Br. J. Cancer (1978) 37, Suppl. III, 54
CYTOTOXICITY OF MISONIDAZOLE AND DNA DAMAGE IN HYPOXIC MAMMALIAN CELLS B. PALCIC AND L. D. SKARSGARD
Fromn the B.C. Cancer Foundation, 2656 Heather Street, Vancouver, B.C. V5Z 3J3, Canada, and the University of British Columbia, Vancouver, B.C. V6T 1 JJ'5, Canada
Summary.-The loss of colony-forming ability and the yield of DNA single strand breaks were studied following exposure of various mammalian cell types to misonidazole. A correlation was observed between cell inactivation and DNA damage. If, after exposure, the cells were washed free of the drug and then incubated at 37°C in growth medium, it was observed that surviving cells repaired their DNA to the point that sedimentation profiles were identical to those of unexposed cells. In nonsurviving cells, however, the DNA was found to be further degraded with postexposure incubation.
AMONG the many nitroheterocyclic compounds which have been shown to be effective radiosensitizers of hypoxic mammalian cells, metronidazole and misonidazole have received the greatest attention. These agents also possess a selective toxicity for hypoxic cells, a property which may complement their action against hypoxic tumour cells. The hypoxic cytotoxicity of misonidazole is a function of concentration of the drug (Moore, Palcic and Skarsgard 1976, Hall and Roizin-Towle, 1975). Cytotoxicity is much more pronounced at higher temperatures; at 0°C the drug is non-toxic (Moore et al. 1976; Stratford and Adams, 1977; Hall and Biaglow, 1977). The cytotoxic effects are also dependent on the cell line (Taylor and Rauth, 1977). The drug freely enters both aerobic and hypoxic mammalian cells. In hypoxic (though not aerobic) cells, misonidazole is metabolized and at least two breakdown products have been isolated to date (Varghese, Gulyas and Mohindra, 1 976; Wong and Whitmore, 1977). Misonidazole and metronidazole have been shown to induce DNA single strand breaks in mammalian cells. Breaks occur only in hypoxic cells, and the yield is dependent upon the concentration of the
drug and the temperature (Palcic and Skarsgard, unpublished work). This paper describes a correlation between DNA damage and cell killing in the cytotoxic effect of misonidazole. We also show that DNA repair is effective in surviving cells, while non-surviving cells degrade their DNA. A model is proposed to explain cytotoxic cell inactivation in terms of DNA damage. METHODS
Two lines of Chinese hamster cells were used in these studies, CH2B2 and CHO. These were cultured according to standard procedures (Moore et al., 1976), CH2B2 as monolayers in MEM F-16 medium and CHO as suspension cultures in of medium, both supplemented with 10% fetal calf serum. Human skin fibroblasts derived from skin biopsy of a normal male were cultured as monolayers in MEM F-15 supplemented with 1500 FCS. These human cells were designated the AN strain and all experiments were performed using cultures in the 10th to 18th passage. For each experiment, log phase cells were harvested either by trypsinization (CH2B2, AN) or by centrifugation (CHO). The cells were then washed once in growth medium and resuspended at 4 x 106 cells/ml in growth medium. After holding them at this
55t
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS
concentration for 10-15 min at room temperature to deplete the oxygen, they were added to pre-gassed glass vessels containing growth medium and a prescribed amount of misonidazole, such that the final concentration of cells was -2 x 105 cells/ml. Misonidazole was freshly prepared in growth medium containing 10% fetal calf serum and 19 ml was loaded into the glass vessels. Purified N2 (containing less than 5 ppm 02) or 02 was flowed over the stirred solutions for 15 min before the addition of 1 ml of concentrated cells. The time of addition of cells was defined as the starting time of exposure. The flow of the respective gasses continued throughout the exposure time. At various times 1 ml samples were withdrawn and immediately diluted 10 times in aerobic growth medium at 0°C. The cells were th'n washed free of misonidazole and either plated for colony formation or used for analysis of DNA breaks. For molecular studies the cell DNA was uniformly labelled with either 14C-TdR (0.05 ,uCi/ml) or 3H-TdR (0-25 ,uCi/ml) by adding labelled nucleotides to the growth medium 24 h before collection of cells. Alkaline sucrose gradients were used to determine the molecular weight of the DNA molecules. This method allows one to calculate the number of DNA single strand breaks induced by a given treatment. The procedure for cell lysis, centrifugation, collection of gradients and analysis of data will be described elsewhere (see Palcic and Skarsgard, 1972).
