Expression and Cyclic AMP-dependent Regulation of a High Affinity ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY

VOl. 266, No. 24, Issue of August 25, pp. 15750-15757, 1991 Printed in U.S.A.

c 1991by The American Society for Biochemistry and Molecular Biology, Inc.

Expression and Cyclic AMP-dependent Regulation aofHigh Affinity Serotonin Transporterin the Human Placental Choriocarcinoma Cell Line (JAR)* (Received for publication, January 2, 1991)

David R. Cool$, Frederick H. LeibachS, Vinod K.BhallaQ, VirendraB. Maheshp, and Vadivel GanapathyST From the DeDartmentsof tBiochemistrv and Molecular Biology and §Physiology and Endocrinology, Medical College of Georgia, Augusta, Geo’rgia 30912-2100

The JAR human placental choriocarcinoma cell lineshown to be present in the plasma membrane of a few cell transports serotonin, accumulating the monoamine in- types such as the neuron, platelet, and endothelium (1-5). side the cell against a concentrationgradient.The The characteristics of this transporterhave been investigated transport is energized by an NaCl gradient. Tricyclic in detail using plasma membrane vesicles isolated from these (imipramine and desipramine) and non-tricyclic (pacells. Recently we provided evidence for the presence of a high roxetineandfluoxetine)antidepressantsinhibitthe affinity serotonin transporter in the brush-border (maternal transporter markedly, but reserpine and 5-hydroxy- facing) membrane of the human placental syncytiotrophotryptophan do not. Ouabain, gramicidin, and nigericin,blast (6). In general the properties of the placental serotonin which reduce or abolish the transmembrane Na+ gradient, and phloridzin, which interferes with glucose transporter appear to be similar to those of the neuronal and platelet serotonin transporters (6-10). transport into the cells, inhibit the transport. PreinTo date, the characterization of the serotonin transporter cubation of the cells with glucose-free medium also causes similar inhibition. The activity of the serotonin in these cells has focused mainly on the modulation of the transporter in thiscell line is stimulated in response to transporter function by ions and ion gradients, the binding of evaluation overnight (16-h) incubation with increasing concen- various antidepressants tothe transporter, and the of the antidepressant-induced inhibition of the transporter trations of choleratoxin (0.1-1,000 ng/ml).Under these conditions the stimulation is maximal at 10 ng/ activity. A few studies have also been directed toward the goal ml cholera toxin (3.1 2 0.2-fold). Cholera toxin inof solubilization, identification, and purification of the transcreases the cAMP content of these cells by several porter in either the platelet or the neuronal tissue (11-17). hundredfold within 2 h. Isobutylmethylxanthine (100 But the regulatory aspects of the serotonin transporter have MM), dibutyryl cAMP (100 PM), and forskolin (100 PM) received comparatively less attention. Very little information mimic the action of cholera toxin, elicitinga 1.6-2.5is available concerning whether or not hormones and/or infold stimulation of the serotonin transporter activity. tracellular second messengers modulate the function of the The stimulatoryeffect of cholera toxin is antagonized transporter. To our knowledge there is only a single study significantly by simultaneous incubation of the cells with 50 PM N-(2-aminoethyl)-5-isoquinolinesulfon-that hasindicated that protein kinase C (C-kinase)is involved amide, a protein kinase inhibitor. The effect of cholera in the regulation of the serotonin transporter incultured toxin on serotonin transport is specific because, under endothelial cells (18).Even though isolated plasma membrane similar conditions, cholera toxin inhibits 3-0-methyl- vesicles offer a number of advantages to characterize certain D-glUCOSe transportand does not influence taurine aspects of the transporter such as ion requirement, kinetics, transport in this cell line. There is also no significant and interaction with antidepressants, they have limitations change in the protein content of the cells after cholera in studiesinvolving the regulation of the transporterfunction. toxin treatment. Kinetic analysis reveals that cholera An intact cell system is undoubtedly a more suitable experitoxin causes an increase in the maximal velocity (7.89mental tool to investigate the involvement of hormones and f 0.67 to 17.55 f 1.06 pmol/mg of protein/5 min) and intracellular second messengers in the regulation of the transa decrease in the Michaelis-Menten constant (0.52 f porter. In addition to platelets, cultured bovine endothelial 0.09 to 0.29f 0.04 PM). These data show that the JAR cells ( B ) , and a mouse neuroblastoma X rat glioma hybrid human placental choriocarcinomacell line expresses a cell line (20), all of which express a serotonin transporter, high affinity serotonin transporter thatis sensitive to there have been three recent reports of the development of inhibition by antidepressants and that the activityof serotonin-transporting cell lines by specific manipulations the transporteris under CAMP-dependent regulation. (21-23). The mouse hypothalamic cell line F7 and the mouse teratocarcinoma cell line F9K4b2, both transformed by simian virus 40, express a serotonin transporter upon induction of A high affinity serotonin transporter that is sensitive to differentiation (21, 22). The mouse fibroblast L-M cells have inhibition by a selective group of antidepressants has been been reported to express stablya high affinityserotonin transporter after transfectionwith human genomic DNA(23). * This work wassupported by National Institutes of Health Grants However, none of these cell systems has yet been used to HD22103 and HD24451. The costs of publication of this article were study the regulation of the serotonin transporter. defrayed in part by the payment of page charges. This article must We are currently inthe process of evaluating the feasibility therefore be hereby marked “aduertisement” in accordance with 18 of using a cell line derived from human placental choriocarU.S.C. Section 1734 solely to indicate this fact. cinoma (JAR) as a model to probe the regulation of several ll To whom correspondence should be sent.

