Role of Chromosome Loss in ras/myc-induceÀSyrian Hamster ... · Role of Chromosome Loss in ras/myc-induceÀSyrian Hamster Tumors Mitsuo Oshimura,1 Minoru Koi,1 Nobuyoshi Ozawa,

[CANCER RESEARCH 48, 1623-1632, March 15, 1988]

Role of Chromosome Loss in ras/myc-induceàSyrian Hamster TumorsMitsuo Oshimura,1 Minoru Koi,1 Nobuyoshi Ozawa, Osamu Sugawara, Patricia W. Lamb, and J. Carl Barrett2

Laboratory of Molecular Carcinogenesis, National Institute of Environmental Health Sciences, NIH, P.O. Box 12233, Research Triangle Park, North Carolina 27709

ABSTRACT

It has been shown previously that normal Syrian hamster embryo cellsare neoplastically transformed by transfection with two cooperating on-cogenes, \--myc plus v-Ha-nw. Karyotypic analyses of the cells from the

tumors revealed a nonrandom chromosome change, monosomy of chromosome 15. In order to clarify the role of chromosome loss in thesetumor cells with defined oncogene alterations, molecular and cytogeneticstudies were performed on hybrids between normal Syrian hamsterembryo cells and ras/myc tumor cells. Following fusion of the tumor cellswith the normal cells which are not immortal, the majority of the cellhybrids senesced after <20 population doublings indicating that immortality was recessive. Some of the hybrids escaped senescence and grewindefinitely. These immortal hybrid cells retained the expected numbersof chromosome 15 indicating that escape from senescence did not involveloss of this chromosome. The tumorigenicity and anchorage-independentgrowth of the nonsenescent hybrids were still suppressed significantly.In these suppressed hybrid cells, RNAs complementary to the v-IIa-ravand \-myc oncogenes were expressed. Furthermore, radioimmune precipitation with a monoclonal antibody to p21"" of |35S]methionine-labeledcell extracts followed by polyacrylamide gel electrophoresis/sodium do-decyl sulfate electrophoresis showed that the suppressed hybrid cellscontained high levels of the mutated ras protein. These results indicatethat tumorigenicity is suppressed in the hybrids even though the oncogenes are expressed. When the hybrid cells were passaged, anchorage-independent variants appeared in the cultures. At this time, morphologicalchanges occurred in the cultures and the cells were tumorigenic. Karyotypic analyses of the transformed segregants versus the parental hybridcells revealed a nonrandom loss of one copy of chromosome 15 in thetransformed segregants. No other nonrandom chromosome change wasobserved. These results suggest that the loss of chromosome 15 resultsin the loss of a cellular tumor suppressor gene which effects a phenotypicchange necessary for expression of neoplastic transformation. In addition,the cellular factors responsible for the senescence of the hybrids mayprovide another mechanism involved in suppressing tumorigenicity.

INTRODUCTION

Nonrandom chromosome changes, including structural andnumerical changes, have been observed in a variety of humanand animal tumors (1-4). Some of the nonrandom changes,particularly translocations, are associated with enhancedexpression and/or deregulation of cellular oncogenes (1-5).Little is understood, however, about the significance of numerical chromosome changes in tumor cells, even though there isconsiderable evidence showing that specific chromosome lossesand/or gains occur nonrandomly in certain tumors (3, 4).Recently, we reported (6) that normal SHE3 cells neoplastically

transformed following transfection with two cooperating oncogenes, v-myc plus v-Ha-ros, were monoclonal, and had anonrandom chromosome change, monosomy of chromosome15. The finding of a nonrandom chromosome loss indicates

Received 6/9/87; revised 11/23/87; accepted 12/10/87.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.

' Present address: Laboratory of Cell Biology, Kanagawa Cancer Center Research Institute, Nakao-chyo 54-2, Yokohama, Japan.

2To whom requests for reprints should be addressed.3The abbreviations used are: SHE, Syrian hamster embryo; HAT, complete

medium with 10~4 M hypoxanthine, 10~5 M aminopterin, and 1.6 x 10~5 Mthymidine; PEG, polyethylene glycol; CFE, colony forming efficiency; SA*, soft

agar positive.

that this alteration plays a role in the carcinogenic process butthe nature and significance of this role are undefined. Since weobserved loss of chromosome 15 in cells derived from thetumors, it is possible that this change was a late event occurringafter the cells had acquired tumorigenicity; tumor cells withthis chromosome loss may have a selective advantage in thetumor population. Alternatively, this chromosome may carry aspecific cellular gene which suppresses the expression of tumorigenicity and has to be lost for neoplastic transformation.The existence of tumor suppressor genes (or antioncogenes) issupported by studies of hybrids between normal and tumor cells(7-10), studies of genetic predisposition to tumors in humans(11-13), and studies of revenants of tumor cells (14, 15).

In order to clarify the role of monosomy 15 in the ras/myctumors, molecular and cytogenetic studies of hybrid cells between the ras/myc tumor cells and normal SHE cells wereconducted. Our results show that the hybrids are nontumori-genic and the loss of chromosome 15 is a necessary event forthe expression of tumorigenicity. Furthermore, analysis of theexpression of the v-Ha-ras and v-myc oncogenes in the nontu-morigenic hybrids reveals that normal cells can suppress thetumorigenicity of the ras/myc tumor cells even though the viraloncogenes are expressed. Thus, the mechanism of tumorsuppression does not appear to involve transcriptional regulation of these oncogenes. These results support the concept thattumor suppressor genes exist in normal cells and control theexpression of tumorigenicity even when multiple oncogenes areactivated in a cell. In addition, the expression of immortalitywas recessive in some hybrids. Escape from senescence of thehybrids did not result from loss of chromosome 15. Thus,another possible mechanism for suppression of carcinogenesisinvolving senescence is indicated by our studies and others (16-19).

MATERIALS AND METHODS

Cells,'Growth Medium, and Cell Hybridization. The cells used for the

fusion studies were normal SHE cells and a tumor-derived cell line(ras/myc^) induced by transfection of SHE cells with the viral v-Ha-ras and \-myc oncogenes (6). Mutants of the tumor cells (designatedras/myc-Tm) were selected for resistance to both 6-thioguanine andouabain (16) to allow selection of cell hybrids. Methods for cell culturehave been described (6, 16). For cell hybridization, IO6cells of each celltype were plated into 75-cm2 flasks and fused 24 h later by treatmentat 37°Cfor 1 min with 5 ml of 41.7% (W/V) PEG 1000 (Baker

Chemical Co.) containing 15% dimethyl sulfoxide, followed by treatment for 2 min with an additional S ml of 25% PEG without dimethylsulfoxide. The cells were washed extensively, grown for 24 h in growthmedium, and then plated (IO4 cells per 100-mm dish) into selective

medium, i.e., HAT plus 1.5 HIMouabain. In some experiments hybridcells were selected for growth in agar 24 h after fusion. In theseexperiments IO5cells/60-mm dishes were plated in agar with selective

growth medium (i.e., HAT plus ouabain). Hybrid colonies formed after3-4 wk on plastic at a frequency of 10~3 to 10~" hybrid colonies perfused cell, whereas no colonies (frequency <10"6) were observed if only

one parental cell type was treated and selected in HAT/ouabain medium.

