Identification of the Key Protein for Zinc Uptake in Hemophilus influenzae *

Very little is known about specific mechanisms for zinc accumulation and transport in bacteria. In this study a putative adhesin B in Hemophilus influenzae, the product of gene HI0119, has been identified as a periplasmic zinc-binding protein (PZP1). A pzp1-deficient mutant has been constructed which is defective for growth under aerobic conditions and grows poorly under anaerobic conditions. The growth defect is specifically rescued by supplementing the growth medium with high concentrations of zinc. Subcellular fractionation was used to localize PZP1 to the periplasmic region in a nontypeable H. influenzae strain and in a transfected recombinant Escherichia coli strain (TApzp1). Recombinant PZP1, purified from a periplasmic extract of E. coli strain TApzp1, contained ∼two zinc atoms/protein molecule as determined by neutron activation analysis and atomic absorption spectroscopy. The zinc atoms could be removed by incubation with EDTA, and, by further addition of zinc, a total of five zinc atoms/PZP1 could be bound. Direct binding of 65Zn to the recombinant protein by Western blot was demonstrated. Taken together, these results provide direct evidence that PZP1 plays a key role in zinc uptake by H. influenzae.

Zinc is essential for all organisms because it plays a critical role in the catalytic activity and/or structural stability of many proteins. More than 300 zinc-dependent enzymes have been identified (1). Several important motifs commonly found in transcriptional regulatory proteins are stabilized by zinc, including the zinc finger, zinc cluster, RING finger, and LIM domain (2). Despite this importance, very little is known about the mechanisms and regulation of zinc transport in bacteria (3). The few studies reported suggest that bacteria appear to possess a specific energy-dependent zinc transport system (4 -7). However, studies concerned with the intracellular accumulation of the metal have been confounded by both the nonspecific binding of zinc to the bacterial surface and the rapid exchange of cellular zinc with zinc in the medium (7)(8)(9). For unknown reasons, the zinc requirement for bacteria seems to be much lower than that for fungi or other eukaryotic cells (8,9). The exceedingly small requirements for zinc have frustrated studies in this field. So far, a specific zinc transport mechanism has not been properly demonstrated, and thus there has been no means to study the regulation of the transport of this metal in prokaryotes (3).
Hemophilus influenzae is a commensal of the human upper respiratory tract and can cause both localized and invasive infections in humans (10,11). Recently, we reported 1 the partial characterization of HI0119, identified from the H. influenzae genomic sequence (13) as a putative adhesin B because of its homology with the adhesin fimA of Streptococcus parasanguis. However, this 37-kDa protein is distinct from fimA because of a central histidine-rich domain, potent celite binding ability, 1 and a COOH-terminal disulfide-bonded domain. Expression of HI0119 is highly conserved in all H. influenzae clinical strains tested. 1 In this study, we demonstrate that this putative adhesin B is in fact, a periplasmic zinc-binding protein (PZP1) which plays a key role in the zinc uptake of H. influenzae. This is the first description of a potential prokaryotic zinc-specific transport protein.

MATERIALS AND METHODS
Clinical strains of H. influenzae were kindly provided by the Department of Microbiology, Hospital for Sick Children. Pfu and Taq polymerase were purchased from Stratagene and Pharmacia Biotech Inc., respectively. Restriction enzymes and buffers used for DNA manipulation were purchased from Pharmacia. DNA and protein standards were purchased from Life Technologies, Inc. and Bio-Rad, respectively. TA cloning kit, pTrc plasmids, and Escherichia coli strains Top10 and INV␣ were purchased from Invitrogen. SpectroPor dialysis tubing (MWCO 12,000 -14,000 kDa) was from Fisher. The tubing was washed extensively with double distilled H 2 O before use. Goat anti-rabbit horseradish peroxidase conjugate was from Bio-Rad. 65 ZnCl 2 was purchased from Amersham. Restriction analysis and plasmid constructions were performed using standard techniques as outlined by Sambrook et al. (14). Amino acid analysis was performed by the Biotechnology Service Center at the University of Toronto.

Bacterial Cultures
Nontypeable H. influenzae strain NTHI6564 was obtained immediately following isolation and stocks stored at Ϫ70°C in glycerol citrate. The strain was plated from frozen stocks onto chocolate agar plates and grown overnight at 37°C under aerobic conditions or anaerobic conditions using a BBL GasPak Plus generator with catalyst (Baxter Healthcare Corporation, Medford, MA). Each strain was subcultured twice before assay. E. coli was grown in Luria-Bertani or BHI broth or agar, aerobically at 37°C and supplemented with 100 g/ml ampicillin where appropriate.