15750

Cyclic AMP-dependent Regulation of Serotonin Transporter

15751

choriocarcinoma cells was measured by a procedure similar to the one described above for serotonin transport except that the preincubation step involving iproniazid was omitted, and imipramine was not present during the washing of the cells. Protein Determinution-Protein was assayed for each experiment using duplicate dishes cultured concurrently and under conditions identical to those dishes used for serotonin uptake measurements. After removing the culturemedium, deionized water (1 ml) was added to each dish. The dishes were frozen and thawed twice after which the dish contents were suspended using a 1-ml syringe and a 25-gauge needle prior to protein estimation. Protein was determined by the method of Lowry et al. (24) using bovine serumalbumin as the standard. Determination of Cell Number-The cell number was determined for each experiment using triplicate dishes cultured concurrently and EXPERIMENTALPROCEDURES under conditions identical to those dishes used for protein and seroMaterials-5-[1,2-3H-]Hydroxytryptamine (serotonin) binoxalate tonin uptakemeasurements. The cells were released by trypsinization, (specific radioactivity, 30.4 Ci/mmol), 3-O-[methyl-3H]methyl-~-glususpended in phosphate-buffered saline, and the cell number was cose (specific radioactivity, 79.0 Ci/mmol), [carbo~y-'~C]inulin (spe- determined using a Coulter counter. cific radioactivity, 1.8 mCi/g), and [2-'H]taurine (specific radioactivDetermination of Intracellular Water Space-Intracellular water ity, 30.3 Ci/mmol) were purchased from Du Pont-New England space was determined from the equilibrium (60-min incubation) disNuclear. Serotonin, 5-hydroxytryptophan, imipramine, desipramine, tribution of radiolabeled 3-O-methyl-~-glucose (25)aftermaking reserpine,iproniazid, gramicidin, nigericin, choleratoxin,IBMX,' necessary corrections for inulin space. The washing buffer contained forskolin, and dibutyryl cAMP were obtained from Sigma. H-9 dihy- 0.5 mM phloridzin to prevent the efflux of the sugarduring the drochloride was obtained from Research Biochemicals, Inc., Natick, washing procedure. Inulin space was determinedunderidentical MA. Paroxetine was a gift from Beecham Pharmaceuticals, Betch- experimental conditions by using radiolabeled inulin. worth, Surrey, United Kingdom, and fluoxetine was a gift from Lilly. Determination of CAMP-After the desired treatment, cells were RPMI 1640 medium, penicillin-streptomycin, Fungizone (amphoter- scraped off from the dish and transferredalong with the medium into icin B), andfetal bovine serum were obtained from Flow Laboratories, a test tube. The tube was placed immediately in a boiling water bath. Mclean, VA. All other chemicals were of analytical grade. After 3 min the content of the test tube was centrifuged, and the Culture of Human Placental Choriocarcinoma Cells(JAR)-The clear supernatant was collected. The cAMP contentof these samples JAR cell line was obtained from the American Type Culture Collec- was determined by radioimmunoassay as described previously (26). tion and cultured in 75-cm2 Corning culture flasks with RPMI 1640 Measurements were made in triplicate with three or four dilutions of medium containing10%fetal bovine serum, 50 IU/mlpenicillinthe samples to ensure parallelism with the standard curve of the streptomycin, and 0.125 /*g/mlFungizone. The cells were maintained assay. at 37 "C and 10% CO,. Confluentcultures were trypsinized with Statistics-Each experimental point was determined with duplicate phosphate-bufferedsaline containing 0.1% trypsinand 0.25 mM or triplicate dishes, and each experiment was repeated two to five EDTA andsubcultures started from the released cells. For the uptake times. The results are expressed as means & S.E. from these replicate experiments the cells were seeded in 35-mm disposable Petri dishes determinations. Statistical significance was determined by Student's (Falcon) at a densityof approximately 1.5 X lo6cells/dish and allowed t test, anda p value < 0.05 was considered significant. Kinetic analysis to grow as a monolayer. The cells received 2 ml of fresh cultured was done using a computer package Statgraphics (STSC, Rockville, medium 24 h after subculturing and were used the next day. When MD). the effects of cholera toxinand otheragents were tested, fresh culture medium containing these agents was added to the cells exactly 24 h RESULTS after subculturing. The time of incubation with these agents varied depending upon the individual experiment, but in most cases it was Nu+-dependent Serotonin Transport-Fig. 1 describes the 16 h. time-dependent accumulation of serotonin in human placenSerotonin Uptake Measurekent-The monolayer cultures were tal choriocarcinoma cells. When NaCl was present in the incubated for 45 min at room temperature (21-22 "C) prior to the measurement of serotonin uptake with a buffer containing 25 mM Hepes/Tris, 140 mM NaCI, 5.4 mM KC1, 1.8 mM CaCl,,0.8 mM MgS04, 5 mM D-ghCOSe, and 0.1 mM iproniazid, pH 7.5 (uptake buffer). Preincubation with iproniazid was necessary to prevent serotonin metabolism via monoamine oxidase. After preincubation, serotonin uptake was initiated by removing the buffer from the dish and adding 1 ml of the uptake buffer containing radiolabeled serotonin. In most cases the final concentration of serotonin was 50 nM. After incubation for a desired time the buffer was removed, and the / cells were quickly washed four times with the uptakebuffer containing 0.1 mM imipramine. The time taken for washing was always less than 30 s. Inclusion of imipramine in the wash buffer prevented efflux of serotonin from the cells during thewashing procedure. After washing, 1 ml of 0.3 N NaOH was added to each dish and allowed to stand for 1 h, at which time the contents from the dish were transferred to a scintillation vial, and the radioactivity of the contents was measured by liquid scintillation spectrometry. In the experiments determining the effects of Na+ on serotonin uptake, choline chloride was substituted for NaCl in the uptake buffer, and where the effects of anions on the uptake were investigated, the sodium salt of the respective anion was used in the uptake buffer. In experiments involving regulation of serotonin transport the medium was replaced 16 h prior to use with fresh medium containing the appropriate concentration of Incubation time (mid the compound involved. Taurine Uptake Measurement-Transport of taurine into the JAR FIG. 1. Time course of serotonin uptake in human placental choriocarcinoma cells. Uptake of serotonin in choriocarcinoma The abbreviations used are: IBMX, 3-isobutyl-1-methylxanthine; cells was measured from a medium containing either NaCl ( O " 0 ) Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid H-9, N- or choline chloride (o"--o). The concentration of serotonin in the (2-aminoethyl)-5-isoquinolinesulfonamide. uptake medium was 50 nM.

transport systems that have been described in the normal human placenta. The data presented in this paper show that the JAR cell line stably expresses a high affinity serotonin transporter under normal conditions with no experimental manipulation and that thecharacteristics of this transporter are in general similar to those of the platelet,brain, and placental serotonin transporters. In addition we have used this cell line to delineate certain aspects of the regulation of the transporter. We provide evidence here which shows for the first time that the serotonin transporter activity is under CAMP-dependent regulation in thiscell line.

