Mutagenesis vol.13 no.4 pp.367-373, 1998
DNA sequence analysis of spontaneous lad"* mutations in
06-alkylguanine-DNA alkyltransferase-proficient and -deficient
Escherichia coli
Antonio Vidal, Nieves Abril and Carmen Pueyo1
Departamento de Bioqufmica y Biologfa Molecular, Universidad de
Cordoba, 14071-Cordoba, Spain
Spontaneous mutagenesis in C^-aJkylguanine-DNA alkyltransferase-proficient and -deficient {ada ogt mutants) Escherichia coli was studied in two ways: in bacteria growing
in nonselective liquid medium and in bacteria resting
on selective agar plates. ATase mutants showed similar
spontaneous mutation rates as ATase proficient bacteria
during growth phase,- an excess of mutants arising in
nondividing cells. The resting-associated mutagenesis in
ada+ ogt+ uvr bacteria was biphasic,* the high sensitive
range being triggered beyond the first 6 days after plating.
Contrarily, spontaneous Lac*1 mutants from ada' ogf uvr~
cells steadily increased over the 8 day period of plate
incubation. These results suggested that, in the absence of
nucleotide excision repair, the repair by both the Ada and
the Ogt ATases is not saturated until the cells have been
resting for 6 days. The spontaneous lacV* mutation spectrum of ada+ ogt+ «vr" bacteria growing in non-selective
liquid medium served as a baseline to determine the
mutation events increased in the ATase-deficient derivative
upon prolonged incubation on selective plates. The percentage of G:C—»A:T transitions, presumably driven by unrepaired 0*-alkylguanine lesions, was increased at the expense
of other mutation types. G:C—>A:T transitions accumulated
with a pronounced 5-PuG bias, suggesting that the endogenous metabolite(s) responsible for this mutation class is
an S,vl type alkylating compound(s). Accordingly, the site
distribution of G:C-»A:T transitions in nondividing ATase
defective bacteria showed similarities with the spectra
induced by alkylnitrosoureas, particularly with those generating bulky alkylated DNA adducts.
Introduction
The mechanisms by which spontaneous mutations occur are
of considerable interest, since it has been hypothesized that
one of the early steps in carcinogenesis involves an elevation
in the rate of spontaneous mutation (Loeb, 1991; Lees and
Harlow, 1995). Numerous studies in Escherichia coli have
established that low spontaneous mutation rates are achieved
by the high accuracy of DNA replication and by prevention
and removal of DNA damage (Friedberg et al, 1995). Nonenzymatic DNA methylation caused by endogenous reactive
molecules is a potential source of spontaneous mutation
(Lindahl, 1993), repair of DNA alkylation damage contributing
to the maintenance of low spontaneous mutation rates in both
prokaryotes (Rebeck and Samson, 1991; Yamada et al, 1995)
and eukaryotes (Xiao and Samson, 1993).
Alkylating agents transfer alkyl groups to the nucleophilic
nitrogens and oxygens in DNA, generating a variety of lesions
(Singer and Grunberger, 1983). Of these, the most relevant to
mutagenesis are C^-alkylguanine and (T'-alkylthymine, which
can mispair during DNA replication, giving rise to G:C—»A:T
and A:T->G:C transitions respectively (Horsfall et al., 1990).
Mutagenesis by C^-alkylguanine is prevented by ubiquitous
(^-alkylguanine-DNA alkyltransferases (ATases) (Friedberg
et al, 1995). These DNA repair proteins directly dealkylate
C^-alkylguanine in cellular DNA and irreversibly transfer the
alkyl group to a cysteine residue at the active site of the
protein and so the reaction is stoichiometric and autoinactivating, further repair requiring de novo synthesis of
more ATase (Pegg, 1990; Samson, 1992).
Escherichia coli has two ATases, a constitutive ogt gene
product and an inducible ada gene product (Potter et al, 1987;
Teo et al., 1984). Both activities repair C^-alkylguanine and
(7*-alkylthymine in DNA, but differ in their efficiencies in
reverting 04-alkylthymine and long chain alkylated DNA bases
(Wilkinson et al., 1989). Wild-type non-adapted E.coli contain
only two to four Ada molecules per cell, but 10-fold more
Ogt protein (Potter et al, 1989; Rebeck et al, 1989; Vaughan
et al, 1991). The prime defense against endogenous mutagenic
DNA alkylation is, therefore, the Ogt protein. Non-toxic levels
of some methylating agents fully induce ada expression by as
much as 1000-fold, conferring to the cells increased protection
against mutation induction (Robins and Cairns, 1979). The
level of Ada protein is ~20-fold higher in stationary phase and
starved cells than in rapidly growing cells (Tavema and
Sedgwick, 1996).
Escherichia coli ada ogt null mutants, which are totally
deficient in ATase activity, display an increased spontaneous
mutation rate, a phenotype that appears to occur predominantly
in non-dividing cells (Rebeck and Samson, 1991). This
suggests generation of endogenous DNA alkylating agents
under non-growth conditions. Putative sources of such agents
are nitrosatable metabolites, such as methylamine and methylurea (Taverna and Sedgwick, 1996), but most probably more
abundant cellular amines, like amino acids, peptides and
polyamines (Sedgwick, 1997). Mackay et al. (1994) have
investigated which base substitution mutations are elevated in
ada ogt null mutants by monitoring the reversion of six
mutated lacZ alleles that revert via each of the six possible
base pair substitutions. Five of the six base substitutions
were elevated in resting ATase-deficient cells versus the
corresponding wild-type version.
It has been frequently argued that no set of strains
containing a small number of revertible point mutations can
encompass the entire set of possible mutagenic events in DNA.
