[CANCER RESEARCH 50, 4779-4786, August 1, I990|

EJ/ras Neoplastic Transformation of Simian Virus 40-immortalized HumanUroepithelial Cells: A Rare Event1

Brian J. Christian, Chinghai Kao, Shi-Qi Wu, Lorraine F. Meisner, and Catherine A. Reznikoff2

Department of Human Oncology; University of Wisconsin Clinical Cancer Center, Madison, Wisconsin 53792 [B. J. C., C. K., C. A. R.], and Cytogenetics Section,University of Wisconsin, State Laboratory of Hygiene, Madison, Wisconsin 53 706 fS-Q. W., L. F. M.J

ABSTRACT

To determine if expression of mutant p21 ras could convert SimianVirus 40-immortalized human uroepithelia) cell line (SV-HUC) to tumorigenicity, SV-HUC cells were transfected with pSV2-neo (a neomy-cin-resistant gene) or PREJ/ras (c-HA-rai-1 with the 12th codon mutation and neo). Seven independent G418-resistant clones (A—»G)wereisolated from each group (SV-HUC/r<w and SV-HUC/neo). SV-HUC/ras clones were morphologically altered, while SV-HUC/neo clones retained a typical SV-HUC epithelial morphology. Electrophoretic analysisof immunoprecipitated ras proteins detected altered p21 ras protein infour of seven SV-HUC/ras clones at passage (P)2 and in five of sevenclones at P12 posttransfection. The relative levels of raíp21 differedamong the clones and appeared to increase with passage in culture. RNAand DNA dot blot analyses showed that clones with more abundantmutant p21 also had higher ras RNA levels and, in one case, increasedras gene copy number. No altered ras protein was detected in any SV-HUC/neo clones, ras- and neo-transfected clones were tested for tumor-¡genicityat P2 posttransfection and again at PI 2 by four s.c. inoculationseach into athymic nude mice. None of 56 inoculations of SV-HUC/neoclones was tumorigenic. None of the SV-HUC/rai clones at P2 gave riseto tumors at all four injection sites. However, two roi-transfected clones,SV-HUC/roi-B and SV-HUC/ras-F, produced one tumor each. Oneclone, SV-HUC/rai-D which produced abundant mutant p21, was negative when inoculated at P2, but produced tumors in four of four siteswhen reinoculated after ten passages in vitro. All tumorigenic clones haddetectable levels of mutant ras p21. However, the relative levels of alteredp21 ras protein among the SV-HUC/ros clones did not directly predicttheir tumorigenic potential, as several nontumorigenic SV-HUC/raiclones had protein levels equal to or higher than the most tumorigenicclone (SV-HUC/ruî-Dat PI 2). Cell lines established from the tumorexpiants exhibited higher ras gene copy numbers, higher RNA levels,and more abundant p21 than was seen in the clones at the time ofinoculation. Therefore, increases in ras protein abundance occurred duringtumor formation in vivo, as well as during passage of cells in culture, andsuch cells apparently had a selective growth advantage. However, expression of abundant mutant ras protein was not in itself sufficient forneoplastic transformation of SV-HUC. These data are consistent with amodel of transformation in which mutant ras protein in combination withat least one additional, rare, and stochastically occurring event contributesto neoplastic transformation of SV-HUC and, in high abundance, providescells with a selective growth advantage in vitro and in vivo.

INTRODUCTION

The high incidence of ras gene alterations in human androdent tumors strongly implicates activation of these genes inmammalian cell carcinogenesis (1-7). However, the presenceof altered raígenes in cancer cells does not per se prove a causalrelationship. Therefore, numerous studies using diverse systemsboth in vivo and in vitro have attempted to address the role ofras gene alterations in multistep neoplastic transformation ofmammalian cells.

Received 8/10/89; revised 2/8/90.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.

' Supported by the Haertle Foundation and by NIH Training Grant 5-T32-CA09474-03 (B. C.) and NIH Grant CA-29525-08 (C. A. R.).

2To whom reprints for reprints should be addressed.

An important approach in addressing this question has beento study the properties of cultured cells after transfection withras genes (3). There are many reports of studies examining thecharacteristics of ras genes transfected into rodent cells inculture. Mutant ras can neoplastically transform some immortal cell lines, including NIH/3T3 and C3H/10T'/2 (8-11). How

ever, mutant ras does not neoplastically transform all established cultures (12, 13), possibly reflecting important differences in the target cells. The quantity of mutant ras proteinaffects the transformed phenotypes (14-18). For example, mutant ras extends the life of diploid rodent cells in cultures, butresults in neoplastic transformation only when expressed inhigh abundance (19). The presence of a second nuclear actingoncogene may cooperate with ras in transforming cells (20-24). For example, ras with myc or adeno-£7/4 neoplasticallytransforms diploid rodent cells, while ras alone is not transforming (20, 21, 23, 24). However, some evidence suggests thattransformation of diploid rodent cells by two cooperating on-cogenes may require an additional event. For example, tumorigenic transformation of diploid hamster cells by myc and rasis highly correlated with segregation of a specific chromosome(25). Normal ras protein, when expressed in high abundance,also has transforming ability for rodent cells, but this is limitedcompared with mutant ras (14, 19, 26).

In general, human cells, particularly epithelial cells, havebeen more difficult to transform in vitro than rodent cells (27).There are a few reports of neoplastic transformation of humanepithelial cells by viral ras oncogenes. For example, humanbronchial epithelial cells were transformed to tumorigenicityafter transfection with v-Ha-ras3 (28). Human keratinocytesimmortalized after infection with an adeno-12-SV40 hybridand human amniocytes immortalized by SV40 converted totumorigenicity after infection with KiSV (29, 30), which contains a ras oncogene related to v-Ha-ras (31). Recently, it hasbeen reported that human keratinocytes immortalized by HPV-16 DNA were converted to tumorigenicity after transfectionwith v-Ha-ras (32). These results taken together support amodel in which mutant ras may be sufficient for transformationof certain immortalized human epithelial cells. In contrast, ithas been reported that \-myc and v-ras do not transform humanlymphocytes (33).

