Cell killing by various monofunctional alkylating agents in Chinese hamster ovary cells

M~ation Research, 177 (1987) 267-276 Elsevier

267

MTR 04315

Cell killing by various monofunctional alkylating agents in Chinese hamster ovary cells R. G o t h - G o l d s t e i n a n d M. H u g h e s Biomedical Division, Lawrence Berkeley Laboratory, University of California, Berkeley, CA 94720 (U.S.A.) (Received 2 May 1986) (Revision received 18 September 1986) (Accepted 20 October 1986)

Kevwords: Cell Killing; Alkylating agents, monofunctional; (Chinese hamster ovary cells); N-Methyl-N'-nitro-N-nitrosoguanidine; N-Methyl-N-nitrosourea; N-Ethyl-N-nitrosourea; Methyl methanesulphonate.

Summary Cell killing by N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), N-methyl-N-nitrosourea (MNU), Nethyl-N-nitrosourea (ENU), and methyl methanesulfonate (MMS) was measured in Chinese hamster ovary (CHO) cells using the colony-formation assay. Cell killing by these agents was determined in exponentially growing asynchronous cells, in synchronous cells as a function of cell-cycle position and in nondividing cells. Distinct differences in the cytotoxic effect of the 4 alkylating agents were found in respect to dose-response, cell cycle phase-sensitivity and growth state. MNNG and MNU showed the same biphasic dose-survival relationship in exponentially growing cells, with an initial steep decline followed by a shallow component. The shallow component disappeared in growth-arrested cells. MNNG and MNU differed, however, in the cell-cycle age response. No cell-cycle phase difference was seen with MNNG, whereas cells in G 1 seemed more sensitive to MNU than cells in S phase. MMS and ENU both showed shouldered dose-response curves for exponentially growing asynchronous cells, and the same cell-cycle pattern for synchronous cultures with cells in early S phase being the most sensitive. However, survival of nondividing cells versus dividing cells was reduced much more by MMS than by ENU. Caffeine, which interferes with the regulation of DNA synthesis and is known to modify cell killing by DNA-damaging agents, enhanced cell killing by all agents. It is concluded that there must be a number of factors which contribute to cell killing by monofunctional alkylating agents, and that besides alkylation of DNA reaction with other cellular macromolecules should be considered.

Alkylating agents react with nucleophilic sites of various cellular components, macromolecules, as well as smaller molecules (Roberts et al., 1971; Topal and Baker, 1982). Attention has focused Correspondence: Dr. Regine Goth-Goldstein, Biomedical Division, Lawrence Berkeley Laboratory, University of California, Berkeley, CA 94720 (U.S.A.).

primarily on the reaction with DNA, because these alterations are presumed to be responsible for the biological effects of alkylating agent, such as mutations and cell killing. DNA is modified by alkylating agents at a number of sites, and the relative amounts of the different alkylated bases formed are characteristic for a specific alkylating agent (Lawley, 1974; Singer, 1979). Alkyl sulfates

0027-5107/87/$03.50 © 1987 Elsevier Publishers B.V. (Biomedical Division)

