Suppression of cell proliferation and regulation of estrogen ...540 nm was measured. Percentage survival was deﬁned as: % survival=100% (O.D. of test sample/O.D. of control). Direct

Suppression of cell proliferation and regulation of estrogen receptor� signaling pathway by arsenic trioxide on human breast cancerMCF-7 cells

Stephanie K Y Chow1, Judy Y W Chan1 and Kwok Pui Fung1,21Department of Biochemistry, Mong Man Wai Building, The Chinese University of Hong Kong, Shatin, Hong Kong, China2Institute of Chinese Medicine, The Chinese University of Hong Kong, Shatin, Hong Kong, China

(Requests for offprints should be addressed to K P Fung, Department of Biochemistry, Room 603, Mong Man Wai Building, The Chinese University ofHong Kong, Shatin, Hong Kong, China; Email: [email protected])

Abstract

In recent years, breast cancers have aroused much concern.Together with a growing incidence all over the world,the development of drug resistance to tamoxifen, themost commonly prescribed chemotherapeutic drug forbreast cancer patients, has highlighted the importance ofdeveloping a new chemotherapeutic drug in combatingbreast cancer. With the aim of treating breast cancers, theanti-tumor effects of arsenic trioxide in MCF-7 cells havebeen studied.

MCF-7 cells are estrogen responsive cells which mimicbreast cancers at the early stage. 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT)assay and direct cell counting were used to measure cellproliferation. The mechanisms of action were elucidatedthrough the measurement of estrogen receptor (ER)binding, mRNA and protein levels of ER� and its activity.

We have demonstrated that arsenic trioxide wascapable of reducing cell survival in MCF-7 cells via the

suppression of the estrogen-induced growth stimulatoryeffects in MCF-7 cells. Arsenic trioxide was shown tosuppress the action of estrogen through the regulation ofthe ER� signaling pathway. Arsenic trioxide could down-regulate ER� mRNA and protein levels without compet-ing with estrogen for ER� binding. Arsenic trioxide alsoinhibited the transcription activity mediated by the ER�signaling pathway and ultimately it down-regulated c-mycprotein expression and inhibited cell entry to S phaseunder estrogen’s stimulation.

In conclusion, arsenic trioxide could inhibit the growthof MCF-7 cells by reducing the growth stimulatory effectof estrogen. As estrogen is a primary risk factor inpromoting the growth of breast tumor cells, the anti-estrogenicity exhibited by arsenic trioxide sheds light onthe therapy of breast cancer.Journal of Endocrinology (2004) 182, 325–337

Introduction

Arsenic trioxide (As2O3) is an arsenic compound existingas an odorless, tasteless, white crystal or powder. In the1970s, the effects of As2O3 on a number of cancers wereinvestigated and it was found to be most effective in killingacute promyelocytic leukemia (APL) cells. Later, anothergroup from Shanghai found promising effects of As2O3 inclinical trials of APL patients. APL accounts for 10–15%of all acute myeloid leukemia in adults (Soignet et al.1998). In earlier times, APL patients were treated withanthracyclines. This cytotoxic chemotherapy achieved70–85% remission but also induced severe complicationsand drug resistance in APL patients (Warrell et al. 1993).It was not until the characterization of the molecularpathologies of APL that a group from Shanghai found thatall-trans retinoic acid (ATRA) was effective in combat-ing the disease. Clinical trials indicated a complete

remission in 85–90% of cases (Huang et al. 1988).Although ATRA treatment showed improvement incoagulopathy and high complete remission in APL patients(Huang et al. 1988), adverse effects were also observed.Hyperleukocytosis developed during the second and thirdweeks of treatment. It is attributed to the activation ofleukocytes (Castaigne et al. 1990). Together with othersyndromes such as respiratory distress, pleural effusions,weight gain, fever and occasionally renal failure, theseadverse effects are collectively called ‘retinoic acid syn-drome’. Moreover, ATRA resistance was observed inpatients following ATRA treatment. Since then, frequentclinical trials have been performed to assess the effective-ness of As2O3 in APL treatment. As2O3 achieved acomplete remission rate of 57–98% in both de novo andrelapsed APL patients. Moreover, patients receivingAs2O3 treatment had improved disseminated intra-vascular coagulation, hyperfibrinolysis and bleeding

325

Journal of Endocrinology (2004) 182, 325–3370022–0795/04/0182–325 � 2004 Society for Endocrinology Printed in Great Britain

Online version via http://www.endocrinology.org

Downloaded from Bioscientifica.com at 06/04/2021 03:16:19PMvia free access

http://www.endocrinology.org

symptoms (Shen et al. 1997, Guo et al. 1999). Studies inNB4 cells have shown that As2O3 induces apoptosis atconcentrations of 1–2 µM. When NB4 cells weretreated with low concentrations of As2O3 (0·1–0·25 µM)over 10 days, condensed chromatin, smaller nuclei, anddecreased numbers of nuclei were observed. Cytoplasmratio and appearance of granules in the cytoplasm werealso observed (Chen et al. 1996). Further investigationshowed the up-regulation of CD52 and BFl-1, thedown-regulation of retinoic acid receptor �, andsynergistic effects in the regulation of protein kinaseCB-1 and small ubiquitin-related modifier 1 (SUMO-1)(Cai et al. 2000). Recent findings showed that As2O3induced acetylation of histones 3 and 4 resulting intranscriptional activation of downstream genes fordifferentiation (Wang 2001).

