EXPERIMENTALCELLRESEARCH

187,143-l@

(1990)

Complete Transformation of Embryonal Rat Fibroblasts by Polyomavirus Occurs during Passage in Vitro INGRID MARTENS,* *Department

TORBJ~RN

RAMQVIST,~

Academic

AND STIG LINDER*,$~

of Medical Virology, Uppsala University Biomedical Centre, Box 584, 75123 Uppsalu, Sweden; TDepartment of Tumor Biology, Karolinska Institute, 104 01 Stockholm, Sweden; and SDepartment of Oncology, Radiumhemmet, Karolinska Institute and Hospital, 104 01 Stockholm, Sweden

mouse polyomavirus encodes three tumor antigens, the large, middle, and small T-antigens (reviewed in Ref. [26]). Expression of the middle T-antigen is sufficient for transformation of established rodent cell lines such as 3T3 cells [32]. The large T-antigen confers on primary cells the ability to grow indefinitely in culture [ 241. For expression of the completely transformed phenotype, the small T-antigen is required as well [ 19,231. It is well accepted that the evolvement of naturally occurring tumors proceeds in several steps. A similar process of tumor progression has also been demonstrated after in vitro transformation of primary rodent cells. Thus, carcinogen-treated embryonal hamster cells [l] and polyomavirus-infected or -transfected embryonal rat fibroblasts developed an increasing tumorigenicity during in vitro passage [2, 341. In this report we extend these previous findings by showing that cells transfected by an origin-deficient polyomavirus are not fully transformed at early times after transfection. During growth in vitro these cells became increasingly transformed and tumorigenic and developed the ability to grow in the absence of large T-antigen. We find that changes occur in the expression of some cellular genes, but not in the expression of the integrated virus, during in vitro propagation.

The tumorigenicity of secondary rat embryo fibroblasts transfected with a plasmid harboring a replication origin-defective polyomavirus was found to increase during in vitro propagation. Thus, polyomavirus-transfected cells were found to be more than lO,OOO-fold more tumorigenic when injected into syngenic rats at 3 months after transfection compared to those injected at an earlier time point. Furthermore, most clones of polyomavirus-transfected cells did not grow in semisolid medium at 52 days after transfection but did grow at 95 days. Addition of glucocorticoid hormones, but not of 25% fetal calf serum, to the growth medium of the early passage cells resulted in limited anchorage-independent growth. An altered level of expression of a number of proteins was found in cells analyzed at different times after transfection. Notably, the expression of a component of the actin filament system, tropomyosin 2, was shown to decrease during growth in vitro. The development of a more fully transformed phenotype at late passages correlated with loss of the requirement for large T-antigen for growth. Thus, cells transfected with a polyomavirus mutant encoding a thermolabile large T-antigen did not grow at the restrictive temperature at 6 weeks after transfection, but grew well at 5 months after transfection. We suggest that these phenomena may be explained by assuming that establishment of rodent fibroblasts, and thereby sensitivity to transformation by middle T-antigen, is not an immediate consequence of expression of large Tantigen but occurs after a period of growth in vitro. 0 1990

TINA DALIANIS,?

Press,

MATERIALS

METHODS

Plasmid DNA. The polyomavirus genomes used were replication defective, and propagated as recombinants with the plasmid pBR322 joined at the BamHI sites in both molecules. The plasmidpdl1023 [14] encodes all three tumor antigens of polyomavirus but contains a 6-bp deletion that inactivates the viral origin of replication. Plasmid ~5~1051 [18] encodes the small and middle tumor antigens. The large tumor antigen mRNA cannot be expressed due to a point mutation in the 5’ splice site. Plasmid pSV2neo [29] confers resistance to the neomycin derivative G418. Plasmid pPytsA contained the tsA mutant of polyomavirus [6] cloned in the BamHI site in pBR322. Cell cultures and transfectiom. Cells were grown in Dulbecco’s modified Eagles’s medium supplemented with 5% fetal calf serum. Rat embryo fibroblasts were prepared from 15- to 16-day Fisher rat embryos by cutting the skin, mincing it, and exposing the pieces to trypsin (0.25%) and collagenase (1 mg/ml) for 30 min at 37°C. Confluent primary cultures were trypsinized and the cells were frozen in aliquots. Cells were thawed and allowed to grow to 50% confluence prior to

Inc.

