Enhancement of retention and cytotoxicity of 2-chlorodeoxyadenosine in cultured human leukemic lymphoblasts by nitrobenzylthioinosine, an inhibitor of equilibrative nucleoside transport.

In leukemic cells exposed to 2-chlorodeoxyadenosine (2-CdA), levels of the nucleoside drug and its phosphate metabolites decay with time in the absence of external 2-CdA; an intrinsic part of this process is the efflux of 2-CdA. The effects of nitrobenzylthioinosine (NBMPR) and of dipyridamole (DPM), both potent inhibitors of es (e, equilibrative; s, sensitive to NBMPR) nucleoside transport processes, were studied in four lines of cultured leukemic lymphoblasts. Suspensions of 2-CdA-loaded cells were diluted 10-fold with 2-CdA-free medium to initiate the cellular 2-CdA decay processes, which followed a biexponential time course. When diluting media contained NBMPR or DPM, intracellular levels of 2-CdA and its metabolites were substantially increased (P < 0.001) compared with cells in media lacking the transport inhibitors, and 2-CdA loss followed a monoexponential time course. As a consequence, the AUCs (area under time-course plots of intracellular 2-CdA and its metabolites) were significantly (P < 0.001) lower in untreated control cells compared to inhibitor-treated cells. These results suggest that nucleoside transport processes contribute to the efflux of 2-CdA from the cultured lymphoblasts. The cytotoxicity of 1-h exposure to 2-CdA of Reh-A2 and CCRF-CEM cells was enhanced three-fold by subsequent exposure to 0.5 microM NBMPR relative to that of control cells subjected to the same manipulations without NBMPR exposure. However, before such a strategy may be considered to have a therapeutic application, careful examination of effects in normal lymphocytes and ex vivo leukemic lymphoblasts must first be undertaken. Leukemia (2000) 14, 52-60.

Intracellular pharmacokinetics of 2-chlorodeoxyadenosine in leukemia cells from patients with chronic lymphocytic leukemia.

2-Chlorodeoxyadenosine (2-CdA) is an important agent in the treatment of hairy cell leukemia and chronic lymphocytic leukemia (CLL). Others have reported that levels of 2-CdA phosphates present in human leukemia cells decline rapidly when the cells are in 2-CdA-free medium (Santana et al. J Clin Oncol 1991; 9: 416-422). In the present study, time-courses of 2-CdA loss from CLL cells were biexponential: the mean half-life of the initial phase was 0.30 +/- 0.18 h; the presence of 0.5 microM nitrobenzylthioinosine (NBMPR, a classical inhibitor of nucleoside transport) in the suspending medium, significantly decreased the initial rate of 2-CdA efflux (mean half-life, 0.43 +/- 0.22 h). As a consequence, AUCs (areas under time-course plots) were significantly higher in the NBMPR-treated cells (4.56 +/- 2.01 pmol.h/10(6) cells, n = 19) than in untreated control cells (3.83 +/- 1.74 pmol.h/10(6) cells; n = 19). 2-CdA was the principal efflux product released into the medium from 2-CdA-loaded CLL cells. We conclude that nucleoside transport processes contribute to the efflux of 2-CdA from CLL cells and that NBMPR may be useful as a retentive agent.

Fibrinogen Zurich I: impaired release of fibrinopeptide A.

Fibrinogen Zurich I is characterized by an abnormal fibrin monomer polymerization. It consists of two fractions of molecules, one with a normal aggregation and one not aggregating at all and interfering with the aggregation of the normal population. Using a radioimmunoassay for fibrinopeptide A, only approximately half of the expected fibrinopeptide A could be recovered after thrombin or Defibrase proteolysis. The defective fibrinopeptide A release could be confirmed by measurement of the N-terminal Gly/Tyr ratio. It is likely that the abnormal fibrin monomer aggregation of the abnormal fraction of fibrinogen Zurich I is due to the defective fibrinopeptide A release of this fraction.

Dephosphorylation of nitrobenzylthioinosine 5'-monophosphate by ecto 5'-nucleotidase of HeLa cells.

