QUARTERLY Review Analogs of diadenosine tetraphosphate (Ap4A) �

This review summarizes our knowledge of analogs and derivatives of diadenosine 5',5"'-P1,P4-tetraphosphate (Ap4A), the most extensively studied member of the dinucleoside 5',5"'-P1,Pn-polyphosphate (NpnN) family. After a short discussion of enzymes that may be responsible for the accumulation and degradation of Np4)N's in the cell, this review focuses on chemically and/or enzymatically produced analogs and their practical applications. Particular attention is paid to compounds that have aided the study of enzymes involved in the metabolism of Ap4A (Np4N'). Certain Ap4A analogs were alternative substrates of Ap4A-degrading enzymes and/or acted as enzyme inhibitors, some other helped to establish enzyme mechanisms, increased the sensitivity of certain enzyme assays or produced stable enzyme:ligand complexes for structural analysis.

The biological roles of Np n N¢ s are rather obscure.Some data suggest that they act as signalling molecules in, for example, regulation of the cell cycle (Grummt, 1978;Nishimura, 1998).Under certain circumstances, e.g. when competing with ATP in ATP-dependent reactions (Rotllan & Miras-Portugal, 1985;Pype & Slegers, 1993) and/or binding with nucleotide receptors (Pintor et al., 1991), Np n N¢ s can be detrimental to the organism (McLennan, 2000).The levels of Np n N¢ s can be precisely regulated by numerous degradative enzymes (Guranowski, 2000).In addition to the non-specific ones, like nucleotide pyrophosphatases/phosphodiesterases (Jakubowski & Guranowski, 1983;Bartkiewicz et al., 1984;Cameselle et al., 1984;Gasmi et al., 1998;Vollmayer et al., 2003), for which Np n N¢ s are very good substrates, there are various specific enzymes.In higher eukaryotes (animals and plants) there is a dinucleoside triphosphatase (EC 3.6.1.29)that preferentially converts Np 3 N¢ s to nucleoside mono-and diphosphates, NMP + N¢ DP and/or N¢ MP + NDP, and a dinucleoside tetraphosphatase (EC 3.6.1.17)that asymmetrically hydrolyzes Np 4 N¢ s to either NTP + N¢ MP or N¢ TP + NMP.In lower eukaryotes -fungi (yeast) and protozoa (Euglena) -dinucleoside tetraphosphates are degraded phosphorolytically, either to N¢ DP + NTP or to NDP + N¢ TP, by Ap 4 A phosphorylases (EC 2.7.7.53).In bacteria, Np 4 N's are hydrolyzed symmetrically to NDP + N¢ DP by a specific Co 2+ -dependent Ap 4 A hydrolase (EC 3.6.1.41).Recently, however, an asymmetrically-acting Np 4 N-ase related to the higher eukaryotic enzyme has been detected in several bacteria (Conyers & Bessman, 1999;Cartwright et al., 1999;Bessman et al., 2001;Lundin et al., 2003).The asymmetrical Np 4 N-ases belong to the "nudix" protein family, comprising enzymes that hydrolyze nucleotides in which a nucleoside diphosphate is attached to one of various groups assigned as x (Bessman et al., 1996;2001).These nudix proteins have a conserved amino+acid sequence that directly participates in catalysis (Harris et al., 2000;Maksel et al., 2001).A search for nudix proteins in various genomes has led to the discovery of hydrolases that prefer Ap 5 A and/or Ap 6 A as substrates in budding yeast (S. cerevisiae) (Cartwright & McLennan, 1999), fission yeast (Schizosaccharomyces pombe) (Ingram et al., 1999) and humans (Safrany et al., 1999).One approach to understanding the biological roles of Np n N¢ s is through the use of structural analogs in biochemical and physiological studies.This review focuses on analogs of diadenosine 5¢ ,5¢ ¢ ¢ -P 1 ,P 4 -tetraphosphate (1)* (Ap 4 A), the most widely investigated member of the Np n N¢ s, and presents our current knowledge of both chemically and enzymatically produced nucleotides.Particular attention is paid to compounds that have been useful in studies of enzymes involved in the metabolism of Ap 4 A. Analogs that behave either as alternative substrates of Ap 4 A-degrading enzymes and/or as enzyme inhibitors have helped to establish the mechanism of action of these enzymes.They have also increased the sensitivity of some enzyme assays, or, by forming stable enzyme:ligand complexes, allowed an analysis of the substrate-binding site in the three-dimensional structures of certain hydrolases.This work partially updates two earlier reviews (Blackburn et al., 1992;Guranowski, 2000) that presented some chemical and biological aspects of both Ap 4 A and Ap 3 A analogs.
