Avian sarcoma virus transformed hamster cells made resistant to ethidium bromide

The phenanthridine dye ethidium bromide (EB) is of particular interest because of its specific binding to closed circular duplex DNA (Crawford & Waring, 1967). As a consequence, mitochondria are a prime target for this drug. Wild-type yeast can be converted to respiratory-deficient mutants by EB treatment (Slonimski, Perrodin & Croft, 1968; Mahler, Mehrotra & Perlman, 1971). In mammalian cells EB treatment causes structural alterations to mitochondria (Lenk & Penman, 1971; McGill, Hsu & Brinklcy, 1973), reduction in activity of several mitochondrial enzymes (Soslau & Nass, 1971; Radsak, Kato, Sato & Koprowski, 1971; Naum & Pious, 1971), inhibition of thymidine incorporation into closed circular mitochondrial DNA and SV40 DNA (Nass, 1970, 1972; Eason & Vinograd, 1971; Klietmann, Kato & Koprowski, 1972), inhibition of mitochondrial RNA synthesis (Zylber, Vesco & Penman, 1970), and various other effects (Perlman & Penman, 1970).

Prolonged treatment of mammalian cells with EB, even at low concentrations, is usually lethal. EB-resistant cells which can grow in the presence of EB concentrations which are normally highly lethal for tissue culture cells are therefore of interest. Such resistant cells could be useful for a study of the nature of EB resistance, and would probably yield information about the localization of genes in somatic cells.

Hamster cells transformed with simian virus 40 and resistant to EB were described recently (Klietmann, Sato & Nass, 1973). The studies reported here concern the properties of hamster cells transformed by avian sarcoma virus and resistant to EB, as well as of cells resistant to both EB and 5-bromodeoxyuridine (BrdU).

General tissue culture techniques were the same as those used previously (Altaner & Temin, 1970). Tissue culture cells were grown in glass Petri dishes at 38 °C in a humidified CO₂ incubator. The standard culture medium consisted of Eagle’s minimum essential medium (E) supplemented with 20% tryptose phosphate broth (T), 10% calf serum (C10), penicillin and streptomycin.

The medium containing ethidium bromide (Serva, Heidelberg) was prepared by addition of EB from freshly made stock solutions of 5 μg/ml to ETC10 medium.

5-BrdU dissolved in dimethylsulphoxide as a stock solution of 0·1 g/ml was added to ETC10 medium to obtain a BrdU concentration of 100μg/ml.

Medium ETC 10 was also supplemented with hypoxanthine (0·1 ITIM), aminomethopterin (20 μM), thymidine (16 /IM), and glycine (10 μM) - (HATG medium) (Littlefield, 1965). The baby hamster kidney cells transformed with Rous sarcoma virus, strain Schmidt-Ruppin, clonal cell line C13/SR7 used in this study was established by Macpherson (1965). These cells are designated Ha(SR) cells in our experiments.

Ha(SR) cells resistant to 100 μ g/ml BrdU and deficient in thymidine kinase activity - Ha(SR)BU100 - have been described (Altaner, Hladka & Schlechte, 1973).

The procedure for determination of cell growth rates has been described previously (Altaner et al. 1973).

[6-³H]thymidine (spec, activity 5 Ci/mM) was used throughout. Thymidine incorporation into DNA was measured as previously described (Altaner & Hlavayova, 1973).

Mitochondrial DNA was extracted by a technique based on that of Hirt (1967).

DNA was determined colorimetrically using the diphenylamine reaction according to Burton (1956).

Tissue culture fluid harvested from cells was concentrated by ultracentrifugation according to the procedure described earlier (Altaner & Temin, 1970). Endogenous RNA-dependent DNA polymerase and exogenous DNA polymerase were estimated by the standard DNA polymerase assay of Temin & Mizutani (1970).

Different numbers of cells were plated into 60-mm tissue culture dishes (Falcon) in medium containing 15% calf serum. The cell colonies were counted 1 week later to determine plating efficiency in liquid medium. The same cell samples in appropriate dilutions in 2 ml of 0 · 3% agar medium were seeded on a base layer of 5 ml of 0 · 5% agar medium in 60-mm Petri dishes (Falcon). The number of colonies was counted 2 weeks later under low magnification with an inverted microscope.

