Stereospecific 2'-amination and 2'-chlorination of adenosine by Actinomadura in the biosynthesis of 2'-amino-2'-deoxyadenosine and 2'-chloro-2'-deoxycoformycin

2'-Amino-2'-deoxyadenosine and 2'-chloro-2'-deoxycoformycin (2'-CldCF) are two nucleoside antibiotics produced by Actinomadura. The biosynthesis of these two nucleoside antibiotics has been studied by the addition of [U-14C]adenosine with or without unlabeled adenine to cultures of Actinomadura. By this experimental approach, it is possible to demonstrate that adenosine is the direct precursor for the biosynthesis of 2'-amino-2'-deoxyadenosine and 2'-CldCF. These conclusions are based on the observation that the percentage distribution of 14C in the aglyconic and pentofuranosyl moieties of 2'-amino-2'-deoxyadenosine and 2'-CldCF were similar to the distribution of 14C in the adenine and ribosyl moieties of the [U-14C]adenosine (i.e., 48:52) added to cultures of Actinomadura. Experimentally, the percentage distribution of 14C in the (i) adenine:2-amino-2-deoxy-beta-D-ribofuranose of 2'-amino-2'-deoxyadenosine is 51:49; (ii) 8-(R)-3,6,7,8-tetrahydroimidazo[4,5-d]-[1,3-diazepin-8-o1]:2 -chloro-2- beta-D-ribofuranose of 2'-CldCF is 45:55; and (iii) adenine:ribose of the adenosine isolated from the RNA of Actinomadura is 42:58. Further proof that adenosine is the direct precursor for the biosynthesis 2'-amino-2'-deoxyadenosine and 2'-CldCF was demonstrated by the addition of 75 mumol of unlabeled adenine together with [U-14C]adenosine to nucleoside-producing cultures of Actinomadura. The percentage distribution of 14C in the aglycon and the sugar moieties of 2'-amino-2'-deoxyadenosine and 2'-CldCF were 46:54 and 47:53, respectively; the percentage distribution of 14C in the adenine and ribose moieties of the adenosine isolated from the RNA of Actinomadura was 51:49. These data show that the hydroxyl on C-2' of the ribosyl moiety of adenosine undergoes a replacement by a 2'-amino or a 2'-chloro group to form 2'-amino-2'-deoxyadenosine or 2'-CldCF with retention of stereconfiguration at C-2'. Finally, Actinomadura can utilize inorganic chloride from the medium as demonstrated by the isolation of [36Cl]2'-CldCF following the addition of [36Cl]chloride to the culture medium. Mechanisms for the regioselective modification of the C-2' hydroxyl group and stereospecific insertion of the amino and chloro groups are discussed.     

The role of platelet-derived endothelial cell growth factor/thymidine phosphorylase in tumor behavior

Platelet-derived endothelial cell growth-factor (PD-ECGF) is similar to the pyrimidine enzyme thymidine phosphorylase (TP). A high TP expression at tumor sites is correlated with tumor growth, induction of angiogenesis, and metastasis. Therefore, high TP is most likely associated with a poor prognosis. TP is not only expressed in tumor cells but also in tumor surrounding tissues, such as tumor infiltrating macrophages. TP catalyzes the conversion of thymidine to thymine and doxyribose-1-phosphate (dR-1-P). The latter in its parent form or in its sugar form, deoxyribose (dR) may play a role in the induction of angiogenesis. It may modulate cellular energy metabolism or be a substrate in a chemical reaction generating reactive oxygen species. L-deoxyribose (L-dR) and thymidine phosphorylase inhibitor (TPI) can reverse these effects. The mechanism of TP induction is not yet completely clear, but TNF, IL10 and other cytokines have been clearly shown to induce its expression. The various complex interactions of TP give it an essential role in cellular functioning and, hence, it is an ideal target in cancer therapy.     

Methanobacterium thermoformicicum thymine DNA mismatch glycosylase: conversion of an N-glycosylase to an AP lyase.

