Conserved Residues and Motifs in the NixA Protein ofHelicobacter pylori Are Critical for the High Affinity Transport of Nickel Ions*

NixA, the high affinity nickel transport protein of Helicobacter pylori, imports Ni2+ ions across the cytoplasmic membrane for insertion into the active site of the urease metalloenzyme, which is essential for colonization of the gastric mucosa. Twelve conserved aspartate (aspartates 47, 49, 55, 194, 231, and 234), glutamate (glutamates 106, 198, and 274), and histidine (histidines 44, 50, and 79) residues were identified by alignment of NixA with homologous transporters. Polymerase chain reaction-generated site-directed mutants of these residues were expressed in E. coli along with the H. pylori urease gene cluster. Mutations in residues within the predicted periplasmic domains of NixA maintained near wild type levels of Ni2+ uptake and urease activity, as did control mutations of conserved positively charged residues (lysines 140 and 268; arginines 162 and 167). Mutations in highly conserved motifs in predicted helices II and III of NixA abolished Ni2+ uptake and urease activity. Mutations in helices V and VI and the cytoplasmic domains decreased Ni2+transport rates by ≥90%. Reduction in rates of Ni2+transport correlated with reduction in urease activities (r = 0.77). Ni2+ transport was inhibited in the presence of Co2+, Cu2+, and Zn2+, indicating that these ions may also be bound or transported by NixA. We conclude that conserved Asp, Glu, and His residues in the transmembrane domains of NixA are critical for the transport of the divalent cations Ni2+, Co2+, Cu2+, and Zn2+ into the cytoplasm of H. pylori.

shown to be highly sensitive to low pH unless urea is present (11). The ammonia produced by ureolysis has been postulated to allow the acid-sensitive organism to survive and colonize the low pH environment of the gastric mucosa by neutralizing gastric acid. Ammonia produced by this reaction may also contribute directly to histological damage (12). An even greater role of urease in H. pylori colonization and pathogenesis has been implied by the observation that urease-negative mutants of H. pylori are unable to colonize animal models of infection in which the pH of the stomach is maintained at neutrality by the administration of a proton pump inhibitor (13,14).
The divalent nickel ion is a requisite cofactor in the active site of urease (15). Since human serum concentrations of Ni 2ϩ average only 2-11 nM (16), H. pylori must be capable of scavenging Ni 2ϩ from presumably similar concentrations present in the gastric mucosa. One mechanism by which this is accomplished is the NixA nickel transport protein (17). When a plasmid carrying nixA was cotransformed into Escherichia coli also expressing the H. pylori urease gene cluster, NixA proved to be a very high affinity Ni 2ϩ importer with a K T ϭ 11.3 nM, well suited to physiological conditions, and allowed synthesis of catalytically active urease independent of growth conditions (17).
NixA is a 331-amino acid, 37-kDa integral membrane protein consisting of eight predicted transmembrane domains. Three one-component Ni 2ϩ -importing homologs have been identified: the HoxN protein of Alcaligenes eutrophus (18), the HupN protein of Bradyrhizobium japonicum (19), and the UreH protein of the thermophilic Bacillus TB90 (20). NixA, HoxN, HupN, and UreH share the amino acid sequence motif HX 4 DH located in transmembrane helix II of NixA with the NikC component of the nonhomologous Nik ABC type Ni 2ϩ transporter of E. coli (18). Significantly, the amino acid sequence motif GX 5 GHSSVV, located in the sequence spanning the second periplasmic loop and helix III of NixA, is also conserved in HoxN, HupN, and UreH (18).
In this report, we present the results of specific site-directed mutagenesis of 12 potential metal-binding residues conserved among NixA and homologous transporters by direct measurement of Ni 2ϩ uptake and the synthesis of catalytically active urease.
Recombinant DNA Techniques-Recombinant DNA methods including restriction endonuclease digestion, ligation, and transformation were performed according to standard protocols (21,23). Plasmid DNA was purified by rapid alkaline lysis (24). Large scale preparations were isolated using Qiagen DNA purification columns according to the manufacturer's instructions.
