TY - JOUR
AU - de Kretser, D.M.
AB - Abstract Activin A, a dimer of the βA-subunit of inhibin, has been shown to have multiple biological activities and sites of production. Follistatin is a high-affinity binding protein for activin, which neutralizes its activity. This study provides the first data, using a cross-sectional design, on the measurement of both these proteins in the maternal circulation of a large cohort of women (6–39 weeks of gestation, n = 2–20 women/time point) during normal pregnancies, and confirms that similar patterns are seen in nine women studied longitudinally during pregnancy. The concentrations of total activin A were measured using a specific two-site enzyme-linked immunosorbent assay (ELISA), and a new radioimmunoassay for measuring total follistatin in serum utilizing dissociating reagents to eliminate the interference of activin is described. At 38–39 weeks gestation, both activin A and follistatin concentrations rose to a peak (4.59 ± 0.54 ng/ml and 72.7 ± 3.31 ng/ml, respectively). The activin A and follistatin concentrations were highly correlated both in the cross-sectional study (P <0.0001) and in individual women in the longitudinal study (P <0.05–0.0001). Concentrations of follistatin showed a greater increase in the second trimester of pregnancy relative to activin A concentrations. The parallel increase in the secretion of these two proteins throughout pregnancy probably reflects feto-placental secretion. activin A, follistatin, human, pregnancy, radioimmunoassay Introduction Activins are homo- and heterodimers of the inhibin βA and βB subunits which form activin A, activin B and activin AB, and were identified initially by their capacity to stimulate the secretion of follicle stimulating hormone (FSH) by pituitary cells in culture (Ling et al., 1986; Vale et al., 1986). Subsequent studies have shown that these proteins have a wide range of actions in a diversity of tissues (Mather et al., 1997), all of which suggest that they should be considered as growth factors, in keeping with their classification into the transforming growth factor β family of proteins (Mason et al., 1985). The involvement of the activins in pregnancy emerged from the identification of mRNA for the α, βA and βB subunits and dimeric proteins in the human placenta (Meunier et al., 1988; Petraglia et al., 1991; de Kretser et al., 1994; Yokoyama et al., 1995) and fetal membranes (Petraglia et al., 1993a). High concentrations of activin A have been shown in ovine (Wongprasartsuk et al., 1994) and human (Petraglia et al., 1993b) amniotic fluid. With the development of a variety of more sensitive and specific assays for the measurement of activin A, elevated concentrations in human pregnancy were demonstrated (Petraglia et al., 1993b, 1994; Harada et al., 1996; Woodruff et al., 1997). A subsequent study with a larger cohort of women confirmed these findings (Muttukrishna et al., 1996). In early pregnancy, activin A concentrations have been shown to be higher than those of cycling women, and studies in women with non-functional ovaries have suggested a feto-placental origin for activin A (Birdsall et al., 1997; Lockwood et al., 1997; Muttukrishna et al., 1997a). Moreover, higher than normal activin A concentrations have been demonstrated in gestational diseases (Petraglia et al., 1995a,b), premature labour (Petraglia et al., 1997) and in pre-eclampsia (Muttukrishna et al., 1997b). The role of activin in various biological processes is modulated by follistatin, a high-affinity activin-binding protein, which can neutralize the majority of the actions of activin (Nakamura et al., 1990; Kogawa et al., 1991). Follistatin exists as two forms termed follistatin 288 (FS288) and the larger form, FS315, which arise through an alternative splicing action (Shimasaki et al., 1988) and a number of other forms which arise from proteolytic cleavage and glycosylation variants (Sugino et al., 1993). Follistatin has been isolated from human placenta (de Kretser et al., 1994) and is also found in fetal membranes, decidua and amniotic fluid (Petraglia et al., 1994; Wongprasartsuk et al., 1994). Recently, serum concentrations of follistatin have been