Forward mutation systems have been offered as the genetic
solution to the problem of mutagen-mutation specificity (RuizRubio et al, 1984). Unlike previous studies based on reversion of specific lacZ mutant alleles, here we determined the
spontaneous mutation spectrum of ATase-deficient {ada ogt
double mutants) bacteria resting on selective agar plates as
'To whom correspondence should be addressed. Tel: +34 57 218695; Fax: +34 57 218688; Email: [email protected]
© UK Environmental Mutagen Society/Oxford University Press 1998
367
A.Vidal, NAbril and C.Pueyo
Table I. Escherichia coli K-12 bactenal strains
Strain
UC831O
UC1178
UC1011
UC1175
UCI263
UC932
UC1264
UC1I53
Chromosomal genotype
araD81 nadlli arg56 A(uvrB-bio) Aipro-lac)
as UC8310 but ada.JnlO ogt-l.:kan*
as UC8310 but ada:TnW A(ogt-fnr)]
as UCII78 but uvr+
as UC1011 but uvr+
araD8l nadl!3 arg56 A(uvrB-bio) A(pro-lac)
as UC932 but ada.JnlO ogt-\::kan*
as UC932 but ado :Tn 10 A(ogt-fnr)l
Episomal genotype*
F
F
F
F
F'
F'
F:
F
+
lacP ZL8 proB
lacP ZL8 proB+
lacP ZL8 proB+
lad'' ZL8 proB +
lacH ZL8 proB +
lacrZT proB+
lactTT prvB+
lacrr proB+
Reference or source
Vidal etal., 1995
This work
Jurado et al, 1995
This work
This work
This work
This work
This work
"The F' lacP ZL8 proB+ episome allows selection of lad plus lacCf forward mutations (Lacc phenotype) The lacZ mutant allele on F' lactZr proB+
episome reverts to the wild-type via G:C—>A:T transitions
compared with corresponding ATase-proficient cells growing
in non-selective liquid medium. To this end we have sequenced
a total of 124 independent forward mutations affecting the
first 180 bp of the lacl gene. This region is the target for
dominant lacl mutations (termed lacl~~^) and is suitable for
studies of mutational specificity since it contains a relatively
high density of detectable mutable sites. The comparison has
been made in a uvrB mutant background, to avoid interference
from the UvrABC nucleotide excision repair system.
Materials and methods
Bacterial strains
The strains used in this study are listed in Table I. All the strains have
a chromosomal deletion of the lactose operon and all of them carry one
of two F' lac episomes. Bacteria carrying the F' lacP ZLS proB+ episome
allow selection of mutants that synthesize ji-galactosidase constitutively.
Lacc mutants are selected on minimal agar plates with 750 mg/1 phenyl(J-D-galactoside (Pgal, Sigma) as the sole carbon source. Bacteria with the
F' lacl~7r proB+ episome allow selection of LacZ + revertants on lactose
minimal agar plates. These revertants arise specifically by G C—>A T
transitions (Cupples and Miller, 1989). The deletion A(ogt-fnr)I covers the
ogt gene and the nearby fnr locus (Abnl tt al., 1994). PI transduction
was used to transfer the wild-type uvrB gene to produce uvr+ derivatives
of UC1178 and UCI0I1. The F' lacTZT proB+ episome was transferred
by conjugation from strain CC102 (Cupples and Miller, 1989) The resulting
exconjugants were selected for proline prototrophy CA7O92.1 [HrfH ara~
ValR Ailac-pro)], used as scavenger cells, was obtained from Dr P.L Foster
(Boston University) CA7092.1 bacteria have a chromosomal deletion of
the lactose operon, therefore, they can neither mutate nor rccombine with
F' lac to produce Lac constitutive mutants or LacZ+ revertants. Media
were as described by Miller (1972) and Davis tt al (1980)
Spontaneous mutagenesis
Spontaneous mutagenesis was studied in bacteria growing in non-selective
liquid medium and in bacteria resting on selective agar plates. In both cases
a single colony was inoculated into 10 ml Luna-Bertani (LB) nutrient medium
and the bacteria were grown to stationary phase (37°C, 90 r.p.m. for 12 h).
The culture was concentrated by centnfugation. washed twice and resuspended
in Vogel-Bonner salts (VB).
Spontaneous mutagenesis in growing bacteria Aliquot suspensions, containing
~IO7 cells (Lacc forward mutation assay) or 109 cells (LacZ+ reversion assay),
from at least 20 independent cultures were combined in 2 ml molten top agar
and then poured onto Pgal (Lacc mutants) or lactose (LacZ+ revertants)
minimal selective plates. The plates were then incubated upside down at 37°C
for 3 (Lacc mutants) or 2 days (LacZ+ revertants). To determine viability at
the time of seeding the overnight cultures were diluted and plated on LB agar
plates. Viable cells were counted 1 day after incubation at 37°C.
Spontaneous mutagenesis in resting bacteria. Approximately 103 cells (Lacc
forward mutation assay) or 109 cells (LacZ + reversion assay) respectively
were plated together with 5 x 108 CA7092.1 scavenger cells on selective plates
for Lacc mutants or LacZ + revertants. Addition of the scavenger cells prevents
bacterial tester strain multiplication on the minimal selective plates The
growth of scavenger cells is prevented in the absence of proline Plates were
sealed individually with Parafilm and incubated at 37°C. With continued
incubation of the plates the Lacc mutants and LacZ + revertants accumulated
and 10 plates were counted daily for 8 days To determine viability of the
368
cells during these prolonged incubation periods -10 3 cells were plated together
with 5X1O8 CA7092.1 scavenger cells on selective plates for both Lacc
mutants and LacZ+ revertants. Every day two of these selective plates were
gently covered with 2 ml top agar containing 1 % glycerol Viable cells were
automatically counted 2 days later using an image analyzer (model 40-10,
Analytical Measuring System Ltd). From 0 to <15 mutant colonies/plate
emerged over the incubation period. These colonies were clearly distinguished
from the viable colonies.
