modulation of aromatase expression in the breast tissue by ... · an oligonucleotide and its...

[CANCER RESEARCH 58. 5695-57IX). December 15, 1998]

Advances in Brief

Modulation of Aromatase Expression in the Breast Tissue by ERRa-1 OrphanReceptor1

Chun Yang,2 Dujin Zhou,2 and Shiuan Chen3

Division of Immunology. Beckman Research Inviane of Â¡heCity of Hope, Duane. California 91010

Abstract

We have previously identified a silencer element (SI) that is situatedbetween promoters 1.3 and II of the human aromatase gene and thatdown-regulates the action of these promoters. We recently applied theyeast one-hybrid approach to screen a human breast tissue hybrid cDNA

expression library for genes encoding the proteins binding to the silencerregion. Most proteins identified from this approach belong to the nuclearreceptor superfamily. Fifty % of the positive clones encode for KKKit-1,and other positive clones include EAR-2, EAR-3 (COUP-TF1), RARy, andpl20E4F. Because ERRa-1 was found to be the major protein interactingwith SI, we decided to examine the regulatory action of ERRa-1 on

promoter 1.3 of the human aromatase gene. Using a reporter plasmid thatincludes the aromatase genomic fragment containing promoter 1.3 and SI,ERRa-1 was found to have a positive regulatory function in breast cancerSK-BR-3 cells. Gel mobility shift assays have confirmed that ERRa-1binds to SI in a dose-dependent manner, and DNase I footprinting analysis has revealed that ERRa-1 binds to a region, 5'-AAGGTCAGAAAT-3', which is within SI and between 96 and 107 bp relative to the tran-

scriptional start site of promoter 1.3. In addition, despite the fact that thenuclear receptor SF1 was shown previously to bind to the same site and tomediate a i A Ml" response in ovary, our yeast one-hybrid screening did not

find any SI -1 clones. Gel mobility shift assays further revealed that SF-1can bind to the silencer element with an affinity comparable with ERRa-1.Because our reverse transcription-PCR analysis was not able to detect SF1mRNA in breast cancer tissue or in SK-BR-3 cells, it is thought that SF1protein is not expressed in breast cancer tissue. Two ERRa-1 RNAvariants with differences at the 5'-end have been reported. Our reverse

transcription-PCR analysis identified the shorter variant in 28 of 32

breast tumor specimens and the longer variant in only 1 specimen. Inaddition, the shorter variant was detected in breast cancer SK-BR-3 cells

as well as in a breast tumor fibroblast line WS3TF. The results suggestthat ERRa-1 is one of the nuclear proteins interacting with SI in breast

cancer tissue. It is thought that the silencer element in the human aromatase gene may function differently in different tissues because of distinct expression patterns of transcription factors.

Introduction

Aromatase converts androgen to estrogen. The control of humanaromatase gene expression is complex in that several promoters directaromatase gene expression in a tissue specific manner (1-5). Thisconclusion is based on the fact that there are multiple tissue-specific

exons 1, exons 1.1, 1.2, 1.4, 1.5, 1.3, and pii, existing in the mRNAsisolated from placenta, placenta, adipose tissue, skin fibroblasts orfetal liver, adipose tissue, and ovary, respectively. Aromatase mRNA

Received 9/23/98: accepted 10/29/98.The costs of publication of Ihis article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

' This research was supported by NIH Grant CA44735 and University of California

BCRP Grant 1RB0118. S. C. and D. Z. are members of the City of Hope Breast CancerProgram (Grant CA 65767). C. Y. is a predoctoral student of the Third Military MedicalUniversity. Chongquing. China.

2 These authors contributed equally to this study.^ To whom requests for reprints should be addressed, at City of Hope Beckman

Research Institute. 1450 East Duane Road, Duarte, CA 91010-0269. Phone: (626)359-8111, extension 2601: Fax: (626) 301-8186: E-mail: [email protected].

in the noncancerous breast regions is exon 1.4-dominant, indicating

aromatase expression in the normal breast tissue is driven by promoter1.4. which has been shown to be glucocortcoid dependent (6). Ourprevious studies on 70 breast tumor specimens using RT-PCR4 tech

nique has revealed that exons 1.3 and pii are the two major exon Ispresent in aromatase mRNAs isolated from breast tumors. Theseresults suggest that promoters 1.3 and II are the major promotersdirecting aromatase expression in breast cancer and surrounding adipose stromal cells (7). This switching of promoter usage, frompromoter 1.4 in normal tissue to promoters 1.3 and II in breast cancertissue, was also shown in a study on 49 Japanese breast cancersamples (8) and a study on 18 breast cancer patients (9).

Our laboratory has identified a silencer element (S l ) that is situatedbetween promoters 1.3 and II and that down-regulates the action ofthese two promoters (10, 11). In addition, recently we found a cAMP-

responsive element (CREaro) that is positioned upstream from promoter 1.3. CREaro was shown to be able to overcome the negativeregulation of S l ,5 On the basis of results generated from our and other

laboratories, we propose that in normal breast adipose stromal cellsand fibroblasts, aromatase expression is driven by promoter 1.4, andthe action of promoters 1.3 and II is suppressed by S l. However, incancer cells and surrounding adipose stromal cells, the level of cAMPincreases, and aromatase promoters are switched to cAMP-dependent

promoters, i.e., 1.3 and II. These studies indicate that SI plays animportant role in regulating aromatase expression in breast tissue.Recently, we used a yeast one-hybrid screening method to search fortrans factors in breast tissue that interact with SI. ERRa-1 orphan

receptor was found to be the major protein binding to this cis element.This report shows the characterization of the interaction of ERRa-1

with SI in the human aromatase gene.

