377
Mutation Research, 41 (1976) 377--386 © Elsevier/North-Holland Biomedical Press
THE FLUCTUATION TEST AS A MORE SENSITIVE SYSTEM FOR DETERMINING INDUCED MUTATION IN L5178Y MOUSE LYMPHOMA CELLS
J. COLE, C.F. ARLETT and M.H.L. GREEN MRC Cell Mutation Unit, University of Sussex, Falmer, Brighton, BN1 9QG (Great Britain) (Received January 30th, 1976) (Revision received June 29th, 1976) (Accepted'August 2nd, 1976)
Summary Luria-Delbriick fluctuation tests for the determination of the spontaneous m u t a t i o n rate to ouabain resistance in cultured L5178Y mouse l y m p h o m a cells gave values in the range 0.44 to 1.03 X 10 -7 mutants per cell per generation. Addition of very low, non-toxic, levels of EMS (0.1 mM) and MMS (0.012 mM) 30--48 h before plating in selective medium gave a highly significant increase in the number of ouabain-resistant mutants. Methods for the calculation of spontaneous and induced m u t a t i o n rates are discussed and a m e t h o d for the c o m p u t a t i o n of induced m u t a t i o n rates is described. It is suggested t h a t the modified fluctuation test is a simpler and considerably more sensitive assay for mutagens than the conventional experimental design. Some of its limitations are discussed.
Introduction The standard technique for measuring induced mutation frequency in cultured mammalian cells has been to treat a population of cells with mutagen, followed by cloning a sample in selective medium to estimate the number of variants, and in non-selective medium to measure survival. To provide sufficient time for newly-induced m u t a n t s to be expressed a period of growth in non-selective medium must precede the addition of the selective agent. For monolayer cultures this can be achieved by plating the cells in non-selective medium immediately after treatment and adding selective agent to the medium at different expression times [2,6,17,23]. This m e t h o d can, in some cases, result in a loss of m u t a n t s at late expression times, probably as a consequence of metabolic cooperation [20,25] between m u t a n t and n o n - m u t a n t cells. The problem can be overcome by either using cells growing in suspension culture, so that the
378 cells are then plated in selective medium in soft agar [1,7,9] or by maintaining a monolayer culture which is trypsinized and replated [19,21] at different times after treatment. In addition very large populations of cells may have to be maintained and plated every day in order to detect significant increases in mutation frequency after low doses of mutagen. The Luria-Delbrilck fluctuation test [16] has been used to estimate the spontaneous m u t a t i o n rate per cell per generation in bacteria [14,24] and cultured mammalian cells [5,10]. Recently Voogd et al. [26] and Green et al. [12] have shown for bacteria that an adaptation of this test can be used as a very sensitive system for detecting the effects of low levels of mutagens, or weak mutagens. In this report, we show that a similar increase in assay sensitivity is possible using the mouse l y m p h o m a cell line L5178Y in a modified fluctuation test. Materials and methods Cells L5178Y mouse l y m p h o m a cells [11] were grown in suspension culture and cloned in soft agar by the procedure previously described [9], except that samples were set up in Fischer's Medium (Gibco) containing 20% horse serum (Gibco) since this was found to give more consistent growth rates from very low cell density. The standard protocol for the fluctuation test was as follows. An exponentially growing culture of L5178Y cells was diluted to 1--2.5 X 101 cells/ml in Fischer's Medium + 20% horse serum, 200 #g/ml sodium pyruvate (Sigma) and 100 IU/ml penicillin/streptomycin (Flow). Twenty ml aliquots were dispensed into 45 replicate 125-ml flat glass bottles which were gassed with 5% CO2 in air, sealed and incubated at 37°C in a horizontal position. The treatment was applied at 100--120 h when the cell density was ~ 5 X 104/ml (~1 X 106 cells per replicate). At this time aliquots (0.1--0.2 ml) were withdrawn from five of these bottles and counted to determine the exact number of cells per replicate. The bottles were divided into three series of 15 replicates each and were treated as follows. Series A (control) 0.2 ml of phosphate buffered saline (PBS) added to each bottle; series B m e t h y l methanesulphonate (MMS) 0.2 ml of PBS containing MMS to give a final concentration of 0.012 mM was added to each bottle; Series C ethyl methanesulphonate (EMS) 0.2 ml of PBS containing EMS to give a final concentration of 0.1 mM was added to each bottle. The bottles were then gassed, sealed and incubated for a further 30--48 h. At the end of this period of growth aliquots (0.1--0.2 ml) were taken from 3--5 bottles from each series in which the cells had been resuspended by vigorous pipetting and counted to determine the mean number of cells per replicate (~ 5 - 1 0 X 106 cells/ml). Aliquots (0.1 ml) were also withdrawn from 3 replicates in each series, and diluted with 30 ml of non,selective cloning medium to be plated at 15 ml per plate in 9-cm bacterial grade vented petri dishes ( N u n c l o n ) t o determine survival. The contents of every replicate bottle were then diluted to a total of 60 ml in selective cloning medium containing 1 mM ouabain/ml and plated [9] at 15 ml per plate in four 9-cm petri dishes.
