- Email: [email protected]
Role of endogenous atrial natriuretic factor in the regulation of fetal cardiovascular and renal function
Role of endogenous atrial natriuretic factor in the regulation of fetal cardiovascular and renal function Cecilia Y. Cheung, PhD La Jolla, California The purpose of the present study was to determine whether endogenous atrial natriuretic factor participates in the maintenance of normal vascular pressure and renal function in ovine fetuses at 128 to 130 days' gestation. Circulating atrial natriuretic factor in the fetus was immunoneutralized by an intravenous bolus injection of an atrial natriuretic factor antiserum at a dilution of 1 : 2000 (low dose, n = 7) or 1 :400 (high dose, n = 6). In the high-dose group, plasma atrial natriuretic factor concentration was significantly reduced by 65 ± 14 pg/ml from basal levels of 165 ± 12 pg/ml within 10 minutes and remained reduced for the 90-minute period after the injection. Fetal arterial pressure acutely and transiently decreaased, but at 50 minutes arterial pressure increased and was elevated for the remainder of the experiment. Urine flow and urinary excretion rates of sodium, potassium, and chloride were reduced within 10 minutes after the injection. Urine flow rate was suppressed for as long as plasma atrial natriuretic factor concentrations were reduced. Fetuses in the low-dose and control groups showed no significant change in cardiovascular or renal function. In response to atrial natriuretic factor antiserum injection, plasma angiotenSin" concentrations were increased, whereas plasma arginine vasopressin concentrations were unchanged. These results suggest that endogenous atrial natriuretic factor is involved in the maintenance of arterial pressure and urinary excretion in the ovine fetus. (AM J OSSTET GYNECOL 1991 ;165:1558-67.)
Key words: Atrial natriuretic factor, fetus, antiserum, urine flow, vascular pressure, angiotensin II, arginine vasopressin In the mammalian fetus, atrial natriuretic factor (ANF) has been implicated in the regulation of blood pressure and fluid volume. Administration of ANF into the fetal circulation decreases arterial pressure and blood volume, while having little effect on venous pressure or heart rate.' The renal effects of ANF are largely similar to those in the adult, in that it increases fetal urine flow rate, glomerular filtration rate, and enhances urinary electrolyte excretions.2.3 However, these effects occur only at high ANF concentrations and appear to be gestational-age dependent. 4 Contrary to that in the adult, the interaction of ANF with other endocrine systems in the fetus is minimal, except at supraphysiologic concentrations when arginine vasopressin and plasma renin activities are increased in response to ANF infusion. 3 . 4 These observations suggest that ANF may not be playing a significant role in the regulation of vascular pressures or fluid dynamics in the fetus. However, ANF concentrations in the fetal circulation are higher than From the Division ofPerinatal Medicine, Department of Reproductive Medicine, University of California, San Diego. Supported in part by National Institute of Child Health and Human Development grant HD20299. Received for publication November 21 , 1990; revised April 25 , 1991 ; accepted May 28,1991. Reprint requests: Cecilia Y. Cheung, PhD, University of California, San Diego, Department of Reproductive Medicine 0802, School of Medicine, La jolla, CA 92093-0802. 611131255
1558
those in the adult:-" and physiologic manipulations of the fetal cardiovascular system can signifiicantly alter fetal plasma ANF concentrations. For example, expansion of the fetal vascular volume' or fetal hyperosmolality" increases plasma ANF levels, whereas reduction in fetal blood volume decreases circulating ANF concentrations. 9 Furthermore, fetal hypoxia elevates plasma ANF levels.'o Therefore, to document in the fetus the contribution of endogenous ANF to the physiologic maintenance of blood pressures, urine flow, and electrolyte excretion and its interaction with other hormones, this study was undertaken to investigate the cardiovascular, renal, and endocrine effects of blockade of endogenous ANF action. Since a physiologic antagonist to ANF is currently not available, an antiserum was used for these studies, to immunoneutralize endogenous ANF in the fetal circulation. Methods
Eleven pregnant sheep with singleton fetuses were used for these studies. The experimental protocol was approved by the University committee on animal research. We followed the guidelines of the University and the National Institutes of Health for the care and use of these animals. With the animals under gas inhalation anesthesia (1 % to 2% halothane in oxygen), chronic vascular catheters were implanted in the fetal descending aorta and inferior vena cava at the level of
Volume 165 Number 5, Part I
the diaphragm. Additional catheters were placed in the urinary bladder and in the amniotic cavity using procedures previously described. II Prophylactic antibiotics were administered to the ewe and into the amniotic fluid for the first 5 postsurgical days. Experiments were carried out at least 5 days after surgery. The gestational ages of the fetuses at the time of experimentation were 128 to 139 days (term 145 to 150 days). Four fetuses were experimented on once, and seven were experimented on twice, with 48 hours allowed for recovery between experiments. During the experiment, arterial pressure, venous pressure, heart rate, amniotic fluid pressure, and urine flow rate were continuously monitored on a polygraph recorder, and data were stored with an on-line computer. 12 The fetal vascular pressures were continuously corrected with the on-line computer, with amniotic fluid pressure used as the zero pressure reference. For data analysis of vascular pressures, heart rate, and urine flow rate, 2-minute averages of the on-line data were taken and stored in a file for statistical evaluation. After a 30-minute control period, an ANF antiserum was injected intravenously as a bolus into the fetal circulation. This was followed by a 90-minute observation period. The antiserum was generated against the carboxyl terminal of human ANF I . 28 (Peninsula Laboratories, Belmont, Calif.). We have successfully used this antiserum for the measurement of ANF concentrations in the pregnant sheep and their fetuses. s The antiserum recognizes ovine ANF and, when used at a dilution of 1: 25,000 for radioimmunoassay (RIA), it has a sensitivity of 3 pg of human ANF I2R • One group of fetuses (low dose, n = 7) was given the antiserum at a dilution of I : 2000 in 2 ml of I % normal rabbit serum diluted in phosphate-buffered saline solution. A second group of fetuses (high dose, n = 6) was given the antiserum at a dilution of 1: 400 in 2 ml of I % normal rabbit serum. A separate group of fetuses were used as controls (n = 5) and were given 2 ml of 1 % normal rabbit serum. Samples of fetal blood at 3.5 ml each and urine at I ml each were taken at 10 minutes after the start of the control period and at I5-minute intervals thereafter. The fetal blood removed by blood sampling was replaced by an equal volume of heparinized maternal blood. In the blood samples, blood gases (IL 1302 blood gas analyzer, Instrumentation Laboratory), hematocrit (in triplicate), plasma protein concentration (American Optical, TS meter), and plasma ANF, arginine vasopressin, and angiotensin II concentrations were measured. In the urine samples, osmolality (Advanced Diagnostic Osmometer 3D2) and electrolyte concentrations (Nova 5 + 5 electrolyte analyzer) were measured. Plasma ANF concentrations were determined by RIA with procedures previously described 5 with the follow-
ANF antiserum in fetus
1559
ing exception: The ANF concentration in the samples was measured by a direct assay without the extraction step. This was necessary because the ANF antibody, when injected into the fetus, became bound to the hormone, thus reducing the concentration of free ANF in the circulation. Extraction ofthese plasma samples with the Sep-Pak procedure would dissociate the antibody from the ANF, thereby freeing the ANF for subsequent RIA. This would result in measurement of total ANF in the plasma instead of free ANF left unbound by the injected antiserum. By using the plasma directly for RIA without extraction, the concentration of free ANF in the plasma not bound by the injected antiserum could be obtained. In a previous study in the rat a similar direct assay procedure was used to determine the level of free ANF in the circulation after antibody administration." We have validated this method by the addition of the antiserum to a pool of fetal plasma before subjecting to direct RIA and obtained antiserum dilution-dependent reductions of ANF concentration in the plasma. Furthermore, when the fetal plasma pool (without antiserum addition) was measured by the direct method, as well as after extraction, the mean value obtained with the direct method was 9% greater than with the extracted method. There were no statistically significant differences between the values obtained by the two methods. Plasma concentrations of arginine vasopressin and angiotensin II were determined by RIA after extraction, as has been reported previously.l4 The data are presented as the mean ± SE. The change in blood volume at a given time was calculated as a percent of control using the mean hematocrit values.15 To explore changes with time in a given variable, a two-factor analysis of variance was used, with time and animal being the two factors. If a significant F value was obtained, individual group means were compared with Fisher's least-significant-difference test. In some instances, when the intraanimal variability was large, a non parametric analysis of variance was used. (Friedman's nonparametric test, which was the equivalent of a two-factor analysis of variance, with time and animal being the two factors, calculated a value for X2.) Data were considered statistically significant if p < 0.05. Urinary excretion rate of each of the electrolytes-sodium, potassium, and chloride-was calculated as the product of urine flow rate and urinary concentration of the electrolyte. Results
Basal conditions. For the fetuses in the control, lowdose, and high-dose groups, the gestational age at the time of experimentation and the control values during basal conditions are given in Table 1. There was no significant difference in any of these variables among the three groups of animals studied.
