• No results found

Mechanism of calcium transport stimulated by chlorothiazide in mouse distal convoluted tubule cells

N/A
N/A
Protected

Academic year: 2020

Share "Mechanism of calcium transport stimulated by chlorothiazide in mouse distal convoluted tubule cells"

Copied!
11
0
0

Loading.... (view fulltext now)

Full text

(1)

Mechanism of calcium transport stimulated by

chlorothiazide in mouse distal convoluted

tubule cells.

F A Gesek, P A Friedman

J Clin Invest. 1992;90(2):429-438. https://doi.org/10.1172/JCI115878.

Thiazide diuretics inhibit Na+ and stimulate Ca2+ absorption in renal distal convoluted tubules. Experiments were performed on immortalized mouse distal convoluted tubule (MDCT) cells to determine the mechanism underlying the dissociation of sodium from calcium transport and the stimulation of calcium absorption induced by thiazide diuretics. Control rates of 22Na+ uptake averaged 272 +/- 35 nmol min-1 mg protein-1 and were inhibited 40% by chlorothiazide (CTZ, 10(-4) M). Control rates of 36Cl- uptake averaged 340 +/- 50 nmol min-1 mg protein-1 and were inhibited 50% by CTZ. CTZ stimulated 45Ca2+ uptake by 45% from resting levels of 2.86 +/- 0.26 nmol min-1 mg protein-1. Bumetanide (10(-4) M) had no effect on 22Na+, 36Cl-, or 45Ca2+ uptake. Control levels of intracellular calcium activity ([Ca2+]i) averaged 91 +/- 12 nM. CTZ elicited concentration-dependent increases of [Ca2+]i to a maximum of 654 +/- 31 nM at 10(-4) M. CTZ reduced intracellular chloride activity ([Cl-]i), as determined with the chloride-sensitive fluorescent dye

6-methoxy-N-(3-sulfopropyl)quinolinium. The chloride channel blocker

5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB, 10(-5) M) abolished the effect of CTZ on [Cl-]i. NPPB also blocked CTZ-induced increases of 45Ca2+. Resting membrane voltage, measured in cells loaded with the potential-sensitive dye 3,3'-dihexyloxacarbocyanine iodide [DiOC6(3)], averaged -72 +/- 2 mV. CTZ hyperpolarized cells in a concentration-dependent and reversible manner. At 10(-4) M, CTZ hyperpolarized MDCT cells by 20.4 +/-[…]

Research Article

Find the latest version:

(2)

Mechanism of Calcium

Transport Stimulated

by Chlorothiazide

in Mouse Distal Convoluted

Tubule Cells

FrankA.Gesekand PeterA. Friedman

Department ofPharmacologyand Toxicology,Dartmouth Medical School,Hanover,NewHampshire03756

Abstract

Thiazidediuretics inhibitNa+ andstimulateCa2+absorptionin renal distal convoluted tubules. Experiments wereperformed

onimmortalizedmousedistal convolutedtubule(MDCIT)cells todeterminethemechanism underlyingthedissociationof

so-dium from calcium transport and the stimulation ofcalcium absorption inducedbythiazide diuretics. Controlratesof 22Na+

uptake averaged 272±35 nmol

min-'

mg protein-' and were

inhibited 40%bychlorothiazide(CTZ,10'-M). Control rates

of

MCl-

uptake averaged340±50nmolmin'mgprotein-'and wereinhibited50% by CTZ. CTZstimulated

4"Ca2"

uptake by 45%from restinglevelsof 2.86±0.26nmolmin'mgprotein'.

Bumetanide (10-4 M) had noeffecton22Na+, -6CI-, or45Ca2+ uptake. Controllevels ofintracellularcalcium activity

([Ca2+1I)

averaged 91±12nM.CTZ elicited concentration-dependent in-creases of

ICa2+J1

toa maximumof 654±31 nM at 10-4M.

CTZreducedintracellularchloride activity

([ClI-I),

as

deter-mined with thechloride-sensitive fluorescentdye

6-methoxy-N-(3-sulfopropyl)quinolinium. Thechloride channel blocker

5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB, 10-5

M) abolishedtheeffect ofCTZ on

[Cl-Ij.

NPPB alsoblocked

CTZ-inducedincreases of 45Ca2+.Restingmembrane voltage,

measured in cellsloaded withthepotential-sensitive dye 3,3'-dihexyloxacarbocyanine iodide[DiOC6(3)J, averaged -72±2

mV. CTZ hyperpolarized cells in a concentration-dependent

and reversible manner. At 1o-4 M, CTZ hyperpolarized MDCT cells by 20.4±7.2mV.Reduction of extracellular

Cl-or addition of NPPB abolished CTZ-induced hyperpolariza-tion.Direct membranehyperpolarization increased45Ca2+

up-take whereasdepolarization inhibited45Ca2+uptake.

CTZ-sti-mulated

45Ca2+

uptake was inhibited by the Ca2+ channel blockernifedipine (10-5 M). Weconclude that thiazide di-uretics block cellular chloride entrymediated byapical mem-brane NaCl cotransport. Intracellular chloride, which under control conditions isaboveitsequilibriumvalue, exits the cell through NPPB-sensitive chloride channels. This decrease of intracellular chloride hyperpolarizes MDCTcells and

stimu-lates

Ca2+

entry byapical membrane,

dihydropyridine-sensi-tive

Ca2+

channels. (J. Clin. Invest. 1992.90:429-438.) Key

Portions ofthis workwerepresentedatthe annualmeetingofthe Amer-icanSociety of Nephrology, 17-20 November1991, Baltimore,MD.

Addresscorrespondence to Peter A. Friedman, Ph. D., Dept.of Pharmacology &Toxicology,Dartmouth MedicalSchool,7650

Rem-sen,Hanover,NH03756.

Receivedfor publication 1I September1991andinrevisedform21

February1992.

words: distal tubule * calcium transport * thiazide diuretics-intracellular calcium*NaClcotransport

Introduction

Despitethewidespread therapeuticuse ofthiazide diureticsin

themanagement of hypertension, congestiveheart failure, and

idiopathic hypercalciuria,thecellular mechanisms underlying their actiononboth sodiumandcalcium transport remain

un-clear. The unique ability of these agentsto reduce renal

cal-ciumexcretion,especiallyuponchronic administration, while simultaneously promoting that of sodium is along-standing observation (1). Renal clearance (2), micropuncture (3-6),

and microperfusion (7, 8) experiments identified the distal convoluted tubule as the nephron site atwhich thiazide

di-ureticsinhibit sodiumtransport.Distal convoluted tubulesare

also the siteatwhich thiazidediureticsstimulate calcium

ab-sorption(6). Uncertainty regarding the mechanism by which

thiazidediureticsexerttheseopposingeffectsoncation trans-port maybeattributabletotheincompletecharacterizationof

sodiumandcalciumabsorption inthis nephronsegment.

Thiazide diuretics decrease distal nephron sodium and

chlo-rideabsorption (6, 8) but enhancenetcalcium absorption (6). These effects require the presence of the drug in the tubular lumen (9, 1

0).

Themechanismby which calciumtransportis stimulated by thiazide diuretics is notknown. Itisgenerally accepted that thiazidediureticsdonot altertransepithelial

volt-age(6). However, recentstudies show thatthiazide diuretics

hyperpolarize plasma membrane voltage with little or no

change oftransepithelialvoltage in homologous distal nephron

structures of the amphibian nephron (11, 12), the flounder bladder( 13),and inpilot studies,the rabbit distal convoluted tubule( 14). Further,

hydrochlorothiazide

increases Ca2+

up-take across luminal membranes of distal convoluted tubule cells ( 15). We recently demonstrated that parathyroid hor-monestimulatescalciumentryinprimarycell cultures

ofcorti-cal thickascendinglimb anddistalconvoluted tubule cellsby activating

dihydropyridine-sensitive

calcium channels (16). Activationof such

dihydropyridine-sensitive

calcium channels inapical plasma membranes is consistent with the

physiologi-cal effect ofparathyroidhormonetoincrease therate

oftransep-ithelial calcium transport( 17, 18). Takentogether,these ob-servationssuggestedaplausible mechanism bywhichthiazide

diuretics might stimulate calcium transport.

Specifically,

we

theorizedthatinhibition of chlorideentrybythiazidediuretics

hyperpolarizes

distal convoluted tubule cells, which in turn

increases calciumentry. Theexperiments reportedin the pres-entworkwereundertakentodeterminethemechanism ofthia-zide-induced

hyperpolarization

andto testthis

hypothesis

us-ing immortalized distal convoluted tubule cells. The results

suggest thatthiazide diuretics inhibit NaCl cotransport, result-ingindiminishedcellular chlorideentry,decreased

intracellu-larchloride, membrane

hyperpolarization,

and increased

cal-Thiazide-stimulatedCa2" Transport byDistal ConvolutedTubule Cells 429

J.Clin.Invest.

©The AmericanSocietyforClinicalInvestigation,Inc.