shown to rise throughout pregnancy, although the numbers of subjects in the studies were limited (Wakatsuki et al., 1996; Woodruff et al., 1997). The high-affinity binding of activin by follistatin can affect the accuracy of the assays for either protein since the formation of the complex between activin and follistatin can interfere in the binding of the proteins to the antisera (de Kretser et al., 1994; McFarlane et al., 1996). This has led to the use of a variety of methods to remove this interference such as analyte denaturation (Knight et al., 1996), dissociating agents or detergents (McFarlane et al., 1996; Evans et al., 1998) and sample extraction (Harada et al., 1996) to enable the measurement of total activin or follistatin. This study utilizes two assays which address the complexities of the measurement of total activin and follistatin, to determine the concentrations of these proteins in a large cohort of pregnant women in whom the outcome of the pregnancy was normal. Materials and methods Patient samples Cross-sectional study: Serum samples were collected as previously described (Yohkaichiya et al., 1991). A single serum sample was collected from each of 220 pregnant women, at various stages of gestation and before the onset of labour, who visited the ante-natal clinic at Monash Medical Centre, Clayton, Victoria, Australia. Each pregnancy and its outcome was followed and samples were only included in the study if they satisfied the following criteria: (i) no gestational complications associated with metabolic, genetic, cardiovascular or hormonal disorders; (ii) gestation length >37 weeks; (iii) no fetal distress before or during labour; (iv) an APGAR score of 7+ and a birth weight of >2500 g; and (v) no operative delivery procedures. Longitudinal study: Sequential samples were taken at 2- to 4-weekly intervals from pregnant women who volunteered at their first ante-natal clinic visit. Nine women whose pregnancy satisfied the above criteria were included in the study. Hormone assays Activin A enzyme-linked immunosorbent assay (ELISA): Total serum activin A concentrations were measured using a two-site enzyme immunoassay specific for activin A as previously described (Knight et al., 1996). Human recombinant (hr) activin A, purified as described previously (Robertson et al., 1992) from material provided by Biotech Australia Pty Ltd (East Roseville, NSW, Australia), was used as standard. The mean intra- and inter-assay coefficients of variation (CV, %) were 9.8% and 12.3% respectively. The minimum detection limit was 0.05 ng/ml. Follistatin radioimmunoassay: Total follistatin concentrations were measured using a heterologous, discontinuous radioimmunoassay. The assay used a rabbit antiserum (#204) raised against purified 35 kDa bovine follistatin and human recombinant follistatin 288 (hrFS288; provided by the National Institute of Diabetes, Digestive and Kidney Disease, Bethesda, MD, USA) was used as standard. [125I]-hrFS288 was used as tracer, iodinated using Iodogen reagent (Pierce, Rockford, IL, USA). Dissociating reagents were used to remove the interference of activin in the assay. A combination of reagents (termed TDS reagent) at final concentrations of 6.67% Tween-20 (Sigma, St Louis, MO, USA), 3.33% sodium deoxycholate (BDH, Poole, Dorset, UK) and 0.13% sodium dodecyl sulphate (SDS; BioRad, Hercules, CA, USA) was used (modified from McFarlane et al., 1996). Immune complex precipitation was achieved by the use of a goat anti-rabbit second antibody (GAR#11; IRD, Monash University). The total serum concentration per tube was equalized using normal human serum (provided by the Australian Red Cross Blood Bank, South Melbourne, Victoria, Australia) immediately after addition of the second antibody. Radioimmunoassay procedure The assay buffer used was 0.1 M phosphate-buffered saline (PBS), pH 7.4 containing 0.5% bovine serum albumin (BSA; Sigma). Standard and samples (100μl) were diluted in assay buffer and incubated with 100 μl antiserum #204 (1/8000 diluted in assay buffer containing 1/1400 normal rabbit serum, NRS; IRD, Monash University) and 100 μl TDS reagent overnight at room temperature. Tracer was added at 10 000 