| Nucleic Acids Research | Pages |
Identification of AP-2 as an interactive target of Rb and a regulator of the G1/S control element of the hamster histone H3.2 promoter
Introduction
Materials And Methods
Construction of plasmids
Electrophoretic mobility shift assays (EMSAs)
In vitro transcription and translation
Glutathione S-transferase (GST) pull-down assays
Transfection assays
Immunoprecipitation and western analysis
Results
The H3core contains an AP-2-like binding site directly 3[prime] to an AP-1-like site
Binding of the transcription factor AP-2 to the H3core
Localization of the AP-2 binding site on the H3core
Interaction of AP-2 with Rb in vitro and in vivo
Regulation of the H3core by AP-2 and Rb
Discussion
Acknowledgements
References
Identification of AP-2 as an interactive target of Rb and a regulator of the G1/S control element of the hamster histone H3.2 promoter
ABSTRACT
INTRODUCTION
The histone gene family encodes a group of small basic proteins (the core histones H2A, H2B, H3, H4 and the linker histone H1) that are essential for the packaging of newly synthesized DNA into nucleosomes. Expression of replication-dependent histone genes is regulated in part by transcriptional activation at the G1/S-phase boundary, coupled with the S-phase specific stabilization of the newly synthesized histone transcripts (1-3). To understand the molecular mechanisms which confer G1/S-phase induction of the mammalian histone genes, key cis-regulatory elements of a variety of histone genes have been identified in both the promoters and coding sequences (1-6).
Within a given histone gene subtype, each gene has conserved control elements. Correspondingly, different trans-acting factors for subtype specific histone gene transcription have been reported. For example, H1TF-1 interacts with a highly conserved AC-box and H1TF-2 with a CCAAT-box in the H1 promoters (7-9). In the case of H2A, it has been reported that E2F binds to a conserved E2F recognition motif in the human H2A.1 promoter and activates its promoter activity (10). For H2B, Oct-1 is a major promoter binding factor (11). For H4, Sp1 is needed for its high level transcription (12) while IRF-2 regulates its induction in S-phase (13). It has also been reported that a transcription complex termed HiNF-D containing CDC2, cyclin A, and a retinoblastoma (Rb)-related protein interacts with multiple recognition motifs in human H1, H3 and H4 promoters (14).
The H3 gene promoters ranging from wheat to human contain a highly conserved motif which resembles but differs from the AP-1 and CRE consensus binding sites (15). This motif is contained within a 32 bp region spanning -241 to -210, which by criterion of 5[prime] deletion, is critical for stringent G1/S-phase regulation of the hamster H3.2 promoter (4). This conserved sequence motif is implicated in the transcriptional control of both plant and vertebrate H3 genes and has been shown to be a nuclear factor binding site (15-18). Using site-directed mutagenesis, we have demonstrated a 10 bp region spanning -240 and -231 is required for the S-phase dependent increase of the hamster H3.2 transcription in cells synchronized by serum starvation or aphidicolin block (15,19). Further in vitro analysis has shown that the 3[prime] flanking sequence of this 10 bp region is also required for the formation of a protein complex which binds to a 40 bp H3.2 promoter subfragment spanning -250 to -211, referred to below as the H3core (15,19). Using nuclear extracts prepared from serum synchronized cells at various stages of the cell cycle in gel mobility shift assays, we have observed cell cycle fluctuation of the H3core binding activities (19). The H3core complex contains a minor component of AP-1, and a major protein species distinct from AP-1 as a high affinity H3core binding factor which is previously unidentified (19).
AP-2, a 52 kDa nuclear protein, is a retinoic acid-inducible and developmentally regulated transcription factor (20-25). AP-2 has a unique dimerization motif and binds to DNA as a dimer (26,27). Both transcriptional activation and repression effects have been postulated to act through AP-2 (28-35). AP-2 exerts its critical function in regulating gene expression in response to a number of different signal transduction pathways (22,23,32,36). Since the discovery of AP-2, three new isoforms (AP-2B, AP-2[beta] and AP-2[gamma]/AP-2.2) have been identified and shown to form dimers (20,37,38). One of them, AP-2B, is a naturally occurring transdominant-negative mutant which lacks the C-terminal DNA-binding domain but retains its dimerization ability with AP-2 (20). For clarity, the originally discovered AP-2 is referred to below as AP-2A. Interestingly, AP-2A activity can be modified by the SV40 large T antigen (34), E1A (35) and c-Myc (30) as well as by Rb when the two factors are brought into close proximity on an active promoter (39). AP-2A and AP-2[beta] knock-out mice exhibit phenotypes of severe developmental abnormality, increased apoptosis and embryonic or neonatal lethality (40-42).
