A New Arrangement of (β/α)8 Barrels in the Synthase Subunit of PLP Synthase
2005; Elsevier BV; Volume: 280; Issue: 30 Linguagem: Inglês
10.1074/jbc.m503642200
ISSN1083-351X
AutoresJianghai Zhu, John W. Burgner, Etti Harms, Boris R. Belitsky, Janet L. Smith,
Tópico(s)Enzyme Production and Characterization
ResumoPyridoxal 5′-phosphate (PLP, vitamin B6), a cofactor in many enzymatic reactions, has two distinct biosynthetic routes, which do not coexist in any organism. Two proteins, known as PdxS and PdxT, together form a PLP synthase in plants, fungi, archaea, and some eubacteria. PLP synthase is a heteromeric glutamine amidotransferase in which PdxT produces ammonia from glutamine and PdxS combines ammonia with five- and three-carbon phosphosugars to form PLP. In the 2.2-Å crystal structure, PdxS is a cylindrical dodecamer of subunits having the classic (β/α)8 barrel fold. PdxS subunits form two hexameric rings with the active sites positioned on the inside. The hexamer and dodecamer forms coexist in solution. A novel phosphate-binding site is suggested by bound sulfate. The sulfate and another bound molecule, methyl pentanediol, were used to model the substrate ribulose 5-phosphate, and to propose catalytic roles for residues in the active site. The distribution of conserved surfaces in the PdxS dodecamer was used to predict a docking site for the glutaminase partner, PdxT. Pyridoxal 5′-phosphate (PLP, vitamin B6), a cofactor in many enzymatic reactions, has two distinct biosynthetic routes, which do not coexist in any organism. Two proteins, known as PdxS and PdxT, together form a PLP synthase in plants, fungi, archaea, and some eubacteria. PLP synthase is a heteromeric glutamine amidotransferase in which PdxT produces ammonia from glutamine and PdxS combines ammonia with five- and three-carbon phosphosugars to form PLP. In the 2.2-Å crystal structure, PdxS is a cylindrical dodecamer of subunits having the classic (β/α)8 barrel fold. PdxS subunits form two hexameric rings with the active sites positioned on the inside. The hexamer and dodecamer forms coexist in solution. A novel phosphate-binding site is suggested by bound sulfate. The sulfate and another bound molecule, methyl pentanediol, were used to model the substrate ribulose 5-phosphate, and to propose catalytic roles for residues in the active site. The distribution of conserved surfaces in the PdxS dodecamer was used to predict a docking site for the glutaminase partner, PdxT. Pyridoxal 5′-phosphate (PLP), 1The abbreviations used are: PLP, pyridoxal 5′-phosphate; IGPS, imidazole glycerol phosphate synthase; PDB, Protein Data Bank; TEV, tobacco etch virus; SeMet, selenomethionine; MPD, 2-methyl-2,4-pentanediol; SAD, single-wavelength anomalous diffraction; TIM, triose phosphate isomerase. the biologically active form of vitamin B6, is an essential cofactor in numerous biochemical reactions. Despite the importance of PLP, its biosynthesis was unknown until recently. Two distinct de novo PLP biosynthetic pathways have been characterized, which do not coexist in organisms that produce PLP (1Tanaka T. Tateno Y. Gojobori T. Mol. Biol. Evol. 2005; 22: 243-250Crossref PubMed Scopus (70) Google Scholar). The first pathway is employed by many eubacteria, such as Escherichia coli, to produce PLP from erythrose 4-phosphate and 1-deoxy-d-xylulose 5-phosphate (2Drewke C. Leistner E. Vitam. Horm. 2001; 61: 121-155Crossref PubMed Google Scholar, 3Hill R.E. Spenser I.D. Neidhardt F.C. Curtiss R.I. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. American Society for Microbiology, Washington, D. C.1996: 695-703Google Scholar). The PLP nitrogen is derived from glutamate by transamination (4Tanaka K. Tazuya K. Yamada K. Kumaoka H. J. Nutr. Sci. Vitaminol. (Tokyo). 