oncogene. When a pool of 4,700 G418-resistant colonies was injected into nude mice, tumors were obtained. These tumors contain a transfected human rasHÂ ...
MOLECULAR AND CELLULAR BIOLOGY, Jan. 1985, p. 268-272 0270-7306/85/010268-05$02.00/0 Copyright C 1985, American Society for Microbiology
Vol. 5, No. 1
New Acceptor Cell for Transfected Genomic DNA: Oncogene Transfer into a Mouse Mammary Epithelial Cell Line N. E. HYNES,* R. JAGGI, S. C. KOZMA, R. BALL, D. MUELLENER, N. T. WETHERALL,t B. W. DAVIS,t AND B. GRONER Ludwig Institute for Cancer Research, Bern Branch Inselspital, 3010 Bern, Switzerland
Received 19 July 1984/Accepted 9 October 1984 A line of mouse mammary epithelial cells (NMuMG) has been characterized for its ability to be stably transfected with exogenous DNA. A transfection frequency of at least 1 cell per 1,000 was obtained with the pSV2neo plasmid. Several thousand G418-resistant NMuMG cell clones can easily be generated in cotransfection of genomic DNA and pSV2neo. The NMuMG cells were isolated from normal mammary glands and do not form malignant lesions when injected into nude mice. We have cotransfected NMuMG cells with pSV2neo and genomic DNA from the human EJ bladder carcinoma line, a cell line which contains an activated c_rasH oncogene. When a pool of 4,700 G418-resistant colonies was injected into nude mice, tumors were obtained. These tumors contain a transfected human rasH gene. Genomic DNA transfection into a line of mouse epithelial cells, in combination with the selection of stable transfectants and tumor induction in nude mice, can be used to screen human tumor DNA for the presence of activated oncogenes.
replating into G418 selective medium is critical. The protocol which we adopted was as follows. NMuMG cells (0.5 x 106) were seeded onto a 10-cm dish, and 24 h later 1 ml of calcium phosphate-DNA precipitate (5, 8) containing 30 pLg of genomic DNA and 2 ,Ig of the pSV2neo plasmid was added to the cells. After 20 h the precipitate-containing medium was replaced by fresh medium. The cells were incubated for 1 to 2 additional days and then trypsinized and plated at a dilution of 1:10 into medium containing 1.0 mg of Geneticin G418 sulphate (GIBCO Laboratories, Grand Island, N.Y.) per ml (ca. 0.5 mg of antibiotic G418 per ml). After 10 to 14 days, one plate was stained, and cell clones were counted. The transfection frequency, i.e., the number of cells on the original plate which stably took up and expressed the pSV2neo plasmid, was at least 1 per 1,000 cells transfected. In our experience the NMuMG cells were as efficient in taking up and expressing foreign DNA as NIH 3T3 cells and various other cell lines previously described
DNA isolated from some established tumor cell lines and primary tumors contains activated oncogenes. These oncogenes have been detected by transfecting genomic DNA into NIH 3T3 mouse fibroblasts and scoring for foci of morphologically transformed cells (12, 13, 16, 18, 20). However, many tumor DNAs tested by this method have failed to give rise to detectable foci (16, 18), and the majority that score positively contain an activated oncogene belonging to the c-ras gene family (3, 7, 15, 19, 22). It may be possible that NIH 3T3 cells are most easily morphologically transformed by a ras oncogene. Alternatively, cell type specificity for oncogene action may exist. Since many human tumors are carcinomas, the use of epithelial cells as acceptor cells for oncogene detection may be important. This could result in a higher number of positively scoring tumor DNAs or allow the isolation of different oncogenes, or both. We have developed an alternative to the standard transformation assay in NIH 3T3 cells. In this paper, we describe our results with a mouse mammary epithelial cell line, NMuMG (14), as the recipient for transfected DNA. Two criteria must be met before a cell line can be used to screen genomic DNA for oncogenes. First, the frequency of stable transfection must be sufficiently high to generate a large number of transfected cell clones. Second, transformation of the acceptor cells by an activated oncogene must result in a detectable and selectable change in phenotype. We show in this paper that the NMuMG cells meet both criteria and may be an alternative to NIH 3T3 cells for screening genomic DNA from carcinoma cells for the presence of oncogenes. The NMuMG cells were tested for their ability to be stably transfected with the plasmid pSV2neo (23). This plasmid contains a dominant selectable marker which confers resistance to the aminoglycoside antibiotic G418. Different procedures were tested, and we found that there were two important parameters. First, an amount of pSV2neo sufficient to transfect all of the competent cells must be added to each plate. Second, the density of the transfected cells when
(4).
