BIOLOGY OF REPRODUCTION 78, 1038–1048 (2008)Published online before print 20 February 2008.DOI 10.1095/biolreprod.107.066340

Regulated Expression of ADAM8 (a Disintegrin and Metalloprotease Domain 8) in theMouse Ovary: Evidence for a Regulatory Role of Luteinizing Hormone, ProgesteroneReceptor, and Epidermal Growth Factor-Like Growth Factors1

Venkataraman Sriraman,2 Ursula Eichenlaub-Ritter,3 Jorg W. Bartsch,4 Andrea Rittger,5 Sabine M.Mulders,6 and JoAnne S. Richards7

Institute of Genetechnology/Microbiology,3 University of Bielefeld, D-33501 Bielefeld, GermanyDepartment of Biochemistry,4 King’s College London, London SE1 9NH, United KingdomBiochemical Institute,5 University of Kiel, 24098 Kiel, GermanyN.V. Organon,6 5340 BH Oss, The NetherlandsDepartment of Molecular and Cellular Biology,7 Baylor College of Medicine, Houston, Texas 77030

ABSTRACT

ADAM8 (a disintegrin and metalloprotease domain 8) isexpressed in immune, neuronal, and bone progenitor cells and isthought to be involved in the tissue-remodeling process. Micro-array analyses indicate that Adam8 is a potential target of theprogesterone receptor (Pgr) in murine ovary. Further studiesdocument that Adam8 mRNA and protein are expressed ingranulosa cells and cumulus cells of periovulatory follicleswhereas expression is significantly reduced in Pgr null mice thatfail to ovulate. There is a reduced expression in granulosa cellsfrom cultured, in vitro ovulated follicles exposed to inhibitors ofprogesterone or epidermal growth factor signaling whileepiregulin induced its expression in the absence of hCG. Invitro studies with primary mouse granulosa cells document thatAdam8 is induced in response to forskolin (Fo) and phorbol ester(PMA) or Fo and Amphiregulin treatment. To understand thetranscriptional regulation of the Adam8, we amplified 1 kb of themouse Adam8 promoter by PCR and subcloned it into a pGL3-luciferase reporter construct. The Adam8 promoter-luciferaseconstructs are induced by Fo and PMA treatment aftertransfection into rat granulosa cells, and cotransfection with aPGR-A expression vector further augment basal and Fo/PMAinducibility. Site-specific mutations within the �615/þ50promoter document that a GC-rich region, NF-1 (nuclearfactor-1) site, and putative TATA box are critical for Adam8promoter activation by Fo/PMA. Thus, ADAM8 is expressed in astage-specific manner and is hormonally regulated in ovulatingfollicles by the coordinate actions of LH and PGR. To ourknowledge, ADAM8 is the first member of the ADAM familyshown to be hormonally regulated.

luteinizing hormone, ovary, ovulation, progesterone receptor

INTRODUCTION

Ovulation is a complex event in mammalian reproductionand resembles an inflammatory reaction [1]. This process isassociated with significant changes in the extracellular matrixof the ovary that control the release of the oocyte from themature preovulatory follicle [1, 2]. Matrix-modifying enzymes,namely, cathepsins [3, 4], ADAMTSs (a disintegrin andmetalloprotease domain with thrombospondin motifs) [5],matrix metalloproteinases [6], and plasmin [7], have beenimplicated in these remodeling events. The nuclear transcrip-tion factor progesterone receptor (Pgr) is critical for ovulation,and mice null for Pgr do not ovulate and are infertile due tosevere defects in the ovary, uterus, and mammary gland [8].The PGR-A isoform is predominantly expressed in granulosacells, and one prevailing hypothesis is that PGR regulatesfollicular rupture by induction of matrix-modifying enzymessuch as ADAMTS1, cathepsin L (Ctsl ) [9], and plasminogenactivator system [7]. In our studies of Pgr null mice (PRKO),Adam8 (a disintegrin and metalloprotease domain 8) wasidentified as a putative PGR target gene, and this suggests thatit may be expressed in a stage-specific manner by PGR action.

The ADAMTS1 protein is one of the first members of themetzincin superfamily that was shown to be essential forcumulus expansion [10]. Targeted disruption of the Adamts1gene leads to subfertility in mice [11]. ADAMTS1 is inducedby the progesterone-signaling pathway [9], while the role andcontrol of the other members of the ADAMs family in theovary remain unknown. ADAM proteins constitute a family oftransmembrane glycoproteins with essential physiological rolesin fertilization, neovascularization, myogenesis, and neurogen-esis [12, 13]. They are composed of several conserved andcharacteristic structural modules, including a prodomain,metalloprotease domain, disintegrin domain, cysteine-richregion, EGF (epidermal growth factor) repeat transmembranedomain, and cytoplasmic tail [12, 13]. Distinct structuraldomains within these proteins confer specific functions in cell-cell fusion, cell-cell interactions, or metalloprotease activity[12, 13].

ADAM8 was originally cloned from mouse macrophages[14] and is expressed in neurons and oligodendrocytes [15],and, with the exception of T-cells, in most human immune celltypes [13]. ADAM8 contains the catalytic consensus sequenceHEXXHXXGXXHD required for the metalloproteinase activ-ity, and the recombinant soluble ADAM8 possesses catalyticactivity [16]. In vitro, it can cleave myelin basic protein (MBP),amyloid polypeptide precursor protein, and FCER2 (Fcfragment of IgE receptor; CD23), a low-affinity IgE receptor

1Supported by NIH grants HD16229 and HD07495 (SCCPIR) (J.S.R.)and by the King’s College (J.W.B.). V.S. is supported in part by theNICHD and Office of Research on Women’s Health as a BuildingInterdisciplinary Research Careers in Women’s Health ScholarK12HD052023 and ASRM/Organon grant.2Correspondence: Venkataraman Sriraman, Department of InternalMedicine,3 Division of Endocrinology, University of Texas MedicalBranch, Galveston, Texas 77555. FAX: 409 772 8709;e-mail: vesriram@utmb.edu

Received: 25 October 2007.First decision: 5 November 2007.Accepted: 1 February 2008.� 2008 by the Society for the Study of Reproduction, Inc.ISSN: 0006-3363. http://www.biolreprod.org

1038

that promotes cell motility and leukocyte infiltration [17].Recently the neural cell adhesion molecule close homologue ofL1, termed CHL1 [18], that promotes neurite outgrowth andsuppresses neuronal cell death has been shown to be a substrateof ADAM8. Expression of catalytically active ADAM8 isassociated with increased invasive activity and/or extracellularmatrix (ECM) remodeling [19]. Thus, ADAM8 can influenceinflammation, cell adhesion, and migration processes that arerelevant to the ovulatory process.

The goal of this study was to understand the hormonalregulation of ADAM8 expression in ovary during folliculardevelopment, ovulation, and luteinization, employing a numberof different experimental approaches to study control ofexpression of Adam8. These included analysis of expressionof Adam8 in ovary of in vivo ovulated follicles and inhormonally stimulated control and PRKO mice. Furthermore,expression was analyzed in granulosa cells from preantralfollicle cultures stimulated by hCG (human chorionic gonad-otropin) and EGF to ovulate in vitro, with and withoutexposure to inhibitors of PGR and EGF receptor during thepreovulatory period. Finally, we studied luciferase expressionfrom a transgene carrying different segments of the Adam8promoter in mammalian primary granulosa cells stimulated byforskolin and phorbol ester that activates protein kinase A andprotein kinase C signaling pathways. Although the function ofAdam8 in the preovulatory follicle in vivo remains elusive, ourstudies document the induction of Adam8 mRNA and proteinin mammalian granulosa cells and of mRNA in cumulus-oocyte-complexes (COCs) of the preovulatory follicle and arole for LH and PGR in the induction of this gene.

