Proc. Natl. Acad. Sci. USA
`Vol. 90, pp. 3904-3907, May 1993
`Genetics
`
`Tandem double CC -> TT mutations are produced by reactive
`oxygen species
`THOMAS M. REID AND LAWRENCE A. LOEBt
`Joseph Gottstein Memorial Cancer Research Laboratory, Department of Pathology, University of Washington School of Medicine, Seattle, WA 98195
`
`Communicated by Irwin Fridovich, January 8, 1993
`
`Oxidative damage to DNA is mutagenic and
`ABSTRACT
`thus may play a role in carcinogenesis. Because of the large
`number of different DNA lesions formed by oxidative species,
`no genetic alteration so far identified is exclusively associated
`with oxygen damage. Tandem double CC -- TT mutations are
`known to occur via UV damage to DNA and are thought to be
`a specific indicator of UV exposure. Using a sensitive reversion
`assay that can detect both single and double mutations within
`the same codon of the M13-encoded lacZa gene, we show that
`treatments that produce reactive oxygen species can also pro-
`duce tandem double CC -k TT mutations. The frequency at
`which these mutations occur is less than that for single base
`mutations by a factor of approximately 30. The induction of
`these mutations is inhibited by treatment that scavenges hy-
`droxyl radicals. This unique mutation provides a marker of
`oxygen free radical-induced mutagenesis in cells that are not
`exposed to UV-irradiation and an indicator for assessing the
`involvement of oxidative damage to DNA in aging and tumor
`progression.
`
`Oxygen free radicals have been implicated in a number of
`degenerative diseases including aging and cancer (1-4).
`These reactive species arise through normal cellular pro-
`cesses, inflammatory events, ischemia, and xenobiotic me-
`tabolism (5, 6). The ability ofoxygen radicals to damage DNA
`is well documented (7, 8). This damage has been hypothe-
`sized to be a factor in the initiation and/or promotion of
`malignancies, but as yet there is no direct proof of this. A
`major problem is that the multitude of lesions that oxygen
`radicals produce in DNA makes it difficult to assign a
`particular mutation to a given oxidative lesion. In contrast,
`many chemical carcinogens produce a limited number of
`DNA lesions and base changes, thus allowing a correlation
`between the mutational spectrum found in commonly mu-
`tated genes oftumors and the proposed etiologic agent (9, 10).
`This use of molecular epidemiology was recently extended
`in studies that showed a significant proportion of individuals
`with squamous cell skin carcinoma had tandem double CC
`TT mutations in the p53 genes of these tumors (11). This
`unique mutation has been thought to be a specific indicator
`of UV-irradiation and provided a means of substantiating the
`involvement of sunlight in the origin of the skin carcinomas.
`Recently however, in studies using metal ions or human
`neutrophils as sources of oxygen radicals (12, 13), we have
`also observed tandem double CC -+ TT mutations, suggest-
`ing that reactive oxygen species and UV light may produce
`a common intermediate in DNA that leads to a tandem double
`mutation.
`To determine the potential for reactive oxygen species to
`produce tandem double CC -+ TT mutations, we have
`adapted a reversion assay (14) that is specific for damage to
`cytosine residues and can detect both single and tandem
`double mutations at the same locus. We have used this assay
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertisement"
`in accordance with 18 U.S.C. §1734 solely to indicate this fact.
`
`to measure the frequency of tandem double CC -- TT
`mutations produced by Fe2+ or a combination of Cul+ or
`Cu2+ plus H202, to demonstrate the inhibition of mutagenesis
`by a hydroxyl radical scavenger and to compare oxygen
`radical-induced mutagenesis with that produced by UV light.
`The data suggest that tandem double CC -* TT mutations
`might provide a marker to assess the involvement of oxida-
`tive damage to DNA during aging or tumor progression.
