Deoxyguanosine (dG) enhances the S phase cytotoxicity of camptothecin (CPT), a topoisomerase I (topo I) inhibitor, but by contrast does not affect the toxicity of VM26, a topoisomerase II inhibitor. The 80% survival of S phase human fibroblasts after a 60 min exposure to 0.2 pM CPT is reduced by half in the presence of 25 μM dG. Gi cells are resistant to CPT toxicity, though the levels of the single-strand DNA breaks induced by the drug are similar in Gi and S phase cells. Higher concentrations of dG retard the recovery of RNA and DNA synthesis and inhibit recovery from the S-G2 cycle block after CPT removal. At 100 μM dG the number of CPT-induced protein-linked single-strand DNA breaks is almost doubled, suggestive of a direct effect of dG on the cellular activity of topo I. In the presence or absence of dG, single-strand breaks disappear within minutes of the removal of CPT.

We found that the inhibition of topo I by CPT induces the formation of double as well as singlestrand breaks in the chromosomal DNA. Previously we have shown, using a pulse-field gel electrophoresis technique, that the double-strand breaks (DSBs) are generated predominantly at sites of replication and not in the bulk DNA. A number of these DSBs are long-lived. The present study shows that dG affects the repair of these DSBs in a dose-dependent manner, and that a higher proportion of the initial lesions induced in nascent DNA remain 24 h after removal of CPT. We suggest that the long-lived double-strand breaks, formed in replicating DNA at the time of CPT exposure, are the lethal drug-induced lesions, which explains both the selective cytotoxicity of CPT towards S phase cells and the enhancement of CPT cytotoxicity by dG.

Topoisomerases are enzymes that regulate the superhelicity of DNA by introducing a transient nick into either one (type I) or both (type II) strands of DNA. Topoisomerases play a central role in transcription, DNA replication, segregation and in suppressing mitotic rDNA recombination (Wang, 1985, 1987; Christman et al. 1988), although the independent contribution of each enzyme to these processes is sometimes difficult to assess (Yanagida and Wang, 1987). Certain antitumour drugs have been found to inhibit either topoisomerase I or II, and are widely used as tools in assessing the role(s) of each enzyme in. vivo (Liu, 1989).

Camptothecin is a cytotoxic alkaloid with strong antitumor activity. In eukaryotes the intracellular target for camptothecin action is DNA topoisomerase I, and the presence of wild-type enzyme in the cell is necessary for its cytotoxic effect (Mattern et al. 1987; Hsiang and Liu, 1988; Eng et al. 1988; Kjeldsen et al. 1988). Camptothecin causes fragmentation of cellular DNA. It inhibits both DNA and RNA synthesis, the extent of inhibition correlating with the level of DNA strand breaks induced in the cellular DNA (Liu, 1989). The drug interferes with the topoisomerase I DNA breakage-reunion reaction. It stabilizes the intermediates in the topoisomerase I action, covalently cross-linked protein-DNA breaks known as ‘cleavable complexes’, and thus prevents the alteration in DNA topology. Treatment of these cleavable complexes with a protein denaturant leads to the exposure of single-strand DNA breaks in which the topoisomerase I is covalently linked to the 3’ end of the broken DNA strand (Hsiang et al. 1985). On the basis of the localization of camptothecin-induced topoisomerase I DNA cleavage sites it has been shown that the enzyme is part of the DNA replication apparatus (Snapka, 1986; Avemann et al. 1988; Hsiang et al. 1989), and is also involved in the transcription of genes (Gilmour and Elgin, 1987; Stewart and Schutz, 1987; Zhang et al. 1988; Culotta and Sollner-Webb, 1988).

Given the proposed mechanism of camptothecin’s action, the presence of active topoisomerase I should be both necessary and sufficient to account for the growth inhibition caused by the drug. Paradoxically, however, several studies have shown that camptothecin cytotoxicity does not correlate with the frequency of induced protein-linked single-strand DNA breaks. For example, camptothecin has been shown to be specifically toxic to cells in S phase, though the level of topoisomerase I appears relatively constant throughout the cell cycle (Kessel et al. 1972; Li et al. 1972; Horwitz and Horwitz, 1973; Heck et al. 1988). The single-strand DNA breaks induced are sealed promptly when the drug is removed even when 90% of topoisomerase I is covalently linked (Hsiang and Liu, 1988). Finally, two ionizing radiation-sensitive mutants, Saccharomyces cerevisiae rad 52 and the human disorder, ataxia telangiectasia, are hypersensitive to camptothecin, but in both cases the activity of topoisomerase I and the frequency of protein-linked single-strand DNA breaks induced by the drug appear to be normal (Eng et al. 1988; Smith et al. 1989). It has been shown that camptothecin treatment can generate short-lived double as well as single-strand breaks in genes that are being transcribed (Gilmour and Elgin, 1987; Zhang et al. 1988; Kroeger and Rowe, 1989), and also at the replication fork of the SV40 minichromosome (Snapka, 1986; Avemann et al. 1988). Recently we have demonstrated (Ryan et al. 1991) that camptothecin cytotoxicity is associated with the generation of double-strand breaks, predominantly at sites of DNA replication. A fraction of these DNA breaks persist even 24 h after camptothecin removal. Aphidicolin, an inhibitor of DNA polymerases alpha and delta, abolishes camptothecin cytotoxicity (Hsiang et al. 1989; Holm et al. 1989), and we have shown (Ryan et al. 1991) that it also prevents the generation of double-strand breaks and thus protects the cells from the lethal effect of camptothecin.

Recently we have shown that fibroblasts taken from individuals with the inherited disease Cockayne’s syndrome (CS) are not only sensitive to UV, but are also hypersensitive to deoxyguanosine (dG) and camptothecin (CPT) (Squires et al. 1990). While it has been established that topoisomerase I is the target for CPT action (reviewed by Liu, 1990), the mechanism of dG toxocity is unknown (Duan et al. 1990). In the present work we address the question of whether dG affects the cellular response to CPT and by implication modifies topoisomerase I activity. We report that dG enhances the toxicity of CPT but not that of the topoisomerase II inhibitor, VM26. CPT induces double-as well as single-strand breaks in the chromosomal DNA. The double-strand DNA breaks (DSBs) are generated predominantly in replicating DNA, and they are long lived. Our results suggest that the DSBs are the cytotoxic lesions induced by CPT, and that dG reduces the extent of their repair and thus enhances CPT toxicity.

Materials

Camptothecin sodium salt (Sigma) and teniposide (VM-26) were dissolved in dimethylsülfoxide and stored in small aliquots at concentrations of 1-10mM at −20°C. The drugs were protected from visible light and were diluted in growth medium immediately before use. Hydroxyapatite was purchased from Boehringer Mannheim Ltd, Eagle’s minimal essential medium (MEM), vitamins and essential amino acids from Life Technologies Ltd and foetal calf serum from ICN Flow. Stocks (50-100 mM) of deoxynucleosides (Sigma) were made in phosphate buffered saline and stored in aliquots at −20°C.

Cell culture and cytoxicity measurements

Three human lines were used: embryonic lung fibroblasts HEL (Life Technologies Ltd), simian virus 40-transformed skin fibroblasts SV40MRC5V1 (a gift from MRC Cell Mutation Unit, Brighton) and a clone of human bladder carcinoma line EJ30/8D 2 TGR. Cells were grown in Eagle’s minimum essential medium (MEM) containing 5% (v/v) foetal calf serum (EJ30/8D) or 10% (HEL and SV40MRC5).

The toxicity of each treatment was assayed by cell proliferation and clonal cell survival. One day after seeding the cells were exposed for 60 min to different concentrations of freshly prepared camptothecin or VM26. After the removal of the drugs the cultures were washed in warm medium or in phosphate buffered saline (PBS), and incubated further in the appropriate supplemented growth medium. The rate of proliferation was determined by means of cell counting (Coulter Electronics Inc) 2-6 days after seeding, and the cell number is expressed as a percentage of the untreated controls. In the clonal survival assay cells from exponentially growing cultures were plated at low density in growth medium with or without one of the four deoxynucleosides (10-25 μM). A day after seeding the cells were exposed for 60 min to various concentrations of the inhibitors of the topoisomerases. After drug treatment the dishes were washed with warm medium and further incubated for 8-14 days to allow colonies to develop. In experiments involving the combined treatment with deoxynucleoside and the topoisomerase inhibitor, the nucleoside was present before, during and after inhibitor treatment. The colony survival curves were constructed for each cell cycle phase by plotting the log of surviving fraction against camptothecin concentration.

