HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rampazzo, C. Articles by Bianchi, V. Search for Related Content PubMed PubMed Citation Articles by Rampazzo, C. Articles by Bianchi, V. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 8, 5409-5415, February 25, 2000

Mammalian 5'(3')-Deoxyribonucleotidase, cDNA Cloning, and Overexpression of the Enzyme in Escherichia coli and Mammalian Cells^*

Chiara Rampazzo^§, Magnus Johansson^¶, Lisa Gallinaro, Paola Ferraro, Ulf Hellman, Anna Karlsson^¶, Peter Reichard^§, and Vera Bianchi^**

From the  Department of Biology, University of Padova, I-35131 Padova, Italy, the ^§ Department of Biochemistry I, Medical Nobel Institute, Karolinska Institutet, SE-17177 Stockholm, Sweden, the ^¶ Division of Clinical Virology F68, Huddinge University Hospital, SE-14186 Huddinge, Sweden, and the  Ludwig Institute for Cancer Research, Box 595, Uppsala University, SE-75142 Uppsala, Sweden

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

5'(3')-Deoxyribonucleotidase is a ubiquitous enzyme in mammalian cells whose physiological function is not known. It was earlier purified to homogeneity from human placenta. We determined the amino acid sequences of several internal peptides and with their aid found an expressed sequence tag clone with the complete cDNA for a murine enzyme of 23.9 kDa. The DNA was cloned into appropriate plasmids and introduced into Escherichia coli and ecdyson-inducible 293 and V79 cells. The recombinant enzyme was purified to homogeneity from transformed E. coli and was found to be identical with the native enzyme. After induction with ponasterone, the transfected mammalian cells showed a gradual increase of enzyme activity. A human expressed sequence tag clone contained a large part of the cDNA of the human enzyme but lacked the 5'-end corresponding to 51 amino acids of the murine enzyme. Several polymerase chain reaction-based approaches to find this sequence met with no success. A mouse/human hybrid cDNA that had substituted the missing human 5'-end with the corresponding mouse sequence coded for a fully active enzyme.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In 1971 two independent reports described the occurrence of a cytosolic 5'-nucleotidase with a preference for deoxyribonucleotides in rat liver (1) and in mouse fibroblasts (2). In contrast to many other enzymes involved in the synthesis of DNA precursors or DNA (3), the enzyme did not show any major variations during the cell cycle (2). Partially purified preparations of the enzyme were studied by Fritzson (summarized in Ref. 4). Besides being a 5'-nucleotidase, the enzyme also dephosphorylated the 2'(3')-phosphates of both ribo- and deoxyribonucleotides and was stimulated "allosterically" by purine deoxyribonucleosides.

In 1990 Höglund and Reichard (5) obtained the human enzyme in homogenous form after a 15,000-fold purification from placenta. The protein had an apparent molecular weight of 45,000 and a subunit mass of 22 kDa. Its catalytic properties were similar to those described earlier for partially purified preparations of the rat enzyme, showing 15-30-fold higher relative V_max/K_m ratios for 5'-deoxyribonucleotides than 5'-ribonucleotides with the following preferences for various bases: Hyp > Ura > Gua > Thy > Ade > Cyt. The pure enzyme also hydrolyzed 2'(3')-nucleotides, establishing that a single protein is responsible for the earlier described activity against both 5'- and 2'(3')-nucleotides. In the latter case there was no preference for deoxyribose over ribose, and the V_max/K_m ratios were actually slightly higher than for 5'-deoxyribonucleotides. Unlike other 5'-nucleotidases the enzyme was neither allosterically activated nor inhibited by ATP. The enzyme has been given several different names¹; here it is called 5'(3')-deoxyribonucleotidase, abbreviated dNT.² dNTs are present in most mammalian cells. Lymphoid cells were reported to show a particularly high activity (6).

Pure dNT clearly differs in catalytic properties and protein nature from two other ubiquitous 5'-nucleotidases that have attained more attention over the years. One of them is a membrane-bound ectonucleotidase (7, 8), which in minor amounts has also been found in the cytoplasm (9). The other enzyme is a purely cytosolic nucleotidase, named high K_m-nucleotidase (10, 11), with a preference for hydroxypurine nucleotides. The cDNAs of both these enzyme have been cloned (12, 13) and were also used for the construction of bacteria and mammalian cells that overproduce the enzymes (8, 14, 15).

In connection with studies of patients with a genetically caused chronic hemolytic anemia, Paglia et al. (16) described the presence in erythrocytes of two "isozymes" of cytoplasmic 5'-nucleotidase. A deficiency in one form was disease-related. The two forms were subsequently separated chromatographically and named (17) P5N-I and P5N-II. P5N-I was the anemia-related enzyme, and P5N-II was suggested to be a dNT.

