- Email: [email protected]
Cisplatin-modification of DNA repair and ionizing radiation lethality in yeast, Saccharomyces cerevisiae
Mutation Research 433 Ž1999. 127–136
Cisplatin-modification of DNA repair and ionizing radiation lethality in yeast, Saccharomyces cereÕisiae J.-A. Dolling a
a,)
, D.R. Boreham a,1, D.L. Brown R.E.J. Mitchel a,4
b,2
, G.P. Raaphorst
c,3
,
Radiation Biology and Health Physics Branch, AECL, Chalk RiÕer, Ontario, Canada K0J 1J0 b Department of Biology, UniÕersity of Ottawa, Ottawa, Ontario Canada, K1N 6N5 c Ottawa Regional Cancer Centre, Ottawa, Ontario, Canada, K1H 8L6 Received 24 June 1998; revised 4 December 1998; accepted 11 December 1998
Abstract Cis-diamminedichloroplatinum II Žcisplatin. is a DNA inter- and intrastrand crosslinking agent which can sensitize prokaryotic and eukaryotic cells to killing by ionizing radiation. The mechanism of radiosensitization is unknown but may involve cisplatin inhibition of repair of DNA damage caused by radiation. Repair proficient wild type and repair deficient Ž rad52, recombinational repair or rad3, excision repair. strains of the yeast Saccharomyces cereÕisiae were used to determine whether defects in DNA repair mechanisms would modify the radiosensitizing effect of cisplatin. We report that cisplatin exposure could sensitize yeast cells with a competent recombinational repair mechanism Žwild type or rad3 ., but could not sensitize cells defective in recombinational repair Ž rad52 ., indicating that the radiosensitizing effect of cisplatin was due to inhibition of DNA repair processes involving error free RAD52-dependent recombinational repair. The presence or absence of oxygen during irradiation did not alter this radiosensitization. Consistent with this result, cisplatin did not sensitize cells to mutation that results from lesion processing by an error prone DNA repair system. However, under certain circumstances, cisplatin exposure did not cause radiosensitization to killing by radiation in repair competent wild type cells. Within 2 h after a sublethal cisplatin treatment, wild type yeast cells became both thermally tolerant and radiation resistant. Cisplatin pretreatment also suppressed mutations caused by exposure to N-methyl-N X-nitro-N-nitrosoguanidine ŽMNNG., a response previously shown in wild type yeast cells following radiation pretreatment. Like radiation, the cisplatin-induced stress response did not confer radiation resistance or suppress MNNG mutations in a recombinational repair deficient mutant Ž rad52 ., although thermal tolerance was still induced. These results support the idea that cisplatin adducts in DNA interfere with RAD52-dependent recombinational repair and thereby sensitize cells to killing by radiation. However, the lesions can subsequently induce a general stress response, part of which is induction of RAD52-dependent error free recombinational
) 1 2 3 4
Corresponding author. Tel.: q1-613-584-3311; Fax: q1-613-584-1713 Tel.: q1-613-584-3311; Fax: q1-613-584-1713. Tel.: q1-613-562-5800; Fax: q1-613-562-5486. Tel.: q1-613-737-7700; Fax: q1-613-247-6811. Tel.: q1-613-562-5800; Fax: q1-613-562-5486.
0921-8777r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 8 7 7 7 Ž 9 8 . 0 0 0 6 9 - X
128
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
repair. This stress response confers radiation resistance, thermal tolerance, and mutation resistance in yeast. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Cisplatin; Ionizing radiation; Radiosensitization; Adaptive response; Yeast DNA repair
1. Introduction A synergistic cell killing effect of combined cisplatin Ža DNA inter- and intrastrand crosslinking agent. and ionizing radiation treatments has been observed in both prokaryotic w1,2x and eukaryotic w3–10x systems, and this effect has been used in cancer therapy. The exact mechanism of radiosensitization by cisplatin has not been established, but it has been postulated that the supra-additive killing effect of cisplatin and radiation is a consequence of cisplatin inhibition of repair of radiation-induced DNA damage. Two DNA repair pathways that are known to respond to radiation-induced DNA damage, and have been extensively characterized in yeast, are excision repair and recombinational repair w11x. The genes involved in excision and recombinational repair pathways belong to the RAD3 and the RAD52 epistasis groups, respectively. rad3 yeast cells are UVlight sensitive and unable to remove pyrimidine dimers and other bulky adducts from their DNA w11,12x, whereas rad52 cells are sensitive to ionizing radiation and deficient in meiotic and mitotic recombination. rad52 yeast cells are unable to repair DNA double strand breaks, resulting in an extremely ionizing radiation sensitive phenotype w13x. The experiments in this work were designed to investigate the involvement of recombination and excision repair processes in the mechanism of radiosensitization by cisplatin, by determining whether defects in repair processes could modify the radiosensitizing effects of cisplatin. While cisplatin DNA lesions may inhibit repair and consequently radiosensitize cells, under certain conditions, these DNA lesions may also signal the induction of an adaptive response whereby cellular repair capacity is enhanced and radiation resistance is increased. In prokaryotes and eukaryotes, DNA repair capacity can be elevated in response to a variety of stresses including ionizing radiation w14x, heat shock w14,15x, mutagenic chemicals w16x and chemotherapeutic drugs w17x. In yeast, it has been
shown that different stresses induce specific sets of genes w17–20x. A critical set of genes induced in yeast by DNA damaging agents is the RAD50-57 gene series w21x. These proteins form the complex machinery w22x induced during the adaptive response to radiation and used to repair DNA lesions via an error free recombinational process w16,21x. Recombinational repair is essential for successful repair of lesions involving both DNA strands, since the damage in either strand cannot be resynthesized using the genetic information encoded by the damaged complementary strand. Instead, the damaged or missing genetic information is copied from a duplicate source Ži.e., homologous chromosome or sister chromatid. w23–25x. Lesions requiring the recombinational repair system include double strand breaks and interstrand crosslinks. Recombinational repair is not, however, restricted to repair of these relatively complex lesions. For example, it has been shown that recombinational repair can restitute chemically generated N-methyl-N X-nitro-N-nitrosoguanidine ŽMNNG. DNA lesions w16,26x. MNNG is a chemical mutagen that methylates DNA at various sites including the N-7 and O-6 positions of guanine w27x. If the methylated guanine is not repaired, it pairs incorrectly with thymine instead of cytosine during replication, resulting in a GC to AT transition w28x. It has been suggested that MNNG exposure induces an error prone system which is responsible for essentially all MNNG-generated mutations w16x. The number of MNNG-generated mutations, however, has been shown to be reduced in a dose dependent manner in yeast cells pre-exposed to ionizing radiation w16,26x or heat w29x but not by pretreatment with MNNG itself w30x. In fact, MNNG induces an error prone system w29x which increases the mutagenic effect of a subsequent treatment with other chemical mutagens like ethyl methanesulfonate w16x. It has been proposed that preexposure of cells to radiation and heat induces an error free recombinational system that competes with the error prone system for repair of these lesions w29x. An increase in cellular error free recombinational repair capacity would lead
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
to a decrease in the number of MNNG lesions processed by the error prone system and result in a net suppression of mutation. It is not known whether DNA crosslinks caused by cisplatin are effective signals for the induction of error free recombinational repair, although in yeast, crosslinks caused by furocoumarins resulted in upregulation of RAD54 transcription w31x. In addition, there is evidence to indicate that the basic signal for induction of recombinational repair is perturbation of DNA topology w32x. Therefore, it is likely that DNA crosslinks which cause base destacking and strand kink angles of 40 to 608 w33x, could Ža. invoke an adaptive response and Žb. lead specifically to an increased error free repair recombinational capacity. In the investigation presented here, the induction of recombinational repair in yeast by cisplatin was tested by pretreatment with cisplatin and subsequent exposure to ionizing radiation, heat or MNNG, and the results were compared with the potentially radiosensitizing effect of cisplatin treatments.
