INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Recently, nucleosides and a variety of chemically synthesized nucleoside analogs have attracted a great deal of interest, as they have antibiotic, antiviral, and antitumoral effects (9). In light of this trend, we conducted studies on the microbial metabolism of nucleosides (5). In this study, we found that a bacterium, Ochrobactrum anthropi, shows a high level of activity in the N-riboside cleavage of purine nucleosides. This reaction is important in the decomposition of purine nucleosides in foodstuffs which cause hyperuricemia, an increasingly common disease in adults (3).

The enzymatic N-riboside cleavage of nucleosides is a common reaction in various organisms (1, 20). This reaction seems to participate in a salvage or assimilation pathway for nucleosides. Two kinds of enzymes, nucleoside phosphorylases (EC 2.4.2.-) and nucleosidases (nucleoside hydrolase; EC 3.2.2.-), are known to catalyze this reaction. Nucleoside phosphorylases catalyze the phosphorolytic cleavage of nucleosides and show ribosyl transferase activity (7). These enzymes play roles mainly in the salvage pathway. Nucleoside phosphorylases have been well studied. In addition, they have been purified from various sources and used as catalysts for the synthesis of nucleoside analogs through base exchange reactions (7, 21). In contrast, nucleosidases catalyze the irreversible hydrolysis of nucleosides and participate mainly in the assimilation pathway. There have been few studies on microbial nucleosidases acting on purine and pyrimidine nucleosides and no reports of homogenously purified bacterial purine and pyrimidine nucleosidases (8, 18, 19).

In this study, we report the purification of a nucleosidase from O. anthropi which catalyzes the N-riboside cleavage of purine nucleosides. The enzyme was a bacterial purine-specific nucleosidase (purine nucleosidase; EC 3.2.2.1), which has not been studied sufficiently. The nucleotide sequence of the gene encoding the enzyme and the mode of regulation by the pyrimidine nucleoside are also presented.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. All nucleoside derivatives were purchased from Sigma (St. Louis, Mo.). Standard proteins for gel filtration and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were purchased from Oriental Yeast Co. (Tokyo, Japan) and Daiichi Pure Chemical Co. (Tokyo, Japan), respectively. All other reagents were obtained from commercial sources and were of analytical grade.

Microorganism and cultivation. O. anthropi 37a (AKU 995; Faculty of Agriculture, Kyoto University) was used as a source of the enzyme. For investigation of enzyme induction, the bacterium was grown in test tubes (16.5 by 160 mm) containing 5 ml of medium A for 72 h at 28°C with shaking (300 strokes/min). For enzyme purification, the bacterium was grown in 2-liter flasks containing 500 ml of medium B for 47 h at 28°C with shaking (120 strokes/min). Medium A was composed of 3 mM concentrations of various nucleosides, 10 g of glucose, 1 g of K₂HPO₄, 1 g of KH₂PO₄, 4 g of NH₄Cl, 0.3 g of MgSO₄ · 7H₂O, 1 mg of thiamine hydrochloride, 2 mg of riboflavin, 2 mg of nicotinic acid, 2 mg of pantothenic acid, 2 mg of pyridoxine hydrochloride, 0.1 mg of biotin, 1 mg of p-aminobenzoate, and 0.1 mg of folic acid in 1 liter of deionized water, pH 7.0. Medium B comprised 5 g of tryptone, 5 g of yeast extract, 1 g of glucose, 1 g of K₂HPO₄, and 3 mM inosine in 1 liter of tap water, pH 7.0.