1560 Cheung
November 1991 Am J Obstet Gynecol
20
o
-20
-40
-60
-80 -30
o
60
30
90
TIME (min) Fig. 1. Changes in fetal plasma ANF concentration (mean ± SE) after ANF antiserum injection at 30 minutes in low-dose (solid squares), high-dose (solid circles), and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes of experiment. Dotted line, Time of ANF antiserum injection.
Table I. Basal values before antiserum injection in three groups of fetuses Control (n = 5) Gestational age (days) Arterial pressure (mm Hg) Venous pressure (mm Hg) Heart rate (beats/min) pH Pco. (mm Hg) Po. (mm Hg) Hematocrit (%) Protein (gm/dl) Urine flow (mIl min) Urine osmolality (mOsm/kg) Urine sodium (mEq/L) Urine potassium (mEq/L) Urine chloride (mEq/L) Plasma ANF (pg/ml) Plasma arginine vasopressin (pg/ml) Plasma angiotensin II (pg/ml)
134 47.9 4.0 167 7.34 54.5 19.4 33.7 3.7 0.82 138 38.3 9.4 29.1 165 8.12
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 2.2 0.6 6
om
0.4 1.3 1.5 0.1 0.17 II 6.4 3.2 2.4 24 2.12
32.5 ± 6.4
After the 30-minute control period, the ANF antiserum was injected into two groups of fetuses. The lowdose group received the antiserum at 1: 2000 dilution, and the high-dose group received the antiserum at 1 : 400 dilution. The control group was given an injection of the diluent. Low-dose group. In the low-dose group administration of the ANF antiserum lowered fetal plasma ANF concentration from control levels of 115 ± 20 pg/ml by 13 ± 7 pg/ml at 10 minutes (Fig. 1). However this decrease was not statistically significant (F = 1.19,
Low (n = 7) 132 42.6 3.2 165 7.33 53.0 19.4 34.4 3.9 1.06 144 34.0 12.8 38.6 115 4.03
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 1.1 0.4 7 0.01 0.9 1.2 1.0 0.1 0.19 14 6.5 4.0 6.9 20 0.64
25.7 ± 5.4
High (n = 6) 134 46.7 3.3 165 7.33 53.9 21.9 33.3 3.7 0.83 125 34.2 5.8 27.6 165 8.13
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 1.4 0.5 4 0.01 0.6 1.4 2.2 0.1 0.17 7 6.0 2.2 2.8 12 5.20
30.3 ± 5.8
p = 0.33; two-way analysis of variance). Fetal arterial pressure decreased sharply by 1.6 ± 0.4 mm Hg (F = 2.02, P = 0.06), returned to basal levels by 10 minutes after this low-dose antiserum injection, and was unchanged for the remainder of the experiment. Fetal heart rate increased transiently by 10 ± 4 beats/min (F = 1.98, P = 0.07) and returned to basal levels within 10 minutes after the injection. Venous pressure did not change significantly at any time after the antiserum administration (F = 1.00, P = 0.42). There was a small increase in fetal blood volume of
ANF antiserum in fetus
Volume 165 Number 5, Part 1
1561
5
o -5
-10 2
'" >
:z: --.
1
:I:
0
H
r.l (,!)
~u
tTt
~
-1 -2 30
~
:I:
:z: H
s::
20
.~
~
......
III (,!) +J 10
10
(Jl
~u
Q)
.c
0 -10 -30
0
30
60
90
TIME (min) Fig. 2. Changes in fetal arterial pressure (AP), venous pressure (VP), and heart rate (RR) (mean ± SE) after ANF antiserum injection at 30 minutes in high· dose (solid circles) and control (open circles) groups. See text for statistical analysis. Dotted line, Time of ANF antiserum injection.
1.8% ± 0.7% at 10 minutes after the i~ection, but the change was not statistically significant (F = 1.73, P = 0.14). By 40 minutes, blood volume had returned toward baseline. Plasma protein concentrations did not change immediately after the low-dose antiserum injection but was elevated at 40 minutes after the injection (F = 3.81, P < 0.01). Thereafter, plasma protein remained elevated until the end of the experiment. Fetal arterial blood gases and pH did not change significantly throughout the experiment. Similarly, fetal urine flow rate; urine osmolality; and urinary sodium, potassium, and chloride excretion rates did not change in response to the low-dose antiserum treatment. High-dose group. In the high-dose group injection of the ANF antiserum significantly reduced fetal
plasma ANF concentration (F = 6.17, P < 0.001) within 10 minutes by 65 ± 14 pg/mlfromcontrollevels of 165 ± 12 pg/ml (Fig. 1). Plasma ANF levels remained reduced for the duration of the experiment. Fetal arterial pressure showed significant changes after the high-dose antiserum injection (F = 4.22,p < 0.01; Fig. 2). There was an acute and transient fall in arterial pressure of 7.6 ± 2.4 mm Hg from control, followed by a return to basal levels at 10 minutes after the injection. At 40 minutes, arterial pressure began to rise and reached a maximum increase of 3.0 ± 1.1 mm Hg from control at 50 minutes after the injection (p < 0.05). Thereafter, arterial pressure was elevated for the remainder of the experiment. Venous pressure showed no consistent change throughout the experi-
1562 Cheung
November 1991 Am J Obstet Cynecol
104
:>
c:Q
*
102
z
H
~
tH>
100
~
~u
98
.10
,,
,,~, 'f ,
o -.10 -30
o
30
60
90
TIME (min) Fig. 3. Changes in fetal blood volume (BV) and protein (PROT) concentrations (mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes. Dotted line, Time of ANF antiserum injection.
0.2
0
tL.
0
z c:
H
~
~
Z
IC(
~"! ~ ~~!~t d! H~lf i l
'rl
J::
.......
-0.2
~~ ~
f~Ui~~~
,
I,
.-4
J::
:x:
u
-0.4
-0.6
-0.8 -30
o
30
60
90
TIME (min) Fig. 4. Changes in fetal urine flow rate (UF, mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. See text for statistical analysis. Dotted line, Time of ANF antiserum injection.
ANF antiserum in fetus
Volume 165 Number 5, Part I
1563
0.2
0
",.""
..
~
0 H
-0.2
r... ~
Z H p:: ~
Z
H
~
l!l Z
.., ~
a r-4 a
-0.4
........
•
• ",.",
...•.. ...
.'.
.'
••
•
-0.6
.ex:
:t:
u
-0.8
•
-1. 0
-80
o
-40
40
CHANGE IN PLASMA ANF (pg/ml)
Fig. 5. Correlation between change in urine Row rate and change in plasma AN} concentrations in low-dose, high-dose, and control groups of fetuses during 30 minutes after ANF antiserum injection (r = 0.64, P < 0.01, n = IS).
