Abstract
Regulation of hairy and enhancer of split homologue-1 (HES-1) by estradiol and all-trans retinoic acid affects proliferation of human breast cancer cells. Here, we identify and characterize cis-regulatory elements involved in HES-1 regulation. In the distal 5′ promoter of the HES-1 gene, we found a retinoic acid response element and in the distal 3′ region, an estrogen receptor α(ER)α binding site. The ERα binding site, composed of an estrogen response element (ERE) and an ERE half-site, is important for both ERα binding and transcriptional regulation. Chromatin immunoprecipitation assays revealed that ERα is recruited to the ERE and associates with the HES-1 promoter. We also show recruitment of nuclear receptor co-regulators to the ERE in response to estradiol, followed by a decrease in histone acetylation and RNA polymerase II docking in the HES-1 promoter region. Our findings are consistent with a novel type of repressive estrogen response element in the distal 3′ region of the HES-1 gene.
Introduction
Hairy and enhancer of split homolog-1 (HES-1) is a basic-helix-loop-helix transcriptional repressor and an effector of the Notch signaling pathway. Studies of the HES-1−/− mouse show that the factor plays an important role during development (Ishibashi et al. 1995). By using breast cancer cells, we have previously shown that expression of HES-1 is downregulated by 17β-estradiol (E2; Strom et al. 2000)and upregulated by all-trans retinoic acid (atRA) in human breast cancer cells. This upregulation is important for the anti-estrogenic effect of atRA (Muller et al. 2002) and appears to affect the expression of the cell cycle regulator E2F-1. Forced expression of HES-1 inhibits an E2-mediated increase in E2F-1 and breast cancer cell proliferation (Hartman et al. 2004). Estrogen response in target cells is mediated by estrogen receptors (ERs). Upon activation, the ERs can bind to a palindromic estrogen-response element (ERE; GGTCAnnnTGACC) or ERE half-sites (GGTCA) in combination with an SP-1 or AP-1 site. Alternatively, ERs can bind cis-regulatory regions by tethering to either SP-1, NF-κB or AP-1 transcription factors (DeNardo et al. 2005). ER binding can lead to the recruitment of both activating and repressing co-regulators (Heldring et al. 2007). Recent genome-wide binding site mapping studies have shown that the majority of ER-binding sites are situated far away from the promoters that they affect (Carroll et al. 2005, 2006, Lin et al. 2007). Most known EREs positively affect transcription, however, ∼ half of the regulated genes in expression analyses are in fact downregulated (Frasor et al. 2003, Carroll et al. 2006, Lin et al. 2007). In this study, we demonstrate a complex regulation of HES-1 by E2 and all-trans retinoic acid in breast cancer cells where the respective response elements are distally located to the promoter region.
Materials and Methods
Cell lines, transient transfection, and luciferase assay
MCF-7, HEK 293, and COS-7 cells were maintained in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS; Saveen Werner AB, Lund, Sweden) and 0.1% gentamicin (Gibco). Transient transfections were performed using Lipofectamine 2000 (Invitrogen). Cells of order 4×105 were seeded on a 35 mm plate and 100 ng of each plasmid DNA in Opti-MEM (Gibco) were used. After 4 h, the medium was replaced with phenol red-free DMEM (Gibco) supplemented with 5% steroid-stripped (charcoal-treated) serum and possible additives, as indicated in the figures. Total cell lysates were harvested 48 h later, and luciferase activity was measured by using Luciferase assay kit (Biothema) in a Luminoskan Ascent luminometer (Labsystem). The luciferase activity was normalized to β-galactosidase activity that was measured by Power Wave X microplate reader (Bio-Tek Instruments Inc, Vinooski, UT, USA). Each transfection experiment was done in triplicate, or more, and presented as mean relative light units (±s.d).
Reporter constructs and chemicals
pGL3 promoter vector (Promega) was used as reporter vector. The 2× RARE vector contains the bps 17 883–17 842 upstream of HES-1 transcriptional start site, duplicated, and cloned into Sma I and Bgl II sites of the reporter vector. The reporter vector pHE contains 452 basepairs (originated from bps 22 858–23 310 downstream of the HES-1 transcriptional start site) cloned from human genomic DNA (Roche) by PCR into the Sma I site of pGL3 promoter vector. pHEΔA contains a deletion of bps 1–138 of the 452 DNA segment, pHEΔB 138–205, and pHEΔC 205–452. 17β-estradiol (E2) (Sigma) was dissolved in ethanol and used at 1 nM concentration. All-trans retinoic acid (atRA) (Sigma) was dissolved in DMSO (Sigma) and used at 2.5 μM concentration.
