NUDT16 is a (deoxy)inosine diphosphatase,and its deficiency induces accumulation of single-strand breaks in nuclear DNA and growth arrest

Nucleotides function in a variety of biological reactions; however, they can undergo various chemical modifications. Such modified nucleotides may be toxic to cells if not eliminated from the nucleotide pools. We performed a screen for modified-nucleotide binding proteins and identified human nucleoside diphosphate linked moiety X-type motif 16 (NUDT16) protein as an inosine triphosphate (ITP)/xanthosine triphosphate (XTP)/GTP-binding protein. Recombinant NUDT16 hydrolyzes purine nucleoside diphosphates to the corresponding nucleoside monophosphates. Among 29 nucleotides examined, the highest k_cat/K_m values were for inosine diphosphate (IDP) and deoxyinosine diphosphate (dIDP). Moreover, NUDT16 moderately hydrolyzes (deoxy)inosine triphosphate ([d]ITP). NUDT16 is mostly localized in the nucleus, and especially in the nucleolus. Knockdown of NUDT16 in HeLa MR cells caused cell cycle arrest in S-phase, reduced cell proliferation, increased accumulation of single-strand breaks in nuclear DNA as well as increased levels of inosine in RNA. We thus concluded that NUDT16 is a (deoxy)inosine diphosphatase that may function mainly in the nucleus to protect cells from deleterious effects of (d)ITP.     

Cloning, expression, and characterization of a human inosine triphosphate pyrophosphatase encoded by the itpa gene

ITP and dITP exist in all cells. dITP is potentially mutagenic, and the levels of these nucleotides are controlled by inosine triphosphate pyrophosphatase (EC ). Here we report the cloning, expression, and characterization of a 21.5-kDa human inosine triphosphate pyrophosphatase (hITPase), an enzyme whose activity has been reported in many animal tissues and studied in populations but whose protein sequence has not been determined before. At the optimal pH of 10.0, recombinant hITPase hydrolyzed ITP, dITP, and xanthosine 5'-triphosphate to their respective monophosphates whereas activity with other nucleoside triphosphates was low. K(m) values for ITP, dITP, and xanthosine 5'-triphosphate were 0.51, 0.31, and 0.57 mm, respectively, and k(cat) values were 580, 360, and 640 s(-1), respectively. A divalent cation was absolutely required for activity. The gene encoding the hITPase cDNA sequence was localized by radiation hybrid mapping to chromosome 20p in the interval D20S113-D20S97, the same interval in which the ITPA inosine triphosphatase gene was previously localized. A BLAST search revealed the existence of many similar sequences in organisms ranging from bacteria to mammals. The function of this ubiquitous protein family is proposed to be the elimination of minor potentially mutagenic or clastogenic purine nucleoside triphosphates from the cell.     

Molecular basis of the antimutagenic activity of the house-cleaning inosine triphosphate pyrophosphatase RdgB from Escherichia coli

Inosine triphosphate pyrophosphatases, which are ubiquitous house-cleaning enzymes, hydrolyze noncanonical nucleoside triphosphates (inosine triphosphate (ITP) and xanthosine triphosphate (XTP)) and prevent the incorporation of hypoxanthine or xanthine into nascent DNA or RNA. Here we present the 1.5-Å-resolution crystal structure of the inosine triphosphate pyrophosphatase RdgB from Escherichia coli in a free state and in complex with a substrate (ITP + Ca^2 +) or a product (inosine monophosphate (IMP)). ITP binding to RdgB induced a large displacement of the α1 helix, closing the enzyme active site. This positions the conserved Lys13 close to the bridging oxygen between the α- and β-phosphates of the substrate, weakening the P_α-O bond. On the other side of the substrate, the conserved Asp69 is proposed to act as a base coordinating the catalytic water molecule. Our data provide insight into the molecular mechanisms of the substrate selectivity and catalysis of RdgB and other ITPases.     

