Escherichia coli YrbI Is 3-Deoxy-d-manno-octulosonate 8-Phosphate Phosphatase
2003; Elsevier BV; Volume: 278; Issue: 20 Linguagem: Inglês
10.1074/jbc.m301983200
ISSN1083-351X
Autores Tópico(s)Glycosylation and Glycoproteins Research
Resumo3-Deoxy-d-manno-octulosonate 8-phosphate (KDO 8-P) phosphatase, which catalyzes the hydrolysis of KDO 8-P to KDO and inorganic phosphate, is the last enzyme in the KDO biosynthetic pathway for which the gene has not been identified. Wild-type KDO 8-P phosphatase was purified from Escherichia coli B, and the N-terminal amino acid sequence matched a hypothetical protein encoded by the E. coli open reading frame, yrbI. The yrbI gene, which encodes for a protein of 188 amino acids, was cloned, and the gene product was overexpressed in E. coli. The recombinant enzyme is a tetramer and requires a divalent metal cofactor for activity. Optimal enzymatic activity is observed at pH 5.5. The enzyme is highly specific for KDO 8-P with an apparent Km of 75 μm and a kcat of 175 s−1 in the presence of 1 mm Mg2+. Amino acid sequence analysis indicates that KDO 8-P phosphatase is a member of the haloacid dehalogenase hydrolase superfamily. 3-Deoxy-d-manno-octulosonate 8-phosphate (KDO 8-P) phosphatase, which catalyzes the hydrolysis of KDO 8-P to KDO and inorganic phosphate, is the last enzyme in the KDO biosynthetic pathway for which the gene has not been identified. Wild-type KDO 8-P phosphatase was purified from Escherichia coli B, and the N-terminal amino acid sequence matched a hypothetical protein encoded by the E. coli open reading frame, yrbI. The yrbI gene, which encodes for a protein of 188 amino acids, was cloned, and the gene product was overexpressed in E. coli. The recombinant enzyme is a tetramer and requires a divalent metal cofactor for activity. Optimal enzymatic activity is observed at pH 5.5. The enzyme is highly specific for KDO 8-P with an apparent Km of 75 μm and a kcat of 175 s−1 in the presence of 1 mm Mg2+. Amino acid sequence analysis indicates that KDO 8-P phosphatase is a member of the haloacid dehalogenase hydrolase superfamily. 3-deoxy-d-manno-octulosonate 3-deoxy-d-manno-octulosonate 8-phosphate purine nucleoside phosphorylase open reading frame molecular weight haloacid dehalogenase 2-(N-morpholino)ethanesulfonic acid 3-Deoxy-d-manno-octulosonate (KDO)1 is an 8-carbon sugar that links the lipid A and polysaccharide moieties of the lipopolysaccharide region in Gram-negative bacteria (1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3281) Google Scholar, 2Raetz C.R. Annu. Rev. Biochem. 1990; 59: 129-170Crossref PubMed Scopus (1030) Google Scholar). It has been demonstrated that an interruption in the biosynthesis of KDO leads to the accumulation of lipid A precursors and subsequent arrest in cell growth (3Rick P.D. Osborn M.J. Proc. Natl. Acad. Sci. U. S. A. 1972; 69: 3756-3760Crossref PubMed Scopus (65) Google Scholar, 4Rick P.D. Osborn M.J. J. Biol. Chem. 1977; 252: 4895-4903Abstract Full Text PDF PubMed Google Scholar, 5Rick P.D. Young D.A. J. Bacteriol. 1982; 150: 447-455Crossref PubMed Google Scholar). Thus, enzymes involved in KDO biosynthesis and/or its incorporation into lipid A are considered attractive targets for the design of novel antibiotics.The biosynthesis and utilization of KDO involves five sequential enzymatic reactions that are catalyzed byd-arabinose 5-phosphate isomerase, 3-deoxy-d-manno-octulosonate 8-phosphate (KDO 8-P) synthase, KDO 8-P phosphatase, cytidine 5′-monophosphate-KDO synthetase, and KDO transferase (Fig. 1) (2Raetz C.R. Annu. Rev. Biochem. 1990; 59: 129-170Crossref PubMed Scopus (1030) Google Scholar). During the past two decades, the genes responsible for the expression of KDO 8-P synthase (6Radaev S. Dastidar P. Patel M. Woodard R.W. Gatti D.L. J. Biol. Chem. 2000; 275: 9476-9484Abstract Full Text Full Text PDF PubMed Scopus (96) Google Scholar, 7Radaev S. Dastidar P. Patel M. Woodard R.W. Gatti D.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 2000; 56: 516-519Crossref PubMed Scopus (15) Google Scholar, 8Woisetschlager M. Hogenauer G. Mol. Gen. Genet. 1987; 207: 369-373Crossref PubMed Scopus (20) Google Scholar, 9Woisetschlager M. Hodl-Neuhofer A. Hogenauer G. J. Bacteriol. 1988; 170: 5382-5384Crossref PubMed Google Scholar), cytidine 5′-monophosphate-KDO synthetase (10Goldman R.C. Bolling T.J. Kohlbrenner W.E. Kim Y. Fox J.L. J. Biol. Chem. 1986; 261: 15831-15835Abstract Full Text PDF PubMed Google Scholar, 11Goldman R.C. Kohlbrenner W.E. J. Bacteriol. 1985; 163: 256-261Crossref PubMed Google Scholar, 12Jelakovic S. Jann K. Schulz G.E. FEBS Lett. 1996; 391: 157-161Crossref PubMed Scopus (27) Google Scholar), and KDO transferase (13Clementz T. Raetz C.R. J. Biol. Chem. 1991; 266: 9687-9696Abstract Full Text PDF PubMed Google Scholar, 14Belunis C.J. Clementz T. Carty S.M. Raetz C.R. J. Biol. Chem. 1995; 270: 27646-27652Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar) have been identified, and their respective enzymes have been studied extensively. More recently, the gene encoding the d-arabinose 5-phosphate isomerase (KpsF) from Neisseria meningitides was identified by Tzeng et al. (15Tzeng Y.L. Datta A. Strole C. Kolli V.S. Birck M.R. Taylor W.P. Carlson R.W. Woodard R.W. Stephens D.S. J. Biol. Chem. 2002; 277: 24103-24113Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar). KDO 8-P phosphatase, therefore, remains the last enzyme in the lipid A-KDO pathway for which a gene has not been assigned.KDO 8-P phosphatase catalyzes the hydrolysis of KDO 8-P to KDO and inorganic phosphate. Gahlambor and Heath (16Ghalambor M.A. Heath E.C. J. Biol. Chem. 1966; 241: 3222-3227Abstract Full Text PDF PubMed Google Scholar) first suggested the existence of a phosphatase for KDO 8-P in 1966. In 1975, Berger and Hammerschmid (17Berger H. Hammerschmid F. Biochem. Soc. Trans. 