INTRODUCTION

DNA tumor viruses have evolved functions to stirnulate the host cells to enter DNA synthesis and thereby sustain viral replication. In nonpermissive cells this interference with normal cellular functions may sometimes lead to tumor transformation. The early region of i To whom reprint

AND

requests should be addressed. 143

0014-4827/90 All

$3.00

Copyright 0 1990 by Academic Press, Inc. rights of reproduction in any form reserved.

144

MARTENS

transfection [35] using 10 c(g of polyomavirus DNA and 1 pg ofpSV2neo per 60-mm dish. At 48 h after addition of the DNA precipitate, the cultures were split 15 and G418 (GIBCO, Paisley, Scotland) was added at 0.4 mg/ml. At 3 weeks after transfection, the colonies (lo-15 in one dish) were trypsinized and all cells were pooled into one culture. The cells were counted at each subsequent passage and lo5 cells were reseeded per 90-mm dish. Dexamethasone (DEX, Sigma Chemical Co.) was stored as a 1 mM stock solution in ethanol and was added to the growth medium as indicated. Autoradiography of [sH]thymidine-labeled cells. Cells were plated on glass coverslips and labeled for 30 h with [3H]thymidine at a concentration of 1 rCi per milliliter. The coverslips were dipped in photographic emulsion (Kodak, Nuclear Track Emulsion, Type NTBS) and exposed for 3 days. Analysis of tumorigenicity. Cells were inoculated subcutaneously in 4-week-old Fisher rats irradiated with 400 rad. The rats were examined weekly for development of tumors. Two-dimensional gel electrophoresis. Cells were labeled for 16 h in methionine-free medium supplemented with 50 &i [35S]methionine/ ml. Total cell extracts were prepared and two-dimensional electrophoresis was performed essentially as described by O’Farrell [20]. Cytoskeletal proteins were purified by washing cells in 0.15 M NaCI, 3 mM MgCl,, 0.5% NonidetP40, 10 mA4 Tris-HCl, pH 7.5. Tropomyosins were extracted from cytoskeleton preparations by the method of Matsumura et al. [16].

ET AL.

S11

s14

RESULTS

Passage-Dependent Tumorigenicity of Rat Embryo Fibroblasts Transfected with Polyomavirus DNA Secondary rat embryo cells were cotransfected with a plasmid harboring a replication origin-deficient mutant of polyomavirus, dZlO23 [ 141 and a neomycin resistance marker. At 4 weeks after transfection, neomycin-resistant colonies were trypsinized and pooled into a mass culture (REF/py cells). Alternatively, individual colonies were isolated and established into cell lines (clones Sl-S16). The growth properties of the REF/py cells and the different clonal cell populations were studied in detail during subsequent in vitro culture. At early passages after the first trypsinization, the REF/py cells exhibited a flat morphology, whereas cells at later passages were smaller and more refractile. The change in morphology occurred between 2 and 3 months after transfection for most clones. Figure 1 shows the morphology of two clonal populations 12 days after plating the cells at low cell density at Days 52 and 98. Both clones grew to higher cell densities at the later passage (compare A and B; and C and D). In order to examine whether the changes in morphology and growth rate correlated with the development of a more tumorigenic potential, early (30-day) and late passage (95day) REF/py cells (derived from a pooled population of clones) were inoculated in syngenic rats. The same batch of cells, frozen at different passages, was used. A more than lO,OOOfold difference in tumorigenicity was observed between the early and late passage REF/py cells (Table 1). The ability of the cells in different clones to grow in semisolid medium was examined. When these clones

FIG. 1. Effect of in vitro passage on the morphology of polyomavirus-transfectedcells. Cells of subclones S12 (A, B) and S14 (C, D) were seeded at a density of 100 cells/g-cm dish 52 days (A, C) and 98 days (B, D) after transfection. Cells were grown in 5% fetal calf serum for I2 days before fixing and staining.