MATERIALS AND METHODS

Reagents

Equine chorionic gonadotropin (eCG) and forskolin (Fo) were purchasedfrom Calbiochem (San Diego, CA) and hCG from American PharmaceuticalPartners Inc. (Schaumburg, IL). Recombinant luteinizing hormone (rLH) andrecombinant follicular stimulating hormone (rFSH) were from Serono(Unterschleißheim, Germany, kindly donated), and recombinant epidermalgrowth factor (rEGF) was from Promega (Mannheim, Germany). Media andcell culture reagents and materials were procured from Life Technologies, Inc.(Grand Island, NY) and Corning Inc. (Corning, NY); fetal bovine serum wasprocured from Hyclone Laboratories, Inc. (Logan, UT). L15 Leibovitz-glutamax medium, fetal calf serum, and a-minimal essential medium withglutamax for preantral follicle culture were from GIBCO Invitrogen (Karlsruhe,Germany). Insulin-transferrin-selenium mixture and progesterone receptorinhibitor RU486 were purchased from Sigma (Taufkirchen, Germany), andAG1478 inhibitor of epidermal growth factor receptor (EGFR) was fromCalbiochem. Epiregulin came from R & D Systems (Wiesbaden, Germany).Trypsin, soybean trypsin inhibitor, deoxyribonuclease, and PMA (phorbol 12-

myristate 13-acetate) were obtained from Sigma Chemical Co. (St. Louis, MO),electrophoresis and molecular biology-grade reagents were obtained from Bio-Rad Laboratories, Inc. (Richmond, CA), and reagents for RT-PCR wereobtained from Promega Corp. (Madison, WI). Amphiregulin (AREG) wasobtained from R&D Systems, Inc. (Minneapolis, MN). Oligonucleotides weresynthesized by Sigma-Genosys (Houston, TX). The generation of rabbitpolyclonal antibodies against ADAM8 has been described [15], and itsapplication in Western blots have been demonstrated earlier [15, 16].

Animals and Hormone Treatment

Immature (23 days old) C57BL/6 mice were obtained from Harlan SpragueDawley, Inc. (Indianapolis, IN) and housed under a 16 h light and 8 h darkschedule in the Center for Comparative Medicine at Baylor College ofMedicine with food and water ad libitum. Mice were injected i.p. with 4 IUeCG to stimulate follicular growth, and after 44 h with 5 IU hCG (i.p.), which,like LH, induces ovulation and luteinization. Ovulation occurs approximately12–16 h after hCG administration in this model [20]. Ovaries were isolatedfrom these hormone-stimulated mice at selected time intervals for extraction ofRNA and protein, or fixed for in situ hybridization. Pgr null mice were used inselected experiments because follicles develop normally in response to eCG butfail to ovulate in response to hCG [9]. Cumulus-oocyte-complexes wereisolated from eCG treated mice (0 h) as well as from preovulatory andovulating follicles of eCG- and hCG-stimulated mice (8 and 12 h). OvulatedCOCs were collected from oviduct at 16 and 24 h.

C57BL/6 mice for preantral follicle culture were maintained at BielefeldUniversity under a 12 h light and 12 h dark cycle with food and water adlibitum, and animal experimentation was conducted under a permit from theuniversity authorities.

Rat granulosa cells were employed for transfection studies to analyzeAdam8 promoter. Intact immature (23 days old) Holtzman Sprague Dawleyfemales (Harlan Sprague Dawley, Inc., Indianapolis, IN) were housed under a16 h light and 8 h dark schedule in the Center for Comparative Medicine atBaylor college of Medicine with food and water ad libitum. The rats wereinjected with estradiol-17b (E

2) for 3 days (1.5 mg E

2/0.2 ml in propylene

glycol), and the ovaries were harvested for isolation of granulosa cells. AllAnimals were treated in accordance with the National Institutes of HealthGuide for the Care and Use of Laboratory Animals, as approved by the AnimalCare and Use Committee at Baylor College of Medicine (Houston, TX).

RNA Isolation from Whole Ovary, Granulosa Cells, andCOCs, and RT-PCR

Total RNA was isolated from whole ovaries, granulosa cells, and COCs ofimmature (23 days old) untreated mice, and from eCG- and hCG-treated mice,using TRIzol reagent (Life Technologies, Inc.) and purified as described in themanufacturer’s instructions. For each RNA sample, whole ovaries, granulosacells, or COCs were pooled separately from two or three animals. RNA wassimilarly prepared from cultured cells. To determine the expression of Adam8in ovary, 2 lg of RNA were reverse transcribed in 20 ll reaction containingoligo-deoxythymidine (Amersham Pharmacia Biotech, Newark, NJ) and AMV(avian myeloblastosis virus) reverse transcriptase (Promega Corp.) at 428C for75 min [20]. cDNA (2 ll) obtained from reverse transcription reaction wassubjected to labeled PCR with [P32]deoxy-CTP (ICN, Los Angeles, CA) usingTaq polymerase [20] and specific primers for Adam8 and Rpl19 (Table 1). All

TABLE 1. List of primers employed for RT-PCR and cloning promoter constructs.

Primer name Sequence

RT-PCRAdam8 Forward 50-aggatattcagcaggtgtagcaa-30

Adam8 Reverse 50-tgctaaaggtatagcaggagtcg-30

Adam8 promoter cloning PCRAdam8 Forward (�1065) 50-gagct agcgttcccacagtaacagac-3 0

Adam8 Forward (�615) 50-gagct agc agga tcag tca gg ttcc-30

Adam8 Forward (�240) 50-gagctagc ggg aga tcaataggagc-30

Adam8 Forward (�106) 50-gagct agcagctacctagaagttacttcg-3 0

Adam8 Reverse (þ50) 50-gactcgagtgatgggatctggagc-3 0

Site directed mutagenesisTATA mutant 50-tggg tgttggtgaagaggtgtgtacttctt-30

NF-1 mutant 50-ccctggaggctgaaagctgccaa-30

GC box I mutant 50-agac tgccctggattctggttgctgcc-30

TATA/NF-1/GC box I mutant 50-ctgccctggattctgaaagctgccaac tgggtgttggtgaagaggtgtgtac-30

REGULATION OF Adam8 GENE EXPRESSION IN OVARY 1039

PCRs were carried out within the experimentally determined linear amplifi-cation range. The ribosomal protein L19 (Rpl19) was employed as an internalcontrol. PCR products were resolved by a 6% PAGE and visualized byautoradiography [20]. The products were quantified using PhosphoImager(Molecular Dynamics, Inc., Sunnyvale, CA) [20]. All PCR products werecloned in TOPO vector, and their authenticity was confirmed by sequencing atthe Baylor College of Medicine Core Facility.