So far, only three Ap 4 A analogs modified in the sugar (ribose) moieties have been synthesized.These are the dinucleotides containing fluorescent N-methylanthraniloyl group(s) (abbreviated here as m) bound to the 2¢ -or 3¢ -hydroxyl of the ribose(s) via an ether linkage: mAp 5 A and mAp 5 mA (26) (Reinstein et al., 1990) and mAp 5 T (Lavie et al., 1998).
The adenine-containing analogs can be successively converted into their hypoxanthine-containing counterparts, Ip n Ns, such as Ip 4 A ( 16) and Ip 4 I by treatment with adenosine-phosphate deaminase (EC 3.5.4.17) from the snail Helix pomatia or the fungus Aspergillus oryzae (Guranowski et al., 1995).
Finally, truncated Ap 4 A derivatives, such as adenosine(5¢ )tetraphospho(5¢ )ribose and ribose(5¢ )tetraphospho(5¢ )ribose, can be produced by ATP N-glycosidase from the marine sponge Axilla polypoides.This unusual enzyme catalyzes hydrolysis of the N-glycosidic bond in any compound containing an adenosine-5¢ -diphosphoryl moiety (Reintamm et al., 2003).In practice, the quantities of Ap 4 A analogs synthesized enzymatically are much lower than those obtained by chemical procedures.Representative structures of chemically and enzymatically generated Ap 4 A analogs are shown in Table 1.
[b-32 P]8-N 3 -Ap 4 A was also synthesized by Chavan & Haley (1994) to study its interaction with acidic fibroblast growth factor.

APPLICATION OF SOME OTHER Ap 4 A ANALOGS; THEIR INTERACTION WITH DIFFERENT PROTEINS Ap 4 A homologs
So far, Ap 5 A has been the most useful of the Ap 4 A homologs.Acting as a bisubstrate analog it strongly inhibits adenylate (EC 2.7.4.3) and adenosine (EC 2.7.1.20)kinases.The lowest K i values, around 30 nM, were estimated for the adenylate kinases from rabbit (Lienhard & Secemski, 1973) and pig skeletal muscle (Feldhaus et al., 1975).Based on this, an Ap 5 A-Sepharose affinity resin was prepared and successfully used for the isolation of adenylate kinase from vertebrate muscle (Feldhaus et al., 1975).Ap 6 A was a much poorer inhibitor, with K i values of 450 nM for the porcine adenylate kinase (Feldhaus et al., 1975) and 55 nM for the rabbit muscle enzyme (Bone et al., 1986b).Co-crystallization of Ap 5 A with adenylate kinases from pig muscle (Pai et al., 1977), baker's yeast (Egner et al., 1987) and E. coli (Müller & Schulz, 1988) has allowed the three-dimensional structures of these enzymes to be studied.The K i values for Ap 5 A acting as a competitive inhibitor of MgATP binding were 73 nM and 400 nM, respectively, for the adenosine kinase from human liver (Bone et al., 1986b) and bovine adrenal me-dulla (Rotllan & Miras-Portugal, 1985).Ap 5 A and Ap 6 A also inhibited calf thymus terminal deoxynucleotidyl transferase (EC 2.7.7.31) with K i values of 1.5 mM and 1.3 mM, respectively.These two compounds were found to be more effective than the diadenosine polyphosphates containing 2-, 3-or 4-phosphate groups.However, only Ap 5 A seems to span both the substrate and primer binding site domains of the enzyme (Pandey et al., 1987;Pandey & Modak, 1987).Ap 5 A and Ap 6 A also inhibit the nucleotide-depleted mitochondrial F 1 -ATPase (EC 3.6.1.34)and have been employed in studies of the orientation of the catalytic and non-catalytic sites of this enzyme (Vogel & Cross, 1991).Finally, Ap 5 A has been shown to inhibit carbamoyl phosphate synthetase (glutamine hydrolyzing) (EC 6.3.5.5), indicating that this enzyme has two separate binding sites for ATP (Powers et al., 1977).Ap 5 A and Ap 6 A also affect some physiological processes.Of the various Ap 2-6 As studied, Ap 5 A appeared to be the strongest inhibitor of ADP-induced human platelet aggregation (Harrison et al., 1977) while both Ap 5 A and Ap 6 A, which occur naturally in platelets, act as vasopressors (Schlüter et al., 1994).

