Plucmtu (1991), 12,521-531
Human Embryo Modulates Placental Functibn in the First Trimester; Effects of Neural Tissues upon Chorionic Gonadotropin and Progesterone Secretion
REVITAL SHURTZ-SWIRSKI, ROBERT J. SIMON, YAEL COHEN EYTAN R. BARNEA”
&
Feto-Placental Endocrine Unit, Rappaport Institute and Department of Obst&cs/Gynaecology Rambam Medical Centre, Technion, Haif, Israel a To whom correspondence and reprint requests should be addressed Paper accqted 164.1991
SUMMARY We investigated
the efect
of embyonal
neural and adrenal
tissues (7-14
weeks
gestational age) upon BhCG semetion by homologous placental explants in staticand
dynamic cultures. In static co-culture si’ijicant inhibition by SC and brain was notedat 7-9 weeks.Similarly, in superjiision,using a novelco-chamberingtechnique there was a significantreductionin the area under the curve but notpeakfrequency of spontaneous pulsatileBhCG secretion.Incubationswith neural tissues11 weeksand abozlecaused a stimulatoryeflectupon BhCG secretionin both models. The effectof adrenal tissuein staticcultureswasd@&nt, namelyslightlyinhibitoryat 7-9 weeks and inhibit0y at 11 weeksand above.In superjiision,the effectof adrenal tissuewas not sign@ant. Extracted neural tissue 7-9 weeks incubated with placental explants exhibited inhibitoryeffectsupon BhCG secretionas well. Buffer-based extractsof neural tissues efiect was more pronounced than alcoholbased extractsregarding BhCG semetion. The effect of extractswas dose-dependent and effectswere notedup until a 2000-fold dilution. In contrast,the buffer SC extract had no efect on progesterone (P) secretion while the alcohol extract eflect was inhibit0y at 7-9 weeksand stimulatoy at > I I weeks. Superfiised explantspatternofl3hCG secretionwasinhibitedby one minutepulseof the SC buffer extract. In conclusion,the human neural tissueofembyonal origin may modulateplacental hCG and P secretionduring early pregnancy. 01~3-4004/91/050521 + 11 $05.00/O
@J 1991 Baillike Tindall Ltd
522
Placenta(1991), Vol. 12
INTRODUCTION A viable trophoblast is a prerequisite for normal embryonic development. However, it has not been established, in the human, whether a viable embryo is necessary for maintaining placental function (Hertig, Rock and Adams, 1956). It has been shown histologically that, cultures of trophoblast obtained shortly after embryonal death appear to be damaged (Hustin and Gaspard, 1977). There is a close temporal association between trophoblastic function, as reflected by the peak of hCG secretion at 9 gestational weeks, and the shift of responsibility for progesterone secretion from the corpus luteum to the placenta. At the same time, with the completion of embryogenesis, the transition from embryo to fetus takes place (Adamsons, 1968). Several strands of evidence suggest that there is an interdependence in mammals between the trophoblast and the embryo. Fetal death in both sheep and rats results in a decline in placental lactogen secretion (Ramsay et al, 1985; Albrecht, Haskins and Pepe, 1984; Taylor et al, 1984; Robertson, Owens and McCoshen, 1984). Despite many investigations upon the endocrine control of the trophoblast, the endogenous compound(s) that actually modulate the widely changing pattern of circulating hCG in the first trimester have not yet been identified. There is some evidence that, in humans, embryonal dysfunction or death modifies placental endocrine function. Indeed, it was recently shown that elevated circulating hCG levels in the second trimester can serve as a screening test for detecting certain chromosomal aberrations in the fetus (Bogart, Pandian and Jones, 1987). However the mechanism behind this abnormal rise is unknown. We have recently reported that extracts of visceral organs of human embryonal origin affect hCG secretion by cultured placental explants (Barnea, Simon and Kol, 1989). In that study it was not established whether the effect was due to compounds present only within the tissue itself or whether compounds were also secreted by the various organs, thus having an endocrine effect. In addition, we have recently shown that in superfused placental explants hCG secretion is episodic and is modulable by various hormones (Barnea and Kaplan, 1989). Therefore, in the present study the effect of neural and adrenal tissue co-cultured with placental explants upon hCG and progesterone secretion was examined. This was carried out in both static cultures where feedback between placenta and embryo is operative and in dynamic co-chambering where, because of the unidirectional flow, only the embryo can influence the placenta without feedback. In addition, the effect of neural extracts upon hCG secretion in the two models was examined. These results show that, at 7-9 weeks, neural tissue inhibited hCG secretion in both static and dynamic cultures, while at 11 weeks and later the effect was stimulatory. This pattern is characteristic to neural tissues, since adrenal affected the endocrinological behavior of the placenta differently.