The standard assay in NIH 3T3 cells depends upon the detection of foci of morphologically transformed cells. Even in NIH 3T3 cells this can be difficult, since there is variation in the extent to which different oncogene products alter the morphology of the cells. Some of the tumor cell DNAs which have been negative in this assay may contain oncogenes which cause subtle changes in cell growth. For epithelial cells, a transformation assay based upon screening for foci of transformed cells is even more difficult. It is generally agreed that there is a lack of characteristic morphological changes associated with the malignant transformation of epithelial cells (1, 10, 25). Therefore, an alternative test for the transformation of NMuMG cells by oncogenes is to assay directly for tumorigenicity in nude mice. The NMuMG cells which were isolated from the mammary glands of Namru mice have epithelial growth characteristics (14) and stain positively for keratin (unpublished data). In the original publication it was shown that when 2 x 106 to 4 x 106 NMuMG cells were injected into the interscapular fat pad of newborn mice, 18 of 28 animals (64%) developed benign cystadenomas. We injected from 1 x 106 to 5 x 106 cells into nude mice and also observed that 7 of 14
* Corresponding author. t Present address: Department of Pathology, Vanderbilt University Medical Center, Nashville, TN 37232.
268
NOTES
VOL. 5, 1985
animals (50%) developed benign cystadenomas (Table 1). Figure 1 shows a cystically dilated neoplasm formed at the site of inoculation of the NMuMG cells. Despite forming these small benign lesions in nude mice, NMuMG cells, unlike NIH 3T3 cells, do not give rise to malignant tumors. Therefore, NMuMG cells are good candidates for use as acceptor cells to screen for transforming sequences by tumor induction. We next tested whether it was possible to transform the NMuMG cells, using a known oncogene, into cells capable of forming a malignant tumor in nude mice. NMuMG cells were transfected with pSV2neo in the presence of a 10-fold excess of a plasmid containing the activated c-rasH oncogene (p-rasH) which has been molecularly cloned from the EJ human bladder carcinoma cell line (21). G418-resistant cell clones were picked and analyzed for tumorigenicity. Table 1 shows that when 106 cells of one of the clones (clone 1) were injected into five mice, all of the animals developed a tumor with a latency period of 2 weeks. Figure 2 shows that the tumor is an invasive carcinoma which infiltrates the underlying skeletal muscle and is composed of rapidly proliferating anaplastic cells with epithelial characteristics. The tumor cells also stain positively for keratin (data not shown) as further evidence supporting the epithelial nature of the transformed cells. Finally, the NMuMG cells were tested for their tumorigenicity after the transfection of EJ bladder carcinoma cell line genomic DNA. The genomic DNA transfection was done in the presence of the selectable pSV2neo marker for two reasons. First, the efficiency of the transfection could be easily monitored. Second, the background of nontransfected NMuMG cells would be killed, allowing only those cells which have taken up exogenous DNA to be injected into the animal. It has previously been shown that 2 x 103 to 3 x 103 kilobases of DNA are taken up into a transfected cell (11, 17). Therefore, one must isolate at least 103 individual G418-resistant cell clones to represent one genomic equivalent of DNA. By transfecting three plates of NMuMG cells with the EJ cell line DNA by the protocol described above, we generated 4,700 G418-resistant cell clones. As a control, we carried out a parallel experiment transfecting human placenta DNA. In this case, 2,000 cell clones were generated. The cell clones from each of the transfections were pooled and injected into nude mice. Table 1 shows that three of the mice injected with the EJ cell line-transfected NMuMG cells developed tumors during the 13-week observation period, whereas none of the mice injected with the placenta DNA-transfected cells developed tumors. Thus, it seemed likely that the activated c-rasH was instrumental in causing the tumors.