Hybrid clones were isolated and grown to a sufficient cell number toanalyze. The hybrid cells were tested for anchorage-independent growthin soft agar as described previously (16). Colonies of >50 cells were

1623

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN rus/mj'oINDUCED TUMORS

scored after 3 wk. Cells (106/site) were tested for tumorigenicity by s.c.

injection into BALB/c-n«/nu mice (Frederick Cancer Research Center).Animals were observed for 30-60 days and the latency periods fortumor formation were determined from the time required for palpabletumors to appear. All tumors grew progressively.

DNA, RNA, and Protein Analysis. RNA dot-blot analysis and DNASouthern analysis of normal SHE cells, ras/myc-T cells, SHE x ras/m.v<-Tin hybrid cells and soft-agar variants isolated from the hybrid

clones were performed as previously described (20) by hybridization tothe 32P-labeIed (1-2 x IO8cpm/Vg) v-Ha-ros insert (450-base pair) fromclone BS9 or \-myc insert (1.5 kilobase) from pSVv-myc. For analysisat the protein level of p21"", cells at <80% confluence in 100-mmdishes were labeled for 4 h with [35S]methionine (1 mCi/dish). Immu-noprecipitation (>1 x IO8cpm) with Y13-259 anti-p21"" monoclonalantibody and sodium dodecylsulfate-polyacrylamide gel electrophoresiswas performed according to the procedures of others (21, 22). Theantibody was kindly provided by Dr. Steve Reynolds of NationalInstitute of Environmental Health Sciences.

Chromosomal Analysis. For conventional chromosome preparations,cultures were treated for 3 h with colcemid (0.2 ¿tg/mlfinal concentration) before the cells were disassociated with phosphate-buffered salinecontaining 0.05% trypsin and 0.02% EDTA. The cells were thencentrifuged (1200 rpm, 5 min) and «suspended in 0.075 M KC1 (20°C,

15 min). Following further centrifugation, the cells were fixed byresuspending in methanohglacial acetic acid (3:1 v/v). Centrifugationand resuspension in fresh fixative were repeated at least 2 times beforethe cells were placed onto clean microscopic slides, air-dried, andstained. Ten banded spreads/clone were photographed and karyotyped,using the Q-banding technique as described previously (6). The analysesof the parental hybrid cells were made at the same time the cells weretested for anchorage-independent growth (generally at passage 3). Thehybrid colonies were continually subcultured until colonies growing inagar were detected. These colonies were then isolated with a sterilepipet placed into a 60-mm Petri dish, and the chromosome analyses ofthe soft-agar clones were made at the first passage after isolation of thecolonies from age.

In Situ Chromosome Preparation. In order to examine the chromosomal constitution of different hybrid colonies at 1-2 wk after selectionin HAT/ouabain medium when the number of cells per colony wasinsufficient to allow conventional chromosome preparations, we developed a new method which permits karyotypic analysis of cells withinan intact colony. Dishes containing colonies of ~ 1000 cells were treatedwith colcemid (0.2 Mg/ml) for 3 h and then the medium was gentlyaspirated from the edge of each dish. Five ml of 0.05 M KC1 hypotoniesolution were carefully added down the inside edge of each dish, and30 min later an equal volume of freshly prepared 3:1 methanokaceticacid (v/v) fixative was added and allowed to stand for 2 min. Afterremoval of half of the fluid from the edge of the dishes, the fixative wasagain added; this process was repeated several times. At least 20 minlater, the fixative was aspirated completely and the dishes were driedquickly in humid air over a 50°Cwater bath. The dishes were then

rinsed with 70% ethanol and were stained with Giemsa solution 1 dayafter fixation. Mètaphase chromosome spreads were analyzed in colonies in which 5 or more metaphases could be analyzed.

RESULTSCell Hybrids between Normal SHE Cells and ras/myc-'ïCells.

We previously reported that following transfection with the v-Ha-ros and v-myc oncogenes, SHE cells were able to formtumors in nude mice (6, 20). The cells derived from thesetumors (designated ras/myc-1 cells) have a nonrandom loss ofchromosome 15 (6). In order to understand the role of thischromosome loss in tumorigenicity, cell hybrids were isolatedfollowing fusion of SHE and ras/myc-1 cells with PEG. Forselection of hybrids, a mutant cell line (ras/myc-lm), whichwas resistant to both 6-thioguanine and ouabain, was isolated.This cell line was highly tumorigenic with a latency period of5-8 days following injection of IO6 cells into nude mice, grew

in soft-agar with a high efficiency (>80%), had a near-diploid

karyotype with monosomy of chromosome 15 (described indetail below), and expressed both the v-Ha-ros and v-myc genes.

Hybrids between SHE and ras/myc-1 cells were formed bytreatment with PEG and selected in HAT/ouabain medium bystandard methods (16). As an initial test of whether the SHEcells could suppress the neoplastic phenotype of the ras/myc-1cells, the ability of the hybrids to grow in soft agar was assayed24 h after fusion (16). The number of hybrid cells which formedon plastic dishes in HAT/ouabain medium was compared tothe number of colonies growing in agar containing the sameselection media. As a control experiment, ras/myc-1m cellswere fused with parental ras/myc-1 cells. A similar number(ratio, 0.89) of ras/myc-1 x ras/myc-1m hybrids was observedwhen the selection was performed on plastic dishes or in agar(Table 1). In contrast, SHE x ras/myc-1m hybrids were detected when the selection was performed on plastic dishes, butno hybrids were detected when the selection was performed inagar (Table 1). This finding suggests that SHE x ras/myc-1hybrids were not anchorage independent, a phenotype closelycorrelated with tumorigenicity of Syrian hamster cells (16).

In order to confirm this finding and to test the tumorigenicityof the hybrids, colonies growing in HAT/ouabain medium onplastic dishes were isolated by cloning cylinders and subculturedinto 35-mm dishes. The colonies selected for isolation werehealthy colonies consisting of >1000 cells. All of the clonesgrew after subculturing to a certain degree, but cells in thecultures began to show signs of senescence as evidenced by cellenlargement and cessation of growth after 4-6 wk. Approximately 50% (15 of 27) of the colonies could be subcultured onlyonce or twice into 60-mm dishes and then growth totally ceased.These colonies were estimated to have undergone <20 population doublings. When colonies of normal SHE cells were isolated, the cells also senesced after <20 population doublings.Twelve SHE x ras/myc-1m hybrid colonies could be subcultured further. The growth in these cultures was slow at first butincreased with subculturing, suggesting that the colonies escaped senescence. The colonies were grown for >10 passages(>50 population doublings) and are therefore considered immortal. In contrast to the SHE x ras/myc-1m hybrids, all ras/myc-1 x ras/myc-1m hybrid colonies were readily isolated andcould be subcultured indefinitely without evidence of a crisisperiod (Table 2).

