Hormones and the Fetus

By J. R. Pasqualini, F. A. Kincl and C. Sumida

Volume One of Hormones and the Fetus contained a generalization of the role of hormones during fetal life. In this second volume, the main topics focus on hormonal binding to fetal and maternal plasma proteins, amniotic fluid and placenta; the presence of hormone receptors in the fetal compartment; and the mechanism of action of various hormones in different target tissues of the fetus and placenta. This up-to-date information is of major importance in the study of hormonal activity in fetal life, and the biological effect of hormones through the receptor. The mechanism of fetal sexual differentiation, in which hormones play an important role, is extensively covered.

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 22, 2013
• ISBN: 9781483286686

Book preview

HORMONES AND THE FETUS

Volume II

JORGE R. PASQUALINI

CNRS Steroid Hormone Research Unit, Foundation for Hormone Research, 26 Boulevard Brune, 75014 Paris, France

The City University of New York, Staten Island, New York, USA

CHARLOTTE SUMIDA

CNRS Steroid Hormone Research Unit, Foundation for Hormone Research, 26 Boulevard Brune, 75014 Paris, France

Table of Contents

Cover image

Title page

Inside Front Cover

Copyright

Preface

Chapter 1: The Binding of Hormones in Maternal and Fetal Biological Fluids

Publisher Summary

Introduction

Conclusions

Chapter 2: Receptors, Mechanism of Action and Biological Responses of Hormones in the Fetal, Placental and Maternal Compartments

Publisher Summary

Introduction

A ESTROGEN RECEPTORS

B PROGESTERONE RECEPTORS

C ANDROGEN RECEPTORS

D GLUCOCORTICOID RECEPTORS

E MINERALOCORTICOID RECEPTORS

F THYROID HORMONE RECEPTORS

G LUTEINIZING (LH) AND FOLLICLE-STIMULATING (FSH) HORMONE RECEPTORS

H PROLACTIN RECEPTORS

I GROWTH HORMONE RECEPTORS

J PLACENTAL LACTOGEN RECEPTORS

K GONADOTROPIN-RELEASING HORMONE (GnRH) RECEPTORS

L ADRENOCORTICOTROPIC HORMONE (ACTH) RECEPTORS

M INSULIN AND GLUCAGON RECEPTORS

N OXYTOCIN RECEPTORS

O RELAXIN RECEPTORS

P ANGIOTENSIN RECEPTORS

Chapter 3: Sex Differentiation and Fetal Endocrinology

Publisher Summary

Introduction

Index

Inside Front Cover

Book Title of Related Interest

PASQUALINI and KINCL

Hormones and the Fetus Volume I

Production, concentration and metabolism during pregnancy

Journal Titles of Related Interest

Cellular Signalling

Current Advances in Physiology

Journal of Steroid Biochemistry and Molecular Biology

Progress in Growth Factor Research

Reproductive Toxicology

Copyright

Copyright © 1991 Pergamon Press plc

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means; electronic, electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without permission in writing from the publishers.

First edition 1991

Library of Congress Cataloging-in-Publication Data

Pasqualini, Jorge R.

Hormones and the fetus.

Includes bibliographies and index.

Vol. II by: Jorge R. Pasqualini, Fred A. Kincl, Charlotte Sumida.

Contents: v. I. Production, concentration and metabolism during pregnancy.—v. II. [without special title]

1. Obstetrical endocrinology. 2. Placental hormones. I. Kincl, Fred A. II. Sumida, Charlotte. III. Title. [DNLM: 1. Hormones—Physiology. 2. Hormones—Biosynthesis. 3. Maternal-fetal exchange. 4. Reproduction. 5. Fetus. W 210.5 P284h]

RG558.5.P37 1985 612′.647 84-344

British Library Cataloguing in Publication Data

Pasqualini, Jorge R.

Hormones and the fetus.

Vol. II

1. Man. Foetuses. Development. Role of hormones

I. Title II. Kincl, Fred A. III. Sumida, Charlotte

612.647

ISBN 0-08-035720-2

Printed in Great Britain by BPCC Wheatons Ltd., Exeter

Preface

Volume I of Hormones and the Fetus covered the quantitative and qualitative aspects of different hormones in the three compartments (fetal, placental and maternal) of the human and other mammalian species, the transfer between the compartments and the qualitative and quantitative hormonal changes involved before and during parturition.

This second volume of Hormones and the Fetus gives a general idea of the mechanism of action of the different hormones during fetal development.

The first chapter of Volume II includes the interaction of different hormones with fetal plasma and other biological fluids of the fetal and placental compartments. Despite the fact that at present the role of this interaction is not very clear, one can accept that the binding of the hormone to these proteins could be a control of the biological activity or to serve as a reserve source for the active hormone. Some proteins bind specifically and with high affinity to limited species: progesterone binding globulin in the guinea-pig and other hystricomorphs; α-fetoprotein, which is present in most species, binds specifically and with high affinity estradiol in only two species, the rat and the mouse; uteroglobin in the uterine fluid of the rabbit.

