Chem. Res. Toxicol. 2003, 16, 171-179
171
Molecular Characterization of Mitomycin C-Induced Large Deletions and Tandem-Base Substitutions in the Bone Marrow of gpt delta Transgenic Mice Akira Takeiri,*,† Masayuki Mishima,† Kenji Tanaka,† Akihumi Shioda,† Otoya Ueda,† Hiroshi Suzuki,†,‡ Makoto Inoue,† Ken-ichi Masumura,§ and Takehiko Nohmi§ Fuji Gotemba Research Laboratories, Chugai Pharmaceutical Co., Ltd., 1-135 Komakado, Gotemba-shi, Shizuoka 412-8513, Japan, and Division of Genetics and Mutagenesis, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku, Tokyo, 159-5801, Japan Received June 7, 2002
Deletion mutations constitute an important class of mutations that may result in a variety of human diseases, including cancer. Although many chemicals and ionizing radiations induce deletions, this class of mutation has been poorly characterized at the molecular level, particularly in vivo. Here we report the molecular nature of deletions as well as base substitutions induced by antitumor antibiotic mitomycin C (MMC) in the bone marrow using a novel transgenic mouse, gpt delta. In this mouse model, deletions and point mutations in lambda DNA integrated in the chromosome are individually selected as Spi- (sensitive to P2 interference) phages and 6-thioguanine-resistant bacterial colonies, respectively. The mice were treated with MMC (1 mg/kg/day) for five consecutive days. One week after the last treatment, lambda phage was rescued from the genomic DNA of the bone marrow by in vitro packaging reactions and subjected to Spi- and 6-thioguanine selections. The mutant frequency of Spiwith large deletions increased more than 20-fold over that of the control. Molecular sizes of the large deletions were mostly more than 2000 base pairs. The large deletions frequently occurred between two short direct repeat sequences from 2 to 6 base pairs, suggesting that they are generated during the end-joining repair of double-strand breaks induced by interstrand cross-links in DNA. In 6-thioguanine selection, tandem-base substitutions, such as 5′-GG-3′ to 5′-AT-3′, were induced. It highlights the relevance of intrastrand cross-links as genotoxic lesions. Previous in vitro studies report the induction of single-base substitutions and singlebase deletions by MMC. However, no such mutations were identified in vivo. Thus, our results strongly caution that in vitro mutation spectra do not necessarily reflect genotoxic events in vivo and emphasize the importance of transgenic rodent genotoxicity assays to examine the roles of DNA adducts in mutagenesis and carcinogenesis.
Introduction (MMC)1
Mitomycin C is a natural cytotoxic and genotoxic agent used in clinical anticancer chemotherapy. It is most often used in combination with other drugs, such as 5-fluorouracil, for tumors of the stomach, pancreas, lung, and cervix (1). MMC requires reductive activation by flavoreductases for the expression of biological and chemical properties (2). The reductive metabolites bind covalently to DNA, which is thought to be responsible for the cytotoxicity and genotoxicity. Interestingly, MMC alkylates DNA in several different ways. It binds to the N2 position of guanine in DNA and forms monoalkylation products. However, after this step, if the carbamate at C-10′′ of MMC is lost, it will give rise to another active site capable of alkylating a second guanine in DNA. * To whom correspondence should be addressed. † Fuji Gotemba Research Laboratories. ‡ Present address: National Research Center for Protozoan Diseases, Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan. § Division of Genetics and Mutagenesis. 1 Abbreviations: MMC, mitomycin C; DSBs, DNA double-strand breaks; Spi-, sensitive to P2 interference; 6-TG, 6-thioguanine; MF, mutant frequency; kb, kilobase pair; i.p., intraperitoneal; bp, base pair.
