| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
-positive Breast Cancer Cells1
Cancer Sciences Division, School of Medicine, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, United Kingdom
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(ER) status. Experimental models have demonstrated that hdm2 overexpression can promote breast cancer development. Here, we show that the elevated level of hdm2 protein in ER+ve breast cancer cell lines such as MCF-7 and T47D is because of transcription from the p53-inducible P2 promoter of hdm2. The P2 promoter is inactive in ER-ve cell lines such as SKBr3. Hdm2-P2 promoter activity in T47D cells is independent of p53, as well as of known regulators of the mouse mdm2-P2 promoter, including ER and ras-raf-mitogen-activated protein/extracellular signal-regulated kinase (MEK) mitogen-activated protein kinase (MAPK) signaling. We show that hdm2-P2 activity in T47D cells is dependent on the integrity of both an evolutionarily conserved composite binding site for AP1 and ETS family transcription factors (AP1-ETS) and a nonconserved upstream (nnGGGGC)5 repeat sequence. Lack of hdm2-P2 activity in ER-ve cells is shown to be a consequence of reduced transcriptional activation through the AP1-ETS element. Overexpression of ETS2 in SKBr3 cells reconstitutes AP1-ETS element-dependent hdm2-P2 promoter activity, resulting in increased levels of hdm2 protein in the cells. Our findings support the hypothesis that the elevated levels of hdm2 expression reported in cancers such as ER+ve breast tumors play an important role in the development of these tumors. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and its transcriptional target, progesterone receptor, has proven to be a useful predictor of prognosis and therapeutic response (3)
. ER is detectable in approximately two-thirds of breast cancers, and its expression correlates with a well-differentiated phenotype (4)
and a dependence on the mitogenic action of estrogen for tumor growth (2)
. Recent gene expression profiling studies have clearly demonstrated that ER status is associated with a distinct pattern of transcription of several hundred genes (5
, 6)
.
In common with several other human tumor types (7)
, resistance to chemotherapy in breast cancer has been correlated with the presence of inactivating mutations in the p53 tumor suppressor gene (8
, 9)
. This is consistent with the central role played by p53 in the stress-induced up-regulation of transcription of genes that induce cell cycle arrest and apoptosis (10)
. P53 gene mutation occurs in 
18% of breast tumors according to a meta-analysis of published reports (11)
and shows a strong correlation with ER-ve tumor status (9
, 12)
. The frequency of p53 mutation in tumors expressing both ER and progesterone receptor has been shown to be as low as 9% (12)
, which compares to an overall rate of 50–55% in all human cancers (13)
, indicating that the selection pressure for the acquisition of p53 mutation is relatively low in these breast tumors (14)
. Several lines of evidence now indicate that this reduced selection pressure is a consequence of multiple genotypic and phenotypic characteristics of these tumor cells. Firstly, it has been demonstrated that in a proportion of these cancers, the p53 protein is sequestered in the cytoplasm of the cell where it is unable to function as a transcription factor (15)
. Secondly, levels of transcription of p53 mRNA are low in some breast cancers because of reduced HOXA5 transcription factor activity (16)
. Furthermore, the p53-dependent apoptotic response is reduced in some breast tumors because of reduced expression of ASPP proteins (17)
. Finally, several reports have also demonstrated that the expression of hdm2, which is the major negative regulator of p53 protein levels and activity in the cell (18, 19, 20, 21)
, is up-regulated in ER+ve breast cancers at the protein and mRNA level (22, 23, 24, 25, 26, 27)
. Mdm2 promotes breast cancer formation in murine models (28)
and overexpression of hdm2 in human tumors such as sarcomas, in which the hdm2 gene is amplified (29)
, results in a reduced rate of p53 mutation in these cancers. The mechanism of increased hdm2 expression in ER+ve breast tumors is currently unknown and, because transcription of hdm2 is itself up-regulated by p53 (30
, 31)
, it is not known whether this increased expression is merely a consequence, rather than a cause, of the retention of wild-type p53 in these tumors.
Hdm2 expression is regulated by transcription from two distinct promoters, P1 and P2 (30)
. Transcription from P1 occurs at low levels in most cells, whereas P2 is highly induced by p53 because of the presence of evolutionarily conserved p53-REs in both the murine mdm2-P2 (32
, 33)
and human hdm2-P2 (30)
promoters. Studies dissecting the mechanisms that control transcription of mdm2 and hdm2 have focused on the P2 promoter, primarily because the human P1 promoter-dependent transcript is poorly translated (34)
. With the exception of the study by Zauberman et al. (30)
, which identified the p53-REs in the human hdm2-P2 promoter, mechanistic studies have, to date, been limited to the analysis of the murine P2 promoter. A number of functional, p53-independent response elements have been identified in this promoter, notably a thyroid hormone response element, which is active in pituitary cells (35)
, and a combination of a 5' ETS binding site and composite AP1-ETS site that is required for activation of the promoter by growth factor-dependent ras-raf-MEK-MAPK-signaling pathways (36
, 37)
. The mechanism whereby mdm2 expression can be regulated by other factors, most notably ER in ras-transformed murine fibroblasts (38)
, has yet to be defined. It is important to note, however, that levels of murine P2 transcript in nonstressed adult tissues has been demonstrated to be essentially p53 independent (39)
.
