Diadenosine Oligophosphates (Apna), a Novel Class of Signalling Molecules?

View metadata,FEBS 20186 citation and similar papers at core.ac.uk FEBS Letters 427 (1998)brought to 157^163 you by CORE

provided by Elsevier - Publisher Connector Minireview

Diadenosine oligophosphates (ApnA), a novel class of signalling molecules?

Lev L. Kisseleva;b;*, Just Justesenc, Alexey D. Wolfson1;d, Lyudmila Yu. Frolovaa;b

aU248 INSERM, Institut Curie, Paris, France bEngelhardt Institute of Molecular Biology, 117984 Moscow, Russia cInstitute of Molecular and Structural Biology, University of Aarhus, Aarhus, Denmark dDepartment of Biochemistry, University of Colorado, Boulder, CO, USA

Received 24 March 1998

as pleiotropically acting alarmones [5]. Further studies dem- Abstract The diadenosine oligophosphates (ApnA) were dis- covered in the mid-sixties in the course of studies on aminoacyl- onstrated a ubiquitous occurrence of ApnA in the whole spec- tRNA synthetases (aaRS). Now, more than 30 years later, about trum of organisms from bacteria to higher eukaryotes. 300 papers have been published around these substances in Although the role of ApnA in stress response remains un- attempt to decipher their role in cells. Recently, ApnA have known for eukaryotes, in bacteria ApnA binds to and inhibit emerged as intracellular and extracellular signalling molecules the oxidative stress-related proteins [6,7]. The earlier phase of implicated in the maintenance and regulation of vital cellular ApnA investigations was summarised in several reviews [8^ functions and become considered as second messengers. Great 14]. variety of physiological and pathological effects in mammalian Till the late eighties the role of ApnA in higher eukaryotes cells was found to be associated with alterations of ApnA levels remained controversial. More recently, due to e¡orts of many (n from 2 to 6) and Ap3A/Ap4A ratio. Cell differentiation and groups, ApnA have emerged as putative low-molecular extra- apoptosis have substantial and opposite effects on Ap3A/Ap4A ratio in cultured cells. A human Ap A hydrolase, Fhit, appeared and intracellular signals implicated in the most essential cel- 3 lular functions. The aim of this review is to summarise brie£y to be involved in protection of cells against tumourigenesis. Ap3A is synthesised by mammalian u synthetase (TrpRS) which in the recent observations on the possible roles of ApnA mainly contrast to most other aaRS is unable to synthesise Ap4A and is in higher eukaryotes. an interferon-inducible protein. Moreover, Ap3A appeared to be If ApnA do play important but unknown roles in cell me- a preferred substrate for 2-5A synthetase, also interferon- tabolism, one should observe at least two phenomena: phys- inducible, priming the synthesis of 2P adenylated derivatives of iological e¡ects on various organisms, cell types, tissues, or Ap A, which in turn may serve as substrates of Fhit. Tumour 3 organs induced by ApnA in vivo and/or in vitro, and depend- suppressor activity of Fhit is assumed to be associated with ence of Ap A concentrations in various cell types or organ- involvement of the FhitWAp A complex in cytokine signalling n 3 isms on external and/or internal environmental factors and pathway(s) controlling cell proliferation. The ApnA family is potentially a novel class of signal-transducing molecules whose physiological or pathological cell status. If both of these pre- functions are yet to be determined. requisites are met, more sophisticated biochemical analysis z 1998 Federation of European Biochemical Societies. would be required to elucidate the role of these molecules in the framework of cellular regulation. Key words: Diadenosine oligophosphate;

Signal transduction; Fhit; 2-5A synthetase; 2. ApnA are physiologically active and respond to various Tryptophanyl-tRNA synthetase; Tumour suppression factors

Examples of the e¡ects observed with ApnA are given in 1. Introduction Table 1. Some interesting features emerge from examination of even this incomplete set of data. Many entirely di¡erent Diadenosine oligophosphates (ApnA) are made up of two tissues or cells respond to ApnA (heart, hyppocampus, sperm, adenosine moieties joined in 5P-5P linkage by a chain from two hepatocytes neutrophils, pancreatic cells, etc.) indicating the to six phosphodiester linkages. These compounds were discov- involvement of ApnA family in wide-spread biochemical ered by P. Zamecnik and his co-workers in the mid-sixties events, not restricted to any specialised cell type or tissue. (reviewed in [1]) but the ¢rst indication of the putative phys- The other important feature of ApnA e¡ects is a great variety iological role of ApnA appeared only 10 years later when a of responses that occurs in various cell types after adminis- correlation between Ap4A concentration and proliferative sta- tration or addition of ApnA (Table 1, left column). Some tus of mammalian cells has been demonstrated [2]. In prokar- e¡ects are closely associated with nuclear functions (e.g. stim- yotes, heat shock and oxidative stress cause ApnA accumula- ulation of DNA synthesis, mitogenic activity, activation of tion [3,4], and for this reason these molecules were considered gene transcription), others involve membranes (e.g. induction 2 of Ca oscillations and release, inhibition of KATP channels, etc.). Some e¡ects occur with cells and tissues as whole entities *Corresponding author. Fax: +7 (095) 135 14 05. (e.g. inhibition of sperm motility, decrease of coronary perfu- E-mail: [email protected] sion pressure, vasoconstrictor and vasodilator actions, preven- 1On leave from Bakh Institute of Biochemistry, Moscow, Russia. tion of aggregation of human platelets induced by ADP).

