| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Biological Chemistry, Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel 91904
Address all correspondence and requests for reprints to: Dr. Joseph Orly, Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, Israel. E-mail: orly{at}vms.huji.ac.il.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
isoform of the human activating protein-2 (AP-2). Specific antibodies tested in EMSA confirmed that AP-2 is the predominant isoform that binds SCC1 in the human placenta, whereas AP-2
is the only mouse placental protein that binds this oligonucleotide. Functional studies showed that coexpression of the rat P450scc promoter (-378/+8 CAT) and AP-2 isoforms ( or ) in human embryonic kidney 293 cells results in a marked activation of chloramphenicol acetyltransferase (CAT) transcription that is dependent on an intact AP-2 motif, GCCTTGAGC. This motif conforms with consensus sequences previously determined for binding of the AP-2 and isoforms. Mutations of the AP-2 element ablated binding of AP-2 to SCC1, as well as severely diminished the promoter activity in primary mouse giant trophoblasts and human choriocarcinoma JAR cells. Collectively, these studies suggest that expression of placental P450scc is governed by AP-2 factors that bind to a cis-element that largely overlaps the sequence required for recognition of SF-1 in other steroidogenic tissues. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In humans, the corpus luteum is active during the first trimester, and the placenta assumes the steroidogenic role of the ovary at 8 wk of pregnancy. The principal steroid-producing cells in the human placenta are the syncytiotrophoblasts that, at term, can secrete up to 300 mg of progesterone per d. The use of antiprogestin mifepristone (RU486) for therapeutic abortion during the second trimester (13), together with the dramatic decrease of placental steroidogenesis observed in women who experienced pregnancy failure during wk 8–12 of gestation (14), suggest that placental progesterone production is essential for maintenance of pregnancy. In this regard, several mechanisms by which placental progesterone affects various myometrial and endometrial activities to prevent premature parturition have been proposed (15, 16, 17). The latter roles of progesterone in sustaining pregnancy are also relevant in rodents. However, the role of placental progesterone production in rodents is less obvious because luteal progesterone secretion is indispensable throughout term (18). Nevertheless, by the time of placentation at midpregnancy, a marked rise in the expression of the steroidogenic genes, including P450scc (19, 20, 21), steroidogenic acute regulatory factor (StAR), 3ß-HSD (22), and P45017 (20, 21), is observed in the giant trophoblast cells. This transient expression of genes required for placental progesterone production has been associated with various speculations related to protection of the conceptus from maternal immunorejection (19, 23), or effects of locally produced steroids on sex determination of the fetus (24). It is noteworthy that after implantation, expression of P450scc and 3ß-HSD has been demonstrated also in decidual cells of the uterine wall (21, 22, 25), suggesting that a local uterine production of steroid hormones is also important during early phases of pregnancy.
It is well accepted that expression of P450scc and other steroidogenic genes in the gonads and the adrenal is provoked by trophic hormones via the intermediary role of cAMP (26, 27, 28). It is less clear what hormones or other signaling cues up-regulate the steroidogenic genes in the placenta. At the transcriptional level, steroidogenic factor 1 (SF-1) (29), also known as Ad4BP (30), is the pivotal tissue-specific factor that mediates hormone-induced expression of P450scc and other steroid hydroxylases in the adrenal and gonads (11, 30, 31, 32, 33, 34). Furthermore, targeted ablation of the mouse SF-1 gene also impairs the embryonic development of those organs (35, 36, 37). Nevertheless, P450scc expression and normal development of SF-1-deficient placenta allow live birth of mutant pups (35, 37). These observations agree with the apparent absence of SF-1 in human trophoblasts (6), rat choriocarcinoma Rcho-1 cells, and midpregnancy rat placenta (38). Collectively, these findings suggest that the expression of placental P450scc does not require SF-1. The present study aimed to find what trans-acting protein(s) could possibly control P450scc transcription in the absence of SF-1 in the mouse and human placenta.
Among several cis-acting DNA sequences that have been implicated in cAMP-mediated transcription of P450scc (32, 39, 40, 41), two proximal regions, named SCC1 and SCC2 (Ref. 33 and Table 1
) harbor SF-1 binding motifs required for functional activity of the promoter in rat ovarian and adrenal cells (31, 33). In particular, the sequence of SCC1 is highly conserved (Table 2) among the rat, mouse, porcine, ovine, bovine, and human genes (31, 39, 42, 43, 44, 45, 46). Therefore, we postulated that the placenta should express a protein that can "substitute" for SF-1 action by associating with a cis-element in the SCC1 region and mimic SF-1 activation of the P450scc promoter in trophoblast cells. Our studies revealed that these criteria are met by a placental nuclear factor, initially called PNF, which was later purified and identified as the activating protein-2 (AP-2).
