0013-7227/92/1301-0490$03.00/0 Endocrinology Copyright 0 1992 by The Endocrine
Vol. 130, No. 1
Printed in U.S.A.
Society
Evidence for a Modulation of Human Chorionic Gonadotropin (hCG) Subunit Messenger Ribonucleic Acid Levels and hCG Secretion by y-Aminobutyric in Human First Trimester Placenta in Vitro PETER Department Germany
LIGHT*, of
PETER
Biochemistry
HARBARTH, II, University
of
AND Heidelberg,
WOLFGANG
Acid
E. MERZt
Im Neuenheimer
Feld 328; 6900 Heidelberg,
at 10 PM. The effect on hCG secretion was mimicked by the GABA-A receptor agonist muscimol (P < 0.002), but under the experimental conditions used (multiple pulses; 1 PM), only the fl mRNA was increased. The GABA-A receptor antagonist bicuculline (two pulses; 10 pM) suppressed basal hCG secretion (P < 0.001) and abolished the episodic secretion pattern observed in the control cultures. Applying a combination of equimolar amounts of GABA and bicuculline, hCG secretion and the episodic secretion pattern were similar as in control cultures. The data seem to suggest a regulation of hCG biosynthesis in human first trimester placenta in which GABA is involved, probably acting via GABA-A-like receptor sites. (Endocrinology 130: 490-496,1992)
ABSTRACT. The cytotrophoblasts are the site of production of liberins and statins in human placenta, whereas the syncytiotrophoblasts synthesize tropic hormones. These placental cell layers seem to interact like the hypothalamus and pituitary. In the central nervous system, y-aminobutyric acid (GABA)-ergic neurons represent one important control mechanism that seems to influence the lutropin biosynthesis indirectly (via gonadoliberin) and directly. It was the objective of the present study to find out whether GABA also may influence the biosynthesis and secretion of hCG by human first trimester placenta. Already one single pulse of GABA (1 h; 0.01-100 MM) stimulated hCG secretion significantly (P < 0.0001). GABA also induced a marked increase in the mRNA levels of both subunits, with an optimum
C
special interest. In the placenta, the presence of a GnRH precursor (1) and also the synthesis of a GnRH molecule that is structurally identical with the hypothalamic decapeptide (2, 3) were shown. Low affinity GnRH receptors have been described in placenta tissue (lo), and there is growing evidence that GnRH may stimulate hCG biosynthesis (11) and hCG secretion (11-14) in a way similar to its action on the synthesis of the pituitary gonadotropins. GnRH release in the CNS seems under the control of a complex pulse-generating system in which y-aminobutyric acid (GABA) is involved as either a stimulatory (15) or an inhibitory neurotransmitter (16), depending on the type of receptor involved (GABA-A and GABAB, respectively). Furthermore, the estrogen concentration seem to modulate the response to GABA (17, 18). Recently, GnRH-independent stimulation of LH release has been reported after treatment of pituitary cells with a GABA agonist (19). So far, only the presence of GABAA receptor sites, endogenous GABA, glutamate decarboxylase (an enzyme that catalyzes one GABA-synthesizing pathway) (20), and peripheral benzodiazepinebinding sites (21) have been described in term placenta. In this paper we present evidence that GABA can
ONSIDERABLE homology seems to exist between the hypothalamic-pituitary endocrine axis and the paracrine interactions of cytotrophoblasts and syncytiotrophoblasts in human placenta. Several peptide hormones commonly known as products of hypothalamic origin also produced by the cytotrophoblasts [e.g. gonadoliberin (GnRH) (l-3), corticoliberin (4), somatostatin (5)]. As in the central nervous system (CNS), these mediators have been suggested to interact with the synthesis and secretion of tropic hormones produced by the syncytiotrophoblasts (hCG, human choriocorticotropin, and human choriosomatomammotropin). Recently, placental opiate-binding sites of the K type (6) were described. Treatment with K-agonists stimulated the hCG secretion (7-9), supporting the thesis of a parallelism between the regulation of central nervous and placental gonadotropin release. With respect to hCG biosynthesis by the syncytiotrophoblasts, the role of GnRH is of Received May 10, 1991. * Present address: Department of Obstetrics and Gynecology, University of Louisville School of Medicine, 438 MDR Building, Louisville, Kentucky 40292. t To whom all correspondence and requests for reprints should be addressed. 490
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA MODULATION
OF hCG mRNA
increase the hCG (Y- and P-subunit mRNA and the amount of secreted hCG. This is probably mediated via GABA-A receptors or GABA-A-like receptors and may take part in endogenous processes regulating placental gonadotropin release.
Materials
and Methods
Source of reagents
Tissueculture medium199(M199) with Hanks’ salts,human transferrin, and Norit-A charcoal were purchasedfrom Serva Feinbiochemica(Heidelberg,Germany); GABA, muscimol,and bicuculline-methiodide from Sigma (Munich, Germany); and penicillin G/streptomycin and amphotericin-B from Biochrom KG (Berlin, Germany), Baclofen was kindly provided by Dr. Unger (Department of Pharmacology,University of Heidelberg, Germany). The Third International Standard for choriogonadotropin (Immunoassay) was obtained from WHO International Laboratories for Biological Standards (London, United Kingdom). Tissue preparation
Human first trimester placenta tissue (8th to 10th week of gestation) was obtained from legal interruptions of pregnancy carried out on social groundsin healthy women and was used for the experiments after written consent of the patients had been obtained. Immediately after curettage, the tissue wastransferred into sterile ice-cold Ml99 containing 0.6 mg/liter sodiumbicarbonate, 0.11 g penicillin G (sodium salt), and 0.2 g streptomycin sulfate. It was transported to the laboratory on ice. Under sterile conditions, the tissue was freed from blood clots and nonvillous tissue,washedthoroughly, and dissectedinto pieces of approximately 20-30 mg wet weight. Superfusiontissue culture The superfusiontissue culture apparatus used in these experiments haspreviously been describedin detail (11). Briefly, in a typical experiment 16 tissue pieceswere perifused simultaneously by serum-free Ml99 with Hanks’ salts containing, per liter, 1.5 g sodium bicarbonate, 100,000IU penicillin G/ streptomycin, 2.5 mgamphotericin-B, and 10mg iron-saturated transferrin. The medium circulated continuously between a 2.5-liter tank and a unit where it wasequilibrated with a sterile mixture of 95% air and 5% CO, before being distributed to the tissue piecesthrough capillary glass tees and Teflon tubing sets. GABA, muscimol, baclofen, and bicuculline methiodide were addedseparately to the medium (1:lO ratio of flow rates) by micrometeringpumps (PLG Desaga,Heidelberg,Germany). Mixing with the medium was performed in mixing units equippedwith stirring magnets(mixer driver, LKB, Bromma, Sweden).The tissuepieceswere located in 0.5 ml Teflon tissue culture chamberson a 100~pmpolyester net and superfused continuously with 3 ml/h by a 16-channel peristaltic pump (IPS-16, ISMATEC Corp., Zurich, Switzerland). The effluent medium was collected by a fraction collector placed in a cold box (2 C; Colora Messtechnik, Larch, Germany). Medium and tissueculture chamberswere kept at 37 C. In the first 24 h of
AND
hCG SECRETION
491
culture, all tissuepieceswere treated with the sameconditions as the control cultures. After that, the cultures (except the control cultures) were treated with 2 pulses(unlessotherwise stated; l-h duration each) of GABA, bicuculline, or a combination of GABA and bicuculline, starting the first pulse 24 h and the second 36 h after the beginning of the superfusion culture. In the interval between the 2 pulses and after the application of the secondpulseuntil the end of the experiments (mostly 46-48 h after the start of the culture), all tissuepieces, including the control tissue, were cultured under equal conditions. Radioimmunological
quantification
of hCG
hCG levels were determined by a RIA, usingpolyclonal goat antibodies against the native hormone, as describedin detail previously (11). Cross-reactivities were 1.6% with the free LYsubunit (wt/wt) and 0.3% with the free P-subunit (wt/wt) (Mayer, M., M. Layer, and W. E. Merz, unpublished).Purified hCG (13,000IU/mg) wasobtained, aspreviously described(22), and labeledwith “‘1 according to the method of Markannen et al. (23), using the chloramine-T method. The specific radioactivity of ‘251-labeledhCG was approximately 50 kBecquerel/ pmol (Hilf, G., and W. E. Merz, unpublished). The RIA was performed in microvessels(no. 731055,Sarstedt, Numbrecht, Germany). Incubation with the antibody was carried out in a humidified closedchamber at 37 C for 2 h. Separation of free and antibody-bound hormone wasperformed according to the method of Herbert et al. (24). Samples were measured in triplicate. The intraassay variability was 5.9%; the interassay variability was 14.1%. The Third International Standard for choriogonadotropin served as referencepreparation. RNA preparation
and Northern
blot
Total RNA was prepared from four pooled placenta pieces derived from the same placenta and treated with the same condition at the end of the superfusionculture (mostly 46-48 h after the start of the culture, which correspondsto lo-12 h elapsed since treatment with the second pulse; see above). Tissue was homogenized (20 set; Ultra-Turrax, Janke and Kunkel, Staufen/Breisgau, Germany) in a buffer at pH 7.0 containing 4 M guanidinium isothiocyanate, 25 mM sodium citrate, 0.1 M 2-mercaptoethanol,and 0.5% N-lauroylsarcosine. Purification wasperformed according to the method of Chomczynski and Sacchi (25). RNA was denatured and blotted on GeneScreen(DuPont-NEN, Dreieich, Germany) accordingto the method of Sernia et al. (26). Hybridization with 32P-labeled cDNAs [hCG-a (27) and ,&subunit (28)] and /3-actin (rat) was performed accordingto the method of Church and Gilbert (29). The cDNAs were labeled by a random primer DNA-labeling kit (Boehringer, Mannheim, Germany), using [a-32P]dCTP (3000 Ci/mM, Amersham Buchler, Braunschweig, Germany). The specific radioactivity of the labeled cDNAs was in the range of 5-10 x 10’ cpm/pg. Radioactivity was detected by autoradiography (18 h -80 C; Kodak XAR-5 intensifying screen,Eastman Kodak, Rochester,NY). Quantitative evaluation was performed by laser densitometry (2202 Ultroscan, LKB, Bromma, Sweden) of x-ray films (three scans each) coupled with computer-supportedevaluation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
492
GABA MODULATION
OF hCG mRNA
AND
hCG SECRETION
Endo.
