Chemistry of the 8-Nitroguanine DNA Lesion: Reactivity, Labelling and Repair

Article abstract of DOI:10.1002/chem.201705541

The 8-nitroguanine lesion in DNA is increasingly associated with inflammation-related carcinogenesis, whereas the same modification on guanosine 3′,5′-cyclic monophosphate generates a second messenger in NO-mediated signal transduction. Very little is known about the chemistry of 8-nitroguanine nucleotides, despite the fact that their biological effects are closely linked to their chemical properties. To this end, a selection of chemical reactions have been performed on 8-nitroguanine nucleosides and oligodeoxynucleotides. Reactions with alkylating reagents reveal how the 8-nitro substituent affects the reactivity of the purine ring, by significantly decreasing the reactivity of the N2 position, whilst the relative reactivity at N1 appears to be enhanced. Interestingly, the displacement of the nitro group with thiols results in an efficient and specific method of labelling this lesion and is demonstrated in oligodeoxynucleotides. Additionally, the repair of this lesion is also shown to be a chemically feasible reaction through a reductive denitration with a hydride source.

Full text of DOI:10.1002/chem.201705541

A Journal of
Accepted Article
Title: Chemistry of the 8-nitroguanine DNA lesion- reactivity, labelling
and repair
Authors: Katie J Alexander, Matthew McConville, Kathryn R Williams,
Konstantin V Luzyanin, Ian A O'Neil, and Richard Cosstick
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201705541
Link to VoR: http://dx.doi.org/10.1002/chem.201705541
Supported by

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
Chemistry of the 8-nitroguanine DNA lesion- reactivity, labelling
and repair
Katie J. Alexander, Matthew McConville, Kathryn R. Williams, Konstantin V. Luzyanin, Ian A. O’Neil
and Richard Cosstick*
Abstract: The 8-nitroguanine lesion in DNA is increasingly
associated with inflammation-related carcinogenesis, whilst the
same modification on guanosine 3’,5’-cyclic monophosphate
generates a second messenger in NO-mediated signal transduction.
tissues from the sites of cancers that are caused by chronic
inflammation9 and the accumulation of 8-nitroG is also
associated with
a poor prognosis of inflammation-induced
cancers.9,10 Thus, 8-nitroG is progressively seen as
a
Very little is known about the chemistry of 8-nitroguanine nucleotides, biochemical marker for DNA damage and as potential indicator
despite the fact that their biological effects are closely linked to their
chemical properties. To this end, a selection of chemical reactions
have been performed on 8-nitroguanine nucleosides and
oligodeoxynucleotides. Reactions with alkylating reagents reveal
how the 8-nitro substituent affects the reactivity of the purine ring, by
significantly decreasing the reactivity of the N2 position, whilst the
relative reactivity at N1 appears to be enhanced. Interestingly, the
displacement of the nitro group with thiols results in an efficient and
specific method of labelling this lesion and is demonstrated in
oligodeoxynucleotides. Additionally, the repair of this lesion is also
shown to be a chemically feasible reaction through a reductive
denitration with a hydride source.
of inflammation-induced cancers. Nitration of the guanine base
also occurs in the ribonucleotide pool and 8-nitroguanosine-3’,5’-
cyclic phosphate11 (8-nitro-cGMP) has attracted attention as an
electrophilic second messenger that is involved in an unusual
post translational modification of proteins. This process of
protein S-guanylation,11 results from the nucleophilic sulfhydryl
group of a cysteine residue displacing the nitro group of 8-nitro
cGMP. More recently H2S, a proposed gaseous transmitter in
mammals, has been shown to react with and sequester 8-nitro-
cGMP, and provides a potential mechanism to shut down this
electrophile-mediated signalling pathway.12
O
O
HN
N
HN
N
H2N
HO
H2N
R1O
NH
N
Introduction
N
N
O
NO2
O
O
The guanine base in DNA is particularly sensitive to chemical
modification1 by reactive oxygen species and reactive nitrogen
species (RNS) and a number of mutagenic lesions, such as 8-
nitro-2’-deoxyguanosine2 (8-nitro-dG, 1a, Figure 1) and the more
widely studied 8-oxo-2’-deoxyguanosine3 (8-oxo-dG, 2a) are
associated with these reactive species. A characteristic of 8-
nitrodG is that the nitro group labilises the glycosidic bond to
release 8-nitroguanine (8-nitroG), leaving an apurinic site in the
DNA.4 The mutagenic nature of 8-nitrodG results from both the
generation of apurinic sites5 in the DNA and the mispairing of 8-
nitrodG6-8 during DNA synthesis; both these events lead to G to
T transversion mutations.