ment (F = 1.05, P = 0.39; Fig. 2). The fetal heart rate changes were reciprocal to those in arterial pressure. Heart rate increased sharply by 17 ± 9 beats/min and gradually returned to basal levels by 20 minutes after the injection (F = 1.22, P = 0.30; Fig. 2). Heart rate continued to fall by a maximum of 7 ± 6 beats/min below basal levels at 50 minutes after the injection and slowly returned to control levels by the end of the experiment. Because of the variability among the animals, these changes in fetal heart rate were not statistically significant. Fetal blood volume increased after the high-dose antiserum injection to a maximum of 2.3% ± 0.7% above control at 25 minutes after the antiserum injection (F = l.50, P = 0.22; Fig. 3) and returned to control levels thereafter. Concurrent with the increase in blood volume, there was a nonsignificant decrease from control in plasma protein concentrations at 25 minutes, which gradually rebounded to levels significantly above baseline at 50 minutes after the injection (F = 3.56, P < 0.01; Fig. 3). There were no consistent changes in fetal arterial blood gases and pH after ANF antiserum injection. In response to the high-dose ANF antiserum, fetal urine flow rate was significantly suppressed within 5 minutes by 0.5 ± 0.1 mUmin from a basal rate of 0.8 ± 0.2 mUmin (F = 2.53, P < 0.05), and remained
low at this level for 30 minutes. Thereafter, although urine flow increased somewhat, it remained below basal levels for the remainder of the experiment (Fig. 4). During the 30 minutes after antiserum administration, the reduction in urine flow was significantly correlated with the decrease in plasma AN:F concentrations (r = 0.64, P < 0.01; Fig. 5). Urine osmolality increased by 33 ± 17 mOsm/kg from control at 25 minutes after the antiserum injection (F = 2.52, P < 0.(5) but returned to baseline at 55 minutes. Urinary excretion rates of sodium, potassium, and chloride decreased significantly at 10 minutes after the injection to 55% ± 12% (Friedman's X' = 15.2, P < 0.01), 52% ± 10% (Friedman's X2 = 17.0, P < 0.01), and 54% ± 12% (Friedman'S X2 = 16.6, P < 0.01) of control values, respectively. However, by 40 minutes after the antiserum iruection, urinary electrolyte excretion rates were not different from basal levels (Fig. 6). In fetuses of the control group, variations in cardiovascular variables during the entire experimental period were within the 95% confidence interval about the mean during the 30-minute control period. There were no significant changes in blood gases, pH, or renal function in these fetuses throughout the experimental period. Hormone responses to ANF antiserum injection. In the low-dose and the high-dose groups, plasma angio-
1564 Cheung
November 1991 Am J Obstet Gyneco1
140 Z 0 H
E-i riI
C)
riI
riI
~
~
100
X E-i ItS
60
Z
20 140 Z 0
H
E-i oV> r.l ~ r.l C) X E-i riI
:.::
~
100
60
*
20 Z
0
H
E-i r.l
C)
r.l
r.l
~
~
100
X E-i
r-i
60
C)
20 -30
o
30
60
90
TIME (min) Fig. 6. Percent changes in sodium (Na), potassium (K), and chloride (el) excretion rates (mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes of experiment. Dotted line, Time of ANF antiserum injection.
tensin II concentrations increased after the ANF antiserum injection (Fig. 7). The angiotensin II responses were similar in the two groups; therefore the data were combined for statistical analysis. At 10 minutes after ANF antiserum injection, angiotensin II concentrations were elevated by 22 ± 9 pg/ml from control levels of 28 ± 4 pg/ml (F = 2.55, P < 0.01). Angiotensin II concentrations returned toward basal levels by the end of the experiment. In both the low-dose and high-dose groups, plasma arginine vasopressin concentrations during the 90~min ute period after ANF a('ltiserum administration were not significantly altered from the levels in the control period (Fig. 7).
Comment
To determine the role of endogenous ANF in the maintenance of fetal cardiovascula,r and renal functions, ovine fetuses were studied after suppression of endogenous ANF. Because of the unavailability of an effective antagonist for ANF, blockade was accomplished by immunoneutralization of endogenous ANF with a specific antiserum. Using similar techniques of immunoneutralization of ANF in adult rats, other investigators have reported that endogenous ANF is consistently involved in the maintenance of urinary output and renal sodium excretion. 13, 16, 17 The effect of ANF on arterial pressure was variable in that ANF antiserum treatment was shown to have either no effece 3 or sig-
ANF antiserum in fetus
Volume 165 Number 5, Part 1
40
H H
~
.-I
H
......
a
20
0'1 r>:I 0-
10
Z
*
30
0
0
Z
0
~ (J
-10 10
1565
"~""""""" ,t""" ----'·t""""""" ----- .''~.' ~"'t·····
,
.... ., .....
---
A.
~ Z
H
.-I
r>:I 'trI 0 0-
~
5
a ......
.'.'
-'-
,,
0
-..
............ -
(J
-5 -30
0
30
60
90
TIME (min) Fig. 7. Changes in plasma angiotensin II (ANG II) and arginine vasopressin (AVP) concentrations
(mean ± SE) after ANF antiserum injection at 30 minutes in control fetuses (open circles, n = 5) and fetuses given both low-dose and high-dose ANF antiserum (solid circles, n = 13). Asterisk, p < 0.05 compared with control values during first 30 minutes. Dotted line. Time of ANF antiserum injection.
nificantly increased arterial pressure}7 In this study, when the antiserum was administered in sufficient quantities as a bolus to the ovine fetus, plasma ANF concentration was significantly and consistently reduced for at least 90 minutes. This reduction in cir, culating ANF levels resulted in significant alterations in fetal arterial pressure, blood volume, and renal excretion of fluid and electrolytes. These findings are consistent with the role of ANF as a physiologic regulator of vascular pressure and fluid balance in the fetus under basal conditions. In the fetus, plasma ANF concentration is normally greater than that in the adult:·6 and the endogenous release rate of ANF is high at 13 ng/min/kg.' The present study demonstrated that a high concentration of ANF antiserum was required to reduce plasma ANF to 58% of basal levels. The use of a lower dose (a fivefold higher dilution) of ANF antiserum reduced plasma concentrations to only 88% of basal levels. In this lowdose group, although ANF levels were below control levels after antiserum injection, at no time was this decrease significant. These observations are consistent with the high production rate of ANF in the fetus. The initial arterial pressure response of the fetus to the high-dose antiserum injection was an abrupt but
transient fall in arterial pressure. This fall in pressure was also seen in the low-dose group. although the decrease was much smaller in magnitude than in the highdose group. This decrease, however. was not detected in the control group of fetuses given I % normal rabbit serum. This acute hypotension was an unexpected finding. It appeared to be related to the presence of the antiserum itself and might not be the result of the blockade of ANF action. Instead. it could be a nonspecific effect of the rabbit antiserum, such as the induction of complement activation by the antibody leading to histamine release. that caused a vasodilation and reduction in arterial pressure. The abrupt fall in arterial pressure would likely affect the fetal cardiovascular system. In studies where fetal hypotension was induced by nitroprusside, arterial pressure fell and remained reduced as long as nitroprusside was infused. 18 In addition, all cardiovascular variables returned to normal within 10 minutes after termination of the nitroprusside infusion. In this study the arterial hypotension was transient; arterial pressure returned to baseline within 10 minutes and then increased to levels above basal conditions. This was contrary to a maintained hypotensive response such as that induced by nitroprusside. The transient nature in the fall in arterial pressure and the quick
1566 Cheung
recovery suggested that it was not caused by the blockade of ANF action, because plasma ANF levels were reduced for a prolonged period. In previous studies of the adult rae s, 16, 17 a transient hypotensive effect after ANF antiserum administration was not reported. However, in those studies arterial pressure was measured at 15- to 3D-minute intervals rather than the continuous recording and display (Fig. 2) used in this study. Thus it is unclear whether the transient fall in arterial pressure was missed by the previous investigators. The reduction of plasma ANF concentrations to levels below normal in the fetus produced a delayed but sustained rise in fetal arterial pressure of 6.2% ± 2.2% that was maintained for as long as plasma ANF concentration was suppressed. These findings are consistent with our previous report that infusion of exogenous ANF into the ovine fetus lowered arterial pressure. 3 An increase in arterial pressure after ANF antiserum treatment has also been reported in the adult rat. 17 The fetal heart rate responses to the ANF antiserum were reciprocal to those of arterial pressure, indicating a baroreceptor mechanism in the maintenance of fetal heart rate. A direct effect of ANF withdrawal on the fetal heart was not apparent. Fetal venous pressure was not consistently affected by the presence of the ANF antiserum. These results support the concept that, in the fetus, ANF normally participates in the maintenance of arterial pressure but not venous pressure or heart rate. It has been demonstrated in the adule 9 and the fetus 3 that administration of ANF increases hematocrit, suggesting a reduction in blood volume. In the ovine fetus this decrease in blood volume is associated with a transient increase in plasma protein concentration. In this study blockade of endogenous ANF in the fetus resulted in a rise in blood volume (although not significant), accompanied by a fall in plasma protein concentration. These observations are consistent with a modulatory role of ANF in the transfer of proteins and fluids across the capillary in the fetus. However, urine flow rate was dramatically reduced at this time; therefore the retention of fluids by the fetal kidney might have contributed to the increase in blood volume. Although the acute, transient fall in arterial pressure might have allowed movement of fluid from the interstitium into the vascular compartment, causing an increase in blood volume, this appeared to be unlikely. This was because blood volume was not elevated until arterial pressure had returned to basal conditions. The elevation in blood volume was not maintained, even though plasma ANF concentrations remained suppressed. This might be partly due to the subsequent increase in arterial pressure in these fetuses. The fetal renal response to ANF antiserum was much more pronounced than the cardiovascular effects. This
