Mutation Research, 243 (1990) 291-298 Elsevier

291

MUTLET 0328

Oncogene alterations in in vitro transformed rat tracheal epithelial cells M a r c J. M a s s a, N a n c y S. S c h o r s c h i n s k y b, Jessica A. L a s l e y b, D i a n e K. B e e m a n a a n d S t e p h e n J. A u s t i n b "Respiratory Carcinogenesis Group, Carcinogenesis and Metabolism Branch ( MD-68 ), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711 (U.S.A.) and ~Environmental Health Research and Testing Inc., Research Triangle Park, NC 27711 (U.S.A.) (Accepted 20 November 1989)

Keywords." Oncogenes; Respiratory tract; Cell transformation; (Rat)

Summary 10 derivations of rat tracheal epithelial (RTE) cells, including normal cells, normal primary cultures, 7 tumorigenic cell lines and 1 nontumorigenic cell line transformed in vitro by treatment with 7,12-dimethylbenz[a]anthracene (DMBA), benzo[a]pyrene (BP) and/or 12-O-tetradecanoylphorbol-13-acetate (TPA) were examined for oncogene alterations. No abnormalities of Ha-ras or Ki-ras were seen that were suggestive of amplification, rearrangement or the presence of RFLPs. Analysis of specific-point mutations in Ha-ras using Pst I digestion (codon 12, GGA to GCA) or Ha-ras and Ki-ras using Xba I (codon 61, CAA to CTA) were negative. In one cell line derived by DMBA treatment, changes in the c-myc restriction digest pattern were seen after incubation with Bam HI and Hind III. Northern analysis revealed consistent differences between normal and transformed cells when probed with Ha-ras; c-myc expression was of low intensity, and the expression of Ki-ras could not be detected. Transfection of RTE cell DNAs into NIH/3T3 cells did not result in the appearance of morphologic transformants. The studies suggest that Ha-ras or Ki-ras codon 61 A to T transversions (CAA to CTA) are not associated with the immortal/tumorigenic phenotype in RTE cells transformed by DMBA or TPA, and are in contrast to results reported in some other biological systems.

Activated oncogenes capable of transforming recipient mammalian cells by transfection have been Correspondence: Dr. Marc J. Mass, Respiratory Carcinogenesis Group, Carcinogenesis and Metabolism Branch (MD-68), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711 (U.S.A.). Abbreviations: BP, benzo[a]pyrene; DMBA, 7,12-dimethylbenz[a]anthracene; RFLP, restriction fragment length polymorphism; RTE, rat tracheal epithelial; TPA, 12-O-tetradecanoylphorbol-13-acetate.

observed in 10-20% of human malignancies of various histologic types (Bishop, 1983; Barbacid, 1985). In some specific cancers such as those of the exocrine pancreas and in follicular thyroid cancers, almost all tumors contain a mutated activated ras oncogene (Almoguera et al., 1988; Lemoine et al., 1988). The proportion of chemically induced rodent cancers with activated oncogenes, specifically ras genes, is also variable with respect to particular sites of tumor induction (Zarbl et al., 1985; Watatani et al., 1989). Establishment of a causal

0165-7992/90/$ 03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

292

relationship between oncogene activation by mutation and neoplasia could conceivably be determined by analysis of oncogene mutation/activation at early stages in the neoplastic process. In the present study, we analyzed DNA from a number of carcinogen-induced immortalized/tumorigenic RTE cell lines, and a nontumorigenic cell line induced by treatment with a tumor promoter in vitro. Because these particular cell lines had never been propagated in vivo, they could be considered to represent relatively early stages in tumor development. We sought to characterize some changes in oncogenes Ha-ras, Ki-ras and c-myc that might be present in these cell lines. The presence or absence of oncogene changes at early stages of neoplastic development could provide evidence to support or refute the role of these genes in the establishment of neoplastic stages. We report these RTE cell lines do not contain transfectable genes that transform N I H / 3 T 3 cells, and rarely have Haras, Ki-ras or c-myc alterations in that no Ha-ras or Ki-ras restriction polymorphisms were found, and only one RFLP in the c-myc gene was found in 1 of 8 cell lines tested. These results suggest that specific alterations in the oncogenes tested are not a common feature of RTE cells in early transformed stages. A preliminary report of these data has been presented (Mass et al., 1989b).