were first tested (at 52 days after transfection), only 2 out of 15 clones could form visible colonies in soft agar (Fig. 2, gray bars). During subsequent culture (between 52 and 98 days after transfection), 9 additional clones acquired the ability for anchorage-independent growth. The two clones that were able to grow in agar already at the earlier passage, increased their plating efficiency. Thus, 11 out of 15 clones tested developed an increased ability to grow in semisolid medium during the 46 days between the two time points. We excluded the possibility that the increase in tumorigenicity was due to a higher level of expression of the polyomavirus early region in late passage cells. Thus, the level of early region transcription was not found to be different in REF/py cells at early and late time points (data not shown). Effect of Serum and Glucocorticoid Hormones on the Growth Properties of REF/py Clonul Cells Our results demonstrate that early passage REF/py cells-but not late passage-grow inefficiently in semisolid medium. We addressed the question of whether the addition of growth factors to early passage cells could induce anchorage-independent growth. We found that increasing the concentration of fetal calf serum to 25%

POLYOMAVIRUS

TABLE Tumorigenicity

of Polyoma-Transformed

1

Cells at Different Times after Transfection No. of animals with tumors

Time after transfection

DNA used for transfection

145

TRANSFORMATION

(days)

lo2

lo3

lo*

lo5

106”

Pwt

30 95

o/2 2/2

o/2 212

o/2 212

o/2 212

012 212

pbclo51

30 52 95

o/2 o/2 l/2

o/2 o/2 212

012 012 2/2

o/2 o/2 w

o/2 w 212

a No. of cells injected in each animal.

did not induce anchorage-independent growth at Day 52 after transfection (data not shown). We also tested dexamethasone, classified as a class III growth factor for human diploid fibroblasts [21], for its ability to induce anchorage-independent growth. Addition of 0.5 pA4 DEX to the soft agar medium resulted in a plating efficiency of 5-10% for 11 out of 13 clones that did not grow in unsupplemented agar medium at Day 52 (Fig. 2A, black bars). As described in the previous section, most REF/py cells were able to grow in semisolid medium at 98 days after transfection (Fig. 2B). Addition of DEX generally increased the plating efficiency of these late passage cells about twofold.

A

Changes in the Pattern of Protein Synthesis We wished to document specific changes in gene expression occurring during in vitro propagation. Furthermore, the observation that DEX treatment leads to an increasing plating efficiency in soft agar of early passage cells led us to examine whether common phenotypic changes resulted from in vitro passage and DEX treatment. We addressed this question by subjecting total cell extracts to two-dimensional gel analysis. Several proteins showing interesting patterns of expression were identified: three of these are indicated in the gel sections shown in Fig. 3. DEX treatment altered the expression of several proteins at both early (Day 35) and late (Day 95) passage.

dex

60

n

+ dcx

S4

S5

? 5 ‘C ‘-

0 Sl

S2

S3

S6

S7

S9 SIOSII

S12Sl3Sl4Sl5Sl6

J 0 SI

S?

S3

S4

S5

S6

S7

S9 SIOSII

SI2Sl3Sl4

SISSl6

Subclone

FIG. 2. Plating efficiency in semisolid medium of clones of polyomavirus-transfected cells. Cells (5 X 103) were seeded in 0.33% agarose at 52 days (A), or at 98 days after transfection (B). The cells were plated in the presence or absence of 0.5 pM DEX. The frequency of colonies with a diameter of more than 0.1 mm was recorded at 2 weeks after plating.

FIG. 3. Changes in the pattern of protein synthesis during in uitro culture. Polyomavirus-transfected cells were labeled with [?S]methionine for 16 h and labeled proteins were extracted and analyzed by twodimensional gel electrophoresis. Cells were labeled at 30 days (top) and 95 days (bottom) after transfection in the presence or absence of 0.5 &4 DEX as indicated. Spots 1-3 represent examples of proteins whose expression changed during culture and/or after hormone treatment. Only the portion of the gels showing 28- to 45-kDa proteins with pIs between 4.5-6 are shown. The thick spot in the top center is actin.

