Formation of 8-oxoguanine in cellular DNA of Escherichia coli strains defective in different antioxidant defences

Abstract

This paper examines the relationship in Escherichia coli between the in vivo content of 8-oxoguanine (8-oxoG) in chromosomal DNA and deficiencies of various key antioxidant defences. The structural genes for catalases (katG and katE), cytosolic superoxide dismutases (sodA and sodB) or formamidopyrimidine-DNA glycosylase (fpg) were inactivated to obtain bacterial strains lacking the scavenger enzymes for H 2 O 2 or O 2 "~ or the DNA repair protein for 8-oxoG. Wild-type bacteria showed 5-fold increased sensitivity to both lethality and mutagenesis by H 2 O 2 in K medium (1% casamino acids and 1% glucose), as compared with nutrient broth. This higher sensitivity was associated with increased chromosomal oxidative damage, estimated as the 8-oxodG content, and with a marked decrease in both catalase and SOD activities. Bacteria lacking both cytosolic SODs (sodA sodB mutant) displayed increased 8-oxodG content in chromosomal DNA (2.8-fold that of the wild-type) when grown under standard aerated conditions. Comparatively, no significant difference in 8-oxodG content was observed in cells grown without aeration. Bacteria totally devoid of catalase activity (katG katE mutant) showed wild-type contents of 8-oxodG in chromosomal DNA when grown under aerated conditions. Nevertheless, the protective role of catalase in preventing formation of 8-oxodG in chromosomal DNA became evident under oxidative stress conditions: growth under hyperoxygenation and, particularly, following H 2 C>2 exposure. Catalase defi-ciency resulted in a dramatic decrease in viability after H 2 O 2 exposure. A deficiency of Fpg protein also sensitized E.coli to H 2 O 2 lethality, though to lesser extent than a deficiency of catalase activity. However, the scavenger enzyme and the DNA repair protein protected equally against 8-oxoG formed in vivo upon H 2 O 2 treatment Introduction

Full text

Page 1

Mutagenesb vol.13 no.6 pp.589-594, 1998
Formation of 8-oxoguanine in cellular DNA of Escherichia coli
strains defective in different antioxidant defences
Jos£ Alhama, Julia Ruiz-Laguna,
Antonio Rodriguez-Ariza, Fermm Toribio,
Juan L6pez-Barea and Carmen Pueyo1
Depaitamento de Bioqufmica y Biologfa Molecular, Universidad de
Cdrdoba, Avenida de Medina Azahara s/n, 14071-C6rdoba, Espafla
This paper examines the relationship in Escherichia coli
between the in vivo content of 8-oxoguanine (8-oxoG)
in chromosomal DNA and deficiencies of various key
antioxidant defences. The structural genes for catalases
(katG and katE), cytosolic superoxide dismutases (sodA and
sodB) or formamidopyrimidine-DNA glycosylase (fpg) were
inactivated to obtain bacterial strains lacking the scavenger
enzymes for H2O2 or O2"~ or the DNA repair protein
for 8-oxoG. Wild-type bacteria showed 5-fold increased
sensitivity to both lethality and mutagenesis by H2O2 in K
medium (1% casamino acids and 1% glucose), as compared
with nutrient broth. This higher sensitivity was associated
with increased chromosomal oxidative damage, estimated
as the 8-oxodG content, and with a marked decrease in
both catalase and SOD activities. Bacteria lacking both
cytosolic SODs (sodA sodB mutant) displayed increased
8-oxodG content in chromosomal DNA (2.8-fold that of the
wild-type) when grown under standard aerated conditions.
Comparatively, no significant difference in 8-oxodG content
was observed in cells grown without aeration. Bacteria
totally devoid of catalase activity (katG katE mutant)
showed wild-type contents of 8-oxodG in chromosomal
DNA when grown under aerated conditions. Nevertheless,
the protective role of catalase in preventing formation of
8-oxodG in chromosomal DNA became evident under
oxidative stress conditions: growth under hyperoxygenation
and, particularly, following H2C>2 exposure. Catalase defi-
ciency resulted in a dramatic decrease in viability after
H2O2 exposure. A deficiency of Fpg protein also sensitized
E.coli to H2O2 lethality, though to lesser extent than a
