<<

. 10
( 51 .)



>>

Several in vitro evidences demonstrated that CRH has vasodilatory effects in a num-
ber of species (Ramirez et al., 1995). In fact, CRH caused dilation of the mesenteric
arteries in vivo (MacCannell et al., 1984) and both in rat and human i.v. administra-
tion of CRH lowers arterial pressure due to peripheral vasodilation caused by a direct
action on vascular smooth muscle (Hermus et al., 1987; Kiang and Wei, 1987; Corder
et al., 1992). However, in most animals and in non-pregnant humans, peripheral
concentrations of CRH are low (Vale et al., 1993), which suggests that CRH may play
a minimal role in the control of vascular tone. In contrast, in the pregnant human,
plasma CRH concentrations rise exponentially, peaking at term (Ramirez et al., 1995;
Challis et al., 2000; Hillhouse and Grammatopoulos, 2002).
The CRH, when administered chronically in pregnant rats, causes a decrease
in blood pressure (Jain et al., 1998), and it is also a potent relaxant of the uterine
artery of pregnant rats, acting both on the endothelium (mediated by NO“cGMP
system) and the vascular smooth muscle (Jain et al., 1999). Animal studies revealed
that reduced uterine blood ¬‚ow and consequent hypoxia induce an increased
expression and secretion of CRH, and consequently of ACTH and cortisol (Sue-
Tang et al., 1992). As CRH acts as vasodilator in placental circulation, increased
CRH could act systemically or be released locally in the placenta as a compensatory
mechanisms to reduce uterine resistance to blood ¬‚ow (Gagnon et al., 1997). In
humans, an impaired placental CRH secretion has been associated with the lack of
uterine artery dilation and the decrease of the utero-placental blood ¬‚ow through
uterine arteries, supporting the concept that at mid-gestation placental CRH is
involved in the control of the uterine artery tone (Florio et al., 2003b).
Recent investigations have shown that CRH is a potent dilatator of the human
fetal“placental circulation (Clifton et al., 1994; 1995; 1996), acting at concentra-
tions comparable to plasma CRH levels in maternal and fetal circulation. Placental
CRH may, therefore, have a signi¬cant physiological role as a regulator of feto-
placental vascular tone. This effect is due to endothelial-independent pathways
(Kiang and Wei, 1987; Corder et al., 1992), but in some species CRH may also operate
via endothelium-dependant mechanism, acting on speci¬c receptors expressed by
endothelium, as in the case of human umbilical vein endothelial cells (HUVEC)
(Simoncini et al., 1999).
In the human fetal“placental circulation CRH causes vasodilation via a NO- and
cGMP-mediated pathway, as the addition of a blocker of NO formation and
48 F. Petraglia et al.


Inhibitors

CRF CGRP

Urocortin



CRF-receptors cAMP-coupled receptors

Placental blood vessels


Vascular smooth muscle Vascular smooth muscle
relaxation relaxation

NO pathway NO pathway

CGMP pathway

Placental
ACTH


Fetal“placental vasodilation

Figure 1.6 Placental CRF, urocortin, NPY, CGRP and PTHrP are involved in the ¬ne control of
the fetal“placental blood ¬‚ow


inhibitors of cGMP formation respectively cause marked attenuation of CRH“
stimulated vasodilation. The addition of CRH to preconstricted placental vessels is
able to attenuate all constrictor mechanisms without variation in CRH potency as
a vasodilator agent. The CRH-induced vasodilation appears to be mediated by a
CRH receptor, as the vasodilatory response to CRH is antagonized in the presence
of a CRH receptor antagonists (Clifton et al., 1994; 1995). The CRH-induced
vasodilation occurred at concentrations comparable to plasma CRH levels found
in the maternal and fetal circulations (Challis et al., 2000; Florio et al., 2002d,
Hillhouse and Grammatopoulos, 2002; Reis et al., 1999), and CRH is approxi-
matately 50 times more potent than prostacyclin (Clifton et al., 1995) (Figure 1.6).
Urocortin has the same effects of CRH: administered i.v. in rats it is more potent
than CRH in causing hypotension (Vaughan et al., 1995; Torpy et al., 1999) and,
with respect to placental circulation, it causes vasodilation, reducing fetal“
placental vascular resistance via CRH type 2 receptors, and being more potent that
CRH (Leitch et al., 1998) (Figure 1.6).
As syncytiotrophoblast is the main source of CRH during pregnancy (Petraglia
et al., 1990b; 1996d; Petraglia, 1991; Reis et al., 1999; 2001; 2002; Florio et al.,
49 Placental expression of neurohormones and other neuroactive molecules


