Transcriptional Regulation, Metal Binding Properties and Structure of Pden1597, an Unusual Zinc Transport Protein from Paracoccus denitrificans*

Background: Bacterial cluster 9 ATP binding cassette (ABC) transporters are responsible for import of Zn2+ and Mn2+. Results: A solute-binding protein of the cluster 9 family is shown to preferentially bind Zn2+ in an unusual coordination sphere. Conclusion: A subclass of Zn2+ transporter has been identified with distinct structural properties. Significance: Cluster 9 ABC transporters are important virulence factors. ATP-binding cassette (ABC) transporters of the cluster 9 family are ubiquitous among bacteria and essential for acquiring Zn2+ and Mn2+ from the environment or, in the case of pathogens, from the host. These rely on a substrate-binding protein (SBP) to coordinate the relevant metal with high affinity and specificity and subsequently release it to a membrane permease for translocation into the cytoplasm. Although a number of cluster 9 SBP structures have been determined, the structural attributes conferring Zn2+ or Mn2+ specificity remain ambiguous. Here we describe the gene expression profile, in vitro metal binding properties, and crystal structure of a new cluster 9 SBP from Paracoccus denitrificans we have called AztC. Although all of our results strongly indicate Zn2+ over Mn2+ specificity, the Zn2+ ion is coordinated by a conserved Asp residue only observed to date as a metal ligand in Mn2+-specific SBPs. The unusual sequence properties of this protein are shared among close homologues, including members from the human pathogens Klebsiella pneumonia and Enterobacter aerogenes, and would seem to suggest a subclass of Zn2+-specific transporters among the cluster 9 family. In any case, the unusual coordination environment of AztC expands the already considerable range of those available to Zn2+-specific SBPs and highlights the presence of a His-rich loop as the most reliable indicator of Zn2+ specificity.

His (20 -26), with the exception of TroA, which uses only 1 Asp and 3 His residues (27)(28)(29). Further complicating the situation is the observation that some of the SBPs are not crystallized with their cognate metal, or the identity of the physiologically relevant metal has not been determined by independent means.
Based on a simple BLAST search, the genome of Paracoccus denitrificans contains 3 ABC transporters of the cluster 9 family (Fig. 1) identified by SBP genes pden1259, pden1597, and pden4140. Pden1259 is homologous to the Mn 2ϩ SBP MtsA from S. pneumoniae (31.6% sequence identity) including the absolute conservation of metal ligands and the absence of a His-rich sequence. Similarly, pden4140 is homologous to the Zn 2ϩ SBP ZnuA from S. enterica (28.7% sequence identity) including absolute conservation of metal ligands and the presence of a particularly large His-rich region. The final gene, pden1597, falls somewhere in between, with slightly higher homology to SpMtsA than StZnuA (26.4 versus 20.2%, respectively) and a severely abbreviated His-rich sequence. Similarly, the presence of a His at position 204 suggests Zn 2ϩ specificity, whereas an Asp at position 279 suggests Mn 2ϩ specificity. Thus, P. denitrificans Pden1597 presents an excellent opportu-nity to test the determinants of metal ion specificity among the cluster 9 family of ABC transporters.
Here we have analyzed expression of the above genes under Zn 2ϩ -and Mn 2ϩ -limiting conditions as well as the in vitro metal binding properties and structure of Pden1597. The expression of pden1597 is increased under conditions of Zn 2ϩ starvation but not during Mn 2ϩ starvation, and heterologously expressed Pden1597 binds Zn 2ϩ with an affinity ϳ100-fold greater than that for Mn 2ϩ . Unexpectedly, the structure of Pden1597 shows that Zn 2ϩ is coordinated by Asp-279, which has never been observed for a Zn 2ϩ -specific SBP. Based on these results, we propose that Pden1597 is a Zn 2ϩ -specific transporter with an unusual metal binding site. These experiments expand the range of flexibility available to the cluster 9 SBPs in regard to metal coordination and highlight the presence of a His-rich sequence as the most reliable indication of Zn 2ϩ specificity within this family.

