Kidney International, Vol. 57 (2000), 2319–2333

Glucocorticoid modulates Na1/H1 exchange activity invascular smooth muscle cells by nongenomic andgenomic mechanisms

SHIGEAKI MUTO, SATORU EBATA, KOJI OKADA, TOSHIKAZU SAITO, and YASUSHI ASANO

Departments of Nephrology and Endocrinology and Metabolism, Jichi Medical School, Tochigi, Japan

caused no effect. Scatchard analysis of [3H]CORTI surfaceGlucocorticoid modulates Na1/H1 exchange activity in vascularbinding to VSMCs showed a single class of CORTI bindingsmooth muscle cells by nongenomic and genomic mechanisms.sites with a Bmax of 876.2 fmol per mg of cell protein and a KdBackground. In vascular smooth muscle cells (VSMCs),of 12.2 nmol/L. RU38486 also inhibited [3H]CORTI surfaceNa1/H1 exchange (NHE) plays an important role in intracellu-binding to VSMCs.lar pH (pHi) regulation. The genomic effect of glucocorticoid

Conclusions. In VSMCs, NHE activity is stimulated by(GC) on NHE activity has been suggested in VSMCs. However,short-term exposure to CORTI, but is inhibited by long-termthe nongenomic and genomic effects of GC on NHE activityexposure to CORTI. The short-term stimulatory effect ofand the underlying intracellular signaling mechanisms have notCORTI on NHE activity is independent of gene transcriptionyet been demonstrated in VSMCs. Also, it is not knownand protein synthesis, is mediated through the CORTI surfacewhether there are specific surface-binding sites of GC to thereceptor, and occurs through a microtubule-dependent process.plasma membrane of VSMCs.The long-term inhibitory effect of CORTI on NHE activityMethods. The effects of short (3 h)- and long (24 h)-termrequires gene transcription and protein synthesis and occursexposure to corticosterone (CORTI) on NHE activity wereonly through the cytosolic GR. The short- and long-term effectsstudied in cultured rat aortic VSMCs by using pHi measurementof CORTI on NHE activity occur via PKC activation. There-with the pH-sensitive fluorescent dye 2979-bis(carboxyethyl)-fore, CORTI differentially modulates NHE activity in VSMCs5(6)-carboxyfluorescein. The NHE activity was calculated fromby nongenomic and genomic mechanisms.the initial rate of Na1-dependent pHi recovery after the acid

load.Results. Short-term exposure of VSMCs to CORTI (1026

mol/L) increased NHE activity, whereas long-term exposure Na1/H1 exchange (NHE) is a plasma membrane trans-to CORTI decreased it. The inhibitors of gene transcriptionport protein found in a broad range of biological systems,(actinomycin D) and of protein synthesis (cycloheximide) didincluding vascular smooth muscle cells (VSMCs) [re-not affect the short-term effect of CORTI on NHE activity,

but inhibited the long-term effect of CORTI on NHE activity. viewed in 1, 2]. In VSMCs, the entry of Na1 into cellsThe cytosolic GC receptor (GR) antagonist (RU38486) inhib- in exchange for an intracellular H1 is the main effectited both the short- and long-term effects of CORTI on NHE of NHE, which therefore is involved in the control ofactivity, but the cytosolic mineralocorticoid receptor antagonist

intracellular pH (pHi), cell volume regulation, and initia-(spironolactone) did not influence either the short- or long-tion of cell growth and proliferation [3–6].term CORTI effects. Two protein kinase C (PKC) inhibitors

(staurosporine A and calphostin C) and PKC down-regulation Glucocorticoids (GCs) play a fundamental role in the[24-h pre-exposure to phorbol 12-myristate 13-acetate (PMA)] control of homeostasis, growth, and differentiation ininhibited both short- and long-term CORTI effects. Exposure VSMCs [7–11]. Long-term exposure of VSMCs to GCsto PMA for three hours mimicked the short-term CORTI ef-

inhibits cell growth [7–9], whereas brief exposure to GCsfect. The short-term CORTI effect was inhibited by the dis-stimulates it [9]. These reports raise the possibility thatruptor of microtubule (colchicine), but not by the disruptor of

filamentous-actin (cytochalasin B). The long-term exposure to GCs affect NHE activity in VSMCs.CORTI decreased NHE (NHE-1) mRNA levels to 0.65 times It is generally accepted that the delayed and slow ac-the control level, whereas the short-term exposure to CORTI tions (in hours or days) of GCs are mediated via the

classic genomic mechanism, which involves the activa-tion of its high-affinity cytosolic receptor and then theKey words: corticosterone, Na1/H1 exchanger, intracellular pH regula-

tion, cell growth, glucocorticoid surface receptor, protein kinase C. translocation of GCs-receptor complex to the cell nu-cleus. The activated GCs-receptor complex binds to spe-Received for publication August 24, 1999cific DNA sequences, termed GC response elements,and in revised form December 17, 1999

Accepted for publication January 20, 2000 resulting in positive or negative modulation of gene tran-scription [reviewed in 12, 13]. Berk et al reported that 2000 by the International Society of Nephrology

2319

Muto et al: Regulation of vascular NHE by GC2320

hydrocortisone stimulates NHE activity of VSMCs after mechanisms for the nongenomic and genomic CORTIeffects. We also determined whether there are specifica latency of at least four hours, presumably by the classic

genomic pathway of the steroid action [8]. However, surface binding sites of CORTI to the plasma membraneof VSMCs.they have not yet determined whether the stimulatory

effect of hydrocortisone on NHE activity occurs via cyto-solic GC receptor (CGR) and/or mineralocorticoid re-

METHODSceptor (MR) and is associated with increased NHE

Culture of rat VSMCsmRNA. Also, the intracellular signaling mechanisms forthe genomic effect of hydrocortisone on NHE activity Primary cultures of rat VSMCs were isolated from

thoracic aortae of male Sprague-Dawley rats (200 toare unclear.Besides these classic genomic mechanisms, it is now 250 g) by the treatment with collagenase, as described

previously in our laboratory [10, 11]. Briefly, the cellspossible to define the characteristics of any steroid actionthat represents a nongenomic effect [reviewed in 12, 14]. were cultured in Eagle’s minimum essential medium

(MEM) supplemented with 10% fetal bovine serumThey are instantaneous or with a very short lag periodand are insensitive to RNA and protein synthesis inhibi- (FBS), 26.1 mmol/L NaHCO3, 100 U/mL penicillin, and

100 mg/mL streptomycin in a 5% CO2 incubator at 378C.tors. They are direct steroid actions, not mediated throughnuclear receptor occupancy. There is increasing evidence After reaching confluence, VSMCs were passaged through

trypsin/ethylenediaminetetraacetic acid (EDTA) andfor rapid nongenomic steroid actions, including neuraleffects after local application of steroids and fast effects seeded at a 1:3 ratio in 100 mm culture dishes (Becton

Dickinson, Lincoln Park, NJ, USA). Experiments wereof steroids on the g-aminobutyric acid A receptor, onluteinizing hormone-releasing peptide secretion, on do- performed with VSMCs, which were grown to confluence

after 4 to 10 passages.pamine release, on oocyte maturation, and on the acro-some reaction in spermatozoa [14]. In our laboratory,

Measurements of pHi in VSMCswe have demonstrated that exposure of VSMCs to aldo-sterone for three hours activates NHE via a nongenomic For pHi measurements, cells were plated on 35 mm

Petri dishes containing a glass coverslip bottom (MatTec,mechanism and that the nongenomic stimulatory effectof aldosterone on NHE activity occurs through protein Ashland, MA, USA) and were used two to seven days

later. Cells were incubated in FBS-free MEM for 48 hourskinase C (PKC) activation and both microtubule- andfilamentous actin-dependent processes [6]. However, it before use. The pHi was continuously measured using the

pH-sensitive fluorescent dye 2979-bis(carboxyethyl)-5(6)-has not yet been established whether GCs actually havenongenomic effects on NHE activity in VSMCs. Further- carboxyfluorescein acetoxymethyl ester (BCECF-AM)

according to the methods described by Berk et al [22, 23]more, little is known about the intracellular signalingmechanisms for the nongenomic effect of GCs on NHE and as modified in this laboratory [6]. Prior to each

experiment, the cells grown to confluence were incu-activity in VSMCs.There have now been numerous reports of steroid- bated for 30 minutes in FBS-free MEM containing

BCECF-AM (2 mmol/L) at 378C. The Petri dish was thenbinding sites to the plasma membranes of many tissues[15–20], including kidney [15], liver [16], and neural tis- placed on the stage of an inverted epifluorescence micro-

scope (IMT-2; Olympus, Tokyo, Japan) and was thensue [18]. The identification of the steroid-binding sitesto the plasma membranes supports the hypothesis of continuously perfused with HEPES-buffered solutions

(described later in this article) by gravity. A water jacketnongenomic steroid action involving plasma membranesurface receptors distinct from the classic cytosolic ste- was used to maintain temperature in the dish at 378C.

