| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Molecular Reproduction Research Laboratory, Clinical Research Institute of Montreal, Montreal, Quebec, Canada H2W 1R7
b Division of Experimental Medicine, Department of Medicine, McGill University, Montreal, Quebec, Canada H3A 1A3
c Faculty of Medicine, University of Montreal, Montreal, Quebec, Canada H3T 1J4
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
follicle-stimulating hormone receptor, gene regulation, granulosa cells, ovary, Sertoli cells, steroid hormone receptors, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Inspection of the sequence of the putative promoter regions of FSHR in four species (human, mouse, rat, and sheep) has shown that they have a few conserved cis-elements, including TGACC-motif, activating protein-1 (AP-1), steroidogenic factor-1 (SF-1)-binding sites, CACC-box, E-box, and initiator region [7–10]. Previous studies suggest that SF-1 binding to the promoter may participate in cell-specific regulation of FSHR expression, and interaction of SF-1 with upstream stimulatory factor (USF) is required for activation of the FSHR gene [9, 11]. Recent experiments found that E2F and GATA-1 also regulated FSHR expression as activators [12]. However, tissue- and stage-specific expression of genes is dependent on the relative levels of cellular repressors and activators and their binding affinities as well as on chromatin remodeling induced by hormones or other transcription factors [13, 14]. In addition, different cellular combinations of transcription factors interact with the same promoter regions, leading to the type-specific transcription in the cells. Moreover, RNA processing and posttranslational modification also determine the functional protein levels of FSHR on the cell membrane [15, 16]. Although some of the mechanisms by which the FSHR gene is down-regulated by FSH stimulation and protein kinase A signal transduction pathway have recently been proposed, little detailed information has been reported so far regarding the factors that lead to suppression of the FSHR gene [11, 16].
In our previous studies, we have analyzed the regulatory function of sequences extending over 1.2 kilobases (kb) of the 5'-flanking region of the ovine FSHR (oFSHR) gene [17]. Promoter-deletion studies revealed several positive and negative regulatory regions. The strongest promoter was localized from -200 to +163 relative to the transcription start site. A 94-base pair (bp) deletion from -200 to -106 of the strongest promoter dramatically attenuated promoter activity, implying the presence of a strong enhancer in this region [17]. Our DNase I-footprinting assays have mapped two regulatory elements, an E-box (+32 to +57) and a CACC repeats (-46 to -67) near the 3' end of the promoter, but failed to detect a protected region from -200 to -106 on both DNA strands, with the exception of a notable hypersensitive site at -185 [10, 17]. Sequence analyses of the 94-bp promoter region revealed the presence of an imperfect, inverted hormone response element with 5-bp spacing located from -197 to -171. This palindromic element exhibits similarity to several sequences previously reported to bind and to mediate orphan nuclear receptor and AP-1 activities [18–20]. A similar palindromic sequence with 2-bp spacing for chicken ovalbumin upstream promoter-transcription factor (COUP-TF) was also noted at +32 to +57, which was previously mapped as an E-box [17].
In the present report, we have characterized the nuclear factors binding to a third cis-element located from -197 to -171 of the strongest promoter region. Among them, we have identified AP-1 proteins, SF-1, USF-1/2, and COUP-TF, and we have studied their involvement in regulating FSHR gene transcription in two gonadal cell lines. We have also demonstrated that COUP-TF- and USF-binding sites are overlapping at +32 to +57 and, consequently, that binding of these factors mutually interferes with each other, providing a antagonistic mechanism to balance the receptor expression.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasmid pGL3-basic (Promega, Madison, WI) and oFSHR promoter/luciferase constructs were described previously [17]. Mouse COUP-TF expression vectors (pCR3.1COUP-TFI, pCR3.1COUP-TFII) were kindly provided by Dr. Ming-Jer Tsai (Baylor College of Medicine, Houston, TX). SF-1 expression vector (pCMV5SF-1) was a generous gift from Dr. Keith L. Parker (University of Texas, Southwestern Medical Center, Dallas, TX). Mouse Sertoli cell line (15P1) was given by Dr. Francois Cuzin (University of Nice, Nice Cedex 2, France) [21]. A porcine ovarian granulosa cell line (JC-410) was established by Dr. Jorge P Chedrese (University of Saskatchewan, Saskatoon, Canada) [22]. Antibody against SF-1 was purchased from Upstate Biotechnology (Lake Placid, NY). Antibody specific to estrogen receptor 
(ER) was a gift from Dr. Geoffrey L. Greene (University of Chicago, Chicago, IL). Antibodies against USF-1, USF-2, c-Fos, c-Jun, COUP-TFI, COUP-TFII, preimmune rabbit immunoglobulin (Ig) G, and preimmune goat IgG were products of Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).
