Increased membrane-associated nucleoside diphosphate kinase activity as a possible basis for enhanced guanine nucleotide-dependent adenylate cyclase activity induced by picolinic acid treatment of simian virus 40-transformed normal rat kidney cells

Arabidopsis NDK1 is a Component of ROS Signaling by Interacting with Three Catalases

Plants sense various environmental stimuli and have specific signaling pathways to respond to these cues. We focused on light responsive components and found that NDKs were phosphorylated specifically after red light irradiation in Pisum sativum [Tanaka et al. (1998) J. Photochem. Photobiol. B 45: 113] and after blue light irradiation in Neurospora crassa [Oda and Hasunuma (1997) Mol. Gen. Genet. 256: 593, Ogura et al. (2001) J. Biol. Chem. 276: 21228]. We performed yeast two-hybrid screening using AtNDK1, the counterpart of NDK-P1 (Pisum sativum NDK1) in Arabidopsis, as bait, and isolated catalase3 (AtCat3). Interactions between AtNDK1-AtCAT1 and AtNDK1-AtCAT2 were also detected with the two-hybrid system. Non-denaturing two-dimensional gel electrophoresis of crude extracts from plants revealed that catalase and NDK activities co-migrated in the same area of the gel. Transgenic plants expressing AtNDK1 under control of the CaMV 35S promoter exhibited tolerance to paraquat and high ability to eliminate exogenous H2O2. These results indicate that AtNDK1 has a role in ROS response.

Extracellular nucleoside diphosphate kinase NM23/NDPK modulates normal hematopoietic differentiation

OBJECTIVE

We previously demonstrated the presence of nucleoside diphosphate kinase NDPK/NM23 in normal human plasma. It also was reported that extracellular NM23 could inhibit differentiation of certain hematopoietic cell lines. We further investigated the extracellular effect of NM23 on hematopoiesis by adding recombinant NM23-H1, NM23-H2, and NM23-H3 proteins to in vitro differentiation assays of normal human hematopoietic progenitors.

MATERIALS AND METHODS

To study the effect on the earlier stages of hematopoietic maturation, NM23 was added to serum-free pre-colony-forming unit (pre-CFU) assays starting from immature CD34++CD38- bone marrow cells. Serum-free CFU assays starting from CD34+ CD38+ bone marrow cells were used as a model for terminal hematopoietic differentiation.

RESULTS

In pre-CFU assays, none of the NM23 isoforms used significantly changed the expansion of CD34++CD38- cells, nor did NM23 alter the CD34++ CD38- cell lineage commitment. In contrast, terminal differentiation of CD34+CD38+ progenitor cells in CFU assays was significantly altered by addition of NM23 protein. More erythroid burst-forming units and fewer macrophage colonies were observed in cultures containing any of the NM23 isoforms examined. Similar effects were observed using the enzymatically inactive H118N mutant of NM23-H1, strongly suggesting that the observed effect is independent of the nucleoside diphosphate kinase activity of NM23.

CONCLUSION

We demonstrated a modulating effect of extracellular NM23 proteins on the terminal stages of normal hematopoietic differentiation. Therefore, the fairly high concentrations of NM23 constitutively present in plasma could have a physiologic role in supporting erythropoiesis and inhibiting excessive macrophage formation.

Evidence supporting the Nucleotide Axis Hypothesis: ATP release and metabolism by coronary endothelium

The Nucleotide Axis Hypothesis, defined and supported herein, proposes that ATP stimulates the release of vasoactive mediators from endothelium, including ATP itself. Here, we show rapid endothelium-dependent, agonist-stimulated ATP elaboration in coronary vessels of guinea pigs. Measurement of extracellular ADP metabolism in intact vessels results in the time- and substrate-dependent formation of ATP in the coronary perfusate in amounts greater than can be accounted for by release from endothelium alone. ATP formation by endothelial cells is saturable (K(M) = 38.5 micromol/l, where K(M) is substrate concentration at which rate is half-maximal.) and trypsin-sensitive, membranes from [gamma-(32)P]ATP-labeled cells support ADP-dependent transphosphorylation by a 20-kDa protein, Western blots reveal the presence of a nucleoside diphosphate kinase (NDPK) of approximately 20 kDa in endothelial membranes, and analysis of NDPK antibody binding by flow cytometry is consistent with the presence of an ecto-NDPK on cardiac endothelial cells. Sequencing of the endothelial cell ecto-NDPK reveals a predicted amino acid sequence with 85% identity to human Nm23-H1 and consistent with a protein whose properties may confer membrane association as well as sites of regulation of activity. Our data underscore the potential importance of a nucleotide axis in cardiac blood vessels.