A single-cell measurement of pHi was performed usingroid receptors. However, it is not known whether thereare specific surface binding sites of GCs to the plasma a microscopic fluorometer (OSP-3; Olympus) as de-

scribed previously in our laboratory [6]. Measurementsmembrane of vascular tissue.In previous studies, dexamethasone and hydrocorti- were made at 3100 magnification, and the diameter of

the beam of light focused on the single VSMC was ap-sone were used as GCs to examine the effects of GCson NHE activity and growth response in cultured proximately 7.5 mm. The light source was a 75 W xenon

lamp. The fluorescent dye was excited alternatively atVSMCs from rat and bovine aortae [7–9]. However, theyare not endogenous GCs in rats. In the present study, 440 and 490 nm by spinning the sector mirror at 300

r.p.m. and was measured at a wavelength of 530 nm.we chose corticosterone (CORTI) as a GC, since thissteroid is the natural and possibly sole GC in rats [21]. Because in this apparatus it takes 10 ms to obtain one

fluorescence ratio (I490/I440), the single VSMC was exposedTherefore, the present study used cultured VSMCsfrom rat aortae to determine whether there are nongeno- to light for one second to obtain one mean I490/I440. We

used only VSMCs that had at least a 20-fold greatermic and genomic effects of CORTI on NHE activityin VSMCs, and to determine the intracellular signaling fluorescence intensity than that of the background. Cali-

Muto et al: Regulation of vascular NHE by GC 2321

bration of the BCECF excitation ratio for each experi- fragment (nucleotides 478 to 1850) of the rat NHE-1cDNA (a generous gift from Dr. John Orlowski, McGillment was determined using the K1/nigericin technique

as described previously [6, 24]. University, Montreal, Canada) [28]. The filters were thenwashed twice for 10 minutes at room temperature in 2 3The NHE activity of VSMCs was assayed as the initial

rate of Na1-dependent pHi increase (dpHi/dt) after acid SSC and 0.1% sodium dodecyl sulfate (SDS) followedby a 10-minute wash at 458C in 0.1 3 SSC and 0.1%load using nigericin, a K1/H1 ionophore, in the absence

of CO2/HCO32 [6, 25, 26]. The composition of solutions SDS. Autoradiography was performed at 2708C and

quantitated by densitometric scanning (Image Analyzerused for the assay was as follows: (1) for HEPES-buf-fered Na1 solution, 130 mmol/L NaCl, 5 mmol/L KCl, TIF-64; Immunomedica, Tokyo, Japan). The filter mem-

branes were also hybridized with an [a-32P]-dCTP–1.5 mmol/L MgCl2, 1 mmol/L CaCl2, and 30 mmol/LHEPES; and (2) for Na1-free solution, Na1 was isosmoti- labeled rat glyceraldehyde-3-phosphate dehydrogenase

(GAPDH) probe. The NHE-1 mRNA area was stan-cally replaced with N-methyl-D-glucamine (NMDG).All solutions were adjusted to a pH of 7.4 at 378C. Once dardized for each lane in the GAPDH area to control

for variability in RNA loading.a stable fluorescence signal was reached, Na1 was re-moved by replacing NMDG. With removal of Na1, the

[1, 2-3H(N)]-CORTI ([3H]CORTI) surface receptorpHi decreased, and when Na1 was readded, there wasbinding to VSMCsrapid recovery of pHi to baseline levels. The dpHi/dt was

calculated from a line drawn tangential to the initial 30- The specific binding of [3H]CORTI to surface recep-tors was determined by using intact VSMCs and wassecond deflection after the return of Na1 to the perfusate.

Intracellular buffering capacity was determined from performed in binding buffer, which consisted of 140mmol/L NaCl, 5 mmol/L KCl, 1 mmol/L CaCl2, 0.5the pHi response to the removal of NH3/NH4

1 using thefollowing formula: intracellular buffering capacity (bI) 5 mmol/L MgCl2, 5 mmol/L glucose, 1 mmol/L Na2HPO4,

10 mmol/L Tris/HCl, and 0.2% bovine serum albuminD[NH41]i/DpHi, where [NH4

1]i is the intracellular con-centration just before NH3/NH4

1 removal, calculated as (BSA), pH 7.4. All steps were carried out at 48C toensure that internalization of the receptor-ligand com-[NH4

1]i 5 [NH41]o 3 10(7.4-pHi), and DpHi is the pHi change

on removal of NH41 [6, 27]. A change in dpHi/dt is plex would not occur. The experiments used confluent

monolayers of VSMCs grown in 35 mm culture dishesinterpreted to reflect a difference in NHE activity whenintracellular buffering capacities in two conditions are (Becton Dickinson). At the beginning of each experi-

ment, the culture medium was aspirated from the dishes,similar.and cells were then washed twice with 2 mL ice-cold

RNA isolation and Northern blot analysis binding buffer. Equilibrium binding studies were per-formed at 48C for 60 minutes in binding buffer containingVascular smooth muscle cells were cultured and grown

in 100 mm culture dishes (Becton Dickinson). Confluent various concentrations of [3H]CORTI from 8 3 10211 to1027 mol/L. Preliminary experiments demonstrated thatVSMCs were placed in serum-free MEM for 48 hours

and were then exposed to CORTI (1026 mol/L) for the under these conditions, equilibrium was reached. Theincubation volume was 0.9 mL. At the end of the incuba-next 24 hours. Thereafter, total RNA was isolated using

a single-step guanidine thiocyanate method as reported tion, the reaction mixture was aspirated, and the cellswere then rinsed four times with 2 mL ice-cold bindingpreviously [10, 11]. In brief, VSMCs were lyzed using

ISOGENe (Nippon Gene, Tokyo, Japan), which con- buffer. The attached cells were then dissolved with 1.5mL of 0.1% SDS and 0.1 N NaOH. Radioactivity intained phenol and guanidine isothiocyanate. The lysate

was extracted with chloroform/isopropanol and was the dissolved cells was counted by a liquid scintilationcounter (Aloka LSC-671, Tokyo, Japan). A portion ofwashed with 75% ethanol. Poly (A)1 RNA was purified

using oligo(dT)-latex column (Takara, Tokyo, Japan), the dissolved cells was used for determining the proteincontent [29]. Nonspecific binding was determined in theas described previously [6]. Equal amounts of Poly (A)1

RNA (5 mg/lane) were size fractionated on 1% denatur- presence of 1025 mol/L unlabeled CORTI. Receptor den-sity (Bmax values) and dissociation constant (Kd) foring agarose-formaldehyde gels and capillary blotted onto

nylon membrane filters (Hybond N; Amersham, Arling- [3H]CORTI were calculated from Scatchard plots [30]. Inaddition, a trypsin-dependent loss of [3H]CORTI surfaceton Heights, IL, USA) in 20 3 standard saline citrate

(SSC; 1 3 SSC contains 0.15 mol/L NaCl and 0.0015 binding was studied following exposure of the VSMCsto 0.25% trypsin for 30 minutes at 378C. Thereafter,mol/L sodium citrate, pH 7.0). The filters were prehy-

bridized for 30 minutes at 688C before hybridization us- VSMCs were incubated with 4 3 1029 mol/L [3H]CORTIin the absence or presence of 1025 mol/L unlabeleding Quick-Hybe (Stratagene, La Jolla, CA, USA), and

were then hybridized for one hour at 688C in the same CORTI at 48C for 60 minutes. After washing the bindingbuffer, the cells were rinsed with the ice-cold bindingsolution with 106 cpm/mL of [a-32P]-dCTP random

primer-labeled probes. The NHE probe was a Pst I-Pst I buffer to remove the unbound radioactive ligand. The

Muto et al: Regulation of vascular NHE by GC2322

count of surface radioligand was then measured aftersolubilizing the cells as described previously in this arti-cle. Specific binding was calculated as the difference be-tween total binding and nonspecific binding. To furthercharacterize the CORTI surface receptors in VSMCs,monolayers of VSMCs were also incubated with the fixed[3H]CORTI (4 3 1029 mol/L) and increasing concentra-tions of unlabeled CORTI or RU38486 (RU) from 10211

to 1025 mol/L at 48C for 60 minutes. After washing thebinding buffer, the cells were rinsed with the ice-coldbinding buffer to remove the unbound radioactive li-gand. The count of surface radioligand was then mea-sured after solubilizing the cells, as described previouslyin this article. All binding experiments were performedin triplicate and were repeated at least twice with similarresults.