To date, little is known about the effect of As2O3 onhuman breast cancer. In fact, public concern and aware-ness about breast cancer have been heightened recently.At present, the cause of breast cancer is still unclear. Thefemale hormone, estrogen, plays an essential role in thegrowth and differentiation of tissues in the female repro-ductive system. Over the past years, both in vitro andin vivo studies have provided support for the role ofestrogen in the development of breast cancer. Estrogen wasfound to be the primary stimulant of the growth of breastcancer (Lippman et al. 1976, Henderson et al. 1982). Themechanism of action of estrogen on breast cancer involvesboth promotion of cell proliferation and prevention ofapoptosis (Kyprianou et al. 1991). Estrogen increased thecell proliferation rate by recruiting non-cycling cells intothe cell cycle, shortening the overall cell cycle timethrough the reduction of the length of the G1 phase andpromoting cell entry to the S phase (Brunner et al. 1989,Clarke et al. 2001). Estrogen was found to promoteresistance of chemotherapeutic drugs on MCF-7 cells byup-regulation of Bcl-2 mRNA and protein levels which isan anti-apoptotic protein (Teixeira et al. 1995). Similarstudies had also shown the inhibition of paclitaxel orultraviolet (UV) radiation-induced apoptosis by estrogenthrough the inhibition of c-Jun N-terminal kinase (JNK)activity, bcl-2 and bcl-xl phosphorylation, activation ofcaspase 9 and ultimately induction of apoptosis. The actionof estrogen was mediated through the estrogen receptor(ER). Two ERs have been identified, estrogen receptor �(ER�) and estrogen receptor � (ER�). ER� was onlyidentified in 1996 (Mosselman et al. 1996). At the onset ofbreast cancer, 46–77% of breast cancers are ER� positive(Dickson & Lippman 1995). Thus, ER� expression inhuman breast tumors is an important prognostic indicatorand marker of the responsiveness of endocrine therapies.Studies of the regulation of ER� and its respectivesignaling pathway will unravel the stimulatory effects ofestrogens on proliferation in breast cancer cells and thusshed light on the development of strategies in treatingbreast cancer.

The present study mainly focused on the estrogenreceptor-dependent signaling pathway. Studies werefocused on using As2O3 in concentrations below 2 µM toelucidate the mechanism of action in mediating theanti-tumor effect on MCF-7 cells.

Materials and Methods

Cell culture

MCF-7 and MDA-MB-231 cell lines were purchasedfrom the American Type Culture Collection (Baltimore,MD, USA). They were maintained in RPMI 1640medium supplemented with 10% dextran-coatedcharcoal-stripped fetal bovine serum and 1% antibiotics(v/v) at 37 �C in a humidified atmosphere of 5% CO2.

Preparation of As2O3The stock solution of As2O3 was prepared by dissolvingAs2O3 powder in PBS at a concentration of 10 mM.

MTT assay

MCF-7 cells and MDA-MB-231 were seeded at1�104 cells/well in 96-well plates. After treatment withthe appropriate concentration of As2O3, suspensionmedium was removed and 30 µl MTT solution wereadded to each well and incubated at 37 �C for 2–4 h.After that, 100 µl DMSO were added to each well andincubated for a further 15 min. Optical density (O.D.) at540 nm was measured. Percentage survival was definedas: % survival=100%�(O.D. of test sample/O.D. ofcontrol).

Direct cell counting by the trypan blue dye exclusion method

MCF-7 cells were seeded at 5�104 cells/well in 24-wellplates. After treatment with the appropriate concentrationof As2O3 and 17�-estradiol for 24, 48 or 72 h, the cellnumber in each well was counted by the trypan bluestaining method using a hematocytometer.

Effect of As2O3 on estrogen binding to ER� by ER�competitive binding assay

The assay was carried out using an ER� competitivescreening kit according to the manual provided by themanufacturer (Wako Chemical, Richmond, VA, USA;catalogue number 295-56301). In brief, human recom-binant ER� was coated in microplates. When incubatingsamples with fluorescence-labeled estrogen, there is com-petition with estrogen for the binding sites on ER�. Afterremoval of unbound substances or fluorescence-labeledestrogen, the retained estrogen was determined by

S K Y CHOW and others · Effect of As2O3 on ER� signaling326

www.endocrinology.orgJournal of Endocrinology (2004) 182, 325–337



measuring the fluorescent intensity. A control offluorescence-labeled estrogen alone was prepared to act asa positive control. Excitation of the sample mixtureemitted high fluorescent intensity. The extent of bindingdepends on the capability of the substance in competingwith estrogen and this is correlated to the affinity of thatsubstance to ER�.

Assessment of the transcriptional activity of ER�

Estrogen response element (ERE) containing vector wasconstructed by insertion of a single stranded ERE oligo-nucleotide into pGL3 basic vector at NheI site upstream ofthe luciferase gene. The sequence of ERE oligonucleotidewas constructed according to the previous report (Cullenet al. 2001).

The sequence is: 5�-CCAGGTCAGAGTGACCTGAGCTAAAATAACACATTCAG-3�. pGL3-controlvector was used as a positive control which emitted strongluminescence upon activation. pGL3-basic vector wasused as a negative control which is lacking in promoter andenhancer sequences. Renilla luciferase reporter plasmidwas prepared to act as an internal control for the determi-nation of transfection efficiency. The treatment schemewas as follows. Cells were seeded at 5�104 cells/well in24-well plates until confluence at 40–60%. Four wellswere prepared for each treatment. Experimental medium

without serum supplement (50 µl) was transferred tomicrotubes to which 300 ng ERE luciferase reporterplasmid, 100 ng Renilla luciferase reporter plasmids and1 µg Fugene 6 transfection reagent were added. Themixture was incubated at room temperature for 30 min.Next, the cells were incubated with transfection mixturefor 6 h at 37 �C. After that, the mixture was discarded.The transfected cells were then treated with 10 nM17�-estradiol, As2O3 at 1 µM, As2O3 at 2 µM or As2O3plus 17�-estradiol. The cells were treated for either 48 or72 h. Luciferase assay was carried out using the Dual-Luciferase Reporter Assay System (Promega, Madison,WI, USA). Procedures were followed according to themanual provided by the manufacturer. The luminescenceof ERE luciferase measured is proportional to the luci-ferase activity of the ER� complex upon treatment. Theluminescence of Renilla luciferase is proportional to theamount of transfected cells.

Normalized luciferase luminescence is calculatedas follows: normalized luciferase luminescence=luminescence (Firefly luciferase)/luminescence (Renillaluciferase). Normalized luciferase activity (% control)=normalized luciferase luminescence (test sample)/normalized luciferase luminescence (untreated control)*100%.

Normalized luciferase activity of untreated control wasexpressed as 100%.

