FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2017146847> ?p ?o ?g. }

• W2017146847 endingPage "971" @default.
• W2017146847 startingPage "967" @default.
• W2017146847 abstract "PNP and PNP bind by x-ray crystallography (View interaction) Purine nucleoside phosphorylase (purine nucleoside orthophosphate ribosyl transferase, EC 2.4.2.1) is the key enzyme in the purine salvage pathway, present in every living cell [1]. It catalyses the reversible phosphorolytic cleavage of the glycosidic bond of 6-oxo-purine nucleosides (trimeric and hexameric PNPs) and some analogues, including 6-amino-purine nucleosides (only hexameric PNPs), as follows: purine nucleoside + orthophosphate ↔ purine base + α-dpentose-1-phosphate. The less specific PNP from Escherichia coli is very useful as a tool for enzymatic synthesis of various nucleosides and offers a basis for enzyme-activating prodrug gene therapy enabling selective killing of tumour cells expressing the E. coli PNP gene [2, 3]. The biologically active form of this enzyme is a homohexamer whose structure can be described as a trimer of dimers [1]. In the unliganded structure all six subunits are structurally equivalent, but in the binary complex with phosphate open- and closed-active site conformations were identified [4]. All but one crystal structure of this enzyme complexed with ligands published thus far are biased by the fact that sulfate, a phosphate analogue and an inhibitor of the enzyme, was used as a precipitating agent [4-6]. In a recently published paper [7] we reported the first structure of the binary complex of this enzyme with phosphate (Pi) obtained in the absence of sulfate. However, to better understand the mechanism of the enzyme action, a ternary complex is necessary. Here we report the first structure of the ternary complex of this enzyme obtained in the absence of sulfate. The enzyme was complexed with phosphate and crystallized using phosphate as the precipitant. Then the crystal was soaked with formycin A, which is the structural analogue of the substrate adenosine, with a C–C bond instead of a C–N glycosidic bond. Because of this change formycin A can not be phoshorolysed and acts as a competitive inhibitor vs. nucleoside substrates [8]. Therefore, the ternary complex of the enzyme with phosphate and formycin A (FA) is expected to model substrate binding. The structure reported here shows some new features in the stoichiometry of ligand binding, mode of binding and of the active site contacts. Recombinant E. coli PNP was obtained as described previously [7]. Crystals of the binary complex PNP/Pi were grown from a 50 mM citric buffer pH 5.2 and 30% ammonium phosphate (1.17 M) with an enzyme concentration 65 mg/ml. In these conditions crystals appeared after a year and were then soaked for 24 h in 5 mM of formycin A dissolved in the mother liquor. The X-ray diffraction data was collected with the BM14 beamline at the ESRF, Grenoble. The data collection and refinement statistics are given in the Table 1 . Fluorescence titrations of the E. coli PNP and tyrosine with phosphate and NaCl were conducted at pH 7.6 (50 mM Tris/HCl) and 5.3 (50 mM citrate buffer) as described previously for phosphate [4, 7]. To achieve a high concentration of phosphate or NaCl in titrations, very concentrated stock solutions were used: sodium phosphate 1.5 M at pH 7.6 and 4 M at pH 5.3, and 5 M of NaCl at both pHs. Protein and tyrosine were both present in these solutions so there was no dilution of the fluorophore in the course of titration. For phosphate above 0.5 M and NaCl above 2 M, to keep a constant concentration of the fluorophore, a sequence of two or three titrations was conducted. The first started with 0 M of a phosphate or NaCl in a cuvette, while the second and third titration started with initial concentration of titrant 1.2 and 2 M, respectively. In each case a control titration of the buffer with the ligand was performed and subtracted from the respective experiment with the protein [9]. These allow introduction of corrections for the inner-filter effect. This effect may be neglected for low phosphate concentrations, up to about 50 mM, for which absorbance of the phosphate solution for wavelengths 250–300 nm is less than 0.05. But for phosphate concentrations as high as 1 M it is very prominent and if not corrected may substantially bias the results. We attempted to fit equations describing several models of ligand binding to the titration data. Enzyme concentration was in the range of 0.5–1.5 μM, while phosphate concentration was up to mM (see below), so we fitted equations derived under the assumption that protein concentration is much lower than ligand concentration. Since we observed three kinds of phosphate binding sites in the crystal structure we attempted to fit the three site model to the full ligand concentration range. However, this was not successful, since no saturation at high ligand concentration was achieved, and we have also shown, by titration of the enzyme with NaCl, that