help button home button Endocrine Society Molecular Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pena, P.
Right arrow Articles by Pestell, R. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pena, P.
Right arrow Articles by Pestell, R. G.
Right arrowPubmed/NCBI databases
*Substance via MeSH
Molecular Endocrinology 13 (8): 1402-1416
Copyright © 1999 by The Endocrine Society

Activator Protein-2 Mediates Transcriptional Activation of the CYP11A1 Gene by Interaction with Sp1 Rather than Binding to DNA

Pilar Pena1, Anne T. Reutens1, Chris Albanese, Mark D’Amico, Genichi Watanabe, Amy Donner, I-Wei Shu, Trevor Williams and Richard G. Pestell

The Albert Einstein Cancer Center, Departments of Medicine and Developmental and Molecular Biology Albert Einstein College of Medicine Bronx, New York 10461
Department of Biology (A.D., T.W.) Yale University New Haven, Connecticut 06520
Division of Endocrinology, Metabolism, and Molecular Medicine (I-W.S.) Northwestern University Medical School Chicago, Illinois 60611


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The ovine P450 side chain cleavage (CYP11A1) enzyme gene, which catalyzes the initial enzymatic step in steroid hormone biosynthesis is transcriptionally regulated in cultured steroidogenic human trophoblastic JEG-3 cells. The ovine CYP11A1 promoter contains two GC-rich footprinted regions referred to as ovine footprints 5 (OF5) and OF3, which are well conserved among the CYP11A1 promoters of different species. These GC-rich sequences resemble activator protein-2 (AP-2)/Sp1 binding sites and were previously implicated in basal and cAMP-regulated activity of the bovine and ovine CYP11A1 promoters. In the current studies, AP-2 induced the ovine CYP11A1 promoter 4.5-fold in JEG-3 cells with full induction requiring the previously defined cAMP-responsive elements. Point mutation of OF3 abolished induction by AP-2, and OF3 was sufficient for induction by AP-2 when linked to a heterologous promoter. AP-2 induction of the CYP11A1 promoter required the basic region (N165-N278) and the carboxy terminus of AP-2 (N413-N437). In the course of investigating the mechanisms by which OF5 and OF3 regulated CYP11A1 transcription, we found that OF5 and OF3 bound Sp1 and Sp3 in JEG-3 cells. AP-2 did not bind OF5 or OF3 directly but rather formed a multiprotein complex with Sp1 in JEG-3 cells. AP-2 associated directly with Sp1 in vitro requiring the AP-2 basic region and the Sp1 carboxy terminus. AP-2 induced Sp1/Sp3 activity independently of AP-2 binding to DNA using a GAL4 paradigm. The Sp1 and Sp3 transactivation domains were linked to the DNA-binding domain of GAL4, and their activity was assessed using a luciferase reporter gene containing only the GAL4 DNA-binding sites linked to the minimal TATA site. AP-2 induced Sp1/Sp3-GAL4 activity 3- to 4-fold, requiring both the amino and extreme carboxy terminus of AP-2. We conclude that AP-2 can bind to and stimulate Sp1 activity and induces the ovine CYP11A1 promoter through conserved Sp1/Sp3-binding sites in JEG-3 cells. The induction of Sp1 activity by AP-2 may contribute to the induction of other genes that bind Sp1.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The P450 side chain cleavage enzyme (CYP11A1) gene encodes the enzyme catalyzing the initial step in steroid biosynthesis, converting cholesterol to pregnenolone (1). The CYP11A1 gene is expressed in all steroidogenic tissues. During development the CYP11A1 gene is also expressed in the central nervous system and the gastrointestinal tract (2, 3). Hormones and peptide growth factors induce the CYP11A1 gene in steroidogenic cells, and expression of the Cyp11A1 gene also increases during differentiation in the trophoblast (4). In the gonads, trophic hormones, including FSH and LH, act through G protein-coupled receptors to induce intracellular cAMP levels and CYP11A1 abundance. ACTH induces expression of the CYP11A1 gene in adrenocortical cells through a cAMP-dependent mechanism, and cAMP induces CYP11A1 abundance in placental JEG-3 cells (1, 5). Peptide growth factors including epidermal growth factor (EGF) and insulin-like growth factor (IGF) also induce CYP11A1 expression and promoter activity in cultured steroidogenic cells (6, 7, 8, 9, 10).

The CYP11A1 gene promoter regions of a number of different species contain conserved GC-rich DNA sequences governing basal and regulated expression (11, 12, 13, 14, 15). The ovine CYP11A1 promoter GC-rich sequences are referred to as ovine footprint 5 (OF5) and OF3 because nuclear proteins from human trophoblastic JEG-3 cells and ovine primary placental and adrenal cells bound these sequences in deoxyribonuclease 1 (DNAse 1) footprinting experiments (11). These GC-rich sequences were shown to be well conserved between species (11, 16). The porcine CYP11A1 promoter contains GC-rich sequences involved in IGF-responsive gene expression (10). One of the human CYP11A1 promoter GC-rich sequences binds Sp1-like proteins and conveys basal level regulatory function (17). The ovine and bovine CYP11A1 promoter GC-rich sequences control basal level and cAMP-responsive function (7, 16). The sequences governing cAMP-regulated expression of the human CYP11A1 gene in JEG-3 cells resembles an activator protein-2 (AP-2) binding site (13, 14, 15). Thus, GC-rich sequences, conserved between the CYP11A1 promoters of several species, play a role in basal and/or cAMP-regulated expression, do not resemble the canonical cAMP response element (CRE) known to bind CREB (18, 19), and may involve AP-2 and/or Sp1 transcription factors.

AP-2 was originally isolated as a protein binding to GC rich basal regulatory sequence of the SV40 and human metallothionein IIA (hMT IIA) promoters (20, 21). AP-2 is expressed in the trophoblast (22) and widely during development (23, 24). Activation of the cAMP pathway has been shown to increase AP-2 activity in Hela cells (21, 25), and AP-2 responsive elements have been identified in the regulatory regions of the genes encoding the murine major histocompatability complex H-2 Kb, collagenase, human GH, proenkephalin, ornithine decarboxylase, keratin 14, CG ß-subunit, and vascular endothelial growth factor (VEGF) (20, 21, 22, 25, 26, 27, 28). AP-2 can also form heterodimeric complexes binding Myc to regulate gene transcription through a Myc-binding site (27), and the AP-2 response element of the VEGF promoter is a GC-rich enhancer that bound Sp1 proteins (28), suggesting that AP-2 may regulate gene transcription through combinatorial interactions. Sp1 is a widely expressed transcription factor of approximately 100 kDa that binds GC-rich sequences through a DNA-binding domain consisting of three C-terminal zinc fingers (29). The activity of Sp1 is generally constitutive but can be regulated (30, 31). Sp1 activity can be induced directly by O-glycosylation (32, 33) or indirectly by other proteins such as the retinoblastoma protein pRB, which can induce Sp1 transactivation function (34).

Because AP-2 can bind to GC-rich sequences (GCC NNN GGC) and has been implicated in basal and cAMP-regulated gene expression and the conserved GC-rich sequences of the CYP11A1 promoters convey basal and/or cAMP-regulatory function, we evaluated a role for AP-2 in regulating the ovine CYP11A1 gene promoter. AP-2 induced the CYP11A1 promoter through the GC-rich conserved DNA sequences OF5 and OF3 in JEG-3 cells. Deletion within the basic helix-span-helix (bH-S-H) region of AP-2 abolished induction of the native CYP11A1 promoter. The AP-2 responsive elements bound Sp1 and Sp3 in JEG-3 cell nuclear extracts but did not bind AP-2. AP-2, however, directly associated with Sp1 in both native JEG-3 cells and in vitro binding assays. Deletion of the bH-S-H domain of AP-2 or the carboxy terminus of Sp-1 abolished binding. When Sp1 and Sp3 were linked to a GAL4 DNA-binding domain and assessed using a heterologous reporter system dependent upon GAL4 DNA binding only, AP-2 induced Sp1 and Sp3 activity. Thus, induction of Sp1/Sp3 transactivation by AP-2 can occur independently of AP-2 binding to DNA. These studies suggest AP-2 may regulate CYP11A1 promoter activity through direct association with Sp1.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
AP-2 Induction of the CYP11A1 Promoter in JEG-3 Cells Requires OF3
In previous studies we and others identified GC-rich sequences in the CYP11A1 promoter that were both basal and cAMP-regulatory elements (7, 16); however, CREB did not bind these sequences, suggesting distinct mechanisms were involved in regulating CYP11A1 gene expression through these sites. Previous studies had shown that cAMP induced AP-2 activity in HeLa cells (21, 25) and that GC-rich sequences bind AP-2. We therefore sought to determine whether the CYP11A1 promoter may respond to AP-2 through these GC-rich sequences. To examine a role for AP-2 in CYP11A1 transcription, cotransfection experiments were conducted using the -2700 bp CYP11A1 promoter (-2700 CYPLUC) and an expression vector encoding wild-type AP-2 in JEG-3 cells. Comparison was made with the empty expression vector. At 48 h CYP11A1 reporter activity was induced 4.5-fold by AP-2 in JEG-3 cells (Fig. 1AGo). The effect of AP-2 on the previously described wild-type AP-2 site (35) was assessed using the heterologous reporter (AP-2)2TKLUC. The reporter (AP-2)2TKLUC, which contains two AP-2 sites in tandem, was induced 6.5-fold by AP-2 in JEG-3 cells (Fig. 1AGo). The vector pA3LUC and the reporter (UAS)5E1BTATALUC were not induced by AP-2 (Fig. 1AGo). The reporter construction (UAS)5E1BTATALUC consists of multimeric binding sites for the GAL4 DNA-binding sites linked to the E1B TATA box, and therefore does not contain AP-2-binding sequences. These studies indicate that induction of the CYP11A1 promoter in the pA3LUC reporter requires specific DNA sequences.



