Cytotoxic and Metabolic Effects of Adenosine and Adenine on Human Lymphoblasts1

Home , Adenine, Adenosine, Adenosine deaminase, Adenosine monophosphate, Pyrimidine, Uracil

[CANCER RESEARCH 38, 2357-2362, August 1978] 0008-5472/78/0038-OOOOS02.00 Cytotoxic and Metabolic Effects of Adenosine and Adenine on Human Lymphoblasts1

Floyd F. Snyder,2 Michael S. Hershfield,3 and J. Edwin Seegmiller

Department of Medicine, University of California, San Diego, La Jolla, California 92093

ABSTRACT and further studies have shown that cytotoxic concentra tions of adenosine reduce pyrimidine ribonucleotides (12, The metabolic and growth inhibitory effects of adeno- 17, 21, 31) and NAD+ (17, 31) and cause an accumulation of sine toward the human lymphoblast line WI-L2 were po orotic acid (12, 31). Adenosine also increases cellular ade tentiated by the adenosine deaminase inhibitors erythro- nine nucleotides (12, 17, 31), including a transient increase 9-(2-hydroxy-3-nonyl) adenine (EHNA) and coformycin. in cAMP4 concentration (34, 35). In addition to studies in EHNA, 5 nu, or coformycin, 3.5 Â¿iM,atconcentrations that cultured cells, increased adenine ribonucleotide pools have inhibited adenosine deaminase activity more than 90%, been found in erythrocytes (1, 25) and lymphocytes (28) of had little effect on cell growth or the metabolic parameters adenosine deaminase-deficient, immune-defective patients. studied. Adenosine, 50 /Â¿M,plusEHNA, 5 JUM,arrested cell The selective toxicity of adenosine to dividing lymphoid growth in both parent and adenosine kinase-deficient cells, including inhibition of both the response of human lymphoblasts, implicating the nucleoside as the mediator peripheral blood lymphocytes to mitogen and the growth of of the cytostatic effect. Adenosine, 50 /KM,in combination lymphoblastoid cell lines, is considered a possible basis for with the adenosine deaminase inhibitors reduced 14C02 the severe combined immunodeficiency disease associated generation from [1-14C]glucose by 38%, depleted 5-phos- with a hereditary absence of adenosine deaminase activity phoribosy 1-1-pyrophosphate by more than 90%, and re (11,24). duced pyrimidine ribonucleotide concentrations. Uridine, It has been suggested that the toxic effects of adenosine 10 or 100 ftM, reversed adenosine plus EHNA growth are caused by an increased cellular adenine nucleotide inhibition in WI-L2 but not in adenosine kinase mutants. pool, but some evidence has suggested a mechanism(s) of Adenine, 500 //M, which may be converted to the same toxicity that does not require conversion of adenosine or intracellular nucleotides as adenosine, reduced the adenine to nucleotides. Certain analogs of adenosine that growth rate by 50% in both parent and adenine phospho- cannot be phosphorylated are cytotoxic (19, 32), and we ribosyltransferase-deficient lymphoblasts. Although ade have found the toxicity of adenosine to persist in adenosine nine also depleted cells of 5-phosphoribosyl-l-pyrophos- kinase-less mutants of the WI-L2 human splenic lympho phate and reduced pyrimidine ribonucleotide by 50%, the blast line; mutants lacking adenine phosphoribosyltransfer- mechanisms of adenine and adenosine toxicity differ. In ase were as sensitive as was their parent line to growth contrast to the ability of uridine to reverse adenosine inhibition by adenine (16). We now report more detailed cytostasis, growth inhibition by adenine was not reversed studies on the biochemical effects of adenosine and ade by uridine, indicating that pyrimidine ribonucleotide de nine on these lymphoblast lines. In these studies we have pletion is not the primary mechanism of adenine toxicity. used 2 potent inhibitors of adenosine deaminase, coformy cin (26) and EHNA (27). We have compared the effects of INTRODUCTION adenosine and adenine since both purines can be con verted to the same intracellular nucleotides (Chart 1) and The mechanism(s) of cytoxicity for the naturally occurring might therefore be expected to have similar mechanisms of nuceoside, adenosine, and its related base adenine is not toxicity if their effects are related only to expanded adenine understood but remains important, with growing interest in nucleotide pools. the use of adenosine deaminase inhibitors in combination chemotherapy. Both adenosine (9, 10, 18, 30) and adenine (20) block mitogen-induced transformation of human lym MATERIALS AND METHODS phocytes. Adenine toxicity toward mouse L-cells was par Biochemicals. Radiochemicals were purchased from tially overcome by certain pyrimidines (2), and adenosine Amersham/Searle Corp., (Arlington Heights, III.): [8- toxicity was reversed by uridine in normal (12) but not in 14C]adenine, 59 mCi/mmol; [8-14C] adenosine, 59 mCi/ adenosine deaminase-deficient fibroblasts (4). Hilz and mmol; [8-14C]hypoxanthine, 59 mCi/mmol; [1-14C]glucose, Kaukel (17) first documented a decrease in intracellular 9.37 mCi/mmol; [6-14C]glucose, 5.0 mCi/mmol; and sodium DTP concentration in adenosine-inhibited HeLa cells (21), [14C]formate, 59 mCi/mmol. Adenine, adenosine, hypoxanthine, uridine, and PP-ribose-P (sodium salt) were pur 1This work was supported by Grants AM-1362, AM-5646, and GM-17702 chased from P-L Biochemicals (Milwaukee, Wis.). EHNA from NIH and grants from the National Foundation and the Kroc Foundation. 2 Present address: Division of Pediatrics and Medical Biochemistry, Fac was provided by Wellcome Research Labs (Research Tri- ulty of Medicine, The University of Calgary, Calgary, Alberta T2N 1N4, Canada. To whom requests for reprints should be addressed. 3 Present address: Department of Medicine, Duke University Medical 4The abbreviations used are: cAMP, cyclic adenosine 3':5'-monophos- Center, Durham, N. C. 27710. phate; EHNA, eryr/ii-o-9-(2-hydroxy-3-nonyl) adenine; PP-ribose-P, 5-phos- Received December 13, 1977; accepted May 11,1978. phoribosyl-1 -pyrophosphate.