269
The DNAs from two of the tumors arising from the
p-ras'-transformed cells and from one tumor arising from
the EJ cell line DNA-transfected cells were analyzed for ras -specific sequences with a standard DNA blot hybridization technique (6; Fig. 3). The DNA from NMuMG cells transfected with p-rasH (clone 1) contains the expected 6.6-kilobase BamHI fragment (lane 1). The additional bands reflect plasmid rearrangements which generally occur during transfection experiments. The two tumors arising from clone 1 cells show the same pattern of rasH-hybridizing sequences (lanes 2 and 3). The DNA isolated from the tumor which developed from NMuMG cells transfected with the EJ cell line DNA is shown in lane 5. The expected 6.6-kilobase BamHI fragment was found in the tumor, but it was amplified in comparison to the band seen in the EJ cell line DNA (lane 4). Lanes 6 to 8 show the intensity of hybridization when 10, 50, and 250 pg of p-rasH DNA were analyzed. Since the intensity of hybridization of 10 and 250 pg of plasmid DNA approximately corresponds to the intensity seen for, respectively, the EJ cell line DNA and the tumor DNA, we concluded that the rasH gene is amplified at least 25-fold in the tumor DNA. We analyzed two other tumors which arose from the injection of EJ cell line-transfected NMuMG cells. In these tumors the rasH gene was rearranged but not amplified (data not shown). Therefore, amplification of the Ha-ras gene did not appear to be a prerequisite for tumor growth. RNAs isolated from the tumors arising from NMuMG cells transfected with either p-rasH or EJ cell line DNA were analyzed for their rasH transcripts (9, 24). All of the tumors contained the expected 1.4-kilobase RNA (data not shown). We have shown in this paper that a mouse mammary epithelial cell line, NMuMG cells, can be used as acceptor cells in screening tumor cell DNA for oncogenes. The cells are capable of taking up and expressing an oncogene which is present as a single copy in genomic DNA. To establish the assay conditions, we have transfected the cells with genomic DNA from the EJ cell line. When the pool of NMuMG cells transfected with EJ genomic DNA is injected into nude mice, malignant tumors containing the activated c-rasH gene are formed. The results are interesting for two reasons. First, the transformed NMuMG cells may be useful in studying the properties of malignant epithelial cells, as the parameters which distinguish transformed cells from their nontransformed counterparts have been most extensively studied in fibroblasts and lymphoid cells. It will also be important to establish the characteristics of malignant epithelial cells per se. Second, since the NMuMG cells can be transfected with a high frequency, they may be used to screen tumor cell genomic DNA for active oncogenes. If
TABLE 1. Tumorigenicity of transfected NMuMG cells in nude mice No. of recipient cells in
Transfecting DNA
inoculuml 106
2 x 106 5 x 106 106 2 x 106 to 3 x 106 2 x 106 to 3 x 106
None None None
pSV2neo/p-rasH (Clone 1) pSV2neo/EJ cell line pSV2neo/Human placenta
Latency period (wks)b
6.5 2.5 2.5 2.5 6
Observation period (wks)
8 10 6.5 4 13 13
Tumor incidence
Mean tumor volume (cm3)'
1/5
0.1
2/5 4/4
5/5 3/5
0.1 4.2 1.8
0/5
a The inoculation of the cells and the tumor histology were performed as described in the legend to Fig. 1. b Time of appearance of first tumor. c Tumor volume was calculated from the formula 4/3 Xab2; a is one-half of the long axis, and b is one-half of the short axis.
Histology" Cystadenoma Cystadenoma Cystadenoma Invasive carcinoma Carcinoma
I
B~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~W r'k,, /'
4.
_
)
F
%%
Ak'A .v'''K, r8 1 e '", "" " J , t'' + _ \ '. &%s %' (V
4'
_
';e.p ,*;
F'
FIG. 1. Charactenization of benign lesions caused by NMuMG cell injection into nude mice. NMuMG cells were trypsinized, centrifuged, and suspended in medium lacking serum, and 0.2 ml containing 1 x 106 to 5 x 106cells was injected subcutaneously into the interscapular region of 3- to 4-week-old nude (nu/nu) mice of Swiss background. (Top) Cystically dilated adenomatous neoplasms formed at the inoculation site in 7 of 14 mice. These were removed, fixed in 4% buffered formaldehyde, and embedded in paraffin. Classification of the lesions was made on sections visualized with hematoxylin and eosin stain (magnification, x 100). (Bottom) The cystic spaces of these adenomas were lined by columnar and pseudostratified columnar epithelia. The underlying stroma is distinct from subcutaneous tissue and consisted of myoepithelium-like cells and small fibroblasts (magnification, x 1,000; hematoxylin and eosin stain).
