829
Role of the Renin-Angiotensin System in the Control
of Vasopressin Secretion in Conscious Dogs
Virginia L. Brooks, Lanny C. Keil, and Ian A. Reid
From the Department of Physiology, University of California, San Francisco, California, and the Ames Research Center,
Moffett Field, California
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
SUMMARY. The present studies were designed to evaluate the physiological significance of
angiotensin II in the control of vasopressin secretion in conscious dogs. They demonstrated that
exogenous angiotensin II (10 ng/kg per min) increased vasopressin secretion more when the
pressor effect of angiotensin II was abolished. The fact that endogenous angiotensin II levels are
normally increased without an increase in arterial pressure suggests that angiotensin II may play
a greater role in the control of vasopressin secretion than was previously thought. The present
study also evaluated the role of endogenous angiotensin II in the control of vasopressin secretion
during sodium depletion, a state in which angiotensin II levels are elevated. Intracarotid infusion
of a low dose of the angiotensin II antagonist, saralasin, decreased plasma vasopressin concentration, suggesting that endogenous angiotensin II acts in an area of the brain perfused by the carotid
arteries to stimulate vasopressin secretion in sodium-deprived dogs. Finally, the present experiments evaluated the role of angiotensin II in baroreceptor reflex control of vasopressin secretion.
Baroreflex function was assessed by examining the relationship between the change in blood
pressure and the log of the change in vasopressin secretion over a range of blood pressure levels.
Exogenous angiotensin II (10 ng/kg per min) altered baroreflex function by causing a shift of this
relationship to a higher pressure level in sodium-replete dogs. In sodium-depleted dogs, inhibition
of the renin-angiotensin system with saralasin or captopril produced an opposite shift. These
results suggest that endogenous angiotensin II may be necessary for the maintenance of normal
baroreflex control of vasopressin secretion during sodium depletion. Collectively, these results
support the hypothesis that endogenous angiotensin II plays a role in the control of vasopressin
secretion. (Circ Res 58: 829-838, 1986)
THERE have been reports that intravenous infusion
of angiotensin II (All) stimulates vasopressin secretion (Bonjour and Malvin, 1970; Unlich et al., 1975;
Padfield and Morton, 1977; Ramsay et al., 1978a;
Reid et al., 1982), but it is not clear whether this
effect of All is physiologically important. The purpose of the present investigation was to examine the
physiological significance of All in the control of
vasopressin secretion. Three experimental approaches were used.
Previous attempts to analyze the role of All in the
control of vasopressin secretion have evaluated
whether doses of All that produce plasma levels
within the phsyiological range are sufficient to stimulate vasopressin release. For example, Ramsay et
al. (1978) reported that physiological concentrations
of All were capable of increasing plasma vasopressin
levels. However, we and others have observed that
supraphysiological doses are required (Padfield and
Morton, 1977; Reid et al., 1982). Possibly, high
levels of AH are needed because the pressor effect
of All, by stimulating baroreceptor pathways
(Schrier et al., 1979), inhibits the release of vasopressin. This possibility was examined recently by
Mitchell et al. (1982), who reported that intravenous
AH injections inhibited the activity of supraoptic
magnocellular neurons. However, after sinoaortic
denervation, All increased the firing rate of these
neurons (Mitchell et al., 1982). These results were
interpreted as evidence that the pressor effect of AH,
by acting via baroreceptors, counteracts the stimulatory action of All on vasopressin release. The first
set of experiments in the present study were designed to test this hypothesis more directly. In these
experiments, the effect of All on plasma vasopressin
concentration was examined in conscious dogs while
the increase in blood pressure produced by All was
eliminated by simultaneous infusion of the vasodilator, nitroprusside.
Another approach that has been used to study the
role of All in the control of vasopressin secretion
involves the administration of antagonists of the
renin-angiotensin system to animals with elevated
plasma All levels. In previous studies, this approach
has provided evidence that All is involved in the
release of vasopressin that follows injection of isoproterenol and furosemide, two agents which increase renin secretion (Ramsay et al., 1978b; Siegel
et al., 1979; Knepel and Meyer, 1980). In addition,
it has been suggested that All may contribute to the
elevated vasopressin levels in dehydrated rats (Yamaguchi et al., 1980, 1982; Keil et al., 1983). How-
Circulation Research/Vol. 58, No. 6, June 1986
830
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ever, All has not been implicated in the increase in
vasopressin levels provoked by hemorrhage (Claybaugh and Share, 1972, 1973). Our second purpose,
in the present study, was to evaluate the role of All
in the control of vasopressin secretion during sodium
depletion, another state in which plasma All levels
are elevated. In these experiments, the All antagonist, saralasin, was infused into the carotid or vertebral arteries of chronically prepared conscious
dogs.
The last group of experiments extended the experiments described above in which saralasin was
infused into the blood supply to the brain of sodiumdepleted dogs. In those experiments, saralasin was
also infused intravenously. As expected, this treatment caused hypotension, emphasizing the important role of AH in the maintenance of blood pressure
during sodium depletion. However, saralasin infusion had no effect on plasma vasopressin concentration, despite the fact that hypotension normally
stimulates vasopressin secretion (Schrier et al.,
1979). Experiments therefore were performed to
determine why plasma vasopressin concentration
did not increase when blood pressure was decreased
by saralasin.
Two hypotheses were tested. One is that baroreflex control of vasopressin secretion is altered by
sodium depletion. Sodium-deprived dogs do not
exhibit tachycardia in response to carotid occlusion
(Rocchini et al., 1977; Szilagyi et al., 1981). In addition, the reflex increase in renal nerve activity
produced by hypotension is reduced in sodiumdepleted dogs (Takishita and Ferrario, 1982). To
evaluate whether baroreflex control of vasopressin
secretion is also affected, we compared the relationship between blood pressure and plasma vasopressin in sodium-deprived and sodium-replete dogs.
