Ionic strength and magnesium affect the specificity of
Escherichia coli and human 8-oxoguanine-DNA
glycosylases
Viktoriya S. Sidorenko
1
, Grigory V. Mechetin
1
, Georgy A. Nevinsky
1,2
and Dmitry O. Zharkov
1,2
1 SB RAS Institute of Chemical Biology and Fundamental Medicine, Novosibirsk, Russia
2 Department of Natural Sciences, Novosibirsk State University, Russia
In all living organisms DNA is subject to ongoing
damage by various environmental and endogenous
factors [1]. One of the most frequently encountered
base lesions is 8-oxo-7,8-dihydroguanine (8-oxoG),
produced by oxidative stress to the steady-state level
of $ 1 · 10
6
guanines in human DNA [2]. 8-oxoG is
mutagenic due to its ability to form a stable Hoog-
sten pair with A [3] and its propensity to direct the
incorporation of dAMP by DNA polymerases [4]. If
left uncorrected, the resulting 8-oxoG:A mispair is
converted to a T:A pair in the next round of replica-
tion, producing a G:C fi T:A transversion mutation,
the type frequently encountered in human cancers
[5,6].
The consequences of 8-oxoG’s appearance in DNA
are counteracted by a three-tier enzymatic ‘GO system’

and caffeine) on the activity and opposite-base specificity of Escherichia coli
Fpg and human OGG1. The activity of both enzymes towards 8-oxoG:A
decreased sharply with increasing salt and Mg
2+
concentration, whereas
the activity on 8-oxoG:C was much more stable, resulting in higher oppo-
site-base specificity when salt and Mg
2+
were at near-physiological concen-
trations. This tendency was observed with both Cl
)
and glutamate as the
major anions in the reaction mixture. Kinetic and binding parameters for
the processing of 8-oxoG:C and 8-oxoG:A by Fpg and OGG1 were deter-
mined under several different conditions. Polyamines, crowding agents,
biotin and caffeine affected the activity and specificity of Fpg or OGG1
only marginally. We conclude that, in the intracellular environment, the
specificity of Fpg and OGG1 for 8-oxoG:C versus 8-oxoG:A is mostly due
to high ionic strength and Mg
2+
.
Abbreviations
8-oxoG, 8-oxo-7,8-dihydroguanine; AP, apurinic ⁄ apyrimidinic; KGlu, potassium glutamate; THF, tetrahydrofuran.
FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS 3747
repair to restore the original G:C pair. Importantly,
both Fpg and OGG1 are much less likely to excise
8-oxoG from 8-oxoG:A substrates, because if these
mispairs are generated by the incorporation of dAMP
opposite 8-oxoG, such excision would immediately fix
the G:C fi T:A transversion. Instead, 8-oxoG:A

include a buffer (often non-physiological, such as Tris
or Good buffers), a salt (usually NaCl or KCl) and
stabilizing agents (usually a metal chelator, a thiol
reagent and glycerol). In living cells, the reactions
catalyzed by DNA-dependent proteins may be
affected by ionic strength, the concentration of diva-
lent cations such as Mg
2+
, the nature of the buffer-
ing agents, the presence of competing polyamines and
other small molecules, and crowding by other macro-
molecules. Their effects on the specificity of 8-oxoG
excision have never been studied. Because these fac-
tors may be important for the efficiency of correct
8-oxoG repair, in this study we address how the rela-
tive efficiency of 8-oxoG excision from pairs with C
and A by Fpg and OGG1 depends on buffer compo-
sition, ionic strength, Mg
2+
concentration and several
other factors.
Results
Effects of ionic strength and divalent cations on
the activity and specificity of Fpg and OGG1
The conditions inside a living cell differ from most
buffer systems in which the activity and specificity of
Fpg and OGG1 have been studied. For example, in
Escherichia coli the intracellular concentrations of
Na
+