Cell Culture and Transient Transfections
Cell culture and transient transfections were carried out as described previously [17, 23]. The cells were transfected with 5 µg of luciferase reporter and 0.5 µg of a pCMVlacZ in the presence or absence of 2.5 µg of pCMV5SF-1 and/or 2.5 µg of pCR3.1COUP-TF. After 48 h of incubation, cells were lysed for both luciferase and ß-galactosidase assays as described previously [17]. The relative luciferase activity was normalized to ß-galactosidase activity and expressed as the fold-induction over the activity of the proper control vector.
Electrophoretic Mobility Shift Assay
All oligonucleotides were commercially synthesized from BioCorp, Inc. (Montreal, QC, Canada). Nuclear extract preparation and electrophoretic mobility shift assay (EMSA) were performed as described previously [24]. Briefly, double-strand DNA (dsDNA) was labeled at the 5' ends using T4 polynucleotide kinase and [
-32P]ATP. Nuclear extract (10 µg) from either 15P1 or JC-410 cells was incubated in a binding buffer [17] containing 50 µg/ml of poly dI-dC and 20 fmole of labeled DNA probe. After addition of the radiolabeled probe, the mixture was incubated at room temperature for another 20 min. For antibody supershift assay, the reaction was incubated at room temperature for an additional 15 min after adding 1 µg of antibody, the source of which has been described above. Excess unlabeled DNA competitors were added 5 min before the radiolabeled DNA probe was added. Gels were dried and visualized by autoradiography.
Site-Directed Mutagenesis
Mutagenesis was performed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). In brief, P363 construct containing a wild-type oFSHR promoter fused to a luciferase gene was used as a template [17]. A pair of synthetic oligonucleotides with a directed mutation was extended by means of Pfu turbo DNA polymerase. The extension product was treated with 10 U of Dpn I endonuclease at 37°C for 1 h. The nicked-vector DNA incorporating a mutation was then transformed into Escherichia coli XL1-Blue Supercompetent Cells. Individual clones were isolated, and the purified DNA was sequenced to confirm the correct mutations.
Immunohistochemistry
Fixed sections from 3-mo-old mouse ovary were first deparaffinized. Antigenic sites were unmasked by microwaving slides in 10 mM sodium citrate (pH 5.65) for 5 min. The sections were treated with 3% (v/v) H2O2 of peroxidase-blocking solution at room temperature for 10 min and then incubated with primary antibody (2 µg/ml) of COUP-TFI, COUP-TFII, retinoic acid receptor X (RXR), and preimmune IgG, respectively, at 4°C overnight. The sections were washed in PBS (pH 7.4) containing 0.1% (v/v) Triton X-100 and processed for immunostaining using the ABC Staining System according to manufacturer's protocols (Santa Cruz Biotechnology) [25].