1992 1
Vol130.No
A
Statistical analysis Analysis of variance, calculation of the standard curve using spline-approximated data, and calculation of the hCG medium concentrations as determined in the RIA were performed with our own software (Haag, W., H.-J. Scheuermann, and W. E. Merz, unpublished) on an HP 9825 computer and printed on an HP 9871 A printer. Statistical analysis of similar multiple experiments was performed with a rank test for complete block design according to the procedure of Haux et al. (30).
600 $ 500 E 2: 400
H
4 300 f E 200
600 1
100
Results A single application of GABA (10 PM; 1 h duration) to first trimester placenta explants in superfusion tissue culture caused a highly significant increase in hCG secretion (P < O.OOOl), which became evident about 4-6 h after the start of the GABA pulses and lasted for at least 18 h (Fig. 1). No significant acute changes in hCG secretion during GABA treatment or immediately after that were detected. GABA also stimulated hCG secretion when a smaller concentration was continuously added to the medium during the entire experiment (2.5 pM; P < 0.004; not shown). Concomitantly with the increase in hCG secretion, we observed a significant elevation of mRNA levels for the (Y- and @-subunits of hCG (Fig. 2). No changes in the mRNA levels of controls, such as @actin (Fig. 2A) and a-glycerophosphate dehydrogenase (not shown), were observed. The GABA-mediated stimulation seemed dose dependent, showing an optimum
0
C0ntr.d
GA0A 10 ,IM
FIG. 2. Specific effects of GABA on the (Y- and B-subunit mRNA content (A) and hCG secretion (B). Placental tissue (eighth week of gestation) was treated with two GABA pulses (10 PM; l-h duration each; 24 and 36 h after the start of the superfusion culture). Four tissue pieces, obtained from the same placenta, were cultivated per condition. A, The mRNA levels were determined by Northern blot analysis of total RNA prepared at the end of the experiment from the four pooled tissue pieces. Northern blots were sequentially hybridized with (Ysubunit cDNA (striated column) and b-subunit cDNA (small dots) and were evaluated by densitometry. The specificity of the GABA effect on hCG biosynthesis was assayed by hybridization with &actin cDNA (large dots). Contents of Q- and j3-subunit and actin mRNAs in control cultures (open column) were set at 100%. For mRNA levels, bars show the mean f SD of three measurements. B, Stimulation of hCG secretion (mean + SEM; n = 4 tissue pieces) determined in the fractions collected during the last 3 h of culture was significantly increased (P < 0.001) in GABA-treated cultures (striated) compared to that in untreated control cultures (open column). The results are representative for seven independent experiments of the same type.
1. Influence of different GABA concentrations on hCG subunit
TABLE
mRNAs GABA (PM) 0 0.01 0.1
1 10 100
42
h
Time course of GABA-induced stimulation of hCG secretion. Placental tissue (eighth week of gestation) was treated in superfusion culture with one pulse of 10 PM GABA for 1 h, 24 h after the start of the culture. The inset shows the changes in hCG secretion during the GABA pulse (horizontal bar; A, control cultures; 0, GABA-treated cultures; means of four cultures each). Medium was collected in 90min fractions (inset, 9-min fractions) and assayed for immunoreactive hCG. Each column represents the mean + SEM hCG secretion of four cultures (Cl, control; n , GABA-treated). The abscissa of the large figure refers to the time elapsed in hours from the beginning of the culture. Stimulation of hCG secretion became evident 4-6 h after the pulse had been applied and lasted for at least 18 h. Similar results were observed in seven independent experiments of the same type. FIG.
a mRNA 100 159.6 f 5.9 254.0
+ 3.8
138.4 f 3.8
j3 mRNA (W) 100 190.0 f 38.3 180.0 f 19.2
&Actin mRNA 106 56.7 f 3.4
90.5 f 15.7
525.0 f 9.5
302.0 f 6.9 882.0 -c 170
96.3 f 15.5
114.9 f 5.4
158.0 rt 9.5
101.0 + 5.9
78.8 + 3.4
Placenta explants (eighth week of gestation) were treated with the same protocol as that described in Fig. 2. Two tissue pieces were used per GABA concentration and control, respectively. The Northern blots were sequentially hybridized with (Y- and @-subunit and actin cDNAs. Values are the mean f SD (n = 3).
1.
value of 10 I.IM GABA (Table 1). The @-actin mRNA levels at the various GABA concentrations are considered not to be different. They reflect, at least partially, differences in the amounts of mRNA applied to the electrophoresis and transferred to the blotting membrane. Treatment of the tissue with pulses of the GABA-A receptor agonist muscimol resulted in a marked increase in hCG secretion (P < 0.002) and in the hCG$-subunit mRNA level, whereas the mRNA level for the a-subunit
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA
MODULATION
OF hCG mRNA
remained unchanged and was decreased in some experiments (Fig. 3). Preliminary experiments with the GABAB receptor agonist baclofen yielded no significant changes in hCG secretion (not shown). Two pulses of bicuculline, a GABA-A receptor antagonist, caused a significant decrease in basal hCG secretion (P C 0.001) and the mRNA levels of both subunits (Fig. 4 and Table 2). The episodic secretion pattern observed in the control cultures (Fig. 5A) was almost completely suppressed in response to treatment of the tissue with bicuculline (Fig. 5B). When equimolar amounts (10 PM) of GABA and bicuculline were simultaneously applied, the hCG concentration in the medium reached at least that in the control cultures, the episodic pattern of secretion was similar to that of basal secretion, and the (Y- and @-subunit mRNA levels were slightly increased (Figs. 4 and 5C and Table 2). Discussion Our results show that GABA can stimulate hCG secretion and hCG biosynthesis in human first trimester placenta tissue in vitro in a superfusion culture. The superfusion system is suitable to study in detail the time course of a tissue response to a treatment and 0
A
;c E 2! g 0 1 2
1000
2000
600
1600
600
1200
400
600
200
LOO
0
COIllId
M”rcmlol
Control
Mus.mlot
AND
hCG SECRETION
a
493
Actin
P
FIG. 4. The GABA-A receptor antagonist bicuculline causes a specific decrease in hCG-a- and &subunit mRNA content in first trimester placental tissue. Placenta explants (one donor) from the 10th week of gestation (four tissue pieces per condition) were treated twice (24 and 36 h after the start of the superfusion culture) for 1 h each with 10 PM bicuculline methiodide or an equimolar mixture (10 1~) of GABA and bicuculline. Total RNA was prepared from a pool of four tissue pieces 8 h after the second application of the drugs (42 h after the start of the culture). The same Northern blot was sequentially hybridized with a32P-labeled cDNAs of the Q- and b-subunit and @-actin (as a control for specificity). Bicuculline decreased the mRNA levels of the a- and the &subunits of hCG, whereas the @-actin mRNA level was maintained. Similar results were obtained in four other experiments using placentae from different donors.