A second hypothesis is that saralasin blocks the
reflex release of vasopressin that is elicited by a fall
in blood pressure. More specifically, it is possible
that endogenous All contributes to the vasopressin
response to hypotension in sodium-depleted dogs,
and that saralasin blocks this action. To test the
hypothesis, we examined the relationship between
blood pressure and vasopressin secretion before and
after blockade of the renin-angiotensin system in
sodium-depleted dogs.
Methods
Mongrel dogs of either sex weighing between 16 and
32 kg were used.
Surgical Preparation
Subcutaneous injections of acepromazine maleate (1
mg/kg; Ayerst Laboratories, Inc.) were administered, and
then the dogs were anesthetized with pentobarbital sodium (15 mg/kg, iv). Silastic cannulas (0.5 mm i.d.) were
placed into a femoral artery and vein and were advanced
into the abdominal aorta and vena cava. The dogs then
were divided into three groups. In some dogs of the first
and third groups, right atnal catheters were placed via a
jugular vein, or left atrial catheters were placed via a left
thoracotomy. In the second group of dogs, both carotid
and both vertebral arteries were exposed, and small Tygon
cannulas (0 1 or 0 30 mm i.d.) were implanted, without
occluding blood flow, into each vessel. Each cannula was
held in place with a purse-string suture (Tevdek 4-0
cardiovascular suture). All catheters were led subcutaneously to an area between the scapulae, where they exited
and were protected by a nylon jacket (Medical Arts).
Vertebral and carotid catheters were flushed daily with
isotonic saline and filled with sodium heparin (1000
U/ml); the remaining catheters were flushed 3 times a
week. In most dogs, a minimum of 7 days elapsed before
the first experiment was performed. Dogs that were subjected to a thoracotomy were allowed at least 2 weeks to
recover. All dogs received penicillin and streptomycin
(Henry Schein, Inc.; 3 ml/day, im) for 5 days after surgery.
After surgery, the dogs were fed one of two diets. Sodiumreplete dogs were fed a diet which provided approximately
70 mEq/day sodium. Sodium-deprived dogs were fed two
cans H/D Prescription diet (Hills Pet Products), which
provided approximately 10 mEq/day sodium, and were
injected im with furosemide (20 mg) on alternate days for
1 week.
Experimental Protocols
On the day of the experiment, the dogs were brought
to the laboratory where they stood, loosely restrained by
a nylon sling. Mean and pulsatile arterial pressure and
heart rate were measured via the femoral arterial cannula
with a Statham Strain gauge and a Grass polygraph. Right
and left atrial pressures also were measured. Transducers
were placed so that the zero pressure point was at the
level of the heart. Because of the difficulty of accurately
determining the zero reference point for atrial pressures,
we analyzed the atrial pressure data by paired f-test or by
analysis of variance for repeated measures, which adjusts
for differences in baseline values between dogs (Winer,
1971), and by analyzing the change in atrial pressure from
control values.
After a 45-minute equilibration period, one of the following three groups of experiments was performed.
Group I. The Effect of AH and Nitroprusside on Vasopressin
Secretion
Protocol I. These experiments were performed in sodium-replete dogs (n = 7) Immediately after collection of
a control arterial blood sample, an infusion (0.5 ml/min)
of All (10 ng/kg per min; Schwarz-Mann or Bachem)
dissolved in 5% dextrose was begun Fifteen minutes later,
another blood sample was collected, and an intravenous
infusion of nitroprusside (Nipride, Roche Laboratories)
dissolved in 5% dextrose at 0.3 Mg/kg per min was begun
while the All infusion was continued. Every 15 minutes,
an arterial blood sample was collected, and the dose of
nitroprusside was increased until a total of four doses had
been given (0.3, 0.6, 1.5, 3.0 Mg/kg per min). All blood
samples (13 ml) were replaced with an equal volume of
isotonic saline. In control experiments, the dogs received
the nitroprusside infusions without All (n = 7), and also
received intravenous infusions (0.5 ml/min) of the drug
vehicle (n = 5).
Protocol 11. Experiments were performed to determine
the effect of eliminating the pressor effect of All with
another vasodilator, hydralazine (Apresoline HC1, Taylor
Pharmacal Co.). Two control arterial blood samples, 15
minutes apart, were collected, and one of three 30-minute
infusions was begun. Dogs received either All alone (10
Brooks et a/./Angiotensin II and Vasopressin Secretion
831
ng/kg per mm) (n = 7) or the same dose of All with either
nitroprusside (3 MgAg per min) or hydralazine (15-20 mg
bolus). Blood samples (13 ml) were collected 15 and 30
minutes after the experimental infusion was begun. Thirty
minutes were allowed after terminating the experimental
infusions, and a recovery sample was collected. All and
nitroprusside were diluted with 5% dextrose.
group I, protocol I. A control arterial blood sample was
collected and an intravenous infusion of saralasin (1 jig/
kg per min) or captopril (20 MgAg per min following a
1 mg/kg pnme; E.R. Squibb & Sons, Inc.) dissolved in 5%
dextrose was begun. Fifteen minutes later, a second blood
sample was collected, and infusion of nitroprusside (0.3
Mg/kg per min) was begun, along with saralasin or captopril. Subsequently, the dose of nitroprusside was increased every 15 minutes until a total of four doses were
given. Blood samples were collected just before the dose
of nitroprusside was changed. In separate experiments,
sodium-deprived dogs also received nitroprusside alone.
The saralasin and nitroprusside experiment was also repeated in a group of sodium-replete dogs
In all experiments, blood was collected into chilled,
heparinized plastic tubes and centrifuged. Plasma was
analyzed for arginine vasopressin by radioimmunoassay
(Keil and Severs, 1977).
Group II. The Effect of lntracarotid, lntravertebral, and
lntranveous Infusion of Saralasin on Plasma Vasopressin
Concentration in Sodium-Deprived Dogs
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Protocol 1. After a control blood sample was obtained
[Sar\Ala8]-AII (saralasin; Bachem), dissolved in isotonic
saline was infused for 30 minutes into both carotid or
both vertebral arteries. Three doses (0.1, 0.3, and 1 /ig/kg
per min per pair of arteries) were infused, each on a
different day. The experimental infusion was followed by
a 45-minute recovery period. Subsequently, the same dose
of saralasin was infused intravenously for 30 minutes, and
the experiment ended after a final 45-minute recovery
period In some dogs, the intravenous infusion was performed first. Blood samples were collected immediately
before and 15 and 30 minutes after the start of the
saralasin infusions, and also at the end of each recovery
period.
