Regulatable atrial natriuretic peptide gene
therapy for hypertension
Kurt J. Schillinger*, Sophia Y. Tsai*, George E. Taffet†, Anilkumar K. Reddy†, Ali J. Marian†, Mark L. Entman†,
Kazuhiro Oka*, Lawrence Chan*, and Bert W. O’Malley*‡
Departments of *Molecular and Cellular Biology and †Medicine, Sections of Cardiovascular Sciences and Cardiology and DeBakey Heart Center,
Baylor College of Medicine, 1 Baylor Plaza, Houston, TX 77030
Contributed by Bert W. O’Malley, August 10, 2005
Hypertension (HTN) is a disease that begins with dysfunctional
renal-sodium excretion and progresses to a syndrome of highly
elevated systolic, diastolic, and mean arterial pressures. Inadequa-
cies in the therapy of HTN have led to the investigation of the gene
therapy of this disease by using systemic overproduction of vaso-
dilatory peptides, such as atrial natriuretic peptide (ANP). How-
ever, gene-therapy approaches to HTN using ANP are limited by the
need for long-term ANP gene expression and, most important,
control of ANP gene expression. Here, we introduce a helper-
dependent adenoviral vector carrying the mifepristone (Mfp)-
inducible gene-regulatory system to control in vivo ANP expres-
sion. In the BPH兾2 mouse model of HTN, Mfp-inducible ANP
expression was seen for a period of >120 days after administration
of vector. Physiological effects of ANP, including decreased systolic
blood pressure, increased urinary cGMP output, and decreases in
heart weight as a percentage of body weight were also under the
control of Mfp. Given these capabilities, this vector represents a
paradigm for the gene therapy of HTN.
adenovirus 兩 mifepristone-inducible
H ypertension (HTN) is a disease that is characterized by
chronically elevated arterial blood pressure that affects
⬇25–35% of the population in developed countries, and it is
one of the leading causes of morbidity and mortality in adults
(1). Development of HTN is likely to involve defects in
renal-sodium handling that are exacerbated by maladaptive
activation of neurohormonal compensatory mechanisms. The
morbidity and mortality caused by the hypertensive state are
a consequence of detrimental effects that manifest predomi-
nantly in the heart, brain, and kidney and are referred to
collectively as hypertensive end-organ disease. Typically, end-
organ disease involves pathological left-ventricular hypertro-
phy, a propensity toward cerebrovascular accidents, and ac-
celerated renal atherosclerosis (1, 2).
The current therapy of HTN has been rationally designed to
address the recognized causes of hypertensive pathophysiology,
including impaired urinary-sodium excretion and neurohor-
monal compensatory-mechanism activation. Also, in recogni-
tion of the fact that hypertensive end-organ disease represents
the true danger of chronically elevated arterial blood pressure,
the main goals of antihypertensive therapy have evolved to
include a reduction in blood pressure as well as prevention of
cardiovascular, cerebrovascular, and renal pathology associated
with elevated blood pressure (2, 3). However, a multitude of
studies have revealed that no available drug therapy for HTN has
the capability to cure the hypertensive state or reverse or prevent
the pathological changes associated with hypertensive end-organ
disease fully (2). As a result, therapeutic approaches to HTN
involving gene therapy have been of recent interest.
Atrial natriuretic peptide (ANP), a 27-aa peptide hormone
with potent vasodilatory, natriuretic, diuretic, and antihypertro-
phic effects, is an attractive candidate gene product for antihy-
pertensive therapy (4, 5). Studies involving the infusion of
synthetic ANP to humans with HTN have demonstrated signif-
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0506807102
icant decreases in blood pressure and brisk natriuresis兾diuresis
associated with peptide administration (6, 7). Likewise, gene-
therapy experiments conducted with adenoviral vectors in ro-
dent models of HTN have shown equally impressive results (8,
9). However, these studies also have revealed the serious limi-
tations that are associated with the use of ANP in a gene-therapy
context. Chief among these limitations are the need for long-
term ANP expression and, importantly, control of ANP gene
expression. Here, we report the development and testing of a
helper-dependent adenoviral (HD-Ad) vector, called iANP.HD,
which was designed to address these current limitations for ANP
gene therapy of HTN. Our strategy combines the long-term
transgene-expression capability of an HD-Ad vector with the
use of the mifepristone (Mfp)-inducible gene-regulatory sys-
tem (MIGRS) for ANP-expression control. By application of
iANP.HD to the BPH兾2 mouse model of HTN, we demonstrate
the ability of this vector to allow Mfp-inducible ANP expression
in vivo at time points ⬎120 days after a single administration of
vector. Also, we demonstrate that certain physiological conse-
quences of ANP overexpression from this vector are also regu-
latable with administration of Mfp.
