Regulation of activin-A production by human amnion, decidua and placenta in vitro by pro-inflammatory cytokines

Placenta (1998), 19, 429-434

Regulation Placenta

J. A. Keelat?,

of Activin-A In Vitro

b Oxford Brookes University, Paper accepted 13 February

by Human

by Pro-inflammatory

N. P. Groomeb

a Department of Pharmacology Auckland, New Zealand

Production

Decidua

and

Cytokines

and M. D. Mitchell”

and Clinical Headington, 1998

Amnion,

Pharmacology, Oxford

University

of Auckland

School

of Medicine,

Park Road, Grafton,

OX3 OBP, UK

Activin-PA subunits are expressed by the human placenta and extraplacental membranes at term and preterm. The regulation of activin-A production by these tissues has not been characterized to date, however. To determine the effects on activin-A production of pro-inflammatory cytokines, amnion, decidual and placental cells were isolated by enzyme dispersion and treated in primary culture with interleukin-10 (IL-l/3) and tumour necrosis factor-a (TNF-a). Activin-A production (determined by ELISA) by amnion, decidual and placental cultures was 1.2 + 0.27, 31.1 * 9.9, and 50.7 * 28.5 pg/pg protein/l6 h, respectively (mean * SEM; n = 5-7 experiments). Both IL-l p and TNF-u stimulated activin-A production in a concentration-dependent fashion in all cultures; maximal stimulation was achieved at 0.25-1.0 rig/ml IL-lp and 25-50 rig/ml TNF-a, respectively. In amnion, decidual and placental cultures IL-lp stimulated activin-A production to 747 f 274, 190 * 11, and 254 * 60.2 per cent of controls, while TNF-a stimulated production to 312 f 81.5, 194 + 22.5, and 193 f 12.5 per cent, respectively (mean f SEM; n=5; PcO.05 by ANOVA). These studies show for the first time that pro-inflammatory cytokines are potent stimulators of activin-A production by intrauterine tissues. This may provide an explanation for the elevated concentrations of activin-A measured in the sera of some 0 1998 W. B. Saunders Company Ltd women in preterm labour. Placenta (1998), 19, 429-434

INTRODUCTION Activin, a glycoprotein hormone related to inhibin and transforming growth factor-p (TGF-/3) (Massague, 1990; de Kretser, 1993), is synthesized by the human placenta (de Kretser et al., 1994; Petraglia et al., 1994; Petraglia, 1997). Messenger RNA for both the PA and PB activin subunits have been localized to the fetal membranes (amnion and chorion), and the maternal decidua (Petraglia et al., 1990, 1991) by Northern analysis and in situ hybridization. More recently, activin-C mRNA expression in the placenta has been reported (Loveland, McFarlane and de Kretser, 1996). Activin-A has been measured by immunoassay in maternal serum through pregnancy, and concentrations have been reported to rise steadily throughout gestation (Petraglia et al., 1993a; Muttakrishnan et al., 1995; Knight, Muttakrishnan and Groome, 1996), with further increase apparent after the onset of term labour (Petraglia, Gallinelli and de Vita, 1994). Concentrations of free activin (i.e. unbound by follistatin) also increase dramatically in maternal serum in the third trimester (Petraglia, Gallinelli and de Vita, 1994). Levels of activin in the maternal circulation of women presenting with preterm labour ’ To whom correspondence should be addressed. E-mail: [email protected] 0143S4004/98/050429+06

$12.00/O

are elevated with respect to normal term deliveries (Petraglia et al., 1995), an observation consistent with the reported increase in activin pulse amplitude found in this group (Gallinelli et al., 1996). The source of much of the activin and inhibin measurable in the circulation in pregnancy is almost certainly the placenta, although the relative contributions from other gestational tissues are not known at present. Indeed, since mRNAs for all inhibin and activin subunits are expressed in the amnion, chorion and decidua (Petraglia et al., 1990, 1991), the potential for synthesis of numerous isoforms of activins and inhibins exist in all these tissues. At this stage it is unclear as to the relative proportions of these proteins that are produced by gestational tissues and whether these proportions change throughout pregnancy. In amniotic fluid inhibin-A concentrations exceed those of inhibin-B in early pregnancy (Petraglia et al., 1993a; Wallace et al., 1997), while amniotic fluid activin-B levels in late pregnancy have been reported to be much higher than those of activin-A (Petraglia et al., 1993a). These findings suggest that the fetal membranes could be a significant source of activins and inhibins during pregnancy and that the activin subunits are differentially expressed, both temporally and spatially, within the placenta and extraplacental membranes. 0