EMSA
COS-7 cells cultured on a 150 mm plate were transiently transfected with 5 μg pSG5–RARα and 5 μg pSG5–RXRα. After 24 h, the nuclear extracts were prepared and protein quantitated by using Bio-Rad's protein assay. The RARE probe (5′-TGCAGATCACTTGAGGTCAGGAGTTCAAGACCAGCCTGGCCAACAC-3′), the mutated RARE probe (5′-TGCAGATCACTTGAGCCCAGGAGTACTAGACCAGCCTGGCCAACAC-3′), the ERE probe HE (5′-CGCGGGAGCCAGGAGAACCTGACCTGGAGATGGGCCCTGCTCCAACCCACTGTGACCTTGGCAA-3′), HEΔA probe (5′-GGGCCCTGCTCCAACCCACTGTGACCTTGGCAA-3′) and HEΔB probe (5′-CGCGGGAGCCAGGAGAACCTGACCTGGAGAT-3′) were labeled with [γ32P]ATP (Amersham Biosciences) using T4 polynucleotide kinase (USB) and purified using a Sephadex G50 Nick column (Amersham Biosciences). The DNA–protein binding reactions were carried out in a 30 μl reaction containing binding buffer (20 mM Hepes-NaOH, pH 7.6, 50 mM NaCl, 10 mM dithiothreitol, 5% glycerol, 0.5 mM EDTA, and 0.3 mg/ml BSA), 1 μg poly (dI–dC; Amersham Biosciences), 15 μg nuclear extract for the RARE EMSA, and 3 pmol recombinant human ERα Invitrogen) for the ERE EMSAs. The reaction was left to incubate for 15 min at room temperature before adding ∼20 000 c.p.m. of the probe and was incubated for 20 min at room temperature. In the competition experiments, a 100-fold excess of unlabeled probe was added before adding the radiolabeled probe and incubated for 5 min. The sequences of the unlabeled probes were: vitellogenin A2 ERE probe (5′-AAAGTCAGGTCACAGTGACCTGATCAA-3′) and the random DNA (5′-AGCGGTGCCGCGTGTCTTGGAGCT-3′). The DNA/protein complexes were separated on a 5% native gel in 1×TBE. Gels were dried and subjected to phosphor imager analysis.
Chromatin immunoprecipitation assays
Chromatin immunoprecipitation assays (ChIP) were performed as described (Metivier et al. 2003). MCF-7 cells were grown in phenol red-free DMEM supplemented with 5% steroid-stripped FBS for a 72 h period. Following estrogen deprivation, a set of cells were incubated with 10 nM E2 for a 45 min period, and another set was mock treated with the ethanol carrier for E2. Chromatin was cross-linked with formaldehyde, and the complexes were immunoprecipitated. Antibodies used for the assays: RARα (sc-551), ERα (HC-20), HA-probe (Y-11), Pol II (N-20), SRC1 (sc-6096), SRC3 (sc-9119), NCoR (sc-1609), and SMRT (sc-20778) were purchased from Santa Cruz Biotechnology. Ac-H3 (06-911), and Ac-H4 (06-762) were from Upstate Biotech, whereas RIP140 (252182) was from ABCAM. DNA was isolated with QIAquick PCR purification kits obtained from Qiagen. The 5′ to 3′ sequences of the primers used for PCR were: HES-1 ERE fw (5′-GGCTGGGTGTGATTGAAAGA) and rev (5′-AAGGTCACAGTGGGTTGGAG); Greb1 ERE fw (5′-AGCAGTGAAAAAAAGTGTGGCAACTGGG) and rev (5′- CGACCCACAGAAATGAAAAGGCAGCAAACT); HES-1 promoter fw (5′-TCCTCCTCCCATTGGCTGAA) and rev (5′-ACGGGGGATTCCGCTGTTAT); pS2 fw (5′-CCATGTTGGCCAGGCTAGTC) and rev (5′-ACAACAGTGGCTCACGGGCT). A range of 30–35 cycles of PCR was performed and the products were visualized on ethidium bromide-stained agarose gels. Quantitative PCR was performed using SybrGreen (Roche) as a marker of DNA amplification on a LightCycler 480 (Roche) with 45 cycles of amplification. Fold changes in estrogen receptor and RNA polymerase II recruitment were determined using the delta–delta CT calculations with input samples used to normalize the data. Wilcoxon signed rank-sum tests were used to determine statistical significance.