E x 4 z 0 'c 0 5 r
0
FIG. 3. Effect of the GABA-A receptor agonist muscimol on the mRNA levels of hCG subunits (A) and hCG secretion (B). In the first 24 h of the superfusion culture, eight tissue pieces (human placenta, one donor, eighth week of gestation) were cultured without any treatment. After that, four tissue pieces were treated with 30-min pulses of 1 PM muscimol, applied every 3 h. The other four tissue pieces remained untreated and served as control cultures. A, Muscimol caused a g-fold increase in fi mRNA levels (small dots), whereas in this experiment the a mRNA content was decreased (striated column) compared to that in the control cultures (open column; set to 100%). Actin mRNA levels (large dots) showed no differences to the control cultures. The (Y- and b-subunit and actin mRNA levels in the control cultures were set at 100%. B, hCG secretion was significantly stimulated by muscimol (P < 0.002; mean of four tissue pieces + SEM). Similar SthUhtiOnS of j3 mRNA and hCG secretion were obtained in three other experiments with the same protocol, using placentae from different donors. The overall a mRNA levels did not significantly differ from the control values.
TABLE 2. Influence of bicuculline content and on the hCG secretion
Treatment
on the (Y- and @subunit mRNA
mRNA content a
a
B-Actin
Control 1.00 1.00 1.00 Bicuculline 0.24 f 0.07 0.26 + 0.02 1.09 + 0.32 Bicuculline 1.36 f 0.32 2.89 f 0.31 0.94 + 0.28 + GABA For experimental conditions, see Fig. 4.
hCG secretion (IU/lO mg. L) 248.0 + 106.0 80.0 f 9.6 273.0 + 112.5
to distinguish transient responses from long-lasting effects on hCG biosynthesis. In addition, we have shown previously that the hCG secretion rate in the superfusion system used is about 4 times higher than that in static cultures due to the improved culture conditions (11). In these studies we have used culture medium containing 10% newborn calf serum (NC@. When we started to investigate the influence of GABA on hCG biosynthesis we performed these experiments, initially in culture medium containing 10% NCS as well as under serum-free conditions (not shown). Since we found neither a qualitative nor a quantitative influence of NCS on the basal or GABA-stimulated hCG secretion, we decided to perform these experiments under serum-free conditions.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
494
GABA
MODULATION
OF hCG mRNA
500 0 1000
M
Time (h)
FIG. 5. Suppression of episodic hCG secretion by the GABA-A receptor antagonist bicuculline and abrogation of this effect by GABA. The hCG levels in the medium of superfusion cultures of three groups of cultures (A-C; four tissue pieces per group) are shown (mean + SD; 90min fractions collected). The tissue pieces derived from the same placenta (16th week of gestation). The first group was left untreated (control cultures; A). The second group (B) was treated with two pulses of 10 PM bicuculline methiodide (1 h each, 24 and 36 h after the beginning of the culture). The third group (C) was treated with a mixture of GABA and bicuculline (10 pM each; same protocol of pulses as in B). In each case hCG secretion during the last 20 h of the experiment (28-48 h after the start of the culture) is depicted. Sometimes, the SD bars coincide with the filled symbols.
It is of interest that a significant transient increase in hCG secretion, as it occurred in acute response to a gonadoliberin (GnRH) pulse (11, 31), was not observed for GABA treatment. This might suggest that the GABA effect is not mediated by GnRH. A GnRH-independent stimulation of the secretion of pituitary gonadotropins by GABA has recently been reported (19). The time course of the response of placental tissue to GABA treatment hints, more likely, at an increase in hCG biosynthesis, rather than at processes that only facilitate hCG secretion. This assumption is confirmed by the specific increase in the (Y- and P-subunit mRNA levels. A noncontinuously episodic secretion has been reported to occur in pregnant women (32) and in normal nonpregnant adults (33). Even in vitro, basal hCG secretion seems to follow an episodic secretion pattern, with a high frequency (31) and a low frequency component (11, 34), both influenced by GnRH treatment (31, 34). The pulse generator(s) of the episodic hCG secretion observed in vivo and in vitro is unknown. As our experiments show, the GABA-A receptor antagonist bicuculline suppressed the episodic secretion pattern observed in the control culturesalmost completely. This might be caused by abolishment of a stimulatory effect of endogenous GABA. Toxic effects of the bicuculline concentration used seem unlikely, since the actin mRNA level was not changed in bicuculline-treated cultures. The hCG
AND
hCG
SECRETION
Endo. 1992 Voll30. No 1
secretion pattern and the hCG subunit mRNA levels also remained undisturbed when GABA and bicuculline were applied simultaneously. The (Y- and ,&subunit mRNA levels were even increased, possibly since the bicuculline concentration was not sufficient to abrogate the effects of endogenous and exogenous GABA. The suppression of the episodic hCG secretion pattern by bicuculline might suggest a direct or indirect influence of GABAmediated mechanisms on the placental pulse-generating system, as it has been shown in the CNS (14,X5). In the placenta, a GABA-mediated signalling must take another route, since the placenta itself is not innervated. The GABA-A receptor agonist muscimol(35,36) could mimic the effects of GABA on hCG secretion and the mRNA levels for the hCG-/3-subunit (for reviews concerning GABA receptor pharmacology, see Ref. 37). No stimulation of hCG-a-subunit mRNA was observed; in some placentae it was even inhibited. We have no explanation for this phenomenon now, but in these experiments we used multiple pulses for application of muscimol, and that might have some influence on the result. For GnRH-induced changes in the LH subunit mRNAs, the frequency and mode of GnRH application were shown to change the ratio of (Y- and P-subunit mRNAs (38, 39). Also, in the GABA-treated cultures the (.y/p mRNA ratios differed greatly between placental tissue from various donors, as is obvious from a comparison of the data shown in Fig. 2 and Table 1. Therefore, the present experiments seem to show only that the CY-and @-subunit mRNA levels are both increased in GABAtreated cultures compared to those in untreated controls, with no clear preference to one subunit mRNA. Also, basal and GABA-induced hCG secretion by placenta tissue of various donors differed in a similar wide range. As yet, we have not been able to find any morphological or biochemical criteria other than those used in the present studies that correlate with the different responsiveness of the tissue to GABA. Since muscimol stimulated and the GABA-A receptor antagonist bicuculline (40) suppressed basal hCG secretion and the mRNA levels of the hCG subunits, we suggest a GABA-A receptor-mediated or GABA-A-like receptor-mediated regulation of hCG biosynthesis in human first trimester placenta. Our results agree with the findings of Erdo et al. (20), who reported the presence of GABA, GABA-A receptor sites, and glutamate decarboxylase, a GABAsynthesizing enzyme, in human term placenta. To our knowledge, nothing is known of whether this also applies for the first trimester placenta. Erdo and co-workers suggested a physiological role for endogenous GABA in the control of placental blood flow. Preliminary experiments with the GABA-B receptor agonist baclofen did not show a significant effect on hCG secretion. However, this does not exclude that other
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA MODULATION
OF hCG mRNA
types of GABA receptors might be involved in regulatory processes in the placenta. That type of GABA-A receptor which mediates the effects of GABA described in this paper may not be of the classical GABA-A type, since bicuculline did not suppress the GABA action when applied simultaneously with GABA. The same effect was observed by Virmani et al. (19) concerning a GABA effect exerted directly on LH biosynthesis and secretion that was mediated by nonclassical GABA-A receptors in the pituitary. This might represent further support for the thesis of analogies between the control mechanisms of gonadotropin synthesis in the pituitary and the placenta. Further investigations are necessary to show whether GABA acts in the placenta directly on hCG biosynthesis or is mediated by an increase in placental GnRH synthesis. Acknowledgments The authors are indebted to Dr. J. C. Fiddes (California Biotechnology, Mountain View, CA) for providing the cDNA clonesfor the hCG subunits,to Dr. D. Werner (German Cancer ResearchCenter, Heidelberg, Germany) for the actin cDNA clone, to Dr. V. Glatz (Neckargemiind, Germany) for his support in the attainment of the placenta tissue, and to E. Mohr for skillful technical assistance.