z
0
() Z >
0t
z
C,,
0
1
2
3
TIME h
FIG. 1. Hypoxic cytotoxicity of mnisonidazole to different cell types. Cells were exposed for various times to 15 mm misonidazole under hypoxic conditions at 37°C in growth medium. Cells were then plated and assayed for colony-forming ability.
dependent on drug concentration and on temperature. It probably reflects the time required for metabolic conversion of the drug and accumulation of lethal damage. RESULTS Colonies of surviving cells were of the The survival response of the three dif- same size as in control samples regardless ferent cell types after hypoxic exposure to of where in the survival response they misonidazole is shown in Fig. 1. CH2B2 were obtained. In contrast, for example, cells were found to be the most sensitive when these cells are exposed to y-rays, and AN human cells most resistant to drug many of the colonies emerging from irratreatment. Responses were somewhat vari- diated cells are much smaller than those able from experiment to experiment, par- of unirradiated cells. ticularly for CH2B2 cells. The yield of DNA single strand breaks Characteristically, the survival curves is presented in Fig. 2. It can be seen that were flat (zero slope) for short exposure the cells showing the more sensitive surtimes. This was not due to the presence of vival response also show a higher yield of molecular oxygen. Much longer pregassing- breaks; thus one finds a good correlation of the cells and of the medium containing between survival and DNA breaks. Note misonidazole before adding the cells to the that induction of DNA breaks begins immedium did not alter the survival response. mediately after the addition of misonidaThe duration of the zero slope portion was zole to the cells.
056
B. PALCIC AND L. D. SKARSGARD 15mM
ments of cell survival and DNA damage and repair were made on cells exposed to misonidazole for various times. At the prescribed time a sample of cells was withdrawn, washed and split into three equal parts. One part was used to determine the
MISONIDAZOLE-
"°Co -Trays; 0 C, O,
AN
rLjt) 0
\)I
I
0
1
2
3
4
5
#-.a
IIC
CHO
6
EMO&M
7
TIME h Fi(o. 2. DNA breaks induced by misonidazole in different cell types. Cells were expo0sed to misonidazole as described in Fig. 1, but were then lysed on the top of alkaline sucrose gradients. The samples were then centrifuged and DNA weight average molecular weights were determined. The ratio of molecular weights of control and exposed cells (M C /ME) was then plotted against the time of exposure. The cell types which displayecl greater sensitivity in survival studies also demonstrated greater DNA damage.
It has been well established that yirradiated mammalian cells effectively repair DNA single strand breaks (e.g. Elkind and Kamper, 1970). An example of this is shown in the upper portion of Fig. 3. When CH2B2 cells were washed after hypoxic exposure to misonidazole and then incubated in growth medium at 37°C (950o air, 5%o C02) no DNA repair was detected in cells that were exposed to the drug for extended times. In fact, the DNA was even further degraded with postexposure incubation times (Fig. 3, bottom graph). This is very different from the situation following y-irradiation in which, even when the initial damage is more extensive, repair of DNA breaks is very effective (upper graph, Fig. 3). For very short hypoxic exposures to misonidazole, however, post-exposure incubation allowed the cells to repair their damaged DNA; no additional breaks were detected in this instance. In Fig. 4 are shown the results of an experiment in which parallel measure-
0
N.F
DOSE k rad 15mM MISONIDAZOLE;N2,370C
CH12B, cells 20
90min repair, 370C c0
n,rp/
F
0
-no
1
2
repair
3
TIME h FIG. 3.- Repair of DNA breaks induced by y-rays or misonidazole. While most of the y-induced DNA breaks were repaired during a 90 min incubation period at 37°C after irradiation (upper diagram), the DNA from misonidazole-treated cells is even further degraded with post-exposure incubation, if cells were exposed to 15 mm misonidazole for a period of 1 h or more.
