Molecular landscape of the interaction between the urease accessory proteins UreE and UreG

Biochimica et Biophysica Acta 1844 (2014) 1662–1674

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap

Molecular landscape of the interaction between the urease accessory proteins UreE and UreG Anna Merloni, Olena Dobrovolska, Barbara Zambelli, Federico Agostini 1, Micaela Bazzani, Francesco Musiani, Stefano Ciurli ⁎ Laboratory of Bioinorganic Chemistry, Department of Pharmacy and Biotechnology, University of Bologna, Italy

a r t i c l e

i n f o

Article history: Received 23 May 2014 Received in revised form 12 June 2014 Accepted 19 June 2014 Available online 27 June 2014 Keywords: Urease UreE UreG Nuclear magnetic resonance Calorimetry Protein–protein docking

a b s t r a c t Urease, the most efficient enzyme so far discovered, depends on the presence of nickel ions in the catalytic site for its activity. The transformation of inactive apo-urease into active holo-urease requires the insertion of two Ni(II) ions in the substrate binding site, a process that involves the interaction of four accessory proteins named UreD, UreF, UreG and UreE. This study, carried out using calorimetric and NMR-based structural analysis, is focused on the interaction between UreE and UreG from Sporosarcina pasteurii, a highly ureolytic bacterium. Isothermal calorimetric protein–protein titrations revealed the occurrence of a binding event between SpUreE and SpUreG, entailing two independent steps with positive cooperativity (Kd1 = 42 ± 9 μM; Kd2 = 1.7 ± 0.3 μM). This was interpreted as indicating the formation of the (UreE)2(UreG)2 hetero-oligomer upon binding of two UreG monomers onto the pre-formed UreE dimer. The molecular details of this interaction were elucidated using highresolution NMR spectroscopy. The occurrence of SpUreE chemical shift perturbations upon addition of SpUreG was investigated and analyzed to establish the protein–protein interaction site. The latter appears to involve the Ni(II) binding site as well as mobile portions on the C-terminal and the N-terminal domains. Docking calculations based on the information obtained from NMR provided a structural basis for the protein–protein contact site. The high sequence and structural similarity within these protein classes suggests a generality of the interaction mode among homologous proteins. The implications of these results on the molecular details of the urease activation process are considered and analyzed. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Urease (E.C. 3.5.1.5) is a non-redox nickel-dependent enzyme [1] that catalyzes urea hydrolysis in the last step of organic nitrogen mineralization [2–9] at a rate fifteen orders of magnitude larger than the spontaneous reaction [10]. Urease is the main virulence factor in a large variety of lethal human pathogens such as Mycobacterium tuberculosis, Yersinia enterocolitica, and Cryptococcus neoformans. Infections of the urinary and gastrointestinal tracts in human and animals by ureolytic bacteria such as Proteus mirabilis can cause kidney stone formation, catheter encrustation, pyelonephritis, ammonia encephalopathy, and hepatic coma. Helicobacter pylori is a bacterium able to survive in the acidic environment of the stomach by exploiting the pH increase caused by the urease activity, and acting as the major cause of pathologies, including cancer induced by gastro-duodenal infections. Control of urease activity thus represents an important goal to be

⁎ Corresponding author at: Laboratory of Bioinorganic Chemistry, Department of Pharmacy and Biotechnology, University of Bologna, Viale Giuseppe Fanin 40, I-40127, Bologna, Italy. Tel.: +39 051 209 6204; fax: +39 051 209 6203. E-mail address: [email protected] (S. Ciurli). 1 Present address: Centre for Genomic Regulation (CRG), Barcelona, Spain.

http://dx.doi.org/10.1016/j.bbapap.2014.06.016 1570-9639/© 2014 Elsevier B.V. All rights reserved.

pursued for the battle against deadly pathologies globally affecting the world population [2,5,11]. Urease is an hetero-polymeric enzyme that possesses multiple and identical active sites, each containing two essential Ni(II) ions bridged by a carbamylated lysine and a hydroxide ion [12–16]. Apo-urease is inactive, and its activation requires several chemical modifications such as GTP hydrolysis, lysine carbamylation through CO2 uptake, and final delivery of Ni(II) into the active site, steps that are typically carried out by four accessory proteins named UreD, UreF, UreG, and UreE. The most currently accepted hypothesis for this process [8,17] involves the sequential binding of UreD, UreF, and UreG, or of a preformed aggregate of UreD, UreF, and UreG (UreDFG), to obtain a pre-activation complex that carbamylates the active site lysine side chain and further binds Ni(II) ions delivered by UreE [18–21] through a route driven by GTP hydrolysis. Even though the structure of urease bound to any of the accessory proteins is not yet available, the crystal structure of the (UreDFG)2 complex from H. pylori has been recently determined [22]. On the other hand, structural information on UreE proteins from various bacteria has been derived from numerous crystallographic studies: UreE from Sporosarcina pasteurii (formerly known as Bacillus pasteurii, SpUreE) [23,24], Klebsiella aerogenes (KaUreE) [25], and H. pylori (HpUreE) [26,

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

27] displays a similar fold made by a symmetric homo-dimer, with each protomer composed of N- and C-terminal domains connected by flexible linkers (Fig. 1). The N-terminal domains of UreE, residing at the periphery of the molecule, consist of two three-stranded β-sheets stacked nearly perpendicularly upon each other, with a short helical region between the two sheets. The C-terminal domains, involved in a head-to-head dimerization, feature a four-stranded anti-parallel βsheet and α-helices organized in a ferredoxin-like βαββαβ fold. The metal-binding site, situated at the edge of the dimerization interface on the surface of the protein, involves two conserved histidines (H100 for SpUreE, H96 for KaUreE, and H102 for HpUreE). The C-terminal region of UreE, not observed by crystallography because of disorder in the solid state, features a variable number of histidines depending on the biological source [28]. This phenomenon causes the observation of different nickel-binding stoichiometries, ranging from one Ni(II) per dimer in the case of HpUreE [27,29,30], to two Ni(II) per dimer for SpUreE [24,31], up to six Ni(II) per dimer associated to KaUreE [32]. Indeed, in the case of SpUreE two histidines located along the C-termini have been found to participate in metal binding, as shown by NMR [33,34] and X-ray crystallography coupled with calorimetry [24]. In addition to its role as a Ni(II) metallo-chaperone, UreE appears to be involved in the enhancement of the GTPase activity of UreG [21]. This process involves the direct interaction of UreE with UreG observed, in the case of the proteins from H. pylori, both in vivo through twohybrid and immunoprecipitation experiments [35] and in vitro using calorimetry and NMR spectroscopy [30]. A structure of the HpUreEHpUreG (HpUreEG) complex was obtained by molecular modeling, suggesting that the protein–protein interface comprises the metal-binding sites on both HpUreE (H102) and HpUreG (C66 and H68) [30]. Consistent with this model, a novel metal site was observed in the HpUreEG

1663

complex that featured a nanomolar binding affinity for Zn(II), ca. three orders of magnitude tighter as compared to the separate proteins [30]. A protein complex between the HpUreEG model complex and the HpUreE-HpUreG (HpUreFG) crystallographic complex has been calculated [36]. However, experimental information on the structure of UreEG complexes from any biological source is still missing. In this paper we describe the results of solution studies aimed at deriving structural information on the complex involving UreE and UreG from S. pasteurii. In particular, we established the stoichiometry and the thermodynamic parameters for this interaction using isothermal titration calorimetry (ITC) and we employed NMR spectroscopy to investigate the chemical shift perturbations induced by the protein–protein interaction. This information was used to map the protein surface on SpUreE involved in the complex formation with SpUreG using NMR-driven docking calculations. Knowledge of the structural details of the protein interaction network that leads to urease activation is a prerequisite for the design and development of new molecules that could interfere with this process, and potentially allow the design of drug for the eradication of human pathogenic ureolytic bacteria. 2. Materials and methods 2.1. SpUreE expression and purification Recombinant wild-type UreE from S. pasteurii (GenBank: AAD55 059.1, 147 residues) was purified using a modification of a previously reported protocol [31]. E. coli BL21(DE3) (Stratagene) cells were transformed by heat shock with the pET-3d::ureE plasmid containing the wild type ureE gene (512 bp), as previously described [31]. The

Fig. 1. Ribbon scheme of the structure of SpUreE homodimer (PDB ID: 1EAR); the two monomers are colored in blue and orange, with the darker and lighter portions representing the Nterminal and C-terminal domains, respectively. The views in the top and bottom panel are rotated by 90° along the horizontal axis. The position of the N- and C-termini, as well as the metal binding conserved histidine, is indicated. Picture produced using Chimera [58].