AUGUST 1978 2357

ATP conditioned to more than 3 months of growth in medium

II supplemented with 10% horse serum. This serum was ADP previously shown to deaminate less than 5 /Â¿molof 50 /XM II adenosine per 25 hr (30). The growth rate of WI-L2 lymphoblasts in this medium was inhibited 50% by approximately

EHNA, Cofonnycln 200 /Â¿Madenosine, and 5000 /UM adenosine completely /If(ADENOSINE Â« INOSINE ' (AD arrested growth (Chart 2). The adenosine deaminase inhib DAP \ t ? TG A itors EHNA, 5 /iM, and coformycin, 3.5 /Â¿M(1 pig/ml), ADENINE HYPOXANTHINE inhibited lymphoblast adenosine deaminase activity more Chart 1. Purine interconversion showing sites of selected lymphoblast than 95% in either extracts or whole cells but had little mutation and enzyme inhibition. DAP, 2,6-diaminopurine resistant and effect on growth. WI-L2 grew at approximately 90% of the deficient in adenine phosphoribosyltransferase; MTV, 6-methylthioinosine resistant and deficient in adenosine kinase; TG . 6-thioguanme resistant and normal growth rate for 8 weeks in the presence of 5 /Â¿M deficient in hypoxanthine-guanine phosphoribosyltransferase; EHNA and EHNA with weekly subculturing and addition of EHNA. coformycin, inhibitors of adenosine deaminase; AMPS, adenylosuccinate. EHNA increased the sensitivity of WI-L2 lymphoblasts to growth inhibition by adenosine greater than 10-fold, such angle Park, N. C.), and coformycin was provided by Dr. H. that 100 ^M adenosine completely arrested growth in the Umezawa, Institute for Microbial Chemistry, Tokyo, Japan. presence of 5 /Â¿MEHNA (Chart 2). Complete growth inhibi Lymphoblasts. The human splenic lymphoblast line Wl- tion by 50 tÂ¿Madenosine and 5 /Â¿MEHNA occurred after 24 L2 (23) was grown in suspension culture supplemented with hr of exposure or approximately 1 cell doubling (Chart 3/4) 2 nriM glutamine and 10% horse or fetal calf serum (Flow and was reversible for at least 72 hr of culture (Chart 30), Laboratories, Rockville, Md.) as previously described (16, demonstrating retention of cell viability despite growth 31). The isolation and characterization of clonal lympho arrest. The combination of 50 /J.M adenosine and 3.5 /Â¿M blast lines derived from WI-L2 deficient in adenosine kinase coformycin also arrested growth of WI-L2 lymphoblasts (EC 3.7.1.20) (MTI), both adenosine kinase and hypoxan after 24 hr. thine-guanine phosphoribosyltransferase (EC 2.4.2.8) (MTI- In studies with lymphoblast extracts, we found the appar TG), and adenine phosphoribosyltransferase (EC 2.4.2.7) ent Km's of adenosine kinase and adenosine deaminase for (DAP) were previously described (16). Cell counts were adenosine to be approximately 2 to 4 and 40 to 50 /Â¿M, measured by a Model ZB, Coulter counter. respectively. The maximal velocity of the deaminase was Metabolic Studies. Adenosine kinase, adenine phospho approximately 10-fold greater than that of the kinase. The ribosyltransferase, and hypoxanthine-guanine phosphori metabolism of extracellular adenosine by lymphoblasts ap bosyltransferase were assayed in lymphoblast extracts as peared to be governed by the characterisitics of these 2 previously described (16, 30). Adenosine metabolism was enzymes. Thus deamination was the principal route of studied in the intact lymphoblast as previously described adenosine metabolism for intact WI-L2 exposed to 80 /Â¿M (30). Intracellular PP-ribose-P concentrations (15, 16, 31) adenosine; the rates of deamination and phosphorylation and de novo purine synthesis (15) were measured as before. were 1035 and 74 pmol/106 cells/min, respectively. At a cAMP concentrations were measured according to the lower concentration of adenosine, 4 Â¿Â¿Morapproximately method of Wastili ef al. (33). Intracellular nucleotides were twice the Kmfor adenosine kinase, phosphorylation was the measured by high-pressure liquid chromatography as pre principal route of adenosine metabolism; the rates of deam viously described (6, 31). ination and phosphorylation were 33 and 65 pmol/106 cells/ The metabolism of [1-l4C]glucose to 14C02 was examined min, respectively. The products of deamination, Â¡nosine or as a measure of hexose monophosphate shunt activity. hypoxanthine (500 /Â¿M),did not inhibit lymphoblast growth. After 24 hr of incubation with test compounds, lymphoblasts were harvested by centrifugation, 120 x g for 3 min, and resuspended in fresh medium lacking glucose with 12 mW sodium phosphate, pH 7.2, and the original concentra tion of test compounds. To 0.5-ml suspensions of 0.5 x 106 cells in stoppered 15- x 80-mm tubes, a final concentration of 2.5 mW [1-14C]glucose or [6-14C]glucose, 5.0 mCi/mmol, was added. After 30 min of incubation at 37Â°,reactions were stopped by injection of 0.5 ml of 5 N H2SOâ€ž;then0.25 ml of Hyamine hydroxide (New England Nuclear, Boston, Mass.) was added to center wells (Kontes Glass, Berkeley, Calif.) containing fluted filter paper. After a further 60-min io- IO'4 IO" IO"' shaking, trapped 14CO2was counted at 70% efficiency by placing center wells in 10 ml Bray's solution (5). ADENOSINE I molar I Chart 2. Potentiation of adenosine-mediated growth inhibition by EHNA. WI-L2 lymphoblasts were cultured in the absence or presence of adenosine ( O) or adenosine plus 5 /Â¿MEHNA (â€¢).Each point represents the reciprocal RESULTS fraction of the mean doubling time of duplicate adenosine-treated cultures measured over a 4- to 5-day growth period with cell number measured every Effects of Adenosine on Human Lymphoblast Growth 24 hr, compared to the mean doubling time in the absence of additions. and Metabolism. The human lymphoblast line WI-L2 and Cells treated with 5 fiM EHNA had a reciprocal relative doubling time of 0.92. Cells were incubated for 30 min with EHNA prior to addition of adenosine. derived mutants used in studies of adenosine toxicity were The initial cell density was 0.5 to 1.0 x 105 cells/ml.