268 like MMS react almost exclusively with ring nitrogens of DNA bases, whereas N-nitroso compounds like MNU or ENU react also with exocyclic oxygens of DNA bases, ethylating nitroso compounds more so than methylating ones. The effect of various monofunctional alkylating agents on the reproductive capacity of cells growing in tissue culture has been investigated in a number of studies, most extensively in Chinese hamster V79 cells (Roberts et al., 1971; Peterson et al., 1979; Friedman and Huberman, 1980; Natarajan et al., 1984) and in CHO cells (Couch and Hsie, 1978; Couch et al., 1978; O'Neill et al., 1979; Beranek et al., 1983). These studies show distinct differences in the dose-survival relationship of various alkylating agents, some giving shouldered responses and some giving straight exponential declines. These differences suggest a different cytotoxic action of various agents and might reflect a different capacity of cells to repair certain types of damage. A survival value measured in an asynchronously growing cell population actually representg the average of individual contributions of cells in different cell-cycle phases. Synchronous populations of cells and division-arrested cells allow us to investigate cell-cycle variations in cellular sensitivity to alkylating agents and the role of DNA replication in a cell's sensitivity. This can provide important information on the mechanisms of cell killing by alkylating agents. It has been well established that caffeine potentiates cell killing by alkylating agents in Chinese hamster cells (Roberts and Ward, 1973), though a prerequisite is that the agent reacts with DNA (Goth-Goldstein, 1982). This potentiating effect of caffeine is possibly due to its interference with the regulation of DNA synthesis (Goth-Goldstein and Painter, 1981). In mouse 10T1/2 cells caffeine enhances cell killing by MNNG, but not by MMS (Smith and Grisham, 1983). This suggests different mechanisms of killing by MNNG and MMS, with MMS producing toxicity not through reaction with DNA, but mainly by damaging cell membranes. In Chinese hamster cells the effect of caffeine on MMS-induced killing has not been reported. The studies reported here were initiated to establish cell killing by various alkylating agents

in CHO cells comparing their effect on asynchronous and synchronous cultures of exponentially growing cells and on growth-arrested cultures. The agents selected for comparison in this study were the alkylsulfate MMS and three Nnitroso compounds with either methylating activity, MNNG and MNU, or ethylating activity, ENU. Also, the effect of caffeine on cell killing by these alkylating agents was examined. Materials and methods

Cell culture. Our CHO cell line (CHO-9) is a subclone of the CHO-KK cell line selected for use on the cell-cycle analyzer. Fresh cultures are started from frozen stocks every 3 months to ensure that genetic changes in the cell lines during growth at 37°C are minimal. The cell line has been shown to be free of PPLO, and has a modal chromosome number of 21. The cells were maintained as monolayer cultures in McCoy's 5A medium supplemented with 7.5% fetal calf serum, antibiotics (100 units/ml penicillin, 100 # g / m l streptomycin) and 1 mM HEPES buffer in open plastic tissue culture flasks in a CO2 incubator. Growth arrested cell populations. Nondividing cultures were obtained by the procedure described by O'Neill (1982): 2 × 105 cells were seeded per 25-cm 2 flask. After 18 h cultures were washed with Pucks Saline A, and the growth medium replaced with serum-free medium. This was renewed daily. Survival of cells held up to 10 days under these conditions was unchanged. DNA synthesis as measured by [3H]thymidine incorporation into acid-insoluble cellular material was reduced to 4.9% and 2.8% of exponentially growing control cultures after 3, respectively 4 days in medium without serum. Cells were treated with alkylating agents after 4 days in serum-free medium as described below. After the long incubation of cells in serum-free medium it was impossible to obtain a single cell suspension by treatment with trypsin. A 4-h incubation of cells in medium with 7.5% serum after the treatment with alkylating agents eliminated the problem. Synchronous cell populations. Synchronous cell populations were obtained by selective mitotic detachment using a cell-cycle analyzer apparatus

269 (Talandic Research Corporation, Duarte, CA) as described previously (Goth-Goldstein and Burki, 1980). Mitotic cells were shaken off at 1-h intervals and collected in 75-cm2 flasks to obtain cohorts of cells spaced 1 h in cycle time. 96-98% of cells collected at shake-off were in mitosis. Progress of cells through the cell cycle was monitored in all experiments by cell-volume spectroscopy using Coulter Counter Model ZBI matched to a Coulter Channelyzer, and periodically by flow cytometry, and [3H]thymidine incorporation into acid-precipitable material as described previously (Burki and Aebersold, 1978; Goth-Goldstein and Burki, 1980). The cell-cycle time was determined by cell-volume spectroscopy to be 11-12 h, and the cell-cycle phases were estimated from flow cytometry and [3H]thymidine uptake as G x, 3.5-4 h; S, 6-6.5 h and G 2 plus M, 1.5 h (Fig. 1).

