Characterization of Messenger Ribonucleic Acid Expression for ...

BIOLOGY OF REPRODUCTION 58, 849-856 (1998)

Characterization of Messenger Ribonucleic Acid Expression for Prostaglandin F2a and Luteinizing Hormone Receptors in Various Bovine Luteal Cell Types' Roni Mamluk, 3 Dong-bao Chen, 4 Yaffa Greber, 3 John S. Davis, 4 ,5 and Rina Meidan 2 ,3 Department of Animal Sciences,3 Faculty of Agriculture, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, Rehovot 76100, Israel Department of Obstetrics and Gynecology,4 Women's Research Institute, University of Kansas School of Medicine, Wichita, Kansas 67214 Veterans Administration Medical Center, 5 Wichita, Kansas 67218 ABSTRACT LH and prostaglandin F20 (PGF 20 ) control the life span and function of the corpus luteum (CL). Nevertheless, identification of the various cell types (steroidogenic and nonsteroidogenic) expressing the receptors for these hormones remains controversial. In this study we characterized LH and PGF 2,, receptor (r) expression in the various luteal cell types using quantitative reverse transcription-polymerase chain reaction. We found, in agreement with previously described functions of PGF 2,,, that the two steroidogenic cell types, as well as luteal endothelial cells, expressed PtGFr. In contrast, LHr was mainly expressed by small luteal cells. A similar pattern of PGFr and LHr expression was observed in steroidogenic cells luteinized in vitro and in cells derived from the mature CL. The expression of these two receptors was inversely affected by increased levels of cAMP (achieved by incubating cells with varying doses of forskolin); LHr expression was down-regulated by 50% in the presence of 10 ILM forskolin (p < 0.05), while an increase was observed in PGFr expression. In granulosa-derived luteal cells, maximal expression of PGFr was higher (approximately by 3-fold, p < 0.05) than in the theca-derived luteal cells. PGF 20, mimicking its in vivo effect, markedly down-regulated LHr expression in thecaderived luteal cells, abolishing expression at a concentration of 100 ng/ml. In summary, these studies depict cAMP and PGF 2,, as major regulators of PGFr and LHr expression in the two steroidogenic cell types. All three major cell types of the CL (steroidogenic and endothelial) express PGFr. LHr mRNA, on the other hand, was detected mainly in small luteal cells. Such broad cellular distribution of PGFr may highlight the significant role played by this prostaglandin in the bovine CL.

cells in the CL [6, 7]. Our recent studies had ascribed a mediatory role to endothelin-1 produced by these cells in PGF2,-induced luteal regression [8, 9]. LH receptors (LHr) of several species have been cloned, and their expression in granulosa and Leydig cells of rodents has been extensively studied [10]. However, in relatively few studies has the regulation of LHr expression during luteal development been examined. Our studies had suggested that in the cow, differences in LHr expression correlated with cellular responsiveness to LH [11]; nevertheless, findings on this issue remain inconclusive. In contrast to the LHr, relatively little is known about the PGF 2 . receptor (PGFr). Bovine, ovine, and rat PGFr cDNAs have recently been cloned and identified as members of the family of G protein-coupled receptors [12-14]. During luteal regression, or after in vivo administration of PGF 2 , LHr and PGFr within the CL have been reported to be markedly reduced [15, 16]. However, the cell types involved in this response have not yet been identified. In fact, the expression and actions of PGFr in the various cell types composing the bovine CL are still highly controversial. This study was undertaken to examine the regulation of LHr and PGFr expression in theca- and granulosa-derived luteal cells and to characterize their expression in luteal endothelial cells and in steroidogenic cell types. We have utilized cells luteinized in vitro and cells separated from the mature CL. MATERIALS AND METHODS Materials Dulbecco's Modified Eagle's medium (DMEM)/Ham's F-12 (1:1, v:v) nutrient mixture was from Gibco BRL Life Technologies (Gaithersburg, MD); penicillin, streptomycin, neomycin, and fetal calf serum (FCS) were from Biological Industries (Beit HaEmek, Israel); bovine LH (USDA bLHB-5) was kindly provided by the USDA Animal Hormone Program (Beltsville, MD); forskolin, insulin, and PGF 2. were from Sigma Chemical Company (St. Louis, MO); dNTPs, random hexamers oligodeoxynucleotides, and RNase inhibitor were from Promega (Madison, WI); Moloney murine leukemia virus StrataScript II reverse transcriptase was from Stratagene (La Jolla, CA); Taq DNA polymerase was from Farmentas (Vilnius, Lithuania); oligonucleotide primers were synthesized by Biotechnology General (Kiryat Weizmann, Rehovot, Israel).