Hybrid analogs
Hybrid dinucleoside tetraphosphates, Ap 4 Ns (where N A), have been tested as alternative substrates for different Ap 4 A-de-grading enzymes (Jakubowski & Guranowski, 1983;Plateau et al., 1985;Brevet et al., 1987;Prescott et al., 1989) and for Ap 3 A hydrolase (Barnes et al., 1996).Their use has revealed the asymmetry of the Np 4 N¢ -binding sites of such enzymes as yeast Ap 4 A phosphorylase (Brevet et al., 1987), asymmetrically acting Ap 4 A hydrolase from Artemia (Prescott et al., 1989), and the human Fhit protein, which is a typical dinucleoside triphosphatase (Barnes et al., 1996).In each case, a degree of preferential bond cleavage was observed for these hybrid molecules rather than random degradation.For example, phosphorolysis of Ap 4 G by yeast Ap 4 A phosphorylase yielded over 7-fold more ATP + GDP than GTP + ADP while hydrolysis of Ap 4 G by the Artemia Ap 4 A hydrolase yielded a 4.5-fold excess of AMP + GTP over GMP + ATP.The Fhit protein degraded Ap 4 G to AMP + GTP (85%) and GMP + ATP (15%).
In addition, hybrid Ap 4 A analogs have been used as typical bisubstrate analogs in studies of various nucleoside and nucleotide kinases.Ap 4 U was shown to be an effective inhibitor of uridine kinase from Ehrlich ascites tumor cells (Cheng et al., 1986) and analogs with N = dN were used as probes for distinguishing between kinetic mechanisms of the appropriate kinases: Ap 4 dT for thymidine kinase (EC 2.7.1.21)(Bone et al., 1986a), and Ap 4 dC, Ap 4 dG and Ap 4 dA for deoxycytidine (EC 2.7.1.74),deoxyguanosine (EC 2.7.1.113)and deoxyadenosine (EC 2.7.1.76)kinases (Ikeda et al., 1986).Ap 4 dT, Ap 5 dT and Ap 6 dT were found to be inhibitors of thymidylate kinase (EC 2.7.4.9) from peripheral blast cells of patients with acute myelocytic leukemia (Bone et al., 1986b) and Ap 5 dT and/or P 1 -(5¢ -adenosyl)-P 5 -(5¢ ¢ ¢ -(3¢ ¢ -azido-3¢ ¢-deoxythymidine)pentaphosphate, AZT-p 5 A, were used in studies of the same kinase from yeast (Lavie et al., 1998a) and E. coli (Lavie et al., 1998b).The crystal structures of the latter enzyme complexed with these compounds have been solved to 2.0-C and 2.2-C resolution.Davies and co-workers (1988) tested Ap 3 dT, Ap 4 dT, Ap 5 dT and Ap 6 dT plus their analogs with a methylene group a,b to the thymidine residue, e.g.ApppCH 2 pdT, as potential inhibitors of thymidine kinase, thymidylate kinase and ribonucleotide reductase (EC 1.17.4.1).Ap 5 dT was the best inhibitor of the thymidine kinase and both Ap 5 dT and Ap 6 dT strongly inhibited the thymidylate kinase and were potent inhibitors of CDP reduction catalyzed by the ribonucleotide reductase from L1210 cells.8-Azido-Ap 4 A (20) was employed to covalently label acidic fibroblast growth factor (FGF-1) (Chavan & Haley, 1994) and sugarmodified analogs, dAp 4 dA and dAp 4 dT, were shown to be a new type of substrates for several DNA polymerases of human, bacterial and viral origin.The strongest activity of those compounds was observed for HIV reverse transcriptase (Victorova et al., 1999).
