Na/H antiporter mRNA expression in single nephron segments of rat kidney cortex

(1)

Amendment history:

Correction

(December 1991)

Na/H antiporter mRNA expression in single

nephron segments of rat kidney cortex.

R Krapf, M Solioz

J Clin Invest.

1991;

88(3)

:783-788.

https://doi.org/10.1172/JCI115377

.

Renal cortical tubules consist of polarized epithelial cells where Na/H antiport activity has

been demonstrated on the apical and/or basolateral membrane. Apical Na/H antiport activity

plays an important role in transcellular bicarbonate (HCO3-) reabsorption, whereas

basolateral Na/H antiport activity could be involved in transcellular HCO3- secretion as well

as cell volume and pH control. To determine whether this heterogeneity in both localization

and function is due to the existence of more than one Na/H antiporter, we studied the tissue

distribution of Na/H antiporter mRNA by use of reverse transcription (RT) and polymerase

chain reaction (PCR) in isolated nephron segments from rat renal cortex. The primers used

were directed against the rat renal cortical Na/H antiporter cDNA which is homologous to

the human growth factor-activatable Na/H antiporter. RT/PCR of beta-actin mRNA were

performed as positive controls. Na/H antiporter mRNA expression in the proximal tubule

was not detectable in S1 and S2 segments from superficial and most midcortical nephrons,

which exhibit exclusively luminal Na/H antiport activity. It was expressed in S1 and S2

segments from juxtamedullary nephrons which have also basolateral Na/H antiport activity.

Beta-actin mRNA was expressed uniformly in all segments […]

Research Article

(2)

Na/H Antiporter mRNA Expression

in

Single Nephron Segments of Rat Kidney Cortex

RetoKrapf and MarcSolioz

With the technicalassistanceofCatherine Fehimann

DepartmentofMedicineand InstituteofClinicalPharmacology,InselUniversity Hospital, CH-3010Berne,Switzerland

Abstract

Renalcortical tubules consist of polarized epithelial cells where

Na/H antiport activity has been demonstrated on the apical and/or basolateral membrane. Apical Na/H antiport activity

plays animportant role in transcellularbicarbonate (HCO3)

reabsorption, whereas basolateralNa/Hantiportactivitycould beinvolvedin transcellularHCO3- secretionaswellascell vol-ume andpH control.

Todetermine whether this heterogeneity in both localiza-tionandfunction isduetotheexistenceofmorethanoneNa/H antiporter, we studied the tissue distribution of

Na/H

anti-porter mRNA by use of reversetranscription (RT)and

polymer-asechain reaction(PCR)inisolated nephronsegmentsfromrat renal cortex. The primers used weredirected against the rat renal corticalNa/H antiportercDNAwhich is homologous to the human growth factor-activatable Na/H antiporter. RT/ PCRof beta-actin mRNA wereperformedaspositivecontrols.

Na/H antiportermRNAexpressionintheproximal tubule

was not detectable in S1 and S2 segmentsfrom superficialand most midcortical nephrons, which exhibitexclusively luminal

Na/H antiport activity.It wasexpressed in

Si

andS2segments from juxtamedullary nephrons which have also basolateral Na/ Hantiport activity. Beta-actin mRNA was expresseduniformly in all segments of theproximal tubule.Na/Hantiporter mRNA was also expressed in cortical thickascending limb and cortical

collecting duct, segments with basolateral Na/H antiport

activity as well as in the glomeruli.

Inconclusion, at least twodifferent Na/Hantiportersexist

in the renal cortex,i.e., the proximal tubule. The close correla-tion betweenfunctional localizationofbasolateralNa/H

anti-portactivityand mRNAexpression suggests that the rat kid-neyNa/Hantiporter DNA homologous to the human growth

factor activatable Na/H antiporter encodes abasolateral

ex-changer.The observedexpressionin aminority of midcortical proximal tubules couldreflecta certainheterogeneity in these nephron segments. (J. Clin. Invest. 1991. 88:783-788.) Key words: Na/H antiporter * mRNA expression *nephron seg-ments * reverse transcription * polymerase chainreaction

Introduction

Inthepolarizedrenal cortical tubule cells,electroneutral,

amil-oride-sensitiveNa/Hantiport activityhas been demonstrated oneither the luminal (apical) and/or the basolateral side.When

located on the apicalmembrane, the Na/H antiport activity

Receivedforpublication 25 March 1991.

effects proton secretion into the luminal fluid. The cellular bi-carbonate ionsthus generatedaretransportedinto the extracel-lular fluid across an electrogenic

Na/CO32-/HCO-3

symporter.

This polarized,inseries location of a luminalNa/Hantiporter and a basolateral

Na/CO32/HCO3

symporter allows efficient transcellular bicarbonatereabsorption in the proximal tubule and the cortical thick ascending limb (1-3). In the proximal

tubule, theluminalNa/Hantiportermediates as much as two-thirds ofproximaltubule bicarbonatereabsorption (4).