Synchronization procedure

Diploid fibroblasts were synchronized by growing the cultures to confluence and releasing the cells from density inhibition by detachment with pancreatin as described (Squires et al. 1987). At different times after release the cells were exposed to a 60 min pulse of camptothecin and were examined after drug removal for colony survival and the induction of protein-linked single-strand DNA breaks. In parallel, the S phase index was determined by labelling the cells for 30 min with tritiated thymidine, followed by autoradiography.

Cell-cycle phase analysis

Cells were seeded into 6-well plates (8×104 cells/well) and incubated for 48 h before a 1-h treatment with camptothecin, followed by two washes in PBS and continued incubation in fresh growth medium. In experiments involving combined treatments deoxyguanosine was added 24 h before camptothecin exposure, and was then present during and after the camptothecin treatment period. At specified times cultures were washed with PBS lacking Ca2+ and Mg2+, detached with trypsin/versene, and resuspended in complete medium. The DNA of permeabilised RNase-digested cells was stained with ethidium bromide using the one-step technique described previously (Smith et al. 1985). Flow cytometric determinations of cellular DNA contents and cell cycle phase distributions were determined as described previously (Watson et al. 1987).

Determination of protein-linked single-strand DNA breaks

The in vivo formation of protein-linked DNA breaks was determined using two assays in parallel in aliquots from the same sample. The frequency of DNA breaks was measured by alkaline unwinding followed by hydroxyapatite chromatography (Squires et al. 1982), and a modified K+-SDS precipitation method (Rowe et al. 1986) for DNA-protein cross-link estimates. A total of 1×105 to 2×105 cells, uniformly prelabelled with [3H]thymidine, were exposed to different concentrations of the inhibitors of both classes of DNA topoisomerases. After incubation the cells were lysed on ice for 25 min in alkaline sucrose solution (5%, w/v, sucrose, 10 mM Na2EDTA, 150 mM NaCl and 100 mM NaOH), before neutralizing with KH2PO4 and reducing the size of the DNA by sonication (MSE Instrument Ltd). The extent of unwinding in alkali is a function of the frequency of DNA breaks, the number of DNA breaks being estimated from the calibration curves of X-irradiated cells as described (Squires et al. 1982). For all experimental points the low break frequencies of the untreated controls were deducted from the values of the drug-treated cells.

The K+-SDS coprecipitation assay for measuring the DNA-protein cross-links is a modification of the assay described’ by Rowe et al. (1986). A 1 ml sample of the sheared cell lysate was transferred to a 1.5 ml Eppendorf tube and 100 mM KC1, 1% SDS and 0.5 mg herring sperm DNA were added. The samples were vortexed vigorously for 15 s, and incubated on ice for 10 min. From this point the protein-DNA complexes were treated as described by Rowe et al. (1986). Radioactivity was determined in each sample before and after K+-SDS coprecipitation, and the counts in the precipitate expressed as a percentage of the total counts. In the proteinase K-treated samples, 0.5 mg of the enzyme was added to each tube before the K+-SDS precipitation and the samples were incubated for 60 min at 50 °C.

Asymmetric field inversion gel electrophoresis (AFIGE) for measurement of double-strand DNA breaks

In experiments for measuring double-strand DNA breaks (DSBs), exponentially growing SV40MRC5 cells were uniformly labelled with [14C]thymidine for 2-3 days. One day before camptothecin (CPT) treatment the medium was replaced with fresh unlabelled MEM, and in those experiments involving combined treatments deoxyguanosine was added. To identify replicating DNA the cells were exposed simultaneously to CPT and l μCiml−1 [3H]thymi-dine for 50 min. The drug and label were removed, cells were washed twice with buffered saline, and the cells assayed for DSBs immediately or after a 5 or 24 h incubation period in fresh medium plus or minus deoxyguanosine.

We have used the gel electrophoresis system developed by Stamato and Denko (1990) for measuring DSBs generated in replicating and bulk DNA, as described by Ryan et al. (1991). Briefly, cells were detached with viokase, washed and resuspended at a concentration of 0.3×107 to 1.0×107cellsml−1 in 0.8% low melting point agarose (Bethesda Research Laboratories Inc.) at 37 °C. The agarose/cell mixture was taken up into 3 mm internal diameter tubing, solidified on ice, and cut into 5 mm pieces. These agarose plugs were incubated overnight at 50°C in 0.5M EDTA, pH8.0, 1% Sarkosyl and 50/zgml−1 proteinase K. Electrophoresis was carried out in 1.5% agarose gels using 45 mM Tris, 45 mM boric acid, 1.5 mM EDTA, pH 8.0, containing 0.025/zgml−1 ethidium bromide. The pulse conditions were 5Vcm−1 for 125 s in the direction of net DNA migration, and lOVcm−1 for 15 s in the reverse direction, for total run time of about 6h. The temperature during electrophoresis was maintained at 10-14°C using cooled recirculating buffer. Under these conditions only DNA that is smaller than 3Mbp (base pairs) enters the gel where it migrates as a discrete band that can be visualized under ultraviolet light. The amount of DNA that enters the gel from an agarose plug is a measure of DSBs. Using radiolabelled precursors, the DSBs in a sample can be determined by excising both the band of migrating DNA and the agarose plug from the gel following AFIGE, then counting the radioactivity present in each, and the results are expressed as a fraction of radioactivity released (%FAR) from the plug into the gel (Stamato and Denko, 1990; Ryan et al. 1991).

It has been shown that for a given dose of X-irradiation, the release of the DNA from the plug is significantly reduced in S phase cells relative to cells in other phases of the cell cycle (Illiakis et al. 1991a). We have also shown (Ryan et al. 1991) that for increasing doses of gamma-irradiation (at least up to 150 Gy) FAR following AFIGE is approximately linear for both bulk and pulsed radiolabelled DNA, but the FAR values for the replicating DNA [3H]TdR (thymidine) in comparison to those for the bulk DNA [14C]TdR are several-fold lower. The relative response of the replicating and bulk DNA in the AFIGE assay can vary in different cell strains and at different times after treatment. To compare between samples and cell types, we expressed drug-induced DSBs in pulse-labelled and bulk-labelled DNA as gray equivalent DSBs (Ryan et al. 1991). This value represents the dose of gamma-irradiation in Gy, which gives an equivalent FAR value to that observed in the experimental sample. To estimate this value for a given sample we subject a few agarose plugs to 100 Gy irradiation, and thereafter determine the FAR value by AFIGE analysis. If, for example, the FAR of the drug-treated sample alone is 10% and that of the sample plus 100 Gy is 30% (i.e. an additional 20% due to the irradiation), then the sample FAR of 10% represents 50 Gy equivalent of DSBs, assuming a linear response to gamma-irradiation (Ryan et al. 1991). For all experimental conditions the FAR values of the mock-treated controls were subtracted from the values of the drug-treated cells.

Growth inhibition caused by camptothecin is enhanced by deoxyguanosine

We have examined the effect of deoxyguanosine (dG) on the camptothecin (CPT)-induced growth inhibition of two human cell lines, an SV40-transformed fibroblast, SV40MRC5, and the bladder carcinoma-derived EJ30. Fig. 1A shows that a 60 min exposure to CPT reduces the rate of cell proliferation, and that the inhibition is much more pronounced when the cultures are cotreated with 25 μM dG; 1 μM CPT inhibits the growth rate by about 25% and the addition of dG reduces growth further to 50% of control rates. In parallel control cultures, growth during the same period is logarithmic with 4-to 10-fold increase in cell density. Increasing camptothecin concentration from 1 to 2.5 μM does not lead to extra inhibition, consistent with previous reports that a subpopulation of cells is resistant to the drug (Li et al. 1972).

Fig. 1.

Camptothecin (CPT) inhibition of cell proliferation is greater in the presence of deoxyguanosine (dG). SV40MRC5 (○, •) and EJ3018D (▵, ▴) cells were treated with various concentrations of CPT with or without 25 μG (A), or with various concentrations of dG plus or minus 0.5 μM CPT (B). 104 cells per well were seeded and 2 days later were treated with CPT for 60 min. After CPT removal the cells were incubated further in fresh medium. For cotreatment dG was present a day before, during and after CPT exposure. Proliferation was determined by cell counting 2 and 5 days after CPT treatment, and is expressed as a percentage of the proliferation of untreated control. The data represent an average of at least two independent experiments; the standard errors are around 10%. Filled symbols in (A) plus 25 μM dG; and in (B) plus 0.5 μM CPT.

Fig. 1.

