ICANCERRESEARCH56. 4063-4070. September 1. 19961

ABSTRACT

The molecular mechanisms by which human cancer cells spread tobone are largely unexplored. The process likely involves cell adhesionmolecules (CAMs) that are responsible for homophilic and heterophiliccell-cell interactions. One relevant CAM may be the calcium-dependenttransmembrane glycoprotein E-cadherin. To investigate the involvementof E-cadherin in breast cancer metastasis to bone, we used an in vivomodel in which osteolytic bone metastases preferentially occur after in

jections ofcancer cells directly into the arterial circulation through the leftventricle of the hearts of nude mice. We have found that E-cadherinnegative human breast cancer cells MDA-MB-231 (MDA-231) developradiographically detectable multiple osteolytic bone metastases and cachexia in this model. However, MDA-231 breast cancer cells that weretransfected with E-cadherin cDNA showed a dramatically impaired Capacity to form osteolytic metastases and induce cachexia. Histological andhistomorphometrical analyses of bones of mice bearing mock-transfectedMDA-231 revealed aggressive metastatic tumor, whereas metastatic tumor burden was significantly decreased in the bones of mice bearingE-cadherin-expressing MDA.231. Nude mice bearing E-cadherin-trans

fected MDA-231 breast cancer cells survived longer than mice bearing

mock-transfected MDA-231 breast cancer cells. Anchorage-dependentand -independent growth in culture and tumor enlargement in the mammary fat pad of nude mice were unchanged between mock-transfected andE-cadherin-expressing MDA-231, suggesting that these differences in metastatic behavior are not due to an impairment of cell growth and tumorigenicity. Our results show the suppressive effects of E-cadherin expression on bone metastasis by circulating breast cancer cells and suggest thatthe modulation of expression of this CAM may reduce the destructiveeffects of breast cancer cells on bone.

INTRODUCTION

E-cadherin (uvomorulin) is a Mr 120,000 transmembrane glycoprotein involved in calcium-dependent cell-cell adhesion. Cell-cell adhesion mediated by E-cadherin is homophilic or heterophilic, causing

homotypic or heterotypic cell aggregation, which is likely importantin controlling embryogenesis, morphogenesis, and tissue-specific immunity (1 , 2). Recently, evidence has accumulated that E-cadherinalso plays a key role in cancer invasion and metastasis. Immunohis

tochemical examination of clinical tumor specimens of the head and

neck (3), prostate (4), stomach (5), kidney (6), and other organs (7)has shown that E-cadherin expression is decreased in invasive andmetastatic cancers and that decreased E-cadherin expression is associated with poor prognosis (6, 8). It has also been shown that poorly

differentiated metastatic breast cancers express lower levels of Ecadherin than do well-differentiated breast cancers (9—11). These

clinical studies suggest that E-cadherin expression is inversely corre

lated with the invasiveness and metastatic potential of cancer. How

Received 4/22/96; accepted 7/2/96.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.

I Supported by NIH Grants P0 1-CA40035, CA58 I 83 (Specialized Program of Re

search Excellence), and CA63628.2 To whom requests for reprints should be addressed, at Department of Medicine/

Endocrinology, University of Texas Health Science Center. 7703 Floyd Curl Drive, SanAntonio, TX 78284-7877. Phone: (210) 567-4900; Fax: (210) 567-6693; E-mail:yoneda@uthscsa.edu

ever, they do not show whether E-cadherin expression affects thebehavior of cancer cells per se during invasion and metastasis.

Treatment of noninvasive epithelial Madin-Darby canine kidneycells with monoclonal antibodies to E-cadherin in vitro rendered thesecells more invasive (12). Expression of the E-cadherin gene in highlyinvasive cancer cells dramatically suppressed their invasiveness and,

conversely, introduction of E-cadherin-specific antisense RNA ren

dered noninvasive epithelial cells invasive in vitro ( 13). E-cadherinwas strongly expressed in breast cancer cells with reduced invasive

ness, whereas relatively low levels of E-cadherin were detected in

highly invasive human breast cancer cells (14). These experimentalresults show the inhibitory effects of E-cadherin on the local inva

siveness of cancer cells and are consistent with the notion that E

cadherin is an “invasionsuppressor gene―( 13). However, they do notdemonstrate whether E-cadherin expression also suppresses cancermetastasis to distant organs. Because multiple sequential steps are

required for cancer cells to advance to achieve distant metastasis, it islikely that the expression of E-cadherin inhibits detachment and

invasion of cancer cells at the primary sites before the step of intrav

asation. Conversely, however, it is also possible that, subsequent to

intravasation, E-cadherin may rather facilitate the arrest of cancer

cells at metastatic sites or coaggregation of cancer cells with plateletsin capillaries, resulting in a promotion of metastasis. Determining

whether this is the case seems to be important to our understanding ofthe role of E-cadherin expression in distant metastasis of cancer.

We have recently developed an experimental metastasis model innude mice in which inoculation of cancer cells into arterial circulationthrough the left cardiac ventricle selectively results in osteolytic bonemetastases (15), based on the method described by Arguello et a!.

(16). Using this model of bone metastasis, we examined the role ofE-cadherin in bone metastasis of human breast cancer because breastcancer frequently colonizes bone (17, 18). The E-cadherin-negative

human breast cancer cell line MDA-23 1â€w̃as transfected with mouseE-cadherin cDNA and studied for the capacity of these cells todevelop osteolytic bone metastases. Our data provide in vivo experimental evidence that functional E-cadherin expression in MDA-23 Ibreast cancer cells caused a marked decrease in osteolytic bone

metastases even after these cells were directly introduced into the

blood stream. The results suggest that E-cadherin might have suppressive effects not only on the local invasiveness of cancer cells atprimary sites but also on the metastasis of circulating cancer cells todistant organs such as bone.

