Extracellular ATP activates polyphosphoinositide breakdown and Ca2+ mobilization in Ehrlich ascites tumor cells

ARCHIVES

OF BIOCHEMISTRY

Vol. 245, No. 1, February

AND

BIOPHYSICS

15, pp. 84-95,1986

Extracellular ATP Activates Polyphosphoinositide Breakdown and Ca2+ Mobilization in Ehrlich Ascites Tumor Cells GEORGE

R. DUBYAK

Department of Biochemistry and Biaphysics, University of Pennsylvania Scholl of Medicine, Philadelphia, Pennsylvania 1910.4 Received

June

7,1985,

and in revised

form

October

9,1985

The effects of extracellular ATP on phosphoinositide metabolism and intracellular Ca2+ homeostasis were studied in Ehrlich ascites tumor cells. Cytosolic [Ca2+] was measured using either quin 2 or the recently described indicator fura 2. Addition of 0.5-25 PM extracellular ATP to intact cells results in a rapid mobilization of Ca2+ from a nonmitochondrial, intracellular Ca2+ store. Likewise, direct addition of 0.2-2 PM myo1,4,5-inositol trisphosphate (IP,) to digitonin-permeabilized Ehrlich cells induces a rapid and reversible release of Ca2+ from a nonmitochondrial pool. Under the same conditions which facilitate intracellular Ca2+ mobilization, extracellular ATP also triggers a rapid breakdown of phosphatidylinositol4,5-bisphosphate (PtdIns(4,5)P2) and accumulation of IPB. A maximal 18% decrease of the polyphosphoinositide is observed 40-60 s after the addition of 25 I.IM ATP; within 5 min PtdIns(4,5)Pz returns to or exceeds the original, prestimulus level. These conditions also trigger a rapid accumulation of phosphatidic acid (1.7-fold increase within 5 min). Paralleling these ATP-induced changes in phospholipid levels is a substantial accumulation of the mono-, bis-, and trisphosphate derivatives of inositol; most significantly, a 2-fold increase in the IP3 level is observed within 30 s after ATP addition. These results suggest that in these tumor cells, extracellular ATP elicits changes in phosphoinositide metabolism similar to those produced by a wide variety of Ca2+ -mobilizing hormones and growth factors. Q 19% Academic PEW IW.

An increasing number of studies have described a variety of actions of extracellular ATP on the function of intact cells (l-9). Most of these effects involve some perturbation of plasma membrane permeability and function. Recently we have demonstrated that treatment of Ehrlich ascites tumor cells with micromolar concentrations of extracellular ATP induces a rapid and transient increase in cytosolic [Ca2+] by simultaneously mobilizing a nonmitochondrial pool of intracellular Ca2+ and selectively increasing the Ca” permeability of the plasma membrane (10). Similar studies describing intracellular Ca2+ mobilization by extracellular ATP have been independently reported by two groups working with isolated hepatocytes (11, 12). These observations suggest that, 0003-9861/86 Copyright All rights

$3.00

Q 1986 by Academic Press, Inc. of reproduction in any form reserved.

in certain cell types, extracellular ATP may act in a manner similar to other Ca2+-mobilizing agents, i.e., by activating breakdown of phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2)l into diacylglycerol and myo-1,4,5-inositol trisphosphate (13). This particular inositol phosphate has been shown to induce Ca2+ release from non’ Abbreviations used: PtdIns, phosphatidylinositol; PtdIns(4)P, phosphatidylinositol 4-phosphate; PtdIns(4,5)P,, phosphatidylinositol 4,5-bisphosphate; Ptd(OH), phosphatidic acid; IPI, inositol trisphosphate; Hepes, N-2-hydroxyethylpiperazine-M-2-ethanesulfonic acid; EGTA, ethylene glycol his@-aminoethyl ether) N,N’-tetraacetic acid; BSS, balanced salt solution; FCCP, carbonyl cyanide ptrifluoromethoxyphenylhydrazone; TPEN, tetrakis(Z-pyridylmethyl)ethylenediamine. 84

PHOSPHOINOSITIDE

BREAKDOWN

TRIGGERED

mitochondrial pools in several cell types (14, 15). We report here that treatment of Ehrlich ascites tumor cells with micromolar concentrations of extracellular ATP does induce a rapid breakdown of PtdIns(4,5)P, and a corresponding accumulation of IPB and phosphatidic acid. Consistent with these latter findings is the demonstration that digitonin-permeabilized Ehrlich cells contain Ca2+ stores which can be rapidly mobilized by direct addition of myo-1,4,5inositol trisphosphate. METHODS C&s. Ehrlich ascites tumor cells (hyperdiploid strain) were cultured using serial intraperitoneal transplantation into 25- to 30-g male Swiss Webster mice. Cells were harvested 6-S days after transplantation. All studies were performed using a balanced salt solution (BSS) containing 120 mM NaCl, 5 mM KCl, 1.5 mM MgClz, 1 mM CaCl,, 25 mM N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (Hepes), and 10 mM n-glucose (adjusted to pH 7.40 with concentrated NaOH). CaClz was omitted from the medium in certain studies as indicated. Cells (approximately 2.5 g wet wt) obtained from one mouse were suspended in 50 ml ice-cold BSS and pelleted by centrifugation (Beckman J-6B) at 250g for 2 min; this was repeated three times. The washed cells were gently resuspended to approximately 5 x lo6 cells/ml and stored on ice. Cytosolic [Ca”‘] wwasurements using quin 2 and fura 2. Portions of the stock cell suspension were preincubated for 10 min at 37°C. Aliquots of quin 2 AM or fura 2 AM dissolved in DMSO were added to final concentrations of 10 and 1 PM, respectively. The suspension was incubated for 10 min at 37°C. The cells were then washed with fresh BSS and incubated for an additional 10 min at 3’7°C to allow complete hydrolysis of the entrapped esters. Loading was then terminated by a fivefold dilution of the cells with icecold BSS followed by centrifugation (2 min at 250g). The cell pellet was resuspended in BSS to 2-5 X lo6 cells/ml and stored on ice, Fluorescence was monitored using a fluorometer designed by the Johnson Foundation Biomedical Instrumentation Group as previously described (10). All intracellular [Ca”] experiments were performed at 37°C with constant mixing. Cytosolic Ca2+ levels were calculated from the quin 2 or fura 2 signals using a previously described calibration protocol (16) subsequent to permeabilization of the cells with 25 pg/ml digitonin. Cafi homeost.a&s in permeabilized Ehrlich Zeus Cells were washed and resuspended (5 X 106/ml) in a medium containing 130 mM KCI, 1 mM MgClz, 0.1 mM

