[CANCER RESEARCH 49. 699-705, Februar) l, 1989|

Gap-Junctional Intercellular Communication in Epidermal Cell Lines from SelectedStages of SENCAR Mouse Skin Carcinogenesis1

Richard C. Mann, ' D. James Fitzgerald, Colette Piccoli, Thomas J. Slaga, and Hiroshi Yamasaki

Programme of Multistage Carcinogenesis, International Agency for Research on Cancer, ISO, cours Albert-Thomas, 69372 Lyon, France fD. J. F., C. P., H. Y.], and TheUniversity of Texas M. D, Anderson Cancer Center, Science Park-Research Division, Smithville, Texas 78957 [R. C. K., T. J. S.J

ABSTRACT

Homologous and heterologous gap-junctional intercellular communication (1C) was characterized in a panel of cell lines derived from selectedstages of SENCAR mouse skin carcinogenesis. This panel included a"carcinogen-altered" cell line, 3PC, obtained from Ca2*-resistant primary

adult keratinocytes after exposure to dimethylhenz(a)anthracene as wellas cell lines obtained from early and late-stage papillomas and a squamouscell carcinoma (CA3/7) generated during standard in vivo initiation/promotion protocols (dimethylbenz(a)anthracene/! 2-O-tetradecanoyl-phurbol-13-acetate). Also studied was a cell line (B66BA) obtained froma metastatic lesion following benzo(o)pyrene-induced skin tumorigenesis.Intercellular communication was measured in low-calcium (0.05 HIM)medium by quantitation of cell-cell transfer of microinjected fluorescentdye Lucifer Yellow CM. Homologous 1C ability diminished progressivelyfrom 68 dye-coupled cells per injection for 3PC cultures, to between 21and 54 dye-coupled cells per injection for three papilloma-derived celllines, to six and three dye-coupled cells per injection for CA3/7 andB66BA cells, respectively. To test communication of these cells withtheir normal counterparts, heterologous 1C was examined in cocultureswith primary adult keratinocytes. Under the conditions used, normal cellsestablished functional communication channels with each cell line tested,showing no selectivity. These results suggest that progressive loss ofhomologous but not heterologous 1C capacity accompanies neoplasticdevelopment in mouse skin carcinogenesis.

INTRODUCTION

The direct cell-cell exchange of low-molecular-weight cellularcomponents (M, < 1000) that occurs via gap junctions isconsidered to play a pivotal role in the fundamental biologicalprocesses of development, tissue homeostasis, and differentiation (1-3). Not surprisingly, then, since cancer is essentially aderangement of these processes, the relationship between tumorigenesis and gap-junctional intercellular communication(1C)3 has been the subject of many studies. In early studies on

tumor tissues, ionic coupling was measured to assay functionalcell-cell communication, and gap junction structure was examined by electron microscopy. These studies revealed that manytumors are deficient in gap-junctional communication (4-6) orhave fewer or qualitatively different gap junctions when compared with surrounding normal tissue (7, 8). These alterationsare not seen in all tumors, however (7), just as tissue culturestudies show that many but not all tumor cells are 1C incompetent (2, 9-11).

Aside from diminished intrinsic 1C competence of neoplastic

Received 8/15/88; revised 10/27/88; accepted 11/2/88.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' This research was supported in part by Grant ROl-CA-40534, National

Cancer Institute, USPHS. Preliminary results of the work were presented at the38th Annual Meeting of the Tissue Culture Association, Washington, DC, 1987(see Ref. 56).

2Supported by the International Research Data Bank Programme underContract N01-CO-65341 (International Cancer Research Technology Transfer),National Cancer Institute, and the International Union Against Cancer. To whomrequests for reprints should be addressed.

3The abbreviations used are: 1C, gap-junctional intercellular communication:DMBA, dimethylbenz(a)anthracene; TPA, 12-O-tetradecanoylphorbol-13-ace-tate; FAV, fibronectin plus albumin plus vitrogen.

cells, other modes of perturbation of 1C include briefly: (a)defective responses of tumor cells to signals that modulate 1C,as shown for mouse mammary tumor cells (12); (b) disturbed1C in normal tissue surrounding a tumor, as in the case ofNovikoff hepatoma (4), implying an influence of some factorproduced by the tumor on 1C of normal cells; (c) alteredspecificity of 1C of tumor cells, as shown by mammary cells(13), among others (14, 15), which acquire with transformationan ability to establish gap junctions with contacting fibroblastsboth in vitro and in vivo, a change that may then allow exploitation of growth-stimulatory signals from this abnormal source;

and (d) establishment of selective 1C. Recent experimentsshowed that in situ foci formed by carcinogen- or oncogene-

transformed BALB/c 3T3 cells display an internal 1C capacitysimilar to that of surrounding normal cells but do not communicate with them (16-18). This selective 1C could be abolishedwith such agents as retinoic acid and dibutyryl cyclic AMP andresulted in regression and disappearance of the transformedfoci (19). Selective 1C was also observed in coculture experiments with tumorigenic rat liver epithelial cells and nontumo-rigenic counterparts (20) and with pairs of normal/abnormalC3H10T'/2 cells (21). These studies suggested that the estab

lishment of selective 1C in an emerging neoplasm may aid theescape of tumor cells from controls on proliferation and differentiation normally imposed within a cell society. In addition tothese findings of alterations of 1C characteristics in neoplasticpopulations, the large body of data demonstrating inhibition of1C by many tumor promoters adds strength to the contentionthat disturbed 1C may play an important role in carcinogenesis(22-25).