Site-directed Mutagenesis-Site-directed mutations were PCR-generated by the overlap extension method of Ho (25), using the mutagenic and flanking primers listed in Table I. First round PCR products were agarose gel-purified to prevent the amplification of wild type nixA. PCR reactions were performed using cloned Pfu DNA polymerase (Stratagene). PCR-generated mutants were cloned into pBluescript II SKϩ as XbaI/HindIII fragments.
Nucleotide Sequencing-Plasmid DNA was sequenced by the dideoxy chain termination method (26). Reactions were run on an Applied Biosystems model 373A DNA sequencer.
Preparation of a Polyclonal Antiserum to NixA-A 149-base pair PCR-amplified fragment corresponding to gene sequences encoding NixA amino acids 133-180 was subcloned into pBluescript SKϩ (Stratagene), excised at PCR-generated NcoI and BglII sites, and ligated into the His 6 tail vector pQE60 (Qiagen). The resulting construct, pQNS99, was transformed into E. coli M15pRep4, and the NixA-His 6 polypeptide was overexpressed by induction of exponential (A 600 ϭ 0.5) phase Luria broth cultures (1 liter) containing 5 mM isopropyl-1-thio-␤-D-galactopyranoside for 4 h at 37°C. Bacteria were harvested by centrifugation (5000 ϫ g, 10 min, 4°C) and resuspended in 30 ml of buffer B (8 M urea, 0.1 M NaH 2 PO 4 , 0.01 M Tris-HCl, pH 8.0). Cells were ruptured in a French press at 20,000 p.s.i., and the resultant lysate was centrifuged (8000 ϫ g, 10 min, 4°C). The supernatant was loaded onto a 4-ml Ni 2ϩ -nitrolotriacetic acid column (Qiagen) pre-equilibrated with buffer B. The column was washed with 200 ml of buffer B and then washed with buffer C (buffer B, adjusted to pH 6.3) until the A 280 of the effluent was Ͻ0.01. The NixA-His 6 polypeptide was eluted with 10 ml of buffer E (buffer B, adjusted to pH 4.5) collected in 1-ml fractions, which were analyzed by SDS-polyacrylamide gel electrophoresis in a 20% polyacrylamide gel. NixA-His 6 -containing fractions were dialyzed overnight against 1 mM Tris-HCl, 0.5 mM EDTA, pH 8.0.
The purified polypeptide was trichloroacetic acid-precipitated, and 100 g was emulsified in Freund's complete adjuvant and used to immunize two New Zealand White rabbits by subcutaneous injection. Three 30-g boosters were administered subcutaneously in Freund's incomplete adjuvant at 3-week intervals. NixA-specific antibodies were affinity-purified by immobilizing the NixA-His 6 polypeptide on a Pierce Aminolink column according to the manufacturer's instructions. Serum (1 ml) was diluted 1:1 in 10 mM Tris-HCl, pH 7.5, and loaded onto the column followed by 200 l of 10 mM Tris-HCl, pH 7.5, and allowed to stand at room temperature for 1 h. The column was washed with 40 ml of 10 mM Tris-HCl, pH 7.5, and then with 40 ml of 0.5 M NaCl, 10 mM Tris-HCl, pH 7.5. Antibodies were eluted with 10 ml of 100 mM glycine, pH 2.5, into a tube containing 1 ml of 1 M Tris-HCl, pH 8.
Western Blotting-Membranes of E. coli SE5000 (pHP808) cotransformed with pUEF204, pBluescript, or site-directed mutants of nixA were isolated from 100-ml overnight cultures. Bacteria from Luria broth cultures (100 ml) of each strain were harvested by centrifugation (5000 ϫ g, 10 min, 4°C) and resuspended in 25 ml of phosphatebuffered saline containing 1 mM phenylmethylsulfonyl fluoride. Cells were ruptured in a French press at 20,000 p.s.i., and the lysate was centrifuged (8000 ϫ g, 10 min, 4°C). The cleared lysate was ultracentrifuged (170,000 ϫ g, 90 min, 4°C). The resultant membrane pellet was resuspended in and washed with phosphate-buffered saline and then ultracentrifuged and washed twice as before. Each sample (5 g of protein) was electrophoresed under denaturing conditions on a 12% SDS-polyacrylamide gel and transferred to Immobilon P as described by Ausubel et al. (23). Blots were probed with antibodies to NixA, washed, and reprobed (23) with a polyclonal goat anti-rabbit alkaline phosphatase conjugate (Sigma). Blots were developed in 5-bromo-4-chloro-3indolyl phosphate/nitro blue tetrazolium substrate solution (Sigma) according to the manufacturer's instructions.
