Mutagenesis vol.14 no.1 pp.135–140, 1999
Mutations at G:C base pairs predominate after replication of
peroxyl radical-damaged pSP189 plasmids in human cells
Philip C.Burcham1 and Louise A.Harkin2
Department of Clinical and Experimental Pharmacology, The University of
Adelaide, Adelaide, SA 5005, Australia
2Present address: Department of Haematology, The Flinders University of
South Australia, Bedford Park, SA 5042, Australia
The mutagenicity of peroxyl radicals, important participants in lipid peroxidation cascades, was investigated using
a plasmid-based mutational assay system. Double-stranded
pSP189 plasmids were incubated with a range of concentrations of the water-soluble peroxyl radical generator
2,29-azobis(2-amidinopropane) hydrochloride (AAPH). Following replication in human Ad293 cells, the plasmids were
screened for supF mutations in indicator bacteria. Exposure
to peroxyl radicals caused strand nicking and a decrease
in transfection efficiency, which was accompanied by a
significant increase in supF mutants. Each of these effects
was abolished in the presence of the water-soluble vitamin
E analogue Trolox. Automated sequencing of 76 AAPHinduced mutant plasmids revealed that substitutions at
G:C base pairs were the most common changes, accounting
for 85.5% of all identified mutations. Of these, most
comprised G:C→T:A transversions (53.5%), with lesser
contributions by G:C→A:T transitions (23.9%) and
G:C→C:G transversions (22.5%). Collectively, these data
confirm our previous findings concerning the spectrum of
mutations produced upon bacterial replication of peroxyl
radical-damaged phage DNA and extend them by showing
that such damage has mutagenic consequences during
replication in more complex eukaryotic systems.
Introduction
During oxidative stress, polyunsaturated lipids in cell membranes decompose to generate a range of chemically reactive
substances, some of which possess genotoxic potential (Vaca
et al., 1988; Esterbauer, 1993; Burcham, 1998). Among these,
various α,β-unsaturated aldehydes, such as malondialdehyde
and 4-hydroxy-2-nonenal, have attracted considerable attention. These and related enals have been shown to cause
both micro- and macromutational changes in bacterial and
mammalian cells (Marnett et al., 1985; Esterbauer et al., 1990;
Esterbauer, 1993; Dittberner et al., 1995; Burcham, 1998).
Furthermore, cyclic adducts formed by such lipid-derived enals
have been detected in DNA from humans and laboratory
animals, suggesting that DNA damage by lipid peroxidationderived genotoxicants accompanies normal cellular metabolism
in vivo (Chaudhary et al., 1994; Wang and Liehr, 1994; Chung
et al., 1996; Burcham, 1998).
In addition to electrophilic aldehydes, lipid peroxidation
generates various oxygen-centred lipid radicals that are potentially genotoxic, including peroxyl radicals which catalyse the
propagation phase of lipid autoxidation cascades (Ingold, 1969;
1To
Marnett, 1987). Peroxyl radicals form upon fragmentation of
lipid hydroperoxides produced during both enzymatic and nonenzymatic lipid peroxidation (Marnett, 1987; Chamulitrat and
Mason, 1989). Since peroxyl radicals are known to form in
nuclear membranes, DNA associated with chromatin attachment sites in such structures seems a likely target for these
species (Cole et al., 1974; Greenley and Davies, 1994).
Lipid hydroperoxides exhibit genotoxicity in a number of
test systems and the fact that their DNA damaging properties
are inhibited by lipophilic peroxyl radical scavengers suggests
a direct role for peroxyl radicals in the DNA damage
(Hruszkewycz, 1988; Cao et al., 1995; Yang and Schaich,
1996). However, attributing any specific effects to peroxyl
radicals is difficult in such systems, since such a wide range
of reactive substances form upon the decomposition of lipid
hydroperoxides (Vaca et al., 1988; Esterbauer, 1993; Burcham, 1998).
To gain a fuller appreciation of the DNA damaging properties
of peroxyl radicals, in recent studies we have used the watersoluble azo initiator 2,29-azobis-(2-amidinopropane) hydrochloride (AAPH) as a source of peroxyl radicals (Harkin
and Burcham, 1997; Harkin et al., 1997). In an aqueous
environment, AAPH thermolyses at a constant rate to generate
two equivalent alkyl radicals, which upon escaping the solvent
cage react rapidly with O2 to generate peroxyl radicals (Niki,
1990). Using the M13 forward mutational assay, we found
that peroxyl radicals generated from submillimolar concentrations of AAPH are indeed genotoxic, generating premutagenic
lesions that induced base pair substitutions upon bacterial
replication (Harkin et al., 1997). The peroxyl radical-induced
mutations exhibited a high selectivity (almost 80%) for guanine
residues and almost exclusively involved G→T transversions.
Due to dissimilarities in efficiencies of repair and replication,
the fate of a given mutagenic lesion is sometimes different
between bacterial and mammalian cells (Tweats, 1995). Therefore, in the present work we have investigated the mutagenic
consequences of peroxyl radical-induced DNA damage in
human cells, using the shuttle vector pSP189 as our experimental tool. Our results indicate that DNA damage produced
by peroxyl radicals is highly mutagenic in human cells, with
a high preference for G:C base pairs.