Subcellular Fractions of H. influenzae
Triton X-114 Extraction of Integral Membrane Proteins-Integral membrane proteins of NTHI6564 were extracted with Triton X-114 essentially as described by Swancutt et al. (15). Briefly, bacterial cultures were grown to an A 600 nm of 0.5, harvested by centrifugation at 4°C, and washed using 1 volume of 200 mM Tris-HCl (pH 8.0). The pellets were suspended in 20 mM Tris-HCl (pH 8.0), 10 mM EDTA, 2% Triton X-114. After incubation for 4 h at 4°C, cellular debris was removed by centrifugation. The supernatants were warmed to 37°C to allow phase separation. After centrifugation at 14,000 ϫ g for 10 min, the aqueous phase was separated from the detergent phase. The detergent phases were washed three times using 20 mM Tris-HCl (pH 8.0), 10 mM EDTA.
Osmotic Shock Extraction of Periplasmic Proteins-Proteins in the periplasmic space were obtained by the osmotic shock procedure described by Ames (16). Briefly, bacterial cultures were grown at 37°C overnight and harvested by centrifugation. The bacterial pellet was gently resuspended in 30 mM Tris-HCl (pH 8.0), 20% sucrose. EDTA was added to a final concentration of 1 mM, and the bacteria were incubated at room temperature for 5-10 min with shaking. The cells were recovered by centrifugation and then resuspended in ice-cold 5 mM MgSO 4 with shaking for 10 min at 4°C. The supernatant, containing periplasmic proteins, was collected as osmotic shock fluid.
Expression and Purification of PZP1 in E. coli PCR 2 amplification of the pzp1 gene and its upstream 590-base pair region was performed as described previously. The primers used in amplification were PZU500, 5Ј-GACTACGTCATTGATGC-3Ј and HIMA2, 5Ј-GAATTCTTATTTAGCTAAACATTCCATGTAGC-3Ј. The 1.6-kb amplified product was ligated into a TA cloning vector, pCRII . The resulting plasmid was designated as pTApzp. The orientation of the insert was determined by restriction enzyme digestion; two clones, containing either orientation, were checked to assess whether the native pzp1 promoter has function in E. coli.
The E. coli transformant containing the plasmid pTApzp was grown overnight in 1 liter of BHI medium with ampicillin. The cells were harvested by centrifugation, and the periplasmic proteins were isolated as described above. The concentrated periplasmic extract was dialyzed overnight against 25 mM imidazole-HCl buffer (pH 7.4) and was applied to a column of Polybuffer exchanger 94 (Pharmacia) equilibrated with the same buffer. Elution was carried out with a degassed solution of Polybuffer 74 (Pharmacia) diluted 1:8 with distilled water and adjusted to pH 4.0 with HCl. Fractions were monitored for absorbance at 280 nm, and for pH, and the PZP1-positive fractions (identified by Western blotting) were pooled and lyophilized. To remove ampholytes, the lyophilized material was dissolved in 1 ml of water and applied to a G-75 fine column (1.5 ϫ 80 cm) equilibrated with 50 mM Tris-HCl (pH 7.2). The column was run at 0.5 ml/min, and fractions were monitored at 254 nm.
A standardized stock solution of PZP1 of known concentration was prepared as follows. Purified protein was dialyzed extensively against water, and the protein concentration was determined by amino acid analysis in triplicate. The absorbance at 280 nm was measured for dilutions of stock solution prepared in 20 mM Tris-HCl (pH 7.1), 20 mM NaCl, and from these values the extinction coefficient of PZP1 was calculated to be 36,800 liters mol Ϫ1 cm Ϫ1 .

