Regulation of cellular levels of ADP-ribose is important in preventing nonenzymatic ADP-ribosylation of proteins. The Escherichia coli ADP-ribose pyrophosphatase, a Nudix enzyme, catalyzes the hydrolysis of ADP-ribose to ribose-5-P and AMP, compounds that can be recycled as part of nucleotide metabolism. The structures of the apo enzyme, the active enzyme and the complex with ADP-ribose were determined to 1.9Å, 2.7Å and 2.3Å, respectively. The structures reveal a symmetric homodimer with two equivalent catalytic sites, each formed by residues of both monomers, requiring dimerization through domain swapping for substrate recognition and catalytic activity. The structures also suggest a role for the residues conserved in each Nudix subfamily. The Nudix motif residues, folded as a loop-helix-loop tailored for pyrophosphate hydrolysis, compose the catalytic center; residues conferring substrate specificity occur in regions of the sequence removed from the Nudix motif. This segregation of catalytic and recognition roles provides versatility to the Nudix family.
ADP-ribose (adenosine 5'-diphosphoribose, ADPR) is produced enzymatically as part of the turnover of NAD+, cyclic ADPR and poly ADPR. High intracellular levels of ADPR can result in nonenzymatic ADP-ribosylation of proteins, a deleterious process that inactivates enzymes and interferes with recognition processes that rely on enzymatic ADP-ribosylation. For example, actin polymerization is inhibited by nonenzymatic ADP-ribosylation at a cysteine residue1. In addition, histone H1 modified by ADP-ribosylation could serve as a site of elongation for polyADP-ribose polymerase2. Enzymatic ADP-ribosylations can also have deleterious effects; for example, they induce apoptosis in some systems3,
4 and mediate the cytotoxic effect of bacterial toxins5. Clearly, the control of ADP-ribose levels is essential for cellular maintenance.
ADP-ribose pyrophosphatases catalyze the hydrolysis of ADP-ribose to AMP and ribose-5-P. This activity has been detected in organisms ranging from bacteria to higher eukaryotes, including Artemia franciscana embryonic cysts6, rat mitochondria7 and human erythrocytes8. So far, nine ADP-ribose pyrophosphatases have been cloned, expressed and characterized9,
10,
11,
12,
13,
14 from archaea, eubacteria, and eukaryotes, including one murine and two human enzymes. Despite the importance of regulating cellular levels of ADPR, a comprehensive picture of the role of this metabolite is just beginning to emerge. In addition, no structural information is available for any ADP-ribose metabolizing enzyme, including ADP-ribose pyrophosphatase.
All cloned ADP-ribose pyrophosphatases belong to the family of Nudix hydrolases (previously the 'MutT family'), a group of phosphoanhydrases that catalyze the hydrolysis of a nucleoside diphosphate linked to another moiety X15. Members of the Nudix family contain the consensus sequence GX5EX7REUXEEXGU (where U represents Ile, Leu or Val, and X represents any amino acid), which forms part of the versatile catalytic site for diphosphate hydrolysis found in at least 20 characterized monomeric and multimeric enzymes from all three kingdoms. More than 450 putative proteins from 90 species contain the Nudix signature sequence. The nucleoside diphosphate linkage is the common feature of the wide range of substrates of the family, which include NADH, dinucleoside polyphosphates, nucleotide sugars and (deoxy)ribonucleoside triphosphates. These substrates are either harmful to the cell or they are metabolic intermediates that require modulation during the cell cycle or during periods of stress; sanitizing the cell is the postulated general role for the Nudix enzymes15. Elimination of free ADPR falls into this category.
Since the Nudix motif is common to enzymes catalyzing highly diverse reactions, determinants of substrate specificity must lie elsewhere in the structure. Sequence alignment reveals additional conserved amino acid residues, distal from the signature sequence, that correlate with particular substrate specificities9. So far, three subfamilies have been identified: the ADPR pyrophosphatases (ADPRase), the adenosine-(p)n-adenosine hydrolases (ApnAase) and the NADH hydrolases (NADHase). The ADPRase subfamily, which includes the E.coli ADP-ribose pyrophosphatase described here, is characterized by a proline residue that occurs 15 residues from the terminal glycine of the Nudix signature sequence.