Materials and methods
Materials
Double-stranded pSP189 plasmid DNA, amplified in the bacterial strain
AB2463, and the indicator bacterial strain Escherichia coli MBM7070 were
a generous gift from Prof. L.J.Marnett (Vanderbilt University, Nashville,
TN). Human embryonic adenovirus-transformed kidney cells (AD293) were
obtained from the American Type Culture Collection (Rockville, MD). Human
cells were cultured in Eagle’s minimum essential medium supplemented with
10% heat-inactivated horse serum and antibiotics (60 mg/l penicillin and 100
mg/l streptomycin). Bacterial cells were cultured in LB medium supplemented
with ampicillin (50 µg/ml). AAPH was obtained from Polysciences Inc.
(Warrington, PA). BRESAspin plasmid isolation kits were obtained from
whom correspondence should be addressed. Tel: 161 8 8303 5287; Fax: 161 8 8224 0685; Email: [email protected]
© UK Environmental Mutagen Society/Oxford University Press 1999
135
P.C.Burcham and L.A.Harkin
Bresatec Pty Ltd (Thebarton, South Australia), while BioRad Prep-A-Gene
DNA purification kits were purchased from BioRad Laboratories (Hercules,
CA). Dye Terminator Cycle Sequencing Ready Reaction kits were supplied
by Perkin Elmer, Applied Biosystems Division (Foster City, CA). All other
reagents were of the highest purity obtainable from standard commercial suppliers.
Peroxyl radical modification of pSP189 DNA
To prepare peroxyl radical-damaged plasmids, double-stranded pSP189 DNA
(0.1 µg/µl) was incubated for 2 h at 37°C in open reaction vessels containing
200 µM AAPH in sodium phosphate buffer (50 mM, pH 7.0). Control DNA
samples were treated similarly in the absence of AAPH. Following incubation,
the DNA was recovered using a BioRad DNA purification kit and then
quantitated via absorbance measurements at 260 nm. In some experiments,
Trolox (200 µM) was also included in the reaction mixtures.
Transfection and rescue of pSP189
The methods used for the mutagenesis experiments were similar to those
reported by Parris and Seidman (1992). Briefly, Ad293 cells were plated at a
density of 33106 cells/90 mm dish ~18 h prior to transfection. Each plate
was transfected with 10 µg modified or unmodified pSP189 DNA by the
calcium phosphate co-precipitation method (Sambrook et al., 1989). After
48 h, the cells were lysed and the plasmid DNA in the resulting supernatant
was purified by phenol/chloroform extraction before it was treated with RNase
A (50 mg/ml, 37°C, 1 h). Following a second phenol/chloroform extraction,
the pellet was digested with DpnI restriction endonuclease (20 U, 37°C, 1 h)
to remove plasmid DNA that had not been replicated by the human cells.
An additional phenol/chloroform extraction and ethanol precipitation were
performed before the DNA was finally dissolved in 40 µl H2O. The DNA
was then quantitated via absorbance measurements at 260 nm.
Bacterial transformation and mutant selection
Escherichia coli MBM7070 cells (2–33109 cells) in 10% glycerol were
transformed by electroporation with 50 ng pSP189 DNA which had been
recovered from Ad293 cells. After incubation on ice for 5–10 min the DNA/
bacteria mixture was transferred to an electroporation cuvette (0.1 cm electrode
gap) and electroporated using a Gene Pulser (BioRad) set at 1.8 kV, 25 µF,
200 Ω. Immediately following electroporation, 1 ml SOC medium was added
and the sample was transferred to another tube and incubated at 37°C for 1 h
to allow time for the cells to develop ampicillin resistance. Transformants
were selected on LB broth agar plates containing ampicillin (50 µg/ml)
overlaid with 2 mg 5-bromo-4-chloro-3-indolyl-galactosidase (X-gal) and
2 µmol isopropyl-D-thiogalactoside, which were incubated overnight at 37°C.
Colourless and pale blue colonies were selected from the background of blue
wild-type colonies and then re-streaked on X-Gal indicator plates to confirm
the mutant phenotype. Plasmid was extracted from confirmed mutants and
purified using a BRESAspin plasmid isolation kit. Transfection efficiencies
(number of colonies/ng DNA) were calculated for each treatment to obtain
estimates of the proportion of plasmids that were replication competent and
expressed as a percentage of that obtained for untreated DNA. The mutation
frequency (number of mutant colonies/wild-type) was also calculated and is
expressed per 13104 transformants.
Analysis of mutants
Purified pSP189 mutant plasmids were sequenced unidirectionally through
the supF gene with an automated ABI Prism 377 DNA sequencer using a
fluorescence dye terminator dideoxynucleotide chain termination method. A
20 base oligodeoxynucleotide (59-GGCGACACGGAAATGTTGAA-39) was
used as sequencing primer (Juedes and Wogan, 1996). Sequencing reactions
were carried out in a single reaction mixture using Dye Terminator Cycle
Sequencing Ready Reaction kits according to the manufacturer’s recommended procedure.