EXPERIMENTAL PROCEDURES
Bacterial Growth Conditions and Harvesting of Cells-A minimal media composition using an organic phosphate source and succinate as a carbon source was developed based on Graham et al. (30) and Wang et al. (31). Bulk  An overnight culture of P. denitrificans PD1222 cells grown in replete media was used to inoculate three replicates of each of the four different media types described above. The growth of cells in all four conditions was monitored at 600 nm using an Agilent Cary 60 UV-visible spectrophotometer and 5 ml of cells at midexponential phase (A 600 nm 0.4 -0.6) were harvested from each condition. To preserve RNA, 2 ml of ice-cold 5% (v/v) phenol in ethanol were added and incubated on ice for 30 min prior to centrifugation and storage at Ϫ80°C until RNA was isolated.
RNA Extraction and qRT-PCR-RNA was extracted and purified from the replicate samples described above using a PureLink RNA Mini Kit (Ambion) according to the manufacturer's instructions. Contaminating DNA was removed by an on-column DNase digestion protocol (Invitrogen). RNA concentration and purity were determined spectrophotometrically using a Nanodrop Spectrophotometer (ND 1000).
Purified RNA (1 g) was reverse transcribed to cDNA using iScript TM cDNA synthesis kit (Bio-Rad) according to the manufacturer's instructions. cDNA was diluted 5-fold in H 2 O before use in qPCR. Primers were designed to amplify 100 -150 bp regions of each gene with a T m ϳ60°C and used at a final concentration of 0.3 M. Quantitative real-time PCR was done using Power SYBR Green PCR Master Mix (Applied Biosystems) and a CFX96 RT-PCR real-time system combined with a C1000 Thermal cycler (Bio-Rad). Relative transcript abundance was normalized to dnaN, encoding the ␤-subunit of DNA-polymerase III (Pden_0970), a housekeeping gene previously used for P. denitrificans RT-PCR experiments (32).
Cloning, Heterologous Expression, and Purification-The entire pden1597 gene was amplified by PCR from genomic DNA using the following primers: 5Ј-GCTGTCCATGGCAA-TGAAAGACTGGCTCTTCCGC-3Ј forward and 5Ј-ACTAT-GGATCCTCAGCGGGCGGCCAGCGC-3Ј reverse. The PCR product was cloned into a pET-45(b) ϩ vector (Novagen) using NcoI and BamHI restriction sites, resulting in a translated protein lacking any affinity tags and containing the N-terminal periplasmic targeting sequence. Plasmid was transformed into BL21(DE3) Escherichia coli cells, which were grown in LB medium containing 100 g/ml of ampicillin at 37°C and 250 rpm to an A 600 ϭ 0.8 -1.0. Overexpression was then induced by addition of isopropyl 1-thio-␤-D-galactopyranoside to 1.0 mM, the temperature was decreased to 20°C, and the cells were grown with shaking overnight. Cells were harvested by centrifuging at 3000 ϫ g for 30 min at 4°C.
The periplasmic fraction was obtained using an osmotic shock protocol adapted from Wang et al. (31). Briefly, the cell pellet was resuspended at 5 ml/g of wet weight cells in 50 mM phosphate, pH 8, 0.5 M sucrose, 0.67 mM EDTA, and 7.5 mg/ml of lysozyme and incubated at 30°C for 15 min. An equal volume of deionized water was added and incubated a further 45 min at 30°C. NaCl and MgCl 2 were then added to 300 and 0.5 mM, respectively. Cell debris was removed by centrifugation at 25,000 ϫ g for 30 min at 4°C, leaving the periplasmic fraction in the supernatant. Polyethyleneimine was added to 0.5% (v/v) to precipitate nucleic acids, which were removed by centrifugation at 25,000 ϫ g for 30 min at 4°C. Finally, the protein was precipitated by addition of ammonium sulfate to 60% (w/v) and centrifuged as above.
The ammonium sulfate pellet was dissolved in 20 mM Tris, pH 8.0, and centrifuged at 20,000 ϫ g at 4°C for 30 min to remove any remaining particulates. The supernatant was loaded onto a HiTrap Q HP column (GE Healthcare) equilibrated with 20 mM Tris, pH 8.0. The NaCl concentration was stepped to 150 mM NaCl and a linear gradient of NaCl applied to 350 mM NaCl. The peak containing Pden1597 eluted at ϳ250 mM NaCl. Fractions containing Pden1597 were combined and concentrated to 500 l, and applied to a HiPrep Sephacryl S-100 HR column (GE Healthcare) equilibrated with 20 mM Tris, pH 8.0, 150 mM NaCl, 0.1 mM DTT. At this point, the protein was highly pure as judged by SDS-PAGE. Protein concentration was determined using an extinction coefficient at 280 nm of 19,691 M Ϫ1 cm Ϫ1 calculated as previously described (33).
