Primed tn Sweden Copyrighr @ 1976 by Academic Press. Inc. in any,formrrserwd All rights of reproducrion
Experimental
ISOLATION CELL
Cell Research 102 (1976) 29&310
AND DETAILED LINES
RESISTANT
CHARACTERIZATION
OF HUMAN
TO D-THREO-CHLORAMPHENICOL
R. L. SIEGEL,’ A. J. JEFFREYS,’ W. SLY* and 1. W. CRAIG’ ‘Genetics Laboratory, Department of Biochemistry, University of Oxford. Oxford OX1 3QV, UK, and %chool of Medicine, Washington tJniversity, St Louis, MO 63110, USA
SUMMARY The isolation and characterization of chloramphenicol resistant derivatives of the human cell line HeLa B is described. Growth of resistant lines was unaffected in the presence of 100 pg/ml Dthreo-chloramphenicol, whereas growth of the parental cells was inhibited at 12.5 pg/ml. The incorooration of P?Slmethionine into mitochondriai orotein of intact resistant cells continued normaliy in the presence of 100 pg/ml chloramphenicbl (cytoplasmic protein synthesis was blocked by addition of 50 &ml emetine). Under these conditions the electroohoretic motile of labelled. presumptive mitochondrially-made proteins was similar to that of the parental cell line labelled in the absence of chloramphenicol. The cell lines selected in the presence of chloramphenicol also showed increased resistance to some other inhibitors of mitochondrial protein synthesis, e.g. carbomycin and mikamvcin. f’~lChloramphenicol was found to have normal access to the interior of resistant cells and it is therefore unlikely that resistance results from altered cell permeability. No modification of the drug by acetylation or glucuronide conjugation mechanisms was observed. The possibilities remain that resistance is mediated by altered permeability of the mitochondrial membrane, or from modification to a component of the mitochondrial protein synthetic system.
Chloramphenicol is one of a group of antibiotics which specifically inhibit mitochondrial and bacterial protein synthesis, but do not inhibit (in the same dose range) protein synthesis on cytoplasmic ribosomes of eukaryotes. Mutants resistant to chloramphenicol (CAP) and/or other inhibitors acting at the level of mitochondria have been isolated in yeast [ 1, 21, Aspergillus [3], Paramecium [4] and Tetrahymena [5]. Some of these have been shown to be cytoplasmically inherited and could therefore result from mutations at the level of mitochondrial DNA (mitDNA). Recent reports [6, 71 have shown that mutants of cultured mammalian cells can be Exp
Cd
Res
102 (1976)
obtained which are resistant to CAP. Human and mouse cells have been isolated which were capable of growth in 25 and 50 pg/ml of the drug. Protein synthesis in isolated mitochondria was resistant, at least in part, to CAP (see Discussion). The resistance in mouse cells could be transferred by enucleated fragments to sensitive cells of the same species. Since these observations, Wallace & Freeman [8] have isolated mouse cell lines which are resistant to the chloramphenicol analogue, Tevenel. and Mitchell et al., have reported the selection of a human cell line capable of growth in 100 pg/ml CAP [9]. Resistance to other drugs known to in-
Human cell lines resistant to D-threo-chloramphenicol hibit mitochondrial function has also been observed in cultured cells. Klietmann et al. [lo] reported the isolation of SV40 transformed hamster cells capable of growth in the presence of normally toxic concentrations of ethidium bromide (EB). However, the mechanisms conferring resistance to EB, or to CAP in cultured cells are poorly understood. We now report in detail on the properties of a human cell line which is resistant to 100 pg/ml D-threo CAP. The cell line was isolated with the intention of subsequently employing it in studies on mitochondrial biogenesis. Several genes for human mitochondrial functions have been assigned to human chromosomes (see [ 11, 121).However, human mitDNA is apparently lost in fusions between human cells and established mouse cell lines [13-141. In contrast, fusions between human cell lines and embryonic mouse cells yielded hybrids capable of propagating mitochondrial genomes of both species. In some cases, the human and rodent sequences were apparently linked, suggesting interspecific recombination of mitDNA [ 151. It was hoped that CAP resistance could be employed as a marker to select for the presence of human mitDNA sequences in hybrid cells. This may provide an opportunity to identify human chromosome(s) possibly necessary for propagation of human mitDNA, or to characterise recombination events between mammalian mitochondrial genomes. A preliminary report of some aspects of this study has appeared previously [ 161.
METHODS Cells, media and growth conditions CAP resistant cell lines were isolated from HeLa B, a subclone derived from the S strain of HeLa (the kind gift of Dr L. Tolmach). Cells were grown initially in
299
Eagles minimum essential medium (GIBCO, Grand Island, N.Y.) supplemented with fetal calf serum (10% v/v), penicillin (100 W/ml) and streptomycin (100 pg/ml). Following early growth in 75 cm* tissue culture flasks (Falcon Plastics, Oxnard, Calif.) cells were grown in glass bottles, or in spinner culture, employing RPM1 1640 medium (Flow Laboratories Limited, Irvine, Scotland) supplemented with serum and antibiotics as above. Pure o-threo-chloramphenicol was kindly supplied by Dr H. E. Machamer (Parke, Davis and Co., Detroit, Mich.). It is important to note that different commercial sources of CAP vary widely in their ability to inhibit mitochondrial protein synthesis. This presumably results from contamination with inactive Lthreo-isomer. The pure o-threo-CAP was about 2.5x as effective in studies with intact cells as that available commerciallv (Siema Co.. London) and L-threoCAP had little detectible inhibitory action up to 200 &ml 1171.Emetine hvdrochloride was obtained from : Stgma ilondon), mikamycin and EB from Calbiochem (Los Angeles, Calif.) and carbomycin from Pfizer Inc. (Groton, Conn.) MEM lacking methionine was obtained from Biocult Laboratories Ltd. (Paisley, Scotland). Phosphate-buffered saline (PBS) was 0.85% (w/v) NaCl, KH,PO, (2.5 mM) and Na2HP0, (10 mM).
Mutagenesis and selection Mutagenesis was performed employing a modification of the procedure described by Spolsky & Eisenstadt [6]. The cells (HeLa B) were plated at 106-2~ 1Ogper 75 cm2 tissue culture flask and treated with EB (0.05 PM) for 24 h. The medium was removed and replaced with one containing N-methyl-N-nitro-N nitrosoguanidine (0.05 pg/ml) for 3 h. After washing, the cells were subsequently selected in medium containing sthreo-chloramphenicol (50 &g/ml). After two weeks the majority of cells began to die. Selection was continued for 6 months during which time some of the remaining cells apparently began to increase in size and finally cell division was observed. In an attempt to facilitate outgrowth of the putative resistant cells, the CAP concentration was reduced to 12.5 p.gl ml. Rapid cell proliferation followed and several independent uncloned populations were isolated. These were tested for the resistance of mitochondrial protein synthesis to chloramphenicol. Those which showed significant resistance were subject to cloning in 100 pg/ml CAP. Two of the cloned isolates, MC63 and MC75, were chosen for detailed investigation. They were routinely maintained in medium containing 50 pglml CAP.
Growth curves Cells were plated in plastic dishes (5.0 cm diam) in normal medium. This was changed 24 h later, either to unmodified medium or to one containing o-threoCAP (50 &ml). At zero time and at subsequent regular intervals cells on duplicate dishes were harvested and counted. Exp Cell Res 102 (1976)
300
Siegel et al.
Assay for mitochondrial protein synthesis in intact cells The assay was based on conditions established oreviously [is, 191.Cells (0.5x106-2.0~ 1oB)were plated in plastic dishes (5 cm diam.) in RPM1 medium 1640 12-17 h before use. The plates were washed twice with PBS and 1 ml of MEM-methionine, supplemented as described previously [18] and containing emetine (SO pg/ml), was added. One hundred ~1 of various antibiotic solutions, or PBS, were added to produce the desired final concentrations (medium without antibiotics served as a control). Plates were preincubated with antibiotics for 30 min at 37°C prior to addition of label. The subsequent steps were similar to those described previously except that the label was 1 &i [YS]-methionine (170-240 Cilmmole) and incornoration of label continued for 2 h. Subseauently, the cells were harvested and suspended in n&ma1 media and centrifuged (500 g 5 min). The wet cell pellet was resuspended in 1 ml NaOH (1 N) and incubated at 37°C for 10 min. One ml HCl (1 N) was added followed by 8 ml trichloracetic acid (TCA 5% w/v) containing L-methionine (10 mM). The suspension was chilled on ice and filtered through GF/Cglass fibre filters (Millipore Ltd, London). Filters were washed with 3~ 10 ml TCA+methionine, airdried and placed in 10 ml scintillation fluid before counting.
Electrophorectic mitochondrially
Mutant
isolation
Mutagenesis and selection of cells resistant to chloramphenicol were carried out as described in Materials and Methods. Attention should be drawn to the length of time required for cell growth (6-9 months). This is consistent with the observations of
lb t
I
II
I
‘.