Thephysiologicalrole ofNa/H antiport activitylocated on the basolateral membrane is less clear. In tubule segments with thecapacityto secretebicarbonateinto the luminal fluid such asthecorticalcollecting duct,bicarbonate exit across the apical membranewouldacidify the cells.This acid load would stimu-late abasolateral Na/Hantiporterto secrete protons intothe extracellularfluid,thuscompletingthe process of transcellular

bicarbonate secretion (5). In tubule segments not known to

secretebicarbonate(e.g.,cortical thickascending limb,SI and

S2segmentsofjuxtamedullary nephrons,seebelow),a basolat-eralNa/H

antiporter

could assume an

important

rolein intrin-siccell

regulation

such as cellpH defenseandcellvolume con-trol as well as beinvolved insignaltransduction(growthfactor

activation).

Inthe mammalian renalcortex, anexclusive apical

loca-tionfor theNa/Hantiporterhas beendemonstrated

function-ally forthe

proximal

convoluted tubuleof

superficial

and mid-cortical nephrons (S1 andS2segments;6-9).Thecortical thick

ascending

limbalso has aluminal Na/H antiporter (2), but,in

addition, shows basolateral Na/Hantiport activity (3). Na/H

antiport

activity restricted to the basolateral membrane was

found in the cortical

collecting

duct(5).Inthe

proximal

tubule,

theexistenceanddistribution ofabasolateral Na/H antiporter is controversial. While Geibel and co-workers described

re-cently that rabbit S1 and S2 segments from juxtamedullary nephrons exhibit basolateral Na/H antiport activity (9), no suchevidencewasfound byBaum(10).Inothercorticaltubule segments(i.e., thedistalconvoluted and connecting tubules)

no direct information about presence and location ofNa/H

antiport

activity is available. Na/H antiport activity has been demonstrated, however, in the nonpolarized mesangial cells of glomeruli (1 1).

Itisnotknown, whether the heterogeneity in distribution

(locationinS1 andS2proximaltubule segmentsfrom superfi-cialandmidcorticalversus

juxtamedullary

nephrons),

polarity

(locationonapicalversusbasolateralmembrane), andfunction

(bicarbonate

transport versus cell pH and volumecontrol)of

Na/Hantiport activityin themammalianrenal cortexisdue to

differential

epithelial sorting

and regulation ofone and the same

antiporter

or to the existence of more than one

anti-porter.

Evidence from

LLC-PKl

cells,immortalizedcellsofrenal tubule

origin,

which exhibit apical and basolateralNa/H

anti-J.Clin. Invest.

(3)

portactivity, raises the possibility of the existence of two differ-ent antiporters. The apical antiporter exhibits lowamiloride sensitivity

_(IC50

for the amiloride analogue ethylisopropyl-amiloride, - 10 ,uM), whereas the basolateral transporter is

verysensitive toethylisopropylamiloride inhibition

_(IC5o

- 20

nM; 12). Thus, based on these inhibition kinetics, the apical form resembles the proximal tubule brush border Na/H

anti-porter, whereas the basolateral transporter behaves like the ubiquitous transporter found in many epithelial and nonepith-elial cells(13). Itwaspostulated, therefore, that LLC-PKl cells

contain two differentNa/H antiporters: a basolateral, "house-keeping" Na/H antiporter involved in cell pH defense and

vol-ume regulation, and an apical Na/H antiporter involved in transcellular proton/bicarbonate and possibly sodium

trans-port. In addition, recent preliminary evidence (14) suggests thatapical and basolateral antiporters in these cells are

regu-lateddifferentlybyCa++/calmodulin-dependent processes.

Im-munostaining withpolyclonal antibodies against amino acid

sequencesofacloned LLC-PKl Na/H

antiporter

with 95% homology to the human growth factor activatable Na/H anti-porter, revealed exclusive localizationofthis transporter on the basolateral membrane in LLC-PKl cells (15). In addition, a human Na/Hantiporter probe was noticed to localize a basolat-eral Na/H antiporter in the polarized intestinal cells, Caco2 and HT29(Huet, C., D. Louvard, C. Sardet, and J. Pouyssegur,

personal communication). However,inthebovinekidney,

im-munostaining of thecortex waspredominantly localized

api-cally,inthebrushborder membraneofproximal tubules(16). Theseimmunohistochemical observations raisethe important

possibility that the expression of two different Na/H anti-porters could be restricted tomutated,immortalized cell lines. Wehave recently cloned and sequenced apartial-lengthrat renal Na/HantiportercDNA(17)which exhibits - 95%

ho-mology to the human, growthfactor-activatable Na/H anti-porter( 18)and have shown that mRNAexpressionofthis trans-porter is enhanced in response to metabolic acidosisin vivo

( 19).Ifthereare twodifferentantiportersinthe renal cortex, our cDNA would beexpected to encode for the basolateral

antiporterbased onthe observations from immunohistochemi-calexperiments. Thus,inthe case of two different antiporters, mRNAcorrespondingto ourcDNAshouldnotbeexpressedin S1andS2proximaltubule segments ofsuperficialand

midcor-tical nephrons. ThismRNAshould beexpressed, however,in

thickascending limbs, cortical collecting ducts, possiblyS1 and

S2

proximal

tubule segments from

juxtamedullary

nephrons

andglomeruli.

