Does or can human placenta produce prostacyclin?

Placenta (1986), 7, 37-42

Does or Can H u m a n Placenta Produce Prostacyclin?

M. J. N. C. KEIRSE a, J. J. H. M. ERWICH & G. KLOK Department of Obstetrics, University of Leiden, Academisch Ziekenhuis, RijnsburgerwegIo, 2333 AA Leiden, The Netherlands To whom correspondence shouM be addressed Paper accepted H.7.I985

INTRODUCTION Ever since Myatt and Elder (I977) described the potent antiaggregatory activity of the human placenta and ascribed this to prostacyclin (PGIz) , there has been considerable argument as to whether the placenta does or does not produce PGI 2(Elder, Myatt and Jogee, i983; Jogee, Myatt and Elder, x983a; Jeremy et al, I983, I985). The question is especially relevant, since some authors (e.g., Lewis, i983) have postulated that this particular prostaglandin evolved in order to 'cope' with placentation, while others reported PGI 2 deficiency in conditions associated with defective placentation such as pregnancy-induced hypertension (reviewed by Keirse, Moonen and Klok, I985), fetal growth retardation (Jogee, Myatt and Elder, i983a ) and other states of 'chronic placental insufficiency' (Stuart et al, I98I). Nevertheless, there is little agreement on the placental capacity for PGI 2 production, as there are unexplained discrepancies between the results from superfused placental tissue (Mitchell et al, I978), from placental cells in culture (Jogee, Myatt and Elder, i983a , I983b), from tissue minces (Jeremy et al, I985) and from broken cell preparations (Demb61&Duchesne et al, I98I; Christensen and Gr&n, i983). This recently led Jeremy et al 0985) to pose the crucial question: does human placenta produce prostacyclin? However, in dealing with this question they in fact encompassed three questions. First, 'does' the placenta produce PGIff Second, 'can' the placenta produce PGIff Third, can the placental antiaggregatory activity in vitro be explained by PGI 2 synthesis? Only one of these questions (the last) was given a firm, albeit negative, answer, and it is thus clear that the in vitro antiaggregatory activity of the placenta cannot be ascribed to PGI 2 synthesis. This observation is consistent with other data (Hutton et al, 198o; Dembel& Duchesne et al, 1982), though it cannot totally exclude any contribution by PGI2, however minimal, to that antiaggregatory activity. The answer to the first question was left delightfully vague (Jeremy et al, t985) as it should be. Indeed, a proper answer would require in vivo data that are difficult to obtain in view of the abundance ofPGI 2 synthase (Moonen, Klok and Keirse, I984) and PGI 2synthesis (Christensen and Green, i983) in the uterus itself on the one hand, and the relative inaccessibility of the placenta on the other. Thus the question of whether the placenta 'can' produce PGI z has remained as relevant as ever. Indeed, while Jeremy et al 0985) reported the absence ofPGI 2 synthesis, as identified by its hydrolysis product 6-oxo-prostaglandin FI~ (6-oxo-PGFl~), from 14C-arachidonic acid in 37

38

Placenta (z986), Vol. 7

placental tissue minces, they also demonstrated detectable endogenous levels of 6--oxo-PGFl~ in this preparation. We therefore investigated whether the human placenta does at least possess the enzyme (PGI 2 synthase) necessary for P G I 2 synthesis, concentrating on the first trimester of pregnancy for two reasons in particular. First, because that period of gestation is characterized by rapid placental development, and second, in order to avoid potential interference from a fully developed placental vasculature.

MATERIAL AND M E T H O D S Placental tissue was obtained with informed consent from 12 women undergoing termination of pregnancy for social reasons between 7 and 12 weeks of gestation. Only pregnancies of certain duration were included in the study. Products were collected, immediately rinsed in cold tapwater, and identified pieces of trophoblast were snap-frozen in liquid nitrogen and stored at - 8 o ~ until analysis. Frozen tissue samples were sliced and homogenized in a #ass-glass potter at 4~ in o. x M Tris-hydrochloric acid buffer, pH 8.o, containing 2 mM magnesium chloride, 0.2 5 i sucrose and i mM diethyldithiocarbamic acid. The homogenate was centrifuged at 5oo g for 5 min; the supernatant was recentrifuged at 8ooog for 15 min and the resulting supernatant again centrifuged at Io5 ooo g for 6o min. These procedures were carried out at 4~ and the lO5 ooo-g pellet (microsomal fraction) was resuspended in sucrose-free homogenization buffer. Protein concentrations of the microsomal suspensions were determined with the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA) following the standard procedure and using bovine serum albumin as reference protein. Quantification of P G I 2 synthase in the microsomal suspensions was based on the doubleantibody immunoradiometric assay of DeWitt and Smith (1982) with the modifications introduced by Moonen, Klok and Keirse (1984). The assay makes use of two monoclonal antibodies which bind to different antigenic sites on the enzymes. One of these (isn-3; Figure i) was iodinated with Bolton-Hunter reagent, monoiodo, 125I, as described (Moonen, Klok and Keirse, 1984). The other (isn-i; Figure i) was linked to Staphylococcus aureus cells via rabbit anti-mouse IgG, and this complex was coated overnight with excess purified rabbit IgG to diminish non-specific precipitation later in the assay (Moonen, Klok and Keirse, 1984). Microsomes isolated from bovine aortae (Moonen, Klok and Keirse, 1984) were used as a standard for PGI 2 synthase. Both standards and samples were solubilized with Triton X-ioo and diluted with assay buffer (o.I M Tris-hydrochloric acid, pH 8.o, containing o. 5 per cent (w/v) Triton X-Ioo) for use in the assay at different dilutions. Standards and samples were incubated with iodinated antibody (isn-3; 75 ooo cpm) for 3~ min, following which the Staphylococcus