0021-9738/92/08/0429/10 $2.00

(3)

ciumentry.The increaseof calciumentryismediatedby

dihy-dropyridine-sensitive calciumchannels.

Methods

Preparation and cell culture of immortalized distal convoluted tubule

cells. Cortical thickascending limbanddistalconvoluted tubule cells were isolated frommice (CDI; Charles RiverBreedingLaboratories

Inc.,Wilmington,MA) withadouble-antibodyprocedure previously described( 19) andestablishedinprimarycellculture. Antibioticswere removed24 hbeforetransformation.Cellswereinfected for24 h with

chimeric adenovirus 12-simian virus40(AD12/SV40; Dr. J.Rhim,

National Cancer Institute).Theuseofthis hybridconstructwas em-ployedpreviouslytoestablishacollectingtubule cellline (20).After mockinfectantshadexpired(10d), dishes oftheinfectedcells were passaged.PreliminaryscreeningwasperformedtomonitorcAMP for-mation in responsetoPTH. Cellsexhibiting PTH-inducedcAMP ac-cumulation 2 5Xbackgroundwereclonedbylimitingcelldilutionand grown toconfluence. Determinationofdistal convolutedtubule

pheno-type wasbasedupon thefollowingcriteria:(a)inhibitionof22Na'and 36C-uptakebychlorothiazide(l0-4 M);

(b) absenceof effect of bu-metanide (10-4M) on22Na+or36CI uptake;and (c)stimulationof

45Ca2+uptakebychlorothiazide.Cellsexhibitingthesecharacteristics

wereestablishedin culture and aredesignatedmouse distalconvoluted

tubule(MDCT)1 cells.Cellswere grown on 60-mm dishes(Coming

GlassWorks, Coming MedicalandScientific, Coming,NY) inDME/

Ham's F12media (DME Nutrient MixtureF-12 Ham; Sigma

Chemi-cal Co., St.Louis, MO) supplementedwith 5%heat-inactivated(560C for 20min)FCS(Sigma Chemical Co.)and PSNantibiotic mixture (50,ggpenicillin,50Mg streptomycin, 100Mgneomycin/100ml media; GibcoLaboratories, Gaithersburg,MD)in ahumidified atmosphereof 95%02/5% CO2at370C.Cellpassages5-21 were used in the present study.

IsotopicNa',Cl-, andCa2` uptakemeasurements.A rapid

filtra-tiontechniquereported inpreviousinvestigations(21)wasmodifiedto

measure uptake of22Na', 'CI-, and45Ca2+by cultured cells. Cells (5-6 X 106cells/60-mm dish)were removedfrom dishes with 0.125%

trypsin,<5 min(Sigma Chemical Co.)andrinsedwithmodifiedKRB

containing (mM): 140NaCl,5KCl, 1 CaC12, 1

MgCl2,

bufferedwith 18 Tris base and 10 Hepes to pH7.40±0.01 and295±1 mosmol/kg H20.Todeterminetheeffectof membranevoltageon45Ca2+uptake, voltagewas altered byvarying extracellularK+(0-100 mM), which wasreplacedisosmoticallywith Na+ andclampedby use of the

iono-phore valinomycin(10-6M). Thechloride dependenceof45Ca2+ up-take wasevaluatedbyisosmoticallyreplacingCl- withgluconate.For theseexperiments,calcium activityof thegluconate-containing solu-tions was measured withacalcium-specificelectrode (model 932000;

Orion Research Inc., Cambridge, MA)and adjusted byaddition of

CaCl2tobeidenticaltothatofchloride-containingsolutionsat pH 7.4.

Buffer,withorwithout further additions,(totalvolume= 100 M1)

was placed in 12X75 mmpolystyrenetubes and a

50-Ml

aliquotofcells was addedfor1 minbefore additionandvortexingof a

50-Mu

aliquotof

22Na'(10,Ci/ml, carrier-free; Amersham Corp., Arlington Heights, IL), 36CP-(58 MCi/ml; ICN Radiochemicals, Costa Mesa, CA) or

45Ca2+(1 MCi/ml,carrier-free;NewEngland Nuclear,Boston, MA) to

initiate isotopeuptake. All sample tubes weremaintainedat

370C

in a

shakingincubator. Controlexperiments indicatedthat traceruptake waslinearfor 2 min. In allexperiments reported, tracer uptake was

terminatedafter 1 min by rapidadditionofice-cold isosmotic, LiCI-Hepes rinse buffer (140 mM LiCl, 10 mM Hepes; pH 7.40, 295±2

1.Abbreviationsused in thispaper:

[Ca2+]j,

intracellularcalcium activ-ity; DiOC6(3),

3,3tdihexyloxacarbocyanine

iodide; MDCT, mouse distal convoluted tubules; NPPB,

S-nitro-2(3-phenylpropylamino)

benzoicacid; SPQ,

6-methoxy-N-(3-sulfopropyl)quinolinium.

mosmol/kg H20)

andfilteredontoWhatmanGF/Cfilters(Whatman

Inc.,

Clifton,

NJ)

using a 12-portmanifold(Millipore Continental

Water

Systems,

Bedford, MA)followedbytwoadditionalrinses with

LiCl-Hepes

buffer. Forexperimentsin which Cl- wasreplaced, an

isosmotic

Li2SO4-Hepes

buffer(140mMLi2SO4, 10 mMHepes;pH

7.40,

295±2

mosmol/kg

H20)

was

used

for rinses.

Filters

containing

cells in which22Na+wasmeasuredwerecounted inaBeckman 8000 gammacounter(BeckmanInstruments Inc.,

Ful-lerton,

CA).

Filterscontainingcells foruptakeof3Cl-or

45Ca2+

were placedin vialscontaining 10 ml ofAquasol2(New England Nuclear).

Radioactivity

wasdeterminedbyscintillationcountingwitha

Beck-manLS3801 counter(BeckmanInstrumentsInc.). Nonspecific

bind-ing

of

'Na+,

36Cl-,

or

45Ca2+

tofilters and cellswasdeterminedin the

presence of

LiCl-Hepes

(or

Li2SO4-Hepes)

buffer andsubtractedto calculate uptake. Protein contentwas measured bythe method of

Lowry

etal.

(22)

on

50-MA

aliquotsof cells. Traceruptakes are

ex-pressedasnanomolesper

minute.

mgprotein(nmolmin-'mg

pro-tein-').

Measurement

of

intracellular

Ca2`

activity ([Ca2+]i).

Cellswere growntonearconfluenceon25-mmglass coverslips(-3d).Cellson

coverslipswererinsed three times with KRB andincubatedwith

I0o-Mfura-2AM(Molecular Probes, Inc.,

Eugene,

OR)in 2.5 ml KRB for 60 minat37°Con arotatingincubator. Thecoverslips wererinsed twice with KRBtoremoveextracellular fura-2 AM and mounted ina

chamberofamicroincubation system,temperature-controlledto37°C

(MS-C;

Narishige

USA, Inc., Greenvale, NY).The chamberwas

af-fixedtothestageofaNikonDiaphotinvertedmicroscope(Nikon Inc.,

Melville, NY). Fluorescenceexcitation and emission intensitywere

measured withaNikonPhotoscan-2(Nikon Inc.).Interferencefilters of 340nmand 380nm

(Omega

OpticalInc.,Brattleboro, VT)were

employed

for fluorescence excitation. Emissionat 510 nm was

fol-lowed

by using

a 510 nmdichroic

mirror (DM510;

NikonInc.)in series witha

510±5

nminterferencefilter

(Omega

OpticalInc.).Fields oftwo tothree cellsonthecoverslipwereselectedby usingashutter assemblymounted in front ofthe

photomultiplier

tube. Anareadevoid of cellsonthecoversliporacoverslipof cellsnotincubatedwith

fura-2

AMwasusedtomeasureand subtract

background

autofluorescencein each experiment.

Background

was

typically

observedtobe <5% of

fluorescenceof cells loaded with fura-2. Beforeadministrationof

chlo-rothiazide,

abaseline

recording

of

[Ca2'+]

wasobtained.Fluorescence

intensity

wassampledatarateof 2

points/s.

Absolutevalues

of[Ca2`]J

werebasedoncalibrationsthatwereperformedasdescribed

previously

(16).

[Ca2+JI

was calculated

using

the equation derived

by

Grynkieiwcz

etal.

(23).

Membrane

voltage.

Fluorescencemeasurementsofmembrane

po-tential were performed with distal convoluted tubule cells grown on coverslips and loaded with the

voltage-sensitive

probe

3,3'dihexyloxacarbocyanine

iodide

(DiOC6(3);

Molecular Probes,

Inc.).A10-mMstockof

DiOC6

(3)inDMSOwaspreparedandfiltered

through

a

O.45-jsm

Ultrafree-MC

cartridge (Millipore

Continental

Water

Systems).