146

MARTENS

ET AL.

extraction of isolated cytoskeletons (Fig. 4E) confirmed that spot 3 was identical to tropomyosin 2 (tm2), using the terminology of Leavitt et al. [ 121. The tm2 level was found to be lower in early passage REF/py cells than in untransformed REF cells. In late passage REF/py cells, tm2 was found to be further reduced compared to that in early passage cells. Treatment of REF/py cells with DEX resulted in a reduction of tm2 in cells labeled at both 1 and 3 months after transfection (Fig. 4) but not in untransformed cells (data not shown). Three other tropomyosin forms were identified (indicated by circles in Fig. 4). Of these, tml was also reduced in polyomavii-us-transformed cells, whereas tm5 expression was slightly increased and tm4 unaffected. FIG. 4. Two-dimensional gel electrophoresis of polypeptides from rat embryo fibroblasts (REF cells; A, B, and E) and from T24Ha-rus-transformed REF cells (C, D). The figure shows two-dimensional gel profiles of total protein extracts (A, C), detergent-washed cytoskeletons (B, D), and heat-extracted cytoskeletons (E). The boxes (l-3) highlight the same spots that are indicated in Fig. 3. The circles show the tropomyosin forms tml, tm2, tm4, and tm5.

Two such examples are the spot 1 (27-kDa) and spot 2 (38-kDa) proteins. The expression of the spot 2 protein was also found to change during in vitro culture. Thus, the spot 2 level was higher at Day 35 than at Day 95. Comparison of the abundance of the spot 2 protein in untransfected REF cells and in T24-Hu-rus-transformed cells showed that this protein was present in much lower levels in the transformed cells (Fig. 4), consistent with the notion that spot 2 represents a transformation-sensitive protein. Other proteins were identified that were differently expressed at early and late passages. The abundance of one such protein, designated spot 3 (37 kDa), decreased during propagation of REF/ py cells (Fig. 3). Interestingly, the expression of the spot 3 protein also decreased after DEX treatment. The abundance of at least three other proteins was similarily altered by passage and DEX treatment. We did not, however, observe that in vitro passage and DEX treatment induced identical changes in the pattern of gene expression-one example of this is the pattern of expression of the spot 2 protein. We were interested in revealing the nature of the spot 3 (37-kDa) protein, the expression of which decreased both by DEX treatment and during passage. This polypeptide was present in high amounts in untransformed REF cells but was almost absent in a REF-derived T24Ha-r-as-transformed cell line (Figs. 4A and 4C). The protein could be demonstrated to be a component of the cytoskeleton of these cells (Figs. 4B and 4D) and was putatively identified as a tropomyosin isoform based on its electrophoretical migration properties. The spot 3 protein was not labeled by [14C]proline (data not shown), an amino acid absent from all known tropomyosins. Heat

Large T-antigen Is Not Required for Maintenance of Transformation at Late Passages We wished to investigate whether the observed changes in cellular growth properties result in a decreased requirement for viral functions. To test whether large T-antigen is required for growth of polyomavirustransformed cells at late passages, we isolated clonal cell lines of rat embryo fibroblasts expressing the tsA mutant of polyomavirus [6]. The tsA mutation leads to an amino acid substitution at the C-terminus of large Tantigen resulting in a thermolabile protein [33], whereas the middle and small T-antigens are unaffected by the mutation [30]. The growth properties of one clonal line, ts A2, is shown in Fig. 5. At 6 weeks after transfection, the ts A2 cells were able to grow at 33°C with a popula8A

A 33°C 0 39°C n 39°C +dcx

6-

c ‘S

B

Days

after

temperature

shift

FIG. 5. Growth of one clonal cell line transfected with the tsA mutant of polyomavirus at the permissive and nonpermissive temperature. The growth rate was analyzed at 6 weeks (A) and 20 weeks (B) after transfection. Cells were grown at 33°C (A) or at 39°C (0, W) in the presence (H) or absence (A, 0) of 0.5 &f DEX. The cells were counted and replated at 5 X 10’ tells/6-cm dish at each passage.