cannot be separated kinetically [23,28]; therefore, unas-
sisted cleavage of substrate DNA by this enzyme was
used as the assay endpoint. The AP lyase activity of
OGG1 proceeds via b-elimination and is much less effi-
cient than its glycosylase activity [25,26]. Two assay
endpoints were used in this case: glycosylase activity
was measured after full thermal degradation of the AP
site left by base excision, whereas the AP lyase activity
was assayed as unassisted cleavage of the substrate by
Table 1. Outline of the factorial design activity experiments.
Series
Varied reaction mixture
components Concentrations
1 KCl 0, 50, 100, 150, 200 m
M
MgCl
2
0, 5, 10, 15, 20 mM
2 KGlu 0, 50, 100, 150, 200 mM
MgCl
2
0, 5, 10, 15, 20 mM
KP
i
0, 25 mM
3 Spermine or spermidine 0, 1, 10, 100, 1000 lM
MgCl
2
0, 5, 10, 15, 20 mM
4 Poly(ethylene glycol)

and G); OGG1 was inhibited only by the highest con-
centrations of KCl and MgCl
2
. Interestingly, the AP
lyase activity of OGG1 on 8-oxoG:A at any given
0
10
20
30
40
50
0
50
100
150
200
0
5
10
15
20
[
P
]
,
n
M
K
C
l

]
,
n
M
K
C
l
,
m
M
M
g
C
l
2
,
m
M
0
10
20
30
40
50
60
0
50
100
150
200

150
200
0
5
10
15
20
[
P
]
,
n
M
KC
l
,
m
M
M
g
C
l
2
,
m
M
0
5
10
15

M
0
5
10
15
20
25
30
35
0
50
100
150
200
0
5
10
15
20
C
/
A
K
C
l
,
m
M
M
g

l
,
m
M
M
g
C
l
2
,
m
M
0
2
4
6
8
10
12
0
50
100
150
200
0
5
10
15
20
[

200
0
5
10
15
20
C
/
A
K
C
l
,
m
M
M
g
C
l
2
,
m
M
AB C
DE F
GH I
Fig. 1. Activity and specificity of Fpg and OGG1 in buffers with different concentrations of KCl and MgCl
2
. (A–C) Fpg, (D–F) glycosylase
activity of OGG1, (G–I) AP lyase activity of OGG1. The extent of cleavage of 8-oxoG:C (A, D, G) or 8-oxoG:A (B, E, H) or the C ⁄ A specificity

levels, making the determination of individual kinetic
constants problematic. The C ⁄ A preference measured
in the factorial design experiments under these con-
ditions was 5.2 for 0 mm MgCl
2
and 16 for 10 mm
MgCl
2
. The results are summarized in Table 2. In the
absence of Mg
2+
, the kinetic constants were in agree-
ment with those reported in the literature [23], with
8-oxoG:C being a better substrate because of its lower
K
M
value. The effect of Mg
2+
on K
M
was not high
($ 1.5-fold); however, in the presence of Mg
2+
, K
M
improved for 8-oxoG:C and worsened for 8-oxoG:A.
By contrast, Mg
2+
reduced k
cat

À1
ES !
k
2
EP !
k
3
E þ P Scheme 1
To independently evaluate the effects of ionic
strength and Mg
2+
on the activity and specificity of
OGG1, we measured the apparent values of k
2
and k
3
under conditions of low salt (KP
i
only) and no Mg
2+
,
low salt and 20 mm Mg
2+
, and high salt
(KP
i
+ 150 mm KCl) and no Mg
2+
. These conditions
were selected to represent regions of the factorial

MgCl
2
or KCl. However, an increase in the ionic
strength of Mg
2+
had a pronounced deleterious effect
on k
2
and k
3
for 8-oxoG:A substrates: 20 mm Mg
2+
decreased k
2
by 42-fold and k
3
by 5.5-fold, whereas
150 mm KCl decreased k
2
by 3.9-fold and k
3
by 6.9-
fold. Therefore, physiological concentrations of ionic
strength and divalent cations enhance both base exci-
sion and the turnover of OGG1 cleaving its proper
substrate, 8-oxoG:C, and prevent cleavage of the
improper 8-oxoG:A substrate.
A well-recognized mechanism by which ionic strength
and divalent cations could modulate the activity of
DNA-dependent enzymes is changes in the affinity of