Cellson

coverslips

werepreparedandincubatedwith 10-6

M DiOC6(3)

for60 min inasimilarmannertothoseincubated

with

fura-2

AMdescribedabove.Anexcitationwavelengthof 490nm

andemission

wavelength

of 510nm wereused.Fluorescence

intensity

wassampledatarateofIpoint/s.Alterationsin membranevoltageare

presentedasrelative

changes

of

DiOC6(3)

fluorescence. Calibrations

were performed by a null-point procedure

using

graded

changes

in

externalpotassiumin thepresenceof

valinomycin

(24).

Intracellular chloride activity. Intracellular chloride

([Cl-j)

was

measuredin MDCT cellsgrownonglass

coverslips by using

the

C1--sensitivedye

6-methoxy-N-(3-sulfopropyl)quinolinium

(SPQ;

Molec-ular Probes, Inc.). A 50-mM stock solution ofSPQ in DMSO was

prepared.

Loading

solutions consistedof 5 mM

SPQ

in KRB.

Loading

of cells with

SPQ

wasaccomplished

by

preincubating

the cellsfor5 hat

37°C. After

rinsing,

the coverslipwasinserted intothe

microincuba-tionchamber describedabove.Fluorescenceintensitywasquantified

(4)

tively,wereemployed.Fluorescenceintensitywassampledat arateof 1 point/s.Attheendofeachexperiment 150mMpotassium thiocyanate

+ I0-5Mvalinomycinwasadded to quenchintracellularfluorescence

(25). SPQ leakage was estimated to be - 12%/h.Alterations in

intra-cellularchloride are presented as relative changes in SPQfluorescence. Preparationofdrugsolutions andstatisticalevaluation.

Chlorothia-zide and bumetanide (a gift from Hoffmann-LaRoche, Nutley, NJ) were prepared so that the molar concentration indicated in the text was thefinal concentrationtowhich cells were exposed. Solutions

contain-ing drugs were prepared fresh daily. Chemicals wereofthehighest

gradecommercially available. All Na+,36CI-,and45Ca2+uptake mea-surementswithinanexperimentwereperformedintriplicate.The data arepresentedasmean±SE,where nindicatesthenumberofseparate

experiments.Intracellularcalcium, ([Ca

2i]i),

intracellular chloride

ac-tivity,and membranevoltageweremeasured induplicateortriplicate

within asingleexperimentandreported as the mean±SEofnseparate

experiments. Comparisonsbetween control anddrugtreatment groups were examined by post hoc analysis ofmultiple comparisons with Newman-Keuls testsusingthestatisticalsoftwareANCOM(SciLab, Guilderland, NY).Values ofP.0.05 were assumedtobesignificant.

Results

Diureticeffectson22Na+and36CV- uptake by culturedMDCT cells. Isolation of cells from mouse kidney cortex using the

double-antibodyprocedure basedonthenephronlocalization

oftheTamm-Horsfall antigen yieldsbothcortical thick ascend-ing limb anddistal convoluted tubule cells(19).Sincemouse

(26, 27) andhuman (28)cortical thick ascending limb and distal convoluted tubule cellspossess PTH receptors,the

distin-guishingfeature between these cell typesis theirunique Na'

transportcharacteristics.Apicalmembrane Na+entryin

corti-calthickascending limbsis mediated byNa:K:2C1cotransport

andis inhibitable bysulfamoyl loop diureticssuchas bumetan-ide or furosemide (29-31). Sodium influx across luminal membranesof distal convoluted tubules is thoughttobe

me-diatedbypotassium-independentNaClcotransport(8, 32, 33)

that isselectivelyinhibitedbythiazide diuretics. This distinct

patternofsensitivitytobumetanideorthiazide diureticsallows

discriminationbetweencortical thickascendinglimb and dis-talconvoluted tubule phenotypes.

Theeffects ofthethiazidediuretic,chlorothiazide, andthe

loop diureticbumetanideonsodium and chloride uptake by MDCT cellsare depictedinFig. 1. At an extracellularNaCl

concentration of 140 mM, and in the absence ofCO2 and bicar-bonate, chlorothiazide (l0-4 M) inhibited 22Na' uptake by

40±5%from272±35to 163±29 nmol

min-'

mg

protein-'

(P

<0.05)and3C1-uptake by 50±5% from 340±50to172±37 nmolmin-' mg

protein-'

(P< 0.01). In contrast,thesame

concentration ofbumetanide had no demonstrableeffect on

either22Na+or

36C1-

uptake. Inhibition ofNa+andCl-

trans-portbychlorothiazide, coupledwith lackof effect of bumetan-ide, stronglysupport the view that MDCT cells express adistal convoluted tubulephenotype. Similar ratesof22Na' uptake

were noted in the absence ofexternal potassium (data not

shown).Theseobservationsareconsistent with the view that NaClcotransportisresponsible foramajorcomponentofNa+

uptakeintheseimmortalized distal convolutedtubule cells and

thatchlorothiazide blocks C1- entry by inhibiting NaCl

co-transport.

Chlorothiazide-induced decreasesofintracellularchloride.

Theaforementionedresultsarecompatiblewiththehypothesis

400--c

300

a

E 200

0

22Na+ 36cl

-T

CONT CTZ

Figure1. Effects ofchlorothiazide(CTZ, l0-4 M)and bumetanide (BUM, 10-4M)on22Na'and

3CP-

uptakebyMDCTcells. Barsare mean±SEoftriplicatemeasurementsinn=5 separateexperiments.

*P<0.05,**P<0.01comparedwith control(CONT).

thatby blockingchlorideentrychlorothiazide decreases intra-cellularchloride.Totestthis

directly,

wemonitored intracellu-larchlorideactivitywith thechloride-sensitive fluorescent dye, SPQ.

SPQ

fluorescence is

quenched

by

chloride and

by

other halide ions (34). Decreases of fluorescence

intensity

are

in-verselyrelatedtointracellular chloride

activity

(25, 34). The results ofarepresentative

experiment

in which

SPQ

fluores-cence wasmeasuredin responsetochlorothiazideare

depicted

inFig.2.Inthisexperiment,

blocking

Cl-entrywith

chlorothia-zide reduced intracellular chloride

activity.

When

chlorothia-zidewaswashedoutof the

bathing solution,

intracellular chlo-ride returned to control levels.

Application

of the chloride channel blocker

5-nitro-2-(3-phenylpropylamino)benzoic

acid(NPPB)resultedin aprofoundelevation of intracellular

chloride.Inthepresence of

NPPB,

chlorothiazide hadno

de-tectable effect onintracellularchloride

activity.

The effect of NPPBwasfully reversible upon

exchanging

thebathforone

devoidof

drugs.

Asecond

application

ofchlorothiazide elicited

asimilar fall of intracellular chloride

and, again,

wasreversible

uponremoval of the

drug.

Finally, application

of

potassium

thiocyanate

(+

valinomycin)

resulted in

virtually complete

quenching

ofintracellular

SPQ

fluorescence.Theseresults

indi-catethat

inhibiting

chloride entry with chlorothiazidereduced

intracellular chloride

activity

and

blocking

chloride exit in-creased intracellular chloride

activity.

These observationsare

consistent with the view that MDCT cells express

plasma

membrane chloride channels that mediate cellular chloride

ef-flux.

Effects

ofchlorothiazide

on45Ca2+

uptake

and

[Ca2"],.

Al-though

renal Ca2+ andNa'

absorption normally

proceed

in

parallel,

thiazide diuretics dissociate Na+ from

Ca2`

move-mentby

inhibiting

sodium transport and

enhancing

calcium

absorption

indistal convoluted tubules

(6). Lajeunesse

etal.

(15)demonstrated that the

hypocalciuric

effect of thiazide

di-uretics is dueto increased

Ca2`

uptake

acrossluminal

mem-branes of distal convolutedtubule cells. To determine if

thia-zide diureticspromote

Ca2`

transportin MDCT cellswe

mea-sured

"5Ca2" uptake

and

[Ca2+]i

in these cells. The results,

illustratedinFig. 3demonstrate thatchlorothiazideincreased

45Ca2`

uptake by 45±4% from2.86±0.26to4.15±0.25 nmol

min-'mg

protein'-

(P

<0.01

).

Bumetanide hadnodetectable

effect on 45Ca2+

uptake.

The resultsof

experiments

in which

the effect ofchlorothiazideon

[Ca2

]i

wasmeasuredareshown

inFigs.4and 5. Basal

[Ca2+

]i averaged

91±12nM.

