reductase activity for Hmp under a low concentration of NO, we measured NO levels in bacterial cells following treatment with 10 μM DETA/NO under anaerobic and microaerobic conditions using Δhcr mutant EHEC. The specific luminescence of the ΔnorV Δhmp Δhcr mutant EHEC NO reporter was lower than that of the ΔnorV Δhmp Δhcp Δhcr mutant EHEC NO reporter (Fig. 8I and J), suggesting that there is another route of electron transfer to Hcp that is independent of Hcr. The specific luminescence of the ΔnorV Δhmp Δhcr mutant EHEC NO reporter was lower than that of the ΔnorV Δhcr mutant EHEC NO reporter under microaerobic conditions (Fig. 8J). However, under anaerobic conditions, the level of specific luminescence of the ΔnorV Δhmp Δhcr mutant EHEC NO reporter was similar to that of the ΔnorV Δhcr mutant EHEC NO reporter (Fig. 8I). This result at least indicated that the NO reductase activity of Hcp in EHEC was suppressed by Hmp in an Hcr-independent manner in the presence of a low concentration of NO under microaerobic conditions. However, we did not obtain any evidence of a direct interaction between the Hmp protein and the Hcp protein by a pulldown assay (data not shown).

DISCUSSION

The data presented in this paper indicate that the NO-induced expression of hcp has a low threshold relative to those of norV and hmp under anaerobic and microaerobic conditions. Moreover, Hcp effectively decreased the NO concentration in bacterial cells at a low concentration of NO under anaerobic and microaerobic conditions. Therefore, the regulation of transcription of hcp in response to a low concentration of NO and the high affinity of Hcp for NO might enable Hcp to reduce NO to N₂O under normal physiological conditions when the concentration of NO in bacterial cells is too low for NorV and Hmp to be expressed or effective. These results are consistent with a previous report (32).

The expressions of hcp and hmp in EHEC were regulated by NsrR (Fig. 4B, C, E, and F), and the promoter regions of both hcp and hmp have NsrR-binding sites in E. coli (Fig. 9) (29, 31). However, the NsrR-regulated expression of hcp was significantly greater than the NsrR-regulated expression of hmp in EHEC. NsrR-binding sites in E. coli fall into two classes: those comprising two copies of an 11-bp inverted repeat with 1-bp spacing and those which have a single copy of an 11-bp element (29). Although the NsrR-binding sites of both hcp and hmp were of the former type, the first and second NsrR-binding sites in hcp satisfied the consensus sequence of the NsrR-binding motif (29). On the other hand, the second NsrR-binding site in hmp was far from satisfying this sequence (29). Therefore, one reason for the differential expression between hcp and hmp in response to NO may have been differences in the sequences of the NsrR-binding sites between hcp and hmp.

FIG 9.

Gene regulation and function of NorV, Hmp, and Hcp in EHEC. Schematic (not to scale) of upstream regions and structures of the norV, hmp, and hcp genes in EHEC. Regulatory proteins and their corresponding cis-acting elements are indicated. The NorR-binding site in the upstream region of norV has been experimentally verified (48). The NsrR-binding sites in the upstream regions of hmp and hcp have been experimentally verified (29). The Fnr-binding sites in the upstream regions of hmp and hcp have been experimentally verified (43, 51). The Fnr-binding site completely overlaps the NsrR-binding site in the hmp promoter. NorV plays a role at a high concentration of NO under anaerobic conditions. Hmp plays a role at a high concentration of NO under anaerobic and microaerobic conditions. Hcp plays a role at a low concentration of NO under anaerobic and microaerobic conditions.

