[CANCER RESEARCH 58. 4383-4390. October I. 1998]

Influence of Melatonin on Invasive and Metastatic Properties of MCF-7 HumanBreast Cancer Cells1

Samuel Cos,2 Rosario Fernández, AndrésGüézmes,and Emilio J. Sánchez-Barceló

Department of Physiology anil Pharmacology. School of Medicine, University of Cantabria. 39011 Saniander /S.C. R. F.. E.J.S-B./. ami Palhnliigical Analiimy Sen'ice,"Sicrrallana " Hospital, 39300 Torrelavega ¡A.G. f. Cantabria. Sftiiin

ABSTRACT

Melatonin, the principal pineal gland hormone, exerts a direct antipro-liferative effect on estrogen-responsive MCF-7 cells in culture. The pur

pose of the current study was to investigate the effects of melatonin on theinvasion capacity of MCF-7 cells.

In vitro, melatonin at physiological doses ( 1 mil reduced (/' < 0.001 ) the

invasiveness of tumoral cells measured in Falcon invasion chambers.Subphysiological (0.1 p\i) and pharmacological concentrations (10 /t\n ofmelatonin failed to inhibit cell invasion. Melatonin was also able to block170-estradiol-induced invasion !/' < 0.001). Pretreatment of MCF-7 cells

with 1 mi melatonin increased the response of tumoral cells to theanti-invasive effects of this indolamine. To explore possible mechanisms

by which melatonin reduces invasiveness, we measured the attachment ofMCF-7 cells to a basement membrane, the chemotactic response of the

cells, and their type IV collagenolytic activity. The presence of melatonin(l n\t) in the culture medium significantly reduced the ability of MCF-7

cells to attach to the basement membrane; this effect was enhanced bypretreating the cells with the same indolamine (/' < 0.001). Melatonin also

counteracts the stimulatory effects of 17ß-estradiol on cell adhesion(P < 0.001). The chemotactic response of MCF-7 cells also decreased inthe presence of l UMmelatonin, and this melatonin-induced reduction of

cell migration was more effective on cells that were previously incubatedfor 5 days with melatonin than it was on nonpretreated cells (/' < 0.001).

The simultaneous addition of 17ß-estradiol and melatonin resulted in asignificantly lower chemotactic response than that of 17ß-estradiol-

treated cells (P < 0.001). However, type IV collagenolytic activity was notinfluenced by melatonin. Our results demonstrate that melatonin reducesthe invasiveness of MCF-7 cells, causing a decrease in cell attachment andcell motility, probably by interacting with the estrogen-mediated mechanisms of MCF-7 cell invasiveness. In addition, we also studied the influ

ence of melatonin on the expression of two cell surface adhesion molecules(E-cadherin and /!, integrin) and an intermediate filament protein (vi-

mentin), the expression of which has been correlated with the relativeinvasive capacity of human breast cancer cells. The culture of tumor cellsin the presence of melatonin (1 nM) increased the membrane staining forE-cadherin and /!, integrin as well as the number of E-cadherin and /!,integrin immunoreactive cells (P < 0.01). Neither control MCF-7 cells nor

those treated with melatonin stained for vimentin.Preliminary in vivo experiments carried out on ovariectomized athymic

nude mice implanted with 170-estradiol pellets and inoculated with5 x 10'' MCF-7 cells in the inguinal mammary fat pad suggest that

melatonin could decrease the tumorigenicity of these tumor cells. However, these results need further confirmation.

Taken together, our results suggest that melatonin shifts MCF-7 human

breast cancer cells to a lower invasive status by increasing the /!, integrinsubunit and E-cadherin expression and promoting the differentiation oftumor cells. Finally, our study points out the existence of the anti-invasive

actions of melatonin as a part of the oncostatic action of melatonin.

Received 4/7/98; accepted 8/3/98.The costs of publication of this 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.

1Supported by Grams PB92-0831 and PM97-0042 from the Spanish General Direction

of Scientific and Technological Investigation (DGICYT) and Ihe Marquésde ValdecillaFoundation 6/98.

2 To whom requests for reprints should be addressed, at Department of Physiology and

Pharmacology. School of Medicine. University of Cantabria. Cardenal Herrera Oria s/n.39011 Santander. Spain.

INTRODUCTION

The role of the pineal gland as an oncostatic gland has beenstudied in different animal models of tumorigenesis. especially inthose concerning the mammary gland. The most common conclusion is that either experimental manipulations that activate thepineal gland or the administration of melatonin, the main pinealhormone, reduces the incidence and development of chemicallyinduced mammary tumors, whereas pinealectomy usually stimulates breast cancer growth (1-4). Two different mechanisms have

been proposed to explain how melatonin could reduce the development of mammary tumors: (a) indirect neuroendocrine mechanisms such as the melatonin regulation of some pituitary andgonadal hormones that control tumor growth (I, 3-5); and (b)modulation by melatonin of the immune response to the presenceof a malignant neoplasm (6) and the action of melatonin as anendogenous hydroxyl radical scavenger (7, 8); on the other hand,direct antiestrogenic melatonin actions at the cellular level havebeen proposed (9, 10). In vitro, concentrations of melatonin (1 nMand 10 pM) corresponding to the physiological levels present inhuman blood during the night exert a direct antiproliferative effecton estrogen-responsive MCF-7 cells, including decreases in cellnumber. DNA content, and thymidine incorporation (9, 11, 12). Todate, all of the studies (above all, in vitro studies carried out onMCF-7 cells) have focused on the antiproliferative actions ofmelatonin (10. 13-16), but some other possible mechanisms involved in the oncostatic properties of this hormone have not beenexplored. The MCF-7 breast carcinoma cell line, which was established from the pleural effusion of a patient with breast adenocar-cinoma (17), contains estrogen receptors and requires estrogens toexpress a great malignant character (18, 19). It is well known that17/3-estradiol stimulates the invasive and metastatic potential ofcancer cells, increasing the ability of MCF-7 cells to both formtumors and produce distant metastasis in nude mice (18, 20), andenhancing the ability of these cells to invade through an artificialreconstituted basement membrane in vitro (19). Because the anti-tumor action of melatonin is partially exerted by an interactionwith 17ß-estradiol(9, 10), we postulate that melatonin could alsoregulate the metastatic behavior of MCF-7 human breast cancercells. To test our hypothesis, we studied: (a) the effects of melatonin on the invasion capacity of MCF-7 cells in vitro by evaluating its influence on cell invasion, cell adhesion, chemotaxis, andtype IV collagenolytic activity of MCF-7 cells, which are associated with the processes involved in the metastatic behavior ofMCF-7 cells; and (b) the effects of melatonin on the tumorigenicityand metastasis formation of MCF-7 cells in \'ivo. Furthermore, we

also investigated the influence of melatonin on the expression oftwo cell surface adhesion molecules. E-cadherin (a calcium-dependent membrane protein responsible for cell-cell contact) and ß,integrin (a subunit of integrins, which are receptors that regulatethe interaction between cells and the extracellular matrix), as wellas on the expression of vimentin (an intermediate filament proteinwhose expression has been previously related with the relativeinvasiveness of human breast cancer cells: Refs. 21-24).

