MOLECULAR REPRODUCTION AND DEVELOPMENT 74:1312–1316 (2007)

Distinct Responses of Human GranulosaLutein Cells After hCG or LH Stimulation in aSpheroidal Cell Culture SystemJULIA BECKER,1 ANDREA WALZ,1 STEFANIE DAUBE,1 CHRISTOPH KECK,1,2

AND DETLEF PIETROWSKI1,3*1Department of Obstetrics and Gynecology, University of Freiburg Medical School, Freiburg, Germany2PAN-Klinik am Neumarkt, Cologne, Germany3Heart-Institute, Lahr/Baden, Germany

ABSTRACT The growth and development ofthe corpus luteum (CL) is regulated by gonadotropichormones. It is formed by granulosa cells (GCs), thecacells, and endothelial cells, and is the primary source ofcirculating progesterone. During early pregnancy onlyhuman chorionic gonadotropin (hCG) but not luteiniz-ing hormone (LH) extends the life span of the CL,although hCG and LH interact with the same receptorand have similar actions on the CL. In this study arecently by our group established spheroidal GC cultureassay served as a model of CL development on whichwe compared the actions of the gonadotropic hor-mones LH and hCG. To find out which signal pathwaystake part in the hormonal regulation of GC westimulated GC-spheroids with modulators of proteinkinases A and C dependent signaling cascades anddetermined their impact on sprout forming activity inGC. Our results indicate that PKA-dependent signalingpathways play a major role in mediating the hormonal-induced signaling cascades leading to sprouting in GC.Furthermore, this study strongly indicates that thedifferent effects of hCG and LH in the maintenance ofthe CL may be reasoned in different signal transductionpathways triggered by hCG or LH. Mol. Reprod. Dev.74: 1312–1316, 2007. � 2007 Wiley-Liss, Inc.

Key Words: corpus luteum; granulosa cells; spher-oid; PKA; PKC; hCG; LH

INTRODUCTION

The human corpus luteum (CL) develops from theruptured follicle after ovulation. It is formed bygranulosa cells (GCs) and theca cells which remainin the follicle cavity and by invading endothelial cells(EC). The development of the CL is controlled by thegonadotropins luteinizing hormone (LH) and follicle-stimulating hormone (FSH). Non-fertile cycles arecharacterized by a regression of the CL after 8–14 daystriggered by lower LH-levels and diminished respon-siveness of GC to LH. In the case of conception the CLremains functional up to 12 weeks and becomes the main

source of progesterone necessary for the establishmentand maintenance of pregnancy. Choriogonadotropin(hCG) which is solely secreted by the growing tropho-blast during early pregnancy is known to be responsiblefor the maintenance of the CL. There have been manystudies about the molecular actions involving hormonebinding and receptor activation as well as the followingactivation of G-Proteins in GC (Bourdage et al., 1984;Dufau, 1998; Ascoli et al., 2002; Bukovsky et al., 2003). Itturned out that although LH and hCG bind and activatethe same G-protein coupled hormone receptor they cancause distinct responses in their target cells. It wasdescribed that there are different binding sites for hCGand LH which might result in an activation of differentprotein kinases following signal transduction via Gs, Gi,or Gq-proteins (Gudermann et al., 1992; Gilchrist et al.,1996; Herrlich et al., 1996; Rajagopalan-Gupta et al.,1999; Muller et al., 2003). The most important signalpathway activated by hCG and LH is the protein kinaseA (PKA) pathway controlling such crucial events as theprogesterone production in the CL (Gilchrist et al., 1996;Ryu et al., 1998; Ji et al., 1997). Another kinase activatedby hCG or LH stimulation of GCs is protein kinase C(PKC). In contrast to the PKA signal only few data existson the hormone-dependent action of PKC in luteiniza-tion (Murphy, 2000).

In the present work the role of PKA and PKCpathways are studied by stimulating GC with a varietyof modulators of PKA and PKC (Davies et al., 2000;Bain et al., 2003). To further investigate the hormone-dependent activation of these two kinases we used athree-dimensional spheroidal culture assay and ana-lyzed the effects of gonadotropic hormones LH and hCGon the sprout forming activity of GC-spheroids.