270
NOTES
VOL. S, 1985
271
4E\ *L
~~ ~
~
IN,
~
~
~ ~ 1~
4, ,'. .*
Neoplasms
FIG. 2. Neoplasms arising from the injection of NMuMG cells transfected with p-rasH. When 106 NMuMG cells transfected with p-rasH were injected into Swiss nude (nulnu) mice, invasive neoplasms which infiltrated underlying skeletal muscle developed (magnification, x 250; hematoxylin and eosin stain).
KB
1 2 3 4 5 6
7 8 9
23.7_
9.9 6.7-
Y
*4
*'
4.2
2.2FIG. 3. Southern hybridization analysis of genomic DNA containing rasH sequences. DNA (5 ,ug) was digested with BamHI, electrophoresed through an 0.8% agarose gel, blotted onto Gene Screen Plus filter paper, hybridized with the 32P nick-translated 6.6-kilobase BamHI fragment containing the human rasH gene (21), and washed as previously described (2). The DNA was isolated from the following sources: a.clone of NMuMG cells cotransfected with pSV2neo and p-rasH (clone 1) (lane i); two individual tumors taken from animals injected with 106 clone 1 cells (lanes 2 and 3); the EJ bladder carcinoma cell line (lanes 4 and 9); a tumor removed from an animal injected with 2 x 106 to 3 x 106 NMuMG cells consisting of a pool of cell clones cotransfected with pSV2neo and genomic DNA from the EJ cell line (lane 5); and 10, 50, and 250 pg of BamHI-digested p-rasH electrophoresed and blotted to estimate the degree of amplification of the rasH gene in the tumor DNA of lane 5 (lanes 6 to 8, respectively).
there is a cell type specificity for oncogene action, screening genomic DNA in an epithelial cell line may bring different results than transfection into NIH 3T3 fibroblasts. We have begun testing DNAs from human carcinoma cells for their ability to transform NMuMG cells. We thank C. Dickson for providing us with the NMuMG cells, R. Weinberg for the plasmid containing the activated c-rasH, and M. L. Carrozza, H. Gronemeyer, and S. Redmond for helpful discussions, and we acknowledge the expert technical assistance of S. Saurer and H. Birk. G. W. Locher, Universitats-Frauenklinik, Bern, provided the animal facilities. We also thank C. Wiedmer for help in preparing the manuscript. LITERATURE CITED 1. Butel, J. S., C. Wong, and D. Medina. 1984. Transformation of mouse mammary epithelial cells by papova virus SV40. Exp. Mol. Pathol. 40:79-108. 2. Felber, B. K., S. Gerber-Huber, C. Meier, F. E. B. May, B. Westley, R. Weber, and G. U. Ryffel. 1981. Quantitation of DNase I sensitivity in XenopuS chromatin containing active and inactive globulin, albumnin and vitellogenin genes. Nucleic Acids Res. 9:2455-2474. 3. Goldfarb, M., K. Shimizu, M. Perucho, and M. Wigler. 1982. Isolation and preliminary characterization of a human transforming gene from T24 bladder carcinoma cells. Nature (London)
296:404-409.
4. Gorman, C., R. Padmanabhan, and B. H. Howard. 1983. High
efficiency DNA-mediated transformation of primate cells. Sci-
ence 221:551-553. 5. Graham, F. L., and A. J. vkn der Eb. 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52:456-467. 6. Groner, B., and N. E. Hynes. 1980. Number and location of mouse mammary tumor virus proviral DNA in mouse DNA of normal tissue and of mammary tumors. J. Virol. 33:1013-1025. 7. Hall, A., C. J. Marshall, N. Spurr, and R. A. Weiss. 1983. The transforming gene in two human sarcoma cell lines is a new
272
8.
