DNA AND CELL BIOLOGY Volume 10, Number 6, 1991 Mary Ann Liebert, Inc., Publishers Pp. 411-421
Gonadotropin-Releasing Hormone and Chorionic Gonadotropin Gene Expression in Human Placental Development AMALIA C. KELLY,* ANNA
RODGERS.t KE-WEN DONG.t NESTOR X. BARREZUETA.t BLUM.t and JAMES L. ROBERTSt
MARIANN
ABSTRACT Control of human chorionic gonadotropin (hCG) synthesis during pregnancy is poorly understood, although in vitro data suggest a role for placental gonadotropin releasing hormone (GnRH) in its regulation. To study GnRH regulation during placental development, placental tissue of different gestational ages was analyzed for GnRH and ßhCG mRNA content. cRNA probes to exonic/intronic sequences of GnRH and ßhCG transcripts were constructed and used to perform solution hybridization/nuclease protection and in situ hybridization assays. The levels of GnRH mRNA were approximately 0.1-1% of that of 0hCG mRNA, in agreement with its suggested paracrine, rather than endocrine, role. While ihC(. mRNA content decreased significantly from first trimester to term (643 to 21.6 pg//¿g RNA), there was no significant change in GnRH mRNA (0.179 to 0.155 pg//¿g RNA). While ßhCG mRNA was localized almost exclusively in syncytiotrophoblasts, GnRH mRNA was present in all cell types of the placenta, including the stroma. In the course of performing sense-strand controls in the in situ hybridization, we noted that the placenta appeared to express more antisense GnRH than sense GnRH mRNA, again, in all cell types. Solution hybridization/nuclease protection analysis with exon 1 and exon 3 probes confirmed this observation, showing that there is two to three times more antisense GnRH RNA than sense GnRH mRNA. These studies suggest that GnRH gene expression and its role in regulating hCG production in human placenta is complex and does not fit a simple model for paracrine regulation of hCG.
INTRODUCTION
(hCG) plays an imrole in the maintenance of the ovarian corpus early pregnancy and may contribute to the regulation of fetal testicular and placental steroidogenesis. Secreted by the placental trophoblast, maternal serum levels of hCG rise rapidly after conception to reach a peak in the first trimester and then fall slowly to a nadir at approximately 18 weeks of gestation where it remains through the remainder of the pregnancy (Braunstein et al, 1976). Several studies have localized hCG to the syncytiotrophoblast by both immunohistochemical (Dreskin et al, 1970; de Ikonicoffand Cedard, 1973; Gaspard et al, 1980) and in situ hybridization (Hoshina et al., 1982, 1983; Wide et al., 1988) techniques. Placental hCG content (Vaitukaitis, 1974) and a-glycoprotein hormone and /3hCG mRNA
Human portant luteum in
chorionic gonadotropin
•Department of
tDr.
Arthur M.
levels (Boothby et al., 1983) decline significantly after the first trimester. The factors regulating placental hCG synthesis have not been clearly defined. Since placental hCG and pituitary luteinizing hormone (LH) bear a close structural and functional resemblance, one postulate is that gonadotropin-releasing hormone (GnRH), the prime hypothalamic regulator of LH secretion, might also play a role in placental regulation of hCG synthesis and secretion. Several reports support this hypothesis. The placenta contains a GnRH decapeptide (Gibbons et al, 1975; Siler-Khodr and Khodr, 1978) identical to hypothalamic GnRH (Tan and Rousseau, 1982) and can synthesize it in vitro (Gibbons et al, 1975; Khodr and Siler-Khodr, 1980). Some (Khodr and Siler-Khodr, 1978a), but not all (Seppala et al, 1980), investigators have localized placental GnRH to the cytotrophoblastic layer, adjacent to the syncytiotrophoblast, which produces hCT.
Obstetrics and Gynecology, Columbia University College of Physicians and Surgeons, New York, NY 10032. Fishberg Research Center for Neurobiology, Mount Sinai School of Medicine, New York, NY 10029.
411
KELLY ET AL.