Methods
Preparation of iANP.HD. A three-piece ligation was performed with
(i) a NotI–AscI fragment from a pBluescript II(⫹) backbone
(Stratagene) containing a modified multicloning site; (ii) a PCR
fragment generated with NotI and NcoI linkers encoding the
promoter of the pEP1422 plasmid (Valentis, Burlingame, CA),
which contains three tandem copies of the Gal4 upstream
activation sequence, a TATA box, and an artificial intron; and
(iii) a PCR fragment generated with NcoI and AscI linkers from
the pUC9-preproANP plasmid (gift from David Vesely, Uni-
versity of South Florida, Health Sciences Center, Tampa), which
APPLIED BIOLOGICAL
contains a cDNA copy of the rat preproANP gene. The resulting
plasmid, PNA-6xANP, contained the rat preproANP cDNA in
SCIENCES
the context of the pEP1422 Mfp-inducible promoter. A PCR
fragment with ClaI and AscI linkers was then generated from the
pCR3.1 plasmid (Stratagene), which contained the bovine-
growth hormone poly(A) signal and was subcloned into the
unique ClaI and AscI sites of PNA-6xANP to generate PNA-
6xANPp(A). A NotI–AscI fragment from PNA-6xANPp(A) was
then ligated to a NotI–AscI fragment from PAP-GH-GLp65 (10)
encoding GLp65 driven by a transthyretin promoter-enhancer
sequence (TTRB) (11) to generate plasmid PAP-6xANP-
GLp65. A NotI fragment from this plasmid, ⬇8.5 kbp, was then
subcloned into the unique NotI site of the C4HSU HD-Ad
backbone (12) to generate plasmid 6xANP.C4HSU. We then
linearized 10 ␮g of 6xANP.C4HSU by digestion with PmeI, and
Abbreviations: HTN, hypertension; ANP, atrial natriuretic peptide; Mfp, mifepristone;
MIGRS, Mfp-inducible gene-regulatory system; HD-Ad, helper-dependent adenoviral; HW兾
BW, heart weight as a percentage of body weight.
‡To whom correspondence should be addressed. E-mail: berto@bcm.tmc.edu.
© 2005 by The National Academy of Sciences of the USA
PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 13789 –13794
recombinant HD-Ad particles, named iANP.HD, were gener-
ated and amplified as described (13).
In Vitro Studies. HepG2 cells were infected with 1 ⫻ 109 particles
iANP.HD and treated with either 100% ethanol or 1 ⫻ 10⫺7 M
Mfp in ethanol as described (10). At 24 h later, medium was
removed from each well and frozen at ⫺80°C until the time of
assay. For assay, 50 ␮l of medium was used directly in a
commercially available enzyme immunoassay for rat ANP (Kit
S-1132; Bachem).
Mouse Studies. Three BPH兾2 brother–sister mating pairs were
acquired from The Jackson Laboratory (strain, BPH兾2J; gen-
eration, F66). Animals were maintained at a constant tempera-
ture with 14:10-h light兾dark cycle and ad libitum access to
standard rodent food and tap water. Male BPH兾2 mice were
housed one animal per cage to avoid fighting. All animal
protocols were approved by the Institutional Animal Care and
Use Committee of Baylor College of Medicine in accordance
with the National Institutes of Health Guide for the Care and
Use of Laboratory Animals.
BPH兾2 mice were infected with 5 ⫻ 1010 particles iANP.HD
by tail-vein injection as described (10). For plasma collection,
retroorbital blood was collected and transferred to Microtainer
EDTA tubes (BD Biosciences). We combined 15 ␮l of plasma
with 250 ␮l of 1⫻ RIA buffer (RIKBUFF; Bachem) and stored
it at ⫺80°C until the time of the assay. For the assay, 100 ␮l of
plasma–RIA buffer mixture was used directly in a commercially
available RIA for rat ANP (Kit S-2039; Bachem).
Urine Studies. Model 650-0100 metabolic cages with mouse
diuresis adaptors (MTB-0322, Nalge) were used for collection
of urine. A Plexiglas divider was inserted into the domicile
portion of the cage to prevent fighting. All collections were
started at 0700 – 0730 hours. At the time of collection, two male
mice were used in each cage (separated by the Plexiglas
divider), and their combined urine was collected. Each mouse
had ad libitum access to water but not to food. We split 9 h of
total urine-collection time into three contiguous 3-h collection
periods. At the end of each collection period, the urine was
centrifuged for 5 min at 17,900 ⫻ g and room temperature.
Volume was recorded, and urine was stored at ⫺80°C until
assay. The feces-separation screen was rinsed under tap water,
dried, and replaced at the end of each 3-h collection period.
For assay, urinary cGMP concentration was determined by
using a commercially available ELISA (Kit 900-013, Assay
Designs, Ann Arbor, MI). Simultaneously, urine samples were
sent to the Pathology Department of Baylor College of Med-
icine Center for Comparative Medicine for determination of
urine sodium and potassium by f lame photometry.