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Figure 1 shows that protein extracts prepared from either mouse embryonic d 9.5 (E9.5) giant trophoblast cells (lane 7), or E14.5 whole placenta (lane 10), formed a slow migrating complex with SCC1 that is clearly distinct from the typical SF-1 binding patterns observed with cell extracts from either testicular Leydig MA-10 mouse cell line, mouse adrenal, or mouse ovary (lanes 1, 2, and 3, respectively). The SCC1-bound trophoblast protein(s) was not affected by SF-1 antiserum (lane 8), whereas this antiserum (47) readily ablated the ovarian SF-1 band shift (lane 4). These results suggested that the placental protein(s) that formed a complex with SCC1 did not include SF-1. An identical DNA-protein band shift was also observed with placenta nuclear extract (lane 12). Therefore, for reasons of convenience we named the putative placental protein that binds SCC1 "placental nuclear factor" (PNF). Control nonspecific antiserum to selective promoter 1 (Sp1) did not affect PNF or SF-1 binding to SCC1 (Fig. 1, lane 9 or 5, respectively). This result indicated that Sp1 is not involved in the PNF-SCC1 complex. The amount of PNF expressed in E7.5 preplacentation sites was very low (lane 6). Screening of various adult mouse tissue extracts for the presence of PNF-like binding activity did not reveal the typical PNF band shift in the lung, intestine, heart, kidney, muscle, brain, and liver (lanes 14–20). These results suggested that, in comparison to the placenta, the maternal tissues do not contain detectable levels of PNF activity.
|
) were tested by EMSA using giant cell extract. PNF did not complex with any mutants of the SF-1 binding element (Fig. 2B, lanes 3–5). Moreover, PNF binding required the presence of an additional 5'-nucleotide triplet (AGC) upstream of the SF-1 site (M5, lane 6). These results were corroborated by competition assays using wild-type (WT) 32P-SCC1 as a probe and a 50-fold molar excess of the unlabeled DNA mutants (lanes 8–14). Additional EMSA studies using ovarian extracts (Fig. 2C) confirmed that the DNA motif required for SF-1 binding is 5'-TCAAGGCTA, whereas PNF binding required an extended element, 5'-AGCTCAAGGCTA (Fig. 2D). At first, this sequence seemed to be a near-consensus variation of a direct repeat with 0 bp spacing element (DR0), AGGTCAAGGTCA. Therefore, we tested whether direct repeat elements recognizing various known nuclear receptors, like chicken ovalbumin upstream promoter transcription factor (COUP-TF) (48), germ cell nuclear factor (GCNF) (49), and retinoid X receptor/retinoic acid receptor (RXR/RAR) (50) can also bind PNF. Figure 3B provides a compelling evidence that COUP-TF, GCNF, and RXR/RAR binding elements cannot compete for PNF binding to SCC1 (lanes 4–9). A similar inability to compete for PNF binding activity was also observed in the presence of molar excess of the following motif sequences (not shown) that bear relevant resemblance to the PNF binding element: nerve growth factor inducible-B response element (NGRE)-Nur77 (51), cAMP response element (52), estrogen response element (53), Sp1 (54), and CCAAT enhancer binding protein-ß (CEBP/ß) (54). Finally, Fig. 3C shows that SCC1 probe in which the putative PNF recognition site (AGCTCAAGGCTA) was replaced with a consensus DR0 motif, AGGTCAAGGTCA (SCC1-DR0), was incapable of PNF binding (lanes 3–6). This result suggested that the minor differences between the sequence of consensus DR0 and the sequence of PNF binding element are, nevertheless, critical for recognition of the placental protein.
|
|
) and nests a second site (GGGAGGTCA), previously proposed to be a putative SF-1 binding motif important for functional activity of the P450scc promoter in Y-1 adrenal cells (31). To test whether SCC2 can associate with PNF, we compared the binding characteristics of SCC2 and SCC1 using E9.5 giant trophoblast cell extract (Fig. 4). Clearly, SCC2 did not complex with PNF (lane 4), nor could it compete for SCC1 binding to PNF (lanes 3). Control experiments with extract of corpora lutea from the same 9.5 postcoitus (p.c.) females showed that SF-1 did not bind the SCC2 probe either (lane 8). However, a large molar excess of SCC2 moderately displaced SCC1 binding to SF-1 (lane 7). The latter result conformed with earlier studies showing a weak SCC2 binding to rat ovarian SF-1 (33). Collectively, these results suggest that SCC2 does not contain sequences of interest related to binding of PNF or other potential factors in the placenta.
|
) includes a minor fraction of unavoidable nongiant placenta cells that underlay the mesometrial giant cell layers. Therefore, to examine in which of the placental cell types the P450scc promoter can be activated, we constructed chimeric genes based on the -378/+8 region of the rat P450scc gene, placed upstream of a green fluorescent protein (GFP) coding sequence (Materials and Methods). Additional GFP reporter plasmids were constructed to include site-specific mutations in the SCC1 core element, aiming to impair the potential binding of PNF (M3 and M5, Figs. 2A and 5B). After transfection, cells were incubated with or without 8-bromo-cAMP (8-Br-cAMP) in serum-free medium and inspected daily by confocal microscopy. Figure 5C shows visual promoter activities in live cells depicted by a pseudocolor-coded presentation of GFP fluorescence levels. The highest expression of GFP was observed in cells containing the WT promoter and treated with cAMP (panel D'). Site-specific mutations M3 and M5 markedly ablated the GFP expression in cells treated with cAMP (panels F' and H'), or incubated without the cyclic nucleotide (panels E' and G'). The low expression of M3 and M5 mutants in the absence of cAMP suggest that PNF is responsible for both the basal and the inducible promoter activity in the placental cells. In all cases, a predominant expression of GFP was observed in the giant cells, whereas GFP staining of other placental cell types was not detectable (panels A'–D').
|
, upper right panel) were ligated into a pCAT-Basic plasmid and transiently expressed in giant trophoblast cells. Because 8-Br-cAMP-treated cells expressed much higher levels of the GFP reporter gene, we examined the CAT constructs in cAMP-induced giant cells. Figure 6A shows that mutations unable to bind PNF (M3 and M5), lost 70–75% of their ability to activate CAT expression in cAMP-treated cells. A similar loss of CAT expression by the M3 and M5 promoter mutants was also observed in control cultures without cAMP, although the overall CAT activities were considerably lower (not shown).
|
). The M3 construct, in which the SF-1 binding motif is mutated, showed a 75% reduction of the promoter activity. By contrast, the promoter plasmid that harbors the M5 mutation, which does not affect the SF-1 binding motif, generated a normal 10-fold response to FSH.