In this study, we have investigated the mechanism underlying the increased levels of hdm2 expression in ER+ve breast tumor cells. We have demonstrated that both P1 and P2 promoter-derived mRNA transcripts are differentially expressed in a panel of breast cancer cell lines and that the P2 promoter is activated by a p53-independent pathway in ER+ve cell lines such as MCF-7 and T47D. T47D cells provide a useful experimental system to study the role of estrogens on the proteins involved in p53-dependent cell cycle control (40)
. We have therefore used these cells to dissect the transcription factor response elements in the human hdm2-P2 promoter that drive hdm2 expression in ER+ve cells and have subsequently identified a transcription factor that is able to restore hdm2-P2 promoter activity in ER-ve cells.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
RNA Analysis.
Total RNA extraction was performed using RNAzol B (Biogenesis, Inc.). For RT-PCR analysis of hdm2 transcripts, 1 µg of RNA was reverse transcribed in a 40-µl volume using Moloney murine leukemia virus-reverse transcriptase (Promega) and O3 primer 5'-CTGCCTCGAGTCTCTTGTTCCGAAGCTGG-3'. Two µl of cDNA product were used as target in 50 µl of PCR reactions containing O1b/O3 or O2/O3 primer pairs for the amplification of hdm2-P1 and hdm2-P2 promoter-derived transcripts, respectively (O1b 5'-CTGGGGAGTCTTGAGGGACC-3', O2 5'-CCTGTGTGTCGGAAAGATGG-3'). PCR conditions were the same for both products (95°C 30 s, 58°C 30 s, 72°C 1 min), except that cycle number was optimized to ensure that amplification was terminated in the exponential phase: P1 transcript, 30 cycles; P2 transcript, 25 cycles. Products were of the predicted molecular size and were verified by subcloning into pGEMTeasy (Promega) and sequencing. RT-PCRs using oligodeoxythymidylic acid and ß actin primers were used to control for levels of input mRNA and were also terminated in the exponential phase of PCR (22 cycles). RPA was performed according to the manufacturers instructions (Ambion), using cloned O2/O3 PCR product as a probe. PCR and RPA results were quantified using a Kodak KDS1D imaging system.
Protein Analysis.
Cells were washed with PBS, pelleted by centrifugation at 1000 x g, snap frozen, and stored at -80°C. For immunoblotting, pellets were lysed for 15 min at 4°C in denaturing urea buffer [7 M urea, 0.1 M DTT, 0.05% Triton X-100, 25 mM NaCl, 20 mM HEPES (pH7.6)] then clarified by centrifugation at 13,000 x g for 10 min. Protein concentration was determined by the method of Bradford (Bio-Rad). Immunoblotting was performed by standard procedures, as previously described (41)
, and membranes were probed for hdm2 using mAb 2A9 or 2A10 (42)
, p53 (mAb DO-1 or DO-12, Serotec) or ETS2 (rabbit polyclonal C-20; Santa Cruz Biotechnology). Equal protein loading was confirmed on all immunoblots using rabbit anti-actin antibody (Sigma). Bands were visualized by chemiluminescence (Supersignal; Pierce) and quantified using a Fluor-S MAX system (Bio-Rad).
Plasmids, Transfections, and Reporter Gene Assays.
Genomic hdm2 sequence was amplified from normal human liver DNA and ligated into pGL3-Basic using the MluI/XhoI sites (Promega) to generate reporter construct hdm2luc01. The sequence of the inserted 895-bp region [(-602) to (+293) relative to the start of exon 2] was identical to that generated by the human genome project (AC026121.10) with the exception of two single bp differences (C-470T, which is a documented polymorphism, and G-133A). Additional constructs containing deletions of the hdm2 promoter (luc23, luc06, luc02, and luc03; Fig. 4A
) were generated by proof-reading PCR of hdm2luc01 using primers containing Mlu1 and Xho1 sites, followed by ligation into pGL3-Basic. Analysis of potential transcription factor binding sites in the hdm2-P2 promoter was performed using MatInspector.4
Mutations and deletions were introduced into hdm2luc01 using the QuickChange mutagenesis kit (Stratagene), and all constructs were verified by sequencing (MWG Biotech). Forward site-directed mutagenesis primers used are as follows (complementary reverse primers not shown): 
EBOX 5'-GGGGCATGGGGCAGGCTTTGCGGAGG-3' (3-bp deletion); ETSb/c 5'-GCTTCGGCGCGGTGATCGCAGGTGCC-3' (10-bp deletion); AP1 5'-GTGGGCAGGTACACTCAGCTTTTC-3' (2-bp substitution); and ETSa 5'-CTCAGCTTTAGCTCTTGAGCTGGTC-3' (2-bp substitution).