1998 W. B. Saunders

Company

Ltd

430

Placenta

Our understanding of the factors that regulate placental activin production is minimal. We have previously reported that the production of activin-PA subunits by trophoblast cells in vitro is increased following treatment with phorbol ester (i.e. protein kinase-C activation) and also, though less consistently, with the adenylate cyclase agonist forskolin (Keelan, Song and France, 1994). Tanimoto et al. (1992) suggested that the stability of PA mRNA might be increased following a rise in intracellular CAMP concentrations in placental trophoblasts. In haematopoietic cells, activin-A production has recently been reported to be increased by the pro-inflammatory cytokines interleukin-ID (IL-l@ and tumour necrosis factor-u (TNF-a) (Shao et al., 1992; Takahashi et al., 1992), giving rise to the naming of activin-A as ‘the new cytokine’ (Yu et al., 1996). While these pro-inflammatory cytokines are produced in gestational tissues (see below), their effects on activin production have not yet been investigated. Intra-amniotic infection is believed to be the cause of a significant proportion of preterm deliveries, particularly in women with preterm labour refractory to tocolytic therapy (Romero et al., 1994). It has been convincingly demonstrated that infection of the fetal membranes (chorioamnionitis) induces prostaglandin biosynthesis within gestational tissues resulting in the initiation of labour, and that this response is mediated, at least in part, by induction of local cytokine release (Mitchell et al., 1991; Dudley and Trautman, 1994). Specifically, the pro-inflammatory cytokines IL-l/3 and TNF-a, which have been identified as products of these tissues in the presence of bacteria or bacterial endotoxins, have been shown to be potent enhancers of prostaglandin production by cells in culture (Romero et al., 1989; Bry and Hallman, 1991; LundinSchiller and Mitchell, 1991; Mitchell, 1994; Shimonovitz et al., 1995). In the present study, we hypothesized that the presence of elevated concentrations of activin-A detected in maternal circulation in pregnancies complicated by preterm labour may be due to enhanced placental/extraplacental production in response to local cytokine production secondary to intrauterine infection. To examine the effects of these inflammatory mediators on activin-A production by gestational tissues, we treated cultured cells of amnion, decidua and placenta with IL-ID and TNF-a, and measured their effects on the production of activin-A by a sensitive and specific enzyme-linked immunosorbent assay (ELISA).

MATERIALS

AND

METHODS

Materials

Culture medium (Ham’s F12/DME) and phorbol 12-myristate, 13-acetate (PMA) were purchased from Irvine Scientific, CA, USA. Fetal calf serum (FCS) and the ELISA signal amplification kits were from Life Technologies, Auckland, New Zealand. Cytokines and the anti-IL-6 antisera were obtained from R&D Laboratories, MN, USA. Disposable tissue culture plasticware was from Nunc, Roskilde, Denmark.

(1998),

Vol. 19

The Avidchrome hydrazide-coated plates for the activin-A assay were purchased from Unisyn Technologies, San Diego, CA, USA, while ‘Easy Wash’ ELISA plates for the IL-6 assay were supplied by Corning Costar Corp., Cambridge, MA, USA.