105
t
Ii
t I
\
of chloramphenicol
[“CID-threo-chloramphenicol [W]CAP was purchased from the Radiochemical Centre, Amersham, with money made available through the good offices of Dr J. A. -L. Gorringe, Director of Cli&al Investigation, Parke Davis & Co., Pontypool. The specific activity of the drug was 27.3 @i/mg and was labelled in the 1-C position of the dichloroacetamido group. Its radiochemical purity was 99%, but no details were available of the contamination with the L-threo isomer. Falcon flasks (25 cme) containing about 10 cells in 3 ml medium containing D-threo*AP (12.5 &ml) plus r4C-labelled CAP (1 pg, 6~ lW dpm/ml) were inExp Cell Res 102 (1976)
RESULTS
separation of synthesized proteins
Cultured cells were washed in PBS, resuspended in incorporation medium to 106-3 x 106cells/ml and preincubated in the presence of emetine (100 pg/ml) and amounts of CAP as described previously [ 181. Thirty $i [YS]methionine/ml were added and incubation continued for 3 h at 37°C. The incorporation was followed by a ‘cold chase’ with 1 mM L-methionine for 20 min and the cells subsequently pelleted and washed with PBS. Cell pellets were stored in liquid nitrogen. Electrophoresis of whole cell extracts, made by solubilizing cells at 100°Cin 2 % (w/v) sodium dodecyl sulphate (SDS) containing 2-mercaptoethanol (5 % w/v), was effected in slab polyacrylamide gels (12%) employing the buffer system of Laemmli [20]. Dried gels were subsequently autoradiographed on Kodak Blue Brand X-ray film.
Chromatography and derivatives
cubated for 1, 2, or 3 days. After incorporation, cells were detached by trypsinization. Cells, trypsin and medium were pooled and extracted with an equal volume of ethyl acetate at 0°C with agitation. The organic phase was separated by centrifugation at 500 g for 5 min. The aqueous phase was re-extracted twice and the combined ethyl acetate extractions were pooled and taken to dryness. Residues were taken up in small volumes of ethyl acetate and spotted on 5x20 cm alumina chromatography plates (Polygram ALOX N/UV 254, Camlab, Surrey). Chromatography was effected according to Shaw [21] CAP and its acetylated derivatives were run as markers. Dried plates were examined under UV and autoradiographed by exposure to Kodak Blue Brand X-ray film at 4°C.
Fig. 1. Abscissa: time (days); ordinate: cell no. Growth of (a) the CA-Pa cell line, Mc63 and (b) the parental cell line HeLa B (A, m, in the precence and-A, 0, in the absence of 50’ pg/ml ‘bthreochloramphenicol. Each cell count represents the mean of two observations (see Methods). Arrow indicates number of cells plated at time zero.
Human cell lines resistant to D-threo-chloramphenicol
301
sistant clones. These are referred to as MC63 and MC75. A subpopulation of MC63 was maintained in medium lacking CAP and various properties were checked over a 9 months period. This line is designated MC63NC. Growth characteristics
I
I 'P
100
&ml CAP; ordinare: % incorporation; A-A, HeLa B; W-W, MC63. Effects of o-threo-chloramphenicol on mitochondrial protein synthesis in intact cells. The 100% value refers to the incorporation observed in the presence of 50 ,ug/ml emetine and in the absence of CAP.
Fig. 2. Abscissa:
Spolsky & Eisenstadt [63 and has been previously interpreted as representing the repopulation of a cell with resistant mitochondria (however, see Discussion). The overall frequency of resistant colonies was of the order, 1 per flask, or 1 per 1.2~10~ cells before mutagenesis (the mutagen pretreatment kills about 20% of the original population of cells). From 100 starting cultures, 14 uncloned isolates were obtained and several of these were checked for resistance to CAP of mitochondrial protein synthesis in intact cells (see Methods). Surprisingly, some isolates grown in the presence of 12.5 ,uglml chloramphenicol showed sensitive, or partially sensitive behaviour when checked for the dose response of mitochondrial protein synthesis to the drug. Those which showed significant resistance when compared with the parental cell line were cloned at 100 pg/ml CAP and studied further. Use of this additional selection step gave rise to two independent (i.e. from different original mass cultures) re20-761809
The growth curve of the resistant line MC63 was similar in the presence or absence of 50 pg/ml CAP, reaching confluence in about 6 days (fig. la). The growth of the parental cell line (HeLa B) began to slow rapidly after 3-4 days exposure to 50 fig/ml CAP (fig. 16). Cell death resulted if exposure to the inhibitor was continued. The mean generation time for MC63 in the presence of 50 pg/ml CAP was similar to that of the parental line in the absence of the drug. Under the conditions of the experiment this was about 29 h. Karyotype
analysis
The distributions of chromosome numbers recorded for cells of both parental and resistant lines were unimodal with a sharp peak at 63. The means and standard deviations for the cell lines examined were: HeLa B, 62.4 (1.2); MC63, 62.6 (1.8) and MC75, 62.9 (0.6) (between 17 and 20 metaphases were examined for each line). Effects of chloramphenicol on mitochondrial protein synthesis
Mitochondrial protein synthesis can be examined in whole cells or in isolated mitochondria (in vitro). The system of choice from our studies was an approach in which cytoplasmic protein synthesis in whole cells was blocked by the presence of 50 pg/ml emetine. At this concentration a qualitatively similar effect to that in the presence of cycloheximide (200 pg/ml) was observed, except that emetine was more efficient in Exp Cell Res IO2 (1976)
302
Siegel et al.
. 1.0
100
10.
20
eliminating residual cytoplasmic protein synthesis [ 181. The dose-response curve relating the incorporation of [?S]methionine into mitochondrially synthesized protein to the concentration of CAP in the medium for the parental cell line (HeLa B) is illustrated in fig. 2. When cytoplasmic protein synthesis is blocked, the residual incorporation of label into alkali-stable, TCA-precipitable counts is sensitive to CAP. Of the total incorporation which is CAP sensitive, 50% is inhibited at 7 pglml. This compares with a concentration of 15 pg/ml calculated from the data of Spolsky & Eisenstadt [6] in studies with isolated mitochondria. In our studies with intact cells we have con-
loo-
. -.
-\
sistently observed that about 30% of the emetine resistant protein synthesis (i.e., that presumed to be independent of cytoplasmic ribosomes) is not inhibited by CAP, nor by other inhibitors of mitochondrial protein synthesis (mikamycin, carbomycin). The nature of these residual labelled products is unknown. No major component resolved by SDS polyacrylamide gel electrophoresis is synthesized by CAP-sensitive cells in the presence of both emetine and chloramphenicol (see fig. 6). The resistant cell line MC75 also gave a similar dose response curve to that of MC63 (not shown) indicating that mitochondrial protein synthesis in intact resistant cells is only slightly inhibited at 200
. . .
.
--\
Fig. 3. Abscissa: pg/ml mikamycin; ordinate: % incorporation. The effect of increasing concentrations of mikamycin on mitochondrial protein svnthesis assaved in intact cells. CM, CAPR cell line-(MC63); A-A, CAPS cell line (HeLA B); CO, MC63 NC was maintained 24 months in the absence of selection before testing. Each value represents the mean of three determinations.
.
4. Abscissa: &ml carbomycin; ordinafe: % incorporation. The effect of increasing concentrations of carbomycin on mitochondrial protein svnthesis assaved in intact cells. m-m, CAPR cell line (MC63); A-A, CAPS cell line (HeLa B). Each value represents the mean of three determinations.
Fig.
20
Exp Cell Res 102 (1976)
1.0
10
A100
Human cell lines resistant to D-threo-chloramphenicol
20
1.0
10
pg/ml and not at all at 100 pglml. The stability of the resistance phenotype was evident from the dose response curve obtained with MC63 cultured in the absence of CAP for 9 months. The profiles for resistant cells maintained in the presence, or absence of the selective agent were indistinguishable within the limits of the assay system. In contrast to the dose response profiles of Spolsky & Eisenstadt [6] and of Wallace & Freeman [S], no residual chloramphenicol sensitive protein synthesis (which could reflect mixed populations of sensitive and resistant mitochondria) was observed. The data from Spolsky & Eisenstadt [6] indicated that about 60 % of the incorporation in mitochondria from CAPR cells showed normal sensitivity to the drug. The cross resistance to other antimitochondrial drugs of CAPR cell lines
Several antibiotics other than chloramphenicol have been reported to inhibit protein synthesis on mitoribosomes (e.g. see [2]). It was therefore of interest to test the sensitivity of mitochondrial protein synthesis to a range of antibiotics. The effects of increasing concentrations of mikamycin and carbomycin on the mitochondrial protein
.