In this paper, we present the first example of transfection ofimmortalized human epithelial cells with a ras oncogene isolated from a human carcinoma, rather than from a retrovirus.For these studies, we have used the well-characterized ras

oncogene from the EJ(T24) human bladder cancer cell line. EJ/ras carries point mutations at the 12th codon (9, 34) and in thelast intron (35). These mutations are thought to be responsiblefor its transforming activity and increased gene expression.Significantly, we have for the first time transfected the EJ/ras

3The abbreviations used are: v-Ha-ras, Harvey vial ras; SV40, Simian Virus40; KiSV, Kirsten sarcoma virus; HPV-16, human papilloma virus 16; SV-HUC,SV40-immortalized human uroepithelial cell; HUC, human uroepithelial cell;PI, passage 1 (P2, P12, and P100 defined similarly); PBS, phosphate-bufferedsaline; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis;FBS, fetal bovine serum.

4779

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/rasHl

gene back into a cell type from which it was originally isolated,namely, human uroepithelial cells. The cell line that was usedis an SV40-immortalized-HOC (SV-HUC) that was clonal inorigin and pseudodiploid (36, 37). We have previously shownthat SV-HUC can be neoplastically transformed after treatmentwith the polycyclic hydrocarbon carcinogen. 3-methylcholan-threne (38), and/or the human bladder carcinogen, 4-aminobi-phenyl (39). In this paper we present an analysis of the capability of the ras oncogene to neoplastically transform SV-HUC.These results show that expression of abundant mutant ras-encoded protein is insufficient to neoplastically transform SV-HUC but, in combination with an additional stochasticallyoccurring event, contributes to neoplastic transformation.

MATERIALS AND METHODS

Cells and Culture Methods. A clonal line of SV-HUC was used inthis study. The characteristics of SV-HUC have been described (36,37). Briefly, SV-HUC was derived from a culture of normal HUC afterinfection at PI with SV40. SV-HUC cells have remained nontumori-genic when tested in nude mice (>150 inoculations using IO6 to IO7

cells/site) after 80 serial passages and more than 2 yr in continuousculture. SV-HUC cells are 100% positive for human epithelial keratinsand for SV40 T-antigen, but do not produce infectious SV40. Southernblot analysis shows a single integration site of SV40 in SV-HUC.4

Cytogenetic analysis shows that SV-HUC maintains a near balancedkaryotype (37). In the present study, SV-HUC were grown on plasticdishes (Corning, Corning, NY) in 1% FBS-F12+, a supplementedmedium developed for human uroepithelial cells (40). The basal medium is Ham's F12 (GIBCO, Grand Island, NY), with 5 Mg/ml of

insulin (Eli Lilly and Co., Indianapolis, IN), l Mg/ml of hydrocortisone(Merck Sharpe and Dohme, West Point, PA), 5 jig/ml of transferrin(Sigma Chemical Co., Saint Louis, MO), 0.1 IHMnonessential aminoacids (Microbiological Associates, Walkersville, MD), 2.0 HIM L-glu-tamine (GIBCO), 15 mM dextrose (Amend Drug and Chemical Co.,Irvington, NJ), 100 units/ml of penicillin (Pfizer, Inc., New York, NY),100 Mg/ml of streptomycin (Pfizer), and 1% FBS (HyClone, Logan,UT) added. Cultures were grown in humidified incubators at 37°Cin

an atmosphere of 5% CO2 and 95% air. The medium was changed 3times per wk. Routine dispersion of the cultures for serial passage wasperformed using 0.1% EDTA (Sigma Chemical Co.) dissolved inHanks' balanced salt solution (GIBCO) as previously described (36).

Plasmids and DNA Transfection Techniques. SV-HUC were trans-fected with either pSV2neo (41) or PREJ/us (42) using the calciumphosphate precipitation method of Graham and van der Eb (43) withthe modifications suggested by Manoharan et al. (10). The neo geneconfers resistance to the neomycin analogue G418 (GIBCO); PREJraicontains the 6.6-kilobase pair BamHl restriction fragment of the mutated c-Ha-ras-1 gene inserted into the BamHl site of pSV2neo. PlasmidDNAs were added to a transfection solution which contained 140 mviNaCl, 5 mM KC1, 0.7 mM Na2HPO4, 6 mM dextrose, 20 mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonicacid, and 120 mM CaCl2. SV-HUC were seeded into 100-mm culture dishes (5 Mg/ml) 24 h prior totransfection at a density of IO6cells/dish. After 24 h, when cells were

beginning the log phase of growth, the medium was changed (4 ml perP100), and the transfection solution (2 ml per P100) contraining 10 Mgof DNA was added. After a 5-h incubation period, the medium containing the DNA was removed. Cells were then exposed to 15% glycerol inPBS for 4 min to enhance the uptake of DNA. The cultures were rinsedtwice with PBS, and 10 ml of fresh growth medium (1% FBS-F12+)were added. Twelve h following transfection, the cultures were dissociated with EDTA and subcultured in 1% FBS-F12+ containing theselective agent G418 (200 Mg/ml). At 2 wk posttransfection, 7 independent G418-resistant colonies (A through G) from each group (SV-HUC/neo and SV-HUC/ras) were ring isolated using a cloning cylinder. Clones were passed first to a 30-mm culture dish and then to a

4 Manuscript in preparation.

100-mm culture dish. At P2, cells were expanded for animal inocula

tions, isolation of DNA and RNA for dot blot assays, and p21 analyses(see below). Cells were serially passaged in G418-containing mediumfor the same analyses at PI 2 (after 5 mo in culture). Aliquots of cellsat both P2 and PI2 were cryopreserved.

Tumorigenicity Assays and Establishment of Tumor Cell Lines. Cellsfrom each independent group [7 clones (A—>G)of pSV2neo- and 7clones (A—>G)of PREJras-transfected SV-HUC] were injected s.c. into4-wk-old female athymic nude mice (nu/AF; Harlan-Sprague-Dawley,Indianapolis, IN) using 2.5 x 10" cells/site in 0.2 ml of 1% FBS-F12+.

As a positive control for tumorigenicity, animals were inoculated withthe tumorigenic cell lines, T24 and HS242. These animals formedtumors at 4 of 4 sites within 6 wk. Some clones (see below) were alsoinoculated into mice i.v. via the tail vein using 1 x IO6cells in 0.1 mlof 1% FBS-F12+ per animal. Mice were housed in sterile laminar flowcages in a temperature (30°C)-and humidity (60%)-controlled room.

