Biochimica et Biophysica Acta, 1138 (1992) 275-281 © 1992 Elsevier Science Publishers B.V. All rights reserved 0925-4439/92/$05.00
275
BBADIS 61130
A procedure for selecting mammalian cells with an impairment in oxidative phosphorylation M.C.P. L o m b a r d o 1,2, J.W. van der Z w a a n 1, S. Brul 1 and J.M. Tager 1,2 l E.C. Slater Institute for Biochemical Research, University of Amsterdam, Academic Medical Centre, Amsterdam (The Netherlands) and : Institute of Medical Biochemistry and Chemistry, University of Bari, Bari (Italy) (Received 19 July 1991)
Key words: CHO mutant; Mitochondrial disease; Oxidative phosphorylation; Cytochrome c oxidase; Complex IV
The mitochondriai encephalomyopathies in man are characterized by heterogeneous defects leading to an impairment in the pathway of aerobic energy production. As a means of investigating the molecular and genetic mechanisms underlying these disorders we have developed a procedure for selecting m a m m a l i a n cell lines with features resembling the h u m a n pathological phenotypes. The principle of the selection is the use of a fluorescent amphiphilic dye, 2,4-(dimethylamino)-l-styrylmethylpyridiniumiodine, a cation showing two main features. Firstly, it is accumulated by mitochondria to an extent correlated with the magnitude of the electrochemical gradient of protons across the mitochondrial inner membrane. Secondly, upon irradiation with UV light, it gives rise to formation of free radicals, which inflict damage to the cell. Mutant cells with an impairment in oxidative phosphorylation will have more chance to survive than wild type cells. The selection procedure was applied to a stock of mutagenized Chinese h a m s t e r ovary cells. After subcioning of the cells which survived the selection procedure, twenty-six independent clones were isolated. Eighteen of the clones had a partial deficiency of cytochrome c oxidase ranging from 30 to 60% of the activity in control cells. The properties of two of the clones are described. One clone has been cultured under non-selective conditions for at least 12 months with retention of the partial deficiency of cytochrome c oxidase.
Introduction The mitochondrial encephalomyopathies comprise a heterogeneous group of diseases in man that are due to defects in the pathway of aerobic energy production in the mitochondrial inner m e m b r a n e [1,2]. These pathological conditions can be divided into two groups: those in which there is a deficiency of mitochondrially encoded subunits of an enzyme and those in which the defect is apparently localized in the nuclear genome, for instance in a gene coding for a protein required for the replication of mitochondrial D N A [3].
Abbreviations: CHO, Chinese hamster ovary; cyt c, cytochrome c; DASPMI, 2,4 (dimethylamino)-l-styrylmethylpiridinium iodine; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulphate; TMPD, tetramethyl-p-phenyldiamine. Correspondence: J.M. Tager, Institute of Medical Biochemistry and Chemistry, University of Bari, Piazza Giulio Cesare, 70124 Bari, Italy.
In order to understand the molecular and genetic mechanisms underlying these disorders, a possible approach is to select mutant cells with features resembling the human pathological phenotype [4]. Such mutants have been isolated from Chinese hamster lung cells [5,6] by a method involving replica plating. In this p a p e r we describe a procedure for the isolation of cells with an impairment in oxidative phosphorylation, i.e. with an impairment in the ability to establish a respiration-dependent electrochemical potential. This procedure makes use of a selection medium in which wildtype ceils do not survive. We have used this procedure to isolate Chinese hamster ovary ( C H O ) cell lines with a deficiency in cytochrome c oxidase. In principle the method could also be used to select mutants in other segments of the respiratory chain. Principle of the selection procedure The procedure used to select mutant cells with a perturbation in oxidative phosphorylation is based on the properties of some cationic amphiphilic dyes which
276 permeate freely through the plasma membrane and accumulate specifically in mitochondria, provided that an electrochemical gradient of protons across the mitochondrial inner m e m b r a n e is present [7]. These compounds have been used for monitoring of the relative mitochondrial m e m b r a n e potential in living cells [7]. It has been shown that the accumulation of such dyes by mitochondria is correlated with the magnitude of the mitochondrial electrochemical gradient and is particularly evident in proliferating cells [7]. Such dyes may act as photosensitizers. Upon irradiation following accumulation of the dyes free radicals, which are cytotoxic, are produced. The effectiveness of the cytotoxic effect of this treatment will depend on the amount of the dye accumulated, i.e. on the magnitude of the mitochondrial m e m b r a n e potential. Indeed, such dyes have been used for selective killing of carcinoma cells in vivo and in vitro [8]. Cultured cells defective in one or more components of the oxidative phosphorylation system are expected to have a decreased capacity to build up an electrochemical potential via the respiratory chain and to have a decreased capacity to accumulate these cations. Such cells can be expected to survive this treatment better than wild-type cells. We have used these principles to develop a selection medium for isolating cells with an impairment in mitochondrial oxidative phosphorylation. For this purpose we have employed the cationic dye 2,4-(dimethylamino)-l-styrylmethylpyridiniumiodine (DASPMI). As shown by Bereiter-Hahn [9], D A S P M I accumulates specifically in mitochondria in cultured cells in response to a m e m b r a n e potential across the mitochondrial inner membrane.
Materials and Methods
Cell culture. Chinese hamster ovary ( C H O ) cells, strain CHO-9, were grown at 37 ° in a 1:1 mixture of Ham-F10 medium (Flow) and D M E M (Dulbecco's modified Minimum Essential Medium; Flow), supplemented with 5% ( v / v ) foetal calf serum, streptomycin (100 / , g / m l ) and penicillin G (100 units/ml). The gas phase was 5% C O 2 - 9 5 % air. H u m a n fibroblasts from a control subject (cell line F5035) and from a patient with a deficiency of cytochrome c oxidase also expressed in fibroblasts (cell line 83RD88) were cultured as above except that the concentration of foetal calfserum was 10% (v/v). Cell line F5035 was kindly provided by Drs. Monique Mathieu and Irene Maire (Lyon) and cell line 83RD88 by Professor Jasper Scholte (Rotterdam). Mutagenization of CHO cells and selection of mutants. CHO-9 cells were mutagenized with ethylmethane-sulphonate exactly as described in Ref. 10.
The mutagcnized cells were platcd with Ill ml nledJunl in 9.4 cm Petri dishes at a density of 10" cells per p l a t e 12 h later 1 ~ M buthionine sulphoxinline was added to the culture medium. After another 12 h, 1() # M oligomycin was added and the cclls werc cultured in this medium for another 24 h. The medium was then removed and the ceils were washed thrcc times with glucose-free Hanks' medium. The washed cells were preincubated for 2 h with glucose-free Hanks' medium (Flow) at 37 °. The preincubation buffer was replaced by fresh Hanks' supplemented with 8 mM sodium suecinate (pH 7.4). After 30 min, DASPMI dissolved in 10 mM Mops (pH 7.4) was added at a final concentration of 30 # M . Then the ceils were irradiated with longwave UV light fl)r 10 min at 12 w a t t s / m -~. The cells that survived this treatment were isolated and subsequently subcloned. Preparation of mitochondria. Mitochondrial fractions were prepared from C H O cells as described by Rickwood et al. [11] for cultured fibroblasts.
