A critical role for PKC zeta in endothelin-1-induced uterine contractions at the end of pregnancy.

We have previously shown that protein kinase C (PKC) zeta and/or PKC delta are necessary for endothelin-1 (ET-1)-induced human myometrial contraction at the end of pregnancy (Eude I, Paris P, Cabrol D, Ferré F, and Breuiller-Fouché M. Biol Reprod 63: 1567-1573, 2000). Here, we report that the selective inhibitor of PKC delta isoform, Rottlerin, does not prevent ET-1-induced contractions, whereas LY-294002, a phosphatidylinositol (PI) 3-kinase inhibitor, affects the contractile response. This study characterized the in vitro contractile response of cultured human pregnant myometrial cells to ET-1 known to induce in vitro contractions of intact uterine smooth muscle strips. Cultured myometrial cells incorporated into collagen lattices have the capacity to reduce the size of these lattices, referred to as lattice contraction. Neither the selective conventional PKC isoform inhibitor, Gö-6976, or rottlerin affected myometrial cell-mediated gel contraction by ET-1, whereas this effect was blocked by LY-294002. We found that treatment of myometrial cell lattices with an inhibitory peptide specific for PKC zeta or with an antisense against PKC zeta resulted in a significant loss of ET-1-induced contraction. Evidence is also presented by using confocal microscopy that ET-1 induced translocation of PKC zeta to a structure coincident with the actin-rich microfilaments of the cytoskeleton. We have shown that PKC zeta has a role in the actin organization in ET-1-stimulated cells. Accordingly, our results suggest that PKC zeta plays a role in myometrial contraction in pregnant women.

Interleukin-1beta induces glycosaminoglycan synthesis via the prostaglandin E2 pathway in cultured human cervical fibroblasts.

The aim of this study was to identify, in cultured human cervical fibroblasts, the mechanisms by which interleukin (IL)-1beta induces the synthesis of glycosaminoglycans (GAG) and to explore the putative role of prostaglandin E(2) (PGE(2)) in this process. Exposure of the cells for 24 h to IL-1beta induced a significant (P < 0.05) dose-dependent increase in GAG synthesis. IL-1beta (1 ng/ml) induced the expression of cyclooxygenase-2 (COX-2) protein 6 h after treatment, accompanied by a 7.5-fold increase in PGE(2) production. We confirmed that NS398, a selective COX-2 inhibitor, dose-dependently blocked PGE(2) augmentation following IL-1beta treatment. AH23848, the selective EP(4) receptor antagonist, completely abolished IL-1beta-induced GAG synthesis, whereas AH6809, an EP(2) receptor antagonist, had no effect on the stimulatory effects of IL-1beta. Furthermore, we demonstrated that 6 h exposure to IL-1beta induced a notable increase in EP(4) receptor mRNA expression and a decrease in EP(1) receptor mRNA but had no effect on the expression of EP(2) and EP(3) receptor transcripts. In conclusion, these findings indicate that IL-1beta not only induced GAG synthesis by increasing COX-2 protein expression and the subsequent PGE(2) production but also enhanced the responsiveness of cervical fibroblasts to PGE(2) by selectively up-regulating EP(4) receptor mRNA expression. These results suggest that PGE(2) may regulate human cervical ripening in an autocrine/paracrine manner via EP(4) receptors.

Is up-regulation of phosphodiesterase 4 activity by PGE2 involved in the desensitization of beta-mimetics in late pregnancy human myometrium?