9.
10.
11.
12. 13.
14. 15.
16.
NOTES member of the ras gene family located on chromosome one. Nature (London) 303:396-400. Hynes, N. E., N. Kennedy, U. Rahmsdorf, and B. Groner. 1981. Hormone responsive expression of an endogenous proviral gene of mouse mammary tumor virus after molecular cloning and gene transfer into cultured cells. Proc. Natl. Acad. Sci. U.S.A. 78:2038-2042. Hynes, N. E., U. Rahmsdorf, N. Kennedy, L. Fabiani, R. Michalides, R. Nusse, and B. Groner. 1981. Structure, stability, methylation, expression and glucocorticoid induction of endogenous and transfected proviral genes of mouse mammary tumor virus in mouse fibroblasts. Gene 16:307-317. Kaighn, M. E., S. K. Narayan, Y. Ohnuki, L. W. Jones, and J. F. Lechner. 1980. Differential properties among clones of simian virus 40-transformed human epithelial cells. Carcinogenesis 1:635-645. Kavathas, P., and L. A. Herzenberg. 1983. Stable transformation of mouse L cells for human membrane T-cell differentiation antigens, HLA and ,B2-microglobulin: selection by fluorescenceactivated cell sorting. Proc. Natl. Acad. Sci. U.S.A. 80:524-528. Krontiris, T. G., and G. M. Cooper. 1981. Transforming activity of human tumor DNAs. Proc. Natl. Acad. Sci. U.S.A. 78:1181-1184. Murray, M. J., B. Z. Shilo, C. Shih, 1). Cowing, H. W. Hsu, and R. A. Weinberg. 1981. Three different human tumor cell lines contain different oncogenes. Cell 25:355-361. Owens, R. B., H. S. Smith, and A. J. Hackett. 1974. Epithelial cell cultures from normal glandular tissue of mice. J. Natl. Cancer Inst. 53:261-266. Parada, L. F., C. J. Tabin, C. Shih, and R. Weinberg. 1982. Human EJ bladder carcinoma oncogene is homologue of Harvey sarcoma virus ras gene. Nature (London) 297:474-478. Perucho, M., M. Goldfarb, K. Shimizu, C. Lama, J. Fogh, and
MOL. CELL. BIOL.
17. 18. 19.
20.
21. 22.
23.
24. 25.
M. Wigler. 1981. Human tumor-derived cell lines contain common and different transforming genes. Cell 27:467-476. Perucho, M., and M. Wigler. 1980. Linkage and expression of foreign DNA in cultured animal cells. Cold Spring Harbor Symp. Quant. Biol. 45:829-838. Pulciani, S., E. Santos, A. V. Lauver, L. K. Lohg, S. A. Aaronson, and M. Barbacid. 1982. Oncogenes in solid human tumours. Nature (London) 300:539-542. Santos, E., S. R. Tronick, S. A. Aaronson, S. Pulciani, and M. Barbacid. 1982. T24 human bladder carcinoma oncogene is an activated form of the normal human homologue of, Balb and Harvey MSV transforming genes. Nature (London) X$8.:343-347. Shih, C., B. Shilo, M. Goldfarb, A. Dannenberg' and R. Weinberg. 1979. Passage of phenotypes of chenmically-transformed cells via transfection of DNA and chromatin. Proc. Natl. Acad. Sci. U.S.A. 76:5714-5718. Shih, C., and R. A. Weinberg. 1982. Isolation of a ttansforming sequence from a human bladder carcinoma cell line. Cell 29:161-169. Shimizu, K., M. Goldfarb, M. Perucho, and M. Wigler. 1983. Isolation and preliminary characterization of the tra 'sforming gene of a human neuroblastoma cell line. Proc. Natl. Acad. Sci. U.S.A. 80:383-387. Southern, P. J., and S. Berg. 1982. Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. Mol. Appl. Genet. 1:327-341. Thomas, P. 1980. Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc. Natl. Acad. Sci. U.S.A. 77:5201-5205. Yuspa, S. H., P. Hawley-Nelson, B. Koehler, and J. R' Stanley. 1980. A survey of transformation markers in differentiating epidermal cell lines. Cancer Res. 40:4694-4703.