|
HEGO (44)
, from which we generated the dominant negative ER mutant S554fs by site-directed mutagenesis, pERE-Tkluc contains three copies of the ER binding element 5' to a minimal promoter, and the mouse mdm2-P2 promoter reporter vector, mdm2luc (30)
. md2.9CAT0 (32)
was used to generate murine mdm2-P2 promoter sequence. For reporter assays, cells in 60-mm dishes were transfected using Transfast reagent (Promega) with 2 µg of reporter plasmid, 0.25 µg of pCDNA3.1mychislacZ and 0.05–1 µg of other expression vectors where indicated. Cells were assayed 44 h later using Luclite reagent and a Topcount plate reader (Packard Bioscience) for luciferase reporter gene activity and a colorimetric ß-galactosidase assay to control for transfection efficiency. Luciferase activity was normalized to ß-galactosidase, and results are expressed as mean RLUs. Each transfection was performed in duplicate, and results are presented either as mean RLU ± SD (n = 2) for a representative experiment, or where indicated, results were calculated as a percentage of the activity of the full-length hdm2luc01 reporter vector in the cell line, and data from multiple experiments were pooled and expressed as mean ± SE. Statistical analysis was performed using a paired two-tailed t test. |
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
+ve Breast Cancer Cell Lines.+ve cell lines, whereas SKBr3 and MDAMB-231 are ER-ve. MCF-7 and ZR75.1 express wild-type p53 protein, whereas the other four lines express p53 protein containing inactivating point mutations at codons 194, 285, 175, and 280 in T47D, BT474, SKBr3, and MDAMB-231 respectively. The study by Gudas et al. (23)
demonstrated that hdm2 protein levels, detected by Western blotting with mAb IF2, were consistently higher in the ER+ve than ER-ve cancer cell lines. Hdm2 protein is known to be highly phosphorylated at a number of sites that can affect its immunoreactivity with several antibodies (45)
, and therefore, we first sought to confirm hdm2 expression levels in the breast cancer cell lines using a second mAb, 2A9 (42)
, the epitope for which (amino acids 155–222) is not known to be sensitive to posttranslational modification. As shown in Fig. 1A
(top panel), levels of the p90 form of hdm2 were highest in the 4 ER+ve cell lines, with MCF-7 cells having the lowest levels of this group and ZR75.1 the highest. Hdm2 protein levels in both of the ER-ve cell lines (SKBr3 and MDAMB-231) were determined, in two independent experiments, to be <10% of levels in ER+ve T47D cells, thus confirming the previous report (23)
. As shown in Fig. 1A
(center panel), in this panel of cell lines, elevated levels of p53 protein correlate with the presence of an inactivating mutation in the p53 gene.
|
. This probe is complementary to the 5' region of the mRNA transcribed from the hdm2-P2 promoter (Fig. 1B)
, and we were therefore able to quantitatively determine the levels of this transcript. The hdm2-P2 transcript was readily detectable in all of the ER+ve cell lines, whereas in the ER-ve cell lines, it was absent (MDAMB-231) or expressed at very low levels (levels in SKBr3 are <5% of those in MCF-7). The transcript from the hdm2-P1 promoter only partially protects the probe and therefore gives rise to a smaller fragment corresponding to part of exon 3 (Fig. 1B
, top panel). This fragment was detected in all of the breast cancer cell lines at higher levels than the P2 transcript. Hdm2-P1 transcript expression was independent of ER status, although P1 levels were 3–4-fold higher in T47D and BT474 cells than any of the other lines. It is theoretically possible that the exon 3 fragment is derived from mRNA other than the expected hdm2-P1 transcript, so semiquantitative exon-specific RT-PCR was used to confirm hdm2 transcript levels in the cell lines (Fig. 1B)
. The RT-PCR data for both the hdm2-P1 and hdm2-P2 transcripts were comparable with that obtained with the RPA, with small differences in the relative levels of the P1 transcript being attributable to the semiquantitative nature of the RT-PCR assay.
These data demonstrate that hdm2 mRNA expression in breast cancer cells can be regulated by differential expression from both the P1 and P2 promoters. It is known, however, that although the P1 mRNA transcript is expressed at higher levels than the P2 transcript in most cell types, it is translated 8-fold less efficiently (34)
and therefore contributes less to hdm2 protein expression levels. Furthermore, we found that expression of the P2 transcript in the ER+ve breast cancer cells correlated with the elevated hdm2 protein levels in these cells, whereas the 3–4-fold increased levels of the P1 transcript in T47D and BT474 compared with MCF-7 and ZR75.1 were not associated with increased hdm2 protein expression. We therefore proceeded to study in detail the mechanisms that regulate hdm2-P2 promoter activity in the breast cancer cell lines.
Of the panel of cell lines we have used, MCF-7 and ZR75.1 are representative of the most common class of breast cancers, which both express ER and retain wild-type p53. However, it is not possible to study p53-independent hdm2-P2 promoter activity in these cells by conventional reporter assays because the transfection procedure induces a DNA damage response and consequent activation of p53-responsive promoters such as hdm2-P2 (46)
. T47D cells, however, express inactive mutant p53 protein and therefore provide a good model to study p53-independent regulation of the hdm2-P2 promoter. To confirm that the mutant p53 in these cells cannot be activated by DNA damage, cells were exposed to 5 Gy 
-irradiation and levels of hdm2-P2 transcript determined 3 h later. Although a strong p53-dependent induction of hdm2-P2 transcripts occurred in MCF-7 cells, levels in T47D cells were unaffected by the radiation (data not shown).
