Growth Inhibition of Transformed Mouse Fibroblasts by Adenine ...

THE JOURNAL OF BIOLOGICAL CHEM~~RY 0 1988 by The American Society for Biochemistry

Vol. 263, No. 25, Issue of

Growth Inhibition of Transformed Nucleotides Occurs via Generation

5, pp. 12367-12372,1966

Printed in U.S.A.

Mouse Fibroblasts by Adenine of Extracellular Adenosine* (Received

Gary A. Weisman$Q!l, Ilan Friedberg **

September

and Molecular Biology, Inc.

Kevin

From the $Section of Biochemistry, 11Department of Microbiology, The

D. LustigSQ,

Elizabeth

Lane&

Ning-na

Molecular and Cell Biology, Cornell University, George S. Wise Faculty of Life Sciences, Tel-Aviv

The growth of transformed mouse fibroblasts (3T6 cells) in medium containing 5% fetal bovine serum was inhibited after treatment with concentrations greater than 50 MM ATP, ADP, or AMP. Adenosine, the common catabolite of the nucleotides, had no effect on cell growth at concentrations below 1 mM. However, the following results indicate that the toxicity of ATP, ADP, and AMP is mediated by serum- and cell-associated hydrolysis of the nucleotides to adenosine. 1) ADP and AMP, but not ATP, were toxic to 3T6 cells grown in serum-free medium or medium in which phosphohydrolase activity of serum was inactivated. Under these conditions, the cells exhibited cell-associated ADPase and 5’-nucleotidase activity, but little ectoATPase activity. 2) Inhibition of adenosine transport in 3T6 cells by dipyridamole or S-(p-nitrobenzyl)-6thioinosine prevented the toxicity of ATP in serumcontaining medium and of ADP and AMP in serumfree medium. 3) A 16-24-h exposure to 125 PM AMP or ATP was needed to inhibit cell growth under conditions where serum- and cell-associated hydrolysis of the nucleotides generated adenosine in the medium continuously over the same time period. In contrast, 125 pM adenosine was completely degraded to inosine and hypoxanthine within 8-10 h. Furthermore, multiple doses of adenosine added to the cells at regular intervals over a 16-h period were significantly more toxic than an equivalent amount of adenosine added in one dose. Treatment of 3T6 cells with AMP elevated intracellular ATP and ADP levels and reduced intracellular UTP levels, effects which were inhibited by extracellular uridine. Uridine also prevented growth inhibition by ATP, ADP, and AMP. These and other results indicate that serumand cell-associated hydrolysis of adenine nucleotides to adenosine suppresses growth by * Work performed at Cornell University was done in the laboratory of Dr. Leon A. Heppel with support from his National Institutes of Health Grant AM-11789 and National Science Foundation Grant PCM 81-21029. 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” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 3 Present address: Dept. of Biochemistry, University of MissouriColumbia, Columbia, MO 65211. 11Supported by Grant Project MO-00493 from the United States Department of Agriculture, Agricultural Experiment Station, and University of Missouri Institutional Biomedical Research Support Grant RR 07053 from the National Institutes of Health. To whom correspondence and reprint requests should be addressed. ** Supported by the National Council for Research and Development (Israel), the Deutsches Krebsforschungszentrum (Heidelberg, Germany), Grant 86-0058 from the United States-Israel Binational Science Foundation (Jerusalem, Israel), and the Israel Cancer Research Fund.

adenosine-dependent

for publication,

HuangS, Ithaca, New University,

pyrimidine

Ilana

February

Belzer

9, 1988)

11,and

York 14853 and the Tel-Aviv, Israel 69978

starvation.

Extracellular adenine nucleotides inhibit the growth of a variety of mammalian cell types (l-7). ATP and AMP were toxic to mouse splenocytes (l), and AMP was shown to decrease DNA synthesis in mouse cells from spleen, lymph nodes, and peripheral blood (2). ATP and ADP inhibited the growth of CAPANand HT-29 (3,4), two human tumor cell lines; and ATP inhibited the growth of transformed mouse fibroblasts (5, 6) and murine erythroleukemia cells infected with Friend virus (7). Evidence has accumulated indicating that cell-surface receptors (purinoceptors) mediate many effects of extracellular adenine nucleotides (8, 9), including growth inhibition (10). Other studies (3, 4) have suggested that transmembrane nucleotide uptake leads to alterations in intracellular nucleotide levels, thereby inhibiting DNA synthesis. This study indicates that growth inhibition of mouse fibroblasts by adenine nucleotides is due to adenosine generated slowly from the nucleotides by enzymes on the cell surface and in serum. EXPERIMENTAL