The thymine DNA mismatch glycosylase from Methanobacterium thermoformicicum, a member of the endonuclease III family of repair proteins, excises the pyrimidine base from T-G and U-G mismatches. Unlike endonuclease III, it does not cleave the phosphodiester backbone by a beta-elimination reaction. This cleavage event has been attributed to a nucleophilic attack by the conserved Lys120 of endonuclease III on the aldehyde group at C1' of the deoxyribose and subsequent Schiff base formation. The inability of TDG to perform this beta-elimination event appears to be due to the presence of a tyrosine residue at the position equivalent to Lys120 in endonuclease III. The purpose of this work was to investigate the requirements for AP lyase activity. We replaced Tyr126 in TDG with a lysine residue to determine if this replacement would yield an enzyme with an associated AP lyase activity capable of removing a mismatched pyrimidine. We observed that this replacement abolishes the glycosylase activity of TDG but does not affect substrate recognition. It does, however, convert the enzyme into an AP lyase. Chemical trapping assays show that this cleavage proceeds through a Schiff base intermediate and suggest that the amino acid at position 126 interacts with C1' on the deoxyribose sugar.     

Reactivity and selectivity in light-induced free radical reactions of 2-propanol with purine and pyrimidine mononucleotides and dinucleoside monophosphates.

Photoalkylation reactions with 2-propanol, initiated with di-tert-butyl peroxide, of a variety of purine and pyrimidine mononucleotides and dinucleoside monophosphates lead to the substitution of an alpha-hydroxyisopropyl group for the H-8 atom of adenosine and the addition of the alcohol across the 5,6-double bond of the pyrimidines. Adenosine moieties blocked at their 3'-hydroxyl group are alkylated faster than those blocked at their 5'-hydroxyl. The reactivity of the uridine moieties of 3'-UMP, 5'-UMP, and uridylyl-(3',5')-uridine is not affected by the location of the phosphate group. However, the uridine moiety of uridylyl-(3',5')-adenosine is modified faster than that of adenylyl-(3',5')-uridine. It is suggested that steric hindrance imposed by the phosphate group determines the reactivity of adenosine moieties, while base stacking involving adenosine determines the reactivity of uridine moieties. These two effects play a major role in controlling the nature and degree of the selectivity of these photoalkylation reactions for either adenosine or uridine. Cytidine has been found to be inert in these reactions.     

Effect of nucleotide substitution on the peptidyltransferase activity of 2'(3')-O-(aminoacyl) oligonucleotides

Seven 2'(3')-O-(aminoacyl) trinucleotides with structures derived from the 3'-terminal C-C-A sequence of aa-tRNA via nucleotide substitutions were investigated as acceptor substrates in the peptidyltransferase reaction and as inhibitors of substrate binding to the peptidyltransferase A site. It was found that all tested compounds were active in both systems, although substitution in the first and second nucleotide position results in some decrease of acceptor activity. Remarkably, replacement of natural cytidylic acid residues in C-C-A-Phe with guanylic acid moieties resulted only in a small decrease of acceptor or binding activity. The results indicate that the acceptor sequence of aa-tRNA is not probably engaged in base pairing with a sequence of 23S RNA during its interaction with the peptidyltransferase A site.     

Mechanism of DNA cleavage and substrate recognition by a bovine apurinic endonuclease

The location of the phosphodiester bond cleaved by homogeneous Mg2+-dependent apurinic endodeoxyribonuclease (EC 3.1.25.2; APE) of bovine calf thymus has been determined by using a 21-mer oligonucleotide containing a single central apurinic site as a substrate. A single product of cleavage consistent with cleavage of the oligonucleotide 5' to the apurinic site, and leaving a 3' hydroxyl group, was identified. This enzyme is, therefore, a class II apurinic endonuclease. The substrate specificities of this enzyme have been determined by using a variety of natural and synthetic DNAs or oligonucleotides containing base-free sites. Calf thymus APE has an absolute requirement for a double-stranded DNA and requires an abasic site as a substrate. The presence of a base fragment such as a urea residue, an alkoxyamine group attached to the C'-1 position of the abasic site, or reduction of the C'-1 aldehyde abolishes the APE activity of this enzyme. Synthetic abasic sites containing either ethylene glycol, propanediol, or tetrahydrofuran interphosphate linkages are excellent substrates for bovine APE. These results indicate that APE has no absolute requirement for either ring-opened or ring-closed deoxyribose moieties in its recognition of DNA-cleavage substrates. The enzyme may interact with the pocket in duplex DNA that results from the base loss or with the altered conformations of the phosphodiester backbone that result from the abasic site.     

Comparison of the interaction of uridine, cytidine, and other pyrimidine nucleoside analogues with recombinant human equilibrative nucleoside transporter 2 (hENT2) produced in Saccharomyces cerevisiae.