In Situ Hybridization

In situ hybridization was performed as described previously and as reportedby our laboratory [20]. The experiments were performed with four ovarianserial sections obtained from three separate hormonal treatments for each timepoint. Eight to ten follicles at each stage were assessed for hybridization. Theriboprobe in vitro transcription system kit (Promega Corp.) was used to make[S35] uridine triphosphate-labeled antisense and sense probes from the mouseAdam8 cDNA. The cDNA probe were generated by RT-PCR amplification andTOPO TA cloning (Invitrogen, Carlsbad, CA); their identity was verified bysequencing, and their specificity was verified by basic local alignment andsearch tool analyses. Ovaries were fixed in 4% paraformaldehyde, embedded inparaffin, and sectioned at 7 lm onto Fisher Brand Superfrost Plus microscopeslides (Fisher Scientific, Pittsburgh, PA). Tissue sections were rehydrated,treated with 20 lg/ml proteinase K and 0.1 M triethanolamine/acetic anhydride,

dehydrated, and incubated with radiolabeled riboprobe overnight at 558C. Thenext day, slides were washed at high stringency, dried, and exposed to X-OMAT film (Kodak, Rochester, NY) overnight to determine the specificity andintensity of the probe. Slides were dipped in photographic NTB-2 emulsion,exposed at 48C for an appropriate length of time, developed with D-19developer and fixer (Kodak), and stained with hematoxylin. Tissue histologywas observed by both light-field illumination and dark-field illumination tovisualize the regions of hybridization.

Western Blot Analyses

Protein extracts from mouse ovaries were prepared by homogenization in 6M urea and 0.1% Triton buffers containing a protease inhibitor cocktail (Sigma,St. Louis, MO) and 0.05 M EDTA [21]. Protein concentrations weredetermined by the Bradford method, and for Western Blot analyses ofADAM8, 20 lg protein were loaded onto gels. Samples were resolved in 10%acrylamide gels under reducing SDS-PAGE conditions and transferred topolyvinylidene difluoride membrane (Immobilon-P, Millipore Corp., Bedford,MA). Membranes were blocked with 5% nonfat milk for 2 h, followed byovernight incubation with ADAM8 antiserum, total ERK (extracellular signal-related kinase) antibody (Cell Signaling Technology, Danvers, MA), orpreimmune serum obtained before immunization of ADAM8 peptide at 1:1000dilution and then washed with TBST (10 mM Tris, pH 7.5, 150 mM NaCl, and

FIG. 1. Regulated expression of Adam8 in ovary. The hormonal regulation Adam8 expression was determined in hormone-primed mice ovaries afteradministration of eCG (44 h; 4IU) and hCG (5 IU; 2, 4, 8, 12, 16, 24, and 48 h) as described in Materials and Methods by semiquantitative RT-PCR (A) andWestern blot analyses (B). Specifically after eCG and hCG treatments, ovaries were harvested and total RNA were isolated using the TRIzol reagent; 2 lg ofthe RNA were subjected to reverse transcription employing AMV-RT and radioactive PCR employing P32-labeled dCTP with 2 ll of the cDNA with specificprimers for Adam8 and Rpl19 within the linear range of amplification. The products were resolved by 6% acrylamide gel and assessed by autoradiography/phosphorimaging. Total ovarian protein was extracted using urea-Triton buffer with protease inhibitors. Twenty micrograms of the protein was resolved ina SDS-PAGE for Western blot analyses with rabbit polyclonal antibodies for ADAM8 at a dilution of 1:1000. The immunoreactive bands were visualizedusing an enhanced chemiluminescent system. The same blot was also probed with a-actin antibody (loading control). Total ERK antibody (loading control)or preimmune serum was also used to probe the same protein extracts on separate Western blots. The proteolytically active, processed form of the ADAM8(;90 kDa) was quantitated from three different blots and represented in the form of a graph (***P , 0.001, Imm vs. eCG, hCG, 4 or 8 h).

1040 SRIRAMAN ET AL.

0.1% Tween 20). Subsequently, blots were incubated with 1:12,500 dilutedhorseradish peroxidase linked to anti-rabbit IgG (Amersham PharmaciaBiotech, Piscataway, NJ) and then washed with TBST. Immunoreactive bandswere detected using enhanced chemiluminescence according to the manufac-turer’s specifications (Pierce Chemical Co., Rockford, IL). The blots that wereprobed with ADAM8 antibody were also probed a-actin antibody (Sigma, St.Louis, MO). The assessment of a-actin and total ERK expression served as aloading control. The immunoreactive bands were quantitated by densitometer(Molecular Dynamics, Inc., Sunnyvale, CA) and Image quant 5.0 software.

Mouse Preantral Follicle Culture

Preantral follicles were manually isolated from 16 to 18 day old C57BL/6mice. Cell culture was essentially as previously described [22–25]. Briefly,preantral follicles of 100–130 lm diameter with two to three layers of granulosacells and some theca cells (3b follicles) [26] were mechanically isolated in L15Leibovitz-glutamax medium with foetal calf serum and antibiotics. Individualfollicles were placed into 50 ll of culture medium consisting of a-minimalessential medium with glutamax, supplemented with 10 lIU/ml rLH, 10 lIU/ml rFSH, foetal calf serum, and an insulin-transferrin-selenium mixture in 96microtiter plates (Costar, Schiphol-Rijk, Netherlands) covered by a layer of 30ll mineral oil. At each fourth day of culture, 20 ll medium containing 10 lIUof rFSH but no rLH was replenished. At Day 12 when antral-like cavities haddeveloped and oocytes had grown and reached maturational competence,oocyte resumption of maturation and in vitro ovulation was induced by adding1.5 IU/ml recombinant hCG (rhCG) and 5ng/ml rEGF. Four and 18 h later,when oocytes had undergone germinal vesicle breakdown (GVBD) or reachedmetaphase II and emitted a first polar body (PB) and when in vitro ovulationhad occurred, respectively, the mural granulosa cells were collected for RNApreparation and analysis of expression of Adam8 by quantitative RT-PCR.

In order to study the influence of PGR- or EGFR-mediated signaling onAdam8 expression in follicles stimulated for in vitro ovulation, 1 lM of RU486(PGR inhibitor) or 10 lM AG1478 (EGFR inhibitor) were added to the antral

follicles together with rhCG and rEGF on Day 12 of culture. In one set ofexperiments, 100 ng/ml epiregulin was added on Day 12 instead of rhCG andrEGF to induce in vitro ovulation.

Oocytes retrieved from in vitro ovulated follicles were vital stained forchromatin with Hoechst 33342 (Sigma, Germany; 20 min in M2 medium) andanalyzed for induction of GVBD or PB formation with a Zeiss (Axiovert 10).Images of in vitro ovulated COCs or isolated oocytes were obtained using aCanon PC1099 camera attached to the Zeiss Axiovert microscope.

RNA Isolation and Quantitative RT-PCR from GranulosaCells of In Vitro Grown Follicles

About 100 ng of total RNA isolated from follicles were transcribed withExpand Reverse Transcriptase (Roche, Mannheim, Germany) in a volume of 20ll; 2 ll of the RT-product was used for PCR in a 20 ll reaction volume withprimers indicated in Table 1. Real-time PCR consisted of a 20 ll reaction withQuantiTecti SYBR Green PCR Kit (Qiagen, Hilden, Germany), and 2 ll of theRT product was used as the template. Specific Adam8 transcript numbers werenormalized to the ribosomal L7 gene (Rpl7) as described elsewhere [19] andpresented as normalized values relative to controls (control¼ 1). Experimentswith control and inhibitors were performed in triplicate with isolation offollicles for RNA isolation from control and treatment group of eachexperimental group. Granulosa cells were pooled from 15 to 20 follicles fromeach plate. Relative increase or decease in Adam8 mRNA/amplification productwas calculated relative to the untreated controls in the studies with inhibitor orin the epiregulin study, comparing data from triplicate RT-PCR assays.