View larger version (38K):
[in this window]
[in a new window]
  		Figure 1. AP-2 Induces the Ovine CYP11A1 Promoter in JEG-3 Cells and HepG2 Cells

Cotransfection experiments were conducted with the AP-2 expression vector or empty expression vector cassette (300–600 ng) in conjunction with the -2700 CYPLUC reporter (4.8 µg), (AP)2TKLUC, pA3LUC, or (UAS)5E1BTATALUC in JEG-3 (panel A) or HepG2 cells (panel B). Transfected cells were harvested after 48 h. The mean ± SEM, of n separate transfections as indicated in parentheses, are shown for the luciferase reporter constructions.

 
Transient expression studies were also conducted in HepG2 cells, which are deficient in AP-2 (36). The CYP11A1 promoter activity was induced 4.5-fold by AP-2 (Fig. 1BGo). The wild-type AP-2-responsive reporter (AP-2)2TKLUC was induced 5-fold by overexpression of AP-2.

To identify AP-2-responsive sequences within the CYP11A1 promoter, several CYP11A1 5'-promoter deletion constructs were examined in the presence of the AP-2 expression vector (Fig. 2Go). The reporter -2700 CYPLUC was induced 4.5-fold, the -117 fragment was induced 7-fold, and the -92 CYPLUC reporter was induced only 2-fold, suggesting elements located between -117 and -92 are also required for optimal induction by AP-2 (Fig. 2AGo). Further deletion to -77 resulted in a fragment that was induced 5-fold by AP-2, and deletion to -55 abolished AP-2 activation of CYP11A1. PCR-directed point mutation of the OF3 sequence within the -77 bp fragment was performed to produce the mutant reporter -77 CYPOF3mutLUC. The OF3 sequence was mutated in the context of the AP-2-responsive -77 bp fragment from CCG CCC TGT to CaG aaC TGT (Fig. 2BGo, inset). The effect of AP-2 on this mutant reporter was examined in JEG-3 cells (Fig. 2AGo). Unlike the wild type -77 CYP reporter, the -77CYPOF3mtLUC reporter was not induced by the overexpression of AP-2 (Fig. 2AGo). These results suggested that AP-2-responsive sequences were located between -117 and -92 and between -77 and -55.



View larger version (32K):
[in this window]
[in a new window]
  		Figure 2. The Ovine CYP11A1 Promoter OF5 and OF3 Are Required for Full Activation by AP-2

A, The CYP11A1 5'-promoter deletions were cotransfected with the AP-2 expression vector into JEG-3 cells. B (inset, Schematic representation of the ovine CYP11A1 promoter showing footprinted regions OF3-OF5. The sequence of OF5 and OF3 oligonucleotides and the point mutation within OF3 were used to create -77 CYPOF3mutLUC). The role of the footprinted regions OF3 in AP-2-dependent activity was examined in JEG-3 cells using a heterologous construction of the CYP11A1 promoter with comparison to the TK promoter. The data are shown as the mean ± SEM for n separate transfections as indicated in parentheses with significant differences from the adjacent 5'-promoter construct (P < 0.05) designated by *.

 
To determine whether the sequences within OF3 were sufficient for AP-2 responsiveness, this element was linked to the minimal thymidine kinase (TK) promoter and examined in cotransfection experiments. The OF3-containing reporter was induced 3.5-fold (Fig. 2BGo).

OF3 Conveys Basal Level Activity in JEG-3 Cells
Previous studies had shown that the CYP11A1 promoter OF5 region contained basal level regulatory elements that could also function as a cAMP-responsive element in JEG-3 cells (7). The homologous sequences of the bovine CYP11A1 promoter and a related GC-rich sequence homologous to OF3 also functioned as a basal regulatory element and cAMP-responsive sequences in murine Y1 adrenal cells (16). We examined the basal level activity of the OF3 sequence in JEG-3 cells (Fig. 3Go). To examine the basal level function of OF3 in the context of the native promoter, comparison was made between the -77 CYPLUC and the -55 CYPLUC constructions. Basal level activity was reduced approximately 50% by deletion from -77 to -55 (Fig. 3AGo). To determine whether OF3 conveyed basal activity in JEG-3 cells in the context of an heterologous promoter, the basal activity of the OF3 reporter was compared with the parental TKLUC reporter. The basal activity of (OF3)2TKLUC was 6-fold higher than TKLUC, and the basal activity of (AP2)2TKLUC was 10-fold higher than TKLUC (Fig. 3BGo). These findings suggest that OF3 conveys basal level activity in JEG-3 cells.



View larger version (26K):
[in this window]
[in a new window]
  		Figure 3. The CYP11A1 Promoter OF3 Conveys Basal Function in JEG-3 Cells

A, Basal activity of OF3 in the context of the native promoter was assessed by comparing activity of the -77 and -55 CYPLUC reporters. B, The role of the footprinted regions OF3 in basal activity was examined in JEG-3 cells using a heterologous construction of the CYP11A1 promoter, compared with the TK promoter normalized to 100%. The data are the mean of n separate transfections as indicated in parentheses. Transfected cells were harvested after 48 h. * Represents a significant differences from the adjacent 5'-promoter construct (P < 0.05).

 
AP-2 Induction of the CYP11A1 Promoter Requires the AP-2 Carboxy Terminus and bH-S-H Domain
In previous studies we identified several functional domains in AP-2 required for full transactivation function (35) (Fig. 4Go). To examine the domains of AP-2 required for regulation of the CYP11A1 promoter, the effects of wild-type and mutant AP-2{alpha} expression vectors were examined. Deletion of the proline-rich (AP-2 Int 31/77) or acid-rich region (AP-2 Int 97/165) did not significantly reduce induction of the -2700 CYPLUC reporter (Fig. 4AGo). Sequential deletion of the AP-2 amino terminus from N165 to N278 (AP-2 {Delta}N165 and AP-2 {Delta}N278) reduced induction of CYP11A1 approximately 6-fold (Fig. 4AGo). These findings suggested that the bH-S-H domain was required for full induction of the CYP11A1 promoter.



View larger version (40K):
[in this window]
[in a new window]
  		Figure 4. The Domains of AP-2 Required for Induction of the CYP11A1 Promoter Include the bH-S-H Region

A, A series of AP-2 expression vectors (35 ) were transfected with the -2700 CYPLUC reporter into JEG-3 cells. The -fold induction by the expression plasmid was derived by comparison with cells in which the parental plasmid without the cDNA was transfected. The mean data ± SEM of at least six separate transfections are shown. B, The induction of the -77 CYPLUC reporter by the wild-type AP-2 expression vector was compared with the AP-2 mutant (AP-2 {Delta}N278). The results are shown as % wild type induction. C, A separate series of experiments were conducted in which the effect of the wild-type AP-2 and mutant AP-2 plasmids were examined in the presence of the (OF3)2TKLUC reporter. The data are shown as mean ± SEM.

 
Overexpression of the plasmid encoding AP-2 {Delta}C390, a carboxy-terminal deletion mutant that does not bind DNA or dimerize, did not induce the CYP11A1 promoter (Fig. 4AGo), and the effect of the AP-2 {Delta}C390 expression plasmid was the same as the expression vector cassette alone. Overexpression of the mutant AP-2 {Delta}C413, which is DNA binding and dimerization competent, did not induce the promoter, indicating that the extreme carboxy terminus of AP-2 is also required for induction of the CYP11A1 promoter (Fig. 4AGo). In previous studies we had demonstrated that the AP-2 {Delta}C413 mutant is of similar stability to the wild-type AP-2 protein (35). In view of this finding it is possible that the AP-2 {Delta}C390 mutant was defective in CYP11A1 promoter activation both because of the loss of the carboxy terminus of AP-2 and the loss of the DNA-binding and dimerization domains. Together these data suggest that the bH-S-H and the carboxy terminus of AP-2 are required for full induction of the CYP11A1 promoter.