POLYOMAVIRUS

tion doubling time of approximately 60 h, whereas growth was arrested after shifting the incubation temperature to 39°C. Cells did grow at the higher temperature, however, if 0.5 PM DEX was added to the medium, which showed that the arrested growth was not simply due to intolerance to the high temperature. The ts A2 cells were then grown at 33°C for 14 weeks. During this period the growth rate of the ts A2 cells increased about threefold. When the incubation temperature was now raised to 39”C, the cells continued to grow exponentially in the absence of hormone (Fig. 5B). We conclude that large T-antigen is no longer required for growth of the ts A2 cells after 5 months in culture. Similar results were obtained with two other clones. In order to examine whether large T-antigen is required for tumorigenicity of polyomavirus-transformed cells at late passages, REF cells were transfected with a plasmid, pbcl051, encoding only the polyomavirus middle and small T-antigens. Polyclonal cultures of neomycin-resistant cells were established. As described previously, these cells did not grow in the absence of DEX during a period between 30 and 50 days after transfection. At later passages, growth was observed also in the absence of hormone. The plating efficiency of REF/ bc1051 cells in soft agar increased from 8 to 62% between 30 and 95 days after transfection (the agar medium was supplemented with DEX to allow growth at early time points). When late passage cells (Day 95) that no longer required dexamethasone for growth were inoculated in syngenic rats, they readily formed tumors (Table 1). It appears, then, that at 3 months after transfection, the large T-antigen is not needed for expression of the malignant phenotype. The Fraction of Nongrowing during in vitro Culture

Cells Decreases

The phenotypic changes observed during culture may be due to a selection of cells that are large T-antigenindependent and tumorigenic. Since tumorigenic cells are established in the apparent absence of a period of crisis, one may predict that nongrowing cells continuously arise within the population whereas the majority of the cells would grow at any one time point. To test this prediction, cells were labeled with [3H] thymidine for 30 h at different time points after transfection and subjected to autoradiography. At 52 days after transfection, 4.7% of the S14 were not labeled by the [3H]thymidine pulse (Table 2). This fraction had decreased to 1.5% at Day 95. On the other hand, only 2.0% of the cells of another clone, S12, that did grow in soft agar at Day 52 did not incorporate [3H]thymidine at this point. We conclude that we in fact do observe that a higher fraction of the REF/py cells are not growing (or very slowly growing) at the earlier passages, consistent with an ongoing selection process in these cell populations.

147

TRANSFORMATION

TABLE 2 [3H]Thymicline Incorporation at Different Time Points after Transfection Cell type REFlpbcl051

REF/pwt

REF/pwt

S3

REF/pwt

S12

REF/pwt

S14

Days after transfection 41 52 95 41 52 95 52 98 52 98 52 98

Unlabeled cells (%I 9.2 6.8 0.4 3.8 2.5 1.1 3.1 1.4 2.0 1.4 4.7 1.5

DISCUSSION We isolated rodent embryo fibroblasts with stably integrated copies of a replication origin-defective polyomavirus by means of selection for antibiotic resistance. We used this procedure to study the phenotype of cells that were not isolated on the basis of their transformed morphology. Furthermore, we wished to minimize the possibility that phenotypic changes occurred during culture due to viral replication. Our observations show that cells isolated by this procedure are not fully transformed at early passages. During growth in vitro, the cells acquire the ability to grow in soft agar and to form tumors. The altered phenotype is not likely to result from viral replication or any changes in viral integration patterns (we observe less than five virus copies/cell; data not shown) and is not due to an increased level of viral gene expression during passage as evidenced by RNase protection analysis of early transcripts (data not shown). We only observe this phenomenon of progression using embryonic (primary or secondary cultures) cells; transfection of established FR3T3 cells with polyomavirus plasmids results in antibiotic-resistant colonies that grow in soft agar also at early time points after transfection (not shown). Our observations of progression of polyomavirustransformed embryo fibroblasts are in accord with previous observations. Thus, Vogt and Dulbecco [34] showed that some foci developing after infection of primary rat cells with polyomavirus did not have the characteristic morphology of polyomavirus-transformed cells. These cells developed into fully transformed cells during culture in vitro. It was suggested that an additional step was required for transformation. Bouchard and Bastin [2] isolated transformed cells by neomycin selection and found that early passage cells were incompletely transformed and only weakly tumorigenic. Multiple changes