Immunoprecipitation of cytochrome c ~)xidase /?om cellular extracts. About 107 cells were suspended in 1 ml of extraction buffer (50 mM Tris-sulphate (pH 7.4), 1 mM EDTA). Laurylmaltoside was added to a final concentration of 1% (mass/vol), together with the proteinase inhibitors phenylmethylsulphonyl fluoride (25 mM), antipain (21 /,M), pepstatin (20 /,M) and leupeptin (38 /,~M). Alter sonification (3 times for 15 s at 7/.Lm amplitude, with intervals of 45 s), the samples were incubated for 31) rain on ice and centrifuged for 5 min in an Eppendorf centrifuge at 4°C. The supernatant was incubated overnight with anti-(eytochrome c oxidase) coupled to protein A-Sepharose 4B beads under rotation at 4°C. The beads were washed seven times with 50 mM Tris-sulphate (pH 7.4), 1 mM EDTA, 0.05% (mass/vol) laurylmaltoside. Cytochrome c oxidase was solubilized by incubation for 2 h at room temperature in sample buffer (4% (mass/vol) sodium dodecyl sulphate and 12% (mass/vol) glycerol in 50 mM Tris-HCl (pH 6.8)). After centrifugation, the supernatant was subjected to sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis (PAGE) and immunoblotting. Electrophoresis and immunoblotting. Electrophoresis was performed as described by Sch~igger and yon Jagow [12]. Western blotting was performed using a semi-dry system (W.E.P. C o m p a n y , Seattle, Washington; Electr-Trans blotter, model 198155) in the presence of transfer buffers according to Kyhse-Anderson [13] but omitting 6-amino-n-hexanoic acid. The nitrocellulose filters were treated as described by Sinjorgo et al. [14]. Immunoreactive polypeptides were visualized using the immunoperoxidase method with 4-chloro-l-naphthol as substrate. Preparation of rabbit antisera against ~ytochrorne c oxidase. The antisera used in this study had been raised
277 in rabbits against cytochrome c oxidase purified from human heart muscle [14]. A TP content of cells. A T P was determined fluorimetrically in an aqueous extract of cells as described in [151. Assay of activities. In order to measure cytochrome c oxidase activity, cells were incubated for 15 rain at 0 ° in a solution containing 15 mM s o d i u m / p o t a s s i u m phosphate (pH 7.4) and 1% ( m a s s / v o l ) laurylmaltoside. After centrifugation for 5 min at 10000 r e v s / m i n at 4°C, enzymic measurements were performed at room temperature in fractions of the supernatant by pulsing reduced horse cytochrome c (15 /xM final concentration) as substrate and following the decrease in absorption of the sample at 550 nm. Cytochrome c oxidase was also assayed polarographically in digitonin-permeabilized cells with tetramethyl-p-phenylenediamine (TMPD)-ascorbate as substrate exactly as described by Schoonen et al. [16]. The K m of cytochrome c oxidase for cytochrome c was measured at high ionic strength with horse cytochrome c as substrate. Succinate dehydrogenase and succinate : cytochrome c oxidoreductase were assayed in mitochondrial preparations obtained from C H O cells as described in Refs. 17 and 18, respectively. Citrate synthase was assayed in cellular extracts by measuring the appearance of free SH groups, as monitored by reduction of 5,5'-dithiobis-(2nitrobenzoate), after addition of acetyl-CoA and oxaloacetate [19]. Lactate dehydrogenase was assayed in cell extracts as described [20]. Catalase activity was measured polarographically according to Ref. 21.
Lactate and pyruvate formation from glucose (see Ref 22). Wild-type cells and mutant cell lines were
+
Buthionine
sulphoximine
90 80 .-&
70
t~
60
t-
50 40 30
+ UV
20 10 0
0
I
!
!
I
20
40
60
80
DASPMI
( ~ M)
Buthionine
A
100
sulphoximine
90
80 70 rJ 113 ¢-
> -i ¢/)
60 50 40
+ UV
30 ~=
.
20 10 0 0
I
I
I
I
20
40
60
80
DASPMI
lO0
( ~. M)
assayed for their ability to form lactate and pyruvate from glucose 3 days after reaching confluency. Production of lactate and pyruvate from glucose (5 mM) was followed at intervals of 1 h in 270 /zl aliquots of the incubation medium neutralized to p H 6 - 7 after removal of proteins by perchloric acid precipitation. Cells were harvested with 0.1 M N a O H to assay the protein content according to Lowry et al. [23].
Fig. 1. Effect of treatment with DASPMI, UV light and buthionine sulphoximine on viability of wild-type Chinese hamster ovary cells as assessed by colony formation. For conditions see Materials and Methods. The number of surviving cells (colonies) is expressed as a percentage of the number surviving in control dishes in which DASPMI and UV treatment were omitted. In the absence of DASPMI, treatment with UV led to a decrease in survival of about 50%.
Results
respiratory chain might not exhibit a decreased electrochemical potential because of a compensating effect of A T P generated during glycolysis, which can energize mitochondria by reacting with A T P synthase [24]. We therefore included oligomycin, which inhibits A T P synthase, in our selection medium. Fig. 1 shows the results of an experiment with Chinese hamster ovary ( C H O ) cells. In the absence of UV treatment, D A S P M I had only a small effect on the survival of the cells; at concentrations of D A S P M I from 5 /zM up to 30 /xM, survival decreased to a minimum value of 70% and then remained constant as the concentration of D A S P M I was increased further. T r e a t m e n t with U V light led to a 50% decrease in
Setting up the selection procedure We incubated C H O cells in culture with D A S P M I in the presence and absence of the respiratory chain inhibitors antimycin and cyanide. No evident difference in fluorescence of the mitochondria could be seen between the two conditions. This is on the one hand in contrast with studies previously performed using mitochondrial suspensions, where respiratory inhibitors clearly decrease the fluorescence of the mitochondria [7], but on the other hand it is in accordance with data obtained in living cells [24]. Starting from this observation, we expected that intact cells with a defective
278 TABLE I t".J~bct ~Tf'treatment with D A S P M I and ultrat:iolet light on the .w~rt il al o['fibn#~lasts from a control suIzk'et and a patient with a d~[Teienev o[ O, tochrome c oxidase (_'ell line F5035 from a control subject was kindly supplied by Drs. Monique Mathieu and Irene Maire (Lyon) and cell line 83KD88 from a patient with a generalized deficiency of cytochrome e oxidase was kindly supplied by Professor H.R. Scholte (Rotterdam). The activity of cytochrome c oxidase was 28(1 nmol cytochrome e cons u m e d / m i n per mg protein in cell line F5035 and 56 nmol cytochrome c c o n s u m e d / m i n per mg protein in cell line 83RD88 (J.W. van der Zwaan and H.R. Scholte. unpublished observations). Values in parentheses are percentages of the number of cells which survived in the absence of DASPMI and UV irradiation. The cells were cultured in the presence of 1 p.M buthionine sulphoximine, in the absence or presence of 10 s M oligomycin and treated further as indicated in the table. Conditions of culture
In the absence of oligomycin no treatment DASPMI 30/J.M UV irradiation DASPMI 30/~M + UV irradiation In the presence of oligomycin no treatment DASPMI30p.M UVirradiation DASPMI 3 0 / ~ M + U V irradiation
Number of surviving cells control (cell line F5035)
patient (cell line 83RD88)
1.6-10 ~' (100) 1.25.106 (78) 0.80.106 (50)
0.45.10 (' (100) 0.49.106 (108) 0.28-106 (62)
0.29.10 ~ (18)
0.20- 10 ¢' (44)
0.80.10 ~' (50) 0.65.10 ~ (41) 0.31.106 (19)
0.21 • 106 (47) 0.17.10 ~' (38) 0.14.106 (31)
I).05.10 ~
0.18.1() c' (40)
(3)
survival. When the cells were treated with D A S P M I prior to exposure to U V light, the survival decreased further, to a minimal value of 20%. Survival was even lower (maximally 10%) when cells were depleted of glutathione by culturing them with buthionine sulphoximine (see below).
patient was about the same as in Ihe absence ¢~1 oligomycin, but only about 5% of the control cells survived the treatment in the presence of oligomycin.
Isolation of CHO mutants C H O cells (strain CHO-9; see Ref. 25) mutagenized with ethylmethane-sulphonate were plated out in Petri dishes at a density of 105 cells per plate. Buthionine sulphoximine was subsequently added, in order to inhibit 7-glutamylcysteine-synthetase and thereby to decrease the glutathione content of the cells [26]. Glutathione normally acts as scavenger ti~r free radicals {27]. In this way the cells became more sensitive to the presence of free radicals. After treatment with DASPMI and irradiation one or two cells per Petri dish survived and were able to form colonies. The colonies were picked up and cultured separately. Colonies were subcloned by limiting dilution and 26 clones were obtained, each of which was derived from a separate Petri dish. The activity of cytochrome c oxidase in the mutants varied from 28-101% of that in the wild-type cells. In 18 of the clones, the activity of cytochrome c oxidase was 60% or less than that of the wild-type cells. The activity of citrate synthase and catalase did not differ significantly from that in the control cells. Some of the clones were analysed by means of immunoblotting experiments with an anti-(human cytochrome c oxidase) antiserum. Fig. 2 shows that the amount of crossreactive material corresponding to subunits II + 11I was lower in the mutant cells tested, particularly in cell line 13 and cell line 8 (see Fig. 2, lane c and e). In addition, there were marked differences in the electrophoretic mobility and the amount of crossreactive material corresponding to the nuclearly encoded subunits V, VI, VII and Vlll. Such differences were much less marked in mutant cell lines 1 and 2 (Fig. 3, lanes d, e, f and g, respectively). Further investigations were performed on the properties of cell lines I and 2.