To examine further the role of the AP-2 bH-S-H domain in regulation of the CYP11A1 promoter, the -77 CYP11A1 promoter fragment was assessed with the AP-2 mutant (AP-2 {Delta}N278). The -77 CYP11A1 promoter fragment was examined because it contained the proximal 5'-promoter sequences induced by AP-2. The induction of -77 CYPLUC reporter activity by AP-2 was reduced 80% by deletion of the amino terminus (Fig. 4BGo). The heterologous reporter containing the OF3 sequence was also examined with either the wild-type or the AP-2 mutant expression vectors to determine the role of the amino terminus that includes the bH-S-H domain (Fig. 4CGo). In these experiments the OF3 reporter was induced 2.7-fold by AP-2 (Fig. 4CGo). The induction by AP-2 was sustained for the AP-2 {Delta}N165 mutant; however, further deletion to AP-2 {Delta}N278 abolished induction of the promoter. The loss of induction with deletion of the bH-S-H domain, rather than the amino-terminal activation domain, distinguishes the induction of the CYP11A1 promoter mechanistically from the induction of the hMT IIA promoter. These studies provide further support for a role of the AP-2 bH-S-H domain in CYP11A1 activation through OF3.

The Basal Regulatory Sequences of OF and OF3 Bind Sp1 and Sp3 in JEG-3 Cells
Electrophoretic mobility shift assays (EMSA) were performed using JEG-3 cell nuclear extracts to characterize nuclear protein interactions with OF5 and OF3 (Fig. 5Go). The mobility of the complex binding the CYP11A1 OF3 and OF5 site was similar (Fig. 5Go, A and B). The OF3-binding complex consisted of three bands (A–C), which were competed by 100-fold excess of either self-competitor or Sp1 sequences (Fig. 5AGo, lanes 2–3) but not by an equimolar excess of cold AP-2 oligodeoxynucleotides (not shown). Addition of Sp1 antibody supershifted the band labeled A (Fig. 5AGo, lane 6), and the Sp3 antibody supershifted complexes B and C (Fig. 5AGo, lane 8). The addition of Sp2 antibody (Fig. 5AGo, lane 7), preimmune serum (lane 5) or AP-2 antibody (lane 10) was without effect. In similar experiments with the OF5 sequence (Fig. 5BGo), the Sp1 antibody supershifted complex A (lane 6), and the Sp3 antibody shifted complexes B and C (lane 8). The OF5 complex was not shifted with antibodies to Sp2, Sp4, or AP-2 (lanes 7, 9, and 10) and was not shifted with preimmune serum (lane 4). An additional faster migrating band was observed with both OF5 and OF3; however, the naure of this complex is currently unknown.



View larger version (69K):
[in this window]
[in a new window]
  		Figure 5. The CYP11A13 OF3/OF5 Regions Bind Sp1 and Sp3 in JEG-3 Cells

The {gamma}32P-labeled CYP11A1 OF3 (panel A), CYP11A1 OF5 (panel B), and wild-type (panel C) probes were incubated with JEG-3 cell nuclear extracts. The addition of either 100-fold excess of wild-type or mutant competitor sequences, supershifting antibodies to AP-2 (AP) or Sp1 (Sp), or preimmune serum (Ctrl Ab) is indicated in the figure. SS- designates the supershifted complex. C, Binding of AP-2 to the hMT IIA site. The {gamma}32P-labeled wild type AP-2 probe (lanes 1 and 3–6) was compared with the CYP11A1 OF3 sequences (lanes 2, 7, and 8) using JEG-3 cell nuclear extracts or synthesized AP-2 protein. The addition of either 100-fold excess of wild-type competitor sequences (lane 4), an AP-2 antiserum (lane 5, AP), preimmune serum (Ctrl in lane 6), or bacterially synthesized protein (AP-2) (Promega Corp.) (lanes 1, 2) is indicated in the figure. The supershifted complex is indicated as ss.

 
To determine whether AP-2 synthesized in vitro was capable of binding to OF3 or OF5, comparison was made with the binding characteristics of the hMT IIA AP-2 site (Fig. 5CGo). AP-2 protein bound the hMT IIA AP-2 site but not OF3 either as purified protein (Fig. 5CGo, lanes 1 and 2) or as in vitro translated protein (data not shown). The hMT IIA AP-2 site formed a complex with JEG-3 cell protein (Fig. 5CGo, lane 3), which was supershifted with the AP-2 antibody (lane 5), and this complex was not shifted with preimmune serum (Fig. 5CGo, lane 6). The JEG-3 nuclear protein complexes binding the OF3 site (Fig. 5CGo, lanes 7 and 8) were of a different mobility compared with the complexes binding the hMT IIA AP-2 site. Thus, although AP-2 is capable of binding to the SV40 early promoter regions recognized by Sp1 (20), under our experimental conditions OF5 and OF3 bound Sp1 and Sp3 but not AP-2 in EMSA.

Endogenous AP-2 and Sp1 Associates in JEG-3 Cells
Our studies indicated that OF5 and OF3 bound Sp1 and Sp3 in JEG-3 cells. We hypothesized that AP-2 may induce CYP11A1 through these sequences by one or more different mechanisms. AP-2 may enhance the transactivation function of Sp1/Sp3. Alternatively, AP-2 may interact with Sp1 or alter the abundance of Sp1. Finally, it was possible that AP-2 may affect one or more of the many proteins that have been shown to interact with Sp1 and thereby indirectly affect Sp1 activity. For example AP-2 might recruit coactivator proteins known to regulate Sp1 activity. We first sought to determine whether AP-2 could directly interact with Sp1. We examined the immunospecificity of the Sp1 and AP-2 antibodies, by introducing expression plasmids encoding the cDNAs for AP-2 and Sp1 into 293 cells. The 293 cell derivative BOSC 23 was chosen because of its high transfection efficiency. Cells were harvested 48 h after transfection with expression vectors for AP-2 and Sp1. Western blotting was performed using specific antibodies for Sp1 and AP-2. Western blotting with the Sp1-specific antibody identified an immunoreactive band at 97 kDa (Fig. 6AGo, upper panel, lane 4). The relative abundance of this band was increased in the cells transfected with the Sp1 expression vector (Fig. 6AGo, upper panel, compare lane 1 and lane 3). This immunoreactive band was of identical mobility to that observed with in vitro translated Sp1 (not shown). Western blotting of the same extracts was also performed using an AP-2-specific antibody. An immunoreactive band was identified at approximately 46 kDa (Fig. 6AGo, middle panel, lanes 2, and 4), which was identical in mobility to in vitro translated AP-2 (Fig. 7Go). Reprobing of the Western blot with anti-{alpha}-tubulin antibody demonstrated equal amounts of protein in each lane (Fig. 6AGo, lower panel).



View larger version (38K):
[in this window]
[in a new window]
  		Figure 6. Association of AP-2 with Sp1 in Cell Extracts

A, Cell extracts were prepared from BOSC 23 cells transfected with: lane 1, expression plasmid vector (RSV vector, 5 µg); lane 2, RSV AP-2 (5 µg); lane 3, CMV-Sp1 (5 µg); and lane 4, RSV-AP-2 (5 µg) with CMV-Sp1 (5 µg). Cell extracts were prepared 48 h after transfection and subjected to electrophoresis on an SDS-10% polyacrylamide denaturing gel. Western blotting was performed of the membrane using antibodies to either Sp1 (JC6, upper panel), AP-2 (C-18, middle panel) or {alpha}-tubulin (5H1, middle panel). B, IP was performed of BOSC 23 cell extracts using an AP-2 antibody (C-18). The immune complexes were electrophoresed on an SDS-polyacrylamide gel with sequential Western blotting using either the Sp1 antibody (upper panel) or an AP-2 antibody (lower panel). The band indicated as n.s. is due to immunoprecipitating antibody binding antirabbit IgG secondary antibody. C, The IP was performed on JEG-3 cellular extracts using: lane 1, control antibody (C20); lane 2, the Sp1 antibody (1C6); and lane 3, the AP-2 antibody (C-18). The immune complexes were electrophoresed on a 10% SDS-polyacrylamide gel, and Western blotting was performed using the Sp1 antibody. The Sp1 immunoreactive band is indicated by the arrow.