Elevation of cAMP content resulting from stimulation of the receptor-adenylyl cyclase complex is involved in maintaining the quiescence of the human myometrium during pregnancy. The magnitude of this elevation is critically influenced by the rate of cAMP hydrolysis by phosphodiesterase (PDE) isoenzymes. In the present study we report that in term myometrium, enhanced cAMP-specific PDE4 activity takes part in the heterologous desensitization to the beta-mimetic, salbutamol. Indeed, pretreatment with a PDE4-selective inhibitor potentiates the relaxant effect of salbutamol on myometrial strips of women at term. Furthermore, the reduced relaxant effect of salbutamol after long-term treatment of myometrial strips with PGE2, a potent myometrial effector, can be reversed by PDE4 inhibition. Using a model of cultured myometrial cells, we also demonstrated that PGE2 is able to up-regulate PDE4 activity, at least through the induction of synthesis of PDE4B and PDE4D short forms, which, in turn, dampen the cAMP accumulation provoked by the stimulation of adenylyl cyclase. Such data suggest that in late pregnancy endogenous PGE2 might up-regulate PDE4 activity and lessen the responsiveness of myometrium to beta-mimetic activation. Accordingly, coapplication of a selective PDE4 inhibitor might greatly improve the usefulness of beta-mimetic in tocolysis.

Potentiation response of cultured human uterine leiomyoma cells to various growth factors by endothelin-1: role of protein kinase C.

OBJECTIVE: Factors responsible for the abnormal proliferation of myometrial cells that accompanies leiomyoma formation are unknown, although steroid hormones and peptide growth factors have been implicated. We hypothesized that endothelin-1 (ET-1) is a physiological regulator of tumor growth. DESIGN: In this study, we investigated the role of ET-1 on growth of human leiomyoma cells and its synergistic effect with growth factors, as well as the signaling pathway involved in this interaction. METHODS: Leiomyoma cell proliferation was assayed by [H]thymidine incorporation and cell number. Protein kinase C (PKC) isoforms were analyzed by Western blot using specific antibodies. RESULTS: ET-1 on its own was unable to stimulate DNA synthesis but potentiated the leiomyoma cell growth effects of basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), IGF-I and IGF-II. The failure of a protein tyrosine kinase (PTK) inhibitor, tyrphostin 51, to affect the potentiating effect of ET-1, supports the hypothesis of non-involvement of PTK in this process. The inhibition of PKC by calphostin C or its down-regulation by phorbol 12,13-dibutyrate (PDB) eliminated the potentiating effect of ET-1, but did not block cell proliferation induced by the growth factors alone. Five PKC isoforms (alpha, beta1, epsilon, delta and zeta) were detected in leiomyoma cells, but only phorbol ester-sensitive PKC isoforms (PKCalpha, epsilon and delta) contribute to the potentiating effect of leiomyoma cell growth by ET-1. CONCLUSIONS: We have demonstrated that ET-1 potentiates leiomyoma cell proliferation to growth factors through a PKC-dependent pathway. These findings suggest a possible involvement of ET-1 in the pathogenesis of leiomyomas.

EP(4) receptors mediate prostaglandin E(2)-stimulated glycosaminoglycan synthesis in human cervical fibroblasts in culture.

The aim of this study was to determine the prostaglandin E (EP) receptors and second messengers implicated in glycosaminoglycan (GAG) synthesis by human cervical fibroblasts in culture. Human cervical fibroblasts were obtained from cervical biopsies in pre-menopausal, cycling women. Cultured cells were incubated with prostaglandin E(2) (PGE(2)) and an array of agonists and antagonists. Glycosaminoglycan synthesis was assayed after extraction by measuring the [(3)H]glucosamine and [(35)S]sulphate incorporated into GAG and cAMP production was determined by radioimmunoassay. PGE(2) significantly stimulated GAG synthesis. Neither 17-phenyl-trinor-PGE(2), the EP(1) selective agonist, nor sulprostone, an EP(3) agonist, had any effect on GAG production. Butaprost, the EP(2) selective agonist, also failed to increase GAG synthesis. AH6809, an EP(2) antagonist, had no effect on PGE(2)-stimulated GAG production. AH23848, an EP(4) antagonist, inhibited the GAG synthesis provoked by PGE(2). PGE(2) and butaprost significantly increased cAMP production. Both AH6809 and AH23848 inhibited the PGE(2)-stimulated cAMP production. H89, a cAMP-dependent protein kinase (PKA) inhibitor, did not inhibit PGE(2)-stimulated GAG synthesis and Sp-cAMPS, a selective PKA activator, failed to increase GAG production. In conclusion, both EP(4) and EP(2) receptors are present and functional in human cervical fibroblasts. Only EP(4) receptors mediate PGE(2) stimulated GAG synthesis in a PKA-independent pathway.