In-gel Trypsin Digests and Mass Spectrometry-In SDS-PAGE gels of Pden1597 purification fractions, the band corresponding to Pden1597 in the periplasmic fraction was excised and subjected to in-gel proteolysis (34) using trypsin (New England Biolabs). Extracted peptides were then evaporated to dryness, reconstituted in 0.1% (v/v) formic acid, and desalted/exchanged into 75:25 (v/v) acetonitrile/water with 0.1% (v/v) formic acid using C18 resin ZipTip pipette tips (Millipore) prior to introduction into the mass spectrometer. Peptide solutions were analyzed by direct-infusion electrospray ionization Fourier transform-ion cyclotron resonance mass spectrometry (Thermo LTQ FT) in positive ion mode. Nanoelectrospray was performed with an Advion NanoMate and the resulting peptide ions were measured in a parent-ion Fourier transform-ion cyclotron resonance scan at a resolving power of m/⌬m 50% ϭ 100,000 at m/z 400. Fragment ion spectra were generated in the linear ion trap with peak picking based on the Fourier transform-ion cyclotron resonance spectrum and dynamic elusion enabled. Parent ion and MS/MS data were manually inspected using Xcalibur software version 2.1.0 to identify and confirm peptides of interest.
Metal Content Analysis-Protein samples at a concentration of 10 -20 M were digested in 4 M HNO 3 overnight at 70°C to ensure their complete degradation. Prior to metal analysis, samples were diluted 2.5-fold with MilliQ water. For depleted media samples, 2.5 ml were combined with 0.5 ml of concentrated HNO 3 and digested overnight at 70°C. Samples were analyzed on a Perkin-Elmer 2100 DV inductively coupled plasma-optical emission spectrometer (ICP-OES), calibrated with a multielement standard (Ricca Chemical). The wavelengths for measuring Mn 2ϩ and Zn 2ϩ were 257.610 and 206.200 nm, respectively, and samples were run in triplicate.
Generation of Apo-proteins-Apo-Pden1597 was generated by dialyzing the isolated protein against two changes of 500 ml of 50 mM NaOAc buffer, pH 4.5, 50 mM EDTA, and 150 mM NaCl. This was followed by a final dialysis against 500 ml of 20 mM Tris buffer, pH 8.0, 150 mM NaCl, and 3.4 g/liter of Chelex. The structural integrity of the protein was determined by the use of size exclusion chromatography, where it ran as a single peak consistent with its intact, monomeric molecular weight. Metal content was assessed by ICP-OES.
Metal Binding Affinity-All of the metal binding affinity experiments were performed in a 20 mM HEPES, pH 7.2, 200 mM NaCl, and 5% (v/v) glycerol buffer. This buffer was run through a Chelex column to remove any divalent metal ion contaminants. The fluorescent dye mag-fura-2 (MF-2) (Invitrogen) was used to analyze the metal binding affinities of apo-Pden1597 for Mn 2ϩ and Zn 2ϩ as previously described (35). All fluorescence measurements were made using a Varian Cary Eclipse fluorescence spectrophotometer with entrance and exit slits set to 10 nm. The protein concentration was measured before each experiment and the MF-2 concentration was determined using an extinction coefficient at 369 nm of 22,000 M Ϫ1 cm Ϫ1 (35). In each experiment, 1.0 -15.0 M apo-Pden1597 and 0.4 -0.8 M MF-2 were titrated with increasing concentrations of MnCl 2 or ZnSO 2 , keeping the total volume of titrant added to less than 10% (v/v). Pden1597 as isolated with stoichiometric Zn 2ϩ was used at 15.0 M and titrated with MnCl 2 in the presence of MF-2 to assess whether Mn 2ϩ and Zn 2ϩ bind to the same site. Fluorescence excitation spectra were scanned from 250 to 450 nm while monitoring emission at 505 nm. Experiments were performed in triplicate and the fluorescence intensities at ex ϭ 330 nm (Zn 2ϩ ) or 360 nm (Mn 2ϩ ) were fit using the program DYNAFIT (36, 37) using scripts adapted from Golynskiy et al. (35). An example of a script used for fitting Mn 2ϩ binding is shown in Scheme 1.
