|Home | About | Journals | Submit | Contact Us | Français|
YeaZ is involved in a protein network that is essential for bacteria. The crystal structure of YeaZ from Thermotoga maritima was determined to 2.5 Å resolution. Although this protein belongs to a family of ancient actin-like ATPases, it appears that it has lost the ability to bind ATP since it lacks some key structural features that are important for interaction with ATP. A conserved surface was identified, supporting its role in the formation of protein complexes.
yeaZ is an essential gene in many bacteria (Zhang & Lin, 2009 ), such as Escherichia coli (yeaZ), Salmonella typhimurium LT2 (yeaZ), Bacillus subtilis (ydiC), Streptococcus pneumoniae (spr0129), Pseudomonas aeruginosa (PA14_16710) and Francisella novicida (FTN_1148). A genome-wide study of the E. coli interaction network revealed that YeaZ forms a complex with YgjD (Butland et al., 2005 ). A recent study further demonstrated that E. coli YeaZ can interact with either YgjD or YjeE, with YgjD as the preferred partner, suggesting that YeaZ is part of a protein network that may be involved in DNA metabolism and cell division (Handford et al., 2009 ). YgjD is homologous to Kae1 (kinase-associated endopeptidase 1), a component of the yeast KEOPS/EKC complex (kinase, endopeptidase and other proteins of small size/endopeptidase-like and kinase associated to transcribed chromatin) that is necessary for telomere maintenance and transcription of essential eukaryotic genes (Downey et al., 2006 ; Kisseleva-Romanova et al., 2006 ). Kae1 and its homologs belong to the ASKHA (acetate and sugar kinase/Hsp70/actin) superfamily (Mao et al., 2008 ; Hecker et al., 2007 , 2008 ). Recent crystal structures of YeaZs from E. coli (EcYeaZ), S. typhimurium (StYeaZ; Jeudy et al., 2005 ; Nichols et al., 2006 ) and Thermotoga maritima (TmYeaZ; this study) indicate that YeaZ is structurally related to Kae1 and thus belongs to the same superfamily.
Here, we report the 2.5 Å resolution crystal structure of TmYeaZ from T. maritima (TM0874) in the light of current knowledge of the involvement of YeaZ in protein complexes, which was not available when the original StYeaZ structure was reported (Nichols et al., 2006 ). The structure of TmYeaZ was determined using the high-throughput pipeline of the Joint Center for Structural Genomics (JCSG; Lesley et al., 2002 ) as part of the National Institute of General Medical Sciences’ Protein Structure Initiative (PSI; http://www.nigms.nih.gov/Initiatives/PSI/). The tm0874 gene of T. maritima encodes a protein with a molecular weight of 22 986 Da (residues 1–206) and a calculated isoelectric point of 6.35.
The gene encoding TmYeaZ (GenBank AAD35955.1; gi:4981408; Swiss-Prot Q9WZX7) was amplified by polymerase chain reaction (PCR) from T. maritima genomic DNA using PfuTurbo (Stratagene) and primers corresponding to the predicted 5′ and 3′ ends (forward primer, 5′-ATGAACGTTCTGGCACTCG-3′; reverse primer, 5′-CTCTTAATTAAGTCGCGTTAGCCCCTTTTCTTTTTTTCCCAG-3′). The PCR product was cloned into plasmid pMH4 (developed at the JCSG), which encodes an expression and purification tag (MGSDKIHHHHHH) at the amino-terminus of the full-length protein. The cloning junctions were confirmed by DNA sequencing. Protein expression was performed in a selenomethionine-containing medium using E. coli strain GeneHogs (Invitrogen). Lysozyme was added to the culture at the end of fermentation to a final concentration of 250 µg ml−1 and the cells were harvested. After one freeze–thaw cycle, the cells were sonicated in lysis buffer [50 mM Tris–HCl pH 7.9, 50 mM NaCl, 10 mM imidazole, 1 mM tris(2-carboxyethyl)phosphine hydrochloride (TCEP)] and the lysate was clarified by centrifugation at 32 500g for 30 min. The soluble fraction was applied to nickel-chelating resin (GE Healthcare) pre-equilibrated with lysis buffer, the resin was washed with wash buffer [50 mM Tris–HCl pH 7.9, 300 mM NaCl, 40 mM imidazole, 10%(v/v) glycerol, 1 mM TCEP] and the protein was eluted with elution buffer [20 mM Tris–HCl pH 7.9, 300 mM imidazole, 10%(v/v) glycerol, 1 mM TCEP]. The eluate was diluted tenfold to 45 ml with buffer Q [20 mM Tris–HCl pH 7.9, 5%(v/v) glycerol, 1 mM TCEP] containing 50 mM NaCl and loaded onto a 6 ml Resource Q column (GE Healthcare) pre-equilibrated with the same buffer. A linear gradient of 50–500 mM NaCl in buffer Q was used to elute the protein and the appropriate fractions were pooled. The protein was concentrated to 1 ml by centrifugal ultrafiltration (Millipore) and diluted to 15 ml with crystallization buffer (20 mM Tris–HCl pH 7.9, 150 mM NaCl, 1 mM TCEP). This process was repeated two more times, resulting in a 3375-fold buffer exchange. The protein was then concentrated to 14 mg ml−1 for crystallization, with its concentration being determined using Coomassie Plus Protein Assay Reagent (Pierce). To determine its oligomeric state, we analyzed TmYeaZ using a 1 × 30 cm Superdex 200 column (GE Healthcare) coupled with miniDAWN static light-scattering and Optilab differential refractive-index detectors (Wyatt Technology). The mobile phase consisted of 20 mM Tris–HCl pH 7.9, 150 mM NaCl and 0.02%(w/v) sodium azide. The molar mass was calculated using ASTRA 5.1.5 software (Wyatt Technology). The protein was crystallized using the nanodroplet vapor-diffusion method (Santarsiero et al., 2002 ) with standard JCSG crystallization protocols (Lesley et al., 2002 ). Sitting drops consisting of 200 nl protein solution and 200 nl crystallization reagent above a 50 µl reservoir were used. Initial screening for diffraction was carried out using the Stanford Automated Mounting system (SAM; Cohen et al., 2002 ) at the Stanford Synchrotron Radiation Lightsource (SSRL; Menlo Park, California, USA). The crystal used for structure solution was obtained in 10%(v/v) 2-methyl-2,4-pentanediol (MPD) and 0.1 M citrate pH 4.0 at 277 K. A rectangular, plate-shaped crystal (~50 × 30 × 15 µm) was harvested after 10 d. For cryoprotection, additional MPD was added to the crystal, bringing the final concentration to 25%(v/v). Diffraction images were indexed in space group C222.
Multi-wavelength anomalous diffraction (MAD) data were collected at SSRL on beamline 11-1 at wavelengths corresponding to the high-energy remote (λ1) and inflection (λ2) of a selenium MAD experiment. The data sets were collected at 100 K using an ADSC Q315 detector. The MAD data were integrated and reduced using XDS and then scaled using the program XSCALE (Kabsch, 1993 ). Selenium sites were located with SHELXD (Sheldrick, 2008 ) and refined using autoSHARP (Bricogne et al., 2003 ). Phase refinement and automatic model building was performed with RESOLVE (Terwilliger, 2003 ). Model completion and refinement were performed with Coot (Emsley & Cowtan, 2004 ) and REFMAC (Winn et al., 2003 ). Loose NCS restraints for both main chains and side chains (positional and thermal weights of 5.0 and 10.0, respectively) were applied between the two monomers. Each monomer was defined as a TLS group. Experimental MAD phases in the form of Hendrickson–Lattman coefficients were used as restraints during refinement. CCP4 programs were used for data conversion and other calculations (Collaborative Computational Project, Number 4, 1994 ). Data-processing and refinement statistics are summarized in Table 1 .