Figure 1 Cell survival of MCF-7 cells following 17�-estradiol treatment for various time intervals. Cells weretreated with different concentrations (0–1000 nM) of 17�-estradiol for 24, 48 and 72 h. The percentagecell survival was measured by MTT assay. Data are presented as means�S.D. of 6 replicate measurements.The percentage survival is expressed relative to control which is defined as 100%. *P

Detection of expression level of ER� by RT-PCR

Cells were seeded at 3�105 cells/well in 6-well plates.Cells were treated with 10 nM 17�-estradiol, 2 µM As2O3or co-treated with both drugs for 24 and 48 h. Controlswere prepared by incubating with experimental mediumonly. After treatment, RNA was isolated using the TRIzolreagent (Invitrogen) according to the procedures suggestedby the manufacturer. cDNA was then synthesized usingSuperscript First-Strand Synthesis System (Invitrogen).In brief, RNA samples of 5 µg were mixed in 0·5 mlmicrotubes with 1 µl 10 mM dNTP mix, 1 µlOligo(dT)12–18 (0·5 µg/µl) in diethy pyrocarbonate(DEPC) treated water to a final volume of 10 µl. Thesamples were incubated at 65 �C for 5 min and then placedon ice for at least 1 min. A reaction mixture composed of1 µl RNase inhibitor, 2 µl 10�RT buffer, 2 µl dithio-threitol (0·1 M), and 4 µl MgCl2 (25 mM) was added toeach sample and mixed gently. The samples were incu-bated at 42 �C for 2 min. One microliter Superscriptreverse transcriptase was added and incubated for 50 minat 42 �C and then for 15 min at 70 �C. The mixture wasthen chilled on ice. Finally, 1 µl RNase H was added andfurther incubated for 20 min at 37 �C. The cDNA samples

were stored at �20 �C until use. PCR was performedwith primers flanking the ER� gene to produce a PCRproduct with a size of 490 base pairs. The sequences of theprimers were as follows: forward 5� CAG GGG TGAAGT GGG GTC TGC TC 3�; reverse 5� ATG CGGAAC CGA GAT GAT GTA GC 3�.

As an internal control, PCR with primers specific forglyceraldehyde phosphate dehydrogenase (GADPH) wascarried out. Samples (10 µl) were mixed with 2 µl6� loading dye and electrophoresed in 1% agarose gelelectrophoresis. The band intensities of PCR productswere analyzed by ImageQuant program (AmershamBiosciences).

Detection of protein expression levels of ER� and c-myc byWestern blot analysis

Cells were seeded at 1�106 cells/well in 100 mm cultureplates and treated with different drugs for 48 h. Treatmentgroups in this assay included a control group with mediumonly, 2 µM As2O3 alone, 10 nM 17�-estradiol aloneand 2 µM As2O3 together with 10 nM 17�-estradiol.After drug treatments, cells were collected and lysed.

Figure 2 Effect of As2O3 on cell survival of MCF-7 cells treated with 10 nM 17�-estradiol, measured byMTT assay. Cells were treated with 10 nM 17�-estradiol alone (E) and various concentrations (0·25–2 �M)of As2O3 simultaneously with 10 nM 17� estradiol (E) for 24, 48 and 72 h. Cells without any treatmentwere used as a negative control for comparison (CTL). The percentage cell survival was measured by MTTassay. Data are presented as means�S.D. of 6 replicate measurements. The percentage survival wasexpressed relative to the untreated control which was defined as 100%. *P

The protein content in each sample was determined byBCA assay (Sigma). Thirty micrograms protein of eachsample were resolved by 10% SDS-PAGE. After electro-blotting, the membrane was probed with anti-ER� anti-body (Oncogene Science, Cambridge, MA, USA) oranti-c-myc antibody (BD Pharmingen, San Diego, CA,USA). Secondary antibody was conjugated with horse-radish peroxidase. Finally, the signal was detected by anenhanced chemiluminescence (ECL) kit (AmershamBiosciences).

Effects of As2O3 on cell cycle distribution of MCF-7 cellsunder estrogen stimulation

Cells at 3�105 cells/well were seeded in 6-well plates andtreated with different drugs for 48 or 72 h. Treatmentgroups in this assay included a control group with mediumonly, 2 µM As2O3 alone, 10 nM 17�-estradiol alone and2 µM As2O3 together with 10 nM 17�-estradiol. Aftertreatment, the cells were washed with PBS and fixed with70% ethanol for at least 30 min. After fixation, cells werewashed with PBS and stained with propidium iodide (PI)solution (2 mg/ml), RNase A (10 mg/ml) and 400 µl PBSfor 30 min at 37 �C under subdued light. Stained cellswere analyzed using a FACSort flow cytometer (BDBiosciences, San Jose, CA, USA). With the CellQuest

program, the cell population was targeted by forward lightscatter (FSC) and side scatter (SSC). The fluorescencesignal of PI was detected at channels of FL-2. Thepercentages of DNA content at different phases of thecell cycle were analyzed with Modfit software (VeritySoftware House, Topsham, ME, USA).

Statistical analyses

Data were expressed as means �standard deviations (S.D.)for three replicate experiments. The Student’s t-test wasused for statistical analyses.

Results

Effect of As2O3 and 17�-estradiol on cell viability of MCF-7cells

The growth stimulatory effects of 17�-estradiol on MCF-7cells in a range of concentrations were determined byMTT assay and direct cell counting. The percentagesurvival of MCF-7 cells increased upon incubation with0·1 nM to 500 nM 17�-estradiol (Fig. 1) compared withthe untreated control. Within 24 h incubation, no obvioussurvival stimulation was observed. It was not until 48 h

Figure 3 Effect of As2O3 on cell survival of MCF-7 cells treated with 10 nM 17�-estradiol, measured by direct cell counting. Cells weretreated with 10 nM 17�-estradiol alone (E) and various concentrations (0·25–2 �M) of As2O3 simultaneously with 10 nM 17�-estradiol(E) for 24, 48 and 72 h. Cells without any treatment were used as negative control for comparison (CTL). The percentage cell survivalwas measured by trypan blue dye exclusion assay. Data are presented as means�S.D. of 4 replicate measurements. *P

incubation that estrogen induced an increase in percentagesurvival from 100% in untreated controls to a maximumof 137% in 10 nM 17�-estradiol-treated cells. Increasingthe incubation time to 72 h further increased the percent-age survival. At this time period, 10 nM 17�-estradiolattained the maximum survival stimulation of 165%with respect to controls. On the other hand, 1000 nM17�-estradiol reduced cell survival from 75% to 49% after24 and 72 h treatment respectively. To assess the effectof As2O3 on 17�-estradiol-treated MCF-7 cells, MCF-7cells were incubated with As2O3 and 17�-estradiolsimultaneously. As seen in Fig. 2, when treated togetherwith As2O3, the percentage survival of MCF-7 cells wasreduced as compared with 17�-estradiol treatmentalone. The reduction was dose- and time-dependent. Forconcentrations between 0·25 and 0·5 µM, As2O3 didnot reduce the percentage survival. At a concentrationof 2 µM As2O3, 25% reduction of survival was inducedafter 24 h treatment, and it was maximized after 72 h

treatment to 52% as compared with the untreated controls(Fig. 3).

