deficiency of connexin43 gap junctions is an independent marker … · deficiency of connexin43 gap...

[CANCER RESEARCH 59, 4104–4110, August 15, 1999]

Deficiency of Connexin43 Gap Junctions Is an Independent Marker forBreast Tumors1

Dale W. Laird, Paulina Fistouris, Gerald Batist, Lesley Alpert, Hung T. Huynh, George D. Carystinos, andMoulay A. Alaoui-Jamali 2

McGill Centre for Translational Research in Cancer, Lady Davis Institute for Medical Research and Departments of Medicine, Pathology, Oncology, and Pharmacology andTherapeutics, Sir Mortimer B. Davis Jewish General Hospital, Faculty of Medicine, McGill University, Montreal, Quebec, Canada H3T 1E2 [D. W. L., G. B., L. A., H. T. H.,G. D. C., M. A. A-J.]; and Department of Anatomy and Cell Biology, University of Western Ontario, London, Ontario, Canada N6A 5C1 [D. W. L., P. F.]

ABSTRACT

Gap junctions are intercellular channels that are formed from mem-bers of a family of proteins, the connexins (Cxs). Gap junctions play animportant role in vital functions, including the regulation of cell growthand cell differentiation. Here, we examined the expression of Cx43, amajor Cx in breast tissue, in 32 surgical specimens obtained from breastcancer patients who underwent a primary surgical resection prior tochemotherapy or radiotherapy treatments. The expression of Cx43 gapjunctions was compared to the levels of estrogen, progesterone, and erbB2tyrosine kinase receptors. In addition, a panel of breast cancer cell linesand a series of normal rat mammary tissues and rat mammary tumorsinduced in vivo by dimethylbenz(a)anthracene were studied. We demon-strated that the lack of Cx43 gap junctions is a common feature of humanmammary cancer tissues compared to nonneoplastic breast tissues sur-rounding primary tumors. Cx43 gap junctions were not observed in ductalcarcinomasin situ, infiltrating ductal carcinomas, and infiltrating lobularcarcinomas, and they seem to be independent of estrogen, progesterone,and erbB2 receptor status. In breast cancer cell lines and rodent mam-mary carcinoma tissues, down-regulation of Cx43 occurs at the mRNAlevel, suggesting a transcriptional mechanism for the decrease of Cx43protein in breast cancer. In summary, this study provides evidence ofdecreased expression of Cx43 gap junctions in breast cancer at variousstages of progression as well as breast cancer cell lines and raises thepossibility that Cx43 may be a useful marker for detecting early oncogen-esis in the breast. Because Cx43 gap junctions are lacking in breast cancerand restoration of Cx43 has been shown to reverse the malignant pheno-type in vitro, pharmacological up-regulation of Cx43 may prove beneficialin cancer therapeutics.

INTRODUCTION

Increasing experimental and clinical evidence supports stepwiseand progressive mechanisms for breast carcinogenesis. This process isoften associated with the (a) accumulation of a wide spectrum ofgenetic abnormalities, (b) acquisition of uncontrolled cell growth, (c)selective survival of transformed cells within the surrounding inter-stitial tissue of normal breast, and (d) the capacity of cells to invadedistant sites. In the clinical situation, this multistage process is man-ifested by progression of a preneoplastic lesion to carcinomain situ toinfiltrating carcinoma and, ultimately, to micro- and macrometastases.

Breast tumor intercellular interactions within the tissue microenvi-ronment play an important role in this multistage carcinogenic proc-ess. In normal conditions, the mammary gland is composed of severalcell types that interact with each other and with other components ofthe tissue microenvironment, including neighboring mesenchymal

cells and the extracellular matrix. These cell-cell interactions involvephysical and dynamic organization mediated by distant signalsthrough secreted growth factors and signaling molecules as well asdirect junctions. The latter include tight junctions mediated by pro-teins such as occludin and zonula occludins (reviewed in Ref. 1), gapjunctions (reviewed in Ref. 2), adherence junctions mediated bycadherins (reviewed in Ref. 3), and desmosomes (reviewed in Ref. 4).Many of these junctions are interdependent and probably are regulatedthrough a coordinated process (5). Alteration of GJIC3 is among theearly changes associated with carcinogenesis and restoration of gapjunctions can suppress tumor cell growth (6–11). Furthermore, phor-bol esters and epidermal growth factor activation of protein kinasesdown-regulate GJIC (12–16). GJIC has been implicated in manybiological functions, including the regulation of cell growth anddifferentiation (reviewed in Ref. 17).

GJIC mediates the transfer of ions, nucleotides, and small regula-tory molecules from cytoplasm to cytoplasm without leakage into theextracellular space (6, 18). Gap junctions are composed of multiplehemichannels (connexons) in the plasma membrane of one cell joinedin mirror symmetry with the same number of hemichannels in theapposing cell membrane. Connexons are formed from members of amultigene family of distinct but functionally related proteins calledCxs (19, 20). In the course of gap junction assembly, individualchannels cluster at the cell surface to form gap junction plaques (2).Gap junctions are dynamic structures with relatively short half-livesof 1–5 h (2, 21). They are tightly regulated by voltage, growth factors,a number of secondary messengers including cAMP (15, 22–24), andretinoids (25) and are subjected to phosphorylation by a number ofprotein kinases (16, 20, 26).