10 -5) was derived from treatment o1' primary RTE cells with TPA (Steele et al., 1984). Cell lines BP6 and DL3, which are tumorigenic, were derived from treatment of a nontumorigenic, TPA-induced cell line with BP or 7,8-dihydro-7,8-dihydroxy-BP, respectively (Steele et al., 1980).

Preparation of nucleic acids. DNA and RNA were isolated simultaneously an isopyknic CsCI gradient containing guanidine hydrochloride generated by ultracentrifugation. DNA was subjected to proteinase K/sarcosyl/EDTA digestion, RNAase treatment, multiple phenol/chloroform extractions, ethanol precipitation and dialysis. RNA was further purified with sodium acet a t e / a m m o n i u m acetate and ethanol precipitation. Ratios of the absorbance of the nucleic acids at 260 and 280 nm were usually 1.8 or greater. The integrity of the nucleic acids was also checked after electrophoresis through 0.8% agarose.

Materials and methods

Preparation of nucleic acid blots. DNA was subjected to appropriate restriction-enzyme digestions by overnight incubations, and then electrophoresed through 0.8% agarose. Agarose gels were stained with ethidium bromide and photographed. DNA was released from agarose gels and transferred to nitrocellulose. Expression of oncogenes was measured using Northern blot analysis of RNA using formaldehyde-containing 0.8% agarose gels.

Source of cells. DNA or RNA preparations from epithelial cells of normal rat tracheas were obtained from 20-30 male Fischer 344 rats, 8-12 weeks in age (Charles River, Inc., Raleigh, NC). DNA and RNA from primary tracheal epithelial cells were obtained by extracting nucleic acids from RTE cells grown on a monolayer of ~-irradiated Swiss 3T3 cells, as described (Gray et al., 1983). After 7-10 days dead 3T3 cells were removed with 0.002% E D T A (Thomassen et al., 1983) and the remaining RTE cells were removed by trypsinization. Tumorigenic clonally derived cell lines D3b, B3j, B3k, B2o and D2z were established by treatment of primary RTE cells with DMBA (Steele et al., 1988). Nontumorigenic cell line TM5 (originally named

Hybridizations. Nitrocellulose filters containing samples were prehybridized and hybridized for 24 h at 42°C in solutions containing 40% deionized formamide, 5 x Denhardt's solution, 100 ~g/ml salmon sperm DNA, 0.1% SDS, and 5 x SSPE. Filters were hybridized with 5-25 ~Ci of oncogene probes 32p-labeled by oligonucleotide-primed, Klenow fragment-directed replication of the template probe (Multiprime; Amersham, Arlington Heights, IL) or by nick-translation; the specific activities of the probes varied from 2 × 108 to 3 × 10 9 dpm//zg. After hybridization, filters containing DNA were washed twice in 2 × SSC/0.1% SDS for 10 min each wash, at ambient temperature; the filters were then washed 4 times in 1 x SSC/0.1%

293 SDS for 30 min each wash at 37°C; high stringency washes where required were performed for 30 min using 0.1 × SSC/0.1% SDS at 55°C. Filters were autoradiographed with K o d a k X - O M A T XAR-5 film at - 7 0 ° C in the presence of intensifying screens; exposures varied from 24 h (for most Southern blots) to a week (for Northern blots).

