Crystal structure of a human mitochondrial deoxyribonucleotidase

5' nucleotidases are ubiquitous enzymes that dephosphorylate nucleoside monophosphates and participate in the regulation of nucleotide pools. The mitochondrial 5'-(3') deoxyribonucleotidase (dNT-2) specifically dephosphorylates dUMP and dTMP, thereby protecting mitochondrial DNA replication from excess dTTP. We have solved the structure of dNT-2, the first of a mammalian 5' nucleotidase. The structure reveals a relationship to the HAD family, members of which use an aspartyl nucleophile as their common catalytic strategy, with a phosphoserine phosphatase as the most similar neighbor. A structure-based sequence alignment of dNT-2 with other 5' nucleotidases also suggests a common origin for these enzymes. Here we study the structures of dNT-2 in complex with bound phosphate and beryllium trifluoride plus thymidine as model for a phosphoenzyme-product complex. Based on these structures, determinants for substrate specificity recognition and the catalytic action of dNT-2 are outlined.     

Effects of adenosine 5'-monophosphate and adenosine 3',5'-cyclic monophosphate on l cells.

The adenine nucleotides, 5'-AMP and 3',5'-cyclic AMP block L cells in the S-phase of the cell cycle. The intracellular level of cyclic AMP is reduced after incubation of cells with 5'-AMP, and rates of uridine transport are increased after incubation with either 5'-AMP or cyclic AMP. On the contrary, cyclic AMP levels are increased and uridine transport decreased in cells treated with an inhibitor of the cyclic AMP phosphodiesterase. This inhibitor partially reverses the growth-inhibitory effect of cyclic AMP, indicating that a breakdown product is the effective inhibitor of growth. The inhibition of cell growth induced by the adenine nucleotides is prevented by uridine, suggesting that the block in S is due to a lack of availability of pyrimidines.     

Human high-Km 5'-nucleotidase effects of overexpression of the cloned cDNA in cultured human cells.

5'-Nucleotidases participate, together with nucleoside kinases, in substrate cycles involved in the regulation of deoxyribonucleotide metabolism. Three major classes of nucleotidases are known, one on the plasma membrane and two in the cytosol. The two cytosolic classes have been named high-Km nucleotidases and 5'(3')-nucleotidases. Starting from two plasmids with partial sequences (Oka, J., Matsumoto, A., Hosokawa, Y. & Inoue, S. (1994) Biochem. Biophys. Res. Commun. 205, 917-922) we cloned the complete cDNA of the human high-Km nucleotidase into vectors suitable for transfection of Escherichia coli or mammalian cells. After transfection, E. coli overproduced large amounts of the enzyme. Most of the enzyme was present in inclusion bodies that also contained many partially degraded products of the protein. Part of the enzyme, corresponding to approximately 2% of the soluble proteins, was in a soluble active form. Stably transfected human 293 cells were obtained with a vector where the 3'-end of the nucleotidase coding sequence is linked to the 5'-end of the green fluorescent protein coding sequence. Several green clones overproduced both mRNA and fusion protein. Two clones with 10-fold higher enzyme activity were analyzed further. The nucleotidase activity of cell extracts showed the same substrate specificity and allosteric regulation as the high-Km enzyme. The growth rate of the two clones did not differ from the controls. The cells were not resistant to deoxyguanosine or deoxyadenosine, and did not show an increased ability to phosphorylate dideoxyinosine. Both ribonucleoside and deoxyribonucleoside triphosphate pools were decreased slightly, suggesting participation of the enzyme in their regulation.     

Hepatitis C Virus Non-Structural Protein 3 Interacts with Cytosolic 5′(3′)-Deoxyribonucleotidase and Partially Inhibits Its Activity

