Ethidium bromide-induced loss of mitochondrial DNA from primar chicken embryo fibroblasts

Vol. 5, No. 5

MOLECULAR AND CELLULAR BIOLOGY, May 1985, p. 1163-1169 0270-7306/85/051163-07$02.00/0 Copyright C 1985, American Society for Microbiology

Ethidium Bromide-Induced Loss of Mitochondrial DNA from Primary Chicken Embryo Fibroblasts PAUL DESJARDINS,l ERIC FROST,2t AND REJEAN MORAIS1* Institut du Cancer de Montreal, H6pital Notre-Dame, Montreal, Quebec H2L 4M1 , and Departments of Biochemistryl and Medicine,2 Universite de Montreal, Montreal, Quebec H3E 3J7, Canada Received 8 November 1984/Accepted 7 February 1985

Chicken embryo fibroblasts in uridine-containing medium are inherently resistant to the growth-inhibitory effect of ethidium bromide. The drug was found to inhibit the incorporation of [3lHlthymidine into mitochondrial DNA circular molecules. Mitochondrial DNA was quantitated by DNA-DNA reassociation kinetics with a probe of chicken liver mitochondrial DNA. A mean number of 604 copies of mitochondrial DNA per cell was found. This number decreased progressively in cells exposed to ethidium bromide, and by day 13 ca. one copy of mitochondrial DNA was detected per cell. When the cells were then transferred to drug-free medium, the number of copies increased very slowly as a function of time. On the other hand, analyses of DNA extracted from cell populations exposed to ethidium bromide for 20 or more days, with or without subsequent transfer to drug-free medium, revealed very little or no mitochondrial DNA by reassociation kinetics or by Southern blot hybridization of AvaI- or HindIII-digested total cellular DNA. As a result of the elimination of mitochondrial DNA molecules, the establishment of cell populations with a respiration-deficient phenotype was confirmed by measuring cytochrome c oxidase activity as a function of the number of cell generations and the absorption spectrum of mitochondrial cytochromes.

EtdBr from the culture medium, the rate of replication of mtDNA per cell cycle could be insufficient to substantially increase the number of mtDNA copies, and the proliferative cells remain respiration deficient. On the other hand, other than decreasing the number of mtDNA copies per cell, EtdBr might also mutate mtDNA in CEF so that the expression of some or of all the genes is prevented upon the removal of the drug from the culture medium. Although there is no direct evidence in the literature to support the latter possibility, the drug is known to induce a structural alteration in covalently closed mtDNA of vertebrate cells, consisting in part of an increased degree of supercoiling (23, 30) and possibly a breakage of circular DNA without reclosing as observed by electron microscopy (23). These changes appear to be reversible when the cells are transferred to EtdBr-free medium. In contrast, EtdBr is known to induce mtDNA mutations in Saccharomyces cerevisiae, converting the cells to respiration-deficient, cytoplasmically inherited petite phenotype mutants (29). This phenotype ([rho-]) usually results from large deletions introduced in wild-type mtDNA by a mechanism which is still obscure. Upon the removal of EtdBr from the medium, the mutated molecules are replicated and possess a buoyant density different from that of wild-type mtDNA (25). On long-term treatment with the drug, respiration-deficient petites devoid of mtDNA ([rho°] phenotype) are produced (14). In the present article, we report a study on the effect of EtdBr on the mtDNA content of CEF as a function of time. It is demonstrated that EtdBr induces a population of CEF which resembles S. cerevisiae of the [rho0] phenotype.