Animals without tumors were sacrificed after 1 yr. When tumorsdeveloped, these were removed when they reached a size of approximately 1 cm in diameter. A representative section of each tumor wasimmediately fixed in phosphate-buffered formalin for histopathology.The liver, lungs, spleen, and kidneys of tumor-bearing animals werealso fixed for histopathology. The remainder of the tumor was dissectedinto expiants using published techniques (44) to initiate tumor cellcultures. These were grown in G418-containing medium to select forthe inoculated human cells. The tumor lines which were derived fromexpiant cultures of tumors produced by SV-HUC/ras clones B and Fand one of the tumors produced by SV-HUC/ras-D were designatedSV-HUC/rai TB, TF, and TD, respectively. Cell lines derived from

tumors were retested for tumorigenicity by s.c. and i.v. inoculation intoathymic nude mice. All the cell lines derived from SV-HUC/ras tumorsgave tumors in all animals at 4 of 4 sites by both routes within 4 wk.

Analysis of p21 ras Proteins. Analysis of the electrophoretic mobilityof immunoprecipitated ras proteins was performed according to themethod of Srivastava et al. (45). Briefly, proliferating clones ofpSV2neo- and PREJras-transfected SV-HUC were labeled for 4 h with200 MCiper ml of [35S]methionine (1100 Ci/mmol; Amersham, Arling

ton Heights, IL). Labeled cells were then lysed and immunoprecipitatedwith antibody Y13-259 (Oncogene Science, Inc., Manhasset. NY) whichbinds to Ha-, K-, and N-ras-encoded p21. The immunoprecipitatedsamples were analyzed by 15% SDS-PAGE and fluorographed. Todemonstrate that p21 with altered mobility could be detected in ourgels, the T24 bladder carcinoma cell line, which has a mutation at the12th codon of the c-Ha-rai-1 gene resulting in a p2I with a slowermobility, and the HS242 lung carcinoma cell line, which has a mutationin the 61st codon of the c-Ha-ras gene resulting in increased mobilityof p21 ras, were used as controls in every gel.

Analyses of ras DNA and RNA by Dot Blot. High-molecular-weightDNA and total cellular RNA for dot blot analyses were isolated asdescribed (46) from 6 confluent P100 cultures each of pSV2neo andPREJrai clones of SV-HUC expanded at P2. Briefly, cells were collected, lysed in guanidine isothiocyanate, and subjected to CsCl2 densitygradient centrifugation. The probe used for dot blot analyses was a 3.0-kilobase pair restriction fragment of ras isolated from a Sac\ digest(New England Biolabs, Beverly, MA) of PREJras. The probe waslabeled with 32P using a nick translation kit (Boehringer Mannheim

Biochemicals, Indianapolis, IN). For dot blot analyses, aliquots of DNA(or RNA) were spotted onto Gene Screen hybridization transfer membranes (Dupont, Boston, MA), and then baked at 90°Cfor 3 h. Blotswere hybridized with the "P-labeled probe (IO6 cpm/ml) using the

conditions specified in the Gene Screen instruction manual. The dotblots were then exposed to XAR film (Eastman Kodak Company,Rochester, NY) in the presence of intensifier screens (E. I. Dupont deNemours and Co., Wilmington, DE). Each dot blot was repeated 3times. As a control for DNA and RNA loading, one of these dot blotswas denatured, washed, and rehybridized with 0-actin using the sametechniques. Quantitation of the dot blot results was determined usingan LKB soft laser beam densitometer (LKB, Rockville, MD).

Karyotypic Analyses. Karyotypic analyses were performed using published techniques (37, 47) on cell lines initiated from tumor expiantsas soon as sufficient cells were available (P2). For comparison, kary-

4780

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/rajH

otypic analyses were also performed on the clones at the passages whichresulted in tumor production. Samples of these clones were retrievedfrom the liquid N2 freezer, cultured, and karyotyped as soon as possible.For each karyotypic analysis, 50 Giemsa-banded metaphase spreads

were analyzed.

RESULTS

Clones of ras-transfected SV-HUC Show MorphologicalChanges. Seven independent G418-resistant clones (A—>G)ofras-transfected SV-HUC were isolated. Compared with theparental clonal cell line which shows a flat, tightly adherentepithelial morphology (Fig. la), SV-HUC/ras clones wereloosely packed, rounded, and more refractile. SV-HUC/raiclones C and G also showed criss-crossing of cells with bipolarmorphology (Fig. Ib). Thus, all ras-transfected cells showed

morphological alterations. In contrast, seven independent

Fig. 1. Phase-contrast photomicrographs showing the morphologies of SV-HUC/neo and SV-HUC/rai cells at P4 posttransfection. SV-HUC/nro clone A(a) appeared as a monolayer of uniformly shaped epithelial cells indistinguishablefrom the SV-HUC parent culture. SV-HUC/ros clones grew in a more looselypacked fashion than did SV-HUC, and some exhibited a polar morphology withcriss-crossing of cells as illustrated by clone C (ft), x 150.

G418-resistant /leo-transfected SV-HUC clones showed morphologies indistinguishable from the parental SV-HUC.

Tumorigenic Transformation of SV-HUC after ras-Transfection Is Rare. The parental clonal SV-HUC cell line hasremained nontumorigenic through 80 passages and more than150 inoculation sites (33, 38, 40). The 7 independent neo-transfected SV-HUC control clones were inoculated at P2 andPI2 into nude mice using 4 sites at each passage. All of the 56control inoculation sites remained negative (Table 1). To determine if mutant ras could convert SV-HUC to tumorigenicity,seven independent clones (A—>G)of ras-transfected SV-HUCwere isolated and inoculated into nude mice using 4 sites pergroup (Table 1). Control inoculations into mice were doneusing tumorigenic cell lines carrying mutant ras genes (HS242and T24), and all gave tumors at 4 of 4 sites (2- to 6-wk latentperiods). However, none of the seven ras-transfected clonesgave tumors at all 4 sites of inoculation. One clone, SV-HUC/ras-B, gave one tumor from 4 sites injected after a 4-wk period.

Another clone gave one tumor from 4 sites inoculated after alonger latent period of 20 wk. Neither of these clones producedtumors at any of the other 3 inoculation sites, and neitherproduced tumors when reinoculated at P12, suggesting thattumors obtained from the first inoculations resulted from raretransformants in the population. Cell lines initiated from thesetumors all produced tumors at P2 with a short latent period atall 4 sites inoculated, which shows that the lack of tumors atall sites in the original inoculations was not related to insensi-tivity in the assay. One clone, SV-HUC/raj clone D which wasnegative for tumorigenicity at P2, produced tumors at 4 of 4sites at P12, all with a short latent period. This result suggeststhat a neoplastic transformant which emerged in this clonebetween P2 and PI 2 became expanded during the period of cellculture, resulting in tumorigenicity at all sites. These resultsare consistent with a model in which neoplastic transformationof SV-HUC by EJ/ras requires a rare additional, randomlyoccurring event.

