Magnesium Uptake by CorA Is Essential for Viability of the Gastric Pathogen Helicobacter pylori

Abstract

We show here that Mg²⁺ acquisition by CorA is essential for Helicobacter pylori in vitro, as corA mutants did not grow in media without Mg²⁺ supplementation. Complementation analysis performed with an Escherichia coli corA mutant revealed that H. pylori CorA transports nickel and cobalt in addition to Mg²⁺. However, Mg²⁺ is the dominant CorA substrate, as the corA mutation affected neither cobalt and nickel resistance nor nickel induction of urease in H. pylori. The drastic Mg²⁺ requirement (20 mM) of H. pylori corA mutants indicates that CorA plays a key role in the adaptation to the low-Mg²⁺ conditions predominant in the gastric environment.

Magnesium (Mg²⁺) is a cofactor of many enzymes involved in central biochemical pathways which are essential for bacterial viability (20, 23). In order to overcome the Mg²⁺ limitation within the human host, pathogenic bacteria express specific Mg²⁺ uptake systems. In Salmonella enterica serovar Typhimurium and Escherichia coli, high-affinity Mg²⁺ uptake is mainly mediated by the inner membrane protein CorA, which also imports nickel and cobalt (16, 23). The E. coli CorA protein was shown to be involved in iron uptake (13). Enterobacteria have multiple Mg²⁺ uptake systems, and thus corA mutants display a reduced Mg²⁺ transport capacity as well as increased cobalt and nickel resistance (23). The cation metabolism of the gastric pathogen Helicobacter pylori (10), which causes various diseases of the upper gastrointestinal tract, is of substantial importance for survival in the hostile and changing environment of the gastric mucosa (18, 21). Mechanisms involved in maintaining cation homeostasis were shown to be required for effective gastric colonization in animal models (25). Although the essential biological functions of Mg²⁺ point towards a relevance of Mg²⁺ acquisition in the adaptation to the gastric environment, proteins involved in H. pylori Mg²⁺ uptake and metabolism have not been studied in detail. A single corA gene homolog (HP1344/JHP1263) was identified in the annotated genome sequences of H. pylori strains 26695 (24) and J99 (1). Furthermore, a possible nickel uptake function of H. pylori CorA might be of substantial importance because H. pylori expresses high amounts of the nickel metalloenzyme urease for neutralization of gastric acid (14, 18, 19). Nickel induces both urease expression and activity (26). Therefore, we investigated H. pylori CorA functions by analysis of corA mutants as well as by overexpression of the H. pylori corA gene in an E. coli corA mutant. For the experiments, H. pylori was routinely grown under microaerobic conditions as described earlier (6, 9). DNA manipulation and transformation were performed according to standard protocols for E. coli (2) and H. pylori (11). The complete corA gene (Fig. 1) was amplified by PCR from isolated DNA of H. pylori strain 26695 using primer oligonucleotides CORA-L1 (ATATAGGGCCTGCGAGCTTG) and -R1 (CGAGCGATCATAGCCAGACC) and cloned into plasmid pZERO-1 (Table 1). The resulting plasmids pCORA+ and pCORA− contain the H. pylori corA gene in and against the direction of the lacZ promoter present in the plasmid, respectively (Fig. 1; Table 1). For the growth experiments or the determinations of the MICs of Mg²⁺ and of various metals, E. coli and H. pylori were precultured to optical densities at 550 nm (OD₅₅₀) and 600 nm (OD₆₀₀) of 1.0 in Luria-Bertani (LB) and brucella broth (Oxoid) supplemented with 5% fetal calf serum (BBF; Gibco), respectively. For growth experiments, bacterial suspensions were subsequently diluted 1:100 in test media containing defined supplements.

FIG. 1.

Genetic organization and mutagenesis of H. pylori corA with overview of the corA region in H. pylori strain 26695. Genes and primer binding sites are marked by gray and black arrows, respectively. The insertion site of the cat cassette (white arrow) is indicated. Genes are numbered according to the annotated genome sequence (24). The orientation of the DNA region cloned in plasmids pCORA+ and pCORA− with respect to the lacZ promoter (P_lacZ) present in the cloning vector pZERO-1 is indicated.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristicsa	Source(s)
Strains
H. pylori
1061	wt, cag−	5, 6, 8
1061-CORA1	1061, corA::cat, Cmr	This study
26695	wt, cag+	24
26695-CORA1	26695, corA::cat, Cmr	This study
26695-CORA2	26695, corA::Pcat, Cmr	This study
E. coli
MC4100	araD139 Δ(argF-lac)169 flhD5301, fruA25 relA1 rpsL150 rbsR22 deoC1	7
H5324	MC4100, corA::Tn10, Tetr	13
Plasmids
pZERO-1	Cloning vector, MCS in lacZ′, zeo, Zeor	Invitrogen
pZERO-2	Cloning vector, MCS in lacZ′, neo, Kmr	Invitrogen
pCORA+	pZERO-1 with the corA gene of H. pylori strain 26695 cloned in the direction of the lacZ promoter, Zeor	This study
pCORA−	pZERO-1 with the corA gene of H. pylori strain 26695 cloned against the direction of the lacZ promoter, Zeor	This study
pCORA-CAT	pZERO-2 carrying a corA::cat fusion, Cmr	This study
pCORA-PCAT	pZERO-2 carrying a corA::Pcat fusion, Cmr	This study
pHel3	H. pylori shuttle vector, neo, Kmr	15
pHel3-CORA	pHel3 without the E. coli origin of replication fused with pCORA+, Kmr Zeor	This study