INTRODUCTION The life span of the corpus luteum (CL) is balanced by luteotrophic and luteolytic stimuli; the primary luteotrophic hormone in cows, as in most species, is LH [1], whereas luteal regression is attributed to prostaglandin F2, (PGF 2,) of uterine origin [2, 3]. The cellular distribution and regulation of receptors for these hormones are expected to have a marked effect on ovarian function. LH, acting via adenylyl cyclase, triggers the differentiation of preovulatory follicular theca and granulosa cells into small and large luteal cells, respectively [4, 5]. In addition, the development of the CL is also characterized by extensive angiogenesis resulting in an elaborate network of capillaries [6]. It is estimated that endothelial cells lining these capillaries constitute approximately one half of all

Cell Cultures

Accepted November 4, 1997. Received July 8, 1997.

Granulosa and theca cells were isolated from healthy bovine preovulatory follicles as previously described [17, 18]. The experimental design consisted of long-term

'D-b.C. is a Lalor Foundation postdoctoral fellow. Correspondence. FAX: 8 9465763; e-mail: [email protected]

2

849

850

MAMLUK ET AL.

cultures in which granulosa and theca cells were cultured for 8 days with the treatments specified below. In the shortterm cultures, luteal cells were incubated for additional 24 h. Long-term cultures. Granulosa and theca cells were cultured for 8 days in 36-mm culture dishes in medium containing 1% FCS (basal medium) alone or supplemented with insulin (2 iKg/ml) and various concentrations of the adenylyl cyclase activator, forskolin (0.1, 1, or 10 p.M). On Day 8, cells were washed twice with ice-cold PBS, and total RNA was extracted. Short-term cultures. These studies were carried out only with cells cultured in the presence of 2 ig/ml insulin and 10 RIM forskolin, conditions previously shown to induce cell luteinization [17, 18]. On Day 8 of culture, cells were washed twice and incubated for 24 h in media containing the various treatments as specified below. At the end of the 24-h incubation period, cells were washed twice with icecold PBS and total RNA was extracted. Luteal Endothelial Cells Microvascular endothelial cells, type III, were kindly provided to us by K. Spanel-Borowski (Univeristy of Leipzig, Leipzig, Germany). These cells, isolated from bovine CL, had been thoroughly characterized and shown to maintain a stable phenotype in culture for at least 20 passages [19, 201. In our laboratory, these cells retain their cobblestone appearance and factor VIII staining for at least 20 passages (data not shown). Cells were grown in DMEM/F12 medium containing 5% FCS on plates coated with 1% collagen type I (Vitrogen 100; Collagen Corp., Palo Alto, CA). Cells utilized were from passages 4-10. Separation of Bovine Luteal Cells CL of early pregnancy were collected from a local slaughterhouse; they were immediately placed in Medium 199 containing 0.1% BSA, 25 mM Hepes, 100 U/ml penicillin, and 100 Rig/ml streptomycin (pH 7.4) (M-199) on ice and transported to the laboratory. The luteal tissue was sliced with a Saddie Riggs (Thomas Scientific, Swedesboro, NJ) apparatus. The sliced tissue (5-7 g) was placed in a spinner flask and washed twice with M-199. The medium was discarded, and the luteal tissue was incubated for 1 h at 35°C in 5 ml Medium 199 per gram of tissue containing collagenase (0.25%), with continuous stirring. After incubation, the tissue/enzyme mixture was triturated with a pasteur pipette, tissue clumps were allowed to settle, and the supernatant was transferred to 50-ml plastic tubes on ice. Dispersed cells were collected by centrifugation (100 x g, 5 min), washed twice with M-199, and then filtered through nylon membranes. Aliquots were taken for determination of cell number using a hemocytometer, and viability was determined by trypan blue exclusion. Luteal cell viability was greater than 95%. Elutriation Collagenase-dissociated cells (100 x 106) were resuspended and subjected to centrifugal elutriation using elutriation medium (Ca 2 +/Mg2+-free DMEM, pH 7.2, 25 mM Hepes, antibiotics, 0.1% BSA, and 0.02 mg/ml deoxyribonuclease) in a Beckman (Palo Alto, CA) J6B centrifuge. The luteal cells were injected into a Sanderson (Beckman) elutriation chamber, and the eluates were collected with continuous flow as follows. A 200-ml fraction (F-1) containing predominantly erythrocytes and endothelial cells