In addition to NsrR-binding sites, hmp and hcp in E. coli were preceded by a binding site of Fnr (Fig. 9) (31, 51). Both hmp and hcp were consititutively expressed at higher levels in the ΔnsrR mutant (Fig. 4B, C, E, and F). Increased expression of hcp, but not hmp, was strongly dependent on Fnr. This indicates that the hcp promoter was activated by Fnr and repressed by NsrR. However, in the case of the hmp promoter, repression by NsrR was the dominant regulatory mechanism in this experiment, although there is some evidence that repression by Fnr also played a role, as reported previously (52, 53). Under our growth conditions, hmp might not have been significantly derepressed in the fnr mutant grown in LB medium, which masked the indirect effect of NsrR. These different effects of the fnr mutant on the expression of hmp may be due to the choice of growth medium (26). At least in part, the consequence is that the absence of an effect or derepression of the hmp promoter in an fnr mutant can be explained by an indirect effect acting through NsrR (26). However, it was previously reported that there is an Fnr-binding site in the hmp promoter that can be bound by Fnr in vitro (Fig. 9) (51). Since the Fnr-binding site completely overlaps the NsrR-binding site in the hmp promoter (Fig. 9) (49), Fnr might not bind to the hmp promoter when NsrR binds to the hmp promoter in vivo. On the other hand, the Fnr-binding site in the hcp promoter does not overlap the NsrR-binding site (Fig. 9) (31).

The expression of norV in the fnr-deficient reporter was largely increased in response to NO. This was consistent with a previous report in which an Δfnr mutant of E. coli caused an increase in norV-norW transcription without NO treatment (54). Analysis of the region upstream of norV shows that this regulation may be achieved by the binding of Fnr to the Fnr-binding site within the norR coding region (Fig. 9) (31, 54).

Macrophages are important components of the innate immune response. One of the antimicrobial systems of macrophages is the iNOS pathway, which is responsible for the generation of NO (16, 17, 55). The concentration of NO that accumulates in bacterial cells within macrophages and during nitrite reduction has not been reported. However, in a previous study, we revealed that the specific luminescences (RLU/CFU) of the deleted norV-type EHEC NO reporter within macrophages were ∼5 × 10⁻¹ to 10⁰ (42), which correspond to the specific luminescences (RLU/CFU) of the norV-deficient EHEC NO reporter after treatment with more than 100 μM and 200 μM DETA/NO under anaerobic and aerobic conditions, respectively (Fig. 5E to H). In addition, since the maximum steady-state concentration of NO in LB medium following treatment with 200 μM DETA/NO under anaerobic conditions was 12 μM (42), the steady-state concentration of NO after treatment with 100 μM DETA/NO under anaerobic conditions might be no higher than 10⁻⁶ M. On the other hand, in the presence of 3 mM nitrite under anaerobic and aerobic conditions, the specific luminescences (RLU/CFU) of the wild-type EHEC NO reporter were ∼5 × 10⁻⁵ to 10⁻⁴, which correspond to treatment with 10 μM DETA/NO under anaerobic and aerobic conditions (Fig. 6E to H). Therefore, these results suggested that the concentrations of NO generated from the NO donor in this study corresponded to the NO concentrations under physiological conditions.

In EHEC O157, the intact norV-type EHEC strains are more virulent than the deleted norV-type EHEC strains (10). In this study, we show that NorV plays an important role for anaerobic growth in the presence of high concentrations of NO. Since the lumen of the intestinal tract is anaerobic, this result suggests that intact norV in EHEC O157 would be involved in determining clinical outcomes. On the other hand, although the hmp deletion type of E. coli was not observed, we revealed that Hmp was also involved in the protection of anaerobic growth under high concentrations of NO. When oxygen is available, Hmp protects bacteria against NO stress by oxidizing NO to nitrate (35, 56). However, under anaerobic conditions, Hmp also catalyzes a slower reduction of NO, followed by the formation of N₂O (35). Therefore, this result indicated that the NO reductase activity of Hmp was important for anaerobic growth in the presence of high concentrations of NO.

Finally, we show that Hcp and Hmp played cooperative roles for NO consumption in EHEC. Although Hmp alone was not effective at decreasing the NO levels in bacterial cells in the presence of a low concentration of NO, Hmp partially repressed the NO reductase activity of Hcp under anaerobic and aerobic conditions. We revealed that the NO consumption activity of Hmp was necessary for the repression of the NO reductase activity of Hcp at low concentrations of NO. However, NorV, which has NO consumption activity, could not repress the NO consumption activity of Hcp. Since the concerted action of NsrR-dependent Hcp for NO consumption and NsrR-dependent YtfE for the repair of damaged iron centers would maintain the cytoplasmic concentration of NO in the nanomolar range (57), repression of the NO reductase activity of Hcp by Hmp might be important for the coordinate regulation of Hcp and Hmp synthesis by NsrR following exposure to NO. The mechanism by which the NO reductase activity of Hcp is partially repressed in the presence of Hmp under a low concentration of NO remains unknown. However, our results may at least suggest that this mechanism does not interfere with either the interaction between Hcp and Hcr or the transfer of electrons from Hcr to Hcp, since Hmp also partially represses the NO reductase activity of Hcp in hcr-deficient EHEC. Furture studies will be required to explain how Hmp represses the NO reductase activity of Hcp at low concentrations of NO.