Amnion

culture

Placentae were obtained after caesareansection at term. Maternal consent was obtained according to the guidelines of the Auckland Human Ethics Committee. Amnion cells were isolated and cultured according to previously published methods (Keelan, Sato and Mitchell, 1997) based on an original procedure by Okita et al. (1983). The amnion was manually separated from the choriodecidua, and the membranes then dissected from the placenta. After washing to remove blood, the isolated amnion was cut into smaller pieces and then minced using two scalpels. Tissue fragments were then placed into digestion medium [0.012 per cent (w/v) collagenase in Ham’s F12/DMEM] and digested for 2 h at 37°C. This was followed by two sequential digestions of 30 min in 1.2 per cent (w/v) trypsin. Dispersed cells obtained from the two trypsin digests were washed and recovered by centrifugation (7 min at 500 g), then plated into 24-well multidishes at 0.3 X lo6 cells/ ml in Ham’s F12/DMEM supplemented with 10 per cent FCS and antibiotics. One day after plating, media were withdrawn and replaced with fresh media. Experiments were conducted on day 4-S of culture, when medium was replaced with fresh medium [without serum, containing 0.1 per cent (w/v) bovine gamma globulin] and test substances. Using this method, >90 per cent of cells on day 3 of culture stained positive for cytokeratin, an epithelial cell-specific antigen. All cultures were incubated in a humid atmosphere of 5 per cent CO,/95 per cent air. Four replicate wells per test treatment were used, and experiments were conducted on tissues from at least five different placentae.

Decidua

culture

Decidual cells were isolated as described previously (Mitchell et al., 1991) with minor modifications. Decidual tissue was washed and dissected as described above then digested for 1 h in 0.012 per cent collagenase, followed by 1 h in 200 ml 0.25 per cent (w/v) dispase in Ham’s F12/DMEM. DNAse was added (0.8 mg/200 ml) 15 min before the completion of the digestion period. The cells were recovered by centrifugation then applied to a discontinuous Percoll gradient (60,40,20 and 5 per cent) and centrifuged for 20 min at 600 g (Mitchell et al., 1991). Cells recovered from the 6040 and 40-20 per cent interfaces were diluted in F12/DMEM, re-centrifuged, and plated out at 0.3 X lo6 cells/ml in Ml99 media containing 10 per cent FCS and antibiotics. Experiments were carried out in serum-free media as detailed above for amnion cultures.

Keelan,

Groome

and Mitchell:

Trophoblast

Regulation

of Activin-A

Production

by Human

protein

ELISA

Activin-A was measured by ELISA as previously described (Knight, Muttakrishnan and Groome, 1996). Standards and sample dilutions were prepared in culture media and assayedin duplicate. The intra-assay precision of the assay was <5 per cent, while the assay sensitivity was approximately 25 pg/ml. IL-6

and Placenta

431

In Vitro

40,

0

0

10

50

PMA

(nM)

250

0

0.2

1.0

Forskolin

5.0

(nM)

Figure I. Stimulation of placental activin-A production by phorbol-12, 13 myristate-acetate (PMA) and forskolin. The results are the mean & SEM of four determinations from a representative experiment that was performed three times. *P
Table

1. Basal activin-A production by placental, decidual and amnion

cells

in vitro

Activin-A

Activin-A (pg/mg

(w/ml) Trophoblast Decidua Amnion

2914 * 773” (n=7) 6848 f 766b (n=4) 137 * 34.3” (n=4)

protein/l6 h)

50.7 * 28.5 (n=7)NS 31.1 f 9.9 (n=4)NS 1.2 f 0.27 (YZ=~)~~

The data shown are from four to seven experiments (mean f s.E.M.), expressed either as concentration of activin in the conditioned media (pg/ml), or as pg activin-A/pg cellular protein over 16 h. a Significantly different from decidua or amnion; b significantly different from trophoblast or amnion; ’ significantly different from decidua or trophoblast (P~0.05 by Student-Newman-Keul’s test); NS, not significant.

