Adenine dinucleotide-mediated activation of glycogen phosphorylase in isolated liver cells

Purinergic signalling in the liver in health and disease

Purinergic signalling is involved in both the physiology and pathophysiology of the liver. Hepatocytes, Kupffer cells, vascular endothelial cells and smooth muscle cells, stellate cells and cholangiocytes all express purinoceptor subtypes activated by adenosine, adenosine 5'-triphosphate, adenosine diphosphate, uridine 5'-triphosphate or UDP. Purinoceptors mediate bile secretion, glycogen and lipid metabolism and indirectly release of insulin. Mechanical stress results in release of ATP from hepatocytes and Kupffer cells and ATP is also released as a cotransmitter with noradrenaline from sympathetic nerves supplying the liver. Ecto-nucleotidases play important roles in the signalling process. Changes in purinergic signalling occur in vascular injury, inflammation, insulin resistance, hepatic fibrosis, cirrhosis, diabetes, hepatitis, liver regeneration following injury or transplantation and cancer. Purinergic therapeutic strategies for the treatment of these pathologies are being explored.

Studies on the effect of diadenlyated nucleotides on calcium mobilization and prostacyclin synthesis in bovine aortic endothelial cells

Extracellular adenine dinucleotides are modulators of blood vessel tone. We have previously demonstrated that Ap(2)A and Ap(4)A induce the synthesis of nitric oxide (NO) from bovine aortic endothelial cells (BAEC) while Ap(3)A and Ap(5)A do not [FEBS Lett. 427 (1998) 320; Arch. Biochem. Biophys. 364 (1999) 280.]. In this communication we determine the effect of Ap(x)As (x=2-5) on prostacyclin (PGI(2)) synthesis and Ca(2+) mobilization in BAEC. Ap(2)A and Ap(4)A significantly enhanced the synthesis of PGI(2) while Ap(3)A and Ap(5)A do not. These data support the notion that Ap(2)A and Ap(4)A are vasodilators. All four dinucleotides significantly enhanced Ca(2+) mobilization over basal levels. Ap(5)A and Ap(3)A enhanced 2.0 and 1.6 times more Ca(2+) release than Ap(4)A, respectively. Since neither Ap(5)A nor Ap(3)A enhanced the synthesis of either PGI(2) or NO but did mobilize Ca(2+), these data support the hypothesis that in BAEC Ca(2+) release is localized or compartmentalized.

p(1),p(4)-diadenosine 5'-tetraphosphate induces the uptake of arginine and citrulline by a pore on the plasma membrane of bovine aortic endothelial cells

We have previously demonstrated that p(1),p(4)-diadenosine 5'-tetraphosphate (Ap(4)A) induces the release of nitric oxide (NO) and modulates the uptake of extracellular L-arginine (L-Arg) and L-citrulline (L-Cit) by bovine aortic endothelial cells (BAEC) [Hilderman, R.H. and Christensen, E.F. (1998) FEBS Lett. 427, 320-324 and Hilderman, R.H., Casey, T.E. and Pojoga, L.H. (2000) Arch. Biochem. Biophys. 375, 124-130]. In this communication we report that extracellular Ap(4)A enhances the uptake of L-Arg and L-Cit through a pore on the plasma membrane of BAEC that is selective for these two amino acids. We also demonstrate that Ap(2)A, which induces NO release, enhances L-Arg uptake while Ap(5)A, a vasoconstrictor, does not enhance the uptake of L-Arg. The potential physiological significance of the uptake of these two amino acids in relation to NO synthesis is discussed.

Dinucleoside polyphosphates-friend or foe?

Despite being known for over 30 years, the functions of the dinucleoside polyphosphates, such as diadenosine 5',5"'-P(1), P(4)-tetraphosphate (Ap(4)A) and diadenosine 5',5"'-P(1), P(3)-triphosphate (Ap(3)A), are still unclear. On the one hand, they may have important signalling functions, both inside and outside the cell (friend), while on the other hand, they may simply be the unavoidable by-products of certain biochemical reactions, which, if allowed to accumulate, would be potentially toxic through their structural similarity to ATP and other essential mononucleotides (foe). Here, the occurrence, synthesis, degradation, and proposed functions of these compounds are briefly reviewed, along with some new data and recent evidence supporting roles for Ap(3)A and Ap(4)A in the cellular decision making processes leading to proliferation, quiescence, differentiation, and apoptosis. Hypotheses are forwarded for the involvement of Ap(4)A in the intra-S phase DNA damage checkpoint and for Ap(3)A and the pFhit (fragile histidine triad gene product) protein in tumour suppression. It is concluded that the roles of friend and foe are not incompatible, but are distinguished by the concentration range of nucleotide achieved under different circumstances.