The chromosomal constitutions of SHE x ras/myc-1m hybrid colonies were examined. In order to compare the senescentversus immortal hybrids, an in situ method for chromosomeanalysis was developed. Intact colonies of cells growing for only2 wk were treated with colcemid for 3 h and then the cells were

Table 1 Direct assay of suppression/expression of anchorage-independentphenotype in hybrids of SHE x ras/myc-Tm cells or ras/myc-T x ras/myc-Tm

cellsCell fusions were performed with the ras/myc-Tm cell line which was resistant

to thioguanine and ouabain. These cells were fused by PEG treatment as describedin "Materials and Methods" with either SHE cells or the parental ras/myc-T cells

that were not thioguanine or ouabain resistant. These latter two cell types grewin HAT but failed to grow in ouabain containing medium. Hence, only the hybridcolonies survived and grew in HAT/ouabain medium. Hybrid frequency wasdetermined from the number of colonies growing on plastic dishes in HAT/ouabain at 2 wk after selection. Frequency of hybrids growing in agar (SA*hybrids) was determined from the number of colonies (>50 cells) which formedin agar containing selective (HAT/ouabain) medium at 3 wk after selection.

Cell linefused to

ras/myc-TmcellsSHEras/myc-TFrequency

ofhybrids on

plastic(x 10s)22

28Frequency

ofhybrids growingin agar (x10s)<0.02

25Ratio

of SA*

hybrids tototalhybrids<0.001

0.89

1624

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN ros/m>'c-INDUCED TUMORS

Table 2 Senescence or indefinite growth of hybrid coloniesHealthy colonies of >1000 cells were selected at 2-3 wk after selection in

HAT/ouabain, isolated by cloning cylinders, and trypsinized. The colonies weresubcultured into 35-mm dishes with HAT/ouabain and grown to confluence.Cells in all dishes grew to confluence or near confluence and were subculturedinto 60-mm dishes. The cells grew in these dishes but some of the cultures ofSHE x ras/myc-Tm hybrids began to senesce. These cultures (senescent colonies)failed to grow to confluence and failed to grow if subcultured. Some of thesehybrids grew slowly when subcultured but continued to grow. As these cells weresubcultured additional times, the growth rate of the cells increased. When thecolonies could be subcultured >10 times they were termed immortal. All of theras/myc-T x ras/myc-Tm hybrids grew rapidly at every passage and could besubcultured indefinitely.

SHE x ras/myc-Tm hybridsTotal colonies examined 27Senescent colonies 15 (56%)Immortal colonies 12(44%)

ras/myc-T x ras/myc-Tm hybridsTotal colonies examined 26Senescent colonies 0Immortal colonies 26(100%)

Table 3 Cellular composition of hybrids based on chromosome analysis followingin situ chromosome analysis

Hybrid colonies growing in HAT/ouabain were treated with colcemid, fixed,and carefully treated with a hypotonie solution as described in "Materials andMethods." The cells were stained with the quinacrine dye and analysis of meta-

phase chromosome spreads of the cells in the in situ colonies was performed. Thecellular composition of the colonies was based on analysis of 5-10 metaphasesper colony for chromosome number, number of marker chromosomes, andnumber of chromosome 15.

Total hybrids (2wk)Total

hybrids (5-6 wk)Cellular

compositionITxIN"2TxIN*ITx2NrITx

IN2TxINITx

2NNo.

examined16408130%of total

colonies8020038620

" Hybrids formed by 1 tumor cell and 1 normal cell.* Hybrids formed by 2 tumor cells and 1 normal cell.' Hybrids formed by 1 tumor cell and 2 normal cells.

fixed directly on the culture dish and treated with hypotoniesolution. The quality of the metaphase chromosome spreads ofcells in these colonies was adequate to allow analysis of thetotal numbers of chromosomes per metaphase, the numbers ofmarker chromosomes per cell, and the numbers of chromosome15 per cell. Analysis of 5-10 cells from each colony alloweddetermination of whether the hybrid colonies were composed

of 1 tumor cell fused with 1 normal cell ( 1Tx 1N), 2 tumor cellsfused with 1 normal cell (2TxlN), or 1 tumor cell fused with 2normal cells (lTx2N). The relative percentages of hybrids withthese different cellular compositions were compared after 2 wkwhen the total number of hybrids included hybrids that eventually senesced and after 4-6 wk when only immortal hybridspersisted. Initially 80% of the hybrids were ITx IN and 20%were 2TxlN (Table 3); colonies of lTx2N hybrids were notobserved. The proportion of hybrids with a 2TxlN cellularcomposition significantly increased (to 62%) in the immortalhybrid colonies, suggesting that immortality may be related tocell, and presumably gene, dosage effects.

The hybrids of SHE x ras/myc-Tm cells which did notsenesce could be grown to a sufficient number of cells for furtheranalysis of tumorigenicity and anchorage-independent growth.Cytogenetic and molecular changes associated with the expression or suppression of neoplastic transformation were alsoexamined. At the earliest passages (p3 or p4), all of the hybridcolonies examined were nontumorigenic and suppressed foranchorage-independent growth (Table 4). The SHE x ras/myc-Tm hybrid colonies grew well on plastic (15-30% CFE), buttheir ability to grow in agar was reduced by 100- to 1000-fold(i.e., the ratio of anchorage-independent colonies to anchorage-dependent colonies was 0.013 to 0.00013). In contrast ras/myc-T x ras/myc-Tm hybrids grew nearly equally well on plasticand in agar and all formed tumors with latency periods of only5-8 days similar to the parental ras/myc-T cells (Table 4).

When the SHE X ras/myc-Tm hybrid cells were grown further, the ability of the hybrids to grow in agar increased,suggesting that transformed variants arose in the populations.These transformed variants quickly overgrew the slower growing, nontransformed cells as evidenced by a rapid increase inthe number of anchorage-independent cells in the population

and the morphological appearances of the cells (Fig. 1). Multiple anchorage-independent clones were isolated from thesecultures and tested for tumorigenicity. All were highly tumori-genie with short latency periods (<8 days) and grew in agarwith CFEs >50%. Comparisons were then made of the kary-otypes and the levels of expression of v-Ha-ras and \-myconcogenes in the parental cells (SHE and ras/myc-1), thesuppressed hybrids (SHE x ras/myc-Tm), and the anchorage-

Table 4 Properties of SHE cells, ras/myc-Tm cells, and their cell hybridsHybrid colonies were isolated and subcultured for 3-4 passages. They were then analyzed for growth in agar. tumorigenicity in nude mice, and chromosome

number. Nude mice injected with 10*cells per site were examined for up to 60 days.

ParentalcellsSHEras/myc-TmSHE

X ras/myc-TmHybridsclone2-2clone2-3clone2-5clone

4-Mras/myc-T

x ras/myc-TmHybridsclone4clone9clone

11clone13Modal

no. ofchromosomes4443-458478115118/848685NDNDCFE

onplastic(%)10-2080-902030301525263035CFE

inagar(%)0.00180-900.0030.050.40.0028.213.81615Ratio

of anchorage-independent growth

to anchorage-dependentgrowth<0.00011.00.000150.00170.0130.000130.330.530.530.43Tumorigenicity

(latent period indays)No

tumors5-8No

tumorsNotumorsNo

tumorsND°6-86-86-86-8

* ND, not determined.