Enzyme assays. The standard assay mixture contained, in a 200-µl volume, 0.1 µmol of adenosine, 10 µmol of potassium-phosphate buffer (pH 7.0), and an appropriate amount of enzyme. The mixture was incubated at 30°C and terminated by adding 20 µl of 15% (vol/vol) perchloric acid. The initial reaction velocities were determined based on single time points in 5- to 60-min reactions depending on the experimental conditions. In this time range, reactions were confirmed to proceed linearly. The reaction mixture was centrifuged at 2,150 × g for 10 min, and the supernatant was analyzed for the decrease in the substrate (adenosine) and the increase in the product (adenine) with a Shimadzu (Kyoto, Japan) LC-6A high-performance liquid chromatograph at 260 nm using a Cosmosil 5C₁₈-AR column (4.6 by 100 mm; Nacalai Tesque, Kyoto, Japan) at a flow rate of 1.0 ml/min, with 0.1 M NaClO₄ containing 0.1% (vol/vol) H₃PO₄ as the eluent. Investigation of the enzymatic activity dependence on pH, temperature, chelators, and metals were carried out essentially under the standard assay conditions with slight modifications described below. One unit of the enzyme was defined as the amount of enzyme catalyzing the consumption of adenosine or the formation of adenine at the rate of 1 µmol/min under the assay conditions described above.

Purification of the nucleosidase. All procedures were carried out at 0 to 10°C, and 0.01 M Tris-HCl (pH 7.4) was used as a buffer. Centrifugation was carried out at 14,000 × g for 30 min unless otherwise specified.

Analytical methods for the nucleosidase. The relative molecular mass of the native enzyme and subunit were determined by high-performance liquid chromatography (HPLC) on a TSK G-3000SW column (0.75 by 60 cm; Tosoh, Tokyo, Japan) and SDS-PAGE, respectively, as described previously (10). The relative molecular masses of the native enzyme and the subunit were determined from the relative mobility of the marker proteins purchased from Oriental Yeast Co. and Daiich Chemical Co. (molecular weight marker III), respectively. SDS-PAGE with 12.5% polyacrylamide gels, protein concentration determination, isolation of internal peptides, and amino acid sequencing of NH₂-terminal and internal peptides were performed as described previously (6, 16).

Cloning and nucleotide sequencing. A degenerate oligonucleotide [5'-GA (C/T)ACIGA(A/G)AA(A/G)ATGATIATIGA(C/T)ACIGA(A/G)TT-3'] corresponding to the NH₂-terminal amino acid sequence of the purified purine nucleosidase was synthesized, labeled with digoxigenin (DIG), and then used as a probe for Southern hybridization with fragmented genomic DNA of O. anthropi. The approximately 2-kb DNA fragment that hybridized to the DIG-labeled probe was recovered from HindIII-digested genomic DNA, ligated with HindIII-digested pUC19, and then transformed into Escherichia coli JM109. From the positive clone obtained on colony hybridization with the DIG-labeled probe, the HindIII fragment was recovered and then digested with a variety of restriction enzymes to give convenient DNA fragments for subcloning into pUC19. DNA sequence analysis was performed with synthetic primers by dideoxy chain termination using a SequiTherm long-read cycle sequence kit-LC (Epicentre Technologies, Madison, Wis.) and an autosequencer (dNA sequencer model 4000L; LI-COR, Lincoln, Nebr.).

Expression in E. coli. A DNA fragment including the coding region for the purine nucleosidase was obtained by PCR. The sense primer comprised an EcoRI recognition site (underlined sequence below) and 23 nucleotides (about 170 bp upstream of the purine nucleosidase coding region, of which the sequence matches nucleotides 115 to 137 of the sequence under GenBank accession number E15229). The antisense primer contained a BamHI recognition site (underlined sequence below) and 23 nucleotides, of which the sequence is complementary to that about 20 bp downstream of the purine nucleosidase coding region (nucleotides 1393 to 1415 of the sequence under GenBank accession number E15229). The two primers were as follows: sense primer, 5'-GCGACAGAATTCACGCAATGACC-3'; antisense primer, 5'-GCCAATTCTAACCTAGGGACCAT-3'. The amplified PCR product was digested with EcoRI and BamHI, separated by agarose gel electrophoresis, and then purified. The amplified DNA was inserted into the EcoRI-BamHI site of pUC19 and then digested with EcoRI and HindIII. The EcoRI-HindIII fragment was inserted into pKK233-3, yielding pKN113, which was then used to transform E. coli JM109 cells.

