| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Department of Cell and Developmental Biology and Anatomy, University of South Carolina School of Medicine, Columbia, South Carolina 29208
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
corpus luteum, granulosa cells, ovary, ovulatory cycle, theca cells
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mammalian ovary is a dynamic organ with follicles in different stages of development. Simple structures like primordial follicles will undergo extensive processes of cell growth, proliferation, and selection before developing into more complex preovulatory antral follicles. Terminal differentiation of postovulatory granulosa cells and theca cells is necessary for the formation of the corpus luteum (CL) and maintenance of pregnancy. These developmental processes are primarily under the regulation of gonadotropins, FSH and LH, and ovarian growth factors. Several growth factors and polypeptide hormones trigger AP-1 activity in other systems [2, 14], and a few studies have been linked to AP-1 function in the ovary. c-Jun and c-Fos mRNAs were induced in diethylstilbestrol (DES)-primed rat granulosa cells with acute gonadotropin, cAMP analogue, or phorbol ester treatment [15]. JunD and Fra-2 expression were induced by hCG and correlate with terminal differentiation in rat granulosa cells, which is associated with increased progesterone production [16]. Prostaglandin F2
(PGF2), a luteolytic agent, decreased ovarian steroidogenic acute regulatory protein (StAR) mRNA, which correlated with increased c-Fos message and overexpression of c-Fos in transfected cells repressed activation of the murine StAR promoter [17]. PGF2 has also been shown to induce the progesterone-catabolizing enzyme, 20-hydroxysteroid dehydrogenase, via phospho-JunD activation of nur77 [18]. In bovine luteal cells, PGF2 has been shown to induce c-jun and c-fos mRNAs [19]. These previous studies implicate activation of different AP-1 family members with the processes of follicular growth, differentiation, and luteal regression.
Although there are few reports directly linking ovarian hormonal and growth factor mechanisms to AP-1 factors, signaling pathways such as protein kinase A (PKA), protein kinase C (PKC), and mitogen-activated protein (MAP) kinases have been shown to utilize AP-1 factors as downstream effectors in other systems [2, 20]. The PKA, PKC, and MAP kinase cascades modulate gonadotropin and/or growth factor effects in ovarian cells. For example, cyclic AMP activation of PKA is the major initial signaling pathway for both FSH and LH, and both of these gonadotropins promote steroidogenesis by activating MAP kinases [21, 22]. On the other hand, ovarian endothelin-1 activates PKC resulting in inhibition of FSH-stimulated progesterone production in cultured porcine granulosa cells [23]. Likewise, PGF2 inhibits gonadotropin-stimulated progesterone production by cultured granulosa-luteal or luteal cells via PKC and MAP kinase-dependent mechanisms [24–26]. In bovine luteal cells, PGF2 induction of c-Jun and c-Fos mRNA is via PKC-dependent MAP kinase activation [19]. The contrasting effects of some local growth factors and gonadotropins may be due to activation of particular kinase isoforms and/or specific AP-1 factors present in certain ovarian cells at different developmental stages.