Selection and sequencing of lac^ mutations
Dominant mutations (/<jc/""d and lacO0) were identified from total Lacc mutants
by following a simple complementation test (Miller, 1972) Before DNA
sequencing lacO° mutations were removed by genetic mapping (Schmeissner
et al., 1977), since previous reports (Schaaper and Dunn, 1991; Sargentim
and Smith, 1994) have revealed a large contribution of lacCf to spontaneous
dominant lac0 mutations in E.coli. The lacl gene was amplified in a DNA
thermal cycler (Perkin-Elmer Cetus) using a Gene Amp Kit (Perkin-Elmer
Cetus) and two specific 25mer primers (supplied by Operon Technologies)
with their 3'-ends at positions -16 and 1186 of the gene respectively. The
PCR mixture was then subjected to 30 cycles of heat treatment at 94, 55 and
72°C for 1, 1 and 3 mm (2 s extension) respectively and 72°C for 10 min.
PCR products were sequenced using the Applied Biosystems Dye Terminator
Cycle Sequencing Ready Reaction Kit followed by analysis in an ABI model
377 DNA sequencer The synthetic oligonucleotide 3'-AGTTGACCCACGGTCGCACC-5' was used as sequencing primer for the first 300 bp of the lacl
gene All lact^ mutants sequenced for mutational spectrum analysis were of
independent origin.
Results
Spontaneous mutation was measured for ATase-proficient
bacteria and for ATase-deficient derivatives in two ways: in
bacteria growing in non-selective liquid medium and in bacteria
resting on selective agar plates. To this end we constructed
new ada ogt defective strains (Table I) carrying an F' lacF
ZL8 proB+ episome that allows selection of mutants that
synthesize (3-galactosidase constitutively (Lacc phenotype) or
an F' episomal IacZ~ mutation that reverts to /acZ+only via
G:C->A:T transitions (Cupples and Miller, 1989). Two ada
ogt double mutants were used, one carrying an o^r-7::kanR
disruption mutation isolated by Rebeck and Samson (1991)
and the other a A(ogt-fnr)] deletion that covers the ogt gene
and the nearby fnr locus (Abril et al., 1994). Strains carrying
one or the other ogt mutant allele showed no differences with
respect to their sensitivity to mutagenesis by monofunctional
alkylating agents (Abril et al, 1994).
Table II shows the spontaneous mutation rates determined for
ATase-proficient and ATase-deficient strains, from the number
of Lacc colonies present 3 days after plating 3=20 independent
cultures that had grown to stationary phase for 12 h in nonselective liquid medium. The spontaneous frequency of Lacc
mutants was not increased in ATase-deficient bacteria as
compared with ATase-proficient counterparts, even though they
were additionally UvrB-deficient to amplify the sensitivity to
Spontaneous mutations and DNA alkyitransferase
Table n. Spontaneous mutagenesis in bacteria growing in non-selective
liquid medium"
Genotypeb
Mutation frequency
LacZ + revertants (XKT9)
Lac'mutants
+
ada* ogt
ada' og/::kanR
odor A(ogt-frr)
Constitutive
Dominants
49 ± 10
39 ± 9
43 ± 8
11 ± 2
17 ± 4
17 ± 3
7+ 3
10 ± 3
11 ± 3
"Bacteria were grown to stationary phase (12 h) in non-selective LB nutrient
medium as indicated in Materials and methods. Numbers are the average ±
SD of mutant colonies/plate from at least 20 independent cultures.
•"Relevant genetic characteristics. Bacterial strains were UC831O (ada*
ogt+), UC1178 (ada- ogr.:km\K) and UC1011 [ada~ ^(ogt-fnr)} for
determination of Lac' foward mutation frequency and UC932 (ada* ogt+),
UC1264 (ada' ogr: :kanR) and UC1153 [ada' A(ogt-frr)] for determination
of LacZ + reversion frequency. All bacteria carry a A(uvrB-bio) mutation.
'Constitutive mutants are defined by growth on Pgal as the sole carbon
source. Dominants are Lacc constitutive mutants identified by negative
complementation in a simple merodiploid assay.
mutagenesis by putative endogenous alkylating agents (Vidal
et al., 1995, 1997; Ferrezuelo et al, 1998a). The test for
dominance revealed a significant (P <0.01), but very small,
increase (~1.5 times) in the frequency of dominant Lacc
mutants in the absence of ATase repair. A similar small
increment was observed by measuring the reversion to LacZ +
via G:C->A:T transitions. No differences were observed
between the two ada ogt double mutants.
Data in Table II are in accordance with the conclusion that
ada ogt mutants have the same rate of spontaneous mutation
as wild-type E.coli during growth phase (Rebeck and Samson,
1991). To support the proposal that excess mutants arise only
in non-dividing cells we followed accumulation of Lacc mutant
colonies in non-lethal selective plates upon prolonged incubation for 8 days (Figure 1A). In this experiment a low plating
density (103 bacteria/plate) was used to ensure the absence of
pre-existing Lacc mutants, but 10 plates in total (104 bacteria)
were counted. CA7092.1 scavenger cells were added at a high
plating density of 5X10 8 to prevent multiplication of bacterial
tester strains on minimal Pgal plates. Escherichia coli ada+
ogt+ uvr and ada~ ogf::kanR uvr strains differed markedly
in their resting-associated spontaneous mutagenesis. Mutagenesis in ada+ ogt+ uvr bacteria was biphasic, with a low
sensitivity range where little mutagenesis occurred (2 Lacc
mutants/day/104 viable cells distributed over 10 plates)
followed by a high sensitivity range (22 Lacc mutants/day/lCr4
viable cells) that was triggered beyond the first 6 days after
plating. In contrast, the number of spontaneous Lacc mutants
from ada' ogf::kanR uvr cells steadily increased over the 8
day period of incubation at a rate of ~30 Lacc mutants/day/
104 viable cells. Nucleotide excision repair reduced restingassociated spontaneous mutagenesis to 57% in ATase-deficient
bacteria (17 versus 30 Lacc mutants/day/104 viable cells for
the ada" ogc uvr+ and ada~ ogr uvr strains respectively).