November 1991 Am J Obstet Gynecol
was because of the observation that, after blockade of endogenous ANF, fetal urine flow rate was abruptly reduced to 41.4% ± 9.6% of control values. The sharp, transient decrease in arterial pressure in response to ANF antiserum injection could have augmented the fall in urine flow rate, because it has been reported that in the ovine fetus hypotension induced by nitroprusside was accompanied by a fall in urine flow rate and urinary sodium excretion. 20 However, in those studies the fall in urine flow was terminated as the hypotension was reversed. In the present study, the suppression of urine flow rate was prolonged even though the hypotension subsided. Thus it was unlikely that the long-term reduction in urine flow was the result of the short-term decrease in arterial pressure. Furthermore, plasma arginine vasopressin levels did not change during the entire experiment, thus eliminating its possible involvement in reducing urine flow. The fall in urine flow was most likely the result of the withdrawal of ANF, because a significant relationship existed between the fall in plasma ANF concentrations and the reduction in urine flow. This implies that by blocking the renal effects of ANF, urine flow was no longer maintained. Thus, under physiologic conditions, normal fetal urine flow appears to be maintained by circulating ANF levels. In support of this concept is the fact that plasma ANF concentration in the fetus is greater than in the adult, and urine flow rate is higher in the fetus than in the adult relative to body weight. In conjunction with the suppression of urine flow, urinary excretion rates of sodium, potassium, and chloride were reduced. This reduction in electrolyte excretion might be the consequence of the blockade of the natriuretic action of ANF. However, the large concomitant reduction in fetal urine flow rate would have contributed to the decrease in urinary electrolyte excretion rates. The diuretic action of ANF is mediated through an increase in glomerular filtration rate at the fetal kidney.2 When ANF levels were reduced, a decrease in glomerular filtration rate would result. In this study, this was manifested as a fall in urine flow rate. With the reduction in glomerular filtration rate, the filtered load through the kidney was reduced, and thus the amount of electrolytes filtered decreased. The result was a fall in urinary excretion of the electrolytes. This fall in electrolyte excretion rate might be augmented by the action of aldosterone, because plasma aldosterone concentration was found to increase in response to ANF antibody injection in the rat. 13 However, this possibility is unlikely because of the observation in the ovine fetus that aldosterone stimulated the reabsorption of sodium but increased the excretion of potassium, leading to a decrease in the sodium-potassium ratio in the urine. 21 This was inconsistent with our finding, where both sodium and potassium excretion rates
Volume 165 Number 5, Part 1
were reduced, Similarly in the studies by Sasaki et al. 17 and Rudd et aI., I3 the urinary excretion rates of both sodium and potassium were reduced after ANF antiserum treatment. Unlike urine flow that was continuously suppressed in parallel with the reduction in plasma ANF concentration, the excretion of electrolytes returned to basal levels, This increase in electrolyte excretion could not be due to a fall in plasma arginine vasopressin concentration, because arginine vasopressin levels did not change at any time after ANF antiserum treatment. This suggests that, although ANF may participate in the regulation of electrolyte excretion, factors other than ANF may contribute to the maintenance of urinary electrolyte excretion rate during normal conditions, Reduction of endogenous ANF concentration in the fetus did not appear to affect plasma arginine vasopressin concentration. The lack of change in face of urinary changes further supports the role of ANF in the regulation of renal function in the ovine fetus. The reduction in circulating ANF levels increased plasma angiotensin II concentration. A similar increase in plasma renin activity in adult rats after ANF antiserum treatment has been reported by Naruse et al. 16 The increase in angiotensin II concentration in the present study could have been due to the short-term fall in arterial pressure after ANF antiserum injection. Hypotension in the mature ovine fetus is a potent stimulus for increasing plasma renin activity. IS Although it is tempting to speculate that ANF is inhibitory to the renin-angiotensin system as seen in the adult,22 it is difficult to separate the effect of hypotension from that of ANF blockade in the stimulation of angiotensin II release. Besides, infusion of ANF into the ovine fetus was not associated with a fall in plasma renin activity.3 In summary, this study demonstrates that in the ovine fetus plasma ANF plays a role in the regulation of normal arterial pressure and blood volume and in the maintenance of basal urinary output. Furthermore, ANF appears to have little effect in the maintenance of fetal heart rate. Finally, an interaction of ANF with the other vasoactive hormones, angiotensin II and arginine vasopressin, appears to be minimal. REFERENCES 1. Brace RA, Cheung CY. Cardiovascular and fluid responses to atrial natriuretic factor in sheep fetus. Am ] PhysioI1987;253:R561-7. 2. Shine P, McDougall ]G, Towstoless MK, Wintour EM. Action of atrial natriuretic peptide in the immature ovine kidney. Pediatr Res 1987;22:11-5. 3. Brace RA, Bayer LA, Cheung CY. Fetal cardiovascular,
ANF antiserum in fetus
4.
5. 6. 7.
8. 9. 10. II. 12. 13.
14. 15.
16.
17.
18.
19. 20. 21. 22.
1567
endocrine, and fluid responses to atrial natriuretic factor infusion. Am] Physiol 1989;257:R580-7. Robillard ]E, Nakamura KT, Varille VA, Andresen AA, Natherne GP, Van Orden DE. Ontogeny of the renal response to natriuretic peptide in sheep. Am ] Physiol 1988; 254: F634-41. Cheung CY, Gibbs DM, Brace RA. Atrial natriuretic factor in maternal and fetal sheep. Am] PhysioI1987;252:E27982. Ervin MG, Ross MG, Castro R, et al. Ovine fetal and adult atrial natriuretic factor metabolism. Am ] Physiol 1988;254:R40-6. Ross MG, Ervin MG, Lam RW, Castro L, Leake RD, Fisher DA. Plasma atrial natriuretic peptide response to volume expansion in the ovine fetus. AM ] OBSTET GVNECOL 1987; 157: 1292-7. Cheung CY, Miner LK, Brace RA. Hyperosmolality elevates plasma atrial natriuretic factor in the ovine fetus. Am] PhysioI1989;257:E466-72. Cheung CY, Brace RA. Hemorrhage-induced reductions in plasma atrial natriuretic factor in the ovine fetus. AM ] OBSTET GVNECOL 1991;165:474-81. Cheung CY, Brace RA. Fetal hypoxia elevates plasma atrial natriuretic factor concentration. AM] OBSTET GvNECOL 1988; 159: 1263-8. Brace RA. Thoracic duct lymph flow and its measurement in chronically catheterized sheep fetus. Am ] Physiol 1989;256:HI6-20. Brace RA, Brittingham DS. Fetal vascular pressure and heart rate responses to non labor uterine contractions. Am ] PhysioI1986;251:R409-16. Rudd A, Plavin S, Hirsch AT, Ingelfinger ]R, Dzau VJ. Atrial natriuretic factor-specific antibody as a tool for physiological studies. Evidence for role of atrial natriuretic factor in aldosterone and renal electrolyte regulation. Circ Res 1989;65:1324-9. Miner LK, Brace RA, Cheung CY. Temporal relationships among fetal urine flow, ANF, and AVP responses to hypertonic infusion. Am] PhysioI1990;258:R469-75. Brace RA. Blood volume in the fetus and methods for its measurement. In: Nathanielsz PW, ed. Animal models in fetal medicine. IV. Monographs in fetal physiology. Ithaca, New York: Perinatology Press, 1984;5:19-36. Naruse M, Obana K, Naruse K, et al. Antisera to atrial natriuretic factor reduces urinary sodium excretion and increases plasma renin activity in rats. Biochem Biophys Res Commun 1985; 132:954-60. Sasaki AA, Kida 0, Kita T, et al. Effects of antiserum against Ol-rat atrial natriuretic polypeptide in spontaneously hypertensive rats. Am ] Physiol 1989;257: H1104-9. Rawashdeh NM, Ray ND, Sundberg DK, Rose ]C. Comparison of hormonal responses to hypotension in mature and immature lambs. Am] PhysioI1988;255:R6772. Laragh]H. The endocrine control of blood volume, blood pressure and sodium balance: atrial hormone and renin system interactions.] Hypertens 1986;4:S143-56. Lumbers ER, Stevens AD. The effects of fusemide, saralasin and hypotension on fetal plasma renin activity and on fetal renal function.] Physiol 1987;393:479-90. Lingwood B, Hardy K], Coghlan]P, Wintour EM. Effect of aldosterone on urine composition in the chronically cannulated ovine fetus.] Endocrinol 1978;76:553-4. BurnettJC Jr, Granger JP, Opgenorth TJ. Effects of synthetic atrial natriuretic factor on renal function and renin release. Am] Physiol 1984;247 :F863-6.