BY EXTRACELLULAR

EGTA, and 25 mrd K-Hepes

ATP

(pH 7.1). Digitonin

85 (50

pg/ml) was added and the cells were incubated at 37°C for 5 min. These conditions have been shown to effect selective permeabilization of the plasma membranes of most cells (17). The cells were pelleted and washed three times with the same medium omitting the EGTA and digitonin; the permeabilized cells, resuspended to a final concentration of 5 X lO’/ml, were stored on ice. Ca2+ homeostasis in suspensions of permeabilized cells was measured using a previously described Ca2+selective electrode and calibration technique (18, 19). All measurements were made at 35°C. Cells were treated with rotenone (2 wg/ml) and oligomycin (2 rg/ml) prior to the addition of mitochondrial substrate (8 mM Tris-succinate) and an ATP-regenerating system (2 mM MgATP, 8 mM phosphocreatine, and 5 units/ml creatine phosphokinase). Myoinositol-trisphosphate, purchased from Sigma Chemical Company (St. Louis, MO.), was used without further purification or analysis; it was assumed that this bovine brainderived fraction is predominantly composed of the 1,4,5-isomer.

Phosphoinositide labeling. Ehrlich cell phosphoinositides were isotopically

labeled using three different

protocols: (1) Phospholipids were q-labeled by incubating cells (5 X 106/ml) in BSS containing 0.5 mM NaPi and 10 &i/ml “Pi (New England Nuclear) for 2 h at 37°C prior to experimental perturbation and extraction. (2) Phosphoinositides were 3H-labeled in vitro by incubation of the cells (lO?/ml) in BSS containing 106 &i/ml rH]myoinositol (Amersham) for 2 h at 37°C. The cells were washed three times in 50 vol of BSS and then incubated (2 X lo6 cells/ml) in BSS containing 10 mM unlabeled myoinositol for 30 min at 37°C. The cells were then washed two times in 20 vol of BSS, resuspended in BSS to a final concentration of 2 X lo6 cells/ml, and stored on ice prior to the experiments. (3) Phosphoinositides were 3Hlabeled in wivo by injecting 100 &i [3H]myoinositol into the peritoneum of a mouse with a 7-day-old tumor. After 18 h, the mouse was sacrificed and the Ehrlich cells were isolated and washed as previously described. These cells were then resuspended (2 X 10’1 ml) in BSS and incubated for 15 min at 37°C prior to lipid extraction. While the polyphosphoinositides labeled by the in viva procedure had a specific activity approximately lo-fold lower than that observed in lipids extracted from cells labeled in vitro, the yield of labeled cells was 100 times greater using the in tivo technique. Phosptirwsitide extraction and analysis. For analysis of phospholipid content, 100~~1aliquots of =P- or 3H-labeled cells were rapidly mixed with 500 ~1 of chloroform:methanol:concentrated HCl (200:100:1). After 15 min the extract was centrifuged for 1 min at 8000g; the chloroform phase was removed and stored while the protein interface and aqueous phase

86

GEORGE

were reextracted for 15 min with 100 ~1 of the acidified chloroform:methanol. After centrifugation, the secondary chloroform phase was pooled with the initial extract; the pooled extract was evaporated to dryness and the residue was dissolved in 35 ~1 of chloroform: methanol (21). Aliquots (25 ~1) were spotted onto oxalate-impregnated silica gel plates (plastic-hacked) which were then developed using a previously described mobile phase (20) consisting of chloroform: acetone:methanol:acetic acid:HaO (160:60:52:48:32). Phosphoinositides and other phospholipids were identified by comparison to the R, values observed with authentic lipid standards. a?- and 3H-labeled phospholipids were also identified using conventional autoradiographic techniques (21); TLC plates containing ‘H-phospholipids were sprayed with En‘Hance (New England Nuclear) prior to autoradiography. Radioactive spots on the TLC plates were cut out, placed in Formula 963 (New England Nuclear), and counted by liquid scintillation chromatography. Inositol phosphate w&action and analysis. Cells labeled for either 2 h in vitro or 18 h in viva with [3H]myoinositol were used for analysis of 3H-labeled inositol phosphates. After labeling and washing, the cell suspensions were preincubated with 10 mM LiCl for 10 min prior to addition of ATP. For extraction of these sugars, 500-~1 aliquots of cell suspension were extracted with 1.9 ml of chloroform:methanol:concentrated HCl (2OO:lOO:l). After 10 min, 0.7 ml chloroform and 0.7 ml Ha0 were added to the extracts, which were then centrifuged at 25009 for 10 min. The top aqueous phase (1.5 ml), mixed with 2.5 ml of 10 mM sodium borate, was then stored on ice until analysis. Extracts were subjected to ion-exchange chromatography using Dowex 1 anion-exchange resin in the formate form (1 ml packed resin/column). The various inositol phosphates were separated using a previously described elution protocol (22). The radioactivity of the various elution fractions (mixed with Formula 963) was then quantified by liquid scintillation counting. Nucleotide labeling and wzeaeuremen& Intracellular nucleotides were extracted by rapid addition of 100 pl of cell suspension to 100 pl 1 N perchloric acid/l rnrd EDTA and kept at 0°C for 30 min. The protein was removed by centrifugation and 150 pl of the supernatant was neutralized with ‘75 ~1 1 N KOH. Adenine nucleotides were separated and quantitated as previously described (23) using a Waters HPLC system employing a 5-Mm Cl8 reverse-phase column (radial compression). In samples from =P-labeled cells, additional lo-r1 aliquots of the extracts were spotted onto polyethylenimine-cellulose thin-layer plates, which were then developed with 0.5 M KHaPO,. The developed plates were subjected to autoradiography to identify the [-r-“PIATP spots. These spots were cut out and analyzed by liquid scintillation counting.

