Reflex control of ACTH, cortisol, AVP, and renin responses to slow hemorrhage in fetal sheep

MISSING IMAGE

Material Information

Title:
Reflex control of ACTH, cortisol, AVP, and renin responses to slow hemorrhage in fetal sheep
Physical Description:
xii, 113 leaves : ill. ; 29 cm.
Language:
English
Creator:
Chen, Hong-Gen
Publication Date:

Subjects

Subjects / Keywords:
Corticotropin -- pharmacology   ( mesh )
Argipressin -- pharmacology   ( mesh )
Hydrocortisone -- pharmacology   ( mesh )
Renin -- pharmacology   ( mesh )
Hemorrhage   ( mesh )
Sheep -- physiology   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 97-112).
Statement of Responsibility:
by Hong-Gen Chen.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 25487736
ocm25487736
System ID:
AA00011191:00001

Full Text











REFLEX CONTROL OF ACTH, CORTISOL, AVP, AND RENIN
RESPONSES TO SLOW HEMORRHAGE IN FETAL SHEEP


















By

HONG-GEN CHEN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1991



































To my parents Qinshan Chen and Xiuying He, and to my husband Ke Wu.















ACKNOWLEDGEMENTS


I wish to express my gratitude and appreciation to Dr. Charles E. Wood, chairman of

my supervisory committee, for his patience, guidance, and encouragement throughout the

fulfillment of this work. He has not only provided an invaluable academic direction in the

research but also set an example of responsibility and dedication.

My sincere thanks are extended to the members of my committee, Drs. Robert M.

Abrams, Willa H. Drummond, Maureen Keller-Wood, and Wendell N. Stainsby. In particular,

I wish to thank Dr. Maureen Keller-Wood for her very helpful suggestions and criticism for

the project and Dr. Willa H. Drummond for introducing me to this program and her continued

encouragement and advice. Special thanks must be also extended to Mr. Curt Kane and Mrs.

Ellen Manlove, and also, to Miss. Christine Taranovich, Miss. Jennifer Johnson, and Dr.

Timothy Cudd for their friendship and expert technical assistance in running hormone assays

or doing surgeries and experiments. Gratefulness also goes to the faculty members and my

fellow students in the Department of Physiology for their concern, advice and support.

Last but not least, I am greatly indebted to my husband, Ke Wu, and my parents for

their love, backing, and understanding. My gratitude to them is beyond description.















TABLE OF CONTENTS



ACKNOWLEDGEMENTS ............................................. iii

LIST OF TABLES .................................................... vi

LIST OF FIGURES .................................................. vii

KEY TO ABBREVIATIONS ............................................ ix

ABSTRACT ........................................................ xi

CHAPTER 1 INTRODUCTION ........................................ 1

CHAPTER 2 BACKGROUND REVIEW .................................. 3

2.1 A Brief Review of the Fetal Circulatory System ..................... 3
2.2 The Development of Autonomic Control of the Heart ................ 4
2.3 The Development of Baroreflex Control of Circulation in Fetuses ....... 6
2.4 The Development of Chemoreflex Control of Circulation in Fetuses ...... 8
2.5 The Fetal ACTH, Cortisol, AVP and Renin Responses ............... 11
2.5.1 ACTH and Cortisol .................................. 11
2.5.2 A V P ............................................. 16
2.5.3 R enin ............................................ 18
2.6 Sum m ary ................................................. 19

CHAPTER 3 GENERAL METHODOLOGY .............................. 21

3.1 Surgical Procedures of Catheterization and Denervation .............. 21
3.2 General Preparation for Experiments and Hemodynamic Data
C collection ................................................ 22
3.3 Handling and Analysis of Blood Samples ......................... 23
3.4 Calculations and Statistical Analyses ............................ 25

CHAPTER 4 THE ROLE OF VAGOSYMPATHETIC AFFERENT FIBERS
IN THE CONTROL OF ACTH, VASOPRESSIN, AND RENIN RESPONSES
TO HEMORRHAGE IN FETAL SHEEP ...................... 26

4.1 Introduction .............................................. 26
4.2 M materials and M ethods ....................................... 27
4.3 R results .................................................. 27
4.4 D discussion ................................................ 39









CHAPTER 5 ACTH, AVP, AND RENIN RESPONSE TO INTRAVENOUS
INFUSION OF HYDROCHLORIC ACID ARE CHEMORECEPTORS
RESPONSIBLE FOR FETAL HORMONE SECRETION? .......... 44

5.1 Introduction .............................................. 44
5.2 M methods ................................................. 45
5.3 R results .................................................. 46
5.4 D discussion ................................................ 53

CHAPTER 6 REFLEX CONTROL OF FETAL ARTERIAL PRESSURE
AND HORMONAL RESPONSES TO SLOW HEMORRHAGE ...... 58

6.1 Introduction .............................................. 58
6.2 M methods ................................................. 58
6.2.1 Experimental Protocol ................................ 59
6.2.2 Calculated Variables ................................. 60
6.3 R results .................................................. 62
6.3.1 Changes in Hemodynamic Variables During Hemorrhage ...... 62
6.3.2 Changes in Blood Gases During Hemorrhage ............... 62
6.3.3 Changes in Plasma Hormone Levels ...................... 64
6.3.4 Blood Volume Restitution ............................. 64
6.4 Discussion ................................................ 67
6.4.1 Alterations in Blood Gases and Vascular Pressures ........... 72
6.4.2 Reflex Hormonal Responses ........................... 74
6.4.3 Defense of Blood Volume ............................. 75

CHAPTER 7 THE ACTH AND AVP RESPONSES TO NORMOXIC
HYPERCAPNIA IN FETAL AND MATERNAL SHEEP .......... 77

7.1 Introduction .............................................. 77
7.2 M methods ................................................. 78
7.3 R results .................................................. 79
7.3.1 Responses in Fetuses ................................. 79
7.3.2 Responses in Ewes ................................... 81
7.4 Discussion ............................................... 88

CHAPTER 8 SUMMARY ............................................ 91

REFERENCES .................................................... 97

BIOGRAPHICAL SKETCH ........................................... 113















LIST OF TABLES


Table 4-1. Values of correlation coefficients relating fetal plasma hormone
concentrations to pHa, MAP, and CVP ......................... 35

Table 5-1. Results of ANOVAs ...................................... 48

Table 5-2. Fetal plasma cortisol concentrations .......................... 51

Table 5-3. Fetal hematocrit ......................................... 54

Table 6-1. Initial and final fetal blood volume and fetal blood volume
restitution. ............................................. 69

Table 7-1. Fetal Means & SEs of HCT, plasma potassium & sodium ........... 82

Table 7-2. Maternal means & SEs of HCT, plasma potassium & sodium ........ 87















LIST OF FIGURES


Figure 4-1.

Figure 4-2.

Figure 4-3.

Figure 4-4.

Figure 4-5.

Figure 4-6.

Figure 4-7.

Figure 4-8.


Figure 5-1.

Figure 5-2.

Figure 5-3.


Figure 5-4.


Figure 6-1.

Figure 6-2.

Figure 6-3.

Figure 6-4.

Figure 6-5.

Figure 6-6.

Figure 7-1.

Figure 7-2.


Fetal MAP and HR during hemorrhage ........................ 28

Fetal CVP during hemorrhage ............................... 30

Fetal PaO2, Paco2, and pHa during hemorrhage .................. 31

Fetal Hct during hemorrhage. ............................... 32

Fetal Plasma ACTH and cortisol concentrations during hemorrhage. 33

Fetal Plasma AVP concentration and PRA during hemorrhage. ....... 34

Relations among pHa and fetal plasma ACTH, AVP and PRA. ....... 37

Distribution of r values of fetal pHa, MAP, and CVP vs
Lg ACTH, AVP, or PRA................................... 38

Fetal pHa, Paco20 and Pao2 during intravenous infusion of HCI ..... 47

Fetal MAP and HR during intravenous infusion of HCI ........... 49

Fetal plasma ACTH, AVP, and PRA during intravenous
infusion of HCI. ......................................... 50

Relationships between arterial H+ concentration, Pao02 and plasma ACTH,
AVP, and HR in 0.50 meq/min group.......................... 52

Fetal MAP, HR and CVP during hemorrhage. ................... 63

Fetal Pa02, pH. and Paco2 during hemorrhage. .................. 65

Fetal plasma ACTH, cortisol, AVP and PRA during hemorrhage ..... 66

Fetal volume restitution. ................................... 68

Fetal Hct, plasma protein, potassium and sodium ................. 70

Percent changes in fetal plasma protein concentration. ............. 71

Fetal pHa, Paco02 and Pa02 in normocapnia and hypercapnia studies. 80

MAP, CVP and HR in normocapnia and hypercapnia studies. ....... 83









Figure 7-3.


Figure 7-4.


Figure 7-5.


Figure 7-6.


Fetal plasma ACTH and cortisol concentration in
normocapnia and hypercapnia ...............................

Fetal plasma AVP concentration in normocapnia and hypercapnia
studies. ................................................

Maternal pHa, Paco2 and Pa02 in normocapnia and hypercapnia
experim ents. ............................................

Maternal plasma ACTH, cortisol and AVP concentration in
normocapnia and hypercapnia studies ..........................















KEY TO ABBREVIATIONS


ACTH adrenal corticotropic hormone

ANOVA analysis of variance

AVP arginine vasopressin

BVo initial blood volume

BVt subsequent blood volume at each time point

CVP central venous pressure

EDTA ethylenediaminetetraacetic acid

FBS fetal bovine serum

FRCV the fractional red cell mass

Hct hematocrit

Hct, initial hematocrits

Hcts subsequent Hct at each time point

HR heart rate

MAP mean arterial pressure

Po the plasma protein concentration before hemorrhage

%APa actual % changes of plasma protein concentration

Paco2 arterial PCO2

Pao2 arterial P02

%APc theoretical %changes plasma protein concentration

Pc plasma protein concentration

pHa arterial pH

POMC proopiomelanocortin

ix









PRA plasma renin activity

Prf the protein concentration in restitution fluid

Pt the protein concentration at each time point

PVN paraventricular nucleus

RBCo initial red blood cell mass

RBCt subsequent red blood cell mass at each time point

%R percent changes of the blood volume restitution

Rt the blood volume restitution during hemorrhage

RIA radioimmunoassay

SEM standard error of the mean

SON supraoptic nucleus

Vr the blood volume removed during hemorrhage















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

REFLEX CONTROL OF ACTH, CORTISOL, AVP, AND RENIN
RESPONSES TO SLOW HEMORRHAGE IN FETAL SHEEP

By

Hong-gen Chen

May 1991


Chairman: Charles E. Wood
Major Department: Physiology

To investigate the afferent pathway for controlling fetal ACTH, AVP and PRA

responses to slow hemorrhage, forty-six chronically catheterized late gestational fetal sheep

were studied.

Experiment I investigated the role of the vagus nerve. One hundred and forty-three

ml blood was withdrawn within two hours from five bilaterally vagotomized and six intact

fetuses. No significant differences in the stimulation of the hormones were found between

groups. The hormone stimulation was better correlated to the changes in pHa than in blood

pressures, suggesting that the responses in the fetus are possibly by chemoreceptors.

Experiment II tested the effect of chemoreceptors. HC1 was infused at three different

rates to fetuses. Infusions at the highest rate (0.50 meq/min) caused severe metabolic

acidemia and significantly stimulated plasma ACTH and AVP but not PRA. It is speculated

that the ACTH and AVP responses to acidemia are chemoreceptor mediated.

In Experiment III, 143 ml blood was withdrawn within two hours from twenty-six

fetuses, which were either intact or bilaterally carotid sinus denervated plus vagotomized, in

order to identify the peripheral chemoreceptors' effect during hemorrhage. Denervated

xi








fetuses were studied with or without simultaneous infusion of phenylephrine. None of the

hormonal responses was attenuated by denervation, and phenylephrine infusion attenuated or

inhibited ACTH and AVP but not PRA when the exaggeration of MAP, pHa and Paco2 in

denervated fetuses was prevented. Thus the hormonal responses to slow hemorrhage was

concluded not being mediated by peripheral chemoreceptors.

Experiment IV quantified fetal plasma ACTH and AVP responses to hypercapnia and

tested the role of peripheral chemoreceptors. Six out of eleven fetuses were bilaterally carotid

sinus denervated and vagotomized. Each subject was studied under normocapnia control and

hypercapnia. Hypercapnia was induced by a CO2 mixed-gas inhaled by the ewe. Hypercapnia

doubled fetal plasma AVP; denervation did not attenuate the response. ACTH was not

stimulated. We conclude that hypercapnia, associated with smaller decreases in pHa than that

in Experiment II, is not a stimulus to ACTH and is a mild stimulus to AVP. No hormone is

controlled by peripheral chemoreceptors. The afferent pathway for mediating ACTH, AVP

and PRA response to slow hemorrhage in fetal sheep needs to be investigated further.















CHAPTER 1
INTRODUCTION


Various intrauterine stresses, such as hemorrhage, hypotension, hypoxia, acidemia and

hypercapnia, proved to stimulate ACTH, cortisol (Alexander et al., 1973; 1974b; Challis et al.,

1989), AVP (Rurak, 1979; Drummond et al., 1977; 1978) and renin (Zubrow et al., 1988;

Robillard et al., 1982) secretions in late gestational fetal sheep. Only a few experiments have

been done to investigate the specific afferent pathways for mediating the hormonal responses

to certain stresses. For example, the ACTH, AVP and renin responses to vena caval

obstruction were significantly correlated to the magnitude of the decrease in blood pressure

(Wood and Rudolph, 1983), and this response was mediated by afferent fibers in the carotid

sinus and/or aortic nerves (Wood, 1989a). The findings that the response of vasopressin to

hypoxemia is greater when the vagi are intact suggest a regulatory effect of the

chemoreceptors (Rurak, 1978).

Hemorrhage has long been used as a stimulus for investigating the cardiovascular and

endocrine responses to stress. In adult animals, the ACTH, cortisol and AVP (Wang et al.,

1983) responses to hypotensive hemorrhage have been shown to be mediated by cardiovascular

receptors located at the junction of the vena caval and atria, in the aortic arch, and in the

carotid vasculature (Brennan et al., 1971; Cryer and Gann, 1973; Gann et al., 1964; Share and

Levy, 1962; and Ludbrook, 1990). For mild or non-hypotensive hemorrhage, the mediator

for the hormonal responses proved to be atrial type-B receptors (Cryer and Gann, 1973; Wang

et al., 1983; Quail et al., 1987).

Fetuses have the ability to compensate more effectively for their volume depletion

than adults. A slow hemorrhage, which usually causes little or no change in arterial blood








2

pressure, can initiate large hormonal responses. It is suggested that these hormonal responses

are more important for fetal survival, because the fetal cardiovascular control mechanisms,

such as autonomic innervation of the heart (which is the major and most rapid regulator of

adult circulation), baroreflex and chemoreflex systems, etc., are not fully developed and fully

functioning in some species, even at term. It is important to further investigate the regulation

of hormonal responses to better understand the mechanism of fetal survival in intrauteral life,

as well as the cardiovascular and endocrine developmental procedures.

Reflex control of the fetal hormone secretion depends upon the adequate development

of related reflex components (sensors, afferent and efferent fibers and central nuclei) and the

maturation of the related endocrine organs or cells. Based on adult studies, the circulatory

reflex control system is very important to the hormonal responses mentioned above. We

assume that this is also important in fetuses.

The fetal afferent pathway for regulating hormonal responses to slow hemorrhage

remains unknown. The purpose of the studies in this dissertation is to investigate the afferent

pathway regulating ACTH, cortisol, AVP and renin responses to slow hemorrhage in late

gestational fetal sheep. Chapter 2 gives a background review of the developmental aspects of

the circulatory control and the maturation of the related endocrine organs. Chapter 3

describes the general methodology of the experiments involved. Chapters 4 and 6 evaluate

the effect of vagotomy (which interferes with the afferent fibers from atrial type B

receptors), and vagotomy plus carotid sinus denervation (which denervates the peripheral

baroreceptor and chemoreceptor afferent fibers) respectively, in fetal hormone responses to

slow hemorrhage; chapters 5 and 7 study the chemoreceptors' effect on hormone responses

to severe and milder hypercapnic acidemia, which was induced by HCI infusion in the former

and maternal hypercapnic mix-gas inhalation in the latter. Chapter 8 is a summary and

conclusion.