An interesting observation was that bacteria carrying the
A(ogt-fnr) deletion were poorly mutated to Lacc throughout
the incubation period, as compared with those having the
og/::kanR mutation (Figure 1A). An ~90% reduction in
spontaneous mutagenesis was thus observed in both the uvr
and uvr+ genetic backgrounds. In contrast, and in agreement
with previous data (Taverna and SedgwicL, 1996), A(ogt-frir)
bacteria were, if anything, somewhat more sensitive to restingassociated spontaneous mutagenesis than isogenic cells carrying the alternative ogr: :kanR mutation when monitoring
reversion to LacZ+ via G:C—»A:T transitions (Figure 2A). In
the presence of scavenger cells the number of viable colonies
from all bacterial strains tested remained practically constant
over the 8 day incubation (Figures IB and 2B). Therefore, the
reduced accumulation of Lacc mutants in A(ogt-fnr) bacteria
seems not to be the result of reduced viability or reduced
growth.
The N-terminal portion of the lad gene, where lact^
mutations occur, is touted as a very sensitive, informational
and economical mutation assay target to study mutagenic
specificity (Schaaper and Dunn, 1991; Sargentini and Smith,
1994). To employ this assay one selects Lacc mutants and then
sorts out the lact^ mutations (see Materials and methods),
which are more valuable for sequencing and analysis. In this
study a total of 124 lact6 forward mutations were subjected
to DNA sequence analysis, in order to compare the spontaneous
mutation spectrum of ATase-deficient (ada' ogf::kanR) bacteria
resting on selective agar plates with that of isogenic ATaseproficient cells growing in non-selective liquid medium. The
comparison has been made in a uvrB mutant background
(UC1178 versus UC8310), to avoid interference from UvrABC
nucleotide excision repair. Mutants selected from ATasedeficient bacteria were those Lacc mutant colonies made visible
from day 6 to day 7 after plating. To ensure the independence
of each lact^ mutation isolated from ATase-proficient bacteria
only one mutant colony per culture was randomly chosen for
DNA sequencing.
All 124 lact6 mutants sequenced were characterized by
either single base pair substitutions, deletions or additions,
therefore, no multiple mutants were isolated. The distribution
by class of mutational events is shown in Table HI. The
spontaneous mutation profile (lacl^ mutation types and their
respective percentages of occurrence) determined in this study
with bacterial strain UC831O (ada+ ogt+ uvr) is in very good
agreement with previous spontaneous lact^ mutation data
from Schaaper and Dunn (1991) and Sargentini and Smith
(1994) for strain NR9102 (ada+ ogt+ uvr+). Base substitutions
were by far the most frequent mutation type (68.8%), followed
by multi-base deletions (17.5%). Among base substitutions,
all six possible pathways were observed. Transitions (42.5%)
were more frequent than transversions (26.3%); among
transitions, G:C—>A:T (33.8%) was decidedly more frequent
than A:T—>G:C (8.7%). The only apparent inconsistency
between this study and previous spontaneous mutation spectra
involves the percentage of single base deletions. We observed
a somewhat higher percentage (13.8%) than that (4%) reported
by Schaaper and Dunn (1991) and Sargentini and Smith
(1994). The spontaneous mutation spectrum of ATase-proficient
bacteria growing in non-selective liquid medium served as a
baseline to determine the type of mutation events increased
in ATase-deficient derivatives upon prolonged incubation on
selective plates. As shown in Table m, the percentage of
spontaneous G:C-»A:T transitions was significantly elevated
(P < 0.005) in the ada ogt double mutant strain compared
with the isogenic ada+ ogt+ parent. This increase is presumably driven by unrepaired C>°-alkylguanine lesions produced
endogenously under non-growth conditions (Mackay et al,
1994). In contrast to G:C-)A:T transitions, no other mutational
events showed a statistically significant difference, as calculated with Fisher's exact test for differences in proportions.
369
A.VMal, N.Abril and C.Pueyo
15.0 _ A
o
QJ
/
|
7-5 h
ada'
unr"
• ada'
'A
ada*
E
ada' A(ogt-fnr) uvr
ada' Mogt-fnr) uvr*
0.0 1
l
1
1
1
1
3 4 5 6 7 8
3 4 5 6 7 8
Days
Days
Fig. 1. Accumulation of spontaneous Lacc mutant colonies in E.coh strains with different DNA repair capabilities. Bacteria (IO3), together with CA7O92.1
scavenger cells (5X 10*), were plated on selective agar. The accumulating Lacc mutants (A) and the viability of the plated cells (B) were determined on the
days indicated. Data are the mean numbers of Lacc mutants or viable colonies/plate, with 10 plates/experiment and three independent determinations. Strains
were as follows: UC8310 ( • ) , UC1178 (O), UC1011 (O), UC1175 (A) and UC1263 (A); relevant genetic characteristics are indicated in (A)
80 _ A
,o
01
o
•
4->
c
fldfl" A(o^t-/nr) ut>r~
/ ,o°
40
flda* ogt::kan R wrr"
1
1
ada* ogt* uvr
.
•
•
*
n
0
*••
i
i
i
1 1 1 1
l
1
1
1
1
2 3 4 5 6 7 8
Days
1 1
2 3 4 5 6 7
Days
Fig. 2. Accumulation of spontaneous Lac + revenants in E.coh strains with different DNA repair capabilities Bacteria (109). together with CA7O92.1
scavenger cells (5x 108). were plated on selective agar. The accumulating Lac + rcvertants (A) and the viability of the plated cells (B) were determined on the
days indicated. Data arc the mean numbers of Lac + rcvertants or viable colonies/plate, with 10 plates/experiment and three independent determinations.
