. 15
( 51 .)



Figure 2.2 Mechanism of progesterone withdrawal at term in the sheep
77 The regulation of human parturition

David Haig from Harvard has cogently argued the Paternal“Maternal Con¬‚ict
hypothesis to explain the rapid evolutionary divergence which has occurred in
reproductive processes (Haig, 1993). Under this hypothesis paternal investment in
any given pregnancy is restricted to that individual fetus to which he has con-
tributed genetic material, any other pregnancy carried by that mother may not be
his progeny. From the maternal point of view all of her offspring current and
future are of equal value. The paternal genome acting through the fetus and pla-
centa therefore has an interest in maximizing the maternal resources contributed
to that particular fetus even at the expense of other potential offspring of that
mother. The mother has a strong interest in the fetus but may wish to modify its
demands to preserve resources for future offspring. This setting of paternal“
maternal con¬‚ict produces rapid evolutionary change, as each participant seeks to
push the seesaw in a different direction. For every metabolically advantageous
mutation developed by the paternal genome, the mother will seek a modifying or
restricting, contrary change. For these reasons extrapolation from experiments
conducted in reproductive processes in one mammal to another are particularly

The endocrinology of parturition in primates

If the sheep is not a particularly good model for the human, surely primates are
better. It is clear that pregnancy among the different primates is more similar than
between primates and other mammalian classes, nevertheless intriguing differ-
ences exist. The neuropeptide CRH is made in the placentas of all primates studied
except the lemurs and not at all in non-primates (Robinson et al., 1989; Bowman
et al., 2001). However the pattern of production of this peptide and its concentra-
tions in maternal plasma vary considerably across the primates. Thus, while an
exponential rise is seen across gestation in apes, baboons show a peak in mid-preg-
nancy and similar changes are seen in production of estradiol (Goland et al., 1992;
Smith et al. 1993; 1999). Additionally while the human possesses a circulating
binding protein for CRH many primates do not (Bowman et al., 2001). Apes pro-
vide a good model of human parturition based on present data, unfortunately
experimental studies in apes are, if anything, harder to perform than those in
humans due to ethical issues, availability of animals, expense and dangers related to
human pathogens present in apes. Animal studies will continue to provide impor-
tant clues for studies of human reproductive physiology but direct extrapolation is
evidently not appropriate. Experimental studies in humans are not ethically pos-
sible, on some occasions nature™s experiments, in the form of naturally occurring
mutations, provide valuable insights into physiology but, in general, recent human
research has progressed through observational studies.
78 R. Smith et al.

CRH and the timing of birth in humans

Recent studies on the regulation of the timing of human birth have addressed two
related but different questions: how is the duration of gestation determined and
how are the events of labor precipitated? The questions have different clinical
corollaries: how can we predict premature birth and how can we prevent preterm
delivery? Effective methods to identify women at high-risk of preterm delivery are
required in order to establish satisfactory trials of methods to prevent preterm
delivery if women at low-risk of preterm delivery are to be saved from needless
exposure to experimental pharmaceuticals.
While many biochemical markers have been examined for their ability to predict
preterm delivery the most extensive studies have been conducted on CRH. CRH is
synthesized in the placenta and released preferentially into the maternal compart-
ment. Production and maternal plasma concentrations increase exponentially
through gestation peaking at the time of delivery (Figure 2.3). Early studies deter-
mined that women in preterm labor had elevated maternal plasma concentrations
compared to gestational-age matched control women (Goland et al., 1986; Wolfe
et al., 1988). Subsequently, prospective longitudinal studies (McLean et al., 1995;
Prickett et al., 2000) revealed that women destined to deliver preterm had more rapid
exponential rises while women who would deliver late had slower rates of rise. A type
of timing mechanism appears to exist in the human placenta which determined
the length of gestation. Several important concepts arose from this work. Firstly, it


CRH (pg/ml)