Key words: Atrial natriuretic factor, fetus, antiserum, urine flow, vascular pressure, angiotensin II, arginine vasopressin In the mammalian fetus, atrial natriuretic factor (ANF) has been implicated in the regulation of blood pressure and fluid volume. Administration of ANF into the fetal circulation decreases arterial pressure and blood volume, while having little effect on venous pressure or heart rate.' The renal effects of ANF are largely similar to those in the adult, in that it increases fetal urine flow rate, glomerular filtration rate, and enhances urinary electrolyte excretions.2.3 However, these effects occur only at high ANF concentrations and appear to be gestational-age dependent. 4 Contrary to that in the adult, the interaction of ANF with other endocrine systems in the fetus is minimal, except at supraphysiologic concentrations when arginine vasopressin and plasma renin activities are increased in response to ANF infusion. 3 . 4 These observations suggest that ANF may not be playing a significant role in the regulation of vascular pressures or fluid dynamics in the fetus. However, ANF concentrations in the fetal circulation are higher than From the Division ofPerinatal Medicine, Department of Reproductive Medicine, University of California, San Diego. Supported in part by National Institute of Child Health and Human Development grant HD20299. Received for publication November 21 , 1990; revised April 25 , 1991 ; accepted May 28,1991. Reprint requests: Cecilia Y. Cheung, PhD, University of California, San Diego, Department of Reproductive Medicine 0802, School of Medicine, La jolla, CA 92093-0802. 611131255
1558
those in the adult:-" and physiologic manipulations of the fetal cardiovascular system can signifiicantly alter fetal plasma ANF concentrations. For example, expansion of the fetal vascular volume' or fetal hyperosmolality" increases plasma ANF levels, whereas reduction in fetal blood volume decreases circulating ANF concentrations. 9 Furthermore, fetal hypoxia elevates plasma ANF levels.'o Therefore, to document in the fetus the contribution of endogenous ANF to the physiologic maintenance of blood pressures, urine flow, and electrolyte excretion and its interaction with other hormones, this study was undertaken to investigate the cardiovascular, renal, and endocrine effects of blockade of endogenous ANF action. Since a physiologic antagonist to ANF is currently not available, an antiserum was used for these studies, to immunoneutralize endogenous ANF in the fetal circulation. Methods
Eleven pregnant sheep with singleton fetuses were used for these studies. The experimental protocol was approved by the University committee on animal research. We followed the guidelines of the University and the National Institutes of Health for the care and use of these animals. With the animals under gas inhalation anesthesia (1 % to 2% halothane in oxygen), chronic vascular catheters were implanted in the fetal descending aorta and inferior vena cava at the level of
Volume 165 Number 5, Part I
the diaphragm. Additional catheters were placed in the urinary bladder and in the amniotic cavity using procedures previously described. II Prophylactic antibiotics were administered to the ewe and into the amniotic fluid for the first 5 postsurgical days. Experiments were carried out at least 5 days after surgery. The gestational ages of the fetuses at the time of experimentation were 128 to 139 days (term 145 to 150 days). Four fetuses were experimented on once, and seven were experimented on twice, with 48 hours allowed for recovery between experiments. During the experiment, arterial pressure, venous pressure, heart rate, amniotic fluid pressure, and urine flow rate were continuously monitored on a polygraph recorder, and data were stored with an on-line computer. 12 The fetal vascular pressures were continuously corrected with the on-line computer, with amniotic fluid pressure used as the zero pressure reference. For data analysis of vascular pressures, heart rate, and urine flow rate, 2-minute averages of the on-line data were taken and stored in a file for statistical evaluation. After a 30-minute control period, an ANF antiserum was injected intravenously as a bolus into the fetal circulation. This was followed by a 90-minute observation period. The antiserum was generated against the carboxyl terminal of human ANF I . 28 (Peninsula Laboratories, Belmont, Calif.). We have successfully used this antiserum for the measurement of ANF concentrations in the pregnant sheep and their fetuses. s The antiserum recognizes ovine ANF and, when used at a dilution of 1: 25,000 for radioimmunoassay (RIA), it has a sensitivity of 3 pg of human ANF I2R • One group of fetuses (low dose, n = 7) was given the antiserum at a dilution of I : 2000 in 2 ml of I % normal rabbit serum diluted in phosphate-buffered saline solution. A second group of fetuses (high dose, n = 6) was given the antiserum at a dilution of 1: 400 in 2 ml of I % normal rabbit serum. A separate group of fetuses were used as controls (n = 5) and were given 2 ml of 1 % normal rabbit serum. Samples of fetal blood at 3.5 ml each and urine at I ml each were taken at 10 minutes after the start of the control period and at I5-minute intervals thereafter. The fetal blood removed by blood sampling was replaced by an equal volume of heparinized maternal blood. In the blood samples, blood gases (IL 1302 blood gas analyzer, Instrumentation Laboratory), hematocrit (in triplicate), plasma protein concentration (American Optical, TS meter), and plasma ANF, arginine vasopressin, and angiotensin II concentrations were measured. In the urine samples, osmolality (Advanced Diagnostic Osmometer 3D2) and electrolyte concentrations (Nova 5 + 5 electrolyte analyzer) were measured. Plasma ANF concentrations were determined by RIA with procedures previously described 5 with the follow-
ANF antiserum in fetus
1559
ing exception: The ANF concentration in the samples was measured by a direct assay without the extraction step. This was necessary because the ANF antibody, when injected into the fetus, became bound to the hormone, thus reducing the concentration of free ANF in the circulation. Extraction ofthese plasma samples with the Sep-Pak procedure would dissociate the antibody from the ANF, thereby freeing the ANF for subsequent RIA. This would result in measurement of total ANF in the plasma instead of free ANF left unbound by the injected antiserum. By using the plasma directly for RIA without extraction, the concentration of free ANF in the plasma not bound by the injected antiserum could be obtained. In a previous study in the rat a similar direct assay procedure was used to determine the level of free ANF in the circulation after antibody administration." We have validated this method by the addition of the antiserum to a pool of fetal plasma before subjecting to direct RIA and obtained antiserum dilution-dependent reductions of ANF concentration in the plasma. Furthermore, when the fetal plasma pool (without antiserum addition) was measured by the direct method, as well as after extraction, the mean value obtained with the direct method was 9% greater than with the extracted method. There were no statistically significant differences between the values obtained by the two methods. Plasma concentrations of arginine vasopressin and angiotensin II were determined by RIA after extraction, as has been reported previously.l4 The data are presented as the mean ± SE. The change in blood volume at a given time was calculated as a percent of control using the mean hematocrit values.15 To explore changes with time in a given variable, a two-factor analysis of variance was used, with time and animal being the two factors. If a significant F value was obtained, individual group means were compared with Fisher's least-significant-difference test. In some instances, when the intraanimal variability was large, a non parametric analysis of variance was used. (Friedman's nonparametric test, which was the equivalent of a two-factor analysis of variance, with time and animal being the two factors, calculated a value for X2.) Data were considered statistically significant if p < 0.05. Urinary excretion rate of each of the electrolytes-sodium, potassium, and chloride-was calculated as the product of urine flow rate and urinary concentration of the electrolyte. Results
Basal conditions. For the fetuses in the control, lowdose, and high-dose groups, the gestational age at the time of experimentation and the control values during basal conditions are given in Table 1. There was no significant difference in any of these variables among the three groups of animals studied.
1560 Cheung
November 1991 Am J Obstet Gynecol
20
o
-20
-40
-60
-80 -30
o
60
30
90
TIME (min) Fig. 1. Changes in fetal plasma ANF concentration (mean ± SE) after ANF antiserum injection at 30 minutes in low-dose (solid squares), high-dose (solid circles), and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes of experiment. Dotted line, Time of ANF antiserum injection.