Another fluorogenic analog, mAp 5 mA (26), was used to measure the binding constants of adenylate kinase-ligand complexes.It was specially designed for the E. coli enzyme, which has no tryptophan residues and therefore no strong intrinsic fluorescence signal to report ligand binding.Moreover, eAp 5 A produces no significant fluorescence enhancement with this enzyme, in contrast to mammalian cytosolic adenylate kinase.However, mAp 5 mA produced an exceptionally high fluorescence enhancement upon binding to E. coli adenylate kinase (about 300%) (Reinstein et al., 1990).
Diinosine polyphosphates (Ip n Is) have been shown to act as selective antagonists at a diadenosine polyphosphate receptor identified in rat brain synaptic terminals.The best was Ip 5 I, which was 6000-fold more selective for the P4 dinucleotide receptor than for the ATP receptor (Pintor et al., 1997).Recently, Ip 5 I and Ip 6 I were proposed as valuable tools for diabetes research.They antagonize Ap 5 A-mediated inhibition of insulin release from insulin-secreting (INS-1) cells (Verspohl et al., 2003).

Ap 4 A analogs modified in the polyphosphate chain
Various methylene and halomethylene analogs of Ap 4 A have been assayed with specific and non-specific Ap 4 A-degrading en-zymes, acting as substrates and/or potent inhibitors.Chronologically, the first were AppCH 2 ppA (2), AppCHBrppA, ApCH 2 pppA (3) and ApCH 2 ppCH 2 pA (4) (Guranowski et al., 1987).None was hydrolyzed by the (symmetrical)Ap 4 A hydrolase from E. coli but all were strong inhibitors of this enzyme, with K i values ranging from 3-fold (for ApCH 2 p-pC-H 2 pA) to 15-fold (for AppCHBrppA) lower than the K m for Ap 4 A (25 mM).The (asymmetrical)Ap 4 A hydrolase from yellow lupin did hydrolyze those analogs with one methylene or halomethylene group.AppCH 2 ppA competitively inhibited the hydrolysis of Ap 4 A with a K i 4-fold lower than the K m for Ap 4 A (0.25 mM versus 1 mM).The same three analogs were substrates for the non-specific phosphodiesterase from yellow lupin.Finally, of the analogs tested, only ApCH 2 pppA was a substrate of the Ap 4 A phosphorylase from yeast.It was degraded 40-fold more slowly than Ap 4 A and was also the strongest inhibitor of this enzyme, with a K i of 24 mM versus the K m of 60 mM for Ap 4 A. Similar measurements were subsequently performed with the same and other bb'and ab,a¢b¢-disubstituted phosphonate analogs of Ap 4 A. AppCF 2 ppA (5) and AppCCl 2 ppA were as potent as AppCH 2 ppA and AppCHBrppA as inhibitors of lupin Ap 4 A hydrolase but were weaker when tested against the E. coli hydrolase (Guranowski et al., 1989).McLennan and coworkers (1989) studied a set of 13 phosphonate Ap 4 A analogs with the (asymmetrical)Ap 4 A hydrolase from Artemia and established that the substrate efficiency of bb' -substituted compounds decreased with decreasing substituent electronegativity (O>CF 2 >CFH>CCl 2 >CClH>CH 2 ).These compounds were competitive inhibitors of this enzyme with K i values that generally also decreased with electronegativity from 12 mM for AppCF 2 ppA to 0.4 mM for AppCH 2 ppA (K m for Ap 4 A was 33 mM).Disubstituted analogs were generally less effective inhibitors.However, they displayed a low and unexpected rate of symmetrical cleavage by the Artemia enzyme.Both sets of analogs were also competitive inhibitors of E. coli Ap 4 A hydrolase with K i values ranging from 7 mM for App-CH 2 ppA to 250 mM for ApCH 2 CH 2 pp-CH 2 CH 2 pA.The only disubstituted analog to be hydrolyzed by the E. coli enzyme was ApCF 2 ppCF 2 pA at 0.2% of the rate of Ap 4 A; however, several of the bb¢-substituted compounds showed a limited degree of asymmetrical cleavage.These results were interpreted in terms of a "frameshift" model for substrate binding in which the oligophosphate chain can position itself in the active site of the enzyme with either P a or P b adjacent to the attacking nucleophile (water) depending on the electronegativity of the substituent.