Control experiment with human fibroblasts In order to test whether our procedure was selective for cells defective in the aerobic energy-producing pathway, an experiment was performed with fibroblasts derived from a control subject and from a patient with a deficiency in the cytochrome c oxidase segment of the respiratory chain. Cells were plated out and subjected to the treatment used for the selection of C H O cells (see below). Table I shows that, when control cells were cultured in the absence of oligomycin, treatment with D A S P M I followed by U V light decreased the survival of the cells to about 20%, whereas about 45% of the cells from the patient survived. The difference was even greater when oligomycin was present in the culture medium. The survival of the cells from the
Characterization of the phenotype of cell lines 1 and 2 The mutant phenotype consists in both cases of a decreased activity of cytochrome c oxidase, whereas the K m for cytochrome c is equal to that in thc control. The residual activity in clone 1 and clone 2 was about 40 and 30%, respectively, in comparison with the control cell line CHO-9 (Table II). These spectrophotometric assays were corroborated by measurements of oxygen uptake in digitonin-permeabilized cells with TMPD-ascorbate as substrate. Further analysis of the respiratory chain was carried out by isolating mitochondria from the cells and using them for spectrophotometric tests in the presence of substrates acting at specific levels of the respiratory chain. Succinate dehy-
279 a
b
c
d
e
f
g
h
T A B L E lI
Activity of cytochrome c oxidase, citrate synthase and catalase in wild-type and mutant CHO cell lines
-
at
dln~
I
T h e mutant clones were isolated from a stock of mutagenized C H O cells using the procedure described in the text for selection of mutants with an impairment in mitochondrial oxidative phosphorylation. For experimental details see Materials and Methods.
II, III
Cell lines
C
C
Fig. 2. Electrophoretic pattern of subunits of cytochrome c oxidase in extracts of Chinese hamster ovary cells. Extracts of cells were subjected to S D S - P A G E and immunoblotted with anti-(human cytochrome c oxidase) as described in Materials and Methods. Lane a, molecular weight markers; lane b, wild-type (107 cells); lane c, clone 13 (14- 106 cells); lane d, clone 10 ( 1 3 106 cells); lane e, clone 8 (10' cells); lane f: clone 22 (13.5.106 cells); lane g, clone 4 (12.106 cells); lane h, cytochrome c oxidase isolated from beef heart.
Activity of
Wild-type Clone 1 Clone 2 Clones 3 - 5 Clones 6 - 8 Clones 9-12 Clones 13-18 Clones 19-22 Clones 23-25 Clone 26
cytochrome c oxidase a
citrate synthase ~'
catalase h
495 _+35 c (50) 230_+ 12 c (32) 175-+ 15 c (45) 140-181 200-209 231-268 278-298 304-361 402-475 500
45 _+4 c (2) 45_+6 c (2) 41-+2 ~ (2) 43-44 41-43 45-46 40-44 42-46 42-46 45
35 _+3 c (2) 3 5 + 3 c (2) 3 2 + 2 c (2) 30-36 31-36 30-35 30-40 33-37 35-39 33
Specific activity in nmol of substrate consumed per min ~ per mg protein ~. b Specific activity in nmol of oxygen produced per rain ~ per mg protein ~. c Means_+ S.D. The n u m b e r of m e a s u r e m e n t s performed using separate cell cultures is shown in parentheses.
d r o g e n a s e and succinate : c y t o c h r o m e c reductase activities were found to be normal in both the mutant clones.
T A B L E IIl
Biochemical characteristics of wild-type CHO cells and clones 1 and 2 Clones 1 and 2 were isolated from a stock of mutagenized C H O cells using the procedure described in the text for selection of mutants with an impairment in mitochondrial oxidative phosphorylation. For experimental details see Materials and Methods. Values are means_+ S.D. with the n u m b e r of m e a s u r e m e n t s performed using separate cell cultures shown in parentheses. Parameter Cytochrome c oxidase -activity a -K m ( ~ M cyt.c) Oxygen consumption b (ascorbate/YMPD) Succinate dehydrogenase ~' Succinate : cytochrome c reductase c Lactate dehydrogenase a fl-hexosaminidase a Lactate in medium (raM) d Pyruvate in m e d i u m (raM) J L a c t a t e / p y r u v a t e ratio J A T P content ( n m o l / m g protein) " b c a
Wild-type
Clone 1
Clone 2
495 40 2.2
230 __+12 (32) 42 -+ 5 (4) 0.95 -+ 0.3 (2)
175 38 0.6
47.6 84.5
__+35 (50) -+ 5 (5) + 0.2 (2) _+ 1.5 __+ 8
1.97 _+ 23 i 2.48 _+ 0.133_+ 19 _+ 0.24 ___
0.16 3 0.4 0.02 1 0.04
(3) (2)
49.1 82.5
(3) (2) (2) (2) (2) (2)
2.01 _+ 22 + 3.71 _+ 0.098_+ 38 _+ 0.25 ±
Specific activity in nmol of substrate consumed m i n - 1 mg protein - 1 Specific activity in /.Lmol of oxygen consumed h - 1 mg protein i. Specific activity in nmol of cytochrome c reduced m i n - 1 mg protein 1. Values after 4 h of incubation with glucose.
_+ 1 _+ 12 0.15 4 0.5 0.03 1 0.08
-+ 15 (45) + 6 (3) + 0.1 (2)
(3) (3)
50.3 85.2
_+ 1.6 __+ 9
(3) (2) (2) (2) (2) (2)
2.21 _+ 25 _+ 2.82 _+ 0.121_+ 23 _+ 0.29 _+
0.2 3 0.1 0.02 2 0.04
(3) (3) (3) (2) (2) (2) (2) (2)
280 a
b
c
d
e
f
g
h
b ....
.~..,-..
~---:: _ _ -
-7:
--
vla,b,c
--
Vlla,b,c
Fig. 3. Electrophoretic pattern of subunits of cytochrome c oxidase in wild-type Chinese hamster ovary cells and in clones 1 and 2. Extracts of cells were subjected to SDS-PAGE and immuno- blotted with anti-(human cytochrome c oxidase) as described in Materials and Methods. Lane a, molecular weight markers; lanes b and c, wild-type (10 7 cells); lanes d and e, clone 1 (11.106 cells); lanes f and g, clone 2 (107 cells); lane h, cytochrome c oxidase isolated from human heart.