Materials and Methods

Materials. The MATCHMAKER One-Hybrid System kit including a

human mammary tissue MATCHMAKER cDNA library was purchased fromClontech (Palo Alto, CA). DNA sequencing kits were from United StatesBiochemical (Cleveland. OH). AMV reverse transcriptase. T4 kinase, T4 DNAligase, and various restriction endonucleases were purchased from New England Biolabs (Beverly. MA) and Boehringer Mannheim (Indianapolis, IN).AmpiTag polymerase was obtained from Perkin Elmer (Norwalk, CT).[14C]Chloramphenicol (D-threo-|dichloroacetyl-l-l4Clchloramphenicol: spe

cific radioactivity, 55 mCi/mmol) was from Amersham Life Science, Inc.(Arlington Heights, IL). The CAT expression vector. pUMSVOCAT, was agift from Dr. K. Kurachi at the University of Michigan (Ann Arbor, MI).Oligonucleotide primers were synthesized in the DNA/RNA chemistry laboratory at the City of Hope. SK-BR-3 cells from American Type Culture

Collection (Rockville. MD). derived from a human breast adenocarcinoma,were maintained in McCoy's 5A medium containing 10% fetal calf serum and

4 The abbreviations used are: RT-PCR. reverse transcription-PCR; CAT. chloramphen-

icol acetyltransferase; nt. nucleotide(s): AMV, avian myeloblastosis virus: pofy(dl-dC).poly(deoxyinosinic-deoxycytidylic acid); ER, estrogen receptor.

5 D. Zhou and S. Chen. Identification and characterization of a cAMP-responsive

element in the region upstream from promoter 1.3 of the human aromatase gene, submittedfor publication.

5695

Research. on September 14, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

MODULATION OF AROMATASE EXPRESSION IN BREAST TISSUE

glutaniine. WS3TF, a tumor fibroblust line derived from breast cancer tissue,was the gift of Dr. R. J. Santen at University of Virginia Health Science Center(Charlottesville, VA) and maintained in Waymouth's MB 752/1 medium with

15% fetal calf serum.Yeast One-Hybrid Screening. An oligonucleotide and its complimentry

strand containing three tandem copies of SI with an EcoRl site at the 5'-endand a Xho\ or Xba\ site at the 3'-ends were synthesized, annealed, and

subcloned into both His reporter. pHISi. and LacZ reporter, pLacZi plasmids.The sequence of the sense strand was: 5'-AATTCACCAAGGTCAGAAAT-

GCTGCAATTCAAGCCAACCAAGGTCAGAAATGCTGCAATTCAAG-CCAACCAAGGTCAGAAATGCTGCAATTCAAGCCAT-3'. The resulting

plasmids. pHISi-3Sl and pLacZi-3Sl. were linearized and used to transform

the YM4271 yeast strain. A dual reporter strain was used to screen the premadelibrary using the procedure provided by Clontech. The fusion protein-express

ing clones binding to the SI elements were isolated. A nucleotide sequence of2(X)-3(X) bp was generated for each positive, and the identity of the positives

was determined through a homology search against known sequences inGenBank. The total nucleolide sequence of the clone with the longest insert foreach protein was then determined and compared with the published sequences.

RT-PCR Analysis. RNA was isolated from human breast cancer tissues

according to a published procedure by Chirgwin et al. (12). Briefly, a block oftissue ( 1 cm') was placed in a mortar tilled with liquid nitrogen and ground

into a powder, and the powder was resuspended in 3.2 ml of a guanidineisothiocyanate (GT) buffer |1(X) mM sodium acetate (pH 5.0), 4 M GT, and 5niM EDTA|. The sample containing GT solutions were layered over 1.8 ml of5 M CsCl in SW-50TÃŒultracentrifuge tubes and centrifuged overnight in aSW-50TÃŒrotor at 35.000 rpm at 20Â°C.The supernatant was carefully removed,

and the bottom of the tube was cut off. The RNA pellet was rinsed three timeswith 70% ethanol and resuspended in 10 mM Tris-HCI, pH 8.0, 1 mM EDTA

buffer. Total RNA from cultured cells was extracted using a procedure described previously (7).

RT-PCR reactions were carried out in a total volume of 50 jil containing 20

min Tris HC1, (pH 8.3), 50 mM KC1, 15 mM MgCI,. 0.01% gelatin, 200 /AMdeoxynucleotide triphosphates, 20 pmol of each primer, 1.5 /xg of total RNA,2.5 units of Taq DNA polymerase. and 2 units of AMV reverse transcriptase.The reverse transcription reaction was carried out at 37Â°Cfor 7 min, and PCRwas performed for 25 cycles using the following temperature profile: 51Â°C,2min (primer annealing): 72Â°C.2 min (primer extension); and 94Â°C. 1 min(denaturation). An additional extension cycle was performed for 8 min at 72Â°Cbefore cooling the reaction mixture to 4Â°C. After separation of the PCR

product on agarose gel (1.5%), the DNA products were transferred to Zeta-probe membranes (Bio-Rad), followed by hybridization using a 12P-labeled

probe derived from the hERRal cDNA sequence. The conditions for hybridization are according to the Bio-Rad instruction manual.

Our ERRa-1 clone is identical to that reported by Yang et al. (13) except fora 3-base insertion (described further in "Results"). The ERRa-1 sequencereported by Yang et al. ( 13) is missing 177 bp from the 5'-end when comparing

to the sequence reported by Giguere et al. (14). In describing the relativepositions of different primers for RT-PCR, the nucleotides are numbered based

on the sequence reported by Yang el al. (13) with the additional sequence ofGiguere el al. (14) at the 5'-end indicated in negative numbers (see Fig. 1).

There are two potential translational start sites at â€”114 and 177 nt. RT-PCR

for detecting the RNA message containing the potential translational start siteat 177 nt was performed using primer a (5'-AGGTGACCAGCGCCATGTC-CAGCCAGGTGGTGGGCATTGAG-3') and a reverse primer d (5'-GGAG-GCAGCGAGTGGGAGCTGCTGGAC-3'). The expected product is 169-bp

long. RT-PCR for detecting the RNA message containing the potential translational start site at â€”114 nt was performed using primer c (5'-GACGAAT-

TCATTGCCATGGGATTGGAGATG-3'), which is located in exon I, and thereverse primer d. The probe for hybridizing has the sequence 5'-AGCCCTG-GCAGTCTGGATGGAAGAGCTTGGGAA-3'. The expected product of RT-

PCR generated with primers c and d is 450-bp long.RT-PCR analyses of SF-1 messages in breast cancer tissues were also

performed using the same set of RNA samples for ERRa-1 message measurement. We are aware that SF-I and ELP (embryonic long terminal repeat-

binding protein) transcripts arise from a single structural gene, and theircDNAs are virtually identical for 1017 bp but diverge at their 5'- and 3'-ends

(15). Therefore, the primers for RT-PCR Southern analysis were derived fromthe SF-1-specific region at the 3'-end. The sequence of the forward primer is

S'-ACAGCCTGGTCCTGCGGGCC-S', and the sequence of the reverseprimer is S'-CGGAATTCTCAAGCCTGCTTGGCCTGCAG-S'. The probefor hybridization has the sequence 5'-TACCCACACTGCGGGGACAA-3'.

