The Prostate 51:201^210 (2002)

K+Channel Currents in RatVentral ProstateEpithelial Cells

Jun Hee Kim,1 Eun-Kyung Hong,2 Hee Sook Choi,2 Seung-Joon Oh,3

Kwang Myung Kim,3 Dae-Yong Uhm,1 and Sung Joon Kim1*1Departmentof Physiology, SungkyunkwanUniversity SchoolofMedicine, Suwon,Korea

2Medvill Central Research Laboratory, Pyungchang-Dong, Seoul,Korea3DepartmentofUrology, SeoulNational UniversityMedical College, Seoul,Korea

BACKGROUND. Electrophysiological function of the normal prostate has not beenextensively studied. In particular, ion channel currents and their regulation have not beenstudied in freshly-isolated prostate cells.METHODS. Rat prostate secretory epithelial (RPSE) cells were isolated by collagenasetreatment. Columnar epithelial cells were used for nystatin-perforated, whole-cell voltageclamp, and the intracellular Ca2þ concentration ([Ca2þ]i) was measured using fura-2.RESULTS. Step-like depolarizing pulses (900 msec) starting from � 90 mV induced out-wardly rectifying Kþ currents without inactivation. ACh (10 mM) or ATP (100 mM) increasedthe outward current and hyperpolarized the cell membrane potential. Ionomycin (0.1 mM),a Ca2þ ionophore, induced a similar increase in the outward current. TEA (5 mM),charybdotoxin (50 nM), and iberiotoxin (30 nM) inhibited the effect of ACh (or ATP) on theoutward current, whereas apamin (100 nM) had no effect. The [Ca2þ]i of RPSE cells wasincreased by ACh, ATP, and UTP.CONCLUSIONS. RPSE cells have iberiotoxin-sensitive Ca2þ-activated Kþ channels that mayplay an important role in the exocrine secretions of the prostate. Prostate 51: 201–210, 2002.# 2002 Wiley-Liss, Inc.

KEY WORDS: prostate epithelium; Kþ channel; voltage clamp; intracellular Ca2þ;muscarinic receptor; purinoceptor

INTRODUCTION

The prostate, a major sex accessory gland of males,is found in essentially all mammalian species, buthas a widely varying morphology depending on thespecies. Its main function is to produce citrate-richfluid that comprises the major portion of seminalplasma [1,2]. Although the microscopic morphology ofthe prostate is not homogeneous, the luminal epithe-lium is mostly composed of columnar epithelial cells[2,3]. Because in humans the prostate surrounds theurethra in a restricted space in the lower pelvis, it is theorgan where the highest number of male urologicalsymptoms arises, benign prostate hyperplasia (BPH).Also, prostate cancer is one of the most frequentlydiagnosed malignancies in the male [3,4].

Compared with the extensive studies of the clinicalproblems of the prostate, the normal physiological

function of the prostate (e.g., the mechanism of fluidsecretion) is still unclear [1,5,6].Wang et al. [6] reportedthat the secretion of prostatic fluid is under neuronalcontrol; adrenergic stimulation triggers smooth mus-cle contraction and expels the fluid, whereas choliner-gic stimulation appears to increase the formation offluid in the gland without contracting the smooth

Grant sponsor: Plant Diversity Research Center of 21st CenturyFrontier Research Program funded by the Ministry of Science andTechnology of the Korean government; Grant number: PF002108-03.

*Correspondence to: Dr. Sung Joon Kim, Department of Physiology,Sungkyunkwan University School of Medicine, Suwon 440-746,Korea. E-mail: sjoonkim@med.skku.ac.krReceived 19 September 2001; Accepted 29 January 2002DOI 10.1002/pros.10090

� 2002 Wiley-Liss, Inc.

muscle. In that study, the secretion induced by thecholinergic stimulation was sustained and not affect-ed by denervation, indicating that cholinergic drugsdirectly stimulate the epithelial secretory mechanism.

For the effective formation of primary secretoryfluid in exocrine glands, the presence of ion channelsand their regulation by secretagogues is essential [7].To the best of our knowledge, however, electrophy-siological studies of prostate epithelial cells have beenconfined to immortalized cancer cell lines that do notreflect the normal secretory function. Using culturedrat prostate epithelial cells, Ouadid-Ahidouch et al.observed a voltage-dependent Kþ current; however,the effect of secretagogues on the Kþ channel was nottested [8]. Kþ channels are important for the main-tenance of negative membrane potential, electricalsignaling, cell volume regulation, and exocrine secre-tion. Because muscarinic stimulation is known to havea role in prostatic fluid formation [6], we investigatedthe effect of ACh on ion channels in the exocrineepithelial cells of prostate.