Characterisation of multiple substrate-specific (d)ITP/(d)XTPase and modelling of deaminated purine nucleotide metabolism

Accumulation of modified nucleotides is defective to various cellular processes, especially those involving DNA and RNA. To be viable, organisms possess a number of (deoxy)nucleotide phosphohydrolases, which hydrolyze these nucleotides removing them from the active NTP and dNTP pools. Deamination of purine bases can result in accumulation of such nucleotides as ITP, dITP, XTP and dXTP. E. coli RdgB has been characterised as a deoxyribonucleoside triphosphate pyrophosphohydrolase that can act on these nucleotides. S. cerevisiae homologue encoded by YJR069C was purified and its (d)NTPase activity was assayed using fifteen nucleotide substrates. ITP, dITP, and XTP were identified as major substrates and kinetic parameters measured. Inhibition by ATP, dATP and GTP were established. On the basis of experimental and published data, modelling and simulation of ITP, dITP, XTP and dXTP metabolism was performed. (d)ITP/(d)XTPase is a new example of enzyme with multiple substrate-specificity demonstrating that multispecificity is not a rare phenomenon.     

Identification of the dITP- and XTP-hydrolyzing protein from Escherichia coli

A hypothetical 21.0 kDa protein (ORF O197) from Escherichia coli K-12 was cloned, purified, and characterized. The protein sequence of ORF O197 (termed EcO197) shares a 33.5% identity with that of a novel NTPase from Methanococcus jannaschii. The EcO197 protein was purified using Ni-NTA affinity chromatography, protease digestion, and gel filtration column. It hydrolyzed nucleoside triphosphates with an O6 atom-containing purine base to nucleoside monophosphate and pyrophosphate. The EcO197 protein had a strong preference for deoxyinosine triphosphate (dITP) and xanthosine triphosphate (XTP), while it had little activity in the standard nucleoside triphosphates (dATP, dCTP, dGTP, and dTTP). These aberrant nucleotides can be produced by oxidative deamination from purine nucleotides in cells; they are potentially mutagenic. The mutation protection mechanisms are caused by the incorporation into DNA of unwelcome nucleotides that are formed spontaneously. The EcO197 protein may function to eliminate specifically damaged purine nucleotide that contains the 6-keto group. This protein appears to be the first eubacterial dITP- and XTPhydrolyzing enzyme that has been identified.     

Effect of nucleotide cofactor structure on recA protein-promoted DNA pairing. 1. Three-strand exchange reaction.

The structurally related nucleoside triphosphates, adenosine triphosphate (ATP), purine riboside triphosphate (PTP), inosine triphosphate (ITP), and guanosine triphosphate (GTP), are all hydrolyzed by the recA protein with the same turnover number (17.5 min-1). The S0.5 values for these nucleotides increase progressively in the order ATP (45 microM), PTP (100 microM), ITP (300 microM), and GTP (750 microM). PTP, ITP, and GTP are each competitive inhibitors of recA protein-catalyzed ssDNA-dependent ATP hydrolysis, indicating that these nucleotides all compete for the same catalytic site on the recA protein. Despite these similarities, ATP and PTP function as cofactors for the recA protein-promoted three-strand exchange reaction, whereas ITP and GTP are inactive as cofactors. The strand exchange activity of the various nucleotides correlates directly with their ability to support the isomerization of the recA protein to a strand exchange-active conformational state. The mechanistic deficiency of ITP and GTP appears to arise as a consequence of the hydrolysis of these nucleotides to the corresponding nucleoside diphosphates, IDP and GDP. We speculate the nucleoside triphosphates with S0.5 values greater than 100 microM will be intrinsically unable to sustain the strand exchange-active conformational state of the recA protein during ongoing NTP hydrolysis and will therefore be inactive as cofactors for the strand exchange reaction.     

ITPase Activity in Dry Blood Spots is Comparable with That in Fresh Erythrocytes

Inosine triphosphate pyrophosphohydrolase (ITPase) catalyzing the pyrophosphohydrolysis of inosine triphosphate, deoxyinosine triphosphate and xanthosine triphosphate is involved in the metabolism and tolerance of thiopurine drugs. ITPase activity plays an important role in the prediction of toxicity to thiopurine therapy. Activities in dry blood spots were compared with fresh erythrocytes. Samples were incubated with inosine triphosphate, then inosine monophosphate was determined by a capillary electrophoresis method. Calculated enzyme activities obtained from dry blood spots were in good accordance with activity in fresh erythrocytes.     