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS
57
15mM MISONIDAZOLE; 370C, N2 CH2B, cells
z
1'
c,
Zid'
>; x
1U
. ~
0
d ~~~
30
I
TIME min
I
60
90
to z 0
C-) -J
;I0
z
r
C-)
FRACTION NUMBER
FRACTION NUMBER
DIRECTION OF SEDIMENTATION Fio(. 4. DNA repair of misonidazole treated cells. 14C-TdR-labelled CH2B2 cells were assayed for colony forming ability, number of DNA breaks and capacity to repair breaks after hypoxic exposure to 15 mm misonidazole. A 24 h repair interval was used. For drug exposure times short enough (30 min or less) that essentially all cells survived (S = 0 * 93) all cells also restored their DNA to its original size after 24 h repair (a). After 60 min exposure the fraction of surviving cells was 0 -15; for this exposure time only about 14% of the total DNA returns to the control size after 24 h repair (see text). This corresponds closely to the observed survival value. DIRECTION OF SEDIMENTATION --
surviving fraction, the second to determine the number of DNA breaks present immediately after exposure to misonidazole and the third was incubated for a further 24 h in growth medium free of the drug, after which time the size of DNA molecules was examined. The survival response in this particular experiment was thus measured with 14C-TdR labelled CH2B2 cells, and the cells were found to be more sensitive than is generally the case with unlabelled cells. For cells exposed to 15 mm misonidazole
for only 30 min the surviving fraction is 0-93 (upper graph). Although the DNA from these cells shows substantial damage when assayed immediately, after 24 h repair essentially all of the DNA has been restored to the control size (Graph (a)). After 60 min exposure to misonidazole cell survival has dropped to 0-15, and in this case the DNA is substantially degraded during the 24 h incubation period (Graph (b). Furthermore, 5200 of the counts are lost from this sample, apparently due to lysis of lethally damaged cells. Thus, al-
58
B. PALCIC AND L. D. SKARSGARD
though 28% of all counts that remain in this profile appear at the control position, this repaired DNA represents only 14%, approximately, of the total DNA originally in this sample. This analysis was confirmed in a separate experiment where all of the DNA (from cells and from supernatant) was recovered and placed on the gradient. In this case 12% of the total DNA returned to the control size after 24 h incubation following a 60 min exposure to 15 mM misonidazole. There is therefore reasonable quantitative agreement between the proportion of cells which survive (15%) and the proportion of DNA which is repaired (12-14%) following this treatment. From these and similar experiments it was concluded that cells which did not survive in terms of colony-forming ability underwent DNA degradation during the 24 h after exposure to the drug. Cells that did survive completely restored their DNA. DISCUSSION
duced. A reduced product binds to or associates with DNA. Repair enzymes then recognize the lesion and cleave the DNA, the damaged site is excised and further repair follows steps similar to those operating for UV damage repair. As long as there are only a few damaged sites in the DNA, cells are capable of repairing it effectively. With more extensive damage, however, close proximity of damaged sites may lead to overlap of excision regions and complementary re-synthesis fails. This then leads to DNA degradation and cell death. The close parallel between DNA breaks and cell inactivation does not, of course, represent proof of the validity of the model. The model may be useful only in that it suggests further experiments which may indicate whether or not DNA in fact is the primary target for the cytotoxic action of this drug. For example, is isolated or viral DNA damaged by misonidazole or its reduced derivatives? Are surviving cells repairing DNA damage by unscheduled DNA synthesis? Are mutant cells, defective in various forms of DNA repair, more sensitive to drug exposure than normal cells? In some preliminary experiments to test the model, XP cells defective in repair of UV-damaged DNA were tested. Of the two types of XP cells which were used (Complimentation Groups A and D), neither showed a more sensitive response than normal AN cells, suggesting that cells with this repair defect, at least, are not predisposed to this damage. We have recently observed that, within a few hours after a brief hypoxic exposure to misonidazole, cells which were ultimately destined to survive could be distinguished from those which did not survive. This may allow a more selective analysis of the role, if any, of DNA damage and repair in the cytotoxic effects of misonidazole.