The phenanthridine dye ethidium bromide (EtdBr) is a well-known inhibitor of mitochondrial DNA (mtDNA) replication and transcription (22, 37). As yet, no study dealing with the long-term effect of the drug on the mtDNA content of vertebrate cells in culture has been reported, mainly because EtdBr limits the growth capacity of these cells to between one and four cell generations (23, 36). Wiseman and Attardi (36) have demonstrated that the growth of human cells in the presence of the drug for ca. three cell doublings reduced the mtDNA content 10-fold, suggesting an arrest of synthesis and progressive dilution of the number of mtDNA copies per cell. On the other hand, it is well documented in cultured vertebrate cell studies that the drug induces, with continuing cell proliferation, the formation of mitochondria with bizarre shapes and reduces substantially their cytochrome b and aa3 content (15, 32). Upon the removal of EtdBr from the culture medium, however, mtDNA synthesis resumes (36), and apparently normal, respiration-competent mitochondria are found in succeeding cell generations (24, 32). In contrast to the other vertebrate cells studied so far, chicken embryo fibroblast (CEF) populations have been found to be inherently resistant to the growth-inhibitory effect of EtdBr when supplied with exogenous pyrimidine nucleosides (21) or nucleotides (13). CEF exposed to EtdBr for ca. 15 days were deficient in cytochrome c oxidase activity and devoid of a functional respiratory chain (19, 21). Furthermore, when the cells were transferred to EtdBr-free medium for 12 generations, they were found to remain respiration deficient. Two possibilities were considered to explain these observations. On the one hand, CEF exposed to EtdBr are unable to replicate mtDNA, so that after successive cell divisions, the number of mitochondrial genomes per cell is severely reduced. Upon the removal of

MATERIALS AND METHODS Cell culture. CEF were prepared from 8- to 9-day-old White Leghorn embryos (Spafas) and passaged as described previously (20). The basic medium for culturing the cells was Ham F12 supplemented with 8% calf serum, 2% inactivated chicken serum, and 4 ,ug of uridine per ml. Penicillin (100

* Corresponding author. t Present address: Centre International de Recherches Mddicales de Franceville, Franceville, Gabon.

1163

1164

DESJARDINS, FROST, AND MORAIS

IU/ml), streptomycin (100 ,ug/ml), and fungizone (0.5 ,ug/ml) were routinely added to the culture medium. The concentration of EtdBr used throughout this study was 0.4 ,ug/ml. Where indicated, the number of cell generations was calculated as previously reported (20). Cytochrome c oxidase, cytochrome spectra, and protein determinations. These assays were performed as previously described (20). Any minor modifications introduced are described in the appropriate table footnotes and figure legends. Labeling of mtDNA with [3H]thymidine. Cells were grown to a density of 2 x 107 to 2.5 x 107 cells per 150 cm2. The monolayer was washed twice with thymidine-free Ham F12 medium, and 5 ml of incubation medium made up with thymidine-free medium and dialyzed calf and chicken sera was added per flask. The cells were incubated at 39°C for 2 h in the presence of 20 ,uCi of [3H]thymidine (78.2 Ci/mmol) per ml. Incubation was terminated by aspirating the medium and rinsing the monolayer twice with phosphate-buffered saline containing 0.01 M thymidine. Cells were then trypsinized and washed twice with phosphate-buffered saline. Pelleted cells (ca. 108 cells) were suspended in 2 ml of 0.01 M Tris-hydrochloride (pH 7.2)-i mM EDTA. After standing on ice for 20 to 30 min, the cells were homogenized with a Potter-Elvehjem homogenizer. Between 20 to 30 up-anddown strokes were usually sufficient to break nearly all of the cells. Two milliliters of 0.01 M Tris-hydrochloride (pH 7.2)-i mM EDTA-0.5 M sucrose was then added to bring the sucrose molarity to 0.25. Cells were fractionated by differential centrifugation as previously described (17), and the mitochondrial fraction was further purified on a sucrose step gradient (30). Radioactivity was determined on samples of total homogenate and purified mitochondria as previously described (20). Analysis of labeled mtDNA on CsCI-EtdBr gradient. The mitochondrial fraction obtained by differential centrifugation was resuspended in 2 ml of 5 mM Tris-hydrochloride (pH 7.2)-1.5 mM MgCI2-50 mM NaCl-0.25 M sucrose and treated with DNase I (100 ,ug/ml) and RNase A (100 ,ug/ml) for 15 min at 37°C. After pelleting, the fraction was washed twice, lysed with 1% sodium dodecyl sulfate-0.01 M EDTA-0.01 M Tris-hydrochloride (pH 7.4), and the proteins were digested with pronase (75 jig/ml; self-digested for 2 h at 37°C) at 37°C for 30 min. mtDNA was separated from degraded nuclear DNA by centrifugation through a two-step CsCI-EtdBr gradient essentially as described previously (34). Radioactivity was determined from 3-drop fractions collected from the bottom of the tube and counted by liquid scintillation. Preparation of mtDNA from chicken liver. White Leghorn chickens, 3 to 6 weeks old, were fasted overnight and decapitated. A liver homogenate was prepared as previously described for rat liver (8). Mitochondria were isolated by differential centrifugation (17), and mtDNA was extracted and centrifuged through a two-step CsCI-EtdBr gradient as indicated above. Fractions (6 drops) were collected from the bottom of the tube, and a portion of each fraction was electrophoresed on 0.7% agarose gels (type I, low EEO, Sigma Chemical Co., St. Louis, Mo.). After staining with EtdBr, fractions showing only mtDNA were pooled, extracted with n-butanol, and dialyzed overnight against 1 mM EDTA-10 mM Tris-hydrochloride (pH 7.4). The mtDNA concentration was determined by optical density, estimated by electrophoresis, or both on 0.7% agarose gels with various concentrations of X DNA as reference standards. Extraction of total cellular DNA. Trypsinized cells (1.5 x 108 to 4 x 108 cells) were washed twice with phosphate-buff-

