Regulation of AP1 (Jun/Fos) Factor Expression and Activation inOvarian Granulosa CellsRELATION OF JunD AND Fra2 TO TERMINAL DIFFERENTIATION*

Received for publication, April 26, 2000, and in revised form, July 3, 2000Published, JBC Papers in Press, August 8, 2000, DOI 10.1074/jbc.M003555200

S. Chidananda Sharma and JoAnne S. Richards‡

From the Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas 77030

AP1 transcription factors control rapid responses ofmammalian cells to stimuli that impact proliferation,differentiation, and transformation. To determinewhich AP1 factors are present in and regulated by hor-mones in ovarian cells during specific stages of prolifer-ation and differentiation, we used both in vitro and invivo models, Western blotting, immunohistochemistry,DNA binding assays, and transfections of AP1 promoter-reporter constructs. The expression patterns of Jun andFos family members in response to hormones (follicle-stimulating hormone (FSH), luteinizing hormone (LH),and cAMP) were distinct. JunB, c-Jun, c-Fos, and Fra2were rapidly but transiently induced by FSH in imma-ture granulosa cells. JunD and Fra2 were induced by LHand maintained as granulosa cells terminally differenti-ated into luteal cells. Forskolin and phorbol myristateacetate acted synergistically to enhance transcription ofan AP1(273COL)-luciferase construct. JunD appears tobe one mediator of this effect, since JunD was a majorcomponent of the AP1-DNA binding complex in granu-losa cells, and menin, a selective inhibitor of JunD,blocked transcription of 273COL-luciferase. Thus, FSHand LH via cAMP induce specific AP1 factors, the AP1expression patterns are distinct, and that of JunD andFra2 correlates with the transition of proliferating gran-ulosa cells to terminally differentiated, non-dividing lu-teal cells.

Growth of ovarian follicles involves regulated proliferationand differentiation of granulosa cells (1, 2). Exponential growthof granulosa cells in preovulatory follicles is followed by theirterminal differentiation and formation of the corpus luteum, atransition initiated by the ovulatory surge of luteinizing hor-mone (LH).1 This transition of proliferating granulosa cells to

non-dividing luteal cells is associated with specific changes inthe expression of cell cycle regulatory molecules that controlspecific kinase cascades. The cell cycle activators, cyclin D2 andcyclin E, are increased by FSH and steroids in rapidly growingpreovulatory follicles but are rapidly turned off by the surge ofLH that terminates follicular growth (1, 2). Conversely, the cellcycle inhibitors, p27KIP1 and p21CIP1, are induced by the LHsurge (1, 2). Thus, the gonadotropins impact cell cycle progres-sion at multiple steps depending on the stage of granulosa celldifferentiation.

The effects of the gonadotropins on ovarian cell proliferationand differentiation are mediated by changes in intracellularcAMP that activates cAMP-dependent protein kinases (proteinkinase A) (3), as well as several other kinases (4, 5). Knowntargets of protein kinase A are the cAMP regulatory element(CRE)-binding protein CREB and the CREB-binding proteinCBP (6). CREs are essential for transcriptional activation ofaromatase (7) and inhibin a (8), as well as the AP1 factors, c-fos(9) and fra2 (10). AP1 factors, in turn, mediate FSH-regulatedexpression of inhibin bA by binding to a variant CRE in thepromoter of the bA gene (11). Members of the AP1 transcriptionfactor superfamily are proto-oncogenes that regulate cell pro-liferation and transformation (12). They have also been asso-ciated with differentiation (13, 14). These observations suggestthat specific AP1 factors or the combination of specific AP1factors may regulate different sets of functions at specificstages of granulosa cell differentiation. Despite the intenseresearch efforts that have analyzed the function of the Jun/Fosfamily in many different cell types, relatively few studies haveanalyzed how hormones regulate the expression of AP1 factorsin granulosa cells or how these factors might impact FSH or LHactivation of specific genes. Therefore, it becomes imperative toknow which AP1 factors are present in granulosa cells andwhich might be regulated or activated by FSH and LH.

The AP1 transcription factor family is composed of Jun fam-ily members (c-Jun, JunB, and JunD) that can form homo- orheterodimers among themselves and bind to AP1 consensusDNA (TGACTCA) (10, 12, 15). Jun proteins also dimerize withFos family members (c-Fos, FosB, Fra1, and Fra2) (10, 12, 16).Thus, the functional activity of AP1 in any given cell is depend-ent in part on the relative amount of specific Jun/Fos factorspresent in that cell. Most AP1 transcription factors are presentat low levels in cells but are rapidly induced and activated inresponse to specific stimuli. For example, the c-fos gene has acomplex promoter that contains a CRE, a serum response ele-ment (SRE), and a Sis-inducible enhancer allowing it to re-spond to multiple hormones, growth factors, and cytokines (10,15, 17). Not surprisingly then, c-fos has been shown to beinduced by FSH in granulosa cells (9). The promoter of the junBgene is also complex with Stat3, CRE, CAAT enhancer (CAAT)binding domains, as well as an interleukin-6 regulatory region

* This work was supported in part by National Institutes of HealthGrants HD-16229 and a Specialized Cooperative Centers Program inReproduction Research Grant HD-07495. The costs of publication ofthis article were defrayed in part by the payment of page charges. Thisarticle must therefore be hereby marked “advertisement” in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

‡ To whom correspondence should be addressed: Dept. of Molecularand Cellular Biology, Baylor College of Medicine, One Baylor Plaza,Houston, TX 77030. Tel.: 713-798-6238; Fax: 713-790-1275; E-mail:joanner@bcm.tmc.edu.

1 The abbreviations used are: LH, luteinizing hormone; FSH, follicle-stimulating hormone; oFSH, ovine FSH; PMA, phorbol myristate ace-tate; EMSA, electrophoretic mobility shift assays; CRE, cAMP regula-tory element; CREB, CRE-binding protein; CBP, CREB-bindingprotein; hCG, human chorionic gonadotropin; PAGE, polyacrylamidegel electrophoresis; WCE, whole cell extract; MAPK, mitogen-activatedprotein kinase; ERK, extracellular signal-regulated kinase; SRE, serumresponse element; OA, okadaic acid; H, hypophysectomized; E, estra-diol; T, testosterone.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 43, Issue of October 27, pp. 33718–33728, 2000© 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org33718

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

(10, 18). Different promoter regions are utilized in different celltypes (18). Although JunB is expressed in ovarian cells (19), thefactors regulating its expression have not been analyzed. Incontrast, the promoter of c-jun is simpler (10, 15), and junD isconstitutively expressed in many cell lines (19).

The functional activation of AP1 factors in mammalian cellsrequires phosphorylation (15). The activated AP1 factors havebeen shown to regulate gene expression by direct as well asindirect mechanisms (12, 16). Direct activation requires thatAP1 factors bind an AP1 regulatory domain present in thepromoters of target genes and that the AP1 factors be phospho-rylated in their activation domain. The AP1 DNA binding do-main was originally known as the 12-O-tetradecanoylphorbol-13-acetate response element, since in response to phorbol estertumor promoters such as 12-O-tetradecanoylphorbol-13-ace-tate, the AP1 factors are phosphorylated, activated, and in-duced transcription (15). However, AP1 factors can be phospho-rylated and activated in response to numerous other agonists,suggesting that their activation profile is more complex thanoriginally thought. In fact, PMA which activates c-Jun N-ter-minal kinase does not activate Fos-regulating kinase, Frk (15).Activated AP1 factors can also enhance gene transcription byindirect mechanisms that involve their binding to other tran-scription factors via protein-protein interactions but do notrequire their binding to DNA. For example, members of theJun/Fos family of transcription factors have recently beenshown to interact with both Sp1/Sp3 and Smad proteins toenhance transcriptional activity of specific promoters with GC-rich and SMAD-binding regulatory elements, respectively (14,20). This may be one attractive way in which FSH or LHregulate the expression of cyclins, p21cip or p27kip, as well assgk. c-Jun has also been shown recently to recruit CBP forenhanced activation of CREB (21). Such a mechanism may becontribute to FSH activation of aromatase. The importance ofc-Jun is underscored by the embryonic lethality of mice null forc-jun (22). In contrast, mice null for c-fos are viable but exhibitpleiotropic effects (23).

