Endothelial Cell Sodium-Potassium-Chloride Cotransport

Val. 266,No. 18,Issue of June 25,pp. 11559-11566,1991 Printed in U,S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

(c)

Endothelial Cell Sodium-Potassium-Chloride Cotransport EVIDENCE OF REGULATION BY Ca2+ANDPROTEINKINASE

C* (Received for publication, June 14, 1990)

Martha E. O’DonnellS From the Department of Human Physiology, School of Medicine, University of California, Davis, California 95616

Previous studies have shown that vascular endothe-exhibit a highly active Na-K-Cl cotransport system. The Nalial cells exhibit a highly active Na-K-Cl cotransport K-Cl cotransporter mediates the major portion of total Kt system that is regulated by a variety of vasoactive influx in cultured bovine aortic endothelial cells (O’Donnell, hormones and neurotransmitters, suggesting that the 1989a). In addition, the magnitude of cotransport-mediated cotransporter may play an important role in endothe- K+ flux in these cells is largerelative to other cell types. lial cell function. In this study, the regulation of endo- Studies in this laboratory have shown that the endothelial thelial cell Na-K-Cl cotransport was further investi- cells express a comparatively high number of Na-K-CI cogated by probing the stimulus-transfer pathway by transporters as determined by [’Hlbumetanide binding which vasoactive agents stimulate the cotransporter. (O’Donnell, 198913). Our previous studies have also demonSpecifically, three peptides previously shown to stim11,vasopressin, strated that endothelialcell Na-K-Cl cotransportis regulated ulate cotransport activity (angiotensin and bradykinin) were evaluated.Na-K-Cl cotransport by a variety of vasoactive hormones and neurotransmitters was assessed in cultured bovine aortic endothelial cells that are known to regulate a number of vascular processes. as bumetanide-sensitive K+ influx. Stimulationof Na- Thus, endothelialcell cotransport is inhibited by atrial natriK-CI cotransport byangiotensin 11, vasopressin, or uretic peptide, acetylcholine, histamine, and norepinephrine 11, bradykinin was found to be reduced either by removal (O’Donnell,1989a). Otheragents,suchasangiotensin of extracellular Ca2+or by treatmentof the cells with vasopressin and bradykinin,have been foundto stimulate Na8-(N,N-diethylamino)octyl-3,4,5-trimethoxybenzoate K-Cl cotransportof these cells (Brock et al., 1986; O’Donnell, or 1,2-bis(o-aminophenoxy)ethane-N,N,Nf,N’-tetra- 1989a). Elevation of intracellular second messengers has also is, elevation acetic acid. In addition, the calmodulin antagonist W- been shown to regulate cotransport activity, that 7 was found to prevent stimulationof endothelial cell of either cyclic AMP or cyclic GMP causes inhibition of NaNa-K-C1 cotransport by the threepeptides. Thesefind- K-Cl cotransport-mediated Kt flux (O’Donnell, 1989a), ings suggest that regulationof endothelial cell cotrans- whereas treatmentof the cells with Ca2+ionophores to elevate port by these vasoactive peptides may be both Ca2+- intracellular Ca2+levels results in stimulationof the cotransand calmodulin-dependent. Angiotensin 11, vasopres- porter (Brock et al., 1986; O’Donnell, 1989a).The finding that sin,andbradykininwere also found to elevate the Na-K-Cl cotransporteris both highly active in endothelial phosphatidylinositol hydrolysis in the cultured endocells and regulatedby agents that modulate vascular function thelial cells. Thus, the possibility that regulation of suggests that the cotransporter may play an important role endothelial Na-K-Cl cotransport by these vasoactive in endothelial cell function. peptides also involves diacylglycerol activation of proIn this study, the regulation of Na-K-C1 cotransport in tein kinase C was investigated. A 10-min exposure of vascular endothelial cells was further investigated. Specifithe endothelial cells to lowdoses of phorbol 12-myristate 13-acetate was found to reduce Na-K-Cl cotrans- cally, the stimulus-transfer pathway whereby angiotensin 11, port whether in the presence or absence of angiotensin vasopressin, and bradykinin elevate cotransport activity was 11, vasopressin, or bradykinin. However, down-regu- examined. Some information regarding putative second meslation of protein kinaseC by a 40-h exposure to higher sengers induced by these vasoactive peptides has been prodoses of the phorbol ester was found to elevate Na-K- vided by previous studies. In cultured vascular endothelial cells, both bradykinin andvasopressin have been reported to C1 cotransport activity under both control and agoniststimulated conditions, indicatingthatactivation of elevate intracellular Ca2+ levels (Brock et al., 1986; Coldenprotein kinaseC results in inhibition of endothelial cell Stanfield et al., 1987; Morgan-Boyd et al., 1987), and bradyNa-K-Cl cotransport. Thus, protein kinase C activation kinin has been found to increase phosphatidylinositol hydrolmay serveas negative feedback in the stimulus-trans- ysis and productionof inositol phosphates (Aiyar et al., 1987; fer pathwayby which these agonists regulate endothe-Lambert et al., 1986). More information about second meslial cell Na-K-CI cotransport. sengers generated by vasopressin and angiotensinI1 has been provided by studies in other cells, including vascular smooth muscle cells. Thus,cultured vascular smooth muscle cells have been shown to respond to vasopressin and angiotensin Vascular endothelial cells previously have been shown to I1 withariseincytosolic Ca2+ concentration and also an increase in phosphatidylinositol hydrolysis (Alexander et al., * This work was supported by National Heart, Lung, and Blood 1985; Colden-Stanfield et al., 1987; Nabika et al., 1985; Smith Institute Grant HL-31959 and the American Heart Association of et al., 1984). These findings suggest that phosphatidylinositol Metropolitan Chicago. The costs of publication of this article were hydrolysis and elevation of cytosolic Ca2+ concentrationmay defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 mediate the effects of angiotensin 11, vasopressin, and bradykinin on Na-K-Cl cotransport of endothelial cells. Previous U.S.C. Section 1734 solely to indicate this fact. $ To whom reprint requests should be addressed. studies from this laboratory have also shown that treatment