Presently, 15Cx genes have been identified in mammalian cells(27, 28). Three Cxs have been detected in normal breast tissue,namely Cx43, Cx26, and Cx32. Cx43 is the predominant Cx in humanbreast epithelium, and it was detected predominantly between myo-epithelial cells (29). There is evidence suggesting that Cxs may playa role in normal mammogenesis, lactogenesis, and involution. It hasbeen shown that Cx26 levels in the epithelium increase substantiallyin the lactating mouse mammary gland epithelium and decline duringinvolution (30), and hormones such as estrogens and progesteroneshave been shown to regulate the expression of theCx43gene (31).

Here, we examined the expression of Cx43-mediated gap junctionsin a panel of human breast cancer tissues and their matched normalcontrols. Results were compared to those from human breast cancercell lines and experimental murine mammary carcinomas.

MATERIALS AND METHODS

Human Breast Specimens.Surgical specimens used in this study wereobtained from patients who underwent a primary surgical resection in theDepartment of Oncology at the Sir Mortimer B. Davis Jewish General Hos-

Received 1/19/99; accepted 6/15/99.The costs of publication of this article were defrayed in part by the payment of page

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

1 This work was supported by the Canadian Breast Cancer Research Initiative-NationalCancer Institute of Canada Grant 009233 (to D. W. L., M. A. A-J., G. B.) and the McGillCentre for Translational Research in Cancer (to D. W. L., L. A., M. A. A-J.).

2 To whom requests for reprints should be addressed, at Lady Davis Institute forMedical Research, Room 523, 3755 Chemin de la Cote-Ste-Catherine, Montreal, Quebec,Canada, H3T 1E2. Phone: (514) 340-8260 ext. 3438; Fax: (514) 340-7576; E-mail:[email protected].

3 The abbreviations used are: GJIC, gap junctional intercellular communication; Cx,connexin; DCIS, ductal carcinomain situ; ILC, infiltrating lobular carcinoma; IDC,infiltrating ductal carcinoma; DMBA, dimethylbenz(a)anthracene; ER, estrogen receptor;PGR, progesterone receptor.

4104

on July 19, 2020. © 1999 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

pital. Patients had not received chemotherapy or radiotherapy prior to resec-tion. Access to breast tissues was approved by the Hospital Research andEthics Committee (Protocol JGH/98-043). Table 1 summarizes clinical infor-mation of human breast tissues studied. Of the 32 tumor tissues studied, 4 caseswere DCIS, 21 were IDC, and 7 were ILC. Control normal tissues, derivedfrom normal breast surrounding primary tumors from the same patient, wereavailable for study in 14 cases. Tissues were fixed in buffered formalin andembedded in paraffin. Five-mm sections were prepared for analysis of variousmarkers by immunohistochemistry or confocal microscopy. Parallel sectionswere stained with H&E for morphological examination.

Induction of Rat Breast Cancer and Tissue Preparation.Rat mammarytumors were induced by gavage feeding of 20 mg of DMBA to 55-day-oldfemale Sprague Dawley rats. Palpable tumors (.0.5 cm) were observed in;75% of animals by day 90 following carcinogen administration. On day 105after DMBA treatment, rats bearing tumors were sacrificed by CO2 exposure,and mammary tumors were harvested, quickly frozen in liquid nitrogen, andstored at280°C for later analysis. Normal mammary tissues were derivedfrom rats treated with the vehicle alone and prepared as described for mam-mary tumor tissues.

Cell Lines and Cell Culture. The cell lines used in this study consisted ofhuman breast carcinoma cell lines Hs578T, MDA-231, MDA-468, HBL-100,T47-D, and ZR-75. All these cell lines were obtained from the American TypeCulture Collection (Manassas, VA). Cells were maintained as monolayer

cultures ina-MEM supplemented with 10% FCS (Life Technologies, Inc.,Grand Island, NY). Confluent stock cultures were trypsinized and plated at1 3 106 cells per 75-cm2 plate ina-MEM supplemented with 10% FCS. After24 h, the cell monolayer was rinsed once with serum-freea-MEM and thenincubated for a further 48 h ina-MEM containing 2.5% FCS.

Immunofluorescent Labeling and Confocal Microscopy.Deparaffinizedsections were blocked with BSA and immunolabeled for 1 h with 5–8mg/mlaffinity-purified anti-Cx43 antibody as described in Lairdet al. (32). Sectionswere washed and immunolabeled with antirabbit IgG conjugated to Texas Redfor 1 h. For double-labeled experiments, sections were first labeled for Cx43,as described above, followed by a mouse anticytokeratin antibody (1:25dilution; Boehringer Mannheim, Laval, Quebec, Canada) and antimouse IgGconjugated to fluorescein. Immunofluorescent labeled tissue sections wereanalyzed using the Zeiss LSM 410 inverted confocal microscope as describedin Laird et al. (32). Briefly, to excite and detect Texas Red, an argon/kryptonmix gas laser that emits a 568-nm line was used, and the resulting fluorescencewas passed through a 590 nm long pass filter. The same laser was used to emita 488 nm line to excite fluorescein, and the resulting fluorescence wascollected after passing through a 515–565-nm band pass filter. All fluorescentsignals were collected on a photomultiplier and stored as digital images.Composite images were printed on a Tektronix color printer. Normal andtumor breast tissue was examined and scored based on the presence of gap

Table 1 Clinical information and expression of Cx43 gap junctions, ER, PGR, and erbB2 in breast tissues

Patient no. Age (yr) Histological type Grade Tissue typeCx43 gapjunctiona ERb PGRb erbB-2c