Precautions were taken to make sure that our RNA preparations did notcontain DNA. The RNA preparations were treated with RNase-free DNase. In

addition, a control such as PCR analysis with RNA, without treating withreverse transcriptase, was performed to assure ourselves that the PCR productswere derived from ERRa-1 mRNA or SF-1 mRNA.

In Vitro Transcription and Translation. A 1269-bp ERRa-1 cDNA

fragment [from 177 nt start codon to 1446 nt stop codon as shown by Yang elal. (13)] and a 1569-bp ERRa-1 cDNA fragment [from - 114 nt start codon to

the stop codon according to the ERR-1 sequence published by Giguere et al.

(14)1 were generated by PCR using primers a+b and primers c+b, respec

tively. These cDNA fragments were subcloned into the pSG5 expressionplasmid at the EcoRl site. The entire coding region of bovine SF-1 was also

generated by PCR and subcloned into expression vector pSG5 at the EcoRTsite. The correct orientations and sequences of ERRa-1 or SF-1 inserts in the

vector were confirmed by both restriction digestion and direct DNA sequencing. Human ERRa-1 and bovine SF-1 proteins were synthesized in vitro usingthe TNT-coupled reticulocyte lysate system (Promega, Madison, WI) withT7-RNA polymerase according to the manufacturer's instructions. The reac

tions were carried out for 90 min at 30Â°C. The relative amounts of thetranslated ERRa-1 and SF-1 proteins were quantified by [35S] methionine

(1000 Ci/mmol from Amersham Life Science, Inc.) incorporated into theexpressed proteins that were separated on 10% SDS-PAGE. The expressed

proteins were visualized on the dried gel after exposure to Kodak BioMax film(Eastman Kodak Co, Rochester, NY) overnight. The translation reaction mixtures (50-ju.l portions) were snap-frozen and stored at â€”70Â°Cuntil further use.

ERRa-1 protein was only detected in the reaction mixture using the expressionplasmid containing the 1269-bp cDNA. The bovine SF-1 cDNA clone was

kindly provided by Dr. Keith L. Parker (Duke University Medical Center,Durham, NC).

DNA Mobility Shift Analysis. The double strand SI oligonucleotide wasend-labeled with [-y-12P]ATP using T4 kinase and used as a probe in the

mobility shift assay. Mobility shift assays were done as described by Singh etal. (16). Briefly, 3 iÂ¿\of in vitro expressed ERRa-1 protein or SF-1 protein wasincubated with 6000 cpm of 32P-labeled probe at room temperature for 30 min

in a mixture (15 n\) containing 10 mM Tris-HCI (pH 7.5), 50 mM NaCl, 1 mM

MgCU 0.5 mM EDTA, 0.5 mM DTT, 4% (v/v) glycerol, and 0.1 mg/mlpoly(dl-dC). The dose-dependent binding assays of ERRa-1 and SF-1 were

performed with the same amount of in vitro translated proteins (adjusted byPhosphorlmager screening) and different concentrations of SI. The reactionmixture was electrophoresed on 6% acrylamide gels, which were then dried

and autoradiographed. For competition experiments, the conditions used forbinding of the ERRa-1 protein to each probe were the same as those described

above, except that the appropriate amount of the unlabeled DNA fragments andpoly(dl-dC) were supplemented in the binding reaction mixture as specific and

nonspecific competitors. The nucleotide sequences of the oligonucleotidesused in this analysis were (only sense strands are shown): SI, 5'-ACCAAG-GTCAGAAATGCTGCAATTCAAGCCAA-3'; 5'-half of SI. 5'-ACCAAG-GTCAGAAATGC-3'; 3'-half of SI, 5'-TGCAATTCAAGCC-3'; SF1,5'-CAAGGTCA-3'; and Bl, S'-AGCTTCTTATAATTTGGCAAGAAATTT-GGCTTTA-3'. All the competitors were double-stranded except S1-5', whichis the sense strand of the 5'-half of S l.

Expression and Purification of hERRa-1 Protein from Escherichia coli.A 1269-bp ERRa-1 cDNA [from 177 nt start codon to 1446 nt stop codon asshown by Yang et al. (13)] was ligated into the expression vector pET-28a at

the Â£cY>RIrestriction site. The constructed plasmid was sequenced to confirmthat the cDNA was ligated in-frame. This expression recombinant was designated as pET-ERRa-1 and used to transform BL2[DE3 cells. Transformedcells were cultured in LB/Kanamycin medium and induced by isopropyl-1-thio-ÃŸ-D-galactopyranoside (at the final concentration of 2 mM) for 2.5 h. TheERRa-1 carrying a (His)6-Tag sequence was prepared by a batch method

according to the pET system manual (Novagen, Inc. Madison, WI), and theeluted proteins were analyzed on 10% SDS-PAGE. The expressed proteins

were confirmed by protein sequencing using mass spectrometry. The proteinconcentration was determined by Bio-Rad kit. and the protein preparation wasaliquoted and stored at -70Â°C until use.

5696




Footprinting Analysis. A 280-bp genomic fragment containing SI region

was chosen as the probe for the footprinting experiment. The probe was labeledby EcoRl digestion of pUC19-280bp. followed by 5' labeling of the coding

strand with (-y-^PJATP and T4 kinase. It was then excised with Xbcil, sepa

rated on agarose gel. purified with a prep-A-gene kit. and precipitated.