4383

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

Fig. 1. Schematic represemalion of the chambers used in the invasion and chemotaxis assays. A modified Boyden's chamber was con

structed by using Falcon mulliwell cell culture plates (Ai and removableFalcon cell culture inserts (ß)When the inserts are placed into the well,two compartments, which are separated by a microporous filter, aredefined (O. The filter was coated with Matrigel (invasion assay) orcollagen IV (chemotaxis assay). In both types of assay, the cells and thesubstances tested were placed in the upper compartment, whereasfibronectin was placed in the lower compartment as a chemoattractant.Cells that traversed the reconstituted basement membrane (invasionstudies) or the collagen IV layer (chemotaxis assays) crossed the mi-

cropore filter, adhered to the lower surface of the filter, and were fixed,stained, and counted •¿�/•'<

Cell Culture Insert

UpperCompartment

LowerCompartment

Plate Well

Chemoattractant (fibronectin)Reconstituted Basement Membrane

(Matrigel) or Collagen IVInvasive Cells

MATERIALS AND METHODS

Cells and Culture Conditions. MCF-7 human breast cancer cells were

purchased from the American Type Culture Collection (Rockville. MD) andmaintained as monolayer cultures in 75-cm' plastic culture flasks in DMEM

(Sigma Chemical Co., St. Louis, MO) supplemented with 10% FBS' (Life

Technologies. Inc.. Eragny. France), penicillin (20 units/ml), and streptomycin(20 fig/ml: Sigma Chemical Co.) at 37°Cin a humid atmosphere containing

5% CO:. Cells were subcultured every 3-4 days by suspension in 5 minNa2-EDTA in PBS (pH 7.4) at 37°Cfor 5 min.

Before each experiment, stock subconfluent monolayers (80%) of MCF-7cells were incubated with 5 mM NarEDTA in PBS (pH 7.4) at 37°Cfor 5 min.

resuspended in DMEM supplemented with 10% FBS. and passed repeatedlythrough a 25-gauge needle to produce a single cell suspension. Cell number

and viability were determined by staining a small volume of cell suspensionwith 0.4% trypan blue saline solution and examining the cells in a hemocy-

tometer.Invasion Assay. Invasion assays were carried out by methods based on

others described previously (25-27) in modified Boyden's chambers con

structed with Falcon multiwell cell culture plates and Falcon cell culture inserts(see Fig. 1). Cell culture inserts were converted into invasion chambers byapplying a layer of basement membrane onto the surface of microporous filterspresent in each unit. Briefly. 6.4-mm-diameter filters (8 /im pore) of Falcon

cell culture inserts were coated with 25 /ig/filter of reconstituted basementmembrane Matrigel (Collaborative Biomedicai Products. Bedford. MA). Ma-trigel was diluted to the desired concentration with ice-cold distilled water,

applied to filters, dried overnight at room temperature, and reconstituted withDMEM for 90 min at room temperature. Uniformity of the coating waschecked by Coomassie Blue staining and low-power microscopic observation.

The concentration of Matrigel used in the experiments (25 /ig/filter) wasdetermined in a previous study in which we varied the amount of Matrigelplaced on the filters (0, 6.25, 12.50, 25, and 50 fig/filter) until finding theconcentration that allowed a discriminating assay (data not shown).

Exponentially growing MCF-7 cells were harvested with 5 mM Na,-EDTA

in PBS (pH 7.4); washed in DMEM + 10% FBS by centrifugaron: resuspended in DMEM supplemented with 10% FBS, penicillin (20 units/ml), andstreptomycin (20 fig/ml); and passed repeatedly through a 25-gauge needle to

produce a single cell suspension. After determination of the cell count andviability in a hemocytometer by the trypan blue exclusion test, the cells wereadded to the upper compartment of the modified Boyden's chamber ( 1.5 x IO5

cells/chamber: see Fig. 1C). Fibronectin (16 /¿g/chamber) was placed in thelower compartment as a chemoattractant. At the beginning of the assay,melatonin (Sigma Chemical Co.; 10 /¿M.I nM. or 0.1 pM). 17ß-estradiol

(Sigma Chemical Co.; l /J.M or 10 nM). and/or the diluent (final ethanolconcentration/plate. 0.0001%) were added to the upper compartment. In someexperiments, the cells were preincubated for 5 days with melatonin (10 /AM,1nM. or 0.1 pM) before being tested for invasiveness. After incubation for 5 days

1The abbreviation used is: FBS. fetal bovine serum.

at 37°Cin 5% CO2. the cells on the upper surface of the filter were completely

removed by wiping with a cotton swab, and the cells that had traversed theMatrigel and attached to the lower surface of the filter were fixed, stained withH&E, and counted in 15 randomly selected microscopic fields (X400) perfilter. Experiments were performed three times with three dishes for eachexperimental condition each time.

Cell Adhesion Assay. The 24-well tissue culture plates were coated with

25 /Ltg/well Matrigel and left to air dry in a hood overnight. The concentrationof Matrigel used in the experiments was determined in a previous study inwhich we varied the amount of reconstituted basement membrane placed onthe plates. To block nonspecific binding sites, all wells were also incubatedwith DMEM containing 0.1% BSA for l h at 37°Cand then washed with the

same medium. MCF-7 cells were suspended at 150,000 cells/ml in DMEMsupplemented with 0.1% BSA. incubated at 37°Cfor I h to allow the restitu

tion of surface proteins, and added to each well in the presence of melatonin(10 /IM. 1 nM, or 0.1 pM), 17ß-estradiol (l /IM or 10 nM), or the hormone

diluent. In some experiments, cells were preincubated for 5 days with melatonin ( 10 fiM, 1 nM. or 0.1 pM) before the adhesion assay. Aliquots (1 ml) ofthe tumor cell suspension were seeded into the Matrigel-coated wells andincubated for 10. 30, or 60 min at 37°Cin 5% CO2. At the end of these periods,

the wells were washed gently three times with PBS to remove the unattachedcells, whereas the attached cells were harvested and counted.