Metal Composition of Purified PZP1
Neutron Activation Analysis-Purified PZP1 with a concentration of 0.56 mg/ml was heat sealed in a clean polyethylene vial. This was irradiated for 5 min at a neutron flux of 1.0 ϫ 10 12 n⅐cm Ϫ2 s Ϫ1 in the SLOWPOKE reactor of the University of Toronto. The irradiated solution was transferred to a clean vial, and after a delay time of 2.5 h to allow Cl 38 to decay sufficiently, its radioactivity was assayed for 15 min with a hyperpure germanium detector-based gamma-ray spectrometer. The manganese content was established by comparison with standards. To determine the chromium, nickel, scandium, iron, zinc, and cobalt contents of the sample, it was irradiated for 16 h at a neutron flux of 2.5 ϫ 10 11 n⅐cm Ϫ2 s Ϫ1 . After a delay time of 6 days to allow 24 Na to decay, the radioactivity in the sample was counted for 24 h and the elemental concentration determined.
Atomic Absorption Analysis-Aliquots of purified PZP1 were incubated in the presence of 1 mM Zn(NO 3 ) 2 or 5 mM EDTA at 4°C overnight. The samples were then dialyzed extensively against 20 mM Tris-HCl (pH 7.1), 20 mM NaCl at 4°C for 4 days. Zinc content was determined by atomic absorption spectroscopy using a Varian SpectrAA 10 spectrometer.
Zinc Blotting-65 Zn blotting was performed according to the method of Schiff et al. (17).

Circular Dichroism Measurements
CD spectra were recorded between 190 and 250 nm on a Jasco 720A spectropolarimeter using a 1-mm quartz cell at 25°C. The concentration of PZP1 was 0.56 mg/ml in 20 mM Tris-HCl (pH 7.1), 20 mM NaCl.

Electrophoresis and Western Blotting
SDS-PAGE and Western blotting were carried out by the methods of Laemmli (18) and Towbin et al. (19), respectively. For Western blots, a rabbit polyclonal antiserum (1/1,000 dilution) raised against a 6 ϫ histidine-tagged PZP1 fusion protein (His-PZP1) 1 was used.

Construction of the pzp1 Isogenic Mutant
A 3.5-kb DNA fragment from the NTHI6564 chromosome, including the pzp1 gene and its 1.2-kb upstream and 1.3-kb downstream region, was amplified by PCR as described previously. 1 The primers used for PCR were HZU5, 5Ј-CGGATCCCTCTTGTAGCAATGGCTTCAGTG-3Ј, and HZD3, 5Ј-GAATTCCATTGGGATGTTGGTCTCAACAG-3Ј. The amplified product was cloned into vector pCRII, generating plasmid pTA3.5. The 2.5-kb BamHI-EcoRI fragment from pTA3.5 was subcloned into vector pTrcA. The resulting plasmid, pTrc2.5, was digested with PstI to remove an 822-base pair internal region of the pzp1 gene and ligated to a 1.2-kb kanamycin-resistant cassette from pUC4k. The resulting plasmid was designated as Ypzp::kan.
The plasmid Ypzp::kan was digested to completion with BamHI and EcoRI, and this digestion mixture was used to transform NTHI6564 cells made competent for transformation with M-IV medium as described previously (20). The transformants were selected on chocolate agar plates containing 20 g of kanamycin/ml under both aerobic and anaerobic conditions. Transformant colonies were screened by direct amplification of chromosomal DNA from single colonies by PCR. Oligonucleotide primers used in the screening PCR were primers P1, 5Ј-GTATAGCATCAGTA-AAACC-3Ј, and P2, 5Ј-TTATTTAGCTAAACATTCCATGTAGC-3Ј.

Growth Curves
Single colonies from NTHI6564 wild-type strain and the pzp1 mutant were grown on chocolate agar plates under anaerobic conditions overnight. The bacteria from chocolate agar plates were suspended in 1 ml 20% Levinthal broth at an A 600 nm of 0.7. 200 l of this bacteria suspension was added into 50 ml of 20% Levinthal broth with or without zinc supplement and aerobically grown at 37°C with shaking at 180 rpm. The A 600 nm was determined at various points along the growth curve.