MATERIALS AND METHODS

MDA-231: A Human Breast Cancer Cell Line. The E-cadherin-negativehuman breast cancer cell line MDA-23l (Ref. 19; kindly provided by Dr. C.

Kent Osborne, University of Texas Health Science Center, San Antonio, TX)was cultured in DMEM (Hazleton Biologics, Inc., Lenexa, KS) supplemented

with 10% FCS (Hyclone Laboratories, Logan, UT) and 1% penicillin-streptomycin solution (Life Technologies, Inc., Grand Island, NY) in a humidified

atmosphere of 5% CO2 in air.

3 The abbreviations used are: MDA-23 I, MDA-MB-23 I human breast cancer cell line:

MMP, matrix metalloproteinase: lBS. Iris-buffered saline.

4063

E-Cadherin Expression in Human Breast Cancer Cells Suppresses the Developmentof Osteolytic Bone Metastases in an Experimental Metastasis Model'

Gabriel Mbalaviele, Cohn R. Dunstan, Akira Sasaki, Paul J. Williams, Gregory R. Mundy, and Toshiyuki Yoneda2Unirersitv of Texas Health Science Center, Department of Medicine. Division of Endocrinology and Metabolism. San Antonio, Texas 78284-7877 1G. M., C. R. D.. P. J. W..G.R.M., T. Y.J,andDepartmentof Oral andMaxillofacialSurgeryII, OkayamaUniversitySchoolof Dentistry.Okayama700,JapanIA. 5.1

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

Transfection of Mouse E-Cadherin eDNA. MDA-231 (1 X l0@)cellsgrown in DMEM supplemented with 10%FCS were cotransfected with 20 @xgof full-length mouse E-cadherin cDNA (in pBATEM2 plasmid driven underthe control ofthe chicken 13-actinpromoter; Ref. 20) and 1 @xgofpSV2neo (21)for neomycin resistance by calcium-phosphate coprecipitation method (22). Asa control, cells were also transfected with 2 @xgof pSV2neo. The E-cadherinexpression plasmid, pBATEM2, was a generous gift of Dr. M. Takeichi(Kyoto University, Kyoto, Japan). The neomycin expression vector, pSV2neo,was purchased from Invitrogen (San Diego, CA). After 24 h of incubation,cells were fed with fresh medium, grown for an additional 24 h, split andseeded into 100-mm dishes, and cultured in the presence of 470 @xg/mlofantibiotic G418 (Life Technologies, Inc.). G4l8-resistant colonies were iso

lated and cloned by limiting dilution twice.Immunofluorescent Staining of E-Cadherin. Cells were fixed with a

fresh solution of 3.7% (w/v) formaldehyde in PBS at room temperature for 30mm, incubated with the primary antibodies (ECCD-2; Zymed Laboratories,Inc., San Francisco, CA) at a 1:200dilution in PBS containing 1%rabbit serumat 37°Cfor 1.5 h in humidified atmosphere, washed three times with PBS, andincubated with 1:80-diluted affinity-purified rabbit anti-rat IgG conjugatedwith fluorescein 5-isothiocyanate (Sigma Chemical Co., St. Louis, MO). Afterwashing with PBS, cells were mounted in PBS-glycerol solution (1:1) and

examined under a fluorescence microscope (Nikon Inc. Instrument Division,

GardenCity,NY).Immunoblotting. Cells were lysed in lysis buffer [20 mMTris (pH 8.0), 2

mM CaCl2, 150 mrsi NaC1, 1% NP4O, 0.1% SDS, and protease inhibitors (20

mM leupeptin, 1 mM phenylmethylsulfonylfluoride, and 1% aprotinin)] at 4°C

and centrifuged at 10,000 X g for 10 ruin at 4°C,and protein amounts weredetermined using the Bio-Rad DC protein assay (Bio-Rad Laboratories, Her

cules, CA). The lysates were boiled for 5 mm, separated on 7.5% SDS-PAGE,

and transferred to polyvinylidene difluoride membranes (Immobilon-P; Millipore, Bedford, MA) in transblotting buffer [20 mM Tris, 150 mist glycine, and

20% methanol (pH 8.0)]. The membranes were incubated overnight with

blocking buffer [5% BSA in 50 mrviTris, 150 mrviNaCI, and 10mMCaCI2]andthen with monoclonal antibodies to mouse E-cadherin (Transduction Laboratories, Lexington, KY) diluted I :5000 in the same buffer for 2 h, washed oncewith TBS + 0.05% Tween 20 (TBST), then twice with TBS, and incubatedwith horseradish peroxidase-conjugated goat anti-mouse IgG (Capel, Durham,NC; diluted 1:1500 in TBS containing 5% nonfat dry milk) for 1 h. After themembranes were washed three times with TBST and then three times withTBS, the signal was visualized with an enhanced chemiluminescence detectionsystem (DuPont NEN, Boston, MA).

Colony Formation Assay. Anchorage-independentgrowth of MDA-231cells was determined as described previously (23). One ml of agarose (SeaPlaque; FMC Bioproducts, Rockland, ME) at 0.4% (wlv) in DMEM supplemented with 10% FCS containing 300 cells was overlaid on a 1-mi bottomlayer of 0.6% agarose in the same culture medium in 35-mm tissue cultureplates with a grid at the bottom (Sarstedt, Newton, NC). Plates were incubatedfor 14 days in a humidified CO2 incubator at 37°C,and colonies larger than100 ,xm in diameter were manually counted under an inverted microscope.