Probes. The following probes were obtained f r o m Oncor, Gaithersburg, MD: c-myc (1.8-kb h u m a n c-myc, third exon); v-Ha-ras (0.73 kb from H a r v e y murine sarcoma virus); v-Ki-ras (0.618 kb f r o m Kirsten murine sarcoma virus). Xba I polymorphisms of Ha-ras were also determined with plasmid pNMU-1 (ATCC, Rockville, MD). In addition, some blots were probed with 3,-actin (0.47-kb chick 7-actin) as a positive hybridization control. Transfection assays. N I H / 3 T 3 cells were obtained f r o m Dr. J.C. Barrett (National Institute of Environmental Health Sciences, Research Triangle Park, NC). The calcium phosphate precipitation method was used for transfection (Wigler et al., 1979). N I H / 3 T 3 cells were seeded at 1.5 x 10 ~ cells per 100-mm dish; 18-20 h after seeding 40 #g of R T E cell D N A was added to each dish in a H E P E S / N a C l - c a l c i u m phosphate buffer and incubated for 18-20 h. Dishes were rinsed of the precipitate. Cells on dishes were grown for 3-5 weeks, stained and scored for the presence o f transformed foci. Results Oncogene amplification, rearrangement, restriction polymorphisms Ha-ras. Restriction-enzyme digests o f RTE cell D N A with Xba I (Fig. 1), which detects a substitution o f T for A in the second position of the codon 61 C A A sequence exhibited identical bands in the 1.5-kb to 4.7-kb region, confirming the absence of C A A to C T A transversions in codon 61 (Vuorio et al., 1988). Restriction-enzyme digests of normal cell and cell line D N A with Pst I were identical.

Marker (kb] 23.1 9.4 6.6 4.4 2.3 2.0 1.4

Fig. 1. Southern-blot analysis for Ha-ras gene digested with Xba I. Probe used was pNMU-1. N, DNA extracted from norreal rat tracheal epithelia/cells; 1o, DNA extracted from primary RTE cell cultures. The molecular weight markers are 3ap_ labeled k and thX174 DNA digested with Hind III and Hae 111, respectively. This enzyme detects the presence of a change of the second G to C in the G G A sequence of codon 12. The Ha-ras gene was probed after restrictionenzyme digests of either Eco RI, Hind III and Bam H I to examine the possible presence of rearrangements, or restriction fragment length polymorphisms (not shown). No significant changes in copy number were discerned between normal RTE cell D N A , D N A from primary cultures, or D N A f r o m 8 cell lines, indicating that amplifications of Ha-ras were not present, as determined by visual comparisons of intensities of bands and ethidium bromide staining intensities. No differences in the size of restriction enzyme-generated fragments produced by the above restriction enzymes were noted between normal cells and those of cell lines obtained f r o m chemical treatments.

Ki-ras. Xba I digests of RTE cell D N A exhibited identical numbers of bands regardless of cell source, indicating the absence of C A A ~ C T A transversions in codon 61 (Fig. 2). Normal and 1 o

294

Barn HI

'I¢o t7

b-

m

m

N

JCI

.~

Hind I!l

Eco RI O

rn

Marker

(kb}

23.1

-

n

o

m

Marker (kb) 23.1 9.4

9.4 6.6

6.6 4.4

~

.....

t

2.3 2.0

2.3 2.0 ~i~iiil

1.4 1.4

Fig, 2. Southern-blot analysis of Ki-ras gene digested with Xba I. Probe used was v-Ki-ras. DNA from normal RTE cells scraped from tracheas and primary cultures are not shown but exhibit identical autoradiographic patterns. The molecular weight markers are 3zP-labeled X and 4~X174 DNA digested with Hind IlI and Hae 111, respectively.

cell D N A (not shown) exhibited identical restriction patterns to those described above. After accounting for variations in D N A concentration applied to agarose gels as assessed by ethidium bromide staining, v-Ki-ras probing of rat tracheal cell D N A digested with Eco RI and Bam H I indicated similar Ki-ras copy numbers (not shown). c-myc. Restriction enzymes Bam H I , Eco RI and Hind III were used to detect alterations in cmyc (Fig. 3). No significant differences in copy n u m b e r were detected after D N A content normalization as assessed by visual comparison of ethidium bromide staining intensities. Normal R T E cells and RTE cell lines had similar hybridiza-

Fig. 3. Southern-blot analysis of c-myc gene digested with Barn HI, Eco R1, or Hind Ill. Cells from normal rat tracheas, primary cultures, cell lines BP6, B3k, DL3, TM5 exhibited restriction patterns identical to those from D3b, B3j and B2o (shown). The molecular weight markers are X2P-labeled ), and 4~X174 DNA digested with Hind Ill and Hae Ill, respectively.