2358 CANCER RESEARCH VOL. 38

Table 1 Effects of adenosine, adenine, hypoxanthine, and uridine on lymphoblast nucleotide concentrations Additions were made to logarithmically growing cultures at a density of 2.5 to 4.0 x 10s cells/ml 24 hr prior to harvesting, extraction, and high-pressure liquid chromatography as described in "Materials and Methods." In Experiments A and B, cells were grown in medium containing 10% horse serum, and in Experiment C cells were grown in medium containing 10% fetal calf serum. AMP and orotic acid were not separated by the Chromatographie system, and peak area was calculated for AMP standard [in a previous report we have shown this area to be >90% AMP for untreated WI-L2 cultures (31)]: IAXP = AMP + ADP + ATP; 2GXP = GMP + GDP + GTP; SUXP = UMP + UDP + UTP + UDP-sugars. Results are the average of duplicate determinations with the following range for control cultures: 2AXP, Â±5%;2GXP, Â±10%; SUXP, Â±9%. Nucleotide concentration (nmol/10* cells)

Additions AMP 2AXP 2GXP 2UXP 24 48 72 96 120 A. WI-L2 None 0.17 5.6 1.5 3.1 HOURS Chart 3. Growth inhibition by adenosine in parent and adenosine kinase- Adenosine, 0.05 mM, + 0.13 9.6 1.4 1.1 deficient lymphoblasts. Lymphoblasts were cultured with no additions (â€¢)or EHNA, 0.005 mM 5 /iM EHNA plus 50 /IM adenosine (O), as described for Chart 2. A. 10 (A) or 100 /Â¿M(V)und ine was added to EHNA-plus-adenosine-treated lymphoblasts B. MTI at 24 hr; B, lymphoblasts exposed to EHNA plus adenosine were resuspended in fresh medium without additions at 24, 48, or 72 hr (A); C', EHNA- None 0.18 4.8 0.9 2.0 plus-adenosine-treated, adenosine kinase-deficient (AK~)lymphoblasts (MTI) Adenosine, 0.05 mM, + 0.14 5.2 0.6 1.2 EHNA, 0.005 mM were cultured in the absence (O) and presence (A) of 100 ^M uridine; D. EHNA and adenosine were added to adenosine kinase-hypoxanthine-guanine phosphoribosyltransferase-deficient (AK~,HGPRT~) lymphoblasts (MTI- C. WI-L2 TG), which were subsequently cultured in the absence (O) and presence (A) None 0.07 4.3 1.3 2.9 of 100 Â¿Â¿Muridine. Hypoxanthine, 0.5 mw 0.15 4.4 1.1 1.4 Adenine, 0.5 mM 0.05 6.2 1.3 1.5 Thus the adenosine deaminase inhibitor blocks the major Adenine, 0.5 mM, + 0.10 4.9 0.8 7.0 route for metabolism of substantial amounts of exoge- uridine, 1.0 mM nously supplied adenosine and could permit excessive conversion of adenosine to nucleotides or simply increase nucleotide analysis (Table 1) was 2.5- to 5-fold greater than the half-life of adenosine, either of which may be a possible that for the growth studies of Charts 2 and 3. Adenosine, 50 basis for the potentiation by the inhibitor of adenosine fj.M, plus EHNA, 5 /Â¿M,depletedtotal uracil ribonucleotides toxicity. to 35% of control, and adenine ribonucleotides, essentially We examined the question of adenosine growth inhibition ATP, increased (Table 1), presumably from the phosphoryl- being mediated by the nucleoside or some phosphorylated ation of adenosine. Exposure of MTI cells to adenosine plus product by using adenosine kinase-deficient lymphoblasts. EHNA also reduced uracil and guanine ribonucleotides, Adenosine kinase-deficient mutants were selected for re whereas total adenine plus guanine ribonucleotide concen sistance to 2 ^M 6-methylthioinosine and were designated trations were essentially unchanged. Adenosine plus EHNA MTI (16). A mutant deficient in both adenosine kinase and treatment reduced the concentration of CTP to 80 and 40% hypoxanthine-guanine phosphoribosyltransferase was de of untreated cells for WI-L2 and MTI, respectively. Thus in rived from MTI by selection for resistance to 10 /Â¿M6- adenosine kinase-deficient cells adenosine plus EHNA de thioguanine and was designated MTI-TG. The specific activ creased pyrimidine nucleotides under conditions where no ities of adenosine kinase and hypoxanthine-guanine phos- significant change in purine ribonucleotide concentrations phoriboxyltransferase in extracts of mutant lymphoblasts had occurred. At a later stage of complete but reversible were less than 0.1 and 2%, respectively, of control WI-L2 growth inhibition of WI-L2, the nucleotide concentrations (16). Both adenosine kinase-deficient lines showed inhibi in 'adenosine-treated'5 cells compared to control were: tion of growth by the combination of 50 ^M adenosine and UTP, 0.05; CTP, 0.17; UDP-sugars, 0.35; NAD*, 0.52; GTP, 5 /nM EHNA (Chart 3, C and D); complete growth arrest 1.14; and ATP, 1.62 (31). The extensive depletion of pyrimi required 48 hr. Adenosine was therefore growth inhibitory dine nucleotides was accompanied by an accumulation of in the absence of appreciable metabolism via phosphoryla- intracellular orotic acid (31). The growth inhibitory effect of tion or deamination. adenosine and EHNA was overcome by addition of 10 or The effect of adenosine plus EHNA on WI-L2 and MTI 100 /U.Muridine (Chart 3A). Uridine, 100 /*M, in contrast to lymphoblast ribonucleotide concentrations was determined its effect on parental cells, failed to reverse adenosine- more directly by high-pressure liquid chromatography of cell extracts (6, 31). At the time of harvesting, cell growth 6 WI-L2 lymphoblasts were incubated for 10 hr with 500 ÃŸMcAMP,which was not completely arrested because the initial concentra was shown to be slowly converted extracellularly to adenosine by horse tion of lymphoblasts exposed to adenosine plus EHNA for serum activities.

