Gynecol. Endocrinol. 6 (1992) 293-300

The human first-term placenta ilz vitro: regulation of hCG secretion by GnRH and its antagonist A . Sziligyi, R. Benz and W. G. Rossmanith

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Department of Obstetrics and Gynecology, University of Ulm, Ulm/Donau, Germany Key words: GONADOTROPIN RELEASING HORMONE,ANTAGONIST, HUMANCHORIONIC GONADOTROPIN, PLACENTA,PERIFUSION, INCUBATION

ABSTRACT Gonadotropin releasing hormone ( G n R H ) has been proposed to play a role in the regulation of human chorionic gonadotropifii (hCG) release from the human placenta. To test this assumption, we utilized an in vitro periiusion system, together with cultures ofplacental explants, to investigate short- and long-term efects o f GnPJrI and its respective antagonist on the k C G secretion from the early human’ placenta. Tissue slices o f human placenta (100 mg), obtained from first-trimester terminations o f pregnancies, were continuously peritused and the egZuent collected infractions 42-20 min. After initial peritusion periods of3&40 min, either G n R H , a G n R H antagonist (SB-7.5; Asta Pharma, Frankfurt, Germany) or both compounds at equimolar concentrations were added to the peritusion medium atfinal concentration o f 1 c41 c8mol/l). Administration was effected either continuously or intermittently in 1 0-minpulses. Further, 50-mg pieces of placental tissue explants were cultured in tissue culture p1ate.r for up to 6 days. During the pentusions, h C G (determined by enzymeimmunoassay) was found to be released spontaneously in a pulsatilefashion. Pulse amplitudes andfrequencies of this episodic h C G secretion were increased in response to G n R H , but not afected by G n R H antagonist. Also, G n R H stimulated the h C G secretion during cultures of placental explants. When pharmacological doses o f G n R H (1c4mol/l) were utilized, this slimulatory eject o f G n R H was no longer

evident, whilepen@sion with medium containing G n R H antagonist at identical concentrationsstimulated the h C G secretion, indicating an intrinsic agonistic activity o f the antagonist. h C G secretion was not afected, when G n R H and GnRHantagonist were co-administered in equimolar concentrations. These observations demonstrate modulatory efects of GnRH and its respective antagonist on the in vitro placental k C G release,further supporting the notion ofpresumably paracrine efects o f G n R H in the humanfirst-trimester placenta.

INTRODUCTION As the conclusion from an increasing amount of in vitro data, the production and release of placental proteins is now known to mimic a miniature model of regulatory conditions comparable to those of the hypothalamic-pituitary-target axis’,’. Gonadotropin releasing hormone (GnRH) has been reported to be present in the human placenta3, with the majority of GnRH produced primarily in the cytotrophoblast layer ofthe placenta’. As a consequence of these findings, further investigation^^.^ have revealed the human placenta to be the source of considerable amounts of GnRH-like activity, with maximal concentration attained during early gestation. Although the characteristics of chorionic

Correspondence: Dr. W. G. Rossmanith, Department of Obstetrics and Gynecology RH-20, University of Washington, Seattle, Washington, ‘98195,USA

293

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Placental h C G secretion

GnRH may be different from its hypothalamic counterpad, addition of GnRH to placental cultures is also capable of stimulating synthesis and release of human chorionic gonadotropin (hCG)'. Addition of a GnRH antagonist could effectively prevent this stimulation', indicating a receptorspecific event. Eventual involvement of GnRH in the regulation of placental hCG secretion has been evaluated by a dual approach. Placental tissue cultures for up to 5-6 days were utilized, presumed to reflect de nouo hCG biosynthesis. Conversely, perifusion studies may relate to the dynamics of acute hCG release'. Using placental perifusion studies, Barnea and demonstrated a spontaneous pulsatile release of hCG from the early human placenta in vitro, and this episodic secretion was effectively stimulated by addition of GnRH. Conversely, other investigations employing placental explant cultures failed to demonstrate stimulatory effects of GnRH on placental hCG secretion'*. Thus, the role of placental GnRH, if any, remained unresolved. We, therefore, used in uitro placental tissue cultures and a perifusion system to evaluate the effect of GnRH and its respective antagonist on the hCG secretion from the first-trimester human placenta.