Human PNF
In view of the fact that the SCC1 region is highly conserved in the P450scc gene of many species, including man (Ref. 46 and Table 2), we questioned whether human placenta contains a PNF-like binding activity. The band shift patterns shown in Fig. 7 indicated that, indeed, human term placenta contains PNF activity and, as expected, a 100-fold molar excess of either SCC1-M3 mutant (lane 3), or a GCNF probe (lane 4), did not compete for SCC1 binding to the human protein. Importantly, excess of SCC1-M3 and GCNF (DR0) oligonucleotides efficiently displaced a nonspecific complex that was observed in the mouse tissues (Figs. 1–4) and became predominant in extracts of the human term placenta (n.s., Fig. 7). Therefore, we used SCC1-DRO to displace binding of this nonspecific protein to the SCC1 affinity column during purification of PNF from human term placenta (Materials and Methods).
|
). Radiolabeled SCC1 formed a typical PNF band shift, which was absent when either SCC1-M3 or SCC1-M5 mutants were used as probes. In accordance, transient transfection assays in JAR cells revealed that SCC1-M3 and SCC1-M5 mutants of the -378/+8 CAT gene were significantly less active (50%) when compared with the basal activity of the WT promoter (Fig. 8B). Altogether, these results suggested that PNF may be relevant for regulation of the P450scc gene in the human placenta.
|
shows that protein-DNA complexes of similar size were cross-linked in the presence of either mouse or human placenta extracts. The size of the predominant complex was 90 kDa, which suggested that a protein of approximately 50–60 kDa was covalently cross-linked to the DNA probe. To test whether the protein cross-linked to SCC1 is the PNF observed in EMSA, we used a 25-fold excess of unlabeled WT SCC1, or SCC1-M3 mutant, added before cross-linking. The presence of WT SCC1, but not SCC1-M3, markedly reduced the labeling of the predominant band of 90 kDa, suggesting that PNF is indeed the protein covalently bound to SCC1. It should be noted that SCC1 labeled additional complexes of higher molecular mass, implying that PNF-SCC1 complex includes more than one polypeptide component.
|
(AP-2). The molecular mass of human AP-2, 48 kDa (56), agrees well with the expected size of human PNF as determined by UV cross-linking experiments (Fig. 9). To test whether the SCC1-PNF complex observed in EMSA contains AP-2, we incubated extracts from human term placenta with monoclonal antiserum to AP-2. Figure 10 shows that this antiserum caused a supershift of the entire SCC1-PNF complex (lanes 1 and 2). By contrast, antiserum to AP-2 barely affected the PNF-DNA band (lane 3), suggesting that the predominant AP-2 isoform in human placenta is AP-2. Furthermore, identical binding characteristics (lanes 4–6) were observed when extract of human placenta was incubated with a 32P-SCC1 sequence of the human P450scc gene (-49/-32). Control nonrelevant antiserum to c-Jun did not affect the protein-DNA complex at all (lane 7). Interestingly, incubation of AP-2 antisera with extract of JAR cells showed that AP-2 is the predominant isoform in these immortalized cells (lanes 8–10). Finally, EMSA studies testing extracts of the mouse E9.5 giant trophoblast cells revealed that murine PNF is solely AP-2 (lanes 11–13).
|
and AP-2, G/CCCNNA/C/GG/AGG/C/T (57). Therefore, we tested a SCC1 oligonucleotide in which we mutated three bases of the formerly considered PNF binding site, so that the putative AP-2 motif remained intact (SCC1-AP-2) (Fig. 11). Indeed, Radiolabeled SCC1-AP-2 oligonucleotide readily associated with murine AP-2 (Fig. 11, lane 4), and a 20-fold molar excess of nonlabeled SCC1-AP-2 DNA competed for binding of the AP-2 to WT 32P-SCC1 (Fig. 11, lanes 1 and 3). These experiments confirmed that AP-2 binds to the GCCTTGAGC motif. However, the competition experiments suggest that WT SCC1 oligonucleotide binds AP-2 with somewhat better affinity than SCC1-AP-2 DNA (Fig. 11, compare lane 2 with lane 3).
|
or AP-2. Both AP-2 isoforms activated the WT promoter 25- to 40-fold over its basal activity (Fig. 12). The promoter activity was abolished when the AP-2 binding element was modified by the M3 or M5 mutations. These functional studies strongly supported the notion that AP-2 proteins can activate transcription of P450scc.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Despite the fact that the PNF-binding element seemed closely related to a DR0 motif (AGGTCAAGGTCA), oligonucleotide containing various direct repeat motifs were unable to compete for PNF binding. This result implied that PNF might not be a member of the nuclear receptor family after all. To identify the placental protein, we purified PNF from human term placenta by use of affinity chromatography approaches. Sequencing of the purified protein revealed that human PNF is the -isoform of the human activating protein-2 (AP-2). Specific antibodies tested in EMSA confirmed that AP-2 is the predominant isoform in human term placenta. This protein readily binds to the SCC1 region in both the human and rat P450scc genes. Similar EMSA also showed that AP-2 is the only isoform that is expressed in the trophoblast cells of the mouse placenta. This observation agrees with earlier studies demonstrating that of the three AP-2 genes known to date, AP-2 is the most abundant one in the mouse placenta (58).