PROCEDURES

Mater&-Insulin, transferrin, and epidermal growth factor were obtained from Collaborative Research (Bedford, MA). Dulbecco’s modified Eagle’s medium (DMEM),’ Ham’s F-12 medium, penicillin, streptomycin, and fetal bovine serum (FBS) were purchased from GIBCO; [2-3H]adenosine, [G-3H]adenosine 5’-monophosphate, [8‘Qadenosine 5’-diphosphate, and [b-“Cladenosine 5’-triphosphate were purchased from Du Pont-New England Nuclear. All nucleotides, nucleosides, and other chemicals were obtained from Sigma. Cell Culture Condition.-3T6 cells (spontaneously transformed derivatives of Swiss mouse 3T3 cells) were propagated and maintained as previously described (11, 12). The cells were grown in 75-cm* polystyrene flasks (Falcon Labware) in DMEM, pH 7.4, supplemented with 5% FBS, 10 mM HEPES, 44 mM NaHC03, 100 units/ ml penicillin, and 100 rg/ml streptomycin. Cell cultures were maintained in a humidified atmosphere of 5% CO* and 95% air at 37 “C. The cells were removed from the flasks at confluence with 0.05% (w/ v) trypsin in phosphate-buffered saline (10 mM sodium phosphate, 4 mM KCI, 137 mM NaCl), pH 7.4, containing 0.3 mM EDTA and subcultured into 35-mm polystyrene dishes (Falcon Labware) with 2.0 ml of growth medium. The plating density was 5 X 10’ tells/35mm dish. In some experiments, the cells were subcultured in growth medium containing 5% FBS which had been heat-treated for at least 2 h at 56 “C or in serum-free medium containing 3 volumes of DMEM, 1 volume of Ham’s F-12 medium, 5 pg/ml insulin, 5 fig/ml transferrin, and 20 rig/ml epidermal growth factor. A lo-fold excess of soybean trypsin inhibitor was used to inactivate the trypsin when subculturing in serum-free medium. 1 The abbreviations medium; FBS, fetal piperazineethanesulfonic sine.

12367

used are: DMEM, Dulbecco’s modified Eagle’s bovine serum; HEPES, 4-(2-hydroxyethyl)-lacid; NBTI, S-(p-nitrobenzyl)-6-thioino-