2. Materials and methods 2.1. Strains Diploid strains of the yeast Saccharomyces cereÕisiae were used as the test organisms in these experiments. Strain MS33 was wild type for DNA repair capacity Ž Rec q r Rec q , Exc q r Exc q .. Strain MS32 was homozygous for rad52-1 and was defective in recombinational repair Ž Rec y r Rec y , Exc q r Exc q .. Strain MS33 was the isogenic parental strain of MS32. Both MS33 and MS32 were also defective for the ATP phosphoribosyltransferase gene Ž his1-7r his1-7 .. Strain STX432 was homozygous for rad3-2 and was defective in excision repair Ž Rec q r Rec q , Exc y r Exc y .. Yeast strains MS33 and MS32 were gifts from Dr. D.P. Morrison, AECL, Chalk River Laboratories. Strain STX432 was purchased from the Yeast Genetic Stock Center ŽBerkeley, CA.. 2.2. Cell growth Cell cultures were grown at 238C to mid-exponential phase Ž2–4 = 10 6 cellsrml. in liquid nutrient
129
medium containing 0.1% yeast extract, 1% yeast nitrogen base with amino acids, 1% sodium succinate, and 2% glucose Žyeast extractrsuccinate medium or YES.. 2.3. Cisplatin treatment Cells Ž2 = 10 7 . were treated for 30 min with cisplatin ŽBristol Myers Squibb. diluted to the appropriate concentration in 238C sodium phosphate buffer. After cisplatin treatment, cells were washed three times with 08C buffer. Control cells were suspended in sodium phosphate buffer without cisplatin, for 30 min at 238C and then treated in the same way as cisplatin treated cell samples. Survival was determined by plating suitable dilutions on solid yeastrpeptonerdextrose ŽYPD. agar plates containing 0.1% yeast extract, 1% yeast nitrogen base with amino acids, 1% sodium succinate, 2% peptone, 2% glucose. Colony formation was scored following five days of incubation at 238C. Some cells were tested for induced thermal or radiation resistance, or resistance to mutation as described below. 2.4. Radiation surÕiÕal Yeast cells were washed three times and suspended Žat 08C. in 20 mmol dmy3 ŽmM. sodium phosphate buffer, pH 7.0, at approximately 2 = 10 6 cellrml. These suspensions were equilibrated by bubbling the cell suspension at 08C with O 2 or N2 ŽMatheson ultrahigh purity. for 10 min before irradiation and continuously during irradiation. Cells were irradiated at 08C with 60 Co gamma-rays ŽGamma cell 220, Atomic Energy of Canada. at a dose rate of about 1 Gyrs. Survival was determined by plating suitable dilutions of irradiated and unirradiated control cells on solid YPD agar plates followed by incubation for five days at 238C. 2.5. Thermal surÕiÕal Thermal resistance was determined by placing 2.5 ml cell aliquots Ž2 = 10 6 cellsrml in YES medium at 08C. in 16 = 100 mm glass test tubes and submerging the glass test tubes into a 528C shaking water bath. At specified time intervals, individual tubes were removed and rapidly cooled to 08C. Control cells were held at 08C for the same duration.
130
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
with phosphate buffer, all at 08C. Control cells not exposed to MNNG were treated similarly. Reversion to histidine independence and survival to treatments was determined as described above. 2.7. Data and figures Each experiment was repeated at least three times from independent cultures. All figures shown represent the combined results of all the experiments. Standard error bars are shown except were the error is less than or equal to the symbol size. Student’s t-tests were performed to determine the statistical significance between the various treatments and pvalues are given when the results were significantly different.
3. Results 3.1. Radiation surÕiÕal Fig. 1. Survival of wild type MS33 Žcircles., recombinational repair deficient MS32 Žsquares. and excision repair deficient STX432 Žtriangles. yeast cells after ŽA. 60 Co gamma irradiation delivered in the presence of nitrogen or ŽB. cisplatin treatment at 238C for 30 min.
Wild type MS33, recombinational repair deficient MS32 and excision repair deficient STX432 yeast cells were irradiated with increasing doses of gamma
Treated and control cells were then plated for survival on YPD plates. 2.6. Mutation and suppression of MNNG induced mutation by cisplatin Cisplatin-induced mutation was determined by measuring reversion to histidine independence. Revertant cells able to form colonies were counted after incubation at 238C for 6 days on solid medium lacking histidine. Total survival after the various treatments was measured by scoring colony formation on YPD plates. Chemical mutagenesis was performed at 238C in 20 mmol dmy3 ŽmM. phosphate buffer ŽpH 7.0. at a cell density of 1 = 0 7 cellsrml. Cells were treated with 5 or 20 mgrml of MNNG ŽAldrich Chemical. for 30 min either immediately after or at the same time as a cisplatin treatment. After MNNG exposure, cells were washed once with 10% Žwrv. sodium thiosulphate to inactivate the MNNG and then twice
Fig. 2. Survival of wild type MS33 yeast cells after 60 Co gamma irradiation delivered in the presence of nitrogen Žcircles. or oxygen Ždiamonds., with Žopen symbols. or without Žclosed symbols. a cisplatin treatment at 238C for 30 min immediately before irradiation.
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
131
rays in the presence of N2 ŽFig. 1A.. MS32 cells were sensitive to killing by gamma radiation relative to wild type MS33 cells. STX432 showed an intermediate level of sensitivity at doses greater than 1 kGy. 3.2. Cisplatin surÕiÕal All three yeast strains were tested for survival against increasing concentrations of cisplatin ŽFig. 1B.. The results were similar to those for radiation survival as shown in Fig. 1A, with wild type MS33 demonstrating the most resistance to killing by cisplatin, MS32 demonstrating the least resistance and STX432 showing an intermediate response. 3.3. Combined treatment of cisplatin and radiation Survival of wild type MS33 yeast cells to gamma radiation after a cisplatin treatment of 400 mgrml
Fig. 4. Survival of ŽA. wild type MS33 and ŽB. recombinational repair deficient yeast cells exposed to 60 Co gamma irradiation delivered in the presence of oxygen. MS33 cells were previously exposed to 400 mgrml cisplatin Ž238C, 30 min. followed by a 2-h incubation at 238C. MS32 cells were treated for 30 min at 238C with 50 mgrml Žtriangles. or 400 mgrml Žsquares. cisplatin, incubated for various time intervals and then irradiated Ž0.1 kGy in the presence of oxygen.. Radiation survival of untreated control cells is represented by closed circles.
Fig. 3. Survival of ŽA. wild type MS33, ŽB. recombinational repair deficient MS32 and ŽC. excision repair deficient STX432 yeast cells after 60 Co gamma irradiation delivered in the presence of nitrogen, without Žclosed symbols. or with Žopen symbols. a 30-min cisplatin treatment Žopen circles: 400 mgrml, open triangles: 50 mgrml, open diamonds: 10 mgrml. immediately before irradiation.
for 30 min is shown in Fig. 2. Cisplatin treatment resulted in radiosensitization when the gamma radiation was delivered in oxygen or nitrogen. At a survival level of 10% ŽD 10 . the enhancement ratios were 1.25 and 1.3 for oxic and anoxic irradiation conditions, respectively, and were not significantly different. In Fig. 3, survival of MS33, MS32 and STX432 yeast cells to gamma radiation Ždelivered in the presence of N2 . immediately after a cisplatin treatment. Cisplatin treatment Ž400 mgrml, 30 min. resulted in radiosensitization of wild-type MS33 and excision repair deficient STX432 cells Ž p - 0.01. but not recombinational repair deficient MS32 cells. At a survival level of 10% ŽD10 ., the enhancement ratios for MS33 and STX432 cells Ž1.3 and 1.4 respectively. were not significantly different. Fig. 3B
132
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
time point was similar for both yeast strains. A smaller dose of cisplatin Ž50 mgrml, 30 min. had no effect on thermal survival of either yeast strain. 3.6. Cisplatin suppression of MNNG-generated mutations Wild type MS33 cells showed a cisplatin dose-dependent reduction in mutations generated by MNNG ŽFig. 6A.. Simultaneous treatment of cells with cisplatin and MNNG did not result in reduction of MNNG-generated mutation until the concentrations of cisplatin exceeded 100 mgrml. However, treatment of cells with cisplatin immediately before MNNG exposure resulted in a reduction of mutation at all cisplatin concentrations tested, although the
Fig. 5. Cisplatin-induced thermal tolerance in ŽA. wild type MS33 and ŽB. recombinational repair deficient MS32 yeast cells. Cells were treated with 50 mgrml Žtriangles. or 400 mgrml Žsquares. cisplatin, incubated for various time intervals and then heated at 528C for 3 min. Control cells are represented by circles.
shows that cisplatin, at any concentration tested, did not change the survival of recombinational repair deficient MS32 cells at any dose of radiation tested. 3.4. Cisplatin-induced radiation resistance Incubation of wild type MS33 cells for 2 h at 238C after cisplatin treatment Ž400 mgrml, 238C, 30 min. resulted in elevated radiation resistance Ž p 0.05. ŽFig. 4A.. However, in recombinational repair deficient MS32 cells, radiation resistance was not induced by cisplatin treatment Ž50 or 400 mgrml, 238C, 30 min. at any time interval tested ŽFig. 4B.. 3.5. Cisplatin-induced thermal tolerance Cisplatin exposure Ž400 mgrml, 30 min. induced thermal tolerance in both wild type MS33 ŽFig. 5A. and recombinational repair deficient MS32 cells ŽFig. 5B.. The level of thermal tolerance reached at each
Fig. 6. Cisplatin suppression of MNNG-generated mutations in ŽA. wild type MS33 or ŽB. recombinational repair deficient MS32 yeast cells. MS33 cells were treated simultaneously with cisplatin Ž400 mgrml. and MNNG Ž20 mgrml. at 238C for 30 min Žclosed circles. or sequentially with cisplatin Ž400 mgrml, 238C, 30 min. followed by MNNG Ž20 mgrml, 238C, 30 min. Žopen circles.. MS32 cells were treated with cisplatin alone at 238C for 30 min Žclosed circles. or simultaneously with MNNG Ž5 mgrml. Žopen circles..