Although the structures of two Nudix enzymes determined by NMR, E.coli MutT16 and Ap4A hydrolase17, provided very useful information, some of the most important questions about the Nudix family remain unanswered. Is the loop-helix-loop motif that contains the characteristic sequence the only structurally conserved feature? Are the dimeric members of the family simply a combination of two single-site monomers? How do subfamily characteristic residues contribute to substrate specificity? What is the structural basis of the versatility of the Nudix family?
In order to provide insight into these questions, we present the structure of the E. coli ADP-ribose pyrophosphatase alone and in complex with its substrate, ADP-ribose. As ADPRase activity requires a divalent cation9, probably Mg2+, the structure of the complex of the enzyme with the Mg2+ analog Gd3+ was also determined.
Overall ADPRase structure E. coli ADPRase elutes as a dimer in gel exclusion chromatography. In the crystal structure, the two identical monomers are related by a noncrystallographic two-fold axis (monomers A and B) and display extensive domain swapping (Fig. 1a). The 209-amino acid monomers are folded into two distinct structural domains: an N-terminal domain involved in dimer stabilization and a C-terminal domain — the Nudix domain — that contains the conserved sequence of the Nudix family. The N-terminal domain, comprising the first 54 residues, starts with a coil and a 310-helix followed by a three-stranded antiparallel -sheet (Fig. 1a,b). The Nudix domain is an + fold formed by a mixed -sheet and an antiparallel -sheet that face each other, flanked by helices on each side (Fig. 1a,b).
a, Stereo ribbon diagram. Orange (monomer A) and yellow (monomer B) sections identify residues that contribute to the area buried in dimer formation. In monomer B, the secondary structure elements are labeled. Residues 155−164 of monomer A, not observed in the crystal structure, are shown as dashed chain. b, ADPRase sequence showing the position of secondary structure elements: -strands (blue arrows), -helices (pink rectangles) and 310-helices (green curls). The sequence of MutT aligned to ADPRase based on structural superposition is also shown. Sequence identities are indicated below the sequences, with the Nudix residues in green.
While the Nudix and N-terminal domains are two structurally independent units, their spatial arrangement can only be understood in terms of dimer formation. Furthermore, the most likely function of the N-terminal domain is in dimer stabilization. Indeed, the structures of the monomeric Nudix enzymes E.coli MutT16 and Ap4A hydrolase17, which share the Nudix domain with the ADPRase, completely lack the N-terminal domain (Fig. 1b).
Nudix motif Residues 97−119 of the E. coli ADPRase contain the characteristic Nudix sequence GX5EX7REUXEEXGU. This segment folds as a loop-helix-loop motif stabilized by a network of hydrogen bonds among conserved residues (Fig. 2a). The C-terminus of helix 1 in the motif is anchored by a hydrogen bond between O1 of Glu 116 and the main chain amide of Ala 96. At the other end of the motif, hydrogen bonds from Glu 103 to Glu 100 and Arg 111 clamp the helix 1 to loop L6. The first conserved residue of the Nudix motif, Gly 97, is on -strand 7, making two -sheet intrasheet hydrogen bonds to Ala 60 of strand 4 directly facing the metal site. The absence of a side chain at Gly 97 creates a large cavity for the metal and allows a close approach of the pyrophosphate to the metal ion. Arg 111, a conserved residue in the Nudix hydrolases, was proposed to be directly involved in catalysis. This is not supported by the structure, which shows that Arg 111 helps to orient Glu 112 to bind the metal through water molecule W2 and is involved in a salt link with Glu 103 of the Nudix signature sequence. Two positions of the Nudix motif that show a trend for hydrophobic amino acids, 113 and 118, help to clamp the end of helix 1 to loop L7. The nonconserved residues of the Nudix signature sequence, positions 98−102, 104−110, 114 and 117 point away from the active site and are solvent-exposed. In particular, positions 98−102 form the end of loop L6 that connects strand 7 to helix 1.
Figure 2. Structure of the Nudix motif and the complete Nudix fold.
a, Nudix motif. The side chains of the conserved residues of the motif are shown in their conformation in the apo structure in all-atom representation. The main chain of the residues comprising the Nudix motif are shown in red. b, Nudix fold. The ribbon diagram shows the Nudix fold and the location of the Nudix loop-helix-loop (red).