Although the latent periods to first appearance of tumorsamong the clones varied, once the tumors appeared, all grewrapidly and progressively, reaching approximately 1-cm diameter in size within 3 wk (Fig. 2a). Likewise, the histopatholog-ical characteristics of all the tumors were similar. These wereclassified as high-grade poorly differentiated transitional cellcarcinomas (Fig. 26). These tumors showed a high mitotic

Table 1 Tumorigenicity of PREJ/ras and pSV2/neo clonesIndependent clones growing in G418 were ring isolated and inoculated into

nude mice at P2 and again at PI 2 after 5-mo growth in vitro. The positive controlsfor tumorigenicity (T24 and HS24) gave tumors at 4 of 4 sites within 6 wk.

Tumors/sitesinoculatedCloneSV-HUC/roiABCDEFGSV-HUC/neoABCDEFGP20/4-1/4

(4)"0/40/40/4-.1/4(20)0/40/40/40/40/40/40/40/4P120/40/40/4-4/4

(3)ND*ND0/40/40/40/40/40/40/40/4

' Numbers in parentheses, latent period (wk).' ND, not determined. These clones were lost at late passage.

4781

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY El/ras"

*f?%&¡

fcgtólfe"Vi

:.

Fig. 2. Characteristics of SV-HUC/ros tumors as illustrated by Tumor D. Subcutaneous tumors grew to approximately 1 cm in diameter within 3 wk after firstappearance (a). The SV-HUC/raj-induced tumors were classified as Grade IV poorly differentiated transitional carcinomas with high mitotic activity (ft). Tumor celllines derived from the rai-induced tumors formed colonies in the lung (c) when inoculated i.v. into athymic nude mice.

index, occasional foci of ischemie necrosis, and invasion intoadjacent muscle and adipose tissue. None of these subcutaneoustumors spontaneously metastasized to the liver, lung, spleen,or kidneys. The cell lines initiated from the tumor expiants(SV-HUC/ras clones B, F, and D) produced tumors with thesame histopathology as the parental tumors at 4 of 4 sites afters.c. inoculation. When inoculated i.v., all three tumor cell linesformed colonies in the lungs within 2 wk (Fig. 2c).

Mutant p21 Contributes to, but Alone Is Insufficient for S\ -HUC Tumorigenicity. To test if differences in expression ofaltered ras-encoded p21 in the different clones explained thedifferences in their tumorigenicity at the different passages, theelectrophoretic mobilities of p21 were examined in SV-HUC/ras and SV-HUC/neo at P2 soon after the clones were ringisolated, and at PI 2 after a 5-mo period of cell culture (Fig. 3).Results showed that 4 of the 7 ras-transfected SV-HUC clones(clones A, B, D, and F) showed mutant p21 with the expecteddecreased mobility at P2 (Fig. 3a). One clone, SV-HUC/ras G,appeared negative for mutant p21, but showed the alteredprotein at PI2 (Fig. 3A), demonstrating an increase in rasexpression after a period of in vitro culture. Altered p21 wasnot detected in SV-HUC/ras C even after passage in vitro (Fig.3A), even though this was one of the two most morphologicallyaltered clones (Fig. 1¿>).Neither the parental SV-HUC cell line,nor any of the SV-HUC/neo clones produced altered p21 at P2or P12. These results were reproducible in 4 gels at eachpassage.

All clones which were tumorigenic expressed detectable rasprotein, but none of the clones which did not express mutantp21 was tumorigenic. However, if simple expression of a detectable level of altered p21 was sufficient for tumorigenicity, 4of the 7 clones should have produced tumors at P2 and onemore at P12. In fact, only 2 clones, SV-HUC/ras-B and SV-HUC/ras-F, produced tumors at P2, and at only one site each.Furthermore, there was no obvious correlation between thelevel of p21 and the tumorigenicity of some clones comparedwith other clones. The presence of altered p21 was barelydetectable in clone B, which produced a tumor with a relativelyshort latent period, while SV-HUC/ras clone F showed a high

4782

B C O E F

SV-HUC/neo P2

A B D E F G

SV-HUC/ras P2

P21

A B C D E F

SV-HUC/neo P12

l tA B C D G

SV-HUC/ras P12

Fig. 3. Electrophoretic mobilities of p2l ras-encoded proteins immunoprecip-itated from ["S]methionine-labeled cell extracts with antibody Y13-259. Theautoradiograms were produced from equal amounts of total labeled protein(30.000 cpm/lane) and compare SV-HUC/nco- and SV-HUC/ras-transfectedcells (clones designated A through G) at P2 posttransfection (a) and PI2 post-transfection (b) with controls. Lane I in a is the SV-HUC parent control culture.The HS242 lung carcinoma cell line (J) and the T24 bladder carcinoma cell line(ÃŽ)served as controls for altered p2l mobilities.

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/rasHi

level of expression, but produced only one tumor with a long20-wk latent period. SV-HUC/ras-A and SV-HUC/ras-D hadprotein levels similar to the tumorigenic clone F and muchgreater than tumorigenic clone B, but neither of these cloneswas tumorigenic. Thus, the abundance of altered p21 at P2 inthe independent clones did not predict their tumorigenic potential. The same result was observed at PI2 (Fig. 36). One clone(SV-HUC/ras-D) showed abundant protein at P12 and gaverise to tumors at 4 of 4 sites. However, SV-HUC/ras clones Aand G showed equally abundant protein, but yielded no tumors.

The levels of p21 were examined in three tumor cell lines(TB, TD, and TF) initiated from expiants of tumors obtainedafter inoculation of ras-transfected clones into nude mice. Inall cases, the level of expression of p21 ras-encoded protein waselevated compared with the cell line which was inoculated (Fig.4). These results are consistent with a model in which increasedexpression of ras occurs in vitro and/or in vivo and renders aselective growth advantage.