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
133
Table 1 Lack of cisplatin suppression of MNNG-generated mutations in recombinational repair deficient MS32 yeast when cells are treated sequentially with cisplatin followed immediately by MNNG Treatment
Number mutationsr 10 6 survivors
Control Žspontaneous. 100 mgrml cisplatin 5 mrml MNNG 100 mgrml cisplatin followed by 5 mgrml MNNG 20 mgrml MNNG 100 mgrml cisplatin followed by 20 mgrml MNNG
15.0 " 1.7 20.4 " .05 23.0 " 3.4 32.5 " 3.0 a 148.0 " 11.2 152.0 " 1.7 a
a
Cisplatin treatment did not suppress MNNG-generated mutation.
reduction was not statistically significant until the concentration exceeded 50 mgrml. At the point of the largest reduction measured Ža 400-mgrml, 30-min pretreatment reduced the number of mutations from 1230 " 100 to 386 " 91 per 10 6 survivors. the mutation rate was still significantly higher than the spontaneous mutation rate of about 4 per 10 6 survivors for wild-type MS33 cells. Under the conditions of the cisplatin treatments used in this study, cisplatin alone did not produce any mutations above the spontaneous level in wild type MS33 cells. In contrast, cisplatin treatment alone produced mutations in recombinational repair deficient MS32 cells ŽFig. 6B, Table 1.. The combined exposures of recombinational repair-deficient MS32 cells to cisplatin and MNNG did not result in a reduction in the number of MNNGgenerated mutations ŽFig. 6B.. This lack of suppression of MNNG-generated mutations was found when the two treatments were given either simultaneously or when cisplatin treatment Ž100 mgrml, 30 min, 238C. immediately preceded the MNNG treatment Ž5 or 20 mgrml, 30 min, 238C. ŽTable 1.. The largest MNNG exposure of MS32 cells Ž20 mgrml, 30 min. generated 148 " 11.2 mutations per 10 6 survivors. This frequency is significantly less than was observed for MS33 cells exposed to the same dose Ž1230 " 108 mutations per 10 6 survivors, Fig. 6A. Ž p - 0.001..
4. Discussion The results presented in this study show that cisplatin used in combination with ionizing radiation
can produce differential cellular responses depending upon the inherent repair competence of the cell and the time interval between cisplatin and radiation exposures. DNA double strand break repair, via RAD52-dependent recombinational repair, is known to be the major pathway which confers radiation resistance in yeast w13x and our MS32 strain, defective in RAD52-dependent recombinational repair, displayed the expected radiosensitivity compared to wild type MS33 cells ŽFig. 1A.. The radiation sensitivity of cells deficient for the RAD3 excision repair protein Žstrain STX432. ŽFig. 1A. showed that excision repair may also be an important pathway which repairs ionizing radiation damage. Radiosensitivity of another strain defective for RAD3 Žstrain D7-3. has been previously demonstrated ŽR.E.J. Mitchel, unpublished data., but in both studies the isogenic wild type parent strain was not available to make direct comparisons. Therefore, it remains possible that the relative radiosensitivity of both excision repair deficient mutants may be due to some other intrinsic characteristic of the strains. Recombination and excision repair processes are involved in repairing cisplatin DNA adducts as previously shown w34x and demonstrated here by the sensitivity to killing by cisplatin in recombinational repair deficient MS32 and excision repair deficient STX432 cells ŽFig. 1B.. The extreme sensitivity of recombinational repair deficient MS32 cells to both radiation and cisplatin indicates that, in yeast cells, recombinational repair involving proteins in the RAD52 gene series ŽRad51, 54, 57. w22x is a dominant mode of repair. Radiation-induced DNA double strand breaks and cisplatin-induced DNA interstrand
134
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
crosslinks are two lesions that are most likely repaired via a recombinational process, since in both cases correct genetic information from a duplicate source Že.g., homologous chromosome or sister chromatid. is required for successfully restoring the damaged sites. Radiosensitization by cisplatin under oxic and anoxic conditions ŽFig. 2. has been previously reported for bacterial and mammalian systems w1,2,5– 7,9,35x. However, several studies have reported that the synergistic effect for cell killing is diminished or absent when cells are irradiated in air w1,2,8x. Results of the present study show similar radiosensitization under both conditions and indicate that, in yeast, modification of lesions by oxygen does not influence radiosensitization by cisplatin. In the present study, excision repair deficient STX432 cells and wild type MS33 cells, but not recombinational repair deficient MS32 cells, were radiosensitized by cisplatin indicating that radiosensitization by cisplatin is absolutely dependent upon functional recombinational repair. Whereas cisplatin treatment immediately before irradiation led to radiosensitization in wild type MS33 yeast cells ŽFig. 2., a similar exposure followed by a 2 h incubation under growth permissive conditions resulted in increased resistance to killing by radiation ŽFig. 4A.. This phenomenon of induced radiation resistance has been well documented in yeast. It has been shown that cellular recombinational repair capacity can be enhanced by exposure to a variety of stresses including radiation w15,16,32x. The magnitude of the increase in radiation resistance is dependent upon the severity of the inducing stress and the time interval between the inducing and challenge exposures Žreviewed in Ref. w36x.. Since induced radiation resistance is absolutely dependent upon recombinational repair capacity w29x, recombinational repair deficient MS32 cells do not exhibit a radiation resistant phenotype after a cisplatin treatment, regardless of the incubation interval between cisplatin and radiation exposures. Cisplatin adducts also signalled induction of thermal tolerance, to the same extent in both wild type MS33 and recombinational repair deficient MS32 yeast cells ŽFig. 6B., indicating that for induced thermal tolerance, the signalling mechanism for induction by cisplatin does not require RAD52. This has previously been shown for induction of thermal
tolerance by heat and radiation w14x. The induction of both radiation and thermal resistance suggests that cisplatin lesions are inducing a general stress response. The interpretation that cisplatin adducts are capable of invoking recombinational repair is further supported by the data indicating suppression of MNNG-generated mutations when wild type MS33 ŽFig. 6A., but not recombinational repair deficient MS32 cells ŽFig. 6B, Table 1., were treated with cisplatin. Suppression of MNNG-generated mutations is believed to be the result of competition between the induced error free recombinational repair system and the error prone repair system w16,29x. MNNG lesions induce the error prone system and cause mutations. If cisplatin adducts signal induction of an error free repair system that is capable of repairing both cisplatin and MNNG lesions, then fewer MNNG-generated mutations would result. Indeed, in this study, increasing doses of cisplatin resulted in more MNNG lesions being processed by the error free repair pathway and a reduction in the frequency of MNNG-generated mutations. In wild type MS33 cells, both cisplatinrMNNG treatment regimes Žsimultaneous or sequential treatment of cisplatin and MNNG. resulted in a reduction in the number of mutations, indicating that cisplatin adducts had caused upregulation of recombinational repair capacity. At the highest inducing dose of cisplatin, the number of residual mutations was still significantly higher than the spontaneous level of 4r10 6 MS33 survivors. This result indicates that induction of the error free system was paralleled by a relative reduction, but not elimination, in the number of lesions processed by the error prone system. Therefore, it is likely that even at maximal induction of the error free system, some MNNG lesions will be repaired by the error prone system w16x. Differences in the kinetics and magnitude of mutation suppression between treatment regimes may reflect differences in the ability to induce recombinational repair by the different spectra of cisplatin adducts present at the time of MNNG treatment. After short cisplatin treatments Ž15 or 30 min., the ratio of mono- to bifunctional adducts is approximately 50:50. During longer incubation times Ž60 min., the ratio changes to 30:70, mono- to bifunctional adducts, and at longer incubation times Ž2 h or
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
more. monofunctional adducts constitute only a few percent of the total cisplatin adducts w33,37x. When cisplatin and MNNG were delivered simultaneously, induction of an error free system was therefore attributed to either cisplatin monofuctional or bifunctional adducts. However, when MNNG was delivered after the cisplatin treatment, there was more time for conversion of cisplatin mono- to bifunctional adducts and it is possible that the presence of bifunctional adducts resulted in a more effective induction of error free recombinational repair capacity. Further, since bifunctional, but not monofuctional, cisplatin adducts produce dramatic topological perturbations in the DNA helix by creating kink angles of 40 to 608 w33x, the observed responses may be due to differences in DNA topological perturbation. In the present study, during the 30 min cisplatin exposure and subsequent 30 min MNNG treatment, some cisplatin monofunctional adducts, which do not appear to distort the DNA helix w33x, would have been converted into helix distorting, bifunctional adducts, and may have resulted in more effective signalling of an error free recombinational repair system. These perturbations in DNA topology may serve as the signal for induction of a radiation resistance mechanism as has been suggested by a study comparing topoisomerase mutants to wild type yeast cells for induction of radiation resistance w26x. It should be noted that cisplatin adducts present simultaneously with MNNG lesions did not increase the mutation frequency ŽFig. 6A. unlike the effect on lethality from radiation exposure. Since cisplatin adducts sensitize cells to killing by radiation but not to mutation by MNNG, we suggest that cisplatin adducts do not impact on the error prone repair system which recognizes MNNG lesions. In summary, radiosensitization by cisplatin depends absolutely upon a functional RAD52-dependent recombinational repair system but not on a functional excision repair system. The data demonstrate that radiosensitization by cisplatin DNA adducts results from inhibition of recombinational repair. There was no significant effect of oxygen on radiosensitization. The results also indicate that cisplatin adducts can signal the induction of an error free recombinational repair process which is dependent upon RAD52 function and that cisplatin bifunctional adducts are better inducers than monofunc-
135
tional adducts. Induction of this error free system in wild type cells is part of a general stress response which resulted in resistance to killing by radiation and heat, and mutation caused by MNNG.