change in the fluorescence signal for the high phosphate concentration is not due to phosphate itself but to the increased ionic strength (also see Section 3). Therefore we restricted fitting to the data obtained up to the ionic strength that does not change fluorescence (14 and 63 mM phosphate at pH 5.3 and 7.0, respectively). The overall structure of the homohexamer can be described as a trimer of dimers with 32 symmetry and it is similar to other E. coli PNP structures reported to date (Fig. 1 ). The space group of the ternary complex is P 61 2 2 with one half of the hexamer in the asymmetric unit. Monomers A and C form a dimer, while monomer B forms a dimer with monomer B’ from the symmetrically related unit (Fig. 1). In agreement with previously reported structures of the enzyme with phosphate [4, 7], two conformations of E. coli PNP active sites could be identified: an open conformation with a continuous helix H8 (residues 214–236), and a closed conformation with helix H8 broken into two segments and the N-terminal part of the helix (residues 214–219) moved towards the active site pocket narrowing the entrance. As in the structure of the binary complex PNP/Pi [7], two active sites (monomers A and A′) in the hexamer are found in the closed, and four (monomers B, C, B′ and C′) in the open conformation. All of the active sites are fully occupied with FA molecules, which are well resolved in the electron density (Fig. 2 ). The positions of FAs are the same as in the previously described structure of the ternary complex obtained with sulfate as a precipitant (1K9S) [4]. However, in the structure presented several new interesting features are observed. In contrast to all previously reported structures of the E. coli PNP, the active sites of monomers B and C contain two phosphate molecules each (2, 3 ). Moreover, neither of the two phosphates is found in the usual position making hydrogen bond contacts to the nucleoside and residues Gly20, Arg87, Ser90 and Arg43 from the adjacent monomer [4-7, 12]. Instead, the two phosphates make two short hydrogen bonds to each other. In this case the usual phosphate position is occupied by three water molecules (Fig. 3) and one of the two phosphate ions makes three hydrogen bonds with them. These three molecules make contacts with the pentose oxygen atoms of FA. The other phosphate molecule is in close contact with two nitrogen atoms from the Arg24 residue, which points away from the active site. This configuration of two phosphate molecules in close contact with each other is common in PDB deposited structures but it was never reported in the structures of PNPs from E. coli. In monomer A, whose active site is in a closed conformation, one phosphate ion is hydrogen bonded with the pentose of the nucleoside [5-7]. In the region around the H8 helix there is however still some residual density which is most prominently close to the main chain oxygen atom of the Arg24 residue. Any attempt to model this residual density failed as it is too close to the oxygen atom and too large to be a solvent molecule. Interestingly, these two positions match very well with the positions where the two phosphates would be found if monomer A was in the conformation observed for monomers B and C (Fig. 2). One possible explanation for this observation is that monomer A is predominantly in the closed conformation with phosphate at the usual position and rarely in the open conformation with two phosphates found in positions similar to those in monomers B and C (Fig. 4 ). Binding of phosphate by the E. coli PNP in solution (pH 7.0), according to the literature [13], is characterised by two binding constants, K d1 = 29.4 μM and K d2 = 1.12 mM. These results were obtained using fluorimetric titrations of the enzyme with phosphate using ligand concentrations up to 40 mM [13], while in the crystallization drop, in which the crystal described here was obtained, the phosphate concentration was 1.17 M and the pH was 5.3. Therefore, we attempted to see what happens at pH 7.6 and pH 5.3, at such a high phosphate concentration, with the fluorescence of the E. coli PNP and tyrosine – the fluorophore responsible for E. coli PNP intrinsic fluorescence, bearing in mind that these could be the result of phosphate itself or of the increase of ionic strength of the medium caused by the addition of phosphate [9]. Therefore, in the control experiments the E. coli PNP was titrated with NaCl. Results are presented in Fig. 5 . Up to about a milimolar phosphate concentration (Fig. 5, inserts) similar results as in [13] were obtained. Binding of phosphate by PNP may be characterised by two different binding constants: at pH 7.6 K d1 = (26 ± 5) μM and K d2 = (1.2 ± 0.2) mM with the relative population of the stronger binding site N 1 = 0.46 ± 0.04; and at pH 5.3 (for which binding causes only 4% quenching of the enzyme intrinsic fluorescence, hence errors are high) K d1 = (10 ± 4) μM and K d2 = (3.3 ± 1.1) mM with the relative population of the stronger binding site