Inhibitory effects of some purinergic agents on ecto-ATPase activity and pattern of stepwise ATP hydrolysis in rat liver plasma membranes

Inhibitory effects of various purinergic compounds on the Mg(2+)-dependent enzymatic hydrolysis of [(3)H]ATP in rat liver plasma membranes were evaluated. Rat liver enzyme ecto-ATPase has a broad nucleotide-hydrolyzing activity, displays Michaelis-Menten kinetics with K(m) for ATP of 368+/-56 microM and is not sensitive to classical inhibitors of the ion-exchange and intracellular ATPases. P2-antagonists and diadenosine tetraphosphate (Ap(4)A) progressively and non-competitively inhibited ecto-ATPase activity with the following rank order of inhibitory potency: suramin (pIC(50), 4.570)>Reactive blue 2 (4.297)&z.Gt;Ap(4)A (3. 268)>pyridoxalphosphate-6-azophenyl-2',4'-disulfonic acid (PPADS) (2. 930). Slowly hydrolyzable P2 agonists ATPgammaS, ADPbetaS, alpha, beta-methylene ATP and beta,gamma-methylene ATP as well as the diadenosine polyphosphates Ap(3)A and Ap(5)A did not exert any inhibitory effects on the enzyme activity at concentration ranges of 10(-4)-10(-3) M. Thin-layer chromatography analysis of the formation of [(3)H]ATP metabolites indicated the presence of other enzyme activities on liver surface (ecto-ADPase and 5'-nucleotidase), participating in concert with ecto-ATPase in the nucleotide hydrolysis through the stepwise reactions ATP-->ADP-->AMP-->adenosine. A similar pattern of sequential [(3)H]ATP dephosphorylation still occurs in the presence of ecto-ATPase inhibitors suramin, Ap(4)A and PPADS, but the appearance of the ultimate reaction product, adenosine, was significantly delayed. In contrast, hydrolysis of [(3)H]ATP in the presence of Reactive blue 2 only followed the pattern ATP-->ADP, with formation of the subsequent metabolites AMP and adenosine being virtually eliminated. These data suggest that although nucleotide-binding sites of ecto-ATPase are distinct from those of P2 receptors, some purinergic agonists and antagonists can potentiate cellular responses to extracellular ATP through non-specific inhibition of the ensuing pathways of purine catabolism.

P(1),P(4)-Diadenosine 5'-tetraphosphate modulates l-arginine and l-citrulline uptake by bovine aortic endothelial cells

We have previously demonstrated that P(1),P(4)-diadenosine 5'-tetraphosphate (Ap(4)A) interacts with high-affinity and low-affinity binding sites on the bovine aortic endothelial cell (BAEC) surface. In this report we demonstrate that Ap(4)A interaction with the lower affinity site modulates l-arginine (l-Arg) and l-citrulline (l-Cit) uptake by BAEC. Competition uptake studies demonstrate that l-Arg and l-Cit uptake occurs through a common transporter system that is sensitive to Ap(4)A. Evidence is also presented that is consistent with Ap(4)A modulating l-Arg uptake by increasing the affinity of l-Arg for the transporter.

Diadenosine polyphosphates and the control of cyclic AMP concentrations in isolated rat liver cells

Extracellular diadenosine polyphosphates (Ap(n)A), through their interactions with appropriate P(2) receptors, influence a diverse range of intracellular activities. In particular, Ap(4)A stimulates alterations in intracellular calcium homeostasis and subsequent activation of glycogen breakdown in isolated liver cells. Here we show that, like ATP, Ap(4)A and other naturally occurring diadenosine polyphosphates attenuate glucagon-stimulated accumulation of cyclic AMP in isolated rat liver cells. The characteristics of Ap(4)A- and ATP-dependent modulation of glucagon-stimulated cyclic AMP accumulation are similar. These results are discussed in the context of the repertoire of intracellular signalling processes modulated by extracellular nucleotides.