Effects of progesterone on prostaglandin E(2)-induced changes in glycosaminoglycan synthesis by human cervical fibroblasts in culture.

Prostaglandins are known to induce cervical ripening and this effect may be mediated by an increase in glycosaminoglycan (GAG) concentration. The aim of this study was to assess the effects of progesterone on prostaglandin E(2) (PGE(2))-induced changes in GAG synthesis by human cervical cells in culture. Human cervical fibroblasts were obtained by cervical biopsies in hormonally active women and cultured. Cells were submitted to an incubation with progesterone or control medium. A second incubation was then performed with increasing concentrations of PGE(2). GAG synthesis by the cervical cells was assayed after extraction, by incorporation of [(3)H]-glucosamine and [(35)S]-sulphate into GAGs. It was found that progesterone alone induced a dose-dependent increase in GAG synthesis. After pre-incubation with progesterone, PGE(2) further increased [(3)H]-glucosamine and [(35)S]-sulphate uptake. However, when expressed as percentage of stimulation, the stimulatory effect of PGE(2) on GAG synthesis was inhibited at high progesterone concentrations. Therefore we concluded that, although high concentrations of progesterone increase the overall synthesis of GAG, they may also play a preventative role against PGE(2)-induced changes in GAG production during pregnancy.

Does high polyunsaturated free fatty acid level at the feto-maternal interface alter steroid hormone message during pregnancy?

Polyunsaturated fatty acids (PUFA) are important in pregnancy, fetal development and parturition. We measured free fatty acids (FFA), albumin and alpha-fetoprotein (AFP) in the maternal and fetal circulations of women undergoing elective Caesarean section at term. We also studied the impact of PUFAs on estrogen (ER) and progesterone receptors (PR) binding properties in vitro in the myometria of pregnant women and ex vivo in human myometrial cells in culture. FFA in intervillous blood (I) (feto-maternal interface) and maternal peripheral blood (M) were similar, while those in the umbilical vein (V) and arteries (A) were 2-4 fold lower (P<0.001). PUFA levels were low in M and 3 fold higher in I, A and V (P< 0.001); consequently C20:4 and C22:6 were most abundant in intervillous space. Albumin was uniformly distributed throughout the maternal-fetal unit, but there was a transplacental gradient in AFP. The AFP in the intervillous space had a special conformation (less immuno-reactive, more anionic), suggesting loading with PUFA. Physiological concentrations of C20:4 stimulated estradiol binding, but inhibited progestin binding. C20:4 inhibited progesterone binding by decreasing the number of binding sites, with no change in apparent affinity, in vitro in myometrial tissue and ex vivo in myometrial cells. Thus PUFA may modulate the steroid hormone message, so that the high C20:4 concentration at the maternal-fetal interface at term may help amplify the estrogen signal and inhibit the progesterone signal.

Selective up-regulation of phosphodiesterase-4 cyclic adenosine 3',5'-monophosphate (cAMP)-specific phosphodiesterase variants by elevated cAMP content in human myometrial cells in culture.

