| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas 77030
Address all correspondence and requests for reprints to: JoAnne S. Richards, Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030. E-mail: joanner{at}bcm.tmc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
ADAMTS-1 is a secreted, active metalloproteinase (7) with multiple potential functions in ovulating follicles. The protein associates with extracellular matrix molecules through C-terminal thrombospondin-like motifs (8), which are additionally thought to be responsible for its antiangiogenic activity (9). This multifunctional protease is capable of cleaving matrix proteogylcans such as versican (10), aggrecan (11), and brevican (12). Versican is expressed at all stages of follicular growth but is dramatically up-regulated in preovulatory follicles (13). Within preovulatory follicles, ADAMTS-1 and versican colocalize to the expanded matrix surrounding cumulus-oocyte-complexes (6). Significantly, cleavage of versican in ovaries of PRKO mice is reduced compared with PR heterozygote (PRHET) mice (6). In addition, a role for ADAMTS-1 in the ovulatory process is supported by the evidence that mice null for ADAMTS-1 exhibit abnormal ovarian morphology and reduced fertility (14, 15). These data suggest that ADAMTS-1 is a PR-induced protease with functional importance in ovulation.
The aim of this study was to determine how LH and PR regulate ADAMTS-1 gene transcription in granulosa cells. LH acts primarily via activation of adenylyl cyclase and the production of cAMP (16). Although the cAMP response element binding protein is a principal target of cAMP-dependent phosphorylation and activation by protein kinase A, recent studies show that cAMP also regulates the expression and activation of other transcription factors (16). These include specificity protein 1 (Sp1), CAAT enhancer binding protein ß (C/EBPß) (17), and early growth response protein 1 (Egr-1), (18, 19) that are relevant to studies described herein. Recently, it has been reported that Sp1/Sp3 binding sites within the PR and cathepsin L promoters are critical for their transactivation by LH (20, 21). It is also likely that LH impacts ADAMTS-1 induction by the phosphorylation and activation of other coregulatory molecules, including cAMP response element binding protein-binding protein (16).
In most reproductive tissues, including ovarian granulosa cells, PR is expressed as two protein isoforms, PRA and PRB. The two isoforms originate from the same gene and are generated by differential translation of a single transcript (22). The two isoforms are identical with the exception of the amino terminus of PRB, which contains an additional 164 amino acids (23). In granulosa cells, the PRA form is predominantly expressed; however, PRB is also present (24, 25). Additionally, mice null for PRA (PRAKO) but not PRB (PRBKO) exhibit defects in ovulation (26). Classically, PRA and PRB mediate gene expression by binding to consensus PR binding elements (PREs) within the promoters of specific genes, as reported for the IGF binding protein-1 and mouse mammary tumor virus (MMTV) (27) genes. In addition, PR has been shown to regulate transcriptional activity of promoters via protein-protein interactions. Specifically, PR interacts with Sp1 and the Sp1-related protein BTEB, to indirectly regulate target genes, i.e. PR (28), cyclin-dependent kinase inhibitor (p21WAF1) (29), and glycodelin (30). Regulation of transcription by PR-protein interactions may not necessarily be ligand dependent as recently described for the cell cycle inhibitor p21, ectodermal-neural cortex 1, PCDH1 (protocadherin; a cell adhesion molecule), and prolactin receptor (31).
Specific effects of both LH and PR may be critical for transcription of the ADAMTS-1 gene after the LH surge. Therefore, this study aimed to identify regulatory regions and associated transcription factors necessary for not only basal activity but also LH- and PR-mediated induction of the proximal murine ADAMTS-1 promoter. For these experiments, we isolated the mouse ADAMTS-1 promoter, generated ADAMTS-1 promoter-luciferase reporter constructs, and transfected these into primary cultures of rat granulosa cells. Our data indicate that the effects of LH and PR are mediated by independent but overlapping pathways that converge on the ADAMTS-1 promoter, and that the regulation of ADAMTS-1 by PR in granulosa cells is mediated by other DNA binding transcription factors.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). ADAMTS-1 mRNA was increased further in response to hCG treatment (4 h) in both PRHET and PRKO mice (14- and 3-fold respectively) as compared with PMSG alone. An additional significant increase (4-fold, P < 0.01 vs. P (PMSG), hCG 4 h PRHET) in ADAMTS-1 mRNA occurred 12 h after hCG treatment in PRHET mice, but not in ovaries from PRKO mice. After 16 h of hCG treatment, ADAMTS-1 message decreased sharply in PRHET ovaries to levels similar to those observed in granulosa cells of the PRKO mice. These data show that ADAMTS-1 expression is increased during follicle growth and development (PMSG) and to a greater extent during the induction of ovulation (P, hCG 12 h). Of note, the response to hCG/LH is biphasic with an immediate PR-independent increase and a delayed, larger PR-dependent component that failed to occur in PRKO ovaries.
|
). The effectiveness of the antagonist was noted first by the marked uterine ballooning observed in the antagonist-treated mice as compared with vehicle treatment. When the ovaries were analyzed, WT mice injected with vehicle exhibited elevated levels of ADAMTS-1 message, whereas mice receiving the PR antagonist showed significantly reduced levels of ADAMTS-1 mRNA (Fig. 1B). In contrast, message levels of PR were increased in response to ZK98299 treatment (Fig. 1B). No overt changes in the morphology of the antagonist-treated follicles was observed, indicating that ADAMTS-1 message was selectively down-regulated by the antagonist without drastic alterations in other aspects of follicular cell function.