MOL. CELL. BIOL.

ered saline and kept at -40°C until used. Cell pellets were suspended in 5 volumes of 0.15 M NaCl-10 mM EDTA (pH 7.0) and lysed in 1% sodium dodecyl sulfate at 60°C for 15 min. The lysate was incubated for 2 h at 37°C with self-digested pronase (200 ,ug/ml), deproteinized twice with chloroform-isoamyl alcohol (24:1), and precipitated overnight at -20°C with 2 volumes of 95% ethanol. The precipitate was dissolved in 0.05 M Tris-hydrochloride (pH 7.7)-0.2 M NaClI-.01 M EDTA and incubated with self-digested pronase (100 p.g/ml) for 2 h at 37°C. After deproteinization and ethanol precipitation as indicated above, the DNA was solubilized in 1 mM EDTA (pH 7.0). The purity of the DNA preparations was monitored by the measurement of optical density and by the diphenylamine method (11). The 260 nm/280 nm ratio of the DNA preparations varied from 1.94 to 2.00. For the measurement of DNA-DNA reassociation kinetics, the DNA preparations were treated with 0.3 M NaOH for 6 h at 37°C. The DNA was precipitated with ethanol, solubilized in 1 mM EDTA (pH 7.0), and depurinated with 0.25 M HCl for 15 min. The solution was then made alkaline and neutralized before the fragments were precipitated with ethanol. The DNA was solubilized in 1 mM EDTA (pH 7.0) to give 6 mg/ml as measured by optical density. For Southern blot hybridization, the DNA preparations were treated with 200 ,ug of heat-treated RNase A per ml (15 min at 100°C) for 2 h at 37°C. After deproteinization and ethanol precipitation, the DNA was solubilized in 1 mM EDTA (pH 7.0) to give 6 mg/ml. Extraction of Escherichia coli DNA. DNA was extracted from E. coli ED-8767 and processed for DNA-DNA reassociation kinetics and Southern blot hybridization essentially as described above. Preparation of mtDNA probes for measurement of DNADNA reassociation kinetics and for Southern blot hybridization. mtDNA was labeled with [32P]dCTP by nick translation (26). mtDNA (20 to 40 ng) was diluted in 50 RI of solution containing 0.1 M Tris-hydrochloride (pH 7.8), 10 mM MgCI2, 20 mM P-mercaptoethanol, 100 ,ug of bovine serum albumin per ml, 0.25 nmol each of dATP, dGTP, and dTTP, and 1 to 25 ,uCi of [a-32PP]dPCTP (3,200 Ci/mmol; New England Nuclear Corp., Boston, Mass.). The reaction was started by adding 10-3 ,ug of DNase I (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) and 1.8 U of E. coli DNA polymerase I (Boehringer Mannheim). The reaction mixture was incubated at 16°C, and portions were taken at intervals to measure the incorporation of labeled nucleotides into trichloroacetic acid-precipitable material. The reaction was stopped by adding 200 p.1 of 0.01 M Tris-0.01 M EDTA (pH 7.0). After deproteinization with chloroform-isoamyl alcohol (24:1), the unincorporated nucleotides were removed by gel filtration on Sephadex G-50 (0.8- by 25-cm column) with 0.01 Tris-0.01 M EDTA (pH 7.0). The labeled fragments were precipated with ethanol and solubilized in 0.01 M Tris-hydrochloride (pH 7.4)-i mM EDTA. The specific activity of the probe ranged from 1 x 107 to 6 x 108 cpm/,ug. The labeled fragments were ca. 500 nucleotides in length as determined by agarose gel electrophoresis. Measurement of DNA-DNA reassociation kinetics. Reassociation kinetics were measured as described by Sharp et al. (27). Denatured 32P-labeled, nick-translated mtDNA was reassociated in the presence and absence of total cellular DNA. A range of 0.1 to 0.6 ng of mtDNA probe was mixed with 0.2 to 30 ,ug of sheared cellular DNA in 18.5 ,u1 of 0.2 M potassium phosphate buffer (pH 6.8)-i M NaCI. Each reaction mixture was completed to 30 ,ug of total DNA with E. coli DNA. Before annealing, the DNA mixtures were dena-