In order to begin to study gene regulation by AP-1 factors and the role of AP-1 in ovarian follicle signaling cascades, it is of primary importance to establish the presence and distribution of these factors in the ovary. Although the presence and tissue distribution of AP-1 factors has been recently reported for the hormonally primed rat ovary [16], there is virtually no information regarding AP-1 factors in the ovaries of other species. The goal of this study was to determine which AP-1 family members are present in the porcine ovary and to determine whether detectable changes in their expression exist between whole-cell extracts of different-sized antral follicles, corpora lutea, or granulosa cells from antral follicles. Also, we evaluated the distribution of these factors between the cytoplasmic and nuclear fractions at certain developmental stages. Furthermore, considering AP-1 function can be cell type specific, immunohistochemical analyses of ovarian tissues collected from pigs at different stages of the estrous cycle (as assessed by ovarian morphology) were performed to determine the cellular distribution of these factors in the ovary. Finally, the DNA-binding capacity of different AP-1 family members was evaluated in selected ovarian cells at different stages of differentiation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Porcine ovaries obtained from a local abattoir were transported on ice to the laboratory. Small (1–2 mm in diameter), medium (3–4 mm), and large (5–10 or 8–10 mm) antral follicles or midluteal phase corpora lutea were dissected from the surrounding tissue. Midluteal CL were classified as described by Akins and Morrissette [27]. Whole-cell extracts (WCE) were obtained by lysis with 7–10 volumes of the following cocktail: 50 mM Tris, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 mM sodium vanadate, and 1 mM sodium fluoride using a Dounce homogenizer. Protease inhibitors were obtained from Sigma (St. Louis, MO). Corpora lutea were first minced before red blood cell lysis and/or homogenization. Granulosa cells were obtained by needle aspiration and scraping the inner lining of the follicle. Granulosa cells in the follicular fluid were pelleted by centrifugation, and cell pellets were lysed in the buffer described previously to obtain WCE. For nuclear and cytoplasmic extracts of granulosa cells, follicles, or CL, red blood cells were lysed by brief exposure of cells to cold sterile water followed by resuspension in Dulbecco phosphate-buffered saline (D-PBS). After red blood cell lysis, CL and follicles for nuclear protein and cytoplasmic were homogenized in buffer A. Granulosa cell pellets were resuspended in approximately 10 volumes of buffer A (50 mM Hepes, pH 7.5, 10 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 0.5 mM PMSF, 20 µg/ml aprotinin, 10 µg/ml leupeptin, 1 µg/ml soybean trypsin inhibitor, 1 mM NaF, 1 mM NaVO4). Triton X-100 (0.8%) was added, and nuclei were pelleted for 30 sec (3000 x g). The supernatant containing the cytoplasmic proteins was removed. Nuclei were resuspended in two volumes of buffer C (same as buffer A except with 400 mM KCl) and allowed to incubate for 20 min on ice before centrifugation (13 000 x g for 2 min). The supernatants containing nuclear proteins were removed. Nuclear and cytoplasmic fractions were snap frozen and stored at -70°C. Protein concentrations were determined using BioRad Dye reagent (Hercules, CA).
Immunoblot Analyses
For immunoblot analyses, equivalent amounts of protein (30–50 µg of WCE and 30 µg of cytoplasmic or nuclear extracts) were mixed with sample buffer, boiled, and electrophoretically separated on 10% or 12% SDS polyacrylamide gels. Proteins were electrotransferred to nitrocellulose Hybond membranes (Amersham Biosciences Corporation, Piscataway, NJ). Membranes were blocked in 5% nonfat milk in Tris-buffered saline containing 0.05% Tween (TTBS) and incubated overnight with primary antibodies against c-Jun (1 µg/ml, sc-45X), JunD (1 µg/ml, sc-74X), JunB (2 µg/ml, sc-8051X), c-Fos (1 µg/ml, sc-52X), Fra-1 (2 µg/ml, sc-183X, sc-605X), Fra-2 (1 µg/ml, sc-604X), FosB (2 µg/ml, sc-48X, sc-7203X), and normal rabbit IgG (sc-2027) (Santa Cruz Biotech, Santa Cruz, CA) at a 1- or 2-µg/ml dilution in 1% milk-TTBS. These antibodies cross-react with multiple animal species. Following washes, membranes were incubated at a 0.24- or 0.12-µg/ml dilution of secondary antibody (horseradish peroxidase [HRP]-goat anti-rabbit IgG, 65-6120; Zymed, South San Francisco, CA) in 5% milk-TTBS. When using primary JunB monoclonal mouse antibody (sc-8051X), HRP-goat anti-mouse IgG1 (0.18 µg/ml, 61-0120, Zymed) was used as the secondary antibody. For JunB immunoblots, blocking was done in 0.06% gelatin in TTBS for 2 h, and both primary and secondary antibodies were diluted in 0.06% gelatin. Immunoreactive bands were detected by enhanced chemiluminescence (Amersham). All experiments were performed using samples isolated from at least three different animals for CL or three different pools of tissue (for granulosa and follicles). After initial probing, WCE membranes were stripped with a buffer containing 0.5 M Tris-Cl, 10% SDS, 100 mM ß-mercaptoethanol at 60°C for 30 min, and then reprobed with actin antibody (1 µg/ml, sc-10731, Santa Cruz Biotech) to check for similar loading. Immunoreactive band specificity was determined by preabsorbing primary antibodies with their respective blocking peptides (Santa Cruz Biotech). Primary antibodies were preabsorbed using a 5-fold excess concentration of blocking peptide for 1.5 h at room temperature prior to addition to blots. A second control blot was probed with normal rabbit or mouse IgG in place of the primary antibody.