Circular Dichroism-Circular dichroism (CD) spectra were recorded at 25°C using a Jasco-810 spectropolarimeter with a cuvette chamber regulated by a PTC-4235 Peltier device (Jasco). Apo-Pden1597 was diluted to 15 M in 5 mM HK 2 PO 4 , pH 8.0, 150 mM NaCl in a 1-mm quartz cuvette. Where appropriate, ZnSO 4 or MnCl 2 were added to 40 M. Spectra were acquired from 190 to 260 nm at 1 nm bandwidth, 2-s response time, 0.5 nm data pitch, and 10 nm/min scan speed. Each spectrum is the average of three accumulations and has been converted to mean residue ellipticity. For thermal stability experiments, the wavelength at 223 nm was monitored from 25 to 90°C every 0.2°C with a constant heating rate of 1.0°C/min. The fraction of folded protein at a given temperature T was determined using the following equation (38).
For the apo-protein, values of at 55 and 75°C were deemed to correspond to 100 and 0% folded state, respectively. After thermal denaturation experiments, the spectra from 190 to 260 nm were again collected after returning to 25°C to determine reversibility.
Crystallization and Structure Determination-Initial crystallization hits were identified using the Haupton Woodward Institute standard screen (39). These were optimized in house and diffraction quality crystals were grown under paraffin oil using a 1:1 ratio of 25 mg/ml of Pden1597 and precipitant solution containing 3.9 M sodium formate and 0.1 M bis-tris propane, pH 7.0, at 292 K. Crystals were cryoprotected with a 1:1 SCHEME 1 An unusual Zn 2؉ -specific Solute-binding Protein mixture of paratone/paraffin prior to cryocooling in liquid nitrogen.
Diffraction data were collected at 100 K on beamline 5.0.2 at the Advanced Light Source at Berkeley National Laboratory, indexed and integrated with iMosfilm (40) and scaled using Aimless (41). A polyalanine model of MtsA (PDB code 3HH8) (26) was used as the search model for molecular replacement using Phaser-MR (42). The initial solution was subjected to several rounds of automated model building, density modification, and refinement using the AutoBuild Wizard in Phenix (43). Manual model building was done in Coot (44) and further rounds of refinement were done in REFMAC (45).
Pden1597 crystallized in space group C2 with two copies of the protein in the asymmetric unit. Near the end of refinement, a single Zn 2ϩ was modeled at full occupancy bound to each protein molecule, and flexible bond restraints between Zn 2ϩ and its ligating residues were imposed at Zn 2ϩ -N (His) ϭ 2.150 Å and Zn 2ϩ -O (Asp) ϭ 2.085 Å. Solvent content of these crystals was unusually high at 62.6%. Atomic coordinates of Znbound Pden1597 have been deposited in the PDB with entry code 4XRV. Figures were prepared using PyMol, which was also used for pairwise structural alignments of Pden1597 with other solute binding proteins.

Differential Expression of pden1597 under Zn 2ϩ and Mn 2ϩ
Starvation-The expression of pden1597 and pden4140 (a ZnuA homologue) were tested using qRT-PCR of P. denitrificans cells grown in Zn 2ϩ -replete (50 M added Zn 2ϩ ), Zn 2ϩdepleted (0 M added Zn 2ϩ ), and Zn 2ϩ -chelated (0 M added Zn 2ϩ , 50 M TPEN) conditions. The third condition was chosen because ICP-OES of Zn 2ϩ -depleted media showed a residual Zn 2ϩ concentration of 350 nM, which is likely insufficient to induce conditions of extreme Zn 2ϩ starvation (46). Consistent with this finding, growth curves for cells grown in Zn 2ϩ -replete and Zn 2ϩ -depleted conditions were virtually identical, whereas growth in the presence of TPEN was significantly inhibited ( Fig.  2A). Zn 2ϩ -depleted conditions gave rise to slight increases in expression levels for both genes relative to Zn 2ϩ -replete conditions with increases in pden1597 and pden4140 expression of 1.4 Ϯ 0.1-and 2.0 Ϯ 0.5-fold, respectively (Fig. 2B). This effect was significantly enhanced in Zn 2ϩ -chelated conditions, with expression levels of pden1597 and pden4140 increased by 4.8 Ϯ 1.2-and 23.3 Ϯ 6.5-fold, respectively, relative to Zn 2ϩ -replete conditions. These results indicate that the expression of both pden1597 and pden4140 are enhanced under conditions of severe Zn 2ϩ limitation.