Due to their ability to bind tightly to Ap 4 A hydrolases, some of the methylene and halomethylene analogs were used to determine the three-dimensional structures of the enzyme-substrate complexes: a non-degradable analog with two types of modification, Ap s pCHClpp s A (9), was complexed with the enzyme from Lupinus angustifolius (Swarbrick et al., 2000) and AppCH 2 ppA with the hydrolase from Caenorhabditis elegans (Bailey et al., 2002).In physiological assays with INS-1 cells, Ap s pCH 2 pp s A and Ap s pCHClpp s A inhibited insulin release to the same degree as Ap 4 A (Verspohl et al., 2003).AppCH 2 ppA, AppCHFppA and AppCHClppA were quite potent inhibitors of rat brain adenosine kinase while AppCCl 2 ppA was approximately 5-times less potent (Delaney et al., 1997).Different methylene and halomethylene analogs of Ap 4 A, eAp 4 A, eAp 4 eA, as well as an imido-analog (AppNHppA) were examined as effectors of the ADP-ribosylation reaction of histone H1 catalyzed by purified bovine thymus poly(ADP-ribose)transferase (EC 2.4.2.30).Of the compounds tested, ApCH 2 pppA and eAp 4 A were shown to be the most effective inhibitors of the enzyme.The imido analog was the least effective (Suzuki et al., 1987).To the best of my knowledge, there are no other reports of effects exerted by the imido analog.
The monophosphorothioates (S p )Ap 4 AaS ( 6) and (R p )Ap 4 AaS (7) were tested as alternative substrates for three specific Ap 4 A-degrading enzymes (£a¿ewska & Guranowski, 1990); the asymmetrically degrading Ap 4 A hydrolase from yellow lupin seeds, symmetrically acting Ap 4 A hydrolase from E. coli and the Ap 4 A phosphorylase from yeast.Generally, the R p isomer was a better substrate for all three enzymes than the S p one.Interestingly, the S p analog was cleaved randomly by the yeast phosphorylase yielding four reaction products: ATP, ATPaS, ADP and ADPaS.Since the latter product retained its configuration at the a-phosphorus, this confirmed formation of a covalent NMP-enzyme intermediate as previously postulated (Guranowski et al., 1988).Analysis of the regiospecificity of diadenosine polyphosphate hydrolysis catalyzed by three specific hydrolases using  (Guranowski et al., 1994).Regardless of their configuration, diphosphorothioates (Ap s ppp s A, 8) were highly resistant to the (asymmetrical)Ap 4 A hydrolase from Artemia while the presence of the P 2 :P 3 methylene bridge in Ap s pCH 2 pp s A afforded at least a further five-fold increase of resistance to hydrolysis.Moreover, it was noticed that the presence of P 1 and P 4 thiophosphates can force a symmetrical cleavage of Ap s ppp s A for at least one of the diastereoisomers to yield ADPaS (Blackburn et al., 1987b).Another Ap 4 A analog with dual modifications, Ap s pCHClpp s A (9), proved to be a promising anti-platelet aggregation agent (Chen et al., 1997).
Recently, strong inhibition of ADP-triggered blood platelet aggregation was also reported for the analog containing a central di(hydroxymethyl)phosphonic acid moiety in which both non-bridging oxygens of the phosphates linked directly to each ter-minal adenosyl residue were replaced with sulfur atoms (compound 15) (Walkowiak et al., 2002).