`
`MATERIALS AND METHODS
`Escherichia coli Strains. E. coli strain MC1061 [hsdR,
`mcrB, araD, 139A(ara-leu), 7679D, lacX74, galU, galK,
`rpsL, thi] was the host strain used for all transfection exper-
`iments. Cells were plated on the indicator strain E. coli
`CSH50 [A(pro-lac) thi, ara/F' traD36, proAB, lacIZAM15).
`The UV dose used to induce the SOS response in E. coli
`MC1061 was 46 J/m2.
`Treatment of DNA. The construction and sequence of
`M13G*1 DNA have been described (14). Single-stranded
`DNA was treated with Cu or Fe as described (12, 15) or
`irradiated with 254-nm light in 5-,l drops in an uncovered
`plastic Petri dish.
`Scoring Mutants and Analysis of DNA Sequences. Trans-
`fection protocols were as described (12, 15), except that
`aliquots of the transfection mixtures were plated to yield
`plaque densities of no more than 3000 plaques per plate.
`Mutation frequency was determined by dividing the number
`ofblue plaques within a given experimental group by the total
`combined number of white and blue plaques. Statistical
`analysis was conducted according to Birnbaum as described
`by Kastenbaum and Bowman (16). After isolation and re-
`plating of mutant plaques to ensure genetic purity of the
`samples, the DNA was either sequenced or the plaques
`themselves were probed with oligonucleotides specific for
`either possible CC -- TT double mutation. Two oligonucle-
`otides were used for hybridization analysis, either 5'-
`GCCAGCTAAGAAATGGG-3' (CTT probe) or 5'-GC-
`CAGCTGAAAAATGGG-3' (TTC probe). After binding of
`the plaques to nylon filters and incubation with 32P-labeled
`oligonucleotides, the filters were washed at 50°C as described
`(17, 18). Purified plaques representing CCC -- TTC or CCC
`-* CTT double mutations were included on each filter as an
`internal control. In control experiments it was determined
`that the oligonucleotide probes did not bind to samples
`containing single base mutations after the 50°C wash. Tan-
`dem double mutations identified by oligonucleotide hybrid-
`ization were confirmed by sequence analysis.
`
`RESULTS
`Principles of the Mutation Assay. The M13G*1 reversion
`assay has previously been used to detect single mutations
`produced by misincorporation across from single 8-hydroxy-
`guanine or N2,3-ethenoguanine lesions inserted at a specific
`
`tTo whom reprint requests should be addressed.
`
`3904
`
`GDX 1015
`
`

`

`Genetics: Reid and Loeb
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`3905
`
`141
`
`ccc
`
`(I
`
`facZ
`
`/
`
`Tran00
`Fe/Cu/UV
`SSOS-Induced
`-u/u
`
`X-gal
`9
`Medium
`
`0
`o0o
`
`M13G*1
`(white)
`
`0
`
`0
`
`0
`
`0
`
`0 0 0
`
`0 V0
`
`SS M13G*1
`DNA
`
`Revertant
`(blue)
`
`Principles of the M13G*1 reversion assay. The CCC
`FIG. 1.
`codon at positions 141-143 of the lacZa gene is the target for
`mutagenesis. Any single base substitution at the first two positions
`of this codon or any tandem double CC -b TT mutation causes a
`reversion in plaque phenotype from white to dark blue. X-Gal,
`5-bromo-4-chloro-3-indolyl f-D-galactoside.
`
`single base substitutions were detected at the first two
`positions of the CCC codon (Fig. 2). However, C -* T
`transitions were by far the most prevalent mutation produced
`by either oxygen damage or UV-irradiation. The high fre-
`quency of C -+ T transitions appears to be characteristic of
`mutagenesis by oxygen-reactive species (12, 15). This bias
`toward C -* T substitutions has also been reported in the cI
`gene of A phage after exposure to y-irradiation (21). Approx-
`imately twice as many single C -- T mutations occurred at the
`first position of the codon as at the second for both oxygen
`and UV treatment. Whether this is due to a difference in the
`degree to which each base is modified or a bias toward
`insertion of a serine (TCC) rather than a leucine (CTC) at this
`position is not known.
`Tandem Double Mutations Due to Oxidative DNA Damage.