Formation of lactate and pyruvate from glucose by the cells was also examined (Table III). In clone 1 the lactate/pyruvate ratio was twice as high as in control cells. In this respect, clone 1 resembles some human myopathic conditions in which an elevated l a c t a t e / pyruvate ratio is found [28,29]. This finding is a consequence of the increase in both the mitochondrial and the cytosolic N A D H / N A D + ratio due to a block in mitochondrial N A D H oxidation, which leads to an increased formation of lactate from pyruvate via lactate dehydrogenase [22]. In the case of clone 2 there was a slight increase in the lactate production and a slight decrease in the concentration of pyruvate in comparison with the control, but the ratio lactate/pyruvate was not significatively increased. The intracellular level of ATP in clones 1 and 2 was found to be comparable with the control level (Table III). There was no change in the activity of lactate dehydrogenase, catalase and /3-hexosaminidase, marker enzymes for the cytosol, the peroxisomes and the lysosomes, respectively, in clone 1 and 2. In addition, the activity of citrate synthase, a mitochondrial matrix enzyme, was normal in the two mutants, indicating that there is no generalized impairment of mitochondrial functions in the mutants. Discussion
The procedure described in this paper provides a means of selecting cells with a deficiency in the ability
to maintain a respiration-dependent electrochemical gradient of protons. The selection medium contains succinate as a respiratory substrate, since DASPMI inhibits N A D H dehydrogenase [9], and oligomycin, lo prevent energization of mitochondria by ATP generated during glycolysis. Thus the electrochemical gradient of protons built up across the mitochondrial men> brane under these conditions is dependent on electron flow from succinate to O~ and includes the cytochromc c oxidase step. During subsequent culture of the cells in the absence of selection medium N A D H dehydrogenase will, of course, not be inhibited. In 18 of the 26 clones isolated from a stock of mutagenized C H O cells there was a decrease in the activity of the enzyme to 60% of the control value or less. In one of the primary cultures derived from a colony of cells which survived the selection medium cytochrome c oxidase was almost completely deficient. The cells are slow-growing, however, and subcloning is still in progress. The stability of the mutations in two of the clones ( 1 and 2) was followed in time by keeping the cells in culture in the absence of DASPMI, UV light, buthionine sulphoximine or other stress conditions. In the case of clone 1 the phenotype was maintained for 8 months (corresponding to about 480 cell divisions), after which reversion to the wild-type phenotype occurred. Clone 2 has now been kept in culture for 12 months without reversion to the wild-type phenotype. Clearly, the phenotype of clones 1 and 2 is not due to a direct effect of D A S P M I on cytochrome c oxidase activity or on the assembly of the enzyme. The cellular ATP content of clones 1 and 2 was the same as that of control cells. In clone I the lactate/ pyruvate ratio in the medium was increased in comparison with the control cells. Thus in this clone glycolysis may play an important role in maintaining the ATP level in the cells. In clone 2, on the other hand, the lactate/pyruvate ratio was about the same as in the controls. Schoonen et al. [16] have recently shown that exposure of C H O cells to normobaric hyperoxia, which can be expected to increase the concentration of free radicals, leads to respiratory failure due to inactivation of N A D H dehydrogenase, succinate dehydrogenase and 2-oxoglutarate dehydrogenase. Interestingly, hyperoxia had no effect on ubiquinone:cytochrome c reductase (complex IlI), cytochrome c oxidase (complex IV), and ATP synthase (complex V). Moreover, in a control experiment performed by us with human fibroblasts (Table I), the residual activity of cytochrome c oxidase in the cell line derived from a patient affected by mitochondrial myopathy was the same before and alter submission to our selection procedure. We can therefore exclude the possibility that the deficiency arose from a damaging action of free radicals on the mere-
281 brane structure which could disturb the assembly of the enzyme. In clones 23-26 the cytochrome c oxidase activity was 80-100% of that found in wild-type cells. Thus some other characteristic of these cells must have enabled them to survive the selection procedure, for instance an impairment in the proton-pumping activity of cytochrome c oxidase or a defect in some other segment of the respiratory chain. A third possibility is that the cells are resistant tot buthionine sulphoxamine or oligomycin due for instance to increased expression of P glycoprotein which is responsible for multidrug resistance in mammalian cells [30]. In order to test whether the phenotype of clones 1 and 2 was due to hypermethylation of DNA rather than to mutations in specific genes, the cells were subjected to treatment with 5-azacytidine, a hypomethylating agent [31]. Surprisingly, treatment with 5azacytidine led to a 2-3-fold increase in cytochrome c oxidase activity in clones 1 and 2 (M.C.P. Lombardo, unpublished observations). Moreover, a similar increase in activity of the enzyme was observed in wildtype cells. This phenomenon may be due to activation by demethylation of nuclear genes for isoforms of cytochrome c oxidase not normally expressed in CHO cells or of nuclear genes affecting the expression of mitochondrial genes for cytochrome c oxidase. The general features of the organization of the nuclear and mitochondrial genes required for the biogenesis of mitochondria and the very high number of copies per cell of the mammalian mitochondrial genome [32] require concomitant regulation of expression of the genes and very tight nuclear-mitochondrial interactions. The comprehension of these interactions, which are very important in the biogenesis of mitochondria, would be facilitated by analysis of cells having a perturbation in oxidative phosphorylation. Moreover, the characterization of these mutant strains can contribute to an understanding of the functional role played by the supernumerary subunits present in segments of the mitochondrial respiratory chain as well as of the meaning of the differential expression of nuclear genes during differentiation [33].
Acknowledgements We thank Professor S. Papa, Dr. A.J. Meijer, Dr. R.J.A. Wanders, Dr. J.J. van den Bogert and Ms. A. Strijland for helpful discussions and advice.
References 1 Morgan-Hughes, J.A. (1986) TINS, 9, 15-19. 2 Lombes, A., Bonilla, E. and Di Mauro, S. (1989) Rev. Neurol. (Paris) 145, 671-689.
3 Zeviani, M., Servidei, S., Gellera, C., Bertini, E., DiMauro, S. and DiDonato, S. (1989) Nature 339, 309-311. 4 Whitfield, C.D. (1985) in Molecular Cell Genetics (Gottesman, M.M., ed.), pp. 545-588, J. Wiley & Sons, New York, pp. 545-588. 5 Malczewski, R.M. and Whitfield, C.D. (1982) J. Biol. Chem. 257, 8137-8142. 6 Malczewski, R.M. and Whitfield, C.D. (1984) J. Biol. Chem. 259, 11103-11113. 7 Johnson, L.V., Walsh, M.L., Beverly, J.B. and Chen, L.B. (1981) J. Cell Biol. 88,5 26-535. 8 0 s e r o f f , A.R., Ohuoha, D., Ara, G., McAuliffe, D., Foley, J. and Cincotta, L. (1986) Proc. Natl. Acad. Sci. USA 83, 9729-9733. 9 Bereiter-Hahn, J. (1976) Biocbim. Biophys. Acta 423, 1-14. 10 Gottesman, M.M. (1987) in Molecular Genetics of Mammalian Cells (Gottesman, M.M., ed.), 151, pp. 3-8, J. Wiley & Sons, New York. 11 Rickwood, D., Wilson, M.T. and Darley-Usmar, V.M. (1987) in Mitochondria. A practical approach (Darley-Usmar, V.M., Rickwood, D. and Wilson, M.T., eds.), p. 6, IRL Press, Oxford. 12 Scb~igger, H. and Von Jagow, G. (1987) Anal. Biochem. 166, 368-379. 13 Kyhse-Anderson, J. (1984) J. Biochem. Biophys. Methods 10, 203. 14 Sinjorgo, K.M.C., Hakvoort, B.M., Durak, I., Draijer, J.W., Post, J.K.P. and Muijsers, A.O. Biochim. Biophys. Acta 850, 144-150. 15 Gille, J.J.P., Van Berkel, C.G.M., Mullaart, E., Vijg, J. and Joenje, H. (1989) Murat. Res. 214, 89-96. 16 Schoonen, W.G.E.J., Handayani Wanamarta, A., Van Der KleiVan Moorsel, J.M., Jacobs, C., and Joenje, H. (1990) J. Biol. Chem. 265, 11118-11124. 17 Schoppink, P.J., Hemrika, W., Reynen, J.M., Grivell, L.A. and Berden, J.A. (1988) Eur. J. Biochem. 173, 115-122. 18 Fischer, J.C., Ruitenbeek, W., Berden, J.A., Trijbels, J.M., Veerkamp J.H. Stadhouders, A.M., Sengers, R.C.A. and Janssen, A.J.M. (1985) Clin. Chim. Acta 153, 23-36. 19 Srere, P.A. (1969) Methods Enzymol. 13, 3-11. 20 Bergmeyer, H.U. (1970) Methoden der Enzymatischen Analyse (Verlag Chemie GmbH, Weinheim/Bergstr.) I, 441-442. 21 Wanders, R.J.A, Kos, M., Roest, B., Meijer, A.H., Schrakamp, G., Tegelaers, W.H.M., Van Den Bosch, H., Schutgens, R.B.H. and Tager, J.M. (1984) Biochem. Biophys. Res. Commun. 123, 1054-1061. 22 Wijburg, F.A., Feller, N., Scholte, H.R., Przyrembel, H. and Wanders, R.J.A. (1989) Biochem. Internatl. 19, 563-570. 23 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 24 Bereiter-Hahn, J., Seipel, K.H., V6th, M. and PIoem, J.S. (1983) Cell Biochem. Funct. 1, 147-155. 25 Westerveld, A., Hoeijmakers, J.H.J., Van Duin, M., De Wit, J., Odiik, H., Pastiuk, A., Wood, R.D. and Bootsma, D. (1984) Nature 310, 425-428. 26 Griffith, O. and Meister A. (1979) J. Biol. Chem. 254, 7558-7560. 27 Meister A. (1983) Science 220, 472-476. 28 Di Mauro, S., Zeviani, M., Servidei, S., Bonilla, E., Miranda, A.F., Prelle, A. and Schon, E.A. (1986) Ann. N. Y. Acad. Sci. 488, 19-32. 29 Behbehani, A.W., Goebel, H., Osse, G., Gabriel, M., Langenbeck, U., Berden, J., Berger, R. and Schutgens, R.B.H. (1984) Eur. J. Pediatr. 143, 67-71. 30 Ferro-Luzzi Ames, G. (1986) Cell 47, 323-324. 31 Jones, P.A. and Taylor, S.M. (1980) Cell 20, 85-93. 32 Robinson, B.H., Ward, J., Goodyer, P. and Baudet, A. (1986) J. Clin. Invest. 77, 1422-1427. 33 Lomax, M.I., Coucouvanis, E., Schon, E.A. and Barald, K.F. (1990) Muscle and Nerve 13, 330-337.