The rat prostate is a popular experimental systemfor andrologic studies. The rat prostatic tissue is atubulo-alveolar gland and consists of exocrine andneuroendocrine epithelium-lined acini surrounded bya stromal matrix [3,9]. It is composed of three lobes,and most studies of the rat prostate have been carriedout with ventral lobe [9]. In a preliminary experiment,we discriminated typical columnar epithelial cellsfrom the ventral lobe after single cell isolation usingcollagenase and gentle agitation. This encouraged usto investigate the membrane currents of prostateepithelial cells and their regulation by agonists.

MATERIALSANDMETHODS

Cell Isolation

All procedures on experimental animals wereperformed following the guidelines of the InstitutionalAnimal Care & Use Committee (IACUC) of Sung-kyunkwan University. A total of 37 rats were usedfor the study. Male Sprague–Dawley rats (350–400 g)were killed by 100% CO2 inhalation and the ventrallobe of prostate was removed rapidly. The fibrouscapsule was removed, and the tissue was cut withscissors into small pieces (1–2 mm3) in a phosphate-buffered Ca2þ-free Ringer’s solution containing145 mM NaCl, 1.6 mM K2HPO4, 0.4 mM KH2PO4,1 mM MgCl2, and 5 mM D-glucose, at pH 7.4. For theenzymatic digestion, collagenase IV (1 mg/ml) (Serva,Heidelberg, Germany), trypsin inhibitor (0.6 mg/ml)(Sigma, St. Louis, MO), and bovine serum albumin (0.5mg/ml) (Sigma) were added and incubated at 358C for25 min in a shaking water bath. After digestion, thetissue chunks were moved to a fresh solution and

gently agitated using a fire-polished Pasteur pipette(1–2 mm tip diameter). Afterwards, the cell suspen-sion was kept at 48C for up to 4 hr in the same solution.Columnar cells with pale cytoplasm could be clearlydistinguished under the view of inverted microscope(IX-70, Olympus, Tokyo, Japan). For hematoxylin–eosin (H–E) staining, some cells were cytospun ontosilane-coated slides (Muto Pure Chemical, Japan), air-dried, and prepared for the staining procedure (Fig. 1).

Patch-ClampMethods

Isolated cells were transferred into a bath chambermounted on the stage of an inverted microscope(IX-70, Olympus). The bath with a volume of approxi-mately 0.3 ml was perfused at a rate of 10 ml/min atroom temperature (22–258C). The tip resistance ofthe patch pipettes was between 3 and 3.5 MO. Thecomposition of the pipette solution was: 105 mM K-gluconate, 30 mM KCl, 0.4 mM NaH2PO4, 1.6 mMNa2HPO4, 0.73 mM CaCl2, 1 mM MgCl2, and 1 mMethyleneglycol-bis-(oxonitrilo)-tetraacetate (EGTA) atpH 7.3 and pCa 7. For the nystatin-perforated whole-cell patch clamp, a stock solution of nystatin indimethylsulfoxide (15 mg/ml) was added to thepipette solution to make a final concentration of0.15 mg/ml. The standard bath solution, a phos-phate-buffered Ringer’s solution, contained 145 mMNaCl, 1.6 mMK2HPO4, 0.4 mMKH2PO4, 1 mMMgCl2,1.3 mM CaCl2, and 5 mM D-glucose, at pH 7.4. Allchemicals used were of the highest grade of purityavailable and were obtained from Sigma and RBI(Natick). The reference electrode was Ag/AgCl, andthe liquid junction voltage was nullified using thecircuit of the amplifier (Axopatch 1-D, Axon Instru-ments, Foster City, CA). pCLAMP software v.7.0 andDigidata-1200A (both from Axon Instruments) wereused for the acquisition of data and applying com-mand pulses. Details of the voltage command proto-

Fig. 1. The morphology of single prostate cells (H^E staining).Photographs from two different sets of experiments are shown.Filled arrows indicate columnar cells (RPSE cells). Open arrowsindicate round cells regarded here to be RPNE cells. Due to thecytospinningprocedure, initially isolated cells came to be clumped.

202 Kimet al.

cols are given in the figure legends. The voltage andcurrent data were low-pass filtered (5 kHz), anddisplayed on a computer monitor. The data wasstored in a computer and analyzed using Originv. 6.1 (Microcal Software Inc., Northampton, MA).

