JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M501849200 on April 27, 2005

J. Biol. Chem., Vol. 280, Issue 27, 25533-25540, July 8, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/27/25533    most recent
M501849200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Briozzo, P.
Right arrow Articles by Gilles, A.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Briozzo, P.
Right arrow Articles by Gilles, A.-M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Structure of Escherichia coli UMP Kinase Differs from That of Other Nucleoside Monophosphate Kinases and Sheds New Light on Enzyme Regulation*

Pierre Briozzo{ddagger}§, Cécile Evrin¶, Philippe Meyer||**, Liliane Assairi{ddagger}{ddagger}, Nathalie Joly{ddagger}, Octavian Barzu¶, and Anne-Marie Gilles¶

From the {ddagger}Unité de Chimie Biologique, UMR 206 Institut National de la Recherche Agronomique, Institut National Agronomique Paris-Grignon, 78850 Thiverval-Grignon, the Unité de Génétique des Génomes Bactériens, URA 2171 CNRS, Institut Pasteur, 75724 Paris Cedex 15, the ||Laboratoire d'Enzymologie et de Biochimie Structurales, UPR 9063 du CNRS, 91198 Gif-sur-Yvette Cedex, and the{ddagger}{ddagger} Institut Curie, Centre Universitaire Paris-Sud, 91405 Orsay, France

Received for publication, February 18, 2005 , and in revised form, March 30, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Bacterial UMP kinases are essential enzymes involved in the multistep synthesis of nucleoside triphosphates. They are hexamers regulated by the allosteric activator GTP and inhibited by UTP. We solved the crystal structure of Escherichia coli UMP kinase bound to the UMP substrate (2.3 Å resolution), the UDP product (2.6 Å), or UTP (2.45 Å). The monomer fold, unrelated to that of other nucleoside monophosphate kinases, belongs to the carbamate kinase-like superfamily. However, the phosphate acceptor binding cleft and subunit assembly are characteristic of UMP kinase. Interactions with UMP explain the high specificity for this natural substrate. UTP, previously described as an allosteric inhibitor, was unexpectedly found in the phosphate acceptor site, suggesting that it acts as a competitive inhibitor. Site-directed mutagenesis of residues Thr-138 and Asn-140, involved in both uracil recognition and active site interaction within the hexamer, decreased the activation by GTP and inhibition by UTP. These experiments suggest a cross-talk mechanism between enzyme subunits involved in cooperative binding at the phosphate acceptor site and in allosteric regulation by GTP. As bacterial UMP kinases have no counterpart in eukaryotes, the information provided here could help the design of new antibiotics.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Nucleoside monophosphate kinases are key enzymes in the metabolism of nucleotides. They represent a relatively homogeneous family of catalysts (1) thought to derive from a common ancestor. They catalyze the reversible transfer of the {gamma}-phosphoryl group from a nucleoside triphosphate, generally ATP, to a particular nucleoside monophosphate. The resulting nucleoside diphosphates will be further phosphorylated (and eventually reduced) to produce nucleoside triphosphates, precursors of the major biological molecules RNA, DNA, and phospholipids. Eukaryotic UMP/CMP kinases represent an exception to the otherwise generally observed specificity of NMP1 kinases for the base moiety of the phosphate acceptor nucleotide; they phosphorylate with comparable efficiency both UMP and CMP. Conversely, bacteria possess two distinct enzymes, specific to UMP or CMP. Bacterial CMP kinases are monomers, like most NMP kinases, and their overall fold is similar to that of other enzymes of the latter family (2). In contrast, bacterial UMP kinases are hexamers (3). They specifically phosphorylate UMP according to the scheme: UMP + Mg2+·ATP {leftrightarrow} UDP + Mg2+·ADP. They are activated by GTP and inhibited by UTP. This contributes to equilibrating the synthesis of purine versus pyrimidine nucleoside triphosphates.

Genes coding for UMP kinases have been identified in all bacterial genomes investigated to date. They have no closely related counterpart in eukaryotes and have proved to be essential for growth in both Gram-negative species (Escherichia coli (4, 5)) and some Gram-positive species (Streptococcus pneumoniae (6)). They code for enzymes that have strong sequence similarity. Therefore, bacterial UMP kinases represent not only an interesting model of activity regulation but also valuable targets for antibacterial drugs.

UMP kinase from E. coli (UMPKeco) has been the most studied enzyme of the family. Overall, its regulation seems to be complex. GTP and UTP act primarily as effectors but also as weak phosphoryl donors; although the effects of these two nucleotides are widely described as allosteric, Mg2+, an obligatory actor in the phosphoryl transfer reaction, dramatically decreases the inhibition by UTP, with no apparent effect on the activation by GTP. Nevertheless, based on kinetic and fluorescence results, it has been suggested that both nucleoside triphosphates bind to the same allosteric site (7).

A three-dimensional structure was needed for a better understanding of enzyme activity regulation. However, UMP kinase from E. coli exhibits low solubility (~0.1 mg·ml–1), and many others have a low stability. This precluded for several years the determination of their crystal structure. Very recently, the UMP kinase from the hyperthermophilic archaeon Pyrococcus furiosus was crystallized, but its structure could not be solved (8). We previously obtained by site-directed mutagenesis a variant of UMPKeco (D159N) similar to the wild-type enzyme in its stability and kinetic properties but significantly more soluble at neutral pH. Its solubility increases with pH and in the presence of magnesium-free UTP (7). Altogether, these properties allowed us to get crystals of UTP-complexed UMPKeco suitable for x-ray crystallography. To further investigate the phosphorylation process, we then solved the structures with the natural substrate UMP and with the reaction product UDP. In addition, we tested the role of 2 residues, which are involved in intersubunit active site contacts according to crystal structure analysis, by site-directed mutagenesis.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals—Nucleotides, restriction enzymes, T4 DNA ligase, VentDNA polymerase, Tfu DNA polymerase, and coupling enzymes were from Roche Applied Science, Sigma, or Qbiogene. Nucleoside diphosphate kinase from Dictyostelium discoideum (2000 units/mg of protein) was kindly provided by I. Lascu.