12368

Growth Inhibition by Extracellular Adenine Nucleotides

20Cell Growth Measurements-Cells were grown at 37"C for 16 h x Control 19following subculturing, and then adenine nucleotides or other compounds were added as indicated in the figure and table legends. Cells 18wereremovedfrom selected culture dishes at various intervals by 17trypsinization (see above), and cell number was determined using a 16Model 112T electronic cell counter (Particle Data, Elmhurst, IL). 15Measurement of Cell-surface NucleotidaseActiuity"3T6 cells were I4 subcultured in 35-mm dishes with various growth media (see above); o 13and after 16 h, [G-3H]adenosine 5'-monophosphate (0.8 Ci/mmol), [S-"C]adenosine 5"diphosphate (0.04 Ci/mmol), or [8-14C]adenosine x 115"triphosphate (0.04 Ci/mmol) was added to a final concentration of 125 p ~ After . an additional 1-24 h, the medium was sampled; and radiolabeled ATP, ADP, AMP, and adenosine were separated by anion-exchange chromatography on polyethyleneimine-coatedcellulose columns as previously described (13, 14). The amount of radioactivity in each fraction was determined by liquid scintillation count6ing. 5Separation of Adenosine and Adenosine Catabolites-Cellswere 4subcultured in 35-mm dishes in serum-free medium. After 24 h, [23H]adenosine (4 Ci/mmol) was added to a final concentration of 125 WM. The medium was sampled at various times, and adenosine was separated from its major catabolites, inosine and hypoxanthine, by anion-exchange chromatography on polyethyleneimine-coated cellu0 24 48 72 lose plates. Chromatography was performed using saturated ammoGrowth (hours1 nium sulfate, 0.05 M potassium phosphate, pH 6.0,isopropyl alcohol (79:19:2, v/v/v). The plates were cut into small squares, and the radioactivity in each fraction was determined by liquid scintillation counting. High Performance Liquid Chromatography of Cell Extracts-Soluble pools were extracted from the cells with ice-cold 7% (w/v) trichloroacetic acid. The trichloroacetic acid wasremoved from the aqueous solution by extraction with 0.5 M tri-n-octylamine in Freon113. The resulting organic and aqueous layers were separated by low speed centrifugation. The trichloroacetic acid-free aqueous layer was brought to pH 5.0-5.5 with 0.1 N ammonium acetate, pH 8.0, and stored at -20 "C until analyzed. ATP, ADP, and UTP in 200 pl of the aqueous layer were separated on a Partisil SAX anion-exchange column (10 pm, Whatman Inc., Hillsboro, OR) using a Model 2152 HPLC analyzer (LKB Instruments, Gaithersburg, MD). The column was eluted at room temperature with a linear gradient comprised of 0.007 M NaHzP04. HzO, pH 4.0 (initial buffer), and 1.0 M NaHzPOl. [Nucleoflde] p M HzO, pH 4.5 (final buffer). The flow rate was 1.5 ml/min, and nucleotide peaks were located and quantitated using a Spectra-PhysFIG. 1. Effect of ATP, ADP, and AMP on 3T6 cell growth ics 4270 chromatography integrator. in medium containing 6% FBS.Cells were subcultured in 35-mm polystyrene dishes in DMEM containing 5 % FBS. After 16 h (0 h on curve), ATP (O), ADP (A), or AMP (0) was added to afinal RESULTS concentration of 100 p~ (upper)or as indicated (lower). Other The growth of 3T6 cells in medium containing 5% serum cultures served as controls (X). Cella were removed from the dishes was completely inhibited after treatment with 100 p~ ATP by trypsinization (see "Experimental Procedures"), and cell number or ADP and partially inhibited after treatment with 100 PM was determined every 24 h over 3 days (upper) or 72 h after nucleotide AMP (Fig. 1, upper). In contrast, the growth of 3T6 cells in addition (lower). Lower, cell growth is expressed as a percentage serum-free medium (see "Experimental Procedures") was not relative to untreated controls and is plotted on a semilogarithmic scale. L

affected by 125 p~ ATP, but was completely inhibited by the addition of 125 p~ AMP and partially inhibited by 125 PM ADP (Fig. 2, upper). Dose-response curves for each nucleotide showed that ATP was a more potent growth inhibitor than ADP or AMP for cells grown in serum-containing medium (Fig. 1, lower) and that AMP was more potent than ADP for cells grown in serum-free medium (Fig. 2, lower). Adenosine, the common catabolite of the nucleotides, slightly inhibited cellgrowth at concentrations above 100 pM in serum-free medium (Fig. 2, lower) and above 1 mM in serum-containing medium (data not shown). Cell growth in serum-free medium was not affected by treatment with 2'-AMP, 3'-AMP, 3'dAMP, 5'-dAMP, IMP, orcatabolites of adenosine including inosine, hypoxanthine, and adenine at concentrations as high as 2 mM (data not shown). The inability of ATP to inhibit cell growth in serum-free medium suggested that some degradative product of ATP may be the inhibitor in serum-containing medium. Therefore, the degradation of 125 GM ATP, ADP, and AMP was monitored in cells incubating in serum-containing orserum-free medium. The datashowed that at least 97% of the extracellular adenine nucleotides were degraded to adenosine and adenosine catab-

olites over a 24-h period by cells incubating in medium containing 5% FBS (Table I). Serum alone contained significant ATPase, ADPase, and 5'-nucleotidase activity. Data with cells grown in serum-free medium showed that 95,51, and 6% of the AMP, ADP, and ATP, respectively, were degraded to adenosine and adenosine catabolites over a 24-h period (Table I), indicating that the cells exhibit substantial ecto-5"nucleotidase and ecto-ADPase activities, but little ecto-ATPase activity, over this time period. Serum-dependent degradation of adenine nucleotides was completely inhibited by incubating the serum with 1 mM dithiothreitol for 1 h or by heating the serum at 56 "C for at least 2 h (data not shown). The growth of cells in dithiothreitol- or heat-treated serum was inhibited by AMP and ADP, but not by ATP (data notshown), results similar to those obtained with cells grown in serum-free medium (Fig. 2). These data suggested that a catabolite of ATP formed by enzymes in serum was responsible for growth inhibition in serum-containing medium and that the lack of significant cell-associated ATPase activity rendered ATP ineffective as a growth inhibitor in media in which ATPase