Thepresentstudiesweretherefore designedtoinvestigate

thesegmental localization ofthe mRNAexpressionofour re-centlycloned ratrenalNa/H

antiporter

inratrenalcortex.We chosethetechnique ofreversetranscription (RT)' of specific mRNAinmicrodissected, isolated nephronsegmentswith sub-sequent amplification of the cDNAbythepolymerasechain

reaction (PCR).

Methods

Allchemicals and materialswere RNAse-free,of molecularbiology

grade and obtained fromSigmaChemicalCo.,St.Louis, MO,unless statedotherwise.

1.Abbreviations usedinthis paper:PCR, polymerasechainreaction;

PCT,proximalconvolutedtubule;RT,reverse

_{transcription.}

Experimental animals and microdissection. Male Wistar rats

weighing80-140 g were obtained from the Animal Breeding Facility at the InselUniversity Hospital,Berne,Switzerland. The animals had free

accessto waterand commercial rat food.

Ratswereanesthetized by intraperitoneal injection ofpentobarbital andweresacrificedby decapitation. The aorta was cannulated with

polyethylenetubing(PE 50) and the left kidney was selectively perfused with 10 ml ofice-colddissection solution containing 1.5 mg/ml of

collagenase(type1).Thedissection solution was bicarbonate-free and contained(inmillimolar):NaCl135,Na2HPO4 1, MgSO4 1.5, CaCl2 2, KCI 5,glucose5, Hepes 5,titrated to pH 7.4 with NaOH (room

temper-ature).Afterperfusion,thekidney was excised and the renal capsule

wasstrippedoff. The kidney was then cut into coronal slices of - I

mm.Theslices were transferred to individual tubes containing 1 ml of thesamecollagenasesolution as used to perfuse the kidney and were

incubatedfor 30-45 min at370Cin a water bath. Before dissection, the sliceswerewashedoncewithice-cold collagenase-free dissection solu-tion.

Tubule dissectionwasperformed in ice-cold dissection solution

(see above) using dissectingforceps (Dumont No. 5) and a dissecting

microscopeat40x magnification and horizontal illumination. Tubule segmentswereidentifiedaccording to established criteria (20). SI and S2 proximalconvoluted tubule segments (PCT) were dissected and studiedseparately accordingtotheir origin from superficial,

juxtame-dullary, or midcortical nephrons.2 For the dissection of superficial

nephrons, glomeruli highin the cortex (within 1 mm of the renal

cap-sule)werelocated. For the dissection ofjuxtamedullary nephrons, glo-meruli located< 1mmfrom the outer edge of the medulla were

cho-sen.Midcortical nephronswerefrom glomeruli located at least 1 mm away from the renalcapsule and at least 1 mm from the outer edge of the medulla. SIPCT segmentswereidentified by their attachment to a

glomerulus.PCT segments extending beyond 1 mm of the glomerulus

wereconsideredtobe S2.Cortical thick ascending limbs and cortical

collectingductsweredissected from the medullary rays. Distal convo-luted tubules and glomeruli were dissected from the cortical labyrinth. All attachedtissue (blood vessels, tubule segments) were removed from

glomeruli beforetransfer.

Thelengthof the tubuleswasbetween 0.7 and 1 mm. Using a bent Pasteurpipette coated with a0.1% BSAsolution, the tubuleswere

transferred from the dissection dishtofresh dissection solutiontorinse the segments free of debris. The tubuleswerethencapturedwith small

glassbeads(diameter, 425-600

Aim)

and transferredasbefore(new

pipette) tothe RT-PCR tube. The beads/tubuleswere rinsed three times with ice-cold dissection solution andfinally suspendedin 20

JAI

of dissection solutioncontaining> 1U/Alof humanplacentalRNAase inhibitor(PromegaBiotec, Madison, WI).Glomeruli,afterrinsingina

washdish,weretransferred withoutglassbeads and washed like the tubule segments. The reaction tubeswere_placedonice until RTwas

begun(maximum60min).