The quality of the refined structure was analyzed using the JCSG Quality Control server, which verifies the stereochemical quality of the model using AutoDepInputTool (Yang et al., 2004 ), MolProbity (Lovell et al., 2003 ) and WHATIF 5.0 (Vriend, 1990 ), the agreement between the atomic model and the data using SFCHECK 4.0 (Vaguine et al., 1999 ) and RESOLVE (Terwilliger, 2003 ), the protein sequence using ClustalW (Thompson et al., 1994 ), the atomic occupancies using MOLEMAN2 (Kleywegt, 2000 ) and the consistency of NCS pairs. It also evaluates the difference in R cryst/R free, expected R free/R cryst and maximum/minimum B values by parsing the refinement log file and PDB header. Analysis of the crystal packing was performed using the PISA server (Krissinel & Henrick, 2007 ). The sequences used for Fig. 3 are the top hits from a BLAST search (Altschul et al., 1997 ) against the nonredundant protein-sequence database using TmYeaZ as a probe; only sequences with lengths between 180 and 250 residues were retained for the analysis (237 sequences). Multiple sequence alignments were performed using MAFFT (Katoh et al., 2005 ). Mapping of sequence conservation onto the protein was performed by CONSURF (Landau et al., 2005 ). Fig. 1(d) was generated using ESPript (Gouet et al., 2003 ) with secondary structures assigned by DSSP (Kabsch & Sander, 1983 ). Fig. 3(c) was generated using WEBLOGO (Crooks et al., 2004 ). All other figures were prepared with PyMOL (DeLano Scientific).
The selenomethionine derivative of full-length TmYeaZ with an N-terminal His tag was expressed in E. coli and purified by metal-affinity chromatography. The crystal structure of TmYeaZ was determined in space group C222 at 2.5 Å resolution using the MAD method. The final TmYeaZ model includes a dimer (residues 1–193 for chain A and 0–187 for chain B (residue 0 is the last residue of the His tag; the remainder of the His tag is disordered; Figs. 1 a and 1 b), two unknown ligands (UNL), which were modeled as discrete O atoms without geometry restraints, and 23 water molecules in the asymmetric unit. The two independent molecules in the asymmetric unit (A, B) are similar to each other, with a root-mean-square difference of 0.53 Å for 187 aligned Cα atoms. The C-termini (194–206 of chain A and 188–206 of chain B) were not modeled owing to a lack of interpretable electron density. The Matthews coefficient (V M; Matthews, 1968 ) for TmYeaZ is 3.25 Å3 Da−1 and the estimated solvent content is 61.9%. The Ramachandran plot produced by MolProbity shows that 95.5 and 99.5% of the residues are in favored and allowed regions, respectively. The two Ramachandran outliers (residues 115 and 151 of chain A) are located in regions of poor electron density. TmYeaZ is composed of nine β-strands (β1–β9), six α-helices (α1–α6) and four 310-helices (η1–η4). The total β-sheet, α-helical and 310-helical content is 31.6, 32.6 and 6.2%, respectively.
The molecular weight of TmYeaZ in solution was determined to be 52 720 Da by analytical size-exclusion chromatography in combination with static light scattering. As a monomer of His-tagged SeMet-TmYeaZ would have a calculated molecular weight of 24 404 Da, it is likely that TmYeaZ exists as a dimer in solution. Analysis of the crystal packing suggests two possible modes of dimerization. In the first possibility (Fig. 1 b), the two independent monomers in the asymmetric unit would form a dimer (AB dimer) with a buried surface of 1240 Å2 per monomer (12.7% of the monomer surface area). The β3 strands from the two N-terminal domains pack together in an antiparallel manner to form an extended β-sheet. In the alternate mode, in which two dimers (A 2 and B 2) are formed by the crystallographic twofold axis [Fig. 1 c, only the B 2 (B–B′) dimer is shown], the α1 and α2 helices of their N-terminal domains are packed into a four-helix bundle, which generates a V-shaped dimer. This dimer interface buries a surface area of 870 Å2 per monomer for the B 2 dimer (9% of the B monomer surface area) and 565 Å2 per monomer for the A 2 dimer (5.7% of the A monomer surface area). Although this second type of dimer interaction is weaker, it is interesting to note that a similar mode of dimerization is conserved in the crystal structures of both StYeaZ and EcYeaZ (Nichols et al., 2006 ).