Based on the roles of AP1 factors in cell proliferation, differ-entiation, and transformation, we sought to determine whichAP1 factors are present in and regulated by hormones in gran-ulosa cells during follicular growth and terminal differentia-tion to luteal cells. For these studies we have used both in vitroand in vivo models, Western blotting, and DNA binding-elec-trophoretic mobility shift assays to identify and quantify Jun(c-Jun, JunD, and JunB) and Fos (c-Fos, FosB, Fra1, and Fra2)family members present in granulosa cells. The hormonal ac-tivation of AP1 factors in granulosa cells was analyzed bytransiently transfecting cells with AP1 promoter-luciferase re-porter constructs. Specific cellular signaling cascades were ac-tivated by adding forskolin, an agonist that stimulates cAMP,as well as the phorbol ester PMA that is a known activator ofAP1 factors in other cell types. Results show that each AP1factor exhibited a specific expression pattern and response tohormones. Rapid induction of JunB, c-Fos, and Fra2 by FSHcorresponds to the pattern of other immediate-early genes ingranulosa cells, whereas rapid induction of junD and fra2 byLH in granulosa cells of preovulatory follicles resulted in theirprolonged, stable expression during the transition to lutealcells. JunD, in addition to c-Jun and c-Fos, was present as amajor Jun factor comprising the AP1-DNA binding complexand controlled transactivation of the AP1-promoter reporterconstruct. Thus, JunD and Fra2 as well as c-Jun and c-Fosappear to be selectively expressed in terminally differentiatedluteal cells indicating the composition of AP1 factors changesduring granulosa cell proliferation and differentiation.

EXPERIMENTAL PROCEDURES

Reagents—Media and cell culture reagents and materials were pur-chased from Life Technologies, Inc., Sigma, Research Organics (Cleve-land, OH), Fisher, Corning Glass, and HyClone (Logan, UT). Trypsin,soybean trypsin inhibitor, DNase, phorbol myristate acetate (PMA),dithiothreitol, 17b-estradiol (E), and propylene glycol were all pur-chased from Sigma. Ovine FSH (oFSH-16) was a gift of the NationalHormone and Pituitary Program (Rockville, MD). Human chorionicgonadotropin (hCG) was from Organon Special Chemicals (West Or-ange, NJ). Forskolin was from Calbiochem. Antibodies for c-Jun (PC06)and c-Fos (PC05) were obtained from Calbiochem. Antibodies for JunB(SC-73 for Western blots and immunohistochemistry; SC46 for EMSA),JunD (SC74x), FosB (SC-48x), Fra1 (SC605x), and Fra2 (SC-604x) werefrom Santa Cruz Biotechnology (Santa Cruz, CA). TRIzol reagent (num-ber 15596) was obtained from Life Technologies, Inc. Electrophoresisand molecular biology grade reagents were purchased from Sigma,Bio-Rad, and Roche Molecular Biochemicals. Oligonucleotides werepurchased from Genosys (The Woodlands, TX). Reverse transcriptase-polymerase chain reaction reagents were from Promega (Madison, WI)except for deoxyribonucleotides (dNTPs; Roche Molecular Biochemi-cals). [a-32P]dCTP was from ICN Radiochemicals (Costa Mesa, Ca).Hyperfilm was purchased from Amersham Pharmacia Biotech. Themenin expression plasmid and antibodies were generously provided byDr. Sunita Agarwal, NIDDK, National Institutes of Health, Bethesda.The Gal4-ELK and 4XGAl-luciferase vectors were generously providedby Dr. Philip Stork (Oregon Health Science Center, Portland OR), andthe 273COL-luciferase vector was provided by Dr. Michael Karin (Uni-versity of California, San Diego, La Jolla, CA).

Animals—Intact and hypophysectomized (H) immature (day 23 ofage) Holtzman Harlan Sprague-Dawley female rats were obtained fromHarlan (Indianapolis, IN) and housed under a 16:8 light:dark schedulein the Center for Comparative Medicine at Baylor College of Medicineand provided food and water ad libitum. Animals were treated inaccordance with the NIH Guide for the Care and Use of LaboratoryAnimals, as approved by the Animal Care and Use Committee at BaylorCollege of Medicine (Houston, TX). To obtain granulosa cells for pri-mary cultures, immature rats were injected with estradiol (E, 1.5 mg/day for 3 days) (7). To analyze the expression of AP1 factors duringfollicular development in vivo, H rats were treated with E (HE, asabove) to stimulate the growth of large preantral follicles. To analyzethe response of granulosa cells to acute exposure to FSH, some HE ratswere injected intravenously with 5 mg of oFSH. Ovaries (for immuno-histochemistry) and granulosa cells (for whole cell extracts) were iso-lated from H and HE rats prior to FSH (0 h) or from HE rats exposed toFSH for 2 and 8 h (HE, FSH 2, 8 h). Other HE rats were injectedsubcutaneously with 1.0 mg of oFSH twice daily for 2 days (HEF, 48 h)to stimulate the growth of preovulatory follicles that express aro-matase, LH receptor, and inhibin-a (24). HEF rats were subsequentlyinjected intravenously with 10 IU of hCG, an LH-like hormone, tostimulate ovulation and luteinization. Ovaries and granulosa cells wereprepared prior to (0 h) and at 2, 4, 8, 12, and 24 h after hCG forimmunohistochemistry and WCE, respectively (shown schematically inFig. 1).

Granulosa Cell Cultures—Granulosa cells were harvested from es-tradiol-primed intact immature (day 25) rats as described previously(7). Briefly, cells were cultured at a density of 1 3 106 cells per 3 ml ofserum-free medium (Dulbecco’s modified Eagle’s medium:Ham’s F-12medium containing penicillin and streptomycin) in multiwell (35-mm)dishes that were serum-coated. Cells were cultured in defined mediumovernight (0 h) followed by the addition of FSH (50 ng/ml) and testos-terone (T; 10 ng/ml), forskolin (10 mM), and other agonists as indicatedin the figures legends. FSH/T were used to stimulate the differentiationof granulosa cells to a preovulatory phenotype in which aromatase (7),LH receptor (25), and inhibin (8) are expressed. Forskolin was added tothese cells to mimic the LH surge and luteinization. Forskolin alonewas used to determine the relative effects of cAMP on specific cellfunctions (4, 7). PMA was used as an agonist for protein kinase C (15,26) (shown schematically in Fig. 1). Hormones, agonists, and antago-nists were added as indicated in the figure legends.

Transfections—After culture overnight, granulosa cells were trans-fected by the calcium phosphate precipitation method (27) using 5 mg ofvector DNA:273COL-luciferase (15, 28), ERE-E1b-luciferase (29), 4X-Gal-luciferase and Gal-4-Elk-1 (30), or menin (31) as indicated in thefigure legends. Four hours later the cells were washed thoroughly infresh medium. At that time (0 h) the cells were stimulated with ago-nists. After 6 h of agonist stimulation, cell lysates were prepared andused to measure luciferase activity according to standard protocols.

AP1 Factors in the Ovary, Regulation by FSH and LH 33719

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

Relative light units were normalized to protein (mini-Bradford assay;Bio-Rad) in each sample. All transfections were run in triplicate, andeach set of experiments was repeated at least three times. Unlessotherwise indicated, values represent the mean 6 S.E. of threeexperiments.

Cell Extracts and Western Blot Analyses—Total cell extracts wereprepared from cultured granulosa cells according to a method of Gintyet al. (32) by adding to each well hot (100 °C) Tris buffer containing 10%SDS and b-mercaptoethanol (4). The cells were rapidly scraped with arubber policeman. The cell extract was immediately transferred to anEppendorf tube for 5 min at 100 °C and then stored at 4 °C. Equalvolumes (50 ml) of samples were analyzed by SDS-polyacrylamide gelelectrophoresis (PAGE). Following SDS-PAGE, proteins were electro-phoretically transferred to nylon filter, washed briefly in phosphate-buffered saline, and blotted with either 3% bovine serum albumin or 5%Carnation milk at room temperature for 1 h. Antibodies were added inthe same blocking solutions at the dilutions indicated in the figurelegends. Immunoreactive proteins were visualized with ECL reagentsaccording to the specification of the supplier (Pierce). Immunoreactivebands were quantified by image analysis of autoradiograms (ECL)using AlphaImager 2000 (3.3), Alpha Innotech Corp., San Leandro, CA.