Our interest in the dNT stems from earlier work concerning the regulation of intracellular dNTP pools by substrate cycles (3, 18-20). We have found that such cycles, relying on the interplay between a deoxynucleoside kinase and a nucleotidase, participate in the regulation of dNTP pools. The nature of the nucleotidase(s) participating in such cycles is unknown, whereas kinases were identified from isotope flow experiments with genetically manipulated cultured cells (19-21). We intend to carry out similar experiments with nucleotidases. To this purpose we now describe the cloning of the cDNA for the murine dNT as well as initial experiments resulting in the transformation of Escherichia coli and cultured cells with the cloned cDNA.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- [³H]dUMP for routine assays of dNT and [³H]thymidine for phosphotransferase assays were from Amersham Pharmacia Biotech. The latter contained a small amount of a radioactive impurity that interfered with the assay and was removed by reversed phase HPLC. Nonlabeled nucleotides were from Sigma. Zeocin was from Invitrogen, and G418 was from Calbiochem.

Bacterial Strains and Mammalian Cell Lines-- E. coli DH5 was used for routine transformations and plasmid preparations according to standard procedures (22). Proteins were overexpressed in E. coli BL21(DE3)plysS from sequences subcloned in pET20b (Novagen). The bacteria were grown in Luria-Bertani liquid medium with the appropriate antibiotics and plated on Luria-Bertani 1% agarose. Plasmids used in this work are described in Table I. All constructs were verified by sequence determination.

293 human embryonal kidney cells and V79 hamster fibroblasts were cultured in Dulbecco's modified Eagle's medium with 7.5 and 5% heat-inactivated fetal calf serum, respectively, and antibiotics at 37 °C in a humidified incubator at 5% CO₂. Absence of mycoplasma contamination was periodically ascertained by the Gen-Probe Mycoplasma T. C. Rapid Detection System (GEN-PROBE Inc., San Diego, CA).

Sequences of Internal Peptides from Human dNT-- The protein was purified from human placenta as described earlier (5). Briefly, crude extracts were precipitated between 35 and 55% ammonium sulfate saturation, followed by chromatography on DE52 (Whatman), on Phenyl-Sepharose (Amersham Pharmacia Biotech), and finally on MonoQ HR5/5 (Amersham Pharmacia Biotech). After each step the fractions with high dNT activity were combined and concentrated in Centricon-10 (Amicon) tubes before the next step. The final material gave a major Coomassie-stained band at 27 kDa together with several minor bands after electrophoresis on a denaturing SDS gel. The major band was excised and treated in situ (23) with modified trypsin (Promega) or endoproteinase LysC (Wako). Peptides were separated by HPLC (23) and sequenced in an Applied Biosystems (Foster City, CA) model 470 sequencer.

Identification of dNT cDNA Sequences-- We made a BLAST search of the GenBank^TM data base to identify human and murine sequences homologous to the dNT peptides. Several human and murine EST clones were homologous to all the seven peptides. The clones deposited by the I.M.A.G.E. consortium (24) were obtained from Research Genetics and double-strand resequenced in an Automatic Laser Fluorescent sequencer (Amersham Pharmacia Biotech). The Gene Jokey program was used for sequence alignment and analysis.

Expression of Recombinant dNT in E. coli-- The complete murine cDNA and a 5'-truncated cDNA starting at the second ATG at nucleotides 37-39 were subcloned in pET20b giving plasmids p1M-dNT and p2M-dNT. The cDNA was amplified from EST clone AA261077 by PCR with forward primer 1M (5'-CATACATATGGCGGTGAAGCGGCC-3') located to the first ATG of the cDNA and reverse primer 4M (5'-CTGAGGATCCTGGGAGGAGCCGGAAAG-3') located to nucleotides 677-697 in the 3'-untranslated region. The primers introduced a 5' NdeI and a 3' BamHI restriction site, respectively. The 5'-truncated cDNA was amplified by a similar procedure using forward primer 2M (5'-CATACATATGGACGGCGTGCTAGCTGA-3') located to the second ATG. The two amplified products were subcloned into pGEM-T (Promega), cut with NdeI and BamHI, and ligated into the bacterial expression vector pET20b. A hybrid murine/human cDNA sequence was prepared and cloned as a NdeI-BamHI fragment in pET20b (plasmid pHyb-mh). The 5'-end of the hybrid sequence was obtained from mouse EST clone AA261077 amplified with primer 1M and reverse primer CR17 (5'-CCAGGTCGGGCCGCAGGGCGCGGTA-3'). The human sequence was obtained from EST clone AA122046 whose 5'-end corresponds to nucleotide 154 of the mouse sequence. This clone contains a noncoding 80-bp insertion after 189 bp. A first PCR was run with forward primer CR16 (5'-TACCGCGCCCTGCGGCCCGACCTGG-3') and reverse primer CR18 (5'-CGGTGGATCCTGTGGCCTGCCCTTCC-3'), and the product was spliced to the mouse 5'-end with primers 1M and CR18. This first hybrid product was used as the template for a second series of PCRs that had the aim of deleting the 80-bp insertion. In this case the first two PCRs were run with primers 1M and CR19 (5'-CCAGCGGTACTTCTCACCCACACAGTG-3') and primers CR20 (5'-AGAAGTACCGCTGGGTGGAGC-3') and CR18, respectively, and the two products were spliced together with primers 1M and CR18. For expression of the cloned dNT, the recombinant plasmids were transformed into E. coli BL21(DE3) plysS. After growth of the culture at 37 °C, overexpression of the proteins was induced for 3 h with 0.4 mM IPTG.