Statistical

analysis

and

representation

of data

Activin-A production is expressed as pg/yg protein/l6 h. Differences between the means of treatment groups were established by ANOVA followed by post hoc testing (Dunnett’s test or Student-Newman-Keul’s test). WO.05 was considered to be significant.

assay

Cellular protein was measured at the completion of experiments using the BCA method (Redinbaugh and Turley, 1986). The assay had a working range of - 10-1000 pg/ml. Bovine serum albumin was used as a standard. Activin-A

Decidua

culture

Approximately 50 g of chorionic villi, dissected free of blood vesselsand connective tissue, were cut into 2-S-mm pieces and washed in media (M199) to remove adherent blood clots. The tissue was divided into two aliquots, and each subjected to eight sequential digestion steps (10 min at 37°C) with 50 ml trypsin solution (2.5 g/l in Locke Ringer’s buffer) containing 2 mg/l DNAse-1 (Yui et al., 1994). After the first two digests, the tissue fragments were allowed to settle, the cell suspension decanted and discarded. The suspension from the remaining six digestions were pooled on ice in 2 per cent FCS. The pooled supernatants were pelleted by centrifugation (1200 g for 10 min), washed in cold PBS, and re-pelleted. The cells were then resuspended in lysis buffer by gentle triturating and incubated at room temperature for 10 min. The cells were again recovered and washed by centrifugation, then centrifuged over a discontinuous percoll gradient (55, 50, 45, 40, 35, 30, and 15 per cent) according to the method of Kliman et al. (1986). The layer of cells containing trophoblasts (migrating with a density of 1.049-1.062 g/l, as determined by density marker beads) were collected, washed, and reconstituted in Ml99 media supplemented with 10 per cent FCS for counting. They were plated out at 1 X IO6 cells/ml in 24-well culture dishes (1 ml/well) in Ml99 media supplemented with 10 per cent FCS, 1 X BME vitamin supplement, insulin (5 pg/ml), transferrin (10 pg/ml), sodium selenite (0.2 KIM), epidermal growth factor (5 rig/ml) and antibiotics. One day after plating, the medium was replaced with fresh medium containing 1 per cent FCS and the remaining supplements. On day 3 of culture, fresh media was added (with 1 per cent FCS) and the cells exposed to the test substances (n=4 wells per treatment point) as described above. The digestion procedure yielded cultures that contained >95 per cent villous trophoblasts (cytokeratin positive), ~2 per cent macrophages (CD68 positive), and <4 per cent mesenchymal/fibroblast cells (vimentin positive). Cellular

Amnion.

ELISA

IL-6 was measured by ELISA as previously described (Keelan, Sato and Mitchell, 1997).

RESULTS

Initial studies were conducted to validate the assay for use on media samples from the tissues being studied. Conditioned media from amnion, decidual and placental cultures assayed at a series of dilutions maintained linearity and parallelism with standards diluted in fresh medium (not shown). To confirm previous results from immunoblotting studies (Keelan, Song and France, 1994) and Northern blotting experiments (Tanimoto et al., 1992), we treated trophoblast cultures with PMA and the adenylate cyclase potentiator forskolin (Figure 1). Increased activin-A production in response to these agents was observed as anticipated, with PMA being the more potent stimuli as previously suggested (Keelan, Song and France, 1994). Activin-A was readily detectable in all amnion, decidual and placental culture media (Table 1). Concentrations of activin-A

43.2

Placenta

60,

(4

*T

50

30 20 10 0/ ^ zi

15

Vol. 19

When these experiments were repeated with amnion cells [Figure 2(B)] and decidual cells [Figure 2(C)], similar effects were observed, although in the decidual cultures the magnitude of stimulation was not as great. In amnion cultures, maximally effective concentrations of IL-lp and TNF-a stimulated activin-A production to 747 * 274 and 312 * 81.5 per cent of controls, respectively (n=S experiments). In decidual cultures these cytokines stimulated production to 190 * 11 and 194 + 22.5 per cent of control, respectively (n=5 experiments). To compare the effects of pro-inflammatory cytokines on activin-A production with those on production of a more orthodox cytokine, we measured IL-6 production in the same cultures. IL-6 was chosen as it is produced by all of the tissues of interest, its production is responsive to stimulation by pro-inflammatory cytokines, and it has been measured in amniotic fluid for its diagnostic potential in predicting infection-associated preterm labour (Hillier et al., 1990; Coultrip et al., 1994; Dudley et al., 1994). Concentrations of activin-A and IL-6 in all the cultures were similar, as were the degrees of stimulation of production observed in response to IL-lfl and TNF-a (results not shown).