Localization of PZP1 in H. influenzae-
We have shown previously that the NH 2 -terminal 24-amino acid sequence of PZP1 forms a signal peptide, 1 which suggests that this protein is extracytoplasmic. Because PZP1 was originally purified from a surface protein (water) extract from an NTHI strain, immunofluorescence was used to determine the possible surface localization of this protein. This study revealed that this protein is not on the H. influenzae cell surface (data not shown). Furthermore, Western blotting showed that after overnight growth of H. influenzae in liquid culture, PZP1 is cell-associated and is not secreted into the culture medium (data not shown). Other members of the fimA family possess an NH 2 -terminal lipid linkage consensus sequence LXXC (21). Although PZP1 contains no such sequence and does not contain obvious membrane-spanning hydrophobic regions, the possibility of membrane anchorage was investigated.
A characteristic of integral membrane proteins, including outer membrane proteins from Gram-negative bacteria, is their selective partitioning into the detergent phase after Triton X-114 extraction (22). NTHI6564 was extracted with Triton X-114; most proteins were insoluble. Triton X-114-soluble proteins were partitioned into a detergent phase and an aqueous phase. PZP1 was found exclusively in the aqueous phase ( Fig. 1, lanes 1-3).
The periplasmic proteins of Gram-negative bacteria can be released selectively by osmotic shock (16). PZP1 was found in the periplasmic extract of NTHI6564 (Fig. 1, lane 4), indicating that this protein was localized in the periplasmic space in H. influenzae.
Expression and Purification of PZP1 in E. coli-We amplified the pzp1 gene and its upstream 590-base pair region from NTHI6564 chromosomal DNA and cloned this 1.6-kb fragment into the TA cloning vector pCRII to generate plasmids pTApzp1 and pTApzp2. In pTApzp1, the pzp1 gene is in the same orientation as the lac promoter but is inverted with respect to this promoter in plasmid pTApzp2. Recombinant E. coli strains harboring either plasmid were found to overexpress this 37-kDa protein at a similar level, suggesting that transcription of this protein is initiated at a H. influenzae promoter that is well recognized by E. coli RNA polymerase. Recombinant PZP1 was found to localize in the periplasmic space of E. coli as it did in H. influenzae and constitutes about 35% of the total protein in the periplasmic shock fluid. Recombinant protein was purified from periplasmic extracts by chromatofocusing ( Fig. 2A) and gel filtration chromatography. PZP1 thus obtained has been purified to Ͼ98% homogeneity as determined by SDS-PAGE with a yield of approximately 7 mg of pure protein from 1 liter of bacterial culture. Recombinant PZP1 was recognized by antisera against fusion protein His-PZP1 on Western blots (Fig.  2B). The observed isoelectric point of this protein, as estimated by chromatofocusing, was 6.4, which is close to the theoretical isoelectric point (pH 6.7) predicted from the amino acid sequence.
Like the native protein, recombinant PZP1 contains a disulfide bond as determined by differential migration by SDS-PAGE under reducing and nonreducing conditions and has potent celite binding capability, which is similar to the native PZP1 (data not shown). Amino acid analysis of purified recombinant PZP1 gave the expected amino acid composition for the mature, processed form of the protein (results not shown).
To obtain initial experimental information about the secondary structure of the purified protein, a CD spectrum was recorded. The strong molar ellipticity at 222 nm suggests that the major secondary structure of PZP1 is ␣-helix (25%) (results not shown).
PZP1 Is a Zinc-binding Protein-Our previous data have suggested that PZP1 may be capable of binding metal(s). To In 65 Zn blotting, the filter was equilibrated for 2 h in metal binding buffer (100 mM Tris-HCl (pH 8.0), 50 mM NaCl) and probed with 65 ZnCl 2 (20 Ci/10 ml) for 1 h. After washing, the filter was exposed at Ϫ80°C to Kodak X-Omat film.
determine the nature and amount of metals bound to PZP1, the purified protein was analyzed by neutron activation. Neutron activation analysis revealed that PZP1 contained about two zinc atoms/protein molecule but did not contain measurable levels of chromium, nickel, scandium, iron, cobalt, or manganese. The zinc content of purified PZP1 was also confirmed by atomic absorption spectroscopy (Table I).
To assess the zinc binding capacity of PZP1, the purified protein was incubated in the presence of 5 mM EDTA or 1 mM Zn(NO 3 ) 2 at 4°C overnight. After exhaustive dialysis, zinc content was determined by atomic absorption spectroscopy. Untreated PZP1 contained 1.58 zinc atoms/protein molecule, whereas EDTA treatment reduced bound zinc to an undetectable level. On incubation with zinc, a total of five zinc atoms/ protein molecule could be bound (Table I). In addition, 65 Zn binding to PZP1 was demonstrated using a 65 Zn blotting technique (Fig. 2D).