Composite

Thiazide-stimulatedCa2+ TransportbyDistalConvoluted Tubule Cells 431

(5)

CTZ NPPB

,Q,^1 ~~~~~~~~~~~KSCN

v

nq*\

i,<

l~~~~~~~~~~~~~~~

\

s

1'

~~~~~~~I

1~ ~ ~

A...

A

1-l , '~ ., |. Is , .

0 5 15 20 25 30 40

Time, min

45

Figure2.Representative recordingofchangesofintracellular

C1-measuredintwo tothree cellsgrownoncoverslipsandloadedwith

thefluorescent indicator, 6-methoxy-N-(3-sulfopropyl)-quinolinium

(SPQ).Decreases of fluorescenceintensityareinversely relatedto intracellular chloride activity. Additionofdrug isindicated bythe

shaded regionofthe barabovetherecording.Chlorothiazide(CTZ, 10-4 M) rapidly decreasedintracellular CP.Aftera10-minexchange

ofthebathingsolutioncontainingCTZ foronedevoidofdrug, intra-cellularCP returnedtocontrol levels.Applicationof theCl--channel blocker5-nitro-2(3-phenylpropylamino)benzoicacid(NPPB, lo-5 M) increased intracellular CI-.Asecond applicationofCTZ inthe

presenceof NPPB blocked thedecrease ofintracellular CIP. Upon

exchangeoftheextracellularbath foronedevoid of CTZ and NPPB, intracellularCP-returnedtocontrol levels.Athirdapplicationof

CTZ resultedinareversiblefall of intracellularchloride comparable

inmagnitudetothatobservedonthe firstapplication.Cellswere

rinsed andpotassium thiocyanate (KSCN, 150 1-0 M)+5 10-6M valinomycinwasaddedtoquench intracellular SPQfluorescence

(25). Comparable resultswereobtainedinthreeadditional

experi-ments.

cJ

50 7S

f00

lt

CTZ, PM

50 75 100 250

10 15 20

Time, min

Figure4. Chlorothiazide-inducedelevations of intracellular calcium

(ICa2+i)inMDCT cells. Recording showstheresultsofa

represen-tativeexperimentinwhichincreasing concentrationsof chlorothia-zide (CTZ; 1-250 AtM)wereaddedtocellsgrownoncoverslips and

loaded withthefluorescent indicator, fura-2 AM. CTZ concentrations 2 10 Mcaused immediate increases of [Ca2+]. These resultsare

typicalofatotalofsevendeterminations.

achievedwith Io-4 Mchlorothiazide.Theresponse appearsto

bebiphasic:atconcentrations below 50 1o-6Mchlorothiazide increased

[Ca2+]i

inasigmoidal fashionbutasharp increase of

[Ca2+]i

was noted at 75 and 100 10-6 M. The origin of this

biphasicpattern, isuncertain.

Hyperpolarizationofmembrane voltage bychlorothiazide.

Thiazide diuretics hyperpolarize basolateralmembranes of

am-phibian distal tubule cells(11, 12)and flounder urinary

blad-der cells(13). Moreover, in thosestudies,hyperpolarizationof basolateralmembrane voltagewasaccompanied by parallel de-creasesofintracellularCl-activity.Theresults showninFigs.1

and 2 and Table I arecompatible with the hypothesis that the

dose-response relations between chlorothiazide and

[Ca2"]i

arepresentedinFig. 5. Chlorothiazide elicited

concentration-dependent increases of[Ca2+]ithatoccurredwith successive

additions of the drug. Maximalincreasesto654±31 nMwere

5-._

a) 2 le1

.5m

Co

0

Ec

E r_

4-

3-

2-

1-

0-I

CONT CTZ BUM

Figure3. Effects of chlorothiazide(CTZ, 10-4M)andbumetanide (BUM, 10-4M)on45Ca2" uptake by MDCTcells.Barsrepresent

themean±SE of triplicatemeasurementsinn=5separate

experi-ments. **P<0.01.

800

700

600

500-Cu

400-

300-200.

100.

10

Chlorothiazide,pM

100

Figure5. Dose-responserelationsof chlorothiazideonintracellular

calcium(QCa2 ],)inMDCT cellsgrown oncoverslipsand loaded

with fura-2AM.Addition of chlorothiazideresultedin

dose-depen-dent increasesof[Ca2"i.Each pointrepresentsthe mean±SEof

sevenindependent experiments.

432 F. A.Gesek andP. A.Friedman cF~

ax

c2

ax

cn

a

EL

**

(6)

Table I. Effectof ChlorideReplacementon Chlorothiazide-stimulated

'Ca2"

Uptake

Chloride Valinomycin Chlorothiazide 45Ca2+uptake

mm nmolmin-'mgprotein-'

o - - 2.19±0.30

0 - + 2.24±0.20

o + - 2.25±0.22

0 + + 2.17±0.23

140 + - 2.41±0.28

140 + + 2.43±0.30

45Ca2+uptakewasmeasured asdescribed in Methods. Isosmotic

chloridereplacementwasachievedwithgluconate.Thefinal

concen-trationofvalinomycinwasIo-'M andchlorothiazidewas10-4M. Values representmean±SE oftriplicatemeasurementsperformedin three separateexperiments.

hyperpolarization induced by chlorothiazide in MDCT cells resultedfrominhibition of luminal membrane NaCl cotrans-port withanattendant decreaseofintracellular Cl-.To deter-mine ifthiazidediureticsaffectmembrane voltagein MDCT cells, voltage was measured with the fluorescent probe DiOC6(3).Intheexperimentdepicted in Fig. 6,wefollowed membrane voltage as afunctionof time in MDCT cells grown

onglasscoverslips. Reductionsin fluorescencecorrespondto

hyperpolarizationof membrane voltage.Chlorothiazide(10-4

CTZ BUM K+

1 2 3 3 4

Time,min

Figure 6. Effects of chlorothiazide (CTZ, l0-' M)and bumetanide (BUM, l0-4 M)onmembranevoltageinMDCT cells.Changesin membrane potentialweremeasured in cellsgrown oncoverslipsand

loaded with the fluorescentindicator, 3,3'-dihexyloxacarbocyanine [DiOC6(3)]. Decreasesof fluorescenceintensityareinverselyrelated

tomembranevoltage.CTZ causedarapid hyperpolarization that

wascompletely reverseduponwashout of thedrug.BUM hadno

no-table effectonDiOC6(3)fluorescence.Anull-point calibration,

per-formedatthe end of eachexperimentwithgradedconcentrations of

externalK+in thepresenceof 10-6Mvalinomycin,isshown at the right-hand portionof therecording.Similar observationsweremade

inatotal offive experiments.

M) decreased fluorescence by - 30%. Upon exchanging the

extracellular bathing solution foronedevoid of the drug,

fluo-rescenceintensity returnedtocontrol levels.Incontrast tothe hyperpolarization elicited by chlorothiazide, bumetanide (10-4 M) hadnodetectable effect on membrane voltage. As

shown in the right-hand portion of the recording, changes in DiOC6(3) fluorescence were calibrated with graded

concentra-tions of extracellular potassium in the presence of the potas-sium ionophore valinomycin (24). Resting membrane voltage was calculatedfrom the average null-point value for extracellu-larpotassium (8.13±0.60 mM, n=8) and the measured intra-cellular potassium (1 19±9 mM, n = 5) and averaged -72±2 mV (n =8). Fig.7shows thatchlorothiazide elicited stepwise decreases in fluorescence with incremental concentrations of thedrug. Detectable changes in fluorescence were noted with concentrations of chlorothiazide aslowas 10-6M. Maximal decreases offluorescence intensity of 35±13% were achieved at chlorothiazideconcentrations of -1 X 10-4to 2.5x 10-4M.

Again, chlorothiazide-induced hyperpolarization was rapidly and completelyreversibleupon removalofthe drug. This ob-servationindicates that the reduction in fluorescence wasnot

due to either some toxic action of the drug or diffusion of the dye. The average maximal reduction of fluorescence

corre-sponds to ahyperpolarization of 20.4±7.2 mV (n =5). These results demonstrate that chlorothiazide hyperpolarizes mam-malian distal convoluted tubule cells.

Todetermine if alterations of membrane voltage directly influence

Ca2"

influx, MDCT cells were hyperpolarized or de-polarized by varying the extracellular potassium concentration in the presence ofvalinomycin (10-5 M). Fig. 8 shows that

45Ca2"

uptakeincreased when cellswerehyperpolarized. Con-versely,45Ca2+uptake decreased whencellsweredepolarized. Asshown intheinset to Fig. 8, independent experiments

dem-onstratethat thesealterationsin K+elicited linear Nernstian responses of DiOC6(3) fluorescence and the appropriate changes in membrane voltage. These results support the view

a)

C

-C

0c

C.

Co

a) 0

15 Time, min

Figure7.Concentration dependenceof chlorothiazide-induced

cellu-larhyperpolarization.Alterationsinmembranevoltagewere

mea-sured withDiOC6(3)asdescribed in Methods. Therecordingshows

thatuponaddition ofgradedconcentrations of chlorothiazide(1-250 ,uM), dose-dependentdecreases of fluorescenceintensitywere

ob-served.UponwashoutofCTZ,membranevoltagereturnedto control

levels.Comparableresultswereobtained inatotalof fiveexperiments.