Camptothecin (CPT) inhibition of cell proliferation is greater in the presence of deoxyguanosine (dG). SV40MRC5 (○, •) and EJ3018D (▵, ▴) cells were treated with various concentrations of CPT with or without 25 μG (A), or with various concentrations of dG plus or minus 0.5 μM CPT (B). 104 cells per well were seeded and 2 days later were treated with CPT for 60 min. After CPT removal the cells were incubated further in fresh medium. For cotreatment dG was present a day before, during and after CPT exposure. Proliferation was determined by cell counting 2 and 5 days after CPT treatment, and is expressed as a percentage of the proliferation of untreated control. The data represent an average of at least two independent experiments; the standard errors are around 10%. Filled symbols in (A) plus 25 μM dG; and in (B) plus 0.5 μM CPT.

We assessed the effect of increasing dG concentrations on cell proliferation at a fixed dose of CPT (0.5 μM). Fig. 1B shows that the growth rate of SV40MRC5 cells is affected only at a concentration of 100/ZM dG; the cell doubling time increases from 24 h to 48 h. At this concentration of dG DNA synthesis is inhibited by 25% (data not shown). Treatment of cells with 25-50 μM dG increases by severalfold the inhibition of growth caused by CPT. However, at the higher concentration of dG CPT hardly increases the growth inhibition caused by 100 /ZM dG given alone.

Deoxy guanosine selectively enhances the cytotoxicity of a topoisomerase I inhibitor

We have examined by means of colony-survival assay whether deoxyribonucleosides other than dG affect the toxicity of camptothecin or VM26 in SV40MRC5 cells after a 1-h exposure. Low concentrations (10-25/ZM) of deoxyribonucleosides, which do not affect the plating efficiency of the cells, were present in the medium before, during and after drug exposure. Of the four deoxyribonucleosides examined (dA, dC, dT, dG), dG is the only one to reduce the camptothecin colony survival; the concentration of camptothecin which inhibits colony growth by 50% (ID50) is 0.50/ZM, and is reduced to 0.26/ZM in the presence of deoxyguanosine. Neither deoxyguanosine, deoxycytidine nor deoxyadenosine affect the ID50 of VM26 (0.52 μM). The specific effect of deoxyguanosine on the cellular response to CPT suggests that a metabolite in the deoxyguanosineguanosine pathway could interact directly with topoisomerase I to modify its activity in vivo. Alternatively, dG may act indirectly by affecting the overall rate of cell proliferation and thus increasing the fraction of cells susceptible to camptothecin toxicity.

Cotreatment with deoxyguanosine increases the frequency of protein-linked single-strand DNA breaks induced by camptothecin

If deoxyguanosine increases the cellular level of DNA damage generated by camptothecin this might help to explain its effect of increasing camptothecin-induced cytotoxicity. We have modified the cell lysis conditions for the K+-SDS precipitation assay (Rowe et al. 1986), and thus are able to quantitate the level of DNA breakage induced by CPT in vivo and to verify that the DNA breaks are protein-linked. After a short exposure to camptothecin the radio-labelled cells are lysed and the DNA unwound in alkali before reducing the size of the DNA by sonication. The alkaline lysis and DNA unwinding procedure appears to have stabilized the labile topoisomerase I-DNA break complexes (Covey et al. 1989), and the sensitivity of this modified assay has allowed us to determine a camptothecin effect over a range of low concentrations from 0.1 to 2 UM. Fig. 2 shows that in SV40MRC5 cells the level of camptothecin-induced protein-linked single-strand DNA breaks is dose-dependent, reaching a plateau at concentrations of 1-2 fiM. These results correspond well with the maximum effect of camptothecin found in human (Fig. 1) and mouse leukaemic cells (Hsiang et al. 1989).

Fig. 2.

Deoxyguanosine increases the levels of camptothecin-induced single-strand DNA breaks and protein-linked DNA breaks. [3H]thymidine-labelled SV40MRC5 cells were seeded into medium with 100 UM dG (filled symbols) or without dG (open symbols). One day after seeding the cells were treated for 15 min with various concentrations of camptothecin (CPT). After drug treatment the cells were lysed and the DNA was unwound in alkaline solution. The sonicated samples were assayed for single-strand DNA breaks by hydroxyapatite chromatography (A) and for protein-linked DNA breaks by K+-SDS precipitation (B), as described in Materials and methods. Samples were treated with proteinase K (+PK), before K+-SDS precipitation. The 3H counts in the K+-SDS precipitate are presented as a percentage of the total [3H]DNA counts per sample. Points and vertical lines are averages ±s.n. for 5 independent experiments. At a concentration of 0.5 /<M CPT the standard errors of the mean break frequencies per 109 daltons (○) ±0.5 and plus dG are (•) ±1.4; and for% 3H counts in the K+-SDS precipitate (○) ±18% and (•) ±13%.

Fig. 2.

Deoxyguanosine increases the levels of camptothecin-induced single-strand DNA breaks and protein-linked DNA breaks. [3H]thymidine-labelled SV40MRC5 cells were seeded into medium with 100 UM dG (filled symbols) or without dG (open symbols). One day after seeding the cells were treated for 15 min with various concentrations of camptothecin (CPT). After drug treatment the cells were lysed and the DNA was unwound in alkaline solution. The sonicated samples were assayed for single-strand DNA breaks by hydroxyapatite chromatography (A) and for protein-linked DNA breaks by K+-SDS precipitation (B), as described in Materials and methods. Samples were treated with proteinase K (+PK), before K+-SDS precipitation. The 3H counts in the K+-SDS precipitate are presented as a percentage of the total [3H]DNA counts per sample. Points and vertical lines are averages ±s.n. for 5 independent experiments. At a concentration of 0.5 /<M CPT the standard errors of the mean break frequencies per 109 daltons (○) ±0.5 and plus dG are (•) ±1.4; and for% 3H counts in the K+-SDS precipitate (○) ±18% and (•) ±13%.

Preincubating the cells with 100 pM deoxyguanosine almost doubles the number of single-strand DNA breaks generated by up to 1 μM camptothecin (Fig. 2A). This DNA precipitates in K+-SDS, indicating that the DNA breaks are cross-linked to protein (Fig. 2B). Neither deoxyadeno sine nor deoxycytidine affects the frequency of DNA breaks or the level of protein-DNA cross-links induced by camptothecin (data not shown). The effect of dG is camptothecin-specific and does not affect DNA break frequencies induced by VM26 (data not shown). The lack of a dG effect on cells treated with VM26, either on colony survival or on DNA breakage, emphasizes the specificity of dG towards topoisomerase I. In contrast to the effect of 100/ZM dG, the level of cleavable complexes induced by camptothecin is not affected by 25/ZM deoxyguanosine (data not shown), though at this concentration dG enhances cell kill. Therefore, we conclude that dG does not enhance camptothecin cytotoxicity by increasing the number of single-strand DNA breaks. The higher level of camptothecin-induced ‘cleavable complexes’ in the presence of 100/ZM deoxyguanosine does suggest, however, that deoxyguanosine interacts in some manner with topoisomerase I and affects its cellular activity.

Deoxyguanosine does not affect the rate of protein-linked DNA strand break sealing after camptothecin removalFig. 3 shows the time course of accumulation of DNA breaks in the presence of 1 /ZM camptothecin, and the rate of their disappearance after drug removal. The induction of cellular DNA breaks and protein-DNA cross-links by camptothecin is rapid and levels off within 30 min of drug treatment. DNA breaks disappear promptly upon the removal of the drug (Fig. 3) as do protein-DNA cross links (data not shown). These time courses for human cells agree with the published results on mouse cells and are consistent with the cleavable complex hypothesis, which suggests that once a DNA break is sealed the topoisomerase I is no longer stably bound to the DNA (Hsiang et al. 1985; Hsiang and Liu, 1988; Mattern et al. 1987)

The addition of 100 μM dG does not inhibit the ligation step, though the initial level of DNA breaks induced is higher (Fig. 3). It is clear from these results that dG does not stabilize the cleavable complexes, and therefore that this is not the mechanism by which dG sensitizes the cells to the toxic effect of camptothecin.

Fig. 3.

Deoxyguanosine increases the level of camptothecin-induced protein-linked single-strand DNA breaks, but not the rate of their disappearance after drug removal. SV40MRC5 cells uniformly labelled with [3H]thymidine were treated for 30 min with 1 /ZM camptothecin, washed twice and incubated further in fresh medium. At the times indicated single-strand DNA breaks were assayed, as described in the legend to Fig. 2. (•) Plus 100 μM dG and (○) without dG.

Fig. 3.

Deoxyguanosine increases the level of camptothecin-induced protein-linked single-strand DNA breaks, but not the rate of their disappearance after drug removal. SV40MRC5 cells uniformly labelled with [3H]thymidine were treated for 30 min with 1 /ZM camptothecin, washed twice and incubated further in fresh medium. At the times indicated single-strand DNA breaks were assayed, as described in the legend to Fig. 2. (•) Plus 100 μM dG and (○) without dG.