Chemotaxis Assay. Chemotactic assays were performed as described forthe chemoinvasion studies, except that the filter surfaces were coated with 5/xg/filter collagen IV (Collaborative Biomedicai Products) instead of Matrigel(Fig. I ). This coats the interstices of the filter but does not form a barrier overthe surface. The cells are thus free to migrate toward the chemoattractantwithout first having to degrade a barrier. Fibronectin was placed in the lowercompartment as the chemoattractant. In a preliminary study, we compared theeffects of two different concentrations of fibronectin (8 or 16 /ig) to establishthe optimal concentration to be used in the experiments. The greater theconcentration of fibronectin, the larger the number of cells that attach to thelower surface of the filter. Because MCF-7 cells are poorly invasive, we chose

the higher concentration (16 /ig) of chemoattractant. which allows enoughmigration of a number of cells to be easily measured by low-power micro

scopic observation. At the beginning of the assay, melatonin ( 10 /AM,1 nM, or0.1 pM), 17/3-estradiol (l /J.M or 10 nM), and/or the diluent (final ethanol

concentration/plate, 0.0001%) were added to the upper compartment. In someexperiments, cells were preincubated for 5 days with melatonin (10 /IM, l nM,or 0.1 pM) before the chemotaxis assay.

Type IV Collagenolysis Assay. Type IV collagenolytic activity of MCF-7

human breast cancer cells was measured by the methods described previously(28, 29), using 'H-labeled type IV collagen (DuPont New England Nuclear,

Boston. MA) as a tracer.To evaluate the type IV collagenolytic activity of MCF-7 cells, aliquots of

'H-labeled type IV collagen solution (5 /ig; 9,000 cpm) in 0.01 N acetic acid

were placed in each well of 24 multiwell tissue culture plates and left in thelaminar air flow hood at room temperature without light overnight to allow thecollagen solutions to dry to films. MCF-7 cells (150,000 cells) in DMEM

4384

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

supplemented with 10% FBS. penicillin (20 units/ml), and streptomycin (20 inoculated directly into the inguinal mammary fat pad of the mouse by means/j.g/ml) were placed in each well and incubated at 37°C in a humidified

atmosphere (957r air:5% CO2) for 5 days in the presence or absence of 1 nMmelatonin. The incubation was terminated by chilling, and the culture supernatant was withdrawn and mixed with 50 /nl of ice-cold 10% trichloroaceticacid and 0.5% tannic acid in a microcentrifuge tube. After a 30-min incubationat 4°C.the mixture was centrifuged at 10.000 x g for 10 min at 4°Cto

precipitate the undigested materials. The supernatant was withdrawn, and the'H activity was counted.

To measure the type IV collagenolytic activity in tumor cell-conditionedmedia. MCF-7 cells were cultured for 5 days in the presence or absence of 1

nM melatonin in the Falcon invasion chamber. After this period, the conditioned media were harvested by centrifugation. Aliquots of the culture supernatant (100 /il) were treated with trypsin (5 /¿g/ml) at 37°Cfor 10 min.

followed by the addition of soybean trypsin inhibitor (40 ¿ig/ml).Aliquots (100/nl) of the trypsinized or nontrypsinized conditioned media were incubatedwith the radioactive type IV collagen (5 ¿tg/9000cpm/tube) in 100 /il of 10mM CaCI2, 0.1 M NaCl, and 50 mM Tris-HCI buffer (pH 7.5) at 37°Cin 5%

CO, for 24 h. The undigested materials were precipitated with 10% trichloroacetic acid and 0.5% tannic acid and then centrifuged. and the radioactivity

in the supernatant was measured in a ßscintillation counter.Immunohistochemical Study of E-Cadherin, ß,Integrin, and Vimentin.

Exponentially growing MCF-7 cells were plated on glass coverslips in DMEM

supplemented with 10% FBS, penicillin (20 units/ml), and streptomycin (20/¿g/ml)for 24 h. To achieve a partial synchronization of the cell cycle, theculture medium was then replaced by fresh supplemented DMEM containing2 mM thymidine. Twenty-four h later, the cells were released from the thymi-

dine block by rinsing the plate twice with DMEM and allowed to grow in

supplemented DMEM containing either 1 nM melatonin or the melatonindiluent (ethanol; final concentration/plate. 0.0001%). At 6. 12, and 24 h afterthe release from thymidine block, the culture media were aspirated, and thecells were washed in PBS and fixed for 10 min with 3.7% paratbrmaldehydein PBS (pH 7.4) at room temperature. Cells were then incubated with thecorresponding primary antibodies: (a) anti-E-cadherin (mouse monoclonalantibody to L-CAM/uvomorulin: Boehringer Mannheim. Indianapolis, IN)

used at a 1:20 dilution; (b) rat monoclonal antibody reactive with the ß,subunit (CD29) common to all members of the ß,integrin family (BectonDickinson. Bedford. MA) used at a 1:200 dilution: or (c) mouse antivimentin(Boehringer Mannheim) used at a 1:10 dilution. Incubations with the primaryantibodies were performed either overnight at 4°C(E-cadherin and ß,integrin)

or for 1 h at room temperature (vimentin). After washing in PBS. coverslipswere incubated for 60 min at room temperature with antimouse FITC-conju-

gated secondary antibody (Jackson ImmunoResearch Laboratories. WestGrove. PA) at a 1:100 dilution and mounted in Vectashield mounting medium(Vector Laboratories. Burlingame. CA). Samples were studied with a confocallaser microscope (Bio-Rad 1024) using an argon ion (488 nm) to excite FITC.

A semiquantitative estimation based on the relative amount of E-cadherin.

vimentin, and/or j3, integrin expressed by cells and a quantification of thenumber of immunoreactive cells were performed. The number of immunore-

active cells was counted in 15 randomly selected microscopic fields (X400)per sample.

In Vivo Tumorigenesis and Metastasis. Homozygous athymic nude mice(nu+/nu +: BALB/c strain) were obtained from Iffa-Credo (Barcelona, Spain).

Animals were housed and maintained under pathogen-limited conditions in

filtered laminar air flow hoods in standard vinyl cages with air filter tops at2I°Cto 25°Cwith a photoperiod of 14 h of light and 10 h of darkness. Cages

and bedding were autoclaved before use. Food and water were autoclaved andprovided ad lihitum. Six-week-old mice were used to initiate the experiments.

Mice were bilaterally ovariectomized by standard surgical techniques 1 weekbefore cell inoculation, and estrogen supplementation was given in the form of17/3-estradiol slow-release s.c. pellets (90-day release: 0.72 mg/pellet: Inno

vative Research, Sarasota, FL) placed in the interscapular region at the time ofcell inoculation.

Cells from subconfluent monolayers of MCF-7 cells were harvested, sus

pended in DMEM supplemented with 10% FBS. washed in DMEM bycentrifugation. resuspended in DMEM. and passed repeatedly through a 25-

gauge needle to produce a single cell suspension. Once the cell number andviability were determined. 5 X IO6 cells in 0.1 ml of culture medium were

of a syringe and a 25-gauge needle.

Melatonin administration (5 /ig/g body weight/day) dissolved in drinkingwater was begun at the time of tumor inoculation. Controls were givendrinking water containing 0.05% ethanol. There were 10 animals/group.

