Expression of SV40 large T antigen, but not small t antigen, is required for the induction of chromosomal aberrations in transformed human cells

VIROLOGY

180, 49-57 (1991)

Expression of SV40 Large T Antigen, but Not Small t Antigen, Is Required for the Induction of Chromosomal Aberrations in Transformed Human Cells NANCY STEWART AND SILVIA BACCHETTI 1 Molecular Virology and immunology Program, Department of Pathology, McMaster University, Hamilton, Ontario, Canada L8N 3Z5 Received June 6, 1990, accepted August 24, 1990 Expression of the Simian virus 40 (SV40) early region in human cells results in the induction of chromosomal aberrations and polyploidy, and in transformation . To understand how genetic damage occurs and what role it plays in transformation, human diploid fibroblasts and embryonic kidney cells were transfected with plasmids encoding wild type or mutant forms of the viral early region, and the neo gene . Clones selected for G418 resistance and expressing viral genes were initially analyzed within 20 cell divisions . Our results demonstrate that expression of the SV40 large T antigen is sufficient for the induction of chromosomal damage and ploidy changes, and that small t does not contribute to these processes . Mutant plasmids lacking the SV40 origin of DNA replication were as proficient as wild type plasmids, indicating that viral DNA replication is not required for cytogenetic damage . We have also shown that chromosome aberrations, but not necessarily polyploidy, increase in frequency and complexity upon subculturing of the clones regardless of whether such populations arrest at crisis or yield immortal lines . Our results are compatible with the hypothesis that large T antigen destabilizes the cellular genome, and that specific mutations arising from this process may contribute to cell immortalization . © 1991 Academic Press, Inc.

INTRODUCTION

on all of the cells . On the other hand, transformed cells, like tumor cells, have abnormal and unstable karyotypes, i .e ., are mutated and constantly undergoing further mutations (Nichols, 1983 ; Chang, 1986) . Studies with human cells infected with simian virus 40 (SV40) or transfected with SV40 or adenovirus DNA have further indicated that genetic alterations can be detected at early times after expression of the viral oncogenes, prior to crisis and the establishment of immortal lines, and can occur independently of selection for any of the transformed properties (Wolman et al., 1964 ; Moorhead and Saksela, 1965 ; Graham et al., 1977 ; Durnam at al., 1986 ; Canaani et al ., 1986 ; Chang et al., 1986 ; Shay and Wright, 1989) . These observations suggest that viral oncogenes might alter, directly or indirectly, the integrity and stability of the cell genome . This genetic damage could in turn give rise to populations of mutants amenable to selection for a variety of phenotypes, including the ability to survive crisis (immortality) and to induce tumors . We have been investigating the induction of chromosomal aberrations and of aneuploidy following expression of the SV40 early region in transfected human cell strains, with the goals of understanding the mechanisms involved in these processes and the role of genetic changes in immortalization . Previous studies have indicated that expression of large T antigen is required for all aspects of cell transformation and for the induction of karyotypic abnormalities, and that small t contributes at least to morphological transfor-

Carcinogenesis is a multifactorial process and a large body of evidence indicates that chromosomal aberrations and karyotypic instability play a causal role in the development and progression of tumors in man (Klein and Klein, 1985 ; Bishop, 1987 ; Yunis, 1987 ; Klein, 1988) . Similarly, immortalization and oncogenic transformation of human cells in vitro require multiple events, and mutations of cellular genes appear likely to contribute to the development of these transformed phenotypes . Several observations support this hypothesis . Human cells, unlike cells from some other species, very rarely acquire immortality and oncogenicity as a seemingly direct response to carcinogens or viral oncogenes (Sack, 1981 ; DiPaolo, 1983 ; Chang, 1986) . Rather, in this system, both phenotypes appear to be recessive (Stanbridge et al., 1982 ; Pereira-Smith and Smith, 1983), and the majority of human cells that are transformed on the basis of altered morphology and growth properties, retain a finite lifespan and are nontumorigenic . Immortal or tumorigenic cells can be derived from these populations at very low frequencies after a period of crisis or after appropriate selection, respectively (Sack, 1981 ; DiPaolo, 1983 ; Sager, 1984 ; Chang, 1986) . Thus, even the acquisition and expression of viral oncogenes appear to be insufficient to confer an unlimited lifespan and oncogenic properties ' To whom requests for reprints should be addressed . 49

0042-6822/91 $3 .00 copyright ® 1991 by Academic Press, Inc . All rights of reproduction in any form reserved.