(< 10 ,um) and a variable degree of small luteal cells (SLC) was harvested using a flow rate of 16 ml/min at 1800 rpm. The following 200-ml fraction (F-2) containing predominantly SLC (10-20 Im) was harvested using a flow rate of 16 ml/min at 1400 rpm. A third fraction (F3) containing small cell clumps mixed with large cells was collected using a flow rate of 24 ml/min at 1200 rpm. The remaining fraction (F-4), a highly enriched large cell fraction (LLC; > 30 iM) containing less than 5% enucleated cells, was collected using a flow rate of 30 ml/min at 680 rpm. The yield and viability of cells in each fraction were determined immediately after elutriation. F-2 contained 81.6 + 2.78% SLC, 0.3 ± 0.22% LLC, and 15.7 ± 2.08% endothelial cells (mean + SEM, n = 3). F-4 contained 6.62% LLC, 21.6 + 5.01% SLC, and 18.1 ± 59.6 SEM, n = 3). Mixed 1.47% endothelial cells (mean luteal cells, counted prior to elutriation, were composed of SLC, 60.1 + 2.23%; LLC, 4.3 ± 0.22%; and endothelial cells, 35.2 _ 2.71% (mean ± SEM, n = 3). RNA was extracted from mixed luteal cells and from cells of F-2 and F-4 as indicated below. RNA Extraction and Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Total RNA was extracted from cells by the guanidinium thiocyanate method [21]. RNA was quantitated by computerized densitometry using a standard RNA sample. Total RNA (1 ,ug) was preheated for 5 min at 70 0C, immediately cooled on ice, and reverse transcribed for 2 h at 37°C in a 20-1l reaction mixture containing single-strength RT buffer (50 mM Tris-HCl, pH 8.3, 75 mM KCI, 3 mM MgCl 2, and 10 mM dithiotreitol), 1 mM of each dNTP, 100 pmol random hexamers oligodeoxynucleotides, and 50 U of RT. The RT reaction was terminated by heating for 5 min at 95C. The resulting cDNA templates were subjected to PCR amplification in a thermal cycler. Each cycle consisted of denaturation for 30 sec at 95°C and annealing for 30 sec at 58°C, and extension was carried out for 1 min at 72°C in a 25-,ul PCR mixture. The reaction mixture contained cDNA derived from the RT tube, 10 pmol of each oligonucleotide primer, 10 mM Tris-HCI (pH 9), 50 mM KCI, 2.5 mM MgC1 2, 0.1% Triton X-100, 0.1 mM of each dNTP, and 2 U Taq DNA polymerase. MgC1 2 concentration, dNTPs, oligonucleotide, and enzyme amounts were precalibrated to ensure maximal reaction efficiency. PCR products (20 1I of 25-pl total reaction volume) were electrophoresed on 2% agarose gels, stained with ethidium bromide, and photographed by using BioSystem (Fuji, Tokyo, Japan). Band intensities were analyzed by computerized densitometry using NIH Image version 1.6 equipped with macros developed by Thomas Seebacher, University of Konstanz, Germany. Oligonucleotide Primers PCR oligonucleotide primer pairs (20-22 nucleotides, 45-60% GC content) were designed based on the cDNA sequence of the various target genes. All primer pairs were designed to span at least one intron to avoid amplification of DNA contaminants. The glyceraldehyde 3-phosphate dehydrogenase (G3PDH) oligonucleotide primer pair (sense 5'-TGTTCCAGTATGATTCCACCC-3'; antisense 5'TCCACCACCCTGTTGCTGTA-3') were synthesized as described by Tsai et al. [22]. LHr primers were designed based on partial bovine cDNA sequence (GeneBank accession no. U87230) recently cloned in our laboratory (sense

PGFr AND LHr EXPRESSION IN LUTEAL CELLS

10000

1001

1000

B

I

* PGFR

7 6

£In

;1

851

10 10

5 4

-

.

3 2

I -1

100I

I

I

10

15

20

25

Amplification Cycles

30

G3PDH LHRPGFR

15 19

17 21

19 23

21 25

23 27

A.mplification Cycles

I

0

100

200

300

400

ng RNA

FIG. 1. Characterization of quantitative RT-PCR of G3PDH, PGFr, and LHr transcripts. Total RNA was extracted from luteinized theca cells, reverse transcribed, and amplified using a thermal cycler. PCR products were electrophoresed on 2% agarose gel, stained with ethidium bromide, and photographed. Inverse images and their densitometric analysis are presented in the upper and lower panels, respectively. A) G3PDH and LHr products of 100 ng RNA input were amplified for different numbers of cycles as indicated. B) G3PDH, PGFr, and LHr products of 100 ng total RNA input were amplified for different numbers of cycles as indicated. C) G3PDH, PGFr, and LHr products of increasing total RNA input were amplified for 19, 23 and 23 cycles, respectively. Data are representative of two similar experiments.