100
303
Fig. 5. Abscissa: pg/ml EB; ordinate: % incorporation. The effect of increasing EB concentrations on mitochondrial protein synthesis assayed in intact cells. Cm, CAPR cell line (MC63); A-A, CAPS cell line (HeLa B). Each value represents the mean of three determinations. Preincubation with EB was continued for 30 min before addition of [3JS]methionine (see text).
synthesis (assayed in intact cells) of CAPR and CAPS cell lines are illustrated in figs 3 and 4. Cell lines which are resistant to chloramphenicol show increased resistance to both of these inhibitors. In contrast to the inhibitors discussed previously, EB is thought to prevent mitochondrial protein synthesis by inhibition of the replication and transcription of mitDNA rather than by direct action at the level of translation. Perlman et al. [223 found complete elimination of detectable mitochondrial mRNA after 25 min preincubation with the drug at 1 pg/ml. Lederman & Attardi [23] demonstrated 90 % reduction in the incorporation of labelled amino acids into mitochondria in vitro after treatment with 1 pg/ml EB for only 10 min. Similarly, in our system, 30 min preincubation with EB was sufficient to inhibit mitochondrial protein synthesis as shown in fig. 5. Both CAPR and CAPS cells appear to show similar sensitivity to the action of this inhibitor. Acrylamide gel analysis of the products of mitochondrial protein synthesis
Resistance to CAP could have arisen by an alteration to, or loss of, a specific mitochondrially synthesized protein. AlternaExp Cell Res 102 (1976)
304
Siegel et al. HeLa B
MC75
MC63
MC63-NC
Front
Fig. 6. Electrophoretic separation of the proteins synthesized by mitochondria in intact cells as identified by autoradiography (see Methods). Profiles of the proteins labelled at 0, 50 and 200 fig/ml CAP are
compared for both CAPR (MC63, MC63NC and MC79 and CAPS (HeLa B) cell lines. MC63 NC was examined after 14 months growth in the absence of chloramphenicol.
tively, the observed resistant incorporation may have been into protein not associated with mitochondria. For these reasons the spectrum of proteins synthesized in the various cell lines in the presence and absence of inhibitors was examined (see Methods). They were compared with the normal pattern of proteins known to be synthesized by human mitochondria in intact cells. The patterns of labelled proteins resolved after parental and resistant cell lines were treated according to the analytical procedure, described in Methods, are
illustrated in fig. 6. Note the complete inhibition of mitochondrial protein synthesis in the HeLa parental line by 50 and 200 pg/ml CAP. A similar response has been observed for a mouse cell line (RAG) not shown. In contrast, the incorporation of label into mitochondrial proteins observed in the resistant cell lines MC63 and MC75 does not appear to be inhibited at 50 or 200 pg/ml. Resistant cells maintained in the absence of the selective agent for 1.5 months also incorporated label into a similar spectrum of proteins. Our observations also in-
Exp CeNRes 102 (1976)
Human cell lines resistant to D-threo-chloramphenicol
Table 1. Uptake of [14C]chloramphenicol CAPR and CAPS cell lines
Cell line
% cell volume freely accessible to (i.e. in equilibrium with) external CAP
D98 (CAPS) HeLa B (CAPS) MC63 (CAPR)
30 29 37
by
3 X 10s cells were incubated in 0.2 ml PBS containing 8 pg/ml [C”]CAP and 350 pg/ml unlabelled D-the0 CAP for 15 min at 37°C. Two ml ice-cold PBS were added to the cell susuension. After 5 min at O’C, suspensions were rapidly filtered onto Whatman GF/C filters and washed with 3 x 10 ml ice-cold PBS. Filters were dried and counted.The cellular volume accessible to CAP was calculated as the volume of the original ambient medium which contained the same amount of PX’lCAP as that associated with a known packed cell-volume. Cell volume has been taken as the volume of packed cells after centrifuging a suspension in PBS for 1 min in a Beckman microfuge.
dicate that the qualitative distribution and relative positions of labelled material resolved by electrophoresis are similar for the non-inhibited parental line and for resistant cell lines incubated in the presence of chloramphenicol. Uptake of [%]chloramphenicol by sensitive and resistant cell lines
D98 cells suspended at 37°C in PBS rapidly assimilated [14C]CAP (present at 8 pg/ml in the external medium). Uptake was estimated from the radioactivity retained in cells after dilution into ice-cold saline (to wash any drug from low-affinity extracellular sites) and subsequent harvesting of the labelled cells on glass fibre filters. CAP uptake reached a plateau in about 10 min and appeared to result from the passive diffusion of drug into cells. Similar uptake kinetics were observed with cells which had been pretreated with cytotoxic agents (0.2% w/v) formaldehyde, or 0.5 mM KCN). When uptake experiments were repeated in the presence of excess unlabelled
305
D-threo-CAP (350 pg/ml), the plateau level of label associated with cells was reduced by about 30%. Thus, 30% of CAP taken up at 8 pg/ml represents binding to saturable sites and can be competed out with excess unlabelled CAP. The remaining 70% of cellular CAP is presumably present in intracellular pool(s) (plus any residual lowaffinity CAP binding sites). The size of this pool was found to correspond to about 30% of the packed volume of CAPS cells (table 1). The presumed intracellular localization of the [14C]CAP accumulated under these conditions is supported by the observation that 65 % (mean of 2 determinations) could be released from CAPS,HeLa B cells by 3 cycles of freezing in ethanol/ice and thawing. After this treatment, 99% of cells could no longer exclude trypan blue, yet were found to be essentially intact on microscopic examination. The kinetics of CAP uptake into CAPR cells (MC63 and MC75) were very similar to those of CAPS cells. Furthermore, the size of the intracellular pool freely accessible to the drug (as well as the number of saturable binding sites/cell) was similar in CAPR and CAPS cells (table 1). Also, more than 65 % of CAP present in this pool could be released from both MC63 and MC75 by freeze-thaw treatment of cells. Thus it appears that CAP can freely permeate the plasma membrane of CAPR cells and has normal access to a substantial proportion of the cells’ interior (see Discussion). High concentrations of chloramphenicol are known to inhibit cytoplasmic synthesis in mammalian cells [24]. If CAPR resulted from altered permeability, the dose response curve for the inhibitory action of the drug on total cullular protein synthesis would also be expected to differ for CAPR and CAPS cells. Fig. 8 demonstrates that the dose response curves relating total proExpCeNRes lOZ(1976)
306
Siegel et al.
Modification of chloramphenicol by acety!ation (as in some bacteria harbouring appropriate R factors) should result in altered chromatographic behaviour of the products [21]. After incubation for 1, 2, or 3 days in the presence of labelled chloramphenicol, cells from test cultures and their respective media were combined and extracted with ethyl acetate. About 97% of
total counts were extracted by this procedure. Samples of the ethyl acetate extracts were chromatographed on alumina thin layer chromatography plates in ascending fashion using 85 : 15 (v/v) benzenemethanol as solvent and the positions of chloramphenicol (and any derivatives) localized by autoradiography. A mixture of [r4C]CAP together with D-threo-CAP and its mono- and di-acetoxy derivatives were chromatographed as markers. The [‘“Clchloramphenicol used in these experiments was resolved by chromatography into one major component (very faint minor faster migrating spots of unknown identity were also detected). Identical chromatographic profiles were obtained for CAP extracted from the sensitive and resistant cell lines and from controls which were treated identically but contained no cells (fig. 8). This suggests that, unless the modification products co-chromatograph with CAP in this system, neither resistant nor parental cell lines modify chloramphenicol to an ethyl acetate soluble form. One of the major routes of chloramphenico1 detoxification in humans is conjugation with glucuronic acid [25]. The glucuronide is insoluble in ethyl acetate [26]. In the studies reported above, no significant increase in the ethyl acetate insoluble frac-
Fig. 8. Thin-layer chromatography of ethyl acetatesoluble material from medium + cells after incubation of CAPS and CAPn cells for 1, 2, or 3 days with [“Cl
CAP. Arrows indicate position of unlabelled CAP and its acetoxy derivates run as markers and detected by fluorescence quenching under IJV light.
.
Fig. 7. Abscissa: pglml CAP; ordinate: % incorporation. Effect of increasing concentrations of chloramphenico1 on total cell protein synthesis in the absence of emetine. A-A, HeLa B; Cm, MC63; O-O, MC 75; CO, MC63 NC.
tein synthesis and chloramphenicol concentration for the parental cell line and resistant variants are indistinguishable, again suggesting that modification to the plasma membrane is unlikely to be the major factor in conferring resistance. The possible modification chloramphenicol
Exp Cell Res 102 (1976)
of
Human cell lines resistant to D-threo-chloramphenicol Table 2. Percentage of total [VJCAP present in aqueous phase after extraction of medium + cells with ethyl acetate
Cell line
Total dpm [W]CAP present
% Total present in aqueous phase
Control (no cells) HeLa B MC63 MC75
228ooo 192000 224000 215000
3.2 3.5 3.8 3.0
Various cell lines were previously incubated for 3 days in the presence of 12.5 &ml D-three-CAP + 1 UK (60000 dpm) [Y]CAP/mi.-Total culture volume was about 3 ml in each case. Extraction in ethyl acetate was performed as described in the text.
tion of [14C]CAP were observed following incubation with CAP resistant cell lines for 3 days (see table 2). It does not therefore seem probable that resistance results from induction of enzymes capable of catalysing the conjugation of chloramphenicol with glucuronic acid. DISCUSSION Chloramphenicol resistant mammalian cell lines have been isolated previously [6, 8, 91 and a comparison of the results presented here with those previously reported provides some points of interest. The mutagenesis and subsequent selection procedure closely followed that of Spolsky & Eisenstadt [6]. Resistant cells remained quiescent for several weeks after the start of selection, prior to the appearance of colonies. This was not the behaviour expected for a typical nuclear mutation but could be interpreted as a mitochondrial mutation, if growth in chloramphenicol required a gradual repopulation of the cell with resistant mitochondria. However, even if the variant results from an alteration at the level of mitDNA, the situation is probably complex. Recent re-
307
ports indicate that HeLa cells have about 9000 mitochondrial genomes distributed between an unknown number of discrete organelles [27]. The initial slow growth of cells in the presence of chloramphenicol could reflect selection in favour of mutant genomes. The overall dose-response to the drug in terms of cell viability and growth would thus be related to the proportion of ‘mutant’ genome copies. The rapid appearance of fast growing colonies resistant to the chloramphenicol analogue, Tevenel, in the mouse cell line LMTK- [8] has been attributed to the presence of a high proportion of cells in the population containing pre-existing mutant mitochondrial genomes. If this interpretation is correct, cells carrying variant mitochondrial genomes of this type do not appear to be selected against in the absence of the antimitochondrial drug. Other workers have employed mitochondrial preparations in vitro to examine the dose-response of mitochondrial protein synthesis to increasing drug concentration. We have examined mitochondrial protein synthesis in whole cells by blocking cytoplasmic protein synthesis with 50 pg/ml emetine. Under these conditions, the products of mitochondrial protein synthesis in parental type cells are sensitive to 50 pg/ml CAP and can be resolved by electrophoresis in SDS-containing polyacrylamide gels into several (about 9) discrete components. Extremely reproducible patterns were obtained and data for inhibition curves with intact cells have been very consistent. We also attempted to compare data for mitochondrial protein synthesis in whole cells with preparations in vitro. However, rather variable results were obtained for isolated mitochondria, even with the parental line. Furthermore, reproducible patterns for the products of mitochondrial protein synthesis Exp CdlRes 102 (1976)
308
Siegel
et ul.