Effect of As2O3 on cell survival of the hormone independentbreast cancer cell line, MDA-MB-231

To compare the cell survival inhibiting effects of As2O3 onMCF-7 cells and MDA-MB-231 cells, IC50 after varioustime treatments was obtained and the results are shown inTable 1. Higher IC50 of As2O3 was obtained in MDA-MB-231 cells as compared with MCF-7 cells. In otherwords, As2O3 was more potent in combating cell survivalin MCF-7 cells than in MDA-MB-231 cells.

Effect of As2O3 on estrogen binding to estrogen receptor �(ER�)

In assessing the competitive binding capacity of As2O3 onER�, an ER� competitor screening kit was applied. Thespecific estradiol binding after incubation with variousconcentrations of 17�-estradiol, As2O3, tamoxifen andpaclitaxel is shown in Fig. 4. 17�-Estradiol in concen-trations of 0·1 nM to 200 nM reduced the specific estradiolbinding dramatically in a concentration-dependentmanner. Concentrations above 200 nM completely pre-vented the binding of fluorescent estradiol to ER�.The estradiol binding remained high in the presence ofpaclitaxel indicating that it did not compete with estradiolfor ER� binding. The result was consistent with the factthat paclitaxel inhibited breast cancer survival throughmechanisms other than ER� signaling. When incubatedwith tamoxifen in concentrations above 0·25 µM, theestradiol binding was reduced by more than 50%. Com-pared with As2O3, estradiol binding remained high in

Table 1 The inhibiting efects of As2O3 on cell survival of MCF-7and MDA-MB-231 cells after different treatment times. The resultsare expressed as IC50 of As2O3 (�M)

IC50 of arsenic trioxide (�M)

MCF-7 MDA-MB-231

Treatment time (h)24 8 1748 1·8 772 1·2 4·896 0·8 3·4

120 0·5 2·1

Figure 4 A graph showing the competitive binding ability of 17�-estradiol, As2O3,tamoxifen and paclitaxel to ER�. Data shown are representative of three separateexperiments in each of which triplicate wells were assayed. Error bars are S.D. of triplicatewells. The wells incubated with solvent only are considered as control whose fluorescentintensity is expressed as 100%.





concentrations below 2 µM; higher concentrations did notincrease the ability of As2O3 to compete with estradiol forER� binding. This result suggested that As2O3 did notcompete with 17�-estradiol for ER� binding.

Regulation of ER� mRNA levels following As2O3 treatment

As ER� is the important mediator of the estrogen-stimulated signaling pathway, the alteration of ER�expression level in MCF-7 cells during As2O3 exposuremay contribute to the interference exerted by As2O3 onthe pathway. In this sense, the effect of As2O3 on ER�

RNA levels was examined. Total RNA samples from theuntreated control and after treatment with 2 µM As2O3alone or simultaneously with 17�-estradiol were isolatedand analyzed by RT-PCR. As seen in Fig. 5, in the first24 h treatment, 10 nM 17�-estradiol down-regulatedtranscription levels by 52% relative to controls and furthertreatment to 48 h down-regulated the level by 72%.Twenty-four- and forty-eight-hour exposure to 2 µMAs2O3 also induced a down-regulation of ER� expressionbut to a lesser extent. When MCF-7 cells were incubatedwith 2 µM As2O3 plus 17�-estradiol, a synergistic effectwas observed at both time treatments (Fig. 5).

Figure 5 Regulation of ER� mRNA level in MCF-7 cells following As2O3 treatment. MCF-7cells were treated with 2 �M As2O3 alone and simultaneously with 10 nM 17�-estradiol for24 and 48 h. After treatment, total RNA was isolated, transferred to cDNA and amplifiedwith ER� primers. (A) Agarose gel of the RT-PCR amplified GADPH cDNA stained withethidium bromide. GADPH with a size of 400 bp was used as internal control fornormalization. (B) Agarose gel of the RT-PCR amplified ER� cDNA with a size of 490 bp.(C) Densitometric analysis of the amplified ER� PCR product after normalization withGADPH. Lanes 1–4, 24 h treatment: lane 1, untreated control; lane 2, As2O3 (2 �M); lane3, 17�-estradiol (10 nM); lane 4, 17�-estradiol (10 nM) and As2O3 (2 �M). Lanes 5–8, 48 htreatment: lane 5, untreated control; lane 6, As2O3 (2 �M); lane 7, 17�-estradiol (10 nM);lane 8, 17�-estradiol (10 nM) and As2O3 (2 �M). Data ware representative of threeindependent experiments.

Effect of As2O3 on ER� signaling · S K Y CHOW and others 331

www.endocrinology.org Journal of Endocrinology (2004) 182, 325–337



Regulation of ER� protein levels following As2O3 treatment

In addition to the RNA levels, we have examined theER� protein levels. Within 48 h of treatment, there weredecreases in the ER� protein expression levels of 42% and22% following treatment with 10 nM 17�-estradiol and2 µM As2O3 when compared with controls without treat-ment. Co-treatment of both drugs resulted in a greaterreduction in protein expression levels by 90%. Thisrevealed that administration of 10 nM 17�-estradiol and2 µM As2O3 together down-regulated ER� protein levelswith greater effect than that induced by 17�-estradiolalone. Synergistic down-regulation was observed followingboth treatments together (Fig. 6).

