. 18
( 51 .)


When adrenalectomized rats are offered sucrose to drink, CRH is lower than in
the animals offered saccharin, and indistinguishable from controls, whereas their
serum insulin is signi¬cantly higher than both groups (Dallman et al., 2003).
94 M. L. Power and S. D. Tardif

Human placental CRH is regulated by cortisol in a similar manner to that of
amygdalar CRH (see chapter by Smith and colleagues for a description). Thus,
increases in maternal or fetal cortisol production are expected to upregulate CRH
messenger ribonucleic acid (mRNA) synthesis.
The available evidence supports the hypothesis that all anthropoid primates
produce placental CRH and most produce CRH-binding protein (CRHbp) during
pregnancy (Bowman et al., 2001). This sets anthropoid primates apart from other
mammals, as, so far, no other mammalian species have been found to produce pla-
cental CRH. Even among anthropoid primates, however, the pattern of placental
CRH and CRHbp production and secretion differs.
CRH mRNA was detected in the placenta, but not in amnion or chorion, in the
rhesus macaque. Levels of CRH peptide and mRNA did not change over the last 18
days of gestation in this species, however, CRH mRNA increased twofold during
both spontaneous and androstenedione-induced labor (Wu et al., 1995). In the
baboon, there is a peak of maternal serum CRH in early-to-mid-gestation, fol-
lowed by a gradual decline. Both maternal and fetal CRH remain elevated until
term, however (Goland et al., 1992; Smith et al., 1993). In the common marmoset,
both CRH and CRHbp are detectable in maternal serum during pregnancy
(Bowman et al., 2001), and the pattern of maternal serum CRH is similar to that
of the baboon. The common marmoset has a long gestation for its body mass (ca.
350 g; term 144 days), however, for the ¬rst 50“55 days post-fertilization there is
little placental or fetal mass accumulation. During this initial quiescent period
of gestation CRH is undetectable in maternal serum, but by approximately 55 days
gestation maternal serum CRH begins to rapidly rise. This rise in CRH is followed
by a rise in maternal cortisol, and shortly thereafter a rise in maternal estradiol.
Maternal serum CRH then declines, but remains detectable throughout gestation
(Figure 3.3; Tardif et al., unpublished data).
Humans share a pattern of exponentially increasing maternal CRH through
pregnancy with our closest relatives, the chimpanzee and gorilla (Smith et al.,
1999). Gorillas would appear to be the most similar to humans (Smith et al., 1999;
Figure 3.4). Maternal CRH levels in chimpanzees are signi¬cantly lower than in
humans or gorillas, and chimpanzees do not show a decline in CRHbp at term,
in contrast to both humans and gorillas (Smith et al., 1999). The exact function of
the early gestational rise in placental CRH production in all anthropoids, and the
signi¬cance of the differences among monkeys, apes and humans are currently
not known.
In humans, placental CRH is secreted into both the maternal and fetal compart-
ments, but cord blood concentrations are signi¬cantly lower than maternal con-
centrations, indicating that it is preferentially secreted into the maternal compartment
(Ruth et al., 1993). This also appears to be true in the rhesus macaque, where
95 Maternal nutrition and metabolic control of pregnancy


Serum CRH concentration (pmol/l) 140







0 20 40 60 80 100 120 140
Estimated gestational age (days)
Figure 3.3 Pattern of maternal serum CRH concentration during pregnancy in the common marmoset

Maternal serum CRH (pmol/l)






40 60 80 100 120 140 160 180 200 220 240 260
Days of gestation

Figure 3.4 The pattern of maternal serum CRH during pregnancy in gorillas. Adapted from Smith
et al. (1999)

fetal CRH concentrations were approximately 1% of maternal concentrations
(Bowman et al., 2001).
Primate placental CRH is as biologically active as CRH produced by the hypo-
thalamus. Placental CRH stimulates ACTH production from the fetal pituitary
gland, which in turn stimulates cortisol and DHEA-S production from fetal adrenal
glands. Placental CRH has been shown to be able to directly stimulate DHEA-S
96 M. L. Power and S. D. Tardif