Infection with hepatitis C virus (HCV) is etiologically involved in liver cirrhosis, hepatocellular carcinoma and B-cell lymphomas. It has been demonstrated previously that HCV non-structural protein 3 (NS3) is involved in cell transformation. In this study, a yeast two-hybrid screening experiment was conducted to identify cellular proteins interacting with HCV NS3 protein. Cytosolic 5′(3′)-deoxyribonucleotidase (cdN, dNT-1) was found to interact with HCV NS3 protein. Binding domains of HCV NS3 and cellular cdN proteins were also determined using the yeast two-hybrid system. Interactions between HCV NS3 and cdN proteins were further demonstrated by co-immunoprecipitation and confocal analysis in cultured cells. The cellular cdN activity was partially repressed by NS3 protein in both the transiently-transfected and the stably-transfected systems. Furthermore, HCV partially repressed the cdN activity while had no effect on its protein expression in the systems of HCV sub-genomic replicons and infectious HCV virions. Deoxyribonucleotidases are present in most mammalian cells and involve in the regulation of intracellular deoxyribonucleotides pools by substrate cycles. Control of DNA precursor concentration is essential for the maintenance of genetic stability. Reduction of cdN activity would result in the imbalance of DNA precursor concentrations. Thus, our results suggested that HCV partially reduced the cdN activity via its NS3 protein and this may in turn cause diseases.     

2',5'-Oligoadenylates and related 2',5'-oligonucleotide analogues. 1. Substrate specificity of the interferon-induced murine 2',5'-oligoadenylate synthetase and enzymatic synthesis of oligomers

The substrate specificity of the interferon-induced mouse L-cell enzyme, 2',5'-oligoadenylate synthetase, was determined with a number of nucleoside 5'-triphosphate analogues. Selected nucleoside 5'-triphosphates were converted to 2',5'-oligonucleotides with the following order of efficiency for the nucleoside: 8-azaadenosine greater than adenosine = 2-chloroadenosine greater than sangivamycin greater than toyocamycin greater than formycin greater than 3-ribosyladenine greater than ribavirin greater than tubercidin greater than adenosine 1-oxide greater than 2-beta-D-ribofuranosylthiazole-4-carboxamide greater than inosine = 1,N6-ethenoadenosine greater than guanosine greater than 8-bromoadenosine = uridine greater than cytidine. Adenosine 5'-((beta, gamma-imidotriphosphate) did not seem to be a recognizable substrate since no detectable product resulted. Either the 2',5'-oligoadenylate synthetase is not as specific as had been previously thought, or there may be more than one 2',5'-oligonucleotide synthetase. The 2',5'-oligonucleotide analogue products in which the adenosine of ppp(A2'P5')nA was replaced by the various nucleoside analogues were separated by DEAE-cellulose column chromatography and the chain length and number of 5'-phosphate residues analyzed by a rapid, efficient high-performance liquid chromatographic (HPLC) system involving ion-pairing C18 reversed-phase column chromatography. Separation of the 5'-mono-, 5'-di-, and 5'-triphosphorylated forms of the 2',5'-oligonucleotide analogue dimers, trimers, tetramers, and pentamers was readily achieved by this useful HPLC system. No 5'-nonphosphorylated forms were detected for any of the 2',5'-oligonucleotide analogue products.     

Productive and nonproductive binding to ribonuclease A: X-ray structure of two complexes with uridylyl(2',5')guanosine

Guanine-containing mono- and dinucleotides bind to the active site of ribonuclease A in a nonproductive mode (retro-binding) (Aguilar CF, Thomas PJ, Mills A, Moss DS, Palmer RA. 1992. J Mol Biol 224:265-267). Guanine binds to the highly specific pyrimidine site by forming hydrogen bonds with Thr45 and with the sulfate anion located in the P1 site. To investigate the influence of the anion present in the P1 site on retro-binding, we determined the structure of two new complexes of RNase A with uridylyl(2',5')guanosine obtained by soaking two different forms of pre-grown RNase A crystals. In one case, RNase A was crystallized without removing the sulfate anion strongly bound to the active site; in the other, the protein was first equilibrated with a basic solution to displace the anion from the P1 site. The X-ray structures of the complexes with and without sulfate in P1 were refined using diffraction data up to 1.8 A (R-factor 0.192) and 2.0 A (R-factor 0.178), respectively. The binding mode of the substrate analogue to the protein differs markedly in the two complexes. When the sulfate is located in P1, we observe retro-binding; whereas when the anion is removed from the active site, the uridine is productively bound at the B1 site. In the productive complex, the electron density is very well defined for the uridine moiety, whereas the downstream guanine is disordered. This finding indicates that the interactions of guanine in the B2 site are rather weak and that this site is essentially adenine preferring. In this crystal form, there are two molecules per asymmetric unit, and due to crystal packing, only the active site of one molecule is accessible to the ligand. Thus, in the same crystal we have a ligand-bound and a ligand-free RNase A molecule. The comparison of these two structures furnishes a detailed and reliable picture of the structural alterations induced by the binding of the substrate. These results provide structural information to support the hypotheses on the role of RNase A active site residues that have recently emerged from site-directed mutagenesis studies.     

