| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Reproductive Endocrinology Center, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco, California 94143-0556
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. DevelopmentIII. RegulationIV. PhysiologyV. SummaryReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. DevelopmentIII. RegulationIV. PhysiologyV. SummaryReferences |
|---|
Development and function of the primate fetal adrenal cortex are distinct from those in other species. During the latter two thirds of gestation in humans and higher primates, the fetal adrenal glands are disproportionately enlarged and exhibit extraordinary growth and steroidogenic activity in a specialized cortical compartment known as the fetal zone (4, 5, 6, 7). As its name implies, the fetal zone exists only during fetal life; it atrophies soon after birth and has no counterpart postnatally. During midgestation, the fetal zone occupies 80–90% of the cortical volume and produces 100–200 mg/day of the androgenic C19 steroid, dehydroepiandrosterone sulfate (DHEA-S), which is quantitatively the principal steroid product of the primate fetal adrenal gland throughout gestation. The primate fetal adrenal cortex also produces cortisol, which promotes the maturation of fetal organ systems, including the lungs, liver, thyroid, and gut, needed for extrauterine life (1). In some species (e.g. sheep, goats, and rabbits), cortisol produced by the fetal adrenals also regulates the timing of parturition (2, 8); however, a similar role in primates is not apparent.
Experiments of nature in humans (e.g. anencephaly) (9, 10, 11) and studies in pregnant rhesus monkeys (12, 13) indicate that ACTH secreted from the fetal pituitary is the principal trophic regulator of the fetal adrenal cortex. However, several observations indicate that ACTH may not be acting directly. During the latter two thirds of gestation, the fetal zone grows rapidly and produces large amounts of steroid even though circulating ACTH concentrations may be decreasing (14). Furthermore, soon after birth the fetal zone rapidly involutes although exposure to ACTH continues. Thus, other factors, possibly specific to the intrauterine environment, appear to play a role in the regulation of fetal adrenal cortical growth and function. Substances produced by the placenta (e.g. CG) have been implicated (10, 15), and peptide growth factors produced locally within the fetal adrenal (16) are thought to influence fetal adrenal cortical growth and function by mediating and/or modulating the tropic actions of ACTH. Specific nuclear receptor transcription factors also appear to be important regulators of adrenal cortical development by influencing early embryonic differentiation of adrenal cortical progenitors and the maintenance of steroidogenic function (17, 18). Thus, regulation of the primate fetal adrenal cortex is a complex process involving the net effect of a cohort of factors that modulate and/or mediate the trophic actions of ACTH.
Our purpose is to review the literature and synthesize the current understanding of the developmental and functional biology of the primate fetal adrenal cortex. In particular, we will discuss the recent literature concerning the role of growth factors and nuclear receptors and factors emanating from the placenta in the regulation of fetal adrenal cortical growth and function. We will also address recent advances in understanding the function of the fetal adrenal cortex in the regulation of fetal development and parturition.
|
II. Development |
|---|
|
TopAbstractI. IntroductionII. DevelopmentIII. RegulationIV. PhysiologyV. SummaryReferences |
|---|
Development of the primate fetal adrenal cortex differs qualitatively
and quantitatively from that of other species and is characterized by
extraordinarily rapid growth, high steroidogenic activity, and a unique
morphological appearance. For much of gestation, the human fetal
adrenal cortex is composed of two morphologically distinct zones: the
fetal zone and the definitive zone (Fig. 1
). The fetal
zone accounts for the bulk (80–90%) of the cortex and is the primary
site of growth and steroidogenesis. The definitive zone (also referred
to as the adult cortex, neocortex, or permanent zone), occupies the
remainder of the cortex and comprises a narrow band of tightly packed
cells surrounding the fetal zone. Excellent studies of the early
embryonic and fetal development of the human adrenal cortex have been
reported (4, 20, 21, 22, 23, 24, 25). The following description is synthesized from
these works.
|
By the eighth week of gestation, the mass of cells migrating into the adrenal blastema organize into anastomosing cords and exhibit ultrastructural characteristics consistent with steroidogenic capability. These cells eventually differentiate into large polyhedral cells and become the primordium of the fetal zone. Uotila (22) reported that the definitive zone is derived about a week later when a separate population of cells in the same area of celomic epithelium migrates to the adrenal blastema and surrounds the primordial fetal zone. Although Crowder (23) also proposed that the fetal and definitive zone are derived from separate populations of cells in the celomic epithelium, that study concluded that the anlage of both zones develops simultaneously in the celomic epithelium and enters the adrenal primordium at the same time but in discrete juxtaposed columns. Conversely, Jirasek (4), concluded that both zones are derived from a single progenitor population and that the entire adrenal blastema is initially made up of small primitive epithelial cells. The cells in the center of the adrenal blastema then differentiate into fetal zone cells, whereas the peripheral cells maintain their small blastematous morphology and eventually form the definitive zone. Despite these differences in the dynamics of fetal and definitive zone formation, it is clear that by the eighth week of human pregnancy the fetus acquires a rudimentary but distinct adrenal cortex made up of two zonal compartments.
At around the ninth week of gestation, the adrenal blastema is completely enclosed by the adrenal capsule, which is composed of specialized mesenchymal cells migrating from the area of Bowman’s capsule. At the same time, an extensive network of sinusoidal capillaries develops between the cords of the fetal zone. This vasculature predominates in the central portion of the fetal zone and persists throughout fetal life (26). Consequently, the adrenal cortex is one of the most highly vascularized organs in the primate fetus. Abundant vascularization is likely required to facilitate access of hormonal products to the circulation.
The medulla is absent from the primate fetal adrenal as a discrete structure throughout most of gestation except for small islands of chromaffin cells scattered through the body of the cortex. Only after the involution of the fetal zone during the first postnatal week do the chromaffin elements coalesce around the central vein and begin to form a rudimentary medulla. By the fourth postnatal week, essentially all of the chromaffin cells have clustered in the center of the gland. However, it is not until 12 to 18 months that the medulla becomes adult-like in appearance (23).
B. Fetal adrenal development
After 10–12 weeks of gestation, the morphology of the adrenal
cortex remains relatively constant. By midgestation (16–20 weeks), the
fetal zone clearly dominates and is composed of large (20–50 µm)
eosinophilic cells that exhibit ultrastructural characteristics typical
of steroid-secreting cells (i.e. large amounts of tubular
smooth endoplasmic reticulum, mitochondria with tubulovesicular
cristae, large Golgi complexes, abundant lipid, and numerous dense
bodies). In the outer regions of the fetal zone, the cells are arranged
in tightly packed cords. However, the cells in the central portion are
more widely spaced into a reticular pattern and separated by many
vascular sinusoids. Clusters of immature neuroblasts that will
aggregate eventually into a functional medulla are also present between
the innermost fetal zone cells (26).