The second chapter of this volume covers the presence of receptors of the various hormones and the biological hormonal responses in different fetal tissues and in the placenta. The discovery in the early seventies of estrogen receptors and mineralocorticoid receptors in the fetal compartment (J.R. PASQUALINI’s Laboratory) has opened new possibilities for the investigation of the mechanism of action of hormones during embryonic and fetal life. The presence of receptors in the fetal compartment was extended to other steroid hormones and polypeptide hormones and interesting correlations were found between the presence of these receptors and the biological responses, indicating that many biological activities of the hormone are initiated during fetal life and their receptors could be an obligatory step in the mechanism of the hormone action.

The third chapter deals with sex differentiation and fetal endocrinology, including gonadal sex and the hormonal control of sexual development. In this chapter, the work on the origin of germ cells and the development of the gonads was carried out by Dr J.E. JIRASEK from Czechoslovakia to whom we would like to express our deepest thanks. This chapter also includes a general idea of the teratological effect of steroid hormones during fetal life.

We would like to thank the staff of Pergamon Press, in particular Mr Richard Marley, and Ms S.Y. MacDonald (CNRS Steroid Hormone Research Unit, Foundation for Hormone Research) for their efficient collaboration in the preparation of this book.

J.R. PASQUALINI

1

The Binding of Hormones in Maternal and Fetal Biological Fluids

Publisher Summary

Many hormones secreted from specific organs are bound to carrier proteins in blood. The interaction is reversible and depends on the type of hormone, the animal species, and physiological conditions. During pregnancy, the binding of hormones to plasma proteins usually increases Some plasma-binding proteins may be found only in specific animal species, for example, progesterone binding globulin (PBG) appears in guinea pig and other hystricomorphs but not in other species. The interaction of various hormones with plasma proteins presents significant differences in the physicochemical characteristics of the binding, particularly in their affinity and specificity. This chapter discusses the main characteristics of the proteins that interact with hormones present in the plasma of the fetal or maternal compartments and in amniotic fluid during the course of gestation in humans and other mammals. The presence of proteins that bind specifically to various steroid hormones is very limited: α-fetoprotein only in mouse and rat, PBG only in the guinea pig, and uteroglobin only in the rabbit.

Contents

INTRODUCTION

1 THE BINDING OF PROGESTERONE AND OF PROGESTERONE DERIVATIVES IN PLASMA DURING GESTATION

1.1 In Humans

1.2 In Other Mammals

1.2.1 Effects of progesterone binding globulin (PBG)

1.2.1.1 Purification of PBG

1.2.1.2 Structure and physicochemical properties of PBG

1.2.1.3 Physicochemical characteristics of the PBG-progesterone complex

1.2.1.4 Steroid conformation, crystal structure and binding

1.2.1.5 Concentration in maternal and fetal plasma of guinea-pig

1.2.1.6 The origin of PBG

2 ESTROGEN BINDING DURING GESTATION

2.1 Alpha-Fetoprotein (AFP)

2.1.1 Purification and physicochemical properties of AFP

2.1.2 Chemical composition of AFP

2.1.3 Binding of AFP to estrogens

2.1.4 Biosynthesis and control of AFP

2.1.5 AFP and the nervous system

2.1.6 AFP concentrations during development

2.1.6.1 In humans

2.1.6.2 In different animal species

2.1.7 Half-life and fetal–maternal transfer of AFP

2.1.7.1 Biological importance of AFP

3 ANDROGEN BINDING PROTEINS

3.1 Sex Steroid-Binding Protein (SBP)

3.1.1 In human pregnancy

3.1.1.1 Maternal plasma

3.1.1.2 Fetal plasma

3.1.1.3 Amniotic fluid

3.1.2 In different animal species

4 CORTICOSTEROID BINDING GLOBULIN (CBG)

4.1 Plasma Concentration of CBG during Gestation in Different Animal Species

4.2 Biosynthesis of CBG

4.3 Control of CBG

4.4 Physicochemical Properties of CBG

5 UTEROGLOBIN

5.1 Physicochemical Properties of Uteroglobin and Binding to Progesterone

5.2 Concentration of Uteroglobin during Gestation of the Rabbit

5.3 Control of Uteroglobin Secretion

5.4 Biosynthesis of Uteroglobin

5.5 Uteroglobin in Other Organs

CONCLUSIONS

REFERENCES

Introduction

Many hormones secreted from specific organs are bound to carrier proteins in blood. The interaction is reversible and depends on the type of hormone, the animal species and physiological conditions. During pregnancy the binding of hormones to plasma proteins usually increases. For example, the plasma concentration of corticosteroid binding globulin (CBG) is many times higher during gestation in many species.

Some plasma binding proteins may be found only in specific animal species, e.g. progesterone binding globulin (PBG) appears in guinea-pigs and other hystricomorphs but not in other species (see Section 1.2.1). Another interesting example is α-fetoprotein, probably present in most mammals, which binds specifically and with high affinity estradiol in only two species: the rat and the mouse.

These interactions between a hormone and a binding protein may have several functions: 1. to control the biological activity of the hormones since only the unbound, unmetabolized hormone exerts biological activity. This aspect is of importance because the production rates of many hormones (progesterone, estrogens, corticosteroids) increase very significantly during pregnancy (see Volume I, Chapter 3); 2. to serve as a reserve of the potentially active hormones, since the enzymatic systems which transform a hormone into an inactive metabolite either do not operate, or do so to a very limited extent, on the hormone–protein complex; 3. to facilitate the transport of hormones, an aspect of probably limited importance since the binding of the same hormone differs from species to species, and the hormones can reach their target organs by simple diffusion from the blood.