FEBS 20186 7-5-98 158 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163

Table 1 Examples of physiological e¡ects associated with ApnA

E¡ect Target cells or tissues ApnA Reference

Inhibition of KATP channels, Ki 14^17 WM Myocardium Ap4A, Ap5A, Ap6A [22] Inhibition of KATP channels, Ki 17 WM Pancreatic L cells (mice) Ap3A, Ap4A [23] High probability of KATP channel opening at Guinea-pig myocardium Ap5A [15] ischemia Activation of Egr-1 gene, mitogenic e¡ect, Cultured renal mesangial cells (rats) ApnA(n = 3^6) [24] stimulation of DNA synthesis 2 Induction of Ca oscillations Hepatocytes (rats) Ap3A, Ap4A [25] Priming of the respiratory burst of oxidase activity Human neutrophils Ap3A, Ap4A, probably non-speci¢c [26] ATP is more active Negative feedback for excitation Hippocampus Ap4A, Ap5A [27] Inhibition of sperm motility Human spermatozoa Ap3A, Ap4A [28] Decrease of coronary perfusion pressure Isolated rabbit hearts Ap3A, Ap4A [29] 2 Regulation of Ca release via ryanodine receptors Hepatocytes, skeletal muscle, cardiac Ap3A, Ap4A [30] muscle Glycogen phosphorylase activation Rat liver cells Ap3A, Ap4A [21] Vasoconstrictor and vasodilator actions Rat mesenteric arteries, isolated rat ApnA(n = 2^6) [31,32] perfused kidney Antitrombolytic e¡ect, antagonists of ADP-induced Human and rabbit platelets Ap4A and its non-hydrolysable [33] aggregation of human platelets analog 2 Increase in intracellular Ca Human neutrophils ApnA(n = 3^6) [19] Stimulation of gluconeogenesis Isolated rat proximal tubules Ap3A, Ap4A, ATP [34] Inhibition of adenosine kinase activity Rat liver ApnA [35] Control of timing of cell division E. coli Ap4A [36] Delay of apoptosis Neutrophils Ap5A, Ap6A [20]

A tight relation between the metabolic state of cardiac nounced protection from apoptosis that was observed during muscle and the level of Ap5A was revealed [15]. Ischemia incubation with either Ap4A or the cytokine. ApnA-induced induced a 10-fold decrease in cardiac Ap5A levels where the e¡ects do not necessarily have the complex physiological na- major sensor of metabolic stress is the ATP-sensitive K ture; for example, rat liver glycogen phosphorylase is acti- channel [16] an opening of which promotes cellular survival vated by ApnA [21]. under ischemic injury. It is suggested that the nucleotide bind- By no means, this list of ApnA e¡ects is incomplete and will ing domains of KATP channelWprotein complex may serve as be extended tremendously in the coming years. But even the targets for Ap5A [15,17]. ApnA evoked inward and outward limited number of examples clearly demonstrates a high var- ionic current £ows by activating purinoreceptors in Xenopus iability of cell and tissue targets accompanied by a wide di- laevis follicular oocytes [18]. Inward currents were mediated versity of the induced e¡ects. The most probable assumption by a suramin-sensitive P2 receptor which showed at 10 WMof that uni¢es these highly diverse patterns is to consider ApnA ApnA the following agonist potency order: Ap4A s ATP s family as the signal transduction molecules. If this hypothesis Ap3A. Outward currents were mediated by a theophylline- suggested by several groups adequately re£ects present-state sensitive P1 receptor which possesses an agonist potency of understanding of ApnA role, then one should expect to ¢nd Ap2A s ATPEAp4A. Control experiments showed that there multiple factors that transmit their signals via changes in the was no breakdown of the ApnA to ATP, ADP, or AMP. All concentration of ApnA in response to their action. Table 2 ApnA tested so far (n from 2 to 6) stimulated an increase in shows that this prediction is supported by many observations. intracellular concentration of Ca2 in human neutrophils. Most remarkably, interferon causes entirely opposite e¡ects This increase was abolished if cells were pretreated with on the Ap4A and Ap3A levels, 10-fold decrease and 3^5- pertussis toxin known to interfere with activity of G proteins fold increase, respectively. Consequently, the Ap4A/Ap3A ra- [19]. tio is changed 30^50-fold. Interferons are well known to cause Neutrophyl apoptosis could be delayed by inoculating the an inhibition of cell proliferation. On the other hand, in pro- cells with ATP, Ap3A, or Ap4A [20]. Moreover, addition of liferating cells Ap4A increases tremendously, opposite to the these compounds together with a cytokine (granulocyte-mac- resting cells. Heat shock induces in chicken erythrocytes a 10- rophage colony-stimulating factor) resulted in more pro- fold increase of Ap4A concentration. ApnA stimulate prolif-

Table 2 Examples of alterations of ApnA concentrations induced by various factors

Inducer Tissue or cells ApnA E¡ect Reference

Interferon L Human cultured cells Ap4A 10-fold decrease [37] Interferons K and Q Human cultured cells Ap3A 3^5-fold accumulation [38] Glucose Murine pancreatic cells Ap3A, Ap4A 30^70-fold accumulation [23] Heat shock Chicken erythrocytes Ap4A 10-fold increase [39] Phorbol ester (TPA) Promyelocytic human cell line HL60 Ap3A 4^5-fold increase [40] VP16 (topoisomerase II inhibitor, inducer Promyelocytic human cell line HL60 Ap3A, Ap4A 3-fold decrease, 4-fold increase [40] of apoptosis) Ischemia Guinea-pig myocardium Ap5A 10-fold increase [15]