N 1 = 0.44 ± 0.03. Higher phosphate concentrations lead to strong quenching of the enzyme fluorescence but with no saturation characteristic expected for binding of the ligand. Moreover, at pH 5.3 the same effect of strong quenching of the enzyme fluorescence is obtained when ionic strength is increased by adding NaCl (Fig. 5, left panel), hence it is not caused by specific binding of phosphate by the enzyme. At pH 7.6 addition of NaCl has a much lower effect on the enzyme intrinsic fluorescence than addition of phosphate, but a high phosphate concentration leads to complete quenching of the tyrosine fluorescence, hence again no specific binding of phosphate to the enzyme is detected. Therefore, it could be concluded that the additional, weakly binding phosphate, site which gives a stoichiometry of two phosphate molecules per monomer, has no direct support in solution studies. There are only two published structures of the E. coli PNP complexed with ligands which are not biased by the presence of the inhibitor – sulfate, as the precipitant. One is the structure of the binary complex of the enzyme with phosphate, PNP/Pi [7]: it shows two active sites in closed conformations, four active sites in open conformations, and a stoichiometry of one phosphate molecule per enzyme subunit, with all ions bound in the “standard” positions already known from previous studies with sulfate as the precipitant [4, 7]. The other is the structure described in the present study: a crystal structure of the same complex but additionally soaked with FA as an analogue of the substrate. The structure has changed substantially, and cannot be described just by adding the FA molecule to the binary complex PNP/Pi [7]. Not only has the stoichiometry of the phosphate binding into the active sites with open conformation changed but also the positions of the phosphates in these sites. In the resulting ternary complex PNP/Pi/FA, only in two (closed) active sites phosphate molecules retained their previously reported positions and stoichiometry of binding, and reside in direct contact with the nucleoside. In the remaining four (open) active sites the stoichiometry has changed. Two, instead of one, phosphate molecules are bound in each site, but none in the position observed in the binary complex. This position in the ternary complex is occupied by three water molecules that bridge the nucleoside and the pair of phosphate molecules. In addition, it seems that a minor part of monomers A is not in the closed conformation, but in the same conformation as the one observed for monomers B and C. This may be interpreted as the closed conformation preceding the re-opening of the active site in the succession of events necessary for the completion of the catalytic cycle. This is in line with the proposed and strongly experimentally evidenced molecular mechanism of this enzyme [4, 7]. The structure of the E. coli PNP presented here shows some unusual features. It is not clear if these features are characteristic only of the ternary complex of this enzyme with phosphate and FA or they bear some relevance on the catalytic mechanism as well. Namely, these features are either the consequence of very high phosphate concentrations, or were present in the structure even before soaking the crystal with FA due to the very long crystallization time. The later seems rather unlikely since the E. coli PNP is a very stable enzyme and does not lose activity even in such a long period, especially in a high protein concentration (65 mg/ml), and in the presence of ligands. To further clarify these issues, new crystallization conditions of the E. coli PNP are needed, free from precipitant that enters the active sites and especially free from phosphate and its analogues. Our attempts to find such conditions have up to now been unsuccessful, although one such case was described in the literature [12]. Coordinates and structure factors for the crystal structure of PNP have been deposited in the PDB with PDB ID: 3UT6. The research leading to this publication has received funding from the European Community's Seventh Framework Program (FP7/2007-2013) under grant agreement No. 226716. This work was supported by the (Grant No. 098-1191344-2943) and Polish Ministry of Science and Higher Education (Grant No. N301 044939)." @default.
• W2017146847 created "2016-06-24" @default.
• W2017146847 creator A5003941320 @default.
• W2017146847 creator A5020210972 @default.
• W2017146847 creator A5060153518 @default.
• W2017146847 creator A5060959571 @default.
• W2017146847 creator A5084630841 @default.
• W2017146847 creator A5085631589 @default.
• W2017146847 date "2012-03-01" @default.
• W2017146847 modified "2023-10-05" @default.
• W2017146847 title "New phosphate binding sites in the crystal structure of <i>Escherichia coli</i> purine nucleoside phosphorylase complexed with phosphate and formycin A" @default.
• W2017146847 cites W1504533757 @default.
• W2017146847 cites W1994415899 @default.
• W2017146847 cites W2006732267 @default.