MATERIAL AND METHODS Placental perifusions Placental tissue obtained from induced terminations of pregnancy at 9-12 weeks of gestation was immediately collected in sterile Medium 199 (Sigma, Munich, Germany). Placental slices (100 mg) were placed in flow-through perifusion chambers (200 pl volume) and continuously perifused with Medium 199 in a perihsion system (Accusyst, Endotronics, Minneapolis, USA). The perifision medium contained sodium bicarbonate (2.2 gA), bovine serum albumin (BSA, 1 g/l) and 1% of an antibiotic solution (Penicillin G 10 000 U/ml and Streptomycin 10 mg/ml; all Sigma, Munich, Germany). This basic perifusion medium was continuously oxygenated with 5% co2/95% 0 2 and maintained at 37 O C by a recirculating waterbath. After a stabllrzationperiod of 120 min, the perifusion effluent was collected in fractions of 2 min for 100 and 120 min, respectively (n = 13). Following an initial perifusion period of 30 or 40 min using basic medium only, GnRH

294

Sziligyi, Benz and Rossmanith

(Sigma, Munich, Germany) or a GnRH antagonist (SB-75, Asta Pharma, Frankfurt, Germany) was added to the medium at final concentrations of 10+10* molA In most experiments (n = 8), basic medium was replaced by medium containing GnRH and GnRH antagonist at concentrations of 10-4-10+ mol/l, and this medium was again changed to basic medium after another 10 min of perifision. Thus, repeated stimulationwith GnRH was effected. In addition, GnRH and the GnRH antagonist were co-administered at equimolar concentrations (lod, lo4 mol/l) in an intermittent fashion (n = 2). The effluent was stored at -20 OC until assayed.

Culture of placental tissue explants Explants (50 mg, n = 42) of placental tissue were placed in multiwell tissue culture plates (Falcon, Lincoln Park, USA). Each well contained 2 ml of the culture Medium 199 prepared under sterile conditions. For the incubation experiments, 10% fetal calf serum and antibiotics (250 pg/ml Amphotericin B; all Biochrom, Berlin, Germany)were added. The culture medium was sterilized by pressuring it through a 0.2-pm 6lter (Schleicher,Dassel, Germany). For each experimental set-up, six wells containing basic medium served as control, while another six wells constitutedthe experimental conditions (n = 36) During the experiments, medium containing either GnRH, GnRH antagonist or both at final concentrations of 10" and lo4 mol/l was added. The placental explants were cultured at 37 "C for 6 days in humidified air containing 5% COZ. One ml of either basic or of GnRHlGnRH antagonist containing culture medium was replaced every day in each well, and the removed supernatant centrihged and stored at -20 "C until assayed.

Assay determination and statistical evaluation Intact hCG was determined in the perifusion fractions and in the incubation aliquots by a commercial microparticle enzymeimmunoassay (IMx System, Abbot, Wiesbaden, Germany). Use of this assay provided a very high sensitivity (0.5 IU hCG/l) with acceptableintra- and interassay coefficients of variation (CV) of 3.4% and 4.5%, respectively. Cross-reactivity with the a-and P-subunits

GynecologicalEndocrinology

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Placental hCsG secretion

Szilhgyi, Ben+ and Rossmanirh

of the hCG molecule was neghgble (< 1%). All fractions from an individual set of experiments were incorporated in a single assay, to minimize interassay variability. To define experimental variability, a cclntrol or calibration perifusion was performed with medium containing hCG (240 IU/l) and running through an empty perifusion chamber. The hCG secretory profiles obtained by serial determinat:ions of all perifusion fractions were searched for significant pulses by use of the cluster pulse analysis developed by Veldhuis and Johnsod3. Criteria to restrain the rate of false-positive pulse determinations were applied to the data series of the control run. For t h s control experiment, a cluster size of 2 for both a test nadir and peak, together with a t-statistic of2.1 for both significant up- and downstroke were chosen. Identical pulse criteria weire also employed for the experimental data series. Since the small number of perifusions in some experimental set-ups prevented statistical assessment, the results wee presented descriptively. For the placental explant cultures, results are presented as means k SEM. To test for significant changes in t:he hCG secretory rates during different culture experimental set-ups, analysis of variance (ANOVA) was used followed by Newman-Keul's rank test when appropriate.