Protocol II. In four dogs, the isotonic saline vehicle was
infused either into both carotid arteries, or into both
vertebral arteries and intravenously, as described m Protocol I.
Protocol III. Experiments were performed to test the
effectiveness of the saralasin infusions. First, AH (4 ng/kg
per min per pair of arteries) was infused for 15 minutes
into either both carotid or both vertebral arteries. After a
15rminute recovery period, an infusion of the lowest dose
of saralasin (0.1 MgAg per min) was begun into the same
vessels After 20 minutes, All (4 ng/kg per min per pair
of arteries) was again infused for 15 minutes, along with
saralasin Blood samples were collected before, at the end
of, and 15 minutes after, each All infusion.
Data Analysis
Groups 1 and III
In expenments in which four doses of nitroprusside
were infused, the relationship between blood pressure and
plasma vasopressin concentration was obtained. Since this
relationship is exponential, a linear relationship is generated when the log (change in plasma vasopressin concentration) is plotted vs. the change in arterial blood pressure.
The lines drawn in the figures were obtained by first
calculating the mean ± SEM of the change in blood pressure
and the log of the mean ± SEM of the change in plasma
vasopressin concentration for each dose of nitroprusside.
A line of best fit then was calculated for these four or five
mean values by least squares analysis, and this is depicted
graphically in the figures. For statistical purposes, a regression equation was calculated by least squares analysis of
the log (change in plasma vasopressin concentration) vs.
the change in blood pressure for all the individual values
collected in each group, and groups were compared by
testing for differences in slope and intercept (Winer, 1971).
In experiments in which one dose of nitroprusside or
hydralazine was administered with AH, differences between groups were determined by Student's Mest (Winer,
1971).
Group 111. Effect of Sodium Depletion and AH Blockade on
the Relationship Between Blood Pressure and Plasma Vasopressin Concentration
These experiments were performed in sodium-deprived
conscious dogs according to the protocol described in
FIGURE 1. Relationship between the change m
blood pressure and the log of the change m plasma
vasopressin concentration during intravenous infusion of nitroprusside (NP) (n = 7) or nitroprusside
and angwtensin II (NP + All) (n = 7). Control
blood pressures were 109 ± 1 (NP) and 105 ± 2
(NP + All) mm Hg. Control vasopressin levels were
2.0 + 0.4 (NP) and 21 ± 0.4 (NP + All) pg/ml. n
= number of dogs.
Change in
plasma
vasopressin
cone.
(pg/ml)
—40
—20
0
20
Change in blood pressure (mm Hg)
Circulation Research/Vol. 58, No. 6, June 1986
832
TABLE 1
Effect of Angiotensin II, Saralasin, and Captopril on
Baroreflex Function Relationships in Sodium-Replete and
-Depleted Conscious Dogs
Protocol
Slope
Intercept
r2
NP (replete)
NP + All (replete)
NP + SAR (replete)
NP (depleted)
NP + SAR (depleted)
NP + CAP (depleted)
-0.081
- 0 031*
-0.025'
- 0 054
- 0 046
- 0 049
-0.237
0 878*
0.260
0 070
- 0 766f
- 1 035f
0.63
0.46
0 26
0.57
0.45
0 54
Values are slopes and intercepts of the linear equation log
(change in plasma vasopressm concentration) = slope-(change in
blood pressure) + intercept All lines have significant slopes with
P < 0 001 NP = nitroprusside, AH = angiotensin II; SAR =
saralasin, CAP = captopril.
• P < 0.001 compared to NP-replete group
f P < 0 05 compared to NP-depleted group
Group H
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
One-way analysis of variance for repeated measures
(Winer, 1971) was used to evaluate differences within
groups. Where indicated, paired and Student's f-tests also
were performed (Winer, 1971).
Results
Group I: Effect of All on Plasma Vasopressin
Concentration during Concurrent Nitroprusside
Infusion
The relationship between blood pressure and
plasma vasopressin concentration obtained from
dogs treated with All and nitroprusside or nitroprusside alone is illustrated in Figure 1. Alone, All
increased blood pressure by 35 ± 4 mm Hg (P <
0.05) and increased plasma vasopressin concentration by 0.7 ± 0.2 pg/ml (P < 0.05). With increasing
doses of nitroprusside, blood pressure fell and
plasma vasopressin levels increased further. With
the highest dose of nitroprusside, blood pressure
had returned to control levels, and plasma vasopressin concentration was increased by 11.9 ± 1.9
pg/ml (P < 0.001). Thus, AH produced a larger
increase in vasopressin levels (P < 0.001) when the
pressor effect of All was abolished. Also illustrated
in Figure 1 is the relationship between blood pressure and plasma vasopressin concentration produced by nitroprusside infusion alone. Plasma vasopressin concentration rose as blood pressure decreased, and the log of the vasopressin change was
linearly related to the blood pressure fall (Fig. 1;
Table 1). All significantly increased the intercept of
this line (P < 0.001) while displacing it to the right
(Table 1).
In another group of experiments, a second vasodilator, hydralazine, was used to eliminate the pressor effect of AIL A 30-minute infusion of All alone
(n = 6) increased blood pressure from 97 ± 2 to 134
± 2 mm Hg (P < 0.001) and increased plasma
vasopressin levels from 3.1 ± 0.3 to 5.1 ± 0.9 pg/
ml (P = 0.05). In the presence of hydralazine (« =
8), All had no effect on arterial pressure (101 ± 3 to
103 ± mm Hg; P > 0.10) and increased plasma
vasopressin concentration from 3.4 ± 0.3 to 38.8 ±
7.3 pg/ml (P < 0.001). Similarly, with nitroprusside
and All infusion (n = 5), blood pressure did not
change (103 ± 2 to 104 ± 2 mm Hg; P > 0.10), and
plasma vasopressin increased from 3.0 ± 0.4 to 19.1
± 5.9 pg/ml (P < 0.05). Thus, All produced a larger
increase in plasma vasopressin concentration when
the pressor response was eliminated by either hydralazine (P < 0.001) or nitroprusside (P < 0.05).