1625

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN ro.!/m>T-INDUCED TUMORS

Fig. I. Morphologies of normal SHE cell (A), ras/myc-Tm cells (A), hybridcells of SHE x ras/myc-Tm at an early passage when the cells were nontumori-genic (C), and hybrid cells of SHE x ras/myc-Tm at a later passage when thecells were tumorigenic (D).

independent variants of the hybrids (SHE x ras/myc-Tm SA+)

which reexpressed tumorigenicity.Oncogene Expression in Parental Cells, Suppressed Hybrids,

and Anchorage-independent Variants. Southern blot analysis ofcellular DNAs showed no differences in the restriction enzymedigest patterns or amplification of the v-Ha-ras or \-myc sequences in ras/myc-T cells, SHE x ras/myc-T hybrids, or thesoft-agar variant clones from these hybrids (Fig. 2).

Immunoprecipitation of the p2T" protein of [35S]methio-

nine-labeled cells with anti-p2raj monoclonal antibody Y13-

259 and polyacrylamide gel electrophoresis analysis allowedcomparisons of the ras protein in the various cell types. Thisantibody recognized both the cellular p21"" in the normal SHEcells and the mutated p2lras in the ras/myc-T cells. These two

ras proteins could be distinguished on the basis of the slowermobility of the viral p21"" (21,22) and the increased expression

(>10 fold) of the viral ras protein in the tumor cells relative tothe normal cells (Fig. 3). In the nontumorigenic SHE x ras/myc-T hybrids, high levels of expression of the mutated p21""

protein were observed and no differences in this expressionwere observed in the variants of the hybrids which were tumorigenic (Fig. 3 and Table 5). The expression of the p2r°*proteinwas not elevated significantly in the soft-agar, tumorigenicvariants of the hybrids with the possible exception of hybrid 2-2 where the p2rfls level was elevated 2-fold in the soft-agarvariant. This clone also had elevated expression of v-Ha-r<Z5

RN A. Two of three other soft agar clones from this hybrid alsoexpressed 2- to 3-fold more v-Ha-ros RNA. Other SHE x ras/myc-1 hybrids had comparable expression of v-Ha-ras RNA inthe parental hybrids suppressed for tumorigenicity and soft-agar clones derived from these hybrids (Table 5). When the v-myc RNA levels were measured, a decreased expression in thehybrids was observed relative to the ras/myc-T parental cells.The reason for this remains unclear. However, when the soft-agar variants were compared to the suppressed hybrids, v-mycRNA levels were comparable in both cell types indicating thatv-myc RNA expression did not correlate with expression oftumorigenicity.

Karotype Analysis of Parental Cells, Nontumorigenic Hybrid,and Tumorigenic Variants of the Hybrids. All of the cells fromthe original ras/myc-1 tumor (tumor l in Ref. 6) were of XXorigin and had two chromosomal abnormalities common to allof the cells: monosomy 15 and a translocation (X1) between an

X chromosome (at the break point Xpa7) and an unidentifiedchromosome. Detailed banding analysis revealed that this unidentified chromosome segment is quite distinct from chromosome 15. Half of the cells had two X1chromosomes; the modalkaryotype of ras/myc-1 was 44,XX'X',-15. The double mutantras/myc-Tm derivative (thioguanine-resistant, ouabainR mutant) used for the cell fusion retained 2 X' markers and mono-

B

Fig. 2. A. Southern blot of cellular DNAsdigested with Pst l and probed with "P-labeledl.S-kilobase v-myc insert from clone pSVv-myc. The l.S-kilobase fragment on which v-myc gene is located in pSVv-myc is observedin all cells and no differences are detected inthe amount of v-myc DNA among the parentalras/myc-Tm cells, SHE x ras/myc-Tm hybrids, and soft-agar variant clones from thehybrids. B, Southern blot of cellular DNAsdigested with Bglil and probed with 32P-la-beled 0.45-kilobase v-Ha-ros insert from cloneBS9. The v-Ha-ras coding sequences are located within the 2-kilobase Ag/II fragment ofthe 111 plasmid used to transform ras/myc-Tm cells. This fragment is present in all cellsand no difference in the amounts of v-Ha-rosDNA among the parental tumor cells (ra.s•myc-Tm), SHE x ras/myc-Tm hybrids, andsoft-agar clones from the hybrid cells is observed.

7 " "N W CU

lì f I I

1626

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN ras/myc-INDUCED TUMORS

Table 5 Relative expression ofp2inu protein, v-ras RNA, and v-myc RNA inparental and hybrid cells

M

SHE

ras/myc-T

ras/myc-T

Hybrid 2-2

Hybrid 2-2 SA

Hybrid 2-3

Hybrid 2-3 SA*

Hybrid 2-4

Hybrid 2-4 SA

Hybrid 2-5

Hybrid 2-5 SA

Fig. 3. [3*S|Methionine-labeled cellular extracts from normal SHE cells (lane1), ras/myc-Tm cells (lanes 2 and 3), SHE x ras/myc-T hybrid clones (lanes 4, 6,8, 10), and soft agar (.VII variants (lanes 5, 7, 9, 11) isolated from the hybrid cellswere immunoprecipitated with ¡miip21"" monoclonal antibody Y13-259. The

immunoprecipitates were then run on sodium dodecyl sulfate/acrylamide gels,transferred to nitrocellulose paper, and autoradiographed (21, 22). Two bandsare observed corresponding to the precursor (slower band) and mature forms ofp21""(21, 22). The v-Ha-ros p21 protein in the ras/myc-Tm cells is overexpressedrelative to the normal p21"" in the SHE cells and also migrates more slowly (21,22). The precursor pii"" band in the ras/myc-Tm cells moves slowest. Themutant p21 ~ mature form and the normal p21 "" precursor move at approximately

the same location.

somy 15 and had acquired additional chromosome changes,i.e., loss of chromosomes 6, 16, and 17, and 4 marker chromosomes (markers A, C, D, and E; the X' chromosome marker

chromosome is chromosome B). The marker chromosome Awas identified as a translocation of an unknown segment ontoa chromosome 17. The C marker was identified as a 6q~. The

D marker, a small met acentric chromosome of unknown origin,was not identified. The E marker was identified as an isochro-mosome of chromosome 16. Although a majority of cells hadthese marker chromosomes (including 2 copies of the chromosome marker B), a Ip anomaly was also observed in a minorcell population. Thus, the modal karyotype of the ras/myc-Tmderivative used for the cell fusion was 45,XX'X',-15,6q-,i(16q),17q*,mar (Fig. 4; Table 6).