-""

v

hCG secretion from perifused placental explants hCG secretion from the first-trimester human placenta was spontaneously pulsatile (Figure 1). The mean hCG secretion was hghly variable (ranging from 4 to 700 mIU/fraction) and was not related to the size or age of the placenta. Pulses occurred at a mean interval of 12-17 min and their magnitude ranged from 11 to 22 mIU (Figure 1). This intrinsically episodic hCG secretion was stimulated by intermittent exposure to GnRH at concentrations of 10" and lo4 mol/l (Figure 2). Ths was attributed particularly to higher hCG secretory amplitudes (30-150 mIU) and higher pulse frequencies (a secretory episode occurring every 4 min) during perifusion with GnRH-containing medium. O n the other hand, continuous administration of GnRH at concentrations of 10" mol/l yielded a single secretory pulse of hCG (Figure 2, lower panel). GnRH antagonist in doses of mol/l did not alter hCG secretion. When the dose of GnRH was increased to mol/l, the action of GnRH no longer appeared to be stimulatory, and the pulse attributes (fi-equencies and amplitudes) remained unchanged (Figure 3).

I

*

I *

2cr

RESULTS

*

*

*

*

*

c3 0 100f

*

* -

* ** *

*

1

I

I

I

1

I

I

0

20

40

60

80

100

120

1' 0

Minutes

--

Experiment 1

+ Experiment 2

+ Control perifusion

Figure 1 hCG concentrations in the perifusate fractions of two placental perifusions and of a control perifusion during whiclh medium containing hCG was running through an empty chamber. Asterisks indicate significant pulses

Gynecological Endocrinology

295

Szilhgyi, Benz and Rossmanitk

Placental hCG secretion

800

' I

0

**

** *

**

I

I

I

40

20

I

+Med. + GnRH 1O-* Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

I

100

80 Minutes

60

-c Med.

I

120

+ GnRH-antag.

1

10-8 rnol/l

$150 \

2 100 E

V

-

1

0

I

I

I

20

40

60

+Med.

I

+ GnRH 3 0-6

-

Med.

I

I

80 Minutes

100

120

+ GnRH-antag.

1

10-6 rnol/l

16

I

0

20

I

40

I

60 Minutes

I

I

80

100

120

-+ Med. + GnRH 10-6 -c Med. + GnRH-antag. 10-6 moll1 Figure 2 hCG concentrations in the perihsate hctions of placental perifusions'during which GnRH and GnRH antagonist (loT6and lo4 mol/l) were added to the p e h s i o n medium, as indicated by the bars. Fractions were collected at 2-min intervals. Asterisks indicate sigdicant pulses

In contrast t o GnRH, the GnRH antagonist at identical high concentration stimulated the h C G

296

secretion by increasing the pulse amplitudes (up to 130 mIU) (Figure 3)

Gynecological Endocrinology

Placental hCG secretion

Szilhgyi, Benz and Rossmanith

I* * *

0

c

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

0

* * * * * *

I

I

I

I

I

I

20

40

60

80

100

120

10

Minutes

+Med + GnRH-antag.

+ Med + GnRH 5 x lO-'mol/l

5 x lO-'mol/l

Figure 3 hCG concentrations in the perifusate fkctions during which GnRH and GnRN antagonist (lo4 mol/l) were added to the perifusion medium, as indicated by the bars. Fraction intervals were 2 min. Asterisks indicate significant pulses. 500 7

i

400 -

3 0

'E 300-

$ 5

3 200-

v

il

0

&

100 0

c 0

I

I

I

I

I

I

20

40

60

80

-100

120

140

Minutes

- Medlum +M + GnRH + GnRH-antag. lo6

+M

+G

M

+ GnRH-antag.