Immunohistochemistry
Porcine ovaries were obtained at a local abattoir and immediately fixed in 4% paraformaldehyde. The developmental stage of the CL was classified based on gross morphological appearance, including size, extent of vascularity, presence or absence of central blood clot, and color [27]. Tissues were embedded in paraffin, and 8-µm sections were placed on poly-L-lysine-coated slides. Sections were dewaxed in xylenes and subsequently rehydrated by processing through a series of decreasing ethanol solutions and then H2O. To quench endogenous peroxidase activity, the sections were incubated in 1% (v/v) hydrogen peroxide for 45 min. Sections were then blocked in PBS containing 10% normal goat serum for 60 min at room temperature. Optimal dilutions for each antibody were determined in preliminary studies. Sections were incubated overnight at 4°C with primary rabbit polyclonal antibodies against c-Jun (4 µg/ml, sc-45X), JunD (2 µg/ml, sc-74X), Fra-2 (2 µg/ml, sc-604X), c-Fos (4 µg/ml, sc-52X), FosB (4 µg/ml, sc-48X, sc-7203X), Fra-1 (2 µg/ml, sc-183X, sc-605X), and normal rabbit IgG (2–4 µg/ml, sc-2027) (Santa Cruz Biotech). HRP-conjugated goat anti-rabbit IgG was used as the secondary antibody (1.2 µg/ml, 65-6120, Zymed). All antibodies were diluted in 10% normal goat serum/PBS. Sections were then covered for approximately 2 min with a 3,3'-diaminobenzidine (DAB) solution according to the instructions for Liquid DAB-Plus substrate kit (Zymed). Sections were then counterstained with Accustain hematoxylin (Sigma) solution for 10 sec in order to visualize nuclei. For JunB an alternate protocol involving heat-induced antigen retrieval was performed as previously described [28] and incubated with 1.6 or 2 µg/ml of primary rabbit anti-JunB antibody (sc-48X) in 10% goat serum/PBS. Nuclei were scored as positive when a brown color was present [29]. Percentages were determined by counting positive nuclei of at least 100 granulosa, theca, or large luteal cell nuclei in the microscopic field (or multiple fields for structures containing less than 100 nuclei). All healthy oocytes per section were evaluated. Counting was performed on at least three different sections from three different animals. The large luteal cells counted exhibited large round nuclei with one or more nucleoli and extensive cytoplasm. For CL, small luteal cells were not evaluated because they are difficult to distinguish from nonsteroidogenic cells. The specificity of immunohistochemical staining was determined by preabsorbing primary antibodies with their respective blocking peptides (Santa Cruz Biotech) as described previously. Also, as a second negative control, primary antibodies were replaced with normal rabbit IgG.