Thiazide-stimulatedCa2" Transport by Distal Convoluted Tubule Cells 433

0)

4-c C co

0

.)

C,) 0

0 CO

(7)

CT7 _

Cl- 140 14 50 140

K+ _ _ | 50 | 5

10.51

315

130150l

/A

_ Ir"//// //1 //I:II11H

0 5 15 20 30 35 50 55 60 Time,min

Figure8. Effects of membranevoltageon

45Ca2"

uptake byMDCT

cells. Cellular voltagewasalteredby varyingextracellular[K+]inthe

presenceof o0-6Mvalinomycin.Reduction of theextracellular [K+] concentrationbelow5 mMincreased

45Ca2"

influx.Conversely, ele-vation oftheextracellular potassiumconcentration above 5 mM de-creased45Ca2"influx. Theinset shows therelations between extracel-lular[K+]andDiOC6(3)fluorescence, usedas anindex of

mem-brane voltage, and confirmsthataltering extracellular[K+]hadits expectedeffectsonmembranevoltage.Pointsrepresentthe mean±SE

of five separateexperimentsoftriplicatedeterminations of 45Ca2+

uptake.

that variations in membrane voltage can directly modulate

Ca2+entry.

Ifchlorothiazide-stimulatedCa2+influxis duetochanges of membrane voltagethat have theirorigininadecrease of the

chloride diffusionpotential,thenreplacementofextracellular CP- should depolarize distal convoluted tubule cells thereby

preventing stimulation of calcium uptake by chlorothiazide. As shown in Fig. 9, in thepresence of 140 mM extracellular NaCl,chlorothiazide hyperpolarized MDCT cells. Thiseffect

wasreversibleuponremoval of the drug. Isosmotic reduction

ofextracellular chlorideto 14 mMresultedindepolarizationof membrane voltage. These resultsprovideadditional evidence for thepresenceof chloride channelsinplasma membranes of MDCT cells. Moreover,atthis reduced extracellular chloride, chlorothiazide hadnodemonstrable effectonmembrane volt-age. As expected, such membrane depolarization inhibited

45Ca2+uptake(Table I) and blocked the stimulatory effect of chlorothiazide noted earlier. These findings support the

hy-pothesisthatthehyperpolarizationinducedby chlorothiazide

occurs as aresultof blockade of NaClcotransportwitha

con-comitant reduction ofintracellular Cl- activity. This conclu-sion isstrengthenedby theadditional observationsthat when membrane voltagewasclamped with valinomycinthe

stimula-tory effect ofchlorothiazide on 45Ca2+ uptake was blocked,

eitherinthepresence orabsence ofextracellularchloride

(Ta-bleI).

Ifcellular chloride efflux is mediated by chloride channels, thenblockade of thesechannelswith NPPB would be expected topreventthehyperpolarizationnormally induced by chloro-thiazide, which inturnshould inhibit 45Ca2+ uptakestimulated

by the diuretic. Fig. 10 shows that, asbefore, application of

chlorothiazide reversiblyhyperpolarizeddistalconvoluted tu-bule cells.Applicationof NPPB ( 10-5 M) alonecausedaslight

Figure 9. Effectsofreducedextracellular chlorideon chlorothiazide-induced membrane hyperpolarization measured with DiOC6(3)in

MDCT cells. Chlorothiazide (CTZ, 10'- M) reversibly hyperpolar-ized membrane voltageasshown in Figs. 6and 7. Exchange of

incu-bation medium containing 14 mMextracellularCP-(balance

gluco-nate)causedamodest depolarization but blocked completely the hyperpolarizingeffect ofasecond applicationofCTZ. Rinsing and

replacementwith 140 mMCl- plus 50mMK+ elicitedapronounced

depolarization.Anull-point calibration,shownonthe right,was

per-formed with graded concentrations of [K+]+ IO-sMvalinomycin.

Similar effectswereobservedinatotal of fiveexperiments.

andtransientdepolarizationof membrane voltage. However, the effect of subsequent addition ofchlorothiazide on

mem-brane voltage was abolished. Thus, both pharmacological

blockade of chloride channels and low extracellular chloride

CTZ NPPB

C

c~

c

a>)Cnu

(D

C._

ch

40 25

Time,min

Figure10. Effectofthechloride channel blocker NPPBon

chlorothi-azide-induced membranehyperpolarizationinMDCT cells. Voltage

wasmeasuredwithDiOC6(3). Thisrecording showsthe

hyperpolar-izingeffect of chlorothiazide (CTZ, l0-4 M)asthedecreaseof

fluo-rescenceintensity.Uponremoval ofCTZmembrane voltage returned

tocontrol levels. Addition of theCl- channel blocker NPPB(I0-5 M) abolishedthehyperpolarizing effect ofasubsequent application

of CTZ. Similar resultswereobtainedwithpolythiazideintwo addi-tionalexperiments.

434 F.A.Gesekand P. A. Friedman

.E 3

u. E-§

wCL

S

0 1._

e)

c

v

0

p

cR

CD

0 a

(8)

prevent membrane hyperpolarization by chlorothiazide. We nexttested theeffect ofNPPB on45Ca2l uptake. These results are shown in Table II and demonstrate that chlorothiazide-stimulated

45Ca2"

uptake was abolished by NPPB.

Cellular mechanism ofthiazide-stimulatedCa2 uptake. We recently reported that PTH stimulatesCa2" entrythrough

di-hydropyridine-sensitive calcium channels in primary cultures

of cortical thick ascending limband distal convolutedtubule

cells (16). To determine if chlorothiazide-stimulated

Ca2"

entryisalso mediated bydihydropyridine-sensitivechannels weexaminedtheeffect ofthe calcium channel antagonist

nifed-ipine (10-5 M) on chlorothiazide-stimulated

45Ca2"

uptake

and on

[Ca2+]i.

As shown in Table III, nifedipine abolished

chlorothiazide-stimulated

45Ca2"

uptake. Similarly, the rise of

[Ca2+]i

evokedbychlorothiazidewascompletely blocked by

nifedipine.Inneither instance did nifedipinehaveany

inhibi-tory action of its own. These results provide additional and directevidencethatcalciumentryindistal convoluted tubule cellselicited bychlorothiazideis mediated by

dihydropyridine-sensitive

Ca2"

channels and isstimulatedbyhyperpolarization.

Discussion

Under most physiological conditions, renal sodium and cal-cium absorption proceed in parallel. However, after certain physiological, pathophysiological, or pharmacological inter-ventions,theabsorption of calcium and sodiummaybe disso-ciated. Ineachofthese situations the dissociation of calcium and sodiumtransport occursin distal convoluted tubules or,

dependingonthespecies,thedistal nephronsegment

respon-siveto PTHorsensitive tothiazide diureticsor toamiloride.

The cellular mechanismofactionof thiazide diureticsonNa' absorption hasbeen attributed toinhibitionof luminal NaCl

cotransport(32, 35) or toparallel

Na+/H'

and

Cl-/HCO3

exchange(12). In contrast,themechanismbywhichthiazide

diureticsstimulate distal tubularCa2' absorptionand the

na-tureofthe relations between NaClabsorptionandCa2+

trans-port have yettobe resolved.

Elucidation ofthecellular mechanism responsible for the

stimulationofCa2' absorption bythiazide diuretics has been hampered by theheterogeneityof distal tubulesegments.This

portion ofthenephroniscomprised ofthree

morphologically

and functionally distinct segments: distal convoluted tubule, connecting tubule,and initial

collecting

tubule

(36-38).

Fur-thermore, the shortlengthof distal convoluted tubules

(<

0.5 mm) limits direct study by

microperfusion

techniques. The

developmentand

application

of immunodissection

techniques

Table II. Effect ofthe

CV

ChannelBlocker NPPB

onChlorothiazide-stimulated'5Ca2" Uptake

Control CTZ CTZ+NPPB NPPB nmolmin'mgprotein-'

45Ca2+uptake 2.71±0.19 3.78±0.17* 2.74±0.27 2.58±0.16

n 8 8 6 8

45Ca2+uptakewasmeasuredasdescribed in Methods. The final

con-centrationofchlorothiazide (CTZ)was1O-4MandNPPB wasIO-'

M.Valuesrepresent themean±SEof triplicate measurements per-formedinnseparateexperiments. *P<0.01 comparedwithcontrol levelof45Ca2"uptake.

( 19, 39, 40) to the isolation and primary cell culture with sub-sequent cell immortalization ofdistal convoluted tubule cells have greatly facilitated the ability to investigate the

mecha-nismsofNa'andCa2" transport. Thedatadepictedin Figs. 1 and 3 indicate that MDCT cells expressNaCltransport that is

selectivelyinhibited and calcium uptake that is stimulated by

chlorothiazide. These results represent a virtual signature of murine distalconvolutedtubule cells. Additional preliminary