Intracellular Ca2þMeasurement

Isolated single cells were loaded with the ace-toxymethyl ester form of fura-2 (2 mM) in the Ringer’ssolution for 20 min at room temperature and thenwashed out with fresh solution. The recording of[Ca2þ]i was performed with a microfluorometricsystem consisting of an inverted fluorescence micro-scope (IX-70, Olympus) with a dry-type fluorescenceobjective lens (40�, NA 0.85), a photomultiplier tube(type R 1527, Hamamatsu, Japan) and a Deltascanilluminator (Photon Technology International Inc.,Lawrenceville, NJ). Light was provided by a 75-Wxenon lamp (Ushino, Japan). A chopper wheel alter-nated the light path to monochromators (340 and380 nm) with a frequency of 5 Hz, and the intensity ofemitted light at 510 nm was measured. As a measureof [Ca2þ]i, the ratio of fluorescence emissions at 340and 380 nm excitation is presented.

Data Presentation

The data are presented as original recordings,current/voltage (I/V) curves and bar graphs ofmean� SEM with number of cells tested. Wherenecessary, paired Student t-test was applied andP< 0.05 was regarded as significant.

RESULTS

Previous histological studies consistently describethe epithelial cells lining the inner wall of prostategland as tall, columnar cells with pale cytoplasm [9].Figure 1 shows H–E-stained prostate cells after singlecell isolation. Although the cells looked flattened be-cause of the fixation process, we could still identifythe columnar cells (Fig. 1, filled arrows). Since thesecolumnar cells were the cells most frequently observedand they had a very similar morphology to thosedescribed as exocrine cells lining the prostate glandwall [9], they were regarded as rat prostate secretoryepithelial (RPSE) cells. Other cells, indicated by emptyarrows in Figure 1, had a rounder shape and darkercytoplasm than the RPSE cells. This group of cells wasnot only distinguishable by the morphology, but alsohad completely different membrane current responsesto voltage clamp stimulation (see Fig. 3). The dark,round cells had positive immunoreactivity to chromo-granin-A, a representative marker of prostate neu-roendocrine cells [3] (data not shown). Therefore, we

regarded them as rat prostate neuroendocrine (RPNE)cells.

RPSE cells were used for the nystatin-perforatedwhole-cell patch clamp experiment, except for thecases in Figure 3. The resting membrane potentialof RPSE cells in the control state was � 18� 1.9 mV(n¼ 21) using a pipette solution of cytoplasm-likecomposition (see Materials and Methods) and underthe zero-current clamp condition. This value is moredepolarized than the value recorded using theconventional intracellular microelectrode techni-que in the rat prostate [10]. To observe the voltage-dependent ion channel activity, membrane voltagewas held at � 90 mV and step pulses of 900 msec wereapplied with an increment step of 10 mV from � 80mV to þ 70 mV (Fig. 2A). Upon depolarization, aninstantaneous activation of outward current wasobserved with little time-dependent change duringthe step pulses. A summary of the current-to-voltagerelationships (I/V curves) of seven RPSE cells showedthat they were outwardly rectifying (Fig. 2B). Whentetraethylammonium (TEA) (1 and 5 mM) was appliedto the bath solution, the amplitude of the outwardcurrent was markedly decreased (Fig. 2A,C).

For comparison, the same voltage pulses wereapplied to the round cells (RPNE cells). In contrast tothe RPSE cells, the membrane current response ofRPNE cells showed complex time-dependent kinetics,an initial sharp outward current with rapid inacti-vation, and a secondary transient outward current(Fig. 3A). With 5 mM TEA in the bath solution,the secondary transient outward current was selec-tively blocked, while the initial transient outwardcurrent persisted (Fig. 3B). Throughout the study,there was a consistent correlation between the shape ofthe cells and the kinetics of the membrane currentresponses.

In the next experiment, we examined the effectof neurotransmitters on the outward current of RPSEcells. In Figure 4A, the same series of step depolarizingpulses was applied before and during the applicationof 10 mM ACh. The amplitude of the outward currentwas markedly increased by ACh and the membranepotential was hyperpolarized to � 35� 2.3 mV(n¼ 15). Since the control current and the ACh-induced outward current did not show a significanttime-dependent change with the step-pulse proto-col, ramp-like pulses (see the legend of Fig. 4) wereapplied every 3 sec to get brief I/V curves of RPSEcells. The current responses to ramp-like pulsesclearly demonstrated the effect of ACh (Fig. 4B).We also tested the effect of norepinephrine (NE)(10 mM), which increased the outward current onlyslightly in rare cases (2 out of 12 tested cells) (data notshown).