R. DUBYAK RESULTS

Mobilization of Intracellular Ca2’ Stores in Intact and Permeabilixed Ehrlich Tumor Cells As previously described (lo), treatment of intact Ehrlich tumor cells with micromolar concentrations of extracellular ATP induces a rapid mobilization of intracellular Ca2+ stores While this initial study was performed using the Ca2+ indicator quin 2 (37), subsequent experiments have employed fura 2, a recently described Ca2+indicator with a number of greatly improved fluorescence and chelation properties (38). These properties facilitate the reliable measurement of cytosolic [Ca”‘] using much lower concentrations of entrapped indicator. Typically, measurements with equivalent signal/noise can be obtained using cells loaded with either 0.05-0.1 mM fura 2 or l-2 mM quin 2. Both indicators have relatively high affinity for Ca2’ and will act as potent Ca2+ buffers. The effects of such exogenous buffering on cytosolic [Ca2+] transients will be exacerbated at high indicator concentrations, i.e., in quin 2-loaded cells. The net effects of such buffering will include both a blunting and a slowing of any given change in [Ca2+]. Another problem which complicates the use of quin 2 has recently been reported by Arslan et al. (16), who demonstrated that the calculated, cytosolic Ca2+ levels in quin 2-loaded Ehrlich (and other) cells are artifactually low due to appreciable quenching of the entrapped indicator by endogenous Zn2+. These investigators also showed that treatment of these cells with tetrakis(2-pyridylmethyl)ethylenediamine (TPEN), a lipid-soluble, heavy-metal chelator, strips the Zn2+ from the intracellular quin 2 and permits calculation of basal [Ca2+] in the range 120150 nM. Figure 1 illustrates the Ca2+ transients elicited by 25 PM extracellular ATP in Ehrlich cells loaded with either 0.1 mM fura 2 (Fig. 1A) or 1.7 mM quin 2 (Fig. 1B). When fura 2-loaded cells were incubated in the absence of extracellular calcium, addition of ATP produced a very rapid (within 5 s),

PHOSPHOINOSITIDE

BREAKDOWN

TRIGGERED

BY EXTRACELLULAR

ATP

87

In fura 2-loaded cells incubated in the presence of 1 mM extracellular Ca2+, 25 FM 7 2016 ATP elicited a rapid, lo- to 15-fold increase -60 u” 696 I mM [Cop*],,, 522in cytosolic [Ca”‘] similar to that observed g rE -60 g = 336. in the absence of extracellular [Ca2+]. This 2242 0.1 pM [ co2+],,, 1493 G transient increase was followed, however, rkIzrt by a decay toward a [Ca2+] 2- to 3-fold 25 ,uM ATP higher than the basal level; this “plateau” phase was maintained for 3-4 min and was followed by a very gradual (complete within 10 min) decline to the original basal level. Likewise, in quin 2-loaded cells, addition of ATP in the presence of extracellular Ca2+ led to an initial, rapid phase of increase identical to that observed in Ca2+free saline. This was followed by a second, 25 ,dd ATP slower phase of increase (Zmin duration) and then a gradual (lo-min duration) reFIG. 1. Mobilization of intracellular Ca*+ in intact turn to basal values. As previously deEhrlich tumor cells by extracellular ATP. (A) Fura scribed (lo), only the secondary phase of e-loaded Ehrlich cells (4 X lo6 cells/ml; 0.1 nmol fura 2/106 cells) suspended in basal salt solution, were the ATP-induced [Ca”+] transients in quin preincubated at 3’7°C for 5 min prior to the addition 2-loaded Ehrlich cells is dependent on exof 25 PM ATP. In a second aliquot of cells, the extratracellular [Ca2+] and can be inhibited by cellular [Ca*+] was reduced to 0.1 pM by the addition 25 FM LaCla. Similar results were observed of 2 ITIM EGTA 1 min prior to addition of the ATP. in the fura 2-loaded cells (data not shown). (B) Same experimental conditions as in A, but using Figure 2 reveals that similar threshold and quin 2-loaded cells (3.4 X lo6 cells/ml; 1.7 nmol quin maximally active ATP concentrations 2/106 cells). characterize the ATP-induced [Ca”‘] transients in fura 2- and quin 2-loaded Ehrlich lo- to 15-fold increase in cytosolic [Ca2+]; cells. Thus, while the kinetic profiles of the ATP-induced [Ca2’] transients are different this increase was followed by a rapid decay in quin 2- and fura S-loaded Ehrlich cells, (complete within 60 s) to the basal level. experiments with the latter indicator reinSimilarly, when quin 2-loaded cells were force the fundamental conclusions of our exposed to ATP in the absence of extraearlier studies with quin 2-loaded cells (10); cellular calcium, a rapid (complete within of extra10 s), 4- to 6-fold increase in cytosolic [Ca”‘] i.e., micromolar concentrations cellular ATP elicit increased cytosolic was observed. However, the subsequent [Ca2+] by activation of plasma membrane decay to the basal level was greatly slowed and was completed only after 3-4 min. Ca2+ channels and by mobilization of nonmitochondrial pool(s) of intracellular Ca2+. Furthermore, comparison of the calculated, basal Ca2+ levels in quin 2- vs fura 2-loaded Previous studies have demonstrated the Ehrlich cells reveals values in the range presence of IPs-releasable Ca2+ pools in he120-150 nM in the latter cells as opposed patocytes (14) and neutrophils (15) which to apparent values in the range 50-80 nM were permeabilized by treatment with digin the quin 2-loaded cells. The use of fura itonin or saponin. Similar experiments 2, with its considerably lower affinity for with digitonin-permeabilized Ehrlich tuheavy metals (38), does not appear to be mor cells (Fig. 2) revealed that these cells complicated by artifactual quenching by also possess Ca2’ pools which may be reendogenous heavy metals, since similar versibly released by exposure to submibasal [Ca”‘] values were calculated in both cromolar levels of myo-1,4,5-inositol triscontrol and TPEN-treated cells (data not phosphate. When the initial [IP,] > 1 PM shown). (Fig. 3A), the [Ca2+] transients were charA. 0.1 mM