Cell Growth in Monolayer. Cells (1 x l05/well, 6-well plate; Falcon;Becton Dickinson Labware, Lincoln Park, NJ) were cultured in DMEMsupplemented with 10% FCS. After the indicated time of culture, cells were

trypsinized and counted on a hemocytometer under an inverted microscope.Intracardiac Injections of MDA-231 Cells in Nude Mice. Intracardiac

injection of MDA-23l cells was performed according to the technique described previously (24). Subconfluent MDA-23l cells were fed with fresh

medium 24 h before intracardiac injections. Cells (1 X l0@)were suspended in0.1 ml of PBS and injected using a 27-gauge needle into the left heart

I 2 3 4 5 6 7 kDa

-195

:@@ -112

-84

-63

A CFig. 1. E-cadherin expression in MDA-231 cellsin Western blotting (top) and immunofluorescence(bottom). Top: Lane 1, MCF-7 as positive control;Lane 2, Neo-MDA-231; Lanes 3-7, E-cad-MDA

231 subclones. Bottom, phase-contrast microscopy(A and C) and immunofluorescence (B and D).Neo-MDA-23l showed no E-cadherin expression(top, Lane 2; bottom, A and B). In contrast, E-cadMDA-231 subclones strongly expressed E-cadherin[top, Lanes 3-7; bottom, C and D (subclone 4)].

4064

— -@Ti

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

Determination of the Number and Area of Bone Metastases The num

ber of osteolytic bone metastases was enumerated on radiographs as described(15, 24). Animals were anesthetized deeply, placed in prone and lateralpositions against the films (22 X 27 cm; X-OMAT AR; KOdak, Rochester,NY), and exposed to an X-ray at 35 kV for 6 s using a Faxitron radiographicinspection unit (Model 43855A; Faxitron X-ray Corporation, Buffalo Grove,IL). Films were developed using a KOdak RP X-OMAT processor (ModelM6b). All of the radiographs of bones in nude mice were evaluated extensivelyand carefully by three individuals (including one radiologist) who had noknowledge of the experimental protocols. On radiographs, the number and area

of osteolytic metastatic foci as small as 0.5 mm in diameter, which wererecognized as demarcated radiolucent lesions in the bone, were quantitativelyassessed using a computer-assisted Jandel Video Analysis (JAVA) imageanalysis system (Jandel Scientific, Corte Madera, CA).

Tumorigenicity of MDA-231 Cells in Nude Mice. Cells (2 X 106)weresuspended in 0.2 ml of the mixture (1:1) of PBS and Matrigel (Collaborative

72 Research,Bedford,MA)andinoculatedintotherightthoracicmammaryfatpad of female nude mice using 23-gauge needles as described previously (26).Four weeks later, the tumors formed were excised and their weights weredetermined.

Histological, Histomorphometrical, and Stereological Examination ofMetastatic Cancer Burden. The details of these methods were describedpreviously (24). In brief, forelimbs and hind limbs from animals in each

treatment group were fixed with 10% formalin in PBS (pH 7.2) and decalcified

in 14% EDTA solution for 2—3weeks. Paraffin sections were made using

conventional methods. The area of metastatic cancer infiltrations was meas

ured in the distal femoral and proximal tibial metaphyses of both limbs inlongitudinal decalcified sections stained by H&E. A level showing the fullwidth of the growth plate and shaft was selected for measurement. Allmetastatic cancer cells from the joint surface to a point 4 mm down the shaftwere measured in both the bone and where it had expanded into the surrounding soft tissues. Metastases in the epiphyses were also included.

Statistical Analysis. All data were analyzed by Mann-Whitney test fornonparametric samples. The statistical difference of survival rate of the ani

— mals was analyzed by generalized Wilcoxon test. All data were presented as

the mean ±SE.

RESULTS

Transfection of E-Cadherin cDNA into MDA-231 Cells. Subsequent to transfection, G4l8 selection, and cloning by limiting dilution,several subclones of MDA-23 1 breast cancer cells were obtained.MDA-231 subclones transfected with E-cadherin cDNA were designated E-cad-MDA-23l, and MDA-23l subclones transfected withpSV2neo were designated neo-MDA-23 1 and used as controls in thefollowing experiments. E-cad-MDA-231 cells expressed E-cadherin

Body weight (g)

Fig. 2. A, growth curve of neo-MDA-23 I and E-cad-MDA-23 1 (subclones 4-7 of Fig.1. top) in monolayer. Cells (1000/cm2) were plated and cultured for 24, 48, and 72 h. Cellnumber was counted manually after trypsinization on a hemocytometer (n = 4). B,anchorage-independent growth of neo-MDA-231 (0) and E-cad-MDA-23I subclone 4(U).Threehundredcellswereplatedin softagarin 35-mmtissueculturedishes,andthenumber of colonies larger than 100 p@min diameter was counted microscopically after 14 .days of culture (n = 6). C, growth of neo-MDA-23 1 (0) and E-cad-MDA-23 I subclone4 (U) in nude mice. Two X 106 cells in the mixture of PBS and Matrigel (1:1) wereinoculated in the right thoracic mammary fat pad of4 week-old female nude mice. Tumorsformed were excised, and their weight was determined 4 weeks after inoculation asdescribed in the text. Each group had four animals. Values shown are mean ±SE.