VOL. 5, 1985

CHICKEN EMBRYO FIBROBLASTS WITHOUT MITOCHONDRIAL DNA

tured by incubation with 0.4 M NaOH for 15 min at room temperature. After neutralization, the reaction mixture was overlaid with paraffin and quickly brought to the incubation temperature (60°C). At time zero and at predetermined intervals thereafter, samples (5 ,ul) were taken, immediately diluted in 500 ,ul of ice-cold 0.01 M potassium phosphate buffer (pH 6.8), and frozen at -20°C before analysis of the reannealed fraction by hydroxylapatite (Bio-Rad Laboratories, Richmond, Calif.) chromatography in the presence of formamide (1). Single-stranded and double-stranded DNA were eluted with 0.09 M potassium phosphate buffer-50% formamide (pH 7.9) and 0.4 M potassium phosphate buffer-50% formamide (pH 7.0), respectively. The effluents were counted by Cerenkov radiation, and the percentage of reassociation was obtained by dividing the counts per minute in the double-stranded fraction by the total counts per minute of the sample. The kinetics of reassociation were found to follow the equation C/C0 = 1/(1 + K cot) in which C is the concentration of unreassociated DNA, C0 is the initial concentration of DNA, t is time, and K is the reassociation constant (5). The time taken for half of the DNA sequences to enter hybrid (t1/2) was calculated from the equation 1/fss = (4t1/2) + 1 (27); these values were used to determine the number of mtDNA copies per cell. The molecular weight of nuclear DNA was taken as 1.5 x 1012 (2.5 pg of DNA per diploid cell), and that of mtDNA was 1.1 x 107 (4). Southern blot hybridization. Restriction enzymes were purchased from Boehringer Mannheim and were used as recommended by the supplier. Restriction enzyme digests were analyzed by gel electrophoresis on 0.7% agarose gels in 0.04 Tris-acetate-2 mM EDTA (pH 8.0) containing 0.5 p.g of EtdBr per ml. Electrophoresis was carried out in a horizontal slab gel apparatus at 25 V for 17 h. The size of the restriction fragments was estimated by comparison with HindlIl digests of X DNA. After electrophoresis, the gel was soaked twice for 15 min in 0.25 M HCI at room temperature (35). The DNA was then denatured by soaking the gel in several volumes of 0.4 M NaOH for 30 min. After neutralization of the gel in several volumes of a 0.5 M Tris-hydrochloride (pH 7.4)-3 M NaCl solution for 30 min, the gel was blotted onto nitrocellulose paper (BA85; Schleicher and Schuell, Inc., Keene, N.H.) by the method of Southern (33). The nitrocellulose filters were moistened with 5 x SSC (1 x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and prehybridized for 2 to 4 h at 62°C in solution containing 200 ,ul of 5x SSC per cm2, 0.1 M potassium phosphate buffer (pH 6.8), and 5 x Denhardt solution (9). Hybridization was carried out overnight at 62°C in the same buffer (50 pU.1cm2) with 40 ng of heat-denatured mtDNA probe. The filters were washed several times at room temperature in 2x SSC-0.1% sodium dodecyl sulfate and several times at 45°C in 0.1x SSC-0. 1% sodium dodecyl sulfate with constant gentle agitation. Filters were blotted dry and autoradiographed at -70°C with Kodak X-Omat RP film and a DuPont Cronex Lightning-Plus intensifying screen. RESULTS AND DISCUSSION Preliminary experiments carried out to demonstrate the inhibitory effect of EtdBr on mtDNA synthesis in CEF revealed that, in cells preincubatd with the drug for 30 min, the extent of incorporation of [3H]thymidine into acid-insoluble material after a 2-h pulse was reduced by 20% in whole cells and by 75% in the mitochondrial fraction purified on a sucrose step gradient. Analysis on CsCI-EtdBr gradients of the material extracted from the mitochondrial fraction ob-