In human myometrium, the modulation of intracellular cAMP content resulting from agonist-mediated stimulation of the receptor-adenylyl cyclase complex is largely influenced by the rate of cAMP hydrolysis by phosphodiesterase (PDE) isoenzymes. We have previously shown that the PDE4 family contributes to the predominant cAMP-hydrolyzing activity in human myometrium and that elevation of the PDE4B2 messenger RNA steady state level occurs in pregnant myometrial tissue. In the present study, we used a model of human myometrial cells in culture to determine whether an elevated cAMP concentration could influence PDE expression. As in myometrial tissue, high levels of PDE4 activity were detected in these smooth muscle cells. Long term treatment with 8-bromo-cAMP or forskolin resulted in a selective induction of PDE4B and of PDE4D short form messenger RNA variants. Concurrently, an increased immunoreactive signal for the PDE4B- and PDE4D-related isoenzymes was detected. This induction was consistent with an observed significant up-regulation of PDE4 activity. Accordingly, our results demonstrate that in human cultured myometrial cells, cAMP-elevating agents manipulate PDE4 activity through selective induction of synthesis of PDE4B and PDE4D short forms. Such a mechanism might have physiological importance during pregnancy by dampening hormonal stimulation and could thereby be involved in tolerance to the tocolytic effect of beta-adrenoceptor agonists.

Effect of pregnancy on PDE4 cAMP-specific phosphodiesterase messenger ribonucleic acid expression in human myometrium.

In light of the important role of the second messengers cAMP and cGMP in the mechanism of relaxation in the human myometrium, specific regulation of the phosphodiesterase (PDE) enzymatic system responsible for cyclic nucleotide inactivation is essential. We previously identified in the human myometrium PDE4 cAMP-specific PDE as by far the most abundant isoform. Here we have studied the expression patterns of mRNAs for the four cloned human PDE4 genes in the myometria of pregnant and non-pregnant women. Concurrent expression of the PDE4A, 4B, 4C and 4D genes is demonstrated. We found that the PDE4D transcripts are the most prominently expressed. PDE4A and PDE4B mRNAs also are markedly abundant, whereas lower expression is observed for PDE4C mRNAs. Interestingly, we showed that transcripts of PDE4B2 are more abundant in the myometria of pregnant women than in non-pregnant women, whereas no difference between the two tissues was detected for PDE4A, 4C and 4D mRNAs. Cultured human myometrial cells, which present a high level of PDE4 activity and express the four PDE4 mRNA subtypes, provide us with an appropriate model to further evaluate whether the level of expression of the PDE 4B2 mRNA subtype is under hormonal regulation.

Role of endothelin-1 in regulating proliferation of cultured human uterine smooth muscle cells.

Endothelin-1 (ET-1) exhibits vasoconstricting and growth-promoting properties in vascular smooth muscle. Whether ET-1 has mitogenic properties in uterine smooth muscle cells, and which ET receptor subtype mediates this response, is unknown. The present study was undertaken to examine the proliferative potential of the ET family on human myometrial cells in culture. ET-1 stimulated DNA synthesis and proliferation of myometrial cells. The absence of a stimulating effect of endothelin-3 (ET-3) or the ETB agonist sarafotoxin 6c (S6c) was observed. The proliferative effect of 100nM ET-1 was blocked by the two ETA antagonists (BQ 123 and FR 139317), whereas the ETB antagonist IRL 1038 was ineffective. These data indicated that ET-1-induced DNA synthesis was mediated only by the ETA receptor subtype. Pertussis toxin (PTX) pretreatment completely abolished this effect, indicating that this pathway was coupled to the ETA receptor via the Gi protein family. PTX treatment partially decreased serum-induced DNA synthesis. This suggests that some factors from serum may operate via the G-protein in initiation of mitogenesis. Insulin-like growth factors (IGFs), epidermal growth factor (EGF) and insulin were found to be mitogens in the absence of serum, and they had no potentiating effect on ET-1-induced DNA synthesis. In the presence of 0.5% serum, EGF alone caused a weak increase in DNA synthesis, while all the growth factors were able to reduce the proliferative effect of ET-1. These findings on human myometrial cells in culture raise the possibility that, under certain conditions, ET-1 may function as a positive or as a negative modulator of smooth muscle proliferation.