Construction of the pzp1 Negative Mutant-To construct a mutant deficient in the expression of PZP1, we amplified a 3.5-kb DNA fragment from NTHI6564 chromosomal DNA by PCR, including a 1.2-kb upstream and 1.3-kb downstream region of the pzp1 gene and cloned this DNA fragment into vector pCRII. The 2.5-kb BamHI-EcoRI fragment was subcloned into vector pTrcA. The internal 822-base pair PstI fragment of the pzp1 gene was replaced with a kanamycin resistance cassette (kan r ) from pUC4K (Fig. 3A). The resulting plasmid was designated as Ypzp1::kan. Restriction endonuclease analysis of Ypzp1::kan confirmed that the kanamycin cassette had been inserted into the expected site of the pzp1 gene. The mutant plasmid was introduced into the NTHI6564 wild type genome by transformation and homologous recombination with M-IV medium as described previously (20). Transformants were selected on chocolate agar plates containing 20 g of kanamycin/ ml, at 37°C, under aerobic or anaerobic conditions. Transformant colonies were screened by direct amplification of chromosomal DNA from single colonies by PCR. Oligonucleotide primers used in the screening PCR were primer P1 and P2 to allow amplification of the coding region of the pzp1 gene (Fig.  3A). A comparison of PCR in Fig. 3B illustrates that an insertion of the 1.2-kb kanamycin cassette into the pzp1 gene has occurred in transformants of Ypzp1::kan, as expected. Western blotting with antisera against the fusion protein His-PZP1 demonstrated that PZP1 was absent in the whole cell extract of the pzp1 mutant (Fig. 4).
Initial attempts to select transformants under aerobic conditions were unsuccessful; however, under anaerobic conditions, small colonies were observed after overnight incubation. Further study demonstrated that the pzp1 mutant does not grow on chocolate agar plates or in 20% Levinthal broth under aerobic conditions but does grow poorly under anaerobic conditions (Fig. 5). This result indicates that the pzp1 gene is essential for growth under aerobic conditions and also is im-portant for anaerobic growth of H. influenzae.
Growth Defect of the pzp1 Mutant Can Be Specifically Rescued by Zinc-Considering the periplasmic location and potent zinc binding capability of PZP1, it seemed reasonable that the growth defect was caused by zinc deficiency. Accordingly, chocolate agar plates were supplemented with ZnCl 2 at 100 M, and the pzp1 mutant was grown on the plates with zinc under aerobic and anaerobic conditions. The growth defect of the pzp1  mutant under both aerobic and anaerobic conditions was rescued by the addition of zinc (Fig. 5), suggesting that the pzp1 mutant is defective in zinc uptake. To assess the substrate specificity of suppression, 100 M CaCl 2 , MgCl 2 , CuCl 3 , NiCl 2 , Cd(NO 3 ) 2 , MnCl 2 , FeCl 3 , ZnCl 2 , or Zn(NO 3 ) 2 was added into 20% Levinthal broth. Only ZnCl 2 and Zn(NO 3 ) 2 rescued the growth defect of the pzp1 mutant under aerobic conditions (Fig.  6A), indicating that suppression of the growth defect is zinc-specific.
The growth characteristics of the pzp1 mutant at different zinc concentrations under aerobic conditions were compared with that of the wild type parent strain (Fig. 6B). In the presence of 5 M ZnCl 2 , the growth defect of the mutant cannot be rescued. At 100 M ZnCl 2 , the growth rate of the mutant was similar to that of the wild type strain, although the cell density at stationary phase was lower than that of the wild type strain (Fig. 6B). DISCUSSION Despite the rapidly increasing knowledge of zinc function at molecular and cellular levels, a zinc-specific transporter system in bacteria has not been properly demonstrated (3). There are several reasons for this discrepancy. First, extremely low zinc concentrations (0.5-1 M) are required for optimal bacterial growth. Second, elimination of zinc from medium using solvents or alumina has generally been unsuccessful (8). Finally, the high electrostatic affinity of zinc for anionic sites on the microbial surface is not readily distinguishable from zinc transport (8,9). In this report, we have identified the pzp1 (HI0119) gene product as a crucial protein for zinc uptake in H. influenzae. Although PZP1 was originally purified from a surface (water) extract 1 and inferred to be an adhesin, our results indicate that PZP1 in H. influenzae is a periplasmic protein and thus likely does not directly contribute to adhesion of H. influenzae.
The pzp1 isogenic mutant described in this report provides an opportunity to study zinc transport in bacteria using biochemical and genetic methods. Furthermore, purified, functional PZP1 has been readily isolated in high yield. Further structural studies on this protein will give insights as to the precise role PZP1 plays in zinc processing in H. influenzae.