1975; 3: 1096-1097Crossref Scopus (11) Google Scholar) reported the isolation of a specific phosphatase fraction from a DEAE-cellulose column that would hydrolyze KDO 8-P but not d-arabinose 5-phosphate orp-nitrophenylphosphate. Ray and Benedict (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar) first purified and characterized the KDO 8-P phosphatase from Escherichia coli in 1980 but did not identify the encoding gene.In the present work, for the first time, the gene for KDO 8-P phosphatase is identified and cloned into an overexpression vector. A wild-type phosphatase that specifically hydrolyzes KDO 8-P to KDO and inorganic phosphate was isolated and N-terminally sequenced. A BLAST search of the E. coli K12 genome data base at the National Center for Biotechnology Information web site with this N-terminal sequence revealed a hypothetical protein encoded by the open reading frame (orf) yrbI. The yrbI gene was cloned, and the gene product was overexpressed in E. coli. The recombinant protein was purified to homogeneity, and its characteristics were consistent with those properties reported for the wild-type KDO 8-P phosphatase.DISCUSSIONKDO is an essential component of the lipopolysaccharide that is present in Gram-negative bacteria but absent in Gram-positive microorganisms. KDO 8-P phosphatase is the only enzyme in the KDO biosynthetic pathway for which the gene responsible for expression has not been identified. The YrbI protein from E. coli has been defined as KDO 8-P phosphatase based on the following findings: (i) the N-terminal sequence of the wild-type KDO 8-P phosphatase isolated in the present work corresponds to the yrbI orf to which no biological function had been previously assigned, (ii) a BLASTP search of the National Center for Biotechnology Information genomic data bases using the YrbI sequence identified matches to only hypothetical proteins from Gram-negative bacteria (there were no matches to any proteins from Gram-positive microorganisms), (iii) the recombinant gene product of yrbI has a low apparent Km of 75 μm for KDO 8-P and a high kcatof 175 s−1 for KDO 8-P hydrolysis, and (iv) no other phosphorylated monosaccharide or compound tested in the present study served as an alternate substrate.The substrate specificity, kinetic constants, divalent metal requirement, as well as the relatively low pH activity optimum of the recombinant KDO 8-P phosphatase (TableIV) are virtually identical to those reported for the wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). The pI values and the native MWs of recombinant KDO 8-P phosphatase and the previously reported wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar) are identical. A discrepancy in the denatured MWs was observed, however, between the wild-type and the recombinant protein. The calculated MW of YrbI (188 amino acids) is 19,866. The MW of the recombinant enzyme is 23 kDa as determined by SDS-PAGE and 19,881 by mass spectrometry. The apparent MW of the wild-type enzyme isolated in this report was 52 kDa as determined from two-dimensional gel, as opposed to 40–43 kDa as determined by Ray and Benedict from SDS-PAGE (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). There are several possible explanations for this discrepancy. One possibility could be that wild-type and recombinant KDO 8-P phosphatase differ in their intersubunit interactions because one is highly overexpressed while the other is expressed at physiological levels. These differences in quaternary structure between the wild-type enzyme and the recombinant enzyme may account for the variation seen in the denaturation states under the conditions used for SDS-PAGE analysis. Another possibility could be that the wild-type and recombinant KDO 8-P phosphatases actually differ in their amino acid sequence length/composition. The recombinant enzyme encoded by yrbI may simply comprise only the N-terminal segment of the wild-type KDO 8-P phosphatase.Table IVComparison of recombinant KDO 8-P phosphatase and wild-type KDO 8-P phosphatase as reported by Ray and Benedict (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar)PropertyRecombinant KDO 8-P phosphataseWild-type KDO 8-P phosphatase reported by Ray and BenedictKinetic constantsKm = 75 μmKm = 91 μmVmax = 500 units/mgVmax = 480 units/mgAlternate substratesNoNoMetal requirementYesYespH optimum5.5–7.05.5–6.5pI4.64.7MW (kDa)89 (gel filtration)80 (gel filtration)23 (SDS-PAGE)40–43 (SDS-PAGE)19.881 (mass spectrometry)NA4-aNA, not applicable.TetramerDimmer4-a NA, not applicable. Open table in a new tab A close examination of the yrb operon reveals that the termination codon of the yrbI orf overlaps the initiation codon of the downstream yrbK orf (ATGA). The yrbK orf encodes a 21-kDa protein (191 amino acids). The yrbI stop codon, TGA, has long been recognized as a dual signal for either termination or frame shifting (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar). Translational frame shifting has been observed in retroviruses, retrotransposons, bacterial insertion sequences, bacterial cellular