We report here that the predominant component of the H3core complex is AP-2. In search of possible mechanism(s) whereby AP-2 can mediate cell cycle regulation of gene expression, we found that Rb can interact with AP-2 in vitro. When over-expressed, in vivo association of AP-2 and Rb can also be detected. In contrast to the majority of Rb binding proteins, the C-terminal domain (amino acids 768-928) of Rb is sufficient to bind AP-2. For AP-2A, the first 262 amino acids of the N-terminal portion are not sufficient for its binding to Rb. Over-expression of AP-2 results in the transactivation of the H3core when linked to a heterologous promoter in a sequence-specific but orientation-independent manner. Co-expression of Rb and AP-2 results in repression of the AP-2-mediated enhancement of the reporter gene activity through the H3core. Thus, AP-2 may contribute to cell cycle regulation of gene expression, at least in part, through its physical and functional interaction with Rb.
MATERIALS AND METHODS
Construction of plasmids
Synthetic oligonucleotides of the wild-type and mutated H3core were ligated into four tandem copies and subcloned in a pBluescript vector. Tandem copies of the H3core oligomers were then inserted into a pBL-tkCAT2 plasmid containing the herpes simplex virus (HSV)-thymidine kinase (tk) minimal promoter linked to the chloramphenicol acetyltransferase (CAT) reporter gene. Constructs containing the H3core either in the sense orientation (pRO[H3core]4CAT and its mutant form) or in the antisense orientation (pWO[H3core]4CAT and its mutant form) were confirmed by DNA sequencing.
Electrophoretic mobility shift assays (EMSAs)
Double-stranded oligonucleotides were end-labeled by T4 polynucleotide kinase using [[gamma]-32P]ATP and purified on 6% non-denaturing polyacrylamide gels in 1× TBE buffer (89 mM Tris-HCl, pH 7.5, 89 mM boric acid and 0.2 mM EDTA). Conditions for EMSAs have been previously described (19). For gel electrophoresis, 4-6% of non-denaturing polyacrylamide gels (with acrylamide to bis-acrylamide ratio of 55:1) were used.
In vitro transcription and translation
Plasmids cut with ScaI (pCMV/hAP-2A), PvuII (pCMV/mAP-2[beta]) (gifts of Dr R. Buettner, University of Regensburg, Germany) or XbaI (pBS/mD2) (gift of Dr H. Kiyokawa, Memorial Sloan-Kettering Cancer Center) were used for the translation of full-length proteins. Alternatively, plasmids cut with BglII (pCMV/hAP-2A) or NsiI (pCMV/hAP-2[beta]) were used for the generation of the N-terminal AP-2A(N262) or AP-2[beta](N252), respectively. Following protease K digestion, RNA transcripts were generated in vitro from 1 µg of linearized DNA templates by using the T3 or T7 mMESSAGE mMACHINE kit (Ambion, Austin, TX). Radiolabeled proteins were translated in vitro with the nuclease treated rabbit reticulocyte lysate system (Promega, Madison, WI) in the presence of [35S]methionine (20 µCi per 25 µl of reaction volume). AP-2 and cyclin D2 proteins were stored in the presence of 1 mM Na3VO4, 10 µg/ml leupeptin and 1 µg/ml aprotinin at -70°C.