HeLa cells as well as human and mouse erythrocytes possess membrane sites which bind the inhibitor of nucleoside transport, nitrobenzylthioinosine (NBMPR), reversibly but tightly (KD, 10(-9)-10(-10) M). Site-specific binding of the ligand correlates with inhibition of nucleoside transport. The present study showed that the 5'-phosphate of NBMPR, NBMPR-P, was not transport inhibitory. Upon exposure to [35S]NBMPR-P or [G-3H]NBMPR-P, HeLa cells retained the isotopic labels virtually exclusively in the form of NBMPR. The dephosphorylation of [G-3H]NBMPR-P by HeLa cells, assayed by the production of extracellular [G-3H]NBMPR, was competitively inhibited by AMP, but was not affected by the presence of 5 microM NBMPR, a concentration sufficient to completely occupy the transport inhibitory sites. Thus, the sites at which dephosphorylation of NBMPR occurs in HeLa cells are separate from and function independently of the high affinity sites which bind NBMPR.

Structural modifications at the 2'- and 3'-positions of some pyrimidine nucleosides as determinants of their interaction with the mouse erythrocyte nucleoside transporter.

Modifications in the sugar moiety of pyrimidine nucleosides may affect their ability to function as permeants of the mouse erythrocyte nucleoside transporter. In this investigation, a number of synthetic uracil and thymine nucleosides which differ from the physiological nucleosides, uridine, deoxyuridine and thymidine, through structural changes at the 2'- and 3'-positions were studied. Interaction of the analogs with the transporter has been assessed in terms of their affinities for an external site on the transporter as well as their abilities to effect trans-acceleration of thymidine efflux. 1-(beta-D-Arabinofuranosyl) uracil (araU) and 1-(beta-D-arabinofuranosyl)thymine (araT) were comparable to thymidine as permeants while nucleosides in which the 3'-hydroxyl was replaced with hydrogen or a halogen had a decreased affinity for the transporter. 3'-Fluoro-3'-deoxy-araU weakly accelerated thymidine efflux while its ribo-isomer and the other 3'-halogeno-3' deoxy-arabino analogs as well as dideoxythymidine inhibited efflux. The absence of 2'- and 3'-carbons in acyclothymidine and acyclouridine strongly decreased the affinities of these nucleosides for the transporter; efflux of thymidine was not accelerated in the presence of these compounds. The conformationally constrained cyclic nucleoside 2,2'-anhydro-araU had a very low affinity for the transporter, and influx of the radiolabeled compound could not be demonstrated. The results suggest that modification at the 3'-position, loss of a portion of the sugar ring, and lack of conformational flexibility are factors which decrease the abilities of some pyrimidine nucleosides to function as permeants. It is suggested that combined effects of substituents which play a role in determining nucleoside conformation should be considered in assessing structural requirements for permeants of the transporter.

Measurement of nitrobenzylthioinosine in plasma and erythrocytes: a pharmacokinetic study in mice.

PURPOSE: Nitrobenzylthioinosine (NBMPR), a potent inhibitor of nucleoside transport in many cell types, modulates the in vivo disposition of several cytotoxic nucleoside analogs. In this study, a radioligand binding assay was developed for measurement of the NBMPR content of plasma and erythrocytes. METHODS: The assay was based on the competition between NBMPR and [3H]NBMPR for high-affinity sites on human erythrocytes membranes. With this assay, we followed in mice changes in the NBMPR content of blood plasma and erythrocytes, following the intraperitoneal injection of the disodium salt of NBMPR 5'-monophosphate (NBMPR-P), a prodrug form of NBMPR. RESULTS: The radioligand binding assay was able to measure precisely as little as 2.5 pmol of NBMPR, allowing the direct determination of NBMPR concentrations in plasma as low as 16 nM. As few as 8 x 10(3) molecules of NBMPR per cell could be determined in erythrocytes. The NBMPR content of plasma from mice injected with NBMPR-P was maximal at about 20 min after injection and declined to < 0.2% of the peak value by 10 h. Erythrocyte-associated NBMPR was also maximal at 20 min, and declined to 11% of the peak value by 10 h after injection. Time courses for the disappearance of NBMPR from plasma and erythrocytes were monoexponential and yielded half-life values of 0.39 h and 0.68 h, respectively, an apparent volume of distribution of 0.61 l/kg, and a clearance of 1.1 l/h per kg. CONCLUSIONS: The radioligand binding assay is a sensitive and facile method for monitoring NBMPR concentrations in mammalian plasma and tissue extracts.

Es nucleoside transporter content of acute leukemia cells: role in cell sensitivity to cytarabine (araC).