1664

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

transformed cells were subjected to a double-selection process aimed at the production of high yields of the recombinant protein [37]. The selected colonies were grown aerobically under vigorous stirring in a medium supplemented with 100 μg/mL ampicillin. The LB medium was used for the unlabeled protein, and the E. coli OD2-DN (2H, 15 N N 98%, Silantes) or the E. coli OD2-CDN (2H ≥ 95%, 13C, 15N N 98%, Silantes) rich growth media were used to obtain the correspondingly labeled SpUreE. For deuterated samples, the cells were grown in media containing increasing amounts of D2O (30%, 60%, 90%, 100%) to allow for adaptation. Expression was obtained by incubating the cells at 37 °C for 16 h after addition of 0.5 mM IPTG when OD600 reached 0.8–0.9. The cells were harvested by centrifugation at 10,000 ×g for 30 min at 4 °C, resuspended in 50 mM Tris–HCl buffer, pH 7.0, containing 10 mM MgCl2, 2 mM dithiothreitol and 20 μg/mL DNase-I, and broken by two passages through a French Pressure cell (SLM-Aminco) operating at 20,000 psi. Cell debris was separated from the supernatant by centrifugation at 75,600 ×g for 30 min at 4 °C, and solid AMS was added to the soluble portion of the cellular extract up to 60% saturation. The clear supernatant, obtained after centrifugation at 75,600 ×g for 30 min at 4 °C, was loaded onto a Phenyl-Sepharose XK 26/10 column (GE Healthcare) pre-equilibrated with 50 mM Tris–HCl buffer, pH 7.0, containing 2 M AMS and 5 mM EDTA. The unbound proteins were washed away using the equilibration buffer until the baseline was stable. The column was eluted using 300 mL of a linear gradient of AMS (from 2 to 0 M) in 50 mM Tris–HCl buffer, pH 7.0, 5 mM EDTA, with a flow rate of 3 mL/min. SpUreE was eluted at about 1.2 M AMS and the collected fractions were dialyzed against 50 mM Tris–HCl buffer, pH 7.0, 5 mM EDTA. The resulting solution was applied onto a Q-Sepharose XK 16/10 column (GE Healthcare) preequilibrated with 50 mM Tris–HCl buffer, pH 7.0, 5 mM EDTA. The column was washed with the same buffer and the protein was eluted with 300 mL of a linear gradient of NaCl from 0 to 1 M, at 3 mL/min, in the same buffer. Fractions containing SpUreE, eluted at about 0.2 M NaCl, were combined, concentrated to 2 mL using an Amicon Ultra-15 centrifugal filter unit (10 kDa molecular weight cut-off, Millipore), and loaded onto a Superdex-75 16/60 size-exclusion column, eluted using 20 mM HEPES buffer, pH 7.0, 150 mM NaCl in order to exchange the buffer and remove EDTA. Protein purity was analyzed using SDS-PAGE and SimplyBlue Safestain (Invitrogen) staining. Protein quantification was performed by absorption spectroscopy, using the theoretical value of 21,430 M− 1 cm− 1 for the extinction coefficient at 280 nm, calculated using the ProtParam tool () and the amino acid sequence of the protein. The final yield was ca. 100 mg pure SpUreE per liter of culture. The purified protein was devoid of metal ions as shown by inductively coupled plasma emission spectrometry (ICP-ES) [31]. 2.2. SpUreG expression and purification SpUreG was expressed and purified following a modification of a previously reported protocol [38]. E. coli BL21(DE3) (Stratagene) cells, transformed with the pET3a::ureG plasmid [38], were grown at 37 °C in LB medium with vigorous stirring until the OD600 reached 0.5–0.6. Protein expression was induced with 0.5 mM IPTG at 20 °C for 16 h. Cells were harvested by centrifugation at 10,000 ×g for 30 min at 4 °C, re-suspended in 50 mM Tris–HCl buffer, pH 8.0, containing 10 mM MgCl2, 1 mM DTT, 5 mM EDTA and 20 μg/mL DNase-I, and lysed as described above for SpUreE. The obtained lysate was centrifuged and the supernatant was loaded onto a Q-Sepharose XK 26/10 column (GE Healthcare), pre-equilibrated with two volumes of 20 mM Tris–HCl buffer at pH 8.0 containing 5 mM EDTA and 1 mM DTT. The unbound proteins were washed away using the equilibration buffer until the baseline was stable. SpUreG was then eluted using 300 mL of a linear gradient of NaCl from 0 to 1 M, at 3 mL/min. Fractions containing SpUreG were collected and solid AMS was slowly added under continuous mixing at 4 °C until 1 M concentration was reached. The resulting

suspension was centrifuged at 75,600 ×g for 30 min at 4 °C and the supernatant was loaded onto a Phenyl Sepharose 16/10 column (GE Healthcare), pre-equilibrated with 20 mM Tris–HCl buffer at pH 8.0 containing 1 M AMS and 1 mM DTT. The column was washed with the same buffer, and the protein was eluted with a linear gradient of 300 mL of AMS from 1 M to 0 M at 3 mL/min. The fractions containing SpUreG were collected, concentrated using an Amicon Ultra-15 centrifugal filter unit (10 kDa molecular weight cut-off, Millipore), and then loaded onto a Superdex 75 16/60 column (GE Healthcare) pre-equilibrated with 20 mM HEPES buffer at pH 7.0, containing 150 mM NaCl and 1 mM TCEP. The fractions containing SpUreG were collected and concentrated as above. TCEP was removed using Zeba Spin Desalting Columns (Thermo Scientific) just before each NMR experiment, using the same buffer to maintain the protein in the reduced state as long as possible to prevent oxidationinduced dimerization. Protein purity and quantification were assayed with procedures analogous to those used for SpUreE, in this case using a theoretical value of 11,460 M− 1 cm− 1 for the extinction coefficient at 280 nm. The purified protein was devoid of metal ions as shown by inductively coupled plasma emission spectrometry (ICP-ES) [31].

2.3. SpUreE–SpUreG interaction followed by ITC (isothermal titration calorimetry) Titration experiments were typically performed at 25 °C using a high-sensitivity VP-ITC microcalorimeter (MicroCal LLC, Northampton, MA, U.S.A.). The calorimeter consists of a reference cell filled with deionized water and a sample cell where the reaction occurs, and the two cells are maintained at the same temperature. The method monitors the fraction of the protein, contained in the reaction cell, which reacts with the partner protein added by multiple injections through a stirring syringe. This is done by probing the heat generated over time, defined as thermal power, necessary to maintain the reaction cell at the same constant temperature of the reference cell. The heat involved in the reaction induced by each addition, calculated by integrating each peak of thermal power over time, is proportional to the molar enthalpy of the reaction and to the moles of protein–protein complex formed, in turn proportional to the total cell volume and to the differential molar concentration of bound titrant. SpUreE and SpUreG, freshly eluted from a Superdex-75 sizeexclusion column using 20 mM Tris–HCl buffer at pH 7.0 containing 150 mM NaCl, were loaded immediately in the ITC instrument. A solution of SpUreG (50 μM monomer) was degassed and loaded into the sample cell (V = 1.4093 mL), while the reference cell was filled with deionized water. The syringe was filled with a 150 μM solution of SpUreE dimer and the temperature of the two cells was set and stabilized at 25 °C. Stirring speed was 300 rpm, and thermal power was monitored every 2 s using high instrumental feedback. A series of 28 injections (10 μL each) of SpUreE solution was carried out at intervals of 300 s, a time necessary to allow the system to reach thermal equilibrium after each injection. A control experiment was carried out by adding the titrating solution into the buffer alone, under identical conditions. Heats of dilution were negligible. The integrated heat data were analyzed using the Origin software package (MicroCal), and fitted using a non-linear least-squares minimization algorithm to theoretical titration curves that involved different binding models. The reduced parameter χ2v (χ2v = χ2/N, where N is the degrees of freedom, N = Nidp − Npar, Nidp = number of points, and Npar = number of parameters floating in the fit) was used to establish the best fit among the tested models. Values for the enthalpy change of reaction (ΔH), the binding affinity constant (Kb) and the number of sites (n) were the parameters of the fit. The reaction entropy was calculated using the equations ΔG = −RTlnKb (R = 1.9872 cal mol− 1 K− 1, T = 298 K) and ΔG = ΔH − TΔS.