50

STEWART AND BACCHETTI

mation (Sack, 1981 ; Monier, 1986 ; Canaani et al ., 1986 ; Chang etal ., 1986 ; Bikel etaL, 1987 ; deRonde et al., 1989 ; Shay and Wright, 1989) . Those studies, however, did not address the questions of whether large T is sufficient for the induction of cytogenetic damage, and whether its role in this process is direct or indirect . Since large T is also required for the regulation of early viral transcription and the initiation of viral DNA synthesis (DePamphilis and Bradley, 1986) it remained possible that expression of small t or DNA replication might be responsible for the induction of aberrations . The reports that on mutants might be less efficient in the induction of genetic damage appeared to support the latter hypothesis (Small et al., 1982 ; Canaani at al ., 1986 ; Neufeld et al., 1987) . In the present study we have used transfection of human cell strains with plasmids encoding wild-type (wt) or mutant early region of SV40 to : (i) define the role of the individual tumor antigens and of viral DNA replication in the induction of polyploidy and of structural chromosome aberrations in the cells ; and (ii) to investigate the evolution of genetic damage upon growth of the transfected clones, and its contribution to cell transformation . Our results demonstrate that expression of large T is sufficient to alter the integrity and stability of the cell genome, and that small t does not contribute to this process . Further, viral DNA replication is not required since wt and on- mutants are equally proficient in the induction of damage . Lastly, we have obtained evidence documenting the gradual accumulation and the evolution of cytogenetic damage during growth of the transfected cells prior to crisis and suggesting that this damage may contribute to immortalization . As this manuscript was being prepared for submission we became aware of a recent publication by Ray at al . (1990) describing results and conclusions similar to the ones reported here .

MATERIALS AND METHODS Cells Human diploid skin fibroblasts (strain 423) (Chang et al., 1986) and human embryonic kidney cells (HEK) were maintained in a-MEM (GIBCO) supplemented with penicillin (100 U/ml), streptomycin (100 µg/ml), and 10% fetal calf serum . The fibroblasts were routinely subcultured at a 1 :4 ratio and used for experiments between 6 and 10 passages . HEK cells were used as secondary cultures and were not subcultured except in the case of transfected clones which were passaged at a 1 :4 ratio .

Plasmids pSV2neo and pSV3neo (encoding respectively the neomycin resistance gene or the neo gene plus the wt SV40 early region ; Southern and Berg, 1982) were originally obtained from Dr. P . Berg, as was plasmid pX-8, which contains an on mutant of SV40 (Canaani et at, 1986) . pd/2005, which contains the neo gene and the SV40 early region with a deletion in the large T intron, was obtained from Dr . M . J . Tevethia (personal communication) . This plasmid expresses only the large T antigen . All plasmids were propagated in Escherichia colt (strain LE392 or DI-15a), and supercoiled DNAwas purified by isopycnic centrifugation in cesium chlorideethidium bromide gradients . Transfections Strain 423 and HEK cells were transfected with plasmid DNA using the calcium phosphate coprecipitation technique of Graham and van der Eb (1973) . The cells were seeded at approximately 2 .5 X 105 cells per 100mm dish and were transfected 24 hr later with 20-40 ag/dish of plasmid DNAand high-molecular-weight carrier DNA from strain 423 fibroblasts . The cells were incubated overnight in the presence of the DNA-calcium phosphate precipitate, then refed with fresh medium . Forty-eight hours after transfection, they were subcultured at a 1 :3 ratio in medium containing G41 8 (Geneticin, GIBCO) at an effective dose of 400 µg/ml . The selective medium was changed at approximately weekly intervals until colonies could be picked with cloning cylinders and expanded . All clones analyzed in this study were independently generated . Cytogenetic analysis Chromosome spreads from human fibroblasts were prepared by seeding cells at low density onto coverslips . After 24- or 48-h growth, the cells were arrested in metaphase by treatment with colcemid (0 .1 yg/ml) for 3-5 hr . The coverslips were then immersed in prewarmed hypotonic solution (0 .075 M KCI) and incubated at 37° for 20 min . Fixation was with cold methanol :acetic acid (3 :1 v/v) . HEK cells were seeded in plates, treated with colcemid as above, and harvested by trypsinization . Hypotonic treatment and fixation were as above, and chromosome spreads were prepared by the drop method . All cytogenetic analysis was performed by light microscopy on chromosomes stained with 5% Giemsa . Immunolabeling and immunoprecipitation For immunolabeling cells were grown on coverslips to 50-70% confluency . Following a wash with PBS and