Expression of Recombinant Murine dNT in Mammalian Cells-- To construct the ecdyson-inducible mammalian expression vector pdNT-m1, the murine dNT coding sequence was amplified by PCR from EST clone AA261077 with forward primer 6M (5'-CATCAAGCTTTGCGGAGACACCG-3') located at nucleotides 22 to 10 and reverse primer 4M. The amplification product was ligated as a BamHI-HindIII fragment into plasmid pIND (Invitrogen). The recombinant plasmid was transfected by calcium phosphate precipitation into cells of clone 293-2-100 (23) and clone V79-20-11. These clones constitutively produce an intracellular ecdyson receptor that can be activated by the synthetic ecdyson analog ponasterone A (Invitrogen). Clone V79-20-11 was obtained from V79 cells transfected with plasmid pVgRXR (Invitrogen) and selected in the presence of 350 µg/ml zeocin. Inducible clones of 293 and V79 cells overexpressing the murine dNT were isolated by selection with 600 and 800 µg/ml G418, respectively, and tested with 4 µM ponasterone A.

To test the time dependence of dNT induction, parallel cultures of an inducible clone from each cell line (293-dNT14 and V79-dNT15) were plated at 0.15 and 0.07 × 10⁶ cells/5-cm dish, respectively. After 6 h, 4 µM ponasterone A was added to the medium. At 48 h half of the cultures were shifted to inducer-free medium and grown for 2 more days. 48 h after the start of induction dNT activity was measured in the crude cellular extracts (14) at 24-h intervals.

Plasmid pdNT-GFP was obtained by subcloning the dNT coding sequence amplified by PCR with forward primer 6M and reverse primer 11M (5'-CGTAGGATCCCACAGGCTGGCTCGC-3') as a BamHI-HindIII fragment into plasmid pIND-GFP (15). This inducible construct codes for a fusion polypeptide of 52 kDa where the dNT is fused via a 7-amino acid linker to the N terminus of the GFP. This plasmid was used in transient transfection experiments to examine the subcellular localization of dNT.

Northern Blotting-- A human multiple-tissue blot (CLONTECH) containing mRNA from heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas was probed according to standard procedures with a [³²P]dCTP-labeled DNA obtained from EST clone AA404645.

Purification of Recombinant Murine dNT-- After induction with IPTG, a 2-liter culture was centrifuged, and the bacterial pellet (5.5 g) was suspended in 30 ml of 50 mM Tris-HCl, pH 7.5, containing 20% glycerol, 2 mM dithiothreitol, and 2 mM EDTA. Cells were lysed by repeated freezing and thawing, 3 ml of 10% streptomycin sulfate and 0.5 ml of 50 mM phenylmethyl sulfonyl fluoride were added, and the suspension was centrifuged at 100,000 × g for 30 min. The clear supernatant solution was purified by the first three steps (ammonium sulfate precipitation and chromatography on DE52 and Phenyl-Sepharose) of the described procedure, resulting in a 40-fold purification of the enzyme. This preparation was used to determine the substrate specificity and other catalytic properties of the recombinant enzyme.

Nucleotidase Assays-- In the routine radiochemical method (5) [³H]dUMP (500 cpm/nmol) was incubated at 37 °C for 30 min in a final volume of 0.02 ml with 20 mM MgCl₂, 5 mM dithiothreitol, 50 mM Tris-maleate, pH 6.0, and 0.2 mg/ml bovine serum albumin. The reaction was terminated by addition of l ml of ice-cold 50 mM acetic acid, and the mixture was passed through a 2-ml column of AG1-X2 to retain unreacted nucleotides. The column was washed with 2 ml of 50 mM acetic acid. The flow-through and wash fractions were combined, and their total radioactivity was determined and used to calculate the amount of deoxyuridine formed by dephosphorylation of dUMP.

