Effect of hypoxia on placental activin A, inhibin A and follistatin synthesis

Placental activin A and inhibin A output is increased in pre-eclampsia, a condition characterized by placental hypoxaemia, whereas follistatin secretion is unaltered. We investigated whether hypoxia was the basis for elevated placental activin A and inhibin A output. First trimester and term placental explants were grown in 5-6% dissolved O(2) (n=10/trimester) and 200 microM cobalt chloride (CoCl(2),n =6/trimester) to simulate environmental and cellular hypoxia respectively, for up to 72 h. Activin A, inhibin A and follistatin production were compared with control cultures grown in standard media at 20% O(2). In first trimester and term placenta, activin A output declined significantly under 5-6% O(2) (P=0.006 and 0.001 after 48 h respectively). Inhibin A declined significantly under 5-6% O(2), mainly in first trimester placenta (P=0.03, 24h). CoCl(2) significantly elevated activin A production in term placenta (P=0.003, 48 h), whereas inhibin A output was unaffected. Neither low O(2) or CoCl(2) altered follistatin output from first trimester or term placenta. These findings suggest that there may be novel O(2) sensing mechanism/s that down regulate activin A and inhibin A in the placenta and that low O(2) is not the mechanism behind increased placental inhibin A or activin A output in pre-eclampsia.

Activin A and activin receptors in gestational tissue from preeclamptic pregnancies

Maternal serum activin A levels are elevated in women with preeclampsia. To explore whether this could be due, at least in part, to increased production by the gestational tissues, we have measured activin A in the serum of women with (n=23) or without preeclampsia (n=62) at 29-40 weeks of gestation and in placenta and fetal membranes from preterm preeclamptic (PT-PE, n=8), term preeclamptic (T-PE, n=10) and healthy term controls (T-C, n=10). We have also explored if there are associated changes in activin receptor Alk2, ActRII and ActRIIB in these tissues. The relative amounts of receptor proteins were measured by densitometry on Western blots and receptors and activin beta(A) subunit localised by immunohistochemistry in PT-PE, T-PE and T-C gestational tissues (n=8-10/group). Maternal serum activin A levels were significantly elevated in women with preeclampsia (multiples of the normal median (MoM)=3.5, P<0.0001, Mann-Whitney U test) compared with healthy women (median MoM=1.0). Compared with control tissues, the activin A content was significantly higher in preeclamptic placentae (P=0.001 and P=0.0005 for PT-PE and T-PE respectively, Mann-Whitney U test), but significantly lower in the amnion (P=0.005 and P=0.014 for PT-PE and T-PE respectively) and choriodecidua (P=0.009 for T-PE). The maternal serum activin A level in women with preeclampsia was significantly correlated with elevated placental production (P=0.01, Pearson's correlation). Receptor Alk2 protein levels were significantly elevated in T-PE placentae (P=0.0006, Mann-Whitney U test), ActRIIB levels were significantly lower in PT-PE placentae (P=0.01) and ActRII levels were significantly lower in PT-PE choriodecidua (P=0.0002) compared with controls. There were no apparent differences in the distribution of the beta(A) subunit and receptors Alk2, ActRII and ActRIIB between control and preeclamptic tissues. These findings suggest that elevated levels of activin A in the maternal circulation in association with preeclampsia are due, at least in part, to increased placental production, and that the regulation of activin synthesis in placenta and fetal membranes is differentially regulated. Further, the differences in activin receptor protein levels between preeclamptic and control placenta and choriodecidua suggest that activin A-induced regulation may be altered in preeclampsia.

Activin betaA-subunit and activin receptors in human myometrium at term and during labour

OBJECTIVE

To measure activin A content and to localise and semi-quantitate activin receptors in human myometrium at term and during labour.

DESIGN

Myometrium was collected from non-pregnant women (n = 6), pregnant women at term not in labour (n = 6) and at term in labour (n = 6).

SETTING

Monash Medical Centre, Melbourne, Australia.

MAIN OUTCOME MEASURES

Tissue lysates of myometrium were analysed for activin A content using an enzyme-linked immunosorbent assay and activin receptor proteins IA, IIA and IIB using Western hybridisation. Activin betaA-subunit and activin receptors were localised in myometrium by immunohistochemistry.