Nucleoside analogs are important components of treatment regimens for acute leukemia in adults. Plasma membrane permeation of the nucleoside analog molecules, the initial event in the cellular conversion of nucleosides to active agents, is mediated by nucleoside-specific membrane transporters. The widely-expressed es nucleoside transporter accepts as substrates diverse nucleoside analogs, including cytarabine (araC), 2-chlorodeoxyadenosine, and fludarabine. The cellular content of es transporter sites has been measured in blasts from patients with acute lymphoblastic leukemia and acute myelogenous leukemia, by a sensitive, quantitative flow cytometry assay that employs the tightly-bound es ligand, SAENTA fluorescein. Values for es transporter expression varied ten-fold among samples from patients with acute myelogenous leukemia. In this article, we review current findings that document, in confocal fluorescence microscopy images and in flow cytometry assays of SAENTA fluorescein-stained cells, the patient-to-patient variance of es transporter expression in leukemic blasts from patients. Our data show a correlation between the expression of es transporters and the in vitro sensitivity to nucleoside drugs of blasts from acute leukemia patients. These findings show that the flow cytometry assay of es expression provides a facile means of predicting resistance of leukemia cells to the cytotoxicity of araC and other nucleosides.

Interaction of [3H]dilazep at nucleoside transporter-associated binding sites on S49 mouse lymphoma cells.

Dilazep, a tertiary amine that is greater than 96% protonated at pH 7.4, is a potent inhibitor of facilitated diffusion (equilibrative) nucleoside transport (NT) in animal cells. In this study, saturable reversible binding of [3H]dilazep was demonstrated at sites on S49 mouse lymphoma cells but not in AE1 cells, an NT-deficient mutant of S49 cells. Mass law analysis of dilazep binding under equilibrium conditions revealed two saturable components, representing binding sites that differed about 50-fold in affinity for dilazep (Kd values of 0.21 and 10 nM). At pH 7.4, the low affinity sites were more abundant (Bmax, 3.5 X 10(5) sites/cell) than the high affinity site (Bmax, 3.0 X 10(4) sites/cell). Binding of dilazep was pH dependent; at pH 9.0, binding at the high affinity sites predominated, whereas, at pH 5.0, the low affinity component predominated, suggesting that these components represented binding of nonprotonated and protonated dilazep molecules, respectively. Nitrobenzylthioinosine (NBMPR) and physostigmine selectively blocked binding of nonprotonated and protonated species of dilazep, respectively, at pH 7.4, yielding Scatchard plots that were similar to control plots obtained at pH 5.0 and 9.0. First-order plots of the dissociation of [3H]dilazep-binding site complexes in the presence of excess nonradioactive dilazep at pH 7.4 were nonlinear and were resolved into rapid (rate constant, 3.4-4.7 min-1) and slow (rate constant, 0.13-0.15 min-1) components. In the presence of site-saturating concentrations of NBMPR or high concentrations of nucleoside permeants, dissociation of site-bound [3H]dilazep was incomplete and only the slow component of dissociation was apparent (rate constant, 0.11-0.19 min-1). The combined presence of nonradioactive dilazep and NBMPR yielded time courses of [3H]dilazep-site dissociation equivalent to those obtained in the presence of nonradioactive dilazep alone. These results are consistent with a model in which protonated and nonprotonated species of dilazep bind at separate sites on S49 cells. The absence of both high and low affinity sites on AE1 cells suggests that, in S49 cells, both populations of sites are associated with NT polypeptides. The high affinity sites that bind nonprotonated species of dilazep appear to overlap with NBMPR binding sites on these cells.

[Resolution of Bbeta and gamma chains from normal human fibrinogen Zürich II into 2 variants with each differing sialic acid content]. / Trennung der Bbeta- und gamma-Ketten von normalem humanem Fibrinogen und von Fibrinogen Zürich II in je 2 Varianten mit verschiedenem Sialsäuregehalt

The gamma- and Bbeta-polypeptide chains of purified human fibrinogen (both pooled and single donor) have each been resolved into 2 major components: gammaL and gammaR, and BbetaL and BbetaR. They are similar in molecular weight (SDS-PAGE), but differ in sialic acid content, which approximates 2 and 1 residues per molecular of polypeptide in the L- and R-components respectively. Tryptic peptide maps of the L-and R- forms of the gamma chain showed differences within the small group of peptides containing the sialic acid residues. No differences between the peptide maps of BbetaL- and BbetaR-chains were found . A larger ratio of L:R in the gamma- and Bbeta-chains of dysfibrinogenemia fibrinogen "Zürich II" explains the higher content of sialic acid measured in the unmodified Zürich II fibrinogen molecule.

Separation of both the Bbeta- and the gamma-polypeptide chains of human fibrinogen into two main types which differ in sialic acid content.