Acknowledgements The authors would like to thank M.-E. Bahen for expert technical assistance and D.P. Morrison for generously supplying the MS32 and MS33 yeast strains and insights into yeast genetics. This work was supported by the CANDU Owners Group and the Ottawa Regional Cancer Centre.
References w1x R.C. Richmond, E.L. Powers, Radiation sensitization of bacterial spores by cis-dichlorodiammineplatinumŽII., Radiat. Res. 68 Ž1976. 251–257. w2x R.C. Richmond, J.D. Zimbrick, D.L. Hykes, Radiation-induced DNA damage and lethality in E. Coli as modified by the antitumor agent cis-dichlorodiammineplatinumŽII., Radiat. Res. 71 Ž1977. 447–460. w3x I. Wodinsky, J. Swiniarski, C.J. Kensler, J.M. Venditti, Combination radiotherapy and chemotherapy for P388 lymphocytic leukemia in vivo, Cancer Chemotherapy Rep. 4 Ž1974. 73–97. w4x M. Zak, J. Drobnik, Effect of cis-dichlorodiammine platinum ŽII. on the post-irradiation lethality in mice after irradiation with X-rays, Strahlentherapie 142 Ž1971. 112–115. w5x E.B. Douple, R.C. Richmond, Platinum complexes as radiosensitizers of hypoxic mammalian cells, Br. J. Cancer 37 Ž1978. 98–102. w6x M.V. Alvarez, G. Cobreros, A. Heras, M.C. Lopez Zumel, Studies on cis-dichlorodiammineplatinumŽII. as a radiosensitizer, Br. J. Cancer 37 Ž1978. 68–72. w7x J. Overgaard, A.R. Khan, Selective enhancement of radiation response in a C3H mammary carcinoma by cisplatin, Cancer Treat. Rep. 65 Ž1981. 501–503. w8x M. Korbelik, K.A. Skov, Inactivation of hypoxic cells by cisplatin and radiation at clinically relevant doses, Radiat. Res. 119 Ž1989. 145–156. w9x K. Skov, S. MacPhail, Interaction of platinum drugs with clinically relevant X-ray doses in mammalian cells: a comparison of cisplatin, carboplatin, iproplatin, and teraplatin, Int. J. Radiat. Oncol. Biol. Phys. 20 Ž1991. 221–225. w10x G.P. Raaphorst, G. Wang, D. Stewart, C.E. Ng, Concomittant low dose-rate irradiation and cisplatin treatment in ovarian carcinoma cell lines sensitive and resistant to cisplatin treatment, Int. J. Radiat. Biol. 69 Ž1996. 623–631. w11x J. Kiefer, J. Muller, Yeast as a Model Object to Study DNA Repair, in: R. Prasad ŽEd.., Yeast Molecular Biology and
136
w12x
w13x
w14x
w15x
w16x
w17x
w18x
w19x
w20x
w21x
w22x
w23x
w24x
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136 Biotechnology, Omega Scientific Publishers, New Delhi, India, 1991, pp. 172–187. P. Unrau, R. Wheatcroft, B.S. Cox, The excision of pyrimidine dimers from DNA of ultraviolet irradiated yeast, Mol. Gen. Genet. 354 Ž1971. 359–362. M. Frankenberg-Schwager, D. Frankenberg, DNA double strand breaks: their repair and relationship to killing in yeast, Int. J. Radiat. Biol. 58 Ž1990. 569–576. R.E.J. Mitchel, D.P. Morrison, An oxygen effect for gamma-radiation induction of radiation resistance in yeast, Radiat. Res. 100 Ž1984. 205–210. R.E.J. Mitchel, D.P. Morrison, Heat-shock induction of ionizing radiation resistance in Saccharomyces cereÕisiae, and correlation with stationary growth phase, Radiat. Res. 90 Ž1982. 284–291. R.E.J. Mitchel, D.P. Morrison, Inducible DNA repair systems in yeast: competition for lesions, Mutation Res. 183 Ž1987. 149–159. J.S. Lazo, B.R. Pitt, Metallothioneins and cell death by anticancer drugs, Annu. Rev. Pharmacol. Toxicol. 35 Ž1995. 635–653. T. McClanahan, K. McEntee, DNA damage and heat shock dually regulate genes in Saccharomyces cereÕisiae, Mol. Cell. Biol. 6 Ž1986. 90–96. S.W. Ruby, J.W. Szostak, Specific Saccharomyces cereÕisiae genes are expressed in response to DNA-damaging agents, Mol. Cell. Biol. 5 Ž1985. 75–84. T. McClanahan, K. McEntee, Specific transcripts are elevated in Saccharomyces cereÕisiae in response to DNA damage, Mol. Cell. Biol. 4 Ž1984. 2356–2363. J.C. Game, DNA double-strand breaks and the RAD50-57 genes in Saccharomyces, Semin. Cancer Biol. 4 Ž1993. 73–83. S.L. Hays, A.A. Fermenich, P. Berg, Complex formation in yeast double-strand break repair: participation of RAD51, RAD52, RAD55, and RAD57 proteins, Proc. Natl. Acad. Sci. U.S.A. 92 Ž1995. 6925–6929. M.A. Resnick, The repair of double-strand breaks in DNA: a model involving recombination, J. Theor. Biol. 59 Ž1976. 97–106. J.W. Szostak, T.L. Orr-Weaver, R.J. Rothstein, F.W. Stahl, The double-strand break model for recombination, Cell 33 Ž1983. 25–35.
w25x A. Sancar, G.B. Sancar, DNA repair enzymes, Annu. Rev. Biochem. 57 Ž1988. 29–67. w26x D.R. Boreham, R.E.J. Mitchel, DNA lesions that signal the induction of radioresistance and DNA repair in yeast, Radiat. Res. 128 Ž1991. 19–28. w27x J.A. Maga, K. McEntee, Response of S. cereÕisiae to Nmethyl-N l-nitro-N-nitrosoguanidine: mutagenesis, survival and DDR gene expression, Mol. Gen. Genet. 200 Ž1985. 313–321. w28x D.B. Yarosh, The role of O 6-methylguanine-DNA methyltransferase in cell survival, mutagenesis and carcinogenesis, Mutation Res. 145 Ž1985. 1–16. w29x R.E.J. Mitchel, D.P. Morrison, Inducible error-prone repair in yeast. Suppression by heat shock, Mutation Res. 159 Ž1986. 31–39. w30x R. Polakowska, G. Perozzi, L. Prakash, Alkylation mutagenesis in Saccharomyces cereÕisiae: lack of evidence for an adaptive response, Curr. Genet. 10 Ž1986. 647–655. w31x A. Averbeck, S. Averbeck, Induction of the genes RAD54 and RNR2 by various DNA damaging agents in Saccharomyces cereÕisiae, Mutation Res. 315 Ž1994. 123–138. w32x D.R. Boreham, A. Trivedi, P. Weinberger, R.E.J. Mitchel, The involvement of topoisomerases and DNA polymerase I in the mechanism of induced thermal tolerance and radiation resistance in yeast, Radiat. Res. 123 Ž1990. 203–212. w33x S.L. Bruhn, J.H. Toney, S.J. Lippard, Biological processing of DNA modified by platinum compounds, Progress in Inorganic Chemistry: Bioinorganic Chemistry 38 Ž1990. 477–517. w34x M.A. Hannan, S.G. Zimmer, J. Hazle, Mechanisms of cisplatin Ž cis-diammine dichloroplatinum II.-induced cytotoxicity and genotoxicity in yeast, Mutation Res. 127 Ž1984. 23–30. w35x A.C. Begg, Additivity versus repair inhibition in fractionated treatments combining drugs and X-rays: a theoretical approach, Int. J. Radiation Oncology Biol. Phys. 13 Ž1987. 921–927. w36x D.R. Boreham, A. Trivedi, R.E.J. Mitchel, Radiation and stress response in Saccharomyces cereÕisiae, in: R. Prasad ŽEd.., Yeast Molecular Biology and Biotechnology, Omega Scientific Publishers, New Delhi, India, 1991, pp. 294–314. w37x A. Eastman, Reevaluation of interaction of cis-dichloroŽethylenediammine.-platinumŽII. with DNA, Biochemistry 25 Ž1986. 3912–3915.