The RNS, which are responsible for the nitrative damage
are generated particularly at sites of inflammation, due to over-
expression of nitric oxide synthase and given the mutagenic
nature of 8-nitro-dG, it is not surprising that this lesion is
increasingly linked with inflammation-related carcinogenesis.9
Specific antibodies raised against 8-nitroG have been used to
show that 8-nitroG is present in surgically-removed human
R2
R1O
R
HO
1a, R1 = R2 = H = 8-nitro-dG
1b, R1 = H, R2 = OH, 8-nitro-rG
1c, R1 = CH3CO, R2 = OAc
1d, R1 = H, R2 = OMe
2a, R = H = 8-oxo-dG
2b, R = OMe
Figure 1.
Collectively these diverse studies have shown that the
chemical properties of 8-nitroguanine nucleotides are inherently
related to their biological function and have encouraged us to
look more closely at their chemistry. We now report our studies
on some chemical reactions performed on 8-nitroguanine
nucleosides and oligonucleotides. The results shed some light
on how the 8-nitro substituent affects the reactivity of the purine
ring, they demonstrate that the 8-nitroguanine lesion in DNA can
be fluorescently labelled and raise the possibility of a repair
mechanism for this lesion.
Dr. K.J. Alexander, Dr. M. McConville, Kathryn R. Williams, Dr. K. V.
Luzyanin, Dr. I. A. O’Neil and Prof. R. Cosstick
Department of Chemistry,
Results and Discussion
University of Liverpool,
Crown Street, Liverpool,
Alkylation reactions- Previous studies have demonstrated that
some chemical properties of guanine nucleosides are
significantly altered by introduction of the 8-nitro group. The nitro
group greatly enhances the hydrolytic lability of the glycosidic
bond4 and also significantly lowers the pKa of the proton on N1
(pKa = 8.4 for 8-nitroguanosine compared to 9.4 for guanosine).8
L69 7ZD, (UK)
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
In addition, we have made attempts to introduce the usual acyl
or formamidine protecting groups at the N2 position of 8-nitroG
nucleosides, but were unable to prepare these derivatives using
routine procedures.8 This failure to derivatise the N2 position
implied that the reactivity of the guanine base is significantly
reduced because of the strongly electron withdrawing nitro
group and suggested that the chemistry of 8-nitroguanine
nucleosides required wider exploration.
bond; a chemical shift of approximately 70ppm would be
expected should the benzylic carbon have been attached to an
oxygen atom. From the 1H NMR it was possible to determine
which nitrogen atom was alkylated. Acquisition of the spectra in
a
non-protic solvent indicated the absence of the signal
corresponding to the N1 proton which was present in the starting
material at 11.34ppm. Additionally, the broad peak at 6.06ppm
corresponding to the N2 signal integrated to 2 protons. These
observations confirm that alkylation occurred at the N1 position.
This assignment was further strengthened when compared to
the N2 alkylated product previously characterised from the
tBu
O
O
tBu
tBu
reaction between deoxyguanosine and
a
para-quinone
1
N
N
methide,13 in which these observations were reversed; the peak
at 10.35ppm associated with the N1 proton was present and the
signal associated with the N2 proton integrated as a single
proton.