5'-CTCAAGCTTTCAGAGGACTT-3'; antisense 5'TCTGGAAGCTTGTGGATGCCTG-3'). Bovine PGFr primers (sense 5'-TTAGAAGTCAGCAGCACAG-3'; antisense 5'-ACTATCTGGGTGAGGGCTGATT-3') were synthesized according to the cDNA sequence cloned by Sakamoto et al. [12]. The expected PCR product lengths were 850 base pairs (bp) for G3PDH, 224 bp for LHr, and 521 bp for PGFr. To avoid amplification of prostaglandin E2 receptor (EP receptor), the reverse primer used to amplify PGFr was derived from a sequence that contains no homology to EP receptor. Computer searches, sequence alignments, and assembly were performed using software from Genetics Computer Group (Madison, WI). Calibration of Semiquantitative Multiple RT-PCR Semiquantitative RT-PCR was carried out using the housekeeping gene, G3PDH, as an internal standard [2224]. G3PDH is constitutively expressed in both granulosaderived luteal cells (LG) and theca-derived luteal cells (LT) and has been used effectively in studies on the regulation of gene expression in ovarian cells [16, 22]. The RT-PCR amplification of LT (100 ng total RNA) was first calibrated using G3PDH and LHr primer pairs (Fig. 1A). The G3PDH product was already in saturation from cycle 21, whereas the LHr product was in its exponential phase of amplification from cycle 21 to cycle 28. Therefore, G3PDH primers were added after four amplification cycles of LHr (primer dropping method; [25]). In addition to the 850-bp fragment of G3PDH, another band of lower molecular weight was amplified when the G3PDH was in saturation. Linear amplification of LHr, PGFr, and G3PDH signals was observed when G3PDH was amplified for 15-21 cycles, while 23 amplification cycles were used for LHr and PGFr (Fig. B). Amplification cycles (19 for G3PDH and 23 for LHr and PGFr) were within the expo-

nential range of the PCR employing 25-200 ng of RNA (Fig. 1C). We employed these conditions (100 ng RNA and 19 or 23 cycles) for the studies reported herein. Statistical Analysis All experiments were repeated 3-4 times. Data are presented as means + SEM. One-way analyses of variance were used to determine statistical difference of individual treatments as indicated in the text. A value of p < 0.05 was considered significant. RESULTS Characterization of LHr and PGFr Expression in the Various Luteal Cell Types The general pattern of both LHr and PGFr expression observed in steroidogenic cells luteinized in vitro or in cells separated from CL tissue was similar (Fig. 2), although expression levels were lower in cells luteinized in vitro. PGFr was expressed in both small and large luteal cells; in the latter, its expression was 2- to 3-fold higher than in small luteal cells. LHr expression was considerably higher in small than in large luteal cells. Large luteal cell preparations enriched from CL expressed small amounts of LHr; however, LHr was undetectable in cells obtained after in vitro luteinization, under similar assay conditions. This observation may suggest that a stronger down-regulation in LHr may have occurred in vitro than in vivo, or it may reflect the contamination of the large cell preparations with small luteal cells expressing high levels of LHr. Previously, we showed that PGF2 has a direct effect on the resident endothelial cells of the bovine CL [9], suggesting the presence of PGFr in these nonsteroidogenic cells. Indeed, data presented in Figure 3 confirm that luteal endothelial cells express PGFr. Using the sensitive RT-PCR

852

MAMLUK ET AL.

FIG. 2. Expression of PGFr and LHr mRNA in luteal cells. Small (SLC) and large (LLC) were enriched from pooled bovine CL as described in Materials and Methods. Mixed cells: nonseparated dispersed cells. Luteinized granulosa (LG) and theca (LT) cells were obtained after in vitro luteinization. Input of 100 ng total RNA of each sample was reverse transcribed and amplified for 19, 23, and 23 cycles (with G3PDH, PGFr, and LHr primers, respectively). PCR products were electrophoresed on 2% agarose gel, stained with ethidium bromide, and photographed. Upper panel: inverse image of RT-PCR; lower panel: densitometric analysis of PGFr and LHr relative to G3PDH expression.

20 E3 PGFR 0

'A

15

~1

10

C! ,

C

'I

5

o

LLC

LG

method, we found that in addition to its being expressed in endothelial cells and CL, PGFr mRNA was also expressed in the kidney but not in liver tissue. LHr mRNA could not be detected in endothelial cells even by increasing RNA input and amplification cycles. Expression of LHr and PGFr in Cells Undergoing Luteinization Activation of adenylyl cyclase is known to induce luteal cell phenotype. We therefore examined the effect of various concentrations of forskolin on LHr and PGFr mRNA expression (Fig. 4). In LG, LHr expression was detected when cells were incubated in medium containing insulin alone

SLC

LT

LLC

LG

SLC

LT

(Fig. 4A, upper panel) and was occasionally detected in LG cultured with 0.1 i M forskolin (not shown). Taken together, these results indicate that LG expressed low levels of LHr. In contrast, PGFr expression in LG was markedly upregulated by forskolin in a dose-dependent manner; 1 IM and 10 jIM forskolin significantly elevated PGFr expression in these cells (4.5- and 7-fold over the control value, respectively, Fig. 4A). LT expressed both LHr and PGFr (Fig. 4B), and the expression of these receptors was inversely affected by forskolin. LHr expression was down-regulated in the presence of 10 jIM forskolin (3-fold lower than in cells incubated with insulin alone, p < 0.05), while PGFr expression was up-regulated (by 2.5-fold) in cells cultured in the presence of 1 and 10 IM forskolin (p < 0.05). Maximal induction of PGFr expression by forskolin (in LG) was higher than in LT (7- and 2.5-fold above control levels, respectively). Regulation of PGFr and LHr in Luteal Cells

FIG. 3. Presence of PGFr mRNA in various tissues. -RT, without addition of reverse transcriptase; L, liver; K, kidney; EC, luteal endothelial cells. Samples from RT reactions of the various tissues were amplified for 30 cycles. A representative image is presented.