~

~

protein A

O. aureus

12sI ~it 1

rabbit k , ~ _ . j ~ isn- 1T M anti-mouse IgG

Figure I. Schemeof the interactionsinvolvedin the immunoradiometricassayof prostacyclinsynthasc. The namesof the monoclonalantibodies(isn-1 and isn-3)referto the hybridomalinesfromwhichthey wereobtained(by courtesyof Professor W. L. Smith, Michigan, USA).

Keirse, Erwich, Klok: Prostacyclin Production in Human Placenta

39

aureus-linked antibody (isn-i; Figure I) was added for a further 3~ min incubation, all at room temperature (Moonen, Klok and Keirse, I984). The entire complexes, schematically represented in Figure l, were then collected by centrifugation, washed with assay buffer, and counted in a gamma counter. Microsomal suspensions were always incubated in duplicate at three different concentrations. The values of the samples, which were within the linear part of the standard curve, were averaged to calculate the amount of 125I precipitated per millilitre of solubilized microsomal suspension. This was then converted into counts per minute per microgram of microsomal protein.

RESULTS Results are shown in Figure 2. PGI 2 synthase was present in the microsomal fraction of all placentae investigated between 7 and 12 weeks of gestation, albeit in variable amounts (mean value: 6o cpm/#g protein). Its concentration per microgram of microsomal protein was iooofold lower than that in the standard preparation from bovine aortae, and also i oo--fold lower than PGI 2 synthase concentrations recorded in both pregnant and non-pregnant myometrium (Moonen, Klok and Keirse, x984). 140

m

_

120

0

e-

"$

1oo

000

0

80 .-~ 60 40 o

n

0

0

-0

0

20 _

0

0

0 -

!

t

7

8

I

I

I

I

9

10

11

12

weeks of gestation Figure z. Concentrations of placental PGI 2 synthase (in cpm/gg microsomal protein) in relation to gestational age. Results are the mean of duplicate measurements at three different concentrations.

DISCUSSION This investigation, using two specific monoclonal antibodies against prostacyclin (PGI2) synthase (DeWitt and Smith, I98z), now provides conclusive evidence that the human placenta 'can' produce PGI2, though it does not show whether the placenta 'will' do so, either in vivo or in vitro. In view of the wide divergence of opinions on placental PGI 2 production, it is worth emphasizing that at least one of these antibodies has been successfully applied to the purification