Increased p21 Correlates with Increased ras DNA and/or RNALevels. Dot blot analyses and quantitative densitometric determinations for DNA and RNA levels were done on SV-HUC/ras and SV-HUC/neo clones to support SDS-PAGE results by

determining if cells with increased mutant p21 also showedincreased ras DNA and/or RNA levels (Fig. 5). DNAs andRNAs were prepared from neo- and ras-transfected clones and

the tumor cell lines as soon as sufficient cells were available.Results of RNA analysis showed a good correlation betweenp21 abundance and ras RNA levels, with clones with the mostabundant protein (i.e., SV-HUC/ras clones A, B, D, and F)showing 1.6 to 3.6 times the ras RNA levels seen in the SV-HUC/neo controls. The SV-HUC/ras TB and TD tumor celllines (which showed even greater p21 abundance than any ofthe 7 original clones) were 9 and 5 times amplified, respectively,for ras RNA when examined at early passage. DNA dot blotanalysis showed that only one of the ras-transfected clones (SV-HUC/ras-D) had an increased ras gene copy number at P2 (6times compared to SV-HUC/weo controls), probably resultingfrom an efficient transfection. This increased gene copy numberwas also seen in the tumor (TD) obtained after inoculation ofthis clone. One tumor line (TB) showed a 5-fold amplificationof ras DNA in the tumor compared with the control. Therefore,

DNA RNA

P2'

I tB TB D D TD

P2 P2 P12F TF

P2Fig. 4. A comparison of quantities of altered p21 ras-encoded proteins in three

clonal SV-HUC/ras tumor cell lines with each SV-HUC/ras precursor clone isshown here. SV-HUC/ras clones B, D, and F are shown at P2 and PI2 posttrans-fection. Tumor cell lines are designated TB. TD, and TF and were tested at P2.Other conditions are as in Fig. 2.

SV-HUC/neo P2

SV-HUC/ras P2

ABCDEFGTD ABCOEFGTD

Fig. 5. Dot blot analysis of ras DNA and RNA in SV-HUC/neo and SV-HUC/ras cell lines at P2 posttransfection, and of two cell lines established fromthe ras-induced tumors at P2 (TB and TD). High-molecular-weight DNA andtotal cellular RNA were isolated from confluent cell cultures. Aliquots of DNA(5.0, 2.5, 0.5. and 0.125 Mg)and RNA (10.0, 5.0, 2.5, 1.0, 0.5, and 0.25 Mg)werespotted on nitrocellulose filters. The filters were hybridized with a 32P nick-translated 3.0-kilobase Sad restriction fragment from PREJras. TB and TD referto DNAs from the tumor cell lines from SV-HUC/ras clones B and D. TB DNAor RNA is in the last lane with the neo samples. TD DNA or RNA is in the lastlane with ras samples. Lanes A lo G refer to clone designations.

the increased p21 observed in the tumors, as well as in SV-HUC/ras clone D, can be explained by increased ras gene copynumber. In the other clones, the increased p21 could be attributed to either the 2- to 3-fold increased transcription reportedfor the EJ/ras gene (35) or increased mRNA stability. Thelevels of ras RNA and DNA in control SV-HUC/neo clonesA—>Gvaried by a maximum of 0.1-fold and 0.3-fold, respectively. When DNA and RNA dot blots were also probed using/3-actin, no increases in /3-actin gene copy number were seen inany clones including SV-HUC/ras-D, which showed a 6-foldincrease in ras gene copy number. Furthermore, there were noincreases in 0-actin RNA levels in the 3 independent SV-HUC/ras clones (B, D, and F) which showed a 1.6-, 3.1-, and 3.6-foldincrease in ras RNA, respectively. One clone (A) which was1.6-fold amplified for raíRNA was also 1.5-fold amplified for

actin RNA.Tumor Cell Lines Show SV-HUC Karyotypes with Deletions.

To confirm that the tumors were human and derived from theinoculated SV-HUC/ras clones, karyotypes of the tumors weredone and were compared with the inoculated clones and theparental SV-HUC cell line. The parental SV-HUC cell lineshowed a monosomic segment, 7q|32-34j and 3 trisomie arms(8p, llq, and 14q). All of the ras- and weo-transfected clonesshowed the same monosomic and trisomie segments, demonstrating the clonal origin of the cells used for transfection. Atearly passage (P2), most of the transfected clones showed nochromosomal losses or gains (Table 2). Thus, gross chromosomal alterations were not associated with the transfection orcloning of SV-HUC. However, after serial passage, all of theclones examined lost one or more chromosomal segments.

The three tumor lines showed additional chromosomal lossescompared with the clones at the time of inoculation. Two ofthese lost the extra 8p seen in the parental line. These samelines both showed a net loss of 1 or Ip, and 3 or 3p (Table 2).Neither of the parental clones showed losses of these chromosomes, which were the only apparently nonrandom chromosomal losses associated with tumorigenicity. The third cell lineSV-HUC/ras TD did not show gross losses on chromosome 1or 3 by cytogenetic analysis, but showed a loss of 21q, which

4783

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/ras"1

Table 2 Chromosomal losses and gains in SV-HUC/ras clones and tumor lines

This table shows cytogenetically visible losses and gains of chromosomalregions after transfection of SV-HUC with E]/ras. The parental SV-HUC cellsused for this experiment showed del 7q[32-34| and trisomies of 8p, 1Iq, and 14q.All of the transfected clones and tumor lines showed these same changes, exceptthat 2 tumor lines lost the extra 8p. Transfected cell lines tended to lose, ratherthan gain, chromosomes, and changes were more numerous after passage inculture. All of the cell lines including the 3 tumor lines were clonal with >90%of cells showing the same changes. The only net chromosomal losses or gainscorrelated with tumorigenicity were deletions of 3 or 3p and Ip. Clones B and Fwere not analyzed at PI2 because they gave a tumor at P2.

SV-HUC/ras

Clones P2 P12 Tumors SV-HUC/r<w, P2

ABCDF08q0000ND"10p,21,y8qNDTBNTNTTDTF-lp,-3

-8p,-lq(42.1-ter),y+22q,4q[21-25],-.2/i+

\3,+ lS,-lp,-3p.-Sp* ND, not done; NT, no tumor.

has also been observed in human bladder cancers (48). Cyto-genetic losses are currently being examined using restrictionfragment length polymorphism analysis; these results will bereported when completed. Cytogenetic analysis of all threetumor cell lines confirmed the clonal origin of each clone andshowed that the tumors were also clonal in origin, as all thecells showed the same chromosomal alterations. These resultsare consistent with a model in which certain chromosomaldeletions may be required for tumorigenic conversion of ras-transfected cells, and cells with these losses are selected by theirtumorigenicity. This model is currently being tested in ourlaboratory.

