Transient inhibition of angiotensinogen production in transgenic mice bearing an antisense angiotensinogen gene

Kidney International, VoL 47 (1995), pp. 1638—1646

Transient inhibition of angiotensinogen production in transgenic mice bearing an antisense angiotensinogen gene THIERRY PEDRAZZINI, PASCAL CouSIN, JEAN-FRANcOIS AUBERT, and FLs-R. BRUNNER Division of Hypertension, Lausanne University Medical Schoo4 Lausanne, Switzerland

Transient inhibition of angiotensinogen production in transgenic mice bearing an antisense angiotensinogen gene. Angiotensinogen is the precursor of the biologically active hormone angiotensin II. Enzyme kinetic parameters suggest that concentrations of plasma angiotensinogen are rate limiting in the renin reaction. It is therefore assumed that a decrease

blockade of the renin-angiotensin system, the safety of a more or less permanently inhibited renin.angiotensin system has not been established. More recently, renin inhibitors have also been tested as a means to specifically block the renin-angiotensin system [9,

in angiotensinogen synthesis in vivo would result in a decrease in angiotensin H plasma levels and then of blood pressure. To test this

10]. Although promising, the short-term effect of these drugs prevents their use in chronic treatments [11]. In this context, angiotensinogen might represent an attractive target to inhibit angiotensin I, and then angiotensin II, production. We therefore

hypothesis, we generated a transgenic mouse line that carries an inducible antisense angiotensinogen gene. Transient inhibition of angiotensinogen synthesis could be demonstrated in these transgenic animals. However, the amounts of liver angiotensinogen message and plasma angiotensinogen concentrations were rapidly back to levels observed in control animals.

The renin-angiotensin system plays a pivotal role in the maintenance of blood pressure, fluid and sodium homeostasis [1]. The glycoprotein angiotensinogen is cleaved by the renin enzyme to generate the inactive angiotensin I peptide. Further cleavage of angiotensin I by the angiotensin converting enzyme (ACE) produces the biologically active hormone, angiotensin II. Although

evaluated the possibility to specifically block the renin-angiotensin

system through a reduction of angiotensinogen synthesis by antisense inhibition. Regulation of gene expression by transcription of an antisense RNA complementary to an endogenous message has first been

described as a naturally occurring repressor mechanism [12]. Expression is inhibited through hybridization of the antisense molecule to the coding sequence by simple base pairing. It has been suggested that the binding could make the mRNA more sensitive to degradation but it is also possible that simple steric angiotensinogen mRNA has been detected in many tissues, blocking could interfere with the subsequent maturation and including the brain and the kidneys, the protein is predominantly translation of the RNA transcript. These observations have been synthetized in hepatocytes and the liver represents the major fully exploited in the antisense oligonucleotide approach to inhibit source of plasma angiotensinogen [2, 3]. Given that the plasma protein synthesis [13]. In parallel, the production of antisense

concentrations of angiotensinogen in rodents are close to the Km RNA's in transfected cells [14, 15] or in transgenic mice [16—19] value of the renin reaction, it is expected that angiotensinogen is has been used successfully to prevent the expression of specific rate limiting in the production of angiotensin I [4]. Therefore, genes in vivo. In this case, an antisense gene is constructed by the small variations in the angiotensinogen plasma levels should adjunction of a particular promoter to a fragment of DNA placed change the activity of the renin-angiotensin system and as a in the opposite direction so that the transcribed antisense moleconsequence blood pressure values. Indeed, it has been shown cule will be complementary to the target messenger RNA. To test the possibility of chronically blocking the renin-angiothat acute infusion of pure angiotensinogen results in a transient tensin system by reducing plasma angiotensinogen concentrations, increase in blood pressure, whereas blockade of angiotensinogen via administration of anti-angiotensinogen antibodies can cause a a transgenic mouse line was generated, harboring an antisense angiotensinogen gene under control of a liver-specific promoter. significant fall in blood pressure [5—7]. The efficacy of ACE inhibitors in the treatment of hypertensive The promoter of the phosphoenolpyruvate carboxykinase (pepck) patients is widely recognized, suggesting that angiotensin II is gene, a key enzyme involved in gluconeogenesis, was chosen for crucial in the pathogenesis of hypertension and in the develop- several reasons. First, although the pepck gene is expressed in a ment of left ventricular hypertrophy [8]. However, because ACE variety of tissues, the fragment of the promoter that was used in inhibitors are not completely selective, there is still a controversy our study contained primarily the cis-acting sequences required whether this is a specific angiotensin-dependent mechanism. for high liver expression [20]. In addition, the level of activity of

Furthermore, since ACE inhibitors do not result in complete the pepck promoter in animals is responsive to diet through a

change of the insulin-to-glucagon ratio [21, 22]. More precisely, a

diet high in protein and devoid of carbohydrate increases the Received for publication October 17, 1994 and in revised form January 9, 1995 Accepted for publication January 9, 1995

© 1995 by the International Society of Nephrology

pepck promoter activity, whereas a marked decrease is observed when the animals are fed a diet high in carbohydrates and low in protein. Finally, the levels of transcription of the pepck gene are low in the liver of fetal animals but increase immediately after

1638

Pedrazzini et al: Antisense angiotensinogen transgenic mice

Al SB SB

B SaBX

1639

1 Kb I

tI pepck enhancer

I

4 anbeonso anglotensinogen cDNA

frglobin IVS2

and promoter

B

a

b

Ca

b

1.1 Kb

Fig. 1. Representation of the antisense angiotensinogen transgene and analysis of transgenic mouse genomic DNA. A. Schematic representation of the 4 Kb