Ni 2ϩ Transport-E. coli SE5000 (pHP808) was cotransformed individually with pUEF204, pBluescript, and each of the site-directed mutants. Overnight cultures in M9 minimal medium were used to inoculate fresh cultures, which were harvested by centrifugation (8000 ϫ g, 5 min, 4°C) upon reaching an A 600 of 0.4. Cell pellets were resuspended in and washed with transport buffer (50 mM Tris, 1 mM MgCl 2 , pH 7.5). Cells were pelleted and resuspended in the same buffer to an A 600 of 0.5, and 63 NiCl 2 (specific activity, 6.35 mCi ml Ϫ1 , 0.46 mg of nickel ml Ϫ1 ; Amersham Corp.) was added to a final concentration of 50 nM at 37°C. Radioactive 63 Ni 2ϩ incorporation was assayed by vacuum filtration as described previously (17). For 63 Ni 2ϩ transport inhibition experiments, washed cells were resuspended in transport buffer to a final A 600 of 1.0 and added to an equal volume of transport buffer containing 100 nM 63 NiCl 2 and from 0 to 500 M CaCl 2 , CuCl 2 , CoCl 2 , MgCl 2 , MnCl 2 , NiCl 2 , or ZnCl 2 .
Urease Assays-E. coli SE5000 (pHp808) cotransformed individually with pUEF204, pBluescript, or each of the site-directed mutants was grown overnight in 100 ml of Luria broth supplemented with 1 M NiCl 2 . Cells were harvested by centrifugation (8000 ϫ g, 5 min, 4°C), resuspended in 2 ml of ice-cold 10 mM NaH 2 PO 4 , pH 6.8, and lysed by passing through a French pressure cell at 20,000 p.s.i. Lysates were centrifuged (8000 ϫ g, 10 min, 4°C), and urea hydrolysis by supernatants was measured by the phenol red assay of Hamilton-Miller and Gargan (27) as previously calibrated for quantitation (28). Protein concentrations were determined by the bicinchoninic acid method using the Pierce BCA assay kit according to the manufacturer's instructions.

RESULTS
Conservation of Amino Acid Residues among NixA and Homologs-A word search of the Swissprot and PIR data bases with the NixA amino acid sequence revealed three significant homologs of NixA that are also single component Ni 2ϩ importers (17). The 38.9-kDa HoxN protein of the Gram-negative soil and aquatic bacterium A. eutrophus shares 40% amino acid sequence identity with NixA (18). HoxN is responsible for the high affinity (K T ϭ 20 nM) import of Ni 2ϩ for incorporation into at least three enzymes including a cytoplasmic NAD-reducing hydrogenase, an electron transport-coupled hydrogenase, and a urease (29). Topological mapping of this integral inner membrane protein suggests eight transmembrane domains, with both the amino and carboxyl termini located in the cytoplasm (29,30). The second homolog, HupN, is a 40-kDa integral membrane protein that is encoded as part of a three-gene operon involved in Ni 2ϩ incorporation into the H 2 -recycling hydrogenase of the N 2 -fixing Gram-negative soybean symbiont B. japonicum (19). The HupN protein shares 41% amino acid identity with NixA; charge distribution and amino acid sequence similarity suggest eight membrane-spanning domains as in HoxN (31). The third homolog, the 25-kDa UreH protein of the thermophilic Bacillus TB90 (20), shares only 13% amino acid identity with NixA but contains sequence signatures of other high affinity Ni 2ϩ transporters (18) and is the only known example of a bacterial Ni 2ϩ transporter encoded within a urease operon (17).