Glutathione S-transferase (GST) pull-down assays
The GST-proteins used were as follows: GST, GST-AP-2A (gifts of Dr P. Kannan, University of Texas at Houston), GST-AP-2[beta](DBD) (gift of Dr R. Buettner, University of Regensberg, Germany) containing only the C-terminal DNA binding domain (DBD) of the murine AP-2[beta], GST-TBP and GST-Rb(379-928) (gifts of Dr D. Dean, Washington University), GST-Rb-A/B(379-792) and GST-Rb-C(768-928) (gifts of Dr J. Y. Wang, University of California, San Diego), GST-Rb-B/C(646-928) (gift of Dr Y. K. Fung, University of Southern California), GST-c-Myc(259-439) (gift of Dr K. Calame, Columbia University), and GST-H-Ras (gift of Dr D. Broek, University of Southern California). Following expression of GST-proteins in Escherichia coli and IPTG induction, GST-proteins, prepared by sonication in PBS plus 1% Triton X-100, were purified by affinity chromatography with glutathione-linked agarose beads (Sigma, St Louis, MO), and the protein yields were verified by Coomassie Blue staining.
Equal amounts of in vitro translated and 35S-labeled AP-2A, AP-2[beta] or cyclin D2 proteins (4 µl of each) were mixed with bacterially expressed GST-proteins bound onto beads (~25 pmol of each). Then the reaction mixtures were incubated in 50 µl of the binding buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 2 mM EDTA, 50 mM NaF, 1 mM Na3VO4 and 0.5% NP-40), in the presence of 1 mg/ml bovine serum albumin (BSA), at 4°C with gentle rocking for 2-4 h. The beads were then washed five times each with 800 µl of NETN buffer (20 mM Tris-HCl, pH 8, 100 mM NaCl, 1 mM EDTA and 0.5% NP-40). Proteins bound on beads were boiled in sodium dodecyl sulfate (SDS) sample buffer for 5 min, subjected to 10% SDS-PAGE, and detected by autoradiography.
Transfection assays
Conditions for transfection into the hamster fibroblast K12 cells were described (43). The AP-2 expression vectors pSV/hAP-2A and pSV/hAP-2B were gifts of Drs M. Tainsky and P. Kannan (University of Texas at Houston); pCMV/hAP-2A and pCMV/mAP-2[beta] were gifts of Dr R. Buettner (University of Regensburg, Germany); the full-length Rb (pH[beta]-Apr-1-Rb) and its vector (pH[beta]-Apr-1-neo) were gifts of Dr Y. K. Fung (University of Southern California, Los Angeles, CA). All transfection reaction mixtures contained either 3 µg of pCMV-lacZ, pRSV-lacZ, or 5 µg of grp78-lacZ (containing 2 kb of the rat grp78 promoter fused to the lacZ gene) which served as an internal control of transfection efficiency.
Immunoprecipitation and western analysis
K12 cells were transiently co-transfected with pCMV/mAP-2[beta] and pH[beta]-Apr-1-Rb using SuperFect reagent (Qiagen, Hilden, Germany). After 48 h, the cells were lysed in ELB buffer (50 mM HEPES, pH 7.0, 250 mM NaCl and 0.1% NP-40) in situ at 4°C for 10 min. The cell lysate was diluted in ELB buffer containing a final concentration of 125 mM NaCl, and precleared by incubating with 50 µl mixture of Protein-A Sepharose CL-4B (Sigma, St Louis, MO) and Protein-G Sepharose-4FL beads (Pharmacia-LKB, Piscataway, NJ) at 4°C for 1 h. Precleared lysates were immunoprecipitated with either normal rabbit or mouse serum (Sigma, St Louis, MO), anti-AP-2 or anti-Rb (IF8) (Santa Cruz, CA), or anti-v-Src (Oncogene, Cambridge, MA) antibody at 4°C for 4-6 h with gentle rocking. Protein complexes were collected by incubation with either Protein-A Sepharose CL-4B or Protein-G Sepharose-4FL beads at 4°C for 2 h followed by a quick spin. Protein complexes bound onto beads were washed in 1 ml of ice-cold buffer containing 50 mM Tris-HCl, pH 7.4, 125 mM NaCl and 0.1% NP-40 five times and resolved by a 10% SDS-PAGE. The western analysis was performed by using an anti-Rb antibody [1:500 dilution in T-TBS (20 mM Tris-HCl, pH 7.6, 137 mM NaCl and 0.1% Tween-20) containing 2% BSA] and the antigenic protein was detected by an enhanced chemiluminescence (ECL) kit (Amersham, Arlington Heights, IL).