CHAPTER 2
BACKGROUND REVIEW



2.1 A Brief Review of the Fetal Circulatory System


The fetal cardiovascular system differs anatomically and functionally from the adult

system to meet the growth and developmental requirements of intrauteral life. There are two

circulations that are in parallel: the systemic and the umbilical-placental circulations. The

umbilical vein and arteries are vessels that connect these two circulations. The ductus venosus

carries most of the placental venous blood flow that bypasses the liver; the foramen ovale is

an opening between the right and left atria located in the interatrial septum; and the ductus

arteriosus is a vessel that connects the pulmonary trunk and the aorta. All these "designs" are

based on the fact that fetal alveoli are fluid-filled; the gastrointestinal tract and the kidneys

are much less functional than they become after birth. The two circulations ensure that the

fetus gets enough oxygen and nutrients to supply it's growth and developmental requirement

and to eliminate waste products.

Compared to the adult, the most remarkable features of fetal cardiovascular functions

are the relatively higher cardiac output and heart rate (HR), which are 500 ml/kg/min and

150-180 bpm, respectively (Rudolph and Heymann, 1972; 1973), and lower mean arterial

pressure (MAP), which is 36-44 mm Hg (Yardley et al., 1983) in late gestational sheep fetuses.

The HR progressively falls and the blood pressure progressively increases in parallel with the

growth and development of the fetus. In fetal lamb, the HR was reported to fall 0.67

beats/min/day and the MAP was rise 0.46 mm Hg/day from 100 days of gestation (0.68 term,

term = 147 days) to term (Boddy et al., 1974b).








4

Fetuses have limited ability to increase their stroke volumes (Rudolph and Heymann,

1970; 1974; Rudolph et al., 1981). Examination of isolated fetal heart indicates that the heart

muscle cells are smaller and the amount of noncontractile mass is considerably greater in

fetuses than in the adults. Fetal sarcomeres are smaller in shape and are not well organized

(Friedman, 1973). The fetal resting or passive tension of the heart muscle is higher and the

compliance is much lower. The "stiffness" of the fetal heart restrict the performance of the

heart. Although previous consumption that fetal heart behaves at the upper limits of its

ventricular function curve (Gilbert, 1980) is proven to be incorrect because the influence of

the stroke volume by after load was ignored (Hawkins et al., 1988), the ability of the

increment of stroke volume in fetuses is still considered to be limited.

Almost half of the fetal cardiac output goes to the placenta through umbilical vessels.

It is known that umbilical vasculature is not under neural control, but vascular resistance can

be influenced by several vasoactive substances, such as angiotensin II, prostaglandins,

bradykinin, etc. (Mott and Walker, 1983). Placenta flow is also influenced indirectly by

systemic vascular tone. When systemic flow decreases or increases by vasoconstriction or

vasodilation, the placental flow increases or decreases respectively. However, under normal

physiological conditions, the fetus can always maintain an optimal flow ratio between these

two circulations. The mechanism by which the fetus maintains the appropriate balance

between placental and systemic circulations is not fully understood.



2.2 The Development of Autonomic Control of the Heart


The development of autonomic innervation of the fetal heart is the basis for the

appearance of baro and chemoreflex control of the circulation. Reports from studies of

rapidly developing chick embryos reveal that there are receptors and inactivating enzymes for

both parasympathetic and sympathetic nerves on day 2 (0.1 term, term=21 days); cholinergic

neuron growth within the sinoatrial node on day 6 (0.28 term); the first detectable release of








5

acetylcholine on day 10 (0.48 term), and the appearance of the functional vagal

neurotransmission, acetylcholine, and the acetylcholine synthetic enzyme, choline

acetyltransferase, on day 12 (0.57 term) (Pappano, 1977). Electrical stimulation of isolated

human fetal atria also caused release of acetylcholine, and depressed contractility of the heart

from 91 days of gestation (0.3 term). In fetal lambs, stimulation of the cervical vagus nerve

produced a small inhibition of the heart as early as 60 days of gestation (0.4 term) (Dawes,

1968). There is also evidence that effective parasympathetic neurotransmission develops in

guinea pig and rabbit fetuses (Vlk and Vincenzi, 1977). Myocardial innervation of the

sympathetic nerve in most species is found to be incomplete at birth (Friedman et al., 1968;

Lebowitz et al., 1972; Lipp and Rudolph, 1972). Sympathetic cardiac innervation in chick

embryos was proved by histochemical technique, which followed the same sequence of

cholinergic innervation, but the appearance of sympathetic nerves and the onset of effective

neurotransmission was delayed (Pappano, 1977). A similar sequence of autonomic

development occurs in man and other mammals, except the rat, which does not have cardiac

sympathetic nerves until about the end of the first week of life (Marvin et al., 1980; Vik,

1979). These species differences of the development of the autonomic nerve system are

thought to correspond roughly to the general maturity of the animal at birth. The more

dependent the species are on their mothers at birth, the less well developed the system is. The

potential autonomic imbalance due to the delayed development of sympathetic nerves might

be compensated for by the earlier development of cardiac adrenergic receptors. In sheep

fetuses, f-adrenergic receptors can be stimulated at about 60 days of gestation (0.4 term) by

appropriate agonists, which are ahead of the reported occurrence of sympathetic innervation

(Barrett et al., 1972) and in parallel to the reported earliest stage of the parasympathetic

stimulation (Dawes, 1968; Dawes et al., 1968). The point that greater development of

parasympathetic than sympathetic nerve control of the heart, even at end of fetal life, has

been very well proven and accepted.












2.3 The Development of Baroreflex Control of Circulation in Fetuses


Many efforts have been made to investigate the developmental aspects of arterial

baroreflex control of fetal circulation and it has long been known that these receptors are

active during fetal life. Two of the most common methods used in in vivo studies are: 1)

demonstrating the reflex HR responses following changes in arterial and/or carotid sinus

pressure, and 2) testing the HR response with or without intact baro-afferent nerve fibers.

As early as the 1940s, Barcroft reported a functional baroreflex in anesthetized and

exteriorized young fetuses in anesthetized ewes by injecting adrenaline and noradrenaline into

lamb fetuses (Barcroft, 1946). He extended his findings to goat, rabbit and sheep fetuses by

lighting their umbilical cords. The hypertension produced was associated with a significant

bradycardia. He also proved that this HR response was abolished by vagotomy. Since then,

numerous experiments in this area have been performed among different species.

It was found that the baroreflex was partially inhibited by carotid denervation and

totally abolished by subsequent aortic denervation during fetal life (Rudolph and Heymann,

1973), which suggests that these two nerve fibers are important for the responses. Rudolph

and his associates developed an unanesthetized, unstressed, and selectively peripheral

baroreceptor denervated fetal model (Itskovitz and Rudolph, 1982). Since then, the function

of these peripheral baroreflex afferent fibers has been better characterized. Baroreceptor

activity was also demonstrated directly by recording the impulse from nerve fibers. Biscoe

et al. (1969) measured phasic electrical activity in small bundles of fibers dissected from the

carotid sinus nerve in 120-147 days of gestation (0.8-0.99 term) fetal sheep, and Ponte and

Purves (1973) measured the activity in vago-afferent fibers in close-to-term sheep fetuses.

They found that the measured nerve activity is synchronous with the arterial pulse, suggesting

a functional baroreflex arc. Blanco et al. reported that spontaneous carotid and aortic








7

baroreceptor discharge is detectable from 85 days of gestation (0.6 term) in sheep fetuses

(Blanco et al., 1985).

However, the earliest stage of the fetal lamb that can be proven to present a functional

baroreflex control of the heart is 81 days of gestation (0.55 term) (MacDonald et al., 1980).

The baroreflex control of the heart rate matures with advancing gestation. This is

demonstrated by increased positive or negative HR responses to brief elevations or reduction

of arterial pressure, respectively. Shinebourne et al. (1972) studied fetal lambs from 85-145

days of gestation (0.6-0.99 term) and found that baroreflex sensitivity progressively increased

with advancing gestation until term. In younger fetuses, a sudden increase in aortic pressure

was associated only sporadically with a baroreflex response. When it was elicited, the response

was quite limited. Other studies also demonstrated that nearly all species experience a

progressive postnatal maturation of the reflex to adult levels (Gootman et al. 1979; Vatner and

Manders, 1979; Walker et al., 1990).

Further investigations tested the physiological role of baroreceptors in intrauterine

cardiovascular regulation. The baroreflex was considered unimportant for fetuses except

under conditions of elevated pressure, because the threshold of the reflex in adults is above

the range of the normal fetal arterial blood pressure. This point was challenged by Yardley

et al. (1979; 1983), who continuously recorded fetal arterial blood pressure during a 24-hour

period in both intact and sino-aortic denervated fetal lambs. He found that denervation

significantly increased the natural variability of arterial pressure. The coefficients of

variations of mean arterial pressure (defined as stand deviation of the mean arterial pressure

over the mean value of the mean arterial pressure) were twice as great in denervated fetuses

compared to intact ones. This result is similar to the result from another experiment

conducted by Cowley et al. (1973), who continuously recorded the BP within 24-hour on

active, unanesthetized adult dogs. Marked fluctuations in fetal basal arterial blood pressure

and heart rate after sinoaortic denervation were also observed by Itskovitiz et al. (1983),








8
although the MAP and HR are not different from control values. These results strongly

suggest that arterial baroreceptors have a natural role in day-to-day fetal arterial pressure

regulation.

Mechanoreceptors in atria, ventricles, and pulmonary arteries are potential sites of

cardiovascular reflexes that have not been systematically investigated in the fetus. In 1975,

Nuwayhid et al. (1975) reported that Bezold-Jarisch reflex could be stimulated by veritrodine

injection in lamb fetuses close to term (Wt > 2700 g). In their experiments, the assumption

that ventricular receptors affect fetal circulation was based on the reflex decrease of HR

response to decrease of arterial pressure (Oberg, 1976). It is known that in adult animals, HR

is accelerated by hemorrhage through arterial baroreceptor stimulation. In adult animals, only

very rapid and severe hemorrhage can produce bradycardia, which is mediated by the

increased firing rate of ventricular receptors that supersede the effect of arterial baroreceptors

(Oberg and White, 1970). In fetuses, tachycardia only occurs when blood pressure is mildly

reduced by hemorrhage or by venous occlusion (Wood et al., 1979; MacDonald et al., 1980).

With greater decrease in pressure, the transient increase in heart rate is reversed and

bradycardia occurs (Toubas et al., 1981; Wood and Rudolph, 1983a). Following atropine

administration the bradycardia is abolished, indicating vagal efferent involvement of the

reflex. These studies suggest that the ventricular receptor is functional during fetal life.




2.4 The Development of Chemoreflex Control of Circulation in Fetuses


In adult life, besides their dramatic effects on respiration, chemoreceptors are also

important in the control of circulation. The peripheral chemoreceptors are located in the

carotid and aortic bodies, and the central chemoreceptors are in the medulla oblongata. The

stimulation of the central chemoreceptors alone always results in a reflex tachycardia and a

very substantial increase in arterial blood pressure. Both effects are thought to be protective,









9
an attempt to circulate more blood through the affected areas in order to bring about a

decrease in carbon dioxide tension or an increase in oxygen tension. Also, it has been

suggested that the primary effect of carotid body stimulation in adult animals is bradycardia

(De Burgh Daly, 1972), and by contrast, the effect of aortic chemoreceptor stimulation

produces tachycardia (Sleight, 1974). The chemoreceptors of the aortic body in adults are

considered to be more important than those of the carotid body in circulatory regulation

(Comroe et al. 1964).

The aortic chemoreceptor was receptive to stimulation at about 100 days of gestation

(0.68 term) (Dawes et al., 1969a; 1969b). An increase in arterial blood pressure, a decrease

in HR, and vasoconstriction in the hind limb occurred when arterial Pao2 and Paco2 was

changed over the physiological range, or when cyanide or nicotine was infused into fetal

lambs. These responses can be abolished either by vagotomy or aortic denervation (Baillie et

al., 1971; Dawes et al., 1969a; 1969b), which have specifically demonstrated the aortic

chemoreceptor's effect. Direct recordings of fetal afferent electrical activity in aortic

chemoreceptor fibers (Ponte and Purves, 1973) reveal that the responses to asphyxia and

chemical stimulation are similar to those in the adults. The point that aortic chemoreceptors

are active and important for fetal cardiovascular regulation has been very well accepted.

Early reports from fetal carotid chemoreceptor studies are controversial. Some claim

that they are not functioning before birth, based on early direct nerve fiber recording studies

and acute hypoxia studies (Cross and Malcolm, 1952; Dawes et al., 1982, Boddy et al., 1974;

Dawes et al., 1969a; Purves and Biscoe, 1966), and some claimed that it is active based on

response to infusion of cyanide (Dawes et al., 1969b). Others observed that carotid

denervation showed no effect on fetal breathing movement, and only large quantities of

chemical stimulation of the carotid body affect fetal breathing movement (Jansen et al., 1981;

Purves, 1981). These results led to a conclusion that carotid chemoreceptor is inactive or

plays only a small part, if any, in intrauteral fetal life regulation. However, later studies, most








10

from chronic animal preparations, suggested that the carotid body was more active prenatally

than previously thought. Blanco et al. (1982) made direct recordings from carotid afferent

fibers in fetal sheep and found that a random activity at about 5 Hz occurred at a Pao2 of 25

mm Hg. The firing rate was increased by retrograde injection of CO2-saturated saline into

the lingual artery. Itskovitz and Rudolph (1987) performed carotid sinus denervation in sheep

fetuses at 120 days of gestation (0.8 term). The cardiorespiratory response to intracarotid

injection of cyanide was eliminated, which also supports the observation that carotid

chemoreceptors are functional during at least the last third of gestation. A resetting of the

chemoreceptor sensitivity on the day of birth and 5-10 days after birth was also reported

(Blanco et al., 1984; Hanson et al., 1986). The carotid chemoreceptor is actually more active

than previously thought.

Some reports indicate that direct stimulation of central chemoreceptors in fetal sheep

increases fetal heart rate (Jansen, 1975; Jansen and Read, 1977; Jansen et al., 1981). However,

most of the studies that test brainstem chemosensitivity of the brainstem in fetal lambs were

related to the regulation of fetal breathing movement. Ventricular-cisternal perfusion with

artificial cerebra-spinal fluid containing various concentrations of HCO3- increased the depth

and incidence of breathing movements as CSF[H+] increased from 45 to 73 neq/L (Bissonnette

et al., 1980). A delayed fetal breathing movement was observed after intravenous infusion

of NH4+ and HCI, which can be attributed to slow penetration of the H+ into the interstitium

of the brain (Hohimer and Bissonnette, 1981). The developmental aspect of central

chemoreceptors and the interaction of the central and peripheral chemoreceptors during fetal

life need to be further studied.










2.5 Fetal ACTH. Cortisol. AVP and Renin Responses


2.5.1 ACTH and Cortisol


The structural and functional development of the hypothalamus-pituitary-adrenal axis

has been intensively investigated since 1960s. The glucocorticoids have been demonstrated

to play a vital role in fetal life in various species. Glucocorticoids are involved in the

maturation of several fetal organs or tissues, including the fetal lungs, adrenal medulla and

small intestine. Glucocorticoids also participate in initiating parturition, and in mediating

stress responses, which are important for fetal survival.

In human fetuses, the hypothalamus and pituitary are structurally showed to exist and

pituitary hormones in granules can be detected at 42 days gestation (0.18 term); the

hypothalamo-hypophysial portal system begins, and the releasing factors are identified at

about 49 days gestation (0.2 term). By 70 to 91 days of gestation (0.29-0.37 term), the

pituitary and hypothalamic tissues can respond in vitro to stimulatory or inhibitory stimuli.

By mid-gestation, the fetal hypothalamic-pituitary axis is functional and the feedback control

mechanism can be demonstrated. The pituitary-portal system presents as early as 49 to 56

days (0.2-0.23 term) and is complete at 126-140 days (0.53-0.58 term) in humans (Decherney

and Naftolin, 1980).