tion patterns, except for the cell line D2z, which had an altered pattern consisting of an extra band migrating more slowly than those seen in other cell lines; this aberration was seen with restriction enzymes Barn H I and Hind IIl, but not with Eco RI. Northern analysis Total RNA was isolated from normal tracheal epithelial cells, primary cultured RTE cells, and all cell lines. Ha-ras expression was not detectable in R N A isolated from normal and primary-derived R T E cells. In contrast, in all samples except RNA f r o m cell line B2o there was increased expression of Ha-ras R N A transcript (Fig. 4). Expression of the Ki-ras oncogene was not detectable in any samples by Northern analysis. Expression of c-myc was not detectable in R N A isolated from normal or primary-derived cells but faintly detectable after 1

295

Marker (kb) 4.4

2.3 2.0 1.4 1.1

-

Fig. 4. Northern-blot analysis of Ha-rasmessage. 15 #g of total RNA were applied to 0.8°7o agarose formaldehyde denaturing gels. The molecular weight markers are 32p-labeledXand ~X174 DNA digested with Hind III and Hae Ili, respectively. Autoradiograms were exposed for 1 week at - 70°C in the presence of intensifying screens. week exposure in 5 of 8 cell lines (not shown).

Transfection o f R TE cell DNA into N I H / 3 T3 cells R T E cell D N A (40/zg) was added to 5-10 replicate dishes containing N I H / 3 T 3 cells to monitor for transfectable transforming genes using a calcium phosphate transfection protocol. Each transfection was performed at least twice. Transformed colonies were not detected in the 3T3 monolayer at a level substantially above the background seen for negative control D N A (calf-thymus D N A or N I H / 3 T 3 cell DNA). Transfection of a positive control [plasmid p H O 6 1 T containing human activated Ha-ras (Spandidos and Wilke, 1984)] resulted in 250-1000 transformed colonies per/~g of D N A transfected. Discussion The aim of this study was to determine if some specific oncogene alterations had taken place dur-

ing in vitro transformation of RTE cell lines elicited by chemical treatment. There is evidence that oncogene activation is a c o m m o n event associated with the appearance of several cancers (Barbacid, 1985; Stowers et al., 1987). In vitro cell transformation systems can permit the determination of some very early genotypic changes and strengthen the evidence that particular genetic changes are responsible for the genesis of certain events in the establishment of neoplasia. O f the 8 RTE cell lines (immortal) that were used in this study, 7 were tumorigenic in nude mice, but all cell lines were derived from in vitro exposures to carcinogens (DMBA or BP) a n d / o r a tumor promoter (TPA). No evidence for amplification of Ha-ras, Ki-ras or c-myc was noted in any cell line. We found RFLPs only in one cell line in Barn H I and Hind III digests probed with c-myc. The c-myc alteration that we observed after Bam H I digestion was also seen by Sawey et al. (1987) in several rat skin tumors induced with radiation. In terms of specific changes, we assessed only two potential Ha-ras point mutations. Digestion with Pst I, which tests for a codon 12 (GGA to GCA) alteration, and digestion with Xba I which tests for the commonly seen codon 61 (CAA to CTA) alteration, were unremarkable. In cell lines generated by D M B A exposure an Ha-ras codon 61 alteration was expected since this alteration has been reported in both early and late-stage tumors elicited by D M B A in both rats and mice (Zarbl et al., 1985; Pelling et al., 1987; Quintanilla et al., 1986; Bizub et al., 1986); D N A from mouse skin papillomas produced by T P A treatment have been shown to contain a C to A transversion mutation in the second nucleotide of codon 61 of Ha-ras (Pelling et al., 1988). The same Xba I restriction site ( T C T A G A ) encompassing codon 61 is produced by A to T transversion in Ki-ras as well as Ha-ras, however, we also found these restriction digest patterns in Ki-ras to be unchanged from controls. In additional experiments we performed PCR/direct D N A sequencing on 10 DMBA-derived RTE cell lines and found no nucleotide sequence changes compared with the normal nucleotide sequence reported for the rat Ha-ras gene in codons 59 through