The gamma- and Bbeta-polypeptide chains of purified human fibrinogen have each been resolved into two major species: gammaL and gammaR and BbetaL and BbetaR. These molecular variants, separable on CM-cellulose, differ from each other in sialic acid content: approximately 2 residues of sialic acid per molecule of polypeptide chain for the L species to 1 residue of sialic acid per molecule for the R species. The two types of each polypeptide are demonstrable in preparations of fibrinogen from single donors as well as in pooled fibrinogen. The L and R forms of the gamma chains or the Bbeta chains do not differ in their electrophoretic mobility in sodium dodecyl sulfate-polyacrylamide gels, suggesting that they are similar in molecular weight. They are also indistinguishable in polyacrylamide gels in the presence of urea at pH 2.7. Maps of ninhydrin-positive tryptic peptides of the L and R forms of the gamma chain displayed differences within a small group of peptides which have been shown to contain the sialic acid residues present in the gamma-polypeptides. No differences between the peptide maps of BbetaL and BbetaR chains were obvious. A larger ratio of L/R in the gamma and Bbeta chains of dysfibrinogenemia fibrinogen "Zürich II" than in those of normal fibrinogen explains the higher content of sialic acid measured in the native Zürich II fibrinogen molecule.

Enantiomeric selectivity of adenosine transport systems in mouse erythrocytes and L1210 cells.

In mediating the entry of adenosine into mouse erythrocytes and mouse leukaemia L1210 cells, nucleoside transport systems were stereoselective, showing a marked preference for the D-enantiomer of adenosine (D-Ado). Inward zero-trans fluxes of the mirror-image isomer, L-adenosine (L-Ado), in those cells were slow relative to those of D-Ado. Contributing to L-Ado fluxes in both cell types were (i) a transporter-mediated process of high nitrobenzylthioinosine-sensitivity and (ii) simple diffusion.

[Fibrinogen Zürich I: defective fibrinopeptide-A release]. / Fibrinogen Zürich I: gestörte Fibrinopeptid-A-Abspaltung.

Fibrinogen Zurich I (FZI) is characterized by delayed fibrin monomer aggregation. It has been previously shown that the patient fibrinogen is composed of two populations of molecules, one aggregating normally and the other not aggregating at all but interfering with the aggregation of the normal population. By use of a radioimmunoassay for fibrinopeptide A (FPA), only approximately half of the expected FPA could be recovered after thrombin and Defibrase proteolysis. Even after exhaustive proteolysis with 2 thrombin units of Defibrase, no further release of FPA could be obtained. The defective FPA release was confirmed by measurement of the N-terminal Gly/Tyr ratio. It may be assumed that the abnormal aggregation of one fraction of FZI is due to the defective FPA release from this fraction.

Interaction of 2'-halogeno-2'-deoxyuridines with the human erythrocyte nucleoside transport mechanism.

The efflux of radioactive thymidine from human erythrocytes at 25 degrees was accelerated in the presence of extracellular 2'-fluoro-2'-deoxyuridine to a maximal velocity 120% of that observed in the presence of extracellular nonradioactive thymidine. Efflux in the presence of 2'-chloro-2'-deoxyuridine and 2'-bromo-2'-deoxyuridine did not exceed 56% and 49%, respectively. 2'-Iodo-2'-deoxyuridine did not accelerate thymidine efflux. In comparison, 2'-fluoro-2'-deoxycytidine and 2'-deoxyuridine accelerated thymidine efflux to maximal velocities of 170% and 91%, respectively. The half-saturation constant for acceleration of thymidine efflux by 2'-fluoro-2'-deoxycytidine was higher (0.90 mM) than those estimated for the other substances (0.22 mM or lower). Influx competition experiments at 25 degrees showed that all of the above nucleosides competitively inhibited influx of thymidine into human erythrocytes. The Km for the zero-trans influx of thymidine was 0.051 +/- 0.008 mM, while the Ki values for 2'-deoxyuridine and the 2'-halogeno-2'-deoxyuridines were similar, ranging from 0.04 to 0.09 mM. The Ki for 2'-fluoro-2'-deoxycytidine was 0.18 mM. These results suggest that, although all nucleosides tested appeared to bind to the same transport site on the external membrane surface, their ease of transport through the membrane was determined by the properties of the halogen substituent at position 2'.

Drug-induced perturbations in the in vivo distribution of oncological radiotracers--II. 5-[125I]iodo-2'-deoxyuridine influenced by nitrobenzylthioinosine-5'-phosphate (NBMPR-P) and acyclothymidine (ACT).

Nitrobenzylthioinosine (NBMPR), a potent inhibitor of facilitated nucleoside transport in vitro and in vivo, and acyclothymidine (ACT), a potent inhibitor of pyrimidine nucleoside phosphorylase in vitro, have been used in an attempt to modulate the biodistribution of 125I-labelled iododeoxyuridine ([125I]IUdR). ACT or NBMPR-P (a water-soluble prodrug of NBMPR) were injected into BDF1 mice bearing implanted Lewis lung tumors, according to protocols which would provide high and low plasma levels of the inhibitor. Compared with controls, both inhibitors induced transient, marginal increases in hepatic, renal and blood levels of [125I]IUdR, and decreased levels in tumors at short time intervals after injection. It is concluded that there is a mild tumor-sparing effect when either NBMPR or ACT are administered together with single i.v. diagnostic doses of [125I]IUdR.

Tumor uptake of radiolabelled pyrimidine bases and pyrimidine nucleosides in animal models--VIII. Synthesis and tissue distribution of N-[2-(hydroxyethoxy) methyl]-5-[3H]methyluracil.

The tritium-labelled acyclonucleoside, N-[2-(hydroxyethoxy)methyl]-5-[3H]methyluracil (3H-3), was synthesized for evaluation as a tumor diagnostic agent. 5-[3H]-Methyluracil, 3H-1, was converted to the 2,4-bis-trimethylsilyl intermediate which was coupled with 2-acetoxyethoxymethyl bromide to afford 1-[(2-acetoxyethoxy)methyl-5-[3H]methyluracil (3H-2). Treatment of 3H-2 with sodium methoxide in methanol afforded 3H-3 (specific activity 188 MBq mmol-1. The tissue distribution of 3H-3 was examined in male BDF1 mice bearing Lewis Lung (LL) carcinomas. Long bone exhibited the highest tumor: tissue ratios. The kidney contained the highest radioactivity level relative to the tumor. This suggested a major urinary route of excretion. The major radioactive blood component (89.21%) was found to have a biological half-life of 0.19 min. The title compound is unsuitable for use as a diagnostic agent for LL carcinoma because of low tumor uptake and rapid urinary elimination of injected radioactivity from the body.

The effect of pH on interaction of nitrobenzylthioinosine and hydroxynitrobenzylthioinosine with the nucleoside transporter of human erythrocyte membranes.

Site-specific binding to human erythrocyte membranes of nitrobenzylthioinosine (NBMPR), un-ionized at physiological pH, was compared with that of hydroxynitrobenzylthioinosine (HNBMPR), pKa 6.4, at graded pH values. Binding of [3H]NBMPR was measured directly, and that of HNBMPR was assayed by competitive inhibition by HNBMPR of [3H]NBMPR binding. Kd and Bmax values for binding of [3H]NBMPR to erythrocyte membranes were independent of pH. Kd values for the competing ligand were determined by mass law analysis of equilibrium binding data using either (a) apparent ligand concentration (dissociated plus undissociated forms of HNBMPR) or (b) the concentration of undissociated HNBMPR. Kd values for HNBMPR calculated with the apparent ligand concentration increased 10-fold as the fraction of HNBMPR molecules present in the dissociated form was increased (by pH changes) from 14 to 88%, whereas Kd values for the undissociated form of HNBMPR were independent of pH. The results presented here demonstrate that the undissociated form of HNBMPR binds more tightly to the transport-inhibitory sites of erythrocytes than NBMPR and suggest that ionization of S6-substituted thiopurine ribonucleosides eliminates or greatly decreases their ability to interact with the binding sites.

Nucleoside permeation in mouse erythrocytes infected with Plasmodium yoelii.

In normal mouse erythrocytes, nucleoside permeation was almost completely blocked in the presence of binding site-saturating concentrations of nitrobenzylthioinosine, whereas permeation in erythrocytes infected with the malarial parasite, Plasmodium yoelii, was substantial under these conditions, suggesting the presence of a permeation mechanism of low sensitivity to nitrobenzylthioinosine in the infected cells. Binding sites for nitrobenzylthioinosine were more numerous on infected erythrocytes than on uninfected cells. When mice infected with P. yoelii were treated with combinations of tubercidin and nitrobenzylthioinosine 5'-monophosphate, progression of parasitemia was delayed and survival times were increased.

[125I]iodohydroxynitrobenzylthioinosine: a new high-affinity nucleoside transporter probe.

[125I]iodohydroxynitrobenzylthioinosine ([125I]IH-NBMPR), a new gamma-labeled nucleoside transport inhibitor, has been prepared at a theoretical specific activity of 2000 Ci/mmol (1 Ci = 37 GBq). IH-NBMPR was more acidic than hydroxynitrobenzylthioinosine (H-NBMPR), having a pKa of 4.6. Site-specific binding of [125I]IH-NBMPR to membrane-enriched fractions (MEF) from S49 mouse lymphoma cells was pH dependent, increasing with the fraction of undissociated molecules present; it was maximal at pH 4.5 and negligible at pH 7.0. Scatchard analysis of specific binding to MEF from S49 cells under equilibrium conditions at pH 5.0 yielded a Kd of 15 nM (equivalent to 4.0 nM for the undissociated fraction of inhibitor molecules) and maximum number of binding sites (Bmax) of 4.9 pmol/mg protein. Specific binding of IH-NBMPR could not be demonstrated in MEF from AE1 cells, a nucleoside transport-deficient mutant of S49 cells. Influx of uridine into mouse erythrocytes at pH 5.0 in the presence of 5 microM IH-NBMPR (1.4 microM undissociated IH-NBMPR) was reduced to about 7% of the control value, indicating that this compound is an effective nucleoside transport inhibitor. Photoactivation of site-bound [125I]IH-NBMPR, following equilibration of the ligand with MEF from S49 cells at pH 5.0, resulted in specific covalent labeling of a polypeptide with a relative molecular mass of 52,000-63,000, identified on sodium dodecyl sulfate-polyacrylamide gels. These results indicate that the new, iodinated ligand is an inhibitor of nucleoside transport and that it binds specifically and with high affinity to nucleoside transporter polypeptides in mammalian cells.

Modulation of the metabolism of beta-L-(-)-2',3'-dideoxy-3'-thiacytidine by thymidine, fludarabine, and nitrobenzylthioinosine.

beta-L-(-)-2',3'-Dideoxy-3'-thiacytidine (3TC) is a cytosine nucleoside analog that potently inhibits the replication of human and duck hepatitis B viruses and human immunodeficiency virus through the activity of its 5'-triphosphate ester metabolite. The present study examined the intracellular decay of 3TC 5'-phosphates and tested strategies for modulating the cellular content of those nucleotides in primary cultures of duck hepatocytes and in human hepatoma 2.2.15 cells and CCRF-CEM T lymphoblasts. Inhibition by deoxycytidine of the 5'-phosphorylation of 3TC in duck hepatocytes confirmed that, as in mammalian cells, deoxycytidine kinase catalyzed 3TC activation. The 5'-mono, 5'-di-, and 5'-triphosphates of 3TC underwent monoexponential elimination from duck hepatocytes and 2.2.15 cells (half-lives, 3.6 to 8.0 h). Thymidine and fludarabine, which are agents that enhance the activity of deoxycytidine kinase, were tested in strategies for increasing the cellular content of 3TC 5'-phosphates. Coordinate treatment of cells with 3TC and thymidine (50 microM) increased the content of 3TC 5'-monophosphate in duck hepatocytes and the content of 3TC 5'-di- and 5'-triphosphates in 2.2.15 cells, but enhancement of 3TC 5'-phosphate levels in CCRF-CEM cells required a higher thymidine concentration (100 microM). Fludarabine (5 microM) did not affect the contents of 3TC 5'-di- and 5'-triphosphates in duck hepatocytes, but modestly increased the contents of those nucleotides in 2.2.15 cells and CCRF-CEM cells. Nitrobenzylthioinosine (NBMPR), an inhibitor of the es facilitated diffusion nucleoside transporter, reduced the level of entry of 3TC into 2.2.15 cells and abolished inward fluxes of thymidine, adenosine, and deoxycytidine. In 2.2.15 cells and CCRF-CEM cells, NBMPR reduced the formation of 3TC 5'-di- and 5'-triphosphates and reversed the thymidine- and fludarabine-induced increases in the formation of those nucleotides. NBMPR protected against the cytotoxicity of 3TC in CCRF-CEM cells, whereas thymidine potentiated that toxicity, apparently by enhancing the formation of 3TC 5'-triphosphate. Taken together, these results indicate that deoxycytidine kinase and the es nucleoside transporter are targets for manipulation of the metabolism and activity of 3TC.