genes, and eukaryotic genes (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar). Frame shifting events at some recoding sites yield two protein products from one orf or one fusion protein in which the N- and C-terminal regions are encoded by two overlapping orfs, respectively. In some instances, this process serves as a control mechanism for the expression of specific genes (27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar, 29Sekine Y. Eisaki N. Ohtsubo E. J. Mol. Biol. 1994; 235: 1406-1420Crossref PubMed Scopus (119) Google Scholar). It is therefore attractive to postulate that the wild-type KDO 8-P phosphatase is the product of a −1 translational frame shifting event that occurred at the overlapping sites of yrbI and yrbK. If this were the case, the cell might express three protein products in varying concentrations: the yrbI product (20 kDa), theyrbK product (21 kDa), and the yrbI-yrbKtransframe fusion protein (41 kDa). This frame shifting event, which would occur under certain physiological circumstances, could be a regulatory mechanism for the production and/or transport of KDO and, therefore, would serve as a control point in the biosynthesis of the lipopolysaccharide region of Gram-negative bacteria. It should be noted that the E. coli cells used for isolating the wild-type KDO 8-P phosphatases in both the present study and Ray and Benedict's study were grown in a glucose minimal medium supplemented with high levels of inorganic phosphate to repress the synthesis of alkaline phosphatase. This phosphate-rich growth condition may have induced the production of the putative high MW yrbI-yrbKtransframe protein. Further studies are under way to distinguish between these two possible explanations as well as other possible scenarios for the observed discrepancies in the denatured MWs.Based on the biochemical characteristics presented here, KDO 8-P phosphatase can be classified as a specific, low molecular weight acid phosphatase. Amino acid sequence homology analysis has also been utilized to classify phosphatase families (30Stukey J. Carman G.M. Protein Sci. 1997; 6: 469-472Crossref PubMed Scopus (221) Google Scholar, 31Thaller M.C. Schippa S. Rossolini G.M. Protein Sci. 1998; 7: 1647-1652Crossref PubMed Scopus (132) Google Scholar). Such analysis places KDO 8-P phosphatase into the haloacid dehalogenase (HAD) superfamily of hydrolases. This family is comprised of haloacid dehalogenases, epoxide hydrolases, ATPases, phosphomutases, and a variety of phosphatases, including phosphoserine phosphatase, phosphoglycolate phosphatase, sucrose-6F-phosphate phosphohydrolase, and trehalose-6-phosphatase (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 33Koonin E.V. Tatusov R.L. J. Mol. Biol. 1994; 244: 125-132Crossref PubMed Scopus (266) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 35Lunn J.E. Ashton A.R. Hatch M.D. Heldt H.W. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 12914-12919Crossref PubMed Scopus (63) Google Scholar). KDO 8-P phosphatase, the translated YrbI orf, shares the three highly conserved motifs generally observed in this superfamily of enzymes (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 36Wang W. Kim R. Jancarik J. Yokota H. Kim S.H. Structure. 2001; 9: 65-71Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar): motif I, DXDX(T/V); motif II, (S/T)XX; and motif III, K(G/S)(D/S)XXX(D/N) (Fig. 6). Although the overall sequence similarity between members of the HAD superfamily is generally low, a comparison of the structures of several members of the family demonstrates a conserved fold and suggests that enzymes in this superfamily most likely evolved from a common ancestor (37Selengut J.D. Biochemistry. 2001; 40: 12704-12711Crossref PubMed Scopus (63) Google Scholar). An additional signature sequence motif (GGXGAXRE), unique to the KDO 8-P phosphatase-like sequences identified in the gene data bank, is located in the C-terminal region (Fig. 6). Whether this signature sequence plays a role in structure and function unique to KDO 8-P phosphatases remains to be determined.Figure 6Alignment of KDO 8-P phosphatase-like sequences. The sequences were aligned using ClustalW (25Thompson J.D. Higgins D.G. Gibson T.J. Nucleic Acids Res. 1994; 22: 4673-4680Crossref PubMed Scopus (55190) Google Scholar). The GenBankTM accession numbers of the sequences are listed in Table II. Absolute conserved residues are marked with anasterisk. Residues that are homologous to conserved residues in motifs I, II, and III of the HAD superfamily are highlighted inblack. Conserved residues in the unique signature sequence of KDO 8-P phosphatase (C-terminal region) are highlighted ingray.View Large Image Figure ViewerDownload (PPT)During the completion of this study, Parsons et al. (38Parsons J.F. Lim K. Tempczyk A. Krajewski W. Eisenstein E. Herzberg O. Proteins. 