296 61 (Mass and Austin, 1989a). The lack of RFLPs and specific ras codon changes in codon 61 as well as the inability of RTE cell D N A to transform N I H / 3 T 3 cells supports the notion that ras codon 61 alterations are not involved in early stages of RTE cell transformation in vitro in DMBA and TPA-induced cell lines. We observed a slight increase in Ha-ras m R N A as well as marginal increases in c - m y c expression in chemically transformed RTE cell lines. These findings correlate with our previous findings of distinct changes in C C G G sequence methylation in the Haras and c - m y c genes and are indicative of alterations c o m m o n to the transformed phenotype that are irrespective of the initiating agent (Mass et al., 1989c). The lack of Ki-ras expression seen here was also reported by C o s m a et al. (1989) and is concordant with studies demonstrating hypermethylation of Ki-ras C C G G sequences in RTE cells (Mass et al., 1989c). C o s m a et al. (1989) have reported enhanced c - m y c expression that appears to correlate with neoplastic stage in RTE cells exposed in vivo to DMBA. In pilot experiments we did not observe increases in transcripts of the f m s oncogene, as observed by Walker et al. (1987) for tumor-derived R T E cells. [This observation is considered to be artifactual due to presence of extraneous SMFeSV viral polymerase sequences in the f i n s probe that hybridized to contaminating viral sequences acquired by RTE cell lines during propagation in nude mice (Walker et al., 1989).] At this time there are no in vitro studies in rat epithelial cells using DMBA as the initiating agent with which to compare this study. This brings into question whether the lack of alterations is a phen o m e n o n whose mechanism involves the in vitro selection process for carcinogen-altered cells (Harper et al., 1987), or whether the lack of changes is a characteristic of this tissue. The present study is concordant with one published by Knowles et al. (1987), where rat urothelial cells were immortalized by exposure to N M U in vitro. The majority of those cell lines were tumorigenic, and the investigators were unable to demonstrate transformation of N I H / 3 T 3 cells by a rat-derived transforming gene after transfection or to demon-

strate significant changes in restriction endonuclease patterns for several oncogenes. In further support of evidence against ras codon 61 activation as an early or late event in DMBAinduced carcinogenesis of RTE cells, cancers derived by in vivo exposure of heterotopic rat tracheal grafts to DMBA are devoid of Ha-ras oncogene codon 61 alterations as assessed by Xba I restriction digestion (Dr. A.J.P. Klein-Szanto and Dr. J. C a a m a n o , Fox Chase Institute, Philadelphia, PA; personal communication), and DNA from cell lines derived from preneoplastic lesions produced by in vivo exposure to DMBA do not have transfectable transforming genes in the N I H / 3 T 3 assay (Cosma et al., 1989). Recent studies by Hochwalt et al. (1988), Garte et al. (1985) and Watatani et al. (1989) imply that oncogenes other than ras may be more relevant to carcinogenesis in certain epithelial tissues of the rat. Alternatively, the deletion of a ' t u m o r suppressor gene' also needs to be considered as a model for carcinogenesis in this tissue, as exemplified by recent studies in h u m a n lung cancers where a consistent chromosome 3p deletion characterizes many of these tumors (Kok et al., 1987; Johnson et al., 1989). Acknowledgements

The authors thank Teresa Tsuji and P a m Wright for technical assistance. Dr. Steve Reynolds, N I E H S , was extremely patient while helping us interpret the N I H / 3 T 3 cell transfection assays. Portions of this work were funded by U . S . E . P . A . contract 68-02-4456 to Environmental Health Research and Testing, Inc. This manuscript has been reviewed by the Health Effects Research Laboratory, U.S. E P A and approved for publication. The views expressed herein are solely those of the authors and do not necessarily reflect those of the U.S. EPA. Mention of trade names and commercial products does not constitute endorsement for use. References