Fura 2

-100

58 ,‘o : ?a o=o: C-P :s 9 n -.

A.

[W].nM

[Fum2]=O.ImM -

- 2016

I\

t t J-1 - 896 - 522 - 336

224 149 96

t

[ATP]+M:

60s

0.5

25

t 25.0

5.0

B. [ouinZ]rl.smM

. - 115 - 76

t t

Y[ATp],pM:

J 1035 460 268 172

-I 50

t

5.0

2.5

0.5

25.0

FIG. 2. Dose dependence characterizing ATP-induced Ca2+ mobilization in Ehrlich cells. (A) Fura 2-loaded cells (4 X lo6 cells/ml; 0.1 nmol fura 2110’ cells) suspended in basal salt solution containing 1 mM CaClz were treated with the indicated concentrations of extracellular ATP. (B) Same experimental conditions as in A, but using quin 2-loaded cells (5 X lo6 cells/ml; 1.5 nmol quin 2110’ cells).

PC0

A.

-8mM

Succinatr

4.8

I

5. 1

5 units/ml

Crsotine

Kinore

I

C. peok PC0 5.90

PC0 5.90 5.98

6.05 6.20 6. 12

emin

i_ =oo

I

2

3

4

5

bsh FIG. 3. Mobilization of intracellular Caz+ stores in permeabilized Ehrlich tumor cells by inositol 1,4,5_trisphosphate. Ehrlich cells were permeabilized with digitonin as described under Methods. A 250-pl aliquot of the cells (5 X lOr/ml) was then added to the thermostated (35°C) Caz+ electrode chamber. (A) Cells were pretreated with rotenone (2 ag/ml) and oligomycin (2 pg/ml), and then mitochondrial Ca2’ uptake was initiated by addition of Tris-succinate. After a steady-state pCa was reached, the suspension was supplemented with the indicated concentrationsef creatine kinase, phosphocreatine, and MgATP. After the second steady-state pCa was achieved, various concentrations of myo-1,4,5-inositol trisphosphate (IP,) were added at the indicated times. (B) Same conditions as in A, but the cells were pulsed with different concentrations of IPr. (C) The peak pCa values of the IPs-induced transients are plotted as a function of the initial IPs concentrations. Similar results were obtained with three additional cell preparations. 88

PHOSPHOINOSITIDE

BREAKDOWN

TRIGGERED

acterized by a rapid release, complete within 5 s, followed by a gradual (several minutes duration) re-uptake. No release of Cazf is observed when IP3 (up to 20 PM) was added either to intact cells or to permeabilized cells incubated in the absence of MgATP (data not shown). The postrelease re-uptake process also required the presence of MgATP and was inhibited (>90%)by 1 mM vanadate (data not shown). While the [Ca2’] transients induced by [IP,] > 2 PM were characterized by identical peak pCa (Fig. 3A), the re-uptake process following addition of 5 PM IP3 was considerably slower than that observed with 2 PM IP3. This most likely reflects the kinetics of the phosphatase reaction responsible for IP3 degradation (24). The similarity between the rapid mobilization of Ca2+ elicited either by ATP in intact cells or by IPB in permeabilized cells strongly suggests that, in intact cells, extracellular ATP triggers the accumulation of intracellular IP3. This, in turn, implies that extracellular ATP induces rapid activation of phosphoinositide metabolism.

30

60

90 120 Minutes

150

BY

EXTRACELLULAR

Phosphoinositide Tumor Cells

ATP

89

Levels in Ehrlich

Prior to investigating the effects of extracellular ATP on phosphoinositide metabolism, it was important to determine the resting levels of the major phosphoinositides in Ehrlich cells. Figure 4 shows the time course of =P incorporation into Ehrlich cell ATP and phospholipids. Throughout the labeling, the total adenine nucleotide pools (Fig. 4A) were maintained into at constant levels. 32P incorporation ATP (Fig. 4B) was characterized by a rapid, initial phase during the first 60 min followed by a much slower rate of increase over the next 2 h; the rapid, initial phase indicates that the y-P of ATP reaches near constant specific radioactivity within the first 60-90 min. Rapid, initial rates also 32P incorporation into the characterize polyphosphoinositides and phosphatidic acid (Fig. 4D). Conversely, 32P incorporation into PtdIns and the other bulk phospholipids was characterized by a slow phase during the initial 60 min, followed

30

60

90 120 Minutes

150

FIG. 4. Time course of ?P incorporation into the ATP and phospholipid cells. Ehrlich cells (4 X lo6 cells/ml) were suspended in a balanced salt NaPi and 10 &i/ml “pi. At the indicated times after addition of the “Pi, for extraction and analysis of nucleotides and phospholipids, as described total adenine nucleotide content of these cells during the labeling. (B) cellular ATP pool. (C, D) mP incorporation into cellular phospholipids. average + range of duplicate determinations.