Blood Pressure Studies. Mice were anesthetized by i.p. injection of
75 mg兾kg pentobarbital (Nembutal, Abbott), and the left carotid
artery was exposed and cannulated with a heparin–saline-filled
catheter, as described in ref. 14. The catheter was composed of
a 2-cm piece of polytetrafluoroethylene (PTFE) tubing (o.d.兾
i.d., 0.41:0.25 mm) (SUBL-160, Braintree Scientific), which was
joined with Superfast epoxy (Elmer’s Products, Columbus, OH)
to a 10-cm-long piece of PE-50 polyethylene tubing (o.d.兾i.d.,
0.038:0.023 inches) (PE50, Braintree Scientific, Braintree, MA).
Before insertion of the PTFE portion of the catheter into the
carotid artery, the entire catheter was filled with a sterile 250
units兾ml heparin–saline solution and connected to a pressure
transducer (Meritrans MER100, Merit Medical Systems, South
Jordan, UT) by 23G blunt tubing adaptor (Becton Dickinson).
The catheter was flushed briefly with ⬍50 ␮l of heparin–saline,
the system was allowed to stabilize for ⬇2 min, and then 10-s
blood pressure recordings were made through a calibrated
13790 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0506807102
Fig. 1. The iANP.HD vector contains two expression cassettes in the C4HSU
HD-Ad backbone, which contains the inverted terminal repeats (ITR) and
packaging signal (␺) of the Ad5 genome but is devoid of all adenoviral coding
regions (12). The first expression cassette encodes the MIGRS and enables
liver-specific expression of the GLp65 transactivator from a transthyretin
promoter (TTRB) (11). The second expression cassette (iANP) encodes rat
preproANP in the context of a Mfp-sensitive promoter (MSP). In the absence
of Mfp, GLp65 exists in a predominantly monomeric, inactive state. Mfp binds
to and activates GLp65, promoting dimerization, DNA binding, and transac-
tivation of rat preproANP expression.
bridge amplifier (band width, 0–2 kHz) on a Dell workstation
with signal-processing software (DSPW, Indus Instruments,
Houston) every 1 min for the next 5 min. At the end of recording,
the catheter was removed, the left carotid artery was perma-
nently ligated, and the neck incision was closed to allow the
animal to recover. Blood pressure data acquired for 10 s during
blood pressure measurement were exported to a worksheet in
EXCEL (Microsoft) at an interval of 0.5 ms (20,000 data points
per 10 s) for accurate determination of systolic blood pressure
and heart rate.
Tissue Preparation. At 14–16 h after blood pressure determina-
tion, animals were killed and weighed. Hearts were removed,
cleaned under a dissecting microsope, rinsed, dried, and
weighed. A cross-sectional ring through both ventricles was cut,
placed in 10% zinc–formalin (Protocol, Fisher Scientific), and
fixed overnight. Heart rings were then cleared through xylene
and embedded in paraffin. Tissues were sectioned at 5 ␮m and
stained with hematoxylin–eosin or Masson’s trichrome (Kit
87019, Richard-Allan Scientific, Kalamazoo, MI).
mMode Echo and Doppler Analysis. Doppler and 2D mMode-
guided echocardiography studies were performed on isoflurane-
anesthetized mice as described (15, 16).
Statistical Analysis. An independent t test was performed on
sample data by using the SPSS 11.0 statistical-analysis software
(Prentice Hall, Upper Saddle River, NJ). Levine test for equal
variance (⬍0.05) was applied to all samples before evaluation
with t test. In the absence of equal variance, statistical signif-
icance was not reported.
Results
Design of iANP.HD. Prior attempts at using ANP for the gene
therapy of HTN suffered from limitations, such as short duration
of ANP expression and lack of control of ANP expression
subsequent to vector delivery (9). iANP.HD, the HD-Ad vector
used throughout the studies described here, was designed to
address the limitations of earlier ANP gene-therapy approaches.
To this end, the MIGRS, which is an expression system that
couples control of a specific target gene to administration of the
compound Mfp, was incorporated into iANP.HD.
Two expression cassettes have been incorporated into iANP.HD
(Fig. 1). In the first cassette, termed the MIGRS cassette, expres-
Schillinger et al.
 per well of vector. Treatment of infected cells with 1 ⫻ 10⫺7 M
Mfp resulted in a 200-fold increase in ANP expression (Fig. 2A).

Testing iANP.HD in Vivo. Preliminary investigation of Mfp-

Fig. 2. iANP.HD enables Mfp-inducible ANP expression in vivo and in vitro.
(A) Infection of HepG2 cells with iANP.HD and subsequent administration of
Mfp resulted in a 213-fold increase in ANP expression. No increase in ANP
expression occurred in cells treated with PBS (control) and Mfp. (B) At 4 weeks
after introduction of 5 ⫻ 1010 particles of iANP.HD to mice (iANP.HD), a single
oral dose of Mfp induced an average 134-fold increase in plasma ANP levels.