Coupled 5' nucleotide recognition and processivity in Xrn1-mediated mRNA decay

Messenger RNA decay plays a central role in the regulation and surveillance of eukaryotic gene expression. The conserved multidomain exoribonuclease Xrn1 targets cytoplasmic RNA substrates marked by a 5' monophosphate for processive 5'-to-3' degradation by an unknown mechanism. Here, we report the crystal structure of an Xrn1-substrate complex. The single-stranded substrate is held in place by stacking of the 5'-terminal trinucleotide between aromatic side chains while a highly basic pocket specifically recognizes the 5' phosphate. Mutations of residues involved in binding the 5'-terminal nucleotide impair Xrn1 processivity. The substrate recognition mechanism allows Xrn1 to couple processive hydrolysis to duplex melting in RNA substrates with sufficiently long single-stranded 5' overhangs. The Xrn1-substrate complex structure thus rationalizes the exclusive specificity of Xrn1 for 5'-monophosphorylated substrates, ensuring fidelity of mRNA turnover, and posits a model for translocation-coupled unwinding of structured RNA substrates.     

Wheat (Triticum vulgare) chloroplast nuclease ChSI exhibits 5' flap structure-specific endonuclease activity

The structure-specific ChSI nuclease from wheat (Triticum vulgare) chloroplast stroma has been previously purified and characterized in our laboratory. It is a single-strand-specific DNA and RNA endonuclease. Although the enzyme has been initially characterized and used as a structural probe, its biological function is still unknown. Localization of the ChSI enzyme inside chloroplasts, possessing their own DNA that is generally highly exposed to UV light and often affected by numerous redox reactions and electron transfer processes, might suggest, however, that this enzyme could be involved in DNA repair. The repair of some types of DNA damage has been shown to proceed through branched DNA intermediates which are substrates for the structure-specific DNA endonucleases. Thus we tested the substrate specificity of ChSI endonuclease toward various branched DNAs containing 5' flap, 5' pseudoflap, 3' pseudoflap, or single-stranded bulged structural motifs. It appears that ChSI has a high 5' flap structure-specific endonucleolytic activity. The catalytic efficiency (k(cat)/K(M)) of the enzyme is significantly higher for the 5' flap substrate than for single-stranded DNA. The ChSI 5' flap activity was inhibited by high concentrations of Mg(2+), Mn(2+), Zn(2+), or Ca(2+). However, low concentrations of divalent cations could restore the loss of ChSI activity as a consequence of EDTA pretreatment. In contrast to other known 5' flap nucleases, the chloroplast enzyme ChSI does not possess any 5'-->3' exonuclease activity on double-stranded DNA. Therefore, we conclude that ChSI is a 5' flap structure-specific endonuclease with nucleolytic activity toward single-stranded substrates.     

Glycosylations of inosine and uridine nucleoside bases and synthesis of the new 1-(beta-D-glucopyranosyl)-inosine-5',6"-diphosphate

Gluco- and ribosylation of the bases of sugar protected inosine and uridine were investigated, obtaining only adducts with beta-configuration at the new glycosidic carbon; stereospecific insertion of a sugar moiety at the 1-N of inosine was achieved either using a Mitsunobu approach (for ribosylation) or by direct coupling of 1-alpha-bromoglucose 13 with 2',3',5'-tri-O-acetylinosine for glucosylation. 1-(beta-D-glucosyl)-inosine, chosen as starting substrate for glucosylated analogs of cyclic IDP-ribose, was phosphorylated at the primary hydroxyls and tested in intramolecular pyrophosphate bond formation.     