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

2.4. NMR spectroscopy data collection and analysis for backbone assignment NMR spectra were acquired at 25 °C on a Bruker AVANCE 900 spectrometer, operating at the proton nominal frequency of 899.2 MHz (21.1 T) on NMR samples containing 0.5 mM of labeled SpUreE dimer in 20 mM HEPES buffer at pH 7.0, containing 150 mM NaCl, in 90% H2O and 10% D2O. At this protein concentration, SpUreE is reported to be a dimer [31,34,39]. The spectrometer was equipped with a TCI 5mm triple resonance cryo-probe with Pulsed Field Gradients along the z-axis. 2D 1H-15N HSQC and 1H-15N TROSY-HSQC, 3D TROSY-HNCO, TROSY-HNcaCO, TROSY-HNCA, TROSY-HNcoCA, TROSY-HNCACB, and TROSY-HNcoCACB experiments on the 2H/15N/13C-labeled SpUreE were used to obtain the sequential backbone resonance assignment of 1 H, 13C, and 15N nuclei (Table 1-SI). In these pulse schemes, water suppression is achieved using selective pulse and transverse signal cancellation with pulsed field gradients associated with a flip-back pulse. The NMR data were processed with the NMRpipe/NMRDraw software package v. 8.1 [40] using a squared cosine function to apodize the data and zero-filled once in all dimensions before Fourier transformation, using forward and backward linear prediction for the indirect dimensions to increase resolution. Spectral analysis for resonance assignment was performed using CARA 1.8.4.2 [41]. The assignment is reported in Table 2-SI. 2.5. SpUreE–SpUreG interaction followed by NMR 2 H/15N-labeled 0.5 mM SpUreE dimer was mixed with unlabeled SpUreG monomer at 1:1, 1:2, 1:3 and 1:4 ratios in 20 mM HEPES buffer at pH 7.0, containing 150 mM NaCl, in 90% H2O and 10% D2O, at 25 °C using a Bruker AVANCE 950 spectrometer, operating at the proton nominal frequency of 950.2 MHz (22.3 T). The spectrometer was equipped with a TCI 5-mm triple resonance cryo-probe with Pulsed Field Gradients along the z-axis. Corresponding 1H-15N TROSY-HSQC spectra were recorded and compared to the 1H-15N TROSY-HSQC spectrum of free SpUreE. The original data were zero-filled four times in F1 and eight times in F2 prior to Fourier transformation, and mild resolution enhancement was achieved by applying a π/3-shifted sine-squared apodization function in both dimensions, using the iNMR software (www.inmr.net). The chemical shifts were determined using the spectrum of SpUreE in the absence and in the presence of four equivalents of SpUreG monomer, using the peak-picker tool in iNMR. Chemical shift perturbations (CSP) were calculated using the formula: Δδ = ΔHN + ΔN/7 (or Δδ = ΔHN + ΔN/5 for glycines), where ΔHN and ΔN are the absolute values of the chemical shift differences (ppm) of the amide proton (HN) and nitrogen (N) resonances, respectively [42,43].

2.6. Protein-protein docking calculations 2.6.1. The model of SpUreG The alignment between the sequences of UreG from H. pylori strain 26695 (HpUreG) and SpUreG was produced using the Promals3D server [44]. The sequence identity between HpUreG and SpUreG is 58%. The crystal structure of the dimeric form of HpUreG in complex with HpUreF and HpUreD (PDB ID: 4HI0) [22], was used as template structure to calculate the model structure of the dimer of SpUreG using MODELLER 9v12 [45]. 100 models were generated imposing symmetry restraints on protein chains and including the GDP molecules found in the crystal structure of HpUreG. The best model was selected on the basis of the lowest value of the DOPE score in MODELLER [46], and was subjected to a refining step of loop optimization using MODELLER. The stereochemical quality of the structures was established using PROCHECK [47]. The results of this analysis (most favored and additionally allowed residues in the Ramachandran plot 92.4% and 7.6%, respectively) confirm the high reliability of the model structure (Fig. 1-SI).

1665

2.6.2. Normal conformational mode analysis of SpUreE The five low-frequency normal conformational modes of the dimeric SpUreE crystal structure (PDB ID: 4L3K) [24], were calculated using the elNémo web-server [48] (http://www.igs.cnrs-mrs.fr/elnemo/) and a recently published protocol [49]. The eleven protein conformations determined using this analysis, which comprised the starting structure as well as the ten structures derived by applying the perturbations consistent with each calculated normal mode to the starting structure, were utilized to build a library of structures to be used in the subsequent docking stages (Fig. 2-SI).

2.6.3. Calculation of the SpUreEG complex The SpUreE structures were docked onto the SpUreG model structure using the data-driven docking program HADDOCK 2.1 [50,51]. HADDOCK (High Ambiguity Driven biomolecular DOCKing) implements an approach that uses biochemical and/or biophysical interactions data to drive the docking process. The calculation was guided by defining selected residues as “active” in the protein–protein interaction. The docking algorithm rewards the complexes that have these active residues on the interaction interface. For SpUreE, the active residues were those identified by NMR chemical shift perturbations (residues 32–49, 100–102 and 144–147). For SpUreG, the residues in the conserved CPH motif [38] and the most conserved residues on the same side of the protein determined using the server ConSurf [52–54] were used to guide the docking (residues 69–72, 74–76, 78, and 111–113). In the first HADDOCK docking round, a rigid body energy minimization was carried out, and 1000 structures were calculated. The 200 best solutions selected based on the intermolecular energy were used for the semi-flexible, simulated annealing followed by an explicit water refinement. The solutions were clustered using a cut-off of 7.5 Å RMSD based on the pair wise backbone RMSD matrix. The best complex was selected on the basis of the HADDOCK score among those in the most populated cluster of structures (Fig. 3SI).

2.6.4. Modeling of SpUreE C-terminal tails The six C-terminal residues of SpUreE, not observed in the crystal structure, were built onto the calculated structure of the complex between SpUreE and SpUreG using Modeller. The best of 100 models generated by imposing symmetry restraints on the protein chains was identified on the basis of lowest score of the Modeller DOPE score. The conformation of the C-terminal tail was then optimized using 1000 Modeller loop optimization runs. The best model was chosen on the basis of a clustering analysis performed using the g_cluster module of Gromacs 4.6 package [55] and using the Gromos algorithm [56]. A 1.0 Å cut-off for the root mean square deviation of the Cα was used to include structures in the same cluster.

2.6.5. Electrostatic properties of the protein surfaces In order to analyze the electrostatic properties of the calculated docked complex, the protonation state of Cε and Cδ of all histidine residues was determined according to the optimization of the hydrogen bonding networks, while the N- and C-terminal residues were charged. The electrostatic color-coding was generated using DelPhi [57]. The program solves the linearized Poisson-Boltzmann equation to obtain the electrostatic potential in and around the protein, while taking the presence of solvent into account as a high dielectric continuum. The protein internal dielectric constant was set to 4 in all calculations, and the solvent dielectric constant was 80. The molecular (solvent-excluded) surface and the electrostatic potentials generated by DelPhi, as well as all the molecular graphics shown in this work were displayed using UCSF Chimera [58].