379
All plates were incubated at 37°C in a humidified incubator gassed with 5% CO2 in air, for 12 days before being scored.
Calculation o f mutation rates (1) The Po m e t h o d o f Luria and Delbriick [16] and Voogd [26] loge 2(--1OgeP0 ) a =
Nt
Where a = spontaneous m u t a t i o n rate/cell/generation; N t = final number of cells/replicate, and Po = number of replicates with no mutants/total number of replicates. (2) The Median m e t h o d o f Lea and Coulson [15] When nearly all the replicates contain mutants, m e t h o d (1) is n o t applicable. The m u t a t i o n rate can then be calculated from ro, the median of the distribution of r, ~vhere r = n u m b e r of mutants per replicate. Using the table given by Lea and Coulson ro/m, and hence m, the mean number of mutations per replicate can be estimated and a can then be calculated from m
Nt
(3) The Mean m e t h o d o f Capizzi and Jameson [4] Capizzi and Jameson [4] have presented a table for estimation of the spontaneous m u t a t i o n rate of cells in culture from F, the mean number of mutants per replicate in a limited number of samples. CaNt a = CNt T where C = total number of samples, and CaN t is tabulated as a function of CF. (4) Calculation o f induced mutation rates per cell per generation Although for comparative purposes, the three methods of calculating the m u t a t i o n rate/cell/generation outlined above have been applied to the mutagenized replicates, these methods m a y be considered not to be valid for experiments of the type described here. The reason is that the mutagen is present for only part of the period of growth during which mutations can occur. Calculations of m u t a t i o n rate by the Luria-Delbriick formula or the median m e t h o d of Lea and Coulson will give an erroneously low value since they assume that the m u t a t i o n rate will be constant t h r o u g h o u t the experimental period whereas in our experiments it will increase when the mutagen is added. It can be shown that given an infinite number of cultures, mutations arising in each generation contribute equally to the final average number of mutants per culture [16]. Normally it is necessary to correct empirically for the fact that mutations in early generations are rare unlikely to be found in an experiment involving a small number of cultures. However, in the special case considered here, induced mutations will only arise when the mutagen is pres-
380 ent, and will be reasonably frequent throughout this period. If No cells are present per culture when the mutagen is added, Nt cells are present finally pi is the mean number of induced mutants per culture (~ {treated series) -- r (control series)) and a I the induced mutation rate per cell per generation, the number of generations will be log Nt/No log 2 the number of mutants contributed by any one generation will be FI log 2 log Nt/No In the last generation, one m u t a t i o n will contribute one m u t a n t and in this generation Nt/2 cell divisions will occur. Hence the m u t a t i o n rate
ai=
~I2 log 2 N t log (Nt/No)
Although Nt/2 should be corrected by 1/loge2 to allow for divisions in progress, a similar correction to ~, the mean number of mutants per culture is required and the two corrections cancel out. (A similar argument holds for methods (2) and (3), but since m e t h o d (1) is based on the absence of mutants, the correction factor of 1]loge2 must be applied to Nt). It must be stressed t h a t this m e t h o d is only valid in the situation where the mutagen is present for a limited period and is likely to produce mutations t h r o u g h o u t the time t h a t it is present. If the mutagen is present t h r o u g h o u t the period of growth, conventional methods of calculation may be used. In c o m m o n with other calculations of m u t a t i o n rate based on the fluctuation test, the calculation given here does n o t allow for expression time (the time required for DNA damage to cause a genetic change, and for this change to be expressed phenotypically). In the case of ouabain resistance this time has been found to be short [3,9].