Electrophoretic Mobility Shift Assays
A double-stranded oligonucleotide containing a single AP-1 consensus sequence was purchased from Promega Corporation (Madison, WI) and used for electromobility shift assays (EMSA). The oligonucleotide was labeled with 
-32P-ATP (3000 Ci/mmol) (Amersham) using T4 polynucleotide kinase (Promega). Nuclear protein extracts were dialyzed as previously described [28]. Five microliters of dialyzed nuclear protein (5–7 µg for granulosa cells and 5–15 µg for CL) were incubated with or without (4 µg) of antibody or 100-fold excess cold oligonucleotide in a buffer containing 10 mM Tris pH 7.5, 0.5 mM dithiothreitol, 5% glycerol, 0.5 mM PMSF, 4 ng/ml aprotinin, 2 ng/ml leupeptin, 80 mM KCl, and 2 µg poly dI-dC (Sigma) and incubated for 30 min on ice. Labeled AP-1 oligonucleotide (50 000 cpm) was added to the mixture and incubated for another 30 min on ice. Samples were mixed with loading buffer (25 mM Tris, pH 7.5, 4% glycerol) and electrophoresed on native 5% polyacrylamide gels for 3 h at 150 V in 0.5x Tris-borate-EDTA running buffer. Gels were dried and exposed to BioMAX MR film (Eastman Kodak Company, Rochester, NY) at -70°C with intensifying screens. The antibodies used for supershift were the same as those used for immunoblotting. Control experiments for supershift studies were performed with normal rabbit or mouse IgG instead of specific antibody.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To determine the presence of AP-1 transcription factor proteins under nonstimulated conditions in the porcine ovary, we performed immunoblot analyses of whole-cell extracts. The developmental stage of porcine granulosa cells and follicles was assessed by the size of the follicle [30, 31]. Whole-cell protein extracts isolated from antral follicles (1–2, 3–4, 5–10 mm diameter) or midluteal phase CL of cycling pigs were separated by SDS-PAGE followed by immunoblotting with specific antibodies against AP-1 factors. Immunoblot analyses confirmed that c-Jun, JunD, JunB, c-Fos, FosB, Fra-1, and Fra-2 immunoreactive proteins were present in whole-cell extracts of all antral follicles and CL (Fig. 1A) and granulosa cells from antral follicles of different sizes (1–2, 3–4, 5–7, 8–10 mm; Fig. 1B). On immunoblots of WCE, c-Jun appeared as predominantly a single 43-kDa isoform, JunB as a 39-kDa band, and JunD as at least two immunoreactive bands (35/40 kDa). c-Fos appears as one or two bands of approximately 55 kDa, Fra-1 as a 37-kDa protein [32], and Fra-2 as two or more immunoreactive bands (43/46 kDa). A band of the correct reported size (46 kDa) [32] was detected for FosB using two different antibodies in 50 µg of both granulosa and whole-follicle WCE. The intensities of c-Jun and c-Fos protein bands were decreased in all samples of CL compared to antral follicles. Although JunD appeared more intense in some samples of CL, this observation was not consistent. Other AP-1 factors appeared to be expressed at similar levels in antral follicles and CL.
|
A similar immunoreactive pattern was seen in whole-cell extracts between granulosa cells from different-sized antral follicles for all AP-1 proteins with the exception of Fra-2. In granulosa cells of 8–10-mm size follicles, which correspond in the pig with preovulatory follicles [27], Fra-2 consistently (n = 3) exhibited a higher-molecular-weight isoform. Preincubation of antibodies with their respective blocking peptides eliminated the immunoreactive bands.
The relative cytoplasmic and nuclear distributions of AP-1 molecules and their isoforms were evaluated in granulosa cells from whole 1–2- and 8–10-mm antral follicles and CL (Fig. 2). Equal amounts of nuclear protein and cytoplasmic protein were loaded in adjacent lanes on gels. c-Jun was more prevalent in the nucleus than the cytoplasm in granulosa cells and follicles, but the difference between fractions was less pronounced in CL. Short film exposures revealed that nuclear c-Jun existed as primarily two immunoreactive bands, and this was also evident in the cytoplasm. The two major JunD isoforms predominated in the nuclear fraction yet exhibited a cytoplasmic presence. JunB tended to have weak expression in most tissues with a slight nuclear predominance. Notably, in contrast to whole-cell extracts, JunB was consistently more abundant in nuclear extracts of 8–10 GC than any other tissue. c-Fos tended to be more intense in the nuclear extracts of granulosa cells and follicles, but the cytoplasmic fraction exhibited greater c-Fos in midluteal phase corpora lutea. A slightly retarded mobility was observed for the c-Fos nuclear band in granulosa and follicles. Fra-1 displayed similar abundance in cytoplasmic and nuclear extracts. Fra-2 resided mostly in the nuclear fraction. Up to four Fra-2 immunoreactive bands could be observed in tissues. As in whole-cell extracts, GC from 8–10-mm follicles exhibited a retarded mobility on gels when compared to GC from 1–2-mm follicles on the same immunoblots. Also with Fra-2, the highest-molecular-weight isoforms predominated in the nucleus, and the lower-molecular-weight isoforms were the prevalent cytoplasmic forms. With the exception of GC from 1–2-mm follicles, which exhibited similar cytoplasmic and nuclear band intensities, FosB predominated in the cytoplasmic fraction of the other tissues evaluated.