FEBS 20186 7-5-98 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163 159

Table 3 Examples of ApnA binding proteins (APNAB)

ApnA Binding protein Comments Reference

Ap3A, Ap4A P2 purine receptors Rat liver cells, guinea-pig vas deferens [18,44,45] Ap4A Glyceraldehyde-3-phosphate dehydrogenase/uracil-DNA glycosylase HeLa cell nuclear extract 37 kDa [41] Ap4A Acidic ¢broblast growth factor Probably non-speci¢c binding [46] Ap4A, Ap5A, Ap6A Surface receptors of murine brain and heart cells 30 kDa [47^50] Ap4A, not Ap3A DNA polymerase associated protein 200 kDa, subunit 22 kDa [51,52] Ap4AAp4A binding protein 90 kDa X. laevis [53] Ap4A Brain membrane receptor Kd 0.71 WM [54] Ap3A, Ap4A, ATP Granulocyte-macrophage colony-stimulating factor Binding causes delay in apoptosis [55] Ap3A Diadenosine triphosphate hydrolase, Fhit Ap3A is a substrate for this enzyme [56] Ap4A Receptor of ApnA Binding is inhibited by tripeptide Arg- [57] Gly-Ser Ap4A Diadenosine tetraphosphate hydrolase Enzyme substrate tomato [58] Ap4A Heat shock protein ClpB E. coli [59] eration of cultured mesangial cells and augment mesangial cell di¡erent a¤nity and in this case the concentrations of various growth induced by other mitogens released from platelets [24]. ApnA become essential. Ap3A accumulation was observed after induction of cell dif- In fact, APNAB do occur in various cell types (Table 3). ferentiation by phorbol ester in promyelocytic cell line HL60. Roughly, APNAB can be categorised into four groups: (a) In contrast, when HL60 cells were induced to undergo apop- receptors located at the cell surface and most probably in- tosis by VP16, an inhibitor of DNA topoisomerase II, Ap3A volved in signal transduction from an environment inside concentration dropped down while Ap4A concentration in- the cell (the structure and function of these receptors have creased (see references in Table 2). been recently reviewed [60]); (b) enzymes utilising ApnAas Glucose at concentrations inducing insulin release from substrates; (c) other enzymes that bind ApnA but do not pancreatic murine cells produces a 30- to 70-fold increase in metabolise them; (d) other proteins, probably most of them Ap3A and Ap4A concentrations in L cells [23]. These concen- yet unidenti¢ed. trations are su¤cient to inhibit ATP-regulated K channels Among APNAB are haemoglobin, glycogen phosphorylase, when applied to the intracellular side of excised membrane glyceraldehyde-3-phosphate dehydrogenase, DNA polymerase patches from these cells in culture. The authors arrive to the K associated protein. Hormones and hormone-like factors are conclusion that Ap3A and Ap4A act as second messengers able to bind ApnA for example to acidic ¢broblast growth mediating a glucose-induced blockage of the ATP-dependent factor and granulocyte-macrophage colony-stimulating factor. potassium channel [23]. These observations imply that ApnA when bound to speci¢c In many physiological manifestations Ap3A and Ap4A pos- ligands may modulate signal transduction induced by these sess similar but not identical activities, in some cases resem- molecules. In some cases the speci¢city of the binding towards bling those of ATP and ADP. However, in other systems, very di¡erent members of ApnA family and other adenosine deriv- di¡erent and even opposite e¡ects were observed for ApnA atives was relatively low, in other instances the spectrum of depending on oligophosphate chain length. For example, compounds tested for binding speci¢city was rather narrow. ApnA with four or more phosphates are vasoconstrictors in The major problem regarding APNAB is related to biological rat mesenteric arteries, while those with three or less phos- signi¢cance of this binding and its functional consequences. It phates are vasodilators [31]. might be of particular interest to isolate proteins with narrow Data summarised in Table 2 are consistent with the signal speci¢city towards ApnA derivatives, i.e. those that bind for transduction hypothesis since classical inducers of di¡erent example Ap4A but not Ap3A and vice versa. If successful, cell status cause well-de¢ned alterations of intracellular con- speci¢c ApnA binders may be very helpful in elucidating the centrations of various members of ApnA family. signal transduction pathway from ApnA to target molecules. The Ap4A binding capacity [41] of uracil-DNA glycosylase/ 3. ApnA binding proteins (APNAB) glyceraldehyde-3-phosphate dehydrogenase (UDG/GAPDH) is particularly interesting, because this enzyme also binds to The major pathway for signal-transducing molecules to per- various RNAs including tRNAs and to mammalian TrpRS form their functions goes through binding to proteins. The [42]. It seems quite possible that Ap3A synthesised by TrpRS classical examples of such kind of interactions are provided (see Sections 4 and 5) could be also bound to UDG/GAPDH. by binding of hormones to their receptors and by GTP/GDP Since both UDG/GAPDH and TrpRS [43] are present in nu- binding to the members of G protein family. If the signal clei apart of being cytosolic enzymes, it is tempting to spec- transduction hypothesis for ApnA family is correct, the structural dissimilarities between the various members of ApnA family caused by the length of the oligophosphate chain may result in the di¡erent patterns of binding proteins for each ApnA molecule. If a single protein binds ApnA non- selectively toward phosphate chain length, signal transduction may be modulated by di¡erent protein conformational states or by charge di¡erences induced by di¡erent Ap A molecules. n Fig. 1. Synthesis of Ap4A and Ap3A by aminoacyl-tRNA synthe- Finally, a single protein may bind several ApnA but with tases (aaRS) in the absence of cognate tRNA.