Alignment of the sequence of these four homologous transporters, Kyte and Doolittle hydropathy predictions, and the results of NixA-LacZ fusions, 2 suggest an eight-transmembrane domain topology for NixA. Furthermore, sequence alignment produced an overall consensus (at least three out of four residues identical) sequence identity of 30%. Among the 95 amino acids of the consensus sequence, we identified 12 conserved Asp, Glu, and His residues (Fig. 1), including residues in the HX 4 DH motif in helix II and the GX 5 GHSSVV motif in helix III (18), which we postulated might be involved in Ni 2ϩ transport. Additionally, 2 conserved Lys and 2 conserved Arg residues were identified to serve as controls in mutagenesis experiments.
Site-directed Mutagenesis-To evaluate the contribution of conserved residues in Ni 2ϩ uptake, single amino acid mutations of NixA were generated by PCR overlap extension (25) using pUEF204 as a template and the primers listed in Table I (32) and is abundant in the wild type protein. First round PCR products were agarose gel-purified to eliminate amplification of wild type sequences in the second round of PCR. Second round PCR products were subcloned into the EcoRV site of pBluescript SKϩ for sequencing and isolation of plasmids carrying the mutant constructs. The nixA gene carrying each mutation was excised at PCR-generated XbaI and HindIII sites and religated into pBluescript SKϩ at these sites to regenerate modified plasmid pUEF204 encoding each mutation.
Ni 2ϩ Transport-NixA site-directed mutations H44I, D49I, and H50I, which lie in the conserved HX 4 DH motif of helix II, and H79I, which lies in the conserved GX 5 GHSSVV motif, showed complete abolishment of 63 Ni 2ϩ uptake when co-expressed in E. coli with the urease operon encoded on plasmid pHP808 (Table II). Transport was also abolished by or was not significantly different from background levels (E. coli transformed with pBluescript) for mutations D47I, D55I, D194I, and D231I, all of which lie within or immediately adjacent to putative transmembrane domains. Mutations E106I (predicted to be located in the periplasm), K140I, R162I, R167I, and K286I (conserved positively charged residues mutated as controls) imported Ni 2ϩ at rates not significantly different from wild type NixA. Mutations E198I, D234I, and E274I, all located within or immediately adjacent to putative transmembrane domains, maintained consistently low (Ͻ10% of wild type) rates of transport that were not statistically different from vector controls.
Urease Activities Confirm Transport Results-To confirm the Ni 2ϩ transport results and investigate the mutations that demonstrated low but measurable transport activity, urease assays were conducted. Cultures were grown overnight in Luria broth containing 1 M NiCl 2 , as opposed to the transport conditions of 50 nM 63 NiCl 2 sampled over 2.5 min to allow constructs that might import Ni 2ϩ at low rates to accumulate sufficiently high ion concentrations to yield synthesis of statistically significant urease activity. Mutations that abolished measurable Ni 2ϩ uptake at 50 nM also lowered urease activity to Ͻ1% of wild type, with the exceptions of D55I and D231I, which retained 5 and 16% of wild type urease activity, respectively (Table II). Mutations that demonstrated wild type uptake yielded wild type or greater urease activities. Interestingly, mutations E198I, D234I, and E274I, which demonstrated low but consistent levels of transport, did have greater than 50% of wild type urease activity in the assay. Significantly, all mutations that retained any measurable transport, with the exception of E106I, which gave inconsistent results, correlated directly (r ϭ 0.77) with observed urease activities (Table II, Fig. 2).