Nudix fold The residues in the Nudix sequence do not account for the entire extent of conserved structure in the family. The Nudix motif is at the center of a larger structure formed by two -sheets packed between -helices, which can be considered the fundamental common fold of the family, the 'Nudix fold' (Fig. 2b). The fold is formed by a four-stranded, mixed -sheet sandwiched between the catalytic helix and an orthogonally oriented -helix and flanked by a second antiparallel three-stranded -sheet. The fold can accommodate sequences of different lengths in the connecting loops and in the antiparallel -sheet. The Nudix signature sequence motif stabilizes the overall structure of the Nudix fold through a series of hydrogen bonds between the helix of the motif and the mixed -sheet.
Metal binding sites To identify the metal site(s), we determined the structure of ADPRase in complex with Gd3+, in place of Mg2+. Gd3+ is a good analog for Mg2+ — that is, it supports catalytic activity to a level of at least 30% of the Vmax observed with Mg2+ at the same concentration (S.B.G., M.J.B. & L.M.A. unpublished results) — and is readily detectable by X-ray diffraction. Two Gd3+ ions, one at the active site of each monomer, were unambiguously identified based on their anomalous diffraction signal. The metal in monomer B is coordinated by three bidentate carboxylates, all from the same monomer: Glu 112 and Glu 115 of the Nudix signature sequence and Glu 164. Two water molecules, W8 and W9, complete the metal coordination (Fig. 3). With the exception of one of the Glu 164 oxygens, the O−Gd distances range from 2.7 Å to 2.9 Å. The coordination of the Gd3+ observed in molecule A is equivalent to that in molecule B, but due to the crystal contacts, Glu 48A' of a symmetry-related adjacent molecule, rather than Glu 164A, acts as a metal ligand.
The side chains of the residues involved in ion coordination are shown in all-atom representation. The corresponding portion of the 2Fo - Fc electron density map is shown in sky blue. The metal ion is shown in green, and the metal-coordinating waters as red spheres.
In the metal-free ADPRase structure, a water molecule coordinated by Glu 112 and Glu 164 occupies the position of the Gd3+ ion. The coordination of Gd3+ in ADPRase is similar to that observed in the crystal structure of gadolinium (III) acetate tetrahydrate (GAT)18,
19, but the ninth coordination position20 is either not well ordered or unoccupied in ADPRase.
The only change required to bind Mg2+, which is typically hexacoordinated, is for some of the glutamic acid residues to reorient and bind Mg2+ in monodentate fashion. The structures suggest that the metal in ADPRase bridges the pyrophosphate of the substrate to the enzyme.
Dimer interactions The ADPRase dimer is stabilized by interactions in four regions (Fig. 1a). The first region contains a pair of equivalent interfaces created by swapping the N-terminal domains of the monomers. As a result, the N-terminal domain of one monomer and the Nudix domain of the other are in intimate contact. The interfaces have a high content of aromatic residues: Phe 11, Tyr 34, Phe 36 and Phe 41 from the N-terminal domain; Tyr 83 and Trp 90 from the Nudix domain and the two Phe 54 at the crossing point between domains. In the second region, loop L5 of one domain and L8 of the other interdigitate such that the loops of the mixed -sheets are flanked by those of the antiparallel -sheet. Significantly, Pro 134, the characteristic residue of the ADPRase subfamily, is at the tip of loop L8, protruding into the second monomer and helping to interlock the core of the two monomers. In the third region of interaction, symmetry-related L1 -hairpins of the N-terminal domains rest on one another with Phe 28 and Phe 29 at their tips. This arrangement resembles the interactions of the flaps of the homodimeric retroviral aspartic proteases21. In the fourth region, the two-fold related helices 3 are parallel and form hydrogen bonds between residues Gln 195, Glu 181 and His 200 and their symmetry-related mates. Strands 8 of symmetry-related monomers also contribute to this region. The interdigitation of the Nudix -sheets, together with the extensive N-terminal domain swapping (Fig. 1a), leads to 7,505 Å2 buried in dimer formation, two-thirds of which involve apolar residues.
The two monomers are highly similar: the root mean square (r.m.s.) deviation for corresponding C atoms (192 out of 209) is 0.44 Å. Only in two regions, both involving crystal contacts, do the structures differ. In monomer B, the electron density is not well defined for the first eight amino acids, but in monomer A they are ordered through a crystal contact with a symmetry-related molecule. For residues 155−164 of loop L9 in molecule A, crystal contacts have the opposite effect; the electron density for these residues was not well-defined (dashed chain in Fig. 1a). Attempts to build these residues in the conformation observed in monomer B led to collisions with strand 1 of the symmetry-related monomer A.