15752

Cyclic AMP-dependent Regulation

of Serotonin Transporter

uptake buffer, uptake of serotonin at a concentration of 50 and nigericin, for this purpose. Preincubation of the cells with nM was very rapid and increased with time. The uptake was these reagents is expected either partially or fully to collapse linear at least up to 20 min. However, when NaCl in the the transmembrane Na' gradient. The results of the experiuptake buffer was replaced with choline chloride, uptake of ment, given in Table I, show that all four reagents caused serotonin was reduced markedly at all time points.The initial marked inhibition of serotonin uptake.Gramicidin and nigeruptake rates measured with a 5-min incubation were only icin, which are ionophores for monovalent cations such as 15% in the presence of choline chloride compared with the Na+ and K', showed the maximal inhibitory activity (about of glucose rates in the presence of NaC1. These data demonstrate clearly 75%). Ouabain inhibitedabout60%.Inhibition the dependence of serotonin uptake on Na+.The presence of entry into thecell by phloridzin, a condition that is expected to result in a reduction of ATP production, also caused a C1- alone was not enough to support the uptake. Intracellular Volume-To assess whetherserotonin was significant inhibition (32%; p < 0.001) in the initial rates of accumulated inside the cells against a concentration gradient serotonin uptake. In addition, removal of glucose from the in thepresence of NaCl we first determined the intracellular medium during preincubation and uptake measurement rewater space. We chose 3-O-methyl-D-glucose,a nonmetabo- duced serotonin uptake by 49%. Taken collectively the data lizable sugar, for this purpose. Uptake of this sugar into the demonstrate that the uphill transport of serotonin in human cells occurs by facilitated diffusion via the glucose transporter, placental choriocarcinoma cells is energized by an Na' graand hence it is not concentrated against a gradient. Thereforedient and thusis dependent on metabolic energy. Anion Specificity-The Na+-dependent serotonin uptake in the intracellular water space can be calculated from the equilibrium distribution of this solute inside the cell and in the platelet (27, 28) and brain (29, 30) plasma membrane vesicles medium. Correction for extracellular space can be made by exhibits an absolute requirement for C1-. Similar resultswere determining inulin space under identical conditions. Initial obtained in our laboratory for the placental serotonin transexperiments showed that uptake of 3-O-methyl-~-glucose intoporter in purified brush-border membrane vesicles (6,8). We choriocarcinoma cells reached equilibrium with a 60-min in- therefore investigated whether the uptake of serotoninin cubation. Under these equilibrium conditions the extracellular human placental choriocarcinoma cells also exhibits similar characteristics with respect to the anion specificity. In these space represented 5.6 & 0.3% of thetotal3-O-methyl-~glucose space (intracellular space + extracellular space). The experiments preincubation as well as uptake measurements intracellular water space in choriocarcinoma cells thus deter- were performed in uptake buffers containing the Na' salt of mined was 4.2 & 0.3 pl/mg of protein or 1.02 ? 0.07 fi1/1O6 different anions. The initial rates (5-min incubation) of secells. With this value for the intracellular water space the rotonin uptake were determined and compared. The results intracellular concentration of serotonin was calculated and are given in Table 11. Serotonin uptake was maximal in the the cel1:medium distribution ratio determined. At a concen- presence ofNaC1. Substitution ofC1- with gluconate, F-, tration of 50 nM the distributionratio for serotonin was SCN-, NO;, or I- drastically reduced the uptake.These greater than 1.0 in the NaC1-containing medium even at as anions supportedthe uptake only to a maximum of 18% when short as 5-min incubation (intracellular concentration, 183 compared with C1-. These data show clearly that theserotonin nM). The ratio increased with time and was 29.8 ? 3.5 at 60 transporter of the choriocarcinoma cells also exhibits an . results show obligatory dependence on C1- for its maximal activity. min (intracellular concentration, 1.5 p ~ )These Effects of Classical Serotonin Uptake Inhibitors-Many that serotonin was actively transportedintothe cell and accumulated against a concentration gradient. The distribu- antidepressants such as imipramine, desipramine, paroxetine, tion ratio was drastically reduced when choline chloride was and fluoxetine act as potentinhibitors of serotonin uptake in present in the uptake medium instead ofNaC1. However, the platelet, the brain, and theplacenta. Therefore we investhere was indication for active transport of serotonin even tigated the effect of these inhibitors on serotonin uptake in from the choline chloride-containing medium because the the JAR human placental choriocarcinoma cell line (Table cel1:medium distribution ratio was 3.8 ? 0.5 at 60-min incu- 111).At a concentrationof 10 p~ the tricyclic antidepressants, imipramine and desipramine, as well as the non-tricyclic bation. Even though the cells were preincubated with choline antidepressants, paroxetine and fluoxetine, inhibited the inichloride medium for 45 min prior to initiation of serotonin tial ratesof serotonin uptake in these cells by about 80%.The uptake, it is possible that these conditions were not enough to remove the endogenous Na+ completely. Traces of endogTABLEI enous Na' leaking out of the cells into the uptake buffer during uptake measurements might support active transport Effect of dissipation of the Na' gradient on the transport of serotonin in human placental choriocarcinoma cells of serotonin into the cells to some extent. Furthermore, we The cells were preincubated in theabsence or presence of inhibitors have shown previously that in human placental brush-border at theindicated concentrations for 30 min in NaC1-containing transmembrane vesicles the Na+-dependent uptakeof serotonin is port buffer. When glucose-free medium was used, the composition of stimulated markedly by extravesicular C1- and also by intra- the medium was maintained as that of the control transport buffer vesicular K+ (6, 8). Since the conditions used for the uptake except that glucose was omitted. After the preincubation the buffer of serotonin into the cells from choline chloride-containing was removed from the dishes and replaced with the transport buffer medium did not alter the concentrations of the intracellular containing 50 nM radiolabeled serotonin and therespective inhibitors. Incubation was carriedout for 5 min after which the cells were K+ and extracellular C1-, these ions might have significantly washed four times with the imipramine-containing buffer, and serostimulated the uptake of serotonin in the presence of traces tonin transport was determined. In the absence of inhibitors, the of extracellular Na+. control transport rate was 0.75 f 0.05 pmol/mg of protein/5 min. The data arefrom three to nine determinations. Effect of Dissipation of the Transmembrane Na' GradientThe role of atransmembraneNa+gradientinserotonin Inhibitor % Inhibition uptake in choriocarcinoma cells was probed further by invesOuabain (2 mM) 60 f 2 tigating the influence of dissipation of the Na' gradient on Gramicidin (30 pg/ml) 75f 1 Nigericin (10 p M ) 73 f 2 the initial rates (5-min incubation) of serotonin uptake. We Phloridzin (2 mM) 32 +. 2 used ouabain, an inhibitor of Na+-K+-ATPase,phloridzin, an Glucose-free medium 49 2 5 inhibitor of glucose transport, andtwo ionophores, gramicidin