Bacterial Strains, Plasmids, Growth Conditions, and DNA Manipulations—General DNA manipulations were performed as described by Sambrook et al. (9). The sequence coding for UMPKeco was amplified by the PCR method using the plasmid harboring the gene with the D159N mutation (7) as template, the Vent DNA polymerase, the dNTPs, and the following primers: 5'-UMPK, 5'-GGAATTCCATATGGCTACCAATGCAAAACCCGT-3', and 3'-UMPK, 5'-CGGCGCTCGAGTTATTCCGTGATTAAAGTCCCTTCT-3'. The PCR products were cloned at the NdeI and XhoI restriction sites of the vector pET28a. The resulting plasmid pLA 2.1.1 harboring the gene coding for UMP kinase D159N with a His tag was introduced into the E. coli strain BL21(DE3)pDIA17 (10). The transformants were grown in the 2YT medium containing kanamycin (70 µg/ml) and chloramphenicol (30 µg/ml) at 37 °C. When the optical density reached at least 1.5 at 600 nm, the expression of the recombinant proteins was induced by the addition of 1 mM isopropyl-{beta}-D-thiogalactopyranoside, and the growth was continued for an additional 3 h at 37 °C. The cells were then pelleted by centrifugation and served as source for protein purification.

The double mutants D159N,T138A and D159N,N140A and the triple mutant D159N,T138A,N140A were constructed by the one-tube PCR-based mutagenesis method (11) using the plasmid pLA2.1.1 harboring the UMP kinase gene as template, Tfu DNA polymerase, the dNTPs, and the following mutagenic oligonucleotides: 3'UMPK N140A, 5'-GGTGGTAAAGAACGGGGCACCTGTACCGGCGGA-3'; 3'-UMPK T138A, 5'-AAAGAACGGGTTACCGGCACCGGCGGAGAGGAT-3'; 3'-UMPK T138A N140A, 5'-AAAGAACGGGGCACCTGCACCGGCGGAGAGGAT-3'. The PCR product was cloned at the NdeI and XhoI restriction sites of the vector pET28a, giving respectively the plasmids pLA2.1.2, pLA2.1.3, and pLA2.1.4.

Purification of UMP Kinase and Activity Assays—The different N-terminal His-tagged variants of E. coli UMP kinase overproduced in the same bacterial strain were purified by nickel-nitrilotriacetic acid affinity chromatography using the Qiagen express system (12) after sonication in a buffer containing 3 mM UTP. The UMP kinase activity was determined at 30 °C (0.5 ml final volume) using a coupled spectrophotometric assay (13). The reaction medium contained 50 mM Tris-HCl, pH 7.4, 2 mM MgCl2, 1 mM phosphoenolpyruvate, 0.2 mM NADH, 1 mM ATP, and 2 units each of lactate dehydrogenase, pyruvate kinase and nucleoside diphosphate kinase. The pure or crude preparation of UMP kinase was then added, followed 2 min later by 0.3 mM UMP. The decrease in absorbance at 340 nm was then recorded and corrected for secondary reactions occurring in the absence of UMP. One unit of UMP kinase corresponds to 1 µmol of product formed per min. Protein concentration was measured according to Bradford (14).

For production of the selenomethionine derivative, selenomethionine was incorporated in place of methionine, whereas the cells were growing in M9 minimum medium containing kanamycin (70 µg/ml) and chloramphenicol (30 µg/ml) at 37 °C. When the optical density at 600 nm reached 0.5, the medium was supplemented with the amino acids lysine, threonine, and phenylalanine at 100 mg/ml and leucine, isoleucine, valine, and selenomethionine at 50 mg/ml. After 30 min, the expression was induced by the addition of 0.5 mM isopropyl-{beta}-D-thiogalactopyranoside for 16 h.

Ion spray mass spectra were used to check the variant proteins and the selenomethionine derivative. They were recorded on a quadrupole mass spectrometer API-365 (PerkinElmer Life Sciences) equipped with an ion spray (nebulizer-assisted electrospray) source.

Crystallization and Data Collection—Three types of crystalline complexes were studied using the histidine-tagged D159N variant: with UMP, with UDP, or with UTP. Crystals were grown at 20 °C in a 50 mM Tris-HCl buffer, pH 8.5, by the vapor diffusion method. The 6-µl hanging drop contained 3 mg/ml UMPKeco, the UXP ligand (5 mM for UMP and UTP, 50 mM for UDP), 0.3 M potassium sodium tartrate, 0.25% (w/v) n-octyl-{beta}-D-glucoside, and (with UMP or UTP) 5% (v/v) 2-methyl-2,4-pentane diol. For the UTP-containing drops, the selenomethionine derivative of the enzyme was used in the presence of 5 mM dithiothreitol. The drop contained ADPNP in the case of UMP-(25 mM, in the presence of 25 mM MgCl2) and UTP-containing complexes (5 mM), but this ATP analogue was not seen later in the corresponding electron density maps. Drops were equilibrated with a reservoir solution containing the precipitant sodium/potassium tartrate (1.7 M with UMP, 1.3 M with UDP, 1.1 M with UTP). Good quality crystals only appeared when the crystallization drop was in contact with the vacuum grease otherwise used for airtightness (with the exception of a unique UDP crystal). They were allowed to grow for several weeks before data collection. Prior to data collection, they were transferred to a cryoprotectant solution (consisting in reservoir solution supplemented with 35% (v/v) glycerol) during 3 min and then frozen in liquid ethane. Data were collected at 100 K on the European Synchrotron Radiation Facility (ESRF) in Grenoble, France, beamline BM30A, for crystals with UDP or UTP and beamline ID14-2 for the crystal with UMP. With UTP, the absorption edge of selenium was determined from fluorescence emission scans, and a data set was collected at the peak wavelength of selenium. All crystal systems are rhombohedral, space group R3, with two molecules per asymmetric unit. Data were processed with DENZO and scaled and reduced with SCALEPACK (15).