SUFFICIENCY OF SV40 T FOR CHROMOSOME DAMAGE

51

fixation with acetone at -20°, the cells were reacted with a mouse monoclonal antibody against SV40 large and small tumor antigens (Mab 419 (Harlow et al., 1981)), then with a rabbit anti-mouse antibody conjugated to horseradish peroxidase (Dakopatts a/s) . Staining was with AEC substrate solution (5% AEC (4% 3-amino-9-ethylcarbazole in N,N-dimethylformamide) in 20 mM sodium acetate, 50 mM acetic acid plus one drop of 30% hydrogen peroxide) and counterstaining with hematoxylin . For immunoprecipitations, cells were labeled overnight with [ 35 S]methionine (100 pCi/ ml) in medium with 1/10 the normal concentration of methionine, harvested, and lysed in RIPA buffer plus proteolytic inhibitors (50 mM Tris-HCI, pH 7 .2, 150 mM NaCl, 0 .1% Na dodecyl sulfate (SIDS), 1 0/o Na deoxycholate, 1 % Triton X-100, 1 mM phenylmethylsulfonylfluoride (Sigma), and 0 .1 1 trypsin inhibitor units of aprotinin (Sigma) per milliliter) . Aliquots of cell lysates were reacted with Mab 419 in the presence of protein A--Sepharose beads (Pharmacia) for 3 hr at 4° . Immunocomplexes, collected by centrifugation, were washed with RIPA buffer, boiled for 3 min in Laemmli sample buffer (Laemmli, 1970), and analyzed by SDSPAGE (12% polyacrylamide) and autoradiography . RESULTS Cytogenetic analysis of cells transfected with the SV40 early region Initial experiments aimed at assessing the amount and type of cytogenetic damage induced by the SV40 tumor antigens were performed with human diploid skin fibroblasts . The cells were transfected with the pSV3neo plasmid, which contains the neo gene and the wt SV40 early region (T', t', on'), or the control pSV2neo plasmid, encoding only the neo gene . Following selection with G418 for expression of the neomycin gene, colonies of surviving cells were isolated from both populations and expanded . Immunolabeling to assess expression of the tumor antigens and cytogenetic analysis were carried out on six independent pSV3neo and six pSV2neo transfectants at passage 2, corresponding to approximately 20 divisions of the parental cell . As expected, control clones generated by transfection with pSV2neo (Fig . 1A) retained a fibroblastic morphology and were negative for expression of the SV40 tumor antigens as determined by their lack of reactivity with the Mab 419 antibody . These cells also retained the growth pattern and growth rate of the untransfected parental fibroblasts (not shown) . pSV3neo transfectants (Fig . 1 B), on the other hand, had altered (epithelioid) morphology and were all positive for expression of the tumor antigens as indicated

FIG . 1 . Expression of SV40 tumor antigens in transfected cells . Clonal populations selected after transfection of fibroblasts with (A) pSV2neo, (B) pSV3neo (T*, 1*), or (C) rd/2005 (T*, t) or of HEK cells with pSV3neo (D) were immunostained with monoclonal antibody 419, which recognizes both large and small tumor antigens .

by staining with Mab 419 . These clones grew at a faster rate and to higher saturation densities than the untransfected parental cells . All of the clones were also analyzed at passage 2 to determine the ploidy of the cells and the frequency and type of chromosomal aberrations . The results of the cytogenetic analysis are presented in Table 1 for the individual clones, and in Fig . 2 for the average of each group of clones . pSV2neo transfectants had retained an essentially normal diploid karyotype : depending upon the clone, polyploid (±4N) cells amounted to 0-4% of the total scored, and structural aberrations (primarily chromatid breaks and fragments) were observed in a maximum of 20% of the cells . In contrast, pSV3neo transfectants exhibited a high degree of polyploidy (22- 100% of the cells) and a high frequency of structural aberrations (58-84% of the cells) . Among



52

STEWART AND BACCHETTI

(7N C m

InN O0 0 O

0LL

mV a L0 ym

Ln 7 LO N N O N a

0 0 LL

vN LD N LDLO

CJ N NCu L1J (0 0 0 LO CO

0 LL

o O O

o O V

0 LL

N CO Cu C7 7 LO

0 N " N C] N MN O O W

0

O N

a U LL

OLn LnO N LP

m f VN CO M O M Ln

U LL

O W V

<-. O f 7 O O O

U LL

000 7 CO O

--Ln LO m 0 . 00 N

0 LO WC1 Om

(O N,- 0 0 0 m7

C'7 o O O M

100 pSV2neo e0 60 • • 40

100 psVOneo 80 60 40 20

t 20 r PA • 1000 100 pd12005 m 80 60 • 60 60 40 40 20 20 i 0 0 FIG . 2 . Cytogenetic analysis of clonal populations generated by transfection of fibroblasts with SV40 plasmids . The average frequency of total aberrant cells (solid bars), of polyplold cells (open bars), and of cells with structural aberrations (shaded bars) for each type of clones was obtained from the data presented in Table 1 .

a

•

•

C 0 0 C M

J IM •

m LL

07 ON COO

mL-

0 oO NO

m LL


m LL

OLO Nd 7

O O N O O Vm

M LL

W N O a N NLO

V CO NNO O LOV

cO 0 v Ln

'ONOOLu

O 0 o v

C N O 0 0 N

cO 0 a 7

N O O O O O N

O 0 N O

O O o 0 0 0 0

CO 0 N V

O O O N O O N

0 o OON

CO

LLL LL

•

NC 0 00 NC M

Q LL

Q LL

0000

the polyploid cells the majority were ±4N but a significant percentage (up to 10%) had ±8N complements . The most frequent structural aberrations consisted of dicentrics, fragments, and double minutes ; occasional chromosome and chromatid breaks, rings, and translocations or duplications were also observed . These results confirm and extend those of Chang et al . (1986) on a comparable series of transfected clones . Both studies indicated that induction of genetic changes in the cells correlates with the expression of the SV40 tumor antigens, and that such changes occur at early times after transfection of the cells . Role of large T and small t in the induction of aberrations To investigate whether induction of aberrations required the expression of both SV40 tumor antigens, additional clones were generated by transfection of fibroblasts with pd/2005, a plasmid which contains the neo gene and the SV40 early region but which expresses only large T due a deletion in the T intron . Following transfection and selection for neo expression, four colonies were expanded and analyzed . These cells (Fig . 1 C) were more epithelial in morphology and grew in a more disorganized pattern than the parental 423 cells or the pSV2neo clones, but in both respects they were not as altered as the pSV3neo transfectants . They did, however, resemble the latter clones in terms of their growth rate . All four pd/2005 colonies were found positive by immunostaining with Mab419 and were further analyzed by immunoprecipitation with this antibody to certify that they expressed large T but not small t . As shown in Fig . 3 for a representative clone (FC3, lane 3), large T was indeed expressed by these cells in amounts comparable to