`While single mutations occurred at a frequency of 41/2000,
`tandem double CC -+ TT mutations were less frequent by a
`factor of about 30. No tandem double mutations were ob-
`served among the 349,000 plaques obtained with DNA that
`was not exposed to agents that generate reactive oxygen
`species (Table 3). In contrast, damage to M13G*1 DNA by
`either Fe2+ or Cu/H202 resulted in the production of CC --
`TT substitutions. Fe2+ tended to produce mutations at the
`first two positions of the CCC codon while Cu/H202 did not
`show this specificity. The diminution of the Fe2+-induced
`mutations by the addition of mannitol provides evidence that
`these mutations are dependent on the production of reactive
`oxygen species. UV-irradiation was -3-fold more efficient
`per lethal event in producing tandem double mutations than
`oxidative damage. Tandem double CC -+ TT mutations
`accounted for -3% of the total mutations produced by
`oxidative damage; -7% of the UV-induced mutations were
`tandem doubles. The lack of inhibition of UV-induced mu-
`tagenesis by mannitol indicates that the UV-induced muta-
`genesis is not mediated by the generation of hydroxyl radi-
`cals.
`
`DISCUSSION
`
`Tandem double CC -* TT mutations have been thought to be
`a specific indicator of UV damage to DNA and have been
`detected after UV-irradiation in bacteria (22-24), yeast (25),
`and primate cells (26). This mutation has also been identified
`in the p53 gene of squamous cell skin carcinomas of human
`subjects (11), presumably as a result of UV exposure. We are
`unaware of any report of this mutation occurring because of
`misincorporation by DNA polymerases or after exposure of
`DNA to chemical carcinogens. In fact, no tandem double CC
`-* TT mutations have been observed in the lacZa gene in at
`least 10,000 sequenced mutants arising from DNA polymer-
`ase errors, the insertion of site-specific chemical adducts, or
`other types of DNA damage. Recently, however, tandem
`double CC -+ TT mutations have been detected in forward
`mutation assays when using either Cu ions (12) or human
`neutrophils as a source of reactive oxygen species (13). The
`goal of this study was to determine whether oxygen free
`radicals were responsible for inducing this mutation and
`
`site in a double-stranded DNA template (14, 19). Essentially,
`the M13G*1 construct has an altered codon (GCC -* CCC)
`within the lacZa gene that yields white plaques rather than
`the dark blue of the parental M13mp2 DNA (Fig. 1). Any
`single base substitution at the first position of this altered
`codon causes a reversion to wild-type dark-blue phenotype.
`In a similar system, Frederico et al. (20) showed that single
`C -* T transitions at the second position ofthis codon are also
`scorable but that mutations at the third position of this codon
`are phenotypically silent. Our analysis of the amino acid
`substitutions that result in reversion to wild type indicated
`that certain tandem double mutations could also be scored,
`including CCC -+ TTC and CCC -* CTT mutations. Thus,
`this DNA construct allows detection of both single and
`double C -- T mutations within the same codon of the lacZa
`gene.
`Mutagenesis Due to Oxidative DNA Damage. In previous
`studies we have shown that Fe and Cu ions are mutagenic and
`that the mutagenicity is dependent upon the generation of
`oxygen-free radicals or other reactive oxygen species (12,
`15). These metal ions also induce mutations in the M13G*1
`reversion assay as summarized in Table 1. Fe2+ or a combi-
`nation of Cul+ or Cu2+ with H202 produced an average 7-fold
`increase in mutation frequency at the CCC locus of M13G*1.
`Each of these treatments also decreased the biological ac-
`tivity of the DNA by -90%. The addition of the hydroxyl
`radical scavenger mannitol to the Fe2+-catalyzed reactions
`produced both a doubling in survival and a halving in overall
`mutation frequency. In parallel experiments, single-stranded
`M13G*1 DNA was irradiated with 254-nm light and assayed
`for mutagenesis. This treatment produced a 9-fold increase in
`mutation frequency. In contrast to the results with Fe2+, the
`presence of mannitol did not decrease mutation frequency for
`UV-irradiated DNA. However, it did increase the biological
`activity of the DNA relative to that without mannitol.