148

MARTENS

have also been reported for the development of a fully transformed phenotype after transfection of primary rat hepatocytes with SV40 DNA [36]. Cuzin et al. [5] also observed phenotypic changes during culture of polyomavirus-transfected rat embryo fibroblasts. In these experiments the viral small T-antigen was found to be required only during initial stages for focus formation in monolayer culture. Our finding that large T-antigen is not required for the growth of late passage transformed cells is consistent with our previous results but has few other precedents in the literature. We previously found that REF cells transfected with a large T-antigen-deficient polyomavirus mutant can be established in the presence of glucocorticoid hormones and that these cells develop hormone independence at late passages in culture [15], consistent with the results of the present study. Rassoulzadegan and co-workers [24] reported that large T-antigen is needed for growth of REF cells transfected with a plasmid encoding a temperature-sensitive large T-antigen. These were, in contrast to the cells of the present study, only expressing large T-antigen. Jat and Sharp [lo] and Crook et al. [4] reported that SV40 large T-antigen and the human papilloma virus E7 protein are required for maintenance of transformation of embryonic rodent cells. The reason for the discrepancies between these published results and our own is unclear, but may be explained by the fact that different combinations of oncogenes were used. The results of the present and previous studies showing progression of polyomavirus-transformed REF cells can be explained by hypothesizing that the growth of polyomavirus-transfected rodent embryo fibroblasts can be divided into two distinct phases. During the first phase (“the establishment phase”), the cells are not yet established and are dependent on the expression of large T-antigen (Fig. 5) or on the addition of glucocorticoid hormones. Such cells may not yet have entered the immortalized state and are therefore not fully susceptible to transformation by middle T-antigen. It has been amply demonstrated that nonestablished rodent cells are resistant to transformation by middle T-antigen [23] and p2l(val)ras [17]. At late passages, the cells have entered a second phase where a large fraction of the cells in the population are established (i.e., in this context are independent of large T-antigen or glucocorticoid hormones for growth). We observe a temporal correlation between the development of anchorage-independent growth and large T-antigen-independent growth; both events appear to occur between Days 50 and 90 after transfection for most clones (Ref. [15] and Fig. 2 in this communication). The length of this phase appears to vary among different clones: some clones grow in agar already at Day 52 whereas others do not at Day 98. Establishment is a fairly frequent phenomenon in cultures of rodent fibroblasts that have not been exposed to tumor viruses [31]. We would like to suggest a model for

ET AL.