Electrophoretic Mobility Shift Assays (EMSA)—Whole cell extracts(WCE) were prepared from hormonally stimulated granulosa cells inculture or from hormonally primed H rats as described previously (29)and in the figure legends. WCE were incubated 20 min on ice with a32P-labeled consensus AP1 oligonucleotide probe and then subjected tonon-denaturing electrophoresis (0.53 TBE, Tris borate-EDTA) at 150V. Where indicated, either unlabeled probe or antibodies to specific AP1transcription factors were added to the reactions 30 min on ice beforethe addition of labeled DNA.

Immunohistochemical Analyses—Immunohistochemical analyses ofAP1 factors in whole ovaries isolated from hormonally primed rats wereperformed as described previously (24). Ovaries were fixed in 4%paraformaldehyde and were paraffin-embedded. Sections (6 microns)were processed according to routine procedures and then blocked with10% non-immune goat serum followed by incubation with specific an-tibodies diluted 1:50 in 10% goat serum overnight at room temperature.After washing in phosphate-buffered saline, biotinylated anti-rabbitantiserum (Vector, Berlingame, CA) was added for 30 min, slides werewashed, and streptavidin-conjugated horseradish peroxidase was ap-plied for 30 min. Sections were incubated with DAB substrate (3,39-diaminobenzidene) for 2 min, dehydrated without counter-staining, andmounted.

Statistical Analyses—Transfection data were analyzed by one-wayanalysis of variance or Student’s t test. Values were considered signif-icantly different if p , 0.05.

RESULTS

Models for Analyzing Hormone-induced Changes inGranulosa Cell Expression of AP1 Factors

The AP1 transcription factor superfamily represents an im-portant signal transduction system in mammalian cells. How-

ever, relatively little is known about which AP1 factors areexpressed or activated in ovarian granulosa cells. Therefore,our initial goals were to determine which members of the AP1family of transcription factors were present in undifferentiatedgranulosa cells and which factors might be regulated by hor-mones in association with proliferation versus differentiation.

Two models were used to analyze hormone-induced changesin AP1 expression in granulosa cells (Fig. 1).

In Vitro Cultures—In the first model, undifferentiated gran-ulosa cells were harvested and cultured in defined mediumovernight. At that time the undifferentiated cells were eitherstimulated acutely (1.5 h) with forskolin to increase cAMP, thephorbol ester PMA (a known activator of AP1 factors), or thecombination of forskolin and PMA. To stimulate the differen-tiation of granulosa cells to a preovulatory phenotype, FSH/Twere added to the cells for 48 h. These differentiated cells(FSH/T), which express aromatase and LH receptor, were alsostimulated acutely with forskolin, PMA, or the combination tomimic the acute effects of the LH surge. Finally, the differen-tiated cells were cultured with forskolin for longer periods tomimic the LH-induced process of luteinization in which gran-ulosa cells cease to divide and are terminally differentiated.

Hormonal Treatment of rats in Vivo—In the second model,hypophysectomized immature rats (H) were injected with es-tradiol (E; HE rats) to stimulate granulosa cell proliferationand the growth of large preantral follicles. Subsequently, FSH(F) was administered to the HE rats (HEF) to examine rapideffects as well as long term effects of the gonadotropins, leadingto the development of preovulatory follicles (see “ExperimentalProcedures” for details). HEF rats were injected with an ovu-latory dose of the LH-like hormone, hCG, to stimulate terminaldifferentiation of granulosa cells. This process of luteinizationleads to the formation of corpora lutea.

Differential Induction of Jun/Fos Transcription Factorsin Undifferentiated Granulosa Cells

To analyze the response of undifferentiated cells, granulosacells were harvested from immature rats and cultured withforskolin to stimulate cAMP production for different time in-tervals (0, 0.33 (20 min), 0.66 (40 min), 1.5, 3, 6, 12, 24, and48 h). Cell extracts were prepared and analyzed by Westernblotting using antibodies specific for individual members of theAP1 transcription factor family as follows: c-Jun, JunB, JunD,c-Fos, FosB, Fra1, and Fra2 (Fig. 2).

Jun Family Members—When undifferentiated granulosa

FIG. 1. Schematic models for ana-lyzing hormone-induced expressionof AP1 factors in ovarian granulosacells. See text for details.

AP1 Factors in the Ovary, Regulation by FSH and LH33720

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

cells were acutely stimulated with forskolin, different patternsof AP1 protein expression were observed (Fig. 2). Among theJun family members, immunoreactive c-Jun was present at lowlevels in granulosa cells cultured overnight in medium alone (0h). The addition of forskolin stimulated 1-, 5-, 6-, and 8-foldincreases in c-Jun at 20, 40, and 90 min and 3 h, respectively.Levels of c-Jun decreased at 6 h and remained at this level atsubsequent time intervals. JunB exhibited a similar pattern ofresponse as that of c-Jun, but the magnitude of induction wasgreater. JunB increased 1-, 3-, 14-, and 16-fold at 20, 40, and 90min and 3 h, respectively. Levels of JunB protein were de-creased at 6 h and remained low. In contrast, JunD was presentin granulosa cells at 0 h, and its expression was relativelyunaffected by the addition of forskolin.

Fos Family Members—Among the Fos family members, im-munoreactive c-Fos increased rapidly in response to forskolin;2- and 6-fold increases occurred at 20 and 90 min, respectively(Fig. 2). However, this increase was transient; immunoreactivec-Fos returned to the base-line (0 h) level between 3 and 6 h andremained low thereafter. Similarly, FosB and Fra1 exhibitedincreases in response to forskolin at 20 and 40 min but declinedto basal or near undetectable levels, respectively, between 6and 24 h of culture. Fra2 showed the most dramatic changes inresponse to forskolin (Fig. 2, bottom panel). Fra2 increased 3-,3-, and 11-fold at 20, 40, and 90 min, respectively, remainedelevated at 3 h, and then declined gradually at 6–12 h and waslow again at 24–48 h.

Collectively, these data indicate that JunD is constitutivelyexpressed in cultured granulosa cells, whereas other AP1 fac-tors, especially c-Jun, JunB, Fra2, and c-Fos, are rapidly but

transiently induced/increased by forskolin/cAMP. Immunocy-tochemical analyses documented that AP1 factors were local-ized to the nuclei of granulosa cells (data not shown).

Induction of AP1 Transcription Factors inUndifferentiated and Differentiated Granulosa Cells,

Comparison of the Effects of Forskolin and PMA

The transcriptional activation of AP1 factors has been shownin other cells types to be stimulated by the phorbol ester, PMA.Therefore, we next determined if the effects of forskolin (cAMP)on AP1 factor expression were similar to, different from, orsynergistic with those of PMA and if the effects were dependenton the stage of granulosa cells differentiation. For these exper-iments, granulosa cells were cultured overnight in defined me-dium. At that time (0 h), the undifferentiated cells were treatedacutely (for 1.5 h) with either forskolin (10 mM), PMA (20 nM), orthe combination of forskolin and PMA. Some cells were exposedto forskolin for 24 h. To examine the responses of differentiatedgranulosa, additional cells were cultured in the presence ofFSH/T for 48 h. The differentiated cells were stimulatedacutely (for 1.5 h) with agonists to mimic the LH surge or wereexposed to forskolin for 24 and 48 h to induce luteinization. Cellextracts were prepared, and the inducibility of two factors,JunB and c-Fos, was analyzed by Western blotting (Fig. 3).