ITPase activity in dry blood spots is comparable with that in fresh erythrocytes

Inosine triphosphate pyrophosphohydrolase (ITPase) catalyzing the pyrophosphohydrolysis of inosine triphosphate, deoxyinosine triphosphate and xanthosine triphosphate is involved in the metabolism and tolerance of thiopurine drugs. ITPase activity plays an important role in the prediction of toxicity to thiopurine therapy. Activities in dry blood spots were compared with fresh erythrocytes. Samples were incubated with inosine triphosphate, then inosine monophosphate was determined by a capillary electrophoresis method. Calculated enzyme activities obtained from dry blood spots were in good accordance with activity in fresh erythrocytes.     

Specificity for guanine nucleotide activation and stabilization of rabbit cardiac adenylate cyclase

These studies examined the structural specificity for guanine nucleotide-facilitated hormonal activation and guanine nucleotide stabilization of cardiac adenylate cyclase. 1. The phosphonate analogues of GTP, p[CH(2)]ppG (guanosine 5'-[betagamma-methylene]-triphosphate) and pp[CH(2)]pG (guanosine 5'-[alphabeta-methylene]triphosphate), were the most effective activators of adenylate cyclase. Other nucleotides producing significant activation (P<0.01) were, in decreasing order of activation: ITP, GDP, GMP, GTP, XTP, CTP, p[NH]ppG (guanosine 5'-[betagamma-imido]triphosphate), dGTP and 2'-O-methyl-GTP. Guanosine, cyclic GMP, UTP and ppppG (guanosine tetraphosphate) had no effect, and 7-methyl-GTP caused a decrease in the activity. 2. Preincubation of membranes at 37 degrees C for 15min before assay at 24 degrees C produced an 80% decrease in adenylate cyclase activity, and preincubation with p[CH(2)]ppG and pp[CH(2)]pG protected and resulted in a net increase in activity. Other nucleotides that completely or partially preserved activity in decreasing order of effectiveness were p[NH]ppG, GDP, GTP, dGTP, ITP, ppppG, 2'-O-methyl-GTP, GMP, CTP and XTP. Several compounds had no effect, including guanosine, cyclic GMP and UTP, whereas preincubation with 7-methyl-GTP produced a further decrease (P<0.05) in activity. 3. The concentration-dependence for activation and stabilization by the naturally occurring guanine nucleotides was examined in the absence of a regenerating system and revealed GMP to have no stabilizing effect and to be less potent than either GDP or GTP in activating adenylate cyclase. 4. A significant correlation (r=0.90) was found between the properties of activation and stabilization for the compounds examined. These findings are consistent with there being a single nucleotide site through which both the activation and stabilization of adenylate cyclase are mediated.     

Effect of dipyridamole on inosine triphosphate pyrophosphohydrolase activity and inosine triphosphate content in fresh human erythrocytes incubated with adenosine

The activity of inosine triphosphate pyrophosphohydrolase (ITPH) in human erythrocytes was found to be 1.50 +/- 0.39 mumol of inosine triphosphate (ITP) hydrolysed x min-1 per g Hb, and no measurable amount of ITP was detected. When dipyridamole was added to the medium composed of adenosine, pyruvate and inorganic phosphate, ITPH activity was 1.18 +/- 0.41, and at the same time ITP accumulation was 0.61 +/- 0.31 mumol/g Hb. The negative correlation between ITPH activity and accumulation of ITP was r = -0.87 at P less than 0.001.     

Inosine nucleosidase from Azotobacter vinelandii

An enzyme catalyzing the hydrolysis of purine nucleosides was found to occur in the extract of Azotobacter vinelandii, strain 0, and was highly purified by ammonium sulfate fractionation, DEAE-cellulose chromatography, hydroxylapatite chromatography and gel filtration on Sephadex G-150. A strict substrate specificity of the purified enzyme was shown with respect to the base components. The enzyme specifically attacked the nucleosides without amino groups in the purine moiety: inosine gave the maximum rate of hydrolysis and xanthosine was hydrolyzed to a lesser extent. The pH optimum of inosine hydrolysis was observed from pH 7 to 9, while xanthosine was hydrolyzed maximally at pH 7. The K _m values of the enzyme for inosine were 0.65 and 0.85 mM at pH 7.1 and 9.0, respectively, and the value for xanthosine was 1.2 mM at pH 7.1.Several nucleotides inhibited the enzyme: the phosphate portions of the nucleotides were suggested to be responsible for the inhibition by nucleotides. Although the inhibition of the enzyme by nucleotides was apparently non-competitive type with respect to inosine, allosteric (cooperative) binding of the substrate was suggested in the presence of the inhibitor. The physiological significance of the enzyme was discussed in connection with the degradation and salvage pathways of purine nucleotides.     