Acyl-CoA synthetase catalyzes the synthesis of diadenosine hexaphosphate (Ap6A)

The synthesis of diadenosine hexaphosphate (Ap6A), a potent vasoconstrictor, is catalyzed by acyl-CoA synthetase from Pseudomonas fragi. In a first step AMP is transferred from ATP to tetrapolyphosphate (P4) originating adenosine pentaphosphate (p5A) which, subsequently, is the acceptor of another AMP moiety from ATP generating diadenosine hexaphosphate (Ap6A). Diadenosine pentaphosphate (Ap5A) and diadenosine tetraphosphate (Ap4A) were also synthesized in the course of the reaction. In view of the variety of biological effects described for these compounds the potential capacity of synthesis of diadenosine polyphosphates by the mammalian acyl-CoA synthetases may be relevant.

Evidence for a G protein-coupled diadenosine-5',5'''-P1,P4-tetraphosphate (Ap4A) receptor binding site in lung membranes from rat

Nucleotide receptors are of considerable importance in the treatment of lung diseases, such as cystic fibrosis. Because diadenosine polyphosphates may also be of significance as signalling molecules in lung, as they are in a variety of tissues, in the present work we investigated the binding sites for [3H]diadenosine-5',5'''-P1,P4-tetraphosphate (Ap4A) in plasma membranes from rat lung and studied their possible coupling to G proteins. We present evidence for a single high-affinity binding site for [3H]Ap4A with similar affinity for other diadenosine polyphosphates ApnA (n = 2 to 6). Displacement studies with different nucleotides revealed that the [3H]Ap4A binding site was different from P2X and P2Y2 receptor binding sites. Pretreatment of lung membranes with GTPgammaS or GTP in the presence of Mg2+ increased the Ki for Ap4A from 91 nM to 5.1 microM, which is indicative of G protein coupling. The putative coupling to G proteins was further confirmed by the enhancement of [35S]GTPgammaS binding (to Galpha proteins) to lung membranes by Ap4A (63% increase over basal) in a concentration-dependent manner. Therefore, our data for the first time provide evidence of a G protein-coupled Ap4A binding site in lung membranes.

Characterization of ATP and P2 agonists binding to the cardiac plasma membrane P1,P4-diadenosine 5'-tetraphosphate receptor

P1,P4-Diadenosine 5'-tetraphosphate (Ap4A) acts as an extracellular modulator through its interaction with purinoceptors. Our laboratory has demonstrated the presence of an Ap4A receptor in cardiac tissue [1,2]. Due to the rapid hydrolysis of ATP by cardiac membranes the relationship of ATP and Ap4A binding to purinoceptors on cardiac membranes has not been characterized. In this communication we used two approaches to determine the relationship of ATP to the Ap4A receptor. Radioligand binding carried out with [alpha-32P]Ap4A and adenosine 5'-O-¿3-thiotriphosphate¿ ([gamma-35S]ATPgammaS) demonstrates the presence of a single high affinity binding site for Ap4A and the presence of two binding sites for ATPgammaS. The second approach utilized immunoaffinity purified Ap4A receptor that was shown to be free of ATPase and Ap4Aase activities. Non-radiolabeled Ap4A and ATPgammaS effectively inhibited photocrosslinking of [alpha-32P]8-N3Ap4A to the receptor polypeptide while ATP was a much less effective inhibitor. Furthermore, on plasma membranes [alpha-32P]8-N3Ap4A photocrosslinked to only a 50 kDa polypeptide. These data are consistent with Ap4A interacting with a homogeneous population of receptors on cardiac plasma membranes but with ATP having a low affinity for the receptor.

Steady-state binding of [3H]ATP to rat liver plasma membranes and competition by various purinergic agonists and antagonists