MATERIALS AND METHODS

Bacterial strains, plasmids, and oligonucleotides.

The bacterial strains and plasmids used in this study are listed in Tables 1 and 2. The oligonucleotides used for this study are listed in Table S1 in the supplemental material.

TABLE 1.

Bacterial strains

Strain	Relevant characteristics	Reference or source
K47	Enterohemorrhagic Escherichia coli O157:H7; stx2, intact norV type	10
K47(−2)S3	K47; deletion of stx2	This study
K47(−2)VE1	K47; deletion of stx2, insertion of luxE downstream of the norV promoter	This study
K47(−2nor)VE1	K47; deletion of stx2 and norR, insertion of luxE downstream of the norV promoter	This study
K47(−2nsr)VE1	K47; deletion of stx2 and nsrR, insertion of luxE downstream of the norV promoter	This study
K47(−2fnr)VE1	K47; deletion of stx2 and fnr, insertion of luxE downstream of the norV promoter	This study
K47(−2)hmpE1	K47; deletion of stx2, insertion of luxE downstream of the hmp promoter	This study
K47(−2nor)hmpE1	K47; deletion of stx2 and norR, insertion of luxE downstream of the hmp promoter	This study
K47(−2nsr)hmpE1	K47; deletion of stx2 and nsrR, insertion of luxE downstream of the hmp promoter	This study
K47(−2fnr)hmpE3	K47; deletion of stx2 and fnr, insertion of luxE downstream of the hmp promoter	This study
K47(−2)hcpE3	K47; deletion of stx2, insertion of luxE downstream of the hcp promoter	This study
K47(−2nor)hcpE1	K47; deletion of stx2 and norR, insertion of luxE downstream of the hcp promoter	This study
K47(−2nsr)hcpE1	K47; deletion of stx2 and nsrR, insertion of luxE downstream of the hcp promoter	This study
K47(−2fnr)hcpE1	K47; deletion of stx2 and fnr, insertion of luxE downstream of the hcp promoter	This study
K47(−2)espE1	K47; deletion of stx2, insertion of luxE downstream of the espA promoter	This study
K47(−2nor)SK1	K47; deletion of stx2 and norR	This study
K47(−2nsr)SK6	K47; deletion of stx2 and nsrR	This study
K47(−2fnr)SK1	K47; deletion of stx2 and fnr	This study
K47(−2V)SC4	K47; deletion of stx2 and norV	This study
K47(−2H)SC2	K47; deletion of stx2 and hmp	This study
K47(−2C)SK5	K47; deletion of stx2 and hcp-hcr	This study
K47(−2VH)KSC2	K47; deletion of stx2, norV, and hmp	This study
K47(−2VC)SCK1	K47; deletion of stx2, norV, and hcp-hcr	This study
K47(−2HC)SCK1	K47; deletion of stx2, hmp, and hcp-hcr	This study
K47(−2VHC)SCK3	K47; deletion of stx2, norV, hmp, and hcp-hcr	This study
K47(−2VC)H100	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(1–300)	This study
K47(−2V)H100	K47; deletion of stx2 and norV, replacement of mutant hmp(1–300)	This study
K47(−2VC)H305	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(1–915)	This study
K47(−2V)H305	K47; deletion of stx2 and norV, replacement of mutant hmp(1–915)	This study
K47(−2HVC)hR43A1	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(R43A)	This study
K47(−2HV)hR43A1	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(R43A)	This study
K47(−2HVC)hR390A1	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(R390A)	This study
K47(−2HV)hR390A1	K47; deletion of stx2, norV, and hcp-hcr, replacement of mutant hmp(R390A)	This study
K47(−2VR)SK2	K47; deletion of stx2, norV, and hcr	This study
K47(−2VHR)SCK1	K47; deletion of stx2, norV, hmp, and hcr	This study
BL21(DE3)	E. coli B; F− dcm ompT hsdS(rB− mB−) gal λ (DE3)	Nippon Gene
DH5α(λpir)	recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 deoR (lacZYA-argF)U169 λpir+	58

TABLE 2.