Effect of nucleotide cofactor structure on recA protein-promoted DNA pairing. 2. DNA renaturation reaction.

We have examined the effects of the structurally related nucleoside triphosphates, adenosine triphosphate (ATP), purine riboside triphosphate (PTP), inosine triphosphate (ITP), and guanosine triphosphate (GTP), on the recA protein-promoted DNA renaturation reaction (phi X DNA). In the absence of nucleotide cofactor, the recA protein first converts the complementary single strands into unit-length duplex DNA and other relatively small paired DNA species; these initial products are then slowly converted into more complex multipaired network DNA products. ATP and PTP stimulate the conversion of initial product DNA into network DNA, whereas ITP and GTP completely suppress network DNA formation. The formation of network DNA is also inhibited by all four of the corresponding nucleoside diphosphates, ADP, PDP, IDP, and GDP. Those nucleotides which stimulate the formation of network DNA are found to enhance the formation of large recA-ssDNA aggregates, whereas those which inhibit network DNA formation cause the dissociation of these nucleoprotein aggregates. These results not only implicate the nucleoprotein aggregates as intermediates in the formation of network DNA, but also establish the functional equivalency of ITP and GTP with the nucleoside diphosphates. Additional experiments indicate that the net effect of ITP and GTP on the DNA renaturation reaction is dominated by the corresponding nucleoside diphosphates, IDP and GDP, that are generated by the NTP hydrolysis activity of the recA protein.     

The effect of ITPA polymorphisms on the enzyme kinetic properties of human erythrocyte inosine triphosphatase toward its substrates ITP and 6-Thio-ITP

The role of inosine triphosphatase (ITPase) in adverse drug reactions associated with thiopurine therapy is still under heavy debate. Surprisingly, little is known about the way thiopurines are handled by ITPase. We studied the effect of ITPA polymorphisms on the handling of inosine triphosphate (ITP) and thioinosine triphosphate (TITP) to gain more insight into this phenomenon. Human erythrocyte ITPase activity was measured by incubation with ITP using established protocols, and the generated inosine monophosphate (IMP) was measured using ion-pair RP-HPLC. Molecular analysis of the ITPA gene was performed to establish the genotype. Kinetic parameters were established for the two common polymorphisms for both ITP and TITP as substrates using the above mentioned protocol. Both ITP and TITP are substrates for ITPase and their enzyme activities are comparable. Substrate binding is not altered in the different ITPA polymorphisms. It is shown that the velocity of pyrophosphohydrolysis is compromised when the c.94C > A polymorphism is present, both in the heterozygous and in the homozygous state. TITP is handled by ITPase in a similar way as for ITP, which implies that TITP will accumulate in the erythrocytes of patients with an ITPase deficiency, resulting in adverse drug reactions (ADRs) on thiopurine therapy. In carriers of ITPA polymorphisms, the matter is more complex and the development of ADR may depend on additional epigenetic factors rather than on the accumulation of thiopurinenucleotides.     

Profiles of purine biosynthesis, salvage and degradation in disks of potato (Solanum tuberosum L.) tubers