`DNA Sequence Analysis of Single Base Substitutions. The
`nucleotide sequence of -200 of the revertants produced by
`exposure to metal ions or UV light was analyzed by sequence
`analysis (Table 2). It is interesting to note that all possible
`Mutation frequency and survival of M13G*1 DNA after oxidative or UV damage
`Plaques scored
`Mutation frequency
`x 10-4
`Total
`Mutants
`348,953
`0.7
`25
`123,657
`6.0
`74
`296,866
`5.1
`150
`5.4
`344,402
`185
`431,382
`2.8
`119
`141,626
`6.4
`91
`247,044
`7.9
`195
`
`Table 1.
`
`Treatment
`
`Control
`Cul+ (2.5 ,uM) + H202 (5 MuM)
`Cu2+ (1.0 gM) + H202 (2.5 uM)
`Fe2+ (5,M)
`Fe2+ (5 MM) + mannitol (0.1 M)
`UV (25 J/m2)
`UV (25 J/m2) + mannitol (0.1 M)
`
`% survival
`100
`12
`10
`10
`22
`3
`9
`
`GDX 1015
`
`

`

`3906
`
`Genetics: Reid and Loeb
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`Analysis of DNA sequence changes in M13G*1 DNA
`Table 2.
`after oxidative or UV damage
`Codon from CCC mutation
`TCC
`CAC
`CGC
`GCC
`ACC
`11
`2
`4
`2
`23
`2
`2
`4
`4
`33
`2
`2
`2
`9
`23
`3
`1
`1
`3
`Total
`90
`9
`9
`7
`18
`% total
`48
`5
`5
`4
`10
`tlncludes Cul+ or Cu2+ in combination with H202-
`
`Sample
`Control
`Cut
`Fe
`UV
`
`CTC
`6
`15
`13
`10
`44
`24
`
`whether the mechanism of its formation is similar to UV-
`induced double mutations. To do this we have utilized a
`reversion assay that can specifically detect mutations at
`cytosine residues in DNA and assayed for tandem double CC
`-- TT substitutions by DNA sequencing or plaque hybrid-
`ization.
`Fig. 3 is a compilation of oxidatively induced tandem
`double mutations observed in this laboratory in previous
`studies (12, 13) as well as the reversion assay described here.
`The ability of Fe, Cu, and phorbol ester-stimulated neutro-
`phils to cause CC -) TT mutations and the multiplicity of sites
`at which they occur suggest that CC -* TT mutations are a
`general manifestation of oxygen damage to DNA. The com-
`mon sequence motif among the sites at which CC -- TT
`mutations have been detected is the presence of at least three
`consecutive cytosines. In fact, there is only one stretch of
`three consecutive cytosines within the target region of the
`lacZa gene in which this mutation has not been detected. In
`contrast, UV-induced tandem double mutations have been
`shown to occur at dicytosine sites or runs of three or more
`cytosines with equal efficiency (23). The dependence of CC
`-+ TT mutations on induction of SOS functions is not known,
`nor is it established that a similar error-prone response occurs
`in human cells. Previous studies have shown an enhancement
`of mutagenicity by oxygen radicals when using SOS-induced
`cells (15, 27). Because of the rarity of CC -* TT mutations,
`we have elected to use SOS-induced cells to increase the
`sensitivity with which they can be detected.
`
`T-T
`T-T
`T-T
`T-T
`T-T
`T-T
`T-T
`T-T
`
`T-T
`T-T
`T-T
`T-T
`T-T
`
`5'-C-C-A-T-T-T- C
`
`C
`
`C -A-G-C-T-G-G-3'
`
`TTTTTTT
`§§TTTTTTT
`I I
`
`T t1
`r
`AAAA
`GGGG
`
`,,
`
`TTTTr
`7i7777T
`717i7i7
`rirlrrT
`lilllll
`AAAAAA
`AAAAAAA
`GGGGGG
`
`Single and tandem double mutations at the CCC locus of
`FIG. 2.