KþCurrents in Rat Prostate Epithelial Cells 203

Extracellular ATP triggers various intracellularsignal transduction cascades through either P2Y-type metabotropic receptors coupled to the G-proteinpathway or P2X-type ionotropic receptors throughwhich cation influx occurs [11]. In RPSE cells, 100 mMof ATP increased the outward current in a mannervery similar to the effect of ACh (Fig. 4C). The cor-responding I/V curves also show that both extra-cellular ATP and ACh increases the outward current;their curves cross the control I/V curve at about� 70 mV, close to the Nernst equilibrium potential of

Kþ (Fig. 4B,C). Similar to the effect of ACh, themembrane potential of the RPSE cells was hyperpolar-ized to � 36� 2.3 mV (n¼ 12) with ATP. Since bothmuscarinic and P2Y receptors can release Ca2þ fromintracellular stores [11], it is likely that the increase ofoutward current was due to an activation of Ca2þ-activated Kþ (K(Ca)) channels. We also tested theeffect of ionomycin, a Ca2þ ionophore. Ionomycin(0.1 mM) slowly increased the outward current andinduced membrane hyperpolarization (Fig. 4D) (n¼2). In another series of experiments, we adopted

Fig. 2. Outward current activated bymembrane depolarization in RPSE cells.A: Representative current traces obtained by incrementalstep-like depolarizing pulses every 5 sec (seepulseprotocol above left traces). Steady-state effects of TEAon the same cell are shown in themiddle (1mM) and right traces (5mM).B:The current-to-voltage relationship (I/V curve) of RPSE cells.Mean current amplitudes from sevenRPSE cells areplotted against the clampvoltages (Vc).C:Mean effects of TEAon three RPSE cells.

204 Kimet al.

a conventional ruptured whole-cell patch clampmethod with a high concentration of EGTA (10 mM)in the pipette solution to prevent changes in [Ca2þ]i.Under this condition, ACh did not change themembrane current, indicating that the increase in theKþ current was due to an increase of [Ca2þ]i (P> 0.05,n¼ 5) (Fig. 4E). Compared to the prominent effects onthe outward current, the inward current at the nega-tively hyperpolarized voltage range was only slightlyincreased by the application of agonists. On theapplication of ACh, the currents were increased to402� 48.8% of control at 0 mV and 152� 39.6% ofcontrol at � 80 mV (P< 0.05, n¼ 12). With ATP, theywere increased to 354� 47.4% and 119� 9.5% at 0 and� 80 mV, respectively (P< 0.05, n¼ 11).

To further investigate the ionic selectivity of theACh-activated current and its voltage dependence,RPSE cells were perfused with high Kþ (125 mM)solution and the same ramp-like pulses were applied(Fig. 4F). With high Kþ solution in the bath, AChevoked an inward current at � 40 mV and the I/Vcurve showed a U-shape, which crossed the controlcurrent at around � 5 mV. The shift of the reversalpotential to a more positive value suggested that theKþ channel was most likely activated by ACh. Also,the decrease of the inward current by hyperpolariza-tion strongly indicated that the ACh-activated Kþ

channels are voltage dependent in RPSE cells (seeDiscussion).

In another series of experiments, we tested whetherACh or ATP could increase the [Ca2þ]i of RPSE cells.Typical responses from three different RPSE cells areshown in Figure 5. The positive effects of ACh andATP could be demonstrated in the same cell as an

increase in the fura-2 fluorescence ratio (n¼ 10). Insome cases, however, the response was limited to oneof the two agonists (ACh or ATP) tested (n¼ 3) (datanot shown). Among the wide variety of P2Y purino-ceptors, the P2Y1 and P2Y2 subtypes have beencommonly observed to be linked with an increase of[Ca2þ]i [11]. The most distinguishing feature of theP2Y2 subtype is that it is equally sensitive to UTP, apyrimidine, while P2Y1 is far less sensitive to UTP.Therefore, we tested the effect of UTP as well as ATP.UTP (100 mM) also increased [Ca2þ]i (n¼ 5) (see themiddle and the lower traces of Fig. 5).

K(Ca) channels are separated into three classesbased on their single channel conductance: large con-ductance (BK or maxi-K), small-conductance (SK), andintermediate conductance (IK or SK4) [12]. As a nextstep, we tested several pharmacological agents knownto block K(Ca) channels with some specificity for eachclass. Apamin (500 nM), a bee venom that blocksSK, had no effect on the control current and did notblock the effect of ATP (n¼ 3) (Fig. 6A). BK channel isknown to be blocked by TEA, charybdotoxin (ChTX),and most specifically by iberiotoxin (IbTx) [12]. Asshown in Figure 6B,C,D, the outward current of RPSEcells was decreased by the above three blockers andthe effects of the agonists were markedly suppressed(n¼ 3 for each blocker). The SK and IK (SK4) familiesare basically voltage independent and open only inresponse to Ca2þ elevations. In contrast, the BK type ofK(Ca) channels has an intrinsic voltage dependencethat is modulated by intracellular Ca2þ [12]. Wehave already shown the voltage dependence of ACh-activated current (Fig. 4E), which has a decreasedconductance when hyperpolarized and produces aU-shape I/V curve. Since the activation of voltage-independent channels would produce a linear I/Vcurve in the symmetrical ionic composition, the volt-age dependence of the ACh-activated current sup-ports the role of BK channels in RPSE cells. We alsoconfirmed that the ACh-activated current in the highKþ condition is not blocked by clotrimazole (CTZ), ablocker of IK [12] (Fig. 6E).