Cells were grown on coverslips. The original cell lines Ha(SR) and Ha(SR)BU100 were incubated for 6 or 12 h in medium with EB 10 μ g/ml, the resistant cell lines in appropriate medium with EB. The coverslips were mounted on microscope slides in the complete medium with or without EB, or in Earle’s tissue culture solution (Difco). The slides were examined by fluorescence microscopy with an HBO 200 lamp using a BG 12 excitation filter.

The cells were fixed in situ with 1-8 % glutaraldehyde in 0·1 M phosphate buffer pH 7·2 for 1 h at 4 °C and postfixed with 2% OsO₄ in 0·1 M phosphate buffer for 1 h. They were scraped off in the osmium solution with a rubber policeman and pelleted. The pellet was dehydrated in ethanol, propylene oxide, and embedded in Epon. Thin sections were stained with uranyl acetate and lead citrate and examined with a Tesla BS 613A electron microscope at 80 kV.

The assay was based on the fusion of chicken secondary embryo cells with hamster cells using inactivated Sendai virus. Various numbers of X-irradiated hamster cells were fused using u.v.-inactivated Sendai virus with chicken embryo cells. After 3 days cultivation of the cell mixture in vitro, the cells were scraped off with a rubber policeman and inoculated into the wing-web of 3-day-old chicks. Induction of tumours in chickens was taken as evidence of the presence of the virus genome in the mammalian cells.

Ethidium bromide-resistant cells were established by cultivation of Ha(SR) and Ha(SR)BUioo cells in media containing EB. Initially the cultures were grown in 0·1μg/ml EB. Gradually the concentration of EB was increased to 10 and 30μg/ml over a period of 16 months. The cultures were usually subcultivated every 5 days. The concentration of EB was adjusted according to the physiological state of the cultured cells during the adaptation period.

The cells resistant to 10 and to 30μg/ml EB were designated Ha(SR)EB10, and Ha(SR)EB30, respectively.

The Ha(SR)BU100 EB10 cells were prepared in the same way from Ha(SR)BU100 cells, i.e. from cells already resistant to 100μg/ml BrdU (Altaner et al. 1973).

The stabilized cells lines of EB-resistant cells were tested for growth properties in media supplemented with ethidium bromide as well as in normal medium (ETC10). The BrdU-resistant cells were tested for ability to grow in a medium containing 100μg/ml BrdU and in HATG medium.

The progenitor clonal cell line Ha(SR) did not grow in medium supplemented with 10 μ g/ml of EB, while the resistant cells Ha(SR)EB10 reached the same saturation density in EB-containing medium as in normal medium (Fig. 1). The Ha(SR)EB30 cells grew in medium containing 30 /*g/ml EB as well as in normal medium.

Growth curves of Ha(SR)BU100 EB10 and Ha(SR)BU100 cells are given in Fig. 2. The parental cells Ha(SR)BU10 grew in normal medium as well as in BrdU-containing medium, but did not grow in HATG medium, and died within 8 days in medium containing 10 μ g/ml EB. The resistant cells Ha(SR)BU100 EB10 proliferated in normal medium, medium with BrdU, and medium with 10 μ g/ml EB, but did not grow in HATG medium.

Therefore the EB resistance did not influence the deficiency of thymidine kinase activity as shown by the inability of cells to grow in HATG medium.

The original cell line (Fig. 3) is a polymorphous fibroblastoid cell population with single partially multinucleated giant cells (about 1 per high-power field). The morphology of BrdU-resistant cells (Fig. 5), EB-resistant cells (Fig. 4), and the doubly resistant line (Fig. 6) did not differ substantially from that of the original cell lines. In all resistant cell lines the number of giant, single or multinucleated cells is increased about 2 –3 times.

To test whether resistance to EB had any influence on the ability of cells to incorporate thymidine, the following experiments were done. Cell cultures in logarithmic growth phase were incubated for 3 and 6 h with [6-³H]thymidine, and the rate of its incorporation into DNA was determined. As shown in Table 1, the Ha(SR)EB10 cells did not differ in this property from Ha(SR) cells. The ability of the Ha(SR)BU100EB10 cells to incorporate thymidine was slightly increased compared with the parental Ha(SR)BU100, but still represented only 3% of the incorporation of the original Ha(SR) cells. Therefore the Ha(SR)BU100EB10 cells lack thymidine kinase (TK) activity. This was confirmed by direct enzyme assay of thymidine kinase. The enzyme level of Ha(SR)BU100EB10 was 2% of the level of TK in Ha(SR) cells.