11559

11560

Endothelial Cell Cotransport Nu-K-C1

Regulation

a 95% air, 5% CO, atmosphere with vitamin-free MEM containing 10% fetal bovine serum and [3H]myoinositol (1pCi/ml). To measure inositol phosphaterelease, [3H]myoinositol-labeled cellswere washed three times with Hanks' balanced salt solution containing Ca2+,M P , NaHC03, and20 mM Hepes (HBSS) and thenwere preincubated for 60 min a t 37 "C in an air atmosphere with HBSS containing10 mM LiCI. Li+ waspresent in this preincubation medium to inhibit inositol1-phosphatase. The medium was then changed to HBSS containing LiCl and 100 nM angiotensin 11, vasopressin, or bradykinin; and the cells were incubated for 5 min. To terminate the assay, cells were rapidly rinsed inice-cold isotonic MgC12, and then10% trichloroacetic acid was added to the dishes. The trichloroacetic acid extracts were washed five times with an equal volume of diethyl ether to remove trichloroacetic acid. The lower phase of each washed extract was EXPERIMENTAL PROCEDURES neutralized by addition of sodium tetraborate (to 5 mM) and then Materials-Bumetanide was the gift of Hoffmann-La Roche. An- transferred toDowex 1formate columns and washed twice with water. giotensin 11, vasopressin, bradykinin, phorbol12-myristate 13-acetate Labeled inositol phosphates present in the extracts (inositol, glycerol (PMA),' 1,2-oleoylacetylglycerol (OAG), 1,2-dioctanoylglycerol phosphoinositol, inositol 1-phosphate, inositol 1,4-bisphosphate, and (diC,), 4a-phorbol 12,13-didecanoate (4a-PDD),andEGTA were inositol 1,4,5-trisphosphate, respectively) were separated by sequenobtained from Sigma. Ouabain was obtained from Boehringer Mann- tial elution with: ( a ) water; (6) 5 mM disodium tetraborate, 60 mM (TMB- sodium formate; (c)100 mM formic acid, 200 mM ammonium formate; heim. 8-(N,N-Diethylamino)octyl-3,4,5-trimethoxybenzoate 8) was purchased from Aldrich. 1,2-Bis(o-aminophenoxy)ethane-N, ( d ) 100 mM formic acid, 500 mM ammonium formate; and (e) 100 N,N',N'-tetraacetic acid (BAPTA) was purchased from Molecular mM formic acid, 1 M ammonium formate. Aliquots of each elution Probes, Inc. (Eugene, OR). The calmodulin antagonists W-7 (N-(6- were evaluated for "H content by liquid scintillation counting. The aminohexyl)-5-chloro-l-naphthalenesulfonamide) andW-5(N-(6protein content of identical dishes was determined fluorometrically aminohexy1)-1-naphthalenesulfonamide) were obtained from Seikausing SDS extracts asdescribed below. gaku America (St. Petersburg, FL).asRb+was purchased from AmerFluorometric Protein Assay-The protein content of the culture sham Corp. Dulbecco's modified Eagle's medium (DMEM) was from dishes containing endothelial monolayers used for the transport exHazelton Biologics (Lenexa, KS), and fetal bovine serum was from periments and the inositol phosphaterelease studies was determined HyClone Laboratories (Logan, UT). using a fluorometric assay (Avruch and Wallach, 1971). Cells were Cells-Cultured bovine aortic endothelial cells were passaged from extracted with 0.2% SDS, and the fluorescence of the extract was calf thoracic aorta as described by Gordon and Martin (1983). Cells assessed at an excitation wavelength of 280 nm and an emission were grown in DMEM containing 10% fetal bovine serum a t 37 "C in wavelength of 340 nm. The protein concentrations of the SDS exa 95% air, 5% COe atmosphere andwere used between passages5 and tracts were calculated from a standard curveusingbovine serum 25. For both transport experiments and inositol phosphate measure- albumin. ments, cells were removed from stock culture dishes (100 mm) by A T P Measurements-ATPwasmeasuredenzymatically (Bergtrypsinization andwere subcultured onto 60-mm dishes, with a split- meyer, 1974). Bovine aortic endothelial cells were subcultured onto ting ratio of -1:5. Cells were used 3-5 days later as confluent mono- 60-mm culture dishes asdescribed above. Cells were incubated for 60 layers. min a t 37 "C in an air atmosphere in Hepes-buffered medium conNa-K-Cl CotransportMeasurements-Na-K-C1 cotransport was taining 5 +M BAPTA/AM and also either0 or 10 mM pyruvate. The measuredasouabain-insensitive,bumetanide-sensitiveK+influx medium was then aspirated, and the dishes were rinsed in ice-cold using "'Rb' as a tracer for K'. Details of this method havebeen cells as described by 0.1 M MgCl,. ATP was extractedfromthe published previously (O'Donnell, 1989a). Endothelial cells were first Flatman (1988) using ice-cold 10 mM EDTA plus 0.2% Triton X-100 equilibrated for 5min withHepes-buffered minimal essentialmedium to lyse the cells, followed by precipitation of protein by 2.8 M per(MEM) in an air atmosphere at 37 "C in a gyratory water bath. The chloric acid. Samples were kept onice throughout theprocedure. The cells were then preincubated for a second 5-min period at 37 "C in ATP contentsof the resulting extracts were then determined spectroHepes-buffered MEM containing1 or 0 mM ouabain +. 10 p~ bume- photometrically. Samples of extract were added to a Tris-buffered tanide and other agents to be tested for their effects on Na-K-C1 solution containing glucose, NADP+, and a mixture of hexokinase cotransport as indicated in the figure legends. To assay Na-K-CI plus glucose-6-P04 dehydrogenase. The ATP content of the sample cotransport, the medium was replaced with identical fresh Hepeswas determined from the increase in absorbance due to generation of buffered MEMcontaining *'Rb+ (1 pCi/ml),andthe cells were NADPH following enzymatic conversion of ATP plus NADP+ to incubated for 5 min. The assay was terminated by aspirating the ADP and NADPH (Bergmeyer, 1974). assay medium and rapidly rinsing the dishes in three 1-liter volumes of ice-cold isotonic MgC1,. After air drying, the contentsof the dishes RESULTS were extractedwith 0.2% sodium dodecyl sulfate (SDS), and the amount of radioactivity present in each extract was determined using Treatment of cultured bovine aortic endothelial cells with a liquid scintillation counter."Rb' uptake was calculated as theslope angiotensin 11, vasopressin, or bradykinin was found to stimof the uptake uersus time plot as described previously (Owen and ulate Na-K-C1 cotransport in a dose-dependent manner, as Prastlin, 1985). "6Rb+uptake by bovine aortic endothelial cells has been demonstrated previously toremainlinear for up to 10 min shown in Fig. 1. Cotransport was assessed in these studies as (O'Donnell, 1989a). Protein contents of the dishes were determined ouabain-insensitive, bumetanide-sensitive K+ influx, with "Rb+ used as a tracerfor K', as described under "Experimenspectrofluorometrically as described below. The K+ influx data are expressed as micromoles of K+/gram of protein/minute. tal Procedures." Angiotensin 11, vasopressin, and bradykinin Inositol Phosphate Release Measurements-Inositol phosphate re- were all found to be potent stimulators of endothelial cell Nalease was measured according to a modification of the method of K-C1 cotransport, with K./, values of -1, -4, and -1 nM, Berridge (1984) as previously described by Owen (1986). Endothelial cell monolayers on 60-mm culture dishes were incubated for 48 h in respectively. For all three agonists, maximal stimulation of

of cultured endothelial cells with activators of protein kinase C , such as phorbol 12-myristate 13-acetate,results in an inhibition of Na-K-C1 cotransport (O'Donnell et al., 1989a). Thus, it is possible that activation of protein kinase C by diacylglycerol, a product of phosphatidylinositol hydrolysis, is also involved in regulation of endothelial Na-K-Cl cotransport by angiotensin 11, vasopressin, and bradykinin. This study was conducted to evaluate the involvement of intracellular Ca2+and protein kinase C in the regulation of cotransport by these vasoactive peptides.