c.p.m./100 μl of assay buffer containing 0.1% Triton X-100 (v/v) (Sigma) on day 2 and incubation continued overnight at room temperature. On day 3, 100 μl of second antibody GAR#11 (1/90) in assay buffer containing 0.05 M EDTA (BDH) was added to precipitate immune complexes and the total serum concentration per tube was equalized at the same time by adding 100 μl of serum/buffer to each tube to give a final concentration of 100 μl serum per tube. Following incubation overnight at 4°C, 2 ml cold 0.9% saline was added and the tubes were centrifuged at 4000 g for 45 min at 4°C. The supernatant was decanted and the pellets counted on a gamma counter. The sensitivity of the radioimmunoassay, based on a +2 logit value, was 1.4 ng/ml with a working range of 1.4 to 110 ng/ml. The intra-assay coefficient of variation based on the mean CV from six assays was 5.7% and the inter-assay CV based on a quality control pool of normal human serum was 9.5%. Radioimmunoassay validation: Samples of male and female human serum were serially diluted and assessed for their dose-dependence and parallelism to the standard. Cross-reactivity of hrFS315 was determined using material kindly provided by Dr H.Sugino, Institute for Enzyme Research, University of Tokushima, Japan. The accuracy of the assay was tested by measuring serum samples `spiked' with known amounts of hrFS288 (3.125–100 ng/ml) to assess quantitative recoveries. Interference in the radioimmunoassay from activin was assessed by pre-incubating increasing amounts (0–50 ng/tube) of hr activin A with normal human serum and assaying these with and without the TDS reagent in the radioimmunoassay. Data analysis Dose–response curves were linearized using logit log-dose transformation and analysed by parallel line statistics. Data from the cross-sectional study were grouped into 2-week intervals and represented as arithmetic mean ± SEM. For analysis, the data were log-transformed and analysed using a one-way analysis of variance (ANOVA) with Bonferroni post test to determine significant differences in concentrations between different stages of gestation. Correlations were calculated using Pearson correlation coefficients. Analyses were carried out using the GraphPad software package (GraphPad Software Inc., San Diego, CA, USA). Results Follistatin radioimmunoassay Validation studies Samples of human male and female serum serially diluted in a dose-dependent manner and were parallel to the standard (Figure 1). A pool of human serum spiked with known amounts of hrFS288 (3.125, 6.25, 12.5, 25, 50 and 100 ng/ml) gave recoveries of 103.4 ± 5.1, 105.9 ± 5.5, 110.6 ± 4.2, 108.1 ± 5.5, 96.9 ± 5.3 and 102.8 ± 8.6% (mean ± SEM, n = 4 assays) respectively. Cross-reactivity of hrFS315 in the radioimmunoassay against a hrFS288 standard was found to be 35.9%. Effect of TDS/interference of activin The interference of activin in the radioimmunoassay and the ability of the TDS reagent to dissociate the activin/follistatin complexes was assessed in an experiment where increasing amounts of activin (0–50 ng/tube) were pre-incubated in normal human serum and then assayed. In the absence of the dissociating reagent TDS, follistatin was undetectable in male human serum (Figure 2). Without dissociation, only free follistatin, and a small amount of complexed follistatin possibly where the epitope is exposed, can potentially be measured. It therefore appears that the majority of follistatin in serum exists in a complex with activin and as such is undetectable. However, in the presence of the TDS reagent, a level of 7.1 ng/ml of follistatin was measured in the serum. This was due to the dissociation of follistatin from the complex with activin, thereby allowing the measurement of `total' follistatin in serum. In the presence of TDS reagent, adding increasing amounts of activin A had no effect on the amount of follistatin measured (Figure 2). However, when no TDS reagent was present, increasing amounts of added activin resulted in an apparent rise in the amount of follistatin measured. This, we interpret, is due to the added free activin binding to the iodinated follistatin tracer and preventing it binding to the follistatin antiserum. This