412
GnRH receptors have also been identified in the placenta ger Mannheim, IN) in 30 ¡A of 10 x SET (10% NaDodS04, (Currie et al., 1981; Belisle et al, 1984), suggesting that all 10 mM Tris HC1 pH 8, 50 mM EDTA), extracted with the machinery is in place for paracrine effects of placental phenol/chloroform, precipitated with ethanol, centriGnRH on hCG production. Work in vitro shows that fuged, dried, and resuspended in 10 mM Tris, 1 mM GnRH stimulates the release of hCG from human placenta EDTA pH 8. Total RNA was quantitated by measuring the in a dose-dependent manner (Khodr and Siler-Khodr, absorbance at 260 nm. The integrity of the total RNA ob1978b; Siler-Khodr and Khodr, 1981; Belisle et al, 1986), tained and absence of chromosomal DNA were verified by and the reverse is seen with a GnRH antagonist (Siler- agarose gel electrophoresis. Khodr et al, 1983), implying that GnRH may be responsible for regulating hCG release in vivo. Both GnRH content Synthesis of probes and reference RNA (Siler-Khodr and Khodr, 1978) and maternal serum levels The phGnRH-7 plasmid (Fig. 1A) was made by subclon(Siler-Khodr et al, 1984) are highest in the first half of pregnancy, the time when placental (Vaitukaitis, 1974) and ing into pGEM part of intron B through exon 3 and part of intron C of the human proGnRH-GAP gene we isolated serum (Braunstein et al, 1976) hCG levels are also highest. Thus, GnRH, synthesized in the cytotrophoblast, may be using synthetic DNA oligomers based on the published núinvolved in regulating hCG synthesis in the syncytiotro- cleo tide sequence (Adelman et al, 1986). An antisense RNA probe was synthesized from linearized phGnRH-7 phoblast. To address the relationship between placental GnRH using SP6 RNA polymerase and labeled with [32P]UTP to mRNA levels and ßhCG mRNA levels, we developed an a specific activity of 1 x 109 cpm/fig. As a result, the assay to quantitate human placental GnRH mRNA, and probe contains vector polylinker sequences (15-20 nucleomeasured GnRH mRNA and /3hCG mRNA throughout tides) on both the 3' and 5' ends, making it longer than the placental development. We also used probes specific for corresponding primary transcript (see Fig. 1A). A second GnRH and hCG mRNA to examine placental tissue in situ plasmid, p5'/3hCG #5, was made by subcloning a fragment to define better the location of their expression during pla- of the #5 figCG gene (Talmadge et al, 1984) spanning 200 cental development. bases of the 5' promoter through exon 1 and 130 bases of intron A into pBluescript (Fig. IB). An antisense probe was synthesized from linearized plasmid using T3 RNA MATERIALS AND METHODS polymerase and labeled with [32P]UTP to a specific activity of 1 x 109 cpm//ig. Sense RNA from each plasmid was Tissue specimens synthesized from the opposite promoter, T7 polymerase Placental tissue was obtained from pregnancies in each for phGnRH-7 and T7 polymerase for p5'/3hCG#5, and trimester of gestation from women undergoing elective used to generate a standard RNA for preparation of the pregnancy termination by dilatation and aspiration, or standard curves for specific RNA quantitation (Blum, from preterm and term deliveries. None of the patients had 1989). significant medical illnesses or fetal anomalies or had received steroid hormone therapy prior to delivery. Memand GnRH mRNA brane-free trophoblastic tissue was immediately dissected, Quantitation of ßhCG frozen by immersion in liquid nitrogen, and stored at /3hCG and GnRH mRNA were quantitated using a solution hybridization/nuclease protection assay (Blum, 1989). -70°C. Approximately 750 pg of probe RNA was mixed with 30 ¡A of hybridization solution (80% formamide, 40 mM PIPES RNA preparation pH 7.0, 400 mM NaCl, 10 mM EDTA pH 8) and 100-150 Total cellular RNA was isolated by the method of fig of total RNA for GnRH mRNA analysis or with 0.5 to Cathala et al. (1983). Individual frozen placental speci- 5.0 ¡ig of total RNA for /3hCG mRNA analysis. For the mens (1-2 grams) were homogenized in 21 ml of ice-cold GnRH mRNA standard curve, an equal amount of probe lysis buffer (5 M guanidine monothiocyanate, 10 mM was mixed with 0, 0.5, 1, 3, 5, or 10 pg of GnRH sense EDTA pH 8, 50 mM Tris HC1 pH 7.5, 8% /3-mercaptoeth- RNA; for the /3hCG mRNA standard curve 0, 5, 10, 25, anol), precipitated with 4 M LiCl at 4CC overnight, col- 50, or 100 pg /3hCG sense RNA was used. After overnight lected by centrifugation at 11,000 x g, for 90 min at 4°C hybridization at 45°C, the mixture was treated with 300 ¡A and washed with 3 M LiCl. Each sample was centrifuged RNase solution (10 mM Tris HC1 pH 7.5, 300 mM NaCl, 5 again at 11,000 x g at 4°C for 60 min, and the pellet was mM EDTA, 40 fig/m\ RNase A, 2 fig/m\ RNase Tl) for 1 resuspended in 15 ml of solubilization buffer (0.1 % hr at 30°C, followed by proteinase K treatment for 15 min NaDodS04, 10 mM Tris HC1 pH 7.5, 1 mMEDTA pH 8). at 37°C. RNA samples were extracted with phenol/chloroFollowing extraction with phenol/chloroform and precipi- form then RNA was precipitated with ethanol, washed, tation with isopropanol, the samples were air-dried and dried, dissolved in 10 ¡A of gel loading buffer, and electrofurther purified by resuspension in 300 ¡A of DNase buffer phoresed through a 5% polyacrylamide nondenaturing gel. (10 mM Tris HC1 pH 7.4, 10 mM NaCl, 2 mM MgCl2, 2 The gel was dried and exposed overnight at room temperamM DTT) and digestion with 40 U of DNase (Worthing- ture to X-ray film; the protected bands were cut from the ton Biochemical Corp., NJ) at 37°C for 5 min. The samgel and the radioactivity quantitated in a scintillation counples were then treated with 3 ¡ig of proteinase K (Boehrin- ter. The amount of radioactivity was plotted against the ~
—
PLACENTAL GnRH/hCH mRNA DURING PREGNANCY
413
A. phGnRH-7 Protected Antisense GnRH Transcript (-271 b)
YSSSSSSSSSSSSS/SSSSSSSSSSSJYSSSjrSSSSSSSSSSSSSSSSSjySSSSSSSSSSSSSSSSSSSSSSSSSS/SSSSSSSSS/S/SSSSS.