2002d; Hillhouse and Grammatopoulos, 2002), placental CRH may access the
fetal“placental circulation to cause dilation by paracrine or endocrine mecha-
nisms. It may be released locally to affect the vascular smooth muscle and endothe-
lium, or it may be secreted into the feto-placental circulation and travel to its site
of action through the placental vascular system. Finally, CRH may maximize the
release of products such as POMC peptides (Florio et al., 2002d; Petraglia et al.,
1987c; 1999a) or PGs (Challis et al., 2000) in vivo, by causing vasodilation of pla-
cental vascular tissue (Clifton et al., 1996) (Figure 1.6).
The NPY is involved in regulation of the local uterine blood ¬‚ow (Fried et al.,
1986; Fallgren et al., 1989; Ekesbo et al., 1991; Fried and Samuelson, 1991). Since
level of NPY in plasma is increased in women with eclampsia, preeclampsia and
hypertension (Fried et al., 1986), perhaps elevated plasma NPY values may play a
key role in the development of these pathologies.

(B) Effect of CGRP
The CGRP is a potent vasodilator in a variety of animal and human systems
(Poyner et al., 2002; Yallampalli et al., 2002) and, the mode of action of CGRP is
still under investigation, as in some tissues it appears to act independently on
endothelium, apparently by interacting directly with coupled cAMP receptors on
smooth muscle (Fiscus et al., 1992), or via mechanisms mediated by the release of
vasodilator NO (Fiscus et al., 1991). On the basis of these evidences, it has been
suggested that CGRP may play a role in the control of vasoactivity of the human
fetal“placental circulation (Mandsager et al., 1994) (Figure 1.6) and, of precon-
stricted human chorionic plate vessels in vitro through two classes of receptors and
also independently of NO pathway (Firth and Pipkin, 1989). The CGRP participates
in regulation of the vascular adaptations occurring during normal pregnancy, and
it appears to be involved in the pathophysiology of preeclampsia (Kraayenbrink
et al., 1993; Schiff et al., 1995). In fact, in rats the coadministration of L-NAME
(a drug that increases blood pressure) and CGRP prevented the gestational (not the
postpartum) L-NAME hypertension and decreased pup mortality to 6.4% but did
not reverse the decreased fetal weight (Yallampalli et al., 1996; Gangula et al., 1997;
Parida et al., 1998). The same phenomenon was evident in the presence of adequate
levels of progesterone in the postpartum period, as if CGRP regulates vascular
adaptations during pregnancy and these effects may be progesterone dependent
(Figure 1.6).
Sex steroid hormones are raised during pregnancy, and the increase of plasma
CGRP levels might be related to steroid hormone levels as levels are higher in women
than in men, and mainly in women taking contraceptive pills (Valdermarsson et al.,
1990). Taken together, we could hypothesize that the relaxant effect of CGRP on
the non-pregnant and pregnant uterus, as well as its vasodilatory actions are, at least
50 F. Petraglia et al.


in part, under progesterone control. In fact, progesterone treatment of ovariectomized
mice resulted in a signi¬cant increase in the responsiveness of the myometrium to
CGRP (Naghashpour and Dahl, 2000) and, recent studies in female rats indicate
that the vasodilator effects of CGRP are sex steroid hormone dependent, as chronic
CGRP administration to pregnant rats increased systolic blood pressure, indicat-
ing a role for CGRP in maintaining a normotensive state during pregnancy
(Gangula et al., 2002). Furthermore, sex hormones increase both the synthesis of
CGRP (Gangula et al., 2000) and the responsiveness to synthetic CGRP in the vas-
culature (Gangula et al., 2001).