Analysis of ADAMTS-1 and PR mRNA Expression in Primary Culture of Rat Granulosa Cells
To determine whether ADAMTS-1 could be induced directly in granulosa cells by specific agonists of LH signaling, primary cultures of rat granulosa cells were used. Some cells were stimulated with FSH + testosterone (T) for short time intervals (2–12 h) to determine the immediate response to these hormones. Other cells were treated with FSH + T for a longer interval (48 h) to study the effects of differentiation. Cells were also treated with forskolin (Fo) plus phorbol myristate acetate (PMA) to mimic LH-like signal transduction (32, 33). ADAMTS-1 mRNA was rapidly but transiently induced after treatment for 2 h with either FSH + T or Fo + PMA (27- and 33-fold, respectively) (Fig. 2A). Expression of ADAMTS-1 decreased between 2 and 4 h and remained low after 8 and 12 h of either FSH + T or Fo + PMA treatment. In differentiated granulosa cells (48 h FSH + T treatment), ADAMTS-1 mRNA levels were low, but further stimulation with Fo + PMA for 2 h induced ADAMTS-1 expression 5-fold over the 48 h FSH + T control. Treatment of these differentiated cells with Fo + PMA for longer time intervals (4, 8, and 12 h) did not sustain ADAMTS-1 expression (data not shown). These data show that ADAMTS-1 mRNA is induced by FSH + T and Fo + PMA signaling cascades, and that this induction is rapid but transient in both immature and differentiated granulosa cells.
|
). The expression of PR peaked after 4 h of Fo + PMA treatment (39-fold over control) and declined after 8 and 12 h. After 48 h of FSH + T, the expression of PR was low but could be reinduced by treatment with Fo + PMA (5-fold over 48 h FSH + T control). These data demonstrate that PR is induced transiently by both FSH + T and Fo + PMA in cultured granulosa cells.
Analysis of ADAMTS-1 Expression in Cultured Granulosa Cells Treated with PR Agonists and Antagonists
To investigate the extent to which endogenous PR mediates the induction of ADAMTS-1 observed in vitro, additional rat granulosa cells were cultured overnight with or without PR antagonist ZK98299 and treated with Fo + PMA for 2 h the next day with or without the PR agonist R5020 (Fig. 3A). Neither R5020 nor ZK98299 significantly altered the induction of ADAMTS-1 in untreated cells. However, treatment of cells overnight with ZK98299 significantly reduced the induction of ADAMTS-1 message in response to Fo + PMA. Because ZK98299 treatment did not completely abolish the effects of Fo + PMA, it is likely that both PR and LH-induced signaling pathways stimulate ADAMTS-1 expression in culture as observed in vivo (Fig. 1A). That R5020 had no effect on the expression of ADAMTS-1 is explained by the fact that the cultured cells secrete significant levels of progesterone (2 ng/ml). Thus, we used a concentration of ZK98299 that was in 1000-fold excess of the progesterone. To ensure that the effect of the antagonist treatment was specific, PR mRNA was measured. PR mRNA levels in untreated and Fo + PMA-treated cells were not affected by ZK98299 treatment (Fig. 3A).
|
). ADAMTS-1 was induced similarly by Fo + PMA in WT and PRKO granulosa cells (Fig. 3B). As expected, PR was expressed and induced (5-fold) by Fo + PMA in WT cells, but not in PRKO cells (Fig. 3B, inset). Thus, a rapid but moderate induction of ADAMTS-1 in cultured granulosa cells, in response to LH/Fo + PMA signaling cascades, is independent of the presence and actions of PR.
Analysis of Transcription Factor Binding Sites on the ADAMTS-1 Promoter
The molecular mechanisms of LH- and PR-regulated transcription of the ADAMTS-1 gene were next investigated by isolating a 5-kb fragment of the gene upstream of the transcriptional start site from a murine genomic library. A smaller fragment (–1186/+266 bp) of the ADAMTS-1 gene was subcloned into the pGL3-Basic vector to create –1186 bp ADAMTS-1-Luc. This promoter fragment includes additional sequence to that previously reported by Kuno et al. (34). No consensus PREs were identified in our analysis of –1186 bp of sequence upstream of the transcriptional start site (Fig. 4A), nor in the analysis of Kuno et al. (34). However, several of the motifs [C/EBPß and guanine cytosine (GC)-I, II, and III] were found to be conserved in mouse, rat, and human (Fig. 4B).