Sense Probe or Standard (-31 Ob)
llllMIItlJIIIIMIl—
Hind IM
Exon 3
Intron B(~100b)
T7
Antisense Probe Protected
Xbal
(96b)
Intron C (~75b)
SP6
(-31 Ob)
Primary Transcript (-271 b)
I I M I I M I I I II I I MM M III I I Mil I I I
Protected mRNA (96b)
B. p5' ßhCG #5 £"**"
EcoR
5' Flank
T7
(~200b)
Intron A
Exon 1
(-130b)
(360b)
Antisense Probe (-71 Ob) I II MM II MM I I II II I II I III I I II I I I II I I I II I I I I I I I I II II I I I
Protected mRNA
(360b)
Schematic representation of cRNA probes used in solution hybridization nuclease protection assays and the protected RNA fragments. A. Portion of the proGnRH GAP gene spanning part of intron B, all of exon 3, and part of intron C. B. Portion of ßhCG gene #5 spanning part of the promoter, all of exon 1, and part of intron A. Protected primary transcript and/or protected mRNA are denoted by the thick-hatched lines. In both cases, the probe has sequences at both ends not complementary to the RNAs due to transcribed vector and polylinker sequences (PL), making it longer than the protected primary transcript. The three different bacteriophage RNA polymerase promoters utilized for synthesizing either sense or antisense RNAs, T3, T7, or SP6, are shown in open boxes with arrows denoting the direction of synthesis.
FIG. 1.
known amounts of reference RNA, yielding a regression line from which the content of protected mRNA in each sample was derived. The results are expressed as picograms of RNase-resistant mRNA per microgram of total RNA, corrected for the full-size mRNAs, namely, 580 bp for /3hCG mRNA and 1,500 bp for human GnRH mRNA (Talmadge et al., 1984; Seeburg and Adelman, 1984).
In situ
hybridization
Frozen
lO-^m sections
cut at -20°C and
of first trimester placenta were thaw-mounted on microscope slides pre-
coated with a 300 /¿g/ml sodium azide, 0.15% gelatin, and 10 mg/ml polylysine solution. The sections were air-dried and immediately fixed for 5 min at room temperature in freshly prepared 4.0% neutral-buffered paraformaldehyde, rinsed for 10 min in 2x SSC, and dehydrated through graded alcohols. Sections were stored dessicated at -70°C until use.
Sections were removed from -70°C, thawed, air-dried temperature, and immersed in 2x SSC for 5 min. Excess SSC was removed and each section was covered with 100 fA of prehybridization buffer [50% deionized formamide, 0.6 M NaCl, 10 mM Tris pH 7.5, 0.04% Ficoll, 0.04% polyvinylpyrrolidone (PVP), 0.1% bovine serum albumin (BSA), 1 mM EDTA, 0.5 mg/ml sonicated salmon sperm DNA, 0.5 mg/ml total yeast RNA, and 0.05 mg/ml yeast tRNA]. After 2 h at 50°C in a humid chamber, the prehybridization solution was blotted off the slides and replaced with 100 ¡A hybridization buffer (50% deionized formamide, 0.6 M NaCl, 10 mM Tris pH 7.5, 0.04% Ficoll, 0.04% PVP, 0.1% BSA, 1 mM EDTA, 0.5 mg/ml sonicated salmon sperm DNA, 10% dextran sulfate (wt/ vol, 0.05 mg/ml total yeast RNA, 0.05 ¡ig/m\ yeast tRNA) containing -5 ng/ml of [35S]UTP labeled cRNA (-1 x 109 cpm/^g) or 5 ng/ml digoxigenin-UTP labeled cRNA (synthesized with 2 mMUTP and 8 mM digoxigenin-UTP, Boehringer Mannheim Biochemicals, Indianapolis, IN). at room
~
KELLY ET AL.
414 at 50° C in a humid chamber for washed in 2x SSC for 30 min at room temperature and incubated with 30 /¿g/ml of RNaseA for 45 min at 37°C to digest any unhybridized cRNA probe. The slides were then washed first in 2 liters of 2 x SSC at 50°C for 2 hr and then in 4 liters of 0.1 x SSC containing 0.05% sodium thiosulfate and 14 mM mercaptoethanol for 3 hr at 50° C, after which this wash was allowed to cool to room temperature and continue washing
RESULTS
Following hybridization 12-18 hr, sections
were
overnight.
hybridized to a [35S]UTP labeled probe were dehydrated through 50%, 70%, and 90% ethanols containing 0.3 M ammonium acetate and vacuum-desiccated Sections
for
a
minimum of 2 hr. The slides
were
then coated with
a
layer of Kodak NTB-2 emulsion (Eastman Kodak Co., Rochester, NY). After exposure times varying from 7 to 28 days, the slides were developed with Kodak D-19 developer, fixed with Kodafix, and counterstained with hematoxylin-eosin. Detection
of digoxigenin labeled hybrids
Characterization of the GnRH and ßhCG mRNA
quantitation
assays
The amount of radioactive antisense probe protected from RNase digestion was linearly related to the known amounts of sense RNA with which it hybridized (Fig. 2A). The placental RNA samples measured 100-400 cpm in the GnRH mRNA assay and fell within the standard curve, above background. An autoradiogram of a typical polyacrylamide gel of a GnRH solution hybridization assay is shown in Fig. 2B. When hybridized with total placental RNA, the GnRH probe yielded two protected bands: one of 96 bp corresponding to the fully protected exon 3 as derived from mRNA and the other of 270 bp, corresponding to the GnRH primary transcript. Quantitation of both protected species in a single sample revealed that, on a molar basis, the primary transcript represents 16% of the mRNA. In a parallel study, a probe to exon 1 of GnRH was used yielding values for GnRH mRNA levels comparable to those obtained with the exon 3 probe when normalized to full-length mRNA (data not shown). The abundance of GnRH mRNA was two to three orders of magnitude less than that of /3hCG mRNA.