The human equilibrative nucleoside transporters I and 2 (hENT1, hENT2) share 50% amino acid identity and exhibit broad selectivities, accepting purine and pyrimidine nucleosides as permeants. The permeant selectivity of hENT2 is less well understood because of the low abundance of the native transporter in cells amenable to functional analysis. Recent studies of hENT2 produced in recombinant form in functional expression systems have shown that it differs from hENT1 in that it transports nucleobases. To further understand the structural requirements for permeant interaction with hENT2, we compared the relative abilities of uridine, cytidine, and their analogues to inhibit transport of [3H]uridine by recombinant hENT1 and hENT2 produced in yeast. hENT1 and hENT2 tolerated halogen modification at the 5 position of the base and the 2' and 5' positions of the ribose moieties of uridine whereas removal of the hydroxyl group at the 3' position of the ribose moiety of uridine eliminated interaction with both transporters. hENT2 displayed a lower ability, compared with hENT1, to interact with cytidine and cytidine analogues, suggesting a low tolerance for the presence of the amino group at the 4 position of the base.     

5-Nitrouridine-monohydrate: crystal structure and conformation.

The crystal structure of 5-nitrouridine was determined by X-ray analysis. The pyrimidine ring is slightly non-planar, showing a shallow boat conformation. The nitro group has no influence on the C4 - O4 bond length as compared to uridine. The ribose shows the C3'-endo conformation and the base is in the anti orientation to the sugar with a torsion angle of 25.6 degrees. This conformation is stabilized by a hydrogen bond from the base to the ribosyl moiety (H6 ... 05'). Stacking interactions between neighboring bases are almost negligible in the crystal. A water molecule is involved in a bifurcated donating hydrogen bond to 04 and to 052 of the nitro group of the one base and an accepting bond from the H3 of the other base. Two more hydrogen bonds are formed between the water molecule and the ribose. The structural aspects of 5-nitrouridine are discussed with respect to the special stacking features found for 5-nitro-1-(beta-D-ribosyluronic acid)-uracil monohydrate in the crystal (1).     

Protection against DNA damage-induced apoptosis by the angiogenic factor thymidine phosphorylase

Thymidine phosphorylase (TP) is involved both in pyrimidine nucleoside metabolism and in angiogenesis. TP also conferred the resistance to hypoxia-induced apoptosis of the cancer cells. In U937 cells, DNA damage-inducing agents significantly enhanced the expression of TP. Cell lines stably transfected with TP cDNA were more resistant to the DNA damage-inducing agents than the mock-transfected cells and showed augmented activity of Akt. The cytoprotective function of TP against DNA damage was independent of its enzymatic activity. The resistance to apoptosis was partially abrogated by treatment with the phosphatidyl inositol 3-kinase (PI3K) inhibitors, suggesting that the cytoprotective function of TP is mediated, at least in part, by regulation of the PI3K/Akt pathway. These findings indicate that TP expression in increased by various stress including DNA damage and that TP molecules confer resistance to DNA damage-induced apoptosis in cancer cells.     

The role of thymidine phosphorylase and uridine phosphorylase in (fluoro)pyrimidine metabolism in peripheral blood mononuclear cells

Thymidine phosphorylase (TP) and uridine phosphorylase (UP) catalyze the (in)activation of several fluoropyrimidines, depending on their catalytic activity and substrate specificity. Blood cells are the first compartment exposed to most anticancer agents. The role of white blood cells in causing toxic side effects and catalyzing drug metabolism is generally underestimated. Therefore we determined the contribution of the white blood cell compartment to drug metabolism, and we investigated the activity and substrate specificity of TP and UP for the (fluoro)pyrimidines thymidine (dThd), uridine (Urd), 5'-deoxy-5-fluorouridine (5' dFUrd) and 5-fluorouracil (5FU) in peripheral blood mononuclear cells (PBMC) and undifferentiated monocytes and differentiated monocytes: macrophages and dendritic cells. PBMC had an IC50 of 742 microM exposed to 5'dFUrd, increasing to > 2000 microM when both TP and UP activities were inhibited. Total phosphorolytic activity was higher with dThd than with Urd, 5'dFUrd or 5FU. Using a specific TP inhibitor (TPI) and UP inhibitor (BAU) we concluded that dThd and Urd were preferentially converted by TP and UP, respectively, while 5'dFUrd and 5FU were mainly converted by TP (about 80%) into 5FU and FUrd, respectively. 5FU was effectively incorporated into RNA. dThd conversion into thymine was highest in dendritic cells (52.6 nmol thymine/h/10(6) cells), followed by macrophages (two-fold) and undifferentiated monocytes (eight-fold). TPI prevented dThd conversion almost completely. In conclusion, PBMC were relatively insensitive to 5'dFUrd, and the natural substrates dThd and Urd were preferentially converted by TP and UP, respectively. TP and UP were both responsible for converting 5'dFUrd/5FU into 5FU/FUrd, respectively.     