|
) includes a region between –1017 bp and –986 bp that is not recorded in the mouse genome sequencing database. Specific features of the ADAMTS-1 promoter that had not been reported previously include a stretch of Ts (–856/–829 bp) as well as repeats of CA and CT dinucleotides (–540/–640 bp). The major transcription factor binding sites identified by computational analysis included a TATA-box at –32/–27 bp from the transcriptional start site and at least three GC-rich regions: –52/–43 bp (GC-I), –77/–67 bp (GC-II), and –136/–125 bp (GC-III) all of which bind Sp1 and Sp3, as confirmed by supershift analysis (Fig. 5) and competition with an unlabeled probe containing the consensus binding sequence for Sp1/3 (data not shown). Additionally, the GC-II site bound Egr-1 as demonstrated by supershift analysis with the corresponding Egr-1 antibody (Fig. 5). Other sites bind a nuclear factor (NF) 1-like factor (–164/–148 bp); the c-Ets factor GA binding protein, GABP (–311/–290 bp) and CAAT/enhancer binding protein ß, C/EBPß (–370/–360 bp), which were confirmed by competition with unlabeled consensus probe or by supershift analysis with the appropriate antibodies (Fig. 5). Both C/EBPß and Egr-1 are induced in response to the LH surge (17, 18, 19), and when induced bind relevant sites of other promoters (17, 18, 21) and that of ADAMTS-1 as shown in Fig. 5. Sp1/Sp3 is not hormonally regulated in granulosa cells (35), and its binding to the promoter is not affected by hormonal stimulation (Fig. 5). Similar to Sp1/Sp3, GABP and the NF1-like factor bind the ADAMTS-1 promoter in the absence and presence of hormone (Fig. 5). These data validate the putative binding sites reported by Kuno et al. (34), and also identify additional Sp1/Sp3 binding sites (GC-I and III), the NF1-like site, as well as the second c-Ets site. Additionally, all of the above-mentioned sites are highly conserved between mouse, rat, and human (Fig. 4B), suggesting that they play important roles in the expression of the ADAMTS-1 gene.
|
). Fo and Fo + PMA induced activity of the –1186kbADAMTS-1-Luc (5- and 6-fold, respectively), whereas PMA alone activated the promoter to a lesser extent. The –375 bp ADAMTS-1-Luc reporter was found to be sufficient for ADAMTS-1 promoter induction in response to Fo and Fo + PMA, and exhibited higher activity than the –1187 bp ADAMTS-1-Luc construct, indicating that the inducible region in granulosa cells is within the –375 bp promoter.
|
). Deletion of the NF1-like region (–286 bp and –151 bp), or the GC-I site (–61/–43 bp) reduced activity. Specific mutation of GC-III, in the context of –151 bp ADAMTS-1-Luc, led to a reduction in the promoter activity by 4-fold (P < 0.05), as compared with the WT –151 bp ADAMTS-1-Luc. Importantly, elimination of all three GC boxes (–43 bp ADAMTS-1-Luc) resulted in a near complete loss of basal promoter activity. Because the GC-II and C/EBPß sites bind transcription factors (Egr-1 and C/EBPß, respectively) that are induced by LH (17, 19) and because the NF1-like, GC-III, and GC-I sites reside in a region of the promoter that appears to be important for basal activity (Fig. 6B), site-specific mutants of these sites were generated within the –375 bp ADAMTS-1-Luc construct and transfected into cultured rat granulosa cells (Fig. 6C). Importantly, single mutations of the NF1-like binding site, GC-III and GC-I, as well as the double mutant (mC/EBPß/mGC-II), resulted in significant reduction of the basal activity (P < 0.001). Furthermore, the GC triple mutant (GCIII, GCII and GCI) reduced basal activity to a level comparable with that observed when the three sites were deleted by truncation (Fig. 6B).
Despite decreased basal activity in selected mutants, all constructs remained inducible by Fo + PMA (Fig. 6, B and C), indicating that multiple sites contribute to this activity and that ablation of just one site is not sufficient to reduce promoter activity in response to agonist stimulation. Taken together, these data indicate that the binding sites for NF1-like factor, Sp1/Sp3 (GC-III, GC-I), and C/EBPß, together with GC-II, contribute to basal activity of the ADAMTS-1 promoter, whereas multiple sites appear to confer activation by Fo + PMA.
Additional truncations of the promoter in which the TATA box and 5'-untranslated region (UTR) were removed reveal that the sequence containing the TATA box is essential for Fo + PMA-induced activity (Fig. 7). Rat granulosa cells were transiently transfected with the WT –375 bp ADAMTS-1-Luc, a construct missing the 5'-UTR (–375/+1), a construct missing the TATA box in addition to the 5'-UTR (–375 bp/–35 bp), or the empty pGL3-Basic vector. Fo + PMA induced the WT construct and –375 bp/+1 bp-Luc by 8- and 4-fold, respectively, whereas the construct lacking the TATA box and the empty vector were only induced 2-fold by Fo + PMA (similar to pGL3-Basic alone; Fig. 7, inset), indicating that without the region containing the TATA box the ADAMTS-1 promoter is no longer inducible by Fo + PMA.
|
). This approach was taken because the role of endogenous PR in the cultured cells is difficult to assess. Not only are the cells synthesizing progesterone, but also Fo + PMA are needed to induce endogenous PR expression. Cotransfection of –375 bp ADAMTS-1-Luc and increasing amounts (0, 10, 50, and 100 ng/ml) of the PRA expression construct shows that PRA stimulated ADAMTS-1 promoter activity in a dose-dependent manner (Fig. 8A). Cells cotransfected with the –375 bp ADAMTS-1-Luc reporter and 10 ng/ml of the PRA expression construct significantly (P < 0.001) increased luciferase activity as compared with cells transfected with the empty vector. To confirm that an optimal amount of PRA expression vector had been used for induction of ADAMTS-1, additional granulosa cells were cotransfected with the –375 bp ADAMTS-1-Luc construct and lower amounts (0, 2, 5, 10, and 20 ng/ml) of the PRA expression vector (Fig. 8B). As shown, 5–10 ng/ml of the PRA expression vector gave the maximal response.
|
), which is similar to that reported in Fig. 6A. Fo + PMA treatment of granulosa cells also transfected with 10 ng/ml of the PRA expression vector resulted in a cooperative activation of the promoter by approximately 2-fold (Fig. 8A), as compared with Fo + PMA-treated cells transfected with the empty vector. Amounts of exogenous PRA above 10 ng/ml (50 and 100 ng/ml) resulted in lower activation of the ADAMTS-1 reporter with or without Fo + PMA stimulation, which may be due to squelching at excessive overexpression levels.