Adenylated derivatives of methanetrisphosphonate, so called "supercharged" analogs of Ap 4 A, have been shown to be effective inhibitors of Ap 3 A hydrolases (Liu et al., 1999).They were, however, poor inhibitors of the human and narrow-leafed lupin (asymmetrical)Ap 4 A hydrolases (Maksel et al., 1999); only the triadenylated compound, App-CH-(ppA)-ppA (10), exerted a significant effect, with IC 50 values estimated in the presence of 50 mM Ap 4 A of 80 mM for the human and 40 mM for the lupin enzyme.A novel family of Ap 4 A analogs in which adenylate or adenosine-5'-phosphorothioate residues are chemically attached to polyols such as glycerol, erythritol and pentaerythritol (Baraniak et al., 1999) have been tested in my laboratory as potential substrates and/or inhibitors of the following enzymes: Ap 3 A hydrolase from yellow lupin seeds, human Ap 3 A hydrolase (Fhit protein), yeast Ap 4 A phosphorylase, (symmetrical)Ap 4 A hydrolase from E. coli and two (asymmetrical)Ap 4 A hydrolases, from narrow-leafed lupin and humans.All the compounds were resistant to the action of these enzymes.However, as inhibitors, they behaved differently.Generally, the adenosine-5'-O-phosphorothioylated polyols were much more potent inhibitors of these enzymes than their adenosine-5'-O-phosphorylated counterparts.Yeast Ap 4 A phosphorylase was the most refractory to inhibition.The Ap 3 A hydrolases were inhibited quite effectively but the estimated K i values did not exceed the range of K m s for Ap 3 A; i.e. they were not lower than 10 -6 M. The strongest inhibitory effects were observed for some of these analogs with the Ap 4 A hydrolases (Guranowski et al., 2003b): 1,4-di(adenosine-5'-O-phosphorothio)erythritol (12) appeared to be the strongest inhibitor of the (asymmetrical)Ap 4 A hydrolase from lupin and humans (K i values of 0.15 mM and 1.5 mM, re-spectively).Of eight adenosine-5'-O-phosphorylated compounds, the same enzyme was inhibited only by 1,4-di(adenosine-5'-O-phospho)erythritol (11).Di(adenosine-5'-O-phosphorothio), di(phosphorothio)pentaerythritol (13) and tri(adenosine-5'-O-phosphorothio), thiophosphoro-pentaerythritol (14) were the most powerful inhibitors of the (symmetrical)Ap 4 A hydrolase from E. coli ever reported (K i values of 0.04 mM and 0.08 mM, respectively).For these two types of enzymes the K i values were lower than the K m s for Ap 4 A. A comparison of the inhibitory effects exerted by di(adenosine-5'-O-phosphorothio)erythritol towards three different (asymmetrical)Ap 4 A hydrolases -lupin, human and nematode (C.elegans), -showed that each enzyme was inhibited to a different extent, probably due to small differences in the shape (topography) of the active sites of these hydrolases.Interestingly, the homologous compound di(adenosine-5¢-O-phosphorothio)glycerol was not inhibitory.To further pursue this structure-activity relationship, two di(adenosine-5'-O-phosphorothio)diols, 1,4-di(adenosine-5'-O-phosphorothio)butanediol and 1,5-di(adenosine-5'-O-phosphorothio)pentanediol, were synthesized.These two compounds inhibited neither the lupin nor the human (asymmetrical)Ap 4 A hydrolase.Thus, one can conclude that not only the distance between the adenosine-5'-O-phosphorothio-residues matters but also the presence of hydroxyls in the chain linking the two nucleosides.
The strongest inhibitors of the (symmetrical)Ap 4 A hydrolase from E. coli were also the strongest when tested with the equivalent symmetrically acting enzyme from Salmonella typhimurium, although the K i values for the latter enzyme were markedly higher from those found for the E. coli counterpart: 0.2 mM versus 0.04 mM for di(adenosine-5'-O-phosphorothio), di(phosphorothio)pentaerythritol and 1.7 mM versus 0.08 mM for tri(adenosine-5'-O-phosphorothio), phosphorothio-pentaerythritol.This result also points to small differences in the binding sites of these two similar enzymes.
Of the enzymes that specifically hydrolyze dinucleoside tri-or tetraphosphates, only the (symmetrical)Ap 4 A hydrolase has not yet had its three-dimensional structure determined.The above analogs may help this goal to be achieved.