� 2007 WILEY-LISS, INC.

*Correspondence to: Detlef Pietrowski, PhD, Universitats-FrauenklinikFreiburg, Hugstetter Str. 55, D-79106 Freiburg; Herzentrum Lahr/Baden, Hohbergweg 2, D-77933 Lahr.E-mail: detlef.pietrowski@heart-lahr.com

Received 12 September 2006; Accepted 15 November 2006Published online 8 February 2007 in Wiley InterScience(www.interscience.wiley.com).DOI 10.1002/mrd.20696

MATERIALS AND METHODS

Cell Culture

Endothelial cell basal medium (ECBM) was pur-chased from PromoCell (Heidelberg, Germany). Fetalcalf serum (FCS) was obtained from Gibco, (Paisley,Scotland).

Human GCs were isolated from follicular fluid ofpatients undergoing IVF therapy after written consentof the patients. They were cultured in Medium 199(Gibco, Paisley, Scotland) containing 10% heat-inacti-vated FCS.

Spheroidal Sprouting-Assay

GC were trypsinized after 48 hr of incubation. 160,000GC were suspended in 40 ml ECBM (PromoCell) andmethocel (4:1). The methocel was produced by dissolving6 g carboxymethylcellulose in 500 ml ECBM. Only theclear supernant was used after centrifugation. In orderto generate spheroids, 400 cells were suspended in eachwell of non-adherent round-bottom 96-well plates(Greiner, Frickenhausen, Germany) and cultured for24 hr at 378C with 5% CO2 and 100% humidity. After24 hr of cultivation, the spheroids were harvested.Methocel (4.5 ml) and FCS (4:1) were put onto 400spheroids and mixed with 4.5 ml of a 48C collagen gelcontaining 4.0 ml acidic collagen type 1, 0.5 ml Medium199 10� (Sigma-Aldrich, Steinheim, Germany) andapproximately 0.5 ml 0.2 N NaOH to adjust the pHto 7.4.

One milliliter of the prepared mixture containing 50spheroids was given into each of the eight prewarmed(378C) middle wells of a 24-well plate. The gels wereallowed to polymerize for 30 min at 378C with 5% CO2

and 100% humidity after which the corresponding testsubstance in ECBM (100 ml/ml) was pipetted ontothe gels. After 24 hr, digital quantification of sproutformation was carried out by measuring the cumulativelength of the sprouts (CSL) deriving from one spheroid.The digital photographs (Camera DP50) were takenwith the help of an inverse microscope (Olympus IX50)and evaluated using the digital imaging softwareanalySIS 3.1 (Olympus, Germany). Measuring wasrepeated for at least 30 different spheroids per experi-mental group and experiment.

Hormones, Reagents, and Concentrations Used

Recombinant human chorionic gonadotropin (10 U/ml) and LH (35 mU/ml) were obtained from Serono(Unterschleissheim, Germany). For the modulation ofPKA and PKC 8-Bromo-cAMP (100 mM), H-89 (500 nM),DHI (10 mM), GO 6976 (7.9 nM), and GO 6983 (10 mM)were purchased from Calbiochem/Merck (Schwalbach,Germany). These substances were either dissolved inwater or in DMSO. The final concentration of DMSOused to stimulate the spheroids did not exceed 0.01%.Control groups were incubated with 100 ml ECBM or10 ml DMSO in 100 ml ECBM to lock out effects causedby DMSO. As there were no significant differences inthe sprouting activity between the DMSO control and

the ECBM control (Fig. 4), ECBM-treated spheroidswere used as negative control in all experiments.

Statistics

The results of this work have been analyzed usingANOVA with paired comparisons. The graphs showmean�SEM. Differences were considered statisticallysignificant at P<0.05.