Although 1C of normal and fully neoplastic populations hasbeen compared in many cancer model systems, little information is available on 1C changes during the progressive stages ofneoplastic development (26, 27). Among the many model systems available, mouse skin presents an attractive choice forsuch studies because of its well-characterized stages of initiationand promotion (28). Recently, we established cell lines frompapillomas and carcinomas of SENCAR mice after applying atreatment protocol of DMBA initiation and TPA promotion.4

We also used a cell line derived from a lymph node metastasis,as well as a "carcinogen-altered" cell line generated after ex

posure of adult mouse primary keratinocytes to DMBA andfollowed by selection in high Ca2+ medium. Acquisition of

resistance to the differentiation-inducing effect of calcium hasbeen suggested to be the critical event associated with initiationof carcinogenesis in mouse skin (29). In this report, we describethe 1C characteristics of these cell lines derived during selectedstages of carcinogenesis in mouse skin. We used fluorescentdye microinjection techniques to assay homologous 1C in thesecell lines and heterologous 1C in cocultures established betweenthem and primary adult keratinocytes.

' R. C. Mann. B. S. Beilue, C. L. Haskin, and T. J. Slaga. Establishment and

characterization of cell lines from SENCAR mouse papillomas and carcinomas,submitted for publication.

699

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

MATERIALS AND METHODS

Cell Culture. The base medium used for primary and established cellline cultures was Eagle's minimal essential medium prepared from

individual ingredients with the omission of CaCl2. Components (SigmaChemical Co., St. Louis, MO) added to this Ca2+-free medium were:

insulin (5 Mg/ml final concentration); epidermal growth factor (5 ng/ml); transferrin (10 Mg/ml); O-phosphorylethanolamine (10 MM);2-

aminoethanol (10 MM);hydrocortisone (1 MM);gentamicin sulfate (50Mg/ml); and glutamine (Gibco; 2 mM). This medium was then supplemented with 8% fetal bovine serum previously Chelex treated (Chelex-100 resin, Na* form; BioRad Labs, Richmond, CA) to remove calcium(30). According to atomic absorption analysis, the Ca2+ concentration

of this medium was approximately 0.03 mM. Calcium chloride wasthen added to give a final Ca2+concentration of 0.05 mM, and this low-calcium medium was used for most experiments. When high-calciummedium was required, CaCl2 was added to a final concentration of 1.2mM.

Cell cultures were established in 60-mm dishes (Falcon) precoatedjust before use with FAV solution, essentially according to Kawamuraet al. (31). To each dish were added 2.0 ml of a solution of bovinefibronectin (10 Mg/ml; Sigma), bovine serum albumin (0.1 mg/ml;Sigma), collagen (30 Mg/ml; Collagen Corp., Palo Alto, CA), and 20mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonicacid in Ca2+, Mg2"1"-

free phosphate-buffered saline. The solution was then aspirated and thedishes left to air dry for 10 to 15 min before cell seeding. All incubationswere conducted at 37°Cin a humidified atmosphere of 5% CÛ2.

Primary Cell Cultures. Cultures of primary keratinocytes were obtained from 6- to 9-wk-old female SENCAR mice (Frederick CancerResearch Facility, Frederick, MD) as described previously (32). Typicalcell yield and viability, determined by trypan blue dye exclusion, were15 x IO6 to 30 x IO6 cells/dorsal skin with viability of 60 to 80%.Viable cells (10 x IO6to 30 x IO6)were seeded in 4 ml of medium onto

FAV-coated dishes; after 24 h, cultures were refed with fresh mediumand used for dye transfer studies 2 to 3 days later.

SENCAR Mouse Epidermal Cell Lines. The cell lines used in thisstudy were considered to represent early, intermediate, and late stagesof mouse skin carcinogenesis. The 3PC cell line was obtained afterexposure of adult primary keratinocytes in vitro to DMBA (0.1 Mg/ml,24 h) and a shift to high-Ca2+ medium to select for Ca2+-resistant cells.Other investigators have similarly generated Ca2*-resistant carcinogen-

altered keratinocytes following in vitro or in vivo carcinogen exposure(29, 32, 33). Papilloma-derived cell lines were obtained from tumorsappearing after DMBA initiation of dorsal skin of 7-wk-old femaleSENCAR mice followed by brief (PI/17) or extended (P2/15, P3/22)promotion with TPA according to the scheme depicted in Fig. 1. TheCA3/7 cell line was obtained from a carcinoma 16 wk after completionof the DMBA/TPA regimen. Upon reintroduction in vivo in implantation chambers, the histological characteristics of the epithelia formedby these tumor cell lines reflected those of the original tumors; thepapilloma-derived lines produced hyperplastic and hyperkeratotic epithelia, P3/22 in addition to displaying limited invasion of granulationtissue, whereas CA3/7 cells produced highly invasive squamous cellcarcinomas.4 The cell line B66BA was obtained from a lymph node

metastasis of a mouse bearing a squamous cell carcinoma followingcomplete carcinogenesis with benzo(a)pyrene. All cell lines were maintained in Iow-Ca2+ medium and were used at passages 7 to 32.