Dephosphorylation of nucleotides other than dUMP was measured with a sensitive colorimetric assay from the formation of inorganic phosphate. In the experiments determining substrate specificity 5 mM of the various nucleotides were incubated in 0.05 ml for 30 min as described for the radiochemical assay. The reaction was terminated by addition of volume of 1.2 M H₂SO₄, and after centrifugation, inorganic phosphate was determined (25). This method was also used for the determination of kinetic parameters. In this case the time of incubation ranged from 5 to 15 min to assure linearity of the reaction. In all experiments less than 20% of the nucleotide was dephosphorylated. The different constants were obtained from regression analyses of Lineweaver-Burk reciprocal plots.

The colorimetric assay was used also to measure stimulation of UMP-3' and dUMP-5' dephosphorylation by 1 and 4 mM concentrations of various nucleosides (dIno, dGuo, dThd, dUrd, Ino, Guo and 2', 3'-dideoxyinosine).

One unit of enzyme activity is defined as 1 µmol of product formed during 1 min. Specific activity is units/mg of protein, determined colorimetrically (26) with crystalline bovine serum albumin as standard.

Phosphotransferase Assay-- We looked for the transfer of the phosphate group from 2 mM dUMP-5' or UMP-3' as donor to highly radioactive 10 mM [³H]thymidine (5000 cpm/nmol) as acceptor. Incubations were as described for the radiochemical assay with 1.5 or 15 milliunits of enzyme in a total volume of 0.1 ml, and samples were removed for assay after 5, 10, and 20 min. The reaction was terminated by immersion of the samples in a boiling water bath and centrifugation. The samples were injected into a Partisil SAX (Whatman) HPLC column and eluted with 20 mM ammonium phosphate, pH 3.7. Samples (1 ml) were collected and analyzed for radioactivity. This method separated all radioactive thymidine cleanly from 5'- and 3'-dTMP. Compared with controls, no radioactivity was found in the nucleotide region.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Identification of Internal Peptide Sequences in Human dNT-- Our strategy was to obtain sequences of random internal peptides of the dNT to be able to identify corresponding DNA sequences in the GenBank^TM data base. The 1700-fold purified preparation of the enzyme was reduced, carboxymethylated, and electrophoresed on a denaturing 12% SDS gel. The following seven internal peptides were identified (23) from a prominent 27-kDa band: ALRPDLADK, VASVYE, TSPLLK, YHHSVGEK, XRWVEQHLG, XVERIILTRPK, and TVVLGDLLIDDK.

Identification of the Murine dNT cDNA-- A search of the GenBank^TM data base revealed several murine cDNAs matching the peptide sequences obtained from the purified human dNT. The clone with accession number AA261077 matched all the seven peptides and appeared to contain the complete coding sequence for the nucleotidase (Fig. 1). Two potential initiation codons separated by 33 nucleotides were present. The deduced amino acid sequence starting at the first ATG codon and ending at a TGA stop codon at nucleotides 601-603 corresponds to a 200-amino acid-long polypeptide with a calculated molecular mass of 23,892 Da and an isoelectric point of 5.0. The cDNA sequence contains 28 bp upstream the first ATG and 167 bp downstream the stop codon, with a polyadenylation signal at positions 765-770 followed by a poly(A) tail. EST clone AA261077 was sequenced completely and used to construct the different expression vectors for the experiments.

View larger version (41K): [in this window] [in a new window]   Fig. 1.   cDNA sequence and deduced amino acid sequence of murine (M.m.) dNT. The deduced partial amino acid sequence of the human dNT (H.s.) is aligned with the murine sequence, starting at amino acid residue 52, with asterisks indicating identity of residues. The sequences of the seven peptides identified after degradation of the human placenta dNT are underlined. The first two ATG codons of the murine sequence are in bold type.

Partial cDNA of the Human Enzyme-- A BLAST search of the GenBank^TM data base revealed a number of human sequences coding for some or all the identified dNT peptides, but in this case the situation was more complex than with the mouse cDNA. All the human cDNAs lacked the 181 bp present at the 5'-end of the mouse cDNA, and some contained intervening stretches of nucleotides that produced frameshifts. The amino acid sequence deduced from the incomplete human cDNA was 83% identical to the deduced mouse sequence (Fig. 1).