Structure Solution, Model Building, and Refinement—The first structure solved was that with UTP, using the single-wavelength anomalous dispersion method. The data were used in a single-wavelength anomalous dispersion routine implemented in the SOLVE/RESOLVE program suite version 2.03. SOLVE (16) was used to find the selenium sites, taking into account the calculated solvent content of 45%. This led to the location of 12 of the 22 expected selenium atoms. Then statistical density modification and automated model building was done using RESOLVE (17), and refinement of the built model was done using REFMAC (18). This allowed constructing 116 of the 241 UMPKeco residues for molecule A and 121 for molecule B. Other residues were then hand-constructed. The structures of UMP- and UDP-UMPKeco complexes were solved by molecular replacement at 3.5 Å resolution with AMoRe (19), using only the protein part of the UTP-UMPKeco model. In both cases, a clear density immediately appeared for the nucleotide.

Models were built using TURBO (20). CNS (21) version 1.1 was used for refinement, which was monitored using a free R factor. In all cases, simulated annealing was initially used, as well as non-crystallographic constraints. Then the two molecules of the asymmetric unit were refined as different models, using individual B factors. In the final steps, we placed water molecules in residual density above 2.5 standard deviations. The final models contain two polypeptide chains A and B corresponding respectively to each of the two molecules from the asymmetric unit. A few residues (2–4, depending on the ligand and the chain) at the N-terminal end and residue 26 from molecule B were missing. For the UMP-containing complex, this is also the case of residues 176–178 in molecule A and for the UTP-containing complex of residue 112. Models were superposed with the procedure implemented in O (22) with defaults, and the relevant root mean square deviations for C{alpha} atoms were calculated using O or DALI. The Protein Data Bank codes of the E. coli UMPK complexes are 2BNE with UMP, 2BND with UDP, and 2BNF with UTP.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Overall Fold and Quaternary Structure—The selenomethionine-substituted D159N variant of UMPKeco in complex with UTP crystallized in the rhombohedral system, space group R3, with two molecules per asymmetric unit. The structure was solved at 2.45 Å resolution in a single-wavelength anomalous dispersion experiment. The crystals of the enzyme complexed to UMP (2.3 Å resolution) or to UDP (2.6 Å resolution) were obtained in the same space group. Corresponding structures were solved by molecular replacement using the UTP-UMP-Keco model. All three structures are very similar (root mean square deviation (r.m.s.d.) for all C{alpha} atoms of 0.48 Å when comparing the UMP structure with that of UDP and of 0.76 Å when comparing the UMP structure with that of UTP), and their stereochemistry is good as indicated by Ramachandran plots. Information on data collection, processing, refinement, and model statistics is given in Table I.


View this table:
[in this window]
[in a new window]
  		TABLE I
Structural data

 
As expected, the enzyme is a hexamer. The UMPKeco subunit has an open {alpha}/{beta}-fold (Figs. 1 and 2A) composed of nine {beta}-strands and eight {alpha}-helices. All the {beta}-strands constitute a unique, mainly parallel {beta}-sheet; at increasing distances from the {alpha}3 helix (pink), strands 4, 3, 5, 2, 1, 6, and 8 are parallel. This {beta}-sheet is sandwiched between two layers of respectively three ({alpha}1, {alpha}3, {alpha}8) and four ({alpha}4, {alpha}5, {alpha}6, {alpha}7) {alpha}-helices. The fold also contains the external helix {alpha}2, and a 3-10 helix, {eta}1.

The largest buried surface between two neighbor monomers of the hexamer is found between the two subunits (named A and B) of an asymmetric unit, which are related by a non-crystallographic two-fold axis (Fig. 2B). The global structure of these two subunits is nearly identical; r.m.s.d. between A and B is 0.81 Å on 233 equivalent C{alpha} atoms (UMP-UMPKeco model). Slightly different positions are observed for the {alpha}6 and {alpha}7 helices, for the loop connecting them, and also for the loop between {beta}6 and {beta}7. The dimer formation buries a surface of ~2500 Å2 (taking into account the sum of the buried surfaces of the two monomers), corresponding for each subunit to ~11% of its total surface. The interface is mostly hydrophobic but also involves two hydrogen bonds connecting the head-and-tail {alpha}3 helices through residues Asn-72 and Asp-93. This helix is by far the longest of the molecule (25 residues) and forms most of the dimer interface, the remaining part of which is mainly represented by the N-terminal side of {alpha}1 and the C-terminal side of {alpha}2.

The three-fold axis of the rhombohedral crystal system applied to the dimer generates a hexamer with a central channel. The interactions between two neighbor dimers are less extended (buried surface ~1430 Å2, corresponding to ~6.5% of the surface) than those between two subunits of a dimer. They mostly involve hydrophobic interactions (mainly with residues from {alpha}3, {alpha}5, or {alpha}7), but also hydrogen bonds; the main-chain oxygen of Thr-138 from one subunit is H-bonded with the side-chain nitrogen of Asn-140 from the neighbor subunit (Fig. 2D). These 2 residues also interact with the base moiety of UMP. As this creates a cross-talk path between the active sites of two neighbor dimers, we call the loop (Fig. 2A, green) containing these 2 residues the cross-talk loop.