Carbocyclic thymidine derivatives efficiently inhibit Plasmodium falciparum thymidylate kinase (PfTMK)

During the course of our research into new anti-malaria drugs, Plasmodium falciparum thymidylate kinase (PfTMK) has emerged as an important drug target because of its unique substrate specificity. Compared with human thymidylate kinase (HsTMK), PfTMK shows broader substrate specificity, which includes both purine and pyrimidine nucleotides. PfTMK accepts both 2'-deoxyguanosine monophosphate (dGMP) and thymidine monosphosphate (TMP) as substrates. We have evaluated the inhibitory activity of seven carbocyclic thymidine analogs and report the first structure-activity relationship for these inhibitors against PfTMK. The 2',3' dideoxycarbocyclic derivative of thymidine showed the most potent inhibition of the enzyme. The K(i)(dTMP) and K(i)(dGMP) values were 20 and 7 μM respectively. Thus, further modifications of carbocyclic thymidine analogs represent a good strategy for developing more powerful thymidylate kinase inhibitors.     

Development and validation of a quantitative method for thymidine phosphorylase activity in peripheral blood mononuclear cells

Abstract

The enzyme thymidine phosphorylase (TP) is important for activation of capecitabine and 5-fluorouracil. Assessment of TP phenotype might be suitable for identification of patients at risk of fluoropyrimidine-induced toxicity. In this paper, we describe the development and validation an assay for TP activity in peripheral blood mononuclear cells (PBMCs). The assay was based on ex vivo conversion of the TP substrate thymidine to thymine. The amount of thymine formed was determined by high-performance liquid chromatography – ultraviolet detection (HPLC-UV) with 5-bromouracil as internal standard. Lymphocytes and monocytes were purified from isolated PBMCs to examine cell-specific TP activity. TP activity in PBMCs demonstrated Michaelis-Menten kinetics. The lower limit of quantification was 2.3?µg PBMC protein and assay linearity was demonstrated up to 22.7?µg PBMC protein. Within-day and between-day precisions were ≤9.2% and ≤6.0%, respectively. Adequate stability TP activity was demonstrated after long-term storage of PBMC dry pellets and lysates at ?80?°C. In monocytes, TP activity was approximately 3 times higher than in lymphocytes. Clinical applicability was demonstrated in samples that were collected from five cancer patients. A simple, precise and sensitive HPLC-UV assay for quantification of TP activity in PBMCs was developed that can be applied for clinical research.     

Biosynthesis of triphosphoribosyl-dephospho-coenzyme A, the precursor of the prosthetic group of malonate decarboxylase

Malonate decarboxylase from Klebsiella pneumoniae consists of four subunits MdcA, D, E, and C and catalyzes the cleavage of malonate to acetate and CO(2). The smallest subunit MdcC is an acyl carrier protein to which acetyl and malonyl thioester residues are bound via a 2'-(5' '-phosphoribosyl)-3'-dephospho-CoA prosthetic group and turn over during the catalytic mechanism. We report here on the biosynthesis of holo acyl carrier protein from the unmodified apoprotein. The prosthetic group biosynthesis starts with the MdcB-catalyzed condensation of dephospho-CoA with ATP to 2'-(5' '-triphosphoribosyl)-3'-dephospho-CoA. In this reaction, a new alpha (1' ' --> 2') glycosidic bond between the two ribosyl moieties is formed, and thereby, the adenine moiety of ATP is displaced. MdcB therefore is an ATP:dephospho-CoA 5'-triphosphoribosyl transferase. The second protein involved in holo ACP synthesis is MdcG. This enzyme forms a strong complex with the 2'-(5' '-triphosphoribosyl)-3'-dephospho-CoA prosthetic group precursor. This complex, called MdcG(i), is readily separated from free MdcG by native polyacrylamide gel electrophoresis. Upon incubation of MdcG(i) with apo acyl carrier protein, holo acyl carrier protein is synthesized by forming the phosphodiester bond between the 2'-(5' '-phosphoribosyl)-3'-dephospho-CoA prosthetic group and serine 25 of the protein. MdcG corresponds to a 2'-(5' '-triphosphoribosyl)-3'-dephospho-CoA:apo ACP 2'-(5' '-phosphoribosyl)-3'-dephospho-CoA transferase. In absence of the prosthetic group precursor, MdcG catalyzes at a low rate the adenylylation of apo acyl carrier protein using ATP as substrate. The adenylyl ACP thus formed is an unphysiological side product and is not involved in the biosynthesis of holo ACP. The 2'-(5' '-triphosphoribosyl)-3'-dephospho-CoA precursor of the prosthetic group has been purified and its identity confirmed by mass spectrometry and enzymatic analysis.     