)1
)
k
cat
⁄ K
M
(nM
)1
Æmin
)1
)
8-oxoG:C 0 26 ± 6 6.1 ± 1.2 0.24
10 17 ± 2 1.2 ± 0.2 0.068
8-oxoG:A 0 490 ± 150 3.3 ± 0.7 0.0066
10 700 ± 340 1.0 ± 0.4 0.0015
Factors affecting the specificity of Fpg and OGG1 V. S. Sidorenko et al.
3750 FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS
tetrahydrofuran (THF) moiety instead of 8-oxoG. THF
is a good ligand for Fpg and OGG1, with their affinity
for THF-containing DNA closely paralleling the affinity
for 8-oxoG-containing DNA [23,26], and these particu-
lar ligands have been successfully used to analyze
stopped-flow kinetics for both enzymes [19,20]. The
results of the fluorescence titration experiments are
summarized in Fig. 2. In the absence of MgCl
2
, the
affinity of Fpg for the THF:C ligand was 1.6-fold higher
than for the THF:A ligand. The presence of Mg
2+

k
2
(min
)1
) k
3
(min
)1
) k
2
(min
)1
) k
3
(min
)1
)
0m
M KCl
0m
M MgCl
2
0mM KGlu
0.50 ± 0.09 0.21 ± 0.11 0.97 ± 0.49 0.011 ± 0.008
0m
M KCl
20 m
M MgCl
2
0mM KGlu

1.5
2.0
CA CA CA CA CA CA
Kd, µM
0
2
4
6
8
1 2
3 4 5
6
A
B
Fig. 2. Binding of Fpg and OGG1 to uncleavable THF:C and THF:A ligands under different conditions. (A) A representative experiment show-
ing fluorescence titration of Fpg with a THF:C ligand in the presence of 0 m
M (black circles) or 10 mM (white circles) MgCl
2
. AU, arbitrary
units. (B) Dissociation constants for binding of Fpg (1, 2) and OGG1 (3–6) to THF:C and THF:A ligands (denoted C and A, and represented by
white and black circles, respectively) determined from the fluorescence titration data. The variable components of the buffers included:
25 m
M KP
i
and 50 mM KCl (1), 25 mM KP
i
,50mM KCl and 10 mM MgCl
2
(2), 25 mM KP
i

with KGlu as the sole salt and buffer; the Mg
2+
con-
centration was varied in the same way as in the KP
i
–
KCl experiments described above (Table 1, Series 2).
For Fpg, the substitution of KGlu for KCl did not
change the overall dependence of the enzyme activity if
KP
i
was present (Fig. 3A–C). The only notable differ-
ence was a higher activity towards 8-oxoG:C at high
Mg
2+
and salt concentrations compared with when
Cl
)
was the major anion (cf. Figs 1A and 3A). As a
consequence, the specificity of Fpg for 8-oxoG:C versus
8-oxoG:A was highest at 150–200 mm KGlu and 5–
20 mm MgCl
2
, conditions that may better resemble the
cellular environment. If KP
i
was absent (Fig. 4A–C),
Fpg had very low activity at 0 mm KGlu and 0 mm
MgCl
2

(0–10 mm), whereas the activity towards
8-oxoG:C was higher. The AP lyase reaction with
8-oxoG:A was efficient only at low Mg
2+
and no
KGlu, whereas with 8-oxoG:C it was in general
agreement with the salt dependence of the glycosylase
reaction; the C ⁄ A specificity was highest at low to
medium KGlu and medium MgCl
2
. In the absence of
KP
i
, the DNA glycosylase and AP lyase activity of
OGG1 towards 8-oxoG:C resembled its activity in the
presence of KP
i
(Fig. 4G,H). However, exclusion of
KP
i
significantly influenced both activities of OGG1
with 8-oxoG:A; a more or less efficient glycosylase
reaction was observed only at 0 mm KGlu and 0–
15 mm MgCl
2
, whereas the AP lyase reaction required
0–100 mm KGlu and 0–5 mm MgCl
2
. The C ⁄ A speci-
ficity of the glycosylase reaction in the absence of KP