1666

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

3. Results and discussion The study of the detailed structural properties of protein–protein complexes must rely on X-ray diffraction crystallography, nuclear magnetic resonance spectroscopy, or computational studies [59]. While computational approaches to elucidate the interaction between UreE and UreG proteins have been previously reported [30], and attempts to co-crystallize UreE and UreG are currently in progress in our laboratory, we undertook an investigation of their solution structural properties using NMR. Initial trials focused on the proteins from H. pylori were unsuccessful: the deuterated form of HpUreE, necessary to improve the spectral resolution for this large 35-kDa protein and its complexes, was unstable, while UreG proteins are intrinsically disordered in solution independently of the biological source [38,60,61], with NMR spectra characterized by large signal broadening and overlap [38,62–65], preventing an extended assignment of their NMR spectra. In particular, HpUreG is the most folded UreG so far investigated [63], but it was unstable in solution over the course of the several days necessary to acquire the spectra for backbone NMR signal assignment. Therefore, we turned our attention onto UreE and UreG from S. pasteurii, known to be much more resilient to denaturation and aggregation. 3.1. SpUreE–SpUreG interaction thermodynamics by ITC The binding stoichiometry and thermodynamic parameters for this protein–protein interaction were investigated using isothermal calorimetric measurements carried out by titrating a concentrated SpUreE solution into a diluted SpUreG solution. The occurrence of a binding event was revealed by the presence of exothermic peaks that followed each addition (Fig. 2A). Integration of the peak areas revealed a binding isotherm (Fig. 2B) that was initially fit considering a direct titration of SpUreE dimer into SpUreG monomer and a single set of identical binding sites. The obtained fitting parameters [χ2v = 6765, n = 0.37 ± 0.01, Kb = (8.3 ± 0.5) × 105 M−1 (Kd = 1.20 ± 0.07 μM), ΔH = − 5.75 ± 0.05 kcal mol−1 and ΔS = +7.80 cal mol−1 K−1] were statistically consistent but showed an unrealistic stoichiometry. An alternative model involving two sets of independent binding sites, even though resulting in a better fit of the data (χ2v = 3050) also showed unrealistic stoichiometries, as well as an unacceptable error in one of the affinity constants, suggesting an over-parameterization of the fitting procedure [n1 = 0.30 ± 0.03, Kb1 = (7.7 ± 0.4) × 105 M− 1 (Kd1 = 1.30 ± 0.07 μM), ΔH1 = − 5.49 ± 0.06 kcal mol−1 and ΔS1 = + 8.52 cal mol− 1 K− 1; n2 = 0.08 ± 0.01, Kb2 = (1.8 ± 2.7) × 108 M− 1 (Kd2 = 6 ± 9 nM), ΔH2 = − 5.48 ± 0.05 kcal mol−1 and ΔS2 = + 19.4 cal mol−1 K− 1]. The fitting model that considers the sequential binding of SpUreE dimers to SpUreG monomers was not applicable in this case, because of the half-integer stoichiometry evident from the binding isotherm. An alternative fitting approach involved a reverse titration model that considers the titration of SpUreG monomers onto SpUreE dimers. The binding scheme involving a single set of sites gave, as before, an unreliable stoichiometry, considering the homo-dimeric nature of SpUreE and the dimerization equilibrium involving SpUreG [χ2v = 6765, n = 2.73 ± 0.02, Kb = (3.0 ± 0.5) × 105 M−1 (Kd = 3.3 ± 0.5 μM), ΔH = − 2.10 ± 0.02 kcal mol−1 and ΔS = + 18.0 cal mol− 1 K−1]. Another model that considers two independent sets of binding sites did not go to convergence and was therefore discarded. Finally, the model involving a sequential binding scheme gave realistic and statistically acceptable fitting parameters [χ2v = 5612, n1 = 1.0, Kb1 = (2.4 ± 0.5) × 104 M−1 (Kd1 = 42 ± 9 μM), ΔH1 = + 2.7 ± 0.9 kcal mol− 1 and ΔS1 = + 29.2 cal mol− 1 K − 1, n2 = 1.0, Kb2 = (6 ± 1) × 105 M − 1 (Kd2 = 1.7 ± 0.3 μM), ΔH2 = − 8.5 ± 0.9 kcal mol− 1 and ΔS2 = − 2.18 cal mol− 1 K− 1] and was thus accepted. This model entails two SpUreG monomers, known to be present in solution from light-scattering experiments (unpublished data from our laboratory) sequentially interacting with a SpUreE dimer with positive cooperativity, and could be explained by considering

Fig. 2. Analysis of SpUreE binding to SpUreG performed using isothermal titration calorimetry. (A) Representative plot of raw titration data of SpUreE dimer (0.15 μM) onto SpUreG monomer (0.05 μM). (B) Integrated heat data as a function of SpUreE/SpUreG molar ratio. The solid line represents the best fit obtained using a model that involves the sequential binding of two monomers of SpUreG onto one SpUreE dimer.

that binding of the second SpUreG monomer onto SpUreE would induce SpUreG dimerization, thus contributing to the binding affinity of the second SpUreG monomer. Similarly, in the case of the UreE– UreG interaction for the proteins from H. pylori, a single event of binding of two monomers of UreG onto the UreE dimer was observed with Kd = 4 μM [30]. In order to derive structural information for this complex, it is necessary to determine the residues on SpUreE involved in the interaction with SpUreG and vice versa. Initial attempts carried out by NMR indicated that the 1H-15N HSQC spectrum of a 15N-labeled SpUreG sample obtained in the presence of equimolar amounts of unlabeled 14 N-SpUreE maintained the characteristics of a disordered protein, analogously to what was observed for the same proteins from H. pylori [30], precluding this approach. The alternative tactic of studying 15N-labeled SpUreE in the presence of unlabeled 14 N-SpUreG was instead achievable provided that no metal ions [Ni(II) or Zn(II)] were present in solution because this condition lead to protein precipitation, likely due to the high protein concentration in the NMR samples. The NMR spectra of SpUreE have been previously studied, and the backbone amide NH signals have

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

been assigned at pH 6.5, 0.5 M NaCl, and 35 °C [33]. However, SpUreG is less soluble and more prone to aggregation at pH lower than 7.0, prompting us to carry out the backbone assignment of SpUreE at the same pH 7.0 utilized for the determination of the binding parameters for the SpUreEG complex in solution by ITC, and at ionic strength similar to physiologic conditions (150 mM NaCl). It has been previously reported [66] that SpUreE slowly precipitates at 35 °C at NaCl concentrations lower than 0.5 M, preventing an extensive NMR study at high temperature and low ionic strength and prompting us to use a temperature of 25 °C, as in the case of ITC. 3.2. Assignment of the NMR spectra of SpUreE Heteronuclear 2D and 3D NMR spectra of SpUreE were recorded and analyzed, and backbone assignments were obtained using the scalar connectivities provided by the triple resonance experiments. As previously reported [33], deuterium random labeling of the protein and TROSY-based pulse sequences were necessary to improve the resolution and sensitivity of the NMR spectra for this 34.8 kDa (147 residues per monomer) homo-dimeric protein [67]. The assigned 2D 1H-15N TROSY-HSQC spectrum and assignments of the amide resonances are shown in Fig. 3. The large spectral dispersion of proton signals unambiguously indicates that the protein is generally well folded, while the

1667

number of cross-peaks is consistent with the number of residues in one subunit, supporting a symmetrical average orientation of the two monomers. The identification of amide 1H and 15N peaks was obtained for 125 out of the expected 144 residues (not counting M1, P83 and P139) following a standard sequential assignment procedure. Nearly complete resonance assignments were achieved for the other backbone nuclei: 85.0% for 13Cα, 84.3% for 13Cβ, and 77.6% for 13CO. The 1HN, 15N, 13 Cα, 13Cβ, and 13CO chemical shifts have been deposited in the BioMagResBank (http://www.bmrb.wisc.edu) under accession number 19820. A comparison with the previous assignment carried out at 35 °C in a different buffer at pH 6.5 (BMRB code 5484) reveals an overall good agreement, except for the first few residues at the N-terminus, probably as a consequence of the different conditions of the protein solution. The 19 unassigned residues in the TROSY-HSQC spectrum (E28, D29, N31, N51, G58, D65, C103, I110, R113, K116, V125, R135-R143) were not observable in the spectrum at pH 7.0. To understand whether this was due to fast exchange with the solvent, intrinsic exchange rates were calculated [68] using the program Sphere (http://www.fccc.edu/research/ labs/roder/sphere/sphere.html) applied to the crystal structure of dimeric SpUreE (PDB ID: 1EAR). The data indicate that residues N31, N51, G58, D65, C103, R113, K116, R135, F136, K137, K141, Y142 and R143 feature exchange rates ≳ 12 s−1, consistently with their position in exposed loops connecting strands in the N-terminal domains, or in

Fig. 3. 900 MHz 1H-15N TROSY-HSQC spectrum of SpUreE in 90% H2O, 10% D2O, pH 7.0, 25 °C. Assigned cross-peaks are labeled with one-letter amino acid type and sequence number.

1668

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

the C-terminal region. On the other hand, amide signals for residues E28, D29, I110, V125, E138 and F140 appear not to be involved in exchange with the solvent, suggesting that they are broadened beyond detection because of conformational exchange phenomena involving mobility with rates in the 102–103 s−1 range. These rates are intermediate between those related to chemical or solvent exchange phenomena

(1–103 s−1) and those due to conformational changes involving small protein segments or slow loop reorientations (106–109 s−1) or local motions involving few atoms (bond librations, 109–1012 s−1), and are typical for large inter-domain motions. Indeed, all these residues, with the exception of V125 located at the end of helix α4 in the C-terminal domain and possibly subject to helix fluctuations, are located at the

Fig. 4. Computed structural disorder for SpUreE. (A) Charge-hydropathy plot for SpUreE calculated using PONDR; data for ordered and disordered proteins are blue and red, respectively, while the position of SpUreE is shown as a green dot; (B) Predicted regions of order and disorder using PONDR VL-XT; sections with scores higher than 0.5, shown in red, are predicted to be disordered; (C) Surface of SpUreE (PDB ID: 1EAR) showing in red the regions predicted as disordered by PONDR VL-XT.