The interaction of various hormones with plasma proteins presents significant differences in the physicochemical characteristics of the binding, particularly in their affinity and specificity. In addition, the interaction of hormones with plasma proteins is influenced by the relative concentration of the other blood proteins. For instance the concentration of albumin, which binds most steroids with low affinity, is 1000 times greater than that of corticosteroid binding globulin (CBG) and thirty times greater than that of α1-acid glycoprotein (AAG) (Table 1.1). Finally, the presence of the other hormones may influence the hormone-plasma protein binding.

TABLE 1.1

Concentrations and Molecular Weights of Some Steroid Hormone Binding Proteins in Human Plasma

In this chapter we give a general outline of the main characteristics of the proteins which interact with hormones present in the plasma of the fetal or maternal compartments and in amniotic fluid during the course of gestation in humans and other mammals. (For a recent and exhaustive review of steroid–protein interaction, except the binding to receptor, see Westphal, 1986.)

1 The Binding of Progesterone and of Progesterone Derivatives in Plasma During Gestation

Progesterone is associated mainly with three plasma proteins: corticosteroid binding globulin (CBG), human serum albumin (HSA), and α1-acid glycoprotein (AAG). The distribution of progesterone binding to various proteins is shown in Table 1.2. Half of the progesterone present is associated with serum albumin which, in spite of its low affinity for progesterone (see Table 1.3) binds large amounts because it is present in very high concentrations (see Table 1.1).

TABLE 1.2

Distribution of Progesterone Bound to Different Proteins During Human Pregnancy

Quoted from Westphal (1966); Rosenthal et al. (1969).

TABLE 1.3

Affinity Constants of Progesterone Binding Proteins in Human Serum

Quoted from Westphal et al. (1977a,b).

1.1 In Humans

The production rate and plasma concentration of progesterone increase very significantly during pregnancy (see Volume I, Chapter 3) but only a small fraction of the native hormone, 5–7% of the total circulating progesterone, is present in an unbound form. The percentages of unbound progesterone, the absolute values during the luteal phase, during the first trimester and at term, and in the umbilical vein are shown in Table 1.4. Batra et al. (1976) found that unbound progesterone increased with the advance of pregnancy and reported values of 6% of the total progesterone at week 24 of gestation and 13% at term (Fig. 1.1). Rosenthal et al. (1969) give values of only 1.8%, while Yannone et al. (1969) report a range between 1.3 and 11%. The discrepancies may be due to the different methods used for the measurement. A significant increase in absolute values of unbound progesterone would influence its biological activity during gestation.

TABLE 1.4

Concentrations of Bound and Unbound Progesterone in Plasma During the Human Menstrual Cycle and in Pregnancy

Quoted from Tulchinsky and Okada (1975).

FIG 1.1 Mean Concentrations of Total and Unbound Progesterone (P) in the Last Trimester of Human Pregnancy. Quoted from Batra et al. (1976).

Immediately after birth (1–3 h), the plasma concentration of progesterone abruptly decreases, while the percentage of the unbound fraction increases. In the newborn the rise in the concentration of unbound progesterone increases, possibly the result of a significant increase in corticosteroids during labor, which compete with progesterone for binding, or the decrease in the concentration of CBG due to the decreased production of estrogens.

1.2 In Other Mammals

1.2.1 EFFECTS OF PROGESTERONE BINDING GLOBULIN (PBG)

In 1967, Heap and Deanesly observed drastic changes in progesterone metabolism during gestation in the guinea-pig, and in 1969 two groups, Diamond et al. and Heap found in the serum of pregnant animals of the same species, a protein which specifically bound progesterone with high affinity. This unique plasma protein has so far been found also in related species (chinchillas, cuis, degu, coypus, viscachas, casiraguas and tuco-tucos) (Heap and Illingworth, 1974; Ackland et al., 1979) during gestation.

1.2.1.1 Purification of PBG

The concentration of PBG can reach a value of 1 g/l (see Table 1.12) in the maternal serum of pregnant guinea-pigs. The protein has been purified by ammonium sulfate precipitation, gel filtration, ion-exchange chromatography and electrophoresis (Lea, 1973; Milgrom et al., 1973; Burton et al., 1974). A practical procedure was described by Stroupe and Westphal (1975a) who used an acid chromatography column of sulfopropyl Sephadex (pH 4.5) which allows the adsorption of most of the other serum proteins. As PBG has a low isoelectric point (pI 2.8), it is eluted from this column in the void volume. The partially purified PBG was then subjected to affinity chromatography using activated Sepharose 4B condensed with the 17-hemisuccinate of 19-nor-testosterone or the hemisuccinate of deoxycorticosterone (Cheng et al., 1976). Purified PBG was obtained after elution from the affinity chromatography column using 10 μM of progesterone or of 5α-pregnane-3,20-dione solutions.

TABLE 1.12

The Concentrations of Progesterone Binding Globulin in Plasma of Different Hystricomorph Rodents During Gestation and its Physicochemical Properties

Quoted from Illingworth et al. (1973); Heap and Illingworth (1974); Ackland et al. (1979); Heap et al. (1981).