The definitive zone is composed of a narrow band of small (10–20 µm) tightly packed basophilic cells that exhibit structural characteristics typical of cells in a proliferative state (i.e. small cytoplasmic volume containing free ribosomes; small, dense mitochondria with lamelliform cristae and scant lipid). Its inner layers form arched cords that send finger-like columns of cells into the outer rim of the fetal zone. Although definitive zone cells are lipid-poor during midgestation, they accumulate some cytoplasmic lipid and begin to resemble steroidogenically active cells with increasing age. By late gestation, the definitive zone cells may be likened to cells of the adult zona glomerulosa (26).
Ultrastructural studies have also demonstrated a third zone between the
fetal and definitive zones, the cells of which have intermediate
characteristics (27). We have referred to this cortical area as the
‘transitional zone’ (28) (Fig. 1
). Studies in our laboratory
(discussed below) indicate that after midgestation, transitional
zone cells may have the capacity to synthesize cortisol and thus be
analogous to cells of the zona fasciculata of the adult adrenal. By the
30th week of gestation, the definitive zone and transitional zone begin
to take on the appearance of the zona glomerulosa and the zona
fasciculata, respectively (19). Thus, by late gestation the fetal
adrenal cortex resembles a rudimentary form of the adult adrenal
cortex.
C. Neonatal adrenal development
Soon after birth, the primate adrenal cortex dramatically
remodels. Keene and Hewer (20) in 1927 reported that during the first 6
weeks of postnatal life "... The whole gland shrinks owing
to the rapid disappearance of the foetal cortex... " The
demise of the fetal zone was thought to occur by necrosis and
hemorrhage. However, more recent studies have suggested that the
necrosis and hemorrhage were probably artifacts related to the cause of
death and time elapsed before tissue fixation. Detailed studies by
Sucheston and Cannon (19) in humans and by McNulty (29) in rhesus
monkeys, demonstrated that the postnatal remodeling of the primate
adrenal cortex involves a complex wave of differentiation such that the
fetal zone atrophies and the zonae glomerulosa and fasciculata develop.
Recent studies by Spencer et al. (30) indicate that fetal
zone remodeling in the human is an apoptotic process.
It has generally been thought that the adult cortical zones develop from the persistent definitive zone. However, there is no evidence of adrenal cortical insufficiency during the perinatal period and the postnatal remodeling process. Thus, it is likely that the nascent adult cortical zones are present and functional before birth. Indeed, morphological studies have identified rudimentary zonae glomerulosa and fasciculata during late gestation (19). This lends support to the notion that the postnatal remodeling of the primate adrenal cortex involves apoptosis of the fetal zone and the simultaneous expansion of preexisting zonae glomerulosa and fasciculata.
D. Growth
Rapid growth of the human fetal adrenal cortex begins at
approximately the 10th week of gestation and continues to term (Fig. 2). The growth is almost entirely due to enlargement of
the fetal zone and, as a consequence, the gland becomes as large as the
fetal kidney by 20 weeks. Between 20 and 30 weeks, the size and weight
of the fetal adrenal gland doubles, achieving a relative size 10- to
20-fold that of the adult adrenal. A further doubling in fetal adrenal
weight occurs after 30 weeks of gestation such that by term the gland
weighs approximately 3–4 g (4, 20).
|
Centripetal migration of lipid-containing cells from the definitive to the fetal zone was first reported by Keene and Hewer (20) and later confirmed by Crowder (23). Jirasek (4) described the daughter cells resulting from mitoses in the definitive zone forming cords that invade into the outer layers of the fetal zone. The disparate level of proliferation between the definitive and fetal zones and evidence of centripetal migration lend support to the migration theory of adrenal cortical cytogenesis and suggest that the definitive zone is the germinal/stem-cell compartment from which the inner cortical zones are derived. Thus, cells proliferate in the definitive zone and then migrate inward to form the fetal zone. This concept is supported by studies in the adult rat and mouse showing that mitotic pressure in the periphery of the adrenal cortex causes centripetal migration of cells from the glomerulosa to the inner cortical compartments (33, 34, 35, 36, 37, 38). The adrenal cortical zones therefore appear to be interdependent and derived from a common pool of cells in the periphery. Thus, growth of the fetal zone not only involves limited proliferation of existing fetal zone cells but also the differentiation and hypertrophy of inwardly migrating cells from the definitive zone.
Apoptosis also may occur in the developing human fetal adrenal cortex.
By morphological criteria, Jirasek (4) detected evidence of cellular
apoptosis primarily in the central portions of the fetal zone and
similarly, Spencer et al. (30) using a technique that
identifies apoptotic cell in situ based on the detection of
increased accumulation of cleaved DNA found that the labeling index of
apoptotic nuclei was greater in the central areas of the fetal zone
than in the definitive zone. In contrast, Albrecht et al.
(39) could not detect evidence of apoptosis during early-, mid-, and
late-gestation in the baboon fetal adrenal. However, in that study
apotosis was assessed by measuring that extent to which genomic DNA
extracted from whole baboon fetal adrenals was cleaved into specific
oligonucleosomes (the DNA-laddering method). The low number of
apoptotic cells detected by Spencer et al. (30) in the
central areas of the cortex probably would not contribute enough
cleaved DNA for detection in the laddering assay. Thus, the primate
(and most likely other species’) fetal adrenal cortex is a dynamic
organ in which cells proliferate in the periphery, migrate
centripetally, differentiate to form the specialized cortical
compartments (and possibly continue to proliferate within the
compartments), and then undergo senescence when they reach the center
of the cortex (Fig. 3). The size of the fetal adrenal
cortex and its constituent zones represents the net effect of forces
that modulate these dynamic parameters of growth.
|
1. Ontogeny of steroidogenic activity.
Morphological analyses
indicate that the human fetal adrenal cortex has steroidogenic
capabilities early in gestation. This is first seen at 6–8 weeks when
the cells in the adrenal blastema differentiate and acquire
steroidogenic characteristics (23, 27). In cord blood from human
fetuses between 10 and 20 weeks of gestation, concentrations of
cortisol (40, 41, 42), corticosterone, and aldosterone (40) are greater in
the umbilical artery than in the umbilical vein, indicating that these
steroids are produced by the fetus. Perfusion of previable human
fetuses with radiolabeled progesterone showed that the fetus has the
capacity to produce cortisol from progesterone as early as the 16th
week of gestation (43, 44, 45). Small amounts of aldosterone have been
detected in human amniotic fluid at 9 weeks of gestation (46); however,
it is not certain whether it originates from the maternal or fetal
adrenals at this stage of gestation. At 20 weeks, low levels of
aldosterone can be produced by the previable human fetus perfused with
radiolabeled corticosterone (47, 48).