production from the fetal adrenal glands (Smith et al., 1998). The primate placenta
lacks the enzyme to convert progesterone to estrogen, and instead converts the
androgen DHEA-S to estrogen. In chimpanzees and gorillas, maternal estradiol
and CRH concentrations were highly correlated (Smith et al., 1999), consistent
with the hypothesis that placental CRH drives placental estrogen synthesis through
its stimulation of the fetal adrenal.
In pregnant women, serum CRH is detectable by the end of the ¬rst trimester,
and exponentially rises until parturition (McLean et al., 1995). Concentrations of
maternal serum CRH quickly rise to levels capable of stimulating the maternal
HPA axis (Sasaki et al., 1989). However, through much of gestation the placenta
also produces a CRHbp that inactivates CRH in maternal circulation. In normal
human pregnancy, the concentration of CRHbp decreases in late gestation
(Perkins et al., 1993; McLean et al., 1995). Preterm birth is associated with not only
a premature rise in maternal serum CRH (Goland et al., 1986), but also an early
decline in CRHbp (Perkins et al., 1993).
The prepartum increase in cortisol production by the fetal adrenal is important
for fetal organ maturation, especially of the lungs and kidneys, and also has effects
on the fetal HPA axis and on extra hypothalamic brain regions. Fetal cortisol pro-
duction can stimulate further CRH production from placenta. In humans (Goland
et al., 1994), chimpanzees and gorillas (Smith et al., 1999), serum cortisol and CRH
are correlated. This is consistent with the hypothesis that glucocorticoids drive pla-
cental CRH production via a feed-forward system linking the placenta with the
fetal pituitary and adrenal glands.
Several lines of evidence suggest that events early in pregnancy may set the tim-
ing of birth and that ˜setting™ may be related to the CRH-cortisol axis. Women who
subsequently enter preterm labor not only have elevated serum CRH at mid-
pregnancy, but their rate of increase of CRH is accelerated from early on (McLean
et al., 1995; Leung et al., 2001; Figure 3.5). Opportunistic studies of pregnancy
duration following large man-made disasters, such as the Dutch Famine in 1944“5
(Stein and Susser, 1975), or natural disasters such as earthquakes (e.g. Glynn et al.,
2001) indicate that gestations in the ¬rst trimester during the event are the most
likely to result in preterm delivery.
Recent, suggestive evidence indicates that an early nutritional insult may also
increase the risk of preterm birth in sheep. A study in which 10 ewes were food
restricted from 60 days before to 30 days after conception (achieving a 15% reduc-
tion in maternal weight), with ad libitum feeding thereafter, resulted in signi¬cantly
shorter gestation lengths compared with control ewes fed ad libitum throughout
(Bloom¬eld et al., 2004). The evidence (precocial surges in cortisol and ACTH)
suggested this early maternal energy restriction resulted in early maturation of the
fetal HPA axis. The evidence did not support limited nutrient availability affecting
97 Maternal nutrition and metabolic control of pregnancy


Rate of increase in maternal serum CRH 0.32








Preterm Term
Figure 3.5 The rate of increase in maternal serum CRH concentration is greater in pregnancies
destined to deliver preterm. Adapted from Leung et al. (2001) with permission

fetal growth, as fetal lambs did not differ in size between the two groups. Thus, an
early nutritional insult that did not appear to affect overall fetal growth apparently
programmed an accelerated maturation of the fetal HPA axis.
In another study (Whorwood et al., 2001), energy restriction of ewes for the ¬rst
half of gestation had no effect on length of gestation or birth weight; however, it
did have tissue speci¬c effects on the expression of glucocorticoid receptor (GR)
and 11 -hydroxy dehydrogenase mRNA in the fetuses. Food restriction during
early-to-mid-gestation resulted in increased expression of GR mRNA in fetal organs
(adrenal, kidneys, liver, lung, and perirenal adipose tissue), increased 11 -hydroxy
dehydrogenase type 1 mRNA in perirenal adipose tissue, and decreased 11 -
hydroxy dehydrogenase type 2 mRNA in adrenals and kidney. These differences
persisted until birth and were evident in the lambs even though the plane of nutri-
tion was increased to ˜normal™ for the last half of gestation. In unrestricted ewes,
11 -hydroxy dehydrogenase type 2 was abundant in the placenta at mid-gestation,
though absent at term; the placentas of energy restricted ewes had lower 11 -
hydroxy dehydrogenase type 2 at mid-gestation.

