Renin mRNA and Secretion

Renin mRNA Quantification Using Polymerase Chain Reaction in Cultured Juxtaglomerular Cells

Short-term Effects of cAMP

^on

Renin mRNA and Secretion

Roberto Della Bruna, Armin Kurtz, Pierre Corvol, Florence Pinet

The aim of the present studywas to setup a method to quantify renin mRNA levels in ^mouse renal juxtaglomerular cells, the main physiological site of renin synthesis. Because ofthe scarcity of the cells,

aquantitative polymerase chain reaction hadtobe developedtomeasurerenin mRNA. Juxtaglomerular cells were isolated andcultured for 2 days under various conditions, and renin mRNA^was measured directly from the cytoplasm of the cultured cells without prior RNA purification. An internal standard consisting ofa mutated renin mRNA with aninsertion of 60 bp was designed to quantify the reaction, ensuring an identical detection and amplification efficiency to the target RNA. Renin mRNA could be precisely quantified between 0.6 and 20 pg,thus allowing its detection in approximately 5000 juxtaglo- merular cells. Forskolin, an activator of adenylate cyclase, led toa concentration-dependent maximal threefold increase in renin mRNA in the cultures after 20 hours of incubation. The half-maximal effective dosewas3^X1O` mol/L. The effect of forskolin wasmimicked by 10-` mol/L isoproterenol, ^a f-receptor agonist, and by10- mol/Lisobutylmethylxanthine. A time-course study showedarapidincrease in renin mRNAwithin 3 hours after forskolin and isoproterenol addition. Renin secretion in the culture medium

wasmeasured in parallel and foundtobestimulated by bothagents.These results show thatquantitative polymerase chain reaction isasuitable tool for studying reningeneexpression in cultured juxtaglomer- ularcells. Our findings indicate that cAMP isapotentand fastactivator of reningenetranscription and renin secretion in renaljuxtaglomerular cells. (Circ Res. 1993;73:639-648.)

KEYWORDs * juxtaglomerular cells ^* renin mRNA * renin secretion * cAMP * quantitative polymerase chain reaction

T he aspartylproteaserenin issynthesized,stored, and released exocytotically by renal juxtaglo- merular (JG) cells, which are metaplastically transformed vascular smooth muscle cells.1-3

The release of renin from the JG cells is the rate- limitingstepinthegeneration ofangiotensin II, which is amajordeterminant of blood pressure and intravascular volume. Renin is synthesized as preprorenin. In the endoplasmic reticulum, a signal peptide is cleaved, yielding prorenin,which is enzymatically inactive. Pro- renin is then packaged in storage granules, where it is activated to renin by proteolytic cleavage of an N-ter- minalprofragment.4-9Therefore,thereareseveral steps of potential control of renin synthesis and secretion.

Evidence obtained in vivo suggests that reninsynthesis and secretion are concomitantly regulated by avariety offactors,includingbloodpressure,'0"'1sodium chloride

intake,12-14sympathetic

nervousoutput, andangiotensin IIlevel.15

Exocytosis of renin is influenced bythe intracellular levels ofcalcium, cAMP, andcGMP,aswellasprotein

Received December10, 1992;acceptedJune29,1993.

FromINSERMUnit 36(R.D.B., P.C.,F.P.), CollegedeFrance, Paris, France, andPhysiologischesInstitutI(A.K.), Regensburg, Germany.

Correspondence to Dr Florence Pinet, INSERM Unit 36, CollegedeFrance,3rued'Ulm, 75005Paris,France.

kinase C

activity.2"16

However, the effects of these second messengers in the regulation of renin synthesis are not well understood. Inparticular, it isnot known how reninsynthesis andrenin secretion in renal JG cells arerelated. This lack ofknowledge ismainlyduetothe lack of experimental models in which renin secretion and renin gene expression could be studied together under controlledconditions.

Recently,wehavedevelopeda mouseprimary JG cell culture modelinwhichreninsynthesisandsecretioncan be studied in parallel.Inthismodel, denovo synthesis and secretion of renin have been shown.'7 Initial find- ingshave indicated that cAMP stimulated bothsynthe- sis and secretion of renin invitro,but itwas notpossible to distinguish whether the stimulatory effect of cAMP on renin synthesis in isolated JG cells was due to an increase in translational activity of renin mRNA or primarilydue toanincrease in renin genetranscription.

Since JG cells can be isolated only in small numbers from thekidneysand donotproliferate inculture,itwas necessary toestablishasensitive and reliable assay that allows quantification of renin mRNA concentration.

Becauseof the low abundance ofthecultured cells and therefore of renin mRNA, common methods used to

semiquantify gene expression, like Northern blots,

RNase protection, or solution hybridization assays, could not be used.

Areversetranscriptase (RT)-polymerasechain reac- tion (PCR) method was therefore developed for pre- cisely quantifying mouse renin mRNA levels directly from thecytoplasm ofapproximatively5000 cultured JG cells. The very low amount of cells required for one singlemRNA measurementmakesthis method suitable for extensive investigation of renin mRNA accumula- tion indifferent experimental conditions.

It has been found that the increase in intracellular levels of cAMP in JG cells led to a time- and dose- dependent increase in renin mRNA levels. Forskolin andisoproterenolsignificantly stimulated renin mRNA levels within 3 hours. In contrast to isoproterenol, forskolin seemed to produce a biphasic stimulation of renin mRNAlevels. Thisrapid effectwasthenfollowed byasecondarydelayed increase in reninmRNAlevels.

These findings suggest that cAMP is a potent and fast-operating regulator of reninmRNAlevels in renal JG cells.

Materials and Methods Isolation ofMouseJG Cells

ThemousestrainC57B16,which carriesasinglerenin gene (Ren-1c),l8 was used in this study. JG cells were isolated aspreviously described.17 In atypical prepara- tion, thekidneys of five male anesthetized mice (age, 4 to 6 weeks) were removed, decapsulated, and minced witha scalpel blade. The minced tissue was incubated withgentlestirringinbuffer 1(mmol/L:NaCl, 130; KCI, 5; CaCl2, 2; MgCl2, 1; glucose, 10; sucrose, 20; and Tris-HCI, 10; pH7.4) supplementedwith0.25%trypsin and 0.1% collagenase at 37°C for 70 minutes. The materialwasthenfilteredthrougha22-,um nylon mesh.

Single cells passing through the nylon were collected andwashed with buffer 1 (at500gfor 10minutes). The washed single cells (final volume, 4 mL) were mixed with 60 mL of 30% isosmotic Percoll solution, equally distributed intwo centrifugation cups, and centrifuged at27OOOgin aJA-20rotor and Beckmancentrifuge for 25 minutes at 4°C. Four cellular bands differing in densitywereapparent.The bandcontaining the highest specificreninactivity(density, 1.07g/mL)wasused for primary culture.