Chromatin immunoprecipitation paired-end ditag analysis
The ChIP-PET analysis spanning 25 kb 3′ of HES-1 gene is a part of a ChIP-PET study presented elsewhere (Lin et al. 2007).
Results and discussion
Gene-regulatory elements are located in the distal promoter and distal sequence after the gene respectively
We have shown earlier that HES-1 is positively regulated by all-trans retinoic acid in MCF-7 cells (Muller et al. 2002). Therefore, we decided to search for retinoic acid response elements in silico within 100 kb of the gene, i.e. 50 kb downstream and 50 kb upstream of the gene. The search resulted in identification of an RARE at 17.8 kb upstream of the transcription start site. This element is a DR2 sequence, composed of AGGTCAggAGTTCA (Fig. 1A). The DR2 consensus retinoic acid response element has been shown to be frequent in the genome, as a result of Alu repeat expansion. Many of the DR2 class response elements have been shown to be located adjacent to known retinoic acid responsive genes (Laperriere et al. 2007). To determine whether the RARE adjacent to HES1 is involved in binding RAR within the genomic context, we performed ChIP experiments using antibodies against RARα. We observed receptor binding in the absence and the presence of atRA, although we observed a slight increase in the binding of the RARE following retinoic acid treatment (Fig. 1B). As expected, in the negative control experiment, RARα did not bind to the HES1 ERE. Taken together with the EMSA and reporter gene assays, these findings suggest the presence of a bona fide RARE that may be involved in regulating the expression of HES1.
Retinoic acid response element located the distal promoter. (A) Location of the RARE in relation to the HES-1 gene. (B) RARα binding to the HES1 RARE was determined following retinoic acid treatment or ethanol vehicle control treatment for 4 h by ChIP experiments followed by quantitative PCR using specific primers flanking the RARE or negative control primers flanking the HES1 ERE. Input controls were included to ensure equal loading of samples prior to the ChIP. The gel image is representative of three independent experiments. Control ChIP experiments using anti-HA antibodies were also carried out and did not show any non-specific binding to the RARE (data not shown). (C) Transient transfection of HEK 293 cells with 100 ng all-trans retinoic acid-responsive reporter (RARE in two copies in front of a SV40 promoter). pGL3 promoter vector was used as negative control. Cells were cultured with all-trans retinoic acid or vehicle (DMSO) for 48 h. The data are expressed as relative light units and normalized for β-galactosidase activity. The bars are the means from an experiment repeated nine times and presented as s.d. (±). (D) EMSA was performed with a probe encompassing RARE and nuclear extracts from RARα and RXRα-transfected COS-7 cells. An excess of unlabeled RARE or mutated RARE were used as competitors. BSA, BSA; NE, nuclear extracts; RARE, retinoic acid response element.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
A reporter vector was generated containing the putative RARE in two copies, and placed in front of the SV40 promoter. The multimerized element showed a 3.5-fold induction of the luciferase activity in response to retinoic acid (Fig. 1C). We also demonstrated in electrophoretic mobility shift assays (EMSAs) that nuclear extract from COS-7 cells transiently transfected with retinoic acid receptor α and retinoid X receptor α were able to bind oligonucleotides encoding the element (Fig. 1D).