1111

40

3'i

* * T

(1998),

25 50

0 0.010.05 0.25 1.25

t(B)

DISCUSSION

100

I

50

0:

LI

01

5 10 25 50

TNF-a

(rig/ml)

0 0.040.2 1 5 10 IL-Q3 (rig/ml)

Figure 2. Stimulation of activin-A production by (A) placental trophoblasts, (B) amnion and (C) decidual cells by interleukin-lp (IL-la) and tumour necrosis factor-a (TNF-a). The results are the mean 4~ SEM of four determinations from a representative experiment. Similar results were obtained in four or five other experiments. “P
in conditioned media from trophoblast and decidual cultures were significantly higher than in amnion-conditioned media, although when production rates were normalized to cellular protein these differences lost statistical significance. Cultured placental trophoblasts were treated with TNF-a and IL-lj3 over a range of concentrations that have been reported in the literature (and confirmed in our laboratory) to stimulate cytokine and prostanoid production by gestational cells in vitro. Both IL-l p and TNF-a stimulated activin-A production in a concentration-dependent fashion [Figure 2(A)]. Maximal stimulation was observed at 0.25-1.0 rig/ml IL-lp and 25-50 rig/ml TNF-a, respectively, effecting an increase of 254 f 60.2 and 193 + 12.5 per cent of controls (mean =t SEM, n= 5 experiments).

The present report is the first quantitative study of activin production in vitro by gestational tissues, and for the first time describes the regulation of activin-A production by proinflammatory cytokines in these tissues. While activin PA mRNA has been localized to the extraplacental membranes in previous studies (Petraglia et al., 1990, 1991), until now the production of activin-A by the fetal membranes and decidua had been a matter of assumption rather than fact. While our study measured immunoreactive activin-A and not performed structural or bioactivity studies, there is little doubt that the measurements accurately reflect total activin production by these cells. The activin-A ELISA has been extensively validated and found to be highly specific due to the epitope targeted by the antibody used and the dissociation step carried out before assay (Muttakrishnan, Knight and Groome, 1996). Linearity/parallelism studies supported the view that the substance being measured in the conditioned media is activin-A. The assay measures total activin-A, irrespective of the activity of binding proteins (i.e. follistatin, a,-macroglobulin) present. Hence the data presented should not be taken as a reflection of ‘free’ bioactive activin levels. Further studies will be required to ascertain if the changes described in this report reflect similar changes in bio-available activin. The present findings suggest that intrauterine activin production is likely to be elevated in areas of local infection and cytokine release, and as such is one of a growing number of cytokines reported to be synthesized by marrow-derived and resident cells of the intrauterine tissues in response to proinflammatory mediators. The data support the possibility that the elevated levels of activin-A measured in the sera of women

Keelan,

Groome

and Mitchell:

Regulation

of Activin-A

Production

by Human

with preterm labour might be, in some cases, due to increased production by intrauterine tissues in response to bacterial infection. The villous placenta is the most likely source of the activin measured in the maternal circulation, due to its ample access to maternal blood supply. While it is conceivable that some of the activin produced by the fetal membranes might enter the maternal circulation, their inaccessibility to the maternal vasculature makes their contribution to maternal plasma levels likely to be relatively minor. Thus, it appears likely that in the majority of cases of ascending intrauterine infection, inflammation of the membranes, accompanied by increased cytokine and activin-A production, would not result in detectable increases in activin-A in the maternal circulation. It remains to be determined to what extent placental infection occurs in casesof chorioamnionitis, although at least one study suggests that it is, in fact, considerable (Halgunset et al., 1994). Hence, it is plausible to hypothesize that elevated maternal activin-A concentrations described in some women with preterm labour might reflect inflammation of the villous placental tissue as a consequence of ascending intrauterine infection. The rise in maternal plasma activin-A concentrations reported after onset of labour (Petraglia, Gallinelli and de Vita, 1994) might reflect elevated placental production associated with an inflammatory reaction as a consequence of the labor process itself. The physiological significance of increased activin-A production in response to cytokines remains to be established. While activin-A has actions on placental hormone production