Comparison with Other Proteins of Known Three-dimensional Structure—The fold of NMP kinases, including bacterial CMP kinases, contains three structurally distinct domains. The {alpha}/{beta}-CORE domain, a parallel {beta}-sheet surrounded by connecting {alpha}-helices, is used as a rigid platform around which the LID and NMP-binding domains move in an induced-fit mechanism, closing upon the phosphate donor and acceptor nucleotides, respectively. This allows the phosphoryl transfer reaction to proceed (23). The majority of NMP kinases are monomers, with a molecular mass around 25 kDa. The UMP-Keco subunit molecular mass (25.8 kDa) is close to that of monomeric NMP kinases, but its fold is not related to any other member of that family of enzymes. A search for structural homologues of the UMPKeco subunit in the Protein Data Bank using the program DALI (24) confirmed the absence of similarity with other NMP kinases. It returned two proteins of the carbamate kinase-like superfamily (25), with a Z-score of 19 (other proposed proteins had significantly lower Z-scores: less than 8). Both are dimeric kinases. The first one is the carbamate kinase from the extremophile P. furiosus (CBMK, Protein Data Bank access codes 1b7b [PDB] and 1e19), with an r.m.s.d. of 4.0 Å over 212 equivalent {alpha}-carbons. Many of the secondary structure elements are equivalent, but there are striking differences. When compared with UMPKeco (Fig. 1), carbamate kinase has two additional insertions: a very long one of 50 residues (between {beta}4 and {alpha}4 from UMPKeco), including three antiparallel {beta}-strands and an {alpha}-helix (26), and a shorter one of 9 residues after Asp-146, forming a {beta}-strand. The second structurally homologous protein is the N-acetyl glutamate kinase from E. coli (NAGK, Protein Data Bank code 1gs5 [PDB] ). Its structure is more similar to that of UMPKeco (r.m.s.d. 2.6 Å on 202 C{alpha} atoms). However, there are important differences; NAGK contains three {beta}-hairpins (Fig. 3, orange) close to the phosphate acceptor that have no equivalent in UMPKeco. As a result, whereas both enzymes have the same number of {alpha}-helices, the former one contains almost twice as many {beta}-strands (16 (27), as opposed to only 9 for UMPKeco).



View larger version (74K):
[in this window]
[in a new window]
  		FIG. 1.
Sequence alignment of UMPKeco with other kinases. Alignment with the amino acid sequences of bacterial UMP kinases from Haemophilus influenzae (Hi) and B. subtilis (Bs), the N-acetyl glutamate kinase from E. coli (NAGK-Ec), and the carbamate kinase from P. furiosus (CBMK-Pf) is shown. A column is framed in red when residues are identical and boxed when at least 4 of its residues (red) are homologues. Correspondence between the amino acid sequence and the secondary structure defined by DSSP (36) is given on top for UMPKeco; helices are shown in blue (except the longest one {alpha}3, emphasized in pink), and {beta}-strands are shown in orange. Residues interacting with UMP are indicated by a magenta asterisk at the bottom, those that in addition interact with UDP and UTP are indicated by green triangles, and those from the cross-talk loop are indicated in green. The figure was drawn with ESPript (37).

 
Although there is a clear parenthood between NAGK and UMPKeco monomers, their quaternary structure assembly is fairly different. In the dimer of UMPKeco observed in the asymmetric unit, the axes of the two long {alpha}3 helices (Fig. 2B, pink) are parallel. Conversely, in the NAGK dimer, the axis of the longest helix situated at the interface and equivalent to {alpha}3 (Fig. 3B, pink) makes an angle of ~65° with that of its counterpart in the neighbor monomer. It is worth noting that the CBMK dimer is arranged very much like that of NAGK. The two latter kinases have therefore close quaternary structures that contrast with that of UMP kinase.

Binding of Ligands in the Phosphate Acceptor Cavity: The UMP Substrate, the Reaction Product UDP, or the Magnesium-free UTP Inhibitor—The interactions between E. coli UMP kinase and UMP are shown on Fig. 4A. Those with uracil involve 2 residues from the cross-talk loop: Thr-138, which makes a bidentate interaction with the carbonyl 04 oxygen, and Asn-140 which is H-bonded to Asn-3 from the uracil ring. The 2'OH from the ribose makes hydrogen bonds with the main-chain nitrogen of Gly-63 and with the side-chain oxygen of Asp-77. All three terminal oxygens of the {alpha}-phosphate interact with enzyme residues: with a terminal nitrogen of Arg-62 side chain; with both the side-chain oxygen and the main-chain nitrogen of Thr-145; and with the glycines 57 and 58. Thus, as is the case for other NMP kinases, several conserved amino acid residues interact with the phosphate acceptor to allow the simultaneous recognition of its base, sugar, and phosphate moieties.

The interactions between the enzyme and the uracil, the ribose, or the {alpha}-phosphate from UDP (not shown) are very similar to those with UMP. Gly-58 binds the {beta}-phosphate instead of the {alpha}-phosphate with the nucleoside monophosphate. The other interactions with the {beta}-phosphate (Fig. 5A) involve the side chain of Arg-62, and in addition, 3 residues that are very close in the sequence: Lys-15, a residue essential for catalytic activity2 through its terminal nitrogen; Ser-17 (only for one molecule of the asymmetric unit) through its side-chain oxygen; and Gly-18.



View larger version (57K):
[in this window]
[in a new window]
  		FIG. 2.
Overall fold and quaternary structure. A, ribbon representation of the UMPKeco monomer fold (using the UMP-containing complex). {beta}-strands (yellow) and helices (blue, except {alpha}3, pink) are numbered. A stick model of UMP is shown in magenta. The cross-talk loop (green) is labeled CT. The loops are smoothed for clarity. A gray line connects residues that delimit a segment in which no clear density can be seen. B, the dimer constituted by the two molecules of an asymmetric unit. View along the non-crystallographic two-fold axis (indicated by a black ellipse symbol). The blue subunit orientation is close to that in panel A. C, ribbon representation of the hexamer viewed along the three-fold crystallographic axis (indicated by a black triangle). A particular color is used for each subunit. For the blue and green dimer, {alpha}3 helices are pink. Corey-Pauling-Koltun space-filling models of UMP are magenta. The three non-crystallographic two-fold axes are shown by dotted lines; they are perpendicular to the three-fold axis. D, magnification of the dimer-dimer interface emphasizing the 2 residues (shown in sticks and labeled) from the cross-talk loop that interact both with their homologues from the facing dimer and with UMP (stick model with carbon atoms in magenta). Hydrogen bonds are shown as red dots. {alpha}-helices are transparized for clarity. The figure was drawn with PyMOL, Version 0.97 (38).