DISCUSSION

In this study, we characterized the Kþ current in thefreshly-isolated secretory epithelial cells of rat pros-tate, the first time this has been done. We alsoprovided firm evidence that natural agonists can elicitincreases in [Ca2þ]i, one of the most importantintracellular signals. The electrophysiological andpharmacological data are consistent with the presenceof the BK type of K(Ca) channel that can be furtheractivated by cholinergic or purinergic stimulation.Although small in its size, a significant increase of the

Fig. 3. Outward current activatedbymembrane depolarizationin RPNE cells.A: Representative current traces obtained by step-like pulses every 10 sec from � 90 mV to � 40, 0, and þ 50 mV.B: In the same cell, 5 mMTEA selectively abolished the slow time-dependent component and revealed fast inactivation of the initialtransientcomponent.

KþCurrents in Rat Prostate Epithelial Cells 205

Fig. 4. Effects of ACh, ATP, and ionomycin on the outward current of RPSE cells.A:The cellmembrane voltagewas held at� 90mV, andincremental step-likepulses from� 60mV toþ 80mV (20mVintervals)were appliedevery 5 sec.Left: control, right: effects ofACh (10mM).B^F:Brief I/Vcurvesobtainedbyramp-likepulses (0.1V/sec) from� 80toþ 50mV(B)or from�100toþ 40mV(C^F).During theinterpulseperiod, themembranevoltagewasheldat� 40mV,andthepulsewasappliedevery3sec.EffectsofACh(10mM)(B,E,F),ATP(100mM)(C), andionomycin (0.1 mM) (D).Original chart traces are also shown in (C) and (D). E:Membrane currents recordedunder the rupturedwhole-cellclamp conditionwith10 mMEGTA. ACh had no effect.Note that TEA (5mM) largely reduced the outward current.F: I/V curves recordedinhighKþ (125mM)bath solution.The I/V curves cross the abscissa at approximately 0mV. ACh (10mM)markedly increased the amplitude ofthe inwardcurrent and shifted the inwardpeakof the I/V curve to a negativevoltage.Washout: I/V curve obtained after thewashoutofACh.

206 Kimet al.

inward current at negative membrane voltage sug-gested that another ionic conductance (e.g., Ca2þ-activated Cl� channel) might be also activated by AChand ATP in RPSE cells (Fig. 3).

Identif|cation of RPSECells

The epithelium of the prostate gland is composedprincipally of luminal epithelial cells, basal epithelialcells, and neuroendocrine cells [3,13]. The basalepithelial cells are generally considered to be a kindof stem cell from which the secretory epithelial cellsand neuroendocrine cells develop. In the presentstudy, we could not distinguish the stem cell-likebasal cells based on their morphology. Ravindranathand Dym isolated prostate cells of different sizes andsedimentation rates by using the ‘‘STAPUT’’ techni-

que [14]. In that study, however, all the freshly isolatedcells came to have a round shape after the procedure.In contrast, following a relatively mild digestionprocedure, we were able to identify columnar cellsthat could be used for the whole-cell patch clampexperiment.

IonChannels in the Prostate Epithelial Cells

A number of studies of ion channels have beencarried out in different prostate cancer cell lines. Thosestudies focused upon the relationship of the ionchannels with apoptosis, metastasis, and cell prolif-eration. The ion channels expressed in prostate cancercells are nonselective cation channels [15], Ca2þ-inhibited Kþ channels [16], Ca2þ-activated Kþ chan-nels [17,18], voltage-activated Kþ channels [19],voltage-activated Naþ channels [20], volume-activatedCl� channels [21], and store-operated Ca2þ channels[22]. In a recent study of cultured rat prostate cells,only the voltage-activated Kþ channel current wasdescribed. The authors increased the [Ca2þ]i, yet noCa2þ-activated Kþ current was induced [8]. Theabsence of the K(Ca) channel is clearly in contrastwith our results. In our study, using freshly isolatedcells, the application of agonists consistently increased[Ca2þ]i and induced an outwardly rectifying K(Ca)channel current and membrane hyperpolarization thatwas sensitively blocked by IbTx, ChTx, or TEA (Figs. 4and 6). The pharmacological properties and thevoltage dependence suggest the presence of BK chan-nels in RPSE cells. The difference might be due tothe specimens used (dorsolateral lobes were used in[8]) or to changes in cell properties during the primaryculture. It has been reported that the differencesbetween the three lobes of the rat prostate are quitelarge not only in their histological features, but also intheir biochemical properties and physiological func-tions [1]. Further studies comparing the lobes of therat prostate under the same cell isolation conditionswould be helpful for elucidating the role of ionchannels in prostate function.