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

interface between the N- and C- terminal domains, supporting the existence of inter-domain flexibility in UreE proteins, as suggested by X-ray crystallography [23] and molecular modeling studies [28]. Increasing the temperature from 25 °C to 35 °C did not lead to the appearance of additional signals, and the lower temperature was thus chosen to avoid sample degradation. 3.3. Secondary structure propensity of SpUreE from NMR chemical shifts Considering charge and hydrophobicity in the Uversky plot [69] obtained using the disorder predictor PONDR [70] (http://www.pondr. com), SpUreE is found to lie close to the boundary between proteins predicted to be folded and unfolded (Fig. 4A). In particular, PONDR VL-XT predicts the presence of three disordered regions covering 43 residues in the segments 25–49, 96–100, and 127–139 (Fig. 4B). The first two regions belong to structural fragments (Fig. 4C) that were predicted to be involved in forming a complex between UreE and UreG [30], while the third segment covers a long final stretch of the protein that is inserted between the N- and C-terminal domains, possibly modulating the

1669

inter-domain flexibility of the protein. This picture of protein disorder fits well with the idea that protein–protein interactions depend on mobile regions at the molecular interface, which is also consistent with molecular dynamics calculations on UreG that indicated the presence of disorder in regions predicted to be interacting with UreE [61], and proven by crystallography to be involved in the interaction between UreG and the UreDF complex in H. pylori [22]. This is consistent with the phenomenon of moonlighting, by means of which a given protein fulfills more than one function by interacting with different partners while adopting different conformations depending on the metabolic process, is well known for intrinsically disordered proteins [71]. In order to validate the disorder prediction, the experimental NMR chemical shifts were used to determine the residue-specific secondary structure propensity for SpUreE in solution using the programs SSP [72] (Fig. 5A) and TALOS + [73] (Fig. 5B). The secondary structure elements estimated using these calculations are in good agreement with those determined by X-ray crystallography in the solid state. In addition, the regions predicted as disordered by PONDR indeed show small or negative SSP scores, suggesting the presence of mobility in these protein

Fig. 5. Secondary structure propensity by NMR. SSP (A) and TALOS + (B) scores calculated using the Cα, Cβ and Hα chemical shifts for SpUreE. A positive score indicates a propensity for αstructure, while a negative score indicates a propensity for β-structure or extended loops, with residues predicted in fully formed α-helices and β-strands given scores of +1 and −1, respectively. The α-helices and β-strands established by UCSF-Chimera [58] (PDB ID: 1EAR) are shown as red and blue bars, respectively.

1670

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

segments. These regions were thus considered as good candidates for protein–protein interactions between SpUreE and SpUreG, considering that conformational equilibria pre-existing in the unbound state of proteins are known to sample the most suitable conformers to optimize binding to the partner and to stabilize complex formation [74,75]. In order to support this hypothesis with direct observations, and to map the residues involved in this interaction, an analysis of the perturbations of the amide 1H and 15N NMR chemical shifts upon protein–protein complex formation was carried out. 3.4. Chemical shift perturbation (CSP) analysis of the SpUreE–SpUreG interaction 1 H-15N TROSY HSQC experiments were acquired for 2H/15N-labeled SpUreE alone and in the presence of an excess of unlabeled SpUreG, at a

stoichiometric ratio of 1 dimer SpUreE : 4 monomer SpUreG. The data were analyzed with respect to changes in chemical shift (Δδ) (Fig. 6A). Addition of SpUreG gave rise to small CSPs, with maximum observed shifts of ~0.04 ppm for 1H and ~0.11 ppm for 15N, translating in ~40 Hz for 1H and ~10 Hz for 15N at 950 MHz proton Larmor frequency. Assuming a diffusion-controlled binding to a sterically available site, which has a kon typically around 109 M−1 s−1, the relationship koff ~ 109 · Kd can be derived [43]. The latter, together with the value of the overall dissociation constant measured by ITC (Kd = 70 μM) provides koff ~ 7 · 104 s−1. This value is consistent with the presence of fast exchange between the free and bound SpUreE on the NMR time scale, and the consequent absence of resolved chemical environments for the free and bound forms of SpUreE. The small CSP values prevented a statistically significant analysis of the data towards the discernment of an allosteric effect for the first and second UreG monomer binding to UreE, as observed by calorimetry.

Fig. 6. NMR-based detection of the SpUreE–SpUreG interaction site. (A) Chemical shift perturbations (CSP) measured for all SpUreE assigned residues. (B) Surface representation of the structure of SpUreE, highlighting regions experiencing CSP between 0.01 and 0.02 mapped in orange and larger than 0.02 colored in red, according to the data in panel A; top and bottom panels are rotated by 90° along the horizontal axis.

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

The amino acids showing the largest perturbations of chemical shift belong to the unstructured regions L2–S16 and G144–H147, while some smaller changes are observed throughout the sequence. Similarly small CSPs are not unprecedented [76–79], and have been ascribed to either weak complexes involving a well-defined patch, or to multiple binding modes [43]. The surface map of the residues affected by CSP, shown in Fig. 6B, indeed reveals a fairly consistent distribution on the protein surface, a result that can be interpreted as indicating that the small CSPs observed are due to the formation of a complex involving a single protein–protein orientation or multiple complexes with small structural deviations from each other. The protein patch identified by NMR as showing the largest and most consistent alterations of the backbone amide chemical shifts is consistent with the surface region on UreE suggested by molecular modeling to be involved in the formation of the complex with UreG in H. pylori [30]. Another region of perturbed residues (E88-M102) constitutes the interface between the two monomers at the core of the SpUreE dimer, and is thus hidden from direct contact with the partner protein. Its perturbation, however, could be induced by SpUreG binding to the SpUreE homo-dimer, leading to a small rearrangement of the two monomers by perturbing the hydrophobic interactions between the α-helices through an allosteric mechanism. The other region shown to be perturbed by the interaction with UreG covers residues G144–H147 and represents the disordered C-terminus of SpUreE that, along with H100, has been shown to be involved in Ni(II) coordination. Their relatively large CSP values suggest that they are also involved in the interaction of UreE with UreG. Interestingly, mutation of the C-terminal histidine (H152) in HpUreE, involved in Ni(II) coordination and corresponding to either H145 or H147 in SpUreE, abolished the high affinity metal binding site built by the UreEG complex [30], supporting the idea that the position of the C-terminal arm of UreE is influenced by its interaction with UreG. Even though the residues preceding the last portion of the protein are not assigned because unobserved (R135–R143), protein docking [30] predicted that these residues assist in binding UreG in the case of the proteins from H. pylori. Supporting this idea, the cross-peak corresponding to R134 experiences a large intensity loss upon addition of UreG, suggesting that the whole C-terminal tail (R135–H147) is involved in the formation of this protein–protein complex. The residues L2–S16 and R56–L61, also undergoing small but significant CSPs, are positioned on the opposite side of the proposed SpUreE–SpUreG interaction surface, which suggests a long-range allosteric structural rearrangement. If the perturbations were the result of direct interaction, a more consistent set of CSPs modifications in this region would be expected, considering the size of the UreG dimer. A larger sequence conservation is observed in the proposed interaction site as compared to this latter loop [28], further supporting the hypothesis that the UreE– UreG protein complex occurs at the K32–R49, H100–M102, and G144–H147 interface. A comparison of Figs. 4B and 6C reveals that the central portion of the C-terminal domain of SpUreE, known to bind Ni(II) and Zn(II), is both affected by predicted disorder and involved in the NMR-detected interaction with SpUreG, consistent with the idea, expressed above, that disordered regions are likely involved in protein–protein interactions. In order to provide structural support to the NMR-based observations, a protein–protein docking protocol was devised. 3.5. NMR-driven SpUreEG docking complex A model structure of the SpUreEG complex was calculated on the basis of the NMR-based findings and using state-of-the-art docking techniques implemented in the program HADDOCK 2.1 [50,51]. This software not only offers the possibility to guide the docking calculation by using experimentally derived constraints, but also includes a final refinement step with molecular dynamics in explicit water, which allows for flexibility of the residues at the interface of the interacting proteins. The model complex involved the latest published structure of SpUreE