ADP-ribose binding The only significant differences between the structures in the presence and absence of substrate are observed in the conformation of side chains that interact with ADPR: Met 98, Phe 28, Phe 29 and Arg 56. In addition, 14 water molecules are displaced upon substrate binding. In contrast to the metal binding sites, the two ADPR binding sites occur at the dimer interfaces (Fig. 4a) and have contributions from loop L1 and L9 of one monomer and loop L8 of the other one.
a, Location of the two equivalent ADPR binding sites in the ADPRase dimer. In each binding site, loop L8 of the opposite monomer is in close proximity to the ribose moiety of ADPR. b, Stereo diagram of one ADPR binding site. Residues of the two monomers contributing to binding are labeled (B: main monomer, A: second monomer). The 2Fo - Fc electron density of the ADPR is shown in light blue. Carbons are gray, oxygens red, nitrogens blue, phosphorous yellow, and sulfur green; bound waters (labeled W3 and W4) are shown as red spheres. The adenosine group of the substrate binds to the enzyme in anti conformation (dihedral glycosylic bond is −143°); the adenine ribose ring has C3'-endo puckering and the terminal ribose binds with C2'-endo puckering. c, Interactions between ADPR and ADPRase. The ADPR molecule is drawn with heavy lines. Hydrogen bonds are shown with dashed blue lines; the distances between donors and acceptors are indicated. Amino acids providing van der Waals interactions are shown as decorated arcs. Residue numbers are followed by a letter (A or B) to indicate the monomer. Water molecules W1 to W4 are shown as spheres.
Bound ADP-ribose (Fig. 4b) resembles a horseshoe; the two ends come together such that the adenine N7 and a -phosphate oxygen make hydrogen bonds to the same water molecule, W3 (Fig. 4c). The adenine base is stacked between Phe 29B and Arg 51A (a conserved residue in the ADPRase subfamily) in a hydrophobic pocket that also includes Phe 28B. Adenine positions N1 and N6, which participate in Watson-Crick base pairing in double stranded DNA, form two hydrogen bonds to the protein in the ADPRase−substrate complex (Fig. 4c). These interactions, also observed in 54 PDB entries for 29 different ATP-binding proteins with different catalytic activities and folds22, provide the structural basis for the experimental observation that ADPRase is highly specific for ADPR: its Km for ADPR is lower by at least two orders of magnitude9 than for other sugar nucleotides.
The pyrophosphate and the terminal ribose are buried (buried area = 500 Å2), with the pyrophosphate group pointing towards loop L6 and helix 1 of the Nudix motif, and form hydrogen bonds to water molecules and main chain atoms. Similar to Bacillus cereus ADP-ribosylating toxin23, ADPRase uses an Arg residue to form hydrogen bonds to the phosphate oxygen rather than the Lys typical of nucleotide binding domains. The charge on the -phosphate is neutralized by Arg 79, and the -phosphate makes hydrogen bonds to the Met 98 amide nitrogen and to a water molecule. The ADPR terminal ribose is stacked against Pro 134, the conserved residue characteristic of the ADPRase subfamily, of the opposite monomer. This Pro is part of a turn in loop L8 and forms a wall against which the ribose packs. In addition, flanking amino acid residues, Ser 133 and Gly 135, make hydrogen bonds to ribose oxygen O1. The unusual structure of this loop is most likely the reason this subfamily requires a Pro residue at this position. Significantly, these residues, which participate directly in recognition of the ribose group, are contributed by the second monomer.
The structure of the complex of ADPRase with bound substrate not only shows unambiguously the location and extent of the binding site, but also explains why members of this subfamily of Nudix hydrolases function as dimers: amino acids of both monomers participate in the recognition of both ends of the substrate. In addition, it suggests separate roles for the different components of the catalytic site; the Nudix sequence binds the metal and participates in catalysis, but specificity for the substrate is provided by other residues in the molecule.
