X‐ray crystal structure of a hypothetical Sua5 protein from <i>Sulfolobus tokodaii</i> strain 7

Abstract

The Sua5-yciO-yrdC domain proteins (pfam01300, pfam database1) are widely distributed in prokaryotes and eukaryotes (http://pfam.janelia.org/cgi-bin/getdesc?name=Sua5_yciO_yrdC) [Fig. 1(A)]. One of the proteins in this family, Escherichia coli YrdC, preferentially binds to double-stranded RNA and DNA.6 It has been predicted to be a rRNA maturation factor.7, 8 The crystal structure of E. coli YrdC has been determined and a possible nucleic acid-binding site has been identified.6 (A) Sequence alignment of Sulfolobus Sua5 (S.Sua5) with Sua5/YrdC/YciO family proteins. Strictly conserved and similar residues are boxed in red and represented by red letters, respectively. Tma, Thermotoga maritima MSB8 (gi:15643615); Sce, Saccharomyces cerevisiae S288C (gi:6321269); Cal, Candida albicans SC5314 (gi:68471205); Eco, E. coli K12 (gi:16131163); and Hsa, Homo sapiens (gi:114325403). The sequences were aligned using Clustal W.2 The secondary structure was predicted with DSSP,3 and the figure was generated with ESPript 2.2.4 (B) Ribbon diagram of Sulfolobus Sua5. The α-helices and β-strands are colored magenta and green, respectively. (C) Superpositioning of the main chain structures of Sulfolobus Sua5 (green) and E. coli YrdC (magenta). (D) Molecular surface representation of Sulfolobus Sua5. Red and blue surfaces represent negative and positive electrostatic potentials (−10 kBT, +10 kBT), respectively. AMP is shown as a stick model, in which yellow, red, blue, and orange represent carbon, oxygen, nitrogen, and phosphorus atoms, respectively. (E) Molecular surface colored as to sequence conservation, and difference electron density around the large concave. Strictly conserved residues, which are boxed in red in A, are colored magenta. The Fo − Fc omit map (gray mesh) was obtained without the ligand molecule at 1.8 Å resolution, with contouring at 4.0 σ. AMP is represented as in D. The electrostatic potentials were calculated using the Adaptive Poisson-Boltzmann Solver (APBS)5 with PyMol APBS tools. B–E were generated with program PyMol (http://www.pymol.org). The Sua5 protein consists of an N-terminal YrdC domain and a C-terminal Sua5 domain.6 The sua5 gene was first identified in Saccharomyces cerevisiae as a suppressor of a translation initiation defect in the leader region of the iso-1-cytochrome c (cyc1) gene, which is caused by the generation of an aberrant ATG start codon upstream of the gene as a consequence of a single base-pair substitution.9, 10 The function and three-dimensional structure of the Sua5 protein remain to be elucidated. In the present study, we determined the crystal structure of the hypothetical Sua5 (ST1526) protein from thermoacidophilic archaeon Sulfolobus tokodaii strain 7,11 which exhibits 49.7% similarity to S. cerevisiae Sua5 [Fig. 1(A)]. The open reading frame of S. tokodaii Sua5 (gi:15921819) was cloned into the pET-21a(+) expression vector (NdeI-BamHI site) (Novagen), with no purification tag. Selenomethionine-substituted proteins were produced in the E. coli methionine auxotroph BL21-CodonPlus (DE3)-RIL-X strain (Stratagene). The cell lysate was heated at 90°C for 10 min. Then the soluble fraction was applied to a Resource ISO column (GE Healthcare Biosciences) equilibrated with 50 mM sodium phosphate buffer (pH 7.0) containing 1.5M (NH4)2SO4, which was eluted with a linear gradient of 1.5–0M (NH4)2SO4. The fractions containing the Sua5 protein were collected and applied to a Resource Q column (GEHealthcare Biosciences) equilibrated with 20 mM Tris-HCl (pH 8.0), which was eluted with a linear gradient of 0–0.5M NaCl. The fractions containing the Sua5 protein were collected and applied to a hydroxyapatite CHT10-I column (Bio-Rad) equilibrated with 10 mM sodium phosphate buffer (pH 7.0), which was eluted with a linear gradient of 10–500 mM sodium phosphate buffer (pH 7.0). The fractions containing the Sua5 protein were collected and applied to a HiLoad 16/60 Superdex 200 column (GE Healthcare Biosciences) equilibrated with 20 mM Tris-HCl (pH 8.0) containing 150 mM NaCl. The purified protein was concentrated to 9.6 mg/mL using a Vivaspin 20 concentrator (10,000 molcular-weight cutoff, Sartorius), and dithiothreitol was added to the sample to a final concentration of 1 mM. The N- terminal amino-acid sequence of the purified protein was TQIIK, which means that the N-terminal methionine had been deleted. The