G) 1991 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 266 .... acetylcholine, and others (Apple and Barefoot, 1988; Weiss and Atlas ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY G) 1991 by The American Society for Biochemistry and Molecular Biology, Inc
Vol. 266, No. 27, Issue of September 25, pp. 17990-17994, 1991 Printed in U.S.A.
ATP Receptor A PUTATIVERECEPTOR-OPERATEDCHANNELINPC-12CELLS* (Received for publication, March 28, 1991)
Dalit Sela, Efrat Ram, and Daphne Atlas From the Department of Biological Chemistry and TheOtto Loewi Center for Neurobiology, The Hebrew University of Jerusalem, Jerusalem 91904, &rue1
ExternalATPinduces [3H]dopamine ([3H]DA) re- ulator of various cellular responses in a wide variety of excitlease inrat pheochromocytoma cells (PC-12 cells). The able and nonexcitable cells. Burnstock (1978) proposed that ATP-induced release is a saturable process with half- ATP exerts its effects via a specific receptor, the purinergic effective concentration of EC5O = 80 KM. ADP is a poor (P2)receptor, which is distinct from the adenosine(PI)recepsecretagogue of [3H]DA (one-sixth of ATP) and AMP tor. is devoid of secretory capabilities. Adenosine and the Since then, many studieshave confirmed an ATP-selective non-hydrolyzableanalogues of ATP, AppNHp and action mediated via a unique ATP receptor (for review see AppCp are ineffective as inducers of [3H]DA, release, Burnstock, 1978, 1979; Gordon, 1986; Bean and Friel, 1990). or as inhibitors of the ATP-induced r3H]DA release. The wide variety of functions attributed to ATP receptor are: The most potent antagonist of ATP-induced release is calcium entry via calcium-specific channels in smoothmuscle Coomassie Blue (ICs0 = 2 5 p ~ ) ,compared to ADPBS cells (Benham et al., 1986; Benham and Tsien, 1986), activa(ICs0 = 500 KM). The overall rank orderof potency is tion of Ca2+-dependent K+ channelsinperitoneal mouse ATP > ADP >> AMP > adenosine, which is charactermacrophages (Hara et al., 1990), increase in intracellular istic of the P2-purinergic receptor. ATP-induced secretion is absolutely Ca2+ dependent, calcium ([Ca”];) in peritoneal mouse macrophages (Sung et cells indicating anexocytotic process and is independent of al., 1985; Greenberg et al., 1988) in Erhrlich acites tumor Mg2+(up to2 mM) suggesting that the activespecies is (Dubyak and De Young, 1985), increase in inositol trisphosnot ATP4-. (a)The ATP-induced45Ca2+influx into the phate in rat hepatocytes (Charest et al., 1985), and prostacycells is ingood correlation to ATP induction of release clin formation in adrenal medullary endothelial cells (Fors(ICs0 = 80 and 90 PM, respectively) andis carried over berg et al., 1986). There is no direct evidence for the role of to ADP which has a diminished ability to induce both ATP-activatedchannels in neuronal cells, although some release and 45Ca2+influx. ( b )Divalent cations (Ba2+> studies in dorsal horn neurons (Fyffe and Perl, 1984; Jahr and Jessell,1983) and sensory neurons in the caudate trigemSr2+> Ln3+ > Mn2+)replace Ca2+ and support ATPinal nucleus (Salt and Hill, 1983) have shown excitation by induced release similar to their effectiveness in supporting bradykinin- and K+ (50 mM)-induced release externally applied ATP. It appears that ATP stimulates a in PC-12cells (Weiss, C., Sela, D., and Atlas, D. (1990) family of relatedCa2+channels via several ATP receptor Neurosci. Lett,119,241-245). Combined together the subtypes. absolute requirementof [Ca2+Iexfor release, inhibition In dorsal root ganglion cells, the activation of a calcium of release byGd3+(ICs0= 100 MM), Ni2+,and Co2+(ICs0 current by ATP (1.5 FM) is rapid with a time constant of 70 = 1 mM), and support of release by Ba2+, Sr2+,and ms and even faster in frog arterial cells (Bean, 1989). The Mn2+, we suggest that ATP induces Ca2+ entry via rapid ATP kinetics strongly suggest that ATP binding to a ligand-operated Ca2+ channels as previously suggested specific receptor is coupled directlyto opening of Ca2+ channel for ATP in smooth muscle cells (Benham,C. D., Bolton, and most likely not via a second messengersystem (Bean and T.B., Byren, N. G., and Large,W. A. (1987)J.Physiol. Friel, 1990). (Lond.) 387, 473-488). No