AUGUST 1978 2359

Table 2 14C02 generation. Adenine produced less than 10% inhibi Effect of adenosine in combination with adenosine deaminase tion, either alone or in combination with coformycin, indi inhibitors upon lymphoblast PP-ribose-P concentration cating a specificity for inhibition by the nucleoside adeno Parent and adenosine kinase-deficient (MTI) lymphoblasts were sine. Thus an adenosine-mediated inhibition of pentose cultured for 24 or 48 hr in the absence or presence of 3.5 /Â¿M coformycin, 5 Â¿iMEHNA, or 50 Â¡J.Madenosine, and PP-ribose-P phosphate synthesis could account in part for the reduction concentrations were measured as described in "Materials and in lymphoblast PP-ribose-P concentration. The generation Methods." Initial cell density was 0.5 to 0.8 x 105 cells/ml. In the of 14CO2from [6-14C]glucose, 2.5 mM, was less than 5% of absence of additions, PP-ribose-P concentrations (pmol/106 cells) the rate of 14CO2generation from [1-14C]glucose, indicating at 24 and 48 hr, respectively, were 133 and 290 for WI-L2 and 140 a low rate of glycolytic metabolism in the amino acid-rich and 190 for MTI. medium. Relative PP-ribose-P concentration Effects of Adenine on Lymphoblast Growth and Metab olism. Adenosine and adenine, although not directly inter- WI-L2AdditionsNone convertable in the human lymphocyte (29, 30), may be hr1.00 hr1.00 hr1.00 hr1.00 converted to the same intracellular nucleotides (Chart 1), and we have therefore compared their effects on WI-L2 Coformycin 1.28 0.84 0.94 1.59 lymphoblasts. Approximately 0.6 mM adenine produced EHNA 0.93 1.08 1.04 0.88 50% growth inhibition of lymphoblasts, whereas 0.5 IDM Adenosine 0.32 0.21 0.15 0.22 hypoxanthine was not growth inhibitory (16). Cell lines Coformycin + adenosine <0.01 0.01 0.29 <0.01 selected from WI-L2 that were resistant to 200 /xM 2,6- EHNA + adenosine24 0.0748 0.07MTI24 0.2448 <0.01 diaminopurine, designated DAP, compared to 50% growth inhibition of the parental strain at 5 /J.Mdiaminopurine, had induced growth inhibition in the adenosine kinase-deficient less than 0.5% of the parental adenine phosphoribosyl lymphoblasts (Chart 3, C and D). transferase activity in lymphoblast extracts (16). In intact The adenosine-mediated depletion of pyrimidine nucleo- adenine phosphoribosyltransferase-deficient lymphoblasts, tides and accumulation of orotic acid (31) suggested a the rate of nucleotide synthesis from 0.5 mM [14C]adenine possible deficiency of PP-ribose-P, which is required for was <0.1% of that of parental cells (16). Despite an inability further metabolism of orotate to pyrimidine nucleotides. to metabolize adenine, the DAP mutant showed no in The PP-ribose-P concentrations of lymphoblasts cultured in creased resistance or sensitivity to the growth inhibitory the presence of adenosine and inhibitors of adenosine effects of adenine (16). Thus adenine, like adenosine, deaminase activity were examined (Table 2). Despite inhibit remains growth inhibitory without being converted to intra ing lymphoblast adenosine deaminase activity by more than cellular nucleotides. 