10-8 mol/l

Figure 4 hCG concentrations in the perifusate 6actions of placental perifusions during which GnRH and GnRH antagonist (1OW6and mol/l) were co-administered in an intermittent fashion (10 min), as indicated by the bars. Asterisks indicate significant pulses

W h e n G n R H and G n R H antagonist were coadministered in an intermittent fashion (Figure 4), their individual effects appeared t o be antagonized. T h e hCG secretory rates remained unchanged during these perifusions.

Gynecological Endocrinology

Cultured placental explants Incubation with basic m e d u m for up t o 6 days resulted in increased h C G secretion (&om 11 t o 30 IU/24 h x 1000). GnRH administered in concentrations of 10" and lo4 mol/l significantly

297

Szilhgyi, Benz and Rossmanith

Placental h C G secretion 60 h

0 0 0 4

2 40 * (v

I \

=1 2 ; 20 0 c 0

2

1

3

4

6

5

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Days -Medium *Mehum

;

- 1

1

-

+ GnRH lod

Medium

+ h a g . 10-6

+ G m -Medium

-+GnFUI

+ htag.

+ GnRH 104mol/l

I

I

I

3

4

5

Days

*Med + GnRH-antag. 10-6

-

Med

I

6

+ GnRH-antag. 10-8 mol/l

Figure 5 hCG concentrations in the culture medium of cultured first-trimester placental explants. Cultures were performed with basic medium or medium containing GnRH, GnRH antagonist or both at equimolar concentrations and lo-' mol/l). The values represent the mean f SEM of six experiments. *p < 0.05, **p < 0.01 vs. corresponding control values

(p < 0.05 or less) increased the hCG secretion fiom placental tissue (Figure 5, upper panel). When GnRH and GnRH antagonist were co-administered in equimolar concentrations to the culture medium, the stimulatory effect of GnRH was antagonized by addition of the GnRH antagonist, thus leaving the hCG secretory rate unchanged during the incubation period (Figure 5, upper panel). In contrast, addition of GnRH antagonist to the culture medium did not change the hCG release, compared to that during control conditions (Figure 5, lower panel).

298

DISCUSSION Most in vitro studies investigating the effect of GnRH on placental hCG secretion have indicated that GnRH stimulates the hCG secreti~n',*,~,'J~. The stirnulatory effect of GnRH on placental hCG secretion was found to be time-dependent and dose-related"I6. GnRH secretion has also been found to be regulated by a series of intraplacental factors15J7.In the early gestational period with a maximum of hCG production, GnRH-induced effects on hCG release have been found to be less pronounced6. The observations of the current

Gynecological Endocrinology

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Placental hCG secretion

investigations relate to the first-trimester human placenta and confirm previous findings, additionally supporting the concept of a stimulatory effect of GnRH both in a perifusion system and in placental tissue explant cultures. Our observations also agree with findings in placental perifusions, during whch the release of hCG has been reported to be spontaneously pulsatile'0,". These and our investigations indicate that t h s episodic hCG release was stimulated both in pulse amplitudes and frequencies by addition of GnRH to the perihsion medlum. However, the enhanced hCG secretion by G n M administration was no longer evident when GnRH doses hlgher mol/l) than those reported during physiological conditions" were utilized'*. As a possible mechanism of GnRH action, changes in the intracellular phosphatidylinosito1 turnover have been proposed". Likewise, GnRH action within the placenta may involve Ca2+ionsz0,similar to the conditions found within the pituitary. Consequently,a h g h dose of GnRH may quickly desensitize the local receptors, a feasible explanation for the lack of GnRH effects. Effects of-differentGnRH antagonists additionally employed to study the effects of GnRH on placental hCG secretion have been found to depend greatly on the experimental condltions. In placental tissue cultures, an inhbitory effect could be demonstrated in the placenta at mid-gestation, but not in the early placenta*.Perifusing the early trimester placenta, Barnea and colleagues" observed that use of a GnRH antagonist prevented the !jtimulatoryeffect of both GnRH and GnRH agonist and decreased the pulse amplitudes ofthe spontaneouslypulsatile hCG release. Administration of a GnRH antagonist during pregnancy in experimental animals decreased the hCG secretion and subsequently caused pregnancy lossz1.In the current perifusion and in vitro culture studies, GnRH antagonist at concentrations of 10-6-10-8 mol/l faaed to modlftr the hCG