Bead Labeling and Coculture (Heterologous Culture) of Cells. Todistinguish cell types in cocultures, one partner was prelabeled bypreincubating subconfluent cells with fluorescent latex beads (Fluores-brite, Bright Blue, 1.17-/im average diameter, 4.55 x 1010/ml; Polysci-

ences, Inc., Warrington, PA). Exposure of cultures at 60 to 80%confluence overnight to medium containing 4.55 x IO7beads/ml caused

most cells to phagocytose at least 10 beads, making them easily identifiable in subsequent cocultures. Cells were then rinsed 5 times withphosphate-buffered saline and dissociated, and Viopart was added tointact cultures (60 to 90% confluent) of the second cell type. Theresulting coculture was examined for heterologous communication after8.5 to 26 h. The cocultures were prepared from prelabeled cell linesadded to primary keratinocytes and from prelabeled primary keratinocytes added to cell line cultures.

CARCINOGEN(Low dose)

PROGRESSION METASTASIS

CARCINOGEN(Low dose)

CULTURED m^^^mCELLS

(Low C.ili-min! (High Calcium) DMBA

Fig. 1. Scheme of protocols used to generate the six cell lines studied andshowing the approximate temporal appearance of the various grade tumors intwo-stage mouse skin carcinogenesis. For initiation/promotion in SENCAR mice,10 nmol of DMBA were followed after 1 wk by twice-weekly TPA doses (2 Mg)for up to 20 wk. The cell lines PI/17, P2/15, and P3/22 were established frompooled papillomas from 1,4, and 5 mice, respectively. The cell line B66 (of whichB66BA is a clone) was obtained from a lymph node metastasis sampled aftercarcinoma formation induced by multiple benzo(a)pyrene (BaP) treatments.

Assay of Intercellular Communication. Gap-junctional intercellularcommunication was assayed using the method of Lucifer Yellow dyemicroinjection (34, 35). Injection needles were prepared from microfil-ament glass (A-M Systems, Everett, WA) using a dual-step puller(Narishige Co., Tokyo, Japan). Needles were then filled with a solutionof 10% Lucifer Yellow CH (Sigma) in 0.33 M LiCl and fixed in thecondensor of an inverted microscope (Olympus Injectoscope IMT-2SYF). With the aid of an automatic (pneumatic) microinjector (Eppen-dorf, Model 5242), individual cells were injected with the fluorescentdye, and the extent of cell-cell spread of the dye was observed underfluorescence 10 to 15 min after injection. Communication in homologous cultures was tested 24 h after addition of fresh medium, and 15to 30 independent injections were made for each experimental point.To study heterologous communication in cocultures, first- or second-order neighbors of a bead-labeled cell were injected, and after 10 to 15min the labeled cell was scored as a dye recipient or not.

RESULTS

Homologous Communication Studies. In choosing the cultureconditions for the present study, two important factors wereculture substrate and medium calcium concentration. Clonalgrowth experiments4 indicated that the tumor-derived cell lines

are not markedly sensitive to the type of substrate (plastic orFAV) or the calcium concentration (0.05 mM or 1.2 mM).Coculture experiments involving primary keratinocytes requiremore exacting conditions, however, because normal adult keratinocytes are difficult to culture on plastic or in the presenceof high-Ca2+ medium (31, 36). In preliminary experiments, the

effect of substrate and calcium concentration on 1C was examined in three of the cell lines; these results are shown in Fig. 2.For 3PC, PI/17, and P2/15 in Iow-Ca2+ medium, 1C was not

significantly affected by substrate, although a slight increasewas apparent for 3PC and PI/17 cultured on FAV-coateddishes. In addition, for 3PC, a shift to high-Ca2* medium did

not alter 1C, regardless of the substrate. Since 1C competenceof these cell lines was generally unaffected by these cultureparameters, we elected to use the conditions favoring primarycell culture, Iow-Ca2+ medium, and FAV substrate for further

experiments.The results shown in Fig. 2 suggested that the intrinsic 1C

capacity of the papilloma lines PI/17 and P2/15 was less thanthat of 3PC, the carcinogen-altered cell line. A comparativestudy of homologous 1C involving all six cell lines is shown inFig. 3. Evident is a progressive decline in 1C competence thatcorrelated with the successive stages of skin carcinogenesis (seealso Fig. 4). Early stage 3PC cells communicated best, transfer-

700

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

100-

M180~|

GO'S«

40-A

•¿�Z20-[Ca2+]

:_JS-1ü,10.05mM 0.05mMFAV

: .^D

3PCü P1/17BP2/15JL1.2

mM 1.2mMi

Culture ConditionFig. 2. Effect of Substrate and calcium concentration on junctional commu

nication in SENCAR mouse skin cell lines. Cells were plated at low density inlow-calcium (0.05 mM) medium on culture dishes with or without FAV coatingand refed after 24 h with low- or high-calcium (1.2 mM) medium. At confluence(4 to 5 days), cell-cell communication was assayed (see "Materials and Methods").