evidence and further characterization of these distal convo-luted tubule cells has been published (41). It should be noted

thatthereis considerable heterogeneityoftransportfunction in mammalian distal nephron segments. Thus, whereas PTH and

thiazide diureticsexerttheireffectsin distal convoluted tubules inmice, rats, and humans (26, 28, 42), similarreceptorsand transportfunctionsare expressed in connecting tubules in the

rabbit(43, 44). As noted above, in the present study

chlorothia-zide directly stimulated

45Ca2"

uptake and a rise of

[Ca2+]i.

Somewhat differentresults were recently reported by Shimizu etal. (45). TheyfoundthatPTHpretreatmentof rabbit

con-nectingtubules was necessary to elicit an effect oftrichlormethi-azide on calcium transport (45), but not sodium absorption

(46).Inourexperiments,cells were grown in culture media in which the serum was heat inactivated. Further, cells were serumstarved for18 hbefore

45Ca2+

uptakeexperimentsor for 1 hbefore the beginning of experimentsin which

[Ca2+]i

was

measured. Hence, thedifference between the two studies is

unlikelytostem from residualPTHactivityin theincubation

media weemployed.

Three hypotheseshave been proposed to accountfor

thia-zide-induced dissociationofsodiumandcalcium transport in

distal nephrons. One explanation advancedbyWalser (47)was

thatthiazidediuretics reduce the transepithelial voltageand, as aresult, decrease therateof passive calcium backflux intothe

tubular lumen.Twolinesofevidence havesince indicated that thedissociation ofNa'from

Ca2+

transportafter administra-tion of thiazide diuretics doesnotoccur by such amechanism. First, thiazide diureticshavebeen shown to havelittle, ifany,

effect on transepithelial voltage in mammalian distal convo-lutedtubules (46) and, second, the passive calcium permeabil-ity of thisnephron segmentis remarkablylow(6).An alterna-tive theory proposed by Costanzo and Windhager (48)

sug-gested that the reciprocal association between sodium and calciumtransport was a consequenceof basolateral membrane

Na+/Ca2'

exchange.Accordingtothisexplanation, by inhibit-ingapicalmembranesodiumentry,thiazide diuretics wouldbe

expectedto reducethe intracellularNa' concentration. This decrease ofintracellular Na+ would inturnincrease the electro-chemical gradient for basolateral sodium entry, thereby

aug-menting calciumeffluxthrough the

Na+/Ca2'

exchanger. Shi-mizuetal. (45) reported studies in which the stimulation of calcium absorption by isolatedmicroperfusedrabbit

connect-ing tubuleswasthoughttobemediated by increased

Na+/Ca2+

exchange. Severallines of evidence introduced in thepresent

study suggest that enhanced

Na+/Ca2'

exchange is not the

primary mechanism by which thiazide diuretics dissociate

so-diumandcalciumabsorptionindistal convoluted tubules.The

observationthatchlorothiazideincreased intracellular calcium activity(Figs.4and5) and stimulated calcium influx

(Fig. 3)

in distal convoluted tubulecellsisoppositetowhatwould be predictedwere

Na+/Ca2+

exchangeresponsibleformediating

thedissociationinducedbythiazide diuretics(45).

The mechanismweproposetoaccountforthe stimulatory

(9)

TableIII.EffectofNifedipineonChlorothiazide-stimulated"5Ca2+Uptakeand

[Ca2+]i

Control CTZ Nifedipine+ CTZ Nifedipine n

45Ca2+uptake(nmol min-'mgprotein-') 2.86±0.25 4.15±0.24* 2.91±0.32 2.84±0.30 5-6

[Ca2'+i

(nM) 102±7 562±8* 103±7 99±8 3

45Ca2+uptakeand

[Ca2+]i

weremeasuredasdescribed in Methods. The final concentration of chlorothiazidewas10-4M and nifedipinewas

10-5

M. For45Ca2'uptakethe values representthemean±SEoftriplicatemeasurementsperformedin the indicated number of experiments.[Ca2+]i was measured in theindicated numberofexperiments. * P<0.01 comparedwith control levels of45Ca2+uptake and[Ca2+]i.

action of thiazide diureticsoncellular calcium influx involves

four elements: (a) thiazide diuretics blockapical membrane chlorideentryby NaClcotransport;(b) continuedbasolateral chlorideefflux throughchloride channels results inadecrease

ofintracellularchlorideactivity; (c)thereduction ofintracellu-lar chloridehyperpolarizes apical and basolateral cell

mem-branes; and(d)membranehyperpolarizationincreasesapical

membrane calcium entry through dihydropyridine-sensitive

calcium channels. Suchamechanism is supported bythe fol-lowing observations.

Thefirstcomponentof the proposed mechanism involves theinhibition bythiazide diuretics of coupled NaCl uptake.

The data presented in Fig. 1 indicate thatchlorothiazide in-hibitssodium uptake by MDCT cells. Approximately40%of sodiumtransportwasinhibitedby l0-4Mchlorothiazide. The magnitude of this inhibitionis comparabletothatreportedby Costanzo(7) inearlydistaltubulesoftheratandShimizuetal.

(46)inrabbitconnecting tubules,butnotby Ellisonetal.(8)

whofound that IO-3Mchlorothiazide abolished NaCl absorp-tion in early distal tubules of the rat. Inhibition of sodium uptake by chlorothiazide may have been incomplete in our

experiments fortwo reasons.First,inseparateexperimentswe

found that, in additiontoNaClcotransport, sodiumentryis mediated byamiloride-sensitive sodium channels (49).

To-gether,thesetwo entrymechanismsaccountfor - 88-90%of

ouabain-suppressiblesodiumaccumulation.Shimizuetal. sim-ilarly found additive inhibitory effects of trichlormethiazide

andamilorideon sodiumtransport (46) andstimulatory ef-fectsoncalciumabsorption(45) insingle microperfused

con-necting tubules. Second, chlorothiazide is a relatively weak blockerof sodiumabsorption(50).More potentthiazide

ana-logs suchaspolythiazide, cyclothiazide, ormetolazone inhibit sodiumtransportbyanadditional 5-10%. Thereasonfor the

partialinhibitionofchloride uptake (Fig. 1) bychlorothiazide

maybemorecomplicated and relatetothenatureof the thia-zide binding siteonthecell membrane. Tran et al. (51)

sug-gested that the NaClcotransporter possessesseparatebinding

sites forNa+ andfor C1-. TheCl- binding siteisshared by

thiazide diureticsand by thiazide-like drugs such as metola-zone.Sodium binding is thoughttoincreasetheaffinity of the

anion site. The binding ofCl-to the transporterpermits Na+

andCl- influx across the cell membrane.Thiazide diuretics competewithCl-forthe anionbindingsite on thetransporter,

thereby preventingthe transport of Na+ andCl-. Indeed, in separate experiments we found that polythiazide inhibited

36CI-

accumulation by 81±9% at reducedextracellular

chlo-ride concentrations(80 mM) (49) but only by 57% at 140 mM

chloride (unpublished observations),consistentwith the

pre-dictionofTran et al. ( 51 ).

Accordingtotheproposed mechanism,inhibition of chlo-ride entry into MDCT cells results in a fall of intracellular chlorideactivity. The fact that chlorothiazide reduced intracel-lularchloride activityis establishedbyexperimentssuchasthat shown inFig. 2. Suchadecreaseimplies first that intracellular chloride is above its electrochemicalequilibrium,andsecond,

thatthereisaneffluxmechanism for chloride. Control intra-cellular chloride estimated from the Goldman-Hodgkin-Katz equationwas22mM, clearlygreaterthan the10mM predicted from its passive distributionattheresting membrane voltage, -72mV. Afteradditionofchlorothiazide,membranevoltage

hyperpolarized by 20mV.Themagnitude ofthis

hyperpolariza-tionas assessedwith the fluorescent probe, DiOC6(3) in the

presentstudy(20mV) is comparabletothatreported by others

usingconventional microelectrodes: 10-26 mV(11-13, 52). Hyperpolarization of20 mVto-90to-92mVwould

corre-spondto acalculatedintracellularchlorideactivityof 5 mM.

Aftertreatmentwithchlorothiazide,theestimateofthe actual change ofintracellular chloride derived fromthe increase in

SPQ fluorescence (Fig. 2)was6.1±0.3 mM(n = 3). Toour

knowledge, these experiments representthe first demonstra-tion that thiazide diuretics decrease intracellular chloride in mammalian renal tubular cells. However, ample precedentfor

suchanaction has been established. Forinstance,Stanton (12)