This study shows a correlation between cell inactivation and the DNA damage which follows cytotoxic exposure of mammalian cells to misonidazole. For both of these endpoints, the damage is much greater in hypoxic cells, it increases with drug concentration and with time of exposure to misonidazole and it is temperature-dependent; exposure of cells to the drug at 0°C does not appear to induce DNA strand breaks (Palcic and Skarsgard, unpublished work). Furthermore, the yield of DNA breaks is higher in cells which display an increased sensitivity in survival studies. The cells that will not survive show an even greater number of DNA breaks with further incubation at 37TC after removal of the drug. One could speculate that DNA breaks are not induced directly but that damaged DNA is enzymatically cleaved by repair enzyme(s). If it is assumed that the target for the hypoxic cytotoxic effect We are grateful to Dr C. E. Smithen who proof misonidazole is DNA, then the following vided the misonidazole used in these experiments model may be proposed to explain the re- and to I. Harrison for capable technical assistance. work was supported by the National Cancer sults reported in this study: the drug freely This Institute of Canada and the British Columbia enters cells where it is metabolically re- Cancer Foundation.
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS REFERENCES ELKIND, M. M. & KAMPER, C. (1970) Two Forms of Repair of DNA in Mammalian Cells Following Irradiation. Biophys. J., 10, 237. HALL, E. J. & BIAGLOW, J. (1977) Ro-07-0582 as a Radiosensitizer and Cytotoxic Agent. Int. J. Radiat. Oncol. Biol. Phys., 2, 521. HALL, E. J. & RoIzIN-TowLE, L. (1975) Hypoxic Sensitizers: Radiobiological Studies at the Cellular Level. Radiology, 117, 453. MOORE, B. A., PALcic, B. & SKARSGARD, L. D. (1976) Radiosensitizing and Toxic Effects of the 2-Nitroimidazole Ro-07-0582 in Hypoxic Mammalian Cells. Radiat. Res., 67, 459. PALCIC, B. & SKARSGARD, L. D. (1972) The Effect of Oxygen on DNA Single-strand Breaks Produced by Ionizing Radiation in Mammalian Cells. Int. J. Radiat. Biol., 21, 417.
59
STRATFORD, I. J. & ADAMS, G. E. (1977) Effect of Hyperthermia on Differential Cytotoxicity of a Hypoxic Cell Radiosensitizer, Ro-07-0582, on Mammalian Cells in vitro. Brit. J. Cancer, 35, 307. TAYLOR, Y. & RAUTH, A. M. (1977) A Comparison of the Hypoxic Cell Specific Toxicity of Ro-070582 Towards HeLa and CHO Cells. (Abstract). Radiat. Res., 70, 702. VARGHESE, A. J., GULYAS, S. & MOHINDRA, J. K. (1976) Hypoxia-Dependent Reduction of 1-(2by Nitro-1-imidazolyl)-3-methoxy-2-propanol Chinese Hamnster Ovary Cells and KHT Tumor Cells in vitro and in vivo. Cancer Res., 36, 3761. WONG, T. W. & WHITMORE, G. F. (1977) A Comparison of Radiation-Sensitizing Ability and Cell Uptake for NDPP and Ro-07-0582. Radiat. Res., 71, 132.