To determine the size and expression pattern of the human dNT mRNA, we performed a multiple tissue Northern blot analysis. The blot revealed two major mRNAs of 1.5 and 1 kilobase in all the human tissues tested. Expression was highest in skeletal muscle, heart, and pancreas. On the basis of these results we tried several different PCR-based approaches to identify the missing 5'-end of the human cDNA. Without success we used commercial cDNA libraries from pancreas and brain and mRNAs from skeletal muscle and placenta. Specific amplification products were always truncated at about the same 5' position present in the EST clones available in the data base.

Expression of the Recombinant Murine dNT in E. coli-- Starting from each of the two potential initiation codons present in the murine cDNA (Fig. 1), we prepared two separate constructs in the bacterial expression vector pET20b. Plasmid p1M-dNT started at the first potential initiation codon, and p2M-dNT started at the second one. The two corresponding proteins were overexpressed in E. coli BL21(DE3)plysS. After induction with IPTG, the SDS-lysed bacteria containing p2M-dNT, but not those containing p1M-dNT, gave a strong band at 25 kDa on denaturing gels, in a position expected from the truncated dNT (Fig. 2, lanes 1 and 2). Extracts of bacteria prepared with Tris-maleate buffer did not show this band (data not shown), suggesting that a major part of the truncated enzyme was present in inclusion bodies.

View larger version (111K): [in this window] [in a new window]   Fig. 2.   SDS gel analyses. Lane 1, complete lysate of bacteria transformed with p1M-dNT (from 0.075 ml of culture); lane 2, bacteria transformed with p2M-dNT; lane 3, bacteria transformed with pET 20b (control): lane 4, buffer extract (5 µg of protein) from bacteria as in lane 1; lane 5, purified recombinant dNT (1.5 µg protein); lane 6, molecular mass markers.

Enzyme analyses of buffer extracts from IPTG-induced bacteria harboring p1M-dNT showed a high dNT activity (specific activity, 4-7 units/mg protein), whereas extracts from bacteria harboring p2M-dNT showed no increase over the activity found in noninduced controls (specific activity, <0.01 unit/mg). The pure enzyme was reported to have a specific activity of 170 (5), suggesting that dNT represented approximately 3% of the soluble protein in p1M-dNT extracts. Our results also show that the first ATG is the correct initiation codon for the enzyme as the truncated protein, starting at the second initiation codon, lacked enzyme activity. After purification, the recombinant protein had a specific activity of 300-400 units/mg of protein and gave a single band at 27 kDa after electrophoresis on denaturing gels (Fig. 2, lane 5).

Catalytic Properties of Recombinant Murine dNT-- The pure human dNT from placenta (5) and a partially purified enzyme from rat liver (1, 27) both had the following catalytic properties that distinguish them from other nucleotidases: (i) an acid pH optimum between 6.0 and 6.5; (ii) a preference for 5'-deoxyribonucleotides over 5'-ribonucleotides; (iii) the ability to dephosphorylate certain 2'- and 3'-ribonucleotides and deoxyribonucleotides; and (iv) stimulation of this dephosphorylation by some deoxyribonucleosides. We tested the recombinant enzyme for these properties.

Table II shows the substrate specificity of the recombinant dNT tested at a high concentration (5 mM) of 5'-, 3'- and 2'-ribonucleotides and deoxyribonucleotides. For comparison published values for the rat and human enzymes obtained under identical conditions of incubation are also given. With one exception, the relative rates of dephosphorylation found with our enzyme agree well with the published values. The exception is dCMP, which gives a higher activity with the recombinant enzyme. We found similar values with two separate preparations of dCMP. Solutions of dCMP showed the correct UV spectrum, excluding the possibility that contamination with dUMP was responsible for our results.

In a series of experiments we determined the influence of substrate concentration of the various 5'-deoxyribonucleotides and UMP-3'. Table III gives K_m and V_max values obtained from Lineweaver-Burk plots. K_m values were in the mM range. The substrate efficiency (V_max/K_m) for the various deoxyribonucleotides was highest for dUMP and dTMP, followed by dGMP, dAMP, and finally dCMP. Also UMP-3' was an excellent substrate. The data are in good agreement with published K_m values for the rat liver enzyme (1). Most values for the human enzyme are slightly higher (5).

The dephosphorylation of UMP-3' was stimulated 2-fold by deoxyinosine or deoxyguanosine but not by the correspondent ribonucleosides. Also thymidine, but not deoxyuridine, gave a small effect (1.7-fold at 4 mM). In contrast, the dephosphorylation of the 5'-deoxyribonucleotide dUMP was not stimulated by any of the tested nucleosides. These results are in good agreement with the earlier data obtained with the native enzymes (5, 27). Finally the pH activity profile for the recombinant enzyme showed a relatively sharp acid pH optimum between 5.5 and 6.0 for dUMP. With UMP-3', the curve was broader and slightly displaced toward neutrality (data not shown).