The early development of the hypothalamus and pituitary is also established in other

species, including sheep. Proopiomelanocortin (POMC), a precursor of ACTH and three other

peptides: pro-'yMSH, P-LPH and i-endorphrin, were detected by an immunocytochemical

staining technique in the pituitary as early as 38 days of gestation (0.26 term) in sheep fetuses

(Mulvogue et al., 1986) and at 34 days (0.3 term, term = 114 days) in the porcine pituitary

(Dacheux, 1984). In the intermediate lobe of sheep fetuses, the POMC cells did not stain to

show positively until 60 days of gestation (0.4 term). From 70 days on (0.48 term), the

number of immunoreactive cells gradually increases, and by 90 days (0.61 term), almost all








12

the pars intermedia cells were intensely stained (Perry et al., 1985). Both "fetal" type

corticotrophs (cells that are unique to the fetal pituitary) and "adult" type corticotrophs, which

both contain POMC, were discovered in the fetal pituitary. Before 130 days (0.88 term), the

"fetal" corticotrophs were predominant cell types; then these cells gradually disappeared and

were replaced by the "adult" corticotrophs before term (Perry et al., 1985). By using ink-

filling techniques, the hypothalamo-hypophysial blood system in sheep fetus was found to be

patent from 45 days gestation (0.31 term), which gives evidence that the pituitary is

potentially able to respond to hypothalamic releasing factors from this age onwards (Levidiotis

et at., 1989). The fetal pituitary is responsive to AVP stimulation, which causes it to secrete

substantial amounts of ACTH before the appearance of hypothalamic CRF. CRF terminals

cannot be identified immunocytochemically in median eminence until around 100 days of

gestation (0.68 term) (Levidiotis et al., 1987). AVP is present from 42 days (0.29 term), with

the majority of AVP fiber terminations being found in the external zone of the median

eminence. Norman and Challis (1987b) intravenously injected equimolar AVP and CRF into

sheep fetuses on days 110-115, 125-130, 135-140 of gestation (0.75-0.78, 0.85-0.88, and 0.92-

0.95 term, respectively). They found that the ovine fetal pituitary responds separately and

synergistically to AVP and CRF on days 110-115 (0.75-0.78 term), but the role of AVP in

stimulating ACTH release decreases with progressive gestational age. Plasma ACTH in sheep

fetuses has been detected as early as 59 days gestation (0.4 term) (Alexander et al., 1973b),

which also indicates the early functioning of the anterior pituitary gland.

The sheep fetal adrenal is morphologically recognizable by day 28 of gestation (0.19

term) (Ravault and Ortavant, 1977), while the zonation was not apparent until 60 days

gestation (0.41 term) (MacIsaac et al., 1989; Webb, 1980). The sheep fetal adrenal gland, a

highly active endocrine organ, is able to synthesize a considerable range of steroid hormones.

There are adrenal structural and functional differences in sheep fetuses as compared to

primates, including a permanent adrenal cortex instead of a fetal zone. This adult zone has









13

both 3-P/HSD and A5 isomerase activities; thus the fetuses do not need to rely on the placental

source of progesterone for cortisol biosynthesis (Anderson et al., 1973). A slow, steady

growth of the fetal adrenal was observed between 80 and 135 days (0.54 and 0.91 term),

followed by explosive growth during the last 10-15 days of fetal life. This increase in weight

is almost entirely due to an increase in the size of the adrenal cortex, and within the cortex

the major growth occurs in the zona fasciculate (Comline and Silver, 1961; Liggins, 1969;

Nathanielsz et al., 1972). When comparing with the weight of the adrenal gland relative to

body size, the adrenals were maximal at an early stage of gestation (40-80 days, 0.27-0.54

term) in sheep fetuses (Wintour et al., 1975), 32 days (0.48 term, term=67 days) in guinea pigs

(Moog and Ortiz, 1957), and 84-140 days (0.35-0.58 term) in humans (Lanman, 1953; 1962).

Wintour et al. (1975) studied the adrenal function from 40 days gestation (0.27 term) in sheep

fetuses. They found that the cortisol output per g body weight was greater in the fetal adrenal

than the term adrenal when the adrenal glands were incubated with ACTH. The ACTH-

stimulated effect on cortisol secretion declined significantly from 91 to 120 days gestation

(0.62 to 0.82 term). Wintour et al. also measured peripheral blood level of cortisol in fetus

between 60 days and term sheep fetuses (0.41 to 0.99 term), plasma cortisol concentration was

significantly lower in 90-120 day fetuses (0.61-0.82 term) than in the younger or older ones.

These findings suggest that the fetal adrenal becomes relatively quiescent and partially loses

its sensitivity to ACTH after an early period of activity from 40 to 90 days gestation (0.27

to 0.61 term). This quiescent period is accompanied by a decrease in 17-hydroxylase and SCC

mRNAs in the inner zone of the cortex (MacIsaac et al., 1989). Studies show that the plasma

basal level of cortisol, accompanied by ACTH, starts to increase again 15-20 days before

parturition (Hennessy, et al., 1982; Maclsaac et al., 1985; Magyar et al., 1980 and Norman et

al., 1985). Some investigators divided the pattern of observed fetal plasma cortisol changes

from 120 days (0.82 term) to term into four phases: phase I is the initial phase of low plasma

cortisol concentration between 120-135 days (0.82-0.92 term) or so of gestation; phase 2 is 5-









14
13 days before delivery, which presents a slow initial increment of cortisol; phase 3 is a more

rapid increase that occurs between 5 days before delivery and in the last 12-24 hour of fetal

life; phase 4 is a very variable, final, stress-induced increment immediately before delivery

(Nathanielsz, 1976).

The dependence of the growth of fetal adrenal on pituitary factor, most likely ACTH,

was indicated by several studies. Hypoplasia of the fetal adrenal cortex (mainly in the zona

fasciculate) and a fall in fetal plasma corticosteroid concentration were observed after

electrocoagulation of the fetal sheep pituitary at 93 days gestation (0.63 term) (Liggins et al.,

1966; 1968; Nathanielsz et al., 1972). Infusion of ACTH into the fetus reversed the

deficiencies caused by hypophysectomy. The feedback regulation of cortisol on ACTH has

also been shown by infusion of cortisol into fetuses, which causes inhibition of ACTH

response to a variety of stimuli (Norman and Challis, 1987; Rose et al., 1985; Wood and

Rudolph, 1983b). Bilateral adrenalectomy and therefore removal of cortisol feedback leads

to an increase in fetal ACTH values after 125 days (0.85 term) (Wintour et al., 1980).

In terms of initiation of parturition, the fetal hypothalamo-hypophysial-adrenal axis

plays a major role in sheep. Experiments that interrupted the normal connection between the

fetal hypothalamus and pituitary by inserting a silicone plate between them, or by surgical

ablation of the fetal pituitary or adrenal, resulted in the prolongation of pregnancy with

failure of parturition to commence. Infusing ACTH or cortisol into the fetuses induced

premature delivery (Challis and Olson, 1988; Liggins, 1969; Liggins et al., 1973; MacDonald

and Porter, 1983; Thorburn and Charllis, 1979).

When the fetal hypothalamus-pituitary-adrenal axis response to stress are assessed

from the increase of plasma ACTH and cortisol concentrations, the other sources of the

hormone should be considered, including the placenta and the transplacental exchange. It is

known that ACTH can be produced by the placenta but the physiological role of placental

ACTH is uncertain. Although no change has been found in placental ACTH content as a








15

function of gestational age, placental ACTH can be stimulated under certain stress situations.

Jones et al. (1988) have suggested that CRF may be released from the ovine placenta in

response to the reduction of placental perfusion that results from /-agonist administration.

Evidence from in vitro studies also indicate that CRF and ACTH are localized in the placenta,

endometrium, and amnion (Challis JRG and Brooks AN, 1989). Carr et al. (1981) observed

that the concentration of ACTH in maternal plasma rose progressively during gestation,

independently of maternal glucocorticoids. This increase in ACTH might be from the

placenta, which lacks an effect on maternal adrenal.

The placenta is impermeable to peptides, including ACTH, AVP, renin, angiotensin

I and II, but is permeable to steroids, including cortisol. In another study, a five-hour

intravenous infusion of cortisol into pregnant ewes raised the maternal plasma cortisol from

12.2+4.2 to 44.0+7.0 ng/ml and the fetal plasma values increased from 3.9+1.5 to 12.0+4.5

ng/ml (Wood and Rudolph, 1984), suggesting a significant transplacental movement of cortisol

from mother to the fetus. A report from Dixon et al. (1970) also supports this conclusion.

However, these ability of cortisol to move transplacentally seems limited, because of the

presence of the diffusion resistance of the placenta to cortisol (Bietins et al., 1970; Liggins et

al., 1973). This diffusion resistance establishes a constant plasma cortisol concentration

gradient between ewes and their fetuses. Liggins et al. (1973) reported the resting gradient

is about 5 fold greater concentration in ewe plasma than in fetal plasma (14 and 3 ng/ml in

the ewe and fetus, respectively) by using isotope dilution techniques. Bietins et al. (1970)

calculated the transplacental movement of cortisol from fetus to ewe and showed that the total

contribution to the mothers by fetal plasma levels of cortisol is small. Since there is no

evidence to show that the cortisol transplacental movement is unidirectional, we assume that

cortisol can diffuse to the mother if the fetal cortisol concentration rises to a certain high level

above the maternal concentration. These data suggest that the fetal plasma cortisol









16
concentration is determined by both fetal adrenal production and the transplacental diffusion

movement, especially when the mother is under stress.



2.5.2 AVP

Vasopressin is synthesized in the supraoptic and paraventricular nuclei (SON and

PVN) of the hypothalamus via a precursor protein, called propressophysin, which is packaged

into neurosecretory vesicles within the neuronal perikaryon and enzymatically processed into

vasopressin, neurophysin, and a C-terminal polypeptide during axonal transport. The early

presence of vasopressin-generating cells can therefore be demonstrated by the precursor and

the other two peptides in addition to vasopressin. The axonal terminals spread along the

supraoptic-hypophyseal tract to the posterior pituitary, which is a storage terminal of the

hormone. Axons containing neurohypophyseal hormones also project to the median eminence,

where hormones can be secreted into portal vessels, and to other areas of the brain and spinal

cord.

Both SON and PVN are formed between 12-14 days gestational (0.57-0.67 term, term

= 21) in rats (Altman and Bayer, 1978a; Ifft et al., 1972) and mice (Shimada and Nakamura,

1973; Karim and Sloper, 1980). The neurons in these two species do not reach their mature

size until the fourth week postnatally (Dellmann et al., 1981; Khachaturian and Sladekm

1980). A quantitative study on the maturation of the rat SON was tested by vasopressin-

neurophysin antibodies (Khachaturian and Sladek 1980), showing that the neurophysin

positive cells were minimal initially, then gradually increased to 27% at 17 days gestation (0.81

term) and reached 46% and 65% at 20 and 22 days gestation (0.95-0.99 term). The cell sizes

also gradually enlarged to more adult-appearing cell bodies at end of gestation.

The ovine hypothalamo-neurohypophysial system is developed anatomically by 37-47

days gestation (0.25-0.32 term) (Diepen, 1941), and AVP is detected in the fetal lamb

neurohypophysis at 55 days (0.37 term) (Rurak, 1979) and in plasma at 59 days (0.40 term)








17

(Drummond et al. 1980). Although AVP concentration in fetal sheep pituitaries was found

to double between 90-143 days gestation (0.61-0.97 term) (Alexander et al., 1973b), the

resting plasma AVP concentration is normally very low, close to or somewhat lower than adult

levels (Alexander, 1974). It is unlikely that AVP is involved in normal circulatory regulation

with such a low plasma concentration. Early data suggest that the clearance rate of AVP is

higher in fetal lambs and rhesus monkeys than in adults (Allexander et al., 1976; Jones and

Rurak, 1976; Robillard and Weitzman, 1980). The placenta was originally considered to be

a major site of metabolism (Chard et al., 1973). However, when AVP metabolic clearance

rates and production rates were measured in chronically catheterized fetal, newborn, and adult

sheep by constant infusion, there was no significant difference among these three groups

(Stegner et al., 1984). Wiriyathian et al. (1983) and DeVane et al. (1982) giving evidence that

AVP was not cleared by the placenta.

It is known that in adult animals, AVP is important as an antidiuretic hormone, a

potent vasoconstrictor, and an ACTH-releasing factor. AVP release is stimulated by rising

blood osmolality or angiotensin II, by hypovolemia, and by a decrease in wall tension of the

left atrium. The antidiuretic, vasoconstrictive and central ACTH-releasing effects of AVP

have received much attention in developmental studies. It has been demonstrated in the past

few years that AVP's antidiuretic and vasoconstrictor effects were mainly important in

extreme situations during fetal life, such as during hemorrhage and hypoxia.

Hemorrhage is one of the most potent stimuli for AVP secretion. AVP can be

stimulated by massive hemorrhage as early as 59 days gestation (0.40 term) (Drummond et al.,

1980). A 10% reduction in blood volume in chronically catheterized fetal sheep aged between

96-124 days gestation (0.65-0.84 term) resulted in a rise in plasma AVP levels from 1.6+0.23

to 9.6+4.0 pg/ml, and a 20% loss produced levels of up to 31.7+14.7 pg/ml (Drummond et al.,

1980). Alexander (1971; 1974b) reported that the maximum AVP responses in fetal sheep can

be stimulated by less than 40% of hemorrhage. Hypotension caused by nitroprusside infusion








18

without changes of blood volume also stimulates AVP secretion (Zubrow et al., 1988).

Infusing the same amount of AVP as is produced by mild hemorrhage increased fetal blood

pressure, decreased fetal heart rate, and redistributed blood flow in chronically catheterized

fetal sheep (Iwamoto et al. 1979b). Kelly et al (1983) pretreated with vasopressin antagonist

A-(CH2)5Tyr(Me)AVP, found that 20% hemorrhage caused a significantly greater fall in BP

than fetuses without an AVP antagonist. But AVP antagonist pretreatment did not change the

basal fetal HR or MAP (Kelly et al., 1983). These results suggest that AVP is important for

fetuses in the regulation of blood pressure following hemorrhage in fetuses. However, the

unchanged basal HR and arterial pressure during AVP antagonist infusion suggest that AVP

may not have much involvement in daily regulation of fetal circulation.


2.5.3 Renin

Renin, an enzyme that can convert angiotensinogen to angiotensin I, is also detected

very early in fetal life. Renin has been extracted from the kidney of a 40-day gestational

fetal lamb (0.27 term) (Wintour et al., 1977) and 22 days (0.19 term) from a pig fetus (Kaplan

and Fridman, 1942). The lamb fetal plasma renin concentration was lower than maternal

levels from 90 to 104 days gestation (0.61-0.70 term) (Carver and Mott, 1975), and

progressively increased toward term. In the last quarter of gestation, the plasma renin

concentration reached a level that is significantly higher in the fetal lambs than in their

mothers (Broughton Pipkin and O'Brien, 1978; Broughton Pipkin et al., 1974a). Plasma renin

activity (PRA), which is determined as the initial production rate of angiotensin I from

angiotensinogen during incubation of plasma in vitro, was found consistently higher in

chronically catheterized fetal lambs than in the ewes (Broughton-Pipkin and O'Brien, 1978;

Broughton Pipkin et al., 1974a; Fleischman et al., 1975; Smith et al., 1974).

The high activity of the renin-angiotensin system in unstressed fetuses suggests that

this system is important in regulating fetal circulation under physiological conditions.

Infusion of Saralasin, a specific angiotensin receptor-blocking agent, into fetal lambs, resulted









19

in a fall in blood pressure, and the changes of blood pressure were linearly correlated to the

initial angiotensin II level (Broughton-Pipkin and O'Brien, 1978). A rapid fall in fetal blood

pressure was also observed when converting enzyme blocker captopril was administered to the

mother. This fall in fetal blood pressure lasted for up to 3 days (Broughton-Pipkin et al.,

1982). Hyman et al. (1975) found that chronic unilateral renal artery constriction in fetal

lambs was associated with a marked elevation in both fetal plasma renin activity and arterial

blood pressure. In addition to the effect on blood pressure, infusion of angiotensin II

inhibitors to unstressed fetuses also decreased placenta flow (Iwamoto and Rudolph, 1982).

Due to the immaturity of the fetal cardiovascular regulation system, it is possible that renin-

angiotensin system plays a more important role in regulating circulation physiologically than

it does in adults.

Hemorrhage in both anesthetized and unanesthetized fetal lambs, even as little as 3%

of the calculated fetal blood volume, has been shown to stimulate PRA in fetuses from 110

days gestation to term (0.75 term) significantly (Broughton-Pipkin et al., 1974a; 1974b; Smith

et al., 1974). In another study, a 12 ml hemorrhage stimulated PRA in 3 out of 5 anesthetized

fetal piglets on 85 days of gestation (0.75 term). By 111 days (0.97 term), PRA increased in

all eight piglets studied (Broughton Pipkin et al., 1981). Iwamoto and Rudolph (1981b)

reported that previous administration of saralasin blocked the fetal lamb's ability to restore

its blood pressure and heart rate following hemorrhage. These studies suggest that

hemorrhage is as least as strong a stimulus for fetal renin secretion as it is in adults.





2.6 Summary


1. The fetal cardiovascular system differs anatomically and functionally from the adult

to meet its growth and developmental requirements in intrauterine life.









20

2. The autonomic system starts to function at least before the end of the first half

gestation, and the parasympathetic nerve system is more developed than the

sympathetic nerve system at term.

3. Fetal baro and chemoreceptors undergo a development sequence and are active at the

second half of gestation.