275
BBADIS 61130
A procedure for selecting mammalian cells with an impairment in oxidative phosphorylation M.C.P. L o m b a r d o 1,2, J.W. van der Z w a a n 1, S. Brul 1 and J.M. Tager 1,2 l E.C. Slater Institute for Biochemical Research, University of Amsterdam, Academic Medical Centre, Amsterdam (The Netherlands) and : Institute of Medical Biochemistry and Chemistry, University of Bari, Bari (Italy) (Received 19 July 1991)
Key words: CHO mutant; Mitochondrial disease; Oxidative phosphorylation; Cytochrome c oxidase; Complex IV
The mitochondriai encephalomyopathies in man are characterized by heterogeneous defects leading to an impairment in the pathway of aerobic energy production. As a means of investigating the molecular and genetic mechanisms underlying these disorders we have developed a procedure for selecting m a m m a l i a n cell lines with features resembling the h u m a n pathological phenotypes. The principle of the selection is the use of a fluorescent amphiphilic dye, 2,4-(dimethylamino)-l-styrylmethylpyridiniumiodine, a cation showing two main features. Firstly, it is accumulated by mitochondria to an extent correlated with the magnitude of the electrochemical gradient of protons across the mitochondrial inner membrane. Secondly, upon irradiation with UV light, it gives rise to formation of free radicals, which inflict damage to the cell. Mutant cells with an impairment in oxidative phosphorylation will have more chance to survive than wild type cells. The selection procedure was applied to a stock of mutagenized Chinese h a m s t e r ovary cells. After subcioning of the cells which survived the selection procedure, twenty-six independent clones were isolated. Eighteen of the clones had a partial deficiency of cytochrome c oxidase ranging from 30 to 60% of the activity in control cells. The properties of two of the clones are described. One clone has been cultured under non-selective conditions for at least 12 months with retention of the partial deficiency of cytochrome c oxidase.
Introduction The mitochondrial encephalomyopathies comprise a heterogeneous group of diseases in man that are due to defects in the pathway of aerobic energy production in the mitochondrial inner m e m b r a n e [1,2]. These pathological conditions can be divided into two groups: those in which there is a deficiency of mitochondrially encoded subunits of an enzyme and those in which the defect is apparently localized in the nuclear genome, for instance in a gene coding for a protein required for the replication of mitochondrial D N A [3].
Abbreviations: CHO, Chinese hamster ovary; cyt c, cytochrome c; DASPMI, 2,4 (dimethylamino)-l-styrylmethylpiridinium iodine; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulphate; TMPD, tetramethyl-p-phenyldiamine. Correspondence: J.M. Tager, Institute of Medical Biochemistry and Chemistry, University of Bari, Piazza Giulio Cesare, 70124 Bari, Italy.
In order to understand the molecular and genetic mechanisms underlying these disorders, a possible approach is to select mutant cells with features resembling the human pathological phenotype [4]. Such mutants have been isolated from Chinese hamster lung cells [5,6] by a method involving replica plating. In this p a p e r we describe a procedure for the isolation of cells with an impairment in oxidative phosphorylation, i.e. with an impairment in the ability to establish a respiration-dependent electrochemical potential. This procedure makes use of a selection medium in which wildtype ceils do not survive. We have used this procedure to isolate Chinese hamster ovary ( C H O ) cell lines with a deficiency in cytochrome c oxidase. In principle the method could also be used to select mutants in other segments of the respiratory chain. Principle of the selection procedure The procedure used to select mutant cells with a perturbation in oxidative phosphorylation is based on the properties of some cationic amphiphilic dyes which
276 permeate freely through the plasma membrane and accumulate specifically in mitochondria, provided that an electrochemical gradient of protons across the mitochondrial inner m e m b r a n e is present [7]. These compounds have been used for monitoring of the relative mitochondrial m e m b r a n e potential in living cells [7]. It has been shown that the accumulation of such dyes by mitochondria is correlated with the magnitude of the mitochondrial electrochemical gradient and is particularly evident in proliferating cells [7]. Such dyes may act as photosensitizers. Upon irradiation following accumulation of the dyes free radicals, which are cytotoxic, are produced. The effectiveness of the cytotoxic effect of this treatment will depend on the amount of the dye accumulated, i.e. on the magnitude of the mitochondrial m e m b r a n e potential. Indeed, such dyes have been used for selective killing of carcinoma cells in vivo and in vitro [8]. Cultured cells defective in one or more components of the oxidative phosphorylation system are expected to have a decreased capacity to build up an electrochemical potential via the respiratory chain and to have a decreased capacity to accumulate these cations. Such cells can be expected to survive this treatment better than wild-type cells. We have used these principles to develop a selection medium for isolating cells with an impairment in mitochondrial oxidative phosphorylation. For this purpose we have employed the cationic dye 2,4-(dimethylamino)-l-styrylmethylpyridiniumiodine (DASPMI). As shown by Bereiter-Hahn [9], D A S P M I accumulates specifically in mitochondria in cultured cells in response to a m e m b r a n e potential across the mitochondrial inner membrane.
Materials and Methods
Cell culture. Chinese hamster ovary ( C H O ) cells, strain CHO-9, were grown at 37 ° in a 1:1 mixture of Ham-F10 medium (Flow) and D M E M (Dulbecco's modified Minimum Essential Medium; Flow), supplemented with 5% ( v / v ) foetal calf serum, streptomycin (100 / , g / m l ) and penicillin G (100 units/ml). The gas phase was 5% C O 2 - 9 5 % air. H u m a n fibroblasts from a control subject (cell line F5035) and from a patient with a deficiency of cytochrome c oxidase also expressed in fibroblasts (cell line 83RD88) were cultured as above except that the concentration of foetal calfserum was 10% (v/v). Cell line F5035 was kindly provided by Drs. Monique Mathieu and Irene Maire (Lyon) and cell line 83RD88 by Professor Jasper Scholte (Rotterdam). Mutagenization of CHO cells and selection of mutants. CHO-9 cells were mutagenized with ethylmethane-sulphonate exactly as described in Ref. 10.
The mutagcnized cells were platcd with Ill ml nledJunl in 9.4 cm Petri dishes at a density of 10" cells per p l a t e 12 h later 1 ~ M buthionine sulphoxinline was added to the culture medium. After another 12 h, 1() # M oligomycin was added and the cclls werc cultured in this medium for another 24 h. The medium was then removed and the ceils were washed thrcc times with glucose-free Hanks' medium. The washed cells were preincubated for 2 h with glucose-free Hanks' medium (Flow) at 37 °. The preincubation buffer was replaced by fresh Hanks' supplemented with 8 mM sodium suecinate (pH 7.4). After 30 min, DASPMI dissolved in 10 mM Mops (pH 7.4) was added at a final concentration of 30 # M . Then the ceils were irradiated with longwave UV light fl)r 10 min at 12 w a t t s / m -~. The cells that survived this treatment were isolated and subsequently subcloned. Preparation of mitochondria. Mitochondrial fractions were prepared from C H O cells as described by Rickwood et al. [11] for cultured fibroblasts.