A total of 12 hybrid clones from 3 different hybridizationexperiments was isolated 2-3 wk after hybridization. One cul

ture consisting of multiple hybrid clones was also studied. Basedon chromosome analysis including chromosome number andthe number of marker chromosomes, five of the 12 hybridclones were near-tetraploid (ITxlN) hybrids. The remaining 7isolated clones were derived from fusions of two tumor cellswith one normal cell (2TxlN). The uncloned culture of hybridclones was a mixture of the two types of hybrids. Chromosomeanalysis was made at the same time as tumorigenicity, plating

Cellsras/myc-TmSHEras/myc-Tm

x SHEhybridsHybrid2-2Hybrid

2-2SAIHybrid2-2SA2Hybrid2-2SA3Hybrid2-2SA4Hybrid

2-3Hybrid2-3SAHybrid

2-5Hybrid2-5SAHybrid

2-6Hybrid2-6SAHybrid4-Mp21~1.0"0.08C0.841.65ND*NDND1.281.331.090.920.931.09NDv-ras

RNA1.0*<0.010.310.800.340.950.980.360.370.991.51.91.11.8v-mycRNAl.O4<0.010.070.110.050.100.130.060.060.180.150.180.410.36

* The ("Sjmethionine-labeled cellular extracts from ras/myc-Tm cells, SHEcells, SHE X ras/myc-Tm hybrid clones, and soft agar variants isolated from thehybrid cells were immunoprecipitated with anti-p21 "" monoclonal antibody Y13-259. The immunoprecipitates were then run on sodium dodecyl sulfate/acrylam-ide gels, transferred to nitrocellulose paper, and autoradiographed (21, 22). Thegels were scanned by a densitometer and the intensity of the two bands of p21'â„¢was determined (Fig. 4). Parental ras/myc-Tm cells were run on each gel and theintensity of p21"~ was determined for the hybrid cells and expressed relative tothe intensity of p2lrmMin the ras/myc-Tm cells.

*Total cytoplasmic RNA was isolated from the cells, diluted onto nitrocellulosedot blots, and hybridized with either "P-labeled \-Ha-ras insert from BS9 plasmidof 32P-labeled or v-myc insert from pSV v-myc plasmid. The relative intensity ofthe v-Ha-ros or v-myc RNA in each cell type was determined by densitometermeasurements of the autoradiograms of the dot blots. No detectable hybridizationoccurred with RNA from SHE cells under the high stringency hybridizationconditions used.

f The p21â„¢protein in the SHE cells was immunoprecipitated by the Y13-259antibody but the mobility of the normal P21â„¢1protein was higher than themutated, viral p21"" protein in the ras/myc-Tm cells. The ratio given is therelative amounts of either form of p21r" in the two cell types.

* ND, not determined.

efficiency on plastic, and growth in soft agar were examined.The 5 near-tetraploid (ITxlN) hybrids and their anchorage-independent variants were analyzed karyotypically in detail.

The anchorage-independentranchorage-dependent growth ratio of the 2-2 clone was 0.000IS and the median chromosomenumber (Fig. 5) was 84 (the expected chromosome number ofa SHE x ras/myc-T'm hybrid is 89). Ten clones were isolatedfrom agar (SA+). The CFEs in agar of these clones were >50%

and their median chromosome numbers varied from 71 to 80(Table 7). Only loss of one copy of chromosome 15 was common in all of the metaphases from the 10 SA ' clones (Figs. 5

and 6). The 2-3 clone had a anchorage-independent:anchorage-dependent growth ratio of 0.017. The parental hybrid clone hada median chromosome number of 78 (Table 8). Whereas 100%of the cells examined contained 3 copies of chromosome 15, allten variants isolated from agar had lost one copy of chromosome 15. None of the SA* clones had a common loss of any

other chromosomes when compared with the parental clone.Loss of chromosomes other than No. 15 was frequent in SA*

clones of specific hybrids (for example, chromosome X inhybrid 2-2 and chromosome 4 in hybrid 2-3). However, thesechromosome losses were not observed in all of the SA* variants

and were not common between independent hybrids.Two hybrid clones (3-5 and 3-6) had relatively high plating

efficiencies (0.4% and 4.8%) in agar at the time of analysis. Amajority of the cells (80-90%) of these hybrids had already lostone copy of chromosome 15 at this passage. All ten SA* clones

derived from these hybrids also had lost one copy of chromosome 15. This indicates that the SA* variants in the parental

cell population represent the majority of the cells in metaphase.Banded karyotypic analysis was performed on 10 metaphases

for each of 10 SA* clones isolated from these four hybrids. All

1627

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN ras/myc-ÃŒNDUCEDTUMORS

1 2 3 4

I U U H »6 7 8 9 10

U ii ti u i11 12 13 14 15

l t ftl till»«19 20 2

Fig. 4. Modal karyotype (45,XX'X',-15,6q-,i(16q),17q-H,+mar) of pseudodi-ploid ras/myc-Tm parental derivative. The marker chromosomes were identifiedas follows: A, 17q*; B, X'; C, 6q~; 0, unidentified marker chromosome; £,i(16q).

See Table 5.

10 metaphases for a given SA+ clone had the same number of

copies of chromosome 15 with the exception of two metaphasesin one of 10 SA+ clones (i.e., 2 of 10 metaphases of SA-7 ofhybrid 3-6 had 3 rather than 2 copies of chromosome 15). Thus,the two metaphases in SA-7 of hybrid 3-6 were the only cellsthat had not lost chromosome 15 in 400 metaphases studiedfrom 40 SA+ subclones of the 4 hybrid clones. This clearly

indicates a nonrandom loss of this chromosome in the transformed variants of these four hybrids.

A different result was observed with the hybrid clone 3-3.This hybrid had plating efficiencies of 15% in agar and 35% onplastic at the passage studied. All of the metaphases of theparental hybrid clone had 3 copies of chromosome 15. Incontrast to the 4 hybrids described above, none of the 80metaphases from SA+ clones of hybrid 3-3 had lost chromo

some 15 (Table 9). However, chromosomes 5 and 11 werecommonly lost in the SA+ clones.

In addition to the chromosome analysis of 5 near-tetraploidITxlN hybrids, 18 SA+ clones from seven 2TxlN hybridclones were partially karyotyped. All of the metaphases in SA"1"

clones had lost one or two copies of chromosome 15. Thus,only hybrid 3-3 among the 12 hybrids studied failed to show anonrandom loss of chromosome 15 associated with reexpres-sion of transformation and tumorigenicity. There was no common loss or gain of marker chromosomes in any of the hybridsstudied, although additional marker chromosomes were acquired in some parental and SA+ clones.

DISCUSSION

We have examined the ability of normal SHE cells to suppress tumorigenicity of a tumor cell line (ras/myc-T) inducedby transfection of the SHE cells with v-Ha-ras plus \-myconcogenes. The results show that the transformed morphology,anchorage-independent growth, and tumor-forming ability ofthe hybrid cells are inhibited. When the hybrid cells werepassaged, anchorage-independent variants appeared in the cultures which were accompanied by morphological changes andtumorigenicity. These transformed segregants of the hybridcells had a nonrandom loss of chromosome 15 in 11 of 12hybrid clones. The results support the idea that the loss ofchromosome 15 results in the loss of a cellular gene whichsuppresses a phenotypic change necessary for neoplastic development. The present findings are consistent with the results ofothers, showing that fusion of normal and tumorigenic cellsresults in decreased tumorigenicity (7-10). Unlike the hamsterx hamster hybrids described in this report, transformed phe-notypes such as growth in agar are not suppressed in nontu-morigenic human x human hybrids (7). The significance of thisinteresting difference is unclear. Normal human cells suppressboth anchorage-independent growth and tumorigenicity of Syrian hamster tumor cells (23, 24), and anchorage-independentgrowth and tumorigenicity correlate more closely in Syrianhamster cells than in human cells (16, 24). These findings donot necessarily indicate any significant difference between human and hamster cells except that anchorage-independentgrowth is a better phenotypic marker for tumorigenicity in thelatter.