To find general metabolic profiles of purine ribo- and deoxyribonucleotides in potato (Solanum tuberosum L.) plants, we looked at the in situ metabolic fate of various ¹⁴C-labelled precursors in disks from growing potato tubers. The activities of key enzymes in potato tuber extracts were also studied. Of the precursors for the intermediates in de novo purine biosynthesis, [¹⁴C]formate, [2-¹⁴C]glycine and [2-¹⁴C]5-aminoimidazole-4-carboxyamide ribonucleoside were metabolised to purine nucleotides and were incorporated into nucleic acids. The rates of uptake of purine ribo- and deoxyribonucleosides by the disks were in the following order: deoxyadenosine > adenosine > adenine > guanine > guanosine > deoxyguanosine > inosine > hypoxanthine > xanthine > xanthosine. The purine ribonucleosides, adenosine and guanosine, were salvaged exclusively to nucleotides, by adenosine kinase (EC 2.7.1.20) and inosine/guanosine kinase (EC 2.7.1.73) and non-specific nucleoside phosphotransferase (EC 2.7.1.77). Inosine was also salvaged by inosine/guanosine kinase, but to a lesser extent. In contrast, no xanthosine was salvaged. Deoxyadenosine and deoxyguanosine, was efficiently salvaged by deoxyadenosine kinase (EC 2.7.1.76) and deoxyguanosine kinase (EC 2.7.1.113) and/or non-specific nucleoside phosphotransferase (EC 2.7.1.77). Of the purine bases, adenine, guanine and hypoxanthine but not xanthine were salvaged for nucleotide synthesis. Since purine nucleoside phosphorylase (EC 2.4.2.1) activity was not detected, adenine phosphoribosyltransferase (EC 2.4.2.7) and hypoxanthine/guanine phosphoribosyltransferase (EC 2.4.2.8) seem to play the major role in salvage of adenine, guanine and hypoxanthine. Xanthine was catabolised by the oxidative purine degradation pathway via allantoin. Activity of the purine-metabolising enzymes observed in other organisms, such as purine nucleoside phosphorylase (EC 2.4.2.1), xanthine phosphoribosyltransferase (EC 2.4.2.22), adenine deaminase (EC 3.5.4.2), adenosine deaminase (EC 3.5.4.4) and guanine deaminase (EC 3.5.4.3), were not detected in potato tuber extracts. These results suggest that the major catabolic pathways of adenine and guanine nucleotides are AMP → IMP → inosine → hypoxanthine → xanthine and GMP → guanosine → xanthosine → xanthine pathways, respectively. Catabolites before xanthosine and xanthine can be utilised in salvage pathways for nucleotide biosynthesis.     

Profiles of purine metabolism in leaves and roots of Camellia sinensis seedlings

To determine the metabolic profiles of purine nucleotides and related compounds in leaves and roots of tea (Camellia sinensis), we studied the in situ metabolic fate of 10 different (14)C-labeled precursors in segments from tea seedlings. The activities of key enzymes in tea leaf extracts were also investigated. The rates of uptake of purine precursors were greater in leaf segments than in root segments. Adenine and adenosine were taken up more rapidly than other purine bases and nucleosides. Xanthosine was slowest. Some adenosine, guanosine and inosine was converted to nucleotides by adenosine kinase and inosine/guanosine kinase, but these compounds were easily hydrolyzed, and adenine, guanine and hypoxanthine were generated. These purine bases were salvaged by adenine phosphoribosyltransferase and hypoxanthine/guanine phosphoribosyltransferase. Salvage activity of adenine and adenosine was high, and they were converted exclusively to nucleotides. Inosine and hypoxanthine were salvaged to a lesser extent. In situ (14)C-tracer experiments revealed that xanthosine and xanthine were not salvaged, although xanthine phosphoribosyltransferase activity was found in tea extracts. Only some deoxyadenosine and deoxyguanosine was salvaged and utilized for DNA synthesis. However, most of these deoxynucleosides were hydrolyzed to adenine and guanine and then utilized for RNA synthesis. Purine alkaloid biosynthesis in leaves is much greater than in roots. In situ experiments indicate that adenosine, adenine, guanosine, guanine and inosine are better precursors than xanthosine, which is a direct precursor of a major pathway of caffeine biosynthesis. Based on these results, possible routes of purine metabolism are discussed.     

Evidence for the involvement of cytosolic 5'-nucleotidase (cN-II) in the synthesis of guanine nucleotides from xanthosine

In this paper, we show that in vitro xanthosine does not enter any of the pathways known to salvage the other three main natural purine nucleosides: guanosine; inosine; and adenosine. In rat brain extracts and in intact LoVo cells, xanthosine is salvaged to XMP via the phosphotransferase activity of cytosolic 5'-nucleotidase. IMP is the preferred phosphate donor (IMP + xanthosine --> XMP + inosine). XMP is not further phosphorylated. However, in the presence of glutamine, it is readily converted to guanyl compounds. Thus, phosphorylation of xanthosine by cytosolic 5'-nucleotidase circumvents the activity of IMP dehydrogenase, a rate-limiting enzyme, catalyzing the NAD(+)-dependent conversion of IMP to XMP at the branch point of de novo nucleotide synthesis, thus leading to the generation of guanine nucleotides. Mycophenolic acid, an inhibitor of IMP dehydrogenase, inhibits the guanyl compound synthesis via the IMP dehydrogenase pathway but has no effect on the cytosolic 5'-nucleotidase pathway of guanine nucleotides synthesis. We propose that the latter pathway might contribute to the reversal of the in vitro antiproliferative effect exerted by IMP dehydrogenase inhibitors routinely seen with repletion of the guanine nucleotide pools.     