- --.- =_J._ --

. . . E.cad-MDA-231ventricles of 4-week-old female BALB/c-nu/nu mice (Harlan Industries, Houston, TX) under anesthesia with pentobarbital (0.05 mg/g). Animals were kept Fig.3. Bodyweightof nudemicebearingneo-MDA-231andE-cad-MDA-231. Data

. . . . . . . shown in Figs. 3, 4, 5, and 6 were obtained from the same experiments and experimentsin our animal facilities for 4-7 weeks as descnbed (25). The weight of animals . .were repeated twice. Four E-cad-MDA-231 subclones (see Fig. I . top) were examined.and excised tumors was determined by using a Lum@O@GramTM(Ohaus Scale Valuesshownaremean±SE(n = 5 X2experiments).*, significantlydifferentfromtheCorporation, Florham Park, NJ). neo-MDA-231group(P < 0.01).

4065

Cell Numberx105

15

0 Neo-MDA-231

00 24 48

CultureTime(h)

T@

A

B

C

Colony/well

200

1@

Tumor Weight (9)

2.0

1',

Neo-MDA-231 E-Cad-MDA-231

I

E-Cad-MDA-231 @,* *

20•

15•

10•

5

0I@ 6 7@

Neo-MDA-231

]E-cad-MDA-231

Neo-MDA-231

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

Neo-MDA-231 E-cad-MDA-231

—@@ @@AI‘L'5'@

4@I p.@ p

E-cad-MDA-231

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

vitro and in vivo results convincingly demonstrate that transfection of

E-cadherin cDNA did not impair the capacity of growth and tumorigenicity of MDA-23 1 cells.

Development of Cachexia and Osteolytic Bone Metastases.Nude mice injected with neo-MDA-23 1 cells into the left ventricle ofthe heart showed profound cachexia (loss of muscle, fat, and bodyweight; Fig. 3 and Fig. 4, top) and multiple osteolytic lesions in theupper and lower extremities (Fig. 4, bottom and Fig. 5) 3—4weeksafter inoculation, as we reported previously (15, 24). On the otherhand, nude mice inoculated with E-cad-MDA-23l cells demonstrateda marked impairment of cachexia (Fig. 3) and a significant decreasein osteolytic bone metastases (Fig. 4, bottom and Fig. 5).

Histological View of Bones with Metastatic MDA-231 Cells.Neo-MDA-23l cells almost completely destroyed the marrow architecture, including the primary and secondary spongiosa of tibial bone,and occupied the entire marrow cavity with frequent extension to thesurrounding soft tissue (Fig. 6). In contrast, E-cad-MDA-23 1 colonization was relatively limited in the marrow cavity, and the majority oftrabecular bone and primary and secondary spongiosa remained intact

(Fig. 6). In both groups, enhanced osteoclastic bone resorption alongthe endosteal surface of bone was observed adjacent to metastaticMDA-231 cells (Fig. 6). The osteoclast number and bone resorptionsurface area determined by histomorphometry were not different inbones of E-cad-MDA-231- and neo-MDA-23l-bearing mice (data notshown).

Histomorphometrical Analysis of Metastatic Cancer Burden inBone. Consistent with the histological view, histomorphometricalanalysis of bones with neo-MDA-23 1 cells showed a larger metastaticcancer burden than bones with E-cad-MDA-23 1 cells (Fig. 7).

Metastasis in Non-Bone Organs. As a characteristic feature ofthis experimental metastasis model and by yet-unexplained mecha

Lesion number10

I

@,

/

_1@-@

4'

--

7.5

5@

2.5

0

Le@5

4

3

2

I

0

Fig. 4. Representative pictures (top) and radiographs (bottom) of nude mice bearingneo (left)- and E-cad (right; subclone 4 of Fig. 1, top)-MDA-231 cancer. Arrows,osteolytic bone metastases. Pictures and radiographs were taken 4 weeks after cellinoculation. Note the differences in cachexia (loss of fat, muscle, and body weight) andthe number of osteolytic bone metastases between both groups. *

-I-

4 5 6

*

* I _*

7

as determined by immunoblotting (Fig. 1, top) and immunofluorescence (Fig. 1, bottom) and showed polygonal shape and increasedcell-cell contact (Fig. 1, bottom), suggesting that the E-cadherintransfected was functional in MDA-231. Neo-MDA-231 cells showedelongated shape and expressed undetectable levels of E-cadherin(Fig.1).

Anchorage-dependent and -independent Growth and Tumorigenicity of E-cad-MDA-231 Cells. To determine whether E-cadhem expression affects the growth of MDA-23l cells, the growthcurve of E-cad-MDA-23l cells was compared to that of neo-MDA23 1 cells in monolayer culture. There was no difference in the growth

pattern between E-cad-MDA-23 1 and neo-MDA-23 1 cells (Fig. 2A).Furthermore, anchorage-independent growth of neo-MDA-231 and

E-cad-MDA-23l cells assessed by colony formation in soft agar wasnot significantly different (Fig. 2B).

More importantly, the weight of E-cad-MDA-23l tumors formed inthe mammary fat pad of nude mice 4 weeks after the inoculation wasnot different from that of neo-MDA-23l tumors (Fig. 2C). These in

14 5 6 7 i

Neo-MDA-231 E-cad-MDA-231

Fig. 5. The number and area of osteolytic bone metastases in nude mice bearing neo(ETh-andE-cad(U)-MDA-23I cancer.Thenumberandareaofosteolyticbonemetastaseswere scored 4 weeks after cell inoculation on radiographs using quantitative imageanalysis. Each group had five mice. Values shown are mean ± SE (si = 10). *.significantly different from the neo-MDA-231 group (P < 0.005).

4066

* *1@

on area (mm @)

+

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

?:. @@4:@;r@ -@ k B@

S@@ •@

. -@,:r@4°'@ •

. .@ :‘‘

@ °@@

-:@ @?. .@

. . ., ., ..“ .@.

,.,‘ , .