Estrogens, particularly estriol, in the maternal circulation are indicative of fetal adrenal steroidogenic activity. The placenta, which is the principal source of estrogens during pregnancy, utilizes exogenous androgens supplied to an increasing extent by the fetal adrenal cortex and to a decreasing extent by the maternal adrenal cortex as gestation proceeds (49), as precursors for estrogen synthesis (see Section IV). Low levels of estriol can first be detected in the maternal circulation at the eighth week of gestation, indicating that DHEA-S is being produced by the fetus at this stage. At around the 12th week of gestation, estriol concentrations in the maternal circulation rapidly increase approximately 100-fold (50). This increase coincides with the initiation of fetal zone hypertrophy and ACTH secretion by the fetal pituitary gland (51). In contrast, estriol levels are markedly decreased and sometimes undetectable in women bearing anencephalic fetuses (49, 52), if fetal death occurs (49), or in placental sulfatase deficiency (53, 54). These observations indicate that the human fetal adrenal cortex produces DHEA-S beginning at around 8–10 weeks of gestation in sufficient quantities to effect increases in maternal estrogen levels. Production of DHEA-S by the fetal adrenal cortex continues for the remainder of pregnancy and during the second and third trimesters increases considerably such that by term the human fetal adrenal produces around 200 mg/day (49).
A major unanswered question regarding function of the primate fetal adrenal cortex is the stage of gestation at which the fetal adrenal cortex begins to produce cortisol. Observations of infants with congenital adrenal hyperplasia (CAH) suggest that the fetal adrenal cortex produces cortisol early in gestation. The most common form of CAH is caused by a deficiency of the 21-hydroxylase enzyme (P450c21) (55) and, as a result, the fetal adrenal cortex cannot synthesize adequate amounts of cortisol. The reduced glucocorticoid negative feedback to the hypothalamus and pituitary leads to a compensatory increase in ACTH secretion, the trophic actions of which cause hyperplasia of the fetal adrenal cortex. The elevated ACTH also increases production of DHEA-S, as biosynthesis of this C19 steroid is not affected by P450c21 deficiency. Consequently, the primary manifestations of P450c21 deficiency are those of androgen excess, which are first expressed in utero, resulting in the masculinization of the external genitalia in female fetuses. Inasmuch as differentiation of external genitalia in both sexes begins at week 7 of gestation and is complete by week 10 (Ref. 56 for review), these effects of androgen excess in P450c21 deficiency most likely occur before week 10 of gestation. These observations imply that, under normal circumstances, the fetal adrenal produces cortisol early in gestation which exerts negative feedback control on fetal pituitary ACTH secretion. Moreover, these observations indicate that the human fetal pituitary produces ACTH before 10 weeks of gestation, which regulates fetal adrenal steroidogenesis. Immunohistochemical studies demonstrate the presence of corticotropes in the human fetal pituitary at about this time (51). It is conceivable that cortisol is produced by the fetal adrenal early in gestation to suppress ACTH secretion and prevent overproduction of adrenal androgens during the androgen-sensitive phase of sexual differentiation (particularly in females).
Studies of human fetal adrenal tissue in vitro have provided
more direct evidence of its steroidogenic capability early in
gestation. Incubations with labeled acetate (57) or progesterone (58, 59) and superfusion of human fetal adrenal tissue in vitro
with media containing ACTH (60) indicate that the human fetal adrenal
cortex is responsive to ACTH and produces corticoids and DHEA-S as
early as the tenth week of gestation. The pattern of glucocorticoid
production by the primate fetal adrenal cortex during the rest of
pregnancy is not clear. Expression of key steroid-metabolizing enzymes
suggests that the human fetal adrenal cortex does not produce cortisol
de novo from cholesterol until around week 30 of gestation
(see below). However, this does not preclude the possibility that
cortisol is produced utilizing progesterone as precursor early in
gestation. MacNaughton et al. (45) infused radiolabeled
progesterone into previable human fetuses between 16 and 18 weeks of
gestation and found that it was metabolized to cortisol and that the
rate of metabolism was increased by ACTH. The abundant amount of
progesterone produced by the placenta provides the fetal adrenal cortex
with a potential source of 
4-C21 steroid
substrate for cortisol and aldosterone production even though the
adrenal may not be sufficiently mature to produce these steroids
de novo.
Mineralocorticoid production by the primate fetal adrenal cortex is very low early in gestation but increases during the third trimester. At term, 80% of the aldosterone in human and rhesus monkey fetal blood appears to originate from the fetal adrenal (47, 61). In 18- to 21-week human fetal adrenals, the mineralocorticoid metabolic pathway is localized to the definitive zone, although its activity is very low and unresponsive to secretagogues (62). The angiotensin-II receptors, AT1 and AT2, are present on human fetal adrenal cortical cells after 16 weeks of gestation (63). The AT2 receptor is localized mainly on definitive zone cells, whereas the AT1 receptor is detectable to a lesser extent in cells from both fetal and definitive zones. Thus, during the first and second trimesters, the ability of the human fetal adrenal cortex to synthesize mineralocorticoids is minimal even though the cells express angiotensin-II receptors.
2. Functional zonation and ontogeny of steroidogenic enzyme
expression.
The contributions of the fetal and definitive zones to
human fetal adrenal steroidogenic potential have been studied in
vitro using tissue perfusion and cell culture techniques (60, 64, 65). In response to ACTH, fetal zone cells produce mainly DHEA-S and
little cortisol, whereas definitive zone cells produce mainly
corticoids and little DHEA-S. It was concluded that, during
midgestation, the fetal zone is the site of DHEA-S synthesis and that
the definitive zone is the site of cortisol synthesis (60).
Another approach to assessing the steroidogenic potential of adrenal
cortical zones has been to determine which steroidogenic enzymes are
expressed by the cells of each zone. As the fate of pregnenolone
metabolism is determined by the branch-point enzymes cytochrome P450
17
hydroxylase/17,20 lyase (P450c17) and 3ß-hydroxysteroid
dehydrogenase/4–5 isomerase (3ßHSD), the
steroidogenic potential of cells may be inferred by the pattern of
expression of these two enzymes. Thus, if both enzymes are expressed,
the cells have the potential for cortisol production, whereas if
3ßHSD is expressed and P450c17 is lacking, steroidogenesis would be
directed toward mineralocorticoid synthesis, and conversely, if P450c17
is expressed and 3ßHSD is lacking, steroidogenesis would be limited
to the 5 pathway and be directed toward DHEA-S
synthesis. Thus, expression of 3ßHSD is particularly relevant for the
de novo synthesis of mineralocorticoids and glucocorticoids,
and expression of P450c17 is essential for C19 steroid
(e.g., DHEA) biosynthesis.