2002; 46: 393-404Crossref PubMed Scopus (50) Google Scholar) published the crystal structure of the YrbI protein fromHaemophilus influenzae (HI1679, MWcal = 19,432) solved to 1.67-Å resolution. The H. influenzae YrbI was tetrameric, and the monomer subunits exhibited an α/β-hydrolase fold. The active site of each monomer was located at the subunit interface. The active site was formed mainly by the three conserved motifs characteristic of the HAD superfamily to which KDO 8-P phosphatase belongs. A cobalt ion, used for crystallization, was coordinated at each of the four active sites. Based on structural and sequence analysis as well as enzymatic assays, the authors tentatively assigned the function of the protein to be that of a small molecule phosphatase. Although the true physiological substrate of their phosphatase was not identified, the authors correctly predicted YrbI to be the sugar phosphatase. The H. influenzae and E. coli YrbI are 39% identical; thus, the present study not only confirms their prediction that the H. influenzae YrbI is a phosphatase but also suggests the substrate for their enzyme.In summary, the E. coli yrbI orf encodes for the protein KDO 8-P phosphatase. This is the first characterized gene in the yrb operon in E. coli and may help provide a clue to the function of other gene products in this operon. Further studies on this enzyme may provide useful information in the study of the evolution and structure/function of the entire HAD superfamily of enzymes. Additional experiments are in progress to better understand the discrepancies between the MWs of the recombinant and wild-type KDO 8-P phosphatases. Based on the importance of KDO in the lipopolysaccharide biosynthetic pathway, inhibition of KDO 8-P phosphatase will present an attractive target for the design of new generation antibiotics. The determination of the three-dimensional structure of the E. coli KDO 8-P phosphatase, now in progress, as well as the structure of various site-directed mutants, in the presence of substrate and/or substrate analogues, will further assist in the elucidation of the mechanism of KDO 8-P phosphatase catalysis, which will prove invaluable in the design of inhibitors. 3-Deoxy-d-manno-octulosonate (KDO)1 is an 8-carbon sugar that links the lipid A and polysaccharide moieties of the lipopolysaccharide region in Gram-negative bacteria (1Raetz C.R. Whitfield C. Annu. Rev. Biochem. 2002; 71: 635-700Crossref PubMed Scopus (3281) Google Scholar, 2Raetz C.R. Annu. Rev. Biochem. 1990; 59: 129-170Crossref PubMed Scopus (1030) Google Scholar). It has been demonstrated that an interruption in the biosynthesis of KDO leads to the accumulation of lipid A precursors and subsequent arrest in cell growth (3Rick P.D. Osborn M.J. Proc. Natl. Acad. Sci. U. S. A. 1972; 69: 3756-3760Crossref PubMed Scopus (65) Google Scholar, 4Rick P.D. Osborn M.J. J. Biol. Chem. 1977; 252: 4895-4903Abstract Full Text PDF PubMed Google Scholar, 5Rick P.D. Young D.A. J. Bacteriol. 1982; 150: 447-455Crossref PubMed Google Scholar). Thus, enzymes involved in KDO biosynthesis and/or its incorporation into lipid A are considered attractive targets for the design of novel antibiotics. The biosynthesis and utilization of KDO involves five sequential enzymatic reactions that are catalyzed byd-arabinose 5-phosphate isomerase, 3-deoxy-d-manno-octulosonate 8-phosphate (KDO 8-P) synthase, KDO 8-P phosphatase, cytidine 5′-monophosphate-KDO synthetase, and KDO transferase (Fig. 1) (2Raetz C.R. Annu. Rev. Biochem. 1990; 59: 129-170Crossref PubMed Scopus (1030) Google Scholar). During the past two decades, the genes responsible for the expression of KDO 8-P synthase (6Radaev S. Dastidar P. Patel M. Woodard R.W. Gatti D.L. J. Biol. Chem. 2000; 275: 9476-9484Abstract Full Text Full Text PDF PubMed Scopus (96) Google Scholar, 7Radaev S. Dastidar P. Patel M. Woodard R.W. Gatti D.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 2000; 56: 516-519Crossref PubMed Scopus (15) Google Scholar, 8Woisetschlager M. Hogenauer G. Mol. Gen. Genet. 1987; 207: 369-373Crossref PubMed Scopus (20) Google Scholar, 9Woisetschlager M. Hodl-Neuhofer A. Hogenauer G. J. Bacteriol. 1988; 170: 5382-5384Crossref PubMed Google Scholar), cytidine 5′-monophosphate-KDO synthetase (10Goldman R.C. Bolling T.J. Kohlbrenner W.E. Kim Y. Fox J.L. J. Biol. Chem. 1986; 261: 15831-15835Abstract Full Text PDF PubMed Google Scholar, 11Goldman R.C. Kohlbrenner W.E. J. Bacteriol. 1985; 163: 256-261Crossref PubMed Google Scholar, 12Jelakovic S. Jann K. Schulz G.E. FEBS Lett. 1996; 391: 157-161Crossref PubMed Scopus (27) Google Scholar), and KDO transferase (13Clementz T. Raetz C.R. J. Biol. Chem. 1991; 266: 9687-9696Abstract Full Text PDF PubMed Google Scholar, 14Belunis C.J. Clementz T. Carty S.M. Raetz C.R. J. Biol. Chem. 1995; 270: 27646-27652Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar) have been identified, and their respective enzymes have been studied extensively. More recently, the gene encoding the d-arabinose 5-phosphate isomerase (KpsF) from Neisseria meningitides was identified by Tzeng et al. (15Tzeng Y.L. Datta A. Strole C. Kolli V.S. Birck M.R. Taylor W.P. Carlson R.W. Woodard R.W. Stephens D.S. J. Biol. Chem. 2002; 277: 24103-24113Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar). KDO 8-P phosphatase, therefore, remains the last enzyme in the lipid A-KDO pathway for which a gene has not been assigned. KDO 8-P phosphatase catalyzes the hydrolysis of KDO 8-P to KDO and inorganic phosphate. Gahlambor and Heath (16Ghalambor M.A. Heath E.C. J. Biol. Chem. 1966; 241: 3222-3227Abstract Full Text PDF PubMed Google Scholar) first suggested the existence of a phosphatase for KDO 8-P in 1966. In 1975, Berger and Hammerschmid (17Berger H. Hammerschmid F. Biochem. Soc. Trans. 