SUFFICIENCY OF SV40 T FOR CHROMOSOME DAMAGE

M 1 2

3 4

3a 4a

53

which were found positive for expression of the tumor antigens by immunostaining . As shown in Table 1 and Fig . 2, polyploidy and chromosomal aberrations were readily detected in the majority of the cells of each clone, and the frequency and type of structural changes observed did not differ significantly from those seen in pSV3neo transfected cells . These results indicated that T antigen-induced viral DNA replication was not required for the induction of aberrations and pointed to the involvement of other properties of the protein . Accumulation and evolution of aberrations prior to crisis

FiG . 3 . lmmunoprecipitations of SV40 tumor antigens from transfaced cells . Equal amounts of lysates from radiolabeled cells were immunoprecipitated with antibody 419, and the immunocomplexes were analyzed by SDS-PAGE and autoradiography . Lane 1, cos- I cells; lanes 2-4, fibroblasts transfected with pSV3neo (2), pd/2005 (3), or pSV2neo (4) ; M, molecular weight markers . Lanes 3a and 4a are longer exposures of lanes 3 and 4 . The large and small arrowheads indicate large 7 and small t, respectively .

those detected in pSV3neo transfectants (lane 2), whereas small t was absent . Even upon longer exposure of the autoradiograph (lane 3a), this antigen could not be detected . Cytogenetic analysis of the pd/2005 clones revealed that they were not significantly different from pSV3neo clones (Table 1 and Fig . 2) . Between 32 and 60* of the cells had 4N or 8N complements, and 30-60% also had structural aberrations, again consisting mainly of dicentrics, fragments, and double minutes . These results indicated that expression of large Twas sufficient for the induction of genetic changes, and suggested that small t might not contribute to this process . Effect of

on

mutants

A previous study by Canaani et al. (1986) on two human cell lines immortalized by transfection with ori SV40 had reported the existence of very few chromosomal aberrations in the cells . This suggested the possibility that viral DNA replication, particularly amplification and excision of integrated viral DNA, which can occur with wt SV40 in semipermissive human cells (Zouzias et al., 1980), might contribute to the induction of aberrations . To test this possibility we cotransfected human fibroblasts with the pX-8 ori - plasmid and pSV2neo, and analyzed five independent colonies

In addition to defining the SV40 proteins required for the induction of genetic damage, our experiments were aimed at assessing the role of this damage in cell immortalization, and at identifying early genetic events (markers) which may contribute to this process . Since fibroblasts proved to be very susceptible to T antigeninduced damage, and had severely abnormal karyotypes even at early times aftertransfection, they did not seem particularly suitable for the detection of initial aberrations and for the analysis of their evolution . For this reason, an additional series of T antigen-positive clones was generated bytransfection of human embryonic kidney (HEK) cells with pSV3neo . We had used these cells in studies with adenoviruses, and found that upon expression of the viral oncogenes they did not exhibit the high frequency of aberrations seen with T antigen-positive fibroblasts (Caporossi and Bacchetti, 1990 ; Schramayr et al., 1990) . In addition, normal HEK cells cannot be subcultured as extensively as normal fibroblasts (Martin et al., 1970 ; Poirier et al., 1988 ; Shay and Wright, 1989), thus allowing for more rapid identification of clones with increased life span . Following selection with G41 8, the initial analysis of these clones was carried out after the same number of cell divisions (about 20) as for the fibroblasts . At this stage HEK cells had acquired changes in morphology and growth properties and were positive for expression of the tumor antigens (Fig . 1 D) . However as shown in Fig . 4, they exhibited a substantially lower amount of cytogenetic damage than did fibroblasts . Up to 92% of the cells retained a diploid and normal karyotype : the maximum frequency of polyploidy was 8% and the maximum frequency of cells with structural aberrations (mainly breaks and gaps) amounted to 16% . HEK clones from control pSV2neo transfections grew only to a limited extent and did not yield a sufficient number of metaphases for a comparable analysis . However, of the metaphases scored, all retained the normal diploid karyotype (data not shown) ; we have also rarely de-