1.2.1.2 Structure and Physicochemical Properties of PBG

Progesterone binding globulin is rich in carbohydrates (71%) (Table 1.5). Using Sephadex G-200 columns, Burton et al. (1974) reported that PBG is composed of two forms: PBG-I, with a molecular weight of 117 300 Da and PBG-II of 78 400 Da. The presence of various carbohydrates in the two forms of PBG are shown in Table 1.6 and the details of amino acid composition (in %) in Table 1.7. The physicochemical properties of PBG and of the two forms PBG-I and PBG-II are summarized in Table 1.8. The low values of pI agree with the predominance of acid components.

TABLE 1.5

Composition of Progesterone Binding Globulin (PBG)

Quoted from Harding et al. (1974); Stroupe and Westphal (1975a); Westphal et al. (1977a,b); Evans et al. (1982).

TABLE 1.6

Composition of Progesterone Binding Globulin: PBG-I and PBG-II

Quoted from Burton et al. (1974).

TABLE 1.7

Amino Acid Composition of PBG-I and PBG-II

Quoted from Burton et al. (1974).

TABLE 1.8

Physicochemical Properties of Progesterone Binding Globulin (PBG)

Quoted from Lea (1973); Milgrom et al. (1973); Burton et al. (1974); Stroupe and Westphal (1975b); Cheng et al. (1976).

1.2.1.3 Physicochemical Characteristics of the PBG–Progesterone Complex

The association constant Ka of the binding of progesterone to PBG is 2.2 × 10⁹ M−1 at 4°C and 0.35 × 10⁹ M−1 at 37°C (Stroupe and Westphal, 1975b). The affinity constants of PBG for different steroids (Table 1.9) indicate that the PBG–steroid interaction decreases with increasing steroid polarity. The data suggest that this interaction is hydrophobic in nature. The affinity for a synthetic progestagen, medrogestone, is two times that of progesterone and the Ka for Cortisol is 1/1000 that of progesterone.

TABLE 1.9

Affinity Constants of the Binding of Various Steroids to Progesterone Binding Globulin

Quoted from Stroupe and Westphal (1975b).

PBG–progesterone complex is relatively resistant to temperature and pH changes. After 30 min of heating at 50°C, 70–80% of progesterone is still bound with high affinity (MacLaughlin and Westphal, 1974) (Table 1.10). The isolation of PBG in pure form allows the measurement of the binding equilibrium between the hormone and the macromolecule. Quenching of the strong fluorescence signal of the interaction of PBG and 3-oxo-4-ene steroids provides a sensitive indicator to study the association and dissociation rate constants. Stroupe and Westphal (1975b) used the stopped-flow fluorometry method and evaluated the association rate constants (k+1) between progesterone and PBG to be 8.6 × 10⁷ M−1 s−1 at 20°C, with a half-life of 22.5 min. The dissociation rate constant (k–1) at the same temperature was 0.053 s−1 with a half-life of 13.1 s. In PBG, as well as in the two forms of PBG (I and II) there is one steroid binding site per molecule (Stroupe and Westphal, 1975b; Westphal et al., 1977a,b).

TABLE 1.10

Physicochemical Properties of the PBG–Progesterone Complex

Quoted from Stroupe and Westphal (1975b).

The kinetic parameters of the steroid hormone–protein complex are of physiological and biological importance. A faster binding of the hormone to plasma proteins facilitates the protective mechanism against an enzymatic attack of the hormone and avoids an excess of the circulating unbound progesterone. A rapid dissociation provides a ready source of the biologically active hormone.

In contrast to plasma protein binding, the dissociation rate of the hormone-receptor complex in the target tissue is about 500 times lower: the half-time of dissociation of rat liver glucocorticoid receptor at 37°C is 13 min (Koblinsky et al., 1972) while that of the PBG-progesterone complex (same temperature) is only 1.8 s. The longer binding of a hormone to a receptor molecule is necessary for the different steps of the hormone action and genome expression.

1.2.1.4 Steroid Conformation, Crystal Structure and Binding

The binding of a steroid to a macromolecule is a function of different parameters: polarity, chemical structure, optimal contact and spatial relationship. In the last factor, the steric hindrance of a substituent in the steroid moiety can significantly alter the affinity of the complex (see for review Westphal, 1986).

In recent years a comparison of the crystallographic steroid conformation and dimensional data has been extensively used to correlate a structure with biological activities (for a review see Duax and Norton, 1975; Duax et al., 1988; Griffin et al., 1984). An important condition required to validate the application of crystallographic steroid conformation to biological function is that the conformational structure revealed by X-ray crystallographic data must be similar to that of the steroid-protein binding in aqueous solution. In the case of PBG–progesterone complex there is a good correlation between the binding affinity and the planar structure, e.g. progesterone and 5α-dihydroprogesterone have a significantly higher affinity for PBG than 5β-dihydroprogesterone (angular structure).

The volume of space of one molecule of progesterone is 448 ų (5.2 × 6.3 × 13.8 Å) (Westphal, 1958). At 40 days of gestation, maternal plasma contains around 50 μg/100 ml of progesterone, indicating that each unbound progesterone molecule is surrounded by a solvent volume of more than 100 million times its own.