Plasmids^a

Plasmid	Relevant characteristic(s)	Reference or source
pTVm2	NorV-His expression plasmid; Ampr; Ptrc	10
pThmp1	Hmp-His expression plasmid; Ampr; Ptrc	This study
pBhcp1	Hcp expression plasmid; Cmr; PBAD	This study
pBnorR2	NorR expression plasmid; Cmr; PBAD	This study
pBnsrR2	NsrR expression plasmid; Cmr; PBAD	This study
pBfnr1	Fnr expression plasmid; Cmr; PBAD	This study
pTrcHis 2C	Expression plasmid; Ampr; Ptrc	Invitrogen
pBAD33	Expression plasmid; p15-ori PBAD; Cmr	National BioResource Project (NIG) (Japan)
pRPL3	NO reporter plasmid; Ampr	10
phmpC10	Suicide plasmid; hmp and loxP-flanked Cmr cassette	This study
phmp100C2	Suicide plasmid; deleted hmp(1–300) and loxP-flanked Cmr cassette	This study
phmp305C15	Suicide plasmid; deleted hmp(1–915) and loxP-flanked Cmr cassette	This study
phmpCF43A1	Suicide plasmid; mutated hmp(F43A) and loxP-flanked Cmr cassette	This study
phmpCF390A1	Suicide plasmid; mutated hmp(F390A) and loxP-flanked Cmr cassette	This study
pRed/ET(amp)	Red/ET expression plasmid; Ampr	Gene Bridges GmbH
pFRT-Kan	Suicide plasmid; FRT-flanked PGK-gb2-neo cassette	Gene Bridges GmbH
ploxCm1	Suicide plasmid; loxP-flanked Cmr cassette	41
ploxSm11	Suicide plasmid; loxP-flanked Smr cassette	This study
pAH144	Source of Smr cassette	National BioResource Project
pCreA1	Thermosensitive Cre expression plasmid; Ampr	41
pFT-A	Thermosensitive FLP expression plasmid; Ampr	National BioResource Project
pLCE19	Suicide plasmid; luxE and loxP-flanked Cmr cassette	41
pluxCDAB3	LuxCDAB expression plasmid; Ampr; Plac	41

^aAmp^r, ampicillin resistant; Cm^r, chloramphenicol resistant; Sm^r, streptomycin resistant.

Culture conditions.

EHEC strains grown overnight were diluted 1:100 with LB broth containing 10 mM HEPES (pH 7.0). For anaerobic conditions, EHEC strains were grown statically at 37°C using an Anaero Pack-Anaero 5% system (Mitsubishi Gas Chemical Company, Tokyo, Japan). For microaerobic conditions, EHEC strains were grown statically at 37°C in an aerobic culture. For aerobic conditions, EHEC strains were grown with shaking at 220 rpm at 37°C in an aerobic culture.

Gene reporter assay, growth inhibition assay, and NO reporter assay.

For the gene reporter assay, mutant EHEC strains containing a single-copy chromosomal target gene-luxE fusion construct were transformed using the LuxCDAB expression plasmid pluxCDAB3 (41). For the growth inhibition assay and NO reporter assay, mutant EHEC strains were transformed using the NO reporter plasmid pRPL3 (10). Relative light units (RLU) and the number of bacteria were measured by using a Glomax 20/20 luminometer (Promega, Madison, WI) and bacterial plate counts (CFU), respectively.

Reagents and media.

NO was generated by an NO donor, namely, diethylenetriamine NONOate (DETA/NO) (Cayman Chemical Company, Ann Arbor, MI). The half-life of DETA/NO is 20 h at 37°C in 0.1 M phosphate buffer (pH 7.4). NaNO₂ was purchased from Wako (Tokyo, Japan).