1 55 DCIS/IDC 3 Tumor 2 NDd ND NDNormal 1

2 41 DCIS 3 Tumor 1/2 1 1 NDNormal 11

3 73 DCIS 2 Tumor 2 1 1 ND4 42 DCIS 2–3 Tumor 2 ND ND ND

Normal 1115 66 IDC 3 Tumor 2 2 2 1116 90 IDC 3 Tumor 2 2 2 27 46 IDC 3 Tumor 2 2 2 1

Normal 1 28 91 IDC 2 Tumor 2 1 1/2 1/29 71 IDC 3 Tumor 2 1 1 2

Normal 1 210 44 IDC 2 Tumor 2 1 1 1

Normal 1 211 38 IDC 3 Tumor 2 1 2 112 66 IDC 2–3 Tumor 2 2 2 113 60 IDC 2–3 Tumor 2 1 1 214 39 IDC 3 Tumor 2 2 2 215 56 IDC 2 Tumor 2 1 1 216 46 IDC 3 Tumor 2 1 1 217 36 IDC 3 Tumor 2 2 2 2

Normal 11 218 40 IDC 2–3 Tumor 2 1 1/2 111

Normal 111 219 57 IDC 3 Tumor 2 1 1 111

Normal 111 220 63 IDC 3 Tumor 2 1 2 111

Normal 111 221 52 IDC 2 Tumor 2 1 2 111

Normal 111 222 42 IDC 3 Tumor 2 2 1/2 2

Normal 111 223 45 IDC 2 Tumor 2 1 1 2

Normal 111 224 46 IDC 1–2 Tumor 2 2 1 2

Normal 1 225 46 IDC 2–3 Tumor 2 1 1 1126 66 ILC/IDC 2–3 Tumor 2 1 1 11

Normal 1 227 80 ILC 2 Tumor 2 1 1 128 70 ILC 1 Tumor 2 2 2 129 67 ILC 1 Tumor 2 1 2 230 80 ILC 2 Tumor 2 1 1 131 41 ILC 1 Tumor 2 1 1 132 46 ILC 2 Tumor ND 1 1 2

a 2, no expression;1, moderate to good expression;11 to 111, high to very high expression.b 2, no expression;1, positive staining$5% of the cells;1/2, ,5% expression.c 2, no expression;1/2, weak expression;1, moderate to good expression;11 to 111, high to very high expression.d ND, not determined.

4105

CX43 GAP JUNCTIONS IN BREAST CANCER



junction plaques. In many cases, both normal and tumor tissue were present onthe same slide.

Immunohistochemistry Analysis. Immunohistochemical analysis wasperformed on deparaffinized slides using the automated Ventana immuno-staining machine (Ventana Medical Systems Inc., Tucson, AZ). This system isbased on localization of bound antibodies by a biotin-conjugated secondaryantibody in conjunction with an avidin/streptavidin-enzyme conjugate. Amouse monoclonal antibody directed against an epitope on human ER (clone6F11; Ventana Medical Systems) and a mouse antibody directed against anepitope on human PGR (clone PGP-1A6; Ventana Medical Systems) wereused to examine the levels of ER and PGR expression, respectively. Theanti-erbB-2 Ab4 antibody (Oncogene Science, Cambridge, MA) was used at adilution of 1:100, whereas ER and PGR antibodies were used undiluted.Normal goat serum was used to block the nonspecific binding sites prior toincubating the sections with primary antibodies in a humidified chamber for18 h at 4°C. Antibody binding was visualized by subsequent incubations withbiotinylated secondary antibody and an avidin-biotin-horseradish peroxidasecomplex, followed by a 4-min incubation in diamine tetrahydrochloride beforecounterstaining with hematoxylin. Scoring of sections was carried out blindedto clinical characteristics. Specimens were assigned one of the five stainingscores, as follows:111, heavy staining on 75–100% of the cells;11,moderate staining on 25–75% of the cells;1, focally positive on 5–25% of thecells;2/1, weak staining on,5% of the cells; or2, negative staining. For ERand PGR, the tumors were considered positive if staining is was found in.5%of the cells.

Northern Blot Analysis. Fresh tissues, frozen immediately in liquidnitrogen after surgery, were used to extract total RNA. Cells were collectedat 70% confluence. Total RNA was isolated using the RNAZol B premixedsolution (Tel-Test, Friendswood, TX). RNA integrity was examined byagarose-formaldehyde gel electrophoresis. Twentymg of RNA were elec-trophoresed through a 1% formaldehyde-agarose gel and blotted ontonitrocellulose membrane for 18 h. Filters were prehybridized for 2 h at42°C in prehybridization buffer [containing 50% (v/v) formamide, 53SSC, 53 Denhardt’s solution, 250mg/ml sonicated calf thymus DNA, and0.5% SDS]. Cx43 cDNA was labeled with [32P]dCTP (13 106 cpm/ml)using an oligolabeling kit (Pharmacia) and was then added to the blots.Hybridization was carried out for 20 h at 42°C in hybridization buffer[dextran sulfate:prehybridization buffer, 1:4 (v/v)]. Membranes werewashed three times for 10 min at room temperature in 13 SSC containing0.1% SDS and three times at 60°C for 10 min in 0.13 SSC containing 0.1%SDS, and then they were subjected to autoradiography. After development,X-ray films were scanned with a spectrophotometer equipped with LKBUltrascan Laser Densitometer. After each probing, filters were stripped ofprobes by three washes in 0.13 SSC and 0.1% SDS at 100°C for 15 minand exposed to X-ray film to ensure that the probe was completelyremoved. The stripped nitrocellulose membranes were reprobed withmouse GAPDH cDNAs, used to normalize for mRNA levels.