4. The ventricular receptors function in late gestational fetal sheep.

5. The fetal hypothalamus-pituitary-adrenal axis is active in the first trimester of fetal

life. The adrenal becomes refractory to ACTH stimulation in the middle third of

gestation, which is accompanied by a decrease in 17-hydroxylase and SCC mRNAs

in the inner zone of the cortex. After this period, the adrenal becomes increasingly

responsive to ACTH, and both ACTH and cortisol concentration in plasma increase

shortly before term and reach their peaks at parturition.

6. The hypothalamo-neurohypophysial system appears to be anatomically developed by

the first quarter of gestation. AVP acts as an early ACTH-releasing factor before and

after the CRF is functional. AVP is important for responses to hypotension, but not

for basal circulatory regulation in fetuses.

7. The renin-angiotensin system is important in regulating fetal circulation under both

physiological and pathological conditions.

8. The placenta can produce ACTH and CRF and is permeable to corticosteroids but not

to ACTH, AVP, and renin.

9. The fetuses respond to hemorrhage with an increase in ACTH, cortisol, AVP, and

PRA. In adult animals, the afferent pathway for the hormone responses to mild- or

non-hypotensive hemorrhage is mediated by atrial type-B receptors which is linked

to vagus afferent fibers. The afferent pathway for fetal hormone secretion to slow

hemorrhage remains unknown.














CHAPTER 3
GENERAL METHODOLOGY



3.1 Surgical Procedures of Catheterization and Denervation


All surgeries were performed at least four days before the experiments. For 24 hours

before surgery ewes were not fed but were allowed free access to water. During surgery ewes

were anesthetized with 1.0-2.5% halothane in oxygen. Using strictly aseptic techniques, we

exposed the uterus with a midline incision. After incising the uterus, a fetal hind limb was

delivered and a polyvinylchloride catheter (0.030 in. i.d., 0.05 in 0.d.) was inserted into the

tibial artery. A larger catheter (0.040 in. id.,0.070 in. o.d.) was inserted into the saphenous

vein. After catheterization of these two vessels, the fetal skin was sutured and the hind limb

was returned to the amniotic cavity. Then the second fetal hind limb was delivered and the

procedure was repeated. A polyvinylchloride catheter (0.050 in.i.d., 0.090 in.o.d.), with side-

holes cut into the tip, was sutured to the skin before returning the second hind limb to the

amniotic cavity. The amniotic catheter was used for measuring amniotic fluid pressure during

experiments. The uterus was closed.

The surgical procedure for vagotomy and carotid sinus denervation was similar to that

described by Itskovitz and Rudolph (1982). Briefly, after making another incision in the

uterus, the fetus' head was delivered. A midline incision was performed in the fetus' neck

near the angle of the jaw, the occipital-carotid arterial junctions were carefully exposed. The

carotid sinus nerves, the IXth cranial nerves and the vagosympathetic trunks were in turn

identified and sectioned as required. The superior thyroid arteries were ligated and cut, as

were all branches of the common carotid arteries between the superior thyroid and lingual









22
arteries, to ensure sections of the common carotid nerves. The walls of the common carotid

arteries in this area, as well as the lingual arteries and common carotid arteries extending 0.5-

1 cm rostral from the lingual-carotid arterial junction, were stripped off all visible nerve

fibers. The fetal skin was closed.

After closing the incision in the uterus, 500 mg ampicillin (Polyflex, Veterinary

Products, Bristol Laboratories, Syracuse, New York) was injected into the amniotic fluid.

Catheters were filled with heparin (1,000 units/ml; Wilkins-Sinn, Cherry Hill, New Jersey),

plugged, and threaded out through a stab wound in the flank where they were protected by

a cloth pouch sutured to the skin. In the case of twins, both fetuses were catheterized.

After the completion of the abdominal surgery, polyvinylchloride catheters (0.050

in.i.d., 0.090 in.o.d.) were inserted into the maternal femoral artery and vein at the level of

the femoral triangle, and the tips were advanced to the abdominal aorta and inferior vena

cava, respectively. The catheters also were filled with heparin, plugged, and routed

subcutaneously to the flank, where they were protected by the same pocket as the fetal

catheters. 500 mg ampicillin was injected intramuscularly into the mother at the time of

surgery as well as once per day for five days after surgery. Fetuses were also treated with 500

mg ampicillin via the amniotic fluid catheter once every day for five days and after each

experiment. All catheters were flushed and reheparinized at least once every 3 days.



3.2 General Preparation for Experiments and Hemodynamic Data Collection


The pregnant ewes used in all the experiments were mixed Western and Florida Native

breeds. In the 2 series of hemorrhage studies, the animals participated in only one

experiment. For those animals participating in more than one experiment in the other two

studies (HC1 infusion and hypercapnia), at least 48 hours were allowed between experiments,

and the order of the experiments was randomized. No fetus was subjected to the same

protocol more than once. All experiments were started between 0900 and 1100 hours to









23
prevent possible rhythm variations between animals in resting hormone concentrations or in

the magnitude of stimulated responses.

On the morning of the experiment, the ewe to be studied was transported from the

Health Center Animal Resources Department to the laboratory. At least one hour was allowed

before the experiment started to let the ewe acclimate to the laboratory environment. During

this time, one fetal femoral artery, vein, and the amniotic fluid catheter were connected to

the transducers (Statham P23 Db or P23Id transducers, Statham Instruments, Oxnard,

California) for measurement of intravascular and amniotic fluid pressures. Pressures were

measured continuously using a Beckman R611 (Beckman Instruments, Schiller Park, Illinois)

or Grass Model 7 (Grass Instruments, Quincy, Massachusetts) direct-writing recorder. Fetal

heart rate was measured using an appropriate Beckman or Grass cardiotachometer triggered

from the arterial pressure signal. Mean femoral arterial pressure was recorded as the damped

mean of the phasic signal. Analog voltage outputs of the amplifiers were sampled by a

Keithley DAS analog-to-digital converter (Keithley System 500, Keithley Data Acquisition

Control, Cleveland, Ohio) and IBM AT microcomputer at a rate of 0.5 Hz. The variables were

sampled and analog-to-digital conversions performed at 2-second intervals and stored on

floppy disk. Intravascular pressures were corrected by subtraction of amniotic fluid pressure

off-line. All fetal intravascular pressures were calculated using amniotic fluid pressure

as zero reference.




3.3 Handling and Analysis of Blood Samples


All blood samples were placed in chilled plastic centrifuge tubes containing 0.05 ml

of 0.3 M Na4EDTA (Sigma Chemical Co, St. Louis, Missouri) per ml blood. All tubes were

kept on ice until the end of the experiment, when they were centrifuged at 3,000 g for 20









24
minutes in a refrigerated (4*C) centrifuge. The plasma was stored at -20*C until hormones

were assayed.

PRA was measured using a modification of the method by Haber et al., (1984). The

kit was from Clinical Assays. For this assay, angiotensin I was generated in buffered (pH 5.7)

plasma in vitro for one hour at 37*C. At the end of the incubation period the angiotensin I

concentration was measured by RIA. This method has been described in detail elsewhere

(Wood et al., 1989b).

Plasma cortisol concentrations were measured by RIA using antiserum No. 1460 (lot

R2) from Radioassay Systems Laboratories (Carson, California) and [3H]-[1,2,6,7]-cortisol

from New England Nuclear Corp. (Boston, Massachusetts) or Amersham Co. (Arlington

Heights, Illinois). Plasma was deproteinized with 50-100 vol of ethanol before assay.

Plasma ACTH was measured by radioimmunoassay (RIA) in either unextracted plasma

(in the first two series of studies) or extracted plasma (in the last two series of studies, using

powdered glass, eluted with HCI and acetone). The former used anti-ACTH antiserum and

hACTH-(1-39) supplied by the National Hormone and Pituitary Program. The 12II-ACTH

used in this assay was made by iodination of the hACTH-(1-39) standard by the chloramine-

T method (Rees et al, 1971). The latter used both anti-ACTH raised in this laboratory and

1-39 hACTH standard, which was supplied by the National Hormone and Pituitary Program

(NIDDK, University of Maryland School of Medicine). This antiserum cross-reacts 100%

with (1-24)-ACTH, 11-24 ACTH, and 1-39hACTH, but does not bind oCRF or methionine

enkephalin.

Plasma vasopressin (AVP) concentrations were measured by RIA using anti-

vasopressin antiserum purchased from Amersham, 1251-vasopressin from New England

Nuclear, and synthetic arginine vasopressin from Sigma.

In the first two series of experiments, plasma assayed for AVP was an extract of

acetone as described by Cowley and co-workers (Cowley et al., 1981). In the last two series








25

studies, the plasma was extracted using bentonite, eluted with a 1:1 mixture of 1 M HCI and

acetone. Extracts were dried using the Speedvac concentrator. The antiserum bound arginine

vasopressin 0.5% relative to 1-39 hACTH.

Fetal blood gases were measured using a Radiometer PHM73 blood gas analyzer and

BMS3Mk2 blood microsystem (Radiometer, Copenhagen, Denmark). The blood gas analyzer

was calibrated as recommended by the manufacturer using two reference gas mixtures, one

containing 8% CO2, 21% 02, and the balance N2, and the other containing 5% CO2 and the

balance N2 (Radiometer). Plasma protein concentration was measured by refractometry in

study 3 (Chapter 6). Plasma sodium and potassium concentrations were measured using a

NOVA 1 sodium/potassium analyzer (Nova Biomedical, Waltham MA).




3.4 Calculations and Statistical Analyses


The data were analyzed by two-way analysis of variance (ANOVA) corrected for

repeated measures in one dimension, time (Winer, 1971). Changes were assessed within

groups in some studies using one-way ANOVA for repeated measures. A posteriori

comparison of individual means was performed using Duncan's multiple range test (Winer,

1971). Initial fetal and maternal variables were either tested by Student's t-test or by one-way

ANOVA. In Experiment I (Chapter 4), correlations among variables were tested using

standard correlations analysis. A significant level of p
hypothesis in all tests.














CHAPTER 4
THE ROLE OF VAGOSYMPATHETIC AFFERENT FIBERS
IN THE CONTROL OF ACTH, VASOPRESSIN, AND RENIN RESPONSES
TO HEMORRHAGE IN FETAL SHEEP



4.1 Introduction1


In adult animals, hemorrhage stimulates secretion of ACTH (Redgate, 1968),

corticosteroids (Farrell et al., 1956; Gann, 1969), renin (Sapirstein et al., 1941), and AVP

(Ginsburg and Heller, 1953) mediated via cardiopulmonary (Cryer and Gann, 1973; Brennan

et al., 1971) and arterial mechanoreceptors (Redgate, 1968; Gann et al., 1964; Share and Levy,

1962). Renin is controlled by mechanoreceptors in those areas and by the intrarenal

baroreceptor and the macula densa (Davis and Freeman, 1976). All three of these hormone

systems also respond to hypoxia and/or hypercapnia, responses that may be at least partially

mediated by the peripheral chemoreceptors (Raff et al., 1983a; 1983b; 1984). Results of

experiments in adult animals suggest that atrial or other cardiac receptors with

vagosympathetic afferent fibers are of primary importance in the control of ACTH, AVP, and

renin responses to small or moderate hemorrhage (Cryer and Gann, 1973; Wang et al., 1983;

Quail et al., 1987).

Fetal sheep respond to hemorrhage with increases in plasma ACTH (Alexander et al.,

1973; 1974; Rose et al., 1978), AVP (Robillard et al., 1979; Rurak, 1979; Ross et al., 1986;

Brace and Cheung, 1986), and renin (Robillard et al., 1982) concentrations. Presumably,

receptor populations in the fetus that control the secretion of these hormones are similar to



This work was in part supported by the National Institute of Health grants HD-20098
and HL-36289.








27
those in the adult. One might predict that, in the fetus as in the adult, the hormonal responses

to a nonhypotensive or mildly hypotensive hemorrhage would be attenuated by interruption

of the vagosympathetic afferent fibers from atrial or other cardiopulmonary receptors. This

study was designed to investigate this possibility.





4.2 Materials and Methods


Ten chronically catheterized pregnant ewes and their fetuses were studied. One sheep

carried twins; the others carried single fetuses. On the day of study, the fetuses were between

128 and 133 days' gestation.

Six fetal sheep (including the twins) were prepared with vascular catheters only. Five

fetal sheep were subjected to bilateral section of the cervical vagosympathetic trunks.

Eleven ml blood samples were withdrawn from a fetal arterial catheter at 10-minute

intervals for 120 minutes. At the beginning (0 minute) and end (120 minutes) of each

experiment, 5 ml blood samples were drawn from the maternal arterial catheter. This

hemorrhage paradigm is similar to that used by Brace and Cheung. Fetal blood gases were

measured in all experiments.




4.3 Results


Initial fetal pHa values were 7.39+.02 and 7.39+.01, Pao2 22.3+1.4 and 24.0+1.5 mm

Hg, and Paco2 43.31.3 and 42.0+1.1 mm Hg in the intact and vagotomized groups,

respectively. None of these values were significantly different between groups.

Overall, hemorrhage decreased MAP (Figure 4-1, left). There was no significant

difference in the MAP response to hemorrhage in the two groups. HR (Figure 4-1, right) was

not significantly altered by the hemorrhage. Vagotomy did alter the CVP response to the























560 iAFcr


6 20 40 60 80 100 120 0 20 40 60 80 100 120
Time (min from onset of hemorrhage)
























Figure 4-1. MAP and HR during hemorrhage in fetal sheep that were intact or
subjected to bilateral section of cervical vagosympathetic trunks.










hemorrhage (Figure 4-2).

CVP significantly decreased in intact fetuses during hemorrhage. Hemorrhage

significantly decreased fetal pHa (Figure 4-3, bottom panel). The pH, decreased more rapidly

in the vagotomized group. In similar fashion, Paco2 (Figure 4-3, middle) increased

significantly, but more rapidly in the vagotomized group. Pao2 (Figure 4-3, top) decreased

similarly in both groups equally.

Hematocrit (Figure 4-4) was significantly lower in the vagotomized fetal sheep at the

beginning of hemorrhage (42+2% vs. 37+2% packed cell volume [PCV] in intact vs.

vagotomized fetuses). The magnitude of the decrease in hematocrit during the course of the

hemorrhage was greater in the intact fetuses (mean change, 6 % PCV) than in the vagotomized

fetuses (mean change, 4 % PCV). Because the fetal sheep does not have a contractile spleen,

acute changes in Hct have been used to calculate changes in blood volume (Brace, 1983). In

the present experiments, the decrease in Hct during the hemorrhage may have been caused

by transplacental and transcapillary refilling of the fetal vascular space. Red cells were

diluted 14% in the intact group and 12% in the vagotomized fetuses, suggesting a more

effective defense of blood volume in the intact group.

Initial fetal plasma ACTH concentrations were 41+8 and 60+7 pg/ml and initial fetal

plasma cortisol concentrations were 9.1+3.3 and 7.8+2.2 ng/ml in the intact and vagotomized

fetuses, respectively (Figure 4-5). Initial values of ACTH and cortisol were not different in

the two groups. Hemorrhage stimulated similar ACTH and cortisol responses in both groups.

Initial fetal plasma AVP concentrations were 2.70.3 and 2.60.3 pg/ml and initial

fetal PRAs were 5.0+1.5 and 4.7+2.2 ng angiotensin I/ml/hr (Figure 4-6). These initial values

of vasopressin concentration and PRA were not different in the two groups. Hemorrhage

stimulated similar AVP and similar PRA responses in the two group correlations of fetal

plasma ACTH and AVP concentrations and PRA to fetal MAP, CVP, and pHa were tested

(Table 4-1). Logarithms of fetal hormone concentrations and activities were used in the


































5-
4-
3-
2-
I

0
-1

-2-
-3


I I I 20I I I I I 100I I 120
0 20 40 60 80 100 120


Time (min from onset of hemorrhage)








Figure 4-2. CVP during hemorrhage in fetal sheep that were intact or subjected to
bilateral section of cervical vagosympathetic trunks.


VAGOTOMY























































7.40


7.35-


0 20 40 60 80 100 120
Time (min from onset of hemorrhage)










Figure 4-3. Pao2, PaCO2, and pHa during hemorrhage in fetal sheep that were intact
(open circles) or subjected to bilateral section of cervical vagosympathetic trunks
(filled circles).


I.


T.o
Il


1.


7n














50





0\0


6 40
0

E





30


100 120


Time (min from onset of hemorrhage)















Figure 4-4. Hct during hemorrhage in fetal sheep that were intact (open circles) or
subjected to bilateral section of cervical vagosympathetic trunks (filled circles).