RESULTS
The H3core contains an AP-2-like binding site directly 3[prime] to an AP-1-like site
The sequence of the H3core spanning -250 to -211 is shown in Figure
Figure 1. Localization of the AP-1-like and AP-2-like sites within the H3core of the hamster H3.2 promoter. (A) The sequence of the H3.2 promoter spanning -250 to -211 is shown. The H3 AP-1 and H3 AP-2 sites are marked by the thin and thick boxes, respectively. Arrows indicate the guanine residues protected by DNA methylation interference assays (19). (B) Sequence homology between the consensus AP-1 or AP-2 site and the H3core AP-1-like and AP-2-like sequences. The mismatched base is indicated by a black square.
Table 1.
Binding of the transcription factor AP-2 to the H3core
Previously, it has been determined that the H3core complex consists of at least two components, a minor amount of AP-1 and a predominant complex not yet identified (15,19). To search for its identity, synthetic oligonucleotide binding sites for E2F, Ets-1, EGR-1, Sp1 and AP-2 were used in competition assays in EMSAs (Fig.
Figure 2. In vitro binding of AP-2 to the H3core. (A) EMSAs were performed with 2 µg of CHO nuclear extracts (CHO NE), 1 ng of H3core as probe (asterisk) and 200 ng of poly dI·dC as non-specific competitor. A 25- or 50-fold molar excess of the various binding site competitors as indicated on top was used. The sequence of the competitor oligomers are shown in Table 1. The positions of the minor AP-1 complex (open arrowhead) and the major AP-2 complex (closed arrowhead) are indicated. (B) EMSAs were performed using 1 ng of bacterially expressed recombinant human AP-2A (rAP-2) (Promega, Madison, WI), 1 ng of the H3core as probe and 200 ng of poly dI·dC. The AP-2 complexes were competed against a 25- or 50-fold molar excess of the consensus wild-type or mutated AP-2 (AP-2m) binding site. The ability of AP-2 to bind the H3core was directly demonstrated by using bacterially expressed recombinant human AP-2A in EMSAs (Fig. 
Localization of the AP-2 binding site on the H3core
To map the AP-2 binding site within the H3core, mutated forms of the H3core oligonucleotide were synthesized (Table 1). The oligomer corem is identical to the wild-type H3core except that the sequence spanning -240 to -231 is mutated such that both the AP-1-like site and one base of its adjacent AP-2-like site are affected; 5[prime]mcore is mutated from -250 to -241, and AP-2mcore from -230 to -215 to destroy all AP-2-like GC-rich sequence motifs. As controls, the consensus site for AP-1 or AP-2, and a mutated form of the AP-2 consensus site (AP-2m ) were also used as competitors (Table 1). The EMSAs were performed using CHO nuclear extracts and radiolabeled H3core as probe. For comparison, radiolabeled consensus AP-2 site was also used in a similar competition assay.
First, our results indicated that the AP-2 binding site resides within the 3[prime] half of the H3core since the 5[prime]m core competed for the H3core complex as efficient as the wild-type H3core (Fig. Figure 3. Mapping of the AP-2 binding site within the H3core. EMSAs were performed with CHO nuclear extracts (CHO NE), using either the H3core (A and C) or the consensus AP-2 binding site (B) as probe (asterisk). The complexes were competed with a 25- or 50-fold molar excess of binding site competitors as indicated on top. In (A and C), the positions of the minor AP-1(open arrowhead) and the major AP-2 (closed arrowhead) complexes are indicated. In (B), the position of the AP-2 complex is indicated. In (C), the effects of various antisera (1.5 or 3 µg) and/or binding site competitors (25- or 50-fold molar excess) on the formation of the AP-1 and AP-2 complexes are shown. The [alpha]-AP-2 (C18) is an affinity-purified, rabbit polyclonal antibody against a carboxyl peptide of human AP-2A and is crossreactive with AP-2[beta] (Santa Cruz Biotechnology, Santa Cruz, CA), NRS is normal rabbit serum, and [alpha]-78 is a rabbit polyclonal anti-hamster GRP78 peptide antibody. The antibodies were pre-incubated with the nuclear extracts on ice for 15 min prior to the addition of probes. In lanes 4-9, a 25-fold molar excess of the consensus AP-1 binding site was added to the reaction prior to the addition of the antisera as indicated on top. Figure 4. Effect of AP-1 site mutation on the formation of the H3core complex. EMSAs were performed using CHO NE and the AP-1mcore or consensus AP-2 site as probe (asterisk). (A) The complexes were competed with a 25- or 50-fold molar excess of binding site competitors as indicated on top. (B) The formation of the complex was competed with increasing amounts (1.5 and 3 µg) of the antisera as indicated. The position of the AP-2 complex is indicated. (C) The EMSA was performed same as (B) except the consensus AP-2 site was used as probe. To dissociate AP-1 and AP-2 protein binding on the H3core, two approaches were used. First, AP-1 protein was removed by binding site competitions and second, the AP-1-like binding site was eliminated by mutation. In the first approach, with H3core as probe in EMSAs, AP-2 protein binding was abolished by an anti-AP-2 antibody (Fig. In the second approach, AP-1 binding activity was eliminated completely by mutation of the AP-1-like site in the H3core to generate an AP-1mcore used in EMSA (Table 1). In this case, the major complex formed with the AP-1mcore probe is AP-2 since competition by molar excess of the consensus AP-2 site alone completely eliminated the complex, whereas the AP-2m site had no effect (Fig. 