FEBS 20186 7-5-98 160 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163

changes the situation and activates substantially the Ap4A synthesis. Majority of aaRS catalyses the synthesis of Ap4A (reviewed in [11,12,61]). Exceptions are ArgRS and GlnRS of any origin. Inability of ArgRS and GlnRS to produce Ap4A is related to the fact that both of these enzymes belong to the group of aaRS capable of aminoacyl adenylate formation only in the presence of cognate tRNA. However, as mentioned above, presence of tRNA in most cases inhibits Ap4A synthesis. It seems very likely that the third member of this group, GluRS also lacks the Ap4A-synthesising activity. Mammalian TrpRS produces only Ap3A due to some peculiarities of its catalytic mechanism [61,62]. Ap4A synthesis by some of aaRS is activated by Zn2 while the activity of the majority of aaRS is not a¡ected by this ion (see [11,12] and references therein). Along with ATP, aaRS also utilise any nucleoside oligophosphates (NDP, NTP, ppGpp, etc.) as substrates resulting in formation of the family of adenylated dinucleoside oligophosphates (reviewed in [11,12]). Non-nucleotidic molecules like inorganic triphosphate or thiamine pyrophosphate may Fig. 2. A hypothesis connecting interferon signalling pathway and also serve as donors of pyrophosphate groups in this reaction, tumour suppression through Ap3A and Fhit hydrolase. FhitWAp3A resulting in synthesis of corresponding adenylated compounds is a protein ligand complex. Ovals indicate proteins. (see [13]). Structural framework for Ap4A formation by SerRS from Thermus thermophilus was deduced from the crystallographic ulate that a ternary TrpRSWAp3AWUDG/GAPDH complex data [63]. A second ATP molecule can bind to enzymeWadeny- may take part both in cytoplasmic and/or nuclear events, late complex with its Q-phosphate in the same position as the for instance, in DNA repair. L-phosphate of the original ATP. This Q-phosphate can attack the seryl adenylate with the formation of Ap4A by the same 4. ApnA metabolism in-line displacement mechanism, which leads to PPi release, but in the reverse direction. A divalent cation is essential for The best known reaction leading to the formation of Ap4A the reaction and may be directly involved in the stabilisation is catalysed by aaRS (Fig. 1). It starts with the formation of of the transition state. an enzyme-bound aminoacyl adenylate, which is the normal Apart from aaRS ¢re£y luciferase [64] and Ap4A hydrolase intermediate in the synthesis of aminoacyl-tRNA. In the ab- [65] were shown to synthesise Ap4A. Luciferase catalyses the sence of tRNA the fate of the enzyme-bound aminoacyl ad- reactions with formation of activated adenylates (luciferyl) as enylate is determined by the competition of several reactions. intermediates. Ap4A hydrolase can synthesise Ap4A in a sim- The reverse reaction results in the formation of ATP and ple back reaction. Apparently, enzymes catalysing the transfer amino acid. Alternatively, aminoacyl adenylate can be at- of nucleoside monophosphate are potential candidates for tacked by the pyrophosphate moiety of ATP leading to for- producing ApnA [66]. mation of Ap4A, or hydrolysed, or dissociated from the en- The ability of aaRS to produce Ap4A in vivo demonstrated zyme. In the absence of substrate tRNA the reaction by the experiments with overproduction of various aaRS in equilibrium is shifted signi¢cantly towards pyrophosphoroly- E. coli [67] resulted in a 3- to 14-fold increase of intracellular sis of the aminoacyl adenylate. Cleavage of PPi formed in the Ap4A concentration. In control, overproduction of the inac- course of amino acid activation by inorganic pyrophosphatase tive mutant aaRS did not a¡ect Ap4A level. In vivo overex-

Table 4 Properties of some eukaryotic ApnA hydrolases 3 Origin Ap4AAp3AFIS50(Ki) Divalent cations MR Protein Reference (WM) family Km kcat Km kcat Activation Inhibition (WM) (s31) (WM) (s31) Human (Fhit) 31.3 0.1 65.0 2.3 ^ Mg2,Mn2,Ca2 Zn2 20642 (K2) HIT [78] S. pombe nd nd 0.16a ^Mg2,Mn2,Ca2 Zn2 20672 (K2) HIT [81] Torpedob 0.48 1 0.54 1.1 1400 Mg2,Mn2,Ca2 nd nd nd [82] Lupin nd nd nd nd 3 Mg2,Mn2 Ca2 22982 (K) MutT [74] Human nd nd ^ ^ 40 nd nd 16829 (K) MutT [83] Porcine 0.8 nd ^ ^ 24 nd nd 16837 (K) MutT [73] Fire£y 1.9 nd ^ ^ 50 nd nd 16000 (K) nd [68] nd, not determined; ^, non-measurable; kcat are given to compare the rates of Ap4A and Ap4A hydrolysis. Rate of Ap4A hydrolysis is 1. aEstimate, catalytic parameters were not measured. b Enzyme activity was characterised using the preparation of the presynaptic membranes. Activity was tested using etheno derivatives of ApnA. The supposed subunit composition of the enzymes is given in brackets.