0 20 40 60 80


I I I I I I I I I I I I













1000

800

600

400

200


--0 INTACT
*---@ VAGOTOMIZED






iA7Li


15

101. 1I041


0 20 40 60 80 100
Time (min from onset of hemorrhage)


Figure 4-5. Plasma ACTH and cortisol concentrations in fetal sheep that were intact
(open circles) or subjected to bilateral section of cervical vagosympathetic trunks
(filled circles).

















o-o INTACT
* ---.* VAGOTOMIZED


0 20 40 60


80 100


Time (minutes after onset of hemorrhage)












Figure 4-6. Plasma AVP concentration and PRA in fetal sheep that were intact (open
circles) or subjected to bilateral section of cervical vagosympathetic trunks (filled
circles).


10 k


10 -

























z 0 a
zo^

i
5J


I I I i
,,,o
0 o 0 0
LA %O (
os tm t-J qw
* 0


Ipo
w-J 0. so


I
P 0 0 P
,m >" 00


0oo '








o >
Co




o2 W




a D



<





aCD
? r
0. u




















-. a




.



0


I I I I
A o
w wid
** *


o o


I
- C 0C
WD t4 -


I I I I I I





I I I
0 P 00o0




I I
pp po0
^~ ^ L >


J0o




I,
5.,

a as


I I
0 0
- OS




"I
SM '


I

SO


I I

wa
-*-


I
1
*


0< 0-


I I






PP
*
I I






pop
o o o








o a La


& S.DL

*


pop
u b s


I I I I

* "



II
0oo


>E
C




,-0







to



0

2. i
M "








*o'
as=
0













CD












o-
0
-t

0'
5n









36
correlations because logarithmic transformation linearized the relations (for example, see

Wood et al., 1982). In one intact fetus, plasma ACTH was significantly related to MAP and

CVP, and PRA was related to MAP. In one other intact fetus, vasopressin was significantly

related to MAP. In one vagotomized fetus, ACTH and vasopressin were significantly related

to both MAP and CVP, and in another vagotomized fetus, ACTH and PRA were significantly

correlated to MAP (Table 4-1). Interestingly, in some of the experiments, ACTH,

vasopressin, or PRA were correlated to increases in central venous pressure (e.g., fetuses

Nos.0071, Y70, and Y64). In contrast to the inconsistent or seemingly inappropriate

(increasing hormone concentrations with increasing intravascular pressures) correlations

between hormone concentrations and arterial or venous pressures, the hormone responses were

much better correlated to the changes in pHa (Table 4-1). In 12 of 14 possible comparisons

in intact fetuses and in 13 of 18 possible comparisons in vagotomized fetuses, correlations

between hormone response and change in pHa were significant. The mean values of fetal

plasma ACTH, AVP, and PRA are graphically summarized in Figure 4-7. The relation

between ACTH or vasopressin concentration or PRA and pHa appeared to differ in the two

groups (Figure 4-7). However, there was more variability in the slopes within groups than

between groups. For this reason, there was no significant difference in the slopes between

the two groups. Shown in Figure 4-8 are the distributions of correlation coefficients relating

pHa, MAP, and CVP to the logarithm of plasma ACTH, AVP, and PRA.

Maternal plasma hormone concentrations were measured at the beginning of each

experiment. Maternal plasma ACTH concentrations were 10239 and 151+14 pg/ml, and

plasma cortisol concentrations were 12.3+4.8 ns 16.1+3.7 ng/ml in the groups with intact and

vagotomized fetuses, respectively, Maternal plasma AVP concentrations were 2.50.1 and

2.90.3 pg/ml, and PRAs were 2.9+0.7 and 2.4 +1.1 ng angiotensin I/ml/hr in the intact and

vagotomy groups, respectively. There were no statistically significant differences in maternal

plasma hormone levels in the two groups (tested by t test for independent groups).









1000
ACTH kl/mI)

I*o \O

100




10 1
10
AVP (pq/mi)
*
o






2
2 -------I ------
100
PRA (ng Al/mI/hr)






1----------.-- ----.9
10





7.300 7.350 7.400
Arterial pH

Figure 4-7. Relations among mean values of pHa and mean values of fetal plasma
ACTH and AVP concentrations and PRA.






















pH MAP CVP



1.-






0






Figure 4-8. Distribution of r values of fetal pH-, MAP, and CVP vs Lg ACTH (open


inside box are the range of distribution and median values.







Figure 4-8. Distribution of r values of fetal pHa, MAP, and CVP vs Lg ACTH (open
circles), AVP (filled circles), or PRA (open triangle). Symboles at the extremes and
inside box are the range of distribution and median values.













4.4 Discussion


The results of this study suggest that the control of hormonal responses to hemorrhage

in fetal sheep is different from that in the adult. Section of the cervical vagosympathetic

trunks, which interrupt afferent fibers from atrial and ventricular mechanoreceptors, did not

alter the magnitudes of the ACTH, AVP or PRA responses to hemorrhage. In adult

anesthetized dogs, acute vagotomy attenuated the vasopressin (Share, 1967) and adrenal

corticosteroid (Gann and Cryer, 1973) responses to nonhypotensive or mildly hypotensive

hemorrhage. ACTH and AVP in the adults are inhibited by afferent impulses from the atrial

mechanoreceptors with afferent fibers in the vagus nerves (Brennan et al., 1971; Baertschi et

al., 1976) and from carotid arterial mechanoreceptors (Gann et al., 1964; Share and Levy,

1962; Wood and Rudolph, 1983; Wood et al., 1985). Renin is controlled by these receptors and

by the intrarenal baroreceptor (reviewed by Davis and Freeman, 1976).

The cardiopulmonary receptors are quite important in the mediation of the ACTH and

AVP responses to slow or otherwise mild hemorrhage in adult animals. Renin responses to

hemorrhage are less dependent on the activity of the cardiopulmonary receptors. Adrenal

corticosteroid responses to 5 ml/kg hemorrhage in anesthetized dogs are attenuated by

simultaneous inflation of a balloon in the right atrium (Cryer and Gann, 1973). Blockade of

cardiopulmonary receptors with intrapericardial injections of procaine blocked the vasopressin

but not the renin response to hemorrhage of a volume of up to 35% of blood volume in

conscious rabbits (Quail et al, 1987). Similarly, cardiac denervation attenuated the vasopressin

response but not the renin response to 10, 20, or 30 ml/kg hemorrhage in conscious dogs

(Wang et al., 1983). While afferent information from arterial mechanoreceptors may

contribute to the ACTH, AVP, and renin responses to hypotensive hemorrhage, the

cardiopulmonary receptors appear to be most important. AVP and renin responses to









40
progressive hemorrhage to 35% of blood volume in conscious rabbits were not attenuated by

sinoaortic denervation (Quail et al., 1987). ACTH and corticosteroid responses to hemorrhage

in sinoaortic denervated lambs were not smaller than responses in intact lambs (Wood et al.,

1985). Chemoreceptors may influence the secretion of ACTH, AVP, and renin in adult

animals during hypoxia and/or hypercapnia (Raff et al., 1983a; 1983b; 1984), however,

chemoreceptor stimulation is unlikely to mediate the responses of these hormones to small or

moderate hemorrhage because carotid sinus denervation (which would eliminate afferent

fibers from carotid arterial chemoreceptors and baroreceptors) does not attenuate the ACTH,

AVP, or renin responses to hemorrhage in postnatal lambs (Wood et al., 1985).

Responses to hemorrhage in fetal animals differ from responses in adult animals in

that moderate hemorrhage produces acidemia in addition to alterations in arterial and/or

venous pressures (Toubas et al., 1981). The fetal acidemia is a respiratory acidemia caused

by a primary increase in fetal Paco0. This condition, analogous to hypoventilation by the

adult, is most probably caused by decreased perfusion of the umbilical-placental circulation.

Microsphere studies have demonstrated that hemorrhage reduces umbilical-placental perfusion

(Itskovitz et al., 1982).

Fetal animals differ from adult animals in that MAP is regulated at a lower level.

Chronic sinoaortic denervation in fetal sheep increased the variability of blood pressure and

heart rate, indicating that fetal arterial baroreceptors are active at the regulated level of

arterial pressure (Itskovitz et al., 1983). However, the firing rates of the fetal arterial

baroreceptors may be at or near the threshold for activation. If this were true, fetal arterial

baroreceptors would be less responsive to decreases than increases in fetal arterial pressure.

Fetal arterial Paos is also regulated at a somewhat lower level than that in the adult. It is

known that fetal peripheral chemoreceptors are active at this low Pa02 and that the sensitivity

of these receptors to changes as arterial oxygen tension resets after birth to adult levels









41
(Blanco et al., 1984). Little is known about the characteristics of atrial receptors in fetal

animals.

Several studies have demonstrated that fetal sheep respond to hemorrhage with

increases in plasma ACTH and cortisol (Rose et al., 1978), AVP (Rurak, 1979), PRA,

angiotensin II, and aldosterone (Robillard et al., 1982). It is known that the secretion of these

hormones is influenced by induced changes in arterial or venous pressure. For example,

ACTH and AVP are stimulated by vena caval obstruction in fetal sheep, and the magnitude

of the responses is related to the severity of the obstruction (Wood et al., 1985), leading to the

speculation that cardiovascular mechanoreceptors influence the secretion of these hormones.

During hypovolemia in fetal sheep, plasma AVP and PRA responses correlated better to the

volume of the hemorrhage than to the induced changes in MAP, leading to the speculation

that volume receptors are important in the control of these two hormones (Robillard et al.,

1979). A significant correlation between plasma AVP and pHa during hypovolemia was noted

in one study (Rurak, 1979). Other investigators have suggested that fetal hypoxia during

hypovolemia may contribute to the vasopressin response (Ross et al., 1986). While several

groups of investigators have studied hormonal responses to hemorrhage in the fetus, the

populations of cardiovascular receptors responsible for mediation of these hormonal responses

have received little attention.

As in studies by other investigators, the fetal hormonal responses to hemorrhage were

not well correlated to the induced changes in arterial pressure (Robillard et al., 1982; Brace

and Cheung, 1986). It is apparent from the results of the present study that fetal ACTH,

AVP, and renin responses to mildly hypotensive hemorrhage in fetal sheep are more closely

associated with the induced changes in arterial blood pH. The correlations with arterial and

central venous blood pressure reveal mostly either nonsignificant relations or relations that

are seemingly inappropriate for cardiovascular mechanoreceptor-hormone relations (increased

hormone concentration with increased blood pressure). In contrast, the correlation









42
coefficients calculated from relations between hormone concentrations and pHa are high in

most experiments. The association of hormone response to changes in pHa would therefore

suggest the possibility that the responses of these hormones to mild hemorrhage in the fetal

sheep are mediated by the peripheral chemoreceptors rather than mechanoreceptors. It is also

possible however, that the hypercapnia produced by fetal hemorrhage stimulated the hormonal

responses via medullary chemoreceptors. Raff and co-workers (Raff et al., 1984)

demonstrated that deafferentation of the carotid chemoreceptors attenuated that ACTH and

corticosteroid response to isocapnic hypoxia but had no effect on the responses to hypercapnic

hypoxia in anesthetized adult dogs.

While vagotomy did not alter the final magnitudes of the fetal ACTH, AVP, or PRA

responses to hemorrhage (Figure 4-5 and 4-6), it did alter the Paco2 and pHa responses

(Figure 4-3). The greater increases in Paco2 and decreases in pHa in the vagotomized group

may reflect a greater decrease in fetal cardiac output and a concomitant greater decrease in

fetal umbilical-placental flow in the vagotomized fetuses.

Vagotomy also altered the CVP responses to the hemorrhage (Figure 4-2). Intact

fetuses defended CVP well during this hemorrhage while vagotomized fetuses responded to

the hemorrhage with a substantial decrease in CVP. The maintenance of CVP in the intact

fetuses may have been dependent on rapid restitution of blood volume. Because fetal sheep

do not release red blood cells into the circulation after acute sympathetic stimulation (Brace,

1983), changes in hematocrit can be used to follow acute changes in blood volume. Assuming

the fetal body weight was approximately 3.3 kg (calculated from the relation of gestational

age to fetal body weight) as described by Robillard and Weitzman (Robillard and Weitzman,

1980) and assuming that fetal blood volume is 110 ml/kg (Brace, 1983), we can estimate that

approximately 33 ml (28% of the volume removed between the 0- and 120-minute samples)

of fluid re-entered the fetal vascular space during the fetal hemorrhage. Using a similar

approach, we can calculate that only 20 ml (17% of the volume removed ) of fluid re-entered









43
the fetal vascular space in the vagotomized fetuses. This apparent difference in the rate of

volume restitution may have been the cause of the greater decrease in CVP in the vagotomized

fetuses. It is also possible that vascular compliance was lower in the vagotomized fetuses. In

any event, the reduced rate of volume restitution and/or decreased vascular compliance may

have contributed to an impairment of cardiac output and umbilical-placental blood flow,

which, in turn, produced the greater increase in Paco2 and decrease in pHa.

The difference in initial hematocrit in the intact and vagotomized groups may have

also reflected alteration of fetal fluid balance by vagotomy. It is tempting to propose that the

vagotomy produced an expanded plasma volume and, therefore, secondarily, a decrease in

hematocrit by dilution of the red blood cells. This remains to be tested.

In summary, we have found that vagotomy altered the blood gas and CVP responses

to slow hemorrhage in the sheep fetus but did not alter the magnitudes of the ACTH, AVP,

or renin responses. The results suggest that the control of these hormonal responses to

hemorrhage in the fetus is fundamentally different from that in the adult. The data suggest

the possibility that hormonal responses to hemorrhage in the fetus are stimulated by

chemoreceptor activity, secondary to the acidemia or hypercapnia of fetal hemorrhage.















CHAPTER 5
ACTH, AVP, AND RENIN RESPONSE TO INTRAVENOUS INFUSION
OF HYDROCHLORIC ACID ARE CHEMORECEPTORS RESPONSIBLE
FOR FETAL HORMONE SECRETION?



5.1 Introduction2


The secretion of ACTH, AVP, and renin are stimulated by hemorrhage, hypotension,

and hypoxia. In adult animals, the responses of these hormone systems to hemorrhage and

hypotension are mediated by arterial and atrial mechanoreceptors (Brennan et al., 1971; Bunag

et al., 1966; Cryer and Gann, 1973; Gann et al., 1964; Redgate, 1968; Share and Levy, 1962)

and, for renin, by the intrarenal baroreceptor and the macula densa (Davis and Freman, 1976).

Because resting mean arterial pressure is much lower in the fetus than in the adult, it is not

clear that similar mechanisms govern the ACTH, AVP, and PRA responses to hemorrhage and

hypotension in fetal sheep.

These endocrine systems were investigated in previous experiments to show their

responses to vena caval obstruction (Wood, 1989a, Wood et al, 1982) and to slow hemorrhage

(Chapter 4). Vena caval obstruction, a stimulus that produces a large and rapid decrease in

fetal arterial pressure increased fetal ACTH, AVP, and PRA to levels that were related to the

induced change in arterial blood pressure. Heart rate was decreased to levels significantly

related to the induced change in arterial blood pressure. Sino-aortic denervation attenuated

the ACTH, AVP, and heart rate responses, suggesting that the afferent pathways for these

reflexes were interrupted by removal of either the mechanoreceptor or the chemoreceptor



2 This work was in part supported by the National Heart, Lung, and Blood Institute
grant HL-36289.









45
afferent fibers from the carotid sinuses or aortic arch (Wood, 1989a). The attenuation of the

bradycardia normally observed during fetal hypotension suggested that these reflex responses

may be mediated by chemoreceptor afferents rather than the mechanoreceptor afferents.

Slow fetal hemorrhage, a stimulus that produces small and prolonged decreases in fetal

arterial pressure, increased fetal Paco2, and decreased fetal pHa; the changes in pHa and

Paco0 were probably secondary to reduced perfusion of the umbilical-placental circulation

(Faber et al., 1973). The fetal ACTH, AVP, and renin responses to the hemorrhage were

highly correlated to the change in pHa but were not correlated to the induced change in fetal

mean arterial or central venous pressure. We performed these experiments to test the

hypothesis that the acute acidemia, in the absence of hypovolemia, stimulates fetal ACTH,

AVP, and renin secretion.



5.2 Methods


Eight chronically catheterized pregnant ewes and their fetuses were studied. Two of

these ewes carried twins; the other six carried singleton fetuses. On the days that the fetuses

were studied, the ages ranged from 132 to 140 days gestation.

Hydrochloric acid (HC1) was infused into the fetal inferior vena cava at one of three

different rates (0.02, 0.10, or 0.50 meq/min) for 60 minutes. All infusions were performed

using a constant infusion pump (Sage Instruments model 341A, Cambridge, MA) at a volume

rate of 0.5 ml/min, with HCI concentrations in the infusate adjusted to 0.04, 0.5, and 1.0

meq/ml. We performed five infusions of 0.02 meq/min, six infusions of 0.10 meq/min, and

five infusions of 0.50/meq/min. Blood samples (5 ml/each) were drawn from a fetal femoral

arterial catheter at the beginning of the infusion and thereafter at 10-min intervals throughout

the infusion period. As each fetal blood sample was drawn, an additional blood sample (1 ml)

was drawn into a heparinized syringe for measurement of fetal hematocrit and blood gases.