8-oxoG:A in the presence of 25 mm KP
i
and 200 mm
KGlu (C ⁄ A specificity 9.0 for the glycosylase reaction,
15 for the AP lyase reaction). As shown in Table 3, in
comparison with KP
i
only, the addition of KGlu had a
minimal effect on either rate constant in the case of
8-oxoG:C (a 1.2-fold increase in k
2
and an 1.5-fold
decrease in k
3
), and even improved the k
2
value for
8-oxoG:A by 1.7-fold. However, this was accompanied
by a 16-fold decrease in k
3
, indicating that the enzyme
turnover on 8-oxoG:A slows significantly, contributing
to a decrease in the efficiency of its cleavage by OGG1.
Moreover, fluorescence titration analysis of OGG1
binding to uncleavable THF:C and THF:A damaged
ligands showed that although KGlu did not affect the
affinity of the enzyme for the THF:C ligand, its affinity
for the THF:A ligand decreased at least 2.3-fold in com-
parison with the reactions (C⁄ A specificity for binding
was 3.5 for 25 mm KP

mine or spermidine (Table 1, Series 3) and varying
concentrations of MgCl
2
. In the absence of Mg
2+
,
spermine slightly ($ 1.5-fold) increased the specificity
of Fpg due to a corresponding decrease in activity on
8-oxoG:A. No significant influence of polyamines on
OGG1 activity was observed, except that AP lyase
activity on 8-oxoG:C was approximately twofold
higher in 1 mm spermine (but not spermidine), possibly
due to the chemical degradation of AP sites by
polyamines [37]. Overall, polyamines had minimal
0
10
20
30
40
50
0
50
100
150
200
0
5
10
15
20

5
10
15
20
[
P
]
,
n
M
K
G
l
u
,
m
M
M
g
C
l
2
,
m
M
0
10
20
30
40

20
30
40
50
0
50
100
150
200
0
5
10
15
20
[
P
]
,
n
M
KG
l
u
,
m
M
M
g
C
l

u
,
m
M
M
g
C
l
2
,
m
M
0
2
4
6
8
10
0
50
100
150
200
0
5
10
15
20
C
/

10
15
20
[
P
]
,
n
M
K
G
l
u
,
m
M
M
g
C
l
2
,
m
M
0
2
4
6
8
0

5
10
15
0
50
100
150
200
0
5
10
15
20
C
/
A
K
G
l
u
,
m
M
M
g
C
l
2
,
m

and found only marginal differences for any of the
enzyme–substrate pairs [Fig. 5C shows an example of
DNA glycosylase activity of OGG1 with the range
of poly(ethylene glycol) 8000 and 0 mm MgCl
2
].
Therefore, macromolecular crowding is likely to be of
little importance for the function of these two
enzymes.
In addition, we analyzed the effect of two low
molecular mass compounds, biotin and caffeine, on
the activity of Fpg and OGG1. Biotin can be regarded
as a structural mimic of 8-oxopurines [41], and avidin,
0
10
20
30
40
50
60
0
50
100
150
200
0
5
10
15
20