pools of Ehrlich tumor solution containing 0.5 mM lOO-~1 samples were taken under Methods. (A) The 8zp incorporation into the Each point represents the

90

GEORGE

by faster rates of incorporation over the subsequent 2 h. These results indicate that the monoester phosphate groups of the polyphosphoinositides rapidly exchange with the 7-P of ATP, while the diester phosphate groups of all three phosphoinositides (and all other phospholipids) presumably would reach isotopic steady state only after many hours. While the results detailed in Fig. 4 indicate that the monoester phosphates of the polyphosphoinositides exchange much more rapidly than diester phosphate, no information regarding the relative specific activities of the 4- and 5-phosphate moieties can be inferred from such data. Hawkins et al. (39) have demonstrated that 32P is unequa 11y distributed between the 4and 5-phosphates of PtdIns(4,5)Pa extracted from erythrocytes incubated with 32P. for 2 h. For this reason, it is not possibie to calculate the relative masses of the polyphosphoinositides from such shortterm 32P-labeling experiments. Long-term in vitro incubation of cells with [3H]inositol will label the phosphoinositides of most cells to near constant specific activity; in cultured GH3 cells (40) or smooth muscle cells (27), 24-36 h are required to achieve steady state. In GH3 cells (41), 95% of the 3H incorporated into the lipid fraction is found in PtdIns, PtdIns(4)P, PtdIns(4,5)P2, and lyso-PtdIns. In the present study, Ehrlich cells were isolated 16-18 h after injection of donor mice with [3H]inositol. Under these conditions, 84% of the 3H associated with the cellular phospholipids was incorporated into PtdIns, 2% into lyso-PtdIns, 5% into PtdIns(4)P, and 8% into PtdIns(4,5)P2.

ATP-Induced Breakdown of Ehrlich Tumor Cell Polyphosphoinositid Treatment of 32P-labeled Ehrlich cells with 25 PM extracellular ATP (Fig. 5) induced a rapid breakdown of PtdIns(4,5)P2 during the first 60 s after ATP addition; only a very small decrease in PtdIns(4)P levels was observed during this time. Between 2 and 5 min after ATP addition,

R. DUBYAK

h 367 x

.c Qo11 seconds

60 post

120 addition

IS0

240

300 I

of 25 $4 ATP

FIG. 5. Time course of polyphosphoinositide breakdown and phosphatidie acid accumulation induced by treatment of Ehrlich tumor cells with extracellular ATP. Ehrlich cells (5.5 X 106/ml) were prelabeled with =Pi (10 &i/ml) for 2 h. The cells were then exposed to 25 pM ATP, lOO-~1 aliquots were taken for extraction at the indicated times. Phospholipids were separated and quantified as described under Methods. Each point represents the average + range of duplicate determinations.

PtdIns(4,5)P2 returned to the original basal level. In parallel with the transient breakdown of PtdIns(4,5)P2, ATP also induced rapid accumulation of phosphatidic acid. Consolidation of the data from four cell preparations revealed that treatment of Ehrlich cells with 25 PM ATP for 60 s produced an 18.2 f 3.0% decrease (mean f. SE) in PtdIns(4,5)Pz, no significant changes in PtdIns(4)P, and a 19.0 +- 11.1% increase in phosphatidic acid. After 5 min, the polyphosphoinositides returned to or slightly overshot the original basal levels (4 f 1.8% increase in PtdIns(4,5)P2; 12.4 f 3.3% increase in PtdIns(4)P), while the phosphatidic acid content was increased by 72.0 f 15.9%.

ATP-Induced Accumulation of In&to1 Phosphates in Ehrlich Tumor Cells Figure 6 shows the time course of accumulation of inositol trisphosphate and inositol bisphosphate after treatment of

PHOSPHOINOSITIDE 5OOr

BREAKDOWN

TRIGGERED

1P,

400

300

200 i 5

100 E

seconds

post addition

of 25pM

ATP

FIG. 6. Time course of inositol tris- and bisphosphate accumulation induced by treatment of Ehrlich tumor cells with extracellular ATP. Ehrlich cells were prelabeled with [‘Hlmyoinositol (100 &i/ml) for either 2 h in vitro (m) or 18 h in viva (0). After washing and preincubation with 10 mM LiCl for 10 min, the cells, either 2 x 106/ml@) or 2 X lO’/ml (a). were exposed to 25 pM ATP; 500-~1 aliquots were taken for analysis at the indicated times. Inositol phosphates were extracted, separated, and quantified as described under Methods. Each point represents the average i range of values determined from the two separate experiments.

Ebrlich tumor cells with 25 @M extracellular ATP. Within 15 s, a 50-70s increase was observed in both IPa and IPB levels. The content of IP3 was nearly maximal within 30 s after ATP addition, while the IP2 content peaks at about 60 s. Nearly identical rates of IPz and IPa accumulation were observed regardless of whether cells labeled in vivo on in vitro are used. Consolidation of the data from nine separate preparations of 3H-labeled cells (six labeled in vitro, three in vivo) demonstrated that 25 PM ATP produced a 10’7 & 23% increase in IPB levels within 30 s. As demonstrated in our previous study (lo), half-maximal mobilization of intracellular Ca2+ was triggered by 2.5 PM extracellular ATP, while 5-7 PM ATP elicited maximal Ca2+ mobilization. Thus, the standard ATP concentration (25 PM) used

BY

EXTRACELLULAR

ATP

91

in the measurements of phosphatidylinositol turnover and IP3 production was approximately five times that required for maximal Ca2+ mobilization. In Fig. ‘7, the dose-response relationship characterizing ATP-induced IP3 accumulation is compared with that characterizing ATP-induced mobilization of Ca2+ in fura 2-loaded cells. While a typical sigmoidal curve characterized the Ca2+ mobilization triggered by 0.5-25 PM ATP, no saturation of the stimulated IPa accumulation was observed in this same concentration range. These data strongly suggest that the doseresponse curve describing the effects of ATP on IP3 accumulation is shifted considerably to the right of that characterizing ATP-induced Ca2’ mobilization. DISCUSSION