Plasmid construction.

To construct plasmids expressing Hmp, hmp was amplified by PCR from the genomic DNA of the EHEC K47 strain using the primer set P1032-P1033 (Table S1). A DNA fragment was cleaved with NcoI and HindIII and then inserted into these sites in pTrcHis 2C to yield pThmp1 expressing His-tagged Hmp (Table 2 and Table S2). To construct plasmids expressing Hcp-Hcr, NorR, NsrR, and Fnr, the hcp, norR, nsrR, and fnr genes were amplified by PCR from the genomic DNA of the EHEC EDL933 strain using the primer sets P1345-P1346, P1339-P1340, P1341-P1342, and P1343-P1344, respectively (Table S1). Each DNA fragment was cleaved with SacI and SalI and then inserted into these sites in pBAD33 to yield pBhcp1 expressing Hcp-Hcr, pBnorR2 expressing NorR, pBnsrR2 expressing NsrR, and pBfnr1 expressing Fnr, respectively (Table 2 and Table S2). To construct a template plasmid for homologous recombination using a Red/ET recombination system (Gene Bridges GmbH, Heidelberg, Germany), a DNA fragment containing a streptomycin-resistant (Sm^r) gene was obtained by PCR from the plasmid DNA of pAH144 using the primer set P730-P731 (Table S1) and cloned into a DNA fragment that was amplified by PCR using the primer set P736-P737 (Table S1) from plasmid DNA of ploxCm1 (41) to yield ploxSm11 (Table 2 and Table S2). To construct template plasmids for hmp mutation, DNA fragments containing intact and deletion hmp genes were amplified by PCR from the genomic DNA of K47 using the primer sets P1296-P1297, P1296-P1298, and P1296-P1299 (Tables S1 and S2). These fragments were cloned by the ligation of a NotI fragment into ploxCm1 to yield the template hmp plasmids phmpC10, phmp100C2, and phmp305C15 (Table 2 and Table S2). In addition, to construct template plasmids for hmp point mutation, a PrimeSTAR mutagenesis kit (TaKaRa Bio Inc., Shiga, Japan) was used with the plasmid DNA of phmpC10 as a template and primer sets P1310-P1311 and P1312-P1313 (Table S1) to yield the template hmp plasmids phmpCF43A1 and phmpCF390A1, respectively (Table 2 and Table S2). These plasmids were confirmed using restriction digestion and DNA sequencing.

Strain construction.

Mutant strains were derivatives of EHEC O157:H7 K47 and were produced using a Red/ET recombination system (Table 1 and Table S3). Briefly, for construction of the deletion and replacement mutants, PCR primers containing 50 bp immediately upstream and downstream, respectively, of the target locus were used to amplify the DNA fragment containing the FLP recombination target (FRT)-flanked PGK-gb2-neo (kanamycin-resistant [Kan^r]) cassette, a loxP-flanked chloramphenicol-resistant (Cm^r) cassette, or a loxP-flanked Sm^r cassette found in the appropriate plasmids (Table S3).

For construction of reporter strains, PCR primers containing 50 bp immediately upstream and downstream, respectively, of the target locus were used to amplify the DNA fragment containing the promoterless RBS-luxE gene and loxP-flanked Cm^r cassette in the pLCE19 plasmid (Table S3). The resulting linear PCR product was electroporated with the appropriate parent strain previously transformed with plasmid pRed/ET(amp) (Table S3). Recombinants containing appropriate antibiotic-resistant elements in place of the target locus were selected on appropriate antibiotic plates and confirmed by PCR. Deletions of antibiotic-resistant elements were accomplished by transformation with plasmid pCreA1 or pFT-A, resulting in antibiotic-sensitive mutants that were verified by PCR. The plasmids pRed/ET(amp), pCreA1, and pFT-A are temperature sensitive and were cured by overnight growth at 37°C.

Statistics.

An unpaired t test was used to determine significant differences between two treatment groups. One-way analysis of variance (ANOVA) with a Student-Newman-Keuls multiple-comparison test was used to analyze significant differences among three or more groups.