 
The UTP molecule is bound in the phosphate acceptor site, similar to UMP and UDP except for slight differences in the nucleotide conformation. Thus, the nucleoside triphosphate inhibitor binds in the active site and not in an allosteric site presumably distant from it, as was suggested by previous publications (3, 6, 7). Overall, the enzyme-UTP interactions (Fig. 4B) are very similar to those observed in the UMPKeco-UDP complex. A noticeable difference is that for one molecule of the asymmetric unit, the side chain of Thr-144 is H-bonded to the {alpha}-phosphate. The residues interacting with the {gamma}-phosphate are Lys-15, (which also binds the {beta}-phosphate), and Ser-17.



View larger version (53K):
[in this window]
[in a new window]
  		FIG. 3.
Comparison with NAGK structure. A, superposition of the ribbon representations of the monomers of UMPKeco (blue, with a stick model of UMP in cyan) and NAGK (yellow, with N-acetyl glutamine and ADPNP in red; the helix homologous to {alpha}3 is pink). The cross-talk loop of UMPKeco is green, the flexible loops close to ADPNP are magenta, and the extra {beta}-hairpins of NAGK are orange. The orientation is close to that in Fig. 2A but slightly modified to better see all ligands. B, the dimer of NAGK. The orientation is slightly different from that in Fig. 2B to give a better view of the interface.

 
Substitution by Site-directed Mutagenesis of Thr-138 or Asn-140 Affects the Stability and Kinetic Properties of UMPKeco— Structure analysis of E. coli UMP kinase suggests that the 2 residues Thr-138 and Asn-140, conserved among bacterial UMP kinases, are involved in cross-talk between two adjacent active sites in the hexameric structure. We therefore expected that substituting these residues by a hydrophobic side chain would disrupt the communication between active sites in the oligomer and consequently disturb some of the regulatory properties and kinetic parameters. As shown in Table II, individual substitution of either of these two amino acids by an alanine decreases the half-denaturation temperature (Tm) of UMP kinase by around 10 °C. Substituting Thr-138, the side chain of which is H-bonded to uracil, results in a four times higher Km for UMP. In contrast, this Km is not altered by the mutation of Asn-140, a residue that only binds uracil through its main-chain carbonyl. Both of these two single-residue mutations induce a moderate loss of sensitivity to UTP inhibition and an important loss in activation by GTP. Simultaneous substitution of Thr-138 and Asn-140 with alanine decreases the Tm of the parent molecule by 25 °C and increases the apparent Km for ATP or UMP by a factor of 2.6 and 12, respectively. For all variants, the Vm value remains unchanged or moderately changed. It suggests that each monomer conserved its three-dimensional structure and that the decrease in thermal stability results from a weakening of the quaternary structure.


View this table:
[in this window]
[in a new window]
  		TABLE II
Thermal stability and kinetic parameters of E. coli UMP kinase variants obtained by site-directed mutagenesis of the D159N reference form.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
A Nucleoside Monophosphate Kinase with an Atypical Regulation Supported by an Atypical Fold—UMPKeco exhibits striking differences with other NMP kinases concerning its activity regulation: activation by the GTP effector and inhibition by UTP, the end product of the phosphorylation pathway. In parallel, its three-dimensional structure is unrelated to that of other NMP kinases. This is in accordance with the fact that bacterial UMP kinases do not share significant sequence similarity with other NMP kinases. Indeed, they are closer to other phosphotransferases like aspartokinases (around 30% sequence identity) for which no crystal structure is known, and to a lesser extent (less than 20% sequence identity), to N-acetylglutamate kinases or carbamate kinases. The monomer fold of UMPKeco is close to that of NAGK and CBMK, which suggests that these three kinases derive from a common ancestor. The latter kinase is the less homologous to UMPKeco as it contains a large insertion of around 50 residues that forms a supplementary protruding subdomain involved in dimer interface (26). NAGK is structurally closer but accommodates a very different phosphate acceptor substrate. In accordance, this kinase contains, in the vicinity of the phosphate acceptor, three supplementary {beta}-hairpins (Fig. 3, orange) involved in dimer formation and N-acetyl-glutamate binding (27). NAGK is a homodimer with no documented allosteric regulation, whereas all studied bacterial UMP kinases are hexamers with an allosteric regulation. In accordance, the quaternary structures are fairly different. The dimer assembly in NAGK (and CMBK) contrasts with that observed for the dimer constituted by the two molecules of an asymmetric unit in UMPKeco crystals. Moreover, in the hexamer of the latter kinase, the interface of two neighboring dimers, and in particular, the cross-talk loop (Fig. 3A, green), corresponds to the zone of supplementary {beta}-hairpins in NAGK. As a result, a hexameric assembly similar to that observed in UMPKeco is totally incompatible with the structure of the NAGK (or CBMK) subunit. Therefore, it appears that evolution modified bacterial UMP kinases to allow their typical hexameric packing. This quaternary structure is required for allosteric regulation, as indicated by site-directed mutagenesis of residues from the cross-talk loop involved in subunit contacts (Thr-138 or Asn-140), which decreases the activation by GTP.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 4.
Schematic representation of the enzyme-nucleotide interactions. A, with UMP (violet sticks). B, with UTP. Nitrogen atoms are blue, oxygen atoms are red, and phosphorus atoms are violet. Green dashed lines indicate H bonds (maximal distance 3.3 Å) that are present in both molecules of the asymmetric unit, and radiating semicircles indicate hydrophobic interactions. Labels of the 3 residues that are not H-bonded to UMP are boxed in green. The figure was drawn with LIGPLOT (39).

 



View larger version (53K):
[in this window]
[in a new window]
  		FIG. 5.
Structural basis of the putative phosphate transfer and its inhibition by UTP. A, the UDP binding cavity and the virtual ATP binding cavity as suggested by the superposition of UDP-UMPKeco structure with that of ADPNP-NAG-NAGK. ADPNP and magnesium are orange, and the phosphates of UDP are yellow. Hydrogen bonds with the {beta}-phosphate of UDP are red, whereas the coordinations (some of which being virtual) of the magnesium associated to ADPNP are orange. B, same representation with UTP-UMPKeco. The water molecule interacting with the {alpha}- and {gamma}-phosphates from the nucleotide is shown in red, and the H bonds it establishes are shown as red dots. To better see Asp-146, a small N-terminal part of {alpha}5 is not represented as a schematic helix.