1671

[24] as well as a model of SpUreG calculated based on its high degree of homology (sequence identity = 58%) with HpUreG, whose crystal structure in complex with HpUreF and HpUreD has been recently published [22]. Despite the fact that UreG model structures previously published by our group were based on low homology templates [38,62,63,65], the root-mean-square-deviation of the Cα atoms between the experimental and these model structures of dimeric HpUreG is small [80]. The modeling and docking protocol involved three steps: i) modeling of SpUreG structure, ii) docking between SpUreG model structure and the SpUreE crystal structure, and iii) modeling of the C-terminal portions of SpUreE not observed in the crystal structure because of disorder. Among the 200 docked complexes, grouped as a function of their internal root mean square deviation, the best was selected as belonging to the most populated cluster, which also features the lowest binding energy (Fig. 3-SI). Fig. 7A reports the structure of the calculated complex between SpUreE and SpUreG. The general features of the complex are similar to those observed in the previously reported complex for the proteins from H. pylori, which was calculated using only information on the CPH conserved motif of UreG and on the metal binding histidine residues at the UreE dimer interface [30]. The main difference with respect to the H. pylori model complex is an anticlockwise rotation of the SpUreE dimer around the vertical axis of about 35° (Fig. 4-SI). This rotation could explain why the N-terminal peripheral domains of UreE, apparently affected by disorder (see Fig. 6C), are less affected by chemical shift perturbations than residues in the C-terminal domain of UreE (see Fig. 4B). The interaction surface between SpUreE and SpUreG is ca. 1420 Å2, accounting for about 10.5% and 8.7% of the surface of SpUreE and SpUreG, respectively) (Fig. 7A). Considering that the observed area of the interaction surface in protein complexes ranges from ca. 600 to ca. 4800 Å2 (representing the 6%–24% of the accessible surface area of the individual monomers, with an average of about 12%) [81,82], the SpUreEG complex appears to be reasonably stable. The SpUreE residues in direct contact with SpUreG (Fig. 7B) are in good agreement with the experimental data, witnessing the correct outcome of the docking simulation. The modeled flexible C-terminal pendant arms of SpUreE containing H145 and H147, known to be involved in metal binding [24], are located in the clefts formed upon interaction of SpUreE with SpUreG, and are in contact with SpUreG, in agreement with NMR chemical shift perturbation data. The SpUreE surface in the metal binding region is mainly hydrophobic, with some positive spots in correspondence of residues R33, K48, K120, and R143. The electrostatic potential mapped on the protein surface of SpUreG shows that the interaction region is mainly negatively charged (Fig. 7B). Therefore, the presence of a divalent cation bound to SpUreE H100 could efficiently change the electrostatic properties of the protein surface, allowing a more favorable interaction with the negatively charged surface of SpUreG, consistently with the stabilization of the UreEG complex in H. pylori upon metal binding [30]. 4. Conclusions A significant number of studies have indicated that the interaction between UreE and UreG is an essential step in Ni(II) trafficking and delivery into the active site of urease. In the present study the structural basis of this interaction was analyzed, providing information on the overall protein interaction network built toward urease activation. The data presented here are consistent with two monomers of SpUreG cooperatively binding at the SpUreE interface forming a heterodimer of dimers (UreE)2–(UreG)2, previously observed for the same proteins from H. pylori [30]. The regions involved in protein–protein interaction, mapped by chemical shift perturbation (CSP) analysis, appear to experience significant mobility in the free state of the protein using disorder prediction algorithms. A similar protein flexibility has been detected, on the basis of molecular dynamics calculations, for the regions of UreG predicted to be involved in the interaction with UreE [61],

1672

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

Fig. 7. Ribbon diagram (A, left panel) and solvent excluded surface (A, right panel) of the model structure of the SpUreEG complex (SpUreE, dark and light blue; SpUreG, red and orange). Ribbons are colored in order to show the monomers composing the complex. Residues found at the interface of the complex and known to be involved in metal binding (SpUreE H100 and SpUreG C68 and H70), as well as the positions of two GDP molecules, are shown as ball and stick models and the atoms are colored accordingly to the atom type. (B) Solvent excluded surfaces of two components of the SpUreEG complex orientated in order to expose the interaction surfaces. In the left panel, the surface is colored according to the distance between the docked proteins: gray, N10 Å; red, 5–10 Å; yellow, 2.5–5 Å; and green, b2.5 Å. In the right panel, the surface is colored according to the surface electrostatic potential.

confirming a role of disorder for molecular recognition in the urease interaction network. The most disordered regions in UreE is the Cterminal portion, containing two histidine residues, H145 and H147, responsible, together with H100, for metal ion binding [24]. The residues covering this part of the protein feature relatively larger CSPs upon addition of SpUreG, showing that their chemical environment is more directly affected by protein–protein contacts. This observation, together with the fact that the SpUreE surface involved in the interaction with

UreG contains H100, responsible for metal binding, suggests that the interaction between the two proteins directly dictates the metal binding properties of the complex and is thus essential for metal ion delivery. The present study implies therefore that the UreE–UreG interaction is important not only for facilitating GTP hydrolysis by UreG, as previously observed [21], but also for directly aiding metal ion delivery by UreE into the urease activation complex. The design of new molecules inhibiting this protein–protein interaction that leads to urease

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

activation will benefit from the knowledge of the structural and biophysical details provided in the present study. Acknowledgments Fabio Calogiuri is thanked for acquiring NMR data at the Center for Magnetic Resonance (Sesto Fiorentino, Italy). A.M. was supported by a Ph.D. fellowship by the University of Bologna; O.D. and M.B. were supported by a fellowship financed by Specialty Fertilizer Products (Leawood, KS, USA). F.M. was funded by CIRMMP (Consorzio Interuniversitario di Risonanze Magnetiche di Metallo-Proteine). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbapap.2014.06.016. References [1] M.J. Maroney, S. Ciurli, Nonredox nickel enzymes, Chem. Rev. 114 (2014) 4206–4228. [2] H.L.T. Mobley, R.P. Hausinger, Microbial urease: significance, regulation and molecular characterization, Microbiol. Rev. 53 (1989) 85–108. [3] P.A. Karplus, M.A. Pearson, R.P. Hausinger, 70 years of crystalline urease: what have we learned? Acc. Chem. Res. 30 (1997) 330–337. [4] S. Ciurli, S. Benini, W.R. Rypniewski, K.S. Wilson, S. Miletti, S. Mangani, Structural properties of the nickel ions in urease: novel insights into the catalytic and inhibition mechanisms, Coord. Chem. Rev. 190–192 (1999) 331–355. [5] R.A. Burne, Y.Y. Chen, Bacterial ureases in infectious diseases, Microbes Infect. 2 (2000) 533–542. [6] S. Ciurli, Urease. Recent insights in the role of nickel, Nickel and its surprising impact in nature, vol. 2, John Wiley & Sons, Ltd, Chichester, UK, 2007, pp. 241–278. [7] B. Krajewska, I. Ureases, Functional, catalytic and kinetic properties: a review, J. Mol. Catal. B Enzym. 59 (2009) 9–21. [8] B. Zambelli, F. Musiani, S. Benini, S. Ciurli, Chemistry of Ni2+ in urease: sensing, trafficking, and catalysis, Acc. Chem. Res. 44 (2011) 520–530. [9] S. Benini, P. Kosikowska, M. Cianci, L. Mazzei, A.G. Vara, Ł. Berlicki, S. Ciurli, The crystal structure of Sporosarcina pasteurii urease in a complex with citrate provides new hints for inhibitor design, J. Biol. Inorg. Chem. 18 (2013) 391–399. [10] B.P. Callahan, Y. Yuan, R. Wolfenden, The burden borne by urease, J. Am. Chem. Soc. 127 (2005) 10828–10829. [11] B. Zambelli, S. Ciurli, Nickel and human health, Met. Ions Life Sci. 13 (2013) 321–357. [12] E. Jabri, M.B. Carr, R.P. Hausinger, P.A. Karplus, The crystal structure of urease from Klebsiella aerogenes, Science 268 (1995) 998–1004. [13] S. Benini, W.R. Rypniewski, K.S. Wilson, S. Miletti, S. Ciurli, S. Mangani, A new proposal for urease mechanism based on the crystal structures of the native and inhibited enzyme from Bacillus pasteurii: why urea hydrolysis costs two nickels, Structure 7 (1999) 205–216. [14] N.-C. Ha, S.-T. Oh, J.Y. Sung, K.A. Cha, M.H. Lee, B.-H. Oh, Supramolecular assembly and acid resistance of Helicobacter pylori urease, Nat. Struct. Biol. 8 (2001) 505–509. [15] A. Balasubramanian, K. Ponnuraj, Crystal structure of the first plant urease from jack bean: 83 years of journey from its first crystal to molecular structure, J. Mol. Biol. 400 (2010) 274–283. [16] A. Balasubramanian, V. Durairajpandian, S. Elumalai, N. Mathivanan, A.K. Munirajan, K. Ponnuraj, Structural and functional studies on urease from pigeon pea (Cajanus cajan), Int. J. Biol. Macromol. 58 (2013) 301–309. [17] M.A. Farrugia, L. Macomber, R.P. Hausinger, Biosynthesis of the urease metallocenter, J. Biol. Chem. 288 (2013) 13178–13185. [18] M.H. Lee, S.B. Mulrooney, M.J. Renner, Y. Markowicz, R.P. Hausinger, Klebsiella aerogenes urease gene cluster: sequence of ureD and demonstration that four accessory genes (ureD, ureE, ureF, ureG) are involved in nickel metallocenter biosynthesis, J. Bacteriol. 174 (1992) 4324–4330. [19] I.S. Park, R.P. Hausinger, Requirement of carbon dioxide for in vitro assembly of the urease nickel metallocenter, Science 267 (1995) 1156–1158. [20] I.-S. Park, R.P. Hausinger, Metal ion interactions with urease and UreD-urease apoproteins, Biochemistry 35 (1996) 5345–5352. [21] A. Soriano, G.J. Colpas, R.P. Hausinger, UreE stimulation of GTP-dependent urease activation in the UreD-UreF-UreG-urease apoprotein complex, Biochemistry 39 (2000) 12435–12440. [22] Y.H. Fong, H.C. Wong, M.H. Yuen, P.H. Lau, Y.W. Chen, K.-B. Wong, Structure of UreG/ UreF/UreH complex reveals how urease accessory proteins facilitate maturation of Helicobacter pylori urease, PLoS Biol. 11 (2013) e1001678. [23] H. Remaut, N. Safarov, S. Ciurli, J. Van Beeumen, Structural basis for Ni2+ transport and assembly of the urease active site by the metallochaperone UreE from Bacillus pasteurii, J. Biol. Chem. 276 (2001) 49365–49370. [24] B. Zambelli, K. Banaszak, A. Merloni, A. Kiliszek, W. Rypniewski, S. Ciurli, Selectivity of Ni(II) and Zn(II) binding to Sporosarcina pasteurii UreE, a metallochaperone in the urease assembly: a calorimetric and crystallographic study, J. Biol. Inorg. Chem. 18 (2013) 1005–1017.