DISCUSSION

In the present study, the ability of the human bladder cancerEJ/T24 ras oncogene to neoplastically transform SV-HUC wastested. A significant aspect of these experiments is that theywere designed to permit assessment of the tumorigenic potentialand levels of raígene expression in independent clones of ras-and weo-transfectants of the clonal SV-HUC cell line both soonafter transfection and then again after an extended (5-mo)passage of cells in culture.

These results support a role for mutant ras p21 in transformation, as no clone without a detectable level of mutant rasp21 was tumorigenic. The results also show that the additionof mutant ras alone is insufficient to neoplastically transformSV-HUC, even when ras p21 is abundantly expressed. If expression of mutant ras alone were capable of transforming SV-HUC, then most of the SV-HUC/ras clones should have beentumorigenic (Fig. 3). Instead, only one of 4 sites in 2 of the 7independent clones gave rise to tumors at early passage, andneither of these clones was tumorigenic at a later passage (Table1). Therefore, our results are more compatible with a model inwhich only a few cells in some of the clones were tumorigenic(Table 1). This tumorigenicity could not have resulted fromaltered ras alone or from a specific integration site of ras withintumor-producing clones, or all of the cells in the expandedprogeny of such clones should have been tumorigenic, even atearly passage. It should be noted that, if the ras-transfected SV-HUC cells were not cloned and were not tested at early passage,it is probable that tumors would have been obtained in allanimals inoculated with pooled ras-transfected cells. Thus, ourresults would simply have confirmed the results of others, whosedata suggest that the ras oncogene is sufficient to neoplasticallytransform immortalized human epithelial cells (28-30, 32).

It could be argued that failure of some clones of SV-HUC/

ras to attain some "threshold level" of mutant ras expression

explains their nontumorigenicity. This argument is consistentwith the observation that none of the SV-HUC/ras clones withlow or undetectable ras protein levels was tumorigenic. However, this result does not explain the nontumorigenicity ofseveral of the ras-transfected clones (A and G), in which thealtered ras protein levels were equivalent to (or exceeded) thoseof the most tumorigenic clone (i.e., clone D at P12). Furthermore, ras expression in the tumors was not significantly higherthan in some of the clones (e.g., clone A at P12) which remainednontumorigenic. Therefore, we cannot say that a threshold levelof altered p21 ras is the sole requirement for tumorigenicconversion of SV-HUC. Nevertheless, since all of the tumorsshowed a higher abundance of p21 ras than the same clonesbefore inoculation, our results do suggest that cells with a higherras expression may have a tumor growth advantage in vivo.These results also show that "spontaneous" amplification of

ras occurred in one tumorigenic clone (SV-HUC/ras-B), andthat these cells were subsequently selected in vivo. This interpretation is consistent with the idea of clonal selection withintumors (49) and with observations that high ras expression iscorrelated with high tumorigenicity (14, 16, 17).

The SV-HUC/ras transformants, although rare, producedmalignant, rapidly growing carcinomas (Fig. 2). In contrast, thenontumorigenic parental SV-HUC cell line and the n^o-trans-fected control clones did not produce tumors or nodules of anynature in any animal on inoculation, even after cloning andserial passage in vitro. The tumor cell lines when reinoculatedinto nude mice rapidly formed tumors at all sites of inoculationwith a short latent period. These results show that the lack oftumorigenicity at all injection sites in the original inoculationsof the SV-HUC/ras clones (Table 1) cannot be explained by alow tumorigenic potential of ras transformants in the nudemouse system or by insensitivity in the assay.

All of the above results taken together suggest a model oftransformation in which mutant ras together with at least oneadditional event, which apparently occurs stochastically andrarely, leads to neoplastic transformation of SV-HUC. It wouldbe of great interest to determine the nature of this event. Thekaryotypic analysis of ras transformants fortuitously may haveprovided a possible clue to such a mechanism. Cytogeneticanalysis shows very few chromosomal alterations associatedwith transfection, cloning, and serial culture of SV-HUC (Table2). It is, therefore, significant that 2 of the 3 independent tumorcell lines showed a net loss of 2 of the same chromosomal arms,namely Ip and 3p, and the third line showed a loss of 21q,which has been described as the sole Cytogenetic change in somebladder cancer subtypes (48). The parental clones from whichthese tumors derived showed no such losses. Losses of regionson chromosomal arms Ip and 3p have been observed in anincreasing number of human carcinomas (50) and, therefore, ithas been hypothesized that these regions may contain tumorsuppressor genes (51). While it is premature to suggest thatloss of such genes is a necessary step in ras transformation ofSV-HUC, it is interesting that similar losses (i.e., 3p) have alsobeen observed in chemically transformed SV-HUC (38, 40).Furthermore, several lines of evidence suggest that suppres-sor(s) of ras transformation may be present in nontumorigeniccells. For example, somatic cell hybrids between tumorigenicras-transfected cells and nontumorigenic cells in our system5

and in other systems (52,53) are nontumorigenic. Furthermore,neoplastic transformation of diploid hamster cells after trans-

' Unpublished data.

4784

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/roî1"

faction with myc and ras is correlated with specific chromosomeloss (25). In addition, there are several reports of failure ofmutant ras to transform immortalized cell lines (12, 13). It ispossible that such cell lines have not lost (or do not easily lose)the hypothesized cancer suppressor gene(s), in contrast to ras-transformable cell lines.

In summary, this paper presents a model of transformationin which mutant ras participates in neoplastic transformationof SV40-immortalized human epithelial cells, but of itself isinsufficient. The data suggest that an additional event is required for tumorigenic conversion. The nature of this event hasnot been identified. However, results suggest a possibility thatloss of suppressor genes may be involved. Studies to addressthis hypothesis are in progress in our laboratory.

ACKNOWLEDGMENTS

We wish to thank Dr. John C. Roder and workers for supplying thePREJ plasmid and Dr. Stuart A. Aaronson for providing us with theHS242 cell line. We are grateful to Dr. Ryoichi Oyasu for histopatho-

logical evaluation of the tumors. We thank William S. Reznikoff forhis help with this manuscript. We thank Karen Blomstrom for herpatience and skill in preparing the manuscript.

REFERENCES

1. Balmain, A. Transforming ras oncogenes and multistage carcinogenesis. Br.J. Cancer, 51: 1-7, 1985.

2. Barbacid, M. Oncogenes in human cancers and in chemically induced animaltumors. Prog. Med. Virol., 32: 86-100, 1985.

3. Barbacid. M. ras Genes. Annu. Rev. Biochem.. 56: 779-827, 1987.4. Shimon. D. J., deKernion, J. B., Verma, I. M., and Cline, M. J. Expression

of cellular oncogenes in human malignancies. Science (Wash. DC), 224:256-262, 1984.