ilii

I

1165

, II

B

l

76-

5-

43-

2-

0

10

20

40 30 FRACTIONS

50

60

70

FIG. 1. Sedimentation patterns in two-step CsCI-EtdBr gradients of DNA extracted from the mitochondrial fraction of CEF pulse-labeled for 2 h with [3H]thymidine. (A) EtdBr- treated cells were preincubated with the drug for 30 min before the addition of [3H]thymidine. The mitochondrial fraction was prepared from 1 x 108 control and 1.2 x 108 EtdBr-treated cells. (B) Cells were treated (A) or not treated (0) with EtdBr for 20 days and then cultivated in drug-free medium for 25 cell generations. The mitochondrial fraction was prepared from 1.1 x 108 control and 7 x 107 drug-treated cells. Also shown are oligomeric (I) and monomeric closed circular (II) and monomeric open circular (III) mtDNA molecules.

tained from control and EtdBr-treated cells supported these observations (Fig. 1A). For control cells, three bands were usually seen in the lower portion of the gradient, and analysis by electron microscopy of the material present in these bands indicated that it consisted mainly of oligomeric and monomeric closed circular and monomeric open circular mtDNA molecules. With 4)X174 DNA as an internal standard, the contour length of the molecules was found to be 5.04 ± 0.03 p.m (25 molecules analyzed), in agreement with values reported by others for chicken mtDNA (3). In the upper part of the gradient, the band consisted of degraded nuclear DNA and possibly also of some degraded mtDNA (7, 34). On 0.7% agarose gels, the latter material was found to migrate much faster than circular mtDNA. EtdBr strongly inhibited the incorporation of [3Hlthymidine in the mtDNA

1166

MOL. CELL. BIOL.

DESJARDINS, FROST, AND MORAIS TABLE 1. Effect of EtdBr on mtDNA copy number of CEF mtDNA copy no./cell'

EtdBr treatment (days)

0 3 6 13 20

Days in EtdBr-free medium

End of

treatment

14

28

604 134 (11) 79 29 (6) 9 2 (4) 0.5 ± 0.2 (2) 0.4 ± 0.3 (7) 0.1 ± 0.1 (5)

209 132 (6) 25 24 (4) 1.6 ± 1.2 (3) 0.1 ± 0.1 (2) ND

242 ± 68 (5) 37 ± 30 (2) 1.2 ± 0.7 (2) 0.9 ± 1.0 (4)

42

412 ± 92 (2) 116 (1) 4(1) 0.1 (1) ND

43b ND "Values represent the mean plus or minus standard deviation of individual experiments. The number of experiments is given in parentheses. ND, Not done. b Mean value from cell populations cultivated in EtdBr-containing medium for 35 to 52 days.