in vitro were not obtained. The problems encountered probably reflect procedural difficulties in reproducibly isolating mitochondria which are relatively uncontaminated by cytoplasmic membranes and ribosomes, but which are still capable of supporting protein synthesis with endogenously generated ATP. In our experience, therefore, the approach with intact cells was more reliable. However, its main disadvantage is that it does not provide information concerning the involvement of the plasma membrane in contributing to the resistant phenotype. We have attempted to obtain information relevant to this possibility by comparative studies on the uptake and release of labelled chloramphenico1from sensitive and resistant cells. About 30% of the packed cell volume of both resistant and parental cell lines was found to be freely accessible to external [14C]CAP, under conditions where the binding of label to high affinity saturable sites was diluted out with excess unlabelled CAP. (Specific D-threo-CAP binding sites have been identified in cultured human cells and are the subject of a separate communication, Jeffreys, Siegel & Craig, in preparation). The estimate of 30% for the proportion of cell volume accessible to CAP must be interpreted as a minimum value; the true total cell volume being less than the volume of the packed cell pellet. The suggestion that radioactivity associated with washed cells represents intracellular CAP is supported by the rapid release of the majority of such label by procedures designed to disrupt cell membranes. Resistant and sensitive cells behaved similarly both in terms of accumulation and subsequent release of labelled chloramphenicol. Therefore, it is unlikely that differences in response to the drug result from alterations in permeability at the plasma membrane. Exp Cell Res 102 (IQ76)
Possible alterations to the permeability of the mitochondrial membrane are less easily investigated. Although attempts have been made to disrupt the normal permeability barrier of mitochondria in vitro with Triton X-100 [6, 8, 93 it is not clear to what extent this would assist the entry of chloramphenicol. Procedures which offer a more direct approach, such as studies on the CAP sensitivity of isolated mitoribosomes in catalysing polyuracil-dependent phenylalanine incorporation [28] or assays of peptidy1 transferase activity [29] require large quantities of purified mitoribosomes and have yet to be reported for cultured cells. Another possible mechanism of CAP resistance is that of modification as observed in drug resistant bacteria [21]. Still another detoxification system exists in mammalian liver [25], namely the conjugation of CAP with glucuronic acid. Resistance in the human cell line could possibly have originated by the induction of certain liver specific functions or by the acquisition of chloramphenicol acetyl transferase activity. However, neither CAPR nor parental HeLa cells appeared to modify chloramphenicol by either of these mechanisms. The cross resistance observed to some other antibiotics also argues against the acquisition of resistance by a specific inactivation mechanism. The observed cross-resistance of mitochondrial protein synthesis in CAPR cell lines to mikamycin and carbomycin contrasts with the behaviour of some other CAP resistant cell lines which show little or no cross resistance to these antibiotics (Eisenstadt, personal communication). Mitochondria from the human CAP resistant cell line described by Mitchell et al. [9] showed extensive cross resistance to mikamycin and carbomycin in vitro and mitochondrial protein synthesis in intact cells
Human cell lines resistant to D-threo-chloramphenicol
exhibited very similar increases in resistance to those reported here. There is a precedent in CAP resistant yeast for mutants which show cross resistance to these antibiotics. This suggests that the presumptive CAP target may share a common or overlapping site with mikamytin and carbomycin. Chloramphenicol resistant yeast strains show variable phenotypes. Whereas some mutants were reported to show no cross resistance to other antibiotics tested [30] it is clear from studies on isolated mitoribosomes that other mutants, selected for CAP resistance, can show increased resistance to antibiotics such as erythromycin [3 11.Other, apparently cytoplasmically inherited, CAP resistant mutants in yeast showed a high level of cross resistance to mikamycin and carbomycin [32]. It was thought that this reflected a changed mitochondrial membrane permeability (however, see [31]). There is little information available for other eukaryotes. Nevertheless, cytoplasmically inherited erythromycin resistant mutants of Paramecium were found to be cross resistant to mikamycin [33] and the binding of [14C]CAP to isolated ribosomes from rat liver mitochondria was partly inhibited by carbomycin [28]. The conclusion from the bulk of these studies is that the observation of cross resistance does not exclude the possibility of a mutant expressed at the level of mitochondrial ribosomes. The profile of mitochondrially made proteins of the resistant mutants observed on SDS polyacrylamide gel electrophoresis was identical to that of the parental cell type. Thus gross alteration to, or elimination of, a specific mitochondrially synthesized protein does not appear to be an explanation of the resistant phenotype. However, minor changes and single amino acid substitution would not have
309
been detected. Neither is any information available concerning possible alterations to mitochondrial ribosomal RNAs. In the absence of further data, it is impossible to discriminate between the possibilities of a modification to mitochondrial membranes or to mitochondrial ribosomes, or perhaps to both. However, it should be pointed out that cross resistance is not without specificity; EB (which is assumed to act by blocking mitochondrial transcription) is equally inhibitory to CAP resistant and to CAP sensitive cells. Implicit in the foregoing discussion has been the assumption that the CAP resistance could be controlled by mitDNA. Bunn et al. [7] have previously documented the transfer of the resistant phenotype by cytoplasmic fragments in mouse L cells. Results consistent with the suggestion that the mutants described in this paper are cytoplasmitally controlled will be presented elsewhere. This work was supported in part by grants from NATO, the MRC and the Cancer Research Campaign. L.S. was supported by grants USPHS, GM 1511 and GM 210%. also M.S.T.P. GM02016.
REFERENCES 1. Gillham, NW, Ann rev genet 8 (1974) 347. 2. Lloyd, D, Mitochondria of microorganisms, p. 285. Academic Press, London (1974). 3. Rowlands, R T & Turner, G, Mol gen genet 126 (1973) 201. 4. Beale, G, Knowles, J & Tait, A, Nature 235 (1972) 3%. 5. Roberts, C T & Orias, E, Genetics 73 (1973) 259. 6. Spolsky, C M & Eisenstadt, J M, FEBS lett 25 (1972) 319. 7. Bun& C L, Wallace, D C & Eisenstadt, J M, Proc natl acad sci US 71 (1974) 1681. 8. Wallace, R B & Freeman; K B, J cell bio165 (1975) 492. 9. Mitchell, C H, England, J M & Attardi, G, Som cell genet l(l975) 215. 10. Klietmann, W, Sato, N & Nass, M M K, J cell bio158 (1973) 11. 11. Craig, I W, Proc 2nd int symp genet ind microorganisms, Academic Press, New York (1976). 12. Gordon, S, J cell bio167 (1975) 257.
.ExpCell
Res 102 (1976)
310
Siegel
et al.
13. Clayton, D A, Teplitz, R L, Nabholz, M, Dovey, H & Bodmer, W F, Nature 234 (1971) 560. 14. Attardi, B & Attardi, G, Proc natl acad sci US 69 (1972) 129.
26. Della Bella. D. Ferrari. V. Marca. G & Bonanomi.
L, Biochem pharmacoi 17(1%8) 2381. 27. Bonenhaeen. D & Clavton. D A. J biol them 249
(1974) 7997.’
*
15. David, I B, Horak, I & Coon, H G, Genetics 78
28. Ibrahim, N G, Burke, J P & Beattie, D S, J biol
(1974)459. 16. Craig, I W, Jeffreys, A J, Siegel, R L & Sly, W,
29. Denslow, N D & O’Brien, T W, Biochem biophys
FEBS 10th congress abstracts p. 158(1975). 17. Siegel, R L, PhD Thesis Washington University, St Louis, MO (1976). In preparation. 18. Jeffreys, A & Craig, I, Biochem j 144 (1974) 161. 19. - Biochem sot transact 3 (1975) 398. 20. Laemmh, U K, Nature 227 (1970) 680. 21. Shaw, W V, J biol them 242 (1967) 687. 22. Perlman, S, Abelson, H T & Penman, S, Proc natl acad sci US 70 (1973) 350. 23. Lederman, M & Attardi, G, J mol biol 78 (1973) 275. 24. Haldar, D & Freeman, K B, Can j biochem 46 (1%8) 1009.
25. Glazko, A J, Antimicrobial agents and chemotherapy, Am sot microbial, p. 655. (1%7).
Exp Cell Res 102 (1976)
them 249 (1974) 6806. res commun 57 (1974) 9. 30. Molloy, P L, Howell, N, Plummer, D T, Linnane,
A W & Lukins, H B, Biochem biouhvs res commun 52 (1973) 9. 31. Grivell, L A, Netter, P, Borst, P & Slonimski, P P, Biochim biophys acta 312 (1973) 358. 32. Bunn, C L, Mitchell, C H, Lukins, H B & Linnane, A W, Proc natl acad sci US 67 (1970) 1233. 33. Queiroz, C & Beale, G H, Genet res 23 (1974) L
233.