Tumor size was measured weekly and at autopsy. Mice were sacrificed onor before week 10 after inoculation, depending on physical condition. Somemice were lost due to unexpected deaths and could not be analyzed. Thevolume of the tumor was calculated according to the following formula:

Tumor volume =(width)1 X length

At autopsy, mice were sacrificed by ether anesthesia, and brain, lymph nodes,heart, adrenal glands, muscle (chest wall), bone (ribs), kidney, liver, lung,spleen, and all proliferating tumors were removed postmortem and fixed in10% formalin for histological examination. The embedding, mounting, andstaining of tissues with H&E were performed in the standard way.

Statistical Analysis. Data were analyzed by a one-way ANOVA followedby the Student-Newman-Keuls multiple comparisons test. Tumoral incidence

in mice and the percentage of immunoreactive cells in the immunohistochem-ical study were analyzed by a nonparametric ANOVA (Mann-Whitney test).Differences among the groups' means were considered significant at P < 0.05.

RESULTS

Effects of Melatonin on the Invasive Activity of MCF-7 Cells.When melatonin at physiological concentrations ( I nM) was added tothe upper compartment of the invasion chamber, the number of cellsthat invaded the Matrigel membrane decreased significantly (—44%;P < 0.001) as compared with that of the control group. Subphysi-

ological (0.1 pin) or supraphysiological (10 /MM)concentrations ofmelatonin lacked this effect (Fig. 2A). Pretreatment of MCF-7 cells

with 1 nM melatonin for 5 days before the invasion assay increased(Fig. 2B) the response to the anti-invasive effects of this indolamine(P < 0.001). The highest anti-invasion effects were found whenMCF-7 cells were first incubated for 5 days in the presence of 1 nM

melatonin in culture media, and then the same concentration ofmelatonin was added to the upper compartment of the invasionchamber. In this way, a reduction of up to 14% of the invasiveness ofMCF-7 cells was observed.

IoB'S

B120

p 10°I 80-

8 60o 4085 20-•¿�Ta b

Melatonin Melatonin Melatonin10,,M 1 nM 0.1 pM

Fig. 2. Effect of melatonin on the invasiveness of MCF-7 cells. A. exponentiallygrowing MCF-7 cells were harvested and counted, and 1.5 X IO' cells were added to each

invasion chamber in the presence of melatonin (IO /AM.1 nM, or O.I pM) or the diluent(elhanol; final concentration. O.OOOI'fr). B. cells were preincuhaled for 5 days with

melatonin (10 ¿IM,1 RM, or O.I pM) before being placed in the invasion chamber andexposed to the same concentration of melatonin. Data are expressed as the percentage ofthe control response (mean ±SE), a. P < 0.001 versus control; b. P < 0.001 versus othermelatonin treatments.

4385

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

2§

Control Melatonin Estradici1 nM 1 „¿�M

Estradici10 nM

Estradici Estradici1 MM 10nM

Melatonin Melatonin1 nM 1nM

Fig. 3. Effect of melatonm and I7ß-estradiol un Ihe invasiveness of MCF-7 cells.MCF-7 cells were placed in Ihe upper compartment of Ihe invasion chambers ( 1.5 X IO5

cells/chamber) in the presence of mcialonin ( IO JÃŒM.I nw.orO.I p\i) and/or !70-es(radiol

(1 /IM or 10 nM). Data are expressed as the percentage of the control response(mean ±SE), a, P < 0.001 versus control; b, P < 0.001 versus l U.M 17/3-eslradiol: c.P < 0.001 versus 10 nM 17/3-estradiol.

The addition of 17ß-estradiol(l /UMor 10 HM)to the upper com

partment of the invasion ehamber (Fig. 3) increased the number ofcells invading the Matrigel membrane, and tnelatonin ( 1 nM) was ableto counteract the stimulatory effects of 17ß-estradiol when both

hormones were added simultaneously (P < 0.001).Effects of Melatonin on the Attachment of MCF-7 Cells to the

Basement Membrane. To evaluate the influence of melatonin on theattachment of MCF-7 cells to the basement membrane, tissue culture

plates were coated or not coated with an extract of basement membrane components (Matrigel). and cells were added to the plates andincubated for 10, 30, or 60 min in the presence of different concentrations of melatonin, 17ß-estradiol, or the hormone diluent. The

number of cells that adhered to the wells was then counted. Asexpected, the nonspecific attachment of MCF-7 cells to Matrigel-

coated wells was higher than that to the uncoated wells (data notshown). Melatonin did not influence the adhesiveness of cells touncoated tissue culture plates, whereas it reduced the adhesion ofcancer cells to the Matrigel basement membrane (P < 0.001). Thisreduction was greater at 1 nM melatonin than it was at 10 /AMor 0.1pM (Fig. 44). The melatonin-induced reduction of the adhesiveness ofMCF-7 cells was significantly higher in cells pretreated with the

indolamine (10 /¿M,1 nM, or 0.1 pM) than it was in nonpretreated cells(P < 0.001; Fig. 4ß). 17ß-Estradiol (I /XMor 10 nM) increased the

adhesion of cells to the Matrigel membrane. The simultaneous addition of 17/3-estradiol and melatonin resulted in a significantly lower(P < 0.001) cell adhesiveness than did the addition of 17/3-estradiol

alone and was similar to that obtained with melatonin alone (Fig. 5).Effects of Melatonin on the Chemotactic Migration of MCF-7

Cells. The introduction of fibronectin as a chemoattractant into thelower compartment of the invasion chamber sped up, in a linearmanner, the movement of the cells from the upper compartment to thelower compartment through the collagen IV membrane separatingboth. This chcmotactic response of MCF-7 cells was inhibited by the

addition of 1 nM melatonin: other concentrations of this indolaminewere ineffective (Fig. 6/1). The actions of melatonin in lowering cellmigration were stronger in cells that were previously incubated withI nM melatonin than they were in nonpretreated cells (P < 0.00! ; Fig.6ß). 17ß-Estradiol (l JAMor 10 nM) stimulated the chemotacticmigration of MCF-7 cells. The simultaneous addition of 17ß-estradiol

and melatonin resulted in a significantly (P < 0.001) lower chcmotactic responsiveness than that of the 17ß-estradiol-treated cells and

was similar to the chemotactic migration observed in control (untreated) cells (Fig. 7).

Effects of Melatonin on Type IV Collagenolytic Activity. Neither the type IV collagenolytic activity of MCF-7 cells nor the activity

of type IV collagenolytic enzymes in the culture media conditioned byMCF-7 cells was significantly different in melatonin-treated andcontrol MCF-7 cells (data not shown).

Effects of Melatonin on Cell Surface Adhesion Molecules (E-Cadherin and /3, Integrin) and Vimentin. E-Cadherin expressionwas detected in MCF-7 human breast cancer cells (Fig. 8,4). Melato

nin (1 nM) increased this expression, as can be seen in the strongstaining of cell membrane and cell-cell junctions in these cells (Fig.8ß).The number of E-cadherin-immunoreactive cells also increasedsignificantly in melatonin-treated cells (P < 0.01; Fig. 8C).