Control mice demonstrated no increase in plasma ANP levels after Mfp. n ⫽ 6,
for all groups.
sion of the Mfp-responsive chimeric transactivator GLp65 is driven
by a liver-specific transthyretin promoter (TTRB) (11). GLp65 is
composed of a truncated form of the human progesterone receptor
ligand-binding domain, the DNA-binding domain of the yeast Gal4
transcription factor, and a transactivation domain derived from the
p65 subunit of human NF-␬B. In the absence of Mfp, GLp65
produced from this cassette exists in a largely monomeric, inactive
state. The second expression cassette, called the iANP cassette,
contains a cDNA of rat preproANP downstream of a Mfp-sensitive
promoter (MSP). Upon binding Mfp, GLp65 undergoes a confor-
mational change that allows the protein to dimerize, bind to the
MSP in the iANP cassette, and transactivate expression of rat
preproANP (Fig. 1). Because both the MIGRS and iANP cassettes
have been subcloned simultaneously into a unique site in the
C4HSU HD-Ad backbone of Sandig et al. (12) to create iANP.HD,
this single vector has the capacity to enable liver-specific, Mfp-
inducible ANP expression.
Testing iANP.HD in Vitro. To test Mfp-inducible ANP expression
from iANP.HD, HepG2 cells were infected with 1 ⫻ 109 particles
inducible ANP expression in vivo from iANP.HD was done by
using two groups of male mice from the BPH兾2 genetically
hypertensive mouse strain (17). All mice in the first group
received 5 ⫻ 1010 particles of iANP.HD by tail-vein injection,
whereas all mice in the second group received only PBS by
tail-vein injection. At 4 weeks after injection, administration of
a single oral dose of Mfp at 330 ␮g兾kg resulted in an average
134-fold increase in plasma ANP levels in mice that had received
iANP.HD (Fig. 2B).
iANP.HD Enables Long-Term, Dose-Dependent ANP Expression in Vivo.
Recognizing that any future potential application of iANP.HD
to the gene therapy of HTN would require demonstration of
constant, controllable expression of ANP, we next investigated
the feasibility of inducing ANP expression over a 60-day period
with the implantation of a single, biodegradable pellet contain-
ing Mfp. Also, to demonstrate true Mfp-regulable ANP expres-
sion, we investigated the possibility of inducing different plasma
levels of ANP by using pellets containing different quantities
of Mfp.
At 8 weeks of age, 30 male BPH兾2 mice were divided into five
groups of six mice. All mice in the first two groups received PBS
by tail-vein injection, whereas mice in the remaining three
groups received iANP.HD. To demonstrate that iANP.HD
would enable inducible ANP expression ⬎100 days after it was
delivered, these long-term expression levels were not started
until 11 weeks after vector delivery. At 11 weeks after tail-vein
injection, all animals had blood drawn to measure baseline
plasma ANP levels. At this time point, no difference in baseline
plasma ANP levels was observed between mice that had received
iANP.HD and mice that had received PBS (Fig. 3A, day 0).
At week 11.5 after injection, ANP expression was induced by
s.c. implantation of timed-release, biodegradable pellets con-
taining Mfp (Innovative Research of America). Implantation of
180-␮g Mfp pellets into mice infected with iANP.HD resulted in
statistically significant increases in average plasma ANP levels

APPLIED BIOLOGICAL
SCIENCES

Fig. 3. iANP.HD enables long-term, Mfp-inducible, bioactive ANP expression. (A) At 11 weeks after injection (day 0), and before implantation of pellets, no
difference in plasma ANP levels was apparent between groups receiving iANP.HD (vector) and groups receiving PBS. Groups receiving both iANP.HD and Mfp
(Vector⫹180 ␮g Mfp and Vector⫹90 ␮g Mfp) showed sustained elevations in plasma ANP for as long as 46 days after pellet implantation. n ⫽ 6, for all groups.
(B) At 10 weeks after injection (day 0), and before implantation of pellets, no difference in urinary cGMP output was apparent between groups receiving iANP.HD
(vector) and groups receiving PBS. Activation of ANP expression by implantation of Mfp pellets in groups receiving iANP.HD (Vector⫹90 ␮g Mfp and Vector⫹180
␮g Mfp) resulted in significant elevations in urinary cGMP output lasting as long as 53 days. n ⫽ 3, for all groups. *, P ⬍ 0.05.

Schillinger et al. PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 13791
observable 4, 18, 32, and 46 days after implantation. Similarly,
implantation of 90-␮g Mfp pellets into mice infected with
iANP.HD resulted in statistically significant increases in average
plasma ANP levels 4 and 18 days after implantation; a statisti-
cally less significant trend in elevated average plasma ANP levels
extended to 32 and 46 days after implantation. Importantly, 4
days after pellet implantation, a statistically significant differ-
ence in average plasma ANP levels existed between the groups
of mice that had received iANP.HD and either a 180- or 90-␮g
Mfp pellet. No elevations in plasma ANP levels were seen at any
time points in mice receiving PBS by tail-vein injection or by mice
that had received iANP.HD but were implanted with a placebo
pellet (Fig. 3A).