ERα binding is detected 22 kb downstream of HES-1 transcriptional start site
To identify the ERE involved in the regulation of HES-1, we analyzed results from a whole-genome ERα binding site mapping experiment (Lin et al. 2007) and found an ERα binding site in the 3′ distal region of HES-1 (Fig. 2A). Five chromatin immunoprecipitated DNA fragments were mapped to 22 kb downstream of the HES-1 transcription start site (Fig. 2B), and the clustering of these fragments to the same region is indicative of a high confidence ERα binding site (Lin et al. 2007). The same binding site was also detected by Carroll and his colleagues in their genome-wide study of the ERα binding sites (Carroll et al. 2006), supporting the validity of the putative ERα binding site adjacent to HES-1. Analysis of this region with the Dragon ERE finder program (Bajic et al. 2003) indicated the presence of three ERE-like sequences denoted A, B, and C, where B was ½ an ERE (Fig. 2C). We then performed a time course measuring HES-1 mRNA levels after the addition of E2 respective atRA +E2 as can be seen in Fig. 2D; where atRA+E2 is added, the mRNA is upregulated at 1 h and stays high during the 5 h of the time course, while where only E2 is added, the mRNA starts to decline at 2 h and is maximally reduced at 3 h after addition of E2. The mRNA is not increased at any time point by E2 indicating that a repression complex is assembled shortly after E2 addition. The ERE containing region was PCR amplified from human genomic DNA and cloned into the pGL3 promoter vector. Transfections of this construct together with an expression vector for ERα into HEK 293 respective MCF-7 cells revealed that a functional ERE was present (Fig. 3A). The activity of the construct was increased more than fivefold after addition of 1 nM E2. Since the element functions as a negative regulatory element in the native gene, the in vitro function of the element as an activating element suggests that certain regulatory sequences, and/or chromatin structures, are lacking in the construct used in transient transfections. This was indeed demonstrated in our subsequent experiments showing the recruitment of nuclear receptor co-repressors and the corresponding decrease in histone deacetylation (Fig. 5) as potentially important mechanisms requiring an intact nucleosomal structure not present in the reporter gene construct. Deletion constructs of the ERE containing region showed that, whereas both ERE A and ERE B were required for activity, ERE C was not functional (Fig. 3A). The activity induced by E2 was completely abolished when ERE A was deleted. A very small estrogenic effect remained after the deletion of ERE B. No change in activity was seen when deleting ERE C. To our knowledge, this is the first time a suboptimal ERE has been shown to require ½ an ERE to be functional. The most common examples of interactions between the ERE and other response elements are with SP-1 or AP-1 sites and nearly half EREs.
ERα binding sequence 3′ of the HES-1 gene identified by the ChIP-PET technique. (A) Location of ERE in relation to the HES-1 gene. (B) A chromatin immunoprecipitation paired-end ditag analysis (ChIP-PET) spanning 25 kb 3′ of the HES-1 gene. The ten horizontal lines within brackets are the ChIP DNA fragments, using anti-ERα specific antibody that corresponds to the genomic location on chromosome 3 shown on top with numbers. Boxed are five of these fragments in a region of ∼ 500 bp. (C) The region found by ChIP-PET analysis contains three elements (bold letters) that show similarity to EREs; two full sites (A and C) and one ½ ERE (B). ERE, estrogen response element. (D) Time course study of HES-1 mRNA after the addition of 10 nM E2 respective 10 nM E2 +1μM atRA.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
Deletion analysis showing importance of putative EREs. (A) Deletion mutants of the region containing EREs in front of a SV40 promoter transiently transfected of HEK 293 cells with pcDNA3 FLAG ERα. Cells were incubated with or without E2, as indicated, for 48 h. The data are expressed as relative light units and normalized for β-galactosidase activity. The bars are the means from an experiment repeated three times and presented as s.d. (±). (B) ERE-containing region in front of a SV40 promoter transiently transfected into MCF-7 cells. (C) EMSA was carried out with a probe spanning the ERE-containing region and purified human ERα. Unlabeled ERE, consensus ERE (Xenopus vitellogenin A2), and random DNA was used in excess for the competition study. BSA, E2, 17β-estradiol; ERE, estrogen response element; p32, radiolabeled phosphate.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