FEBS 20186 7-5-98 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163 161 pression of mammalian TrpRS correlates with the elevated mutagenesis of the Fhit protein [56]. The three-dimensional levels of Ap3A, not Ap4A [38]. Possible participation of the structure of Fhit is close to that of protein kinase C interact- luciferase in the synthesis of Ap4A in vivo is indirectly sup- ing protein (PKCI), also a member of HIT family, possessing ported by the unusually high level of Ap4A hydrolase activity two purine-nucleotide binding sites per dimer. However, in ¢re£y lanterns [68]. PKCI is unable to hydrolyse Ap3A but it slowly hydrolyses Although ApnA can be non-speci¢cally degraded by a vari- ADP [76,79,80]. ety of phosphodiesterases and nucleotidases, the major role in their degradation belongs to the speci¢c enzymes. In E. coli, 5. Perspectives Ap4A degradation occurs mainly through symmetrical hydrolysis into two ADP molecules (see [13]). All ApnA are the In spite of obvious heterogeneity of the data on ApnA substrates for this enzyme, if n s 3. At least one of the reac- functions some interesting consequences follow from the re- tion products should be NDP. No speci¢c Ap4A hydrolytic cent developments in this ¢eld. In most studies on ApnA the activity was detected in S. cerevisiae. Instead, two isoenzymes attention was drawn to the absolute concentrations of various catalysing Ap4A phosphorolysis have been uncovered (see ApnA derivatives and often to only one of them, e.g. solely to [12,13]). Recently, the same activity was found in cyanobac- Ap4A. Now it has become evident that the whole set of ApnA teria in addition to the symmetrical hydrolase activity [69,70]. from 2 to 6 phosphates possesses biological activity and there- Reaction catalysed by Ap4A phosphorylase is described as fore the analysis should be limited to one or two derivatives. Ap4A+PiIATP+ADP. The number of phosphate residues in ApnA bound to AP- Most of the higher eukaryotes metabolise Ap4A molecules NAB may modulate signal transduction. For instance, if AP- by asymmetrical hydrolysis, resulting in formation of ATP NAB alternatively binds Ap4AorAp3A it resembles the bind- and AMP (reviewed in [13]). Among the lower eukaryotes ing of GTP and GDP (also one phosphate group di¡erence) the same activity was found in ¢ssion yeast S. pombe [77]. to G proteins, known to be signal transducers. In G proteins Ap4A hydrolases are speci¢c for tetraphosphates, although the GTP/GDP ratio is regulated by GTPase activity of the G they can also cleave dinucleotides with ¢ve and six phos- proteins while in APNAB it might be regulated non-enzymati- phates. ATP should be one of the reaction products and cally by di¡erent a¤nity constants towards ApnA and Ap3A is poor substrate. There is a speci¢c Ap3A hydrolysing Apn31A for the same APNAB. If so, a selective synthesis activity, found in many higher eukaryotes. or/and degradation of a certain type of ApnA may cause Properties of some ApnA hydrolases are listed in Table 4. the shift in proportion of various ApnA derivatives, which According to their primary structures these enzymes are di- may be utilised by the cell as a signal. TrpRS speci¢cally vided into two groups. Human Ap3A hydrolase Fhit (fragile produces Ap3A [61] while Fhit degrades it and therefore the histidine triad) and S. pombe Ap4A hydrolase belong to the Ap4A/Ap3A ratio can be modulated by changing the activity HIT family and possess 52% sequence identity [56,71]. of TrpRS/Fhit pair. Possibly, other aaRS can form analogous Although these two enzymes di¡er in their speci¢city toward pairs with Ap4A hydrolases regulating the Ap4A level as well. 2 Ap4A and Ap3A both activities are inhibited by Zn but not Therefore, the Ap4A/Ap3A ratio may be controlled by many 3 by F ions which are inhibitory toward other Ap4A hydrol- proteins. The idea that Ap3A/Ap4A ratio is essential for func- ases [72]. Sequences of human and porcine Ap4A hydrolases tional activity of these ApnA was already put forward [84]. are highly conserved (88% identity) [73] and they display sig- The protective e¡ect of Fhit towards tumourigenicity [77] ni¢cant homology to lupin enzyme [74]. Enzymatic properties implies that Ap3A and/or Fhit is directly implicated in control of these three enzymes are similar. Although Ap4A hydrolase of cell growth and proliferation. No mechanism was suggested from Torpedo presynaptic plasma membranes di¡ers from the so far to explain the antitumourigenic role of Fhit. Since others by the lack of discrimination between Ap3A and Ap4A, apoptosis in human cultured cells is associated with a decrease nevertheless it seems to belong to the same group. in Ap3A level [40], Fhit by binding or hydrolysing Ap3A may Discovery of the FHIT gene and Ap3A hydrolysing activity induce apoptosis of transformed cells. As mentioned above of Fhit [56,71] is probably one of the most interesting ¢ndings Fhit is a close structural analog of PKCI, a well-known reg- providing a clue for the in vivo function of Ap3A. Fhit be- ulator of cellular signalling pathways, therefore, it may appear longs to the large and ubiquitous group of proteins named that binding of Ap3A to PKCI can modulate the PKCI ac- HIT family due to a conserved His-X-His-X-His-X-X se- tivity as well. Furthermore, Ap3A serves as preferred substrate quence motif (where X is a hydrophobic amino acid residue) (primer) [85] for the interferon-inducible enzyme 2-5A synthe- [75,76]. The FHIT gene was identi¢ed in fragile locus p14.2 of tase and the adenylated derivatives, Ap3A(pA)n may retain human chromosome 3 [71]. Exhaustive genetic analysis of this their substrate properties toward Fhit and after hydrolysis locus showed deletions and other abnormalities in several generate ppA(pA)n (n = 1^6) which in turn may activate the cancer cell lines and primary tumours, leading to the proposal third interferon-inducible enzyme, the RNase L [85]. Appar- that the FHIT gene belongs to the family of the tumour sup- ently, Fhit via Ap3A is tightly interrelated with a cytokine pressor genes and that the Fhit protein may play an essential signalling pathway known to possess strong antiproliferative role in the oncogenesis [77]. activity. The recent important observation [77] that Fhit re- The crystal structures of Fhit and its complexes with sub- mained active as a tumour suppressor even when His96Asn strate analogs were resolved by multiwavelength anomalous mutant lacking hydrolytic activity was introduced into cancer di¡raction [78,79]. This data suggest a metal-independent cells ¢ts nicely to our model where FhitWAp3A complex is mechanism of catalysis involving the formation of a covalent assumed to be a signal (Fig. 2) rather than Fhit or Ap3A nucleotidyl phosphohistidyl intermediate. Proposed mecha- on their own. The hypothesis that tumour suppressor activity nism and the assignment of the functional roles of the con- is associated with Fhit involvement in interferon and/or apop- served His residues ¢t well with the results of site-directed totic and/or PKCI regulatory circuits via FhitWAp3A complex

