An ABC Transporter and a TonB Ortholog Contribute to Helicobacter mustelae Nickel and Cobalt Acquisition

H. mustelae strain NCTC 12198 (ATCC 43772) and its isogenic mutants (see Tables S1 and S2 in the supplemental material) were routinely grown at 37°C in a microaerobic atmosphere of 5% O₂, 7.5% CO₂, 7.5% H₂, and 80% N₂ on Dent agar plates consisting of BBL Columbia agar (Becton Dickinson) supplemented with 7% saponin-lysed defibrinated horse blood (BioTrading), 0.004% triphenyltetrazolium chloride (Sigma), and Dent selective supplement (Oxoid). Broth cultures of H. mustelae for urease assays, Northern hybridization, and metal content determinations were grown in brucella broth (Becton Dickinson) supplemented with 3% newborn calf serum (Biotrading) and Dent supplement. Iron restriction of brucella broth was achieved by the addition of deferoxamine B (Desferal; Sigma) to a final concentration of 20 μM, whereas iron-replete medium was obtained by supplementing deferoxamine B-treated medium with FeCl₃ to a final concentration of 100 μM. The effect of metals on H. mustelae growth was assessed by using CoCl₂ and NiCl₂ supplementation of defined Ham's F-12 tissue culture medium (Kaighn's modification; Invitrogen) with 0.2% β-cyclodextrin (Fluka) and Dent supplement. Broth cultures were shaken at 70 rpm and incubated at 37°C for a maximum of 24 h. Escherichia coli strains DH5α and M147 were cultured in Luria-Bertani medium (Biotrading) (27). When appropriate, growth medium was supplemented with chloramphenicol (20 μg ml⁻¹), kanamycin (20 μg ml⁻¹), or erythromycin (250 μg ml⁻¹) for the selection of E. coli transformants. H. mustelae mutants were selected by growth on medium containing chloramphenicol (10 μg ml⁻¹), kanamycin (10 μg ml⁻¹), or erythromycin (10 μg ml⁻¹). Concentrations of antibiotics given represent their final concentration in the growth medium.

The H. mustelae nikR and fur mutants were described previously (33, 34). For the insertional mutagenesis of the other genes studied here (see Tables S1 and S2 in the supplemental material), a ∼1-kb fragment of the gene of interest was PCR amplified from DNA from H. mustelae strain NCTC 12198 by using knockout (KO) primers (Table S2). The amplicons were cloned into the pGEM-T Easy vector (Promega), and the cloned sequences were interrupted by the insertion of an antibiotic resistance cassette obtained from plasmids pJMK30 (kanamycin), pAV35 (chloramphenicol), and pDH20 (erythromycin) (16, 40). The antibiotic resistance cassettes were inserted into unique restriction sites in the cloned fragments: BglII for tonB1, tonB2, nixA, and ceuE; BclI for frpB1; EcoRV for frpB2 and fecA; KpnI for cfrA; HindIII for feoB; and Eco47III for fecD. The constructs were then introduced into H. mustelae strain NCTC 12198 and selected mutant strains by natural transformation (8), and the correct replacement by homologous recombination of the gene was confirmed by PCR using primers (Table S2) located outside the recombination side (ORS).

Cells obtained from cultures grown overnight were centrifuged for 10 min at 4,000 × g and resuspended in phosphate-buffered saline to a final optical density at 600 nm (OD₆₀₀) of 10. Bacteria were lysed by sonication for 15 s with an MSE Soniprep 150 set at an amplitude of 6. Whole-cell proteins were separated on a 6% or 15% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) (27). The assessment of UreB and UreB2 urease subunit expression was based on immunoblots incubated with a 1:10,000 dilution of antibodies raised against H. felis urease (3, 25, 33).