54

STEWART AND BACCHETTI

100 HAI

∎ Id n coynod 0 vmh monaLo4s

so 60 40

20 0 100 HA2

so 60 40 20

∎PAhhh

N

g

c

0 100 HA3 60

m

60

0 40 W

hitIIl

26 C N U

o

p 100 HAI 80 60 40 2O

D ,

100 'HAS 80' 60

fraction of polyploid cells also increased upon passage, except in the case of one clone (HA -1) . The growth rate of four of the five HEK clones declined progressively after passage 25, but no sign of cell death was detected . By passage 30 all of these clones had ceased to grow and no immortal populations could be derived from any of them . One clone, HA1, however, continued to grow vigorously at all times without any visible sign of crisis ; HAI cells have currently been passaged over 80 times and can be considered immortal . Cytogenetic analysis of HAI was also performed at passage 51 (about 120 divisions) . As shown in Fig . 4, by this stage the frequency of cells with structural aberrations had further increased, whereas that of polyploid cells had remained essentially constant and at levels significantly lower than those observed for the other clones . The latter observation suggests the possibility that high levels of polyploidy might interfere with immortalization, but the small sample size does not permit a definitive conclusion . Of interest was the finding that the most frequent rearrangement (66ek of the total aberrations) in HAI cells consisted of a marker chromosome most likely involving a member of group C . Retrospective analysis of all the transfectants revealed the presence of the same marker in two of the HEK clones with limited lifespan and in six of the fibroblast clones . This rearrangement therefore does not ensure acquisition of the immortal phenotype, but might contribute to it by conferring an initial growth advantage .

20'

DISCUSSION

20

X

0

2036

466

56

66 120

cell divisions FIG. 4 . Cytogenetic analysis of HEK clones . Clonal populations of HEK cells transfected with pSV3neo and expressing SV40 antigens were analyzed at the indicated number of cell divisions after transfection . The frequency of total aberrant cells (solid bars), of polyploid cells (open bars), and of cells with structural aberrations (shaded bars) are plotted versus the age of the populations .

tected polyploidy or aberrations in untransfected HEK cells (Caporossi and Bacchetti, 1990) . pSV3neo HEK transfectants were kept in culture and screened for karyotypic aberrations at passages 10, 15, 20, and 25 (corresponding approximately to 36, 46, 56, and 66 cell divisions) . This analysis revealed that, upon passaging of the cells, complex structural aberrations (predominantly dicentrics and fragments) developed and increased in number, until by passage 20-25 they reached levels comparable to those observed in the pSV3neo fibroblast transfectants at passage 2 . The

The observation that tumor cells contain chromosomal rearrangements and unstable genomes formed the basis of the mutation theory for the origin of cancer (Boveri, 1914) . Support for this hypothesis has since been provided by innumerable studies at the cytogenetic and molecular levels which have shown that : (i) cells from a variety of tumors contain mutations or chromosomal rearrangements which often involve cellular protooncogenes or tumor suppressor genes ; (H) particular genetic abnormalities are characteristic of particular malignancies, and attest to the clonal origin of the tumors ; and (iii) inherited susceptibility to cancer is associated with congenital cytogenetic defects or with impaired ability to repair DNA (Nowell, 1983 ; Klein and Klein, 1985 ; Bodmer, 1986 ; Bishop, 1987 ; Mitelman, 1988) . The common occurrence of genetic abnormalities and chromosomal instability in cells transformed in vitro can be similarly taken as suggesting that mutations of cellular genes might contribute to the development of the immortal and oncogenic phenotypes in this system (Chang, 1986) . This is particularly likely in the

SUFFICIENCY OF SV40 T FOR CHROMOSOME DAMAGE case of transformation by chemical and physical agents which do not introduce into the cells exogenous genetic information, and which are often proven mutagens (Evans, 1984) . In virus-induced transformation, the acquisition by the cells of viral oncogenes which are known to interfere with the control of cell growth could theoretically bypass the requirement for cellular mutations (Bishop, 1987) . Yet virus-transformed cells acquire extensive genetic damage and are similar, in this respect, to transformants induced by other agents (Chang, 1986) . In addition, viruses have long been known to be capable of inducing mutations and rearrangements in the cell genome, and more recent evidence has linked these processes to the expression of the viral oncogenes (Nichols, 1983 ; Durnam et al., 1986, Chang et at, 1986 ; Caporossi and Bacchetti, 1990 ; Schramayr et al ., 1990) . In the case of SV40, studies with human cells have shown that synthesis of the large T antigen is required for the induction of chromosomal aberrations and ploidy changes (Chang, 1986 ; Chang at al., 1986) . Large T is also essential for initiation of viral DNA synthesis, regulation of viral transcription, and cell transformation (DePamphilis and Bradley, 1986, Monier, 1986 ; Salzman et al ., 1986) . The protein is multifunctional and has the ability to bind to DNA, to function as an ATPase and helicase, to complex with cellular proteins, and to induce host gene expression and DNA synthesis (Butel and Jarvis, 1986) . The present study on human cells transfected with plasmids encoding the SV40 early region was undertaken to determine whether one or both of the viral oncogenes were required for the induction of karyotypic abnormalities in the cells, to begin to define oncogene functions involved in this process, and to examine the progression of genetic changes during growth of the transfected cells prior to crisis . Although previous reports had shown that expression of the large T antigen was required for induction of genetic damage in the cells, these studies had not addressed the question of sufficiency of this protein nor that of the contribution of T antigen-mediated viral DNA replication to the cytogenetic effect (Canaani et al ., 1986 ; Chang et al ., 1986 ; Shay and Wright, 1989) . The results of our experiments indicate that large T antigen is sufficient for the induction of aberrations and of genomic instability, since cells transfected with the pd/2005 mutant (which does not express small t) had an abnormal karyotype . Our results further suggest that small t antigen does not contribute to these processes, since the levels of damage in pd/2005 transfectants were not significantly different from those detected in cells expressing the wt early region (pSV3neo clones) . This conclusion parallels that from previous