Adam8 Promoter Constructs and Mutants

The mouse Adam8 promoter (�1065/þ50) was amplified using specificprimers containing Nhe and XhoI sites and ligated to the pGL3 luciferasereporter plasmid using PCR cloning techniques. PCR primers were designed

FIG. 2. Expression of Adam8 in granulosa cells and COCs. The spatiotemporal expression of Adam8 was determined by in situ hybridization (A) inovarian sections (S An, small antral follicle; PO, preovulatory follicle) and RT-PCR analyses (B). Ovarian paraffin sections obtained from hormone-primedmice ovaries were assessed for Adam8 mRNA expression employing a specific riboprobe generated by RT-PCR. Granulosa cells were isolated from ovariesof eCG- and hCG-treated mice, or whole ovarian samples were collected (24-h post-hCG). COCs were collected from the ovary or oviduct (16 and 24 h)to determine Adam8 expression after hormone treatment. Total RNA was extracted and subjected to RT-PCR analysis. The products were resolved in a 6%acrylamide gel and quantified by phosphorimaging (R, residual ovarian tissue). Duplicates of triplicate RT-PCR reaction products per treatment areshown. (***P , 0.001, eCG vs. eCG, hCG, 8 h.)

REGULATION OF Adam8 GENE EXPRESSION IN OVARY 1041

based on previous studies that characterized the Adam8 gene and putativetranscription factor binding sites. In addition, three other Adam8 promoterregions that extend from �615 /þ50.�240/þ50, and �106/þ50 that span thetranscription start site þ1 were amplified and subcloned. The authenticity ofeach construct was verified by sequence analysis in both directions. Site-specific mutations in the Adam8 promoter constructs were generated using thegene editor site-directed mutagenesis kit from Promega Corp. (Madison, WI).Oligonucleotides used for creating site-specific mutations at the criticalnucleotides necessary for transcription factor binding to the GC boxes, NF-1(nuclear factor-1) site, and the TATA box are listed in Table 1. The mutantsobtained were confirmed by sequencing.

Granulosa Cell Culture and Transfection

Rat granulosa cells were harvested by needle puncture of the ovaries andprocessed described previously [3]. Briefly, cells were cultured in 12-well cultureplates at a density of 0.5 3 106 cells/1.5 ml in 5% serum containing medium(DMEM:F12 containing penicillin and streptomycin). Two hours after plating,cells were transiently transfected with 500 ng of the respective promoter-reporterconstructs with Fugene 6 (Roche Molecular Biochemicals) overnight. On thenext day, cells were washed and cultured in fresh, serum-free medium containing10 lM Fo, 20 nM PMA, 100 ng/ml AREG, Fo/PMA, or Fo/AREG, and after 4 hharvested with lysis buffer (0.2 M Tris, pH 8.0, containing 0.1% Triton X-100).An additional group of cells were transfected with pSCT empty vector or thePGR-A expressing pSCT-PGR-A construct (10 ng). Fo and PMA have been usedpreviously to mimic the effects of the LH surge and optimal induction of Ptgs2and Pgr in cultured rat granulosa cells [27, 28]. Luciferase activity in the extractswas analyzed according to a standard protocol [29]. In brief, a 40 ll aliquot of thecell lysate was mixed automatically with 100 ll luciferase assay reagent (20 mMTris, pH 8.0, containing 4 mM MgSO

4, 0.1 mM EDTA, 30 mM dithiothreitol, 0.5

mM ATP, 0.5 mM luciferin, and 0.25 mM coenzyme A), and each reaction wasmonitored 20 sec in a luminometer. The protein concentrations were determined

by mini-Bradford assay (Bio-Rad Laboratories, Inc.). Data are expressed basedon the amount of protein in each sample: light-specific units (LSUs) permicrogram of protein (mean 6 SD). Transfection of empty pGL3 vector ingranulosa cells showed basal values that were less than unstimulated wild-typepGL3�1065/þ50 construct. The inducibility of pGL3 by Fo/PMA was nevermore than 1.5-fold in each experiment.

Primary mouse granulosa cells isolated from eCG-primed mice ovaries byneedle puncture was employed to investigate agonist-induced Adam8expression. The isolated cells were cultured overnight in DMEM:F-12containing 1% serum, and on the next day cells were treated with Fo/PMAor Fo/AREG for 4 h. At the end of the treatment, the cells were harvested fortotal RNA isolation and semiquantitative RT-PCR analyses.

Statistics

Results are presented as means 6 SD from at least three differentexperiments. Significant differences between groups were analyzed byANOVA followed by the Neuman-Keuls test employing GraphPad Prismsoftware (San Diego, CA).

RESULTS

Adam8 mRNA Is Hormonally Regulated and DifferentiallyExpressed in Ovaries of Wild-Type and Pgr Null Mice

To analyze Adam8 expression, total RNA was preparedfrom whole ovaries of immature mice at specific times prior toand after hormone treatments to stimulate preovulatory follicledevelopment and ovulation. Semiquantitative RT-PCR analysisrevealed that Adam8 mRNA was expressed in ovaries ofimmature mice and that treatment with eCG did not cause a

FIG. 3. Reduced expression Adam8 inPRKO mice. The expression of Adam8mRNA in PRKO mice was determined byRT-PCR. Immature Pgr null heterozygotes(Hetþ/�) and homozygotes (PRKO�/�)were treated with eCG and hCG as de-scribed in Material and Methods. A) Ovarieswere harvested at indicated time points(eCG, hCG: 8, 12, and 16 h) for RNAextraction, and the relative expression ofAdam8 mRNA in these samples was deter-mined; shown are duplicates of triplicate ortriplicate RT-PCR reaction products pertreatment. B) The relative expression ofAdam8 in wild-type (WT þ/þ), Het (þ/�),and PRKO (�/�) mice is represented in theform a bar graph that shows the loss of Pgrin each allele (**P , 0.01, ***P , 0.001).

1042 SRIRAMAN ET AL.

significant change in Adam8 mRNA levels. Upon administra-tion of an ovulatory dose of hCG, the levels of Adam8 mRNAincreased rapidly at 2 h and reached a maximum level by 4 h(Fig. 1A). Adam8 expression remained elevated until 16 h butdecreased markedly by 24 h. Changes in the ovarian levels ofADAM8 protein followed a similar expression pattern (Fig.1B). Western blot analyses of ADAM8 in other tissuestypically reveal three immunoreactive bands at 120, 90, and60 kDa, representing the unprocessed, processed, and remnantforms of the protein, respectively [16]. Identical bands wereobserved and were found to increase in ovarian samples 4–8 hpost-hCG (Figure 1B). Processing of the ;120 kDa proform ofADAM8 by autocatalysis to ;90 kDa is required forproteolytic activity [16], and the ;60 kDa remnant form isabundant in all tissues and cell types that express ADAM8[15]. a-Actin and total ERK served as a loading control for theWestern blots. No specific bands were observed on Westernblot analysis with preimmune serum.

In situ hybridization confirmed induction of Adam8 mRNAat 8 h post-hCG and localized its expression to granulosa cellsof preovulatory follicles. Little or no signal was observed ineCG-treated ovaries (Fig. 2A, upper left). To verify andquantify Adam8 expression, semiquantitative RT-PCR analyseswere carried out using RNA prepared from granulosa cells andCOCs isolated at different time intervals after eCG-hCGtreatment. The results confirmed that induced expression ofAdam8 mRNA occurred in granulosa cells and COCs 8 h afterhCG treatment (Fig. 2B). The expression of Adam8 mRNAwas also analyzed in ovaries of Pgr null mice that exhibit an

anovulatory phenotype. RT-PCR analyses of total RNAsamples prepared from Pgr null homozygote (�/�) andheterozygote (þ/�) mouse ovaries at 8, 12, and 16 h afterhCG treatment documented significantly reduced levels ofAdam8 mRNA in PRKO samples compared to samples fromheterozygote at 8 and 12 h but not at 16 h (Fig. 3A).Interestingly, the decline in Adam8 mRNA at 8 h post-hCGcorrelated with the decrease in number of Pgr alleles (Fig. 3B)relative to wild type mice.