Aat II (A)-Xho I (X) antisense angiotensinogen transgene composed of a 0.9 Kb Sac I (S) fragment of the pepck promoter coupled to the 1.9 Kb angiotensinogen cDNA inserted in reverse position into the Sal I (Sa) of the vector. The IVS2 intron sequences from the rabbit 13-globin gene are also present. Barn HI (B) sites, which are relevant for Southern analysis, are indicated. Position of the two probes used for either Southern analysis (hatched box) or dot blotting (black box) are shown underneath. Arrows indicate the positions of the primers used for the RT-PCR detection of transgene expression. B. Southern analysis of genomic DNA from either a normal (a) or a transgenic (b) mouse. Liver genomic DNA was digested with Barn HI, separated on a 0.7% agarose gel, transferred on nitrocellulose and hybridized with the 0.3 Kb Aat Il-Sac I fragment described above, C. Dot blot analysis

of the F2 progeny. Denatured tail DNA was transferred on nitrocellulose and hybridized sequentially with the 0.9 Kb Sac I fragment of the pepck promoter (a) and a control probe recognizing the renin-1 gene (b).

birth [22]. Therefore, this promoter combines tissue specificity and inductibility and, because of its low activity before birth, minimizes the chance of lethal transgene expression during the fetal life. The pepck promoter was then coupled to an angiotensinogen cDNA in reverse position. The present report describes the effects of the expression of an antisense angiotensinogen transgene on the endogenous levels of angiotensinogen mRNA and protein. Although a significant reduction of angiotensinogen mRNA accumulation and protein synthesis could be demonstrated in transgenic animals, the inhibition was only transient. In adult transgenic mice, the amounts of message and protein were even higher than those measured in normal mice.

with 235 bp of an enhancer element that confers preferential liver

Methods

digest of the pGEM-Angio plasmid (a gift of Dr. Allan R. Brasier, University of Texas, Galveston, TX, USA) [24, 25]. This fragment was introduced into the Sal I site of p6ROSE after linker addition. Appropriate restriction analysis identified a plasmid in which the sequence was subcloned in reverse orientation. This plasmid was named p6ROSE-Angio(-).

Plasmid construction

The plasmid p6ROSE containing the rat pepck promoter was constructed as follows. The fragment of the pepck promoter, composed of 660 bp of regulatory sequences of the pepck gene

expression, was obtained after Hae II/Bgl II digestion of the pPL9DP23 plasmid (a gift of Dr. Roger Chalkley, Vanderbilt University, Nashville, TN, USA) [20]. The pSTX556 expression vector (provided by Dr. Sandro Ruskoni, Institute of Biochemistry, Fribourg, Switzerland) [23], in which the CMV promoter and the rat growth hormone cDNA were removed by Sac I digestion, provided the backbone sequences plus the rabbit 3-g1obin IVS2 and polyadenylation site. The pepck promoter fragment was then ligated to the pSTX556 sequences after Sac I linker addition to produce the p6ROSE vector in which a unique Sal I site is present

immediately downstream to the pepck promoter. The eDNA encoding the rat angiotensinogen was purified from an EcoRI

1640

0.

-D

Brain

Heart

Lung

Kidney

Spleen

Liver

Brain

Heart

Lung

Kidney

Spleen

Liver

pa

ra 0 a-

0pa a V

C

0

Pedrazzini et al: Antisense angiotensinogen transgenic mice

Fig. 2. Tissue-specific expression of the antisense angiotensinogen transgene. A. PCR detection of the antisense angiotensinogen RNA in liver

C,

from either a normal (a) or a transgenic mouse (b). Total RNA was purified from liver, reverse transcribed and subject to specific amplification using primers spanning the angiotensinogen eDNA and the /3-globin sequences (see experimental procedures for exact positions of the primers). The reverse transcription reaction was omitted in lane c and the amplification of transgenic mouse genomic DNA is shown in lane d. B. Tissue-specific expression of the antisense angiotensinogen gene. Total RNA from a normal or a transgenic mouse, isolated from various tissues, was amplified by PCR using transgene-specific primers, and analysed on agarose gel. C. Control amplification of the

Normal

Transgenic

different RNA's using /3-actin-specific primers.

0.5% SDS containing 500 xg/ml proteinase K. DNA was purified by phenol/chloroform extraction and ethanol precipitated. DNA DNA for microinjection was excised from the plasmid was then denatured in 0.125 M NaOH, 0.125 X SSC and transp6ROSE-Angio(-) by Xmn I/Xho I digestion, which removes most ferred on a GeneScreen Plus nitrocellulose membrane (Dupont of the vector sequences. DNA was purified by agarose electroNEM) using a Schleicher and Schuell manifold apparatus. The phoresis and electroelution, precipitated and resuspended at a nitrocellulose filter was then UV cross linked, prehybridized six concentration of 3 j.tg/mI in 5 mivi Tris-HC1 pH: 7.4, 0.1 mM hours and hybridized overnight at 42°C in 50% formamide, 5X EDTA. Fertilized eggs were obtained from superFertilized egg injection