95%, coformycin or EHNA alone had little effect on either The effect of adenine and hypoxanthine on PP-ribose-P cell growth or PP-ribose-P concentration. Adenosine, 50 concentration, de novo purine synthesis, and ribonucleo- Â¿IM,in combination with either coformycin of EHNA re tide concentrations was examined. Adenine, 100 /Â¿M,which duced the PP-ribose-P concentration by more than 90% in just begins to slow growth, and hypoxanthine, 500 /Â¿.M, both parent and adenosine kinase-deficient lymphoblasts. which does not inhibit growth, both strongly inhibit de novo Complete PP-ribose-P depletion in MTI cells ocurred at 48 purine synthesis throughout log-phase growth and cause hr, when the increase in cell density had ceased (Chart 3C). prolonged reduction in the intracellular concentration of We have attempted to elucidate the biochemical site of PP-ribose-P (Table 4). The effects of 500 /Â¿Madenine or adenosine-mediated reduction in pyrimidine nucleotides and PP-ribose-P. Adenosine, 50 to 500 (J.M, in the absence or presence of EHNA, 5 /LC.M,has no significant effect on Table 3 Effect of adenine, adenosine, and coformycin on [ifC]glucose orotate phosphoribosyltransferase (F. F. Snyder, unpub lished data) or PP-ribose-P synthetase (31) activities in Wl- metabolism in lymphoblasts Lymphoblasts, 4 x 105 cells/ml, were cultured for 24 hr in the l_2 extracts. Because our results suggest that growth inhi absence or presence of 50 /J.M adenine, 50 /Â¿Madenosine, 3.5 ^M bition is mediated by adenosine and since this nucleoside coformycin, or 5 Â¿tMEHNA; harvested; and resuspended in fresh is known to increase cAMP concentrations of lymphoid medium with additions but lacking glucose. After 30 min of incu bation in fresh medium, either [1-14C]glucose or [6-14C]glucose, cells (34, 35), cAMP concentrations were measured. No 2.5 mM (5.0 mCi/mmol), was added, and 14CO2 generation was significant change in lymphoblast cAMP concentration was measured as described in "Materials and Methods." In the absence produced by 24 hr of culture in the absence or presence of of further additions, the generation of 14CO2was 53, 940, and 1500 50 UM adenosine plus 5 Â¡J.MEHNA; control cells had 0.4 cpm/106 cells/30 min from [1-14C]glucose and [6-14C]glucose, re pmol of cAMP per 106 cells. spectively. The oxidative branch of the pentose phosphate pathway Relative 14CO2generation from [1-14C]glucose is an important route for PP-ribose-P synthesis in the Additions mammalian cell (13). As an estimate of oxidative pentose None 1.00 phosphate synthesis from glucose, the generation of '"CO., Adenine 0.92 from [1-14C]glucose was measured after 24 hr of exposure Adenosine 0.93 0.88 of lymphoblasts to adenosine or adenine in the absence EHNA EHNA + adenosine 0.79 and presence of coformycin or EHNA (Table 3). The growth Coformycin 0.86 inhibitory combination of adenosine and coformycin or Coformycin + adenine 0.94 adenosine and EHNA produced the greatest inhibition of Coformycin + adenosine 0.62