The metabolism of radiolabeled progesterone to cortisol by previable
human fetuses indicates that the fetal adrenal cortex expresses the
enzymes downstream from 3ßHSD [i.e. P450c17,
21-hydroxylase (P450c21), and 11ß-hydroxylase (P450c11)] needed for
cortisol synthesis early in gestation. However, studies in the late
gestation fetal rhesus monkey (66) and baboon (67) show that the
conversion of placental progesterone to cortisol by the fetus is
minimal. Expression of 3ßHSD by the primate fetal adrenal cortex is
therefore the critical step in the metabolism of pregnenolone, as it
confers on cells the ability to convert 5-3ß
hydroxysteroids to 4–3-ketosteroids essential for
mineralocorticoid and glucocorticoid production. Data regarding the
expression of 3ßHSD by the human fetal adrenal cortex early in
gestation are conflicting. Goldman et al. (68) first
indicated that 3ßHSD activity is present in the human fetal adrenal
cortex as early as 12 weeks of gestation and that it is localized to
the definitive zone. However, the specificity of their assay is
uncertain. Voutilainen et al. (69) examined 3ßHSD
expression using the highly sensitive technique of RT-PCR and could not
detect mRNA encoding 3ßHSD (either type I or type II) in human fetal
adrenals between 12 and 22 weeks of gestation but could detect its
expression after 22 weeks. In contrast, Parker et al. (70),
using immunohistochemistry, detected 3ßHSD staining exclusively in
the definitive zone between 11 and 15 weeks of gestation. However,
between 15 and 24 weeks, 3ßHSD expression in the human fetal adrenal
cortex decreased to undetectable levels and after 24 weeks was again
detectable in the definitive zone. Thus, although masculinization of
the female fetus with P450c21 deficiency indicates glucocorticoid
production by the fetal adrenal cortex before 10 weeks of gestation,
evidence of its de novo synthesis (based on 3ßHSD
expression) is uncertain.
The steroidogenic potential of each cortical zone in the primate fetal adrenal also has been examined by determining the in situ pattern of cholesterol side chain cleavage (P450 scc) and P450c17 and 3ßHSD expression during midgestation (16–24 weeks) in the human and late-gestation rhesus monkey fetal adrenal by in situ hybridization and immunocytochemistry (28). In those studies, the late-gestation fetal rhesus monkey was used as a model for the late-gestation human fetus, as the growth, structure, and function of its adrenals closely resemble those of the human (26). Moderate levels of P450scc expression were detected in all cortical zones of the human and rhesus monkey fetal adrenals at all gestational ages. Consistent with other studies (70, 71, 72), 3ßHSD was not expressed in any cortical zone in midgestation, i.e. 16–22 weeks. At 22–24 weeks, 3ßHSD staining was detected in the outermost layer of definitive zone cells. Dupont et al. (72) and Parker et al. (70) showed that by 28 weeks, this staining extended inward to encompass all of the definitive and transitional zone cells (cells at the interface between the fetal and definitive zones). At no time in gestation was 3ßHSD expression detected in the fetal zone. This ontogenetic pattern of 3ßHSD expression suggests that the human fetal adrenal cortex cannot synthesize cortisol de novo between 16 and 22 weeks because it cannot convert pregnenolone to progesterone due to the lack of 3ßHSD.
Interestingly, expression of P450c17, although highly abundant in the transitional and fetal zones, was lacking in the definitive zone at all gestational ages (28, 63, 71). The lack of P450c17 in the definitive zone was unexpected and implies that these cells cannot synthesize cortisol in vivo either de novo or from progesterone. Expression of P450c17 was highest in the transitional zone cells which, late in gestation, also expressed 3ßHSD. Therefore, the transitional zone may be the site of de novo glucocorticoid production by the human fetal adrenal cortex and may acquire this capability late in gestation when its cells begin to express 3ßHSD.
Recently, Coulter and Jaffe (C. L. Coulter and R. B. Jaffe, unpublished
data) examined the ontogenetic localization of P450c21 and P450c11
expression in the human (13–24 weeks) and rhesus monkey (109 days to
term) fetal adrenal cortex using immunohistochemistry. In the human and
rhesus fetal adrenals at all gestational ages, P450c21 immunoreactive
staining was detected in all of the cortical zones. Staining for
P450c21 was less intense in the fetal zone than in the definitive and
transitional zones and was increased in adrenals from rhesus monkey
fetuses treated with metyrapone, which was administered to increase
endogenous ACTH secretion. P450c11 peptide was detected with a
polyclonal antibody that also detects P450c11AS (aldosterone synthase).
Staining for P450c11 was detected early in gestation only in the
transitional zone. Later in gestation, P450c11 staining was found in
the definitive and transitional zones and was increased by metyrapone
treatment. These data suggest differential regulation of P450c21 and
P450c11 expression in the primate fetal adrenal cortex. The presence of
P450c21 throughout gestation in all cortical zones suggests that this
is not a rate-limiting step in steroidogenesis. The colocalization of
P450c21 and P450c11 in the transitional zone further implicates this
cortical compartment as the site of de novo glucocorticoid
synthesis. The onset of P450c11 in the definitive zone late in
gestation indicates that this zone does not have the capacity to
synthesize aldosterone until near term. The onset of 3ßHSD expression
in definitive and transitional zone cells late in gestation (73),
coupled with the presence of P450c21 and P450c11, suggests that these
cells acquire the capacity to synthesize mineralocorticoids and
glucocorticoids, respectively. Thus, the ontogeny of 3ßHSD expression
in specific zones allows the functional differentiation of the primate
fetal adrenal cortex. Studies in the fetal rhesus monkey in
vivo suggest that the ontogeny of 3ßHSD in the fetal adrenal
cortical zones is regulated by ACTH (73) (Fig. 4).
Future studies of the ontogeny and localization of P450c21 and P450c11
in the human fetal adrenal cortex will provide valuable information and
should help resolve the issue of whether cortisol is produced de
novo or derived from progesterone early in gestation.
|
5-steroid
production, particularly DHEA-S (Fig. 5).
|
). Fetal zone cells can be induced to synthesize
cortisol de novo if they are exposed to a sufficiently high
concentration of ACTH. Thus, it is possible that, although fetal zone
cells exhibit some differentiation when placed under culture
conditions, their expression of 3ßHSD in vitro may be due
to exposure to supraphysiological concentrations of ACTH.
|
Ligand-induced up-regulation of ACTH receptor expression may be an important adaptive process directed toward optimizing adrenal responsiveness to ACTH in concert with physiological requirements for hypothalamic-pituitary-adrenal activity. This is particularly important for the physiological response to stress and the maintenance of metabolic homeostasis in which the adrenals play a pivotal role. Inhibition, by ligand-induced receptor down-regulation, of mechanisms involved in the response to stress would be detrimental, whereas enhancement by ligand-induced up-regulation would be advantageous by permitting a more efficient and rapid response. Whether such a process is advantageous for fetal development is not clear. It is possible that this up-regulation is part of the process by which the fetus prepares for responding to the stresses of delivery and the perinatal period. Clearly, an efficient and functionally responsive fetal adrenal cortex is essential for survival during the perinatal period.