Fig. 1. Effects of corticosterone (CORTI; 1026 mol/L) on baseline in-Materials tracellular pH (pHi) in vascular smooth muscle cells (VSMCs). VSMCs

were exposed to control Eagle’s minimum essential medium (MEM;All chemicals were obtained from Sigma Chemicalh) or CORTI-supplemented MEM ( ) for 3 or 24 hours before the

Co. (St. Louis, MO, USA) unless noted as follows. MEM, measurement of pHi. pHi measurements were carried out in the nominalabsence of CO2/HCO3

2. Data are expressed as means 6 SE. *P ,penicillin, streptomycin, and FBS were from Flow0.01; **P , 0.001 compared with control. The numerals in parenthesesLaboratories (Irvine, Scotland, UK). Ethylisopropylami- indicate number of experiments.

loride (EIPA) was from Research Biochemicals Incor-porated (Natick, MA, USA). HEPES was from Dojindo(Kumamoto, Japan). [a-32P]-dCTP was from Amersham.

RESULTS[3H]CORTI was from New England Nuclear (Boston,MA, USA), and BCECF-AM was from Molecular Effect of CORTI on pHi in VSMCsProbes Inc. (Eugene, OR, USA). RU 38486 (RU) was a We first examined the effects of CORTI on baselinegenerous gift from Roussel Uclaf (Romainville, France). pHi in VSMCs. VSMCs were placed in serum-free MEM

CORTI was dissolved in ethanol as a 1022 mol/L stock for 48 hours before the addition of the hormone andsolution and then diluted to final concentrations of 10211

were then exposed to CORTI (1026 mol/L) for 3 hoursto 1025 mol/L. RU was also dissolved in ethanol as a (short-term exposure) or 24 hours (long-term exposure).1021 mol/L stock solution and then diluted to final con- The results are illustrated in Figure 1. After short-termcentrations of 10211 to 1025 mol/L. Colchicine (COL) exposure of VSMCs to CORTI, the baseline pHi (7.12 6was dissolved in ethanol at 0.1% final concentration. 0.02, N 5 8, P , 0.01) was significantly greater than thatCycloheximide (CHX) was prepared as a 4 mg/mL stock in cells exposed to control MEM (7.04 6 0.03, N 5 8).solution in distilled water and diluted to a final concen- In contrast, after continuous exposure to CORTI for 24tration of 20 mg/mL. Actinomycin D (ACD) was pre- hours, the basal pHi of VSMCs (6.91 6 0.04, N 5 8, P ,pared as a 1 mg/mL stock solution in distilled water and 0.001) was significantly decreased when compared withdiluted to a final concentration of 4 mg/mL. EIPA was that in cells exposed to control MEM (7.06 6 0.01, N 5dissolved in methanol at 1021 mol/L and diluted to a 8). Thus, the short-term exposure to CORTI alkalinizedfinal concentration of 100 mmol/L. Spironolactone (SPR), VSMCs, whereas the long-term exposure to CORTIstaurosporine A (ST), calphostin C (CAL), phorbol 12- acidified them.myristate 13-acetate (PMA), and cytochalasin B (CYTO)

Effect of CORTI on NHE activitywere dissolved in dimethyl sulfoxide at 0.1% final con-centrations. Equivalent concentrations of vehicle were Since the NHE is an important mechanism for theadded as a control for individual protocols. maintenance and regulation of pHi in VSMCs [4], we

next examined whether the CORTI-induced changes inStatistical analysis pHi depend on changes in NHE activity. For this purpose,

Results are expressed as mean 6 SEM. Comparisons VSMCs were placed in serum-free MEM for 48 hourswere performed by Student’s t-test and one-way analysis before the addition of the hormone and were then ex-of variance (ANOVA) in combination with Fisher’s pro- posed to CORTI (1026 mol/L) for 3 or 24 hours. Cellstected least significant difference, as needed. P values were first bathed in the HEPES-buffered Na1 solution

(130 mmol/L Na). The solution was then changed to theof , 0.05 were considered significant.

Muto et al: Regulation of vascular NHE by GC 2323

Fig. 2. Typical fluorescent tracing of pHi showing Na1-dependent pHi recovery (dpHi /dt) in VSMCs exposed to CORTI (1026 mol/L) for (A) 3or (B) 24 hours in the absence or presence of 100 mmol/L EIPA. pHi measurements were carried out in the nominal absence of CO2/HCO3

2.Symbols are: (s) control; (d) CORTI; (h) control 1 EIPA; (j) CORTI 1 EIPA.

Na1-free solution (0 mmol/L Na) containing 10 mmol/Lnigericin for four minutes, which caused pHi to decreaseto ,5.9 to 6.1. Nigericin was then removed by resus-pending the cells in the Na1-free solution containing1% (wt/vol) BSA (which binds the ionophore) for twominutes. Cells were then bathed in the Na1-free solutionfor one minute. Thereafter, a subsequent addition ofNa1 to the extracellular solution caused a rapid increasein pHi, as shown in Figure 2; the rise was more rapidand greater in VSMCs exposed to CORTI (0.84 6 0.06pH/min, N 5 8, P , 0.01) for three hours than in cellsexposed to control MEM (0.64 6 0.05 pH/min, N 5 8).In sharp contrast, the Na1-dependent pHi increase afteracid load was inversely slower and smaller in VSMCs

Fig. 3. Time course of Na1-dependent pHi recovery (dpHi /dt) inducedexposed to CORTI (0.51 6 0.03 pH/min, N 5 8, P , by CORTI in VSMCs. VSMCs were exposed to control MEM or CORTI

(1026 mol/L)-supplemented MEM for various terms before the measure-0.01) for 24 hours than in cells exposed to control MEMment of dpHi/dt after a nigericin-induced intracellular acidosis. pHi(0.70 6 0.04 pH/min, N 5 8). The Na1-dependent pHi measurements were carried out in the nominal absence of CO2/HCO3

2.recoveries after acid load in cells treated with both con- Data are expressed as means 6 SE of eight separate experiments. *P ,

0.005; **P , 0.001 compared with control.trol MEM and with CORTI-supplemented MEM for 3and 24 hours were completely inhibited by 100 mmol/LEIPA, a specific NHE inhibitor (Fig. 2). On the otherhand, there were no significant differences of intrinsic

treated with CORTI for 3 hours, but was inhibited inbuffering capacity at pHi of 6.1 between control andVSMCs treated with CORTI for 24 hours.CORTI-treated VSMCs at 3 hours [control, 19.6 6 0.7

We next examined the time-dependent alterations in(mmol/L)/H1 · pH, N 5 5; CORTI, 18.8 6 0.9 (mmol/the NHE activity induced by CORTI (1026 mol/L). TheL)/H1 · pH, N 5 5] and 24 hours [control, 19.5 6 0.8results are shown in Figure 3. Upon the addition of(mmol/L)/H1 · pH, N 5 5; CORTI, 19.2 6 0.7 (mmol/CORTI to VSMCs, the dpHi/dt after acid load rapidlyL)/H1 · pH, N 5 5]. Therefore, the Na1-dependent pHi

increased to 1.3 times the control by as early as threerecovery after an acid load was dependent on the NHEactivity, and the NHE activity was stimulated in VSMCs hours. However, the CORTI-induced increase in dpHi/dt

Muto et al: Regulation of vascular NHE by GC2324

was not sustained. A significant difference of the dpHi/dtin cells treated with CORTI for 12 hours was not ob-served. Further long-term exposure of VSMCs toCORTI for 24 hours inversely caused a significant de-crease in dpHi/dt to 0.7 times the control level. Thus, theshort-term exposure to CORTI stimulated NHE activity,whereas the long-term exposure to CORTI inhibited it.

Dose-dependent effect of CORTI on NHE activity

We next examined the dose-dependent effects ofCORTI on NHE activity. VSMCs were placed in serum-free MEM for 48 hours and were then exposed to CORTIat various concentrations for 3 or 24 hours. As shownin Figure 4, both a CORTI-induced increase and de-crease in NHE activities were dose-dependent. Signifi-cant effects of CORTI on NHE activity were observedat 1027 to 1026 mol/L. In rats, plasma concentrations ofCORTI are in the range of 1027 to 1026 mol/L [31].Therefore, CORTI at physiological concentrations af-fects NHE activity. In the following experiments, wefocused on the mechanisms for the short- and long-termeffects of CORTI on NHE activity.