MATERIALS
AND METHODS
Placental and embryonic material Twenty 7-14-week-old placentas and embryos were studied. After obtaining appropriate consent, elective pregnancy terminations by vacuum curretage were performed at the Rambam Medical Center. Collected tissue was rinsed several times in 0.9 per cent NaCl to remove all blood, followed by three further rinses with culture medium [Dulbecco’s
Shurt..-Swirski et al: Human Embyonal
Organs Effect on Placenta
Modified Eagle’s Medium (DMEM), Beit Haemek, Israel], U/ml, streptomycin 20 yg/ml and amphotericin B 50 pug/ml.
523
containing
penicillin
5000
Neural and adrenal tissue preparation Embryos were removed aseptically and gestational ages were confirmed by crown-rump measurements. The SC, which was removed from the spinal column, the brain and adrenal tissues were washed in a large excess of 0.9 per cent saline solution followed by several rinses with media containing 1 per cent antibiotics. The neural tissue was cut to 3-4 mm size fragments and was added to placental explant culture. The entire adrenal gland was cocultured with placental explants. Neural tissue extraction In other experiments the SC was extracted by sonication for 60 set (Triacregler-Ystral GmbH D-7801, Dottringhen) with one ml of (a) absolute ethyl alcohol, or (b) phosphate buffer 0.05 14, pH 7.4 containing 10 mu dithiothreitol (DTT) and 2 mu phenylmethylsulfonylfluoride (PMSF). Following sonication the particulated matter was sedimented by centrifugation at 800g for 10 min. The supernatant fraction was removed and saved. The supernatant from the alcohol extracts was dried and later resuspended in 1 ml of distilled water. The extracts were stored at -20°C until used. Explant cultures Using previously reported procedures (Barnea et al, 1989; Barnea et al, 1990), explants 5070 mg wet weight were dissected and rinsed in 1 per cent antibiotic solutions. For culture, explants were placed in DMEM media with or without the tested organs or organ extracts. At least three replicates per test agent or control were plated at 37°C in an atmosphere of 9.5 per cent air/5 per cent CO*. After 24 h of incubation, the media were collected and stored at - 20°C until assay. The tissue was saved for protein analysis. Dishes containing only vehicle served as controls. Additional culture dishes containing DMEM and neural tissue or adrenal, but not trophoblast explants were cultured and the media was collected and assayed for PhCG, hCG and progesterone (P). In addition, the /3hCG content of neural tissue extracts was also determined. Superfusion studies The method was recently published (Barnea and Kaplan, 1989; Barnea et al, 1990). Briefly, a superfusion apparatus (Accusyst, Endotronics, St. Paul, MN) with a multichannel peristaltic pump and fraction collector (model 272, ISCO, Durham, NC) were used to study the shortterm dynamics of hCG secretion. The explants (200-300 mg wet wt) were placed in the culture chambers and a DMEM solution containing HEPES (18 mM) was washed through in an atmosphere of 5 per cent COJ95 per cent air at 37°C. Experiments were conducted for a 2 h period; a I ml sample from the effluent was collected every 2.4 min for hCG measurements. In each experiment, one channel served as control and four served as experimental channels. At given intervals, a 1 ml pulse of SC extract was given through a peristaltic pump equipped with a digital flowmeter (ISMATECH DD, Chicago, IL). Co-chambering experiments In some experiments, placental explants were co-chambered with SC fragments by placing the two tissues serially in a double chamber separated by a metal mesh which allowed the
524
Placenta(1991), Vol. I2
medium to tlow unidirectionally from SC to placenta during superfusion. Media collected were stored at -20°C until assayed. Assays
As reported previously, /3hCG was measured by radioimmunoassay using hCG-MAIA clone assay (Serono) (Barnea, Simon and Kol, 1989; Barnea et al, 1986b; Barnea et al, 1990), with an intra-assay variability of 1.7 per cent. The interassay variability was 3.2 per cent. The lit of assay sensitivity was 11 weeks. This can be considered as evidence of the extractibility of the factors, high potency, and specificity, i.e. PBS-based extract was more potent than the alcohol-based extract on /?hCG secretion. In contrast, the aqueous extract had no effect on P secretion while alcohol effect was significant. The decline in /?hCG production was also seen when the placental explant and the SC tissue were co-chambered during superfusion, as evidenced by the decreased area under the curve of significant pulses of/3hCG secretion (using PULSAR analysis). A delayed inhibitory effect of SC was also noted when PhCG pulsatility was examined