To assess the metal specificity of the above effect, the expression levels of pden1597 and pden4140 were compared with those of pden1259 (the putative Mn 2ϩ transporter SBP) under conditions of Mn 2ϩ limitation. Unlike Zn 2ϩ , simple omission of Mn 2ϩ from the culture medium resulted in a concentration of less than 40 nM by ICP-OES, sufficient to induce a relatively mild but significant growth defect in P. denitrificans ( Fig. 2A). Under these conditions, neither pden1597 nor pden4140 were significantly up-regulated, whereas expression of pden1259 increased by 5.6 Ϯ 0.6-fold relative to Mn 2ϩ -replete conditions (Fig. 2C). These results suggested that like the ZnuA homo- logue pden4140, pden1597 is a Zn 2ϩ -specific transporter of the cluster 9 family of ABC transporters and prompted further studies on purified Pden1597 protein.
Expression and Purification of Pden1597-A plasmid bearing the full-length pden1597 gene, including the N-terminal 22 residues comprising the predicted periplasmic localization signal peptide (47), was transformed into E. coli and overexpressed (Fig. 3A). SDS-PAGE identified a band of the correct size for full-length Pden1597 in the induced whole cell fraction, whereas a band of slightly smaller apparent M r was enriched in the periplasmic fraction. This was subsequently identified by in-gel trypsin digest and mass spectrometry as Pden1597. A non-tryptic cleavage site between Ala-24 and Glu-25 was observed (Fig. 3C), indicating that the signal sequence is recognized by E. coli and cleaved upon export into the periplasm. The processed protein was purified to homogeneity from the periplasmic fraction by ion exchange and size exclusion chro-matography (Fig. 3B). Size exclusion chromatography results further show that Pden1597 migrates exclusively as a monomer, with the molecular weight determined by elution time closely matching that calculated from primary sequence. Analysis of metal content by ICP-OES revealed that Pden1597 was isolated with nearly stoichiometric Zn 2ϩ (0.95 equivalents) and undetectable levels of Mn 2ϩ consistent with a specific role in Zn 2ϩ import.
Metal Binding Affinities-Apo-Pden1597 was generated as described under "Experimental Procedures" and confirmed by ICP-OES to contain no more than 0.05 eq of Zn 2ϩ . Careful optimization of the pH used during metal removal dialysis was required to prevent aggregation of the protein. Only preparations that ran as a single, monomeric peak during size exclusion chromatography were used for metal binding analysis. The chelating fluorophore MF-2 forms well characterized 1:1 complexes with transition metals and competition assays using this  MAY (19, 30, 35, 48 -51). The affinities of MF-2 for Mn 2ϩ and Zn 2ϩ have been determined at 0.97 M and 36 nM, respectively, using competition assays against more weakly binding metals (35). These assays were reproduced to validate our experimental conditions, and essentially identical values (K d , Mn 2ϩ ϭ 0.97 M; K d , Zn 2ϩ ϭ 33 nM) were obtained and used for fitting analyses.