In untreated granulosa cells, immunoreactive JunB in-creased 44-, 19-, or 52-fold, respectively, by exposure to forsko-lin, PMA, or the combination for 1.5 h (Fig. 3, upper panel). Theinduction by forskolin was transient, and JunB protein re-turned to basal levels by 24 h. JunB was also low in theFSH/T-differentiated cells but could be induced by forskolin,PMA, or the combination 28-, 21-, and 34-fold, respectively.These results indicate that JunB is rapidly but transientlyinduced by forskolin in control and differentiated granulosacells. PMA alone or in combination with forskolin also in-creased JunB. However, since the effects were not additive orsynergistic, these two agonists appear to induce JunB by sim-ilar or overlapping mechanisms. Likewise, c-Fos was inducedequally well by forskolin or PMA in undifferentiated granulosacells, and the combined treatment was additive (Fig. 3, lowerpanel). In the FSH/T-differentiated cells, the response to fors-kolin appeared less in this experiment than in subsequentexperiments (compare results in Fig. 3 and Fig. 4). Thesedifferences may be due to the rapid and transient nature of theresponse. c-Fos was consistently increased by PMA or PMA andforskolin.

FIG. 2. The expression of AP1 factors in granulosa cells isregulated by forskolin/cAMP. Granulosa cells were isolated fromE-primed immature rats and cultured in defined medium overnight (0h). Forskolin was added, and cell extracts were prepared at the timeintervals indicated. Equal volumes of extract were resolved by SDS-PAGE and electrophoretically transferred to nylon filters. AP1 factorswere identified by specific antibodies and ECL detection. Antibodies toc-Jun, JunB, and JunD were diluted 1:250, 1:500, and 1:3000, respec-tively. Antibodies to c-Fos were diluted 1:500; antibodies to FosB, Fra1,and Fra2 were diluted 1:3000.

FIG. 3. AP1 factors are increased by forskolin and PMA inundifferentiated and differentiated granulosa cells. Granulosacells were isolated and cultured overnight (0 h) as in Fig. 1. At 0 h, theundifferentiated cells were stimulated either with forskolin (10 mM),PMA (20 nM), or the combination for 1.5 h or with forskolin for 24 h.Other cells were cultured with FSH (50 ng/ml) and testosterone (T; 10ng/ml) for 48 h to stimulate differentiation. At that time, agonists wereadded to the differentiated cells as indicated. Immunoreactive JunBand c-Fos were detected by Western blotting as in Fig. 1.

AP1 Factors in the Ovary, Regulation by FSH and LH 33721

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

Differential Induction of AP1 Factors in DifferentiatedGranulosa Cells

To analyze the effects of forskolin and PMA on the expressionof other AP1 factors in the differentiated cells and to determineif the tumor promoter, okadaic acid (OA) (33), selectively af-fected their expression in ovarian cells, additional cultureswere studied (Fig. 4). Granulosa cells were either culturedovernight in medium alone (0 h) or with FSH/T (48 h) ascontrols. Forskolin was added to the differentiated cells, andextracts were prepared 1.5 and 24 h later. In addition, FSH/T-treated granulosa cells were exposed to PMA for 2 h or OA for3 h.

As in previous experiments (Figs. 1 and 2), immunoreactivec-Jun was low at 0 and 48 h (Fig. 4, lanes 1 and 2). In theFSH/T-differentiated cells, forskolin increased c-Jun 2.5- and5.5-fold at 1.5 and 24 h, respectively (Fig. 4, lanes 3 and 4).More dramatic was the 14-fold increase in c-Jun stimulated byPMA (Fig. 4, lane 5). In contrast, OA had little effect on c-Jun(Fig. 4, lane 6). Immunoreactive JunB was also low at 0 and48 h after FSH/T but increased 7- and 3.5-fold in response toforskolin at 1.5 and 24 h, respectively (lanes 3 and 4). Thisresponse was similar to that of JunB at 0 h (Figs. 2 and 3).JunB increased 3-fold in response to PMA (lane 5) and 6-fold

with OA (lane 6). As noted above, JunD was not affected bygranulosa cell differentiation to the preovulatory phenotype(FSH/T 48 h; lane 2), PMA (lane 5) or OA (lane 6). However,JunD was increased in luteinizing cells (i.e. those exposed toFSH/T 48 h followed by forskolin for 24 h; Fig. 4, lane 4).

Immunoreactive c-Fos was low in the control and FSH/T-treated cells (Fig. 4, lanes 1 and 2) but increased 5- and 2.5-foldwith forskolin at 1.5 and 24, respectively (lanes 3 and 4), 4-foldin response to PMA (lane 5), and 5-fold in response to OA (lane6). In contrast, levels of FosB and Fra1 decreased from 0 to 48 h(lanes 1 and 2). FosB was increased with OA (lane 6), whereasFra1 increased with forskolin, 1.5 h (lane 3); otherwise theirexpression remained low and unchanged. As in previous exper-iments, Fra2 was low in control cells but increased 8-fold afterFSH/T, 48 h. Most dramatic were the 12- and 36-fold increasesin Fra2 in response to forskolin (lanes 3 and 4) and the 12-foldincreases in response to either PMA (lane 5) or OA (lane 6).Collectively, these results indicate that in the FSH/T-treatedgranulosa cells induction of JunB, c-Fos, FosB, and Fra2 aremost sensitive to OA, a potent inhibitor of phosphatase (PP1and PP2A) activity. c-Jun and c-Fos are induced selectively byPMA. JunB and Fra2 are most responsive to FSH/T and fors-kolin (cAMP) with Fra2 reaching highest levels in the lutein-ized cells.

Components of the AP1-DNA Binding Complex Changeduring Granulosa Cells Differentiation

To determine if the AP1 factors present in granulosa cellswere capable of binding to AP1 consensus DNA, WCE wereprepared from granulosa cells cultured overnight in mediumalone (control, undifferentiated cells; 0 h) or in the presence offorskolin, PMA, or the combination for 1.5 h. Additional ex-tracts were prepared from FSH/T-differentiated cells (48 h)followed by forskolin for 24 h, a regimen shown (Fig. 2) toincrease selectively levels of immunoreactive JunD and Fra2.WCE were incubated with a labeled AP1 consensus oligonu-cleotide probe alone, a 100-fold excess of unlabeled oligonucleo-tide, or in the presence of antibodies specific for c-Jun, JunD,c-Fos, FosB, Fra1, and Fra2. As shown in Fig. 5, two protein-DNA complexes, one major and one minor, were formed whenWCE from control (Fig. 5, 0 h; lane 1) granulosa cells wasincubated with probe alone. Both complexes were reduced inthe presence of a 100-fold excess of unlabeled DNA (lane 2).Antibodies to c-Jun (lane 3), JunD (lane 4), c-Fos (lane 5), FosB(lane 6), and Fra2 (lane 8) all caused a supershift of the majorprotein-DNA complex but did not alter the migration of theminor complex. Antibodies to Fra1 (lane 7) did not stimulate asupershift of the complex present in control granulosa cells.

When WCE of differentiated granulosa cells (FSH/T 1 fors-kolin) were analyzed by EMSA, the major protein-DNA com-plex was markedly increased (Fig. 5, lane 9). Supershift anal-yses indicated that c-Jun (lane 10), JunD (lane 11), c-Fos (lane12), and Fra2 (lane 15) were the AP1 components of the com-plex since each antibody caused a marked shift in the majorprotein-DNA complex. In contrast, antibodies to FosB (lane 13)and Fra1 (lane 14) did not cause supershifts, consistent withtheir lower abundance in extracts of differentiated cells (Fig. 3).Although no antibody alone caused a complete supershift of thecomplex, the addition of two or more antibodies to the reactionmixture caused a more complete supershift of the complex(data not shown). These data indicate that c-Jun, JunD, c-Fos,and Fra2 are more abundant in the differentiated granulosacells than in the undifferentiated cells.

To analyze the binding activity of JunB, extracts of undiffer-entiated granulosa cells were prepared at 0 h or after treat-ment with forskolin, PMA, or forskolin and PMA for 1.5 h. The

FIG. 4. Expression of AP1 factors in differentiated cells is ag-onist- and time-dependent. Granulosa cells were cultured as in Fig.2. Following 48 h of culture with FSH/T, granulosa cells were exposedto forskolin (Fo) for 1.5 or 24 h, PMA (P) for 2 h, or okadaic acid (OA; 10nM) for 3 h. Western blotting was performed as in Fig. 1. Immunoreac-tive bands were detected using an AlphaImager 2000 and plotted aspercent of control, 0 h values. The data are representative of twoseparate experiments.