55

studies with adenoviruses in which we detected a requirement for only one of the viral oncogenes in the induction of genetic damage in human cells (Caporossi and Bacchetti, 1990 ; Schramayr at al ., 1990) . The observations that the growth rate of pd/2005 cells was similar to that of pSV3neo cells, whereas their morphology was less altered, are compatible with the notion that these two phenotypes are affected by different processes . As already suggested by others, large T alone appears capable of altering cell growth, whereas small t contributes to cell morphology (Sleigh et al ., 1978 ; Bikel et al., 1987 ; deRonde et al ., 1989 ; Donahue and Stein, 1989 ; Jat and Sharp, 1989 ; Radna et al ., 1989) . Of the possible mechanisms by which T antigen might alter the cell genome, its ability to initiate viral DNA replication by binding to the replication origin seemed among the most likely (DePamphilis and Bradley, 1986) . Autonomous replication, as well as integration, amplification, and excision of viral DNA, can occur in human cells infected or transfected with wt SV40 (Zouzias et al., 1980), and the latter three processes in particular might induce considerable damage in the cell chromosomes (Gish and Botchan, 1987 ; Ruiz and Wahl, 1990) . The inability of on mutants to replicate and excise and reports that such mutants are less efficient in the induction of aberrations supported this hypothesis (Small at aL, 1982 ; Canaani et al ., 1986 ; Gish and Botchan, 1987 ; Neufeld et al ., 1987) . However the results of our experiments with cells transfected with these mutants have indicated that they are as efficient as wt SV40 at inducing aberrations . Thus T antigenmediated processes other than viral DNA replication must play a role in damage induction . Of note is the observation that viral DNA replication is also not required for induction of damage by adenoviruses (Caporossi and Bacchetti, 1990 ; Schramayr at al ., 1990) . In a recently published report, Ray et al . (1990) have analyzed the karyotype of human fibroblasts electroporated with pSV3neo or with a mutant plasmid encoding large T but lacking small t sequences and the origin of DNA replication . Our results and theirs are in full agreement in demonstrating that large T is sufficient and that small t and viral DNA replication are dispensable for induction of cytogenetic damage . Our studies with pSV3neo-transfected HEK cells have also provided several interesting observations . The amount and complexity of genetic damage detected in these cells at early times after expression of the tumor antigens were much less than those seen in fibroblasts at the same stage of growth . Indeed HEK cells initially contained almost exclusively gaps and breaks and were polyploid only at a low frequency . They were, however, distinctly altered in morphology



56

STEWART AND BACCHETTI

and growth rate, compared to control cells . These findings suggest inherent differences in the cytogenetic response of different cell types to the SV40 oncogenes . They also indicate that alterations in morphology and growth properties might be independent of genetic damage, and might result from more direct effects of the viral proteins . Upon passaging of the HEK clones, chromosomal damage increased in both frequency and complexity and ultimately reached levels comparable to those in fibroblasts . This progression is consistent with the possibility that the breaks and gaps detected at early times might serve as substrates for the rearrangements present at later passages . Whether T antigen expression is continually required for the evolution of the aberrant karyotype or whether the protein is responsible only for the initial damage and/or destabilization of the genome could not be established by our experimental protocol . Our findings with both fibroblasts and HEK clones are compatible with either possibility . Control HEK clones, generated by transfection with pSV2neo, grew for only a very limited period, as expected, whereas all T antigen-positive clones acquired an extended lifespan . The majority (4/5) of these clones, however, ceased growing after about 60 divisions, and only one continued to grow into an immortal line . This low frequency of immortalization is in agreement with results reported by others for human cells (Chang, 1986 ; Shay and Wright, 1989) . Thus expression of T antigen per se, the occurrence of karyotypic abnormalities, or even both phenomena together do not appear sufficient to ensure the acquisition of the immortal phenotype by all of the cells . This is entirely consistent with a mutagenic mechanism of transformation which would predict that : (i) genetic damage generates random rearrangements or mutations ; and (ii) the majority of these events are null for transformation and only a subset of them, affecting critical genes, will contribute to the transformed phenotype . If this hypothesis is correct, identification of chromosomal rearrangements common to immortal cells generated by studies such as those reported here may lead to an understanding of the mechanisms underlying the development of the immortal phenotype . ACKNOWLEDGMENTS We are particularly grateful to P . L . Chang for drawing our attention to her studies and for encouraging us to pursue them, and to M . 1 . Tevethia for her gifts of SV40 plasmids and for her advice throughout this work . We also thank J . R . Smiley for help with computer graphics . F . L . Graham and J . R . Smiley are acknowledged for revision of the manuscript. This study was supported by a grant from the National Cancer Institute of Canada (NCIC) . S .B . is a Terry Fox Cancer Research Scientist of the NCIC .