A

C@ @;-:@t@@

I

.@

, ‘S

@v@1.:* , ,_

@,@,w..,@

Bone

Fig. 6. Histological view of the proximal tibiae of nude mice bearing no cancer (A), neo-MDA-23l (B), and E-cad-MDA-23l (C; subclone 4 of Fig. 1, top). Almost all the primaryandsecondaryspongiosawerereplacedby metastaticbreastcancer(7) in miceinoculatedwithneo-MDA-231(B)ascomparedto theboneof non-tumor-bearingnormalmice(A).Incontrast, colonization of E-cad-MDA-23l (7) in the tibiae was localized, and the majority of the marrow cavity remained intact (C). with an appearance similar to that ofnon-tumor-bearing normal mice. H&E staining, X40. Osteoclastic bone resorption along the endosteal surface of bone of nude mice bearing neo-MDA-23 1 (D) and E-cad-MDA-231(5) at highermagnification.Osteoclasts(arrows)arepresentbetween metastaticMDA-231 cells (7) and bone. H&E staining, X200.

nisms, very few metastases in other organs such as lungs, kidney, died of cancer with osteolytic bone metastases (with analogous radiobrain (data not shown), and liver were detected in both groups graphical and histological views to those of neo-MDA-23 1-bearing(Table 1). Therefore, it was not possible to statistically analyze the animals) 8 weeks after the inoculation. The results indicate thateffects of E-cadherin expression on MDA-23l metastasis to non-bone E-cad-MDA-23l cells did not lose the capacity to develop bone

organs. metastases, but that the capacity is impaired, presumably due toSurvival of Animals. Nude mice inoculated with E-cad-MDA-23l E-cadherin expression.

cells survived significantly longer than animals injected with non- Production of Bone-destroying Activity. To determine whethertransfected MDA-231 cells or neo-MDA-231 cells (Fig. 8). However, E-cad-MDA-23l cells produce greater levels of osteolytic activitiesit should be noted that all animals bearing E-cad-MDA-231 cells also than do neo-MDA-231 cells, serum-free culture supernatants were

4067

.@. , ••1

.,.@ I °@‘

,t@@

@:

@:@@:%j.

%@@ $e*@

•1

-‘ 11Z@; ‘

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

Table 1 Dissemination of MDA-231 in the o

MDA-23 1 cells (1 X lOs) were inoculated into the left cardiac ventricle. Four weeks after the inocmetastasis-positive: (—),metastasis-negative; (±),occasionally metastasis-positive.rgans

of nud

ulation, eache

mice

organwas examined macroscopically and microscopically. (+),Muscle

adjacentBoneto bone AdrenalglandsBrainKidneyLungLiverOvaryNeo-MDA-23l

(n = 15) +++ ++±—---E-cad-MDA-231(n = 15) + ± ——————

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

E-cadherin in E-cadherin-negative MDA-23l human breast cancercells suppresses metastasis to bone even after the cells have enteredthe systemic circulation, diminishes loss of body weight, and prolongsthe life span of MDA-231 breast cancer-bearing animals. The resultsprovide experimental evidence that E-cadherin has a suppressiveaction on distant metastasis of breast cancer cells to bone in vivo.

The mechanisms by which E-cadherin expression decreases osteolyric bone metastases in this animal model remain to be elucidated. Inthis study, several subclones of MDA-23 1 cells that were transfectedwith E-cadherin cDNA exhibited decreased formation of osteolyticbone metastases, indicating that the results obtained here reflect thephenotype of E-cadherin-expressing MDA-231 cells and are not dueto the clonal difference. Our data also demonstrate that E-cadherinexpression had no effects on the growth and tumorigenicity of MDA231 cells. Frixen and Nagamine (31) have reported that E-cadherinexpression is associated with decreased secretion of proteolytic enzymes. In our study, however, there was no change in MMP secretionand bone-destroying activity in the absence or presence of E-cadherinexpression. Subsequent to entry into the systemic circulation, there arestill multiple diverse and complex steps including directional migration, escape from host immune cell surveillance, and extravasationthrough which cancer cells must progress before they arrest in bone.

One mechanism may involve the alteration of cell-cell adhesionbetween MDA-23 I cells. E-cad-MDA-23 I cells in culture demon

strated changes in morphology and increased cell-cell contact. Clump

ing or clustering of MDA-23l cells may affect dispersion and migration capacity in the circulation and reduce cancer cell passage throughbone capillary beds. Recent studies have described that much strongershear forces are required to dissociate the aggregates of E-cadherinpositive breast cancer cells than E-cadherin-negative breast cancercells (32) and that E-cadherin expression is inversely correlated withcell motility (33). E-cadherin has been shown to mediate not onlyhomophilic but also heterophilic and heterotypic cell-cell contact (34)and cell-substratum adhesion (35). Therefore, it is possible that theinteractions of MDA-23 1 cells with host cells and microenvironmentsare modulated by E-cadherin expression, leading to decreased bonemetastases. To further study the role of E-cadherin in MDA-23lmetastasis to bone, we are currently examining chemotactic migration, invasiveness, and adherence to various types of extracellularmatrix in neo- and E-cad-MDA-231 cells, using Boyden chamberassay, Matrigel, and extracellular matrix-coated culture plates,respectively.