Data from this study demonstrated the potential of iANP.HD
to enable long-term, Mfp-inducible ANP expression and sug-
gested the existence of a dose-dependence relationship between
ANP expression levels and dose of Mfp administered. Also, note
that Mfp-inducible ANP expression was observed 123 days after
a single administration of the iANP.HD vector to mice. This
length of expression after vector administration is 74 days longer
than any expression that has been reported (8).
ANP Produced from iANP.HD Is Bioactive. Measurement of urinary
cGMP output serves as an accurate, noninvasive means of
monitoring the bioactivity of plasma ANP (18). To take advan-
tage of this fact, the mice used above to investigate long-term,
Mfp-inducible ANP expression were used simultaneously to
investigate the bioactivity of Mfp-inducible ANP.
At 11, 25, 39, and 53 days after pellet implantation, statistically
significant elevations in urinary cGMP output were observed in
mice receiving iANP.HD and a 180-␮g Mfp pellet. Similarly, 11,
25, and 39 days after pellet implantation, statistically significant
elevations in urinary cGMP output were observed in mice
receiving iANP.HD and a 90-␮g Mfp pellet. In a manner
analogous to that seen with plasma ANP levels, a trend toward
dose dependence developed in the urinary cGMP output levels
between mice receiving iANP.HD and either a 180- or 90-␮g
Mfp pellet. Although not statistically significant because of
limited sample size and variance, this trend appeared to be
maintained throughout the study, days 11–67 (Fig. 3B).
Controlling the Hypotensive Effects of ANP with Mfp. Recognition
that uncontrollable expression of a hypotensive gene product
could have serious unintentional consequences has led to the
recent observation that relevant application of gene therapy to
HTN will require a means to control expression of therapeutic
gene products (19). The primary reason, then, for incorporating
the MIGRS into iANP.HD was to link control of the hypotensive
effects of ANP to the administration of Mfp.
To investigate the effects of Mfp-inducible ANP expression
on blood pressure, a study was conducted that was designed to
complement the studies of long-term, Mfp-inducible ANP
expression described above. In the first part of this study, 24
male BPH兾2 mice were divided into four groups of six mice.
All mice in the first two groups received PBS by tail-vein
injection, whereas all mice in the remaining two groups
received iANP.HD. As before, 80 days after injection, mice
were implanted s.c. with 180-␮g Mfp pellets or placebo pellets.
At 17 days after pellet implantation, a time point that corre-
sponded to maximum ANP expression levels and maximum
urinary cGMP output in the previous study, systolic blood
pressure was measured in all mice through carotid artery
catheterization. At this time point, a statistically significant
decrease in systolic blood pressure was observed between mice
receiving iANP.HD and a 180-␮g Mfp pellet and all control
mice (Fig. 4A).
The second part of this blood pressure study was conducted
with four additional groups of six BPH兾2 mice in a manner
13792 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0506807102
Fig. 4. Mfp-induced ANP expression decreases systolic blood pressure. (A) At
17 days after pellet implantation (14 weeks after infection), average systolic
blood pressure in mice receiving iANP.HD and a 180-␮g Mfp pellet (Vec-
tor⫹180 ␮g Mfp) was significantly lower than systolic blood pressure in all
controls. *, P ⬍ 0.05. (B) At 57 days after pellet implantation (19.5 weeks after
infection), results were similar to those obtained at 17 days after pellet
implantation. *, P ⬍ 0.05.
nearly identical to the first part of the study. However, blood
pressure was measured 57, rather than 17, days after pellet
implantation in all four groups. This time point was chosen to
correlate with minimum ANP expression levels and minimum
urinary cGMP output levels observed in the previous ANP
expression study. Results at the day-57 time point were analo-
gous to those seen at the day-17 time point. Average systolic
blood pressure in animals receiving iANP.HD and a 180-␮g Mfp
pellet was significantly lower than average systolic blood pres-
sure in all control groups (Fig. 4B).
Taken together, results from this study indicate that iANP.HD
enables long-term, Mfp-inducible control of the hypotensive
effects of ANP. Interestingly, whereas ANP expression levels
previously demonstrated dose dependence at both 17 and 57
days after pellet implantation, systolic blood pressure did not
demonstrate a similar trend. This result is likely to be caused by
the potency of ANP, as even picogram quantities of this peptide
can significantly decrease systolic blood pressure.
Supraphysiological ANP Expression Decreases the Heart Weight as a
Percentage of Body Weight (HW兾BW) Ratio. Characterization of the
BPH兾2 mouse model of genetic HTN by Schlager and Sides (17)
revealed an increased HW兾BW ratio in this mouse strain when
compared with normotensive controls. Using the BPH兾2 model,
Leckie (20) demonstrated a small decrease in this elevated
HW兾BW ratio after 7 days of treatment with the angiotensin-
converting enzyme inhibitor captopril. Given the importance of
addressing hypertensive end-organ disease, such as left-
ventricular hypertrophy, by antihypertensive therapy, the effect
of Mfp-inducible ANP expression on the HW兾BW ratio of
BPH兾2 mice was investigated. To study this effect, mice in all
blood pressure measurement groups were killed 14–16 h after
blood pressure measurement and HW兾BW was determined.