150
200
0
5
10
15
20
[
P
]
,
n
M
K
G
l
u
,
m
M
M
g
C
l
2
,
m
M
0
10

m
M
0
10
20
30
40
50
0
50
100
150
200
0
5
10
15
20
[
P
]
,
n
M
K
G
l
u
,
m

K
G
l
u
,
m
M
M
g
C
l
2
,
m
M
0
5
10
15
20
25
30
35
0
50
100
150
200
0
5

200
0
5
10
15
20
[
P
]
,
n
M
K
G
l
u
,
m
M
M
g
C
l
2
,
m
M
0
2
4

l
2
,
m
M
0
10
20
30
40
50
60
0
50
100
150
200
0
5
10
15
20
C
/
A
K
G
l
u
,

through mechanisms that are not fully understood
[43]. We analyzed the ability of Fpg and OGG1 to
cleave their substrates in the presence of up to 20 mm
biotin or caffeine. However, no effect was found
except for a slight ($ 30%) inhibition of Fpg at the
highest caffeine concentration used (data not shown).
Therefore, biotin and caffeine are unlikely to influence
the activities of these enzymes in vivo.
Discussion
The substrate specificity of DNA glycosylases has been
subject to a number of studies, yet the results are often
conflicting. For example, Fpg has been reported to
excise more than 20 different damaged bases from oli-
gonucleotide substrates [44], whereas the excision from
damaged genomic DNA has been reported only for
8-oxoG, 4,6-diamino-5-formamidopyrimidine, 2,4-dia-
mino-6-oxo-5-formamidopyrimidine and 2,4-diamino-
6-oxo-5N-methyl-5-formamidopyrimidine [45,46]. The
relative activity of Fpg on substrates containing
different bases opposite 8-oxoG also seemingly varies
depending on the assay used [19,23]. It is clear that
when an enzyme can process several substrates with
comparable efficiencies, as is the case for almost all
DNA glycosylases [47], the preferences for each sub-
strate may depend on the reaction conditions to differ-
ent degrees. The influence of the reaction conditions
on various aspects of substrate specificity of DNA
glycosylases has been given little attention, but there
are reasons to believe that the impact of ionic strength
and divalent cations may be significant. In one recent

concentration. In general, the
specificity of both enzymes was highest when these fac-
tors approached physiological values. The reason for
the increase in specificity was a pronounced decrease
in the activity of Fpg and OGG1 on 8-oxoG:A at high
Mg
2+
and ionic strength, whereas most of the activity
on 8-oxoG:C was retained under these conditions.
At least for OGG1, we observed only a modest
decrease in the affinity for both THF:C and THF:A
uncleavable ligands with increasing salt concentration.
In the case of Fpg, it has been reported that binding
0
1
10
100
1000
% Activity
0
50
100
150
200
Spermine, µM
0
1
10
100
1000

the lesion is available, the structures of both enzymes
complexed with undamaged DNA show that initial
recognition of the lesion involves mostly weak non-
specific interactions partly mediated through a water
layer [51,52]. Such protein–DNA interactions are easily
competed out by small cations [53]. Because the initial
recognition complexes exist for longer during process-
ing of 8-oxoG:A by either Fpg or OGG1, whereas
with 8-oxoG:C the reaction quickly proceeds to its
catalytic steps [19,20], the effect of electrostatic screen-
ing by higher ionic strength may be more pronounced
with 8-oxoG:A substrates; it is also possible that
electrostatic interactions may stabilize catalytically
inactive conformations of the enzyme in the case of
8-oxoG:A.
Unlike 8-oxoG:C pairs that exist in a conventional
anti ⁄ anti conformation [54], the 8-oxoG deoxynucleo-
tide in 8-oxoG:A mispairs prefers a syn conformation
due to steric repulsion between the O
8
atom and the
5¢-phosphate [3,55]. Thermodynamic and modeling
studies suggest that 8-oxoG may exist in a syn⁄ anti equi-
librium when paired with A [56,57]. Mg
2+
is known to
induce conformational transitions in nucleic acids, pos-
sibly by selective stabilization ⁄ destabilization of one of
the conformations; for example, submillimolar concen-
trations of Mg

than its DNA glycosylase activity.
The nature of anions in the reaction mixture did not
affect the general specificity, although some details were
notably different between reactions performed in the
presence of KCl and KGlu. For example, KGlu sup-
ported the DNA glycosylase activity of OGG1 on
8-oxoG:A over a much wider range of salt and Mg
2+
concentrations than KCl did, and tended to sustain the
activity of Fpg and the AP lyase activity of OGG1 at
low Mg
2+
better than KCl. At the same time, high
KGlu concentrations selectively decreased the turnover
of OGG1 on the 8-oxoG:A substrate and the affinity of
OGG1 for the uncleavable THF:A ligand. Some anions
are known to influence the activity of DNA glycosylases;
for example, phosphate is a competitive inhibitor of Fpg
with K
i
$ 10 mm [61], and we observed that the activity
of Fpg on 8-oxoG:C in the absence of KP
i
was generally
higher than in its presence. However, it seems that the
overall differences between Cl
)
and glutamate are not
decisive for the opposite-base specificity of Fpg and
OGG1. Also, although the buffering capacity of KGlu