Cyclic AMP-dependent Regulation of Serotonin Transporter TABLE I1

15753

ity. Thus, whether cAMP and A-kinase are involved in the regulation of the serotonin transporter in the JAR cell line can be studied by investigating the effect of cholera toxin on serotonin transport. For this purpose, the JAR cells were incubated for 16 h with increasing concentrations of cholera of toxin, from 0 to 1,000 ng/ml, andthetransportrate serotonin in the cells was subsequently determined (Fig. 2). Serotonin transport was found to be stimulated markedly by Anion transport Serotonin % cholera toxin. Maximal stimulation occurred between 10 and pmollmg protein15 min 1,000 ng/ml cholera toxin. Under these conditions, the transChloride 0.65 f 0.04 100 port rate incholera toxin-treated cells was 1.78 f 0.09 pmol/ Iodide 0.12 f 0.02 18 mg of protein/5 min whereas the transport rate in control 0.09 f 0.01 13 Thiocyanate cells was 0.58 f 0.06 pmol/mg of protein/5 min. Thus, cholera 12 0.08 f 0.01 Fluoride toxin at concentration of 10-1,000 ng/ml caused a 3.1 f 0.20.05 f 0.02 7 Gluconate Nitrate 0.04 +. 0.01 6 fold stimulation of serotonin transport activity in these cells. Effects of Forskolin, IBMX, and Dibutyryl CAMPon Serotonin Transport-To substantiate the stimulation of seroTABLE 111 tonin transportby cholera toxin we investigated the influence Inhibition of serotonin transport in human placental choriocarcinoma on serotonin transport in these cells of three other agents cells by antidepressants Initial transport rates(5-min incubation) of serotonin were meas- that are known to elevate intracellular cAMP levels. These ured in choriocarcinoma cells in the presence or absence of various agents areforskolin, an activator of adenylate cyclase; IBMX, inhibitors. The Concentration of serotonin was 50 nM, and thatof the an inhibitor of cAMP degradation by inhibiting cyclic nucleoinhibitors, when present, was 10 FM. In the absence of inhibitors, tide phosphodiesterases; and dibutyryl CAMP, a membranecontrol transport rate was 0.91 f 0.06 pmol/mg of protein/5 min. permeable cAMP derivative. The cells were incubated overThe data arefrom four to six determinations. night (16 h) with these agents, each at a concentration of 100 Inhibitor % Inhibition PM, and theserotonin transport ratewas determined in these cells after the incubation. The effects of theseagents on Imipramine 81 f 1 Desipramine 81 f 1 serotonin transport were compared with that of cholera toxin Paroxetine 85 f 1 (10 ng/ml). The results are given in Fig. 3. Forskolin stimuFluoxetine 78 f 5 lated serotonin transport to approximately the same level as 5-Hydroxytryptophan 1 f 6 cholera toxin (forskolin, 2.5 & 0.2-fold cholera toxin, 2.3 & Reserpine 10 f 6 0.2-fold). IBMX and dibutyryl cAMP also stimulated serotonin transport (1.6 f 0.1-and 1.8 f 0.1-fold, respectively), serotonin transporter of the choriocarcinoma cell thus appears but the stimulation was less compared with the effects of to be similar tothetransporter described in the plasma cholera toxin and forskolin. membrane of the platelet and the brain and in the brushThe length of time necessary for cholera toxin and forskolin border membrane of the placenta with respect to sensitivity to elicit their effects on serotonin transport was determined to various inhibitors. The transporteralso appears to be very by incubating the cells for increasing periods of time in a specific for serotonin because 5-hydroxytryptophan, a precur- medium containing either 10 ng/ml cholera toxin or 100 PM sor for serotonin, failed to inhibit serotonin uptake in these forskolin (Fig. 4).With both compounds a significant stimucells. lation was observed only after a lag period of 8 h. After this Platelets and serotonergic neurons possess two types of time the stimulation increased gradually with time and was serotonin transport mechanisms: one localized in theplasma persistent even after 24 h. membrane, and the otherin the storage granules. These two Effect of CholeraToxin on Cellular CAMP Content-To mechanisms are distinct and differ in many respects. The plasma membrane serotonin transporter is many-fold more sensitive to inhibition by imipramine, desipramine, paroxetine, and fluoxetine than is the storage granule transporter. On the contrary, reserpine is a specific inhibitor of the storage granule serotonin transporter with no or minimal effect on the plasma membrane transporter. In fact, just like imipramine and paroxetine are used as specific, high affinity ligands for the plasma membrane serotonin transporter (3-5), reserpine is used as a specific, high affinity ligand for the transporter in the storage granules (31). We therefore studied the effect of reserpine on serotonin uptake in choriocarcinoma cells. Serotonin uptake in these cells was found to be insensitive to reserpine, thereby indicating that the measured se* rotonin uptake was caused by the action of the reserpinej 0 0.1 1 10 100 1000 insensitive plasma membrane transporter present in these Cholera toxin concentration (ng/mL) cells. FIG. 2. Dose response of cholera toxin-induced stimulation Effects of CholeraToxin on SerotoninUptak-Cholera toxin is known to be responsible for the ADP-ribosylation of of serotonin uptake. The choriocaricinoma cells were incubated the a subunit of the stimulatory G protein (G,) which results overnight (16 h) with varying concentrations of cholera toxin (01,000 ng/ml), after which serotoninuptake into these cells was in the stimulation of adenylate cyclase to generate CAMP. measured from the NaC1-containing medium using a 5-min incubaThe resulting rise in intracellular cAMP levels would then tion. The concentration of serotonin in the uptake medium was 50 potentiate CAMP-dependent protein kinase (A-kinase) activ- nM. Anion specificity for serotonin transport in human placental choriocarcinoma cells Transport buffers were prepared by substituting 140 mM NaCl with an equal concentration of the Na+ salts of different anions. Preincubation as well as transport measurements were performed in these buffers. The concentration of serotonin was 50 nM, and the incubation time was 5 min. The ~ a t are a from four determinations.