ovulated, six-week-old NMRI female mice (IFFA CREDO,

ssc, lox Denhardt, 50 m phosphate buffer pH: 7.9, 10 mM

EDTA, 1% SDS containing 100 xg/ml denatured salmon sperm pronuclei were injected with the DNA solution at a constant DNA. The Aat IT/Sac I fragment used as probe is described in Figure 1 and was labeled with [a-32PjdCTP (Amersham) using a pressure of 2 psi and viable embryos were reimplanted in the random primed DNA labeling kit (Boehringer). The same condioviduct of pseudopregnant NMRI mice. tions were used to analyze restriction fragments of genomic liver DNA's by Southern blotting except that the digested DNA's were Identification of transgenic mice transferred, after separation in a 0.7% agarose gel, on Hybond-N Transgenic mice were identified first by dot blot analysis of tail filter (Amersham) by capillary blot. The blot was hybridized with DNA. At three weeks of age, 2 cm of tail were cut and digested the [a-32P}dCTP labeled pepck promoter fragment described in overnight at 55°C in 0.7 ml of Tris-HCI pH: 8.0, 100 mM EDTA, Figure 1. L'Arbresles, France) and kept in M2 medium until used. The male

1641

Pedrazzini et al: Antisense angiotensinogen transgenic mice

A 0

z

0.60

0.30

E c

0(Ti cli

0

0.20

0 0.00 (Ti

C Ti)

0)

0C

z

0.07 E 0.06 C

C))

C

0.10

a)

0

0) C

C.)

9,

0

0.05

0.04

C 0.03

Kidney

Fig. 3. Quantitative analysis of angiotensinogen mRNA in normal and transgenic mice. Total RNA, isolated from liver, kidney or brain, from either normal () or transgcnic mice (U) was analyzed by RNase protection assay and the levels of angiotensinogen and /3-actin mRNA were determined. The results are presented as the ratio of angiotensinogen to /3-actin mRNA. * Significant difference compared to normal mouse control, P < 0.005 (SuperANOVA).

a)

0C

0.02

C 0.01 (I)

0

0) C

0.00 0.07 0.06 0.05

Mice

0.04

Adult, either normal or transgenic, male NMRI mice were used at 10 to 14 weeks of age at the beginning of the experiment. The age of newborn mice on the day of sacrifice is indicated.

0.03 0.02 0.01

Reverse-transcription polymerase chain reaction (RT-PCR)

Total RNA was purified from tissues by homogenization of a

frozen sample in 4 M guanidine isothiocyanate, 2% N-lauryl sarcosine, 25 m sodium acetate pH: 6.0, 120 mM J3-mercaptoethanol. The tissue lysate was layered on top of a CsCl cushion composed of 5.7 M CsC1 in 25 m sodium acetate. After centrifugation at 100,000 g for 18 hours, the pellet was resuspended in 0.3 M sodium acetate, 0.1% SDS and ethanol precipitated. Transgene and J3-actin expression was detected by RT-PCR using the RNA GeneAmp kit (Perkin Elmer/Cetus). The conditions were

according to the manufacturer's recommendations. Oligo dT primer was used in the reverse transcription reaction. For the

0.00 Normal diet

Days on high protein diet

Fig. 4. Quantitative analysis of angiotensinogen mRNA in normal and transgenic mice fed a high protein diet. Liver, (A) kidney (B) or brain (C) total RNA, from either normal (ii) or transgenic mice (•) that were kept for various times on high protein diet, was analyzed by RNase protection assay and the levels of angiotensinogen and f3-actin mRNA were determined. The results are presented as the ratio of angiotensinogen to /3-actin mRNA. * Significant difference compared to the normal mouse control, P < 0.005 (SuperANOVA).

PCR amplification, two sets of primers wcre used. The transgene

specific forward primer ASF1 is homologous to the antisense gen rnRNA in tissues. The procedure was as indicated by the angiotensinogen sequence and the backward primer ASB4 to the rabbit 13-globin gene. In addition, two mouse /3-actin specific

primers named BACF (forward) and BACB (backward) were used in control experiments. The oligonucleotide sequences are as follows. ASF1: 5'-ATC TOT GGA CTF GCT TCT GTG TGT C; ASB4: 5'-GAT CTC AGT GGT AlT TGT GAG CCA G; BACF: 5'-TGG CAC CAC ACC TIC TAC AAT GAG; BACB: 5'-GCT TCT CTF TGA TGT CAC GCA CG.

manufacturer. To synthetize the RNA probes, an angiotensinogen template DNA was constructed by insertion of a BarnHI/Bgl II fragment of the mouse angiotensinogen coding sequences spanfling the first exon into the Barn HI site of the pGEM transcription vector (Promega). A Pvu II digestion produced a linear template. The MAXlscript kit from Ambion Inc. was used to synthetize the angiotensinogen RNA probe from the T7 promoter in the pres-

Ribonuclease protection assay (RPA)

ence of [a-32P]dUTP (Amersham). A linear mouse actin transcription vector is included in the kit and the actin-specific RNA probe was generated from the T3 promoter. The hybridization

Total RNA was obtained from tissues as described above. The RPA II kit from Ambion Inc. was used to quantitate angiotensino-

solution contained a molar excess of the two RNA probes together with 10 ig of the RNA sample. Protected fragments

1642

Pedrazzini et a!: Antisense angiotensinogen transgenic mice

A

B

600

600

550

550

C

0 0 C 0

0

500 450

450

400

400

350

350

300

300

()

0) C

Days on high protein diet

Normal diet Days on high protein diet

0 0

C.)

ples were collected, frozen in liquid nitrogen and kept in a —70°C

-1

*

0 0

radioimmunoassay [261.