Regulation of cyclic AMP metabolism and lipolysis in isolated rat fat cells by insulin, N6-(phenylisopropyl) adenosine and 2',5'-dideoxyadenosine.

The increases in cyclic AMP accumulation and lipolysis by rat fat cells incubated in the presence of catecholamines were abolished by N6-(phenylisopropyl) adenosine. The same inhibition of cyclic AMP accumulation was seen in the presence of 2',5'-dideoxyadenosine but lipolysis was unaffected. In contrast, insulin inhibited lipolysis without affecting cyclic AMP accumulation by norepinephrine plus adenosine deaminase. These results suggest that there are either multiple pools of cyclic AMP or that ther exists some other mechanism which is involved in the regulation of lipolysis by hormones.     

Double-stranded RNA causes synthesis of 2',5'-oligo(A) and degradation of messenger RNA in interferon-treated cells

Interferon-treated HeLa cells were incubated with [3H]uridine to label mRNA and were then exposed to the double-stranded RNA poly(inosinic acid).poly(cytidylic acid) (In.Cn). The incubation with In.Cn greatly enhanced the decay of mRNA. When the cells were incubated in this way in the presence of cycloheximide, which blocks ribosome movement along mRNA, extensive polysome degradation was detected in interferon-treated cells. Products of degradation of mRNA were recovered from monosomes which were presumably formed as a result of endonucleolytic breaks of mRNA. This endonucleolytic activity was correlated with the formation of 2',5'-oligo(A) by an enzyme induced by interferon and activated by double-stranded RNA; the 2',5'-oligo(A) was previously shown to activate an endonuclease in cell extracts. The 2',5'-oligo(A) levels in cells were measured by a competition-binding assay. Details of the procedure used are described, including synthesis of highly radioactive (2'-5')pppA3[32P]cytidine 3',5'-diphosphate, separation of 2',5'-oligo(A) binding from degrading activities, and specificity of the assay.     

Adenosine(5')tetraphospho(5')adenosine-binding protein of calf thymus

An adenosine(5')tetraphospho(5')adenosine (Ap4A) binding protein has been purified from calf thymus. The protein is comprised of a single polypeptide of Mr 54000 and is capable of high-affinity (Kd = 13 microM) binding of Ap4A with great substrate specificity. The Ap4A binding protein has been isolated in two forms: a 'free', or non-polymerase-bound, form which predominates, and a similar form which copurifies with DNA polymerase alpha, but which can be resolved from it. The free form of Ap4A binding protein contains associated adenosine(5')tetraphospho(5')adenosine phosphohydrolase (Ap4Aase) activity, while the form resolved from DNA polymerase alpha contains no such activity. The Ap4Aase activity, which catalyzes the phosphohydrolysis of Ap4A to ATP and AMP, is strongly inhibited by low levels (50-100 microM) of Zn2+ without any effect on the Ap4A binding protein activity. This difference in associated Ap4Aase activity between free and polymerase-bound forms of the protein, plus the copurification mentioned above, indicate a specific association between Ap4A binding protein and DNA polymerase alpha.     

Anabolism versus catabolism of [5-3H]uridine and its relationship to ribonucleic acid labelling in mouse liver after partial hepatectomy

The balance between anabolism and catabolism of [5-(3)H]uridine was studied in the mouse after partial hepatectomy. Labelling of RNA and UDP-glucose was determined and evaluated in relation to changes in the specific radioactivity of UTP. The amounts of labelled catabolic products of uridine were increased several-fold in liver and blood after partial hepatectomy. The specific radioactivity of RNA decreased to about 60% of the control value at 6h and was in the same range as that of control liver at 24h after operation. Decreased labelling of RNA and UDP-glucose was attributable to decreased specific radioactivity of UTP. No changes in the size of the UTP pool or in the balance between uridine anabolism and catabolism were found that could explain the decreased specific radioactivity of UTP. Rather, the alterations in the labelling of this metabolite induced by the partial hepatectomy may be related to decreased phosphorylating capacity in the liver cells and/or dilution of the labelled precursor in an expanded uridine pool. The enhanced amounts of uridine catabolic products in liver and blood were probably a consequence of accumulation and altered incorporation of the metabolites from the blood into the liver cells. Despite the increased amounts of labelled catabolic products and the decreased labelling of RNA, the results reported here actually suggest decreased uridine catabolism and slightly increased RNA synthesis in mouse liver after partial hepatectomy. The results stress the importance of proper controls in determination of nucleic acid synthesis and in metabolic studies by use of labelled precursors.     