Role of KþChannels in the Secretory Epithelium

In the rat prostate gland, Szatkowski et al. recordedtransepithelial potentials using a conventional micro-electrode technique (mean, � 4.2 mV) [10]. A lumen-negative transepithelial potential suggests that thesecretory mechanism found in other exocrine glands,electrogenic anion secretion, might also operate in theprostate gland [7,23,24]. Recently, great interest hasfocused upon the secretion of citrate by the prostatebecause the prostatic fluid contains enormous amountof citrate [1]. Since citrate is a trivalent anion, thesecretion of citrate could be facilitated by the mem-

Fig. 5. Cholinergic or purinergic stimulation increases [Ca2þ]i inRPSEcells.The fluorescenceratio fromdual excitationof intracellu-lar fura-2 (R340/380) isplottedagainst time.Each tracewas obtainedfromadifferentRPSEcell.Thebathwas continuouslyperfused, andagonists were applied as indicated in the figure. In the middle andlower traces, UTP as well as ATP increased the [Ca2þ]i of RPSEcells, but to a differentextent.

KþCurrents in Rat Prostate Epithelial Cells 207

brane hyperpolarization due to Kþ channel activation.In addition, Kþ channels can provide a Kþ-recyclingpathway, namely the efflux of Kþ taken up during theoperation of Naþ/Kþ ATPase and other transporters

(e.g., Naþ2Cl�Kþ-cotransporter) [7]. In the prostategland, the luminal membrane is strongly stained withthe anti-a1 subunit of Naþ/Kþ ATPase [5]. The exactmeaning of this localization is not yet known, but it is

Fig. 6. Pharmacology of the Kþ current in RPSE cells. I/Vcurves were obtained by the ramp pulse protocol (0.1 V/sec).A: Effect of ATP after treatment with apamin (500 nM). Notethat apamin alone had no effect on the membrane current.B,C: TEA (5 mM) or ChTx (50 nM) markedly decreased the out-ward current, and in this condition, the addition of ATP inducedonly a small increase in the outward current. (D).The outwardcurrentwas greatly increased by ACh, after which the additionof IbTx (30 nM) abolished most of the outward current. E: I/Vcurves recorded in high Kþ (125 mM) bath solution. AChincreased both inward and outward currents after pretreat-mentwithCTZ (1 mM).

208 Kimet al.

likely that the Kþ channels co-localize in the luminalmembrane for the efficient operation of Naþ/Kþ

ATPase. The activation of Kþ channels and membranehyperpolarization by the Ca2þ- or cAMP-mediatedsecretagogue is reported in many cases of exocrinecells [7,23,24]. The role of the BK type of K(Ca) channelin luminal Kþ secretion or in electrogenic Cl� secretionhas been suggested in various exocrine cells, includingthe pigmented ciliary epithelium of rabbit [25], thepancreatic acinar cells of pig [26], human vas deferensepithelial cells [27], and the acinus of the mouse saliv-ary gland [28]. The IK type of K(Ca) channel also playsan important role in the electrolyte-fluid secretion ofthe pancreatic acini of some species [24] and of thecolonic crypt [23,29].

Role ofMuscarinic Receptors andPurinoceptors in RPSECells

In many kinds of exocrine epithelia, muscarinicstimulation is linked to PLC/IP3 signaling pathwaysand releases stored Ca2þ. The subsequent increase in[Ca2þ]i regulates many kinds of Ca2þ-activated ionchannels, including Ca2þ-activated Cl� channels [24]and K(Ca) channels [23–28]. The prostate is no ex-ception to the concept that the male reproductivetract receives abundant innervation from both thesympathetic and parasympathetic divisions. Althoughless dense than the adrenergic nerves, cholinergicnerves have been found in the prostate and are closelyrelated to the glandular epithelium, suggesting a secre-tory function [30]. The anatomical distribution ofcholinergic nerves is consistent with the physiologicaldata provided by Wang et al. [6]. We also tested theeffect of noradrenaline on RPSE cells, but could notobserve a significant change in the membrane currentor [Ca2þ]i (data not shown). Besides the secretoryregulation, the denervation studies have providedevidence that both autonomic systems may playimportant roles in the growth and maturation of theprostate [30].