Sections hybridized to a digoxigenin-UTP-labeled probe immersed overnight in 2x SSC, 0.05% Triton X-100, and 2% normal sheep serum followed by a 10-min wash in 100 mM Tris-HCl pH 7.5, and 150 mM NaCl. The Levels of ßgCG and GnRH mRNA during sections were then incubated with a 1:500 dilution of polydevelopment placental clonal sheep anti-digoxigenin antibody conjugated to either alkaline phosphatase or rhodamine (Boehringer The mean levels ± SEM for each placenta, as well as for Mannheim Biochemicals, Indianapolis, IN) in 100 mM each trimester, are shown in Fig. 3. 0hCG mRNA levels Tris-HCl pH 7.5, 150 mM NaCl, 1% normal sheep serum were highest in the first trimester and decrease markedly and 0.3% Triton X-100 for 5 h at room temperature. (30-fold) by the third trimester. The large variation in firstSections incubated with the rhodamine conjugated anti- trimester levels of /3hCG mRNA could not be explained on body were washed for 10 min in 100 mM Tris-HCl pH 7.5 clinical grounds, since none of the patients had evidence of and 150 mM NaCl. Coverslips were affixed with an aque- multiple gestation (higher levels), molar pregnancy (higher ous mounting medium, and the slides were examined and levels), or failing pregnancy (lower levels). GnRH mRNA photographed with a Leitz Orthoplan microscope levels did show any difference from first to third trimester equipped with a band pass excitation filter in the 515- to (Fig. 3). 560-nm range and a barrier filter at 580 nm. Sections incubated with the alkaline phosphatase-conjuLocalization of GnRH and ßhCG mRNA in placenta gated antibody were reacted with nitroblue tetrazolium salt Both ßhCG and GnRH cRNA probes were used to local(75 mg/ml), 5-bromo-4-chloro-3-indolyl phosphate, toluidinium salt (50 mg/ml), and levamisole (0.24 mg/ml) in ize their respective mRNAs in first-trimester placental tis100 mM Tris-HCl pH 9.5, 100 mM NaCl, and 50 mM sue. As previously reported (Hoshina et al., 1982, 1983; MgCl2 for 16 hr. The reaction was stopped by immersing Wide et al, 1988, 1989), 0hCG mRNA was localized exthe sections in 10 mM Tris-HCl 8.0 and 1 mM EDTA. clusively in the syncytiotrophoblasts giving an appreciable Coverslips were affixed over the sections with an aqueous signal within 6 days of exposure [Fig. 4, hCG(-)]. In conmounting medium, and the slides were examined for the trast, GnRH mRNA is more evenly distributed throughout presence of a blue-brown precipitate and photographed the syncytiotrophoblasts, cytotrophoblasts, and stromal cells of the placenta [Fig. 4, GnRH(-)] when examined using visible light. with a radiolabeled probe. In agreement with a lower level of GnRH mRNA relative to ßhCG mRNA, a 14-day expoData analysis sure gave only a mild signal by in situ hybridization. Consections using sense strand ßhCG figure [Fig. 4, in trol from each placenta were analyzed multiple Samples solution hybridization assays for both /3hCG and GnRH hCG(+)] yielded only background levels of grains. Control sections using sense strand GnRH [Fig. 4, and the mean ± SE for each patient for /3hCG and GnRH mRNA levels was calculated. Differences across trimesters GnRH(+)] showed a level of silver grains two- to threein RNA levels were analyzed using one way analysis of var- fold greater than that seen with antisense probe, suggesting iance. The level of significance was set at 0.05. transcription of the opposite strand in this tissue or possiwere
PLACENTAL GnRH/hCH mRNA DURING PREGNANCY
415
pg RNA
B
Samples
GnRH standards
pg/lane 0
0.5 1.0 3.0 5.0 10
309-*
1
2
3
•Primary transcript (-280 bp)
242-*>
Protected
message
(96 bp) FIG. 2.
A. Typical hGnRH-7 standard curve generated from radioactivity present in protected reference RNA hybrids plotted against known quantities of reference RNA. r 0.998. B. Autoradiogram of polyacrylamide gel of GnRH solution hybridization/nuclease protection assay. The diffuse band seen at 96 bp in several of the GnRH standard lanes represents an unidentified artifact that was not present in all standards or all analyses. M, end-labeled Msp I-cut pBR322 as molecular weight marker. The GnRH standards run at approximately 300 bases, representing additional complementarity at the 5' and 3' ends due to plasmid vector elements; GnRH standard curve 0-10 pg/lane. Samples: 1, first trimester; 2, second timester; 3, third trimester. =
416
KELLY ET AL.