Total RNA was isolated by using Trizol (Gibco) according to the manufacturer's instructions. RNA was separated on 2% formaldehyde-1.5% agarose gels in 20 mM sodium phosphate buffer (pH 7), transferred onto positively charged nylon membranes (Roche Diagnostics), and covalently bound to the membrane by UV cross-linking (33, 34). RNA was visualized by methylene blue staining, and RNA concentrations were normalized based on 16S and 23S rRNA band intensities. Probes were amplified with the primers listed in Table S2 in the supplemental material. The PCR fragments contained a T7 promoter sequence and were used for the production of an antisense RNA probe labeled with digoxigenin by transcription with T7 RNA polymerase (Roche Diagnostics). Northern hybridization and stringency washes were performed at 68°C, and bound probe was visualized with the digoxigenin (DIG) detection kit (Roche Diagnostics) and the chemiluminescent substrate CPD-Star (Amersham Pharmacia) as described previously (33, 34).

Medium was prepared by mixing 7 g brucella broth (Becton Dickinson) and 3.75 g Bacto agar (Becton Dickinson) in 250 ml water, followed by sterilization by autoclaving. The medium was cooled down to 56°C, and a solution containing 7.5 ml heat-inactivated newborn calf serum (Biotrading), 250 μl 20 mM deferoxamine B (Sigma), and 250 μl Dent selective supplement (Oxoid) was added, together with triphenyl tetrazolium chloride (Sigma) to a final concentration of 0.004%. The medium was distributed over 9-cm petri dishes, at 17.5 ml per dish. H. mustelae strain NCTC 12198 and isogenic mutants (see Tables S1 and S2 in the supplemental material) were grown for 24 h at 37°C on Dent agar, harvested by centrifugation, and diluted in phosphate-buffered saline (PBS) to an OD₆₀₀ of 0.2. This suspension was spread equally across the agar plate by using a cotton swab. Blanc paper discs (Becton Dickinson) were saturated with 30 μl of either 1% heme (Sigma) solution in 0.1 N NaOH or 0.5 mM FeCl₃. PBS was used as a negative control. The discs were placed onto the agar plate and incubated for 48 h at 37°C in a microaerobic atmosphere. The diameter of the zone of growth was measured, and each experiment was performed at least three times independently.

The enzymatic activity of the H. mustelae UreAB urease was determined by measuring ammonia production from the hydrolysis of urea by using the Berthelot reaction as described previously (33, 34). Cells were lysed by sonication and incubated for 30 min at 37°C in buffer consisting of 100 mM sodium phosphate (pH 7.5), 10 mM EDTA, and 50 mM urea, and the ammonia produced was measured after the addition of phenol nitroprusside and alkaline hypochlorite (Sigma Diagnostics) by measuring the OD₅₇₀. Protein concentrations were determined by the bicinchoninic acid method (Pierce) using bovine serum albumin as a standard. Urease enzyme activity was expressed as units representing mmol of urea hydrolyzed per min and is expressed as U mg⁻¹ of total protein. The average data from three independent experiments are shown. Only the activity of the H. mustelae UreAB urease was measured, since sonication inactivates the UreA2B2 urease of H. mustelae (33).

H. mustelae strains were grown as broth cultures for 24 h, and cells were harvested by centrifugation. The cell pellet was washed three times with 1 ml PBSE solution, consisting of PBS supplemented with EDTA to a final concentration of 1 mM. Cell density was assessed by measurement of the OD₆₀₀, and cells were subsequently lysed by using 500 μl HNO₃ (65% Suprapur; Merck) and 250 μl of a 30% H₂O₂ solution (Suprapur; Merck). Metal concentrations were determined by high-resolution induction-coupled plasma mass spectrometry (HR-ICPMS) (23, 36). Levels of iron, cobalt, and nickel in the wild-type strain were 4.31 × 10⁶ (±1.21 × 10⁶), 2.08 × 10³ (±3.19 × 10³), and 5.94 × 10⁶ (±0.98 × 10⁶) atoms/cell, respectively, and values for the respective mutants are expressed as a percentage of the levels found in the wild-type strain. The number of cells was calculated by using the following conversion formula: OD₆₀₀ = 1 equals 5 × 10⁸ cells/ml. Results shown are the averages of data from three independent experiments.