 


View larger version (74K):
[in this window]
[in a new window]
  		Figure 7. In Vitro Binding Assay Association of AP-2 with Sp1

A, IPs were performed of coincubated synthesized proteins as detailed in the figure using the AP-2 antibody (C-18). Immunoprecipitated proteins were resolved by 8% SDS-PAGE, and 35S-labeled proteins were visualized by autoradiography (lanes 1–3) or Western blotting (lanes 4 and 5) as described in panel A above. B, Sp1 and AP-2 proteins were synthesized, coincubated, and then immunoprecipitated with the Sp1 antibody (1C6) as described in Materials and Methods. The immune complexes were subject to electrophoresis on an 8% SDS-polyacrylamide gel. Autoradiography of the 35S-labeled proteins (lanes 1–3) or Western blotting using the AP-2 antibody 5E4 (lanes 4 and 5) is shown. C, Equal amounts of either GST-Sp1 (lane 2), GST-AP-2 (lanes 3 and 5), or GST (lanes 1, 4, and 6) were incubated with in vitro translated 35S-labeled proteins for AP-2 or Sp1 for 1 h at 4 C and with protein A-sepharose beads for 6–12 h. Beads were washed four times, and bound proteins were solubilized in sample buffer and resolved by 8% SDS-PAGE. Gels were dried and 35S-labeled proteins that were bound to the fusion proteins were visualized by autoradiography. A component of the GST pull-down was also transferred to nitrocellulose, and Western blotting was performed using the AP-2 antibody 5E4 (24 ). D, Sp1 binding to wild-type AP-2 produced by in vitro translation (IVT AP-2) (lane 3) was compared with GST (lane 2) and with the effect of amino-terminal deletion of the AP-2 protein (AP-2 N278, shown as IVT AP-2{Delta}) (lanes 4–6). No binding was observed between the in vitro translated AP-2 N278 (lanes 4) and wild-type GST-Sp1 (lane 6). The effect of carboxy-terminal deletion of Sp1 (Sp1AB) on binding to AP-2 was assessed in lanes 7–10. Note the reduction in binding between the wild-type Sp1 (lane 8) and the carboxy-terminal Sp1 fusion protein (lane 9).

 
To determine whether Sp1 may associate directly with AP-2 in cultured cells, cellular extracts were subjected to immunoprecipitation (IP)-Western blotting. The AP-2 antibody was used to immunoprecipitate AP-2 from BOSC 23 cells, and the IP was subjected to SDS-PAGE. Western blotting of the immunoprecipitate was performed with either the AP-2 antibody C18 (Fig. 6BGo, lower panel) or the monoclonal AP-2 antibody 5E4 (24). An AP-2 immunoreactive band was identified as expected at 46 kDa (Fig. 6BGo, lower panel). An additional nonspecific band (n.s.) corresponding to immunoprecipitating antibody binding to the antirabbit IgG secondary antibody was found running with slower mobility (Fig. 6BGo, lower panel). Western blotting of the AP-2 immunoprecipitated material with the Sp1 antibody revealed an additional 97-kDa immunoreactive band (Fig. 6BGo, upper panel).

It was important to determine whether the association observed in 293 cells through overexpression occurred normally in JEG-3 cells. To determine whether Sp1-immunoreactive material was associated with endogenous AP-2 in JEG-3 cells, IP was performed with either an Sp1 or an AP-2 antibody. (Unlike the 293 cell experiments, the JEG-3 cells were not transfected with expression plasmids encoding either Sp1 or AP-2). The immunoprecipitates were electrophoresed on an 10% SDS-PAGE and subjected to Western blotting with an Sp1 antibody. An Sp1-immunoreactive band was identified at 97 kDa in cell extracts immunoprecipitated with either the Sp1 or the AP-2 antibody (Fig. 6CGo, lanes 2 and 3). The specificity of the IP was confirmed by the finding that equal amounts of a nonspecific antibody (Fig. 6CGo, lanes 1) or IgG (not shown) did not precipitate any Sp1 immunoreactive material.

In Vitro Protein-Binding Assays
Together, these studies indicate that an AP-2-immunoreactive antibody is capable of coprecipitating Sp1 in JEG-3 cells but do not demonstrate whether a direct interaction occurs between these two proteins. To extend these observations, a glutathione-S-transferase (GST) fusion protein for Sp1 was incubated with in vitro synthesized AP-2, and IP was performed with an AP-2-specific antibody (Fig. 7AGo, lane 1). 35S-labeled AP-2 was immunoprecipitated by the AP-2 antibody as expected (Fig. 7Go, lane 1). When GST-AP-2 was coincubated with in vitro translated Sp1 and immunoprecipitated with the AP-2 antibody (C-18, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), the band corresponding to 35S-labeled Sp1 was observed at 97 kDa (Fig. 7AGo, lane 2). This interaction was not observed when GST-AP-2 was incubated with an unrelated in vitro translated protein (Fig. 7AGo, lane 3). To confirm the presence of AP-2 in the GST-AP-2 used to immunoprecipitate the in vitro translated Sp1, Western blotting was performed on a component of the IP using the AP-2-specific antibody 5E4 (24), and the AP-2 immunoreactive band was identified in Fig. 7AGo, lane 4.

The reciprocal coimmunoprecipitation also provided further evidence for an interaction between Sp1 and AP-2 (Fig. 7BGo, lane 1). GST-Sp1 was coincubated with 35S-labeled AP-2 and immunoprecipitated with the Sp1 antibody. The immune complexes were resolved on a denaturing gel. An 35S-labeled band corresponding to AP-2 was identified in the Sp1 immunoprecipitate (Fig. 7BGo, lane 1). In vitro translated Sp1 was immunoprecipitated with the Sp1 antibody (Fig. 7BGo lane 3), and immunoreactive AP-2 was identified in the Sp1 immunoprecipitate by Western blotting (Fig. 7BGo, lane 4). Western blotting with the AP-2 antibody 5E4 confirmed the presence of GST-AP-2 associated with in vitro translated Sp1 in the IP performed with the Sp1 antibody (lane 4), but not with an unrelated control in vitro translate (lane 5). Together, these studies provide evidence for an association between AP-2 and Sp1 using in vitro binding assays.

To characterize this interaction further, in vitro binding assays were performed using GST-Sp1 and in vitro translated AP-2. Comparison was made with equal amounts of either GST protein or [35S]methionine-labeled control proteins. The full-length GST-Sp1 bound in vitro translated AP-2 (Fig. 7CGo, lane 2). The specificity of this interaction was indicated by the undetectable binding of in vitro translated AP-2 to equal amounts of GST protein (Fig. 7CGo, lane 1). In vitro translated Sp1 also bound the full-length GST-AP-2 (Fig. 7CGo, lane 3), and the specificity of this interaction was supported by the finding that no binding was observed between in vitro translated Sp1 and GST protein (Fig. 7CGo, lane 4). Western blotting of the GST pull-down product of GST-AP-2 and in vitro translated Sp1 using the AP-2- specific antibody 5E4 demonstrated an AP-2-immunoreactive band in lane 5.

To identify the domains of AP-2 and Sp-1 required for this protein-protein interaction, experiments were conducted with either wild-type or mutant expression plasmids shown schematically in Fig. 7DGo. The programmed lysate product for AP-2 (PL) (lane 1) and AP-2 {Delta}N278 (lane 4) were incubated with GST-Sp1 proteins. In vitro translated wild-type AP-2 bound to wild-type GST-Sp1 as shown in Fig. 7DGo (lane 3). To identify the region of AP-2 required for binding to Sp1, the amino-terminal deletion mutant AP-2 {Delta}N278 was produced by in vitro translation (IVT AP-2{Delta}, lane 4) and its abundance assessed on SDS-PAGE, and the protein was coincubated with GST-Sp1 protein. In contrast with the binding of wild-type AP-2 to Sp1, no binding was observed between Sp1 and the AP-2 {Delta}N278 deletion mutant (Fig. 7DGo, lane 6). A 5-fold longer exposure did not show any evidence of specific binding (data not shown).

The carboxy-terminal domain of Sp1 had previously been implicated in protein-protein interactions with GATA-1, YY1, and E2F-1. To determine whether the Sp1 carboxy-terminus was required for AP-2 binding, an Sp1 deletion mutant that lacks the carboxy-terminal zinc finger domain (GST-Sp1AB) was used (37) (Fig. 7DGo). Equal amounts of wild-type or mutant Sp1 GST fusion protein (GST-Sp1AB) was incubated with equal amounts of in vitro translated AP-2 protein. We confirmed that the amount of GST-Sp1AB protein was equal to the amount of wild type Sp1 protein through GST Western blotting of the protein transferred from an SDS-PAGE (data not shown). The binding of AP-2 to Sp1 was reduced approximately 90% by deletion of the Sp1 carboxy terminus (Fig. 7DGo, lane 8 vs. 9). Together, these findings demonstrate that the carboxy terminus of Sp1 is required for optimal binding to AP-2.