|
Differential Localization of AP-1 Transcription Factors
Immunohistochemical analysis of ovarian sections from animals at different stages of the estrous cycle was performed to localize Jun/Fos family proteins. The distribution of Jun family (Fig. 3) and Fos family (Fig. 4, except FosB) members was evaluated in healthy primordial, unilaminar primary, and multilaminar preantral follicles; small (1–2 mm), medium (3–4 mm), and large (5–10 mm) antral follicles; and corpora hemorrhagica (CH), midluteal phase corpora lutea, and corpora albicantia (CA). All the factors exhibited nuclear localization as expected [8]; therefore, the percentages of positive nuclei were determined as described in methods. A summary of the average frequency of immunopositive nuclei in different cell types for these AP-1 factors is shown in Table 1. In primordial and primary unilaminar follicles, all AP-1 factors studied except for FosB were detected in granulosa cell nuclei. Granulosa and theca cell nuclei of primary multilaminar follicles were immunopositive for all factors, with a low expression of FosB. In antral follicles, granulosa cells exhibited nuclear staining for all factors, as did thecal cell nuclei with the exception of FosB. For all stages examined, FosB could be detected in only a few granulosa and theca cell nuclei. Large antral follicles tended to show lower levels of granulosa and theca nuclear staining compared with other antral follicles for all factors (data not shown). All AP-1 factors were present in the oocyte nuclei of primordial and primary unilaminar follicles. In oocyte nuclei of multilaminar preantral follicles, only JunD, JunB, c-Fos, Fra-1, Fra-2, and FosB could be detected. In oocyte nuclei of antral follicles, JunD, JunB, c-Fos, and Fra-2 were detected. Cells in corpora hemorrhagica and lutea exhibited intense nuclear staining for JunD and Fra-2. All other factors were present in cell nuclei of CH and CL, but to a lesser extent. For each AP-1 factor, the intensities of staining for both CH and CL were similar. In corpora albicantia, lower levels of staining were seen for all factors compared to CH and CL, but also few cells with large luteal cell phenotype were present. IHC with FosB antibodies yielded weak nuclear staining in CH and CA. Cytoplasmic staining was observed to some extent for all AP-1 factors in all granulosa, theca, and luteal cells, and intense cytoplasmic staining was observed for c-Jun and c-Fos. FosB immunoreactivity was weak but tended to be mostly cytoplasmic. Stromal cell nuclear staining was observed with most antibodies, but no analysis of particular cell types was made. Preabsorption of the primary antibody with its respective blocking peptide or substitution of normal IgG for the primary antibody abolished almost all staining (cytoplasmic and nuclear) with the exception of weak oocyte and vascular smooth muscle cell cytoplasmic staining.
|
|
|
On examination of the ovarian sections, ovarian blood vessels, especially small arteries and arterioles, were immunopositive for most AP-1 factors. We analyzed the expression of AP-1 factor in ovarian blood vessels (Fig. 5 and Table 1). The majority of vascular smooth muscle cell (VSMC) nuclei expressed intense staining for JunD, JunB, Fra-1, and Fra-2. Less intense staining for c-Jun and c-Fos was observed in VSMC. Weak FosB staining of blood vessels appeared to be nonspecific, as it appeared on control slides (data not shown). Preabsorption of JunB antibody with its blocking peptide removed all follicular staining but only reduced the intense VSMC staining. Endothelial cell nuclei expressed c-Jun, JunD, JunB, c-Fos, and Fra-1.