The HR-ICPMS equipment used for quantification was an Element XR (Thermo Scientific) equipped with PC³ (Elemental Scientific [ESI]) and nickel interface cones. The following operating conditions for HR-ICPMS were used: radio frequency (RF) power of 12.5 kW and gas flow rates of 16 liter min⁻¹, 0.7 liter min⁻¹ (auxiliary), and 1.1 liter min⁻¹ (nebulizer). PC³ consists of a Peltier cooled inlet system, which incorporates the ESI cyclonic spray chamber, the ESI PFA-ST nebulizer, and the ESI Fast introduction system. A 0.25-ml loop was used on the Fast introduction system, and the PFA-ST nebulizer was fed by an ESI syringe system at a constant flow rate of 112 μl/min. The analytical masses (m/z) used for the analysis were m/z 56 for Fe, m/z 59 for Co, and m/z 60 for Ni. The medium resolution (MR) (greater than 4,000 m/Δm) was used because these elements were heavily interfered with by polyatomic ion species produced by a combination of isotopes coming from plasma, matrix, and reagents. Samples were diluted 10 times before analysis, and only ultrapure acids and ultrahigh-quality (UHQ) water (18.2 MΩ) were used. The calibration range was adapted to the element concentrations in the unknown samples. In order to correct for eventual instrumental drifts and matrix effects, Ge 73 was used as an internal standard. In order to check the quality of the analysis, the certified reference samples SPS-SW1, SPS-SW2, and TMDA-52.3 were analyzed in between the H. mustelae samples and were within their certified values.

We first assessed whether the transcription of the proposed ferric citrate and enterochelin uptake genes fecD and ceuE is regulated by Fur and/or iron, as is commonly observed for iron acquisition systems (1). RNA was isolated from H. mustelae wild-type and fur mutant strains cultured under iron-restricted and iron-replete conditions. As in H. pylori (13, 15, 39), the transcription of the ceuE and fecD genes was not affected by iron or the absence of Fur (Fig. 1A), suggesting a role in the acquisition of metals other than iron. As a control we included the frpB2 gene, which displayed the characteristic iron-repressed transcription in the wild-type strain but displayed derepressed and iron-independent transcription in the fur mutant (Fig. 1A).

We also investigated whether the two H. mustelae TonB orthologs are involved in iron acquisition by growth promotion assays with heme and ferric chloride as the sole iron sources. Iron was chelated from serum-supplemented brucella medium by using deferoxamine B (39), resulting in the absence of growth without an external iron source, whereas the addition of ferric chloride restored the growth of wild-type H. mustelae (Fig. 1B). H. mustelae was capable of using heme as the sole iron source; however, an inhibition zone around the heme-supplemented disc indicated heme toxicity at higher concentrations, as reported recently (30). The inactivation of the tonB1 gene prevented the utilization of heme as the sole iron source (Fig. 1B) but did not affect growth using iron chloride. In contrast, the inactivation of tonB2 did not affect the utilization of heme as the single iron source (Fig. 1B). The absence of both TonB orthologs gave the same results as those with the tonB1 mutant for heme and the tonB2 mutant for iron chloride (Fig. 1B). The inactivation of any of the seven putative TonB-dependent outer membrane transporter genes (see Table S1 in the supplemental material) did not affect the ability to use heme or iron chloride as the sole iron source (data not shown). Overall, this suggests that the tonB2, fecD, and ceuE genes are not involved in iron acquisition but may function in the acquisition of other metals or compounds.