The recovery of RNA and DNA synthesis after camptothecin treatment is slower in the presence of deoxyguanosine

Work in other laboratories has shown that the extent of inhibition of RNA and DNA synthesis by CPT correlates with the level of breaks induced in cellular DNA (reviewed by Liu, 1989). We have examined the effect of dG on the extent of inhibition of RNA and DNA synthesis by CPT, and on the rate of the recovery of synthesis after drug removal. In Fig. 4 we show that camptothecin severely inhibits both RNA and DNA synthesis in SV40MRC5 cells. Cotreatment with dG does not affect the degree of inhibition of either RNA or DNA synthesis. Fig. 4A shows, however, that while the inhibition of RNA synthesis is fully reversed 90 min after CPT removal, in the presence of 100 μM dG the recovery of RNA synthesis is much slower: an additional 2h are required to reach the same rate of synthesis as is achieved without dG. Moreover, the surge in RNA synthesis in the camptothecin-treated cultures, seen 3-5 h after drug removal, is only very slight when dG is present. The time course of recovery of RNA synthesis after VM26 treatment is similar to that after camptothecin, but dG does not affect the rate of recovery after drug removal (data not shown).

Fig. 4.

Inhibition of RNA and DNA synthesis by camptothecin (CPT); deoxyguanosine (dG) retards nucleic acid synthesis recovery after CPT removal. Logarithmically growing SV40MRC5 cultures were exposed for 60 min to CPT two days after seeding. For RNA synthesis (A) the cells were treated with 5 /ZM, and for DNA (B) with 2.5 /ZM CPT. Cells were pulse-labelled for 30 min with 2/zCiml’1 of [3H]uridine or 1/zCiml−1 [3H]TdR at various times as indicated. The points correspond to amounts of acid-insoluble radioactivity in CPT-treated versus untreated cells, normalized for cell number by means of the radioactivity present in the equilibrated acid-soluble pool of each sample. dG was present one day before, and also during and after CPT treatment. (○) Without dG; (▴) with 25 μM or 100 μM dG (▪).

Fig. 4.

Inhibition of RNA and DNA synthesis by camptothecin (CPT); deoxyguanosine (dG) retards nucleic acid synthesis recovery after CPT removal. Logarithmically growing SV40MRC5 cultures were exposed for 60 min to CPT two days after seeding. For RNA synthesis (A) the cells were treated with 5 /ZM, and for DNA (B) with 2.5 /ZM CPT. Cells were pulse-labelled for 30 min with 2/zCiml’1 of [3H]uridine or 1/zCiml−1 [3H]TdR at various times as indicated. The points correspond to amounts of acid-insoluble radioactivity in CPT-treated versus untreated cells, normalized for cell number by means of the radioactivity present in the equilibrated acid-soluble pool of each sample. dG was present one day before, and also during and after CPT treatment. (○) Without dG; (▴) with 25 μM or 100 μM dG (▪).

After removal of CPT the recovery of DNA synthesis is also delayed and rate of synthesis reaches only about 80% of that of the untreated cells 6h after drug removal (Fig. 4B). Cotreatment with 25 or 100 μM dG retards the rate of recovery; 6 h after drug removal the rate of DNA synthesis is still only 40% to 50% of the untreated control.

Cell cycle analysis after camptothecin treatment: the effect of deoxy guanosine on the rate of recovery after drug removal

A 1-h exposure to CPT has a rapid effect on the cell cycle behaviour of the population, so that cells accumulate in S phase up to 12 h following treatment, a response that is relatively independent of dose over the range of concentrations studied (0.1-2.5 μM). Increasing concentrations of CPT act to block cells at progressively earlier points during replication, although the total percentage of cells trapped in S phase is constant. Fig. 5 shows representative results for the effects of 1 gM CPT. The delay in S phase exit is mirrored by a depletion of cells from Gi, reaching a minimum 12 h after drug removal (Fig. 5A). This indicates that in addition to the slowing of cells in S phase (Fig. 5B) there is a further block to cell division involving inhibition of G2 emptying (Fig. 5C). The S phase recovery is indicated already at 12 h post-treatment with the arrival of a cohort of S phase-delayed cells in G2 (Fig. 5C) and the resupply of cells to Gi by 24 h post-treatment (Fig. 5A).

Fig. 5.

Deoxyguanosine at 100 μM increases the cell cycle perturbation induced by camptothecin. Logarithmically growing SV40MRC5 cultures were treated 48 h after seeding with 1 μM camptothecin (filled symbols). One day after seeding the medium was changed to complete medium without (---,•), with 25 μM deoxyguanosine (dG) (▵, ▴) and with 100 μM dG (▫, ▪). The cell cycle distribution was monitored as described in Materials and methods at 6, 12, 24 and 48 h after camptothecin removal.

Fig. 5.

Deoxyguanosine at 100 μM increases the cell cycle perturbation induced by camptothecin. Logarithmically growing SV40MRC5 cultures were treated 48 h after seeding with 1 μM camptothecin (filled symbols). One day after seeding the medium was changed to complete medium without (---,•), with 25 μM deoxyguanosine (dG) (▵, ▴) and with 100 μM dG (▫, ▪). The cell cycle distribution was monitored as described in Materials and methods at 6, 12, 24 and 48 h after camptothecin removal.

The effect of 25 and 100 μM dG on the cell cycle distribution after camptothecin removal was examined (Fig. 5). To maximize the effect of dG the cultures were pretreated 24 h before CPT exposure; 25 gM dG does not affect the cell distribution of either the CPT-treated or untreated cultures, though, as we showed above, it delays the recovery of DNA synthesis (Fig. 4B) and enhances CPT cytotoxicity after drug removal. However, treatment of cultures with 100 μM dG alone has an effect on cell cycle distribution, and this effect is more marked with camptothecin (Fig. 5A-C). The addition of 100 μM dG to the growth medium results in a progressive depletion of cells in Gi phase and an accumulation of cells in S phase, the effects only becoming significant after 24 h of exposure. This partial block in the flow of cells through the cycle is probably the result of a 25% reduction in the rate of DNA synthesis seen after dG was added (data not shown); 100 gM dG has no effect on the initial 1 gM camptothecin-induced S phase accumulation, but it significantly delays (by 12-24 h) the resolution of the block to Gi entry and S phase traverse (Fig. 5A,B). The cotreated cultures show enhanced long-term accumulation of cells in G2, and the accumulation is progressive from the 6 to 48 h after CPT treatment (Fig. 5C).

A direct interpretation of the results is that the recovery from camptothecin-induced delay acts to generate a cohort of cells that normally traverses G2 at 12-24 h following treatment. At 100 μM dG, the block to G2 exit results in the capture of this cohort of cells in G2.

Effects of camptothecin on DNA break induction and cell survival during the cycle; deoxyguanosine enhances the killing of mid and late S phase cells

If 25 gM dG sensitizes the normally resistant population of non-S phase cells to CPT killing this might explain the increased growth inhibition of cotreated asynchronous cultures. To test this hypothesis we studied the effect of dG on camptothecin cytotoxicity in synchronized human fibroblast cultures (Fig. 6). Human fibroblasts in different stages of the cell cycle were treated with camptothecin and cytotoxicity was measured by colony survival. Synchronization was achieved by releasing quiescent cultures from density inhibition by enzymatic detachment and replating the cells at low density in fresh medium with or without 25 μM dG. At different times after release (5 to 30 h) cells were exposed to camptothecin for 60 min. The progression of cells through the cycle was monitored autoradiographically in parallel samples by counting the proportion of cells labelled during a 30 min exposure to tritiated thymidine (Table 1). Fig. 6 shows that: (a) S phase but not Gi phase cells are sensitive to the cytotoxic effect of CPT with or without dG; (b) mid to late S, or possibly even early G2 cells, are most sensitive to the drug, results that are in good agreement with data on synchronized Chinese hamster DON cells (Li et al. 1972); (c) cotreatment with 25 /ZM dG increases camptothecin-induced cell lethality only in the mid to late S phase population. This increase of cell lethality attributable to dG is twofold, from 80 to 40% survival at 0.2 μM CPT. Therefore, the dG enhancement of CPT cytotoxicity (Fig. 1) is not due to sensitization of Gi cells but to the hypersensitization of S phase (and possibly G2) cells to the toxic effects of camptothecin.

Table 1.