Primary Culture ofIsolated Mouse JG Cells

The cellswererid of Percoll by washing once with50 mLofbuffer 1. The cellswerethen suspended in 3 mL culture medium (RPMI 1640 supplemented with 0.66 U/mLpenicillin, 100,ug/mL streptomycin, and 2% fetal calf serum) and distributed in

100-p

^L aliquots into 96-wellplates. The cultures^wereincubated^at37°C in a humidified atmosphere containing 5% CO2in air.

Experiments were performed after 20 hours of pri- maryculture, unless otherwise indicated.After 20 hours ofprimary culture, the culture medium was removed, and the cultures were washed once with culture me- dium. Fresh culture medium, supplemented as de- scribed above,was then added alongwith thedrugs to be tested or theirrespective solvent controls.

Renin Immunostaining

Reninimmunostainingwasperformedin cultures the second day after seeding. The culture medium was removed, and the cells were washed with phosphate-

bufferedsaline(PBS).Thecellswerefixed withBouin's solution for 15 minutes. Afterwards, the cells were washed three times for 2 minutes withPBS,andendog-

enousperoxidase activitywasblockedby incubatingthe cellsin0.5% H202inmethanol for 30 minutesatroom

temperature. The cells were then permeabilized by adding methanol and keeping them at -20°C for 4 minutes.The cells wererehydrated by washingtwice for 15 minutes with PBS and then incubated in blocking buffer (PBS supplementedwith 1%bovine serum albu- min and10%fetal calfserum)for 1 hour. Subsequently, the cells were incubated with rat antiserum directed against mouse renin19 at 4°C overnight (final dilution, 1:500inblocking buffer). Afterthisincubation, thecells were washed five times with PBS, and biotinylated rabbitimmunoglobulin Gdirected against ratimmuno- globulin G (mouse adsorbed) (final dilution, 1:500 in blocking buffer)^wasadded for 2 hoursatroomtemper-

ature.Thecells were washed twice with PBS buffer,and immunoreactivity was detected by the avidin-biotin- peroxidase complex using diaminobenzidine/H202 as a substrate. Control cells were preparedwith normal rat seruminstead of antireninserum.

Northern BlotExperiments

ForNorthern blot experiments, theJG cellsfrom40 mice were isolated anddistributedin two25-cm2culture flasks. After 20 hours ofprimaryculture, forskolin

(10`

mol/L) was added to one flaskfor 20 hours; the other was used as a control. At the end of the experiment,

total RNA was extracted by the acid guanidinium thiocyanate-phenol-chloroform method ofChomczyn- skyandSacchi,20 yielding13 and 14

gg

oftotalRNAfor theforskolin-treated andthecontrol cells, respectively.

Glyoxal-dimethyl sulfoxide (5 ,ug)-treatedRNAwas electrophoresed and transferred to Hybond N filters accordingto the method ofThomas.2' RNAblotswere exposedtoUVlight,baked for 1 hour at80°C, andthen prehybridized in 50% formamide, 5x standard saline citrate (SSC; lx SSC consists of 0.15 mol/L NaCland 0.015 mol/L sodium citrate), 50mmol/L sodium phos- phate (pH 6.5), 0.1% sodiumdodecylsulfate(SDS). lx Denhardt's solution (0.02% each of Ficoll 400, polyvi- nylpyrrolidine, and nuclease-free bovine serum albu- min), and 250 ,ug/mL heat-denatured salmon sperm DNA for 24 hours at42°C. The blotswerethenhybrid- ized with a ['2P]-labeled 1427-bp mouse renin cDNA4 for 24hoursat42°C. The filterswerewashed fourtimes at roomtemperaturewith 2x SSC and0.1% SDS, then twice at52°C with0.1 x SSCand0.1% SDS, and finally twice at 55°C with0.lx SSC and 0.1% SDS.

To normalize Northern blot results, the total RNA blots weredehybridized by heatingat 70°Cfor 2 hours with 0.005mol/LTris-HCl (pH 8.0),0.002 mol/LEDTA (pH 8.0), and 0.1x Denhardt's solution and rehybrid- ized witha 28SrRNAoligonucleotideprobe, according tothemethod ofBarbu andDautry.22 Theadvantage of this probe over other probes such as f3-actin derives from observations showing that the abundance of28S rRNA is directly related to the amount of total RNA present in the sameblot.

Densitometric scanning of the autoradiograms was performed with a densitometer (model 620CCD, Bio- RadLaboratories,Richmond, Calif) coupledwith aI-D analyst (MacIntosh Data).

Measurementof Renin mRNA by Quantitative RT-PCR

Renin mRNA quantification required setting up an RT-PCR method and the construction of an internal standard that couldbeeasilydetected. For thispurpose, a synthetic RNA containing a 60-bp insertion was designed.

Choice of the primers. To avoid coamplification of genomic DNA coding for renin, two oligonucleotide primers,onespanning theexon6/exon7border and the otherlocatedonexon8 ofthe reningene, werechosen, thus amplifying a 194-bp sequence. Sense primer (5'- ATG AAG GGG GTG TCT GTG GGG TC-3') and antisenseprimer(5'-ATGTCG GGG AGG GTGGGC ACCTG-3') were synthesized with aPCR-mate DNA synthesizer(AppliedBiosystemsInc,FosterCity,Calif).

Both primerswerechecked for the absenceof fortuitous homology to other known sequences in the GenBank data base.

Internal standard preparation. An internal standard was synthesized by in vitro transcription from plasmid pISMR. Thisplasmidwas constructed from clone pRn 1-44: complete renin cDNA 1427 nucleotides in length was isolated usingPst Irestriction sites and subcloned into the Pst I restriction site of Bluescript BS-SK (Stratagene Inc,LaJolla, Calif). Subsequently, a60-bp linker fragment corresponding to HindIII/Sac I poly- linker derived from BS-SK was inserted into blunt- ended Bcl Isite renin cDNA(942nucleotides). The Bcl I restriction site is in the amplified renin fragment.

Renin cRNAwasprepared as a sensestrand from the Sac I-digested pISMR template using T7 RNA poly- merase according tothe invitro transcription protocol (Stratagene).After thereaction,thetemplatewascom- pletelydigested with RNase-freeDNaseI. Renin cDNA could not be amplified in the absence ofreverse tran- scriptase, thus confirming that the internal standard only consisted of renin RNA.

RTreaction on cultured cells.