deficiency of catalase activity. However, the scavenger
enzyme and the DNA repair protein protected equally
against 8-oxoG formed in vivo upon H2O2 treatment
Introduction
Hydrogen peroxide (H2O2), one of the reactive oxygen species
generated in many enzymatic and non-enzymatic reactions of
dioxygen, is detoxified in Escherichia coli mainly by two
distinct species of catalase. The katG gene encodes a bifunc-
tional catalase, hydroperoxidase I (HPI) (Loewen etal., 1985),
which is transcriptionally induced by OxyR as part of the
genetic response to H2O2 (Storz et al., 1990). The katE gene
codes for the monofunctional HPII (Loewen, 1984) and is
under the control of rpoS, an alternative a factor which is
activated in stationary phase cells (Sak et al, 1989). Recent
studies have also shown an OxyR-independent regulation of
HPI by rpoS as exponentially growing cells enter stationary
phase (Ivanova et al., 1994). Weak acids such as acetate,
which accumulate in the supernatant of stationary phase
cultures, have been implicated in the RpoS-dependent induction
of both the katE and katG genes (Schellhorn and Stones, 1992;
Mukhopadhyay and Schellhorn, 1994).
Superoxide radical (O2*~) is scavenged by superoxide dismut-
ases (SODs). Escherichia coli possesses two cytosolic SODs
that employ manganese (MnSOD) or iron (FeSOD) as the
catalytic metal. It has recently been discovered that mutants
lacking both MnSOD and FeSOD are not truly SOD nulls;
they contain modest levels (~2%) of a periplasmic copper-
zinc SOD (CuZnSOD) which is expressed only in late-
stationary phase culture (Benov and Fridovich, 1994; Imlay
and Imlay, 1996). The genes for MnSOD, FeSOD and
CuZnSOD are designated sodA, sodB and sodC respectively.
The three SODs from E.coli are differently regulated, possibly
according to their respective biological functions (Touati,
1997). CuZnSOD might play a role during starvation and
defence against external oxidative stress. Due to its recent
discovery, regulation of CuZnSOD is not yet fully understood,
although it increases 32-fold upon transition to aerobiosis
(Benov and Fridovich, 1994). FeSOD seems to be responsible
for housekeeping defence against inactivation of several [FeS]-
containing enzymes during the transition to aerobiosis, before
MnSOD is induced. The expression of FeSOD is independent
of oxygen and unaffected by a range of environmental perturba-
tions (Touati, 1997). MnSOD is a multi-inducible enzyme that
allows the bacterium to face various oxidative stress conditions.
Several global regulators interact to control sodA transcription
as a member of the soxRS regulon (Jamieson and Storz, 1997).
Cellular DNA damage by H2O2 arises from the highly
reactive hydroxyl radicals (OH*) generated in the Fenton
reaction between H2O2 and metal ions, such as Fe2"1" (Epe,
1995). It has been recently proposed that O2*~ would accelerate
DNA oxidation through its ability to release Fe2+ from
superoxide-sensitive [FeS] clusters (Keyer et al., 1995). This
new idea is considered a more satisfying physiological explana-
tion than the conventional role envisioned for O2"~ as a
reductant for iron (Touati, 1997). 7,8-Dihydro-8-oxoguanine
(8-oxoG) is one of the multiple oxidative damages induced in
DNA by OH* (Halliwell, 1993; Epe, 1995). This lesion
represents >30-50% of the total base modification products
induced by different OH'-producing experimental models, thus
being considered as a 'fingerprint' of OH* attack on DNA
(Dizdaroglu et al, 1991). Development of a HPLC-based
technique, coupled with highly sensitive electrochemical detec-
tion (EC), for the quantification of 8-oxodG in hydrolyzed
DNA has led to the popularization of this stable oxidized
nucleoside as an index of oxidative DNA damage (Floyd et al,
1986). 8-oxoG contributes to the oxidative mutational load by
inducing targeted GC-»TA transversions upon DNA replication
(Wood et al., 1990). In E.coli, 8-oxoG positioned opposite
'To whom correspondence should be addressed. Tel: 57 218695; Fax: 57 218688; Email: bblpucuc@uco.es
© UK Environmental Mutagen Society/Oxford University Press 1998
589