2000; 46: 55-57Crossref PubMed Scopus (15) Google Scholar). All E. coli genes encoding enzymes in this pathway have been identified, the pathway has been reconstituted in a cell-free system, and structures of most of the enzymes have been determined (5Cane D.E. Chimia. 2003; 57: 75-76Crossref Google Scholar). In the second B6 biosynthetic pathway (Fig. 1), which exists in some eubacteria and in archaea, fungi, plants, plasmodia, and some metazoa, the carbon skeleton of PLP is synthesized from an intact five-carbon pent(ul)ose unit and an intact three-carbon triose unit (6Zeidler J. Gupta R.N. Sayer B.G. Spenser I.D. J. Org. Chem. 2003; 68: 3486-3493Crossref PubMed Scopus (18) Google Scholar, 7Gupta R.N. Hemscheidt T. Sayer B.G. Spenser I.D. J. Am. Chem. Soc. 2001; 123: 11353-11359Crossref PubMed Scopus (26) Google Scholar). Very recently these substrates were shown to be ribulose 5-phosphate (or ribose 5-phosphate) and glyceraldehyde 3-phosphate (or dihydroxyacetone phosphate) (8Burns K.E. Xiang Y. Kinsland C.L. McLafferty F.W. Begley T.P. J. Am. Chem. Soc. 2005; 127: 3682-3683Crossref PubMed Scopus (106) Google Scholar). The PLP nitrogen is derived from the amide group of glutamine (9Tazuya K. Adachi Y. Masuda K. Yamada K. Kumaoka H. Biochim. Biophys. Acta. 1995; 1244: 113-116Crossref PubMed Scopus (35) Google Scholar). Only two genes have been implicated in this pathway, which we refer to as the PdxS/PdxT pathway for the names of the corresponding gene products in Bacillus subtilis (10Belitsky B.R. J. Bacteriol. 2004; 186: 1191-1196Crossref PubMed Scopus (70) Google Scholar). Homologs of pdxS are known as yaaD, pdx1, PDX1, SNZ, SOR1, PYROA, and HEVER, and pdxT homologs are known as yaaE, pdx2, PDX2, SNO, and SNZB (11Bean L.E. Dvorachek Jr., W.H. Braun E.L. Errett A. Saenz G.S. Giles M.D. Werner-Washburne M. Nelson M.A. Natvig D.O. Genetics. 2001; 157: 1067-1075Crossref PubMed Google Scholar, 12Ehrenshaft M. Bilski P. Li M.Y. Chignell C.F. Daub M.E. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9374-9378Crossref PubMed Scopus (256) Google Scholar, 13Ehrenshaft M. Daub M.E. J. Bacteriol. 2001; 183: 3383-3390Crossref PubMed Scopus (70) Google Scholar, 14Osmani A.H. May G.S. Osmani S.A. J. Biol. Chem. 1999; 274: 23565-23569Abstract Full Text Full Text PDF PubMed Scopus (150) Google Scholar, 15Rodriguez-Navarro S. Llorente B. Rodriguez-Manzaneque M.T. Ramne A. Uber G. Marchesan D. Dujon B. Herrero E. Sunnerhagen P. Perez-Ortin J.E. Yeast. 2002; 19: 1261-1276Crossref PubMed Scopus (81) Google Scholar, 16Seack J. Perovic S. Gamulin V. Schroder H.C. Beutelmann P. Muller I.M. Muller W.E. Biochim. Biophys. Acta. 2001; 1520: 21-34Crossref PubMed Scopus (24) Google Scholar, 17Wrenger C. Eschbach M.L. Muller I.B. Warnecke D. Walter R.D. J. Biol. Chem. 2005; 280: 5242-5248Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 18Sakai A. Kita M. Katsuragi T. Ogasawara N. Tani Y. J. Biosci. Bioeng. 2002; 93: 309-312Crossref PubMed Scopus (27) Google Scholar). The pdxS- and pdxT-like genes are proximal or adjacent to one another in most organisms where the PdxS/PdxT pathway exists. PdxS-like proteins have among the most highly conserved sequences to emerge from proteome comparisons (19Galperin M.Y. Koonin E.V. Mol. Microbiol. 1997; 24: 443-445Crossref PubMed Scopus (33) Google Scholar, 20Padilla P.A. Fuge E.K. Crawford M.E. Errett A. Werner-Washburne M. J. Bacteriol. 1998; 180: 5718-5726Crossref PubMed Google Scholar). Our analysis of PdxS-like sequences shows that, among 91 sequences from all kingdoms of life (archaea, eubacteria and eukarya), all pairs of sequences are at least 60% identical, and ∼22% of the residue positions are strictly conserved. Sequence searches with PdxS and its homologs do not identify proteins of known structure, although PdxS homology to some members of the β/α barrel superfamily was predicted (19Galperin M.Y. Koonin E.V. Mol. Microbiol. 