Therefore a total of 42 ml was drawn from the fetal circulation during these experiments.












5.3 Results


Intravenous infusion of HCl at a rate of 0.5 meq/min significantly decreased fetal PHa

and increased Paco2 (Figure 5-1). Results of two-way ANOVAs are reported in Table 5-1.

Pa02 was significantly increased over all groups, but this increase was not significantly related

to the rate of infusion. Infusion of HCl at a rate of 0.02 meq/min did not significantly alter

fetal blood gases; however, infusions of 0.1 and 0.5 meq/min significantly decreased pHa and

increased Paco2 beginning 20 and 10 min after the onset of infusion, respectively.

Infusion of HCl did not significantly alter fetal arterial blood pressure at any of the

rates tested (Figure 5-2, Table 5-1). Fetal heart rate was increased during infusion of 0.5

meq/min but not during infusion of 0.02 or 0.10 meq/min. Fetal heart rate was significantly

increased over the initial level 54-61 min after the onset of infusion in the 0.5 meq/min

group. Initial values of fetal blood pressure, and blood gases were not different among

groups.

Fetal plasma ACTH and AVP were stimulated by infusion of HCI at a rate of 0.5

meq/min (Figure 5-3, Table 5-1). Fetal plasma cortisol concentration appeared to increase

in all groups (Table 5-1 and 5-2). A posteriori comparison of individual means by Duncan's

multiple range test indicated that in the 0.5 meq/min group, ACTH and AVP were

significantly increased (relative to 0 min) at 30, 40, 50 and 60 min and that cortisol was

significantly increased at 50 and 60 min (Table 5-2). Initial values of plasma ACTH and AVP

concentrations were not different among groups). In Figure 5-4 we present the relationships

between fetal arterial H' concentrations, Paco2 and the plasma concentrations of ACTH and

AVP in the 0.5 meq/min group. Figure 5-4 shows that both hormones are increased at H+

concentrations > 60 nM and Paco2 > 50 mm Hg. Because of the consistent relationship

between blood H+ and ACTH and AVP, there were significant correlations between pHa and

the logarithm of plasma ACTH (r=0.843, n=35) and AVP (r=0.773, n=28). The logarithms of











7.500- A


7.400 -
7.300-
7.200-
7.100-
7.000


60-



50-



40-


I I I
0 10 20


I I 1 1
30 40 50 60


Time (minutes)


Figure 5-1. Mean fetal pHa (A), Paco2 (B), and Pa02 (C) during intravenous infusion
of HCI at rates of 0.02 (filled circles), 0.10 (filled squares), and 0.50 (filled triangles)
meq/min.


E
E
C\J
0
O0
CL


0
-.


23-

21
19-

17-
15-








Table 5-1. Results of ANOVAs


F(df)
Variable
Do" Time Dose x time
pH 42.99" (2,13) 118.56- (6.78) 69.61" (12,78)
Paco, 8.00* (2,13) 36.04" (6,78) 12.65- (12,78)
Pao, 0.13 (2,13) 5.40' (6,78) 1.78 (12.78)
Hct 1.48 (2,13) 6.81* (6,78) 1.32 (12,78)
Mean arterial 0.87 (2,13) 1.35 (60,780) 1.00 (120,780)
pressure
Heart rate 0.12 (2,13) 3.95" (60,72U) 1.63- (120,720)
ACTH 4.53* (2,13) 9.86* (6,78) 5.70' (12,78)
Cortisol 0.46 (2,12) 5.50' (6.72) 1.87 (12.72)
AVP 2.33 (2,10) 2.87" (6.60) 3,65- (12.60)
PRA 2.82 (2,13) 3.81" (6,78) 1.12 (12,78)
Paco, and Pao,, arterial partial pressure of COz and 02. respect ively;
Hct, hematocrit; AVP, arginine vasopressin; PRA, plnsma renin activ-
ity; ANOVA, analysis of variance; df, degrees of freedom. P < 0.05.


















0.50 mEq/min


45




50-



45



55-



50-



45-


0.02 mEq/min


0 1 I 0 30 40 50
0 10 20 30 40 50 60


Time (minutes)
















Figure 5-2. Fetal MAP (left) and HR (right) during intravenous infusion of HC1 at
rates of 0.02 (bottom), 0.10 (middle), and 0.50 (top) meq/min.


-200

-180

-160

-140


mEq/min -200


-180



140


I I I i I I I
0 10 20 30 40 50 60


-200

-180

-160

-140









1000


%.


<


< E
Q-c'


C
CL


100


10.0-
5.0-



1.0 --


0.5


10-
8-
6
4 -
2-
0-


I I I I I I---1


0 20 40 60


Time (minutes)











Figure 5-3. Mean fetal plasma ACTH (A), AVP (B), and PRA (C) during intravenous
infusion of HCI at rates of 0.02 (filled circles), 0.10 (filled squares), and 0.50 (filled
triangles) meq/min.









Table 5-2. Fetal plasma cortisol concentrations


Time, Infusion Rate, nmeq/iin
miln 0.02 0.10 0.50

0 268 232 195
10 2610 224 204
20 236 204 195
30 296 224 256
40 35110 246t 276
50 317 224 t 316*f
60 33:9 194 337*t
Values are means SE. Significantly different from initial time
point in that group. t Significantly different from same time point in
0.02-meq/min group. t Significantly different from same time point in
0.1-meq/min group.
















1000 -
800-
600-
400-
200-
0-


E



I
I-
0
<








>1
0i

n


250-


200-


150-

itoo ....


I**
**


1 I I I I


S
S
0

0.


0
0
0
* 1.
0@
0
0


4 -- -


I I I I
20 60 100

H+ (nM)


p
...oo

il o J


I I T 1 T1 I


S

0

'p


* *


* O


* *


I I I I ,
20 40 60


-1000
-800
--6UU
-400
-200
-0


-14
-12
-10


-2
-4
-2


-250


-200


-150


100
80


PoC02 (mm Hg)


Figure 5-4. Relationships between arterial H+ concentration (left), Paco2 (right) and
plasma ACTH (top), AVP (middle), and HR (bottom) in 0.50 meq/min group. All
points within the 0.50 meq/min group are shown.


14
12
10-
8-
6-
4-
2-
0-


0
--
I

1

3


3-
-U


I


100








53

fetal plasma ACTH (r=0.742, n=35) and AVP (r=0.754, n=28) were also significantly

correlated to the changes in Paco2. As shown in Figure 5-4, the relationships between heart

rate and pHa and Paco2 were much less consistent. This may be related to the inherent

variability in fetal heart rate.

Fetal PRA was not stimulated by H' (Figure 5-3, Table 5-1). It is likely that PRA

increased equally in all groups in response to the blood sampling. The logarithm of fetal PRA

was not correlated significantly to the induced changes in pHa (r=-0.007, n=35) or Paco2 (r=-

0.238, n=35) during infusion of 0.5 meq/min. Initial values of PRA were not different among

groups (1-way ANOVA).

Fetal hematocrit was also significantly altered by the infusions, although the effect

was small (Table 5-3). A posteriori comparison of individual means using Duncan's multiple

range test indicates a small but significant increase in hematocrit in the 0.02 meq/min group

at 50 min relative to the initial value in that group (mean change in hematocrit=-2%) and a

significant increase in hematocrit at 30 min in the 0.5 meq/min group at 30 min relative to

the initial value in that group (mean change in hematocrit=+1%). The hematocrit values were

significantly lower in the 0.02 meq/min group at all time points compared with the other two

groups.




5.4 Discussion


The results of this study demonstrate that fetal ACTH and AVP secretion are

stimulated by fetal metabolic acidemia in fetal sheep between 132 and 140 days gestation. In

a previous study, we found that ACTH, AVP, and renin responses to slow hemorrhage were

significantly correlated to decreases in pHa and increases in Paco2 in fetal sheep between 128

and 132 days gestation (Chapter 4). The results of these two studies cannot be directly

compared because of the differences in gestational ages of the fetuses. However the results








Table 5-3. Fetal hematocrit


Time, Infusion Rate. meq/min
nrin 0.02 0.10 0.50

0 3841 412 411
10 381* 412 401
20 38+1 412 410
30 381* 402 42lt
40 371* 402 401
50 362*t 402 400
60 37t1* 402 400
Values are means SE. Significantly different from same time
point in other groups. t Significantly different from initial value in
that group.








55
of the present study suggest that fetal ACTH and AVP responses to fetal hemorrhage might

be at least partially mediated by increases in fetal Paco2 or decreases in pHa.

In adult animals, acute metabolic acidemia (produced by intravenous or intraduodenal

infusion of H') increases plasma concentrations of both glucocorticoid and mineralocorticoids

(Augustinsson and Johansson, 1986; Perez et al., 1979). Chronic metabolic acidosis, produced

by diabetic ketoacidosis, is also associated with increased plasma concentrations of cortisol,

aldosterone, AVP, and renin (Christlieb, 1976; Christlieb et al., 1975; Gerich et al., 1971;

Walsh et al., 1979). However, the hormonal responses to diabetic ketoacidosis may be in part

related to the associated hypovolemia (Waldhausl et al., 1979).

In fetal animals, acute acidemia is known to increase plasma concentrations of AVP

and epinephrine (Faucher et al., 1987). In that study, the investigators infused NH4C1 into

fetal sheep at 137+4 (SD) days gestation at a rate of 0.382 meq/min for 120 min. That rate

of acid infusion decreased fetal pHa from 7.37+0.01 to 7.04+0.05 at the end of the 120-min

infusion but did not alter fetal Pao2 and Paco2. The acidemia stimulated an increase in fetal

plasma AVP from 2.85 + 0.23 to 5.26 + 1.111/U/ml during the 120-min infusion period. In

the present study, the infusion of HCI at a rate of 0.5 meq/min for 60 min. The rate of HCI

infusion increased Paco2, as well as decreasing pHa, and increased plasma AVP. We found

that this infusion of HCI increased plasma AVP from 1.1+0.3 to 7.0+3.0 pg/ml, a nearly seven

fold increase. While it is not valid to compare absolute values for plasma AVP between

laboratories, it is apparent that Faucher and co-workers (Faucher et al., 1987) observed only

an approximately twofold increase in plasma AVP during their acid infusion study. The

major difference between our study and theirs is that our infusion of HCI increased Paco2,

while their infusion of NH4CI did not. This was probably related to the higher rate of

infusion and the stronger acid used in our experiments. Together, these studies may indicate

that there is a significant interaction between acidemia and hypercapnia in the control of AVP

secretion in the fetus. Whether this apparent effect of hypercapnia is mediated by the








56
peripheral or central chemoreceptors is not known. Another difference between these studies

is the rate of change of pHa. In the present study, pHa decreased at an apparently linear rate

of 0.36 pH units/h, and in the Faucher study pHa decreased at an apparently linear rate of

0.17 pHa units/h. The possible influence of the rate of change of pHa on AVP secretion is

unknown at the present time.

Fetal heart rate was increased by infusion of HCI at a rate of 0.5 meq/min.

Stimulation of peripheral fetal chemoreceptors by fetal intravenous cyanide injection or

occlusion of the maternal descending aorta produces reflex bradycardia (Iskovitz and

Rudolph, 1982). Analogously, chemoreceptor stimulation in adult animals produces

bradycardia if ventilation is maintained constant(i.e., if the reflex tachypnea is blocked). If

reflex tachypnea is allowed in adult animals after stimulation of the chemoreceptors, one

observes tachycardia, indicating an interaction between the chemoreceptor and pulmonary

afferent fibers in the control of heart rate. It is possible that in the present experiments the

fetal sheep developed tachycardia during HCI infusion of 0.5 meq/min because they also

started fetal breathing movements. We did not measure fetal breathing movement in this

study; however, it is known from other studies that hypercapnia stimulates fetal breathing

movement (Walker, 1984b). In summary, it is therefore possible that infusion of HCI at a rate

of 0.5 meq/min stimulated increases in fetal heart rate because of increased stimulation of

chemoreceptors combined with stimulation of fetal breathing movement. Fetal heart rate

during acid infusion was also measured in a study by Widness and co-workers (1986). In that

study, infusion of lactic acid at a rate of 0.58-0.75 meq/min decreased fetal arterial pH from

7.37 to 7.21 over the course of 4.5 hour. The rate of change of pHa was lower (0.04 pH

units/h) than in the present study. There were no significant changes in fetal heart rate or

Paco2* It is possible that the changes in heart rate and Paco2 in the present study were

partially dependent on the rate of change of pHa.









57
Infusion of acid into the fetal circulation increased fetal Paco2. In adult animals with

spontaneous ventilation, arterial acidemia reflexly increases ventilation, decreasing Paco2.

The fetal animal controls gas exchange by altering umbilical-placental blood flow. In the

present experiments, it is apparent that the Paos2 increased because umbilical-placental blood

flow did not increase accordingly.

We conclude that acute acidemia is a potent stimulus to ACTH and AVP secretion in

fetal sheep between 132 and 140 days gestation. The lack of response of PRA to the acidemia

may indicate insensitivity of renin to control by changes in pHa or Paco2 or may have been

the result of secondary inhibition of renin secretion by the increased Cl- load at the fetal

macula densa (Share and Levy, 1962). These results support the hypothesis that the acidemia

produced during hemorrhage might partially stimulate the ACTH and AVP responses to the

hemorrhage.

















CHAPTER 6
REFLEX CONTROL OF FETAL ARTERIAL PRESSURE
AND HORMONAL RESPONSES TO SLOW HEMORRHAGE



6.1 Introduction3


Fetal sheep respond to slow hemorrhage with increases in fetal plasma ACTH and

AVP concentrations and PRA (Alexander et al., 1974; Brace and Cheung, 1986; Drummond

et al., 1980; Ross et al., 1986). From Experiment I (Chapter 4), we found that bilateral

vagotomy, which interrupts afferent fibers from atrial type B receptors did not alter the

ACTH, AVP or PRA responses to hemorrhage. The receptors mediating the hormonal

responses to mildly hypotensive hemorrhage are therefore different in the fetus than in the

adult (Cryer and Gann, 1974; Share and Levy, 1962). We and others have demonstrated that

the ACTH, AVP, and renin responses to hemorrhage were significantly correlated to changes

in pH., but were not correlated to changes in arterial or venous pressure. From Experiment

II, we found that hypercapnic acidemia, produced by infusion of hydrochloric acid increased

fetal ACTH and AVP (Chapter 5). Together, the results of these studies suggest that ACTH

and AVP responses to slow hemorrhage might be chemoreceptor-mediated.

The present study was designed to investigate the role of peripheral chemoreceptors

in controlling the fetal hormone secretion during slow hemorrhage. To test this hypothesis,

we compared responses to slow hemorrhage in intact fetuses to responses in fetuses subjected

to bilateral vagotomy combined with bilateral carotid sinus denervation. The combined



3 This work was supported by the National Institute of Health grant HD-20098.

58








59
denervation impaired the ability of the fetus to defend blood pressure during the period of

hypovolemia, exaggerating the changes in fetal arterial blood gases usually produced by

hemorrhage. We therefore added a third group, denervation plus infusion of phenylephrine

(an oc-adrenergic agonist drug) to support blood pressure and umbilical-placental perfusion

during the period of hypovolemia. Comparison of these three experimental groups allows

assessment of the roles of carotid sinus and vagal afferents together in the control of reflex

hormonal and hemodynamic responses to slow hemorrhage in the late-gestation fetal sheep.



6.2 Methods


6.2.1 Experimental Protocol


Twenty-six chronically catheterized fetal sheep of 121-138 days of gestation were

studied. Four of them were twins. Nineteen of these fetuses were subjected to bilateral

section of cervical vagosympathetic trunks and the carotid sinus nerves in order to deafferent

peripheral chemoreceptors and baroreceptors.

All fetuses underwent the same 120-minute hemorrhage procedure, and each fetus was

studied once. At ten-minute intervals, 11 ml blood was withdrawn from a fetal arterial

catheter. One ml of the blood was drawn anaerobically for analysis of fetal pHa, Pa02, Paco2,

plasma protein, plasma sodium, potassium and Hct and 10 ml were used for hormone analysis.

Therefore, the total volume of blood withdrawn was 143 ml. One maternal arterial blood

sample was withdrawn at the beginning of each experiment. Each fetus was studied once.