RESULTS

Activin A was detected by ELISA in non-pregnant, pregnant and labouring myometrium. Levels were significantly higher in labouring myometrium. The three activin receptors IA, IIA and IIB were detected in all myometrial samples by Western hybridisation. Receptor IA was expressed in significantly higher levels in pregnant myometrium. Receptor IIA was very weakly expressed throughout. The expression of receptor IIB was similar in all three groups. Activin betaA-subunit and all three receptors were localised to the endothelial cells of myometrial blood vessels. Neither activin betaA-subunit nor any of the three activin receptors were immunolocalised to myometrial smooth muscle cells in the three groups. This result was confirmed by Western blotting for expression of activin receptors in isolated myometrial smooth muscle and microvascular endothelial cells.

CONCLUSION

The myometrium is not a target for activin A during late pregnancy or labour. However, activin A may have a role in the regulation of microvascular endothelial cell function in the myometrium.

Physiological and regulatory roles of activin A in late pregnancy

Unexplained fetal death in utero in late pregnancy represents an increasing proportion of perinatal deaths. It has been assumed that critical hypoxia is the likely mechanism underlying these losses, but the lack of a physiological marker has hampered both confirmation and prediction which could lead to timely intervention. In this paper, we report studies on hypoxia that we have performed in chronically cannulated late pregnant sheep, complemented by parallel investigations undertaken in human pregnancies. Our initial studies were directed towards determining activin secretion in the fetus and mother during late gestation, and immediately after fetal surgery using a sheep model. This led us to propose that there may be a relationship between hypoxia and activin A, follistatin and prostaglandin (PG) release from the feto-placental unit. Subsequent studies have been directed towards examining this potential relationship in sheep and in humans with compromised pregnancies. As a result of these studies, we have identified a potential mechanism by which activin A may be involved in regulating the response of the fetus to hypoxic insult. Activin A and follistatin concentrations increased in late gestation in ovine maternal plasma and in fetal fluids. Feto-placental hypoxemia or maternal isocapnic hypoxemia, leading to fetal hypoxia, were specific triggers for an acute increase in fetal activin A and follistatin concentrations during late gestation. The source and secretion of activin A, follistatin, and the associated release of PGE(2,) from within the feto-placental unit varied according to the site of the insult. The concomitant secretion of activin A and PGE(2) into the fetal circulation and amniotic fluid during reduced uterine blood flow provides an insight into the physiological regulatory mechanisms that might be involved. Changes observed in maternal activin A concentrations in mid and late gestation in the human may also be associated with fetal compromise. In human pregnancies, elevated activin A concentrations were observed in maternal plasma in mid and late gestation, in association with severe pre-eclampsia and with severe fetal growth restriction, compared to those observed in pregnancies with constitutionally small, healthy fetuses. Activin A was also elevated in maternal and arterial cord plasma in women at term during labour and immediately prior to undergoing emergency Caesarean section for failure to progress. These findings offer exciting new possibilities to gain insights into the mechanisms that underlie the maintenance of fetal wellbeing and provide a rationale for the potential that activin A may prove to be a useful clinical marker of fetal distress.

Maternal serum total activin A and follistatin in pregnancy and parturition

OBJECTIVE

To examine changes in maternal serum levels of activin A and follistatin during pregnancy and labour.

DESIGN

In three cross sectional and three longitudinal studies venous blood was collected from women during pregnancy, spontaneous labour, labour induction and prior to elective caesarean section for the measurement of activin A and follistatin.

SETTING

Monash Medical Centre, Clayton, Victoria, Australia.

POPULATION

One hundred and twenty-three women participated in a cross sectional study in pregnancy, 18 women in two longitudinal pregnancy studies, 36 women in a cross sectional labour study, nine women in a longitudinal study of labour induction. Ten women undergoing elective caesarean section were also studied.

METHODS

Activin A and follistatin were measured using two sensitive and specific enzyme-linked immunosorbent assays.

RESULTS

In the cross sectional study of pregnancy, mean (SEM) maternal serum activin A and follistatin levels increased towards term (2.4 ng/mL (0.3) and 1.8 ng/mL (0.3) in first trimester to 18.9 ng/mL (3.8) and 5.3 ng/mL (0.9) at term, respectively), but the longitudinal study revealed that levels plateau in the last three weeks of pregnancy (16.0 ng/mL (2.6) and 6.2 ng/mL (1.4) at 37 weeks and 16.6 ng/mL (3.5) and 6.2 ng/mL (0.5) before labour for activin A and follistatin, respectively). There was no difference in levels of activin A and follistatin between women delivered by caesarean section and labouring women at term (14.9 ng/mL (2.8) vs 11.0 ng/mL (0.93) and 5.95 ng/mL (0.67) vs 5.71 ng/mL (0.63), respectively) and levels of both proteins did not alter throughout spontaneous or induced labour.

CONCLUSIONS

We believe that these data argue against activin A playing an acute role in the initiation or regulation of human parturition.