1.2.1.5 Concentration in Maternal and Fetal Plasma of Guinea-pig

Westphal (1971) measured progesterone binding activity in the serum of pregnant and lactating guinea-pigs using the combining affinity C, a concept introduced by Daughaday in 1958a,b. The combining affinity is defined as:

where Sbd is the concentration of the bound fraction of the steroid, S the unbound, and Pt the total protein.

The equilibrium dialysis method using tritiated progesterone reveals that at 40–50 days of gestation the levels of PBG reach a maximal concentration in the maternal plasma of more than 1 g/l (1.2 × 10−5 mol) (Fig. 1.2(a)). Significant amounts are found from 20 days post-coitum, and there is a sharp decrease after 50–55 days of gestation, and particularly after birth (Lea et al., 1976; Evans et al., 1981, 1982). Figure 1.2(b) shows that there is a parallel increase in progesterone levels and PBG concentration. Quantitatively, there is a large molar excess of PBG over progesterone; in early gestation the PBG ratio to progesterone is about ten and increases many times at 40–50 days of gestation. As a result, the percentage of circulating unbound progesterone, the physiologically active hormone, is very low.

FIG 1.2 Progesterone Binding Globulin (PBG) and Progesterone (P) Plasma Concentrations in Guinea-pig during Gestation. Quoted from Lea et al. (1976).

PBG is also found in the fetal plasma of guinea-pig (40–60-day-old fetus) (Castellet and Pasqualini, 1973; Millet and Pasqualini, 1978; Perrot and Milgrom, 1978). The progesterone receptor is present in this compartment mainly in the uterus and ovaries (Pasqualini and Nguyen, 1980). Comparison of the physicochemical properties of the interaction of progesterone with the receptor and with PBG shows differences in the characteristics of these two complexes (see Table 2.18, Chapter 2, Section 1.1.1.1). PBG is mainly found in the plasma; it does not penetrate and is not synthesized in the progesterone target organs, such as the uterus.

The PBG concentrations in maternal and fetal plasma, in the umbilical artery and vein and amniotic fluid are indicated in Table 1.11. Using an immunoenzymatic assay, only very low levels of PBG were found in nonpregnant females and in males (Perrot and Milgrom, 1978). It is interesting to note that PBG is present in the milk of lactating guinea-pigs with a concentration of 26.5 ± 12 nM on day 1 post-partum (Raymoure and Kuhn, 1980). The plasma concentrations of PBG in other hystricomorph rodents are indicated in Table 1.12. The table also shows the binding capacities and the association constants of PBG in these species.

TABLE 1.11

Concentration of Progesterone Binding Globulin (PBG) in Pregnant and Nonpregnant Guinea-pigs and in the Amniotic Fluid

ND: Not detectable

*1 g PBG ∼ 1.2 × 10−5 mol.

Quoted from Diamond et al. (1969); Heap and Illingworth (1974); Perrot and Milgrom (1978); Evans et al. (1981).

As a result of the high concentrations of PBG, the metabolic clearance rate of progesterone decreases drastically in pregnant guinea-pigs, from a value of 1128 ± 7.0 to 8.3 ± 0.81 plasma/day/kg (Heap, 1970; Illingworth et al., 1970). The high affinity binding influences progesterone extractibility from plasma; in nonpregnant animals most of the hormone (95%) is extracted with ether, but in pregnant guinea-pig only 16–18% is obtained with that procedure and for a complete extraction, plasma proteins must be denatured with NaOH.

1.2.1.6 The Origin of PBG

PBG is synthesized most likely by the placenta. Metz et al. (1977) used immunohistochemical techniques to localize PBG in the synctiotrophoblast of the guinea-pig placenta and Perrot-Applanat and David-Ferreira (1982) located the protein in different organelles: the rough endoplasmic reticulum, Golgi apparatus and perinuclear space of the same organ. The latter authors were unable to localize PBG in the liver, muscles, heart, lungs, kidneys, ovaries or uterus. The origin of the PBG present in fetal plasma remains to be elucidated.

In the guinea-pig the concentrations of PBG rise sharply by the period of 15–20 days of gestation, which coincides with the time when the definitive placenta is established and the developing allantois establishes close contact with the chorion. However, in the viscacha PBG rises 1–2 weeks before the time of formation of the definitive placenta and in the casiragua the increase in PBG is 10 days after the formation of the allantochorionic placenta. These data also indicate that the origin and the mechanism of control of PBG remain to be elucidated.

What is the physiological role of PBG which is present in a limited number of species and only during gestation? Despite the fact that this protein can control progesterone activity, the reason for the very high concentrations present in both maternal and fetal compartments is an important aspect to be elucidated.

2 Estrogen Binding During Gestation

2.1 Alpha-Fetoprotein (AFP)

In 1956, Bergstrand and Czar discovered in the human fetal plasma a protein called α-fetoprotein (AFP), which was not detected in the plasma of children or adults. This protein accounted for 10% of the total fetal serum proteins (Bergstrand and Czar, 1957). It was demonstrated that the physicochemical properties of this protein differed from those of another fetal protein ‘the fetuin’ found previously by Pedersen in 1944 in the fetal sera of bovine and other animal species (Pedersen, 1947; Putnam, 1965). It can be remarked that fetuin throughout the gestational period of different species (e.g. sheep, cattle, pigs, goats) is quantitatively one of the most important proteins. It constitutes up to 5 g/l of fetal plasma (Dziegielewska et al., 1980).