REFERENCES BIKEL, I ., MONTANO, X ., AGHA, M . E ., BROWN, M ., McCORMACK, M ., BoLTAx, J ., and LIVINGSTON, D . M . (1987) . SV40 small t antigen enhances the transformation activity of limiting concentrations of SV40 large T antigen . Cell 48, 321-320 . BISHOP, 1 . M . (1987) . The molecular genetics of cancer . Science 235, 305-311 . BODMER, W . F . (1986) . Inherited susceptibility to cancer . In "Introduction to the Cellular and Molecular Biology of Cancer" (L . M . Franks and N . Teich, Eds .), pp . 93-110 . Oxford Scientific, Oxford . BOVERI, T . (1914) . "Zur Frage der Entstehung maligner tumoren ." Gustav Fischer Verlag, Jena . ("The Origin of Malignant Tumors ." Williams & Wilkins Co ., 1929) BUTEL, 1 . S ., and JARVIS, D . L. (1986) . The plasma-membrane associated form of SV40large tumour antigen : Biochemical and biological properties . Biochim . Biophys . Acta 865, 171-195 . CANAANI, D ., NAIMAN, T., TEITZ, T ., and BERG, P . (1986) . Immortalization of Xeroderma pigmentosum cells by simian virus 40 DNA having a defective origin of DNA replication . Somatic Cell Mol. Genet. 12, 13-20 . CAPOaossi, D ., and BACCHErn, S . (1990) . Definition of adenovirus type 5 functions involved in the induction of chromosomal aberrations in human cells . J. Gen . Virol. 71, 801-808 . CHANG, P . L ., GUNBY, J . L., TOMKINS, D . J ., MAK, I ., ROSA, N . E ., and MAK, S . (1986) . Transformation of human cultured fibroblasts with plasmids carrying dominant selection markers and immortalizing potential . Exp. Cell Res . 167, 407-416 . CHANG, S . E . (1986) . In vitro transformation of human epithelial cells . Biochim . Biophys . Acts 823, 161-194 . DEPAMPHILS, M ., and BRADLEY, M . K . (1986) . Replication of SV40 and polyoma chromosomes . In "The Papovaviridae," Vol . 1, "The Polyomaviruses" (N . P . Salzman, Ed .), pp . 99-246 . Plenum, New York/London . DERONDE, A., SOL, C . J . A., VAN STRIEN, A., TER SCHEGGET, J ., and VAN DER NooeDAA, J . (1989) . The SV40 small t antigen is essential for the morphological transformation of human fibroblasts . Virology 171,260-263 . DIPAOLO, J . A. (1983) . Relative difficulties in transforming human and animal cells in vitro . JNCI 70, 3-8 . DONAHUE, L . M ., and STEIN, G . H . (1989) . Coordinate control of growth arrest states in normal human diploid fibroblasts : Effect of cloned SV40 tumour antigen genes . Exp . Cell Res. 184, 297-303 . DURNAM, D . M ., SMITH, P. P ., MENNINGER, J . C ., and McDOUGALL, 1 . K . (1986) . The El region of human adenovirus type 12 determines the sites of virally induced chromosomal damage . Cancer Cells 4, 349-354 . EVANS, H . J . (1984) . Genetic damage and cancer . In "Genes and Cancer" (J . M . Bishop, 1 . D . Rowley, and M . Greaves, Eds .), pp . 3-18 . A . R . Liss, New York . GISH, W . R ., and BOTCHAN, M . R . (1987) . Simian virus 40-transformed human cells that express large T antigens defective for viral DNA replication . 1. Virol. 61, 2864-2876 . GRAHAM, F. L., and VAN DER ER, A . J . (1973). A new technique for the assay of infectivity of human adenovirus 5 DNA . Virology 54, 456467 . GRAHAM, F . L ., SMILEY, J . R ., RUSSELL, W . C ., and NAIRN, R . (1977) . Characteristics of a human cell line transformed by DNA from human adenovirus type 5 . J. Gen. Viral. 36, 59-72 . HARLOW, E ., CRAWFORD, L . V., PIM, D . C ., and WILLIAMSON, N . M . (1981) . Monoclonal antibodies specific for simian virus 40 tumour antigens . J. Virol. 39, 861-869 . JAT, P . S ., and SHARP, P . A. (1989) . Cell lines established by a temperature-sensitive simian virus 40 large-T-antigen are growth re-

SUFFICIENCY OF SV40 T FOR CHROMOSOME DAMAGE

stricted at the nonpermissive temperature . Met. Cell. 1672-1680 .