Treatment of Ehrlich ascites tumor cells loaded with the Cazt indicators quin 2 (37)

450Y F-7

i2’00 z

FIG. 7. Comparison of dose-response relationships characterizing ATP-induced Ca” mobilization and ATP-induced IP, accumulation. Ebrlich cells were prelabeled with [aH]myoinositol for 18 h in zrivo. After washing and preincubation at 37°C with 10 mM LiCl for 10 min, the cells (2 X 107/ml) were exposed to the indicated concentrations of extracellular ATP. After 30 s, 500-4 aliquots were taken for extraction and analysis of IPa content. Each point (0) represents the mean (*SE) of five determinations from two separate experiments. Also plotted are the peak cytosolic (M) [Ca”] values measured during the ATP-induced [Caz’] transients in fura Z-loaded cells as illustrated in Fig. 2A.

92

GEORGE R. DUBYAK

and fura 2 (38) with low concentrations of extracellular ATP induces a rapid mobilization of intracellular Ca2+ followed by increased entry of Ca” across the plasma membrane (Figs. 1 and 2 and Ref. (10)). The results of several experiments described in the present study indicate that the fundamental mechanism underlying this action of ATP is similar to that which mediates the effects of more conventional Ca*+-mobilizing agents, i.e., a rapid stimulation of phosphoinositide turnover, a corresponding accumulation of IPa, and finally, an IPa-induced release of Ca2+ from intracellular stores. Direct treatment of digitonin-permeabilized Ehrlich cells with 1,4,5-inositol trisphosphate induces a rapid release of Ca2+ from a nonmitochondrial compartment (Figs. 3A, B). The dose response characterizing this IPs-induced release (Fig, 3C) is similar to that observed with permeabilized neutrophils (15) and hepatocytes (14), i.e., half-maximal release at about 400 nM IPB. The IPa-triggered Ca” release and the subsequent re-uptake phase follow approximately the same time course as that which describes the cytosolic [Ca2’] transients induced by extracellular ATP in intact cells (Fig. 1). Addition of 25 PM ATP to intact cells also triggers a rapid, but transient, breakdown of PtdIns(4,5)P2 and a sustained accumulation of phosphatidic acid (Fig. 5). The results of previous studies characterizing phosphoinositide turnover indicate that there are substantial quantitative and qualitative differences in the responses of different cell types to various Ca2+-mobilizing agents. In hepatocytes (25) and adrenal glomerulosa cells (26), the respective responses to vasopressin and angiotensin II are characterized by very substantial (30-50%) and sustained (32 min in duration) decreases in the levels of both PtdIns(4)P and PtdIns(4,5)P2. Conversely, in cultured arterial smooth muscle cells (27), treatment with angiotensin II produces only asmall (lO%)and transient (~20 see in duration) decrease in PtdIns(4,5)P2, followed by a sustained 15% overshoot of the prestimulus level; the PtdIns(4)P levels

show only a monophasic increase after angiotensin II exposure. This diversity of cellular response probably results from the fact that most Ca2+-mobilizing agonists stimulate not only the breakdown of phosphoinositides, but also their resynthesis (13, 28, 29). These relative rates of breakdown and synthesis may vary substantially between cell types. The transient nature of the ATP-induced decrease in the PtdIns(4,5)P, levels and the increased labeling of PtdIns(4)P suggest that ATP is also activating both the breakdown and the synthesis of the polyphosphoinositides in Ehrlich tumor cells. In addition to decreasing Ehrlich cell polyphosphoinositide levels, extracellular ATP also induces a corresponding accumulation of the phospholipid-derived breakdown products, viz., the inositol phosphates (Fig. 6). The relative time courses characterizing accumulation of the bis- and trisphosphate esters indicate that the trisphosphate form accumulated somewhat faster than the bisphosphate derivative; this suggests, but does not prove, that ATP is primarily enhancing breakdown of the PtdIns(4,5)P2 precursor. Similar time courses characterize the accumulation of the inositol phosphates in other cell types treated with various Ca2’mobilizing hormones (26, 27, 30). The exact isomeric form of the IP3 accumulated in Ehrlich cells at various times after ATP stimulation remains to be determined. Very recently, Irvine et al. (42) demonstrated that the 1,4,5-isomer was the predominant species accumulated in rat parotid glands only during the first 5 s after carbachol stimulation. Thereafter, the 1,3,4-isomer of IPB rose rapidly while the 1,4,5-IP3 levels dropped; 1,3,4-IP3 was eventually accumulated to levels lo-20 times that of the 1,4,5-isomer. These findings are very significant, since only the 1,4,5-isomer of IPB has definitively been shown to induce Ca2’ release of intracellular stores (14,15). Accumulation of diacylglycerol, the other product of polyphosphoinositide breakdown, was not directly monitored in this study, but a rapid and sustained ac-