Received March 1, 1976 Accepted May 6, 1976
-
Experimental
ISOLATION CELL
Cell Research 102 (1976) 29&310
AND DETAILED LINES
RESISTANT
CHARACTERIZATION
OF HUMAN
TO D-THREO-CHLORAMPHENICOL
R. L. SIEGEL,’ A. J. JEFFREYS,’ W. SLY* and 1. W. CRAIG’ ‘Genetics Laboratory, Department of Biochemistry, University of Oxford. Oxford OX1 3QV, UK, and %chool of Medicine, Washington tJniversity, St Louis, MO 63110, USA
SUMMARY The isolation and characterization of chloramphenicol resistant derivatives of the human cell line HeLa B is described. Growth of resistant lines was unaffected in the presence of 100 pg/ml Dthreo-chloramphenicol, whereas growth of the parental cells was inhibited at 12.5 pg/ml. The incorooration of P?Slmethionine into mitochondriai orotein of intact resistant cells continued normaliy in the presence of 100 pg/ml chloramphenicbl (cytoplasmic protein synthesis was blocked by addition of 50 &ml emetine). Under these conditions the electroohoretic motile of labelled. presumptive mitochondrially-made proteins was similar to that of the parental cell line labelled in the absence of chloramphenicol. The cell lines selected in the presence of chloramphenicol also showed increased resistance to some other inhibitors of mitochondrial protein synthesis, e.g. carbomycin and mikamvcin. f’~lChloramphenicol was found to have normal access to the interior of resistant cells and it is therefore unlikely that resistance results from altered cell permeability. No modification of the drug by acetylation or glucuronide conjugation mechanisms was observed. The possibilities remain that resistance is mediated by altered permeability of the mitochondrial membrane, or from modification to a component of the mitochondrial protein synthetic system.
Chloramphenicol is one of a group of antibiotics which specifically inhibit mitochondrial and bacterial protein synthesis, but do not inhibit (in the same dose range) protein synthesis on cytoplasmic ribosomes of eukaryotes. Mutants resistant to chloramphenicol (CAP) and/or other inhibitors acting at the level of mitochondria have been isolated in yeast [ 1, 21, Aspergillus [3], Paramecium [4] and Tetrahymena [5]. Some of these have been shown to be cytoplasmically inherited and could therefore result from mutations at the level of mitochondrial DNA (mitDNA). Recent reports [6, 71 have shown that mutants of cultured mammalian cells can be Exp
Cd
Res
102 (1976)
obtained which are resistant to CAP. Human and mouse cells have been isolated which were capable of growth in 25 and 50 pg/ml of the drug. Protein synthesis in isolated mitochondria was resistant, at least in part, to CAP (see Discussion). The resistance in mouse cells could be transferred by enucleated fragments to sensitive cells of the same species. Since these observations, Wallace & Freeman [8] have isolated mouse cell lines which are resistant to the chloramphenicol analogue, Tevenel. and Mitchell et al., have reported the selection of a human cell line capable of growth in 100 pg/ml CAP [9]. Resistance to other drugs known to in-
Human cell lines resistant to D-threo-chloramphenicol hibit mitochondrial function has also been observed in cultured cells. Klietmann et al. [lo] reported the isolation of SV40 transformed hamster cells capable of growth in the presence of normally toxic concentrations of ethidium bromide (EB). However, the mechanisms conferring resistance to EB, or to CAP in cultured cells are poorly understood. We now report in detail on the properties of a human cell line which is resistant to 100 pg/ml D-threo CAP. The cell line was isolated with the intention of subsequently employing it in studies on mitochondrial biogenesis. Several genes for human mitochondrial functions have been assigned to human chromosomes (see [ 11, 121).However, human mitDNA is apparently lost in fusions between human cells and established mouse cell lines [13-141. In contrast, fusions between human cell lines and embryonic mouse cells yielded hybrids capable of propagating mitochondrial genomes of both species. In some cases, the human and rodent sequences were apparently linked, suggesting interspecific recombination of mitDNA [ 151. It was hoped that CAP resistance could be employed as a marker to select for the presence of human mitDNA sequences in hybrid cells. This may provide an opportunity to identify human chromosome(s) possibly necessary for propagation of human mitDNA, or to characterise recombination events between mammalian mitochondrial genomes. A preliminary report of some aspects of this study has appeared previously [ 161.
METHODS Cells, media and growth conditions CAP resistant cell lines were isolated from HeLa B, a subclone derived from the S strain of HeLa (the kind gift of Dr L. Tolmach). Cells were grown initially in
299
Eagles minimum essential medium (GIBCO, Grand Island, N.Y.) supplemented with fetal calf serum (10% v/v), penicillin (100 W/ml) and streptomycin (100 pg/ml). Following early growth in 75 cm* tissue culture flasks (Falcon Plastics, Oxnard, Calif.) cells were grown in glass bottles, or in spinner culture, employing RPM1 1640 medium (Flow Laboratories Limited, Irvine, Scotland) supplemented with serum and antibiotics as above. Pure o-threo-chloramphenicol was kindly supplied by Dr H. E. Machamer (Parke, Davis and Co., Detroit, Mich.). It is important to note that different commercial sources of CAP vary widely in their ability to inhibit mitochondrial protein synthesis. This presumably results from contamination with inactive Lthreo-isomer. The pure o-threo-CAP was about 2.5x as effective in studies with intact cells as that available commerciallv (Siema Co.. London) and L-threoCAP had little detectible inhibitory action up to 200 &ml 1171.Emetine hvdrochloride was obtained from : Stgma ilondon), mikamycin and EB from Calbiochem (Los Angeles, Calif.) and carbomycin from Pfizer Inc. (Groton, Conn.) MEM lacking methionine was obtained from Biocult Laboratories Ltd. (Paisley, Scotland). Phosphate-buffered saline (PBS) was 0.85% (w/v) NaCl, KH,PO, (2.5 mM) and Na2HP0, (10 mM).
Mutagenesis and selection Mutagenesis was performed employing a modification of the procedure described by Spolsky & Eisenstadt [6]. The cells (HeLa B) were plated at 106-2~ 1Ogper 75 cm2 tissue culture flask and treated with EB (0.05 PM) for 24 h. The medium was removed and replaced with one containing N-methyl-N-nitro-N nitrosoguanidine (0.05 pg/ml) for 3 h. After washing, the cells were subsequently selected in medium containing sthreo-chloramphenicol (50 &g/ml). After two weeks the majority of cells began to die. Selection was continued for 6 months during which time some of the remaining cells apparently began to increase in size and finally cell division was observed. In an attempt to facilitate outgrowth of the putative resistant cells, the CAP concentration was reduced to 12.5 p.gl ml. Rapid cell proliferation followed and several independent uncloned populations were isolated. These were tested for the resistance of mitochondrial protein synthesis to chloramphenicol. Those which showed significant resistance were subject to cloning in 100 pg/ml CAP. Two of the cloned isolates, MC63 and MC75, were chosen for detailed investigation. They were routinely maintained in medium containing 50 pglml CAP.
Growth curves Cells were plated in plastic dishes (5.0 cm diam) in normal medium. This was changed 24 h later, either to unmodified medium or to one containing o-threoCAP (50 &ml). At zero time and at subsequent regular intervals cells on duplicate dishes were harvested and counted. Exp Cell Res 102 (1976)
300
Siegel et al.
Assay for mitochondrial protein synthesis in intact cells The assay was based on conditions established oreviously [is, 191.Cells (0.5x106-2.0~ 1oB)were plated in plastic dishes (5 cm diam.) in RPM1 medium 1640 12-17 h before use. The plates were washed twice with PBS and 1 ml of MEM-methionine, supplemented as described previously [18] and containing emetine (SO pg/ml), was added. One hundred ~1 of various antibiotic solutions, or PBS, were added to produce the desired final concentrations (medium without antibiotics served as a control). Plates were preincubated with antibiotics for 30 min at 37°C prior to addition of label. The subsequent steps were similar to those described previously except that the label was 1 &i [YS]-methionine (170-240 Cilmmole) and incornoration of label continued for 2 h. Subseauently, the cells were harvested and suspended in n&ma1 media and centrifuged (500 g 5 min). The wet cell pellet was resuspended in 1 ml NaOH (1 N) and incubated at 37°C for 10 min. One ml HCl (1 N) was added followed by 8 ml trichloracetic acid (TCA 5% w/v) containing L-methionine (10 mM). The suspension was chilled on ice and filtered through GF/Cglass fibre filters (Millipore Ltd, London). Filters were washed with 3~ 10 ml TCA+methionine, airdried and placed in 10 ml scintillation fluid before counting.
Electrophorectic mitochondrially
Mutant
isolation
Mutagenesis and selection of cells resistant to chloramphenicol were carried out as described in Materials and Methods. Attention should be drawn to the length of time required for cell growth (6-9 months). This is consistent with the observations of
lb t
I
II
I
‘.