resulted in a smaller amount of iodinated follistatin bound to the antiserum being precipitated, which in turn resulted in an apparently higher concentration of follistatin being calculated. This effect was first seen when 0.625 ng/tube of activin was added and the amount of follistatin measured subsequently increased linearly as the amount of activin added was increased. Activin A and follistatin in pregnancy Cross-sectional study Mean serum activin A concentrations rose 69-fold (P < 0.001) throughout pregnancy from 0.07 ± 0.02 ng/ml at weeks 6–7 to a peak of 4.59 ± 0.54 ng/ml at 38–39 weeks (Figure 3). Activin A concentrations remained detectable but stable from 6 to 14 weeks (P > 0.05), but then showed a small but significant rise at 22–23 weeks (P < 0.05). Subsequently, there was a 6.1-fold increase (P < 0.001) from weeks 24–25 to 30–31 weeks where concentrations plateaued slightly and then rose steeply from weeks 32–33 to a peak at 38–39 weeks (3.0-fold, P < 0.001). Follistatin showed a progressive increase from early pregnancy through the second and third trimesters to term (Figure 3). Mean concentrations rose 5.8-fold (P < 0.001) from 7.4 ± 0.8 ng/ml at weeks 6–7 to a brief plateau at 30–31 to 32–33 weeks and then rose rapidly 1.9-fold (P < 0.001) from 32–33 weeks to a peak of 72.7 ± 3.3 ng/ml at 38–39 weeks. Both activin and follistatin concentrations showed a brief plateau at 30–31 to 32–33 weeks and then concentrations rose sharply and in a linear fashion towards term. The activin A and follistatin concentrations throughout pregnancy are highly correlated (r = 0.7796, P < 0.0001). Longitudinal study In all nine women followed longitudinally, activin and follistatin concentrations rose in parallel, beginning in the third trimester and rising to a peak towards term (Figure 4a–i). The activin A and follistatin concentrations in each woman were highly correlated across pregnancy in all nine women (r = 0.867–0.998, P < 0.05 to < 0.0001). The concentrations of follistatin and activin A during pregnancy in these nine women fell within the normal range established by the cross-sectional study (Figure 5a and b), except for one woman whose activin A concentrations were higher than the normal range until week 30 when the concentrations remained within the normal range until term, while her follistatin concentrations fell within the normal range. Discussion This study demonstrates the changes in the concentrations of follistatin and activin A in the maternal circulation throughout normal human pregnancy in a large cohort of women using a cross-sectional design, and in a more limited longitudinal study. In particular, the concentrations of both proteins were measured in the same set of samples from women in whom the course and outcome of the pregnancy, in terms of the mother and baby, were normal. These concentrations are of considerable significance since there is an increasing number of reports suggesting that the concentrations of activin A are abnormal in premature labour, gestational diseases and in pre-eclampsia (Petraglia et al., 1995a,b, 1997; Muttukrishna et al., 1997b), and the measurement of this protein may be required for diagnostic purposes. As such, the provision of normal ranges is crucial and these must be based on a sufficiently large cohort which is representative of the normal population at any stage of pregnancy. The assays used in this study have been carefully characterized and provide measurements of the total activin A and total follistatin. While activin does not interfere in the follistatin radioimmunoassay, it is recognized that the concentrations of follistatin may be an underestimate since cross-reactivity of hrFS315, which probably represents the major circulating form, is only 36% with reference to the hrFS288 assay standard. However, the results described here in the last trimester of pregnancy are in fact only slightly lower than the results of Wakatsuki et al. (1996) who appear to have the only assay reported which is directed specifically to the measurement of FS315. The activin A ELISA used in this study is not compromised by the presence of follistatin (Knight et al., 1996), in contrast to a