CELLLYSATE PREPARATION.At the end of the incuba- tionexperiments,primarycultured cellswereplacedon ice, the culture mediumwas removed, and the cultures werewashedoncewith ice-cold PBS buffer. Then,cells werelysed bytheaddition of 40,uLofanice-cold buffer containing 2% Nonidet (NP 40), 10 mmol/L Tris-HCl (pH 8), 10mmol/L NaCl, 3 mmol/L MgCl2, and 0.7%

(vol/vol) mercaptoethanol to each culture well and shakingfor 10 minutesat4°C.The cellularlysatesoftwo culture wellswerepooledfor thedetermination of renin mRNA. Cell nuclei were removed bycentrifugation at 12000gand4°C for 3 minutes. Three microliters of the cytoplasmic fraction was immediately used forreverse

transcription; the remaining partwas stored at -20°C forcellular proteinmeasurement.

RTREACrION.Tenmicroliters ofanice-cold solution containing 10pmol ofeach PCR primer, ¹ ggofyeast tRNA, and 5pgof the internal standardwasaddedto3 1,L of thecytosolicfraction. Sampleswere heatedfor5 minutesat65°Cand thenchilledonice.Then, 10,Lof solutionAwas added,and the samples^wereincubated for1hourat37°C.SolutionAconsisted of22gLofa25 mmol/Lsolution ofdeoxyribonucleotides, 45 ,uLof 5x RTbuffer (supplied with the RT kit), 6 ,L of bovine

Corp, Indianapolis, Ind), 6 ML of the RNase inhibitor rRNasin (40 000 U/mL, Promega Corp, Madison, Wis), 9

ML

of

murine-Moloney

leukemia virus RT

(200

U/ML,

GIBCO-BRL), and 20 ,L of 0.1 mol/L dithiothreitol.

After incubation, the reaction ^was stopped by heating the samples for 2 minutes at95°C.

PCR. To 3 ML of the cDNA solution, 2.5 MLof10x

PCR buffer (supplied with Taq polymerase [Boehringer Mannheim]), 1 ML (10 pmol) of each primer, 4.5 ML of

a25mmol/L MgCl2 solution, 14 ML H20, and 10 MCi of [a-32P]dCTP were added. Samples were overlayered with mineral oil, denatured at94°C for 5 minutes, and cooledto65°C for⁵ minutes. Afteraddition of 1 ML of 25 mmol/L deoxyribonucleotide solution and 1.25 U Taq polymerase, 30 PCR cycles consisting of denatur- ationat94°C (1 minute), annealingat60°C (1 minute), and extension at 72°C (30 seconds) ^were performed.

PCR ^was completed by ^a final extension step of 10 minutes at 72°C.

After PCR, the amplification products originating from renin mRNA^orfrom the internal standard^were separated bypolyacrylamide gel electrophoresis. N,N'- Methylene-bis-acrylamide was replaced by dihydroxy- ethylene-bis-acrylamide. After autoradiography, the bandswereexcised, solubilized in 0.025 mol/Lperiodic acid at 50°C, and counted in ^a beta counter. The radioactivity incorporated by renin mRNA ^{was rou-} tinely expressed ^{as a} percentage of the radioactivity incorporated by the 5pgof internal standard. Absolute values for renin mRNA were extrapolated from the standard curve generated with the internal standard: ^a theoretical correction factor of 1.3 accounted for the minor size of renin mRNA (194 bp)compared with the internal standard (254 bp). Absolute values^were used only for data expressing the decrease of renin mRNA levels withthe duration ofJG cells in culture.

Forsequence analysis, the amplified DNA fragment of interest (194 bp)wasextracted from2% agarosegel and purified with ^a Spin-X filter (Costar Corp, Cam- bridge, Mass). Both strands of DNAwerethen directly sequenced by the dideoxy termination method using ^a thermostablepolymerase derived from Thermusaquat- icus(TAQuencekitversion2.0, United States Biochem- icalCorp, Cleveland, Ohio).

Determination of Renin Secretion

Reninsecretion rateswere estimated from the ^mea- surement of renin activity appearing in the culture medium at different times. Renin activity ^was deter- mined by its ability^to generateangiotensin ^Ifrom the plasma of bilaterallynephrectomizedrats,as described.

Angiotensin ^Iwas measuredbyradioimmunoassay.23 MeasurementofProtein

Cellular protein concentration was determined ^ac- cordingtothe method ofMcKnight.24 Briefly, 20 MLof the cellular lysate ^was layered ^{on a} GFC membrane (Whatman Inc, Clifton, NJ) and precipitatedwith tri- chloroacetic acid. Proteins were stained with 0.25%

Coomassieblue,7.5% aceticacid, and 5% methanol in water.

After washing the membrane with 7.5% acetic acid and^5% methanol inwater,specific staining^waseluted in80% methanol,and absorbance was readat590nm.

serum albumin (20 mg/mL, Boehringer Mannheim Bovineserum alb

c C~~~f'

¹

e~~~~~~~~~~~~~~4

e S!

*~' ^{Se W} A

'

m>

Ro jl ^{0 0}

t 4P'

r 5>

5"1

J4 p~i

W ;

9I

.1

*A

4 S

FIG 1. Photomicrographs show immunoperoxi- dasestaining ofmousejuxtaglomerularcells on the secondday

ofprimary

culture. Specificreninimmu- noreactivity was observed in 70% to 80% ofthe cells using a rat anti-mouse renin serum (A). No stainingwasobserved with normalratserum (B).

Q4 44'! 5:0

1-4

.#9

^*

^fw*-

^{4 S A9 '-}

Statistics

Levels of significance^werecalculatedusing Student's t test. A value of P<.05 was considered significant. To

compare the time course of renin mRNA and renin stimulation, ^one way analysis variance (ANOVA) ^was carried out, and a value of P<.05 was considered significant.

Results

Enriched cultures of JG cellswereobtained by using

a Percoll density gradient separation ofsingle-cell ^sus- pensions from ^mouse kidney. On the second day of primary culture, renin-positive cells represented ap-

proximately 70% to 80% of the cultured cells, ^as assessed bythe immunoperoxidase method using ^a rat antiserum against ^mouse renin (Fig 1).

In a preliminary experiment, theeffect of intracellu- lar cAMPon reninmRNAlevel in primary cultures of JG cells ^was determined by Northernblot experiments (Fig 2). The cultured cells were incubated with the receptor-independent activator of adenylate cyclase, forskolin(10` mol/L), for 20 hours.AsingleRNA band of 1.4 kb, which had the known size of renal renin mRNA, hybridized with the^mouserenincDNAprobe.

Densitometric analysis of the autoradiogram demon-

B

stratedatwofold increase in reninmRNAby forskolin, ascompared with the control value.