Immunoprecipitation of cytochrome c ~)xidase /?om cellular extracts. About 107 cells were suspended in 1 ml of extraction buffer (50 mM Tris-sulphate (pH 7.4), 1 mM EDTA). Laurylmaltoside was added to a final concentration of 1% (mass/vol), together with the proteinase inhibitors phenylmethylsulphonyl fluoride (25 mM), antipain (21 /,M), pepstatin (20 /,M) and leupeptin (38 /,~M). Alter sonification (3 times for 15 s at 7/.Lm amplitude, with intervals of 45 s), the samples were incubated for 31) rain on ice and centrifuged for 5 min in an Eppendorf centrifuge at 4°C. The supernatant was incubated overnight with anti-(eytochrome c oxidase) coupled to protein A-Sepharose 4B beads under rotation at 4°C. The beads were washed seven times with 50 mM Tris-sulphate (pH 7.4), 1 mM EDTA, 0.05% (mass/vol) laurylmaltoside. Cytochrome c oxidase was solubilized by incubation for 2 h at room temperature in sample buffer (4% (mass/vol) sodium dodecyl sulphate and 12% (mass/vol) glycerol in 50 mM Tris-HCl (pH 6.8)). After centrifugation, the supernatant was subjected to sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis (PAGE) and immunoblotting. Electrophoresis and immunoblotting. Electrophoresis was performed as described by Sch~igger and yon Jagow [12]. Western blotting was performed using a semi-dry system (W.E.P. C o m p a n y , Seattle, Washington; Electr-Trans blotter, model 198155) in the presence of transfer buffers according to Kyhse-Anderson [13] but omitting 6-amino-n-hexanoic acid. The nitrocellulose filters were treated as described by Sinjorgo et al. [14]. Immunoreactive polypeptides were visualized using the immunoperoxidase method with 4-chloro-l-naphthol as substrate. Preparation of rabbit antisera against ~ytochrorne c oxidase. The antisera used in this study had been raised
277 in rabbits against cytochrome c oxidase purified from human heart muscle [14]. A TP content of cells. A T P was determined fluorimetrically in an aqueous extract of cells as described in [151. Assay of activities. In order to measure cytochrome c oxidase activity, cells were incubated for 15 rain at 0 ° in a solution containing 15 mM s o d i u m / p o t a s s i u m phosphate (pH 7.4) and 1% ( m a s s / v o l ) laurylmaltoside. After centrifugation for 5 min at 10000 r e v s / m i n at 4°C, enzymic measurements were performed at room temperature in fractions of the supernatant by pulsing reduced horse cytochrome c (15 /xM final concentration) as substrate and following the decrease in absorption of the sample at 550 nm. Cytochrome c oxidase was also assayed polarographically in digitonin-permeabilized cells with tetramethyl-p-phenylenediamine (TMPD)-ascorbate as substrate exactly as described by Schoonen et al. [16]. The K m of cytochrome c oxidase for cytochrome c was measured at high ionic strength with horse cytochrome c as substrate. Succinate dehydrogenase and succinate : cytochrome c oxidoreductase were assayed in mitochondrial preparations obtained from C H O cells as described in Refs. 17 and 18, respectively. Citrate synthase was assayed in cellular extracts by measuring the appearance of free SH groups, as monitored by reduction of 5,5'-dithiobis-(2nitrobenzoate), after addition of acetyl-CoA and oxaloacetate [19]. Lactate dehydrogenase was assayed in cell extracts as described [20]. Catalase activity was measured polarographically according to Ref. 21.
Lactate and pyruvate formation from glucose (see Ref 22). Wild-type cells and mutant cell lines were
+
Buthionine
sulphoximine
90 80 .-&
70
t~
60
t-
50 40 30
+ UV
20 10 0
0
I
!
!
I
20
40
60
80
DASPMI
( ~ M)
Buthionine
A
100
sulphoximine
90
80 70 rJ 113 ¢-
> -i ¢/)
60 50 40
+ UV
30 ~=
.
20 10 0 0
I
I
I
I
20
40
60
80
DASPMI
lO0
( ~. M)
assayed for their ability to form lactate and pyruvate from glucose 3 days after reaching confluency. Production of lactate and pyruvate from glucose (5 mM) was followed at intervals of 1 h in 270 /zl aliquots of the incubation medium neutralized to p H 6 - 7 after removal of proteins by perchloric acid precipitation. Cells were harvested with 0.1 M N a O H to assay the protein content according to Lowry et al. [23].
Fig. 1. Effect of treatment with DASPMI, UV light and buthionine sulphoximine on viability of wild-type Chinese hamster ovary cells as assessed by colony formation. For conditions see Materials and Methods. The number of surviving cells (colonies) is expressed as a percentage of the number surviving in control dishes in which DASPMI and UV treatment were omitted. In the absence of DASPMI, treatment with UV led to a decrease in survival of about 50%.
Results
respiratory chain might not exhibit a decreased electrochemical potential because of a compensating effect of A T P generated during glycolysis, which can energize mitochondria by reacting with A T P synthase [24]. We therefore included oligomycin, which inhibits A T P synthase, in our selection medium. Fig. 1 shows the results of an experiment with Chinese hamster ovary ( C H O ) cells. In the absence of UV treatment, D A S P M I had only a small effect on the survival of the cells; at concentrations of D A S P M I from 5 /zM up to 30 /xM, survival decreased to a minimum value of 70% and then remained constant as the concentration of D A S P M I was increased further. T r e a t m e n t with U V light led to a 50% decrease in
Setting up the selection procedure We incubated C H O cells in culture with D A S P M I in the presence and absence of the respiratory chain inhibitors antimycin and cyanide. No evident difference in fluorescence of the mitochondria could be seen between the two conditions. This is on the one hand in contrast with studies previously performed using mitochondrial suspensions, where respiratory inhibitors clearly decrease the fluorescence of the mitochondria [7], but on the other hand it is in accordance with data obtained in living cells [24]. Starting from this observation, we expected that intact cells with a defective
278 TABLE I t".J~bct ~Tf'treatment with D A S P M I and ultrat:iolet light on the .w~rt il al o['fibn#~lasts from a control suIzk'et and a patient with a d~[Teienev o[ O, tochrome c oxidase (_'ell line F5035 from a control subject was kindly supplied by Drs. Monique Mathieu and Irene Maire (Lyon) and cell line 83KD88 from a patient with a generalized deficiency of cytochrome e oxidase was kindly supplied by Professor H.R. Scholte (Rotterdam). The activity of cytochrome c oxidase was 28(1 nmol cytochrome e cons u m e d / m i n per mg protein in cell line F5035 and 56 nmol cytochrome c c o n s u m e d / m i n per mg protein in cell line 83RD88 (J.W. van der Zwaan and H.R. Scholte. unpublished observations). Values in parentheses are percentages of the number of cells which survived in the absence of DASPMI and UV irradiation. The cells were cultured in the presence of 1 p.M buthionine sulphoximine, in the absence or presence of 10 s M oligomycin and treated further as indicated in the table. Conditions of culture
In the absence of oligomycin no treatment DASPMI 30/J.M UV irradiation DASPMI 30/~M + UV irradiation In the presence of oligomycin no treatment DASPMI30p.M UVirradiation DASPMI 3 0 / ~ M + U V irradiation
Number of surviving cells control (cell line F5035)
patient (cell line 83RD88)
1.6-10 ~' (100) 1.25.106 (78) 0.80.106 (50)
0.45.10 (' (100) 0.49.106 (108) 0.28-106 (62)
0.29.10 ~ (18)
0.20- 10 ¢' (44)
0.80.10 ~' (50) 0.65.10 ~ (41) 0.31.106 (19)
0.21 • 106 (47) 0.17.10 ~' (38) 0.14.106 (31)
I).05.10 ~
0.18.1() c' (40)
(3)
survival. When the cells were treated with D A S P M I prior to exposure to U V light, the survival decreased further, to a minimal value of 20%. Survival was even lower (maximally 10%) when cells were depleted of glutathione by culturing them with buthionine sulphoximine (see below).
patient was about the same as in Ihe absence ¢~1 oligomycin, but only about 5% of the control cells survived the treatment in the presence of oligomycin.
Isolation of CHO mutants C H O cells (strain CHO-9; see Ref. 25) mutagenized with ethylmethane-sulphonate were plated out in Petri dishes at a density of 105 cells per plate. Buthionine sulphoximine was subsequently added, in order to inhibit 7-glutamylcysteine-synthetase and thereby to decrease the glutathione content of the cells [26]. Glutathione normally acts as scavenger ti~r free radicals {27]. In this way the cells became more sensitive to the presence of free radicals. After treatment with DASPMI and irradiation one or two cells per Petri dish survived and were able to form colonies. The colonies were picked up and cultured separately. Colonies were subcloned by limiting dilution and 26 clones were obtained, each of which was derived from a separate Petri dish. The activity of cytochrome c oxidase in the mutants varied from 28-101% of that in the wild-type cells. In 18 of the clones, the activity of cytochrome c oxidase was 60% or less than that of the wild-type cells. The activity of citrate synthase and catalase did not differ significantly from that in the control cells. Some of the clones were analysed by means of immunoblotting experiments with an anti-(human cytochrome c oxidase) antiserum. Fig. 2 shows that the amount of crossreactive material corresponding to subunits II + 11I was lower in the mutant cells tested, particularly in cell line 13 and cell line 8 (see Fig. 2, lane c and e). In addition, there were marked differences in the electrophoretic mobility and the amount of crossreactive material corresponding to the nuclearly encoded subunits V, VI, VII and Vlll. Such differences were much less marked in mutant cell lines 1 and 2 (Fig. 3, lanes d, e, f and g, respectively). Further investigations were performed on the properties of cell lines I and 2.