Effects of CHX and ACD on CORTI-inducedchanges in NHE activity

To determine whether protein synthesis and gene tran-scription are required for the CORTI-induced changesin NHE activity, VSMCs were incubated with CHX, aninhibitor of protein synthesis, or ACD, an inhibitor ofgene transcription. For these experiments, VSMCs wereplaced in serum-free MEM for 48 hours and were thenexposed to MEM treated with CHX (20 mg/mL) or ACD(4 mg/mL) plus CORTI (1026 mol/L) for 3 or 24 hours.This concentration of CHX (20 mg/mL) inhibited proteinsynthesis by 95%, as determined by [3H]leucine incorpo-ration [10], and the results are shown in Figure 5A (forCHX-treated VSMCs) and Figure 5B (for ACD-treatedVSMCs). CHX or ACD alone had no effect on NHEactivity. Either CHX or ACD did not influence the short-term stimulatory effect of CORTI on NHE activity,whereas both of them completely prevented the long- Fig. 4. Dose-dependent effects of CORTI on NHE activity in VSMCs.

VSMCs were exposed to CORTI at various concentrations for (A) 3term inhibitory effect of CORTI on NHE activity.or (B) 24 hours. pHi measurements were carried out in the nominalabsence of CO2/HCO3

2. Data are expressed as the percentage of dpHi/dtEffects of RU and SPR on CORTI-induced changescompared with those observed in VSMCs exposed to control MEM.

in NHE activity The number of experiments done on VSMCs treated with CORTI for3 and 24 hours is eight and five, respectively. *P , 0.05; **P , 0.005;To examine whether the effects of CORTI on NHE†P , 0.001 compared with control value.

activity occurs via cytosolic GR and/or MR, VSMCs wereplaced in serum-free MEM for 48 hours and were thenexposed to the cytosolic GR antagonist RU (1025 mol/L) NHE activity, whereas SPR did not affect either theor the cytosolic MR antagonist SPR (1024 mol/L), plus short-term or the long-term CORTI effects.CORTI (1026 mol/L) for 3 or 24 hours. The results were

Effect of PKC pathway modulation onshown in Figure 6A (for RU-treated VSMCs) and FigureCORTI-induced changes in NHE activity6B (for SPR-treated VSMCs). RU or SPR alone caused

no effects on NHE activity. RU completely inhibited Protein kinase C has been shown to play an importantrole in the regulation of NHE activity in many tissuesboth the short- and the long-term effects of CORTI on

Muto et al: Regulation of vascular NHE by GC 2325

Fig. 5. Effects of inhibitors of protein synthesis, cycloheximide (CHX), or gene transcription, actinomycin D (ACD), on CORTI-induced changesin NHE activity in VSMCs. pHi measurements were carried out in the nominal absence of CO2/HCO3

2. (A) VSMCs were exposed to controlMEM or CORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours in the absence or presence of CHX (20 mg/mL). Data are expressed as thepercentage of dpHi/dt compared with those observed in CORTI-untreated VSMCs without CHX (control). The number of experiments done onVSMCs treated with CHX for 3 and 24 hours is six and eight, respectively. (B) VSMCs were exposed to control MEM or CORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours in the absence or presence of ACD (4 mg/mL). Data are expressed as the percentage of dpHi/dt comparedwith those observed in CORTI-untreated VSMCs without ACD (control). The number of experiments done on VSMCs treated with ACD for 3and 24 hours is six and eight, respectively.

[1, 32], including VSMCs [6, 23, 33, 34]. Therefore, we 1027 mol/L), plus CORTI (1026 mol/L) for 3 or 24 hours.The results for ST-treated VSMCs are shown in Figureexamined whether PKC activation is responsible for the

CORTI-induced changes in NHE activity. For this pur- 7A and for CAL-treated VSMCs in Figure 7B. ST orCAL alone had no effect on NHE activity. On the otherpose, VSMCs were placed in serum-free MEM for 48

hours and were then exposed to MEM treated with one hand, ST and CAL completely inhibited the short-termstimulatory effect of CORTI on NHE activity and com-of two PKC inhibitors, ST (1027 mol/L) or CAL (5 3

Muto et al: Regulation of vascular NHE by GC2326

Fig. 6. Effects of the glucocorticoid receptor antagonist RU 38486 (RU) or the mineralocorticoid receptor antagonist spironolactone (SPR) onCORTI-induced changes in NHE activity in VSMCs. pHi measurements were carried out in the nominal absence of CO2/HCO3

2. (A) VSMCswere exposed to control MEM or CORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours in the absence or presence of RU (1025 mol/L).Data are expressed as the percentage of dpHi/dt compared with those observed in CORTI-untreated VSMCs without RU (control). The numberof experiments done on VSMCs treated with RU for 3 and 24 hours is six and eight, respectively. (B) VSMCs were exposed to control MEM orCORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours in the absence or presence of SPR (1024 mol/L). Data are expressed as the percentageof dpHi/dt compared with those observed in CORTI-untreated VSMCs without SPR (control). The number of experiments done on VSMCstreated with SPR for 3 and 24 hours is six and eight, respectively.

pletely prevented the long-term inhibitory effect of VSMCs [34]. The VSMCs were then exposed to eithercontrol MEM or CORTI-supplemented MEM for 3 orCORTI on NHE activity. The role of PKC was further

examined in PKC-depleted cells. For this experiment, 24 hours in the presence of PMA. The results are shownin Figure 7C. Pre-exposure of VSMCs to PMA aloneVSMCs were placed in serum-free MEM for 48 hours

and were then pre-exposed to PMA (a specific activator had no effect on NHE activity. On the other hand, itcompletely inhibited both short- and long-term effectsof PKC; 1027 mol/L) for 24 hours. This treatment has

been reported to down-regulate PKC activity in cultured of CORTI on NHE activity. We next examined whether

Muto et al: Regulation of vascular NHE by GC 2327

Fig. 7. Effects of protein kinase C (PKC) inhibitors or PKC down-regulation on CORTI-induced changes in NHE activity in VSMCs. pHi

measurements were carried out in the nominal absence of CO2/HCO32. (A) VSMCs were exposed to control MEM or CORTI (1026 mol/L)-

supplemented MEM for 3 or 24 hours in the absence or presence of staurosporine A (ST). Data are expressed as the percentage of dpHi/dtcompared with those observed in CORTI-untreated VSMCs without ST (control). The number of experiments done on ST-treated VSMCs for 3and 24 hours is eight. (B) VSMCs were exposed to control MEM or CORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours in the absenceor presence of calphostin C (CAL; 5 3 1027 mol/L). Data are expressed as the percentage of dpHi/dt compared with those observed in CORTI-untreated VSMCs without CAL (control). The number of experiments done on CAL-treated VSMCs for 3 and 24 hours is eight. (C) VSMCswere exposed to control MEM or CORTI (1026 mol/L)-supplemented MEM for 3 or 24 hours with or without 24-hour pre-exposure of phorbol12-myristate 13-acetate (PMA; 1027 mol/L). Data are expressed as the percentage of dpHi/dt compared with those observed in CORTI-untreatedVSMCs without PMA (control). The number of experiments done on PMA-treated VSMCs for 3 and 24 hours is five. (D) VSMCs were exposedto control MEM or PMA (1027 mol/L)-supplemented MEM for three hours. Data are expressed as the percentage of dpHi/dt compared with thoseobserved in PMA-untreated VSMCs (control). The number of experiments done on PMA-treated VSMCs is four.

exposure of VSMCs to PMA for three hours actually then exposed to MEM treated with either COL (1024

mol/L) or CYTO (1024 mol/L), plus CORTI (1026 mol/L)increases NHE activity. As shown in Figure 7D, treat-ment of VSMCs with PMA (1027 mol/L) for three hours for three hours. The results for COL-treated VSMCs are

shown in Fig. 8A and for CYTO-treated VSMCs in Fig.caused NHE activity to increase to 1.4 times the controllevel. 8B. COL or CYTO alone had no influence on NHE

activity. COL completely inhibited the short-term stimu-Role of cytoskeletal elements in mediating latory effect of CORTI on NHE activity, but CYTO hadCORTI-induced increase in NHE activity no effect.