Reversetranscription(RT). Reaction tubeswerecentrifugedfora

few seconds(12,000rpm)topelletthe

beads/tubules

andglomeruli. Thedissection solutionwascarefullyremoved and 10

/sl

of

lysis

buffer containing2%TritonX-100

(Boehringer-Mannheim

GmbH,

Mann-heim, FRG),> 1.5U/,lof RNAse inhibitor and 5mMdithiothreitol

wereadded. Thecontentsweregentlymixedby

tapping

and thereverse

transcriptasereagentswereadded

(final amounts/concentrations;

reac-tionvolume,20

Al):

antisense

primer,

40pmol; deoxynucleotides,200

Amol; Tris-HCI,10mM,pH 8.3; KC1, 50 mM;MgCI2,2.5mM;DTT,

1mM;BSA,1mg/ml;reverse

transcriptase,

2.5

UIAI (Moloney

murine leukemiavirus;NewEngland

Biolabs,

Beverly,

MA).

Reaction tubes

wereincubatedat37°Cfor 60 min.Negativecontrol reactions

contain-2.S3proximaltubuleswerealsostudied in

preliminary experiments.

However,the resultsarenotincluded here because the cortical

portion

of S3 segments is ill defined. Infouroutof four tested S3segments,

(4)

ing all the reagents save the reverse transcriptasewereperformed in parallelfor each cortical tubule segment andglomeruli.

Polymerase chain reaction (PCR).For theNa/Hantiportergene, primer 1 (antisense) was 5'-ATCTGGTTCCAGGCITCCTCG-TAGG-3' defined by bases 1997-1973 andprimer2(sense) was 5'-CAAGAGACGAAGCGCTCCATCAACG-3' definedbybases 1534-1558 (baseenumeration according to full-length human antiporter

cDNA; 19, 20). Thus, the cDNAamplification productwaspredicted to be 464bp in length. Theamplified regionof theratrenalNa/H

antiporter cDNA has 98% homologyto thegrowthfactor-activated humanNa/Hantiporter (17, 18). While theintron/exon boundariesof the ratNa/Hantiporter geneare notyetdefined,thehigh homology

between human andratcDNA suggestsinterspecies conservationof intron/exonboundaries. Based on thiscomparison,theprimersused amplify a cDNA segmentencompassingthreeintrons(21).An addi-tional oligonucleotide (sense) 5'-GGTTGAGCTTGTCCTTCCAG-3'

(defined by bases 1636-1655)was usedas anamplification specific

probe. RT and PCRof ratcytoplasmicbeta-actin servedas apositive

control. The primerswere definedbythefollowingcDNA base

se-quences (22): primer 1 (antisense), bases 3078-3053, sequence,

5'-ACCTTCAACACCCCAGCCATGTACG-3'; primer2(sense), bases 2168-2193, sequence,

5'-CTGATCCACATCTGCTGGAAGGTGG-3'. Theseprimersspantwointrons and result ina703-bpcDNA ampli-fication product(22).

Alltheprimerswerecheckedfor theabsenceof fortuitous homol-ogy toany sequence in theEMBLdata bank.

After RT, 30 Ml of PCR reagent mixwereaddedtoeachtube di-rectly.The PCRreactions (50ul)contained thefollowingcomponents (final amounts/concentrations): Taq polymerase 2 U

(Boehringer-Mannheim GmbH),primers(40 pmoleach),deoxynucleotides (200

MM),Tris-HCI(10 mM), pH8.3(roomtemperature),KCI(50mM),

MgCl2(2mM),gelatine (1 mg/ml), Triton X-100(1%).Thereaction tubes wereoverlayered with 50,lofmineral oiltopreventevaporation.

PCR wasperformedinanautomated thermalcycler(EasyCycler; Eri-comp, SanDiego,CA) whichwasprogrammedasfollows.First, incu-bation at92°C for 2 min(initial melt), followedby 30 PCRcycles:

RTenzyme - + + - + + + + _ +

92°C for 10 (denature), 55°C for 60 (anneal), and 72°C for 100 s (ex-tend). PCR was completed by a final extension (72°C, 10 min).

RTof Na/Hantiporter and actin RNA from different tissues with subsequentamplification by PCR was also performed. TotalRNAwas isolated from rat kidney cortex, liver, skeletal muscle, heart, lungs, and spleen usingguanidiniumisothiocyanate (4 M) and cesium chloride gradient centrifugation (19). For RT, 100 ng total RNA were used. Reversetranscriptase concentration was 2 U/tube. All other reagents for RT and PCR were as described above.

Foranalysis of the PCR products, 40 Ml of reaction volume were ethanol-precipitated. DNA was resuspended, and the PCR product size-fractionated on 2% agarose gels and stained with ethidium bro-mide (23).

For Southern blot analysisof PCR products, gels were denatured (NaOH),electrotransferred, and blotted onto Zeta-probe Nylon mem-branes(Bio-Rad Laboratories, Inc.,Richmond,CA) as recommended by the supplier. To test for rat renal Na/H antiporter specific PCR amplification products, 1 pmol of the amplification product specific primer (20 mer oligonucleotide) was labeled with 32P at the 5' end using T4polynucleotide kinase (24) and used as a hybridization probe.