AFP gained considerable importance in the studies on the mechanism of action of steroid hormones when it was observed that AFP bound estradiol with high affinity in two species: rat and mouse, like the binding of estrogens to the receptor molecule (Soloff et al., 1971; Nunez et al., 1971a; Uriel et al., 1972; Savu et al., 1972).

Another attractive aspect of this protein is its presence in the serum of mice bearing primary hepatomas (Abelev, 1963, 1968) and in the serum of humans with primary hepatocellular carcinoma (Tatarinov, 1965). Elevated serum concentrations of AFP are not only found in tumors of gonadal or extragonadal origin (Abelev et al., 1967; Masopust et al., 1968; Nørgaard-Pedersen and Axelsen, 1978), but also in some major abnormalities of intrauterine life, including neural tube defects, intrauterine fetal distress and following fetal death (Aliau et al., 1973; Brock and Sutcliffe, 1972; Brock and Scrimgeour, 1972; Nørgaard-Pedersen et al., 1975; Weiss et al., 1978; Milunsky et al., 1980).

2.1.1 PURIFICATION AND PHYSICOCHEMICAL PROPERTIES OF AFP

Purification of AFP was achieved by different methods including column chromatography on Sephadex C-50, hydroxyapatite, DEAE-Sephadex A-25, electrophoresis on polyacrylamide gels and affinity chromatography on immunoadsorbent columns of Sepharose coupled with rabbit antibodies against rat adult serum proteins.

The examination on analytical polyacrylamide-gel electrophoresis of AFP from rat amniotic fluid revealed a microheterogeneity of this protein composed of two bands: α1-fetoprotein, moving slowly and corresponding to about two-thirds of the total protein, and α2-fetoprotein moving fast (Versee and Barel, 1978a,b). This property of AFP was already reported in fetal serum of human (Alpert et al., 1972) and rat (Aussel et al., 1973).

The molecular weight of AFP is around 70 000 Da and is similar for the two forms. A molecular weight of 65 000 Da was found in human fetal plasma and of 70 000 Da in the plasma of hepatoma patients and fetal mice (Nishi, 1970; Watabe, 1974).

Further studies using Ricinus communis agglutinin fractionation (Kerckaert et al., 1977) and concanavalin-A-Sepharose (Soloff et al., 1976) revealed the presence of at least nine molecular variants of AFP.

Guinea-pig AFP can be separated into three electrophoretic variants in nondenaturating polyacrylamide gel with respective isoelectric points of 5.0, 5.12 and 5.54 (Gourdeau and Belanger, 1983).

AFP has a tendency to aggregate in a dimer form with a molecular weight of 140 000 Da or in trimeric polymer. All of these oligomeric forms are a consequence of the experimental conditions (Ruoslahti and Seppala, 1971; Yachnin et al., 1977).

AFP has a sedimentation coefficient of 4.5–4.8 S, an isoelectric point (pI) of 4.7–5.0, a diffusion constant (D20,w) of 5.7–6.6 (10−7 cm²/s) and an absorbance E¹%1 cm of 4.15–5.30 (at 278 nm).

2.1.2 CHEMICAL COMPOSITION OF AFP

About 93% of the molecule 65 000 Da, corresponds to the peptide portion and 5300 Da to carbohydrates. Studies on the amino acid composition revealed a similarity (qualitatively and quantitatively) among different species: in the human, rat and mouse, as well as in the chicken (Table 1.13). The carbohydrate fraction contains mainly hexose, hexosamine and sialic acid (Table 1.14).

TABLE 1.13

Amino Acid Composition of α-Fetoprotein in Different Animal Species (expressed as mol/1000 moles)

Quoted from:

¹Nishi (1975)

²Watabe (1974)

³Hassoux et al. (1977)

⁴Ido and Matsuno (1982).

TABLE 1.14

Carbohydrate Composition of α-Fetoprotein in the Fetal Serum of Different Species (in percentage of total composition)

Quoted from

¹Ruoslahti and Seppala (1971)

²Aliau et al. (1978)

³Versee and Barel (1978a,b)

⁴Zimmerman et al. (1976).

2.1.3 BINDING OF AFP TO ESTROGENS

Although the presence of AFP has been described in the fetal serum of most mammalian species, the protein binds estrogens specifically and with high affinity only in the rat and mouse (Soloff et al., 1971; Nunez et al., 1971b; Raynaud et al., 1971; Savu et al., 1974b). No specific binding of AFP to these hormones was found in the fetal plasma of human (Swartz and Soloff, 1974), cow, rabbit, chicken (Attardi and Ruoslahti, 1977) or guinea-pig (Pasqualini et al., 1976; Gourdeau and Belanger, 1983). In the rat, the binding affinity for estrogens is higher at 5 days of age and disappears at 3–4 weeks of age, and in the mouse maximal affinity values are found in the 18-day-old fetus. The [³H]-estrogen-AFP complex has a sedimentation coefficient of 4.5 S. Maximal values are found in the plasma of 20-day-old rat fetuses (Raynaud et al., 1971). There is one molecule of estrogen bound per AFP molecule (Aussel and Masseyeff, 1977; Versee and Barel, 1978b).