FEBS 20186 7-5-98 162 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163 is experimentally testable. If proven, it will lead to signi¢cant [29] Pohl, U., Ogilvie, A., Lamontagne, D. and Busse, R. (1991) Am. advancement in our understanding of Ap A functions in gen- J. Physiol. 260, 1692^1697. n [30] Holden, C.P., Padua, R.A. and Geiger, J.D. (1996) J. Neuro- eral and tumour suppression in particular. chem. 67, 574^580. In conclusion, it seems possible to consider the whole ApnA [31] Ralevic, V., Hoyle, C.H. and Burnstock, G.J. (1995) J. Physiol. family as a new class of signalling molecules used by eukary- Lond. 483, 703^713. otic cells to regulate many house-keeping and specialised func- [32] van der Giet, M., Khattab, M., Borgel, J., Schulter, H. and Zidek, W. (1997) Br. J. Pharmacol. 120, 1453^1460. tions. This class of molecules de¢nitely deserves further and [33] Zamecnik, P.C., Kim, B., Gao, M.J., Taylor, G. and Blackburn, more detailed biochemical analysis. G.M. (1992) Proc. Natl. Acad. Sci. USA 89, 2370^2373. [34] Edgecombe, M.H.S., Craddock, D.C., Smith, A.G., McLennan, Acknowledgements: L.L.K. and L.Yu.F. acknowledge the support via A.G. and Fisher, M.J. (1997) Biochem. J. 323, 451^456. Ecole Normale Supeèrieure (Chaire Internationale Blaise Pascal award [35] Delaney, S.M., Blackburn, G.M. and Geiger, J.D. (1997) Eur. by Reègion Ile-de-France). We thank Kyrill Turpaev for helpful com- J. Pharmacol. 332, 35^42. ments. [36] Nishimura, A., Moriya, S., Ukai, H., Nagai, K., Wachi, M. and Yamada, Y. (1997) Genes Cells 2, 401^413. [37] Weinmann-Dorsch, C., Hedle, A., Grummt, I., Albert, W., Fer- References dinand, F., Friis, R.R., Pierron, G., Moll, W. and Grummt, F. (1984) Eur. J. Biochem. 138, 179^185. [1] Zamecnik, P. and Stephenson, M. (1969) in: H.M. Kalkar, H. [38] Vartanian, A., Narovlyansky, A., Amchenkova, A., Turpaev, K. Klenow, A. Munch-Petersen, M. Ottesen and J.M. Thaysen and Kisselev, L. (1996) FEBS Lett. 381, 32^34. (Eds.), The Role of Nucleotides for the Function and Conforma- [39] Bonaventura, C., Cashon, R., Colacino, J.M. and Hilderman, tion of Enzymes, Munksgaard, Copenhagen, pp. 276^291. R.H. (1992) J. Biol. Chem. 267, 4652^4657. [2] Rapaport, E. and Zamecnik, P. (1976) Proc. Natl. Acad. Sci. [40] Vartanian, A., Prudovsky, I., Suzuki, H., Dalpra, I. and Kisselev, USA 73, 3984^3988. L. (1997) FEBS Lett. 415, 160^162. [3] Lee, P.C., Bochner, B.R. and Ames, B.N. (1983) J. Biol. Chem. [41] Baxi, M.D. and Vishwanatha, J.K. (1995) Biochemistry 34, 9700^ 258, 6827^6834. 9707. [4] Lee, P.C., Bochner, B.R. and Ames, B.N. (1983) Proc. Natl. [42] Filonenko, V.V., Beresten, S.F., Rubikaite, B.I. and Kisselev, Acad. Sci. USA 80, 7496^7500. L.L. (1989) Biochem. Biophys. Res. Commun. 161, 481^488. [5] Varshavsky, A. (1983) Cell 34, 711^712. [43] Popenko, V.I., Cherny, N.E., Beresten, S.F., Ivanova, J.L., Filo- [6] Baker, J.C. and Jacobson, M.K. (1986) Proc. Natl. Acad. Sci. nenko, V.V. and Kisselev, L.L. (1993) Eur. J. Cell Biol. 62, 248^ USA 83, 2350^2352. 258. [7] Johnstone, D.B. and Farr, S.B. (1991) EMBO J. 10, 3897^3904. [44] Westfall, T.D., McIntyre, C.A., Obeid, S., Bowes, J., Kennedy, [8] Zamecnik, P. (1983) Anal. Biochem. 134, 1^10. C. and Sneddon, P. (1997) Br. J. Pharmacol. 121, 57^62. [9] Perret, J., Hepburn, A., Cochaux, P., Van Sande, J. and Du- [45] Edgecombe, M., McLennan, A.G. and Fisher, M.J. (1996) Bio- mont, J.E. (1990) Cell. Signal. 