40

Placenta (1986), Vol. 7

of PGI 2 synthase by immunoaffmity chromatography (DeWitt and Smith, I983). For the same reason, it may be necessary to indicate that the presence of an enzyme, and its antigenic determinants in particular, does not necessarily imply that it will be present in an active and noninhibited form. Since there are large differences in protein content between early and late placentae (Keirse, Erwich and Klok, 1985) , we have, for methodological reasons, limited our study to the first trimester. However, it has been shown that PGI 2 production by cultured placental cells is greater at term than in the first trimester (Jogee, Myatt and Elder, I983b). In addition, we may now exclude the argument that the findings can be attributed to contamination with gross placental vasculature (Jeremy et al, I983), though there is a great deal of other evidence to indicate that PGI 2 synthase is not confined to vascular structures (Smith, DeWitt and Allen, 1983; Keirse, Moonen and Klok, 1984). Our observation that PGI 2synthase is definitely present in the placenta may now reconcile the apparent contradictions between previous findings. When summarized, previous findings succeeded in detecting PGI 2 in various placental preparations only when measured in terms of inhibition of platelet aggregation (Myatt and Elder, 1977) or by 6-oxo-prostaglandin F t~ (6-oxo-PGFI~) radioimmunoassay (Mitchell et al, I978; Jogee, Myatt and Elder, I983a, I983b; Jeremy et al, 1985). Both of these methods are known to be particularly prone to errors in specificity (Hutton et al, I980; Demb~l~-Duchesne et al, 198I). On the other hand, when investigated in terms of conversion of radioactive substrate, whether it be arachidonic acid (Christensen and Green, 1983; Jeremy et al, 1985) or a prostaglandin endoperoxide (Demb~l~Duchesne et al, 198 I), into 6-oxo-PGFl~ the general impression has been that no such bioconversion took place (Jeremy et al, 1985). Thus far, only Christensen and G r i n (1983) found some conversion, and this was limited to only 5 of 22 placentae investigated and to yields that were less than o. i per cent of the added substrate. However, conversion of arachidonic acid to PGI 2 and thenceforth to 6oxo-PGFl~ not only requires PGI 2 synthase activity but also cyclo-oxygenase (prostaglandin endoperoxide synthase) activity, which is known to be lower in the placenta than in any other tissue from the pregnant uterus (Keirse, 1975; Christensen and G r i n , 1983). Furthermore, the placenta is known to possess several enzymes that can utilize the intermediate prostaglandin endoperoxide (Demb~l~-Duchesne et al, 1981; Christensen and Green, 1983). Although the low level of placental PGI 2 synthase, in comparison with other tissues (Moonen, Klok and Keirse, I984) , may in itself explain why so little bioconversion into PGI 2 and thenceforth into 6-oxo-PGFl~ is observed, the properties of the enzyme need to be considered too. PGI 2 synthase is a ferrihaemoprotein that belongs to the family of cytochrome P-45 o enzymes (UIMch, Castle and Weber, 198t; DeWitt and Smith, 1983; Ullrich and Graf, 1984) and is hence particularly prone to inactivation. It is thus strongly inhibited by the lipoxygenase product, is-hydroperoxy-arachidonic acid (Moncada et al, I976), and by some other hydroperoxy fatty acids (DeWitt and Smith, i983), and it also shows self-inactivation (DeWitt and Smith, I983). Such processes can easily occur when cellular integrity is disturbed during the preparation of placental samples for subsequent bioconversion studies. Thus the low level of placental PGI 2 synthase and the properties of the PGI 2 synthase enzyme can easily explain the apparent discrepancies between the results of bioconversion and other in vitro investigations. On the whole, they certainly emphasize the difficulty of finding an in vitro model that will appropriately reflect the situation in vivo. In conclusion, when correctly interpreted, our data show no contradiction with any of the previous studies. On the contrary, they appear to indicate that even among these studies the contradictions are more apparent than real. Therefore, it should now be possible to reach general agreement that the placenta 'can' produce PGI 2 and that its potential for doing so is relatively

Keirse, Erwich, Klok: Prostacyclin Production in Human Placenta

4x

limited. Regrettably, however, it remains entirely uncertain whether the h u m a n placenta actually 'does' or 'does not' produce P G I 2 in vivo.

SUMMARY I n order to resolve contradictory data on the capacity of the h u m a n placenta to produce prostacyclin (PGI2) , we have quantified P G I 2 synthase in i2 placentae obtained in the first trimester o f pregnancy. U s i n g a specific immunoradiometric assay and two monoclonal antibodies against the enzyme, we found P G I 2 synthase to be present in the microsomal fraction of all placentae investigated, albeit in concentrations that were xooo-fold lower than in bovine aortal microsomes and too-fold lower than in both pregnant and n o n - p r e g n a n t myometrium. Comparison of these data with previous reports on placental P G I 2 production suggests that the contradictions between previous data are more apparent than real. W e conclude that the h u m a n placenta possesses the potential for P G I 2 production, that it therefore 'can' produce PGI2, b u t that it remains uncertain whether the placenta actually 'does' or 'does not' produce P G I 2 in vivo.