The fetuses were divided into three groups: 1) intact fetuses (n=7); 2) denervated

fetuses (n= 12); 3) denervated fetuses infused intravenously with phenylephrine hydrochloride

(Winthrop-Breon Laboratories, Division of Sterling Drug, Inc.) using a syringe pump (Sage

Instruments model 341, Cambridge, MA). The rate of phenylephrine infusion was adjusted

to maintain the fetal blood gases at pre-hemorrhage levels. Infusion was started during the









60
period between 30 to 40 minutes after the onset of the hemorrhage; the rate of infusion was

varied between 0.01 ml/min and 0.38 ml/min (0.2-7.6 pg/ml). All ewes and fetuses were

sacrificed and the fetuses were weighed immediately after the experiments.




6.2.2 Calculated Variables


All of the following calculations were performed as described by Brace and coworkers

(Brace and Cheung, 1986; Brace, 1983).

Initial blood volume (BVo) was assumed to be approximately 110 ml/kg, and

subsequent blood volume at each time point (BVt) was calculated from BVo, Hot of each

sample (Hctt), and volume of blood removed from the fetus as measured by Brace (Brace and

Cheung, 1986). Therefore,



BVo = 110 ml/kg x Wt [1]

and

BVt = BVo x FRCV x Hcto/Hctt [2]



where Hcto and Hctt are initial and subsequent hematocrits, respectively, and FRCV is the

fractional red cell mass. In this study, FRCV was calculated as follows:

First, calculate the initial red blood cell mass (RBCo):



RBCo = BVo x Hcto/100



Then, subsequent RBC remaining (RBCt) at each 10-min interval of hemorrhage was

calculated as:










RBCt = Initial RBC RBC removed during hemorrhage

= RBCo 11 x (Hcto + Hctj +...+ Hctt)/100


Thus,


FRCV = RBCt / RBCo


Vr and Rf are referred to as the blood volume removed and the blood volume

restitution during hemorrhage. We have



Vr = 11 x n


Rf= Vr- (BVo BVt)


where n is the number of 11 ml samples. Therefore, the %R after hemorrhage was calculated

using the following equations (Brace, 1983):


%R = Rf/Vr x 100


Assuming that the fluid entering the circulation contained no protein, the percent

change of the calculated or theoretical plasma protein concentration (%APc) was calculated

(Brace, 1983) by


%APc = 100 x (Hcto Hctt)/(100 Hcto)


The %APa was calculated by


%APa = (Po Pt)/Po x 100









62
where Po is the plasma protein concentration before hemorrhage and Pt is the protein

concentration at each time point.

The protein concentration in restitution fluid (Pf) was calculated by



Prf = (Pt Pc) x BVt/Rf [6]

where

Pc = (1 + %AP') x Po




6.3 Results


Initial values of all fetal blood gases, blood pressures, HR and fetal and maternal

hormone concentrations were similar to values previously reported for healthy fetuses and

ewes in this laboratory (See previous chapters), suggesting that the fetuses had recovered from

surgery. Initial values of variables were not significantly different among experimental

groups as tested by one-way ANOVA.

Maternal plasma ACTH concentrations in intact, denervated, and denervated plus

phenylephrine infusion groups were 103+40, 167+39, and 120+41 pg/ml, respectively, and

maternal plasma cortisol concentrations in these groups were 11.7+2.5, 13.3+2.1, and 9.6+1.5

ng/ml, respectively. Maternal plasma vasopressin concentrations in the three groups were

0.80.2, 4.3+1.2, and 3.4+1.2 pg/ml, and plasma renin activities were 2.7+1.1, 1.7+0.2, and

2.00.8 ng Ang I/mlehr, respectively. None of the measured maternal plasma hormone levels

differed significantly among groups, as tested by one-way ANOVA.


6.3.1 Changes in Hemodvnamic Variables During Hemorrhage


Hemorrhage decreased fetal MAP in both intact (Group 1) and denervated (Group 2)

fetuses. The decrease in MAP was more severe in the denervated fetuses (Figure 6-1, top



















Group 3 (Denervated + Phe)


240
1210
-D
180
leo--
tr 150
S120


t I


0 40 80


120 0 40 80 120 0 40 80 120


Time (min from onset of hemorrhage)









Figure 6-1. MAP (top), HR (middle) and CVP (bottom) during hemorrhage in fetal
sheep which were intact (left), subjected to bilateral sino-aortic and vagosympathetic
denervation (middle), and subjected to denervation plus phenylephrine infusion
(right).


Group 2 (Denervoted)


Group 1 (Intact)


U**WF "Sf^ Y^I A


Ormel drOr


Y~c~o(rgl~Jli









64
panel). Infusion of phenylephrine increased arterial blood pressure during hemorrhage in

denervated fetuses (Group 3).

Heart rate increased in all groups during hemorrhage (Figure 6-1, middle panel).

There were no significant differences among groups. Fetal CVP was significantly decreased

during hemorrhage in intact fetuses (Figure 6-1, bottom panel).


6.3.2 Changes in Blood Gases During Hemorrhage


The initial values of Paco2 and pHa were not significantly different. Hemorrhage

significantly decreased fetal pHa and increased Paco2. The changes of pHa and Paco2

produced by the hemorrhage were greatly exaggerated by denervation; infusion of

phenylephrine restored the pHa response to a magnitude that was not different from that in

intact fetuses. Pa02 was not significantly altered by the hemorrhage (Figure 6-2, top panel).


6.3.3 Changes in Plasma Hormone Levels


Hemorrhage stimulated fetal ACTH and cortisol secretion in all three groups compared

to either intact or denervated fetuses infused with phenylephrine (Figure 6-3, left panel).

The responses were greatest in denervated (Group 2) fetuses. Infusion of phenylephrine

attenuated these responses such that the magnitudes of the responses were similar to those of

the intact fetuses. Plasma ACTH concentration was significantly higher than the other two

groups after 70 min in denervated fetuses. Plasma ACTH concentration in Group 3

(denervated plus infusion of phenylephrine) was lower than the concentration in Group 1

(intact) at 120 min. Plasma cortisol concentration was higher in denervated fetuses after 20

min than in either of the other two groups. There were no significant increases in cortisol

above initial concentrations in either group I or group 3.

Hemorrhage stimulated fetal AVP secretion in both intact and denervated fetuses.

Phenylephrine infusion abolished the AVP response to hemorrhage in denervated fetuses














Pa02 (mm Hg)


PHa




T I


PaC02 (mm Hg)


60-
55-
50


0 40 80 120
Time (min from onset of hemorrhage)


Figure 6-2. Pao2, pHa and Paco2 during hemorrhage in fetal sheep which were intact
(filled circles), denervated (filled triangles) and denervated plus phenylephrine
infusion (filled squares).


7.430

7.380-


.3304


7.280

7.230-

7.180-
























3000 Fetal Plasma ACTH (pg/ml)
*-* hItact
2000 A-A' D"enrvat.
*-U Den + phe iv /

10004





18 Fetal Plasma Cortisol
(ng/mi) 1 i-

12 -


6


25


0 40 80 120
Time (min from onset of


Fetal Plasma AVP
(pg/mi)




i- v4il


Fetal Plasma PRA
(ng/ml/hr)


(14 p


0 40 80 120
hemorrhage)


Figure 6-3. Fetal plasma ACTH (top left), cortisol (bottom left), AVP (top right) and
PRA (bottom right) during hemorrhage in fetal sheep which were intact (filled
circles), denervated (filled triangles) and denervated plus phenylephrine infusion.


O=









67
(Figure 6-3, right top panel). There was a significant suppression of AVP in the denervated

fetuses infused with phenylephrine after 110 min relative to the denervated fetuses not

infused with phenylephrine.

Hemorrhage stimulated similar PRA responses in all 3 groups (Figure 6-3, right

bottom panel).




6.3.4 Blood Volume Restitution


All fetuses were bled the same absolute volume of blood. The ratio of hemorrhage

volume to estimated initial blood volume was not different among groups: 47.0 0.1%,

47.0 0.1% and 53.7 0.1% for Groups 1,2, and 3, respectively. Figure 6-4 shows that intact

fetuses restored blood volume better than the fetuses in the two denervation groups. The

effect of the denervation on blood volume restitution is summarized in Table 6-1.

Denervation reduced the final restitution of blood volume 21.2% (calculated from volumes

expressed as ml/kg) or 23.4% (calculated from volumes expressed as the percentage of

hemorrhage volume). Infusion of phenylephrine did not restore the rate of blood volume

restitution observed in the intact fetuses.

Plasma protein concentrations decreased in all three groups, (Figure 6-5, left bottom).

The decrease in plasma protein concentration suggests that transcapillary refilling occurred

with fluid containing protein at a concentration less than that of plasma. Figure 6-6 illustrates

the percentage change in plasma protein that was measured and the calculated change

assuming there is no protein in the fluid entering the vascular space. The values were

significantly different between measured and calculated values within each group. This

analysis can be used to calculate the mean protein concentration in the fluid entering the

vascular space. These values are reported in Table 6.1. The protein concentrations range

from 2.7 0.4 to 3.7 0.2 g/dl, and are not significantly different among groups.

















20

16--

12--


0 40 80 120


Time (min from onset of hemorrhage)






Figure 6-4. Fetal volume restitution, which is represented as ml/kg (top) and
percentage changes of hemorrhage volume (bottom). Filled circles, triangles and
squares are represented intact, denervated and denervated plus phenylephrine infusion
groups.










Table 6-1. Initial and final fetal blood volume and fetal blood volume restitution.


INTACT DEN@ DEN+PHE@



INITIAL BV (ml) 309.0 + 15.0 319.0 + 22.0 273.0 + 17.0
(estimated)

FINAL BV (ml) 211.0 + 14.0 213.0 + 22.0 156.0 + 18.0
(estimated)

FINAL HEMORRHAGE 47.0 + 0.1 47.0 + 0.1 53.7 + 0.1
RATIO (% of BVo)

FINAL BV 16.5 + 1.9 #13.0 + 1.2 #*10.7 + 1.3
RESTITUTION (ml/kg)

FINAL BV 31.5 + 2.6 25.4 + 2.2 #*18.3 + 2.4
RESTITUTION
(% of hem. volume)

FINAL PROTEIN CONC. 3.7 + 0.2 3.5 + 0.6 2.7 + 0.4
IN RESTITUTION FLUID (g/dl)



# Duncan's multiple test shows significant difference from intact fetuses.

One-way ANOVA shows significant difference among groups.

@ DEN: denervated fetuses; DEN + Phe: denervated plus phenylephrine infusion






























4.300T Plasma Potassium (mEq/L)


4.100+


3 3.900-

2 3.700


3.500-


Plasma Protein (g/dl)


148 Plasma Sodium (mEq/L)


,10


0 40 80 120


0 40 80 120


Time (min from onset of hemorrhage)

















Figure 6-5. Fetal Hct (top left), plasma protein (bottom left), potassium (top right)
and sodium (bottom right) in fetuses which were intact, denervated and denervated
plus phenylephrine infusion (filled circles, triangles and squares, respectively).


5.700-

5.500


5.300

5.100

4.900


Plasma Hemotocrit (% PCV)


144

142































-30J
10-


0-


-10


-20


-30
10'


-20+


% Protein Concentration



0--0 6---



-
0.-6 _
1 ,.T T

GROUP I T 1 -.T













GROUP 2
(n=11)





--- .
""-r':. --i- --l-l_







'[3"- ? ,


GROUP 3
(n=7)


Time (min from onset of hemorrhage)








Figure 6-6. Percent changes in fetal plasma protein concentration. Open symbols
represent the theoretical changes and the filled symbols represent the actual changes.









72
Plasma potassium and sodium concentrations were only measured in 7 intact (Group

1) and 11 denervated (Group 2) fetuses (Figure 6-5, right). There was a tendency for plasma

potassium concentration to decrease during hemorrhage, but the apparent changes were not

statistically significant. There were also no significant differences in plasma sodium

concentrations between groups.




6.4 Discussion


This study was designed to test the role of peripheral chemoreceptors and cardio-

vascular mechanoreceptors in the control of the hormonal and cardiovascular responses to slow

hemorrhage in the late gestation fetal sheep. We found that section of the vagosympathetic

trunks plus bilateral denervation of the carotid arterial baroreceptors and chemoreceptors had

a profound effect on the responses to hemorrhage. We successfully controlled the exaggerated

changes in blood gases during hemorrhage by infusing phenylephrine (an a-adrenoreceptor

agonist) into the denervated fetuses during hemorrhage to constrict the fetal systemic

vasculature, and to increase placental flow to maintain changes in fetal arterial blood gases

that were similar to those observed in the intact fetuses. The infusion of phenylephrine did

not, however, improve BV restitution in denervated fetuses. The phenylephrine infusion

attenuated the ACTH and cortisol responses and blocked the AVP response, but did not

change the PRA response.


6.4.1 Alterations in Blood Gases and Vascular Pressures


The greater decreases in MAP and pHa and the greater increase in Paco2 suggested

that the afferent fibers sectioned in the denervated compared to intact fetuses are important

in the reflex vasomotor responses to hemorrhage. In a previous study, we compared responses

to hemorrhage in intact and vagotomized only fetuses and the later were not significantly









73
different from the responses in intact fetuses. Therefore, it is likely that afferent fibers from

the carotid sinus are critical for maintaining arterial pressure and blood gases during

hemorrhage. It is also possible that interruption of fibers important for maintenance of

arterial blood pressure during hypovolemia requires both vagotomy and carotid sinus

denervation, since we did not test the response to hemorrhage in fetuses subjected to carotid

sinus denervation alone.

The large changes in fetal arterial blood gases in the denervated fetuses not infused

with phenylephrine was most probably the result of reduced flow in the umbilical-placental

vascular bed secondary to reduced arterial blood pressure. Umbilical-placental flow is

influenced by the relative distribution of fetal combined ventricular output between the

umbilical-placental vascular bed and other vascular beds in the fetus (Itskovitz et al., 1982;

Faber et al., 1973; Perez et al., 1989). Reflex control of blood pressure and blood gases during

hemorrhage involves vasoconstriction in the fetal body and redistribution of combined

ventricular output to the placenta. The results of these and other experiments demonstrate

that the umbilical-placental vascular bed is less sensitive to alpha-adrenergic stimulation than

other vascular beds in the fetus (Berman, 1978; Oakes et al., 1980). Thus, infusion of

phenylephrine during hemorrhage in the denervated fetuses redistributes flow to improve gas

exchange.

Central venous pressure did not decrease as much in the totally denervated fetal sheep

as much as in the intact fetal sheep. It is possible that, if the denervation partially blocked

the reflex sympathetic efferent response to the hemorrhage, cardiac contractility would not

have been as high in the denervated fetuses as in the intact fetuses. We therefore speculate

that central venous pressure does not fall during hemorrhage in the totally denervated fetuses

because of a reduced of cardiac output compared to the intact fetuses.










6.4.2 Reflex Hormonal Responses


ACTH, AVP, and PRA were increased in the intact fetuses. The magnitudes of the

responses appeared related to the changes in pHa and Paco2, confirming results of previous

studies by us and others (Challis et al, 1989; Faucher et al, 1987; Chapter 4 and 5). However,

none of the responses were inhibited or attenuated by denervation, which suggests that the

sectioned fibers were not the sole mediators of the hormonal responses. In other experiments,

severe hypercapnic acidemia caused by HCl infusion in intact fetuses (pHa from 7.397+0.008

to 7.038+0.038; Paco2 from 43.1+1.8 to 57.5+3.2 mm Hg) increased both ACTH and AVP

secretion (Chapter 5). Therefore, assuming that the responses to hypotension in the intact

fetus are not mediated by cerebral ischemia, central chemoreceptors might mediate part of

the AVP response to slow hemorrhage. Together, these results prove that severe hypercapnic

acidemia is a stimulus to ACTH and AVP secretion. It is likely that peripheral

chemoreceptors do not mediate either the AVP or the ACTH response to slow hemorrhage.

The role of central chemoreceptors in the regulation of fetal ACTH and AVP secretion

might be greater during hemorrhage than during metabolic acidemia. Central chemoreceptors

might have been greatly stimulated by the combination of the exaggerated increase in PaCO2

and a possible reduction in cerebral blood flow in the denervated group caused by

hypotension. The augmented ACTH and the unchanged AVP responses in the denervated

group were either attenuated or inhibited after phenylephrine infusion. It is possible that

phenylephrine removed the remaining stimulus to central chemoreceptors by normalizing the

blood gas responses and increasing cerebral blood flow during the hemorrhage. Alternatively,

it is also possible that cardiac mechanoreceptors with non-vagal afferent fibers contributed

to the ACTH and AVP response to hemorrhage after denervation might reflect the different

sensitivity of each receptor not deafferented by the vagotomy and carotid sinus denervation.