The Domains of AP-2 Required for Induction of Sp1 Transactivation Function
Because AP-2 did not bind the AP-2 response element of the CYP11A1 promoter, the current studies supported a model in which AP-2 activated CYP11A1 independently of binding to DNA. To examine the possibility that AP-2 could induce Sp1/Sp3 activity independently of AP-2 binding to DNA, we used an heterologous reporter system (UAS)5E1BTATALUC (38), which consists of multimeric GAL4 DNA-binding sites linked to a luciferase reporter gene. The Sp1 and Sp3 transactivation domains were linked to the GAL4 DNA-binding domain (Fig. 8AGo), and the basal and AP-2-regulated activity of Sp1 and Sp3 were examined in conjunction with the heterologous reporter construction. Sp1 and Sp3 conveyed basal enhancer function, and overexpression of the wild-type AP-2 expression vector enhanced Sp1 and Sp3 activity approximately 3-fold (P < 0.05) (Fig. 8CGo). AP-2 did not induce the (UAS)5E1BTATALUC reporter in the absence of the GAL4-Sp constructions (as shown in Fig. 1AGo). Because the (UAS)5E1BTATALUC reporter does not contain AP-2 binding sequences, these studies demonstrate that AP-2 can activate Sp1/Sp3 transactivation function without binding to DNA.



View larger version (32K):
[in this window]
[in a new window]
  		Figure 8. The Domains of AP-2 Required for Induction of Sp1 Transactivation

The reporter (UAS)5E1BTATALUC (4.8 µg) was transfected with expression vectors for GAL4-Sp1 or GAL4-Sp3 (panel A) and the expression vector encoding either wild type AP-2 (RSV-AP-2 (300–600 ng) or RSV vector (panel B). In panel C the effect of AP-2 on the activity of GAL4-Sp1 and GAL4-Sp3 is shown. Comparison was made between the effect of the AP-2 expression vector and equal amounts of the RSV parental vector on the activity of the GAL4 construction. The % increase in luciferase reporter activity is shown ± SEM for n separate experiments as indicated in the figure. In panel D the effects of the AP-2 wild type, the deletion mutant N278, and the carboxy-terminal AP-2 mutant expression plasmid (AP-2 {Delta}C413) were compared for their effects on the activity of the GAL4-Sp1 vector with the (UAS)5E1BTATALUC (4.8 µg) reporter. Basal activity of the GAL4-Sp1 plasmid with (UAS)5E1BTATALUC (4.8 µg) was normalized to 100%. The data are shown for 10–17 separate experiments.

 
To determine the domains of AP-2 required for induction of Sp-1 transactivation function, cotransfection experiments were conducted with the AP-2 expression vectors AP-2 {Delta}N278 and AP-2 {Delta}C413 (Fig. 8BGo). In contrast with the 3- to 4-fold induction of Sp1 transactivation function with the wild-type AP-2 plasmid, the AP-2 mutant AP-2 {Delta}N278 did not induce GAL4-Sp1 activity (0.85 ± 0.09-fold induction, n = 5) (Fig. 8DGo). These findings are consistent with the observation that the GST-Sp1 had reduced binding to the AP-2 {Delta}N278 mutant (Fig. 7DGo, lanes 3 vs. 6). The AP-2 {Delta}C413, which had been defective in activation of the CYP11A1 promoter (Fig. 4AGo), was also defective in activation of GAL4-Sp-1 activity, increasing activity only 50% (Fig. 8DGo). Together these studies indicate that specific domains of AP-2, both the amino and carboxy terminus, are required for full induction of GAL4-Sp1 activity in cultured cells.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
The current studies demonstrate that the optimal induction of the ovine CYP11A1 promoter by AP-2 required sequences between -77 and -55. Point mutation of OF3, in the context of the AP-2 responsive -77 CYPLUC reporter, abolished AP-2-induction, strongly suggesting the importance of the OF3 region. The OF3 sequences functioned as a basal level and AP-2-responsive element. In the current studies, OF3 and OF5 bound Sp1 and Sp3 in JEG-3 cells. The in vitro interaction between Sp1 and AP-2 required the amino terminus of AP-2 and the carboxy terminus of Sp1. AP-2 induced transactivation of Sp1 and Sp3 when they were linked to an heterologous GAL4 DNA-binding domain. Because this reporter system contains only GAL4-binding sites, these studies indicate that AP-2 can activate Sp1/Sp3 transactivation independently of an AP-2-binding DNA sequence. Induction of Sp1 transactivation function required the AP-2 amino terminus. Together, these findings are consistent with a model in which AP-2 may regulate the CYP11A1 promoter through an indirect mechanism involving Sp1-binding sites. These studies extend the repertoire of proteins previously shown to interact with Sp1 and provide evidence for an alternate mechanism by which AP-2 may regulate a variety of target genes containing Sp1-binding sites.

AP-2 can activate gene expression through binding a DNA sequence identified in the promoter region of many different genes (35). The DNA sequences implicated in signaling by AP-2, however, do not always resemble an AP-2 site (27), and the proline-rich activation domain of AP-2 is a poor transactivator when positioned in a distal position (39), suggesting AP-2 may function through combinatorial mechanisms. Our findings that AP-2 formed a protein-protein complex with Sp1 to regulate CYP11A1 promoter activity are consistent with recent studies showing that AP-2 forms a heteromeric protein-protein interaction with c-Myc (27) and E1A (40) to regulate gene transcription. AP-2 heterodimerized with c-Myc to activate the ornithine decarboxylase gene through a Myc-binding site (27). As with our studies in which induction of the CYP11A1 promoter required the AP-2 basic helix-span-helix (bH-S-H) domain and extreme carboxy terminus, the AP-2 bH-S-H domain was also required for the interaction with c-Myc (27) and the adenovirus E1A protein in vitro (40). AP-2 homodimerizes in solution or when bound to DNA via a H-S-H motif (35). Heterodimerization with AP-2 can occur either with other AP-2 isoforms or with the products of the related AP-2ß and AP-2{gamma} genes (41, 42). The bH-S-H motif of AP-2 may form a surface capable of interacting with several different proteins as this region was also necessary for binding to the amino terminus of the adenovirus E1A protein (40). The carboxy terminus of AP-2 was required for regulation of c-Myc transactivation function (27) and Sp1 activity when linked to a GAL4 DNA-binding domain. These studies demonstrate that AP-2 is capable of regulating gene transcription indirectly through forming heteromeric protein-protein interactions and suggest a common surface of AP-2 may be involved.

When Sp1 and Sp3 were linked to a GAL4 DNA-binding domain and their transactivation function assayed using a heterologous reporter containing only GAL4 DNA-binding sequences, AP-2 induced their activity 3- to 4-fold. Previous studies had demonstrated that Sp1 transactivation function can be induced by several different proteins including the retroviral oncoprotein v-Rel (43) and the retinoblastoma protein pRB (34). Induction of Sp1 transactivation function in our studies required the AP-2 bH-S-H domain and the extreme carboxy terminus. The physical interaction between Sp1 and AP-2 that we observed in native JEG-3 cell extracts required the Sp1 carboxy terminus when examined using in vitro pull-down assays. The carboxy-terminal zinc finger DNA-binding domain of Sp1 was also involved in binding to GATA-1 (44), YY1 (45), E1A (46), and E2F-1 (47). It has been proposed that the interactions between Sp1 and these additional transcription factors facilitate activity of specific target genes. It remains to be determined whether the AP-2/Sp1 complex functions as a molecular bridge to coactivator proteins and the basal transcription apparatus.