number of other assays reported in the literature. However, the activin A concentrations are significantly less than those shown in the study by Muttukrishna et al. (1996) using the same ELISA with a different standard preparation. They report concentrations of 25 ng/ml (compared with 4.6 ng/ml) at the end of the third trimester, but the patterns of activin A concentrations across pregnancy are similar. There is a need for the establishment of international standards for the measurement of these proteins in conditions where the concentrations may be of clinical importance. This is of particular importance for the measurement of activin A since our studies of this protein in allantoic fluid have shown that preparations of activin A may run with different mobilities on high-performance liquid chromatography and reflect significantly differing potencies when measured in bioassays, ELISA and radioimmunoassay, all of which were proven to be specific for this protein (Foulds et al., 1998). The principal finding in this study is a parallel rise in the circulating concentrations of total follistatin and activin A as pregnancy progresses, confirming and extending earlier data (Petraglia et al., 1993b; Harada et al., 1996; Muttukrishna et al., 1996; Woodruff et al., 1997). The peripheral concentrations of follistatin and activin A are highly correlated throughout pregnancy, but there is a greater increase in follistatin concentrations in the second trimester and a steeper rise in activin A in the third trimester. The significance of this observation is unclear, as are the mechanisms through which the increases occur. The principal source of activin A and follistatin during pregnancy is likely to be the feto-placental unit, since these proteins have been isolated from the term placenta (de Kretser et al., 1994; Yokoyama et al., 1995) and activin A concentrations in maternal blood decline rapidly in the puerperium (Harada et al., 1996; Petraglia et al., 1997). Further, activin A concentrations were significantly higher in multiple pregnancies, wherein the placental mass is greater, compared with singleton pregnancies (Lockwood et al., 1997). Total activin A concentrations show a rapid decline 4 h after surgical termination of pregnancy (Muttukrishna et al., 1997a). However, the activin A concentrations had not returned to luteal phase concentrations, suggesting the ovary as a possible source for some of the activin in early pregnancy, or a low clearance rate of activin. The specific function of the elevated concentrations of activin A during pregnancy remains a matter of conjecture. Given the capacity for activin to cause immunomodulation, the elevated concentrations arising from a placental source may form part of the immune mechanisms protecting the fetus. In this regard, there is evidence that the early pregnancy trophoblast and fetal membranes express the βA subunit and may be related to the establishment of the placenta (Petraglia et al., 1993a). In this context, there are increasing data to support the view that follistatin may be elevated as part of the body's acute-phase reaction as it can be controlled by the cytokine interleukin-1β (Phillips et al., 1996) and can be stimulated by lipopolysaccharide (Klein et al., 1996; Michel et al., 1996). Given the very significant stress and tissue remodelling that occurs in the postpartum period, the continued elevation of follistatin concentrations in this period (Wakatsuki et al., 1996) may well be driven by these mechanisms. The greater elevation of both proteins in the second and third trimesters probably reflects placental secretion, but whether they have a role to play in the mother is unclear. Given that the rise in serum follistatin concentrations occurs earlier than activin, this may represent a protective measure since the pleiotrophic actions of activin could affect a large number of physiological processes throughout the body. The capacity of follistatin to bind and neutralize the actions of activin make such a role possible, and recent data indicate that this binding may direct the follistatin–activin complex into the cell and to a lysosomal pathway (Hashimoto et al., 1997). Additionally, previous