We next examined the maintenance of PGFr and LHr in luteal cells. As can be seen in Figure 5, PGFr expression in both luteal cell types was reduced by 50% or more when forskolin was withdrawn from the culture medium. LHr expression in LT (Fig. 5B) was up-regulated when forskolin was removed, whereas in LG it remained undetectable. These findings concur with data from the previous experiments indicating inverse effects of elevated cAMP levels on LHr and PGFr expression. Insulin (with or without forskolin) did not appreciably influence PGFr expression in either cell type (Fig. 5). The in vivo administration of PGF 2 is known to reduce LHr and PGFr [13, 14, 16]. To examine whether this is a


800-

300 *p

°= ,

600-

250

- _ ca -

m

400-

· p OL

-

200

o U < o

*

150 l

100-

200-

E

C

.

0_

up

-

xo

< 0

E

853

50 W

l

Forskolin Insulin

-

-

+

0.1 +

A II

{

{

|

1 +

10 +

Forskolin Insulin

I

I

I

-

-

0.1

I

1

I

10

-+ + + FIG. 4. Effect of forskolin on the expression of mRNA for PGFr and LHr in luteal granulosa (LG) and theca (LT) cells. Cells were incubated for 8 days in basal medium (control) or in basal medium containing insulin alone (2 pxg/ml) or insulin and increasing concentrations of forskolin (0-10 p.M). RNA was extracted, and 100 ng total RNA of each sample was reverse transcribed and amplified for 19, 23, and 23 cycles (with G3PDH, PGFr, and LHr primers, respectively). PCR products were electrophoresed on 2% agarose gel, stained with ethidium bromide, and photographed. Upper panel: inverse image of representative RT-PCR; lower panel: densitometric analysis of PGFr and LHr relative to G3PDH expression. Data (mean SEM) are from three separate experiments. Asterisk indicates a statistically (p < 0.05) significant difference from control (basal media).

A.

B. LG

LT

200

200

150

S

. .

100

lcn Iu

6 1; 100

i: l Eg 50

50

0

Forskolin (10p1M) Insulin (2pg/ml)

+ +

+ +

-

n v Forskolin (101pM) Insulin (2gg/ml)

+ +

+ +

FIG. 5. Maintenance of PGFr and LHr mRNA expression in luteal cells. On Day 8 of culture, luteal granulosa (LG; A) and theca (LT; B) cells were washed and incubated for 24 h with various treatments as specified in the figure. RT-PCR was performed as described in the legend to Figure 4. Data (mean SEM) are derived from densitometric analysis of four separate experiments and are depicted as percentage of initial expression (expression in the presence of insulin 2 g/ml and forskolin 10 IM). Asterisk indicates a statistically (p < 0.05) significant difference from initial expression.

854

MAMLUK ET AL.

A.

B. LG

LT

c

.9

0 Ma

U,

>

5a

~ Y i

,

¢ r W.

¢

W5

3W D

0

10 100 PGF2(ng/ml)

1000

0

10

100

a

1000

PGF2,(ng/ml)

FIG. 6. Effect of PGF2, on PGFr and LHr mRNA expression. On Day 8 of culture, luteal granulosa (LG; A) and theca (LT; B) cells were washed and incubated for 24 h without PGF2 . (0) or with various concentrations of PGF 2,. RT-PCR was performed as described in the legend to Figure 4. Data (mean SEM) are derived from densitometric analysis of three separate experiments and are depicted as arbitrary units of PGFr and LHr relative to G3PDH expression. Asterisk indicates a statistically (p < 0.05) significant difference from control.

direct effect that can be reproduced in vitro, PGF 2. was added to the incubation medium of luteal steroidogenic cells (Fig. 6). In LG expressing high levels of PGFr mRNA (see also Fig. 2), PGF 2, induced a dose-dependent decrease in the expression of its receptor (Fig. 6A, p < 0.05). In LT, PGF2 . down-regulated both LHr and PGFr expression, and a concentration of 100 ng/ml was sufficient to significantly reduce receptor expression (Fig. 6B, p < 0.05). The PGF 2,-induced reduction of LHr expression appeared to be more pronounced than the reduction of PGFr expression (Fig. 6B). DISCUSSION In this study we demonstrate that 1) the PGFr was expressed by all three major cell populations of the bovine CL, i.e., the two steroidogenic cell types as well as luteal endothelial cells, while the LHr was mainly expressed by the small luteal cells. The pattern of PGFr and LHr expression observed in cells luteinized in vitro or in cells separated from the mature bovine CL was similar. 2) PGFr expression in the steroidogenic cells was positively correlated with forskolin (and possibly cAMP) levels. 3) Incubation of steroidogenic luteal cells with PGF 2. reduced LHr and PGFr mRNA. The LH surge alters the pattern of gene expression in ovarian cells. Most notable are changes in expression of various steroidogenic enzymes and hormone receptors [1, 5]. In whole ovarian tissue, LH/hCG administration has been shown to transiently down-regulate LHr expression and hormone binding in the postovulatory follicle, with recovery occurring 36-48 h later in the CL [26, 27]. However, whether this represents coordinated changes in LHr expression in the two steroidogenic cell types is unclear. After an 8-day incubation with forskolin, we observed that LT, but not LG, expressed LHr. Similar data were obtained when cells were only briefly (12 h) exposed to high levels of LH [11]. Likewise, in the porcine ovary, LHr expression was detected only in theca-derived luteal cells [28]. In LT, as reported for Leydig cells also [29], levels of LHr expression in vitro were inversely related to the concentra-