It has been reported that the interactions betweenEffect of CORTI on NHE mRNA levels in VSMCsthe cytoskeleton and the plasma membrane regulate the

activity of many ion channels and transport proteins, These experiments suggest that transcription andtranslation are involved only in the long-term effect ofincluding NHE [6, 35, 36]. To determine whether the

cytoskeleton is involved in the short-term effect of CORTI on NHE activity. Therefore, we examined theshort- and long-term effects of CORTI on NHE (NHE-1)CORTI on NHE activity, VSMCs were incubated with

COL, a disruptor of microtubule, or CYTO, a disruptor mRNA levels in VSMCs. Representative Northern blotsare shown in Figure 9. The NHE-1 cDNA probe hybrid-of filamentous actin (F-actin). For this purpose, VSMCs

were placed in serum-free MEM for 48 hours and were ized to a single band of 4.8 to 5.0 kb, as previously

Muto et al: Regulation of vascular NHE by GC2328

Fig. 8. Effects of cytoskeleton disruptors, colchicine (COL), or cytochalasin B (CYTO) on the short-term CORTI-induced increase in NHEactivity in VSMCs. VSMCs were exposed to control MEM or CORTI (1026 mol/L)-supplemented MEM for three hours in the absence or presenceof COL (1024 mol/L) or CYTO (1024 mol/L). pHi measurements were carried out in the nominal absence of CO2/HCO3

2. Data are expressed asthe percentage of dpHi/dt compared with those observed in CORTI-untreated VSMCs without COL or CYTO (control). The number of experimentsdone on COL- and CYTO-treated VSMCs is five.

described for NHE-1 mRNA in cultured VSMCs [6]. respectively. Therefore, treatment with 0.25% trypsinUpon addition of CORTI (1026 mol/L) to VSMCs for significantly inhibited [3H]CORTI-specific binding (con-three hours, NHE-1 mRNA levels were not significantly trol cells, 3937.0 6 244.2 cpm, N 5 3; trypsin-treatedaltered. On the other hand, when VSMCs were exposed cells, 763.6 6 172.0 cpm, N 5 3, P , 0.005). Figure 10Bto CORTI for 24 hours, NHE-1 mRNA levels decreased shows the surface receptor binding of [3H]CORTI withto 0.65 times the control level. There was no significant unlabeled CORTI or RU to VSMCs. At 4 3 1029 mol/Ldifference in GAPDH mRNA levels between the control [3H]CORTI, approximately 71% of the total counts wereand CORTI-treated VSMCs. bound, with nonspecific binding (binding in the presence

of 1 3 1025 mol/L nonradioactive CORTI) representing[3H]CORTI surface receptor binding to VSMCs 29% of the total counts. The presence of unlabeled

Scatchard analysis of the binding data revealed a linear CORTI or RU displaced bound [3H]CORTI to VSMCsplot demonstrating a single class of CORTI binding sites in a similar and competitive manner. The curve in RU-with a Bmax of 876.2 fmol per mg of cell protein and a treated VSMCs was slightly shifted to right, as comparedKd of 12.2 nmol/L (Fig. 10A). To determine whether with the curve in CORTI-treated VSMCs, indicating that[3H]CORTI receptor binding was sensitive to trypsin, RU has a slightly lower affinity for [3H]CORTI surfacewhich removes surface receptors [37, 38], the effect of receptors than nonradioactive CORTI.exposure to 0.25% trypsin on [3H]CORTI binding toVSMCs was studied. In control VSMCs, at 4 3 1029

DISCUSSIONmol/L [3H]CORTI, the total and nonspecific (in the pres-The present study was designed to determine whetherence of 1 3 1025 mol/L nonradioactive CORTI) counts

CORTI modulates NHE activity in VSMCs and to deter-were 9868.6 6 375.3 (N 5 3) and 5931.6 6 209.1 cpmmine the mechanisms for CORTI-induced changes in(N 5 3), respectively. On the other hand, in trypsin-NHE activity. Our data demonstrate that CORTI in-treated cells, the total and nonspecific counts were

1101.6 6 230.4 (N 5 3) and 338.0 6 59.4 cpm (N 5 3), creases VSMCs NHE activity as early as three hours.

Muto et al: Regulation of vascular NHE by GC 2329

Fig. 9. Effects of CORTI on steady-state NHE (NHE-1) mRNA levelsin VSMCs. VSMCs were exposed to control MEM or CORTI (1026

mol/L)-supplemented MEM for 3 or 24 hours. Five micrograms ofpoly (A)1 RNA from control and CORTI-treated VSMCs were sizefractionated by 1% agarose-formaldehyde gel electrophoresis, trans-ferred to nylon membrane, and hybridized rat NHE-1 (upper panel)and GAPDH (lower panel) cDNA probes. Data are representative offive separate experiments that gave similar results.

However, this effect was not sustained. Twenty-four–hour exposure to CORTI showed a significant decrease

Fig. 10. (A) Scatchard plot of [3H]CORTI surface receptor binding toin NHE activity. Both short- and long-term effects ofVSMCs. The experimental protocols were described in detail in theCORTI on NHE activity are dose dependent. Short-Methods section. Each point represents mean of three separate experi-

term (3 hours) stimulatory effect of CORTI on NHE ments. (B) Influence of unlabeled CORTI (h) or RU (e) on the surfacebinding of [3H]CORTI to VSMCs. The experimental protocols wereactivity occurs independently of gene transcription anddescribed in detail in the Methods section. Each point representsprotein synthesis, occurs through the CORTI surfacemean 6 SEM (N 5 3).

receptor, is mediated by microtubule-dependent process,and is not accompanied by increased NHE-1 mRNAlevels. In sharp contrast, a long-term (24 hours) inhibi-tory effect of CORTI on NHE activity requires gene activity was not affected by inhibitors of transcriptiontranscription and protein synthesis, occurs only through (ACD) or translation (CHX). Therefore, the short-termthe cytosolic GR, and is associated with decreased NHE- effect is related to a nongenomic effect. This is further1 mRNA levels. Both short- and long-term effects of supported by the fact that short-term exposure toCORTI on NHE activity are PKC dependent. This study CORTI did not increase NHE-1 mRNA levels (Fig. 9).was the first to demonstrate dual modulation of NHE Recently, our laboratory reported that exposure ofactivity by GC in VSMCs. VSMCs to aldosterone for three hours also activates

NHE through a nongenomic mechanism [6].Short-term stimulation of NHE activity by CORTI Since the previously mentioned findings suggest that

the short-term CORTI effect is mediated through surfaceThe present study found that short-term exposure toCORTI increased basal levels of pHi and that this in- receptors, we performed binding studies of [3H]CORTI

to surface receptors in VSMCs. Our results demonstratecrease in pHi is caused by NHE activation. Furthermore,the short-term stimulatory effect of CORTI on NHE the presence of specific surface binding sites for CORTI

Muto et al: Regulation of vascular NHE by GC2330

in VSMCs. The Scatchard analysis indicates that completely inhibited the nongenomic stimulatory effectof CORTI on NHE activity (Fig. 6A). Accordingly, RU[3H]CORTI binds to one type of binding site in VSMCs

(Fig. 10A). The Kd and Bmax values for the CORTI bind- could bind to the CORTI surface receptor in the plasmamembrane of VSMCs to inhibit the stimulatory effecting sites found in this work are similar to those in

[3H]CORTI binding to rat kidney plasma membrane [15] of CORTI on NHE activity. Similarly, Liu and Chenreported that RU partially blocked the rapid inhibitoryand to rat liver plasma membrane [16], but are 10- and

8-fold greater than those in [3H]CORTI binding to bo- effect of CORTI on release of arginine vasopressin fromhypothalamic slices in rats [45]. It is suggested that therevine aortic cytosol, respectively [39]. Thus, we can de-

duce that the specific [3H]CORTI binding observed in is probably some homology between the membranebinding site of CORTI and cytosolic GR, since RUthe present study is not to the cytosolic receptor for

several reasons. First, in the present experiments, the blocked not only the classic genomic actions at the cyto-solic GR, but also nongenomic effects at the surfacebinding of [3H]CORTI was determined by use of intact

VSMCs at 48C. At this temperature, internalization/en- receptor of the plasma membrane.We demonstrate that two dissimilar inhibitors of PKCdocytosis is effectively arrested, and ligand binding is

held to represent surface binding sites for many physio- activity (ST and CAL) or prior down-regulation of PKCactivity by 24-hour pre-exposure to PMA prevented thelogically important macromolecules that enter cells by

receptor-mediated endocytosis [40, 41]. Reported bind- short-term stimulatory effect of CORTI on NHE activ-ity. We also found that short-term exposure of VSMCsing studies of vasoconstrictor peptide hormones (such

as arginine vasopressin, angiotensin II, and endothelin) to PMA mimicked the stimulatory effect of CORTI onNHE activity. Taken together, our results are consistentto surface receptors of the plasma membrane have been

performed at 48C [37, 38, 41, 42]. Second, trypsin treat- with the notion that the short-term stimulatory effectof CORTI on NHE activity is mediated through PKCment, which moves cell surface receptor [37, 38], de-

creased specific binding of [3H]CORTI to VSMCs by activation, although we have not yet determined whetherCORTI actually stimulates PKC activity in VSMCs. Sim-81%, indicating the protein nature of this binding site.