Agarose gel electrophoresis
pSP189 plasmids (1 µg) were loaded onto a 0.8% agarose gel containing
0.5 µg/ml ethidium bromide. The DNA was resolved for ~2 h at 65 V, after
which it was visualized under UV light. The supercoiled (form I) and relaxed
(form II) DNA was then quantitated via densiometry using a Kodak DC40
digital camera and Digital Science electrophoresis analysis software.
Results
Strand-nicking of pSP189 plasmids by AAPH
In recent work, we found that exposure of single-stranded
phage DNA to the peroxyl radical generator AAPH generated
genetic damage that was mutagenic upon replication in SOSinduced E.coli (Harkin et al., 1997). To assess the mutagenic
136
Fig. 1. Loss of covalently closed circular DNA (form I DNA) via strand
nicking during exposure of pSP189 plasmids to various concentrations of
AAPH (0–400 µM). To prepare peroxyl radical-modified DNA, pSP189
DNA (0.1 µg/µl final concentration) was incubated in 50 mM sodium
phosphate buffer (pH 7.0) for 2 h at 37°C. AAPH was added as a freshly
prepared solution to give the final concentrations shown. Following
incubation, the reaction products were resolved on a 0.8% ethidium
bromide-containing agarose gel, before the DNA was quantitated via
densitometric analysis. Each data point represents the mean 6 SE of three
determinations.
properties of such damage in human cells, the shuttle vector
pSP189 was used in this present study. In addition to containing
SV40 sequences that enable replication in mammalian cells,
pSP189 possesses a unique 8 bp signature element that
enables ready identification of sibling mutants (Parris and
Seidman, 1992).
Previous research has shown that high concentrations of
AAPH cause extensive nicking of double-stranded DNA
(Hiramoto et al., 1993). We thus sought to determine the
maximal concentration of AAPH that could be used to prepare
modified DNA for use in our mutagenesis experiments. For
this, pSP189 plasmids were incubated with a range of AAPH
concentrations (0–400 µM) for 2 h, after which they were
resolved on an ethidium bromide-containing agarose gel. The
data in Figure 1 show that AAPH caused concentrationdependent nicking of the plasmids, evidenced by a loss
of covalently closed circular DNA (form I) (Figure 1). A
corresponding increase in the amount of singly nicked plasmids
(form II) accompanied the loss of form I DNA (data not
shown). The presence of the water-soluble vitamin E analogue
Trolox completely prevented nicking of the DNA (Figure 1).
This concentration of Trolox was chosen on the basis of its
ability to completely protect single-stranded M13 DNA against
AAPH (Harkin et al., 1997). The ability of Trolox to suppress
the genotoxicity of AAPH implicates peroxyl radicals, rather
than carbon-centred alkyl radicals, in the induction of genetic
damage under our conditions.
Since the intent was to achieve a significant level of
DNA damage while avoiding extensive plasmid degradation,
a 200 µM concentration of AAPH was used in subsequent
mutagenesis experiments. Although this concentration of
AAPH caused an ~60% loss of form I DNA, the amount of
remaining closed circular DNA was judged sufficient to permit
high levels of plasmid replication in the cells.
AAPH-induced mutagenesis in double-stranded pSP189
plasmids
For the mutagenesis experiments, AAPH-modified pSP189
plasmids were introduced into human A293 cells via the
calcium phosphate method and then 48 h was allowed for
Radical-damaged pSP189 plasmids in human cells
Table I. Effect of exposure to 200 µM AAPH in the presence and absence
of 200 µM Trolox on the frequency of supF mutations in pSP189 plasmids
following replication in human Ad293 cells
Treatment
Colonies screened
No. mutants
MFa
Control
AAPH
AAPH 1 Trolox
88634
259065
85192
4
146
3
0.451
5.64
0.352
aMutation
frequency (no. mutants/104 transformants). Modified plasmids
were prepared using 200 µM AAPH, which was also the concentration of
Trolox used. More details concerning these experiments are contained in
Materials and methods.
DNA replication and mutation fixation. The plasmid DNA
was recovered and then electroporated into MBM7070 indicator
cells for the detection of supF mutations. By expressing the
number of plaques formed relative to the amount of DNA
electroporated, transfection efficiencies were obtained for each
treatment. Treatment with 200 µM AAPH decreased transfection efficiencies by ~66% compared with unmodified plasmids
(data not shown). Furthermore, the presence of 200 µM Trolox
during the reactions with AAPH prevented the loss of plasmid
capable of replication so that transfection efficiencies were
comparable with those of unmodified DNA templates (data
not shown).
In addition to the loss of replication competent DNA induced
by peroxyl radicals, an increase in supF mutants in the
surviving plasmid population was seen (Table I). The phenotype
of all suspected mutants was confirmed by restreaking infected
colonies on X-gal indicator plates. Any ‘mutants’ that failed
this checking procedure were discarded. A .12-fold increase
in the supF mutation frequency was seen for plasmids that
had been exposed to 200 µM AAPH compared with the
spontaneous mutation rate in unmodified DNA (Table I).