Nucleotides regulate the binding affinity of the recombinant type A cholecystokinin receptor in CHO K1 cells

Cholecystokinin (CCK) receptors on rat pancreatic acinar cells display two binding affinity states in the presence of adeninine and guanine triphosphates with the effect of ATP mediated by the enzyme nucleoside diphosphate kinase. To determine whether this behavior was intrinsic to a single receptor protein we studied the binding affinity of CHO cells stably transfected with a cloned rat CCKA receptor. 125I-CCK binding to intact cells at 37 degrees C revealed two affinity states for CCK of Kd values 20 pM and 2.4 nM. Membranes prepared from these cells displayed a single affinity state for CCK but two affinity states could be restored in the presence of GTP[gamma S], ATP and ATP[gamma S] but not AMP-PCP. ATP and ATP[gamma S] but not AMP-PCP were substrates for nucleoside diphosphate kinase present in CHO cell membranes and transferred their terminal phosphate to GDP. These findings indicate that the interconvertible affinity states of the CCK receptor are inherent in a single receptor protein and that nucleoside diphosphate kinase mediates the effect of ATP to regulate these two affinity states.

Inhibition by cromoglycate and some flavonoids of nucleoside diphosphate kinase and of exocytosis from permeabilized mast cells

1. The anti-allergic compound, cromoglycate, is reported to possess affinity for, and to suppress the autophosphorylation of a 72kDa protein having the sequence of nucleoside diphosphate kinase (NDPK). 2. We have tested the ability of cromoglycate, and a panel of ten structurally related flavonoids of plant origin, to inhibit the NDPK reaction and the exocytotic process of permeabilized mast cells. The conditions of permeabilization (use of an isotonic medium based on sodium glutamate) were selected to ensure that NDPK activity would be an essential component in the induction of Ca(2+)-induced exocytosis in which ATP is required for generation of GTP. For comparison, we also measured the inhibition of exocytosis induced by GTP-gamma-S; this proceeds in the absence of ATP and bypasses the need for NDPK activity. 3. We found that cromoglycate does not discriminate between Ca2+ and GTP-gamma-S-induced exocytosis and is a poor inhibitor of NDPK activity. Concentrations in the millimolar range are required for inhibition of all these functions. By comparison, many of the flavonoids are effective at concentrations in the micromolar range. 4. While we were unable to discern any systematic relationships between their ability to inhibit the three functions, two compounds, quercetin and genistein, inhibit Ca(2+)-induced, but not GTP-gamma-S-induced exocytosis. Inhibition of the late stages of the stimulus-response pathway in mast cells by these compounds is therefore likely to be due to inhibition of NDPK and the consequent failure to generate GTP.

Effect of retinoic acid on Nm/23 nucleoside diphosphate kinase and components of cyclic adenosine monophosphate-dependent signalling in human neuroblastoma cell lines

The effects of retinoic acid on components of the cAMP-dependent signalling system were examined in two related human neuroblastoma cell lines SK-N-SH-F (SHF) and SK-N-SH-N (SHN). Retinoid treatment for a week significantly increased the concentration of intracellular cAMP and the levels of activity of protein kinase A and adenylate cyclase in both cell lines. Retinoic acid treatment also caused a very marked translocation of nucleoside diphosphate kinase from the cytosol to the membrane fraction. The increases in cyclic nucleotide and protein kinase A activity were observed to occur as early as within 1 and 2 days respectively and preceded or were concurrent with the onset of observable morphological differentiation. Results also indicated that agents which elevated intracellular cAMP caused neuronal differentiation and blunted retinoic acid-induced melanocytic differentiation in SHF cells. However, increases in cAMP brought about by treatment of SHF cells with retinoic acid alone were several-fold smaller and thus insufficient to induce neuritogenesis in these cells. The results as a whole indicate that one overall effect of retinoic acid treatment is to upgrade the activity of components of the cAMP-dependent signalling system in both neuroblastoma cell lines. However, retinoic acid causes the SH-F and SH-N cell lines to differentiate along different routes which means that the upgrading responses may be related to more general aspects of differentiation rather than to specific phenotype expression.