Table I. Basal values before antiserum injection in three groups of fetuses Control (n = 5) Gestational age (days) Arterial pressure (mm Hg) Venous pressure (mm Hg) Heart rate (beats/min) pH Pco. (mm Hg) Po. (mm Hg) Hematocrit (%) Protein (gm/dl) Urine flow (mIl min) Urine osmolality (mOsm/kg) Urine sodium (mEq/L) Urine potassium (mEq/L) Urine chloride (mEq/L) Plasma ANF (pg/ml) Plasma arginine vasopressin (pg/ml) Plasma angiotensin II (pg/ml)
134 47.9 4.0 167 7.34 54.5 19.4 33.7 3.7 0.82 138 38.3 9.4 29.1 165 8.12
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 2.2 0.6 6
om
0.4 1.3 1.5 0.1 0.17 II 6.4 3.2 2.4 24 2.12
32.5 ± 6.4
After the 30-minute control period, the ANF antiserum was injected into two groups of fetuses. The lowdose group received the antiserum at 1: 2000 dilution, and the high-dose group received the antiserum at 1 : 400 dilution. The control group was given an injection of the diluent. Low-dose group. In the low-dose group administration of the ANF antiserum lowered fetal plasma ANF concentration from control levels of 115 ± 20 pg/ml by 13 ± 7 pg/ml at 10 minutes (Fig. 1). However this decrease was not statistically significant (F = 1.19,
Low (n = 7) 132 42.6 3.2 165 7.33 53.0 19.4 34.4 3.9 1.06 144 34.0 12.8 38.6 115 4.03
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 1.1 0.4 7 0.01 0.9 1.2 1.0 0.1 0.19 14 6.5 4.0 6.9 20 0.64
25.7 ± 5.4
High (n = 6) 134 46.7 3.3 165 7.33 53.9 21.9 33.3 3.7 0.83 125 34.2 5.8 27.6 165 8.13
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1 1.4 0.5 4 0.01 0.6 1.4 2.2 0.1 0.17 7 6.0 2.2 2.8 12 5.20
30.3 ± 5.8
p = 0.33; two-way analysis of variance). Fetal arterial pressure decreased sharply by 1.6 ± 0.4 mm Hg (F = 2.02, P = 0.06), returned to basal levels by 10 minutes after this low-dose antiserum injection, and was unchanged for the remainder of the experiment. Fetal heart rate increased transiently by 10 ± 4 beats/min (F = 1.98, P = 0.07) and returned to basal levels within 10 minutes after the injection. Venous pressure did not change significantly at any time after the antiserum administration (F = 1.00, P = 0.42). There was a small increase in fetal blood volume of
ANF antiserum in fetus
Volume 165 Number 5, Part 1
1561
5
o -5
-10 2
'" >
:z: --.
1
:I:
0
H
r.l (,!)
~u
tTt
~
-1 -2 30
~
:I:
:z: H
s::
20
.~
~
......
III (,!) +J 10
10
(Jl
~u
Q)
.c
0 -10 -30
0
30
60
90
TIME (min) Fig. 2. Changes in fetal arterial pressure (AP), venous pressure (VP), and heart rate (RR) (mean ± SE) after ANF antiserum injection at 30 minutes in high· dose (solid circles) and control (open circles) groups. See text for statistical analysis. Dotted line, Time of ANF antiserum injection.
1.8% ± 0.7% at 10 minutes after the i~ection, but the change was not statistically significant (F = 1.73, P = 0.14). By 40 minutes, blood volume had returned toward baseline. Plasma protein concentrations did not change immediately after the low-dose antiserum injection but was elevated at 40 minutes after the injection (F = 3.81, P < 0.01). Thereafter, plasma protein remained elevated until the end of the experiment. Fetal arterial blood gases and pH did not change significantly throughout the experiment. Similarly, fetal urine flow rate; urine osmolality; and urinary sodium, potassium, and chloride excretion rates did not change in response to the low-dose antiserum treatment. High-dose group. In the high-dose group injection of the ANF antiserum significantly reduced fetal
plasma ANF concentration (F = 6.17, P < 0.001) within 10 minutes by 65 ± 14 pg/mlfromcontrollevels of 165 ± 12 pg/ml (Fig. 1). Plasma ANF levels remained reduced for the duration of the experiment. Fetal arterial pressure showed significant changes after the high-dose antiserum injection (F = 4.22,p < 0.01; Fig. 2). There was an acute and transient fall in arterial pressure of 7.6 ± 2.4 mm Hg from control, followed by a return to basal levels at 10 minutes after the injection. At 40 minutes, arterial pressure began to rise and reached a maximum increase of 3.0 ± 1.1 mm Hg from control at 50 minutes after the injection (p < 0.05). Thereafter, arterial pressure was elevated for the remainder of the experiment. Venous pressure showed no consistent change throughout the experi-
1562 Cheung
November 1991 Am J Obstet Cynecol
104
:>
c:Q
*
102
z
H
~
tH>
100
~
~u
98
.10
,,
,,~, 'f ,
o -.10 -30
o
30
60
90
TIME (min) Fig. 3. Changes in fetal blood volume (BV) and protein (PROT) concentrations (mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes. Dotted line, Time of ANF antiserum injection.
0.2
0
tL.
0
z c:
H
~
~
Z
IC(
~"! ~ ~~!~t d! H~lf i l
'rl
J::
.......
-0.2
~~ ~
f~Ui~~~
,
I,
.-4
J::
:x:
u
-0.4
-0.6
-0.8 -30
o
30
60
90
TIME (min) Fig. 4. Changes in fetal urine flow rate (UF, mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. See text for statistical analysis. Dotted line, Time of ANF antiserum injection.
ANF antiserum in fetus
Volume 165 Number 5, Part I
1563
0.2
0
",.""
..
~
0 H
-0.2
r... ~
Z H p:: ~
Z
H
~
l!l Z
.., ~
a r-4 a
-0.4
........
•
• ",.",
...•.. ...
.'.
.'
••
•
-0.6
.ex:
:t:
u
-0.8
•
-1. 0
-80
o
-40
40
CHANGE IN PLASMA ANF (pg/ml)
Fig. 5. Correlation between change in urine Row rate and change in plasma AN} concentrations in low-dose, high-dose, and control groups of fetuses during 30 minutes after ANF antiserum injection (r = 0.64, P < 0.01, n = IS).
ment (F = 1.05, P = 0.39; Fig. 2). The fetal heart rate changes were reciprocal to those in arterial pressure. Heart rate increased sharply by 17 ± 9 beats/min and gradually returned to basal levels by 20 minutes after the injection (F = 1.22, P = 0.30; Fig. 2). Heart rate continued to fall by a maximum of 7 ± 6 beats/min below basal levels at 50 minutes after the injection and slowly returned to control levels by the end of the experiment. Because of the variability among the animals, these changes in fetal heart rate were not statistically significant. Fetal blood volume increased after the high-dose antiserum injection to a maximum of 2.3% ± 0.7% above control at 25 minutes after the antiserum injection (F = l.50, P = 0.22; Fig. 3) and returned to control levels thereafter. Concurrent with the increase in blood volume, there was a nonsignificant decrease from control in plasma protein concentrations at 25 minutes, which gradually rebounded to levels significantly above baseline at 50 minutes after the injection (F = 3.56, P < 0.01; Fig. 3). There were no consistent changes in fetal arterial blood gases and pH after ANF antiserum injection. In response to the high-dose ANF antiserum, fetal urine flow rate was significantly suppressed within 5 minutes by 0.5 ± 0.1 mUmin from a basal rate of 0.8 ± 0.2 mUmin (F = 2.53, P < 0.05), and remained
low at this level for 30 minutes. Thereafter, although urine flow increased somewhat, it remained below basal levels for the remainder of the experiment (Fig. 4). During the 30 minutes after antiserum administration, the reduction in urine flow was significantly correlated with the decrease in plasma AN:F concentrations (r = 0.64, P < 0.01; Fig. 5). Urine osmolality increased by 33 ± 17 mOsm/kg from control at 25 minutes after the antiserum injection (F = 2.52, P < 0.(5) but returned to baseline at 55 minutes. Urinary excretion rates of sodium, potassium, and chloride decreased significantly at 10 minutes after the injection to 55% ± 12% (Friedman's X' = 15.2, P < 0.01), 52% ± 10% (Friedman's X2 = 17.0, P < 0.01), and 54% ± 12% (Friedman'S X2 = 16.6, P < 0.01) of control values, respectively. However, by 40 minutes after the antiserum iruection, urinary electrolyte excretion rates were not different from basal levels (Fig. 6). In fetuses of the control group, variations in cardiovascular variables during the entire experimental period were within the 95% confidence interval about the mean during the 30-minute control period. There were no significant changes in blood gases, pH, or renal function in these fetuses throughout the experimental period. Hormone responses to ANF antiserum injection. In the low-dose and the high-dose groups, plasma angio-
1564 Cheung
November 1991 Am J Obstet Gyneco1
140 Z 0 H
E-i riI
C)
riI
riI
~
~
100
X E-i ItS
60
Z
20 140 Z 0
H
E-i oV> r.l ~ r.l C) X E-i riI
:.::
~
100
60
*
20 Z
0
H
E-i r.l
C)
r.l
r.l
~
~
100
X E-i
r-i
60
C)
20 -30
o
30
60
90
TIME (min) Fig. 6. Percent changes in sodium (Na), potassium (K), and chloride (el) excretion rates (mean ± SE) after ANF antiserum injection at 30 minutes in high-dose (solid circles) and control (open circles) groups. Asterisk, p < 0.05 compared with control values during first 30 minutes of experiment. Dotted line, Time of ANF antiserum injection.