Western Blot Analysis. Cells were grown in serum supplemented mediumuntil they were 70% confluent in 75-cm2 flasks. Cells were washed twice incold PBS and then lysed directly using lysis buffer [1% Triton X-100, 10 mM

Tris-HCl (pH 8.0), 60 mM KCl, 1 mM EDTA, 1 mM DTT, 0.5% NP40, 0.5 mMphenylmethylsulfonyl fluoride, 0.01 mg/ml leupeptin, 0.01 mg/ml pepstatin,0.01 mg/ml aprotonin, and 5 mM sodium orthovanadate]. The lysate wasincubated for 10–15 min at 4°C and centrifuged at 14,0003 g at 4°C for 15min. The supernatant was used to determine protein concentration. Cell ex-tracts were resolved by 10% SDS-PAGE, transferred to nitrocellulose (Costar,Cambridge, MA), blocked in 5% milk in PBS, and then incubated with ananti-Cx43 rabbit polyclonal antibody (21). Immune complexes were detectedby horseradish peroxidase conjugates using the enhanced chemiluminescencedetection reagents (Amersham). Blots were subsequently stripped in 100 mM

2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl (pH 6.7) at 50°C for 30min and then immunoblotted with monoclonal anti-GAPDH (clone 6C5;Cedarlane Laboratories, Hornby, Ontario, Canada).

RESULTS

We immunofluorescently examined Cx43 expression and localiza-tion in human breast tumor specimens from 32 cases with DCIS, ILC,or IDC. For 15 cases, control tissues derived from normal breast

surrounding the primary tumors were examined on the same slide. Thethree types of breast histological phenotypes were identified, exam-ined for Cx43 gap junctions, and compared to normal breast tissuefrom the same specimen. Transmitted light images (Fig. 1,A, C, E,andG) show clear morphological changes between normal and cancertissue. The punctate Cx43 immunostaining pattern seen in normaltissue (Fig. 1B,inset,arrows) was lost in IDC (Fig. 1D), ILC (Fig.1F), and mixed DCIS/IDC (Fig. 1H). Fig. 1 (A–D) shows differentregions of the same tissue section, and fluorescent images (Fig. 1,BandD) were collected under identical imaging conditions.

To examine the phenotypic origin of the normal and tumor cellscontained within the breast biopsies, we immunolabeled tissue sec-tions for cytokeratin. Tissue sections of a normal human breast (Fig.2, A–C) and an invasive ductal carcinoma (Fig. 2,D–F) taken fromdifferent regions of the same breast were immunofluorescently dou-ble-labeled for Cx43 and an intermediate filament protein, cytokera-tin. Punctate Cx43 immunolabeling was seen in normal (Fig. 2,A andC) but not tumor (Fig. 2,D andF) tissue. Cytokeratin labeling, whichis used as a marker for cells of epithelial origin, was seen in bothnormal (Fig. 2,B andC) and tumor (Fig. 2,E andF) tissue, confirm-ing the epithelial phenotype of the breast carcinoma.

Table 1 summarizes the expression of Cx43 gap junctions in allhuman breast tissues tested. In all cancer tissues, regardless of theirgrade, distinct Cx43 gap junctions were undetectable, although theywere always present, at various levels, in normal tissues. It is plausiblethat some of these tumors were expressing a low level of Cx43, but theCx43 was clearly not assembled into classical gap junctions. Theexpression of Cx43 gap junctions in normal tissue varies from mod-erate to high expression. To examine whether the absence of Cx43 gapjunctions is correlated with the expression of ER, PGR, and/or erbB-2tyrosine kinase receptor, we immunostained tissue sections from thesame paraffin blocks to evaluate the presence of these three commonprognostic markers associated with breast cancer. No correlationbetween Cx43 gap junctions and ER, PGR, and/or erbB-2 status wasfound in the breast tissues used (Table 1).

To confirm and further examine Cx43 gap junction expression inbreast cancer, we performed both Northern and Western blots on apanel of human breast cancer cell lines. As indicated in Fig. 3, bothCx43 protein and mRNA were significantly down-regulated in allbreast cancer cells lines except Hs578T and MDA-231. This lack, orlow level, of Cx43 mRNA expression in MDA-468, HBL-100,T47-D, and ZR-75 cells was not due to gross gene rearrangement asmeasured by Southern blot analysis (data not shown).

To examine whether Cx43 down-regulation occurs in murine mam-mary cancerin vivo, we examined the expression of Cx43 in sixindependent rat mammary tumors induced by DMBA. As indicated inFig. 4, both Cx43 mRNA and protein are barely detectable in cancertissues, but they are highly expressed in normal rat mammary tissues.

DISCUSSION

Primary breast cancer is generally comprised of tumor cells andsurrounding connective tissue. This arrangement creates multiple cell-cell interactions among tumor cells and between tumor cells andnormal neighboring stromal cells. Among various patterns of cell-cellinteractions, GJIC involving Cx43 is considered among the earliestalterations associated with malignant cell transformation (6–11). Lossof Cx43 has been shown to correlate with tumorigenesis (33), andup-regulation of Cxs has been shown to restore normal phenotypesinvitro and reduce tumor growthin vivo (11, 34, 35). Both Cx26 andCx43 suppressed the cancer phenotype in MDA-MB-435 humanmammary carcinoma cells and suppressed cell growth in culture and

4106




in animal models (35). Together, these studies suggest that at leastsome Cxs act as tumor suppressors (11, 34, 35).