(p-QM)
NO2
1c
p-QM, 1 eq
DMF, 37oC, 1h
N
HO
H2N
N
R
tBu
3
68% yield
o-QM, 4 eq
DMF, 37oC,
10 min
We also examined the alkylation of acetylated 8-
nitroguanosine with an ortho-quinone methide (o-QM) produced
O
OH
from
2-bromomethyl-O-tert-butyldimethyl-silylphenol
using
(o-QM)
potassium fluoride (Scheme 1).15,16 In an initial experiment with
one equivalent of the o-QM, two products were observed by
HPLC; a more polar product (~40%) and a less polar product
(~5%). When four molar equivalents of o-QM were used a more
even distribution of the products was obtained and the more
polar and less polar products were isolated in yields of 40% and
27%, respectively. Mass spectrometry revealed the more polar
(4) and less polar (5) products to be mono-alkylated and di-
alkylated species, respectively (Scheme 1). NMR techniques
revealed the mono-alkylated product (4) to be entirely consistent
with the N1-ortho quinone methide, based on similarities to the
N1-para quinone methide product (3).
6
O
N
OH
O
N
N
1
N
N
N
OH
N
NO2
NO2
2
N
H
H2N
N
R
R
5
4
dialkylated
27% yield
monoalkylated
40% yield
R = 2,3,5,-tri-O-acetyl-β-D-ribofuranosyl-1-yl
Scheme 1. Reactions of 2’,3’,5’-tri-O-acetyl-8-nitroguanosine with quinone
methides. Experimental details in the Supplementary Information.
For the dialkylated product (5), the presence of two non-
nucleosidic CH2 groups with chemical shifts of 67.36ppm and
40.00ppm in the 13C NMR spectrum, was established. These
resonances are characteristic of a benzylic carbon-oxygen bond
and a benzylic carbon-nitrogen bond, respectively and therefore
indicated that O6 is one of the alkylation sites.14,17 With regard to
the second alkylation site on nitrogen, alkylation at the N1
position seemed unlikely. This was supported by 1H NMR
spectra recorded in a non-protic solvent revealed the presence
of a single NH signal which integrated to one proton.
We chose to investigate alkylation reactions on 2’,3’,5’-tri-
O-acetyl-8-nitroguanosine (1c, Figure 1), as in this substrate the
electron withdrawing acetyl groups, particularly on the 2’-OH,
help to further stabilise the glycosidic bond. Given the seemingly
inert nature of the 8-nitroguanine base, ortho- and para-quinone
methides were selected as alkylating agents as they possess a
benzylic carbon which is highly electron deficient and extremely
susceptible to nucleophilic attack.13,14 In addition, the products of
the reactions of these quinone methides with 2’-deoxyguanosine
have been well characterised.13,14
The results of alkylation reactions performed here with
2’,3’,5’-tri-O-acetyl-8-nitroguanosine
(1c) show significant
differences to those previously reported for 2’-deoxyguanosine
(dG), both in terms of the products formed and their distribution.
For example, the sole nucleoside product isolated from
alkylation of (1c) with p-QM is the N1 adduct (3), whereas
previous results obtained with dG show N2 to be the favoured
site of alkylation.14 That the 8-nitro group leads to preferential
reaction at N1 rather than N2 is not so surprising given that the
lone pair on N2 is directly conjugated through to the nitro group
(Scheme 2) and that reaction at N1 of the nitrated nucleoside
might be enhanced by deprotonation, due to its significantly
lower pKa.