Amnion,

Decidua

and Placenta

In Vitro

433

(Petraglia et al., 1987, 1993b; Petraglia, Vaughan and Vale, 1989; Qu and Thomas, 1993; Florio et al., 1995; Song, Keelan and France, 1997) and first trimester trophoblast proliferation (Canaggia, Lye and Cross, 1997), these seem unlikely to be of major significance in the face of an intrauterine infection. It is possible, however, that the low levels of cytokines produced by gestational tissues in normal pregnancy might have a role to play in regulating production of placental hormones, activin-A included. To date there have been no studies published describing the actions of activin on the fetal membranes or decidua. Prostaglandin production by an amnion-derived epithelial cell line has been reported to be stimulated by activin (Petraglia et al., 1993b), but this finding has yet to be confirmed and we have been unable to reproduce it in our laboratory (Keelan and Mitchell, unpublished). In contrast, we have observed a concentration-dependent stimulation of prostanoid production in placental cells in vitro (Mesnage, Hansen, Mitchell & Keelan, unpublished), supporting a role for activin in modulating the production of inflammatory mediators in this organ. In conclusion, our studies have shown for the first time that pro-inflammatory cytokines are potent stimulators of activin-A production by intrauterine tissues. The significance of these findings to pregnancy and parturition remain to be determined, although they may provide an explanation for the elevated concentrations of activin-A measured in the sera of some women with preterm labour.

ACKNOWLEDGEMENTS The assistance of the nursing and theatre was funded by grants from the Auckland Foundation

staff at National Women’s Hospital, Auckland, with the collection of placentae is gratefully acknowledged. This work University Research Committee, New Zealand Lottery Health Grants Board, and the Auckland Medical Research

REFERENCES Bry

K & Hallman M (1991) Synergistic stimulation of amnion cell prostaglandin E, synthesis by interleukin-1, tumor necrosis factor, and products of activated human granulocytes. Prostaglandins, Leukotrienes and Essential Fatty Acids, 42, 167-169. Canaggia I, Lye SJ & Cross JC (1997) Activin is a local regulator of cytotrophoblast differentiation. Endocrinolog),, 138, 3976-3986. Coultrip LL, Lien JM, Gomez R, Kapernick P, Kboury A & Grossman JH (1994) The value of amniotic fluid interleukin-6 determination in patients with preterm labor and intact membranes in the detection of microbial invasion of the amniotic cavity. Am 3’ Obstet Gynecol, 171, 901-911. de Kretser DM (1993) Inhibin, activin, and follistatin: placental production and physiology. In Molecular Aspects of Placental and Fetal Membrane Autocoids (Ed.) Rice GE and Brennecke SP, CRC Press, pp. 239-250. de Kretser DM, Foulds LM, Hancock M & Robertson DM (1994) Partial characterization of inhibin, activin and follistatin in the term human placenta. 3 Clin Endocrinol Metab, 79, 502-507. Dudley DJ & Trautman MS (1994) Infection, inflammation, and contractions: the role of cytokines in the pathophysiology of preterm labor. Sem Reprod Endocrinol, 12, 263-273. Dudley DJ, Hunter C, Mitchell MD & Varner MW (1994) Clinical value of amniotic fluid interleukin-6 determinations in the management of preterm labour. Br 3 Obstet Gynaecol, 101, 592-597. Florio P, Lombard0 M, Gallo R, Di Carlo C, Sutton S, Genazzani AR & Petraglia F (1995) Activin-A, corticotropin releasing factor, and prosta-

glandin F,, increase immunoreactive oxytocin release from cultured human placental cells. Placenta, 17, 3077311. Gallinelli A, Gallo R, Genazzani AD, Matteo ML, Caruso A, Woodruff TK & Petraglia F (1996) Episodic secretion of activin A in pregnant women. Eur3 Endocrmol, 135, 340-344. Halgunset J, Johnsen H, Kjollesdal AM, Qvigstad E, Espevik T & Austgulen R (1994) Cytokine levels in amniotic fluid and inflammatory changes in the placenta from normal deliveries at term. EurJ7 Obstet Gynecol Reprod Bzol, 56, 1533160. Hillier SL, Witkin SS, Krohn MA, Watts DH, Kiviat NB, & Eschenbach DA (1990) The relationship of amniotic fluid cytokines and preterm delivery, amniotic fluid infection, histologic chorioamnionitis, and chorioamnion infection. Obstet Gynecol, 81, 941-948. Keelan JA, Song Y & France JT (1994) Comparative regulation of inhibin, activin and human chorionic gonadotrophin production by placental trophoblast cells in culture. Placenta, 15, 803-818. Keelan JA, Sato T & Mitchell MD (1997) Interleukin (IL)-6 and IL-8 production by human amnion cells in vitro: Regulation by cytokines, growth factors, glucocorticoids, phorbol esters and bacterial lipopolysaccharide. Biol Reprod, 57, 1438-1444. Kliman HJ, Nestler JE, Sermasi E, Sanger JM & Strauss III JF (1986) Purification, characterization, and in vitro differentiation of cytotrophoblasts from human term placentas. Endocrinology, 118, 1577-1582. Knight PG, Muttukrishnan S & Groome NP (1996) Development and application of a two-site enzyme immunoassay for the determination of ‘total’ activin-A concentrations in serum and follicular fluid. 3 Endocrinol, 148, 267-279.

434 Loveland KL, McFarlane JR & de Kretser DM (1996) Expression of activin beta C subunit mRNA in reproductive tissues. JMol Endocn’nol, 17, 61-65. Lundin-Schiller S & Mitchell, MD (1991) Prostaglandin production by human chorion laeve cells in response to inflammatory mediators. Placenta, 12, 353. Massague J (1990) The transforming growth factor-beta family. Anntuzl Reviews of Cell Biology, 6, 597-641. Mitchell MD (1994) Eicosanoid biosynthesis and its regulation during human pregnancy and parturition. In Textbook of Fetal Physiology (Ed.) Thorburn GD & Harding R, pp. 430-444. Oxford: Oxford University Press. Mitchell MD, Branch DW, Lundin-Schiller S, Romero RJ, Daynes RA & Dudley DJ (1991) Immunological aspects of preterm labour. Sem Perinatol, 15, 210-224. Muttukrishna S, George L, Fowler PA, Groome NP & Knight PG (1995) Measurement of serum concentrations of inhibin-A during human pregnancy. Clin Endocrinol, 42, 391-397. Okita JR, Sagawa N, Casey ML & Snyder JM (1983) A comparison of human amnion tissue and amnion cells in primary culture by morphological and biochemical criteria. In Vitro,9, 117-126. Petraglia F (1997) Inhibin, activin and follistatin in the human placenta-a new family of regulatory proteins. Placenta, 18, 3-8. Petraglia F, Vaughan J & Vale W (1989) Inhibin and activin modulate the release of gonadotropin-releasing hormone, human chorionic gonadotropin, and progesterone from cultured human placental cells. Proc Nat Acad Sci USA, 86, 51145117. Petraglia F, Gallinelli A & de Vita D (1994) Activin at parturition: changes of maternal serum levels and evidence for binding sites in placenta and fetal membranes. Obstet Gynecol, 84, 278-282. Petraglia F, Sawchenko P, Lim AT, Rivier J & Vale W (1987) Localization, secretion, and action of inhibin in human placenta. Science, 237, 187-189. Petraglia F, Calza L, Garuti GC, Abrate M, Giardino L, Genazzani AR, Vale W & Meunier H (1990) Presence and synthesis of inhibin subunits in human decidua. 3 Clin Endocrinol Metab, 71, 487-492. Petraglia F, Garuti GC, Calza L, Roberts V, Giardino L, Genazzani AR, Vale W & Meunier H (1991) Inhibin subunits in human placenta: localization and messenger ribonucleic acid levels during pregnancy. Am 3 Obstet Gynecol, 165, 750-758. Petraglia F, Garg S, Florio P, Sadick M, Gallinelli A, Wong WL, Krummen L, Cornitini G, Mather J & Woodruff TK (1993a) Activin-A and activin-B measured in maternal serum, cord blood serum and amniotic fluid during human pregnancy, Endocrinol 3, 1, 3233327. Petraglia F, Anseschi MM, Calza L, Garuti GC, Fusaro P, Giardino L, Genazzani AR & Vale W (1993b) Inhibin and activin in human fetal membranes: evidence for a local effect on prostaglandin release. 3 Clin Endocrinol Metab, 77, 542-548.