Functional studies showed that coexpression of the AP-2 isoforms with the rat P450scc promoter (-378/+8 CAT) in a heterologous HEK293 cell system results in a marked activation of transcription. Such activity requires an intact AP-2 recognition site present in the promoter. As predicted, this AP-2 site is nested (underlined) within the previously perceived PNF element (sense 5'-TAGCCTTGAGCT) and fulfills the requirements for specific binding of the AP-2 isoforms and to a consensus sequence, G/CCCNNA/C/G G/AGG/C/T (57).
The identification of AP-2 proteins as regulators of P450scc expression in the placenta is consistent with PNF characteristics and previously reported notions about the role of AP-2 in the placenta. First, the AP-2 gene family encodes for 48- to 52-kDa proteins that bind to DNA in dimers (59, 60, 61, 62, 63). This can explain the slow migration of the PNF-SCC1 complex in EMSA, when compared with SF-1 that binds SCC1 as a monomer. Second, AP-2 isoforms are expressed in the placenta where they regulate transcription of genes essential for placental development (58) and endocrine functions. Examples are the genes encoding both human CG (hCG) subunits (64, 65, 66), the hCG receptor (67), and placental lactogens in humans and sheep (68, 69). Most relevant to our studies is the fact that hCG expression levels increase concomitantly with the dual onset of P450scc and AP-2 expression during the differentiation of cytotrophoblasts to syncytiotrophoblasts in first trimester human placenta (64, 70).
Strongly supportive of our findings, a recent study by Peng and Payne (71) showed that AP-2 is one of three factors involved in the transcription of placenta-specific isoform VI of the murine 3ß-HSD gene. Thus, the involvement of AP-2 in the regulation of 3ß-HSD VI and P450scc can explain the coordinated spatio- temporal expression of both genes observed during the second half of mouse pregnancy (22). Lastly, the third gene product involved in progesterone synthesis is StAR. Because the human placenta does not express StAR (72), it was proposed that the placental MLN64 protein replaces the role of StAR function in this organ (73). Current studies in our laboratory show that AP-2 is also implicated in the regulation of the human MLN64 promoter, which is strongly activated in primary giant trophoblast cells (Sher, N., and J. Orly, unpublished results). Collectively, these observations suggest that all the genes required for synthesis of steroid hormones in the human placenta are commonly regulated by AP-2.
The fact that the AP-2 element-dependent activation of the P450scc promoter in the giant trophoblasts was cAMP inducible is in agreement with the well established data that cAMP and protein kinase A signaling are involved in AP-2-dependent promotion of transcription (66, 74, 75). Especially relevant to our findings are studies showing that cAMP signaling activates AP-2-dependent transcription of hCGß and hCG in JEG-3 cells and primary villous trophoblasts (65, 76). Therefore, it is conceivable that P450scc expression in murine giant cells is also mediated by a cAMP/AP-2 signaling pathway triggered by hormonal cues yet to be discovered.
The pursuit of tissue-specific transcription factors regulating P450scc expression in the placenta started earlier on, when several investigators focused on the -155/-131 region of human P450scc, a region that was sufficient to confer transcription activation when positioned in a minimal heterologous promoter tested in JEG-3 cells (32, 41). Later on, this DNA element was shown to bind several factors, not necessarily of placental origin (77, 78). Other investigators used rat choriocarcinoma Rcho-1 cells as a model to study the regulation of rat P450scc (38). In contrast to our findings with similar promoter constructs, stable or transient transfections of Rcho-1 cells with a series of promoter segments showed that the proximal -379/+37 region was hardly active at all. Instead, more distal sequences of this promoter (-639/-428) were implicated in the developmental activation of the P450scc gene during differentiation in culture. No further identification of particular cis-acting elements required for activation of the P450scc promoter in the Rcho-1 cells was ever reported. Collectively, the above studies imply that the regulatory mechanism of P450scc expression in the placenta requires cis- elements that are distinct from those that control the expression of the gene in the adrenal cortex and gonads. Our findings do not support this conclusion. We show herein that AP-2, described as a helix-span-helix transcription factor (79), and SF-1, which belongs to the zinc-finger nuclear receptor family, both bind to largely overlapping DNA sequences (Fig. 11). Therefore, it seems that transcription of P450scc presents a unique example of flexible adaptation of transcription machinery, by which a single modular sequence element is compatible to interact with different tissue-specific factors: AP-2 in the ephemeral placenta and SF-1 expressed in the fetus and adult steroidogenic tissues.