molecular forms sedimenting in the lower portion of the gradient. Assuming that the recovery of mtDNA was the same for both control and EtdBr-treated cells, the labeling of the mtDNA forms was at least 90% inhibited by the drug. Similar observations have been reported by several investigators in mouse, human, and hamster cell lines at EtdBr concentrations ranging from 0.1 to 1.0 g/ml (18, 23, 30). When cells were treated with EtdBr for 13 and 20 days, respectively, and then transferred to drug-free medium for various periods of time (19 to 55 days), the extent of incorporation of [3H]thymidine in whole cells was found to be comparable to that of control cells. In sharp contrast, the amount of [3H]thymidine present in the purified mitochondrial fraction of the drug-treated cells was 7 and 3.5% of that of control cells. Analysis of the material extracted from mitochondrial fractions on CsCl-EtdBr gradients supported these observations. Cells treated for 20 days and then transferred to drug-free medium for 35 days appeared unable to incorporate labeled thymidine into circular mtDNA-like molecules. (Fig. 1B). These results were interpreted to indicate that very little if any synthesis of mtDNA was taking place in CEF after these cells had been exposed to EtdBr for long periods of time. mt DNA content of CEF treated with EtdBr. To investigate the effect of EtdBr on the mtDNA content of CEF, total cellular DNA was extracted from control and EtdBr-treated cells, and the mtDNA content was measured by reassociation kinetics. An alternative method of quantitation based on the prior isolation of mitochondria, followed by the extraction of mtDNA (2), was excluded mainly on the basis of variation in the recovery of mitochondrial proteins from respiration-deficient CEF. The ultrastructure of mitochondria from EtdBr-treated cells has been shown to be disorganized (19), and it is our experience that these organelles are more fragile than those from normal CEF. The mtDNA content of control cells was determined in populations maintained in culture for periods of 7 to 45 days. TABLE 2. Cytochrome EtdBr treatment

EtdBrtreaymen (dayS)b 3 (3) 6(4) 13 (5) 20 (3)

c

A mean number of 604 mtDNA copies per cell was found (Table 1), a value slightly lower than that found in permanently aneuploid mouse fibroblast cell lines (ca. 1,000 copies [2]) and substantially lower than that found in various primary human fibroblast cell lines (ca. 2,000 to 6,000 copies [28]). Variation in the number of mtDNA copies per cell was seen in various CEF populations from various initial isolates as has been reported for human primary fibroblast cell populations (28). In fact, it was observed that the number of copies in CEF populations varied from 450 to 850. However, for a given cell preparation, the number of mtDNA copies per cell determined at early and mid-to-late cell passages was found to remain approximately the same (data not shown). This observation differs from that made by Shmookler Reis and Goldstein (28), who reported that the number of mtDNA copies per cell increases during successive passage of human primary fibroblasts. EtdBr reduced the number of mtDNA copies per cell (Table 1). It would be expected that if the replication of mtDNA were severely impaired by the drug, then after n cell doublings the amount of mtDNA per cell should be reduced by 1/2 n. Assuming a mean number of 600 mtDNA copies per cell, three, five, and nine cell generations should reduce the mtDNA content to ca. 75, 18, and 1 copy per cell. The theoretical values are close to those found in CEF treated with EtdBr for 3, 6, and 13 days, that is, after a mean 2.5, 4.5, and 8.7 cell doublings (Table 2, column 2). In cultures treated with EtdBr for 20 and 43 days, equivalent to ca. 13 and 26 cell generations, the populations should contain about one copy of mtDNA per 1 x 101 and 1 x 104 cells, respectively, values below the limit of detection of the present method. Very little or no mtDNA were detected in these cell populations. The same observation was made for cells treated for 20 days and then transferred to drug-free medium for up to 42 days. Progressive dilution of the mtDNA content, down to 10% of control cell values after three doublings, has also been

oxidase activity of respiration-deficient CEF cultivated in EtdBR-free medium"

No. of cell generations aat the end of treatment

theneratindson At thle end of

2.40 0.3 4.65 0.6 8.67 ± 1.7 13.81 ± 2.4

o

treatment

10 to 20

21 to 30

31 to 35

29.2 16 40.9 6 52.3 10 5.1 5 15.9 13 28.4 15