RESULTS

There have been several studies which demonstratethat PKA and PKC are the main signal pathwaystriggering the hormone signal of hCG and LH in GC(Gudermann et al., 1992; Herrlich et al., 1996; Ryu et al.,1998). The aim of this study was to survey the role ofPKA and PKC on the cellular sprout formation activityof GCs in dependency of gonadotropic hormones. Weused a recently established three-dimensional spheroi-dal sprouting assay as a model of CL developmentin vitro in order to investigate the actions of gonado-tropic hormones and modulators of PKA and PKC in acell culture system. As demonstrated by Walz et al.(2005) and confirmed in this experiments GC reveal ahigh sprout forming activity when they are not stimu-lated with hCG or LH (Fig. 2a, bar A). In contrasttreatment of GCs with hCG or LH reduced their sproutforming activity twofold (Fig. 2a, bar C and bar E; andalso Fig. 1a,b).

To investigate if sprout formation in GCs is regulatedvia a PKA dependent pathway we used 8-Bromo-cAMPfor activation of PKA. As shown in Figure 2a thesprouting activity of spheroids stimulated with 8-Bromo-cAMP (bar B) was inhibited to a level whichwas similar to that reached by hCG or LH. The CSL ofhCG-stimulated spheroids was decreased when the cellswere additionally treated with 8-Bromo-cAMP. How-ever, the CSL of hCG- and LH-treated cells were notsignificantly decreased suggesting that in hormone-treated cells which show a minimum of sprout formingactivity the CSL could not be further reduced byactivation of PKA. After having shown that activationof PKA decreases sprouting in GCs comparable with theeffect reached by treatment of the cells with hCG alonewe were interested to know if the inhibition of PKApathways led to an increment of sprouting in our model.

For these experiments we used H-89, a well-estab-lished inhibitor of PKA-dependent pathways (Davieset al., 2000). As shown in Figure 2b, the incubation ofGC-spheroids with H-89 enhanced only the CSL of hCG-treated cells but not the CSL of LH treated cells,indicating that the hormonal-induced repression ofsprouting via LH or hCG is postponed by a differentsignaling mechanism.

Data by other authors have suggested a role for PKCin the regulation of growth and development of GCs(Herrlich et al., 1992; Gilchrist et al., 1996; Ji et al.,1997). We therefore were interested if also PKC-dependent pathways may play a role in the regulationof sprouting in GCs. We used 6-[N-decylamino]-4-

Molecular Reproduction and Development. DOI 10.1002/mrd

GONADOTROPIN RESPONSE IN HUMAN GRANULOSA CELLS 1313

hydroxymethylinole (DHI) as a categorical PKC activa-tor for stimulation of PKC-dependent pathways. Asshown in Figure 3a, this experiment demonstrates analmost threefold decrease of the CSL compared to thecontrol group (compare bar A and bar B). In contrast,there was only a slight, but not significant inhibitionwhen spheroids were treated with either hormonetogether with DHI (compare bar C, bar D, and bar F).This suggests that PKC is involved in sprout formationof GC, but PKC activation alone was not sufficient toabrogate the hormone-induced sprout reduction in GCas was demonstrated for the PKA-dependent pathways.

The family of to date known PKC consists of variousisoenzymes which are differentially involved in cellularprocesses. In order to further investigate PKC subtypedependent sprout formation we used the PKC selectiveinhibitors GO 6983 which selectively inhibits the PKC-isoenzymes PKCa, PKCb, PKCg, and PKCd (Gschwendtet al., 1996) and GO 6976 which inhibits the PKC-isoenzymes PKCa und PKCb1 (Martiny-Baron et al.,1993).

In Figure 3b,c it is shown that neither GO 6976 nor GO6983 affected untreated or hormone-treated GCs whichsupports the assumption that PKC plays a minor role incontrolling sprout formation in GC.

DISCUSSION

In the present study we investigated the sproutmodulating effects of PKA- and PKC-dependent signalpathways on GC-spheroids.

Until now there has been no final explanation howhCG in contrast to LH is able to rescue the CL duringearly pregnancy in humans and primates (Zeleznik,1998). Therefore attention is led in this work towardsthe effect of the gonadotropins and their signal trans-duction pathways in GC, especially on possible differ-ences between the effects of hCG and LH.