CYTOTOXICITY OF MISONIDAZOLE AND DNA DAMAGE IN HYPOXIC MAMMALIAN CELLS B. PALCIC AND L. D. SKARSGARD
Fromn the B.C. Cancer Foundation, 2656 Heather Street, Vancouver, B.C. V5Z 3J3, Canada, and the University of British Columbia, Vancouver, B.C. V6T 1 JJ'5, Canada
Summary.-The loss of colony-forming ability and the yield of DNA single strand breaks were studied following exposure of various mammalian cell types to misonidazole. A correlation was observed between cell inactivation and DNA damage. If, after exposure, the cells were washed free of the drug and then incubated at 37°C in growth medium, it was observed that surviving cells repaired their DNA to the point that sedimentation profiles were identical to those of unexposed cells. In nonsurviving cells, however, the DNA was found to be further degraded with postexposure incubation.
AMONG the many nitroheterocyclic compounds which have been shown to be effective radiosensitizers of hypoxic mammalian cells, metronidazole and misonidazole have received the greatest attention. These agents also possess a selective toxicity for hypoxic cells, a property which may complement their action against hypoxic tumour cells. The hypoxic cytotoxicity of misonidazole is a function of concentration of the drug (Moore, Palcic and Skarsgard 1976, Hall and Roizin-Towle, 1975). Cytotoxicity is much more pronounced at higher temperatures; at 0°C the drug is non-toxic (Moore et al. 1976; Stratford and Adams, 1977; Hall and Biaglow, 1977). The cytotoxic effects are also dependent on the cell line (Taylor and Rauth, 1977). The drug freely enters both aerobic and hypoxic mammalian cells. In hypoxic (though not aerobic) cells, misonidazole is metabolized and at least two breakdown products have been isolated to date (Varghese, Gulyas and Mohindra, 1 976; Wong and Whitmore, 1977). Misonidazole and metronidazole have been shown to induce DNA single strand breaks in mammalian cells. Breaks occur only in hypoxic cells, and the yield is dependent upon the concentration of the
drug and the temperature (Palcic and Skarsgard, unpublished work). This paper describes a correlation between DNA damage and cell killing in the cytotoxic effect of misonidazole. We also show that DNA repair is effective in surviving cells, while non-surviving cells degrade their DNA. A model is proposed to explain cytotoxic cell inactivation in terms of DNA damage. METHODS
Two lines of Chinese hamster cells were used in these studies, CH2B2 and CHO. These were cultured according to standard procedures (Moore et al., 1976), CH2B2 as monolayers in MEM F-16 medium and CHO as suspension cultures in of medium, both supplemented with 10% fetal calf serum. Human skin fibroblasts derived from skin biopsy of a normal male were cultured as monolayers in MEM F-15 supplemented with 1500 FCS. These human cells were designated the AN strain and all experiments were performed using cultures in the 10th to 18th passage. For each experiment, log phase cells were harvested either by trypsinization (CH2B2, AN) or by centrifugation (CHO). The cells were then washed once in growth medium and resuspended at 4 x 106 cells/ml in growth medium. After holding them at this
55t
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS
concentration for 10-15 min at room temperature to deplete the oxygen, they were added to pre-gassed glass vessels containing growth medium and a prescribed amount of misonidazole, such that the final concentration of cells was -2 x 105 cells/ml. Misonidazole was freshly prepared in growth medium containing 10% fetal calf serum and 19 ml was loaded into the glass vessels. Purified N2 (containing less than 5 ppm 02) or 02 was flowed over the stirred solutions for 15 min before the addition of 1 ml of concentrated cells. The time of addition of cells was defined as the starting time of exposure. The flow of the respective gasses continued throughout the exposure time. At various times 1 ml samples were withdrawn and immediately diluted 10 times in aerobic growth medium at 0°C. The cells were th'n washed free of misonidazole and either plated for colony formation or used for analysis of DNA breaks. For molecular studies the cell DNA was uniformly labelled with either 14C-TdR (0.05 ,uCi/ml) or 3H-TdR (0-25 ,uCi/ml) by adding labelled nucleotides to the growth medium 24 h before collection of cells. Alkaline sucrose gradients were used to determine the molecular weight of the DNA molecules. This method allows one to calculate the number of DNA single strand breaks induced by a given treatment. The procedure for cell lysis, centrifugation, collection of gradients and analysis of data will be described elsewhere (see Palcic and Skarsgard, 1972).