Western Blots Confirm the Presence of Full-length NixA in the Membrane-To show that changes in Ni 2ϩ transport and urease activity of NixA containing site-directed mutations were not the result of synthesis of a truncated protein, failure of the protein to be inserted in the membrane, or improper insertion or misfolding of the protein that might render it more susceptible to proteolysis, membrane preparations of E. coli transformed with pBluescript, pUEF204, or each site-directed mutation were assayed by Western blot with affinity-purified polyclonal antiserum raised against an internal polypeptide of NixA (Fig. 3). All NixA site-directed mutants produced a band of the appropriate size, at least equal in intensity to pUEF204 encoding wild type NixA. While some constructs appear to have bands that might correspond to degradation products of unstable NixA mutations, these same lanes tend to have more intense signals corresponding to full-length protein, indicating that the samples may contain more total membrane protein and less remaining cytoplasmic protein (each lane contains 5 g of protein). It is, nevertheless, possible that some constructs that possessed low transport but measurable urease activity, such as E198I, are altered in conformation and that the charges they encode are involved in formation of the "pore" or structure of the permease and not directly in binding and translocation of Ni 2ϩ ions.
Co 2ϩ , Cu 2ϩ , and Zn 2ϩ Are Also Bound or Transported by NixA-To establish the ion specificity of NixA, transport of 63 Ni 2ϩ was investigated in the presence of various cations (Table III). Co 2ϩ , Cu 2ϩ , and Zn 2ϩ each inhibited Ni 2ϩ transport, indicating that they are bound and possibly imported by NixA. Co 2ϩ inhibited transport most efficiently, decreasing uptake by 50% when present between 10-and 100-fold excess. Cu 2ϩ and Zn 2ϩ produced 50% inhibition of Ni 2ϩ uptake when present in 100 -1000-fold excess. Ca 2ϩ , Mg 2ϩ , and Mn 2ϩ had no measurable effect on 63 Ni 2ϩ uptake over the range of concentrations tested. DISCUSSION H. pylori is a serious human pathogen whose ecological niche is highly restricted to the gastric mucosa of humans, and production of an abundant urease is required for colonization. To synthesize catalytically active urease, Ni 2ϩ must be present in the cytosol for incorporation into the nascent subunits as they are folded and assembled into the 550-kDa multimeric enzyme (10). As normal human serum concentrations of Ni 2ϩ range only from 2 to 11 nM, it is predictable that H. pylori must have systems capable of scavenging Ni 2ϩ from presumably similar concentrations in the gastric mucosa. One such system has been shown to be the NixA nickel transport protein, which has a K T of 11.3 nM, well suited to these low physiological Ni 2ϩ concentrations (17). Isogenic NixA-deficient H. pylori mutants possess only 30% of the nickel-importing capacity of the parent strains and demonstrated a 42% decrease in urease activity, even when grown on rich medium (Brucella agar supplemented with 10% sheep blood) (33). While this may imply that H. pylori possesses alternate mechanisms of Ni 2ϩ incorporation, the reduction in urease activity of NixA-deficient mutants and the similarity of the K T of NixA to the normal serum concentration clearly underscore the importance of NixA in the synthesis of catalytically active urease, particularly under physiological conditions.
To investigate the role of 12 conserved negatively charged or histidine residues in the transport of Ni 2ϩ across the periplasmic membrane, we constructed specific site-directed mutations of these residues and measured the effect of these single amino acid changes directly on Ni 2ϩ uptake and on the resultant urease activities of recombinant constructs. MgCl 2 (10 mM) was included in the transport buffer to inhibit the nonspecific binding and transport of Ni 2ϩ by the CorA, MgtA, and MgtB Mg 2ϩ transport systems. The E. coli NikABCD transporter has a significantly lower affinity for Ni 2ϩ and a lower rate of Ni 2ϩ transport (35) compared with NixA and would not be relevant at the low Ni 2ϩ concentrations used in this study. As expected, mutations H44I, D49I, and H50I, which comprise the HX 4 DH motif of helix II abolished Ni 2ϩ uptake and urease activity. Recent studies of the two corresponding His residues in HoxN (30) confirm the importance of these residues in Ni 2ϩ binding and transport. Mutation of Asp 47 , which is conserved in three of the four homologs and lies in the variable region of the helix II motif in NixA, also abolished Ni 2ϩ transport and urease activity.