after removing the SC, following an overnight co-incubation. This suggests that the effect of SC lasts beyond the direct contact with the tissue. This was shown by the significant decrease in the pulse amplitude as well. It is likely that the more pronounced effect seen was obtained because of the longer incubation time (16 h compared with 2 h seen in the former case). Pre-exposure to 1 ,MJIcyclohexamide (a specific protein synthesis inhibitor in this concentration; Barnea et al, 1986), markedly reduced /?hCG secretion in superfused placental explants (data not shown). This was seen by reductions in mean peak amplitude, area under the curve and frequency of pulses. At high cyclohexamide concentration (> 100 PM), the effect was toxic and pulsatility was completely lost. In contrast, SC did not modify/3hCG pulse frequency, an indication for the specificity of SC action. The mechanism(s) of action and identity of these placental hCG and P secretion modulatory products are currently unknown. A rapid inhibition in spontaneous /3hCG secretion was seen when the diluted SC extract was added for 1 min to superfused explants. The increase in circulating hCG levels is linear until 8 weeks of gestation, plateauing at 910 weeks, and declining thereafter. The factors involved in the physiologic decline in hCG secretion after 9 weeks have not been identified. Our data provide indirect evidence that soluble component(s) of human embryonic neural tissues (and not of adrenal) may have a role in limiting trophoblastic hCG secretion with advancing gestation. Whether these principles reach the placenta in vivo remains to be established. Interestingly, after 11 weeks, when hCG
530
Plucenta(1991), Vol. 12
levels are lower, an augmentation of/?hCG and hCG secretion could be detected, indicating a reversal of the situation by neural elements. In our view, this shift is a result of changes occurring in the neural tissue itself, which undergoes major changes during embryogenesis and through fetal development. Further investigations have revealed the presence of two separate factors-one stimulatory while the other having inhibitory effects upon hCG secretion which their concentration could change in the favor of the latter, following the hCG plateau i.e. 10-l 1 weeks (unpublished observation). In vivo, the SC and brain may secrete these modulatory factors which reach the trophoblast through diffusion, a process which is operative in early placenta, since placental vascularization is a later developmental event (Hustin and Schaaps, 1987). These compounds may reach through the amniotic fluid which is a rich source of active compounds among them neurotransmitters (Genazzani et al, 1984). Alternatively, these compounds could reach the trophoblast via the embryonal vascular system which becomes operative by the sixth week (Hytten and Leitch, 1971). The finding that the PBS extract was more potent than the alcohol-based extract suggests that the hCG modulatory compounds are likely to be water soluble, suggestive of their proteinecious nature. In contrast, the alcohol extracts were only effective in this case. A further indication for extracts specificity and the multifactorial control that is involved in placental hormone secretion. Studies for identifying these factors are currently carried out. The modulatory effect of neural tissues upon hCG and P secretion is not unique to this tissue since we have recently reported that other embryonal visceral tissue, like lung and kidney of the same gestational age, affect trophoblastic function, following a similar agedependent effect inhibitory until week 9 and stimulator-y thereafter (Bamea, Simon and Kol, 1989). However, the superfusion model reveals differences in response between neural and adrenal tissues. This work also shows that the compounds are also secreted by and are not only present in the tissue, strongly suggesting that we are dealing with a physiological effect and not only an effect produced by cellular structural element(s). Thus, our data suggest that the embryo in situ has a role in regulating placental function with respect to both needs of the embryo and the maternal environment (mainly decidua and corpus luteum). Our data is in accord with previous studies on the regulatory role that the early embryo may have. Indeed, placental explants obtained from patients shortly after embryonal death grow poorly in culture (Hustin and Gaspard, 1977). This dependence is likely to be gestational age-dependent since the absence of a viable fetus during second trimester or later does not appear to impair placental function (Schermers, 1984), a strong indication that placental independence is increasing from the early second trimester onwards.