TmYeaZ belongs to the actin-like ATPase superfamily, which typically contains a duplication of ribonuclease H-like domains (Andreeva et al., 2004 ). In TmYeaZ, the first ribonuclease H-like domain is composed from both the N-terminus (residues 1–100) and the C-terminus (residues 160–193) of the protein. The second domain (residues 101–159) can be superimposed onto the N-terminal portion of the first domain with an r.m.s.d. of 3.18 Å for 47 aligned Cα atoms; however, it lacks the segment corresponding to the α1–β4 region of the first domain. The top hits from DALI (Holm & Sander, 1995 ) identified two bacterial YeaZ proteins: EcYeaZ (PDB code 1okj; Z = 21.0, r.m.s.d. = 2.1 Å for 188 aligned Cα atoms, 22% sequence identity; C. Abergel, S. Jeudy & J. M. Claverie, unpublished work) and StYeaZ (PDB code 2gel; Z = 20.6, r.m.s.d. = 2.3 Å for 187 aligned Cα atoms, 21% sequence identity; Nichols et al., 2006 ). Since detailed structural comparisons of these YeaZ homologs and ASKHA proteins have been reported previously (Nichols et al., 2006 ), we only briefly summarize the new results related to TmYeaZ. A structure-based sequence alignment of YeaZs is shown in Fig. 1 (d). Compared with the other two YeaZs, TmYeaZ lacks the 20-residue αβ insertion between β9 and α5 in the second domain. Overall, the substantial structural similarities among these YeaZs suggest a common function. TmYeaZ also displays strong structural similarities to Kae1, especially in the first domain (PDB code 2ivp; r.m.s.d. = 2.1 Å for 110 aligned Cα atoms, 18% sequence identity; Hecker et al., 2007 ). The second domain of Kae1 has an additional helical domain inserted between β8 and α4 of TmYeaZ, as well as an αβ insert between β9 and α5. Most strikingly, the orientation of the second domain of YeaZs with respect to the first domain differs significantly from that of Kae1 and other ASKHA proteins (Fig. 2 a; Hecker et al., 2007 ; Nichols et al., 2006 ). We were unable to identify a conserved ATP-binding site at the domain interface, although it has previously been proposed that YeaZ may still bind nucleotides through significant rearrangement of its two domains (Nichols et al., 2006 ). YeaZs also lack the metal-binding motif of Kae1. In ASHKA proteins, the second domain plays an important role in stabilizing the adenosine base of the bound ATP. However, the substructure that interacts with the base in Kae1 is absent in TmYeaZ (Fig. 2 b). Thus, it is unclear how a suitable environment for nucleotide binding could be assembled within TmYeaZ.
EcYeaZ forms a stable complex with YgjD, but can also interact with the YjeE ATPase in a mutually exclusive manner (Handford et al., 2009 ). In order to identify sites that are potentially important for TmYeaZ function, we studied the sequence-conservation pattern of the YeaZ and Kae1 families in the context of the YeaZ structures (Fig. 3 ). The most prominent common feature of these proteins is the prevalence of the sequence motif GPGXXTGXR located at the N-terminus of a helix (α2 of TmYeaZ). This motif, which is reminiscent of a phosphate-binding motif, is close to the ATP-binding site in Kae1, but does not directly interact with ATP in Kae1. The arginine in this motif is exposed on the face of a helix in both Kae1 and YeaZs. Other conserved residues of the YeaZ homologs that are clustered around this motif include residues from the N-terminus of α1 (Lys32 and His33), the C-terminus of β1 (Asp7 and Thr8), the C-terminal portion of β6 (Arg112, Ala113 and Arg114), β7 (Tyr119) and the C-terminus of α6 (Pro188, Tyr190 and Gln192). This cluster of conserved residues indicates that this region is likely to be directly involved in the function of YeaZ. In TmYeaZ, a prominent positively charged electrostatic surface overlaps with this conserved surface; however, this feature is not conserved in EcYeaZ or StYeaZ. We speculate that part of or the entire conserved surface is involved in mediating protein–protein interactions between YeaZ and YgjD (TM0145) or YjeE (TM1632) in T. maritima.