Hdm2-P2 Promoter Activity Is Confined to ER+ve Cell Lines but Is Independent of ER Function.
Our first aim was to establish whether hdm2-P2 transcript levels correlate with ER expression because the P2 promoter is dependent on ER activity, or whether it may belong to the class of genes that are estrogen independent, but coexpressed with ER (5)
. We therefore generated a luciferase reporter construct (hdm2luc01) containing 895 bp of DNA sequence flanking the hdm2-P2 promoter, including part of exon 1 through to sequence 3' to the start ATG in exon 3 (Fig. 4A)
. As shown in Fig. 2
Ai, the activity of this promoter mirrored the cell type-specific expression of the P2-derived mRNA transcript, the reporter gene being efficiently expressed in ER+ve T47D cells but not in the ER-ve SKBr3 line. To confirm that the reporter vector was able to function in SKBr3 cells if relevant activating transcription factors were present, we performed cotransfections with wild-type p53, which binds the tandem p53-REs in the promoter, and demonstrated that the promoter was active in SKBr3 cells when wild-type p53 was expressed (Fig. 2
Aii).
|
antagonist ICI 182780 is able to effectively inhibit transcription from a bona fidae estrogen responsive promoter in T47D cells (Fig. 2B
, pERE-Tkluc), whereas we demonstrate that it has no effect on expression from the hdm2luc01 vector (Fig. 2B)
. It is known, however, that in addition to its effect on promoters containing estrogen response elements, ER is also able to enhance transcription through its ability to interact with and increase the activity of AP1 transcription factors, and this activity is not repressed by ER antagonists (47)
. Indeed, this function of ER has previously been implicated in up-regulating mdm2 expression in transformed mouse fibroblasts (38)
. We therefore generated a dominant negative mutant of human ER (ER-S554fs), which inhibits this activity of ER and also down-regulates mdm2 expression in murine fibroblasts (38)
. ER-S554fs had no effect on the activity of hdm2luc01 in T47D cells, despite blocking expression from the known ER-responsive reporter in the same cells (Fig. 2C)
. Finally we demonstrate that forced overexpression of ER in the ER-ve SKBr3 cell line has no significant effect (P > 0.05) on hdm2-P2 promoter activity, despite it being functionally active in the cells, as demonstrated by its ability to activate transcription from the bona fidae estrogen responsive reporter vector (Fig. 2D)
.
Role of ras-raf-MEK-MAPK Signaling in Regulating hdm2-P2 Promoter Activity in T47D Cells.
Previous studies have identified a role for growth factor signaling through the ras-raf-MEK-MAPK cascade in the p53-independent regulation of mdm2/hdm2 expression (36
, 37)
. The mechanism of raf-induced mdm2 expression has been dissected in the context of the murine mdm2 promoter and is dependent on the integrity of both a 5' ETS site (ETSA) and a composite ras-response element (AP1-ETSB; Ref. 36
). An alignment analysis of the hdm2 and mdm2 promoter regions (Fig. 3)
showed that although the AP1-ETSB element is conserved between species (labeled as AP1-ETSa in the human promoter), the ETSA site, which was demonstrated by Ries et al. (36)
to be necessary for raf-responsiveness, is not conserved. Consistent with this lack of conservation, we found (data not shown) that treatment of T47D cells with inhibitors of MEK (U0126 and PD98059) inhibited neither the expression of the endogenous hdm2-P2 transcript, nor transcription from the hdm2luc01 reporter vector. Expression from a murine mdm2 promoter vector was inhibited by U0126 in T47D cells. Consistent with the findings of Ries et al. (36)
, however, we did find that U0126 reduced levels of expression of hdm2 protein in human cancer cells, and we are currently examining the mechanism whereby this occurs.
|
.
Deletion mapping of the hdm2luc01 vector (Fig. 4B)
determined that 55% of the total activity was lost when the 5' region (-602 to -376) was deleted (compare hdm2luc01 with hdm2luc06, P = 0.006). Deletion of the region –375 to –133 resulted in a smaller, but significant reduction in activity (compare hdm2luc02 with hdm2luc03, P = 0.001), whereas the reduction in activity observed by the deletion of sequences 3' of +33 did not reach the level of statistical significance (compare hdm2luc06 with hdm2luc02, P = 0.063). Therefore, although it is clear that multiple sequence elements are required for the full activity of the hdm2-P2 promoter, at least one positive acting element is present between -602 and -376. An additional deletion mutant (hdm2luc23) located this element to the sequence between -418 and -376 (Fig. 4C
, compare hdm2luc23 with hdm2luc06, P = 0.006). This 42-bp sequence consists almost entirely of a (nnGGGGC)5 repeat sequence (-416 to -381). A potential EBOX (CACGTG) is also present (-381 to -376), however, destruction of this element by an internal 3-bp deletion had no effect on promoter activity in either the hdm2luc01 (Fig. 4C
, hdm2luc01EBOX) or hdm2luc23 (data not shown) backgrounds, and therefore the positive-acting element is contained within the (nnGGGGC)5 repeat.