• W2017146847 cites W2028689547 @default.
• W2017146847 cites W2034698551 @default.
• W2017146847 cites W2035354931 @default.
• W2017146847 cites W2050511169 @default.
• W2017146847 cites W2064570453 @default.
• W2017146847 cites W2065338657 @default.
• W2017146847 cites W2072246683 @default.
• W2017146847 cites W2092126394 @default.
• W2017146847 cites W2353173941 @default.
• W2017146847 doi "https://doi.org/10.1016/j.febslet.2012.02.039" @default.
• W2017146847 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/22569248" @default.
• W2017146847 hasPublicationYear "2012" @default.
• W2017146847 type Work @default.
• W2017146847 sameAs 2017146847 @default.
• W2017146847 citedByCount "7" @default.
• W2017146847 countsByYear W20171468472015 @default.
• W2017146847 countsByYear W20171468472017 @default.
• W2017146847 countsByYear W20171468472018 @default.
• W2017146847 countsByYear W20171468472019 @default.
• W2017146847 countsByYear W20171468472020 @default.
• W2017146847 countsByYear W20171468472021 @default.
• W2017146847 crossrefType "journal-article" @default.
• W2017146847 hasAuthorship W2017146847A5003941320 @default.
• W2017146847 hasAuthorship W2017146847A5020210972 @default.
• W2017146847 hasAuthorship W2017146847A5060153518 @default.
• W2017146847 hasAuthorship W2017146847A5060959571 @default.
• W2017146847 hasAuthorship W2017146847A5084630841 @default.
• W2017146847 hasAuthorship W2017146847A5085631589 @default.
• W2017146847 hasBestOaLocation W20171468471 @default.
• W2017146847 hasConcept C104317684 @default.
• W2017146847 hasConcept C107824862 @default.
• W2017146847 hasConcept C181199279 @default.
• W2017146847 hasConcept C185592680 @default.
• W2017146847 hasConcept C204779464 @default.
• W2017146847 hasConcept C2776476023 @default.
• W2017146847 hasConcept C2776543447 @default.
• W2017146847 hasConcept C2777132085 @default.
• W2017146847 hasConcept C547475151 @default.
• W2017146847 hasConcept C55493867 @default.
• W2017146847 hasConcept C71240020 @default.
• W2017146847 hasConcept C71615608 @default.
• W2017146847 hasConceptScore W2017146847C104317684 @default.
• W2017146847 hasConceptScore W2017146847C107824862 @default.
• W2017146847 hasConceptScore W2017146847C181199279 @default.
• W2017146847 hasConceptScore W2017146847C185592680 @default.
• W2017146847 hasConceptScore W2017146847C204779464 @default.
• W2017146847 hasConceptScore W2017146847C2776476023 @default.
• W2017146847 hasConceptScore W2017146847C2776543447 @default.
• W2017146847 hasConceptScore W2017146847C2777132085 @default.
• W2017146847 hasConceptScore W2017146847C547475151 @default.
• W2017146847 hasConceptScore W2017146847C55493867 @default.
• W2017146847 hasConceptScore W2017146847C71240020 @default.
• W2017146847 hasConceptScore W2017146847C71615608 @default.
• W2017146847 hasIssue "7" @default.
• W2017146847 hasLocation W20171468471 @default.
• W2017146847 hasLocation W20171468472 @default.
• W2017146847 hasLocation W20171468473 @default.
• W2017146847 hasOpenAccess W2017146847 @default.
• W2017146847 hasPrimaryLocation W20171468471 @default.
• W2017146847 hasRelatedWork W1985654998 @default.
• W2017146847 hasRelatedWork W1999889393 @default.
• W2017146847 hasRelatedWork W2008487863 @default.
• W2017146847 hasRelatedWork W2035104364 @default.
• W2017146847 hasRelatedWork W2040317364 @default.
• W2017146847 hasRelatedWork W2040680298 @default.
• W2017146847 hasRelatedWork W2086038894 @default.
• W2017146847 hasRelatedWork W2570025094 @default.
• W2017146847 hasRelatedWork W4225414450 @default.
• W2017146847 hasRelatedWork W2048360890 @default.
• W2017146847 hasVolume "586" @default.
• W2017146847 isParatext "false" @default.
• W2017146847 isRetracted "false" @default.
• W2017146847 magId "2017146847" @default.
• W2017146847 workType "article" @default.