References 1. Seeburg PH, Adelman JP 1984 Characterization
of cDNA for
precursorof luteinizing hormonereleasinghormone.Nature 311:666-666
2. Khodr GS, Siler-Khodr TM 1980 Placental luteinizing hormonereleasing factor and its synthesis. Science 207:315-317 3. Tan L, Rousseau P 1982 The chemical identity of the immunoreactive LHRH-like peptide biosynthesized in the human placenta. Biochem Biophys Res Commun 109:1061-1071 4. Sasaki A, Liotta AS, Luckey MM, Margioris AN, Suda T, Krieger DT 1984 Immunoreactive corticotropin-releasing factor is present in human maternal plasma during the third trimester of pregnancy. J Clin Endocrinol Metab 59:812-814 5. Watkins WB, Yen SSC 1980 Somatostatin in cytotrophoblast of the immature human placenta: localization by immunoperoxidase cytochemistry. J Clin Endocrinol Metab 50:969-971 6. Porthe G, Valet& A, Cros J 1981 Kappa opiate binding sites in human placenta. Biochem Biophys Res Commun lOl:l-6 7. Belisle S, Peptit A, Gallo-Payet N, Bellabarba D, Lehoux JG, Lemaire S 1988 Functional opioid receptor sites in human placentas. J Clin Endocrinol Metab 66:283-289 8. Valette A. Tafani M. Porthe G. Pontonnier G. Cros J 1983 Placental kappa’binding site: interaction with dynohhin and its possible imp&ion in hCG secretion. Life Sci [Sippl-I] 33:523-526 9. Cemerikic B. Genbacev 0. Sulovic V. Beaconsfield R 1988 Effect of morphine’on hCG release by first trimester human trophoblast in vitro. Life Sci 42:1773-1779 10. Iwashita M, Evans MI, Catt KJ 1986 Characterization of a gonadotropin-releasing hormone receptor site in term placenta and chorionic villi. J Cl& Endocrinol Metab 62:127-133. 11. Merz WE. Erlewein C. Licht P. Harbarth P 1991 The secretion of human chorionic gonadotropin as well as the (Y- and @mRNA levels are stimulated by exogenous gonadoliberin pulses applied to first trimester placenta in a superfusion culture system. J Clin Endocrinol Metab 73:84-92 12. Biitzow R 1982 Luteinizing hormone-releasing factor increases
AND
hCG SECRETION
release of human chorionic gonadotropin in isolated cell columns of normal and malignant trophoblaste. Int J Cancer 29:9-11 13. Mathialagan N, Jagannadha Rao A 1986 Gonadotropin releasing hormone (GnRH) stimulates both secretion and svnthesis of human chorionic gonadotropin by first trimester placental minces in vitro. Biochem Int 13:757-765 14. Siler-Khodr TM, Khodr GS, Valenzuela G, 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 34245-254 15. Pass KA, Ondo JG 1977 The effect of y-aminobutyric acid on prolactin and gonadotropin secretion in the unanesthetized rat. Endocrinology 100:1437-1442 16. Masotto C, Wisniewski G, Negro-Vilar A 1989 Different r-aminobutyric acid receptor subtypes are involved in the regulation of opiate-dependent and independent luteinizing hormone-releasing hormone secretion. Endocrinology 125:548-553 17. Fliigge G, Oertel WH, Wuttke W 1986 Evidence for estrogenreceptive GABAergic neurons in the preoptic/anterior hypothalamic area of the rat brain. Neuroendocrinology 43:1-5 18. Jarry H, Perschl A, Wuttke W 1988 Further evidence that preoptic anterior GABAergic neurons are part of the GnRH pulse and surge generator. Acta Eidocrinol (Copenh) 118:573 19. Virmani MA. Stoiilkovic SS. Catt KJ 1990 Stimulation of luteinizing hormone release by yiaminobutyric acid (GABA) agonists: mediation by GABA-A-type receptors and activation of chloride and voltage-sensitive calcium channels. Endocrinology 126:24992505 20. Erdij SL, Laszlo A, Kiss B, Zsolnai B 1985 Presence of y-aminobutyric acid and its specific receptor binding sites in human term placenta. Gynecol Obstet Invest 20:199-203 21. Fares F, Gavish M 1986 Characterization of peripheral benzodiazepine binding sites in human term placenta. Biochem Pharmacol 35:227-230 22. Merz WE, Hilgenfeld U, Dijrner M, Brossmer R 1974 Biological, immunological and physical investigations on human chorionic gonadotropin. Hoppe Seylers Z Physiol Chem 355:1035-1045 23. Markkanen S, Tiillikkii K, Jiiliskeliiinen K, Rajaniemi H 1980 Effect of radioiodination on the structural intergrity, receptor and antibody-binding activity and circulatory behavior of human choriogonahotropin:Horm &?s 12:35-45 24. Herbert V. Lau KS. Gorrlieb CW. Bleicher SJ 1965 Coated charcoal immunoassay of insulin. J Clin Endocrinol Metab 25:1375-1384 25. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159 26. Sernia C, Clement8 JA, Funder JW 1989 Regulation of liver angiotensinogen mRNA by glucocorticoids and thyroxine. Mol Cell Endocrinol61:147-156 27. Fiddes JC, Goodman HM 1979 Isolation, cloning and sequence analysis of cDNA for the a-subunit of human chorionic gonadotropin. Nature 281:351-356 28. Fiddes JC, Goodman HM 1980 The cDNA for the &subunit of human chorionic gonadotropin suggests evolution of a gene by readthrough into the untranslated region. Nature 286:684-687 29. Church GM, Gilbert W 1984 Genomic sequencing. Proc Nat1 Acad Sci USA 81:1991-1995 30. Haux R, Schumacher M, Weckesser G 1984 Rank tests for complete block designs. Biomath J 5567-582 31. Barnea ER, Kaplan M 1989 Spontaneous, gonadotropin-releasing hormone-induced and progesterone-inhibited pulsatile secretion of human chorionic gonadoGopin in the first trimester placenta in vitro. J Clin Endocrinol Metab 69215-217 32. Owens O’DM, Ryan KJ, Tulchinsky D 1981 Episiodic secretion of human chorionic gonadotropin in early pregnancy. J Clin Endocrinol Metab 53:1307-1309 33. Ode11 WD, Griffin J 1987 Pulsatile secretion of human chorionic gonadotropin in normal adults. N Engl J Med 317%X38-1691 34. Merz WE. Erlewein C. Licht P. Waener TOF 1989 Evidence for episodic GnRH-modulated hCG secretion by placental tissue in vitro. Acta Endocrinol [Suppl l] (Copenh) 120:103
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
496
GABA
MODULATION
OF hCG mRNA
35. Sigel E, Mamalaki C, Barnard EA 1982 Isolation of a GABA receptor from bovine brain using a benzodiazepine affinity column. FEBS Lett 147:45-48 36. Sigel E, Stephenson FA, Mamalaki C, Barnard EA 1983 A yaminobutyric acid/benzodiazepine receptor complex from bovine cerebral cortex. Purification and partial characterization. J Biol Cbem 258:6965-6971 37. Stephenson FA 1988 Understanding the GABA-A receptor: a chemically gated ion channel. Biochem J 24921-32 38. Haisenleder DJ, Katt JA, Ortolano GA, El-Gewely MR, Duncan
AND
hCG SECRETION
Endo. Voll30.
1992 No 1
JA, Marshall JC 1988 Influence of gonadotropin-releasing hormone pulse amplitude, frequency, and treatment duration on the regulation of luteinizing hormone (LH) subunit messenger ribonucleic acids and LH secretion. Mol Endocrinol2: 338-343 39. Weiss J, Jameson JL, Burrin JM, Crowley Jr WF 1990 Divergent responses of gonadotropin subunit messenger RNAs to continuous versus pulsatile gonadotropin-releasing hormone in vitro. Mol Endocrinol4:557-564 40. Turner AJ, Whittle SR 1983 Biochemical dissection of the yaminobutyrate synapse. Biochem J 20929-41
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
Vol. 130, No. 1
Printed in U.S.A.