Steady-state analysis of nucleotide-binding sites on rat liver plasma membranes was carried out using 3H-labelled ATP as radioligand under complete inhibition of ecto-ATPase activity by excess EDTA. Binding of [3H]ATP to the membranes is saturable, reversible and apparently involves one population of specific binding sites with Kd of about 90 nM and binding capacity (Bmax) of 15 pmol/mg protein. A broad spectrum of purinergic agonists and antagonists was examined as potential inhibitors of the measured binding. The displacement studies showed the following rank order of inhibitory potency for [3H]ATP-binding sites (pIC50 values in parentheses): ATPgammaS (7.49)>2-MeSATP (7.18)>ATP (6.91)>ADPbetaS (6.64)>/=ADP (6.56)>>RB2 (6.14)>>suramin (5.40)>>Ap4A (4. 57)>alpha,beta-MeATP (4.19)>/=beta,gamma-MeATP (3.97). AMP, adenosine, Ap5A, PPADS, beta-glycerophosphate as well as non-adenine nucleoside triphosphates GTP, UTP and CTP did not exert any effect on the measured binding at concentration ranges of 10-6-10-4 M. In order to ascertain whether ATP and its analogues are capable of interacting with the same binding domain, 2-MeSATP and ADP were treated as alternative ligands that could compete with unlabelled ATP for its binding sites. A 2-fold increase of Kd value for ATP-receptor interaction was observed in the presence of 2-MeSATP (60 nM) or ADP (250 nM) without any modulation of Bmax value, confirming that inhibitory effects of these compounds are competitive in nature. These studies demonstrate that ATP and its analogues are able to interact with a single binding domain on liver plasma membranes, which may be identified as ligand-binding component of P2 purinoceptors of the P2Y1 subtype.

Diadenosine polyphosphate-mediated activation of phospholipase D in isolated rat liver cells

Diadenosine polyphosphates (ApnAs) can, through interaction with appropriate purinoceptors, affect a range of cellular activities. Ap4A, the most prominent naturally occurring diadenosine polyphosphate, stimulates alterations in intracellular calcium homeostasis and subsequent activation of glycogen breakdown in isolated liver cells. Here we show that Ap4A, and other naturally occurring diadenosine polyphosphates, also stimulates phospholipase D (PLD) activity in isolated rat liver cells. The characteristics of Ap4A-mediated activation of PLD are similar to those for the activation of PLD by extracellular ATP. These results are discussed in the context of the relation between diadenosine polyphosphate- and adenine mononucleotide-mediated cellular signalling processes.

P1,P4-diadenosine 5' tetraphosphate induces nitric oxide release from bovine aortic endothelial cells

Since the infusion of P1,P4-diadenosine 5' tetraphosphate (Ap4A) into animal models induces vasodilation [1,2], the present study was performed to determine whether Ap4A induces the release of nitric oxide (NO) from endothelial cells. Ap4A induced NO release was 4.2-fold greater than the amount of NO released under basal condition. Ap4A induced NO release was inhibited by NG-nitro-L-arginine (L-NNA) and this inhibition was reversed by L-Arg. In addition, EGTA inhibits Ap4A induced NO release. These data are consistent with Ap4A inducing the release of NO from endothelial cells through the activation of endothelial nitric oxide synthase.

Effects of diadenosine triphosphate and diadenosine tetraphosphate on rat liver cells. Differences and similarities with ADP and ATP

Liver cells possess multiple types of purinoceptors that mediate the effects of extracellular nucleotides. Like ADP and ATP, the dinucleotides diadenosine triphosphate (Ap3A) and diadenosine tetraphosphate (Ap4A) fully activated glycogen phosphorylase, with ED50 values of 0.31 microM and 1.3 microM, respectively. At variance with ATP, neither the dinucleotides nor ADP significantly increased the levels of IP3.Ap4A (and also ADP) moderately increased IP3 (+/- 72%) whereas Ap3A was completely ineffective. Like ATP, Ap3A, Ap4A, and ADP inhibited the cAMP increase after glucagon. Phorbol-12-myristate-13-acetate (PMA) pretreatment of the hepatocytes clearly inhibited the glycogenolytic potency of Ap3A and ADP, but had only a minor effect on the potency of Ap4A or ATP. It is concluded that, depending upon the effect studied (glycogenolytic effect with or without PMA, increasing IP3 potency, or inhibition of cAMP increase), different analogies between the agonists studied emerged, indicating the complexity of the interaction of ATP and its analogues with liver purinoceptors and/or of the transduction mechanism(s) initiated by the different nucleotides.

Characterization of the binding of diadenosine 5',5'''-P1,P4-tetraphosphate (Ap4A) to rat liver cell membranes