Page 2

J.AIhama et al
cytosine in DNA is repaired by formamidopyrimidine-DNA
glycosylase (Fpg), encoded by the fpg gene (Chung et al,
1991; Tchon et al., 1991).
We have previously demonstrated that E.coli strains lacking
catalases or cytosolic SODs due to inactivation of the katG,
katE, sodA or sodB genes are extremely sensitive to both
the lethal and mutagenic actions of H2O2, H2O2-generating
mixtures of compounds and O2*~-dependent reactants (Abril
and Pueyo, 1990; Prieto-Alamo et al., 1993). Here we examine
the relationship between this increased sensitivity to oxidative
stress and the in vivo formation of 8-oxo-deoxyguanine in
chromosomal DNA.
Materials and methods
Reagents
H2O2 was purchased as a 30% solution from Sigma Chemical Co (St Louis,
MO). The Genomics DNA isolation kit was obtained from Talent (Trieste,
Italy) and the ASAP kit from Boehnnger (Indianapolis, IN). All other reagents
used in this study were of analytical grade
Bacterial strains
Escherichia colt K-12 strain UC575 (Abril et al., 1992) is a catalase- and
SOD-proficient strain, considered as the parental wild-type in this study.
UC1058 (catalase-deficient) and UC1084 (SOD-deficient) are derivauve strains
isolated in this work These two strains are equivalent to UC498 [katG
katEr.TnlO double mutant] (Abril and Pueyo, 1990) and UC628 [(j<*i4:.Mud
PR13)25 (sodB-kan)l-A2 double mutant] (Pneto-Alamo et al., 1993) respect-
ively, but Uvr+ and without the mutagenesis-enhancing plasmid pKMlOl
UCI057 is as UC575 but fpg-1: KnR Genetic manipulations were carried out
as described by Miller (1992). Successful transfer of the fpg-l KnR mutation
was confirmed by assaying Fpg activity in cell-free extracts, as described by
Boiteux and Huisman (1989) Successful transfer of the uvrB+ gene and loss
of pKMlOl was confirmed by screening for mutagenesis by 4-mtroquinoline-
1-oxide
Media
Nutrient broth (NB) nutrient broth powder 13 g/1 Nutrient agar NB solidified
with Difco agar (17 g/1). K medium: 1 % glucose, 1 % casamino acids, thiamine,
nicotinic acid and arginine in M9 minimal medium (Gerhardt et al., 1994).
Selective plates for L-arabinose-resistant mutants (AraR): Vogel-Bonner
minimal medium E (Gerhardt et al., 1994) containing Difco agar (17 g/1),
glycerol, L-arabinose, D-biotin, thiamine, nicotinic acid, adenine, arginine,
methionine and casamino acids (0 25 g/1). Sugars were at 2 g/1, amino acids
at 40 (ig/ml, vitamins at 5 Hg/ml and adenine at 100 ng/ml Top agar Difco
agar (6 g/1), NaCI (5 g/1) and D-glucose (0 25 g/1).
Enzymatic assays
Cell-free extracts were prepared by ultrasonic treatment (30 s pulses and 15 s
pauses until lysis was observed) (Prieto-Alamo el al., 1993) Catalase activity
was assayed spectrophotometncally by measuring H2O2 decomposition at
240 nm (Beers and Sizer, 1952). The peroxidase activity of HPI was assayed
in dialyzed extracts by monitoring H2O2 decomposition at 460 nm with
o-dianisidine as hydrogen donor (Mukhopadhyay and Schellhom, 1994). One
unit of catalase or peroxidase activity is defined as the amount of enzyme
that decomposes 1 (Jjnol F^Oj/min at 25°C. Protein concentration was
quantified as described by Bradford (1976). Superoxide dismutase activity
was assayed in dialyzed extracts by the xanthine/xanthine oxidase/cytochrome
c assay at pH 7 8. One SOD unit is the amount of enzyme inhibiting the rate
of cytochrome c reduction by 50% under the conditions specified (McCord
and Fridovich. 1969).
Mutagenesis and survival assays
Bacteria were grown at 37°C for 14 (nutrient broth) or 16 h (K medium) with
gentle shaking (90 r.p.m.). Cells were harvested by centrifugation and
resuspended in 50 mM Tris-HCI buffer (pH 8). Samples of 10 ml of this
bacterial suspension in 100 ml Erienmenyer solution and the appropriate dose
of H2O2 were incubated at 37°C for 30 min with shaking (150 r.p.m.). After
centrifugation. 0.1 ml aliquots containing ~5X 107 bacteria were added to 2 ml
molten top agar containing 0.5 mg D-glucose and poured on selective plates
for AraR mutants. For survival determinations, 0.1 ml aliquots containing