Current pathogenetic and molecular concepts in viral liver carcinogenesis

Hepatocellular carcinoma (HCC) is one of the most frequent malignancies in humans and in most cases a consequence of chronic infection of the liver by hepatotropic viruses (Hepatitis B Virus (HBV) and possibly Hepatitis C Virus (HCV)). Formation of HCC results from a stepwise process involving different preneoplastic lesions that reflect multiple genetic events, like protooncogene activation, tumor suppressor gene inactivation, and growth factor over- or reexpression. Recent investigations have gained new insights into how these factors are activated and may interact. In addition, improved knowledge of the molecular biology of HBV has led to better understanding of its pleiotropic effects on induction and progression in hepatocarcinogenesis.

A Drosophila gene that is homologous to a mammalian gene associated with tumor metastasis codes for a nucleoside diphosphate kinase

The product of the abnormal wing discs (awd) gene of Drosophila is 78% identical to the product of the nm23 gene of mammals, which is differentially expressed in certain metastatic tumors. We present evidence that the awd gene codes for a nucleoside diphosphate kinase (NDP kinase) and that this Awd/NDP kinase is microtubule associated. Neuroblasts in Drosophila larvae homozygous for a null mutation in the awd gene are arrested in metaphase, indicating that microtubule-associated Awd/NDP kinase plays a critical role in spindle microtubule polymerization.

Evidence for complex formation between GTP binding protein(Gs) and membrane-associated nucleoside diphosphate kinase

When the Gs in rat liver membranes was prelabeled with [32P]NAD and cholera toxin, solubilized with octylglucoside, and then analyzed by sucrose density gradient centrifugation, it was fractionated into two peaks with approximate molecular sizes of 12-13S and 3-4S. Pretreatment without or with GDP beta S of the labeled membranes resulted in a larger peak in the high molecular weight region, whereas pretreatment with glucagon plus GTP gamma S caused almost equal peaks in both regions. The affinity-purified anti-nucleoside diphosphate (NDP) kinase antibodies only precipitated the Gs in high molecular weight region. Under the same condition, small but significant NDP kinase activity was associated with the high molecular weight Gs region although a large portion of the enzyme activity was recovered in fractions where it alone should appear (6.2S). Both Lubrol-PX and digitonin solubilized the Gs in forms insensitive to immunoprecipitation by anti-NDP kinase antibodies although the latter detergent was able to solubilize the Gs in a high molecular weight form, that is, a ternary glucagon-receptor-G protein complex. These results demonstrate that Gs and membrane-associated NDP kinase may exist in part in a complexed form in membranes. Physiological relevance of the complex formation in membrane signal transduction is discussed.

Prevalence of nucleoside diphosphate kinase autophosphorylation in human colon carcinoma versus normal colon homogenates

The G-regulatory proteins of adenylate cyclase, tubulin, and the ras oncogene protein product require the production of GTP from ATP in order to exert their effects within the cell. This implies that the activity of nucleoside diphosphate kinase (NDPK) plays a major role in the regulation of cellular events requiring GTP and that the level of activity of this enzyme is critical. This report presents a simple method for trapping a specific isozyme of NDPK in its high-energy phosphorylated form (NDPK approximately P) using EDTA and demonstrates that this NDPK approximately P is tenfold higher in malignant colon tumor tissue than in normal colon tissue. This autophosphorylation of the 21,000 and 24,000 Mr subunits of NDPK occurs rapidly at 0 degrees C, will use either [gamma-32P]ATP, [gamma-32P]GTP, or the corresponding 8-azidopurine photoprobes, is intramolecular, displays saturation effects, and is prevented from forming if GTP gamma S is added. Dephosphorylation in the homogenate occurs rapidly upon addition of Mg2+ or any nucleoside-5'-diphosphate. The subunits autophosphorylated in the homogenates are mostly in the soluble phase, and they comigrate with the subunits of pure NDPK from human erythrocytes. Cross-addition of normal and malignant homogenates does not decrease the level of autophosphorylation of NDPK, which indicates that the level of NDPK approximately P may be a quantitative measure of the level of this specific NDPK isozyme form. Assays for NDPK activity show correspondingly elevated levels in the malignant homogenates. Using western blot and photoaffinity labeling techniques, we distinguished the NDPK approximately P subunits from two closely migrating GTP-binding proteins. These were identified as the ras gene protein product and a 20,000 Mr protein, which comigrates identically with ADP-ribosylating factor (ARF). The ARF also comigrates in a tight band that is phosphorylated by [gamma 32P]ATP or [gamma-32P]GTP when Mg2+ is present.