Consistent with its ability to prevent strand nicking and the
decline in transfection efficiency, the presence of 200 µM
Trolox completely suppressed the increase in supF mutation
frequency produced by AAPH (Table I).
Distribution of AAPH induced supF mutations
To determine the type of mutations produced during replication
of AAPH-damaged plasmids, the supF gene of 76 AAPHinduced mutants was analysed via automated DNA sequencing.
Due to the low number of spontaneous mutants detected
(4/88634), none of these were sequenced. When the signature
elements for all plasmids with mutations at the same site were
compared, only two pairs of siblings were found. One of each
was omitted from the results.
In a previous study of the bacterial mutagenicity of peroxyl
radical-induced DNA damage, over 75% of AAPH-induced
mutations in the lacZα gene of M13mp19 occurred at G:C
base pairs (Harkin et al., 1997). The results obtained in the
present study coincide closely with the earlier work and suggest
that the mutagenic properties of peroxyl radical-induced DNA
damage is similar in human cells to in bacteria. The types of
mutations observed are summarized in Table II, while Figures
2 and 3 show the location of the mutations within the suppressor
tRNA sequence of the supF gene. The majority of the AAPHinduced mutants were base substitutions (73/83) that occurred
predominantly at G:C base pairs (Table II). The most common
were G:C→T:A transversions (38/73), followed by G:C→A:T
transitions and G:C→C:G transversions, which made up 23.3
and 21.9% of all base substitutions, respectively. Substitutions
Table II. Summary of sequence change in the tRNA supF gene of AAPHdamaged pSP189 plasmids following replication in human Ad293 cells
Alteration
No.
Base pair substitutions
G:C→T:A
G:C→A:T
G:C→C:G
A:T→G:C
A:T→T:A
A:T→T:A
Insertions
Deletions
Double mutations
Triple mutations
Tandem mutations
38
17
16
1
1
0
2
8
8
2
1
at A:T base pairs were only very minor contributors to the
mutational spectrum (Table II). AAPH-induced base substitutions occurred at many sites within the target gene, although
47% were located at one of six hotspots, all of which involved
G:C base pairs (positions 104, 126, 129, 150, 156 and 164;
Figure 2).
In addition to base pair substitutions, a number of frameshift
mutations occurred (Figure 3). Two insertion mutations were
detected, one of which involved addition of a C:G base pair
within a run of five consecutive C:G pairs while the other
involved duplication of a 6 bp sequence (GGTTCC) (Figure 3).
Eight deletion mutants were found, ranging in size from 16 to
127 bp (Figure 3).
Of all the supF mutants sequenced, eight contained two
mutations within the target sequence, while two contained
three mutations. Multiple base substitutions are relatively
common among inactivating mutations in the supF gene
(Kraemer and Seidman, 1989). Only a single tandem mutation
was found (GG→AA, position 104/5), while one mutant
contained a 24 bp insertion which was located downstream of
the supF gene (positions 239–245). For two of the sequenced
mutants, the sequence change could not be specified as the
break points could not be defined.
Discussion
The present findings extend recent work in our laboratory that
has investigated the mutagenic consequences of interactions
between DNA and peroxyl radicals (Harkin and Burcham,
1997; Harkin et al., 1997). While previous studies have
indicated a role for free radicals in induction of the DNA
damage associated with lipid peroxidation, the use of the azo
initiator AAPH as a chemically defined source of peroxyl
radicals removes some of the ambiguity associated with earlier
work. Since studies of reaction rates between peroxyl radicals
and various organic substrates have shown that altering the
substituent attached to the dioxygen group has only minor
consequences for their reactivity, alkylperoxyl radicals derived
from AAPH are a useful surrogate for their lipid counterparts
(Neta and Hule, 1990; Niki, 1990). Consequently, peroxyl
radicals generated upon thermolysis of AAPH have proved
useful in the study of reactions with a range of cell constituents,
including lipid, DNA and protein (Hiramoto et al., 1993; Viner
et al., 1997; Garner et al., 1998).
As with the earlier findings in the M13 forward mutational
assay, peroxyl radical-induced mutational events in the structural region of the supF gene were heavily represented at G:C
137
P.C.Burcham and L.A.Harkin
Fig. 2. Distribution of AAPH-induced base pair substitutions within the suppressor tRNA sequence (bases 99–183) of the supF gene in pSP189 plasmids
following replication in human Ad293 cells. Tandem mutations are underlined.
Fig. 3. Distribution of AAPH-induced deletions and insertions within the supF tRNA gene (bases 99–183) in pSP189 plasmids following replication in human
Ad293 cells. Deletion break points were as shown while arrows indicate that the break point lies outside the target sequence. Insertions are shown above the
sequence.
base pairs (Table II). Indeed, out of the 73 base substitutions
detected in the present study, all but two involved G:C sites.
Due to the use of double-stranded DNA in this study, it is
possible that base damage in either of the strands could have
caused the sequence changes (i.e. damage at G or C residues).