Theactin-specificamplification product was analyzed on ethidium bromide stained agarose gels. No attempt was made to quantitate

RT/PCR.

Results

The presence of Na/H antiporter mRNA was first analyzed by RT/PCR in single Sl and S2 proximal convoluted tubules from juxtamedullary nephrons, acortical collecting duct

seg-mentand sixglomeruli. For all thesesegments,there is func-tional evidence for thepresenceofaNa/H antiporter. Fig. 1

shows the 2% agarose gel (left) and the corresponding Southern blot analysis (right) of the amplification products.An amplifi-cation product ofthe predicted sizewasevidentintheRT/PCR reactions from all tubule segments tested. The Southern blot of

basepairs

703

464

_I)_ NlUN CN

uw u Uw u z wn a

c) ^4.

T (

U Z U)

.,

en U) U)n ( co cn O

464 bp

z

a

z

aON.

U 'm U Z

Figure 1. Product analysis of RT/PCR for Na/H antiporter mRNA in the proximal tubule ofjuxtamedullary nephrons

(SI

and S2segments),

glomeruli (G,n =_{6) and}_{cortical collecting duct (CCD).}

_(Left)

_{an ethidium bromide stained 2% agarose gel; (right) an autoradiogram of the}

correspondingSouthern blot. Results are shown fromexperimentsperformed in the presence (RT-enzyme +) and absence (RT-enzyme -, neg-ative controls) of reverse transcriptase. The amplification product of PCR of Na/H antiporter DNA and the RT/PCR product of beta-actin

(5)

RT enzyme - + + + +

-1353 1078 872

603 464

: z a

U. OUO

Figure 2.Product analysis forRT/PCR for Na/H antiportermRNAexpression in cortical thick ascending limb (cTAL), cortical collecting duct (CCD), glomeruli(G,n=8) and distalconvoluted tubule(DCT). (Left)Anethidium bromide stained2%agarosegel; (right) autoradiogramof

the corresponding Southern blot.Resultsareshown fromexperimentsinthepresence(RT-enzyme +)and absence(RT-enzyme-,negative

control)ofreversetranscriptase.

thegel demonstrates specificbinding of the independent oligo-nucleotidetothe product, thus confirming its identity.

Nega-tive control reactions using eitheran S1 or an S2 proximal

convolutedtubule from ajuxtamedullary nephron in the ab-senceofreversetranscriptase but with all other RT/PCR

re-agentsarealso shown. Nospecific bandwasdetectedoneither

theagarose gelor the Southern blot. Na/H antiporter DNA

(- 0.5 ng)wasamplified inapositive control reaction for the

PCR and yielded an identical amplification product as RT/

PCR from the nephronsegments.Theamplification product of RT/PCR of beta-actin mRNA inaSI proximal tubule froma

juxtamedullary nephron convoluted tubule is shown for

com-parisonontheagarosegel and servedasanadditional positive

control for theRT/PCR reactionsequence.

Fig. 2 extends the analysistothe cortical thick ascending limb ofHenle's loop and thedistal convoluted tubule. Results from additional glomeruli (eight glomeruli in one RT/PCR

reaction) and acortical collecting ductarealso shown. Both

apical and luminal Na/H antiporters have been found func-tionallyinthecortical thickascending limb, whereas there isno

direct functional evidence forthe presence ofa Na/H

anti-porter in the distal convoluted tubule. Again, amplification products of the predicted lengthareshownontheagarosegel

(Fig. 2, left). Their identitywas confirmedby highly specific binding of the independent oligonucleotideon aSouthern blot (Fig. 2, right). The negative control reaction using acortical

thickascending limb didnotshowanybands. Theidentity and

specificity of the RT/PCR product is further illustrated by the positive control reactionsonFigs. 1 and2.

Thenextexperimentsweredesignedtospecifically analyze the distribution ofNa/H antiporter mRNA in the six

subseg-mentsof the proximal tubule (SI andS2segmentsfrom

super-ficial, midcortical, and juxtamedullary nephrons, respectively). Based on functional studies, proximal tubulesare heteroge-neouswithrespect tothe distribution of Na/H antiport activ-ity. S1 and S2 segments of superficial and midcortical

nephrons have apical, butnobasolateral Na/H antiport

activ-ity (6-9), whereas SI and S2 segments from juxtamedullary nephrons have been showntohave apical and basolateral

Na/

Hantiport activity, although thepresenceof basolateralNa/H activity is controversial (9, 10). Fig. 3 illustrates that Na/H antiportermRNAwasabsent in SI and S2segmentsfrom

su-perficial and midcortical nephrons,whereas itwasexpressedin

SIand S2segments fromjuxtamedullary nephrons.These

re-sults thusprovidestrong evidence for thepresenceofatleast two differentantiportersinthe proximaltubule. Ourresults

and observationsinstudies about the functional distributionof

Na/H antiportersuggest that the mRNA detectedbytheseRT/ PCR reactions encodesthebasolateralNa/H antiporter.