Feeding, fasting, cortisol, and CRH

Even short-term starvation is known to increase glucocorticoid secretion (Dallman
et al., 2003), and energy restriction that results in modest weight loss is known to
change the circadian pattern of glucocorticoid secretion (Krieger, 1974). Human
pregnancy is associated with a state of hyper secretion of insulin with peripheral
98 M. L. Power and S. D. Tardif

insulin resistance and a relative hypoglycemia. Pregnant women are more vulner-
able to ketonemia after a brief period of fasting (Felig and Lynch, 1970). This is
true of both lean and obese pregnant women (Metzgar et al., 1982). Women appear
to have a shorter ˜starvation time™ when pregnant.
Fasting appears to be an independent risk factor for preterm birth (Hobel and
Culhane, 2003). Habitually going more than 13 h without eating was associated
with a threefold greater risk of delivering preterm (Siega-Riz et al., 2001).
Herrmann and colleagues (2001) examined maternal serum CRH in regard to fast-
ing in 237 pregnancies. They found that women who habitually went 13 h or more
without food had signi¬cantly higher serum CRH concentrations. In addition,
they found an inverse linear relationship between maternal serum CRH and gesta-
tional age at delivery. Thus, fasting and an established metabolic marker for preg-
nancies at risk of delivering preterm (elevated CRH) have now been linked.
Whether the elevated CRH is causal of preterm labor, or a marker of other events,
perhaps accelerated placental“fetal axis maturation, or merely re¬‚ects the activation
of the maternal HPA axis, is unknown.


Leptin is a molecule intimately linked with nutrition and feeding. Leptin is also an
excellent example of the value of animal models in inducing new research path-
ways. The obese mouse model (ob/ob mouse) was developed over 50 years ago.
The evidence quickly supported the hypothesis that the ob/ob mouse lacked a
humoral factor that led to unregulated food intake, and thus obesity. However, that
humoral factor (leptin) was not identi¬ed until recently (Zhang et al., 1994).
Adding back leptin to the ob/ob mouse reduced food intake and led to weight loss;
but leptin had another effect as well. The obese mouse model was infertile; adding
back leptin also reversed the infertility (Chehab et al., 1996). Leptin is now believed
to have important functions in many reproductive processes (Castracane and
Henson, 2002). This illustrates another biological truism; biologically active
molecules often have multiple functions, and are active in many physiological
In addition to its role as a regulator of energy intake and adipososity, leptin
appears to have important functions regarding reproduction, though much of the
data is open to interpretation. These functions include an association with the onset
of puberty, a role in fertility for males and females, a role in ovarian folliculogene-
sis, and in implantation of the fertilized ovum. Leptin also appears to have impor-
tant roles in fetal growth and developmental processes. In many instances, such as
puberty, the role of leptin may be permissive rather than required. Leptin may
serve as a signal to the central nervous system with information on the critical
99 Maternal nutrition and metabolic control of pregnancy

amount of adipose tissue stores that is necessary for gonadotropin-releasing hormone
(GnRH) secretion and pubertal activation of the hypothalamic“pituitary“gonadal
axis. Leptin also acts at the periphery, directly on the ovary and testis where it may
control steroidogenesis (Baldelli et al., 2002).
As leptin is strongly associated with a measure of maternal nutritional status (fat
mass), it is a plausible candidate for being an important metabolic signal for the
maintenance and duration of pregnancy. Low leptin levels are associated with
pregnancy loss in humans. Leptin levels may be abnormally high in pregnancies
complicated by conditions such as diabetes mellitus and pre-eclampsia. Leptin is
considered to be permissive of pregnancy, but not required. It may serve as a signal
that maternal condition is satisfactory for reproduction (Castracane and Henson,
2002; Dumali and Messinis, 2002).
Leptin is produced by the placenta in many species, including humans, baboons,
bats, rodents, pigs and sheep. Signi¬cant differences in leptin regulation and func-
tion during pregnancy exist between rodents and primates. Placental leptin pro-
duction is greater in primates. In rodents the placenta largely secretes leptin into
the fetal compartment, minimally into the maternal compartment. In humans
(and baboons) leptin is produced on both sides of the placenta; that is, placental
production contributes to both maternal and fetal leptin concentrations (Henson
and Castracane, 2002).
In humans, maternal serum leptin concentration is highest at mid-gestation,
and then declines. Pregnancy is considered to be a state of hyperleptinaemia with
leptin resistance; that is, high maternal leptin does not decrease food intake. Maternal


. 18
( 51 .)