Regulation of ER� transcriptional activity following As2O3treatment

In this section, the regulation of the estrogen activatedtranscriptional activity by As2O3 was studied by examin-ing the effect of As2O3 on ER� binding to ERE. In thestudy, a dual luciferase reporter system was adopted.Renilla luciferase control reporter vector and pGL3reporter vector were co-transfected to MCF-7 cells.Results are expressed as percentage activation. Thenormalized luciferase activity of 10 nM 17�-estradiol,As2O3 at 1 µM and 2 µM concentrations, and As2O3

simultaneously added with 17�-estradiol are shown inFig. 7. The luciferase activity was significantly enhanced3·5-fold by 10 nM 17�-estradiol as compared with thecontrols. Upon treatment with both 1 µM and 2 µMAs2O3, dose- and time-dependent reductions in percent-age activity were observed. When exposed to both17�-estradiol and 2 µM As2O3 over 72 h, the 3·5-foldinduction by 10 nM 17�-estradiol was decreased to28·7%. The results suggested that 2 µM As2O3 exerted aninhibitory effect on ER�-mediated luciferase activationover 48 and 72 h of treatment. In other words, As2O3 at2 µM down-regulated the transcriptional activity of ER�.Most importantly, it also counteracted the transcriptionalactivity of ER� induced by 10 nM 17�-estradiol (Fig. 7).

Regulation of c-myc protein level by As2O3By Western blot analysis, the protein expression of thec-myc gene was examined and the results are shownin Fig. 8. 17�-Estradiol up-regulated c-myc proteinexpression by 65% after 48 h treatment with respect to theuntreated control. The up-regulation was partially sup-pressed by co-treatment with 2 µM As2O3 in MCF-7 cellsover the same period of time, such that the protein levelafter co-treatment remained higher (131%) than that ofcontrols (100%). After treatment with 2 µM As2O3 alone,the protein level of c-myc was down-regulated over 60%as compared with the controls without any treatment.

Effects of As2O3 on cell cycle distribution of MCF-7 cellsunder estrogen stimulation

The effect of As2O3 on cell cycle distribution of MCF-7cells was analyzed after the cells were treated with 2 µMAs2O3, 17�-estradiol, and 17�-estradiol simultaneouslywith 2 µM As2O3. Figure 9(A-D) shows the cell cycledistribution of MCF-7 cells after 48 h treatment. Com-pared with controls, 48 h treatment with 2 µM As2O3markedly increased the proportion of cells in the G1 phasefrom 46·7% to 67·4% while it reduced the proportion ofcells in the S and G2/M phases from 18·5% to 10·2% andfrom 34·6% to 22·4% respectively. Consistent results wereobtained at 72 h treatment, with a greater percentagechange in the phase distribution. Thus, As2O3 induced G1phase arrest in MCF-7 cells by inhibiting cell cycleprogression to the S and G2/M phases. Following incu-bation of MCF-7 cells with the same concentration of17�-estradiol alone, a reduction of 35·5% of the cellpopulation in the G1 phase occurred while G2/M phaseand S phase cell populations were decreased by 9% and79·8% respectively. Following incubation with 2 µMAs2O3 together with 17�-estradiol, an increase in the cellpopulation in the G1 phase and a reduction in the cellpopulation in the G2/M phase and S phase over the same

Figure 6 Regulation of ER� protein level in MCF-7 cells followingAs2O3 treatment for 48 h. MCF-7 cells were treated with 2 �MAs2O3 alone and in combination with 10 nM 17�-estradiol.(A) Western blot analysis of ER�. (B) Relative protein levels ofdifferent samples after quantification. Lane 1, control; lane 2,As2O3 (2 �M); lane 3, 17�-estradiol (10 nM); lane 4, 17�-estradiol(10 nM) and As2O3 (2 �M). Data are the means�S.D. of threeindependent experiments and are expressed relative to controldefined as 100%.





treatment time was also seen. This implied that As2O3induced G1 phase growth arrest of MCF-7 cells stimulatedby 10 nM 17�-estradiol.

Discussion

It has been widely concluded that estrogen exposure is aprominent risk factor in breast cancer and its stimulatoryeffects can potentiate the growth of breast tumor cells(Henderson et al. 1988). In an attempt to treat breastcancer, estrogen withdrawal is one of the strategies used.To see whether As2O3 is able to block estrogen-stimulatedcell growth in MCF-7 cells, the effect of As2O3 inestrogen withdrawal was assessed. Prior to the study,the growth stimulatory effect of estrogen was assessed inwhich 10 nM 17�-estradiol could induce a most dramaticcell growth stimulation. Different concentrations of 17�-estradiol have been used in studies of ER-regulatedpathways or as a model for anti-estrogenicity. The con-centrations range from 0·1 nM to 100 nM. Concentrationshigher than this range induce toxic effects. Here, we

observed that 1 nM 17�-estradiol could stimulate cellgrowth which was at its maximum at a concentration of10 nM. So, 10 nM 17�-estradiol was used in the followingexperiments as a positive control (Fig. 1). With respect tothe untreated controls, the percentage survival of MCF-7cells was reduced from 160% with 10 nM 17�-estradioltreatment alone to 50% in co-treatment of 2 µM As2O3together with 10 nM 17�-estradiol (Fig. 2). In addition,As2O3 was also found to suppress the stimulation of cellgrowth by estradiol in MCF-7 cells (Fig. 3).

The MDA-MB-231 cell line is an estrogen-independent cell line that does not depend on estrogen forgrowth and survival. As2O3 in concentrations below 2 µMwas also shown to suppress the survival of MDA-MB-231following 72 h exposure (Table 1). The lower sensitivity ofMDA-MB-231 cells to As2O3 suggested an association ofestrogen receptor status with growth inhibitory potency.Recent studies also revealed that tamoxifen inducedapoptosis in both ER+ and ER� cell lines via differentmechanisms (Salami & Karami-Tehrani 2003). Only ER+cells could respond to low concentrations of tamoxifen. Itmight be concluded that there exists ER+ and ER�

Figure 7 Regulation of estrogen-activated ER� transcriptional activity following As2O3 treatment for 48 and72 h. MCF-7 cells co-transfected with luciferase reporter plasmid were treated with various concentrationsof As2O3 alone, 10 nM 17�-estradiol alone and 2 �M As2O3 simultaneously with 10 nM 17�-estradiol.Cells without any treatment (control) were prepared for comparison. Each value was normalized by Renillaluciferase control vector. The luciferase activity was expressed as the percentage of untreated controldefined as 100%. The experiment was repeated three times and a single representative experiment isshown. Error bars represent the S.D. of four wells of each sample.





pathways for the induction of apoptosis. Further study isneeded to ascertain the mechanism of action of arsenictrioxide on ER� cells.