Growth Inhibition

by Extracellular Adenine Nucleotides

12369

nase and nucleosidase (19). The slow catabolism of adenosine in the presence of dipyridamole and NBTI (Fig. 4) may be e due to nonmediated entry of adenosine or to an inhibitorinsensitive component of the adenosine transport system (16, 20). In the absence of dipyridamole and NBTI, 125 PM adenosine was almost completely catabolized in 8 h by cells in serum-free medium (Fig. 4). The rapid degradation of adenosine by 3T6 cells may explain its low toxicity as compared to adenine nucleotides. In contrast,serum-and cell-associated hydrolysis of adenine nucleotides caused continuous generation of extracellular adenosine and its breakdown products over 24 h (Table I) under conditions where ATP, ADP, and AMP were toxic to 3T6 cells (Figs. 1 and 2). Experiments suggested that it was critical to maintain a significant nucleotide concentration in the growth medium for longer than 8 h to observe growth inhibition (Fig. 5). Cells in serum-free medium were treated with 125 p~ AMP, and after various time periods, the medium 1 I I I I was replaced with similar medium conditioned by cells grown 96 I20 72 48 0 24 Growth (hours) in the absence of AMP for equal lengths of time. The data indicated that a 16-24-h exposure to AMP was necessary to 120,induce maximal growth inhibition and that treatment of the 0 ATP 0 cells for 8 h caused only 5-15% inhibition (Fig. 5). Similar 0 0 U e results were obtained with ATP in serum-containing medium 100 and with AMP in medium containing heat-treatedserum (Fig. 5). These findings support the idea that growth inhibition by 2 s ao- A adenine nucleotides is mediated by continuous generation of z u \@Adenosine adenosine over a 16-24-h period and that adenosine at equimolar concentrations is not toxic because it is depleted from f 60the media in 8 h. Other evidence supporting the theory that e a sustained level of extracellular adenosine is required for 0 5 40 growth inhibition came from studies in which adenosine was u added in small amounts to3T6 cell cultures periodically over 16 h. Adding increments of 5.0 pM adenosine to cells every 1 2o 20 h or10.0 p~ every 2 h for 16 h resulted in substantiallygreater inhibition (Fig. 6) than adding one initial dose of250 ptM I I I I adenosine (Fig. 2, lower). Furthermore, adenosine added to 25.0 O45& 50.0 75.0 125.0 250.0 the medium in multiple doses to a total concentration of 80 Concentration (pM) p~ was nearly as toxic to cell growth as the addition of one FIG. 2. Effect of ATP, ADP, AMP, and adenosine on 3T6 cell growth in serum-free medium. Cells were subcultured in dose of 80 PM AMP (Fig. 2, lower). The biochemical basis of adenosine toxicity has been studserum-free medium (see “Experimental Procedures”) in 35-mm dishes. After 16 h (0 h on curve), ATP (O), ADP (A), or AMP (0) ied in a number of cell lines (15, 16), including 3T6 cells (17). was added to a final concentration of 125 p~ (upper) or as indicated Extracellular adenosine has been reported to induce altera(lower). Lower,adenosine (0)was also added at theindicated concen- tions in intracellular ATP and ADP levels which subsequently trations.Othercultures served as controls (X). Cell number was inhibit either orotate phosphoribosyltransferase (21) or the determined every 24 h over 5 days (upper) or 120 h after nucleotide addition (lower).Lower, cell growth is expressed as a percentage accumulation of the substratefor the enzyme, phosphoribosyl de mu0 UTP and relative to untreated controls and is plotted on a semilogarithmic pyrophosphate (22),therebyinhibiting scale. DNA synthesis (15-17). Addition of uridine to cells prior to treatment with adenosine has, in some cases, prevented growth inhibition presumably by by-passing the inhibited step activity was absent or inactivated. The nature of the growth inhibitory substance for cells in pyrimidine biosynthesis (15-17). Several experimental retreated with adenine nucleotides was investigated. Adenosine, sults suggested that adenine nucleotides also caused inhibition the common breakdown product of ATP, ADP, and AMP, of 3T6 cell growth through pyrimidine starvation. Uridine, a t , the growth inhibitory inhibited the growth of a number of cell lines (15-17), al- concentrations above 2.5 p ~prevented though, in 3T6 cells, adenosine was considerably less toxic effects of 125 pM AMP in serum-free medium (Fig. 7) and of than adenine nucleotides (see Fig. 2, lower). Nevertheless, 125 I . ~ MATP and ADP in serum-containing medium (data not inhibitors of the adenosine translocator in mammalian cells, shown). Uridine, which competitively inhibits adenosine dipyridamole and S-(p-nitrobenzyl)-6-thioinosine(NBTI) transport in many cell types (23), had no significant effect on the uptake of 125 pM adenosine in 3T6 cells at uridine (16, 18), prevented the toxic effects on 3T6 cells of 125 p~ AMPandADP in serum-free medium (Fig. 3, upper and concentrations below 15 p~ (data not shown). Furthermore, ) nearly a 2-fold increase lower) and of 125 pM ATP in serum-containing medium (data extracellular AMP (125 p ~ induced not shown). Dipyridamole and NBTI significantly inhibited in intracellular levels of ATP and ADP and a 40% decrease the cell-associated breakdown of adenosine to inosine and in the intracellular level of UTP over a 48-h period (Table hypoxanthine (Fig. 4), consistent with the finding that these 11). AMP-induced changes in intracellular UTP levels were compounds inhibited adenosine uptake and subsequent deam- not seen if the cells were pretreated with 10 p~ uridine (Table ination and deribosylation by intracellular adenosine deami- 11), a level sufficient to prevent the growth inhibitory effect