To see whether the transformed phenotype, as expressed by the ability of cells to grow in soft agar (Macpherson & Montagnier, 1964), was influenced by resistance to EB, the parental clonal cell line Ha(SR), and Ha(SR)BU100, and the resistant cells were tested for their ability to form colonies in soft agar and liquid medium. The results are shown in Table 2. The ability of EB-resistant cells Ha(SR)EB10 to form colonies in agar was reduced 4-fold, the plating efficiency of Ha(SR)EB30 cells was lowered both in fluid and agar medium, but was unchanged in EB-supplemented medium. The plating efficiency of Ha(SR)BU100EB10 cells was greatly decreased compared with the original Ha(SR)BU100 cells, but was the same in media supplemented with either EB or BrdU.

To determine whether the reduced plating efficiency of Ha(SR)EB10 in agar medium was correlated with the presence of the virus genome in the cells, fusion experiments with chicken embryo cells were carried out (Table 3). There was no change in the number of resistant cells needed for tumour induction in chicks in comparison with control cells. Both types of cells behave as virogenic cells; the number of cells needed for induction of tumours in chicks after fusion with permissive chick embryo cells was about 10³.

The EB-resistant cells and doubly resistant cells (EB, BrdU), as well as both parental cell types, produce neither viral particles in tissue culture medium, nor particles having endogenous RNA-dependent DNA polymerase, or exogenous DNA polymerasc. Nevertheless, in both single EB-resistant cell lines R-type virus-like particles (Bernhard & Tournicr, 1964) have been observed (Figs. 10–12).

To determine whether the EB resistance of Ha(SR)EB10 cells is genetically stable, the cells were cultivated in medium without EB for 8 cell passages (42 days). The cells were then seeded again in media containing EB, and growth curves determined. There was no change in the growth of these cells in EB-containing medium compared with normal medium. The resistance to EB was maintained and seems to be genetically determined.

To find whether the EB in the medium is inactivated by resistant cells, the EB containing medium was exposed to heavy cultures of resistant cells for 24 h. This medium was then supplemented with fresh calf serum, and used for determination of growth curves of parental Ha(SR) cells. No differences in survival of Ha(SR) cells were observed in medium with freshly added EB compared with medium already used on resistant cells. Therefore no inactivation of EB by resistant cells took place.

The drug resistance of mammalian cells could be related to a change in the cell membrane causing a decrease of the amount of drug that penetrates into the cell, as was found in the case of actinomycin D-rcsistant mouse lymphoma cells (Bosmann, 1971). It was found that Ha(SR)EB30 cells were unable to grow in a medium supplemented with 10μg/ml of actinomycin D. Therefore there is no common cell membrane permeability barrier for these 2 drugs.

Several experiments were performed to test whether there is a block to penetration in EB-resistant cells, using the finding (Burns, 1972) that EB can be localized in the cell by fluorescence microscopy. There were no substantial differences in the quantity or quality of fluorescence in the parental and resistant cells. In all cell lines only isolated cells or groups of cells showed orange fluorescence, corresponding to the rounded-up cells (Figs. 7–9). In these cells the most intense fluorescence was found over the nucleoli and heterochromatin. The nucleoplasm and, less consistently, the cytoplasm were faintly fluorescent. The other cells were without detectable fluorescence. These findings did not differ with various mounting media used. Therefore fluorescence microscopy did not reveal any significant changes in EB penetration into the resistant cells.

To test whether the incorporation of thymidine into nuclear and mitochondrial DNA was influenced by EB resistance the following experiments were done. Resistant and parental cells were labelled with [³H]thymidine (20 μCi/ml) for 48 h. DNA was isolated from the labelled cells according to the method of Hirt (1967). It was found (Table 4) that all kinds of EB-resistant cells incorporated thymidine into mitochondrial DNA to the same extent as did the parental cells. The ratio of incorporation of thymidine into nuclear and mitochondrial DNA of doubly resistant cells was different, due to the deficiency in thymidine kinase activity. The specific activity of mitochondrial DNA was higher in these cells, showing that the incorporation found was due to the activity of mitochondrial thymidine kinase (Kit, Kaplan, Leung & Trkula, 1972).