Expression of Adam8 Is Induced in Granulosa Cells fromPreantral Follicle Culture upon Stimulation by hCG and EGF

To further examine the control of Adam8 expression inperiovulatory follicles, a follicle culture system was used (Fig.4A). Preantral follicles with two to three layers of granulosacells (100–130 lm diameter) were manually dissected from theovaries of prepubertal mice and cultured for 12 days in thepresence of rFSH to promote follicle and oocyte growth, antrumformation, and granulosa cell differentiation in vitro. At Day 12,medium was replenished, and in vitro ovulation was induced byrhCG and rEGF. Within 18 h of culture to Day 13, the majority(on average .95%) of the antral follicles undergo in vitroovulation. Granulosa cells were harvested to analyze expressionof Adam8 by RT-PCR prior to (0 h) and following (4 and 18 h)follicle exposure to rhCG/rEGF (Fig. 4H). There was a low levelof expression at Day 12. The relative abundance of transcriptincreased dramatically at 4 h and remained high at 18 h afterstimulation of in vitro ovulation by rhCG/rEGF. Expression wasalso compared between the 18 h rhCG/rEGF-stimulated

FIG. 4. Induction of Adam8 mRNA ingranulosa cells of preantral follicle culturestimulated by hCG and EGF for in vitroovulation. Preantral follicle culture (A)starting with follicles with two to threelayers of granulosa cells, a basal membrane,and some theca cells that are isolated andcultured in medium with rFSH and rLH.After attachment (Day 1; B), granulosa cellsproliferate (Day 4; C) and differentiate intocumulus (arrowhead) and mural granulosacells (Day 8; D). They form follicles withantral-like cavities (white arrow; Day 12; E)when cultured in medium containing rFSH.On Day 12, oocytes still contain a GV(black arrow; e), but 4 h after replenishmentof medium with rhCG and rEGF (F),germinal vesicle breakdown occurs (f ).Eighteen hours after stimulation (Day 12 þ18 h, corresponding to Day 13 of culture;G) of in vitro ovulation, a PB oocyte withinthe expanded COC is found (g) (ZP: zonapellucida). H) Total RNA was isolated at 0,4, and 18 h after the rEGF/rhCG treatmentto determine the expression of Adam8 byRT-PCR. Rpl7 expression analysis by RT-PCR served as an internal control. Bar ¼ 50lm (A–C and e–g); 100 lm (D–G).

REGULATION OF Adam8 GENE EXPRESSION IN OVARY 1043

follicles cultured in the absence or presence of 10 lM AG1478(EGFR inhibitor) or 1 lM RU486 (PGR inhibitor) (Fig. 5A) orstimulated by 100 ng/ml epiregulin instead of rhCG/rEGF (Fig.5B). AG1478 as well as RU486 significantly reduced theexpression of Adam8 mRNA whereas the relative abundance ofmessage was significantly increased over the rhCG/rEGFcontrols by stimulation with epiregulin.

Hormonal Regulation of the Adam8 Promoter Activity inGranulosa Cells Depends on Multiple Promoter Elements

To understand the transcriptional regulation of Adam8, weconfirmed the role of LH signaling in induction of Adam8mRNA by the addition of Fo/PMA or Fo/AREG to primarygranulosa cell cultures. Treatment of primary mouse granulosacells with Fo/PMA for 4 h induced the Adam8 expression 4–6fold over the basal levels (Fig. 6A). AREG (100 ng/ml)increased the expression of Adam8 ;1.4 fold. However, whenAREG was combined with Fo, Adam 8 mRNA levels wereincreased 3–4 fold over basal levels (Fig. 6B).

To further determine the molecular mechanisms by whichAdam8 is induced in granulosa cells, mouse Adam8 promoter-luciferase constructs were generated based on a previous studyby Kataoka et al. [30]; they documented that ;1 kb of theAdam8 promoter contained the sites that are necessary to conferlipopolysaccharide and interferon-c inducibility. In addition,their studies determined the transcription start site and predictedseveral putative transcription factor binding sites. Therefore, wecloned 1 kb (�1065/þ50) of the murine Adam8 promoter alongwith 50 bases of exon I. Based on the predicted transcription

factor binding sites [30], we also cloned shorter constructs(�615/þ50, �240/þ50, and �106/þ50) by a PCR-basedapproach. Each promoter fragment was subcloned into pGL3-luciferase reporter constructs. Transfection assays were carriedout in primary rat granulosa cells (that can be easily transfectedbut lack LH-receptor) to determine the inducibility of thepromoter by Fo/PMA. The�240 Adam8 promoter-luc constructexhibited the highest basal activity as well as the greatestinducible (15–20 fold) activity (Fig. 7A) that decreased withdeletion to �106 Adam8-luc. These results indicated that the�240/þ50 region of the promoter contained elements that arecritical for the inducibility of the promoter whereas the regionsdistal to it have putative repressor elements. All the promoterregions were responsive to Fo, PMA, or Fo/PMA, with the lattershowing an additive effect for the �106 Adam8 promoter-lucconstruct (Fig. 7B). Thus, the �106 Adam8 promoter-lucconstruct appeared to contain the minimal elements that arerequired for Fo/PMA-induced activation of the promoter.

Adam8 expression is reduced in ovaries of the Pgr nullmice, indicating that it is regulated in part by PGR. Therefore, aPGR-A expression vector was cotransfected with the �106Adam8-luc promoter construct in granulosa cells. Expression ofPGR-A increased the basal, Fo, PMA, and Fo/PMA inducibil-ity 2–3 fold compared to the empty vector controls showing astimulatory role of PGR for induction of this gene (Fig. 7B).Similar results were obtained with cotransfection of PGR-Aand �240 Adam8-luc promoter-reporter construct (data nowshown). AREG (100 ng/ml) alone also consistently increased�106 Adam8-luc promoter activity ;1.7 fold over the basallevels and exerted an additive effect when combined with Fo

FIG. 5. Influence of inhibition of EGFR or PGR signaling or epiregulin on expression of Adam8 in granulosa mouse cells during ovulation in vitro. Theregulatory role of PGR- and EGF-like growth factors in induction of Adam8 in follicle culture was determined by treatment with 10 lM AG1478 or 1 lMRU486 for the last 18 h with rhCG/EGF. A) At the end of the incubation period, total RNA was isolated and Adam8 expression was determined by RT-PCR.Similarly, the role of EGF-like growth factor in the expression of Adam8 was determined by the treatment of follicle cultures with 100 ng/ml epiregulin orrhCG/rEGF at the last 18 h. B) At 18 h after treatment, RNA was isolated and expression of Adam8 was determined by RT-PCR. (**P , 0.01; ***P , 0.001.)