Structure of the substrate complex of thymidine kinase from Ureaplasma urealyticum and investigations of possible drug targets for the enzyme

Thymidine kinases have been found in most organisms, from viruses and bacteria to mammals. Ureaplasma urealyticum (parvum), which belongs to the class of cell-wall-lacking Mollicutes, has no de novo synthesis of DNA precursors and therefore has to rely on the salvage pathway. Thus, thymidine kinase (Uu-TK) is the key enzyme in dTTP synthesis. Recently the 3D structure of Uu-TK was determined in a feedback inhibitor complex, demonstrating that a lasso-like loop binds the thymidine moiety of the feedback inhibitor by hydrogen bonding to main-chain atoms. Here the structure with the substrate deoxythymidine is presented. The substrate binds similarly to the deoxythymidine part of the feedback inhibitor, and the lasso-like loop binds the base and deoxyribose moieties as in the complex determined previously. The catalytic base, Glu97, has a different position in the substrate complex from that in the complex with the feedback inhibitor, having moved in closer to the 5'-OH of the substrate to form a hydrogen bond. The phosphorylation of and inhibition by several nucleoside analogues were investigated and are discussed in the light of the substrate binding pocket, in comparison with human TK1. Kinetic differences between Uu-TK and human TK1 were observed that may be explained by structural differences. The tight interaction with the substrate allows minor substitutions at the 3 and 5 positions of the base, only fluorine substitutions at the 2'-Ara position, but larger substitutions at the 3' position of the deoxyribose.     

Breaking pseudo-twofold symmetry in the poliovirus 3'-UTR Y-stem by restoring Watson-Crick base pairs

The previously described NMR structure of a 5'-CU-3'/5'-UU-3' motif, which is highly conserved within the 3'-UTR Y-stem of poliovirus-like enteroviruses, revealed striking regularities of the local helix geometry, thus retaining the pseudo-twofold symmetry of the RNA helix. A mutant virus with both pyrimidine base pairs changed into Watson-Crick replicated as wild type, indicating the functional importance of this symmetry relation in viral RNA replication. Here we investigated the effect of changing only one of the two pyrimidine base pairs to Watson-Crick. We determined the NMR structures of two Y-stem variants: one containing the 5'-CU-3'/5'-AU-3' motif, which has been found in wild-type virus isolates as well, and the other containing a 5'-CU-3'/5'-UG-3' motif, which is not present in any enterovirus sequenced to date. Both structures show single pyrimidine mismatches with intercalated bases. In the 5'-CU-3'/5'-AU-3' motif a C-U Watson-Crick-type base pair is formed that retains the pseudo-twofold symmetry, while in the 5'-CU-3'/5'-UG-3' motif a single asymmetric U-U mismatch breaks the twofold symmetry. Surprisingly, for the nonnatural variant no effect of the single base-pair replacement was observed on polioviral RNA replication using an in vitro replicon assay.     

Role of superoxide in radiation-killing of Escherichia coli and in thymine release from thymidine

The role of superoxide and hydroxyl radicals in gamma-radiation-killing of Escherichia coli K12 was studied in aerated suspensions supplemented with formate, phosphate, superoxide dismutase, catalase and saturated with nitrous oxide. Nitrous oxide, which converts e-aq to .OH, caused decreased radiosensitivity. On the other hand, formate, which results in conversion of .OH to .O2-, resulted in an increased radiosensitivity. The results implicated .O2- as a major cause of radiation-mediated cell-killing. The addition of the enzymes, superoxide dismutase or catalase to the E. coli suspensions prior to and during irradiation had no effect on cell survival, indicating that the biologically significant site of generation and action of .O2- is an intracellular one. Further studies were undertaken to examine the role of superoxide in DNA damage. The release of thymine from the DNA base, thymidine was studied as a result of gamma-irradiation and of chemically generated superoxide (using KO2 in dimethyl sulfoxide). Thymine was identified by HPLC and mass spectrometry. C-13 NMR analysis of the reaction mixture of thymidine with KO2 in dimethyl sulfoxide provided evidence for attack of .O2 at the ribosyl Cl' atom.     