Alkylation of two guanine bases on the same DNA strand results in the formation of intrastrand cross-links, and the alkylation of guanine bases on opposite DNA strands leads to the formation of interstrand cross-links (3-7). DNA interstrand cross-links are formed exclusively between two guanine bases in the 5′-CG-3′/3′-GC-5′ sequences (2). Since MMC induces a variety of DNA adducts, attention has been paid to their biological consequences. It has been demonstrated that the monoalkylation products are preferentially formed at G in 5′-CG-3′ sequences and DNA synthesis was terminated at the nucleotide 3′ to the DNA adducts (8, 9). Studies using a shuttle vector replicating in cultured human cell indicate that singlebase substitutions and single-guanine-base deletions are induced by the monoalkylation adducts and that tandembase substitutions at 5′-GG-3′ site arise from the intrastrand cross-links (10, 11). In addition to the in vitro studies, the genotoxicity of MMC in vivo has been investigated using several transgenic mice. A 2-fold increase in the mutant frequency (MF) of lacZ is induced in the bone marrow of Muta Mouse after five daily intraperitoneal (i.p.) injection of 2 mg/kg MMC (12, 13).
10.1021/tx0255673 CCC: $25.00 © 2003 American Chemical Society Published on Web 01/29/2003
172
Chem. Res. Toxicol., Vol. 16, No. 2, 2003
When treated by gavage, single treatment of 4 mg/kg induces nearly a 2-fold increase in the MF of lacZ in the small intestine (14). However, the lacZ mutations are not characterized at the molecular level so that it is unclear whether the mutations observed in vitro (10, 11) are induced in vivo. Besides point mutations, MMC induces gross genetic alterations, such as chromosome aberrations and sister chromatid exchanges, which are induced by DNA double-strand breaks (DSBs) (15, 16). In fact, 2-fold induction of inversion events is induced in the spleen of pKZ1 mice by single i.p. injection of MMC at a dose of 4 mg/kg (17). However, molecular nature of the inversion events is not reported. gpt delta mice (see below) are treated by single i.p. injection at a dose of 4 mg/kg (18). Although some of the mutants are indeed deletions, the treatment does not significantly enhance MF in the bone marrow. No base substitutions (gpt) are analyzed in the study. Thus, it is desired to thoroughly characterize both genetic alterations and point mutations at the molecular level to establish the relationships between MMC adduct structures and genotoxicity in vivo. We have established gpt delta transgenic mice to facilitate the detection and analysis of mutations in vivo (19-21). A feature of this mutation assay is its ability to detect deletions and point mutations. In this mouse model, about 80 copies of lambda EG10 shuttle vector DNA carrying the red/gam genes of lambda phage and the gpt gene of Escherichia coli are integrated in chromosome 17 of the C57BL/6J background (22). Lambda phage is rescued from the genomic DNA by in vitro packaging reactions after treatments with mutagens. Deletions in the red/gam genes and point mutations in the gpt gene are individually identified by Spi- selection (sensitive to P2 interference) and 6-thioguanine (6-TG) selection, respectively (22, 23). The Spi- selection takes advantage of the fact that only mutant lambda phages that are deficient in gam and redBA gene functions can grow well in P2 lysogens and display the Spi- phenotype. Simultaneous inactivation of both the gam and the redBA gene is induced by deletions in the region (24). Using Spiselection, Nohmi et al. detected 5-15-fold increases in the MF of Spi- over the background in the spleen of mice irradiated with γ-irradiation, and they demonstrated that all the Spi- mutants are indeed deletions in the lambda EG10 DNA (19, 23). In this study, we treated gpt delta mice with MMC and characterized both deletions (Spi-) and point mutations (gpt) in the bone marrow at the molecular level. The results indicate that more than half of MMC-induced deletions are generated between two short homologous sequences and the deletion sizes are more than 2 kilobase pairs (kb). We also revealed using 6-TG selection that tandem-base substitutions are induced by MMC, but single-base substitutions and single deletions, which are induced by MMC in vitro, are not identified in vivo. Possible mechanisms leading to DSBs and tandem-base substitutions are discussed.