Na-K-C1 cotransport was observed at a concentration of 100 ' The abbreviations used are: PMA, phorbol 12-myristate 13-ace- nM. This finding confirms and extends a previous report by tate; OAG, 1,2-oleoylacetylglycerol;dlCs, 1,2-dioctanoylglycerol; 401- Brock et al. (1986) that bradykinin and vasopressin stimulate ouabain-insensitive, furosemide-sensitive K+ influx in endoPDD, 4a-phorbol 12,13-didecanoate; TMB-8, 8-(N,N-diethylamino)octyl-3,4,5-trimethoxybenzoate;EGTA, [ethylenebis(oxythelial cells. The magnitude of cotransport stimulation proethylenenitrilo)]tetraacetic acid BAPTA, 1,2-bis(o-aminophenduced by either vasopressin or bradykinin was found to be 0xy)ethane-N,N,N',N'-tetraacetic acid AM, acetoxymethylester; greater than thatproduced in response to angiotensin 11. This DMEM, Dulbecco's modified Eagle's medium; MEM, minimal essential medium; SDS, sodium dodecyl sulfate; Hepes, N-2-hydroxyethyl- relatively smaller stimulation of Na-K-C1cotransport induced by angiotensin I1 treatment of the endothelial cells was a piperazine-N'-2-ethanesulfonicacidHBSS,Hanks'balancedsalt consistent observation of this study. solution.

11561

Bll*llI

Y

0

11

10 9 8 7 Vasopressin (-log M)

6 -

Ang II

VP

BK

FIG.2. Effectsof TMB-8 and Ca2"-freemedium on enilothelial cell Na-R-Cl cotransport stimulated by angiotensin E, vasopressin, and bradykinin. Endothelia1 cells were subcultured onto 60-mm culture dishes as described under experimental Procedures." Cells were preincubated for 5 min in ~epes- buffer^ MEM c o n ~ i n i 0n or ~ 50 NM TMB-8. The cells were then p r e ~ c u b a ~ford a second €"in period either with Ca2+-freeHepes-buffered MEM plus 1 mM EGTA or with regular Wepes-buffered MEM (with 1.3 mM Ca2+)& 50 p~ TMB-8. Both media also contained 1 mM ouabain t 10 p~ bumetanide and 0 or 100 nM a n ~ o t e n s ~IIn (Ang TI), vasopressin (VP), or bradykinin (BIC) as indicated, Cells were assayed for 5 min by replacing the preincubation medium withfresh identical medium containing ffiRb+(1pcifml). Data are expressed as ouabaininsensitive, bumetanide-sensit~~e K+ influx. Values represent means f S.E. of ~ a d ~ p l i c a ~ d e ~ r ~from ~ a five t i oseparate ns experiCa2"-free; a, TMB-8.

basal cotransport, by 73%. Na-K-Clcotranspo~activity was also found to be reduced by Ca2'-free medium or TMB-8 in the presence of 100 nM angiotensin 11,vasopressin, or bradyf ~ fflepes-buffered kinin. In medium containing 1.3 l ~ Ca2+ MEM), cotransport activity was stimulated from 19.8 to 27.3, 37.6, and 32.9 @molof K+/g of p~tein/minby angio~nsin11, vasopressin, and bradyk~in,respectively. Treatment of cells with the Caz+-free medium reduced cotransport activity in the presence of each of these agonists. However, Ca2+-freemedium reduced Na-K-C1 c o t r ~ s p o ractivity t in the presence of vasopressin or bradykin~by 81 and 83%, respectively, but Y only decreasedcotransport by 38% in the presence of angio0 tensin 11. In contrast, TW-8 was found to cause a 68% Bradyki~in(-log M) decrease in cotransport activity in thepresence of angio~nsin RG. 1, Doas r e ~ p oof~ es n d ~ h cell e ~Na-B-Cl ~ atrans- 11, but only 18 and 17% decreases in the presence of vasoport stirnulatian by angiotensin If, vasopressin, and brady- pressin and bradykinin, respectively. One possible explanakinin. Endothelial cells were cultured and subcultured onto 60-mm tion of these data is that vasopressin and bradykinin may culture dishes as described under "Experimental Procedures," Cells promote Ca2+influx through plasma me~brane ea2+ chanwere preincuba~dfor 5 min in ~epes-bufferedMEM c o n ~ n 1~ g nels, whereas an~otensin II may have little or no effect on mM ouabain zl: 10 phA bume~nideand various concentrations of angiotensin I1 ( A ) ,vasopressin ( B ) ,or bradykinin (C) as indicated. transmembrane Ca'+ flux. To investigate this possibility, the effect of Caz+-free mediumon Na-K-CI cotransport was exT h e mediumwas then replaced with fresh medium of identical composition but c o n ~ i n ~ n=Rb+ g (1 pCi/ml), and the cells were amined in the presence of the divalent cation io~ophore assayed for 6 min. Na-K-CI cotransport was measured as ouabain- A23187. Whencells were treated withCa2+-free medium insen~tive, b~etanide-~nsitive K+ influx (with "Rb" as a tracer). containing EGTA,100 n&i angiotensin 11, and also 1 p~ Values represent means & S.E. of quadruplicate determinations from A23187, Na-K-Cl cotransport activity was reduced to the a representative experiment. same level as observed in the presence of Ca2+-free medium with EGTA plus vasopressin or b r a d y ~ i ~(data n not shown). As an initial approach to assess the involvement of Ca'" in The same was found to occur with Ca2+-free medium plus agonist-induced stimulation of endothelial cell Na-K-C1 co- EGTA and A23187 under basal conditions (i.e. in absence of transport, the effect of removing ca'+ from the extracellular the peptides). In contrast, addition of A23187 to cells in Ca2+medium on cotransport-mediated K+ influx was evaluated. free medium with EGTA plus vasopressin or bradykininwas To do this, the cells were preincubated and assayed in a Ca2+- not found to cause afurther decrease in c o t r a n s activity. ~~ free H e ~ s - b ~ f e r medium ed containing 1 mM EGTA. The Flatman (1987,1988) has shown recentlythat in ferret red effect of treating the cells with TMB-8,an agent reported to blood cells, Na-K-C1cotransport is regulated by the level of inhibit mobilization of intracellular Ca2+ (Chiouand Mala- free intracellul~M$+, and not Ca2+. Thus, to determine godi, 1975), was also evaluated. The results of these studies whether the decreased endotheli~cell cotransport activity are shown in Fig. 2. Under basal conditions ( i e .in the absence observed in Caz+-freemedium containing EGTA could bethe of the peptide agonists), incubating the cellsinCa2+-free result of decreased levelsof M e , rather than a lack of Ca2+, medium reduced catransport activity by 25%. Treatment of the effects of M$+-free medium werealso evaluated. Specifithe cells with 50 FM TMB-8 was found to markedly reduce cally, Na-K-Cl cotranspo~of the endothelial cells was as-