Induction of phosphodiesterase by cyclic adenosine 3':5'-monophosphate in differentiating Dictyostelium discoideum amoebae.

Cyclic adenosine 3':5'-monophosphate added to the starvation media of Dictyostelium discoideum amoebae induces both intracellular and extracellular phosphodiesterase activities of these cells. The induced enzyme activity appears several hours earlier than that in starved cells which have not been induced with cyclic nucleotide. In both cases, the appearance of enzyme is inhibited by cycloheximide, and actinomycin D, and daunomycin. The KmS for the extracellular enzyme(s) of nucleotide-induced and uninduced control cells are identical. The induction of enzyme activity seems specific for cyclic adenosine 3':5'-monophosphate since cyclic guanosine 3':5'-monophosphate, as well as other nucleotides, have no effect. No differences in the activity or excretion of either N-acetylglucosaminidase or the inhibitory of the extracellular phosphodiesterase are observed between cyclic adenosine 3':5'-monophosphate-induced and control cells. A direct activation of phosphodiesterase by cyclic adenosine 3':5'-monophosphate can be excluded, since the addition of this nucleotide to cell lysates has no effect on the enzyme activity.     

Dynamics of the dATP pool in cultured mammalian cells.

Conditions for labeling the dATP pool of V79 and 3T3 cells from [3H]deoxyadenosine (salvage) or [3H]adenine (via ribonucleotide reduction) were established. With deoxyadenosine the specific radioactivity of dATP reached a constant value after 60 min. In resting 3T3 cells this value was 30 times higher than in S-phase cells. Turnover of dATP and absolute rates of DNA synthesis and excretion of breakdown products of dATP were determined from the accumulation of isotope in various compartments and the specific activity of dATP. In S-phase cells the dATP pool had a half-life of 4 min, identical to that of dTTP determined earlier. Deoxyadenosine was the major breakdown product of dATP in the presence of an inhibitor of adenosine deaminase. The rate of deoxyadenosine excretion of V79 cells amounted to 4% of the rate of dATP incorporation into DNA. Inhibition of DNA replication increased deoxyadenosine excretion 5- to 10-fold, demonstrating a continued de novo synthesis of dATP, albeit at a slightly reduced rate. Our results fit a model involving a substrate cycle between dAMP and deoxyadenosine regulating the dATP pool, similar to the model of substrate cycles involved in the regulation of pyrimidine deoxyribonucleotide pools developed earlier.     

2'5'-Oligoadenylate Polymerase Activity in the Rat Small Intestine

2'5'-Oligoadenylates are potent protein synthesis inhibitors: they are synthesized by a polymerase which was first described in interferon-treated cells. This system may also be involved in a normal process of cell proliferation and differentiation, in the absence of any viral infection. The small intestine enterocyte has been investigated as a model to test this hypothesis. The presence of 2'5'-oligoadenylate polymerase activity is demonstrated in the intestinal mucosa of normal adult rats. The distribution of this enzyme, and of the enzymes degrading 2'5' oligoadenylates have been investigated on enterocyte pools selectively extracted, under mild conditions, from the different parts of the rat small intestine. Similar results were obtained with enterocytes extracted from the mucosa of germ-free animals.     

Activation of ribonuclease L by (2'-5')(A)4-poly(L-lysine) conjugates in intact cells