Measurement of blood pressure and heart rate in conscious mice

o 0

o 0

(0

until used. Aliquots of plasma were diluted in 20 mM phosphate buffer pH: 6.0 containing 30 mivi EDTA and 5 mM phenanthroline. Samples were incubated in the presence of a large excess of semi-purified mouse submaxillary gland renin for two hours at 37°C. Preliminary experiments demonstrated that the amounts of angiotensinogen present in the plasma samples are completely cleaved in these conditions. The concentrations of angiotensin I produced were then determined using a sensitive freezer

-'4

Plasma angiotensinogen concentrations, rIM

Normal diet

Fig. 5. Angiotensinogen concentrations in plasma of transgenic mice. Normal (i) or transgenic mice were kept for different times on high protein diet and plasma angiotensinogen concentrations were determined by indirect radioimmunoassay following generation of angiotensin I. Results represent the mean plus standard error of five to seven mice per group. * Significant difference compared to the control group fed a normal chow, P < 0.005. • Significant difference compared to the normal mouse group, P < 0.005 (SuperANOVA).

Mice were anesthetized (3% halothane in oxygen) and a catheter, formed by a piece of stretched PE-lO tubing, was

Day after birth Fig. 6. Angiotensinogen concentrations in plasma of newbom mice. Angio-

tensinogen concentrations were determined by indirect radioimmunoassay in the plasma of normal (fl) or transgenic (•) newborn mice different days after birth. Results represent the mean plus standard error of ten mice per group. * Significant difference compared to the age-matched controls, P < 0.005 (SuperANOVA).

were separated on a 6% polyacrylamide gel. The radioactivity associated with each band was accurately determined using an Instant Imager detector (Packard Instruments, Meriden, CT, USA). Results are expressed as the ratio of the angiotensinogen to the actin mRNA.

introduced into the femoral artery. The catheter was then passed subcutaneously to exit at the neck. The next day, pulsatile arterial pressure and heart rate were determined during a one hour period by the mean of a pressure transducer connected to a physiograph recorder. Results Generation of a transgenic mouse line harboring an antisense angiotensinogen gene In order to test the possibility of blocking the renin-angiotensin

system through an inhibition of angiotensin I production, we decided to use the antisense approach to specifically inhibit angiotensinogen synthesis in transgenic animals. A transgene, composed of a fragment of the pepck gene promoter fused to a full-length angiotensinogen eDNA placed in reverse position, was microinjected into fertilized eggs. A schematic representation of

the transgene is depicted in Figure 1A. Transgenic mice were identified by dot blot analysis of tail DNA using a Aat TI/Sac I Angiotensinogen concentrations were measured in plasma us- probe (Fig. 1). This report describes the characterization of the ing an indirect method. Briefly, blood was drawn from the Tsg(Angio-)4 transgenic line which results from the breeding of a retro-orbital vein (adult mice) or harvested following decapitation male founder bearing ten copies of the antisense transgene. (newborn mice), and EDTA was immediately added to reach a Figure lB shows a Southern blot analysis of BamHI restriction final concentration of 10 m. After centrifugation, plasma sam- fragments of liver genomic DNA's from normal and transgenic Plasma angiotensinogen assay

Pedrazzini et al: Antisense angiotensinogen transgenic mice

116 —

B 800 780 760 -

114 .

740 -

112 —

720

110 —

70o. 680

A 120 —

118 H

108 —

106-

Fig. 7. Blood pressure and heart rate in

660

104 —

640-

102 —

620 —

Normal Transgenic

1643

Normal Transgenic

conscious mice. Normal (E) or transgenic (•) mice were anesthesized and a catheter was introduced in the femoral artery to the abdominal aorta. The next day, blood pressure (A) and heart rate (B) were recorded via a pressure transducer during a one-hour period. The results represent the mean plus standard error of eight mice per group.

mice. The purified pepck promoter segment was used as probe. specific expression of the antisense gene. RNA samples from The endogenous mouse pepck gene is detected as a high molec- different tissues were submitted to analysis and the results are ular weight band (approximately 10 kb) in the genomic DNA from presented in Figure 2B. No transgene expression could be obeither normal or transgenic mice. However, the two transgene- served in tissues from a normal mouse. In contrast, high liver specific bands of respectively 1.1 and 0.9 kb were found in the expression was manifest in the transgenic mouse. In addition, DNA of the founder offspring only. The 0.9 kb band is likely to be significant expression could be seen in the brain of transgenic

generated by the presence of a BamHI site upstream of the transgene integration site. The transmission to progeny was followed by dot blot hybridization of tail DNA's. Figure 1C shows

the result of a representative experiment that identifies normal, heterozygote and homozygote mice as judged by the intensity of the hybridization signal. The proportion of mice that segregate in transgene-negatives (29.2%), heterozygotes (43.2%) and homozygotes (27.6%), as well as the equal separation of the progeny between male and female (52.3% vs. 47.7%), suggests an autosomal Mendelian type of inheritance (total number of mice tested = 65). No obvious phenotypic differences could be observed in the transgenic population and the development appeared normal. Antisense gene expression

To facilitate the detection of the antisense RNA molecule,

animals. Finally, the result of the amplification of the mouse /3-actin mRNA from the different tissue is presented in Figure 2C, demonstrating the presence of RNA in each sample. Levels of angiotensinogen message in transgenic animals

The liver, the brain and the kidney have been described as containing significant amount of angiotensinogen mRNA. Since the antisense gene appeared to be expressed in detectable amount in the liver and in the brain of transgenic animals, the effect on the endogenous angiotensinogen message was investigated in these tissues. We also tested the effect on the angiotensinogen mRNA concentrations in the kidney since, in some experiments, a faint expression of the transgene was detected in this tissue (data not

shown), and also because the pepck promoter was previously reported to direct the expression in the kidney [21]. Figure 3

transgene expression was followed using RT-PCR. The transgenespecific forward primer was selected in the antisense angiotensinogen sequence whereas the backward primer was complementary

shows the levels of angiotensinogen message detected by quantitative RNAse protection assay in the liver, brain and kidney of

to the /3-globin gene. To exclude the possible amplification of contaminating genomic DNA, the primers were selected so that they were separated by the /3-globin intron sequences. As shown in Figure 2A, the amplification of the expected 220 bp fragment, which demonstrates the presence of the antisense RNA, was evident only in the amplified products from the liver RNA of a transgenic animal (lane b). No fragment could be seen when liver RNA from normal mice was tested (lane a) or when the reverse transcription reaction was omitted before amplification of RNA from a transgenic mouse (lane c). In addition, Figure 2A shows the 820 bp product obtained after amplification of genomic DNA from a transgenic mouse (lane d). The product size is augmented by the 600 bp of the intron sequences. The RT-PCR method was then used to determine the tissue-

published observations, the amounts of mRNA in the kidney and in the brain of normal animals were approximately a fifth of those observed in the liver. Surprisingly, the amount of angiotensinogen

either normal or transgenic mice. In accordance to previously

message in the liver of transgenic mice was not significantly diminished but was even increased as compared to normal animals. A similar result was observed in the brain whereas the kidney showed a slight decrease in the mRNA concentrations (Figs. 3 and 4). Levels of angiotensinogen mRNA in mice fed a high protein diet

Since the pepck promoter is responsive to diet, the animals were fed a regimen devoid of carbohydrate (high protein diet) for

different times and the amounts of angiotensinogen mRNA in tissues were determined (Fig. 4). The first striking observation was

1644

Pedrazzini et a!: Antisense angiotensinogen transgenic mice

a marked increase in the levels of the liver angiotensinogen long period of a high protein diet were ineffective in changing mRNA in normal mice kept on a high protein diet as compared to

normal animals fed a normal chow. This increased amount of message in the liver seemed constant during the entire course of the experiment. In contrast, the effect of the antisense expression

in transgenic animals appeared to blunt the angiotensinogen mRNA accumulation during the first twenty days. However, the level of message in the liver gradually increased to reach values observed in normal animals around thirty days after starting the diet. The situation in the kidney very much reflected that seen in the liver. On the contrary, the high protein diet did not seem to induce any reduction of the levels of angiotensinogen message in the brain. Plasma angiotensinogen concentrations in normal vs. transgenic adult mice

these parameters in transgenic animals (data not shown). Discussion In this report, we describe a transgenic mouse line that carries an antisense angiotensinogen gene. Transient inhibition of angiotensinogen mRNA accumulation in tissues was demonstrated in

this model. In addition, significant reduction of the levels of plasma angiotensinogen protein was also observed in transgenic animals in which the expression of the transgene was suddenly activated. However, a long-term blockade of the renin-angiotensin system appears difficult using this approach. The antisense gene used in this study is composed of the rat pepck gene promoter coupled to a full-length rat angiotensinogen cDNA. The species difference does not represent a major obstacle

since the mouse and the rat sequences share long stretches of Plasma angiotensinogen concentrations were also determined complete homology, with an overall identity of more than 90% in normal and transgenic mice (Fig. 5). As expected from the [25, 27, 28], and more importantly, because this combination was amounts of mRNA found in the liver, the angiotensinogen plasma successfully used to inhibit angiotensinogen production by hepalevels were significantly higher in transgenic animals as compared toma cells that were stably transfected with an antisense angioto normal mice. Then, normal and transgenic mice were kept for tensinogen gene [15]. In transgenic animals, the fragment of the

different times on high protein diet before measuring angiotensinogen concentrations. Similarly to what was seen at the mRNA level, a high protein diet gradually increased angiotensinogen concentrations in the plasma of normal animals. Angiotensinogen accumulation peaked 80 days after starting the diet. In contrast, the regimen in transgenic mice led to a slight reduction in plasma angiotensinogen levels. Although modest, this decrease was significant after thirty days on a high protein diet

(P < 0.005). Plasma angiotensinogen concentrations in normal versus transgenic newborn mice

The compensation that operates in transgenic animals seems to take place over a period of several weeks. We therefore postulate that a sudden switch from no expression to high expression of the

antisense transgene should facilitate the demonstration of the antisense inhibition. As already mentioned, the pepck gene promoter is not active during the fetal life but the expression reaches

its maximum just after birth. We decided then to measure the plasma angiotensinogen concentrations in either normal or trans-

genic newborn animals. Immediately after birth, normal mice showed a strong rise in their plasma angiotensinogen levels (Fig. 6). A gradual reduction was then observed and plasma angiotensinogen values approached those measured in adult animals. In transgenic animals, this peak was very much reduced as the result of the antisense expression. However, the difference in plasma angiotensinogen concentrations between control and

pepck promoter has been shown to direct the expression of

different protein to liver and kidney tissues [21]. In the present work, the antisense angiotensinogen transgene expression was detected in the liver, the brain and possibly the kidney. Expression

in the central nervous system under the control of the pepck promoter was not reported previously. It is possible that, depend-

ing on the site of transgene integration in the mouse genome, leaky expression could emerge. Although transgene expression

could be readily demonstrated using a PCR assisted technique, we were not able to detect the presence of the antisense RNA using the quantitative RNA protection assay. The reason might be that a competition binding to the antisense RNA takes place between the endogenous angiotensinogen mRNA and the RNA probe, both of them being complementary to the antisense molecule, during the hybridization step of the ribonuclease assay. The amounts of angiotensinogen mRNA were then determined in the different tissues that express the transgene. Surprisingly, no dramatic reduction in the levels of angiotensinogen mRNA was observed in transgenic mice. In fact, the levels of angiotensinogen message appeared even increased in the liver of transgenic mice whereas changes in the brain or in the kidney were marginal (Fig.

3). Although expression of the antisense transgene is barely detectable in the kidneys, the pepck promoter has been shown to be active in this tissue [21]. This might explain the small reduction

in the levels of renal angiotensinogen mRNA. The fact that the brain levels appeared slightly augmented could be a reflection of the different regulation of the angiotensinogen gene in different tissue [2]. Correspondingly to what was seen at the mRNA level,

transgenic animals progressively diminished, and the values mea- the plasma angiotensinogen was also increased in transgenic sured in transgenic mice even exceeded normal levels by day 5 animals fed a normal diet (Fig. 5). The amount of message in the after birth. Thereafter, the difference between the values of the kidney and in the brain represents approximately 20% of that transgenic versus the normal group was similar to that observed in found in the liver (Fig. 3) [2, 3]. Moreover, a parallel increase in adult animals. the amounts of hepatic angiotensinogen mRNA and the levels of the plasma protein was observed between normal and transgenic Blood pressure and heart rate in normal versus transgenic mice animals (Figs. 3 and 5). It is therefore unlikely that the kidney or Blood pressure and heart rate were also examined in normal the brain constituted a significant source that would be responsiand transgenic mice and the results of a representative experiment ble for the increase in the angiotensinogen plasma levels observed are depicted in Figure 7. No difference in these two parameters in transgenic animals. could be observed between the two groups of animals. Short or To boost the expression of the antisense transgene, mice were

1645

Pedrazzini et al: Ant isense angiotensinogen transgenic mice

fed a high protein diet [21, 22]. This diet also dramatically RNA's. Although the cellular approach has proved to be successincreased the amount of endogenous angiotensinogen message in the liver of normal mice, although it did not change the levels of angiotensinogen mRNA in the brain or in the kidney as previously described by others (Fig. 4) [29, 30]. The possible reason for this liver-specific increase is unknown, but differential regulation of angiotensinogen production in various tissues has been reported.

For instance, glucocorticoids, estrogens and thyroid hormones

ful [14, 15], targeted antisense expression in transgenic animals led to mixed results [161. We were indeed able to show inhibition of angiotensinogen message and protein synthesis in a transgenic mouse line that carries an antisense angiotensinogen gene. However, the antisense inhibitory effect was only transient because a compensation took place that brought the angiotensinogen message and plasma concentrations back to control values. Renin

modify differently liver, kidney and brain angiotensinogen mRNA expression in the kidneys and plasma renin activity are not

production [2, 31, 32]. In addition, insulin has been shown to different in the two group of mice (data not shown). As a negatively regulate angiotensinogen expression in liver-derived consequence, it is not surprising that blood pressure and heart cells [33]. On a high protein diet, the levels of circulating insulin are dramatically reduced. It is therefore possible that the absence of this potentially negative signal resulted in an increased transcription of the angiotensinogen gene.

The apparent discrepancy between the level of reduction

observed in the angiotensinogen message and that seen in plasma protein in transgenic animals could reflect decreased steady-state

levels of angiotensinogen message in the face of an increased mRNA turnover in response to the antisense inhibition. Alternatively, one cannot exclude the possibility that different tissues could contribute to the plasma angiotensinogen concentrations upon antisense activation. However, it is unlikely that the kidney would play a major role since a reduction of angiotensinogen message similar to that seen in the liver, was observed in this tissue (Fig. 4). The amounts of brain angiotensinogen mRNA on the other hand did not appear to be significantly diminished by the expression of the antisense transgene, even when the animals were placed on a high protein diet. On the contrary, the levels of brain

message seemed slightly augmented compared to control. Whether or not this difference could explain the 20% rise in plasma levels remain to be elucidated. Other tissues have been described to transcribe the angiotensinogen gene. However, the presence of angiotensinogen message in the heart, the lung or the spleen was not detected (data not shown). Finally, a high protein diet has been shown to activate plasma renin activity [34, 35], which in turn could increase circulating concentrations of angiotensin II. This peptide is thought to increase angiotensinogen gene transcription and to therefore participate in a feedback mechanism that restores angiotensinogen plasma levels [36—38].

The antisense inhibition was further documented in newborn animals in which a sudden switch from no activity of the pepck promoter (and therefore of transgene transcription) to maximum activity naturally occurs immediately after birth [22]. The study of

the antisense effect appeared also facilitated by the rise in the amount of liver angiotensinogen message as well as in the plasma angiotensinogen concentrations that occur during the first postnatal days [39—41]. This enhanced angiotensinogen expression is probably secondary to an increase in plasma steroid and thyroid

hormones, which are known to rise after birth and stimulate angiotensinogen gene transcription [2]. A clear decrease of the angiotensinogen plasma levels was then observed in two- to four-day-old transgenic mice compared to normal animals. Again,

the antisense inhibition was overcome by a compensatory phenomenon that was even able to increase plasma concentrations five days after birth. At that time, the difference between normal control and transgenic mice was similar to that seen in adult animals.

Different attempts have been made to partially knock out gene expression in transfected cells or in transgenic mice with antisense

rates do not appear to be modified by the presence of the transgene (Fig. 7). Finally, the exact molecular mechanisms that

account for the compensatory mechanism remain to be elucidated, but it is of interest that the levels of angiotensinogen protein in transgenic mice were remarkably stable over time. The antisense RNA might have introduced an extra level of regulation. Acknowledgments This study was supported in part by grant from the Swiss National Science Foundation (grant no. 3200—033741) and by the Emma Muschamp Foundation, Switzerland. We are grateful to Dr. H. van der Putten

for the introduction to the transgenic technology. We thank Dr. J. Nussberger for his help in establishing the angiotensinogen assay. Reprint requests to Thieriy Pedrazzini, Ph.D., Division of Hypertension, Lausanne University Medical School, CH-1OJ1 Lausanne, Switzerland.

References 1. ROBERTSON JIS: Renin and angiotensin: A historical review, in The Renin-Angiotensin System, edited by ROBERTSON JS, NICHOLLS M, London, Gower Medical Publishing, 1993, p 1.1 2. CAMPBELL DJ, HABENER JE: Angiotensinogen gene is expressed and

differentially regulated in multiple tissue of the rat. J Clin Invest 78:31—39, 1986 3. OHKUBO H, NAKAYAMA K, TANAKA T, NAKANISHI S: Tissue distribu-

tion of rat angiotensinogen mRNA and structural analysis of its heterogeneity. J Biol Chem 261:319—323, 1986 4. REID IA, MORRIS BJ, GANONG WJ: The renin-angiotensin system. Ann Rev Physiol 40:377—410, 1978 5. MENARD J, EL AMRANI AIK, SAv0IB F, BOUHNIK J: Angiotensinogen:

An attractive and underrated participant in hypertension and inflammation. Hypertension 18:705—706, 1991 6. WEAVER D, SKINNER 5, WALKER L, SANGSTER M: Phenotipic inhibi-

tion of the renin-angiotensin system, emergence of the ren-2 gene, and adaptive radiation of mice. Gen Adap Endocrinol 83:306—315, 1991 7. GARDES J, BOUHNIK J, CLAUSER E, CORVOL P, MENARD J: Role of

angiotensinogen in blood pressure homeostasis. Hypertension 4:185— 189, 1982 8. HANNSON L, DAHLOF B, HIMMELMANN A, SVENSSON A: Angiotensin-

converting enzyme inhibitors in the treatment of essential hypertension, in The Renin-Angiotensin System, edited by ROBERTSON JS, NIcHoLLS M, London, Gower Medical Publishing, 1993, p 91.1 9. NUSSBERGER J, DELABAYS A, DE GASPARO M, CUMIN F, WAEBER B,

BRUNNER HR, MEND J: Hemodynamic and biochemical consequences of renin inhibition by infusion of CGP 38560A in normal volunteers. Hypertension 13:948—953, 1989 10. CAMENZIND E, NUSSBERGER J, JUILLERAT L, MUNAFO A, FISCULI W, CoAssoLo P, VAN BRUMMELEN P, WAEBER B, BRUNNER HR: Effect of

the renin response during renin inhibition; oral RO-425892 in normal humans. J Cardiovasc Pharmacol 18:299—307, 1991 11. ZUSMAN RM, HUI KY, NUSSBERGER J, CHRISTENSEN DM, HIGGINs J,

CARLSON WD, SCHOENFELD D, HABER E: R-PEP-27, a potent renin

inhibitor, decreases plasma angiotensin II and blood pressure in normal volunteers. Am J Hypertens 7:295—301, 1994 12. GREEN PJ, PINES 0, INOUYE M: The role of antisense RNA in gene regulation. Ann Rev Biochem 55:569—597, 1986

1646

Pedrazzini et a!: Antisense angiotensinogen transgenic mice

13. SmIN CA, COHEN JS: Oligodeoxynucleotides as inhibitors of gene

ular cloning of the mouse angiotensinogen gene. Genomics 2:240—248,

expression: A review. Cancer Res 48:2659—2668, 1988 14. STOUT JT, CASKEY CT: Antisense RNA inhibition of HPRT synthesis. Somat Cell Mol Genet 16:369—382, 1990 15. CLOUSTON WM, LLOYD Cl, RICHARDS RI: Inducible antisense RNA for angiotensinogen in stably transformed hepatoma cell lines. J Mo! Endocrinol 4:107—117, 1990

1988 28. TAMURA K, TANIM0T0 K, TAKAHASHI 5, SAGARA M, FUKAMIZU A,

16. MUNIR MI, ROSSITER IF, CASKEY CT: Antisense production in transgenic mice. Somat Cell Mo! Genet 16:383—394, 1990 17. HAN L, YUN JS, WAGNER TE: Inhibition of Moloney murine leukemia virus-induced leukemia in transgenic mice expressing antisense RNA complementary to the retroviral packaging sequences. Proc NatlAcad Sci USA 88:4313—4317, 1991 18. PEPIN M-C, POTHIER F, BARDEN N: Impaired type II glucocorticoid-

receptor function in mice bearing antisense RNA transgene. Nature 355:725—728, 1992 19. KATSUKI M, SATO M, KJMURA M, YOKOYAMA M, KOBAYASHI K, NOMURA T: Conversion of normal behavior to shiverer by myelin basic protein antisense cDNA in transgenic mice. Science 241:593—595, 1988 20. I YT, PooN D, STONE D, GRANNER DK, CHALKEY R: Interaction of

a liver-specific factor with an enhancer 4.8 kilobases upstream of the phosphoenolpyruvate carboxykinase gene. Mo! Cell Biol 10:3770— 3781, 1990 21. SHORT MK, CLOUTHIER DE, SCHAEFFER IM, HAMMER RE, MAGNU-

SON MA, BEALE EG: Tissue-specific, developmental, hormonal, and dietary regulation of rat phosphoenolpyruvate carboxykinase-human

growth hormone fusion genes in transgenic mice. Mo! Cell Biol 12:1007—1020, 1992 22. MCGRANE MM, YUN JS, PATEL YM, HANSON RW: Metabolic control

of gene expression in vivo studies with transgenic mice. TIBS 17:40— 44, 1992 23. SEVERNE Y, WIELAND S, SCHAFFNER W, RUSCONI 5: Metal binding

finger structures in the glucocorticoid receptor defined by sitedirected mutagenesis. EMBO J 7:2503—2508, 1988 24. BRASIER AR, PHILIPPE J, CAMPBELL DJ, HABENER JF: Novel expres-

sion of the angiotensinogen gene in a rat pancreatic islet cell line. Transcriptional regulation by glucocorticoids. JBiol Chem 261:16148— 16154, 1986 25. OHKUBO H, KAGEYAMA R, UJIHARA M, HIROSE T, INAYAMA S, NAKANISHI S: Cloning and sequence analysis of eDNA for rat angiotensinogen. Proc Nat! Acad Sci USA 80:2196—2200, 1983

26. NUSSBERGER J, BRUNNER HR: Measurement of angiotensins in plasma, in The Renin-Angiotensin System, edited by ROBERTSON JS, NICHOLLS M, London, Gower Medical Publishing, 1993, p 15.1 27. CLOUSTON WM, EvANS BA, HARALAMBIDIS J, RICHARDS RI: Molec-

MURAKAMI K: Structure and expression of the mouse angiotensinogen gene. Jpn HeartJ 33:113—124, 1992 29. ROSENBERG ME, CHMIELEWSKI D, HOSTETTER TH: Effect of dietary

protein on rat renin and angiotensinogen gene expression. J Clin Invest 85:1144—1149, 1990

30. CORREA-ROrrER R, HOSTETFER TH, ROSENBERG ME: Effect of

dietary protein on renin and angiotensinogen gene expression after renal ablation. Am J Physiol 262:F631—F638, 1992 31. HONG-BROWN LQ, DESCHEPPER CF: Regulation of the angiotensino-

gen gene by estrogens in rat liver and different brain regions. P S E M B 203:467—473, 1993 32. HONG-BROWN LQ, DESCHEPPER CF: Effects of thyroid hormones on angiotensinogen gene expression in rat liver, brain, and cultured cells. Endocrinology 130:1231—1237, 1992 33. CI-w'tc E, PERLMAN AJ: Angiotensinogen mRNA. Regulation by cell cycle and growth factors. J Biol Chem 263:5480—5484, 1988

34. DANIELS BS, HOSTETFER TH: Effects of diatery protein intake on vasoactive hormones. Am J Physiol 258:R1095—R1100, 1990 35. ROSENBERG ME, SwANsoN JE, THOMAS BL, HOSTETrER TH: Glomer-

ular and hormonal responses to dietary protein intake in human renal disease. Am J Physiol 253:F1083—F1090, 1987 36. NAKAMURA A, IWAO H, FUKUI K, KIMURA 5, TAMAKI T, NAKANISHI 5,

ABE Y: Regulation of liver angiotensinogen and kidney renin mRNA levels by angiotensin II. Am J Physiol 258:E1—E6, 1990 37. SCHUNKERT H, INGELFINGER JR, JACOB H, JACKSON B, BOUYOUNES B,

DZAU VJ: Reciprocal feedback regulation of kidney angiotensinogen and renin mRNA expressions by angiotensin II. Am J Physiol 263: E863—E869, 1992 38. KLETF C, NOBILING R, GIERSCHIK P, HACKENTHAL E: Angiotensin II

stimulates the synthesis of angiotensinogen in hepatocytes by inhibiting adenylylcyclase activity and stabilizing angiotensinogen mRNA. J Biol Chem 268:25095—25107, 1993

39. KALINYAK JE, HOFFMAN AR, PERLMAN AJ: Ontogeny of angiotensinogen mRNA and angiotensinli receptors in rat brain and liver. J Endocrinol Invest 14:647—653, 1991 40. GOMEZ AR, CASSI5 L, LYNCH KR, CHEVALIER RL, WILFONG N, CAREY RM, PEACH MJ: Fetal expression of the angiotensinogen gene. Endocrinology 123:2298—2302, 1988 41. JELINEK J, HACKENTHAL R, HILGENFELDT U, SCHAECHTELIN G, HACK-

ENTHAL E: The renin-angiotensin system in the perinatal period in rats. J Develop Physiol 8:33—41, 1986