Materials and Methods Treatment of Animals and Preparation of Bone Marrow Cells. Seven- to 25-week-old gpt delta mice maintained in Fuji Gotemba Research Laboratories, Chugai Pharmaceutical Co., Ltd. were used. The mice received five consecutive i.p. injections of MMC (CAS no. 50-07-7, Kyowa Hakko Kogyo, Japan) at a daily dose of 1 mg/kg body weight for 5 days. Bone
Takeiri et al. marrow cells were collected from the femurs of the three male and the three female mice both 1 and 4 weeks after the last injection. These groups are referred to as MMC 1-week expression time and MMC 4-week expression time groups, respectively. Control bone marrow cells were also collected from four male and three female gpt delta mice, which received five consecutive injections of saline and were sacrificed 4 weeks after the last treatment. The harvested cells were immediately frozen in liquid nitrogen and stored at -80 °C. Mutation Analysis. The MFs in the gpt and red/gam genes derived from the bone marrow cells were determined by 6-TG selection and Spi- selection as previously described (19, 20, 22, 23). DNA sequences of the mutants isolated from the control bone marrow cells and the MMC 1-week expression time group were determined. Four or five randomly chosen gpt mutants from each mouse and all Spi- mutants were analyzed. Briefly, the DNA fragment containing the gpt gene was PCR-amplified using primer-1 (forward) and primer-2 (reverse) (25), which were filtered through Microcon YA-100 (Millipore, Bedford, MA) and then sequenced by the Dye Terminator Cycle Sequencing System (Perkin-Elmer, Wellesley, MA) using sequencing primers as described (25). For the sequence analysis of Spi- mutants, deletion loci were predicted on the migration of PCR fragments in agarose gel electrophoresis. The PCR fragments were prepared with various combinations of forward primers and reverse primers as previously described (23, 24): The forward primers were follows: primer 1, 5′-CACTCTCCTTTGATGCGAATGCCAGCGTCAGAC-3′; primer 5, 5′AAACAGGCGCGGGCATCAGCGTGGTCTGA-3′; primer s001, 5′-GTTACAGAGGTTCGTCCGGGAAC-3′; primer s002, 5′-GAAATCACTCCCGGGTATATGAA-3′; primer s003, 5′-ATTTGATTTCAATTTTGTCCCACTC-3′; primer s101, 5′-CCACTTCATTTCGCATAAATCACCAACT-3′; primer s102, 5′-AATCCAAACTCTTTACCCGTCCTTGGGT-3′; primer s201, 5′-CGCTTCGATAACTCTGTTGAATGGCTCT-3′; primer s202, 5′-TGAGTACGGTCATCATCTGACACTACAG-3′; primer s203, 5′-AGTTCCAGCACAATCGATGGTGTTACCA-3′; primer s301, 5′-GGTGTGAATCCCATCAGCGTTACCGTTT-3′; primer s302, 5′-AGTGATTGCGCCTACCCGGATATTATCGTG-3′; primer s401, 5′AGCCACACCGGTGCAAACCTCAGCA-3′; primer s402, 5′GGCTTTATGACGTAACATCCGTTTGGG-3′; primer s403, 5′CCATGCCGACACGTTCAGCCAGCTTCCCAG-3′. The reverse primers are as follows: primer 2, 5′-CAGGAGTAATTATGCGGAACAGAATCATGCCTGGTG-3′; primer 12, 5′CGCGGCCGGTCGAGGGACCTAATAACTTCGTA-3′; primer sA, 5′-CCACTTTTATTGGCGATGAAAAGATGTTTC-3′. Then the closest primer to the expected deletion region was selected as a sequencing primer from the PCR primers.