observed with Fpg, which is also not inhibited by free
8-oxoG [21,60], and OGG1, which has been reported to
be inhibited by 8-oxoG [60].
Factors affecting the specificity of Fpg and OGG1 V. S. Sidorenko et al.
3756 FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS
The problem of discriminating against the DNA ele-
ments that are close to the substrates but should not be
processed is not unique for DNA glycosylases. Well-
known phenomena are the misinsertion of an incorrect
dNMP by DNA polymerases [67,68] or ‘star activity’ of
restriction endonucleases [69]. In all these cases, the
fidelity of the enzymatic reaction may be significantly
influenced by the presence of various compounds in the
reaction mixture. Generally, lower ionic strength, high
crowding and divalent cations other than Mg
2+
tend to
facilitate non-specific reactions, whereas high ionic
strength and Mg
2+
increase the fidelity. In this respect,
some intracellular factors (ionic composition) favor cor-
rect, physiologically relevant substrates, whereas the
others (crowding) act in the opposite direction. As our
results indicate, inside the cell, the preference of Fpg
and OGG1 for processing substrates with C as
compared with those containing A opposite the lesion,
significantly relies on the high intracellular ionic
strength and Mg
2+

substrate, 25 mm KP
i
(pH 7.5; omitted in some specified
cases), 1 mm dithiothreitol and Fpg (5 nm) or OGG1
(10 nm for the glycosylase activity assay, 100 nm for the
AP lyase activity assay). The enzymes were diluted to
10-fold working concentrations in 12.5 mm KP
i
(pH 7.5)
supplemented with 0.5 mgÆmL
)1
BSA. The varied compo-
nents of the reaction mixture included KCl, KGlu, MgCl
2
,
spermine or spermidine and poly(ethylene glycol) (4000 or
8000); the components included in complete factorial design
experiments are summarized in Table 1. When KGlu was
used as a component of the reaction mixture, its stock was
first brought to pH 7.5 by titration with glutamic acid and
then used to prepare the reaction mixtures; the concen-
trations are reported with respect to glutamate anion. If
necessary, the reaction mixture was supplemented with
0.1–20 mm biotin or caffeine. The reaction was initiated by
adding the enzyme (1 ⁄ 10 volume of the reaction mixture)
and allowed to proceed for 5 min (Fpg) or 10 min (OGG1).
In the case of Fpg, the reaction was terminated by mixing
with 1 ⁄ 2 vol. of formamide loading dye [71] and heating
for 5 min at 95 °C. To measure the glycosylase activity of
OGG1, the reaction was terminated by heating for 30 min

mine the product release rate constant (k
3
) for OGG1 in the
burst rate assay, the reaction mixture included 50 nm
substrate and 10 nm enzyme. The reaction mixture was
incubated at 37 °C for 0.5–5 min and the glycosylase activity
was assayed. All reported kinetic constants were obtained by
non-linear least-square fitting using sigmaplot v. 8.0 soft-
ware (Systat Software, San Jose, CA, USA).
Fluorimetric titration
Binding of Fpg and OGG1 to the THF-containing ligands
was assayed by following tryptophan fluorescence of the
enzymes using an SFM-25 spectrofluorimeter (Kontron,
V. S. Sidorenko et al. Factors affecting the specificity of Fpg and OGG1
FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS 3757
Munich, Germany), equipped with a 80 lL cuvette thermo-
stated at 15 °C. The fluorescence was excited at k
ex
=
290 nm, and the emission was recorded at k
em
= 300–
360 nm. The reaction mixtures were the same as described
under ‘Kinetic constants determination’ except the enzyme
concentrations were different (1 lm Fpg, 3.6 lm OGG1)
and the oligonucleotide ligand was added in small incre-
ments until the binding curve was saturated. K
d
values were
extracted by fitting the fluorescence data to a single-site