Statistical treatment The non-Poisson distribution of the data [16] precludes the use of tests based upon the actual values for numbers of mutants per replicate. The appropriate analysis utilizes the rank ordering of the number of mutants per replicate [18]. Thus in a given comparison the two series to be compared (i.e. 15 control and 15 treated replicates) were ranked in order of ouabaln-resistant colonies scored per replicate, with the lowest rank number being the replicate with the highest number of resistant colonies. The ranks were assigned from 1--30, tied values being giving the mean rank. A table was set up of the rank numbers of the control and treated replicates and the rank total and mean rank number of each series calculated. The significance of the difference between the rank totals was analysed by Students t test, where t = difference of the mean ranks/ S.E. of the difference.
381 Results
Choice of selective agent On the basis of conventional mutagenesis experiments [9] using L5178Y cells, we have considered three selective agents, 1.65 mM thymidine, 0.18 mM 6-thioguanine (TG), and 1 mM ouabain for use in the fluctuation experiments. In fluctuation experiments performed as described here, the cell density in selective medium (1--2 × 10S/ml) was about 10-fold higher than in the conventional experiments and while OUA and TG at the given molarity were completely inhibitory at this cell density the use of thymidine was ruled out because the growth of sensitive wildtype cells was n o t completely inhibited. It was anticipated that use of TG as a selective agent in these fluctuation experiments might present difficulties because of the long time ( 9 6 - - 1 9 2 h) [9,13] required for TG resistance to be expressed in these cells. Thus, during the time course of our experiments, only mutants arising in the first 48 h, when very few cells are present ( - 5 × 103) could be expressed by 144 h. This was confirmed in two experiments when no TG Res colonies were found in any replicate.
TABLE I DESIGN AND RESULT OF A MODIFIED FLUCTUATION MMS A N D E M S Time 0
TEST FOR OUABAIN RESISTANCE WITH
4 5 R e p l i c a t e s at 2 X 102 cells per replicate in 20 ml medium
T i m e at w h i c h m u t a g e n a d d e d ( N 0) = 115 h M e a n n u m b e r o f cells p e r r e p l i c a t e at N O = 1 . 5 9 X 106 T i m e o f p l a t i n g in s e l e c t i v e m e d i u m (N t) = 144 h M e a n n u m b e r o f cells p e r r e p l i c a t e at N t O U A Res colonies per replicate a
Per c e n t survival b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Series A 15 replicates Control
Series B 15 r e p l i c a t e s 0.012 mM MMS
Series C 15 replicates 0.1 m M EMS
9 . 5 5 X 106
7 . 8 8 X 106
8 . 1 5 X 106
6 1 1 2 5 0 1 2 7 0 0 8 0 0 5
10 2 9 5 4 6 3 7 15 4 5 5 7 4 16
13 25 11 22 17 13 20 14 26 28 21 19 26 19 22
100
100
100
M e a n O U A R e s p e r r e p l i c a t e (r-)
2.5
6.8
19.7
Variance
7.1
15.9
26.3
a 4 plates b Sttrvival selective between
per replicate. w a s d e t e r m i n e d b y d i l u t i o n o f 0 . 1 m l t a k e n f r o m 3 r e p l i c a t e s per series a n d p l a t e d in n o n c l o n i n g m e d i u m ( 2 p l a t e s p e r r e p l i c a t e ) . T h e r e w a s n o s i g n i f i c a n t d i f f e r e n c e s in c l o n i n g a b i l i t y t h e series.