AP1 Factors in the Ovary, Regulation by FSH and LH33722

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

protein-DNA complexes formed with WCE from control cellswere not supershifted with JunB antibody, whereas extractsfrom the forskolin-treated cells (containing high levels of im-munoreactive Jun B; Figs. 1 and 2) were supershifted with theJunB antibody (data not shown) Thus, JunB is an induciblecomponent of the AP1 complex in cells exposed to acute stim-ulation by forskolin.

Forskolin and PMA Act Synergistically to InduceExpression of the 273COL-luciferase Transgene

To examine the ability of granulosa cells to transactivatepromoters containing AP1 regulatory domains, three promoter-luciferase reporter constructs were initially tested. The pro-moters contained either 273 base pairs of the human collagen-ase gene (273COL), 4 concatamers of a consensus AP1-bindingsite (4XAP1), or a single AP1 domain (1XAP1). Since in initialtests each responded in a similar manner to forskolin, PMA, orthe combination, we have used the 273COL-luciferase con-struct for the studies described herein. When the 273COL-luciferase construct was transfected into control granulosacells (0 h), forskolin stimulated a 9-fold increase, whereas PMAstimulated only a 3-fold increase. The combination of forskolinand PMA induced a 30-fold increase in luciferase activity,indicating they exert a synergistic response (Fig. 6).

To determine if the effects of forskolin and PMA on the AP1promoter elements were specific, we tested the response of twoother promoter-reporter constructs to forskolin and PMA (Fig.6). We chose an estrogen receptor response element E1b-lucif-erase construct since we have previously shown that it re-sponds to forskolin (29), and others have reported activation of

estrogen receptor by MAPK (34). When this vector was trans-fected into granulosa cells, forskolin alone induced a 10-foldincrease, PMA a 7-fold increase, and the combination a 12-foldincrease. Thus the effects of forskolin and PMA were additivebut not synergistic.

Since c-Jun and c-Fos were preferentially increased by PMAalone, we sought to determine if a PMA-specific pathway wasoperative in granulosa cells. For this, we tested the functionalactivation of Elk-1, a ternary complex factor that binds serumresponse factor and enhances transactivation from SREs.Granulosa cells were co-transfected with a 4X-Gal-luciferasereporter construct and an expression vector containing a chi-meric gene in which the Gal-4 DNA binding domain was ligatedto the activation domain of Elk-1 (Fig. 6, 15). Luciferase activ-ity was low in the absence of agonists, increased 2-fold withforskolin, 31-fold with PMA, and 39-fold with forskolin andPMA (Fig. 6).

Finally, since JunD was present in the highest concentra-tions in unstimulated granulosa cells, we sought to determineif the activity of the AP1 complex could be altered by menin, aspecific inhibitor of JunD. Additional cells were co-transfectedwith the 273COL-luciferase reporter construct and a meninexpression vector or empty vector (Fig. 6). As shown, meninmarkedly decreased (75%) the forskolin 1 PMA-mediated lu-ciferase activity but did not affect the control or forskolin-stimulated effects. Thus, JunD, mostly likely in complex withFosB or Fra2 (factors rapidly induced by forskolin and PMA),composes the functional AP1 complex in forskolin-stimulated(undifferentiated) granulosa cells (Fig. 6).

FSH and LH Regulate Transient Versus Stable AP1Expression in Granulosa Cells in Vivo

To determine if the expression of AP1 factors was regulatedby hormones in vivo, hypophysectomized (H) rats were admin-istered estradiol (E), FSH, and LH/hCG to stimulate thegrowth of antral follicles in which granulosa cells are prolifer-ative, preovulatory follicles in which granulosa cells have ac-quired a transitional state of differentiation and luteinizedfollicles in which the granulosa cells have terminally differen-tiated to non-dividing cells, respectively (Fig. 1 and “Experi-mental Procedures”).

Western Blots—Each AP1 factor exhibited its own specificpattern of expression in granulosa cells of growing, ovulating,and luteinizing follicles (Fig. 7). Immunoreactive JunB wasnegligible in granulosa cells of H and HE rats. However, JunBincreased dramatically (13-fold) within 2 h of exposure to FSH.Three distinct immunoreactive bands of JunB indicate thatthey are either phosphorylated forms or degraded products ofJunB. The latter may be more likely since JunB was no longerdetected at 8 h. JunB was also induced rapidly by hCG ingranulosa cells of preovulatory follicles; JunB was highest at8 h (24-fold increase) but was non-detectable at 24 h. Multiplebands were also observed in the 8-h sample, indicating thatJunB is rapidly synthesized and processed (phosphorylated ordegraded).

Fra2, like JunB, was negligible in granulosa cells of H andHE rats but was increased rapidly and transiently by acuteexposure to FSH (Fig. 7). However, the expression of Fra2 ingranulosa cells of preovulatory follicles was distinct from thatof JunB. Fra2 was rapidly induced by hCG, but the elevatedlevels were then sustained in luteinizing granulosa cells (HEF-hCG, 4–12 h) and in corpora lutea (HEF-hCG, 24 h; corpusluteum). In this manner, Fra2 is similar to JunD.

Immunoreactive JunD was absent in immature granulosacells of H and HE rats but was increased rapidly (2-fold) by 2 hfollowing acute injection of FSH. The increase was transient

FIG. 5. Granulosa cell AP1 factors bind an AP1 consensus DNA.WCE were prepared from undifferentiated granulosa cells culturedovernight in defined medium (0 h) or from differentiated granulosa cellsthat had been cultured in the presence of FSH/T for 48 h followed byforskolin for 24 h to stimulate luteinization. Extracts (5 mg of protein)were incubated with labeled AP1 oligonucleotide with or without unla-beled competitor DNA or were preincubated with antibody prior to theaddition of labeled probe. Protein-DNA complexes were resolved byPAGE in 0.53 TBE. Complex I contained AP1 factors, whereas theminor complex II appeared to be nonspecific. Supershifted complexesare depicted by the brackets. Note that in the differentiated cell ex-tracts, the increased binding activity appeared to be composed of JunDand Fra2, as well as c-Jun and c-Fos.

AP1 Factors in the Ovary, Regulation by FSH and LH 33723

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

since levels of JunD returned to control levels at 8 h. JunD alsoincreased rapidly in granulosa cells of HEF rats following ad-ministration of hCG. In contrast to the transient response ofimmature HE granulosa cells to FSH, the response of differen-tiated HEF granulosa cells to hCG was sustained from 4 to24 h. The 11-fold increase in JunD at 8 h was maintained in theovaries of HEF-hCG, 24-h rats that are composed mostly ofcorpora lutea.

The JunD inhibitor, menin, exhibited a different pattern ofexpression (Fig. 7). Levels of menin were low in granulosa cellsof H, HE, and HEF rats but were increased in luteinizedovaries of HEF-hCG,24 h rats, and in corpora lutea isolatedfrom pregnant rats in early pregnancy (day 7 of gestation) butlow in later pregnancy (day 22). Thus, menin is also regulatedand may modify the functional activity of JunD at specificstages of granulosa cells differentiation.

EMSAs—Changes in the relative binding of AP1 factors to aconsensus AP1 oligonucleotide exhibited patterns similar tothose observed by Western blotting. AP1 factor binding activitywas low in WCE of granulosa cells from H and HE rats (Fig.8A). AP1 binding increased rapidly (11-fold) but transiently inresponse to acute exposure to FSH at 2 h. Low levels of AP1 ingranulosa cells of preovulatory follicles (HEF; lane 5) wereincreased by acute stimulation with hCG and were then main-tained at levels between 11- and 30-fold above that observed inthe H control (Fig. 8A, lane 1).