The role of ER� in breast cancer is not clearly known.Elucidating the role of ER� in the pathogenesis of breastcancer is important because many human breast tumorsexpress both ER� and ER�. Recent studies showed thatintroducing ER� into MCF-7 cells caused an inhibition ofproliferation in vitro and prevented tumor formation inresponse to estradiol in a mouse xenograft model(Paruthiyil et al. 2004). ER� inhibits the proliferation ofMCF-7 cells by repressing c-myc, cyclin D1, and cyclin Agene transcription, and increases the expression of p21 andp27, which lead to a G2 cell cycle arrest (Paruthiyil et al.2004). These results demonstrated that ER� and ER�produced opposite effects on cell proliferation and tumorformation in MCF-7 cells. Here, we focused on ER� in

elucidating the mechanisms of how As2O3 inhibited breastcancer cell survival. The effects of As2O3 on ER� needfurther investigation.

Anti-estrogens, such as tamoxifen, elicited an estrogenwithdrawal effect mainly by competitive binding to thehormone binding domain of ER� and subsequent altera-tion of the conformation necessary for recruitment oftranscription co-activators to transcription activation func-tion 2 (AF2). Our study of the competitive binding abilityof As2O3 on MCF-7 cells indicated that As2O3 did notcompete with 17�-estradiol for ER binding (Fig. 4). Thespecific ER� binding required hydrogen bond formationbetween an aromatic ring in the ligands and residues in thedomain and a water molecule. The remaining residues inthe binding cavity interacted with a variety of differenthydrophobic groups of ligands (Brzozowski et al. 1997,Pike et al. 2001). The structure of As2O3 shows that themetal elements, arsenic and oxygen, are arranged in a polarring structure. It was believed that the inability of As2O3in binding to the ER� ligand binding site could beattributed to its non-hydrophobic structure.

By RT-PCR and Western blot analysis, both ER�gene transcription and protein expression levels weredown-regulated following 10 nM 17�-estradiol treatment(Figs 5 and 6). Our results are consistent with other studiesin which various concentrations of 17�-estradiol were used(Berthois et al. 1990, Alarid et al. 1999). Some reports haveindicated an association of ER� mRNA levels and proteinlevels after 17�-estradiol and other antiestrogen treatments(Saceda et al. 1988, Santagati et al. 1997). ER� proteindegradation was reported to be dependent on estrogen-induced proteasome-mediated proteolysis instead of DNAtranscription (Alarid et al. 1999). Here, we demonstratedthe down-regulation of ER� mRNA levels followed bythe down-regulation of ER� protein levels following 48 htreatment with As2O3. The mechanism of action remainselusive but there is no doubt that following 2 µM As2O3treatment together with 10 nM 17�-estradiol, both ER�protein levels and mRNA levels were further down-regulated as compared with treatment with 17�-estradiolalone (Figs 5 and 6). The result might be responsible forthe reduced estrogen stimulatory effect in MCF-7 cells byreducing the number of ER� sites available for 17�-estradiol binding and subsequent activation of the ER�signaling pathway.

After binding to ER�, the estrogen–ER� complex willtranslocate to the target DNA binding site called estrogenresponsive element (ERE) in the promoter region of thetarget gene for gene transcription activation (Klinge 2000).

Figure 8 Regulation of c-myc protein levels in MCF-7 cellsfollowing As2O3 treatment for 48 h. MCF-7 cells were treated with2 �M As2O3 alone and in combination with 10 nM 17�-estradiol.(A) Western blot analysis of c-myc protein. (B) Relative proteinlevels of different samples after quantification. Lane 1, 10 nM17�-estradiol; lane 2, untreated control; lane 3, 10 nM17�-estradiol+2�M As2O3; lane 4, 2 �M As2O3.

Figure 9 Regulation of cell cycle distribution of MCF-7 cells following treatment with 2 �M As2O3 and 10 nM 17�-estradiol. MCF-7 cellswere treated with various concentrations of As2O3 alone or together with 10 nM 17�-estradiol for 48 and 72 h. Cell cycle distributionof MCF-7 cells was studied by flow cytometry with propidium iodide (PI) staining. The Y-axis represents the number of events(total=10 000), whereas the X-axis represents the fluorescence intensity of PI (FL2-A). (A–D) Distribution after 48 h treatment;(E–H) distribution after 72 h treatment. (A) and (E) Untreated control; (B) and (F) 2 �M As2O3; (C) and (G) 10 nM 17�-estradiol;(E) and (H) 2 �M As2O3 and 10 nM 17�-estradiol. The data are from a representative experiment of three independent experiments.





Here, we showed the suppressing effect of As2O3 on10 nM 17�-estradiol-stimulated transcription activation byusing a luciferase reporter system containing the EREelement. Following 10 nM 17�-estradiol treatment, theactivation was dramatically enhanced indicating that theaction of the 10 nM 17�-estradiol dose on MCF-7 cellswas mediated by the estrogen–ER� complex binding toDNA and the stimulation of target gene transcription.When treated with 2 µM As2O3 simultaneously with10 nM 17�-estradiol, transcription activation was notenhanced. Instead, it was suppressed twofold and fivefoldover the 48 and 72 h treatment periods respectively, ascompared with that of 10 nM 17�-estradiol treat-ment alone (Fig. 7). Thus, As2O3 at 2 µM suppressed17�-estradiol-induced transcription of ERE bearing targetgenes.