tions of cAMP. Interestingly, LHr expression in LG did not recover upon withdrawal of a cAMP-elevating agent (forskolin) in vivo or in vitro (this study and [28]), suggesting different transcriptional regulatory mechanisms in the two luteal cell types. The low LHr mRNA levels found in granulosa-derived large luteal cells could explain their low LH responsiveness, as only superphysiological levels (> 1 ig/ ml) are capable of eliciting a response [30]. Regression of the CL, initiated by PGF 2 . [3], is essential for normal cyclicity, as it enables the development of a new ovulatory follicle. However, despite the key physiological role of PGF 2 , details of its mechanism of action and even its target cells within the CL have not yet been clearly identified. In the cow, both small and large luteal cells are responsive to PGF2,; these responses include changes in progesterone production, cAMP production, Ca2 + ion effluxes, and inositol phosphate formation [18, 30-32]. A study by Chegini et al. [33] indicated that both types of steroidogenic luteal cells bind PGF2 ., while in a more recent study, Sakamoto et al. [12] detected PGFr expression mainly in the large luteal cells. Results obtained by in situ hybridization [12] are qualitative and may not accurately detect lower levels of mRNA, as demonstrated for PGFr expression in small luteal cells in our study. The findings described in this paper are consistent with results reported in the functional studies mentioned above, showing that both bovine small and large luteal cells express PGFr. Different results have been reported in the sheep, where PGF 2. function and binding sites were shown to be localized to large luteal cells [34]. The reason for this difference in such evolutionarily close species is unclear. Elevated cAMP levels dose-dependently up-regulated PGFr mRNA expression in luteal cells (in vivo and in vitro). Neither LG nor LT acquired PGFr unless a cAMPelevating agent was added to the culture medium (this study and [22]); moreover, following forskolin withdrawal, a reduction in PGFr levels was observed. These findings also concur with those reported for human luteinized granulosa cells [35]. PGF 2,-induced luteal regression is accompanied by re-


duction in LHr and PGFr expression in the CL [13, 14, 16]. Since these reductions are evident only after the decline in progesterone, PGF2 ,,-induced regression is considered to be part of structural luteolysis. Our findings suggest that the reduction of LHr and PGFr mRNA is due to a direct effect of PGF2a on both types of steroidogenic luteal cells. Our previous studies had shown that PGF2, stimulated endothelin-1 production in CL in vivo; by using luteal endothelial cells in vitro, we now extend this observation to show that endothelial cells express PGFr mRNA. This indicates that the abrupt rise in endothelin-1 observed during late luteal phase or after PGF2 administration is at least partially attributable to the direct action of PGF 2a on endothelial cells [8, 9]. Together, these findings lend support to the physiological relevance of endothelial cells and endothelin-1 as mediators of PGF2 ,-induced luteolysis. PGF2. is a vasoactive compound that acts on nonsteroidogenic cells also; one such well-documented effect is the contraction of various smooth muscle cells [36]. Recently, specific PGFr have been found in endothelial cells of the jugular vein in rabbits [37]. It will be critical to determine the mechanism of action of PGF2 in endothelial cells and its relation to the control of luteal function. Using in situ hybridization and immunohistochemistry, LHr was detected in endothelial cells of uterine arteries [38]. In our hands, using RT-PCR, LHr mRNA expression could not be detected in type III microvascular endothelial cells of bovine CL, or in endothelial cells of large vessels such as the bovine aorta (unpublished results). The expression of both LHr and PGFr in cells luteinized in vitro was highly comparable to that of cells separated from the CL. These findings further validate the use of in vitro-luteinized granulosa and theca cells as a model for bovine large and small luteal cells [17, 18, 39]. Nevertheless, mRNA expression levels were lower in cultured cells relative to those of CL-derived cells, suggesting the contribution of extracellular matrix and intercellular communication in luteal cell function. The broad cellular distribution of PGFr expression (which is in sharp contrast to that of LHr expression) suggests that this prostaglandin may also mediate effects other than luteolysis. It is worth noting that cells exposed to elevated cAMP levels, which are therefore expected to produce more progesterone, also express higher levels of PGFr. Also consistent with this notion are the findings that PGF 2a [40] and its receptor [22] are induced by the ovulatory surge of LH and that highest concentrations of PGF2 , are found during early luteal phase [41]. While the LH surge (and cAMP) elevates PGF2 , and PGFr, it down-regulated its own receptor, suggesting perhaps that the primary luteotrophic role of LH is confined to the periovulatory phase while that of PGF 2. emanates in early luteal phase. The involvement of PGF 2a in these functions awaits further investigation. REFERENCES 1. Hansel W, Blair RM. Bovine corpus luteum: a historic overview and implications for future research. Theriogenology 1996; 45:1267-1294. 2. McCracken JA, Carlson JC, Glew ME, Goding JR, Baird DT, Green K, Samuelsson B. Prostaglandin F2 identified as a luteolytic hormone in sheep. Nature 1972; 238:129-134. 3. Milvae RA, Hinckley ST, Carlson JC. Luteotropic and luteolytic mechanisms in the bovine corpus luteum. Theriogenology 1996; 45: 1327-1349. 4. Alila HW, Hansel W. Origin of different cell types in the bovine corpus luteum as characterized by specific monoclonal antibodies. Biol Reprod 1984; 31:1015-1025.