Group II: Effect of Intravertebral, Intracarotid,
and Intravenous Saralasin Infusion on Plasma
Vasopressin Concentration in Sodium-Deprived
Dogs
Neither intravertebral nor intravenous infusion of
any dose of saralasin decreased plasma vasopressin
TABLE 2
Effect of Intravertebral, Intracarotid, and Intravenous Saralasin Infusion on Plasma Vasopressin
Concentration (pg/ml) in Conscious, Sodium-Deprived Dogs
Dose of
saralasin
per min)
Control
0 1(6)
0.3 (6)
10(4)
2 0 ± 0.5
2.7±08
28± 1 0
Control
0 1(5)
0.3 (6)
10(4)
1 2 ±0.1
2.1 ± 0 4
1 7 + 01
Control
0 1(9)
0.3 (9)
1.0(7)
2.5 ± 0 5
2.9 ± 0 4
2.5 + 0 7
Intravertebral
22 ± 0 6
2 1 ±0.6
23 ± 0 7
2.4 ± 0.7
3.0 ± 0 7
2 2 ± 0.5
Intracarotid
2.5 ±1.1
2.4 ± 0 5
1 1±02
20±07
2.1+0.3
1.2 ±0.2
Intravenous
2.4 ±0.7
2.7 ±0.4
30 ± 1.0
24 + 06
2.8 ± 0 4
3.3 ±1.1
Recovery
P
2 8 ± 0.8
3.4 ± 0.7
3.6 ± 0 8
<0 05
NS
NS
Recovery
18±03
2.6 ± 05
1.6 ± 0.2
NS
NS
<0 05
Recovery
3.3 ± 0 9
3.1 ± 0 6
3.1 ± 0.9
NS
NS
NS
Values are means ± SEM of plasma vasopressin concentration (pg/ml) Numbers in parentheses are
the number of dogs in each group NS = no significant change (P > 0 05).
Brooks et a/./Angiotensin II and Vasopressin Secretion
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
concentration (Table 2). The two lowest doses of
saralasin infused into the carotid arteries also failed
to affect vasopressin levels; however, the highest
dose (1 MgAg per min) produced a significant decrease (Table 2) (P < 0.05). The effect of intravertebral, intracarotid, and intravenous saralasin infusion on mean arterial pressure is summarized in
Table 3. Saralasin produced a dose-dependent decrease in blood pressure with each route of administration (Table 3).
The saline vehicle was infused intravenously and
into the carotid arteries (two dogs) or intravenously
and into the vertebral arteries (two dogs). Neither
intracarotid, intravertebral, nor intravenous saline
infusion altered plasma vasopressin concentration
or blood pressure.
We performed experiments in five dogs to verify
adequate All blockade with the doses of saralasin
used in these experiments. Infusion of All (4 ng/kg
per min) into the carotid arteries increased plasma
vasopressin levels from 1.9 ± 0.6 to 3.7 ± 1.1 pg/
ml (P = 0.01). After a 20-minute infusion of the
lowest dose of saralasin (0.1 MgAg per min), All had
no effect on plasma vasopressin concentration (2.0
± 0.5 to 2.1 ± 0.7 pg/ml; P > 0.10).
Group III: The Effect of Sodium Depletion and
Blockade of the Renin-Angiotensin System on
the Relationship between Blood Pressure and
Plasma Vasopressin Concentration
The relationship between plasma vasopressin
concentration and blood pressure obtained in sodium-deprived dogs from group II treated with three
intravenous doses of saralasin is compared to the
relationship obtained from sodium-replete dogs
treated with nitroprusside (Fig. 2). Although blood
pressure fell by up to 25 mm Hg, vasopressin levels
did not increase with saralasin infusion. This result
is in marked contrast to the increase in vasopressin
833
A
100
10
Change
V
1
in plasma
I
vasopressin
;
cone
(PO/ml)
n
L
or
i
I
—20
-10
B
—2
—30
—20
—10
Change in blood pressure (mm Hg)
FIGURE 2. Relationship between the change m blood pressure and the
change m plasma vasopressin concentration in (panel A) sodiumreplete dogs (n = 7) infused intravenously with four doses of nitroprusside (0.3, 0 6, 1 5 and 3 0 ng/kg per mm) and in (panel B) sodiumdepleted dogs (n = 7-9) infused intravenously with three doses of
saralasin (0.1, 0 3, and 1 0 ng/kg per min). Control values for panel
are A as in Figure 1. Control blood pressures and plasma vasopressin
concentrations for panel B in Tables 2 and 3 n = number of dogs
produced by nitroprusside in sodium-replete dogs
(group I) (Figs. 1 and 2).
The final groups of experiments examined why
vasopressin did not rise as blood pressure fell with
saralasin in sodium-depleted dogs. One hypothesis
TABLE 3
Effect of Intravertebral, Intracarotid, and Intravenous Saralasin Infusion on Arterial Blood
Pressure (mm Hg) in Conscious, Sodium-Deprived Dogs
Dose of
saralasin
(fig/kg
per min)
0 1(6)
0 3(6)
10(4)
Control
103 ± 1
104 ± 2
104 ± 2
Control
0 1(5)
0 3(6)
10(4)
102 ± 5
105 ± 3
103 ± 5
Control
0 1(9)
0 3(9)
10(7)
103 ±3
105 ±2
100 + 3
Intravertebral
98 + 2
96 ± 3
90 + 3
96 ± 3
90 ± 3
85 ± 3
Intracarotid
Recovery
P
104 ± 4
105 ± 2
101 ± 3
<0 05
<0 001
<0 001
Recovery
101 ± 5
98 + 4
92 ± 5
96 ± 4
93 ± 4
83 ± 5
Intravenous
97 ± 3
93 ± 3
81 ± 2
0
94 ± 3
88 ± 3
75 ± 3
101 ± 5
104 ± 4
101 ± 3
<0 05
<0 001
<0 01
Recovery
101 ± 2
101 ± 2
101 ± 3
<0 001
<0.001
<0 001
Values are means ± SEM of blood pressure (mm Hg) Numbers in parentheses are the number of
dogs in each group
834
Circulation Research/Vol 58, No. 6, June 1986
100 r
Change in 10
plasma
vasopressin
cone.