z
0
() Z >
0t
z
C,,
0
1
2
3
TIME h
FIG. 1. Hypoxic cytotoxicity of mnisonidazole to different cell types. Cells were exposed for various times to 15 mm misonidazole under hypoxic conditions at 37°C in growth medium. Cells were then plated and assayed for colony-forming ability.
dependent on drug concentration and on temperature. It probably reflects the time required for metabolic conversion of the drug and accumulation of lethal damage. RESULTS Colonies of surviving cells were of the The survival response of the three dif- same size as in control samples regardless ferent cell types after hypoxic exposure to of where in the survival response they misonidazole is shown in Fig. 1. CH2B2 were obtained. In contrast, for example, cells were found to be the most sensitive when these cells are exposed to y-rays, and AN human cells most resistant to drug many of the colonies emerging from irratreatment. Responses were somewhat vari- diated cells are much smaller than those able from experiment to experiment, par- of unirradiated cells. ticularly for CH2B2 cells. The yield of DNA single strand breaks Characteristically, the survival curves is presented in Fig. 2. It can be seen that were flat (zero slope) for short exposure the cells showing the more sensitive surtimes. This was not due to the presence of vival response also show a higher yield of molecular oxygen. Much longer pregassing- breaks; thus one finds a good correlation of the cells and of the medium containing between survival and DNA breaks. Note misonidazole before adding the cells to the that induction of DNA breaks begins immedium did not alter the survival response. mediately after the addition of misonidaThe duration of the zero slope portion was zole to the cells.
056
B. PALCIC AND L. D. SKARSGARD 15mM
ments of cell survival and DNA damage and repair were made on cells exposed to misonidazole for various times. At the prescribed time a sample of cells was withdrawn, washed and split into three equal parts. One part was used to determine the
MISONIDAZOLE-
"°Co -Trays; 0 C, O,
AN
rLjt) 0
\)I
I
0
1
2
3
4
5
#-.a
IIC
CHO
6
EMO&M
7
TIME h Fi(o. 2. DNA breaks induced by misonidazole in different cell types. Cells were expo0sed to misonidazole as described in Fig. 1, but were then lysed on the top of alkaline sucrose gradients. The samples were then centrifuged and DNA weight average molecular weights were determined. The ratio of molecular weights of control and exposed cells (M C /ME) was then plotted against the time of exposure. The cell types which displayecl greater sensitivity in survival studies also demonstrated greater DNA damage.
It has been well established that yirradiated mammalian cells effectively repair DNA single strand breaks (e.g. Elkind and Kamper, 1970). An example of this is shown in the upper portion of Fig. 3. When CH2B2 cells were washed after hypoxic exposure to misonidazole and then incubated in growth medium at 37°C (950o air, 5%o C02) no DNA repair was detected in cells that were exposed to the drug for extended times. In fact, the DNA was even further degraded with postexposure incubation times (Fig. 3, bottom graph). This is very different from the situation following y-irradiation in which, even when the initial damage is more extensive, repair of DNA breaks is very effective (upper graph, Fig. 3). For very short hypoxic exposures to misonidazole, however, post-exposure incubation allowed the cells to repair their damaged DNA; no additional breaks were detected in this instance. In Fig. 4 are shown the results of an experiment in which parallel measure-
0
N.F
DOSE k rad 15mM MISONIDAZOLE;N2,370C
CH12B, cells 20
90min repair, 370C c0
n,rp/
F
0
-no
1
2
repair
3
TIME h FIG. 3.- Repair of DNA breaks induced by y-rays or misonidazole. While most of the y-induced DNA breaks were repaired during a 90 min incubation period at 37°C after irradiation (upper diagram), the DNA from misonidazole-treated cells is even further degraded with post-exposure incubation, if cells were exposed to 15 mm misonidazole for a period of 1 h or more.