1997; 24: 443-445Crossref PubMed Scopus (33) Google Scholar, 21Bauer J.A. Bennett E.M. Begley T.P. Ealick S.E. J. Biol. Chem. 2004; 279: 2704-2711Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar). PdxT is a homolog of the glutaminase domains (or subunits) of several glutamine amidotransferases (19Galperin M.Y. Koonin E.V. Mol. Microbiol. 1997; 24: 443-445Crossref PubMed Scopus (33) Google Scholar), in accord with glutamine being the nitrogen source for the PdxS/PdxT pathway. PdxT is a member of the "Triad" family of glutaminase domains, characterized by a conserved Cys-His-Glu triad in the active site (22Zalkin H. Smith J.L. Purich D.L. Advances in Enzymology and Related Areas of Molecular Biology. Interscience Publishers, New York1998: 87-144Google Scholar), and is most similar to the glutaminase domain of imidazole glycerol phosphate synthase (IGPS), an enzyme from the histidine biosynthetic pathway. The expected Triad structure was observed in crystal structures of PdxT from B. subtilis (21Bauer J.A. Bennett E.M. Begley T.P. Ealick S.E. J. Biol. Chem. 2004; 279: 2704-2711Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar) and from its thermophilic cousin Geobacillus stearothermophilus (PDB code 1Q7R). Considerable experimental evidence demonstrates that PdxS/PdxT is a heteromeric glutamine amidotransferase. PdxS and PdxT from B. subtilis form a physical complex (10Belitsky B.R. J. Bacteriol. 2004; 186: 1191-1196Crossref PubMed Scopus (70) Google Scholar), as do their homologs from yeast (20Padilla P.A. Fuge E.K. Crawford M.E. Errett A. Werner-Washburne M. J. Bacteriol. 1998; 180: 5718-5726Crossref PubMed Google Scholar, 23Dong Y.X. Sueda S. Nikawa J. Kondo H. Eur. J. Biochem. 2004; 271: 745-752Crossref PubMed Scopus (60) Google Scholar). Glutaminase activity was detected for B. subtilis PdxT (10Belitsky B.R. J. Bacteriol. 2004; 186: 1191-1196Crossref PubMed Scopus (70) Google Scholar) and yeast SNO1 (23Dong Y.X. Sueda S. Nikawa J. Kondo H. Eur. J. Biochem. 2004; 271: 745-752Crossref PubMed Scopus (60) Google Scholar), and this activity required the presence of the second subunit, B. subtilis PdxS or yeast SNZ1. The recent observation of PLP synthase activity in vitro with glutamine as a nitrogen substrate shows that the PdxS/PdxT complex, now known as PLP synthase, is indeed a glutamine amidotransferase having glutaminase (PdxT) and synthase (PdxS) subunits (8Burns K.E. Xiang Y. Kinsland C.L. McLafferty F.W. Begley T.P. J. Am. Chem. Soc. 2005; 127: 3682-3683Crossref PubMed Scopus (106) Google Scholar). Glutamine amidotransferases are a group of about 20 enzymes that remove the amide nitrogen from glutamine and add it to another substrate (22Zalkin H. Smith J.L. Purich D.L. Advances in Enzymology and Related Areas of Molecular Biology. Interscience Publishers, New York1998: 87-144Google Scholar). The active sites for glutamine hydrolysis and ammonia addition are chemically distinct and are carried on different enzyme domains or subunits (24Binda C. Bossi R.T. Wakatsuki S. Arzt S. Coda A. Curti B. Vanoni M.A. Mattevi A. Struct. Fold. Des. 2000; 8: 1299-1308Abstract Full Text Full Text PDF Scopus (76) Google Scholar, 25Tesmer J.J. Klem T.J. Deras M.L. Davisson V.J. Smith J.L. Nat. Struct. Biol. 1996; 3: 74-86Crossref PubMed Scopus (211) Google Scholar, 26Krahn J.M. Kim J.H. Burns M.R. Parry R.J. Zalkin H. Smith J.L. Biochemistry. 1997; 36: 11061-11068Crossref PubMed Scopus (187) Google Scholar, 27Thoden J.B. Holden H.M. Wesenberg G. Raushel F.M. Rayment I. Biochemistry. 