-1

9r

Control

PAT.

I

L

t

I

Growth Inhibition by Extracellular Adenine Nucleotides

12370

TABLE I Adenine nucleotide hydrolysis by 3T6 cells in serum-containingand serum-free media Cells were subcultured in 35-mm dishes in DMEM containing 5% FBS (“Cells serum”) or in serum-free medium (“Cells + serum-free medium”). After 16 h, 125 p~ radiolabeled ATP (A), ADP (B), or AMP (C) was added to themedium, and adenine nucleotide hydrolysis was followedover 24 h (see “Experimental Procedures”). Nucleotide concentrations areexpressed as a percentage of the totallabel present inthe medium at thetime points indicated. Adenosine and derivative percentages include label in the form of inosine, hypoxanthine, and adenine. Extracellular label trapped intracellularly over the course of the experiment was less than 5% of the initial label. No hydrolysis of nucleotides was detected in serum-free medium over 24 h (data not shown).

+

Cells Oh”

+ serum

4l hh

8h

Cells + serum-free medium 24h

Serum (no cells)

O h2 844hhhl h

Oh2 844hhh lh % remaining

A. ATP ADP AMP Adenosine and derivatives 63 20 3

94 1 1 4

60 7 30

B. ADP AMP Adenosine and derivatives

93 7 0

62 29 9

28 34 40

C. AMP Adenosine and derivatives a Incubation time.

100 0

76 24

20 80

0 0.10 0.25

0.50

46 6 28

22 1 14

0 08 2 98

95 72 9077 2 2 0 1 2

79 19 0 2

18 2 3

20 2 4 6

94 1 32 1 4

61 8 27

11 8 81

0 3 97

89 5 6

84 10 6

71 11 18

62 28 10

44 5 51

93 7 0

4 96

1 99

94 6

84 16

62 38

24 76

5 95

94 6

I.o [Dipyrldarnole] pt4

IO

0

17

31 0 28 41

0 1 1 98

61 27 12

28 36 36

11 19 70

0 0 100

80 20

36 64

9 91

1 99

50 1

‘0

-

,\:“----* 2 .o

a 6.0

4.0

8.0

10.0

T h e (hour)

FIG. 4. Effect of dipyridamole and NBTI on adenosine transport in 3T6 cells. Cells were subcultured in 35-mm dishes in serum-free medium (see “Experimental Procedures”). After 24 h, the medium was removed, and 1.0 ml of serum-free medium containing 125 p~ [3H]adenosine (4.0 Ci/mmol) was added. The percentage of adenosine remaining in the medium was determined at the times indicated in the presence of dipyridamole (A) or NBTI (0)or in the absence of the compounds (0).Adenosine in the medium was separated from inosine, hypoxanthine, and other catabolites using thinlayer chromatography on polyethyleneimine-coatedcellulose plates (see “Experimental Procedures”). 25 0

12.5

50

100 125 [ N B T I ] nM

..