< rr
"rö
S
3 rr
E
I rr
ü Q.
Gonadotropin-Releasing Hormone and Chorionic Gonadotropin Gene Expression in Human Placental Development AMALIA C. KELLY,* ANNA
RODGERS.t KE-WEN DONG.t NESTOR X. BARREZUETA.t BLUM.t and JAMES L. ROBERTSt
MARIANN
ABSTRACT Control of human chorionic gonadotropin (hCG) synthesis during pregnancy is poorly understood, although in vitro data suggest a role for placental gonadotropin releasing hormone (GnRH) in its regulation. To study GnRH regulation during placental development, placental tissue of different gestational ages was analyzed for GnRH and ßhCG mRNA content. cRNA probes to exonic/intronic sequences of GnRH and ßhCG transcripts were constructed and used to perform solution hybridization/nuclease protection and in situ hybridization assays. The levels of GnRH mRNA were approximately 0.1-1% of that of 0hCG mRNA, in agreement with its suggested paracrine, rather than endocrine, role. While ihC(. mRNA content decreased significantly from first trimester to term (643 to 21.6 pg//¿g RNA), there was no significant change in GnRH mRNA (0.179 to 0.155 pg//¿g RNA). While ßhCG mRNA was localized almost exclusively in syncytiotrophoblasts, GnRH mRNA was present in all cell types of the placenta, including the stroma. In the course of performing sense-strand controls in the in situ hybridization, we noted that the placenta appeared to express more antisense GnRH than sense GnRH mRNA, again, in all cell types. Solution hybridization/nuclease protection analysis with exon 1 and exon 3 probes confirmed this observation, showing that there is two to three times more antisense GnRH RNA than sense GnRH mRNA. These studies suggest that GnRH gene expression and its role in regulating hCG production in human placenta is complex and does not fit a simple model for paracrine regulation of hCG.
INTRODUCTION
(hCG) plays an imrole in the maintenance of the ovarian corpus early pregnancy and may contribute to the regulation of fetal testicular and placental steroidogenesis. Secreted by the placental trophoblast, maternal serum levels of hCG rise rapidly after conception to reach a peak in the first trimester and then fall slowly to a nadir at approximately 18 weeks of gestation where it remains through the remainder of the pregnancy (Braunstein et al, 1976). Several studies have localized hCG to the syncytiotrophoblast by both immunohistochemical (Dreskin et al, 1970; de Ikonicoffand Cedard, 1973; Gaspard et al, 1980) and in situ hybridization (Hoshina et al., 1982, 1983; Wide et al., 1988) techniques. Placental hCG content (Vaitukaitis, 1974) and a-glycoprotein hormone and /3hCG mRNA
Human portant luteum in
chorionic gonadotropin
•Department of
tDr.
Arthur M.
levels (Boothby et al., 1983) decline significantly after the first trimester. The factors regulating placental hCG synthesis have not been clearly defined. Since placental hCG and pituitary luteinizing hormone (LH) bear a close structural and functional resemblance, one postulate is that gonadotropin-releasing hormone (GnRH), the prime hypothalamic regulator of LH secretion, might also play a role in placental regulation of hCG synthesis and secretion. Several reports support this hypothesis. The placenta contains a GnRH decapeptide (Gibbons et al, 1975; Siler-Khodr and Khodr, 1978) identical to hypothalamic GnRH (Tan and Rousseau, 1982) and can synthesize it in vitro (Gibbons et al, 1975; Khodr and Siler-Khodr, 1980). Some (Khodr and Siler-Khodr, 1978a), but not all (Seppala et al, 1980), investigators have localized placental GnRH to the cytotrophoblastic layer, adjacent to the syncytiotrophoblast, which produces hCT.
Obstetrics and Gynecology, Columbia University College of Physicians and Surgeons, New York, NY 10032. Fishberg Research Center for Neurobiology, Mount Sinai School of Medicine, New York, NY 10029.
411
KELLY ET AL.