1673

[25] H.-K. Song, S.B. Mulrooney, R. Huber, R.P. Hausinger, Crystal structure of Klebsiella aerogenes UreE, a nickel-binding metallochaperone for urease activation, J. Biol. Chem. 276 (2001) 49359–49364. [26] R. Shi, C. Munger, A. Asinas, S.L. Benoit, E. Miller, A. Matte, R.J. Maier, M. Cygler, Crystal structures of spo and metal-bound forms of the UreE protein from Helicobacter pylori: role of multiple metal binding sites, Biochemistry 49 (2010) 7080–7088. [27] K. Banaszak, V. Martin-Diaconescu, M. Bellucci, B. Zambelli, W.R. Rypniewski, M.J. Maroney, S. Ciurli, Crystallographic and X-ray absorption spectroscopic characterization of Helicobacter pylori UreE bound to Ni2+ and Zn2+ reveals a role for the disordered C-terminal arm in metal trafficking, Biochem. J. 441 (2012) 1017–1026. [28] F. Musiani, B. Zambelli, M. Stola, S. Ciurli, Nickel trafficking: insights into the fold and function of UreE, a urease metallochaperone, J. Inorg. Biochem. 98 (2004) 803–813. [29] S. Benoit, R.J. Maier, Dependence of Helicobacter pylori urease activity on the nickelsequestering ability of the UreE accessory protein, J. Bacteriol. 185 (2003) 4787–4795. [30] M. Bellucci, B. Zambelli, F. Musiani, P. Turano, S. Ciurli, Helicobacter pylori UreE, a urease accessory protein: specific Ni2+ and Zn2+ binding properties and interaction with its cognate UreG, Biochem. J. 422 (2009) 91–100. [31] M. Stola, F. Musiani, S. Mangani, P. Turano, N. Safarov, B. Zambelli, S. Ciurli, The nickel site of Bacillus pasteurii UreE, a urease metallo-chaperone, as revealed by metalbinding studies and X-ray absorption spectroscopy, Biochemistry 45 (2006) 6495–6509. [32] M.Y. Lee, H.S. Pankratz, S. Wang, R.A. Scott, M.G. Finnegan, M.K. Johnson, J.A. Ippolito, D.W. Christianson, R.P. Hausinger, Purification and characterization of Klebsiella aerogenes UreE protein: a nickel binding protein that functions in urease metallocenter assembly, Protein Sci. 2 (1993) 1042–1052. [33] Y.-H. Lee, H.-S. Won, H.-C. Ahn, S. Park, H. Yagi, H. Akutsu, B.-J. Lee, Backbone NMR assignments of the metal-free UreE from Bacillus pasteurii, J. Biomol. NMR 24 (2002) 361–362. [34] H.S. Won, Y.H. Lee, J.H. Kim, I.S. Shin, M.H. Lee, B.J. Lee, Structural characterization of the nickel-binding properties of Bacillus pasteurii urease accessory protein (Ure)E in solution, J. Biol. Chem. 279 (2004) 17466–17472. [35] P. Voland, D.L. Weeks, E.A. Marcus, C. Prinz, G. Sachs, D. Scott, Interactions among the seven Helicobacter pylori proteins encoded by the urease gene cluster, Am. J. Physiol. Gastrointest. Liver Physiol. 284 (2003) G96–G106. [36] F. Biagi, F. Musiani, S. Ciurli, Structure of the UreD-UreF-UreG-UreE complex in Helicobacter pylori: a model study, J. Biol. Inorg. Chem. 18 (2013) 571–577. [37] A. Sivashanmugam, V. Murray, C. Cui, Y. Zhang, J. Wang, Q. Li, Practical protocols for production of very high yields of recombinant proteins using Escherichia coli, Protein Sci. 18 (2009) 936–948. [38] B. Zambelli, M. Stola, F. Musiani, K. De Vriendt, B. Samyn, B. Devreese, J. Van Beeumen, P. Turano, A. Dikiy, D.A. Bryant, S. Ciurli, UreG, a chaperone in the urease assembly process, is an intrinsically unstructured GTPase that specifically binds Zn2+, J. Biol. Chem. 280 (2005) 4684–4695. [39] S. Ciurli, N. Safarov, S. Miletti, A. Dikiy, S.K. Christensen, K. Kornetzky, D.A. Bryant, I. Vandenberghe, B. Devreese, B. Samyn, H. Remaut, J. van Beeumen, Molecular characterization of Bacillus pasteurii UreE, a metal-binding chaperone for the assembly of the urease active site, J. Biol. Inorg. Chem. 7 (2002) 623–631. [40] F. Delaglio, S. Grzesiek, G.W. Vuister, G. Zhu, J. Pfeifer, A. Bax, NMRPipe: a multidimensional spectral processing system based on UNIX pipes, J. Biomol. NMR 6 (1995) 277–293. [41] R.L.J. Keller, Optimizing the process of nuclear magnetic resonance spectrum analysis and computer aided resonance assignment, Swiss Federal Institute of Technology, Swiss Federal Institute of Technology of Zurich, Zurich, 2004. [42] R.A. Williamson, M.D. Carr, T.A. Frenkiel, J. Feeney, R.B. Freedman, Mapping the binding site for matrix metalloproteinase on the N-terminal domain of the tissue inhibitor of metalloproteinases-2 by NMR chemical shift perturbation, Biochemistry 36 (1997) 13882–13889. [43] M.P. Williamson, Using chemical shift perturbation to characterise ligand binding, Prog. Nucl. Magn. Reson. Spectrosc. 73 (2013) 1–16. [44] J. Pei, B.H. Kim, N.V. Grishin, PROMALS3D: a tool for multiple protein sequence and structure alignments, Nucleic Acids Res. 36 (2008) 2295–2300. [45] M.A. Marti-Renom, A.C. Stuart, A. Fiser, R. Sanchez, F. Melo, A. Sali, Comparative protein structure modeling of genes and genomes, Annu. Rev. Biophys. Biomol. Struct. 29 (2000) 291–325. [46] M.Y. Shen, A. Sali, Statistical potential for assessment and prediction of protein structures, Protein Sci. 15 (2006) 2507–2524. [47] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, PROCHECK: a program to check the stereochemical quality of protein structures, J. Appl. Crystallogr. 26 (1993) 283–291. [48] K. Suhre, Y.H. Sanejouand, ElNemo: a normal mode web server for protein movement analysis and the generation of templates for molecular replacement, Nucleic Acids Res. 32 (2004) W610–W614. [49] F. Agriesti, D. Roncarati, F. Musiani, C. Del Campo, M. Iurlaro, F. Sparla, S. Ciurli, A. Danielli, V. Scarlato, FeON-FeOFF: the Helicobacter pylori Fur regulator commutates iron-responsive transcription by discriminative readout of opposed DNA grooves, Nucleic Acids Res. 42 (2014) 3138–3151. [50] C. Dominguez, R. Boelens, A.M. Bonvin, HADDOCK: a protein–protein docking approach based on biochemical or biophysical information, J. Am. Chem. Soc. 125 (2003) 1731–1737. [51] S.J. de Vries, A.D. van Dijk, M. Krzeminski, M. van Dijk, A. Thureau, V. Hsu, T. Wassenaar, A.M. Bonvin, HADDOCK versus HADDOCK: new features and performance of HADDOCK2.0 on the CAPRI targets, Proteins 69 (2007) 726–733. [52] H. Ashkenazy, E. Erez, E. Martz, T. Pupko, N. Ben-Tal, ConSurf 2010: calculating evolutionary conservation in sequence and structure of proteins and nucleic acids, Nucleic Acids Res. 38 (2010) W529–W533.