CANCER RESEARCH VOL. 38 2360

Table 4 adenosine by lymphoid cells is insufficient to account for Effects of growth in adenine or hypoxanthine on purine synthesis immune dysfunction in adenosine deaminase-deficient pa de novo and PP-ribose-P concentration in parent and adenine tients. Other metabolic (7) and immunological (8) evidence phosphoribosyltransferase-deficient lymphoblasts also supports these conclusions. The potentiation of aden Parallel cultures of WI-L2 and a DAP clone (<0.1% adenine phosphoribosyltransferase activity) were incubated at 37Â°ingrowth osine effects by adenosine deaminase inhibitors may be medium (10% dialyzed fetal calf serum) containing the indicated understood in terms of their having blocked the principal additions. After 26 hr of exposure, the rate of labeling of Â¡ntracel- route for converting exogenously supplied adenosine to lular purines with sodium [14C]formate (2.17 mw, 4.6 Â¿tCi//Â¿mol) nontoxic metabolites. and the concentration of PP-ribose-P were determined as "Materi als and Methods." Pyrimidine nucleotide depletion was postulated to be the primary cause of adenosine toxicity on the basis of adeno- syn sine-mediated growth inhibition being reversed by uridine thesis (cpm/ (12). In adenosine kinase-deficient lymphoblasts, however, 30min/106 P (pmol/ AdditionsNone lineWI-L2 cells)4575 10Â«cells)236 uridine failed to reverse adenosine toxicity in the pres ence of EHNA (Chart 3). The inability of uridine to reverse DAPWI-L2 4590160 384<232 adenosine plus EHNA growth inhibition in the adenosine kinase-deficient cells may be interpreted as follows. Under Adenine, 0.5 rriMCell conditions of adenosine-mediated PP-ribose-P depletion, DAPPurine 1040PP-ribose- cells may have limited PP-ribose-P-dependent synthesis of

Hypoxanthine, 0.5 WI-L2 215 both purine and pyrimidine nucleotides, and this view is mw supported by the observed decrease in pyrimidine and guanine nucleotides. The inability of MTI cells to phospho- hypoxanthine in the absence and presence of 1 rriM uridine rylate adenosine may further limit purine nucleotide synthe sis in these cells as compared to WI-L2. Although there may on intracellular concentrations of adenine, guanine, and uracil ribonucleotides were examined (Table 1). Both ade be other effects of adenosine in adenosine deaminase and adenosine kinase-deficient cells, limitation of PP-ribose-P- nine and hypoxanthine cause a 50% reduction in the con centration of uracil ribonucleotides, but neither causes an dependent reactions is sufficient to account for growth inhibition caused by adenosine in the adenosine kinase- accumulation of orotic acid. In the presence of 1 nriM uridine, pyrimidine nucleotide concentrations were greater deficient cell. than in the control cells (Table 1); nevertheless, uridine had Adenine remained toxic to mutant lymphoblasts deficient virtually no effect on the growth inhibition caused by in adenine phosphoribosyltransferase activity. The bio adenine (16). These experiments indicate that: (a) dimin chemical effects of growth inhibitory concentrations of ished PP-ribose-P concentration per se is not growth inhib adenosine and adenine on human lymphoblasts are quali itory since adenine, which is toxic, and hypoxanthine, tatively similar in that both decrease pyrimidine nucleotides which is not toxic, both decreased PP-ribose-P concentra and PP-ribose-P concentrations. The mechanism by which tion to the same extent; and (o) diminished pyrimidine adenosine reduced pyrimidine nucleotide concentrations nucleotide concentration is not the sole basis of adenine appears to differ from that of adenine and hypoxanthine in toxicity since hypoxanthine causes a similar decrease in that adenosine (31) but not the purine bases (Table 1) intracellular pyrimidine ribonucleotides, and, more directly, causes an accumulation of orotic acid despite nearly equiv alent depression of PP-ribose-P concentration (Tables 2 repletion of these pools with uridine does not alter inhibi tion of growth by adenine. and 4). Also, uridine reversed adenosine but not adenine growth inhibition in WI-L2. The PP-ribose-P-dependent conversion of purine bases to nucleotides can deplete PP- ribose-P, whereas adenosine in the presence of adenosine DISCUSSION deaminase inhibitors is not readily converted to purine Both the growth inhibitory and metabolic effects of aden- bases (Chart 1) and therefore appears to deplete PP-ribose- osine persisted when adenosine metabolism via phospho- P by another mechanism, perhaps by inhibiting PP-ribose- rylation or deamination was blocked by the adenosine P formation. Thus in a previous report we found no evi kinase mutation and the adenosine deaminase inhibitors. dence for adenosine-mediated inhibition of PP-ribose-P Thus growth inhibitory combinations of adenosine and synthetase activity in lymphoblast extracts (31), but adeno EHNA or coformycin depleted PP-ribose-P and pyrimidine sine plus coformycin or EHNA partially inhibited pentose ribonucleotide concentrations in WI-L2 and adenosine ki- phosphate synthesis under conditions of incomplete nase-deficient mutants. These observations are important growth arrest (Table 3). Adenine at a very high concentra in understanding the mechanism of adenosine toxicity. For tion (0.5 rriM) and after 26 hr of incubation (Table 4) did example, effects analogous to those of adenosine in reduc lower PP-ribose-P concentration in adenine phosphoribo- ing PP-ribose-P concentrations have been reported for syltransferase-deficient cells, but we have not determined 2',5'-dideoxyadenosine (32) and S'-deoxyadenosine (19), whether this occurred by a mechanism involving pentose which cannot be phosphorylated to the 5'-nucleotide. EHNA shunt inhibition. However, because growth inhibition by also inhibits PP-ribose-P-dependent nucleotide synthesis adenine does not produce as severe depletion of pyrimidine (14). The minimal effect of low concentrations of the aden ribonucleotides as does adenosine at comparable growth osine deaminase inhibitors alone in these and other studies inhibition, adenine and hypoxanthine appear to depress (3, 9, 22, 30) suggest that the endogenous formation of primarily the steady state concentration of PP-ribose-P