1975; 3: 1096-1097Crossref Scopus (11) Google Scholar) reported the isolation of a specific phosphatase fraction from a DEAE-cellulose column that would hydrolyze KDO 8-P but not d-arabinose 5-phosphate orp-nitrophenylphosphate. Ray and Benedict (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar) first purified and characterized the KDO 8-P phosphatase from Escherichia coli in 1980 but did not identify the encoding gene. In the present work, for the first time, the gene for KDO 8-P phosphatase is identified and cloned into an overexpression vector. A wild-type phosphatase that specifically hydrolyzes KDO 8-P to KDO and inorganic phosphate was isolated and N-terminally sequenced. A BLAST search of the E. coli K12 genome data base at the National Center for Biotechnology Information web site with this N-terminal sequence revealed a hypothetical protein encoded by the open reading frame (orf) yrbI. The yrbI gene was cloned, and the gene product was overexpressed in E. coli. The recombinant protein was purified to homogeneity, and its characteristics were consistent with those properties reported for the wild-type KDO 8-P phosphatase. DISCUSSIONKDO is an essential component of the lipopolysaccharide that is present in Gram-negative bacteria but absent in Gram-positive microorganisms. KDO 8-P phosphatase is the only enzyme in the KDO biosynthetic pathway for which the gene responsible for expression has not been identified. The YrbI protein from E. coli has been defined as KDO 8-P phosphatase based on the following findings: (i) the N-terminal sequence of the wild-type KDO 8-P phosphatase isolated in the present work corresponds to the yrbI orf to which no biological function had been previously assigned, (ii) a BLASTP search of the National Center for Biotechnology Information genomic data bases using the YrbI sequence identified matches to only hypothetical proteins from Gram-negative bacteria (there were no matches to any proteins from Gram-positive microorganisms), (iii) the recombinant gene product of yrbI has a low apparent Km of 75 μm for KDO 8-P and a high kcatof 175 s−1 for KDO 8-P hydrolysis, and (iv) no other phosphorylated monosaccharide or compound tested in the present study served as an alternate substrate.The substrate specificity, kinetic constants, divalent metal requirement, as well as the relatively low pH activity optimum of the recombinant KDO 8-P phosphatase (TableIV) are virtually identical to those reported for the wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). The pI values and the native MWs of recombinant KDO 8-P phosphatase and the previously reported wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar) are identical. A discrepancy in the denatured MWs was observed, however, between the wild-type and the recombinant protein. The calculated MW of YrbI (188 amino acids) is 19,866. The MW of the recombinant enzyme is 23 kDa as determined by SDS-PAGE and 19,881 by mass spectrometry. The apparent MW of the wild-type enzyme isolated in this report was 52 kDa as determined from two-dimensional gel, as opposed to 40–43 kDa as determined by Ray and Benedict from SDS-PAGE (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). There are several possible explanations for this discrepancy. One possibility could be that wild-type and recombinant KDO 8-P phosphatase differ in their intersubunit interactions because one is highly overexpressed while the other is expressed at physiological levels. These differences in quaternary structure between the wild-type enzyme and the recombinant enzyme may account for the variation seen in the denaturation states under the conditions used for SDS-PAGE analysis. Another possibility could be that the wild-type and recombinant KDO 8-P phosphatases actually differ in their amino acid sequence length/composition. The recombinant enzyme encoded by yrbI may simply comprise only the N-terminal segment of the wild-type KDO 8-P phosphatase.Table IVComparison of recombinant KDO 8-P phosphatase and wild-type KDO 8-P phosphatase as reported by Ray and Benedict (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar)PropertyRecombinant KDO 8-P phosphataseWild-type KDO 8-P phosphatase reported by Ray and BenedictKinetic constantsKm = 75 μmKm = 91 μmVmax = 500 units/mgVmax = 480 units/mgAlternate substratesNoNoMetal requirementYesYespH optimum5.5–7.05.5–6.5pI4.64.7MW (kDa)89 (gel filtration)80 (gel filtration)23 (SDS-PAGE)40–43 (SDS-PAGE)19.881 (mass spectrometry)NA4-aNA, not applicable.TetramerDimmer4-a NA, not applicable. Open table in a new tab A close examination of the yrb operon reveals that the termination codon of the yrbI orf overlaps the initiation codon of the downstream yrbK orf (ATGA). The yrbK orf encodes a 21-kDa protein (191 amino acids). The yrbI stop codon, TGA, has long been recognized as a dual signal for either termination or frame shifting (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar). Translational frame shifting has been observed in retroviruses, retrotransposons, bacterial insertion sequences, bacterial cellular genes, and eukaryotic genes (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar). Frame shifting events at some recoding sites yield two protein products from one orf or one fusion protein in which the N- and C-terminal regions are encoded by two overlapping orfs, respectively. In some instances, this process serves as a control mechanism for the expression of specific genes (27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar, 29Sekine Y. Eisaki N. Ohtsubo E. J. Mol. Biol. 1994; 235: 1406-1420Crossref PubMed Scopus (119) Google Scholar). It is therefore attractive to postulate that the wild-type KDO 8-P phosphatase is the product of a −1 translational