Synthesis of glycosaminoglycans by human cervical fibroblasts in culture: effects of prostaglandin E2 and cyclic AMP.

OBJECTIVE: To study the mechanism of action of prostaglandin E2 (PGE2) and its analogue sulprostone leading to production of glycosaminoglycans (GAGs) in the human uterine cervix. STUDY DESIGN: We analysed the effects of PGE2 and its analogue sulprostone upon production of adenosine 3',5'-monophosphate (cAMP), in human cultured fibroblasts. We also studied the effects of PGE2, sulprostone and a cAMP analogue (8-Bromo-cAMP), on the incorporation of [3H]glucosamine into GAGs in human cervical fibroblasts in culture. RESULTS: Following treatment with PGE2 (10(-4)-10(-6) M), we observed a significant increase in the production of cAMP from 96.3 +/- 8.4 pmol/10(6) cells without phosphodiesterase inhibitor 3-isobutyl-methylxanthine (IBMX) to 325 +/- 63 pmol/10(6) cells with 10(-4) M IBMX (Spearman correlation test; P < 0.05). Under the same conditions, the effects of sulprostone (10(-6) M) were limited (from 8.1 +/- 1.5 to 51.3 +/- 14.1 pmol/10(6) cells without and with IBMX, respectively; not significant). Both PGE2 and 8-bromo-cAMP (from 10(-12) to 10(-4) M) increased [3H]glucosamine uptake into GAGs (Spearman correlation test; P < 0.05). Sulprostone (10(-12)-10(-4) M) was unable to reproduce such an effect even after a 24 or 48 h treatment. CONCLUSION: Since firstly, PGE2 acts through EP1, EP2 and EP3 specific receptors, whereas the action of sulprostone is only mediated by EP1 and EP3, and secondly EP2 receptor is coupled with cAMP production, we conclude that cAMP is involved in mediating the action of PGE2 upon GAG synthesis by human cultured cervical fibroblasts.

Inhibition of prostaglandin E2 production in myometrial and amniotic cells in culture by human amniotic fluid. Loss of inhibition after intra-uterine fetal death.

OBJECTIVE: To study the effects of amniotic fluid obtained during early second trimester of normal human pregnancies on prostaglandin production in cultured myometrial and amniotic cells. STUDY DESIGN: Cultured human myometrial and amniotic cells were incubated with human amniotic fluid obtained by amniocentesis and fractionated by ultrafiltration. Prostaglandin E2 was measured in the incubation media. In one case, amniotic fluid was obtained during the days following intra-uterine fetal death (IUFD). RESULTS: PGE2 production in myometrial and amniotic cells was significantly decreased when incubated with the fraction of amniotic fluid containing molecules of molecular weight between 3 and 30 kD. This inhibition was still present after heating. After IUFD, the inhibitory activity of amniotic fluid was persistent for the first 3 days but had disappeared 6 days after IUFD. CONCLUSION: These data suggest that a factor contained in amniotic fluid, with a molecular weight within the 3- to 30-kD range, and possibly produced or controlled by the fetus, inhibits PG synthesis. Further work is necessary to characterize this factor and evaluate its physiological role in human parturition.

Myosin heavy chain isoform expression in human myometrium: presence of an embryonic nonmuscle isoform in leiomyomas and in cultured cells.