It is of interest to examine the mechanism of howAs2O3 suppressed 17�-estradiol-stimulated cell growthand even elicited growth inhibition in MCF-7 cells. In thepresent study, c-myc protein expression was studied.c-myc is an oncogene responsible for cell growth (Escotet al. 1986, Dang 1999). Previous studies have reportedthat c-myc expression was elevated in estrogen-treatedcells (Dubik et al. 1987, Dubik & Shiu 1988). Moreover,down-regulation of c-myc expression was sufficient toblock 17�-estradiol-stimulated cell growth. This result wasalso observed in our study as c-myc protein expression wasenhanced by 10 nM 17�-estradiol after 48 h treatment. Atthis time, a dramatic increase in the percentage survival ofMCF-7 cells following 10 nM 17�-estradiol treatmentover 48 h was observed (Fig. 8). So, c-myc was responsiblefor the growth stimulation of MCF-7 cells as demonstratedin the previous study. Upon exposure to 2 µM As2O3together with 10 nM 17�-estradiol, the expression level ofc-myc was reduced as compared with that stimulated by10 nM 17�-estradiol alone (Fig. 8). From our study,survival of MCF-7 cells under the same conditions wasreduced to 70% of control. Following a longer incubationto 72 h, As2O3 suppressed cell survival to 48% of control(data not shown). Therefore, it was possible that theblocking of the 17�-estradiol-stimulated cell growth wasassociated with down-regulation of c-myc protein expres-sion. As the promoter region upstream of c-myc geneincluded half the ERE, the down-regulation of c-mycprotein expression might be attributed to the inhibition ofER� transcription activation.

The growth inhibition induced by As2O3 in MCF-7cells was further explored by studying the effects of As2O3on the estrogen-regulated cell cycle. 17�-Estradiol stimu-lated cell growth by accelerating G1–S phase progressionand recruiting cells from the Go phase to enter the cellcycle (Sutherland et al. 1983). Here, we showed that thecell population after 10 nM 17�-estradiol treatmentshowed an increase in numbers in the S phase and adecrease in numbers in the G1 phase, i.e. the stimulatorycell cycle progression in MCF-7 cells (Fig. 9). When

co-treated with 2 µM As2O3 and 17�-estradiol, thecell population accumulated in the G1 phase, indicatingthat 2 µM As2O3 opposed 17�-estradiol-induced cellcycle progression to S phase in MCF-7 cells. Theopposition may be due to the regulation of cell cycleproteins in the G1 and G1/S phases such as down-regulation of cyclin D1 mRNA and protein expression, asup-regulation of cyclin D1 levels were shown to beresponsible for 17�-estradiol-stimulated cell cycle progres-sion (Prall et al. 1997, Charpentier et al. 2000). By cDNAmicroarray analysis, cyclin D1 gene was found to beregulated by estrogen with the ERE sequence in thepromoter regions (Gruvberger et al. 2001). So, As2O3induced cell growth inhibition in MCF-7 cells by G1phase arrest in MCF-7 cells; this might be related to cyclinD1 and awaits further studies.

As estrogen is a primary risk factor of promoting thegrowth of MCF-7 cells and estrogen was also found topromote resistance of chemotherapeutic drugs in MCF-7cells (Teixeira et al. 1995), our finding that As2O3 couldexhibit anti-estrogenicity on MCF-7 cells may shed lighton the therapy of breast cancer in the initial stages of tumordevelopment.

Funding

This study was supported by direct grants from ResearchGrants Council, Hong Kong and the Department ofBiochemistry, The Chinese University of Hong Kong,Hong Kong. There is no conflict of interest that wouldprejudice its impartiality of this paper.

References

Alarid ET, Bakopoulos N & Solodin N 1999 Proteasome-mediatedproteolysis of estrogen receptor: a novel component in autologousdown-regulation. Molecular Endocrinology 13 1522–1534.

Berthois Y, Dong XF, Roux-Dossetto M & Martin PM 1990Expression of estrogen receptor and its messenger ribonucleic acidin the MCF-7 cell line: multiparametric analysis of its processingand regulation by estrogen. Molecular and Cellular Endocrinology 7411–20.

Brunner N, Bronzert D, Vindelov LL, Rygaard K, Spang-Thomsen M& Lippman ME 1989 Effect on growth and cell cycle kinetics ofestradiol and tamoxifen on MCF-7 human breast cancer cells grownin vitro and in nude mice. Cancer Research 49 1515–1520.

Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T,Engstrom O, Ohman L, Greene GL, Gustafsson JA & Carlquist M1997 Molecular basis of agonism and antagonism in the oestrogenreceptor. Nature 389 753–758.

Cai X, Shen YL, Zhu Q, Jia PM, Yu Y, Zhou L, Huang Y, ZhangJW, Xiong SM, Chen SJ, Wang ZY, Chen Z & Chen GQ 2000Arsenic trioxide-induced apoptosis and differentiation are associatedrespectively with mitochondrial transmembrane potential collapseand retinoic acid signaling pathways in acute promyelocyticleukemia. Leukemia 14 262–270.

Castaigne S, Chomienne C, Daniel MT, Ballerini P, Berger R,Fenaux P & Degos L 1990 All-trans retinoic acid as adifferentiation therapy for acute promyelocytic leukemia. I. Clinicalresults. Blood 76 1704–1709.





Charpentier AH, Bednarek AK, Daniel RL, Hawkins KA, Laflin KJ,Gaddis S, MacLeod MC & Aldaz CM 2000 Effects of estrogen onglobal gene expression: identification of novel targets of estrogenaction. Cancer Research 60 5977–5983.

Chen GQ, Zhu J, Shi XG, Ni JH, Zhong HJ, Si GY, Jin XL, TangW, Li XS, Xong SM, Shen ZX, Sun GL, Ma J, Zhang P, ZhangTD, Gazin C, Naoe T, Chen SJ, Wang ZY & Chen Z 1996In vitro studies on cellular and molecular mechanisms of arsenictrioxide (As2O3) in the treatment of acute promyelocytic leukemia:As2O3 induces NB4 cell apoptosis with downregulation of Bcl-2expression and modulation of PML-RAR alpha/PML proteins.Blood 88 1052–1061.

Clarke R, Leonessa F, Welch JN & Skaar TC 2001 Cellular andmolecular pharmacology of antiestrogen action and resistance.Pharmacological Reviews 53 25–71.

Cullen R, Maguire TM, McDermott EW, Hill ADK, O’Higgins NJ& Duffy MJ 2001 Studies on oestrogen receptor-alpha and -betamRNA in breast cancer. European Journal of Cancer 37 1118–1122.