ACKNOWLEDGEMENTS We are grateful to ProfessorW. L. Smith, Department of Biochemistry,Michigan State University, USA, for a generous donation of antibodies used in this study, and to Ms H. Wittenberg for assistancewith the manuscript. The research was supported by the Netherlands Foundation for Medical Research (FUNGO, grant 13-44-42)and the Praeventiefonds, The Hague (grant 28-iii8). REFERENCES Christensen, N. J. & Gr6en, K. (x983) Bioconversionof arachidonic acid in human pregnant reproductive tissues. Biochemical Medicine, 3o, i6z-tSo. Demb~16-Duchesne, M. J., Thaler-Dao, H., Chavis, C. & Crastes de Paulet, A. (I98I) Some new prospects in the mechanism of control of arachidonate metabolism in human placenta and amnion. Prostaglandins, ~2, 979-Ioo2. Demb~16-Duchesne, M. J., Thaler-Dao, H., Chavis, C. & Crastes de Paulet, A. (1982) The human placental antiaggregating factor is neither prostacyclin nor a prostacyclin metabollte. Prostaglandins, 24, 7ox-7149 DeWitt, D. L. & Smith, W. L. 0982) Monoclonal antibodies against PGI 2synthase: an immunoradiometricassay for quantitating the enzyme. Methods in Enzymology, 86, 24o-246. DeWitt, D. L. & Smith, W. L. 0983) Purification of prostacyclin synthase from bovine aorta by immunoaffinity chromatography. Journal of Biological Chemistry, 258, 3285-3293. Elder, M. G., Myatt, L. & Jogee, M. 0983) Decreased prostacyclin production by placental cells in culture from pregnancies complicated by fetal growth retardation. British Journal of Obstetrics and Gynaecology, 9o, Io98-xo99. Hutton, R. A., Dandona, P., Chow, F. P. R. & Craft, I. L. (~98o) Inhibition of platelet aggregation by placental extracts. Thrombosis Research, x7, 465-47 x. Jeremy, J. g., Barradas, M. A., Mikhailidis, D. P. & Dandona, P. 0983) Decreased prostacyclin production by placental cells in culture from pregnancies complicated by fetal growth retardation. British Journal of Obstetrics and Gynaecology, 90, xo97-xo98. Jeremy, J. Y., Barradas, M. A., Craft, I. L., Mikhailidis, D. P. & Dandona, P. 0985) Does human placenta produce prostacyclin? Placenta, 6, 45-52. Jogee, M., Myatt, L. & Elder, M. G. (x983a) Decreased prostacyclin production by placental cells in culture from pregnancies complicated by fetal growth retardation. British Journal of Obstetrics and Gynaecology, 9o, 247-250. Jogee, M., Myatt, L. & Elder, M. G. (I983b) Prostacyclinproduction by human placental cells in short-term culture. Placenta, 4, 219-23o. Keirse, M.J.N.C. (I975) Studies onprostaglandinsin relation to humanparturition. D.Phil. thesis, University of Oxford. Keirse~ M. J. N. C., Erwich, J. J. H. M. & Klok, G. 0985) Increase in placental 15-hydroxy-prostaglandin dehydrogenase in the first half of human pregnancy. Prostaglandins, 3o, I31-14o. Keirse, M. J. N. C., Moonen, P. & Klok, G. (1984) Prostacyclin synthase in pregnant human myometrium is not confined to the utero-placental vasculature. 1RCS Medical Science, 12, 824-825.

42

Placenta (s986), Vol. 7

Keirse, M. J. N. C., Moonen, P. & Kiok, G. (1985) Control of prostacyclin synthesis in pregnancy-induced hypertension. Prostaglandins, 29, 643~5 o. Lewis, P. J. (1983) Does prostacyclin deficiency play a role in preeclampsia? In Prostacyclinin Pregnancy(Ed.) Lewis, P. J., Moncada, S. & O'Grady, J. pp. 2tS-2t8. New York: Raven Press. Mitchell, M. D., Bibby, J. G., Hicks, B. R. & Turnbull, A. C. 0978) Possible role for prostacyclin in human parturition. Prostaglandins, 16, 931-937. Moncada, S., Cn-yglewski, R. J., Bunting, S. &Vane, J. R. ( 1976) A lipid peroxide inhibits the enzyme in blood vessel microsomes that generates from prostaglandin endoperoxides the substance (prostaglandin X) which prevents platelet aggregation. Prostaglandins, x2, 715~737 . Moonen, P., Klok, G. & Keirse, M. J. N. C. (I984) Increase in concentrations of prostaglandin endoperoxide synthase and prostacyclin synthase in human myometrium in late pregnancy. Prostaglandins, 28, 3o9-321. Myatt, L. & Elder, M. (1977) Inhibition of platelet aggregation by a placental substance with prostacyclin-like activity. Nature, ~68, 159-i6o. Smith, W. L., DeWitt, D. L. & Allen, M. L. (x983) Bimodal distribution of the prostaglandin I z synthase antigen in smooth muscle cells. Journal of Biological Chemistry, 258 , 5922-5926. Stuart, M. J., Clark, D. A., Sunderji, S. G. et al (I98I) Decreased prostacyclin production: a characteristic of chronic placental insufficiency syndromes. Lancet, i, 1126-1128. Ulirich, V. & Graf, H. 0984) Prostacyclin and thromboxane synthase as P-45 o enzymes. Trends in Pharmacological Science, 5, 352-355. Ullrieh, V., Castle, L. & Weber, P. (i98i) Spectral evidence for the cytochrome P45o nature ofprostacyclin synthase. Biochemical Pharmacology, 30, 2o33-2o36.