This problem was avoided in our earlier work through the use
of single-stranded M13 DNA as the template for replication
(Harkin et al., 1997). In that study, a high proportion of base
substitutions in lacZα mutants occurred at G residues (79/92)
(Harkin et al., 1997). In contrast, transversions and transitions
involving C residues comprised just 11% of all base substitutions in the single-stranded AAPH-induced mutants, although
they were common among spontaneous mutants (Harkin et al.,
1997). On the basis of the selectivity for G residues seen in
that study, it is concluded that modifications at C residues are
probably only minor contributors to the substitutions at G:C
base pairs seen in the present study.
At least two modifications to deoxyguanosine residues are
known to occur upon exposure of DNA to lipid hydroperoxides,
namely oxygenation at the C-8 of guanine and depurination
to form an abasic site (Inouye, 1984; Ueda et al., 1985;
Akasaka and Yamamoto, 1994; de Kok et al., 1994). In
138
earlier work, we were unable to detect formation of 8-oxo-29deoxyguanosine in AAPH-treated 29-deoxyguanosine solutions, a finding that was recently confirmed by Valentine et al.
(1998). Furthermore, Valentine and associates were unable to
detect digestion of peroxyl radical-damaged DNA with Fpg
protein, a DNA glycosylase which recognizes a number of
oxidized purines, including 8-oxo-29-deoxyguanosine (Krokan
et al., 1997; Valentine et al., 1998).
In contrast, we found that the mutagenicity of AAPHinduced DNA damage was susceptible to alkaline treatment,
which suggested the presence of abasic sites (Harkin et al.,
1997). Our subsequent work using microbial AP endonucleases
with distinct substrate specificities suggested that the major
lesion in peroxyl radical-damaged DNA was an unconventional
abasic site in which the deoxyribose exists as a C1’-oxidized
moiety (Harkin and Burcham, 1997). Interestingly, although
they reported a bacterial mutational spectrum that resembled
our own in that transversions at G residues were the most
common sequence changes, Valentine et al. (1998) were unable
to detect abasic sites in peroxyl radical-modified 32P-endlabelled DNA templates, despite using alkali and two AP
endonucleases to reveal the lesions. The reasons for the
Radical-damaged pSP189 plasmids in human cells
difference in findings between the two laboratories are puzzling,
especially with regard to the lack of effect of alkali treatment.
However, the inability of Valentine et al. to demonstrate
enzymatic digestion of peroxyl radical-damaged DNA may
have been due to their use of Class I or 39-acting AP
endonucleases only, namely Nth and Fpg (Valentine et al.,
1998). Work by others indicates that 39-AP endonucleases
show little or no catalytic activity towards DNA that contains
C1’-oxidized sugars (Haring et al., 1994). Rather, these lesions
are recognized by Class II or 59-acting AP endonucleases such
as E.coli exonuclease III, an enzyme that we have shown can
digest peroxyl radical-treated plasmids (Harkin and Burcham,
1997). We plan to conduct new experiments to determine
whether methodological factors account for the differences
between our findings and those of Valentine et al. (1998).
If C1’-oxidized sugar residues are confirmed to be key
products of reactions between DNA and peroxyl radicals, then
recent work on the genotoxicity of various oxometal species
such as certain carcinogenic chromium(V) complexes may be
pertinent to our findings (Goyne and Sigman, 1987; Sugden
and Wetterhahn, 1997). Damage by the latter species was
recently suggested to occur via hydrogen abstraction from the
C1’ position of the sugar, followed by a hydration reaction
and two β-elimination reactions that result in base loss and
strand scission (Bose et al., 1998). Although all four bases
(A, G, T and C) were released during oxidation of calf thymus
DNA by Cr(V), the concentration of G released exceeded that
of the other bases by some 2.5- and 3.3-fold (Bose et al.,
1998). A similar selectivity for G residues has been seen with
osmium and ruthenium oxo complexes that act as ‘chemical
nucleases’ via an initial attack on the C1’ of deoxyribose
(Cheng et al., 1995). Such a heightened susceptibility of the
sugar attached to G residues to C1’ oxidation may also account
for the mutagenic selectivity for G:C sites seen in the present
study. Furthermore, a mechanism proceeding via C1’ oxidation
of deoxyribose followed by β-elimination reactions and strand
scission may also explain the DNA nicking caused by AAPH
(Figure 1; 25). A mechanism involving hydrogen abstraction
from the C1’ position of the sugar is also consistent with the
known ability of peroxyl radicals to catalyse H abstraction
(Ingold, 1969; Marnett, 1987). An effort is currently under
way to gain more definitive evidence for an attack at the C1’
of the sugar moiety of deoxyguanosine residues by peroxyl
radicals generated by AAPH and also lipid hydroperoxides.