AP-2 is capable of inducing promoter activity through sequences that are responsive to cAMP (21, 22, 48, 49). The site of cAMP-induced expression of the acetyl-CoA carboxylase gene, for example, binds AP-2 (50). The human CG ß (hCGß) subunit gene promoter also contains an important basal level enhancer and cAMP-responsive region (CRE) that bind AP-2 (22, 50). The GC-rich OF5 sequence was previously shown to convey basal level and cAMP responsiveness in JEG-3 cells (7). The ovine CYP11A1 promoter is induced by cAMP in murine adrenal Y1 cells (11, 51), and the homologous regions corresponding to OF5 and OF3 of the bovine CYP11A1 promoter bound Sp1 from Y1 cell extracts and conveyed cAMP responsiveness (16). The bovine CYP17 gene CRE bound Sp1 in thecal and luteal cells; however, the binding affinity to these sites was unchanged by forskolin treatment (52), and the binding of Sp1 and Sp3 to OF5 and OF3 was unchanged after 24 h of cAMP treatment of JEG-3 cells (P. Pena and R.G. Pestell, unpublished). Although the expression of AP-2 is increased by cAMP in primary astrocytes (41) and in JEG-3 cells (P. Pena, R.G. Pestell, unpublished), consistent with a role for AP-2 as an intermediary protein in cAMP-induced gene expression, the role of AP-2 in cAMP-mediated activation of the ovine CYP11A1 gene remains to be determined. We are currently investigating whether cAMP regulates AP-2 binding to Sp1 in JEG-3 cells or whether cAMP recruits the coactivator PC4 to AP-2. In this regard it is of interest that the mutants of AP-2 that were defective in binding PC4 (53) were also defective in activating the -2700CYPLUC reporter. The promoter regions of a number of steroidogenic genes including the CYP21, CYP17, CYP19, CYP11A1, and ferrodoxin gene (7, 52, 54, 55) contain GC-rich sequences that function as important basal level regulatory elements. In some circumstances these GC-rich sequences convey cAMP-mediated enhancer activity (7, 52, 54, 55). As many of the steroidogenic CYP genes are coordinately induced by hormonal stimuli (1, 56), it will be of interest to determine the role of AP-2 in regulating the other CYP genes.

The current studies suggest a model in which the Sp1/Sp3 binding sites of the CYP11A1 promoter are required for induction by AP-2 (Fig. 9Go). Adrenal and ovarian steroidogenic cells contain SF-1, which can both bind DNA sequences in the CYP11A1 promoters of several species (57, 58, 59) and bind to Sp1 in two-hybrid assays (57). In steroidogenic cells containing SF-1, the coactivator p300 is brought to the basal apparatus through SF-1 (60). Placental cells, including JEG-3 and RCHO-1, are deficient in SF-1 (4, 61). In the current studies, AP-2 activated the CYP11A1 promoter through OF3, which bound Sp1, but not AP-2, in EMSA using JEG-3 cell extracts. Because there is flexibility in the requirement for G or C sequences in the consensus AP-2 site (GCC NNN GGC), the core of OF3 (CGC CCT GTC) differs primarily by the eighth nucleotide from the AP-2 site, being T rather than G/C. Mutation of OF3 (from 5'-CGC CCT GTC-3' to 5'-aGa aCT GTC-3') abolished induction by AP-2. Endogenous AP-2 in JEG-3 cells bound Sp1 in IP-Western blot analysis. AP-2 activated Sp1 when it was linked to a GAL4 DNA-binding site in a reporter system that contained only GAL4 DNA-binding sequences. These studies indicate that AP-2 is capable of inducing Sp1 activity independently of an AP-2 binding site. These studies suggest induction of CYP11A1 by AP-2 can occur independently of binding DNA, but do not formally exclude the possibility that AP-2 can bind to other sequences in the CYP11A1 promoter. We propose that AP-2 may form a bridge to Sp1 in JEG-3 cells, which contain AP-2 but lack SF-1, and thereby recruit coactivator (s) such as PC4 to the basal apparatus (Fig. 9Go). In addition to placental JEG-3 cells, AP-2 may also regulate CYP11A1 in other SF-1-deficient tissues, including the CNS and primitive gut, that express CYP11A1 (referenced in Refs. 3, 61, 62). In recent studies AP-2 was shown to directly bind the coactivator PC4 (53). PC4 stimulates transcriptional activity of several different activation surfaces, including the glutamine-rich activation surface of Sp1 (63). A recent model suggests that several PC4 molecules stabilize interactions between multiple proteins in the preinitiation complex through pairwise interaction (64), raising the possibility that through binding Sp1, AP-2 may recruit PC4 to the CYP11A1 promoter preinitiation complex (Fig. 9Go).



View larger version (107K):
[in this window]
[in a new window]
  		Figure 9. Hypothetical Model for Tissue-Specific Regulation of the CYP11A1 Promoter

In cells containing SF-1 (adrenal and ovary), Sp1 bound to OF5 or OF3 forms a secondary bridge to SF-1 (57 ), and SF-1 is capable of binding to p300 (60 ) linking activity to the basal apparatus. In cells deficient in SF-1 (placental trophoblastic cells), AP-2 binds Sp1 (53 ) and recruits PC4 to the basal apparatus (64 ).

 

   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Reporter Genes and Expression Vectors
The ovine CYP11A1 5'-promoter constructs linked to the luciferase gene in the vector pA3LUC (2700 CYPLUC, -183 CYPLUC, -117 CYPLUC, -92 CYPLUC, and -77 CYPLUC) were previously described (7). The -55 CYPLUC reporter was constructed by PCR using an oligodeoxyribonucleotide probe directed against the published sequence (11). The CYP11A1 OF3 sequence (5'-GGA GGA AGC TGA CCG CCC TGT TCA-3') was synthesized as complementary oligodeoxyribonucleotide strands to the CYP11A1 promoter region from -80 to -57 (11) and was cloned into the TK81pA3LUC reporter to form (OF3)2TKLUC. For comparison, the AP-2 sites from the human metallothionein IIA promoter were also linked to the minimal TK promoter to form (AP-2)2TKLUC. Point mutation was performed of the OF3 sequence within the context of the -77 bp CY11A1 reporter (-77 CYPOF3mutLUC) using PCR by changing the bases within OF3 from CCG CCC TGT to CaG aaC TGT. The reporter (UAS)5E1BTATALUC (38) consists of the (UAS)5E1BTATA sequences from Gal5CAT (35) cloned into the reporter pA3LUC.

The wild-type and mutant RSV-driven mammalian AP-2 expression vectors were previously described (35) (Fig. 4Go). The vector AP-2 {Delta}C390 has a deletion of the carboxy terminus between 390 and 437, which abolishes DNA binding; AP-2 {Delta}C413 contains a 24-amino acid carboxy-terminal deletion. AP-2 Int 97/165 has an internal deletion of the acid-rich activation domain. AP-2 INT/31/77 contains a deletion of the proline-rich activation domain. AP-2 {Delta}N51 and AP-2 {Delta}N278 encode sequential amino-terminal deletions of AP-2. In previous studies nuclear and cytoplasmic extracts were prepared from transfected HepG2 cells, and the presence of AP-2 derivatives was determined by immunoblotting (35). The data indicated that the wild-type and mutant constructs containing an intact DNA-binding domain were all expressed and translocated to the nucleus (35). CMV-Sp1 encodes the full-length Sp1 protein and was a gift from Dr. G. Gill and Dr. R. Tjian. The Sp1 and Sp3 activation domains were linked to the GAL4 DNA-binding domain to form GAL4-Sp1 (GAL4-Sp1(83–621) (29, 43) (a gift from Dr. G. Gill), and GAL4-Sp3(1–382) (a gift from Dr. J. Horwitz). The wild-type AP-2 cDNA (35) was cloned into either pT7ßSal to form pT7ßSal AP-2 or the GST expression vector pGEX2TK to form pGEX2TKAP-2 wt. The amino-terminal deletion fragment of AP-2 N278 was cloned in frame into pT7ßSal to form pT7ßSal AP-2 {Delta}N278. GST-Sp1 and the C-terminal deletion fragment of Sp-1 (Sp-1 AB) (37) were a gift from Dr. J. Horwitz, and the in vitro expression plasmid T3-FL-Sp1 plasmid was a gift from Dr. R. Tjian. The luciferase T7 control DNA and luciferase T3 control DNA were from Promega Corp. (Madison, WI).

Cell Culture, DNA Transfection, and Luciferase Assays
Cell culture, DNA transfection, and luciferase assays were performed as previously described (7, 65). JEG-3 choriocarcinoma cells (American Type Culture Collection, Manassas, VA) and CV-1 cells (a green monkey kidney cell line) were cultured in DMEM with 10% FCS, 1% penicillin, and 1% streptomycin. The 293 cell line derivative, BOSC 23, was a gift from Dr. D. Baltimore (66). Cells were transfected by calcium phosphate precipitation, the medium was changed after 6 h, and luciferase activity was determined after 48 h. At least two different plasmid preparations of each construct were used. In cotransfection experiments, a dose response was determined in each experiment with 300 ng and 600 ng of expression vector and the CYP11A1 promoter reporter plasmids (4.8 µg). Luciferase assays were performed at room temperature using an Autolumat LB 953 (EG&G Berthold, Gaithersburg, MD). Luciferase content was measured by calculating the light emitted during the initial 10 sec of the reaction, and the values are expressed in arbitrary light units (ALU). Background activity from cell extracts was typically less than 150 ALU/10 sec. The -fold effect was determined for 300–600 ng of AP-2 expression vector with comparison made to the effect of the empty expression vector cassette (RSV) or the mutant vector {Delta}C390, and statistical analyses were performed using the Mann-Whitney U test. Significant differences were established as P < 0.05.