15754

TABLEIV 250

-a? - 200

Effect of cholera toxin on cAMP content in human placental choriocarcinoma cells Monolayer cultures were incubated with 1 ml of culture medium with or without 10 ng/ml of cholera toxin at 37 "C for different time intervals. After the incubation, cells were scraped off from the dish and transferred along with the medium into a test tube. The tubewas

A

c

4

150

a

.I

immediately placed in a boiling water bath. After 3 min the content was centrifuged, and the clear supernatant was collected. The concentration of cAMP in the supernatant was determined by radioimmunoassay. Thedata are from three determinations done in quadruplicate.

2 100 2

4

50

Control

Forskolin

Dibutyryl

cAMP Cholera toxin

Incubation time

IBMX

FIG. 3. Effects of forskolin, IBMX, and dibutyryl cAMP on serotonin uptake. The choriocarcinoma cells were incubated overnight (16 h) with or without cholera toxin (10 ng/ml), forskolin (100 p M ) , IBMX (100 pM), or dibutyryl cAMP (100 p M ) . After the incubation,uptake of serotonin into these cells was measuredin the presence of NaC1. Uptake incubation was for5 min, andthe serotonin concentration was50nM. Control uptake was 0.56 f 0.04 pmol/mg of protein/5 min, which was taken as 100%.

FIG. 4. Time dependence of the stimulation of serotonin uptake by cholera toxin and forskolin. Thechoriocarcinoma cells were incubated with or without 10 ng/ml cholera toxin or 100 WM forskolin for varying time periods (4-24 h), after which serotonin uptake into these cells was measured inthe presence of NaC1. Uptake incubation wasfor 5 min, and the serotonin concentration was50 nM. Slashed bars,cholera toxin;cross-hatched bars, forskolin.

h

0.5 1 2 4

cAMP content Control

Cholera toxin pmol/l@ cells

0.34 f 0.02 0.48 rf: 0.05 0.68 f 0.05 0.44 f 0.05

0.39 f 0.05 8.38 rf: 1.02 306.63 f 36.00 185.06 f 20.94 1.ok

C

FIG. 5. Effect of cholera toxin on the uptake of taurine and 3-O-methyl-~-glucoseand on the protein content. The choriocarcinoma cells wereincubatedovernight(16 h) withorwithout cholera toxin (10 ng/ml), after which uptake of taurine (25 nM) and 3-0-methyl-D-glucose (63nM) into these cells as well as the protein content of the cells were measured. The uptake medium contained NaC1, and the uptake incubation was for 3 min.

was found to have no significant effect on taurine transport ( p > 0.6) (Fig. 5 A ) . Under similar conditions, transport of 3substantiate that the stimulatory effect of cholera toxin on 0-methyl-D-glucose also was not stimulated but was rather serotonin uptake was caused by the elevation of cAMP levels significantly ( p < 0.001) inhibited (Fig. 5 B ) . There was also between in these cells we determined the effect of cholera toxin (10 no significant( p > 0.4) change in the protein content ng/ml) on thecellular content of CAMP. The resultsgiven in control and cholera toxin-treated cells (Fig. 5 C ) . These data Table IV show that cholera toxin increased thecellular con- demonstrate that the stimulatory effect of cholera toxin on serotonin transportobserved in the JAR choriocarcinoma cell centration of cAMP by several hundredfold. line was not a general phenomenon common to all transport Effects of Cholera Toxin on Other Transport System-To establish whether or not the cholera toxin-induced stimula- systems in thesecells. Effect of a Protein Kinase Inhibitor on the Cholera Toxintion in this cell line was specific to the serotonin transporter we investigated the influence of cholera toxin (10 ng/ml) on induced Stimulation of Serotonin Transport-Since cholera two other transport systems (taurine and 3-O-methyl-~-gh- toxin and other agents that elevate the intracellular levels of A-kinase is most likely cose) as well as on the protein content of the cells. Taurine is cAMP stimulated serotonin transport, involved in this regulatory process. To test the involvement transported into human placental brush-border membrane vesicles in a manner similar to that of serotonin, requiring a of the A-kinase, we determined whether H-9, an inhibitor of transmembrane NaCl gradient (32, 33). The JAR choriocar- protein kinases witha greater specificity for A-kinaseand G cinoma cells also express a taurine transporter whose char- kinase than for C-kinase (35), was able to antagonize the acteristics are similar to those of the transporter found in the stimulatory effect of cholera toxin (Fig. 6). The JAR chorionormal placenta (34). Thus, taurine transport in this cell line carcinoma cells were incubated overnight with or without presence or absence of H-9, and serotonin occurs by an active mechanism. On the contrary, transport of cholera toxin in the 3-O-methyl-~-glucose in thesecells is mediated by an Na+- transport was subsequently measured in these cells. In the independent glucose transporter, a nonactive transport proc- absence of H-9, cholera toxin at a concentration of 10 ng/ml ess. Exposure of the cells to 10 ng/ml cholera toxin for 16 h stimulated serotonin transportby 2.3 & 0.2-fold. H-9 alone a t