The ATP-binding protein cassette system is involved in the transport of a diverse array of macromolecules across the cytoplasmic membranes of bacteria and eukaryotes (23). This system consists of three basic parts: one or two ATPases, one or two integral membrane proteins, and one substrate-specific binding protein. In Gram-negative bacteria the binding protein is soluble and periplasmic, but in Gram-positive bacteria the binding protein is lipid linked to the cytoplasmic membrane. Usually these three components are encoded together in one operon in bacteria (24,25). PZP1 of H. influenzae, a periplasmic zinc-binding protein, is 23.7% identical and 47.8% similar to fimA of S. parasanguis. 1 fimA of S. parasanguis is a lipoprotein that is involved in adherence of these bacteria to the salivary pellicle of dental surfaces (26 -28). DNA sequence data showed that the S. parasanguis fimA locus encodes an ATP-binding membrane transport system (29). Interestingly, a fimA isogenic mutant did not display an obvious growth defect in vitro but was found to be less virulent in animal models (28). Our previous studies have suggested that there is functional heterogeneity between PZP1 of H. influenzae and fimA of S. parasanguis, since PZP1 has a central histidine-rich domain and a COOH-terminal disulfide-bonded domain which are absent in the fimA protein of S. parasanguis. 1 We speculate that fimA of S. parasanguis may be involved in an ATP transport system similar to that proposed for PZP1 but have different substrate binding specificity. Furthermore, a recent BLAST search showed that PZP1 has 49.2% identity and 59.4% similarity to an unidentified protein (YebL) in E. coli, a 31.1-kDa protein in the msbB-ruvB intergenic region precursor. Compared with the sequence of PZP1, YebL in E. coli seems to have a similar domain structure including a central potential metal binding domain of 21 amino acid residues and two conserved cysteines in the COOH-terminal region which may form a disulfide bond. The YebL locus also appears to be organized in an operon. YebL of E. coli may thus have a function similar to that of PZP1 of H. influenzae, which is involved in the transport of zinc.
In general, there is little sequence conservation between the binding proteins for different substrates among ATP-binding protein cassette transport systems (23). However, the pairs of periplasmic binding proteins that interact with a common membrane receptor have extensive homology (23). In the H. influenzae genome, another putative adhesin B (HI0362) has 21% identity to PZP1, and the gene HI0362 locus seems to have a genetic organization similar to that of typical ATP-binding protein cassette systems (13). Even though the pzp1 locus does not appear to be part of an operon as is usually found for comparable transporters, it is possible that PZP1 and the product of the HI0362 gene may interact with the same core transmembrane complex.
Our results demonstrate that the pzp1 mutant cannot grow under aerobic conditions and grows poorly under anaerobic conditions. Only zinc can suppress the growth defects of the pzp1 mutant. This suggests that the metabolic process under aerobic conditions may be more dependent on zinc than that under anaerobic conditions in H. influenzae. Alternatively, there may be an additional, lower affinity zinc transport system operating when H. influenzae grows under anaerobic conditions. Western blotting has shown that the expression level of PZP1 was not found to be decreased during anaerobic growth (results not shown), supporting the former explanation.
PZP1 of H. influenzae contains an unusual histidine-rich domain of about 47 amino acids. 1 This domain is extremely rich in potentially metal-binding amino acids, including 23 histidines, 10 aspartic acids, and 6 glutamic acids. We expect that the zinc binding sites may be located at this domain. Our results showed that purified PZP1 contained an average of 1.6 -1.9 zinc atoms/protein molecule, although this protein has potential to bind up to five zinc atoms. This finding is consistent with a zinc accumulation and transport role, since it would be expected that the PZP1 population would contain molecules at various stages of substrate delivery to the membrane bound component of the transporter.
Recent studies showed that zinc uptake in yeast Saccharomyces cerevisiae is transporter-mediated by at least two systems, one with high affinity and second with lower affinity. The transporters that are responsible for both uptake systems have been identified because of their significant similarity to IRT1, an Fe(II) transporter gene from the plant Arabidopsis thaliana (30). The zrt1 gene encodes the transporter protein of the high affinity system (31), whereas the zrt2 gene encodes the transporter of the low affinity system (12). Based on protein sequences, ZRT1 and ZRT2 were predicted to be integral membrane proteins containing eight potential transmembrane domains. However, the zrt1/zrt2 double mutant is viable, indicating the existence of additional zinc uptake pathways (12).
In this study, we demonstrate that PZP1 is a highly soluble periplasmic zinc-binding protein. Our results suggest that unlike in yeast, there may be only one zinc uptake system in H. influenzae.