|
Variation and Composition of AP-1 DNA-Binding Complexes
In order to determine whether AP-1 family members had the capacity to bind an AP-1 DNA consensus sequence, nuclear extracts from granulosa cells at two developmental stages (1–2-mm small follicles and 8–10-mm preovulatory follicles; Fig. 6, A and B) and midluteal phase CL tissue were investigated using electrophoretic mobility shift assays (Fig. 6C). The composition of AP-1 binding complexes was probed by addition of specific antibodies for each Jun and Fos family member. Addition of antibodies led to one or more supershifts of the DNA-protein complexes. Similar AP-1 complexes were observed in three different samples of each tissue. The composition of the binding complexes was similar in granulosa cells from 1–2- and 8–10-mm follicles and consisted predominantly of c-Jun, JunD, JunB, c-Fos, and Fra-2. The strongest supershifts were observed for Fra-2 and JunD. Binding of Fra-1 or FosB was minimal or not detected using two different antibodies for each protein. In CL, binding of c-Jun, JunD, and JunB was present. Fra-2 was the predominant Fos family member in CL AP-1 complexes. c-Fos was not detected as a component of the AP-1 DNA-binding complex in CL nuclear extracts.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our data indicate that although most AP-1 factors are present throughout the follicular cycle, specific Jun/Fos proteins are more characteristic for a particular stage of development than for another. JunD and Fra-2 are present in all follicles and granulosa cells indifferent of their size and also in corpus luteum. Our immunoblot data with whole-cell extracts revealed that granulosa cells of 8–10-mm-diameter (preovulatory) follicles showed a retardation in the gel mobility for Fra-2 protein compared to other stages of granulosa. This observation was confirmed in cytoplasmic and nuclear preparations also. This change is specific for granulosa cells of preovulatory follicles and may be due to posttranslational modifications [8, 33]. The appearance of higher-molecular-weight isoforms might reflect a developmental change in Fra-2 activity, as this factor is regulated through phosphorylation that converts it from an inefficient transcriptional activator to a more active one [34]. Preovulatory follicles are responsive to LH and after the midcycle LH surge undergo extensive differentiation to a highly steroidogenic phenotype [35]. It is therefore possible that phosphorylation of Fra-2 in preovulatory follicles may be important for initiation of differentiation and be a result of LH activity. Corpus luteum nuclear extracts also exhibited predominantly the higher-molecular-weight isoforms of Fra-2 suggesting these modifications may be important for ongoing luteal function.
Immunoblot data coincided with IHC results, as Fra-2 and JunD immunopositive nuclei were seen in both the granulosa and the theca cell layer of follicles of all sizes. Our findings in pig follicles are in contrast to the study of hormonally primed rat model in which JunD was low and Fra-2 negligible in follicles of ovarian sections prior to hCG injection [16]. Immunopositive JunD and Fra-2 appeared as the prevalent forms of nuclear AP-1 in the large cells of midluteal phase corpora lutea. The presence of JunD and Fra-2 in luteal cell nuclei was also evident in earlier corpora hemorrhagica. These data are consistent with the report in rat in which differentiation induced by hCG increased the nuclear JunD and Fra-2 immunoreactivity of luteinizing follicular cells and luteal cells [16]. The predominance of JunD and Fra-2 associated with differentiation has been observed previously in other cell types such as osteoblasts, embryonic carcinoma, and muscle cells [36–38]. Also, overexpression of JunD has been found to result in slower cell growth and an increase in the percentage of cells in G0/G1 phase of the cell cycle [39], which would be consistent with a less proliferative and more differentiated cellular phenotype. Our data suggest that JunD and Fra-2 presence in large luteal cells may be related with the need to maintain or promote a differentiated status.