11562

Endothelial Cell Na-K-Cl Cotransport Regulation

sessedinmediumnominallyfree of Mg2+, but containing A Ca2+. Under these conditions, cotransport activity was not found to be reduced below the control levels shown in Fig. 2, whether in thepresence or absenceof the peptides, that is, in medium that is nominally M$+-free but contains 1.3 mM Ca'+, cotransport was found tobe 24.1 f 0.3 pmol of K+/g of protein/min under basal conditions and 27.0 f 2.0, 35.3 f 1.7, and 36.2 f 1.0 pmol of K+/g of protein/min in the presence of angiotensin 11, vasopressin, and bradykinin, respectively (data are from two separate experiments with quadruplicate determinations). The observation that TMB-8 has less of an inhibitoryeffect on cotransport in the presence of vasopressin or bradykinin than in the presence of angiotensin I1 could be due to using submaximal concentrations of the inhibitor. Thus, the potency of TMB-8 for inhibition of cotransport was evaluated in the presence of each of the three peptides. The results of these studies are shown in Fig. 3. The maximal inhibition of Na-K-Cl cotransport inducedby TMB-8 wasfound to be greater in the presence of angiotensin I1 than in the presence of vasopressin or bradykinin.K,,?values for TMB-8 inhibition of cotransport were determined t o be 0.7, 0.2, and 0.2 p~ in the presence of angiotensin 11, vasopressin, and bradykinin, respectively. Maximal inhibition of cotransport in vasopresTMB-8 (-log M) sin- and bradykinin-treated cells was found to occur with TMB-8 concentrations between 3 and 10 PM, well below the concentration of TMB-8 used in the studies of Fig. 2. To further examine the dependence of endothelial Na-KC1 cotransport activity on intracellular Ca2+, the effect of loading cells with the Ca2+buffer BAPTA to lower intracellular Ca'+ was evaluated, asshown in TableI. In these studies, endothelial monolayers were loaded with BAPTAby pretreating them for 60 min with 5 p~ BAPTA/AM, the permeant acetoxymethyl ester form of BAPTA. Previous studies have shown that intracellular cytoplasmic esterases convert BAPTA/AM to the nonpermeant BAPTA, trapping Ca'+ the X chelator within thecell (Tsien, 1980). When endothelial cells O L ' 7 ' 6 " 4 5 were loaded with BAPTA and then rinsed free of extracellular T M B I (-log M) BAPTA and subsequently assayed for Na-K-C1 cotransport FIG. 3. Dose responses of TMB-8 inhibition of endothelial activity in Ca2+-containingmedium, cotransport activitywas cell Na-K-C1 cotransport stimulated by angiotensin 11, vasofound to be reduced relative to control. BAPTA-loadedcells pressin, and bradykinin. Endothelial cells were subcultured onto were also found to exhibit significantly reduced cotransport 60-mm culture dishesas described under "Experimental Procedures." activity in thepresence of 100 nM angiotensin 11, vasopressin, Cells were preincubated for 5 min inHepes-buffered MEM containing The cells were then incubated TMB-8 at the indicated concentrations. or bradykinin relative to control (nonloadedcells). medium of the samecomposition but also containing It ispossible that loading the cells with BAPTAcould cause for 5 min in fresh reduced cellular levels of ATP due to inhibition of hexokinase 1mM ouabain +- 10 p~ bumetanide and 0 or 100 nM angiotensin (A 1, vasopressin ( B ) ,or bradykinin ( C ) .Finally, the medium was replaced by formaldehyde released during cleavage of BAPTA/AM to with identical freshmedium with R6Rb+(1pCi/ml), and thecells were BAPTA. Reduction of cellular ATP to low levels has been assayed for 5 min. Data are expressed as ouabain-insensitive, bumeshown to inhibit Na-K-C1 cotransport in a number of cell tanide-sensitive K+ influx. Values represent means f S.E. of quadruplicate determinations from a representative experiment. The K, types (Chipperfield, 1986). Thus, the BAPTA-induced inhibition of cotransport could be mediated in partby decreased values for TMB-8 inhibition of Na-K-Cl cotransport are 0.7 pM in the presenceof angiotensin 11, 0.5 pM in the presence of vasopressin, ATP levels. To evaluate thispossibility, Na-K-C1 cotransport and 0.3 p~ in the presenceof bradykinin. activity was also assessed in cells loaded with BAPTA in the presence of 10 mM pyruvate to prevent reduction of cellular the endothelialcells with BAPTA/AMwas not found tocause ATP levels. Pyruvate was found to have no effect on the a significant reduction in the levels of cell-associated ATP, BAPTA-inducedinhibition of cotransport,whetherunder whether in the presenceor absence of 10 mM pyruvate. basal conditions or in the presence of the three peptides. The above observations suggest that stimulation of Na-KCotransport activities of BAPTA plus pyruvate-treated cells C1 cotransport activity by angiotensin 11, vasopressin, and were found to be 17.0 f 1.2 pmol of K+/g or protein/min bradykinin occurs via a Ca2+-dependentprocess. To further under basal conditions and17.1 f 1.3, 22.9 f 1.0, and 22.4 f investigate the signal transduction pathway by which these 0.4 pmol of K+/g of protein/min in the presence of angiotensin agents stimulate Na-K-C1 cotransport activity in the endo11, vasopressin, and bradykinin, respectively ( n = 4). These thelial cells, the effect of calmodulin antagonists on cotransvalues are not significantly different from those found for port activitywere examined. In these studies, the effect of the cells treated with BAPTA alone. Additional experiments were potent naphthalenesulfonamide calmodulin antagonist W-7 was compared to the effect of W-5, a structurally similar done to evaluate effect the of BAPTA and pyruvate cellular on effect on calmodulinlevels of ATP. As shown in Table 11, a 60-min treatment of analog reported to have little or no