There is a characteristic Step ifl bone metastasis that is not relevantin metastasis to non-bone organs. Because the majority of boneconsists of calcified matrix and cancer cells do not apparently possessan intrinsic capacity to destroy bone, the cooperative assistance ofosteoclasts, the only cells capable of resorbing bone, is required forcancer cells to grow progressively in bone (17, 18). Thus, as amechanism specific for bone metastasis, a functional associationbetween metastatic cancer cells and osteoclasts that are modulated by

either soluble factors or direct cell-cell contact or both in the bonemicroenvironment seems to be necessary for the development of

osteolytic metastases. Along this line, it is noteworthy that our recentdata demonstrate that osteoclasts express E-cadherin (27). However,

*1'

0'4

Neo-MDA-231 E-cad-MDA-231Fig. 7. Histomorphometrical analysis of metastatic cancer burden of two ([1)- and

E-cad (U; subclone 4 of Fig. I, top)-MDA-231 in bone (proximal tibia). Values shown aremean ±SE (n = 20 bones).*, significantlydifferentfromthe neo-MDA-231group(P < 0.005).

harvested and examined for their activity in bone resorption assay

using 45Ca-labeled fetal rat long bones, as described (27). There wasno difference in the secretion of bone-destroying activity betweenE-cad-MDA-231 and neo-MDA-23l cells (data not shown). Moreover, we also examined MMP activity in the culture supernatants byzymography. However, we did not observe a significant change in

MMP secretion between E-cad-MDA-231 and neo-MDA-23l cells(data not shown).

DISCUSSION

Previous experimental and clinicopathological studies have mdi

cated that E-cadherin impairs the local invasion of cancer cells intosurrounding tissue at the primary sites (3, 7, 11—14,28—30),whichmay eventually result in the suppression of distant metastasis. However, to date, little is known in regard to whether E-cadherin decreasesthe metastasis of circulating cancer cells that have achieved intravasation. We attempted to address this question specifically for breastcancer metastasis to bone because bone is the most common site ofmetastasis in breast cancer (17, 18). To study this, we took advantageof an experimental metastasis model in which the inoculation ofcancer cells into arterial circulation through the left heart ventricle offemale nude mice causes selective development of metastases in bone.Nakai et a!. ( 15) and Sasaki et a!. (24) have already reported that, asa characteristic feature of this model, cancer cells inoculated using thistechnique preferentially metastasize to bone and develop osteolyticlesions, whereas metastasis to other organs such as lung, liver, andkidney is rare. Although the mechanisms underlying the preference

for bone to non-bone organs in this experimental metastasis modelremain unknown, this technique allows us to study the role of Ecadherin in the development of osteolytic bone metastases by humanbreast cancer cells that are directly introduced into the arterial bloodstream. The data presented here clearly show that the expression of

4068

Areaof Individualtumor4 I@ —r-

3'

2'

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

adhesion molecule E-cadherin is reduced or absent in high-grade prostate cancer.Cancer Res., 52: 5104—5109,1992.

5. Mayer. B.. Johnson, J. P.. Leitl, F., Jauch, K. W., Heiss. M. M.. Schildberg. F. W..Birchmeier, W., and Funke, I. E-cadherin expression in primary and metastatic gastriccancer: down-regulation correlates with cellular differentiation and glandular disintegration. Cancer Res.. 53: 1690—1695. 1993.

6. Katagiri. A., Watanabe, R.. and Tomita, Y. E-cadherin expression in renal cell cancerand its significance in metastasis and survival. Br. J. Cancer, 71: 376—379, 1995.

7. Takeichi, M. Cadherins in cancer: implications for invasion and metastasis. Cuff.Opin. Cell. Biol., 5: 806—811, 1993.

8. Umba, R., Isaacs. W. B., Bringuier. P. P., Schaafsma, H. E., Karthaus. H. F. M.,Oosterhof, G. 0. N., Debruyne, F. M. J., and Shalken, J. A. Decreased E-cadherinexpression is associated with poor prognosis in patients with prostate cancer. CancerRes., 54: 3929—3933,1994.

9. Shiozaki, H., Tahara, H., Oka, H., Miyata, M., Kobayashi. K.. Tamura. S., Lihara,K.. Doki, Y., Hirano, S., Takeichi, M., and Mon. T. Expression of immunoreactive E-cadherin adhesion molecules in human cancer. Am. J. Pathol., 139: 17—23,I991.

10. Gamolla, C., Palacios, J., Suarez, A., Pizarro, A., Navarro, P., Quintanilla, M., andCano, A. Correlation of E-cadherin expression with differentiation grade and histological type in breast carcinoma. Am. J. Pathol.. 142: 987—993, 1993.

I I . Oka, H., Shiozaki, H., Kobayashi. K., Inouc, M.. Tahara. H., Kobayashi. T.,Takatsuka, Y., Matsuyoshi. N.. Hirano, S., Takeichi. M., and Mon. I. Expression ofE-cadherin cell adhesion molecules in human breast cancer tissues and its relationshipto metastasis. Cancer Res.. 53: 1696—1701, 1993.

12. Behrens, J., Marcel, M. M., Van Roy. F. M., and Birchmeier, W. Dissecting tumorcell invasion: epithelial cells acquire invasive properties after the loss of uvomorulinmediated cell-cell adhesion. J. Cell Biol., 108: 2435—2447, I989.

I3. Vleminckx, K., Vakaet, L., Marcel, M., Fiers, W., and Van Roy, F. Genetic manipulation of E-cadherin expression by epithelial tumor cells reveals an invasion suppressor role. Cell, 66: 107—119, 1991.

14. Sommers, C. L., Thompson. E. W.. Toni, J. A., Kemler. R.. Gelmann. E. P.. andByers. S. W. Cell adhesion molecule uvomorulin expression in human breast cancercell lines: relationship to morphology and invasive capacities. Cell Growth & Differ..2: 365—371.1991.