Andres et al also reported that specific CORTI binding ilar findings have been reported in VSMCs treated withaldosterone for three hours [6] and PMA-treated innerto calf adrenal cortex plasma membrane was reduced by

more than 70% when membranes were pretreated with medullary collecting duct cells from rat kidneys [46].There are several potential mechanisms to explain theproteases [17]. Third, the specific binding site determined

in VSMCs is different from the characterized binding of PKC-dependent NHE activation induced by CORTI.One possibility is that PKC could phosphorylate the reg-corticosteroid-binding globulin to rat cell membrane

[43]. This binding was a little specific at 48C, and the ulatory protein of NHE or the NHE itself directly [23,33, 47, 48] and subsequently activate NHE. In supportmaximal binding was obtained at 378C. The kinetic pa-

rameters were Kd of 0.84 mmol/L and a Bmax of 39 of this concept, Sardet et al first described in hamsterfibroblasts that the NHE is rapidly phosphorylated inpmol/mg protein. The values of Kd and Bmax are 70- and

50-fold greater than those obtained for the CORTI bind- response to various mitogens, and that this phosphoryla-tion of the NHE is temporally correlated with its activa-ing sites in this work, respectively. Thus, the specific

[3H]CORTI binding to VSMCs is not due to a corticoste- tion [47]. That the magnitude of phosphorylation couldregulate the rate of NHE is also suggested by the findingroid-binding globulin. Taken together, our data suggest

that CORTI binds to the surface receptor of the plasma that vanadate, an inhibitor of phosphatases, activatesNHE in A431 cells [48]. Additional experiments supportmembrane in VSMCs to activate NHE and that trypsin

is a potent inhibitor of [3H]CORTI binding to plasma the hypothesis that PKC is one of the kinases responsiblefor this phosphorylation in VSMCs [23, 49]. Grinsteinmembrane of VSMCs. The existence of a plasma mem-

brane receptor for steroid hormones has been exten- et al have shown in rat thymic lymphocytes that directPKC activation by PMA increases NHE activity by caus-sively reported in many tissues [15–20]. Further studies

are required to characterize the CORTI receptor in the ing an alkaline shift in the pHi dependence of the antiport[50, 51]. Thus, it is also possible that the NHE activationplasma membrane of VSMCs.

RU has anti-GC activity and high binding affinity for could occur via a PKC-dependent alkaline shift in thepHi responsiveness of an allosteric modifier site on thea cytosolic GR [44]. In bovine aortic cytosol, [3H]CORTI

bound mainly to the RU-blockable component [39]. On cytoplasmic surface of the cell membrane [51, 52].As shown in Figure 8, COL, but not CYTO, inhibitsthe other hand, RU has a slightly lower affinity for

[3H]CORTI binding to the surface receptor of VSMCs the short-term stimulatory effect of CORTI on NHEactivity in VSMCs. These findings raise the possibilitythan unlabeled CORTI (Fig. 10B). Thus, the inhibitory

characteristic of RU on [3H]CORTI binding to the sur- that microtubule, but not F-actin, may regulate the exo-cytic insertion of NHE protein into the plasma mem-face receptor of VSMCs in this study is strikingly distinct

from that of the bovine aortic cytosol. Moreover, RU brane in CORTI-treated VSMCs. Similarly, Bloch et al

Muto et al: Regulation of vascular NHE by GC 2331

reported that angiotensin II stimulates NHE activity in cates that the long-term inhibitory effect of CORTI onNHE activity occurs only through the cytosolic GR.the rabbit proximal tubule and that this stimulation is

inhibited by COL [35]. Evidence suggests that microtu- Taken together, the long-term inhibitory effect ofCORTI on NHE activity occurs at transcriptional levels.bules also participate in the regulation of other mem-

brane transport processes, including renal H1 pump [53] The present study also demonstrates that the long-term inhibitory effect of CORTI on NHE activity wasand water channels [54]. Schwartz and Al-Awqati re-

ported that treatment with COL inhibits exocytic inser- prevented by two different PKC inhibitors (ST andCAL) and by PKC down-regulation with continuous ex-tions of H1 pump into apical membrane of renal proxi-

mal tubular cell [53]. In addition, COL has been shown posure to PMA. These findings suggest that the long-term CORTI effect, in addition to the short-term CORTIto inhibit the vasopressin-induced insertion of water

channels into apical membrane, consistent with a role for effect, occurs through PKC activation. Similarly, Vrtovs-nik et al reported in opossum kidney cells that GC inhib-microtubules in the delivery of the vesicles that contain

water channels to the apical membrane [54]. On the its Na1/phosphate cotransport activity through PKC acti-vation [56]. It is unlikely that PKC down-regulation mayother hand, in VSMCs, the short-term (3 h) stimulatory

effect of aldosterone on NHE activity is inhibited by lead to decreased NHE activity since two PKC inhibi-tors or PKC down-regulation with continuous exposureboth COL and CYTO [6]. Since COL did not affect basal

NHE activity in control VSMCs (Fig. 8), the data indicate to PMA did not mimic the long-term inhibitory effectof CORTI on NHE activity (Fig. 7 A–C). The intracellu-that the short-term exposure to CORTI activates a COL-

sensitive process, which activates NHE. However, the lar signaling mechanism responsible for the CORTI-induced decrease in NHE-1 mRNA abundance has notmechanisms by which CORTI activates NHE via COL-

sensitive process are still unclear. It is possible that NHE been defined in the present study. Nevertheless, the factthat the decreased NHE activity is dependent on bothprotein may be directly shuttled to the plasma membrane

via microtubules to stimulate NHE activity in CORTI- CORTI-induced PKC activation and gene transcriptionsuggests that CORTI-induced PKC activation may signaltreated VSMCs. We also propose that an activator of

NHE might be shuttled to the plasma membrane to stim- the decrease in NHE-1 mRNA levels. These results arein line with the reported effect of 1, 25-dihydroxyvitaminulate silent carriers. For example, PKC has been shown

to rapidly move from a cytoplasmic position to the D3, another steroid hormone on transcriptional regula-tion of PKC in HL-60 cells [57] and the direct transcrip-plasma membrane following stimulation of cells with

phorbol ester [55]. This possibility is also supported by tional induction by dexamethasone of a serine/threonineprotein kinase gene in cultured cells [58]. However, thethe fact that the short-term stimulatory effect of CORTI

on NHE activity occurs through PKC activation (Fig. 7). mechanisms whereby CORTI may repress transcriptionof NHE-1 gene via transcriptional activation of PKC are

Long-term inhibition of NHE activity by CORTI not known.In sharp contrast to the short-term exposure to

Dual modulation of NHE activity by CORTICORTI, the long-term exposure to CORTI decreasedin VSMCsbasal pHi values, and the CORTI-mediated pHi decrease

is caused by inhibition of NHE activity. Recently, our There are several reports on the time course of NHEactivity [8, 59] and cell growth in GC-treated VSMCslaboratory reported that exposure of VSMCs to aldoste-

rone for 24 hours increased basal values of pHi and that [7–9]. Christ et al have shown that an addition of VSMCsto hydrocortisone at 10–6 mol/L for four minutes had nothis increase in pHi is a result of increased NHE activity