Deoxyguanosine enhances the S phase specificity of camptothecin cytotoxicity; lack of correlation with the frequency of induced single-strand DNA breaks

Deoxyguanosine enhances the S phase specificity of camptothecin cytotoxicity; lack of correlation with the frequency of induced single-strand DNA breaks
Deoxyguanosine enhances the S phase specificity of camptothecin cytotoxicity; lack of correlation with the frequency of induced single-strand DNA breaks
Fig. 6.

Cytotoxicity of camptothecin is exaggerated in populations enriched in S phase cells; deoxyguanosine enhances the lethality of mid and late S phase human fibroblasts treated with camptothecin. Quiescent fibroblasts were obtained 3-4 weeks after seeding with at least 3 medium changes during the interim period. Synchronized resting cells were released from contact inhibition by seeding the cells in fresh medium with 25 μM deoxyguanosine (filled symbols) or without (open symbols). At different times after release from contact inhibition the cells were exposed to various camptothecin concentrations for 60 min. Clonal survival was assayed as described in Materials and methods and the flow of cells through the cycle is given in Table 1. The% survival was determined in duplicates from two or more independent experiments. The symbols for the times in hours after release from contact inhibition are as follows; (○,•) 5h (G1); (◊, ♦) 16.5 (‘early S’); (▵, ▴) 23 (‘mid S’); (▫, ▪) 30 (‘late S’). Values for% survival at 0.1 μM CPT with their standard errors are as follows: (○) 103±12, (•) 97±5, (◊) 100±5, (♦ 93±10, (▵) 76±15, (▴) 41±3, (▫) 68+12, (▪) 43±7.

Fig. 6.

Cytotoxicity of camptothecin is exaggerated in populations enriched in S phase cells; deoxyguanosine enhances the lethality of mid and late S phase human fibroblasts treated with camptothecin. Quiescent fibroblasts were obtained 3-4 weeks after seeding with at least 3 medium changes during the interim period. Synchronized resting cells were released from contact inhibition by seeding the cells in fresh medium with 25 μM deoxyguanosine (filled symbols) or without (open symbols). At different times after release from contact inhibition the cells were exposed to various camptothecin concentrations for 60 min. Clonal survival was assayed as described in Materials and methods and the flow of cells through the cycle is given in Table 1. The% survival was determined in duplicates from two or more independent experiments. The symbols for the times in hours after release from contact inhibition are as follows; (○,•) 5h (G1); (◊, ♦) 16.5 (‘early S’); (▵, ▴) 23 (‘mid S’); (▫, ▪) 30 (‘late S’). Values for% survival at 0.1 μM CPT with their standard errors are as follows: (○) 103±12, (•) 97±5, (◊) 100±5, (♦ 93±10, (▵) 76±15, (▴) 41±3, (▫) 68+12, (▪) 43±7.

The number of camptothecin-induced DNA breaks in the synchronized fibroblasts at different stages in the cycle is also shown in Table 1. Levels of single-strand DNA breaks induced by CPT are similar in Gi and S phase, and the small fluctuations seen do not correlate with the cycle-related sensitivity of the cells to the drug. Our measurements of CPT-induced DNA breakage during the cycle are in agreement with the reported constant levels of topoisomerase I enzyme in cells from different phases of the cycle (Heck et al. 1988).

Camptothecin induces double-strand breaks in newly synthesized DNA; deoxyguanosine reduces the repair of CPT-damaged nascent DNA

It has been suggested (Hsiang et al. 1989) that camptothecin induces both double- and single-strand DNA breaks at replication forks, the former being lethal to the proliferating cell. Recently we have shown (Ryan et al. 1991) that double-strand breaks (DSBs) are induced during camptothecin exposure in newly synthesized DNA of human cells, and persist for hours after drug removal. We have proposed that camptothecin cytotoxicity, which is biased towards S phase cells, is associated with the generation of the long-lived DSBs at the replication fork during drug exposure.

In this report we have analysed the effect of dG on CPT-induced double-strand breaks. We have used the pulsefield gel electrophoresis system (AFIGE) developed by Denko et al. (1989), where the amount of DNA migrating into the gel following damage is proportional to the number of DSBs (Stamato and Denko, 1990). We have compared DSBs arising in newly replicated DNA (pulsed labelled with tritiated thymidine) and those in bulk DNA (uniformly pre-labelled with [14C]thymidine). Using the AFIGE system, Iliakis et al. (1991a,b) have shown that for a given dose of X-irradiation, the release of DNA from the plug fluctuates throughout the cell cycle to reach a minimum FAR value in mid S phase. Their results show that the electrophoretic mobility of a replicating DNA molecule is retarded and that the differences in the FAR per Gy throughout the cell cycle are a reflection of the presence of S phase cells. We have recently shown that in X-irradiated cells the release of replicating DNA ([3H]TdR) from the plug is significantly reduced in comparison to bulk DNA ([14C]TdR) (Ryan et al. 1991). The relative difference in the release of radioactivity from the plug between replicating and bulk DNA is cell specific; in SV40MRC5 cells the FAR of nascent DNA is 6.7 (±1.4)-fold lower than that of the bulk DNA, in both the control untreated and the CPT-treated samples. After a 24-h chase following [3H]TdR removal, the FAR values of X-irradiated plugs are the same for the DNA that was replicating at the time of exposure and bulk DNA, indicating that the nascent DNA has now matured in both control untreated and CPT-treated samples. To compare between different cell types and at different times after treatment, the camptothecin-induced DSBs in replicating DNA and bulk DNA, measured as [3H]- and [14C]FAR (%), respectively, are expressed as gray equivalent DSBs as described in Materials and methods and by Ryan et al. (1991).

In this set of experiments the 14C-prelabelled SV40MRC5 cells are treated for 50 min with 2.5 μM CPT in the presence of 1 μCiml−1 of [3H]TdR. The drug and the label are then removed, and the fraction of radioactivity released into the gel is measured in samples taken immediately after exposure, and again 5 and 24 h after drug removal. Fig. 7 shows that during CPT treatment DSBs are generated almost exclusively in the 3H-labelled nascent DNA. At 24 h after CPT removal the majority of the DSBs have disappeared, presumably repaired, and only about 3 gray equivalents remain. The relative longevity of CPT-induced DSBs, a fraction of which persists for up to 24 h, contrasts with the extremely fast and complete repair of single-strand DNA breaks (Fig. 3).

Fig. 7.

Camptothecin (CPT) induces double-strand breaks (DSBs) in nascent DNA; deoxyguanosine (dG) affects the induction and persistence of these breaks in a dose-dependent manner. The DNA in SV40MRC5 cells was uniformly labelled with [14C]thymidine for two days. The label was then removed and the cells were incubated further in standard medium or medium supplemented with 25 μM, 50 μM or 100 μM dG. After 16 h the cells were treated simultaneously with 2.5 μM CPT and 1 μC ml*1 of [3H]TdR for 50 min. dG was present before, during and after CPT treatment. Cells were embedded in agarose immediately or after chasing for various periods of time in unlabelled medium without CPT. DSBs were detected by AFIGE as described in Materials and methods. The DSBs generated in the DNA that was replicating at the time of drug exposure (3H-labelled) and in the bulk DNA (14C-labelled) are expressed in gray equivalent. 2.5 UM CPT (▫), CPT plus 25 μM dG (▨), CPT plus 50 μM dG (▨), CPT plus 100 μM dG (▪). The FAR% was determined from 2-7 independent experiments. The values for DSBs (Gy equiv.) in 3H-labelled DNA with their standard errors at time 0 and at 24 h (in parenthesis) are as follows: (▫) 44±14 (3±0.9); (▨) 33±7.4 (5±0.9), (▨) 28±5.6, (8±2.5), (▪) 22±3, (15+0.6). The values in 14C-labelled DNA of DSBs (Gy equiv.) at 24 h after CPT removal are: (▫) 0.8±0.8; (▨) 1±1, (▨) 2.1±1, (▪) 3.2±2.

Fig. 7.

Camptothecin (CPT) induces double-strand breaks (DSBs) in nascent DNA; deoxyguanosine (dG) affects the induction and persistence of these breaks in a dose-dependent manner. The DNA in SV40MRC5 cells was uniformly labelled with [14C]thymidine for two days. The label was then removed and the cells were incubated further in standard medium or medium supplemented with 25 μM, 50 μM or 100 μM dG. After 16 h the cells were treated simultaneously with 2.5 μM CPT and 1 μC ml*1 of [3H]TdR for 50 min. dG was present before, during and after CPT treatment. Cells were embedded in agarose immediately or after chasing for various periods of time in unlabelled medium without CPT. DSBs were detected by AFIGE as described in Materials and methods. The DSBs generated in the DNA that was replicating at the time of drug exposure (3H-labelled) and in the bulk DNA (14C-labelled) are expressed in gray equivalent. 2.5 UM CPT (▫), CPT plus 25 μM dG (▨), CPT plus 50 μM dG (▨), CPT plus 100 μM dG (▪). The FAR% was determined from 2-7 independent experiments. The values for DSBs (Gy equiv.) in 3H-labelled DNA with their standard errors at time 0 and at 24 h (in parenthesis) are as follows: (▫) 44±14 (3±0.9); (▨) 33±7.4 (5±0.9), (▨) 28±5.6, (8±2.5), (▪) 22±3, (15+0.6). The values in 14C-labelled DNA of DSBs (Gy equiv.) at 24 h after CPT removal are: (▫) 0.8±0.8; (▨) 1±1, (▨) 2.1±1, (▪) 3.2±2.