 
Specificity for the Phosphate Acceptor Substrate—Unlike eukaryotes, bacteria use two different enzymes that separately phosphorylate UMP and CMP. How does UMPKeco achieve UMP specificity? This substrate is accommodated in a rather short cavity, which seems too small for purine nucleobases, thus reducing the specificity problem to the pyrimidines. This contrasts with the eukaryotic UMP/CMP kinases from yeast (28) and slime mold (29), the crystal structures of which show a cavity for the phosphate acceptor that is large enough for purine nucleotides to enter. The discrimination between UMP and CMP, characteristic of bacteria, must be related to the two chemical differences between their base moieties. The first difference is the fact that the N3 nitrogen is bound to a hydrogen atom in uracil. This allows the hydrogen bonding observed with the main-chain carbonyl of Asn-140 from the enzyme. The second difference is the presence of the O4 carbonyl on the base ring from UMP (instead of an amino group in CMP). This carbonyl is H-bonded with the main-chain nitrogen of Thr-138. These interactions could not take place with cytosine. Overall, the discrimination between the two nucleoside monophosphates involves main-chain H-bonded atoms. This contrasts with bacterial CMP kinases that achieve this discrimination through side-chain atoms (30).

In the structures published here, the 2'OH interacts with 2 residues. The first one is an aspartate (Asp-77), as frequently observed for ribose-NMP kinases interactions. Site-directed mutagenesis of this residue greatly increases the Km for UMP and lowers the Vmax (31). The second one is a glycine (Gly-63), a residue rarely encountered in nucleotide kinase-ribose interactions. The hydrogen bond it establishes here is favored by that mentioned above with Asp-77. The presence of the bidentate interaction with the 2'OH and the absence of interaction with the 3'OH could explain the inability of UMPKeco to phosphorylate 2'-deoxy-UMP. Conversely, E. coli CMP kinase, which interacts with the 3'OH of the CMP ribose, efficiently phosphorylates 2'-deoxy-CMP.

Insights into Phosphate Transfer—At present, there is no known structure of UMPKeco bound to the phosphate donor ATP or its analogues. However, although the fold of the enzyme subunit differs from that of its structural homologue NAGK in the phosphate acceptor site, the remaining parts of the two proteins are very similar. Therefore, their phosphate donor binding sites should be equivalent. When the structures of UMP-UMPKeco and ADPNP-NAG-NAGK are superimposed (Fig. 3A), the UMP position is close to that of the N-acetyl-glutamate (NAG), and its {alpha}-phosphate points toward the {gamma}-phosphate of the non-hydrolyzable ATP analogue ADPNP. Although the {alpha}5 helix from UMP-Keco has a position very close to that of its NAGK counterpart, the neighboring helices {alpha}6 and {alpha}7 are less well superposed. In particular, the {alpha}6 position results in a more open putative ATP binding site. In the crystal structure, it adopts different positions in the two monomers of an asymmetric unit, suggesting that it could be mobile. The loop connecting {alpha}6 and {alpha}7 (Fig. 3A, magenta loop over the ATP analogue) is in a position that would result in clashes between the side chain of Lys-198 and the adenine moiety of the putative ATP. However, its high B factors and its relatively weak electronic density indicate that this loop is mobile. These signs of mobility also apply to the loop connecting {beta}6 and {beta}7 (Fig. 3A, magenta, at the bottom of ADPNP), which forms the other side of the putative phosphate donor binding cleft. We suggest that ATP binding could imply a shift of these loops and a closure of the {alpha}6 helix upon the phosphate donor. As shown in Fig. 5A, the Asp-201 residue from the loop connecting {alpha}6 and {alpha}7 helices is in a favorable position to bind the catalytically essential magnesium ion coordinated to the ATP analogue. It also applies to Asp-146 from the {alpha}5 helix. Substitution of any of these 2 aspartate residues dramatically decreases kcat (31).

Many proteins that bind ATP have a type A Walker sequence motif (G/A)XXXXGK(T/S). In particular, most NMP kinases contain a classical mononucleotide binding fold with a sequence pattern GXXGXGK (32), corresponding to a phosphate binding loop close to their N-terminal end. This catalytic motif protects the phosphate transferred from the donor ATP from water. In the structure of UMPKeco bound to UDP, 3 residues that are very close in sequence (Lys-15, Ser-17, and Gly-18) are involved in the binding of the {beta}-phosphate (Fig. 5A). They are conserved among bacterial UMP kinases. Lys-15 and Gly-18 residues are homologous to Lys-8 and Gly-11 from NAGK, which interact with the {gamma}-phosphate in the structure of the latter kinase complexed to NAG and ADPNP. It has also been reported that Lys-8 stabilizes the transition state (33). Moreover, the side chain position of Lys-15 in UMPKeco structure is similar to that of Lys-8 in NAGK. The Lys-15, Ser-17, and Gly-18 residues from UMP kinase probably help transfer the phosphate exchanged during catalysis, playing the same role as those of a canonical phosphate binding loop. However, they are not included in such a loop as Lys-15 is included in the first N-terminal {beta}1 strand and Gly-18 is included in the subsequent {eta}1 helix.

Enzyme Activity Regulation—Based on previous kinetic and fluorescence studies on E. coli UMP kinase, it has been suggested that UTP binds with high affinity to the allosteric site under Mg2+-free form (3). Recent biochemical studies on UMP kinases from the Gram-positive bacteria Bacillus subtilis (34) and S. pneumoniae (6) also suggested that this nucleoside triphosphate acts as an allosteric inhibitor. Thus, it was a surprise to see in our crystal structure UTP bound in the phosphate acceptor site. In fact, it is the first time that a nucleoside triphosphate can be accommodated on the acceptor site of a NMP kinase. This shows that the inhibition by UTP is of competitive nature, in agreement with those experiments in which an excess of UMP reverses the inhibiting effect of magnesium-free UTP (7). It could be hypothesized anyway that there is a second binding site for UTP, distinct from the active site, which happened not to be bound in our crystallization conditions. However, the unusual selectivity for magnesium-free UTP on the phosphate acceptor site seen in the structure is in accordance with the biochemical data on inhibition by this nucleotide. This suggests that the UTP binding site observed here is equivalent to the allosteric site previously proposed; UTP binding in the phosphate-acceptor binding site could in turn induce cooperative effects. The decrease of UTP inhibition observed with the substitution of Asn-140, the side chain of which does not interact directly with the nucleoside triphosphate, could be related to the loss of the previously described cooperativity in UTP binding (3).