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous promoter deletion studies suggested a possible presence of important regulatory elements between -200 and -106 of the oFSHR promoter [17]. To identify these cis-elements, we labeled a 94-bp polymerase chain reaction fragment encompassing from -199 to -106 of the oFSHR promoter (FSHRP-199/-106) and performed EMSA. Using this probe, our EMSAs detected two apparent protein-DNA complexes with the nuclear extract from JC-410 cells (Fig. 1). These bindings appeared to be sequence-specific, because they were fully competed by unlabeled FSHRP-199/-106 dsDNA but not by an unrelated glucocorticoid response element (GRE) (Fig. 1). However, complex I was partially displaced by estrogen response element (ERE) but unaffected by SF-1 response element (SFRE), whereas complex II was greatly diminished in the presence of SFRE (Fig. 1). Other smaller bands appeared to be nonspecific or to have resulted from degraded protein.
|
This experiment suggested that the sequence motif of 
that is highly homologous to ERE and SFRE might participate in the protein-DNA interactions. A close inspection of the sequence revealed similarity to a CACnNTG (E-box) motif, AP-1-binding site (TGAC/GTACA), and an inverted hormone response element with 5-bp spacing. It was reasonable to assume that nuclear factors, including many orphan receptors, helix-loop-helix family, and AP-1 proteins, may be involved in the observed protein binding. Therefore, we synthesized oligonucleotides that spanned sequences from -197 to -171 of the regulatory element 3 (RE-3) and carried out gel competition experiments (Fig. 2). Two predominant protein-DNA complexes were formed in the presence of nuclear extract from JC-410 cells, but only one was detected with nuclear extract from 15P1 cells (Fig. 2). The results suggest that 15P1 cells do not express endogenous complex II protein but, rather, have similar proteins present in complex I. A close examination indicated that the complex I contained at least two retarded bands overlapping each other (Fig. 2). The complex Ia, migrating a little slower, was displaced by unlabeled RE-3 and AP-1 regulatory element (Ap-1), partially by ERE and cAMP response element (CRE), but remained unchanged in the presence of retinoic acid response element (RARE), retinoic acid-related orphan receptor response element (RORE), SFRE, and AP-2 regulatory element (Ap-2) (Fig. 2). The complex Ib was competed by ERE and unlabeled RE-3 but not by RARE, RORE, CRE, SF-1, Ap-1 and Ap-2. The DNA-protein complex II virtually disappeared in the presence of 200-fold molar excess of SFRE, but it remained unchanged with other nonspecific competitors (Fig. 2). The band that migrated faster than the complex II appeared to be nonspecific, because it was not competed by any of the competitors (Fig. 2).
|
To identify nuclear proteins that bound to RE-3 sequence, we incubated nuclear extract with antibodies against SF-1, ER, USF-1/2, c-Fos/c-Jun, and preimmune rabbit IgG (Fig. 3). The DNA-protein complex Ia was supershifted in the presence of either anti-c-Fos or antibody specific to phosphorylated c-Jun, but it was unaffected by rabbit IgG or antibody H222 against ER (Fig. 3). It was notable that antibody specific to c-Fos supershifted almost all the complex Ia, leaving the complex Ib distinguishable (Fig. 3, lane 6). The complex Ib was supershifted by both USF-1 and USF-2 antibody, whereas DNA-protein complex II disappeared only in the presence of antibody against SF-1 (Fig. 3). These data would be consistent with USF-1/USF-2 heterodimer, c-Fos/c-Jun heterodimer, and SF-1 monomer independently binding to the promoter, because respective antibodies individually shifted three complexes (Fig. 3). However, the majority of proteins that bound to this particular composite RE-3 seemed to be c-Fos/c-Jun as a heterodimer. The smaller band was concluded to be nonspecific, and its intensity varied in some lanes.
|
COUP-TFI and COUP-TFII Bind to RE-3 and E-Box Region of the Promoter
Previous studies demonstrated that SF-1-, USF-, and AP-1-binding sites were also recognized by COUP-TFs or retinoic acid receptors (RARs) [26–28]. To test this possibility, we performed the same gel supershift assays but with nuclear extract from JC-410 cells that overexpressed COUP-TFs. Our experiments showed that COUP-TFI and COUP-TFII bound to RE-3 of the promoter, because a part of DNA-protein complex I migrated slower after adding anti-COUP-TFI or anti-COUP-TFII antibody (Fig. 4). However, the same complex was not affected by anti-RAR, anti-RARß, or anti-RAR antibody, and both complex I and II were greatly reduced in the presence of anti-RXR antibody, which cross-reacts with all three isotypes (, ß, and ) (Fig. 4, lane 6). The experiments suggest that COUP-TFs preferentially bind to RE-3 as homodimers, which are thought to be more stable than COUP-TF/RXR heterodimer [29]. A similar orphan receptor response element (ORRE) could be found at +32 and +54, which was previously mapped as footprinting 1 (FP1) of oFSHR promoter [17]. Using 15P1 cells, we were able to provide evidence that COUP-TFs also bound to FP1 region by showing that the retarded bands were supershifted by both COUP-TFI and COUP-TFII antibodies (Fig. 5).