AUGUST 1978 2361

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1978 American Association for Cancer Research. F. F. Snyder et al. without completely blocking its availability for nucleotide 1976. 16. Hershfield. M. S., Snyder, F. F., and Seegmiller, J. E. Adenine and synthesis. Our studies suggest that pentose phosphate Adenosine Are Toxic to Human Lymphoblast Mutants Defective in Purine precursors other than glucose should be examined as Salvage Enzymes. Science, 797: 1284-1287, 1977. alternatives to uridine in counteracting adenosine toxicity. 17. Hilz, H., and Kaukel, E. Divergent Action Mechanism of cAMP and Dibutyrl cAMP on Cell Proliferation and Macromolecular Synthesis in HeLa S3 Cultures. Mol. Cellular Biochem.. 7. 229-239, 1973. 18. Hovi, T., Smyth, J. F., Allison, A. C., and Williams, S. C. Role of ACKNOWLEDGMENTS Adenosine Deaminase in Lymphocyte Proliferation. Clin. Exptl. Immu nol., 23: 395-403, 1976. We thank the laboratory of Dr. S. E. Mayer for cAMP assays and Inga 19. Hunting, D., and Henderson. J. F. Inhibition of Nucleotide Synthesis by Jansen and Michael Cruikshank for excellent technical assistance. 5'-Deoxyadenosine. Proc. Am. Assoc. Cancer Res.. 18: 133, 1977. 20. Ito, K., and Uchino, H. Control of Pyrimidine Biosynthesis in Human Lymphocytes. J. Biol. Chem.,257: 1427-1430, 1976. REFERENCES 21. Kaukel. E., Fuhrmann, U., and Hilz. H. Divergent Action of cAMP and Dibutyryl cAMP on Macromolecular Synthesis in HeLa S3 Cultures. 1. Agarwal. P. R., Crabtree. G. W.. Parks, R. E., Jr., Nelson, J. A., Biochem. Biophys. Res. Commun.,48: 1516-1524, 1972. Keightley, R., Parkman. R., Rosen, R. S., Stern, R. C.. and Polmar, S. H. 22. Lapi. L., and Cohen, S. S. Toxicities of Adenosine and 2'-Deoxy-Adeno- Purine Nucleoside Metabolism in the Erythrocytes of Patients with sine in L Cells Treated with Inhibitors of Adenosine Deaminase. Bio Adenosine Deaminase Deficiency and Severe Combined Immunodefi chem. Pharmacol.. 26: 71-76, 1977. ciency. J. Clin. Invest.. 57; 1025-1035. 1976. 23. Levy, J. A., Virolainen. M.. and Defendi. V. Human Lymphoblastoid 2. Aranow. L. Reversal of Adenine Toxicity by Pyrimidine Nucleosides. Lines from Lymph Node and Spleen. Cancer, 22. 517-524, 1968. Biochim. Biophys. Acta, 47: 184-185. 1961. 24. Meuwissen, H. J., Pickering, R. J.. Pollara, B.. and Porter. I. H. (eds.). 3. Ballet, J. J., Insel, R.. Merler, E., and Rosen, F. S. Inhibition of Combined Immunodeficiency Disease and Adenosine Deaminase Defi Maturation of Human Precursor Lymphocytes by Coformycin, an Inhib ciency. A Molecular Defect, pp. 73-83. New York: Academic Press, Inc., itor of the Enzyme Adenosine Deaminase. J. Exptl. Med., 743.- 1271- 1975. 1276.1976. 25. Mills. G. C., Schmalstieg, F. C., Trimmer, K. B., Goldman, A. S., and 4. Benke, P. J.. and Dittmar, D. Purine Dysfunction in Cells from Patients Goldblum, R. M. Purine Metabolism in Adenosine Deaminase Defi with Adenosine Deaminase Deficiency. Pediat. Res., 10: 642-646, 1976. ciency. Proc. Nati. Acad. Sei. U. S., 73: 2867-2871, 1976. 5. Bray. G. A. A Simple Efficient Liquid Scintillator for Counting Aqueous 26. Ohno, M., Yagisawa, N., Shibahara, S., Kondo, S.. Maeda, K., and Solutions in a Liquid Scintillation Counter. Anal. Biochem., 1: 279-285, Umezawa, H. Synthesis of Coformycin. J. Am. Chem. Soc., 96: 4326- 1960. 4327, 1974. 6. Brenton, D. P.. Astrin, K. H., Cruikshank, M. K., and Seegmiller, J. E. 27. Schaeffer, H. J., and Schwender, C. F. Enzyme Inhibitors. 