Nuclear receptor co-factors are recruited to the HES-1 ERE following estrogen treatment. Histone H3 and H4 acetylation status and RNA polymerase II binding at the HES-1 promoter following estrogen treatment. (A) MCF-7 cells were treated with E2 for 45 min or 2 h, and the ERα-SRC1 complex was first isolated by anti-ERα ChIP, followed by an anti-SRC1 ChIP (re-ChIP) using specific antibodies against the receptor and co-activator. SRC1 recruitment was determined by qPCR using HES-1 ERE or control pS2 ERE primers, as indicated. Data are shown as a fold change in binding in E2-treated cells as compared with mock-treated (E2) controls. (B) Recruitment of SRC3 was determined using methods similar to the SRC1 experiments. The results represent average values from three replicate experiments, and the error bars denote standard error of means from the replicate data. (C) Following estrogen treatment for 45 min or 2 h, the MCF-7 cells were lysed and the ERα–NCoR complex was isolated by sequential ChIP assays using anti-ERα and anti-NCoR antibodies. Recruitment of NCoR to the HES-1 ERE and the control pS2 ERE was determined by qPCR. (D) RIP140 recruitment studies were performed in a similar manner as the NCoR experiments. (E) Recruitment of SMRT was determined also by sequential ChIP experiments described above. Results represent average values from three replicate experiments, and the error bars denote standard error of means from the replicate data. (F) Histone H3 acetylation was determined by ChIP assays using anti-acetylated histone H3 antibodies and qPCR with primers specific for the HES-1 promoter region. Acetylation status of histone H3 in the pS2 promoter was included as a positive control. (G) Acetylation status of histone H4 was determined in the same manner as histone H3 acetylation. (H) The ChIP assays using anti-RNA polymerase II (Pol II) antibodies were used to monitor polymerase binding following estrogen treatment. Average values from three replicate qRT-PCR experiments, using primers specific for the HES-1 promoter region, are shown in the graphs, and the error bars represent standard error of means from the replicate data. E2, 17β-estradiol; ERE, estrogen response element.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
Next, we investigated whether binding to the EREs corresponded to the observed transcriptional activity. Human recombinant ERα, with or without E2 (Fig. 3B, lane 2, and 3), bound to the ERE containing sequence and binding was abolished by non-labeled ERE (lane 4) or by non-labeled vitellogenin A2 ERE (lane 5) (consensus ERE) but not by random DNA (lane 6). Deletion of either A (ΔA) or B (ΔB) caused a complete loss of binding indicating that binding to the ERE-containing sequence correlated well with activity (lane 7 and 8). Since both elements are required for binding, this probably represents an example of cooperative binding. Cooperative binding can occur if there is an even number of DNA turns between the elements and this fits well with the current situation since one turn is 10 nucleotides and there are 39 nucleotides between the middle of the ERE consensus, A, and the ½ ERE, B.
ERα is recruited to the HES-1 ERE and associates with the promoter region following estrogen treatment
To validate and further characterize the in vivo interactions between ERα and the HES-1 cis-regulatory regions, we performed ChIP assays against ERα in MCF-7 breast cancer cells. Initially, to determine the dynamics of receptor recruitment, we assayed ERα binding at four time points following E2 treatment (45 min, 2, 12, and 24 h). ERα was recruited to the HES-1 ERE in a ligand-dependent manner following E2 treatment (Fig. 4A). The recruitment of ERα was found to be the most significant, compared with mock-treated controls, at 45 min and 2 h post E2 treatment (16.7 and 14.1-fold increase respectively; Fig. 4B). The increase in recruitment to the HES-1 ERE paralleled the recruitment of ERα to the positive control GREB1 ERE at the same time points (Fig. 4B). ERα recruitment diminished to basal levels as time progressed, and a result most probably was caused by estrogen-mediated proteasomal degradation of ERα (Nawaz et al. 1999). Based on this observation, we conclude that the 45 min and 2 h time points would be the most optimal in analyzing the regulation of gene expression using ChIP analyses; and thus all subsequent experiments focused on these two time points. To determine whether ERα, bound to HES-1 ERE, was physically associated with the promoter region of HES-1, we performed real-time PCR experiments using promoter region-specific primers following anti-ERα ChIP assays (Fig. 4A). At 45 min and 2 h, ERα also appeared to be associated with the HES-1 promoter region, though the binding was about 1/3 of that observed for the ERE.