(pg/ml)
1
FIGURE 3. Relationship between the change in blood
pressure and the log of the change m plasma vasopressin concentration in sodium-replete [NP (replete)] (n =
7) and sodium-depleted [NP (deplete)] (n = 6) dogs
receiving intravenous infusions of nitroprusside. Control blood pressure values were 109 ± 1 (replete) and
110 ± 3 (depleted) mm Hg Control vasopressin levels
were 2.0 ± 0.4 (replete) and 1 9 + 0.5 (depleted) pg/
ml n = number of dogs
NP (deplete)
I
o1
—40
—20
0
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Change in blood pressure (mm Hg)
was that sodium depletion alters baroreflex control
of vasopressin secretion. The relationship between
blood pressure and plasma vasopressin concentration following nitroprusside in sodium-replete and
sodium-depleted dogs is summarized in Figure 3.
Hypotension induced similar increases in vasopressin in both groups of animals (Fig. 3), and the
regression lines were not different (Table 1).
A second hypothesis was that saralasin affects
baroreceptor control of vasopressin secretion such
that a smaller rise in vasopressin is elicited for a
given decrease in blood pressure. To test this hypothesis, we examined the relationship between
blood pressure and vasopressin in the presence and
absence of saralasin (Fig. 4). In saralasin-treated
dogs, hypotension elicited a rise in vasopressin secretion; however, the relationship between blood
pressure and vasopressin concentration was shifted
to the left (Fig 4) compared to dogs that received
nitroprusside alone. There was a significant decrease
in intercept without a change in slope (Table 1).
Inhibition of All production with captopril had a
similar effect on the line relating blood pressure to
the log of the change in vasopressin concentration
(Fig. 5; Table 1). Saralasin did not affect the intercept
of this line in sodium-replete dogs (Table 1; Fig. 6),
although the slope was slightly depressed (Table 1).
Discussion
Does the Pressor Effect of Infused All Inhibit
Vasopressin Secretion?
Bonjour and Malvin (1970) presented the first
direct evidence that increases in plasma All levels
stimulate vasopressin secretion. Numerous subsequent reports demonstrated that mtracerebroventricular injection of All also increases vasopressin release (Mouw et al., 1971; Keil et al., 1975); however,
100^
Change in
plasma
vasopressin
cone.
(pg/ml)
10
NP + SAR
01
—60
—40
FIGURE 4. Relationship between the change w blood
pressure and the log of the change in plasma vasopressin concentration in sodium-depleted dogs receiving
intravenous infusion of nitroprusside (NP) (n = 6) or
nitroprusside and saralasin (NP + SAR) (n = 7). Control blood pressures were 110 ± 3 (NP) and 108 ± 3
(NP + SAR) mm Hg. Control vasopresstn levels were
19 + 05 (NP) and 25 ± 05 (NP + SAR) pg/ml n =
number of dogs.
v
—20
Change in blood pressure (mm Hg)
Brooks et a/./Angiotensin II and Vasopressin Secretion
835
100
Change in 10
plasma
vasopressin
cone.
FIGURE 5. Relationship between the change m blood
pressure and the log of the change m plasma vasopressin concentration in sodium-depleted dogs receiving
intravenous infusion of nitroprusside (NP) (n = 6) or
mtroprusside and captopnl (NP + CAP) (n = 5) from
control blood pressures of 110 ± 3 (NP) and 107 ± 3
(NP + CAP) mm Hg and control vasopressin concentrations of 1.9 ± 0.5 (NP) and 2 6 ± 0 4 (NP + CAP)
pg/ml n = number of dogs.
NP + CAP
0.1
-60
I
-40
—20
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Change in blood pressure (mm Hg)
the stimulatory effect of systemic All administration
originally observed by Bonjour and Malvin has
proved difficult to duplicate. Part of the inconsistency may be due to the fact that some experiments
have been performed in anesthetized animals. In
these experiments, intravenous infusion of a wide
range of doses of All has consistently failed to
stimulate vasopressin secretion (Claybaugh et al.,
1972; Shade and Share, 1975; Cadnapaphornchai et
al., 1975). Another explanation may be species differences, since rats, for example, appear to be relatively insensitive to All (Knepel and Meyer, 1980).
More consistent results have been obtained in
experiments performed in conscious dogs and human subjects. Most investigators agree that intravenous AH infusion can increase vasopressin release
(Bonjour and Malvin, 1970; Ramsay et al., 1978;
Uhlich et al., 1975); however, in some cases, large
doses of All were required (Reid et al., 1982; Padfield and Morton, 1977), making the physiological
significance of this effect uncertain. It is possible
that high levels of exogenous All are needed to
increase vasopressin secretion because the pressor
effect of AH counteracts a direct stimulatory action
of All on vasopressin secretion. A major finding of
the present study is that intravenous All infusion
produces a larger increase in plasma vasopressin
concentration when the pressor effect of All is eliminated by a simultaneous nitroprusside infusion.
Thus, the rise in blood pressure does appear to
interfere with a stimulatory action of All on vasopressin release. Since increases in endogenous All
levels usually occur without increases in blood pressure, the stimulatory effect of All on vasopressin
100 r
Change in
plasma
vasopressin
cone,
(pg/ml)
10
FIGURE 6. Relationship between the change in
blood pressure and the log of the change in
plasma vasopressin concentration m sodiumreplete dogs receiving intravenous infusion of
nitroprusside (NP) (n = 7) or nitroprusside and
saralasm (NP + SAR) (n = 6) Control values
for blood pressure were 109 ± 1 (NP) and 102
± 4 (NP + SAR) mmHg and for plasma vasopressin concentration were 20 ± 0.4 (NP) and
3 5 ± 0 6 (NP + SAR) pg/ml. n = number of
dogs
NP + SAR
1
J
0.1
—40
—20
Change in blood pressure (mm Hg)
0
836
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
secretion would normally be unopposed. These results therefore suggest that All may play a greater
role in the control of vasopressin secretion than was
previously thought.