Why is magnesium-bound UTP unable to inhibit the enzyme activity? In the structure with UTP, a water molecule (Fig. 5B, red) is H-bonded to the {alpha}-and {gamma}-phosphates, to the main-chain nitrogen atom of Asp-146, and to the side-chain oxygen of Thr-145. The distances and geometry with these neighbors are in accordance with the identification of the electronic density to a water molecule. Nucleoside triphosphates are often stabilized by a magnesium ion, interacting with two or three of the phosphates, which plays a catalytic role in phosphate transfer (35). We suggest that the water molecule shown in Fig. 5B occupies a position equivalent to that of the bound magnesium atom in Mg-UTP. Therefore, a possible hypothesis is that magnesium-bound UTP could not be accommodated in the phosphate acceptor binding site of the enzyme as the vicinity with the main-chain NH of Asp-146 and with Lys-15 (probably positively charged) would have a repulsive effect toward the magnesium ion. It is conceivable that evolution avoided such a binding, which could facilitate the loss of the {gamma}-phosphate toward a facing nucleotide acceptor or water, a reaction counterproductive for inhibition. When bound to magnesium, UTP is no more an inhibitor but behaves as a phosphoryl donor (although much weaker than ATP) that binds in the phosphate donor site (3).


   FOOTNOTES
 
* This work was supported by grants from the Centre National de la Recherche Scientifique (Grants UPR 9063, URA 2185, and URA 2171), the Institut National de la Recherche Agronomique (Grant UMR 206), the Institut Pasteur (Grant AC02), and AstraZeneca R & D, Boston, Inc. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and structure factors (code 2BNE, 2BND, 2BNF) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). Back

§ To whom correspondence may be addressed. Tel.: 33-1-30-81-54-73; Fax: 33-1-30-81-53-73; E-mail: briozzo{at}grignon.inra.fr. ** A postdoctoral fellow of the Association pour la Recherche contre le Cancer. To whom correspondence may be addressed. Tel.: 33-1-69-82-42-49; Fax: 33-1-69-82-31-29; E-mail: meyer{at}lebs.cnrs-gif.fr.

1 The abbreviations used are: NMP, nucleoside monophosphate kinases; UMPK, UMP kinase; UMPKeco, E. coli UMPK; CBMK, carbamate kinase; NAG, N-acetyl-glutamate; NAGK, NAG kinase; r.m.s.d., root mean square deviation; ADPNP, 5'-adenylylimidodiphosphate. Back