Supershift analyses identified the AP1 factors that werepresent in granulosa cells of preovulatory HEF follicles asJunD and c-Jun (data not shown). Neither JunB nor Fra2 wasdetected (data not shown). In contrast, WCE of granulosa cellsfrom HEF, hCG 8-h rats contained c-Jun, JunD, c-Fos, andFra2 (Fig. 8B, lanes 4, 5, 7, and 9). JunB appeared to be a minorcomponent of the AP1-DNA complex (Fig. 8B, lane 3) consistentwith data obtained by Western blots (Fig. 7). FosB and Fra1were not detected (lanes 6 and 8). Similar but higher levels ofAP1 factors were present in the HEF, hCG 24-h extracts (Fig.8B, lanes 10–17).

Collectively, these results show that the hormonal regulationof AP1 factor expression in granulosa cells of growing folliclesin vivo is similar to that in cultured granulosa cells (Fig. 10).Western blots and EMSAs confirm that JunB is rapidly but

transiently induced by FSH (and forskolin) in undifferentiatedas well as by LH in differentiated cells. FosB and Fra1 arepresent in undifferentiated cells but not in luteal cells. Incontrast, Fra2 as well as c-Fos and c-Jun are rapidly buttransiently induced in undifferentiated cells but are expressedat elevated levels in luteinizing granulosa cells and luteal cells.JunD is expressed in granulosa cells of growing follicles and iselevated in luteinized cells. In vivo, JunD is also regulated byFSH and LH. Thus, JunD and Fra2 as well as c-Jun and c-Fosare likely different sets of AP1-regulated genes during thetransition of proliferating granulosa cells to terminally differ-entiated luteal cells.

Immunohistochemical Localization of AP1 Factors inOvarian Cells

Immunohistochemical data confirm the hormonal regulationand nuclear localization of JunD and Fra2 in granulosa cellsand further demonstrate regulation of these factors in thecacells (Fig. 9). JunD was present but low in ovaries of HEF rats.JunD increased in theca cells of preovulatory follicles in re-sponse to hCG at 2 h. By 4 h immunoreactive JunD wasdetected in granulosa cells and remained elevated and in nucleiof granulosa cells during luteinization. Immunoreactive Fra2(and JunB, data not shown) was negligible in granulosa cellsand theca cells of preovulatory follicles. Immunoreactive Fra2and JunB (not shown) was increased first in theca cells at 2 and4 h after hCG and then appeared in granulosa cells at 4 h afterwhich it reached maximal levels between 8 and 12 h. Immuno-reactive Fra2, unlike JunB (not shown), was present in nucleiof corpora lutea of immature rats 24 h after hCG as well as ofpregnant rats on day 7 of gestation. Thus, JunB, JunD, andFra2 are low in preovulatory follicles but are increased rapidlyin response to hCG, appearing first in the theca cells and thenin the granulosa cells. JunD and Fra2, but not JunB, remainpersistent in luteal cell nuclear suggesting that they are selec-tively associated with terminal differentiation of the granulosacells.

DISCUSSION

The FSH-, LH-, and forskolin-induced changes in the expres-sion and activation of AP1 factors in granulosa cells in vivo and

FIG. 6. Forskolin and PMA act syn-ergistically to transactivate the273COL-luciferase reporter con-struct but not other promoters. Gran-ulosa cells were cultured overnight in de-fined medium (0 h) at which time the cellswere transfected with specific promoter-reporter constructs as follows: 273COL-luciferase vector, ERE-E1b-luciferase vec-tor, Gal4(4X)-luciferase in combinationwith expression vectors for Gal4-ELK andthe 273COL-luciferase vector, and eithera menin expression plasmid or empty vec-tor. Following 6 h of transfection, the cellswere washed and stimulated with forsko-lin (Fo), PMA, or the combination (C).Data represent the mean 6 S.E. of threeseparate experiments; LSU, light specificunits.

AP1 Factors in the Ovary, Regulation by FSH and LH33724

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

in vitro indicate that the AP1 signaling pathways are impor-tant downstream targets of cAMP in granulosa cells. Since theFSH-induced AP1 complex in proliferating granulosa cells isdistinct (in composition and temporal pattern) from that of theLH-induced complex in terminally differentiated, non-dividingluteal cells, these changes likely impact specific AP1 targetgenes during this transitional period (Fig. 10).

In proliferating (undifferentiated) granulosa cells of smallfollicles, induction by FSH (forskolin) of c-Jun, JunB, c-Fos,FosB, Fra1, and Fra2 is rapid but transient; the most dramaticincreases occurred for JunB (16-fold) and Fra2 (11-fold). Theincreases in these AP1 factors relate temporally to the expres-sion of other immediate-early genes such as sgk (24) and egr-1(12, 35). JunB, c-Jun, c-Fos, and FosB were also induced ingranulosa cells by the tumor promoters, PMA and okadaic acid.That the hormone (as well as cAMP-, PMA-, and okadaic acid-induced)-induced increases in AP1 factors are transient ingranulosa cells during the normal progression of folliculargrowth may be important for preventing granulosa cell trans-formation. In numerous transformed cell lines, several AP1factors (c-Jun, JunB, Fra1, and Fra2) are expressed at highlevels; hence their designation as proto-oncogenes. Of potentialrelevance to these studies, JunB is expressed at high levels inovarian cancer cell lines (36). Moreover, the increased levels ofJunB in the transformed cells have been associated with sus-tained activation of MAPK by the Ras/Raf pathway (37). TheRas/Raf MAPK pathways may also mediate the effects of FSH

FIG. 7. FSH and LH regulate the expression of AP1 factors ingranulosa cells of growing follicles and during luteinization.Immature hypophysectomized (H) rats were treated with estradiol (E;HE), FSH (F; HEF) and hCG HEF-hCG to stimulate follicular growthand luteinization (see Fig. 1 and “Experimental Procedures”). WCEwere prepared from granulosa cells isolated from the ovaries of H andHE rats and from HE rats 2 and 8 h after an intravenous injection ofFSH or 48 h after twice daily injections subcutaneously in FSH. WCEwere also prepared from granulosa cells 2, 4, 8, 12, and 24 h afterinjection of hCG (subcutaneously). Western blots document the pres-ence and hormonal regulation of JunB, Fra2, and JunD in the extractsof granulosa cells from H rats. Menin was also regulated. For Fra2 andJunD, 20 mg of WCE protein were loaded in each lane; for JunB 35 mgof protein was needed and for menin 75 mg were used.

FIG. 8. AP1 binding activity is regulated by hormones in gran-ulosa cells in vivo. Immature hypophysectomized (H) rats weretreated with hormones to stimulate follicular growth and luteinizationas above (Figs. 1 and 7). EMSAs were run using 2.5 mg of WCE proteinand a labeled AP1 oligonucleotide as the probe (A). The AP1 factors thatwere present in WCE of HEF-hCG, 8 and 24 h granulosa cells (B) wereidentified by supershift assays. Antibodies to the AP1 factors wereadded 0.5 h prior to the addition to the labeled probe. CL, corpusluteum.

AP1 Factors in the Ovary, Regulation by FSH and LH 33725

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

on AP1 factor expression and activation in granulosa cells.Recent evidence shows that FSH/cAMP can impact the ERK,p38 MAPK, and phosphatidylinositol 3-kinase pathways (5).Importantly, the effects of FSH/cAMP on these pathways areindependent of protein kinase A. Likely, cAMP activates anewly identified group of cAMP-regulated proteins, cAMP-gua-nine nucleotide exchange factors, that activate Ras-like mole-cules (38, 39). Thus, FSH may modify AP1 factor expressionand activity by diverse signaling pathways. Although junBexhibits high sequence homology to c-jun, and like c-jun has nointrons, the promoter of junB is more complex than that ofc-jun (10). As shown herein, the hormonal induction of JunB ingranulosa cells is far more dramatic and transient than that ofc-Jun. What factors contribute to JunB induction in the ovary,

or its expression in Sertoli cells in the testis (45), remain to bedetermined. Based on regulatory regions in the junB promoter,these factors could be CREB or related family members (18),Smads (40, 41), C/EBPb (18), Stat factors (18), or a combinationof these all of which are expressed and regulated by hormonesin ovarian granulosa cells.