Nucleotide sequence accession number. The nucleotide sequence obtained for the gene fragment encoding the purine nucleosidase of O. anthropi appears in the DDBJ, EMBL, and GenBank databases under accession no. E15229.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Induction and purification of the enzyme. The specific activity in purine nucleoside N-glucoside cleavage in a cell extract of O. anthropi was greatly enhanced when the bacterium was cultivated in the medium containing a purine nucleoside. Inosine, guanosine, xanthosine, and adenosine in medium A (3 mM) increased the specific activity 25-, 16-, 5.8-, and 5.1-fold, respectively, above that without a purine nucleoside (1.7 × 10² U/mg). The enzyme was purified from cells cultivated with inosine. Dialysis of the cell extract against a buffer (0.01 M Tris-HCl, pH 7.4) to remove phosphate did not depress but rather enhanced the activity (Table 1), suggesting that the enzyme is not a phosphorylase but a nucleosidase. The increase in total activity before the first DEAE-Sephacel column chromatography indicated the existence of endogenous inhibitors in the cell extract. The enzyme was purified approximately 900-fold (Table 1). The purified enzyme gave a single band with more than 99% purity on SDS-PAGE, corresponding to a subunit molecular mass of 40 kDa. The homogeneity of the enzyme was confirmed by HPLC elution, with a single symmetrical peak corresponding to a native molecular mass of 172 kDa being obtained, suggesting a homotetrameric enzyme.

Substrate specificity and kinetic properties. The enzyme showed high levels of activity and affinity toward adenosine (Table 2). All other tested purine nucleosides, i.e., guanosine, inosine, and xanthosine, served as preferred substrates in addition to adenosine. Other than these, only 3'-AMP and 3'-deoxyadenosine served as substrates, but these were far less suitable, their relative activities being 0.67 and 0.0033%, respectively, of that of adenosine. Other purine nucleotides, deoxypurine nucleosides, pyrimidine nucleosides, pyrimidine nucleotides, deoxypyrimidine nucleosides, -NAD⁺, -NADP⁺, and -nicotinamide mononucleotide did not serve as substrates for the enzyme.

Inhibition by pyrimidine nucleosides. The effects of nonsubstrate nucleosides and nucleotides on purine nucleoside hydrolysis by the purified enzyme were examined. Among various nucleosides and nucleotides tested at a final concentration of 0.5 mM under the standard assay conditions, the pyrimidine nucleosides of cytidine and uridine were found to be inhibitory (Table 3). Kinetic analysis by Lineweaver-Burk plots with enough reliability (r² values were more than 0.913) revealed mixed inhibition by cytidine. The K_i (dissociation constant for free enzyme) and K_i' (dissociation constant for enzyme-substrate complex) values obtained by K_m/V_max versus [I] and 1/V_max versus [I] plots, respectively, are presented in Table 3.

Effects of chemicals. The effects of various compounds and metal ions were examined at a final concentration of 1 mM in the standard reaction mixture. Among the various metal ions tested, only Ca²⁺ slightly enhanced the enzyme activity to 120% of the original level. Cu²⁺, Mn²⁺, Hg²⁺, and Ag⁺ were rather inhibitory (57, 34, 61, and 68% inhibition, respectively). The enzyme exhibited sensitivity towards metal ion chelators, such as EDTA, 8-hydroxyquinoline, and o-phenanthroline (46, 93, and 23% inhibition, respectively). After the enzyme was preincubated with 8-hydroxyquinoline for 10 min and then dialyzed, the effects of divalent metal ions (Fe²⁺, Mg²⁺, Ca²⁺, Ba²⁺, Zn²⁺, Co²⁺, Ni²⁺, Cu²⁺, Mn²⁺, Sn²⁺, and Pb²⁺) on the reactivation of 8-hydroxyquinoline-treated enzyme were tested. A concentration of 3 mM Zn²⁺ or Co²⁺ restored the activity inhibited by 1 mM 8-hydroxyquinoline (70% inhibition) to the initial level, suggesting that such divalent metal ions are involved in the activity. NaCN, NaF, NaN₃, and NaAsO₂ also showed mild inhibition (41, 38, 27, and 31% inhibition, respectively).

Enzyme stability and activity. After 30 min of incubation at various temperatures in 100 mM potassium-phosphate buffer, pH 6.0, the enzyme was found to be stable up to 40°C. At 50°C, about 50% of its activity remained but only 20% remained at 60°C. The pH stability was tested by 30 min of incubation of the enzyme in 100 mM concentrations of buffers of various pHs at 30°C. The enzyme retained more than 80% of its initial activity between pHs 6.0 and 8.5 but only 40 and 50% of its activity at pHs 4.0 and 9.0, respectively.

Gene cloning, sequencing, and expression in E. coli. By colony hybridization with an oligonucleotide corresponding to the NH₂-terminal amino acid sequence as a probe and successive gene walking, a 1.5-kb fragment encoding the complete nucleosidase was obtained and sequenced. The open reading frame encodes a protein of 363 amino acids (exact molecular mass, 39,935.63 Da) containing amino acid sequences identical to the NH₂-terminal and internal amino acid sequences of the purified purine nucleosidase (Fig. 1). A leader peptide of 22 amino acids was found before the NH₂ terminus of the purified nucleosidase. A predicted ribosome-binding site (Shine-Dalgarno sequence, AGGAGGA [17]) was found 8 bp before the start codon of the gene. An apparent promoter exhibiting homology to the E. coli consensus promoter (14) was found in the 56- to 102-bp upstream region of the start codon.

View larger version (77K): [in this window] [in a new window]   FIG. 1.   Alignment of the deduced amino acid sequences of the nucleosidase from O. anthropi (O. anthropi Pu-N), inosine-adenine-guanosine N-ribohydrolase from T. brucei brucei (T. brucei IAG-N), inosine-uridine nucleoside hydrolase from C. fasciculata (C. fasciculata IU-N), and inosine-uridine-preferring nucleoside hydrolase from L. major (L. major IU-N). Brackets indicate the regions of significant similarity. Arrows indicate putative Ca2+-binding Asp residues. Symbols: *, amino acids common to all sequences; , Gln-190 and Asn-192 located in the proposed catalytic site pocket; , Phe-262 involved in the leaving-group specificity. The underlined sequences represent the amino acid sequences determined previously from those of NH2-terminal and internal peptides.

4

Homology. Using the BLASTP program on the NR-AA database, homology of the deduced amino acid sequence with those of three well-characterized nucleosidases from protozoan parasites, i.e., inosine-uridine-preferring nucleoside hydrolase from Leishmania major (15) (29% identical over 334 amino acids), inosine-uridine nucleoside hydrolase from Crithidia fasciculata (2, 4, 11, 12) (28% identical over 335 amino acids), and inosine-adenosine-guanosine N-ribohydrolase from Trypanosoma brucei brucei (13) (29% identical over 189 amino acids), was found. The homologues with more than 23% identity over more than 300 amino acids were found in bacteria (E. coli, Pseudomonas aeruginosa, Bacillus halodurans, Campylobacter jejuni, Desulfurolobus ambivalens, and Mycobacterium tuberculosis), fungi (Saccharomyces cerevisiae and Schizosaccharomyces pombe), plants (Arabidopsis thaliana), mosozoa (Caenorhabditis elegans), and insects (Drosophila melanogaster).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Only a few studies have been performed so far on the characterization of bacterial nucleosidases (EC 3.2.2.1). Two bacterial nucleosidases were partially purified from Lactobacillus delbrueckii (18) and Pseudomonas fluorescens (19). The enzymes from L. delbrueckii and P. fluorescens hydrolyze both purine and pyrimidine nucleosides, although their preferred substrates are purine and pyrimidine nucleosides, respectively. In this study, a bacterial nucleosidase was first purified to homogeneity from O. anthropi. It showed strict specificity to purine nucleosides and one order smaller K_m values than those of the nucleosidases from L. delbrueckii and P. fluorescens. These results together with the induction of the enzyme by purine nucleosides suggested that the O. anthropi enzyme plays a specific role in purine nucleoside metabolism. The purine nucleoside hydrolyzing activity of the nucleosidase from O. anthropi was inhibited by pyrimidine nucleosides, especially cytidine, with quite small K_i and K_i' values. These pyrimidine nucleosides may be endogenous inhibitors in cell extracts which were removed by dialysis and successive DEAE-Sephacel column chromatography. The pyrimidine nucleoside hydrolyzing activity of the P. fluorescens nucleosidase was reported to be inhibited by purine nucleosides (19). These results indicate that the activities of bacterial purine and pyrimidine nucleosidases are mutually controlled by each other's substrates, probably to maintain the proper balance of purine and pyrimidine nucleoside concentrations in cells.

A considerable number of studies have been performed on nucleosidases from protozoan parasites such as C. fasciculata (2, 4, 11, 12). T. brucei brucei (13), and L. major (15). Because they are deficient in de novo nucleoside synthesis and rely on a salvage pathway, inhibitors of their nucleosidases are potential pharmaceuticals for trypanosomes and related pathogens. The nucleosidases from trypanosomes are classified based on their substrate specificities into inosine-adenosine-guanosine-preferring (IAG), inosine-uridine-preferring (IU), and guanosine-inosine-preferring (GI) nucleosidases. The amino acid sequence of the nucleosidase from O. anthropi exhibits homology with those of trypanosome nucleosidases. Regions of significant amino acid sequence similarity include the 25 NH₂-terminal amino acids (first region) and the regions comprising amino acids 189 to 200 (second region) and 262 to 285 (third region) (Fig. 1). The NH₂-terminal region includes highly conserved Asp residues. The X-ray crystal structure of the IU nucleosidase from C. fasciculata revealed that the NH₂-terminal Asp residues are clustered near the catalytic site and tightly hold Ca²⁺ (2). The second group of conserved amino acids, 189 to 200, contains the AEXNXXXDPXAA motif. Judging from the crystal structure of the IU nucleosidase, the amino acids corresponding to Gln-190 and Asn-192 in the O. anthropi nucleosidase are also located in the proposed catalytic site pocket (2, 13). The third conserved region, 262 to 285, in the O. anthropi nucleosidase contains Asp-263, which corresponds to Asp-242 in the IU nucleosidase, a residue that is in contact with the catalytic site, Ca²⁺ (2, 13). Thus, the conserved regions are associated with catalytic site elements despite the substantial differences in substrate specificity. No amino acid corresponding to His-241 of the IU nucleosidase, which was demonstrated to be the proton donor involved in leaving group activation, was found, like in the IAG nucleosidase from T. brucei brucei, indicating that different amino acids are involved in the case of aglycone activation in purine-preferring nucleosidases (2, 13). Phe-262 of the O. anthropi nucleosidase likely corresponds to Trp-242 of the IAG nucleosidase, which seems to be involved in its leaving group specificity for purine as a result of base stacking with Phe or Trp (13, 15). The X-ray crystal structure of the IU nucleosidase from L. major revealed the existence of tightly bound Ca²⁺ (15), but no kinetic evidence for the requirement of added metal ions was obtained. The O. anthropi nucleosidase has a conserved Ca²⁺ binding sequence, and Zn²⁺ and Co²⁺ restore the activity inhibited by metal ion chelators, indicating that some metal ion is involved in the activity.

Homologous sequences with that of O. anthropi nucleosidase are widely distributed not only in bacteria but also in fungi, protozoa, mosozoa, arthropoda, and plants, indicating that such nucleoside hydrolyzing activity, as well as nucleoside phosphorylases, plays an important role in many organisms.