2 C. Evrin and L. Assairi, unpublished data. Back


   ACKNOWLEDGMENTS
 
We thank Drs. Jean-Luc Ferrer, Philippe Carpentier (beamline BM-30A), and Carlos Petosa (beamline ID14-2) from the ESRF for attentive assistance and Christophe Caillat for participation in refinement.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Yan, H., and Tsai, M. D. (1999) Adv. Enzymol. Relat. Areas Mol. Biol. 73,103 –133[Medline] [Order article via Infotrieve]
  2. Briozzo, P., Golinelli-Pimpaneau, B., Gilles, A. M., Gaucher, J. F., Burlacu-Miron, S., Sakamoto, H., Janin, J., and Bârzu, O. (1998) Structure 6,1517 –1527[Medline] [Order article via Infotrieve]
  3. Serina, L., Blondin, C., Krin, E., Sismeiro, O., Danchin, A., Sakamoto, H., Gilles, A. M., and Bârzu, O. (1995) Biochemistry 34,5066 –5074[CrossRef][Medline] [Order article via Infotrieve]
  4. Yamanaka, K., Ogura, T., Niki, H., and Hiraga, S. (1992) J. Bacteriol. 174,7517 –7526[Abstract/Free Full Text]
  5. Smallshaw, J., and Kelln, R. A. (1992) Genetics 11,59 –65
  6. Fassy, F., Krebs, O., Lowinski, M., Ferrari, P., Winter, J., Collard-Dutilleul, V., and Salahbey Hocini, K. (2004) Biochem. J. 384,619 –627[CrossRef][Medline] [Order article via Infotrieve]
  7. Serina, L., Bucurenci, N., Gilles, A. M., Surewicz, W. K., Fabian, H., Mantsch, H. H., Takahashi, M., Petrescu, I., Batelier, G., and Bârzu, O. (1996) Biochemistry 35,7003 –7011[CrossRef][Medline] [Order article via Infotrieve]
  8. Marco-Marin, C., Escamilla-Honrubia, J. M., and Rubio, V. (2005) Biochim. Biophys. Acta 1747,271 –275[Medline] [Order article via Infotrieve]
  9. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, Second Ed., Spring Harbor Laboratory, Cold Spring Harbor, NY
  10. Munier, H., Gilles, A. M., Glaser, P., Krin, E., Danchin, A., Sarfati, R. S., and Bârzu, O. (1991) Eur. J. Biochem. 196,469 –474[Medline] [Order article via Infotrieve]
  11. Picard, V., and Bock, S. (1997) Methods Mol. Biol. 67,183 –188[Medline] [Order article via Infotrieve]
  12. Crowe, J., Döbeli, H., Gentz, R., Hochulu, E., Stüber, D., and Henco, K. (1994) Methods Mol. Biol. 31,371 –387[Medline] [Order article via Infotrieve]
  13. Blondin, C., Serina, L., Wiesmüller, L., Gilles, A. M., and Bârzu, O. (1994) Anal. Biochem. 220,219 –221[CrossRef][Medline] [Order article via Infotrieve]
  14. Bradford, M. M. (1976) Anal. Biochem. 72,248 –254[CrossRef][Medline] [Order article via Infotrieve]
  15. Otwinowski, Z., and Minor, W. (1993) DENZO, a Film Processing Program for Macromolecular Crystallography, Yale University Press, New Haven, CT
  16. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55,849 –861[CrossRef][Medline] [Order article via Infotrieve]
  17. Terwilliger, T. (2002) Acta Crystallogr. Sect. D Biol. Crystallogr. 59,34 –44
  18. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. Sect. D Biol. Crystallogr. 53,240 –255[CrossRef][Medline] [Order article via Infotrieve]
  19. Navaza, J. (1994) Acta Crystallogr. Sect. A 50,157 –163[CrossRef]
  20. Roussel, A., and Cambillau, C. (1989) Silicon Graphics Geometry Partners Directory, Vols.77–78 , Silicon Graphics, Mountain View, CA
  21. Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., and Grosse-Kunstleve, R. W. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54,905 –921[CrossRef][Medline] [Order article via Infotrieve]
  22. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallog. Sect. A 47,110 –119[CrossRef]
  23. Vonrhein, C., Schlauderer, G. J., and Schulz, G. E. (1995) Structure 3,483 –490[Medline] [Order article via Infotrieve]
  24. Holm, L., and Sander, C. (1996) Science 273,595 –603[Abstract/Free Full Text]
  25. Murzin, A. G., Brenner, S. E., Hubbard, T., and Chothia, C. (1995) J. Mol. Biol. 247,536 –540[CrossRef][Medline] [Order article via Infotrieve]
  26. Ramòn-Maiques, S., Marina, A., Uriarte, M., Fita, I., and Rubio, V. (2000) J. Mol. Biol. 299,463 –476[CrossRef][Medline] [Order article via Infotrieve]
  27. Ramòn-Maiques, S., Marina, A., Gil-Ortiz, F., Fita, I., and Rubio, V. (2002) Structure 10,329 –342[Medline] [Order article via Infotrieve]
  28. Müller-Dieckmann, H. J., and Schulz, G. E. (1994) J. Mol. Biol. 236,361 –367[CrossRef][Medline] [Order article via Infotrieve]
  29. Scheffzek, K., Kliche, W., Wiesmüller, L., and Reinstein, J. (1996) Biochemistry 35,9716 –9727[CrossRef][Medline] [Order article via Infotrieve]
  30. Bertrand, T., Briozzo, P., Assairi, L., Ofiteru, A., Bucurenci, N., Munier-Lehmann, H., Golinelli-Pimpaneau, B., Bârzu, O., and Gilles, A. M. (2002) J. Mol. Biol. 315,1099 –1110[CrossRef][Medline] [Order article via Infotrieve]
  31. Bucurenci, N., Serina, L., Zaharia, C., Landais, S., Danchin, A., and Bârzu, O. (1998) J. Bacteriol. 180,473 –477[Abstract/Free Full Text]
  32. Schulz, G. E. (1992) Curr. Opin. Struct. Biol. 2,61 –67[CrossRef]
  33. Gil-Ortiz, F., Ramòn-Maiques, S., Fita, I., and Rubio, V. (2003) J. Mol. Biol. 331,231 –244[CrossRef][Medline] [Order article via Infotrieve]
  34. Gagyi, C., Bucurenci, N., Sîrbu, O., Labesse, G., Ionescu, M., Ofiteru, A., Assairi, L., Landais, S., Danchin, A., Bârzu, O., and Gilles, A. M. (2003) Eur. J. Biochem. 270,3196 –3204[Medline] [Order article via Infotrieve]
  35. Tock, M. R., Frary, E., Sayers, J. R., and Grasby, J. A. (2003) EMBO J. 22,995 –1004[CrossRef][Medline] [Order article via Infotrieve]
  36. Kabsch, W., and Sander, C. (1983) Biopolymers 22,2577 –2637[CrossRef][Medline] [Order article via Infotrieve]
  37. Gouet, P., Courcelle, E., Stuart, D. I., and Metoz, F. (1999) Bioinformatics (Oxf.) 115,305 –308
  38. DeLano, W. L. (2002) The Pymol Molecular Graphics System, version 0.97, DeLano Scientific, San Carlos, CA
  39. Wallace, A. C., Laskowski, R. A., and Thornton, J. M. (1995) Protein. Eng. 8,127 –134[Abstract/Free Full Text]
  40. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallog. 24,946 –950

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
D. Shi, V. Sagar, Z. Jin, X. Yu, L. Caldovic, H. Morizono, N. M. Allewell, and M. Tuchman
The Crystal Structure of N-Acetyl-L-glutamate Synthase from Neisseria gonorrhoeae Provides Insights into Mechanisms of Catalysis and Regulation
J. Biol. Chem., March 14, 2008; 283(11): 7176 - 7184.
[Abstract] [Full Text] [PDF]

Home page
 Infect. Immun.Home page
S. E. Lee, S. Y. Kim, C. M. Kim, M.-K. Kim, Y. R. Kim, K. Jeong, H.-J. Ryu, Y. S. Lee, S. S. Chung, H. E. Choy, et al.
The pyrH Gene of Vibrio vulnificus Is an Essential In Vivo Survival Factor
Infect. Immun., June 1, 2007; 75(6): 2795 - 2801.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
C. Evrin, M. Straut, N. Slavova-Azmanova, N. Bucurenci, A. Onu, L. Assairi, M. Ionescu, N. Palibroda, O. Barzu, and A.-M. Gilles
Regulatory Mechanisms Differ in UMP Kinases from Gram-negative and Gram-positive Bacteria
J. Biol. Chem., March 9, 2007; 282(10): 7242 - 7253.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
N. Devroede, N. Huysveld, and D. Charlier
Mutational Analysis of Intervening Sequences Connecting the Binding Sites for Integration Host Factor, PepA, PurR, and RNA Polymerase in the Control Region of the Escherichia coli carAB Operon, Encoding Carbamoylphosphate Synthase
J. Bacteriol., May 1, 2006; 188(9): 3236 - 3245.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/27/25533    most recent
M501849200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend