(CANCER RESEARCH 49, 4983-4989, September 15. 1989]

Mechanism of Adenosine Triphosphate Catabolism Induced by Deoxyadenosine andby Nucleoside Analogues in Adenosine Deaminase-inhibited HumanErythrocytes1

FrançoiseBontemps2 and Georges Van den Berghe'

Laboratory of Physiological Chemistry, International Institute of Cellular and Molecular Pathology, ana University of Louvain, B-1200 Brussels, Belgium

ABSTRACT

The mechanism of the depletion of ATP, recorded in the erythrocytesof adenosine deaminase-deficient children and of leukemia patientstreated with deoxycoformycin, »asinvestigated in normal human erythrocytes treated with this inhibitor of adenosine deaminase. Deoxyadenosine, which accumulates in both clinical conditions, provoked a dose-dependent accumulation of dATP, depletion of ATP, and increases in theproduction of inosine plus hypoxanthine. Concomitantly, there was anincrease of AMP and IMP, but not of adenosine, indicating that catabo-lism proceeded by way of AMP deaminase. A series of nucleosideanalogues (9-/3-D-arabinofuranosyladenine, A*-methyladenosine, 6-meth-ylmercaptopurine ribonucleoside, tubercidin, ribavirin, and ¿V-I-ribosyl-5-aminoimidazole-4-carboxamide riboside) also stimulated adenine nu-cleotide catabolism and increased AMP and IMP to various extents. Theeffects of deoxyadenosine and of the nucleoside analogues were preventedby 5'-iodotubercidin, an inhibitor of adenosine kinase. Strikingly, they

were reversed if the inhibitor was added after the accumulation ofnucleotide analogues and initiation of adenine nucleotide catabolism.Further analyses revealed linear relationships between the rate of phos-phorylation of deoxyadenosine and nucleoside analogues and the increasein AMP and between the elevation of the latter above a thresholdconcentration of 10 UM and the rate of adenine nucleotide catabolism.Kinetic studies with purified erythrocytic AMP deaminase, at physiological concentrations of its effectors, showed that the enzyme is nearlyinactive up to 10 UMAMP and increases in activity above this threshold.We conclude that the main mechanism whereby deoxyadenosine andnucleoside analogues stimulate catabolism of adenine nucleotides by wayof AMP deaminase in erythrocytes is elevation of AMP, secondary tothe phosphorylation of the nucleosides.

INTRODUCTION

Accumulation of dATP has been documented in erythrocytes(1-6), in lymphocytes and bone marrow cells (7), and in platelets (8) of children with ADA4 deficiency. It is also found inRBC (9-11) and lymphoblasts (12, 13) of leukemia patientstreated with the ADA inhibitor, dCF. In addition, dramaticdepletions of ATP were recorded in the erythrocytes (9, 10) andlymphoblasts (13) of patients treated with dCF, whereas smallerdecreases were observed in the erythrocytes of some childrenwith ADA deficiency (3-6). Bagnara and Hershfield (14), investigating ADA-inhibited human lymphoblastoid cells, have

Received 2/27/89; revised 6/1/89; accepted 6/14/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1Supported by Grant 3.4539.87 of the Fund for Medical Scientific Research(Belgium) and by the Belgian State-Prime Minister's Office-Science Policy Pro

gramming.! To whom requests for reprints should be addressed, at Laboratory of Phys

iological Chemistry. UCL-ICP 75.39, Avenue Hippocrate, 75, B-1200 Brussels,Belgium.

1Director of Research of the Belgian National Fund for Scientific Research.4The abbreviations used are: ADA. adenosine deaminase (EC 3.5.4.4); dAdo.

2'-deoxyadenosine; AMP-DA, AMP deaminase (EC 3.5.4.6); KRB, Krebs-Ringerbicarbonate; dCF, 2'-deoxycoformycin; ITu, 5'-iodotubercidin; 2,3-BPG, 2,3-bisphosphoglycerate; ara-A, 9-fi-D-arabinofuranosyladenine; AICA riboside, ¿V-l-ribosyl-5-aminoimidazole-4-carboxamide: N'-MeAdo, A*-methyladenosine;MMPR, 6-methylmercaptopurine ribonucleoside: Tu, lubercidin (7-deazaadeno-sine); Rbv, ribavirin (l-/i-n-ribofuranosyl-l.2,4-triazole-3-carboxamide); HPLC,high-performance liquid chromatography.

explained this ATP depletion as follows: (a) dAdo, whichaccumulates when ADA is deficient or inhibited, is phospho-rylated to dAMP and subsequently to dATP, a process whichutilizes ATP and generates ADP and AMP; (¿>)the accumulation of dATP, which is greatly facilitated by the fact that dAMPis a poor substrate for AMP-DA (15, 16), stimulates both thedeamination of AMP to IMP and the dephosphorylation ofIMP to inosine (14, 17), thereby provoking adenine ribonucle-otide catabolism. This mechanism is, however, difficult to reconcile with two observations: (a) dATP and ATP are equipotentas stimulators of erythrocyte (15) and lymphoblast (14) AMP-DA; (b) the sum of dATP and ATP in ADA-deficient (4, 5) orin ADA-inhibited cells (9, 10, 13) is not or only slightly higherthan that of ATP in control cells. These data and uncertaintiesabout the concentration of AMP in ADA-deficient or -inhibitedcells prompted a reinvestigation of the mechanism of the depletion of ATP induced by dAdo. This study was performed withnormal human erythrocytes in which ADA was inhibited bydCF. The effect of dAdo has been compared to that of otherpurine nucleoside analogues which are also substrates of adenosine kinase. Our results provide evidence that all these nucleosides induce catabolism of the adenine nucleotides by a common mechanism, namely elevation of AMP. Part of this workhas been presented at a symposium (18).

MATERIALS AND METHODS

Chemicals. [t/-uC]Adenine (270 Ci/mol) and [8-MC]AMP (55 Ci/

mol) were purchased from the Radiochemical Centre (Amersham,Buckinghamshire, England). ITu was from RBI (Natick, MA) and dCFwas from Warner Lambert (Detroit, MI). Adenine nucleotides andadenosine were from Boehringer (Mannheim, Germany). Other nucleotides, nucleosides, and GTP-agarose were from Sigma (St. Louis, MO).The sources of all other chemicals have been given (19, 20).

Incubation of Erythrocytes. Fresh blood taken from a cubital vein ofhealthy human volunteers was collected on heparin. Isolation andwashing of erythrocytes were performed in KRB, pH 7.4, containing 5mM glucose and gassed with 95% O2-5% CO2 as described in Ref. 19.The RBC were resuspended in the same medium as a 20% hematocritand their adenine nucleotides were labeled by a 60-90-min prcincuba-tion at 37°Cwith 2-3 (¡M[U-'4C|adenine. This was followed by two

washes and resuspension as a 20% hematocrit in KRB with 5 mMglucose. Unless given otherwise, the concentration of P¡in the KRBbuffer was 1.2 HIM.In the experiments in which this concentration wasincreased to 10 mM. that of Ca-+ was reduced from 2.5 to 1.25 m\i.

Incubations were performed in carefully regassed and stoppered vials.In all experiments, suspensions were preincubated for 20 min with 1^M dCF before addition of nucleosides, in order to allow tight bindingof the inhibitor with ADA. We have shown previously that dCF has noeffect by itself on erythrocytic adenine nucleotide catabolism (19). Insome experiments, 10 »\iITu was used to inhibit adenosine kinase(21).

Assessment of Inhibitor Requirements in Intact Erythrocytes. Preliminary studies were performed in which the nucleotides and the catabo-lites produced from the added nucleosides were measured by HPLC, asdescribed below. These showed that the addition of IMMdCF and 10/aMITu inhibited by more than 90% the deamination of up to 0.5 mM

4983

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MECHANISM OF NUCLEOSIDE-1NDUCED ATP CATABOL1SM

adenosine, dAdo, and ara-A and prevented any detectable synthesis ofnucleotides from the latter nucleosides and from Tu, MMPR, N6-

MeAdo, Rbv, and AICA riboside.Assay and Partial Purification of AMP Deaminase. AMP-DA activity

was measured by the production of [S-'TJIMP from |8-'4C]AMP.Incubations were performed at 37°Cin a medium containing 50 HIM

A'-2-hydroxyethyl-piperazine-A'-2-ethanesulfonic acid buffer (pH 7.2),100 mM KCI, 5 mM MgCl2, [8-uC)AMP (0.05-0.1 ßCi/lest),and AMP

at the concentrations indicated, in a total volume of 50 or 100 n\. Atappropriate time intervals, 10 M'of the incubation medium were spottedon polyethyleneimine cellulose thin-layer chromatography plates onwhich carrier solutions (50 mimi) of AMP and IMP had been applied.After development in 1.4 M LiCl, activity was calculated from theradioactivity appearing in IMP. In crude preparations, AMP-DA activity was measured at 5 mM AMP in the absence of Mg2* to prevent thedephosphorylation of AMP or IMP by S'-nucleotidase(s). Insignificant

amounts of inosine, hypoxanthine, and adenosine (less than 5% ofAMP or IMP) were produced during the assay under these conditions.

AMP-DA was purified approximately 50,000-fold, with a yield of10-20%, by a two-step procedure. The hemolysate, prepared from 1unit of fresh blood collected on acid-citrate-dextrose, was washed freeof hemoglobin and subjected to DEAE-Trisacryl chromatography in20 mM Tris-HCI buffer, pH 7.2, containing 1 mM dithiothreitol (bufferA), as described in Ref. 20. Fractions containing AMP-DA activityeluted at about 100 mM NaCl from the DEAE-Trisacryl column. Theywere pooled and 5-ml portions were applied to a 0.4- x 6-cm affinitychromatography column containing 0.1 ml of GTP-agarose, equilibrated with buffer A. After a rinsing with 5 x 1 ml of buffer A,containing 0.5 M NaCl, AMP-DA was eluted by 10 mM 2,3-BPG. Theeluate, collected in albumin (1 mg/ml) to stabilize the enzyme, wasfiltered on Sephadex G-25 (fine) to remove 2,3-BPG which is a stronginhibitor of AMP-DA (22). Specific activity, measured at 5 mM AMP.was above 200 nmol/min/mg of protein. The enzyme was free ofmyokinase, 5'-nucleotidase, and adenosine kinase activities.

Analytical Methods. The extraction of metabolites was performed asdescribed in Ref. 19. All extracts were neutralized within 30 min toavoid artifactual increases in AMP, which represents only 0.5-1% ofATP under control conditions. ATP, dATP, ADP, AMP, IMP, andthe nucleotides produced from nucleoside analogues were separated byHPLC on a 12.5-cm Partisphere 5 SAX column by the method ofHartwick and Brown (23). Detection was performed at 254 nm, exceptfor the Rbv nucleotides which were analyzed at 205 nm. All thenucleotides formed from nucleoside analogues were clearly separated,except those generated from N6-MeAdo, which slightly overlapped with

the adenine nucleotides. Concentrations of the nucleotides were calculated by comparison with nucleotide standards when available, or whennot, as in the case of MMPR, by comparison with the absorbance ofthe nucleoside. The radioactivity in the adenine nucleotides and inadenosine, inosine, and hypoxanthine was determined by thin-layerchromatography as given previously (19). The concentration of thenucleosides and base was calculated from the specific radioactivity ofthe adenine nucleotides, which remained constant over the duration ofthe experiments.

For determination of the nucleosides added to the erythrocyte suspensions, these were briefly centrifuged, and the supernatant mediumwas deproteinized by heating for 1 min at 100°C.This extraction

method avoided acid hydrolysis of dAdo to adenine. The concentrationsof nucleosides were measured by reversed phase HPLC on a 3.9-mm x30-cm Ci8-¿iBondapak column (Waters-Millipore, Bedford, MA),eluted at a flow rate of 1 ml/min with 10 mM NaH2PO4, pH 5.5, anda 0-15% gradient of methanol over 15 min. MMPR was eluted iso-cratically with 30% methanol. The methods used for the determinationof other metabolites have been given (19).

RESULTS

Effects of Deoxyadenosine on the Concentrations of dATP andATP and on the Production of Purine Catabolites. Incubation ofADA-inhibited erythrocytes with dAdo resulted in a dose-dependent accumulation of dATP (Fig. 1, left); dAdo, even at the

lowest concentration, was not completely utilized over 3 h;dATP represented 80 to 90% of the total amount of deoxyri-bonucleotides synthesized (not illustrated). Concomitantly,there was a decrease of ATP. At all concentrations of dAdo,the sum of both triphosphates remained approximately equalor became slightly higher than the initial concentration of ATP.The addition of dAdo also induced an accumulation of theterminal purine catabolites, hypoxanthine and inosine (Fig. 1,rifflit). the latter increasing only at high concentrations of dAdo(see Fig. 2). The very small accumulation of purine catabolites,recorded in the control condition, was about doubled uponaddition of 20 /¿MdAdo, the lowest concentration used, and30- to 50-fold increased in the presence of 0.5 mivi dAdo. Thepurine catabolites nearly completely accounted for the loss ofATP. All these results are qualitatively similar to those recordedby Bagnara and Hershfield (14) in ADA-inhibited lymphoblas-toid cells.

Fig. 2 depicts the time course of the effect of various concentrations of dAdo. Concurrently with the depletion of ATP,dose-dependent increases of ADP, AMP, and IMP were recorded. Despite inhibition of ADA, no adenosine could bedetected. Together with the increase in IMP, this indicates thatthe accumulation of inosine and/or hypoxanthine results fromdeamination rather than from dephosphorylation of AMP.

Influence of Inhibition of the Phosphorylation of dAdo. Similarly to that of adenosine, the phosphorylation of dAdo can beinhibited by the addition of ITu, a potent inhibitor of adenosinekinase (21). When 10 ^M ITu was added at the beginning ofthe incubation, together with dAdo (Fig. 3, ¡eft),the accumulation of dATP as well as the depletion of ATP and theelevations of AMP and IMP were prevented, while the production of inosine plus hypoxanthine was inhibited by more than90%. These results accord with the observation that dAdo didnot stimulate adenine nucleotide catabolism in an adenosinekinase- and deoxycytidine kinase-deficient cell line (14). Likewise, addition of ITu 90 min after dAdo, halfway through theincubation (Fig. 3, right), completely arrested the build-up ofdATP. Strikingly, however, the addition of ITu also broughtabout a complete arrest of the degradation of ATP. This wasaccompanied by a return of the concentration of AMP to nearlyits basal value, by a decrease in IMP and, after a 30-min latency,

0.2 0.5 0 0.2

[ dAdo ] (mM)

0.5

Fig. 1. Effect of the concentration of dAdo on the synthesis of dATP and onthe catabolism of ATP. Various concentrations of dAdo were added to erythrocytes that had been preincubated successively with ['4C]adenine, in order to label

their adenine nucleotides, and with dCF. in order to inhibit adenosine deaminase,as described under "Materials and Methods.** Metabolites were measured after 3

h of incubation. ATP and dATP were determined by HPLC. whereas inosine(INO) plus hypoxanthine (HX) were determined radiochemically and their concentrations were calculated from the specific radioactivity of the adenine nucleotides. Results shown are means of two experiments.

4984

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MFCHAMSM OF MTl.EOSIDE INDITED ATP CATABO1 ISM

12301 23

Time (h) after addition of dAdo

Fig. 2. Time course of the effect of various concentrations of dAdo on theconcentration of ATP and of its catabolites. Erythrocytes were preincubated asin Fig. 1 and incubated thereafter in the absence or in the presence of dAdo at0.1, 0.2, or 0.5 HIMconcentration. Results shown are means of two experiments.INO, inosine; HX, hypoxanthine.

can be phosphorylated intracellularly. As shown in Fig. 4,adenosine and a series of nucleoside analogues, added at 0.5HIMconcentration, all stimulated, although to various extents,the catabolism of the adenine ribonucleotides into inosine andhypoxanthine. Maximal stimulation was obtained by Tu, followed by, in order, N6-MeAdo, dAdo, MMPR, AICA riboside(not illustrated), Rbv, adenosine, and ara-A. Owing to thepresence of dCF, it could be concluded that, as with dAdo, theinosine and hypoxanthine produced from the prelabeled adenine nucleotides resulted from the deamination of AMP intoIMP rather than from AMP dephosphorylation. In the presenceof Tu and of adenosine, some dephosphorylation of AMPoccurred, as evidenced by the appearance of labeled adenosine.

To verify the similarity of the mechanism of action of thenucleoside analogues with that of dAdo, further experimentswere performed with ITu. Results obtained with two of them,MMPR (Fig. 5, left), which is converted for 90% into a mon-ophosphate, and Tu (Fig. 5, right), which is mainly convertedinto a triphosphate, are shown. Similarly to that induced bydAdo, the catabolism of the adenine nucleotides provoked bythese nucleosides was arrested by the subsequent addition ofITu. This arrest was associated with an interruption of thesynthesis of MMPR monophosphate, which reached the concentration of about 1.2 /¿mol/mlof packed cells after 90 min,and of Tu triphosphate which reached about 0.8 ^mol/ml ofpacked cells after 60 min (not illustrated). As observed withdAdo, addition of ITu also provoked a decrease of AMP andIMP, as well as of the production of inosine and hypoxanthine.Similar results were obtained with adenosine, Rbv, N6-MeAdo,

and AICA riboside (not illustrated). These experiments indicatethat all the nucleosides provoke an increase in the activity ofAMP-DA by a common mechanism, which is independent ofthe level of their phosphorylation.

Influence of Supraphysiological Concentrations of 1',. Because

Parks and Brown (24) had not observed a depletion of ATPafter addition of various nucleosides to human erythrocytesincubated in 30 mivi P¡,the influence of an elevation of theconcentration of P¡was investigated. When P¡in the incubationmedium was increased from its physiological level of 1.2, to 10HIM.the effects on adenine nucleotide catabolism of adenosine,dAdo, ara-A, and MMPR, all at 0.5 HIMconcentration, werecompletely suppressed (results not shown). The effect of 0.5

01 2301 23

Time (h) after addition of 0.5 mM dAdo

Fig. 3. Effect of ITu on the synthesis of dATP and on the catabolism of theadenine nucleotides induced by dAdo. Erythrocytes were preincubated as in Fig.1. Incubations were performed in the presence of 0.5 mM dAdo alone (O, D) ortogether with 10 /AI ITu (•,•¿�).which was added to the cell suspension, eitherinitially, simultaneously with 0.5 mM dAdo (left), or 90 min later (right). Resultsshown are means of duplicates. INO, inosine; HX, hypoxanthine.

by a marked decrease of the production of inosine plus hypoxanthine. These observations indicate that accumulation ofdATP is not sufficient by itself to explain the catabolism of theadenine nucleotides. In addition, they suggest that the elevationof AMP may play a determining role in the induction of thiscatabolism.

Effect of Nucleoside Analogues. The results depicted in Fig.3 led to the hypothesis that the ability of dAdo to induce adeninenucleotide catabolism may be shared by other nucleosides that

4985

0123

Time (h) after addition

Fig. 4. Effect of various nucleosides on the production of inosine (INO) plushypoxanthine (HX) from prelabeled adenine nucleotides. Erythrocytes were preincubated as in Fig. 1. Incubations were started with the addition of the nucleosidesat 0.5 mM concentration. Inosine plus hypoxanthine were determined radio-chemically and their concentrations were calculated from the specific radioactivityof the adenine nucleotides. Ado, adenosine; Co, control.

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MECHANISM OF NUCLEOS1DE-1NDUCED ATP CATABOLISM

012301 2

Time (h) after addition of nucleosides

Fig. 5. Effect of ITu on the catabolism of adenine nucleotides induced byMMPR and by Tu. Erythrocytes were preincubated as in Fig. 1. Incubations werestarted with the addition of 0.5 ITIMMMPR (left) or 0.5 mivi Tu (right). O, D,values measured in the absence of ITu; •¿�,•¿�values measured after the additionof 10/IM ITu at the times indicated. Results are means of duplicates.//VO, inosine;HX, hypoxanthine.

IHMTu was strongly delayed, and at 0.1 HIMit was completelyabolished. The accumulation of the nucleotide derivatives ofthe nucleosides mentioned was slightly increased in the presenceof 10 IHMP¡(not shown).

Effects of dAdo and Other Nucleosides on the Concentrationof the Effectors of AMP Deaminase in Erythrocytes. HumanRBC AMP-DA is known to be stimulated by ATP and alkalimetal cations and inhibited by P¡and 2,3-BPG (15, 16, 22, 25).Both inhibitors were therefore measured in ADA-inhibitederythrocytes in order to assess if a decrease of their concentration could explain the increased activity of AMP-DA inducedby the addition of various nucleosides. The concentration of2,3-BPG, which was about 4-5 ^mol/ml of packed cells incontrol conditions, was not modified by the addition of thenucleosides. However, the concentration of intracellular P¡,depending on the added nucleoside, decreased by 30 to 60%(results not shown). Nevertheless, when in experiments similarto those depicted in Figs. 3 and 5, the ATP degradation inducedby dAdo (Fig. 6, left) or by Tu (Fig. 6, right) was arrested bythe addition of ITu, intracellular P¡reincreased only by 20%.The reason for the decrease in P¡is not clear. It does not seemlinked exclusively to the phosphorylation of the nucleosidessince it was also observed, although to a smaller extent, whenthe synthesis of dATP was prevented by the addition of ITu atthe beginning of the incubation (not illustrated).

Correlation between the Concentration of AMP and the Rateof Catabolism of the Adenine Ribonucleotides. With all nucleosides studied, the concentration of AMP increased when theirphosphorylation was allowed to proceed and adenine nucleotidecatabolism was induced, and AMP concentration decreasedwhen both processes were switched off by ITu. The correlationbetween the concentrations of AMP and the rates of productionof inosine plus hypoxanthine, recorded under control conditions and with various nucleosides, was therefore investigated.As shown in Fig. 7, below a threshold concentration of AMPof 2 nmol/ml of suspension (corresponding to 10 nmol/ml of

1.0

0.8

So 0.6

to —¿�u on»_

i~0.4

I 0.2a.

-ITu

0123 0123

Time (h) after addition of nucleosides

Fig. 6. Influence of dAdo. Tu, and ITu on the concentration of intracellularPÃŒ.Erythrocytes were preincubated as in Fig. 1. Incubations were started with theaddition of 0.5 mM dAdo (left), or 0.5 mM Tu (righi). O, D, values measured inthe absence of ITu; •¿�,•¿�values measured after the addition of 10 /IM ITu at thetimes indicated. Results are means of duplicates.

Control0.02-0.5mM dAdo

O.SmM AdoO.SmM Ara-A

O.SmM MMPRO.SmM Tu

12

(nmol/ml of suspension)

Fig. 7. Correlation between the concentration of AMP and the rate of production of inosine (INO) plus hypoxanthine (HX) from prelabeled adeninenucleotides. AMP concentrations are means of three measurements, performedafter 30, 60, and 120 min of incubation, in the absence (O) or in the presence ofnucleosides as indicated. Productions of inosine plus hypoxanthine were calculated over the 120-min duration of the incubations. Results are from 6 separateexperiments, r = 0.95; P < 0.001. Ado, adenosine.

packed cells), there was almost no production of inosine andhypoxanthine. Beyond this value, a steep linear relationshipbetween the concentration of AMP and the rate of productionof inosine plus hypoxanthine was observed; a 2-fold increase inAMP enhanced the production of these catabolites by a factorof 10. The AMP threshold was not due to extracellular AMP,since its concentration in the medium was below 0.4 nmol/ml,and the slight hemolysis recorded under control conditions wasnot increased by the addition of the various nucleosides.

Fig. 8 shows that the concentrations of AMP obtained withthe various nucleosides were closely related to their rates ofphosphorylation, with the exception of MMPR. This may belinked to the fact that MMPR is converted to a monophosphatenucleoside, in contrast with the other nucleosides which aremetabolized to triphosphates.

Kinetic Properties of Erythrocyte AMP Deaminase at Physiological Concentrations of AMP. The kinetic properties ofpurified erythrocytic AMP-DA, including its stimulation byATP and K+, and inhibition by 2,3-BPG and P¡,have been

studied extensively (15,16,22,25). However, to our knowledge,4986

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MECHANISM OF NUCLEOSIDE-INDUCED ATP CATABOLISM

Q-

<2 -

Control•¿�0.02-0.5mM dAdo* O.SmM Ado•¿�O.SmM Ara-A* O.SmM TuT O.SmM MMPR

0 20 40 60 80 100 120 140

Rate ol nucleotide synthesis(nmol/h x ml of suspension)

Fig. 8. Correlation between the rate of synthesis of nucleotide analogues andthe concentration of AMP. AMP concentrations are means of three measurementsas in Fig. 7. Rates of synthesis of nucleotides from the added nucleosides werecalculated from the concentrations of the nucleotides measured at 30, 60. and120 min of incubation, r (excluding the MMPR data) = 0.96; P < 0.001. Ado,adenosine.

150

100

IQ.

< 50

500

1mM ATP3mM 2,3-BPG

1mM Pi

L

0.02 0.05

AMP ] (mM)

0.1

Fig. 9. Influence of effectors on the substrate saturation curve of AMPdeaminase. The activity of the enzyme was measured in the absence of additions(O) or in the presence of 1 mM ATP (•)and of 1 mM ATP, 3 mM 2,3-BPG, and1 mM I', (•).

the kinetics of the enzyme had not been investigated at themicromolar concentrations of AMP which prevail in intactRBC. The influence of both the increase in AMP and thedecrease in intracellular P¡,recorded during the metabolism ofnucleosides, was therefore assessed. In the presence of 100 HIMKC1, but in the absence of other effectors AMP-DA displayedhyperbolic kinetics with a S0.5 for AMP of 0.4 HIM (Fig. 9).Addition of 1 mM ATP, the concentration of the stimulator incontrol erythrocytes, decreased the S0.s for AMP to 0.15 mM.Further addition of the inhibitors at their physiological concentrations, namely 1 mM for Pi and 3 mM for unbound 2,3-BPG,rendered the substrate saturation curve sigmoid and the enzymenearly inactive up to 10 UM AMP. At 50 ¿IMAMP and in thepresence of 1 HIM ATP and 3 HIM 2,3-BPG, P¡inhibited theenzyme activity by 40% at 1 HIM and by 95% at 10 HIMconcentration (Fig. 10, left). A decrease in P¡from 1 HIMto 0.5mM, as recorded in ADA-inhibited erythrocytes incubated with

= 30E3E

a 20

Q.

< 10

60

40

20

I PI ] (mM)

10 0 2 5 10

[ ATP ] or I dATP ] (mM)

Fig. 10. Influence of P¡,ATP, and dATP on AMP deaminase activity. AMP-DA activity was measured at 50 ¿IMAMP. The effect of P, (left) was measured inthe presence of 1 mM ATP and 3 mM 2,3-BPG; the effects of ATP (•)and dATP(A) (right) were measured in the presence of 1 mM P¡and 3 mM 2,3-BPG.

0.5 mM dAdo or Tu, would thus increase the activity of AMP-DA by no more than 20%. The influence of ATP and dATP onthe activity of the enzyme in the presence of physiologicalconcentrations of its other effectors is shown in Fig. 10 (right).As described by others (14, 15), dATP and ATP were equallyefficient as stimulators, the half-maximal effect being obtainedat approximately 1 mM in the presence of 1 mM P¡and 3 mM2,3-BPG. Under the same conditions, ara-ATP was a 2-foldless potent stimulator than ATP and dATP (not shown). Theeffect of other triphosphate nucleosides was not tested becausethey were not available. It was also verified that dAMP, whichis not a substrate of AMP-DA (15, 16), was not a positiveeffector of the enzyme at low concentrations of AMP. Tumonophosphate and ara-AMP were also without effect on theactivity of AMP-DA.

DISCUSSION

Several mechanisms have been proposed to explain the lossof ATP accompanying accumulation of dATP in erythrocytesof leukemia patients treated with dCF and of children withADA deficiency. Simmonds et al. (4) have claimed that eryth-rocytic ATP is dependent, at least partly, on adenosine produced by the transmethylation pathway and that this productioncould be impaired in ADA-deficient or -inhibited cells, becauseof inactivation of S-adenosylhomocysteine hydrolase by dAdo(Ref. 6, and references therein). This mechanism is, however,difficult to reconcile with the finding of increased concentrations of adenosine in the plasma and urine of ADA-deficientpatients (1,7). Recently, Snyder et al. (26) have proposed thatadenosine causes a substrate inhibition of adenosine kinaseand/or a decreased phosphoribosylation of adenine. On theother hand, the study of Bagnara and Hershfield (14) in ADA-inhibited lymphoblasts and our study in ADA-inhibited erythrocytes show that dAdo increases the rate of degradation ofATP and of total adenine nucleotides. The effect is already seenat 20 MMdAdo, a concentration which is in the range of theconcentrations found in leukemia patients treated with dCF(11, 12). Both studies also show that the stimulation of thedegradation of adenine nucleotides by dAdo results from anincreased activity of AMP-DA. Indeed, catabolism of AMP viadephosphorylation to adenosine can be ruled out because, de-

4987

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MECHANISM OF NUCLEOS1DE-1NDUCED ATP CATABOLISM

spite inhibition of ADA, adenosine does not accumulate, andinosine and hypoxanthine are produced. The accumulation ofinosine, recorded concomitantly with the high rates of build-upof hypoxanthine, indicates that the activity of purine nucleosidephosphorylase becomes limiting under these conditions, mostlikely owing to the low intracellular concentration of P¡.

We show in this work that the main factor responsible forthe increased activity of AMP-DA in erythrocytes incubatedwith dAdo is neither the accumulation of one of its stimulators,dATP, nor a decrease of its inhibitors, P¡and 2,3-BPG, but anincrease in AMP brought about by the utilization of ATP inthe phosphorylation of dAdo. That the elevation of AMP playsa major role in this catabolism is indicated by the observationthat the arrest of the synthesis of dATP by ITu (Fig. 3) reversesthe elevation of AMP and stops the catabolism of ATP, despitethe presence of 0.5 mM dATP in the cells. ITu, an inhibitor ofadenosine kinase, has no effect by itself on AMP-DA (19, 27).The role of the elevation of AMP is validated by our resultsobtained with other nucleosides, which are also substrates ofadenosine kinase (28, 29). Similarly to dAdo, the nucleosideanalogues induced an elevation of AMP and an acceleration ofthe catabolism of the adenine ribonucleotides, which were bothcounteracted by the subsequent inhibition of adenosine kinaseby ITu.

The relation between the intracellular concentration of AMPand the rate of production of purine catabolites (Fig. 7) showsthat erythrocytes are very sensitive to an elevation of AMPabove a threshold concentration of 5-10 fiM. One hypothesisto explain this threshold, is that part of AMP, similarly toother erythrocytic metabolites (30), is bound to hemoglobinand thus not accessible to AMP-DA. We, however, favor theexplanation that AMP-DA is not active at 5-10 MMAMP,owing to the kinetic characteristics of its substrate saturationcurve in the presence of physiological concentrations of effectors (Fig. 9). The kinetic data given in Fig. 10 indicate thatboth the decrease of intracellular PEfrom 1 to 0.5 mM and thesubstitution of ATP by dATP are not the triggering factors ofthe catabolism of the adenine nucleotides induced by dAdo.The reason why nucleosides induce a decrease of intracellularP¡(Fig. 6) is not clear, since it occurs even when the synthesisof nucleotides is prevented by a prior addition of ITu.5 The factthat AMP-DA is strongly inhibited at supraphysiological concentrations of P¡(Fig. 10) most likely explains why the degradation of the adenine nucleotides induced by nucleosides canbe prevented by incubation in a high-P¡medium. Indeed, at 10mM extracellular P¡intracellular P¡increases from 1 to 4-5 mM(19). This probably also explains why Parks and Brown (24),who incubated erythrocytes in a high-P¡medium, did not observe a depletion of ATP with various nucleoside analogues.Why Henderson et al. (31 ) recorded nucleoside-induced adeninenucleotide catabolism in the presence of 25 mM P¡is, however,not clear.

The increase of the concentration of AMP provoked bynucleosides is clearly related to the rate of their phosphorylationor, more precisely, to the amount of high-energy phosphaterequired for the synthesis of analogue nucleotides (Fig. 8). Thisis evidenced by the lower increase of AMP induced by MMPR(which becomes a monophosphate nucleotide) than by Tu(which becomes a triphosphonucleotide), although both nucleosides are phosphorylated at the same rate. These results alsoindicate that the rate of glycolysis in erythrocytes is not quitesufficient to provide high-energy phosphates upon addition of

* Unpublished results.

nucleosides, although we have observed a small (10-20%) stimulatory effect of their addition on the production of láclate.5

The rates of accumulation of nucleotides, shown in Fig. 8,although potentially also influenced by their concomitant degradation which may vary from one analogue to another, reflectapproximately the affinities and substrate efficiencies of therespective nucleosides toward adenosine kinase from rabbit liver(29). Adenosine is, however, an exception since, although it isthe best substrate of adenosine kinase, it induced only a limitedelevation of adenine nucleotides, including AMP. This is mostlikely due to the known inhibition of adenosine kinase by excessof substrate (26, 29).

The relationships between the rate of phosphorylation ofnucleosides and the elevation of AMP and between the latterelevation and the rate of AMP catabolism most likely explainwhy the depletion of ATP is usually smaller in erythrocytes ofADA-deficient children (3-6) than in patients treated with dCF(9-11). Indeed, in the latter, higher concentrations of dAdo inthe plasma have been recorded (11, 12).

Taken together, our results suggest that catabolism of theadenine nucleotides of human erythrocytes can be induced byall nucleosides, and perhaps by other compounds, that arephosphorylated at a sufficient rate. The degradation is causedby an elevation in erythrocytic AMP and proceeds by way ofAMP-DA. This catabolism may be responsible for the anemiarecorded not only in ADA deficiency (6) and dCF treatment (9,32) but also with other nucleoside analogues used in anticancerand in antiviral therapy (33). Whether a similar nucleoside-induced catabolism occurs in other tissues remains to be established. Its existence may depend on various factors, amongwhich are the kinetic characteristics of the AMP-DA isozymeof the tissue. For liver AMP-DA for instance, the concentrations of the effectors of the enzyme seem to play a moreimportant role than that of its substrate (27, 34).

ACKNOWLEDGMENTS

We thank A. Delacauw for excellent technical assistance.

REFERENCES

1. Cohen, A., Hirschhorn, R., Horowitz, S. D., Rubinstein. A., Polmar, S. H.,Hong, R., and Martin, D. W., Jr. Deoxyadenosine triphosphate as a potentially toxic metabolite in adenosine deaminase deficiency. Proc. Nati. Acad.Sci. USA, 75: 472-476, 1978.

2. Coleman, M. S., Donofrio, J., Mutton, J. J., Hahn, L., Daoud, A., Lampkin,H.. and Dyminski, J. Identification and quantitation of adenine deoxynucleo-tides in erythrocytes of a patient with adenosine deaminase deficiency andsevere combined immunodeficiency. J. Biol. Chem., 253: 1619-1626, 1978.

3. Simmonds, H. A., Sahota, A., Potter, C. F., Perrett, D., Hugh-Jones, K.,and Watson, J. G. Purine metabolism in adenosine deaminase deficiency.CIBA Found. Symp., 68: 255-262, 1979.

4. Simmonds, H. A., Levinsky, R. J., Perrett, D., and Webster, D. R. Reciprocalrelationship between erythrocyte ATP and deoxy-ATP levels in inheritedADA deficiency. Biochem. Pharmacol., 31: 947-951, 1982.

5. Simmonds, H. A., Webster, D. R., Perrett, D., Reiter, S., and Levinsky, R.J. Formation and degradation of deoxyadenosine nucleotides in inheritedadenosine deaminase deficiency. Biosci. Rep., 2: 303-314, 1982.

6. Hershfield, M. S., Kurtzberg, J., Aiyar, V. N., Suh, E. J., and Schiff, R.Abnormalities in S-adenosylhomocysteine hydrolysis, ATP catabolism, andlymphoid differentiation in adenosine deaminase deficiency. Ann. NY Acad.Sci., 451: 78-86, 1985.

7. Donofrio, J., Coleman, M. S., Hutton, J. J., Daoud, A., Lampkin, B., andDyminski, J. Overproduction of adenine deoxynucleosides and deoxynucleo-tides in adenosine deaminase deficiency with severe combined immunodeficiency disease. J. Clin. Invest., 62: 884-887, 1978.

8. Simmonds, H. A., Goday, A., Morris, G. S., Fairbanks, L. D., and Levinsky,R. J. dATP accumulation and APT depletion in platelets in adenosinedeaminase deficiency: significance for the immune response? Biosci. Rep., 4:809-818, 1984.

9. Siaw, M. F. E., Mitchell, B. S., Koller, C. A., Coleman, M. S., and Hutton,J. J. ATP depletion as a consequence of adenosine deaminase inhibition inman. Proc. Nati. Acad. Sci. USA, 77: 6157-6161. 1980.

4988

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

MECHANISM OF NUCLEOSIDE-INDUCED ATP CATABOLISM

10. Grever, M. R., Siaw, M. F. E., Jacob, W. F., Neidhart, J. A., Miser, J. S.,Coleman, M. S., Hutton, J. J., and Balcerzak, S. P. The biochemical andclinical consequences of 2'-deoxycoformycin in refractory lymphoprolifera-tive malignancy. Blood, 57: 406-417, 1981.

11. Koller, C. A., and Mitchell, B. S. Alterations in erythrocyte adenine nucleo-tide pools resulting from 2'-deoxycoformycin therapy. Cancer Res., 43:1409-1414, 1983.

12. Mitchell, B. S-, Koller, C. A., and Heyn, R. Inhibition of adenosine deaminaseactivity results in cytotoxicity to T lymphoblasts in vivo. Blood, 56:556-559,1980.

13. Yu, A. L., Bakay, B., Kung, F. H., and Nyhan, W. L. Effects of 2'-

deoxycoformycin on the metabolism of purines and the survival of malignantcells in a patient with T-cell leukemia. Cancer Res., 41: 2677-2682, 1981.

14. Bagnara, A. S., and Hershfield, M. S. Mechanism of deoxyadenosine-inducedcatabolism of adenine ribonucleotides in adenosine deaminase-inhibited human T lymphoblastoid cells. Proc. Nati. Acad. Sci. USA, 79: 2673-2677,1982.

15. Lian, C.-Y., and Harkness, D. R. The kinetic properties of adenylate deaminase from human erythrocytes. Biochim. Biophys. Acta, 341: 27-40, 1974.

16. Yun, S.-L., and Suelter, C. H. Human erythrocyte 5'-AMP aminohydrolase.Purification and characterization. J. Biol. Chem. 253: 404-408, 1978.

17. Carson, D. A., and Wasson, D. B. Characterization of an adenosine 5'-triphosphate- and deoxyadenosine 5'-triphosphate-activated nucleotidasefrom human malignant lymphocytes. Cancer Res., 42:4321-4324, 1982.

18. Bontemps, F., and Van den Berghe, G. Mechanism of ATP catabolisminduced by deoxyadenosine and other nucleosides in adenosine deaminase-inhibited human erythrocytes. Pediatr. Res., 24: 113, 1988.

19. Bontemps, F., Van den Berghe, G., and Hers, H. G. Pathways of adeninenucleotide catabolism in erythrocytes. J. Clin. Invest., 77: 824-830, 1986.

20. Bontemps, F., Van den Berghe, G., and Hers, H. G. 5'-Nucleotidase activitiesin human erythrocytes. Identification of a purine 5'-nucleotidase stimulatedby ATP and glycerate 2,3-bisphosphate. Biochem. J., 250:687-696, 1988.

21. Henderson, J. F., Paterson, A. R. P., Caldwell, I. C., Paul, B., Chan, M. C.,and Lau, K. F. Inhibitors of nucleoside and nucleotide metabolism. CancerChemother. Rep. Part 2, 3: 71-85, 1972.

22. Askari, A., and Rao, S. N. Regulation of AMP deaminase by 2,3-diphos-phoglyceric acid: a possible mechanism for the control of adenine nucleotidemetabolism in human erythrocytes. Biochim. Biophys. Acta, 151: 198-203,1968.

23. Hartwick, R. A., and Brown, P. R. The performance of microparticle chemically-bonded anion-exchange resins in the analysis of nucleotides. J. Chro-matogr., 112: 651-662, 1975.

24. Parks, R. E., Jr., and Brown, P. R. Incorporation of nucleosides into thenucleotide pools of human erythrocytes. Adenosine and its analogues. Biochemistry, 12: 3294-3302, 1973.

25. Askari, A. Modifying effects of anions on the alkali-cation-activated AMPdeaminase of human erythrocyte. Mol. Pharmacol., 2: 518-525, 1966.

26. Snyder, F. F., Dyer, C, Seegmiller, J. E., Goldblum, R. M., Mills, G. C,and Schmalstieg, F. C. Substrate inhibition of adenosine phosphorylation inadenosine deaminase deficiency and adenosine-mediated inhibition of PP-ribose-P dependent nucleotide synthesis in hypoxanthine phosphoribosyl-transferase deficient erythrocytes. J. Inherited Metab. Dis., //: 174-183,1988.

27. Bontemps, F., Van den Berghe, G., and Hers, H. G. Evidence for a substratecycle between AMP and adenosine in isolated hepatocytes. Proc. Nati. Acad.Sci. USA, SO:2829-2833, 1983.

28. Lindberg, B., Klenow, H., and Hansen, K. Some properties of partiallypurified mammalian adenosine kinase. J. Biol. Chem., 242: 350-356, 1967.

29. Miller, R. L., Adamczyk, D. L., Miller, W. H., Koszalka, G. W., Rideout, J.L., Beacham, L. M., Ill, Chao, E. Y., Haggerty, J. J., Krenitsky, T. A., andElion, G. B. Adenosine kinase from rabbit liver. II. Substrate and inhibitorspecificity. J. Biol. Chem., 254: 2346-2352, 1979.

30. Hamasaki, N., and Rose, Z. B. The binding of phosphorylated red cellmetabolites to human hemoglobin. A. J. Biol. Chem., 249:7896-7901,1974.

31. Henderson, J. F., Zombor, G., Burridge, P. W., Barankiewicz, G., and Smith,C. M. Relationships among purine nucleoside metabolism, adenosine tri-phosphate catabolism, and glycolysis in human erythrocytes. Can. J.Biochem., 57: 873-878, 1979.

32. Smith, C. M., Belch, A., and Henderson, J. F. Hemolysis in mice treatedwith deoxycoformycin, an inhibitor of adenosine deaminase. Biochem. Pharmacol., 29:1209-1210, 1980.

33. Canonico, P. G., Kastello, M. D., Spears, C. T., Brown, J. R., Jackson, E.A., and Jenkins, D. E. Effects of ribavirin on red blood cells. Toxicol. Appi.Pharmacol., 74:155-162, 1984.

34. Van den Berghe, G., Bronfman, M., Vanneste, R., and Hers, H. G. Themechanism of adenosine triphosphate depletion in the liver after a load offructose. A kinetic study of liver adenylate deaminase. Biochem. J., 702:601-609, 1977.

4989

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1989;49:4983-4989. Cancer Res   Françoise Bontemps and Georges Van den Berghe  Deaminase-inhibited Human ErythrocytesDeoxyadenosine and by Nucleoside Analogues in Adenosine Mechanism of Adenosine Triphosphate Catabolism Induced by

  Updated version

  http://cancerres.aacrjournals.org/content/49/18/4983

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/18/4983To request permission to re-use all or part of this article, use this link

on May 20, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from