Direct interaction between membrane-associated nucleoside diphosphate kinase and GTP-binding protein(Gs), and its regulation by hormones and guanine nucleotides

In previous studies we have proposed that the membrane-associated nucleoside diphosphate kinase (m-NDP kinase) may play a role as a GTP channeling machinery for adenylate cyclase regulation by hormones. In this study, whether the m-NDP kinase has a direct interaction with the component (GTP-binding protein (Gs)) of the glucagon- and beta-adrenergic agonist-sensitive adenylate cyclase systems in rat liver membranes was examined by extraction with octylglucoside, followed by immunoprecipitation by affinity-purified monospecific anti-NDP kinase antibodies. The results demonstrated that the m-NDP kinase and the Gs were extractable as a complexed form and that the complex formation was reversibly regulated, through cell surface receptors, by hormones which had an ability to cause activation of the rat liver adenylate cyclase. Also, it was suggested that guanine nucleotides rather than hormones were primary regulators of the m-NDP kinase-Gs interaction. These results were discussed in relation to the regulatory cycle of the Gs of adenylate cyclase system.

Physiological correlation between nucleoside-diphosphate kinases and the 21-kDa guanine-nucleotide binding proteins copurified with the enzymes from the cell membrane fractions of Ehrlich ascites tumor cells

The physiological correlation between nucleoside-diphosphate kinases (NDP-kinases) and the 21-kDa guanine nucleotide-binding proteins (G1 and G2) which are copurified with the enzymes from the cell membrane fractions of Ehrlich ascites tumor cells has been biochemically investigated in vitro. We found that: incubation of the phosphoenzyme (enzyme-bound high-energy phosphate intermediate) of NDP-kinases (F-I and F-II) with one of the nucleoside 5'-diphosphates in the presence of 1 mM Mg2+ or 0.25 mM Ca2+ results in the rapid formation of nucleoside 5'-triphosphates without strict base specificity; GDP on the guanine nucleotide-binding proteins (G1, G2 and recombinant v-rasH p21) acts as a phosphate acceptor for the high-energy phosphates of the phosphoenzyme in the presence of 0.25 mM Ca2+; and [32P]GTP is preferentially formed from the 32P-labelled phosphoenzyme F-I and GDP-bound G1 or GDP-bound recombinant v-rasH p21 protein, even if any other nucleoside 5'-diphosphates are present in the reaction mixture. Although [32P]GTP formed was bound with the guanine nucleotide-binding proteins, it was immediately hydrolyzed by the proteins themselves in the presence of 5 mM Mg2+, but not in the presence of 0.25 mM Ca2+. Available evidence suggests that NDP-kinase may be responsible for the activation of the guanine nucleotide-binding proteins (G1, G2 and p21 proteins) through phosphate transfer by the enzyme.

Physiological correlation between nucleoside-diphosphate kinase and the enzyme-associated guanine nucleotide binding proteins

The physiological correlation between NDP-kinase and the enzyme-associated guanine nucleotide binding proteins (G1 and G2) has been studied in vitro. It was found that incubation of the phosphoenzyme (enzyme-bound high-energy phosphate intermediate) of NDP-kinases with one of the nucleoside 5'-diphosphates (NDPs) in the presence of divalent cations (Mg2+ and Ca2+) results in the formation of nucleoside 5'-triphosphates (NTPs) within 40 sec even at low temperatures (below 4 degrees C) without strict base-specificity; and high-energy phosphates on the phosphoenzyme can transfer preferentially to GDP on the guanine nucleotide binding proteins (G1, G2 and r-p21 protein) in the presence of 0.25 mM Ca2+ or 1 mM Mg2+ even if any other NDPs are present in the reaction mixtures. These observations suggest that NDP-kinase may be responsible for the phosphate-transfer between GDP on the guanine nucleotide binding proteins and its phosphoenzyme.