1044 SRIRAMAN ET AL.

FIG. 7. Induction of Adam8 promoter inresponse Fo/PMA or Fo/AREG treatment andcotransfection of PGR-A. The hormonesresponsiveness and the functional regionsthat are critical for Adam8 promoter in-duction were determined by cloning differ-ent lengths of promoter regions by PCRapproaches into pGL3 vector that extendfrom �1065, �615, �240, and �106 toþ50 relative to the transcription start site(þ1). The cloned promoter constructs (500ng) were transfected into primary rat gran-ulosa cells with Fugene 6 to determine theirinducibility in response to Fo, PMA, and Fo/PMA. A) Specifically, the cells were trans-fected overnight and treated with Fo, PMAor Fo/PMA (FP) for 4 h, lysed, and extractedto determine the luciferase activity. Thevalues were expressed as light-specific units(LSU) per lg of protein. B) In addition, thecells were cotransfected with empty vectorpSCT or pSCT-PGR-A expression construct(10 ng) to determine the role of PGR inregulation of the inducibility of the con-struct under basal and Fo/PMA-treatedconditions. C) Similarly, the inducibility ofthe promoter in response to AREG, Fo, orboth was also determined. (*P , 0.05, ***P, 0.001, Fo and Fo/PMA, respectively,between the empty vector and PGR-Atransfected treatments.)

FIG. 6. Induction of Adam8 mRNA inprimary mouse granulosa cells in responseto Fo and PMA treatment. The role of LHand EGF signaling in induction of Adam8 inprimary mouse granulosa cells were deter-mined by treatment with Fo/PMA (A) andFo/AREG or AREG alone (B). The cells weretreated with the agonists for 4 h, and totalRNA was isolated to determine the Adam8induction by a nonradioactive semiquanti-tative RT-PCR. Amplification of Rpl19served as an internal control. The productswere resolved in a 2% agarose gel and therelative induction of Adam8 in these reac-tions was quantitated by gel documentationsystem. (***P , 0.001, control (Con) vs. Fo/PMA and Con vs. Fo/AREG.)

REGULATION OF Adam8 GENE EXPRESSION IN OVARY 1045

(;14 fold) (Fig. 7C). A similar trend of induction wasobserved with the�240 Adam8-luc promoter-reporter constructon AREG or Fo treatment (data now shown).

Multiple Regions of the Adam8 Promoter Are Critical for ItsInducibility in Granulosa Cells

Our previous studies with PGR target genes, namely,Adamts1 and Ctsl, documented that GC-rich Sp1/Sp3 bindingsites are critical for mediating Fo/PMA and PGR effects [3, 31].Therefore, when computer analyses of the Adam8 promoter

identified the presence of three GC boxes—II, III, and IV(�296, �277, and �231, respectively)—within the �615 to�220 region (Fig. 8A), these GC boxes were mutated togenerate a triple mutant in the�615/þ50 version of the Adam8promoter. However, these mutations did not alter the basal andFo/PMA inducibility of the promoter (Fig. 8B). Since the regionextending between the�106 bp and the transcription start sitewas enough to confer Fo/PMA responsiveness, we reasonedthat this region might contain the key regulatory sites for theinduction of the promoter. Studies by Katoaka et al. [30]predicted the presence of binding sites for NF-1 and IRF-1 as

FIG. 8. Determination of cis-acting ele-ments that mediate Fo/PMA induction ofAdam8 promoter. A) Adam8 promoter DNAsequence representing the transcription startsite and the putative transcription factorbinding sites in the promoter. B) The criticalregion of Adam8 promoter that confers Fo/PMA (FP) inducibility within the �106/þ50region was determined by site-directedmutagenesis employing the editor kit. GCbox I, NF-1 site, and TATA boxes weremutated individually. In addition, GC box I–TATA box double mutant and GC box I–NF-1 site–TATA box triple mutant were gener-ated. The mutants were confirmed bysequencing and transfected (500 ng) intoprimary rat granulosa cells employing Fu-gene 6 to determine the Fo/PMA inducibilitycompared to the wild type (Wt) controls asdescribed earlier. (*P , 0.05, ***P , 0.001;analyzed between Wt and mutants.) C)Sequence alignment of Sp1 sites inAdamts1, Pgr, and Adam8 (GC box I)promoter aligned to the consensus Sp1/Sp3binding sequence.

1046 SRIRAMAN ET AL.

well as a TATA box in this regions (Fig. 8B); however theinterferon regulatory factor-1 (IRF-1) site did not match with theIRF-1 consensus sequence [32]. In addition to the above-mentioned sites, we noted another GC box adjacent to theputative NF-1 site in the proximal promoter (GC box I; Fig, 8, Aand B). The GC box I (GGAGG) is identical to SP1/SP3 bindingsites found in the promoters of the Pgr and Adamts1 genes [28,31] (Fig. 8C). Mutation of the GC box I and TATA boxindividually did not alter the basal or Fo/PMA inducibility of thepromoter; however, a double mutant that disrupted the GC box Iand TATA box exhibited ;25% reduction in basal and Fo/PMAinducibility of the promoter (Fig. 8B). Mutation of NF-1 sitereduced the inducibility of the promoter by ;25%. Subse-quently, when all three sites (GC box I, NF-1, and TATA) weremutated, the basal and Fo/PMA response declined by more than75% (Fig. 8B). Thus, transcription activation of Adam8 isuniquely regulated by a combinatorial set of factors distinctfrom other LH- and PGR-regulated genes that are primarilyregulated by SP1/SP3 [33].

DISCUSSION

Our results document for the first time that the metal-loprotease ADAM8 is expressed in granulosa cells of smallfollicles and is selectively increased in cumulus cells andgranulosa cells of preovulatory follicles of the murine ovary.Specifically, Adam8 mRNA is induced in vivo by the ovulatorysurge of LH/hCG. The effects of LH appear to be mediated, inpart, by its ability to induce PGR and EGF-like growth factors[34, 35]. In mice null for Pgr, the levels of Adam8 mRNA werereduced markedly prior to ovulation and at 8 and 12 h post-hCG, but not following ovulation at 16 h, indicating that Adam8is present in granulosa cells but is reduced as the cells luteinize.

As observed in vivo, low levels of Adam8 mRNA weredetected in murine granulosa cells of periovulatory folliclesfrom preantral follicle culture. Likewise, within 4 h after theaddition of rhCG/rEGF on Day 12 (to mimic the LH surge invivo), there was a dramatic increase in Adam8 mRNA levels,and these high levels persisted until 18 h, when in vitroovulation occurred. Moreover, as observed in the Pgr null mice,a PGR inhibitor reduced Adam8 mRNA in granulosa cells fromantral follicles cultured in the presence of rhCG/rEGF. Lastly,the positive effects of epiregulin and the inhibitory effects ofEGFR blockers on Adam8 expression in the in vitro ovulationmodel support the notion that the EGFR pathway affect theexpression of this protease both in vitro and in vivo. However,the fact that AREG alone did not markedly induce Adam8 inprimary murine or rat granulosa cell cultures indicates that thefollicle milieu provides additional critical regulatory molecules.