412
GnRH receptors have also been identified in the placenta ger Mannheim, IN) in 30 ¡A of 10 x SET (10% NaDodS04, (Currie et al., 1981; Belisle et al, 1984), suggesting that all 10 mM Tris HC1 pH 8, 50 mM EDTA), extracted with the machinery is in place for paracrine effects of placental phenol/chloroform, precipitated with ethanol, centriGnRH on hCG production. Work in vitro shows that fuged, dried, and resuspended in 10 mM Tris, 1 mM GnRH stimulates the release of hCG from human placenta EDTA pH 8. Total RNA was quantitated by measuring the in a dose-dependent manner (Khodr and Siler-Khodr, absorbance at 260 nm. The integrity of the total RNA ob1978b; Siler-Khodr and Khodr, 1981; Belisle et al, 1986), tained and absence of chromosomal DNA were verified by and the reverse is seen with a GnRH antagonist (Siler- agarose gel electrophoresis. Khodr et al, 1983), implying that GnRH may be responsible for regulating hCG release in vivo. Both GnRH content Synthesis of probes and reference RNA (Siler-Khodr and Khodr, 1978) and maternal serum levels The phGnRH-7 plasmid (Fig. 1A) was made by subclon(Siler-Khodr et al, 1984) are highest in the first half of pregnancy, the time when placental (Vaitukaitis, 1974) and ing into pGEM part of intron B through exon 3 and part of intron C of the human proGnRH-GAP gene we isolated serum (Braunstein et al, 1976) hCG levels are also highest. Thus, GnRH, synthesized in the cytotrophoblast, may be using synthetic DNA oligomers based on the published núinvolved in regulating hCG synthesis in the syncytiotro- cleo tide sequence (Adelman et al, 1986). An antisense RNA probe was synthesized from linearized phGnRH-7 phoblast. To address the relationship between placental GnRH using SP6 RNA polymerase and labeled with [32P]UTP to mRNA levels and ßhCG mRNA levels, we developed an a specific activity of 1 x 109 cpm/fig. As a result, the assay to quantitate human placental GnRH mRNA, and probe contains vector polylinker sequences (15-20 nucleomeasured GnRH mRNA and /3hCG mRNA throughout tides) on both the 3' and 5' ends, making it longer than the placental development. We also used probes specific for corresponding primary transcript (see Fig. 1A). A second GnRH and hCG mRNA to examine placental tissue in situ plasmid, p5'/3hCG #5, was made by subcloning a fragment to define better the location of their expression during pla- of the #5 figCG gene (Talmadge et al, 1984) spanning 200 cental development. bases of the 5' promoter through exon 1 and 130 bases of intron A into pBluescript (Fig. IB). An antisense probe was synthesized from linearized plasmid using T3 RNA MATERIALS AND METHODS polymerase and labeled with [32P]UTP to a specific activity of 1 x 109 cpm//ig. Sense RNA from each plasmid was Tissue specimens synthesized from the opposite promoter, T7 polymerase Placental tissue was obtained from pregnancies in each for phGnRH-7 and T7 polymerase for p5'/3hCG#5, and trimester of gestation from women undergoing elective used to generate a standard RNA for preparation of the pregnancy termination by dilatation and aspiration, or standard curves for specific RNA quantitation (Blum, from preterm and term deliveries. None of the patients had 1989). significant medical illnesses or fetal anomalies or had received steroid hormone therapy prior to delivery. Memand GnRH mRNA brane-free trophoblastic tissue was immediately dissected, Quantitation of ßhCG frozen by immersion in liquid nitrogen, and stored at /3hCG and GnRH mRNA were quantitated using a solution hybridization/nuclease protection assay (Blum, 1989). -70°C. Approximately 750 pg of probe RNA was mixed with 30 ¡A of hybridization solution (80% formamide, 40 mM PIPES RNA preparation pH 7.0, 400 mM NaCl, 10 mM EDTA pH 8) and 100-150 Total cellular RNA was isolated by the method of fig of total RNA for GnRH mRNA analysis or with 0.5 to Cathala et al. (1983). Individual frozen placental speci- 5.0 ¡ig of total RNA for /3hCG mRNA analysis. For the mens (1-2 grams) were homogenized in 21 ml of ice-cold GnRH mRNA standard curve, an equal amount of probe lysis buffer (5 M guanidine monothiocyanate, 10 mM was mixed with 0, 0.5, 1, 3, 5, or 10 pg of GnRH sense EDTA pH 8, 50 mM Tris HC1 pH 7.5, 8% /3-mercaptoeth- RNA; for the /3hCG mRNA standard curve 0, 5, 10, 25, anol), precipitated with 4 M LiCl at 4CC overnight, col- 50, or 100 pg /3hCG sense RNA was used. After overnight lected by centrifugation at 11,000 x g, for 90 min at 4°C hybridization at 45°C, the mixture was treated with 300 ¡A and washed with 3 M LiCl. Each sample was centrifuged RNase solution (10 mM Tris HC1 pH 7.5, 300 mM NaCl, 5 again at 11,000 x g at 4°C for 60 min, and the pellet was mM EDTA, 40 fig/m\ RNase A, 2 fig/m\ RNase Tl) for 1 resuspended in 15 ml of solubilization buffer (0.1 % hr at 30°C, followed by proteinase K treatment for 15 min NaDodS04, 10 mM Tris HC1 pH 7.5, 1 mMEDTA pH 8). at 37°C. RNA samples were extracted with phenol/chloroFollowing extraction with phenol/chloroform and precipi- form then RNA was precipitated with ethanol, washed, tation with isopropanol, the samples were air-dried and dried, dissolved in 10 ¡A of gel loading buffer, and electrofurther purified by resuspension in 300 ¡A of DNase buffer phoresed through a 5% polyacrylamide nondenaturing gel. (10 mM Tris HC1 pH 7.4, 10 mM NaCl, 2 mM MgCl2, 2 The gel was dried and exposed overnight at room temperamM DTT) and digestion with 40 U of DNase (Worthing- ture to X-ray film; the protected bands were cut from the ton Biochemical Corp., NJ) at 37°C for 5 min. The samgel and the radioactivity quantitated in a scintillation counples were then treated with 3 ¡ig of proteinase K (Boehrin- ter. The amount of radioactivity was plotted against the ~
—
PLACENTAL GnRH/hCH mRNA DURING PREGNANCY
413
A. phGnRH-7 Protected Antisense GnRH Transcript (-271 b)
YSSSSSSSSSSSSS/SSSSSSSSSSSJYSSSjrSSSSSSSSSSSSSSSSSjySSSSSSSSSSSSSSSSSSSSSSSSSS/SSSSSSSSS/S/SSSSS.