Control

Melatonin 10 (iM

CZI Melatonin 1 nM

[33 Melatonin 0.1 pM

oo

o

sra

0)O

B

oSmjaoO

10 Min 30 Min 60 Min

Fig. 4. Effect of melatonin on the adhesion of MCF-7 cells to the basement membrane.A, 1.5 X 10s cells were added io plates coated with Matrigel (25 /zg) and incubated for

IO. 30. or 60 min in the presence of 10 ¿IM,l nM. or O.l pM melatonin. B, cells werepreincuhated for 5 days with melatonin (IO ¿IM,I nM, or O.I pM) before being added toplates coated wilh Matrigel and exposed to the same concentrations of melatonin for IO,30, or 60 min. Attached cells were counted. Data are expressed as mean ±SE. a, P < 0.05versus control; b, P < O.OOl versus control: c, P < 0.05 versus other melatonintreatments: d, P < O.OOl irr.v//.v other melatonin treatments.

120

g 100

x. 80

l 60U

I 40tS 20-

0

•¿�Control _ Estradici 1(iM ,—,Estradici 10 nMCU Melatonin 1nM Melatonin 1nM Melatonin 1nM

CU Estradici 1i.M •¿�EstradioMOnM •¿�

I10 min 30 min 60 min

Fig. 5. Effect of nielalonin and I7ß-csm\diol on the adhesion of MCF-7 human breastcancer cells to the hasement membrane. Cells (1.5 X IO5) were added to plates coated

with Matrigel (25 (¿g>and incubated for 10. 30, or 60 min in the presence of nielalonin(10 /AM, 1 nM, or 0.1 pM) and/or I7ß-estradiol (I /AM or 10 nM). Attached cells were

counted. Data are expressed as mean ±SE. u. P < 0.05 versus control; b. P < 0.05 versusI JIM 17ß-estradiol: c, P < 0.05 versus 10 nM I7ß-estradiol.

4386

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

120 n

Oo'Sss

B

ou•¿�sss

Melatonln0.1 pM

Fig. 6. Effect of melatonin on the chemolactic migration of MCF-7 cells. A. cells( 1.5 X 105 cells/chamber) were placed in the upper compartment of the invasion chambers

in the presence or absence of melatonin ( l O¡UM,l nM.or 0.1 pM) and left to migrate towardthe lower compartment of the invasion. B. cells were preincubated for 5 days withmelatonin ( 10 ¡ÕM.I nM, or 0.1 pM) before being placed in the chamber and exposed to thesame concentration of melatonin during the chemotaxis assay. Data are expressed as thepercentage of the vehicle-treated controls (mean ±SE), a, P < O.OOl versus control; h.P < 0.05 versus control: c. P < O.OOl versus other melatonin treatments.

We also studied the cell surface expression of the ß,subunit of theintegrins. As shown in Fig. 8D. the ß,subunit was expressed onMCF-7 cells. The culture of tumor cells in the presence of melatonin

increased the expression of these cell surface adhesion molecules (Fig.8£)and the number of ß,integrin-immunoreactive cells (P < 0.005;

Fig. 8F).Neither the control MCF-7 cells nor those treated with melatonin

stained for vimentin.Influence of Melatonin on the Tumorigenicity and Metastasis

Formation of MCF-7 Cells in Nude Mice. Ten weeks after MCF-7cell inoculation into the inguinal fat pad of ovariectomized athymicnude mice implanted with I7ß-estradiolslow-release s.c. pellets, 75%

of the intact athymic mice had tumors of more than 3 mm in diameter.Oral melatonin reduced tumor formation, and only 20% of the mela-tonin-treated mice developed tumors. Melatonin also reduced the size

of tumors in comparison with those of control animals. Of the intactmice killed 10 weeks after cell inoculation, 34% had métastasesin thelungs, liver, and lymph nodes. In contrast, only 20% of the melatonin-treated animals developed distant metastasis. A total of 67% of theathymic nude mice inoculated with MCF-7 cells died within 10 weeks

after cell inoculation. In contrast, the mice that were given oralmelatonin survived longer than control mice, and only 44% of theanimals died within 10 weeks of the experiment. All of these in vivoresults represent preliminary data because the number of animals thatdied throughout the experimental period was higher than expected;consequently, the small number of mice that survived at the end of theexperiment makes a correct analysis difficult.

DISCUSSION

effects of melatonin, the current hypothesis is that the oncostaticactions of melatonin are mediated via its effects on the tumor cells'

estrogen-response pathway (IO, 11, 13). The link between the anti-proliferative effect of melatonin on the growth of MCF-7 cells and theestrogen-response pathway is further supported by: (a) the ability ofmelatonin to block the mitogenic effect of 17ß-estradiol (9, 10) in

different culture systems (monolayer and clonogenic soft agar); (b)the melatonin blockade of the estrogen rescue of tamoxifen-inhibitedcells in clonogenic agar and monolayer culture (10); (c) the down-regulation by melatonin of estrogen receptor expression in MCF-7cells (11, 30); and (d) melatonin modulation of estrogen-regulatedproteins, growth factors, and proto-oncogenes in human breast cancer

cells (31, 32). This concept is also supported by a study in whichwomen with estrogen receptor-positive breast cancer were found to

have decreased nocturnal plasma levels of melatonin when comparedwith women with estrogen receptor-negative breast tumors and withage-matched controls (33). Because human breast cancer cells show a

pleiotropic response to estrogens (34, 35), and, as noted previously,the malignant character of these cells increases with estrogens, weconsidered it of interest to study whether melatonin may or may notmodify the metastatic behavior of breast cancer cells.

The present study demonstrates that melatonin at physiologicaldoses ( 1 nM) reduces the invasiveness of tumor cells in vitro, and thisanti-invasive effect is increased when cells are pretreated with the

same indolamine. On the contrary, subphysiological (0.1 pM) andpharmacological concentrations (10 /AM)of melatonin failed to inhibitcell invasion. It is interesting to emphasize that the optimal melatoninconcentration to reduce MCF-7 invasiveness is the same as that whichgives the highest antiproliferative effects in an anchorage-dependent

culture system (9, 13). As reported previously (19), estrogens inducea marked change in the interaction of MCF-7 cells with the basement

membrane components, changes that are characteristic of the malignant phenotype. Estrogen-treated cells show a greater attachment to

the basement membrane, a greater ability to migrate toward laminin,a greater proliferation in culture in the presence of the basementmembrane matrix, and a much greater ability to invade the barriers ofthe reconstituted basement membrane (19). These changes have beenrelated with an increase in 17ß-estradiol in the number of lamininreceptors in MCF-7 cells (19). In our invasion model, 17ß-estradiol

also increased the number of cells that invaded the reconstitutedbasement membrane, and, interestingly, melatonin was able to counteract the stimulatory effects of 17ß-estradiol.The effects of someantiestrogenic drugs on MCF-7 cell invasiveness have been studied

140

120

100 J

80 -

"5 60 -

40 -

20 -

Control Melatonin Estradiol Estradici Estradici Estradici1 nM 1 uM 10 nM 1 iiM 10 nM

Melatonin Melatonin1 nM 1 nM

Fig. 7. Effect of melatonin and I7ß-estradiol on the chemotactic migration of MCF-7cells. In the same conditions as indicated in the legend to Fig. 9. cells migrated in theinvasion chambers in the presence of melatonin (IO /IM. l nM, or O.I pM) and/or17ß-estradiol( f /AMor 10 nM). Data are expressed as the percentage of the vehicle-treatedcontrols (mean ±SE), a. P < 0.05 versus control; b. P < O.OOl versus I /AM 17/3-

gen receptors have been found to be responsive to the antimitogenic estradioi.