In initial experiments, acetylated 8-nitroguanosine (1c) was
reacted with a molar equivalent of the para-quinone methide (p-
QM) prepared from 2,6-bis(tert-butyl)-4-methylphenol by
oxidation with silver (I) oxide (Scheme 1).13 Under these
conditions TLC and HPLC showed that a single nucleoside
product (3) was formed and was subsequentlty isolated in 68%
yield following column chromatography. Mass spectrometry
ascertained that (3) was a monoalkylated product and N1 was
established as the position of alkylation through NMR
techniques. DEPT 135 established the presence of two CH2
groups and through HSQC and COSY experiments benzylic
protons from the quinone methide adduct were identified as a
singlet at 5.11ppm. In the 13C NMR spectrum the benzylic signal
resonated at 45.70ppm, characteristic of a carbon-nitrogen
The reaction of excess o-QM with nitrated nucleoside (1c)
once again gave the N1 adduct (4) as the major product (40%),
together with the O6,N2-dialkylated product (5, 27%). Alkylation
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
at N1 was expected based on the result obtained with p-QM,
particularly since the use of excess potassium fluoride to
generate the o-QM in unbuffered conditions, would be expected
to lead to mildly basic conditions that favour reaction at N1 due
to deprotonation. The O6,N2-dialkylated product was
unexpected, since no O6 or N2 mono-alkylated products were
observed. The ortho geometry of the o-QM means that
hydrogen-bonding of NH1 with the carbonyl group of the
quinone methide could assist nucleophilic attack by either O6 or
N2 (Scheme 2). Given, that N2 appears to be highly deactivated,
attack by O6 is likely to be favoured from the hydrogen-bonded
complex. In order to explain why no O6 mono-alkylated product
is observed, the subsequent alkylation on N2, would have to be
relatively fast and might be assisted by the alkylated and
therefore electron donating O6. In contrast, studies on the
reaction between dG and o-QM have been shown to give N2
Reductive denitration of 8-nitroguanosine- Aromatic
nitro compounds such as nitroimidazoles have emerged as
potent anti-TB pro-drugs that exert their activity through the
release of nitric oxide (NO), which is the effective agent.19 NO
released from these drugs is initiated by nucleophilic attack of a
hydride equivalent from a deazaflavin cofactor that is present in
a deazaflavin-dependent nitroreductase and the overall reaction
is a denitration in which NO2 is replaced by H.19 A related
reaction performed on the 8-nitroG lesion in DNA, in which a
hydride equivalent was delivered to the C8 position, could
remove the nitro group through an addition-elimination reaction,
thus suggesting a potential DNA repair mechanism by directly
reversing the lesion. Relatively few DNA lesions are repaired by
direct chemical reversal, but this route is well established for the
repair of UV-induced pyrimidine dimer lesions20 and the repair of
guanine O6 alkylation products by O6-alkylguanine DNA
alkyltransferase.21 We were therefore interested in investigating
the chemical feasibility of reducing 8-nitroguanosine back to
guanosine as this could suggest a potential repair mechanism
for this lesion by direct chemical reversal.
alkylation as either the sole product18 or to favour N2 alkylation
over N1 by a factor of about 5.14
A
There are a few examples where borohydride has been
used to effect aromatic denitration and these have been
reported for both polysubstituted nitrobenzenes22,23 and
heterocyclic systems.24 These reactions take place through an
addition elimination mechanism in which nitrite is displaced by
hydride.23 Earlier studies25 on the characterisation of the 8-
nitroguanine base have shown that reducing agents such as:
sodium hydrosulfite; zinc-HCl; hydrazine hydrate-Raney nickel,
all yield the expected 8-aminoguanine, but to our knowledge the
reaction has not been studied with hydride reducing agents.
6
O
N
O
O6-alkylated
H
N
N
N
O
NO2
2
N
H
H
dialkylated (5)
Sugar
B
O
N
O
N
N
O
N
N
O
O
HN
HN
O
O
N
N
N
N
N
H2N
O
N
N
HN
H2N
H2N
HN
NaBH4
H2O
NH2
SugarSugar
SugarSugar
8-nitro-rG
1b
N
N
H2N
O
R
R
Scheme 2. Panel A shows O6 versus N2 hydrogen bond-directed alkylation,
resulting in initial alkylation at the presumably more reactive O6 position and
subsequent alkylation at N2, to give dialkylated 5. Panel B shows reduced
nucleophilicity of N2 because of conjugation through to the nitro group.