Recombinant dNT Has No Phosphotransferase Activity-- Some 5'-nucleotidases catalyze not only the dephosphorylation of nucleotides but also a transfer of the phosphate group from one nucleotide donor to a second nucleoside acceptor. The high K_m nucleotidases belong to this category (10), and also the two 5'-nucleotidases from erythrocytes (P5N-I and P5N-II) were reported to have phosphotransferase activity (28). Pure dNT isolated from placenta was reported to be devoid of phosphotransferase activity (5). We now tested the recombinant enzyme for the ability to transfer the phosphate group from dUMP-5' or UMP-3' to labeled thymidine as described under "Experimental Procedures." In these experiments we also measured simultaneously the dephosphorylation of the presumptive phosphate donors. The sensitivity of the method was such that we would easily have detected an amount of isotope transfer corresponding to 0.5% of the nucleotidase activity. This was not the case, and we conclude that the recombinant dNT lacks phosphotransferase activity.

Expression of Hybrid Murine/Human dNT in Bacteria-- A mouse/human hybrid cDNA starting with the 5'-end of the murine cDNA (nucleotides 1-153) fused to the human cDNA coding for the entire amino acid sequence shown in Fig. 1 (plasmid pHyb-mh) was transformed into E. coli. Production of the corresponding hybrid murine/human dNT was induced with IPTG. Extracts of the bacteria had a dNT specific activity of 3 units/mg of protein, similar to that of extracts from bacteria transformed with the murine cDNA.

Expression of the Murine dNT cDNA in Mammalian Cells-- The ecdyson-inducible mammalian expression system (29) was used to create stably transfected cell lines in which the dNT level could be modulated by a synthetic ecdyson analog such as ponasterone A. Seven human 293 and five hamster V79 clones were isolated whose dNT activity was increased 2-7-fold after 24 h of induction with 4 µM ponasterone A. One clone from each cell line (293-dNT14 and V79-dNT15) was chosen for the following experiment.

To study the time course of induction, cultures of each clone were treated with 4 µM ponasterone for up to 96 h. Cell growth and the level of dNT activity in the crude cellular extracts (Fig. 3, A and B) were measured at 24-h intervals. In both cell lines dNT activity increased with time and eventually reached a plateau. In the human 293 cells enzyme activity increased rapidly and showed a final 10-15-fold increase, whereas in hamster V79 cells the increase was slower and after 96 h was less than 3-fold. Cell growth was unaffected in both cell lines (data not shown). After 48 h induction half of the cultures were shifted to medium without ponasterone and grown for 2 more days. During this period dNT activity declined (Fig. 3, A and B). The higher level of induction in 293-dNT14 cells made possible a closer analysis of the decrease in enzyme activity (Fig. 3C). After removal of ponasterone the total amount of enzyme in the culture remained constant, suggesting that dNT induction was rapidly reversible but that, once formed, the enzyme remained in the culture. The decrease in dNT specific activity in Fig. 3A after removal of the inducer then depended on the appearance of cells in the culture that no longer overproduced the enzyme. During the chase period cell number increased 4-fold, and dNT specific activity decreased from 125 to 25. 

View larger version (8K): [in this window] [in a new window]   Fig. 3.   Induction of dNT activity in mammalian cells. A, 293-dNT14 cells. Cell cultures were induced with 4 µM ponasterone A for 96 h with determinations of dNT activity () as described under "Experimental Procedures." After 48 h, ponasterone was removed from half of the cultures (). Specific activity, units/mg protein. B, as in A but with V79-dNT15 cells. C, after 48 h, total induced dNT activity was calculated in 293-dNT14 cells from the experiment in A by multiplying the values for the dNT specific activity above background with the number of cells/culture divided by 106. Symbols are as in A.