To assess the influence of dG on CPT-induced DSBs, cells were treated with dG for one day before CPT exposure and then subsequently. Fig. 7 shows that dG reduces the initial level of DSBs induced by CPT in a dose-dependent manner. At 200 μM dG the effect is even more pronounced (data not shown). The mechanism by which dG affects the production of DSBs in replicating DNA is not yet clear, but at concentrations of 100 gM or greater it is likely to be a consequence of S phase inhibition and the reduced rate of cell proliferation (Figs 1 and 5). In addition, there is a marked dG concentration-dependent increase in the number of nascent (i.e. nascent at the time of CPT treatment) DNA DSBs 24 h after CPT removal, 4.5 gray equivalents in the presence of 25 gM dG, 6 and 15 gray equivalents with 50 and 100 gM dG, respectively. In the absence of dG there is a steady decrease in the number of nascent DNA DSBs during the 24 h following CPT treatment. In the presence of dG, however, it is clear that, despite the reduction in initial break numbers, these lesions are more slowly repaired. Most striking is the reappearance of DSBs in the 100 gM dG sample between 5 and 24 h, indicating the continued susceptibility of the camptothecin-targeted nascent DNA to further damage at a later time (Fig. 7). Increasing doses of dG lead to increased FAR of the [14C]TdR-labelled bulk DNA at 24 h post CPT treatment. The origin of these DSBs is at the present time unknown, but they may represent [14C]TdR-labelled parental strand DNA associated with the [3H]TdR-pulsed labelled, or inter-replicon [14C]TdR-labelled DNA released due to the DSBs induced in adjacent replicons. However, the possibility cannot be ruled out that some of these DSBs may be a product of partial DNA degradation in dying cells, although this is not apparent on standard agarose gel electrophoresis (H. L. Strutt, unpublished). If at all, the contribution of DSBs that originated in the bulk DNA due to DNA degradation is only minor, and the majority of the DSBs are in the DNA that was replicating at the time of CPT exposure.

We have shown that aphidicolin, an inhibitor of DNA polymerases alpha and delta, prevents the generation of DSBs and thus protects the cells from the lethal effects of CPT (Ryan et al. 1991). Despite the finding that deoxyguanosine reduces the initial level of CPT-induced DSBs, it does not abolish aphidicolin protection (data not shown), thus indicating that dG does not by-pass the initial step of DSB induction by CPT.

In the present study we demonstrate that while campto-thecin induces similar levels of protein-linked singlestrand DNA breaks throughout the cycle it kills predominantly S phase cells. The S phase specificity of camptothecin cytotoxicity implies that it is the arrest of topoisomerase I activity involved in replication rather than in transcription that is responsible for the generation of the critical cytotoxic lesions, and that these probably lie at or close to the replication fork (Kessel et al. 1972; Li et al. 1972; Horwitz and Horwitz, 1973; Hsiang et al. 1989). Studies on the inhibitory effect of camptothecin on SV40 DNA replication suggest that when the drug traps topoisomerase I in a cleavable complex at the replication fork one likely outcome is a double-strand DNA break (Snapka, 1986; Avemann et al. 1988; Hsiang et al. 1989).

Using pulse-field gel electrophoresis to detect low levels of DNA fragmentation (Ager et al. 1990), we have demonstrated (Ryan et al. 1991 and in this paper) that camptothecin induces double-strand DNA breaks almost exclusively in newly synthesized DNA, and that a fraction of these breaks persist for up to 24 h. These long-lived double-strand breaks are, therefore, likely to represent major cytotoxic lesions. There is much evidence for the cytotoxicity of double-strand DNA breaks (Blocher and Pohlit, 1982). Examples can be found in yeast and in mammalian cell mutants deficient in double-strand break repair; both sets of mutants are highly sensitive to cell killing by X-rays (Frankenberg et al. 1981; Resnick and Martin, 1976; Jeggo and Kemp, 1983).

Deoxyguanosine, and to a much lesser extent guanosine, augment camptothecin cytotoxicity. The effect is specific to the guanine nucleosides but they have no effect on the action of a topoisomerase II inhibitor. The enhancement of killing by camptothecin is reached at concentrations of dG that affect neither cell proliferation nor the level of the protein-linked single-strand DNA breaks. But these lower concentrations of dG do reduce the rate of recovery of DNA synthesis after camptothecin removal, and they also increase the selective S phase killing by the drug. These results presumably reflect the existence of additional potentially cytotoxic lesions.

The present work provides evidence for the way in which deoxyguanosine exaggerates the lethal effect of camptothecin on S phase cells. We show that in the presence of dG there is a slower rate of DSB repair during the 24 h period after the removal of CPT. As with CPT alone these breaks are induced almost exclusively in the fraction of DNA that was replicating at the time of drug treatment. The higher level of DSBs associated with dG is not the result of more breaks being generated by the co-treatment; on the contrary, deoxyguanosine reduces the initial level of DSBs generated in nascent DNA during drug treatment. We conclude therefore that the enhancement of CPT cytotoxicity by dG is mediated by the higher level of persistent replication-associated DSBs. Aphidicolin, an inhibitor of replication fork polymerases, abolishes CPT toxicity (Holm et al. 1989; Hsiang et al. 1989), and it was suggested by Hsiang et al. that CPT toxicity involves an interaction between the replication machinery and a drug-mediated topoisomerase I-DNA cleavable complex. We have recently shown that one outcome of this interaction is a replication-associated DSB, since both camptothecin-induced DSBs and cytotoxicity are prevented by aphidicolin (Ryan et al. 1991). Deoxyguanosine does not prevent the aphidicolin protection of CPT toxicity (data not shown), suggesting that DSBs are generated by the same mechanism. The exacerbation by dG of camptothecin’s damaging effect on replicating DNA is clear, but the mechanism is not. Preventing the repair of X-ray-induced DSBs by ara A (9-β-D-arabinofuranosyladenine) results in greater cell killing (Bryant and Illiakis, 1984). By analogy we suggest that in the presence of dG the rate of repair of the DSBs is slower, leading to a higher level of persistent S phase-dependent DSBs at the replication fork, which in turn results in enhanced cell killing.

Little is known about the regulation of topoisomerase I activity in vivo and our results with dG are unexpected. Deoxyguanosine alone, at low concentrations, is cytotoxic to human lymphocytes taken from individuals with an inborn deficiency of purine nucleoside phosphorylase (PNP) (Cohen et al. 1978) and also to fibroblasts taken from individuals with the hereditary disease Cockayne’s syndrome (Squires et al. 1990, and unpublished data). dG cytotoxicity to human and mouse PNP-deficient mutants was thought to be associated with feedback inhibition of the ribonucleotide reductase by dGTP, the accumulated product (Ullman et al. 1979). But dG at higher, though still physiological concentrations, was also found to be toxic to normal human T and B lymphocytes and to mouse lymphoma cells, via the accumulation of guanine ribonucleotides and not dGTP, indicating that the inhibition of ribonucleotide reductase by dGTP may not be the primary target for dG toxicity (Scharenberg et al. 1985; Duan et al. 1990). The mechanism of dG toxicity is unclear and Duan et al. (1990) have suggested that a possible early target of dG toxicity is RNA transcription, though they show that the inhibition of RNA synthesis by dG is not related to the inhibition of RNA polymerase at least in a cell-free homogenate. In the present work we have shown that higher concentrations of dG, which partially inhibit DNA synthesis and lead to the accumulation of SV40MRC5 cells in early S phase, almost double the number of protein-linked single-strand DNA breaks induced by CPT and inhibit the recovery from the S-G2 cell cycle block after drug removal. These results suggest that dG has a direct effect on the cellular activity of topoisomerase I, either activating it in vivo, or causing a change in the cleavage sequence specificity of topoisomerase I leading to more DNA breakage (Christiansen et al. 1987; Champoux and Aronoff, 1989). In vitro topoisomerase I activity can be up-regulated by phosphorylation or down-regulated by ADP-ribosylation (Durban et al. 1983; Ferro and Olivera, 1984). Perhaps dG affects a component that is involved in the repair of the DSBs generated at the replication sites by camptothecin. Alternatively, nucleoside triphosphate analogues have been shown to inhibit the nicking step of topoisomerases (type I and II) in vitro and if topoisomerases have nucleotide-binding sites (Liu et al. 1989), it is possible that dGTP or a derivative is involved in topoisomerase I regulation.

