Proc. Natl. Acad. Sci. USAVol. 86, pp. 5155-5158, July 1989Medical Sciences

Extra-renal transcription of the renin genes in multiple tissues ofmice and rats

(renin-angiotensin system/polymerase chain reaction)

MARC EKKER, DIANA TRONIK, AND FRANMOIS ROUGEON*Unite de Gdndtique et Biochimie du D6veloppement, Laboratoire associd Centre National de la Recherche Scientifique 361, DIpartement d'Immunologie,Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France

Communicated by Frangois Jacob, April 17, 1989 (received for review February 9, 1989)

ABSTRACT Expression of the mouse renin genes (Ren-)and Ren-2) and of the unique rat renin gene was determined inseveral extra-renal tissues of mice and rats by primer-directedenzymatic amplification of cDNAs. In addition to the adrenalglands, testis, and ovaries, renin tlnscripts are detected in theliver, whole brain, and hypothalamus and, at lower levels, inspleen, thymus, lung, and prostate. Expression of the rat reningene correlates with that of the mouse Ren-) gene with thenotable exception of the submaxillary gand where renin tran-scripts are found only in mice. The levels of renin transcriptsin the liver of females from both species are higher than inmales. In mice, the relative levels of Ren-I and Ren-2 tran-scripts vary widely in different tissues. These results supportthe hypothesis of a local renin-angiotensin system in multipleextra-renal sites and imply the existence of complex mecha-nisms of regulation of the renin gene, previously thought to beexpressed in a tissue-specific manner.

Renin plays an important role in the control ofblood pressureby cleaving angiotensinogen into angiotensin I. Circulatingrenin, in its active and inactive pro-renin forms, is producedby the juxtaglomerular cells of the kidney. In mice, thesubmaxillary gland (SMG) constitutes another major sourceof renin. The observation of renin activity in several otherextra-renal tissues (for review, see ref. 1) has led to thesuggestion of extra-renal renin-angiotensin systems. How-ever, activity measurements are sometimes ambiguous due tononspecific renin-like activities of other proteases and thepossible contribution of blood-borne renin in tissue samples(1). The synthesis of renin in extra-renal sites has beensupported by the detection of renin mRNA in tissues includ-ing the adrenal glands, testis, ovaries, brain, and heart (2-6).However, in the latter studies, renin mRNA levels are oftenclose to the limit of detection by RNA blot analysis. Fur-thermore, the methods of detection used in studies of extra-renal renin gene expression have not allowed discriminationbetween transcripts of the two mouse renin genes, Ren-J andRen-2.

Strains of laboratory mice can be divided in two categorieson the basis of SMG renin expression: Strains, such asBALB/c or C3H, that produce small amounts of SMG reninpossess a single renin gene, designated Ren-), which encodesfor the circulating enzyme. Strains, such as DBA/2 or Swiss,which have an additional copy of the renin gene, designatedRen-2, exhibit 100-fold higher SMG renin (7-10) resultingfrom the higher SMG transcription- of Ren-2. Renin activityis higher in the SMG of males and transcriptional regulationof Ren-) and Ren-2 by androgens has been demonstrated(11). In contrast to the SMG, Ren-) and Ren-2 are transcribedat the same level in kidney juxtaglomerular cells (12).

To elucidate the mechanisms underlying tissue-specificexpression of the mouse and rat renin genes and the differ-ential expression of the two mouse genes, several tissueshave been examined for the presence of renin transcriptsusing a specific and much more sensitive method based uponamplification of renin mRNA sequences by the polymerasechain reaction (PCR) (13, 14).

MATERIALS AND METHODSAmplification of Renin cDNA. Total RNA was prepared

from tissue as described (15). Five micrograms of total RNAwas used for cDNA first-strand synthesis (15) in a totalvolume of 15 A.l using 0.1 pug of (dT),2.18 as primer andMoloney murine leukemia virus reverse transcriptase. Afteran incubation period of 1 hr at 37C, the PCR with Thermusaquaticus DNA polymerase (14) was carried out directly onthe entire cDNA reaction mixture or on a 1.5-Al aliquot in afinal reaction volume of 100 p.l. The concentration of the twoPCR primers was in excess over the oligo(dT) and the meanlength of the latter limited annealing at the temperature usedfor amplification. Therefore, only the amplified productprimed by the two specific primers was obtained. Three dryblocks were used for the enzymatic amplification. Each cycleconsisted of heating for 2 min at 950t, followed by annealingat 520C for 2.5 min, and elongation at 700C for 2 min. This wasrepeated for 30 cycles. The last cycle was followed by anadditional incubation period of 8 min, at 700C. After gradualcooling to room temperature, the DNA was precipitated withethanol and dissolved in 100 ul of water.