Treatment with alkylating agents and determination of plating efficiency. N-Methyl-N'-nitro-Nnitrosoguanidine ( M N N G ) and N-methyl-Nnitrosourea (MNU) were obtained from Aldrich Chemical Co., Milwaukee, WI, N-ethyl-N-nitrosourea (ENU) from Pfaltz and Bauer, Flushing, NY, and methyl methanesulfonate (MMS) from Eastman Kodak Co., Rochester, NY. M N N G was dissolved in 100 mM acetate buffer, p H 5 and M N U and ENU in dimethyl sulfoxide. Drug treatment was performed in 25-cm2 flasks with about 106 cells or in case of synchronous cells in 75-cm2 flasks with about 2 x 106 cells. Appropriate amounts of the agents were added to the flasks containing serum-free medium, and cells were incubated for 1 h at 37°C. After cells had J 8C ¥ 6C

E E u

4c

i

i

i

T

I

~5c

/

c

~°

~5o

i8 sc

2O

4

1'2

' ~, ; 1'2 Hours efter mitotic detochment

Fig. 1. Cell-cycleprogress analysis of CHO cells after mitotic detachment by (a) cell volume spectroscopy;(b) flow cytometry; (c) uptake of [3H]thyrnidine.

been rinsed thoroughly with saline, they were trypsinized, diluted and counted on a Coulter Counter. Survival was assessed using the colony formation assay. 100-50000 cells were plated per 90-mm tissue-culture dishes, 5 dishes per experimental point. Colonies were stained with 1% methylene blue after 10 days. The plating efficiency of control cells was 75-85%. In the case of asynchronous growing and growth-arrested cultures, 2 - 4 agents were evaluated concurrently; in the case of synchronous cultures, only 1-2 agents were tested at one time. Results

Cell killing in asynchronous cells Two distinct types of cell survival curves were observed for exponentially growing asynchronous C H O ceils exposed to graded concentrations of each of the 4 monofunctional alkylating agents (Fig. 2): (1) with M N N G and MNU, a biphasic curve with little or no shoulder but an initial steep decline to about 2% survival followed by a more shallow decline; (2) with MMS and ENU, a shoulder followed by an exponential decline. The shoulder width of the survival curves varied considerably among the 4 agents. MMS gave a very broad shoulder, whereas M N N G and M N U had hardly any. The shallow component of the biphasic survival curve with M N N G did not seem to result from a resistant subpopulation in our CHO line, because any clone derived from a single cell of the CHO line again gave a biphasic response to M N N G (Goth-Goldstein, in preparation). Other methylating N-nitroso compounds, such as streptozotocin and methylazoxymethanol acetate had the same biphasic dose-survival relationship as M N N G and M N U (data not shown). All other alkylating agents tested, like ethyl methanesulfonate and N-ethyl-N'-nitro-N-nitrosoguanidine, gave the second type of survival curve, the sigmoidal response. When comparing equitoxic concentrations of the 4 alkylating agents, M N N G was the most potent agent and ENU the least effective, a result reported by Couch and Hsie (1978). In such a comparison one should consider the amount of drug decomposed during the exposure to cells. To determine the stability of the alkylating agents

270

100

I

I

I

I MNNG

~"',,.,,

I

I

MIMs

10

"'i i

"6 :>

J

05

> ~oc D u0

"~I

l

_ , ~

o

I

1 l

1.5 r

I

2 l

•

40

I

I

80

I

120 I

MNU

I

160 EINU

o

,i

1C

I

5

I

10

I

15

I

20

I

I

25 30 250 Concentration (pg/ml)

1

750

I

1250

Fig. 2. Cell survival of exponentially growing asynchronous CHO cells after a 1-h treatment with MNNG, MNU, MMS or ENU (normalized to control plating efficiency). Different symbols indicate independent repeat experiments. In a few instances duplicate cultures were utilised indicated by the same symbol at a certain dose.