The binding reaction conditions were the same as those described for theDNA mobility shift analysis. The appropriate amount of expressed ERRÂ«-1protein was incubated with 10,(KK)cpm of <2P-labeled probes at room tem

perature for 30 min in 15 ^tl of reaction mixture containing 10 mM Tris-HCI

(pH 7.5). 50 mM NaCl. 1 mM MgCK, 0.5 mM EDTA. 0.5 mM DTT, 4%glycerol. and 0.1 mg/ml poly(dl-dC). The appropriate amount of DNase I in 20

/Â¿Iof Tris buffer (10 mM. pH 7.5) containing CaCl, (5 mM) and MgCl2 ( 10 mM)was then added, and the incubations were continued for the indicated time at

room temperature. The action of DNase I was then stopped by adding Trisbuffer (10 min. pH 7.5) containing 100 mM NaCl. 20 mM EDTA, 1% SDS. and50 /j.g/ml yeast tRNA. The reaction mixtures were extracted with phenol:

chloroformiisoamyl alcohol (25:24:1) and precipitated with ethanol. The re

action products were separated on 6% polyacrylamide sequencing gels andsubjected to autoradiopraphy. The sequencing of the probes was performed as

described by Maxam and Gilbert (17). and the sequencing mixtures wereloaded adjacent to the DNase I cut samples.

Cell Transfection and CAT Assay. SK-BR-3 cells were transfected with

Lipofectin (Life Technologies. Inc.) according to the instructions provided.The cotransfection experiments were performed 20 to 24 h after seeding~4 X 10s cells/60-mm tissue culture dish using 10 jug of the test plasmid and

3 /ig of the plasmid pSV-/3-Gal. which was used to normalize the transfection

efficiency. After overnight incubation, media containing lipofectin and DNAwere removed, and the cells were cultured in the regular growth medium. After24-h incubation, the cells were harvested from the plates by scraping, pelletedby centrifugation, resuspended in 0.25 M Tris-HCI (pH 8.0), and disrupted by

freeze-thawing four times. Aliquots of the lysate were used for assay ofÃŸ-galactosidase activity (18). CAT activity in the cell extract containing an

equal amount of ÃŸ-galactosidase activity from each sample was determined by

the liquid scintillation counting method (19). Briefly, the appropriate amountof cell extracts was incubated in a reaction containing MC-labeled chloram-

phenciol and H-butyryl coenzyme A. The reaction products were extracted with

a small volume of xylene. The xylene phase was mixed with scintillant andcounted in a scintillation counter. The CAT activity was expressed as relative

activity compare with that of the pUMSVOCAT construct ( 1.0) and shown asthe mean Â±SE of three independent transient transfection experiments per

formed for each construct.

Results

Identification of trans Factors in Human Breast Tissue ThatBind to SI by the Yeast One-Hybrid Screening Method. Asdescribed in "Materials and Methods," three tandem copies of S l were

introduced into two reporter vectors, pHISi and pLacZi. The resultingplasmids pHISi-3Sl and pLacZi-3Sl were linearized and used to

transform YM4271 yeast strain. A dual reporter strain was used toscreen 1 X IO6 independent clones from a premade MATCHMAKER

mammary tissue cDNA library. Twenty-five positive clones wereobtained, of which 13 encoded for ERRa-1. Additional positiveclones encoded for EAR-2 (four clones), EAR-3 (COUP-TF1, three

clones), RAR y (three clones), p 120E4F (one clone), as well as a novelclone. Because ERRa-1 is clearly the dominant clone, we decided to

focus our attention first on the interaction of this orphan receptor withSI.

Despite the fact that the nuclear receptor SF-1 was shown previ

ously to bind to the same region in the human aromatase gene and tofunction as a cAMP-responsive element in ovary (33), our yeastone-hybrid screening did not find any SF-1 clones. Because ourRT-PCR analysis was not able to detect SF-1 mRNA in breast cancertissue nor in SK-BR-3 cells or tumor fibroblasts (results not shown),it is thought that SF-1 protein is absent in breast cancer tissue.

Ã„UG AUG-177-114 1 177

Fig. I. Scheme of the cDNA for human ERR-I (from nt -177 to 1433; Ref. 14) andERRa-1 (from nt 1 to 1433; rÃ©f.13). -144 ni and 177 ni are potential transnational startsites for ERRI and ERRal. respectively, a, b, c. and (/ are primers used in RT-PCRstudies as described in "Materials and Methods."

Structural Characterization and Expression of ERRa-1 inBreast Cancer Tissue. The longest insert of our ERRa-1 clones was2170-bp long containing 720-bp of the 3'-untranslated sequence. We

have found that the nucleotide sequence of our clone matches with theERRa-1 sequence reported by Yang et al. (13) except for a three-base(CAG) insertion between 745 and 746 nt. This three-base insertion

introduces an additional amino acid Gin and does not change thereading frame. The ERRa-1 sequence reported by Yang et al. (13) is177-bp shorter at the 5'-end than that of ERR-1 which was published

in 1988 (14). To better describe our findings, the region missing fromERRa-1 is indicated in negative number (Fig. 1). In addition, theERRa-1 sequence is different from that of ERR-1 in three areas, 1-17nt, 745-746 nt, and 1206-1238 nt (13). There are three additionalbases, CAG, between 745 and 746 nt of the ERR-1 clone, the same asour ERRa-1 clone. The significance of the three-base differencebetween our ERRa-1 cDNA and that of Yang et al. (13) has not yet

been determined.Two potential translational start sites for ERRa-1/ERR-1 RNA.