In this study, extracellular ATP increased [Ca2þ]iand subsequently augmented the K(Ca) channel inRPSE cells. The effect of extracellular ATP is a novelfinding in prostate cells, but not an unexpected oneconsidering the ubiquitous role of purinoceptors invarious epithelial cells [11,31]. ATP can be secretedfrom nerve terminals or from various epithelial cells ina paracrine or autocrine manner [11]. For example,the urinary bladder epithelium releases ATP under thehydrostatic distension, which can be used to sensethe filling of the bladder [32]. Since the ducts of theprostate gland can be distended depending uponthe autonomic control state [6], a release of ATPsimilar to that in the urinary bladder might occur in

the prostate. In a prostate cell line (DU-145), inter-cellular Ca2þ waves spread from cells triggered bymechanical strain, and it was suggested that thisintercellular communication is due to the paracrine re-lease of ATP and the activation of purinoceptors [33].

Since the present study does not provide anyevidence consistent with the presence of P2X-typenonselective cation channels (e.g., transient activationof inward cationic current), we conclude that P2Ytype receptors linked to a PLC/IP3 pathway mightbe present in RPSE cells andmight induce Ca2þ releasefrom intracellular stores. Although weaker than theeffect of ATP at the same concentration, the significanteffect of UTP suggests that the P2Y2 subtype may bepresent in RPSE cells (Fig. 5). The precise identificationof P2Y receptor subtypes, however, will need rigor-ous pharmacological study to determine the relativepotencies of purine and pyrimidine analogues [11].

CONCLUSIONS

We recorded voltage- and Ca2þ-activated Kþ

current in freshly-isolated RPSE cells. Since the Kþ

channels are strongly regulated by physiologicalagonists via changes in [Ca2þ]i, they may play animportant role in the exocrine secretions of theprostate gland.

REFERENCES

1. Costello LC, Franklin RB. Citrate metabolism of normal andmalignant prostate epithelial cells. Urology 1997;50:3–12.

2. Setchell BP, Brooks DE. Anatomy, vasculature, innervation,and fluids of the male reproductive tract. In: Knobil E, Neill J,editors. The physiology of reproduction. New York: RavenPress; 1988. p 753–836.

3. Xue Y, Smedts F, Verhofstad A, Debruyne F, de la Rosette D,Schalken J. Cell kinetics of prostate exocrine and neuroendo-crine epithelium and their differential interrelationship: newperspectives. Prostate 1998;8(Suppl):62–73.

4. McNeal JE. Cancer volume and site of origin of adenocarcinomain the prostate: relationship to local and distant spread. HumPathol 1992;23:258–266.

5. Mobasheri A, Oukrif D, Dawodu SP, Sinha M, Greenwell P,Stewart D, Djamgoz MB, Foster CS, Martın-Vasallo P, Moba-sheri R. Isoforms of Naþ/Kþ ATPase in human prostate:specificity of expression and apical membrane polarization.Histol Histopathol 2001;16:141–154.

6. Wang J-M, McKenna KE, Lee C. Determination of prostaticsecretion in rats. Prostate 1991;18:289–301.

7. Greger R. The membrane transporters regulating epithelialNaCl secretion. Pflugers Arch 1996;432:579–588.

8. Ouadid-Ahidouch H, Van Coppenolle F, Le Bourhis X, BelhajA, Prevarskaya N. Potassium channels in rat prostate epithelialcells. FEBS Lett 1999;459:15–21.

9. Lee C, Holland JM. Anatomy, histology, and ultra-structure(correlation with function), Prostate, Rat. In: Jones TC, Mohr U,Hunt RD, editors. Genital system (monographs on pathology oflaboratory animals). Berlin, Heidelberg: Springer-Verlag; 1987.p 239–251.

KþCurrents in Rat Prostate Epithelial Cells 209

10. Szatkowski M, Mycielska M, Knowles R, Kho A-L, DjamgozMB. Electrophysiological recordings from the rat prostate glandin vitro: identified single-cell and transepithelial (lumen)potentials. BJU Int 2000;86:1068–1075.

11. Ralevic V, Burnstock G. Receptors for purines and pyrimidines.Pharmacol Rev 1998;50:413–492.

12. Wallner M, Meera P, Toro L. Calcium-activated potassiumchannels in muscle and brain. In: Kurachi Y, Jan LY, LazdunskiM, editors. Potassium ion channels: molecular structure,function, and disease. San Diego: Academic Press; 1999.p 117–140.

13. Bonkhoff H. Neuroendocrine cells in benign and malignantprostate tissue: morphogenesis, proliferation, and androgenreceptor status. Prostate 1998;8(Suppl):18–22.