Analyses of Adam8 promoter-luciferase constructs providedadditional support that factors downstream of LH, includingPGR and EGF-like factors, regulate transcription of the Adam8gene in vivo. Specifically, transient transfection analyses ofvarious promoter regions from�1 to 0.1 kb were responsive toFo/PMA, suggesting that this minimal region is sufficient toconfer PRG, AREG, and Fo/PMA responsiveness in mamma-lian granulosa cells. Since many of the genes that are inducedduring the periovulatory period, including Pgr [28], Sgk1 [36],Ctsl [3], Egr1 [37], Adamts1 [31], and Areg [38], have SP1/SP3 sites that are critical for induction, we originally predictedthat a triple mutant of the three GC boxes at�296,�277, and�231bp in Adam8 promoter would abolish promoter activity.However, mutation of these GC boxes did not alter the basal orFo/PMA inducibility of the Adam8 promoter. Within the 100bp region, the predicted IRF-1 site [30] did not match withreported consensus AANNGAA [32]; hence, we focused on the

NF-1 and TATA site. Mutation of the putative NF-1 site alonereduced the Fo/PMA inducibility of the promoter, suggestingthat NF-1 or an NF-1-like factor [31] may be involved. TheAdam8 promoter TATA box (TATAGAG) did not match theconsensus TATA box TATA(A/T)A(A/T) [39], suggesting thatthe Adam8 TATA is an unusual or weak TATA box.Moreover, mutation of the TATA box alone did not alter theFo/PMA inducibility of the promoter. Although the GC box (I)had a consensus SP1/SP3 site binding sequence similar to thatcharacterized in the Pgr and Adamts1 promoters [28, 31],mutation of this region alone also did not alter Adam8 promoteractivity (Fig. 8B). However, a double mutation of GC box Iand the TATA box and a triple mutant (GC box I, NF-1 site,and TATA) reduced basal and agonist-regulated Adam8promoter activity 20% and 80%, respectively. Thus, transcrip-tion factors binding to these three regions exert a cooperativeeffect, and any two of the three sites appear sufficient forAdam8 promoter activation. In conclusion, the interaction ofPGR with NF-1 and/or SP1/SP3 in these regions of the Adam8promoter provides a novel regulatory mechanism of inductionof a gene in the metzincin family in addition to that observedwith the Adamts1 promoter [31].

Although the precise function of ADAM8 in the ovaryremains to be determined, its expression in the embryonicgonad [40, 41] and its induction in periovulatory follicles invivo and in vitro as well as in the proestrous uterus [42]indicate that ADAM8 may have more than one function in thedeveloping ovarian follicles and in fertility. As a metal-loprotease, ADAM8 could influence the ovulatory process bymodifying the extracellular matrix. Specifically within thefollicle, ADAM8 may be critical for the proteolytic release ofmatrix-associated factors (ectodomain shedding) [12, 13, 43]involved in the expansion of the cumulus oocyte complex andcell migration, which would be similar to its activities inglioma cells [19]. Ectodomain shedding that causes the releaseof EGF-like growth factors from the matrix following LH surgemay also facilitate LH action during ovulation [35]. Prelimi-nary studies document the expression of FCER2 (CD23) andCHL1, known substrates of ADAM8, in the ovary (data nowshown). Cumulus cells of COCs have been shown to expressgenes that are highly relevant to neuronal and immunefunction, including MBP, another known ADAM8 substrate[44]. Since MBP is known to stabilize membrane structures[45], cleavage of MBP by ADAM8 may alter cumulus cellfunctions. Despite these putative roles for ADAM8, mice nullfor Adam8 do not exhibit an overt anovulatory phenotype andare fertile, possibly because of potential functional redundan-cies with many other proteases expressed and/or activatedduring ovulation. Collectively, our results document for thefirst time that Adam8 is expressed in granulosa cells ofprimary/small preantral follicles within the adult murine ovaryand is induced selectively in granulosa/cumulus cells duringthe periovulatory period in vivo and in culture by thecombinatorial effects of LH, PGR, and EGF-like factors. Toour knowledge ADAM8 is the first member of the ADAMfamily that has been shown to be hormonally regulated.

REFERENCES

1. Richards JS, Russell DL, Ochsner S, Espey LL. Ovulation: newdimensions and new regulators of the inflammatory-like response. AnnuRev Physiol 2002; 64:69–92.

2. Richards JS, Russell DL, Ochsner S, Hsieh M, Doyle KH, Falender AE,Lo YK, Sharma SC. Novel signaling pathways that control ovarianfollicular development, ovulation, and luteinization. Recent Prog HormRes 2002; 57:195–220.

3. Sriraman V, Richards JS. Cathepsin L gene expression and promoteractivation in rodent granulosa cells. Endocrinology 2004; 145:582–591.

REGULATION OF Adam8 GENE EXPRESSION IN OVARY 1047

4. Oksjoki S, Soderstrom M, Vuorio E, Anttila L. Differential expressionpatterns of cathepsins B, H, K, L and S in the mouse ovary. Mol HumReprod 2001; 7:27–34.

5. Richards JS, Hernandez-Gonzalez I, Gonzalez-Robayna I, Teuling E, LoY, Boerboom D, Falender AE, Doyle KH, LeBaron RG, Thompson V,Sandy JD. Regulated expression of ADAMTS family members in folliclesand cumulus oocyte complexes: evidence for specific and redundantpatterns during ovulation. Biol Reprod 2005; 72:1241–1255.

6. Liu K, Wahlberg P, Leonardsson G, Hagglund AC, Ny A, Boden I,Wibom C, Lund LR, Ny T. Successful ovulation in plasminogen-deficientmice treated with the broad-spectrum matrix metalloproteinase inhibitorgalardin. Dev Biol 2006; 295:615–622.

7. Pall M, Mikuni M, Mitsube K, Brannstrom M. Time-dependent ovulationinhibition of a selective progesterone-receptor antagonist (Org 31710) andeffects on ovulatory mediators in the in vitro perfused rat ovary. BiolReprod 2000; 63:1642–1647.

8. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CAJr, Shyamala G, Conneely OM, O’Malley BW. Mice lacking progesteronereceptor exhibit pleiotropic reproductive abnormalities. Genes Dev 1995;9:2266–2278.

9. Robker RL, Russell DL, Espey LL, Lydon JP, O’Malley BW, Richards JS.Progesterone-regulated genes in the ovulation process: ADAMTS-1 andcathepsin L proteases. Proc Natl Acad Sci U S A 2000; 97:4689–4694.

10. Shimada M, Nishibori M, Yamashita Y, Ito J, Mori T, Richards JS. Down-regulated expression of A disintegrin and metalloproteinase withthrombospondin-like repeats-1 by progesterone receptor antagonist isassociated with impaired expansion of porcine cumulus-oocyte complexes.Endocrinology 2004; 145:4603–4614.

11. Shindo T, Kurihara H, Kuno K, Yokoyama H, Wada T, Kurihara Y, ImaiT, Wang Y, Ogata M, Nishimatsu H, Moriyama N, Oh-hashi Y, Morita H,Ishikawa T, Nagai R, Yazaki Y, Matsushima K. ADAMTS-1: ametalloproteinase-disintegrin essential for normal growth, fertility, andorgan morphology and function. J Clin Invest 2000; 105:1345–1352.

12. Wolfsberg TG, Primakoff P, Myles DG, White JM. ADAM, a novelfamily of membrane proteins containing A Disintegrin And Metal-loprotease domain: multipotential functions in cell-cell and cell-matrixinteractions. J Cell Biol 1995; 131:275–278.

13. Yamamoto S, Higuchi Y, Yoshiyama K, Shimizu E, Kataoka M, Hijiya N,Matsuura K. ADAM family proteins in the immune system. ImmunolToday 1999; 20:278–284.

14. Yoshida S, Setoguchi M, Higuchi Y, Akizuki S, Yamamoto S. Molecularcloning of cDNA encoding MS2 antigen, a novel cell surface antigen stronglyexpressed in murine monocytic lineage. Int Immunol 1990; 2:585–591.

15. Schlomann U, Rathke-Hartlieb S, Yamamoto S, Jockusch H, Bartsch JW.Tumor necrosis factor alpha induces a metalloprotease-disintegrin,ADAM8 (CD 156): implications for neuron-glia interactions duringneurodegeneration. J Neurosci 2000; 20:7964–7971.

16. Schlomann U, Wildeboer D, Webster A, Antropova O, Zeuschner D,Knight CG, Docherty AJ, Lambert M, Skelton L, Jockusch H, Bartsch JW.The metalloprotease disintegrin ADAM8. Processing by autocatalysis isrequired for proteolytic activity and cell adhesion. J Biol Chem 2002; 277:48210–48219.