and Duffeyand Frizzell (13) showed that thiazide diuretics reduceintracellularchloride inAmphiumadistal tubules and

flounder bladder, respectively. Furthermore, loop diuretics

such as bumetanide or furosemide have been shown to

de-crease intracellular chloride in cortical thickascending limb

cells(31, 53, 54).

The observation that the chloride channelblocker NPPB

prevented thehyperpolarizing effect ofchlorothiazide (Fig. 10) is consistent with theviewthat basolateral chloride channels mediate chloride effilux. Further, insofaras NPPB increased intracellular chloride activity and abolishedtheeffect of

chloro-thiazide on intracellularchlorideactivity (Fig. 2), thesedata suggestthatchloridechannels,atleastinpart,mediate chloride

effiluxin MDCTcells.Abasolateralchloride conductancehas

been reported in rabbit distal convoluted tubules (55). The

observation that membrane voltage decreasedto a greater ex-tentupon alog-foldchange in extracellular potassiumthan to anequivalentchange in extracellular chloride (Fig. 9)suggests

that MDCT cells also possess plasma membrane potassium channelsandthat, undercontrolconditions, potassium

con-ductanceisgreaterthanchlorideconductance. Thepresenceof plasma membranechlorideandpotassiumchannelsprovidesa

plausibleexplanation for thehyperpolarization of membrane

voltageinduced by chlorothiazidein MDCT cells (Figs. 6 and

7), the thirdcomponentofourmodel. Inparticular, because

(10)

intracellularchlorideis above its electrochemicalequilibrium,

chlorothiazide-induced blockade of chlorideentrymediatedby

the NaCl cotransporter results in a decrease of intracellular chloride since it continues to leavethe cell through chloride channels.Thefall of intracellular chloride inturndecreases the

magnitude ofthe chloride equilibriumpotential and the cell hyperpolarizes toward the potassium equilibrium voltage. Thus, the mechanismby which thiazide diuretics inhibit NaCl

cotransportindistal convoluted tubule cellssuggestedhere fol-lowsexplicitlythe model proposedby Wittneretal.(31) forthe

hyperpolarizationof membranevoltageand theinhibitory

ac-tion of loopdiureticsonNa:K:2Clcotransportinthick ascend-ing limbs.

The fourthelement ofthe mechanismwepropose forthe

stimulatory action of thiazide diuretics oncalcium transport

requires that the induced hyperpolarization increase calcium

entryby distal convoluted tubule cells. The evidence insupport

of thiscontentionin shown inFig. 8 and inTablesIandII. In somerespectsthis is themostdifficultpointtoestablish,

there-fore, five independent approacheswere employed. First,

pre-venting thiazide-induced hyperpolarization with the chloride channel blockerNPPBabolished thestimulatory effectof chlo-rothiazideon

45Ca2"

uptake (Table II). Second, depolarizing

membranevoltage by chloridereplacementabolishedthe stim-ulatory action ofchlorothiazide on45Ca2+ uptake (Table I). Third, clamping the membrane voltage with valinomycin

blocked the inhibitory effect of chloride removal on 45Ca2+

uptake (Table I). Fourth, clamping the membrane voltage

withvalinomycinin thepresenceofnormal extracellular

chlo-rideblocked thestimulatoryeffectof chlorothiazideon

45Ca2+

uptake (Table I). Fifth, directly

manipulating

the membrane voltage byvaryingextracellularpotassiumcaused

proportion-ate changes in

45Ca2+

uptake

(Fig.

8). Thus,

hyperpolarizing

the membranevoltagestimulated45Ca2+ uptakewhereas

depo-larizing the membrane reduced it.

In previousexperiments we found that calciumtransport

activatedby PTHinprimaryculturesof cortical thick ascend-ing limb plus distal convoluted tubule cells was mediatedby

dihydropyridine-sensitive

calcium channels

(16).

The results ofexperimentsshown in TableIII indicate that calciumentry

stimulatedby chlorothiazide in MDCT cells is also mediated

by

dihydropyridine-sensitive

calcium channels.Inrecent

pilot

experiments,wehavebeen abletodemonstrate thepresenceof

calcium channels in

apical

plasma membranes of these cells

(56). Further,theseobservationsare

entirely

compatible

with resultsreported byLajeunesseetal.( 15),whofoundthat thia-zide diuretics increased calcium uptake by distal tubule lu-minal membrane vesicles.

In summary, ourresults show thatchlorothiazideinhibits sodium and chlorideentry,stimulatescalcium

uptake,

and

in-creasesintracellularcalcium

activity

inimmortalized MDCT cells. Inhibition of chlorideentrywasassociated withafallof

intracellular chlorideactivitythat resultedfrom continued ef-flux in theabsenceofadditional chlorideentry.Cellular chlo-ride effluxwasmediatedbychloride channels.Asintracellular chlorideactivity decreased,thecell membrane

hyperpolarized.

Membrane hyperpolarization, in turn, stimulated calcium

entry.Chlorothiazide-stimulated calciumentrywasmediated

by dihydropyridine-inhibitable calcium channels. These

re-sults

explain

themechanism

by

whichthiazidediuretics aug-mentcalciumtransport.

Acknowledaments

The excellent technical assistanceofMs. B.Coutermarsh isgreatly appreciated.The authors also thank Mr. CharlesFanghella and the NikonCorporationfor support and loans ofequipment.

Thesestudiesweresupportedby National Institutes ofHealth grant GM-34399. F. A. Gesek received support from National Institutesof

Health grant DK-07301.

References

1.Lamberg, B.-A.,and B.Kuhlback.1959.Effect ofchlorothiazideand hydro-chlorothiazideontheexcretion of calcium in urine. Scand.J.Clin.Lab.Invest.

11:351-357.

2.Suki,W.,F.C. Rector, Jr.,andD. W.Seldin. 1965.Thesiteof action of furosemideandothersulfonamide diuretics.J.Clin.Invest.44:1458-1469.

3.Meng,K.1967.Mikropunktionsuntersuchungenuberdiesaluretische wir-kungvonhydrochlorothiazid,acetazolamide undfurosemid.NSArch.

Pharma-col. 257:355-371.

4.Kunau,R.T.,Jr.,D. R.Weller,and H. L. Webb.1975.Clarification ofthe siteof action of chlorothiazidein theratnephron.J.Clin.Invest.56:401-407.

5.Edwards,B.R.,P.G.Baer,R. A.L.Sutton,and J. H.Dirks.1973.

Micro-puncturestudyof diureticeffectsonsodiumandcalciumreabsorptionin thedog nephron.J. Clin.Invest.52:2418-2427.

6.Costanzo,L.S.,andE. E.Windhager.1978.Calciumandsodium transport bythedistal convoluted tubuleoftherat.Am.J.Physiol.235:F492-F506.

7.Costanzo,L.S.1985. Localization of diuretic actioninmicroperfusedrat

distaltubules:Ca and Na transport. Am. J.Physiol.248:F527-F535.

8.Ellison,D.H., H.Velfizquez,and F. S.Wright. 1987.Thiazide-sensitive sodiumchloride cotransportinearlydistaltubule.Am.J.Physiol.

253:F546-F554.

9.Costanzo,L.S.,and I. M.Weiner. 1974. Onthehypocalciuricaction of chlorothiazide. J.Clin. Invest.54:628-637.

10.Lavender,A.R.,and T. N. Pullman.1961.The renaleffectsof chlorothia-zide.J.Pharmacol. Exp.Ther. 134:281-285.

11.Hansen, L.L.,A.R.Schilling,and M. Wiederholt. 1981.Effect of cal-cium,furosemideandchlorothiazideonnetvolumereabsorptionandbasolateral

membranepotentialof distaltubule.PfluegersArch.Eur.J.Physiol. 389:121-126.

12. Stanton,B. A. 1990. Cellular actionsof thiazide diureticsin the distal tubule. J.Am.Soc.Nephrol.1:832-836.

13.Duffey,M.E.,and R.A.Frizzell. 1984.Flounderurinarybladder mecha-nism ofinhibitionbyhydrochlorothiazide (HCT).Fed.Proc. 43:444.(Abstr.)

14.Velazquez, H.,and R.Greger.1986. Influencesonbasolateral K

conduc-tanceofcellsofearlydistalconvoluted tubule.KidneyInt. 29:394a.(Abstr.)

15.Lajeunesse, D.,M.G.Brunette,and J.Mailloux. 1991.Thehypocalciuric effect of thiazides: subcellular localization ofthe action.PfluegersArch.Eur.J. Physiol.417:454-462.

16.Bacskai,B.J.,and P.A.Friedman.1990.Activationof latentCa2" chan-nels inrenalepithelialcellsbyparathyroidhormone. Nature(Lond.). 347:388-391.

17.Bourdeau,J.E.,and M. B.Burg.1980.Effect ofPTHoncalcium transport across the cortical thick ascending limb ofHenle's loop. Am. J. Physiol.

239:F121-F126.

18.Costanzo,L.S.,and E. E.Windhager.1980.Effects ofPTH, ADH,and cyclicAMPondistal tubular Ca and Nareabsorption.Am.J.Physiol.