Results Mutant Frequencies in the Spi- and 6-TG Selections. To examine the genotoxic effects of MMC in the bone marrow, 12 mice were injected with MMC at 1 mg/ kg for five successive days, and six mice each were sacrificed 1 week and 4 weeks after the last treatment. As a control group, seven mice were treated with saline and sacrificed 4 weeks after the last injection. In both Spi- and 6-TG selections, the MFs increased significantly 1 week after MMC treatment (Figure 1). The increases in MFs were 2.9-fold in Spi- selection (5.2 × 10-6 versus 1.8 × 10-6) and 1.7-fold in 6-TG selection (14.1 × 10-6 versus 8.2 × 10-6) over the control levels. The increases were statistically significant (p < 0.005 in Spi- selection and p < 0.0001 in 6-TG selection). The MFs decreased to the vehicle control levels 4 weeks after the last treatment. Characteristics of the Spi- Mutation Spectrum. To clarify the specificity of MMC-induced mutations, 33 mutants from the MMC 1-week expression time group
Molecular Nature of Mitomycin C-Induced Mutations
Figure 1. Mutant frequencies in the bone marrow of MMCtreated and vehicle-treated gpt delta mice: (A) Spi- mutant frequencies; (B) gpt mutant frequencies. The samples were taken 1 week after the last treatment (MMC 1 week) or 4 weeks after the last treatment (MMC 4 weeks). MMC was administered by i.p. injection at a dose of 1 mg/kg for 5 consecutive days. Control cells were collected 4 weeks after the mice received five consecutive injections of saline (Control). p values calculated by Chi-square test were p < 0.005 (*) and p < 0.0001 (**). Bars represent mean values and standard deviations.
and 14 mutants from the control group were subjected to molecular analysis. These mutants were classified into five classes based on the deletion sizes and the sequence characteristics of the junction region, as described below. In the MMC-treated group, class 1 included 10 large deletion mutants that possessed short homologous sequences [2-6 base pairs (bp)] in their junctions and one of the short homologous sequences were eliminated when two DNA ends were joined (Figure 2). The deletion sizes were more than 2 kb but less than 7 kb. Class 2 included five large deletions that did not exhibit short homologous sequences at the junctions. These deletions ranged from 110 bp to about 8 kb. Interestingly, three of five mutants had 4 or 5 bp with short homologous sequences in the vicinity of the junctions. Unlike the short homologous sequences in the class 1 mutants, the homologous sequences in the vicinity of the junctions in this class were not eliminated when two DNA ends were joined. One
Chem. Res. Toxicol., Vol. 16, No. 2, 2003 173
mutant having a 110 bp deletion possessed two TGTT sequences: one in the junction and the other 4 bp away from the junction, i.e., 5′-TGTTcaccTGTT-3′. Another mutant having a 6768 bp deletion possessed two GGCA sequences at the junction and in its vicinity, i.e., 5′GGCA(ctctcag)taactctcaGGCA-3′. It had a 7-bp inserted sequence, shown in parentheses, whose sequence was the same as the 3′-sequence that is underlined. The other mutant having a 8326 bp deletion possessed two GCGCA sequences, both in the vicinity of the junction, and had a 5 bp insertion at the junction, i.e., 5′-GCGCAaga(tagat)ttGCGCA-3′, where the inserted sequence is shown in parentheses. In this case, no sequences similar to or the same as the inserted sequence were found in the vicinity of the junction. The mutant having a 110 bp deletion could make plaques on recA- mutants of E. coli (see below). Class 3 included three single-bp deletions that occurred in nonrepetitive sequences in the gam gene (Figure 3). Two of them had both single-bp deletions and base substitutions, such as 5′-gCC-3′ to 5′-gA-3′ and 5′CAGGat-3′ to 5′-TTTat-3′. Class 4 included nine singlebp deletions that occurred in the repetitive sequences in the gam gene. Three of them occurred in nucleotides 227-231, i.e., 5′-AAAAA-3′; two of them each in nucleotides 295-300, i.e., 5′-AAAAAA-3′; and two of them each in nucleotides 286-289, i.e., 5′-GGGG-3′. Class 5 included six mutants that had base substitutions and/or insertions but without deletions in the gam gene (Figure 3). Base