The reexpression of tumorigenicity in cell hybrids has beenassociated with the loss of specific chromosomes in otherstudies (25-28). In human x human cell hybrids, the segregation of tumorigenic hybrids from initially nontumorigenic hybrids results after the loss of specific chromosomes. For example, in HeLa x normal human fibroblasts hybrids, reexpressionof tumorigenicity is associated with the loss of a single copy ofhuman chromosome 11 (25). Stanbridge and coworkers have

Table 6 Chromosomal constitution of SHE cells, ras/myc-Tm cells, and SHE x ras/myc-Tm hybrids

Cells

Normal chrosomes Marker chrosomes

Modalchromo

someno.° l 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 B C D E

SHE

ras/myc-TmSHE x ras/myc-

Tm hybrid (expected)

44

43-4587-89

222222222222222222222 2or

2222 1222221222222221 l l444443444444443334444 3

or

orl

lor0

l 2l 2

l l ll l l

' The observed or expected number of total, normal, or marker chromosomes for each cell type is given.

1628

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN roí/m^r-INDUCED TUMORS

Fig. 5. Modal karyotype of near-tetraploidhybrid 2-2. Note that one copy of the markerchromosomes except for the B marker in theras/myc-Tm parental cell is present along withone additional marker and that 3 copies ofchromosome 15 are present. This indicatesthat the hybrid was formed by fusion of oneras/myc-Tm and one normal SHE cell. SeeTable 6.

Table 7 Karyotypic analysis of SHE x ras/myc-Tm hybrid 2-2 and anchorage-independent (SA) variants of the hybrid

Modalchromo-

Clone some no.

Normal chromosomes" Marker chromosomes

7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 X Y A B C D E Others

Hybrid2-2SA-1!>A-3SA-6SA-7SA-9

SA-108480767276773443334333442423143432

23333334333443423243432

11 23333334333442323244442 1113343334333443423244432

1 24343334432442222243442 11 1 1

' Q-banded metaphases of at least 10 metaphases per clone were analyzed and the highest number of each normal or marker chromosome observed was determined.

Fig. 6. Modal karyotype of an anchorage-independent variant from the hybrid 2-2. Notethat several chromosomes are missing and twoadditional markers are present. Loss of chromosome 15 was the only common chromosome change among the 10 anchorage-independent variants examined. See Table 6.

¡ííMr M , M• eVC il1 2 34«icifl'cn67

" 12..

;l »i»u ».i.16 17 is

2-2 SA6

further shown that the chromosome 11 from the normal cells, tumorigenicity (26). These findings provide strong support fornot the HeLa cells, is lost and reintroduction of a single human the concept that a gene on this chromosome suppresses tumor-chromosome 11 by microcell transfer results in suppression of igenicity. Other human tumors are suppressed for tumorigenic-

1629

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN raj/m^c-INDUCED TUMORS

Table 8 Karyotypic analysis of SHE x ras/myc-Tm hybrid 2-3 and anchorage-independent (SA) variants of the hybridNormal chromosomes0 Marker chromosomes

Modalchromo-

Clone some no. 10 11 12 13 14 15 16 17 18 19 20 21 X B C D E Others

Hybrid2-3SA-1SA-3SA-5

SA-6SA-8SA-9

SA-10787075

7-169

78776571333

*3

33244443442344343434244234432343324322443434342442344243434244223323

333424423323322222 2233222

22233333£433334432333322

1111112

11112111

2211

22 1121

1311 2

1 1 1°Q-banded metaphases of at least 10 metaphases per clone were analyzed and the highest number of each normal or marker chromosome observed was determined.

Table 9 Karyotypic analysis of SHE x ras/myc-Tm hybrid 3-3 and anchorage-independent (SA) variants of the hybridNormal chromosomes0 Marker chromosomes

Modalchromo-

Clone some no. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 BCD Others

Hybrid3-3SA-2

SA-3SA-4

SA-5SA-78575

8083

83793 333 323 433 433 333

3333

3332

33

33

334

343334334

343

42

33

3333333333333332233323344 3333 33312111211 12111111

12111211l;:

°Q-bandedmetaphases of at least 10 metaphases per clone were analyzed and the highest number of each normal or marker chromosome observed was determined.

ity in cell hybrids with normal cells but the tumor suppressorgene involved may be on a different chromosome. In HT 1080fibrosarcoma x normal human fibroblasts hybrid, loss of chromosome 1 and/or chromosome 4 is observed in the hybridswhich reexpress tumorigenicity (27). Studies on hybrids between normal human fibroblasts and Syrian hamster tumorcells also indicate that human chromosome 1 is important insuppressing tumorigenicity of this interspecies hybrid (24). Inmouse cell hybrids, chromosome 4 of the mouse appears to beimportant in suppressing tumorigenicity (28). Chromosome 15of the Syrian hamster may contain a gene that is analogous, atleast in function, to the tumor suppressor gene(s) mapped tothese human and mouse chromosomes.

An important role for chromosome loss in neoplastic development is also suggested by studies of dominantly inheritedpredisposition to specific cancers in humans. By analysis ofrestriction fragment length polymorphisms, it can be shownthat loss of chromosome 13 occurs frequently in ret ¡noblasto-mas and osteosarcomas (13, 29, 30) and loss of chromosome11 in Wilms' tumor, rhabdomyosarcoma, hepatoblastomas, and

bladder tumors (31, 32). These tumors, however, do not usuallyexhibit monosomies for these chromosomes (30, 33). In part,this is because chromosome loss followed by duplication of theremaining copy of the chromosome occurs (3, 13, 29). This hasbeen shown to result in homozygosity of inherited mutationsin both the Wilms' and retinoblastoma loci (31, 32). Other

mechanisms, such as somatic recombination, deletions, geneconversion, or point mutations may also result in homozygosityor hemizygosity of these loci, but chromosome loss with orwithout reduplication is clearly an important mechanism inthese human tumors (13). In the case of ros/wyc-induced Syrianhamster tumor cells, loss of chromosome 15 frequently occurswithout reduplication of the remaining chromosome. Double

nondisjunction does occur for other chromosomes in chemicallytransformed Syrian hamster cell lines (3, 34). The factorsdetermining whether a chromosome remains monosomic or isduplicated by a second nondisjunction event are unclear; butthe occurrence of double nondisjunction is clearly more frequent than recognized and may obscure the importance ofchromosome loss in some tumors (31, 32).