TableIsummarizesthe results from all tubulestested for

Na/H antiporter and beta-actin mRNAexpression. None of

the SI and S2 proximal tubule segments from superficial nephrons didexpressNa/H antiporter mRNA,whereasactin

mRNA expression was uniform. In contrast, all SI and S2

proximal tubule segmentsfromjuxtamedullary nephronswere

positiveforNa/H antiportermRNA. The table further

illus-trates thataclearmajorityofS1andS2segmentsfrom midcor-ticalnephronsdidnotexpressNa/H antiportermRNA.

How-e w.

464 bp

z 0

u u

a

u C5

C) c

0 Z

(6)

RTenzyme . + + + + + + +

-2176 1229 1033

464 234

z

U) -2 U) 2 -I ₂ ₂ I.

.

C4C4 -

--_

(n cO cn Vn cn C U) Z

(0)

C')-,

N 04~

(fl (s

* * 464bp

z

(flu 2 U₂

U) 2 .) 2 2 -. {

in n N, N N I

W

() _(0(cf)0 _Z

Figure3. Productanalysisof RT/PCR for Na/HantiportermRNAexpressioninS1 andS2proximaltubule segments ofsuperficial(SS), mid-cortical (MC), andjuxtamedullary (JM) nephrons.(Left)Anethidium bromide stained 2% agarosegel;(right)anautoradiogramof the

corre-spondingSouthern blot. Resultsareshownfromexperimentsperformedin the presence(RT-enzyme +)and absence(RT-enzyme-,negative

control) ofreversetranscriptase.PCRofNa/HantiporterDNA servedasapositivecontrol.

ever, a

minority

ofsegments(oneoutof10SI segmentsand

two outof12S2 segments)expressed Na/HantiportermRNA. Afalse positiveresultinthese tubulesisexcludedbecausewe

performed negative

controls in all experiments (see Methods, Figs. 1-3). Thus, if the dissection criteria used (see Methods) reliably identify midcortical nephrons,thereisapparent

hetero-geneity

in these midcortical SI andS2 segments withrespect to

Na/Hantiporter mRNA expression.

Inadditional experiments,wewishedtodetermine whether theRT/PCR reaction detected Na/H

antiporter

mRNAin sev-eral organs. If the Na/H antiporter mRNA detected by our

RT/PCR encodedforthe"housekeeping"Na/Hantiporter in-volvedprimarily incell pHand volumecontrol, thismRNA

wouldbe expected to beexpressedubiquitously in extrarenal tissues. RT/PCRwasperformedon 100 ngof totalRNA iso-latedfrom liver, skeletalmuscle,lung,heart,andspleen.

Am-plification products ofthe predicted sizes were detected for Na/H

antiporter

mRNA and cytoplasmic beta-actin in all these

organs(datanot

shown).

TableLNa/HAntiporterandBeta-Actin mRNAExpression

inRatRenal Cortex

Na/Hantiporter Beta-actin

(positive/total) (positive/total)

Proximal tubule

SI superficial 0/6 6/6

S2superficial 0/6 6/6

S1midcortical 1/10 8/8

S2midcortical 2/11 8/8

S1juxtamedullary 6/6 6/6

S2juxtamedullary 6/6 6/6

Corticalthick ascending limb 7/7 7/7 Distal convoluted tubule 5/5 5/5 Corticalcollectingduct 8/8 8/8

Glomeruli 6/6 6/6

Discussion

These studiesinvestigatedthe tissue distribution of mRNA ex-pression ofa ratNa/Hantiporter homologoustothe human growthfactor-activatableantiporter. The mainfindingswere: (a)thisNa/H antiportermRNA isexpressedinglomeruliand all the cortical tubule segments that show basolateral

Na/H

antiport activity (i.e., S1 and S2 segments of

juxtamedullary

nephrons, cortical thick ascending limbs, cortical collecting duct); (b)thisantiportermRNA is not expressed in tubule seg-ments with functional evidence only for apical, but not for basolateral Na/H antiport activity (i.e., S1 and S2 proximal

tubule segments of superficial and the majority ofmidcortical nephrons);and(c)the antiporter mRNA expressed in segments with basolateral Na/H antiporter activity isindistinguishable

fromanantiportermRNAexpressedinavarietyof

nonepithe-lial, nonbicarbonate transporting tissues (i.e., spleen, heart,

lung, skeletal muscle).