We had previously found no myosin heavy chain (MHC) changes in expression during pregnancy in human myometrium. In the present work, we compared the MHC pattern of expression in normal human myometrium, pregnant and non-pregnant, to that in benign tumors of the uterine musculature and in cultured myometrial cells. We used a high-resolution gel electrophoretic system and monoclonal antibodies directed against smooth muscle and nonmuscle MHCs. Smooth muscle MHCs (SM1, 204 kDa, and SM2, 200 kDa, MHCs) and a nonmuscle MHC of 196 kDa (NM MHC) were detected in pregnant and nonpregnant human myometrium. Pregnant myometrium was found to differ from nonpregnant myometrium by its slightly lower content in NM MHC, whereas the ratio of SM1/SM2 was equivalent. In leiomyomas and in cultured cells grown from human myometrium explants, SM1, SM2, and NM MHCs were also expressed. In addition, a nonmuscle MHC of 198/200 kDa (SMemb MHC), which was present in a fetal human uterus but not in adult normal tissue, was observed in leiomyomas and in cultured cells. Expression of SM1 and SM2 MHCs was variable in the different leiomyomas studied. In cultured cells, SM1 and SM2 MHC content was low, but it was enhanced by suppression of serum after cell confluency. Present results confirm that pregnancy-associated smooth muscle cell hypertrophy is not accompanied by major changes in MHCs. In contrast, cell culturing and cell hyperplasia leading to leiomyoma formation induce substantial modifications in MHCs, including the occurrence of a second type of nonmuscle MHC.

Induction of labor and cervical maturation using mifepristone (RU 486) in the late pregnant rat. Influence of a cyclooxygenase inhibitor (Diclofenac).

The mechanism of action of RU 486 (Mifepristone), an antiprogesterone compound, on labor induction and on cervical maturation, is still not well documented. We have investigated the effect of RU 486, alone and in association with a cyclooxygenase inhibitor (Diclofenac) on the induction of preterm delivery and on concomitant changes in the distribution of cervical glycosaminoglycans (GAGs) in pregnant Wistar rats: a control group (n = 18), a RU 486 treated group (n = 36), and a RU 486 and Diclofenac treated group (n = 15). The results of this study confirm the ability of this antiprogesterone treatment to induce preterm delivery in the rat. This effect was antagonized by cyclooxygenase inhibition, suggesting that the action of RU 486 on labor induction could be mediated by prostaglandins. The absence of an increase in plasma prostaglandin E2 (PGE2) levels in RU 486 treated animals could be explained by local uterine changes in prostaglandin concentrations. Mifepristone also induced some of the biochemical features of cervical maturation (i.e. increased hydration and hyaluronic acid concentration). This effect was not inhibited in Diclofenac treated animals suggesting that factors other than prostaglandins play a role in this phenomenon.

Cyclooxygenase and lipoxygenase inhibitors--induced changes in the distribution of glycosaminoglycans in the pregnant rat uterine cervix.

In order to study the hormonal control mechanisms of cervical maturation, we investigated cyclooxygenase and 5-lipoxygenase inhibitors-induced changes in the distribution of glycosaminoglycans (GAG) in pregnant Wistar rat uterine cervices at term. The GAG were measured in a control (n = 11), in a Diclofenac (cyclooxygenase inhibitor) treated group (n = 8), in a BW 755C (dual inhibitor of cyclooxygenase and 5-lipoxygenase) treated group (n = 6), and a L 651392 (5-lipoxygenase inhibitor) treated group (n = 9). The results of these studies suggest, that cervical hyaluronic acid metabolism and cervical hydration are controlled in association by prostaglandins and leukotrienes (and perhaps by other phospholipids metabolites), whereas heparan sulphate metabolism is obviously controlled by prostaglandins. Nevertheless complete and normal cervical maturation is probably controlled in association by arachidonic acid metabolites and other factors (steroids and peptides).

Prostaglandin E2-induced changes in the distribution of glycosaminoglycans in the isolated rat uterine cervix.