Analysis of the hearts of these mice revealed decreases in
HW兾BW ratio with Mfp-induced ANP expression greater than
any that have been published to our knowledge. At 17 days after
implantation of a 180-␮g Mfp pellet into mice that had received
iANP.HD, a 15% decrease in HW兾BW was evident (Fig. 5A).
Schillinger et al.

Fig. 5. Mfp-induced ANP expression decreases heart size. (A) At 17 days after
pellet implantation, HW兾BW was significantly decreased in mice receiving
iANP.HD and a 180-␮g Mfp pellet (Vector⫹180 ␮g Mfp) when compared with
all controls. **, P ⬍ 0.01. (B) Decreases in HW兾BW in mice receiving both
iANP.HD and a 180-␮g Mfp pellet (Vector⫹180 ␮g Mfp) was more pronounced
57 days after pellet implantation than 17 days after implantation. **, P ⬍ 0.01.
More impressively, 57 days after activation of ANP expression in
mice receiving iANP.HD and a 180-␮g Mfp pellet, a 20%
decrease in HW兾BW was seen (Fig. 5B). Also, histological
evaluation of hearts from mice receiving iANP.HD and a 180-␮g
Mfp pellet for 57 days revealed that cardiac architecture was
unchanged, with no evidence of abnormal collagen deposition
(data not shown). Histological studies performed on the livers of
these mice also indicated no difference between controls and
mice receiving iANP.HD in terms of architecture and inflam-
matory cell infiltrates (data not shown).
In a separate but parallel experiment designed to study the
functional consequences of this decrease in heart size, 10 BPH兾2
male mice at 8 weeks of age were injected with either iANP.HD
or PBS, followed by implantation of a 180-␮g Mfp pellet. Thirty
days later, assessment of cardiac function and left-ventricular
dimensions was made through Doppler and 2D-guided mMode
echocardiographic studies. Mfp-inducible ANP expression was
accompanied by decreases in peak aortic flow velocity and mean
and peak aortic flow acceleration. After 30 days of supraphysi-
ological ANP expression, left-ventricular stroke volume was also
decreased significantly. These results are likely to be caused by
a combination of chronic unloading of the myocardium through
ANP-mediated decreases in peripheral vascular resistance and
intravascular volume, and direct antihypertrophic, growth-
inhibitory effects of ANP on individual cardiomyocytes (21, 22).
This demonstration of the remarkable potency of ANP on the
myocardium at supraphysiological levels also underscores the
importance of control of ANP gene expression in a gene-therapy
context.
Discussion
The strength of ANP as a potential antihypertensive agent arises
from the multifactorial nature of its effects on blood pressure
homeostasis, fluid-electrolyte balance, and cardiomyocyte hy-
pertrophy. Current antihypertensive therapies such as ␤-blockers
are capable of addressing single mechanistic dysfunctions in the
pathophysiology of HTN. ANP, by contrast, has the capacity to
address multiple simultaneously dysfunctional systems. In the
peripheral vasculature, ANP promotes vasodilation by both
Schillinger et al.
relaxing vascular smooth muscle cells (VSMC) and attenuating
vascular reactivity to vasoconstrictive agents such as endothelin
1 and angiotensin II (AngII) (23–25). In the kidney, ANP
selectively vasodilates afferent glomerular arterioles and inhibits
passive and active sodium reabsorption in the renal intramed-
ullary collecting duct. These properties of ANP make it the only
currently known agent capable of simultaneously decreasing
blood pressure while increasing the glomerular filtration rate
(26). In addition to these antihypertensive effects, ANP also
directly inhibits the sympathetic nervous system (SNS) and the
renin–angiotensin–aldosterone (RAA) axis, two neurohor-
monal systems strongly implicated in the pathophysiology of
HTN. The sympatholytic effects of ANP have been demon-
strated in humans at the level of autonomic ganglion transmis-
sion and are thought to play a critical role in neuromodulation
of both cardiovascular tone and renal SNS activity (21, 27). With
respect to the RAA axis, ANP directly inhibits release of renin
from the juxtaglomerular apparatus, attenuates the effects of
AngII in VSMCs and cardiomyocytes, and directly abrogates the
synthesis and release of aldosterone from the zona glomerulosa
(26, 28). It is these multisystem effects of ANP that make this
peptide hormone a very attractive potential antihypertensive
agent.
Practical application of ANP to HTN is limited by a plasma
half-life of 30 s and a requirement for intravascular administra-
tion. Consequently, gene therapy represents a treatment modal-
ity in which ANP could be applied to a chronic disorder such as
HTN. ANP gene therapy has been studied in rat models of HTN
with positive but limited results (8, 9). Certainly, physiologic
results presented here are in accord with results obtained in
these previous studies, as significant decreases in systolic blood
pressure were seen in mice expressing ANP from iANP.HD.