2, 57^65. chem. J. 314, 687^693. [10] Baxi, M.D. and Vishwanatha, J.K. (1995) J. Pharmacol. Toxicol. [46] Chavan, A.J., Haley, B.E., Volkin, D.B., Mar¢a, K.E., Verticelli, Methods 33, 121^128. A.M., Bruner, M.W., Draper, J.P., Burke, C.J. and Middaugh, [11] Plateau, P. and Blanquet, S. (1992) in: A.G. McLennan (Ed.), C.R. (1994) Biochemistry 33, 7193^7202. Ap4A and Other Dinucleoside Polyphosphates, CRC Press, Boca [47] Hilderman, R.H., Lilien, J.E., Zimmerman, J.K., Tate, D.H., Raton, pp. 63^81. Dimmick, M.A. and Jones, G.B. (1994) Biochem. Biophys. [12] Plateau, P. and Blanquet, S. (1994) Adv. Microbiol. Physiol. 36, Res. Commun. 200, 749^755. 81^107. [48] Walker, J., Bossman, P., Lackey, B.R., Zimmerman, J.K., Dim- [13] Guranowski, A. and Sillero, A. (1992) in: A.G. McLennan (Ed.), mick, M.A. and Hilderman, R.H. (1993) Biochemistry 32, 14009^ Ap4A and Other Dinucleoside Polyphosphates, CRC Press, Boca 14014. Raton, pp. 205^228. [49] Walker, J. and Hilderman, R.H. (1993) Biochemistry 32, 3119^ [14] Ogilvie, A., Blasius, R., Schulze-Loho¡, E. and Sterzel, R.B. 3123. (1996) J. Auton. Pharmacol. 16, 325^328. [50] Walker, J., Lewis, T.E., Pivorun, E.P. and Hilderman, R.H. [15] Jovanovic, A., Jovanovic, S., Mays, D.C., Lipsky, J.J. and (1993) Biochemistry 32, 1264^1269. Terzic, A. (1998) FEBS Lett., in press. [51] Baxi, M.D., McLennan, A.G. and Vishwanatha, J.K. (1994) Bio- [16] Bryan, J. and Aguilar-Bryan, L. (1997) Curr. Opin. Cell Biol. 9, chemistry 33, 14601^14607. 553^559. [52] Vishwanatha, J.K. and Wei, Z. (1992) Biochemistry 31, 1631^ [17] Jovanovic, A., Alekseev, A.E. and Terzic, A. (1997) Biochem. 1635. Pharmacol. 54, 219^225. [53] Zourgui, L., Baltz, D., Baltz, T., Oukerro, F. and Tarrago-Lit- [18] Pintor, J., King, B.F., Miras-Portugal, M.T. and Burnstock, G. vak, L. (1988) Nucleic Acids Res. 16, 2913^2929. (1996) Life Sci. 59, PL179^184. [54] Hilderman, R.H., Martin, M., Zimmerman, J.K. and Pivorun, [19] Gasmi, L., McLennan, A.G. and Edwards, S.W. (1997) Immu- E.B. (1991) J. Biol. Chem. 266, 6915^6918. nology 90, 154^159. [55] Gasmi, L., McLennan, A.G. and Edwards, S.W. (1996) Blood 87, [20] Gasmi, L., McLennan, A.G. and Edwards, S.W. (1996) Br. 3442^3449. J. Haematol. 95, 637^639. [56] Barnes, L.D., Garrison, P.N., Siprashvili, Z., Guranowski, A., [21] Keppens, S. (1996) Biochem. Pharmacol. 52, 441^445. Robinson, A.K., Ingram, S.W., Croce, C.M., Ohta, M. and [22] Jovanovic, A., Zhang, S., Alekseev, A.E. and Terzic, A. (1996) Huebner, K. (1996) Biochemistry 35, 11529^11535. Eur. J. Physiol. 431, 800^802. [57] Liu, G., Bryant, R.T. and Hilderman, R.H. (1996) Biochemistry [23] Ripoll, C., Martin, F., Manuel Rovira, J., Pintor, J., Miras-Por- 35, 197^201. tugal, M.T. and Soria, B. (1996) Diabetes 45, 1431^1434. [58] Feussner, K., Guranowski, A., Kostka, S. and Wasternack, C. [24] Schulze-Loho¡, E., Zanner, S., Ogilvie, A. and Sterzel, R.B. (1996) Z. Naturforsch. C 51, 477^486. (1995) Hypertension 26, 899^904. [59] Fuge, E.K. and Farr, S.B. (1993) J. Bacteriol. 175, 2321^2326. [25] Green, A.K., Cobbold, P.H. and Dixon, C.J. (1995) Biochem. [60] Burnstock, G. (1997) Neutopharmacology 36, 1127^1139. J. 310, 629^635. [61] Kisselev, L. and Wolfson, A. (1994) Progr. Nucleic Acids Res. [26] Gasmi, L., McLennan, A.G. and Edwards, S.W. (1994) Biochem. Mol. Biol. 48, 83^142. Biophys. Res. Commun. 202, 218^224. [62] Merkulova, T., Kovaleva, G. and Kisselev, L. (1994) FEBS Lett. [27] Klishin, A., Lozovaya, N., Pintor, J., Miras-Portugal, M.T. and 350, 287^290. Krishtal, O. (1994) Neuroscience 58, 235^236. [63] Belrhali, H., Yaremchuk, A., Tukalo, M., Berthet, C.C., Rasmus- [28] Chan, P.J., Su, B.C. and Tredway, D.R. (1991) Arch. Androl. 27, sen, B., Bosecke, P., Diat, O. and Cusack, S. (1995) Structure 3, 103^108. 341^352.