To determine whether exogenous PRB could also induce ADAMTS-1 promoter activity, granulosa cells were cotransfected with the –375 bp ADAMTS-1 construct in addition to increasing amounts of a PRB expression vector (0, 0.5, 2, 5, and 10 ng/ml) (Fig. 8C). Exogenous PRB activated the ADAMTS-1 promoter-reporter even at the lowest plasmid concentration (0.5 ng/ml). Additionally, the empty pGL3-Basic vector was cotransfected with 10 ng/ml of the PRA and PRB expression constructs to show that the effect of PR was specific to the ADAMTS-1 promoter and not due to background activation of the vector (Fig. 8C, inset).
To analyze further whether the regulation of ADAMTS-1 by exogenous PRA is dependent on functional PR, granulosa cells were cotransfected with –375 bp ADAMTS-1-Luc reporter constructs and either the empty or the PRA expression vector and treated with R5020 for 5 h or with ZK98299 overnight (Fig. 8D). Cells transfected with PRA exhibited significant (P < 0.05) promoter activity, as seen in Fig. 8A. R5020 did not significantly alter promoter activity with or without PRA coexpression. Treatment of cultured granulosa cells with ZK98299 overnight resulted in decreased levels (P < 0.0005) of promoter activity as compared with control cells, indicating that the induction of ADAMTS-1 is dependent on functional PR.
To determine whether ligand-dependent PR transactivation occurs in granulosa cells and whether exogenous PR is functional, the activation of the PR-responsive reporter, GRE2-TATA-Luc, was analyzed in similar cultures (Fig. 8E). Cells cotransfected with the GRE2-TATA-Luc and empty vector exhibited negligible reporter activity regardless of treatment. In cells cotransfected with PRA or PRB, luciferase activity increased dramatically (P < 0.001) compared with the empty vector transfected cells. Promoter activity was even further induced in the presence of exogenous ligand (R5020) and was decreased to levels equal with that of cells lacking exogenous PRA by overnight treatment with the PR antagonist ZK98299. Thus, in rat granulosa cell cultures, exogenous PRA and PRB are more highly transcriptionally active than the endogenous receptors. Furthermore, these results show that endogenous progesterone secreted by the granulosa cell cultures likely serves as a near saturating ligand that can be effectively competed by antagonist.
PR Regulation of the ADAMTS-1 Promoter Is Specific
To confirm that the activation of the ADAMTS-1 promoter by PRA and PRB is specific and not a consequence of nuclear receptor overexpression, granulosa cells were cotransfected with the –375 bp ADAMTS-1-Luc and increasing amounts of estrogen receptor 
(ER) (Fig. 9A). ER is activated by Fo + PMA signaling in granulosa cells (36); however, in cells cotransfected with increasing amounts (0, 2, 10, and 50 ng/ml) of the ER expression vector and the ADAMTS-1 promoter-reporter construct, no enhancement of promoter activity was detected in cells treated with or without Fo + PMA, as compared with the respective controls (0 ng/ml ER± Fo + PMA). To determine whether exogenous ER was functional in granulosa cells the ERE-E1b-Luc construct was cotransfected with the ER expression vector and cells were treated with or without Fo + PMA. Only cells expressing exogenous ER and treated with Fo + PMA exhibited enhanced promoter activation, as observed previously (36). Similar results were found when overexpressing the androgen receptor (data not shown).
|
). COX-2 is an LH-regulated gene in granulosa cells and is the rate-limiting enzyme in prostaglandin biosynthesis (37). The COX-2 promoter has been described previously and is known to contain a functional CRE and E-Box (38). Fo + PMA treatment of cells transfected with the COX-2 promoter-reporter significantly induced promoter activity (Fig. 9B), indicating that it is an LH-regulated gene. However, cotransfection of the PRA expression vector had no significant effect on the activity of the COX-2 promoter (Fig. 9B), indicating that COX-2 is not regulated by PR and that the regulation of the ADAMTS-1 promoter by PR is specific.