15755

20k

4U J \

.0 c

2a

c

1.0

bo

a

2

-B 4 0.5

3

.-m 0 * 2

I

4

Control Cholera

toxin

H-9

Cholera

\

o\ I

~

toxin

+

H-9

FIG. 6. Effect of H-9, a protein kinase inhibitor, on cholera toxin-induced stimulation of serotonin uptake. The choriocarcinoma cells were treated overnight (16 h) with or without cholera . the toxin (10 ng/ml) in the presence or absence of H-9 (50 y ~ ) After treatment, uptake of serotonin (50 nM) was measuredin the presence of NaCl, using a 5-min incubation.

\ .

1 V/B

30 10 20 40 50 (pmoI/mg of proteindmin/~M)

FIG. 7. Kinetic analysis of serotonin uptake in control and cholera toxin-treated choriocarcinoma cells. The choriocarcior) without noma cells were treated overnight (16 h) with (M (W) cholera toxin (10 ng/ml). After the treatment, initial uptake rates (5-min incubation) of serotonin were measured in these cells from an NaC1-containing medium over a serotonin concentration of 0.1-1.5 p ~ The . results are given as an Eadie-Hofstee plot (initial uptake rate/serotonin concentration versus initial uptake rate).

a concentration of 50 p~ showed a 33% inhibitionof serotonin transport compared with control cells. However, compared with the cells treated with H-9, serotonin transport in cells treated with cholera toxin plus H-9 was stimulated only by transporter that we described previously in the brush-border 1.6 & 0.2-fold. Thus, the stimulatory effect of cholera toxin membrane of the normal term human placenta (6-10). These was significantly ( p < 0.015) antagonized by H-9, suggesting characteristics include energization by a transmembrane Na' the involvement of the A-kinase in the cholera toxin-inducedgradient, an obligate dependence on C1-, specificity and high stimulation. affinity for serotonin, and sensitivity to inhibition by tricyclic Influence of Cholera Toxin on the Kinetic Parameters of and nontricyclic antidepressants. In addition, we have used Serotonin Transport-The kinetic parameters of serotonin this cell linetoinvestigatethe regulatory aspects of the transport were determined by studying the relationship be- serotonin transporter. The results presented in this paper tween the initial transport rates (5-min incubation) and sedemonstrate clearly that the serotonin transport activity in rotonin concentrations in cells incubated overnight in the the JARcell line is under regulationby CAMP.Cholera toxin, presence and absenceof 10 ng/ml cholera toxin. The concen- which is known to elevate the levels of intracellular CAMP, tration of serotonin was varied over a range of 0.1-1.5 p ~ At . stimulates the serotonin transport activityby approximately each concentration the transport buffer contained 0.1 p~ 3-fold. The participation of cAMP in this stimulation was radiolabeled serotonin, and the remainder was unlabeled se- substantiated by demonstrating that cholera toxin increases rotonin. Non-carrier-mediated transport was determined by the concentrationof cAMP in thesecells by several hundredmeasuring the transportof the radiolabel in the presenceof 1 fold. The effect of cholera toxin on serotonin transport can mM unlabeled serotonin and was subtracted from the total be mimicked by forskolin, IBMX, and dibutyryl CAMP. The transport to calculate the carrier-mediated transport. When stimulatory effect of cAMPis specific for theserotonin these data were analyzed according to the method of Eadie- transporter because, undersimilarconditions, two other Hofstee (initial transport rate/serotonin concentration versus transport systems we studied in this cell line are not stimuinitial transport rate), linear plots (? > 0.86) were obtained lated. H-9,a protein kinase inhibitor,significantly attenuates in control as well as cholera toxin-treated cells, indicating the thestimulationcaused bycholera toxin, suggesting that involvement of a single transport system in the transport of CAMP-dependent protein phosphorylation is involved in the serotonin in both the cases (Fig. 7).However, the kinetic cholera toxin-mediated stimulation. Kinetic analysis reveals constants differed significantly between the control cells and that the stimulation is caused by an increase in themaximal the cholera toxin-treated cells. In controlcells, the Michaelis- velocity of the transport system aswell as an increase in the Mentenconstant was 0.52 f 0.09 p ~ and , themaximal affinity of the system for serotonin. The mechanism responvelocity was7.89 f 0.67 pmol/mg of protein/5 min. In cholera sible for the stimulationof serotonin transportby cAMP was toxin-treated cells, the correspondingvalues were 0.29 f 0.04 not investigated in this study. However, since the elevation PM and 17.55 k 1.06 pmol/mg of protein/5 min. Thus, the of cAMP content in these cells by cholera toxin or forskolin cholera toxin-induced stimulation of serotonin transport in was observed within 2 h whereas the stimulatory effect on these cells was caused by a n increase in the maximal velocity serotonin transport was not evident until 8 h, it islikely that of the transport system as well as an increase in the affinity theunderlyingmechanism was not CAMP-induced phosof the transporterfor serotonin. phorylation of the transporter but rather CAMP-induced stimulation of de novo synthesis of the transporter. DISCUSSION The significance of these results is many-fold. The obserIn the present paper we have provided unequivocal evidence vation that the serotonin transporter is regulated by cAMP t o show that the JAR human placental choriocarcinoma cell is importantbecause there is currentlyvery little information line expresses a highaffinity serotonin transporter. The char- available on the regulatory aspects of the serotonin transacteristics of this transporter are very similar to thoseof the porter in any tissue. Ours is the first report on involvement the