guanosine
8-aminoguanosine
N
NH
O
N
N
NH2
O
R = β-D-ribofuranosyl-1-yl
The alkylation reactions reported here have been
performed on 2’,3’,5’-tri-O-acetyl-8-nitroguanosine in order to
reduce depurination and were also optimised to generate good
yields of the products to aid their characterisation, therefore the
reaction outcomes cannot be directly compared with quantitative
studies previously obtained with dG. However, given that the
sites of alkylation for DNA bases with quinone methides remain
OH
8,5'-O-cyclo-guanosine
HO
Scheme 3. Reduction of 8-nitroG with sodium borohydride.
Prior to studying the reduction, 8-nitroguanosine (1b,
Figure 1) was purified by reverse-phase (RP)-HPLC, although
as can be seen in Figure 2, a small amount of the depurination
product 8-nitroguanine was unavoidably introduced into the
starting samples of 8-nitroguanosine during evaporation of the
fractions. It was also necessary to develop RP-HPLC conditions
that would separate the two most likely products; namely 8-
aminoguanosine and guanosine. An extensive investigation of
eluents revealed that 8-aminoguanosine and guanosine were
readily separated using a gradient of acetonitrile in aqueous
constant
in
their
deoxynucleoside,
single-stranded
oligodeoxynucleotide and duplex polydeoxynucleotide forms,18
the presence of the acetyl protecting groups on the sugar is not
expected to alter the site of alkylation. Thus, the results
presented here indicate that the reactivity of the guanine base is
significantly altered by the introduction of the 8-nitro group; the
reactivity of the N2 position significantly decreases whilst N1
appears to be enhanced, presumably due to its considerably
lower pKa.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
acetic acid (0.1% AcOH) and these elution conditions were also
compatible with HPLC-MS.
Further proof for the direct displacement of the nitro group
by hydride was obtained by performing the reaction using NaBD4
in H2O, and in this case HPLC-MS revealed the guanosine
fraction to have the correct mass for the incorporation of one
deuterium atom (see Table 1).
As expected, the peak
corresponding to 8-aminoguanosine showed no inclusion of
deuterium as any incorporated during reduction would exchange
through protonation.
In an attempt to increase the proportion of guanosine, the
reduction was investigated under a variety of conditions. Using
buffered aqueous conditions between pHs 6.0 – 9.0 (0.1M
ammonium acetate) there was little difference in terms of the
product distribution from those performed in water, although the
reaction at lower pH did progress slightly faster. It is not
surprising that all the aqueous reduction reactions gave similar
results as the very large excess of borohydride, which raises the
pH as it is hydrolysed, resulted in all these reactions becoming
alkaline as the reaction proceeded (pH ~ 9.6 at the end of the
reaction). No reduction products were observed when the
weaker reducing agent sodium cyanoborohydride was used at
pH 6.0. The work of Lamson23a has established that the
denitration of polysubstituted nitrobenzenes proceeded more
rapidly in DMSO than in protic solvents. However, when the
reaction was performed in DMSO we were unable to detect any
reduction products, even during a prolonged reaction (>1 week)
and only a trace amount of 8,5’-O-cycloguanosine was formed.
Whilst we have been unable to increase the proportion of
guanosine above about 12%, its formation by displacement of
the nitro group by a hydride equivalent, in aqueous solution,
does clearly show that a direct repair of the 8-nitroguanine lesion
is a chemically feasible reaction.
Figure. 2. RP-HPLC analysis of the reduction of 8-nitroguanosine with sodium
borohydride. Details of the reaction conditions are described in the
Experimental section. HPLC chromatograms were obtained using an elution
gradient of 0-20% acetonitrile in 0.1% aqueous acetic acid over 20 min, with a
flow rate of 0.5 ml/min. Chromatograms were recorded at 254 nm.
Table 1. High-resolution mass data for the borohydride (either NaBH4 or
NaBD4) reduction of 8-nitroguanosine as shown in Scheme 3. Accurate mass
measurements were obtained by LC-MS as described in the experimental
section. Mass spectra for individual compounds are shown in the SI.