(C) Effect of PTHrP
The PTHrP is a potent vasodilator (Winquist et al., 1987; Nickols et al., 1990)
through the activation of myometrial cells of adenylyl cyclase and its expression in
vascular smooth muscle cells increases in response to hypertension, vasoconstric-
tor agents, increased ¬‚ow, shear stress and stretch. The PTHrP has vasodilator effect
in the human fetal“placental circulation (Macgill et al., 1997), and is 100 times
more potent than PTH (Mandsager et al., 1994). The expression of PTHrP in utero
may be stimulated by hormones including estradiol, prolactin and placental lacto-
gen (Dvir et al., 1995), by cytokines and growth factors (Casey et al., 1992; Dvir
et al., 1995) and mechanical stimuli (Daifotis et al., 1992).


Inhibin-related peptides

Inhibins and activins (Box 1.8) are growth factors belonging to the transforming
growth factor- (TGF- ) superfamily, and composed by two subunits. Inhibins
are heterodimers of a subunit with a subunit (inhibin A A; inhibin
B B) while activins are omodimer of subunit (activin A A A; activin
B B B; and activin AB A B) (Vale et al., 1988). They were originally iden-
ti¬ed from gonads as factors acting antagonistically in the endocrine regulation of
pituitary follicle-stimulating hormone (FSH) production, but successive descriptions
of their expression in numerous cells types and tissues outside the hypothalamic“
pituitary“gonadal axis indicate different functions, particularly as modulators of cell
growth, differentiation and apoptosis (Luisi et al., 2001). The A subunit is also
expressed from the ¬rst trimester of gestation, with the highest value at term, while
the B-subunit, present in the outer syncytial layer, is detected only at term
(Petraglia et al., 1991; 1992b). The A-subunit is localized in the external syncytial
layer of placental villi, in some structure of the stroma, in maternal decidual cells
and in some amnion and chorionic cells (Petraglia et al., 1990d; 1991; 1993c;
1994c) but, in term trophoblasts, also in endothelial cells within the placental villi
(Schneider-Kolsky et al., 2002).
51 Placental expression of neurohormones and other neuroactive molecules



Box 1.8 Inhibin-related peptides

Inhibins and activins are growth factors belonging to the TGF- superfamily,
and composed by two subunits. Inhibins are heterodimers of a subunit with
a subunit (inhibin A A; inhibin B B) while activins are omodimers
of subunit (activin A A A; activin B B B; activin AB A B) (Vale
et al., 1988).


Expression and localization
Human placenta synthesizes inhibins and activins (Petraglia, 1997; Florio et al.,
2001a). The subunit mRNA in the human trophoblast is expressed from the
¬rst trimester of gestation, with the highest values at term (Petraglia et al.,
1991), and it is localized within the structure of placental villi (cyto- and inter-
mediate trophoblast, mesenchymal cells) (Petraglia et al., 1987b; 1991; 1992b),
maternal decidua (Petraglia et al., 1990d), amnion and chorionic cells (Petraglia
et al., 1993c).


Receptors and binding proteins
Activins signal through a heteromeric complex of receptor serine kinases which
include at least two type I (IA and IB) and two type II (IIA and IIB) receptors.
These receptors are all transmembrane proteins, composed of a ligand-binding
extracellular domain, a transmembrane domain and a cytoplasmic domain
with predicted serine/threonine speci¬city (Gray et al., 2002).