Almoguera, C., D. Shibata, K. Forrester, J. Martin, N. Arn-

297 heim and M. Perucho (1988) Most human carcinomas of the exocrine pancreas contain mutant c-K-ras genes, Cell, 549-554. Barbacid, M. (1985) Oncogenes in human cancers and in chemically induced animal tumors, Prog. Med. Virol., 32, 86-100. Bizub, D., A.W. Wood and A.M. Skalka (1986) Mutagenesis of the Ha-ras oncogene in mouse skin tumors induced by polycyclic aromatic hydrocarbons, Proc. Natl. Acad. Sci. (U.S.A.), 83, 6048-6052. Cosma, G.N., A.C. Marchok and S.J. Garte (1989) Oncogene activation in cell lines from rat tracheal implants exposed in vivo to 7,12-dimethylbenz[a]anthracene (DMBA), Mol. Carcinogen., in press. Garte, S.J., A.T. Hood, A.E. Hochwalt, P. D'Eustachio, C.A. Snyder, A. Segal and R.E. Albert (1985) Carcinogen specificity in activation of transforming genes by directacting alkylating agents, Carcinogenesis, 6, 1709-1712. Gray, T.E., D.G. Thomassen, M.J. Mass and J.C. Barrett (1983) Quantitation of cell proliferation, colony formation, and carcinogen-induced cytotoxicity of rat tracheal epithelial cells grown in culture on 3T3 feeder layers, In Vitro, 19, 559-570. Harper, J.R., S.H. Reynolds, D.A. Greenhalgh, J.E. Strickland, J.C. Lacal and S.H. Yuspa (1987) Analysis of the rasn oncogene and its p21 product in chemically induced skin tumors and tumor-derived cell lines, Carcinogenesis, 8, 1821-1825. Hochwalt, A.E., I. Wirgin, M. Felber, D.D. Currie and S.J. Garte (1988) Detection of novel non-ras oncogenes in rat nasal squamous cell carcinomas, Mol. Carcinogen., 1, 4-6. Johnson, B.E., A.Y. Sakaguchi, A.F. Gazdar, J.D. Minna, D. Burch, A. Marshall and S.L. Naylor (1989) Restriction fragment length polymorphism studies show consistent loss of chromosome 3p alleles in small cell lung cancer patient's tumors, J. Clin. Invest., 82, 502-507. Knowles, M.A., M.E. Eydmann, A. Proctor, R.A. Padua and J. Roberts (1987) N-Methyl-N-nitrosourea-induced transformation of rat urothelial cells in vitro is not mediated by activation of ras oncogenes, Oncogene, 1, 143-148. Kok, K., J. Osinga, B. Carritt, M.B. Davis, A.H. van der Hout, A.Y. van der Veen, R.M. Landsvater, L.F.M.H. Leij, H.H. Berendsen, P.E. Postmus, S. Poppema and C.H.C.M. Buys (1987) Deletion of DNA sequence at the chromosomal region 3p21 in all major types of lung cancers, Nature (London), 330, 578-581. Lemoine, N.R., E.S. Mayall, F.S. Wyllie, C.J. Farr, D. Hughes, A. Padua, V. Thurston, E.D. Williams and D. Wynford-Thomas (1988) Activated ras oncogenes in human thyroid cancers, Cancer Res., 48, 4459-4463. Mass, M.J., and S.J. Austin (1989a) Absence of mutations in codon 61 of the Ha-ras oncogene in epithelial cells transformed in vitro by 7,12-dimethylbenz[a]anthracene, Biochem. Biophys. Res. Commun., in press. Mass, M.J., J.A. Lasley, S.J. Austin, D.K. Beeman and N.S. Schorschinsky (1989b) Lack of expected ras codon 61 changes