REFERENCES Adamsoq R. (Ed.) (1968) Diagnosis and treatment of fetal disorders. New York: Springer-Verlag. Albrecht, E. D., Ha&ins, A. L. & Pepe, G. J. (1984) The intluence of feteetomy at midgestation upon the serum concentrations of progesterone, estrone and estradiol in baboons. Enubinolog~, 107,766-799. Borneo, E. R., Bischof, P., Bran+ C. & Sunyal, M.. K. (1986) PAPP-A release from isolated first trimester trophoblastic cells. Archives ofGynedogy, 237,187-190. Barnes, E. R., Lavi, G., F&h, H. & DeCerney, A. H. (1986a) The role of ACTH in placental steroidogenesis. Placenta, 7,3033 11. Barnes, E. R., Ohmer, G., Benveniste, R., Romero, R. & DeCherney, A. H. (1986b) Progesterone, estradiol, and alpha-human chorionic gonadotropin secretion in patients with ectopic pregnancy. ~oburnal of Clinical EndocrinologyandMetabolism, 62,529-533. Bamea, E. R. & Kaplan, M. (1989) Spontaneous gonadotropin-releasing hormone induced, and progesterone
Shzrrtz-Smirski et al: Human Embyonal
Organs Effect on Placenta
531
inhibited pulsatile secretion of human chorionic gonadotropin in the first trimester placenta in vitro. 3ozmal~1’ Clinical Endocrinolog) and Metabolism, 69,215-217. Bamea, E. R., Simon, R. J. & Kol, S. (1989) Human embryonal extracts modulate placental function in the first trimester: effects of visceral tissues upon chorionic gonadotropin and progesterone secretion. Placenta, 10,331344. Barnes, E. R., Feldman, D., Kaplan, M. & Morrish, D. W. (1990) The dual effect of epidermal growth factor upon human chorionic gonadotropin secretion by- the first trimester placenta in vitro. 3oburnal of Clinical Endocrinology and.Metabolisnt, 71, 923-928. Bogart, M. H., Pandian M. R. &Jones, 0. W. (1987) Abnormal Maternal serum chroionic gonadotropin levels in pregnancies with fetal chromosome abnormalities, Prenatal Diagnosis, 7,623-630. Genazzani, A. R., Petraglia, F., Parrini, D., Nasi, A., Angioni, G., Facchinetti, F., Facchini, V. & Volpe, A. (1984) Lack of correlation between amniotic fluid and maternal plasma contents of beta endorphin, betalipotropin and adrenocorticotropic hormone in normal and pathological pregnancies.Ametican3or~rnalof Obstetrics and G.jmecology, 148, 198-203. Hertig, A. T., Rock, J. & Adams, E. C. (1956) A description of 34 human ova within the first 17 days of development. American ToburnalofAnatomy, 98,435-439. Hustin, J. & Gaspard, u. (1977) Comparison of histological changes seen in placental tissue cultures and in olacentae obtained after fetal death. British ?oumal of Obstetrics and Gvnecolotrv. 84.210-216. H&tin, J. & Schaaps, J. P. (1987) Echocariographic and anatomic studies zfthe’maternal trophoblastic border during the first trimester of pregnancy. American3oumal of Obstetrics and CynecoloD, 157, 162-168. Hytten, F. E. & Leitch, I. (1971) In: Physiolog): of human pregnang: 2nd ed., Oxford: Blackwell. Lowry, 0. H., Rosenbrough, N. J., Farr, A. L. & Randall, R. J. (1953) Protein measurements with the pholin phenol reagent.Joumal ofBiological Chemisty, 193, 265-269. Ramsay, T. G., Sheahan, J. A., Hausman, G. J. &Martin, R. J. (1985) Effects of fetal decapitation upon porcine placental metabolism. Biological Neonate, 47,42-53. Robertson, M. C., Owens, R. E. & McCoshen, J. A. (1984) Ovarian factors inhibit and fetal factors stimulate the secretion of rat placental lactogen. Endocrinoloa, 114, 22-30. Taylor, M. J., Jenkin, G., Robinson, J. S. & Friesen, H. C. (1984) Effect of intrauterine death and fetectomy on ovine placental lactogen production. Research Veterinay Sciences, 35, 22-31. Schermers, J. P. (1984) Ectopic pregnancy: A morphologic and endocrine study, pp. l-87. Amsterdam: Boek Enoffsetdrukkenj los-Naarden.