I

1000

5000

FIG. 3. Effect of adenosine transport inhibitors on 3T6 cell growth in presence of AMP. Cells were subcultured in 35-mm dishes in serum-free medium (see “Experimental Procedures”); and 16 h later, dipyridamole (upper) or S-(p-nitrobenzyl)-6-thioinosine (lower)was added at theindicated concentrations. After 5 min, AMP (X) was added to a final concentration of 125 pM (upper) or 250 p M (lower),whereas other cultures served as controls (0).Cell number was determined 120 h after theaddition of AMP and is expressed as a percentage relative to untreated controls.

of AMP (Fig. 7). Further support for the idea that thetoxicity of extracellular adenine nucleotides is mediated by elevation of intracellular ATP and ADP levels came from experiments with a 3T6 cell mutant deficient in cellular adenosine kinase activity (obtained from Howard Green, Department of Physiology and Biophysics, Harvard University Medical School).

The addition of up to 200 p~ ATP, ADP, or AMP to this adenosine kinase-deficient mutanthad no effect on cell growth in medium containing 5% FBS (data not shown). DISCUSSION

The purpose of this study was to examine the mechanism by whichextracellular adenine nucleotides inhibit the growth of 3T6 mouse fibroblasts in physiological media. The results indicate that the toxic effects of ATP, ADP, and AMP on 3T6 cells are mediated by their common catabolite, adenosine, even though the cells are less sensitive to equimolar levels of adenosine added directly to the medium. The results also suggest that adenosine derived from extracellular adenine nucleotide hydrolysis induces growth inhibition by elevating intracellular adenine nucleotide levels and reducing intracellular UTP levels. One plausible explanation for the greater toxicity of aden-

Growth Inhibition by Extracellular Adenine

Nucleotides

12371

TABLEI1 Effect of AMP on intracellular kvels of ATP, ADP, and UTP Cells were subcultured in 35-mm dishes in DMEM containing heat-treated 5% FBS (see “Experimental Procedures”). After 24 h, uridine (10 p ~ and ) AMP (125 PM) were added, whereas other dishes received only AMP (125 p ~ ) After . an additional 48 h, soluble pools were obtained from the cells by extraction with ice-cold 7% (w/v) trichloroacetic acid. Intracellular levels of ATP, ADP, and UTPwere determined after separation by anion-exchange chromatography (see “Experimental Procedures”). The data are expressed as the mean f S.D. of results from four experiments. Additions

0

I

I

I

8

16

24

32

- 1

72

None 125 pM AMP 125 pM AMP UM uridine

+ 10

UTP ADP

ATP

nmol/lo‘ cells

nmol/lO’ cells

nmol/lO‘ cells

0.88.2 f 0.2 1.4 f 0.2 0.9 f 0.2

0.8 0.8 8.1 f 1.4

3 f 0.3 1.8 f 0.3 3 f 0.3

& 15 &

Nucleotide Exposure Time(hr)

FIG.5. Effect of limited exposure to ATP and AMP on 3T6 cell growth. Cells were subcultured in 35-mm dishes in DMEM containing 5% FBS (X) or 5% FBS heat-treated for 20 h at 56 “C (A) or inserum-free medium containing epidermal growth factor, insulin, and transferrin ( 0 )(see “Experimental Procedures”). After 16 h, 125 p~ ATP ( X ) or AMP (A, 0 ) was added; and at the indicated times, the mediumwas aspirated and replaced by conditioned medium obtained from cultures grown for identical lengths of time in the absence of nucleotides. Cell number was determined 72 h after the addition of nucleotides and is expressed as a percentage relative to untreated controls.

0 - “

j -x

X

I

50 [Uridine]

pM

FIG.7. Effect of uridine on 3T6 cell growth in.presence of AMP. Cells were subcultured in 35-mm dishes in serum-free medium; and after 16 h, uridine was added at the indicated concentrations. After 5 min, AMP (0)was added to a final concentration of 125 p ~ whereas othercultures served as controls (X). Cell number was determined 120 h after the addition of AMP and is expressed as a percentage relative to untreated controls.