Sziligyi, Benz and Rossmanith

secretion. However, when co-adrmnistered with GnRH to the medium, it antagonized the stimulatory effect of GnRH on hCG release. Interestingly, GnRH antagonist stimulated hCG release when used in a pharmacological concentration (lo" mol/l). We relate thls finding to the intrinsic agonistic activity, inherent to many, if not all, competitive receptor antagonists. Findlngs &om our and other investigations'.2 suggest that intraplacental GnRH may play a key role in the regulation ofhCG secretion during early pregnancy. Modulation of secretory activity can be demonstrated both in a perihsion model visualizing spontaneously pulsatile hCG secretion and during long-term tissue cultures relating to the de novo hCG biosynthesis. Several studies have been conducted to define a role ofplacental GnRH, but have yielded codhcting results2.I2.Controversies in the observed findlngs may be reconciled by variabhty related to the species studied, differences in the experimental condltions, and presumably also to the fetal gender12. Further, other not yet determined factors modulating the intraplacental GnRH action may be considered. As already previously suggested2z,the effects of GnRH may be constrained to intraplacental paracrine processes, since acute systemic adrmnistration of GnRH to primates does not significantly affect any of the circulating hormones during pregnancy23.In conclusion, observations from the current studies support a role of GnRH in the intraplacental regulation of hCG secretion during early gestation.

ACKNOWLEDGEMENTS A.S. is supported by the Alexander von Humboldt Foundation. W. G. R. is a recipient of a Deutsche Forschungsgemeinschaft grant (Ro 657/2-1). We thank Mrs. E. Ambach for her excellent technical assistance.

REFERENCES 1. Petraglia, R., Volpe, A., Genazzani, A. R., River, J., Sawchenko, P. E. and Vale, W. (1990). Neuroendociinology of the human placenta. Frontiers Neuroendom'nol., 11, 6 2. Petraglia, F., Calza, L., Garauti, G. C., Giardino, L., De Ramundo, B. M. and Angioni, S. (1990).

GynecologicadEndocrinology

New aspects of placental endocrinology. J . Endocrinol. Invest., 13,353 3. Gibbons, J. M., Mitnick, M. and Chieffo, V. (1975).In vitro biosynthesis of TSH- and LH-releasing factor by human placenta. A m . J . Obstet. Gynecol., 121, 127

299

Gynecol Endocrinol Downloaded from informahealthcare.com by McMaster University on 11/04/14 For personal use only.