-5X103 bacteria were plated on nutrient agar. Viable colonies were counted
after 24 h and mutants after 72 h at 37°C. except for the sodA sodB double
mutant, for which mutant colonies were counted after 144 h incubation at
37°C. Bacterial colonies were counted automatically (model 40-10; Analytical
Measuring System Ltd) Each mutagenesis and survival assay was repeated
100
10
oar>-o-»-«
rt
C/5
0,1
0,01
0,001
0,0001
NB
(LD . 265 mM)
K medium
100
1000
10 100 1000
H2O2 (mM)
Fig. 1. Effect of growth medium on sensitivity of the wild-type strain
(UC575) to the lethal and mutagenic actions of H2O2 The survival
percentages (left) and the numbers of AraR mutants per selective plate
(right) were plotted as function of the H2O2 dose. Bacteria were grown to
stationary phase in NB (•) or K medium (O) Lethal and mutagenic
potencies are shown in parentheses.
on at least two separate occasions with a wide range of H2O2 concentrations.
The mutagenic potency was expressed as the number of AraR mutants induced
per H2O2 dose (mM). This value was estimated as the slope of the linear
regression line fitted to the linear section of the corresponding dose-response
curve. The lethal potency of H2O2 was expressed as the dose giving 50%
survival (LD50).
DNA purification and enzymatic hydrolysis
Cells from stationary (10 ml) or exponential phase cultures (20 ml) were
pelleted, washed with 10 ml 50 mM Tris-HCI buffer (pH 8 0), centrifuged
again and resuspended in 0.5 ml of the same buffer containing 10% methanol.
Cell lysis and RNA digestion were carried out by 1 h incubation at 37°C with
lysozyme (2 mg, 40 |il) and RNase A (0 1 mg, 5 |il) Proteins were then
digested by incubation for 1 h at 50°C with I 6 mg proteinase K dissolved in
80 ul 10 mM Tris-HCI buffer (pH 7.4) in the presence of 80 nl 10% Tnton
X-100. DNA was consecutively extracted with 0.6 ml phenol, 1.2 ml
phenol-chloroform:isoamyl alcohol (25:24 1) and 0 6 ml chloroform:isoamyl
alcohol (24:1). The DNA was precipitated from the final aqueous phase by
adding 195 (imol sodium acetate buffer (pH 5.0, 65 u.1), 16 |ig glycogen
(10 |il) and 1 3 ml ice-cold ethanol. The mixture was kept at -20°C for
30 min, centnfuged at 10 000 g for 15 min at 4°C and the pellet washed three
times with 1 ml 70% ethanol The DNA (50-200 u\g) was dissolved in 100 |il
20 mM sodium acetate buffer (pH 4.8) containing 10% methanol. Hydrolysis
to nucleotides was carried out by digestion with nuclease PI (10 |ig, 3 (il)
for 1 h at 37°C. The pH of the mixture was adjusted with 10 |imol Tns-HCl
buffer (pH 7.5, 10 (il) and hydrolyzed to nucleosides with 2 U calf intestine
alkaline phosphatase (2 |il) for 2 h at 37°C. The deoxynucleoside mixture
was cleared by centrifugation in Microcon-10 units (Amicon Inc.. Beverly,
MA, USA) at 14 000 g for 15 mm at 4°C.
HPLC separation and 8-oxodG quantification
Deoxynucleosides were analyzed in a liquid chromatograph fitted with solvent
delivery module 126. injection valve 210A. a 20 |ll injection loop and analog
interface 406 (all from Beckman, San Ram6n, CA) The system was controlled
with a Deskpro 386/2Oe computer (Compaq, Houston. TX) using System
Gold 6.0 software (Beckman). Deoxyguanosine was quantified at 290 nm in
a diode array detector (model 168: Beckman) 8-oxodG was quantified in a
Coulochem EC detector (ESA, Bedford, MA), fined with a guard cell 5020
and an analytical cell 5011, connected after the UV detector. Separation was
achieved in a Spherisorb ODS2 column (250X4 mm, 3 urn particles) fitted
with a C-18 pre-column (Tracer). Isocrarjc elunon was performed with 50 mM
potassium phosphate buffer (pH 5.5):methanol (928) as the mobile phase at
room temperature and a 0 75 ml/min flow rate. The potential settings of the
EC detector were- guard cell. +0.5 V; detector 1, +0.1 V; detector 2, +0.4 V
Results are expressed as number of 8-oxodG/105 dG residues, based on
calibration curves obtained with dG and authentic 8-oxodG.
Results
Figure 1 compares the sensitivity to both lethality and muta-
genesis induced by H2O2 in the wild-type parent strain grown
590