CONCLUDING REMARKS AND PERSPECTIVES
As has been shown in this review, Ap 4 A analogs helped to increase our knowledge about the enzymes involved in the metabolism of Ap 4 A and other nucleotides.In particular, the hybrid analogs, Ap 4 Ns, proved to be alternative substrates of different Ap 4 A-degrading enzymes.The preference of cleavage of these asymmetrical compounds as well as other asymmetrically modified Ap 4 A derivatives, e.g.2'-or 3'-adenylylated Ap 4 As, showed that the active sites of the Ap 4 A hydrolases and Ap 4 A phosphorylases are also asymmetric.Various Ap 4 Ns and an Ap 4 A homolog, Ap 5 A, were applied as useful tools in studies of different nucleoside and/or nucleotide kinases.Ap 4 A analogs modified in the oligophosphate chain, particularly the non-degradable ones, such as AppCH 2 ppA and Ap s pCHClpp s A, form stable complexes with (asymmetrical)Ap 4 A hydrolases and this allowed the determination of the three-dimensional structures of these enzymes from the nematode C. elegans (Bailey et al., 2002) and higher plant L. angustifolius (Swarbrick et al., 2000), respectively.The very strong inhibition of (symmetrical)Ap 4 A hydrolases by some adenylylated and adenosine-5'-phosphorothioated polyols (Guranowski et al., 2003b) encourages one to use these analogs for determination of the three-dimensional structure of a (symmetrical)Ap 4 A hydrolase.P a -chiral analogs of Ap 4 A allowed the mechanism of Ap 4 A degradation catalyzed by (asym-metrical)Ap 4 A hydrolase (Dixon & Lowe, 1989) and Ap 4 A phosphorylase (Łażewska a& Guranowski, 1990) to be elucidated.
The extracellular, physiological effects of Ap n As (Lüthje & Ogilvie, 1988;Miras-Portugal et al., 1999;Campbell et al., 1999;Hoyle et al., 2001) call for more detailed studies on the interaction of these compounds with purine-nucleoside receptors (P1), purine-mononucleotide receptors (P2X, P2Y) and specific receptors termed dinucleotide or P4 receptors, and on the enzymes located on the cell surface, such as ecto-nucleotide pyrophosphatases/phosphodiesterases, NPP1, NPP2 and NPP3 (Vollmeyer et al., 2003).Ap 4 A analogs should help to discriminate between these proteins and to better understand the role of Ap 4 A and other Np n N's as signal molecules.
Finally, I anticipate studies on new Ap n A-derivatives, adenosine(5')polyphospho(5')riboses and ribose(5')polyphospho(5')riboses, that can be produced by the use of a novel enzyme, ATP N-glycosidase (Reintamm et al., 2003).Testing of these truncated Ap 4 A derivatives as potential substrates and/or inhibitors of different Ap 4 A-degrading enzymes will shed new light on the requirements of those enzymes with respect to the structure of their substrates.
showed that, for the (symmetrical)Ap 4 A hydrolase from E. coli, attack by water took place only at the b-phosphorus at the unmodified end of the molecule, producing [b-18 O]ADP and unlabeled ADPaS and coumarate:CoA ligase(Pietrowska-Borek et al., 2003)can therefore be employed for the synthesis of different Ap 4 Ns using ATP (or ADP in the case of the Ap 4 A phosphorylase) as an adenylate donor and various NTPs as adenylate acceptors.In addition to NTPs, ADP can also act as an adenylate acceptor (but only in the case of aminoacyl-tRNA ligases) to yield Ap 3 A, while p 4 A or p 4 G give rise to Ap 5 A and Ap 5 G, respectively.ATPaS used as an adenylate acceptor yields a monophosphorothioate derivative of Ap 4 A (Ap 4 AaS or Ap s pppA) (6,7) and b,g[CH 2 ]ATP and a,b[CH 2 ]ATP allow the synthesis of AppCH 2 ppA (2) and ApCH 2 pppA (3), respectively.eATP used as the only NTP in the reaction mixture yields eAp 4 eA (17), while ATPgS, being a poor adenylate acceptor, yields only small amounts of Ap 4 AbS (App s ppA) (Günther Sillero