molecular mass of the purified protein estimated on light scattering photometry was 38.5 kDa, suggesting that it exists as a monomer in solution. Crystallization of the Sua5 protein was performed by the sitting drop vapor diffusion method by mixing 1 μL of a protein solution with an equal volume of a reservoir solution comprising 0.1M sodium acetate (pH 4.6) and 4M sodium nitrate (Salt RX, no.46, Hampton Research) at 20°C. The crystal in the mother liquor containing 20% glycerol was cryo-cooled in a nitrogen-gas stream. Multiple-wavelength anomalous dispersion (MAD) data were collected at the RIKEN Structural Genomics Beamline II (BL26B2) at SPring-8 (Hyogo, Japan) utilizing the anomalous scattering from Se atoms. The data sets were collected at 1.8 Å resolutions using a Jupiter 210 CCD detector (Rigaku MSC). The collected data were processed with the HKL2000 program suite.12 The atomic positions of the Se atoms in the unit cell were determined with program SOLVE,13 and then density modification was performed with program RESOLVE.14 The automatic-tracing procedure in program ARP/wARP15 was utilized to build the initial model. The model refinement, initial picking, and manual verifying of water molecules were carried out using programs CNS and Xtalview/Xfit.16, 17 In order to compensate for excess electron density of the N-terminal domain, AMP was presumptively used as a ligand [Fig. 1(E)]. According to PROCHECK in the CCP4 suite,18 91.9% of the residues in the final model are in the most favored region of a Ramachandran plot, with no residues in the disallowed regions. The data collection statistics and processed data statistics are presented in Table I. The coordinates are available in the Protein Data Bank, under accession code 2EQA. The overall structure of Sulfolobus Sua5 is shown in Figure 1(B). Note that there are two disordered regions, residues 196 and 215–233, which are not included in the model. Further, the side chains of Glu190 and Tyr234 could not be observed. The N-terminal domain of Sulfolobus Sua5 (residues 2–214) consists of eight α-helices and eleven β-strands. The overall fold of the N-terminal domain is similar to that of E. coli YrdC6 that exhibits 42.9% similarity in amino acid sequence with the N-terminal domain of Sulfolobus Sua5, the Z-score being 21.3 and the r.m.s.d. value being 2.4 Å [Fig. 1(C)].19 A large concave surface exhibiting a positive electrostatic potential, which is similar to that in YrdC, where nucleic acids are predicted to interact,6 was found in Sulfolobus Sua5 [Fig. 1(D)]. Interestingly, excess electron density that might be due to an E. coli-derived nucleotide was observed on this concave surface of Sulfolobus Sua5 [Fig. 1(D,E)]. The C-terminal Sua5 domain (residues 234–352) of Sulfolobus Sua5 consists of three α-helices, α9, residues 249–259; α10, residues 291–307; and α11, residues 324–335, and five β-strands, β12, residues 242–246; β13, residues 264–270; β14, residues 281–286; β15, residues 322–326; and β16, residues 338–340, which adopt a Rossmann fold. The C-terminal domain structure was compared with the previously determined structures in the PDB database, using the DALI server.19 The closest structure was that of Methanocaldococcus jannaschii HypB, a GTP-binding protein that regulates metal binding20 (PDB code: 2HF9 chain A), the Z-score being 7.7 and the r.m.s.d. value being 3.2 Å, despite the low sequence identity (IDE = 14%). The structural homologs that showed high Z-scores were hypothetical protein TM0796 from Thermotoga maritima (PDB code: 2G0T chain A, Z = 7.4, r.m.s.d. = 2.8 Å, IDE = 11%), the GTP-binding domain of Fth from Thermus aquaticus21 (PDB code: 1NG1, Z = 6.4, r.m.s.d. = 3.1 Å, IDE = 10%), and E. coli cytidine triphosphate synthetase22 (PDB code: 1S1M chain A, Z = 5.6, r.m.s.d. = 2.6 Å, IDE = 8%). Thus, the C-terminal Sua5 domain might be involved in nucleotide binding. The three-dimensional structure of Sulfolobus Sua5 showed that both the N- and C-terminal domains were possibly involved in nucleotide binding or metabolism, which is supported by the observation that Sulfolobus Sua5 showed ATP hydrolysis activity, AMP being produced, and the apparent rate constants were 0.04 and 0.1 s−1 at 37 and 60°C, respectively, in the presence of 1 mM ATP, 5 mM MgCl2, 50 mM HEPES, and 100 mM KCl, pH 7.5. The authors wish to thank Emiko Matsunaga for protein purification, Takeshi Nagira for crystallization, and Seiki Baba and Yoshiaki Kitamura for data collection at SPring-8.