Endothelial Cell Na-K-Cl Cotransport Regulation

11563

TABLEI Effect of BAPTA on agonist-stimulated Na-K-C1 cotransport in endothelial cells Bovine aortic endothelial cells were subcultured onto 60-mm culture dishes as described under “Experimental Procedures.” Na-K-CIcotransport was determined as ouabain-insensitive, bumetanide-sensitiveK’ influx (with 86Rb+as tracer). Cells were loaded with BAPTA by pretreating them for 60 min at 37 “C in an air atmosphere in Hepes-buffered MEM containing 5 p~ BAPTA/AM. The cells were then washed once with Hepes-buffered MEM lacking BAPTA/AM and incubatedfor 5 min in the BAPTA-free medium.The medium was subsequently replaced with Hepes-buffered MEM containing 1 mM ouabain f 10 p M bumetanide and 0 or 100 nM angiotensin I1 (Ang II), vasopressin (VP), or bradykinin (BK); and the cells were incubated for a second 5-min period. To assay NaK-C1 cotransport, the medium was aspirated, and the cells were incubated for 5 min with fresh Hepes-buffered MEM identical in composition to the previous medium but also containing 86Rb+(1 pCi/ml). Values represent means f S.E. of quadruplicate determinationsfrom five separate experiments. All values for BAPTA-loaded cells are statistically different from control (nonloaded) cellsby the Student’s t test ( p < 0.01). K’ influx Basal

Ang I1

f 1.3 18.7 f 1.0

27.9 f 0.5 16.9 k 0.8

VP

BK

39.1 f 0.9

36.7 ? 1.5

25.0 f 1.2

21.1 ? 1.0

pmol/g proteinlmin

23.2 Control +BAPTA

TABLEI1 Effect of BAPTA and pyruvate onendothelial cell A T P levels Bovine aortic endothelial cells were subcultured onto 60-mm culture dishes as described under “Experimental Procedures.”Cells were loaded with BAPTA by pretreating them for 60 min at 37 “C in an air atmosphere in Hepes-buffered MEM containing 5 p M BAPTA/ AM and also either 0 or 10 mM pyruvate. ATP content of the cells was determined spectrophotometrically as described under “Experimental Procedures.” Values represent means k S.E. ( n = 6).

Previousstudies from thislaboratory haveshown that treatment of endothelial cells withthe biologically active phorbol ester PMA causes a reduction inNa-K-C1 cotransport activity (O’Donnell, 1989a).One interpretationof this finding is that PMA-induced activation of protein kinase C results in inhibition of cotransport activity. Exposure of the cells to angiotensin 11, vasopressin, and bradykinin would be predicted to alsocause an activationof protein kinase C through diacylglycerol release occurring with phosphatidylinositol hyATP drolysis; yet these three agonists all are potent stimulators of -Pyruvate +Pyruvate Na-K-Cl cotransport in endothelialcells. Thus, studies were pmol/dish cells conducted to further investigate the possible role of protein Control 479 f 78 512 f 77 kinase C in regulation of endothelial cell Na-K-C1 cotrans+BAPTA 478 f 57 435 f 69 port. The results of these studies are shown in Table IV. If PMA causes an inhibition of cotransport through activation dependent processes (Hidaka et al., 1981; Sakata et al., 1987). of protein kinase C, then other agentsknown to activate the T h e results of these studies are shown in Fig. 4. Na-K-Cl kinase should have similareffects on cotransport. Whencells cotransport was assessed in endothelial cells incubated with were incubated with thediacylglycerol analogs, OAG or diC8, Na-K-Cl cotransport activitywas found to be reduced below 100 nM angiotensin 11, vasopressin, or bradykinin plus increasing concentrations of W-7 or W-5. In eachcase, Na-K- basal levels. Significantinhibition of cotransport was obC1 cotransport activity was markedly reduced by W-7, from served in cells treated with PMA, OAG, or diCs at a concenthe agonist-stimulated level to values well below basal co- tration of 1 nM. Increasing the concentration of these three transport activity. The KIAvalues for w - 7 inhibition of co- agents to 50 nM was found to produce a further reduction in transport in the presence of angiotensin 11, vasopressin, or cotransport activity by diCs, but not by PMA or OAG. The bradykinin were determined from these data to be -30 PM in effect of a biologically inactive phorbol ester, 4a-PDD (Blumeach case. In contrast, W-5 was found to have no significant berg et al., 1984), was also evaluated. If the inhibitory effect result protein kinase effect on cotransport activity at concentrations up to 30 PM of PMA on Na-K-Cl cotransport is the of C activation, then phorbol esters thatdo not activate protein and a relatively small inhibitory effect at 50 PM. The possibility that angiotensin 11, vasopressin, and bra- kinase C would be predicted to have no effect on cotransport. dykinin alterNa-K-C1 cotransport activity through a pathway Consistent with this prediction, no reduction of cotransport involving protein kinase C was also evaluated in this study. activity was observed in the presenceof either 1 or 50 nM 4aBradykinin has been reported previously to increase phospha- PDD. tidylinositol hydrolysis in endothelial cells (Derian and MosT o further investigate whether PMA inhibition of endothekowitz, 1986; Lambert et al., 1986). Angiotensin I1 and vaso- lial Na-K-Cl cotransport is mediated by activation of protein pressin have been shown to elevate inositol phosphate release kinase C, the effects of protein kinaseC down-regulationwere et al., 1986, 1987; examined. It has been reported in vascular smooth musclecells(Aiyar previously that whereas shortNabika et al., 1985; Smith et al., 1984), although the effects term exposure of cells to PMA causes activation of protein of these two agonists on endothelial cell production of inositol kinase C, long-term exposure tohigh concentrations of PMA phosphates have not beenclarified. Thus,angiotensin 11, results in loss of protein kinase C activity as PMA-induced vasopressin, and bradykinin were evaluated for their effects association of protein kinase C with the plasma membrane on production of inositol phosphates by the cultured bovine appears to promote degradationof the kinase (Ballester and aortic endothelial cells. As shown in Table 111, whencells Rosen, 1985; Hepler et al., 1988; Rodriquez-Pena and Rozenwere treated with any of the three peptides, inositol phosphate gurt, 1984; Sakata et al., 1987). Thus, endothelial cells were release was elevated significantly above basal levels. Under treated either with400 nM PMA for 40 h or with 10nM PMA the experimental conditions employed in these studies, angio- for 10 min; and then Na-K-C1 cotransport activity was astensin 11, vasopressin, and bradykinin were found to cause sessed in the presenceor absence of angiotensin 11, vasopresincreases in inositol phosphate production of 70-109%. sin, and bradykinin. The results of these studies areshown in