ERα is recruited to the HES-1 ERE following estrogen treatment. (A) Chromatin immunoprecipitation (ChIP) assays to detect ERα binding at the HES-1 ERE were carried out in 10 nM E2 and ethanol- (EtOH) treated MCF-7 cells at 45 min, 2 h, 12 h, and 24 h following treatment. Interaction of ERα and HES-1 promoter region was assayed at 45 min and 2 h following E2 treatment. ChIP assays using the influenza hemagglutinin (HA) antibody were included as negative controls. (B) The recruitment of ERα to the HES-1 and GREB1 ERE was quantified with real-time PCR on DNA extracted from the ChIP assays. Fold change differences between E2-treated and mock-treated cells were calculated from the results normalized with input controls. E2, 17β-estradiol; ERE, estrogen response element.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
Both co-activators and co-repressors are recruited to the HES-1 ERE
Since ERα was recruited to the HES-1 ERE in response to E2, we next wanted to investigate which nuclear receptor co-regulators were in complex with ERα. We first performed re-ChIP assays using antibodies against ERα and then reprecipitated the complex using antibodies against co-activators SRC1 and SRC3 and showed that SRC1 was recruited to the HES-1 ERE at 2 h after the addition of E2, similar to the recruitment observed for the positive control pS2 ERE (Fig. 5A). The recruitment of SRC1 to the HES-1 ERE is in contrast to the recent observations by Kininis et al. who reported that the SRC proteins were not recruited to repression-associated EREs (Kininis et al. 2007). We note, however, that they only examined EREs in the proximal promoter regions of target genes and the mechanisms utilized by distally located sites, such as the one examined in this study, may differ in their nuclear receptor co-regulator repertoire. Furthermore, they utilized pan-SRC antibodies for their ChIP assays and their result may have differed if they also utilized the anti-SRC1 antibodies used in these studies. SRC3 was highly recruited at 45 min after E2 addition and remained bound at 2 h. SRC3 was recruited to the pS2 ERE at 45 min, and its recruitment further increased at 2 h (Fig. 5B). For the co-repressors, we also performed re-ChIP experiments to examine the recruitment of N-CoR, SMRT, and RIP140. N-CoR was recruited at 45 min but recruitment was strongly reduced at 2 h. By contrast, N-CoR recruitment to the pS2 ERE was reduced at 45 min and strongly reduced at 2 h after E2 addition (Fig. 5C). RIP140 was strongly recruited at 45 min to the HES-1 ERE but recruitment was reduced at 2 h, in comparison with the pS2 ERE where recruitment was reduced at 45 min and strongly reduced at 2 h (Fig. 5D). SMRT was modestly recruited at 45 min but was strongly recruited at 2 h. SMRT binding to the pS2 ERE was similar to the profiles observed for N-CoR where both co-repressors were expelled from the ERα complex following E2 treatment (Fig. 5E), and this is consistent with the negative results obtained by Metivier and colleagues in their time course study of co-regulator recruitment to the pS2 ERE (Metivier et al. 2003).
Histone acetylation and RNA polymerase II docking at the HES-1 promoter are decreased in response to E2 treatment
To investigate the effect of the ERE binding on the HES-1 proximal promoter, we examined by ChIP both the histone acetylation and RNA polymerase II (Pol II) docking. We found that the acetylation of histone H3 decreased after 45 min and was further decreased after 2 h, in strong contrast to the pS2 promoter where histone acetylation was increased after 45 min and 2 h and the same pattern was found for histone H4, consistent with the known associations of these modifications with transcriptional repression and activation respectively (Fig. 5F and G). In agreement with decreased histone acetylation of the HES-1 promoter, we also found that the interaction between ERα and the HES-1 ERE affected the recruitment of RNA polymerase II (Pol II) to the promoter region. We found a decreased recruitment of Pol II at both 45 min and 2 h in response to the treatment with E2 compared with no treatment controls while the recruitment of Pol II to the pS2 promoter was increased at 45 min and further increased at 2 h (Fig. 5H). These results suggest that the recruitment of ERα to the HES-1 ERE is associated with a decrease in Pol II docking in the promoter region of HES-1 and corresponds to the repression of HES-1 expression. The decrease in RNA Pol II recruitment in response to estradiol occurs despite the recruitment of co-activators. The role of the individual components in the repression complex assembled on the HES-1 ERE remains to be investigated as well as whether the distance between the HES-1 ERE and the promoter are important for repression by the HES-1 ERE.
In this report, we present molecular mechanisms underlying the regulation of HES-1 expression by retinoic acid and estrogen. We have found a novel negative ERE, novel in the sense that both an ERE and ½ an ERE are essential for ERα binding and transcriptional regulation. In addition, we show that both the co-repressors and co-activators are recruited to this repressor element bound by ERα following estrogen treatment. Based on these observations, a model is outlined in Fig. 6. Since both the co-activators and co-repressors were recruited to the HES-1 ERE, the function of the co-activators might be to assemble the repressor complex in an E2-dependent manner. Taken together, our findings provide insights into the transcriptional regulatory mechanisms of HES-1 expression and suggest potential molecular targets in the modulation of breast cancer cell proliferation.
Models of ERα complex formation at the HES-1 ERE and alterations in histone acetylation and RNA polymerase II binding in the promoter region at 45 min and 2 h following E2 treatment. E2, 17β-estradiol; ERE, estrogen response elements.