Strains were as follows. UC932 ( • ) , UC1264 (O), and UC1153 (O). relevant genetic characteristics are indicated in (A)
The distribution within the DNA target of spontaneous
G:C—>A:T transitions is presented in Table IV. Mutations
were not randomly distributed, hence they were recovered at
nine (ada+ ogt+) and 14 {ada~ ogr) out of 24 potentially
available sites for G:C—»A:T transitions in the A/-terminal
region of the lacl gene (Pienkowska et ai, 1993; Jurado et ai,
370
1995). The ATase-deficient strain accumulated significantly
more G:C->A:T mutations at positions 95 (P < 0.05) and 174
(P < 0.005) than the ATase-proficient counterpart. Furthermore, assuming a Poisson distribution, sites 84, 93 and 198
in the ATase-proficient spectrum and site 174 in the ATasedeficient spectrum can be considered as real hotspots
Spontaneous mutations and DNA alkyitransferase
Table HL Classification of spontaneous lact^ mutations"
Mutational class
Change
Base-pair substitutions
Transitions
G:C-»AT
A:T-»G:C
Subtotal
Transversions
G:C->T:A
G:C->C:G
A:T->T:A
A:T->C:G
Subtotal
Other mutational events
Single base deletions
Multiple base deletions
Multiple base additions
Mutants collected
Table IV. Site specificity of spontaneous G:C-»A:T transitions*
ada+ ogt+
(growing)
ado" ogn:kanR
(resting)
Siteb
Sequence0
ada+ ogt+
(growing)
ada~ ogr.:kanR
(resting)
55 (68.8)
37 (84.1)
42
56
57
75
80
84
90
92
93
95
104
120
129
140
173
174
179
185
186
188
191
198
201
206
Total
ACG
CGC
GCA
TCT
TCA
ACC
TCC
CCG
CGC
CGT
CCA
TCT
ACG
AGT
TCC
CCC
CCG
GGC
GCA
ACA
ACA
GCG
GGC
ACA
0 (0.0)
0 (0.0)
0 (0.0)
2 (2.5)
0 (0.0)
4 (5.0)
0 (0.0)
3 (3.8)
5 (6.2)
0 (0.0)
3 (3.8)
2 (2.5)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
2 (2.5)
0 (0.0)
0 (0.0)
4 (5.0)
2 (2.5)
0 (0.0)
80 (100)
0 (0.0)
0 (0.0)
2 (4.5)
2 (4.5)
1(2.3)
2 (4.5)
2(4.5)
1 (2.3)
1 (2.3)
3 (6.8)d
0 (0.0)
2 (4.5)
0 (0.0)
0 (0.0)
2 (4.5)
6 (13.6)e
0 (0.0)
1 (2.3)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
1 (2.3)
2 (4.5)
44(100)
b
27 (33.8)
7 (8.7)
34 (42.5)
28 (63.6)
1 (2.3)
29 (65.9)c
9(11.2)
3 (3.8)
7 (8.8)
2 (2.5)
21 (26.3)
25(31.3)
11 (13.8)
14 (17.5)
0 (0.0)
80(100)
4(9.1)
2(4.5)
1 (2.3)
1 (2.3)
8(18.2)
7 (15.9)
2 (4.5)
4(9.1)
1 (2.3)
44(100)
mutations are classified by mutation type based on DNA sequence
analysis. Values are the numbers of occurrences; the percentages are in
parentheses. Bacterial strains are UC8310 (ada+ ogt+) and UC1178 (ada~
ogr:kan R ); both strains carry a uvrB mutation.
b
P < 0.005 by Fisher's exact test for comparison between strains.
C
P < 0.05 by Fisher's exact test for comparison between strains.
(P < 0.05). Position 174 is a common hotspot in mutagenesis
by alkylating agents (Jurado et al, 1995; Vldal et al, 1995,
1997), therefore suggesting that all these differences might
reflect the influence of unrepaired C^-alkylguanine lesions on
the spontaneous G:C—»A:T mutation spectrum in the ATasedeficient background. In fact, the two G:C—>A:T spectra were
statistically different, with a P value of <10~3, according to a
computer program for comparison of mutation spectra (Cariello
et al, 1994). DNA sequence analysis is difficult to perform
in a double-stranded target without knowing the exact nature
of the mutagenic damage. To this end we assumed that guanines
are the bases from which G:C—>A:T transitions arise in resting
ATase-deficient bacteria (Table V). No DNA strand preference
was observed, while a highly significant (P <0.005) bias in
favor of mutations at G residues preceded by a 5'-Pu base
could be inferred.
Discussion
Previous studies based mostly on reversion (or suppression)
of specific mutant alleles have shown that a deficiency in
repair by the E.coli ATases results in a mutator phenotype,
particularly in non-dividing cells (Rebeck and Samson, 1991;
Mackay et al, 1994; Taverna and Sedgwick, 1996). Here we
used a non-specific forward mutation assay system to study
spontaneous mutagenesis in ATase-deficient bacteria resting
on selective agar plates as compared with isogenic ATaseproficient cells growing in non-selective liquid medium.
Lac constitutive mutant colonies from ada~ ogr uvr E.coli
accumulated steadily upon prolonged incubation for 8 days on
non-lethal Pgal selective plates. In contrast, accumulation of
Lacc mutants from isogenic ada+ ogt+ uvr bacteria only
began 6 days after plating. This observation is in agreement
with the role attributed to both the Ada and the Ogt ATases
in protecting non-dividing cells against endogenously produced
DNA alkyl lesions (Rebeck and Samson, 1991; Mackay et al.,
1994; Taverna and Sedgwick, 1996). Our data suggest further
that in the absence of nucleotide excision repair the repair
•Values are the number of mutants recovered at each site. Percentages are in
parentheses. Bacterial strains are UC8310 (ada+ ogt+) and UC1178 (ada~
ogt: :kanR); both strains carry a uvrB mutation.
b
Nucleotide numbering of the lad gene according to Farabaugh (1978). All
sites known from other studies to generate the LacH phenotype via a
G:C—>A:T transition are included.