Human Embryo Modulates Placental Functibn in the First Trimester; Effects of Neural Tissues upon Chorionic Gonadotropin and Progesterone Secretion
REVITAL SHURTZ-SWIRSKI, ROBERT J. SIMON, YAEL COHEN EYTAN R. BARNEA”
&
Feto-Placental Endocrine Unit, Rappaport Institute and Department of Obst&cs/Gynaecology Rambam Medical Centre, Technion, Haif, Israel a To whom correspondence and reprint requests should be addressed Paper accqted 164.1991
SUMMARY We investigated
the efect
of embyonal
neural and adrenal
tissues (7-14
weeks
gestational age) upon BhCG semetion by homologous placental explants in staticand
dynamic cultures. In static co-culture si’ijicant inhibition by SC and brain was notedat 7-9 weeks.Similarly, in superjiision,using a novelco-chamberingtechnique there was a significantreductionin the area under the curve but notpeakfrequency of spontaneous pulsatileBhCG secretion.Incubationswith neural tissues11 weeksand abozlecaused a stimulatoryeflectupon BhCG secretionin both models. The effectof adrenal tissuein staticcultureswasd@&nt, namelyslightlyinhibitoryat 7-9 weeks and inhibit0y at 11 weeksand above.In superjiision,the effectof adrenal tissuewas not sign@ant. Extracted neural tissue 7-9 weeks incubated with placental explants exhibited inhibitoryeffectsupon BhCG secretionas well. Buffer-based extractsof neural tissues efiect was more pronounced than alcoholbased extractsregarding BhCG semetion. The effect of extractswas dose-dependent and effectswere notedup until a 2000-fold dilution. In contrast,the buffer SC extract had no efect on progesterone (P) secretion while the alcohol extract eflect was inhibit0y at 7-9 weeksand stimulatoy at > I I weeks. Superfiised explantspatternofl3hCG secretionwasinhibitedby one minutepulseof the SC buffer extract. In conclusion,the human neural tissueofembyonal origin may modulateplacental hCG and P secretionduring early pregnancy. 01~3-4004/91/050521 + 11 $05.00/O
@J 1991 Baillike Tindall Ltd
522
Placenta(1991), Vol. 12
INTRODUCTION A viable trophoblast is a prerequisite for normal embryonic development. However, it has not been established, in the human, whether a viable embryo is necessary for maintaining placental function (Hertig, Rock and Adams, 1956). It has been shown histologically that, cultures of trophoblast obtained shortly after embryonal death appear to be damaged (Hustin and Gaspard, 1977). There is a close temporal association between trophoblastic function, as reflected by the peak of hCG secretion at 9 gestational weeks, and the shift of responsibility for progesterone secretion from the corpus luteum to the placenta. At the same time, with the completion of embryogenesis, the transition from embryo to fetus takes place (Adamsons, 1968). Several strands of evidence suggest that there is an interdependence in mammals between the trophoblast and the embryo. Fetal death in both sheep and rats results in a decline in placental lactogen secretion (Ramsay et al, 1985; Albrecht, Haskins and Pepe, 1984; Taylor et al, 1984; Robertson, Owens and McCoshen, 1984). Despite many investigations upon the endocrine control of the trophoblast, the endogenous compound(s) that actually modulate the widely changing pattern of circulating hCG in the first trimester have not yet been identified. There is some evidence that, in humans, embryonal dysfunction or death modifies placental endocrine function. Indeed, it was recently shown that elevated circulating hCG levels in the second trimester can serve as a screening test for detecting certain chromosomal aberrations in the fetus (Bogart, Pandian and Jones, 1987). However the mechanism behind this abnormal rise is unknown. We have recently reported that extracts of visceral organs of human embryonal origin affect hCG secretion by cultured placental explants (Barnea, Simon and Kol, 1989). In that study it was not established whether the effect was due to compounds present only within the tissue itself or whether compounds were also secreted by the various organs, thus having an endocrine effect. In addition, we have recently shown that in superfused placental explants hCG secretion is episodic and is modulable by various hormones (Barnea and Kaplan, 1989). Therefore, in the present study the effect of neural and adrenal tissue co-cultured with placental explants upon hCG and progesterone secretion was examined. This was carried out in both static cultures where feedback between placenta and embryo is operative and in dynamic co-chambering where, because of the unidirectional flow, only the embryo can influence the placenta without feedback. In addition, the effect of neural extracts upon hCG secretion in the two models was examined. These results show that, at 7-9 weeks, neural tissue inhibited hCG secretion in both static and dynamic cultures, while at 11 weeks and later the effect was stimulatory. This pattern is characteristic to neural tissues, since adrenal affected the endocrinological behavior of the placenta differently.