tensin II concentrations increased after the ANF antiserum injection (Fig. 7). The angiotensin II responses were similar in the two groups; therefore the data were combined for statistical analysis. At 10 minutes after ANF antiserum injection, angiotensin II concentrations were elevated by 22 ± 9 pg/ml from control levels of 28 ± 4 pg/ml (F = 2.55, P < 0.01). Angiotensin II concentrations returned toward basal levels by the end of the experiment. In both the low-dose and high-dose groups, plasma arginine vasopressin concentrations during the 90~min ute period after ANF a('ltiserum administration were not significantly altered from the levels in the control period (Fig. 7).
Comment
To determine the role of endogenous ANF in the maintenance of fetal cardiovascula,r and renal functions, ovine fetuses were studied after suppression of endogenous ANF. Because of the unavailability of an effective antagonist for ANF, blockade was accomplished by immunoneutralization of endogenous ANF with a specific antiserum. Using similar techniques of immunoneutralization of ANF in adult rats, other investigators have reported that endogenous ANF is consistently involved in the maintenance of urinary output and renal sodium excretion. 13, 16, 17 The effect of ANF on arterial pressure was variable in that ANF antiserum treatment was shown to have either no effece 3 or sig-
ANF antiserum in fetus
Volume 165 Number 5, Part 1
40
H H
~
.-I
H
......
a
20
0'1 r>:I 0-
10
Z
*
30
0
0
Z
0
~ (J
-10 10
1565
"~""""""" ,t""" ----'·t""""""" ----- .''~.' ~"'t·····
,
.... ., .....
---
A.
~ Z
H
.-I
r>:I 'trI 0 0-
~
5
a ......
.'.'
-'-
,,
0
-..
............ -
(J
-5 -30
0
30
60
90
TIME (min) Fig. 7. Changes in plasma angiotensin II (ANG II) and arginine vasopressin (AVP) concentrations
(mean ± SE) after ANF antiserum injection at 30 minutes in control fetuses (open circles, n = 5) and fetuses given both low-dose and high-dose ANF antiserum (solid circles, n = 13). Asterisk, p < 0.05 compared with control values during first 30 minutes. Dotted line. Time of ANF antiserum injection.
nificantly increased arterial pressure}7 In this study, when the antiserum was administered in sufficient quantities as a bolus to the ovine fetus, plasma ANF concentration was significantly and consistently reduced for at least 90 minutes. This reduction in cir, culating ANF levels resulted in significant alterations in fetal arterial pressure, blood volume, and renal excretion of fluid and electrolytes. These findings are consistent with the role of ANF as a physiologic regulator of vascular pressure and fluid balance in the fetus under basal conditions. In the fetus, plasma ANF concentration is normally greater than that in the adult:·6 and the endogenous release rate of ANF is high at 13 ng/min/kg.' The present study demonstrated that a high concentration of ANF antiserum was required to reduce plasma ANF to 58% of basal levels. The use of a lower dose (a fivefold higher dilution) of ANF antiserum reduced plasma concentrations to only 88% of basal levels. In this lowdose group, although ANF levels were below control levels after antiserum injection, at no time was this decrease significant. These observations are consistent with the high production rate of ANF in the fetus. The initial arterial pressure response of the fetus to the high-dose antiserum injection was an abrupt but
transient fall in arterial pressure. This fall in pressure was also seen in the low-dose group. although the decrease was much smaller in magnitude than in the highdose group. This decrease, however. was not detected in the control group of fetuses given I % normal rabbit serum. This acute hypotension was an unexpected finding. It appeared to be related to the presence of the antiserum itself and might not be the result of the blockade of ANF action. Instead. it could be a nonspecific effect of the rabbit antiserum, such as the induction of complement activation by the antibody leading to histamine release. that caused a vasodilation and reduction in arterial pressure. The abrupt fall in arterial pressure would likely affect the fetal cardiovascular system. In studies where fetal hypotension was induced by nitroprusside, arterial pressure fell and remained reduced as long as nitroprusside was infused. 18 In addition, all cardiovascular variables returned to normal within 10 minutes after termination of the nitroprusside infusion. In this study the arterial hypotension was transient; arterial pressure returned to baseline within 10 minutes and then increased to levels above basal conditions. This was contrary to a maintained hypotensive response such as that induced by nitroprusside. The transient nature in the fall in arterial pressure and the quick
1566 Cheung
recovery suggested that it was not caused by the blockade of ANF action, because plasma ANF levels were reduced for a prolonged period. In previous studies of the adult rae s, 16, 17 a transient hypotensive effect after ANF antiserum administration was not reported. However, in those studies arterial pressure was measured at 15- to 3D-minute intervals rather than the continuous recording and display (Fig. 2) used in this study. Thus it is unclear whether the transient fall in arterial pressure was missed by the previous investigators. The reduction of plasma ANF concentrations to levels below normal in the fetus produced a delayed but sustained rise in fetal arterial pressure of 6.2% ± 2.2% that was maintained for as long as plasma ANF concentration was suppressed. These findings are consistent with our previous report that infusion of exogenous ANF into the ovine fetus lowered arterial pressure. 3 An increase in arterial pressure after ANF antiserum treatment has also been reported in the adult rat. 17 The fetal heart rate responses to the ANF antiserum were reciprocal to those of arterial pressure, indicating a baroreceptor mechanism in the maintenance of fetal heart rate. A direct effect of ANF withdrawal on the fetal heart was not apparent. Fetal venous pressure was not consistently affected by the presence of the ANF antiserum. These results support the concept that, in the fetus, ANF normally participates in the maintenance of arterial pressure but not venous pressure or heart rate. It has been demonstrated in the adule 9 and the fetus 3 that administration of ANF increases hematocrit, suggesting a reduction in blood volume. In the ovine fetus this decrease in blood volume is associated with a transient increase in plasma protein concentration. In this study blockade of endogenous ANF in the fetus resulted in a rise in blood volume (although not significant), accompanied by a fall in plasma protein concentration. These observations are consistent with a modulatory role of ANF in the transfer of proteins and fluids across the capillary in the fetus. However, urine flow rate was dramatically reduced at this time; therefore the retention of fluids by the fetal kidney might have contributed to the increase in blood volume. Although the acute, transient fall in arterial pressure might have allowed movement of fluid from the interstitium into the vascular compartment, causing an increase in blood volume, this appeared to be unlikely. This was because blood volume was not elevated until arterial pressure had returned to basal conditions. The elevation in blood volume was not maintained, even though plasma ANF concentrations remained suppressed. This might be partly due to the subsequent increase in arterial pressure in these fetuses. The fetal renal response to ANF antiserum was much more pronounced than the cardiovascular effects. This
November 1991 Am J Obstet Gynecol