The -375 to +293 region of the promoter (hdm2luc06 vector) retains 35–55% activity, and therefore we used a candidate site approach to identify the positive acting elements in this region. On the basis of studies of the murine P2 promoter (36)
, three separate mutations were made in the hdm2luc01 vector, either singly or in combination, to inactivate potential transcription factor binding sites (Fig. 4A)
. The mutated sites were: (a) the conserved AP1 site (2-bp substitution); (b) the adjacent ETS response element (ETSB in mouse, ETSa in human, Fig. 3
; 2-bp substitution), which together form an AP1-ETS ras response element in the murine promoter; and (c) the nonconserved (ETSb/c) sequence containing two potential ETS binding sites (10-bp deletion). As shown in Fig. 4D
, deletion of the ETSb/c element had no effect on promoter activity in T47D cells, whereas mutation of the AP1 site resulted in a 52.5% loss of promoter activity (hdm2luc01 versus hdm2luc01AP1, P < 0.001), and the ETSa site deletion resulted in a loss of 29.6% of activity (hdm2luc01 versus hdm2luc01ETSa, P = 0.003). A vector containing mutations in both AP1 and ETSa elements had the same activity as the AP1 site mutant (46.6 versus 47.5%).
Having determined that the AP1-ETSa site in the hdm2-P2 promoter confers approximately one-half of its p53-independent promoter activity in T47D cells, we were then able to determine to what extent transcriptional activation through this element contributes to the level of hdm2-P2 promoter activity in the MCF-7 cell line (Fig. 4D)
. Although overall hdm2luc01 activity was 40-fold higher in MCF-7 cells than T47D (in Fig. 4D
, promoter activity is presented as a percentage of hdm2luc01 activity in each cell line), due, in part, to the activation of p53 by the transfection process (data not shown), inactivation of the AP1-ETSa element also reduced promoter activity to 50% in MCF-7 cells.
ETS2 Overexpression Reconstitutes hdm2-P2 Promoter Activity in SKBr3 Cells through the Same Elements that Drive Constitutive hdm2-P2 Expression in T47D Cells.
Having identified the cis-acting elements necessary for hdm2-P2 promoter activity in ER+ve T47D and MCF-7 cells, we then wished to determine why the promoter is inactive in the ER-ve cell lines. Expression from the murine P2 promoter is known to be dependent on the levels of transcriptional activation through the AP1-ETS element and its activity can be induced by the transient overexpression of AP1 or ETS transcription factors (36)
. We therefore transfected SKBr3 cells with an expression vector encoding ETS2 to determine whether the lack of hdm2-P2 transcription in these cells is attributable to limiting activity of the AP1-ETS binding transcription factors. ETS2 is known to be able to activate transcription of the murine mdm2-P2 promoter when overexpressed in fibroblasts (36)
and is normally expressed in all of the breast cancer cell lines used in this study at similar levels (Fig. 5A
; Ref. 48
). Transfection of SKBr3 cells with the ETS2 expression vector resulted in an elevation in ETS2 protein expression levels (Fig. 5A)
and a significant (P = 0.007) 5-fold activation of the hdm2luc01 reporter vector (Fig. 5B)
. ETS2 also activated the hdm2luc01 reporter vector in the other ER-ve cell line, MDAMB-231 (data not shown). Hdm2-P2 activation by ETS2 in SKBr3 cells was dependent of the integrity of the AP1-ETS element in the promoter as there was no significant activation of hdm2luc01AP1ETSa by ETS2 (P = 0.073; Fig. 5B
). Consistent with the requirements for basal hdm2-P2 promoter activity in T47D cells, effective ETS2 activation of the promoter in SKBr3 cells was also dependent on the GC-rich repeat sequences present in hdm2luc01 and -23 but not -06 reporter vectors (Fig. 5B)
.
|
-ve cells was capable of inducing the expression of functional hdm2 protein. SKBr3 cells were therefore transfected with increasing amounts (0–1 µg) of pRKETS2 plasmid, and levels of endogenous hdm2 and p53 proteins determined by Western blotting (Fig. 5C)
. ETS2 transfection resulted in a clear increase in hdm2 protein levels and a concomitant decrease in the levels of mutant p53 protein, consistent with increased rates of hdm2-dependent degradation of p53 by the proteosome. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
+ve status in a panel of six breast cancer cell lines. We have therefore used these cell lines to investigate the mechanisms whereby hdm2 expression is regulated in breast cancer cells. We firstly determined that high hdm2 protein expression correlated with increased levels of transcription from the hdm2-P2 promoter and identified T47D as a cell line in which the p53-inducible hdm2-P2 promoter is constitutively active, despite p53 being expressed in a functionally compromised, mutant form in the cells. We show that this activity is dependent on at least three different predicted transcription factor binding sites: an AP1 site; an ETS site, which together form a bi-partite AP1-ETS element at -120 to -99; and a series of five consecutive nnGGGGC repeats at -415 to -381. We also demonstrate that transcriptional activation by the AP1-ETS element drives expression of the hdm2-P2 promoter in the ER+ve cell line MCF-7. These cells express wild-type, functional p53 protein and are widely used as a model system that is representative of the majority of human breast cancers, which are both ER+ve and p53 wild type.