Immunoblot data with whole-cell extracts showed c-Fos was present in preovulatory follicles and corresponding granulosa layer but lower in CL compared to antral follicles. Summation of c-Fos immunopositive nuclei revealed that the CL decrease in c-Fos immunoblot abundance is not a reduction in the number of immunopositive cells. Immunoblot evaluation of cytoplasmic versus nuclear fractions revealed that c-Fos predominated in cytoplasmic extracts of CL, whereas the nuclear fraction was more abundant in preovulatory structures. The decrease in nuclear abundance of c-Fos proteins may account for the minimal c-Fos DNA binding in the CL nuclear extracts. At least two studies support the concept of shuttling c-Fos between the cytoplasmic and nuclear fractions. In the frog Rana esculata, analysis of brain and testis extracts showed a redistribution of c-Fos DNA-binding activity from the cytoplasm to the nucleus with the onset of reproductive activity in these tissues [40, 41]. Our c-Fos DNA-binding data in CL differ from that reported for the hCG-primed rat, in which luteal WCE exhibited intense c-Fos binding to an AP-1 DNA consensus probe [16]. One interpretation for this difference may be that the cytoplasmic component of the rat luteal extracts contained the c-Fos binding activity; alternatively, this may represent a species- or model-specific difference. In addition, the phosphorylation status of c-Fos in CL may alter its DNA-binding ability since phosphorylated c-Fos has increased stability and transactivating ability [42]. In our immunoblots, c-Fos had a slightly retarded mobility in nuclear fractions compared to cytoplasmic extracts in preovulatory structures, but not CL. This retarded mobility may represent a phosphorylated isoform of c-Fos that facilitates DNA binding. However, the phosphorylation of c-Fos in these tissues remains to be determined.
c-Jun had a similar pattern of expression as c-Fos, with lower whole-cell protein expression in CL; however, EMSA results showed that c-Jun binding was present in both granulosa cells and in CL. Therefore, in CL, c-Jun appears to form DNA complexes with AP-1 factors other than c-Fos. Follicular development prior to the LH surge entails extensive proliferation, which might explain the predominant presence of c-Jun and c-Fos in preovulatory stages, as positive activators of proliferation as observed for other cells [2]. The lower protein expression of c-Fos and c-Jun in CL (immunoblot data) and reduced abundance of nuclear c-Fos DNA binding in CL (EMSA data) suggest a decreased need for these factors as differentiation processes dominate over proliferation.
Fra-1 immunolocalization was similar to c-Jun and c-Fos, but whole-cell, cytoplasmic, and nuclear protein abundance appeared developmentally unregulated by immunoblot analyses. Cytoplasmic protein extracts were similar to nuclear extracts in immunoreactive band intensity at all tissues assessed, yet nuclear extracts lacked the capacity to bind to the AP-1 consensus oligonucleotide. Although Fra-1 lacks transactivating potential, it still has the capacity to bind DNA [42], so a possible explanation may be that other, more abundant or higher-affinity Jun/Fos family members may exclude Fra-1 from binding these AP-1 sites or that posttranslational modifications may alter is DNA-binding ability.
FosB immunoreactive staining was present in only few granulosa and theca nuclei with little variation between different stage follicles. In CL, immunohistochemical staining of large luteal cell nuclei for FosB was very weak. In all ovarian structures, a consistent although weak cytoplasmic staining for FosB was observed. This observation is consistent with our cytoplasmic/nuclear fraction immunoblot analyses that showed that with the exception of granulosa cells from small antral follicles, FosB was localized primarily to the cytoplasmic fraction. DNA-binding activity of FosB in granulosa cell and CL nuclear extracts was minimal. Taken together our data suggest that there is little nuclear FosB present in the ovarian structures analyzed. The minimal DNA-binding activity of FosB is most likely a nuclear abundance issue since FosB heterodimers are usually very stable [43].