they were preincubated in medium without serum at 37°C for various lengths of time before exposing cells to them for I h. From the colony-forming ability with and without preincubation of the drug its rate of decomposition could be determined. Thus, the biological half-lives under the experimental conditions were determined as 19 rain for MNU, 15 rain for ENU and 5.5 h for MMS. The half-life of M N N G was 70 min in medium without cells, but it was only 20 rain in medium containing cells at the density used for determination of survival (10 6 cells/25-cm 2 flask), indicating that cells either catalyze the breakdown of M N N G or sequester the agent from the medium. In the case

of nitroso compounds, not the parent compounds themselves, but their decomposition products, primarily the carbonium ions, react with cellular components and are therefore responsible for the biological effects. From the half-lives determined it follows that about 90% of the initial amount of MNU, M N N G and ENU has decomposed during a 1-h exposure and that the effective dose to cells during the exposure interval is essentially identical with the total dose. In the case of MMS, only a fraction of the total MMS reacts with cellular components during a 1-h exposure, so that the effective dose to cells is much smaller than the given total dose.

271

Cell killing in synchronous cells To further characterize the cellular response to alkylating agents, cell killing was measured as a function of the position of cells in the cycle at the time of treatment. Synchronous populations of cells spaced 1 h in cycle time were treated with either of the 4 agents (Fig. 3). Cell survival after MNNG was fairly constant throughout the cell cycle, varying only by a factor of 2. Cell survival after MNU, on the other hand, was low all through the G 1 and eady-S phase, and increased in late S. This pattern was evident also in one experiment using a lower concentration of MNU (5 #g/ml) that reduced average survival to 72% (data not shown). The cell-cycle pattern of killing by MMS and ENU differed from those of MNNG and MNU and were very similar to each other (Fig. 3, also compare Goth-Goldstein and Burki, 1980). Maximum sensitivity was exhibited in the early-S phase with almost an order of magnitude dif-

ference between the most sensitive and least sensitive cell-cycle phase.

Cell killing in growth-arrested cells Another aspect investigated in the process of cell killing by alkylating agents was whether cell killing differed in dividing and division-arrested cells. In initial experiments the growth-arrested state was achieved by letting cultures grow to confluence and then replacing the growth medium with serum-free medium for one day (Jostes, 1981). However, DNA synthesis was reduced only to 25% of an exponentially growing culture by this protocol. In addition, it yielded misleading results in the case of MNNG. Growth-arrested confluent cells were much more resistant to the cytotoxic action of MNNG than exponentially growing cells. The different sensitivity to MNNG was not due, however, to the different state of growth of cells, but merely to the different cell density at the time

100

MMS

MNNG

10

>

U9

100

I

I

I

I

I

1

I

I

t

I

I

I

I

I MNU

I

I

6

8

""

I

I

I

I

I

I

I

IEN U

10

L-

2

4

l

10 12 2 Hours a f t e r m i t o t i c d e t a c h m e n t

l

4

I

6

I

8

I

10

I

12

Fig. 3. Cell survival of synchronous CHO cells after a 1-h treatment with MNNG (0.35 ~g/ml), MNU (18 /~g/ml), MMS ( l l 0 # g / m l ) or ENU O, 400 g g / m l ; A, 700 #g/ml) at various times throughout the cell cycle. The different symbols indicate independent repeat experiments.

272

of MNNG treatment. As already pointed out by Jacobs and DeMars (1978), the killing effect of MNNG is also a function of cell population density. In the case of the other agents (MNU, ENU, MMS), cell survival was independent from cell density during drug treatment. Still, division-arrested cultures were obtained uniformly by keeping cells for 4 days in serum-free medium as described in the Methods section. 3 separate experiments comparing exponentially growing and division-arrested cells were performed with MNU, MMS and ENU. Though the absolute survival at a given concentration of each agent did vary slightly from experiment to experiment, the characteristic difference seen for each agent between exponentially growing and growtharrested cells was very consistent (Fig. 4). With ENU, nondividing cells were slightly more sensitive than dividing cells over the whole dose range tested, leading to a reduction in the shoulder of the survival curve. In the case of MNU, a difference between nondividing and dividing cells became apparent only at survival levels below 1%.