382 OUA resistance has a short expression time (maximum from 24--48 h) [3,9]. OUA (1 mM) was completely inhibitory at the high cell concentration and OUA ~es colonies were large and could easily be scored by eye. We have shown that spontaneous and induced OUA Re~ clones of L5178Y [9] have normal growth rates and cloning efficiency, maintain resistance in culture over a period of months, and can be recovered in reconstruction experiments from 1--5 × 105 sensitive cells. OUA resistance was therefore chosen for further study.
Detection of mutagenesis by EMS and MMS using the fluctuation test Table I shows the protocol and results for a typical fluctuation experiment. It can be seen that the number of OUA aes colonies is increased by both MMS and EMS treatment. Table II summarizes the results of four such experiments with m u t a t i o n rates calculated by the methods described above. It can be seen, firstly, that the spontaneous m u t a t i o n rates to OUA resistance per cell per generation were very similar when calculated from four experiments by methods (1) to (3) (in the range 0.44 to 1.03 × 10 -7) and are of a similar order of magnitude to those found by other workers using various cultured mammalian cell lines and selective agents [5,8,10,22]. Secondly, when induced mutation rates were calculated using m e t h o d (4) values were generally higher than with other methods (1)--(3). This was to be expected and in our description of m e t h o d (4) we explain why we believe the higher value to be more realistic. It is clear that both methods (3) and (4) are vulnerable to "jackpots" of spontaneous mutants occurring in either control or treated series. Such a jackpot occurred in the control series Expt. 3 where one replicate contained 42 m u t a n t colonies and thus raised the mean threefold. The presence of the jackpot had little effect on calculations by methods (1) or (2) but did have a substantial effect on the calculations by methods (3) and (4). An additional disadvantage of m e t h o d (3) is that a different value for induced m u t a t i o n rate is obtained depending upon whether the correction for spontaneous mutation is made upon the calculated mutation rates (b in Table II), or the mean number of mutants per replicate (c in Table II). These values ought to be identical. Finally, 0.1 mM EMS and 0.012 mM MMS gave a highly significant increase in the number of OUA Res mutants in every experiment, except for MMS in Expt. 1, where the number of cells treated was about 10-fold lower. Effect of "y irradiation In a large number of experiments using both L5178Y {Cole and Arlett, unpublished observations) and Chinese hamster V79 cells [3], we have never been able to induce mutation to OUA resistance with 7 irradiation. Table III shows a similar negative result in a single fluctuation experiment. In this experiment, the ~-irradiated series received 50 rad per replicate 48 h before plating. The "false" negative for ~,-induced m u t a t i o n confirms the specificity of the ouabain selective system. The mechanism of this specificity m a y reside in the nature of the mutational event which gives rise to the resistant phenotype [3]. Thus, deletion mutations, or operationally deletion mutations, would prove lethal, and ~/mutation may fall into this category. We have been able to show 7induced m u t a t i o n in L5178Y using Tdr and TG as selective agents (Cole and
3.07
(a) (b)
2. M e d i a n m e t h o d of Lea and Coulson
(c)
6115
24.85
105.9
1.0
4.26
1.42 X 107
144
9115
5.28
6.6
0
1.27
7.61 X 106
144
1.03 . .
.
0.63 .
. .
.
.
1.0
. .
0.81
0.79 . .
. .
.