This study demonstrates that down-regulation of Cx43 gap junc-tions, which is expressed in normal breast epithelium, is a commonfeature of breast cancer. Cx43 gap junction down-regulation wasobserved in tissue at various stages of tumorigenesis, including DCIS,invasive IDC, and ILC, supporting earlier studies that suggested thatthe loss of Cx43 is an early event in carcinogenesis (reviewed in Ref.17). Furthermore, data obtained with rat mammary carcinoma inducedby DMBA also demonstrate that the loss of Cx43 gap junctions is acommon feature of mammary neoplastic transformation. Down-regu-

lation of Cxs has been reported in other tumor cells, including prostate(7), liver (36), and lung carcinomas.4

Consistent with our study, Wilgenbuset al. (37) failed to identifyCx43 in the parenchymal components of the seven breast tumors theyexamined. However, Jamiesonet al. (38) found mainly heterogeneouscytoplasmic Cx43 immunostaining in carcinoma cells of;50% of thetumors examined, and in some cases, Cx43 was organized into whatappeared to be gap junctions. In our studies, the samples were scored

4 Unpublished data.

Fig. 1. Expression of Cx43 gap junctions in hu-man breast tumorsin vivo. Human breast biopsieswere fixed, embedded in paraffin, sectioned, andimmunofluorescently labeled for Cx43. Neoplasticbreast tissue was examined for Cx43 expression, andcompared to normal breast tissue. Transmitted lightimages (A,C, E, andG) show clear morphologicalchanges among normal (A), IDC (C), ILC (E), andDCIS (G). The punctate Cx43 immunostaining pat-tern seen in normal tissue (B) was lost in IDC (D),ILC (F), and mixed DCIS/IDC (H) carcinomas. Im-ages inA–D were obtained from different regions ofthe same tissue section, and fluorescent images (BandD) were collected under identical imaging con-ditions.Scale bar, 25mm.

4107




based on the presence of assembled gap junctions and not on thepossibility that the carcinoma cells may express low levels of unas-sembled Cx43 in the cytoplasm. In essence, if some human breasttumors express Cx43in vivo but fail to assemble gap junctions, thesetumors would not be capable of Cx43-mediated cell-cell communica-tion. It is possible that the defect in some breast tumor cells is theirinability to traffic and assemble Cx43 into gap junctions at the cellsurface. Together, these studies suggest that the majority of breastcarcinoma cells are unlikely to communicate via Cx43 gap junctions.Although Cx26 has been demonstrated to be a minor Cx in normalhuman breasts, a study by Jamiesonet al. (38) provided evidence thatit might be up-regulated in some breast tumors. However, similar toCx43, the immunostaining of Cx26 was mostly cytoplasmic. Jamiesonet al. (38) suggested that tumor cells may not be communicating viaCx26 gap junctions. Nevertheless, in future studies, it will be impor-tant to examine the regulation of Cx26 in breast carcinomas.

There are several markers associated with breast cancer progressionand/or poor prognosis, including loss of heterozygosity at specificloci, loss of ERs, overexpression of specific member of the erbBtyrosine kinase receptors, and p53 mutations (reviewed in Refs. 39–41). In our study, the absence of Cx43 gap junctions was observed inbreast cancer tissues, regardless of their ER, PGR, or erbB-2 status,suggesting that Cx43 may be an independent marker for breast car-cinogenesis. A recent study suggested that the loss of p53 function

occurs at early stages of breast cancer progression, whereas overex-pression of ras may be a late-stage event (42). On the basis of ourstudy of precursors for invasive breast cancer, Cx43 regulation mayoccur at an early stage because Cx43 gap junctions were not observedin the DCIS samples examined. Jamiesonet al. (38) reported Cx43immunostaining in the myoepithelium of four DCIS tumors, but nostaining of the carcinoma cells themselves was observed. However,additional screening of a large panel of DCIS is required to draw afinal conclusion. The lack of any correlation between ER, erbB-2,PGR, and Cx43, however, does not rule out a possible interactionbetween these signaling pathways. Indeed, overexpression of erbB-2as well as several growth factors, including epidermal growth factor,has been reported to regulate Cx43 expression (43, 44). Hormonalregulation of GJIC has been demonstrated in numerous studies (45–48). In myometrium, estrogen up-regulates the expression of Cx43mRNA, whereas progesterone antagonizes this increase (45, 46).Examination of the upstream noncoding region of Cx43 has revealedputative estrogen responsive elements (46). Estrogens induce tran-scriptional up-regulation of Cx43 (31, 48) but down-regulate erbB2mRNA and protein and inhibit erbB-2 transcription in ER-positivebreast tumor cells (49, 50), suggesting possible cross-talk among theseregulatory pathways.

Although the molecular mechanisms by which Cx43 gap junctionsare down-regulated in cancer are unknown, data obtained on cell lines

Fig. 2. Human breast carcinomas express cyto-keratin. Tissue section of a normal human breast(A–C) and an invasive ductal carcinoma (D–F) takenfrom different regions of the same slide were im-munofluorescently double-labeled for Cx43 and anintermediate filament protein, cytokeratin. Cx43 im-munolabeling was seen in normal (A andC) but nottumor (D and F) tissue. Cytokeratin labeling wasseen in both normal (B andC) and tumor (E andF)tissue, confirming the epithelial origin of the breastcarcinoma.Scale bar, 25mm.