Columns, mean of recipient cells per injection; bars, SE.

la.oSCE

ia

80-

70-

60-

50-

40-

30-

20-

10-

n SPCEl P1/17H P2/15ta P3/22E CA3/7D) B66BA

lnitiated...Papilloma...CarcinomaFig. 3. Junctional communication in cell lines from successive stages of mouse

skin carcinogenesis. Homologous 1C was assayed in low-calcium medium inconfluent cultures 24 h after medium change. Columns, mean of recipient cellsper injection; bars, SE.

ring dye to an average of 68 neighboring cells. Dye transfer incultures of papilloma-derived cells nearly matched this level inearly stage PI/17 papilloma cells (54 recipients per injection),whereas cells from progressively later stages showed a decreasefrom 28 dye-coupled cells per injection in the P2/15 culturesto 22 recipients for the P3/22 cell line. Cell lines representingthe last stages of the process of carcinogenesis, squamous cellcarcinoma (CA3/7) and metastasis (B66BA), showed markedreduction of communication to six and three recipients perinjection, respectively. When assayed with this methodology,primary cultures of normal mouse keratinocytes demonstratedan intermediate 1C capacity, 15 to 30 recipients per injection.Directly comparing this to the 1C of the cell lines may beinappropriate, however, because, in addition to the obviousdifference in their adaptation to culture conditions, the primarycultures also comprise a heterogeneous population of cells.

We further characterized the homologous 1C of cell lines3PC and CA3/7 by assaying communication during the growthof cultures to confluence. As shown in Fig. 5,1C was differentially modulated by degree of confluence in these two cell lines.At subconfluence, CA3/7 cells were communication competent,but this property was progressively lost as the cells approachedthe confluent state. In 3PC cells, conversely, 1C capacity wasinitially relatively low but increased with time.

Heterologous Communication Studies. We next examined thecell lines' ability to communicate with normal keratinocytes.

Considering that preneoplastic and neoplastic cells coexist withnormal cells in vivo, we sought to establish the 1C relationshipbetween cancerous or potentially cancerous cells and surrounding normal cells in vitro. In these coculture experiments it wasnecessary to tag one of the cell types because normal keratinocytes and cell lines in culture cannot be distinguished by definitive morphological characteristics. Tagging was done as described in "Materials and Methods." Labeled and dispersed cell

lines were added to primary cell cultures and allowed variouscoculture intervals. Primary cells located near tagged cells werethen microinjected with Lucifer Yellow, and the latter werenoted as having received the dye or not. As a control, labeledcells were replated alone, and their ability to establish homologous junctional contacts was assessed after the same incubation periods used for the cocultures. In all controls, homologouscommunication was evident. As shown in Fig. 6 and Table 1,these communication-competent labeled cells also displayedheterologous communication with the surrounding lawn ofnormal cells in the cocultures. Dye transfer was observed at 85to 100% of the heterologous sites (Table 1).

To see whether the same was true for heterologous populations containing an excess of cell-line cells rather than normalcells, cocultures were prepared with labeled primary keratinocytes added to near-confluent cultures of 3PC or CA3/7 cells.The results shown in Table 2 indicated that, upon LuciferYellow introduction into transformed cells, nearby normal cellssubsequently received the dye. Thus, under these experimentalconditions, the establishment of heterologous gap junctions wasevident regardless of the ratio of the cell types in coculture orthe nature of the cell type receiving the dye injection. In similarstudies, we also observed heterologous communication in cocultures of normal and transformed human keratinocytes.5

DISCUSSION

The principal findings in the present comparative study ofhomologous 1C in these mouse epidermal cell lines suggestedthat communication capacity decreases in a progressive mannerduring skin carcinogenesis. The communication assay used,however, measured only the number of dye-coupled cells aftereach microinjection. Since the total area of dye transfer wasnot quantitated, the same result could also be obtained if therewere major differences in cell size at confluence among the celllines. In a comparison of two cell populations of widely differentcell size and a similar area of dye transfer, the smaller-celledpopulation would seem to be the more communicative. This isunlikely to be true of our results, however, since no majordifferences in cell size were observed. Moreover, visual assessment of dye-transfer area showed the same differences amongthe cell lines as we noted using the cell number method (seeFig. 4). Future studies using computerized image analysis indetermining the rates of cell-cell dye transfer should clarifyambiguities (37-39).

The implications of the observed pattern of decreased 1C incells from later stages of carcinogenesis are currently not clear.Presumably, as a neoplasm progresses, its constituent cellsexpress greater independence and irreconcilability (23, 40).Insofar as gap-junctional communication is concerned, reducedcross-talk between neoplastic and normal cells could weaken

5 D. J. Fitzgerald et al., unpublished observations.

701

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

Fig. 4. Representative photomicrographs of SENCAR mouse epidermal cell lines showing intrinsic gap-junctional communication capacities in low-calciummedium, a, b, c, phase-contrast photographs of 3PC, Pi/22, and CA3/7 cultures, respectively; d, e,f, corresponding fields viewed under fluorescence 10 min aftermicroinjection of Lucifer Yellow CH into a single cell (*). Bar, 40 /Jin.

the normalizing influences emanating from the surroundingtissue. Such normalizing effects have been observed in variousin vitro and in vivo studies of normal cells placed in contactwith premalignant or malignant cells (41-48). A disruption ofthese influences could occur in situ if heterologous communication between tumor and host cells were abolished, or ifhomologous communication among the cells of an emergingneoplasm were reduced. Our current data are consistent withthe latter hypothesis; however, whether progressive reductionof 1C, as seen in the cell lines studied, represents a cause oreffect of epidermal carcinogenesis cannot be determined from

the results of this study. Decreased homologous 1C capacitywas recently shown to correlate with the degree of transformation in rat liver epithelial cell lines (10) and the degree ofmetastatic potential of rat mammary adenocarcinoma cell lines(49).