1997; 36: 6305-6316Crossref PubMed Scopus (306) Google Scholar, 28Knochel T. Ivens A. Hester G. Gonzalez A. Bauerle R. Wilmanns M. Kirschner K. Jansonius J.N. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9479-9484Crossref PubMed Scopus (105) Google Scholar, 29Chaudhuri B.N. Lange S.C. Myers R.S. Chittur S.V. Davisson V.J. Smith J.L. Structure (Camb.). 2001; 9: 987-997Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Ammonia generated by the glutaminase domain of the enzyme is sequestered by the protein and transferred through an internal tunnel between the active sites. The working model for PLP synthase catalysis is that ribulose 5-phosphate or ribose 5-phosphate and glyceraldehyde 3-phosphate or dihydroxyacetone phosphate bind in the active site of PdxS where they combine with ammonia generated by PdxT to form PLP. Here we report the 2.2-Å crystal structure of PdxS from G. stearothermophilus. The PdxS dodecamer observed in the crystal exists in solution in equilibrium with a hexamer. Fortuitous binding of a crystal cryoprotectant molecule provides clues about the mode of ribulose 5-phosphate binding to PdxS and assignment of catalytic roles for some residues. A model of the PdxS/PdxT complex is proposed. Cloning of G. stearothermophilus pdxS—Escherichia coli strains JM107 (30Yanisch-Perron C. Vieira J. Messing J. Gene (Amst.). 1985; 33: 103-119Crossref PubMed Scopus (11465) Google Scholar), XL1-Blue (Stratagene), and BL21(DE3) were used for plasmid construction and gene expression. For crystallization experiments, plasmids expressing two versions of the His6-tagged PdxS were constructed. pBB1186 encodes PdxS with a C-terminal His6-tag. The modified pdxS gene was synthesized by PCR, using the chromosomal DNA of G. stearothermophilus strain 10 (www.genome.ou.edu/bstearo.html) as a template, and oligonucleotides oBB126 (5′-CTTTTGCATGCCTCGCAAAGC) and oBB127 (5′-TTATTAAGCTTAGTGGTGGTGGTGGTGGTGCCAGCCGCGTTCTTGCATC), as 5′- and 3′-primers, respectively (the SphI and HindIII sites are underlined, and the histidine codons are in bold). The PCR fragment was cloned between the SphI and HindIII sites of the expression vector pBAD18 containing the inducible E. coli araBAD promoter (31Guzman L.M. Belin D. Carson M.J. Beckwith J. J. Bacteriol. 1995; 177: 4121-4130Crossref PubMed Scopus (3960) Google Scholar). pET0881 encodes PdxS with an N-terminal tag, including His6 and 26 other amino acids that provide thrombin and tobacco etch virus (TEV) protease cleavage sites (residue numbering of PdxS is relative to Met1 of the native protein). To create this plasmid, the modified pdxS gene was synthesized using pBB1186 as a template and oligonucleotides (5′-AAACATATGGCTAGCACAGGTACGGACCGCGTC) and (5′-TTTTTATTAAGCTTACCAGCCGCGTTCTTGCATCC) as forward and reverse primers, containing the NheI and HindIII sites, respectively (underlined). The PCR product was cloned between the NheI and HindIII sites of pETTEV281. Expression plasmid pETTEV281, a derivative of pET28a (Novagen), encodes a recognition site for TEV protease following the N-terminal His-Tag and the thrombin cleavage site. pETTEV281 was constructed as follows. First, the NdeI site upstream of the T7-Tag was replaced by a KpnI site by site-directed mutagenesis (QuikChange by Stratagene) using the primer 5′-CGCGCGGCAGCGGTACCGCTAGCATGACTGGTGGAC-3′ and its complement. A DNA sequence was designed to encode the TEV protease recognition site (32Carrington J.C. Dougherty W.G. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 3391-3395Crossref PubMed Scopus (172) Google Scholar) and to insert between the newly created KpnI site and the BamHI site in the multiple cloning region. A new NdeI site upstream and adjacent to the BamHI site was created for convenient transfer of DNA fragments between vectors. The annealed product of the designed oligonucleotides, had sticky ends (bold) for insertion between KpnI and BamHI sites (NdeI site is underlined, and the TEV protease cleavage site is indicated by ↓). The oligonucleotides were phosphorylated in a 120-ml reaction mixture (600 pmol of oligomer, 1 mm ATP, 40 units of T4 polynucleotide kinase (New England Biolabs), 70 mm Tris-HCl, pH 7.6, 10 mm MgCl2, 5 mm dithiothreitol) and incubated at 37 °C for 30 min. The reaction was quenched by addition of 25 mm EDTA (final concentration) and the kinase removed by phenol extraction. The phosphorylated oligonucleotides were annealed at 95 °C and slowly cooled to room temperature. The annealed duplex was ligated into modified vector pET28a (no NdeI site) that had been digested by KpnI and BamHI under standard conditions. The ligation product was transformed into competent XL1-Blue cells (Stratagene). Plasmid DNA from colonies was screened by restriction digestion for the new NdeI site, originating from the insert. DNA with the expected restriction pattern was sequenced at the Purdue Genomics Center to verify the construct and the plasmid was named pETTEV281. We were able to produce crystals using either C-terminal or N-terminal His6-tagged versions of PdxS. Below we describe the procedures used for purification and crystallization of the N-terminal His6 form of PdxS. Purification of PdxS—Cells of E. coli BL21(DE3) were transformed with plasmid pET0881 and grown in Luria-Bertani medium containing 50 μg/ml kanamycin. Expression of pdxS was induced by addition of IPTG (final concentration 400 μm) when the A600 reached 0.8 and the culture was incubated at 37 °C for 6 h. All subsequent steps were carried out at 4 °C unless otherwise noted. Cells were harvested by centrifugation, resuspended in 25 ml of extraction buffer (20 mm imidazole, 20 mm Tris-Cl, pH 7.9, 500 mm NaCl), and lysed by two passes through a French press at 10,000 p.s.i. The lysate was centrifuged at 18,000 × g for 30 min. The clarified cell extract was loaded onto a Ni2+-charged immobilized metal affinity chromatography column (HiTrap™ Chelating HP, Amersham Biosciences) pre-equilibrated with the extraction buffer. A gradient of 20-500 mm imidazole (in 20 mm Tris-Cl, pH 7.9, 500 mm NaCl) was applied to the column. The N-terminal His-tagged PdxS eluted at ∼250 mm imidazole. Purified PdxS was dialyzed at room temperature against 20 mm Tris-Cl, pH 8.3, and concentrated to 20 mg/ml. The purified PdxS was extremely sensitive to pH and precipitated in solutions below pH 8. SeMet-PdxS was produced by the same protocol, except cells were grown in the SeMet minimal medium (33Guerrero S.A. Hecht H.J. Hofmann B. Biebl H. Singh M. Appl. Microbiol. Biotechnol. 2001; 56: 718-723Crossref PubMed Scopus (82) Google Scholar) plus 50 mg of l-SeMet per liter. No reducing agent was used during purification. Purified SeMet-PdxS was dialyzed at room temperature against 20 mm Tris-Cl, pH 8.3, and concentrated to 10 mg/ml. Crystallization—PdxS was crystallized at 20 °C by vapor diffusion from a 1:1 mixture of protein (5 mg/ml PdxS, 20 mm Tris-Cl, pH 8.1, 20 mm GDP, 20 mm MgCl2) and well solution (10-12% polyethylene glycol 8000, 100 mm sodium cacodylate, pH 7.0, 200 mm (NH4)2SO4). (GDP, which was used in our original trials, was found later to be unnecessary for PdxS crystallization.) Small crystals were washed in well solution, crushed, and diluted 105-fold for microseeding (1 μl) into drops containing 4 μl of protein solution and 4 μl of well solution. Large crystals (250 × 250 × 250 μm) grew in 3-5 