In contrast to JunB and c-Jun, JunD was not regulatedmarkedly by either forskolin or PMA in undifferentiated gran-ulosa cells in culture. Based on the lack of regulation in cul-tured cells, it was surprising to observe that hormones regu-lated JunD expression in vivo. Notably, FSH increased JunDtransiently in granulosa cells of small follicles, whereas LHincreased JunD in granulosa cells of preovulatory follicles, aresponse that persisted as the cells luteinized. The junD pro-

FIG. 9. Immunohistochemical localization of JunD and Fra2 in hormonally stimulated rat ovaries. Immature hypophysectomized (H)rats were treated with hormones to stimulate follicular growth and luteinization as above (Figs. 1 and 7). At selected time interval ovaries wereisolated, fixed in 4% paraformaldehyde, and embedded in paraffin. Sections (6 micron) were processed for immunohistochemistry by standardprocedures. Sections were incubated with JunD and Fra2 antibodies diluted 1:50 as well as JunB and c-Fos (data not shown). Granulosa cells (GC),theca cells (TC), and oocytes are identified. The * in each panel demarcates immunopositive cells. The magnification of the JunD photomicrographsis 3 20 and that of Fra2 is 3 40.

FIG. 10. Schematic of AP1 expres-sion patterns at specific stages ofgranulosa cell differentiation. See“Discussion.”

AP1 Factors in the Ovary, Regulation by FSH and LH33726

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

moter contains several potential regulatory elements includinga CRE capable of binding CREB, a GC-rich region with poten-tial binding sites for Sp1/Egr-1, an AP1 site, and a CAAT site towhich NF-Y binds (10). The complexity of the junD promoter,as well as the presence of multiple hormones and growth fac-tors present in vivo, may explain the more complex pattern ofJunD expression in vivo compared with in vitro.

The molecular mechanisms by which hormones, forskolin(cAMP), and PMA regulate expression of Fos family membersin proliferating (undifferentiated) granulosa cells are equallydiverse (10, 12, 17, 37). For example, although the gene struc-ture and promoter regulatory regions of the c-fos and fosBgenes are highly conserved, their regulation in granulosa cellsby cAMP and the tumor promoters differs markedly. In theundifferentiated granulosa cells, c-Fos increased more in re-sponse to forskolin (and PMA) than did FosB. In contrast, bothwere increased by okadaic acid. The cAMP response elements(CREs) and AP1 elements within the c-fos promoter are likelytargets of cAMP induction in granulosa cells (15, 17). In con-trast, the SRE of the c-fos promoter is likely to be the target ofthe PMA response, especially since PMA, but not forskolin/cAMP, was a potent stimulator of transactivation of the Elk-Gal and Gal-luciferase reporter system in granulosa cells.These two different promoter sites and the preferential activa-tion of protein kinase A by FSH and ERK by PMA appear tocontribute to the additive effects of forskolin and PMA oninduction of c-Fos in the undifferentiated cells. The mecha-nisms by which Fra2 is induced clearly involves cAMP andprotein kinase A since H89 completely blocked induction ofFra2 by FSH or forskolin (data not shown) confirming thestrong induction of Fra2 by cAMP in other cell types (10).

The LH/hCG (forskolin)-induced expression patterns of theAP1 factors in differentiated granulosa cells of preovulatoryfollicles differed markedly from that in undifferentiated cells.Despite the similarities of the c-fos and fosB promoters, expres-sion of FosB but not c-Fos was turned off in response to hCG. Inmarked contrast, induction of c-Jun and c-Fos, as well as JunDand Fra2, was sustained and stable as granulosa cells termi-nally differentiated to luteal cells. This change from transientto stable expression of JunD and Fra2 suggests that they exertspecific functions during granulosa cell proliferation and dif-ferentiation. That JunD is a functional component of the AP1-DNA binding complex in differentiated granulosa cells sug-gests that it is an important regulator of specific AP1-responsive genes. In other systems, JunD can act as aninhibitor or an activator of transcription, dependent on thecomposition of the heterodimeric complex, the promoter ele-ment, and the cell type. For example, c-Fos/c-Jun or c-Fos/JunB, but not c-Fos/JunD, activates the 273COL-luciferaseconstruct in vascular endothelial cells (42). In contrast, a Fra2-JunD complex was more effective than c-Fos/c-Jun in activat-ing the promoter of the oncostatin gene in ROS 17/2.8 osteo-sarcoma cells (43). Thus, based on the composition of AP1factors in differentiating granulosa cells, JunD/Fra2 het-erodimers may activate one set of AP1-responsive genes,whereas c-Fos-c-Jun complexes regulate other AP1-responsivegenes. Furthermore, other factors such as menin, a specificJunD inhibitor, can alter the activity of JunD. In support ofthis, co-expression of menin blocked forskolin-induced tran-scription of 273COL-luciferase, suggesting that JunD is afunctional component of the AP1 complex in granulosa cells. Interminally differentiating granulosa cells, JunD may play acritical role in terminating cell proliferation. Specifically, JunDhas been shown to be low in certain ovarian cancer cell lines,and overexpression of JunD in these cells can suppress cellgrowth in a cell line-specific manner. JunD and Fra2 have

recently been shown to increase during osteoblast differentia-tion (43). Thus, the increase in JunD and Fra2 in granulosacells of preovulatory follicles exposed to an LH surge maycontrol the transcription of specific genes that regulate the exitof granulosa cells from the cell cycle, thereby terminating gran-ulosa cell proliferation (1, 2). Of note, JunD is low in the testis(19), perhaps because the number of non-proliferative, JunD-positive Sertoli cells (data not shown) is low compared with thenumber of proliferating germ cells.

There are likely to be many genes regulated in granulosacells by members of the Jun/Fos family of transcription factors.First, AP1 factors appear to control their own expression (10,15). Other genes in which AP1 factor regulation has beendetermined include inhibin bA (11), the GnRH receptor (26),TIMP-1 (44, 45), and as already mentioned p21cip (1, 2, 14).Each of these genes is hormonally regulated in granulosa cells,but only the promoter of the inhibin bA gene has been specif-ically examined in granulosa cells. The inhibin bA promotercontains a variant CRE that binds AP1-like factors and isinducible by forskolin and PMA as well as by overexpression ofJunB and FosB proteins in granulosa cells (11). These syner-gistic effects of forskolin and PMA on the inhibin bA promotermimic that observed herein for activation of the AP1 site withinthe human collagenase gene. Synergism between cAMP andPMA has also been observed recently to control induction ofprogesterone receptor mRNA in granulosa cells (46). Based onthe results described herein, the synergy in granulosa cellscould involve the selective induction by PMA of c-Jun and c-Fosand by cAMP of Fra2, JunB, and JunD. In addition to theinduction of AP1 factors, forskolin and PMA activate severalkinase cascades (protein kinase A, p38 MAPK, protein kinaseC, and ERKs) that would lead to the phosphorylation andactivation of specific AP1 factors and their co-activators. Forexample, c-Jun but not JunB is a target of N-terminal c-Junkinase; c-Fos is activated by Frk (15). Finally, AP1 factors mayregulate the expression of ovarian genes by indirect mecha-nisms that involve protein-protein interactions with other tran-scription factors. Of note, c-Jun has been shown to interactwith the cell cycle regulator Rb (retino blastoma protein) (47).c-Jun also interacts with CBP to activate CREB (21), withSp1/Sp3 to activate p21cip1 gene (14), and with Smads (20) toregulate specific chimeric reporter genes. Thus, hormonal reg-ulation of AP1 factors in the ovary has far reaching effects onmany cellular signaling cascades that regulate proliferationand differentiation. The studies herein provide the specificobservation that JunD and Fra2, as well as c-Jun and c-Fos,may be critical for regulating specific sets of AP1-responsivegenes that control the terminal differentiation of granulosacells, their exit from the cell cycle, and the prevention of gran-ulosa cell transformation associated with elevated levels ofAP1.