|
|
To assess whether COUP-TFI, COUP-TFII, and RXR are naturally expressed in granulosa cells, we performed immunohistochemistry with the ovarian sections from 3-mo-old mice (Fig. 6). Antibody against RXR, COUP-TFI, or COUP-TFII specifically recognized the corresponding protein expressed in the granulosa cells of mouse ovarian follicles (Fig. 6, B–D). In contrast, preimmune IgG from either rabbit or goat did not show any positive staining (Fig. 6, A and F). Similar results were also observed using sheep ovarian sections (data not shown). These data, together with earlier observations that COUP-TFs were expressed in the granulosa and Sertoli cells from other species, imply that orphan receptor COUP-TFs are able to regulate a granulosa- or Sertoli cell-specific gene expression [18, 20, 30–33].
|
SF-1, AP-1, and USF Mediate Basal and Phorbol 12-Myristate 13-Acetate-Induced Promoter Activities
Sequence analyses found that, in the RE-3 sequence, the central nucleotides of CACAGG were thought to be important for AP-1, SF-1, and USF binding [34–36]. Among them, the first CA is crucial for AP-1 binding, and the second AGG is required for both USF and SF-1/COUP-TF binding. An individual mutation for AP-1, SF-1, or USF was found to be difficult, because these binding sites are overlapping each other and mutation of one particular binding site seriously affects the other's binding affinity (data not shown). Therefore, we made a block replacement by mutating the central CA to gg and AGG to cGc, and we incubated the mutated RE-3 probe (RE-3mt) with nuclear extract to test protein binding (Fig. 7). Our gel competition experiments found that RE-3mt as well as RORE were ineffective in competing for wild-type probe binding (RE-3) even at 200-fold molar excess (Fig. 7). In contrast, the specific competitor unlabeled RE-3 prevented binding (Fig. 7). Moreover, 32P-labeled RE-3mt probe was not retarded by the nuclear proteins (Fig. 7), suggesting that extra flanking sequences upstream and downstream of the core RE-3 sequence, which were designed for site-directed mutagenesis, could not affect the competition results. Hence, we incorporated the same mutations as the RE-3mt into an FSHR promoter/luciferase reporter P363 to generate a mutant reporter P363m2. We transiently transfected wild-type (P363) as well as mutant (P363m2) reporters into JC-410 and 15P1 cells to analyze the luciferase activities after phorbol 12-myristate 13-acetate (TPA) stimulation (Fig. 8). In agreement with the protein-binding assays shown in Figures 3 and 7, luciferase activity in JC-410 cells transfected with P363 was increased by 100% on 100 nM TPA stimulation (Fig. 8A). However, the mutation of RE-3 in the core promoter not only completely abolished TPA-mediated induction but also diminished basal promoter activity by 80% (Fig. 8A). A similar TPA stimulation was also seen when the P363 reporter was transfected into 15P1 cells that lack SF-1 (Fig. 8B). Consistent with the observation in JC-410 cells, the reporter P363m2 containing a mutant RE-3 but an intact E-box downstream of the transcription start site failed to respond to TPA treatment, indicating that AP-1 mediated TPA activation (Fig. 8). The difference in luciferase activity from the same reporters between JC-410 and 15P1 cells could be cell-line dependent, as discussed previously [17], because 15P1 cells lack expression of endogenous SF-1. These results, together with previous findings, suggested that USF-1/2, c-Fos/c-Jun, and SF-1 might be activators, and that some of these factors mediated protein kinase C agonist-mediated stimulation.