Sense Probe or Standard (-31 Ob)
llllMIItlJIIIIMIl—
Hind IM
Exon 3
Intron B(~100b)
T7
Antisense Probe Protected
Xbal
(96b)
Intron C (~75b)
SP6
(-31 Ob)
Primary Transcript (-271 b)
I I M I I M I I I II I I MM M III I I Mil I I I
Protected mRNA (96b)
B. p5' ßhCG #5 £"**"
EcoR
5' Flank
T7
(~200b)
Intron A
Exon 1
(-130b)
(360b)
Antisense Probe (-71 Ob) I II MM II MM I I II II I II I III I I II I I I II I I I II I I I I I I I I II II I I I
Protected mRNA
(360b)
Schematic representation of cRNA probes used in solution hybridization nuclease protection assays and the protected RNA fragments. A. Portion of the proGnRH GAP gene spanning part of intron B, all of exon 3, and part of intron C. B. Portion of ßhCG gene #5 spanning part of the promoter, all of exon 1, and part of intron A. Protected primary transcript and/or protected mRNA are denoted by the thick-hatched lines. In both cases, the probe has sequences at both ends not complementary to the RNAs due to transcribed vector and polylinker sequences (PL), making it longer than the protected primary transcript. The three different bacteriophage RNA polymerase promoters utilized for synthesizing either sense or antisense RNAs, T3, T7, or SP6, are shown in open boxes with arrows denoting the direction of synthesis.
FIG. 1.
known amounts of reference RNA, yielding a regression line from which the content of protected mRNA in each sample was derived. The results are expressed as picograms of RNase-resistant mRNA per microgram of total RNA, corrected for the full-size mRNAs, namely, 580 bp for /3hCG mRNA and 1,500 bp for human GnRH mRNA (Talmadge et al., 1984; Seeburg and Adelman, 1984).
In situ
hybridization
Frozen
lO-^m sections
cut at -20°C and
of first trimester placenta were thaw-mounted on microscope slides pre-
coated with a 300 /¿g/ml sodium azide, 0.15% gelatin, and 10 mg/ml polylysine solution. The sections were air-dried and immediately fixed for 5 min at room temperature in freshly prepared 4.0% neutral-buffered paraformaldehyde, rinsed for 10 min in 2x SSC, and dehydrated through graded alcohols. Sections were stored dessicated at -70°C until use.
Sections were removed from -70°C, thawed, air-dried temperature, and immersed in 2x SSC for 5 min. Excess SSC was removed and each section was covered with 100 fA of prehybridization buffer [50% deionized formamide, 0.6 M NaCl, 10 mM Tris pH 7.5, 0.04% Ficoll, 0.04% polyvinylpyrrolidone (PVP), 0.1% bovine serum albumin (BSA), 1 mM EDTA, 0.5 mg/ml sonicated salmon sperm DNA, 0.5 mg/ml total yeast RNA, and 0.05 mg/ml yeast tRNA]. After 2 h at 50°C in a humid chamber, the prehybridization solution was blotted off the slides and replaced with 100 ¡A hybridization buffer (50% deionized formamide, 0.6 M NaCl, 10 mM Tris pH 7.5, 0.04% Ficoll, 0.04% PVP, 0.1% BSA, 1 mM EDTA, 0.5 mg/ml sonicated salmon sperm DNA, 10% dextran sulfate (wt/ vol, 0.05 mg/ml total yeast RNA, 0.05 ¡ig/m\ yeast tRNA) containing -5 ng/ml of [35S]UTP labeled cRNA (-1 x 109 cpm/^g) or 5 ng/ml digoxigenin-UTP labeled cRNA (synthesized with 2 mMUTP and 8 mM digoxigenin-UTP, Boehringer Mannheim Biochemicals, Indianapolis, IN). at room
~
KELLY ET AL.
414 at 50° C in a humid chamber for washed in 2x SSC for 30 min at room temperature and incubated with 30 /¿g/ml of RNaseA for 45 min at 37°C to digest any unhybridized cRNA probe. The slides were then washed first in 2 liters of 2 x SSC at 50°C for 2 hr and then in 4 liters of 0.1 x SSC containing 0.05% sodium thiosulfate and 14 mM mercaptoethanol for 3 hr at 50° C, after which this wash was allowed to cool to room temperature and continue washing
RESULTS
Following hybridization 12-18 hr, sections
were
overnight.
hybridized to a [35S]UTP labeled probe were dehydrated through 50%, 70%, and 90% ethanols containing 0.3 M ammonium acetate and vacuum-desiccated Sections
for
a
minimum of 2 hr. The slides
were
then coated with
a
layer of Kodak NTB-2 emulsion (Eastman Kodak Co., Rochester, NY). After exposure times varying from 7 to 28 days, the slides were developed with Kodak D-19 developer, fixed with Kodafix, and counterstained with hematoxylin-eosin. Detection
of digoxigenin labeled hybrids
Characterization of the GnRH and ßhCG mRNA
quantitation
assays
The amount of radioactive antisense probe protected from RNase digestion was linearly related to the known amounts of sense RNA with which it hybridized (Fig. 2A). The placental RNA samples measured 100-400 cpm in the GnRH mRNA assay and fell within the standard curve, above background. An autoradiogram of a typical polyacrylamide gel of a GnRH solution hybridization assay is shown in Fig. 2B. When hybridized with total placental RNA, the GnRH probe yielded two protected bands: one of 96 bp corresponding to the fully protected exon 3 as derived from mRNA and the other of 270 bp, corresponding to the GnRH primary transcript. Quantitation of both protected species in a single sample revealed that, on a molar basis, the primary transcript represents 16% of the mRNA. In a parallel study, a probe to exon 1 of GnRH was used yielding values for GnRH mRNA levels comparable to those obtained with the exon 3 probe when normalized to full-length mRNA (data not shown). The abundance of GnRH mRNA was two to three orders of magnitude less than that of /3hCG mRNA.