Endothelial Cell Na-K-Cl Cotransport Regulation

11564

0

10

20 30 40 [Antagonist], pM

50

60

TABLE111 Effect of angiotensin 11, vasopressin, and bradykinin on formationof inositol phosphates i n endothelial cells Bovine aortic endothelial cells were subcultured onto 60-mm culture dishes. Three to five days later, cells were labeled with [3H] myoinositol for 48 h and then preincubated in Hanks’ balanced salt solution containing 10 mM LiCl for 60 min as described under “Experimental Procedures.” Angiotensin 11, vasopressin, and bradykinin were then added to the preincubation medium, and the cells were incubated for 5 min in a gyratory water bath at 37 “C. The incubation was terminated, inositol phosphates were extracted with 10% trichloroacetic acid and separated on aDowex 1formate column; and [3H]inositolphosphates were quantitated in aliquid scintillation counter. Dataare presented as total inositol phosphates(InsPs) (inositol phosphate, inositol bisphosphate, and inositol trisphosphate). Values represent means f S.E. of quadruplicate determinations from five separate experiments. Values obtained in the presence of the agonists are significantly different from basal values by the Student’s t test ( p < 0.01). Condition

Formation of InsPs

Basal Angiotensin 11 (100 nM) Vasopressin (1002273 nM) nM) Bradykinin (1002251

1088 f 100 1853 f 207 f 276 f 286

cpmldish

.o

lb

20 30 40 [Antagonist], pM

50

60

TABLEIV Effect of phorbol esters and diacylglycerol analogs on Na-K-Cl cotransport of endothelial cells Bovine aortic endothelial cells were subcultured onto 60-mm culture dishes, and Na-K-Cl cotransport was assessed as ouabain-insensitive, bumetanide-sensitive ffiRb+influx as described under “Experimental Procedures.” Cells were preincubated with PMA, OAG, diC8, or 401-PDD at the concentrations indicated in a Hepes-buffered medium for 5 min and then assayed for Na-K-C1 cotransport in the presence of the phorbol esters or diacylglycerol analogs as indicated. Values represent means f S.E. of quadruplicate determinations from two separate experiments. K’ influx

Condition

$ 1 00

10

20

30

40

50

pmoljg proteinfmin

60

[Antagonist], p M

FIG. 4. Effects of calmodulin antagonists on endothelial cell cotransport stimulated by angiotensin 11, vasopressin, and bradykinin. Endothelial cells were cultured as described under “Experimental Procedures.” Cells on 60-mm culture dishes were preincubated for 5 min in Hepes-buffered MEM containing various concentrations of W-7 (0)and W-5 (0)as indicated and also 1 mM ouabain f 10 PM bumetanide. The medium was then replaced with fresh medium of the same composition but also containing =Rb+ (1 +Ci/ml) and 0 or 100 nM angiotensin I1 (A), vasopressin ( B ) , or bradykinin ( C ) ; and the cells were incubated for 5 min. Data are expressed as ouabain-insensitive, bumetanide-sensitive K+ influx. Values represent means +. S.E. of quadruplicate determinations from a representative experiment.

Fig. 5. Under basal conditions (ie. in absence of the peptide agonists), a 40-h exposure of cells to 400 nM PMA was found to cause elevation of Na-K-Cl cotransport activity. Consistent with the data of Table IV, exposure of cells to 10 nM PMA for 10 min resulted in decreased cotransport activity. Similar responses were observedin the presence of the threeagonists, that is, cotransport was reduced by short-term PMA treatment in the presence of angiotensin 11, vasopressin, or bradykinin. However, long-term exposure of the cells to 400 nM PMA, even in the presence of the agonists, caused a significant elevation of cotransport activity above that observed with agonist alone. Thus, under conditions previously shown t o down-regulate protein kinase C, Na-K-C1 cotransport ac-

Basal 21.23 f 0.65 PMA 1 nM 16.60 +. 0.74“ 50 nM 16.01 f 1.26“ OAG 1nM 16.20 f 1.26” 50 nM 14.29 f 0.63” diC, 1 nM 18.17 f 0.47” nM 50 15.71 f 0.67” 4a-PDD 10 nM 21.52 f 1.56 50 nM 21.43 f 2.01 “Values differ significantly from basal values by the Student’s t test ( p < 0.05).

tivity is elevated, whether inthe presence or absence of vasoactive agents that stimulate the cotransporter. DISCUSSION

Inthis study, evidence is provided that stimulation of endothelial cell Na-K-C1 cotransport by angiotensin 11, vasopressin, and bradykinin may occur via a Ca2+dependent pathway. Each of these three peptides was found to be a potent stimulator of Na-K-C1 cotransport activity when cells were assayed in a Hepes-buffered MEM containing Ca2+. When cells were placed in a Ca2+-free medium containing EGTA, however, cotransport activity was reduced below the basal level, whether in the presence or absence of angiotensin