An unusual Zn 2؉ -specific Solute-binding Protein
Binding of Zn 2ϩ to MF-2 causes a shift in the fluorescence excitation peak from ϳ360 to ϳ330 nm (Fig. 4A). Comparison of titrations of 0.4 M MF-2 with Zn 2ϩ in the presence and absence of 1.0 M apo-Pden1597 (Fig. 4B) show that the protein competes very effectively against MF-2 for Zn 2ϩ binding, indicative of a significantly greater affinity. Using the determined K d value of MF-2 for Zn 2ϩ allowed us to calculate a K d of Pden1597 for Zn 2ϩ of 0.29 Ϯ 0.13 nM and stoichiometry of n ϭ 0.94 Ϯ 0.19 from three independent experiments. Similar experiments were conducted with Mn 2ϩ , which causes a quenching of fluorescence intensity (Fig. 4C). In this case, the presence of apo-Pden1597 has a more subtle, yet nonetheless, significant effect on the titration curve for MF-2 (Fig. 4D). This suggests that Pden1597 is also able to compete with MF-2 for Mn 2ϩ . Indeed, fitting analysis yields a K d of Pden1597 for Mn 2ϩ of 52 Ϯ 34 nM and n ϭ 0.54 Ϯ 0.03 from three independent experiments. The reason for the low value of n is currently unclear. To confirm Mn 2ϩ binding by Pden1597 in the presence of MF-2, we performed the experiment with an excess of apo-Pden1597 over MF-2 and obtained very similar results with K d ϭ 35 nM, n ϭ 0.52 (Fig. 4, E and F). Finally, we repeated this experiment with 15 M Pden1597 as isolated with nearly stoichiometric Zn 2ϩ to assess whether Zn 2ϩ and Mn 2ϩ compete for the same binding site. The results show that Mn 2ϩ does not bind to the protein as isolated, indicating that it cannot displace Zn 2ϩ from its binding site. It is therefore evident that Pden1597 binds both Zn 2ϩ and Mn 2ϩ at the same site with high affinity, but binds Zn 2ϩ with roughly 100-fold greater affinity, consistent with its likely role as a Zn 2ϩ -specific transporter.
Holo-and Apo-protein Stability-Circular dichroism (CD) spectroscopy was used to observe the effect of Zn 2ϩ and Mn 2ϩ binding on the thermal stability of Pden1597 (Fig. 5). CD spectra for the apo-protein were virtually identical in the presence or absence of an excess of the relevant metal, indicating that metal binding does not result in any significant rearrangement of secondary structure (Fig. 5A). These samples were then subjected to increasing temperatures up to 90°C, while monitoring the CD absorbance at 223 nm. The apo-protein exhibits a sharp transition and loss of ellipticity around T m ϭ 63°C (Fig. 5B). The presence of Mn 2ϩ had virtually no effect on melting temperature, increasing the T m by no more than 2°C. However, the Zn 2ϩ -bound protein is remarkably stable, maintaining its CD spectrum up to 90°C, indicating that the Zn 2ϩ -bound form is by far the most stable form of Pden1597 studied. Regeneration of the native spectra after returning samples to 25°C suggested that the loss of secondary structure was reversible in each case, although restoration of metal binding activity was not assessed.
Structure of Zn 2ϩ -bound Pden1597-X-ray diffraction data of Pden1597 as purified from E. coli was collected and used to solve the crystal structure of Pden1597 to a final resolution of 2.72 Å (Table 1, Fig. 6). The overall structure is highly similar to those of other cluster 9 SBPs, with two structurally similar domains connected by a long ␣-helix. Two cysteine residues, Cys-158 and Cys-165 are positioned to form a disulfide bond and likely do so in the oxidizing environment of the periplasm. However, there is no electron density between the two sulfur atoms in the current structure, which is likely a result of radiation damage causing reduction of the disulfide bond (52).
The Zn 2ϩ binding site consists of His-61 and His-138 from the N-terminal domain, and His-204 and Asp-279 from the C-terminal domain in a slightly distorted tetrahedral arrangement (Fig. 6, B and C) with average bond angles of 109 Ϯ 13°b etween the two independent metal sites in the asymmetric unit. Asp-279 is modeled as a monodentate ligand in Pden1597, as this preserves the tetrahedral symmetry of the complex and matches the coordination of the Glu ligand observed in other Zn 2ϩ -specific SBP structures of higher resolution (18,19). However, it should be noted that the resolution of the structure  does not preclude the possibility of a bidentate arrangement. Similarly, the presence of a weakly bound water ligand cannot be absolutely ruled out. However, there is no electron density consistent with a water ligand, and its presence would disrupt the preferred 4-coordinate environment of the Zn 2ϩ site (53). Second sphere hydrogen bonding interactions between the N␦ atoms of His-138 and His-204 with the carboxylate oxygen atoms of conserved Asp-136 and Glu-253 are observed for Pden1597 as well as other cluster 9 SBPs and likely pre-organize the site for metal binding.