PHOSPHOINOSITIDE

BREAKDOWN

TRIGGERED

cumulation of phosphatidic acid was observed in the ATP-treated cells (Fig. 5). While phosphatidic acid may be derived from de nova synthesis or from direct hydrolysis of phospholipids, a substantial portion is also formed by phosphorylation of accumulated diacylglycerol via the action of diacylglycerol kinase (31). This possibility is supported by the observation that the rapid phase of phosphatidic acid accumulation coincides with the period of transient PtdIns(4,5)P, breakdown (Fig. 5). A general feature of receptors which are coupled to the inositol phospholipid signaling system is that the agonist concentrations required to maximally activate PtdIns(4,5)Pz breakdown and IP3 release are invariably higher than those required for maximal stimulation of Ca2+ mobilization and Ca2+-dependent cellular functions (30,43-45). Recently, Lynch et al. (46) have compared the relationship among receptor binding capacity, IP3 accumulation, Ca2+ mobilization, and phosphorylase activation for three receptor classes (&i-adrenergic, vasopressin, and angiotensin II) in rat hepatocytes. In all cases, the dose response for agonist-induced IP3 accumulation was shifted to the right of the Ca2+ mobilization curve. The smallest shift (approximately 0.5 log unit) was observed with norepinephrine, while the largest disparity (1.5 log units) was found with vasopressin. The degree of disparity could be directly correlated with the number of the particular receptor since al-adrenergic receptors had the smallest binding capacity, while the vasopressin receptors had the highest capacity. A similar disparity in dose-response relationships appears to characterize the capacity of extracellular ATP to induce IPS accumulation and Ca2+ mobilization in Ehrlich cells (Fig. 7). While a complete dose response for ATP-induced IP3 release was not obtained, the present data strongly suggest that this dose-response curve is shifted, at the very least, 0.75 log unit to the right of the Ca2’ mobilization curve. This observation further underscores the similarity between the actions of extracellular ATP in Ehrlich cells and the actions of well-characterized ag-

BY

EXTRACELLULAR

ATP

93

onists for receptors coupled to the inositol phospholipid signaling cascade. While these various results strongly suggest that the Ca2+-mobilizing action of ATP is mediated via activation of phospholipase C, there remain at least three critical questions concerning the observed actions of extracellular ATP in Ehrlich tumor cells. First, is phospholipase C activated subsequent to the occupation of specific ATP receptors on the surface membrane? Second, are similar effects of extracellular ATP seen in other cell types? Third, do the observed effects of ATP have physiological or pathological significance? While the ATP-induced effects on Ehrlich cell phosphoinositol turnover and Ca2+ homeostasis appear very similar to those induced by the interaction of various agents with cell-surface receptors, we have at present no evidence for or against the presence of specific ATP receptors on the surface membrane of these cells. Burnstock (32, 33) has postulated the existence of purinergic receptors, while Cockcroft and Gomperts (34) have argued that the multiple effects of ATP on mast cell function may be mediated by a common receptor for the free acid form of ATP. As noted in our previous study (lo), the purine nucleotide, ITP, and more significantly the pyrimidine nucleotide, UTP, are as effective as ATP in eliciting Ca2+ mobilization in Ehrlich cells. Conversely, adenosine, AMP, ADP, and nonhydrolyzable ATP analogs (in the range 0.2-200 FM) do not trigger Ca2+ mobilization. These data indicate that if a putative receptor for ATP exists on the surface membrane of these cells, it does not fit the description of the so-called purinergic receptors. Effects of extracellular ATP on Ca2+ mobilization and/or phosphoinositide metabolism have recently been characterized in at least three other cell types. Watts and Borovay (35) reported that both ATP and the nonhydrolyzable analog /3-y-methylene ATP stimulated phosphatidylinositol turnover and tension development in isolated arterial smooth muscle. Haynes et al. (11) found that extracellular ADP and ATP (0.1-l PM) and AMP (l-100 pM) trigger

94

GEORGE

rapid elevation of cytosolic [Ca’+] in rat hepatocytes. Similar effects in rat hepatocytes have also been observed by Charest et al. (12), who additionally reported that adenine nucleotides increase the cellular levels of inositol trisphosphate. In both studies, the observed responsiveness of hepatocytes to ADP and AMP, as well as ATP, suggests the presence of receptors which fit the Pa-purinergic classification of Burnstock (32, 33). The earlier work of Creba et al. (45) on the rapid breakdown of PtdIns(4)P and PtdIns(4,5)Pz in rat hepatocytes included ATP (along with vasopressin, angiotensin, and cul-adrenergic agents) among the effective agonists. In our own preliminary studies, we have not observed Ca’+-mobilizing effects of extracellular ATP in murine and rat thymocytes or murine S49 lymphoma cells. Conversely, in the DDTl cell line derived from a leiomyosarcoma of hamster vas deferens (36), we (Dubyak and Reynolds, unpublished data) have characterized effects of extracellular ATP and UTP on Ca” mobilization which are very similar to those observed in Ehrlich ascites tumor cells. Furthermore, in these latter cells, the Ca2+-mobilizing activity of extracellular ATP is very similar to that produced by occupation of ai-adrenergic receptors (47). Thus, while it is evident that extracellular ATP can activate Ca2+ mobilization (and presumably phosphoinositide turnover) in a number of both normal and transformed cells, further work is required to determine the nature and potential significance of the putative receptor type(s) responsible for these effects. ACKNOWLEDGMENTS Thanks are due to Dr. Antonio Scarpa for his provision of laboratory space and facilities, Francine Stathopulos for technical assistance, and Dan Brannen for preparation of the manuscript. This work was supported in part by NIH Grant HL-15835 to the Pennsylvania Muscle Institute and by Grant IN-135 from the American Cancer Society. REFERENCES 1. SNEDDON, P., AND WESTFALL, sid 347,561-580.