4108




indicate that the frequent absence of Cx43 protein is correlated withdecreased mRNA expression and the loss of GJIC (i.e., HBL-100cells).4 Interestingly, Hs578T cells express abundant levels of Cx43,which is assembled into functional gap junctions, whereas MDA-231cells express significant amounts of Cx43, but these cells assemblefew gap junctions and remain communication-deficient.4 These re-sults suggest that there is considerable GJIC heterogeneity in humanbreast carcinomas cellsin vitro. It is possible that MDA-231 cellsrepresent a subtype of human carcinomasin vivo where low levels ofCx43 are expressed but inefficiently assembled into gap junctions. Nogene rearrangement was found in the various human breast cancer celllines tested (data not shown); however, this does not exclude the

occurrence of mutations that alter transcription of the gene. Bothtranscriptional and posttranscriptional regulation of Cx43 have beenreported (Ref. 16, reviewed in Ref. 20). Cx43 can be phosphorylatedby protein kinase C and mitogen-activated protein kinases, whereasthe phosphorylation of Cx43 by v-Src tyrosine protein kinase resultsin the inhibition of GJIC (16). The ras oncoprotein is also known todisrupt GJIC, possibly through the activation of mitogen-activatedprotein kinase (16). These cell signaling pathways are coupled toseveral growth factor receptors including erbB-2, and it is possiblethat coordinated regulation pathways play a role in Cx43 down-regulation.

In summary, this study demonstrates decreased expression of Cx43gap junctions in breast cancer at various stages of progression. Thelack of Cx43 gap junctions in breast cancer may serve as a diagnosticmarker for breast carcinogenesis and may act as a potential therapeu-tic target. Earlier studies indicated that restoration of Cx43 in cellslacking Cx43 can reverse the transformed phenotype (34, 35), sug-gesting that a strategy to express functional Cx43 may be of thera-peutic value. Furthermore, we and others have reported that therestoration of Cx43 expression or a chemical induction of Cx43 intumor cells can enhance the bystander effect in gene therapy ap-proaches using suicide gene prodrug targeting (51, 52). Together,these observations indicate that specific modulators of Cx43 may havetherapeutic implications in cancer.

REFERENCES

1. Anderson, J. M., Bald, M. S., and Fanning, A. S. The structure and regulation of tightjunctions. Curr. Opin. Cell Biol.,5: 772–778, 1993.

2. Laird, D. W. The life cycle of a connexin: gap junction formation, removal, anddegradation. J. Bioenerg. Biomembr.,28: 311–318, 1996.

3. Gallin, W. J. Evolution of the classical cadherin family of cell adhesion molecules invertebrates. Mol. Biol. Evol.,15: 1099–1107, 1998.

4. Schwartz, M. A., Owaribe, K., Kartenbeck, J., and Franke, W. W. Desmosomes andhemidesmosomes: constitutive molecular components. Annu. Rev. Cell Biol.,6:461–491, 1991.

5. Simionescu, N., and Simionescu, M. Endothelial transport macromolecules: transcy-tosis and endocytosis. Cell Biol. Rev.,25: 5–80, 1991.

6. Loewenstein, W. R. Junctional intercellular communication. The cell to cell mem-brane channel. Physiol. Rev.,61: 829–913, 1981.

7. Mehta, P. P., Lokeshwar, B. L., Schiller, P. C., Bendix, M. V., Ostenson, R. C.,Howard, G. A., and Roos, B. A. Gap-junctional communication in normal andneoplastic prostate epithelial cells and its regulation by cAMP. Mol. Carcinog.,15:18–32, 1996.

8. Mehta, P. P., Bertram, J. S., and Loewenstein, W. R. Growth inhibition of trans-formed cells correlates with their junctional communication with normal cells. Cell,44: 187–196, 1986.

9. Nicolson, G. L. Cancer metastasis: tumor cell and host organ properties important inmetastasis to specific secondary sites. Biochim. Biophys. Acta,948: 175–224, 1988.

10. Zhu, D., Caveney, S., Kidder, G. M., and Naus, C. C. G. Transfection of C6 gliomacells with connexin43 cDNA: analysis of expression, intercellular coupling, and cellproliferation. Proc. Natl. Acad. Sci. USA,88: 1883–1887, 1991.

11. Naus, C. C. G., Elisevich, K., Zhu, D., Belliveau, D. J., and Del Maestro, R. F.In vivogrowth of C6 glioma cells transfected with connexin43 cDNA. Cancer Res.,52:4208–4213, 1992.

12. Brissette, J. L., Kumar, N. M., Gilula, N. B., and Dotto, P. G. The tumor promoter12-O-tetradecanoylphorbol-13-acetate and therasoncogene modulate expression andphosphorylation of gap junction protein. Mol. Cell. Biol.,11: 5364–5373, 1991.

13. Oh, S. Y., Grupen, C. G., and Murray, A. W. Phorbol ester induces phosphorylationand down regulation of connexin 43 in WB cells. Biochim. Biophys. Acta,1094:243–245, 1991.

14. Berthoud, V. M., Ledbetter, M. L. S., Hertzberg, E. L., and Saez, J. C. Connexin43in MDCK cells: regulation by a tumor-promoter phorbol ester and Ca11. Eur. J. CellBiol., 57: 40–50, 1992.

15. Lau, A. F., Kanemitsu, M. Y., Kurata, W. E., Danesh, S., and Boynton, A. L.Epidermal growth factor disrupts gap-junctional communication and induces phos-phorylation of connexin43 on serine. Mol. Biol. Cell,3: 865–874, 1992.

16. Lau, A. F., Kurata, W. E., Kanemitsu, M. Y., Loo, L. W. M., Warn-Cramer, B. J.,Eckhart, W., and Lampe, P. D. Regulation of connexin43 function by activatedtyrosine protein kinases. J. Bioenerg. Biomembr.,28: 359–368, 1996.