If the formation of oxidized abasic sites by the mechanism
outlined above represents the major route to the formation of
premutagenic DNA damage by peroxyl radicals, the question
arises as to whether the observed mutations are consistent with
the known miscoding properties of such lesions? In general,
our finding of a preponderance of G:C→T:A transversions,
accompanied by lesser proportions of G:C→AT transitions and
G:C→C:G transversions, are consistent with the expectation
that the oxidized abasic sites would be processed as classic
non-instructional lesions according to the so-called ‘A-rule’
(Kunkel, 1984; Lawrence et al., 1990; Takeshita and Eisenberg,
1994). As has been pointed out, however, this rule only holds
true for certain non-coding lesions, while the preference for
A insertion is strongly influenced by the local sequence context
(Loechler, 1996). Indeed, such reservations are consistent with
the findings of Neto and co-workers, who have shown that
adenine insertion is only a minor contributor to base misinsertion when abasic lesions are replicated in mammalian cells
(Gentil et al., 1992; Neto et al., 1992; Cabral-Neto et al.,
1994). One reason for the apparent obedience to the so-called
‘A-rule’ in our present work could be that oxidized abasic
sites are processed and/or repaired differently to conventional
abasic sites, thus explaining the difference between our findings
and those of Neto et al. However, a problem with this
conclusion arises from the limitations inherent in ‘random
modification’ mutagenesis experiments of the type used in
this study, namely that the possibility that multiple lesions
contributed to mutagenesis cannot be ruled out. Ultimately,
clarification of the mutagenic properties of C1’-oxidized abasic
sites and of the influence of neighbouring bases on their
mutagenicity/repair will not be possible until methods for the
incorporation of these lesions into a defined site in a viral
genome or oligodeoxynucleotide are developed.
The question as to whether oxidized abasic sites generated
by peroxyl radicals undergo repair in mammalian cells is
presently unanswerable. As mentioned above, our recent work
indicates that these lesions are substrates for the AP endonuclease activity associated with exonuclease III from E.coli,
but not for that associated with T4 endonuclease V (Harkin
and Burcham, 1997). Whether C1’-oxidized abasic sites are
cleaved by the major AP endonuclease in human cells, variously
named HAP1, Ape, APEX and Ref-1, is a question that awaits
resolution (Singer and Hang, 1997). However, since the human
enzyme is a Class II AP endonuclease with the same catalytic
action as exonuclease III (i.e. it cleaves 59 of an abasic site),
it is reasonable to expect that C1’-oxidized abasic sites will
be recognized by the former enzyme. Nevertheless, since C1’oxidized abasic sites are likely to have a finite lifetime in
chemical terms due to the readiness with which they undergo
β-elimination to form strand breaks, it is possible that enzymatic
repair is not a major determinant of their genotoxic potential
in vivo.
In addition to causing point mutations, various frameshift
mutations were detected upon sequencing of peroxyl radicalinduced supF mutants (Figure 3). The eight sizeable deletion
mutations are particularly interesting, since they occurred at a
much higher frequency than in our earlier work in the bacterial
system (Harkin et al., 1997). Two of the deletions occurred at
direct repeats (the 16 and 39 bp deletions shown in Figure 3),
which suggests they may be the result of adduct-induced
template misalignment and slippage during plasmid replication
(Kunkel, 1990). The remaining large deletions presumably
arose as a consequence of random single-strand breaks induced
by the peroxyl radicals.
In conclusion, the present study has extended our prior work
in bacteria, by showing that DNA damage caused by peroxyl
radicals generated by the azo initiator AAPH is mutagenic in
human cells. The DNA damage was strongly suppressed in
the presence of Trolox, indicating the involvement of oxygencentred rather than carbon-centred radicals in the induction of
DNA damage. Mutational events, or at least the predominant
base pair substitutions, occurred mainly at G:C sites. Such
selectivity is consistent with work from other laboratories
concerning the selectivity of DNA oxidants whose initial site
of attack is the C1’ of deoxyribose. Future work in our
laboratory will determine the extent to which such DNA
damage is associated with the induction of lipid peroxidation
in intact biological systems.
Acknowledgements
The authors are grateful to Prof. L.J.Marnett and L.J.Niedernhofer for the
provision of bacterial strains and practical advice concerning the mutagenesis
139
P.C.Burcham and L.A.Harkin
experiments. The work was supported by a Priming Grant from the National
Health and Medical Research Council of Australia.
References
Akasaka,S. and Yamamoto,K. (1994) Mutagenesis resulting from DNA damage
by lipid peroxidation in the supF gene of Escherichia coli. Mutat. Res.,
315, 105–112.
Bose,K.N., Fonkeng,B.S., Moghaddas,S. and Stroup,D. (1998) Mechanisms
of DNA damage by chromium(V) carcinogens. Nucleic Acids Res., 26,
1588–1596.
Burcham,P.C. (1998) Genotoxic lipid peroxidation products: their DNA
damaging properties and role in formation of endogenous DNA adducts.
Mutagenesis, 13, 287–305.
Cabral-Neto,J.B., Cabral,R.E., Margot,A., Le Page,F., Sarasin,A. and Gentil,A.
(1994) Coding properties of a unique apurinic/apyrimidinic site replicated
in mammalian cells. J. Mol. Biol., 240, 416–420.