`M13G*1. The sequence shown is the viral strand of M13G*1 from
`position 135 to 149, where position 1 is the first transcribed base.
`Single base substitutions are shown below the sequence. Tandem
`double mutations are shown above the sequence. Single and double
`mutations produced by exposure to Fe2+ are shown in italicized bold
`type; those produced by Cu/H202 are shown in regular type.
`
`Table 3. Tandem double mutations in M13G*1 DNA after
`oxidative or UV damage
`
`Codon from
`CCC
`mutation
`Plaques
`Total
`CTT
`TTC
`scored
`Treatment
`-
`0
`348,953
`Control
`6t
`Cu*
`3
`3
`385,979
`St
`1
`4
`344,402
`Fe
`2
`1
`1
`431,382
`Fe + mannitol
`7t
`5
`2
`141,626
`UV
`13t
`UV + mannitol
`5
`8
`247,044
`*Includes Cul+ or Cu2+ in combination with H202-
`tSignificantly greater than control DNA (P c 0.05).
`tSignificantly greater than control DNA (P c 0.01).
`
`Mutation
`frequency
`X 10-5
`<0.3
`1.6
`1.5
`0.5
`4.9
`5.3
`
`Oxygen radicals cause a large number of modifications to
`cytosine residues, and single C -* T transitions are among the
`most common mutations produced by most (12, 15) but not
`all (27) reactive oxygen species. Chemical modifications
`produced by reactive oxygen at cytosine residues include
`addition to the 5,6 double bond of the pyrimidine ring,
`formation of glycols, hydantoins, deamination products, and
`species with rearrangements of the pyrimidine ring (7, 8).
`Based on the frequency of two independent hits, it is unlikely
`that the same adduct causes both the single and double
`substitutions. The frequency of CCC -* TTC mutations
`produced by Cu/H202 is 1000-fold higher than would be
`expected from two random independent events (i.e., CCC -*
`TCC plus CCC -+ CTC). Treatments designed to induce
`deamination of cytosine residues have been shown to cause
`tandem double CC -- TT mutations (28). However, it should
`be noted that the tandem double mutations observed in that
`study were dependent on the use of host cells that do not
`possess a functional uracil glycosylase gene. All tandem
`double CC -) TT mutations observed in the present study
`were produced in cells containing a functional uracil glyco-
`sylase. Moreover, we have evidence that the frequency of
`both single and double C -* T mutations produced by reactive
`oxygen species is independent of uracil glycosylase activity
`(data not shown). Thus, it seems more likely that a single
`oxygen-radical-induced modification at adjacent cytosine
`residues is responsible for the CC -* TT mutations and that
`this modification is different from those that elicit single C
`T substitutions.
`A primary candidate for an alteration in DNA that could
`produce the double mutation is the crosslinking of two
`adjacent cytosines. There is evidence that exposure of dinu-
`cleotides to oxidative damage mediated by Cu/H202 results
`in a crosslink between the two bases, although cytosine
`crosslinks were not assayed (29). The finding that CC -> TT
`substitutions are observed after damage to DNA by both
`reactive oxygen species and UV-irradiation suggests that
`both types of agents may produce a similar intermediate,
`leading to a stable mutagenic adduct. However, the ability of
`the hydroxyl radical scavenger mannitol to decrease the
`frequency of oxidatively induced tandem double mutations
`but not those induced by UV light argues against the pro-
`duction of classical UV-induced cyclobutane pyrimidine di-
`mers and (6-4) pyrimidine-pyrimidone photoproducts by
`reactive oxygen species. A more likely modification is sug-
`gested by studies on the effect of hydroxyl radicals generated
`by -y-irradiation of cytosine (30). Fifty percent of the products
`detected in N20-saturated aqueous solution were dimerized
`cytosines structurally distinct from UV-induced photoprod-
`ucts. It remains to be established whether this particular
`modification is responsible for CC -- TT mutations.