substitutions in the four mutants created a stop codon. All the mutants in this class made plaques on recA- mutants of E. coli, whereas none of the other Spi- mutants classified into classes 1-4 made plaques except for the mutant having 110 bp deletion. In the control group, the most predominant mutation was single-bp deletion (12/14). Most of the deletions (11/ 12) occurred at repetitive sequences in the gam gene (class 4). Of 11 mutants, five of them occurred in nucleotides 295-300, three mutants occurred in nucleotides 227-231, two mutants occurred in nucleotides 286-289, and one mutant occurred in nucleotides 316318, i.e., 5′-TTT-3′. A deletion mutant occurring in a nonrepetitive sequence had a change of 5′-cAg-3′ to 5′cg-3′ (class 3). Others (2/14) were single-base substitutions that occurred in a nonrepetitive sequence in the gam gene, i.e., 5′-taC-3′ to 5′-taA(Stop)-3′ (class 5) and a large deletion mutation of 4439 bp (class 2). Specific MF of Large Deletions. To further characterize the deletion mutations induced by MMC, we compared the specific MF of large deletions (classes 1 and 2) and single-bp deletions (classes 3 and 4) (Table 1). We did not calculate the specific MF of class 5 (base substitutions) because they had no deletions. Although the total MF of the Spi- in the MMC 1-week expression time group was 2.9 times higher than that of the Spi- in the control group, the specific MF of large deletions was 24 times higher than that of the control. In contrast, the specific MF of single-bp deletions was not changed significantly by MMC treatment. These results strongly suggest that MMC induces large deletions but not single-bp deletions in the bone marrow. Characteristics of the gpt Mutation Spectrum. To characterize the point mutations induced by MMC, 29 and 30 gpt mutants from the bone marrow of the control group and the MMC 1-week expression time group, respectively, were subjected to DNA sequence analysis.
174
Chem. Res. Toxicol., Vol. 16, No. 2, 2003
Takeiri et al.
Figure 2. Large deletions (classes 1 and 2 mutants) recovered from MMC-treated and vehicle-treated mice and a partial genetic map of the lambda EG10 transgene, including the gam and the redBA target region of Spi- selection. Horizontal open bars represent regions deleted in Spi- mutants. Class 1 includes 10 large deletions that exhibit short homologous sequences (2-6 bp) in their junctions. One of two short homologous sequences in the junctions that are underlined is deleted when two DNA fragments are joined. Class 2 includes six large deletions (five from the MMC-treated group and one from the control) that do not exhibit homology at the junctions. Short homologous sequences in the vicinity of the junctions are highlighted in italicized bold capitals. Inserted bases are highlighted in nonitalicized bold capitals. The sequence that is the same as the inserted DNA is highlighted with a dotted underline. Junctions are indicated as the space between the left and right sequences. Table 1. Spi- MF of Large Deletions (Class 1 and 2) and 1-bp Deletions (Classes 3 and 4) Recovered from the Bone Marrow of MMC-Treated and Vehicle-Treated gpt delta Mice large deletionsa
1-bp deletionsb
specific MF increase above specific MF increase above (× 10-6)c the control (× 10-6) the control control MMCd
0.1 2.4e
1.0 24
1.5 1.9 f
1.0 1.3
Large deletions include class 1 and 2 mutants. b 1-bp deletions include class 3 and 4 mutants. c Specific MF was calculated by multiplying the MF by the ratio of the number of class 1 and 2 (or class 3 and 4) mutants among the total number of Spi- mutants. d MMC stands for the group sacrificed 1 week after the last treatment (MMC 1-week expression time group). e Significantly different from the vehicle control (p < 0.001, Chi-square test).f No significant difference from the vehicle control (p ) 0.79, Chi-square test).
fied only three such mutants, we are not certain that this class of mutation is specific to the MMC-treated group. The complex mutations included two double-base substitutions within two or four neighboring bp and a sequence substitution, i.e., the substitution of a 20 bp sequence to a different 18 bp sequence with direct repeat sequences.