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS
57
15mM MISONIDAZOLE; 370C, N2 CH2B, cells
z
1'
c,
Zid'
>; x
1U
. ~
0
d ~~~
30
I
TIME min
I
60
90
to z 0
C-) -J
;I0
z
r
C-)
FRACTION NUMBER
FRACTION NUMBER
DIRECTION OF SEDIMENTATION Fio(. 4. DNA repair of misonidazole treated cells. 14C-TdR-labelled CH2B2 cells were assayed for colony forming ability, number of DNA breaks and capacity to repair breaks after hypoxic exposure to 15 mm misonidazole. A 24 h repair interval was used. For drug exposure times short enough (30 min or less) that essentially all cells survived (S = 0 * 93) all cells also restored their DNA to its original size after 24 h repair (a). After 60 min exposure the fraction of surviving cells was 0 -15; for this exposure time only about 14% of the total DNA returns to the control size after 24 h repair (see text). This corresponds closely to the observed survival value. DIRECTION OF SEDIMENTATION --
surviving fraction, the second to determine the number of DNA breaks present immediately after exposure to misonidazole and the third was incubated for a further 24 h in growth medium free of the drug, after which time the size of DNA molecules was examined. The survival response in this particular experiment was thus measured with 14C-TdR labelled CH2B2 cells, and the cells were found to be more sensitive than is generally the case with unlabelled cells. For cells exposed to 15 mm misonidazole
for only 30 min the surviving fraction is 0-93 (upper graph). Although the DNA from these cells shows substantial damage when assayed immediately, after 24 h repair essentially all of the DNA has been restored to the control size (Graph (a)). After 60 min exposure to misonidazole cell survival has dropped to 0-15, and in this case the DNA is substantially degraded during the 24 h incubation period (Graph (b). Furthermore, 5200 of the counts are lost from this sample, apparently due to lysis of lethally damaged cells. Thus, al-
58
B. PALCIC AND L. D. SKARSGARD
though 28% of all counts that remain in this profile appear at the control position, this repaired DNA represents only 14%, approximately, of the total DNA originally in this sample. This analysis was confirmed in a separate experiment where all of the DNA (from cells and from supernatant) was recovered and placed on the gradient. In this case 12% of the total DNA returned to the control size after 24 h incubation following a 60 min exposure to 15 mM misonidazole. There is therefore reasonable quantitative agreement between the proportion of cells which survive (15%) and the proportion of DNA which is repaired (12-14%) following this treatment. From these and similar experiments it was concluded that cells which did not survive in terms of colony-forming ability underwent DNA degradation during the 24 h after exposure to the drug. Cells that did survive completely restored their DNA. DISCUSSION
duced. A reduced product binds to or associates with DNA. Repair enzymes then recognize the lesion and cleave the DNA, the damaged site is excised and further repair follows steps similar to those operating for UV damage repair. As long as there are only a few damaged sites in the DNA, cells are capable of repairing it effectively. With more extensive damage, however, close proximity of damaged sites may lead to overlap of excision regions and complementary re-synthesis fails. This then leads to DNA degradation and cell death. The close parallel between DNA breaks and cell inactivation does not, of course, represent proof of the validity of the model. The model may be useful only in that it suggests further experiments which may indicate whether or not DNA in fact is the primary target for the cytotoxic action of this drug. For example, is isolated or viral DNA damaged by misonidazole or its reduced derivatives? Are surviving cells repairing DNA damage by unscheduled DNA synthesis? Are mutant cells, defective in various forms of DNA repair, more sensitive to drug exposure than normal cells? In some preliminary experiments to test the model, XP cells defective in repair of UV-damaged DNA were tested. Of the two types of XP cells which were used (Complimentation Groups A and D), neither showed a more sensitive response than normal AN cells, suggesting that cells with this repair defect, at least, are not predisposed to this damage. We have recently observed that, within a few hours after a brief hypoxic exposure to misonidazole, cells which were ultimately destined to survive could be distinguished from those which did not survive. This may allow a more selective analysis of the role, if any, of DNA damage and repair in the cytotoxic effects of misonidazole.