Society
Evidence for a Modulation of Human Chorionic Gonadotropin (hCG) Subunit Messenger Ribonucleic Acid Levels and hCG Secretion by y-Aminobutyric in Human First Trimester Placenta in Vitro PETER Department Germany
LIGHT*, of
PETER
Biochemistry
HARBARTH, II, University
of
AND Heidelberg,
WOLFGANG
Acid
E. MERZt
Im Neuenheimer
Feld 328; 6900 Heidelberg,
at 10 PM. The effect on hCG secretion was mimicked by the GABA-A receptor agonist muscimol (P < 0.002), but under the experimental conditions used (multiple pulses; 1 PM), only the fl mRNA was increased. The GABA-A receptor antagonist bicuculline (two pulses; 10 pM) suppressed basal hCG secretion (P < 0.001) and abolished the episodic secretion pattern observed in the control cultures. Applying a combination of equimolar amounts of GABA and bicuculline, hCG secretion and the episodic secretion pattern were similar as in control cultures. The data seem to suggest a regulation of hCG biosynthesis in human first trimester placenta in which GABA is involved, probably acting via GABA-A-like receptor sites. (Endocrinology 130: 490-496,1992)
ABSTRACT. The cytotrophoblasts are the site of production of liberins and statins in human placenta, whereas the syncytiotrophoblasts synthesize tropic hormones. These placental cell layers seem to interact like the hypothalamus and pituitary. In the central nervous system, y-aminobutyric acid (GABA)-ergic neurons represent one important control mechanism that seems to influence the lutropin biosynthesis indirectly (via gonadoliberin) and directly. It was the objective of the present study to find out whether GABA also may influence the biosynthesis and secretion of hCG by human first trimester placenta. Already one single pulse of GABA (1 h; 0.01-100 MM) stimulated hCG secretion significantly (P < 0.0001). GABA also induced a marked increase in the mRNA levels of both subunits, with an optimum
C
special interest. In the placenta, the presence of a GnRH precursor (1) and also the synthesis of a GnRH molecule that is structurally identical with the hypothalamic decapeptide (2, 3) were shown. Low affinity GnRH receptors have been described in placenta tissue (lo), and there is growing evidence that GnRH may stimulate hCG biosynthesis (11) and hCG secretion (11-14) in a way similar to its action on the synthesis of the pituitary gonadotropins. GnRH release in the CNS seems under the control of a complex pulse-generating system in which y-aminobutyric acid (GABA) is involved as either a stimulatory (15) or an inhibitory neurotransmitter (16), depending on the type of receptor involved (GABA-A and GABAB, respectively). Furthermore, the estrogen concentration seem to modulate the response to GABA (17, 18). Recently, GnRH-independent stimulation of LH release has been reported after treatment of pituitary cells with a GABA agonist (19). So far, only the presence of GABAA receptor sites, endogenous GABA, glutamate decarboxylase (an enzyme that catalyzes one GABA-synthesizing pathway) (20), and peripheral benzodiazepinebinding sites (21) have been described in term placenta. In this paper we present evidence that GABA can
ONSIDERABLE homology seems to exist between the hypothalamic-pituitary endocrine axis and the paracrine interactions of cytotrophoblasts and syncytiotrophoblasts in human placenta. Several peptide hormones commonly known as products of hypothalamic origin also produced by the cytotrophoblasts [e.g. gonadoliberin (GnRH) (l-3), corticoliberin (4), somatostatin (5)]. As in the central nervous system (CNS), these mediators have been suggested to interact with the synthesis and secretion of tropic hormones produced by the syncytiotrophoblasts (hCG, human choriocorticotropin, and human choriosomatomammotropin). Recently, placental opiate-binding sites of the K type (6) were described. Treatment with K-agonists stimulated the hCG secretion (7-9), supporting the thesis of a parallelism between the regulation of central nervous and placental gonadotropin release. With respect to hCG biosynthesis by the syncytiotrophoblasts, the role of GnRH is of Received May 10, 1991. * Present address: Department of Obstetrics and Gynecology, University of Louisville School of Medicine, 438 MDR Building, Louisville, Kentucky 40292. t To whom all correspondence and requests for reprints should be addressed. 490
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA MODULATION
OF hCG mRNA
increase the hCG (Y- and P-subunit mRNA and the amount of secreted hCG. This is probably mediated via GABA-A receptors or GABA-A-like receptors and may take part in endogenous processes regulating placental gonadotropin release.
Materials
and Methods
Source of reagents
Tissueculture medium199(M199) with Hanks’ salts,human transferrin, and Norit-A charcoal were purchasedfrom Serva Feinbiochemica(Heidelberg,Germany); GABA, muscimol,and bicuculline-methiodide from Sigma (Munich, Germany); and penicillin G/streptomycin and amphotericin-B from Biochrom KG (Berlin, Germany), Baclofen was kindly provided by Dr. Unger (Department of Pharmacology,University of Heidelberg, Germany). The Third International Standard for choriogonadotropin (Immunoassay) was obtained from WHO International Laboratories for Biological Standards (London, United Kingdom). Tissue preparation
Human first trimester placenta tissue (8th to 10th week of gestation) was obtained from legal interruptions of pregnancy carried out on social groundsin healthy women and was used for the experiments after written consent of the patients had been obtained. Immediately after curettage, the tissue wastransferred into sterile ice-cold Ml99 containing 0.6 mg/liter sodiumbicarbonate, 0.11 g penicillin G (sodium salt), and 0.2 g streptomycin sulfate. It was transported to the laboratory on ice. Under sterile conditions, the tissue was freed from blood clots and nonvillous tissue,washedthoroughly, and dissectedinto pieces of approximately 20-30 mg wet weight. Superfusiontissue culture The superfusiontissue culture apparatus used in these experiments haspreviously been describedin detail (11). Briefly, in a typical experiment 16 tissue pieceswere perifused simultaneously by serum-free Ml99 with Hanks’ salts containing, per liter, 1.5 g sodium bicarbonate, 100,000IU penicillin G/ streptomycin, 2.5 mgamphotericin-B, and 10mg iron-saturated transferrin. The medium circulated continuously between a 2.5-liter tank and a unit where it wasequilibrated with a sterile mixture of 95% air and 5% CO, before being distributed to the tissue piecesthrough capillary glass tees and Teflon tubing sets. GABA, muscimol, baclofen, and bicuculline methiodide were addedseparately to the medium (1:lO ratio of flow rates) by micrometeringpumps (PLG Desaga,Heidelberg,Germany). Mixing with the medium was performed in mixing units equippedwith stirring magnets(mixer driver, LKB, Bromma, Sweden).The tissuepieceswere located in 0.5 ml Teflon tissue culture chamberson a 100~pmpolyester net and superfused continuously with 3 ml/h by a 16-channel peristaltic pump (IPS-16, ISMATEC Corp., Zurich, Switzerland). The effluent medium was collected by a fraction collector placed in a cold box (2 C; Colora Messtechnik, Larch, Germany). Medium and tissueculture chamberswere kept at 37 C. In the first 24 h of
AND
hCG SECRETION
491
culture, all tissuepieceswere treated with the sameconditions as the control cultures. After that, the cultures (except the control cultures) were treated with 2 pulses(unlessotherwise stated; l-h duration each) of GABA, bicuculline, or a combination of GABA and bicuculline, starting the first pulse 24 h and the second 36 h after the beginning of the superfusion culture. In the interval between the 2 pulses and after the application of the secondpulseuntil the end of the experiments (mostly 46-48 h after the start of the culture), all tissuepieces, including the control tissue, were cultured under equal conditions. Radioimmunological
quantification
of hCG
hCG levels were determined by a RIA, usingpolyclonal goat antibodies against the native hormone, as describedin detail previously (11). Cross-reactivities were 1.6% with the free LYsubunit (wt/wt) and 0.3% with the free P-subunit (wt/wt) (Mayer, M., M. Layer, and W. E. Merz, unpublished).Purified hCG (13,000IU/mg) wasobtained, aspreviously described(22), and labeledwith “‘1 according to the method of Markannen et al. (23), using the chloramine-T method. The specific radioactivity of ‘251-labeledhCG was approximately 50 kBecquerel/ pmol (Hilf, G., and W. E. Merz, unpublished). The RIA was performed in microvessels(no. 731055,Sarstedt, Numbrecht, Germany). Incubation with the antibody was carried out in a humidified closedchamber at 37 C for 2 h. Separation of free and antibody-bound hormone wasperformed according to the method of Herbert et al. (24). Samples were measured in triplicate. The intraassay variability was 5.9%; the interassay variability was 14.1%. The Third International Standard for choriogonadotropin served as referencepreparation. RNA preparation
and Northern
blot
Total RNA was prepared from four pooled placenta pieces derived from the same placenta and treated with the same condition at the end of the superfusionculture (mostly 46-48 h after the start of the culture, which correspondsto lo-12 h elapsed since treatment with the second pulse; see above). Tissue was homogenized (20 set; Ultra-Turrax, Janke and Kunkel, Staufen/Breisgau, Germany) in a buffer at pH 7.0 containing 4 M guanidinium isothiocyanate, 25 mM sodium citrate, 0.1 M 2-mercaptoethanol,and 0.5% N-lauroylsarcosine. Purification wasperformed according to the method of Chomczynski and Sacchi (25). RNA was denatured and blotted on GeneScreen(DuPont-NEN, Dreieich, Germany) accordingto the method of Sernia et al. (26). Hybridization with 32P-labeled cDNAs [hCG-a (27) and ,&subunit (28)] and /3-actin (rat) was performed accordingto the method of Church and Gilbert (29). The cDNAs were labeled by a random primer DNA-labeling kit (Boehringer, Mannheim, Germany), using [a-32P]dCTP (3000 Ci/mM, Amersham Buchler, Braunschweig, Germany). The specific radioactivity of the labeled cDNAs was in the range of 5-10 x 10’ cpm/pg. Radioactivity was detected by autoradiography (18 h -80 C; Kodak XAR-5 intensifying screen,Eastman Kodak, Rochester,NY). Quantitative evaluation was performed by laser densitometry (2202 Ultroscan, LKB, Bromma, Sweden) of x-ray films (three scans each) coupled with computer-supportedevaluation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
492
GABA MODULATION
OF hCG mRNA
AND
hCG SECRETION
Endo.