Direct antitumor actions of melatonin at the cellular level have beendescribed using the in vitro antiproliferative actions of melatonin as abasis. Because only human breast cancer cell lines that express estro-

4387

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

1_»

12 h 2-1h

Fig. X. lmmunohisUK-hcmic.il staining of cell surface E-cadherin and ß,intcgrin. Partially synchronized MCF-7 cells were incubated for 24 h with the melatonin diluent {A and I)}or 1 nM melatonin (/f and A")and fixed and stained with the mouse monoclonal anli-E-cadherin antibody (A and ß)or the rat monoclonal anli-ß, integrin antibody (D and E").Quantification of the number of H-cadhenn-immunoreactive cells (O and /3¡¡ntegrin-immunorcactivc cells (/•")in the presence (G) or absence (•)of 1 nM melatonin after 6, 12, and

24 h of incubation is shown. Data are expressed as the percentage of the control response (mean ±SE), a, P < 0.05; h, P < 0.01; c, P < 0.005 versus control.

with different results that were apparently related to differential estrogen agonism. Tamoxifen and 4-hydroxytamoxifen. which are

known to be partial agonists, increase the invasive behavior of tumorcells, whereas the antiestrogen ICI I64384, which lacks estrogenagonism. reduces the invasive behavior of tumor cells (36) and is ableto counteract the estrogen invasion-stimulatory effect in a manner

similar to that which we found with melatonin.Having found that melatonin reduces the invasiveness of tumor

cells, our aim was to define the role of this indolamine within eachstep of the metastatic process. Tumor invasion is a complex biologicalprocess that begins with the detachment of tumor cells from the maintumor mass and the subsequent invasion of the adjacent tissue and thesurrounding blood and lymphatic vessels. A small but significantproportion of these detached tumor cells finally arrests in distantorgans by attachment to their basement membrane, by the secretion ofen/ymes that cause a degradation of this membrane barrier, and by themigration of tumor cells into the target tissue, forming secondarytumor deposits (metastasis). Thus, we studied the influence of melatonin on some of these steps: (a) the attachment of MCF-7 cells to the

basement membrane; (/>)the secretion of enzymes by the tumor cells;and (c) the chemotactic response of the tumor cells in the target tissue.

We show here that melatonin reduces the ability of MCF-7 cells to

attach to the basement membrane, with this effect being greater at aconcentration of I nM than at 10 /¿Mor 0.1 pin. It has been demonstrated that estrogen treatment induces a marked rearrangement of thecytoskeleton and adhesion structures and enhances the attachment ofMCF-7 cells to laminin, a basement membrane component (19, 37). Inagreement with these previous studies. I7ß-estradiol. in our system,

increased the adhesion of the cells to the basement membrane; at a

physiological concentration, melatonin completely abolished this effect of 17/3-estradiol.

The chemotactic response of MCF-7 cells toward fibronectin (used

as a chemoattractant) was also reduced by melatonin at physiologicalconcentrations (1 nM). It is well known that 17ß-estradiolstimulatesthe chemotactic migration of MCF-7 cells (19); the simultaneousaddition of 17ß-estradiol and melatonin resulted in a significantlylower chemotactic response than that seen in 17ß-estradiol-treated

cells.Although previous studies indicate that the regulation of the inva

siveness of MCF-7 cells by antiestrogens may be mediated by an

increase in collagenase IV activity (36), we did not find significantchanges in the collagenolytic activity in cells treated with melatonin,which indicates that this is not among the mechanisms involved in itsanti-invasive effects.

Taken together, our results demonstrate that melatonin reduces notonly the proliferation of MCF-7 cells in vitro but also their invasive-

ness, causing a decrease in cell attachment and cell motility probablyby interacting with mechanisms of MCF-7 cell invasion mediated byI7j3-estradiol.

Motility and attachment are two key cellular functions for theprocess of tumor metastasis. The metastatic spread of tumors isdependent on the motility and invasiveness of the cells as well as ontheir adhesive properties toward the extracellular matrix. The inhibition of one or more of these cellular functions may lead to thedecreased metastatic potential of the tumor. Tumor cell motility andinvasion are adhesion-dependent phenomena related to the presenceof cell surface adhesion molecules for both cell-cell and cell-matrixinteractions. A down-regulation or a loss of expression of these cell

4388

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

I basement membrane

Estrogen receptor +i .IE-cadherin

Estrogen receptor -

Well-differentiated Poorly-differentiated

Fig. 9. Schematic representation of the hypothetical phenotypic changes associatedwith the invasive progression of human hreast cancer cells. Human breast cancer cellsrepresent a spectrum of epithelial differentiation from well-differentiated estrogen receptor-positive cells (left) to poorly differentiated estrogen receptor-negative cells (righi).Well-differentiated estrogen receptor-positive cells may progressively lose estrogen receptor expression, followed by a loss of functional E-cadherin and, finally, by the

acquisition of the intermediate filament protein vimentin. ß,integrin expression is moreprevalent in the less-invasive cells. Invasive status is depicted in cartoon form. Wedgesindicate the amount of E-cadherin, ß,integrin. and vimentin expression.

surface adhesion molecules correlates with an increase in the inva-

siveness of tumor cells as well as with poor cell differentiation andbad prognosis of the tumor process (21-24, 38-40). The expressionof E-cadherin, a calcium-dependent membrane protein responsible forcell-cell contact, has been inversely correlated with in vitro invasion

and tumor cell differentiation (23, 40). Here we found an increase incell surface E-cadherin expression after melatonin treatment. In ad

dition, the expression of the ß,integrin (a subunit of integrins, whichare receptors that regulate interaction between cells and the extracellular matrix) has been correlated with cell differentiation and isgenerally down-regulated in breast cancer, especially in its invasive

components (21, 41). An important finding in our study was thatmelatonin increases ß,integrin expression in MCF-7 cells.