105
t
Ii
t I
\
of chloramphenicol
[“CID-threo-chloramphenicol [W]CAP was purchased from the Radiochemical Centre, Amersham, with money made available through the good offices of Dr J. A. -L. Gorringe, Director of Cli&al Investigation, Parke Davis & Co., Pontypool. The specific activity of the drug was 27.3 @i/mg and was labelled in the 1-C position of the dichloroacetamido group. Its radiochemical purity was 99%, but no details were available of the contamination with the L-threo isomer. Falcon flasks (25 cme) containing about 10 cells in 3 ml medium containing D-threo*AP (12.5 &ml) plus r4C-labelled CAP (1 pg, 6~ lW dpm/ml) were inExp Cell Res 102 (1976)
RESULTS
separation of synthesized proteins
Cultured cells were washed in PBS, resuspended in incorporation medium to 106-3 x 106cells/ml and preincubated in the presence of emetine (100 pg/ml) and amounts of CAP as described previously [ 181. Thirty $i [YS]methionine/ml were added and incubation continued for 3 h at 37°C. The incorporation was followed by a ‘cold chase’ with 1 mM L-methionine for 20 min and the cells subsequently pelleted and washed with PBS. Cell pellets were stored in liquid nitrogen. Electrophoresis of whole cell extracts, made by solubilizing cells at 100°Cin 2 % (w/v) sodium dodecyl sulphate (SDS) containing 2-mercaptoethanol (5 % w/v), was effected in slab polyacrylamide gels (12%) employing the buffer system of Laemmli [20]. Dried gels were subsequently autoradiographed on Kodak Blue Brand X-ray film.
Chromatography and derivatives
cubated for 1, 2, or 3 days. After incorporation, cells were detached by trypsinization. Cells, trypsin and medium were pooled and extracted with an equal volume of ethyl acetate at 0°C with agitation. The organic phase was separated by centrifugation at 500 g for 5 min. The aqueous phase was re-extracted twice and the combined ethyl acetate extractions were pooled and taken to dryness. Residues were taken up in small volumes of ethyl acetate and spotted on 5x20 cm alumina chromatography plates (Polygram ALOX N/UV 254, Camlab, Surrey). Chromatography was effected according to Shaw [21] CAP and its acetylated derivatives were run as markers. Dried plates were examined under UV and autoradiographed by exposure to Kodak Blue Brand X-ray film at 4°C.
Fig. 1. Abscissa: time (days); ordinate: cell no. Growth of (a) the CA-Pa cell line, Mc63 and (b) the parental cell line HeLa B (A, m, in the precence and-A, 0, in the absence of 50’ pg/ml ‘bthreochloramphenicol. Each cell count represents the mean of two observations (see Methods). Arrow indicates number of cells plated at time zero.
Human cell lines resistant to D-threo-chloramphenicol
301
sistant clones. These are referred to as MC63 and MC75. A subpopulation of MC63 was maintained in medium lacking CAP and various properties were checked over a 9 months period. This line is designated MC63NC. Growth characteristics
I
I 'P
100
&ml CAP; ordinare: % incorporation; A-A, HeLa B; W-W, MC63. Effects of o-threo-chloramphenicol on mitochondrial protein synthesis in intact cells. The 100% value refers to the incorporation observed in the presence of 50 ,ug/ml emetine and in the absence of CAP.
Fig. 2. Abscissa:
Spolsky & Eisenstadt [63 and has been previously interpreted as representing the repopulation of a cell with resistant mitochondria (however, see Discussion). The overall frequency of resistant colonies was of the order, 1 per flask, or 1 per 1.2~10~ cells before mutagenesis (the mutagen pretreatment kills about 20% of the original population of cells). From 100 starting cultures, 14 uncloned isolates were obtained and several of these were checked for resistance to CAP of mitochondrial protein synthesis in intact cells (see Methods). Surprisingly, some isolates grown in the presence of 12.5 ,uglml chloramphenicol showed sensitive, or partially sensitive behaviour when checked for the dose response of mitochondrial protein synthesis to the drug. Those which showed significant resistance when compared with the parental cell line were cloned at 100 pg/ml CAP and studied further. Use of this additional selection step gave rise to two independent (i.e. from different original mass cultures) re20-761809
The growth curve of the resistant line MC63 was similar in the presence or absence of 50 pg/ml CAP, reaching confluence in about 6 days (fig. la). The growth of the parental cell line (HeLa B) began to slow rapidly after 3-4 days exposure to 50 fig/ml CAP (fig. 16). Cell death resulted if exposure to the inhibitor was continued. The mean generation time for MC63 in the presence of 50 pg/ml CAP was similar to that of the parental line in the absence of the drug. Under the conditions of the experiment this was about 29 h. Karyotype
analysis
The distributions of chromosome numbers recorded for cells of both parental and resistant lines were unimodal with a sharp peak at 63. The means and standard deviations for the cell lines examined were: HeLa B, 62.4 (1.2); MC63, 62.6 (1.8) and MC75, 62.9 (0.6) (between 17 and 20 metaphases were examined for each line). Effects of chloramphenicol on mitochondrial protein synthesis
Mitochondrial protein synthesis can be examined in whole cells or in isolated mitochondria (in vitro). The system of choice from our studies was an approach in which cytoplasmic protein synthesis in whole cells was blocked by the presence of 50 pg/ml emetine. At this concentration a qualitatively similar effect to that in the presence of cycloheximide (200 pg/ml) was observed, except that emetine was more efficient in Exp Cell Res IO2 (1976)
302
Siegel et al.
. 1.0
100
10.
20
eliminating residual cytoplasmic protein synthesis [ 181. The dose-response curve relating the incorporation of [?S]methionine into mitochondrially synthesized protein to the concentration of CAP in the medium for the parental cell line (HeLa B) is illustrated in fig. 2. When cytoplasmic protein synthesis is blocked, the residual incorporation of label into alkali-stable, TCA-precipitable counts is sensitive to CAP. Of the total incorporation which is CAP sensitive, 50% is inhibited at 7 pglml. This compares with a concentration of 15 pg/ml calculated from the data of Spolsky & Eisenstadt [6] in studies with isolated mitochondria. In our studies with intact cells we have con-
loo-
. -.
-\
sistently observed that about 30% of the emetine resistant protein synthesis (i.e., that presumed to be independent of cytoplasmic ribosomes) is not inhibited by CAP, nor by other inhibitors of mitochondrial protein synthesis (mikamycin, carbomycin). The nature of these residual labelled products is unknown. No major component resolved by SDS polyacrylamide gel electrophoresis is synthesized by CAP-sensitive cells in the presence of both emetine and chloramphenicol (see fig. 6). The resistant cell line MC75 also gave a similar dose response curve to that of MC63 (not shown) indicating that mitochondrial protein synthesis in intact resistant cells is only slightly inhibited at 200
. . .
.
--\
Fig. 3. Abscissa: pg/ml mikamycin; ordinate: % incorporation. The effect of increasing concentrations of mikamycin on mitochondrial protein svnthesis assaved in intact cells. CM, CAPR cell line-(MC63); A-A, CAPS cell line (HeLA B); CO, MC63 NC was maintained 24 months in the absence of selection before testing. Each value represents the mean of three determinations.
.
4. Abscissa: &ml carbomycin; ordinafe: % incorporation. The effect of increasing concentrations of carbomycin on mitochondrial protein svnthesis assaved in intact cells. m-m, CAPR cell line (MC63); A-A, CAPS cell line (HeLa B). Each value represents the mean of three determinations.
Fig.
20
Exp Cell Res 102 (1976)
1.0
10
A100
Human cell lines resistant to D-threo-chloramphenicol
20
1.0
10
pg/ml and not at all at 100 pglml. The stability of the resistance phenotype was evident from the dose response curve obtained with MC63 cultured in the absence of CAP for 9 months. The profiles for resistant cells maintained in the presence, or absence of the selective agent were indistinguishable within the limits of the assay system. In contrast to the dose response profiles of Spolsky & Eisenstadt [6] and of Wallace & Freeman [S], no residual chloramphenicol sensitive protein synthesis (which could reflect mixed populations of sensitive and resistant mitochondria) was observed. The data from Spolsky & Eisenstadt [6] indicated that about 60 % of the incorporation in mitochondria from CAPR cells showed normal sensitivity to the drug. The cross resistance to other antimitochondrial drugs of CAPR cell lines
Several antibiotics other than chloramphenicol have been reported to inhibit protein synthesis on mitoribosomes (e.g. see [2]). It was therefore of interest to test the sensitivity of mitochondrial protein synthesis to a range of antibiotics. The effects of increasing concentrations of mikamycin and carbomycin on the mitochondrial protein
.
100
303
Fig. 5. Abscissa: pg/ml EB; ordinate: % incorporation. The effect of increasing EB concentrations on mitochondrial protein synthesis assayed in intact cells. Cm, CAPR cell line (MC63); A-A, CAPS cell line (HeLa B). Each value represents the mean of three determinations. Preincubation with EB was continued for 30 min before addition of [3JS]methionine (see text).
synthesis (assayed in intact cells) of CAPR and CAPS cell lines are illustrated in figs 3 and 4. Cell lines which are resistant to chloramphenicol show increased resistance to both of these inhibitors. In contrast to the inhibitors discussed previously, EB is thought to prevent mitochondrial protein synthesis by inhibition of the replication and transcription of mitDNA rather than by direct action at the level of translation. Perlman et al. [223 found complete elimination of detectable mitochondrial mRNA after 25 min preincubation with the drug at 1 pg/ml. Lederman & Attardi [23] demonstrated 90 % reduction in the incorporation of labelled amino acids into mitochondria in vitro after treatment with 1 pg/ml EB for only 10 min. Similarly, in our system, 30 min preincubation with EB was sufficient to inhibit mitochondrial protein synthesis as shown in fig. 5. Both CAPR and CAPS cells appear to show similar sensitivity to the action of this inhibitor. Acrylamide gel analysis of the products of mitochondrial protein synthesis
Resistance to CAP could have arisen by an alteration to, or loss of, a specific mitochondrially synthesized protein. AlternaExp Cell Res 102 (1976)
304
Siegel et al. HeLa B
MC75
MC63
MC63-NC
Front
Fig. 6. Electrophoretic separation of the proteins synthesized by mitochondria in intact cells as identified by autoradiography (see Methods). Profiles of the proteins labelled at 0, 50 and 200 fig/ml CAP are
compared for both CAPR (MC63, MC63NC and MC79 and CAPS (HeLa B) cell lines. MC63 NC was examined after 14 months growth in the absence of chloramphenicol.