Immunoblot data revealed the presence of JunB in follicular, granulosa, and CL whole-cell extracts. No notable differences were found between different stage follicles in whole-cell extracts. This was consistent with our immunohistochemistry data and our EMSA results. However, when nuclear and cytoplasmic fractions of granulosa cells from 8–10-mm antral follicles were isolated, they consistently showed an increase in the nuclear abundance of JunB compared to other structures on the same blots. This discrepant observation between WCE and the nuclear extracts may be that the nuclear fraction typically is enriched for transcription factors. A strong signal in both granulosa cell and CL nuclei was detected by immunohistochemistry. Also, a shift was observed from the AP-1-DNA complex with anti-JunB antibody in granulosa nuclear extracts as well as CL nuclear extracts. Few theca cells were immunopositive for JunB. A previous report showed that granulosa JunB protein expression can be dramatically and transiently (within 2–4 h) induced by FSH and hCG in hypophysectomized-estradiol-primed rats but was negligible in unchallenged follicles [16]. Our sampling protocol most likely would not detect short-term changes as observed in the aforementioned rat study. Taken together, our data indicate a constant presence of JunB in the nuclei of granulosa and postovulatory luteal cells, with a nuclear increase only in the immediate preovulatory period.
The presence of different AP-1 factors in ovarian blood vessels has not been previously investigated. Our immunohistochemical findings revealed that JunD, JunB, and Fra-2 are strongly expressed in ovarian VSMC. c-Jun, c-Fos, and Fra-1 were also expressed but less frequently. A previous report in cultured rat aortic VSMC [44] detected c-Fos, c-Jun, and JunB expression in response to ERK pathway agonist, with complexes formed between c-Fos and JunB as the most prevalent. Another study of rat VSMC detected Fra-1 and Fra-2 proteins [45]. Our results showed that endothelial cell nuclei staining was weak and variable for JunD, c-Jun, and c-Fos. This is consistent with a previous report in which endothelial cells in culture presented JunD and Fra-2 under nonstimulated conditions and all AP-1 factors after protein kinase C agonist stimulation [46].
By immunohistochemistry, we found that JunD is consistently present in oocyte nuclei throughout the follicular cycle. Literature on AP-1 factors in porcine oocytes is limited to a recent study showing the presence of c-Fos mRNA [47]. The oocyte stained for other AP-1 factors, including c-Jun, JunB, c-Fos, Fra-1, Fra-2, and FosB, but this was less consistent. Some of the oocyte staining may be nonspecific, as oocytes stained weakly when IgG was substituted for the primary antibody or when blocking peptide was utilized.
In conclusion, our results suggest that expression and DNA-binding activity of AP-1 factors in follicular structures varies with the stage of estrous cycle. Perhaps ovarian signaling cascades that use similar intermediate kinases can be better understood by knowing the relative abundance and developmental distribution of these AP-1 factors. Future studies that address the in vivo and in vitro posttranslational modifications (such as phosphorylation status) of this transcription factor family are needed to help elucidate the role of AP-1 in ovarian gene regulation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Holly A. LaVoie, Department of Cell and Developmental Biology and Anatomy, School of Medicine, Building 1, University of South Carolina, Columbia, SC 29209. FAX: 803 733 3212; hlavoie{at}med.sc.edu
Received: 4 December 2002.
First decision: 28 December 2002.
Accepted: 5 February 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. E. Beshay, J. C. Havelock, R. Sirianni, P. Ye, T. Suzuki, W. E. Rainey, and B. R. Carr The Mechanism for Protein Kinase C Inhibition of Androgen Production and 17{alpha}-Hydroxylase Expression in a Theca Cell Tumor Model J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4802 - 4809. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Cobellis, M. Lombardi, D. Scarpa, G. Izzo, G. Fienga, R. Meccariello, R. Pierantoni, and S. Fasano Fra1 Activity in the Frog, Rana esculenta, Testis: A New Potential Role in Sperm Transport Biol Reprod, May 1, 2005; 72(5): 1101 - 1108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.-S. Son, K. Y. Arai, K. F. Roby, and P. F. Terranova Tumor Necrosis Factor {alpha} (TNF) Increases Granulosa Cell Proliferation: Dependence on c-Jun and TNF Receptor Type 1 Endocrinology, March 1, 2004; 145(3): 1218 - 1226. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