Finally, it is not likely that osmoreceptors stimulated AVP secretion in current study.









75
Although we did not measure plasma osmolality, plasma sodium concentration did not change,

suggesting little change in plasma osmolality.

PRA responses to the hemorrhage were not altered by denervation or by infusion of

phenylephrine in the denervated fetuses. These results do not, by themselves, rule out the

control of renin secretion by peripheral or central chemoreceptors. However, changes in pHa

and Paco2 are not potent stimuli to renin secretion (Wood et al., 1989). We have demonstrated

that, although PRA responses to hypercapnia, isocapnic hypoxia, or hypercapnic hypoxia are

statistically significant, the responses are smaller in magnitude than the responses to

hemorrhage. PRA responses to hypercapnia are attenuated by sinoaortic denervation (Wood

et al., 1989). Together, the results of this and our previous studies suggest that the renin

response to hemorrhage is only partially mediated by peripheral chemoreceptors.

Alternatively, it is possible that PRA responses are also influenced by other factors, such as

humoral substances like atrial natriuretic peptide (Scheuer, 1987), and other cardiovascular

mechanoreceptors that are not identical with the carotid sinus or aortic arch

mechanoreceptors, and do not have afferent fibers in the vagosympathetic trunks.


6.4.3 Defense of Blood Volume


The reduction of fetal hematocrit during hemorrhage reflects the restitution of blood

volume. Since under normal conditions the placenta is quite impermeable to molecules even

as small as sodium (Schroder, 1982), the restitution fluid is thought to come mostly from the

fetal interstitial space (Robillard et al., 1979; Brace, 1983). The calculated protein

concentration in restitution fluid and the lack of change in the plasma sodium and potassium

concentrations in the present study support this notion. However, it is also possible that

during hypovolemia the permeability of the placenta increases, allowing transplacental inflow

of some fluid from the mother to the fetus (Manku et al., 1975). The high protein









76
concentration in restitution fluid is also consistent with increased lymphatic return or

decreased protein leakage from the circulation.

The average volume restitution in the intact group, expressed as the percentage of the

hemorrhage volume, is about 31.5% (ranged from 28.9% to 34.1%). These values are less than

those calculated in Brace's studies, approximately 50% (Brace and Cheung, 1986; Brace, 1989).

This might be related to the shorter observation period in the present experiments, and related

to the probable underestimation of fetal body weight before hemorrhage.

In conclusion, the results of the present experiment demonstrate that carotid sinus

baroreceptors or chemoreceptors are important for the maintenance of blood pressure and

blood gases during hemorrhage. The results of this and other studies from this laboratory

indicate that PRA responses to slow hemorrhage might be partially mediated by peripheral

chemoreceptors, and that AVP and ACTH responses are not mediated by peripheral

chemoreceptors. Elucidation of the possible role of the central chemoreceptors or other

baroreceptor populations needs further investigation.















CHAPTER 7
THE ACTH AND AVP RESPONSES TO NORMOXIC HYPERCAPNIA
IN FETAL AND MATERNAL SHEEP





7.1 Introduction4


The fetal hypothalamus-pituitary-adrenal axis and neurohypophysis are functional

beginning early in fetal life. The activity of these endocrine axes is increased by various

hemodynamic stimuli, starting at approximately 90 days gestation (Alexander et al., 1972) and

are important for intrauteral survival during fetal stress.

From Experiment I and III (Chapters 4 and 6, respectively), we have demonstrated

that slow hemorrhage stimulates plasma ACTH, cortisol and vasopressin secretion in late

gestation fetal sheep. In Experiment I, bilateral vagotomy showed no effect on hormonal

responses; the hormonal responses were significantly correlated to changes in arterial pHa, but

not to changes in arterial or venous pressure, which suggested that the ACTH and AVP

responses during hemorrhage might be chemoreflex-mediated. In Experiment II, severe

hypercapnic acidemia, produced by intravenous infusion of hydrochloride acid, stimulated

ACTH, cortisol and AVP secretion, suggesting that chemoreceptors might play an important

role in mediating fetal hormonal responses to hemorrhage (Chapter 5). While this assumption

was disproved by Experiment III (Chapter 6): 1) a more complete peripheral

chemodenervation did not attenuate the ACTH, cortisol, AVP and PRA responses to

hemorrhage; 2) when the fetal blood pHa and Paco2 changes during hemorrhage were


4 This work was supported by the National Heart, Lung, and Blood Institute
grant HL-36289 and American Heart Association.

77









78
balanced by infusion of an alpha-agonist, the ACTH, cortisol and AVP increases in

denervated fetuses were either attenuated or inhibited. Results from Experiment Series III

suggests that peripheral chemoreceptors do not mediate ACTH, AVP or PRA responses to

hemorrhage, and a central chemoreceptors was suggested.

This conclusion made us question the assumption we made from Experiment II

(Chapter 5), which is: chemoreceptors might mediate ACTH and AVP responses to

hemorrhage. Based on the facts that HCI infusion caused metabolic acidemia while slow

hemorrhage caused respiratory acidemia, and the former caused much severe acidemia than

slow hemorrhage did, We speculated that there must be a threshold for initiating

chemoreceptor regulated hormonal responses, and this threshold might not be reached during

slow hemorrhage.

The present study was designed to investigate whether mild hypercapnic acidemia,

similar in magnitude to the changes during hemorrhage but without the associated

hypovolemia, stimulates ACTH and AVP secretion in fetal sheep, and whether these responses

are mediated by peripheral chemoreceptors.



7.2 Methods


Twelve fetal sheep of 123-137 days of gestation were studied. All fetuses and their

mothers were chronically prepared with catheters and six of these fetuses had been subjected

to bilateral section of the cervical vagosympathetic trunks and the carotid sinus nerves in

order to deafferent peripheral chemoreceptors and baroreceptors as well as to remove other

vagal afferent fibers.

The ewes with their fetuses were subjected to two experiments: normocapnia control

and hypercapnia.

A five ml blood samples was withdrawn from both fetal and maternal arterial catheters

every 20 minutes for 1 hour. Immediately after drawing the first fetal and maternal blood









79
samples, a polyethylene bag was fitted over the ewe's head and neck, held loosely in place

with elasticized cloth. Normocapnia control was achieved by passing room air through the

bag. Hypercapnia was achieved by using a gas mixture of 21% 02 and between 5% and 8%

CO2 (adjusted to increase fetal Paco2 at least 10 mm Hg). Data from one of the experiments

in each group was eliminated because of the abnormal baseline levels of fetal blood gases or

hormones. The order of the experiments performed was randomized.

CVP was not measured in one of the intact fetuses in normocapnia experiment because

of the failure of the venous catheters.




7.3 Results


Initial values of all fetal and maternal blood gases, hormones and electrolyte

concentrations, fetal blood pressures and HR were similar to values previously reported for

healthy fetuses and ewes in this laboratory, suggesting that the fetuses and ewes had recovered

from surgery.


7.3.1 Responses in Fetuses


During normocapnia, blood gases, Hct and electrolytes did not change significantly

(Table 7-1).

During hypercapnia, pHa decreased from 7.390 + 0.001 and 7.398 + 0.015 to 7.257 +

0.011 and 7.281 0.010, Paco2 increased from 36.8 + 1.4 and 39.8 + 1.4 to 55.2 + 1.8 and 55.9

2.2 mm Hg in intact and denervated fetuses, respectively. Pao2 was increased from 25.5

1.4 and 26 + 1.5 to 31.6 + 1.9 and 33.6 1.8 mm Hg in intact and denervated fetuses

respectively. Changes in blood gases were not significantly different in intact and denervated

fetuses (Figure 7-1).



























PHO


n==*=6


60

55-


45

T*


4-


I I I I0
0 20 40 60


55--

50-

45-

40-

35-

30-I
0 20 40 60


Pa02 (mm Hg)






~Jo-4


20+


SI 61









I I I i
0 20 40 60


TIME (MINUTES)














Figure 7-1. Mean fetal pHa (left), Paco2 (middle), and Paos (right) in normocapnia
(upper) and hypercapnia (lower) studies. Open and filled symbols represent intact and
denervated fetuses respectively.


7.450-



7.350-


7.250-



7.150-

7.450-r


z
< 7.350
C-)
0
LJi
>-
I 7.250



7.150


PaC02 (mm Hg)









81
Plasma potassium concentration significantly increased during hypercapnia from 3.37

+ 0.07 and 3.42 + 0.11 to 3.62 + 0.10 and 3.63 + 0.19 meq/L in intact and denervated fetuses

respectively (Table 7-1).

There was also a small but significant increase in plasma sodium concentration from

142.3 + 0.8 to 143.7 + 0.7 meq/L in intact fetuses during hypercapnia (Table 7-1).

HR was significantly increased by hypercapnic acidemia only in intact fetuses (Figure

7-2). During normocapnia, there was a significant overall difference in HR between intact

and denervated fetuses, probably reflecting different initial values. CVP was decreased

slightly but significantly after 30 min during normocapnia in the denervated fetuses. CVP

did not change significantly in the other groups. There seems to be some fluctuations of both

MAP and CVP in both intact and denervated fetuses. MAP decreased at 20 and 25 minutes

in denervated fetuses compared to the MAP at 0 minutes, and then returned to normal (Figure

7-2).

There were no significant changes in plasma ACTH concentration in either normoxia

or hypercapnia experiments (Figure 7-3, upper panel). Plasma cortisol concentration was

equally and significantly increased in both intact and denervated fetuses at 40 and 60 minutes

(Figure 7-3, lower panel). The increases in plasma cortisol were equal in the intact and

denervated groups. The plasma concentration of AVP was approximately doubled during

hypercapnia; denervation did not attenuate the AVP response (Figure 7-4).


7.3.2 Responses in Ewes


Hypercapnia significantly decreased maternal pHa from 7.509 + 0.008 to 7.353 + 0.011;

Paco2 increased from 28.4 + 1.1 to 44.2 + 0.9 mm Hg and Pao2 from 100.2 + 2.6 to 119.9 + 4.2

mm Hg, respectively (Figure 7-5). Plasma potassium concentration also increased during

hypercapnia from 3.99 + 0.10 to 4.27 0.10 meq/L (Table 7-2). Plasma sodium concentration

increased slightly during hypercapnia at 20 and 60 minute points compared to the 0 minute











Table 7-1. Means & SEs of fetal HCT, plasma potassium & sodium


NORMOCAPNIA


HYPERCAPNIA


HCT (% PCV)
0
20
40
60


PLASMA POTASSIUM
(meq/L)
0 3.36 +0.11
20 3.41 + 0.11
40 3.41 + 0.12
60 3.47 +0.11

PLASMA SODIUM
(meq/L)
0 144.0 + 2.5
20 143.2 + 1.6
40 141.9 0.8
60 141.6+0.5


3.72 + 0.11
3.68 + 0.12
3.73 + 0.10
3.77 + 0.11



144.7 + 0.9
145.0 +1.1
144.3 + 0.8
144.3 + 0.7


3.37 + 0.07
3.51 + 0.08
3.57 + 0.08*
3.62 + 0.10*


3.4 +0.11
3.4 + 0.14
3.6 +.19*
3.6 + .24*


142.3 + 0.8 144.7 + 0.5
143.2 + 0.8* 145.5 + 0.5*
143.7 + 0.7* 145.0 + 0.5
143.2 + 0.7* 144.6 + 0.4


INT
29.1 + 3.1
29.4 + 3.2
28.8 3.2
30.1 + 3.7


DEN
27.0 + 2.0
27.2 + 2.0
27.1 1.9
27.1 + 2.0


INT
27.5 + 2.1
27.8 + 2.0
27.5 + 2.1
27.2 + 2.0


DEN
26.5 + 0.5
26.3 + 0.6
26.0 + 3.7
26.0 + 0.7


* Significant difference between the variables of certain time point and
their 0 minute measurements.



















MAP (mm Hg)


CVP (mm Hg)


2201

200


, lfS


T

ArIf, hI


-4-


I I- i- i- -- -10i 1 I I I -i 1 2
0 10 20 30 40 50 60 0 10 20 30 40 50 60


220 r


zoo I


TSO. T TLa T T'r T

140

120 1 1
0 10 20 30 40 50 60


TIME (minutes)










Figure 7-2. MAP (left), CVP (middle) and HR (right) in normocapnia (upper) and
hypercapnia (lower) studies. Open and filled symbols represent intact and denervated
fetuses respectively.


HR (bp/min)





















NORMOCAPNIA

0--0 Intact
0-0 Denervoted




T


O-009
O /0


35

30


S T


I


HYPERCAPNIA

A--A Intact
A-A Denervated


0 20 40 6C


0 0

Time (minutes)


20 40 60


Figure 7-3. Fetal plasma ACTH (upper) and cortisol (lower) concentration in
normocapnia (right) and hypercapnia (left) studies. Open and filled symbols represent
intact and denervated fetuses respectively.


250--


200--


150-


100--


50--


I


!I
















NORMOCAPNIA


0-0 Intact
*-* Denervated






bd11_i


HYPERCAPNIA


A-A Intact
A-A Denervated



T
A-^ _

A --------- ^""I


Time (minutes)










Figure 7-4. Fetal plasma AVP concentration in normocapnia (upper) and hypercapnia
(lower) studies. Open symbols represent intact fetuses, filled symbols represent
denervated fetuses.











0-0 Normocapnia (ewes)
*-- Hypercapnia (ewes)


.i -0-


T
TU
T
A/ rT 1
, -- --------I


I I


Time (minutes)
Figure 7-5. Maternal pHa (upper), Paco2 (middle) and Pao2 (lower) in normocapnia
(open squares) and hypercapnia (filled squares) experiments.


7.550-


7.450 -


7.350 -


7.250-
50-


251
140-


130--


120--


110-


100











Table 7-2. Maternal means & SEs of HCT, plasma potassium & sodium


NORMOCAPNIA


HYPERCAPNIA


HCT (% PCV)
0
20
40
60


PLASMA POTASSIUM
(meq/L)
0
20
40
60

PLASMA SODIUM
(meq/L)
0
20
40
60


21.8 + 0.7
21.4 + 0.7
21.5 + 0.7
21.2 + 0.7


3.93 0.05
3.89 + 0.05
3.92 + 0.05
3.89 + 0.06


148.0 + 0.4
146.5 + 0.8
147.8 + 0.5
148.0 + 0.5


22.0 + 0.9
23.8 + 1.2
24.3 0.9
23.7 + 0.9


3.99 + 0.10
4.17 +0.11
4.20 + 0.11
4.27 + 0.10


148.8 + 1.6
149.1 + 0.7
148.8 + 0.5
149.2 + 0.3


# significant difference between groups by Duncan's multiple range test.
* Significant difference between the variables of certain time point and
their 0 minute measurements.










variable (Table 7-2). Hct increased slightly during hypercapnia (Table 7-2).

Plasma ACTH and AVP increased significantly at 40 and 60 minutes and plasma

cortisol significantly increased at 20 and 40 minutes compared to their 0 minute values (Figure

7-6).


7.4 Discussion


The results from this and other studies in this laboratory (Chapter 4) suggest that

decreases in pHa and/or increases in Paco0 stimulated fetal ACTH and AVP secretion. In

Experiment II (Chapter 5), we intravenously infused HCI into fetuses causing a decrease of

pHa from 7.400 to 7.033 and an increase of Paco0 from 37.8 to 55.2 mm Hg. In the present

study, we produced approximately the same degree of hypercapnia (37.8 to 55.2 mm Hg) but

less acidemia (pHa decreased from 7.390 to 7.257). ACTH was not stimulated and AVP

responses to hypercapnic acidemia in the present study were milder than those during infusion

of HC1, suggesting that hypercapnia alone is a mild stimulus to AVP and ACTH. Further

reductions of pHa via metabolic acidosis appears greatly increase the strength of the stimulus.

This is consistent with the results of a study from the laboratory of Daniel (Daniel et al.,

1983), in which AVP was more highly correlated to pHa than to Paco2. The design of the

present study does not, however, confirm that H' is the only stimulus to AVP secretion during

hypercapnia, as suggested by other authors (Wang et al., 1984; Faucher et al., 1987).

Acid infusion experiments in dogs, goats, and rats (Augustinsson and Forslid, 1989;

Perez et al., 1979; Welbourne, 1976) suggest that H+ itself can act as a stimulus to ACTH

secretion. In anesthetized dogs, Richards (1957a) dissociated the decrease of pHa from an

increase of Paco2 by HCI or NaHCO, infusion, and demonstrated that changes in either H+

or Paco2 alone could stimulate the hypothalamus-pituitary-adrenalcortical axis. Results of

studies in which adult animals were subjected to ventilatory hypercapnia are not consistent.

In some studies, ACTH was increased (Augustinson et al., 1989; Bloom et al., 1977; Raff et