However, our studies were designed to take ANP gene therapy
one step closer to practical application. To this end, we focused
on addressing both the chronic nature of HTN, as well as the
need for control of ANP expression. Our results demonstrate
that iANP.HD enabled expression of ANP for 125 days after a
single administration of vector. These data represent ANP
expression that is 74 days longer than any data that have been
reported, to our knowledge. Also, our data convincingly dem-
onstrate that ANP expression from iANP.HD could be con-
trolled by administration of Mfp, and that the physiological
effects of ANP are also coupled to Mfp administration.
Given that the true morbidity and mortality from HTN are
associated with hypertensive end-organ disease, results from this
study showing decreases in the HW兾BW ratio with ANP expres-
APPLIED BIOLOGICAL
sion are encouraging. The antihypertrophic and growth inhibi-
tory effects of ANP have been well described in vitro and in vivo
SCIENCES
and involve increases in intracellular cGMP concentrations (5,
22, 29). The decreases in HW兾BW ratios reported in this article
are similar in many ways to the cardiac changes reported with
administration of the PDE5A inhibitor sildenafil, which blocks
the breakdown of intracellular cGMP (30). However, whereas
PDE5A inhibition increased cardiac functional performance in
the context of acute pressure-overload hypertrophy, prolonged
ANP expression in our study slightly reduced cardiac perfor-
mance. This difference was likely because ANP has a potent
after-load reduction capability involving decreases in both pe-
ripheral vascular resistance and overall circulatory volume.
Consequently, it is likely that the significant decreases in
HW兾BW ratio and the concomitant decreases in cardiac func-
tion observed with ANP expression in our experiments arose as
a result of both decreased load and a direct antihypertrophic
effect of ANP on the cardiomyocytes. These effects were more
potent than those that have been described in any antihyperten-
sive gene-therapy experiment of which we are aware, and they
underscore the critical importance of regulating the expression
of ANP in a gene-therapy context.
PNAS 兩 September 27, 2005 兩 vol. 102 兩 no. 39 兩 13793
 The true advantage of iANP.HD as a gene-therapy vector for
HTN results from combining an HD-Ad vector with the MIGRS.
HD-Ad vectors have significant advantages over their predeces-
sors, including the capacity for up to 36 kbp of foreign DNA,
remarkably reduced cytotoxicity and immunogenicity in vivo,
and greatly prolonged target gene expression (31, 32). Such
characteristics of HD-Ad vectors make them potentially suitable
for the gene therapy of HTN, given the chronic nature of this
disease. However, it remains unclear whether chronic injection
of HD-Ad vectors over years will be tolerated immunologically.
Similarly, the MIGRS is well suited to gene-therapy applications
and has shown considerable promise when delivered with for-
mulated direct DNA injection (33). Also, Mfp has been ap-
proved for use in humans at doses 10- to 100-fold higher than
those that are required currently to induce transgene expression
1. Staessen, J. A., Wang, J., Bianchi, G. & Birkenhager, W. H. (2003) Lancet 361,
1629–1641.
2. Chobanian, A. V., Bakris, G. L., Black, H. R., Cushman, W. C., Green, L. A.,
Izzo, J. L., Jr., Jones, D. W., Materson, B. J., Oparil, S., Wright, J. T., Jr., et al.
(2003) J. Am. Med. Assoc. 289, 2560–2572.
3. Staessen, J. A., Gasowski, J., Wang, J. G., Thijs, L., Den Hond, E., Boissel,
J. P., Coope, J., Ekbom, T., Gueyffier, F., Liu, L., et al. (2000) Lancet 355,
865– 872.
4. D’Souza, S. P., Davis, M. & Baxter, G. F. (2004) Pharmacol. Ther. 101,
113–129.
5. Hayashi, D., Kudoh, S., Shiojima, I., Zou, Y., Harada, K., Shimoyama, M., Imai,
Y., Monzen, K., Yamazaki, T., Yazaki, Y., et al. (2004) Biochem. Biophys. Res.
Commun. 322, 310–319.
6. Predel, H. G., Schulte-Vels, O., Glanzer, K., Meyer-Lehnert, H. & Kramer,
H. J. (1991) Am. J. Hypertens. 4, 871–879.
7. Cusson, J. R., Thibault, G., Cantin, M. & Larochelle, P. (1990) Clin. Exp.
Hypertens. A 12, 111–135.
8. Lin, K. F., Chao, J. & Chao, L. (1995) Hypertension 26, 847–853.
9. Lin, K. F., Chao, J. & Chao, L. (1999) Hypertension 33, 219–224.
10. Burcin, M. M., Schiedner, G., Kochanek, S., Tsai, S. Y. & O’Malley, B. W.
(1999) Proc. Natl. Acad. Sci. USA 96, 355–360.
11. Yan, C., Costa, R. H., Darnell, J. E., Jr., Chen, J. D. & Van Dyke, T. A. (1990)
EMBO J. 9, 869–878.