434.
5 Hollstein M, Sidransky D, Vogelstein B & Harris CC
(1991) p53 mutations in human cancers. Science 253 ,
49–53.
6 Li D, Firozi PF, Zhang W, Shen J, DiGiovanni J, Lau
S, Evans D, Friess H, Hassan M & Abbruzzese JL
(2001) DNA adducts, genetic polymorphisms, and
K-ras mutation in human pancreatic cancer. Mutat
Res 513, 37–48.
7 Michaels ML & Miller JH (1992) The GO system
protects organisms from the mutagenic effect of the
spontaneous lesion 8-hydroxyguanine (7,8-dihydro-
8-oxoguanine). J Bacteriol 174, 6321–6325.
8 Tchou J & Grollman AP (1993) Repair of DNA
containing the oxidatively damaged base, 8-oxoguanine.
Mutat Res 299, 277–287.
9 Tsuzuki T, Nakatsu Y & Nakabeppu Y (2007) Signifi-
cance of error-avoiding mechanisms for oxidative DNA
damage in carcinogenesis. Cancer Sci 98, 465–470.
10 Wilson DM III & Bohr VA (2007) The mechanics of
base excision repair, and its relationship to aging and
disease. DNA Repair 6, 544–559.
11 Tajiri T, Maki H & Sekiguchi M (1995) Functional
cooperation of MutT, MutM and MutY proteins in
preventing mutations caused by spontaneous oxidation
of guanine nucleotide in Escherichia coli. Mutat Res
336, 257–267.
12 Sakumi K, Tominaga Y, Furuichi M, Xu P, Tsuzuki T,
Sekiguchi M & Nakabeppu Y (2003) Ogg1 knockout-
associated lung tumorigenesis and its suppression by

YN, Nevinsky GA, Douglas KT & Fedorova OS (2007)
Pre-steady-state kinetic study of substrate specificity of
Escherichia coli formamidopyrimidine-DNA glycosylase.
Biochemistry 46, 424–435.
20 Kuznetsov NA, Koval VV, Nevinsky GA, Douglas KT,
Zharkov DO & Fedorova OS (2007) Kinetic conforma-
Factors affecting the specificity of Fpg and OGG1 V. S. Sidorenko et al.
3758 FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS
tional analysis of human 8-oxoguanine-DNA glycosy-
lase. J Biol Chem 282, 1029–1038.
21 Tchou J, Kasai H, Shibutani S, Chung M-H, Laval
J, Grollman AP & Nishimura S (1991) 8-Oxoguanine
(8-hydroxyguanine) DNA glycosylase and its
substrate specificity. Proc Natl Acad Sci USA 88,
4690–4694.
22 Castaing B, Geiger A, Seliger H, Nehls P, Laval J,
Zelwer C & Boiteux S (1993) Cleavage and binding of
a DNA fragment containing a single 8-oxoguanine
by wild type and mutant FPG proteins. Nucleic Acids
Res 21, 2899–2905.
23 Tchou J, Bodepudi V, Shibutani S, Antoshechkin I,
Miller J, Grollman AP & Johnson F (1994) Substrate
specificity of Fpg protein: recognition and cleavage of
oxidatively damaged DNA. J Biol Chem 269, 15318–
15324.
24 Auffret van der Kemp P, Thomas D, Barbey R,
de Oliveira R & Boiteux S (1996) Cloning and
expression in Escherichia coli of the OGG1 gene of
Saccharomyces cerevisiae, which codes for a DNA
glycosylase that excises 7,8-dihydro-8-oxoguanine and