0.98
0.53
0.44
. 0.77
--
0.46
1
1.24
2.48 1.45 1.44
--
1.61 0.98
9/15
3.07
3.5
0
>0.40
1.13
2.18 X 106
168
1.75 X 105
120
200
15
4.11
2.53 1.53 1.78
3.05 2.24
---
0/15
2.34
15.9
5.0
Mutation Research, 41 (1976) 377--386 © Elsevier/North-Holland Biomedical Press
THE FLUCTUATION TEST AS A MORE SENSITIVE SYSTEM FOR DETERMINING INDUCED MUTATION IN L5178Y MOUSE LYMPHOMA CELLS
J. COLE, C.F. ARLETT and M.H.L. GREEN MRC Cell Mutation Unit, University of Sussex, Falmer, Brighton, BN1 9QG (Great Britain) (Received January 30th, 1976) (Revision received June 29th, 1976) (Accepted'August 2nd, 1976)
Summary Luria-Delbriick fluctuation tests for the determination of the spontaneous m u t a t i o n rate to ouabain resistance in cultured L5178Y mouse l y m p h o m a cells gave values in the range 0.44 to 1.03 X 10 -7 mutants per cell per generation. Addition of very low, non-toxic, levels of EMS (0.1 mM) and MMS (0.012 mM) 30--48 h before plating in selective medium gave a highly significant increase in the number of ouabain-resistant mutants. Methods for the calculation of spontaneous and induced m u t a t i o n rates are discussed and a m e t h o d for the c o m p u t a t i o n of induced m u t a t i o n rates is described. It is suggested t h a t the modified fluctuation test is a simpler and considerably more sensitive assay for mutagens than the conventional experimental design. Some of its limitations are discussed.
Introduction The standard technique for measuring induced mutation frequency in cultured mammalian cells has been to treat a population of cells with mutagen, followed by cloning a sample in selective medium to estimate the number of variants, and in non-selective medium to measure survival. To provide sufficient time for newly-induced m u t a n t s to be expressed a period of growth in non-selective medium must precede the addition of the selective agent. For monolayer cultures this can be achieved by plating the cells in non-selective medium immediately after treatment and adding selective agent to the medium at different expression times [2,6,17,23]. This m e t h o d can, in some cases, result in a loss of m u t a n t s at late expression times, probably as a consequence of metabolic cooperation [20,25] between m u t a n t and n o n - m u t a n t cells. The problem can be overcome by either using cells growing in suspension culture, so that the
378 cells are then plated in selective medium in soft agar [1,7,9] or by maintaining a monolayer culture which is trypsinized and replated [19,21] at different times after treatment. In addition very large populations of cells may have to be maintained and plated every day in order to detect significant increases in mutation frequency after low doses of mutagen. The Luria-Delbrilck fluctuation test [16] has been used to estimate the spontaneous m u t a t i o n rate per cell per generation in bacteria [14,24] and cultured mammalian cells [5,10]. Recently Voogd et al. [26] and Green et al. [12] have shown for bacteria that an adaptation of this test can be used as a very sensitive system for detecting the effects of low levels of mutagens, or weak mutagens. In this report, we show that a similar increase in assay sensitivity is possible using the mouse l y m p h o m a cell line L5178Y in a modified fluctuation test. Materials and methods Cells L5178Y mouse l y m p h o m a cells [11] were grown in suspension culture and cloned in soft agar by the procedure previously described [9], except that samples were set up in Fischer's Medium (Gibco) containing 20% horse serum (Gibco) since this was found to give more consistent growth rates from very low cell density. The standard protocol for the fluctuation test was as follows. An exponentially growing culture of L5178Y cells was diluted to 1--2.5 X 101 cells/ml in Fischer's Medium + 20% horse serum, 200 #g/ml sodium pyruvate (Sigma) and 100 IU/ml penicillin/streptomycin (Flow). Twenty ml aliquots were dispensed into 45 replicate 125-ml flat glass bottles which were gassed with 5% CO2 in air, sealed and incubated at 37°C in a horizontal position. The treatment was applied at 100--120 h when the cell density was ~ 5 X 104/ml (~1 X 106 cells per replicate). At this time aliquots (0.1--0.2 ml) were withdrawn from five of these bottles and counted to determine the exact number of cells per replicate. The bottles were divided into three series of 15 replicates each and were treated as follows. Series A (control) 0.2 ml of phosphate buffered saline (PBS) added to each bottle; series B m e t h y l methanesulphonate (MMS) 0.2 ml of PBS containing MMS to give a final concentration of 0.012 mM was added to each bottle; Series C ethyl methanesulphonate (EMS) 0.2 ml of PBS containing EMS to give a final concentration of 0.1 mM was added to each bottle. The bottles were then gassed, sealed and incubated for a further 30--48 h. At the end of this period of growth aliquots (0.1--0.2 ml) were taken from 3--5 bottles from each series in which the cells had been resuspended by vigorous pipetting and counted to determine the mean number of cells per replicate (~ 5 - 1 0 X 106 cells/ml). Aliquots (0.1 ml) were also withdrawn from 3 replicates in each series, and diluted with 30 ml of non,selective cloning medium to be plated at 15 ml per plate in 9-cm bacterial grade vented petri dishes ( N u n c l o n ) t o determine survival. The contents of every replicate bottle were then diluted to a total of 60 ml in selective cloning medium containing 1 mM ouabain/ml and plated [9] at 15 ml per plate in four 9-cm petri dishes.