17. Naus S, Reipschlager S, Wildeboer D, Lichtenthaler SF, Mitterreiter S,Guan Z, Moss ML, Bartsch JW. Identification of candidate substrates forectodomain shedding by the metalloprotease-disintegrin ADAM8. BiolChem 2006; 387:337–346.

18. Naus S, Richter M, Wildeboer D, Moss M, Schachner M, Bartsch JW.Ectodomain shedding of the neural recognition molecule CHL1 by themetalloprotease-disintegrin ADAM8 promotes neurite outgrowth andsuppresses neuronal cell death. J Biol Chem 2004; 279:16083–16090.

19. Wildeboer D, Naus S, Amy Sang QX, Bartsch JW, Pagenstecher A.Metalloproteinase disintegrins ADAM8 and ADAM19 are highlyregulated in human primary brain tumors and their expression levels andactivities are associated with invasiveness. J Neuropathol Exp Neurol2006; 65:516–527.

20. Sriraman V, Rudd MD, Lohmann SM, Mulders SM, Richards JS. Cyclicguanosine 50-monophosphate-dependent protein kinase II is induced byluteinizing hormone and progesterone receptor-dependent mechanisms ingranulosa cells and cumulus oocyte complexes of ovulating follicles. MolEndocrinol 2006; 20:348–361.

21. Russell DL, Doyle KM, Ochsner SA, Sandy JD, Richards JS. Processingand localization of ADAMTS-1 and proteolytic cleavage of versicanduring cumulus matrix expansion and ovulation. J Biol Chem 2003; 278:42330–42339.

22. Cortvrindt RG, Smitz JE. Follicle culture in reproductive toxicology: atool for in-vitro testing of ovarian function? Hum Reprod Update 2002; 8:243–254.

23. Sun F, Betzendahl I, Shen Y, Cortvrindt R, Smitz J, Eichenlaub-Ritter U.Preantral follicle culture as a novel in vitro assay in reproductivetoxicology testing in mammalian oocytes. Mutagenesis 2004; 19:13–25.

24. Van Wemmel K, Gobbers E, Eichenlaub-Ritter U, Smitz J, Cortvrindt R.Ovarian follicle bioassay reveals adverse effects of diazepam exposureupon follicle development and oocyte quality. Reprod Toxicol 2005; 20:183–193.

25. Smitz J, Cortvrindt R, Van Steirteghem AC. Normal oxygen atmosphere isessential for the solitary long-term culture of early preantral mousefollicles. Mol Reprod Dev 1996; 45:466–475.

26. Pedersen T, Peters H. Proposal for a classification of oocytes and folliclesin the mouse ovary. J Reprod Fertil 1968; 17:555–557.

27. Morris JK, Richards JS. Hormone induction of luteinization andprostaglandin endoperoxide synthase-2 involves multiple cellular signal-ing pathways. Endocrinology 1993; 133:770–779.

28. Sriraman V, Sharma SC, Richards JS. Transactivation of the progesteronereceptor gene in granulosa cells: evidence that Sp1/Sp3 binding sites in theproximal promoter play a key role in luteinizing hormone inducibility. MolEndocrinol 2003; 17:436–449.

29. Sharma SC, Clemens JW, Pisarska MD, Richards JS. Expression andfunction of estrogen receptor subtypes in granulosa cells: regulation byestradiol and forskolin. Endocrinology 1999; 140:4320–4334.

30. Kataoka M, Yoshiyama K, Matsuura K, Hijiya N, Higuchi Y, YamamotoS. Structure of the murine CD156 gene, characterization of its promoter,and chromosomal location. J Biol Chem 1997; 272:18209–18215.

31. Doyle KM, Russell DL, Sriraman V, Richards JS. Coordinate transcriptionof the ADAMTS-1 gene by luteinizing hormone and progesteronereceptor. Mol Endocrinol 2004; 18:2463–2478.

32. Fujii Y, Shimizu T, Kusumoto M, Kyogoku Y, Taniguchi T, HakoshimaT. Crystal structure of an IRF-DNA complex reveals novel DNArecognition and cooperative binding to a tandem repeat of core sequences.EMBO J 1999; 18:5028–5041.

33. Sriraman V, Richards JS. Sp1: A critical transcription factor of ovulation-related genes?. Biol Reprod 2003; 68 (suppl 1):229.

34. Natraj U, Richards JS. Hormonal regulation, localization, and functionalactivity of the progesterone receptor in granulosa cells of rat preovulatoryfollicles. Endocrinology 1993; 133:761–769.

35. Park JY, Su YQ, Ariga M, Law E, Jin SL, Conti M. EGF-like growthfactors as mediators of LH action in the ovulatory follicle. Science 2004;303:682–684.

36. Alliston TN, Maiyar AC, Buse P, Firestone GL, Richards JS. Folliclestimulating hormone-regulated expression of serum/glucocorticoid-induc-ible kinase in rat ovarian granulosa cells: a functional role for the Sp1family in promoter activity. Mol Endocrinol 1997; 11:1934–1949.

37. Russell DL, Doyle KM, Gonzales-Robayna I, Pipaon C, Richards JS. Egr-1 induction in rat granulosa cells by follicle-stimulating hormone andluteinizing hormone: combinatorial regulation by transcription factorscyclic adenosine 30,50-monophosphate regulatory element binding protein,serum response factor, sp1, and early growth response factor-1. MolEndocrinol 2003; 17:520–533.

38. Shao J, Evers BM, Sheng H. Prostaglandin E2 synergistically enhancesreceptor tyrosine kinase-dependent signaling system in colon cancer cells.J Biol Chem 2004; 279:14287–14293.

39. Babenko VN, Kosarev PS, Vishnevsky OV, Levitsky VG, Basin VV,Frolov AS. Investigating extended regulatory regions of genomic DNAsequences. Bioinformatics 1999; 15:644–653.

40. Coveney D, Ross AJ, Slone JD, Capel B. A microarray analysis of the XXWnt4 mutant gonad targeted at the identification of genes involved intestis vascular differentiation. Gene Expr Patterns 2007; 7:82–92.

41. Kelly K, Hutchinson G, Nebenius-Oosthuizen D, Smith AJ, Bartsch JW,Horiuchi K, Rittger A, Manova K, Docherty AJ, Blobel CP. Metal-loprotease-disintegrin ADAM8: expression analysis and targeted deletionin mice. Dev Dyn 2005; 232:221–231.

42. Kim J, Kim H, Lee SJ, Choi YM, Lee SJ, Lee JY. Abundance of ADAM-8, -9, -10, -12, -15 and -17 and ADAMTS-1 in mouse uterus during theoestrous cycle. Reprod Fertil Dev 2005; 17:543–555.

43. Schlondorff J, Becherer JD, Blobel CP. Intracellular maturation andlocalization of the tumour necrosis factor alpha convertase (TACE).Biochem J 2000; 347Pt 1:131–138.

44. Hernandez-Gonzalez I, Gonzalez-Robayna I, Shimada M, Wayne CM,Ochsner SA, White L, Richards JS. Gene expression profiles of cumuluscell oocyte complexes during ovulation reveal cumulus cells expressneuronal and immune-related genes: does this expand their role in theovulation process? Mol Endocrinol 2006; 20:1300–1321.

45. DeBruin LS, Haines JD, Wellhauser LA, Radeva G, Schonmann V,Bienzle D, Harauz G. Developmental partitioning of myelin basic proteininto membrane microdomains. J Neurosci Res 2005; 80:211–225.

1048 SRIRAMAN ET AL.