Affinity constants for the binding of different estrogens to rat AFP are indicated in Table 1.15. The affinity for estrone is higher than that for estradiol and that for diethylstilbestrol is very weak. This difference in binding of AFP to natural and synthetic estrogens is extensively used to differentiate AFP from the estrogen receptor protein, which in general binds most estrogens with high affinity. Another synthetic estrogen, RU-2858, which binds the estrogen receptor with very high affinity shows no specific binding to AFP (Raynaud, 1973). This estrogen was used recently to characterize estrogen receptors in the fetal uterus and ovary of the rat (Nguyen et al., 1988).

TABLE 1.15

Association Constants (Ka) and Competition of Estrogens and Other Steroids for Estradiol Bound to Rat α-Fetoprotein

Quoted from Laurent et al. (1975); Aussel and Masseyeff (1978).

The affinity constant does not vary significantly with temperature. For example the estrone–AFP complex has at 5°C a Ka of 3.0 × 10⁸ M−1; at 23°C, 2.4 × 10⁸ M−1 and at 37°C, 1.7 × 10⁸ M−1 (Aussel and Masseyeff, 1978). The data indicate the thermostability of the AFP–estrogen complex, which is in opposition to that for the estrogen–receptor complex which is very sensitive to temperature. The association rate constants (k+ 1) for estrone–AFP and estradiol–AFP complexes are respectively: 1.4 × 10⁶ and 1.1 × 10⁶M/s and the dissociation rate constants (k–1): 3.1 × 10−3 and 4.6 × 10−3 s−1 (Keel and Abney, 1984).

An interesting observation was made by Benassayag et al. (1977, 1979) that nonesterified unsaturated fatty acids can compete with estradiol for the AFP binding site. These fatty acids include linoleic, oleic and arachidonic acids (Aussel and Masseyeff, 1983a,b). The inhibitory effect of unsaturated fatty acids on the interaction with the estrogen–AFP complex is dose-dependent. It is suggested that these fatty acids, binding in the vicinity of the estrogen binding sites, can release the hormone and induce specific cellular responses.

AFP can also bind tryptophan methylester and related compounds with high affinity which compete with estrogens. The binding is stereoselective and pH-dependent, suggesting that the protease substrate binding site on AFP is spatially close to the estrogen binding site (Baker et al., 1980). The substitution of p-nitrophenyl for the methyl group in the acetyltryptophan methyl ester results in a 10⁵ increase in affinity for AFP, consequently N-benzyloxycarbonyl-tryptophan p-nitrophenyl ester is bound to AFP with a Kd of 3.9 × 10−9 M (Baker et al., 1982).

2.1.4 BIOSYNTHESIS AND CONTROL OF AFP

AFP is synthesized in the embryonic liver (Gitlin and Boesman, 1967) of different species and in the yolk sac (Gitlin and Perricelli, 1970). Kekomaki et al. (1971) demonstrated that perfusion of isolated liver of 14–20-week-old human fetuses resulted in a release of AFP of 19–26 μg/min. Using the autoradiography method and labeled estrogens, Uriel et al. (1973) demonstrated the intracellular localization of AFP in the liver of fetal and newborn rats. In the fetal liver the synthesis of AFP is localized in a small population of the parenchymal hepatocytes (Tuczeck et al., 1981). This origin of AFP was confirmed by the presence of AFP-mRNA in the yolk sac (Miura et al., 1979) and in the fetal liver of the mouse (Koga et al., 1974).

A variant of AFP with a molecular weight of 65 000 daltons was reported by culture of a mutant-transformed rat fetal liver cell line (the SV 40 tsA). This protein is encoded by a mRNA of 16 S, while the mature AFP is encoded by a mRNA of 20 S (Yang Chou and Savitz, 1986). The data indicated that transcriptional regulation is responsible for the changes in AFP in transformed cells.

Comparison of the primary amino acid sequences of albumin and AFP of several mammalian species revealed the presence of three closely related domains with identical structure (Gorin et al., 1981). The data support the hypothesis that these two proteins arose in evolution as the consequence of a duplication in a common tripartite ancestral gene (Kioussis et al., 1981). All this information leads to the conclusion that there exist different variations of AFP which depend on the animal species and the experimental conditions for their biosynthesis.

Different toxic agents provoke a significant increase in AFP; for instance, administration of carbon tetrachloride to rats during liver regeneration stimulates the production of AFP five-fold (Aussel et al., 1980). Hepatochemical carcinogens also produce an important increase in AFP production, e.g. N-Z-fluorenylacetamide (Sell et al., 1981) and 3′-methyl-4-dimethyl-amino-azo-benzene (Woods, 1983; Yang Chou and Savitz, 1986).

Steroid hormones, particularly estrogens, can play an important role in the control of AFP levels; this was demonstrated after a dose of 10 mg of estriol administered to adult mice increased the plasma concentration from 20 ng/ml (nontreated animals) to 12 500 ng/ml after 5 days of treatment (Kotani et al., 1987). The data could be related to the fact that the administration of large doses of estrogen to adult mice provokes an intense proliferation of hepatocytes (Fujii and Kotani, 1986). On the other hand, glucocorticoids (e.g. dexamethasone) can suppress serum AFP levels (Gourdeau and Belanger, 1983).