[6]. Thus, in VSMCs, the long-term effect of CORTI on effect on NHE activity [59]. Berk et al observed thatthere was no significant difference in basal- or agonist-NHE activity is opposite to that of aldosterone. Further-

more, the inhibitory effect of CORTI on NHE activity stimulated NHE activity after one or four hours of hydro-cortisone (1026 mol/L) treatment, but a significant in-was prevented by inhibitors of transcription (ACD) and

of translation (CHX), and was associated with decreased crease in NHE activity occurred 12 hours after treatmentwith the hormone, with a maximum increase at 24 hoursNHE-1 mRNA levels. Accordingly, the inhibitory effect

of CORTI is related to a genomic effect. In the present [8]. They also observed that exposure of VSMCs to hy-drocortisone for 24 hours induces a phenotypic changestudy, long-term exposure to CORTI decreased NHE

activity to 0.6 to 0.7 times the control level, in concor- of VSMCs and inhibits proliferation of cultured rat aorticVSMCs, if proliferation is induced by 10% fetal calfdance with a decrease in NHE-1 mRNA abundance (0.65

times the control level) by magnitude. These findings serum. Similar growth-inhibiting effects have been re-ported in bovine aortic smooth muscle cells treated withare consistent with the notion that CORTI inhibits NHE

activity by decreasing NHE-1 mRNA levels. In addition, dexamethasone (1026 mol/L) for five days [7]. However,the hydrocortisone-induced inhibition of VSMC growthour observation that the long-term effect of CORTI was

inhibited only by the cytosolic GR antagonist (RU) indi- cannot be explained by increased NHE activity, since

Muto et al: Regulation of vascular NHE by GC2332

growth factors activate NHE to stimulate growth in REFERENCESVSMCs [1, 2, 22, 23, 49]. Our present findings that short- 1. Grinstein S, Rothstein A: Mechanisms of regulation of the

Na1/H1 exchanger. J Membr Biol 90:1–12, 1986term exposure to CORTI stimulates NHE activity but2. Tse M, Levine S, Yun C, Brant S, Coubillon LT, Pouyssegurlong-term exposure to CORTI inhibits NHE activity are

J, Donowitz M: Structure/function studies of the epithelial iso-in agreement with reports of Kawai et al, who have forms of the mammalian Na1/H1 exchanger gene family. J Membr

Biol 135:93–108, 1993shown in A7r5 rat aortic smooth muscle cell line that3. Frelin C, Vigne P, Lazdunski M: The Na1/H1 exchange systemgrowth of VSMCs is stimulated by brief exposure to in vascular smooth muscle cells. Adv Nephrol 19:17–29, 1990

dexamethasone, but is inhibited by long-term exposure 4. Wessberg PL, Little PJ, Cragoe EJ, Bobik A: Na-H antiportin cultured rat aortic smooth muscle: Its role in cytoplasmic pHto the hormone [9]. The dual effects of GC accordingregulation. Am J Physiol 253:C193–C198, 1987to duration of exposure may be responsible for the com- 5. Little PJ, Cragoes ED, Bobik A: Na-H exchange is a majorpathway for Na influx in rat vascular smooth muscle. Am J Physiolplexity of pro-atherogenic and anti-atherogenic actions251:C707–C712, 1986of GC in vitro and in vivo. GC has been considered

6. Ebata S, Muto S, Okada K, Nemoto J, Amemiya M, Saito T,potentially anti-atherogenic, based on its inhibitory ef- Asano Y: Aldosterone activates Na1/H1 exchange in vascular

smooth muscle cells by nongenomic and genomic mechanisms.fect on VSMC proliferation in vitro [7]. However, in theKidney Int 56:1400–1412, 1999clinical setting, GC is believed to promote atherogenesis

7. Longenecker JP, Kilty LA, Johnson LK: Glucocorticoid influ-[60]. In patients with Cushing’s syndrome and experi- ence on growth of vascular wall cells in culture. J Cell Physiol

113:197–202, 1982mental models with GC-induced hypertension, pressor8. Berk BC, Vallega G, Griendling KK, Gordon JB, Cragoe EJ,responses to angiotensin II and norepinephrine are en- Canessa M, Alexander RW: Effects of glucocorticoids on

hanced [reviewed in 61]. However, the mechanisms for Na1/H1 exchange and growth in cultured vascular smooth musclecells. J Cell Physiol 137:391–401, 1988the GC-induced hypertension still remain unknown. The

9. Kawai Y, Hayashi T, Eguchi K, Asazuma K, Masamura K, Iwa-dual modulation of NHE activity by GC may be involved muro A, Takano Y, Tada H, Matsukawa S, Miyamori I: Effects

of brief glucocorticoid exposure on growth of vascular smoothin the mechanisms of GC-induced hypertension. Futuremuscle cells in culture. Biochem Biophys Res Commun 245:493–studies are necessary to determine the role of NHE in496, 1998

GC-induced hypertension and atherogenesis. 10. Muto S, Nemoto J, Ohtaka A, Watanabe Y, Yamaki M, Kawa-kami K, Nagano K, Asano Y: Differential regulation of Na1-K1-In conclusion, we clearly demonstrate in VSMCs thatATPase gene expression by corticosteroids in vascular smoothshort-term exposure to CORTI stimulates NHE activity muscle cells. Am J Physiol 270:C731–C739, 1996

through a nongenomic mechanism, but long-term expo- 11. Muto S, Nemoto J, Ebata S, Kawakami K, Asano Y: Corticoste-rone and 11-dehydrocorticosterone stimulate Na,K-ATPase genesure to CORTI inhibits NHE activity through a genomicexpression in vascular smooth muscle cells. Kidney Int 54:492–508,mechanism. 1998

12. Funder JW: Regulation of transepithelial Na transport by steroidand peptide hormones. Semin Nephrol 12:6–11, 1992ACKNOWLEDGMENTS

13. Meato M: Gene regulation by steroid hormones. Cell 56:335–344,1989This work was supported by a grant from the Salt Science Founda-

14. McEwen BS: Non-genomic and genomic effects of steroids ontion, the Japanese Kidney Foundation (Jinkenkyukai), the Scienceneural activity. Trends Pharmacol Sci 12:141–147, 1991Research Promotion Fund of the Japan Private School Promotion

15. Ibarrola I, Ogiza K, Marino A, Macarulla JM, Trueba M:Foundation, and Grants-in-Aid for Scientific Research from the Minis-Steroid hormone specifically binds to rat kidney plasma membrane.try of Education, Science, and Culture in Japan. A portion of thisJ Bioenerg Biomembr 23:919–926, 1991work was presented at the Annual Meeting of the American Society

16. Trueba M, Ibarrola I, Ogiza K, Marino A, Macarulla JM:of Nephrology in San Antonio, TX, USA, in 1997, and was publishedSpecific binding sites for corticosterone in isolated plasma mem-in abstract form (J Am Soc Nephrol 8:401A, 1997).brane from rat liver. J Membr Biol 120:115–124, 1991

17. Andres M, Marino A, Macarulla JM, Trueba M: Characteriza-Reprint requests to Shigeaki Muto, M.D., Department of Nephrology,tion of specific corticosterone binding sites in adrenal cortex plasmaJichi Medical School, Minamikawachi, Tochigi 329-0498, Japan.membrane and their localization by autoradiographic studies. CellE-mail: smuto@jichi.ac.jpMol Life Sci 53:673–680, 1997

18. Orchinik M, Murray TF, Moore FL: A corticosteroid receptorin neuronal membranes. Science 252:1848–1851, 1991

APPENDIX 19. Blondeau JP, Baulieu EE: Progesterone-receptor characterizedby photoaffinity-labeling in the plasma membrane of Xenopus

Abbreviations used in this article are: ACD, actinomycin D; laevis oocytes. Biochem J 219:785–792, 1984BCECF-AM, 2979-bis(carboxyethyl)-5(6)-carboxyfluorescein acetoxy- 20. Wehling M, Christ MM, Theisen K: Membrane receptors formethyl ester; BSA, bovine serum albumin; CAL, calphostin C; CHX, aldosterone: A novel pathway for mineralocorticoid action. Am Jcycloheximide; COL, colchicine; CORTI, corticosterone; CYTO, cyto- Physiol 263:E974–E979, 1992chalasin B; dpHi/dt, Na1-dependent pHi increase; EIPA, ethylisopropyl- 21. Bondy PK: The adrenal cortex, in Duncan’s Disease of Metabolismamiloride; FBS, fetal bovine serum; GAPDH, glyceraldehyde-3-phos- (7th ed), edited by Bondy PK, Rosenberg LE, Philadelphia, W.B.phate dehydrogenase; GC, glucocorticoid; GR, glucocorticoid receptor; Saunders, 1974, p 1115Kd, dissociation constant; MEM, Eagle’s minimum essential medium; 22. Berk BC, Vallega G, Muslin AJ, Gordon HM, Canessa M,MR, mineralocorticoid receptor; NHE, Na1/H1 exchange; NMDG, N- Alexamder RW: Spontaneously hypertensive rat vascular smoothmethyl-d-glucamine; pHi, intracellular pH; PKC, protein kinase C; muscle cells in culture exhibit increased growth and Na1/H1 ex-PMA, phorbol 12-myristate 13-acetate; RU, RU38486; SDS, sodium change. J Clin Invest 83:822–829, 1989dodecyl sulfate; SPR, spironolactone; SSC, standard saline citrate; ST, 23. Berk BC, Aronow MS, Brock TA, Cragoe E, Gimbrone MA,

Alexander RW: Angiotensin II-stimulated Na1/H1 exchange instaurosporine A; VSMC, vascular smooth muscle cell.