In summary, functional development of the primate fetal adrenal cortex
is a complex process involving the temporal and spatial expression of
specific steroid-metabolizing enzymes and changes in responsiveness to
ACTH. Early in gestation (8–12 weeks), the glands appear to be capable
of cortisol and DHEA-S synthesis. However, it is not clear whether
cortisol is produced from progesterone or whether the full complement
of enzymes necessary for de novo cortisol synthesis are
expressed. Later in gestation, the ontogenetic expression of 3ßHSD,
first in the definitive zone and subsequently in the transitional zone,
appears to reflect the functional maturation of these zones as analogs
of the zonae glomerulosa and fasciculata, respectively. Throughout
gestation, the fetal zone expresses high levels of P450c17 but lacks
3ßHSD, consistent with its continued production of DHEA-S (Fig. 4).
|
III. Regulation |
|---|
|
TopAbstractI. IntroductionII. DevelopmentIII. RegulationIV. PhysiologyV. SummaryReferences |
|---|
A. The fetal pituitary and ACTH
It is not unexpected that the extraordinary growth and
steroidogenic activity of the primate fetal adrenal cortex are
dependent on an intact fetal pituitary gland as it produces ACTH, the
primary tropic regulator of the adrenal cortex, postnatally. Several
observations firmly establish the pivotal role of ACTH secreted by the
fetal pituitary in primate adrenal cortical regulation: 1) Disruption
of hypothalamic-pituitary function in the human fetus (e.g.
in anencephalics or associated with maternal glucocorticoid treatment)
results in failure of the fetal zone to develop beyond the size
attained at approximately the 15th week of gestation (9, 10, 88, 89)
and is associated with dramatically reduced estrogen concentrations in
the maternal circulation (49, 90). 2) Administration of dexamethasone
to normal human fetuses in utero reduces maternal estrogen
levels (91), and dexamethasone treatment of pregnant rhesus monkeys
during late gestation causes atrophy of the fetal zone and a marked
decrease in maternal plasma estradiol and estrone levels and fetal
plasma cortisol levels (12, 13). Similarly, experimental anencephaly
(produced by fetal decapitation at around day 80 of gestation;
term = 165 ± 5 days) in the fetal rhesus monkey also caused
marked atrophy of the fetal adrenal cortex and decreased maternal
estrogen levels (12, 92); 3) In utero administration of ACTH
to normal human fetuses (91) increases maternal estrogen levels and in
anencephalics (89), partially restores adrenal size. 4) In congenital
adrenal hyperplasia, the fetal zone is markedly enlarged, and adrenal
androgen concentrations are elevated to virilizing levels (93). 5) In
fetal rhesus monkeys, administration of ACTH increases fetal cortisol
and maternal estrone concentrations (12), and increased endogenous ACTH
secretion (induced by metyrapone treatment) increases fetal adrenal
cortical growth and advances functional maturation, as assessed by
3ßHSD expression (Refs 32 and 73 and see Fig. 4).
The mechanism by which ACTH regulates fetal adrenal cortical growth and function is not clearly understood. The actions of ACTH are mediated via its interaction with specific receptors on the cell surface of adrenal cortical cells. A human ACTH receptor has been cloned and characterized (94). This receptor is coupled to heterotrimeric guanine nucleotide-binding proteins that activate adenylate cyclase leading to an increase in intracellular cAMP that activates protein kinase A and initiates the cascade of intracellular signaling events. Consequently, the actions of ACTH can be mimicked by cAMP analogs (e.g. 8-bromo-cAMP) or substances that increase adenylate cyclase activity (e.g. forskolin). The signaling events downstream from the activation of protein kinase A have not been characterized. The contribution of other signaling pathways in adrenal cortical steroidogenic activity has been reviewed by Pepe and Albrecht (87).
Although the principal regulator of fetal adrenal cortical development appears to be ACTH, several observations support the concept that human fetal adrenal growth and function also are influenced by factors that act independently from, or in conjunction with, ACTH. These lines of evidence are as follows: 1) human fetal adrenal cortical growth and steroidogenic activity are maximal during mid- to late gestation even though circulating ACTH concentrations in the fetus appear to be decreasing (14); 2) before 10–15 weeks of gestation, adrenal development in anencephalic fetuses (presumably with markedly reduced ACTH) is normal, but thereafter the fetal zone fails to develop and does not exhibit its characteristic growth and steroidogenic activity (9, 10), indicating that early in gestation fetal zone growth and function are independent of ACTH; 3) in contrast, the definitive zone appears normal in anencephalic fetuses despite the absence of ACTH stimulation (9, 10, 89), suggesting that its growth is not dependent on ACTH at any stage in gestation, although its functional maturation appears to be regulated by ACTH (73); 4) the fetal zone rapidly involutes once the newborn is delivered and separated from the placenta despite relatively unchanged exposure to ACTH (29), indicating that fetal zone growth and function are maintained by a factor(s) specific to intrauterine life, and 5) ACTH is not a growth factor per se and is not a mitogen for adrenal cortical cells in vitro (95, 96). However, it stimulates proliferation of adrenal cortical cells in vivo, suggesting that its proliferative actions may be mediated by growth factor(s).
B. Growth factors
Specific growth factors, acting in an autocrine and/or paracrine
fashion, are likely candidates as mediators and/or modulators of the
trophic actions of ACTH on the primate fetal adrenal cortex. This
notion is not without precedent, as several classic growth-promoting
hormones are known to act through the stimulation of local
autocrine/paracrine growth factors. Studies of growth factor
involvement in the regulation of fetal adrenal development have
addressed: 1) the effects of growth factors on proliferation and
function of cultured fetal adrenal cortical cells; 2) growth factor
expression by the fetal adrenals, and 3) the regulation of this growth
factor expression by the fetal adrenals.
1. Basic fibroblast growth factor (bFGF).
Basic FGF is a
peptide mitogen that stimulates the proliferation of mesodermal- and
neurectodermal-derived cells and also is a potent angiogenic and
neurotrophic agent. It is a member of a family of related peptides that
include acidic FGF, FGF-6, FGF-7, keratinocyte growth factor,
hst, and int (97). Four different FGF receptors
have been identified, each with intrinsic tyrosine kinase activity.
Differential mRNA splicing results in multiple isoforms of each
receptor (98).
The mitogenic effects of bFGF on adrenal cortical cells was first observed in the Y-1 mouse adrenal cortical cell line (99). Gospodarowicz et al. (100) and Hornsby and Gill (101) subsequently reported that bFGF is a potent mitogen for primary cultures of bovine adult adrenal cortical cells, and both groups suggested that it mediates the growth-stimulatory actions of ACTH in vivo. Basic FGF was later purified from adult bovine adrenals (102) and shown to be synthesized by cultured bovine adrenal cortical cells (103). More recently, bFGF has been shown to be involved in the compensatory adrenal growth response to unilateral adrenalectomy in the rat (104).