Molecular hybrids were synthesized by coupling (2'-5')(A)n oligoadenylates or 2-5A, an intracellular mediator involved in antiviral activity of interferons (IFNs), with poly(L-lysine) used as a membrane carrier. (2'-5')(A)n in its free form was not taken up by cells, probably because of its ionic character. Conjugation with the polypeptide carrier overcame this problem and enabled its pharmacological properties to be developed. The alpha-glycol group of individual (2'-5')(A)n oligomers was oxidized by periodate oxidation and conjugated by an amino reductive reaction to poly(L-lysine), Mr 14 000, in a molar ratio of 5:1. These hybrid molecules left the biologically active 5' end moiety of the (2'-5')(A)n molecule unchanged, and in particular its triphosphate group, and stabilized the molecule by increasing its resistance to phosphodiesterase hydrolysis. A dose-dependent inhibition of virus growth was observed on concomitant incubation of (2'-5')(A)n-poly(L-lysine) conjugates with vesicular stomatitis virus infected L1210 cell cultures. This was a result of the activation of the (2'-5')(A)n-dependent endoribonuclease (RNase L) by intracellularly delivered (2'-5')(A)n as in some IFN-treated virus-infected cells. Indeed, (2'-5')(A)n-poly(L-lysine) conjugates bind RNase L effectively as can be seen from their ability to compete with authentic (2'-5')(A)n in a cell-free radiobinding assay. Moreover, (2'-5')(A)n-poly(L-lysine) conjugates promote transient inhibition of protein synthesis and a characteristic cleavage pattern of ribosomal RNAs in intact cells.(ABSTRACT TRUNCATED AT 250 WORDS)     

5' stem-loop of collagen alpha 1(I) mRNA inhibits translation in vitro but is required for triple helical collagen synthesis in vivo

The 5' stem-loop is a conserved sequence element found around the translation initiation site of three collagen mRNAs, alpha1(I), alpha2(I), and alpha1(III). We show here that the 5' stem-loop of collagen alpha1(I) mRNA is inhibitory to translation in vitro. The sequence 5' to the translation initiation codon, as a part of the 5' stem-loop, is also not efficient in initiating translation under competitive conditions. This suggests that collagen alpha1(I) mRNA may not be a good substrate for translation. Since the 5' stem-loop binds protein factors in collagen-producing cells, this binding may regulate its translation in vivo. We studied in vivo translation of collagen alpha1(I) mRNA after transfecting collagen alpha1(I) genes with and without the 5' stem-loop into Mov 13 fibroblasts. The mRNA with the alpha1(I) 5' stem-loop was translated into pepsin-resistant collagen, which was secreted into the cellular medium. This mRNA also produced more disulfide-bonded high molecular weight collagen found intracellularly. The mRNA in which the 5' stem-loop was mutated, but without affecting the coding region of the gene, was translated into pepsin-sensitive collagen and produced only trace amounts of disulfide-bonded collagen. This suggests that the 5' stem-loop is required for proper folding or stabilization of the collagen triple helix. To our knowledge this is the first example that an RNA element located in the 5'-untranslated region is involved in synthesis of a secreted multisubunit protein. We suggest that 5' stem-loop, with its cognate binding proteins, targets collagen mRNAs for coordinate translation and couples translation apparatus to the rest of the collagen biosynthetic pathway.     

A calcium channel blocker activates both ecto-5(')-nucleotidase and NO synthase in HUVEC

Since amlodipine, a long-acting Ca channel blocker, increases both NO and adenosine production in canine hearts, we investigated that amlodipine activates both ecto-5(')-nucleotidase responsible for adenosine production and NO synthase (NOS) for NO production in human umbilical venous endothelial cells (HUVECs), and its cellular signaling. We measured activities of ecto-5(')-nucleotidase and NOS in HUVECs in the condition with additions of xanthine (100 microM)+xanthine oxidase (1.6 x 10(-3)U/ml) in the presence or absence of amlodipine (1 x 10(-9)-1 x 10(-6)M). Amlodipine increased both ecto-5(')-nucleotidase and NOS activities. Xanthine+xanthine oxidase deactivated both NOS and ecto-5(')-nucleotidase, and amlodipine increased both activities of NOS and ecto-5(')-nucleotidase by 117+/-33% and 48+/-6%, respectively. Amlodipine phosphorylated p38MAP kinase and that an inhibitor of p38MAP kinase inhibited the amlodipine-induced activation of both NOS and ecto-5(')-nucleotidase. Furthermore, amlodipine increased both adenosine and NO production in the canine ischemic hearts. We concluded that amlodipine activates both NOS and ecto-5(')-nucleotidase via p38MAP kinase in vitro and enhances both NO and adenosine production in vivo.