Cisplatin-modification of DNA repair and ionizing radiation lethality in yeast, Saccharomyces cereÕisiae J.-A. Dolling a
a,)
, D.R. Boreham a,1, D.L. Brown R.E.J. Mitchel a,4
b,2
, G.P. Raaphorst
c,3
,
Radiation Biology and Health Physics Branch, AECL, Chalk RiÕer, Ontario, Canada K0J 1J0 b Department of Biology, UniÕersity of Ottawa, Ottawa, Ontario Canada, K1N 6N5 c Ottawa Regional Cancer Centre, Ottawa, Ontario, Canada, K1H 8L6 Received 24 June 1998; revised 4 December 1998; accepted 11 December 1998
Abstract Cis-diamminedichloroplatinum II Žcisplatin. is a DNA inter- and intrastrand crosslinking agent which can sensitize prokaryotic and eukaryotic cells to killing by ionizing radiation. The mechanism of radiosensitization is unknown but may involve cisplatin inhibition of repair of DNA damage caused by radiation. Repair proficient wild type and repair deficient Ž rad52, recombinational repair or rad3, excision repair. strains of the yeast Saccharomyces cereÕisiae were used to determine whether defects in DNA repair mechanisms would modify the radiosensitizing effect of cisplatin. We report that cisplatin exposure could sensitize yeast cells with a competent recombinational repair mechanism Žwild type or rad3 ., but could not sensitize cells defective in recombinational repair Ž rad52 ., indicating that the radiosensitizing effect of cisplatin was due to inhibition of DNA repair processes involving error free RAD52-dependent recombinational repair. The presence or absence of oxygen during irradiation did not alter this radiosensitization. Consistent with this result, cisplatin did not sensitize cells to mutation that results from lesion processing by an error prone DNA repair system. However, under certain circumstances, cisplatin exposure did not cause radiosensitization to killing by radiation in repair competent wild type cells. Within 2 h after a sublethal cisplatin treatment, wild type yeast cells became both thermally tolerant and radiation resistant. Cisplatin pretreatment also suppressed mutations caused by exposure to N-methyl-N X-nitro-N-nitrosoguanidine ŽMNNG., a response previously shown in wild type yeast cells following radiation pretreatment. Like radiation, the cisplatin-induced stress response did not confer radiation resistance or suppress MNNG mutations in a recombinational repair deficient mutant Ž rad52 ., although thermal tolerance was still induced. These results support the idea that cisplatin adducts in DNA interfere with RAD52-dependent recombinational repair and thereby sensitize cells to killing by radiation. However, the lesions can subsequently induce a general stress response, part of which is induction of RAD52-dependent error free recombinational
) 1 2 3 4
Corresponding author. Tel.: q1-613-584-3311; Fax: q1-613-584-1713 Tel.: q1-613-584-3311; Fax: q1-613-584-1713. Tel.: q1-613-562-5800; Fax: q1-613-562-5486. Tel.: q1-613-737-7700; Fax: q1-613-247-6811. Tel.: q1-613-562-5800; Fax: q1-613-562-5486.
0921-8777r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 8 7 7 7 Ž 9 8 . 0 0 0 6 9 - X
128
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
repair. This stress response confers radiation resistance, thermal tolerance, and mutation resistance in yeast. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Cisplatin; Ionizing radiation; Radiosensitization; Adaptive response; Yeast DNA repair
1. Introduction A synergistic cell killing effect of combined cisplatin Ža DNA inter- and intrastrand crosslinking agent. and ionizing radiation treatments has been observed in both prokaryotic w1,2x and eukaryotic w3–10x systems, and this effect has been used in cancer therapy. The exact mechanism of radiosensitization by cisplatin has not been established, but it has been postulated that the supra-additive killing effect of cisplatin and radiation is a consequence of cisplatin inhibition of repair of radiation-induced DNA damage. Two DNA repair pathways that are known to respond to radiation-induced DNA damage, and have been extensively characterized in yeast, are excision repair and recombinational repair w11x. The genes involved in excision and recombinational repair pathways belong to the RAD3 and the RAD52 epistasis groups, respectively. rad3 yeast cells are UVlight sensitive and unable to remove pyrimidine dimers and other bulky adducts from their DNA w11,12x, whereas rad52 cells are sensitive to ionizing radiation and deficient in meiotic and mitotic recombination. rad52 yeast cells are unable to repair DNA double strand breaks, resulting in an extremely ionizing radiation sensitive phenotype w13x. The experiments in this work were designed to investigate the involvement of recombination and excision repair processes in the mechanism of radiosensitization by cisplatin, by determining whether defects in repair processes could modify the radiosensitizing effects of cisplatin. While cisplatin DNA lesions may inhibit repair and consequently radiosensitize cells, under certain conditions, these DNA lesions may also signal the induction of an adaptive response whereby cellular repair capacity is enhanced and radiation resistance is increased. In prokaryotes and eukaryotes, DNA repair capacity can be elevated in response to a variety of stresses including ionizing radiation w14x, heat shock w14,15x, mutagenic chemicals w16x and chemotherapeutic drugs w17x. In yeast, it has been
shown that different stresses induce specific sets of genes w17–20x. A critical set of genes induced in yeast by DNA damaging agents is the RAD50-57 gene series w21x. These proteins form the complex machinery w22x induced during the adaptive response to radiation and used to repair DNA lesions via an error free recombinational process w16,21x. Recombinational repair is essential for successful repair of lesions involving both DNA strands, since the damage in either strand cannot be resynthesized using the genetic information encoded by the damaged complementary strand. Instead, the damaged or missing genetic information is copied from a duplicate source Ži.e., homologous chromosome or sister chromatid. w23–25x. Lesions requiring the recombinational repair system include double strand breaks and interstrand crosslinks. Recombinational repair is not, however, restricted to repair of these relatively complex lesions. For example, it has been shown that recombinational repair can restitute chemically generated N-methyl-N X-nitro-N-nitrosoguanidine ŽMNNG. DNA lesions w16,26x. MNNG is a chemical mutagen that methylates DNA at various sites including the N-7 and O-6 positions of guanine w27x. If the methylated guanine is not repaired, it pairs incorrectly with thymine instead of cytosine during replication, resulting in a GC to AT transition w28x. It has been suggested that MNNG exposure induces an error prone system which is responsible for essentially all MNNG-generated mutations w16x. The number of MNNG-generated mutations, however, has been shown to be reduced in a dose dependent manner in yeast cells pre-exposed to ionizing radiation w16,26x or heat w29x but not by pretreatment with MNNG itself w30x. In fact, MNNG induces an error prone system w29x which increases the mutagenic effect of a subsequent treatment with other chemical mutagens like ethyl methanesulfonate w16x. It has been proposed that preexposure of cells to radiation and heat induces an error free recombinational system that competes with the error prone system for repair of these lesions w29x. An increase in cellular error free recombinational repair capacity would lead
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
to a decrease in the number of MNNG lesions processed by the error prone system and result in a net suppression of mutation. It is not known whether DNA crosslinks caused by cisplatin are effective signals for the induction of error free recombinational repair, although in yeast, crosslinks caused by furocoumarins resulted in upregulation of RAD54 transcription w31x. In addition, there is evidence to indicate that the basic signal for induction of recombinational repair is perturbation of DNA topology w32x. Therefore, it is likely that DNA crosslinks which cause base destacking and strand kink angles of 40 to 608 w33x, could Ža. invoke an adaptive response and Žb. lead specifically to an increased error free repair recombinational capacity. In the investigation presented here, the induction of recombinational repair in yeast by cisplatin was tested by pretreatment with cisplatin and subsequent exposure to ionizing radiation, heat or MNNG, and the results were compared with the potentially radiosensitizing effect of cisplatin treatments.