PR Regulates the ADAMTS-1 Promoter through Multiple Binding Sites
To determine which sites are regulated by PR within the ADAMTS-1 promoter, rat granulosa cells were cotransfected with truncated promoter-reporter constructs and the PRA expression plasmid (Fig. 10A). The truncated constructs –85 bp, –61 bp ADAMTS-1-Luc, and –43 bp ADAMTS-1-Luc exhibited reduced PRA activation compared with –375 bp ADAMTS-1-Luc (Fig. 10, A and B). These data indicate that elements upstream of –43 bp mediate the PR-regulation of the ADAMTS-1 gene. However, because the basal activity of the truncated constructs was also reduced, site-specific promoter-reporter mutant constructs were used to identify regions of the promoter selectively regulated by PR (Fig. 10C). All transcription factor binding site mutants exhibited significantly (P < 0.001) reduced PRA mediated promoter activity when compared with the WT –375 bp ADAMTS-1-Luc control. These data indicate that PR regulates the ADAMTS-1 gene through multiple transcription factor sites including the C/EBPß, NF1-like factor, and Sp1/Sp3 (GC-I, II, and III) binding sites, each of which has the potential to recruit PR to the ADAMTS-1 promoter via protein-protein interactions.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We show that the PR-independent induction of ADAMTS-1 expression by LH is rapid and sustained for at least 16 h in granulosa cells of PRKO mice in vivo. The induction of ADAMTS-1 by FSH/T or Fo + PMA in granulosa cell cultures is also rapid and occurs equally well in granulosa cells obtained from WT or PRKO mice. Thus, the molecular mechanism by which LH induces ADAMTS-1 uses divergent signal transduction pathways acting through and independently of PR. That cAMP signaling cascades can impact expression of ADAMTS-1 is supported by the observations that PTH induces ADAMTS-1 in the osteoblasts (41).
Important regions for basal promoter activity of the ADAMTS-1 promoter involve C/EBPß, NF1-like factor, three GC-rich regions (GC-I, -II, and -III) and a TATA box because deletions or mutations of these sites resulted in significant decreases in basal activity of the promoter. Like ADAMTS-1, C/EBPß is induced by the LH surge (17), indicating that coordinate regulation of these genes occurs during ovulation. Egr-1, which binds GC-rich sequences such as GC-II within the ADAMTS-1 promoter, is also induced by the LH surge (18, 19), whereas the NF1-like factor (data not shown) and Sp1/Sp3 are constitutively expressed in granulosa cells during this period (20, 35). The important role of Sp1/Sp3 in regulating ADAMTS-1 places it among a growing number of genes (42, 43, 44) including those induced in granulosa cells by the ovulatory LH signal via GC-rich promoter elements. These include PR (20), serum and glucocorticoid-responsive kinase (35), Egr-1 (18, 19), and cathepsin L (21). Because single, double, and triple GC site mutations were not sufficient to alter promoter activation by Fo +PMA, multiple sites and factors appear to be involved. This combinatorial type of regulation has been reported for other promoters namely, Egr-1 (18), cathepsin L (21), and PR (20). Additionally, the ADAMTS-1 TATA box was found to be critical for Fo + PMA induction of the promoter. Other genes in which the TATA box has been reported to be crucial for cAMP regulation include CYPIIAI and kit ligand (45, 46). Significantly, like ADAMTS-1 the kit ligand promoter does not contain a functional CRE site but does contain critical GC-rich regions involved in cAMP regulation of the gene (46). Interestingly, the GC boxes do not replace the need for a TATA box on the ADAMTS-1 promoter, which has been reported for other promoters that contain GC-rich regions (20, 21). Our studies indicate that, for proper regulation and expression of the ADAMTS-1 gene, the TATA box as well as the GC-rich regions must be intact.
Induction of the ADAMTS-1 gene in granulosa cells of preovulatory follicles by LH also depends on its induction of PR. This is indicated by the marked secondary increase in ADAMTS-1 message that occurs in vivo near the time of ovulation (P, hCG 8–12 h) and is only observed in mice where expression of PR is intact. However, the amount or functional state of PR that is expressed in cultured granulosa cells is more difficult to evaluate because treatment of cells with Fo + PMA was necessary to induce PR and also ADAMTS-1. Importantly, treatment of cultured rat granulosa cells with the PR agonist R5020 did not significantly affect ADAMTS-1 mRNA expression, most likely because the cultured cells secrete critical amounts of progesterone. Thus, it is likely that the endogenous PR that is induced by Fo + PMA is saturated with ligand, and therefore the PR-regulated expression of ADAMTS-1 is already maximal before the addition of R5020. This hypothesis is supported by the observation that the PR antagonist ZK98299 reduced expression of ADAMTS-1 message in these cultured cells. Therefore, to study the effects of PR on ADAMTS-1 expression, we cotransfected PRA and PRB expression vectors, which resulted in the activation not only of a GRE reporter but also the ADAMTS-1 reporter constructs. Other investigators have also found that overexpression of certain NFs is necessary to study their function in vitro (43, 47). Furthermore, coexpression of PRA (or PRB) with ADAMTS-1 promoter-Luc reporter constructs significantly stimulated activity in the absence of Fo + PMA. Thus, in cultured granulosa cells PR exerted its effects independently of LH-like signaling cascades. The effect of PR on the ADAMTS-1 promoter is specific because coexpression of other nuclear receptors (ER and androgen receptor) do not alter promoter activity. In addition, the LH-regulated COX-2 promoter is not induced in response to exogenous PRA in granulosa cells, indicating that the regulation of ADAMTS-1 by PR is specific.