Fig. 9 depicts the evolution of human breast cancer cells frompoorly to highly invasive cells in terms of the acquisition or loss ofmarkers that represent the differentiation status of the cell. The progression traverses a spectrum of invasiveness from the poorly invasivecells (estrogen receptor-positive cells) to the highly invasive cells that

are poorly differentiated cells, which lose estrogen receptors andE-cadherin expression, have a lower expression of integrins, and

acquire vimentin expression. Our results suggest that melatonin shiftsMCF-7 human breast cancer cells to a lower invasive status, increasing ß,integrin subunit expression and E-cadherin expression and

promoting the differentiation of tumor cells, as has been demonstratedpreviously by morphological and morphometric studies of our group(42).

It has been shown that treating mice bearing established MCF-7

tumors with estrogens and antiestrogens modulates tumor growth invivo. Estrogen deprivation or antiestrogen treatment inhibits tumorgrowth (18, 43). Our results suggest that the in vitro anti-invasive

effect of melatonin could also be correlated with an in vivo decreasein the tumorigenicity of MCF-7 cells induced by melatonin. However,

because of the high mortality of nude mice and the consequently lownumber of animals that survived until the end of the experiment, westill could not demonstrate this hypothesis. In this regard, Das Guptaand Terz (44) demonstrated that the ablation of the pineal gland inSyrian hamsters increases the growth and spread of malignant melanoma. A recent study suggests that melatonin may amplify the therapeutic efficacy of tamoxifen in women with metastatic breast cancerand induce objective tumor regression in patients who have notresponded to previous therapy with tamoxifen alone (45). The anti-

invasive actions of melatonin could play an important role in stopping

the progression of the disease in these patients. Finally, we canconclude that our study points out the importance of the anti-invasive

actions of melatonin as a part of the oncostatic action of melatonin.Melatonin could delay cancer progression not only via inhibition ofthe proliferation of tumor cells, but also as a direct antagonist ofmetastatic cell functions.

REFERENCES

19.

20.

21.

22.

23.

24.

25.

Blask. D. E. The pineal: an oncostatic gland? In: R. ¡.Reiter (ed.). The Pineal Gland.pp. 253-284. New York: Raven Press, 1984.Blask, D. E.. and Hill, S. M. Melatonin and cancer: basic and clinical aspects. In: A.Miles, D. R. S. Philbrick. and C. Thompson (eds.). Melatonin Clinical Perspectives,pp. 128-173. New York: Oxford University Press, 1988.Sánchez-Barceló, E. J., Cos, S., and Mediavilla, M. D. Influence of pineal glandfunction on the initiation and growth of hormone-dependent breast tumors. Possiblemechanisms. In: D. Gupta. A. Attanasio. and R. J. Reiter (eds.). The Pineal Gland andCancer, pp. 221-232. Tübingen,Germany: Brain Research Promotion. 1988.Sánchez-Barceló. E. J.. Mediavilla. M. D., and Cos. S. Effects of melatonin onexperimental mammary cancer development. In: S. M. Webb, M. Puig-Domingo, M.

Möller,and P. Pevet (eds.). Pineal Update: From Molecular Mechanisms to ClinicalImplications, pp. 361-368. New York: PJD Publications Limited. 1997.Blask, D. E., Hill, S. M., Pelletier. D. B., Anderson, J. M., and Lemus-Wilson, A.

Melatonin: an anlicancer hormone of the pineal gland. In: R. J. Reiter and S. F. Pang(eds.). Advances in Pineal Research. Vol. 3. pp. 259-263. London: John Libbey,

1989.Maestroni. G. J. M. The immunoneuroendocrine role of melatonin. J. Pineal Res., 14:1-10, 1993.

Poeggeler. B., Reiter. R. J.. Tang, D. X.. Chen. L. D.. and Manchester, L. C.Melatonin. hydroxyl radical-mediated oxidative damage and aging: a hypothesis. J.Pineal Res.. 14: 151-168, 1993.

Tan, D. X., Chen. L. D.. Poeggeler, B., Manchester. L. C.. and Reiter, R. J. Melatonin:a potent endogenous hydroxyl radical scavenger. Endocr. J., /: 57-60. 1993.Blask. D. E.. and Hill, S. M. Effects of melatonin on cancer: studies on MCF-7 humanbreast cancer cells in culture. J. Neural Transm., 21 (Suppl.).- 433-449, 1986.Cos, S., Blask, D. E., Lemus-Wilson, A., and Hill. A. B. Effects of melatonin on thecell cycle kinetics and "estrogen rescue" of MCF-7 human breast cancer cells in

culture. J. Pineal Res., 10: 36-43, 1991.

Hill. S. M.. Spriggs. L. L., Simon, M. A.. Muraoka, H.. and Blask. D. E. The growthinhibitory action of melatonin on human breast cancer cells is linked to the estrogenresponse system. Cancer Lett., 64: 249-256, 1992.Cos, S., Fernandez, F.. and Sánchez-Barceló,E. J. Melatonin inhibits DNA synthesisin MCF-7 human breast cancer cells m vitro. Life Sci.. 5S: 2447-2453, 1996.

Hill, S. M.. and Blask, D. E. Effects of the pineal hormone melatonin on theproliferation and morphological characteristics of human breast cancer cells (MCF-7)in culture. Cancer Res., 48: 6121-6126, 1988.Cos, S., and Sánchez-Barceló. E. J. Differences between pulsatile or continuousexposure to melatonin on MCF-7 human breast cancer cell proliferation. Cancer Lett..85: 105-109. 1994.Cos. S., and Sánchez-Barceló. E. J. Melatonin inhibition of MCF-7 human breastcancer cells growth: influence of cell proliferation rate. Cancer Lett.. 93: 207-212,

1995.Cos, S., Recio, J., and Sánchez-Barceló. E. J. Modulation of the length of the cellcycle time of MCF-7 human breast cancer cells by melatonin. Life Sci.. 58: 811-816.

1996.Soûle.H. D.. Vazquez. J., Long. A.. Albert, S.. and Brennan. M. J. A human cell linefrom a pleural effusion derived from a breast carcinoma. J. Nati. Cancer Inst., SI:1409-1412. 1973.

Osborne. C. K.. Hobbs, K.. and Clark, G. M. Effect of estrogens and antiestrogens ongrowth of human breast cancer cells in athymic nude mice. Cancer Res.. 45: 584-590,

1985.Albini. A.. Graf, J., Kitten, G. T.. Kleinman. H. K.. Martin. G. R.. Veillette, A., andLippman, M. E. 17ß-Estradiol regulates and v-Ha-ras transfection constitutivelyenhances MCF-7 breast cancer cell interactions with basement membrane. Proc. Nail.Acad. Sci. USA, »3:8182-8186. 1986.Shafie, S.. and Liotta. L. A. Formation of metastasis by human breast carcinoma cells(MCF-7) in nude mice. Cancer Lett.. //: 81-87. 1980.

Gui, G. P. H.. Puddefoot. J. R.. Vinson. G. P., Wells, C. A., and Carpenter, R. Alteredcell-matrix contact: a prerequisite for breast cancer metastasis? Br. J. Cancer, 75:623-633. 1997.