This study shows a correlation between cell inactivation and the DNA damage which follows cytotoxic exposure of mammalian cells to misonidazole. For both of these endpoints, the damage is much greater in hypoxic cells, it increases with drug concentration and with time of exposure to misonidazole and it is temperature-dependent; exposure of cells to the drug at 0°C does not appear to induce DNA strand breaks (Palcic and Skarsgard, unpublished work). Furthermore, the yield of DNA breaks is higher in cells which display an increased sensitivity in survival studies. The cells that will not survive show an even greater number of DNA breaks with further incubation at 37TC after removal of the drug. One could speculate that DNA breaks are not induced directly but that damaged DNA is enzymatically cleaved by repair enzyme(s). If it is assumed that the target for the hypoxic cytotoxic effect We are grateful to Dr C. E. Smithen who proof misonidazole is DNA, then the following vided the misonidazole used in these experiments model may be proposed to explain the re- and to I. Harrison for capable technical assistance. work was supported by the National Cancer sults reported in this study: the drug freely This Institute of Canada and the British Columbia enters cells where it is metabolically re- Cancer Foundation.
MISONIDAZOLE CYTOTOXICITY AND DNA BREAKS REFERENCES ELKIND, M. M. & KAMPER, C. (1970) Two Forms of Repair of DNA in Mammalian Cells Following Irradiation. Biophys. J., 10, 237. HALL, E. J. & BIAGLOW, J. (1977) Ro-07-0582 as a Radiosensitizer and Cytotoxic Agent. Int. J. Radiat. Oncol. Biol. Phys., 2, 521. HALL, E. J. & RoIzIN-TowLE, L. (1975) Hypoxic Sensitizers: Radiobiological Studies at the Cellular Level. Radiology, 117, 453. MOORE, B. A., PALcic, B. & SKARSGARD, L. D. (1976) Radiosensitizing and Toxic Effects of the 2-Nitroimidazole Ro-07-0582 in Hypoxic Mammalian Cells. Radiat. Res., 67, 459. PALCIC, B. & SKARSGARD, L. D. (1972) The Effect of Oxygen on DNA Single-strand Breaks Produced by Ionizing Radiation in Mammalian Cells. Int. J. Radiat. Biol., 21, 417.
59
STRATFORD, I. J. & ADAMS, G. E. (1977) Effect of Hyperthermia on Differential Cytotoxicity of a Hypoxic Cell Radiosensitizer, Ro-07-0582, on Mammalian Cells in vitro. Brit. J. Cancer, 35, 307. TAYLOR, Y. & RAUTH, A. M. (1977) A Comparison of the Hypoxic Cell Specific Toxicity of Ro-070582 Towards HeLa and CHO Cells. (Abstract). Radiat. Res., 70, 702. VARGHESE, A. J., GULYAS, S. & MOHINDRA, J. K. (1976) Hypoxia-Dependent Reduction of 1-(2by Nitro-1-imidazolyl)-3-methoxy-2-propanol Chinese Hamnster Ovary Cells and KHT Tumor Cells in vitro and in vivo. Cancer Res., 36, 3761. WONG, T. W. & WHITMORE, G. F. (1977) A Comparison of Radiation-Sensitizing Ability and Cell Uptake for NDPP and Ro-07-0582. Radiat. Res., 71, 132.