It is possible that nitroprusside infusion increased
the vasopressin response to All by some action other
than by counteracting the pressor response. This
seems unlikely, however, since hydralazine, another
vasodilator, also enhanced the vasopressin response.
Furthermore, in preliminary experiments, we observed that the vasopressin response to nitroprusside is eliminated by denervation of both high pressure (carotid sinus and aortic arch) and low pressure
(cardiopulmonary) baroreceptors: plasma vasopressin concentration increased during nitroprusside infusion from 5.7 ± 0.9 to 70 ± 25 pg/ml in intact
dogs (P < 0.05, n = 4) but was unaltered in denervated dogs [3.9 ± 0.9 to 7.9 ± 2.7; P > 0.10, n = 3
(Brooks, Quillen, Keil, and Reid, unpublished results)]. Thus, it appears that nitroprusside increases
vasopressin secretion by lowering blood pressure
and by activation of baroreceptor pathways. The
enhanced response was also not due to further
increases in plasma All levels produced by nitroprusside-induced decreases in blood pressure.
Plasma All levels were measured during All infusion
and were not altered by simultaneous nitroprusside
infusion (Brooks and Reid, 1986). It was shown that
nitroprusside lowers right atrial pressure below control levels, both when infused alone and when infused in combination with AH (Brooks and Reid,
1986). Thus, it is possible that the additional increase
in vasopressin secretion produced by nitroprusside
during All infusion was due to stimulation of right
atrial stretch receptors. Although this possibility cannot be definitely ruled out, in most cases the smallest
dose of nitroprusside decreased right atrial pressure
to the lowest level, and higher doses did not increase
pressure further As a result, there was no correlation
between the changes in right atrial pressure and
plasma vasopressin concentration during either nitroprusside (r22 = 0.08, P > 0.10) or nitroprusside
and All (r2 = 0.09, P > 0.10) infusion. Because a
significant correlation was observed between the
changes in arterial pressure and vasopressin concentration, this suggests, in agreement with the studies
by Brennan et al. (1971) and Schultz et al. (1982),
that decreases in right atrial pressure may not have
an important influence on vasopressin secretion.
The lack of a correlation also suggests that the
enhanced vasopressin response to All produced by
nitroprusside was not due to a fall in right atrial
pressure Finally, changes in left atrial pressure
tended to parallel changes in mean arterial pressure.
All increased left atrial pressure, and simultaneous
nitroprusside infusion decreased left atrial pressure
to control levels (Brooks and Reid, 1986). Since
changes in left atrial pressure may have a profound
influence on vasopressin release (Schultz et al.,
1982), it is probable that nitroprusside amplified the
Circulation Research/Vol. 58, No. 6, June 1986
vasopressin response by lowering both elevated arterial and left atrial pressures back to control levels.
A second finding of the present study is that the
relationship between blood pressure and plasma
vasopressin concentration produced by nitroprusside infusion is shifted to the right during All infusion. This suggests that the effects of All to reset
baroreflex control of vasopressin secretion to a
higher pressure level are similar to its effects on
baroreflex control of heart rate and ACTH (Brooks
and Reid, 1986). The slope of the function curve
also was depressed during All infusion, but this may
be because baroreflex function curves are sigmoid,
and a different range of pressures was studied during All and nitroprusside compared to nitroprusside
alone.
The mechanism of this shift with All is not clear,
but we offer three possibilities. First, it has been
demonstrated that arterial stretch receptors rapidly
reset to sustained changes in blood pressure, with a
shift in the function curves relating blood pressure
to baroreceptor activity in the direction of the pressure change (Dorward et al., 1982; Munch et al.,
1983; Heesch et al., 1984). Therefore, it is possible
that the pressor effect of All caused the shift in the
function curves. If this is the case, it would be
expected that All would produce a larger increase
in plasma vasopressin concentration at control pressure in the experiment in which the pressor effect
of AH was slowly reduced with increasing doses of
nitroprusside, compared to the experiment in which
All and the highest dose of nitroprusside were infused simultaneously, and pressure never changed.
However, similar increases in plasma vasopressin
concentration were observed, suggesting that the
vasopressin rise and the shift of the function curves
were not dependent on the blood pressure increase.
Second, it is possible that All acts on the brain to
stimulate vasopressin secretion, and as blood pressure falls when nitroprusside is simultaneously infused, there is an additional and independent stimulation which is mediated by the baroreceptors.
Thus, at any given blood pressure, vasopressin secretion would be higher and the relationship between blood pressure and vasopressin would be
shifted to the right.
Another possibility is that All acts somewhere
within the baroreflex loop to alter baroreceptor control of vasopressin release. The fact that All also
affects baroreceptor control of heart rate and ACTH
(Brooks and Reid, 1986) and baroreflex control of
lumbar sympathetic activity (Guo and Abboud,
1984; Schmid et al., 1985) is consistent with this
possibility, and suggests that All may act early in
the baroreflex loop. However, there is evidence that
All does not exert a direct action on either the carotid
sinus or aortic arch baroreceptors (McCubbin et al.,
1957; Lumbers et al., 1979; Guo and Abboud, 1984).