Endothelial Cell Nu-K-Cl Cotransport Regulation

11565

fect of Ca2+-freemedium (with EGTA) in cells exposed to angiotensin 11, but not in cellsexposed to vasopressin or bradykinin (data not shown). This finding is consistent with the hypothesis that the observed differences in endothelial cell Na-K-Cl cotransport responses to Ca2+-free medium and TMB-8 in the presence of the three agonists may be due to differences in the extent of agonist-induced Ca2+channel opening. The observation that the maximal stimulation of cotransport induced by angiotensin I1 is lower than that induced by vasopressin and bradykinin isalso consistent with the possibility that the lattertwo peptides induce Ca2+influx Basal Ang II VP BK FIG. 5. Effect of PMA on Na-K-CI cotransport of endothe- to a greater extent thandoes angiotensin 11. The finding that TMB-8 treatment reduces endothelial cell lial cells in presence of angiotensin 11, vasopressin, and bradykinin. Endothelial cells were culturedandsubculturedas de- cotransport also suggests that Na-K-C1 cotransport activity scribed under "Experimental Procedures." At 40 h prior to assay of is dependent on intracellular Ca2+. This is further supported Na-K-CIcotransport,theculture medium (DMEM + 10% fetal by the observation that loading endothelial cells withBAPTA bovine serum) was replaced with freshmedium containing either0 or 400 nM PMA. On the day of the experiment,cells were preincubated t o chelate intracellular Ca2+ abolished stimulation of Na-KC1 cotransport by the threevasoactive peptides in the presence for 5 min in Hepes-buffered MEM containing1 mM ouabain f 10 p M bumetanide f 10 or 400 nM PMA and also 0 or 100 nM angiotensin of extracellular Ca2+ and also reducedbasal cotransport activI1 (Ang IZ), vasopressin ( V P ) ,or bradykinin ( B K ) .The medium was ity. Thus, the results of this study indicate that angiotensin then replaced with identical medium containing =Rb+, and the cells 11, vasopressin, and bradykinin stimulate Na-K-Cl cotranswere assayed for 5 min. Data are expressed as ouabain-insensitive, portactivity of endothelial cellsvia a mechanism that is bumetanide-sensitive K+ influx. Values represent means ? S.E. of quadruplicatedeterminations from four separate experiments. 0, dependent on intracellular Ca2+. Whether the stimulus-transfer pathway specifically involves a rise in intracellular free control; 0, PMA (10min); @, PMA (40 h). [Ca"] remains to beclarified. This studyalso provides evidencethat stimulationof endo11, vasopressin, or bradykinin. This finding suggests that both stimulation of Na-K-C1 cotransport activityby these peptides thelial cell Na-K-Cl cotransport by angiotensin 11, vasopresand maintenance of basal cotransport activity are Ca2+-de- sin, and bradykinin is calmodulin-dependent. Treatment of endothelial cells with the calmodulin antagonist W-7 caused pendent. a dose-dependent inhibition of Na-K-C1 cotransport activity, The reduction in cotransport activityobserved in Ca"-free of angiotensin 11, vasopressin, or medium was more pronounced in the presence of vasopressin whether in the presence bradykinin. In contrast, W-5, a structurally similar analogof or bradykinin than either in the presence of angiotensin I1 or was found tohave little under basal conditions. Treatment of the cells with TMB-8 W-7 without anti-calmodulin activity, t o block mobilization of intracellular Ca2+,however, caused a or no effect on Na-K-Cl cotransport,suggesting that cotransmarked reduction of cotransport activity under basal or an- port inhibition by W-7 is due to its anti-calmodulinactivity. giotensin 11-stimulated conditions, butonly a modest decrease Other calmodulin antagonists (trifluoperazine, chlorpromain the presence of vasopressin or bradykinin. It is possible zine, and imipramine) were also found to inhibit endothelial that the differenceobserved intheresponse of Na-K-Cl cell Na-K-C1 cotransport activity (data not shown). These cotransport to the three peptides in the presence of Ca2+-free agents inhibited cotransport in the presenceof 100 nM vasomedium uersw TMB-8 is due to differing effects of the three pressin with a potency order of trifluoperazine > chlorpromof cotranspeptides on Ca'+ channel conductances of the cells. Several azine > imipramime, suggesting that the inhibition previous studies support this possibility. Angiotensin I1 and port by these agents is due to their anti-calmodulin activities (Owen and Villereal, 1982). Thus, elevationof endothelial cell vasopressin have been reported t o increase Ca2+ influx via Ca2+ channels in vascular smooth muscle cells (Smith and Na-K-Cl cotransportby angiotensin 11, vasopressin, and braSmith, 1987; Wallnoffer et al., 1987). Bradykinin has been dykinin appears to involve Ca2+- and calmodulin-dependent reported to increase Caz+ influx via channels in endothelial processes. As occurs in a number of other Ca'+- and calmodcells (Colden-Stanfieldetal., 1987; Johns etal., 1987; Morgan- ulin-dependent systems (Palfrey et al., 1982), it is possible that stimulation of Na-K-Cl cotransport in the endothelial Boyd et al., 1987; Schilling et al., 1988), althoughlittleis may involve phosphorylation by Ca2+/calmodulinknown about theeffects of angiotensin I1 and vasopressin on cells endothelial cell Ca2+ channels.It is possible that vasopressin dependent kinase. Studies of Na-K-C1 cotransport in other and bradykinin could increase endothelial cell Ca2+ channel cell types have provided evidence suggesting that regulation conductance but that angiotensin I1 may have little or no of cotransport involves phosphorylation of regulatory proteins effect on Ca2+channels.Underthesecircumstances,it is or of the transporter itself (Haas, 1989). The possible involvement of protein kinase C in regulation predictedthatexposure of the cells t o Ca2+-freemedium containing EGTA in the presenceof vasopressin or bradyki- of endothelial cell Na-K-CIcotransportactivity was also nin would cause a decrease in intracellular Ca2+levels as the investigated. First, an evaluationof the effects of angiotensin 11, vasopressin, and bradykininverified that all three peptides ion exited the cell down its concentration gradient through Ca2+ channels. In the presence of angiotensin 11, intracellular elevatephosphatidylinositol hydrolysis intheendothelial Ca2+levels would not be lowered to the same extent (due to cells. Thus, it is predicted that these peptides would both less efflux via Ca2+ channels). In addition, TMB-8 inhibition elevate intracellular Ca'+ levels from inositol trisphosphateof intracellular Ca2+mobilization would be predicted to have sensitive intracellular pools and also cause activation of proteinkinase C via elevation of diacylglycerol, ashas been little effect in the presence of vasopressin and bradykinin because intracellular Ca2+ levels would be elevated as Ca2+ demonstrated ina variety of other cell types (Berridge, 1984). entered the cell through Ca'+ channels. Additional studies In the cultured aortic endothelial cells, it appears that actidone to investigate this possibility showed that the divalent vation of protein kinase C causes inhibition of cotransport cation ionophore A23187 greatly increases the inhibitory ef- activity. This is supported by the observation that agents