The mechanisms by which PR mediates ADAMTS-1 activation appears to involve sites other than a consensus PRE because no such sites were identified within the –375/+266 bp promoter fragment that was activated when coexpressed with PRA or PRB. These results indicate that PR regulates the promoter through an indirect mechanism such as protein-protein interactions with other DNA binding transcription factor(s), induction/modification of an intermediate transcription factor, or another indirect mechanism. In support of the first possibility, PR and Sp1 have been shown to interact in the regulation of other target genes, including PR (28), and cyclin-dependent kinase inhibitor (p21WAF1) (29). Because GC-rich Sp1/Sp3 binding sites are functionally important in the ADAMTS-1 promoter and because mutations of the GC-I, -II, and -III sites reduced PR activation, similar mechanisms are potentially operative in this promoter. Additionally, we demonstrate that PR regulates the promoter through the C/EBPß and NF1-like factor sites. Evidence for NF1 and PR transcriptional interaction was reported by Di Croce et al. (48) in which these factors together regulate the MMTV promoter. Therefore, the NF1-like factor, which may be one of the many splice variants of NF1 or one of the four family members, may also interact with PR. Likewise, PR and C/EBPß interact to regulate the decidual prolactin promoter (49). Herein, we show that PR regulates the ADAMTS-1 gene through the binding sites for C/EBPß, NF-1-like factor, and Sp1/Sp3 (GC-I, -II, and -III) binding sites. Because only a few promoters that are PR responsive have been shown to have consensus PREs [i.e. MMTV and IGF binding protein-1 (27)], these consensus sites may represent the exception rather than the rule. Thus PR may, in many situations including its transactivation of ADAMTS-1, act more as a coregulatory molecule.
In summary, this report is the first to characterize the complex hormonal and molecular mechanisms whereby LH and PR regulate ADAMTS-1 expression in rodent ovarian granulosa cells. The effects of LH- and PR-mediated regulation of ovarian ADAMTS-1 are separable but act cooperatively to induce maximal expression of the ADAMTS-1 gene in vivo and in culture. Transcription of the ADAMTS-1 gene by PR appears to be mediated indirectly through its interaction with the transcriptional regulators C/EBPß, NF1-like factor, and Sp1/3. As such, PR appears to play the role of an inducible coregulator of the ADAMTS-1 gene in granulosa cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and GABPß were gifts from Dr. Thomas Brown (Pfizer, Groton, CT) (50).
Animals
PRKO mice were obtained from Dr. John Lydon (Baylor College of Medicine). WT C57BL/6 mice, hypophysectomized (H) rats and immature intact Holtzman Sprague Dawley female rats were purchased from Harlan Sprague Dawley, Inc. (Indianapolis, IN). Animals were housed under a 12-h light, 12-h dark schedule, provided food and water ad libitum and were treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Protocols were approved by the Institutional Animal Care and Use Committee, Baylor College of Medicine.
Immature (21–23 d old) WT and PRKO female mice were injected ip with 4 IU of PMSG to stimulate follicular growth and 44 h later with 5 IU of hCG, an LH analog, to trigger ovulation and luteinization. Mice were euthanized at the indicated times, ovaries extirpated and total RNA isolated. For studies using the PR antagonist, immature mice were injected twice daily with 100 µl of (384 µg) ZK98299 dissolved in 100% ethanol, beginning with an injection of 4 IU of PMSG. hCG (5 IU) was injected to each animal 44 h after PMSG treatment along with ZK98299. Mice were euthanized 10 h post hCG treatment and ovaries extirpated and either total ovarian RNA was isolated or ovaries were fixed in 4% paraformaldehyde for histological analysis.
Ovarian follicular growth and differentiation were stimulated in H rats by hormonal treatments as described previously (51, 52, 53). H rats received daily sc injections of 17ß-estradiol (E; 1.5 mg/0.2 ml propylene glycol) for a consecutive 3 d (HE). FSH (F; 1 µg/0.1 ml PBS) was then injected (sc) twice daily for 2 d (HEF). Luteinization was induced in HEF rats by tail vein injection of 10 IU hCG (HEF/hCG). In addition, intact immature Holtzman Sprague Dawley rats (23–25 d old) were primed with 10 IU of PMSG and 44 h later stimulated with 10 IU of hCG to induce ovulatory genes. At selected intervals, rats were euthanized, ovaries extirpated, and granulosa cells were isolated by needle puncture (54, 55) for preparation of whole cell extracts (WCE). Each treatment group included four rats except for the H group in which 12 rats were used.
For primary cultures of rat granulosa cells, intact immature Holtzman Sprague Dawley rats (23–25 d old) were primed with a single sc injection of PMSG (10 IU). Granulosa cells were either harvested 40–44 h later by needle puncture and cultured as described below.
Cell Culture
Rat Granulosa Cells
Granulosa cells from PMSG-primed rats were cultured overnight at a density of 5.0 x 105 cells per 1 ml of DMEM:F12 medium containing penicillin-streptomycin and 5% fetal bovine serum in multiwell dishes 22.6-mm. The next morning, cells were washed in serum-free medium and then cultured in the same media alone or with various agonists and/or antagonists. FSH (50 ng/ml) and T (10 ng/ml) were added to induce granulosa cell differentiation. Fo (10 µM) and PMA (20 nM) were used to mimic LH-induced second messenger pathways. To stimulate PR activity, the progestin R5020 (10 nM) was added to cell cultures. To inhibit PR activity, the PR antagonist ZK98299 (100 µM) was added to cell cultures overnight.