Placental h C G secretion 4. Khodr, G. S. and Siler-Khodr, T. M. (1980). Placental LRF and its synthesis. Science, 217,315 5. Siler-Khodr, T. M. and Khodr, G. S. (1978). Luteinizing hormone releasing content of the human placenta. A m . J. Obstet. Gynecol., 130, 216 6. Siler-Khodr, T. M., Kang, I. A. and Khodr, G. S. (1991). Current topic: symposium on placental endocrinology. 1. Effects of chorionic GnRH on intrauterine tissues and pregnancy. Placenta, 12, 91 7 . Siler-Khodr, T. M., Khodr, G. S., Valenzuela, G. and Rhode, J. (1986). Gonadotropin-releasing hormone effects on placental hormones during gestation. I. Alpha-human chorionic gonadotropin, human chorionic gonadotropin and human chorionic somatomammotropin. Biol. Reprod., 34, 245 8. Siler-Khodr, T. M., Khodr, G. S., Rhode, J., Vickery, B. H. and Nestor, J. JJr. (1987). Gestational age related inhibition of placental hCG, hCG and steroid hormone release in vitro by a GnRH antagonist. Placenta, 8, 1 9. Cemerikic, B. and Genbacev, 0. (1989). Tissue explant technique in the study of human chorionic gonadotropin (HCG) production in vitro. In Genbacev, O., Hopper, A. and Beaconsfield, R. (eds.) Placenta as a Model and Source, p. 93.( New York: Plenum Press) 10. Barnea, E. R. and Kaplan, M. (1989). Gonadotropin releasing hormone-induced and progesterone-inhibited pulsatile secretion of human chorionic gonadotropin in the first trimester placenta in vitro. J. Clin. Endom’nol. Metab., 69,215 11. Bamea, E. R., Kaplan, M. and Naor, Z. (1991). Comparative stimulatory effect of gonadotrophin releasing hormone (GnRH) and GnRH agonist upon pulsatile human chorionic gonadotrophin secretion in superfused placental explants: reversible inhibition by a GnRH antagonist. Hum. Reprod., 6, 1063 12. Haning, R. V.Jr., Breault, P. H., DeSilva, M. V., Hackett, R. J. and Pouncy, C. L. (1988). Effects of fetal sex, stage of gestation, dibutyryl cyclic adenosine monophosphate, and gonadotropin releasing hormone on secretion of human chorionic gonadotropin by placental explants in vitro. A m . J. Obstet. Gynecol., 159, 1332 13. Veldhuis,J. D. and Johnson, M. L. (1986). Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. A m . J. Physiol., 250, E486

300

Sziligyi, Benz and Rossmanith 14. Kim, S. J., Namkoong, S. E., Lee, J. W., Jung, J. K., Kang, B. C. and Park, J. S. (1987). Response of human chorionic gonadotropin to luteinizing hormone-releasing hormone stimulation in the culture media of normal human placenta, choriocarcinoma cell lines and in the serum of patients with gestational trophoblastic disease. Placenta, 8, 257 15. Petragha, F., Lim, A. T. W. andvale, W. (1987). Adenosine 3’5’-monophosphate. prostaglandins, and epinephrine stimulate the secretion of immunoreactive gonadotropin-releasing hormone &om cultured human placental cells. J. Clin. Endocrinol. Metab., 65, 1020 16. Siler-Khodr, T. M. andKhodr, G. S. (1981). Dose response analysis of GnRH stimulation of HCG release &om human term placenta. Biol. Reprod., 25,353 17. Petragha, F., Vaughan, J. and Vale. W. (1989). Inhibin and activin modulate the release of GnRH, hCG and progesterone from cultured human placental cells. Roc. Natl. Acad. Sci. U S A , 86,5114 18. Currie, A. J.. Fraser, H. M. and Scharpe, R. M. (1981). Human placental receptors for luteinizing hormone releasing factor. Biochem. Biophys. Res. Commun., 99,332 19. Conn, P. M., Staley, D., Harris, C., Andrews, W. V., Gorospe, W. C., McArdle, C. A., Huckle, W. R. and Hansen,J. (1986). Mechanism ofaction of gonadotropin releasing hormone. Ann. Rev. Physiol., 48, 495 20. Mathialagan, N. and Rao, A. J. (1989). A role of calcium in gonadotrophin-releasing hormone (GnRH) stimulated secretion ofchorionic gonadotrophin by first trimester human placental minces in vitro. Placenta, 10, 61 21. Siler-Khodr, R . M., Kuehl, T. J. and Vickery, B. H. (1984). Effects of a gonadotropin-releasing hormone antagonist on hormonal levels in the pregnant baboon and on fetal outcome. Fertil. Steril., 41,448 22. Petragha, F., Woodruff, T. K., Botticelli, G., Botticelli, A., Genazzani, A. R., Mayo, K. E. and Vale, W. (1992). Gonadotropin-releasing hormone, inhibin and activin in human placenta: evidence for a common cellular localization.J . Clin. Endocrinol. Metab., 74, 1184 23. Sopelek,J. N. and Hodgen, G. D. (1987).Infusion of gonadotropin releasing hormone agonist during pregnancy: maternal and fetal responsesin primates. A m . J . Obstet. Gynecol., 156, 755

Gynecological Endocrinology