26. Bridging Measurement of Free Nucleotides in Cultured Human Lymphoid Cells Hydrophobie and Hydrophyllic Regions on Adenosine Deaminase with Using High Pressure Liquid Chromatography. Biochem. Med.. 17: 231- Some9-(2-Hydroxy-3-alkyl)adenines. J. Med. Chem.. 77: 6-8, 1974. 247,1977. 28. Schmalstieg. F. C., Nelson. J. A., Mills. G. C.. Monahan, T. M.. Goldman, 7. Brox. L. W.. and Henderson, J. F., The 'Adenosine-Cycle' Is Not a A. S., and Goldblum, R. M. Increased Purine Nucleotides in Adenosine Significant Route of Purine Metabolism in Mammalian Cells. Can J. Deaminase-Deficient Lymphocytes. J. Pediat.,97: 48-51, 1977. Biochem..54. 200-202, 1976. 29. Scholar, E. M., and Calabresi, P. Identification of the Enzymatic Path 8. Burridge, P. W.. Paetkau, V.. and Henderson, J. F. Studies of the ways of Nucleotide Metabolism in Human Lymphocytes and Leukemia Relationship between Adenosine Deaminase and Immune Function. J. Cells. Cancer Res., 33: 94-103, 1973. Immunol., 119: 675-678, 1977. 30. Snyder, F. F., Mendelsohn, J., and Seegmiller, J. E. Adenosine Metabo 9. Carson, D. A., and Seegmiller. J. E. Effect of Adenosine Deaminase lism in Phytohemagglutinin Stimulated Human Lymphocytes. J. Clin. Inhibition upon Lymphocyte Blastogenesis. J. Clin. Invest., 57: 274-282, Invest..58: 654-666. 1976. 1976. 31. Snyder, F. F., and Seegmiller. J. E. The Adenosine-like Effect of 10. Fox, I. H., Keystone, E. C., Gladman, D. D., Moore, M., and Cane, D. Exogenous Cyclic AMP upon Nucleotide and PP-ribose-P Concentra Inhibition of Mitogen Mediated Lymphocyte Blastogenesis by Adeno tions of Cultured Human Lymphoblasts. Federation European Biochem. sine. Immunol. Commun., 4. 419-427, 1975. Soc. Letters. 66: 102-106. 1976. 11. Giblett. E. R.. Anderson, J. E., Cohen. F., Pollara. B., and Meuwissen, 32. Tyrsted, G., and Sartorelli. A. C. Inhibition of the Synthesis of 5- H. J. Adenosine Deaminase Deficiency in Two Patients with Severely Phosphoribosyl-1-pyrophosphate by 3'-Deoxyadenosine and Structur Impaired Cellular Immunity. Lancet, 2. 1067-1069, 1973. ally Related Nucleoside Analogs. Biochim. Biophys. Acta, 755. 619-622. 12. Green, H., and Chan, T. S. Pyrimidine Starvation Induced by Adenosine 1968. in Fibroblasts and Lymphoid Cells: Role of Adenosine Deaminase. 33. Wastili, W. B., Stull, J. T., Mayer, S. E., and Walsh, D. A. Measurement Science, 782. 836-837. 1973. of Cyclic 3'.5'-Adenosine Monophosphate by the Activation of Skeletal 13. Henderson, J. F. Regulation of Purine Biosynthesis, pp. 50-59. Washing Muscle Protein Kinase. J. Biol. Chem.. 246: 1996-2003. 1971. ton, D. C.: American Chemical Society, 1972. 34. Wolberg, G., Zimmerman. T. P., Hiemstra, K.. Winston, M., and Chu, L. 14. Henderson, J. F.. Brox, L., Zombor, G.. Hunting, D., and Lomax, C. A. C. Adenosine Inhibition of Lymphocyte-Mediated Cytolysis Possible Specificity of Adenosine Deaminase Inhibitors. Biochem. Pharmacol.. Role of Cyclic Adenosine Monophosphate. Science, 787: 957-959, 1975. 26: 1967-1972, 1977. 35. Zenser, T. V. Formation of Adenosine 3',5'-Monophosphate from Aden 15. Hershfield, M. S., and Seegmiller, J. E. Regulation of de Novo Purine osine in Mouse Thymocytes. Biochim. Biophys. Acta. 404. 202-213, Biosynthesis in Human Lymphoblasts. J. Biol. Chem., 257: 7348-7354, 1975.

2362 CANCER RESEARCH VOL. 38

Floyd F. Snyder, Michael S. Hershfield and J. Edwin Seegmiller

Cancer Res 1978;38:2357-2362.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/38/8/2357

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/38/8/2357. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.