In order to study the hormonal control mechanisms of cervical ripening, we investigated the prostaglandin E2 (PGE2)-induced changes in the distribution of glycosaminoglycans (GAG) using hysterectomized and ovariectomized rats, leaving the vascularized uterine cervix in situ, as an animal model. In the first series of experiments, the GAG were measured in a control (n = 22 Wistar rats) and in a PGE2-treated group (n = 20 Wistar rats) without steroid supplementation. In the second series of experiments, the GAG were measured in a control (n = 19) and in a PGE2-treated group (n = 18) receiving estradiol and progesterone supplements. After PGE2 treatment in the two series of experiments, and despite being surgically isolated from the uterine corpus, the cervix was still able to undergo some of the structural changes associated with normal ripening (increased hydration and hyaluronic acid concentration). This suggests that PGE2, acting directly on the cervix, could be, at least in part, a modulator of biochemical events which underlie normal cervical maturation. The animal model described here seems to be suitable for studying the hormonal mechanisms of cervical ripening and the regulatory relationship between cervical maturation and myometrial contractility, which are probably subject to concordant endocrine regulation.

Pregnancy-related changes in the distribution of glycosaminoglycans in the cervix and corpus of the human uterus.

We investigated the variations in the distribution of glycosaminoglycans (GAG) in the cervical, isthmic and corporeal stroma of the pregnant and non-pregnant human uterus. The GAG were measured in non-pregnant (n = 9) and pregnant (n = 14) cervices, in non-pregnant (n = 8) and pregnant (n = 6) uterine isthmus, and in non-pregnant (n = 12) and pregnant (n = 5) uterine corpus. Heparan sulfate is abundant in the isthmic and corporeal stroma. This has to be related to a larger content in blood vessels and muscular cells at this level than in the cervix. The distribution of hyaluronic acid and dermatan sulfate is the same in the cervix as in the corpus, but the amount of these GAGs is lower in the former, which could be due to a lower connective tissue content. Two of the main features of cervical maturation, increased hydration, and decreased dermatan sulfate concentration, were found also in the pregnant corporeal stroma. These results suggest that connective tissue maturation is not only a cervical phenomenon and could play a role in active and/or passive mechanical properties of the myometrium during late pregnancy and labor.

[Attempt at determining fetal sex by the simultaneous study of the HY antigen and testosterone in the amniotic fluid]. / Essai de détermination du sexe foetal par l'étude simultanée de l'antigène H-Y et de la testostérone dans le liquide amniotique.

Attempts have been made to determine fetal sex early by a simultaneous study of H-Y antigen and testosterone in amniotic fluid. H-Y (male) antigen was serologically detected in the amniotic cells from both sexes, but the results obtained from 473 amniotic fluids showed no significant sex differences either in fresh cells or after culture. Testosterone assays done from 263 samples of liquor confirm that the results obtained in cases of male fetuses were significantly higher than in females. 55% of all males and 47% of females could thus be detected to a level of 100% certainty. The probability of having a girl decreases as an inverse function of testosterone levels and comes out at 85% for testosterone less than 200 pg/ml. On the other hand the probability of a male fetus is 95% when the values are above 240 pg/ml. The determination of the testosterone/DHA ratio in order to try to eliminate the influence of the variations in volume has not improved the accuracy of this diagnosis.

[Hydramnios and fetal malformations. Use of estimating the level of alpha-fetoprotein and of bilirubin in the amniotic fluid during the last trimester of pregnancy (author's transl)]. / Hydramnios et malformations foetales. Intérêt du dosage de l'alpha-foetoprotéine et de la bilirubine amniotique au cours du dernier trimestre de la grossesse.

Alpha-fetoprotein (AFP) levels in amniotic fluid have been measured by electro-immunodiffusion methods in the last trimester of pregnancy. 122 samples have been taken by abdominal amniocentesis during 58 pregnancies without hydramnios and 51 pregnancies with hydramnios. The level of AFP was always less than 1.2 mcg/ml when the children were normal in the latter group. On the other hand levels were higher than 1.2 mcg/ml in 17 out of 19 abnormal children. These had 5 neural tube abnormalities, 1 hydrocephalus, 3 duodenal atresias, 1 oesophageal atresia, 1 hare lip, 1 omphalocoele and 5 different malformations. The two abnormal children who had levels of AFP within normal limits were hydrocephalic. The diagnostic value of measuring bilirubinaemia which was also measured in 22 cases has been demonstrated.