The shallow component of the biphasic curve disappeared in division-arrested cells, so that these cells had virtually an exponential dose-response. The greatest increase in sensitivity of division-arrested cells versus cells in the cycle was found with MMS. The shoulder of the survival curve was reduced and the slope of the curve seemed slightly increased in division-arrested cells. In contrast to the very reproducible results with MNU, ENU and MMS, the dose-response of growth-arrested cells to MNNG was not consistent. In 2 Expts. there appeared to be no difference between exponentially growing and growth-arrested cells (data of one of these experiments is represented by square symbols in Fig. 4), but a response similar to that with MNU (a disappearance of the shallow component of the survival curve in growth-arrested cells) was seen in 5 other experiments (two of which are shown in Fig. 4, circle and triangle symbols). The reason for this inconsistency is unknown. With none of the agents tested were division-arrested cells more resistant than exponentially growing cells, but in all cases,

100

I MNNG

I ENU

MMS

MNU

')

10

u

O'

I

05

I

1

I

15

2

2

J

•

10

j

2O

l,

3O 5O Concentratton (Mg/ml)

f

100

I

150

I

200

1

600

I

1000

Fig. 4. Cell survival in exponentially growing (open symbols) versus division-arrested (closed symbols) CHO cells after a 1-h treatment with MNNG, MNU, MMS or ENU. Different symbols indicate independent repeat experiments. The line was drawn by eye through the circle symbols.

273 100

ENU-

IC

I

2

75 10 20 30 ConcentrQtton Jglrnl)

150

500

1000

Fig. 5. Effect of caffeine (1 mM) on survival of exponentially growing CHO cells treated for 1 h with MNNG, MNU, MMS or ENU. O, no caffeine; O, 1 mM caffeine in post-treatment medium.

some aspect of resistance, shoulder or shallow component, seemed to be reduced in the growtharrested state.

Effect of caffeine We examined how caffeine modulated cell killing by the 4 alkylating agents in CHO cells. Cells were treated with the agents for 1 h and cell survival determined with or without 1 mM caffeine (a non-toxic dose to control cells) in the post-treatment growth medium. We did not attempt to establish extensive dose survival curves, but it is clear from the data presented in Fig. 5 that caffeine enhanced cell killing by all 4 agents. A dose that reduced survival of cells growing in medium without caffeine to 10%, reduced survival of cells growing in medium containing caffeine to 0.3-1%, depending on the agent. Discussion

The intent of the experiments presented here was to assess how the growth state of cells influences the cytotoxic effect of various alkylating agents, which might allow conclusions on the mechanism of killing by these agents. In exponentially growing cells we found two types of cell survival curves. With MMS and ENU, after an initial shoulder region, there was an exponential relationship between the dose of the agent and the fraction of cells surviving, whereas the dose-survival relationship seemed more complex for

MNNG and MNU and involved a two-component curve with little or no shoulder region. Our data for survival of asynchronous CHO cells compare well with reports in the literature. A biphasic survival curve of CHO cells with MNNG as described here has been reported previously (Barranco and Humphrey, 1971; O'Neill et al., 1979). Others have found a broad shoulder in the dose-response with MMS, and virtually no shoulder in the dose-response with MNU (Couch et al., 1978; Couch and Hsie, 1978; Beranek et al., 1983). Couch and Hsie (1978) did not find a shouldered survival curve with ENU (compare Fig. 1 and Goth-Goldstein and Burki, 1980), but in their study reduction of survival was measured only down to 25% and a shoulder could have been missed. The shoulder region of survival curves has been considered evidence of a cell's capacity to accumulate some damage before that damage becomes lethal (Elkind and Sutton, 1959). Our data suggest that CHO cells can accumulate a fair amount of damage caused by MMS, some damage by ENU, but only very little after treatment with MNNG and MNU. This could be either because cells have the ability to repair a certain amount of damage induced by MMS and ENU, but not by MNNG and MNU, or because multiple targets are involved in cell killing by ENU and especially MMS. For example inactivation of proteins could contribute to cell killing by MMS. Smith and Grisham (1983) suggested that MMS damages primarily cell membranes. MMS and ENU gave not only similarly shaped cell-survival curves for exponentially growing asynchronous cells, but also a similar cell-cycle pattern for synchronous cells. With both agents, cell survival was lowest in early-S phase. With MNNG no cell-cycle phase difference was seen, whereas G 1 cells were most sensitive to MNU. Cell killing by MNNG, MNU and ENU was similar in dividing and nondividing populations down to a survival level of 1%. Cell killing by MMS, however, was significantly increased in nondividing compared to dividing cells. This has been reported earlier for V79 cells (Crathorn and Shackleton, 1976). These authors showed that the overall reaction of MMS with DNA, RNA and proteins, and the initial formation of single-strand