26 28 30 32 34 36 38 40 labour
Gestational age (weeks)
Figure 2.3 CRH increases exponentially in pregnant women™s plasma. Adapted from Campbell et al.
79 The regulation of human parturition

established that, for at least a proportion of women, it is possible to predict the tim-
ing of delivery months in advance. This reveals the possibility of developing useful
diagnostic tests to predict women at high-risk of preterm delivery and facilitate the
establishment of therapeutic trials of treatment to prevent preterm birth. Secondly,
the work established that events early in pregnancy had an in¬‚uence on the later tim-
ing of birth. Understanding the regulation of placental CRH expression may there-
fore provide insights into the determination of gestational length. Recent work in
animals has suggested that the nutritional state of the mother at conception can
in¬‚uence the length of gestation (Young et al., 1996; Bloom¬eld et al., 2003). This is
the type of clue such comparative studies can provide; we should not expect the sit-
uation in humans to be identical but parallels may exist.

Regulation of placental CRH production

Regulation of CRH production has been explored in human placental tissue
(Petraglia et al., 1989). CRH is produced by syncytial cells which can be created in
vitro by fusion of puri¬ed cytotrophoblast cells. Using cultured placental cells
and radioimmuno assays, Robinson et al. (1988) demonstrated a consistent effect of
glucocorticoids in stimulation of CRH secretion. Interestingly Mazoub et al. have
demonstrated that the exponential increase observed in human pregnancy can be
well reproduced using a model which incorporates positive feed-forward between
CRH and glucocorticoids (Emanuel et al., 1994). This ¬nding was surprising as
glucocorticoids inhibit CRH secretion within the hypothalamus. Using transfections
of CRH promoter constructs the stimulatory mechanism has been partially eluci-
dated (Figure 2.4). In placental tissue glucocorticoids stimulate CRH gene expres-
sion by interacting with proteins which bind to the cyclic adenosine monophosphate
(cAMP) response site (cAMP regulatory element, CRE) of the CRH promoter
(Cheng et al., 2000). Evidence indicates that the difference in behavior of the CRH
gene in the placenta and hypothalamus is due to the expression of different tran-
scription factors, co-activators and co-repressors in these two tissues (King et al.,
2002). In the placenta the transcription factor Jun is found binding to the CRE while
in the pituitary cell line AtT10 (in which glucocorticoids stimulate CRH expression)
Fos is more prominent in its binding. Estrogens have been shown to inhibit CRH
secretion and nitric oxide inhibits CRH secretion but not synthesis (Ni et al., 1997;
2002). cAMP analogues are very potent stimulators of CRH production but it is not
clear what external signals may be driving cAMP stimulated CRH production.
Presently it appears that conditions at the beginning of pregnancy determine the tra-
jectory of CRH production by the placenta (McGrath et al., 2002). Once established
this trajectory of exponential increase is maintained by a positive feed-forward sys-
tem involving glucocorticoids possibly damped by estrogens. The production of
80 R. Smith et al.

cAMP Synergistic
Negative interaction


(a) (b) Glucocorticoids
Figure 2.4 Schematic models of CRH gene regulation. Regulatory interactions on the CRH gene
promoter are shown for the placenta (a: placental model) and the hypothalamus
(b: hypothalamic model). The nGRE is a negative glucocorticoid regulatory element, CRE
is the cAMP regulatory element, CDXRE is caudal type homeobox response element,
GRR represents the region located between 213 and 99 bps that is stimulated by
glucocorticoids in the hypothalamic model, and TATA is the TATA box binding site for basal
transcriptional proteins. Stimulatory ( ) and inhibitory ( ) regulatory effects by cAMP and
glucocorticoids through the different elements are shown. The regulatory proteins
identi¬ed, so far, are represented by different shapes containing their names