Interestingly, Pena et al. (80) were the first to suggest that AP-2 is involved in the regulation of ovine P450scc gene expression. These studies, using human JEG-3 cells, showed that AP-2 does not bind directly to the promoter but rather enables transactivation of the basal transcription machinery by protein-protein interactions with Sp1. In this regard, the present study shows that anti-Sp1 serum, which readily ablates Sp1 binding to other genes (54), did not affect the AP-2 binding to SCC1 (Fig. 1), suggesting that AP-2 binds directly to the promoter and Sp1 is not involved in this complex formation. Also, we did not observe Sp1 binding to the rat SCC2 oligonucleotide (Fig. 4), which is somewhat homologous to the OF3 region of the ovine gene, shown to be involved in transcriptional activation by Sp1 (80).
It is possible that AP-2 is not the sole regulator of P450scc transcription in the placenta. In situ hybridization and ribonuclease protection assays have shown that onset of AP-2 expression in the mouse uterus commences on d 6.5 p.c. (58), whereas migrating giant trophoblast cells do not express P450scc before d 8 (21). Also, AP-2 is expressed in all cells of the placenta (58), while we have shown that giant trophoblasts are certainly the only placental cell type expressing P450scc (21, 22). Lastly, whereas AP-2 expression is sustained through d 14.5 of pregnancy (Ref. 58 and our results), P450scc expression ceases as of d 11.5 (20, 22). These differences in the patterns of P450scc and AP-2 expression predict that, although AP-2 may be essential for control of the P450scc gene in the mouse trophoblast cells, additional positive and negative regulatory proteins are likely to be involved in the spatio-temporal control of this gene during placentation. Such regulatory devices remain to be resolved in future studies.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
IgG (sc-897x) and monoclonal antibody to AP-2 (sc-12726x), anti-c-Jun (sc44x), and anti-Sp1 (sc-059x) IgGs were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Oligonucleotides and PCR primers were synthesized by Genset Oligos (Paris, France) or Sigma-Genosys (Cambridgeshire, UK).
Animals
Female Sprague Dawley rats (21 d old) and C57xBALB/c F1 mice (6–8 wk) were obtained from Harlan Laboratories (Jerusalem, Israel) and maintained under a schedule of 16 h light, 8 h dark with food and water ad libitum. Animals were treated in accordance with the NIH Guide for the Care and Use of Laboratory Animals. All protocols had the approval of the Institutional Committee on Animal Care and Use, The Alexander Silverman Institute of Life Sciences, The Hebrew University of Jerusalem. Pregnant female mice were obtained by hormonal hyperstimulation as previously described (21). Noon time on the day after mating was considered d 0.5 p.c., or E0.5. Rat ovarian granulosa cells were prepared from estradiol-primed animals as described previously (81).
Tissue Extracts
Cell and tissue extracts for EMSA were obtained as described previously (54). Briefly, cells were homogenized using a Dounce homogenizer in 3–4 volumes of buffer A [400 mM KCl; 10 mM NaH2PO4, pH 7.4; 10% glycerol; 1 mM EDTA; 1 mM dithiothreitol (DTT); 5 mM NaF; 1 mM sodium ortho- vanadate; 5 mg/ml aprotinin; 2 mM pepstatin; 1 mM PMSF; and 1% vol/vol protease inhibitor cocktail], and the protein slurry was freeze-thawed three times in liquid nitrogen and a 37 C bath. Finally, the cell lysates were centrifuged for 3 min at 14,000 x g, and the protein content in the supernatants was determined by a modified Bradford assay (82). Extracts were kept at -70 C until use. Nuclear extracts were obtained as previously described (83).
EMSA Studies
Protein extract (10–30 µg) was incubated with 2–5 ng of double-stranded DNA previously labeled by a fill-in reaction using Klenow fragment (Promega Corp., Madison, WI) and [-32P]-dCTP (Amersham Pharmacia Biotech, Little Chalfont, UK). Binding assay was performed using a final volume of 30 µl containing 100 mM KCl, 15 mM Tris-HCl (pH 7.5), 10 mM DTT, 1 mM EDTA, 12% glycerol, and 0.75- 4.5 µg poly(dI-dC). After incubation for 35 min at room temperature, the binding products were resolved on a native prerun polyacrylamide gel (5%) using TBE running buffer (50 mM Tris; 50 mM boric acid; and 10 mM EDTA, pH 8.3). When competition experiments were conducted, the protein extract was added last to the reaction mixture. When antibodies were used for detection of a given protein-DNA complex, the reaction cocktail without the probe was preincubated with 2–8 µg of the antibody for 25 min at room temperature/4 C before addition of the labeled DNA. The dried gels were analyzed using a FLA-3000 Bio-Imaging analyzer (Fuji Photo Film Co., Ltd., Tokyo, Japan). Gels were also exposed to Super RX medical x-ray film (Fuji Photo Film Co., Ltd.) for 2–24 h at -70 C and developed by an X-Omat processor (Curix 60, Agfa, Munchen, Germany).