IS. Nakai, M., Mundy, G. R., Williams, P. J.. Boyce. B., and Yoneda. T. A syntheticantagonist to laminin inhibits the formation of osteolytic metastases by humanmelanoma cells in nude mice. Cancer Res., 52: 5395—5399, 1992.

16. Arguello. F.. Fulanetto, R. W., Baggs. R. B.. Graves, B. I.. Harwell. S. E.. Cohen.H. J., and Frantz, C. N. Incidence and distribution of experimental metastases inmutant mice with defective organ microenvironment (genotype siisr' and WIW').Cancer Res.. 52: 2304—2309. 1992.

17. Yoneda, T., Sasaki, A., and Mundy. G. R. Osleolytic bone disease in breast cancer.Breast Cancer Res. Treat., 32: 73—84,1994.

I8. Mundy, G. R., and Yoneda, I. Facilitation and suppression of bone metastasis. Clin.Orthop. Relat. Res., 312: 34—44,1995.

19. Cailleau, R., Young, R.. Olive, M., and Reeves, W. J. Breast tumor cell lines frompleural effusions. J. NatI. Cancer Inst.. 53: 661—674. 1974.

20. Nagafuchi. A., Shirayoshi, Y.. Okazaki. K.. Yasuda. K., and Takeichi. M. Transformation of cell adhesion properties by exogenously introduced E-cadherin cDNA.Nature (Land.), 329: 341—343,1987.

21. Southern, P. J., and Berg, P. Transformation of mammalian cells to antibioticresistance with a bacterial gene under a control of the SV40 early region promoter. J.Mol. Appl. Genet.. 1: 327—341,1982.

22. Sambrook, J.. Fritsch, E. F., and Manitis, I. Molecular Cloning: A LaboratoryManual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. 1989.

23. Arteaga. C. L., Hurd, S. D., Winnier. A. R., Johnson, M. D., Fendly, B. M.. andForbes. J. I. Anti-transforming growth factor (TGF)-)3 antibodies inhibit breastcancer cell tumorigenicity and increase mouse spleen natural killer cell activity:implications for a possible role of tumor cell/host TGF-@ interactions in human breastcancer progression. J. Clin. Invest., 92: 2569—2576, 1993.

24. Sasaki, A., Boyce, B. F., Story, B., Wright. K. R., Chapman, M., Boyce, R., Mundy.G. R., and Yoneda. Y. Bisphosphonate risedronate reduces metastatic human breastcancer burden in bone in nude mice. Cancer Res., 55: 3551—3557, 1995.

25. Yoneda, T.. Aufdemorte, T. B., Nishimura, R., Nishikawa, N., Sakuda, M., Alsina,M. M., Chavez, J. B., and Mundy, G. R. Occurrence of hypercalcemia and leukocytosis with cachexia in a human squamous cell of the maxilla in athymic nude mice:a novel experimental model of three concomitant paraneoplastic syndromes. J. Clin.Oncol., 9: 468—477, 1991.

26. Price, J. E.. Polyzos, A., Zhang, R. D., and Daniels, L. M. Tumorigenicity andmetastasis of human breast carcinoma cell lines in nude mice. Cancer Res., 50:717—721,1990.

27. Mbalaviele, G., Chen. H., Boyce, B. F., Mundy. G. R.. and Yoneda. T. The role ofcadherin in the generation of multinucleated osteoclasts from mononuclear precursorsin murine marrow. J. Clin. Invest., 95: 2757—2765, 1995.

28. Doki, Y., Shiozaki, H., Tahara, H., Inoue, M., Oka, H., lihara, K., Kadowaki, T.,Takeichi, M., and Mori, I. Correlation between E-cadherin expression and invasiveness in iitro in a human esophageal cancer cell line. Cancer Res.. 53: 3421—3426.1993.

29. Frixen, U. H., Behrens, J., Sacks, M., Eberle, G., Voss, B., Warda, A., Lochner, D.,and Birchmeier, W. E-cadherin-mediated cell-cell adhesion prevents invasiveness ofhuman carcinoma cells. J. Cell Biol., 113: 173—185,1991.

30. Navarro. P., Gomez, M., Pizarro, A., Gamallo, C., Quintanilla, M., and Cano. A. Arole for the E-cadherin cell-cell adhesion molecule during tumor progression ofmouse epidermal carcinogenesis. J. Cell Biol., 115: 517—533,1991.

3 1. Frixen, V. W., and Nagamine, Y. Stimulation of urokinase-type plasminogen activa

Weeks

4069

Survival Rate (%)

Fig. 8. Survival of nude mice bearing neo (broken line)- and E-cad (solid line: subclone4 of Fig. I, top)-MDA-231 cancer. There were 14 mice in each group. *. significantlydifferent from the neo-MDA-231 group (P < 0.01).

our results demonstrated that E-cadherin expression did not alter theproduction of osteolytic activity and MMP secretion in MDA-23 Icells in culture. Thus, it seems unlikely that E-cadherin expressionmodulates the Secretion of soluble factors by metastatic MDA-23 Ibreast cancer cells that stimulate osteoclastic bone destruction.

Enhanced cell-cell aggregation in E-cad-MDA-23 1 cells also mdicates that the transfected E-cadherin cDNA is functional, and thatdownstream signaling molecules, including catenins (36), are intact inMDA-231 cells. In fact, a recent study has demonstrated moderateexpression of a- and @-cateninin MDA-23 I cells (37). Becausecadherin-mediated cell-cell adhesion requires an association of cadhem with catenins (36) and evidence that catenins also play a criticalrole in cancer invasiveness and metastasis is accumulating (38—41),the involvement of catenins in MDA-23 1 metastasis to bone needs tobe explored.