It is worth noting that the conserved residues in α6 and α1 of TmYeaZ display large conformational differences compared with StYeaZ and EcYeaZ (Fig. 4 a). The residues corresponding to α6 of TmYeaZ are not in a helical conformation in StYeaZ and EcYeaZ despite being highly conserved in primary sequence. In the crystal structures of EcYeaZ and StYeaZ, α1 is packed more closely to α2 in an arrangement similar to that in Kae1. α1 and β2 are also stabilized by a hydrogen bond involving a conserved histidine (His34) from the N-terminus of α1 and the β1–β2 loop and a potential disulfide bond between two cysteines (Cys13 and Cys30 in StYeaZ). A corresponding interaction was not observed in TmYeaZ, since α1 is more distant from α2 and, as a result, the conserved histidine (His33 in TmYeaZ) is exposed and further from the conserved cluster (Fig. 4 a). The wider gap between α1 and α2 in TmYeaZ partially exposes the hydrophobic interior of the β1 and β2 strands, which are buried in the other YeaZ structures. Interestingly, a section of unaccounted-for electron density was observed in the resulting exposed groove of TmYeaZ and was modeled as an unidentified ligand (UNL; Fig. 4 b). This UNL could be a lipid or MPD, but it was not possible to unambiguously identify its nature. As the UNL is close to the conserved surface identified above, it may have functional implications; an alternative explanation that cannot be ruled out at present is that it is a crystallization artifact. Overall, the large conformational differences in these conserved residues may indicate that they are flexible in solution and may adopt a more rigid conformation when TmYeaZ is bound in the complex.
One of the dimers observed in the crystal lattice is likely to have physiological significance. EcYeaZ also dimerizes (Handford et al., 2009 ). However, it is not clear from the crystal structure which one of the two possible dimers represents the physiologically relevant form. The mode of dimerization affects the placement and exposure of the conserved surface identified above. The conserved surfaces of each monomer are fully exposed at either end in the AB dimer, while they are clustered together in the A 2 (or B 2) dimer (Figs. 3 d and 3 e). The A 2 dimer imposes very strict steric restrictions on the size of and the mode of interaction with its partners. Based on the size of interface, the AB dimer may be more stable. Given that the first domains of YgjD and YeaZ are structurally similar, the observed crystal-packing interaction seen in the A 2 dimer may mimic the complex between YeaZ and YgjD.
The specific function of YeaZ is still unknown. In Gram-positive organisms such as B. subtilis, YdiB (the YjeE homolog), YdiC (the YeaZ homolog), YdiD and YdiE (the Kae1 or Gcp homolog) are located within the same operon. The four corresponding proteins in T. maritima [TM1632 (TmYjeE), TM0874 (TmYeaZ), TM0577 and TM0145 (TmYgjD)] are dispersed across the genome. It has been suggested that YeaZ mediates the proteolysis of YgjD (Handford et al., 2009 ). However, no active site mimicking those of known peptidases can be identified based on the crystal structure. Thus, its mechanism remains unclear. YeaZ may fulfill its function by contributing one or more critical functional groups to a bipartite active site in a heterodimeric complex with YgjD or YjeE. Since YgjD and YjeE are both ATPases, it is possible that YeaZ functions as an ATPase inhibitor or activator. Further experiments are needed to elucidate the function of YeaZ and its possible partners.
The crystal structure of TmYeaZ (TM0874), a YeaZ homolog that is an essential protein in bacteria, has been elucidated. Based on its structure, TmYeaZ by itself is not likely to be an ATPase owing to the absence of a clearly defined ATP-binding site. A potential interface for mediating protein–protein interaction was identified. It remains possible that TmYeaZ may play a role in nucleotide binding or hydrolysis in a complex involving YgjD.
Essential gene products are excellent targets for antibacterial drugs. Unlike Kae1, YeaZ could be a potential drug target since it only seems to be present in bacteria. The crystal structure and the information presented here should be valuable for further biochemical characterization of this important bacterial protein. Additional information about TmYeaZ is available from TOPSAN (Krishna et al., 2010 ) http://www.topsan.org/explore?PDBid=2a6a.
This work was supported by the National Institute of General Medical Sciences Protein Structure Initiative (http://www.nigms.nih.gov) grants P50 GM62411 and U54 GM074898. Portions of this research were carried out at the Stanford Synchrotron Radiation Laboratory (SSRL). The SSRL is a national user facility operated by Stanford University on behalf of the US DOE, OBES. The SSRL Structural Molecular Biology Program is supported by the DOE, OBER and by NIH (NCRR, BTP and NIGMS). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of General Medical Sciences or the National Institutes of Health.