0

10

20

30

40

50

TOTALCONCENTRATION

60

70

80

(yM)

FIG.6. Effect of multiple doses of adenosine on 3T6 cell growth in serum-free medium. Cells were subcultured in 35-mm dishes in serum-free medium (see “Experimental Procedures”). After 16 h, some cells were treated with one dose of adenosine (ADO; +) or AMP (0)at theindicated concentrations. Other cells received doses of0.5-10 p~ adenosine every 1 (m) or 2 (0)h to obtain the total adenosine concentrations as indicated after 16 h. Cell number was determined 120 h after the first additions were made.

osine derived from adenine nucleotide hydrolysis as compared to adenosine added directly is that hydrolysis of adenine nucleotides maintained adenosine in the medium for 16 h, whereas adenosine added directly was completely metabolized in 8 h (Fig. 4 and Table I). A 16-24-h exposure period to AMP or ATP was needed to observe cytotoxicity (Fig. 5). Furthermore, 80 ~ L Madenosine, a concentration that was not toxic if added in one dose, significantly inhibited the growth of 3T6 cells when added in multiple doses of 5.0-10.0 p~ over a 16-h period (Fig. 6). The finding that AMP was a less effective growth inhibitor in serum-containing medium as compared t o serum-free medium (Figs. 1 and 2) was likely due to the fact that, in the presence of serum, hydrolysis of AMP to adenosine occurred more rapidly (Table I),and significant adenosine levels were not maintained in the me-

dium over a 16-h period. Similarly, ATP was presumably a more effective growth inhibitor thanADP and AMP in serumcontaining medium (Figs. 1 and 2) since hydrolysis of ATP to adenosine was slower than hydrolysis of ADP and AMP (Table I), and significant adenosine levelswere generated from ATP over a longer period of time. Adenosine, transported into cells via an adenosine translocator, is metabolized by several enzymes, including adenosine kinase and adenosine deaminase (24). Adenosine kinase, the enzyme which presumably mediates the growth inhibitory effects of adenosine in 3T6 cells by catalyzing the formation of intracellular adenine nucleotides (Table 11), is substrateinhibited at concentrations exceeding several micromolar (25, 26). Thus, the low toxicity of adenosine concentrations higher than 125 p~ is likely due to substrate inhibition of adenosine kinase as well as to the rapid deamination of adenosine by adenosine deaminase, which exhibits a50-100-fold higher K, for adenosine (24). In contrast,serum-and cell-mediated hydrolysis of extracellular adenine nucleotides resulted in the continuous generation of micromolar concentrations of adenosine which would bepreferentially phosphorylated. Another contributory factor to the greater toxicity of adenosine generated from adenine nucleotides as compared to higher concentrations of adenosine added in one dose may be that cellsurface enzymes generate adenosine from adenine nucleotides in close proximity tothe plasma membrane adenosine translocator. This hypothesis is supported by recent studies (27, 28) showing that higher levels of labeled adenosine were

,

12372

Growth Inhibition Extrac by

found in myocardial cells incubated with [3H]AMPthan with [3H]adenosine, results suggesting that the ecto-5’-nucleotidase is physically or functionally coupled to the adenosine translocator. Our conclusions differ from those of otherstudies that postulate a direct role for adenine nucleotides in cell growth inhibition. Studies (3, 4) with human adenocarcinoma cells suggested that ATP crossed the plasma membrane and directly alteredintracellularadenineand uridine nucleotide levels, thereby affecting DNA synthesis. Although ATP uptake systems have been postulated (29), no direct evidence for such a translocator exists, and the plasma membrane is generally considered to be impermeant to nucleotides under physiological conditions (9). Another study (7) with Friend erythroleukemia cells suggested that inhibition of cell growth by ATP was mediated through a reduction in the plasma membrane potential due to activation of transmembrane ion channels. Similar effects on membrane potentialand ion fluxes were induced by ATP in 3T6 cells incubated in an isotonic buffer low in divalent cations (13, 30). These effects led to an increase in the plasma membrane permeability to nucleotides and other normally impermeant molecules. However, alterations in intracellular Na+ andK’ levels and plasma membrane potential in 3T6 cells incubated in growth medium in the presence of extracellular adenine nucleotides were not detected over a 24-h period (data not shown), and the conditions used in thisstudy were inhibitory to ATP-induced channel formation in 3T6 cells (13,30-34). Acknowledgments-We thank R. Brucker, Juliette Charles, Sharon Johnston, and John Yuan for technical assistance and Vicki Shaff and Lori Gherke for secretarial assistance. We are also grateful to Dr. Leon A. Heppel (Cornel1 University) and Dr. Michael Newman (Roche Institute) for valuable advice on this project.