1992 1
Vol130.No
A
Statistical analysis Analysis of variance, calculation of the standard curve using spline-approximated data, and calculation of the hCG medium concentrations as determined in the RIA were performed with our own software (Haag, W., H.-J. Scheuermann, and W. E. Merz, unpublished) on an HP 9825 computer and printed on an HP 9871 A printer. Statistical analysis of similar multiple experiments was performed with a rank test for complete block design according to the procedure of Haux et al. (30).
600 $ 500 E 2: 400
H
4 300 f E 200
600 1
100
Results A single application of GABA (10 PM; 1 h duration) to first trimester placenta explants in superfusion tissue culture caused a highly significant increase in hCG secretion (P < O.OOOl), which became evident about 4-6 h after the start of the GABA pulses and lasted for at least 18 h (Fig. 1). No significant acute changes in hCG secretion during GABA treatment or immediately after that were detected. GABA also stimulated hCG secretion when a smaller concentration was continuously added to the medium during the entire experiment (2.5 pM; P < 0.004; not shown). Concomitantly with the increase in hCG secretion, we observed a significant elevation of mRNA levels for the (Y- and @-subunits of hCG (Fig. 2). No changes in the mRNA levels of controls, such as @actin (Fig. 2A) and a-glycerophosphate dehydrogenase (not shown), were observed. The GABA-mediated stimulation seemed dose dependent, showing an optimum
0
C0ntr.d
GA0A 10 ,IM
FIG. 2. Specific effects of GABA on the (Y- and B-subunit mRNA content (A) and hCG secretion (B). Placental tissue (eighth week of gestation) was treated with two GABA pulses (10 PM; l-h duration each; 24 and 36 h after the start of the superfusion culture). Four tissue pieces, obtained from the same placenta, were cultivated per condition. A, The mRNA levels were determined by Northern blot analysis of total RNA prepared at the end of the experiment from the four pooled tissue pieces. Northern blots were sequentially hybridized with (Ysubunit cDNA (striated column) and b-subunit cDNA (small dots) and were evaluated by densitometry. The specificity of the GABA effect on hCG biosynthesis was assayed by hybridization with &actin cDNA (large dots). Contents of Q- and j3-subunit and actin mRNAs in control cultures (open column) were set at 100%. For mRNA levels, bars show the mean f SD of three measurements. B, Stimulation of hCG secretion (mean + SEM; n = 4 tissue pieces) determined in the fractions collected during the last 3 h of culture was significantly increased (P < 0.001) in GABA-treated cultures (striated) compared to that in untreated control cultures (open column). The results are representative for seven independent experiments of the same type.
1. Influence of different GABA concentrations on hCG subunit
TABLE
mRNAs GABA (PM) 0 0.01 0.1
1 10 100
42
h
Time course of GABA-induced stimulation of hCG secretion. Placental tissue (eighth week of gestation) was treated in superfusion culture with one pulse of 10 PM GABA for 1 h, 24 h after the start of the culture. The inset shows the changes in hCG secretion during the GABA pulse (horizontal bar; A, control cultures; 0, GABA-treated cultures; means of four cultures each). Medium was collected in 90min fractions (inset, 9-min fractions) and assayed for immunoreactive hCG. Each column represents the mean + SEM hCG secretion of four cultures (Cl, control; n , GABA-treated). The abscissa of the large figure refers to the time elapsed in hours from the beginning of the culture. Stimulation of hCG secretion became evident 4-6 h after the pulse had been applied and lasted for at least 18 h. Similar results were observed in seven independent experiments of the same type. FIG.
a mRNA 100 159.6 f 5.9 254.0
+ 3.8
138.4 f 3.8
j3 mRNA (W) 100 190.0 f 38.3 180.0 f 19.2
&Actin mRNA 106 56.7 f 3.4
90.5 f 15.7
525.0 f 9.5
302.0 f 6.9 882.0 -c 170
96.3 f 15.5
114.9 f 5.4
158.0 rt 9.5
101.0 + 5.9
78.8 + 3.4
Placenta explants (eighth week of gestation) were treated with the same protocol as that described in Fig. 2. Two tissue pieces were used per GABA concentration and control, respectively. The Northern blots were sequentially hybridized with (Y- and @-subunit and actin cDNAs. Values are the mean f SD (n = 3).
1.
value of 10 I.IM GABA (Table 1). The @-actin mRNA levels at the various GABA concentrations are considered not to be different. They reflect, at least partially, differences in the amounts of mRNA applied to the electrophoresis and transferred to the blotting membrane. Treatment of the tissue with pulses of the GABA-A receptor agonist muscimol resulted in a marked increase in hCG secretion (P < 0.002) and in the hCG$-subunit mRNA level, whereas the mRNA level for the a-subunit
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA
MODULATION
OF hCG mRNA
remained unchanged and was decreased in some experiments (Fig. 3). Preliminary experiments with the GABAB receptor agonist baclofen yielded no significant changes in hCG secretion (not shown). Two pulses of bicuculline, a GABA-A receptor antagonist, caused a significant decrease in basal hCG secretion (P C 0.001) and the mRNA levels of both subunits (Fig. 4 and Table 2). The episodic secretion pattern observed in the control cultures (Fig. 5A) was almost completely suppressed in response to treatment of the tissue with bicuculline (Fig. 5B). When equimolar amounts (10 PM) of GABA and bicuculline were simultaneously applied, the hCG concentration in the medium reached at least that in the control cultures, the episodic pattern of secretion was similar to that of basal secretion, and the (Y- and @-subunit mRNA levels were slightly increased (Figs. 4 and 5C and Table 2). Discussion Our results show that GABA can stimulate hCG secretion and hCG biosynthesis in human first trimester placenta tissue in vitro in a superfusion culture. The superfusion system is suitable to study in detail the time course of a tissue response to a treatment and 0
A
;c E 2! g 0 1 2
1000
2000
600
1600
600
1200
400
600
200
LOO
0
COIllId
M”rcmlol
Control
Mus.mlot
AND
hCG SECRETION
a
493
Actin
P
FIG. 4. The GABA-A receptor antagonist bicuculline causes a specific decrease in hCG-a- and &subunit mRNA content in first trimester placental tissue. Placenta explants (one donor) from the 10th week of gestation (four tissue pieces per condition) were treated twice (24 and 36 h after the start of the superfusion culture) for 1 h each with 10 PM bicuculline methiodide or an equimolar mixture (10 1~) of GABA and bicuculline. Total RNA was prepared from a pool of four tissue pieces 8 h after the second application of the drugs (42 h after the start of the culture). The same Northern blot was sequentially hybridized with a32P-labeled cDNAs of the Q- and b-subunit and @-actin (as a control for specificity). Bicuculline decreased the mRNA levels of the a- and the &subunits of hCG, whereas the @-actin mRNA level was maintained. Similar results were obtained in four other experiments using placentae from different donors.