5. Spandidos, D. A., and Agnantis, N. J. Human malignant tumours of thebreast, as compared to their respective normal tissue, have elevated expressionof the Harvey raj oncogene. Anticancer Res., 4: 269-272, 1984.

6. Spandidos, D. A., and Kerr, I. B. Elevated expression of the human rajoncogene family in premalignant and malignant tumours of the colorectum.Br. J. Cancer. 49: 681-688, 1984.

7. Yokota, J., Tsunetsugu-Yokota, Y., Battifora, H., LeFevre, C., and Cline,M. J. Alterations of myc, myb, and raj"' proto-oncogenes in cancers arefrequent and show clinical correlation. Science (Wash. DC), 231: 261-265,1986.

8. Egan, S. E., McClarty, G. A., Jarolim, L., Wright, J. A., Spiro, I., Hager,G., and Greenberg, A. H. Expression of H-raj correlates with metastaticpotential: evidence for direct regulation of the metastatic phenotype in 10T1/: and NIH 3T3 cells. Mol. Cell. Biol., 7: 830-837, 1987.

9. Goldfarb, M., Shimizu, K., Perucho, M., and Wigler, M. Isolation andpreliminary characterization of a human transforming gene from T24 bladdercarcinoma cells. Nature (Lond.), 296: 404-409, 1982.

10. Mancharan, T. H., Burgess, J. A., Ho, D., Newell, C. L., and Fahl, W. E.Integration of a mutant c-Ha-raj oncogene into C3H/10T'/2 cells and itsrelationship to tumorigenic transformation. Carcinogenesis (Lond.), 6:1295-1301, 1985.

11. Storer, R. D., Stein, R. B., Sina, J. F., DeLuca, J. G., Allen, H. L., andBradley. M. O. Malignant transformation of a preneoplastic hamster epidermal cell line by the EJ c-Ha-raj-oncogene. Cancer Res., 46: 1458-1464,1986.

12. Franza, B. R., Jr., Maruyama, K., Garrels, J. 1., and Ruley, H. E. In vitroestablishment is not a sufficient prerequisite for transformation by activatedraj oncogenes. Cell. 44: 409-418, 1986.

13. Ricketts, M. H., and Levinson, A. D. High-level expression of c-Ha-raj-1fails to fully transform rat-1 cells. Mol. Cell. Biol. S: 1460-1468, 1988.

14. Pulciani, S., Santos. E.. Long. L. K.. Sorrentino, V., and Barbacid, M. rajGene amplification and malignant transformation. Mol. Cell. Biol., 5: 2836-2841, 1985.

15. Redmond, S. M. S., Reichmann, E., Müller,R. G.. Friis, R. R.. Groner, B.,and Hynes, N. E. The transformation of primary and established mousemammary epithelial cells by p21-roj is concentration dependent. Oncogene,2:259-265,1988.

16. Reynolds, V. L., Lebovitz, R. M., Warren, S., Hawley, T. S., Godwin, A. K.,and Lieberman, M. W. Regulation of a metallothionein-ras T24 fusion geneby zinc results in graded alterations in cell morphology and growth. Oncogene, /: 323-330, 1987.

17. Sistonen, L., Keski-Oja, J., Dimanen, I., Hölttä,E., Wikgren, B-J., andAlitalo, K. Dose effects of transfected c-Ha-rajVmllzoncogene in transformedcell clones. Exp. Cell Res., 168: 518-530, 1987.