The ultrastructure of the parental cell line Ha(SR) was characterized by moderately developed cell organelles with a considerable number of free ribosomes. In the nucleus in about one third of the cells there were single or multiple nuclear bodies of granular type (Bouteille, Kalifat & Delarue, 1967). Annulate lamellae were seldom present in the cytoplasm. In some cells there were intracytoplasmic R-type virus-like particles, either solitary or in small groups, budding into or free in the cisternae of the rough endoplasmic reticulum, or the perinuclear cisternae (Figs. 10, 11). The mitochondria were mostly of the condensed type. There was variation both in mitochondrial shape, and in the arrangement of the cristae. In some cells the mitochondria had circularly arranged or tubular cristae, and occasionally a honeycomb pattern of cristae.

Both EB-resistant cells showed no ultrastructural differences and also contained the intracytoplasmic R-type virus-like particles (Fig. 12). The number and structure of mitochondria were similar to those in the original cell line (Fig. 13). The proportion of cells containing mitochondria with clear and dense matrix varied in different cultures investigated. The cells usually had normal-appearing mitochondria (Fig. 15), but a small proportion contained the atypical form, as in the original line.

In the BrdU-resistant and the doubly resistant cells (BrdU, EB), the intracytoplasmic R-type virus-like particles were not found. In the majority of these cells the mitochondria appeared normal (Figs. 14, 16), a smaller proportion of cells had polymorphic and atypical mitochondria as in the original cell line. In solitary cells, enlarged mitochondria (Fig. 17) as described in L cells (Soslau & Nass, 1971) were observed.

Thus EB resistance was not accompanied by significant ultrastructural differences in mitochondria as shown by the electron-microscopic findings.

The results reported here present evidence that hamster cells transformed with avian sarcoma virus can be adapted to a high concentration of ethidium bromide. The resistant cells grow in the presence of 30 μg/ml EB which is normally highly lethal for mammalian cells. The cells represent the first cells resistant to EB that contain an RNA tumour virus genome. Recently hamster cells transformed with SV-40 virus resistant to EB have been described by Klietman et al. (1973).

This paper also presents evidence that cells already resistant to BrdU and lacking thymidine kinase activity can be made resistant to EB. The deficiency of thymidine kinase activity was not influenced by ethidium bromide during long-term cultivation in medium with EB in the absence of BrdU. The doubly resistant cells are very useful, as they have several genetic markers: thymidine kinase activity deficiency, EB resistance, and RNA tumour virus genome. These properties make them useful for selective preparation of interspecific cell hybrids. Thus the nature of EB resistance could be studied at the genetic level by this means.

It is generally accepted that the main target of ethidium bromide in cells is the mitochondria. The resistant cells, after selection in EB-containing media show no significant morphological changes in mitochondria, nor is mitochondrial DNA influenced by the EB resistance. The EB resistance is genetically stable. On the basis of our fluorescence microscopy findings and some preliminary results with ¹⁴C-labelled ethidium bromide, it seems that resistance to EB is not due to a cell permeability barrier. It might have been caused by membrane changes in some cell organelle, arising by mutation during the adaptation period, which survived under the selective pressure of EB. Similarly, it has been suggested that the stable EB resistance in yeast could be envisaged as a mutation in the lipoprotein components of the mitochondrial membrane which has altered the charge and conformation of the EB-binding sites to exclude EB molecules (Bech-Hansen & Rank, 1972). However, the precise nature of EB resistance remains to be elucidated.

The EB resistance influenced neither the avian sarcoma virus genome in the cells, nor the intracytoplasmic R-type virus-like particles (Shipman, Weide & Ma, 1969), which are frequently found in BHK-21 cells from which the Ha(SR) cells were derived. It seems that the production of R-type virus-like particles is suppressed in BrdU and BrdU/EB-resistant cells, but our study was not critically oriented in this direction. It is of interest to mention that the avian sarcoma virus genome in Ha(SR) cells resistant to 8-azaguaninc (Hladka & Altaner, 1974) was found to be strongly suppressed. Therefore the EB resistance differs substantially from the 8-azaguanine induced resistance.

A possible experimental approach to study of the genetic nature of EB resistance in mammalian cells using cell hybridization is now being investigated.