Molecular Reproduction and Development. DOI 10.1002/mrd

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 174,21873304,33759304,89059381,6813275,44276702,56696

Control8-Br-cAMP

100µMhCG

10U/mlhCG+8-Br-

cAMPLH

35mU/mlLH+8-Br-

cAMP

FEDCBA

**

*

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 248,61022211,8506570,44347377,14378638,54898624,10853

ControlH-89

500nMhCG

10U/mlhCG+H-

89LH

35mU/mlLH+H-89

FEDCBA

**

*

a

b

Fig. 2. Three-dimensional sprouting assay with GC-spheroidsembedded in collagen gel and stimulated with modulators of PKApathway. The diagram shows the means�SEM of the cumulativesprout lengths (CSL). At least three independent experiments wereperformed measuring 10 spheroids per experimental group. * Indicatessignificant differences of P<0.05 between the groups. a: GC-spheroidswithout stimulation (bar A), and after stimulation with 100 mM PKA-activator 8-Bromo-cAMP (bar B), 10 U/ml hCG (bar C) and 35 mU/mlLH (bar E), as well as hCG and 8-Bromo-cAMP (bar D), respectivelyLH and 8-Bromo-cAMP (bar F) with the same concentrations of thesubstances. b: GC-spheroids without stimulation (bar A), and afterstimulation with 500 nM PKA-inhibitor H-89 (bar B), 10 U/ml hCG(bar C), and 35 mU/ml LH (bar E), as well as hCG and H-89 (bar D),respectively LH and H-89 (bar F) with the same concentrations of thesubstances. [See color version online at www.interscience.wiley.com.]

Fig. 1. a: GC-spheroids were embedded in collagen gel and treated with ECBM only (control) (a) or 10 U/ml hCG (b). The high density of long sprouts indicates a high sprouting activity of the control (a) comparedto only a few sprouts and therefore a low sprouting activity after hCG treatment (b). [See color versiononline at www.interscience.wiley.com.]

1314 J. BECKER ET AL.

We demonstrated that in GC the activation of PKA-dependent signaling cascades inhibited sprouting sig-nificantly whereas the inhibition of PKA stimulatedsprouting. These findings indicate that the formation

of sprouts in GC-spheroids is regulated by changes inthe intracellular cAMP-concentration in which risingcAMP-levels reduce sprouting. Comparable observa-tions were made by different authors, who describe thatgonadotropin and other substances, which increaseintracellular cAMP, cause downregulation of the actin-cytoskeleton in GC (Ben Ze’ev and Amsterdam, 1987;Baum et al., 1990). The finding that the PKA-inhibitorH-89 did not increase the CSL compared to the controlcan be explained by an already preexisting maximalactivation of unstimulated spheroids, so that a furtherenhancement is not possible.

Interestingly, 8-Bromo-cAMP, the PKA-activator, ledto a further increase of the gonadotropic effect whereasH-89 antagonized only the effect of hCG. We concludedthat hCG performs its inhibiting effect on the sproutformation of GC-spheroids via cAMP/PKA while LHseems to use other signal pathways besides PKA toinhibit sprouting in GC. Although it is well known thatthe PKA signal pathway is dominant in the gonado-tropin-dependent regulation of steroidogenesis (Nis-wender et al., 2000), we could demonstrate for the firsttime a direct participation of the PKA pathway on thehormonal regulation of a structure forming process inGC-spheroids suggesting the involvement of cytoskele-tal proteins.

Altogether the intervention in the PKC pathway hadless prominent effects compared to the modulation ofPKA. The PKC-activator DHI clearly inhibited GC-spheroids, but the effect of hCG and LH was notsignificantly increased. We reasoned that PKC plays arole in the regulation of sprouting, but does not take partin the hormone-dependent inhibition of GC-spheroidswhich confirms observations by Murphy (2000), whodescribe that PKC is responsible for cell proliferationand differentiation, but its interplay with gonadotropichormones is not fully understood and seems not to be acrucial process for the development of the CL.

As shown in our experiments, the specific inhibition ofPKCa and PKCb1 by GO 6976 and especially PKCa,PKCg, and PKCd by GO 6983 did not influence sproutformation in GC, so that we can conclude that theseisoenzymes of PKC are not triggering LH and hCGsignals to control the formation of sprouts.

Molecular Reproduction and Development. DOI 10.1002/mrd

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 781,1238019833,8968704

DMSO 10µl/mlControl

Fig. 4. GC-spheroids were treated with ECBM (bar A) or DMSO inECBM (10 ml/ml) (bar B). Four independent experiments wereperformed measuring 10 spheroids per experimental group. Nosignificant difference appeared. [See color version online at www.interscience.wiley.com.]