Interaction of AP-2 with Rb in vitro and in vivo
To elucidate the mechanism whereby AP-2 can affect the transcription of the replication-dependent H3.2 gene, we examined whether AP-2 associates with the cell cycle regulatory molecules. Previously, it has been reported that Rb can repress the transactivation activity of a truncated AP-2A in the form of a fusion protein (39). To test whether there is direct physical interaction between AP-2 and Rb, [35S]methionine-labeled AP-2A and AP-2[beta] were prepared by in vitro transcription and translation. The radiolabeled proteins were then mixed with bacterially expressed GST-AP-2A, GST-Rb(379-928) or GST protein alone. The GST-Rb(379-928) fusion protein contains a truncated N-terminal end but retains the functionally important pocket and C-terminal domains. The GST-bound proteins were resolved on SDS-PAGE and detected by autoradiography. While confirming the previous observations that AP-2A and AP-2[beta] can form homo- or hetero-dimeric complexes, this in vitro assay demonstrated that both AP-2 proteins can interact with Rb (Fig.
Figure 5. Interaction of AP-2 proteins with Rb in vitro. GST pull-down assays were performed with in vitro translated and 35S-labeled (A) AP-2A or (B) AP-2[beta]. Lane 1, 10% of the input radiolabeled protein used for the assay; lane 2, the 35S-labeled AP-2 protein was mixed with GST; lane 3, with GST-AP-2A; and lane 4, with GST-Rb(379-928). The proteins bound onto the GST-beads were eluted, applied to a 10% SDS-PAGE, and detected by autoradiography. The positions of the protein size markers (in kDa) run in parallel are indicated. To map the interaction domains between Rb and AP-2, different GST-proteins were expressed in E.coli, affinity-purified, and their concentration verified by SDS-PAGE and Coomassie Blue staining (data not shown). In vitro GST pull-down assays were performed with 35S-labeled and in vitro transcribed/translated full-length (Fig. To map the region of AP-2 necessary for Rb interaction, AP-2A and AP-2[beta] protein fragments, containing the N-terminal 262 and 252 amino acids respectively, were used in GST pull-down assays (Fig. To determine whether AP-2 and Rb can associate in vivo, co-immunoprecipitation experiments were performed. For this purpose, K12 fibroblasts were co-transfected with expression vectors for AP-2[beta] and Rb. The cell lysates were first immunoprecipitated with either normal rabbit or mouse serum, anti-AP-2, anti-Rb or anti-Src antibody. The immunoprecipitates were then resolved on SDS-PAGE, and western blotted against the anti-Rb antibody. A protein band with molecular weight of ~105-110 kDa was detected by the anti-Rb antibody in immunoprecipitates of anti-AP-2 and anti-Rb but not by the normal control serum or anti-Src antibody (Fig. 
Regulation of the H3core by AP-2 and Rb
To test the effect of AP-2 and Rb on transcription activity directed by the H3core, four tandem copies of the H3core were inserted into a pBL-tkCAT2 vector which contains a minimal TATA promoter sequence derived from the HSV-tk gene linked to a CAT reporter gene. The H3core tandem repeat unit was ligated either in the sense or antisense orientation to the HSV-tk minimal promoter. In parallel, a mutated form of H3core (corem, Table 1) unable to compete efficiently for AP-2 binding as verified by EMSA (Fig.