Progesterone levels of the original cell culture media, conditioned media from rat granulosa cells cultured overnight (16 h), and conditioned media from cultures in which the media was changed after overnight culture and treated with R5020 for 4 h were measured by the University of Virginia Core Ligand and Assay Laboratory.
Mouse Granulosa Cells
Granulosa cells were harvested from immature (22–24 d old) PRKO and WT mice, cultured overnight in serum-coated multiwell dishes (16 mm) and treated as described for rat granulosa cells above.
RNA Isolation and RT-PCR
Total RNA was prepared from intact mouse ovaries and cultured granulosa cells from rats and mice by extraction in TRIzol reagent and purified as specified by the manufacturer. Semiquantitative RT-PCR was performed as described previously (56), using specific primer pairs for ADAMTS-1 (5), mouse PR (5), rat PR (1), and the internal control ribosomal protein L19 (RP-L19) (5). Total RNA (500 ng) was reverse transcribed by using oligo poly-(deoxythymidine) and AMV reverse transcriptase at 42 C for 75 min, followed by 95 C for 5 min. DNA products were amplified by adding [32P]dCTP, Taq polymerase, and Thermocycle buffer to the reaction mixtures for 20 (rat and mouse ADAMTS-1) and 26 (mouse and rat PR) cycles, respectively, at 94 C for 1 min, 60 C for 2 min, and 72 C for 2 min. The cycle number was chosen by determining the linear range of amplification for each gene. The amplified cDNA products were resolved on a 5% polyacrylamide gel, which was dried and exposed to x-ray film. The radioactive PCR products were quantified by using a Storm 860 PhosphorImager software (Molecular Dynamics, Sunnyvale, CA).
Cloning of the Mouse ADAMTS-1 Promoter, Construction of Promoter-Reporter Plasmids, and Expression Plasmids
A mouse 129SvJ genomic DNA library (Stratagene, La Jolla, CA) was screened using a 500-bp probe corresponding to the previously reported 5'-untranslated region plus 100 bp of the ADAMTS-1 coding sequence (34). Several phage clones were purified to homogeneity. Southern blot analyses of Not-1 digested phage DNA using both the 5'-probe used to screen the library and a probe directed to the 3'-UTR of the ADAMTS-1 gene identified one clone, which contained the entire coding sequence as well as 5' and 3' ADAMTS-1 genomic sequence. Direct primer extension sequencing from the T7 polymerase initiation site present in one phage arm demonstrated that the clone ends within the known 3'-UTR sequence (8494 bp). Because the complete mouse gene encoding ADAMTS-1 has been shown to be 9.2 kb in length (34), we conclude that the 5' end of clone contains up to 5 kb of 5'-regulatory sequence. Computer-based homology analyses using online software at the Protein Information Resource of the National Biomedical Research Foundation (http://pir.Georgetown.edu/pirwww) indicated that the mouse promoter was highly similar to rat and to human. Likewise, ADAMTS-1 protein sequence is highly homologous among rat, mouse, human, and the horse (40).
A –1186/+266 bp fragment of the ADAMTS-1 gene (Genbank accession no. AY423552) was released from the phage vector by restriction enzyme (KpnI/NheI) digestion and subcloned into the multiple cloning cassette of the pGL3basic vector (Promega, Madison, WI). To obtain 5' promoter sequence, direct sequencing of the phage clone was performed using a primer corresponding to the end of the previously reported 5'-UTR. Truncations of the ADAMTS-1 promoter were generated by PCR cloning strategies, whereas site-specific mutations of the promoter were created either by using the GeneEditor In Vitro Site-Directed Mutagenesis System (Promega) or by PCR cloning strategies. All constructs were sequenced to verify their authenticity. The sense strand sequences of the oligonucleotide primers used to create mutants are listed below from 5' to 3' (WT sequences can be found under the description of EMSAs). Core binding elements are underlined, and mutations are in lowercase:
C/EBPß mutant–375AGGAATGTTGAGGttTCTTTTCAATT–350 bp
Mutant GC-III–139GCTGCCCCCTCCaaCTTCAGGCCCCGAGG–111 bp
Mutant GC-II–89GACTGAGCTCAGGGaaCGGTGTCG –66 bp
Mutant GC-I–56 GGGAAGGAaaGGCTCCTATGTGG–33 bp
PSCT, PSCT-PRA, PSCT-PRB and GRE2-TATA-Luc plasmids were gifts from Dr. Rainer Lanz and Dr. Neil McKenna (Baylor College of Medicine). The ER expression vector and the ERE-Elb-Luc construct were gifts from Dr. Carolyn Smith (Baylor College of Medicine). The COX-2 promoter-reporter construct (pTIS-371LUC) was provided by Dr. Harvey R. Herschman (UCLA School of Medicine, Los Angeles, CA).
Whole Cell Extracts
Granulosa cells isolated directly from ovaries of hormonally treated (PMSG + hCG) rats by needle puncture or after culture in DMEM/F-12 were collected by microcentrifugation (3000 x g). Cells were resuspended in WCE buffer (10 mM Tris-buffer containing 1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, 400 mM potassium chloride, 1 mM vanadate, 1 mM PMSF, 5 µg/ml PicI and PicII) (57) and lysed by three rapid freeze/thaw cycles. The solubilized proteins were collected in the supernatant by microcentrifugation (14,000 x g). Protein concentrations of soluble extracts were measured by the mini-Bradford assay (Bio-Rad, Richmond, CA).