Oligonucleotide Probes
Sequences of the SCC1 probe and its mutants M1–M6 are shown in Fig. 2; sequence of SCC1-AP2 is shown in Fig. 11; other oligonucleotide probes (sense strand), including overhanging restriction site sequences (lowercase) were:
SCC1-DR0, 5'-gatcGCTCCTCTCTTGACC TTGACCTAGTTAccta;
human SCC1, 5'-gatcGTATGGCCTTGAGCTGGTccta (45);
SCC2, 5'-gatcGGAGGGGGGAGGTCAACACTCCATCAccta;
COUP-TF, 5'-gatcTTAGGGGTCAAAGGTCAAATGGAccta (48);
GCNF, 5'-gatcTCGAACATTTAGGTCAAGGTCAGTCCccta (49);
RXR/RAR, 5'-agctGTCACAGGTCACAGGTCACAGGTCACAGTTCAAGCTccta (50);
NGRE, 5'-agcttGAGTTTTAAAAGGTCATGCTCAATTTccta (51);
cAMP response element, 5'-gctcgagCCTTGGCTGACGTCAGAGGccta (52);
Estrogen response element, 5'-gatcACGGGT AGAGGTCACTGTGACCTCACCCGccta (53);
Sp1, 5'-gatcCGATCG GGGCGGGGCGAGCccta (54);
CEBP/ß, 5'-CTGCAGGATGAGGCAATC ATTCCATCCTccta (54).
UV Cross-Linking
For UV cross-linking studies we labeled a 50 mer probe with [-32P]-dCTP by a fill-in reaction, using the -71/-22 region of the rat P450scc antisense strand (5'-ACTTATAACCACTAACTAGCTCAAGGCTAA GAGAGGAGCTGATGGAGTGT) as template and the -71/-54 of the sense strand as a primer (5'-ACACTCCATCAGCTCCTC). Pooled EMSA reactions (120 µl each, details above) were resolved in single lanes of 5% polyacrylamide gel. After electrophoresis, the gel was placed under a UV source (254 nm, UVC 515, Ultra-LUM Inc., Carson, CA) and irradiated for 10 min. The wet gel was then exposed (15 min) to a PhosphorImager screen and the resulting pattern of the gel radioactivity assisted in the excision of slices containing the protein-DNA complexes. These gel slices were placed in dialysis bags and electroeluted under 100 V for 1 h. The eluates were concentrated by miniconcentrators (Fugisep, Intersep, Filtration systems, Wokingham, UK) and analyzed by 10% SDS-PAGE (84). Dried gels were exposed to hypersensitive film (Kodak Biomax MS, Eastman Kodak Co., Rochester NY) for 6 d at -80 C. Molecular mass of the protein-DNA complexes was determined by standard markers (Multimark, Novex, San Diego, CA).
Expression Plasmids
Plasmids AP2/pcDNA3.1(+) and AP2/pcDNA3.1 were kindly provided by Ronald J. Weigel (Stanford, CA).
Promoter-Reporter Constructs
The -378/+8 region of the rat P450scc (33) was cloned by a PCR-based approach (85) using rat genomic DNA (86) as template. 5'-HindIII and 3'-XbaI cloning sites (lowercase) were included in addition to ggcc in the forward (5'-ggccaagcttGAGTTATAAAGGGCCTGGGGAC) and the reverse (5'-ggcctctagaGTGTCTCTGCCCCAAACCTCCA) primers, respectively. Mutations of SCC1 were obtained by use of a larger reverse primer (-60/+8, 5'-ggcctctagaGTGTCTCTGCCCCAAACCTCCAGAGCCACACTTATAACCACTAACTAGCTCAATACTAAGAGAGGAGC) in which point mutations M3 and M5 were introduced as shown in Figs. 5 and 6. The PCR products were digested with HindIII and XbaI before ligation into promoter-less pCAT-Basic vector (Promega Corp.).
Cloning of -378/+8 P450scc into pEGFP-1 vector (CLONTECH Laboratories, Inc., Palo Alto, CA) was performed after SmaI and HindIII digest of the vector, followed by 5'-dephosphorylation of the DNA by calf intestinal alkaline phosphatase (Promega Corp.). The various -378/+8 CAT constructs were digested by XbaI and filled in by use of Klenow fragment. After precipitation, the digested plasmids were further digested by HindIII, and ligation (overnight at 16 C) was performed in the presence of a 10-fold molar excess of insert over vector DNA. All the generated plasmids were sequenced before use.
Cell Cultures
Giant trophoblast cells were prepared from E9.5 mouse uteri. To this end, uterine myometrium layers were removed from individual implantation sites, and the trophoblast layers were separated from the overlying decidua by a gentle scraping of longitudinally cut half-sites. The giant cells (20 mg tissue/ml) were enzymatically dispersed by a modification of a previous procedure (24) using serum-free DMEM-F12 medium containing 10 mg/ml BSA, 4 mg/ml collagenase-dispase, and 20 µg/ml DNase I. After 15 min incubation at 37 C, the cells were suspended by pipetting. This process was repeated twice before the enzymes were washed with serum-containing medium (10%), and the cell pellet was collected after a 2-min centrifugation at 200 x g. Cells were suspended in DMEM/F-12 growth medium containing 10% fetal bovine serum, 180 U/ml penicillin (Teva, Petach-Tikva, Israel), 0.24 mg/ml streptomycin (Sigma), and 25 mM HEPES (pH 7.4) and seeded (60–100 cells) into individual wells of a 24-well plate (Nunc, Copenhagen, Denmark). Cultures were maintained at 37 C in a humidified incubator, 95% air and 5% CO2.