D. P. (1984)

J. Phy-

R. DUBYAK 2. OKADA, Y., YADA, T., OHNO-SHOSAKU, T., OIKI, S., UEDA, S., AND MACHIDA, I. (1984) Exp. Cell Bes. 152,552-557. 3. CHAHWALA, S. R., AND CANTLEY, L. C. (1984) J. Bid Chem 259,1371’7-13722. 4. KRELL, H., ER~~ISCH, N., KASPAREK, S., AND PFAFF, E. (1983) Eur. J. Biochem 131,247-254. 5. KRISHTAL, 0. A., MAREHENKO, S. M., AND PIDOPLICHKO, V. I. (1983) Neurosti Lett. 35,41-45. 6. DEN HERTOG, A. (1982) J. Physiol 325,423-439. 7. VERHAEGE, R. H. (1977) Amer. J. PhysioL 233, H114-H128. 8. AITEN, J. F., AND LAMB, J. F. (1975) J. PhysioL248, 14P-15P. 9. LANDRY, Y., AND LEHNINGER, A. L. (1976) Bi~~herx J. 158.427-438. 10. DUBYAK, G. R., AND DE YOUNG, M. B. (1985) J. Biol. Chem 260,10653-10661. 11. HAYNES, R. C., JR., PICKING, R. A., AND &STARE, F. D. (1985) Fed Proc. 44,483. 12. CHAREST, R., BLACKMORE, P. F., AND EXTON, J. H. (1985) Fed PVJC 44,1473. 13. WILLIAMSON, J. R., COOPER, R. H., JOSEPH, S. K., AND THOMAS, A. P. (1985) Amer. J. Physiol. 248, C203-C216. 14. JOSEPH, S. K., THOMAS, A. P., WILLIAMS, R. J., IRVINE, R. F., AND WILLIAMSON, J. R. (1984) J. Biol Chem. 259,3077-3081. 15. PRENTKI, M., WOLLHEIM, C. B., AND LEW, P. D. (1984) J. BioL Chem 259,13777-13782. 16. ARSLAN, P., DI VIRGILIO, F., BELTRAME, M., TSIEN, R. Y., AND POZZAN, T. (1985) J. Bid Chem 260, 2719-2727. 17. FISKUM, G. (1985) Cell Calcium 6,25-37. 18. AFFOLTER, H., AND SIGEL, E. (1979) And Biock 97,315-319. 19. BERS, D. M. (1982) Amer. J. Physiol. 242, C404(3408. 20. JOLLES, J., WIRY, K. W. A., SCHOTMAN, P., AND GISPEN, W. H. (1979) FEBSL&. 105,110-114. 21. LASKEY, R. A., AND MILLS, A. D. (1977) FEBSLett. 82,314-316. 22. WATSON, S. P., MCCONNELL, R. T., AND LAPETINA, E. G. (1984) J. Bid Chem 259,13199-13203. 23. DUBYAK, G. R., AND SCARPA, A. (1983) Biochmnistry 22,3531-3536. 24. JOSEPH, S. K., AND WILLIAMS, R. J. (1985) FEBS Lett. 180,150-154. 25. THOMAS, A. P., MARKS, J. S., COLL, K. E., AND WILLIAMSON, J. R. (1983) J. Biol. Chew. 258, 57165725. 26. KOJIMA, I., KOJIMA, K., KREUTTER, D., AND RASMUSSEN, H. (1984) J. BioL Chin. 259, 1444814457. 27. SMITH, J. B., SMITH, L., BROWN, R. E., BARNES, D., SABIR, M. A., DAVIS, J. S., AND FARESE, R. V. (1984) Proc. Nat1 ACUQ! Sci USA 81.7812-7816.

PHOSPHOINOSITIDE 28.

MICHELL,

BREAKDOWN

B&him

R. H. (1975)

Biuphys.

TRIGGERED

Acta 415,

39.

81-147. 29.

KIRK,

C. J., MICHELL,

THOMAS,

31.

J. R. (1984) J. Bid KANOH, H., KONDOH,

A. P., ALEXANDER,

Chem.

32. 33.

BURNSTOCK, BURNSTOCK,

34.

COCKCROFT,

35.

WATTS,

Biochem

D. A. (1981)

R. H., AND HEMS,

40.

REBECCHI, GORN,

41.

IMAI,

J. 194,X5-165.

30.

J., AND

Chem H.,

WILLIAMSON,

AND

ONO,

T. (1983)

J.

258,1767-1774. G. (1976) J. Theor. Biol. 62,491-503. G. (1981) J. Physiol. 313,1-35. S.,

AND

GOMPERTS,

B.

D.

P. T., MICHELL,

BOROVAY,

RINK,

GRYNKIEWICZ, G., POENIE, M., AND (1985) J. Biol Chem 260.3440-3450.

TSIEN,

KIRK,

C. J.

M. J., KOLESNICK, R.N., AND GERSHENM. C. (1983) J. BioL Chem 258,227-234.

A., AND

GERSHENGORN,

J. Biol

M. C. (1985)

260,10536-10540.

43.

CHAREST, MORE,

(1980)

Proc J. Bid

T. J. (1982)

AND

IRVINE, R. F., ANGGARD, E. E., LETCHER, A. J., AND DOWNES, C. P. (1985) B&hem J. 229,505-

45. J. S. (1982)

R. H.,

J 218,785-793.

42.

44.

Fed

M. A. (1983)

95

ATP

511.

J. M&789-798.

R. P., AND

EXTRACELLULAR

Biochem

Chem

259,5574-5584.

42,724. 36. CORNETT, L. E., AND NORRIS, Chem. 257,694-697. 37. TSIEN, R. Y., POZZAN, T., AND Cell Biol 94,325-334. 38.

HAWKINS, (1984)

Biochem

Biol.

BY

R., PRPIC, P. F. (1985)

R. Y.

J. H., AND

Biochem.

J. 227,79-90.

COCKCROFT, S., BARROWMAN, PERTS, B. D. (1985) FEBS CREBA,

J. A.,

BREWSTER, J.

V., EXTON,

DOWNES,

M.

lktt.

C. P.,

G., MICHELL,

M.,

BLACK-

AND

GOM-

l&259-263. HAWKINS,

R. H., AND

P. T., KIRK,

C. J.

(1983) B&hem. J. 212,733-747. 46. LYNCH, C. J., BLACKMORE, P. F., CHAREST, R., AND EXTON, J. H. (1985) MoL Pha ?-?mcol 28.93-99. 47. REYNOLDS, E. E., AND DUBYAK, G. R. (1985)

Biochem

Biophys.

Res. Commun

130,627-632.