studies have shown that the fast form of α2-macroglobulin can bind both activin A and follistatin (Phillips et al., 1997) and may serve to direct the complex into a degradative pathway through the lipoprotein receptor pathway. Depending on the balance between the relative concentrations of activin and its binding protein, follistatin, it is possible that differences in free activin concentrations between women may be reflected in differences in the timing of the cascade of events that initiate parturition. In this context, Draper et al. (Draper et al., 1997) have shown that uterine smooth muscle represents a binding site for labelled activin A injected into the circulation of rats in late gestation, and they suggested that this may be involved in the initiation of uterine contractility. Further studies are necessary to determine the role of these proteins in pregnancy. In summary, the results presented in this study provide total follistatin and activin A concentrations in the circulation of a large cohort of patients during a normal pregnancy and provide the basis for the exploration of patterns of secretion of these proteins during abnormal pregnancy. Figure 1. View largeDownload slide Linearity and parallelism of human male (▪) and female (▵) serum compared with human recombinant FS288 standard (○). Figure 1. View largeDownload slide Linearity and parallelism of human male (▪) and female (▵) serum compared with human recombinant FS288 standard (○). Figure 2. View largeDownload slide The ability of the follistatin radioimmunoassay to measure total follistatin in serum, with (○) and without (•) the TDS dissociating reagent, in the presence of increasing amounts of human recombinant (hr) activin A. Figure 2. View largeDownload slide The ability of the follistatin radioimmunoassay to measure total follistatin in serum, with (○) and without (•) the TDS dissociating reagent, in the presence of increasing amounts of human recombinant (hr) activin A. Figure 3. View largeDownload slide Mean (± SEM) concentrations of total activin A (○) and total follistatin (•) in the maternal circulation during human pregnancy (6–39 weeks) from a cross-sectional study. The number above each point indicates the number of samples per time point. Figure 3. View largeDownload slide Mean (± SEM) concentrations of total activin A (○) and total follistatin (•) in the maternal circulation during human pregnancy (6–39 weeks) from a cross-sectional study. The number above each point indicates the number of samples per time point. Figure 4. View largeDownload slide Profiles of serum concentrations of total activin A (○) and total follistatin (FS, •) in the circulation of nine women (a–i) from samples taken longitudinally across their pregnancies. The dotted line indicates parturition. Figure 4. View largeDownload slide Profiles of serum concentrations of total activin A (○) and total follistatin (FS, •) in the circulation of nine women (a–i) from samples taken longitudinally across their pregnancies. The dotted line indicates parturition. Figure 5. View largeDownload slide Peripheral serum profiles of (a) total activin A and (b) follistatin from nine women (samples taken longitudinally across their pregnancies) versus the normal range (mean ± 2SD, indicated by the bold line) of activin A and follistatin established from the cross-sectional study. Figure 5. View largeDownload slide Peripheral serum profiles of (a) total activin A and (b) follistatin from nine women (samples taken longitudinally across their pregnancies) versus the normal range (mean ± 2SD, indicated by the bold line) of activin A and follistatin established from the cross-sectional study. 