EMSA
Putative transcription factor binding sites were determined by analyzing –1186 bp of the ADAMTS-1 promoter using web-based transcription factor prediction programs: TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html) and WWW Signal Scan IMD Search Service (http://bimas.dcrt.nih.gov/molbio/matrixs/). Double-stranded oligonucleotides containing the putative transcription factor binding sequences of the ADAMTS-1 promoter were end labeled by incubation for 1 h with [32P]dCTP and Klenow and used in EMSAs as previously described (58). Briefly, 50,000 cpm of 32P-labeled oligonucleotides and poly(deoxyinosinic-deoxycytidylic) acid (1.4 µg) were incubated for 30 min at RT with whole cell extract protein in a final buffer volume of 20 µl containing 15 M Tris-HCl (pH 7.5), 100 mM KCl, 5 mM dithiothreitol, 1 mM EDTA, 5 mM MgCl2, and 12% glycerol. For supershift and cold competition experiments, antibodies or unlabeled oligonucleotides, respectively, were incubated with protein extracts for 30 min on ice before addition of the labeled probe. Protein/DNA complexes were separated by 5% acrylamide gel electrophoresis. Gels were dried and exposed to x-ray film. The sense strand sequences of the oligonucleotide probes used are listed below from 5' to 3'. Core binding elements are underlined:
C/EBPß–375AGGAATGTTGAGGAATCTTTTCAATT–350 bp
Dual c-ETS–316GGGCACAGGAAGGGCGACAGGAAGCAGGGTG–386 bp
NF-1–164GCTGTGGCGTGAGGCCAGGG–145 bp
NF1 consensus GTTTTGGATTGAAGCCAATATGATAA
GC-III–139GCTGCCCCCTCCCCCTT–122 bp
GC-II–89GACTGAGCTCAGGGGGCGGTGTCG–66 bp
GC-I–56 GGGAAGGAGGGGCTCCTATGTGG–33 bp
Transient Transfection and Luciferase Reporter Assay
Granulosa cells from PMSG-primed rats were cultured as described above. Transient transfections were carried out 3 h after plating cells. Cells were transiently transfected with 0.5 µg of the indicated promoter-reporter constructs. Cotransfections with either empty or PRA expression vector used between 5 and 50 ng of the plasmid, as indicated. Transfections were carried out overnight in DMEM:F12 containing 5% fetal bovine serum using Fugene transfection reagent (Roche Molecular Biochemicals, Indianapolis, IN) according to the manufacturer’s instructions. The next morning, cells were washed with serum-free medium then placed in the same media containing hormone, agonist, and/or antagonist as indicated. After 6 h of treatment, cells were harvested in lysis buffer [0.2 M Tris (pH 8.0) containing 0.1% Triton X-100]. Cytosolyic protein concentrations were determined by the mini-Bradford assay. Luciferase activity was analyzed according to a standard protocol. In brief, a 40-µl aliquot of the cell lysate was mixed with 100 µl of the Firefly luciferase substrate [20 mM Tris (pH 8.0) containing 4 mM MgSO4, 0.1 mM EDTA, 30 mM dithiothreitol, 0.5 mM ATP, 0.5 mM luciferin, and 0.25 mM coenzyme A] and each reaction was monitored by a Dynex Technologies, Inc. (Chantilly, VA) MLX Luminometer. Data were normalized to the cellular protein concentration in each sample: relative light units/protein concentration (mean ± SEM). Each experiment was performed in triplicate at least three times.
Statistical Analysis
The values represented are mean ± SEM. Data were analyzed by ANOVA and Newman-Keuls multiple comparison test to determine significance. Values were considered significantly different if P < 0.05.
|
ACKNOWLEDGMENTS |
|---|
expression vector and the ERE-Elb-Luc construct. |
FOOTNOTES |
|---|
This work was supported by National Institute of Child Health and Human Development (Specialized Cooperative Centers Program in Reproduction Research) Grant U54-HD07495 (Baylor College of Medicine) and U54-HD28934 University of Virginia Center for Research in Production Ligand Assay Core.
Abbreviations: ADAMTS-1, A disintegrin and metalloproteinase with thrombospondin-like motifs; C/EBPß, CAAT enhancer binding protein ß; COX-2, cyclooxygenase 2; Egr-1, early growth response protein 1; ER, estrogen receptor; Fo, forskolin; GABP, c-Ets factor GA binding protein; GC, guanine cytosine; GRE2-TATA-Luc, PR-responsive reporter; H, hypophysectomized; hCG, human chorionic gonadotropin; HE, hypophysectomized, estradiol; HEF, hypophysectomized, estradiol, FSH; MMTV, mouse mammary tumor virus; NF, nuclear factor; P, PMSG; PMA, phorbol myristate acetate; PMSG, pregnant mare serum gonadotropin; PR, progesterone receptor; PRA and PRB, two protein isoforms of PR; PRE, PR binding element; PRHET, PR heterozygote; PRKO, mice null for PR; Sp1/3, specificity protein 1 or 3; T, testosterone; RP, ribosomal protein; UTR, untranslated region; WCE, whole cell extract; WT, wild-type.
Received for publication September 30, 2003. Accepted for publication July 6, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