17. Yamasaki, H., and Naus, C. C. G. Role of connexin genes in growth control.Carcinogenesis (Lond.),17: 1199–1213, 1996.

18. Flagg-Newton, J., Simpson, T., and Loewenstein, W. R. Permeability of the cell-to-cell membrane channels in mammalian cell junctions. Science (Washington DC),205: 404–407, 1997.

19. Bruzzone, R., White, T. W., and Paul, D. L. Connections with connexins: themolecular basis of direct intercellular signaling. Eur. J. Biochem.,238: 1–27, 1996.

Fig. 3. Expression of Cx43 in human breast cancer cell lines. mRNA and total proteinswere isolated from Hs578T (Lanes 1), MDA-231 (Lanes 2), MDA-468 (Lanes 3),HBL-100 (Lanes 4), T47-D (Lanes 5), and ZR-75 (Lanes 6) cell lines, as described in“Materials and Methods.”A, a Northern blot probed for Cx43 mRNA expression.B, thesame blot probed for GAPDH mRNA.C, a Western blot immunolabeled for Cx43 proteinexpression.D, the same blot probed for the expression of glyceraldehyde-3-phosphatedehydrogenase protein.

Fig. 4. Expression of Cx43 in rat mammary cancer. Rat mammary cancer was inducedin individual animals by exposure to the chemical carcinogen DMBA. Mammary cancertissue and normal mammary tissue were used to isolate proteins and mRNA, as describedin “Materials and Methods.”A, a Northern blot for Cx43 mRNA expression in normal(Lanes 1and 2) and cancer tissues (Lanes 3–8).B, the same blot probed for GAPDHmRNA. C, a Western blot immunolabeled for Cx43 protein expression in normal (Lanes1 and2) and cancer tissues (Lanes 3–8). D, the same blot probed for the expression ofGAPDH protein. Each lane represents an individual animal.

4109




20. Goodenough, D. A., Goliger, J. A., and Paul, D. L. Connexins, connexons, andintercellular communication. Annu. Rev. Biochem.,65: 475–502, 1996.

21. Laird, D. W., Puranam, K., and Revel, J-P. Turnover and phosphorylation dynamicsof connexin 43 gap junction protein in cultured cardiac myocytes. Biochem. J.,272:67–72, 1991.

22. Wang, Y., and Rose, B. Clustering of Cx43 cell to cell channels into gap junctionplaques: regulation by cAMP and microfilaments. J. Cell Sci.,108:3501–3508, 1995.

23. Atkinson, M. M., Lampe, P. D., Lin, H. H., Kollander, R., Li, X-R., and Kiang, D. T.Cyclic AMP modifies the cellular distribution of connexin43 and induces a persistentincrease in the junctional permeability of mouse mammary tumor cells. J. Cell Sci.,108: 3079–3090, 1995.

24. De Mello, W. Effect of intracellular injection of cAMP on the electrical coupling ofcardiac cells. Biochem. Biophys. Res. Commun.119: 1001–1007, 1984.

25. Rogers, M., Berestecky, J. M., Hossain, M. Z., Guo, H., Kadle, R., Nicholson, B. J.,and Bertram, J. S. Retinoid-enhanced gap junctional communication is achieved byincreased levels of connexin43 mRNA and protein. Mol. Carcinog.,3: 335–343,1990.

26. Crow, D. S., Beyer, E. C., Paul, D. L., Kobe, S. S., and Lau, A. F. Phosphorylationof connexin43 gap junction protein in uninfected and Rous sarcoma virus-transformedmammalian. Mol. Cell. Biol.,10: 1754–1763, 1990.

27. Kumar, N. M., and Gilula, N. B. The gap junction communication channel. Cell,84:381–388, 1996.

28. Fishman, G. I., Eddy, R. L., Shows, T. B., Rosenthal, L., and Leinwand, L. A. Thehuman connexin gene family of gap junction proteins: distinct chromosomal locationsbut similar structures. Genomics,10: 250–256, 1991.

29. Monaghan, P., Clarke, C., Perusinghe, N. P., Moss, D. W., Chen, X-Y., and Evans,W. H. Gap junction distribution and connexin expression in human breast. Exp. CellRes.,223: 29–38, 1996.

30. Monaghan, P., Perusinghe, N., Carlile, G., and Evans, W. H. Rapid modulation of gapjunction expression in mouse mammary gland during pregnancy, lactation and invo-lution. J Histochem. Cytochem.,42: 931–938, 1994.

31. Hendrix, E. M., Myatt, L., Sellers, S., Russell, P. T., and Larsen, W. J. Steroidhormone regulation of rat myometrial gap junction formation: effects on Cx43 levelsand trafficking. Biol. Reprod.,52: 547–560, 1995.

32. Laird, D. W., Castillo, M., and Kasprzak, L. Gap junction turnover, intracellulartrafficking and phosphorylation of connexin43 in brefeldin A-treated rat mammarytumor cells. J. Cell Biol.,131: 1193–1203, 1995.

33. Naus, C. C. G., Bechberger, J. F., Caveney, S., and Wilson, J. X. Expression of gapjunction genes in astrocytes and C6 glioma cells. Neurosci. Lett.,112: 33–36, 1991.

34. Mehta, P. P., Hotz-Wagenblatt, A., Rose, B., Shalloway, D., and Loewenstein, W. R.Incorporation of the gene for a cell-cell channel protein into transformed cells leadsto normalization of growth. J. Membr. Biol.,124: 207–225, 1991.