significant inhibition by 1 Pheochromocytoma cells (PC-12) which can be induced to I . ~ Mverapamil, 10 I.IM nifedipine, or 2 mM Cd2+argues release dopamineandnoradrenaline by differentsecretaagainstATPactivation of voltage-dependentCa2+ gogues suchas high K+, bradykinin,nicotine, muscarinic channels as similarly shown for ATP-induced [3H]noracetylcholine, and others (Apple and Barefoot, 1988; Weiss adrenalinerelease (Inoue, K., Nakazawa, K., Fujiand Atlas, 1991) were recently stimulated by ATP to release moro, K., and Takanaka, A. (1989) Neurosci. Lett. [3H]n~radrenalineby activating voltage-insensitive Ca2+ 106, 294-299).Thus,the widely distributedATP channels (Inoue et al., 1989). In other reports the response to Ca2+ receptor might play an essential role inhomeostaexternal ATP in PC-12 cells was suggested to activate both sis of the cell by introducing Ca2+ into the cell via voltage-sensitive calcium channels and other cation channels specific ligand-gatedCa2+channels. (Fasolato et al., 1990). In addition, direct electrophysiological studies suggested ATP-operated Ca2+ channels in PC-12 cells (Nakazawa et al., 1990). In this report we explore the evidence to support the sugATP released from cells was suggested as a possible modgestion that ATPis a receptor-operated channel,by correlatstudies of ATP-induced 45Ca2+influx and ATP-induced * This work was partly funded by the Charles E. Smith Family ing [“IDA release in PC-12 cells. Foundation Laboratory for collaborative research in Psychobiology. T h e costs of publication of this article were defrayed in part by the payment of pagecharges. Thisarticlemusttherefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
EXPERIMENTAL PROCEDURES
Cell Growth-Cells were obtained from Dr. R. Stein of the TelAviv University, Israel. Growth medium consisted of Dulbecco’s
17990
Channel in PC-12 Cells ATP Receptor-Operated Ca'+
17991
modified Eagle's medium with high glucose (GIBCO) supplemented with 10% horse serum, 5% fetal calf serum (Bet Haemek, Israel),130 E units/ml penicillin, and 0.1 mg/ml streptomycin. % For experiments, cellswereremoved from the flask with 1 mM 4 EDTA, washed and plated on collagen-coated 12-well Costar plates, w and assayed when the cells reached a density of 0.5-1.0 X lO'/well. dcz Assay of t'H]DA Release-Release was carried out essentially as previously described (Weiss and Atlas, 1991). Cells were incubated for 1.5 h at 37 "C with %ml of growth medium and 0.85 pl/ml ['HI DA (41 Ci/mmol), followed by extensive washings with medium (3 X 1 ml) and buffer A (5 X 1 ml). Buffer A consisted of (mM): 130NaC1, 6 KC1, 0.8 MgSO,, 24 NaHCOi1, 1 NaH2PO4, 10 glucose, 1 ascorhic acid, and 1.8 CaCl,, exceptinexperimentstestinginhibition by divalent cations where NaH2P04 andNaHCO:, were replaced by 25 mM HEPES,pH 7.4. In a typical experiment, cells were incubated with 0.5 ml of buffer A for five consecutive incubation periods of 3 min each a t 37 "C. 0 10'~ 10.~ 10" Spontaneous ["HIDA release wasmeasured by collecting the medium [ATPI, M released by cells successively for X2 3-min intervals prior to the third FIG. 1. ATP-induced [3H]DA release. Cells grown on monoperiod, where the buffer included ATP (or any other stimulant as layers were preincubated with [3H]DA for 60 min a t 37 "C followed indicated), in order to assess the induced release. Two post-stimula- by washings in release buffer (see "Experimental Procedures"). After tion periods ensured that thecells returned to their resting state. The 2 X 3-min interval of spontaneous release, cells were stimulated for remaining ["HIDA was extracted from the cells by acidic extraction a 3-min stimulation period by ATP at increasing concentration as overnight with 0.5 ml of 0.1 N HCI. indicated. The results represent net fractional release of 3min of Data Analysis of Release Experiments-["HIDA release during the stimulation after subtracting basalrelease, and each point represents 3-min intervals was expressed as a percentage of the total tritium the meank S.E. of three independent experiments. Inset, each column content of the cells. Net evoked release was calculated from ["HIDA represents release of ["HIDA from thecells during a 3-min incubation released