Control experiment with human fibroblasts In order to test whether our procedure was selective for cells defective in the aerobic energy-producing pathway, an experiment was performed with fibroblasts derived from a control subject and from a patient with a deficiency in the cytochrome c oxidase segment of the respiratory chain. Cells were plated out and subjected to the treatment used for the selection of C H O cells (see below). Table I shows that, when control cells were cultured in the absence of oligomycin, treatment with D A S P M I followed by U V light decreased the survival of the cells to about 20%, whereas about 45% of the cells from the patient survived. The difference was even greater when oligomycin was present in the culture medium. The survival of the cells from the
Characterization of the phenotype of cell lines 1 and 2 The mutant phenotype consists in both cases of a decreased activity of cytochrome c oxidase, whereas the K m for cytochrome c is equal to that in thc control. The residual activity in clone 1 and clone 2 was about 40 and 30%, respectively, in comparison with the control cell line CHO-9 (Table II). These spectrophotometric assays were corroborated by measurements of oxygen uptake in digitonin-permeabilized cells with TMPD-ascorbate as substrate. Further analysis of the respiratory chain was carried out by isolating mitochondria from the cells and using them for spectrophotometric tests in the presence of substrates acting at specific levels of the respiratory chain. Succinate dehy-
279 a
b
c
d
e
f
g
h
T A B L E lI
Activity of cytochrome c oxidase, citrate synthase and catalase in wild-type and mutant CHO cell lines
-
at
dln~
I
T h e mutant clones were isolated from a stock of mutagenized C H O cells using the procedure described in the text for selection of mutants with an impairment in mitochondrial oxidative phosphorylation. For experimental details see Materials and Methods.
II, III
Cell lines
C
C
Fig. 2. Electrophoretic pattern of subunits of cytochrome c oxidase in extracts of Chinese hamster ovary cells. Extracts of cells were subjected to S D S - P A G E and immunoblotted with anti-(human cytochrome c oxidase) as described in Materials and Methods. Lane a, molecular weight markers; lane b, wild-type (107 cells); lane c, clone 13 (14- 106 cells); lane d, clone 10 ( 1 3 106 cells); lane e, clone 8 (10' cells); lane f: clone 22 (13.5.106 cells); lane g, clone 4 (12.106 cells); lane h, cytochrome c oxidase isolated from beef heart.
Activity of
Wild-type Clone 1 Clone 2 Clones 3 - 5 Clones 6 - 8 Clones 9-12 Clones 13-18 Clones 19-22 Clones 23-25 Clone 26
cytochrome c oxidase a
citrate synthase ~'
catalase h
495 _+35 c (50) 230_+ 12 c (32) 175-+ 15 c (45) 140-181 200-209 231-268 278-298 304-361 402-475 500
45 _+4 c (2) 45_+6 c (2) 41-+2 ~ (2) 43-44 41-43 45-46 40-44 42-46 42-46 45
35 _+3 c (2) 3 5 + 3 c (2) 3 2 + 2 c (2) 30-36 31-36 30-35 30-40 33-37 35-39 33
Specific activity in nmol of substrate consumed per min ~ per mg protein ~. b Specific activity in nmol of oxygen produced per rain ~ per mg protein ~. c Means_+ S.D. The n u m b e r of m e a s u r e m e n t s performed using separate cell cultures is shown in parentheses.
d r o g e n a s e and succinate : c y t o c h r o m e c reductase activities were found to be normal in both the mutant clones.
T A B L E IIl
Biochemical characteristics of wild-type CHO cells and clones 1 and 2 Clones 1 and 2 were isolated from a stock of mutagenized C H O cells using the procedure described in the text for selection of mutants with an impairment in mitochondrial oxidative phosphorylation. For experimental details see Materials and Methods. Values are means_+ S.D. with the n u m b e r of m e a s u r e m e n t s performed using separate cell cultures shown in parentheses. Parameter Cytochrome c oxidase -activity a -K m ( ~ M cyt.c) Oxygen consumption b (ascorbate/YMPD) Succinate dehydrogenase ~' Succinate : cytochrome c reductase c Lactate dehydrogenase a fl-hexosaminidase a Lactate in medium (raM) d Pyruvate in m e d i u m (raM) J L a c t a t e / p y r u v a t e ratio J A T P content ( n m o l / m g protein) " b c a
Wild-type
Clone 1
Clone 2
495 40 2.2
230 __+12 (32) 42 -+ 5 (4) 0.95 -+ 0.3 (2)
175 38 0.6
47.6 84.5
__+35 (50) -+ 5 (5) + 0.2 (2) _+ 1.5 __+ 8
1.97 _+ 23 i 2.48 _+ 0.133_+ 19 _+ 0.24 ___
0.16 3 0.4 0.02 1 0.04
(3) (2)
49.1 82.5
(3) (2) (2) (2) (2) (2)
2.01 _+ 22 + 3.71 _+ 0.098_+ 38 _+ 0.25 ±
Specific activity in nmol of substrate consumed m i n - 1 mg protein - 1 Specific activity in /.Lmol of oxygen consumed h - 1 mg protein i. Specific activity in nmol of cytochrome c reduced m i n - 1 mg protein 1. Values after 4 h of incubation with glucose.
_+ 1 _+ 12 0.15 4 0.5 0.03 1 0.08
-+ 15 (45) + 6 (3) + 0.1 (2)
(3) (3)
50.3 85.2
_+ 1.6 __+ 9
(3) (2) (2) (2) (2) (2)
2.21 _+ 25 _+ 2.82 _+ 0.121_+ 23 _+ 0.29 _+
0.2 3 0.1 0.02 2 0.04
(3) (3) (3) (2) (2) (2) (2) (2)
280 a
b
c
d
e
f
g
h
b ....
.~..,-..
~---:: _ _ -
-7:
--
vla,b,c
--
Vlla,b,c
Fig. 3. Electrophoretic pattern of subunits of cytochrome c oxidase in wild-type Chinese hamster ovary cells and in clones 1 and 2. Extracts of cells were subjected to SDS-PAGE and immuno- blotted with anti-(human cytochrome c oxidase) as described in Materials and Methods. Lane a, molecular weight markers; lanes b and c, wild-type (10 7 cells); lanes d and e, clone 1 (11.106 cells); lanes f and g, clone 2 (107 cells); lane h, cytochrome c oxidase isolated from human heart.