Implications for the catalytic mechanism Since both phosphate groups of ADPR are bound to the 5' position of a ribose, their chemical environments are equivalent. Thus, hydrolysis of ADPR to AMP and ribose-5-P could proceed by nucleophilic attack at either - or -phosphate. The structures reported here suggest two possible mechanisms of hydrolysis. In one possible mechanism, the water molecule W4, which makes hydrogen bonds to Glu 116 and Glu 164 (Fig. 4c), is at the proper distance (3.6 Å) to attack the P atom, although the water−phosphorus−oxygen angle is not ideal. In this scenario, Glu 116 is the catalytic base. The ribophosphoryl transfer to the water molecule is facilitated by the metal cation, which neutralizes the charge of the AMP leaving group. Arg 79B, which makes hydrogen bonds to the P phosphate, would stabilize the negative charge on the ribose 5-P.
For an attack at P, water molecule W2 is at the correct attacking water−phosphorus−oxygen angle, although its distance to P (4.6 Å) is longer than expected. W2 makes hydrogen bonds to the O2 atom of Glu 112B, to the NH2 of Arg 111B of the Nudix signature sequence and to the carbonyl oxygen of Met 98B (Fig. 4c). In this mechanism Glu 112B, acting as the catalytic base, would dissociate from the metal to receive the proton from W2. Arg 79B stabilizes the negative charge on the departing ribose-5-P.
Comparison of ADPR pyrophosphatase with MutT MutT, a monomeric Nudix GTPase, has 14% sequence identity with ADPRase over 116 aligned amino acids24. Alignment of the structures of E.coli MutT16 and ADPRase (1.9 Å r.m.s. deviation) results in the sequence alignment shown in Fig. 1b. The N-terminal domain (53 residues), the last 10 residues of the C-terminus and insertions in loops L5, L9 (10), and L7 of the ADPRase largely account for the difference in molecular mass between ADPRase (23.7 kD) and MutT (15 kD). The major structural differences between the enzymes are found in three regions. The well-defined antiparallel -sheet in ADPRase formed by strands 5, 6 and 10 contrasts with the single strand equivalent to 5 present in MutT. Of the three C-terminal helices in ADPRase, only one — equivalent to helix 3 — is observed in MutT. Also, the first loop of the Nudix motif adopts different conformations in MutT and ADPRase.
The loop that includes the subfamily specific residue (L8 in ADPRase) has no significant sequence homology with the equivalent region of MutT (Fig. 1b). In ADPRase L8 has Pro 134 at its tip and interacts with the terminal ribose. In the MutT structure the equivalent residue, Phe 75, forms the back wall of the hydrophobic pocket where the base of the substrate binds. Significantly, this residue, which confers substrate specificity, occurs at the same relative position in the amino acid sequences but contributes differently to the binding site in the two enzymes: it interacts with the substrate of the opposite monomer in ADPRase and with the substrate of the only monomer in MutT.
The two enzymes also differ in the conformation of Gly 97, the first residue on the Nudix signature sequence. In ADPRase, Gly 97 is part of a well-defined -strand making hydrogen bonds that may help anchor helix 1 to the characteristic -sheet; these hydrogen bonds are not apparent in the MutT structure.
The metal binding site of ADPRase includes glutamic acid residues 112, 115, 116 and 164, equivalent to glutamic acid residues 53, 56, 57 and 98 of MutT. Of these, a different subset in each enzyme is involved in direct metal coordination: residues 112, 115 and 164 in ADPRase and residues 56, 57 and 98 in MutT. Glu 164 of ADPRase and Glu 98 of MutT are the only residues involved in metal coordination that do not align in the primary sequence. Since these residues align structurally, it is likely that different members of the Nudix family use different sequence positions to provide the fourth position in the Mg2+ binding site.
Although the structure of the Nudix motif is similar in the two proteins, both the mode of substrate binding25 and the specific substrate contacts are different. Only the diphosphate groups of the two substrates contact the same point in the structural motif. The rest of the substrate, most notably the bases, bind at opposite sides of the catalytic site.
Summary and conclusions The structures presented here provide the first experimental framework to understand the catalytic mechanism and substrate specificity of the ADPRase subfamily of the Nudix nucleoside diphosphohydrolases. The Nudix sequence, which forms the core of a larger conserved structure, the Nudix fold, is a structurally conserved loop-helix-loop motif that creates a scaffold for metal binding and pyrophosphatase chemistry, but whose conserved residues do not interact directly with the substrate. In ADPRase, domain swapping brings residues from other regions of the sequence close to the Nudix motif and creates a unique substrate recognition site. This segregation of catalytic activity and substrate specificity provides the basis for the versatility of the Nudix enzyme family.