Residues 117-133 that comprise the abbreviated His-rich loop in Pden1597 are not resolved in the current structure (Fig.  7), nor has this loop been observed in any other structure of a Zn 2ϩ -SBP to date. However, the loop may not be completely disordered in Pden1597. Significant electron density can be observed in this region, although it is insufficient to model with confidence at the current resolution. Nevertheless, it is clear that the relatively short His-rich loop of Pden1597 is positioned near the Zn 2ϩ binding site and may play a role in Zn 2ϩ acquisition or interaction with other ABC transporter components as has been suggested for other Zn-SBPs as discussed below.

DISCUSSION
The cluster 9 family of bacterial ABC transporters has been extensively studied due to its role in virulence (3). However, despite intense interest, the structural features mediating metal specificity and transport among this family remain imperfectly understood. We have chosen to study Pden1597 because it possesses attributes intermediate between Zn 2ϩ -specific and Mn 2ϩ -specific ABC transporters. To determine structural attributes critical for metal specificity in this protein, it was first essential to unambiguously assign that specificity. For this reason, we have determined the expression profiles for pden1597 under conditions of Mn 2ϩ and Zn 2ϩ starvation as well as the in vitro binding affinities of the recombinant protein for these metals.
We have demonstrated using qRT-PCR that transcription of pden1597 is sensitive to the availability of Zn 2ϩ but not Mn 2ϩ . This is likely due to the presence of a putative binding site for the Zn 2ϩ -specific transcriptional regulator Zur in the promoter region of pden1597 (54). We also determined dissociation constants for Pden1597 in the subnanomolar and low nanomolar ranges for Zn 2ϩ and Mn 2ϩ , respectively. This would suggest that Pden1597 could function as both a Zn 2ϩ and Mn 2ϩ carrier.  The ability to bind both metals at high affinity is shared by the homologues TroA (55), PsaA (24), and SitA (15). TroA and PsaA exhibit preferences for Mn 2ϩ of ϳ3and ϳ70-fold, respectively, whereas the binding affinities of SitA for these metals appear to be similar, although precise K d values have not been determined. In the case of Pden1597, there is a Ͼ100-fold preference for binding Zn 2ϩ over Mn 2ϩ . Combined with the sensitivity of pden1597 transcription to Zn 2ϩ but not Mn 2ϩ availability, and the observation that recombinantly expressed Pden1597 contains Zn 2ϩ exclusively, our data strongly suggests that Zn 2ϩ is the physiologically relevant substrate for Pden1597.
CD spectra of Pden1597 are consistent with a predominantly ␣ helical secondary structure that is essentially unchanged by Zn 2ϩ or Mn 2ϩ binding. However, the Zn 2ϩ -bound form of Pden1597 was remarkably stabilized to thermal denaturation relative to the apo form, whereas Mn 2ϩ binding had almost no effect. This trend has been observed for both Zn 2ϩ -and Mn 2ϩspecific SBPs (19,24,55). In fact, imperfect coordination of Mn 2ϩ relative to Zn 2ϩ in PsaA leading to binding reversibility was shown to be essential for Mn 2ϩ delivery to the membrane permease by PsaA (20). Thus, our CD results should not be taken as an indication of specificity. Rather, they show that Zn 2ϩ binding to Pden1597 is very tight, with an essentially negligible off-rate. This in turn indicates that some conformational change accompanying binding to the membrane permease or other component of the ABC transporter system must be required for Zn 2ϩ release.
It has been pointed out that specificity in the cluster 9 SBPs seems to depend upon three structural features (19). 1) A His at position 204 (Pden1597 numbering) indicates Zn 2ϩ specificity, whereas a Glu at this position indicates Mn 2ϩ specificity. 2) Coordination by Asp-279 indicates Mn 2ϩ specificity. 3) The presence of a His-rich loop indicates Zn 2ϩ specificity. To our knowledge, the structures of at least 10 different cluster 9 SBPs from various organisms have been determined by x-ray crystallography. Based on the above criteria, only the ZnuA homologues (14, 16 -19) are predicted to be specific for Zn 2ϩ , whereas the SitA (15), MntC (21,22,25), PsaA (20,23,24), and MtsA (26) homologues should be specific for Mn 2ϩ . TroA is the only member of the group that violates these rules, having both His-205 (Zn-specific) and Asp-289 (Mn-specific) ligands and no His-rich loop (Mn-specific) (27)(28)(29). Although Mn 2ϩ specificity has been demonstrated by troA knock-out studies in S. suis (56), it has a His 3 , 1 carboxylate coordination sphere only observed thus far in Zn 2ϩ -specific SBPs and binds Zn 2ϩ with only slightly lower affinity than Mn 2ϩ .