Tlie 5'—>3' sequence of the non-transcribed strand of lad wild-type
sequence with the mutated base in the central position.
i
P < 0.05 by Fisher's exact test for comparison between strains.
C
P < 0.005 by Fisher's exact test for comparison between strains.
process performed by these two suicide DNA alkyltransferases
is not saturated until the cells have been resting for 6 days.
Spontaneous Lacc mutant colonies from adar ogr uvr E.coli
made visible from day 6 to day 7 after plating on selective
agar were screened for dominant mutations; those mutations
located in the N-terminal portion of the lad gene were then
characterized at the molecular level by DNA sequencing
analysis. The spontaneous lacr4 mutation spectrum of ada+
ogt+ uvr bacteria growing in non-selective liquid medium
served as a reference in order to determine the precise nature
of the mutational alterations and the contextual nature of the
DNA sites of mutation driven from unrepaired alkyl lesions
generated under non-growth conditions.
The spontaneous mutation profile of resting ATase-deficient
bacteria showed a significant increment in the percentage of
G:C—»A:T transitions, at the expense of other mutation types.
This result agrees with previous data from reversion of specific
lacZ mutant alleles, showing that while G:C—»A:T transitions
accumulate during the first week after plating, accumulation
of other types of base pair substitutions only begins 7-9 days
after plating (Mackay et al., 1994). Since it is well established
that C^-alkylguanine can mispair with thymine (reviewed by
Horsfall et al., 1990), we infer that endogenously produced
O6-alkylguanine lesions are responsible for the extra G.C—»A:T
transition mutations. The DNA sequence analysis of this
mutation class in resting ATase-deficient cells revealed a
pronounced 5'-PuG bias, suggesting that the endogenous
371
A.Vldal, N.AbrU and C.Pueyo
Table V. Influence of the 5'-flanking base and the DNA strand on the distribution of spontaneous G C—>A T transitions from resting ada ogr bacteria*
5 Base or strand
No. of sites
Observed mutations
Expected mutations
Average per site
Purine
Pyrimidine
9
15
18
10
10.5
175
2.0
0.1
Non-transcnbed
Transcribed
6
18
6
22
7.0
21.0
10
1 2
X-
I*
86
<0.005
0.012
>0.9
R
"Bacterial strain is UC1178 (ada ogr::kan ), these strains carry also a uvrB mutation. It is assumed that G residues are the bases from which G:C—>A:T
transitions arise.
''Calculated with 1 degree of freedom.
metabolite(s) responsible for the burst of spontaneous
G:C—>A:T mutations is an S^l type alkylating compound(s).
We have thoroughly examined the distribution along the
lad target of G:C—>A:T transitions induced by serial alkylnitrosoureas in ATase- and Uvr-deficient E.coli as compared
with the corresponding ATase-proficient parent (Jurado et ai,
1995; Vidal et ai, 1997; Ferrezuelo et ai, 1998a; Ferrezuelo
et ai, 1998b, in press). The site distribution of G:C->A:T
transitions in non-dividing ATase-deficient bacteria showed
basic similarities with the alkylnitrosourea-induced spectra,
e.g. the highest percentage (13.6%) of this mutational change
was at position 174, which is a hotspot common to all
alkylnitrosourea-induced spectra, accumulating as an average
of 10% of all mutations. Furthermore, that site distribution
showed greater similarities to the spectra induced by complex
alkylnitrosoureas (like PNU, BNU, and CCNU) than to those
induced by simple methylating or ethylating counterparts
(MNU and ENU). Thus G:C->A:T transitions at positions 80,
95 and 173 have been observed with PNU, BNU and/or CCNU
(Jurado et ai, 1995; Ferrezuelo et ai, 1998a) but not with
MNU or ENU (Vidal et ai, 1997). In fact, position 173 is
insensitive to mutagenesis by simple monofunctional alkylating
agents (Horsfall et ai, 1990; Vidal et ai, 1995, 1997); as
single events mutations at this site have been observed only
with the bifunctional alkylnitrosourea CCNU (Jurado et ai,
1995). Likewise, only one G:C—>A:T out of 44 mutations
(2.3%) was detected at position 185 in resting ATase-deficient
bacteria, in concordance with the relatively low sensitivity of
this site to mutagenesis by PNU, BNU and CCNU (around
4% of all induced mutations) and in contrast to the fact that
185 is a hotspot common to both MNU- (15.6 %) and
ENU-induced (9.5%) spectra. All of these observations form
compelling experimental support for the idea that endogenous
alkylating mutagens generate bulky DNA adducts rather than
simple methylated or ethylated lesions (Mackay et ai, 1994;
Sedgwick, 1997). As indicated in this work, the bulky adducts
should be removed from DNA by both ATase and nucleotide
excision repair activities.
A miscoding lesion in the transcribed strand of DNA
could, by inducing a transcriptional error, trigger a round of
DNA replication, thus increasing the probability of being
permanently fixed as a mutation. The generation of spontaneous
mutants from C^-alkylguanine lesions in non-dividing ada ogt
double mutant E.coli has been explained by such a mechanism
(Rebeck and Samson, 1991). Nevertheless, here we do not
observe any significant DNA strand bias in the distribution of
spontaneous G:C—»A:T transitions from resting ATase-deficient
bacteria, thus suggesting that C^-alkylguanine lesions occurring
on the transcribed strand of the mutation sites are not preferen372
tially fixed as mutations. DNA replication under non-dividing
conditions has been postulated for fixation of mutations of
those lesions arising in the non-transcribed strand of DNA
(Foster and Cairns, 1992). The synthesis of DNA under
starvation conditions seems to be indisputable (Bridges, 1996),
though the extent and nature of such synthesis is under debate,
in particular for those genetic targets located on an F' episome
(see for example Caims, 1995; Galitski and Roth, 1995;
Radicella et ai, 1995).