2.1.5 AFP AND THE NERVOUS SYSTEM

Using immunohistochemical methods, studies by various groups demonstrated the intracellular localization of AFP in the neural crest and neural tube derivatives of mammals (Benno and Williams, 1978; Mollgard et al., 1979; Pineiro et al., 1979; Trojan and Uriel, 1979; Uriel et al., 1981b) and in birds (Moro and Uriel, 1981), during the period of their differentiation. In the rat the localization of AFP in brainstem nuclei and intracranial ganglia precedes that of the cerebral cortex and hippocampus (Trojan and Uriel, 1980).

Studies in vivo (Pineiro et al., 1982; Villacampa et al., 1984; Trojan and Uriel, 1986) and in vitro (Schachter and Toran-Allerand, 1982) strongly suggest that the intracellular presence of AFP in developing neurons results from exogenous protein uptake rather than in situ production. The data of the selective accumulation of AFP in the fetal nervous system after injection of labeled AFP ([¹²⁵I]-AFP) into the maternal compartment (see Fig. 1.3(a)) gives support to the hypothesis that the presence of AFP in the developing nervous system of mammals and birds is primarily of exogenous origin (Villacampa et al., 1984). These authors observed that the maximum uptake of radioactivity in the rat is found in the fetal brain before day 16 of fetal development and rapidly declines with the progress of gestation, and they suggest that maternal blood AFP, after crossing the placental barrier, enters into the fetal circulation and accumulates in the cerebrospinal fluid due to the high permeability for AFP (and other serum proteins) of the immature choroid plexus (Mollgard et al., 1979).

FIG 1.3 Radioactivity Ratios (in %) of Brain to other Tissues of Fetal and Post-natal Rats 4 h (ˆ) and 24 h (•) after Administration of [¹²⁵I]-α-Fetoprotein Quoted from Villacampa et al. (1984) with the permission of Developmental Brain Research.

Villacampa et al. (1984) observed a bimodal pattern when plotting the radioactivity ratios of brain to other tissues after administration of [¹²⁵I]-AFP (see Fig. 1.3(b–d)). They suggest that this corresponds to two different periods of neural growth and differentiation (day 16 of fetal development and day 8 post-natal) which can be associated with regional areas of brain development, and they concluded that the preferential localization of AFP in a given area is dependent on the maturity of the area at the time of the observation. Kovaru et al. (1985) found that the maximal intracellular localization of AFP in the brain of the fetal pig is found in the middle of gestation, whereas in the fetal thymus the highest values are found at the end of gestation. Since AFP is present during the organizing process of neural differentiation, it is concluded that this protein can play an important role during the different steps of this period.

2.1.6 AFP CONCENTRATIONS DURING DEVELOPMENT

In humans and various animal species, AFP appears very early in fetal development, increases significantly to a maximal concentration during this period and decreases rapidly during the perinatal phase. The rat is the only species that conserves high levels of AFP during a relatively long post-natal period.

The concentration of AFP in biological fluids (plasma, amniotic fluid) or in organs is currently evaluated by radio-immunoassay (RIA) using an anti-AFP antibody prepared by immunization of rabbits with purified human AFP. A sensitivity of 0.1–1 ng/ml can be obtained using highly diluted antibodies. AFP can also be measured with a great variety of other methods including Immunoelectrophoresis, double diffusion, electroimmunoosmophoresis, immunoautoradiography, latex-agglutination, passive hemagglutination, enzyme-immunoassay; however, RIA is one of the most sensitive and practical techniques (for details see Caballero et al., 1977; Delpre and Gilat, 1978; Wong et al., 1979; Brummund et al., 1980; Gardner et al., 1981; Yamamoto et al., 1986). Very sensitive determinations of AFP were obtained with RIA using monoclonal antibodies (Nomura et al., 1983).

2.1.6.1 In Humans

In fetal serum, AFP achieves a maximum concentration of 2–3 mg/ml at 14 weeks of intrauterine life and declines gradually to 15–100 ng at birth and further to a mean normal adult level of 2–3 ng/ml by two years old (Seppala and Ruoslahti, 1976; Sykes and Dennis, 1977). In amniotic fluid, considerably lower concentrations of AFP parallel those of fetal serum with a significant decrease during the third trimester of pregnancy (Fig. 1.4).

FIG 1.4 α-Fetoprotein (AFP) in Fetal and Maternal Sera and in Amniotic Fluid during Human Pregnancy. Quoted from Seppala (1975); Caballero et al. (1977) and Ruoslahti et al. (1978).

In the maternal serum, AFP increases progressively as pregnancy advances (Fig. 1.4). The evaluation of AFP in maternal plasma, as well as in amniotic fluid, is of particular importance as high values of AFP in the fetus reflect an open neural tube or certain other birth defects (see the Introduction of this section).

The possible correlation between fetal sex and AFP levels has been studied by different authors with conflicting results: Sowers et al. (1983), between 16 and 19 weeks of gestation, and Lardinois et al. (1972), during the third trimester, found higher values of AFP in mothers bearing a male than those bearing a female fetus, whereas Milunsky et al. (1980) found no differences in maternal serum AFP levels between 12 and 32 weeks of gestation when comparing the two sexes.

Caballero et al. (1977) observed that at birth the fetal