Page 3

DNA 8-oxodG content and antloridant defences in E.coli
either in NB or K medium. Growth in K medium resulted in
a 5-fold higher sensitivity to H2O2, as shown by the values of
lethal and mutagenic potencies. The increased sensitivity of
wild-type bacteria to the lethal and mutagenic actions of H2O2
in K medium as compared with NB prompted us to evaluate
the occurrence of oxidative damage in chromosomal DNA.
The 8-oxodG content was assessed as an index of oxidative
DNA damage (Floyd et al., 1986).
Several methods for DNA isolation were compared, namely
that adapted for gram-negative bacteria (Chen and Kuo, 1993),
the Genomics (Bessho et al, 1992) or ASAP kits and that
described in Materials and methods. In our hands, the first
three methods produced lower DNA yields, more complex
chromatograms and/or higher amounts of 8-oxodG than
standard phenol extraction (data not shown). In contrast to
previous reports (Claycamp, 1992), our results agree with the
idea that the phenol method does not artificially increase the
levels of 8-oxodG extracted from DNA if carefully used
(Harris et al., 1994).
Conditions of DNA hydrolysis were also optimized. Hydro-
lysis to deoxynucleotides was completed after 20 min incuba-
tion at 37°C with 10 |ig nuclease PI (results not shown);
nevertheless, we used a 1 h incubation in all subsequent
analyses. Figure 2 shows the time course of hydrolysis to
a b
8 c
38
100
80
60
40
20
0
n-
(.
u
0 30 60
Time (min)
90 120 150 180
Fig. 2. Deoxynucleoside release from DNA hydrolyzates by alkaline
phosphatase. Salmon testis DNA (500 ng in 100 u.1), previously treated with
nuclease PI, was cleaved to deoxynucleosides by incubation at 37°C with
2 U AP (see Materials and methods). At the times shown, the contents of
dG (•) and 8-oxodG (O) were analyzed by HPLC of aliquots taken from
the incubation mixture. The results are the means of two independent
hydrolyses and are expressed as percentages of the maximum values of
deoxynucleosides released.
deoxynucleosides by 2 U alkaline phosphatase (AP). Initial
deoxynucleoside release was very fast, the rate slowing with
time as nucleotides were depleted. 8-oxodG was released more
slowly than dG, suggesting that AP preferred the normal
nucleotide for hydrolysis. Nevertheless, 8-oxodG was fully
released after 2 h and the level did not increase at longer times
of incubation. The same level of 8-oxodG was observed using
up to 16 U AP. In consequence, 2 U AP and 2 h incubation
were routinely used for subsequent analyses. Production of
8-oxodG at a constant rate by E.coli AP treatment has been
previously reported; subtraction of such a value has been
recommended to calculate the 8-oxodG content in DNA (Kasai
et al., 1986; Tajiri et al., 1995). Our data have not been
corrected, since the kinetics of 8-oxodG release by the eukary-
otic AP we used (Figure 2) do not indicate a constant rate of
production of 8-oxodG, in contrast to the artifactual oxidization
of guanine reported by Tajiri et al. (1995) when using a 26-
fold higher amount of E.coli AP. Though we cannot exclude
some oxidative damage during DNA preparation, the lowest
8-oxodG contents determined in this study were in the back-
ground range (0.5-2.0 8-oxodG residues/105 dG) quoted for
different cells and organisms (Kasai et al., 1986; Floyd, 1990).
The influence of different growth media and of different
growth phases on oxidative damage to DNA was studied
(Table I). Stationary wild-type cells grown in K medium
showed a 2.3-fold higher 8-oxodG content in genomic DNA
than those grown in NB. A similar difference was found when
wild-type cells growing exponentially in NB medium were
compared with those grown to stationary phase. Exponentially
growing wild-type bacteria displayed 8.6% (3.92 versus
45.82 U/mg) of total catalase activity (HPI plus HPII) and
20.7% (10.3 versus 49.8 mU/mg) of HPI peroxidase activity
of stationary cultures, in agreement with previous data (Visick
and Clarke, 1997). Therefore, changes in catalase activity
could explain, at least in part, the 8-oxodG differences in wild-
type cells grown under different conditions. Actually, in
contrast to the wild-type, non-statistically significant variations
in 8-oxodG content were observed in the katG katE double
null mutant (Table I).
The levels of total catalase and SOD activities were com-
pared in wild-type, katG katE and sodA sodB bacteria grown
to stationary phase in both NB and K medium (Table II). As
expected, null mutations of both the katG and katE catalase
genes eliminated all catalase activity. The SOD activity
remaining in the sodA sodB strain, with both structural genes
Table I. Effect of growth conditions on the 8-oxo-deoxyguanine content of chromosomal DNA of different E.coli strains
Medium
Nutrient broth
Stationary
Exponential
K medium
Stationary
Wild-type OJC575)
Viable cells/ml
(X109)
3.3
0.8
2.5
8-oxodG/lO5
1.0 ± 05 (n
25 ± 0.6 (n
2.4 ± 0.5 (n
dGb
= 5)
= 3)
= 6)
0.01
0.001
fc3(Gta/£OJC1058)
Viable cells/ml
(X109)
3.0
05
25
8-oxodG/105 dGb
1.2 ± 1.3 (n = 6)
2.3 ± 05 in = 3)
2.6 ± 1.0 (n = 3)
[X
0.2411'
0.06^
"Stationary cultures were obtained after 14 fNB) or 16 h (K medium) growth. Exponential phase cultures were at ODjgg mn 0.6-0.7. Bacteria were grown at
37°C with gentle shaking (90 r.p.m.).
b8-oxodG/105dG measurements are the mean value t SDofn DNA extractions.
'Statistical significances (Student's r-test) are from comparisons with the values of stationary cultures in NB medium. Non-statistically significant comparisons
are indicated by cr:
591