Crickard et al. (105) examined the effect of bFGF on the proliferation of fetal and definitive zone cells from midgestation human adrenals. Proliferation of both cell types was stimulated by bFGF, a finding that was later confirmed by Hornsby et al. (106). Interestingly, bFGF elicited a greater proliferative response (relative to control) in definitive zone cells (4-fold increase) than in fetal zone cells (2-fold increase). This suggests that the definitive zone may be more sensitive than the fetal zone to the mitogenic actions of bFGF and that bFGF may preferentially influence definitive zone development in vivo.
In light of its potent mitogenic actions on human fetal adrenal cortical cells, we hypothesized that bFGF is a local mediator of the stimulatory effects of ACTH (16). Therefore, we determined whether the human fetal adrenals express bFGF and, if so, whether this expression is regulated by ACTH. Basic FGF bioactivity and mRNA were detected in protein and total RNA extracts, respectively, from midgestation human fetal adrenals. Messenger RNA encoding bFGF also was detected in cultured human fetal adrenal cortical cells and interestingly, its abundance was increased 2- to 3-fold by ACTH. Thus, bFGF, a potent mitogen for human fetal adrenal cortical cells, is expressed by these cells and regulated by ACTH. Therefore, bFGF may be an important mediator of ACTH action in human fetal adrenal development. It is of interest that bFGF also is a potent angiogenic factor (107) and that the adrenal cortex, particularly the fetal zone, is one of the most highly vascularized organs in the human fetus (26, 27). It is possible that bFGF not only affects the growth of the fetal adrenal cortex but also its vascularization and may be an important local mediator of these events in response to ACTH. In this regard, our preliminary data also indicate that vascular endothelial growth factor, a potent stimulator of angiogenesis, may promote vascularization of the developing adrenal cortex (108).
2. Epidermal growth factor (EGF).
The effects of EGF on
primary adrenal cortical cell cultures was first examined by
Gospodarowicz et al. (100) who reasoned that because EGF is
a potent mitogen for granulosa cells (109), and granulosa and adrenal
cortical cells are derived from the same embryonic germ layer,
i.e. the celomic epithelium, it also may be mitogenic for
adrenal cortical cells. Their studies of cultured adult bovine adrenal
cortical cells, however, revealed that EGF was not mitogenic, a finding
that was later confirmed by Hornsby et al. (106). In
contrast to these negative results in bovine adrenal cortical cells,
Crickard et al. (105) found EGF to be a potent mitogen for
cultured fetal and definitive zone cells from midgestation human fetal
adrenals, a finding that was also confirmed by Hornsby et
al. (106). Interestingly, as with their response to bFGF,
definitive zone cells were more responsive to the proliferative action
of EGF than were fetal zone cells, suggesting that EGF (like bFGF)
preferentially regulates definitive zone growth. High-affinity
surface-binding sites, characteristic of EGF receptors, were detected
in both cell types (105). These findings imply that the human fetal
adrenal cortex is a target for EGF (or other EGF receptor ligands).
The effect of EGF treatment on fetal adrenal development in the late gestation rhesus monkey in vivo has been examined (110). Treatment with EGF significantly increased adrenal weight and the width of the definitive zone. Interestingly, Coulter et al. (110) found that cell density in the definitive zone of EGF-treated animals was less than that of controls, indicating that definitive zone enlargement in EGF-treated animals was due to cellular hypertrophy rather than hyperplasia. This finding indicates that EGF stimulated hypertrophy and not hyperplasia of definitive zone cells, an unexpected effect given the potent mitogenic action of EGF on definitive zone cells in vitro. Thus, in vivo EGF may not be a direct mitogen for definitive zone cells but instead may affect its growth by modulating the hypothalamic-pituitary axis and possibly increasing ACTH secretion and/or ACTH responsiveness. EGF has been shown to affect the fetal hypothalamic-pituitary-adrenal axis in other species: in fetal sheep, EGF stimulates secretion of cortisol from the fetal adrenals, corticotropin releasing hormone (CRH) from the hypothalamus, and ACTH from the pituitary (111, 112). Therefore, EGF could affect fetal adrenal development indirectly by modulating the fetal hypothalamic-pituitary axis. This also is suggested by the finding of Coulter et al. (110) that EGF treatment in vivo increased the amount of 3ßHSD protein in definitive and transitional zone cells of fetal rhesus monkeys. Similar effects were reported in fetal rhesus monkeys in which endogenous ACTH secretion was increased by administration of metyrapone (32). Thus, in addition to its potential direct effect on adrenal cortical cell proliferation, EGF also may modulate adrenal growth and functional maturation by affecting the hypothalamic-pituitary-adrenal axis.
EGF is a member of a large family of peptide growth factors that
includes transforming growth factor- (TGF), vaccinia virus growth
factor, amphiregulin, heparin-binding EGF-like factor, and betacellulin
(113). Each of these peptides shares considerable sequence homology
with EGF and, because they all bind to the EGF receptor, have similar
biological activities. Therefore, studies of EGF action on adrenal
development must take into account the multiple ligands for the EGF
receptor and that, although EGF may have effects on cultured cells, it
may not be a biologically significant ligand in vivo. Sasano
et al. (114), aware of this issue, examined the expression
of EGF, TGF, and the EGF receptor in human adult adrenals. They
found that TGF and the EGF receptor were expressed, whereas EGF was
not, and concluded that TGF is the significant locally produced
ligand for the EGF receptor in adult human adrenals. Similar
experiments in human fetal adrenal glands, in which the expression of
EGF, TGF, and the EGF receptor were examined, yielded similar
findings (115). Immunostaining for TGF was greatest in the fetal and
transitional zones and less intense in the definitive zone.
Immunostaining for the EGF receptor was detected in each of the
cortical zones with equal intensity. These data indicate that TGF
may be the predominant locally produced ligand for the EGF receptor in
the developing human fetal adrenal cortex. These observations indicate
that activation of the EGF receptor on human fetal adrenal cortical
cells may be an important component in the regulation of fetal adrenal
development.
3. Insulin-like growth factors I and II (IGF-I and IGF-II).
IGF-I and IGF-II affect growth and function in a wide variety of cell
types and can act as autocrine, paracrine, or endocrine factors (116).
IGF-I (formerly known as somatomedin-C) mediates many of the
somatotropic actions of GH (117). Although the role of IGF-II is less
defined, it is thought to be involved in the regulation of fetal
development because its circulating and tissue levels are highest
during fetal life and decrease postnatally (118). Two IGF receptors,
designated type I and type II, have been identified (119). The type I
receptor is structurally related to the insulin receptor and binds both
IGF-I and IGF-II with high affinity and insulin with lower affinity.
The type II receptor, also known as the mannose-6-phosphate receptor,
binds IGF-II with high affinity but will not bind IGF-I or insulin.
Most of the known actions of IGF-I and -II appear to be mediated via
activation of the type I receptor. Effects mediated via the type II
receptor are not clearly understood, although this receptor is known to
be involved in targeting lysosomal enzymes. The IGF system is made more
complex by the presence of six high-affinity IGF-binding proteins that
associate with the IGF peptides and modulate their biological activity
(120).