Therefore, All may act at an area of the brain which
is involved in cardiovascular control and which is
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Brooks et a/./Angiotensin II and Vasopressin Secretion
837
accessible to blood-borne All. The area postrema, a
circumventricular organ that lacks a blood-brain
barrier and communicates with the nucleus tractus
solitarius, an area known to be involved in cardiovascular regulation, is a possible candidate. In this
regard, it is interesting that infusion of All into the
vertebral arteries, which perfuse the area postrema,
increases blood pressure and left atrial pressure
(Lowe and Scroop, 1969; Ferrario et al., 1970; Reid
et al., 1982). Increases in pressure should inhibit
vasopressin secretion (Schrier et al., 1979); however,
we reported previously that vasopressin levels either
were unaffected or increased (Reid et al., 1982),
suggesting that baroreflex control of vasopressin
release is altered during intravertebral AH infusion.
fects the reflex release of vasopressin elicited by a
fall in blood pressure in sodium-deprived dogs. This
possibility is supported in the present experiments
by the finding that All blockade with either saralasin
or captopril produced a parallel shift to the left in
the relationship between blood pressure and the log
of plasma vasopressin concentration without affecting the slope or sensitivity of the relationship. Thus,
during All blockade, blood pressure was required to
fall to a lower level before vasopressin release was
elicited, suggesting that endogenous All acts to reset
the baroreflex during sodium depletion.
Endogenous AH levels are increased in sodiumdepleted animals; therefore, it may seem contradictory that resting vasopressin secretion is not increased, and that baroreflex function curves are not
shifted to higher pressure levels in sodium-depleted
dogs compared to sodium-replete dogs (Fig. 3),
thereby mimicking the effects of exogenous All infusion (Fig. 1). This result suggests that some other
factor is shifting the curves to a lower pressure level
during sodium depletion, and that All acts to maintain a normal relationship between blood pressure
and vasopressin by shifting the relationship back to
the right. The factors that offset baroreflex activity
during sodium depletion were not investigated in
the current study, but they probably are related to
the decreased blood volume of these animals, and
may include exaggerated vagal afferent activity
(Szilagyi et al., 1981; Takishita and Ferrario, 1982).
Thus, the role that the chronically elevated levels of
All play in maintaining normal baroflex control of
vasopressin secretion is very similar to the role All
plays in maintaining normal blood pressure via its
vasoconstrictor effect.
Saralasin slightly depressed the slope of the relationship between blood pressure and vasopressin
concentration in sodium-replete dogs. The antagonist appeared to diminish the reflex release of vasopressin observed with the largest falls in blood
pressure, suggesting that increases in plasma All
levels resulting from large decreases in blood pressure may contribute to the vasopressin rise produced
by hypotension in sodium-replete dogs.
The mechanism of the shift produced by saralasin
or captopril in sodium-depleted animals is unknown. It was shown that when blood pressure is
reduced by nitroprusside, plasma All levels increase
(Brooks and Reid, 1986). These elevated All levels
may then act on the brain to release vasopressin. It
is possible that saralasin blocked this action and
reduced the vasopressin response to a given decrease
in pressure. The result would be a shift to the left in
the relationshiop between blood pressure and vasopressin.
A second possible mechanism for the shift produced by All blockade is that the elevated levels of
All in sodium-depleted dogs act within the baroreflex loop to maintain normal control of vasopressin
release, and that saralasin and captopril block this
Does Endogenous Angiotensin II Affect
Vasopressin Levels and Baroreflex Control of
Vasopressin Secretion in Sodium-Depleted
Dogs?
It previously has been shown that intracarotid All
infusion stimulates vasopressin secretion, whereas
intravertebral infusion of All is either ineffective or
increases plasma vasopressin levels (Reid et al.,
1982). A second purpose of the present experiments
was to evaluate whether the elevated All levels in
sodium-depleted dogs act via either the carotid or
vertebral circulation to affect vasopressin release. In
these experiments, the competitive All antagonist
saralasin was infused into the carotid or vertebral
arteries of conscious sodium-deprived dogs.
Intracarotid infusion of the two lowest doses of
saralasin had no effect on plasma vasopressin concentration; however, the highest dose significantly
reduced plasma vasopressin levels. Because neither
intravenous nor intravertebral infusion of the same
dose of saralasin depressed plasma vasopressin levels, it is possible that chronic elevations in endogenous All stimulate vasopressin secretion in sodiumdeprived dogs by acting in an area of the brain
perfused by the carotid arteries.
It is significant that intravertebral and intravenous
infusion of saralasin generally did not affect vasopressin concentration. This result was somewhat
surprising, since saralasin also produced large decreases in blood pressure, and hypotension normally
stimulates vasopressin secretion (Schrier et al.,
1979). The final group of experiments examined
why vasopressin secretion did not increase during
intravenous saralasin infusion in sodium-depleted
conscious dogs.
One possibility is that some aspect of sodium
depletion alters baroreflex function. However, a
comparison of the relationship between blood pressure and plasma vasopressin concentration revealed
no differences between sodium-replete and sodiumdepleted dogs. Thus, the release of vasopressin induced by decreases in blood pressure is unchanged
by sodium depletion.
Another possibility is that saralasin infusion af-
Circulation Research/Vol. 58, No. 6, June 1986
838
effect. Thus, in the presence of All blockade, the
baroreflex setpoint is decreased and blood pressure
must fall to a lower level before vasopressin secretion is stimulated. Further experiments are required
to distinguish between these possibilities.
Supported by N1H Grant HL 29714 and American Heart Association, California Affiliate.