was because of the observation that, after blockade of endogenous ANF, fetal urine flow rate was abruptly reduced to 41.4% ± 9.6% of control values. The sharp, transient decrease in arterial pressure in response to ANF antiserum injection could have augmented the fall in urine flow rate, because it has been reported that in the ovine fetus hypotension induced by nitroprusside was accompanied by a fall in urine flow rate and urinary sodium excretion. 20 However, in those studies the fall in urine flow was terminated as the hypotension was reversed. In the present study, the suppression of urine flow rate was prolonged even though the hypotension subsided. Thus it was unlikely that the long-term reduction in urine flow was the result of the short-term decrease in arterial pressure. Furthermore, plasma arginine vasopressin levels did not change during the entire experiment, thus eliminating its possible involvement in reducing urine flow. The fall in urine flow was most likely the result of the withdrawal of ANF, because a significant relationship existed between the fall in plasma ANF concentrations and the reduction in urine flow. This implies that by blocking the renal effects of ANF, urine flow was no longer maintained. Thus, under physiologic conditions, normal fetal urine flow appears to be maintained by circulating ANF levels. In support of this concept is the fact that plasma ANF concentration in the fetus is greater than in the adult, and urine flow rate is higher in the fetus than in the adult relative to body weight. In conjunction with the suppression of urine flow, urinary excretion rates of sodium, potassium, and chloride were reduced. This reduction in electrolyte excretion might be the consequence of the blockade of the natriuretic action of ANF. However, the large concomitant reduction in fetal urine flow rate would have contributed to the decrease in urinary electrolyte excretion rates. The diuretic action of ANF is mediated through an increase in glomerular filtration rate at the fetal kidney.2 When ANF levels were reduced, a decrease in glomerular filtration rate would result. In this study, this was manifested as a fall in urine flow rate. With the reduction in glomerular filtration rate, the filtered load through the kidney was reduced, and thus the amount of electrolytes filtered decreased. The result was a fall in urinary excretion of the electrolytes. This fall in electrolyte excretion rate might be augmented by the action of aldosterone, because plasma aldosterone concentration was found to increase in response to ANF antibody injection in the rat. 13 However, this possibility is unlikely because of the observation in the ovine fetus that aldosterone stimulated the reabsorption of sodium but increased the excretion of potassium, leading to a decrease in the sodium-potassium ratio in the urine. 21 This was inconsistent with our finding, where both sodium and potassium excretion rates
Volume 165 Number 5, Part 1
were reduced, Similarly in the studies by Sasaki et al. 17 and Rudd et aI., I3 the urinary excretion rates of both sodium and potassium were reduced after ANF antiserum treatment. Unlike urine flow that was continuously suppressed in parallel with the reduction in plasma ANF concentration, the excretion of electrolytes returned to basal levels, This increase in electrolyte excretion could not be due to a fall in plasma arginine vasopressin concentration, because arginine vasopressin levels did not change at any time after ANF antiserum treatment. This suggests that, although ANF may participate in the regulation of electrolyte excretion, factors other than ANF may contribute to the maintenance of urinary electrolyte excretion rate during normal conditions, Reduction of endogenous ANF concentration in the fetus did not appear to affect plasma arginine vasopressin concentration. The lack of change in face of urinary changes further supports the role of ANF in the regulation of renal function in the ovine fetus. The reduction in circulating ANF levels increased plasma angiotensin II concentration. A similar increase in plasma renin activity in adult rats after ANF antiserum treatment has been reported by Naruse et al. 16 The increase in angiotensin II concentration in the present study could have been due to the short-term fall in arterial pressure after ANF antiserum injection. Hypotension in the mature ovine fetus is a potent stimulus for increasing plasma renin activity. IS Although it is tempting to speculate that ANF is inhibitory to the renin-angiotensin system as seen in the adult,22 it is difficult to separate the effect of hypotension from that of ANF blockade in the stimulation of angiotensin II release. Besides, infusion of ANF into the ovine fetus was not associated with a fall in plasma renin activity.3 In summary, this study demonstrates that in the ovine fetus plasma ANF plays a role in the regulation of normal arterial pressure and blood volume and in the maintenance of basal urinary output. Furthermore, ANF appears to have little effect in the maintenance of fetal heart rate. Finally, an interaction of ANF with the other vasoactive hormones, angiotensin II and arginine vasopressin, appears to be minimal. REFERENCES 1. Brace RA, Cheung CY. Cardiovascular and fluid responses to atrial natriuretic factor in sheep fetus. Am ] PhysioI1987;253:R561-7. 2. Shine P, McDougall ]G, Towstoless MK, Wintour EM. Action of atrial natriuretic peptide in the immature ovine kidney. Pediatr Res 1987;22:11-5. 3. Brace RA, Bayer LA, Cheung CY. Fetal cardiovascular,
ANF antiserum in fetus
4.
5. 6. 7.
8. 9. 10. II. 12. 13.
14. 15.
16.
17.
18.
19. 20. 21. 22.
1567
endocrine, and fluid responses to atrial natriuretic factor infusion. Am] Physiol 1989;257:R580-7. Robillard ]E, Nakamura KT, Varille VA, Andresen AA, Natherne GP, Van Orden DE. Ontogeny of the renal response to natriuretic peptide in sheep. Am ] Physiol 1988; 254: F634-41. Cheung CY, Gibbs DM, Brace RA. Atrial natriuretic factor in maternal and fetal sheep. Am] PhysioI1987;252:E27982. Ervin MG, Ross MG, Castro R, et al. Ovine fetal and adult atrial natriuretic factor metabolism. Am ] Physiol 1988;254:R40-6. Ross MG, Ervin MG, Lam RW, Castro L, Leake RD, Fisher DA. Plasma atrial natriuretic peptide response to volume expansion in the ovine fetus. AM ] OBSTET GVNECOL 1987; 157: 1292-7. Cheung CY, Miner LK, Brace RA. Hyperosmolality elevates plasma atrial natriuretic factor in the ovine fetus. Am] PhysioI1989;257:E466-72. Cheung CY, Brace RA. Hemorrhage-induced reductions in plasma atrial natriuretic factor in the ovine fetus. AM ] OBSTET GVNECOL 1991;165:474-81. Cheung CY, Brace RA. Fetal hypoxia elevates plasma atrial natriuretic factor concentration. AM] OBSTET GvNECOL 1988; 159: 1263-8. Brace RA. Thoracic duct lymph flow and its measurement in chronically catheterized sheep fetus. Am ] Physiol 1989;256:HI6-20. Brace RA, Brittingham DS. Fetal vascular pressure and heart rate responses to non labor uterine contractions. Am ] PhysioI1986;251:R409-16. Rudd A, Plavin S, Hirsch AT, Ingelfinger ]R, Dzau VJ. Atrial natriuretic factor-specific antibody as a tool for physiological studies. Evidence for role of atrial natriuretic factor in aldosterone and renal electrolyte regulation. Circ Res 1989;65:1324-9. Miner LK, Brace RA, Cheung CY. Temporal relationships among fetal urine flow, ANF, and AVP responses to hypertonic infusion. Am] PhysioI1990;258:R469-75. Brace RA. Blood volume in the fetus and methods for its measurement. In: Nathanielsz PW, ed. Animal models in fetal medicine. IV. Monographs in fetal physiology. Ithaca, New York: Perinatology Press, 1984;5:19-36. Naruse M, Obana K, Naruse K, et al. Antisera to atrial natriuretic factor reduces urinary sodium excretion and increases plasma renin activity in rats. Biochem Biophys Res Commun 1985; 132:954-60. Sasaki AA, Kida 0, Kita T, et al. Effects of antiserum against Ol-rat atrial natriuretic polypeptide in spontaneously hypertensive rats. Am ] Physiol 1989;257: H1104-9. Rawashdeh NM, Ray ND, Sundberg DK, Rose ]C. Comparison of hormonal responses to hypotension in mature and immature lambs. Am] PhysioI1988;255:R6772. Laragh]H. The endocrine control of blood volume, blood pressure and sodium balance: atrial hormone and renin system interactions.] Hypertens 1986;4:S143-56. Lumbers ER, Stevens AD. The effects of fusemide, saralasin and hypotension on fetal plasma renin activity and on fetal renal function.] Physiol 1987;393:479-90. Lingwood B, Hardy K], Coghlan]P, Wintour EM. Effect of aldosterone on urine composition in the chronically cannulated ovine fetus.] Endocrinol 1978;76:553-4. BurnettJC Jr, Granger JP, Opgenorth TJ. Effects of synthetic atrial natriuretic factor on renal function and renin release. Am] Physiol 1984;247 :F863-6.