18. Winter. E., and Perucho, M. Oncogene amplification during tumorigenesis

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

of established rat fibroblasts reversibly transformed by activated human rajoncogenes. Mol. Cell. Biol., 6: 2562-2570, 1986.Spandidos, D. A., and Wilkie, N. M. Malignant transformation of earlypassage rodent cells by a single mutated human oncogene. Nature (Lond.),310: 469-475, 1984.Land, H., Parada, L. F., and Weinberg, R. A. Tumorigenic conversion ofprimary embryo fibroblasts requires at least two cooperating oncogenes.Nature (Lond.), 304: 596-602, 1983.Land, H., Chen, A. C., Morgenstern. J. P., Parada, L. F., and Weinberg, R.A. Behavior of myc and raj oncogenes in transformation of rat embryofibroblasts. Mol. Cell. Biol.. 6: 1917-1925, 1986.Newbold, R. F., and Overell, R. W. Fibroblast immortality is a prerequisitefor transformation by EJ c-Ha-raj oncogene. Nature (Lond.), 304:648-651,1983.Ruley, H. E. Adenovirus early region 1A enables viral and cellular transforming genes to transform primary cells in culture. Nature (Lond.). 304: 602-606, 1983.Storer, R. D., Allen, H. L., Kraynak, A. R., and Bradley, M. O. Rapidinduction of an experimental metastatic phenotype in first passage rat embryocells by cotransfection of EJ c-Ha-roj and c-myc oncogenes. Oncogene, 2:141-147, 1988.Oshimura, M., Koi, M., Ozawa, N., Sugawara, O.. Lamb, P. W., and Barrett,J. C. Role of chromosome loss in raj/myc-induced Syrian hamster tumors.Cancer Res., 48: 1623-1632, 1988.Chang, E. H., Furth, M. E., Scolnick, E. M., and Lowy, D. R. Tumorigenictransformation of mammalian cells induced by a normal human gene homologous to the oncogene of Harvey murine sarcoma virus. Nature (Lond.),297:479-483, 1982.Harris, C. C. Human tissues and cells in carcinogenesis research. CancerRes., 47: 1-10, 1987.Yoakum, G. H., Lechner, J. F., Gabrielson, E. W., Korba, B. E., Malan-Shibley, L., Willey, J. C, Valerio, M. G., Shamsuddin, A. M., Trump, B. F.,and Harris, C. C. Transformation of human bronchial epithelial cells transfected by Harvey raj oncogene. Science (Wash. DC), 227: 1174-1178, 1985.Rhim, J. S.. Jay, G., Arnstein, P., Price, F. M., Sanford. K. K., and Aaronson,S. A. Neoplastic transformation of human epidermal keratinocytes by AD 12-SV40 and Kirsten sarcoma viruses. Science (Wash. DC), 227: 1250-1252,1985.Walen, K. H., and Arnstein, P. Induction of tumorigenesis and chromosomalabnormalities in human amniocytes infected with simian virus 40 and Kirstensarcoma virus. In Vitro Cell. Dev. Biol., 22: 57-65, 1986.Der, C. J., Krontiris, T. G-, and Cooper, G. M. Transforming genes of humanbladder and lung carcinoma cell lines are homologous to the raj genes ofHarvey and Kirsten sarcoma viruses. Proc. Nati. Acad. Sci. USA, 79: 3637-3640, 1982.DiPaolo, J. A., Woodworth, C. D., Popescu, N. C., Notario, V., and Doniger,J. Induction of human cervical squamous cell carcinoma by sequential trans-fection with human papillomavirus 16 DNA and viral Harvey raj. Oncogene,4:395-399, 1989.Stevenson, M., and Volsky, D. J. Activated v-myc and v-raj oncogenes donot transform normal human lymphocytes. Mol. Cell. Biol. 6: 3410-3417,1986.Reddy, E. P., Reynolds, R. K., Santos, E., and Barbacid, M. A point mutationis responsible for the acquisition of transforming properties by the T24human bladder carcinoma oncogene. Nature (Lond.), 300: 149-152, 1982.Cohen, J. B., and Levinson, A. D. A point mutation in the last intronresponsible for increased expression and transforming activity of the c-Ha-roj oncogene. Nature (Lond.), 334: 119-129, 1988.Christian, B. J., Loretz, L. J., Oberley, T. D., and Reznikoff, C. A. Characterization of human uroepithelial cells immortalized in vitro by Simian Virus40. Cancer Res., 47: 6066-6073. 1987.Meisner, L. F., Wu, S-Q., Christian, B. J., and Reznikoff, C. A. Cytogeneticinstability with balanced chromosome changes in an SV40 transformedhuman uroepithelial cell line. Cancer Res., 48: 3215-3220, 1988.Reznikoff, C. A., Loretz, L. J., Christian, B. J., Wu, S-Q., and Meisner, L.F. Neoplastic transformation of SV40-immortalized human urinary tractepithelial cells by in vitro exposure to 3-methylcholanthrene. Carcinogenesis(Lond.), 9: 1427-1436, 1988.Bookland, E. A., Swaminathan, S., and Reznikoff, C. A. Neoplastic transformation of SV-40 immortalized human uroepithelial cells by treatmentwith the human bladder carcinogen (ABP) and its metabolite, yv-hydroxy-4-acetylaminobiphenyl (HAABP). Proc. Am. Assoc. Cancer Res., 30: 612,1989.Reznikoff, C. A., Loretz, L. J., Pesciotta, D. M., Oberley. T. D., andIgnjatovic, M. M. Growth kinetics and differentiation in vitro of normalhuman uroepithelial cells on collagen gel substrates in defined medium. J.Cell. Physiol., 131: 285-301, 1987.Southern, P. J., and Berg, P. Transformation of mammalian cells to antibioticresistance with a bacterial gene under control of the SV40 early regionpromoter. J. Mol. Appi. Genet.. /: 327-341, 1982.Trimble, W. S., Johnson, P. W., Hozumi, N., and Roder, J. C. Induciblecellular transformation by a metallothionein-raj hybrid oncogene leads tonatural killer cell susceptibility. Nature (Lond.), 321: 782-784, 1986.Graham, F. L., and Van Der Eb, A. J. A new technique for the assay ofinfectivity of human adenovirus 5 DNA. Virology, 52: 456-467, 1973.Reznikoff, C. A., Gilchrist, K. W., Norback, D. H., Cummings, K. B., Ertiirk,E., and Bryan, G. T. Altered growth patterns in vitro of human papillarytransitional carcinoma cells. Am. J. Pathol., Ill: 263-272, 1983.

4785

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

HUMAN UROEPITHELIAL CELL TRANSFORMATION BY EJ/ros1"

45. Srivastava, S. K., Yuasa, Y.. Reynolds, S. H., and Aaronson, S. A. Effects of Genetic tagging of tumor cells with retrovirus vectors: donai analysis oftwo major activating lesions on the structure and conformation of human ras tumor growth and metastasis in vivo. Mol. Cell. Biol.. 8: 3143-3149. 1988.oncogene products. Proc. Nati. Acad. Sci. USA, «2:38-42, 1985. 50- Mitclman. F. Catalog of Chromosome Aberrations in Cancer, Ed. 3. New

46. Davis, L. G., Dinner. M. D.. and Battey, J. F. Basic Methods in Molecular York: Alan R- Liss- l988-Biology, pp. 130-135, New York: Elsevier. 1986. 5 ' •nl£'"',9„T1h<e,aQPP1I?«"Toll™ °f"^ 'Um°rsuppressor genes'Science (Wash'

47. Wu, S-Q.. Christian, B. J., Reznikoff. C. A., and Meisner, L. F. Marker _, ÌT!' „ .., j t „«.... , . . . . , . ,., 52. Craig. R. \V., and Sager. R. Suppression of tumorigemcily in hvbnds ofchromosome stab.luy associated with neoplasty transformauon of human )£*, ^ Oncogene.?ransforme5PCHEF cells. Proc. Nati. Acad. Sci. USA,

uroepithelial cells. Cancer Genet. Cytogenet.. 36: 77-87. 1988. g¿.2062-2066 198548. Babu. V. R., Miles, B. J., Cerney, J. C., Weiss, L.. and Van Dyke. D. L. 5, Geiser A G Dcr c; j Marshall, C. J.. and Stanbridge, E. J. Suppression

Chromosome 21q22 deletion: a specific chromosome change in a new bladder of tumorigenicity with continued expression of the c-Ha-ros oncogene in EJcancer subgroup. Cancer Genet. Cytogenet., 38: 127-129, 1989. bladder carcinoma-human fibroblast hybrid cells. Proc. Nati. Acad. Sci. USA,

49. Korczak. B., Robson. I. B., Lamarche. J. C.. Bernstein. A., and Kerbel, R. S. S3: 5209-5213. 1986.

4786

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1990;50:4779-4786. Cancer Res   Brian J. Christian, Chinghai Kao, Shi-Qi Wu, et al.   40-immortalized Human Uroepithelial Cells: A Rare Event

Neoplastic Transformation of Simian VirusrasEJ/

  Updated version

  http://cancerres.aacrjournals.org/content/50/15/4779

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/50/15/4779To request permission to re-use all or part of this article, use this link

on August 5, 2020. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from