Dr. Brooks' present address is Department of Physiology, Oregon
Health Sciences University Portland, Oregon 97201 (to whom correspondence and reprint requests should be addressed)
Received March 26, 1985, accepted for publication March 14,
1986
References
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Bonjour JP, Malvm RL (1970) Stimulation of ADH release by the
renin-angiotensin system Am J Physiol 218: 1555-1559
Brennan LA, Malvin RL, Jochim KE, Roberts DE (1971) Influence
of right and left atnal receptors on plasma concentrations of
ADH and renin Am J Physiol 221: 273-278
Brooks VL, Reid IA (1986) Interaction between angiotensin II and
the baroreceptor reflex in the control of ACTH secretion and
heart rate in conscious dogs Circ Res 58: 816-828
Cadnapaphornchai P, Boykm J, Harbottle JA, McDonald KM,
Schner RW (1975) Effect of angiotensin II on renal water
excretion Am J Physiol 228: 155-159
Claybaugh JR, Share L (1972) Role of the renin-angiotensin
system in the vasopressin response to hemorrhage Endocrinology 90: 453-460
Claybaugh JR, Share L (1973) Vasopressin, renin and cardiovascular responses to continuous slow hemorrhage. Am J Physiol
224: 519-523
Claybaugh JR, Share L, Shimizu K (1972) The inability of infusions of angiotensin to elevate the plasma vasopressin concentration in the anesthetized dog Endocrinology 90: 1647-1652
Dorward PK, Andresen MC, Burke SL, Oliver JR, Korner PI (1982)
Rapid resetting of the aortic baroreceptors in the rabbit and its
implications for short-term and longer term reflex control Circ
Res SO: 428-439
Ferrano CM, Dickinson CJ, McCubbin JW (1970) Central vasomotor stimulation by angiotensin Clin Sci 39: 239-245
Guo GB, Abboud FM (1984) Angiotensin II attenuates baroreflex
control of heart rate and sympathetic activity Am J Physiol
246: H80-H89
Heesch CM, Thames MD, Abboud FM (1984) Acute resetting of
carotid sinus baroreceptors I. Dissociation between discharge
and wall changes Am J Physiol 247: H824-H832
Keil LC, Severs WB (1977) Reduction in plasma vasopressin levels
of dehydrated rats following acute stress Endocrinology 100:
30-38
Keil LC, Summy-Long J, Severs WB (1975) Release of vasopressin
levels by angiotensin II Endocrinology 96: 1063-1065
Keil LC, Barbella YR, Dundore RL, Wurpel JND, Severs WB (1983)
Vasopressin release induced by water deprivation: Effects of
centrally administered saralasin Neuroendocnnology 37: 401405
Knepel W, Meyer DK (1980) Role of the renin-angiotensin system
in isoprenalme-induced vasopressin release J Cardiovasc Pharmacol 2: 815-824
Lowe RD, Scroop GC (1969) The cardiovascular response to
vertebral artery infusions of angiotensin in the dog. Clin Sci
37: 593-603
Lumbers ER, McCloskey DI, Potter EK (1979) Inhibition by an-
giotensin II of baroreceptor-evoked activity in cardiac vagal
efferent nerves in the dog J Physiol (Lond) 294: 69-80
McCubbin JW, Page IH, Bumpus FM (1957) Effect of synthetic
angiotensin on the carotid sinus circulation Circ Res 5: 458460
Mitchell LD, Barron K, Brody MJ, Johnson AK (1982) Two possible
actions for circulating angiotensin II m the control of vasopressin release Pephdes 3: 503-507
Mouw D, Bonjour JP, Malvin RL, Vander A (1971) Central action
of angiotensin in stimulating ADH release. Am J Physiol 220:
239-242
Munch PA, Andresen MC, Brown AM (1983) Rapid resetting of
aortic baroreceptors in vitro. Am J Physiol 244: H672-H680
Padfield PL, Morton JJ (1977) Effects of angiotensin II on argininevasopressin in physiological and pathological situations in man
J Endocnnol 74: 251-259
Ramsay DJ, Keil LC, Sharpe MC, Shinsako J (1978a) Angiotensin
II infusion increases vasopressin, ACTH, and 11-hyroxycorhcosteroid secretion Am J Physiol 234: R66-R71
Ramsay DJ, Reid IA, Keil LC, Ganong WF (1978b) Evidence that
the effects of isoproterenol on water intake and vasopressin
secretion are mediated by angiotensin Endocrinology 100: 5459
Reid IA, Brooks VL, Rudolph CD, Keil LC (1982) Analysis of the
actions of angiotensin on the central nervous system of conscious dogs. Am J Physiol 243: R82-R91
Rocchini AP, Cant JR, Barger AC (1977) Carotid sinus reflex in
dogs with low- to high-sodium intake. Am J Physiol 233: H196H202
Schmid PG, Guo GB, Abboud FM (1985) Different effects of
vasopressin and angiotensin II on baroreflexes. Fed Proc 44:
2388-2392
Schner RW, Berl T, Anderson RJ (1979) Osmotic and nonosmotic
control of vasopressin release Am J Physiol 236: F321-F332
Schultz HD, Fater DC, Sundet WD, Geer PG, Goetz KL (1982)
Reflexes elicited by acute stretch of atnal vs pulmonary receptors in conscious dogs. Am J Physiol 242: H1065-H1076
Shade RE, Share L (1975) Vasopressin release during nonhypotensive hemorrhage and angiotensin II infusion. Am J Physiol
228: 149-154
Siegel SR, Weitzman RE, Fisher DA (1979) Endogenous angiotensin stimulation of vasopressin in the newborn lamb. J Clm
Invest 63: 287-293
Szilagyi JE, Masaki Z, Brosnihan KB, Ferrano CM (1981) Neurogenic suppression of carotid sinus reflexes by vagal afferents
in sodium-depleted dogs Am J Physiol 241: H255-H262
Takishita S, Ferrano CM (1982) Altered neural control of cardiovascular function in sodium-depleted dogs Hypertension
4(suppl II): 75-182
Uhhch E, Weber P, Eigler J, Groschel-Stewart V (1975) Angiotensin stimulated AVP-release in human. Clin Wochenschr 53:
177-180
Winer BJ (1971) Statistical Principles in Experimental Design.
New York, McGraw
Yamaguchi K, Hama H, Sakaguchi T, Negoro H, Kamoi K (1980)
Effects of intraventncular injection of Sar'-AIa'-angiotensin II
on plasma vasopressin level increased by angiotensin II and by
water depnvahon on conscious rats. Acta Endocnnol (Kbh) 93:
407-412
Yamaguchi K, Sakaguchi T, Kamoi K (1982) Central role of
angiotensin on the hyperosmolality- and hypervolaemia-induced vasopressin release in conscious rats. Acta Endocnnol
(Kbh) 101: 524-530
INDEX TERMS: Renin-angiotensin system • Baroreceptor reflex
• Saralasin • Captopnl • Brain • Carotid and vertebral artenes •
Nitroprusside
Role of the renin-angiotensin system in the control of vasopressin secretion in conscious
dogs.
V L Brooks, L C Keil and I A Reid
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1986;58:829-838
doi: 10.1161/01.RES.58.6.829
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1986 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/58/6/829
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/