Bacterially catalyzed nitrosation has been implicated in
producing an endogenous alkylating agent(s) during prolonged
incubation under starvation conditions (Taverna and Sedgwick,
1996; Sedgwick, 1997). Possible candidates for the nitrosatable
metabolite(s) include methylamine and methylurea, which arise
as catabolites and on nitrosation generate potent mutagenic
methylating agents (Taverna and Sedgwick, 1996). However,
amines that occur more frequently in cells may be more
significant candidates. Among them, nitrosated polyamines,
like spermidine and azaserine (a model compound for nitrosated
peptides), have been shown to be mutagenic to E.coli ada ogt
mutants, particularly in a uvr-deficient genetic background
(Sedgwick, 1997). The results reported in this paper strengthen
the relevance that bulky alkylated DNA adducts from nitrosated
spermidine and other polyamines might have in spontaneous
mutagenesis and carcinogenesis.
Three types of E.coli mutants, narG, fnr and moa, are
known to be deficient in nitrosation. Taverna and Sedgwick
(1996) have shown that a moa mutation, but not an fnr
mutation, eliminates much of the spontaneous mutagenesis in
ada ogt mutants, as measured by reversion via G:C—»A:T
transition of a specific lacZ mutant allele. They explain the
greater effectiveness of the moa mutation because it eliminates
all three known E.coli nitrate reductase activities; in contrast,
the fnr mutation eliminates only the major nitrate reductase.
Here we confirm the results of Taverna and Sedgwick (1996)
when measuring accumulation of LacZ+ revertants. but in
contrast we show that a deletion covering the fnr locus reduces
~90% of the spontaneous mutagenesis in ada ogt mutants
when using a non-specific forward mutation assay system
(Lacc mutant selection). The reason for such an apparent
discrepancy might be the fact that a non-specific forward
mutation assay should be sensitive to any mutagen generated
during prolonged incubation of the bacteria, while the specific
reversion system detects only a portion of all possible mutagen
specificities.
Acknowledgements
This work was subsidized by CICYT grants SAL91-0842-CE and PB91-0842.
EC contract EV5V-CT91-OO12 and by the Junta de Andaluci'a (group no
Spontaneous mutations and DNA alkyltransferase
3167). A.V. is the recipient of a fellowship from the Ministerio de Educaci6n
y Ciencia (Spain).
References
Abril,N., Hera,C, AJejandreJE., RaffertyJ.A., Margison,G.P. and Pueyo,C.
(1994) Effect of ogt expression on mutation induction by methyl-, ethyland propylmethanesulphonate in Escherichia coli K-12 strains. Mol Gen.
Genet., 242, 744-748.
Bridges,B.A. (1996) Elevated mutation rate in mutT bacteria during starvation:
evidence for DNA turnover? J. Bacterial., 178, 2709-2711.
CairnsJ. (1995) Adaptive mutation and sex. Science, 269, 288.
Cariello^JJ., Piegorsch,W.W., Adams,W.T. and Skopek,T.R. (1994) Computer
program for the analysis of mutational spectra: application to pS3 mutations.
Carcinogenesis, 15, 2281-2285.
Cupples.C.G. and MillerJ.H. (1989) A set of lacZ mutations in Escherichia
coli that allow rapid detection of each of the six base substitutions. Proc.
Natl Acad. Sci. USA, 86, 5345-5349.
Davis,R.W., BotsteinJ). and RothJ.R. (1980) A Manual for Genetic
Engineering. Advanced Bacterial Genetics. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY.
Farabaugh,PJ. (1978) Sequence of the lad gene. Nature, 274, 765-769.
FerrczueloJ1., Prieto-Alamo,MJ., JuradoJ. and Pueyo.C. (1998a) Influence of
DNA repair by (A)BC excinuclease and Ogt alkyltransferase on the
distribution of mutations induced by ^-propyl-N-nitrosourea in Escherichia
coli. Environ. Mol Mutagen., 31, 82-91
FerrezueloJ1., Prieto-Alarao,MJ., JuradoJ. and Pueyo.C. (1998b) Role of
DNA repair by (A)BC excinuclease and Ogt alkyltransferase on the
final distribution of LacI^ mutations induced by A/-butyl-Ar-nitrosourea in
Escherichia coli. Mutagenesis, 13, in press.
FosterJ?.L. and CairnsJ. (1992) Mechanisms of directed mutation. Genetics,
131, 783-789.
Friedberg E.C., Walker G.C. and Siede W. (1995) DNA Repair and
Mutagenesis. American Society for Microbiology, Washington, DC.
Galitski,T. and RothJ.R. (1995) Evidence that F plasmid transfer replication
underlies apparent adaptive mutation. Science, 268, 421-423.
Horsfall.MJ., Gordon.AJ.E., Bums.P.A., Zielenska^l., van der Vliet,G.M.E.
and Glickman.B.W. (1990) Mutational specificity of alkylating agents and
the influence of DNA repair. Environ. Mol. Mutagen., 15, 107-122.
JuradoJ., Ferrezuelo.F. and Pueyo.C. (1995) Mutational specificity of l-(2chloroethyl)-3-cyclohexyl-l-nitrosourea in the Escherichia coli loci gene
of (Aalkylguanine-DNA alkyltransferase-proficient and -deficient strains.
Mol. Carcinogen., 14, 233-239.
Lees,E.M. and Harlow,E. (1995) Cancer and the cell cycle. In Hutchison.C.
and Glover.D.M. (eds), Cell Cycle Control. Oxford University Press, New
York, NY, pp. 228-263.