|
|
The effect of SF-1 on promoter activity was examined in 15P1 cells by cotransfecting oFSHR promoter containing wild-type or mutant RE-3 with or without SF-1 expression vector. In transient transfection assays, overexpression of SF-1 stimulated FSHR promoter activity by 46% (Fig. 9). Consistent with the results of DNA-binding assays (Figs. 2 and 7), RE-3 mutation in the core promoter significantly diminished promoter activity by approximately 75% (Figs. 8 and 9A), as described before, and abolished SF-1-mediated activation (Fig. 9A).
|
COUP-TF Represses SF-1, AP-1, and USF Activation of oFSHR Promoter
To study COUP-TF function, transfections were conducted using P363 reporter together with SF-1 and/or COUP-TF expression vectors as illustrated in Figure 9B. Consistent with other studies in which COUP-TFs antagonized SF-1 activation on various gene promoters in Sertoli and granulosa cells [13, 37, 38], overexpression of either COUP-TFI or COUP-TFII inhibited the promoter activity by approximately 60% in 15P1 cells without SF-1 (Fig. 9B; compare bars 1, 3, and 4) and by 50% in the cells with SF-1 (Fig. 9B; compare bars 2, 5, and 6). However, overexpression of SF-1 slightly overcame COUP-TFI or COUP-TF-II repression by increasing basal level of the promoter activity (Fig. 9B; compare bars 1, 3, and 5).
Our gel supershift assays demonstrated that COUP-TFs also bound to FP1 region (Fig. 5) [17]. To assess their relative contributions, we tried to individually mutate either core E-box sequence or ORRE while leaving the other intact. We mutated GT to ca and AC to tg within ORRE (GGcaCACGTGtg), and the central CG of E-box to at (GGTCAatTGACC) [17] (Fig. 10). To distinguish the small amount of COUP-TFI protein from the abundant USFs that were present, we carried out an antibody supershift assay as described in Figure 5. Our experiments found that COUP-TFI failed to bind to FP1m1 oligonucleotides but bound efficiently to FP1 and mildly to FP1m2 probes (Fig. 10). However, mutation of FP1m2 completely abolished USF binding, whereas mutation of the ORRE sequence adjacent to E-box (FP1m1) also mildly reduced USF binding (Fig. 10).
|
We then incorporated the same mutations as the FP1m1 into P363 and P363m2 reporters and tested their functions in 15P1 cells (Fig. 11). In agreement with the protein binding in the EMSAs, RE-3 mutation resulted in 80% reduction of promoter activity, whereas ORRE mutation diminished promoter activity by 60%, compared to the luciferase activity driven by wild-type promoter (Fig. 11). The greater reduction in luciferase activity driven by the promoter containing ORRE mutation may be due to reduced USF binding and disrupted functional interaction between USF and other proteins [17, 39]. Interestingly, the promoter containing a mutant RE-3 was still responsive to overexpression of COUP-TFI in the transfected 15P1 cells by showing a decreased basal promoter activity (Fig. 11). Obviously, this remaining repression could be attributable to the downstream ORRE (Fig. 5). Similarly, a mutation of ORRE alone in the strongest promoter did not abolish promoter response to COUP-TFI (Fig. 11). However, double mutations of RE-3 and ORRE in P363 completely alleviated COUP-TFI repression (Fig. 11). Because the two regulatory elements were recognized by both positive and negative transcription factors, simple mutation of the negative part alone in the complex element was impossible (data not shown). However, it was notable that a constant 50–60% inhibition of transcription by overexpression of COUP-TFI in Sertoli cells was seen with one or two intact composite sites (Fig. 11), suggesting that both embedded COUP-TF-binding sites are required to balance transcription in these cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Direct characterization of the oFSHR promoter in primary granulosa and Sertoli cells from the same species is difficult, as discussed previously [17]. However, use of two gonadal cell lines from other species has provided us with a feasible alternative because of the conservation of transcription factors and regulatory elements of the promoters among species [9, 10, 40]. Evidence supporting this was strengthened by the present observation that the E-boxes located in the core promoters of human, mouse, and rat FSHR genes are highly conserved and believed to bear similar composite sequences overlapping COUP-TF- and AP-1-binding sites (