Sections hybridized to a digoxigenin-UTP-labeled probe immersed overnight in 2x SSC, 0.05% Triton X-100, and 2% normal sheep serum followed by a 10-min wash in 100 mM Tris-HCl pH 7.5, and 150 mM NaCl. The Levels of ßgCG and GnRH mRNA during sections were then incubated with a 1:500 dilution of polydevelopment placental clonal sheep anti-digoxigenin antibody conjugated to either alkaline phosphatase or rhodamine (Boehringer The mean levels ± SEM for each placenta, as well as for Mannheim Biochemicals, Indianapolis, IN) in 100 mM each trimester, are shown in Fig. 3. 0hCG mRNA levels Tris-HCl pH 7.5, 150 mM NaCl, 1% normal sheep serum were highest in the first trimester and decrease markedly and 0.3% Triton X-100 for 5 h at room temperature. (30-fold) by the third trimester. The large variation in firstSections incubated with the rhodamine conjugated anti- trimester levels of /3hCG mRNA could not be explained on body were washed for 10 min in 100 mM Tris-HCl pH 7.5 clinical grounds, since none of the patients had evidence of and 150 mM NaCl. Coverslips were affixed with an aque- multiple gestation (higher levels), molar pregnancy (higher ous mounting medium, and the slides were examined and levels), or failing pregnancy (lower levels). GnRH mRNA photographed with a Leitz Orthoplan microscope levels did show any difference from first to third trimester equipped with a band pass excitation filter in the 515- to (Fig. 3). 560-nm range and a barrier filter at 580 nm. Sections incubated with the alkaline phosphatase-conjuLocalization of GnRH and ßhCG mRNA in placenta gated antibody were reacted with nitroblue tetrazolium salt Both ßhCG and GnRH cRNA probes were used to local(75 mg/ml), 5-bromo-4-chloro-3-indolyl phosphate, toluidinium salt (50 mg/ml), and levamisole (0.24 mg/ml) in ize their respective mRNAs in first-trimester placental tis100 mM Tris-HCl pH 9.5, 100 mM NaCl, and 50 mM sue. As previously reported (Hoshina et al., 1982, 1983; MgCl2 for 16 hr. The reaction was stopped by immersing Wide et al, 1988, 1989), 0hCG mRNA was localized exthe sections in 10 mM Tris-HCl 8.0 and 1 mM EDTA. clusively in the syncytiotrophoblasts giving an appreciable Coverslips were affixed over the sections with an aqueous signal within 6 days of exposure [Fig. 4, hCG(-)]. In conmounting medium, and the slides were examined for the trast, GnRH mRNA is more evenly distributed throughout presence of a blue-brown precipitate and photographed the syncytiotrophoblasts, cytotrophoblasts, and stromal cells of the placenta [Fig. 4, GnRH(-)] when examined using visible light. with a radiolabeled probe. In agreement with a lower level of GnRH mRNA relative to ßhCG mRNA, a 14-day expoData analysis sure gave only a mild signal by in situ hybridization. Consections using sense strand ßhCG figure [Fig. 4, in trol from each placenta were analyzed multiple Samples solution hybridization assays for both /3hCG and GnRH hCG(+)] yielded only background levels of grains. Control sections using sense strand GnRH [Fig. 4, and the mean ± SE for each patient for /3hCG and GnRH mRNA levels was calculated. Differences across trimesters GnRH(+)] showed a level of silver grains two- to threein RNA levels were analyzed using one way analysis of var- fold greater than that seen with antisense probe, suggesting iance. The level of significance was set at 0.05. transcription of the opposite strand in this tissue or possiwere
PLACENTAL GnRH/hCH mRNA DURING PREGNANCY
415
pg RNA
B
Samples
GnRH standards
pg/lane 0
0.5 1.0 3.0 5.0 10
309-*
1
2
3
•Primary transcript (-280 bp)
242-*>
Protected
message
(96 bp) FIG. 2.
A. Typical hGnRH-7 standard curve generated from radioactivity present in protected reference RNA hybrids plotted against known quantities of reference RNA. r 0.998. B. Autoradiogram of polyacrylamide gel of GnRH solution hybridization/nuclease protection assay. The diffuse band seen at 96 bp in several of the GnRH standard lanes represents an unidentified artifact that was not present in all standards or all analyses. M, end-labeled Msp I-cut pBR322 as molecular weight marker. The GnRH standards run at approximately 300 bases, representing additional complementarity at the 5' and 3' ends due to plasmid vector elements; GnRH standard curve 0-10 pg/lane. Samples: 1, first trimester; 2, second timester; 3, third trimester. =
416
KELLY ET AL.
< rr
"rö
S
3 rr
E
I rr
ü Q.