274 breaks were comparable for both growth states of cells. One might conclude from this, that cells have recovery processes for MMS-induced damage which are more efficient in cycling cells. In the growth-arrested cultures used in these experiments DNA synthesis was reduced to about 3% of exponentially growing cells (see Method section). This means that progression of cells through the cycle was minimal in these cultures with most cells in an extended G t state (Hahn and Little, 1972). The response of growth-arrested G 1 cells differed from that of G 1 cells obtained by mitotic detachment in that growth-arrested G 1 cells were more resistant to MNU and more sensitive to MMS than G1 cells of cycling cultures. There are two possible explanations for this discrepancy. (1) While the t e r m G 1 has been defined by the DNA content of a cell and is used for growth-arrested and cycling cells, it is unlikely that the G 1 s t a t e of resting and cycling cells is biochemically totally equivalent (for discussion see Hahn and Little, 1972). (2) Cycling cells and growth-arrested cells which are released from this state after treatment with alkylating agents might differ in their regulation of DNA synthesis and inhibition of DNA synthesis following treatment with DNA-damaging agent. Various attempts have been made to correlate the genotoxic effects of alkylating agents on mammalian cells in culture with specific DNA-alkylation products (Newbold et al., 1980; Beranek et al., 1983; Natarajan et al., 1984). In all of these studies a good correlation was found between the level of O6-alkylguanine formed by an agent and its mutagenic efficiency, whereas no single DNAalkylation product related to cell killing. In our study MMS and ENU showed similarities in their cytotoxic activity and differed considerably from the methylating N-nitroso compounds M N N G and MNU. This grouping of the 4 alkylating agents does not correlate with their chemical reactivity, that is, with the spectrum of alkylated bases which they produce in DNA (Beranek et al., 1980). Whereas MMS methylates D N A bases almost exclusively at nitrogens, ENU has a greater propensity than MNU and M N N G to react with exocyclic oxygens of DNA bases. M N N G and MNU break down to the same alkylating species, a methyl carbonium ion, and therefore produce the

same proportions of methylated DNA bases, but they decompose under physiological conditions by somewhat different mechanisms. Whereas MNU breaks down by a base-catalysed reaction, M N N G breakdown can be accelerated by thiols (Lawley and Thatcher, 1970). The discrepancy between the known DNA adduct formation by these 4 alkylating agents and the distinctive survival response seen for each agent indicates that factors other than alkylation of DNA contribute to cell killing by these agents. When comparing the data on cell killing by alkylating agents in CHO cells (this paper) with those for another rodent fibroblast line, the 10T1/2 cells (Smith and Grisham, 1983), one finds many differences. 10T1/2 cells show a shouldered dose-response curve for both MMS and MNNG, are more sensitive to killing by M N N G in S phase versus G 1, and show caffeine-enhanced killing by MNNG, but not by MMS. One important difference between these two cell lines is that whereas CHO cells do not repair O6-methylguanine (Goth-Goldstein, 1980), 10T1/2 cells do, and repair of this lesion is reduced during S phase (Smith et al., 1980). O6-Methylguanine is considered not only a potentially mutagenic, but also a potentially lethal lesion (Scudiero et al., 1984). A cell's capacity to repair this lesion might result in a shouldered survival curve, and a reduced repair during S-phase could explain the increased sensitivity in this cell-cycle phase. The difference in the effect of caffeine on cell killing by MMS in CHO cells and 10T1/2 cells is not readily explained. One can propose a different mechanism of killing by MMS in these two lines or a different action of caffeine. In another murine line, L60T, caffeine does potentiate cell killing by MMS (Walker and Reid, 1971). In CHO cells the enhancement of cell killing by caffeine was the only uniform response to the 4 different alkylating agents. As damage to DNA seems to be a prerequisite for the potentiating effect of caffeine (Goth-Goldstein, 1982), all 4 alkylating agents must mediate their cytotoxic effects in part through reaction with cellular DNA. In conclusion, we found distinct differences in the cytotoxic effect of various alkylating agents in respect to dose-response and the variation with cell-cycle phase and growth state. This indicates that there is not one particular mechanism of cell