15756

Cyclic AMP-dependent Regulation

of cAMP in the modulation of the serotonin transporter. A role for C-kinase has been suggested recently in the modulation of the serotonin transporter function in endothelial cells (18). In this case, the effect of C-kinase is to reduce the transporter activity. Our findings that A-kinase is involved in the stimulation of the transporter would then indicate that the activity of the transporter is under reciprocal control by A-kinase and C-kinase. Most of the attention on the serotonin transporter has so far been focused on therole of the transporter in serotonergic transmission. That platelets expresses a similar serotonin transporter hasprompted many investigatorsto use platelets as a model to study the mechanisms of serotonergic transmissionin neurons. However, there appears to bea growing interest in recent years on the function and regulation of the serotonin transporter in nonneuronal systems such as the endothelium and theplacenta. Moreover, recent studies have indicated that theserotonin transporters of the neuronal and nonneuronal tissues may be different. This is true even for the platelet, which has often been used as a model to study the neuronal serotonin transporter. One of the predominant differences between the neuronal and nonneuronal serotonin transporters is in the interaction with Hf. There is evidence that the function of the platelet plasma membrane serotonin transporter is coupled to transmembrane H+ gradients (36). The presence of an outwardly directed H' gradient stimulated the NaC1-dependent serotonin transport many-fold in purified platelet plasma membrane vesicles. We have reported recently that theserotonin transporter of the human placental brush-border membrane is also modulated by pH and pH gradients in a similar manner (8).On the contrary, the serotonin transporter in synaptosomal membrane vesicles is not stimulated but is rather inhibited by an outwardly directed H+ gradient (30). Furthermore, there are indications that the platelet and theneuronal serotonintransporters may be under separate genetic control (37). Therefore, the serotonin transporters of the neuronal and nonneuronal tissues may very likely be different proteins and hence may be regulated differently by intracellular second messengers. The serotonin transporter of the JARcell line reported in thepresent paper can be taken as a model to investigate the characteristics of the nonneuronal type transporter. Since species differences in theproperties of the serotonin transporter cannotbe totally excluded, use of the JAR human placental choriocarcinoma cell line as well as brush-border membrane vesicles from term human placentas offers a unique opportunity to characterize the nonneuronal serotonin transporter of human origin. Human platelets are an alternate,viable model for this purpose, but there are limitations this in case such as tissue availability. In addition, platelets, being anucleated cells, are unsuitable for investigations involving genetic analysis of the transporter. There is another unique aspect of the human placental serotonin transporter. Unlike platelets, the human placental syncytiotrophoblast is a polarized cell, having two distinct regions of the plasma membrane, the brush-border or maternal facing membrane, and thebasal or fetal facing membrane. However, the serotonin transporter is present only in the brush-border membrane. The polarity of the distribution of the transporter in the syncytiotrophoblast suggests that serotonin in the maternal circulation can be transported preferentially into the cell. The two poles of the syncytiotrophoblast plasma membrane differ markedly in thedistribution of hormone receptors in these membranes. The maternalfacing brush-border membrane possesses receptors for insulin, epidermal growth factor, and insulin-like growth factor Iwhereas

of Serotonin Transporter

the fetal facing basal membrane preferentially contains the &adrenergic receptor (38). Adenylate cyclase, responsible for the intracellular generation of CAMP, is present only in the basal membrane. Our observations that cAMP has a profound stimulatory effect on the serotonin transporter are thusvery important to the understanding of the physiological regulation of the transporter function in the normal placenta. These findings suggest that theactivity of the serotonin transporter in the maternal facing membrane can be regulated by hormones of fetal and/or cytotrophoblast origin which interact with the receptors in the basalmembrane and affect the intracellular levels of cAMP throughmodulation of the adenylate cyclase activity. Cell lines that stably express the serotonin transporter have enormous potential in studiesinvolving the regulation of the transporter. A number of cell lines have been shown to possess the serotonin transporter activity, but all of them areof mouse origin, and, moreover, these cell lines express the transporter only after experimental manipulations suchas viral transformation and/or induction of differentiation. In addition, since it is likely that species differences in the properties of the serotonin transporter exist, the results from the experiments using the mouse cell lines cannot be directly extrapolated to the human serotonin transporter. A recent report described a mouse cell line that expresses a serotonin transporter upon transfection with human genomic DNA (23). However, it is not known whether the expressed transporter is indeed the human serotonin transporter because there is a possibility that thetransfection maneuver activated the mouse gene and made the cell line express the mouse serotonin transporter. Even if the expressed transporter is of human origin it is not clear whether the transporter is of the neuronal or nonneuronal type. The JAR human choriocaricnoma cell line provides an unambiguous experimental tool to investigate the regulatory and molecular aspects of the human nonneuronal serotonin transporter. Pharmacologically, the placental transporter exhibits characteristics that are very similar to those of the neuronal transporter. The antidepressants such as fluoxetine, paroxetine, imipramine, and desipramine inhibit the placental as well as the neuronal serotonin transporters. The JARcell line can be used for rapid screening of potential antidepressants for their potency to inhibit the serotonin transporter. There is absolutely no informationavailable on the possible role of serotonin in placental function. Human placenta is a highly vascularized tissue, and serotonin is known to have profound effects on the vascular tone (39, 40). Moreover, serotonin has also been shown to be a growth factor for certain types of cells (41, 42). Therefore, it is very likely that this monoamine plays an important role in the physiology of the human placenta and that the serotonin transporter present in the brush-border membrane of the syncytiotrophoblast is involved in the regulation of serotonin levels in the placental tissue. The presence of the serotonin transporterinthis polarized cell is indeed unique because the brush-border membranes of other polarized cells such as intestinal and renal tubular cells do not possess the transporter. That cAMP modulates the activity of the transporterin the placental cell line underscores the physiological importance of the transporter because cAMP is a well known second messenger in thehuman placentalsyncytiotrophoblast, modulating the synthesis and/or secretion of various hormones (43-45). Acknowledgment-We thank Marcia D. Lewis for excellent secretarial assistance.