Muto et al: Regulation of vascular NHE by GC 2333

cultured vascular smooth muscle cells: Evidence for protein kinase 43. Singer CJ, Khan MS, Rosner W: Characteristics of the bindingof corticosteroid-binding globulin to rat cell membranes. Endocri-C-dependent and -independent pathway that is Ca21-dependent

and pertussis toxin-sensitive. J Biol Chem 262:5073–5078, 1987 nology 122:89–96, 198844. Moguilewsky M, Philibert D: Potent antiglucocorticoid activity24. Thomas JA, Buchsbaum RN, Zimniak A, Racker E: Intracellular

pH measurements in Ehrlich ascites tumor cells utilizing spectro- correlated with strong binding to the cytosolic glucocorticoid recep-tor followed by an impaired activation. J Steroid Biochem 20:271–scopic probes generated in situ. Biochem 18:2210–2218, 1979

25. Grinstein S, Cohen S, Rothstein A: Cytoplasmic pH regulation 276, 198445. Liu X, Chen Y-Z: Membrane-mediated inhibition of corticoste-in thymic lymphocytes by an amiloride-sensitive Na1/H1 antiport.

J Gen Physiol 83:341–369, 1984 rone on the release of arginine vasopressin from hypothalamicslices. Brain Res 704:19–22, 199526. Cano A, Preisig P, Alpern RJ: Cyclic adenosine monophosphate

acutely inhibits and chronically stimulates Na/H antiporter in OKP 46. Slotki IN, Schwartz JH, Alexander EA: Na1-H1 exchange isstimulated by protein kinase C activation in inner medullary col-cells. J Clin Invest 92:1632–1638, 1993

27. Roos A, Boron WF: Intracellular pH. Physiol Rev 61:296–434, lecting duct cells. Am J Physiol 259:F666–F671, 199047. Sardet C, Counillon L, Franchi A, Pouyssegur J: Growth factors1981

28. Orlowski J, Kandasamy RA, Shull GE: Molecular cloning of induce phosphorylation of the Na1/H1 antiporter, a glycoproteinof 110 kD. Science 247:723–726, 1990putative members of the Na/H exchanger gene family. cDNA clon-

ing, deduced amino acid sequence, and mRNA tissue expression 48. Sardet C, Fafournoux P, Pouyssegur J: a-thrombin, epidermalgrowth factor, and okadaic acid activate the Na1/H1 exchanger,of the rat Na/H exchanger NHE-1 and two structurally related

proteins. J Biol Chem 267:9331–9339, 1992 NHE-1, by phosphorylating a set of common sites. J Biol Chem266:19166–19171, 199129. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ: Protein

measurement with Folin phenol reagent. J Biol Chem 193:265–275, 49. Berk BC, Elder E, Mitsuka M: Hypertrophy and hyperplasiacause differential effects on vascular smooth muscle cell Na1/H11951

30. Scatchard G: The attraction of proteins for small molecules and exchange and intracellular pH. J Biol Chem 265:19632–19637, 199050. Grinstein S, Cohen S, Goetz JD, Rothstein A, Gelfand EW:ions. Ann NY Acad Sci 51:660–672, 1949

31. Martin RS, Jones WJ, Hayslett JP: Animal model to study of Characterization of the activation of Na1/H1 exchange in lympho-cytes by phorbol esters: Change in cytoplasmic pH dependence ofthe effects of adrenal hormones on epithelial function. Kidney Int

24:386–391, 1983 the antiport. Proc Natl Acad Sci USA 82:1429–1433, 198551. Grinstein S, Goetz JD, Rothstein A: 22Na1 fluxes in thymic32. Nishizuka Y: Studies and perspectives of protein kinase C. Science

233:305–312, 1986 lymphocytes. II. amiloride-sensitive Na1/H1 exchange pathway;reversibility of transport and asymmetry of the modifier site. J Gen33. Hatori N, Fine BP, Nakamura A, Cragoe E, Aviv A: Angiotensin

II effect on cytosolic pH in cultured rat vascular smooth muscle Physiol 84:585–600, 198452. Aronson PS, Nee J, Suhm MA: Modifier role of internal H1 incells. J Biol Chem 262:5057–5064, 1987

34. Williams B, Schrier RW: Characterization of glucose-induced in activating the Na1-H1 exchanger in renal microvillus membranevesicles. Nature 299:161–163, 1982situ protein kinase C activity in cultured vascular smooth muscle

cells. Diabetes 41:1464–1472, 1992 53. Schwartz GJ, Al-Awqati Q: Carbon dioxide causes exocytosisof vesicles containing H1 pumps in isolated perfused proximal and35. Bloch RD, Zikos D, Fisher KA, Schleicher L, Oyama M, Cheng

JC, Skopicki HA, Sukowski EJ, Cragoe EJ, Peterson DR: Activa- collecting tubules. J Clin Invest 75:1638–1644, 198554. Pearl M, Taylor A: Role of the cytoskeleton in the control oftion of proximal tubular Na1-H1 exchange by angiotensin II. Am

J Physiol 263:F135–F143, 1992 transcellular water flow by vasopressin in amphibian urinary blad-der. Bio Cell 55:163–172, 198536. Watson AJM, Levine S, Donowitz M, Montrose MH: Serum

regulates Na1/H1 exchange in Caco-2 cells by a mechanism which 55. Wolf M, LeVine H, May WS, Cuatrecasas P, Sahyoun N: Amodel for intracellular translocation of protein kinase C involvingis dependent on F-actin. J Biol Chem 267:956–962, 1992

37. Fishman JB, Dickey BF, Bucher NLR, Fine RE: Internalization, synergism between Ca and phorbol esters. Nature 317:546–548,1985recycling, and redistribution of vasopressin receptors in rat hepato-

cytes. J Biol Chem 260:12641–12646, 1985 56. Vrtovsnik F, Jourdain M, Cherqui G, Lefebvre J, FriedlanderG: Glucocorticoid inhibition of Na-Pi cotransport in renal epithelial38. Resink T, Scott-Burden T, Boulanger C, Weber E, Buhler FR:

Internalization of endothelin by cultured vascular smooth muscle cells is mediated by protein kinase C. J Biol Chem 269:8872–8887,1994cells: Characterization and physiological significance. Mol Pharma-

col 38:244–252, 1990 57. Obeid LM, Okazaki T, Karolak LA, Hannun YA: Transcrip-tional regulation of protein kinase C by 1,25-dihydroxyvitamine39. Hayashi T, Kornel L: Cortisol-21-sulfate (FS) is a specific ligand

for intracellular transcortin: Demonstration of three types of high D3 in HL-60 cells. J Biol Chem 265:2370–2374, 199058. Webster MK, Goya L, Ge Y, Maiyar AC, Firestone GL: Charac-affinity corticosteroid binders in bovine aortic cytosol by a com-

bined use of FS and RU 28362. Endocrinology 126:307–316, 1990 terization of sgk, a novel member of the serine/threonine proteinkinase gene family which is transcriptionally induced by glucocorti-40. Limbird LE: Receptor binding kinetics, in Cell Surface Receptors:

A Short Course on Theory and Methods, edited by Limberd LE, coids and serum. Mol Cell Biol 13:2031–2040, 199359. Christ M, Diuwes K, Eisen C, Bechtner G, Theisen K, WehlingBoston, Martinus Nijhoff, 1986, pp 159–194

41. Wileman T, Harding C, Stahl P: Receptor-mediated endocytosis. M: Rapid effects of aldosterone on sodium transport in vascularsmooth muscle cells. Hypertension 25:117–123, 1995Biochem J 232:1–14, 1985

42. Ullian ME, Linas SL: Role of receptor cycling in the regulation 60. Nashel DJ: Is a atherosclerosis a complication of long-term corti-costeroid treatment? Am J Med 80:925–929, 1986of angiotensin II surface receptor number and angiotensin II up-

take in rat vascular smooth muscle cells. J Clin Invest 84:840–846, 61. Walker BR, Williams BC: Corticosteroids and vascular tone:mapping the messenger maze. Clin Sci 82:597–605, 19921989