FEBS 20186 7-5-98 L.L. Kisselev et al./FEBS Letters 427 (1998) 157^163 163

[64] Sillero, M.A.G., Guranowski, A. and Sillero, A. (1991) Eur. Petsko, G.A. and Lowenstein, J.M. (1997) Nature Struct. Biol. 4, J. Biochem. 202, 507^513. 231^238. [65] Guranowski, A., Just, G., Holler, E. and Jakubowski, H. (1988) [77] Siprashvili, Z., Sozzi, G., Barnes, L.D., McCue, P., Robinson, Biochemistry 27, 2959^2964. A.K., Eryomin, V., Sard, L., Tagliabue, E., Greco, A., Fusetti, [66] Guranowski, A., Sillero, M.A.G. and Sillero, A.S. (1990) FEBS L., Schwartz, G., Pierotti, M.A., Croce, C.M. and Huebner, K. Lett. 271, 215^218. (1997) Proc. Natl. Acad. Sci. USA 94, 13771^13776. [67] Brevet, A., Chen, J., Jeveque, F., Plateau, P. and Blanquet, S. [78] Lima, C.D., Klein, M.G. and Hendrickson, W.A. (1997) Science (1989) Proc. Natl. Acad. Sci. USA 86, 8275^8279. 278, 286^290. [68] McLennan, A.G., Mayers, E., Walker, S.I. and Chen, H. (1995) [79] Lima, C.D., D'Amico, K.L., Naday, I., Rosenbaum, G., West- J. Biol. Chem. 270, 3706^3709. brook, E.M. and Hendrickson, W.A. (1997) Structure 5, 763^ [69] McLennan, A.G., Meyers, E. and Adams, D.G. (1996) Biochem. 774. Soc. Trans. 24, 417S. [80] Lima, C.D., Klein, M.G., Weinstein, I.B. and Hendrickson, W.A. [70] McLennan, A.G., Mayers, E. and Adams, D.G. (1996) Biochem. (1996) Proc. Natl. Acad. Sci. USA 93, 5357^5362. J. 320, 795^800. [81] Huang, Y., Garrison, P.N. and Barnes, L.D. (1995) Biochem. [71] Ohta, M., Inoue, H., Cotticelli, M.G., Kastury, K., Ba¡a, R., J. 312, 925^932. Palazzo, J., Siprashvili, Z., Mori, M., McCue, P., Druck, T., [82] Mateo, J., Rotllan, P., Marti, E., Gomoz De Aranda, I., Solsona, Croce, C.M. and Huebner, K. (1996) Cell 84, 587^597. C. and Miras-Portugal, M.T. (1997) Biochem. J. 323, 677^684. [72] Guranowski, A. (1990) FEBS Lett. 262, 205^208. [83] Thorne, N.M.H., Hankin, S., Wilkinson, M.C., Nunez, C., Bar- [73] Hankin, S., Wintero, A.K. and McLennan, A.G. (1997) Int. raclough, R. and McLennan, A.G. (1995) Biochem. J. 311, 717^ J. Biochem. Cell Biol. 29, 317^323. 721. [74] Maksel, D., Guranowski, A., Ilgoutz, S.C., Moir, A., Blackburn, [84] Kisselev, L., Haenni, A.-L. and Frolova, L. (1993) Trends Bio- M.G. and Gayler, K.R. (1998) Biochem. J. 329, 313^319. chem. Sci. 18, 263^267. [75] Robinson, K. and Aitken, A. (1994) Biochem. J. 304, 662^664. [85] Turpaev, K., Hartmann, R., Kisselev, L. and Justesen, J. (1997) [76] Brenner, C., Garrison, P., Gilmour, J., Peisach, D., Ringe, D., FEBS Lett. 408, 177^181.

FEBS 20186 7-5-98