Enhancement of proliferation in cultures of Chinook salmon embryo cells by interactions between inosine and bovine sera

The influence of inosine on DNA synthesis by Chinook salmon embryo cells (CHSE-214) was investigated because previously cell number was shown to increase from six- to thirtyfold if inosine was added to the basal medium (L-15) supplemented with either dialyzed fetal bovine serum (dFBS), calf serum (CS), or dCS. Relative to L-15, ³H-thymidine incorporation was inhibited by these sera alone but elevated in nondialyzed (intact) FBS. Inosine at 10 μM stimulated ³H-thymidine incorporation from ten- to seventyfold in dFBS, CS, and dCS but was only slightly stimulatory in FBS and in L-15 alone. As well as inosine, hypoxanthine, cIMP, IMP, IDP, and ITP were just as stimulatory, but the nonsalvageable purines (xanthine, xanthosine, and XMP) were not. The stimulatory action of inosine was highest in low density cultures. Dipyridamole and S-(p-nitrobenzyl)-6-thioinosine (NBTI), inhibitors of facilitated nonconcentrative nucleoside transport, did not completely block the enhancement of cell number by inosine and by themselves increased proliferation in CS and dCS. Overall, these results suggest that exogenous inosine promoted CHSE-214 proliferation by overcoming factors in the nondialyzable fraction of sera that led to purine loss and by raising intracelular purine nucleotides to levels necessary for cells to respond to growth factors in the nondialyzable fraction of sera. © 1994 Wiley-Liss, Inc.     

Reprogramming the purine nucleotide cofactor requirement of Drosophila P element transposase in vivo.

Guanosine triphosphate (GTP)-binding proteins are involved in controlling a wide range of fundamental cellular processes. In vitro studies have indicated a role for GTP during Drosophila P element transposition. Here we show that P element transposase contains a non-canonical GTP-binding domain that is critical for its ability to mediate transposition in Drosophila cells. Moreover, a single amino acid substitution could switch the nucleotide binding-specificity of transposase from GTP to xanthosine triphosphate (XTP). Importantly, this mutant protein could no longer function effectively in transposition in vivo but required addition of exogenous xanthine or xanthosine for reactivation. These results suggest that transposition may be controlled by physiological GTP levels and demonstrate that a single mutation can switch the nucleotide specificity for a complex cellular process in vivo.     

Nucleoside triphosphate specificity of tubulin

We have determined the binding affinity for binding of the four purine nucleoside triphosphates GTP, ITP, XTP, and ATP to E-site nucleotide- and nucleoside diphosphate kinase-depleted tubulin. The relative binding affinities are 3000 for GTP, 10 for ITP, 2 for XTP, and 1 for ATP. Thus, the 2-exocyclic amino group in GTP is important in determining the nucleotide specificity of tubulin and may interact with a hydrogen bond acceptor group in the protein. The 6-oxo group also makes a contribution to the high affinity for GTP. NMR ROESY experiments indicate that the four nucleotides have different average conformations in solution. ATP and XTP are characterized by a high anti conformation, ITP by a medium anti conformation, and GTP by a low anti conformation. Possibly, the preferred solution conformation contributes to the differences in affinities. When the tubulin E-site is saturated with nucleotide, there appears to be little difference in the ability of the four nucleotides to stimulate assembly. The critical protein concentration is essentially identical in reactions using the four nucleotides. All four of the nucleotides were hydrolyzed during the assembly reaction, and the NDPs were incorporated into the microtubule. We also examined the binding of two gamma-phosphoryl-modified GTP photoaffinity analogues, p(3)-1, 4-azidoanilido-GTP and p(3)-1,3-acetylanilido-GTP. These analogues are inhibitors of the assembly reaction and bind to tubulin with affinities that are 15- and 50-fold lower, respectively, than the affinty for GTP. The affinity of GTP is less sensitive to substitutions at the gamma-phosphoryl position that to changes in the purine ring.