14. Ravindranath N, Dym M. Isolation of rat ventral prostate basaland luminal epithelial cells by the STAPUT technique. Prostate1999;41:173–180.

15. Wissenbach U, Niemeyer BA, Fixemer T, Schneidewind A,Trost C, Calvalie A, Reus K, Meese E, Bonkhoff H, Flockerzi V.Expression of CaT-like, a novel calcium-selective channel,correlates with the malignancy of prostate cancer. J Biol Chem2001;276:19461–19468.

16. Skryma R, van Coppenolle F, Dufy-Barbe L, Dufy B, Prevars-kaya N. Characterization of Ca2þ-inhibited potassium channelsin the LNCaP human prostate cancer cell line. ReceptorsChannels 1999;6:241–253.

17. Gutierrez AA, Arias JM, Garcıa L, Mas-Oliva J, Guerrero-Hernandez A. Activation of a Ca2þ-permeable cation channelby two different inducers of apoptosis in a human prostaticcancer cell line. J Physiol (Lond) 1999;517:95–107.

18. Rane SG. The growth regulatory fibroblast IK channel is theprominent electrophysiological feature of rat prostatic cancercells. Biochem Biophys Res Commun 2000;269:457–463.

19. Fraser SP, Grimes JA, Djamgoz MB. Effects of voltage-gated ionchannel modulators on rat prostatic cancer cell proliferation:comparison of strongly and weakly metastatic cell lines.Prostate 2000;44:61–76.

20. Laniado ME, Fraser SP, Djamgoz MB. Voltage-gated Kþ

channel activity in human prostate cancer cell lines of markedlydifferent metastatic potential: distinguishing characteristics ofPC-3 and LNCaP cells. Prostate 2001;46:262–274.

21. Shuba YM, Prevarskaya N, Lemonnier L, Van Coppenolle V,Kostyuk PG, Mauroy B, Skryma R. Volume-regulated chloride

conductance in the LNCaP human prostate cancer cell line.Am J Physiol 2000;279:C1144–C1154.

22. Skryma R, Mariot P, Bourhis XL, Van Coppenolle F, Shuba Y,Abeele FV, Legrand G, Humez S, Boilly B, Prevarskaya N. Storedepletion and store-operated Ca2þ current in human prostatecancer LNCaP cells: involvement in apoptosis. J Physiol (Lond)2000;527:71–83.

23. Barrett KE, Keely SJ. Chloride secretion by the intestinalepithelium: molecular basis and regulatory aspects. Ann RevPhysiol 2000;62:535–572.

24. Petersen OH. Stimulus-secretion coupling: cytoplasmic calciumsignals and the control of ion channels in exocrine acinar cells.J Physiol (Lond) 1992;448:1–51.

25. Ryan JS, Tao QP, Kelly ME. Adrenergic regulation of calcium-activated potassium current in cultured rabbit pigmentedciliary epithelial cells. J Physiol (Lond) 1998;511:145–157.

26. Maruyama Y, Peterson OH, Flanagan P, Pearson GT. Quanti-fication of Ca2þ-activated Kþ channels under hormonal controlin pig pancreas acinar cells. Nature 1983;305:228–232.

27. Sohma Y, Harris A, Wardle CJ, Gray MA, Argent BE. Maxi Kþ

channels on human vas deferens epithelial cells. J Membr Biol1994;141:69–82.

28. Hayashi T, Poronnik P, Young JA, Cook DI. The ACH-evoked,Ca2þ-activated whole-cell Kþ current in mouse mandibularsecretory cells. Whole-cell and fluorescence studies. J MembrBiol 1996;152:253–259.

29. Warth R, Hamm K, Bleich M, Kunzelmann K, Hahn TV,Schreiber R, Ullrich E, Mengel M, Trautmann N, Kindle P,Albrecht S, Greger R. Molecular and functional characterizationof the small Ca2þ-regulated Kþ channel (rSK4) of colonic crypts.Pflugers Arch 2000;438:437–444.

30. McVary KT, McKenna KE, Lee C. Prostate innervation. Prostate1998;8(Suppl):2–13.

31. Schwiebert EM, Kishore BK. Extracellular nucleotide signalingalong the renal epithelium. Am J Physiol 2001;280:F945–F963.

32. Ferguson DR, Kennedy I, Burton TJ. ATP is released from rabbiturinary bladder epithelial cells by hydrostatic pressure changes:a possible sensory mechanism? J Physiol (Lond) 1997;505:503–511.

33. Sauer H, Hescheler J, Wartenberg M. Mechanical strain-induced Ca2þ waves are propagated via ATP release andpurinergic receptor activation. Am J Physiol 2000;279:C295–C307.

210 Kimet al.