275 killing by monofunctional alkylating agents or one p a r t i c u l a r l e t h a l lesion, b u t t h a t t h e r e are several contributing factors which differ from one agent to a n o t h e r . B e s i d e s a l k y l a t i o n of D N A r e a c t i o n with other cellular macromolecules should be considered.

Acknowledgement T h i s w o r k was s u p p o r t e d E S 0 1 9 1 6 a n d ES03603.

by

NIH

grants

References Barranco, S.C., and R.M. Humphrey (1971) The response of Chinese hamster cells to N-methyl-N'-nitro-N-nitrosoguanidine, Mutation Res., 11, 421-429. Beranek, D.T., R.H. Heflich, R.L. Kodell, S.M. Morris and D.A. Casciano (1983) Correlation between specific DNAmethylation products and mutation induction at the HGPRT locus in Chinese hamster ovary cells, Mutation Res., 110, 171-180. Burki, H.J., and P.M. Aebersold (1978) Bromodeoxyuridine induced mutations in synchronous Chinese hamster cells: temporal induction of 6-thioguanine and ouabain resistance during DNA replication, Genetics, 90, 311-321. Couch, D.B., and A.W. Hsie (1978) Mutagenicity and cytotoxicity of congeners of two classes of nitroso compounds in Chinese hamster ovary cells, Mutation Res., 57, 209-216. Couch, D.B., N.L. Forbes and A.W. Hsie (1978) Comparative mutagenicity of alkylsulfate and alkanesulfonate derivatives in Chinese hamster ovary cells, Mutation Res., 57, 217-224. Crathorn, A.R., and J. Shackleton (1976) Repair processes and the response of dividing and non-dividing cells to methyl methanesulfonate and dimethyl sulfate, Chem.-Biol. Interact., 15, 117-130. Elldnd, M.M., and H. Sutton (1959) X-Ray damage and recovery in mammalian cells in culture, Nature (London), 184, 1293-1295. Friedman, J., and E. Huberman (1980) Postreplication repair and the susceptibility of Chinese hamster cells to cytotoxic and mutagenic effects of alkylating agents, Proc. Natl. Acad. Sci. (U.S.A.), 77, 6072-6076. Goth-Goldstein, R. (1980) Inability of Chinese hamster ovary cells to excise O6-alkylguanine, Cancer Res., 40, 2623-2624. Goth-Goldstein, R. (1982) Cell killing by various nitrosoureas and the potentiating effect of caffeine, Mutation Res., 94, 237-244. Goth-Goldstein, R., and J.H. Burki (1980) Ethylnitrosourea-induced mutaganesis in asynchronous and synchronous Chinese hamster ovary cells, Mutation Res., 69, 127-137. Goth-Goldstein, R., and R.B. Painter (1981) Effect of caffeine on cell killing, mutation induction, DNA repair, and DNA synthesis after treatment with ethylnitrosourea, Carcinogenesis, 2, 1267-1271. Hahn, G.M., and J.B. Little (1972) Plateau-phase cultures of