Theresults ofourstudiestherefore providestrong evidence forthe existenceof atleasttwodifferentNa/Hantiporter pro-teins within the kidney cortex, i.e., the proximal tubule.Several

observations from our study deserve special emphasis. First, in

S1 andS2proximal tubules fromsuperficialand the majority of midcortical nephrons, there was noexpression of mRNA corresponding to the Na/H antiporter cDNA with 95%

homol-ogy tothe human growth factor-activatable Na/Hantiporter.

Inthesesegments,thereisunequivocal evidencefor a luminal Na/H antiporterthat has a quantitatively important role in

proximaltubule NaCl andbicarbonate reabsorption,whereas a basolateral Na/H antiporter has not been detected.Thus,this

proximal tubuleluminalNa/Hantiporter isadifferent trans-porter than the basolateral Na/Hantiporterand does not corre-spond to the mRNA detected by our method. It remainstobe established whether this luminal Na/H antiporter is the

prod-uct of adifferent gene or of an isoform of thesamegene. Our

results exclude posttranslational modification as the mecha-nism for theproductionof adistinct luminalNa/Hantiporter. basepairs

AD

(7)

Differential RNA splicing remains a possibility because we

transcribedandamplifiedonly a 464-bpfragment ofthe Na/H

antiportermRNA.

Thedetection ofNa/H antiporter mRNAexpressionin a

minority of

Sl

andS2 proximal tubulesegmentsfrom midcor-tical nephrons is

interesting.

Itis possible that basolateralNa/H

antiport activity could have been overlooked in functional

stud-iesbecause mRNAexpressionwas found in only- 10%ofall

tubules.Inaddition, Na/H antiport activitymay be very lowin thesesegmentsmaking functional detection difficult. Further

studiesareneededtogainmoreinsight intotheprecise distri-bution,levelof activity,and

physiological

roleofabasolateral Na/H antiporter in the proximal tubule of midcortical nephrons.

Second, it is striking thatmRNAexpression of thisNa/H

antiporter closely parallels the functional evidence forthe

tis-suedistribution ofabasolateralNa/Hantiporter.Inaddition,

anidenticalmRNAcould bereversely transcribedand ampli-fied in several nonepithelial tissues (heart and skeletalmuscles,

spleen, lung) and glomeruli. In these tissues, Na/H antiport

maybe predominantly involved in cell pH defense and cell volume regulation as well asin signal transduction (growth

factor activation). Thus, the mRNA detected andamplified withour primersisprobably the gene product of renal Na/H

antiporter responsible for these "housekeeping" functions within the kidney.Itispresentlyunclearwhetherhousekeeping functions are effectedby one or more antiporters. The gene

expression

of theantiporter detected byourmethodhasbeen showntobeenhanced inresponse toinvivometabolic, butnot

respiratory, acidosis (19). Itremains to be seen whetherthe luminalNa/Hantiporter is regulated similarly.

Third,ourresults predict the existence ofa Na/Hantiporter in cellsfrom the distal convoluted tubule.Inthecortical thick ascending limb,ourresults donotpermitus todecide whether theNa/H antiportermRNAstudied istranslatedinto the

api-cal orbasolateralNa/H antiporter.However, becausethe api-cal Na/H

antiporter

is involved in transcellular bicarbonate

transport, it ismore likely that the Na/H antiportermRNA

studiedhereis translated intothebasolateral

protein.

Inconclusion, ourstudy showsthat mRNAsfromatleast

two

different

Na/H

antiporters

are

expressed

in theratrenal

cortex. A DNA/RNA with 95% homology to the human growth factor activatable Na/H

antiporter

encodes fora

basolat-eralNa/H

antiporter

thatis

probably

involved in

"housekeep-ing" functions suchascellpH

defense, signal transduction,

and volumeregulation.

Acknowledgments

Theauthors thankArlynGarcia-Perez,National Institutes ofHealth,

Bethesda, MD, for her help inadaptingthetechniqueofRT/PCRto

isolatednephronsegments.

This workwassupportedby Swiss National Foundationgrants 31-25370.88to R.Krapfand 31-28577.90toM.Solioz.

References

1.Alpem,R.J.,and M.Chambers. 1986. CellpHin the ratproximal

convo-luted tubule. J.Clin.Invest.78:502-5 10.

2.Good,D. W.1985.Sodium-dependentbicarbonateabsorption bycortical thickascendinglimbofratkidney.Am.J.Physiol.248(Renal Electrolyte

Phys-iol.17):F821-F829.

3.Krapf,R.1988. Basolateral membraneH'/OH-/HCO-transport in therat

corticalthickascendinglimb.J.Clin.Invest.82:234-241.

4.Preisig,P.A.,H. E.Ives,E. J.Cragoe,R. J.Alpern,and F. C.Rector,Jr. 1987. RoleoftheNa+/H+antiporterintheratproximaltubulebicarbonate absorption.J.Clin.Invest.80:970-978.

5.Chaillet,J.R.,A.G. Lopes,and W. F. Boron. 1985.Basolateral Na-H

exchangein therabbitcorticalcollectingtubule.J.Gen.Physiol.86:795-812. 6. Ives,H.E.,J. V.Yee,and D. G.Warnock. 1983.Asymmetric distribution

of theNa+/H+ antiporterin the renalproximaltubuleepithelialcell. J. Biol. Chem. 258:13513-13516.

7. Sabolic,I.,and G.Burckhardt. 1983. Protonpathwaysin theratrenal brush-border and basolateral membranes.Biochim.Biophys.Acta.734:210-220. 8.Krapf, R.,R.J.Alpern,F.C.Rector, Jr.,andC.A.Berry.1987. Basolateral membraneNa/base cotransport isdependentonCOSHCO- intheproximal

convoluted tubule. J. Gen.Physiol.90:833-853.

9.Geibel, J.,G.Giebisch,and W. F. Boron. 1989. Basolateral

sodium-cou-pled acid-basetransport mechanismsoftherabbitproximaltubule.Am.J.

Phys-iol. 257(Renal Fluid Electrolyte PhysPhys-iol.26):F790-F797.

10.Baum, M. 1990. Neonatal rabbitjuxtamedullary proximalconvoluted tubuleacidification.J.Clin.Invest. 85:499-506.

1 1.Boyarski, G.,M.B.Ganz,E. J.Cragoe,and W. F. Boron. 1990. Intracellu-larpH dependenceof Na-Hexchangeandacid-loadinginquiescentandarginine

vasopressin-activated mesangialcells. Proc.Nati.Acad.Sci. USA.87:5921-5924.

12. Haggerty, J. G.,N. Agarwal,R.F. Reilly,E.A.Adelberg,and C.W.

Slayman. 1988.PharmacologicallydifferentNa/Hantiportersontheapicaland basolateral surfacesofculturedporcine kidneycells(LLC-PKI).Proc.Nati.Acad.

Sci. USA.85:6797-6801.

13.Friedrich, T.,J.Sablotni,andG. Burckhardt. 1987.Speciesdifferences betweenratandrabbitrenalNa+/H+ exchangers. Biochem.Biophys.Res.

Com-mun. 144:869-875.

14.Bums,K.D.,T.Homma,andR. C. Harris. 1990. Calmodulin antago-nisminhibits basolateralandactivatesapical Na+/H+ exchangeinLLC-PK,/CL4.

JASN. 1:714.(Abstr.)

15.Reilly,R.F.,D.Biemesderfer,F.Hildebrandt,M.Exner,C.Sardet,J.

Pouyssegur,C.Slayman,and P. S. Aronson. 1990.Cloningand

immunolocaliza-tionofaNa+/H+ exchangerinLLC-PK,cells. JSAN. 1:743.(Abstr.)

16. Ross,W. R.,-P. Breen, M. Hammerman,andA.R. Morrison. 1990.

Immunohistochemical localization of the bovine Na+/H+ exchanger. JASN. 1:730.(Abstr.)

17.Pearce, D.,R.Krapf,F.C.Rector, Jr.,andT. L.Reudelhuber. 1990.

Cloningofapartial lengthratkidneycDNAhomologoustothe humanNa/H

antiporter.KidneyInt.33:233.(Abstr.)

18.Sardet,C.,A.Franchi,and J.Pouyssegur.1989. Molecularcloning,

pri-mary structure,andexpression of the humangrowth factor-activatableNa/H

antiporter. Cell.56:271-280.

19. Krapf, R., D. Pearce, C. Lynch, X.-P. Xi, T. L. Reudelhuber, J. Pouyss6gur,and F. C.Rector,Jr. 1991.ExpressionofratrenalNa/H antiporter

mRNAlevels in responsetorespiratoryandmetabolic acidosis. J. Clin.Invest. 87:747-751.

20.Kaissling, B.,andW.Kriz.1979. Structuralanalysisof the rabbitkidney.

Ad.Anat.Embryol.Cell Biol. 56:1-121.

21. Miller, R. T., R. Lifton, L. Cunillon, G. Pages, C. Sardet, and J.

Pouyssegur. 1990. Structure of the humanNa/Hantiportergene and 5'flanking

regulatoryregion.JASN. 1:742.(Abstr.)

22.Nudel, U.,R.Jahut,M.Shani,S.Neuman,Z.Levy,andD.Yaffe. 1983. The nucleotide sequence of theratcytoplasmicbeta-actingene. Nucleic Acids

Res.11:1759-1771.

23.Sambrook, J.,E. F.Fritsch,andT.Maniatis. 1989. MolecularCloning. Second ed. ColdSpringHarborLaboratory Press,ColdSpringHarbor,NY.