E x 4 z 0 'c 0 5 r
0
FIG. 3. Effect of the GABA-A receptor agonist muscimol on the mRNA levels of hCG subunits (A) and hCG secretion (B). In the first 24 h of the superfusion culture, eight tissue pieces (human placenta, one donor, eighth week of gestation) were cultured without any treatment. After that, four tissue pieces were treated with 30-min pulses of 1 PM muscimol, applied every 3 h. The other four tissue pieces remained untreated and served as control cultures. A, Muscimol caused a g-fold increase in fi mRNA levels (small dots), whereas in this experiment the a mRNA content was decreased (striated column) compared to that in the control cultures (open column; set to 100%). Actin mRNA levels (large dots) showed no differences to the control cultures. The (Y- and b-subunit and actin mRNA levels in the control cultures were set at 100%. B, hCG secretion was significantly stimulated by muscimol (P < 0.002; mean of four tissue pieces + SEM). Similar SthUhtiOnS of j3 mRNA and hCG secretion were obtained in three other experiments with the same protocol, using placentae from different donors. The overall a mRNA levels did not significantly differ from the control values.
TABLE 2. Influence of bicuculline content and on the hCG secretion
Treatment
on the (Y- and @subunit mRNA
mRNA content a
a
B-Actin
Control 1.00 1.00 1.00 Bicuculline 0.24 f 0.07 0.26 + 0.02 1.09 + 0.32 Bicuculline 1.36 f 0.32 2.89 f 0.31 0.94 + 0.28 + GABA For experimental conditions, see Fig. 4.
hCG secretion (IU/lO mg. L) 248.0 + 106.0 80.0 f 9.6 273.0 + 112.5
to distinguish transient responses from long-lasting effects on hCG biosynthesis. In addition, we have shown previously that the hCG secretion rate in the superfusion system used is about 4 times higher than that in static cultures due to the improved culture conditions (11). In these studies we have used culture medium containing 10% newborn calf serum (NC@. When we started to investigate the influence of GABA on hCG biosynthesis we performed these experiments, initially in culture medium containing 10% NCS as well as under serum-free conditions (not shown). Since we found neither a qualitative nor a quantitative influence of NCS on the basal or GABA-stimulated hCG secretion, we decided to perform these experiments under serum-free conditions.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
494
GABA
MODULATION
OF hCG mRNA
500 0 1000
M
Time (h)
FIG. 5. Suppression of episodic hCG secretion by the GABA-A receptor antagonist bicuculline and abrogation of this effect by GABA. The hCG levels in the medium of superfusion cultures of three groups of cultures (A-C; four tissue pieces per group) are shown (mean + SD; 90min fractions collected). The tissue pieces derived from the same placenta (16th week of gestation). The first group was left untreated (control cultures; A). The second group (B) was treated with two pulses of 10 PM bicuculline methiodide (1 h each, 24 and 36 h after the beginning of the culture). The third group (C) was treated with a mixture of GABA and bicuculline (10 pM each; same protocol of pulses as in B). In each case hCG secretion during the last 20 h of the experiment (28-48 h after the start of the culture) is depicted. Sometimes, the SD bars coincide with the filled symbols.
It is of interest that a significant transient increase in hCG secretion, as it occurred in acute response to a gonadoliberin (GnRH) pulse (11, 31), was not observed for GABA treatment. This might suggest that the GABA effect is not mediated by GnRH. A GnRH-independent stimulation of the secretion of pituitary gonadotropins by GABA has recently been reported (19). The time course of the response of placental tissue to GABA treatment hints, more likely, at an increase in hCG biosynthesis, rather than at processes that only facilitate hCG secretion. This assumption is confirmed by the specific increase in the (Y- and P-subunit mRNA levels. A noncontinuously episodic secretion has been reported to occur in pregnant women (32) and in normal nonpregnant adults (33). Even in vitro, basal hCG secretion seems to follow an episodic secretion pattern, with a high frequency (31) and a low frequency component (11, 34), both influenced by GnRH treatment (31, 34). The pulse generator(s) of the episodic hCG secretion observed in vivo and in vitro is unknown. As our experiments show, the GABA-A receptor antagonist bicuculline suppressed the episodic secretion pattern observed in the control culturesalmost completely. This might be caused by abolishment of a stimulatory effect of endogenous GABA. Toxic effects of the bicuculline concentration used seem unlikely, since the actin mRNA level was not changed in bicuculline-treated cultures. The hCG
AND
hCG
SECRETION
Endo. 1992 Voll30. No 1
secretion pattern and the hCG subunit mRNA levels also remained undisturbed when GABA and bicuculline were applied simultaneously. The (Y- and ,&subunit mRNA levels were even increased, possibly since the bicuculline concentration was not sufficient to abrogate the effects of endogenous and exogenous GABA. The suppression of the episodic hCG secretion pattern by bicuculline might suggest a direct or indirect influence of GABAmediated mechanisms on the placental pulse-generating system, as it has been shown in the CNS (14,X5). In the placenta, a GABA-mediated signalling must take another route, since the placenta itself is not innervated. The GABA-A receptor agonist muscimol(35,36) could mimic the effects of GABA on hCG secretion and the mRNA levels for the hCG-/3-subunit (for reviews concerning GABA receptor pharmacology, see Ref. 37). No stimulation of hCG-a-subunit mRNA was observed; in some placentae it was even inhibited. We have no explanation for this phenomenon now, but in these experiments we used multiple pulses for application of muscimol, and that might have some influence on the result. For GnRH-induced changes in the LH subunit mRNAs, the frequency and mode of GnRH application were shown to change the ratio of (Y- and P-subunit mRNAs (38, 39). Also, in the GABA-treated cultures the (.y/p mRNA ratios differed greatly between placental tissue from various donors, as is obvious from a comparison of the data shown in Fig. 2 and Table 1. Therefore, the present experiments seem to show only that the CY-and @-subunit mRNA levels are both increased in GABAtreated cultures compared to those in untreated controls, with no clear preference to one subunit mRNA. Also, basal and GABA-induced hCG secretion by placenta tissue of various donors differed in a similar wide range. As yet, we have not been able to find any morphological or biochemical criteria other than those used in the present studies that correlate with the different responsiveness of the tissue to GABA. Since muscimol stimulated and the GABA-A receptor antagonist bicuculline (40) suppressed basal hCG secretion and the mRNA levels of the hCG subunits, we suggest a GABA-A receptor-mediated or GABA-A-like receptor-mediated regulation of hCG biosynthesis in human first trimester placenta. Our results agree with the findings of Erdo et al. (20), who reported the presence of GABA, GABA-A receptor sites, and glutamate decarboxylase, a GABAsynthesizing enzyme, in human term placenta. To our knowledge, nothing is known of whether this also applies for the first trimester placenta. Erdo and co-workers suggested a physiological role for endogenous GABA in the control of placental blood flow. Preliminary experiments with the GABA-B receptor agonist baclofen did not show a significant effect on hCG secretion. However, this does not exclude that other
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
GABA MODULATION
OF hCG mRNA
types of GABA receptors might be involved in regulatory processes in the placenta. That type of GABA-A receptor which mediates the effects of GABA described in this paper may not be of the classical GABA-A type, since bicuculline did not suppress the GABA action when applied simultaneously with GABA. The same effect was observed by Virmani et al. (19) concerning a GABA effect exerted directly on LH biosynthesis and secretion that was mediated by nonclassical GABA-A receptors in the pituitary. This might represent further support for the thesis of analogies between the control mechanisms of gonadotropin synthesis in the pituitary and the placenta. Further investigations are necessary to show whether GABA acts in the placenta directly on hCG biosynthesis or is mediated by an increase in placental GnRH synthesis. Acknowledgments The authors are indebted to Dr. J. C. Fiddes (California Biotechnology, Mountain View, CA) for providing the cDNA clonesfor the hCG subunits,to Dr. D. Werner (German Cancer ResearchCenter, Heidelberg, Germany) for the actin cDNA clone, to Dr. V. Glatz (Neckargemiind, Germany) for his support in the attainment of the placenta tissue, and to E. Mohr for skillful technical assistance.