The measurements of homologous 1C were conducted inconfluent cultures. Although it is not known which in vitrosituation best simulates intact skin, we assume that, in terms ofcell-cell contact, confluence in vitro is a fair representation.Nonetheless, homologous 1C of 3PC and CA3/7 cells was alsotested at subconfluence (Fig. 5). Under these conditions the

702

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

120-

I 100-

f "H

Õ «Ho0 40-

1 »-

TD3PCB CA3/7

3456Day of Culture

Fig. 5. Effect of culture density on junctional communication in 3PC andCA3/7 cells. Cells were plated at low density in low-calcium medium on FAV-coated dishes and tested at various times thereafter. Test days were chosen tofollow 1C under conditions of from 30 to 100% confluence. In low-densitycultures, the number of contacting cells did not limit the quantitation of dyetransfer.

carcinoma-derived cells were communication competent butrapidly lost 1C ability as confluence was reached. In strikingcontrast, 3PC cultures displayed increased 1C with increasedcell density. This finding clearly indicated that these cell linesdiffer greatly in the regulation of communication. Althoughthese differences cannot be explained on the basis of the findingspresented, an association with carcinogenesis is possible. Thetransition from good to poor communication in CA3/7 cellsmay involve a fundamental pathway of 1C regulation, for example, increased synthesis of some factor, knowledge of which

could help to uncover new methods of cancer prevention. Preliminary experiments have suggested that CA3/7 cells cocul-tured with a large excess of 3PC cells can decrease the 1Ccapacity of the latter. Whether this decrease is mediated by anagent that is secreted, transmitted through gap junctions, or isacting through some other mechanism is not yet known. Sucha phenomenon was described by Loewenstein and Kanno (4),who found reduced 1C capacity in normal cells surroundinginduced or transplanted liver tumors.

Our experiments suggested progressive reduction of homologous 1C capacity during skin carcinogenesis, but heterologous1C with normal keratinocytes was maintained in all cell lines.Theoretically, to explain the suppression of clonal expansionof initiated cells in situ (23, 24), 3PC, the carcinogen-alteredcell line, would be expected to communicate with normal cells.For the tumor-derived cell lines, selective communication maybe unnecessary because reduced homologous 1C may weakennormal cell influences sufficiently to allow expression of thetransformed phenotype. Recent studies with C3H10T'/2 (22) or

BALB/c 3T3 fibroblasts (19, 46) demonstrated, however, thatthese transformed cells avoid growth suppression by selectivecommunication that excludes their nontransformed neighbors.These results, however, may not accurately reflect the patternof 1C occurring during carcinogenesis when analyzed usingprimary cultures of normal and tumor tissues, because of thevery abnormal nature of these heteroploid cell lines.

Alternatively, this disparity may reflect a fundamental difference between carcinogenesis in epithelial and mesenchymal cellsystems. Transformed foci of BALB/c 3T3 cells display no

Fig. 6. Photomicrographs of cocultured normal and transformed mouse keratinocytes showing heterologous communication in low-calciummedium, a, c, phase-contrast (a) and fluorescence photographs (c) of heterologous communication observed upon microinjection of a bead-labeled PI/17 cell (*) after 24-h coculture withprimary keratinocytes. ft, d, phase contrast (ft)and fluorescence photographs (¡I)of communication observed upon microinjection of a primary keratinocyte (if) 25 h after addition ofbead-labeled P2/15 cells (*) to primary cultures.Note dye spread to surrounding normal cells andP2/15 cells. Bar, 40 /jm.

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

Table 1 Heterologous communication in cocultures of bead-loaded cell linesadded to primary keratinocyte cultures

Tissue-cultured cells were latex bead prelabeled, dissociated, and added topriman keratinocyte cultures. After varying periods of coculture in low-calciummedium, primary cells were microinjected with Lucifer Yellow CH to determinethe heterologous communication capacity (see "Materials and Methods").

Cell line cocul-lured with nor

malcells3PCPI/17

P2/15P3/22CA3/7Coculture

period(h)8.5

2425

25.526Heterologous

communication"20/20

(8)*

24/25(15)28/31 (15)32/36(18)24/29(15)

°Number of cell line cells receiving dye from normal cells/number of cell line

cells in immediate vicinity of the injected normal cells.* Numbers in parentheses, total number of normal cells injected.

Table 2 Heterologous communication in cocultures of bead-loaded primarykeratinocytes added to cultures of established cell lines

Primary keratinocytes were latex bead prelabeled, dissociated, and added tocultures of cell lines. After 1 day of coculture in low-calcium medium, cell linecells were microinjected with Lucifer Yellow CH, and the heterologous communication capacity was determined (see "Materials and Methods").

Cell line incoculture

Cocultureperiod (h)

Heterologouscommunication"

3PCCA3/7

2425

18/19(19)*

8/9 (9)" Number of normal keratinocytes receiving dye from cell line cells/number of

normal cells in immediate vicinity of the injected cell line cells.b Numbers in parentheses, total number of cell line cells injected.

diminution of intrinsic 1C but do not communicate with surrounding "normal" cells (16, 17). Other reports also suggested

that reduced homologous 1C is associated more with transformation of epithelial cells (9, 10). Perhaps, then, selective communication is related more closely to fibroblast transformationthan to transformation in epithelial cells. Caution is warranted,however, in considering such a conclusion. Although we do notyet know how best to simulate in vivo heterologous cell contacts,we do know that tumors emerge within normal tissue in situ.Considering this, carcinogen-induced focus formation in fibroblast cultures may represent a more germane model. Unfortunately, no in vitro system yet exists for the in situ productionof transformed epithelial cells with which the question of selective communication could be answered more confidently. Thus,we used synthetic cocultures in the studies described but arecurrently exploring methods of imitating the in vivo process ofepidermal carcinogenesis better.

Interpretation of the in vitro coculture data is further complicated by the use of Iow-Ca2+ medium. Used to maintain the

primary keratinocytes in a proliferative mode, without terminaldifferentiation, this medium further removes these cultured cellpopulations from the balance between proliferation and differentiation that exists in vivo. Pasti et al. (50) have recentlyreported that primary cultures of newborn mouse keratinocytesshow decreased 1C when shifted from low to high Ca2+. Since

alterations of differentiation responsiveness are a hallmark ofcarcinogen-altered and tumor-derived cell populations in theskin (Ref. 29; Footnote 4), the use of Iow-Ca2+ medium may

further reduce the relevance of this coculture model to eventsoccurring in vivo during carcinogenesis. We are currently examining whether heterologous communication may be sensitiveto Ca2+ concentration.

Finally, the communication assay we used may not allow thedetection of selectivity in this model system simply because oftechnical considerations. Since we are examining only the cell-cell passage of Lucifer Yellow CH, a M, 467 molecule, it ispossible that selective communication is present but only forsome gap-junctional permeant or permeants of a larger size.

This could be so if some stricture of the gap-junction channelshad occurred, perhaps as a result of defective alignment ofheterologous junction moieties. Probes of different sizes (up toM, 1000) could be used to test this (51).

The system we used in these studies offers an excellentopportunity to identify changes associated with tumor development, since neoplastic and preneoplastic cells from definedstages of initiation and promotion can be manipulated foranalysis. Recently, much work has been done to characterizegenetic changes related to mouse skin carcinogenesis, particularly the activation of the H-ras oncogenes by DMBA (52-55).We are currently exploring how these alterations relate to gap-junctional communication changes. In addition, it will be important to determine the homologous and heterologous cell-cell communication characteristics of emerging tumors in vivo,and such studies are in progress.

ACKNOWLEDGMENTS

We thank Aurora Viaje for provision of the 3PC cell line; M. Imbert,Service Central de Microanalyse, Solaise, France, for atomic absorptiondeterminations of medium calcium concentrations; and C. Fuchez forpreparing the manuscript.

NOTE ADDED IN PROOF

Further evidence for reduced 1C in the late stages of skin carcinogenesis comes from recent studies in our laboratory using three additionalcarcinoma-derived cell lines, CA 4/5, CA 8/29 and CA 9/20. Thesecells demonstrate homologous 1C of 3.7 ±0.3, 13.2 ±0.9 and 2.8 ±0.4 recipients per injection, respectively.

REFERENCES

1. Pitts, J. D. The gap junction. J. Cell Sci. (suppl.), 4: 239-266, 1986.2. Loewenstein, W. R. Junctional intercellular communication and the control

of growth. Biodi im. Biophys. Acta, 605: 33-91, 1979.3. Lee, S., diluía.N. B., and Warner, A. E. Gap Junctional communication and

compaction during preimplantation stages of mouse development. Cell, 57:851-860, 1987.

4. Loewenstein, W. R., and Ranno. Y. Intercellular communication and thecontrol of tissue growth: lack of communication between cancer cells. Nature(Lond.), 209: 1248-1249, 1966.

5. Jamakosmanovic, A., and Loewenstein, W. R. Cellular uncoupling in cancerous thyroid epithelium. Nature (Lond.), 218: 775, 1968.

6. Ramio. Y., and Matsui, Y. Cellular uncoupling in cancerous stomach epithelium. Nature (Lond.), 218: 775-776, 1968.

7. Weinstein, R. S., Merk, F. B., and Alroy, J. The structure and function ofintercellular junctions in cancer. Adv. Cancer Res., 23: 23-89, 1976.

8. Tachikawa, T., Yamamura, T., and Yoshiki, S. Changes occurring in plasmamembranes and intercellular junctions during the process of carcinogenesisand in squamous cell carcinoma. Virchows Arch. B, 47: 1-15, 1984.

9. Borek, C., Higashino, S., and Loewenstein, W. R. Intercellular communication and tissue growth. IV. Conductance of membrane junctions of normaland cancerous cells in culture. J. Membr. Biol., 1: 274-293, 1969.

10. Mesnil, M., Montesano, R., and Yamasaki, H. Intercellular communicationof transformed and non-transformed rat liver epithelial cells. Modulation byTPA. Exp. Cell Res., 165: 391-402, 1986.

11. Kavanagh, T. J., Chang, C-C., and Trosko, J. E. Characterization of a humanteratocarcinoma cell assay for inhibitors of metabolic cooperation. CancerRes., 46: 1359-1366, 1986.

12. Miller, F. R., Mclnerney, D., and Miller, B. E. Metabolic cooperation invitro: differential ability of ouabain to uncouple normal, preneoplastic, andneoplastic mouse mammary cells. Biochem. Biophys. Res. Commun., 140:955-961, 1986.

13. Fentiman, I. S., Hurst, J., Ceriani, R. L., and Taylor-Papadimitriou, J.Junctional intercellular communication pattern of cultured human breastcancer cells. Cancer Res., 39:4739-4743, 1979.

14. Pitts, J. D., and Kam, E. Communication compartments in mixed cellcultures. Exp. Cell Res., 756:439-449, 1985.

15. Kam, E. K., Melville, L., and Pitts, J. D. Patterns of Junctional communication in skin. J. Invest Dermatol., 87: 748-753, 1986.

16. Enomoto, T., and Yamasaki. H. Lack of intercellular communication betweenchemically transformed and surrounding nontransformed BALB/c 3T3 cells.Cancer Res., 44: 5200-5203, 1984.

17. Yamasaki, H., Hollstein, M., Mesnil, M., Martel, N., and Aguelon, A-M.

704

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CELL-CELL COMMUNICATION IN MOUSE EPIDERMAL CELLS

Selective lack of intercellular communication between transformed and non-transformed cells as a common property of chemical and oncogene transformation of BALB/c 3T3 cells. Cancer Res., 47: 5658-5664, 1987.

18. Yamasaki, H., and Fitzgerald, D. J. The role of selective junctional communication in cell transformation. In: R. Langenbach, J. C. Barrett, and E.Elmore (eds.), Tumor Promoters: Biological Approaches for MechanisticStudies and Assay Systems pp. 131-147. New York: Raven Press, 1988.

19. Yamasaki, H., and Katoh, I . Further evidence for the involvement of gap-junctional intercellular communication in induction and maintenance oftransformed foci in BALB/c 3T3 cells. Cancer Res., 48: 3490-3495, 1988.

20. Mesnil, M., and Yamasaki, H. Selective gap-junctional communication capacity of transformed and non-transformed rat liver epithelial cell lines.Carcinogenesis (Lond.), 9: 1499-1502, 1988.

21. Mehta, P. P., Bertram, J. S., and Loewenstein, W. R. Growth inhibition oftransformed cells correlates with their junctional communication with normalcells. Cell, 44: 187-196, 1986.

22. Murray, A. W., and Fitzgerald, D. J. Tumor promoters inhibit metaboliccooperation in cocultures of epidermal and 3T3 cells. Biochem. Biophys.Res. Commun., 91: 395-401, 1979.

23. Yotti, L. P., Chang, C. C., and Trosko, J. E. Elimination of metaboliccooperation in Chinese hamster cells by a tumor promoter. Science (Wash.DC), 206: 1089-1091, 1979.

24. Fitzgerald, D. J., Knowles, S. E., Ballard, F. J., and Murray, A. W. Rapidand reversible inhibition of junctional communication by tumor promotersin a mouse cell line. Cancer Res., 43: 3614-3618, 1983.

25. Yamasaki, H. The role of cell-to-cell communication in tumor promotion.In: T. J. Slaga and B. Butterworth (eds.), Nongenotoxic Mechanisms inCarcinogenesis, pp. 297-309. Cold Spring Harbor, NY: Cold Spring HarborPress, 1987.

26. Dermietzel, R., Yancey, S. B., Traub, O., Willecke, K., and Revel, J.-P.Major loss of the 28-kD protein of gap junction in proliferating hcpatocy tes.J. Cell Biol., 105:1925-1934, 1987.

27. Beer, D. G., Neveu, M. J., Paul, D. L., Rapp, U. R., and Pilot, H. C.Expression of the c-raf protooncogene, •¿�y-glutamyltranspeptidase,and gapjunction protein in rat liver neoplasms. Cancer Res., 48:1610-1617, 1988.

28. Berenblum, I. Sequential aspects of chemical Carcinogenesis: skin. In: F. E.Becker (ed.), Cancer: a Comprehensive Treatise, Vol. 1, Ed. 2, pp. 451-484.New York: Plenum Press, 1982.

29. Yuspa, S. H., and Morgan, D. L. Mouse skin cells resistant to terminaldifferentiation associated with the initiation of Carcinogenesis. Nature(Lond.), 295:72-74, 1981.

30. Brennan, J. K., Mansky, J., Roberts, G., and Lichtman, M. A. Improvedmethods for reducing calcium and magnesium concentrations in tissue culturemedium: application to studies of lymphe»blastproliferation in vitro. In Vitro,//: 354-360, 1975.

31. Kawamura, H. J., Strickland, E., and Yuspa, S. H. Association of resistanceto terminal differentiation with initiation of Carcinogenesis in adult mouseepidermal cells. Cancer Res., 45: 2748-2752, 1985.

32. Miller, D. R., Viaje, A., Aldaz, C. M., Conti, C. J., and Slaga, T. J. Terminaldifferentiation-resistant epidermal cells in mice undergoing two-stage Carcinogenesis. Cancer Res., 47: 1935-1940, 1987.

33. Kulesz-Martin, M., Kilkenny, A. E., Holbrook, K. A., Digernes, V., andYuspa, S. H. Properties of carcinogen altered mouse epidermal cells resistantto calcium-induced terminal differentiation. Carcinogenesis (Lond.), 4:1367-1377, 1983.

34. Stewart, W. W. Functional connections between cells as revealed by dye-coupling with a highly fluorescent naphthalimide tracer. Cell, 14: 741-759,1978.

35. Enomoto, T., Martel, N., Ranno, Y., and Yamasaki, H. Inhibition of cellcommunication between BALB/c 3T3 cells by tumor promoters and protection by cAMP. J. Cell Physiol., 121: 323-333, 1984.

36. Hennings, H., Michael, D., Chang, C., Steinert, P., Holbrook, K., and Yuspa,S. H. Calcium regulation of growth and differentiation of mouse epidermalcells in culture. Cell, 19: 245-254, 1980.