Figure 6. Mapping the interaction domains of Rb with AP-2. GST pull-down assays were performed with GST-proteins and 35S-labeled full-length (A) AP-2A, (B) AP-2[beta] or (C) cyclin D2 proteins. Lanes 1 and 8, 10% of the input protein; lane 2, GST alone; lane 3, GST-Rb(379-938); lane 4, GST-Rb-A/B(379-792); lane 5, GST-Rb-B/C(646-928); lane 6, GST-Rb-C(768-928); and lane 7, GST-H-Ras. Arrows indicate the positions of input or interacting proteins. These reporter gene constructs were co-transfected into hamster K12 cells with expression vector for human AP-2A (Fig. Further analysis showed that among the several members of AP-2 family tested, the highest stimulation (5-8-fold) was observed with the CMV-driven mouse AP-2[beta] (Fig. Figure 7. Identification of the interacting domain of AP-2 with Rb. GST pull-down assays were performed with GST-proteins and 35S-labeled N-terminal fragments of (A) AP-2A(N262) or (B) AP-2[beta](N252) or (C) full-length cyclin D2 proteins. Lanes 1 and 9, 10% of input protein; lane 2, GST alone; lane 3, GST-AP-2A; lane 4, GST-AP-2[beta](DBD); lane 5, GST-c-Myc(259-439); lane 6, GST-Rb(379-928); lane 7, GST-TBP; and lane 8, GST-H-Ras. Arrows indicate the positions of input or interacting proteins.
Histone and other replication-dependent genes have provided important model systems for studying the control of gene expression during the cell cycle (45,46). Our investigation into the factors interacting with the H3core, the G1/S control element of the hamster H3.2 promoter, led to the discovery that AP-2 is a major DNA binding component. Further, we showed here that AP-2 is a novel target for Rb, and its regulation of the H3core transcriptional activity can be modulated by Rb. While AP-2 is clearly important in mediating gene expression during embryonic morphogenesis and adult cell differentiation (40-42,47), increasingly, AP-2 has been linked to the cell cycle control and tumor progression. AP-2A interacts with both viral and cellular cell cycle regulators such as the SV40 large T antigen (34), E1A (35) and the transcription factor c-Myc (30). In addition, both AP-2A and AP-2[beta] have been shown to be over-expressed in cancer cell lines (48), and are able to transform cells when over-expressed (20,48,49). AP-2 binding sites have been identified in the promoters of a wide variety of cellular genes associated with cell growth and apoptosis. In various gene and cell systems, AP-2 is able to exert a wide range of effects, through both activation and repression of specific gene activity. For example, AP-2 can bind the promoter of the human growth hormone gene and activate its transcription (50). In human fibroblasts, AP-2-mediated transactivation contributes to the constitutively high expression of the human insulin-like growth factor binding protein-5, which can potentiate the effect of insulin growth factors on fibroblast growth (51). Other reports imply that AP-2 acts as a negative regulator of c-Myc (30), and inhibits cancer cell growth through the activation of the p21WAF1/CIP1 expression (52). The N-ras oncogene causes AP-2 transcriptional self-interference which leads to transformation (49). In mammary carcinoma cells, when c-erbB-2 is over-expressed, the AP-2 protein is expressed at elevated levels (48). Further, E1A-mediated repression of a matrix metalloprotease gene implicated in tumor metastasis has been correlated with its ability to bind and inactivate AP-2 (35). Recently, it was reported that loss of AP-2 results in up-regulation of a cell surface glycoprotein MCAM and an increase in tumor growth and metastasis of human melanoma cells (53). Nonetheless, as many promoters which contain an AP-2 site are not regulated in a growth-dependent manner, it is evident that the mechanism for the AP-2-mediated regulation of cellular promoters is complex and likely to involve other co-factors and regulatory signals unique to each gene system (30,32,33). Figure 8. Co-immunoprecipitation of AP-2 with Rb in vivo. Following transient co-transfections of mouse AP-2[beta] (4 µg) and Rb expression vectors (6 µg) in K12 cells, the cell lysates were immunoprecipitated (IP) with