Acknowledgments—We thank Dr. Philip Stork (Oregon Health Sci-ences Center) for the Gal-4-luciferase reporter construct and Elk-Galexpression vector; Dr. Michael Karin (UCSD) for the 273COL-lucifer-ase vector; and Dr. Sunita Agarwal (NIDDK, National Institutes ofHealth) for the menin vectors.

REFERENCES

1. Robker, R. L., and Richards, J. S. (1998) Biol. Reprod. 59, 476–4822. Robker, R. L., and Richards, J. S. (1998) Mol. Endocrinol. 12, 924–9403. Richards, J. S. (1994) Endocr. Rev. 15, 725–7514. Gonzalez-Robayna, I. J., Alliston, T. N., Buse, P., Firestone, G. L., and

Richards, J. S. (1999) Mol. Endocrinol. 13, 1318–13375. Gonzalez-Robayna, I. J., Falender, A. E., Ochsner, S., Firestone, G. L., and

Richards, J. S. (2000) Mol. Endocrinol. 14, 1283–13006. Habener, J. F., Miller, C. P., and Vallejo, M. (1995) Vitam. Horm. 51, 1–577. Carlone, D. L., and Richards, J. S. (1997) Mol. Endocrinol. 11, 292–3048. Pei, L., Dodson, R., Schoderbek, W. E., Maurer, R. A., and Mayo, K. E. (1991)

Mol. Endocrinol. 5, 521–5349. Mukherjee, A., Park-Sarge, O.-K., and Mayo, K. E. (1996) Endocrinology 137,

3234–3245

AP1 Factors in the Ovary, Regulation by FSH and LH 33727

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

10. Herdegen, T., and Leah, J. D. (1998) Brain Res. Rev. 28, 170–19011. Ardekani, A. M., Romanelli, J. C. D., and Mayo, K. E. (1998) Endocrinology

139, 3271–327912. Morganm, J. I., and Curran, T. (1991) Annu. Rev. Neurosci. 14, 421–45113. Vojtek, A. B., and Der, C. J. (1998) J. Biol. Chem. 273, 19925–1992814. Kardassis, D., Papakosta, P., Pardali, K., and Moustakas, A. (1999) J. Biol.

Chem. 274, 29572–2958115. Karin, M. (1995) J. Biol. Chem. 270, 16483–1648616. Morgan, J. I., and Curran, T. (1995) Trends Neurosci. 18, 66–6717. Robertson, L. M., Kerppola, T. K., Vendrell, M., Luk, D., Smeyne, R. J.,

Bocchiaro, C., Morgan, J. I., and Curran, T. (1995) Neuron 14, 241–25218. Tjin Tham Sjin, R. M., Lord, K. A., Abdollahi, A., Hoffman, B., and

Liebermann, D. A. (1999) J. Biol. Chem. 274, 28697–2870719. Hirai, S.-I., Ryseck, R.-P., Mechta, F., Bravo, R., and Yaniv, M. (1989) EMBO

J. 8, 1433–143920. Zhang, Y., Feng, X.-H., and Derynck. R. (1998) Nature 394, 909–91321. Hu, P. P.-C., Harvat, B. L., Hook, S. S., Shen, X., Wang, X.-F., and Means, A. R.

(1999) Mol. Endocrinol. 13, 2039–204822. Johnson, R. S., Lingen, V., Papaioannou, V., and Spiegelman, B. M. (1993)

Genes Dev. 7, 1309–131723. Johnson, R. S., Spiegelman, B. M., and Papaioannou, V. (1992) Cell 71,

577–58624. Alliston, T. N., Gonzalez-Robayna, I. J., Buse, P., Fireston, G. L., and Richards,

J. S. (2000) Endocrinology 141, 385–39525. Segaloff, D. L., Wang, H., and Richards, J. S. (1990) Mol. Endocrinol. 4,

1856–186526. White, B. R., Duval, D. L., Mulvaney, J. M., Roberson, M. S., and Clay, C. M.

(1999) Mol. Endocrinol. 13, 566–57727. Fitzpatrick, S. L., and Richards, J. S. (1994) Mol. Endocrinol. 8, 1309–131928. Angel, P., Baumann, I., Stein, B., Delius, H., Rahmsdorf, H. J., and Herrlich,

P. (1987) Mol. Cell. Biol. 7, 2256–226629. Sharma, S. C., Clemens, J. W., Pisarska, M. D., and Richards, J. S. (1999)

Endocrinology 140, 4320–433430. Vossler, M. R., Yao, H., York, R. D., Pan, M.-G., Rim, C. S., and Stork, P. J. S.

(1997) Cell 89, 73–8231. Agarwal, S. K., Guru, S. C., Heppner, C., Erdos, M. R., Collins, R. M., Park,

S. Y., Saggar, S., Chandrasekharappa, S. C., Collins, F. S., Spiegel, A. M.,Marx, S. J., and Burns, A. L. (1999) Cell 96, 143–152

32. Ginty, D. D., Kornhauser, J. M., Thompson, M. A., Bading, H., Mayo, K. E.,Takahashi, J. S., and Greenberg, M. E. (1993) Science 260, 238–241

33. Rosenberger, S. F., Finch, J. S., Gupta, A., and Bowden, G. T. (1999) 274,1124–1130

34. Smith, C. L. (1998) Biol. Reprod. 58, 627–63235. Espey, L. L., Ujoka, T., Russell, D. L., Skelsey, M., Vladu, B., Robker, R. L.,

Okamura, H., and Richards, J. S. (2000) Endocrinology 141, 2385–239136. Neyns, B., Teugels, E., Bourgain, C., Birrer, M., and DeGreve, J. (1999) Int. J.

Cancer 82, 687–69337. Cook, S. J., Aziz, N., and McMahon, M. (1999) Mol. Cell. Biol. 19, 330–34138. Kawasaki, H., Springett, G. M., Mochizuki, N., Toki, S., Nakaya, M., Matsuda,

M., Housman, D. E., and Graybiel, A. M. (1998) Science 282, 2275–227939. de Rooij, J., Zwartkruis, F. J. T., Verheijen, M. H. G., Cool, R. H., Nijman,

S. M. B., Wittinghofer, A., and Bos, J. L. (1998) Nature 396, 474–47740. Hashimoto, M., Gaddy-Kurten, D., and Vale, W. (1993) Endocrinology 133,

1934–194041. Jonk, L. J. C., Itoh, S., Heldin, C.-H., ten Dijke, P., and Kruijer, W. (1998)

J. Biol. Chem. 273, 21145–2115242. Rao, G. N., Katki, K. A., Madamanchi, N. R., Wu, Y., and Birrer, M. J. (1999)

J. Biol. Chem. 274, 6003–601043. McCabe, L. R., Banerjee, C., Kundu, R., Harrison, R. J., Dobner, P. R., Stein,

J. L., Lian, J. B., and Stern, G. S. (1996) Endocrinology 137, 4398–440844. Botelho, F. M., Edwards, D. R., and Richards, C. D. (1998) J. Biol. Chem. 273,

5211–521845. Hagglund, A. C., Ny, A., Leonardsson, G., and Ny, T. (1999) Endocrinology

140, 4351–435846. Richards, J. S., Robker, R. L., Russell, D., Sharma, S. C., Espey, L. E., Lydon,

J., and O’Malley, B. W. (2000) Steroids, in press47. Nishitani, J., Nishinaka, T., Cheng, C.-H., Rong, W., Yokoyama, K. K., and

Chiu, R. (1999) J. Biol. Chem. 274, 5454–5461

AP1 Factors in the Ovary, Regulation by FSH and LH33728

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from

S. Chidananda Sharma and JoAnne S. RichardsDIFFERENTIATION

Granulosa Cells: RELATION OF JunD AND Fra2 TO TERMINAL Regulation of AP1 (Jun/Fos) Factor Expression and Activation in Ovarian

doi: 10.1074/jbc.M003555200 originally published online August 8, 20002000, 275:33718-33728.J. Biol. Chem. 

  10.1074/jbc.M003555200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/275/43/33718.full.html#ref-list-1

This article cites 45 references, 13 of which can be accessed free at

by guest on January 10, 2021http://w

ww

.jbc.org/D

ownloaded from