Calculated
Measured
Nucleoside
Formula
exact mass for
[M-H]
accurate mass
for [M-H]
Detection of 8-nitroguanine through the displacement of the
nitro group by thiols- Methods have been developed to detect
the 8-nitroguanine base in tissues9,10 and urine,26 but these
methods are only applicable once the 8-nitroguanine base has
been hydrolytically released from DNA and they give no
information as to the steady-state levels of this lesion in DNA
samples. We have recently shown that surface-enhanced
Raman spectroscopy is a very sensitive technique for detecting
this lesion due to the nitro group and can distinguish between
the 8-nitroguanine that is covalently bound to DNA or has been
released by hydrolysis.27
-
-
Guanosine
C10H13N5O5
C10H12DN5O5
C10H11N5O5
C10H14N6O5
282.0844
283.0907
280.0687
297.0953
282.0827
283.0906
280.0693
297.0963
8-Deuteroguanosine
8,5’-O-Cycloguanosine
8-Aminoguanosine
Reduction reactions (Scheme 3) were generally performed
in unbuffered distilled water using a very large excess of sodium
borohydride that was repeatedly added at regular intervals to
compensate for hydrolysis. Under these conditions a slow
reduction occurred (Figure 2) that gave 8-aminoguanosine as
the major product together with a small amount (8-10%) of
guanosine (Scheme 3). The identity of both products was
confirmed by co-elution with authentic samples of guanosine
and 8-aminoguanosine and also by HPLC-MS analysis (see
Figure 2, Table 1 and SI). In addition to the reduction products,
an additional nucleoside product (~20%) was observed that was
identified by HPLC-MS as 8,5’-O-cycloguanosine and
presumably results from displacement of the nitro group by the
5’-hydroxyl group. Formation of the cyclo-nucleoside is probably
facilitated by the rise in pH caused by hydrolysis of the
borohydride.
The displacement of the nitro group in 8-nitroguanine
nucleosides by sulfur nucleophiles is well established and
specific examples include the capture of 8-nitroguanosine by the
nucleoside derivative nitroG-Grasp,28 sulfhydration12 and protein
S-guanylation29 of 8-nitro-cGMP. The mild conditions for these
displacement reactions suggested that it might be possible to
develop
a detection procedure for this lesion based on
displacement of the nitro group with a fluorescent thiol. In theory,
each lesion site would be converted to a fluorescently labelled
guanine base that would be stable to depurination (due to the
loss of the strongly electron withdrawing nitro group) and thus
create a permanent and quantifiable record of the lesion sites,
which in theory could be identifiable through sequencing
methods.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
As 8-nitro-dG (1a), which is the naturally occurring lesion
in DNA readily undergoes depurination, our initial experiments to
explore this labelling technique were performed on 8-
nitroguanosine (1b, Figure 1) and 8-nitro-2’-O-methylguanosine
(1d, , Figure 1). Both of these ribonucleosides have a more
stable glycosidic bond than their 2’-deoxy counter part, due to
the electronegative 2’-oxygen. The 2’-O-methylriboside is also
ideally suited as a means to incorporate this lesion into
oligonucleotides, as no protection of the 2´-position is necessary
during DNA synthesis and it has previously been used as a
model for 8-nitro-dG in oligonucleotides.8 We chose to use N-
(tert-butoxycarbonyl)cysteine ethyl ester (6, Scheme 4) as a
model thiol as its hydrophobicity was expected to significantly
RP-HPLC. A relatively smooth reaction was observed with
greater than 90% conversion to the less polar cysteinyl 13-mer
after 6 hours, together with formation of a small amount of the
depurinated 13-mer, which runs slightly faster than the nitro-13-
mer (Figure 3, Table 2 for retention times.). The cysteinyl 13-mer
was isolated by RP-HPLC and was shown by negative mode
electrospray mass spectrometry to have the expected mass
(Table 3 in the Experimental section).