weeks. Crystals of PdxS are in space group C222 (a = 162.5 Å, b = 163.9 Å, c = 186.8 Å) with six polypeptides in the asymmetric unit and ∼65% solvent. Crystals of SeMet-PdxS were grown under similar conditions at a slightly lower concentration of polyethylene glycol 8000. A mercury derivative crystal was produced by overnight soaking in well solution with 1 mm ethylmercuric phosphate, followed by a 1-min back-soak during cryoprotection. All crystals were cryoprotected by rapid sequential transfer through 5, 10, 15, and 17.5% 2-methyl-2,4-pentanediol (MPD) in well solution before flash freezing in a 100 K N2 gas stream. Data Collection and Processing—A 2.2 Å native data set was collected at APS beamline 19-ID (360 images of 0.5° oscillation width) and was indexed and integrated using HKL2000 (34Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38572) Google Scholar). A three-wavelength SeMet MAD data set was collected from one crystal at APS beamline 19-BM using the inverse beam method and was indexed and integrated by HKL2000. A 3.0 Å mercury SAD data set was collected from one crystal at APS beamline 14-ID. The wavelength of peak absorption at the mercury LIII edge (1.0075 Å) was used to collect 360° of data as 0.5° oscillation images, which were indexed and integrated using MOSFLM (35Leslie A.G.W. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography, No. 26. Daresbury Laboratory, Warrington, UK1992Google Scholar) in the CCP4 suite (36Bailey S. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (41) Google Scholar). Integrated intensities for all data sets were scaled using SCALA (36Bailey S. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (41) Google Scholar). Data processing statistics are shown in Table I.Table ICrystallographic summaryNativeSeMetHg derivative, peakPeakInflectionRemoteData collectionBeamlineAPS 19IDAPS 19BMAPS 19BMAPS 19BMAPS 14IDWavelength (Å)0.979510.980020.980270.965011.00750Space groupC222C222C222C222C222Unit cell (Å)a162.5163.3163.7163.8162.6b163.9164.5164.8164.9164.6c186.8187.1187.3187.4186.9dmin (Å)2.22.83.23.63.0Completeness (%)99.9 (99.9)aValues in parentheses pertain to the outermost shell of data99.8 (99.8)99.5 (99.3)99.6 (99.5)100.0 (100.0)Redundancy7.3 (6.8)9.5 (8.7)9.2 (9.0)8.3 (8.4)14.6 (14.6)RmergebRmerge=∑n∑hkl|Ihkln−〈Ihkl〉|∑n〈Ihkl〉 where Ihkln is the nth observation of reflection hkl, and 〈Ihkl〉 is the average intensity for all observations of reflection hkl (%)0.076 (0.55)0.09 (0.4)0.09 (0.45)0.11 (0.39)0.13 (0.45)I/σ (last shell)22.5 (3.7)20.9 (3.0)19.9 (3.1)15.3 (3.5)13.8 (3.6)Structure refinementData range (Å)50.0-2.2Protein atoms (#)11,034Water molecules (#)1507Ligand atoms (#)108Reflections in refinement119,560Reflections in test set6320RcR=∑hkl|Fhklobs−Fhklcalc|∑hklFhklobsRfree was computed for 5% of the data (%)15.9RfreecR=∑hkl|Fhklobs−Fhklcalc|∑hklFhklobsRfree was computed for 5% of the data (%)19.8Root mean square deviationBonds (Å)0.017Angles (°)1.56Average B factor (Å2)Main chain38.2Side chain39.9Ligand45.2Water48.9Ramachandran outliers (no.)0a Values in parentheses pertain to the outermost shell of datab Rmerge=∑n∑hkl|Ihkln−〈Ihkl〉|∑n〈Ihkl〉 where Ihkln is the nth observation of reflection hkl, and 〈Ihkl〉 is the average intensity for all observations of reflection hklc R=∑hkl|Fhklobs−Fhklcalc|∑hklFhklobsRfree was computed for 5% of the data Open table in a new tab Structure Determination—We first attempted to solve the PdxS structure by SeMet MAD. As the PdxS polypeptide contains 15 Met residues, 90 selenium sites were expected in the asymmetric unit. However, we were unable to solve the selenium substructure using several programs (SOLVE (37Terwilliger T.C. Berendzen J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 849-861Crossref PubMed Scopus (3220) Google Scholar), SHELX (38Schneider T.R. Sheldrick G.M. Acta Crystallogr. Sect. D Biol. Crystallogr. 