â€”114nt and 177 nt, have been suggested (13, 14). Using specific 5'primers (as described in "Materials and Methods"), we examined the

presence of ERRa-1 mRNA in breast cancer tissues by the RT-PCRmethod. RT-PCR products were generated in 28 of 32 breast tumorspecimens when primers a and d were used. The RT-PCR product was

detected in only one specimen when primers c and d were used. Theseresults indicate that the major mRNA species in breast cancer tissue isthe shorter variant, which has a potential translational start site at 177nt. In addition, ERRa-1 was found to be expressed in SK-BR-3 breast

cancer cells and in a tumor fibroblast line WS3TF. Because theERRa-1 clone was identified from a normal human breast tissuelibrary, it is thought that ERRa-1 is also present in normal breasttissue. Preliminary results obtained from RT-PCR analysis of RNA

samples isolated from seven pairs of breast cancer and adjacentnoncancerous tissues suggest that ERRa-1 is expressed at a higherlevel in cancer tissue than in noncancerous tissues.6

Confirmation of the Binding of ERRa-1 to SI by DNA MobilityShift Assay and DNase I Footprinting Analysis. We preparedERRa-1 protein for the DNA mobility shift analysis using an in vitro

translation method and for DNase I footprinting analysis using an E.coli expressed preparation. We successfully produced ERRa-1 using

expression constructs with the translational start site at 177 nt butwere not successful in the expression of protein starting from theupstream site (â€”114 nt). These results suggest that nucleotide 177 is

the true translational start site, agreeing with reports on ERRa-1expression in RL95-2 cells (13).

We have demonstrated that ERRa-1 binds to SI in a dose-dependent manner (Fig. 2), in a manner similar to SF-1. Our results suggestthat SF-1 binds to SI with a binding affinity comparable withERRa-1. Such findings help us to conclude that the reason for notdetecting SF-1 by our yeast one-hybrid screening method is due to theabsence of SF-1 in breast tissue rather than differences in binding

s Unpublished results.

5697




A Blx 2x 4x 8x 16x80x Ix 2x 4x 8x 16x80x103cpm

Fig. 2. Gel mobility shift assays of oligonuclcotidc SI in the presence of in vitrotranslated Ml (-4) and ERRa-1 (B}. Binding reactions contain a constant amount of invitro translated proteins and the increasing amount of radiolabeled SI probe as indicatedabove each lane. SI sequence: 5'-ACCAAGGTCAGAAATGCAATTCAAGCCAA-3'.

HERRI

competitor _ _

CAT activity in a dose-dependent manner with 2.5-fold maximum

induction at a concentration of 5 p.g (Fig. 5). Cotransfection ofERRa-1 had no effect on CAT expression in the cells transfected with

CAT plasmid containing only the promoter 1.3 region without SI,demonstrating that the transactivation function of ERRa-1 on the

aromatase promoter is achieved through specific binding to SI in thearomatase promoter 1.3 region.

Discussion

The cDNA for ERR-1 was first isolated by screening cDNA libraries with probes corresponding to the DNA-binding domain of thehuman ERa (14). Sequence alignment of ERR-1, ERa, and ERÃŸrevealed a high similarity. In the DNA-binding domain, ERR-1 shares

68% amino acid homology with ERa and 70% with ERÃŸ.In thesteroid-binding region, the amino acid sequence of ERR-1 shows 36%identity when compared with ERa and 34% to ERÃŸ.However, ERR-1

does not bind to any of the major classes of steroids, includingestrogens and androgens (14). As indicated in "Results," ERRa-1 is

an isoform of ERR-1 and first reported by Yang et al. (13). Because

the physiological significance of the structural and functional differences between ERR-1 and ERRa-1 has not yet been clearly defined,

we discuss the previous findings on these proteins in a collectivemanner. The human ERR-1/ERRa-1 appears to be widely distributed

although more abundant in the brain (14), heart (21), skeletal muscle(21, 22), and brown adipose tissue (23). A role of ERR-1/ERRa-1 in

Fig. 3. Characicn/alion ol Ihe KRRÂ»-I binding MICh\ DNA mobilii} shiti ,ixvi\-. Invitro Iranslaled ERRa-1 prolcin was incubated wilh '"P-labeled, double-stranded SI probe

in (he absence (Lane 2} or presence of 50-fold molar excess of compeling oligonucleotides(all competitors are double-stranded excepl SI-5') as indicated (Lanes 3-9). SI, 5'-ACCAAGGTCAGAAATGCTGCAATTCAAGCCAA-.V; 5'-half of SI. 5'-ACCAAG-GTCAGAAATGC--V; J'-halfafSI. 5'-TGCAATTCAAGCCAA-3'; SF-I. 5'-CCAAG-GTCA-3'; Bl contains the sequences of promoter 1.3.

ERRa-1 pig)

AG o S S $

r

affinities between SF-1 and ERRa-1. The binding of ERRa-1 to SIcan be competed with the 5'-half probe of SI but not 3'-half probe,indicating that the binding site is situated in the 5'-half region of SI

(Fig. 3). As a control, the single-strand DNA probe could not competewith SI for the binding of ERRa-1. The exact binding site wasdetermined by DNase I footprinting analysis. ERRa-1 binds to aregion within the 5'-half region of SI, 5'-AAGGTCAGAAAT-3',

between 96 and 107 bp relative to the transcriptional start site ofpromoter 1.3 (Fig. 4). This sequence contains the AGGTCA motif,which is present in the ERRa-1 binding elements in other genes (20).

Interestingly, several intensive bands appeared in footprinting analysisperformed with a high concentration of ERRa-1. Such results suggesta conformational change of DNA upon binding of ERRa-1.

Functional Analysis of hERRa-1 on the Promoter 1.3 Activity.To address the biological significance of ERRa-1 expression onaromatase function, we cotransfected the SK-BR-3 human breastcancer cells with an expression vector for ERRa-1 along with a CAT

reporter plasmid containing the aromatase genomic fragment havingpromoter 1.3 and SI. Human ERRa-1 induced promoter I.3-driven

"â€¢rr+107

11 +96

Fig. 4. DNase I footprinling analysis of Ihe SI-containing DNA fragment in Ihepresence of ERRa-1 prolein. The end-labeled 280-bp fragment thai contains SI sequence

was subjected to the Maxam and Gilben sequencing reactions (AG) or DNase 1digeslion(2 units of DNase I for 1 mini in the absence (0) or presence of 10. 15. or 20 fig ofhERRal, which was expressed and purified from E. coli. The protected region, 5'-

AAGGTCAGAAAT-.V. belween 96 bp and 107 bp relative to the transcriptional start site

of promoter 1.3 is indicated by an open hoi.