either normal rabbit serum (NRS), rabbit polyclonal [alpha]-AP-2 antibody, normal mouse serum (NMS), mouse monoclonal [alpha]-Rb (IF8) or [alpha]-Src mouse antibody. The immunoprecipitates were applied onto a 10% SDS-PAGE and subjected to western blot (WB) analysis using the [alpha]-Rb (IF8) antibody. With the exception of NRS or NMS, the immunoprecipitation experiments were performed in duplicates. The positions of Rb (closed arrowhead) and the protein size markers are indicated. In search of possible mechanism(s) whereby AP-2 can contribute to regulation of cell cycle gene expression, we found that AP-2 is a target of Rb. Through mapping of the interaction domains of these two proteins, we further determined that in contrast to the majority of Rb binding proteins, the C-terminal domain of Rb alone is sufficient for AP-2 binding. This novel finding is potentially significant since it implies that the binding of AP-2 to Rb may resemble that of E2F, a key regulator of cell cycle (54,55). While our manuscript was in preparation, it was reported that Rb and c-Myc activation of E-cadherin gene expression in epithelial cells acted through interaction with AP-2 (47). Studies of Batsche et al. (47) showed that the N-terminal domain of AP-2 and the oncoprotein binding domain and the C-terminal domain of Rb are required for the interaction. Our analysis demonstrated that the N-terminal portion of AP-2 may be required but is not sufficient for its binding to Rb (Fig. In the histone H3.2 promoter, the H3core sequence is required for its G1/S regulation. Here we provide evidence that AP-2 is a major binding component of the H3core complex. Through protein-protein interaction, Rb could be recruited onto the H3core and cooperate with AP-2 to regulate H3.2 expression during the cell cycle. Depending on the promoter, Rb can both activate or suppress gene expression through interaction with other regulatory proteins. Suppression of transcription by Rb can be achieved through sequestration of activating transcription factors (46,56,57), or by binding and inactivating transcription factors at the promoter (39). Furthermore, for some specific genes, Rb can also repress transcription through its association with histone deacetylase 1 (58). Additionally, Rb-mediated suppression of AP-2 activity through the H3core may involve changes in post-translational modifications of these proteins. Future investigations into Rb regulation of AP-2 activity will address these important issues. Figure 9. Transactivation of the H3core linked to HSV-tk promoter by AP-2A. (A) Five µg of pRO[H3core]4CAT reporter were co-transfected into K12 cells with various amounts (in µg) of the pSV/hAP-2A or pSV/hAP-2B expression vector as indicated on top. In all transfections, the total amount of DNA was adjusted to be the same with pBluescript and empty vector DNA. Each transfection was performed in duplicate. The autoradiograms of the CAT assays are shown. The CAT activity of the cells transfected with the empty vector was set as one. The fold of stimulation by the AP-2 expression vectors is indicated below. (B) Sequence-specific but orientation-independent transactivation of the H3core by AP-2A. The wild-type or mutant H3core element cloned in both sense and antisense orientations linked to HSV-tkCAT were co-transfected with 0.5 µg of pSV/hAP-2A expression vector (+) or its empty vector (-) into K12 cells. The autoradiograms of the CAT assays are shown. The promoter activity of pRO[H3core]4CAT in the absence of AP-2A was set as one. The relative fold of stimulation is indicated below each reaction. The diverse and pleiotrophic effect of AP-2 on transcription strongly argues it is likely to be cell-type and promoter specific. It has been shown that in complex cell cycle regulated promoters, the function of a particular control element is highly dependent on its interaction with other control elements as well as the basal transcriptional machinery (59). The H3.2 promoter is also highly complex and contains multiple positive and negative regulatory elements with opposing activities adjacent to the H3core (60). The mechanisms whereby