tively, the observed resistant incorporation may have been into protein not associated with mitochondria. For these reasons the spectrum of proteins synthesized in the various cell lines in the presence and absence of inhibitors was examined (see Methods). They were compared with the normal pattern of proteins known to be synthesized by human mitochondria in intact cells. The patterns of labelled proteins resolved after parental and resistant cell lines were treated according to the analytical procedure, described in Methods, are
illustrated in fig. 6. Note the complete inhibition of mitochondrial protein synthesis in the HeLa parental line by 50 and 200 pg/ml CAP. A similar response has been observed for a mouse cell line (RAG) not shown. In contrast, the incorporation of label into mitochondrial proteins observed in the resistant cell lines MC63 and MC75 does not appear to be inhibited at 50 or 200 pg/ml. Resistant cells maintained in the absence of the selective agent for 1.5 months also incorporated label into a similar spectrum of proteins. Our observations also in-
Exp CeNRes 102 (1976)
Human cell lines resistant to D-threo-chloramphenicol
Table 1. Uptake of [14C]chloramphenicol CAPR and CAPS cell lines
Cell line
% cell volume freely accessible to (i.e. in equilibrium with) external CAP
D98 (CAPS) HeLa B (CAPS) MC63 (CAPR)
30 29 37
by
3 X 10s cells were incubated in 0.2 ml PBS containing 8 pg/ml [C”]CAP and 350 pg/ml unlabelled D-the0 CAP for 15 min at 37°C. Two ml ice-cold PBS were added to the cell susuension. After 5 min at O’C, suspensions were rapidly filtered onto Whatman GF/C filters and washed with 3 x 10 ml ice-cold PBS. Filters were dried and counted.The cellular volume accessible to CAP was calculated as the volume of the original ambient medium which contained the same amount of PX’lCAP as that associated with a known packed cell-volume. Cell volume has been taken as the volume of packed cells after centrifuging a suspension in PBS for 1 min in a Beckman microfuge.
dicate that the qualitative distribution and relative positions of labelled material resolved by electrophoresis are similar for the non-inhibited parental line and for resistant cell lines incubated in the presence of chloramphenicol. Uptake of [%]chloramphenicol by sensitive and resistant cell lines
D98 cells suspended at 37°C in PBS rapidly assimilated [14C]CAP (present at 8 pg/ml in the external medium). Uptake was estimated from the radioactivity retained in cells after dilution into ice-cold saline (to wash any drug from low-affinity extracellular sites) and subsequent harvesting of the labelled cells on glass fibre filters. CAP uptake reached a plateau in about 10 min and appeared to result from the passive diffusion of drug into cells. Similar uptake kinetics were observed with cells which had been pretreated with cytotoxic agents (0.2% w/v) formaldehyde, or 0.5 mM KCN). When uptake experiments were repeated in the presence of excess unlabelled
305
D-threo-CAP (350 pg/ml), the plateau level of label associated with cells was reduced by about 30%. Thus, 30% of CAP taken up at 8 pg/ml represents binding to saturable sites and can be competed out with excess unlabelled CAP. The remaining 70% of cellular CAP is presumably present in intracellular pool(s) (plus any residual lowaffinity CAP binding sites). The size of this pool was found to correspond to about 30% of the packed volume of CAPS cells (table 1). The presumed intracellular localization of the [14C]CAP accumulated under these conditions is supported by the observation that 65 % (mean of 2 determinations) could be released from CAPS,HeLa B cells by 3 cycles of freezing in ethanol/ice and thawing. After this treatment, 99% of cells could no longer exclude trypan blue, yet were found to be essentially intact on microscopic examination. The kinetics of CAP uptake into CAPR cells (MC63 and MC75) were very similar to those of CAPS cells. Furthermore, the size of the intracellular pool freely accessible to the drug (as well as the number of saturable binding sites/cell) was similar in CAPR and CAPS cells (table 1). Also, more than 65 % of CAP present in this pool could be released from both MC63 and MC75 by freeze-thaw treatment of cells. Thus it appears that CAP can freely permeate the plasma membrane of CAPR cells and has normal access to a substantial proportion of the cells’ interior (see Discussion). High concentrations of chloramphenicol are known to inhibit cytoplasmic synthesis in mammalian cells [24]. If CAPR resulted from altered permeability, the dose response curve for the inhibitory action of the drug on total cullular protein synthesis would also be expected to differ for CAPR and CAPS cells. Fig. 8 demonstrates that the dose response curves relating total proExpCeNRes lOZ(1976)
306
Siegel et al.
Modification of chloramphenicol by acety!ation (as in some bacteria harbouring appropriate R factors) should result in altered chromatographic behaviour of the products [21]. After incubation for 1, 2, or 3 days in the presence of labelled chloramphenicol, cells from test cultures and their respective media were combined and extracted with ethyl acetate. About 97% of
total counts were extracted by this procedure. Samples of the ethyl acetate extracts were chromatographed on alumina thin layer chromatography plates in ascending fashion using 85 : 15 (v/v) benzenemethanol as solvent and the positions of chloramphenicol (and any derivatives) localized by autoradiography. A mixture of [r4C]CAP together with D-threo-CAP and its mono- and di-acetoxy derivatives were chromatographed as markers. The [‘“Clchloramphenicol used in these experiments was resolved by chromatography into one major component (very faint minor faster migrating spots of unknown identity were also detected). Identical chromatographic profiles were obtained for CAP extracted from the sensitive and resistant cell lines and from controls which were treated identically but contained no cells (fig. 8). This suggests that, unless the modification products co-chromatograph with CAP in this system, neither resistant nor parental cell lines modify chloramphenicol to an ethyl acetate soluble form. One of the major routes of chloramphenico1 detoxification in humans is conjugation with glucuronic acid [25]. The glucuronide is insoluble in ethyl acetate [26]. In the studies reported above, no significant increase in the ethyl acetate insoluble frac-
Fig. 8. Thin-layer chromatography of ethyl acetatesoluble material from medium + cells after incubation of CAPS and CAPn cells for 1, 2, or 3 days with [“Cl
CAP. Arrows indicate position of unlabelled CAP and its acetoxy derivates run as markers and detected by fluorescence quenching under IJV light.
.
Fig. 7. Abscissa: pglml CAP; ordinate: % incorporation. Effect of increasing concentrations of chloramphenico1 on total cell protein synthesis in the absence of emetine. A-A, HeLa B; Cm, MC63; O-O, MC 75; CO, MC63 NC.
tein synthesis and chloramphenicol concentration for the parental cell line and resistant variants are indistinguishable, again suggesting that modification to the plasma membrane is unlikely to be the major factor in conferring resistance. The possible modification chloramphenicol
Exp Cell Res 102 (1976)
of
Human cell lines resistant to D-threo-chloramphenicol Table 2. Percentage of total [VJCAP present in aqueous phase after extraction of medium + cells with ethyl acetate
Cell line
Total dpm [W]CAP present
% Total present in aqueous phase
Control (no cells) HeLa B MC63 MC75
228ooo 192000 224000 215000
3.2 3.5 3.8 3.0
Various cell lines were previously incubated for 3 days in the presence of 12.5 &ml D-three-CAP + 1 UK (60000 dpm) [Y]CAP/mi.-Total culture volume was about 3 ml in each case. Extraction in ethyl acetate was performed as described in the text.
tion of [14C]CAP were observed following incubation with CAP resistant cell lines for 3 days (see table 2). It does not therefore seem probable that resistance results from induction of enzymes capable of catalysing the conjugation of chloramphenicol with glucuronic acid. DISCUSSION Chloramphenicol resistant mammalian cell lines have been isolated previously [6, 8, 91 and a comparison of the results presented here with those previously reported provides some points of interest. The mutagenesis and subsequent selection procedure closely followed that of Spolsky & Eisenstadt [6]. Resistant cells remained quiescent for several weeks after the start of selection, prior to the appearance of colonies. This was not the behaviour expected for a typical nuclear mutation but could be interpreted as a mitochondrial mutation, if growth in chloramphenicol required a gradual repopulation of the cell with resistant mitochondria. However, even if the variant results from an alteration at the level of mitDNA, the situation is probably complex. Recent re-
307
ports indicate that HeLa cells have about 9000 mitochondrial genomes distributed between an unknown number of discrete organelles [27]. The initial slow growth of cells in the presence of chloramphenicol could reflect selection in favour of mutant genomes. The overall dose-response to the drug in terms of cell viability and growth would thus be related to the proportion of ‘mutant’ genome copies. The rapid appearance of fast growing colonies resistant to the chloramphenicol analogue, Tevenel, in the mouse cell line LMTK- [8] has been attributed to the presence of a high proportion of cells in the population containing pre-existing mutant mitochondrial genomes. If this interpretation is correct, cells carrying variant mitochondrial genomes of this type do not appear to be selected against in the absence of the antimitochondrial drug. Other workers have employed mitochondrial preparations in vitro to examine the dose-response of mitochondrial protein synthesis to increasing drug concentration. We have examined mitochondrial protein synthesis in whole cells by blocking cytoplasmic protein synthesis with 50 pg/ml emetine. Under these conditions, the products of mitochondrial protein synthesis in parental type cells are sensitive to 50 pg/ml CAP and can be resolved by electrophoresis in SDS-containing polyacrylamide gels into several (about 9) discrete components. Extremely reproducible patterns were obtained and data for inhibition curves with intact cells have been very consistent. We also attempted to compare data for mitochondrial protein synthesis in whole cells with preparations in vitro. However, rather variable results were obtained for isolated mitochondria, even with the parental line. Furthermore, reproducible patterns for the products of mitochondrial protein synthesis Exp CdlRes 102 (1976)
308
Siegel
et ul.