Biol. 9,

KLEIN, G . (1988) . The approaching era of tumor suppressor genes . Science 238, 1539-1545 . KLEIN, G ., and KLEIN, E . (1985) . Evolution of tumours and the impact of molecular oncology . Nature (London). 315, 190-195 . LAEMMLI, U . K . (1970) . Cleavage of structural proteins during the assembly of the head of bacteriophage T4 . Nature (London) 227, 680-685 . MARTIN, G . M ., SPRAGUE, C . A ., and EPSTEIN, C . 1 . (1970) . Replicative life span of cultivated human cells . Lab . Invest. 23, 86-92 . MITELMAN, F . (1988) . "Catalog of Chromosome Aberrations in Cancer,'' 3rd ed ., A . R . Liss, New York. MONIER, R . (1986) . Transformation by SV40 and polyoma . In "The Papovaviridae," Vol . 1, "The Polyomaviruses" (N . P . Salzman, Ed .), Plenum, New York/London, 247-294 . MOORHEAD, P . S ., and SAKSELA, E . (1965) . The sequence of chromosome aberrations during SV40 transformation of a human diploid cell strain . Hereditas 52, 271-284 . NEUFELD, D . S ., RIPLEY, S ., HENDERSON, A ., and OZER, H . L . (1987) . Immortalization of human fibroblasts transformed by origin-defective simian virus 40 . Mot. Cell. Biol. 7, 2794-2802 . NICHOLS, W . W . (1983) . Viral interactions with the mammalian genome relevant to neoplasia . In ''Chromosome Mutations and Neoplasia" (1 . German, Ed .), pp . 317-332 . A. R . Liss, New York . NOWELL, P . C . (1983) . Tumor progression and clonal evolution : The role of genetic instability . In "Chromosome mutations and neoplasia" (J . German, Ed .), pp . 397-412 . A . R . Liss, New York . PEREIRA-SMITH, O . M ., and SMITH, J . R . (1983) . Evidence forthe recessive nature of cellular immortality . Science 221, 964-966 . POIRIER, V ., TYLER, S .1 ., BROWN, K . W ., SHAW, A . P. W ., and MAITLAND, N .1 . (1988) . SV40 transfection of human kidney epithelial cells and stability of chromosome 11 . Int. l. Cancer 42, 887-894 . RADNA, R . L ., CATON, Y., JHA, K . K ., KAPLAN, P ., LI, G ., TRAGANOS, F ., and OZER, H . L . (1989) . Growth of immortal simian virus 40 tsAtransformed human fibroblasts is temperature dependent . Mot. Cell . Blot, 9, 3093-3096 . RAY, F . A ., PEABODY, D . S ., COOPER, J . L ., CRAM, L . S ., and KRAEMER, P . M . (1990) . SV40 T antigen alone drives karyotype instability that

57

precedes neoplastic transformation of human diploid fibroblasts . J. Cell . Biochem. 42, 13-31 . Rulz, J . C ., and WAHL, G . M . (1990) . Chromosomal destabilization during gene amplification . Mot. Cell. Biol. 10, 3056-3066 . SACK, G . H . (1981) . Human cell transformation bysimian virus 40-A review . In Vitro 17, 1-19 . SAGER, R . (1984) . Resistance of human cells to oncogenic transformation . Cancer Cells 2, 487-493 . SALZMAN, N . P ., NATARoJAN, V ., and SELZER, G . B . (1986). Transcription of SV40 and polyoma virus and its regulation . In "The Papovaviridae," Vol . 1, "The Polyomaviruses'' (N .P . Salzman, Ed .), pp . 27-98 . Plenum, New York/London . SCHRAMAYR, S ., CAPOROSSI, D ., MAK, I ., JELINEK, T ., and BACCHETn, S . (1990) . Chromosomal damage induced by human adenovinus type 12 requires expression of the E1B 55-kilodalton viral protein . J. Virol. 64, 2090-2095 . SHAY, J . W ., and WRIGHT, W . E . (1989) . Quantitation of the frequency of immortalization of normal human diploid fibroblasts by SV40 large T-antigen . Exp . Cell Res. 184, 109-118 . SLEIGH, M . J ., Topp, W . C ., HANISH, R ., and SAMBROOK, J . F . (1978) . Mutants of SV40 with an altered small t protein are reduced in their ability to transform cells . Cell 14, 79-88 . SMALL, M . B ., GLUZMAN, Y., and OZER, H . L . (1982) . Enhanced transformation of human fibroblasts by origin-defective simian virus 40 . Nature (London) 296, 671-672 . SOUTHERN, P . J ., and BERG, P . (1982) . Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. Mot. Appl. Genet. 1, 327--341 . STANBRIDGE, E . J ., DER, C . J ., DOERSEN, G-J ., NISHIMI, R . Y ., PEEHL, D . M ., and WEISSMAN, B . E . (1982) . Human cell hybrids : Analysis of transformation and tumorigenicity . Science 215, 252-259 . WOLMAN, S . L ., HIRSCHHORN, K ., and ToDARO, G . J . (1964) . Chromosomal changes in SV40-infected human fibroblast cultures . Cytogenetics 3, 45-61 . YuNIS, J . J . (1987) . Multiple recurrent genomic rearrangements and fragile sites in human cancer . Somatic Cell Mot. Genet. 25, 149159 . ZOUZIAS, D ., JHA . K ., MULDER, C ., BASILICO, C ., and OZER, H . L . (1980) . Human fibroblasts transformed by the early region of SV40 DNA : Analysis of "free" viral DNA sequences . Virology 104, 439-453 .