Methods Protein expression, purification and crystallization of the ADPRase. Overexpression and purification of E.coli ADPRase were carried out as described9. Size exclusion chromatography was carried out on a Sephadex G100 column using lactalbumin, ovalbumin and bovine serum albumin as molecular weight standards. ADPRase in 50 mM TrisCl pH 7.5, 1 mM EDTA was concentrated to 20 mg ml-1 and crystallized by hanging-drop vapor diffusion at 18 °C. The reservoir contained 10% (w/v) PEG 8000 and 10% (w/v) PEG 1000, and the drops had initially a ratio of protein to reservoir solution of 2:1. Long orthorhombic bars (Table 1) grew to 0.2 0.03 0.03 mm3 in 5−7 d. For data collection, crystals were cryoprotected in a solution of 30% (w/v) PEG 8000 and 30% (w/v) PEG 1000 and flash frozen at 95 K.
Table 1. Statistics for data collection and refinement
X-ray data collection and structure determination. Native data were collected to 1.9 Å resolution on a CCD detector (Brandeis B4) at beamline X25 of the National Synchrotron Light Source (NSLS) (Brookhaven National Laboratory). Diffraction data were processed with DENZO26 and SCALEPACK26. The asymmetric unit of the crystal contains two ADPRase monomers with 42% solvent.
Two derivatives (Table 1) were prepared by soaking preformed crystals in 0.5 mM KAuCl4 for 24 h and in 0.5 mM GdCl3 for 1−3 d. After the heavy atom sites were determined by visual inspection of difference Patterson maps, phases calculated to 2.5 Å resolution with the program SOLVE27 had an overall figure of merit of 0.41 (20−2.5 Å; 0.35 in the last resolution shell) using all the data and including anomalous differences for the gadolinium. Phases were improved by solvent flattening, histogram matching and multi-resolution modification using the program DM28. The densities of the two independent monomers in the asymmetric unit were averaged with the program RAVE29,
30. The model of the ADPRase was built in an electron density map calculated with these phases and was used to trace the chain using the program O31,
32 (residues 1−154 and 165−209 for molecule A and 8−209 for molecule B). Refinement was performed using the Crystallography & NMR System (CNS)33 with a residual target that did not include noncrystallographic symmetry restraints. Rebuilding and correction of the model was guided by A-corrected 2Fo - Fc electron density maps. R and Rfree (calculated with randomly selected 10% of the reflections) were used to monitor refinement of the model34.
ADP-ribose bound to ADPRase. Native crystals were soaked for 12−18 h in mother liquor containing 0.5 mM Na ADPR. Data to 2.2 Å resolution were collected on beamline X25 at NSLS. An electron density map calculated with native phases showed well-resolved density that allowed fitting all portions of the ADPR. A model consisting of native coordinates plus those of ADPR was refined as described for the native. The refinement and model statistics are shown in Table 1.
Other computations. The quality of the structures was assessed with the program PROCHECK35. Buried areas were calculated with CNS33 using a 1.4 Å probe. Figures were drawn with MOLSCRIPT36, BOBSCRIPT37 and RASTER3D38.
Coordinates. Coordinates of the structures have been deposited in the Protein Data Bank with accession codes 1G0S, 1GA7, 1G9Q for the apo enzyme, the metal bound enzyme and the complex with the substrate, respectively.
Received 3 November 2000; Accepted 14 February 2001
REFERENCES
Just, I., Wollenberg, P., Moss, J. & Aktories, K. Cysteine-specific ADP-ribosylation of actin. Eur. J. Biochem.221, 1047−1054 (1994). | Article | PubMed | ISI | ChemPort |
Ueda, K., Kawaichi, M., Okayama, H. & Hayaishi, O. Poly(ADP-ribosy)ation of nuclear proteins. Enzymatic elongation of chemically synthesized ADP-ribose-histone adducts. J. Biol. Chem.254, 679−687 (1979). | PubMed | ISI | ChemPort |
Han, M.K., Cho, Y.S., Kim, Y.S., Yim, C.Y. & Kim, U.H. Interaction of two classes of ADP-ribose transfer reactions in immune signaling. J. Biol. Chem.275, 20799−20805 (2000). | Article | PubMed | ISI | ChemPort |
Moss, J. & Zahradka, P. (guest editors) ADP-ribosylation: metabolic effects and regulatory function. Mol. Cell. Biochem.138, 1−253 (1994).