Formation of lactate and pyruvate from glucose by the cells was also examined (Table III). In clone 1 the lactate/pyruvate ratio was twice as high as in control cells. In this respect, clone 1 resembles some human myopathic conditions in which an elevated l a c t a t e / pyruvate ratio is found [28,29]. This finding is a consequence of the increase in both the mitochondrial and the cytosolic N A D H / N A D + ratio due to a block in mitochondrial N A D H oxidation, which leads to an increased formation of lactate from pyruvate via lactate dehydrogenase [22]. In the case of clone 2 there was a slight increase in the lactate production and a slight decrease in the concentration of pyruvate in comparison with the control, but the ratio lactate/pyruvate was not significatively increased. The intracellular level of ATP in clones 1 and 2 was found to be comparable with the control level (Table III). There was no change in the activity of lactate dehydrogenase, catalase and /3-hexosaminidase, marker enzymes for the cytosol, the peroxisomes and the lysosomes, respectively, in clone 1 and 2. In addition, the activity of citrate synthase, a mitochondrial matrix enzyme, was normal in the two mutants, indicating that there is no generalized impairment of mitochondrial functions in the mutants. Discussion
The procedure described in this paper provides a means of selecting cells with a deficiency in the ability
to maintain a respiration-dependent electrochemical gradient of protons. The selection medium contains succinate as a respiratory substrate, since DASPMI inhibits N A D H dehydrogenase [9], and oligomycin, lo prevent energization of mitochondria by ATP generated during glycolysis. Thus the electrochemical gradient of protons built up across the mitochondrial men> brane under these conditions is dependent on electron flow from succinate to O~ and includes the cytochromc c oxidase step. During subsequent culture of the cells in the absence of selection medium N A D H dehydrogenase will, of course, not be inhibited. In 18 of the 26 clones isolated from a stock of mutagenized C H O cells there was a decrease in the activity of the enzyme to 60% of the control value or less. In one of the primary cultures derived from a colony of cells which survived the selection medium cytochrome c oxidase was almost completely deficient. The cells are slow-growing, however, and subcloning is still in progress. The stability of the mutations in two of the clones ( 1 and 2) was followed in time by keeping the cells in culture in the absence of DASPMI, UV light, buthionine sulphoximine or other stress conditions. In the case of clone 1 the phenotype was maintained for 8 months (corresponding to about 480 cell divisions), after which reversion to the wild-type phenotype occurred. Clone 2 has now been kept in culture for 12 months without reversion to the wild-type phenotype. Clearly, the phenotype of clones 1 and 2 is not due to a direct effect of D A S P M I on cytochrome c oxidase activity or on the assembly of the enzyme. The cellular ATP content of clones 1 and 2 was the same as that of control cells. In clone I the lactate/ pyruvate ratio in the medium was increased in comparison with the control cells. Thus in this clone glycolysis may play an important role in maintaining the ATP level in the cells. In clone 2, on the other hand, the lactate/pyruvate ratio was about the same as in the controls. Schoonen et al. [16] have recently shown that exposure of C H O cells to normobaric hyperoxia, which can be expected to increase the concentration of free radicals, leads to respiratory failure due to inactivation of N A D H dehydrogenase, succinate dehydrogenase and 2-oxoglutarate dehydrogenase. Interestingly, hyperoxia had no effect on ubiquinone:cytochrome c reductase (complex IlI), cytochrome c oxidase (complex IV), and ATP synthase (complex V). Moreover, in a control experiment performed by us with human fibroblasts (Table I), the residual activity of cytochrome c oxidase in the cell line derived from a patient affected by mitochondrial myopathy was the same before and alter submission to our selection procedure. We can therefore exclude the possibility that the deficiency arose from a damaging action of free radicals on the mere-
281 brane structure which could disturb the assembly of the enzyme. In clones 23-26 the cytochrome c oxidase activity was 80-100% of that found in wild-type cells. Thus some other characteristic of these cells must have enabled them to survive the selection procedure, for instance an impairment in the proton-pumping activity of cytochrome c oxidase or a defect in some other segment of the respiratory chain. A third possibility is that the cells are resistant tot buthionine sulphoxamine or oligomycin due for instance to increased expression of P glycoprotein which is responsible for multidrug resistance in mammalian cells [30]. In order to test whether the phenotype of clones 1 and 2 was due to hypermethylation of DNA rather than to mutations in specific genes, the cells were subjected to treatment with 5-azacytidine, a hypomethylating agent [31]. Surprisingly, treatment with 5azacytidine led to a 2-3-fold increase in cytochrome c oxidase activity in clones 1 and 2 (M.C.P. Lombardo, unpublished observations). Moreover, a similar increase in activity of the enzyme was observed in wildtype cells. This phenomenon may be due to activation by demethylation of nuclear genes for isoforms of cytochrome c oxidase not normally expressed in CHO cells or of nuclear genes affecting the expression of mitochondrial genes for cytochrome c oxidase. The general features of the organization of the nuclear and mitochondrial genes required for the biogenesis of mitochondria and the very high number of copies per cell of the mammalian mitochondrial genome [32] require concomitant regulation of expression of the genes and very tight nuclear-mitochondrial interactions. The comprehension of these interactions, which are very important in the biogenesis of mitochondria, would be facilitated by analysis of cells having a perturbation in oxidative phosphorylation. Moreover, the characterization of these mutant strains can contribute to an understanding of the functional role played by the supernumerary subunits present in segments of the mitochondrial respiratory chain as well as of the meaning of the differential expression of nuclear genes during differentiation [33].
Acknowledgements We thank Professor S. Papa, Dr. A.J. Meijer, Dr. R.J.A. Wanders, Dr. J.J. van den Bogert and Ms. A. Strijland for helpful discussions and advice.
References 1 Morgan-Hughes, J.A. (1986) TINS, 9, 15-19. 2 Lombes, A., Bonilla, E. and Di Mauro, S. (1989) Rev. Neurol. (Paris) 145, 671-689.
3 Zeviani, M., Servidei, S., Gellera, C., Bertini, E., DiMauro, S. and DiDonato, S. (1989) Nature 339, 309-311. 4 Whitfield, C.D. (1985) in Molecular Cell Genetics (Gottesman, M.M., ed.), pp. 545-588, J. Wiley & Sons, New York, pp. 545-588. 5 Malczewski, R.M. and Whitfield, C.D. (1982) J. Biol. Chem. 257, 8137-8142. 6 Malczewski, R.M. and Whitfield, C.D. (1984) J. Biol. Chem. 259, 11103-11113. 7 Johnson, L.V., Walsh, M.L., Beverly, J.B. and Chen, L.B. (1981) J. Cell Biol. 88,5 26-535. 8 0 s e r o f f , A.R., Ohuoha, D., Ara, G., McAuliffe, D., Foley, J. and Cincotta, L. (1986) Proc. Natl. Acad. Sci. USA 83, 9729-9733. 9 Bereiter-Hahn, J. (1976) Biocbim. Biophys. Acta 423, 1-14. 10 Gottesman, M.M. (1987) in Molecular Genetics of Mammalian Cells (Gottesman, M.M., ed.), 151, pp. 3-8, J. Wiley & Sons, New York. 11 Rickwood, D., Wilson, M.T. and Darley-Usmar, V.M. (1987) in Mitochondria. A practical approach (Darley-Usmar, V.M., Rickwood, D. and Wilson, M.T., eds.), p. 6, IRL Press, Oxford. 12 Scb~igger, H. and Von Jagow, G. (1987) Anal. Biochem. 166, 368-379. 13 Kyhse-Anderson, J. (1984) J. Biochem. Biophys. Methods 10, 203. 14 Sinjorgo, K.M.C., Hakvoort, B.M., Durak, I., Draijer, J.W., Post, J.K.P. and Muijsers, A.O. Biochim. Biophys. Acta 850, 144-150. 15 Gille, J.J.P., Van Berkel, C.G.M., Mullaart, E., Vijg, J. and Joenje, H. (1989) Murat. Res. 214, 89-96. 16 Schoonen, W.G.E.J., Handayani Wanamarta, A., Van Der KleiVan Moorsel, J.M., Jacobs, C., and Joenje, H. (1990) J. Biol. Chem. 265, 11118-11124. 17 Schoppink, P.J., Hemrika, W., Reynen, J.M., Grivell, L.A. and Berden, J.A. (1988) Eur. J. Biochem. 173, 115-122. 18 Fischer, J.C., Ruitenbeek, W., Berden, J.A., Trijbels, J.M., Veerkamp J.H. Stadhouders, A.M., Sengers, R.C.A. and Janssen, A.J.M. (1985) Clin. Chim. Acta 153, 23-36. 19 Srere, P.A. (1969) Methods Enzymol. 13, 3-11. 20 Bergmeyer, H.U. (1970) Methoden der Enzymatischen Analyse (Verlag Chemie GmbH, Weinheim/Bergstr.) I, 441-442. 21 Wanders, R.J.A, Kos, M., Roest, B., Meijer, A.H., Schrakamp, G., Tegelaers, W.H.M., Van Den Bosch, H., Schutgens, R.B.H. and Tager, J.M. (1984) Biochem. Biophys. Res. Commun. 123, 1054-1061. 22 Wijburg, F.A., Feller, N., Scholte, H.R., Przyrembel, H. and Wanders, R.J.A. (1989) Biochem. Internatl. 19, 563-570. 23 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 24 Bereiter-Hahn, J., Seipel, K.H., V6th, M. and PIoem, J.S. (1983) Cell Biochem. Funct. 1, 147-155. 25 Westerveld, A., Hoeijmakers, J.H.J., Van Duin, M., De Wit, J., Odiik, H., Pastiuk, A., Wood, R.D. and Bootsma, D. (1984) Nature 310, 425-428. 26 Griffith, O. and Meister A. (1979) J. Biol. Chem. 254, 7558-7560. 27 Meister A. (1983) Science 220, 472-476. 28 Di Mauro, S., Zeviani, M., Servidei, S., Bonilla, E., Miranda, A.F., Prelle, A. and Schon, E.A. (1986) Ann. N. Y. Acad. Sci. 488, 19-32. 29 Behbehani, A.W., Goebel, H., Osse, G., Gabriel, M., Langenbeck, U., Berden, J., Berger, R. and Schutgens, R.B.H. (1984) Eur. J. Pediatr. 143, 67-71. 30 Ferro-Luzzi Ames, G. (1986) Cell 47, 323-324. 31 Jones, P.A. and Taylor, S.M. (1980) Cell 20, 85-93. 32 Robinson, B.H., Ward, J., Goodyer, P. and Baudet, A. (1986) J. Clin. Invest. 77, 1422-1427. 33 Lomax, M.I., Coucouvanis, E., Schon, E.A. and Barald, K.F. (1990) Muscle and Nerve 13, 330-337.