during the stimulation period after subtracting basal ["HI period. The first two periods represent basal fractional release, the DA release in the preceding, base-line period (Schwartz and Atlas, third is the stimulation period with the indicated ATP-concentration, 1989; Weiss and Atlas, 1991). and the last two reflect return to basalrelease. Each period represents "Cd+ Influ~-~'Ca~+ influx in PC-12 cells was assayed as follows. the mean I?r S.E. of two to three experiments. Net fractional release Cells were preincubated for 1 h and ['Hladenine (1pCi/ml), and then is calculated by subtracting the base-line release, as represented by cells were washed three times with 1-ml aliquots of buffer B (mM): fractional release in period two, from stimulated release in the third 108 NaC1,0.5 CaC12,4.7 KCI, 1.2 MgS04, 10 glucose,and 25 HEPES,' period. pH 7.4.45CaC1 (2pCi/ml) was addedtothe cells for 3 min of incubation at 37 "C with additions as indicated. The "Ca2+ flux was terminated by extensivewashing of the cells with ice-cold 0.1 M MgSO, solution. The washed cells were extracted with 0.1% sodium dodecyl sulfate and counted. Correction for the number of cells was made by counting the ["Hladenine uptake during the preincubation period. Materials--[~"H]DA (41 Ci/mmol), "CaC1 (10 mCi/ml), and ["HI adenine (1 mCi/ml) were purchasedfromDuPont-NewEngland Nuclear, ATP and its analogues, CTP,collagen type VII, and verapami1 were purchased from Sigma. ADPDS was from Fluka and Coomassie Blue G was from Merck. Nifedipine was a gift from Bayer (West Germany). All salts were of analytical grade. All buffers were prepared from deionized distilled water. 0' 0.0
RESULTS
Dose Response of ATP andADP-induced PHIDAReleaseATP and ADP stimulated["HIDA secretion in PC-12cells in a dose-dependent manner, as measured from 0.02-1 mM. A typical pattern of release is presented inFig. 1 (inset), where spontaneous release is monitored for 2 X 3 min followed by 3 min of stimulation and two other periods of monitoring the return to basal level. The half-maximal effective concentration (EC5")for induction of release by ATP was 80 ~ L M(Fig. 1) and exceeded 2 mM for ADP(datanotshown).Basal release of ["HIDA, measured from unstimulated cells, was consistent across the experiments withaverage values of 0.52.0%net fractional release. Calcium Dependence of ATP-induced Release-Calcium dependence of ATP (0.5 mM)-induced ["IDA release was determined asa function of [Ca2+Iex added to release the buffer, (data not shown). Absolutely no release was observed when Ca2+was omitted from the buffer. (EGTA was not added, and no effort was taken toremove residualCa2+from the medium.)
0.4
0.8
1.2
1.6
2.0
I
[Mg '+I, m M
FIG. 2. Mg2+ dependence of ATP-induced [3H]DA release. Cells were preloaded with ["HIDA (60 min a t 37 "C) and washed extensively with releasebuffer (see "Experimental Procedures"). Basal release was monitored, followed by stimulation with ATP (0.2 mM). MgClz attheindicatedconcentrations was present in the medium before stimulation ( 2 X 3 min) and during the stimulation period (3 min X 1).The results represent net fractional release of 3 min of stimulation with MgCl,, concentration as indicated, and 1.8 mMCa". Each value isthemean ? S.E. of threeindependent experiments.
The Ca'+ dependence of the ATP-inducedrelease showsmaximal release already at 1.0 mM Ca2+. Wehave used physiological Ca2+ concentration(1.8 mM) throughout this study. Mg2+Dependence of ATP-induced Release-Increasing concentrations of MgC12 (0.2-2 mM) in the release buffer did not affect the ATP (500 pM)-induced release (Fig. 2). This result strongly argues against the involvement of the Mg'+-free ATP form as the activespecies. Induction of r3H]DA Release by Various Ligands-As illusI The abbreviations used are: HEPES 4-(2-hydroxyetbyl)-l-pipertrated in Fig. 3 only ATP, and toa smaller extent also ADP, azineethanesulfonic acid; EGTA, [ethylenebis(oxyethylenenitrilo)] induced ["HIDA release in PC-12cells. 5-AMP was ineffective tetraacetic acid AppCp, p,y-methylene ATP;ADPPS, adenosine-5'(6-thi0)diphosphate; [Ca"],, intracellular calcium, [Ca2+Iex, extracel- (up to5 mM) and so was AppCp (1.0 mM), a non-hydrolyzable lular calcium; IP:
