Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
[CANCER RESEARCH 46, 5923-5932, November 1986] Role of Cytoskeleton Changes and Expression of the H-ray Oncogene during Promotion of Neoplastia Transformation in Mouse Epidermal JB6 Cells Kiyoshi Takahashi, Ursula I. Heine,1 James L. Junker, Nancy H. Colburn, and Jerry M. Rice Laboratories of Comparative Carcinogenesis fK. T., U. I. H..J.L. ]., 1. M. R.¡and Viral Carcinogenesis [N. H. C.J, National Cancer Institute-Frederick Cancer Research Facility, Frederick, Maryland 21701-1013 ABSTRACT The relationship between cytoskeletal changes and oncogene expres sion in initated cells during exposure to a tumor promoter was investigated in the phorbol ester-sensitive murine epidermis-derived cell line JB6 (P* cells) and its promotion-insensitive variant (P~ cells) using immunocytochemical methods, soft agar assays, and tumorigenicity tests in nude mice. Cytoskeletal changes in P* and P~ cells induced by short-term incubation with 12-0-tetradecanoylphorbol-13-acetate (TPA) were sim ilar. Prolonged incubation with TPA allowed P cells to regain their original appearance and resulted in growth inhibition; however, the extended presence of TPA produced in P* cells persistent alterations in the distribution of actin, vinculin, and fibronectin. P* cells proceeded to develop multilayered foci. Using monoclonal antibodies, we detected the lì-raxoncogene-encoded M, 21,000 protein (p21) exclusively in focusforming cells. Both the observed morphological changes and the expres sion of p21 were reversible in !" cells when TPA exposure was terminated soon after foci had developed. In order for TPA-treated P* cells to grow as tumors in nude mice, multiple cycles of exposure to TPA in conjunction with clonal expansion in agar were necessary. The results indicate that there exists during promotion of the I" JB6 cells a relationship between expression of the 11-rav gene product p21 and enhanced proliferation with focus formation and that both expression of p21 and focus formation depend on the continuous presence of the promoting agent. INTRODUCTION According to the initiation-promotion theory of Carcinogen esis, initiation is an irreversible, genetic alteration that can be effected by a single exposure to a low dose of carcinogen that by itself causes no (or few) tumors. In contrast, for tumors to develop from initiated cells, repeated treatment with tumor promoters is necessary, and it has been suggested that, during this time span, initiated cells are induced to give rise to new phenotypes that are first expanded as clones which ultimately produce tumors. Because tumor promoters are not mutagenic and many of their effects during promotion are reversible (reviewed in Refs. 1 and 2), altered gene regulation mechanisms have been suggested for promotion. In order to overcome normal growth restrictions during the promotion process, ini tiated cells must undergo certain changes that result in the loss of cell contacts and the acquisition of elevated growth potential. The reduction of cell-cell and cell-substrate contacts is usually intimately associated with both the rearrangement of cytoskel etal structures, especially actin and vinculin (3, 4), and the loss of extracellular matrix proteins, such as fibronectin (5). Such morphological changes, coupled with stimulation of growth and cell division, may result from expression of the protein products of certain oncogenes, as has been suggested for the M, 60,000 phosphoprotein src oncogene product (Ref. 6; reviewed in Ref. 7) and for the Harvey ras p21 protein (8, 9). In support of this, it has been reported that primary papillomas as well as Received 1/20/86; revised 7/24/86; accepted 8/4/86. 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 with 18 U.S.C. Section 1734 solely to indicate this fact. 1To whom requests for reprints should be addressed, at the Ultrastructural Studies Section, Laboratory of Comparative Carcinogenesis, NCI-Frederick Can cer Research Facility. Bldg. 538. Frederick, MD 21701-1013. squamous cell carcinomas induced in mice by an initiationpromotion protocol contain the activated H-ras oncogene (10, 11). It has recently been documented that tumor promoters en hance the transformation of cell cultures transfected with an activated c-H-ras ocogene (12, 13). Thus, as has been suggested by these authors and others (14, 15), it is especially attractive to consider that, during multistage Carcinogenesis, tumor pro moters might interact synergistically with activated or cellular oncogenes, either by altering the quantitative and temporal expression of such genes or by inducing cooperative genes (16, 17) to increase cellular proliferation and bring about neoplastic transformation. The most active known tumor promoter, isolated from croton oil, is TPA.2 TPA exerts a wide range of biological and bio chemical effects, including loss of actin cables (18), decrease in fibronectin (19, 20) and procollagen (21), and interruption of intercellular communication due to the loss of gap junctions (22), thus modulating the original phenotype to assume in many respects a transformed appearance. However, because these effects are not specific for initiated cells, are often transient in the presence of TPA, and are often reversible upon removal of TPA, it is not clear which of the TPA-mediated effects are essential to promote the latent, initiated cells. To examine possible interactions between a tumor promoter, phenotypic modulation, and oncogene expression during pro motion, we used cloned cells from the JB6 mouse epidermisderived cell line (23, 24). Clones of this cell line are especially useful for such work because they include the promotioncompetent (P+) cells (clone 41) that can be irreversibly trans formed by treatment with TPA alone and therefore are thought to represent initiated or postinitiated cells. A promotion-incom petent (P~) clone (clone 30) served as control. Cells from this clone do not transform in response to TPA and thus are thought to be earlier preneoplastic cells. Moreover, as previously re ported (25), transfection of JB6 P~ cells with activated viral Hras DNA transformed these cells to anchorage independence. Thus, these JB6 cells presumably do not express activated p21. A second control consisted of permanently transformed cells, clone RT 101, a TPA-transformed derivative of JB6 P+ clone 41. It is noteworthy that while TPA has many effects on many kinds of cells, JB6 cells retain specificity for promoters of squamous epithelium and fail to respond to promoters of, e.g., hepatic parenchyma! neoplasia (26). In the present study, we investigated structural alterations such as rearrangement of cytoskeletal components, including filamentous actin, microtubules, and vinculin, and changes in cell substrate behavior, as shown by an altered pattern of fibronectin distribution, due to short-term exposure to TPA in P+ and P~ cells. We compared the results to those obtained in long-term studies in which P+ cells were repeatedly exposed to TPA in semisolid medium (clonal expansion) and the resulting 2 The abbreviations used are: TPA, 12-O-tetradecanoylphorbol-13-acetate; FITC, fluorescein isothiocyanate; PKC, protein kinase C; PBS, phosphate-buff ered saline; p 21, M, 21,000 protein. 5923 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES anchorage-independent clones were used for further propaga tion. Cell cultures developed by this regimen were tested for their capacity to show anchorage independence in TPA-free medium and to produce tumors in nude mice. The expression of the H-ras oncogene product p21 was monitored in both short-term and long-term studies. Our findings show that the morphological alterations result ing from short-term exposure to TPA persist in P+ cells cultured in the continued presence of TPA; however, under the same experimental conditions, these effects are only transient in P~ cells. We show also that continuous exposure of P* cells to TPA results, after confluence, in the formation of numerous multilayered foci, whereas P~ cells show density-dependent inhibition of growth even in the continued presence of TPA. We detected the H-ras gene product p21 only in focus-forming P+ cells cultured in the continuous presence of TPA; the onco gene product was not detected in flat cells of the TPA-treated monolayers or in permanently transformed P+ cells produced by repeated exposure to TPA. These findings relate the expres sion of an oncogene-encoded protein specifically to focus for mation that results from exposure to a tumor promoter. The results of our long-term studies show that the irreversible expression of the transformed phenotype proceeds stepwise and involves permanent reduction in actin-containing stress fibers and cellular fibronectin. Progression to the transformed cell type depends on repeated exposure to TPA in conjunction with clonal expansion in agar and leads to increased anchorage independence and, finally, tumorigenicity. In each subclone, after confluence, the formation of foci in the presence of TPA was associated with the transient expression of H-ras p21, thus indicating a requirement for the transient expression of the oncogene product during promotion, but not in maintenance of the resulting tumor cell phenotype. MATERIALS AND METHODS Chemicals. TPA (Chemicals for Cancer Research, Inc., Eden Prairie, MN) was initially dissolved in dimethyl sulfoxide at a concentration of 1 mg/ml. The stock solution was maintained at -20°C and diluted in culture medium before use. Cell Culture. The origin and maintenance of BJ6 cells and their promotion-sensitive clones have been described (23, 24). Promotable (P+) cells (clone 41, passage 80), nonpromotable (P~) cells (clone 30, passage 100), and TPA-transformed cells (clone RT 101) were used. Cells were seeded at a concentration of 5 x 10s cells/60-mm Petri dish in Eagle's minimum essential medium containing 5% heat-inactivated fetal calf serum supplemented with penicillin and streptomycin. TPA was used at a concentration of 0.1 Mg/ml. Cells were examined by phase-contrast and immunofluorescence microscopy. In short-term ex periments, examination was carried out between 2 min and 10 days after addition of TPA, using appropriate controls. Evaluation of Anchorage-independent Growth by Soft Agar Assay. Cells were suspended in culture medium supplemented with 10% fetal calf serum containing 0.33% Difco agar at 40°Cto which solvent alone (0.01% dimethyl sulfoxide) or TPA (0.1 Mg/ml) had been added. The suspension, 1.5 ml containing IO4 cells/dish, was layered over 0.5% agar base in standard medium. Assays were carried out in duplicate without refeeding; colonies consisting of 10 or more cells were counted after 14 days. Repeated TPA Treatment and Establishment of Subclones. I' cells in 0.33% Difco agar medium containing TPA (0.1 Mg/ml) were layered at a concentration of 5 x IO2 cells/dish over 0.5% agar base. After culturing for 3 weeks with refeeding, resulting colonies were isolated and grown as monolayers in the absence of TPA. Among 20 clonal lines derived from soft agar colonies of I' ' cells, the cell line that showed highest anchorage-independent growth in TPA-free agar medium, des AND H-ras P2I IN JB6 CELLS ignated C, (implying one cycle of TPA exposure), was used for a second treatment with TPA according to the schedule described above. Similar selection for anchorage-independent growth yielded clonal line C2. By 3, 4, and 5 successive intervals of exposure to TPA and cloning, we established clonal cell lines C3 to ( .. Assay for Tumorigenicity. Cells of clones <', through C3, as well as the original cells of clone 41 (C0), were suspended in serum-free medium at a concentration of 1 or 5 x IO6 cells/0.2 ml and injected s.c. into nude mice. RT 101 cells, a clone of JB6 cells consisting of permanently transformed cells, were injected into nude mice to serve as a positive control. The resulting tumors were examined by routine histopathological methods. Fluorescence Cytochemistry. Cells used for fluorescence microscopy were grown on glass coverslips. Staining was performed by inverting the cover slips onto 30-50 til of the appropriate solution on Parafilm (American Can Co., Greenwich, CT). To detect filamentous actin, cells were fixed in 4% formaldehyde (TEM grade; Tousimis, Rockville, MD) in PBS, extracted in acetone at -15"C for 5 min, and incubated with 0.17 MMfluorescein-phalloidin (Molecular probes, Inc., Junction City, OR) at room temperature for 30 min. For indirect immunofluorescence experiments, antibodies were di luted with PBS containing 0.1-1 % bovine serum albumin; incubations were for 30 min at room temperature, and washes following fixation and antibody incubations were for 10 min each. Stained cells were mounted in Elvanol (Monsanto, Springfield, MA) (27) and examined on a Leitz Ortholux microscope fitted for epifluorescence. Specifics of the various staining protocols were as follows. For fibronectin, cells were fixed and extracted according to the actin staining protocol, incubated with rabbit anti-human fibronectin (Seragen, Boston, MA) diluted 1:50, followed by FITC-labeled goat anti-rabbit IgG (Cappel Laboratories, Cochranville, PA) diluted 1:50. For vinculin, cells were fixed with 2% formaldehyde in PBS for 5 min, followed by permeabilization with 0.2% Triton X-100 in PBS for 5 min, incubated with rabbit anti-vinculin (Transformation Research, Inc., Framingham, MA) di luted 1:30, followed by FITC-labeled goat anti-rabbit IgG diluted 1:50. For tubulin, cells were treated with stabilization buffer consisting of 0.1 M piperazine-A',A'/-bis(2-ethanesulfonic acid) sodium salt adjusted to pH 6.9 with potassium hydroxide, 1 mM ethyleneglycol bis(2-aminoethyl ether)-jV,yV'-tetraacetic acid, and 4% polyethylene glycol 6000; and permeabilized using 0.5% Triton X-100 (28); fixed in 4% formal dehyde; and incubated in undiluted sheep anti-bovine tubulin (Caabco, Houston, TX), followed by FITC-labeled rabbit anti-sheep IgG (Cappel) diluted 1:50. For keratin and vimentin, cells were fixed for 5 min in methanol at — 15°C,incubated in either DAKO rabbit anti-keratin (Accurate Chemical and Scientific, Westbury, NY), guinea pig antiprekeratin (Miles Scientific, Naperville, IL), mouse monoclonal ami cytokeratin (Labsystems, Chicago, IL), or mouse monoclonal ami vimentin (Labsystems) diluted 1:20, followed by incubation with an appropriate, labeled second antibody. In order to detect II rav p21, three rat monoclonal antibodies were used, YA6-172, Y13-238, and Y13-259 (gifts from Dr. T. Shih, Labo ratory of Molecular Oncology, National Cancer Institute, Frederick, MD). These antibodies recognize different antigenic determinants of p21; YA6-172 and Y13-238 are specific for H-ras p21, and Y13-259 reacts with both H-ras p21 and Kristen ras p21 (29). Cells were fixed for 10 min in methanol at — 15°Cand incubated in antibody (10 Mg/ ml), followed by FITC-labeled anti-rat IgG (Cappel) diluted 1:40. For controls the specific antibody was replaced with nonimmune serum at the same dilution as the primary antiserum (for polyclonal antibodies) or with IgG diluted to the same concentration as primary antibody (for monoclonal antibodies). As an additional control in the p21 studies, antibody Y13-238, which had been adsorbed for 24 h with excess purified H-ras p21 synthesized in Escherichia coli (a gift from Dr. S. Hattori, Laboratory of Molecular Oncology, National Cancer Institute, Frederick, MD), was used in the primary antibody step. The staining protocol for p21 was confirmed using Harvey sarcoma virustransformed NIH 3T3 cells, strain 13A (a gift from Dr. Shih), as a positive control and using uninfected NIH 3T3 cells as a negative control. 5924 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS RESULTS Short-Term Exposure to TPA In these experiments, cells of logarithmically growing cul tures were exposed to TPA (0.10 Mg/ml) for different times ranging from 2 min to several days. Morphological Alterations in the Presence of TPA In subconfluent cultures, the morphological changes resulting from the addition of TPA were similar in P+ and P~ cells. After 30 min, cells were reduced in size and formed long cytoplasmic processes. Scanning electron microscopy showed ruffles at the cell periphery (data not shown). However, after confluence (about 5 to 10 days after addition of TPA), marked differences between P+ and P~ cells were evident. P~ cells regained their original appearance and showed density-inhibited growth (Fig. \A); however, their saturation density was slightly increased («30%)as compared to untreated P~ cells (data not shown). In contrast, P+ cells were highly refractile (Fig. IB), and numerous foci of mult ¡layeredround cells developed in these cultures (Fig. 1C). Actin Distribution. Untreated P+ and P~ cells contained a prominent system of actin cables in their cytoplasm (Figs. "LA and 3/<), that began to disappear upon exposure to TPA for only 5 min. At the same time, peripheral actin staining (in ruffles) became prominent (Fig. 2B). The process was com pleted about 15 min after addition of TPA (Fig. 2C). Although the initial reaction of P+ and P~ cells to TPA was similar (Figs. 2D and 35), maintenance of the cells in TPA-containing me dium until (and after) confluence resulted in reversal of the morphological alterations only in P~ cells (Fig. 3C). About 5 t days after addition of TPA, stress fibers began to reappear in considerable number in these cells culminating in complete reconstruction of the stress fiber system. In contrast, P+ cells .£»»* did not reorganize their stress fiber system (Fig. 2E). Vinculin Distribution. In untreated cultures of P* and P~ cells, the vinculin distribution was comparable and followed a stand ard pattern; i.e., vinculin was concentrated in small adhesion plaques at the cell periphery where it demarcated regions of cell-substrate contacts and indicated the termination points of actin stress fibers (Figs. 2F and 3D). After addition of TPA for 10 min and shortly after actin cable loss was pronounced, vinculin was lost from adhesion plaques (Fig. 2, G and //); however, ruffles at the cell surfaces stained positively (Figs. 21 and 3E). As shown for actin staining, by 10 days vinculin staining was again detectable at stress fiber termination points in P~ cells (Fig. 3F) whereas in P+ cells only an occasional spot of vinculin staining could be detected between adjacent cells (Fig. 2K). Fibronectin Distribution. As illustrated in Figs. 2L and 3G, untreated P+ and P~ cells exhibited a thick matrix of fibronectin. In the presence of TPA, the density of this matrix decreased slowly in cultures of P+ and P~ cells until it was markedly reduced by about 24 h after addition of TPA (Figs 2, M-O and 3//). By 10 days after addition of TPA, there was a further reduction in fibronectin in cultures of P+ cells (Fig. 2P); in contrast, P~ cells reacquired a fibronectin matrix similar in density to that present before treatment (Fig. 37). Distribution of Intermediate Filaments and Tubulin. Neither polyclonal antibodies raised against bovine epidermal keratin, Fig. 1. Phase-contrast microscopy of nonpromotable (P~) and promotable (P*) JB6 cells grown in the presence of TPA. A, confluent culture of P~ (clone 30) cells, 5 days. Cells are flat and cell borders are obscure. B, confluent culture of P+ (clone 41) cells, 5 days. Cells are round with distinct borders. C, confluent culture of P* cells showing multilayered foci (arrows), 10 days. Approximately x 100. which stain suprabasal as well as some basal cell layers (DAKO, data sheet information), nor monoclonal antibodies raised against porcine kidney epithelial cells grown in vitro, which do not stain epidermal cells, were able to detect keratin in P+ or P~ cells. However, cells of both lines contained numerous vimentin filaments, which upon exposure to TPA did not alter their pattern of distribution. Microtubules were present throughout the cytoplasm, preferentially radiating from the cell center towards the periphery. They, too, did not change their distribution during exposure to TPA. 5925 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS Fig. 2. Changes in the distribution of actin (A-E), vinculin (F-K), and cellular fibronectin (L-M) with TPA treatment of P* cells. In the presence of TPA these changes persist. Actin cables disappear rapidly following exposure to TPA (A-E). x 290. A, untreated; B, TPA-treated for 5 min; C, TPA-treated for 10 min; D, TPAtreated for 24 h; E, TPA-treated for 10 days. Loss of vinculin plaques follows loss of actin cables (F-K). F, untreated; G, TPA-treated for 5 min; H, TPA-treated for 10 min; /, TPA-treated for 24 h; K, TPA-treated for 10 days. The density of the fibronectin matrix is gradually reduced (L-P). L, untreated; M, TPA-treated for 20 min; N, TPA-treated for 40 min; O, TPA-treated for 24 h; P, TPA-treated for 10 days. Distribution of the Harvey ras Gene-encoded Protein p21 In order to detect H-ras p21, we performed indirect immunofluorescence using three monoclonal antibodies. Our controls showed that (a) all three monoclonal antibodies gave rise to intense specific fluorescence of the plasma membranes of Harvey sarcoma virus-transfomed NIH 3T3 cells, line 13A; (b) untransformed NIH 3T3 cells did not stain with these antibod ies; and (c) neither normal rat IgG nor Y13-328 antibody preincubated with p21 gave rise to specific immunofluorescence in 13A cells. A specific staining for H-ras p21 was characteristic for all mitotic cells of untreated as well as subconfluent, TPA-treated cultures of both P+ and P~ cells (Fig. 4A). After the cultures became confluent, however, a positive staining was observed also at membranes of the round interphase cells that accumu lated in foci of P+ cell cultures in the presence of TPA (Fig. 4B). This is consistent with a previous report in which specific 5926 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS Fig. 3. Variations in the staining patterns of P cells for actin (A-C), vinculin (D-F), and fibronectin (G-l) with duration of TPA treatment. TPA-induced changes are initially similar to those seen in P+ cells, but these changes do not persist even in the continued presence of TPA. x 3000. A-C, actin staining. A, untreated; B, TPA-treated for 24 h; C, TPA-treated for 10 days. D-F, vinculin staining. D, untreated: E, TPA-treated for 24 h; F, TPA-treated for 10 days. G-l, fibronectin staining. G, untreated; H, TPA-treated for 24 h, /, TPA-treated for 10 days. Fig. 4. H-ras p21 staining of P* cells using monoclonal antibody Y13-238. A, untreated P* cell. Mitotic cell stains positively, x 300. B, multilayered focus of P* cells treated with TPA for 10 days shows p21 staining at cell membrane, x 150. C, control. Multilayered focus of P* cells treated with TPA for 10 days stained with Y13-238 preincubated with excess amount of v-H-ras p21 synthesized in E. coli, x 150. D, untreated P~ cell. Mitotic cell stains positively, x 150. E, P~ cell treated with TPA for 10 days. Only mitotic cells stain positively, x 150. 5927 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-roi P21 IN JB6 CELLS staining of the 11 »v/.v gene product adjacent to the membrane was detected in transformed cells (30). Nondividing flat cells in the monolayers adjacent to foci were not stained by the anti body. Likewise, interphase cells of confluent P~ cell cultures (Fig. 4, D and E) showed no staining for p21 in the presence of TPA. Reversal of the Morphological Changes after Removal of TPA from r Cells The effects of TPA on the morphology, cytoskeleton, and extracellular matrix of P+ cells grown in monolayers were reversible upon removal of the promoter from the culture medium. Cells in the monolayers and especially the round cells in the foci became flat and returned to their original appearance with abundant actin cables, vinculin-containing adhesion plaques and dense fibronectin mats. This reversal, which re quired at least 96 h for cells that had been exposed to TPA (0.1 ¿¿g/ml) for as long as 10 days, resulted in the complete disap pearance of foci from the cultures. We isolated foci from confluent, TPA-treated cultures of P+ cells and reseeded the cells in medium lacking TPA in order to confirm that round focus-forming cells could return to their original morphology. The reseeded cells established monolay ers, were density inhibited, and resembled the original P+ cells in both general morphology and cytoskeletal and matrix ar rangements. They were negative for H-ras p21. Cells isolated from five foci were also tested for anchorage-independent growth in 0.33% agar medium free of TPA. As shown in Table 1, with the exception of focus 3, these cells yielded a very low number of colonies. Our findings clearly indicate that the short-term exposure to TPA (up to 10 days) causes P+ cells to change phenotype and growth pattern remarkably, both aspects becoming similar to those of transformed cells; but these changes were reversible upon removal of TPA from the medium. However, the en hanced anchorage independence of focus 3 indicates that pro gression to transformation is under way in a subpopulation of cells. Long-Term Experiments Enhancement of Anchorage-independent Growth As described in "Materials and Methods," we established by repeated selection rounds clonal cell lines, Ci through C5, each derived from a colony that developed in TPA-containing soft agar. As shown in Fig. 5, during the passage of Ci to C5, the capacity for anchorage-independent growth in TPA-free semisolid medium increased in cultures of P+ cells. Fig. 5 also indicates that, in the presence of TPA, anchorage-independent Table 1 Colony formation in standard agar medium of clone 41 cells after isolation from foci obtained in the presence of TPA No. of colonies/ IO4 Focus no.° cells 50 1 2 70 3 180 4 50 5 SO Untreated clone 41 40 (controls) * Foci were obtained after exposure of confluent monolayers to TPA (100 ng/ ml) for 5 days. Single foci were plucked from monolayers, trypsinized, and seeded in TPA-free 0.33% agar medium to obtain colonies. C0 C, C2 Clonal C3 C4 C5 Cell Line Fig. 5. Number of anchorage-independent colonies of C0 to C5 cells/dish in the absence (O) and in the presence (•)of TPA. Anchorage-independent growth in the presence of TPA is to a large extent reversible. Anchorage-independent growth in the absence of TPA increases stepwise in C0 to C5 cells. Table 2 Presence of H-ras p21 staining in foci of TPA-treated monolayers Untreated Exposed to 0.1 jig/ml TPA 24 h lOdays Otime 10 days (P~)Clone Clone 30 Nofoci+'++++ (P*)C, 41 (P+)Cj (P*)C3 (P*)RT 101 (T)" Positive staining for H-ras p21, as shown in Fig. 4/f. was present in all foci of TPA-treated monolayers. growth of P+ cells in semisolid medium was remarkably in creased at higher passage levels. This increase was partially reversible upon return to standard medium lacking TPA. Morphological Analysis ofC,-C5 Cell Cultures When grown under standard culture conditions without TPA, Ci cells isolated after a single cycle of TPA-induced anchorage independence resembled untreated P+ cells (C0) in growth pat tern, configuration, and distribution of actin, \ inculili, and fibronectin. In contrast, C2-C5 cells, isolated as clones after 2 to 5 cycles, were small and highly refractile (Fig. 6). Fig. 6 indicates that the increase in both size and number of anchor age-independent colonies in TPA-free semisolid medium is accompanied by morphological transformation in monolayer cultures which is clearly seen in Ci-Cj cells. The number of actin cables in these cells were remarkably decreased (compare Fig. 1A with Fig. 7Aand Fig. 7Q. The amount of cell-associated fibronectin in cultures of C2 cells was only slightly decreased, but that of Cj-Ci cells was significantly reduced (Fig. 7, D to F). We found no significant changes in the amount and distri bution of vinculin in cultures Ci through C5. Cells in these cultures did not express the \\-ms p21 protein nor did they form foci. However, after confluence and in the presence of TPA, the cultures proceeded to form foci and simultaneously the focus-forming cells were again capable of expressing the II raspll (Table 2). Tumorigenicity ofC\-Ci Cells To determine the tumorigenicity of Ci-C3 cells, we injected 1 or 5 x IO6 cells of each clonal cell line s.c. into nude mice (Table 3). Untreated P+ cells and cells of passage Ci did not 5928 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS Fig. 6. Phase-contrast microscopy of cells of C» (CI 41) (A), C, («),C2 (C), C3 (D) (X 150) and soft agar colonies of C„ (E), C, (F), C2 (G), and C, (H) cells grown in TPA-free medium (x 40). Morphological transformation and anchorage-independent growth are increasingly evident in C2-C3 cells. yield tumors. After 3 months, 4 of 8 animals that had been given injections of C2 cells and 3 of 8 animals that had been injected with Cj cells developed tumors that histologically were undifferentiated carcinomas. All TPA-transformed JB6 clone 41 P+-derived RT 101 cells, the positive controls obtained by six cycles of TP A exposure in the absence of cell division (31), gave rise to tumors of similar morphology when injected into nude mice. DISCUSSION Phorbol esters can induce a variety of biological effects that result over a period of time in conversion of initiated mouse epidermal cells and their progeny to precancerous and cancer ous cells. Recently the idea was put forward that, during mul tistage promotion, tumor promoters might interact with cellular or activated oncogenes and by elevating their expression might 5929 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS SJf 'A Fig. 7. Irreversible changes in actin (A-C), and fibroneetin (D-F), following repeated exposure to TPA and clonal expansion. Actin cables are prominent in C| cells (A) but are markedly reduced in C2 (B) and C, (C) cells. The fibroneetin matrix is dense in cultures of C¡cells (D), slightly reduced in cultures of C2 cells (£), and markedly diminished in cultures of C3 (F) cells, x 380. Table 3 Tumorigenicity of subclones Ca-C3 after injection into athymic NCrNu/ Nu mice Tumor-bearing mice/ injected mice Concentration (cells/animal) (X ID6) Time after Injection (days) notypic modulation, and oncogene expression we have made use of the mouse epidermis-derived JB6 cell line. We have compared the changing patterns of cell morphology and of Hras oncogene expression of promotable (P+) cells of clone 41 with those of nonpromotable (P~) cells of clone 30 during Histology CoC,C2c,RT exposure to TPA. We demonstrated that the immediate effects of the partially trans P~ cells and are com SO,148, of TPA, namely, the reversible induction 18265, formed cell type, are similar in P+ and 9021UndifferentiatedcarcinomaUndifferentiatedcarcinomaUndifferentiatedcarcinoma 64, 1010/80/84/83/85/5S111546, yield a population of continually dividing cells (12-14). More over, it was also shown that tumor promoters may alter the expression of cooperating genes with consequent progression to tumor cell phenotype (32). To examine the interaction between tumor promoters, phe- parable to those reported previously (2). They consist mainly of a loss of actin cables, rearrangement of vinculin-containing adhesion plaques, and decrease in cell-associated fibroneetin. These changes are thought to be initiated at the plasma mem brane by the interaction of TPA with its receptor, PKC (33). Interestingly, Woolen and Wrenn (34) have shown recently that, during PKC activation by TPA, there is a translocation of PKC from the soluble cell fraction to the cell membrane. PKC is well known to have a pivotal role in regulating protein functions, such as phosphorylation of the adhesion protein 5930 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS vinculin (35). Alterations of vinculin and, possibly, other at tachment proteins such as a-actinin (36) may bring about disruption of actin cable anchorage at the cell membrane re sulting in disorganization of actin bundles (4). The cascade of events is further enhanced since, as shown here and in previous reports (37-39), the loss of actin cables is followed by a decrease in cell-associated fibronectin. Our findings support the sug gested transmembrane relationship between actin cables and fibronectin matrix (40-42). Our investigation shows that, in the presence of TPA, the morphological changes are only transient in P~ cells, thus indicating that these cells are capable of overriding the TPAinitiated signals. The transient nature of various morphological changes induced by TPA, such as a change in cell shape and cell adhesion (21, 43, 44), loss of actin cables (18), and a decrease in the number of gap junctions (22), has been reported for various types of cells. However, our results also clearly indicate that there exist fundamental differences between P+ and P~ cells. In contrast to P~ cells, P+ cells apparently lack the ability to acquire tolerance to TPA. P+ cells retain the altered phenotype and proceed to form foci during the extended presence of TPA. Although experiments to date have not dem onstrated any differences between P* and P~ cells in PKC activity or substrates (45), considering the central role of PKC (and the cascade of events that follows PKC activation) in effecting the TPA signals, the hypothesis remains attractive that P+ cells may lack capability to reduce the activation of the enzyme at the cell membrane to normal levels during prolonged exposure to TPA, whereas P~ cells may have such a mechanism available. This is in agreement with recent results showing a loss of intercellular communication between BALB/c 3T3 cells in the presence of TPA during growth phase; however, after confluence, these cells establish normal communications. In contrast, the transformation-sensitive variant of these cells was not capable of doing so (46). It was suggested that the under lying mechanisms are mediated by the action of PKC and its intermediates (47). If an inhibitory effect of TPA on intercel lular communication and gap junction formation could be detected in the JB6 cell system, it would be of great interest to determine whether the inhibitory effect is persistent in P+ cells and transient in P~ cells and whether it is regulated by the action of PKC. In evaluating our findings in relation to pub lished data, it is tempting to speculate that TPA acts on initiated or postinitiated cells by a sustained activation of PKC and that the cascade of events which follows selectively provides preneoplastic cells with proper intracellular, cell-cell, and cell-sub strate conditions to escape normal growth control and progress to a tumor cell phenotype. In this respect it is especially interesting that we found enhanced expression of the H-ras oncogene product p21 only in those interphase P+ cells that formed foci in the presence of TPA and after confluence, thus identifying the oncogene expres sion as a relatively late event. p21 has been detected previously in cells escaping from G0-Gi, the resting phase of growth, and in mitotic cells implying an important role of the H-ras gene in cell division and growth. It has been shown that microinjection of either transforming p21 or normal p21 into quiescent cells results in rapid alteration of cell morphology and stimulation of DNA synthesis and cell division, although the activity of normal p21 is much lower than that of transforming p21 (8, 9). At present, it seems unlikely that P+ cells contain the oncogenic form of the H-ras gene, because DNA extracted from P+ cells does not transform NIH 3T3 cells in transfection assays (25), and p21 is not expressed in permanently transformed P* cells (this study). Our findings imply that the transient stimulation of the (cellular) H-ras gene may be necessary to trigger, in the preneoplastic and morphologically altered cells, enhanced di vision and proliferation thus enabling these cells to overcome the growth-inhibiting effects of the surrounding normal cells (13). Because normal p21 has a low activity with respect to mitogenesis and morphological transformation, we consider in agreement with Weinstein et al. (48) that the combined effects of TPA on the cytoskeleton (actin cables), adhesion plaques, extracellular matrix (fibronectin), H-ras gene expression and, possibly, cell communication are necessary for preneoplastic P+ cells to yield foci or accomplish clonal expansion. Our results also find support in recently published data by Yuspa et al. (49) indicating that in keratinocytes an activated ras gene provides the proliferative signal for tumor formation, yet exposure to tumor promoters is necessary to selectively reverse the quiescent growth state of the cells that contain the activated ras gene. An alternative possibility not excluded by our data is that p21 expression is not causally related to but rather is indicative of TPA-induced focus formation. Our present study showed that the initial morphological changes, focus formation, and induction of p21 expression in monolayers are reversible upon removal of TPA from P+ cells. Previous results (24, 50) had shown acquisition of tumorigenicity after one cycle of TPA-induced anchorage-independent transformation, harvesting the largest clones for further culti vation. However, we found it necessary to subject these cells to two or more cycles in order to achieve neoplastic transformation as shown by tumor production. Interestingly, neoplastic behav ior as shown by morphological transformation, anchorageindependent growth, and tumorigenicity increases stepwise in these subclones. For example, the ability to assemble a network of actin cables and an extracellular fibronectin matrix and to establish adhesion plaques was progressively diminished, actin cables having been the first to disappear permanently. Such progressive and stepwise acquisition of the transformed phe notype of P+ cells during the repeated treatment with TPA is consistent with skin tumor promotion in vivo. The production of papillomas and carcinomas requires long-term and frequent application of the promoter. Recently, it has been proposed that tumor promoters produce clonal expansion of preneoplastic cells, while additional muta tion appears to be required for malignant conversion (51). Our results indicate that clonal expansion during TPA treatment of P+ cells may play a role in their escape from normal growth restriction. However, it is unlikely that P+ cells have to undergo additional mutation for the /'/; vitro equivalent of malignant conversion because (a) of the high frequency of TPA-induced neoplastic transformation (24) and (b) of the fact that P+ cells remained untransformed despite numerous passages. The latter implies that the incidence of such spontaneous mutations is low. Thus, our data suggest that TPA plays a direct role in preneoplastic progression. In this aspect, it needs to be empha sized that tumor promoters have been thought for many years to act as gene regulators (1, 45, 48, 52). Recently a number of genes have been shown to confer sensitivity to TPA induced transformation, among them T24, ras (13) myc, PyLT, and EIA (16). Of special interest are the pro genes that have been isolated from promotion-sensitive (P+) JB6 cells and have been found to be capable of conferring promotion sensitivity to otherwise promotion insensitive (P~) cells (17, 53, 54). The mechanism by which the expression of pro genes in cooperation with TPA might contribute to the expression of H-ras gene 5931 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. CYTOSKELETAL CHANGES AND H-ras P21 IN JB6 CELLS product p21, described in this report, and to neoplastic trans formation in general needs further exploration. ACKNOWLEDGMENTS We thank Drs. T. D. Gindhart (deceased), A. B. Roberts, and S. Sukumar for stimulating discussions. We thank Dr. T. Shih for the generous gift of rat monoclonal antibodies to detect H-ras p21 and Dr. S. Hattori for purified H-ras p21 synthesized in E. coli. We appreciate the excellent technical assistance of S. Kenney. REFERENCES 1. Boutwell, R. K The function and mechanism of promoters of carcinogenesis. CRC Crit. Rev. Toxicol., 2:419-443, 1974. 2. Slaga, T. .1.. and Andres, P. B. Cellular and biochemical changes during multistage skin tumor promotion. In: H. Fujiki et al. (eds), Cellular Interac tions by Environmental Tumor Promoters, pp. 291-301. Utrecht: VNU Science Press, 1984. 3. Geiger, B., Tokuyasu, K. T., Dutton, A. H., and Singer, S. J. Vinculin, an intercellular protein localized at specialized sites where microfilament bun dles terminate at cell membranes. Proc. Nati. Acad. Sci. USA, 77: 41274131, 1980. 4. Schliwa, M., Nakamura, T., Porter, K. R., and Euteneuer, U. A tumor promoter induces rapid and coordinate reorganization of actin and vinculin in cultured cells. J. Cell Biol., 99: 1045-1059, 1984. 5. Hynes, R. O., Destree, A. T., and Wagner, D. C. Relationship between microfilaments, cell-substratum adhesion and fibronectin. Cold Spring Har bor Symp. Quant. Biol., 46: 459-470, 1981. 6. Rohrschneider, L. R., Eiseman, R. N., and I filch, C. R. Identification of a Rous sarcoma virus transformation-related protein in normal avian and mammalian cells. Proc. Nati. Acad. Sci. USA, 76: 4479-4483, 1979. 7. Rohrschneider, L. R., and Gentry, L. E. Subcellular localizations of retroviral transforming proteins define multiple mechanisms of transformation. Adv. Oncol., 4: 269-305, 1984. 8. Feramisco, J. R., Gross, M., Kamata, T., Rosenberg, M., and Sweet, R. W. Microinjection of the oncogene form of the human H-ras (T-24) protein results in rapid proliferation of quiescent cells. Cell, 38: 109-117,1984. 9. Stacey, D. W., and Kung, H.-F. Tranformation of NIH 3T3 cells by microin jection of Ha-ros p21 protein. Nature (Lond.), 5/0: 508-511, 1984. 10. Balmain. A., and Pragnell, I. B. Mouse skin carcinomas induced in vivo by chemical carcinogens have a transforming Harvey-raj oncogene. Nature (Lond.), 303: 72-74, 1984. 11. Balmain, A., Ramsden, M., Bowden, G. T., and Smith, J. Activation of the mouse cellular Harvey-nu gene in chemically induced benign papillomas. Nature (Lond.), 307: 658-660, 1984. 12. Hsiao, W. L. W., Gattoni-Celli, C., and Weinstein, I. B. Oncogene-induced transformation of C3H 10T'/2 cells is enhanced by tumor promoters. Science (Wash. DC), 226: 552-555, 1984. 13. Dotto, G. P., Parada, L. F., and Weinberg, R. A. Specific growth response of ras-transformed embryo fibroblasts to tumour promoters. Nature (Lond.), 318:472-475, 1985. 14. Greenberg, M. E., and /ill. E. B. Stimulation of 3T3 cells induces transcrip tion of the c-fos proto-oncogene. Nature (Lond.), 311: 433-438, 1984. 15. Müller,R., Bravo, R., Burckhardt, J., and Curran, T. Induction of c-fos gene and protein by growth factors precedes activation of c-myc. Nature (Lond.), 5/2:716-720,1984. 16. Connan, G., Rassoulzadegan, M., and Cuzin, I. Focus formation in rat fibroblasts exposed to a tumour promoter after transfer of polyoma pit and myc oncogenes. Nature (Lond.), 314: 277-279, 1985. 17. Lerman, M. I., Hegamyer, G. A., and Colburn, N. H. Cloning and charac terization of putative genes that specify sensitivity to neoplastic transforma tion by tumor promoters. Int. J. Cancer, 37: 293-302, 1976. 18. Rifkin, D. B., and Crowe, R. M. Tumor promoters induce changes in the chick embryo fibroblast. Cell, 18: 361-368, 1979. 19. Blumberg, P. M., Driedger, P. E., and Rossow, P. W. Effect of a phorbol ester on a transformation-sensitive surface protein of chick fibroblasts. Na ture (Lond.), 264:446-447, 1976. 20. Zerlauth, G., and Wolf, G. Release of fibronectin is linked to tumor promo tion: response of promotable and nonpromotable clones of a mouse epidermal cell line. Carcinogenesis (Lond.), 6: 73-78, 1985. 21. Dion, L. D., Bear, J., Bateman, J., DeLuca, L. M., and Colburn, N. Inhibition by tumor-promoting phorbol estes of procollagen synthesis in promotable JB6 mouse epidermal cells. J. Nati. Cancer Inst., 69: 1147-1154, 1982. 22. Kalimi, G. H., and Sirsat, S. M. Phorbol ester tumor promoter affects the mouse epidermal gap junction. Cancer Lett., 22: 343-350, 1984. 23. Colburn, N. H., Vorder Bruegge, W. F., Bates, J. R., Gray, R. H., Rossen, J. D., Kelsey, W. H., and Shimade, T. Correlation of anchorage-independent growth with tumorigenicity of chemically transformed mouse epidermal cells. Cancer Res., 38:624-634, 1978. 24. Colburn, N., Former, B. F., Nelson, K. A., and Yuspa, S. H. Tumor promoter induced anchorage independence irreversibly. Nature (Lond.), 281:589-591, 1979. 25. Colburn, N. H., Lerman, M. I., Hegamyer, G. A., and Gindhart, T. D. A transforming activity not detected by DNA transfection of NIH 3T3 cells is detected by JB6 mouse epidermal cells. Mol. Cell. Biol., 5:890-893, 1985. 26. Diwan, B. A., Ward, J. M., Rice, J. M., Colburn, N. H., and Spangler, E. F. Tumor-promoting effects of di(2-ethylhexyl)phthalate in JB6 mouse epider mal cells and mouse skin. Carcinogenesis (Lond.), 6: 343-347, 1985. 27. Heimer, G. V., and Taylor, C. E. Improved mountant for immunofluorescence preparations. J. Clin. Pathol., 27: 254-256, 1974. 28. Osborn, M., Webster, R. E., and Weber, K. Individual microtubules viewed by immunofluorescence and electron microscopy in the same PCK 2 cell. J. Cell Biol., 77: R27-R34, 1978. 29. Furth, M. E., Davis, L. J., Fleurdelys, B., and Scolnick, E. M. Monoclonal antibodies to the p21 products on the transforming gene of Harvey murine sarcoma virus and of the cellular ras gene family. J. Virol., 43: 294-304, 1982. 30. Willingham, M. C., Pastan, I., Shih, T. Y., and Scolnick, E. M. Localization of the src gene product of the Harvey strain of MS V to plasma membrane of transformed cells by electron microscopic immunocytochemistry. Cell, 19: 1005-1014, 1980. 31. Colburn, N. H., Wendel, E., and Aborizzo, G. Dissociation of mutagenesis and late-stage promotion of tumor cell phenotype by phorbol esters: mitogen resistant variants are sensitive to promotion. Proc. Nati. Acad. Sci. USA, 78: 6912-6916,1981. 32. Colburn, N. H. The genetics of tumor promotion. In: }. Carl Barrett (ed.). Mechanisms of Environmental Carcinogenesis, Vol. 1. Boca Raton, FL: CRC Press, Inc., in press, 1986. 33. Nishizuka, Y. The role of protein kinase C in cell surface signal transduction and tumor promotion. Nature (Lond.), 308:693-698, 1984. 34. Wooten, M. W., and Wrenn, R. W. Redistribution of phospholipid/calciumdependent protein kinase and altered phosphorylation of its soluble and paniculate substrate protein in phorbol ester-treated rat pancreatic acini. Cancer Res., 45: 3912-3917, 1985. 35. Werth, D. K., Niedel, J. E., and Pastan, I. Vinculin, a cytoskeletal substrate of protein kinase C. J. Biol. Chem., 25«:11423-11426, 1983. 36. Meigs, J. B., and Wang, Y. L. Reorganization of a-actinin and vinculin induced by a phorbol ester in living cells. J. Cell Biol., 102: 1430-1438, 1986. 37. Mautner, V., and Hynes, R. O. Surface distribution of LETS protein in relation to the cytoskeleton of normal and transformed cells. J. Cell. Biol., 75:743-758, 1977. 38. Ali, I. V., and Hynes, R. O. Effect of cytochalasin B and colchicine on attachment of major surface protein of fibroblasts. Biochim. Biophys. Acta, 471: 16-24, 1977. 39. Kurkinen, M., Wartiovaara, J., and Vaheri, A. Cytochalasin B releases a major surface-associated glycoprotein from cultured fibroblasts. Exp. Cell Res., ///: 127-137, 1978. 40. Singer, I. I. The fibronexus: a transmembrane association of fibronectincontaining fibers and bundles of 5nm microfilaments in hamster and human fibroblasts. Cell, 16: 675-685, 1979. 41. Hynes, R. O., and Yamada, K. M. Fibronectins: multifunctional modular glycoproteins. J. Cell Biol., 95:369-377, 1982. 42. Chen, W. T., and Singer, S. J. Immunoelectron microscopic studies of the site of cell-substratum and cell-cell contacts in cultured fibroblasts. J. Cell Biol., 95: 205-222, 1982. 43. Boreiko, C., Mondai, Narayan, K. S., and Heidelberger, C. Effect of 12-Otetradecanoylphorbol-13-acetate on the morphology and growth of C3H/ 10T'/2 mouse embryo cells. Cancer Res., 40:4709-4716, 1982. 44. Nagel, D. S., and Blumberg, P. M. Activity of phorbol ester tumor promoters on enucleated Swiss 3T3 cells. Cancer Res., 40: 1066-1072, 1980. 45. Gindhart, T. D.. Nakamura, Y., Stevens, L. A., Hegameyer, G. A., West, M. W., Smith, D. M., and Colburn, N. H. Genes and signal transduction in tumor promotion: conclusions from studies with promoter resistant variants of JB-6 mouse epidermal cells. Carcinog. Compr. Surv., 8: 341-367, 1985. 46. Yamasaki, H., Enomoto, T., Shiba, Y., Kamm. Y., and Kakunaga, T. Inter cellular communication capacity as a possible determinant of transformation sensitivity of BALB/c 3T3 clonal cells. Cancer Res., 45:637-641, 1985. 47. Enomoto, T., and Yamasaki, H. Rapid inhibition of intercellular communi cation between BALB/c 3T3 cells by diacylglycerol, a possible endogenous functional analogue of phorbol ester. Cancer Res., 45: 3706-3710, 1985. 48. Weinstein, I. B., Gattoni-Celli, S., Kirschmeier, P., Lambert, M., Hsiao, W., Backer, J., and Jeffrey, A. Multistage carcinogenesis involves multiple genes and multiple mechanisms. In: Cancer Cells I/The Transformed Phenotype, pp. 229-237. Cold Spring Harbor Laboratory, New York: 1984. 49. Yuspa, S. H., Kilkenny, A. E., Stanley, J., and Lichti, U. Keratinocytes blocked in phorbol ester-responsive early stage of terminal differentiation by sarcoma viruses. Nature (Lond.), 314:459-62, 1985. 50. Colburn, N. H. Tumor promoter produces anchorage independence in mouse epidermal cells by an induction mechanism. Carcinogenesis (Lond.), /: 951954, 1980. 51. Hennings, H., and Yuspa, S. H. Two-stage tumor promotion in mouse skin; an alternative interpretation. J. Nati. Cancer Inst., 74:735-740, 1985. 52. Hoffman-Liebermann, B., Liebermann, D., and Sachs, L. Regulation of gene expression by tumor promoters. III. Complementation of the developmental program in myeloid leukemic cells by regulating mRNA production and in RNA translation. Int. J. Cancer, 28:615-620, 1981. 53. Colburn, N. H., Lerman, M. I., Hegamyer, G. A., Wendel, E., and Gindhart, T. D. Genetic determinants of tumor promotion: studies with promoter resistant variants of JB6 cells. In: Genes and Cancer, pp. 137-155. New York: Alan R. Liss, Inc., 1984. 54. Colburn, N. H., et al., in press, 1986. 5932 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research. Role of Cytoskeleton Changes and Expression of the H-ras Oncogene during Promotion of Neoplastic Transformation in Mouse Epidermal JB6 Cells Kiyoshi Takahashi, Ursula I. Heine, James L. Junker, et al. Cancer Res 1986;46:5923-5932. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/46/11/5923 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1986 American Association for Cancer Research.