Discussion
a
Although the gpt MF of the MMC-treated group was only 1.7 times higher than that of the control group, the spectra of mutations were strikingly different (Table 2, Figure 4). All the mutations observed in the control group were either single-bp substitutions (transitions and transversions) or single-bp deletions. The specific MFs of single-bp substitutions and single-base deletions were very similar between the control and MMC-treated groups, and about 70% of base substitutions occurred at G:C bp, both in the control and the MMC-treated groups. In contrast, tandem-base substitutions were only observed in the MMC-treated group, suggesting that this type of mutation was specifically induced by MMC. The tandem-base substitutions were predominantly observed at the 5′-GG-3′ site (4/7 ) 57.1%). Other tandem substitutions were induced at the 5′-CG-3′, 5′-AC-3′, and 5′TG-3′ sites. Complex mutations were also only observed in the MMC-treated group. However, because we identi-
In this study, we have examined the in vivo mutations induced by MMC and demonstrated that MMC induces large deletions and tandem-base substitutions mutations in the bone marrow. The mice were treated with MMC at a dose of 1 mg/kg for five successive days, and the mutations were characterized 1 week after the last treatment. Although the total MF of Spi- in the MMCtreated group was only about three times higher than that in the control (Figure 1), the specific MF of classes 1 and 2 mutants in the MMC-treated group (large deletions) was 24 times higher than that of the control (Table 1). Even if we regarded the repetitive four mutants of 2482 bp deletion from mouse m3 as a result of clonal expansion, the specific MF of large deletions was still more than 20 times higher than that of the control (MMC-treated group, 2.2 × 10-6; control, 0.1 × 10-6). The molecular sizes of large deletions were more than 2 kb except in one mutant (Figure 2). In addition, we revealed that MMC specifically induces tandem-base substitutions, mainly in 5′-GG-3′ sequences (Table 2). It should be noted that no single-base transitions and transversions or single-guanine-base deletions were induced by the treatments (see below). To our knowledge, this is the first report in which MMC-induced deletions and point mutations in vivo are systematically characterized at the molecular level.
Molecular Nature of Mitomycin C-Induced Mutations
Chem. Res. Toxicol., Vol. 16, No. 2, 2003 175
Figure 3. Spi- mutants class 3, 4, and 5 in the gam gene. The sequence from top to bottom represents the coding region of the gam gene. Mutations shown below the strands were detected in MMC-treated mice, and those above the strands were detected in control mice. (∆) Single-base deletions. (∧) One-base insertion. The numbers in parentheses indicate the class of the each mutant. Five complex base substitutions are underlined: in nucleotides 139-142, the sequence CAGG was changed to TTT; in nucleotides 177178, the sequence CC was changed to aA; in nucleotides 290-291, the sequence CC was changed to A; in nucleotides 311, the G was changed to CT; in nucleotide 366-370, the sequence CATGG was changed to aATCT. The small letters in the substituted sequences formed stop codon with adjoining two bases. Two of five -A mutations in nucleotides 295-300 in vehicle control group were identified in the same mouse. Two -G mutations in nucleotides 286-289 in vehicle control and MMC-treated groups were identified in the same mice. Table 2. Summary of the gpt Mutations Recovered from the Bone Marrow of MMC-Treated and Vehicle-Treated gpt delta Mice MMCa
control no. of mutants (%)
specific MFb (× 10-6)
no. of mutants (%)
specific MF (× 10-6)
single base transitions G:C to A:T A:T to G:C single base transversions G:C to T:A G:C to C:G A:T to T:A A:T to C:G deletions -A -G tandem base substitutions GG:CC to AT:TA GG:CC to TA:AT GG:CC to TT:AA CG:GC to AA:TT AC:TG to TA:AT TG:AC to CT:GA complexd
8 7 1 17 9 3 2 3 4 3 1 0 0 0 0 0 0 0 0
(27.6) (24.1) (3.4) (58.6) (31.0) (10.3) (6.9) (10.3) (13.8) (10.3) (3.4) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
2.3 2.0 0.3 4.8 2.5 0.8 0.6 0.8 1.1 0.8 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6 4 2 12 8 2 1 1 2 2 0 7 2 1 1 1 1 1 3
(20.0) (13.3) (6.7) (40.0) (26.7) (6.7) (3.3) (3.3) (6.7) (6.7) (0.0) (23.3) (6.7) (3.3) (3.3) (3.3) (3.3) (3.3) (10.0)
2.8 1.9 0.9 5.6 3.8 0.9 0.5 0.5 0.9 0.9 0.0 3.4 0.9 0.5 0.5 0.5 0.5 0.5 1.4
total
29 (100)
8.2
30 (100)
14.1
p valuec 0.49 0.15
0.37