239:F478-F485.

19.Pizzonia,J.H.,F.A.Gesek,S. M.Kennedy,B. A.Coutermarsh,B.J. Bacskai,and P.A.Friedman. 1991.Immunomagnetic separation, primary

cul-tureandcharacterizationof corticalthickascendinglimbplusdistal convoluted tubule cellsfrommousekidney.In VitroCell. &Dev.Biol.27A:409-416.

20.Arend,L.J.,J. S.Handler,J.S.Rhim,F.Gusovsky,and W. S.Spielman.

1989. Adenosine-sensitivephosphoinositideturnoverinanewlyestablishedrenal

cell line.Am.J.Physiol.256:F1067-F1074.

21.Gesek,F.A.,andA.C. Schoolwerth. 1990. Hormonal interactionswith

theproximal Na+-H+exchanger.Am.J.Physiol.258:F514-F521.

22.Lowry,0. H., N. J.Rosebrough,A.L.Farr,andR. J.Randall. 1951. Protein measurementwith theFolinphenolreagent.J.Biol. Chem. 193:265-275.

23.Grynkieiwcz, G.,M.Poenie,andR. Y.Tsien. 1985.Anewgenerationof

Ca2" indicators withgreatlyimprovedfluorescenceproperties. J. Biol. Chem. 260:3440-3450.

24.Zeidel,M.L.,D.Kikeri,P.Silva,M.Burrowes,andB. M.Brenner. 1988.

Atrial natriureticpeptidesinhibitconductive sodiumuptake byrabbit inner med-ullarycollectingductcells.J.Clin.Invest. 82:1067-1074.

(11)

25. Verkman, A.S. 1990. Development and biological applications of

chlo-ride-sensitive fluorescent indicators.Am.J.Physiol.259:C375-C388. 26. Brunette, M. G., D. Chabard&s, M. Imbert-Teboul, A. Clique, M.

Mont6gut,and F. Morel. 1979.Hormone-sensitiveadenylate cyclase along the nephronofgenetically hypophosphatemic mice.KidneyInt. 15:357-369.

27.Chabardes,D., M.Imbert-Teboul,M.Gagnon-Brunette,and F. Morel. 1978.Differenthormonal target sitesalongthe mouse and rabbitnephrons.In

Biochemical Nephrology.W.G.Guder,editor. HansHuber, Bern,Switzerland. 447-454.

28.Chabardes, D., M. Gagnon-Brunette, M. Imbert-Teboul, 0. Gontcharev-skaia, M.Mont6gut,A.Clique,and F.Morel. 1980.Adenylate cyclase

responsive-ness tohormones in variousportionsof the humannephron. J. Clin. Invest.

65:439-448.

29. Haas, M. 1989. Propertiesanddiversityof(Na-K-Cl)cotransporters. Annu. Rev.Physiol. 51:443-457.

30. Molony, D. A., W. B. Reeves, and T. E. Andreoli. 1989. Na+:K+:2Cl-cotransport and the thickascendinglimb.KidneyInt. 36:418-426.

31.Wittner, M., A. Di Stefano, P. Wangemann, and R. Greger. 1991. How do loopdiureticsact? Drugs.41(Suppl.3):1-13.

32.Velizquez,H., D. W.Good,and F.S.Wright.1984. Mutualdependence of sodium and chloride absorption by renal distal tubule. Am. J. Physiol.

247:F904-F91 1.

33. Velazquez, H., and F. S. Wright. 1986. Effects of diureticdrugsonNa,Cl,

and K transportbyratrenal distal tubule. Am. J.Physiol.250:F1013-F1023. 34. Verkman, A. S., R. Takla, B.Sefton,C.Basbaum, and J. H. Widdicombe. 1989.Quantitativefluorescencemeasurementof chloridetransportmechanisms

inphospholipid vesicles. Biochemistry. 28:4240-4244.

35.Stokes,J. B. 1984.Sodiumchlorideabsorption bytheurinarybladder of thewinter flounder.Athiazide-sensitive, electricallyneutral system. J. Clin. In-vest. 74:7-16.

36.Kriz, W.,and L. Bankir. 1988. A standard nomenclature for the structures

ofthekidney.Am.J.Physiol. 254:F1-F8.

37.Madsen, K. M., and C. C. Tisher. 1986. Structure-functionalrelationships alongthedistalnephron.Am.J.Physiol.250:Fl-F15.

38. Morel, F., and D. Chabardes. 1985. Functionalsegmentation ofthe nephron.InTheKidney:PhysiologyandPathophysiology.D. W.SeldinandG. Giebisch,editors. RavenPress,NewYork. 519-580.

39.Smith,W.L., andA.Garcia-Perez. 1985.Immunodissection: useofmono-clonal antibodiestoisolatespecifictypesofrenal cells.Am.J.Physiol.248:F1-F7. 40.Allen,M.L.,A.Nakao,W. K.Sonnenburg,M.Burnatowska-Hledin,

W. S.Spielman, and W. L. Smith. 1988. Immunodissection ofcortical and medul-lary thickascendinglimb cellsfrom rabbitkidney.Am.J.Physiol. 255:F704-F710.

41.Friedman,P.A., B. A. Coutermarsh, J. S.Rhim,and F.A.Gesek. 1991.

Characterizationof immortalized mousedistalconvoluted tubule cells. J. Am. Soc.Nephrol.2:737.(Abstr.)

42. Morel, F. 1981. Sites of hormone action in the mammalian nephron. Am. J.Physiol. 240:F159-F164.

43.Chabardes, D., M. Imbert, A. Clique, M.Mont6gut,and F.Morel. 1975. PTHsensitiveadenyl cyclase activityindifferentsegmentsof therabbitnephron. Pfluegers Arch.Eur. J.Physiol. 324:229-239.

44. Morel,F., D.Chabardes,and M.Imbert. 1976. Functional segmentation

oftherabbitdistal tubulebymicrodeterminationofhormone-dependent adenyl-atecyclase activity.KidneyInt. 9:264-277.

45.Shimizu, T., M. Nakamura, K. Yoshitomi, and M. Imai. 1991.

Interac-tionof trichlormethiazideoramiloridewith PTH in stimulating calcium

absorp-tionin therabbit connectingtubule. Am. J.Physiol.261:F36-F43.

46.Shimizu,T., K.Yoshitomi,M.Nakamura, and M.Imai. 1988. Site and mechanismofaction of trichlormethiazide in rabbit distal nephron segments

perfusedinvitro.J.Clin.Invest.82:721-730.

47. Walser, M. 1971. Calcium-sodium interdependence in renal transport. In RenalPharmacology. J. W. Fisher, editor. Appleton-Century-Crofts, New York. 2141.

48.Costanzo,L.S., and E. E.Windhager. 1986. Transport functions of the

distalconvoluted tubule. InPhysiology ofMembraneDisorders.T. E.Andreoli,

J. F.Hoffman,D. D.Fanestil, and S. G. Schultz, editors. Plenum Medical Book

Company,New York.727-750.

49.Gesek, F. A., B. A. Coutermarsh, and P. A. Friedman. 1991. Thiazide and

amiloride-sensitive Na'uptakemechanisms in distal convoluted tubule cells. J. Am.Soc.Nephrol.2:737. (Abstr.)

50.Friedman, P. A. 1988. Biochemistry and pharmacology of diuretics. Se-min.Nephrol 8:198-212.

51. Tran, J. M., M. A. Farrell, and D. D. Fanestil. 1990. Effect of ions on

bindingof thethiazide-type diuretic metolazone to kidney membrane. Am. J.

Physiol.258:F908-F915.

52.Velfizquez, H., and R. Greger. 1985. K &Clpermeabilities in cells of rabbit early distal convoluted tubule.KidneyInt. 27:322. (Abstr.)

53.Salomonsson,M., E.Gonzalez, P. Westerlund, and A. E. G. Persson. 1991. Chloride concentration in macula densa and cortical thick ascending limb cells.KidneyInt. 39(Suppl. 32):S5I-S54.

54.Greger, R., H. Oberleithner, E. Schlatter, A. C. Cassola, and C. Weidtke. 1983. Chloride activity in cells of isolated perfused cortical thick ascending limbs

of rabbit kidney. Pfluegers Arch.Eur.J.Physiol.399:29-34.

55.Yoshitomi,K., T.Shimizu, J. Taniguchi, and M. Imai. 1989.

Electrophysi-ologicalcharacterization of rabbit distal convoluted tubule cell.PfluegersArch. Eur.J.Physiol.414:457463.

56. Matsunaga, H., P. A.Friedman, N. Kizer, and B. A. Stanton. 1991. Cal-cium channels inapicalmembranesof distal convoluted tubule cells. J. Am. Soc.

Nephrol.2:627. (Abstr.)

References

Related documents

Clustering of catchments by their response behaviour as well as by their physical properties leads to five clusters, each rep- resenting different facets of flow regimes or types

Long-run growth patterns within Asian NIEs: Empirical analysis based on the panel unit root test, allowing the heterogeneity of time trend and endogenous.. multiple

Abstract: This paper uses affiliate level data from Swedish multinationals to examine the impact of tax treaties on both overall affiliate sales and the composition of those

In this paper novel approach is based on hybrid support vector machine for content-based image retrieval by using modified SVM. It supports more accurate

DNVR has a number of miscellaneous features that makes it diverse from other reactive routing protocols like depletion of probes for proficient recognizing of network,

Almohammad [17] proposed the performance of both gray scale and color versions of a given cover image when they are used with a given steganography method, the

This study evaluates the growth performance of Arthrospira platensis in deep ocean water (DOW), surface ocean water (SOW), and artificial ocean water (AOW) with and without

The aim of this paper is double first, Hyper Quad Trees algorithm is applied for finding the cluster centers which will be an input to the K-Means