The near-tetraploid SHE x ras/myc-T hybrids that reex-pressed tumorigenicity had lost only a single copy of chromosome 15. Thus, these tumorigenic hybrids have 2 copies ofchromosome 15 compared to only 1 copy in the parental ras/myc-T cells. This observation can be explained by one of twopossibilities. Only one copy of the putative tumor suppressorgene on chromosome 15 may be active in the normal SHE cells.Therefore, loss of only a single chromosome 15 (the one withthe active alíele)is necessary in the ras/myc tumors or the SHEx ras/myc-T hybrids. The results from studies of Wilms' tumor

and retinoblastoma are consistent with this possibility. However, in these tumors, the inactive alíeleis acquired by a germline mutation. In the SHE cells the identification of the mechanism responsible for the inactivation of one alíeleon chromosome 15 is a key omission in this hypothesis. Alternatively,gene dosage effects may be operative in tumor suppression; asingle gene in a near-diploid ras/myc tumor cell or 2 genes ina near-tetraploid cell hybrid may not be sufficient to overcomethe neoplastic effects of the positive growth signals in the cell.The existence of gene dosage phenomena in controlling expression of tumorigenicity is supported by the observations ofBenedict et al. (27) who showed that near-diploid HT 1080tumor cells are suppressed in cell hybrids with normal cellswhereas near-tetraploid derivatives of HT 1080 tumor cells arenot suppressed in cell hybrids.

We did observe that in one of 12 hybrids examined, loss of1630

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN ras/myc-MDUCED TUMORS

chromosome 15 was not associated with reexpression of tu-morigenicity. We have also observed that a few exceptional ras/myc tumors do not exhibit monosomy 15.4 Muschel et al. (35)

have also reported that rat cells transfected with cooperatingoncogenes may remain diploid. These findings are not necessarily inconsistent with our hypothesis that a gene on a specificchromosome suppresses tumorigenicity. It is possible that thegene on this chromosome becomes inactive by a mutation notvisible by chromosomal banding methods or it is possible thatthis gene is inactivated by nonmutational, epigenetic mechanisms.

One of the key findings of our studies is that tumorigenicityis suppressed in the SHE x ras/myc-T cell hybrids even thoughthe viral v-Ha-ras and \-myc oncogenes are expressed in thecells. The mutated p21"" protein was expressed at high levels

in the nontumorigenic cell hybrids and in general did notincrease significantly when the cells reexpressed tumorigenicity.This is consistent with the observations of hybrids betweennontumorigenic Chinese hamster cells fused with EJ-ras transformed Chinese hamster cells (36), hybrids of EJ bladder carcinoma cells and normal human fibroblasts (37), and hybridsbetween flat revertants of Kirsten murine sarcoma virus-transformed cells and other mouse cell lines transformed by rajoncogenes (15). Similar to our results, these authors reportedthat tumorigenicity of the hybrids is suppressed despite thecontinued expression of the p2\"" protein. Thus, in these stud

ies, the tumor suppressor gene does not appear to act byinterfering with transcription or translation of the ras oncogeneproduct. Tumor suppressor gene functions may interact with,modulate, or compete with the oncogene product at a differentpoint within a chain of events leading to neoplastic cell growth.Until the exact mechanisms of action of these normal genes areelucidated, they should be referred to as tumor suppressorgenes, a functional description, rather than antioncogenes, aterm which may connote a specific mechanism.

Our studies on the control of v-myc expression in the hybridswere complicated by the observations that \-myc expressionwas decreased in the hybrids. However, reexpression of tumorigenicity did not result from an elevation of the \-myc RNAlevels. Further studies on the role of \-myc in these hybrids areneeded to fully understand these observations.

In addition to the putative tumor suppressor gene on hamsterchromosome 15, tumorigenicity may also be controlled byfactors or genes resulting in cellular senescence. We observedthat approximately half of the SHE x ras/myc-T hybrids se-nesced as the normal cells indicating that immortality may berecessive. Some of the hybrids escape this senescence but lossof chromosome 15 and expression of tumorigenicity do notoccur at this stage. Pereira-Smith and Smith have also reportedthat immortality is recessive in human x human hybrids (18),and we have observed similar findings with SHE cells fusedwith chemically transformed hamster cells (16). Control ofoncogene expression remains a possible mechanism for suppressing immortality. Identification of a specific chromosomelost during escape from senescence may allow mapping of apossible senescence gene to this chromosome. We agree withthe hypothesis of Sager (17) that cellular senescence is anotherpossible mechnism for tumor suppression.

In summary, we have shown that tumorigenicity of rasi myctumor cells may be suppressed by normal SHE cells. Thissuppression may be the result of genes or factors involved incontrol of senescence. However, hybrid cells which escape se

nescence are still nontumorigenic despite elevated expressionsof activated oncogenes. Loss of one copy of chromosome 15correlates with reexpression of tumorigenicity of these hybridsindicating that this chromosome may carry a tumor suppressorgene. Understanding the function of this tumor suppressor genemay provide new insights into negative control mechanismsinvolved in normal and neoplastic cell growth.

ACKNOWLEDGMENTS

We would like to thank Dr. Steve Reynolds for the gift of the Y13-259 antibody, Lois Annab for expert technical assistance in the performing of the radioimmune precipitation, Dr. Tona Gilmer for adviceand helpful discussions, Dr. Roger Wiseman, Dr. Anton Jetten, andDr. Jud Spalding for critical reading of the manuscript, and HelenaBonner for expert typing.

REFERENCES

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

' Manuscript in preparation.