[9, 40]. In addition, the imperfect direct repeats/palindromic sequences containing a TGACC-motif are also found to embed AP-1-binding sites, which are putative E-boxes in the promoters of other species [8, 9, 40, 41]. For example, a direct repeat 
and an inverted repeat 
are located from -384 to -399 and from -678 to -694, respectively, relative to the first nucleotide of the translation initiation codon of the mouse FSHR promoter [42]. A direct repeat 
from -210 to -222 in the human FSHR promoter and multiple TGACC-like repeats 
in a region from -639 to -664 of the rat FSHR promoter seem to be conserved [8, 41]. The existence of a common cognate site by the proteins with antagonistic functions implies that the net regulation of the FSHR gene may result from the relative availability of repressors and activators in a given physiological state. This may also contribute to the differential expression of the FSHR gene in gonadal and nongonadal tissues in a stage-dependent manner in many species, as reported for several other genes [20, 37, 43].
The expression of SF-1 is restricted to the anterior pituitary gland and the steroidogenic tissues, including the ovaries, testes, and adrenal glands [44–46]. Although previous studies showed that SF-1 might contribute to cell-specific activity of mouse FSHR promoter [9], SF-1 alone may not be the sole mediator governing a Sertoli or granulosa cell-specific expression [19, 47]. Additional elements and factors must synergize or antagonize SF-1 to direct the expression in a developmental manner [11, 12, 16]. In the present study, we showed that three activators of USF, AP-1, and SF-1 bound a composite cis-element (Figs. 3 and 4). Our experiments failed to detect direct protein-protein associations, because antibodies against AP-1, USF, COUP-TF, and SF-1 appeared to supershift separate complexes (Figs. 3–5). However, based on functional studies of USFs with AP-1 and SF-1 [11, 35, 48], it is reasonable to postulate that they could functionally synergize via integrators such as cAMP response element binding protein (CREB)/p300 to confer optimal promoter activity, as summarized by the proposed model in which the COUP-TFs are absent or excluded from their recognition sites (Fig. 12A).
|
The COUP-TFs are ubiquitously expressed and seem to play complex roles in the regulation of genes important for cell differentiation, embryonic development, and metabolic homeostasis [49]. The COUP-TFs generally seem to function as strong transcriptional repressors [18, 30], but they have also been shown to positively regulate gene expression in some cases [50]. In the present study, we have shown that, in gonadal cells, COUP-TFI and COUP-TFII did not activate transcription of oFSHR gene via RE-3 and E-box/ORRE but did interfere with activation mediated by c-Fos/c-Jun, USF-1/2, and SF-1 (Figs. 9 and 11). Previous studies have shown that most cells that do not express SF-1 do express COUP-TFs [18, 19, 49]. Therefore, it is possible that, in these cells, the repressor activity of COUP-TFs are not antagonized by any other activator, thereby suppressing gene expression [18]. This transcriptional inhibition of COUP-TFs could be exerted by several mechanisms [49]. However, in the case of oFSHR promoter, we believe that it is more likely to result from the direct silencing function of COUP-TF rather than from competing with, quenching, or titrating out a hormonal partner, because a dramatic reduction of the promoter activity was seen in our transfection studies even though only a small amount of COUP-TF binding was detected in the EMSAs (Figs. 4, 5, and 11). Of course, their repressive functions may be enhanced by interaction with two common nuclear receptor corepressors: NcoR (nuclear receptor corepressor), and SMRT (silencing mediator of retinoic acid and thyroid hormone receptor). This is illustrated in Figure 12B, where COUP-TFs occupy the composite regulatory sites [48].