Page 4

J.AIhama et al
Table II. Enzymatic activities of different E.coli
phase in different culture media1
Strain
Wild-type (UC575)
katG katE (UO058)
sodA sodB (UCI084)
Nutrient broth
Catalase
36.6 ± 5.9
0
46.4 ± 2.0
SOD
55.6 ±
53.7 ±
0.4 ±
4.7
6.9
04
strains grown
K medium
Catalase
5.6 ± 0 2
0
6 2 ± 0.03
to stationary
SOD
6.8 ±
7.6 ±
1 1 ±
02
07
0.4
"Data are units enzymatic activity/mg protein, as defined in Materials and
methods, they represent the mean ± SD of two to four bacterial cultures.
Table III. Chromosomal DNA 8-oxo-deoxyguanine content of different
E.coli strains grown to stationary phase in NB
Strain
Mutants/
107 viable cells"
DNA
extractions
8-oxodG/
105 dG
Wild-type
(UC575)
katG katE
(UC1058)
sodA sodB
(UC1084)
34 ± 9 (n = 5) 15
91 ± 27 (n = 5) 12
155 ± 56 (n = 5) 15
0.7 ± 0.3
0 8 ± 0.4 0.43b
2.1 ± 1.1 <0.000lb,
0.00!c
"Spontaneous mutauon rate to L-arabinose resistance, expressed as the mean
value of mutants/106 viable cells ± SD of n cultures.
bSlatistical significance (Student's r-lest) for comparison with wild-type.
'Statistical significance (Student's (-test) for comparison with katG katE.
insertionally inactivated, is attributed to sodC-encoded peri-
plasmic CuZnSOD (Benov and Fridovich, 1994). This activity
accounted for =£2% of total SOD activity in wild-type extracts
(0.43-1.06 versus 55.61 U/mg), in agreement with previous
reports (Benov and Fridovich, 1994; Imlay and Imlay, 1996).
To our knowledge, no sodC mutant strain is yet available,
since CuZnSOD has only recently been discovered in E.coli
(Benov and Fridovich, 1994). Wild-type bacteria grown in K
medium retained =£15% of total catalase and SOD activities
of cells grown in NB medium. Similar reductions in SOD and
catalase activities were observed in the katG katE and sodA
sodB mutant strains respectively.
At conditions of maximal expression of both catalase and
SOD activities (stationary cultures in NB), the 8-oxodG content
of chromosomal DNA of wild-type bacteria was compared
with those of catalase- or cytosolic SOD-deficient bacteria
(Table III). While the katG katE double mutant showed wild-
type levels of 8-oxodG, the sodA sodB mutant derivative
displayed a 2.8-fold increase. In addition to increased DNA
oxidative damage, sodA sodB mutant cells showed a higher
rate of spontaneous mutagenesis than katG katE mutant bac-
teria. When grown under non-aerated conditions, neither the
katG katE nor the sodA sodB mutant showed increased
chromosomal 8-oxodG content, as compared with the wild-
type (Figure 3A). In contrast, when grown under hyperoxygen-
ation, the 8-oxodG content of the katG katE mutant was double
that of the wild-type (Figure 3B).
The relative importance of catalase and Fpg glycosylase in
protecting E.coli from oxidative damage by H2O2 was evaluated
by comparing strains lacking either catalase (katG katE mutant)
or Fpg (fpg mutant) with the isogenic parent with wild-type
levels of both enzymes (Figure 4). As previously reported
(Prieto-Alamo et al.. 1993), cataJase deficiency resulted in a
dramatic decrease in viability after H2O2 treatment; H2O2 was
• 5
wt kitC katE sodA sodB
Fig. 3. Effect of growth without aeration (A) or with hyperoxygenauon (B)
on the 8-oxodG content of chromosomal DNA in different E.coli strains
Data are the means ± SD of 8-oxodG/10s dG residues of n DNA
extractions. To minimize air exposure, in (A) bacteria were inoculated into
250 ml NB contained in a 250 ml screw cap bottle and grown at 37°C for 5
days Periodically, the cultures were gently mixed by inverting the container.
Saturated cultures had the following numbers of viable cells (X lO'Vml'
wild-type, 0.44; katG katE, 0.36; sodA sodB, 0.21. In (B), bacteria were
grown to stationary phase under continuous bubbling of pure oxygen
(Prieto-Alamo et al., 1993) P values indicate the statistical significance
(Student's (-test) for comparison with the wild-type. Bacterial strains were
UC575 (wild-type), UC1058 (katG katE mutant) and UC588 (sodA sodB
mutant).
g
ival
•urvi
100
10
1
0.1
0.01
o.orn
r
r Y
\ \ .
- katG katE
10
20
H2O2 (mM)
0
20
Fig. 4. DNA content of 8-oxodG and lethal sensitivities of different E.coli
strains after H2O2 treatment. Strains UC575 (A, wild-type), UC1057 (•,
fpg) and UC1058 (O. katG katE ) were grown in K medium to stationary
phase. H2O2 treatments were as described in Materials and methods for
mutagenesis and survival assays Data show, as a function of the H2O2
dose, the survival percentages (left) and the 8-oxodG content (right) after
subtracting the values for untreated wild-type, fpg and katG katE strains,
1.7, 3.7 and 1.8 residues 8-oxodG/lO5 dG respectively.
highly cytotoxic in the katG katE double mutant at doses non-
lethal for the other two strains. Bacteria containing the fpg
mutant allele were practically as resistant as the wild-type
below 10 mM H2O2, while at higher peroxide doses the fpg
mutant showed decreased survival. The 8-oxodG levels in
DNA from both catalase- and Fpg-deficient strains increased
with H2O2 dose (Figure 4). Interestingly, despite their differ-
ences in sensitivity to lethality by H2O2, formation of 8-oxodG
in genomic DNA upon H2O2 treatment was roughly the same
for the katG katE and fpg mutant strains (0.44 and 0.34
residues/105 dG/mM H2O2 respectively). Nevertheless, fpg
bacteria showed an increased 8-oxodG background level (3.7
residues/105 dG) as compared with the other two strains (1.7
and 1.8 residues/105 dG for wild-type and katG katE bacteria
respectively). Under the experimental conditions of Figure 4
592