3 Present address: Division of Animal Physiology, University of New England, Armidale, NSW 2351, Australia 4 Present address: Pfizer Pharmaceuticals Inc., Shinjuku-ku, Tokyo 163-0461, Japan 5 To whom correspondence should be addressed This work was funded by the National Health and Medical Research Council of Australia. References Birdsall, M., Ledger, W., Groome, N. et al. ( 1997) Inhibin A and activin A in the first trimester of human pregnancy. J. Clin. Endocrinol. Metab. , 82, 1557–1560. Google Scholar de Kretser, D.M., Foulds, L.M., Hancock, M. et al. ( 1994) Partial characterization of inhibin, activin, and follistatin in the term human placenta. J. Clin. Endocrinol. Metab. , 79, 502–507. Google Scholar Draper, L.B., Chong, H., Wang, E. et al. ( 1997) The uterine myometrium is a target for increased levels of activin A during pregnancy. Endocrinology , 138, 3042–3046. Google Scholar Evans, L.W., Muttukrishna, S. and Groome, N.P. ( 1998) Development, validation and application of an ultra-sensitive two-site enzyme immunoassay for human follistatin. J. Endocrinol. , 156, 275–282. Google Scholar Foulds, L.M., de Kretser, D.M., Farnworth, P. et al. ( 1998) Ovine allantoic fluid contains high concentrations of activin A: partial dissociation of immunoactivity and bioactivity. Biol. Reprod. , 59, 233–240. Google Scholar Harada, K., Shintani, Y., Sakamoto, Y. et al. ( 1996) Serum immunoreactive activin A levels in normal subjects and patients with various diseases. J. Clin. Endocrinol. Metab. , 81, 2125–2130. Google Scholar Hashimoto, O., Nakamura, T., Shoji, H. et al. ( 1997) A novel role of follistatin, an activin-binding protein, in the inhibition of activin action in rat pituitary cells. J. Biol. Chem. , 272, 13835–13842. Google Scholar Klein, R., Clarke, I.J., Hedger, M.P. et al. ( 1996) Plasma follistatin concentrations increase following lipopolysaccharide administration in sheep. Clin. Exp. Pharm. Physiol. , 23, 754–755. Google Scholar Knight, P.G., Muttukrishna, S. and Groome, N.P. ( 1996) Development and application of a two-site enzyme immunoassay for the determination of `total' activin-A concentrations in serum and follicular fluid. J. Endocrinol. , 148, 267–279. Google Scholar Kogawa, T., Nakamura, T., Sugino, H. et al. ( 1991) Activin-binding protein is present in pituitary. Endocrinology , 128, 1434–1440. Google Scholar Ling, N., Ying, S.-Y., Ueno, N. et al. ( 1986) Pituitary FSH is released by a heterodimer of the β-subunits from the two forms of inhibin. Nature , 321, 779–782. Google Scholar Lockwood, G.M., Ledger, W.L., Barlow, D.H. et al. ( 1997) Measurement of inhibin and activin in early human pregnancy: demonstration of fetoplacental origin and role in prediction of early-pregnancy outcome. Biol. Reprod. , 57, 1490–1494. Google Scholar Mason, A.J., Hayflick, J.S., Ling, N. et al. ( 1985) Complementary DNA sequences of ovarian follicular fluid inhibin show precursor structure and homology with transforming growth factor-β. Nature , 315, 659–663. Google Scholar Mather, J.P., Moore, A. and Li, R.-H. ( 1997) Activins, inhibins, and follistatins: further thoughts on a growing family of regulators. Proc. Soc. Exp. Biol. Med. , 215, 209–222. Google Scholar McFarlane, J., Foulds, L., Piscotta, A. et al. ( 1996) Measurement of activin in biological fluids by radioimmunoassay, utilizing dissociating agents to remove the interference of follistatin. Eur. J. Endocrinol. , 134, 481–489. Google Scholar Meunier, H., Rivier, C., Evans, R.M. et al. ( 1988) Gonadal and extragonadal expression of inhibin, βA and βB subunits in various tissues predicts diverse functions. Proc. Natl Acad. Sci. USA , 85, 247–251. Google Scholar Michel, U., Schneider, O., Kirchho, F. et al. ( 1996) Production of follistatin in porcine endothelial cells: differential regulation by bacterial compounds and the synthetic glucocorticoid RU 28362. Endocrinology , 137, 4925–4934. Google Scholar Muttukrishna, S., Fowler, P.A., George, L. et al. ( 1996) Changes in peripheral serum levels of total activin A during the human menstrual cycle and pregnancy. J. Clin. Endocrinol. Metab. , 81, 3328–3334. Google Scholar Muttukrishna, S., Childs, T.G., Groome, N.P. et al. ( 1997) Source of circulating levels of inhibin A, pro alpha C-containing inhibins and activin A in early pregnancy. Hum. Reprod. , 12, 1089–1093. Google Scholar Muttukrishna, S., Knight, P.G., Groome, N.P. et al. ( 1997) Activin A and inhibin A as possible endocrine markers for pre-eclampsia. Lancet , 349, 1285–1288. Google Scholar Nakamura, T., Takio, K., Eto, Y. et al. ( 1990) Activin-binding protein from rat ovary is follistatin. Science , 247, 836–838. Google Scholar Petraglia, F., Garuti, G.C., Calza, L. et