Transporter Cyclic AMP-dependent Regulation of Serotonin REFERENCES 1. Sneddon, J. M. (1973) Prog. Neurobiol. 1, 151-198 2. Ross, S. B. (1982) in Biology of Serotonergic Transmission ( 0 s borne, N. N., ed) pp. 159-195, John Wiley & Sons, New York 3. Davis, A. (1984) Ezperientia 40, 782-794 4. Graham, D. & Langer, S. Z. (1988) in Neuronal Serotonin (Osborne, N. N. & Hamon, M., eds) pp. 367-391, John Wiley & Sons, New York 5. Marcusson, J. 0.& Ross, S. B. (1990) Psychopharmacology 102, 145-155 6. Balkovetz, D. F., Tiruppathi, C., Leihach, F. H., Mahesh, V.B. & Ganapathy, V. (1989) J. Biol. Chem. 264,2195-2198 7. Ganapathy, V., Kulanthaivel, P., Tiruppathi, C., Mahesh, V. B. & Leibach, F. H. (1989) J. Pharrnacol. Exp. Ther. 251,9-15 8. Cool, D. R., Leihach, F. H. & Ganapathy, V. (1990) Biochemistry 29,1818-1822 9. Cool, D. R., Leibach, F. H. & Ganapathy, V. (1990)Am. J . Physiol. 259, C196-C204 10. Cool, D. R., Leihach, F. H. & Ganapathy, V. (1990) Biochem. Pharmacol. 40, 2161-2167 11. Hahert, E., Graham, D. & Langer, S . Z. (1986) Eur. J. Pharmacol. 122,197-204 12. Talvenheimo, J. & Rudnick, G. (1980) J. Biol. Chem. 255,86068611 13. Davis, A,, Morris, J. & Tang, S . W. (1983) Eur. J . Pharrnacol. 86,353-359 14. Cesura, A. M., Muller, K., Peyer, M. & Pletscher, A. (1983) Eur. J. Pharmacol. 96, 235-242 15. Mellerup, E. T., Plenge, P. & Nielsen, M. (1984) Eur. J. Pharmacol. 106,411-413 16. Wennogle, L. P., Ashton, R. A., Schuster, D. I., Murphy, R. B. & Meyerson, L. R. (1985) EMBO J.4,971-977 17. Biessen, E. A. L., Horn, A. S. & Robillard, G. T. (1990) Biochemistry 29,3349-3354 18. Myers, C.L., Lazo, J. S . & Pitt, B. R. (1989) Am. J. Physiol. 257, L253-L258 19. Lee, S. L. & Fanburg, B. L. (1986) Am. J. Physiol. 250, C761C765 20. Hilihrand, A. S., Richardson, U. I., Blusztajn, J. K. & Wurtman, R. J. (1987) Neurochem. Znt. 10, 185-189 21. DeVitry, F., Catelon, J., Dubois, M., Thihault, J., Barritault, D., Courty, J., Bourgoin, S. & Hamon, M. (1986) Neurochem. Znt. 9,43-53 22. Buc-Caron, M. H., Launay, J. M., Lamblin, D. & Kellermann, 0. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 1922-1926

15757

23. Chang, A. S. S., Frnka, J . V., Chen, D. & Lam, D. M. K. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 9611-9615 24. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 25. Kletzien, R. F., Pariza, M. W., Becker, J. E. & Potter, V. R. (1975) Anal. Bwchem. 68,537-544 26. Bhalla, V. K., Powell, M. M., Rojas, F. J. & Rajan, V. P. (1980) in Functional Correlates of Hormone Receptors in Reproduction (Mahesh, V. B., Muldoon, T. G., Saxena, B. B. & Sadler, W. A,, eds) pp. 419-436, Elsevier/North Holland, New York 27. Rudnick, G. (1977) J. Biol. Chem. 252,2170-2174 28. Nelson, P. J. & Rudnick, G. (1982) J. Biol. Chem. 257, 61516155 29. O’Reilly, C. A. & Reith, M. E. A. (1988) J. Biol.Chem. 263, 6115-6121 30. Reith, M. E. A., Zimanyi, I. & O’Reilly, C. A. (1989) Biochem. Pharmacol. 38, 2091-2097 31. Stern-Bach, Y., Greenberg-Ofrath, N., Flechner, I. & Schuldiner, S. (1990) J. Biol. Chem. 265,3961-3966 32. Miyamoto, Y., Balkovetz, D. F., Leibach, F. H., Mahesh, V. B. & Ganapathy, V. (1988) FEBS Lett. 231,263-267 33. Kulanthaivel, P., Leibach, F. H., Mahesh, V. B. & Ganapathy, V. (1989) Biochim. Biophys. Acta 985, 139-146 34. Kulanthaivel, P., Cool, D. R., Ramamoorthy, S., Mahesh, V. B., Leibach, F. H. & Ganapathy, V. (1991) Biochem. J. 277, 5358 35. Hidaka, H., Inagaki, M., Kawamoto, S. & Sasaki, Y. (1984) Biochemistry 23,5036-5041 36. Keyes, S. R. & Rudnick, G. (1982) J. Biol.Chem. 257, 11721176 37. Ieni, J . R., Tobach, E., Zukin, S. R., Barr, G. A. & Van Pragg, H. M. (1985) Eur. J. Pharmacol. 112, 261-264 38. Kelley, L. K., Smith, C. H. & King, B. F. (1983) Biochim. Biophys. Acta 734,91-98 39. Houston, D. S. & Vanhoutte, P. M. (1986) Drugs 31, 149-163 40. Hollenberg, N. K. (1988) Annu. Reu. Pharmacol. Toxicol. 28,4159 41. Takuwa, N., Ganz, M., Takuwa, Y., Sterzel, R. B. & Rasmussen, H. (1989) Am. J . Physiol. 257, F431-F439 42. Seuwen, K. & Pouyssegur, J. (1990) Biochem. Pharmacol. 39, 985-990 43. Feinman, M. A., Kliman, H. J., Caltabiano, S. & Strauss, J . F. (1986) J. Clin. Endocrinol. & Metab. 63, 1211-1217 44. Harman, I., Costello, A., Sane, A. & Handwerger, S . (1987) Endocrinology 121, 59-63 45. Benoit, J., Rodway, M., Yuen, B. H. & Leung, P. C. K. (1988) Am. J. Obstet. Gynecol. 158, 328-332