Page 5

DNA 8-oxodG content and antioxidant defences in E.coli
the wild-type strain was totally resistant to both lethality and
DNA oxidative damage by H2O2.
Discussion
Here we demonstrate that both the lethal and the mutagenic
responses of E.coli to H2O2 challenges are markedly increased
in bacteria grown in K medium (1% casamino acids and 1%
glucose), as compared with those grown in NB. This result is
in agreement with the observation that specific components of
the growth medium can modulate the lethal response of both
E.coli and cultured mammalian cells to H2O2 (Brandi et al.,
1992). Of particular relevance to our work is the bacterial
sensitization produced by certain amino acids, like L-histidine
and L-cysteine, particularly in the presence of glucose (Brandi
et al, 1992; Muller and Janz, 1993). Therefore, it seems likely
that the glucose and at least part of the casamino acids might
cause the stronger responses to H2O2 of bacteria grown in
K medium.
Growth in K medium and also exponential phase growth
sensitized wild-type E.coli to DNA oxidative damage, as
indicated by the increased number of 8-oxodG/lO5 dG residues
observed under such conditions. The increased sensitivity of
wild-type cells grown in K medium to lethality and mutagenesis
by H2O2 and to DNA oxidative damage was further associated
with a marked decrease in both total catalase and SOD
activities (=£15% of those grown in NB). Previous studies
have correlated increments in both HPI and HPII catalase
activities with extracellular accumulation of products of glu-
cose catabolism, specifically
(Schellhorn and Stones, 1992; Mukhopadhyay and Schellhorn,
1994). Acetate utilization would result in increased respiratory
activity and, hence, would cause an increase in the intracellular
flux of toxic oxygen intermediates, including H2O2. In contrast,
both catalase and SOD synthesis are repressed, in a cAMP-
independent mode, when glucose is used as the main carbon
source (Hassan and Fridovich, 1977; Meir and Yagil, 1990).
Bacteria lacking both cytosolic SODs (sodA sodB mutant)
displayed an increased 8-oxodG content in genomic DNA
(estimated as 2.8-fold that of the wild-type) when grown under
standard aerated conditions in NB medium. This result is in
agreement with the reported increment in the steady-state
concentration of O2"~ detected in SOD-deficient bacteria: a
[O2'~]ss of 2.4X10"11 M was estimated in aerobic log phase
E.coli containing wild-type levels of SOD and 3X1(T10 M in
a sodA sodB mutant (Gardner and Fridovich, 1992). Therefore,
SOD, in coping with O2*~ generated during aerobic growth,
protects chromosomal DNA against oxidative damage. In fact,
no significant increment in 8-oxodG content was observed in
bacteria grown under non-aereated conditions.
As previously reported (Prieto-Alamo et al, 1993) and in
agreement with the mutagenic potential of 8-oxoG (Michaels
and Miller, 1992), the sodA sodB double mutant exhibited, in
parallel with an increased level of this oxidized base in
chromosomal DNA, an increased spontaneous frequency of
forward mutations to L-arabinose resistance. In a previous
study, Tajiri et al. (1995) related the mutator phenotype of
bacteria simultaneously lacking Fpg glycosylase (removing
8-oxoG from DNA) and MutT protein (removing 8-oxoGTP
from the nucleotide pool) with the 8-oxodG increment quanti-
fied in chromosomal DNA (estimated as 2.5-fold that of the
wild-type). In addition to 8-oxoG, other oxidative lesions, such
weak acids like acetate
as 8-oxoadenine, 2-oxoadenine, 5-hydroxycytidine, etc., are
known to be mutagenic (Epe, 1995) and could also contribute
to the increased mutagenesis in bacteria lacking antioxidant
enzymes. Nevertheless, the existing data on the mutagenic
potential of the various oxidative lesions cannot be easily
compared, giving only a rough estimate of their respective
mutagenic risks (Epe, 1995).
Though under standard aerated conditions of growth the
8-oxodG content of chromosomal DNA of bacteria totally
devoid of catalase (KatG katE mutant) was not significantly
different from that of wild-type bacteria, differences were
readily appreciated under oxidative stress. Thus, the protective
role of catalase in preventing 8-oxodG formation in genomic
DNA became evident after growing the cells under hyperoxy-
genation and particularly after exposure to H2O2.
Different modes of killing by H2O2 have been described in
E.coli (Imlay and Linn, 1986); that requiring active metabolism
is attributed mainly to single-strand breaks, while that which
does not require active growth is due to uncharacterized cell
damage. In contrast to previous data (Boiteux and Huisman,
1989), we found that the E.coli fpg mutant displays increased
sensitivity to cytotoxicity by H2O2 as compared with the wild-
type, suggesting that formamidopyrimidines, the presumptive
lethal lesions repaired by Fpg, are formed to a significant extent
after H2O2 treatment, contributing to its lethality (Boiteux
et al, 1992). Catalase, in removing H2O2, should prevent
cytotoxicity resulting from any lethal damage caused by
exposure to this peroxide, in accord with the fact that bacteria
lacking catalase (katG katE mutant) exhibited much higher
sensitization to lethality by H2O2 than those lacking Fpg
protein. However, the scavenger enzyme catalase and the
oxidative repair enzyme Fpg protected to similar extents
against the premutagenic lesion 8-oxo-deoxyguanine formed
in vivo upon H2O2 exposure.
Acknowledgements
This work was subsidized by grants PB95-0557-CO2-01 and PB95-0557-
CO2-02 from die DGES, Ministerio de Educaci<5n y Cultura, and by the Junta
de Andalucfa (CVI 0187 and CVI 0151). J.A. was a re-incorporation
postdoctoral fellow of the Ministerio de Educaci6n y Cultura (Spain) and
J.R.-L. was the recipient of a fellowship from the Junta de Andalucfa.
References
Abril.N. and Pueyo.C. (1990) Mutagenesis in E. coli lacking catalase. Environ.
Mol. Mutagen., 15, 184-189.
Abril.N., Roldan-Arjona,T., Prieto-Alamo.MJ., van Zeeland,A.A. and
Pucyo.C. (1992) Mutagenesis and DNA repair for alkylation damage in
Escherichia coli K-12. Environ. Mol. Mutagen., 19, 288-2%.
Beers.R.FJr and Sizer.I.W. (1952) A spectrophotometric method for measuring
the breakdown of hydrogen peroxide by catalase. /. Biol. Chem., 195,
133-139.
Benov.LT. and Fridovich,1. (1994) Escherichia coli expresses a copper- and
zinc-containing superoxide dismutase. /. Biol. Chem., 269, 25310-25314.
Bessho.T., Tano.K., Kasai.H. and Nishimura,S. (1992) Deficiency of 8-
hydroxyguanine DNA endonuclease activity and accumulation of the 8-
hydroxyguanine in mutator mutant (MutM) of Escherichia coli Biochem.
Biophys. Res. Commxin., 188, 372-378.
Boiteux.S. and Huisman.O. O989) Isolation of a formamidopyrimidine-DNA
glycosylase (fpg) mutant of Escherichia coli. Mol. Gen. Genet., 215,
300-305.
Boiteux.S., Gajewsky.E., LavalJ. and Dizdaroglu.M. (1992) Substrate
specificity of the Escherichia coli Fpg protein (formamidopyrimidine-DNA
glycosylase): excision of purine lesions in DNA produced by ionizing
radiation or photosenMtization. Biochemistry; 31, 106-110.
Bradford,M.M. (1976J A rapid and sensitive method for the quantification of
microgram quantities of protein utilizing the principle of protein-dye
binding. Anal Biochem., 72, 248-254.
593