al. ( 1991) Inhibin subunit in human placenta: localization and messenger ribonucleic acid levels during pregnancy. Am. J. Obstet. Gynecol. , 165, 750–758. Google Scholar Petraglia, F., Anceshi, M.M., Calza, L. et al. ( 1993) Inhibin and activin in human fetal membranes: evidence for a local effect on prostaglandin release. J. Clin. Endocrinol. Metab. , 77, 542–548. Google Scholar Petraglia, F., Garg, S., Florio, P. et al. ( 1993) Activin A and activin B measured in maternal serum, cord blood serum, and amniotic fluid during pregnancy. Endocrine J. , 1, 323–327. Google Scholar Petraglia, F., Gallinelli, A., De Vita, D. et al. ( 1994) Activin and parturition: changes of maternal serum levels and evidence for binding sites in placenta and fetal membranes. Obstet. Gynecol. , 84, 278–282. Google Scholar Petraglia, F., Aguzzoli, L., Gallinelli, A. et al. ( 1995) Hypertension in pregnancy: changes in activin A maternal serum concentration. Placenta , 16, 447–454. Google Scholar Petraglia, F., De Vita, D., Gallinelli, A. et al. ( 1995) Abnormal concentration of maternal serum activin A in gestational diseases. J. Clin. Endocrinol. Metab. , 80, 558–561. Google Scholar Petraglia, F., Di Blasio, A.A., Florio, P. et al. ( 1997) High levels of fetal membrane activin βA and activin receptor IIB mRNAs and augmented concentration of amniotic fluid activin A in women in term or preterm labor. J. Endocrinol. , 154, 95–101. Google Scholar Phillips, D.J., Hedger, M.P., McFarlane, J.R. et al. ( 1996) Follistatin concentrations in male sheep increase following sham castration/castration or injection of interleukin-1β. J. Endocrinol. , 151, 119–124. Google Scholar Phillips, D.J., McFarlane, J.R., Hearn, M.T. et al. ( 1997) Inhibin, activin and follistatin bind preferentially to the transformed species of α-2 macroglobulin. J. Endocrinol. , 155, 65–71. Google Scholar Robertson, D.M., Foulds, L.M., Prisk, M. et al. ( 1992) Inhibin/activin β-subunit monomer: isolation and characterization. Endocrinology , 130, 1680–1687. Google Scholar Shimasaki, S., Koga, M., Esch, F. et al. ( 1988) Primary structure of the human follistatin precursor and its genomic organization. Proc. Natl Acad. Sci. USA , 85, 4218–4222. Google Scholar Sugino, K., Kurosawa, N., Nakamura, T. et al. ( 1993) Molecular heterogeneity of follistatin, an activin-binding protein: higher affinity for the carboxyl-terminal truncated forms for heparan sulfate proteoglycans on the ovarian granulosa cell. J. Biol. Chem. , 268, 15579–15587. Google Scholar Vale, W., Rivier, C., Vaughan, J. et al. ( 1986) Purification and characterization of a FSH releasing protein from porcine ovarian fluid. Nature , 321, 776–779. Google Scholar Wakatsuki, M., Shintani, Y., Abe, M. et al. ( 1996) Immunoradiometric assay for follistatin: serum immunoreactive follistatin levels in normal adults and pregnant women. J. Clin. Endocrinol. Metab. , 81, 630–634. Google Scholar Wongprasartsuk, S., Jenkin, G., McFarlane, J.R. et al. ( 1994) Inhibin and follistatin concentrations in fetal tissues and fluids during gestation in sheep: evidence for activin in amniotic fluid. J. Endocrinol. , 141, 219–229. Google Scholar Woodruff, T.K., Sluss, P., Wang, E. et al. ( 1997) Activin A and follistatin are dynamically regulated during human pregnancy. J. Endocrinol. , 152, 167–175. Google Scholar Yohkaichiya, T., Polson, D., O'Connor, A. et al. ( 1991) Concentrations of immunoactive inhibin in serum during human pregnancy: evidence for an ovarian contribution. Reprod. Fertil. Steril. , 3, 671–678. Google Scholar Yokoyama, Y., Nakamura, T., Nakamura, R. et al. ( 1995) Identification of activins and follistatin proteins in human follicular fluid and placenta. J. Clin. Endocrinol. Metab. , 80, 915–921. Google Scholar © European Society of Human Reproduction and Embryology
TI - Serum activin A and follistatin concentrations during human pregnancy: a cross-sectional and longitudinal study
JF - Human Reproduction
DO - 10.1093/humrep/14.3.827
DA - 1999-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/serum-activin-a-and-follistatin-concentrations-during-human-pregnancy-zHiX0bb9UT
SP - 827
EP - 832
VL - 14
IS - 3
DP - DeepDyve
ER -