Southern Blot Analysis. A 10-lI aliquot of the amplifiedcDNA was digested with Alu, I (mouse samples only) beforeloading on a 2.5% agarose gel. The DNA was transferred bycapillarity action to a Zeta-bind nylon membrane. The inter-nal probe was end-labeled in the presence of radiolabeledATP by T4 polynucleotide kinase. The probe hybridizationwas carried out at 550C in 6x SSC/10x Denhardt's solution/0.7% SDS/25 mM sodium phosphate, pH 7.0/2 mM EDTA/sonicated salmon sperm DNA (100 pug/ml). The blot waswashed in 3x SSC/0.5% SDS at 550C. (lx SSC = 0.15 MNaCl/0.015 M sodium citrate, pH 7.0; lx Denhardt's solu-tion = 0.02% polyvinylpyrrolidone/0.02% Ficoll/0.02% bo-vine serum albumin.)

RESULTSFor amplification of renin mRNA sequences, one cDNAstrand was synthesized from total RNA in a small volume.The PCR was then directly performed on the cDNA reactionmixture. As shown in Fig. 1, the amplification primerscorresponded to sequences located in a region of the gene

Abbreviations: SMG, submaxillary gland; PCR, polymerase chainreaction.*To whom reprint requests should be addressed.

5155

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

May

27,

202

1

5156 Medical Sciences: Ekker et al.

AAlu 1 (Ren-2 only)

EXON 9EXON 7

B- PAAA

cDNA first strand synthesis

cDNA first strand amplification

Ren-1 + Ren-2

Alu (Ron-2)

Alu I Digestion

Ren-I 161 bp

Ren-2 109 bp

FIG. 1. Schematic illustrationofthe procedure for the enzymaticamplification of mouse renin

kA Ron-1 + Ren-2 cDNA sequences. (A) Structure ofrTr mFAs the mouse Ren-) and Ren-2 geneswith arrows indicating the positionof oligonucleotide primers. Theprimers used for amplification of

Ren-1 + Ren-2 rat renincDNA occupy equivalentcDNA first strand positions in the unique rat renin

gene whose structure is similar tothat of Ren-) and Ren-2. (B) Pro-cedure for enzymatic amplifica-tion of cDNA sequences. Thehatched box represents the inter-nal probe used for Southern anal-

161 bp ysis of amplified products for bothmouse and rat. Sequences of theoligonucleotides used in this workare as follows: mouse exon-7primer, TCATGCAAGCCCTGG-GAGCC; mouse exon-9 primer,CTTGTCTCTCCTGTTGGGAT;rat exon-7 primer, TCATGCAA-GCCCTGGGAGTC; rat exon-9primer, GTCATCGTTCCTGAA-GGGAT; and internal probe, CG-TAGTCCGTACTGCTGAGTGT-GTAGGCCCTGCCTCCC. kb,Kilobase(s); bp, base pair(s).

that is common to the two mouse genes and the unique ratrenin gene. Each primer was specific for sequences found intwo exons of the genes to avoid confusion of the signalcorresponding to a renin transcript with that resulting from apossible contamination of the RNA sample with DNA. Forthe mouse genes, the primers corresponded to sequences thatare identical for Ren-) and Ren-2 and amplified fragments ofequal sizes (161 base pairs) were obtained. However, an AluI restriction site unique to Ren-2 contained within the am-plified fragment (Fig. 1) allowed discrimination between thetranscripts of the two genes on the basis of size. Finally, theamplified cDNAs were hybridized to an internal probe toincrease both the specificity and the sensitivity of themethod.When kidney or adrenal RNA samples from mice of the