^aMCS, multiple cloning site; cag, cag pathogenicity island.

Isolation of H. pylori corA mutants in the presence of Mg²⁺.

For insertional inactivation of the H. pylori corA gene, we constructed the plasmids pCORA-PCAT and pCORA-CAT, which contain the chloramphenicol acetyltransferase gene cat with (Pcat) and without its own promoter inserted in the corA gene, respectively (Fig. 1; Table 1). The Pcat and cat sequences amplified by PCR with primers CAT-L1 (5′-TCCGGTTTTTGTTAATCCGCC) or CAT-L2 (TCCGAGATTTTCAGGAG) in combination with primer CAT-R1 (TTACGCCCCGCCCTGCCA), respectively, were fused to corA upstream and downstream DNA regions by megaprimer PCR (22). Therefore, corA flanking regions were amplified from DNA of H. pylori strain 26695 with primers carrying 5′ extensions complementary to the 5′ and 3′ ends of the Pcat and cat cassettes (underlined below), respectively. The PCR products generated with primer pairs CORA-L1 and CATCORA-R1 (CTCCTGAAAATCTCGGAAAGTCCCATAATAGATCGTG) or PCATCORA-L1 (GGCGGATTAACAAAAACCGGAAGTCCCATAATAGATCGTG) and CATCORA-L1 (TGGCAGGGCGGGGCGTAATGAAGACGGGTTTGATATTC)/CORA-R1 were purified and subsequently mixed with PCR-amplified Pcat or cat cassettes to work as megaprimers in a second PCR containing the flanking primers CORA-L1 and -R1. Megaprimer PCR products carrying Pcat or cat inserted into the corA gene were cloned into plasmid pZERO-2 (Invitrogen), resulting in plasmids pCORA-PCAT and pCORA-CAT, respectively (Fig. 1). Due to the pronounced interstrain variability of H. pylori we analyzed corA functions in H. pylori strains 26695 and 1061, which contain (24) and lack (8) the entire cag pathogenicity island, respectively (Table 1). To achieve inactivation of corA by site-directed mutagenesis, the plasmids pCORA-PCAT and pCORA-CAT (Table 1; Fig. 1) were introduced into H. pylori strains 26695 and 1061 by electroporation and by natural transformation, respectively. The fact that repeated transformation and subsequent selection for transformants did not result in colonies displaying a Cm^r phenotype provided evidence that the corA gene is essential for H. pylori viability in vitro. To avoid a possible killing of corA mutants by Mg²⁺ depletion, we supplemented the Dent agar with 20 mM Mg²⁺ (catalog no. 105833; Merck). Under these conditions, H. pylori corA::cat and corA::Pcat mutants could be selected. After PCR analysis revealed correct insertion of the cat and Pcat cassettes in the corA gene, these mutants (Table 1) were further investigated.

Determination of the Mg²⁺ requirement of H. pylori corA mutants.

The Mg²⁺ requirement caused by the corA mutation was further investigated by growth experiments. The corA mutants of both H. pylori strains did not grow in unsupplemented broth, indicating that the Mg²⁺ concentration of 1.7 mM in BBF (determined with a Sigma kit; catalog no. 595-M) is not sufficient to restore Mg²⁺ acquisition (Fig. 2). Supplementation of BBF with 20 mM MgCl₂ enabled the corA mutants to grow, whereas CaCl₂ and NaCl added at identical concentrations as controls had no effect (data not shown). Precultivation in BBF with 20 mM Mg²⁺ and subsequent monitoring of growth in BBF with defined Mg²⁺ concentrations revealed that the corA mutation causes in both H. pylori strains 1061 and 26695 a drastic Mg²⁺ requirement in the range of 20 mM (Fig. 2). The strong promoter in front of the cat gene had no effect on the growth behavior of the H. pylori 26695 corA mutants (Fig. 2B), providing evidence that the Mg²⁺ requirement is caused by the corA mutation and not by polar effects on the downstream genes (Fig. 1). However, as the corA gene is localized in an operon-like structure (Fig. 1), possible polar effects on neighboring genes were further excluded by complementation of the corA mutation with the intact corA gene cloned in the H. pylori plasmid pHel3 (Table 1). If introduced into the corA mutant of strain 1061, the resulting plasmid pHel3-CORA enabled the transformants to grow on Dent agar without Mg²⁺ supplementation.