In contrast, however, the hdm2-P2 transcript was not efficiently expressed in the ER-ve cell lines SKBr3 or MDAMB-231. Several lines of evidence demonstrate that this lack of expression is attributable to a loss of transcriptional activation through the AP1-ETS element in these cell lines: (a) in contrast to results from T47D cells, mutation of the AP1-ETS element has no effect on the low basal activity of the hdm2-P2 promoter in SKBr3 cells, demonstrating that there is no activation of transcription through this element in SKBr3; (b) removal of the nnGGGGC repeats does result in a loss of the residual promoter activity in SKBr3 cells (Fig. 5B
, hdm2luc23 versus hdm2luc06), demonstrating that the mechanism whereby promoter activity is stimulated through these repeats is functional in SKBr3; and (c) the overexpression of an ETS factor, ETS2, is able to reconstitute hdm2-P2 promoter activity in SKBr3 cells in an AP1-ETS element-dependent manner. As was found to be the case for the constitutive hdm2-P2 promoter activity in T47D cells, maximal ETS2-induced P2 activity in SKBr3 cells was dependent on the integrity of both the AP1-ETS and nnGGGGC repeat sequence. The precise mechanism whereby this nnGGGGC element contributes to transcription remains to be elucidated. Although there is no clearly identifiable transcription factor binding element in the sequence, it bears similarity to the consensus for factors such as SP1, which binds a direct repeat of GGGGC without a spacer (49)
. Both SV40 LT and PyLT viral proteins direct transcription through GGGGC direct repeats similar to the one in the hdm2-P2 promoter and can inhibit the expression of cellular genes by displacing factors such as the recently identified Rnf6 from such elements (50)
.
Although overexpression of ETS2 is able to reconstitute hdm2-P2 promoter activity in SKBr3 cells, we show that the lack of P2 activity in these cells is unlikely to be simply as consequence of lack of ETS2 protein expression because ETS2 is expressed at similar levels in all of the breast cancer cell lines examined. Activation of transcription through composite AP1-ETS elements normally involves the cooperative activation by both AP1 and ETS family members. Not only are these families comprised of multiple members with differing activities (>20 in the case of ETS; Ref. 51
), many of which can form heterodimers with other family members, but both protein families are subject to control by posttranslational modification and the interaction with both positive and negative regulatory factors (52, 53, 54)
. We have considered pathways that have been previously shown to regulate mdm2-P2 promoter activity to determine whether they may account for the activity of the AP1-ETS element in the hdm2-P2 promoter in ER+ve but not ER-ve cells. However, the activity in ER+ve cells is neither because of the stimulation of AP1 factors by interaction with ER (38)
or the activation of AP1 or ETS factors by ras-raf-MEK signaling (36)
. Additionally, using electrophoretic mobility shift analysis with the AP1-ETS sequence as a probe, we detect specific binding complexes of similar mobilities in the nuclear lysates of both T47D and SKBr3 cells (data not shown). The expression of a large number of genes is known to be differentially regulated between ER+ve and ER-ve breast tumors (6)
, and both AP1 (data not shown; Ref. 55
) and ETS factor (56)
expression shows considerable variation between different breast cancer cells. On the basis of our cell line study and a recent analysis of breast tumor samples (27)
, the hdm2-P2 promoter may be included in the group of genes in which the expression is coordinately regulated with ER in breast cancers.
During the course of our investigations, we have also identified striking differences between human and murine hdm2/mdm2 in the ability of the ras-raf-MEK-MAPK signaling cascade to regulate P2 promoter activity. This finding is critical to our understanding of how hdm2 expression is regulated in many normal and malignant human cells. Specifically, although mdm2 mRNA and protein expression can be induced by this signaling pathway in murine fibroblasts (36) , and in human cancer cells both hdm2 protein expression and transcription from the murine mdm2-P2 promoter reporter vector are sensitive to inhibition of MEK activity, transcription from the human hdm2-P2 promoter is not MEK dependent in the human cancer cell lines. This species-specific difference can be explained by a lack of conservation of the transcription factor binding site in the two promoters because a single ETS factor binding site (ETSA) that is required for ras-raf- MEK-MAPK induction of the murine promoter (36) is not present in the human, and instead, constitutive activity of the human promoter in T47D cells is dependent on the nnGGGGCC repeat element. Ras-raf-MEK-MAPK signaling must therefore use a hdm2-P2 promoter-independent mechanism to regulate hdm2 protein levels in human cells, as has recently been described for the inhibition of hdm2 expression by hypoxia (57) .