Placenta

(1998),

Vol. 19

Petraglia F, Gallinelli A, Grande A, Florio P, Ferrari S, Genazzani AR, Ling N & De Paulo LV (1994) Local production and actions of follistatin in human placenta. 3 Clin Endocrinol Metab, 78, 205-210. Petraglia F, de Vita D, Gallinelli A, Aguzzoli L, Genazzani AR, Romero R & Woodruff TK (1995) Abnormal concentrations of maternal serum activin-A in gestational diseases. 3 Clin Endocrinol Metab, 8, 558-561. Qu J & Thomas K (1993) Regulation of inhibin secretion in human placental cell culture by epidermal growth factor, transforming growth factors, and activin. 3 Clin Endocrinol Metab, 77, 925-931. Redinbaugh MG & Turley RB (1986) Adaption of the bicinchoninic acid protein assay for use with microtiter plates and sucrose gradient fractions. Anal Biochem, 153, 267-271. Romero R, Durum S Dinarello CA, Ovarzun E, Hobbins JC & Mitchell MD (1989) Interleukin-1 stimulates prostaglandin biosynthesis by human amnion. Prostaglandins, 37, 13-22 Romero R, Munoz H, Gomez R, Galasso M, Sherer DM, Cotton D & Mitchell MD (1994) Does infection cause premature labor and delivery? Sem Reprod Endocrinol, 12, 227-239. Shao L-E, Frigon Jr NL, Sehy DW, Yu AL, Lofgre J, Schwa11 R & Yu J (1992) Regulation of production of activin-A in human marrow stromal cells and monocytes. Exp Hematol, 20, 1235-1242. Shimonovitz S, Yagel S, Anteby E, Finci-Yeheskel Z, Adashi EY, Mayer M & Hurwitz A (1995) Interleukin-1 stimulates prostaglandin E production by human trophoblast cells from first and third trimesters. 3 Clin Endocrinol Metab, 80, 1641-1646. Song Y, Keelan JA & France JT (1995) Activin-A stimulates, while transforming growth factor-a inhibits, chorionic gonadotrophin production and aromatase activity in cultured human placental trophoblasts. Placenta, 17, 603-610. Takahashi S, Uchimaru K, Harigaya K, Asano S & Yamashita T (1992) Tumour necrosis factor and IL-l induce activin gene expression in a human bone marrow stromal cell line. Biochem Biophys Res Comm, 188, 310-317. Tanimoto K, Tamura K, Ueno N, Usuki S, Murakami K & Fukamizo A (1992) Regulation of activin PA mRNA level by CAMP. Biochem Biophys Res Comm, 182, 773-778. Wallace EM, Riley SC, Crossley JA, Ritoe SC, Horne A, Shade M, Ellis PM, Aitken DA & Groome NP (1997) Dimeric inhibins in amniotic fluid, maternal serum, and fetal serum in human pregnancy. 3 Clirt Endocrinol Metab, 82, 218-222. Yu J, Shao L-E, Frigon Jr NL, Lofgren J & Schwa11 R (1996) Induced expression of the new cytokine, activin A, in human monocytes: inhibition by glucocorticoids and retinoic acid. Immunology, 88, 368-374. Yui J, Garcia-Lloret M, Brown AJ, Berdan RC, Morrish DW, Wegmann TG & Guilbert LJ (1994) Functional, long-term cultures of human term trophoblasts purified by column-elimination of CD9 expressing Cells. Placenta, 9, 231-246.