We subsequently investigated the possible role of the two tonB genes and the fecDE and ceuE genes in nickel acquisition, similar to what was shown previously for the H. pylori TonB2 system (28) and the H. mustelae nikH gene (34). Urease enzyme activity was initially used as a reporter system for the intracellular nickel concentration (34, 45). As a positive control we included the H. mustelae nikH gene, since the inactivation of nikH results in a ∼50% reduction of urease activity in unsupplemented brucella medium (34). The inactivation of tonB2, fecD, and ceuE resulted in a ∼50% reduction of urease activity compared to that of the wild-type strain, whereas the inactivation of tonB1 and feoB did not affect urease activity (Fig. 2A). Surprisingly, the inactivation of the nixA gene, encoding the H. mustelae ortholog of the H. pylori inner membrane transporter for nickel (19, 46), did not result in a reduction of urease activity, suggesting the presence of an alternative pathway for nickel transport across the inner membrane. The inactivation of the fecD gene in combination with nixA, nikH, or tonB2 or the inactivation of the ceuE gene in combination with nixA or nikH resulted in a virtual absence of urease activity (Fig. 2A). In contrast, the inactivation of either tonB2 or nikH in combination with a nixA mutation did not show a further reduction of urease activity. This finding suggests that NikH and NixA are involved in the same nickel acquisition pathway. Finally, the supplementation of brucella broth with nickel to a final concentration of 100 μM fully restored urease activity in the fecD mutant, similar to what was described previously for an H. mustelae nikH mutant (34), but only partially restored urease activity in the nixA fecD double mutant (see Fig. S1 in the supplemental material).

High-resolution induction-coupled mass spectrometry (HR-ICPMS) (23, 36) was used to determine the cellular levels of nickel and cobalt in the wild-type strain and eight of the H. mustelae mutants previously tested for urease activity (Fig. 2B). There was a very good match between cellular nickel content and urease activity, since reduced urease activity in the nikH, tonB2, fecD, ceuE, and fecDnixA mutants was reflected in similar decreases in nickel contents (Fig. 2B). Iron levels were not significantly different between the tested strains, ranging from 95% to 128% of the levels in the wild-type strain (see Fig. S2 in the supplemental material). We also used the regulation of the expression of the two H. mustelae urease systems to confirm the effect of mutations on nickel levels in the cell, since the UreAB urease is nickel induced and the UreA2B2 urease is nickel repressed, both via the NikR regulator (33, 34). When grown in nickel-supplemented Ham's F-12 tissue culture medium, only the UreB subunit was detected in the wild-type strain, consistent with an absence of UreA2B2 expression (Fig. 2C) (33). The inactivation of nixA did not alter this expression profile, whereas the inactivation of fecD in both the wild-type strain and the nixA mutant resulted in the expression of both UreB and UreB2 (Fig. 2C), consistent with nickel-limited intracellular conditions in these mutants, leading to the simultaneous expression of both the UreAB and UreA2B2 ureases.

To assess whether the putative nickel acquisition systems identified here are specific for nickel, we also determined cellular cobalt levels by HR-ICPMS. Relative cobalt levels matched the trend in cellular nickel levels, since the inactivation of fecD, ceuE, tonB2, and nikH resulted in a significant decrease in cobalt levels, whereas the inactivation of nixA did not affect cobalt levels (Fig. 2B). We also noticed that cobalt supplementation of Ham's F-12 medium resulted in growth inhibition, already at low concentrations (Fig. 3A), and hence also assessed the cobalt resistance of the nixA and fecD mutants and the nixA fecD double mutant. The inactivation of the fecD gene resulted in a significant increase in cobalt resistance, with growth unaffected at 10 μM cobalt, where the wild type and nixA mutant showed virtually no increase in the OD₆₀₀ compared to that of the inoculum at 10 μM cobalt (Fig. 3A). The overall growth rate of the nixA mutants was reduced, but the pattern of cobalt resistance was not altered by the nixA mutation compared to the wild-type strain and the fecD mutant background (Fig. 3A). In contrast, there was no difference between the mutants in nickel sensitivity, as a decrease in the final OD₆₀₀ was observed only at 1,000 μM nickel but not at 10 and 100 μM nickel (see Fig. S3 in the supplemental material).

Cobalt also affected the expression of the two H. mustelae ureases (Fig. 3B), since growth in Ham's F-12 medium supplemented with 1 μM cobalt resulted in the repression of UreB2 expression, similar to the effect of nickel supplementation (33). The inactivation of the fecD gene abolished the cobalt-dependent repression of UreB2 expression (Fig. 3B), similar to what was observed for an H. mustelae nikR mutant (data not shown), supporting a role for the FecD protein in cobalt acquisition.