5698



MODULATION OF AROMATASF. EXPRESSION IN BREAST TISSUE

2(XK)

loon

pSG5-ERRa-l 0 1 |ig 2 \u) 4 |ig 5 ng 6 M9 8 jig 10 |ig

Fig. 5. Functional analysis of ERRa-1 on the promoter 1.3 of the human aromatasegene. SK-BR3 cells were cotransfected with a CAT reporter plasmid (5 /ig; whichcontains promoter 1.3 and SI) and ERRa-1 expression plasmid. pSG5-hERRal. at theindicated concentrations. CAT activities were determined 72 h after transfection and areplotted as the radioactivity (CPMÃŒof the acetylated chloroamphenicol. The graph represents an average of three independent experiments; harx. SD.

bone development (24), skeleton formation (22), and fat metabolism(23) has been suggested. Recent studies from our laboratory haverevealed that ERRa-1 is expressed in breast tissue and may modulate

aromatase expression/estrogen biosynthesis in this tissue.ERRa-1 is thought to be a member of the orphan nuclear receptor

family. Our in vitro translation experiment indicates that the transla-

tional start site is at 177 nt; therefore, the protein is predicted to be 423amino acids long. The calculated molecular weight of translatedERRa-1 is Mr 45,000. However, SDS-PAGE analysis of the ex

pressed protein showed a band at M, 53,000. It is thought that themigration rate of the expressed ERRa-1 on SDS gels can be affected

by its amino acid composition. Our laboratory showed previously thatdifferent migration rates of human, rat, and mouse forms of DT-

diaphorase could be explained by differences in one amino acidamong three forms (25). ERR-l/ERRa-1 has been demonstrated tobind as a monomer, with a high-affinity binding site containing theextended half-site sequence 5'-TCAAGGTCA-3', including some that

were shown previously to function as estrogen-responsive elements(20). Our DNase I footprinting analysis revealed that ERRa-1 recognized the sequence 5'-AAGGTCAGAAAT-3' in the human aro

matase gene. ERR-l/ERRa-1 was shown to function through the cis

element as a transcriptional activator for several promoters, includingosteopontin promoter (24) and medium-chain acyl coenzyme A de-

hydrogenase promoter (23), or as a transcriptional represser for theSV40 late promoter in both cell culture and cell-free transcriptionsystems (20, 26). The expression of ERRa-1 itself was found to beup-regulated by estrogen in mouse uterus (27). In addition, recent datashow that, like COUP-TF, in addition to contacting the basal transcriptional machinery by directly contacting with TFIIB (20), ERR-1

modulates the activation effect of estrogen on a number of genepromoters by both direct DNA-binding competition and through ER-ERR-1 protein-protein interaction (13, 20).

It is well accepted that both estrogen and ER are critical for normalbreast development and play an important role in the pathogenesis andmaintenance of breast cancer. As discussed, ERRa-1 is expressed in

breast tissue and, therefore, can form a heterodimer with ER orcompete with ER for binding to the estrogen response element. We

have now found that ERRa-1 can also modulate aromatase expres

sion/estrogen biosynthesis in breast tissue. It has been shown thataromatase is expressed at a higher level in breast cancer tissue than innoncancerous tissue (7, 8, 28, 29), and in situ aromatase/estrogenbiosynthesis is critical for breast cancer development (30). Promoters1.3 and II are the major promoters driving aromatase expression inbreast tumor (7), and the functions of these promoters are regulated bySI, the ERRa-1 binding site. Therefore, it is logical to propose thatERR-1 may play a role during breast cancer development by modu

lating estrogen receptor action as well as estrogen synthesis. Asindicated, preliminary results obtained from RT-PCR analysis of RNA

samples isolated from paired breast cancer and adjacent noncanceroustissues suggest that ERRa-1 is expressed at a higher level in cancertissue than noncancerous tissues.6

The present study indicates that ERRa-1 has a positive regulatory

effect by interacting with SI. Although it is possible that SI functionsas a positive regulatory element through an interaction with ERRa-1,

SI was demonstrated previously to be a negative regulatory elementin several cell lines examined in our laboratory (10, 11). Our DNaseI footprinting analysis has revealed that ERRa-1 binds to the 5'-half

region of SI (as shown in this report), and at least four proteins wereseen in UV cross-linking experiments of SI using nuclear extract

preparations from breast cancer cells and adipose stromal cells (11).The negative regulatory action of SI may result from an interaction ofERRa-1 with other nuclear receptors such as EAR-3 (COUP-TF 1),

which is known to have a negative regulatory function (31) or withcorepressor proteins. As indicated, we have identified EAR-3 duringyeast one-hybrid screening, and initial experiments suggest it has anegative regulation of promoter 1.3 in the human aromatase gene.6

Gene silencing by EAR-3 is mediated by transcriptional corepressorsN-CoR and SMRT (32). Corepressor proteins such as N-CoR and

SMRT have a molecular weight of Mr 150,000. Indeed, we havedetected proteins with A/r 150,000 binding to S l as demonstrated byUV cross-linking experiments (11). The negative action of SI mayresult from complex formation among ERRa-1, other nuclear receptors, and corepressor proteins. It was found previously that a nucle-otide with the sequence 5'-CCAAGGTCA-3' (a sequence recognized

normally by SF-1 or ERRa-1 ), at a 50-fold molar excess, was not able

to compete with SI for nuclear protein binding (11). Interestingly, a50-fold molar excess of SI, but not the nucleotide 5'-CCAAGGTCA-3', was found to compete effectively for nuclear protein binding to aradioactive probe with the sequence of 5'-CCAAGGTCA-3'. These

results suggest that additional regions of SI are involved in theinteraction with the silencer protein complex. This results in nuclearproteins having a higher affinity for S l than the region recognized byERRa-1 or SF-1. These interactions are being carefully evaluated inour laboratory. It is our present hypothesis that ERRa-1 is a transcrip

tional activator when binding alone to SI, but its modulating activitycan be changed by interacting with coregulatory proteins.