or transfectants from 7,12-dimethylbenz[a]anthraceneexposed rat tracheal epithelial cells in vitro, Proc. Am. Assoc. Cancer Res., 30, 185. Mass, M.J., N.S. Schorschinsky, J.A. Lasley, D.K. Beeman and S.J. Austin (1989c) Consistent oncogene methylation changes in epithelial cells chemically transformed in vitro, Biochem. Biophys. Res. Commun., 164, 693-699. Pelling, J.C., S.M. Fischer, R. Neades, J. Strawhecker and L. Schweickert (1987) Elevated expression and point mutation of the Ha-ras proto-oncogene in mouse skin tumors promoted by benzoyl peroxide and other promoting agents, Carcinogenesis, 8, 1481-1484. Pelling, J.C., R. Neades and J. Strawhecker (1988) Epidermal papillomas and carcinomas induced in uninitiated mouse skin by tumor promoters also contain a point mutation in the 61 st codon of the Ha-ras oncogene, Carcinogenesis, 9, 665-667. Quintanilla, M., K. Brown, M. Ramsden and A. Balmain (1986) Carcinogen-specific mutation and amplification of Ha-ras during mouse skin carcinogenesis, Nature (London), 322, 78-80. Sawey, M.J., A.T. Hood, F.J. Burns and S.J. Garte (1987) Activation of c-rnyc and c-K-ras oncogenes in primary rat tumors induced by ionizing radiation, Mol. Cell. Biol., 7, 932-935. Spandidos, D.A., and N.M. Wilke (1984) Malignant transformation of early passage rodent cells by a single mutated human oncogene, Nature (London), 310, 469-475. Steele, V.E., A.C. Marchok and G.M. Cohen (1980) Transformation of rat tracheal epithelial cells by benzo[a]pyrene and its metabolites, Cancer Lett., 8, 291-298. Steele, V.E., D.K. Beeman and P. Nettesheim (1984) Enhanced induction of the anchorage-independent phenotype in initiated rat tracheal epithelial cell cultures by the tumor promoter 12-O-tetradecanoylphorbol-13-acetate, Cancer Res., 44, 5068-5072. Steele, V.E., J.T. Arnold and M.J. Mass (1988) In vivo and in vitro characteristics of early carcinogen-induced premalignant phenotypes in cultured rat tracheal epithelial cells, Carcinogenesis, 9, 1121-1127. Stowers, S.J., R.R. Maronpot, S.H. Reynolds and M.W. Anderson (1987) The role of oncogenes in chemical carcinogenesis, Environ. Health Perspect., 75, 81-86. Thomassen, D.G., T.E. Gray, M.J. Mass and J.C. Barrett (1983) High frequency of carcinogen-induced early, preneoplastic changes in rat tracheal epithelial cells in culture, Cancer Res., 43, 5956-5963. Vuorio, T., A. Warri, M. Sandberg, K. Alitalo and E. Vuorio (1988) Expression of the c-Ha-ras and neu oncogenes in DMBA-induced, anti-estrogen-treated rat mammary tumors, Int. J. Cancer, 42, 774-779. Walker, C., P. Nettesheim, J.C. Barrett and T.M. Gilmer (1987) Expression of a f m s - r e l a t e d oncogene in carcinogeninduced neoplastic epithelial cells, Proc. Natl. Acad. Sci. (U.S.A.), 84, 1804-1808. Walker, C., P. Nettesheim, J.C. Barrett, F.R. Jirik, J. Sorgem,

298 M. Joyce and T. Gilmer (1989) Mouse retroviral sequences acquired by cell lines after passaging through nude mice detected by hybridization of the fins probe pSM3, Cancer Res., 49, 625-628. Watatani, M., A.O. Perantoni, C.D. Reed, T. Enomoto, M.L. Wenk and J.M. Rice (1989) Infrequent activation of K-ras, H-ras, and other oncogenes in hepatocellular neoplasms initiated by methyl(acetoxymethyl)nitrosamine, a methylating agent, and promoted by phenobarbital in F344 rats, Cancer Res., 49, 1103-1109.

Wigler, M., A. Pellicer, S. Silverstcin, R. Axel, G+ Urlaub and K. Chasin (1979) DNA-mediated transfer of the adenine phosphoribosyl-transferase locus into mammalian cells, Proc. Natl. Acad. Sci. (U.S.A.), 76, 1373-1376. Zarbl, H., S. Sukumar, A.V. Arthur, D. Martin-Zanca and M. Barbacid (1985) Direc! mutagenesis of Ha-ras-I oncogenes by nitrosomethylurea during initiation of mammary carcinogenesis in rats, Nature (London), 315, 382-385. Communicated by J.W. Allen