REFERENCES 1. Gregory, S.,and Kern, M. (1978) Biochem. Biophys. Res. Commun. 83,1111-1116 2. Ikehara, S., Pahwa, R. N., Lunzer, D. G., Good, R.A., and Modak, M. J. (1981) J. Zmmunol. 1 2 7 , 1834-1838 3. Rapaport, E. (1983) J. Cell. Physiol. 1 1 4 , 279-283 4. Rapaport, E., Fishman, R. F., and Gercel, C. (1983) Cancer Res. 43,4402-4406 5. Weisman, G. A., and De, B. K. (1984) Fed. Proc. 43,520 6. Belzer, I,, and Friedberg, I. (1984) Zsr. J. Med.Sci. 2 0 , 483 (abstr.)

:ellularAdenine Nucleotides 7. Chahwala, S. B., and Cantley, L. C. (1984) J. Biol. Chem. 269, 13717-13722 8. Burnstock, G. (1978) in Cell Membrane Receptors for Drugs and Hormones: A Multidisciplinary Approach (Smith, R. W., and Bolis, L., eds) pp. 107-118, Raven Press, New York 9. Burnstock, G., and Brown, C. M. (1981) in Receptors and Recognition (Burnstock, G., ed) pp. 1-44, Chapman Press, New York 10. Gordon, J. L. (1986) Biochem. J. 2 3 3 , 309-319 11. Todaro, G. J., and Green, H. (1963) J. Cell Biol. 17,299-313 12. Freshney, R. I. (1983) Culture of Animal Cells: A Manual of Basic Techniques, pp. 99-118, Alan R. Liss, Inc., New York 13. Weisman, G. A., De, B. K., Friedberg, I., Pritchard, R. S., and Heppel, L. A. (1984) J. Cell. Physiol. 119,211-219 14. Magnusson, R. P., Portis, A. R., Jr., and McCarty, R. E. (1976) Anal. Bwchem. 72,653-657 15. Henderson, J. F., and Scott, F. W. (1981) Pharmucol. Ther. 8, 539-571 16. Fox, I. H., and Kelley, W.N. (1978) Annu. Rev. Biochem. 4 7 , 655-686 17. Ishii, K., and Green, H.(1973) J. Cell Science 13,429-439 18. Aronow, B., Allen, K., Patrick, J., and Ullman, B. (1985) J. Biol. Chem. 260,6226-6233 19. Plagemann, P. G. W., and Woehlheuter, R. M. (1980) Curr. Top. Membr. Tramp. 14,225-330 20. Paterson, A. R. P., Kim, S. C., Bernard, O., and Cass, C. E. (1975) Ann. N. Y. Acad. Sci. 266,402-410 21. Planet, G., and Fox, L. H. (1977) Clin. Res. 2 5 , 323 22. Snyder, F. F., and Seegmiller, J. E. (1976) FEBS Lett. 66, 102106 23. Berlin, R. D., and Oliver, J. M. (1975) Znt. Rev. Cytol. 4 2 , 287336 24. Lum, C . T.,Marz, R., Plagemann, P. G., and Wohlhueter, R. M. (1979) J. Cell. Physwl. 1 0 1 , 173-200 25. Archer, S.,Juranka, P. F., Ho, J. H., and Chan, V.-L. (1985) J. Cell. Physiol. 1 2 4 , 226-232 26. Yamada, Y., Goto, H., and Ogasawara, N. (1980) Biochim. Biophys. Acta 6 1 6 , 199-207 27. Frick, G. P., and Lowenstein, J. M.(1978) J. Biol. Chem. 253, 1240-1244 28. Sasaki, T., Abe, A., and Sakagami, T. (1983) J. Biol. Chem. 268, 6947-6951 29. Chaudry, I. (1982) Yale J. Biol. Med. 65.1-10 30. Dicker, P., Heppel, L. A., and Rozengurt, E. (1980) Proc. Natl. Acad. Sci. U. S. A . 77,2103-2107 31. Rozengurt, E., and Heppel, L. A. (1975) Biochem. Biophys. Res. Commun. 67,1581-1588 32. Rozengurt, E., Heppel, L. A., and Friedberg, I. (1977) J. Biol. Chem. 252,4584-4590 33. Rozengurt, E., and Heppel, L. A. (1979) J. Biol. Chem. 264,708714 34. De, B. K., and Weisman, G . A. (1984) Biochim. Biophys. Acta 776,381-388