A significant number of nuclear receptors are known to recognizethe 5'-AGGTCA-3' motif (26). Our yeast one-hybrid screening has

identified some but not all of the proteins known to recognize thisnucleotide sequence. It is not surprising that several transcriptionalfactors can compete for the same regulatory element. Our yeastone-hybrid screening revealed that ERRa-1 is the major nuclear

receptor interacting with SI in breast tissue. We will determinewhether ERRa-1 has a higher affinity for S l or expresses at a higherlevel in breast tissue than other nuclear receptors such as EAR-2 andEAR-3. We have presented results to show that ERRa-1 and SF-1bind to SI with similar affinities; only ERRa-1. but not SF-1, isexpressed in the breast tissue. On the other hand, although SF-1 is

present in ovarian tissue, Northern blot analysis by Shi et al. (21)revealed that the expression level of ERRa-1 in ovary is very low.

5699




Differential expression of transcriptional factors could be an important mechanism to modulate the regulatory function of a cis element.An investigation from Simpson's laboratory (33. 34) suggests that an

Ad4BP/SFl element (which overlaps with the site recognized byERRa-1 ) and a CRE-like sequence upstream from promoter II are

critical for the cAMP induction of promoter II in ovary tissue. Theinteraction of these elements and the mechanism of cAMP inductionof promoter II are similar to those described for the transcriptionalcontrol of aromatase expression in rat granulosa cells (35). Considering the fact that SF-1 is not expressed in breast tissue, the regulatory

mechanism of promoter II in this tissue may not be the same as thatdescribed in the ovary.

In summary, data present here suggest an important function forERRa-1. We found that this nuclear receptor is expressed in human

breast tissue and can regulate the expression of aromatase. The exactmechanism, including its interaction with other nuclear receptors andcofactors, has yet to be characterized. However, considering its abilityto interact with ER and to modulate aromatase expression/estrogenbiosynthesis, ERRa-1 could be critical for normal breast development

and play an important role in the pathogenesis and maintenance ofbreast cancer.

References

1. Mahcndroo. M. S., Means, G. D.. Mendelson, C. R., and Simpson. E. R. Tissue-

specific expression of human P450arom. The promoter responsible for expression inadipose (issues is different from thai utilised in placenta. J. Biol. Chem.. 266:11276-1128, 1991.

2. Means, G. D.. Kilgore, M. W.. Mahendroo. M. S.. Mendelson, C. R., and Simpson.E. R. Tissue-specific promoters regulate aromatase cytochrome P450 gene expressionin human ovary and fetal tissues. Mol. Endocrino]., 5: 2005-2013, 1991.

3. Horada. N. A unique aromatase (P-450arom) mRNA formed by alternative use oftissue-specific exons 1 in human skin fibroblasts. Biochem. Biophys. Res. Commun..789: 1001-1007. 1993.

4. Mahendroo, M. S., Mendelson, C. R.. and Simpson. E. R. Tissue-specific and

hormonally controiled alternative promoters regulate aromatase cytochrome P450gene expression in human adipose tissue. J. Biol. Chem., 268: 19463-19470. 1993.

5. Harada, N., Ulsumi. T., and Takagi, Y. Tissue-specific expression of the humanaromatase cylochrome P-450 gene by alternative use of multiple exons I and promoters, and switching of tissue-specific exons I in carcinogenesis. Proc. Nati. Acad.Sci. USA, 90: 11312-11316, 1993.

6. Zhao. Y.. Mendelson, C. R.. and Simpson. E. R. Characterization of the sequences ofthe human CYP 19 (aromatase) gene that mediate regulation by glucocorticoids inadipose stromal cells and fetal hepatocyctes. Mol. Endocrinol., 9: 340-349. 1995.

7. Zhou. C.. Zhou. D.. Esteban. J.. Murai, J.. Sincri. P. I.. Wilczynski. S., and Chen. S.Aromatse gene expression and its exon I usage in human breast tumors. Detection ofaromalase messenger RNA by reverse transcription-polymerase chain reaction (RT-PCR). J. Steroid Biochem. Mol. Biol.. 59: 163-171, 1996.

8. Harada. N. Aberrant expression of aromatase in breast cancer tissue. J. SteroidBiochem. Mol. Biol.. 61: 175-184, 1997.

9. Agarwal. V. R.. Bulun. S. E.. Leitch. M.. Rohrich, R., and Simpson. E. R. Use ofalternative promoters to express the aromatase cytochrome p450 (CYP 19) gene inbreast adipose tissues of cancer-free and breast cancer patients. J. Clin. Endocrinol.Metab.. HI: 3843-3849. 1996.

10. Wang. J.. and Chen, S. Identification of a promoter and a silencer at the 3'-end of the

first intron of the human aromatase gene. Mol. Endocrinol.. 6: 1479-1488. 1992.

11. Zhou. D.. and Chen. S. Characterizalion of a silencer element in the human aromatasegene. Arch. Biochem. Biophys.. 353: 213-220. 1998.

12. Chirgwin. J. M.. Przybyia, A. E.. MacDonald. R. J.. and Rutter. W. F. Isolation ofbiologically active ribonucleic acid from sources enriched in rebonuclease. Biochemistry, 18: 5294-5299. 1997.

13. Yang. N.. Shigeta. H.. Shi, H.. and Teng, C. T. Estrogen-related receptor. hERRl.modulates estrogen receptor-mediated response of human lactoferrin gene promoter.J. Biol. Chem.. 271: 5795-5804. 1996.

14. Giguere, V.. Yang, N., Segui, S., and Even. R. M. Identification of a new class ofsteroid hormone receptors. Nature (Lond.l, 331: 91-94, 1988.

15. Ikeda, Y., Lala, D. S., Luo, X.. Kim, E., Moisan. M-P.. and Parker. K. L. Characterization of the mouse FTZ.-F1 gene, which encodes a key regulator of steroidhydroxylase gene expression. Mol. Endocrinol., 7: 852-860. 1993.