McDonald, L.J. & Moss, J. Enzymatic and nonenzymatic ADP-ribosylation of cysteines. Mol. Cell. Biochem.138, 221−226 (1994). | Article | PubMed | ISI | ChemPort |
Fernandez, A. et al. Specific ADP-ribose pyrophosphatase from Artemia cysts and rat liver: effects of nitroprusside, fluoride and ionic strength. Biochim. Biophys. Acta1290, 121−127 (1996). | PubMed | ISI | ChemPort |
Ribeiro J.M., Costas, M.J. & Cameselle J.C. ADP-ribose Pyrophosphatase-I partially purified from livers of rats overdosed with acetaminophen reveals enzyme inhibition in vivo reverted in vitro by dithiothreitol. J. Biochem. Mol. Tox.13, 171−177 (1999). | Article | ISI | ChemPort |
Kim, J. et al. Purification and characterization of adenosine diphosphate ribose pyrophosphatase from human erythrocytes. Int. J. Biochem. Cell Biol.30, 629−638 (1998). | Article | PubMed | ISI | ChemPort |
Dunn, C.A., O'Handley, S.F., Frick, D.N. & Bessman, M.J. Studies on the ADP-ribose pyrophosphatase subfamily of the nudix hydrolases and tentative identification of trgB, a gene associated with tellurite resistance. J. Biol. Chem.274, 32318−32324 (1999). | Article | PubMed | ISI | ChemPort |
Gasmi, L., Cartwright, J.L. & McLennan, A.G. Cloning, expression and characterization of YSA1H, a human adenosine 5'-diphosphosugar pyrophosphatase possesing a MutT motif. Biochem. J.344, 331−337 (1999). | Article | PubMed | ISI | ChemPort |
O'Handley, S.F., Frick, D.N., Dunn, C.A. & Bessman, M.J. Orf186 represents a new member of the Nudix hydrolases, active on adenosine(5')triphospho(5') adenosine, ADP-ribose and NADH. J. Biol. Chem.273, 3192−3197 (1998). | Article | PubMed | ChemPort |
Raffaelli, N. et al. Synechocystis sp. s1r0787 protein is a novel bifunctional enzyme endowed with both nicotinamide mononucleotide adenylyltransferase and Nudix hydrolase activityies. FEBS Lett.444, 222−226 (1999). | Article | PubMed | ISI | ChemPort |
Sheikh, S., O'Handley, S.F., Dunn, C.A. & Bessman, M.J. Identification and characterization of the Nudix hydrolase from the Archaeon, Methanococcus jannaschii , as a highly specific ADP-ribose pyrophosphatase. J. Biol. Chem.273, 20924−20928 (1998). | Article | PubMed | ISI | ChemPort |
Yang, H. et al. Cloning and characterization of a new member of the nudix hydrolases from human and mouse. J. Biol. Chem.275, 8844−8853 (2000). | Article | PubMed | ISI | ChemPort |
Bessman, M.J., Frick, D.N. & O'Handley, S.F. The MutT proteins or 'Nudix' hydrolases, a family of versatile, widely distributed, 'housecleaning' enzymes. J. Biol. Chem.271, 25059−25062 (1996). | Article | PubMed | ISI | ChemPort |
Abeygunawardana, C. et al. Solution structure of the MutT enzyme, a nucleoside triphosphate pyrophosphohydrolase. Biochemistry34, 14997−15005 (1995). | Article | PubMed | ISI | ChemPort |
Swarbrick, J. et al. The three dimensional structure of the Nudix enzyme diadenosine tetraphosphate hydrolase from Lupinus angustifolius L. J. Mol. Biol.302, 1165−1177 (2000). | Article | PubMed | ISI | ChemPort |
Favas, M.C., Kepert, D.L. & Skel, B.W. Crystal structure of gadolinium (III) acetate tetrahydrate. Sterochemistry of the nine-coordinate [M(bidentate ligand)3(unidentated ligand)3] system. J. Chem. Soc. 454−458 (1980). | ChemPort |
Allen, F.H. & Kennard, O. 3D search and research using the Cambridge Structural Database. Chemical Design Automation News8, 31−37 (1993).