MATERIALS
AND METHODS
Placental and embryonic material Twenty 7-14-week-old placentas and embryos were studied. After obtaining appropriate consent, elective pregnancy terminations by vacuum curretage were performed at the Rambam Medical Center. Collected tissue was rinsed several times in 0.9 per cent NaCl to remove all blood, followed by three further rinses with culture medium [Dulbecco’s
Shurt..-Swirski et al: Human Embyonal
Organs Effect on Placenta
Modified Eagle’s Medium (DMEM), Beit Haemek, Israel], U/ml, streptomycin 20 yg/ml and amphotericin B 50 pug/ml.
523
containing
penicillin
5000
Neural and adrenal tissue preparation Embryos were removed aseptically and gestational ages were confirmed by crown-rump measurements. The SC, which was removed from the spinal column, the brain and adrenal tissues were washed in a large excess of 0.9 per cent saline solution followed by several rinses with media containing 1 per cent antibiotics. The neural tissue was cut to 3-4 mm size fragments and was added to placental explant culture. The entire adrenal gland was cocultured with placental explants. Neural tissue extraction In other experiments the SC was extracted by sonication for 60 set (Triacregler-Ystral GmbH D-7801, Dottringhen) with one ml of (a) absolute ethyl alcohol, or (b) phosphate buffer 0.05 14, pH 7.4 containing 10 mu dithiothreitol (DTT) and 2 mu phenylmethylsulfonylfluoride (PMSF). Following sonication the particulated matter was sedimented by centrifugation at 800g for 10 min. The supernatant fraction was removed and saved. The supernatant from the alcohol extracts was dried and later resuspended in 1 ml of distilled water. The extracts were stored at -20°C until used. Explant cultures Using previously reported procedures (Barnea et al, 1989; Barnea et al, 1990), explants 5070 mg wet weight were dissected and rinsed in 1 per cent antibiotic solutions. For culture, explants were placed in DMEM media with or without the tested organs or organ extracts. At least three replicates per test agent or control were plated at 37°C in an atmosphere of 9.5 per cent air/5 per cent CO*. After 24 h of incubation, the media were collected and stored at - 20°C until assay. The tissue was saved for protein analysis. Dishes containing only vehicle served as controls. Additional culture dishes containing DMEM and neural tissue or adrenal, but not trophoblast explants were cultured and the media was collected and assayed for PhCG, hCG and progesterone (P). In addition, the /3hCG content of neural tissue extracts was also determined. Superfusion studies The method was recently published (Barnea and Kaplan, 1989; Barnea et al, 1990). Briefly, a superfusion apparatus (Accusyst, Endotronics, St. Paul, MN) with a multichannel peristaltic pump and fraction collector (model 272, ISCO, Durham, NC) were used to study the shortterm dynamics of hCG secretion. The explants (200-300 mg wet wt) were placed in the culture chambers and a DMEM solution containing HEPES (18 mM) was washed through in an atmosphere of 5 per cent COJ95 per cent air at 37°C. Experiments were conducted for a 2 h period; a I ml sample from the effluent was collected every 2.4 min for hCG measurements. In each experiment, one channel served as control and four served as experimental channels. At given intervals, a 1 ml pulse of SC extract was given through a peristaltic pump equipped with a digital flowmeter (ISMATECH DD, Chicago, IL). Co-chambering experiments In some experiments, placental explants were co-chambered with SC fragments by placing the two tissues serially in a double chamber separated by a metal mesh which allowed the
524
Placenta(1991), Vol. I2
medium to tlow unidirectionally from SC to placenta during superfusion. Media collected were stored at -20°C until assayed. Assays
As reported previously, /3hCG was measured by radioimmunoassay using hCG-MAIA clone assay (Serono) (Barnea, Simon and Kol, 1989; Barnea et al, 1986b; Barnea et al, 1990), with an intra-assay variability of 1.7 per cent. The interassay variability was 3.2 per cent. The lit of assay sensitivity was 11 weeks. This can be considered as evidence of the extractibility of the factors, high potency, and specificity, i.e. PBS-based extract was more potent than the alcohol-based extract on /?hCG secretion. In contrast, the aqueous extract had no effect on P secretion while alcohol effect was significant. The decline in /?hCG production was also seen when the placental explant and the SC tissue were co-chambered during superfusion, as evidenced by the decreased area under the curve of significant pulses of/3hCG secretion (using PULSAR analysis). A delayed inhibitory effect of SC was also noted when PhCG pulsatility was examined