2002; 58: 1772-1779Crossref PubMed Scopus (1578) Google Scholar), ACORN (39Yao J.X. Acta Crystallogr. Sect. D Biol. Crystallogr. 2002; 58: 1941-1947Crossref PubMed Scopus (27) Google Scholar), SAPI (40Hao Q. Gu Y.X. Yao J.X. Zheng C.D. Fan H.F. J. Appl. Crystallogr. 2003; 36: 1274-1276Crossref Scopus (9) Google Scholar), and BnP (41Weeks C.M. Blessing R.H. Miller R. Mungee R. Potter S.A. Rappleye J. Smith G.D. Xu H. Furey W. Z. Kristallogr. 2002; 217: 686-693Crossref Scopus (54) Google Scholar)), presumably due to the poor quality of the data. We next turned to the mercury SAD data set and solved the mercury substructure with both SHELX and BnP. Six pairs of mercury sites were found in the asymmetric unit. The PdxS polypeptide has two Cys residues in the sequence Cys-Gly-Cys. Each cysteine residue bound one mercury atom. The six pairs of sites were consistent with the self-rotation function, which revealed 6-fold noncrystallographic symmetry perpendicular to the crystallographic c axis. A 3 Å unaveraged map generated by the PHASES routine in BnP was interpretable. A Fourier map using Bijvoet differences from the SeMet-PdxS peak wavelength and mercury-SAD phases revealed most of the selenium sites, confirming the correctness of the phasing and also aiding chain tracing. The 6-fold redundancy was exploited to extend the experimental 3.0 Å SAD phases to 2.2 Å in the native data using RESOLVE (37Terwilliger T.C. Berendzen J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 849-861Crossref PubMed Scopus (3220) Google Scholar). An 80%-complete model was auto-traced by ARP/wARP (42Perrakis A. Morris R. Lamzin V.S. Nat. Struct. Biol. 1999; 6: 458-463Crossref PubMed Scopus (2564) Google Scholar) in the phase-refined 2.2 Å map. All manual model building was done in program O (43Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13011) Google Scholar). The model was refined with REFMAC5 (44Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13870) Google Scholar). Water sites were identified automatically by ARP_WATERS in the CCP4 suite, refined, and verified manually. Density corresponding to MPD was initially modeled as unknown atomic sites, which were refined without non-bonded contact restraints using the dummy-atom feature of REFMAC5. Models for both enantiomers of MPD were fit to the resulting five dummy-atomic sites. Refinement statistics are shown in Table I. The final model is available in the Protein Data Bank with accession code 1ZNN. Structure superpositions were performed with program O, sequence alignment was with CLUSTALX (45Thompson J.D. Gibson T.J. Plewniak F. Jeanmougin F. Higgins D.G. Nucleic Acids Res. 1997; 25: 4876-4882Crossref PubMed Scopus (35495) Google Scholar), and molecular figures were prepared with PyMOL (46DeLano W.L. The PyMOL Molecular Graphics System. DeLano Scientific LLC, South San Francisco, CA2002Google Scholar). Analytical Ultracentrifugation—Sedimentation velocity experiments were performed in an Optima XL-I ultracentrifuge (Beckman-Coulter, Fullerton, CA). A double-sector charcoal-filled Epon centerpiece with sapphire windows and 1.2-mm pathlength was used. The reference sector was filled with 420 μl of dialysis buffer (20 mm Tris-Cl, pH 8.1, without or with sulfate or phosphate) and the sample sector with an equivalent volume of sample containing ∼1 mg/ml protein. Solvent densities and viscosities were calculated with SEDNTERP (47Laue T.M. Shah B.D. Ridgeway T.M. Pelletier S.L. Harding S.E. Rowe A.J. Horton J.C. Analytical Ultracentrifugation in Biochemistry and