Western Blots
For the detection of AP-2, Sp1, CYP11A1, GST, and {alpha}-tubulin protein, cell extracts were prepared as previously described (67). Western blotting was performed using antibodies to the rat Cyp11a1 (51, 68), the AP-2 monoclonal antibody 5E4 (24), an {alpha}-tubulin monoclonal IgM antibody (5H1) (69), and several antibodies from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) including antibodies to AP-2 (C-18), GST (B-14), Sp1 (SC-1C6), and E2F-1 (C-20). Reactive proteins were visualized using an antirabbit horseradish peroxidase second antibody for CYP11A1 and AP-2 and an antimouse antibody for {alpha}-tubulin and AP-2 (5E4) and the enhanced chemiluminescence system (Kirkegaard & Perry Laboratories, Gaithersburg, MD). The abundance of immunoreactive protein was quantified by phosphoimaging using a Computing Densitometer (Image Quant version 1.11, Molecular Dynamics, Inc., Sunnyvale, CA).

EMSA and DNAase 1 Footprinting
The oligodeoxyribonucleotides used in EMSA correspond to the CYP11A1 promoter OF3 region from -80 to -57 (5'-TGG AGG AAG CTG ACC GCC CTG TCA-3'), the OF5 region (-112 5' GTT TGG GAG GAG CTG TGT GGG CTG-3'), the wild-type AP-2 site (Promega Corp.), 5'-AGC TGA CCG CCC GCG GCC CGT-3'), and the wild-type Sp1 site (Promega Corp.), (5'-ATT CGA TCG GGG CGG GGC GAG-C3'). Nuclear extracts from JEG-3 cells were prepared as previously described (70). Five micrograms of nuclear extract protein were mixed with binding buffer (20 mM HEPES, pH 7.9, 80 mM KCl, 5 mM MgCl2, 2% Ficoll, 5% glycerol, 0.1 mM EDTA), 25 ng/µl single-strand (ss)DNA, and competitor or antibodies, on ice for 1 h. For competition assay, 10 pmol of unlabeled competitor oligonucleotide were used. In the case of antibody supershift experiment, the incubation conditions were identical, and 1 µl of anti-Sp1 (1C6), anti-Sp2 (K-20), anti-Sp3 (D-20), anti-Sp4 (V-20), or anti-AP-2 (C-18) (from Santa Cruz Biotechnology, Inc.) was used in each reaction. The 32P-labeled probe (10,000 cpm) was added after 1 h and incubated at room temperature for 30 min. The protein-DNA complexes were analyzed by electrophoresis through a 5% polyacrylamide gel in a 0.5 x TBE buffer (0.045 M Tris-borate, 0.001 M EDTA) and 2.5% glycerol. Autoradiography was performed at -70 C using XAR5 film (Eastman Kodak Co., Rochester, NY) with an intensifying screen.

Interactions between AP-2 and Sp1 in Cells
BOSC 23 cells were transfected with RSV-AP-2, CMV-Sp1 expression plasmid, or RSV control vector. After 36 h, cells were rinsed with PBS and harvested by scraping, and cell pellets were lysed for 10 min on ice in RIPA buffer [150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 7.5, with 1 mM sodium orthovanadate (Sigma Chemical Co., St. Louis, MO), 1 mg/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride (PMSF)]. Extracts were cleared by centrifugation and immunoprecipitated with rabbit polyclonal antibody (1 µg of IgG) specific for AP-2 (Santa Cruz Biotechnology, Inc.) using protein A-Sepharose as directed in the Santa Cruz product bulletin. Beads were washed four times with RIPA buffer and boiled in SDS sample buffer, and released proteins were resolved by 10% SDS-PAGE. The gel was transferred to nitrocellulose and Western blotting was performed using the AP-2 or Sp1 antibody.

In Vitro Protein-Protein Interaction
Coupled transcription-translation reactions were performed as previously described (7) using either T7 polymerase [pT7ßSal AP-2 (35)] or T3 polymerase (T3-FL-Sp1) and either T7 or T3 from the plasmid pGEMLUC as a control reaction. AP-2 and Sp1 proteins were labeled with [35S]methionine by coupled transcription-translation with a Promega Corp. TNT reticulocyte lysate kit using 1.5 µg of plasmid DNA in a total of 50 µl (7).

GST fusion proteins were prepared as previously described (71). GST fusion plasmids were used to transform Escherichia coli DH5{alpha}, and cells were grown to an absorbance of 0.5–0.7 at 600 nm. Fusion proteins were induced for 3 h with 0.2 M isopropyl-ß-D-thiogalactopyranoside, and crude lysates were prepared at 4 C. Cell pellets were spun down, resuspended in 20 mM Tris-HCL, pH 8.0, 1 mM EDTA, 1 mM PMSF, 1 mM leupeptin. Lysozyme (1 mg/ml) was added and incubation was performed on ice for 15–30 min. After the addition of 1/6 volume of 5 M NaCl and 1% Tween 20, and a 10-min incubation on ice, the cells were spun at 35,000 rpm for 15 min at 4 C. To the supernatant, 0.5 ml bed volume of GST beads (Pharmacia Biotech, Piscataway, NJ) was added, and the mixture was rotated for 6–12 h at 4 C, after which time centrifugation was performed at 1500 rpm for 5 min. The supernatant was discarded, and the pellet was resuspended in NETN buffer (0.5% NP-40, 20 mM Tris-HCL, pH 8, 100 mM NaCl, 1 mM EDTA). Glutathione elution buffer (0.5 ml of 10 mM glutathione in 50 mM Tris-HCl, pH 8) was added, and incubation was performed at 4 C for 1–4 h. The beads were separated through centrifugation, and the amount of protein was determined by Bio-Rad Laboratories, Inc. (Richmond, CA) assay and Coomassie Blue staining of an SDS-polyacrylamide gel. In addition, to ensure equal amounts of wild-type and mutant GST fusion proteins were used, Western blotting was performed using the GST antibody on serial dilutions of the fusion protein electrophoresed on an SDS-polyacrylamide gel.

In vitro protein-protein interactions were performed as previously described (72). The in vitro translated protein (6 µl of AP-2 or Sp1) was added to either 2 µg of GST, GST-AP-2, or GST-Sp1 in 200 µl of binding buffer (20 mM HEPES, pH 7.9, 1 mM MgCl2, 40 mM KCl, 0.1 mM EDTA, 0.1% NP-40, 1 mM PMSF, 1 mM leupeptin) and rotated for 1 h at 4 C. GST bead slurry (50 µl) was added for GST pull-down experiments. Protein A sepharose (50 µl) and 1 µg of AP-2 or Sp1 antibody (Sp1, PEP2-G, Santa Cruz Biotechnology, Inc.) was added in IP experiments. Samples were rotated for 6–12 h at 4 C and pelleted, and the beads were washed four times with binding buffer. Twenty five microliters of 2 x loading buffer were added, and the sample was electrophoresed on an SDS-polyacrylamide gel (8–10%). The electrophoresed samples were then transferred to nitrocellulose for 1 h and probed with antibodies to AP-2 (5E4 or C-18) and Sp1 (SC-1C6) as directed by the supplier. The gel was then fixed with 25% isopropanol and 10% acetic acid for 15 min, washed with Amplify (Amersham Pharmacia Biotech, Arlington Heights, IL) for 30 min to enhance the 35S signal, and dried for 1 h at 80 C, after which autoradiography was performed at -70 C using XAR5 film (Eastman Kodak Co.) with an intensifying screen.


   ACKNOWLEDGMENTS
 
We are grateful to Dr. G. Gill, Dr. R. Tjian, Dr. D. Baltimore, Dr. L. Binder, and Dr. J. Horwitz for plasmids, antibodies, and cell lines.


   FOOTNOTES
 
Address requests for reprints to: Richard G. Pestell, The Albert Einstein College of Medicine Cancer Center, Department of Developmental and Molecular Biology and Department of Medicine, Chanin 302, 1300 Morris Park Avenue, Bronx, New York, 10461.