2. Materials and methods 2.1. Strains Diploid strains of the yeast Saccharomyces cereÕisiae were used as the test organisms in these experiments. Strain MS33 was wild type for DNA repair capacity Ž Rec q r Rec q , Exc q r Exc q .. Strain MS32 was homozygous for rad52-1 and was defective in recombinational repair Ž Rec y r Rec y , Exc q r Exc q .. Strain MS33 was the isogenic parental strain of MS32. Both MS33 and MS32 were also defective for the ATP phosphoribosyltransferase gene Ž his1-7r his1-7 .. Strain STX432 was homozygous for rad3-2 and was defective in excision repair Ž Rec q r Rec q , Exc y r Exc y .. Yeast strains MS33 and MS32 were gifts from Dr. D.P. Morrison, AECL, Chalk River Laboratories. Strain STX432 was purchased from the Yeast Genetic Stock Center ŽBerkeley, CA.. 2.2. Cell growth Cell cultures were grown at 238C to mid-exponential phase Ž2–4 = 10 6 cellsrml. in liquid nutrient
129
medium containing 0.1% yeast extract, 1% yeast nitrogen base with amino acids, 1% sodium succinate, and 2% glucose Žyeast extractrsuccinate medium or YES.. 2.3. Cisplatin treatment Cells Ž2 = 10 7 . were treated for 30 min with cisplatin ŽBristol Myers Squibb. diluted to the appropriate concentration in 238C sodium phosphate buffer. After cisplatin treatment, cells were washed three times with 08C buffer. Control cells were suspended in sodium phosphate buffer without cisplatin, for 30 min at 238C and then treated in the same way as cisplatin treated cell samples. Survival was determined by plating suitable dilutions on solid yeastrpeptonerdextrose ŽYPD. agar plates containing 0.1% yeast extract, 1% yeast nitrogen base with amino acids, 1% sodium succinate, 2% peptone, 2% glucose. Colony formation was scored following five days of incubation at 238C. Some cells were tested for induced thermal or radiation resistance, or resistance to mutation as described below. 2.4. Radiation surÕiÕal Yeast cells were washed three times and suspended Žat 08C. in 20 mmol dmy3 ŽmM. sodium phosphate buffer, pH 7.0, at approximately 2 = 10 6 cellrml. These suspensions were equilibrated by bubbling the cell suspension at 08C with O 2 or N2 ŽMatheson ultrahigh purity. for 10 min before irradiation and continuously during irradiation. Cells were irradiated at 08C with 60 Co gamma-rays ŽGamma cell 220, Atomic Energy of Canada. at a dose rate of about 1 Gyrs. Survival was determined by plating suitable dilutions of irradiated and unirradiated control cells on solid YPD agar plates followed by incubation for five days at 238C. 2.5. Thermal surÕiÕal Thermal resistance was determined by placing 2.5 ml cell aliquots Ž2 = 10 6 cellsrml in YES medium at 08C. in 16 = 100 mm glass test tubes and submerging the glass test tubes into a 528C shaking water bath. At specified time intervals, individual tubes were removed and rapidly cooled to 08C. Control cells were held at 08C for the same duration.
130
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
with phosphate buffer, all at 08C. Control cells not exposed to MNNG were treated similarly. Reversion to histidine independence and survival to treatments was determined as described above. 2.7. Data and figures Each experiment was repeated at least three times from independent cultures. All figures shown represent the combined results of all the experiments. Standard error bars are shown except were the error is less than or equal to the symbol size. Student’s t-tests were performed to determine the statistical significance between the various treatments and pvalues are given when the results were significantly different.
3. Results 3.1. Radiation surÕiÕal Fig. 1. Survival of wild type MS33 Žcircles., recombinational repair deficient MS32 Žsquares. and excision repair deficient STX432 Žtriangles. yeast cells after ŽA. 60 Co gamma irradiation delivered in the presence of nitrogen or ŽB. cisplatin treatment at 238C for 30 min.
Wild type MS33, recombinational repair deficient MS32 and excision repair deficient STX432 yeast cells were irradiated with increasing doses of gamma
Treated and control cells were then plated for survival on YPD plates. 2.6. Mutation and suppression of MNNG induced mutation by cisplatin Cisplatin-induced mutation was determined by measuring reversion to histidine independence. Revertant cells able to form colonies were counted after incubation at 238C for 6 days on solid medium lacking histidine. Total survival after the various treatments was measured by scoring colony formation on YPD plates. Chemical mutagenesis was performed at 238C in 20 mmol dmy3 ŽmM. phosphate buffer ŽpH 7.0. at a cell density of 1 = 0 7 cellsrml. Cells were treated with 5 or 20 mgrml of MNNG ŽAldrich Chemical. for 30 min either immediately after or at the same time as a cisplatin treatment. After MNNG exposure, cells were washed once with 10% Žwrv. sodium thiosulphate to inactivate the MNNG and then twice
Fig. 2. Survival of wild type MS33 yeast cells after 60 Co gamma irradiation delivered in the presence of nitrogen Žcircles. or oxygen Ždiamonds., with Žopen symbols. or without Žclosed symbols. a cisplatin treatment at 238C for 30 min immediately before irradiation.
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
131
rays in the presence of N2 ŽFig. 1A.. MS32 cells were sensitive to killing by gamma radiation relative to wild type MS33 cells. STX432 showed an intermediate level of sensitivity at doses greater than 1 kGy. 3.2. Cisplatin surÕiÕal All three yeast strains were tested for survival against increasing concentrations of cisplatin ŽFig. 1B.. The results were similar to those for radiation survival as shown in Fig. 1A, with wild type MS33 demonstrating the most resistance to killing by cisplatin, MS32 demonstrating the least resistance and STX432 showing an intermediate response. 3.3. Combined treatment of cisplatin and radiation Survival of wild type MS33 yeast cells to gamma radiation after a cisplatin treatment of 400 mgrml
Fig. 4. Survival of ŽA. wild type MS33 and ŽB. recombinational repair deficient yeast cells exposed to 60 Co gamma irradiation delivered in the presence of oxygen. MS33 cells were previously exposed to 400 mgrml cisplatin Ž238C, 30 min. followed by a 2-h incubation at 238C. MS32 cells were treated for 30 min at 238C with 50 mgrml Žtriangles. or 400 mgrml Žsquares. cisplatin, incubated for various time intervals and then irradiated Ž0.1 kGy in the presence of oxygen.. Radiation survival of untreated control cells is represented by closed circles.
Fig. 3. Survival of ŽA. wild type MS33, ŽB. recombinational repair deficient MS32 and ŽC. excision repair deficient STX432 yeast cells after 60 Co gamma irradiation delivered in the presence of nitrogen, without Žclosed symbols. or with Žopen symbols. a 30-min cisplatin treatment Žopen circles: 400 mgrml, open triangles: 50 mgrml, open diamonds: 10 mgrml. immediately before irradiation.
for 30 min is shown in Fig. 2. Cisplatin treatment resulted in radiosensitization when the gamma radiation was delivered in oxygen or nitrogen. At a survival level of 10% ŽD 10 . the enhancement ratios were 1.25 and 1.3 for oxic and anoxic irradiation conditions, respectively, and were not significantly different. In Fig. 3, survival of MS33, MS32 and STX432 yeast cells to gamma radiation Ždelivered in the presence of N2 . immediately after a cisplatin treatment. Cisplatin treatment Ž400 mgrml, 30 min. resulted in radiosensitization of wild-type MS33 and excision repair deficient STX432 cells Ž p - 0.01. but not recombinational repair deficient MS32 cells. At a survival level of 10% ŽD10 ., the enhancement ratios for MS33 and STX432 cells Ž1.3 and 1.4 respectively. were not significantly different. Fig. 3B
132
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
time point was similar for both yeast strains. A smaller dose of cisplatin Ž50 mgrml, 30 min. had no effect on thermal survival of either yeast strain. 3.6. Cisplatin suppression of MNNG-generated mutations Wild type MS33 cells showed a cisplatin dose-dependent reduction in mutations generated by MNNG ŽFig. 6A.. Simultaneous treatment of cells with cisplatin and MNNG did not result in reduction of MNNG-generated mutation until the concentrations of cisplatin exceeded 100 mgrml. However, treatment of cells with cisplatin immediately before MNNG exposure resulted in a reduction of mutation at all cisplatin concentrations tested, although the
Fig. 5. Cisplatin-induced thermal tolerance in ŽA. wild type MS33 and ŽB. recombinational repair deficient MS32 yeast cells. Cells were treated with 50 mgrml Žtriangles. or 400 mgrml Žsquares. cisplatin, incubated for various time intervals and then heated at 528C for 3 min. Control cells are represented by circles.
shows that cisplatin, at any concentration tested, did not change the survival of recombinational repair deficient MS32 cells at any dose of radiation tested. 3.4. Cisplatin-induced radiation resistance Incubation of wild type MS33 cells for 2 h at 238C after cisplatin treatment Ž400 mgrml, 238C, 30 min. resulted in elevated radiation resistance Ž p 0.05. ŽFig. 4A.. However, in recombinational repair deficient MS32 cells, radiation resistance was not induced by cisplatin treatment Ž50 or 400 mgrml, 238C, 30 min. at any time interval tested ŽFig. 4B.. 3.5. Cisplatin-induced thermal tolerance Cisplatin exposure Ž400 mgrml, 30 min. induced thermal tolerance in both wild type MS33 ŽFig. 5A. and recombinational repair deficient MS32 cells ŽFig. 5B.. The level of thermal tolerance reached at each
Fig. 6. Cisplatin suppression of MNNG-generated mutations in ŽA. wild type MS33 or ŽB. recombinational repair deficient MS32 yeast cells. MS33 cells were treated simultaneously with cisplatin Ž400 mgrml. and MNNG Ž20 mgrml. at 238C for 30 min Žclosed circles. or sequentially with cisplatin Ž400 mgrml, 238C, 30 min. followed by MNNG Ž20 mgrml, 238C, 30 min. Žopen circles.. MS32 cells were treated with cisplatin alone at 238C for 30 min Žclosed circles. or simultaneously with MNNG Ž5 mgrml. Žopen circles..