Human choriocarcinoma JAR cells were grown in the same growth medium specified for the primary giant cells. HEK293 cells were grown in DMEM containing 10% fetal bovine serum, 2 mM glutamine, and the same antibiotics as in the primary giant cell medium.
Primary rat ovarian granulosa cells were prepared from E2-primed animals and grown in serum-free cultures exactly as described before (81).
Transfections
Lipofectamine.
On the morning after seeding, giant cells in each well were transfected with 250 ng DNA and lipofectamine reagent in serum-free medium according to the basic protocol provided by the manufacturer. Six hours after onset of transfection, the medium was replaced by a standard serum-supplemented culture medium, as specified above. On the next morning, the DNA was washed and fresh serum-free medium was added before a 8-Br-cAMP treatment (0.5 mM) commencing 5 h later. CAT analyses were performed after a 24-h incubation with or without 8-Br-cAMP. Cells expressing GFP were monitored by live confocal microscopy as of d 4 after transfection.
Electroporation.
Transfection of granulosa cell suspension (4 x 105 cells/0.8 ml) was performed by electroporation (54, 81) in the presence of -378/+8 CAT constructs (20 µg DNA), and the cells were seeded into four wells, each containing 0.5 ml medium (24-well plate). Extracts for CAT activity were made after a 6-h incubation of the cells with FSH (100 ng/ml).
Calcium Phosphate (Ca-Pi).
Twenty-four hours before transfection, JAR cells were harvested and reseeded in a 24-well plate at 50% confluence (0.5 ml/well). Transfection by Ca-Pi was conducted according to standard procedures (41) using 2.5 µg DNA per well. After 20 h incubation with the DNA precipitate, cells were washed two to three times and further incubated with fresh growth medium for an additional 24 h before harvesting and CAT analysis.
PEI.
Twenty-four hours before transfection, HEK293 cells were harvested and reseeded in a six-well plate at 50% confluence. On the day of transfection, DNA was diluted in 150 mM NaCl (2 µg/ml). Concomitantly, PEI was diluted in 150 mM NaCl to a final concentration of 2 mM. After a 10-min incubation at room temperature, both solutions were mixed (1:1 vol/vol ratio) and allowed to stand for another 15 min. During this incubation, the cell culture medium was changed to a serum- and antibiotics-free DMEM (2 ml/well). DNA-PEI mixtures were added to cell cultures (0.5 ml/well) and were allowed there for 5 h. Afterward, cells were washed and replaced with regular medium for another 24 h, after which cells were harvested for CAT assay.
Confocal Microscopy
To document the activity of -378/+8GFP constructs, giant cells were seeded onto a glass slide (12 mm diameter) that were hand glued at the bottom of a 10-mm hole drilled at the center of a 35-mm tissue culture dish (Falcon 3001, Becton Dickinson and Co., Plymouth, UK). These dishes allowed scanning of live cells by use of an inverted microscope attached to a 1024 confocal workstation (Bio-Rad Laboratories, Inc., Hercules, CA). After the indicated treatments, cells were viewed using a 40x oil immersion objective (numerical aperture 1.3). The excitation wavelength was 488 nm, and the emission of GFP was collected by use of a 525 ± 20 nm filter. All cells were scanned using same scanning conditions (laser power, gain of photo multiplier tube, iris size, and zoom). The acquired images were processed using Image Pro Plus 4.1 (Media Cybernetics, Silver Spring, MD) and displayed in pseudocolor images equivalent to 50–255 gray levels (scale provided in Fig. 5).
CAT Assay
After the indicated treatments, cell lysates were prepared and CAT activity was analyzed as previously described (54, 81). If not otherwise indicated, data are presented as percent of [14C]-chloramphenicol (Amersham International, Little Chalfont, UK) converted to its acetylated products (per protein and time of assay). Each figure presents either the mean ± SE of several independent transfections or depicts a distribution-free presentation to accommodate the large variation of the CAT activities observed in the primary giant cell cultures.