Cao,E.H., Liu,X.Q., Wang,L.G., Wang,J.J. and Xu,N.F. (1995) Evidence that
lipid peroxidation products bind to DNA in liver cells. Biochim. Biophys.
Acta, 1259, 187–191.
Chamulitrat,W. and Mason,R.P. (1989) Lipid peroxyl radical intermediates in
the peroxidation of polyunsaturated fatty acids by lipoxygenase. Direct
electron spin resonance investigations. J. Biol. Chem., 264, 20968–20973.
Chaudhary,A.K., Nokubo,M., Reddy,G.R., Yeola,A.N., Morrow,J.D., Blair,I.A.
and Marnett,L.J. (1994) Detection of endogenous malondialdehyde–
deoxyguanosine adducts in human liver. Science, 265, 1580–1582.
Cheng,C.-C., Goll,J.G., Neyhart,G.A., Welch,T.W., Singh,P. and Thorp,H.H.
(1995) Relative rates and potentials of competing redox processes during
DNA cleavage: oxidation mechanisms and sequence-specific catalysis of
the self-inactivation of oxo-metal oxidants by DNA. J. Am. Chem. Soc.,
117, 2970–2980.
Chung,F.-L., Chen,H.J.,C. and Nath,R.G. (1996) Lipid peroxidation as a
potential endogenous source for the formation of exocyclic DNA adducts.
Carcinogenesis, 17, 2105–2111.
Cole,A. Cooper,W.G., Shonka,F., Cory,P.M., Humphrey,R.M. and Ansevin,A.T.
(1974) DNA scission in hamster cells and isolated nuclei studies by low
voltage electron beam irradiation. Radiat. Res., 60, 1–33.
de Kok,T.M., ten Vaarwerk,F., Zwingman,I., van Maanen,J.M. and
Kleinjans,J.C. (1994) Peroxidation of linoleic, arachidonic and oleic acid
in relation to the induction of oxidative DNA damage and cytogenetic
effects. Carcinogenesis, 15, 1399–1404.
Dittberner,U., Eisenbrand,G. and Zankl,H. (1995) Genotoxic effects of the
alpha,beta-unsaturated aldehydes 2-trans-butenal, 2-trans-hexenal and
2-trans, 6-cis-nonadienal. Mutat. Res., 335, 259–265.
Esterbauer,H. (1993) Cytotoxicity and genotoxicity of lipid peroxidation
products. Am. J. Clin. Nutr., 57 (suppl.), 779S–786S.
Esterbauer,H., Eckl,P. and Ortner,A. (1990) Possible mutagens derived from
lipids and lipid precursors. Mutat. Res., 238, 223–233.
Garner,B., Witting,P.K., Waldeck,A.R., Christison,J.K., Raftery,M. and
Stocker,R. (1998) Oxidation of high density lipoproteins. I. Formation of
methionine sulfoxide in apolipoproteins AI and AII is an early event that
accompanies lipid peroxidation and can be enhanced by alpha-tocopherol.
J. Biol. Chem., 273, 6080–6087.
Gentil,A., Cabral-Neto,J.B., Mariage-Samson,R., Margot,A., Imbach,J.L.,
Rayner,B. and Sarasin,A. (1992) Mutagenicity of a unique apurinic/
apyrimidinic site in mammalian cells. J. Mol. Biol., 227, 981–984.
Goyne,T.E. and Sigman,D.S. (1987) Nuclease activity of 1,10-phenathrolinecopper ion. Chemistry of deoxyribose oxidation. J. Am. Chem. Soc., 109,
2846–2848.
Greenley,T.L. and Davies,M.J. (1994) Direct detection of radical generation
in rat liver nuclei on treatment with tumour-promoting hydroperoxides and
related compounds. Biochim. Biophys. Acta, 1226, 56–64.
Haring,M., Rudiger,H., Demple,B., Boiteux,S. and Epe,B. (1994) Recognition
of oxidized abasic sites by repair endonucleases. Nucleic Acids Res., 22,
2010–2015.
Harkin,L.A. and Burcham,P.C. (1997) Formation of novel C1-oxidised abasic
sites in alkylperoxyl radical-damaged plasmid DNA. Biochem. Biophys.
Res. Commun., 237, 1–5.
Harkin,L.A., Butler,L.M. and Burcham,P.C. (1997) Role of G→T transversions
in the mutagenicity of alkylperoxyl radicals: induction of alkali-labile sites
in bacteriophage M13mp19. Chem. Res. Toxicol., 10, 575–581.
Hiramoto,K., Johkoh,H., Sako,K. and Kikugawa,K. (1993) DNA breaking
activity of the carbon-centered radical generated from 2,29-azobis
(2-amidinopropane) hydrochloride (AAPH). Free Radic. Res. Commun.,
19, 323–332.
Hruszkewycz,A.M. (1988) Evidence for mitochondrial DNA damage by lipid
peroxidation. Biochem. Biophys. Res. Commun., 153, 191–197.
140
Ingold,K.U. (1969) Peroxyl radicals. Acc. Chem. Res., 2, 1–9.
Inouye,S. (1984) Site-specific cleavage of double-strand DNA by
hydroperoxide of linoleic acid. FEBS Lett., 172, 231–234.