Local sequence dependence of rate of base replacement in mammals.

I have analysed the local sequence context of base replacement changes in 78 processed pseudogenes. Transversions occur more often than transitions in a ratio of 3.37 to 1, and G:C is replaced 1.4 times more frequently than A:T. In addition, the bases to the 5' and 3' of the mutating base also influence the rate at which bases change, purine:pyrimidine and pyrimidine:purine pairs changing 1.2 times as fast as purine:purine and pyrimidine:pyrimidine pairs. I discuss implications of this for the mechanism of DNA polymerization in mammals.     

Mechanism of damage recognition by Escherichia coli DNA photolyase

Escherichia coli DNA photolyase binds to DNA containing pyrimidine dimers with high affinity and then breaks the cyclobutane ring joining the two pyrimidines of the dimer in a light- (300-500 nm) dependent reaction. In order to determine the structural features important for this level of specificity, we have constructed a 43 base pair (bp) long DNA substrate that contains a thymine dimer at a unique location and studied its interaction with photolyase. We find that the enzyme protects a 12-16-bp region around the dimer from DNase I digestion and only a 6-bp region from methidium propyl-EDTA-Fe (II) digestion. Chemical footprinting experiments reveal that photolyase contacts the phosphodiester bond immediately 5' and the 3 phosphodiester bonds immediately 3' to the dimer but not the phosphodiester bond between the two thymines that make up the dimer. Methylation protection and interference experiments indicate that the enzyme makes major groove contacts with the first base 5' and the second base 3' to the dimer. These data are consistent with photolyase binding in the major groove over a 4-6-bp region. However, major groove contacts cannot be of major significance in substrate recognition as the enzyme binds equally well to a thymine dimer in a 44-base long single strand DNA and protects a 10-nucleotide long region around the dimer from DNase I digestion. It is therefore concluded that the unique configuration of the phosphodiester backbone in the strand containing the pyrimidine dimer, as well as the cyclobutane ring of the dimer itself are the important structural determinants of the substrate for recognition by photolyase.     

Biosynthesis of 9-beta-D-arabinofuranosyladenine: hydrogen exchange at C-2' and oxygen exchange at C-3' of adenosine

The data presented here describe new findings related to the bioconversion of adenosine to 9-beta-D-arabinofuranosyladenine (ara-A) by Streptomyces antibioticus by in vivo investigations and with a partially purified enzyme. First, in double label in vivo experiments with [2'-18O]- and [U-14C]adenosine, the 18O:14C ratio of the ara-A isolated does not change appreciably, indicating a stereospecific inversion of the C-2' hydroxyl of adenosine to ara-A with retention of the 18O at C-2'. In experiments with [3'-18O]- and [U-14C]-adenosine, [U-14C]ara-A was isolated; however, the 18O at C-3' is below detection. The adenosine isolated from the RNA from both double label experiments has essentially the same ratio of 18O:14C. Second, an enzyme has been isolated and partially purified from extracts of S. antibioticus that catalyzes the conversion of adenosine, but not AMP, ADP, ATP, inosine, guanosine, or D-ribose, to ara-A. In a single label enzyme-catalyzed experiment with [U-14C]adenosine, there was a 9.9% conversion to [U-14C]ara-A; with [2'-3H]-adenosine, there was a 8.9% release of the C-2' tritium from [2'-3H]adenosine which was recovered as 3H2O. Third, the release of 3H as 3H2O from [2'-3H]adenosine was confirmed by incubations of the enzyme with 3H2O and adenosine. Ninety percent of the tritium incorporated into the D-arabinose of the isolated ara-A was in C-2 and 8% was in C-3. The enzyme-catalyzed conversion of adenosine to ara-A occurs without added cofactors, displays saturation kinetics, a pH optimum of 6.8, a Km of 8 X 10(-4) M, and an inhibition by heavy metal cations. The enzyme also catalyzes the stereospecific inversion of the C-2' hydroxyl of the nucleoside antibiotic, tubercidin to form 7-beta-D-arabinofuranosyl-4-aminopyrrolo[2,3-d]pyrimidine. The nucleoside antibiotic, sangivamycin, in which the C-5 hydrogen is replaced with a carboxamide group, is not a substrate. On the basis of the single and double label experiments in vivo and the in vitro enzyme-catalyzed experiments, two mechanisms involving either a 3'-ketonucleoside intermediate or a radical cation are proposed to explain the observed data.