frame shifting event that occurred at the overlapping sites of yrbI and yrbK. If this were the case, the cell might express three protein products in varying concentrations: the yrbI product (20 kDa), theyrbK product (21 kDa), and the yrbI-yrbKtransframe fusion protein (41 kDa). This frame shifting event, which would occur under certain physiological circumstances, could be a regulatory mechanism for the production and/or transport of KDO and, therefore, would serve as a control point in the biosynthesis of the lipopolysaccharide region of Gram-negative bacteria. It should be noted that the E. coli cells used for isolating the wild-type KDO 8-P phosphatases in both the present study and Ray and Benedict's study were grown in a glucose minimal medium supplemented with high levels of inorganic phosphate to repress the synthesis of alkaline phosphatase. This phosphate-rich growth condition may have induced the production of the putative high MW yrbI-yrbKtransframe protein. Further studies are under way to distinguish between these two possible explanations as well as other possible scenarios for the observed discrepancies in the denatured MWs.Based on the biochemical characteristics presented here, KDO 8-P phosphatase can be classified as a specific, low molecular weight acid phosphatase. Amino acid sequence homology analysis has also been utilized to classify phosphatase families (30Stukey J. Carman G.M. Protein Sci. 1997; 6: 469-472Crossref PubMed Scopus (221) Google Scholar, 31Thaller M.C. Schippa S. Rossolini G.M. Protein Sci. 1998; 7: 1647-1652Crossref PubMed Scopus (132) Google Scholar). Such analysis places KDO 8-P phosphatase into the haloacid dehalogenase (HAD) superfamily of hydrolases. This family is comprised of haloacid dehalogenases, epoxide hydrolases, ATPases, phosphomutases, and a variety of phosphatases, including phosphoserine phosphatase, phosphoglycolate phosphatase, sucrose-6F-phosphate phosphohydrolase, and trehalose-6-phosphatase (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 33Koonin E.V. Tatusov R.L. J. Mol. Biol. 1994; 244: 125-132Crossref PubMed Scopus (266) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 35Lunn J.E. Ashton A.R. Hatch M.D. Heldt H.W. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 12914-12919Crossref PubMed Scopus (63) Google Scholar). KDO 8-P phosphatase, the translated YrbI orf, shares the three highly conserved motifs generally observed in this superfamily of enzymes (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 36Wang W. Kim R. Jancarik J. Yokota H. Kim S.H. Structure. 2001; 9: 65-71Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar): motif I, DXDX(T/V); motif II, (S/T)XX; and motif III, K(G/S)(D/S)XXX(D/N) (Fig. 6). Although the overall sequence similarity between members of the HAD superfamily is generally low, a comparison of the structures of several members of the family demonstrates a conserved fold and suggests that enzymes in this superfamily most likely evolved from a common ancestor (37Selengut J.D. Biochemistry. 2001; 40: 12704-12711Crossref PubMed Scopus (63) Google Scholar). An additional signature sequence motif (GGXGAXRE), unique to the KDO 8-P phosphatase-like sequences identified in the gene data bank, is located in the C-terminal region (Fig. 6). Whether this signature sequence plays a role in structure and function unique to KDO 8-P phosphatases remains to be determined.During the completion of this study, Parsons et al. (38Parsons J.F. Lim K. Tempczyk A. Krajewski W. Eisenstein E. Herzberg O. Proteins. 2002; 46: 393-404Crossref PubMed Scopus (50) Google Scholar) published the crystal structure of the YrbI protein fromHaemophilus influenzae (HI1679, MWcal = 19,432) solved to 1.67-Å resolution. The H. influenzae YrbI was tetrameric, and the monomer subunits exhibited an α/β-hydrolase fold. The active site of each monomer was located at the subunit interface. The active site was formed mainly by the three conserved motifs characteristic of the HAD superfamily to which KDO 8-P phosphatase belongs. A cobalt ion, used for crystallization, was coordinated at each of the four active sites. Based on structural and sequence analysis as well as enzymatic assays, the authors tentatively assigned the function of the protein to be that of a small molecule phosphatase. Although the true physiological substrate of their phosphatase was not identified, the authors correctly predicted YrbI to be the sugar phosphatase. The H. influenzae and E. coli YrbI are 39% identical; thus, the present study not only confirms their prediction that the H. influenzae YrbI is a phosphatase but also suggests the substrate for their enzyme.In summary, the E. coli yrbI orf encodes for the protein KDO 8-P phosphatase. This is the first characterized gene in the yrb operon in E. coli and may help provide a clue to the function of other gene products in this operon. Further studies on this enzyme may provide useful information in the study of the evolution and structure/function of the entire HAD superfamily of enzymes. Additional experiments are in progress to better understand the discrepancies between the MWs of the recombinant and wild-type KDO 8-P phosphatases. Based on the importance of KDO in the lipopolysaccharide biosynthetic pathway, inhibition of KDO 8-P phosphatase will present an attractive target for the design of new generation antibiotics. The determination of the three-dimensional structure of the E. coli