Levels in biological ¬‚uids
During the ¬rst trimester of pregnancy, maternal serum levels of inhibin A and
activin A are higher than in non-pregnancy (Florio et al., 2001a), while inhibin B
(Petraglia et al., 1997b) does not signi¬cantly differ from levels detected during
the menstrual cycle. At this gestational period, coelomatic ¬‚uid activin A and
inhibin B levels are higher than in maternal serum and amniotic ¬‚uid (Luisi
et al., 1998). In amniotic ¬‚uid, inhibin A is not detectable (Riley et al., 1996). In
this trimester of pregnancy, the feto-placental unit is the main source of circu-
lating activin A and inhibin A (Muttukrishna et al., 1997). During the second
trimester of pregnancy, inhibin A and activin A further increase in maternal
serum and amniotic ¬‚uid, whilst inhibin B increases only in amniotic ¬‚uid
(Petraglia et al., 1995a, c; 1999b; Muttukrishna et al., 1996; Wallace et al., 1997;
Muttukrishna et al., 2000). At term, maternal serum levels of inhibin A (Florio
et al., 1999a; Muttukrishna et al., 2000), inhibin B (Petraglia et al., 1997b) and
52 F. Petraglia et al.



activin A (Petraglia et al., 1994a; Florio et al., 1999a; Muttukrishna et al., 2000)
and those in amniotic ¬‚uid (Wallace et al., 1997) are at their highest. Inhibins A
and B and activin A are also measurable in umbilical cord blood (Wallace et al.,
1997; Florio et al., 1999a).



Inhibin and activin secretion from cultured placental cells is controlled by both
positive and negative regulatory mechanisms involving hormones and growth fac-
tors (Petraglia et al., 1996d). The FSH, human chorionic gonadotropin (hCG), PGs,
epidermal growth factor (EGF) and TGF- are potent stimulators, while TGF-
and activin A are suppressors for inhibin production in cultured trophoblast cells
(Petraglia et al., 1996d; 1987b). Furthermore, it was found that the addition of
gonadotropin-releasing hormone (GnRH) and glucocorticoids induces an increases
in the release of the inhibin in cultured human trophoblast cells (Keelan et al.,
1994). With respect to activin A, GnRH, inhibin, TGF- , dexamethasone, cAMP and
IL-1 have no effect on its production in cultured trophoblast cells, while its pro-
duction can be stimulated by phorbol ester (Rabinovici et al., 1992; Keelan et al.,
1994), tumor necrosis factor (TNF)- , IL-1 and granulocyte and monocyte colony
stimulating factor (GMCSF) (Keelan et al., 1998; Mohan et al., 2001) (Figure 1.7).
First and second trimester placentae express the various receptor proteins in the
syncytium, whereas at term the distribution is con¬ned to vascular endothelial
cells of villous blood vessels. In the fetal membranes they are localized to some
epithelial cells, mesenchime and chorionic trophoblast (Schneider-Kolsky et al.,
2002).
The activity of activin A is tightly regulated by follistatin, a structurally unre-
lated protein, that binds with high af¬nity to activin and neutralizes its activity
(Luisi et al., 2001). This af¬nity is similar to that for activin receptors, thus it plays
a major role in regulating activin bioavailability on target tissues and functions.
Recently, a new binding protein of 70 amino acids for activin A has been identi¬ed,
namely follistatin-related gene (FLRG), closely related to follistatin (Hayette et al.,
1998), that interacts physically with activin A and, preventing the binding on
ActRs, regulates activin A functions (Hayette et al., 1998; Tsuchida et al., 2001).
Follistatin and FLRG are both present in trophoblast, decidua and fetal mem-
branes amnion and chorion), but FLRG protein immunolocalization differs from
that shown for follistatin. FLRG is predominantly present in the walls of decidual
and placental blood vessels (Ciarmela et al., 2003), whilst follistatin is more local-
ized in cyto- and syncytiotrophoblast cells (Petraglia et al., 1994c).
The measurement of inhibin-related proteins during pregnancy may have
important clinical implications. In fact, at ¬rst/second trimester serum inhibin A
53 Placental expression of neurohormones and other neuroactive molecules


GnRH
FSH
hCG
EGF
TGF-±
TGF-β
IL-1β
activin A

<<

. 10
( 51 .)



>>