In conclusion, we have found in the present study that theexpression of E-cadherin inhibits the metastasis of circulatinghuman breast cancer cells to bone in vivo. Inhibition of aj33integrin by peptide antagonists or monoclonal antibodies has beendemonstrated to promote primary tumor regression by inhibitingangiogenesis (42). In addition to integrins, the modulation ofexpression of another cell adhesion molecule, E-cadherin, by administering agents that up-regulate E-cadherin expression may bea potential therapeutic approach to inhibit the spread and destructive effects of breast cancer in bone. Our results support thefeasibility of such a strategy.

ACKNOWLEDGMENTS

We thank Drs. M. Takeichi and R. Kemler for providing us with plasmidpBATEM2 and DECMA-l antibodies, respectively, and Thelma Barrios forsecretarial assistance.

REFERENCES

I . Takeichi, M. Cadherin cell adhesion receptors as a morphogenetic regulator. Science(WashingtonDC),251:1451—1454,1991.

2. Cepek, K. L., Shaw, S. K., Parker, C. M., Russell, G. J., Morrow, J. S., Rimm, D. L.,andBrenner,M.B.AdhesionbetweenepithelialcellsandT lymphocytesmediatedbyE-cadherin and the aE(37 integrin. Nature (Lond.), 372: 190-193, 1994.

3. Schipper. J. H., Frixen, U. H., Behrens. J., Unger, A., Jahnke, K., and Birchmeier, W.E-cadherin expression in squamous cell carcinomas of head and neck: inversecorrelation with tumor dedifferentiation and lymph node metastasis. Cancer Res.. 51:6328—6337, 1991.

4. Umbas, R., Schalken, J. A., Aalders, T. W., Carter, B. S., Karthaus, H. F. M.,Schaofsma, H. E., Debruyne. F. M. J.. and Isaacs, W. B. Expression of the cellular

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

E-CADHERIN IN BREAST CANCER METASTASIS TO BONE

tor expression by blockage of E-cadherin-dependent cell-cell adhesion. Cancer Res.,53: 3618—3623, 1993.

32. Byers. S. W., Sommers, C. L, Hoxter, B., Mercurio, A. M., and Tozeren, A. Role ofE-cadherin in the response of tumor cell aggregates to lymphatic, venous, and arterialflow: measurement of cell-cell adhesion strength. J. Cell Sci., 108: 2053—2064,1995.

33. Simard, D., and Nabi, I. R. Inverse relation of autocrine motility factor receptor andE-cadherin expression following MDCK epitheial cell transformation. Biochem.Biophys. Res. Commun., 219: 122—127.1996.

34. Cepek, K. L., Shaw, S. K., Parker, C. M.. Russell, G. J., Marrow, J. S., Rimm, D. L.,and Brenner, M. B. Adhesion between epithelial cells and T lymphocytes mediated byE-cadherin and the a@j3@integrin. Nature (Land.), 372: 190-193, 1994.

35. Miyaki, M., Tanaka, K., Kikuchi-Yamashita, R., Muraoka, M., Komshi, M., andTakeichi, M. Increased cell-substratum adhesion and decreased gelatinase secretionandcellgrowthinducedby E-cadherintransfectionof humancoloncarcinomacells.Oncogene, 11: 2547—2552, 1995.

36. Gumbiner, B. M. Cell adhesion: the molecular basis of tissue architecture andmorphogenesis. Cell, 84: 345—357,1996.

37. Piercall, W. E., Woodard, A. S., Morrow, J. S., Rimm, D., and Fearon, E. R. Frequent

alterations in E-cadhenn and a- and (3-catenin expression in human breast cancer celllines. Oncogene, 11: 1319—1326, 1995.

38. Shiozaki, H., Lihara, K., Oka, H., Kadowaki, T., Matsui, S., Gofuku, J., lnoue, M.,Nagafuchi, A., Tsukita, S., and Mori, T. Immunohistochemical detection of a-catemnexpression in human cancer. Am. J. Pathol., 144: 667—674,1994.

39. Watabe, M., Nagafuchi, A., Tsukita, S., and Takeichi, M. Induction of polarizedcell-cell association and retardation of growth by activation of the E-cadherin-catemnadhesion system in dispersed carcinoma line. J. Cell Biol., 127: 247—256,1994.

40. Vermeulen, S. J., Bruyneel, E. A., Bracke, M. E., DeBruyne, G. K., Vennekens,K. M., Vleminckx, K. L., Berx, G. J., Roy, F. M., and Marcel, M. M. Transition fromthe noninvasiveto the invasivephenotypeand loss of a-cateninin humancoloncancers. Cancer Res., 55: 4722—4728.1995.

41. Takayama, T., Shiozaki, H., Shibamoto, S., Oka, H., Kimura, Y., Tamura, S., Inoue,S., Monden,T., Ito, F., and Monden,M. /3-cateninexpressionin humancancers.Am. J. Pathol., 148: 3—46,1996.

42. Brooks, P. C., Montgomery, A. M. P., Rosenfeld, M., Reisfeld, R. A., Hu, T., Klier,G.. and Cheresh, D. A. Integrin aj33 antagonists promote tumor regression byinducing apoptosis of angiogenic blood vessels. Cell, 79: 1157—1164, 1994.

4070

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

1996;56:4063-4070. Cancer Res   Gabriel Mbalaviele, Colin R. Dunstan, Akira Sasaki, et al.   an Experimental Metastasis ModelSuppresses the Development of Osteolytic Bone Metastases in E-Cadherin Expression in Human Breast Cancer Cells

  Updated version

  http://cancerres.aacrjournals.org/content/56/17/4063

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/56/17/4063To request permission to re-use all or part of this article, use this link

Research. on January 16, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from