To investigate whether the dNT protein contains any subcellular localization signal, we used a cell line (293-2-100) that expresses the ecdyson receptor and transfected the cells with a plasmid (pdNT-GFP) in which the dNT coding sequence is fused in frame 5' to the sequence for GFP. The transfected cells fluoresced both in the cytoplasm and the nucleus, similar to control cells transfected with a plasmid (pIND-GFP) that codes for the GFP only. This suggests that the dNT polypeptide is not targeted to any specific organelle.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

From the sequences of seven internal peptides of human dNT we could identify a mouse EST clone containing the complete coding sequence of the mouse dNT. After identification of the correct starting ATG codon, we transferred the DNA to appropriate plasmids and used those to transform E. coli and to transfect mammalian cells. In both cases the cells overproduced dNT activity after induction. The recombinant enzyme obtained by purification from E. coli had properties similar to those published for dNT purified from human (5) or rat cells (4). Also the molecular mass of 23.9 kDa calculated from the deduced amino acid composition is in good agreement with the polypeptide mass of 22 kDa reported for the pure placenta enzyme. Altogether, the evidence is compelling that the recombinant murine enzyme is indeed identical to native dNT.

We were not able to clone the complete cDNA for the human dNT by the same approach. All available human EST clones had no 5'-ATG starting codon and apparently lacked approximately 180 bp at the 5'-end present in the murine sequence. Various PCR-based approaches to obtain the 5'-end were unsuccessful, suggesting that the native mRNA contains some particularly unfavorable secondary structure that interfered with first strand in vitro synthesis even by reverse transcriptases working at high temperatures. A hybrid murine/human dNT in which the missing amino acids of the human enzyme were substituted by the corresponding 5'-end of the murine enzyme showed good activity in the routine nucleotidase assay. Together with the high degree of identity between the two polypeptides (Fig. 1), this strongly suggests that the known part of the human cDNA codes for the correct human dNT sequence.

Our results strengthen the conclusion (17) that P5N-II, one of the two forms (P5N-I and P5N-II) of cytoplasmic nucleotidases in human erythrocytes, is a dNT. Both forms were recently purified to apparent homogeneity (28). Surprisingly the relatedness between P5N-II and the human placenta enzyme (5) was not mentioned. It was claimed that both forms had phosphotransferase activity, whereas our recombinant enzyme, as well as the human placenta dNT, was not a phosphotransferase. Further work is required to clarify this discrepancy.

The lack of phosphotransferase activity suggests that the stimulation of nucleotidase activity by some deoxynucleosides observed with the recombinant and native dNT does not depend on transfer of phosphate from UMP to the activator nucleoside. The underlying mechanism may be uncovered once the tridimensional structure of the dNT is solved.

The availability of cloned mammalian cell lines that can be induced to give a graded increase of the intracellular concentration of dNT adds an important tool to future experiments concerning the nature of the nucleotidase participating in substrate cycles. Earlier work has provided such cell lines for the high K_m nucleotidase (14, 15) as well as for the ectonucleotidase (8, 30). There is now a complete set of cell lines overproducing members of each of the three major groups of 5'-nucleotidases, and it should be possible to carry out the required physiological isotope flow experiments to establish the relative importance of the three groups for the regulation of DNA precursors.

In addition to their function in deoxyribonucleotide metabolism, dNTs are also of interest in connection with the use of nucleoside analogs in the therapy of viral and neoplastic diseases. These drugs must be phosphorylated to nucleotides before they exert their clinical effect via inhibition of nucleic acid synthesis. dNT may counteract the attainment of an efficient concentration of the nucleotide, and the affinity of the phosphorylated analog for dNT may be one parameter to be considered in the choice of suitable analogs. An example of such a situation was recently found³ for two nucleoside analogs used to treat HIV infection.

	FOOTNOTES

^* This work was supported in part by two short-term fellowships from the European Molecular Biology Organization (to C. R.) and by funds from the Istituto Superiore di Sanitá (AIDS Project), Associazione Italiana per la Ricerca sul Cancro, and the Italian Ministry of Research (60% Projects) (to V. B.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF078840 and AF154829.

^** To whom correspondence should be addressed. Tel.: 39-049-8276282; Fax: 39-049-8276280; E-mail: vbianchi@civ.bio.unipd.it.

¹ Fritzson has employed the following names in different publications: 5'(3')-nucleotidase, deoxyinosine-activated nucleotidase, deoxyriboside-activated nucleotidase, and deoxynucleotidase. The name 5'(3')-deoxyribonucleotidase used here stresses the preference of the enzyme for deoxyribonucleotides as well as the ability to dephosphorylate both 5'- and 3'-nucleotides.