This work was supported by NIH Grants R29CA-70897, RO1CA-75503, and RO1CA-77552–01; The Irma T. Hirschl Charitable Trust; and The Monique Weill-Caulier Charitable Trust (to R.G.P.). Work at the Albert Einstein College of Medicine was supported by Cancer Center Core NIH Grant 5-P30-CA-13330–26. G.W. was supported in part by a Travel Fellowship from the Aichi Health Promotion Foundation, Owari Kenyu Committee, and the Takasu Foundation. R.J.L. was supported by NIH Training Grant 5 T32 GM-0852–10. A.T.R. was supported by a P.F. Sabotka Postgraduate Scholarship from the University of Western Australia, and M.D. was supported by NIH Training Grant CA-09475–12. I-W.S. was supported by a David Shemin Undergraduate Fellowship at Northwestern University. P.P. was supported by a Fullbright Fellowship. T.W. is a Pew Scholar in the Biomedical Sciences and was also supported by American Cancer Society Grant RPG-98–096-01-MG.

1 These authors contributed equally to the manuscript. Back

Received for publication January 5, 1999. Revision received May 19, 1999. Accepted for publication May 20, 1999.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Waterman MR 1994 Biochemical diversity of cAMP-dependent transcription of steroid hydroxylase genes in the adrenal cortex. J Biol Chem 269:27783–27786[Free Full Text]
  2. Compagnone NA, Bulfone A, Rubenstein JL, Mellon SH 1995 Expression of the steroidogenic enzyme P450scc in the central and peripheral nervous systems during rodent embryogenesis. Endocrinology 136:2689–2696[Abstract]
  3. Keeney DS, Ikeda Y, Waterman MR, Parker KL 1995 Cholesterol side-chain cleavage cytochrome P450 gene expression in the primitive gut of the mouse embryo does not require steroidogenic factor 1. Mol Endocrinol 9:1091–1098[Abstract]
  4. Yamamoto T, Roby KF, Kwok SCM, Soares MJ 1994 Transcriptional activation of cytochrome p450 side chain cleavage enzyme during trophoblast cell differentiation. J Biol Chem 269:6517–6523[Abstract/Free Full Text]
  5. Miller WL 1988 Molecular biology of steroid hormone synthesis. Endocr Rev 9:295–318[Medline]
  6. Ritvos O, Voutilainen R 1992 Regulation of the cholesterol side-chain cleavage cytochrome p-450 and adrenodoxin mRNAs in cultured choriocarcinoma cells. Mol Cell Endocrinol 84:195–202[CrossRef][Medline]
  7. Pestell RG, Albanese C, Watanabe G, Johnson J, Eklund N, Lastowiecki P, Jameson JL 1995 Epidermal growth factor and c-Jun act via a common DNA regulatory element to stimulate transcription of the ovine P-450 cholesterol side chain cleavage (CYP11A1) promoter. J Biol Chem 270:18301–18308[Abstract/Free Full Text]
  8. Pestell RG, Albanese C, Watanabe G, Lee RJ, Lastowiecki P, Zon LI, Ostrowski M, Jameson JL 1996 Stimulation of the P-450 side chain cleavage (CYP11A1) promoter through ras- and Ets-2-signaling pathways. Mol Endocrinol 10:1084–1094[Abstract]
  9. Penhoat A, Chatelain PG, Jaillard C, Saez JM 1988 Characterization of insulin-like growth factor I and insulin receptors on cultured bovine adrenal fasciculata cells. Role of these peptides on adrenal cell function. Endocrinology 122:2518–2526[Abstract]
  10. Urban RJ, Shupnick MA, Bodenburg YH 1994 Insulin-like growth factor-1 increases expression of the porcine P-450 cholesterol side chain cleavage gene through a GC-rich domain. J Biol Chem 269:25761–25769[Abstract/Free Full Text]
  11. Pestell RG, Hammond V, Crawford R 1993 Molecular cloning and characterisation of the cAMP responsive ovine CYP11A1 (P-450-SCC) gene promoter: DNase 1 protection of conserved consensus elements. J Mol Endocrinol 10:297–311[Abstract]
  12. Momoi K, Waterman MR, Simpson ER, Zanger UM 1992 3',5'-cyclic adenosine monophosphate-dependent transcription of the CYP11A (cholesterol side chain cleavage cytochrome p450) gene involves a DNA response element containing a putative binding site for transcription factor Sp1. Mol Endocrinol 6:1682–1690[Abstract]
  13. Inoue H, Watanabe N, Higashi Y, Fujii-Kuriyama Y 1991 Structures of regulatory regions in the human cytochrome P-450 scc (desmolase) gene. Eur J Biochem 195:563–569[Medline]
  14. Huim DW, Staels B, Miller WL 1993 Basal transcriptional activity and cyclic adenosine 3',5'-monophosphate responsiveness of the human cytochrome p450scc promoter transfected into MA10 leydig cells. Endocrinology 132:546–552[Abstract]
  15. Moore CCD, Hum DW, Miller WL 1992 Identification of positive and negative placenta-specific basal elements and a cyclic adenosine 3'-5'-monophosphate response element in the human gene for P450scc. Mol Endocrinol 6:2045–2058[Abstract]
  16. Venepally P, Waterman MR 1995 Two Sp1-binding sites mediate cAMP-induced transcription of the bovine CYP11A gene through the protein kinase A signaling pathway. J Biol Chem 270:25402–25410[Abstract/Free Full Text]
  17. Chou S-J, Lai K-N, Chung B-c 1996 Characterization of the upstream sequence of the human CYP11A1 gene for cell type-specific expression. J Biol Chem 271:22125–22129[Abstract/Free Full Text]
  18. Meyer TE, Habener JF 1993 Cyclic adenosine 3',5'-monophosphate response element binding protein (CREB) and related transcription-activating deoxyribonucleic acid binding proteins. Endocr Rev 14:269–290[CrossRef][Medline]
  19. Montminy MR, Gonzalez GA, Yamamoto KK 1990 Regulation of cAMP-inducible genes by CREB. Trends Neurosci 13:184–188[CrossRef][Medline]
  20. Mitchell PJ, Wang C, Tjian R 1987 Positive and negative regulation of transcription in vitro: enhancer binding protein AP-2 is inhibited by SV40 T antigen. Cell 50:847–861[CrossRef][Medline]
  21. Imagawa M, Chiu R, Karin M 1987 Transcription factor AP-2 mediates induction by two different signal-transduction pathways: protein kinase C and cAMP. Cell 51:251–260[CrossRef][Medline]
  22. Johnson W, Albanese C, Handwerger S, Williams T, Pestell R, Jameson JL 1997 Regulation of the human chorionic gonadotropin {alpha} and ß-subunit gene promoters by AP-2{alpha}. J Biol Chem 272:15405–15412[Abstract/Free Full Text]
  23. Mitchell PJ, Timmons PM, Hebert JM, Rigby PW, Tjian R 1991 Transcription factor AP-2 is expressed in neural crest cell lineages during mouse embryogenesis. Genes Dev 5:105–119[Abstract/Free Full Text]
  24. Zhang J, Hagopian-Donaldson S, Serbedzija G, Elsemore J, Plehn-Dujowich D, McMahon AP, Flavell RA, Williams T 1996 Neural tube, skeletal and body wall defects in mice lacking transcription factor AP-2. Nature 381:238–241[CrossRef][Medline]
  25. Hyman SE, Comb M, Pearlberg J, Goodman HM 1989 An AP-2 element acts synergistically with the cAMP and phorbol ester-inducible enhancer of the human proenkephalin gene. Mol Cell Biol 9:321–324[Abstract/Free Full Text]
  26. Leask A, Byrne C, Fuchs E 1991 Transcription factor AP2 and its role in epidermal-specific gene expression. Proc Natl Acad Sci USA 88:7948–7952[Abstract/Free Full Text]
  27. Gaubatz S, Imhof A, Dosch R, Werner O, Mitchell P, Buettner R, Eilers M 1995 Transcriptional activation by Myc is under negative control by the transcription factor AP-2. EMBO J 14:1508–1519[Medline]
  28. Gille J, Swerlick RA, Caughman SW 1997 Transforming growth factor-{alpha}-induced activation of the vascular permeability factor (VPF/VEGF) gene requires AP-2-dependent DNA binding and transactivation. EMBO J 16:750–759[CrossRef][Medline]
  29. Courey AJ, Tjian R 1988 Analysis of Sp1 in vivo reveals multiple transcriptional domains, including a novel glutamine-rich activation motif. Cell 55:887–898[CrossRef][Medline]
  30. Saffer JD, Jackson SP, Annarella MB 1991 Developmental expression of Sp1 in the mouse. Mol Cell Biol 11:2189–2199[Abstract/Free Full Text]
  31. Saffer JD, Jackson SP, Thurston SJ 1990 SV40 stimulates expression of the transacting factor Sp1 at the mRNA level. Genes Dev 4:659–666[Abstract/Free Full Text]