35. Hirschi, K. K., Xu, C., Tsukamoto, T., and Sager, R. Gap junction genesCx26andCx32individually suppress the cancer phenotype of human mammary carcinoma cellsand restore differentiation potential. Cell Growth Differ.,7: 861–870, 1996.

36. Krutovskikh, V., Mazzoleni, G., Mironov, N., Omori, Y., Agueloni, A-M., Mesnil,M., Berger, F., Partensky, C., and Yamasaki, H. Altered homologous and heterolo-gous gap-junctional intercellular communication in primary liver tumors associatedwith aberrant protein localization but no gene mutation of connexin 32. Int. J. Cancer,56: 87–94, 1994.

37. Wilgenbus, K. K., Kirkpatrick, C. J., Knuechel, R., Willecke, K., and Traub, O.Expression of Cx26, Cx32 and Cx43 gap junction proteins in normal and neoplastichuman tissues. Int. J. Cancer,51: 522–529, 1992.

38. Jamieson, S., Going, J. J., D’arcy, R., and George, D. Expression of gap junctionproteins connexin26 and connexin43 in normal human breast and in breast tumors.J. Pathol.,184: 37–43, 1998.

39. Allred, D., O’Connell, P., Fuqua, S., and Osborne, C. Immunohistochemical studiesof early breast cancer evolution. Breast Cancer Res. Treat.,32: 13–18, 1994.

40. O’Connell, P., Pekkel, V., Fuqua, S., Osborne, C., and Allred, D. Molecular geneticstudies of early breast cancer evolution. Breast Cancer Res. Treat.,32: 5–12, 1994.

41. Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., and McGuire,W. L. Human breast cancer: correlation of relapse and survival with amplification ofthe HER2/neuoncogene. Science (Washington DC),235: 177–182, 1987.

42. Shackney, S. E., Pollice, A. A., Smith, C. A., Janocko, L. E., Sweeney, L., Brown,K. A., Singh, S. G., Gu, L., Yakulis, R., and Lucke, J. F. Intracellular coexpressionof epidermal growth factor receptor, Her-2/neu, and p21ras in human breast cancers:evidence for the existence of distinctive patterns of genetic evolution that are commonto tumors from different patients. Clin. Cancer Res.,4: 913–928, 1998.

43. Madhukar, B. V., Oh, S. Y., Chang, C. C., Wade, M., and Trosko, J. E. Alteredregulation of intercellular communication by epidermal growth factor, transforminggrowth factor-b and peptide hormones in normal human keratinocytes. Carcinogen-esis (Lond.),10: 13–20, 1989.

44. Hofer, A., Saez, J. C., Chang, C. C., Trosko, J. E., Spray, D. C., and Dermietzel, R.C-erbB2/neu transfection induces gap junctional communication incompetence inglial cells. J. Neurosci.,16: 4311–4321, 1996.

45. Lye, S. J., Nicholson, B. J., Mascarenhas, M., MacKenzie, L., and Petrocelli, T.Increased expression of connexin43 in the rat myometrium during labor is associatedwith an increase in the plasma estrogen: progesterone ratio. Endocrinology,132:2380–2386, 1993.

46. Petrocelli, T., and Lye, S. J. Regulation of transcripts encoding the myometrial gapjunction protein, connexin43, by estrogen and progesterone. Endocrinology,133:284–290, 1993.

47. Lefebve, D. L., Piersanti, M., Bai, X. H., Chen, Z. Q., and Lye, S. J. Myometrialtranscriptional regulation of the gap junction gene, connexin 43. Reprod. Fertil. Dev.,7: 603–611, 1995.

48. Yu, W., Dahl, G., and Werner, R. The connexin 43 gene is responsive to oestrogen.Proc. R. Soc. Lond. Ser. B. Biol. Sci.,255: 125–132, 1994.

49. Antoniotti, S., Taverna, D., Perroteau, I., and De Bortoli, M. Inhibition of c-erbB-2oncogene expression by estrogens in human breast cancer cells. Oncogene,5:1001–1006, 1990.

50. Russell, K. S., and Hung, M. C. Transcriptional repression of the neu proto-oncogeneby estrogen stimulated estrogen receptor. Cancer Res.,52: 6624–6629, 1992.

51. Carystinos, G. D., Katabi, M., Laird, D., Galipeau, J., Chan, H., Alaoui-Jamali, M. A.,and Batist, G. Cyclic-AMP induction of gap junctional intercellular communicationincreases bystander effect in suicide gene therapy. Clin. Cancer Res.,5: 61–68, 1999.

52. Dilber, M. S., Abedi, M. R., Christensson, B., Bjorkstrand, B., Kidder, G. M., Naus,C. C. G., Gahrton, G., and Smith, C. I. E. Gap junctions promote the bystander effectof herpes simplex virus thymidine kinasein vivo. Cancer Res.,57: 1523–1528, 1997.

4110




1999;59:4104-4110. Cancer Res Dale W. Laird, Paulina Fistouris, Gerald Batist, et al. Marker for Breast TumorsDeficiency of Connexin43 Gap Junctions Is an Independent

Updated version

http://cancerres.aacrjournals.org/content/59/16/4104

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/59/16/4104.full#ref-list-1

This article cites 51 articles, 15 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/59/16/4104.full#related-urls

This article has been cited by 21 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/59/16/4104To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/59/16/4104.full#ref-list-1

http://cancerres.aacrjournals.org/content/59/16/4104.full#related-urls

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

mailto:[email protected]



deficiency of connexin43 gap junctions is an independent marker … · deficiency of connexin43 gap...

Documents