12. Sandig, V., Youil, R., Bett, A. J., Franlin, L. L., Oshima, M., Maione, D., Wang,
F., Metzker, M. L., Savino, R. & Caskey, C. T. (2000) Proc. Natl. Acad. Sci. USA
97, 1002–1007.
13. Oka, K., Pastore, L., Kim, I. H., Merched, A., Nomura, S., Lee, H. J.,
Merched-Sauvage, M., Arden-Riley, C., Lee, B., Finegold, M., et al. (2001)
Circulation 103, 1274–1281.
14. Reddy, A. K., Taffet, G. E., Madala, S., Michael, L. H., Entman, M. L. &
Hartley, C. J. (2003) Ultrasound Med. Biol. 29, 379–385.
15. Dewald, O., Frangogiannis, N. G., Zoerlein, M., Duerr, G. D., Klemm, C.,
Knuefermann, P., Taffet, G., Michael, L. H., Crapo, J. D., Welz, A. & Entman,
M. L. (2003) Proc. Natl. Acad. Sci. USA 100, 2700–2705.
13794 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0506807102
from iANP.HD in mice, and it has been well tolerated by
patients. By linking ANP gene expression to the administration
of Mfp, iANP.HD allows the multisystem effects of this peptide
to occur through administration of a single drug, a feat that is
impossible with current antihypertensive therapy. Also, incor-
porating everything into a HD-Ad vector or a formulated direct
DNA injection obviates the need for repeated i.v. delivery of
synthetic ANP and ensures the persistence of the antihyperten-
sive properties of ANP.
We thank Mark Burcin and Thuy Pham for technical assistance, Ming-
Jer Tsai for helpful discussion, Gael Elliston for critical reading of the
manuscript, and Valentis for providing plasmid pEP1422. This work was
supported by National Institutes of Health Grants HD-7857 (to B.W.O.)
and HL-59314 (to L.C.).
16. Hartley, C. J., Reddy, A. K., Madala, S., Martin-McNulty, B., Vergona, R.,
Sullivan, M. E., Halks-Miller, M., Taffet, G. E., Michael, L. H., Entman, M. L.
& Wang, Y. X. (2000) Am. J. Physiol. 279, H2326–H2334.
17. Schlager, G. & Sides, J. (1997) Lab. Anim. Sci. 47, 288–292.
18. Beltowski, J. & Wojcicka, G. (2002) Med. Sci. Monit. 8, RA39–RA52.
19. Phillips, M. I. (1999) Hypertension 33, 8–13.
20. Leckie, B. J. (2001) J. Hypertens. 19, 1607–1613.
21. Melo, L. G., Steinhelper, M. E., Pang, S. C., Tse, Y. & Ackermann, U. (2000)
Physiol. Genomics 3, 45–58.
22. Zahabi, A., Picard, S., Fortin, N., Reudelhuber, T. L. & Deschepper, C. F.
(2003) J. Biol. Chem. 278, 47694–47699.
23. Barbee, R. W., Harrison-Bernard, L. M. & Carmines, P. K. (1992) Peptides 13,
1181–1185.
24. Opgenorth, T. J. & Novosad, E. I. (1990) Eur. J. Pharmacol. 191, 351–357.
25. Kohno, M., Yasunari, K., Yokokawa, K., Murakawa, K., Horio, T. & Takeda,
T. (1991) J. Clin. Invest. 87, 1999–2004.
26. Tremblay, J., Desjardins, R., Hum, D., Gutkowska, J. & Hamet, P. (2002) Mol.
Cell. Biochem. 230, 31–47.
27. Floras, J. S. (1995) Am. J. Physiol. 269, R406–R412.
28. Chiu, T. & Reid, I. A. (1996) J. Pharmacol. Exp. Ther. 278, 793–799.
29. Silberbach, M., Gorenc, T., Hershberger, R. E., Stork, P. J., Steyger, P. S. &
Roberts, C. T., Jr. (1999) J. Biol. Chem. 274, 24858–24864.
30. Takimoto, E., Champion, H. C., Li, M., Belardi, D., Ren, S., Rodriguez, E. R.,
Bedja, D., Gabrielson, K. L., Wang, Y. & Kass, D. A. (2005) Nat. Med. 11,
214–222.
31. Morral, N., Parks, R. J., Zhou, H., Langston, C., Schiedner, G., Quinones, J.,
Graham, F. L., Kochanek, S. & Beaudet, A. L. (1998) Hum. Gene Ther. 9,
2709–2716.
32. Morsy, M. A., Gu, M., Motzel, S., Zhao, J., Lin, J., Su, Q., Allen, H., Franlin,
L., Parks, R. J., Graham, F. L., et al. (1998) Proc. Natl. Acad. Sci. USA 95,
7866–7871.
33. Abruzzese, R. V., Godin, D., Mehta, V., Perrard, J. L., French, M., Nelson, W.,
Howell, G., Coleman, M., O’Malley, B. W. & Nordstrom, J. L. (2000) Mol. Ther.
2, 276–287.
Schillinger et al.