However, a single experiment required 40 mice (ap- proximately 1.6x107 cells corresponding to 2.25 mgof JG cell protein) ^to obtain a sufficient amountof renin mRNA; therefore, a more sensitive technique ^was de- veloped to measure renin mRNA levels.

Detection and Characterization ofRenin mRNA by RT-PCRReaction

A eDNA fragment was amplified after RT reaction and 30 PCRcycles performed directly^onthe cell lysate ofapproximatively 5000 JG cells on the second dayof primaryculture.Foursetsof results showedtheidentity of theamplified product with reninmRNA: (1)The size ofthe PCRproduct, resolvedbyagarosegel electropho- resis and ethidium bromide staining, corresponded to the size of 194 bp predicted for a cDNA originating from renin mRNA (Fig 3A). (2) The amplified cDNA was completely digested by Sac I, producing two frag- ments of 100and 94bp (Fig 3A), as expected from the existence ofaSac I restriction site located 100bp3' of the sense primer in renin cDNA. (3) The amplified

DNAhybridizedwith a 32P-labeledoligonucleotide

(5'-

GCA TGATCAACT TCA GGG AGC TCG

TA-3')

internal to the amplified mRNA and derived from the

..j+.

A: .v.

7,:.

A 1 2

6.5 4.4

B

_1

FIc 2. Renin mRNA expression (A) and28SRNA signals(B) in primary culture of^mousejuxtaglomerularcells. TotalRNAs (5 ^{pg per}lane) from control cells(1)andforskolin-treated cells (2) werehybridized witha 2P-labeledmouserenin cDNA probe (A). Lanes^I and2weresubsequently hybridized with^aradiolabeled28Scomplementaryoligonucleotide probe for normalization (B). Thesize markers, which^areshown inthe leftmargin, areexpressedinkilobases.

mouse renin cDNA ^sequence (Fig 3B). (4) Direct sequencing of the amplified fragment confirmed the identity that ^mouse renin mRNA was amplified (not shown).

Quantification of theRT-PCR Reaction

Atleast 28 PCRcycles^werenecessarytodetect renin mRNAwhenstartingfrom the cellularlysateof 5000JG cells on the second day of primary culture. Incorpo- ration of

[a-32P]dCTP

during the PCR reaction showed thatthePCR product^wasexponentially generated from the 28th to the 32nd cycle (Fig 4). Therefore, all experimentswere performed at30cycles.

An internal standard consisting ofa synthetic renin cRNA with ^an insertion of 60 bp was constructed to

quantify renin mRNA. In an earlier experiment, renin mRNA and the internalstandard were amplified sepa-

rately to investigate whether both ^were detected with the same efficiency (Fig 5A and 5B). After reverse

transcription and 30 PCR cycles, a linear relation be- tweentheincorporated radioactivityandtheamountof startingmaterial^wasobserved from 0.45to 15^pgof the internal standard (Fig SA)and from 0.3 to10^ngof total RNAextracted from JG cells (Fig 5B). The noninter- ference of the lysis buffer ^wasstudied. The ^same serial dilutions of total RNA isolated from JG cells and of internal standard were carried out in lysis buffer. The

same slope as observed in Fig 5A and 5B with PCR buffer wasobtained (not shown).

Astandard curve wasthen generated by plotting the radioactivity incorporated by differentamountsof inter- nal standard. To ^ensure that ^no competition existed between amplification of internal standard and renin mRNA,serial dilutions of total RNA(from 0.25to8 ng) extracted from the JG cells were reverse-transcribed

1 2 1 2

** ^~.-194bp

194bp - lOObp --

A

and coamplified with a fixed amount of internal stan-

dard (5 pg). Atitration curve was carriedout usingan

increasing amount of internal standard and fixed amount of JG cell mRNA. Ten picograms of internal standard did not affect the amplification of the native mRNA (not shown). The amounts ofradioactivity in- corporated by renin mRNA^were givenas the percent-

age of the radioactivity incorporated by the internal standard and plotted against the total RNAconcentra- tion (Fig 5C). The autoradiogram showed that 5 pgof internal standard was not attenuated with increasing

amounts of total RNA. No significant difference was

observed in theslopeof thethree dilutioncurvesinFig

5,showingthat internalstandard and renin mRNAwere

amplifiedwith the^same efficiency.

On the second dayofprimary culture, renin mRNA levelinthe JG cellswas4.70+0.40pgrenin mRNAper

microgramcellularprotein (n =20). ReninmRNAlevels decreased steadily and significantly (P<.05) to 17±3%

(n=4) and to 7+5% (n=2) of the initial value on the 10th and the 17th day from onset of culture, respec-

tively.Fromthe 30th cultureday,noreninmRNAcould be detected.

Regulation of Renin mRNA bycAMP

The effect ofcAMP on renin mRNA in isolated JG cells was examined by different experiments: (1) a

receptor-independent activator of adenylate cyclase (forskolinfrom 3x108 to10` mol/L)andisoproterenol

(10` mol/L), which stimulated thecyclase by f-adren- ergic receptor interaction, and (2) the phosphodi- esterase inhibitor 3-isobutyl-1-methylxanthine (IBMX, 10-mol/L).Inatypicalexperiment,the JG cells of five micewereisolated and distributedinto30culture wells.

On the second day of primary culture, the cellular FIG 3. Characterization ofthepolymerase chain re-

action (PCR) product. A shows ethidium bromide staining ofthe PCRproduct before (lane 1) andafter (lane 2) digestion by Sac^L Thesampleswere electro- phoresed^{on a}2%agarosegel.B showshybridization of the PCR product with a 32P-labeled oligonucleotide derivedfrom exon 7ofthe mouserenin^gene. Reverse

transcriptase PCR wasperformed on total RNA ex-

tractedfrom the juxtaglomerular cells (lane 1) and directlyonthecytoplasmofjuxtaglomerularcells (lane 2)^onthe secondday ofprimaryculture.

1 2 2.0h 2.3

1.3

B

lOQOO&0

n

fl

.1

^10,000

1,000.

1an]^I-

28 29 30 31 32 3s

numberofcycles

FIG 4. Analysis of the amplification conditions. Renin cDNA was obtained by reverse transcription ofrenin RNA from 5000 cultured juxtaglomerular cells. The radioactivity incorporated byrenin mRNA wasplottedagainstthenumber ofamplification cycles. Data are mean-'-SEM (n 4).

protein content averaged 9 gg for each culture well.

After lysis of the cells and separationofthe cell nuclei by centrifugation, mRNA was reverse-transcribed and PCR-amplified. A specific amplification product was detectedfromanaliquot of each lysate correspondingto 5000 cultured cells.

As shown in Fig 6, forskolin, isoproterenol, and IBMX caused statistically significant increases in renin

200

tab 0

FIG 6. Bar graph shows regulation of renin mRNA by cAMP. Thejuxtaglomerular cellswereincubated for 20 hours with10` moliLofforskolin (n=li), isoproterenol(n-9), or 3-isobutyl-1-methylxanthine (IBMX, n=4). Values were ex- pressedas percentage of the controL Data are mean +SEM.

On the ordinate, 100% corresponded to 4.7+0.40pg renin mRNAper5000 cells (n=20).

mRNA levels in the cultured cells after 20 hours of incubation (P<.05). The strongeststimulation of renin mRNA was observed with forskolin. To determine the maximal stimulation of renin mRNA by cAMP, the concentration dependence of the effect of forskolin on renin mRNA wasexamined. As shown in Fig7, forsko- lin produced a clear dose-related effect on the renin

A C

a

I,c

.1U

-internal standard 0 -total RNA

0.25 0.5 1 2 4 8 total RNA (ng)

10 15 2

internal standard (pg)

B

K.2

300-

totalRNA (ng)

10

0 2 4 6 8 10 12

total RNA (ng)

FIG 5. Quantitativeanalysis ofrenin mRNAlevelsfromculturedjuxtaglomerularcells. A, Increasing concentrations(O.45to15 pg) of the internal standard were plotted against the radioactivity of their polymerasechain reaction (PCR) products. Data are mean+SEM (n =4). B, Increasingconcentrations (0.3to10ng) of total RNA extractedfrom juxtaglomerularcellswereplotted againsttheradioactivityof their PCR products. Data are mean±SEM(n-5). C, Increasing concentrations (0.25to8ng) oftotal RNAextractedfrom juxtaglomerularcells werecoamplifiedwith 5 pgoftheinternal standard. At the top is anautoradiogram of

asingleexperiment.Inthe graphatthe bottom, valueswereexpressedas percentageof theradioactivity incorporated by theinternal standard andplotted against total RNA concentration. Absolute values for renin mRNA were thenextrapolated from this curve and corrected by a factor of1.3 because of the minor size of renin mRNA compared with the internal standard. Data are mean±SEM(n=4).

luu F, ---.

V .

20.000-

10

1 2 3 4 5

00

0W9

^-* internal standard

-total RNA

500=

1 400-

a) 0

10-7 10-6 i6-5

forskolinconcentratdon (M)

FIG 7. Dose-response effect offorskolin on renin mRNA levels during 20 hours of incubation. At the top is an

autoradiogram of^atypicalexperiment (lane1, control; lanes 2through5,3^X10-8 10-7, 3x10-7, and1 6-h mol/L forsko- lin, respectively). Although the^sensepolymerase chain reac-

tionprimer^waschosentobe cDNAspecific, amplification of genomicDNAoriginatingfromthehybridization of 15 nucle- otides of the sense primer with exon 7 ofrenin DNA was

occasionally observed, probably ^{as a} consequence of the partial lysis of cell nuclei during the preparation ofcytoplasm, but amplification ofgenomicDNA (454 bp) was much less than amplification ofrenin mRNA (194 bp), and their size difference allowedeasyidentification. At the bottomarerenin mRNA levelsplotted ^against different forskolin concentra- tions. Values were expressed as percentage of the control value.Dataare meansof2different experiments for3x108, 3^{x 10} 7 and 3x10 ⁶mol/L forskolin, 4differentexperiments

for 10` moliL forskolin, 10 different experiments for 106 moliLforskolin, and11differentexperimentsfor10-5molIL forskolin. Bars represent the SEM in case more than 2 experiments were carriedout. On the ordinate, 100% ^corre- spondedto4.7--0.40pgreninmRNA^per5000cells (n -20).

mRNA level, which^was maximally increased by 200%

over the basal level at 10` mol/L

(10'5,

10'6 and 3x 10 mol/L of forskolin ^were statistically significant atP<.05). ED50for this effect was 3x10` mol/L.

A finalsetofexperiments^wasperformedto^assessthe timedependence of forskolin, isoproterenol, and IBMX effects ^on reninmRNA levels and also on renin secre-

tion in the cultured JG cells (Fig 8). Two-factor ANOVA showed that forskolin (Fig 8A) increased renin mRNAlevels (F, 7.8;P<.001) and renin secretion (F, 40; P<.001) toa higher extentthan did isoprotere- nol (Fig 8B). Forskolin seemed to produce ^a biphasic increase in renin mRNA. Values of renin mRNA stim- ulation by forskolin ^were statistically different for each

time of incubation tested

(F,

3.05;P<.05). Arapid and significant elevation ofrenin mRNAlevelswaspresent at3 and 6hoursof incubation (P<.05). This effectwas then followed by a secondary delayed increase in mRNA levels at20 hours (P<.05). Forisoproterenol, a significant elevation of renin mRNA levels compared with the control appeared within 3 hoursofincubation (P<.05)until 20 hours ofincubation (P<.05). Thesame level of stimulation was observed for isoproterenol during the time course. For IBMX (data not shown), a significant elevation of renin mRNA levels compared with the control value appeared after 6 hours of incu- bation (P<.05). The level ofrenin mRNA stimulation wassignificantlyhigherat 16and20hoursof incubation compared with 1 and 3 hours (P<.05). The renin secretion rate was increased and statistically different for each time of incubation by forskolin (F, 29.5;

P<.001), isoproterenol (F, 27.9;P<.001),and IBMX(F, 4.84;P<.05). Arapid and significant secretion ofrenin was observed during the first 3 and 6 hours of culture for forskolin and isoproterenol, respectively (P<.05).

This was followed by a much slower secretion rate (approximatively two times less) before a secondary increase for both forskolin and isoproterenol between 12 and 20 hours of incubation (P<.05).

Discussion

The aim of the present studywas to investigate the effect of cAMP on renin secretion and renin gene transcription in renal JG cells by measuring small amounts of renin mRNA. The model chosen was pri- mary cultures of mouse kidney cells that were highly enriched in JG cells. Thesecells,whichrepresent0.01%

to0.1% ofthekidneycells,3canbe isolated onlyinsmall numbers and do not proliferate under in vitro condi- tions. However, theystorerenin and can regulate renin secretion in response to various stimuli such as varia- tions in intracellular levels ofcAMP or cGMP.16 They can also synthesize renin, as recently shown by immu- noprecipitation of specific newly synthesized [35SJme- thionine-labeled renin.'7 Forskolin and isoproterenol appeared tobe potent stimulators ofrenin synthesisin this model. '7

To further explore the mechanisms leading to renin synthesis, evaluation of renin mRNA accumulation is necessary. Common methods for measuring mRNAlev- els, eg, Northernblots andRNAseprotectionorsolution hybridization assays, have been used to semiquantify mRNAs but require micrograms of total RNA and are therefore not suitable for large-scale investigations of reningeneexpressionin cultured JGcells.Indeed,the JG cells of at least 20 mice were required, in the present study,forasingleNorthern blotexperiment.Moreover,to quantify differences in Northern blot experiments, we have used densitometric analysis, which is limitedbythe saturation ofautoradiographed film.

In recent years, PCR has been shown to be an

extremely sensitive tool to detect low levels of mRNA.2526 It also allows precise and absolute quanti- fication ofspecificmRNA whenanappropriateinternal standard is used.27 PCR hasalreadybeen used todetect thepresenceof reninmRNA in thesingleJGapparatus of therabbit28orin extrarenal tissues of mice andrats.29 In a recent studyv Okura et al313 reported a method based on PCRto measure rat renin mRNAinextrare- ht

A

It

5 10 15 20

(hours)time

FIG 8. Graphs showreninmRNAlevels(0)and renin secretion(o)fromculturedjuxtaglomerularcellsafterthe additionof 10-5 moliL forskolin (A)and 10 ⁵mol/Lisoproterenol (B).Renin mRNA dataforforskolinrepresentthemean, and the barsrepresent the SEM of 3experimentsfor12and 16 hoursofincubation, 5experiments for1hourof incubation, 7experiments for3 and6 hoursofincubation, and 11 experiments for20 hoursofincubation. Renin mRNA datafor isoproterenolrepresent themean, and thebarsrepresentthe SEMof3experiments for12and 16 hoursincubation, 4experiments for 1, 3, and 6 hoursof incubation, and9 experimentsfor20 hours ofincubation. Renin secretion data represent the^mean, and the bars represent the SEMof3 experimentsforbothforskolin andisoproterenol. At theonsetoftheexperiment, control renin mRNAlevelsaveraged9+1.3pg RNA^per5000cells (n =4)andspontaneouslydecreasedto4.9+0.3pg per5000 cells(n =9) after20 hours.After1hour,released reninactivity^was0.93 ±0.04ngangiotensin Iper hourpermicrogramcellprotein (n =3)andspontaneouslyincreasedto1.29±0.05

ngangiotensin I^perhour^permicrogram cellprotein (n =3). Control values^weresetto 100%ateach time.

naltissues. Different levels of renin mRNA^werefound indifferent^organs, butprecise quantification could ^not be performed because of the lack of^an internal stan- dard. To ensure identical detection and amplification efficiency, the internal standard should ideally be ^re- verse-transcribed togetherwith theendogenous mRNA andcoamplifiedand should resemble asmuch aspossi- ble the native mRNA. Recently, a competitive PCR method^wasdescribedbyIwai andInagami31toquantify

rat renin mRNA in extrarenal tissues. In this study, renin mRNA^wasquantified by amplifying the RNA of interest with differentamountsof^aninternalstandard.

A quantitative PCR method, based on the data reported by Wangeta127toquantify several lymphokine mRNAs, ^was developed inthe present study forrenin mRNA. In the present study, ^a defined amount of ^a syntheticRNAwascoamplified bythe^sameprimers as

the RNA ofinterest and usedasreference. Thismethod

was appliedto primary cultures ofmouseJG cells.

Optimal amplification conditions for renin mRNA

weresetupusing totalRNAextracted from cultured JG cells. To amplify renin mRNA, 1 ^ngof total RNA^was sufficient. RT-PCR detection of renin mRNA was

therefore approximately 5000-fold ^more sensitive than the Northernblottechnique. RNA extraction, however, required relativelylargeamountsof cells: the JG cells of 20mice yield 10 to 15 ,g of total RNA. Therefore, a

technique was developed in which the cultured cells

were lysed andwhich allowed renin mRNA detection directly from the cytoplasm of the cells, without prior RNA purification. The ^very small amount of cells required for ^a single mRNA measurement makes the method suitable for extensive investigation of renin

gene expression in differentexperimental conditions.

The identity of the amplified product with renin mRNA was demonstrated by its correct size, by its complete digestion by Sac ^{I, as} predicted, and by its hybridization withan oligonucleotide derived from the

mouserenincDNA^sequence. Messenger RNA quanti-

fication^wasachievedbythe^useofaninternalstandard added to each sample before the RT reaction. To normalize differencesamongthedifferentexperimental conditions, renin mRNA levels ^were routinely ^ex- pressed aspercentages of the internal standard. Serial dilutions of renin mRNA and of the internal standard confirmed that they were both detected and amplified withthe^same efficiency. By usingtheexternallygener-

ated standard curve, absolute values for renin mRNA could be extrapolated. Absolute quantification of renin mRNA was important because the sensitivity of the PCR method allows detection of minimal amounts of RNAs inmanycelltypeswhosephysiologicalrelevance is questionable.32 Absolute quantification of renin mRNA inisolated JG cells allows thecomparison of the presentresults with those obtained in otherexperiments

or inother laboratories. A linear relation between the incorporated radioactivity and the initial ^amount of RNA^was observed in the range from 0.25 to 8 ng for total RNA and from 0.45 to 15 pg for the internal standard. As estimated from Fig 5C, renin mRNA represented approximately 2000 ^pg renin mRNA ^per microgram of totalRNApresent inthe JG cells at the second culture day, ie, 0.2% of total RNA. This value represents a400-fold enhancement compared with the 5.2 ^{pg per} microgram RNA (0.0005%) estimated for

mouse total kidney RNA by Paul et al,33 who used a

solutionhybridization^assay,andmaycorrespondtothe enrichment in renin synthesizing cells in this cell cul- ture. The dilution ^{curves were} used to optimize the initial quantity ofcellular lysate and internal standard toavoid substratedepletions, which would impair^quan- tification. This method was then applied to primary cultures ofmouseJG cells.

Ourfindingsshowed thatconditionsincreasingintra- cellularcAMPconcentration,such^asreceptor-indepen- dent activation ofadenylate cyclase by forskolin, jB-ad- renergic receptor stimulation by isoproterenol, and inhibition of cAMPdegradation by IBMX, clearlystim- 410

-30

I2O

100

i m'

B

(hours)time

ulatedrenin mRNA levels by a factor of 2 to 3 after 20 hours of incubation.

The quantification of renin mRNA by PCR was consistent with the semiquantitative result obtained by Northernblot,whereforskolin (10-5 mol/L) stimulated renin mRNA by a factor of 2 after 20 hours. These results are also in agreement with the data recently reported by Rayson34 and Everett et al,35 who used Northern blot techniques and estimated a 5- and a 1.6-fold increase of renin mRNA levels in response to increased cAMP levels in rat JG cells cultured on a basement membrane and in rat renal microvessels, respectively. The quantitative PCR technique allowed the construction of a dose-response curve. Forskolin produced a clear dose-related effect on renin mRNA levels, reaching a maximal threefold stimulation be- tween 3x10-6 and 10-5 mol/L,with an estimated ED50 of 3xl10` mol/L. A similar dose-response pattern to forskolinhas been reported forrenin secretionin rabbit renal cortical slices36 and for prorenin secretion in cultured human transfected JG cells.37 The ED5s value of 3x10`mol/L obtained for forskolinonreninmRNA levels in the present study is similar to the ED50 of 5 x10`'mol/LreportedbyPinetetal37forforskolin on cAMPlevels in human transfected JGcells.

Byuse of quantitative PCR, itwas possible to com- parefor the first time the timecourse of the forskolin-, isoproterenol-, and IBMX-induced increases in renin mRNAandrenin secretion in cultured JG cells. Renin mRNAlevelswereincreased within 1 hourafter forsko- lin or isoproterenol addition and became significant after 3 hours. ResponsetoIBMX was delayed because increase in renin mRNA levels became significant after 6hours.In contrast toisoproterenol andIBMX,forsko- lin seemed to produce a biphasic increase in renin mRNAlevels,which reachedamaximum after 3 hours.

A secondary delayed increase in renin mRNA was observed between 15 and 20hours.

Thisbiphasicprofileofhormone mRNA expression is also found in other secretory cell types. For example, after acute forskolin treatment, a strikingly similar time-course profilehas been described for thesomato- statin mRNA increase in rat pancreatic islet cells38:

afterarapid stimulatoryphase,asecondarydelayedbut more potent increase of somatostatin mRNAwas ob- served.Theauthorsexplained thisrapidrisebycAMP- mediated stimulation of somatostatingenetranscription via protein kinase A-dependent activation of cAMP responsive element-binding protein and suggested the possibility of a delayed stimulation of new protein synthesis as a compensatory response to replenish cel- lularsomatostatin storagepools. Itwould therefore be interesting to investigate whether increased intracellu- lar cAMP levels are a prerequisite for the second stimulatoryphaseof renin mRNAlevelsorwhetherthis secondaryphase representsaresponse toreplenishthe renin pool after the initial large secretion by a yet- unidentified mechanism.Atthesame time,reninsecre- tion rates were determined and compared with renin mRNA regulation. The rate of renin secretion was nonlinear,buttherewas noclearexistenceofabiphasic secretionin response tocAMP. Whereasrenin mRNA levels plateaued after 3 hours before a subsequent increase at approximatively 12 hours, an acute rise in renin secretion was observed until 5 to 6 hours after

forskolin addition. This rapid stimulation was then followedby a slower steady increase lasting at least 20 hours. The stimulatory profileofisoproterenol on renin mRNA levels and renin secretion was similar to that observed forforskolin.

The comparison ofthe present resultswith our pre- vious reported data, showing a threefold increase in renin synthesis over 20 hours of forskolin action,17 suggests that, at the level ofJGcells, changes in renin synthesisratereflect changes in reninmRNAlevels. In agreement with this interpretation is the parallelism between our results, showing a maximal threefold in- creaseinrenin mRNA levels,andthose of Nakamuraet a139 and Horiuchi et al,40 showing a 2.5- to 3-fold increase in the promoter activity of the mouse Ren-ld renin gene in response tocAMP.

Taken together, ourfindingssuggest thatcAMP is a potent and fast-operating stimulator of renin mRNA levels and renin secretion rates in cultured renal JG cells. Renin mRNA levels and renin secretion were rapidly increased in response to short-term stimula- tions. The different stimulatory profiles suggest, how- ever,the existence ofindependent regulatorypathways for both processes. Recently, using reverse hemolytic plaque assays, Everettet a135observed that the cAMP- mediated stimulation of renin release indeveloping rat renal microvessels was essentially due to the recruit- mentofnewrenin secreting cells. Theamplitude of the observed responses in our cell culture model suggests that, in the adultmouse, a similar phenomenon could onlypartiallyaccount for the cAMP-mediated stimula- tionof renin secretion and renin mRNAlevels.

Acknowledgments

Thisstudywasin partsupportedbyaEuropean Molecular Biology Organizationshort-termfellowship andbySwissNa- tional Science Foundation Grant 31-26381.89. We wish to thank Dr Florent Soubrier for helpful discussions and Dr ShelleyKlemm and Dr EleanorDavies for critical readingof the manuscript. We wish to thank G. Masquelier and A.

Boisquillonfor artwork.

References

1. Barajas L. Anatomy of the juxtaglomerular apparatus. Am J Physiol. 1979;236:F333-F343.

2. HackenthalE, PaulM,GantenD,Taugner R.Morphology,phys- iology,and molecularbiologyof renin secretion.PhysiolRev. 1990;

70:1067-1116.

3. Taugner R, BuehrleCP, HackenthalE, Mannek E, NobilingR.

Morphologyofthejuxtaglomerularapparatusand secretory mech- anisms.In:Berlyne Giovanetti S,ed.ContributionsofNephrology.

Basel,Switzerland: SKarger,AG; 1984:76-101.

4. Panthier JJ, Foote S, Chambraud B, Strosberg AD, Corvol P, RougeonF.Completeaminoacid sequenceand maturationof the mousesubmaxillaryglandreninprecursor. Nature. 1982;298:90-92.

5. CatanzaroDF,MullinsJJ,Morris BJ.Thebiosyntheticpathwayof renin in mouse submandibular gland. JBiol Chem. 1983;258:

7364-7368.

6. ImaiT,MiyazakiH,HiroseS,Hori H,HayashiT,KageyamaR, Ohkubo H, Nakanishi S, Murakami K. Cloning and sequence analysisof cDNA for human reninprecursor.ProcNatlAcad Sci U S A. 1983;80:7405-7409.

7. Pratt RE, Ouellette AJ, Dzau VJ. Biosynthesis ofrenin: multi- plicityofactiveandintermediate forms. ProcNatlAcadSci U S A.

1983;80:6809-6813.

8. GalenFX,DevauxC,HouotAM,MenardJ,CorvolP,CorvolMT, Gubler MC, Mounier F, Camilleri JP. Renin biosynthesis by human tumoraljuxtaglomerular cells: evidences forareninpre- cursor.JClin Invest. 1984;73:1144-1155.

9. Hirose S, Kim SJ, Hotoshi M, ParkYS, Murakami K. In vitro bio- synthesis of human renin and identification ofplasmainactive renin as anactivationintermediate. J Biol Chem.1985;260:16400-16405.

10. Moffet RB, Mc Gowan RA, Gross KW. Modulation of kidney renin messenger RNA levelsduringexperimentallyinducedhyper- tension. Hypertension. 1986;8:874-882.

11. Samani NJ, Godfrey NP, Major JS, Brammar WJ, Swales JD.

Kidney renin mRNA levels in the early and chronic phases of two-kidney,oneclip hypertensionin the rat. JHypertens. 1989;7:

105-112.

12. Morimoto S, Abe R,FukuharaA,Tanaka K, Yamamoto K. Effect ofsodium restrictiononplasmareninactivityand reningranulesin theratkidney.Am JPhysiol. 1979;237:F367-F371.

13. Nakamura N, Soubrier F, Menard J, Panthier JJ, Rougeon F, Corvol P.Nonproportional changes inplasmareninconcentration, renalrenin content, and rat renin messenger RNA. Hypertension.

1985;7:885-859.

14. Ludwig G, Ganten D, Murakami K, Fasching U, Hackenthal E.

Relationship between renin mRNA and renin secretion in adre- nalectomized, salt depleted, or converting enzyme inhibitor- treated rats.MolCell Endocrinol. 1987;50:223-229.

15. Dzau VJ, Burt DW, Pratt RE. Molecular biology of the renin- angiotensinsystem. Am JPhysiol. 1988;255:F563-F573.

16. Kurtz A. Cellular control of renin secretion. RevPhysiol Biochem Pharmacol. 1989;113:1-40.

17. DellaBruna R, Pinet F,CorvolP,Kurtz A. Regulation of renin secretion and renin synthesis by second messengers in isolated mousejuxtaglomerularcells.Cell PhysiolBiochem. 1991;1:98-110.

18. PanthierJJ,HolmI,Rougeon F.The mouse Rnlocus: S allele of the renin regulator gene results from a single structural gene duplication.EMBO J. 1982;1:1417-1421.

19. Michel J-B, Sayah S, Guettier C, Nussberger J, Philippe M, GonzalesM-F, CarelliC, Galen F-X, Menard J, Corvol P. Physi- ologicalandimmunopathologicalconsequences of active immuni- zation of spontaneously hypertensive and normotensive rats against murine renin. Circulation. 1990;81:1899-1910.

20. Chomczynski P, Sacchi N. Single step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction.Anal Biochem. 1987;162:156-159.

21. Thomas PS. Hybridization of denatured RNA and small DNA fragmentstransferred to nitrocellulose. Proc Natl AcadSciU S A.

1980;77:5201-5205.

22. Barbu V, Dautry F. Northern blot normalization with a 28S rRNA oligonucleotideprobe.Nucleic AcidsRes. 1989;17:7115.

23. Menard J, Catt K. Measurement of renin activity, concentration andsubstrate in ratplasma by radioimmunoassay of angiotensinI.

Endocrinology. 1972;90:422-430.

24. McKnight GS. A colorimetric method for the determination of submicrogramquantities of protein. Anal Biochem. 1977;78:86-92.

25. Erlich HA, Gelfand D,Sninsky JJ. Recent advances in thepoly- merase chain reaction. Science. 1991;252:1643-1651.

26. Saiki RK, Gelfand DH, Stoffel S, ScharfSJ,Higuchi R, Horn GT, Mullis KB,Erlich HA.Primer-directed enzymaticamplificationof DNA with athermostable DNA polymerase. Science. 1988;239:

487-491.

27. Wang AM,Doyle MV,Mark DF.Quantitationof mRNAbythe polymerase chain reaction. Proc Natl Acad Sci U S A. 1989;86:

9717-9721.

28. Trivedi BK, Briggs JP, Killen PD.Applicationofpolymerase chain reaction techniques to study of rabbit renin gene expression.

Kidney Int. 1991;39:s23-s27.

29. EkkerM, TronikD, RougeonF.Extra-renal transcriptionof the renin gene inmultipletissues of mice and rats. Proc Natl Acad Sci USA. 1989;86:5155-5158.

30. Okura T, Kitami Y, Iwata T, Hiwada K. Quantitative mea- surement of extra-renal renin mRNA by polymerase chain reaction. BiochemBiophysResCommun. 1991;179:25-31.

31. Iwai N,InagamiT.Quantitativeanalysisofrenin geneexpression in extrarenal tissues by polymerase chain reaction method.

JHypertens. 1992;10:717-724.

32. Chelly J, Concordet J, Kaplan JC, Kahn A. Illegitimate tran- scription:transcriptionof any gene in anycell type. ProcNatlAcad SciUSA. 1989;86:2617-2621.

33. PaulM, Wagner D, Metzger R,Ganten D, Lang RE, Suzuki F, Murakami K, Burbach JHP, Ludwig G. Quantification ofrenin mRNAinvarious mouse tissues by a novel solution hybridization assay. JHypertens. 1988;6:247-252.

34. Rayson BM. Juxtaglomerular cells cultured on a reconstituted basement membrane. Am JPhysiol. 1992;262:C563-C568.

35. Everett AD, Carey RM, Chevalier RL, Peach MJ, Gomez RA.

Reninrelease and gene expression in intact ratkidneymicrovessels andsinglecells. JClinInvest. 1990;86:169-175.

36. ParkSC,Sigmon DH, Han SD, Honeyman TW, Fray JCS. Control of renin secretionby Ca 2+ andcyclicAMPthrough two parallel mechanisms. Am JPhysiol. 1986;251:R531-R536.

37. Pinet F, Mizrahi J, LaboulandineI, Menard J, Corvol P. Regu- lation of prorenin secretion in human transfected juxtaglomerular cells. JClin Invest. 1987;80:724-731.

38. Patel YC, Papachristou DN, Zingg HH, Farkas EM. Regulation of isletsomatostatinsecretion and gene expression: selective effects ofadenosine3',5'-monophosphate andphorbol esters in normal islets of Langerhans and in a somatostatin-producing rat islet clonal cell line 1027 B2.Endocrinology. 1991;128:1754-1762.

39. Nakamura N, Burt DW, Paul M, Dzau VJ. Negative control elements and cAMP responsive sequences in the tissue-specific expression of mouse renin genes. ProcNatl Acad Sci U S A.

1989;86:56-59.

40. Horiuchi M, Nakamura N, Tang SS, Barrett G, Dzau VJ.

Molecular mechanism of tissue-specific regulation of mouse renin geneexpression by cAMP. J Biol Chem. 1991;266:16247-16254.