Citation: Journal of Endocrinology 200, 3; 10.1677/JOE-08-0094
Declaration of interest
P Müller, K W Merrell, J D Crofts, C Rönnlund, C-Y Lin, and A Ström have nothing to declare. J-Å Gustafsson is a consultant and shareholder of KaroBio AB.
Funding
This work was supported by Magnus Bergvall's Foundation and by the Swedish Cancer Fund.
Acknowledgements
We thank Eckardt Treuter and Laure Plantard for their discussions and review of the manuscript.
References
Bajic VB, Tan SL, Chong A, Tang S, Strom A, Gustafsson JA, Lin CY & Liu ET 2003 Dragon ERE Finder version 2: a tool for accurate detection and analysis of estrogen response elements in vertebrate genomes. Nucleic Acids Research 31 3605–3607.
Carroll JS, Liu XS, Brodsky AS, Li W, Meyer CA, Szary AJ, Eeckhoute J, Shao W, Hestermann EV & Geistlinger TR et al. 2005 Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122 33–43.
Carroll JS, Meyer CA, Song J, Li W, Geistlinger TR, Eeckhoute J, Brodsky AS, Keeton EK, Fertuck KC & Hall GF et al. 2006 Genome-wide analysis of estrogen receptor binding sites. Nature Genetics 38 1289–1297.
DeNardo DG, Kim HT, Hilsenbeck S, Cuba V, Tsimelzon A & Brown PH 2005 Global gene expression analysis of estrogen receptor transcription factor cross talk in breast cancer: identification of estrogen-induced/activator protein-1-dependent genes. Molecular Endocrinology 19 362–378.
Frasor J, Danes JM, Komm B, Chang KC, Lyttle CR & Katzenellenbogen BS 2003 Profiling of estrogen up- and down-regulated gene expression in human breast cancer cells: insights into gene networks and pathways underlying estrogenic control of proliferation and cell phenotype. Endocrinology 144 4562–4574.
Hartman J, Muller P, Foster JS, Wimalasena J, Gustafsson JA & Strom A 2004 HES-1 inhibits 17beta-estradiol and heregulin-beta1-mediated upregulation of E2F-1. Oncogene 23 8826–8833.
Heldring N, Pike A, Andersson S, Matthews J, Cheng G, Hartman J, Tujague M, Strom A, Treuter E & Warner M et al. 2007 Estrogen receptors: how do they signal and what are their targets. Physiological Reviews 87 905–931.
Ishibashi M, Ang SL, Shiota K, Nakanishi S, Kageyama R & Guillemot F 1995 Targeted disruption of mammalian hairy and enhancer of split homolog-1 (HES-1) leads to up-regulation of neural helix-loop-helix factors, premature neurogenesis, and severe neural tube defects. Genes and Development 9 3136–3148.
Kininis M, Chen BS, Diehl AG, Isaacs GD, Zhang T, Siepel AC, Clark AG & Kraus WL 2007 Genomic analyses of transcription factor binding, histone acetylation, and gene expression reveal mechanistically distinct classes of estrogen-regulated promoters. Molecular and Cellular Biology 27 5090–5104.
Laperriere D, Wang TT, White JH & Mader S 2007 Widespread Alu repeat-driven expansion of consensus DR2 retinoic acid response elements during primate evolution. BMC Genomics 8 23.
Lin CY, Vega VB, Thomsen JS, Zhang T, Kong SL, Xie M, Chiu KP, Lipovich L, Barnett DH & Stossi F et al. 2007 Whole-genome cartography of estrogen receptor alpha binding sites. PLoS Genetics 3 e87.
Metivier R, Penot G, Hubner MR, Reid G, Brand H, Kos M & Gannon F 2003 Estrogen receptor-alpha directs ordered, cyclical, and combinatorial recruitment of cofactors on a natural target promoter. Cell 115 751–763.
Muller P, Kietz S, Gustafsson JA & Strom A 2002 The anti-estrogenic effect of all-trans-retinoic acid on the breast cancer cell line MCF-7 is dependent on HES-1 expression. Journal of Biological Chemistry 277 28376–28379.
Nawaz Z, Lonard DM, Dennis AP, Smith CL & O'Malley BW 1999 Proteasome-dependent degradation of the human estrogen receptor. PNAS 96 1858–1862.
Strom A, Arai N, Leers J & Gustafsson JA 2000 The Hairy and Enhancer of Split homologue-1 (HES-1) mediates the proliferative effect of 17beta-estradiol on breast cancer cell lines. Oncogene 19 5951–5953.