³ J. Balzarini, S. Aquaro, T. Knispel, C. Rampazzo, V. Bianchi, C.-F. Perno, E. De Clercq, and C. Meier, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: dNT, 5'(3')-deoxyribonucleotidase; IPTG, isopropyl thio--D-galactoside; GFP, green fluorescent protein; EST, expressed sequence tag; PCR, polymerase chain reaction; HPLC, high pressure liquid chromatography; bp, base pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				K. Wallden, P. Stenmark, T. Nyman, S. Flodin, S. Graslund, P. Loppnau, V. Bianchi, and P. Nordlund Crystal Structure of Human Cytosolic 5'-Nucleotidase II: INSIGHTS INTO ALLOSTERIC REGULATION AND SUBSTRATE RECOGNITION J. Biol. Chem., June 15, 2007; 282(24): 17828 - 17836. [Abstract] [Full Text] [PDF]

				L. R. Chiarelli, P. Bianchi, E. Fermo, A. Galizzi, P. Iadarola, A. Mattevi, A. Zanella, and G. Valentini Functional analysis of pyrimidine 5'-nucleotidase mutants causing nonspherocytic hemolytic anemia Blood, April 15, 2005; 105(8): 3340 - 3345. [Abstract] [Full Text] [PDF]

				M. Proudfoot, E. Kuznetsova, G. Brown, N. N. Rao, M. Kitagawa, H. Mori, A. Savchenko, and A. F. Yakunin General Enzymatic Screens Identify Three New Nucleotidases in Escherichia coli: BIOCHEMICAL CHARACTERIZATION OF SurE, YfbR, AND YjjG J. Biol. Chem., December 24, 2004; 279(52): 54687 - 54694. [Abstract] [Full Text] [PDF]

				A. S. Al-Madhoun, J. Johnsamuel, R. F. Barth, W. Tjarks, and S. Eriksson Evaluation of Human Thymidine Kinase 1 Substrates as New Candidates for Boron Neutron Capture Therapy Cancer Res., September 1, 2004; 64(17): 6280 - 6286. [Abstract] [Full Text] [PDF]

				V. Bianchi and J. Spychala Mammalian 5'-Nucleotidases J. Biol. Chem., November 21, 2003; 278(47): 46195 - 46198. [Full Text] [PDF]

				L. Gallinaro, K. Crovatto, C. Rampazzo, G. Pontarin, P. Ferraro, E. Milanesi, P. Reichard, and V. Bianchi Human Mitochondrial 5'-Deoxyribonucleotidase. OVERPRODUCTION IN CULTURED CELLS AND FUNCTIONAL ASPECTS J. Biol. Chem., September 13, 2002; 277(38): 35080 - 35087. [Abstract] [Full Text] [PDF]

				C. M. Galmarini, K. Graham, X. Thomas, F. Calvo, P. Rousselot, A. El Jafaari, E. Cros, J. R. Mackey, and C. Dumontet Expression of high Km 5'-nucleotidase in leukemic blasts is an independent prognostic factor in adults with acute myeloid leukemia Blood, September 15, 2001; 98(6): 1922 - 1926. [Abstract] [Full Text] [PDF]

				A. M. Marinaki, E. Escuredo, J. A. Duley, H. A. Simmonds, A. Amici, V. Naponelli, G. Magni, M. Seip, I. Ben-Bassat, E. H. Harley, et al. Genetic basis of hemolytic anemia caused by pyrimidine 5' nucleotidase deficiency Blood, June 1, 2001; 97(11): 3327 - 3332. [Abstract] [Full Text] [PDF]

				J. Balzarini, S. Aquaro, T. Knispel, C. Rampazzo, V. Bianchi, C.-F. Perno, E. De Clercq, and C. Meier Cyclosaligenyl-2',3'-didehydro-2',3'-dideoxythymidine Monophosphate: Efficient Intracellular Delivery of d4TMP Mol. Pharmacol., November 1, 2000; 58(5): 928 - 935. [Abstract] [Full Text]

				C. Rampazzo, L. Gallinaro, E. Milanesi, E. Frigimelica, P. Reichard, and V. Bianchi A deoxyribonucleotidase in mitochondria: Involvement in regulation of dNTP pools and possible link to genetic disease PNAS, July 18, 2000; 97(15): 8239 - 8244. [Abstract] [Full Text] [PDF]

				C. Gazziola, P. Ferraro, M. Moras, P. Reichard, and V. Bianchi Cytosolic High Km 5'-Nucleotidase and 5'(3')-Deoxyribonucleotidase in Substrate Cycles Involved in Nucleotide Metabolism J. Biol. Chem., February 23, 2001; 276(9): 6185 - 6190. [Abstract] [Full Text] [PDF]

				S. Allegrini, A. Scaloni, L. Ferrara, R. Pesi, P. Pinna, F. Sgarrella, M. Camici, S. Eriksson, and M. G. Tozzi Bovine Cytosolic 5'-Nucleotidase Acts through the Formation of an Aspartate 52-Phosphoenzyme Intermediate J. Biol. Chem., August 31, 2001; 276(36): 33526 - 33532. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire