(Received for publication, January 22, 1997)
From the Medical Research Council of Canada
Multidisciplinary Research Group on Hypertension, Clinical Research
Institute of Montreal, Montreal, Quebec, Canada H2W 1R7 and
¶ Hypertension and Vascular Research Division, Henry Ford
Health Sciences Center, Detroit, Michigan 48202
Angiotensin II (Ang II) is an octapeptide generated by the sequential proteolytic action of renin and angiotensin converting enzyme on the glycoprotein angiotensinogen. While numerous mammalian tissues have been shown to express some or all of the components of the renin-angiotensin system (RAS), the function of most of these tissue RAS remains a matter of conjecture. To test for tissue-specific functions of Ang II and as an alternative to co-expressing all the components of RAS, we have engineered a fusion protein that leads to direct Ang II release within specific tissues. The angiotensin peptide is cleaved from the fusion protein within the secretory pathway by the ubiquitous endoprotease furin and is released from the cell by constitutive secretion. Direct injection of an expression vector encoding such a fusion protein into rat cardiac ventricles results in a highly localized expression of atrial natriuretic peptide mRNA (an angiotensin responsive marker of cardiac hypertrophy), demonstrating the utility of this approach for local targeting of mature peptides to tissues in animal models.
Angiotensin II (Ang II)1 is the peptide product of the renin-angiotensin system (RAS) and is generated via two sequential proteolytic steps. First, renin, a circulating aspartyl protease, cleaves the decapeptide angiotensin I (Ang I) from the amino terminus of the hepatic glycoprotein angiotensinogen. Angiotensin-converting enzyme then removes two amino acids from the carboxyl terminus of Ang I to release the vasoactive peptide Ang II. This enzymatic cascade that occurs in the circulation has also been suggested to take place within certain tissues. Brain, kidney, adrenal and pituitary glands, heart, vasculature and reproductive tissues have been shown to express protein and/or mRNA for many of the components of the RAS including angiotensin receptors (1-6). Activity of local or tissue RAS has been implicated in a variety of physiologic pathways and pathophysiologic conditions including sympathetic nerve transmission, pituitary hormone secretion, migration of eggs in the oviduct, renal development, hypertension, end-stage renal disease, cardiac hypertrophy, and restenosis following vascular injury (reviewed in Refs. 7-10). However, two findings make it difficult to differentiate between the possibilities that the Ang II mediating these activities is either generated and acts locally, comes from the circulation, or is synthesized locally and acts elsewhere to generate the observed effects. First, in tissues, components of the RAS are often expressed in different cell types or are present in only extremely low levels, making it difficult to be certain that all of the necessary components would encounter each other in biologically relevant concentrations. Second, renin is synthesized as a zymogen that is activated before secretion from the juxtaglomerular cells of the kidney. Removal of the kidneys results in virtual disappearance of renin from the circulation, whereas its precursor, prorenin, remains (11-14). This finding has raised the question about whether non-renal tissues have the capacity to activate prorenin and thereby carry out the first reaction in the RAS. For these reasons and despite much circumstantial evidence, the functions of tissue RAS remain a matter of conjecture.
By using transgenic animals, it is possible to test for the biological effect of tissue expression of the RAS by inducing either loss of function or gain of function mutations. Mice lacking all RAS activity have been generated via homologous recombination by insertional mutagenesis of the angiotensinogen gene (the only known substrate for generation of the angiotensin peptides) (15, 16). These mice are hypotensive and exhibit some defects in the development of the kidney, but these experiments shed little light on the normal physiologic functions of tissue RAS. On the other hand, generalized overexpression of components of the RAS in transgenic mice and rats is clearly linked to an increase in blood pressure (17-20). To discriminate between circulatory and local effects of transgene expression, it is necessary to target expression of the RAS transgenes to specific tissues. However, this approach is complicated by the fact that all of the RAS components (i.e. angiotensinogen, prorenin, angiotensin-converting enzyme, Ang II receptors, and a prorenin convertase) need to be expressed in the target tissue. In addition, the Ang II generated can elicit biological responses at other sites.
As an alternative to co-expressing RAS components, we have engineered a
fusion protein that leads to the direct release of an Ang II peptide
within specific tissues. This peptide is a natural variant of Ang II
found in the skin of the Australian frog Crinia georgiana
(fsAngII (21)) which has been reported to have pressor activity similar
to that of mammalian Ang II (22). Its amino-terminal alanine residue
presents a favorable P1 cleavage site for proteolytic release of
fsAngII from the fusion protein by the ubiquitous protease furin (23).
In this study, the receptor binding and biological activity of the
fsAngII peptide were characterized, and experiments were performed to
test for the effect of local over-expression of fsAngII in
vivo. Direct injection of the expression vector into the apex of
rat hearts leads to local expression of fsAngII and induction of atrial
natriuretic peptide (ANP) expression, a biochemical molecular marker of
cardiac hypertrophy in adult cardiac ventricles (24, 25). These results
demonstrate for the first time the feasibility of using an engineered
fusion protein to deliver a peptide with local biological activity in
whole animals.
Expression vector pIg contains the signal peptide of human
prorenin followed by a fragment of the heavy chain constant region of
mouse IgG2b. The Ig fragment encodes 221 amino acids from the CH2 and
CH3 domains and lacks the BIP binding, hinge, and intermolecular disulfide bridge regions (26) while maintaining the protein A binding
site. Construction of this plasmid is described elsewhere (27). This
construct is used as a negative control (Fig. 1A).
The expression vector pIgPfsAngII (Fig. 1B)
contains the human prorenin signal peptide and Ig fragment from pIg
linked to a portion of the human prorenin prosegment followed by a
furin consensus cleavage site and the coding sequence of the fsAngII peptide. pIgPfsAngII was constructed as follows. A cDNA fragment encoding the signal peptide of human prorenin following the Ig
fragment was amplified from pIg by polymerase chain reaction. The
sequences of the forward and reverse primers were
5-CGCCATTTGACCATTCAC-3
and 5
-AAGGCCTACCCGGAGACC-3
. A
StuI restriction site in the reverse primer is underlined. A
portion of the human prorenin cDNA encoding the first 35 amino
acids of the prosegment was generated by polymerase chain reaction from
the expression vector pRhR1100 (28) using the following primers:
forward, 5
-AAGGCCTCTCCCGACAGAC-3
; reverse, 5
-GAAGATCTGGGACCAAGCCTG-3
(StuI and
BglII restriction sites in the amplifying oligonucleotides
are underlined). The StuI-BglII restriction
fragment containing the human prorenin prosegment was ligated to
oligonucleotides containing a 5
BglII site, a sequence
encoding the furin consensus cleavage site (RVRTKR (29, 30)), the
coding sequence for fsAngII (APGDRIYVHPF), a stop codon, and a 3
EcoRI site and inserted at the 3
end of the Ig coding sequence.
The forward oligonucleotide sequence was
5
-GAAGATCTCGCGTACGCACTAAACGCGCCCCCGGGGACCGCATCTACGTACACCCCTTTTGAGAATTCC-3
. The entire coding sequence was placed in the expression vector Rous sarcoma virus-globin where it is under the transcriptional control
of the Rous sarcoma virus-long terminal repeat promoter/enhancer. Transcribed RNA is provided with an intron and polyadenylation signal
from a portion of the rabbit
-globin gene placed 3
of the fusion
protein coding sequence (28).
The expression vector pIgfsAngII (Fig. 1C) is constructed
as pIgPfsAngII except that the portion of the human
prorenin cDNA encoding the first 35 amino acids of the prosegment
is missing, and the 3 splice and polyadenylation signal comes from the
simian virus 40 T-antigen. All recombinant plasmid constructions were verified by DNA sequencing.
To test for binding to the Ang II AT1 receptor, COS-1 cells were transfected with an expression vector for rat AT1b. COS-1 cells were plated in 150-mm Petri dishes and grown to confluence in Dulbecco's modified Eagle's medium supplemented with 5% (v/v) fetal calf serum and gentamicin at 10 µg/ml (Life Technologies, Inc.) in a humidified incubator (5% CO2, 95% air) at 37 °C. The cells were transfected by the DEAE-dextran method (31) using a rat AT1b cDNA ((32) a gift from Dr. Tadashi Inagami) inserted into the mammalian expression vector pCMV5 (a gift from Dr. David Garbers). Cells were collected 48 h after transfection, and the membrane fraction was prepared for AT1 receptor binding assays.
To test for fsAngII binding to the AT2 receptor, PC12W cells, which naturally express only the AT2 receptor, were plated in 150-mm Petri dishes and grown to confluence in a mixture of Dulbecco's modified Eagle's medium supplemented with 5% (v/v.) fetal calf serum, 10% heat-inactivated horse serum, and gentamicin at 10 µg/ml (Life Technologies, Inc.) in a humidified incubator (5% CO2, 95% air) at 37 °C.
Membranes were prepared from transfected COS-1 cells or from confluent cultures of PC12W cells by homogenization and centrifugation (33). Radioligand binding assays to Ang II type 1 (AT1 from transfected COS cells) or type 2 (AT2 from PC12W cells) receptors were carried out as described previously using 125I-labeled [Sar1,Ile8]Ang II as a ligand (34).
Biological Activity of fsAngIIFor intracellular free
calcium concentration ([Ca2+]i) measurements,
COS-1 cells transfected with the rat AT1b receptor were
re-plated 24 h after transfection onto laminin-coated round glass
coverslips (25 mm diameter) in multi-well dishes.
[Ca2+]i measurements were performed 24 h
after re-plating using the ratiometric fluorescent dye fura-2AM
according to previously described methods (35, 36). Briefly, cells were
washed three times with 2 ml of modified Hanks' buffered saline
solution containing (mmol/liter) 137 NaCl, 4.2 NaHCO3, 3 Na2HPO4, 5.4 KCl, 0.4 KH2PO4, 1.3 CaCl2, 0.5 MgCl2, 0.8 MgSO4, 10 glucose, 5 Hepes (pH 7.4). The washed cells were loaded with fura-2AM (4 µmol/liter) (dissolved in dimethyl sulfoxide with 0.02% pluronic F-127) and incubated for 25 min at 37 °C. The loaded cells were washed three times with the
Hanks' buffer and allowed to equilibrate at room temperature for 15 min. [Ca2+]i was measured by fluorescent digital
imaging using the Axiovert 135 inverted microscope and Attofluor
Digital fluorescence System (Zeiss, Germany) using double excitatory
wavelengths of 343 and 380 nm. Video images of fluorescence at 520-nm
emission were obtained using an intensified CCD camera system (Zeiss,
Germany) with the output digitized to a resolution of 512 × 480 pixels. Fluorescence ratios were obtained by dividing, pixel by pixel, the 343-nm image after background subtraction of the 380-nm image. Both
in vitro and in situ calibrations were performed
according to our previously described methods (36, 37). Effects of
varying concentrations of Ang II and fsAngII
(1012-10
5 mol/liter) were determined in
multiple cells. Cells were exposed to one concentration only, and
repetitive determinations were not made. The maximum peak ratio
recorded was considered as the maximal response of the agonist.
Rat pituitary GH4C1 cells were grown in a mixture of Dulbecco's modified Eagle's medium supplemented with 5% (v/v) fetal calf serum, 0.1% serXtend (Irvine Scientific, CA), and gentamicin at 10 µg/ml (Life Technologies, Inc.) in a humidified incubator (5% CO2, 95% air) at 37 °C. GH4C1 cells, plated at 8 × 105 cells per 35-mm dish, were transfected 20 h after plating with Lipofectin (Life Technologies, Inc.) in serum-free medium using 20 µg of plasmid DNA per dish. The cells were transfected with an expression vector for fsAngII (pIgPfsAngII) or its control counterpart pIg (Fig. 1). After 48 h, cells were transferred to 12-well plates at a concentration of 4 × 105 cells/well. Twenty four hours later parallel wells of transfected cells were depleted of methionine for 1 h in methionine-free Dulbecco's modified Eagle's medium containing 10% dialyzed fetal calf serum. The cells were then labeled for 30 min with 300 µCi of [35S]methionine/well, washed with complete medium, and incubated for 2.5 h. Culture supernatants were immunoprecipitated with protein G-agarose (Life Technologies, Inc.). Immunoprecipitated proteins were fractionated by SDS-polyacrylamide gel electrophoresis, and gels were subjected to fluorography.
Expression of fsAngII in Transfected MyocytesVentricular myocyte cultures from 4-day-old Sprague-Dawley rats (Charles River, St-Constant, Quebec, Canada) were prepared as described elsewhere for atrial myocyte cultures (33). Twenty hours before transfection, cardiac myocytes cells were plated at 1.5 × 106 per 35-mm Primaria-coated dish (Falcon). The cells were grown in complete serum-free medium (CSFM-1, Life Technologies, Inc.) supplemented with 10% (v/v) fetal calf serum and ampicillin-streptomycin-fungizone (Life Technologies, Inc.) at 10 µg/ml in a humidified incubator (5% CO2, 95% air) at 37 °C. The cells were transfected with pIgPfsAngII or pIg as a control. For transfection, a DNA-calcium phosphate precipitate (250 µl) formed with 10 µg of plasmid DNA was added, and the cells were incubated at 37 °C. After 24 h the medium was changed, and the cells were incubated an additional 48 h. Supernatants were collected and cells were rinsed twice with phosphate-buffered saline and sonicated in 2 ml of ice-cold extraction medium (5% formic acid, 15% trifluoroacetic acid, 10% hydrochloric acid, 1% sodium chloride). Lysates were cleared by centrifugation, and pellets were retained for protein quantitation by the method of Bradford (38). Ang II content of cell lysates and supernatants from transfected cardiac myocytes was extracted on disposable octadecyl cartridges (Fisher) and measured by radioimmunoassay as described elsewhere (39). An antiserum against Ang II (designated CD4) was used. This antiserum displayed 50-60% cross-reactivity to fsAngII (Ala-Pro-Gly-[Ile3,Val5]Ang II), purchased from Bachem, and less than 1% cross-reactivity to Ang I (data not shown).
In Vivo Injection of DNA Into Rat HeartsThis protocol was approved by the Care of Experimental Animals Committee of Henry Ford Hospital and is in accordance with federal guidelines. 250-300-g rats (Harlan Sprague Dawley, Charles River, Kalamazoo, MI) were anesthetized, intubated, and ventilated. Following thoracotomy, the heart was injected with 95 µg of DNA in 100 µl of phosphate-buffered saline as described previously (40). Rats were randomly divided into two groups, one group receiving injection of pIgPfsAngII and the other receiving pIg as a control. These animals were used to generate data on plasma Ang II levels, heart weight to body weight ratios, and blood pressure. The two DNA constructions were also injected into a separate group of rats for Northern blot analysis of injected hearts. Systolic blood pressure of conscious rats was measured by the tail-cuff method. Four to seven determinations were made for each rat. Measurements were made twice prior to DNA injection and on day 6 post-injection. All other parameters were determined at the time of sacrifice (7 days post-injection).
Northern Blot AnalysisTotal RNA was extracted by homogenization of hearts in RNA STAT-60 (Tel-Test, Friendswood, TX). Northern blot analysis for atrial natriuretic peptide (ANP) and glyceraldehyde-3-phosphate dehydrogenase has been described previously (41). ANP mRNA was quantitated using scanning densitometry of autoradiograms.
Binding of frog skin Ang II
(fsAngII) to the AT1 Ang II receptor was tested using
membrane preparations from transfected COS-1 cells. As shown in
Fig. 2A, binding of 125I-labeled
[Sar1,Ile8]Ang II was completely displaced
with cold Ang II with an apparent Kd of 0.66 ± 0.25 nM. fsAngII also completely displaced the radiolabel,
but the apparent affinity was approximately 43-fold lower than that of
Ang II. Losartan, an AT1-specific antagonist, completely
displaced binding, confirming that all of the receptors in this
membrane preparation are of the AT1 subtype. Using
membranes prepared from PC12W cells, we have determined that the
fsAngII peptide also bound the AT2 receptor with an
affinity only slightly lower than that of native Ang II (apparent
Kd of fsAngII = 6.18 ± 1.01 nM; apparent Kd of Ang II = 1.94 ± 0.30 nM) (Fig. 2B). PD123319, an
AT2-specific antagonist, completely displaced binding in
PC12W membranes, confirming that all of the receptors in this membrane
preparation are of the AT2 subtype. Thus fsAngII binds to
both the AT1 and AT2 Ang II receptor subtypes; however, binding to the AT1 receptor occurs with roughly
40-fold less affinity than native Ang II.
Biological Activity of fsAngII
Intracellular free calcium
concentration ([Ca2+]i) measurements were
conducted to determine the AT1 receptor-mediated biological
response of cells to fsAngII (Fig. 3). Basal
[Ca2+]i in AT1-transfected COS-1
cells was 76.5 ± 2 nmol/liter (n = 36 experiments
comprising 180 cells). Ang II and fsAngII induced a rapid and acute
[Ca2+]i response that reached peak levels within
20 s after addition of the agonist. Both peptides increased
[Ca2+]i in a dose-dependent manner,
with [Ca2+]i sensitivity to Ang II and fsAngII
being similar (pD2 for Ang II = 8.7 ± 0.3;
pD2 for fsAngII = 8.6 ± 0.7; pD2 = log(EC50 (mol/liter)). Ang II at concentrations greater
than 10
9 mol/liter induced significantly larger
[Ca2+]i responses than corresponding
concentrations of fsAngII (Fig. 3A). The time course of
[Ca2+]i recovery to basal levels following
agonist stimulation was significantly prolonged in cells stimulated
with fsAngII. Whereas Ang II-induced [Ca2+]i
returned to near basal levels within 300 s after the peak
response, fsAngII-induced [Ca2+]i levels remained
persistently elevated above basal levels 300 s after the
peak response (Fig. 3B).
Expression of fsAngII in Transfected Cells
To achieve direct
expression of angiotensin peptides, an fsAngII-encoding peptide was
linked to the 3 end of an engineered secretory protein through a
consensus cleavage site for furin, a ubiquitous protease (Fig. 1). To
test for cell-mediated release and cleavage of the fsAngII fusion
protein, GH4C1 cells were transfected with pIgfsAngII and
its control counterpart, pIg, which lacks the fsAngII coding sequences.
Immunoprecipitation of the Ig fragment from cells transfected with
either pIg or pIgfsAngII shows that the fusion protein detected in both
cases migrated as a single band with pIgfsAngII having a slightly
higher molecular weight than the pIg counterpart (data not shown). This
result is consistent with a lack of cleavage of the fsAngII peptide
from the Ig fragment in this fusion protein. In the
pIgfsPAngII construct, the 43-amino acid human prorenin
prosegment was inserted between the Ig fragment and the furin cleavage
site to act as a molecular "spacer" which would favor accessibility
of the furin cleavage site. Immunoprecipitation of the Ig fragment from
GH4C1 cells transfected with pIgPfsAngII shows
two bands of similar intensity (Fig. 4,
arrows). Deduced molecular weights suggest that the higher
band represents the intact fusion protein and the lower band the Ig
fragment linked to the prorenin prosegment following removal of the
fsAngII peptide. These results demonstrate that fsAngII can be cleaved
from an appropriate fusion protein in GH4C1 cells with an
efficacy of roughly 50% and be secreted into the extracellular
space.
To test for cardiocyte-mediated release of fsAngII, primary rat ventricular myocytes were transfected with pIgPfsAngII and its control counterpart, pIg. Ang II content from cell lysates and supernatants was measured by radioimmunoassay (Table I). Introduction of pIgPfsAngII into neonatal rat primary cardiocytes leads to expression of fsAngII. fsAngII was detected both within the cells and in cell lysates, demonstrating that fsAngII can be cleaved from an appropriate fusion protein and be secreted into the extracellular space.
|
To test for in vivo synthesis and biological effects of fsAngII, either pIg or pIgPfsAngII was directly injected into the apex of the rat cardiac ventricle. To test whether expressed fsAngII had local or systemic biological effects in vivo, we measured circulating plasma Ang II levels, blood pressure, heart weight, and ANP mRNA, a molecular marker of cardiac hypertrophy. Data in Table II show that there were no differences in the heart to body weight ratios and blood pressures of pIgPfsAngII- and pIg (control)-injected rats. The pIgPfsAngII-injected group also had no significant elevation in plasma Ang II levels versus controls (data not shown). To examine the effects of overexpression of fsAngII on ANP mRNA, total RNA was isolated from the lower one-third of injected hearts (injection site and surrounding apex and left ventricular free wall) and from the upper one-third of the left ventricle (control tissue, distal to injection area). ANP mRNA was induced 4-fold in the pIgPfsAngII-injected hearts as compared with the pIg-injected hearts (Fig. 5, A and B; p < 0.01, n = 8). Enhanced ANP gene expression was not detected from the upper one-third of the left ventricle of pIgPfsAngII- and pIg-injected hearts (data not shown). Thus, our data suggest that local overexpression of fsAngII in cardiac myocytes leads to a highly localized induction of ANP mRNA, a molecular marker of cardiac hypertrophy.
|
We have engineered a fusion protein that leads to the direct
release of an Ang II analog within transfected tissues. The angiotensin peptide used in this study is released by a single cleavage, effected by the processing protease furin, leading to the constitutive secretion
of the released peptide. Furin, a mammalian homolog of the yeast
precursor-processing Kex2 endoprotease, is a Golgi-anchored convertase
expressed in all examined tissues and cell lines (42-44). Sequence
requirements for efficient processing of precursors by furin have been
extensively studied, and the RXRXKR amino acid combination located from position 6 to
1 relative to the cleavage site has been shown to lead to the higher cleavage efficiency in
cultured cells (29, 30). fsAngII,
Ala-Pro-Gly-[Ile3,Val5]Ang II, is an
undecapeptide isolated from the skin of the Australian frog C. georgiana (21), which was chosen because its amino-terminal alanine residue is a good substrate for efficient processing by furin
(23). Our results show that efficient cleavage and secretion of fsAngII
from the fusion protein requires the presence of a molecular spacer
between the Ig fragment and the angiotensin peptide presumably due to
steric constraints near the cleavage site. The choice of the prorenin
prosegment as a spacer was dictated by our previous success with the
engineered furin cleavage in this peptide (27). The presence of the Ig
fragment in the fusion protein allows for a rapid assessment of
cleavage efficiencies with various peptide-containing fusion proteins
in cell cultures (Fig. 4 and data not shown) by the simple use of
protein A- or G-coupled matrices. Evidence for the cleavage-mediated
release of fsAngII peptide is also provided by the finding that
introduction of pIgPfsAngII into primary cultures of rat
neonatal ventricular myocytes leads to intracellular accumulation and
secretion of immunoreactive fsAngII (Table I).
Our results demonstrate that fsAngII binds to both the AT1 and AT2 Ang II receptor subtypes. While fsAngII binds the AT2 receptor with a similar affinity to that of mammalian Ang II (native Ang II), binding to the AT1 receptor, which is thought to mediate most of the cardiovascular effects of Ang II (34, 45), occurs with 40-fold lower affinity than native Ang II (Fig. 2A). However, even though fsAngII binding elicits only approximately 65% of the AT1 receptor-mediated release of intracellular calcium, its duration of action on target cells is significantly longer than that of native Ang II (Fig. 3). The combination of these properties may explain the apparent similarities in effective biological concentrations of Ang II and fsAngII on transfected cells. Indeed, Khosla and co-workers (22) have previously reported that the pressor activity of the synthetic fsAngII in rats was 90.6 ± 5.0% that of human Ang II.
Our in vivo injection experiments, using pIgPfsAngII-injected rat cardiac ventricles, demonstrate for the first time that an engineered fusion protein precursor (Ig-prosegment-fsAngII) can be processed to a mature peptide (fsAngII) in vivo and have local biological effects. We are unsure as to the origin of the cells secreting the fsAngII peptide in injected hearts as immunostaining for either the Ig or angiotensin peptides was inconclusive (perhaps due to a low level of expression and/or the rapid secretion of the fusion protein from the cells). However, as the viral promoter/enhancer used in our expression vector has a broad cell specificity, it could conceivably be expressed in either cardiac myocytes or fibroblasts. Analyses of injected rats revealed no significant differences in blood pressure, total heart weight to body weight ratios (Table II), or in circulating Ang II levels (data not shown) between the control and pIgPfsAngII-expressing groups. In contrast to the absence of systemic fsAngII effects, fsAngII induced a significant increase in ANP mRNA levels in the injected portion of the heart (Fig. 5). Enhanced ANP gene expression was not detected distal to the injection site. Coupled with the failure to detect an increase in heart to body weight ratios, these data suggest that the expression of fsAngII is highly localized to cells along the injection site. These data support the in vitro results of Sadoshima et al. (46) who demonstrated the autocrine effects of Ang II in cardiac myocyte hypertrophy in vitro and demonstrate that in vivo local overexpression of fsAngII in the rat heart leads to a highly localized cardiac hypertrophic phenotype.
In conclusion, we have described a novel expression vector that can serve as an alternative to co-expressing all the components of the RAS to generate Ang II. This type of approach could also be extended with the use of tissue-specific genes in transgenic animals to test the importance of other bioactive peptides on organ physiology.