379
All plates were incubated at 37°C in a humidified incubator gassed with 5% CO2 in air, for 12 days before being scored.
Calculation o f mutation rates (1) The Po m e t h o d o f Luria and Delbriick [16] and Voogd [26] loge 2(--1OgeP0 ) a =
Nt
Where a = spontaneous m u t a t i o n rate/cell/generation; N t = final number of cells/replicate, and Po = number of replicates with no mutants/total number of replicates. (2) The Median m e t h o d o f Lea and Coulson [15] When nearly all the replicates contain mutants, m e t h o d (1) is n o t applicable. The m u t a t i o n rate can then be calculated from ro, the median of the distribution of r, ~vhere r = n u m b e r of mutants per replicate. Using the table given by Lea and Coulson ro/m, and hence m, the mean number of mutations per replicate can be estimated and a can then be calculated from m
Nt
(3) The Mean m e t h o d o f Capizzi and Jameson [4] Capizzi and Jameson [4] have presented a table for estimation of the spontaneous m u t a t i o n rate of cells in culture from F, the mean number of mutants per replicate in a limited number of samples. CaNt a = CNt T where C = total number of samples, and CaN t is tabulated as a function of CF. (4) Calculation o f induced mutation rates per cell per generation Although for comparative purposes, the three methods of calculating the m u t a t i o n rate/cell/generation outlined above have been applied to the mutagenized replicates, these methods m a y be considered not to be valid for experiments of the type described here. The reason is that the mutagen is present for only part of the period of growth during which mutations can occur. Calculations of m u t a t i o n rate by the Luria-Delbriick formula or the median m e t h o d of Lea and Coulson will give an erroneously low value since they assume that the m u t a t i o n rate will be constant t h r o u g h o u t the experimental period whereas in our experiments it will increase when the mutagen is added. It can be shown that given an infinite number of cultures, mutations arising in each generation contribute equally to the final average number of mutants per culture [16]. Normally it is necessary to correct empirically for the fact that mutations in early generations are rare unlikely to be found in an experiment involving a small number of cultures. However, in the special case considered here, induced mutations will only arise when the mutagen is pres-
380 ent, and will be reasonably frequent throughout this period. If No cells are present per culture when the mutagen is added, Nt cells are present finally pi is the mean number of induced mutants per culture (~ {treated series) -- r (control series)) and a I the induced mutation rate per cell per generation, the number of generations will be log Nt/No log 2 the number of mutants contributed by any one generation will be FI log 2 log Nt/No In the last generation, one m u t a t i o n will contribute one m u t a n t and in this generation Nt/2 cell divisions will occur. Hence the m u t a t i o n rate
ai=
~I2 log 2 N t log (Nt/No)
Although Nt/2 should be corrected by 1/loge2 to allow for divisions in progress, a similar correction to ~, the mean number of mutants per culture is required and the two corrections cancel out. (A similar argument holds for methods (2) and (3), but since m e t h o d (1) is based on the absence of mutants, the correction factor of 1]loge2 must be applied to Nt). It must be stressed t h a t this m e t h o d is only valid in the situation where the mutagen is present for a limited period and is likely to produce mutations t h r o u g h o u t the time t h a t it is present. If the mutagen is present t h r o u g h o u t the period of growth, conventional methods of calculation may be used. In c o m m o n with other calculations of m u t a t i o n rate based on the fluctuation test, the calculation given here does n o t allow for expression time (the time required for DNA damage to cause a genetic change, and for this change to be expressed phenotypically). In the case of ouabain resistance this time has been found to be short [3,9].