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
133
Table 1 Lack of cisplatin suppression of MNNG-generated mutations in recombinational repair deficient MS32 yeast when cells are treated sequentially with cisplatin followed immediately by MNNG Treatment
Number mutationsr 10 6 survivors
Control Žspontaneous. 100 mgrml cisplatin 5 mrml MNNG 100 mgrml cisplatin followed by 5 mgrml MNNG 20 mgrml MNNG 100 mgrml cisplatin followed by 20 mgrml MNNG
15.0 " 1.7 20.4 " .05 23.0 " 3.4 32.5 " 3.0 a 148.0 " 11.2 152.0 " 1.7 a
a
Cisplatin treatment did not suppress MNNG-generated mutation.
reduction was not statistically significant until the concentration exceeded 50 mgrml. At the point of the largest reduction measured Ža 400-mgrml, 30-min pretreatment reduced the number of mutations from 1230 " 100 to 386 " 91 per 10 6 survivors. the mutation rate was still significantly higher than the spontaneous mutation rate of about 4 per 10 6 survivors for wild-type MS33 cells. Under the conditions of the cisplatin treatments used in this study, cisplatin alone did not produce any mutations above the spontaneous level in wild type MS33 cells. In contrast, cisplatin treatment alone produced mutations in recombinational repair deficient MS32 cells ŽFig. 6B, Table 1.. The combined exposures of recombinational repair-deficient MS32 cells to cisplatin and MNNG did not result in a reduction in the number of MNNGgenerated mutations ŽFig. 6B.. This lack of suppression of MNNG-generated mutations was found when the two treatments were given either simultaneously or when cisplatin treatment Ž100 mgrml, 30 min, 238C. immediately preceded the MNNG treatment Ž5 or 20 mgrml, 30 min, 238C. ŽTable 1.. The largest MNNG exposure of MS32 cells Ž20 mgrml, 30 min. generated 148 " 11.2 mutations per 10 6 survivors. This frequency is significantly less than was observed for MS33 cells exposed to the same dose Ž1230 " 108 mutations per 10 6 survivors, Fig. 6A. Ž p - 0.001..
4. Discussion The results presented in this study show that cisplatin used in combination with ionizing radiation
can produce differential cellular responses depending upon the inherent repair competence of the cell and the time interval between cisplatin and radiation exposures. DNA double strand break repair, via RAD52-dependent recombinational repair, is known to be the major pathway which confers radiation resistance in yeast w13x and our MS32 strain, defective in RAD52-dependent recombinational repair, displayed the expected radiosensitivity compared to wild type MS33 cells ŽFig. 1A.. The radiation sensitivity of cells deficient for the RAD3 excision repair protein Žstrain STX432. ŽFig. 1A. showed that excision repair may also be an important pathway which repairs ionizing radiation damage. Radiosensitivity of another strain defective for RAD3 Žstrain D7-3. has been previously demonstrated ŽR.E.J. Mitchel, unpublished data., but in both studies the isogenic wild type parent strain was not available to make direct comparisons. Therefore, it remains possible that the relative radiosensitivity of both excision repair deficient mutants may be due to some other intrinsic characteristic of the strains. Recombination and excision repair processes are involved in repairing cisplatin DNA adducts as previously shown w34x and demonstrated here by the sensitivity to killing by cisplatin in recombinational repair deficient MS32 and excision repair deficient STX432 cells ŽFig. 1B.. The extreme sensitivity of recombinational repair deficient MS32 cells to both radiation and cisplatin indicates that, in yeast cells, recombinational repair involving proteins in the RAD52 gene series ŽRad51, 54, 57. w22x is a dominant mode of repair. Radiation-induced DNA double strand breaks and cisplatin-induced DNA interstrand
134
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
crosslinks are two lesions that are most likely repaired via a recombinational process, since in both cases correct genetic information from a duplicate source Že.g., homologous chromosome or sister chromatid. is required for successfully restoring the damaged sites. Radiosensitization by cisplatin under oxic and anoxic conditions ŽFig. 2. has been previously reported for bacterial and mammalian systems w1,2,5– 7,9,35x. However, several studies have reported that the synergistic effect for cell killing is diminished or absent when cells are irradiated in air w1,2,8x. Results of the present study show similar radiosensitization under both conditions and indicate that, in yeast, modification of lesions by oxygen does not influence radiosensitization by cisplatin. In the present study, excision repair deficient STX432 cells and wild type MS33 cells, but not recombinational repair deficient MS32 cells, were radiosensitized by cisplatin indicating that radiosensitization by cisplatin is absolutely dependent upon functional recombinational repair. Whereas cisplatin treatment immediately before irradiation led to radiosensitization in wild type MS33 yeast cells ŽFig. 2., a similar exposure followed by a 2 h incubation under growth permissive conditions resulted in increased resistance to killing by radiation ŽFig. 4A.. This phenomenon of induced radiation resistance has been well documented in yeast. It has been shown that cellular recombinational repair capacity can be enhanced by exposure to a variety of stresses including radiation w15,16,32x. The magnitude of the increase in radiation resistance is dependent upon the severity of the inducing stress and the time interval between the inducing and challenge exposures Žreviewed in Ref. w36x.. Since induced radiation resistance is absolutely dependent upon recombinational repair capacity w29x, recombinational repair deficient MS32 cells do not exhibit a radiation resistant phenotype after a cisplatin treatment, regardless of the incubation interval between cisplatin and radiation exposures. Cisplatin adducts also signalled induction of thermal tolerance, to the same extent in both wild type MS33 and recombinational repair deficient MS32 yeast cells ŽFig. 6B., indicating that for induced thermal tolerance, the signalling mechanism for induction by cisplatin does not require RAD52. This has previously been shown for induction of thermal
tolerance by heat and radiation w14x. The induction of both radiation and thermal resistance suggests that cisplatin lesions are inducing a general stress response. The interpretation that cisplatin adducts are capable of invoking recombinational repair is further supported by the data indicating suppression of MNNG-generated mutations when wild type MS33 ŽFig. 6A., but not recombinational repair deficient MS32 cells ŽFig. 6B, Table 1., were treated with cisplatin. Suppression of MNNG-generated mutations is believed to be the result of competition between the induced error free recombinational repair system and the error prone repair system w16,29x. MNNG lesions induce the error prone system and cause mutations. If cisplatin adducts signal induction of an error free repair system that is capable of repairing both cisplatin and MNNG lesions, then fewer MNNG-generated mutations would result. Indeed, in this study, increasing doses of cisplatin resulted in more MNNG lesions being processed by the error free repair pathway and a reduction in the frequency of MNNG-generated mutations. In wild type MS33 cells, both cisplatinrMNNG treatment regimes Žsimultaneous or sequential treatment of cisplatin and MNNG. resulted in a reduction in the number of mutations, indicating that cisplatin adducts had caused upregulation of recombinational repair capacity. At the highest inducing dose of cisplatin, the number of residual mutations was still significantly higher than the spontaneous level of 4r10 6 MS33 survivors. This result indicates that induction of the error free system was paralleled by a relative reduction, but not elimination, in the number of lesions processed by the error prone system. Therefore, it is likely that even at maximal induction of the error free system, some MNNG lesions will be repaired by the error prone system w16x. Differences in the kinetics and magnitude of mutation suppression between treatment regimes may reflect differences in the ability to induce recombinational repair by the different spectra of cisplatin adducts present at the time of MNNG treatment. After short cisplatin treatments Ž15 or 30 min., the ratio of mono- to bifunctional adducts is approximately 50:50. During longer incubation times Ž60 min., the ratio changes to 30:70, mono- to bifunctional adducts, and at longer incubation times Ž2 h or
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136
more. monofunctional adducts constitute only a few percent of the total cisplatin adducts w33,37x. When cisplatin and MNNG were delivered simultaneously, induction of an error free system was therefore attributed to either cisplatin monofuctional or bifunctional adducts. However, when MNNG was delivered after the cisplatin treatment, there was more time for conversion of cisplatin mono- to bifunctional adducts and it is possible that the presence of bifunctional adducts resulted in a more effective induction of error free recombinational repair capacity. Further, since bifunctional, but not monofuctional, cisplatin adducts produce dramatic topological perturbations in the DNA helix by creating kink angles of 40 to 608 w33x, the observed responses may be due to differences in DNA topological perturbation. In the present study, during the 30 min cisplatin exposure and subsequent 30 min MNNG treatment, some cisplatin monofunctional adducts, which do not appear to distort the DNA helix w33x, would have been converted into helix distorting, bifunctional adducts, and may have resulted in more effective signalling of an error free recombinational repair system. These perturbations in DNA topology may serve as the signal for induction of a radiation resistance mechanism as has been suggested by a study comparing topoisomerase mutants to wild type yeast cells for induction of radiation resistance w26x. It should be noted that cisplatin adducts present simultaneously with MNNG lesions did not increase the mutation frequency ŽFig. 6A. unlike the effect on lethality from radiation exposure. Since cisplatin adducts sensitize cells to killing by radiation but not to mutation by MNNG, we suggest that cisplatin adducts do not impact on the error prone repair system which recognizes MNNG lesions. In summary, radiosensitization by cisplatin depends absolutely upon a functional RAD52-dependent recombinational repair system but not on a functional excision repair system. The data demonstrate that radiosensitization by cisplatin DNA adducts results from inhibition of recombinational repair. There was no significant effect of oxygen on radiosensitization. The results also indicate that cisplatin adducts can signal the induction of an error free recombinational repair process which is dependent upon RAD52 function and that cisplatin bifunctional adducts are better inducers than monofunc-
135
tional adducts. Induction of this error free system in wild type cells is part of a general stress response which resulted in resistance to killing by radiation and heat, and mutation caused by MNNG.