Diadenosine polyphosphates present in the extracellular environment can, through interaction with appropriate purinoceptors, influence a range of cellular activities. Here we have investigated the nature of the ligand:receptor interactions involved in diadenosine 5',5'''-P1, P4-tetraphosphate (Ap4A)-mediated stimulation of glycogen breakdown in isolated rat liver cells. [2-3H]Ap4A showed specific binding to both intact isolated liver cells and plasma membrane fractions prepared from isolated liver cells. HPLC analysis confirmed that binding was mediated by intact Ap4A and not by potential breakdown products (e.g. ATP, adenosine, etc.). Binding of [2-3H]Ap4A, to isolated liver cell plasma membrane preparations, was successfully displaced by a range of both naturally occurring and synthetic diadenosine polyphospates with the rank order potency Ap4A > or = Ap5A > Ap6A > Ap3A > Ap2A. [2-3H]Ap4A binding was not displaced by P1 effectors but was successfully displaced by a range of P2 effectors with the rank order potency 2-methylthio-ATP > ATP > ATP > or = adenosine 5'-[alpha beta-methylene]triphosphate > adenosine 5'-[beta gamma-methylene]triphospate. These findings are consistent with the interaction of Ap4A with a P2y-like subclass of purinoceptor and are discussed in relation to (1) the known purinoceptor populations in liver cell plasma membranes and (2) observations concerning the binding of diadenosine polyphosphates to purinoceptors in other tissues.

Cytosolic calcium responses to extracellular adenosine 5',5" '-P1,P4-tetraphosphate in PC12 cells

Binding of adenosine 5',5" '-P1,P4-tetraphosphate (Ap4A) to a purinoceptor on nerve growth factor-differentiated (NGF) pheochromacytoma (PC12) cells modulated cytosolic Ca2+ levels. Both Ap4A and ATP elicited an influx of extracellular Ca2+, but both the sensitivity of the response and the flux profile were different. Preincubation of the PC12 cells with the compounds adenosine 5'-0-(2-thio)diphosphate (ADP-beta-S) and periodate-oxidized ATP had differential effects upon the Ap4A and ATP-induced response. These results indicate that Ap4A and ATP were either interacting with distinct purinoceptor subclasses or with the same purinoceptor with differing affinities. Simultaneous depolarization and application of either Ap4A or ATP to the PC12 cells induced an additive effect on the calcium flux. Preincubation with verapamil negated the effects of depolarization without significantly modifying the ligand-elicited Ca2+ fluxes, suggesting the presence of Ap4A ligand-gated channels that may function as modulators of PC12 cell function.

Cytosolic free Ca2+ oscillations induced by diadenosine 5',5"'-P1,P3-triphosphate and diadenosine 5',5"'-P1,P4-tetraphosphate in single rat hepatocytes are indistinguishable from those induced by ADP and ATP respectively

Diadenosine 5',5"'-P1,P3-triphosphate (Ap3A) and diadenosine 5',5"'-P1,P4-tetraphosphate (Ap4A) induce distinctive patterns of [Ca2+]i oscillations in single rat hepatocytes. We show here that [Ca2+]i oscillations induced by Ap3A and ADP are indistinguishable and that [Ca2+]i oscillations induced by Ap4A closely resemble those induced by ATP. These similarities embrace the following: (1) ADP and Ap3A invariably induce [Ca2+]i transients of short duration (approx. 9 s). Ap4A, like ATP, can induce, depending upon the individual cell, either transients of short duration (approx. 9 s), transients of much longer duration or a mixture of short and long transients within a single response. We show here that the pattern of oscillations induced by Ap4A is similar to that induced by ATP in the same hepatocyte. (2) Elevated intracellular cyclic AMP concentration modulates Ap3A-induced transients, like ADP-induced transients, through an increase in both the peak [Ca2+]i and the frequency of the transients. In contrast, Ap4A-induced transients, like ATP-induced transients, develop an increased duration or a sustained rise in [Ca2+]i, with no rise in peak [Ca2+]i. (3) Ap3A-induced transients, like ADP-induced transients, are abolished by low concentrations of the phorbol ester 4 beta-phorbol 12,13-dibutyrate (PDB; 5-10 nM), whereas long Ap4A-induced transients, like long ATP-induced transients, are refractory to high concentrations of PDB (100 nM). We propose that the [Ca2+]i oscillations induced in rat hepatocytes by Ap3A are mediated by the same purinoceptor that mediates the effects of ADP, whereas the oscillations induced by Ap4A are mediated by the same purinoceptor(s) that mediate the effects of ATP.

Adenine dinucleotide-mediated cytosolic free Ca2+ oscillations in single hepatocytes

Single rat hepatocytes microinjected with aequorin respond to Ca(2+)-mobilizing agonists, including ADP and ATP, with oscillations in cytosolic free Ca2+. We show here that single rat hepatocytes also respond to the adenine dinucleotides Ap3A and Ap4A with Ca2+ oscillations which resemble those induced by ADP and ATP.