Cotton, F. & Wilkinson, G. In Advanced inorganic chemistry, 986−987 (John Wiley & Sons, London; 1980).
Navia, M.A. et al. Three dimensional structure of aspartyl protease from human imnunodeficiency virus HIV-1. Nature16, 615−620 (1989). | Article |
Denessiouk, K.A. & Johnson, M.S. When fold is not important: a common structural framework for adenine and AMP binding in 12 unrelated protein families. Proteins38, 310−326 (2000). | Article | PubMed | ISI | ChemPort |
Han, S., Craig, J.A., Putnam, C.D., Carozzi, N.B. & Tainer, J.A. Evolution and mechanism from structures of an ADP-ribosylating toxin and NAD complex. Nature Struct. Biol.6, 932−936 (1999). | Article | PubMed | ISI | ChemPort |
Altschul, S.F. et al. Gapped BLAST and PSI-BLAST a new generation of protein database search programs. Nucleic Acids Res.25, 3389−3402 (1997). | Article | PubMed | ISI | ChemPort |
Lin, J., Abeygunawardana, C., Frick, D.N., Bessman, M.J. & Mildvan, A.S. Solution structure of the quaternary MutT-M2+−AMPCPP-M2+ complex and mechanism of its pyrophosphohydrolase action. Biochemistry36, 1199−1211 (1997). | Article | PubMed | ISI | ChemPort |
Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymol.277, 307−326 (1997).
Terwilliger, T.C. & Berendzen, J. Automated structure solution for Mir and Mad. Acta Crystallogr. D55, 849−861 (1999). | Article | PubMed | ISI |
Cowtan, K.D. An automated procedure for phase improvement by density modification. Joint CCP4 and ESF-EACBM Newslett. Protein Crystallogr.31, 34−38 (1994).
Kleywegt, G.J. & Jones, T.A. Halloween...Masks and Bones. in From first map to final model (eds Bailey, S., Hubbard, R. & Waller, D.) 59−66 (SERC, Daresbury Laboratory, Warrington; 1994).
Jones, T.A., Zou, J.Y., Cowan, S.W. & Kjelgaard, M. Improved methods for binding protein models to electron density maps and the location of errors in these models. Acta Crystallogr. A42, 110−119 (1991). | Article |
Brunger, A.T. et al. Crystallography,NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D54, 905−921 (1998). | Article | PubMed | ISI |
Brunger, A.T. Free R value: Cross-validation in crystallography. Methods Enzymol.277, 366−396 (1997). | ISI | ChemPort |
Laskowski, R., MacArthur, M., Moss, D. & Thornton, J. PROCHECK: a program to check the sterochemical quality of protein structures. J. Appl. Crystallogr.26, 283−291 (1993). | Article | ISI | ChemPort |
Kraulis, J. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structure. J. Appl. Crystallogr.24, 946−950 (1991). | Article | ISI |
Esnouf, R.M. Further additions to MOLSCRIPT version 1.4, including reading and countouring of electron density maps. Acta Crystallogr. D 55 , 938−940 (1999). | ChemPort |
Merrit, E. & Bacon, D. Raster3D: Photorealistic molecular graphics. Methods Enzymol.277, 505−524 (1997). | ChemPort |
Acknowledgments Support was provided by NIGMS grants to L.M.A. and M.J.B. S.B.G was supported by an NSF graduate fellowship. Beamlines X25, X8C and X9B of National Synchrotron Light Source, Brookhaven National Laboratory are gratefully acknowledged. We thank D. Leahy for carefully reading the manuscript.
90 species contain the Nudix signature sequence. The nucleoside diphosphate linkage is the common feature of the wide range of substrates of the family, which include NADH, dinucleoside polyphosphates, nucleotide sugars and (deoxy)ribonucleoside triphosphates. These substrates are either harmful to the cell or they are metabolic intermediates that require modulation during the cell cycle or during periods of stress; sanitizing the cell is the postulated general role for the Nudix enzymes
-sheet (
+ 
1 of Glu 116 and the main chain amide of Ala 96. At the other end of the motif, hydrogen bonds from Glu 103 to Glu 100 and Arg 111 clamp the helix 
0.03 
A-corrected 2Fo - Fc electron density maps. R and Rfree (calculated with randomly selected 10% of the reflections) were used to monitor refinement of the model