Multiple positive factors competing for the same element of FSHR promoter may provide an alternative means to regulate the basal and inducible transcription in a given physiological condition (Fig. 12A). This is supported by observations that SF-1 expression and binding to DNA were significantly increased in hormone-stimulated differentiated granulosa cells of mature follicles compared to the undifferentiated cells of immature follicles [20, 37]. As opposed to SF-1, an increase in the expression and DNA binding of COUP-TF was seen during luteal differentiation [20, 37]. Interestingly, these expression patterns correlated with the stage-specific expression of FSHR during folliculogenesis [4, 6]. However, the dramatic up- and down-regulation of FSHR transcription cannot be explained by the exclusive binding of SF-1 or COUP-TF. Interactions of SF-1 and/or COUP-TF with other regulatory proteins, or specific modification of the transcription factors, is needed for the high up-regulation of the FSHR gene. In addition, functional changes are likely to involve the selective phosphorylation of either factor or of a coregulatory molecule. Our experiments have demonstrated that the TPA induction of oFSHR promoter is mediated by the RE-3 site and that the extent of stimulation is modulated by factors that bind to this region (Fig. 8). Therefore, phosphorylation of SF-1 and AP-1 proteins would be predicted to recruit coactivators and to increase the activity of oFSHR promoter (Fig. 12A). Such protein modifications in differentiated gonadal cells may be a critical event in increasing both basal as well as stimulated expression of the FSHR gene. In addition, it is possible that CREB also binds to RE-3, as suggested in our competition experiments (Fig. 2). However, further study is needed to understand the role of this protein in regulation of the oFSHR promoter.
In conclusion, we have demonstrated, to our knowledge for the first time, that the oFSHR gene is tightly regulated by composite cis-elements that are recognized by both positive and negative factors. These mutually antagonistic interactions may contribute to tissue- and stage-specific expression of the FSHR gene. However, the ultimate test of this hypothesis requires examination of the activity of promoter containing a mutation of the key element in transgenic mice.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by grants from the Canadian Institutes of Health Research (CIHR). W.X. and N.D. hold doctoral fellowship awards of the CIHR. 
2 Correspondence: M. Ram Sairam, Molecular Reproduction Research Laboratory, Clinical Research Institute of Montreal, 110 Pine Avenue West, Montreal, QC, Canada H2W 1R7. FAX: 514 987 5585; sairamm{at}ircm.qc.ca
Accepted: December 20, 2001.
Received: September 17, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-macroglobulin (2M) gene in rat ovarian granulosa cells. Mol Endocrinol 1998; 12:1393-1409 or TR ß in intact cells. Nucleic Acids Res 1995; 23:4143-4150-promoter by steroidogenic factor-1 and cyclic adenosine 3',5'-monophosphate. Mol Endocrinol 2000; 14:66-81 A-crystallin gene: involvement of Pax-6, USF, CREB and/or CREM, and AP-1 proteins. Mol Cell Biol 1994; 14:7363-7376This article has been cited by other articles:
|
|
 |
|
  S. Mazaud Guittot, A. Tetu, E. Legault, N. Pilon, D. W. Silversides, and R. S. Viger The Proximal Gata4 Promoter Directs Reporter Gene Expression to Sertoli Cells During Mouse Gonadal Development Biol Reprod, January 1, 2007; 76(1): 85 - 95. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Luo and M. C. Wiltbank Distinct Regulation by Steroids of Messenger RNAs for FSHR and CYP19A1 in Bovine Granulosa Cells Biol Reprod, August 1, 2006; 75(2): 217 - 225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. P. Hermann and L. L. Heckert Silencing of Fshr Occurs through a Conserved, Hypersensitive Site in the First Intron Mol. Endocrinol., August 1, 2005; 19(8): 2112 - 2131. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Xing and M. R. Sairam Retinoic Acid Mediates Transcriptional Repression of Ovine Follicle-Stimulating Hormone Receptor Gene via a Pleiotropic Nuclear Receptor Response Element Biol Reprod, July 1, 2002; 67(1): 204 - 211. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||