Statistical treatment The non-Poisson distribution of the data [16] precludes the use of tests based upon the actual values for numbers of mutants per replicate. The appropriate analysis utilizes the rank ordering of the number of mutants per replicate [18]. Thus in a given comparison the two series to be compared (i.e. 15 control and 15 treated replicates) were ranked in order of ouabaln-resistant colonies scored per replicate, with the lowest rank number being the replicate with the highest number of resistant colonies. The ranks were assigned from 1--30, tied values being giving the mean rank. A table was set up of the rank numbers of the control and treated replicates and the rank total and mean rank number of each series calculated. The significance of the difference between the rank totals was analysed by Students t test, where t = difference of the mean ranks/ S.E. of the difference.
381 Results
Choice of selective agent On the basis of conventional mutagenesis experiments [9] using L5178Y cells, we have considered three selective agents, 1.65 mM thymidine, 0.18 mM 6-thioguanine (TG), and 1 mM ouabain for use in the fluctuation experiments. In fluctuation experiments performed as described here, the cell density in selective medium (1--2 × 10S/ml) was about 10-fold higher than in the conventional experiments and while OUA and TG at the given molarity were completely inhibitory at this cell density the use of thymidine was ruled out because the growth of sensitive wildtype cells was n o t completely inhibited. It was anticipated that use of TG as a selective agent in these fluctuation experiments might present difficulties because of the long time ( 9 6 - - 1 9 2 h) [9,13] required for TG resistance to be expressed in these cells. Thus, during the time course of our experiments, only mutants arising in the first 48 h, when very few cells are present ( - 5 × 103) could be expressed by 144 h. This was confirmed in two experiments when no TG Res colonies were found in any replicate.
TABLE I DESIGN AND RESULT OF A MODIFIED FLUCTUATION MMS A N D E M S Time 0
TEST FOR OUABAIN RESISTANCE WITH
4 5 R e p l i c a t e s at 2 X 102 cells per replicate in 20 ml medium
T i m e at w h i c h m u t a g e n a d d e d ( N 0) = 115 h M e a n n u m b e r o f cells p e r r e p l i c a t e at N O = 1 . 5 9 X 106 T i m e o f p l a t i n g in s e l e c t i v e m e d i u m (N t) = 144 h M e a n n u m b e r o f cells p e r r e p l i c a t e at N t O U A Res colonies per replicate a
Per c e n t survival b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Series A 15 replicates Control
Series B 15 r e p l i c a t e s 0.012 mM MMS
Series C 15 replicates 0.1 m M EMS
9 . 5 5 X 106
7 . 8 8 X 106
8 . 1 5 X 106
6 1 1 2 5 0 1 2 7 0 0 8 0 0 5
10 2 9 5 4 6 3 7 15 4 5 5 7 4 16
13 25 11 22 17 13 20 14 26 28 21 19 26 19 22
100
100
100
M e a n O U A R e s p e r r e p l i c a t e (r-)
2.5
6.8
19.7
Variance
7.1
15.9
26.3
a 4 plates b Sttrvival selective between
per replicate. w a s d e t e r m i n e d b y d i l u t i o n o f 0 . 1 m l t a k e n f r o m 3 r e p l i c a t e s per series a n d p l a t e d in n o n c l o n i n g m e d i u m ( 2 p l a t e s p e r r e p l i c a t e ) . T h e r e w a s n o s i g n i f i c a n t d i f f e r e n c e s in c l o n i n g a b i l i t y t h e series.