Since the helix II motif is highly conserved among NixA and its homologs, as well as in the NikC protein of the E. coli Nik nickel-importing ABC transporter and the Ni 2ϩ binding site of human serum albumin (18), it is very tempting to assign the role of Ni 2ϩ specificity determinant to this motif. However, the recent discovery of a cobalt-importing transporter NhlF in Rhodococcus rhodochrous (36), which is similar in size and topology to NixA and also contains this motif, implies that the HX 4 DH motif is not the sole determinant of specificity. As further support for this concept, we found that the presence Co 2ϩ in 10 -100-fold excess produced 50% inhibition of 63 Ni 2ϩ transport by NixA. The presence of 500-fold excess Ni 2ϩ decreased 57 Co 2ϩ transport by approximately 90% in NhlF, although direct uptake of 63 Ni 2ϩ by NhlF could not be demonstrated (36).
NixA mutation H79I in the GX 5 GHSSVV motif also abolished Ni 2ϩ uptake and urease activity. While this motif is conserved perfectly among NixA and homologous transporters, it is present as a slight variant (GX 5 GHSTVV) in helix III of the the cobalt importer of R. rhodochrous (36). Despite this slight change in consensus, it also seems unlikely that this sequence alone is sufficient for ion specificity.
The fact that mutation D194I completely abolished Ni 2ϩ transport and urease activity, as did the mutations in the helix II and III motifs, implies that these residues are directly involved in binding and translocation of nickel ions. Notably, Western blotting confirmed that none of these mutations truncated the synthesis of the protein or impaired the insertion of the transporter into the membrane, and no significant increase in proteolytic degradation products were observed in relation to total amount of NixA.
Mutations of residues that are located in predicted periplasmic or cytoplasmic loops (E106I, K140I, R162I, R167I, and K286I) all exhibited Ni 2ϩ transport rates not significantly different than the wild type. Similarly, they also exhibited wild FIG. 2. Urease activities of NixA site-directed mutants correlate directly with Ni 2؉ transport. Maximum rates of Ni 2ϩ transport (at 50 nM) were measured as described in Table II for for each mutation, as well as for wild type and vector control constructs. Urease activities of whole cell lysates of each construct grown in Luria broth supplemented with 1 M NiCl 2 were measured as described in Table II. The percentage of wild type urease activity of each mutant correlates (r ϭ 0.77) with percentage of wild type Ni 2ϩ transport for each mutant. type or greater than wild type urease activity. Mutation of residues located in or immediately adjacent to predicted transmembrane domains (D55I, D231I, E198I, D234I, and E274I), but not in the helix II or III motifs, reduced 63 Ni 2ϩ transport to Ͻ10% of wild type. These constructs, however, showed significant urease activities due to the sensitivity of the urease assay conditions (1 M NiCl 2 ). It appears, therefore, that the mutated conserved residues outside of the helix II and III motifs are not absolutely essential for Ni 2ϩ transport.
Urease activities correlated directly with observed rates of transport, with the exception of mutation E106I, which gave inconsistent transport results, often exhibiting Ͼ150% of the wild type transport. It may be reasonable to assume that residues that maintained low transport rates and significant urease activities are involved in the formation of the "pore" or precise alignment of helices within the membrane, rather than direct binding and translocation of nickel ions. More conservative replacements, multiple concomitant replacements, or spectroscopic analysis would be required to confirm this hypothesis.
Alternatively, it may be the combination of one or more of these residues with residues that clearly abolish all transport and urease activity when mutated that defines the Ni 2ϩ specificity determinant of NixA and the other high affinity onecomponent transporters. The possibility of additional periplasmic binding proteins or highly ion specific surface proteins that interact with NixA and other transporters also cannot be ruled out (17).
The fact that the high affinity NixA permease may also transport lower levels of Co 2ϩ , Cu 2ϩ , and Zn 2ϩ is not altogether surprising. While each of these ions would be toxic in high concentration, specific exporters of these ions (37) that could serve as compensatory mechanisms have been described in the literature.  3. Detection of NixA mutants in the membrane by Western blotting. Membrane fractions were isolated from E. coli (pHp808) cotransformed with pBluescript, pUEF204 (encoding nixA), or each site-directed mutant. Samples (5 g) were separated by SDS-polyacrylamide gel electrophoresis, transferred to polyvinylidene difluoride, and probed with antibodies to NixA.