Deoxyguanosine is not alone in showing a synergistic cytotoxicity and elevated levels of single-strand DNA break induction in combination with inhibitors of topoisomerases. For example, 3-aminobenzamide, an inhibitor of poly(ADP-ribose)polymerase, and tumor necrosis factor have both been reported to act in this way with both classes of topoisomerase (Mattern et al. 1987; Utsugi et al. 1990). Other agents, such as caffeine or beta-Lapachone, also appear to enhance cell killing by increasing the abundance of double-strand breaks in damaged DNA (Roberts and Kotsaki-Kovatsi, 1986; Musk et al. 1990; Boothman and Pardee, 1989), and Boothman and Pardee (1989) have suggested that beta-Lapachone inhibits DNA repair of X-ray damage and radiation-induced neoplastic transformation through activation of DNA topoisomerase type I. Our results suggest that deoxyguanosine might be used to enhance the antitumor action of very low concentrations of camptothecin.

We are grateful to the Cancer Research Campaign of which R.T.J. is a Research Fellow, for their support; to Dr S. Bouffier for helpful suggestions on the manuscript and to Mrs J. Northfield and Mrs P. Pawley for excellent assistance.

Ager
,
D. D.
,
Dewey
,
W. C.
,
Gardiner
,
K.
,
Harvey
,
W.
,
Johnson
,
R. T.
and
Waldren
,
C. A.
(
1990
).
Measurement of radiation-induced DNA double-strand breaks by pulsed-field gel electrophoresis
.
Radiat Res
.
122
,
181
187
.
Avemann
,
K.
,
Knippers
,
R.
,
Koller
,
T.
and
Sogo
,
J. M.
(
1988
).
Camptothecin, a specific inhibitor of type I DNA topoisomerase, induces DNA breakage at replication forks
.
Molec. cell. Biol
8
,
3026
3034
.
Blocher
,
D.
and
Pohlit
,
W.
(
1982
).
DNA double-strand breaks in Ehrlich ascites tumour cells at low doses of X rays II. Can cell death be attributed to double-strand breaks
.
Int J. radiat. Biol
.
42
,
329
338
.
Boothman
,
D. A.
and
Pardee
,
A. B.
(
1989
).
Inhibition of radiation-induced neoplastic transformation by /2-lapachone
.
Proc. natn. Acad. Sci. U S.A
.
86
,
4963
4967
.
Bryant
,
P. E.
and
Iliakis
,
G.
(
1984
).
Possible correlations between cell killing, chromosome damage and DNA repair after X-irradiation
.
Nucl. Acids Symp. Ser
.
13
,
291
308
.
Champoux
,
J. J.
and
Aronoff
,
R.
(
1989
).
The effects of Camptothecin on the reaction and the specificity of the wheat germ Type I Topoisomerase
.
J. biol. Chem
.
264
,
1010
1015
.
Christiansen
,
K.
,
Bonven
,
B J
and
Westergaard
,
O.
(
1987
).
Mapping of sequence-specific chromatin proteins by a novel method; Topoisomerase I on Tetrahymena ribosomal chromatin
J. molec. Biol
.
193
,
517
525
.
Christman
,
M. F.
,
Dietrich
,
F. S.
and
Fink
,
G. R.
(
1988
).
Mitotic recombination in the rDNA of S. cerevisiae is suppressed by combined action of DNA topoisomerase I and II
.
Cell
55
,
413
425
.
Cohen
,
A.
,
Guda
,
L. J.
,
Ammann
,
A. J.
,
Staal
,
G. E.
and
Martin
,
D. W.
Jr
(
1978
).
Deoxyguanosine triphosphate as a possible toxic metabolite in the immunodeficiency associated with purine nucleoside phosphorylase deficiency
.
J. chn. Invest
.
61
,
1405
1409
.
Covey
,
J. M
,
Jaxel
,
C.
,
Kohn
,
K. W.
and
Pommier
,
Y
(
1989
)
Protein-linked DNA strand breaks induced in mammalian cells by CPT an inhibitor of topoisomerase I
.
Cancer Res
.
49
,
5016
5022
.
Culotta
,
V
and
Sollner-Webb
,
B.
(
1988
).
Sites of topoisomerase I action on X. laevis ribosomal chromatin. Transcriptionally active rDNA has a 200 bp repeating structure
.
Cell
52
,
585
597
.
Denko
,
N.
,
Giaccia
,
A.
,
Peters
,
B.
and
Stamato
,
T. D.
(
1989
).
An asymmetric field inversion gel electrophoresis method for the separation of large DNA molecules
.
Analyt. Biochem
.
178
,
172
176
.
Duan
,
D.-S.
,
Nagashima
,
T.
,
Hoshino
,
T.
,
Waldman
,
F.
,
Pawlak
,
K.
and
Sadee
,
W.
(
1990
).
Mechanism of 2’-deoxyguanosine toxicity in mouse T-lymphoma cells with purine nucleoside phosphorylase deficiency and resistance to inhibition of ribonucleotide reductase by dGTP
.
Biochem. J
.
268
,
725
731
.
Durban
,
E.
,
Mills
,
J S.
,
Roll
,
D
and
Busch
,
H.
(
1983
).
Phosphorylation of purified Novikoff hepatoma topoisomerase I
.
Biochem biophys. Res. Commun
.
111
,
897
905
.
Eng
,
W.-K.
,
Faucette
,
L.
,
Johnson
,
R K.
and
Sternglanz
,
R.
(
1988
).
Evidence that DNA Topoisomerase I is necessary for the cytotoxic effects of Camptothecin
.
Molec Pharm
.
34
,
755
760
.
Ferro
,
A. M.
and
Olivera
,
B. M.
(
1984
).
Poly (ADP-ribosylation) of DNA topoisomerase I from calf thymus
.
J. biol. Chem
.
259
,
547
554
.
Frankenberg
,
D.
,
Frankenberg-Schwager
,
M.
,
Blocher
,
D
and
Harbich
,
R.
(
1981
).
Evidence of DNA double-strand breaks as the critical lesions in yeast cells irradiated with sparsely or densely ionizing radiation under oxic or anoxic conditions
.
Radiat. Res
.
88
,
524
532
.
Gilmour
,
D. S.
and
Elgin
,
S. C. R.
(
1987
).
Localization of specific Topoisomerase I interactions within the transcribed region of active heat shock genes by using the inhibitor camptothecin
.
Molec. cell. Biol
.
7
,
141
148
.
Heck
,
M. M. S.
,
Hittelman
,
W. N.
and
Earnshaw
,
W. C.
(
1988
).
Differential expression of DNA topoisomerases I and II during the eukaryotic cell cycle
.
Proc. natn. Acad. Sci. U.S.A
.
85
,
1086
1090
.
Holm
,
C.
,
Covey
,
J. M.
,
Kerrigan
,
D.
and
Pommier
,
Y.
(
1989
).
Differential requirement of DNA replication for the cytotoxicity of DNA topoisomerase I and II inhibitors in Chinese hamster DC3F cells
.
Cancer Res.
49
,
6365
6368
.
Horwitz
,
S. B.
and
Horwitz
,
M. S.
(
1973
).
Effects of Camptothecin on the breakage and repair of DNA during the cell cycle
.
Cancer Res
.
33
,
2834
2836
.
Hsiang
,
Y.-H.
,
Hertzberg
,
R
,
Hecht
,
S
and
Liu
,
L. F.
(
1985
).
Camptothecin induces protein-linked DNA breaks via mammalian DNA Topoisomerase I
.
J. biol. Chem
.
260
,
14873
14878
.
Hsiang
,
Y.-H.
,
Lihou
,
M. G.
and
Liu
,
L. F.
(
1989
).
Arrest of replication forks by drug-stabilized topoisomerase I-DNA cleavable complexes as a mechanism of cell killing by camptothecin
.
Cancer Res
.
49
,
5077
5082
.
Hsiang
,
Y.-H.
and
Liu
,
L. F.
(
1988
).
Identification of mammalian DNA topoisomerase I as an intracellular target of the anticancer drug Camptothecin
.
Cancer Res
.
48
,
1722
1726
.
Iliakis
,
G. E.
,
Cicilioni