1674

A. Merloni et al. / Biochimica et Biophysica Acta 1844 (2014) 1662–1674

[53] F. Glaser, T. Pupko, I. Paz, R.E. Bell, D. Bechor-Shental, E. Martz, N. Ben-Tal, ConSurf: identification of functional regions in proteins by surface-mapping of phylogenetic information, Bioinformatics 19 (2003) 163–164. [54] M. Landau, I. Mayrose, Y. Rosenberg, F. Glaser, E. Martz, T. Pupko, N. Ben-Tal, ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures, Nucleic Acids Res. 33 (2005) W299–W302. [55] D. Van Der Spoel, E. Lindahl, B. Hess, G. Groenhof, A.E. Mark, H.J. Berendsen, GROMACS: fast, flexible, and free, J. Comput. Chem. 26 (2005) 1701–1718. [56] X. Daura, K. Gademann, B. Jaun, D. Seebach, W.F. van Gunsteren, A.E. Mark, Peptide folding: when simulation meets experiment, Angew. Chem. Int. Ed. 38 (1999) 236–240. [57] B. Honig, A. Nicholls, Classical electrostatics in biology and chemistry, Science 268 (1995) 1144–1149. [58] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng, T.E. Ferrin, UCSF Chimera - A visualization system for exploratory research and analysis, J. Comput. Chem. 25 (2004) 1605–1612. [59] J. Janin, A.M. Bonvin, Protein–protein interactions, Curr. Opin. Struct. Biol. 23 (2013) 859–861. [60] P. Neyroz, B. Zambelli, S. Ciurli, Intrinsically disordered structure of Bacillus pasteurii UreG as revealed by steady-state and time-resolved fluorescence spectroscopy, Biochemistry 45 (2006) 8918–8930. [61] F. Musiani, E. Ippoliti, C. Micheletti, P. Carloni, S. Ciurli, Conformational fluctuations of UreG, an intrinsically disordered enzyme, Biochemistry 52 (2013) 2949–2954. [62] B. Zambelli, F. Musiani, M. Savini, P. Tucker, S. Ciurli, Biochemical studies on Mycobacterium tuberculosis UreG and comparative modeling reveal structural and functional conservation among the bacterial UreG family, Biochemistry 46 (2007) 3171–3182. [63] B. Zambelli, P. Turano, F. Musiani, P. Neyroz, S. Ciurli, Zn2+-linked dimerization of UreG from Helicobacter pylori, a chaperone involved in nickel trafficking and urease activation, Proteins 74 (2009) 222–239. [64] B. Zambelli, N. Cremades, P. Neyroz, P. Turano, V.N. Uversky, S. Ciurli, Insights in the (un)structural organization of Bacillus pasteurii UreG, an intrinsically disordered GTPase enzyme, Mol. Biosyst. 8 (2012) 220–228. [65] R. Real-Guerra, F. Staniscuaski, B. Zambelli, F. Musiani, S. Ciurli, C.R. Carlini, Biochemical and structural studies on native and recombinant Glycine max UreG: a detailed characterization of a plant urease accessory protein, Plant Mol. Biol. 78 (2012) 461–475. [66] Y.-H. Lee, H.-S. Won, M.-H. Lee, B.-J. Lee, Effects of salt and nickel ion on the conformational stability of Bacillus pasteurii UreE, FEBS Lett. 522 (2002) 135–140.

[67] K.H. Gardner, L.E. Kay, The use of 2H, 13C, 15N multidimensional NMR to study the structure and dynamics of proteins, Annu. Rev. Biophys. Biomol. Struct. 27 (1998) 357–406. [68] Y. Bai, J.S. Milne, L. Mayne, S.W. Englander, Primary structure effects on peptide group hydrogen exchange, Proteins 17 (1993) 75–86. [69] V.N. Uversky, J.R. Gillespie, A.L. Fink, Why are natively unfolded proteins unstructured under physiologic conditions? Proteins 41 (2000) 415–427. [70] P. Romero, Z. Obradovic, X. Li, E.C. Garner, C.J. Brown, A.K. Dunker, Sequence complexity of disordered protein, Proteins 42 (2001) 38–48. [71] P. Tompa, C. Szasz, L. Buday, Structural disorder throws new light on moonlighting, Trends Biochem. Sci. 30 (2005) 484–489. [72] J.A. Marsh, V.K. Singh, Z. Jia, J.D. Forman-Kay, Sensitivity of secondary structure propensities to sequence differences between alpha- and gamma-synuclein: implications for fibrillation, Protein Sci. 15 (2006) 2795–2804. [73] Y. Shen, F. Delaglio, G. Cornilescu, A. Bax, TALOS+: a hybrid method for predicting protein backbone torsion angles from NMR chemical shifts, J. Biomol. NMR 44 (2009) 213–223. [74] D. Tobi, I. Bahar, Structural changes involved in protein binding correlate with intrinsic motions of proteins in the unbound state, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 18908–18913. [75] D.D. Boehr, R. Nussinov, P.E. Wright, The role of dynamic conformational ensembles in biomolecular recognition, Nat. Chem. Biol. 5 (2009) 789–796. [76] P. Rajagopal, E.B. Waygood, J. Reizer, M.H. Saier Jr., R.E. Klevit, Demonstration of protein–protein interaction specificity by NMR chemical shift mapping, Protein Sci. 6 (1997) 2624–2627. [77] J.A. Worrall, W. Reinle, R. Bernhardt, M. Ubbink, Transient protein interactions studied by NMR spectroscopy: the case of cytochrome C and adrenodoxin, Biochemistry 42 (2003) 7068–7076. [78] J. Morrison, J.C. Yang, M. Stewart, D. Neuhaus, Solution NMR study of the interaction between NTF2 and nucleoporin FxFG repeats, J. Mol. Biol. 333 (2003) 587–603. [79] J.Y. Guan, J.M. Foerster, J.W. Drijfhout, M. Timmer, A. Blok, G.M. Ullmann, M. Ubbink, An ensemble of rapidly interconverting orientations in electrostatic protein-peptide complexes characterized by NMR spectroscopy, Chembiochem 15 (2014) 556–566. [80] B. Zambelli, A. Berardi, V. Martin-Diaconescu, L. Mazzei, F. Musiani, M.J. Maroney, S. Ciurli, Nickel binding properties of Helicobacter pylori UreF, an accessory protein in the nickel-based activation of urease, J. Biol. Inorg. Chem. 19 (2014) 319–334. [81] J. Janin, S. Miller, C. Chothia, Surface, subunit interfaces and interior of oligomeric proteins, J. Mol. Biol. 204 (1988) 155–164. [82] C. Kleanthous, Protein–protein recognition, Oxford University Press, New York, 2000.