855

5. Smith ME Mechanisms associated with corpus luteum development. J Anim Sci 1994; 72:1857-1872. 6. Zheng J, Redmer DA, Reynolds LR. Vascular development and heparin-binding growth factors in the bovine corpus luteum at several stages of the estrous cycle. Biol Reprod 1993; 49:1177-1189. 7. O'Shea JD, Rodgers RJ, D'Occhio MJ. Cellular composition of cyclic corpus luteum of the cow. J Reprod Fertil 1989; 85:483-487. 8. Girsh E, Milvae RA, Wang W, Meidan R. Effect of endothelin-l on bovine luteal cell function: role in PGF2, induced anti-steroidogenic action. Endocrinology 1996; 137:1306-1312. 9. Girsh E, Wang W, Mamluk R, Arditti E Friedman A, Milvae RA, Meidan R. Regulation of endothelin- I expression in the bovine corpus luteum: elevation by prostaglandin F2,. Endocrinology 1996; 137: 5192-5241. 10. Segaloff DL, Ascoli M. The lutropin/choriogonadotropin receptor... 4 years later. Endocr Rev 1993; 14:324-347. 11. Mamluk R, Wolfenson D, Meidan R. Expression of LH receptors and cytochrome P450 side chain cleavage in bovine luteal cells luteinized by LH or forskolin. Domest Anim Endocrinol 1997; (in press). 12. Sakamoto K, Miwa K, Ezashi T, Okuda-Ashitaka E, Okuda K, Houtani T, Sugimoto T, Ito S, Hayaishi O. Expression of mRNA encoding the prostaglandin F2. receptor in bovine corpora lutea throughout the estrous cycle and pregnancy. J Reprod Fertil 1995; 103:99-105. 13. Graves PE, Pierce KL, Bailey TJ, Rueda BR, Gil DW, Woodward DE Yool AJ, Hoyer PB, Regan JW. Cloning of a receptor for prostaglandin F2 . from the ovine corpus luteum. Endocrinology 1995; 136: 3430-3436. 14. Olofsson JI, Leung CHB, Bjurulf E, Ohno T, Selstam G, Peng C, Leung PCK. Characterization and regulation of a mRNA encoding the prostaglandin F2, receptor in the rat ovary. Mol Cell Endocrinol 1996; 123:45-52. 15. Rueda BR, Botros IW, Pierce KL, Reagan JW, Hoyer PB. Comparison of the mRNA levels for the PGF2a receptor (FP) during luteolysis and early pregnancy in the ovine corpus luteum. Endocrine 1995; 3:781787. 16. Smith GW, Gentry PC, Roberts RM, Smith ME Ontogeny and regulation of luteinizing hormone receptor messenger ribonucleic acid within the ovine corpus luteum. Biol Reprod 1996; 54:76-83. 17. Meidan R, Blum O, Girsh E, Aberdam E. In vitro differentiation of bovine granulosa and theca cells into large and small luteal like cells: morphological and functional characteristics. Biol Reprod 1990; 43: 913-921. 18. Meidan R, Aberdam E, Afialo L. Steroidogenic enzyme content and progesterone induction by cAMP-generating agents and prostaglandin F2 a in bovine theca and granulosa cells luteinized in vitro. Biol Reprod 1991; 46:786-792. 19. Mayerhofer A, Spanel-Borowski K, Watkins S, Gratzl M. Cultured microvascular endothelial cells derived from the bovine corpus luteum possess NCAM-140. Exp Cell Res 1992; 201:545-548. 20. Spanel-Borowski K, Fenyves A. The heteromorphology of cultured microvascular endothelial cells. Arzneimittelfoschung 1994; 44:385391. 21. Chomoczynsky P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162:156-159. 22. Tsai SJ, Wiltbank MC, Bodensteiner KJ. Distinct mechanisms regulate induction of messenger ribonucleic acid for prostaglandin (PG) G/H synthase-2, PGE (EP-3) receptor, and PGF2, receptor in bovine preovulatory follicles. Endocrinology 1996; 137:3348-3355. 23. Orly Y, Rei Z, Greenberg NM, Richards JS. Tyrosine kinase inhibitor AG18 arrests follicle-stimulating hormone-induced granulosa cell differentiation: use of reverse transcriptase-polymerase chain reaction assay for multiple messenger ribonucleic acids. Endocrinology 1994; 134:2336-2346. 24. Takamiyagi A, Nakashima Y, Irifune H, Uezato H, Nonaka S. Quantitative analysis of ferrochelatase mRNA in blood cells of erythropoietic protoporphyria patients. J Dermatol Sci 1996; 11:154-160. 25. Wong H, Anderson WD, Cheng T, Riabowol KT Monitoring mRNA expression by polymerase chain reaction: the "primer-dropping" method. Anal Biochem 1994; 223:251-258. 26. LaPolt PS, Jia XC, Sincich C, Hsueh AJW. Ligand-induced downregulation of testicular and ovarian luteinizing hormone (LH) receptors is preceded by tissue-specific inhibition of alternatively processed LH receptor transcripts. Mol Endocrinol 1991; 5:397-403. 27. Hu ZZ, Tasi-Morris CH, Buczko E, Dufau ML. Hormonal regulation of LH receptor mRNA and expression in the rat ovary. FEBS Lett 1990; 274:181-184.