polyomavirus-mediated immortalization of rodent fibroblasts where the role of the virus is to extend the life span of the cells, thereby increasing the probability of immortalization. Both large and middle T-antigens have mitogenic activities [13, 271 that may stimulate the growth of the cells resulting in an extended life span. The large T-antigen has the strongest immortalizing activity [24], but middle T-antigen alone has also been reported to immortalize rat embryo cells at a low frequency [3]. An increased life span is in fact what is observed after transformation of human fibroblasts by SV40 [9]. During the postulated “establishment phase,” about 4% of the cells can be demonstrated to be growth arrested (or grow very slowly) at any specific time point. We would like to suggest that these cells have reached the end of their life span and that their growth cannot be further stimulated by the viral early proteins. Further work is required to prove this point, however. An indirect role of the virus in the immortalization process is in accord with the fact that polyomavirus is a lytic virus. It is difficult to understand how the ability to immortalize cells is beneficial to the spreading of the virus. Human polyomaviruses such as JC and BK viruses clearly spread efficiently in their host population without the ability to immortalize human cells. Glucocorticoid hormones can substitute for immortalizing oncogene products, provided that the cells express transforming gene products [15, 371. We have not observed any mitogenic effect by these hormones on untransformed rat embryo cells although others have reported that glucocorticoids are growth factors in certain defined media [21]. In this study we demonstrate that DEX has the capacity to induce anchorage-independent growth of early passage polyomavirus-transfected cells. It is interesting to compare these effects with those of the oncogene v-erbA. This gene encodes a mutated version of a receptor for thyroid hormone that has lost its ability to bind hormone [25]. In analogy with glucocorticoid hormones, the v-erbA product has no transforming ability on its own, but cooperates with other oncogene products such as the tyrosine kinases or the v-Ha-ras protein [ll]. We do not understand the molecular basis of the glucocorticoid stimulation, although it seems likely that regulation of cellular gene expression is involved. Addition of DEX to polyomavirus-transformed cells does not alter the expression of polyomavirus early transcripts [15]; neither does it affect middle T-antigenassociated phosphorylation [22]. The effect exerted by the steroid hormone could not be mimicked by the addition of serum. Our 2D gel analysis of the pattern of gene expression revealed that the expression of some proteins was similarly altered by DEX and passage in vitro, suggesting that the steroid hormone may be able to affect the activity of some genes that are expressed differently in ageing and immortal cells. Interestingly, one protein that was downregulated both by glucocorticoids and dur-

POLYOMAVIRUS

ing in vitro culture was tropomyosin 2. The major high affinity actin-binding forms (1 and 2) of tropomyosin are known to be downregulated in transformed cells [El]. We observed an altered pattern of tropomyosin synthesis typical of transformed cells in late passage cells, consistent with the development of a more transformed phenotype in these cells. We considered the possibility that the development of aneuploidi occuring during culture may lead to the observed phenomenon of tumor progression. Flow cytometric measurements showed, however, that the DNA content of our transfected cells did not change during culture (data not shown). We regard the possibility that specific chromosomal rearrangements or mutations may occur that lead to immortalization as unlikely, since we find that the majority of subcloned cells are able to progress to anchorage independence during the limited time period of 45 days. The changes leading to progression thus cannot be very rare. We considered gene amplification as a possible mechanism to explain our findings. A correlation between amplification of the myc oncogene and a high degree of tumorigenicity has been observed both in polyomavirus-transformed mouse cells [28] and in human tumor cells [7]. However, we have not observed an increase in c-myc copy number between early and late passage cells (I.M., unpublished data). In summary, the development of the fully transformed phenotype of polyomavirus-transfected cells occurs in more than one step. We would like to suggest that the mitogenic properties of the polyomavirus T-antigens may result in stimulation of cellular growth leading to an increased life span. During this period of extended growth, cells become “immortalized” and susceptible to transformation by middle T-antigen. We thank Professor Goran Magnusson for generous support and for helpful comments on this manuscript. This project was supported by grants from the Swedish Cancer Society to S.L. and to Goran Magnusson and by a grant from Gustav V:s jubileumsfond to S.L. T.R. was supported by a fellowship from the Cancer Research Institute and the Concern Foundation.

REFERENCES 1.

Barrett, J. C., and Ts’o, P. 0. P. (1978) Proc. Natl. Acad. Sci. USA 75,376l.

2.

Bouchard, L., and Bastin, M. (1986) Virology 155, 1. Cowie, A. J., de Villiers, J., and Kamen, R. (1986) Mol. Cell. Biol. 6,4344.

3. 4.

Crook, T., Morgenstern, EMBO J 8,513.

5.

Cuzin, F., Rassoulzadegan, Cells 2,109.

6.

Fried, M. (1965) Proc. Natl. Aead. Sci. USA 53, 486.

Received September 6,1989

J. P., Crawford,

TRANSFORMATION

18. 19.