KDO 8-P phosphatase, now in progress, as well as the structure of various site-directed mutants, in the presence of substrate and/or substrate analogues, will further assist in the elucidation of the mechanism of KDO 8-P phosphatase catalysis, which will prove invaluable in the design of inhibitors. KDO is an essential component of the lipopolysaccharide that is present in Gram-negative bacteria but absent in Gram-positive microorganisms. KDO 8-P phosphatase is the only enzyme in the KDO biosynthetic pathway for which the gene responsible for expression has not been identified. The YrbI protein from E. coli has been defined as KDO 8-P phosphatase based on the following findings: (i) the N-terminal sequence of the wild-type KDO 8-P phosphatase isolated in the present work corresponds to the yrbI orf to which no biological function had been previously assigned, (ii) a BLASTP search of the National Center for Biotechnology Information genomic data bases using the YrbI sequence identified matches to only hypothetical proteins from Gram-negative bacteria (there were no matches to any proteins from Gram-positive microorganisms), (iii) the recombinant gene product of yrbI has a low apparent Km of 75 μm for KDO 8-P and a high kcatof 175 s−1 for KDO 8-P hydrolysis, and (iv) no other phosphorylated monosaccharide or compound tested in the present study served as an alternate substrate. The substrate specificity, kinetic constants, divalent metal requirement, as well as the relatively low pH activity optimum of the recombinant KDO 8-P phosphatase (TableIV) are virtually identical to those reported for the wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). The pI values and the native MWs of recombinant KDO 8-P phosphatase and the previously reported wild-type KDO 8-P phosphatase (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar) are identical. A discrepancy in the denatured MWs was observed, however, between the wild-type and the recombinant protein. The calculated MW of YrbI (188 amino acids) is 19,866. The MW of the recombinant enzyme is 23 kDa as determined by SDS-PAGE and 19,881 by mass spectrometry. The apparent MW of the wild-type enzyme isolated in this report was 52 kDa as determined from two-dimensional gel, as opposed to 40–43 kDa as determined by Ray and Benedict from SDS-PAGE (18Ray P.H. Benedict C.D. J. Bacteriol. 1980; 142: 60-68Crossref PubMed Google Scholar). There are several possible explanations for this discrepancy. One possibility could be that wild-type and recombinant KDO 8-P phosphatase differ in their intersubunit interactions because one is highly overexpressed while the other is expressed at physiological levels. These differences in quaternary structure between the wild-type enzyme and the recombinant enzyme may account for the variation seen in the denaturation states under the conditions used for SDS-PAGE analysis. Another possibility could be that the wild-type and recombinant KDO 8-P phosphatases actually differ in their amino acid sequence length/composition. The recombinant enzyme encoded by yrbI may simply comprise only the N-terminal segment of the wild-type KDO 8-P phosphatase. A close examination of the yrb operon reveals that the termination codon of the yrbI orf overlaps the initiation codon of the downstream yrbK orf (ATGA). The yrbK orf encodes a 21-kDa protein (191 amino acids). The yrbI stop codon, TGA, has long been recognized as a dual signal for either termination or frame shifting (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar). Translational frame shifting has been observed in retroviruses, retrotransposons, bacterial insertion sequences, bacterial cellular genes, and eukaryotic genes (26Baranov P.V. Gurvich O.L. Fayet O. Prere M.F. Miller W.A. Gesteland R.F. Atkins J.F. Giddings M.C. Nucleic Acids Res. 2001; 29: 264-267Crossref PubMed Scopus (46) Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar). Frame shifting events at some recoding sites yield two protein products from one orf or one fusion protein in which the N- and C-terminal regions are encoded by two overlapping orfs, respectively. In some instances, this process serves as a control mechanism for the expression of specific genes (27Tate W.P. Mansell J.B. Mannering S.A. Irvine J.H. Major L.L. Wilson D.N. Biochemistry. 1999; 64: 1342-1353PubMed Google Scholar, 28Engelberg-Kulka H. Schoulaker-Schwarz R. Mol. Microbiol. 1994; 11: 3-8Crossref PubMed Scopus (20) Google Scholar, 29Sekine Y. Eisaki N. Ohtsubo E. J. Mol. Biol. 1994; 235: 1406-1420Crossref PubMed Scopus (119) Google Scholar). It is therefore attractive to postulate that the wild-type KDO 8-P phosphatase is the product of a −1 translational frame shifting event that occurred at the overlapping sites of yrbI and yrbK. If this were the case, the cell might express three protein products in varying concentrations: the yrbI product (20 kDa), theyrbK product (21 kDa), and the yrbI-yrbKtransframe fusion protein (41 kDa). This frame shifting event, which would occur under certain physiological circumstances, could be a regulatory mechanism for the production and/or transport of KDO and, therefore, would serve as a control point in the biosynthesis of the lipopolysaccharide region of Gram-negative bacteria. It should be noted that the E. coli cells used for isolating the wild-type KDO 8-P phosphatases in both the present study and Ray and Benedict's study were grown in a glucose minimal medium supplemented with