Sandberg, A. A. A chromosome hypothesis of oncogenesis. Cancer Genet.Cytogenet., «/.•277-285, 1983.Sasaki, M. Role of chromosomal mutation in the development of cancer.Cytogenet. Cell Genet., 33: 160-168, 1982.Oshimura, M., and Barrett, J. C. Chemically induced aneuploidy in mammalian cells: mechanisms and biological significance in cancer. Environment.Mutagen., 8: 129-159, 1986.Evans, H. J. Neoplasia and cytogenetic abnormalities. In: V. L. Dellarco, P.E. Voytek, and Hollander (eds.), Aneuploidy: Etiology and Mechanisms, pp.165-178. New York: Plenum Press, 1985.Klein, G., and Klein, E. Oncogene activation and tumor progression. Carci-nogenesis (Lond.). 5: 429-435, 1984.Oshimura, M., Gilmer, T. M., and Barrett, J. C. Nonrandom loss of chromosome 15 in Syrian hamster tumours induced by v-Ha-ros plus \-myconcogenes. Nature (Lond.), 316: 636-639, 1985.Stanbridge, E. J., Der, C. J.. Doersen, C.-J.. Nishimi. R. Y., Peehl, D. M..Weissman, B. E., and Wilkinson, J. E. Human cell hybrids: analysis oftransformation and tumorigenicity. Science (Wash. DC), 215: 252-259,1982.Harris, H., Miller, O. J., Klein, G., Worst, P., and Tachibana, T. Suppressionof malignancy by cell fusion. Nature (Lond.). 223: 363-368, 1969.Sager, R., and Kovac, P. E. Genetic analysis of tumorigenesis. I. Expressionof tumor-forming ability in hamster hybrid cell lines. Somat. Cell Genet., 4:375-392, 1978.Weissmann, B. E. Suppression of tumorigenicity in mammalian cell hybrids.In:S. C. Barrett (ed.). Mechanisms of Environmental Carcinogenesis. Vol. I.Epigenetic Changes, pp. 31-45. Boca Raton. EL: CRC Press, 1987.Knudson, A. G. Mutation and human cancer. Adv. in Cancer Res., 17: 317-352, 1973.Murphree, A. L., and Benedict, W. F. Retinoblastoma: clues to humanoncogenesis. Science (Wash. DC), 223: 1028-1033. 1984.Cavenee, W. K., Dryja, T. P. Phillips, R. A.. Benedict. W. F., Godbout, R..Gallic, B. L.. Murphree, A. L.. Strong, L. C., and White. R. L. Expressionof recessive alíelesby chromosomal mechanisms in retinoblastoma. Nature(Lond.), 305: 779-784, 1983.Noda, M., Selinger, /.. Scolnick, E. M., and Bassin, R. H. Flat revertantsisolated from Kirsten sarcoma virus-transformed cells are resistant to theaction of specific oncogenes. Proc. Nati. Acad. Sci., USA, 80: 5602-5606,1983.Bassin, R. H., Noda, M., Scolnick, E. M., and Selinger, Z. S. Study ofpossible relationships among retrovirus oncogenes using flat revertants isolated from Kirsten sarcoma virus-transformed cells. In: G. F. VandeWoode,A. J. Levine, W. C. Topp, and J. D. Watson (eds.), Cancer Cells 2, pp. 2463-2471. Cold Spring Harbor. NY: Cold Spring Harbor Laboratory, 1984.Koi, M., and Barrett, J. C. Loss of tumor suppression function duringchemically induced neoplastic progression of Syrian hamster embryo cells.Proc. Nati. Acad. Sci. USA, 83: 5992-5996, 1986.Sager, R. Genetic expression of tumor formation: new frontier in cancerresearch. Cancer Res., 46: 1573-1580, 1986.Pereira-Smith, O. M., and Smith, J. R. Evidence for the recessive nature ofcellular immortality. Science (Wash. DC), 221: 963-966, 1983.Lumpkin, C. K., McClung, J. K., Pereira-Smith, D. M., and Smith, J. R.Existence of high abundance antiproliferative mRNA's in senescent humandiploid fibroblasts. Science (Wash. DC), 232: 393-395, 1986.Thomassen, D. G., Gilmer, T. G., Annab, L. A., and Barrett, J. C. Evidencefor multiple steps in neoplastic transformation of normal and preneoplasticSyrian hamster embryo cells following transfection with Harvey murinesarcoma virus oncogene (v-Ha-ros). Cancer Res., 45: 726-732, 1985.Furth, M. E., Davis, L. J., Fleurdelys, B., and Scolnick, E. M. Monoclonalantibodies to the p21 products of the transforming gene of Harvey murinesarcoma virus and of the cellular ras gene family. J. Virol., 43: 294-304,1982.

1631

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

ROLE OF CHROMOSOME LOSS IN roi/myc-INDUCED TUMORS

22. Srivastava, S. K., Yuasa, Y., Reynolds, S. H., and Aaronson, S. A. Effects oftwo major activating lesions on the structure and conformation of human ra.voncogene products. Proc. Nati. Acad. Sci. USA, 82: 38-42, 1985.

23. Stoler, A., and Bouck, N. Identification of a single chromosome on thenormal human gene essential for suppression of hamster cell transformation.Proc. Nati. Acad. Sci. USA, 82: 570-574, 1985.

24. Bouck, N., Stoler, A., and Polverini, P. J. Coordinate control of chromosome1 in hamster-human hybrids. Cancer Res., 46: 5101-5105, 1986.

25. Stanbridge, E. J., Flandermer, R. R., Daniels, D. W., and Nelson-Rees, W.A. Specific chromosome loss associated with the expression of tumorigenicityin human cell hybrids. Somat. Cell Genet., 7: 699-712, 1981.

26. Saxon, P. J., Srivatsan, E. S., and Stanbridge, E. J. Introduction of humanchromosome 11 via microcell transfer controls tumorigenic expression ofHeLa cells. EMBO J., 5: 3461-3466, 1986.

27. Benedict, W. F., Weissman, B. E., Mark, C., and Stanbridge, E. J. Tumorigenicity of human HT1080 fibrosarcoma x normal fibroblast hybrids: chromosome dosage dependency. Cancer Res., 44: 3471 -3479, 1984.

28. Evans, E. P., Burtenshaw, M. D., Brown, B. B., Hennion, R., and Harris, H.The analysis of malignancy by cell fusion IX. Re-examination and clarification of the cytogenetic problem. J. Cell Sci., 56: 113-130, 1982.

29. Hansen, M. F., Kontos. A., Gallie, B. L., Phillips, R. A., Fodstad, O.,Brogger, A., Gedde-Dahl, T., and Cavenee, W. K. Osteosarcoma and reti-noblastoma: a shared chromosomal mechanism revealing recessive predisposition. Proc. Nati. Acad. Sci. USA, «2:6216-6220, 1985.

30. Benedict, W. F., Banerjee, A., Corey, M., and Murphee, A. L. Nonrandomchromosomal changes in untreated retinoblastoma. Cancer Genet. Cytoge-net., 70:311-333, 1983.

31. Koufos, A., Hansen, M. F., Copeland, N. G., Jenkins, N. A., Lampkin, B.C., and Cavenee, W. K. Loss of heterozygosity in three embryonal tumourssuggests a common pathogenetic mechanism. Nature (Lond.), 316:330-334,1985.

32. Fearon, E. R., Feinberg, A. P., Hamilton, S. H., and Vogelstein, B. Loss ofgenes on the short arm of chromosome 11 in bladder cancer. Nature (Lond.),Ì18:377-380, 1985.

33. Mitelman, F. Catalog of Chromosome Aberrations in Cancer. New York:Alan R. Liss, Inc., 1985.

34. Oshimura, M., and Barrett, J. C. Double nondisjunction during karyotypicprogression of chemically induced Syrian hamster cell lines. Cancer Genet.Cytogenet., 18:131-139, 1985.

35. Muschel, R. J., Nakahara, K., Chu, E., Pozzatti, R., and Liotta, L. A.Karyotypic analysis of diploid or near diploid metastatic Harvey ras transformed rat embryo fibroblasts. Cancer Res., 46:4104-4108, 1986.

36. Craig, R., and Sager, R. Suppression of tumorigenicity in hybrids of normaland oncogene-transformed CHEF cells. Proc. Nati. Acad. Sci. USA, 82:2062-2066, 1985.

37. Geiser, A. G., Der, C. J., Marshall, C. J., and Stanbridge, E. J. Suppressionof tumorigenicity with continued expression of the c-11¡i-ra.voncogene in I Jbladder carcinoma-human fibroblast hybrid cells. Proc. Nati. Acad. Sci. USA,«3:5209-5213,1986.

1632

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

1988;48:1623-1632. Cancer Res   Mitsuo Oshimura, Minoru Koi, Nobuyoshi Ozawa, et al.   Tumors

-induced Syrian Hamsterras/mycRole of Chromosome Loss in

  Updated version

  http://cancerres.aacrjournals.org/content/48/6/1623

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

  [email protected] at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/48/6/1623To request permission to re-use all or part of this article, use this link

Research. on November 23, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from