Gui. G. P. H., Puddefoot. J. R., Vinson. G. P., Wells, C. A., and Carpenter, R. In vitroregulation of human breast cancer cell adhesion and invasion via integrin receptors tothe extracellular matrix. Br. J. Surg., 82: 1192-1196. 1995.Sommers. C. L., Byers. S. W.. Thompson, E. W„Torri, J. A., and Gelmann, E. P.Differentiation state and invasiveness of human breast cancer cell lines. Breast CancerRes. Treat., 31: 325-335, 1994.

Thompson, E. W., Torri, J., Sabol, M.. Sommers, C. L., Byers, S., Valverius. E. M..Martin, G. R.. Lippman. M. E.. Stampfer. M. R., and Dickson, R. B. Oncogene-induced basement membrane invasiveness in human mammary epithelial cells. Clin.Exp. Metastasis, 12: 181-194, 1994.

Albini, A., Iwamoto, Y.. Kleinman. H. K., Martin. G. R.. Aaronson, S. A., Kozlowski.J. M.. and McEwan. R. N. A rapid in vitro assay forquantitating the invasive potentialof tumor cells. Cancer Res.. 47: 3239-3245. 1987.

4389

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

MELATONIN AND TUMOR INVASION

26. Hendrix. M. J. C.. Seflor. E. A.. Seflor. R. E. B., and Fidler, I. A simple quantitativeassay for studying the invasive potential of high and low human metaslatic variants.Cancer Lett.. J8: 137-147. 1987.

27. Repesh. L. A. A new in vitro assay for quantitating tumor cell invasion. InvasionMetastasis. 9: 192-208. 1989.

28. Nakajima. M.. Welch. D. R.. Belloni. P. N.. and Nicolson. G. L. Degradation ofbasement membrane type IV collagen and lung subendothelial matrix by rat mammary adenocarcinoma cell clones of differing metastatic potentials. Cancer Res.. 47:4869-4876, 1987.

29. Nakajima, M.. Lotan, D., Baig, M. M., Carralero, R. M., Wood, W. R., Hendrix.M. J. C., and Lotan. R. Inhibition by retinoic acid of type IV collagenolysis andinvasion through reconstituted basement membrane by metastatic rat mammaryadenocarcinoma cells. Cancer Res.. 4<i: 1698-1706. 1989.

.10. Molis. T. M., Walters. M. R.. and Hill. S. M. Melatonin modulation of estrogenreceptor expression in MCF-7 human breast cancer cells. Int. J. Oncol.. 3: 687-694,

1993.31. Cos. S.. and Blask. D. E. Melatonin modulates growth factor activity in MCF-7

human breast cancer cells. J. Pineal Res.. 17: 25-32. 1994.32. Molis. T. M., Spriggs. L. L.. Jupiter. Y.. and Hill. S. M. Melatonin modulation of

estrogen-regulated proteins, growth factors, and proto-oncogenes in human breastcancer. J. Pineal Res.. 18: 93-103, 1995.

33. Tamarkin, L., Danforth. D.. Lichter. A.. Demoss. E.. Cohen. M.. Chahner. B.. andLippman, M. Decreased nocturnal plasma melatonin peak in patients with estrogenreceptor-positive breast cancer. Science (Washington DC). 216: 1003-1005. 1982.

34. Lippman. M. E., Bolán.G., and Huff. K. The effects of estrogens and antiestrogenson hormone-responsive human breast cancer in long-term culture. Cancer Res.. 36:4610-4618. 1976.

35. Soûle,H. D.. and McGrath. C. M. Estrogen responsive proliferation of clonal humanbreast carcinoma cells in athymic mice. Cancer Lett.. 10: 177-189, 1980.

36. Thompson, E. W.. Reich, R.. Shima, T. B., Albini, A., Graf, J.. Martin. G. R.,Dickson. R. B.. and Lippman. M. E. Differential regulation of growth and invasive-

ness of MCF-7 breast cancer cells by antiestrogens. Cancer Res., 48: 6764-6768,

1988.37. Sapino. A.. Pietribiasi, F.. Bussolati. G.. and Marchisio, P. C. Estrogen- and tamox-

¡fen-induced rearrangement of cytoskeletal and adhesion structures in breast cancerMCF-7 cells. Cancer Res., 46: 2526-2531, 1986.

38. Millón. R.. Nicora, F., Muller. D.. Eber, M.. Klein-Soyer, C., and Abecassis, J.Modulation of human breast cancer cell adhesion by estrogens and antiestrogens.Clin. Exp. Metastasis. 7: 405-415, 1989.

39. Hendrix. M. J. C.. Seftor. E. A.. Seftor, R. E. B.. and Trevor. K. T. Experimentalco-expression of vimentin and keratin intermediate filaments in human breast cancer

cells results in phenotypic interconversion and increased invasive behavior. Am. J.Palhol.. ISO: 483-495, 1997.

40. Yagasaki. R.. Noguchi. M.. Minami. M., and Earashi. M. Clinical significance ofE-cadherin and vimentin co-expression in breast cancer. Int. J. Oncol., 9; 755-761,

1996.41. Zutter, M. M., Ma/.oujian. G.. and Santoro. S. A. Decreased expression of ¡ntegrin

adhesive protein receptors in adenocarcinoma of the breast. Am. J. Pathol., IÌ7:863-870, 1990.

42. Crespo, D., Femández-Viadero. C.. Verduga. R.. Ovejero. V.. and Cos. S. Interaction

between melatonin and estradici on morphological and morphometric features ofMCF-7 human breast cancer cells. J. Pineal Res., 16: 215-222, 1994.

43. Osborne, C. K. Effects of estrogens and antiestrogens on human breast cancer cellproliferation: in vitro studies in tissue culture and in vivo studies in alhymic mice. In:V. P. Hollander (ed.), Hormonally Responsive Tumors, pp. 93-113. Orlando. FL:Academic Press. 1985.

44. Das Gupta. T. K.. and Terz. J. Influence of pineal gland on the growth and spread ofmelanoma in the hamster. Cancer Res., 27: 1306-1311, 1967.

45. Lissoni, P., Bami. S.. Meregalli. S., Fossati, V., Cazzaniga. M., Esposti. D.. andTancini. G. Modulation of cancer endocrine therapy by melatonin: a Phase II study oftamoxifen plus melatonin in metastatic breast cancer patients progressing undertamoxifen alone. Br. J. Cancer, 71: 854-856. 1995.

4390

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

1998;58:4383-4390. Cancer Res   Samuel Cos, Rosario Fernández, Andrés Güézmes, et al.   MCF-7 Human Breast Cancer CellsInfluence of Melatonin on Invasive and Metastatic Properties of

  Updated version

  http://cancerres.aacrjournals.org/content/58/19/4383

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

  .pubs@aacr.orgDepartment 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/58/19/4383To request permission to re-use all or part of this article, use this link

Research. on January 17, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from