high levels of inorganic phosphate to repress the synthesis of alkaline phosphatase. This phosphate-rich growth condition may have induced the production of the putative high MW yrbI-yrbKtransframe protein. Further studies are under way to distinguish between these two possible explanations as well as other possible scenarios for the observed discrepancies in the denatured MWs. Based on the biochemical characteristics presented here, KDO 8-P phosphatase can be classified as a specific, low molecular weight acid phosphatase. Amino acid sequence homology analysis has also been utilized to classify phosphatase families (30Stukey J. Carman G.M. Protein Sci. 1997; 6: 469-472Crossref PubMed Scopus (221) Google Scholar, 31Thaller M.C. Schippa S. Rossolini G.M. Protein Sci. 1998; 7: 1647-1652Crossref PubMed Scopus (132) Google Scholar). Such analysis places KDO 8-P phosphatase into the haloacid dehalogenase (HAD) superfamily of hydrolases. This family is comprised of haloacid dehalogenases, epoxide hydrolases, ATPases, phosphomutases, and a variety of phosphatases, including phosphoserine phosphatase, phosphoglycolate phosphatase, sucrose-6F-phosphate phosphohydrolase, and trehalose-6-phosphatase (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 33Koonin E.V. Tatusov R.L. J. Mol. Biol. 1994; 244: 125-132Crossref PubMed Scopus (266) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 35Lunn J.E. Ashton A.R. Hatch M.D. Heldt H.W. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 12914-12919Crossref PubMed Scopus (63) Google Scholar). KDO 8-P phosphatase, the translated YrbI orf, shares the three highly conserved motifs generally observed in this superfamily of enzymes (32Aravind L. Galperin M.Y. Koonin E.V. Trends Biochem. Sci. 1998; 23: 127-129Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 34Ridder I.S. Dijkstra B.W. Biochem. J. 1999; 339: 223-226Crossref PubMed Scopus (105) Google Scholar, 36Wang W. Kim R. Jancarik J. Yokota H. Kim S.H. Structure. 2001; 9: 65-71Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar): motif I, DXDX(T/V); motif II, (S/T)XX; and motif III, K(G/S)(D/S)XXX(D/N) (Fig. 6). Although the overall sequence similarity between members of the HAD superfamily is generally low, a comparison of the structures of several members of the family demonstrates a conserved fold and suggests that enzymes in this superfamily most likely evolved from a common ancestor (37Selengut J.D. Biochemistry. 2001; 40: 12704-12711Crossref PubMed Scopus (63) Google Scholar). An additional signature sequence motif (GGXGAXRE), unique to the KDO 8-P phosphatase-like sequences identified in the gene data bank, is located in the C-terminal region (Fig. 6). Whether this signature sequence plays a role in structure and function unique to KDO 8-P phosphatases remains to be determined. During the completion of this study, Parsons et al. (38Parsons J.F. Lim K. Tempczyk A. Krajewski W. Eisenstein E. Herzberg O. Proteins. 2002; 46: 393-404Crossref PubMed Scopus (50) Google Scholar) published the crystal structure of the YrbI protein fromHaemophilus influenzae (HI1679, MWcal = 19,432) solved to 1.67-Å resolution. The H. influenzae YrbI was tetrameric, and the monomer subunits exhibited an α/β-hydrolase fold. The active site of each monomer was located at the subunit interface. The active site was formed mainly by the three conserved motifs characteristic of the HAD superfamily to which KDO 8-P phosphatase belongs. A cobalt ion, used for crystallization, was coordinated at each of the four active sites. Based on structural and sequence analysis as well as enzymatic assays, the authors tentatively assigned the function of the protein to be that of a small molecule phosphatase. Although the true physiological substrate of their phosphatase was not identified, the authors correctly predicted YrbI to be the sugar phosphatase. The H. influenzae and E. coli YrbI are 39% identical; thus, the present study not only confirms their prediction that the H. influenzae YrbI is a phosphatase but also suggests the substrate for their enzyme. In summary, the E. coli yrbI orf encodes for the protein KDO 8-P phosphatase. This is the first characterized gene in the yrb operon in E. coli and may help provide a clue to the function of other gene products in this operon. Further studies on this enzyme may provide useful information in the study of the evolution and structure/function of the entire HAD superfamily of enzymes. Additional experiments are in progress to better understand the discrepancies between the MWs of the recombinant and wild-type KDO 8-P phosphatases. Based on the importance of KDO in the lipopolysaccharide biosynthetic pathway, inhibition of KDO 8-P phosphatase will present an attractive target for the design of new generation antibiotics. The determination of the three-dimensional structure of the E. coli KDO 8-P phosphatase, now in progress, as well as the structure of various site-directed mutants, in the presence of substrate and/or substrate analogues, will further assist in the elucidation of the mechanism of KDO 8-P phosphatase catalysis, which will prove invaluable in the design of inhibitors. We thank Dr. George A. Garcia for kindly providing E. coli BL21 genomic DNA, Dr. Michael Bly for performing two-dimensional gel electrophoresis, and Sherry Williams for performing N-terminal amino acid sequencing. We also thank other members of the Woodard group for helpful discussions.
Referência(s)