`
`GDX 1015
`
`

`

`Genetics: Reid and Loeb
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`3907
`
`TT
`TT
`TT
`TT
`TT
`TT
`TT
`TT
`TT
`TT
`Ur
`TT
`TT
`TT
`rr
`TT
`TT
`TT
`TT
`TT TT
`TT
`15'-GAAAACCCTGGCGGTACCCAACTTAATCGCCTTGCAGCACATCCCCCATTTCCCAGCTGG-3'1
`I TT
`140 TT
`120
`100
`TT
`TT
`TT
`
`FIG. 3. Tandem double mutations in the lacZa gene. The sequence shown is the viral strand of M13 from position 90 to 149, where position
`1 is the first transcribed base. The G -- C change that produces M13G*1 is at position 141. The tandem double mutations include those observed
`in previous studies produced by Cu/H202 at positions 107 and 108 (12) and positions 131 and 132 (unpublished results); by phorbol
`ester-stimulated neutrophils at positions 96 and 97 and 135 and 136 (13); and by Fe and Cu/H202 in the present study at positions 141-143.
`
`Carcinogenesis appears to be a multistep process involving
`the sequential accumulation of mutations that allows cells to
`gain a selective growth advantage and undergo clonal expan-
`sion (10). It is frequently hypothesized that many of these
`mutations are generated by cellular processes and in partic-
`ular by oxygen free radicals (31). Seventy percent of human
`tumors have been shown to contain mutations in the p53
`gene, and specific mutations in this gene have been used to
`identify environmental agents that are associated with the
`genesis of specific tumors (9, 10). To date there are only two
`reports of tandem double CC -f TT mutations occurring in
`the p53 gene. They have been detected in skin carcinomas
`(11), where UV light is the presumed agent, and in a non-
`small-cell lung carcinoma (32), where UV light would not be
`expected to play a role. It is interesting to note that the CC
`-* TT mutations observed in the skin carcinomas all occurred
`at dipyrimidine sites in the gene, whereas the CC -k TT
`change observed in the p53 gene of the non-small-cell lung
`tumor occurred within a run of three consecutive cytosines
`similar to the mutations observed in this study. The lack of
`CC -> TT mutations in the p53 gene obtained from most
`human tumors suggests that mutagenesis by oxygen free
`radicals is not a major contributor to the overall accumulation
`of mutations that characterizes tumor progression. However,
`it should be noted that while there are 14 tracts of three or
`more consecutive cytosines within the region of the p53 gene
`most commonly mutated in tumors (9), it is not likely that
`mutations at each of these sites would give rise to a dysfunc-
`tional p53 gene product or one that would lead to a selective
`growth advantage. In fact, of the 22 codons encompassed in
`the tracts of three or more consecutive cytosines, only 7 have
`been shown to be altered in tumors, and only 2 of these
`TT mutations
`changes could be due to tandem double CC
`(9). Thus, the paucity of tandem double CC
`TT mutations
`in the p53 gene is consistent with the rarity of this mutation
`and the relatively low number of target sites compared with
`single base substitutions. In cells not exposed to UV the
`presence of this rare mutation would provide strong evidence
`for the involvement of oxidative damage.
`
`These studies were supported by National Institutes of Health
`Grants CA08855 (to T.M.R.) and CA39903 and AG01751 (to L.A.L.).
`
`1.
`2.
`3.
`4.
`
`Ames, B. N. (1983) Science 221, 1256-1264.
`Joenje, H. (1989) Mutat. Res. 219, 193-208.
`Cerruti, P. A. (1985) Science 227, 375-381.
`Totter, J. R. (1980) Proc. Natl. Acad. Sci. USA 77, 1763-1767.
`
`5.
`6.
`
`7.
`
`8.
`9.
`
`10.
`11.
`
`12.
`
`13.
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`23.
`
`Klebanoff, S. J. (1980) Ann. Int. Med. 93, 480-489.
`Fridovich, I. (1983) Annu. Rev. Pharmacol. Toxicol. 23, 239-
`257.