The role of coordination geometry in Zn 2ϩ or Mn 2ϩ specificity in the cluster IX SBP remains somewhat enigmatic. A relatively recent analysis of non-redundant metalloprotein crystal structures shows a strong preference of Zn 2ϩ and Mn 2ϩ for 4-and 6-coordinate environments, respectively (53). Pden1597 shares an identical tetrahedral coordination sphere with TroA (Fig. 7), yet has a roughly 100-fold greater affinity for Zn 2ϩ over Mn 2ϩ , whereas TroA has nearly equal affinities for both metals. Given that all TroA and Pden1597 structures are in complex with Zn 2ϩ rather than Mn 2ϩ , it is possible that the binding site is sufficiently flexible that the coordi-nation sphere of Mn 2ϩ in these proteins may differ from that of Zn 2ϩ . However, structures of SitA (15) and PsaA (20) bound to both metals have been solved and show only very subtle changes in coordination. Indeed, non-ideal geometry around the Mn 2ϩ ion in PsaA is thought to facilitate its release to the membrane permease, whereas the Mn 2ϩ ion in SitA is in a nearly perfect octahedral geometry. Thus, it is clear that factors other than the coordination sphere have significant impacts on the specificity of metal binding and release in the cluster 9 family.
Pden1597 appears to have primary structural attributes consistent with both Mn 2ϩ and Zn 2ϩ specificity, including the presence of an abbreviated His-rich loop (Zn-specific) and coordination by Asp-279 (Mn-specific). To our knowledge, Pden1597 is the first SBP for a cluster 9 ABC transporter system that has confirmed specificity for Zn 2ϩ , yet uses Asp-279 as a metal ligand. This would suggest that the most reliable determinant of Zn 2ϩ specificity identifiable from the primary sequence alone is the presence of a His-rich loop near the metal binding site. This feature is proposed to be important in Zn 2ϩ recruitment, mediating structural changes allowing metal transfer to the membrane permease, and/or Zn 2ϩ sensing or scavenging from the periplasm (14,17,19,57). Indeed, a second Zn 2ϩ ion was identified bound to a His residue from the Hisrich loop of E. coli ZnuA (19). However, we see no evidence for a second Zn 2ϩ binding site in Pden1597. Therefore, the unstructured loop of Pden1597 may play a more significant role in the recognition of other components of the ABC transporter system than in metal acquisition.
The unstructured loop of Pden1597 is unusual not only for its brevity but also its sequence. There are relatively few His and Asp/Glu residues. Also, two Tyr residues are present that are not typically seen in these loop features. There are a few close homologues of Pden1597 with sequence identities greater than 50%, which include representatives from the human pathogens Klebsiella pneumonia and Enterobacter aerogenes. Among these, the sequence of the unstructured loop is well conserved and the residues of the metal binding site are absolutely conserved. None of these homologues have been characterized at any level to our knowledge, so it remains to be seen whether these will also be specific for Zn 2ϩ binding. However, simply on the basis of sequence, Pden1597 and its homologues seem to form a subclass within the cluster 9 SBPs.
Here we have described the metal-dependent transcriptional regulation, metal binding properties, and structure of Pden1597, a previously uncharacterized SBP of the cluster 9 family of ABC transporters. On the basis of qRT-PCR results and the significant preference for Zn 2ϩ binding over Mn 2ϩ , we propose that the physiologically relevant metal for this ABC transporter system is Zn 2ϩ . Due to the presence of a znuABC homologue in P. denitrificans, we suggest the name aztABC (auxiliary Zn 2ϩ transporter) for the operon composed of pden1595-1597. Whether there are specific roles for the znuABC and aztABC operons in Zn 2ϩ homeostasis or if they are simply redundant systems remains to be seen and experiments to determine this are currently underway in our laboratory.