Lindahl.T. (1993) Instability and decay of the primary structure of DNA.
Nature, 362, 709-715.
Loeb.L.A. (1991) Mutator phenotype may be required for multistage
carcinogenesis. Cancer Res., 51, 3075-3079.
Mackay.WJ., Han,S. and Samson.L.D. (1994) DNA alkylation repair limits
spontaneous base substitution mutation in Escherichia coli. J. Bacleriol.,
176, 3224-3230.
MillerJ.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
PeggAE. (1990) Mammalian (Aalkylguanine-DNA alkyltransferase:
regulation and importance in response to alkylating carcinogenic and
therapeutic agents. Cancer Res., 50, 6119-6129.
Pienkowska^M., Glickman,B.W., Ferreira^A., Andersonjvl. and Zielenska>l.
(1993) Large-scale mutational analysis of EMS-induced mutation in the
lacl gene of Escherichia coli. Mutat. Res., 288, 123-131.
Potter.P.M., Wilkinson,M.C., FittonJ., Can\FJ., BrennandJ., CooperJ3.P. and
Margison.G.P. (1987) Characterisation and nucleotide sequence of ogt O6alkylguanine-DNA-alkyltransferase gene of E.coli. Nucleic Acids Res., 15,
9177-9193.
Potter.P.M., Kleibl.K., CawkwellJ-. and Margison.G.P. (1989) Expression of
the ogt gene in wild-type and ada mutants of E.coli. Nucleic Acids Res.,
17, 8047-8060.
RadicellaJ.P., ParkJ'.U. and Fox>I.S. (1995) Adaptive mutation in Escherichia
coli: a role for conjugation. Science, 268, 418^-420.
Rebeck.G.W. and Samson,L. (1991) Increased spontaneous mutation and
alkylation sensitivity of Escherichia coli strains lacking the ogt O6methylguanine DNA repair methyltransferase. J. Bacterial., 173,2068-2076.
RebeclcG.W., SmiuYCM., Goad£>.L. and SamsonJ^. (1989) Characterization
of the major DNA repair methyltransferase activity in unadapted Escherichia
coli and identification of a similar activity in Salmonella typhimurium. J.
Bacterial., 171, 4563-4568.
RobinsJ5. and CairnsJ. (1979) Quantitation of the adaptive response to
alkylating agents. Nature, 280, 74-76.
Ruiz-Rubio,M., Hera,C. and Pueyo,C. (1984) Comparison of a forward and
reverse mutation assay in Salmonella typhimurium measuring L-arabinose
resistance and histidine prototrophy. EMBO J., 3, 1435-1440.
SamsonJ-. (1992) The suicidal DNA repair methyltransferases of microbes.
Mol. Microbiol., 6, 825-831.
Sargentini,NJ. and Smith.K.C. (1994) DNA sequence analysis off-radiation
(anoxic)-induced and spontaneous lacf^ mutations in Escherichia coli K12. Mutat. Res.. 309, 147-163.
Schaaper,R.M. andDunn,R.L. (1991) Spontaneous mutation in the Escherichia
coli lacl gene. Genetics, 129, 317-326.
Schmeissner.U., Ganem,D. and Miller, J.H. (1977) Genetic studies of the lac
repressor. II. Fine structure deletion map of the lacl gene, and its correlation
with the physical map. J. Mol. Biol., 109, 303-326.
Sedgwick,B. (1997) Nitrosated peptides and polyamines as endogenous
mutagens in C^-alkylguanine-DNA alkyltransferase deficient cells.
Carcinogenesis, 18, 1561—1567.
Singer.B. and Grunberger.D. (1983) Molecular Biology of Mutagens and
Carcinogens. Plenum Press, New York, NY, pp. 55-68.
TavernaJ'. and SedgwickJ3. (1996) Generation of an endogenous DNAmethylating agent by nitrosation in Escherichia coli. J. Bacleriol., 178,
5105-5111.
TeoJ., SedgwiclcB., Demple3-, Li,B. and Lindahl.T. (1984) Induction of
resistance to alkylating agents in E.coli: the ada+ gene product serves both
as a regulatory protein and as an enzyme for repair of mutagenic damage.
EMBO J., 3, 2151-2157.
VaughanJ5., Sedgwick,B., HallJ., GannonJ. and LindahlJ. (1991)
Environmental mutagens that induce the adaptive response to alkylating
agents in Escherichia coli. Carcinogenesis, 12, 263-268.
Vidal^A., Abril.N. and Pueyo.C. (1995) DNA repair by Ogt alkyltransferase
influences EMS mutational specificity. Carcinogenesis, 16, 817-821.
Vidal,A., AbrilJJ. and Pueyo.C. (1997) The influence of DNA repair by Ogt
alkyltransferase on the distribution of alkylnitrosourea-induced mutation in
Escherichia coli. Environ. Mol Mutagen., 29, 180-188.
Wilkinson^M.C, Potter.P.M., Cawkwell.L., GeorgiadisJ5., Patel.D., Swann.P.F.
and Margison.G.P. (1989) Purification of the E.coli ogt gene product to
homogeneity and its rate of action on O6-methylguanine, C^-ethylguanine
and O4-methylthymine in dodecadeoxyribonucleotides. Nucleic Acids Res.,
17, 8475-8484.
Xiao.W. and Samson.L. (1993) In vivo evidence for endogenous DNA
alkylation damage as a source of spontaneous mutation in eukaryotic cells.
Proc. Natl Acad. Sci. USA, 90, 2117-2121.
YnmaHa M Sedgwick,B., Sofuni.T. and Nohmi,T. (1995) Construction and
characterization of mutants of Salmonella typhimurium deficient in DNA
repair of (Amethylguanine. J. Bacterial., IT7, 1511-1519.
Received on November 3, 1997; accepted on January 7, 1998
373