Fukumoto, M., Shevrin, D. H., and Roninson, I. B. (1988) Proc. Natl. Acad. Sci. USA 86,6846. Hendricks, M., and Weintraub, H. (1981) Proc. Natl. Acad. Sci. USA 78.5633. Huschtscha, L. I., and Holliday, R. (1983) J. Cell Sci. 63,77. Jat, P. S., and Sharp, P. A. (1989) Mol. Cell. Biol. 9,1672. Kahn, P., Frykberg, L. C., Brady, I., Stanley, H., Beug, B., VennStrom, and Graf, T. (1986) Cell 45,349. Leavitt, J. G., Latter, G., Lutomski, L., Goldstein, D., and Burbeck, S. (1986) Mol. Cell. Biol. 6,272l. Liang, T. J., Carmichael, G. G., and Benjamin, T. L. (1984) Mol. Cell. Biol. 4,2774. Luthman, H., Nilsson, M.-G., and Magnusson, G. (1982) J. Mol. Biol. 161,533. Martens, I., Nilsson, M., Magnusson, G., and Linder, S. (1988) Proc. Natl. Acad. Sci. USA 85,557l. Matsumura, F., Lin, J.-C., Yamashiro-Matsumura, S., Thomas, G. P., andTopp, W. C. (1983) J. Biol. Chem. 258,13954. Newbold, R., and Overell, R. W. (1983) Nature (London) 304, 648. Nilsson, S. V., and Magnusson, G. (1983) EMBO J2,2095. Noda, T., Satake, M., Yamaguchi, Y., and Ito, Y. (1987) J. Virol.

20. 21.

O’Farrell, P. H. (1975) J. Biol. Chem. 250,4007. Phillips, P. D., and Cristofalo, V. J. (1988) Enp. Cell Res. 175,

22. 23.

Raptis, L., and Bolen, J. B. (1989) J. Virol. 63,753. Rassoulzadegan, M., Cowie, A., Carr, A., Glaichenhaus, N., Kamen, R., and Cuzin, F. (1982) Nature 300,23. Rassoulzadegan, M., Naghashfar, Z., Cowie, A., Carr, A., Grisoni, M., Kamen, R., and Cuzin, F. (1983) Proc. Natl. Acad. Sci. USA 80,4354. Sap, J., Munoz, A., Damm, K., Goldberg, Y., Ghysdael, J., Leutz, A., Beug, H., and Vennstrom, B. (1986) Nature (London) 324, 635.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

61.2253.

396.

24.

25.

26. 27. 28.

Schaflhausen, B. (1982) Crit. Reu. Biochem. 13,215. Schlegel, R., and Benjamin, T. L. (1978) Cell 14,587. Schwab, M., Ramsay, G., Alitalo, K., Varmus, H. E., Bishop, J. M., Martinsson, T., Levan, G., and Levan, A. (1985) Nature (London) 3 15,345.

29. 30.

Southern, P. J., andBerg, P. (1982) J. Mol. Appl. Genet. 1,327. Thomas, T., Vollmer, P., and Volk, W. R. (1980) J. Virol. 37, 1094.

31. 32.

Todaro, G. J., and Green, H. (1963) J. Cell Biol. 17,299. Treisman, R., Novak, U., Favaloro, J., and Kamen, R. (1981) Nature (London) 292,595. Trejo-Avila, L., Gaudray, P., and Cuzin, F. (1981) Virology 114, 501.

33. 34. 35.

L., and Banks, L. (1989) 36.

M., and Lemieux,

149

L. (1984) Cancer 37.

Vogt, M., and Dulbecco, R. (1963) Proc. Natl. Acad. Sci. USA 49, 171. Wigler, M., Pellicier, A., Silverstein, S., Axel, R., Urlaub, G., and Chasin, L. (1979) Proc. Natl. Acad. Sci. USA 76,1373. Woodworth, C. D., Kreider, J. W., Mengel, L., Miller, T., Meng, Y., and Isom, H. C. (1988) Mol. Cell. Biol. E&4492. Yamashita, T., Kato, H., and Fujinaga, K. (1988) Znt. J. Cancer 42,930.

Complete transformation of embryonal rat fibroblasts by polyomavirus occurs during passage in vitro.

The tumorigenicity of secondary rat embryo fibroblasts transfected with a plasmid harboring a replication origin-defective polyomavirus was found to i...
3MB Sizes 0 Downloads 0 Views