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 330,591397,8364201,5037333,8102247,7496709,8152

DHI 10µMControlhCG

10U/mlhCG+DHI

LH35mU/ml

LH+DHI

**

*

FEDCBA

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 622,9437518,7028297,6844280,3989607,867735,6674

ControlGÖ 6976

7,9nMhCG

10U/mlhCG+GÖ

6976LH

35mU/mlLH+GÖ

6976

**

A CB FED

µm

0

200

400

600

800

1000

1200

1400

CS

L (

µm

)

µm 645,9795518,7028337,3088280,3989695,9387735,6674

ControlGÖ 6983

10µMhCG

10U/mlhCG+GÖ

6983LH

35mU/mlLH+GÖ

6983

** *

A EDCB F

c

b

a

Fig. 3. Three-dimensional sprouting assay with GC-spheroidsembedded in collagen gel and stimulated with modulators of PKCpathway. The diagram shows the means�SEM of the cumulativesprout lengths (CSL). At least three independent experiments wereperformed measuring 10 spheroids per experimental group. * Indicatessignificant differences of P< 0.05 between the groups. a: GC-spheroidswithout stimulation (bar A), and after stimulation with 10 mM PKC-activator DHI (bar B), 10 U/ml hCG (bar C), and 35 mU/ml LH (bar E),as well as hCG and DHI (bar D), respectively LH and DHI (bar F) withthe same concentrations of the substances. b: GC-spheroids withoutstimulation (bar A), and after stimulation with 7.9 nM PKC-inhibitorGO 6976 (bar B), 10 U/ml hCG (bar C), and 35 mU/ml LH (bar E), as wellas hCG and GO 6976 (bar D), respectively LH and GO 6976 (bar F) withthe same concentrations of the substances. c: GC-spheroids withoutstimulation (bar A), and after stimulation with 10 mM PKC-inhibitorGO 6983 (bar B), 10 U/ml hCG (bar C), and 35 mU/ml LH (bar E), as wellas hCG and GO 6983 (bar D), respectively LH and GO 6983 (bar F) withthe same concentrations of the substances. [See color version online atwww.interscience.wiley.com.]

GONADOTROPIN RESPONSE IN HUMAN GRANULOSA CELLS 1315

With this work we were able to shed further light onthe hormonal regulation of GC. The spheroidal cultureassay turned out to be a good model to study the functionof CL development. Our results demonstrate thatmainly PKA is responsible for the development of theCL and that it is the main effector of hCG, thus playingan important role in maintaining the CL in earlypregnancy.

REFERENCES

Ascoli M, Fanelli F, Segaloff DL. 2002. The lutropin/choriogonado-tropin receptor, a 2002 perspective. Endocr Rev 23:141–174.

Bain J, McLauchlan H, Elliot M, Cohen P. 2003. The specificities ofprotein kinase inhibitors: An update. Biochem J 371:199–204.

Baum G, Suh BS, Amsterdam A, Ben-Zeev A. 1990. Regulation oftropomyosin expression in transformed granulosa cell lines withsteroidogenic ability. Dev Biol 142:115–128.

Ben Ze’ev A, Amsterdam A. 1987. In vitro regulation of granulosacell differentiation. Involvement of cytoskeletal protein expression.J Biol Chem 262:5366–5376.

Bourdage RJ, Fitz TA, Niswender GD. 1984. Differential steroidogenicresponses of ovine luteal cells to ovine luteinizing hormone andhuman chorionic gonadotropin. Proc Soc Exp Biol Med 175:483–486.

Bukovsky A, Indrapichate K, Fujiwara H, Cekanova M, Ayala ME,Dominguez R, Candle MR, Wimalsena J, Elder RF, Copas P, FosterJS, Fernando RI, Henley DC, Upadhyaya NB. 2003. Multipleluteinizing hormone receptor (LHR) protein variants, interspeciesreactivity of anti-LHR mAb clone 3B5, subcellular localization ofLHR in human placenta, pelvic floor and brain, and possible role forLHR in the development of abnormal pregnancy, pelvic floordisorders and Alzheimer’s disease. Reprod Biol Endocrinol 1:46.