mammalian cells: an in vitro model for human cancer, Curr. Top. Rad. Res. Quart., 8, 39-83. Jacobs, L., and R. DeMars (1978) Quantification of chemical mutagenesis in diploid human fibroblasts: induction of azaguanine-resistant mutants by N-methyl-N'-nitro-Nnitrosoguanidine, Mutation Res., 53, 29-53. Jostes, R.F. (1981) Sister-chromoatid exchanges but not mutations decrease with time in arrested Chinese hamster ovary cells after treatment with ethylnitrosourea, Mutation Res., 91,371-375. Lawley, P.D. (1974) Some chemical aspects of dose-response relationships in alkylating mutagenesis, Mutation Res., 23, 283-295. Lawley, P.D., and C.J. thatcher (1970) Methylation of deoxyribonucleic acid in cultured mammalian cells by Nmethyl-N'-nitro-N-nitrosoguanidine, Biochem. J., 116, 693-707. Natarajan, A.T., J.W.J.M. Simons, E.W. Vogel and A.A. van Zeeland (1984) Relationship between cell killing, chromosomal aberrations, sister-chromatid exchanges and point mutations induced by monofunctional alkylating agents in Chinese hamster cells, A correlation with different ethylation products in DNA, Mutation Res., 128, 31-40. Newbold, R.F., W. Warren, A.S.C. Medcalf and J. Amos (1980) Mutagenicity of carcinogenic methylating agents is associated with a specific DNA modification, Nature (London), 283, 596-599. O'Neill, J.P. (1982) Induction and expression of mutations in mammalian cells in the absence of DNA synthesis and cell division, Mutation Res., 106, 113-122. O'Neill, J.P., R.L. Schenley and A.W. Hsie (1979) Cytotoxicity and mutagenicity of alkylating agents in cultured mammalian cells (CHO/HGPRT system): mutagen treatment in the presence or absence of serum, Mutation Res., 63, 381-385. Peterson, A.R., H. Peterson and C. Heidelberger (1979) Oncogenesis, mutagenesis, DNA damage, and cytotoxicity in cultured mammalian cells treated with alkylating agents, Cancer Res., 39, 131-138. Roberts, J.J., and K.N. Ward (1973) Inhibition of post-replication repair of alkylated DNA by caffeine in Chinese hamster cells but not HeLa cells, Chem.-Biol. Interact., 7, 241-264. Roberts, J.J., J.M. Pascoe, J.E. Plant, J.E. Sturrock and A.R. Crathorn (1971) Quantitative aspects of the repair of alkylated DNA in cultured mammalian cells, I. The effect on HeLa and Chinese hamster cell survival of alkylation of cellular macromolecules, Chem.-Biol. Interact., 3, 29-47. Scudiero, D.A., S.A. Meyer, B.E. Clatterbuck, M.R. Mattern, C.H.J. Ziolkowski and R.S. Day 1II (1984) Sensitivity of human cell strains having different abilities to repair O 6methylguanine in DNA to inactivation by alkylating agents including chloroethylnitrosourea, Cancer Res., 44, 24672474. Singer, B. (1979) N-Nitroso alkylating agents: formation and persistence of alkyl derivatives in mammalian nucleic acids as contributing factors in carcinogenesis, J. Natl. Cancer Inst., 62, 1329-1339. Smith, G.J., and J.W. Grisham (1983) Cytotoxicity of mono-

276 functional alkylating agents - - methyl methanesulfonate and methyl-N'-nitro-N-nitrosoguanidine have different mechanisms of toxicity for 10T1/2 cells, Mutation Res., 111,405-417. Smith, G.J., D,G. Kaufman and J.W. Grisham (1980) Decreased excision of O(6)-methylguanine and N(7)-methylguanine during the S phase in 10T1/2 cells, BBRC, 92, 787-794.

Topal, M.D., and M.S. Baker (1982) DNA precursor pool: a significant target for N-methyl-N-nitrosourea in C3H/ 10T1/2 clone 8 cells, Proc. Natl. Acad. Sci. (U.S.A.), 79, 2211-2215. Walker, J.G., and B.D. Reid (1971) Caffeine potentiation of the lethal action of alkylating agents on L-cells, Mutation Res., 12, 101-104.