Polymer Science. The Royal Society of Chemistry, Cambridge, UK1992: 90-125Google Scholar). Following thermal equilibration in the centrifuge at 20 °C for at least 1 h at 0 rpm, the velocity experiment was initiated by accelerating the rotor to 40,000 rpm. Boundary positions were determined using both the Rayleigh interference, and the absorbance (280 nm) optics with scans taken every 3-5 min. Data were processed with the programs SEDFIT 8.9 (48Schuck P. Biophys. J. 2000; 78: 1606-1619Abstract Full Text Full Text PDF PubMed Scopus (3064) Google Scholar) and SEDPHAT 3.0 (49Balbo A. Minor K.H. Velikovsky C.A. Mariuzza R.A. Peterson C.B. Schuck P. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 81-86Crossref PubMed Scopus (111) Google Scholar). PdxS Fold—PdxS has a classic (β/α)8 barrel fold, consisting of eight parallel β-strands (β1-β8) alternating with eight α helices (α1-α8) along the polypeptide chain (Fig. 2a). The barrel architecture is decorated by two major insertions, which participate in subunit contacts (see below). The first insertion is a large helical protrusion from the side and bottom of the barrel, formed by helix α6′ (residues 179-189), helix α6″ (residues 193-202), and the C terminus (residues 166-176) of helix α6. The second insertion is a C-terminal helix (α8″; residues 261-268), which lies on the side of the barrel such that the polypeptide C terminus is unconventionally at the top of the barrel. Two smaller deviations from the prototypical barrel include a small 310 helix, which we name α8′ (residues 237-240), on the top of the barrel, and a short 310 helix (residues 117-121), which we name α4 by convention, in place of barrel helix α4. Based on the sequence alignment of PdxS and 90 homologs, all of these deviations from the canonical barrel architecture are conserved. Several residues are disordered and missing from the structure, including 33 from the top of the barrel (residues 46-55 in the β2-α2 loop; residues 272-294 at the C terminus) and 17 from the bottom (residues 1-17 preceding β1). PdxS Quaternary Structure—PdxS crystallized as an unusual dodecamer in which the subunits form a cylinder 90 Å tall and 110 Å in diameter (Fig. 2b). The cylinder walls are 35 Å thick, creating an internal space 40 Å in diameter. The dodecamer is constructed of two opposing hexameric rings. The solid waist of the cylinder is formed by interdigitation of the helical protrusions from the two hexamers. Each end of the cylinder is a ring of six (β/α)8 barrels. The top of each (β/α)8 barrel points inward and the bottom toward the outside of the cylinder. The barrel axis of each subunit does not point directly inward but is tipped 25° toward the nearest end of the cylinder. Subunits in the hexameric ring have extensive hydrophobic contacts that exclude water and bury 17% of the subunit surface. The helical insertion α8″ is part of this interface. A strictly conserved salt bridge from Asp220 to Arg83 and His86 of a neighboring subunit is buried in the otherwise hydrophobic interface. By criteria of interface size, hydrophobicity, packing, and conservation, the hexameric ring appears to be a feature of PdxS and its 90 closest relatives. In contrast to contacts within the hexamer, subunit contacts between hexameric rings are more polar, less conserved, and include buried solvent. Twelve sulfate ions, which may mimic more physiologically relevant phosphate ions, are bound at the interface of hexamers. Each sulfate ion forms salt bridges with three strictly conserved residues from a subunit in one hexamer (His115, Arg137, and Arg138) and conserved Lys187 from a subunit of the opposing hexamer. Conservation of a dodecamer quaternary structure among
Referência(s)