382 OUA resistance has a short expression time (maximum from 24--48 h) [3,9]. OUA (1 mM) was completely inhibitory at the high cell concentration and OUA ~es colonies were large and could easily be scored by eye. We have shown that spontaneous and induced OUA Re~ clones of L5178Y [9] have normal growth rates and cloning efficiency, maintain resistance in culture over a period of months, and can be recovered in reconstruction experiments from 1--5 × 105 sensitive cells. OUA resistance was therefore chosen for further study.
Detection of mutagenesis by EMS and MMS using the fluctuation test Table I shows the protocol and results for a typical fluctuation experiment. It can be seen that the number of OUA aes colonies is increased by both MMS and EMS treatment. Table II summarizes the results of four such experiments with m u t a t i o n rates calculated by the methods described above. It can be seen, firstly, that the spontaneous m u t a t i o n rates to OUA resistance per cell per generation were very similar when calculated from four experiments by methods (1) to (3) (in the range 0.44 to 1.03 × 10 -7) and are of a similar order of magnitude to those found by other workers using various cultured mammalian cell lines and selective agents [5,8,10,22]. Secondly, when induced mutation rates were calculated using m e t h o d (4) values were generally higher than with other methods (1)--(3). This was to be expected and in our description of m e t h o d (4) we explain why we believe the higher value to be more realistic. It is clear that both methods (3) and (4) are vulnerable to "jackpots" of spontaneous mutants occurring in either control or treated series. Such a jackpot occurred in the control series Expt. 3 where one replicate contained 42 m u t a n t colonies and thus raised the mean threefold. The presence of the jackpot had little effect on calculations by methods (1) or (2) but did have a substantial effect on the calculations by methods (3) and (4). An additional disadvantage of m e t h o d (3) is that a different value for induced m u t a t i o n rate is obtained depending upon whether the correction for spontaneous mutation is made upon the calculated mutation rates (b in Table II), or the mean number of mutants per replicate (c in Table II). These values ought to be identical. Finally, 0.1 mM EMS and 0.012 mM MMS gave a highly significant increase in the number of OUA Res mutants in every experiment, except for MMS in Expt. 1, where the number of cells treated was about 10-fold lower. Effect of "y irradiation In a large number of experiments using both L5178Y {Cole and Arlett, unpublished observations) and Chinese hamster V79 cells [3], we have never been able to induce mutation to OUA resistance with 7 irradiation. Table III shows a similar negative result in a single fluctuation experiment. In this experiment, the ~-irradiated series received 50 rad per replicate 48 h before plating. The "false" negative for ~,-induced m u t a t i o n confirms the specificity of the ouabain selective system. The mechanism of this specificity m a y reside in the nature of the mutational event which gives rise to the resistant phenotype [3]. Thus, deletion mutations, or operationally deletion mutations, would prove lethal, and ~/mutation may fall into this category. We have been able to show 7induced m u t a t i o n in L5178Y using Tdr and TG as selective agents (Cole and
3.07
(a) (b)
2. M e d i a n m e t h o d of Lea and Coulson
(c)
6115
24.85
105.9
1.0
4.26
1.42 X 107
144
9115
5.28
6.6
0
1.27
7.61 X 106
144
1.03 . .
.
0.63 .
. .
.
.
1.0
. .
0.81
0.79 . .
. .
.
0.98
0.53
0.44
. 0.77
--
0.46
1
1.24
2.48 1.45 1.44
--
1.61 0.98
9/15
3.07
3.5
0
>0.40
1.13
2.18 X 106
168
1.75 X 105
120
200
15
4.11
2.53 1.53 1.78
3.05 2.24
---
0/15
2.34
15.9
5.0