Our study has a number of important implications regarding how levels of p53 activity may be regulated during the development of breast and, most probably, other malignancies. Firstly, inactivation of p53 by its exclusion from the nucleus, which occurs in 30% of breast cancers and also in normal breast tissue (15)
, is known to be dependent on the expression of hdm2 (58)
, and it is therefore probable that the p53-independent transcription of hdm2 we have described will promote the nuclear exclusion of wild-type p53 in ER+ve breast cancer cells. Secondly, the widely reported correlation between the presence of inactivating mutations of p53 and elevated p53 protein levels is a consequence of the inactive p53 mutant protein no longer driving the expression of the hdm2 protein required to target their own degradation (59
, 60)
. The loss of this correlation, which is observed in significant numbers of breast cancers (8
, 9)
, is likely to be due, in part, to p53-independent expression of hdm2 in a proportion of these tumors.
Finally, our demonstration that the elevated levels of hdm2 expression in ER+ve breast cancer cells are not merely a consequence of transcriptional activation by wild-type p53 confirms the presence of a p53-independent mechanism whereby the p53-hdm2 negative-regulatory feedback loop is modified in these cells to reduce levels of cellular p53 activity. These data clearly strengthen the previously uninvestigated hypothesis (14)
that differences between cancer cells such as cell lineage- or differentiation-dependent activity of transcription factors such as AP1 or ETS may determine the frequency at which p53 mutations are observed in individual tumor types through the regulation of hdm2 expression levels.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the Association for International Cancer Research Grant 01-070 and a grant from the Wessex Cancer Trust. 
2 To whom requests for reprints should be addressed, at Cancer Sciences Division, School of Medicine, University of Southampton, MP 824, Southampton General Hospital, Southampton SO16 6YD, United Kingdom. Phone: 44-0-23-8079-4582; Fax: 44-0-23-8079-5152; E-mail: j.p.blaydes{at}soton.ac.uk
3 The abbreviations used are: ER estrogen receptor AP1-ETS, composite binding site for AP1 and ETS family transcription factors; p53-RE, p53 response elements; MEK, mitogen-activated protein/extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase; RT-PCR, reverse transcription-PCR; RPA, ribonuclease protection assay; mAb, monoclonal antibody; RLU, relative luciferase unit.
4 Internet address: www.genomatix.de/cgi-bin/matinspector/matinspector.pl.
Received 10/29/02. Accepted 3/19/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
to oncogenic K-Ras-mediated NIH3T3 cell transformation and its implication for escape from senescence by modulating the p53 pathway. J. Biol. Chem., 277: 11217-11224, 2002. and ERß at AP1 sites. Science (Wash. DC), 277: 1508-1510, 1997.This article has been cited by other articles:
|
|
 |
|
  S. Pikkarainen, R. A. Kennedy, A. K. Marshall, E. L. Tham, K. Lay, T. A. Kriz, B. S. Handa, A. Clerk, and P. H. Sugden Regulation of Expression of the Rat Orthologue of Mouse Double Minute 2 (MDM2) by H2O2-induced Oxidative Stress in Neonatal Rat Cardiac Myocytes J. Biol. Chem., October 2, 2009; 284(40): 27195 - 27210. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-C. Ku, S.-K. Chiu, Y.-J. Chen, H.-H. Huang, W.-G. Wu, and Y.-J. Chen Complementary Quantitative Proteomics Reveals that Transcription Factor AP-4 Mediates E-box-dependent Complex Formation for Transcriptional Repression of HDM2 Mol. Cell. Proteomics, September 1, 2009; 8(9): 2034 - 2050. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Toffoli, P. Biason, A. Russo, E. De Mattia, E. Cecchin, C. M. Hattinger, M. Pasello, M. Alberghini, C. Ferrari, K. Scotlandi, et al. Effect of TP53 Arg72Pro and MDM2 SNP309 Polymorphisms on the Risk of High-Grade Osteosarcoma Development and Survival Clin. Cancer Res., May 15, 2009; 15(10): 3550 - 3556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wade and G. M. Wahl Targeting Mdm2 and Mdmx in Cancer Therapy: Better Living through Medicinal Chemistry? Mol. Cancer Res., January 1, 2009; 7(1): 1 - 11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Bartel, J. Jung, A. Bohnke, E. Gradhand, K. Zeng, C. Thomssen, and S. Hauptmann Both Germ Line and Somatic Genetics of the p53 Pathway Affect Ovarian Cancer Incidence and Survival Clin. Cancer Res., January 1, 2008; 14(1): 89 - 96. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Hu, G. Jin, L. Wang, F. Chen, X. Wang, and H. Shen MDM2 Promoter Polymorphism SNP309 Contributes to Tumor Susceptibility: Evidence from 21 Case-Control Studies Cancer Epidemiol. Biomarkers Prev., December 1, 2007; 16(12): 2717 - 2723. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Hu, Z. Feng, L. Ma, J. Wagner, J. J. Rice, G. Stolovitzky, and A. J. Levine A Single Nucleotide Polymorphism in the MDM2 Gene Disrupts the Oscillation of p53 and MDM2 Levels in Cells Cancer Res., March 15, 2007; 67(6): 2757 - 2765. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Li, Z. Zhang, D. L. Hill, H. Wang, and R. Zhang Curcumin, a Dietary Component, Has Anticancer, Chemosensitization, and Radiosensitization Effects by Down-regulating the MDM2 Oncogene through the PI3K/mTOR/ETS2 Pathway Cancer Res., March 1, 2007; 67(5): 1988 - 1996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. W. Jackson, L. E. Patt, G. A. LaRusch, D. B. Donner, G. R. Stark, and L. D. Mayo Hdm2 Nuclear Export, Regulated by Insulin-like Growth Factor-I/MAPK/p90Rsk Signaling, Mediates the Transformation of Human Cells J. Biol. Chem., June 16, 2006; 281(24): 16814 - 16820. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lacroix, R.-A. Toillon, and G. Leclercq p53 and breast cancer, an update. Endocr. Relat. Cancer, June 1, 2006; 13(2): 293 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L. Bond, K. M. Hirshfield, T. Kirchhoff, G. Alexe, E. E. Bond, H. Robins, F. Bartel, H. Taubert, P. Wuerl, W. Hait, et al. MDM2 SNP309 Accelerates Tumor Formation in a Gender-Specific and Hormone-Dependent Manner. Cancer Res., May 15, 2006; 66(10): 5104 - 5110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Babbage, C. H. Ottensmeier, J. Blaydes, F. K. Stevenson, and S. S. Sahota Immunoglobulin heavy chain locus events and expression of activation-induced cytidine deaminase in epithelial breast cancer cell lines. Cancer Res., April 15, 2006; 66(8): 3996 - 4000. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Kashatus, P. Cogswell, and A. S. Baldwin Expression of the Bcl-3 proto-oncogene suppresses p53 activation Genes & Dev., January 15, 2006; 20(2): 225 - 235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Yeretssian, M. Lecocq, G. Lebon, H. C. Hurst, and V. Sakanyan Competition on Nitrocellulose-immobilized Antibody Arrays: From Bacterial Protein Binding Assay to Protein Profiling in Breast Cancer Cells Mol. Cell. Proteomics, May 1, 2005; 4(5): 605 - 617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Phelps, A. Phillips, M. Darley, and J. P. Blaydes MEK-ERK Signaling Controls Hdm2 Oncoprotein Expression by Regulating hdm2 mRNA Export to the Cytoplasm J. Biol. Chem., April 29, 2005; 280(17): 16651 - 16658. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Göhler, S. Jäger, G. Warnecke, H. Yasuda, E. Kim, and W. Deppert Mutant p53 proteins bind DNA in a DNA structure-selective mode Nucleic Acids Res., February 18, 2005; 33(3): 1087 - 1100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Slack, Z. Chen, R. Tonelli, M. Pule, L. Hunt, A. Pession, and J. M. Shohet The p53 regulatory gene MDM2 is a direct transcriptional target of MYCN in neuroblastoma PNAS, January 18, 2005; 102(3): 731 - 736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Paajarvi, M. Viluksela, R. Pohjanvirta, U. Stenius, and J. Hogberg TCDD activates Mdm2 and attenuates the p53 response to DNA damaging agents Carcinogenesis, January 1, 2005; 26(1): 201 - 208. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Pohnke, T. Schneider-Merck, J. Fahnenstich, R. Kempf, M. Christian, K. Milde-Langosch, J. J. Brosens, and B. Gellersen Wild-Type p53 Protein Is Up-Regulated upon Cyclic Adenosine Monophosphate-Induced Differentiation of Human Endometrial Stromal Cells J. Clin. Endocrinol. Metab., October 1, 2004; 89(10): 5233 - 5244. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Onel and C. Cordon-Cardo MDM2 and Prognosis Mol. Cancer Res., January 1, 2004; 2(1): 1 - 8. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Ganguli and B. Wasylyk p53-Independent Functions of MDM2 Mol. Cancer Res., December 1, 2003; 1(14): 1027 - 1035. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
indicates the p53-REs. The deletion series of hdm2 reporter vectors is shown relative to the full construct hdm2luc01. Potential transcription factor response elements where mutations were introduced are indicated. The sequence of the 42-bp region at the 5' end of luc23 (difference between luc06 and luc23) contains 5 GC-rich repeats adjacent to an EBOX site (italicized) as follows: GC(TGGGGGC)(TCGGGGC)(GCGGGGC)(GCGGGGC)(ATGGGGC)ACGTG. The (nnGGGGC)5 repeats are represented by a solid triangle on the diagram. B, T47D cells were transfected with hdm2luc01 reporter construct and the hdm2luc06, hdm2luc02, and hdm2luc03 deletion constructs as indicated. Results are expressed as a percentage of hdm2luc01 activity (mean ± SE), with data pooled from five independent experiments. C, T47D cells were transfected with the reporter constructs indicated. D, T47D (
represents the results from a separate experiment in which MCF-7 cells were transfected with hdm2luc01 or hdm2luc01