16. Singh. H.. Sen. R.. Baltimore. D.. and Sharp. P. A. A nuclear factor that binds to aconserved sequence motif in transcriptional control element of immunoglobulingenes. Nature (Lond.), 319: 154-158. 1986.

17. Maxma. A. M., and Gilbert, W. Methods Enzymol., 65: 499-560. 1980.

18. Rosenthal, N. Identification of regulatory element of cloned genes with functionalassays. Methods Enzymol. 752: 705-720, 1987.

19. Seed, B., and Sheen, J. Y. A simple phase-extraction assay for chloramphenicolacyltransferase activity. Gene (Amsl.), 67: 271-277, 1988.

20. Johnston, S.. Liu, X., Zuo. F., Eisenbraun. T. L.. Wiley. S. R.. Kraus. R. J.. and Mertz,J. E. Estrogen-related receptor al functionally binds as a monomer to extendedhalf-site sequences including ones contained within estrogen-response elements. Mol.Endocrinol.. //: 342-352. 1997.

21. Shi. H.. Shigeta. H.. Yang, N.. Fu. K.. O'Brian, G., and Teng. C. T. Human estrogen

receptor-like l (ESRL1) gene: genomic organization, chromosomal localization, andpromoter characterization. Genomics, 44: 52-60, 1997.

22. Bonnelye, E., Vanacker, J. M., Spruyt, N., Alric, S., Fournier, B., Desbiens. X., andLaudet, V. Expression of the estrogen-related receptor 1 (ERR-1 ) orphan receptorduring mouse development. Mech. Dev.. 65: 71-85. 1997.

23. Sladek, R.. Bader. J. A., and Giguere. V. The orphan nuclear receptor estrogen-relatedreceptor a is a transcriptional regulator of the human medium-chain acyl coenzymeA dehydrogenase gene. Mol. Cell. Biol.. 17: 5400-5409. 1997.

24. Bonnelye. E.. Vanacker, J. M.. Dittmar. T., BÃ¨gue,A.. Desbiens. X.. Denhardt. D. T..Aubin. J. E.. Laudet. V., and Fournier. B. The ERR-1 orphan receptor is a transcriptional activator expressed during bone development. Mol. Endocrinol., //: 905-916.

1997.25. Chen, S.. Knox. R., Lewis, A. D., Friedlos. F., Workman, P.. Deng. P. S-K., Fung. M..

Ebenstein, D., Wu, K.. and Tsai. T-M. Catalytic properties of NAD(P)H:quinoneacceptor oxidoreductase: study involving mouse, rat, human, and mouse-rat chimericenzymes. Mol. Pharmacol.. 47: 934-939, 1995.

26. Zho. F.. and Mertz, R. J. Simian virus 40 late gene expression is regulated bymembers of the steroid/thyroid hormone receptor superfamily. Proc. Nati. Acad. Sci.USA, 92: 8586-8590. 1995.

27. Shigeta. H., Zuo, W., Yang, N.. Diaugustine. R.. and Teng. C. T. The mouse estrogenreceptor-related orphan receptor a 1: molecular cloning and estrogen responsiveness.J. Mol. Endocrinol., 19: 299-309, 1997.

28. Esteban. J., Warsi. Z., Haniu, M.. Hall. P.. Shiverly, J. E., and Chen, S. Detection ofintratumoral aromatase in breast carcinomas. Am. J. Pathol., 140: 337-343. 1992.

29. Santner, S.. Pauley. R. T.. Tait. L.. Kaseta. J., and Samen, R. J. Aromalase activity andexpression in breast cancer and benign breast tissue stromal cells. J. Clin. Endocrinol.Melab.. 82: 200-208, 1997.

30. Sun. X-Z., Zhou. D.. and Chen. S. Autocrine and paracrine actions of breast tumoraromalase. A three-dimensional cell culture study involving aromatase transfectedMCF-7 and T-47D cells. J. Steroid Biochem. Mol. Biol., 63: 29-36, 1997.

31. Klinge. C. M., Silver. B. F., Driscoll, M. D.. Sathya, G., Bambara, R. A., and Hilf. R.Chicken ovalbumin upstream promoter-transcription factor interacts with estrogenreceptor, binds to estrogen response elements and half-sites, and inhibits estrogen-induced gene expression. J. Biol. Chem., 272: 31465-31474. 1997.

32. Shibata, H.. Nawaz, Z.. Tsai, S. Y., O'Malley. B. W., and Tsai, M-J. Gene silencing

by chicken ovalbumin upstream promoter-transcription factor I (COUP-TF1) ismediated by transcriptional compressors, nuclear receptor-corepressor (N-CoR) and

silencing mediator for relinoic acid receptor and thyroid hormone receptor (SMRT).Mol. Endocrinol.. //: 714-724. 1997.

33. Michael. M. D.. Kilgore. M. W.. Morohashi. K-I., and Simpson, E. R. Ad4BP/SF-lregula(es cyclic AMP-induced transcriplion from the proximal promoter (PII) of thehuman aromatase P450 (CYP I9| gene in Ihe ovary. J. Biol. Chem.. 270: 13561-13566. 1995.

34. Michael, M. D., Michael, L. F., and Simpson, E. R. A CRE-like sequence lhat bindsCREB amd contributes to cAMP-dependenl regulalion of the proximal promoter ofthe human aromatase P450 (CYP 19) gene. Mol. Cell. Endocrinol., 134: 147-156,

1997.35. Carlone, D. L., and Richards. J. S. Functional interactions, phosphorylation. and

levels of 3'.5'-cyclic adenosine monophosphale-regulatory element binding prolein

and steroidogenic factor-1 mediate hormone-regulated and constitutive expression ofaromatase in gonadal cells. Mol. Endocrinol., //: 292-304, 1997.

5700



1998;58:5695-5700. Cancer Res Chun Yang, Dujin Zhou and Shiuan Chen

-1 Orphan ReceptorαERRModulation of Aromatase Expression in the Breast Tissue by

Updated version

http://cancerres.aacrjournals.org/content/58/24/5695

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/24/5695To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



modulation of aromatase expression in the breast tissue by ... · an oligonucleotide and its...

Documents