,
O.
and
Metzger
,
L.
(
1991b
).
Measurement of DNA double-strand breaks in CHO cells at various stages of the cell cycle using pulsed field gel electrophoresis: calibration by means of 1”I decay
.
Int. J. radiat. Biol
.
59
,
343
357
.
Iliakis
,
G. E.
,
Metzger
,
L.
,
Denko
,
N.
and
Stamato
,
T. D.
(
1991a
).
Detection of DNA double-strand breaks in synchronous cultures of CHO cells by means of asymmetric field inversion gel electrophoresis
.
Int. J. radiat. Biol
.
59
,
321
341
.
Jeggo
,
P. A.
and
Kemp
,
L. M.
(
1983
).
X ray-sensitive mutants of Chinese hamster ovary cell line: Isolation and cross-sensitivity to other DNA-damaging agents
.
Mutât. Res
.
112
,
313
327
.
Kjeldsen
,
E.
,
Bonven
,
B. J.
,
Andoh
,
T.
,
Ishii
,
K.
,
Okada
,
K.
,
Bolund
,
L.
and
Westergaard
,
O.
(
1988
).
Characterization of a Camptothecin-resistant human DNA Topoisomerase I
.
J. biol. Chem
.
263
,
3912
3916
.
Kessel
,
D.
,
Bosmann
,
H. B.
and
Lohr
,
K.
(
1972
).
Camptothecin effects on DNA synthesis in munne leukemia cells
.
Biochim. biophys. Acta
269
,
210
216
.
Kroeger
,
P. E.
and
Rowe
,
T. C.
(
1989
).
Interaction of topoisomerase I with the transcribed region of the Drosophila HSP 70 heat shock gene
Nucl. Acid Res
17
,
8495
8509
.
Li
,
L. H.
,
Fraser
,
T. J.
,
Olin
,
E. J.
and
Bhuyan
,
B. K.
(
1972
).
Action of camptothecin on mammalian cells in culture
.
Cancer Res
.
32
,
2643
2650
.
Liu
,
L. F.
(
1989
).
DNA topoisomerase poisons as antitumor drugs
.
A. Rev. Biochem
.
58
,
351
375
.
Liu
,
S.-Y.
,
Hwang
,
B.-D.
,
Liu
,
Z.-C.
and
Cheng
,
Y.-G.
(
1989
).
Interaction of several nucleoside triphosphate analogues and 10-Hydroxycamptothecin with human DNA topoisomerases
.
Cancer Res
.
49
,
1366
1370
.
Mattern
,
M. R.
,
Mong
,
S.-M.
,
Bartus
,
H. F.
,
Mirabelli
,
C. K.
,
Crooke
,
S. T.
and
Johnson
,
R. K.
(
1987
).
Relationship between the intracellular effects of camptothecin and the inhibition of DNA toposomerase I in cultured L1210 cells
.
Cancer Res
.
47
,
1793
1798
.
Musk
,
S. R. R.
,
Pillidge
,
L
,
Johnson
,
R. T.
and
Downes
,
C. S.
(
1990
).
Action of caffeine on DNA replication after ultraviolet irradiation in Indian muntjac cells, no connection with action on cell cycle delay
.
Biochim. biophys. Acta
1052
,
53
62
.
Resnick
,
M. A.
and
Martin
,
P.
(
1976
).
The repair of double-strand breaks in the nuclear DNA of Saccharomyces cerevisiae and its genetic control
.
Molec. gen. Genet
.
143
,
119
129
.
Roberts
,
J. J.
and
Kotsaki-Kovatsi
,
V. P.
(
1986
).
Potentiation of sulphur mustard or cisplation-induced toxicity by caffeine in Chinese hamster cells correlates with formation of DNA double-strand breaks during replication on a damaged template
.
Mutât. Res
.
165
,
207
220
.
Rowe
,
T. C.
,
Chen
,
G. L.
,
Hsiang
,
Y.-H.
and
Liu
,
L. F.
(
1986
)
DNA damage by antitumor acridines mediated by mammalian DNA Topoisomerase II
.
Cancer Res
.
46
,
2021
2026
.
Ryan
,
A. J.
,
Squires
,
S.
,
Strutt
,
H.
and
Johnson
,
R. T.
(
1991
)
Camptothecin cytotoxicity in mammalian cells is associated with the induction of persistent double strand breaks in replicating DNA
.
Nucl. Acids Res
.
19
,
3295
3300
.
Scharenberg
,
J. G. M.
,
Spaapen
,
L. J. M.
,
Rijkers
,
G. T.
,
Wadman
,
S. K.
,
Staal
,
G. E. J.
and
Zegers
,
B. J. M.
(
1985
).
Mechanisms of deoxyguanosine toxicity for human T and B lymphocytes
.
Adv. exp. Med. Biol
.
195
,
191
199
.
Smith
,
P. J.
,
Anderson
,
C. O.
and
Watson
,
J. V.
(
1985
).
Abnormal retension of X-irradiated ataxia telangiectasia fibroblasts in G2 phase of the cell cycle cellular content, chromatin stability and the effects of 3-aminobenzamide
.
Int. J. radiat. Biol
.
47
,
701
712
.
Smith
,
P. J.
,
Makinson
,
T. A.
and
Watson
,
J. V.
(
1989
).
Enhanced sensitivity to camptothecin in Ataxia-telangiectasia cells and its relationship with the expression of DNA topoisomerase I
.
Int. J. radiat. Biol
.
55
,
217
231
.
Snapka
,
R. M.
(
1986
).
Topoisomerase inhibitors can selectively interfere with different stages of Simian Virus 40 DNA replication
.
Molec. cell. Biol
.
6
,
4221
4227
.
Squires
,
S.
,
Elliott
,
G. C.
and
Johnson
,
R. T.
(
1987
).
Slow DNA rejoining in ultraviolet-irradiated human diploid fibroblasts treated with the mitogens trypsin and insulin
.
Cancer Res.
47
,
4378
4385
.
Squires
,
S.
,
Johnson
,
R. T.
and
Collins
,
A. R. S.
(
1982
).
Initial rates of DNA incision in UV-irradiated human cells. Differences between normal, xeroderma pigmentosum and tumour cells
Mutât. Res
.
95
,
389
404
.
Squires
,
S.
,
Oates
,
D. J.
and
Johnson
,
R. T.
(
1990
).
Cockayne’s Syndrome cells show enhanced sensitivity to deoxyguanosine and camptothecin
.
Mutagenesis
5
,
76
.
Stamato
,
T. D.
and
Denko
,
N.
(
1990
).
Asymmetric field inversion gel electrophoresis: A method for detecting DNA double-strand breaks in mammalian cells
.
Radiat. Res
.
121
,
196
205
.
Stewart
,
A. F.
and
Schutz
,
G.
(
1987
).
Camptothecin-induced tn vivo Topoisomerase I cleavages in the transcriptionally active tyrosine aminotransferase gene
.
Cell
50
,
1109
1117
Ullman
,
B.
,
Gudas
,
L. J.
,
Clift
,
S. M.
and
Mortin
,
D. W.
(
1979
).
Isolation and characterization of purine-nucleoside phosphorylase-deficient T-lymphoma cells and secondary mutants with altered immunodeficiency disease
.
Proc. natn. Acad. Sci. U.S.A
.
76
,
1074
1078
.
Utsugi
,
T.
,
Mattern
,
M. R.
,
Mirabelli
,
C. K.
and
Hanna
,
N.
(
1990
).
Potentiation of topoisomerase inhibitor-induced DNA strand breakage and cytotoxicity by tumor necrosis factor: enhancement of topoisomerase activity as a mechanism of potentiation
.
Cancer Res
.
50
,
2636
2640
.
Wang
,
J. C.
(
1985
).
DNA Topoisomerases
.
A. Rev. Biochem
.
54
,
665
697
.
Wang
,
J. C
(
1987
).
Recent studies of DNA topoisomerases
.
Biochim. biophys Acta
909
,
1
9
.
Watson
,
J. V.
,
Chambers
,
S. H.
and
Smith
,
P. J.
(
1987
).
A pragmatic approach to analysis of DNA histograms with a definable GI peak
.
Cytometry
8
,
1
8
.
Yanagida
,
M.
and
Wang
,
J. C.
(
1987
).
Yeast DNA topoisomerase and their structural genes
.
Nucl. Acids molec. Biol
.
1
,
196
209
.
Zhang
,
H.
,
Wang
,
J. C.
and
Liu
,
L F.
(
1988
).
Involvement of DNA topoisomerase I in transcription of human ribosomal RNA genes
.
Proc. natn. Acad Sci. U.S.A
.
85
,
1060
1064
.