after removing the SC, following an overnight co-incubation. This suggests that the effect of SC lasts beyond the direct contact with the tissue. This was shown by the significant decrease in the pulse amplitude as well. It is likely that the more pronounced effect seen was obtained because of the longer incubation time (16 h compared with 2 h seen in the former case). Pre-exposure to 1 ,MJIcyclohexamide (a specific protein synthesis inhibitor in this concentration; Barnea et al, 1986), markedly reduced /?hCG secretion in superfused placental explants (data not shown). This was seen by reductions in mean peak amplitude, area under the curve and frequency of pulses. At high cyclohexamide concentration (> 100 PM), the effect was toxic and pulsatility was completely lost. In contrast, SC did not modify/3hCG pulse frequency, an indication for the specificity of SC action. The mechanism(s) of action and identity of these placental hCG and P secretion modulatory products are currently unknown. A rapid inhibition in spontaneous /3hCG secretion was seen when the diluted SC extract was added for 1 min to superfused explants. The increase in circulating hCG levels is linear until 8 weeks of gestation, plateauing at 910 weeks, and declining thereafter. The factors involved in the physiologic decline in hCG secretion after 9 weeks have not been identified. Our data provide indirect evidence that soluble component(s) of human embryonic neural tissues (and not of adrenal) may have a role in limiting trophoblastic hCG secretion with advancing gestation. Whether these principles reach the placenta in vivo remains to be established. Interestingly, after 11 weeks, when hCG
530
Plucenta(1991), Vol. 12
levels are lower, an augmentation of/?hCG and hCG secretion could be detected, indicating a reversal of the situation by neural elements. In our view, this shift is a result of changes occurring in the neural tissue itself, which undergoes major changes during embryogenesis and through fetal development. Further investigations have revealed the presence of two separate factors-one stimulatory while the other having inhibitory effects upon hCG secretion which their concentration could change in the favor of the latter, following the hCG plateau i.e. 10-l 1 weeks (unpublished observation). In vivo, the SC and brain may secrete these modulatory factors which reach the trophoblast through diffusion, a process which is operative in early placenta, since placental vascularization is a later developmental event (Hustin and Schaaps, 1987). These compounds may reach through the amniotic fluid which is a rich source of active compounds among them neurotransmitters (Genazzani et al, 1984). Alternatively, these compounds could reach the trophoblast via the embryonal vascular system which becomes operative by the sixth week (Hytten and Leitch, 1971). The finding that the PBS extract was more potent than the alcohol-based extract suggests that the hCG modulatory compounds are likely to be water soluble, suggestive of their proteinecious nature. In contrast, the alcohol extracts were only effective in this case. A further indication for extracts specificity and the multifactorial control that is involved in placental hormone secretion. Studies for identifying these factors are currently carried out. The modulatory effect of neural tissues upon hCG and P secretion is not unique to this tissue since we have recently reported that other embryonal visceral tissue, like lung and kidney of the same gestational age, affect trophoblastic function, following a similar agedependent effect inhibitory until week 9 and stimulator-y thereafter (Bamea, Simon and Kol, 1989). However, the superfusion model reveals differences in response between neural and adrenal tissues. This work also shows that the compounds are also secreted by and are not only present in the tissue, strongly suggesting that we are dealing with a physiological effect and not only an effect produced by cellular structural element(s). Thus, our data suggest that the embryo in situ has a role in regulating placental function with respect to both needs of the embryo and the maternal environment (mainly decidua and corpus luteum). Our data is in accord with previous studies on the regulatory role that the early embryo may have. Indeed, placental explants obtained from patients shortly after embryonal death grow poorly in culture (Hustin and Gaspard, 1977). This dependence is likely to be gestational age-dependent since the absence of a viable fetus during second trimester or later does not appear to impair placental function (Schermers, 1984), a strong indication that placental independence is increasing from the early second trimester onwards.