Juedes,M.J. and Wogan,G.N. (1996) Peroxynitrite-induced mutation spectra
of pSP189 following replication in bacteria and in human cells. Mutat.
Res., 349, 51–61.
Kraemer,K.H. and Seidman,M.M. (1989) Use of supF, an Escherichia coli
tyrosine suppressor tRNA gene, as a mutagenic target in shuttle-vector
plasmids. Mutat. Res., 220, 61–72.
Krokan,H.,E., Standal,R. and Slupphaug,G. (1997) DNA glycosylases in the
base excision repair of DNA. Biochem. J., 325, 1–16.
Kunkel,T.A. (1984) Mutational specificity of depurination. Proc. Natl Acad.
Sci. USA, 81, 1494–1498.
Kunkel,T.A. (1990) Misalignment-mediated DNA synthesis errors.
Biochemistry, 29, 8003–8011.
Lawrence,C.W., Borden,A., Banerjee,S.K. and LeClerc,J.E. (1990) Mutation
frequency and spectrum resulting from a single abasic site in a singlestranded vector. Nucleic Acids Res., 18, 2153–2157.
Loechler,E.L. (1996) The role of adduct site-specific mutagenesis in
understanding how carcinogen–DNA adducts cause mutations: perspective,
prospects and problems. Carcinogenesis, 17, 895–902.
Marnett,L.J. (1987) Peroxyl free radicals: potential mediators of tumour
initiation and promotion. Carcinogenesis, 8, 1365–1372.
Marnett,L.J., Hurd,H.K., Hollstein,M.C., Levin,D.E., Esterbauer,H. and
Ames,B.N. (1985) Naturally occurring carbonyl compounds are mutagens
in Salmonella tester strain TA104. Mutat. Res., 148, 25–34.
Neta,P. and Hule,R.E. (1990) Rate constants for reactions of peroxyl radicals
in fluid solutions. J. Phys. Chem. Ref. Data, 19, 413–498.
Neto,J.B., Gentil,A., Cabral,R.E. and Sarasin,A. (1992) Mutation spectrum of
heat-induced abasic sites on a single-stranded shuttle vector replicated in
mammalian cells. J. Biol. Chem., 25, 19718–19723.
Niki,E. (1990) Free radical initiators as source of water- or lipid-soluble
peroxyl radicals. Methods Enzymol., 186, 100–108.
Parris,C.N. and Seidman,M.M. (1992) A signature element distinguishes
sibling and independent mutations in a shuttle vector plasmid. Gene, 117,
1–5.
Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A
Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
Singer,B. and Hang,B. (1997) What structural features determine repair
enzyme specificity and mechanism in chemically modified DNA? Chem.
Res. Toxicol., 10, 713–732.
Sugden,K.D. and Wetterhahn,K.E. (1997) Direct and hydrogen peroxideinduced chromium(V) oxidation of deoxyribose in single-stranded and
double-stranded calf thymus DNA. Chem. Res. Toxicol., 10, 1397–1406.
Takeshita,M. and Eisenberg,W. (1994) Mechanism of mutation on DNA
templates containing synthetic abasic sites: study with a double strand
vector. Nucleic Acids Res., 22, 1897–1902.
Tweats,D.J. (1995) Mutagenicity. In Ballantyne,B., Marrs,T. and Turner,P.
(eds), General and Applied Toxicology, Abridged Edition. Macmillan,
Basingstoke, UK, pp. 807–872.
Ueda,K., Kobayashi,S., Morita,J. and Komano,T. (1985) Site-specific DNA
damage caused by lipid peroxidation products. Biochim. Biophys. Acta,
824, 341–348.
Vaca,C.E., Wilhelm,J. and Harms-Ringdahl,M. (1988) Interaction of lipid
peroxidation products with DNA. A Review. Mutat. Res., 195, 137–149.
Valentine,M.,R., Rodriguez,H. and Termini,J. (1998) Mutagenesis by peroxy
radical is dominated by transversions at deoxyguanosine. Evidence for the
lack of involvement of 8-oxo-dG and/or abasic site formation. Biochemistry,
37, 7030–7038.
Viner,R.I., Ferrington,D.A., Aced,G.I., Miller-Schlyer,M., Bigelow,D.J. and
Schoneich,C. (1997) In vivo aging of rat skeletal muscle sarcoplasmic
reticulum Ca-ATPase. Chemical analysis and quantitative simulation by
exposure to low levels of peroxyl radicals. Biochim. Biophys. Acta, 1329,
321–335.
Wang,M.Y. and Liehr,J.G. (1994) Lipid hydroperoxide-induced endogenous
DNA adducts in hamsters: possible mechanism of lipid hydroperoxidemediated carcinogenesis. Arch. Biochem. Biophys, 316, 38–46.
Yang,M.H. and Schaich,K.M. (1996) Factors affecting DNA damage caused
by lipid hydroperoxides and aldehydes. Free Radic. Biol. Med., 20, 225–236.
Received on July 17, 1998; accepted on August 28, 1998