37. Biegon, R. P., Atkinson, M. M., Liu, T-F., Kam, E. Y., and Sheridan, J. D.Permeance of Novikoff hepatoma gap junctions: quantitative video analysisof dye transfer. J. Membr. Biol., 96: 225-233, 1987.

38. Evans, M. G., and Trosko, J. E. Concentration/response effect of2,2',4,4',5,5'-hexabromobiphenyl on cell-cell communication in vitro: assessment by fluorescence redistribution after photobleaching ("FRAP"). CellBiol. Toxicol., 4: 163-171, 1988.

39. Evans, M. G., El-Fouly, M. H., Trosko, J. E., and Sleight, S. D. Anchoredcell analysis/sorting coupled with the scrape-loading/dye transfer techniqueto quantify inhibition of gap-junctional intercellular communication in WB-F344 cells by 2,2',4,4',5,5'-hexabromobiphenyl. J. Toxicol. Environ. Health,24:261-271, 1988.

40. Trosko, J. E., Chang, C. C., and Medcalf, A. Mechanisms of tumor promotion: potential role of intercellular communication. Cancer Invest., 1: 511-526, 1983.

41. Sivak, A., and Van Duuren, B. L. Phenotypic expression of transformation:induction in cell culture by a phorbol ester. Science (Wash. DC), 157: 1443-1444, 1967.

42. Bertram, J. S., and Faletto, M. B. Requirements for and kinetics of growtharrest of neoplastic cells by confluent 101',: fibroblasts induced by a specificinhibitor of cyclic adenosine 3':5'-phosphodiesterase. Cancer Res., 45:1946-

1952, 1985.43. Borek, C., and Sachs, L. The difference in contact inhibition of cell replication

between normal cells and cells transformed by different carcinogens. Proc.Nati. Acad. Sci. USA, 56:1705-1711, 1966.

44. Stoker, M. Regulation of growth and orientation in hamster cells transformedby polyoma virus. Virology, 24: 165-174, 1974.

45. Weiss, R. A. The influence of normal cells on the proliferation of tumor cellsin culture. Exp. Cell Res., 63: 1-18, 1970.

46. Bignami, M., Rosa, S., Falcone, G., Tato, F., Katoh, F., and Yamasaki, H.Specific viral oncogenes cause differential effects on cell-to-cell communication, relevant to the suppression of the transformed phenotype by normalcells. Molecular Carcinogenesis, /: 67-75, 1988.

47. Hennings, H., Jung, R., Michael, D., and Yuspa, S. H. An in vitro model forinitiated mouse epidermis. Proc. Am. Assoc. Cancer Res., 27: 139, 1986.

48. Strickland, J. E., Greenhalgh, D. A., Koceva-Chyla, A., and Yuspa, S. H.Development of murine epidermal cell lines which contain an activated ras"'

oncogene and form benign papillomas in skin grafts on athymic nude mousehosts. Proc. Am. Assoc. Cancer Res., 28: 167, 1987.

49. Nicolson, G. L., Dulski, K. M., and Trosko, J. E. Loss of intercellularjunctional communication correlates with metastatic potential in mammaryadenocarcinoma cells. Proc. Nati. Acad. Sci. USA, 85:473-476, 1988.

50. Pasti, G., Riveda!, E., Yuspa, S. H., Herald, C. L., Pettit, G. R., andBlumberg, P. M. Contrasting duration of inhibition of cell-cell communication in primary mouse epidermal cells by phorbol 12,13-dibutyrate and bybryostatin 1. Cancer Res., 48:447-451, 1988.

51. Loewenstein, W. R. Junctional intercellular communication: the cell-to-cellmembrane channel. Physiol. Rev., 61: 829-913, 1981.

52. Balmain, A., Ramsden, M., Bowden, G. T., and Smith, J. Activation of themouse cellular Harvey-ros oncogene in chemically induced benign skin papillomas. Nature (Lond.), 307:658-660, 1984.

53. Harper, J. R., Reynolds, S. H., Greenhalgh, D. A., Strickland, J. E., Lacal,J. C., and Yuspa, S. H. Analysis of the ras" oncogene and its p21 product inchemically induced skin tumors and tumor-derived cell lines. Carcinogenesis(Lond.),«: 1821-1825, 1987.

54. Pelling, J. C., Ernst, S. M., Strawhecker, J. M., Johnson, J. A., Nairn, R. S.,and Slaga, T. J. Elevated expression of H-ras is an early event in two-stageskin Carcinogenesis in SENCAR mice. Carcinogenesis, 7: 1599-1602, 1986.

55. Yamasaki, H., Hollstein, M., Martel, N., Cabrai, J. R. P., Calendo, D., andTomatis, T. Transplacental induction of a specific mutation in fetal Ha-rasand its critical role in post-natal Carcinogenesis. Int. J. Cancer, 40:818-822,1987.

56. Klann, R. C., Slaga, T. J., Fitzgerald, J., and Yamasaki, H. Cell-cell junctionalcommunication in cell lines from SENCAR mouse skin. In Vitro, 23: 17a,1987.

705

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1989;49:699-705. Cancer Res   Richard C. Klann, D. James Fitzgerald, Colette Piccoli, et al.   CarcinogenesisLines from Selected Stages of SENCAR Mouse Skin Gap-Junctional Intercellular Communication in Epidermal Cell

  Updated version

  http://cancerres.aacrjournals.org/content/49/3/699

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/49/3/699To request permission to re-use all or part of this article, use this link

on June 23, 2018. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from