`Teebor, G. W., Boorstein, R. J. & Cadet, J. (1988) Int. J.
`Radiat. Biol. 54, 131-150.
`Halliwell, B. & Aruoma, 0. I. (1991) FEBS Lett. 281, 9-19.
`Hollstein, M., Sidransky, D., Vogelstein, B. & Harris, C. C.
`(1991) Science 253, 49-53.
`Harris, C. C. (1991) Cancer Res. 51 (Suppl.), 5023s-5044s.
`Brash, D. E., Rudolph, J. A., Simon, J. A., Lin, A., McKenna,
`G. J., Baden, H. P., Halperin, A. J. & Ponten, J. (1991) Proc.
`Natl. Acad. Sci. USA 88, 10124-10128.
`Tkeshelashvili, L. K., McBride, T., Spence, K. & Loeb, L. A.
`(1991) J. Biol. Chem. 266, 6401-6406.
`Reid, T. M. & Loeb, L. A. (1992) Cancer Res. 52, 1082-1086.
`Cheng, K. C., Preston, B. D., Dosanjh, M. K., Singer, B. &
`Loeb, L. A. (1991) Proc. Natl. Acad. Sci. USA 88,9974-9978.
`McBride, T., Preston, B. D. & Loeb, L. A. (1991) Biochem-
`istry 30, 207-213.
`Kastenbaum, M. A. & Bowman, K. 0. (1970) Mutat. Res. 9,
`527-549.
`Gupta, P. K., Johnson, D. L., Reid, T. M., Lee, M.-S., Ro-
`mano, L. J. & King, C. M. (1989) J. Biol. Chem. 266, 20120-
`20130.
`Reid, T. M., Lee, M.-S. & King, C. M. (1990) Biochemistry 29,
`6153-6161.
`Cheng, K. C., Cahill, D. S., Kasai, H., Nishimura, S. & Loeb,
`L. A. (1992) J. Biol. Chem. 267, 166-172.
`Frederico, L. A., Kunkel, T. A. & Shaw, B. R. (1990) Bio-
`chemistry 29, 2532-2537.
`Tindall, K. R., Stein, J. & Hutchinson, F. (1988) Genetics 118,
`551-560.
`Miller, J. H. (1985) J. Mol. Biol. 182, 45-68.
`Schaaper, R. M., Dunn, R. L. & Glickman, B. W. (1987) J.
`Mol. Biol. 182, 198-202.
`24. Wood, R. D., Skopek, T. R. & Hutchinson, F. (1984) J. Mol.
`Biol. 173, 273-291.
`Armstrong, J. D. & Kunz, B. A. (1990) Proc. Natl. Acad. Sci.
`USA 87, 9005-9009.
`Protic-Sabljic, M., Tuteja, N., Munson, P. J., Hauser, J.,
`Kraemer, K. H. & Dixon, K. (1986) Mol. Cell. Biol. 6, 3349-
`3356.
`McBride, T. M., Schneider, J. E., Floyd, R. A. & Loeb, L. A.
`(1992) Proc. Natl. Acad. Sci. USA 89, 6866-6871.
`Moyer, R. A., Briley, J. D., Johnsen, A. K., Stewart, U. &
`Shaw, B. R. (1992) Proc. Am. Assoc. Cancer Res. 33, 179
`(abstr.).
`Carmichael, P. L., She, M. N. & Phillips, D. H. (1992) Car-
`cinogenesis 13, 1127-1135.
`Dizdaroglu, M. & Simic, M. G. (1984) Radiat. Res. 100, 41-46.
`Ames, B. N. (1989) Mutat. Res. 214, 41-46.
`Suzuki, H., Takahashi, T., Kuroishi, T., Suyama, M., Ari-
`yoshi, Y., Takahashi, T. & Ueda, R. (1992) Cancer Res. 52,
`734-736.
`
`25.
`
`26.
`
`27.
`
`28.
`
`29.
`
`30.
`31.
`32.
`
`GDX 1015
`
`