The IGF peptides and their receptors have been identified in normal and neoplastic adrenals (121). Growth and differentiated function in several steroidogenic cell types, including granulosa (122, 123), Leydig (124, 125), and adrenal cortical cells (126), are modulated by IGFs. In adult bovine (127) and fetal ovine (128) adrenal cortical cells, IGF-I increases proliferation and enhances the steroidogenic responsiveness to ACTH. IGF-I augments responsiveness of adult bovine adrenal cortical cells to ACTH by increasing ACTH receptors (127). Similar findings were reported by Pham-Huu-Trung et al. (129), who showed that IGF-I modulates steroidogenesis in cultured human adult adrenal cortical cells by enhancing responsiveness to ACTH and the activity of key steroidogenic enzymes, including P450c17. These effects of IGF-I on adrenal cortical cells were most likely mediated via the type I receptor, which has been identified in bovine (126) and human (121, 130, 131) adrenal cortical cells. Penhoat et al. (132) showed that IGF-I is synthesized by adult bovine adrenal cortical cells and that its secretion is enhanced by ACTH, suggesting that it may be a local paracrine/autocrine mediator of ACTH action.
The role of the IGFs in human fetal adrenal development has also been examined. Han et al. (133, 134) studied IGF-I and IGF-II expression in a variety of midgestation human fetal tissues and showed that, consistent with other species, IGF-II is quantitatively the predominant IGF expressed during fetal life. In the fetal adrenals, abundance of mRNA encoding IGF-II was high (second only to the liver), whereas mRNA encoding IGF-I was low. Ilvesmäki et al. (135), using RT-PCR analysis of total RNA extracted from whole adrenals, demonstrated that all of the components of the IGF system (i.e. IGFs, receptors, and binding proteins) are expressed by human fetal adrenals. Interestingly, Han et al. (133), using in situ hybridization analysis, found that mRNAs encoding the IGFs were generally restricted to mesenchymal cells, and in the adrenals were detected only in the capsule. These findings led Han et al. to hypothesize that IGFs produced by the mesenchymal component regulates the growth of associated epithelial elements.
Expression of IGF-I and IGF-II in cultured cortical cells from midgestation human fetal adrenals was first examined by Voutilainen and Miller (136) who found that the high level of IGF-II expression reported by Han et al. (133) in vivo also was present in vitro and could be stimulated by ACTH and factors that increase intracellular cAMP, a finding that we later confirmed (137). This effect of ACTH is inconsistent with IGF-II expression by capsular fibroblasts as suggested by Han et al. (133), insasmuch as fibroblasts are not known to be responsive to ACTH, and their presence in the cell culture preparations was minimal. Instead, the data suggest that IGF-II was expressed by the ACTH-responsive cortical cells. In light of this, we reexamined the localization and regulation of IGF-I and IGF-II expression in midgestation human fetal adrenals (137).
In situ hybridization analysis revealed that IGF-II is expressed by all cortical cells in relatively high abundance, whereas IGF-I is only detectable in the adrenal capsule. Similarly, in cultured human fetal adrenal cortical cells, mRNA encoding IGF-II is highly abundant in the cortical cells and not present in the contaminating fibroblasts, and its abundance in cortical cells is markedly up-regulated by ACTH. In contrast, IGF-I mRNA was not detected in cultured fetal adrenal cortical cells and could not be stimulated with ACTH. These findings recently have been confirmed in the fetal rhesus monkey in vivo in which endogenous ACTH secretion was increased by administration of metyrapone (32). Adrenals of metyrapone-treated fetuses were larger than controls and expressed higher levels of IGF-II but not IGF-I. Taken together, these data strongly implicate IGF-II as an important local regulator of fetal adrenal development and a possible mediator of at least some of the trophic actions of ACTH. Interestingly, cultured adrenal cortical cells from a 6-week human neonate responded to ACTH with increased cortisol production but failed to express IGF-II (138), suggesting that expression of IGF-II by the human adrenal cortex and its regulation by ACTH are unique to fetal life.
The role of the IGFs in human fetal adrenal development was further characterized in studies of their effects on the growth and function of cultured human fetal adrenal cortical cells (137). Both IGF-I and IGF-II are specific mitogens for human fetal adrenal cortical cells. Interestingly, the IGFs act cooperatively with bFGF and EGF, other known mitogens for human fetal adrenal cortical cells (see above), resulting in an additive effect on cell proliferation. That both IGF-I and IGF-II stimulated proliferation in an almost equipotent fashion suggests that their mitogenic actions are mediated through a common receptor, most likely the type-I IGF receptor. In other cell types for which IGF-II is a mitogen, its actions have been found to be mediated via the type-I IGF receptor (139, 140).
In conjunction with its mitogenic activity, IGF-II also affects the
differentiated function of human fetal adrenal cortical cells. In
cultured fetal zone cells, IGF-II augments ACTH-stimulated cortisol and
DHEA-S production (Fig. 7) and ACTH-stimulated
expression of the steroidogenic enzymes P450scc, P450c17, and 3ßHSD
(74, 141) (Fig. 8). As with its mitogenic actions, the
effects of IGF-II on steroid production were mimicked by IGF-I, again
suggesting that the actions of both peptides were mediated by the
type-I IGF receptor. Both IGF-I and IGF-II directly up-regulated basal
expression of P450c17 as assessed by mRNA abundance, but did not affect
basal expression of P450scc or 3ßHSD (Fig. 8). A similar effect of
IGFs on P450c17 activity and expression has been reported in human
adult adrenal cortical cells (129, 142). The increased P450c17 activity
suggests that IGFs may be important regulators of adrenal androgen
production. Activation of the type-I IGF receptor by either IGF-I
(postnatally) or IGF-II (prenatally) may directly augment adrenal
androgen synthetic capacity by augmenting P450c17 expression and
activity. This may be an important mechanism by which adrenal androgen
production is regulated during fetal and postnatal life. Interestingly,
the onset of adrenarche coincides with an increase in circulating IGF-I
(143).
|
|
), indicating that their stimulatory effects on ACTH
responsiveness were exerted at some point distal to ACTH receptor
binding and activation.
4. Activin/inhibins.
Activin and inhibin are homodimeric
(ßA-ßA, ßB-ßB, or ßA-ßB) and heterodimeric (-ßA or
-ßB) glycoproteins, respectively, which are structurally related
to other members of the TGFß family of peptides. Both inhibin and
activin originally were isolated from follicular fluid based on their
ability to modulate FSH secretion from the pituitary (145). Inhibin
suppresses, while activin stimulates, FSH secretion. It is now apparent
that activin’s actions extend beyond the pituitary-gonadal axis to
affect many key biological functions. The amino acid sequence of
activin is strongly conserved between species and is closely related to
that of TGFß, mullerian inhibiting substance, fly decapentaplegic
complex, and bone morphogenesis proteins, all of which are regulators
of development and differentiation of a variety of tissues. Activin may
subserve different functions in the organism depending on the stage of
development. For example, activin induces mesoderm development in
Xenopus during early embryogenesis (146), whereas later in
life it appears to be a significant regulator of ovarian
folliculogenesis (147). A family of activin receptors has been
characterized according to subunit structure. These include the type-I
activin receptor and two homologous type-II receptors (IIA and IIB)
(148).