11566

Endothelial Cell Cotransport Nu-K-C1 Regulation

known to activate protein kinase C, such as PMA, OAG, and dice, all inhibitcotransport activity, whereas 4a-PDD,a phorbol ester that does not activate protein kinase C, has no effect on cotransport activity. Furthermore, down-regulation of protein kinase C by prolonged exposure of the endothelial cells to PMA resulted in elevation of Na-K-C1 cotransport activity, whether under basal conditions or in thepresence of the vasoactive peptides. These findings indicate that activation of protein kinase C does indeed lead to an inhibition of endothelial cell Na-K-C1 cotransport activity. Thus,it is predicted that the actions of angiotensin 11, vasopressin, and bradykinin on endothelial cells would be 2-fold. Elevation of intracellular Caz+ levels via production of inositol trisphosphate would stimulate cotransport activity, and activation of protein kinase C via diacylglycerol would tend to inhibit the cotransporter. In this way, protein kinase C may exert negative feedback in the stimulus-transfer pathway by which these peptide agonists regulate endothelial cell Na-K-C1 cotransport. Consistent with this is the observation that even when cotransport is maximally stimulated by any of thethree peptides, down-regulation of protein kinase C causes a further enhancement of cotransport activity. The mechanism by which protein kinase C activation could cause Na-K-Cl cotransport inhibition is unknown. Studies in vascular smooth muscle cells have shown that phorbol ester-induced activation of protein kinase C can inhibit angiotensin 11- and vasopressin-induced phosphatidylinositol hydrolysis (Aiyar et al., 1987; Brocket al., 1985).Thus, itis possible that theinhibitory effect of protein kinase C activation on endothelial cell cotransport maybe due, in part,to a decrease in inositol trisphosphate-induced mobilization of intracellular Ca2+. However, it is also possible that Na-K-Cl cotransportactivity could be inhibited by activation of protein kinase C through phosphorylation of either the transporteritself or a regulatory protein. In this regard, it has been demonstrated previously in this laboratory that brief exposure of endothelial cells to PMA causes a reduction in the number of [3H]bumetanidebinding sites (O’Donnell, 1989b). This suggests that activation of protein kinase C in the endothelial cells has a direct effect on the Na-K-C1 cotransporter, either decreasing the number of transporters present in the plasma membrane or reducing the ability of bumetanide to bind to thetransporter. Recent studies from this laboratory have shown that endothelial cell Na-K-C1cotransport is also stimulated by elevated extracellular tonicity (O’Donnell, 1989~).Whether the signal transduction pathway by which hypertonicity stimulates the cotransporter may also occurvia a Ca2+- and calmodulindependent mechanism is currently under investigation. Acknowledgment-I wish to thank Kim Steutermann for her excellent technical assistance.

REFERENCES Aiyar, N., Nambi, P., Stassen, F. L., and Crooke, S. T. (1986) Life Sci. 39,37-45

Aiyar, N., Nambi, P., Whitman, M., Stassen, F. L., and Crooke, S. T. (1987) Mol. Pharmacol. 31, 180-184 Alexander, R. W., Brock, T. A., Gimbrone, M. A., Jr., and Rittenhouse, S. E. (1985) Hypertension (Dallas) 7,447-452 Biochim. Biophys. Acta 2 3 3 , Avruch, J., and Wallach, D. F. (1971) H. 234-237 Ballester, R., and Rosen, 0. M. (1985) J. Bid. Chem. 260, 1519415199 Bergmeyer, H. U. (1974) Methods of Enzymatic Analysis, Vols. 1-4, Academic Press, New York Berridge, M. J. (1984) Biochem. J . 220, 345-360 Blumberg, P. M., Jaken, S., Konig, B., Sharkey, N. A,, Leach, K. L., Jeng, A. J., and Yeh, E.(1984) Biochem. Pharmacol. 33, 933-940 Brock, T. A., Rittenhouse, S. E., Powers,C. W.,Ekstein, L. S., Gimbrone, M. A., Jr., and Alexander, R. W. (1985) J. Bid. Chem. 260, 14158-14162 Brock, T . A., Brugnara, C., Canessa, M., and Gimbrone, M. A., Jr. (1986) Am. J. Physiol. 250, C888-C895 Chiou, C. Y., and Malagodi, M. H. (1975) Br. J . Pharmacol. 53,279285 Chipperfield, A. R. (1986) Clin. Sci. (Oxf.) 7 1,465-476 Colden-Stanfield, M., Schilling, W. P., Ritchie, A. K., Eskin, S. G., Navarro, T., and Kunze, D. L. (1987) Circ. Res. 61,632-640 Derian, C. K., and Moskowitz,M. A. (1986) J . Bid. Chem. 2 6 1 , 3831-3837 Flatman, P. W. (1987) J. Physiol. (Lond.)386, 407-423 Flatman, P. W. (1988) J . Physiol. (Lond.)397, 471-487 Gordon, J. L., and Martin, W.(1983) Br. J. Pharmacol,. 79,531-541 Haas, M. (1989) Annu. Reu. Physiol,. 51, 443-457 Hepler, J. R., Earp, H. S., and Harden, T. K. (1988) J. Biol. Chem. 263, 7610-7619 Hidaka, H., Sasaki,Y., Tanaka, T., Endo,T., Ohno, S., Fujii, Y., and Nagata, T. (1981) Proc. Natl. Acad. Sci. U. S. A. 78,4354-4357 Johns, A,, Lategan, T. W., Lodge, N. J., Ryan, U. S., Van Breeman, C., and Adams, D. J. (1987) Tissue & Cell 1 9 , 733-745 Lambert, T. L., Kent, R. S., and Whorton,A. R. (1986) J . Bid. Chem. 261,15288-15293 Morgan-Boyd, R., Stewart, J.M., Vavrek, R. J., and Hassid, A. (1987) Am. J. Physiol. 2 5 3 , C588-C598 Nabika, T., Velletri, P. A., Lovenberg, W., and Beaven, M. A. (1985) J . Biol. Chem. 260,4661-4670 O’Donnell, M. E. (1989a) Am. J. Physiol. 2 5 7 , C36-C44 O’Donnell, M. E. (1989b) J. Biol. Chem. 264,20326-20330 O’Donnell, M. E. (1989~)J. Cell Biol. 1 0 9 , 314a (abstr.) Owen, N. E. (1986) J. Cell Biol 103,2053-2060 Owen, N. E., and Prastein, M. L. (1985) J . Bid. Chem. 2 6 0 , 14451451 Owen, N. E., andVillereal, M. L. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,3537-3541 Palfrey, H. C., Schiebler, W., and Greengard, P. (1982) Proc. Natl. Acad. Sci. U. S. A. 7 9 , 3780-3784 Rodriquez-Pena, A., and Rozengurt, E. (1984) Biochem. Biophys. Res. Commun. 1 2 0 , 1053-1059 Sakata, A,, Ida, E., Tominaga, M., and Onoue, K. (1987) Biochem. Biophys. Res. Commun. 1 4 8 , 112-119 Schilling, W. P., Ritchie, A. K., Navarro, L. T., and Eskin, S. G. (1988) Am. J. Physiol. 2 5 5 , H219-H227 Smith, J . B., and Smith, L. (1987) J . Biol. Chem. 262,17455-17460 Smith, J. B., Smith, L., Brown, E. R., Barnes, D., Sabir, M.A., Davis, J . S., and Farese, R. V. (1984) Proc. Natl. Acad. Sci. U. S. A. 8 1 , 7812-7816 Tsien, R. (1980) Biochemistry 19, 2396-2404 Wallnoffer, A., Cauvin, C., and Ruegg, U. T. (1987) Biochern.Biophys. Res. Commun. 148,273-278