in vitro were not obtained. The problems encountered probably reflect procedural difficulties in reproducibly isolating mitochondria which are relatively uncontaminated by cytoplasmic membranes and ribosomes, but which are still capable of supporting protein synthesis with endogenously generated ATP. In our experience, therefore, the approach with intact cells was more reliable. However, its main disadvantage is that it does not provide information concerning the involvement of the plasma membrane in contributing to the resistant phenotype. We have attempted to obtain information relevant to this possibility by comparative studies on the uptake and release of labelled chloramphenico1from sensitive and resistant cells. About 30% of the packed cell volume of both resistant and parental cell lines was found to be freely accessible to external [14C]CAP, under conditions where the binding of label to high affinity saturable sites was diluted out with excess unlabelled CAP. (Specific D-threo-CAP binding sites have been identified in cultured human cells and are the subject of a separate communication, Jeffreys, Siegel & Craig, in preparation). The estimate of 30% for the proportion of cell volume accessible to CAP must be interpreted as a minimum value; the true total cell volume being less than the volume of the packed cell pellet. The suggestion that radioactivity associated with washed cells represents intracellular CAP is supported by the rapid release of the majority of such label by procedures designed to disrupt cell membranes. Resistant and sensitive cells behaved similarly both in terms of accumulation and subsequent release of labelled chloramphenicol. Therefore, it is unlikely that differences in response to the drug result from alterations in permeability at the plasma membrane. Exp Cell Res 102 (IQ76)
Possible alterations to the permeability of the mitochondrial membrane are less easily investigated. Although attempts have been made to disrupt the normal permeability barrier of mitochondria in vitro with Triton X-100 [6, 8, 93 it is not clear to what extent this would assist the entry of chloramphenicol. Procedures which offer a more direct approach, such as studies on the CAP sensitivity of isolated mitoribosomes in catalysing polyuracil-dependent phenylalanine incorporation [28] or assays of peptidy1 transferase activity [29] require large quantities of purified mitoribosomes and have yet to be reported for cultured cells. Another possible mechanism of CAP resistance is that of modification as observed in drug resistant bacteria [21]. Still another detoxification system exists in mammalian liver [25], namely the conjugation of CAP with glucuronic acid. Resistance in the human cell line could possibly have originated by the induction of certain liver specific functions or by the acquisition of chloramphenicol acetyl transferase activity. However, neither CAPR nor parental HeLa cells appeared to modify chloramphenicol by either of these mechanisms. The cross resistance observed to some other antibiotics also argues against the acquisition of resistance by a specific inactivation mechanism. The observed cross-resistance of mitochondrial protein synthesis in CAPR cell lines to mikamycin and carbomycin contrasts with the behaviour of some other CAP resistant cell lines which show little or no cross resistance to these antibiotics (Eisenstadt, personal communication). Mitochondria from the human CAP resistant cell line described by Mitchell et al. [9] showed extensive cross resistance to mikamycin and carbomycin in vitro and mitochondrial protein synthesis in intact cells
Human cell lines resistant to D-threo-chloramphenicol
exhibited very similar increases in resistance to those reported here. There is a precedent in CAP resistant yeast for mutants which show cross resistance to these antibiotics. This suggests that the presumptive CAP target may share a common or overlapping site with mikamytin and carbomycin. Chloramphenicol resistant yeast strains show variable phenotypes. Whereas some mutants were reported to show no cross resistance to other antibiotics tested [30] it is clear from studies on isolated mitoribosomes that other mutants, selected for CAP resistance, can show increased resistance to antibiotics such as erythromycin [3 11.Other, apparently cytoplasmically inherited, CAP resistant mutants in yeast showed a high level of cross resistance to mikamycin and carbomycin [32]. It was thought that this reflected a changed mitochondrial membrane permeability (however, see [31]). There is little information available for other eukaryotes. Nevertheless, cytoplasmically inherited erythromycin resistant mutants of Paramecium were found to be cross resistant to mikamycin [33] and the binding of [14C]CAP to isolated ribosomes from rat liver mitochondria was partly inhibited by carbomycin [28]. The conclusion from the bulk of these studies is that the observation of cross resistance does not exclude the possibility of a mutant expressed at the level of mitochondrial ribosomes. The profile of mitochondrially made proteins of the resistant mutants observed on SDS polyacrylamide gel electrophoresis was identical to that of the parental cell type. Thus gross alteration to, or elimination of, a specific mitochondrially synthesized protein does not appear to be an explanation of the resistant phenotype. However, minor changes and single amino acid substitution would not have
309
been detected. Neither is any information available concerning possible alterations to mitochondrial ribosomal RNAs. In the absence of further data, it is impossible to discriminate between the possibilities of a modification to mitochondrial membranes or to mitochondrial ribosomes, or perhaps to both. However, it should be pointed out that cross resistance is not without specificity; EB (which is assumed to act by blocking mitochondrial transcription) is equally inhibitory to CAP resistant and to CAP sensitive cells. Implicit in the foregoing discussion has been the assumption that the CAP resistance could be controlled by mitDNA. Bunn et al. [7] have previously documented the transfer of the resistant phenotype by cytoplasmic fragments in mouse L cells. Results consistent with the suggestion that the mutants described in this paper are cytoplasmitally controlled will be presented elsewhere. This work was supported in part by grants from NATO, the MRC and the Cancer Research Campaign. L.S. was supported by grants USPHS, GM 1511 and GM 210%. also M.S.T.P. GM02016.
REFERENCES 1. Gillham, NW, Ann rev genet 8 (1974) 347. 2. Lloyd, D, Mitochondria of microorganisms, p. 285. Academic Press, London (1974). 3. Rowlands, R T & Turner, G, Mol gen genet 126 (1973) 201. 4. Beale, G, Knowles, J & Tait, A, Nature 235 (1972) 3%. 5. Roberts, C T & Orias, E, Genetics 73 (1973) 259. 6. Spolsky, C M & Eisenstadt, J M, FEBS lett 25 (1972) 319. 7. Bun& C L, Wallace, D C & Eisenstadt, J M, Proc natl acad sci US 71 (1974) 1681. 8. Wallace, R B & Freeman; K B, J cell bio165 (1975) 492. 9. Mitchell, C H, England, J M & Attardi, G, Som cell genet l(l975) 215. 10. Klietmann, W, Sato, N & Nass, M M K, J cell bio158 (1973) 11. 11. Craig, I W, Proc 2nd int symp genet ind microorganisms, Academic Press, New York (1976). 12. Gordon, S, J cell bio167 (1975) 257.
.ExpCell
Res 102 (1976)
310
Siegel
et al.
13. Clayton, D A, Teplitz, R L, Nabholz, M, Dovey, H & Bodmer, W F, Nature 234 (1971) 560. 14. Attardi, B & Attardi, G, Proc natl acad sci US 69 (1972) 129.
26. Della Bella. D. Ferrari. V. Marca. G & Bonanomi.
L, Biochem pharmacoi 17(1%8) 2381. 27. Bonenhaeen. D & Clavton. D A. J biol them 249
(1974) 7997.’
*
15. David, I B, Horak, I & Coon, H G, Genetics 78
28. Ibrahim, N G, Burke, J P & Beattie, D S, J biol
(1974)459. 16. Craig, I W, Jeffreys, A J, Siegel, R L & Sly, W,
29. Denslow, N D & O’Brien, T W, Biochem biophys
FEBS 10th congress abstracts p. 158(1975). 17. Siegel, R L, PhD Thesis Washington University, St Louis, MO (1976). In preparation. 18. Jeffreys, A & Craig, I, Biochem j 144 (1974) 161. 19. - Biochem sot transact 3 (1975) 398. 20. Laemmh, U K, Nature 227 (1970) 680. 21. Shaw, W V, J biol them 242 (1967) 687. 22. Perlman, S, Abelson, H T & Penman, S, Proc natl acad sci US 70 (1973) 350. 23. Lederman, M & Attardi, G, J mol biol 78 (1973) 275. 24. Haldar, D & Freeman, K B, Can j biochem 46 (1%8) 1009.
25. Glazko, A J, Antimicrobial agents and chemotherapy, Am sot microbial, p. 655. (1%7).
Exp Cell Res 102 (1976)
them 249 (1974) 6806. res commun 57 (1974) 9. 30. Molloy, P L, Howell, N, Plummer, D T, Linnane,
A W & Lukins, H B, Biochem biouhvs res commun 52 (1973) 9. 31. Grivell, L A, Netter, P, Borst, P & Slonimski, P P, Biochim biophys acta 312 (1973) 358. 32. Bunn, C L, Mitchell, C H, Lukins, H B & Linnane, A W, Proc natl acad sci US 67 (1970) 1233. 33. Queiroz, C & Beale, G H, Genet res 23 (1974) L
233.
Received March 1, 1976 Accepted May 6, 1976
-