References 1. Seeburg PH, Adelman JP 1984 Characterization
of cDNA for
precursorof luteinizing hormonereleasinghormone.Nature 311:666-666
2. Khodr GS, Siler-Khodr TM 1980 Placental luteinizing hormonereleasing factor and its synthesis. Science 207:315-317 3. Tan L, Rousseau P 1982 The chemical identity of the immunoreactive LHRH-like peptide biosynthesized in the human placenta. Biochem Biophys Res Commun 109:1061-1071 4. Sasaki A, Liotta AS, Luckey MM, Margioris AN, Suda T, Krieger DT 1984 Immunoreactive corticotropin-releasing factor is present in human maternal plasma during the third trimester of pregnancy. J Clin Endocrinol Metab 59:812-814 5. Watkins WB, Yen SSC 1980 Somatostatin in cytotrophoblast of the immature human placenta: localization by immunoperoxidase cytochemistry. J Clin Endocrinol Metab 50:969-971 6. Porthe G, Valet& A, Cros J 1981 Kappa opiate binding sites in human placenta. Biochem Biophys Res Commun lOl:l-6 7. Belisle S, Peptit A, Gallo-Payet N, Bellabarba D, Lehoux JG, Lemaire S 1988 Functional opioid receptor sites in human placentas. J Clin Endocrinol Metab 66:283-289 8. Valette A. Tafani M. Porthe G. Pontonnier G. Cros J 1983 Placental kappa’binding site: interaction with dynohhin and its possible imp&ion in hCG secretion. Life Sci [Sippl-I] 33:523-526 9. Cemerikic B. Genbacev 0. Sulovic V. Beaconsfield R 1988 Effect of morphine’on hCG release by first trimester human trophoblast in vitro. Life Sci 42:1773-1779 10. Iwashita M, Evans MI, Catt KJ 1986 Characterization of a gonadotropin-releasing hormone receptor site in term placenta and chorionic villi. J Cl& Endocrinol Metab 62:127-133. 11. Merz WE. Erlewein C. Licht P. Harbarth P 1991 The secretion of human chorionic gonadotropin as well as the (Y- and @mRNA levels are stimulated by exogenous gonadoliberin pulses applied to first trimester placenta in a superfusion culture system. J Clin Endocrinol Metab 73:84-92 12. Biitzow R 1982 Luteinizing hormone-releasing factor increases
AND
hCG SECRETION
release of human chorionic gonadotropin in isolated cell columns of normal and malignant trophoblaste. Int J Cancer 29:9-11 13. Mathialagan N, Jagannadha Rao A 1986 Gonadotropin releasing hormone (GnRH) stimulates both secretion and svnthesis of human chorionic gonadotropin by first trimester placental minces in vitro. Biochem Int 13:757-765 14. Siler-Khodr TM, Khodr GS, Valenzuela G, 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 34245-254 15. Pass KA, Ondo JG 1977 The effect of y-aminobutyric acid on prolactin and gonadotropin secretion in the unanesthetized rat. Endocrinology 100:1437-1442 16. Masotto C, Wisniewski G, Negro-Vilar A 1989 Different r-aminobutyric acid receptor subtypes are involved in the regulation of opiate-dependent and independent luteinizing hormone-releasing hormone secretion. Endocrinology 125:548-553 17. Fliigge G, Oertel WH, Wuttke W 1986 Evidence for estrogenreceptive GABAergic neurons in the preoptic/anterior hypothalamic area of the rat brain. Neuroendocrinology 43:1-5 18. Jarry H, Perschl A, Wuttke W 1988 Further evidence that preoptic anterior GABAergic neurons are part of the GnRH pulse and surge generator. Acta Eidocrinol (Copenh) 118:573 19. Virmani MA. Stoiilkovic SS. Catt KJ 1990 Stimulation of luteinizing hormone release by yiaminobutyric acid (GABA) agonists: mediation by GABA-A-type receptors and activation of chloride and voltage-sensitive calcium channels. Endocrinology 126:24992505 20. Erdij SL, Laszlo A, Kiss B, Zsolnai B 1985 Presence of y-aminobutyric acid and its specific receptor binding sites in human term placenta. Gynecol Obstet Invest 20:199-203 21. Fares F, Gavish M 1986 Characterization of peripheral benzodiazepine binding sites in human term placenta. Biochem Pharmacol 35:227-230 22. Merz WE, Hilgenfeld U, Dijrner M, Brossmer R 1974 Biological, immunological and physical investigations on human chorionic gonadotropin. Hoppe Seylers Z Physiol Chem 355:1035-1045 23. Markkanen S, Tiillikkii K, Jiiliskeliiinen K, Rajaniemi H 1980 Effect of radioiodination on the structural intergrity, receptor and antibody-binding activity and circulatory behavior of human choriogonahotropin:Horm &?s 12:35-45 24. Herbert V. Lau KS. Gorrlieb CW. Bleicher SJ 1965 Coated charcoal immunoassay of insulin. J Clin Endocrinol Metab 25:1375-1384 25. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159 26. Sernia C, Clement8 JA, Funder JW 1989 Regulation of liver angiotensinogen mRNA by glucocorticoids and thyroxine. Mol Cell Endocrinol61:147-156 27. Fiddes JC, Goodman HM 1979 Isolation, cloning and sequence analysis of cDNA for the a-subunit of human chorionic gonadotropin. Nature 281:351-356 28. Fiddes JC, Goodman HM 1980 The cDNA for the &subunit of human chorionic gonadotropin suggests evolution of a gene by readthrough into the untranslated region. Nature 286:684-687 29. Church GM, Gilbert W 1984 Genomic sequencing. Proc Nat1 Acad Sci USA 81:1991-1995 30. Haux R, Schumacher M, Weckesser G 1984 Rank tests for complete block designs. Biomath J 5567-582 31. Barnea ER, Kaplan M 1989 Spontaneous, gonadotropin-releasing hormone-induced and progesterone-inhibited pulsatile secretion of human chorionic gonadoGopin in the first trimester placenta in vitro. J Clin Endocrinol Metab 69215-217 32. Owens O’DM, Ryan KJ, Tulchinsky D 1981 Episiodic secretion of human chorionic gonadotropin in early pregnancy. J Clin Endocrinol Metab 53:1307-1309 33. Ode11 WD, Griffin J 1987 Pulsatile secretion of human chorionic gonadotropin in normal adults. N Engl J Med 317%X38-1691 34. Merz WE. Erlewein C. Licht P. Waener TOF 1989 Evidence for episodic GnRH-modulated hCG secretion by placental tissue in vitro. Acta Endocrinol [Suppl l] (Copenh) 120:103
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
496
GABA
MODULATION
OF hCG mRNA
35. Sigel E, Mamalaki C, Barnard EA 1982 Isolation of a GABA receptor from bovine brain using a benzodiazepine affinity column. FEBS Lett 147:45-48 36. Sigel E, Stephenson FA, Mamalaki C, Barnard EA 1983 A yaminobutyric acid/benzodiazepine receptor complex from bovine cerebral cortex. Purification and partial characterization. J Biol Cbem 258:6965-6971 37. Stephenson FA 1988 Understanding the GABA-A receptor: a chemically gated ion channel. Biochem J 24921-32 38. Haisenleder DJ, Katt JA, Ortolano GA, El-Gewely MR, Duncan
AND
hCG SECRETION
Endo. Voll30.
1992 No 1
JA, Marshall JC 1988 Influence of gonadotropin-releasing hormone pulse amplitude, frequency, and treatment duration on the regulation of luteinizing hormone (LH) subunit messenger ribonucleic acids and LH secretion. Mol Endocrinol2: 338-343 39. Weiss J, Jameson JL, Burrin JM, Crowley Jr WF 1990 Divergent responses of gonadotropin subunit messenger RNAs to continuous versus pulsatile gonadotropin-releasing hormone in vitro. Mol Endocrinol4:557-564 40. Turner AJ, Whittle SR 1983 Biochemical dissection of the yaminobutyrate synapse. Biochem J 20929-41
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 January 2017. at 04:13 For personal use only. No other uses without permission. . All rights reserved.