FIG. 2.

Determination of the Mg²⁺ requirement of H. pylori corA mutants. (A) Mg²⁺ requirement caused by the corA mutation in H. pylori strain 1061. The growth of H. pylori strain 1061 (black bars) and of the corA mutant 1061-CORA1 (gray bars) in BBF was determined by measuring the OD₆₀₀. Bacteria were precultured to an OD₆₀₀ of 1.0 in BBF medium with 20 mM Mg²⁺ and then subcultured in BBF supplemented with Mg²⁺ at increasing concentrations, indicated on the x axis. (B) Mg²⁺ requirement caused by the corA mutation in H. pylori strain 26695. The growth of H. pylori strains 26695 (black bars), 26695-CORA1 (corA::cat; gray bars) and 26695-CORA2 (corA::Pcat; white bars) was determined by measuring the OD₆₀₀. Bacteria were precultured to an OD₆₀₀ of 1.0 in BBF medium with 20 mM Mg²⁺ and then subcultured in BBF supplemented with Mg²⁺ at increasing concentrations, indicated on the x axis. The data shown represent mean values from three independent determinations. Standard deviations are indicated.

Analysis of Mg²⁺ and metal transport functions of H. pylori CorA in E. coli.

The analysis of CorA transport functions in H. pylori was restricted by the drastic Mg²⁺ requirement of the corA mutants. Therefore, we studied functions of H. pylori CorA in Mg²⁺ and metal uptake in the corA mutant H5324 of the E. coli strain MC4100 (wild type [wt]) (Table 1). Strain H5324 displays classical corA phenotypes, namely, increased resistance to cobalt, nickel, and Mg²⁺, as determined in preliminary growth experiments (Fig. 3A). Cation concentrations toxic for the parental strain but not for the corA mutant were determined by growth inhibition experiments in LB medium supplemented with Mg²⁺, nickel (catalog no. N5756; Sigma), cobalt (catalog no. 2539; Merck), iron (catalog no. F2877; Sigma), and copper (catalog no. C6641; Sigma) at increasing concentrations. Under these growth conditions a restoration of E. coli H5324 metal sensitivity indicated CorA-mediated cation import (Fig. 3A). The subsequent analysis of corA phenotypes in strain H5324 transformants carrying intact versions of H. pylori corA on plasmids pCORA+ and pCORA− (Table 1) showed that plasmid pCORA+ restores Mg²⁺, nickel, and cobalt sensitivity of E. coli H5324 to wild-type (wt) levels (Fig. 3A). Similar to CorA of enterobacteria, H. pylori CorA mediates transport of all three elements. Copper resistance was not influenced by CorA. After 12 h of growth, the iron resistance of strain H5324(pCORA+) was reduced to 50% of that of strain H5324. This transient effect, which was not observed after 24 h of growth (Fig. 3A), suggests that CorA mediates low-affinity iron transport, as proposed earlier (27). The plasmid pCORA− did not suppress the corA-mediated phenotypes of E. coli H5324, providing evidence that the expression of the H. pylori corA gene in plasmid pCORA+ is driven by the lacZ promoter (Fig. 1). Determinations of the residual Mg²⁺ content of the media after bacterial growth revealed that H. pylori corA does completely restore the Mg²⁺ transport defect in E. coli H5324 (Fig. 3B). Furthermore, additive supplementation with Mg²⁺ in equimolar concentrations abolished the nickel and cobalt sensitivities of strain H5324(pCORA+), indicating that Mg²⁺ is the dominant substrate of H. pylori CorA (data not shown). Finally, nickel-hexaamine (Ni-HA), a potent inhibitor of S. enterica serovar Typhimurium CorA (17), also affected E. coli and H. pylori CorA transport functions, as additive Ni-HA supplementation of media abolished nickel sensitivity of the E. coli strains MC4100 and H5324(pCORA+) in a concentration-dependent manner (Fig. 3C).

FIG. 3.