Davies SP, Reddy H, Caivano M, Cohen P. 2000. Specificity andmechanism of action of some commonly used protein kinaseinhibitors. Biochem J 351:95–105.

Dufau ML. 1998. The luteinizing hormone receptor. Annu Rev Physiol60:461–496.

Gilchrist RL, Ryu KS, Ji I, Ji TH. 1996. The luteinizing hormone/chorionic gonadotropin receptor has distinct transmembrane con-ductors for cAMP and inositol phosphate signals. J Biol Chem 271:19283–19287.

Gschwendt M, Dieterich S, Rennecke J, Kittstein W, Mueller HJ,Johannes FJ. 1996. Inhibition of protein kinase C m by variousinhibitors. Differentiation from protein kinase C isoenzymes. FEBSLett 392:77–80.

Gudermann T, Birnbaumer M, Birnbaumer L. 1992. Evidence for dualcoupling of the murine luteinizing hormone receptor to adenylylcyclase and phosphoinositide breakdown and Ca2þ mobilization.Studies with the cloned murine luteinizing hormone receptorexpressed in L cells. J Biol Chem 267:4479–4488.

Herrlich P, Ponta H, Rahmsdorf HJ. 1992. DNA damage-inducedgene expression: signal transduction and relation to growthfactor signaling. Rev Physiol Biochem Pharmacol 119:187–223.

Herrlich A, Kuhn B, Grosse R, Schmid A, Schultz G, Guderman T. 1996.Involvement of Gs and Gi proteins in dual coupling of the luteinizinghormone receptor to adenylyl cyclase and phospholipase C. J BiolChem 271:16764–16772.

Ji TH, Ryu KS, Gilchrist R, Ji I. 1997. Interaction, signal generation,signal divergence, and signal transduction of LH/CG and thereceptor. Recent Prog Horm Res 52:431–453.

Kuhn B, Schmid A, Harteneck C, Gudermann T, Schultz G. 1996. Gproteins of the Gq family couple the H2 histamine receptor tophospholipase C. Mol Endocrinol 10:1697–1707.

Martiny-Baron G, Kazanietz MG, Mischak H, Blumberg PM, Kochs G,Hug H, Marme D, Schachtele C. 1993. Selective inhibition of proteinkinase C isoenzymes by the indolocarbazole GO 6976. J Biol Chem263:9194–9197.

Murphy BD. 2000. Models of Luteinization. Biol Reprod 63:2–11.Muller T, Gromoll J, Simoni M. 2003. Absence of exon 10 of the human

luteinizing hormone (LH) receptor impairs LH, but not humanchorionic gonadotropin action. J Clin Endocrin Metabol 88:2242–2249.

Niswender GD, Juengel JL, Silva PJ, Rollyson MK, McIntush EW.2000. Mechanisms controlling the function and life span of the corpusluteum. Physiol Rev 80:1–29.

Rajagopalan-Gupta RM, Mukherjee S, Zhu X, Ho YK, Hamm H,Birnbaumer M, Birnbaumer L, Hunzicker-Dunn M. 1999. Roles of Giand Gq/11 in mediating desensitization of the luteinizing hormone/choriogonadotropin receptor in porcine ovarian follicular mem-branes. Endocrinology 140:1612–1621.

Ryu RS, Gilchrist RL, Koo YB, Ji I, Ji TH. 1998. Gene, interaction,signal generation, signal divergence and signal transduction of theLH/CG receptor. Int J Gynaecol Obstet 60:9–20.

Walz A, Keck C, Weber H, Kissel C, Pietrowski D. 2005. Effects ofluteinizing hormone and human chorionic gonadotropin on corpusluteum cells in a spheroid cell culture system. Mol Reprod Dev 72:98–104.

Zeleznik AJ. 1998. In vivo responses of the primate corpus luteum toluteinizing hormone and chorionic gonadotropin. Proc Natl Acad SciUSA 95:11002–11007.

Molecular Reproduction and Development. DOI 10.1002/mrd

1316 J. BECKER ET AL.