estrogens may be regulated by CRH stimulation of dehydroepiandrosterone sulfate
(DHEA-S) synthesis by cells of the fetal zone of the adrenal which exhibits CRH
receptors (Smith et al., 1998).
Some evidence suggests that the trajectory of CRH production may be increased
by an adverse fetal intrauterine environment. Elevated maternal CRH has been
observed in pregnancies complicated by pre-eclampsia, reduced umbilical artery
¬‚ow as re¬‚ected in Doppler ¬‚ow studies and where fetal distress has lead to elective
preterm delivery (Giles et al., 1996). Whether these increases are due to increased fetal
or maternal cortisol production is unclear. Such increases in maternal CRH may have
a protective effect since CRH is a powerful vasodilator in both the maternal and pla-
cental vascular trees (Clifton et al., 1995). CRH appears to regulate endothelial func-
tion by stimulating mast cell degranulation and increasing release of nitric oxide. The
length of gestation may therefore be determined by factors that set the initial rate of
production of CRH or by factors later in pregnancy which alter the trajectory of
CRH. Not all cases of preterm delivery are associated with elevated concentrations
of CRH. It seems likely that the pathway to delivery can be activated independently of
CRH. Infection does not appear to be associated with increased CRH production.
For these reasons maternal plasma CRH has a relatively high speci¬city but lower
sensitivity (Inder et al., 2001; Ellis et al., 2002). That is if CRH is high it is likely to be
associated with preterm delivery but a low CRH does not preclude preterm birth.
81 The regulation of human parturition

Mechanisms proposed to link CRH to the process of parturition

While the association of maternal plasma CRH with preterm birth is robust in
published studies it is unclear how CRH may be directly linked to the onset of
labor. CRH receptors have been identi¬ed on the myometrium, however these are
predominantly associated with G -s proteins which activate adenylyl cyclase and
lead to increased cAMP and pathways which promote relaxation rather than con-
traction (Grammatopoulos et al., 1998). Work on CRH receptors has suggested
that different receptor isoforms may be expressed at the end of pregnancy which
are less ef¬cient at stimulating cAMP formation and may therefore move the bal-
ance within the myometrial cell towards contraction. Alternatively placental CRH
released into the fetal circulation may act on the fetal pituitary to stimulate ACTH
production, thereby increasing cortisol synthesis and driving parturition in a man-
ner analogous to that seen in the sheep (possibly by increasing prostaglandin pro-
duction in the fetal membranes) (Patel et al. 2003). Finally CRH may act on the
fetal adrenal, and perhaps the maternal adrenal, to drive DHEA-S production.
DHEA-S is an obligate precursor for placental estradiol formation. This mechanism
may drive a progressively increasing concentration of estrogen which activates
contraction associated genes. However it is also possible that rising concentrations
of CRH merely represent a marker of progressive fetoplacental maturation which
is itself, through other pathways, associated with the onset of labor. Evidence for
the ¬nal pathways of human myometrial activation is gradually accumulating
through a number of different experimental approaches.

Activation of the human myometrium

In recent years several groups have begun to examine myometrial tissues obtained
at caesarian section either prior to, or after, the onset of labor. Using these tissues,
and comparing protein and gene expression in the presence and absence of labor,
progress in understanding the mechanisms of human labor has occurred. An early
report identi¬ed a reduction in the expression of the Gas subunit required for path-
ways leading to myometrial relaxation (Europe-Finner et al., 1993). This suggested
a change in the balance of contractile versus relaxatory forces with the onset of
labor. A key dif¬culty in understanding human labor is to determine how labor
could occur despite the continued presence of high concentrations of circulating
progesterone at the end of pregnancy which would be expected to suppress labor
(Figure 2.5). In most mammals labor is associated with a profound fall in circulat-
ing progesterone concentrations but this does not occur in humans or other great
apes. This conundrum has recently been addressed by Mesiano et al., in Australia
and Phil Bennett™s group in London, England (Pieber et al., 2001; Mesiano et al.,
82 R. Smith et al.

Fetus Placenta Mother



ACTH Myometrial contractility,
cervical softening






. 15
( 51 .)