Page 6

J.AIhama et al
Brandi,G , Marrot.L , Giacomoni.PU , Sestili.R, Hussain.S , Cartabeni,F. and
Cantoni.O (1992) The role of extracellular medium components and specific
amino acids in the cytotoxic response of Escherichia coh and Chinese
hamster ovary cells to hydrogen peroxide. Free Radical Res Commun., 16,
41^9
Chen.W. and Kuo.T (1993) A simple and rapid method for the preparation
of gram-negative bacterial genomic DNA. Nucleic Acids Res.. 21. 2260.
Chung.M.H., Kasai.H., Jones.D.S , Inoue.H , Ishikawa,H , Ohtsuka,E and
Nishimura,S (1991) An endonuclease activity of Eschenchia coli that
specifically removes 8-hydroxyguanine residues from DNA. Mutat Res.,
254, 1-12
Claycamp.H.G (1992) Phenol sensitization of DNA to subsequent oxidative
damage in 8-hydroxyguanine assays. Carcmogenesis, 13, 1289-1292
Dizdaroglu.M., Rao,G., Halliwell.B. and Gajewski.E (1991) Damage to the
DNA bases in mammalian chromatin by hydrogen peroxide in the presence
of feme and cupric ions. Arch. Biochem. Biophys., 285, 317-324.
Epe.B. (1995) DNA damage profiles induced by oxidizing agents. Rev Physiol.
Biochem. Pharmacol., 127, 223-249.
Floyd,R A (1990) The role of 8-hydroxyguanine in carcmogenesis.
Carcinogenesis, 11, 1447-1450.
Floyd.R.A, WatsonJJ., Wong.P.K , Altmiller.D.H. and Rickard.R C (1986)
Hydroxyl free radical adduct of deoxyguanosine: sensitive detection and
mechanisms of formation Free Radical Res. Commun., 1, 163—172.
Gardner.P.R. and Fndovich.I. (1992) Inaclivauon-reactivation of aconitase in
Escherichia coir a sensitive measure of superoxide J. Biol. Chem , 267,
8757-8763.
Gerhardt.P, Murray,R.G.E., Wood.W.A and Kneg,N.R. (1994) Methods for
General and Molecular Bacteriology. American Society for Microbiology,
Washington, DC.
Halhwell.B. (1993) Oxidative DNA damage: meaning and measurement. In
Halhwell.B. and Aruoma.O I (eds), DNA and Free Radicals Ellis Horwood,
New York, NY
Hams.G., Bashir.S and Winyard.P.G.
deoxyguanosine present in DNA is not simply an artefact of isolation.
Carcmogenesis, 15, 411 —413-
Hassan,H.M. and Fridovich.l (1977) Regulation of superoxide dismutase
synthesis in Escherichia coli glucose effect J. Baclenol., 132, 505-510.
Imlay.K.R C and Imlay.J.A (1996) Cloning and analysis of sodC, encoding
the copper-zinc superoxide dismutase of Escherichia coh. J. Bacteriol..
178, 2564-2571.
ImlayJ.A. and Linn.S (1986) Bimodal pattern of killing of DNA-repair-
defective or anoxically grown Escherichia coh by hydrogen peroxide
J Bacteriol., 166, 519-527
Ivanova,A., Miller.C, Glinsky.G and Eisenstark.A (1994) Role of rpoS
(katF) in oxy>?-independent regulation of hydroperoxidase I in Escherichia
coh. Mol Microbiol. 12, 571-578
Jamieson,D.J. and Storz.G. (1997) Transcriptional regulators of oxidative
stress response. In ScandaliosJ G (ed.). Oxidative Stress and the Molecular
Biology of Antioxidant Defenses. Cold Spring Harbor Laboratory Press.
Cold Spring Harbor, NY
Kasai.H.. Crain.P.F., Kuchino.Y, Nishimura,S., Ootsuyama,A. and Tanooka,H
(1986) Formation of 8-hydroxyguanine moiety in cellular DNA by agents
producing oxygen radicals and evidence for its repair. Carcinogenesis, 7,
1849-1851
Keyer,K., Gort.A.S. and ImlayJ. (1995) Superoxide and the production of
oxidative DNA damage. J Bacteriol, 177, 6782-6790.
Loewen.PC. (1984) Isolation of catalase-deficient Escherichia coh mutants
and genetic mapping of katE. a locus that affects catalase activity /.
Bactenol, 157. 622-626.
Loewen.PC, Triggs.B L . George.C S and Hrabarchuk.B E (1985) Genetic
mapping of katG, a locus that affects synthesis of the bifunctional catalase-
peroxidase hydroperoxidase I in Eschenchia coli. J Bactenol, 162.
661-667.
McCordJ.M. and Fndovich.I (1969) Superoxide dismutase. an enzymic
function for erythrocuprein. J Biol Chem.. 244. 6049-6055
Meir.E. and Yagil.E. (1990) Regulation of Escherichia coh catalase by
anaerobiosis and catabolite repression Curr Microbiol.. 20. 139-143.
Michaels,MX. and MillerJ.H. (1992) The GO system protects organisms
from the mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-
dihydro-8-oxoguanine). J. Bactenol, 174, 6321-6325.
MillerJ.H. (1992) A Short Course in Bactenal Genetics A Laboratory Manual
and a Handbook for Escherichia coh and Related Bactena Cold Spring
Harbor Laboratory Press. Cold Spring Harbor. NY
Mukhopadhyay.S. and Schellhom.H. (1994) Induction of Eschenchia coh
hydroperoxidase I by acetate and other weak acids J Bactenol. 176.
2300-2307.
MUllerJ. and Janz.S. 11993) Modulation of the H:O2-induced SOS response
(1994) 7,8-Dihydro-8-oxo-2'-
in Escherichia coli PQ300 by amino acids, metal, chelators, antioxidants,
and scavengers of reactive oxygen species. Environ. Mol. Mutagen., 22,
157-163.
Prieto-Alamo.M.J., Abnl,N. and Pueyo.C. (1993) Mutagenesis in Eschenchia
coli K-12 mutants defective in superoxide dismutase or catalase
Carcinogenesis, 14, 237-244.
SaJc.B.D., Eisenstark.A and Touati.D. (1989) Exonuclease III and the catalase
hydroperoxidase II in Eschenchia coh are both regulated by the katF gene
product Proc. Nail Acad Sci. USA, 86, 3271-3275.
Schellhom.H E. and Stones.VL (1992) Regulation of katF and katE in
Escherichia coli K-12 by weak acids J Bacteriol., 174, 4769^4776.
Storz.G , Tartaglia.L A and Ames.B N (1990) TranscriptionaJ regulator of
oxidative stress-mducible genes: direct activation by oxidation. Science,
248, 189-194.
Tajin.T, Maki.H and Sekiguchi.M. (1995) Functional cooperation of MutT,
MutM and MutY proteins in preventing mutations caused by spontaneous
oxidation of guanine nucleotide in Escherichia coh. Mutat Res., 336,
257-267.
TchouJ., Kasai.H., Shibutani.S., Chung.M H., LavalJ., Grollman.A.P. and
Nishimura,S. (1991) 8-Oxoguanine (8-hydroxyguanme) DNA glycosylase
and its substrate specificity. Proc. Nail Acad Sci. USA, 88, 4690-4694.
Touati.D. (1997) Superoxide dismutases in bactena and pathogen protists. In
ScandaliosJ.G. (ed.), Oxidative Stress and the Molecular Biology of
Antioxidant Defenses. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY
VisicM.E. and Clarke.S. (1997) RpoS- and OxyR-independent induction of
HPI catalase at stationary phase in Escherichia coh and identification of
rpoS mutations in common laboratory strains. J. Baclenol, 179,4158^4163.
Wood.M.L.. Dizdaroglu.M., Gajewki.E
Mechanistic studies of ionizing radiation and oxidative mutagenesis: genetic
effects of a single 8-hydroxyguanine (7-hydro-8-oxoguanine) residue
inserted at a unique site in a viral genome Biochemistry, 29, 7024—7032
and Essigmann.J.M (1990)
Received on January 5. 1998, accepted on February 20, 1998
594