AP-2 and Rb act in concert with the other regulatory factors to regulate H3.2 expression remain to be determined. Nonetheless, our findings that AP-2 binds to the G1/S regulatory domain of a replication-dependent H3.2 promoter and interacts with Rb provide a novel mechanistic explanation for a plausible role of AP-2 in cell cycle control. Figure 10. Repression of AP-2 transactivation through the H3core by Rb. (A) Five µg of the pRO[H3core]4CAT or pRO[H3corem]4CAT reporter as indicated on top were co-transfected into K12 cells with various amounts (in µg) of expression plasmid for full-length Rb or its vector. The relative promoter activities were determined and the wild-type pRO[H3core]4CAT activity in the absence of Rb co-expression was set as one. (B) The CAT reporter genes as indicated on top were co-transfected with either the empty vector (v), or the expression vector pCMV/hAP-2A or pCMV/mAP-2[beta] (0.5 µg of each), in the presence or absence of the expression vector for Rb (2.5 µg of each). The promoter activity of each reporter gene in the presence of the empty vector alone was set as one. The fold of induction with standard deviation is indicated. We are grateful to Drs David Ann, Daniel Broek, Reinhard Buettner, Kathryn Calame, Gadi Gazit, Douglas Dean, Yuen Kai Fung, Perry Kannan, Hiroaki Kiyokawa, Gregory S. Naeve, Binayak Roy, Michael Tainsky and Jean Y. J. Wang for the reagents and constructs. We thank Dr Douglas Dean for communication of unpublished results. We thank Edmund Kim and Julie Lau for technical assistance, and members of the Lee laboratory for their helpful discussions. F.W. is the recipient of the Charles Heidelberger Memorial Predoctoral Scholarship Award in Cancer Research. This research is supported by National Institutes of Health grant GM31108 awarded to A.S.L. 
DISCUSSION
ACKNOWLEDGEMENTS
REFERENCES
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 17 Oct 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
N. Iwamoto, S. Abe-Dohmae, M. Ayaori, N. Tanaka, M. Kusuhara, F. Ohsuzu, and S. Yokoyama
ATP-Binding Cassette Transporter A1 Gene Transcription Is Downregulated by Activator Protein 2{alpha}: Doxazosin Inhibits Activator Protein 2{alpha} and Increases High-Density Lipoprotein Biogenesis Independent of {alpha}1-Adrenoceptor Blockade
Circ. Res.,
July 20, 2007;
101(2):
156 - 165.
[Abstract]
[Full Text]
[PDF]
Y. Fu, J. Li, and A. S. Lee
GRP78/BiP Inhibits Endoplasmic Reticulum BIK and Protects Human Breast Cancer Cells against Estrogen Starvation-Induced Apoptosis
Cancer Res.,
April 15, 2007;
67(8):
3734 - 3740.
[Abstract]
[Full Text]
[PDF]
Y.-H. Cheng, B. J. Aronow, S. Hossain, B. Trapnell, S. Kong, and S. Handwerger
Critical role for transcription factor AP-2{alpha} in human trophoblast differentiation
Physiol Genomics,
June 17, 2004;
18(1):
99 - 107.
[Abstract]
[Full Text]
[PDF]
M. Campillos, M. A. Garcia, F. Valdivieso, and J. Vazquez
Transcriptional activation by AP-2{alpha} is modulated by the oncogene DEK
Nucleic Acids Res.,
March 1, 2003;
31(5):
1571 - 1575.
[Abstract]
[Full Text]
[PDF]
J. Zhao, B. K. Kennedy, B. D. Lawrence, D. A. Barbie, A. G. Matera, J. A. Fletcher, and E. Harlow
NPAT links cyclin E-Cdk2 to the regulation of replication-dependent histone gene transcription
Genes & Dev.,
September 15, 2000;
14(18):
2283 - 2297.
[Abstract]
[Full Text]
M. Li, P. Baumeister, B. Roy, T. Phan, D. Foti, S. Luo, and A. S. Lee
ATF6 as a Transcription Activator of the Endoplasmic Reticulum Stress Element: Thapsigargin Stress-Induced Changes and Synergistic Interactions with NF-Y and YY1
Mol. Cell. Biol.,
July 15, 2000;
20(14):
5096 - 5106.
[Abstract]
[Full Text]
F. Wu and A. S. Lee
YY1 as a Regulator of Replication-dependent Hamster Histone H3.2 Promoter and an Interactive Partner of AP-2
J. Biol. Chem.,
January 5, 2001;
276(1):
28 - 34.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (220K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (21)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Wu, F.
Articles by Lee, A. S.
Search for Related Content
PubMed
PubMed Citation
Articles by Wu, F.
Articles by Lee, A. S.
Social Bookmarking
What's this?