856

MAMLUK ET AL.

28. Meduri G, Vu Hai MT, Jolivet A, Takemori S, Kominami S, Driancourt MA, Milgrom E. Comparison of cellular distribution of LH receptors and steroidogenic enzymes in the porcine ovary. J Endocrinol 1996; 148:435-446. 29. Chuzel E Schteingart H, Vigier M, Avallet O, Saez JM. Transcriptional and post-transcriptional regulation of luteotropin/chorionic gonadotropin receptor by the agonist in Leydig cells. Eur J Biochem 1995; 229:316-325. 30. Alila HW, Dowd JP, Corradino RA, Hanis V, Hansel W. Control of progesterone production in small and large bovine luteal cells separated by flow cytometry. J Reprod Fertil 1988; 82:645-655. 31. Alila HW, Corradino RA, Hansel W. Differential effects of luteinizing hormone on intracellular free Ca 2+ in small and large luteal cells. Endocrinology 1989; 124:2314-2320. 32. Davis JS, Alila HW, West LA, Corradino RA, Weakland LL, Hansel W. Second messenger systems and progesterone secretion in the small cells of the bovine corpus luteum: effects of gonadotropins and prostaglandin F2 a. J Steroid Biochem 1989; 32:643-649. 33. Chegini N, Lei ZM, Rao CV, Hansel W. Cellular distribution and cycle phase dependency of gonadotropin and eicosanoid binding sites in bovine corpora lutea. Biol Reprod 1991; 45:506-513. 34. Wiltbank MC, Diskin MG, Niswender GD. Differential actions of second messenger systems in the corpus luteum. J Reprod Fertil Suppl 1991; 43:65-75.

35. Ristimaki A, Jaatinen R, Ritvos O. Regulation of prostaglandin F2Q receptor expression in cultured human granulosa-luteal cells. Endocrinology 1997; 138:191-195. 36. Molnar M, Hertelendy E Signal transduction in rat myometrial cells: comparison of the actions of endothelin 1, oxytocin and prostaglandin F2 .. Eur J Endocrinol 1995; 133:467-474. 37. Chen J, Champa-Rodriguez ML, Woodward DE Identification of a prostanoid FP receptor population producing endothelium-dependent vasorelaxation in the rabbit jugular vein. Br J Pharmacol 1995; 116: 3035-3041. 38. Toth P, Li X, Rao CV, Lincoln SR, Sanfilippo JS, Spinnato II JA, Yussman MA. Expression of functional human chorionic gonadotropin/human luteinizing hormone receptor gene in human uterine arteries. J Clin Endocrinol & Metab 1994; 79:307-315. 39. Aflalo L, Meidan R. The hormonal regulation of side chain cleavage cytochrome P450, adrenodoxin and their messenger ribonucleic acid expression in bovine small-like and large-like luteal cells: relationship with progesterone production. Endocrinology 1993; 132:410-416. 40. Siriois J. Induction of prostaglandin endoperoxide synthase-2 by human chorionic gonadotropin in bovine preovulatory follicles in vivo. Endocrinology 1994; 135:841-848. 41. Milvae RA, Hansel W. Prostacyclin, prostaglandin F2. and progesterone production by bovine luteal cells during the estrous cycle. Biol Reprod 1983; 29:1063-1068.