AP-endonuclease: potential coordination of the initial
steps in base excision repair. Nucleic Acids Res 29, 430–
438.
31 Sidorenko VS, Nevinsky GA & Zharkov DO (2007)
Mechanism of interaction between human 8-oxo-
guanine-DNA glycosylase and AP endonuclease. DNA
Repair 6, 317–328.
32 O’Connor TR, Graves RJ, de Murcia G, Castaing B &
Laval J (1993) Fpg protein of Escherichia coli is a zinc
finger protein whose cysteine residues have a structural
and ⁄
or functional role. J Biol Chem 268, 9063–9070.
33 Paz-Elizur T, Takeshita M & Livneh Z (1997) Mecha-
nism of bypass synthesis through an abasic site analog
by DNA polymerase I. Biochemistry 36, 1766–1773.
34 Moinard C, Cynober L & de Bandt J-P (2005) Poly-
amines: metabolism and implications in human diseases.
Clin Nutr 24, 184–197.
35 Kun E, Kirsten E, Mendeleyev J & Ordahl CP (2004)
Regulation of the enzymatic catalysis of poly(ADP-
ribose) polymerase by dsDNA, polyamines, Mg
2+
,
Ca
2+
, histones H
1
and H
3
, and ATP. Biochemistry 43,

ylases. DNA Repair 2, 839–862.
45 Boiteux S, O’Connor TR & Laval J (1987) Formamido-
pyrimidine-DNA glycosylase of Escherichia coli: cloning
and sequencing of the fpg structural gene and over-
production of the protein. EMBO J 6, 3177–3183.
46 Karakaya A, Jaruga P, Bohr VA, Grollman AP &
Dizdaroglu M (1997) Kinetics of excision of purine
lesions from DNA by Escherichia coli Fpg protein.
Nucleic Acids Res 25, 474–479.
V. S. Sidorenko et al. Factors affecting the specificity of Fpg and OGG1
FEBS Journal 275 (2008) 3747–3760 ª 2008 The Authors Journal compilation ª 2008 FEBS 3759
47 Zharkov DO & Grollman AP (2005) The DNA track-
walkers: principles of lesion search and recognition by
DNA glycosylases. Mutat Res 577, 24–54.
48 Adhikari S, Toretsky JA, Yuan L & Roy R (2006)
Magnesium, essential for base excision repair enzymes,
inhibits substrate binding of N-methylpurine-DNA
glycosylase. J Biol Chem 281, 29525–29532.
49 Eide L, Luna L, Gustad EC, Henderson PT, Essig-
mann JM, Demple B & Seeberg E (2001) Human
endonuclease III acts preferentially on DNA damage
opposite guanine residues in DNA. Biochemistry 40,
6653–6659.
50 Radicella JP, Dherin C, Desmaze C, Fox MS & Boiteux S
(1997) Cloning and characterization of hOGG1, a human
homolog of the OGG1 gene of Saccharomyces
cerevisiae. Proc Natl Acad Sci USA 94, 8010–8015.
51 Banerjee A, Yang W, Karplus M & Verdine GL (2005)
Structure of a repair enzyme interrogating undamaged
DNA elucidates recognition of damaged DNA. Nature

5
dC)Æpoly(dG-m
5
dC). Proc Natl Acad Sci
USA 78, 1619–1623.
59 Subirana JA & Soler-Lo
´
pez M (2003) Cations as hydro-
gen bond donors: a view of electrostatic interactions in
DNA. Annu Rev Biophys Biomol Struct 32, 27–45.
60 Morland I, Luna L, Gustad E, Seeberg E & Bjøra
˚
sM
(2005) Product inhibition and magnesium modulate the
dual reaction mode of hOgg1. DNA Repair 4, 381–387.
61 Ishchenko AA, Vasilenko NL, Sinitsina OI, Yamkovoy
VI, Fedorova OS, Douglas KT & Nevinsky GA (2002)
Thermodynamic, kinetic, and structural basis for recog-
nition and repair of 8-oxoguanine in DNA by Fpg pro-
tein from Escherichia coli. Biochemistry 41, 7540–7548.
62 Krokan H & Wittwer CU (1981) Uracil DNA-glyco-
sylase from HeLa cells: general properties, substrate
specificity and effect of uracil analogs. Nucleic Acids
Res 9, 2599–2613.
63 Thomas L, Yang CH & Goldthwait DA (1982) Two
DNA glycosylases in Escherichia coli which release pri-
marily 3-methyladenine. Biochemistry 21, 1162–1169.
64 Bjelland S & Seeberg E (1987) Purification and charac-
terization of 3-methyladenine DNA glycosylase I from
Escherichia coli. Nucleic Acids Res 15, 2787–2801.