Acknowledgements The authors would like to thank M.-E. Bahen for expert technical assistance and D.P. Morrison for generously supplying the MS32 and MS33 yeast strains and insights into yeast genetics. This work was supported by the CANDU Owners Group and the Ottawa Regional Cancer Centre.
References w1x R.C. Richmond, E.L. Powers, Radiation sensitization of bacterial spores by cis-dichlorodiammineplatinumŽII., Radiat. Res. 68 Ž1976. 251–257. w2x R.C. Richmond, J.D. Zimbrick, D.L. Hykes, Radiation-induced DNA damage and lethality in E. Coli as modified by the antitumor agent cis-dichlorodiammineplatinumŽII., Radiat. Res. 71 Ž1977. 447–460. w3x I. Wodinsky, J. Swiniarski, C.J. Kensler, J.M. Venditti, Combination radiotherapy and chemotherapy for P388 lymphocytic leukemia in vivo, Cancer Chemotherapy Rep. 4 Ž1974. 73–97. w4x M. Zak, J. Drobnik, Effect of cis-dichlorodiammine platinum ŽII. on the post-irradiation lethality in mice after irradiation with X-rays, Strahlentherapie 142 Ž1971. 112–115. w5x E.B. Douple, R.C. Richmond, Platinum complexes as radiosensitizers of hypoxic mammalian cells, Br. J. Cancer 37 Ž1978. 98–102. w6x M.V. Alvarez, G. Cobreros, A. Heras, M.C. Lopez Zumel, Studies on cis-dichlorodiammineplatinumŽII. as a radiosensitizer, Br. J. Cancer 37 Ž1978. 68–72. w7x J. Overgaard, A.R. Khan, Selective enhancement of radiation response in a C3H mammary carcinoma by cisplatin, Cancer Treat. Rep. 65 Ž1981. 501–503. w8x M. Korbelik, K.A. Skov, Inactivation of hypoxic cells by cisplatin and radiation at clinically relevant doses, Radiat. Res. 119 Ž1989. 145–156. w9x K. Skov, S. MacPhail, Interaction of platinum drugs with clinically relevant X-ray doses in mammalian cells: a comparison of cisplatin, carboplatin, iproplatin, and teraplatin, Int. J. Radiat. Oncol. Biol. Phys. 20 Ž1991. 221–225. w10x G.P. Raaphorst, G. Wang, D. Stewart, C.E. Ng, Concomittant low dose-rate irradiation and cisplatin treatment in ovarian carcinoma cell lines sensitive and resistant to cisplatin treatment, Int. J. Radiat. Biol. 69 Ž1996. 623–631. w11x J. Kiefer, J. Muller, Yeast as a Model Object to Study DNA Repair, in: R. Prasad ŽEd.., Yeast Molecular Biology and
136
w12x
w13x
w14x
w15x
w16x
w17x
w18x
w19x
w20x
w21x
w22x
w23x
w24x
J.-A. Dolling et al.r Mutation Research 433 (1999) 127–136 Biotechnology, Omega Scientific Publishers, New Delhi, India, 1991, pp. 172–187. P. Unrau, R. Wheatcroft, B.S. Cox, The excision of pyrimidine dimers from DNA of ultraviolet irradiated yeast, Mol. Gen. Genet. 354 Ž1971. 359–362. M. Frankenberg-Schwager, D. Frankenberg, DNA double strand breaks: their repair and relationship to killing in yeast, Int. J. Radiat. Biol. 58 Ž1990. 569–576. R.E.J. Mitchel, D.P. Morrison, An oxygen effect for gamma-radiation induction of radiation resistance in yeast, Radiat. Res. 100 Ž1984. 205–210. R.E.J. Mitchel, D.P. Morrison, Heat-shock induction of ionizing radiation resistance in Saccharomyces cereÕisiae, and correlation with stationary growth phase, Radiat. Res. 90 Ž1982. 284–291. R.E.J. Mitchel, D.P. Morrison, Inducible DNA repair systems in yeast: competition for lesions, Mutation Res. 183 Ž1987. 149–159. J.S. Lazo, B.R. Pitt, Metallothioneins and cell death by anticancer drugs, Annu. Rev. Pharmacol. Toxicol. 35 Ž1995. 635–653. T. McClanahan, K. McEntee, DNA damage and heat shock dually regulate genes in Saccharomyces cereÕisiae, Mol. Cell. Biol. 6 Ž1986. 90–96. S.W. Ruby, J.W. Szostak, Specific Saccharomyces cereÕisiae genes are expressed in response to DNA-damaging agents, Mol. Cell. Biol. 5 Ž1985. 75–84. T. McClanahan, K. McEntee, Specific transcripts are elevated in Saccharomyces cereÕisiae in response to DNA damage, Mol. Cell. Biol. 4 Ž1984. 2356–2363. J.C. Game, DNA double-strand breaks and the RAD50-57 genes in Saccharomyces, Semin. Cancer Biol. 4 Ž1993. 73–83. S.L. Hays, A.A. Fermenich, P. Berg, Complex formation in yeast double-strand break repair: participation of RAD51, RAD52, RAD55, and RAD57 proteins, Proc. Natl. Acad. Sci. U.S.A. 92 Ž1995. 6925–6929. M.A. Resnick, The repair of double-strand breaks in DNA: a model involving recombination, J. Theor. Biol. 59 Ž1976. 97–106. J.W. Szostak, T.L. Orr-Weaver, R.J. Rothstein, F.W. Stahl, The double-strand break model for recombination, Cell 33 Ž1983. 25–35.
w25x A. Sancar, G.B. Sancar, DNA repair enzymes, Annu. Rev. Biochem. 57 Ž1988. 29–67. w26x D.R. Boreham, R.E.J. Mitchel, DNA lesions that signal the induction of radioresistance and DNA repair in yeast, Radiat. Res. 128 Ž1991. 19–28. w27x J.A. Maga, K. McEntee, Response of S. cereÕisiae to Nmethyl-N l-nitro-N-nitrosoguanidine: mutagenesis, survival and DDR gene expression, Mol. Gen. Genet. 200 Ž1985. 313–321. w28x D.B. Yarosh, The role of O 6-methylguanine-DNA methyltransferase in cell survival, mutagenesis and carcinogenesis, Mutation Res. 145 Ž1985. 1–16. w29x R.E.J. Mitchel, D.P. Morrison, Inducible error-prone repair in yeast. Suppression by heat shock, Mutation Res. 159 Ž1986. 31–39. w30x R. Polakowska, G. Perozzi, L. Prakash, Alkylation mutagenesis in Saccharomyces cereÕisiae: lack of evidence for an adaptive response, Curr. Genet. 10 Ž1986. 647–655. w31x A. Averbeck, S. Averbeck, Induction of the genes RAD54 and RNR2 by various DNA damaging agents in Saccharomyces cereÕisiae, Mutation Res. 315 Ž1994. 123–138. w32x D.R. Boreham, A. Trivedi, P. Weinberger, R.E.J. Mitchel, The involvement of topoisomerases and DNA polymerase I in the mechanism of induced thermal tolerance and radiation resistance in yeast, Radiat. Res. 123 Ž1990. 203–212. w33x S.L. Bruhn, J.H. Toney, S.J. Lippard, Biological processing of DNA modified by platinum compounds, Progress in Inorganic Chemistry: Bioinorganic Chemistry 38 Ž1990. 477–517. w34x M.A. Hannan, S.G. Zimmer, J. Hazle, Mechanisms of cisplatin Ž cis-diammine dichloroplatinum II.-induced cytotoxicity and genotoxicity in yeast, Mutation Res. 127 Ž1984. 23–30. w35x A.C. Begg, Additivity versus repair inhibition in fractionated treatments combining drugs and X-rays: a theoretical approach, Int. J. Radiation Oncology Biol. Phys. 13 Ž1987. 921–927. w36x D.R. Boreham, A. Trivedi, R.E.J. Mitchel, Radiation and stress response in Saccharomyces cereÕisiae, in: R. Prasad ŽEd.., Yeast Molecular Biology and Biotechnology, Omega Scientific Publishers, New Delhi, India, 1991, pp. 294–314. w37x A. Eastman, Reevaluation of interaction of cis-dichloroŽethylenediammine.-platinumŽII. with DNA, Biochemistry 25 Ž1986. 3912–3915.