two-gene DBA/2 strain were subjected to 30 cycles ofamplification,- digestion with Alu I, and separation by agarosegel electrophoresis, fragments of the size expected for themouse Ren-) and Ren-2 transcripts were observed afterethidium bromide staining of the gel (data not shown). Thefragments hybridized to the internal probe after Southerntransfer (Fig. 2A, lanes 4 and 7). As expected, kidney orSMGRNA samples from the one-gene BALB/c strain analyzedwith the same procedure produced only the fragment corre-sponding to Ren-J (Fig. 2A, lane 5, and data not shown). Thefragment corresponding to Ren-1 transcripts in the testis ofDBA/2 mice was observed after staining of the gel butSouthern transfer to a nylon membrane and hybridizationwith the internal probe was necessary to clearly show thepresence of Ren-2 transcripts in the testis (Fig. 2A, lane 3).Similarly, Ren-) transcripts were also found in the liver andbrain of male and female mice, as well as in ovaries (Fig. 2).

In brain and ovaries, Ren-2 transcripts were found at levelssimilar to those of Ren-), a situation similar to that found inthe kidney or adrenal glands, whereas liver Ren-2 expression

A1 2 3 4 5 6 7 8

s

B 1 2 3 4 5 6 7 89 5 6 8 9

a S.- Ren-1- Ren-2

FIG. 2. Detection of renin transcripts in mouse tissues by enzy-matic amplification ofcDNAs. (A) Southern analysis of amplificationproducts from tissues of male mice. All samples originate from miceof the DBA/2 strain unless otherwise indicated. Lanes: 1, brain; 2,liver; 3, testis; 4, kidney; 5, kidney [BALB/c mice; 1 ng of poly(A)+RNA]; 6, heart; 7, adrenal glands; 8, spleen. The number of cycleswas 30. Exposure time was 1 hr (lanes 1-5) or 24 hr (lanes 6-8). (B)Southern analysis of amplification products from tissues of femalemice (DBA/2 strain). Lanes: 1, adrenal glands; 2, liver; 3, brain; 4,ovaries; 5, spleen; 6, kidney; 7, heart; 8, lung; 9, thymus. The numberof cycles was 30. Exposure time was 2 hr (lanes 1-9) or 24 hr (lanes5'-9').

Proc. Natl. Acad Sci. USA 86 (1989)

Dow

nloa

ded

by g

uest

on

May

27,

202

1

Proc. Natl. Acad. Sci. USA 86 (1989) 5157

1 2 3 4 5 6 7 8 9 10 11 12 K T L S1 10 100 1 10 100 1 10 100 1 10

*I- p. S * 0 *

FIG. 3. Detection ofrenin transcripts in rat tissues (Wistar strain)by cDNA amplification. Lanes: 1, male SMG; 2, female SMG; 3,male liver; 4, female liver; 5, brain; 6, prostate; 7, testis; 8, spleen;9, kidney; 10, hypothalamic region; 11, ovaries; 12, pituitary. Thenumber of cycles was 30. Exposure time was 5 hr Qanes 1-11) or 16hr (lane 12).

was 1ower than that of Ren-J. In addition to these tissues,weaker signals were obtained in the spleen and lung whereonly Ren-1 can be detected and in the thymus where bothgenes are expressed at similar levels. Surprisingly, we couldnot detect renin transcripts in the heart although reninexpression has been reported for this tissue (4).

Analysis of extra-renal renin transcripts was extended tothe unique rat renin gene. The coding sequence ofthe rat generesembles more closely that of the mouse Ren-) than that ofRen-2 (16). As shown in Fig. 3, tissue distribution of the ratrenin transcripts reflected that of Ren-1 with the notableexception of the SMG. Here, transcripts were undetectablein the rat, whereas they were easily detectable in mousemales and females of the low-producing BALB/c strain (ref.17 and unpublished observations). Renin transcripts werealso found in the rat prostate, the dissected hypothalamicregion, and the pituitary. There appeared to be high levels ofrenin transcripts in the rat hypothalamus since the amplifiedrenin fragment was observed after staining of the agarose gel(data not shown).

Sex-related differences in SMG renin expression is well-documented (11). Renin mRNA levels in the SMG of malemice are about 5 times higher than in females. Of all tissuesthat were analyzed in this study, the liver was the only otherorgan where a sex-related difference in levels of renin tran-scripts was apparent. In about a dozen liver samples from rator different strains of mice (DBA/2 or Swiss), the intensity ofthe hybridization signal was always greater in females than inmales (some of the data can be seen in Figs. 2 and 4). It isnotable that the mRNA levels of angiotensinogen, the reninsubstrate, whose major site of production is the liver, arestimulated by estrogens in this organ (18).To obtain an estimate of the relative levels of renin

transcripts in different tissues, serial dilutions of the cDNAsynthesis reaction mixtures from kidney, testis, liver, andspleen were subjected to the PCR. As can be seen in Fig. 5,for a given tissue, the intensity of signals resulting from serialdilutions does not decrease in the same ratio as the dilutionsthemselves. In this experiment, the number of amplificationcycles was kept constant, and the efficiency of the procedure

1 2 3 4

Ren-1

Ren-2

FIG. 4. Amplified renin cDNA from mouse liver (Swiss strain).Lanes: 1 and 2, males; 3 and 4, females. The number of cycles was30. Exposure time was 6 hr.

FIG. 5. Enzymatic amplification of serial dilutions of cDNAsfrom mouse tissues (DBA/2 strain). K, kidney; T, testis; L, liver; S,spleen. The dilution factor is indicated. A dilution factor of 1represents the amplification products obtained from cDNA synthe-sized from 0.5 ,ug of total RNA. The number of cycles was 30.Exposure time was 5 hr.

probably decreases as the amount of starting material in-creases. This was especially true for the kidney and testissamples, which show high levels of transcripts. Thus, to beable to estimate the relative levels of transcripts in varioustissues, signals of similar intensities must be compared. Theintensity of the hybridization signal obtained with a 1:100dilution of the kidney sample was equivalent to that obtainedwith the undiluted testis sample. Thus, total renin transcriptsin the kidney (Ren-) plus Ren-2) may be more abundant thanin the testis by about two orders of magnitude. Comparisonof signals from serial dilutions of the testis, liver, and spleensamples indicates that testis have higher levels of renintranscripts than the liver by at least one order of magnitudeand that the liver has more than the spleen by a comparablefactor. These comparisons are very approximative and onlyreflect total tissue renin RNA levels regardless of the natureand number of cells in a given tissue that express the reningene. For instance, the juxtaglomerular cells of the kidneyrepresent only a small proportion of total kidney cells.Similarly, the low levels observed in tissues such as spleen,lung, and thymus might reflect either relatively higheramounts of transcripts in a cell type of low abundance in thattissue or low levels in the most abundant cell type. Therefore,on a cellular basis, relative renin mRNA levels may differsignificantly from those deduced from data presented in Fig.5.

DISCUSSIONPrimer-directed amplification of cDNA sequences with thePCR has allowed us to detect renin mRNAs in multipletissues ofmice and rats. A procedure similar to that describedin the present work has allowed determination of dystrophintranscripts in human muscle and non-muscle tissues (19). Aconsiderable increase in sensitivity is the obvious advantageof the amplification method compared to RNA blot analysis,primer extension, S1 nuclease mapping, or in situ hybridiza-tion. The major drawback compared to in situ hybridizationis the inability to identify which cells in a given tissue expressthe renin gene. Amplification of sequences from differentexons, hybridization of the amplified fragment to the internalprobe and generation ofDNA fragments of the expected sizeafter digestion with Alu I (Ren-2) confirm that the signaldetected corresponds to renin transcripts. The inability todetect renin transcripts in the heart or the rat SMG make itappear unlikely that the small levels of transcripts detected intissues such as the spleen, thymus, lung, and prostate are theresult of a very low but existent "background" transcriptionof the genes or due to the presence of renin transcripts in acell type common to all tissues. In this respect, renin-likeactivities in the arterial wall tissue of many species has beenreported by several authors (1). Therefore, a very sensitive

Medical Sciences: Ekker et al.

#* Affik

&

Dow

nloa

ded

by g

uest

on

May

27,

202

1

5158 Medical Sciences: Ekker et al.

method such as that used in this work should enable detectionofrenin transcripts in every tissue sample due to the presenceof blood vessels in the sample. However, such was not thecase and, therefore, the presence of renin or renin-likeactivity in blood vessel walls may not be a demonstration ofthe synthesis of the protein in this tissue.Although comparison of levels of amplified fragments only

yields an approximation of the relative levels of transcripts indifferent tissues, one can clearly see that the major site ofproduction of renin is the kidney (along with the submaxillarygland in mice). However, expression ofthe renin gene occursin several other tissues. The testis and adrenal glands can beconsidered good sources of transcripts. Comparatively, theliver, ovaries, and brain show lower levels that are in turndistinctively higher than those observed in the spleen, lung,thymus, and prostate. As stated above, due to the inability ofthe method to tell in which cells of a given tissue the gene istranscribed, we cannot determine if the low levels observedin the spleen, lung, thymus, and prostate are the result ofhigher expression in a rare cell type ofthose tissues or ofverylow (perhaps less than one molecule of renin mRNA per cell)in a major cell type. Because the latter possibility cannot beentirely excluded at this time, the physiological significanceof renin transcription in such organs still needs to be ad-dressed with caution.The presence of renin transcripts in several extra-renal

sites argues for the existence of local renin-angiotensinsystems. Physiological roles for a local renin-angiotensinsystem have been suggested in processes such as the inflam-matory process, the regulation of the local vascular tone inendocrine organs to increase or decrease blood flow whenhormone release needs to be modulated (20) or simply theproduction of angiotensin II for local physiological effect(21). A local renin-angiotensin system would require a lowersynthesis of renin than that of the kidney, which mustmaintain circulating renin levels. This could explain therelatively lower levels of transcripts in extra-renal tissues.Another site of regulation of angiotensin II production mightbe found at the level of angiotensinogen gene expression.Angiotensinogen mRNA has been detected in tissues otherthan the liver, its major site of production, but it is not knownif they are found in all tissues that produce renin transcripts.Measurement of angiotensinogen transcripts by the PCRshould help to further clarify this issue.The fate of the pro-renin synthesized in extra-renal tissues

remains to be elucidated. Pathways for the proteolytic pro-cessing and secretion of renin in renal and extra-renal cellsare not yet fully understood. Different pathways may betaken by the renin precursor according to the cell typeespecially in tissues such as the liver that lack the regulatedpathway of secretion. Thus the possibility that most of theextra-renal renin exits the cells in the proenzyme form or isdegraded cannot be excluded. For instance, it is known thatRen-2 is transcribed in the kidney to the same extent as Ren-)but that the protein product of Ren-2 cannot be detected toa significant extent (11, 22).The tissue distribution of the rat renin transcripts reflects

that of the mouse Ren-J gene with the major difference thatrenin transcripts are absent from the rat SMG. Previousattempts to detect renin transcripts in the SMG of rats byNorthern blot or primer-extension analysis were unsuccess-ful (17). Analysis of the promoter regions of mouse, rat, andhuman renin genes carried out in this laboratory has yieldedclues about the gene elements responsible for renin expres-sion in the SMG (17).

Analysis of extra-renal expression of the renin genessuggests numerous sites of genetic regulation. In addition to

the variable levels of renin expression found in variousorgans of mice and rats that might depend upon cis- andtrans-acting elements, different balances between levels ofRen-) and Ren-2 transcripts and different male/female ratiosare observed in mouse tissues. With the notable exception ofthe SMG, the levels of Ren-2 transcripts are either similar tothose ofRen-) (e.g., in the kidney, adrenal glands, brain, andovaries) or lower (e.g., in testis, liver, spleen, and lung).Transcriptional control of the two mouse renin genes byandrogens in the SMG implies the presence of androgenresponse elements within the genes, and the higher reninexpression in the liver of females from both mice and ratssuggests the presence of other hormone response element(s).

Finally, the results of this study of renin gene expressionsupport the hypothesis of local renin-angiotensin systems inmultiple extra-renal tissues. This regulation is achieved bythe interaction of multiple elements of a system whosecomplexity goes far beyond that of the simple example oftissue-specific expression that the renin genes were oncethought to be.

We are grateful to Isabelle Chupin and Ruth Ladenheim for kindlyproviding the rat RNA samples and to Gordon Langsley, GeorgesLutfalla, Roger Ollo, and Charles Roth for critical reading of themanuscript. This work was supported by grants from the Associationpour la Recherche sur le Cancer (ARC), the Institut National de laSantd et de la Recherche Mddicale, and the Institut Pasteur de Paris.M.E. holds a Research Fellowship from the Medical ResearchCouncil of Canada.

1. Deboben, A., Inagami, T. & Ganten, D. (1983) in Hypertension, eds.Genest, J., Kuchel, O., Hamet, P. & Cantin, M. (McGraw-Hill,New York), pp. 194-209.

2. Field, L. J., McGowan, R. A., Dickinson, D. P. & Gross, K. W.(1984) Hypertension 6, 597-603.

3. Dzau, V. J., Ingelfinger, J., Pratt, R. E. & Ellison, K. E. (1986)Hypertension 8, 544-548.

4. Dzau, V. J., Ellison, K. E., Brody, T., Ingelfinger, J. & Pratt, R. E.(1986) Endocrinology 120, 2334-2338.

5. Kim, S. J., Shinjo, M., Fukamizu, A., Miyazaki, H., Usuki, S. &Murakami, K. (1987) Biochem. Biophys. Res. Commun. 142, 169-175.

6. Deschepper, C. F., Mellon, S. H., Cumin, F., Baxter, J. D. &Ganong, W. F. (1986) Proc. Natl. Acad. Sci. USA 83, 7552-7556.

7. Piccini, N., Knopf, J. L. & Gross, K. W. (1982) Cell 30, 205-213.8. Panthier, J. J., Holm, I. & Rougeon, F. (1982) EMBO J. 1, 1417-

1421.9. Mullins, J. J., Burt, D. W., Windass, J. P., McTork, P., George, H.

& Brammar, W. J. (1981) EMBO J. 1, 1461-1466.10. Wilson, C. M., Cherry, M., Taylor, B. A. & Wilson, J. D. (1981)

Biochem. Genet. 19, 509-523.11. Wilson, C. M. & Taylor, B. A. (1982) J. Biol. Chem. 257, 217-223.12. Field, L. J. & Gross, K. W. (1985) Proc. Natl. Acad. Sci. USA 82,

6196-6200.13. Saiki, R. K., Scharf, S., Faloona, F., Mullis, K. B., Horn, G. T.,

Erlich, H. A. & Arnheim, N. (1985) Science 230, 1350-1354.14. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R.,

Horn, G. T., Mullis, K. B. & Erlich, H. A. (1988) Science 239,487-491.

15. Davis, L. G., Dibner, M. D. & Battey, J. F. (1986) in Basic Meth-ods in Molecular Biology (Elsevier, Amsterdam), pp. 130-135.

16. Burnham, C. E., Hawelu-Johnson, C. L., Frank, B. M. & Lynch,K. R. (1987) Proc. Natl. Acad. Sci. USA 84, 5605-5609.

17. Tronik, D., Ekker, M. & Rougeon, F. (1988) Gene 69, 71-80.18. Campbell, D. J. & Habener, J. F. (1986) J. Clin. Invest. 78, 31-39.19. Chelly, J., Kaplan, J.-C., Maire, P., Gautron, S. & Kahn, A. (1988)

Nature (London) 333, 858-860.20. Goldblatt, P. J. (1986) Arch. Pathol. Hab. Med. 110, 1128-1130.21. Urata, H., Khosla, M. C., Bumpus, F. M. & Husain, A. (1988)

Proc. Natl. Acad. Sci. USA 85, 8251-8255.22. Inagami, T., Murakami, K., Yokosawa, H., Matoba, T., Yokosawa,

N., Takii, Y. & Misono, K. (1980) in Enzymatic Release of Vaso-active Peptides, eds. Gross, F., Vogel, H. G. (Raven, New York),pp. 7-30.

Proc. Natl. Acad. Sci. USA 86 (1989)

Dow

nloa

ded

by g

uest

on

May

27,

202

1