REFERENCES Adamsoq R. (Ed.) (1968) Diagnosis and treatment of fetal disorders. New York: Springer-Verlag. Albrecht, E. D., Ha&ins, A. L. & Pepe, G. J. (1984) The intluence of feteetomy at midgestation upon the serum concentrations of progesterone, estrone and estradiol in baboons. Enubinolog~, 107,766-799. Borneo, E. R., Bischof, P., Bran+ C. & Sunyal, M.. K. (1986) PAPP-A release from isolated first trimester trophoblastic cells. Archives ofGynedogy, 237,187-190. Barnes, E. R., Lavi, G., F&h, H. & DeCerney, A. H. (1986a) The role of ACTH in placental steroidogenesis. Placenta, 7,3033 11. Barnes, E. R., Ohmer, G., Benveniste, R., Romero, R. & DeCherney, A. H. (1986b) Progesterone, estradiol, and alpha-human chorionic gonadotropin secretion in patients with ectopic pregnancy. ~oburnal of Clinical EndocrinologyandMetabolism, 62,529-533. Bamea, E. R. & Kaplan, M. (1989) Spontaneous gonadotropin-releasing hormone induced, and progesterone
Shzrrtz-Smirski et al: Human Embyonal
Organs Effect on Placenta
531
inhibited pulsatile secretion of human chorionic gonadotropin in the first trimester placenta in vitro. 3ozmal~1’ Clinical Endocrinolog) and Metabolism, 69,215-217. Bamea, E. R., Simon, R. J. & Kol, S. (1989) Human embryonal extracts modulate placental function in the first trimester: effects of visceral tissues upon chorionic gonadotropin and progesterone secretion. Placenta, 10,331344. Barnes, E. R., Feldman, D., Kaplan, M. & Morrish, D. W. (1990) The dual effect of epidermal growth factor upon human chorionic gonadotropin secretion by- the first trimester placenta in vitro. 3oburnal of Clinical Endocrinology and.Metabolisnt, 71, 923-928. Bogart, M. H., Pandian M. R. &Jones, 0. W. (1987) Abnormal Maternal serum chroionic gonadotropin levels in pregnancies with fetal chromosome abnormalities, Prenatal Diagnosis, 7,623-630. Genazzani, A. R., Petraglia, F., Parrini, D., Nasi, A., Angioni, G., Facchinetti, F., Facchini, V. & Volpe, A. (1984) Lack of correlation between amniotic fluid and maternal plasma contents of beta endorphin, betalipotropin and adrenocorticotropic hormone in normal and pathological pregnancies.Ametican3or~rnalof Obstetrics and G.jmecology, 148, 198-203. Hertig, A. T., Rock, J. & Adams, E. C. (1956) A description of 34 human ova within the first 17 days of development. American ToburnalofAnatomy, 98,435-439. Hustin, J. & Gaspard, u. (1977) Comparison of histological changes seen in placental tissue cultures and in olacentae obtained after fetal death. British ?oumal of Obstetrics and Gvnecolotrv. 84.210-216. H&tin, J. & Schaaps, J. P. (1987) Echocariographic and anatomic studies zfthe’maternal trophoblastic border during the first trimester of pregnancy. American3oumal of Obstetrics and CynecoloD, 157, 162-168. Hytten, F. E. & Leitch, I. (1971) In: Physiolog): of human pregnang: 2nd ed., Oxford: Blackwell. Lowry, 0. H., Rosenbrough, N. J., Farr, A. L. & Randall, R. J. (1953) Protein measurements with the pholin phenol reagent.Joumal ofBiological Chemisty, 193, 265-269. Ramsay, T. G., Sheahan, J. A., Hausman, G. J. &Martin, R. J. (1985) Effects of fetal decapitation upon porcine placental metabolism. Biological Neonate, 47,42-53. Robertson, M. C., Owens, R. E. & McCoshen, J. A. (1984) Ovarian factors inhibit and fetal factors stimulate the secretion of rat placental lactogen. Endocrinoloa, 114, 22-30. Taylor, M. J., Jenkin, G., Robinson, J. S. & Friesen, H. C. (1984) Effect of intrauterine death and fetectomy on ovine placental lactogen production. Research Veterinay Sciences, 35, 22-31. Schermers, J. P. (1984) Ectopic pregnancy: A morphologic and endocrine study, pp. l-87. Amsterdam: Boek Enoffsetdrukkenj los-Naarden.