Activin appears to play a role in the regulation of adrenal cortical
development and function. This is not unexpected as activin has
profound effects on the growth and function of granulosa cells (149),
which originate from the same germ layer as adrenal cortical cells. The
subunits of activin and inhibin (, ßA and ßB) have been
identified in the adrenal glands of the adult rat (150) and the fetal
and adult sheep (151). Interestingly, inhibin knockout mice develop
adrenal tumors, suggesting that inhibin acts as a tumor suppressor gene
in adrenal cortical cells (152). Activin/inhibin subunit localization,
as well as the mitogenic and steroidogenic actions of activin and
inhibin in human fetal and adult adrenals, has been examined (153, 154). Each of the activin/inhibin subunit proteins and their mRNAs were
detected in fetal and adult adrenals by immunohistochemistry. In
midgestation fetal adrenals, specific immunostaining for the
subunits was localized in a scattered pattern in both the fetal and
definitive zones, whereas specific staining for intact activin-A
(ßA-ßA homodimer) was detected predominantly in the definitive and
transitional zones. In cultured fetal adrenal cortical cells, ACTH
stimulated secretion of immunoreactive -subunit. This suggests that
ACTH stimulates inhibin production by fetal adrenal cortical cells, as
the -subunit is only present in the inhibin molecule. ACTH enhances
the abundance of mRNAs encoding the - and ßA-subunits, but not the
ßB-subunit, in cultured fetal adrenal cortical cells. Thus, during
midgestation, the human fetal adrenals express each of the
activin/inhibin subunits and appear to produce immunoreactive activin-A
in the definitive and transitional zones. ACTH stimulates expression of
the - and ßA-subunits, suggesting that fetal adrenal production of
activin and inhibin is under trophic hormone regulation.
In cultured human luteinizing granulosa cells activin stimulates proliferation and steroidogenesis, whereas inhibin has no effect (149). In contrast to its mitogenic affects on granulosa cells, recombinant human activin-A inhibits proliferation of fetal zone cells and, at the same time, increases ACTH-stimulated cortisol production (153, 154). Activin has no effect on DHEA-S production by fetal zone cells or growth and steroidogenesis in definitive zone or adult adrenal cortical cells. Recombinant human inhibin has no effect on proliferation or function of any of the adrenal cortical cell types. Thus, activin acts directly and specifically on fetal zone cells to inhibit their rate of growth and enhance their capacity for cortisol production in response to ACTH. Activin may inhibit fetal zone cell growth by stimulating cellular apoptosis and therefore may be involved in the postnatal demise of the fetal zone, a process involving apoptosis (30). In addition, activin may coordinately stimulate the differentiation of other fetal zone cells into a cortisol-producing phenotype.
5. TGFß.
TGFß is the prototypical peptide of a large
family of growth factor proteins including activin, inhibin, mullerian
inhibiting substance, bone morphogenic protein, and several closely
related proteins designated TGFß2, TGFß3, TGFß4, and TGFß5
(155). Specific receptors for TGFß have been identified on almost all
mammalian cells. The effect of TGFß is variable and appears to be
dependent on the cell type. In general, TGFß stimulates proliferation
of cells of mesenchymal origin and inhibits proliferation of cells of
epithelial or neurectodermal origin (156). TGFß also modulates the
differentiated function of cells and, in particular, has marked effects
on the function of steroid-producing cells. In adult bovine and ovine
adrenal cortical cells, TGFß inhibits basal and ACTH-stimulated
cortisol and agonist-stimulated aldosterone production (157, 158, 159, 160) and
reduces ACTH receptor binding (161). TGFß is produced by, and
interacts with, specific receptors on adult bovine adrenal cortical
cells (162), suggesting that it acts as an intraadrenal
autocrine/paracrine factor.
Several studies have indicated that TGFß is involved in the regulation of human fetal adrenal development. Riopel et al. (163) investigated the effects of TGFß on growth and function of cultured fetal zone cells and found that it significantly inhibits fetal zone cell proliferation but had no effect on steroidogenesis. Subsequently, Spencer et al. (154) confirmed these inhibitory effects of TGFß on fetal zone cell proliferation. Parker et al. (164) later reported that TGFß also inhibits proliferation of definitive zone cells and, in a more detailed study of the effects of TGFß on fetal adrenal steroid production, Stankovic et al. (165) found that both basal and ACTH-, forskolin-, and cAMP-stimulated DHEA-S and cortisol production and expression of P450c17 by fetal and definitive zone cells were inhibited by TGFß. They also showed that TGFß binds to specific sites on human fetal adrenal cortical cells and that these binding sites are regulated by ACTH (166). Lebrethon et al. (167) reported similar findings and, in addition, showed that TGFß has no effects on ACTH receptor and P450scc expression but enhances ACTH-stimulated expression of 3ßHSD, an intriguing finding in light of its inhibitory effects on steroid production. Thus, TGFß inhibits proliferation and steroid production by fetal and definitive zone cells, likely via interaction with specific cell surface-binding sites. Whether TGFß is expressed by human fetal adrenal cortical cells is uncertain. Taken together, these data indicate that TGFß-related peptides (particularly TGFß and activin) are significant negative regulators of human fetal adrenal growth and may play an important role in balancing the positive effects of other growth factors during adrenal development.
C. Nuclear receptors/transcription factors
Nuclear receptors are essential elements in cellular regulation as
they mediate the link between an extracellular signal and the
transcriptional response (Ref. 168 for review). Examples of well
characterized nuclear receptors for which the ligand is known include
the steroid hormone receptors, thyroid hormone receptor, retinoic acid
receptor, and vitamin D receptor. Each of these proteins, when bound to
its cognate ligand, acts on specific sequences of DNA (response
elements) to either promote or inhibit transcription of target genes.
Many transcription factors that share sequence homology (especially in
the DNA- and ligand-binding domains) with classical nuclear receptors
have been identified, but their ligands have not been identified. These
are referred to as "orphan" nuclear receptors. Two of these,
steroidogenic factor-1 (SF-1) and DAX-I, appear to play major roles in
adrenal development and function.
SF-1 is a transcription factor that regulates the expression of genes encoding steroidogenic enzymes [Refs. 169169 and 170170 and see review by Parker and Schimmer (170a)170A in this issue]. Consensus-binding sites for SF-1 have been identified in the promoter regions of genes for most steroidogenic enzymes. In the mouse, SF-1 is expressed in all steroidogenic tissues, including the adrenal gland. Interesting