Analysis of H. pylori corA transport functions in the E. coli corA mutant H5324. (A) Sensitivity to Mg²⁺ and to various metals mediated by H. pylori corA in E. coli. The growth of E. coli strains MC4100 (black bars), H5324 (gray bars), and H5324 with plasmid pCORA+ carrying H. pylori corA (white bars) in the presence of Mg²⁺, nickel (Ni), cobalt (Co), iron (Fe), and copper (Cu) was determined by measuring the OD₅₅₀. The Mg²⁺ and metal concentrations which were toxic for the parental strain but not for the corA mutant are indicated on the x axis. The data represent mean values from three independent determinations. Standard deviations are indicated. (B) Determination of Mg²⁺ uptake mediated by H. pylori CorA. The E. coli strains MC4100 (black bars), H5324 (gray bars), and H5324(pCORA+) carrying the H. pylori corA gene (white bars) grown for 24 h in LB medium were removed by centrifugation and the Mg²⁺ concentrations in the supernatants were determined. Uptake of Mg²⁺ was calculated from the Mg²⁺ concentration in the LB medium prior to inoculation. The data represent mean values from two independent determinations. Standard deviations are indicated. (C) The influence of Ni-HA onCorA-mediated nickel transport. The E. coli strains MC4100, H5324(pCORA+), and H5324 were grown in LB medium supplemented with 3.5 mM nickel in the absence (black bars) or presence (shaded bars) of 2 mM Ni-HA. Growth was monitored by measuring the OD₅₅₀. The growth of the E. coli corA mutant, which was not affected by nickel or by the hexaamine, is shown as a control. The data represent mean values from three independent determinations. Standard deviations are indicated.

Influence of CorA on cobalt and nickel metabolism in H. pylori.

To test if CorA-mediated nickel and cobalt uptake contributes to H. pylori metal metabolism, we determined the MICs of both metals. The results of growth inhibition experiments in broth revealed that the corA mutation had no effect on the nickel and cobalt MICs of 3 mM and 20 μM, respectively, in strain 26695, indicating that CorA is not involved in maintaining metal homeostasis. The possible effect of corA on nickel metabolism was further investigated by analysis of urease expression and activity. Immunoblot analysis with the UreA-specific monoclonal antibody HPM-5021-5 (IBT, Reutlingen, Germany) revealed that neither basal expression nor nickel induction of H. pylori urease were influenced by Mg²⁺ supplementation or by the corA mutation (results not shown). Similar results were obtained for urease activity, which was determined with the Berthelot reaction as described earlier (3, 4, 26). After growth in BBF medium supplemented with 20 mM Mg²⁺ and 10 μm NiCl₂ the H. pylori strains 26695 and 26695-CORA1 displayed nearly identical urease activities of 48.7 ± 2.7 and 49.3 ± 7.8 U in the wt strain and in the corA mutant, respectively, indicating that the corA mutation does not influence H. pylori nickel metabolism.

Conclusions.

The complete growth deficiency in media without Mg²⁺ supplementation and the drastic Mg²⁺ requirement in the range of 20 mM displayed by corA mutants demonstrates that H. pylori CorA is essential for Mg²⁺ acquisition required for survival in low-Mg²⁺ environments. These findings underline the role of H. pylori cation metabolism in maintaining metabolic functions and highlight for the first time a substantial importance of Mg²⁺ acquisition in gastric adaptation. A role of CorA-mediated Mg²⁺ uptake in H. pylori colonization and/or survival in the gastric mucosa is supported by the Mg²⁺ concentration in human gastric juice, which at 0.7 mM (12) is far below the values required for growth of corA mutants. Thus, it seems very unlikely that H. pylori corA mutants can persist in the gastric mucosa for extended time periods. The observation that neither Mg²⁺ supplementation nor the corA mutation influenced H. pylori cobalt and nickel resistance indicates that Mg²⁺ is the dominant CorA substrate. This suggests that H. pylori Mg²⁺ acquisition and metal ion uptake functions are strictly separated. This was further supported by the observations that additive Mg²⁺ supplementation of growth media completely inhibited nickel and cobalt sensitivity mediated by H. pylori CorA in E. coli, and that Mg²⁺ at concentrations as high as 20 mM did not influence either metal resistance or induction of urease by nickel (J. Pfeiffer and S. Bereswill, unpublished observations). The separation of metal and Mg²⁺ uptake functions prevents a competitive inhibition of nickel and cobalt uptake by Mg²⁺, as in most natural environments Mg²⁺ is about 10²- to 10³-fold more abundant than the trace metals. Finally, the lack of alternative Mg²⁺ uptake system homologs (23) in H. pylori (1, 24) and the absence of CorA homologs in eukaryotes (23) offer the possibility of using CorA as a target for the development of new H. pylori-specific drugs. The corA phenotypes observed suggest that the corA mutation could be used as a marker for genetic manipulation and, similar to essential genes in S. enterica serovar Typhimurium (28), they might facilitate the construction of H. pylori security strains for the production of live vaccines.