Purification of Human PNF
Purification of hPNF was conducted by a three-step procedure: 1) Human term placental cotyledons (200 g) were extracted in EMSA buffer A as described above (Tissue Extracts). Ammonium sulfate (dry form) was added to the resulting extract to attain 30% saturation. After chilling for 30 min at 4 C, the suspension was centrifuged at 15,000 x g for 30 min at 4 C, and the pellet was redissolved in 20 ml of buffer B [25 mM HEPES (pH 7.9), 10% (vol/vol) glycerol, 0.1 M KCl, 0.2 mM EDTA, 1 mM DTT, 1 mM PMSF, and 1 mM benzamidine, 1 mM sodium ortho-vanadate, 1 mM NaF, and 5 mM ß-glycerophosphate (87)]. 2) The placental proteins were diluted in 200 ml of buffer C [10 mM NaH2PO4 (pH 7.4), 0.1 M KCl, 10% glycerol, 1 mM EDTA, 0.5 mM DTT, 2.5 µM NaF, 0.5 mM sodium ortho-vanadate, 5 mg/ml aprotinin, 0.5 mM PMSF, 2.5 mM ß-glycerophosphate, and 0.5 mM benzamidine] and 50 ml of heparin-Sepharose 6 beads (fast flow, Amersham Pharmacia Biotech, Uppsala, Sweden) were added. The slurry was stirred at 4 C for 2 h, and bound proteins were eluted by buffer C containing a linear gradient of 0.1–1.5 M KCl. Eluted fractions were tested for binding activity by EMSA. 3) Final purification of PNF was obtained by the use of a DNA affinity column constructed by binding of biotinylated SCC1 to streptavidin agarose beads (Sigma). Active fractions of the heparin column (550–650 mM KCl) were pooled, and the buffer was adjusted to attain the composition of EMSA reaction cocktail (see above). Nonspecific double-stranded DNA (SCC1-DR0, 65 µg/ml) was added, and the proteins were allowed to adsorb to the SCC1 resin (0.4 ml) for 3 h at 4 C. Readsorption of unbound proteins was repeated (overnight) with a fresh aliquot of DNA resin. The DNA resin was stepwise washed with buffer D (glycerol 15%; 1 mM EDTA; 1 mM DTT; 0.1% Triton X-100; 10 mM NaH2PO4, pH 7.4) containing 150 and 300 mM KCl, and PNF activity was eluted by 650 mM of KCl in buffer D. The latter active fraction was repurified on SCC1 resin using less SCC1-DR0 DNA (33 ng/ml) before it was precipitated by adding deoxycholate (200 µg/ml, 30 min at 4 C) and then cold trichloroacetic acid (10%). After an overnight incubation at 4 C, the protein suspension was centrifuged at 14,000 rpm for 15 min at 4 C, and the pellet was dissolved in SDS-PAGE sample buffer. The ability of the purified protein to bind SCC1 was confirmed by Southwestern analysis performed as described previously (88) using 10% of the protein content. The rest of the purified protein was resolved by electrophoresis on a 7.5% minigel and stained by SimplyBlue SafeStain (Invitrogen, Groningen, The Netherlands, catalog no. LC6060). The stained band was excised, and its protein content was reduced (10 mM DTT) and modified with 100 mM iodoacetamide in 10 mM ammonium bicarbonate. Each gel piece was then treated with 50% acetonitrile in 10 mM ammonium bicarbonate to remove the stain from the protein, and dried. After rehydration with 10 mM ammonium bicarbonate containing trypsin (0.1 g), the protein band was incubated overnight at 37 C, and the resulting peptides were recovered with 60% acetonitrile with 0.1% trifluoroacetate.
The tryptic peptides were resolved by reverse-phase chromatography on 0.1 x 300-mm fused silica capillaries (internal diameter 100 µm, J&W Science, Folsom, CA), hand-filled with porous R2 (Roche Molecular Biochemicals). The peptides were eluted using an 80-min linear gradient of 5–95% acetonitrile with 0.1% acetic acid in water at a flow rate of about 1 µl/min. The liquid from the column was electrosprayed into an ion-trap mass spectrometer (LCQ, Finnigan MAT, San Jose, CA). Mass spectrometry was performed in the positive ion mode using repetitively full mass spectrometric scan followed by collision-induced dissociation of the most dominant ion selected from the first mass spectrometry scan. The mass spectrometry data were compared with simulated proteolysis and collision-induced dissociation of the proteins in the nonredundent database (NCBI) using Sequest software (Eng, J., and J. Yates, University of Washington, and Finnigan MAT).
Data Presentation and Statistical Analysis
CAT activities were normalized per protein and time of assay, and the data are presented as percent of the amount of 14C-chloramphenicol converted to the acetylated products. Results of CAT assays performed in granulosa cells, JAR, and HEK293 cells are presented as the mean ± SD of population of several independent transfections as indicated in each figure. Student’s unpaired two-tailed t test was performed using Excel ’97 (Microsoft Corp., Redmond, WA) statistical analysis functions. Differences between the activities of the indicated constructs were considered statistically significant at P < 0.05. Regarding the E9.5 cells, results of each individual experiment were presented whereas medians and confidence horns were used as statistical parameters (89).
|
ACKNOWLEDGMENTS |
|---|
and AP-2 expression plasmids. We thank Dr. U. Motro, of this institute, for assistance with some of the statistical analyses, and N. Sher for helpful discussions of this manuscript. |
FOOTNOTES |
|---|
Abbreviations: AP-2, Activating protein-2; 8-Br-cAMP, 8-bromo-cAMP; CAT, chloramphenicol acetyltransferase; COUP-TF, chick ovalbumin upstream promoter transcription factor; DNase, deoxyribonuclease; DR0, direct repeat with 0 bp spacing element; DTT, dithiothreitol; E9.5, embryonic d 9.5; GCNF, germ cell nuclear factor; GFP, green fluorescent protein; hCG, human choriogonadotropin; HEK, human embryonic kidney; 3ß-HSD, 3-ß-hydroxysteroid dehydrogenase; p.c., postcoitus; PEI, polyethyleneimine; PMSF, phenylmethyl sulfonyl fluoride; PNF, placental nuclear factor; P450scc, cholesterol side-chain cleavage cytochrome P450; RAR, retinoic acid receptor; RXR, retinoid X receptor; SF-1, steroidogenic factor-1; Sp1, selective promoter 1; StAR, steroidogenic acute regulatory; WT, wild-type.
Received for publication February 4, 2002. Accepted for publication April 18, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
5-4 isomerase and differential expression in gonads, adrenal glands, liver, and kidneys of both sexes. Proc Natl Acad Sci USA 88:8870–8874-2 on the release of PGF2 and PGE from epithelial cells of human pr