Dang CV 1999 c-Myc target genes involved in cell growth, apoptosis,and metabolism. Molecular and Cellular Biology 19 1–11.

Dickson RB & Lippman ME 1995 Growth factors in breast cancer.Endocrine Reviews 16 559–589.

Dubik D & Shiu RP 1988 Transcriptional regulation of c-myconcogene expression by estrogen in hormone-responsive humanbreast cancer cells. Journal of Biological Chemistry 263 12705–12708.

Dubik D, Dembinski TC & Shiu RP 1987 Stimulation of c-myconcogene expression associated with estrogen-induced proliferationof human breast cancer cells. Cancer Research 47 6517–6521.

Escot C, Theillet C, Lidereau R, Spyratos F, Champeme MH, Gest J& Callahan R 1986 Genetic alteration of the c-myc protooncogene(MYC) in human primary breast carcinomas. PNAS 83 4834–4838.

Gruvberger S, Ringner M, Chen Y, Panavally S, Saal LH, Borg A,Ferno M, Peterson C & Meltzer PS 2001 Estrogen receptor statusin breast cancer is associated with remarkably distinct geneexpression patterns. Cancer Research 61 5979–5984.

Guo W, Zhu J & Wang H 1999 Effects of all-trans retinoic acid,arsenic trioxide and daunorubicin on tissue factor expression inNB4 cells. Zhonghua XueYeXueZa Zhi 20 453–455.

Henderson BE, Ross RK, Pike MC & Casagrande JT 1982Endogenous hormones as a major factor in human cancer. CancerResearch 42 3232–3239.

Henderson BE, Ross R & Bernstein L 1988 Estrogens as a cause ofhuman cancer: the Richard and Hinda Rosenthal Foundationaward lecture. Cancer Research 48 246–253.

Huang ME, Ye YC, Chen SR, Chai JR, Lu JX, Zhoa L, Gu LJ &Wang ZY 1988 Use of all-trans retinoic acid in the treatment ofacute promyelocytic leukemia. Blood 72 567–572.

Klinge CM 2000 Estrogen receptor interaction with co-activators andco-repressors. Steroids 65 227–251.

Kyprianou N, English HF, Davidson NE & Isaacs JT 1991Programmed cell death during regression of the MCF-7 humanbreast cancer following estrogen ablation. Cancer Research 51162–166.

Lippman M, Bolan G & Huff K 1976 The effects of estrogens andantiestrogens on hormone-responsive human breast cancer inlong-term tissue culture. Cancer Research 36 4595–4601.

Mosselman S, Polman J & Dijkema R 1996 ER beta: identificaitonand characterization of a novel human estrogen receptor. FEBSLetters 392 49–53.

Paruthiyil S, Parmar H, Kerekatte V, Cunha GR, Firestone GL &Leitman DC 2004 Estrogen receptor beta inhibits human breastcancer cell proliferation and tumor formation by causing a G2 cellcycle arrest. Cancer Research 1 423–428.

Pike AC, Brzozowski AM, Walton J, Hubbard RE, Thorsell AG, LiYL, Gustafsson JA & Carlquist M 2001 Structural insights into themode of action of a pure antiestrogen. Structure 9 145–153.

Prall OW, Sarcevic B, Musgrove EA, Watts CK & Sutherland RL1997 Estrogen-induced activation of Cdk4 and Cdk2 duringG1-S phase progression is accompanied by increased cyclin D1expression and decreased cyclin-dependent kinase inhibitorassociation with cyclin E-Cdk2. Journal of Biological Chemistry 27210882–10894.

Saceda M, Lippman ME, Chambon P, Lindsey RL, PonglikitmongkolM, Puente M & Martin MB 1988 Regulation of the estrogenreceptor in MCF-7 cells by estradiol. Molecular Endocrinology 21157–1162.

Salami S & Karami-Tehrani F 2003 Biochemical studies of apoptosisinduced by tamoxifen in estrogen receptor positive and negativebreast cancer cell lines. Clinical Biochemistry 36 247–253.

Santagati S, Gianazza E, Agrati P, Vegeto E, Patrone C, Pollio G &Maggi A 1997 Oligonucleotides squelching reveals the mechanismof estrogen receptor autologous down-regulation. MolecularEndocrinology 11 938–949.

Shen ZX, Chen GQ, Ni JH, Li XS, Xiong SM, Qiu QY, Zhu J,Tang W, Sun GL, Yang KQ, Chen Y, Zhou L, Fang ZW, WangYT, Ma J, Zhang P, Zhang TD, Chen SJ, Chen Z & Wang ZY1997 Use of arsenic trioxide (As2O3) in the treatment of acutepromyelocytic leukemia (APL). II. Clinical efficacy andpharmacokinetics in relapsed patients. Blood 89 3354–3360.

Soignet SL, Maslak P, Wang ZG, Jhanwar S, Calleja E, Dardashti LJ,Corso D, DeBlasio A, Gabrilove J, Scheinberg DA, Pandolfi PP &Warrell RP Jr 1998 Complete remission after treatment of acutepromyelocytic leukemia with arsenic trioxide. New England Journalof Medicine 339 1341–1348.

Sutherland RL, Reddel RR & Green MD 1983 Effects of oestrogenson cell proliferation and cell cycle kinetics. A hypothesis on the cellcycle effects of antioestrogens. European Journal of Cancer and ClinicalOncology 19 307–318.

Teixeira C, Reed JC & Pratt MA 1995 Estrogen promoteschemotherapeutic drug resistance by a mechanism involving Bcl-2proto-oncogene expression in human breast cancer cells. CancerResearch 55 3902–3907.

Wang ZY 2001 Arsenic compounds as anticancer agents. CancerChemotherapy and Pharmacology 48 (Suppl 1) S72–S76.

Warrell RP Jr, de The H, Wang ZY & Degos L 1993 Acutepromyelocytic leukemia. New England Journal of Medicine 329177–189.

Received 7 April 2004Accepted 7 May 2004





Documents

Suppression of cell proliferation and regulation of estrogen ...540 nm was measured. Percentage survival was deﬁned as: % survival=100% (O.D. of test sample/O.D. of control). Direct