Characterization of nucleoside-diphosphate kinase-associated guanine nucleotide-binding proteins from HeLa S3 cells

Nucleoside-diphosphate (NDP) kinase-associated [alpha-32P]GTP-incorporating proteins from HeLa S3 cells have been biochemically characterized. Two distinct NDP-kinases (F-I and F-II) had been partially purified from HeLa S3 cells by Sephacryl S-300 gel filtration and DEAE-cellulose column chromatography. The [alpha-32P]GTP-incorporating proteins (approx. Mr 20,000) could be separated from NDP-kinases (approx. Mr 80,000) by 5-25% glycerol density-gradient centrifugation analysis after treatment with 7 M urea in the presence of 1 mM EDTA. [alpha-32P]GTP incorporation into these two proteins (G1 and G2) from NDP-kinases required 5 mM Mg2+ and was highly inhibited by either GDP or GTP analogues, such as guanylyl imidodiphosphate and guanylyl methylenediphosphate. [3H]GDP, but no other nucleoside 5'-diphosphates, was also bound to these two proteins in the presence of Mg2+ (5 mM). Moreover, incubation of [alpha-32P]GTP with either G1 or G2 in the presence of Mg2+ (5 mM) resulted in the formation of [32P]GDP and Pi. The data presented here indicated that the guanine nucleotide-binding activity, the GTPase activity, and the molecular weight (approx. Mr 20,000) of NDP-kinase-associated proteins from HeLa S3 cells are similar to those reported for ras oncogene products (p21 proteins).

GTP-activated GTP binding protein(Gs) in membranes achieved by hormone plus GDP does not serve as a substrate for ADP-ribosylation by cholera toxin

Adenylate cyclase in the presence of GTP became active by the addition of cholera toxin irrespective of the presence of glucagon, and under the same condition the Gs of these activated enzymes were good acceptor of an ADP-ribose moiety. On the other hand, the cyclase in the presence of GDP remained inactive with cholera toxin but became active by the further addition of glucagon. However, neither of these Gs served as a cholera toxin substrate. Glucagon reduced an inhibitory action of added GDP for cholera toxin plus GTP-stimulated adenylate cyclase activity but did not for toxin plus GTP-enhanced ADP-ribosylation of Gs. These results demonstrate that Gs-GTP complex formation alone is not sufficient for Gs to serve as a cholera toxin substrate, and suggest an additional GTP binding site responsible for ADP-ribosylation by the toxin. Hormone dependent preferential interaction between the GTP binding site on Gs coupled with adenylate cyclase regulation and membrane-associated nucleoside diphosphate kinase is discussed.

Differential susceptibility to GTP formed from added GDP via membrane-associated nucleoside diphosphate kinase of GTP-sensitive adenylate cyclases achieved by hormone and cholera toxin

GTP-sensitive adenylate cyclases in liver membranes achieved by glucagon and by cholera toxin pretreatment displayed similar responses to added GTP in assay with respect to magnitude and sensitivity. However, their susceptibility to GTP formed during incubation from added GDP catalyzed by membrane-associated nucleoside diphosphate kinase(mNDPK) was different. Adenylate cyclase pretreated with cholera toxin was essentially unaffected by added GDP, while further addition of glucagon produced activation. GTP-stimulated adenylate cyclase activity in toxin-treated membranes was inhibited by added GDP, whereas glucagon addition reduced the inhibitory action of GDP by two orders of magnitude. Since neither pretreatment with toxin nor glucagon addition altered GTP formation by mNDPK, these observations suggest a possible presence of a mechanism by which hormone makes adenylate cyclase susceptible to the GTP formed via mNDPK for activation.

Adenosine 5'-(beta, gamma-imino)triphosphate and guanosine 5'-O-(2-thiodiphosphate) do not necessarily provide non-phosphorylating conditions in adenylate cyclase studies

Commercial preparations of adenosine 5'-(beta, gamma-imino)triphosphate (App(NH)p) were found to be contaminated with a GTP-like substance(s) as well as a phosphate donor(s) for GDP. Thus, when these preparations were used as substrate with no purification, GDP was as effective as GTP in promoting PGE1 stimulation of human platelet adenylate cyclase. With purified App(NH)p as substrate, the effect of PGE1 with GDP was reduced but still observable, while that with GTP was unaltered. PGE1 also caused a stimulation in the presence of guanosine 5'-o-(2-thiodiphosphate)(GDP beta S) with ATP as substrate. Both of the PGE1-stimulated activities observed with GDP and its analog were completely lost by the addition of UDP, thereby, inhibiting GTP formation catalyzed by membrane-associated nucleoside diphosphate kinase. The results demonstrate that the stimulatory effects of PGE1 observed with GDP and App(NH)p, and with GDP beta S and ATP were transphosphorylation dependent and, therefore, the analogs must be used with special caution in adenylate cyclase studies.