H
N
COOEt
tBuO
O
S
NO2
6
d(GCGTACG2'OMeCATGCG)
d(GCGTACG2'OMeCATGCG)
8-cysteinylG 13-mer
increase
the
retention
time
of
the
resulting
nucleoside/oligonucleotide adducts on reverse-phase HPLC and
its low extinction coefficient at 260 nm, was not expected to
significantly interfere with the HPLC analysis. Thus, reaction of
8-nitroguanosine with a slight excess of cysteine thiol (6) in
aqueous buffer (pH 7.5), gave the S-guanylated cysteine
8-nitroG 13-mer
Scheme 5. Reaction of nitro-13-mer with cysteine nucleophile 6.
derivative (7b) in good yield (69%).30 Analysis of H-13C HMBC
1
contour map revealed a long range H-13C correlation between
1
0.8
0.4
0
the C8 resonance and the protons adjacent to the sulfur atom,
thus confirming that the sulfur atom had displaced the nitro
group. A similar yield (72%) was obtained for the conversion 8-
nitro-2’-O-methylguanosine (1d) to its corresponding 8-S-
cysteinyl derivative (7d, Scheme 4).
0
5
10
15
20
25
Minutes
O
SH
O
O
HN
N
HN
N
OtBu
NH
H2N
HO
H2N
HO
tBuO
NO2
N
N
H
(6)
COOEt
N
1
0.5
0
O
N
S
N
O
O
COOEt
OR
HO
OR
HO
0
5
10
15
20
25
7b, R = H (69%)
7d, R = Me (72%)
1b, R = H
1d, R = Me
HS
Minutes
Fluor
(8)
O
HN
Figure 3. RP-HPLC chromatogram showing reaction of 8-nitro-13-mer with
cysteine derivative 6. Top panel shows the nitro-13-mer starting material and
the bottom panel shows the reaction mixture after 6 hours. RP-HPLC was
performed as described in the experimental section using a linear gradient of
0% to 65% eluent B (blue line) in eluent A, over 19 min, with a flow rate of 1.0
ml/min. Eluent A = aq TEAB (0.1M), eluent B = 40% CH3CN in aq TEAB
(0.1M). Chromatograms were recorded at 254 nm.
H2N
O
N
O
N
Fluor
N
N
S
HO
H
O
O
OR
9d, R = Me (45%)
HO
HO
O
OH
Fluor
The cysteine nucleophile (6) was also used in
displacement reactions on the longer 27-mer and 37-mer
sequences, which are more representative of the naturally
occurring 8-nitroG lesion in DNA. These reactions were carried
out under the same conditions as those described for the nitro
13-mer and although they proceeded more slowly, >90%
conversion was observed after 11 hours. Both the nitro 27-mer
and nitro 37-mer were predominantly converted to a single
product with the expected longer retention times when
compared to the nitro-containing oligonucleotides (see Table 2
for retention times). As we were unable to obtain mass
spectrometry data for the products from these longer oligomers,
Scheme 4. Reaction of thiols with 8-nitroguanosine nucleosides. Compound 8
is shown as a mixture of the 5(6)-regioisomers.
To apply this labelling technique to model DNA sequences,
8-nitro-2’-O-methylguanosine (1d) was incorporated into 13-mer,
27-mer and 37-mer oligodeoxynucleotides. These model
sequences each contained a single 8-nitroguanine base and
were prepared using previously reported procedures.8 Reaction
of the nitro 13-mer sequence (Scheme 5) with a 100-fold excess
of cysteine thiol (6), in Tris-HCl buffer (pH 8.5) was monitored by
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
the presence the 8-S-cysteinyl group in the 27-mer and 37-mer
was confirmed by enzymatic digestion. Treatment of the product
oligomers with phosphodiesterase and alkaline phosphatase (pH
8.5), followed by HPLC analysis showed the presence of 7d in
the digestion mix, together with the four naturally occurring 2’-
deoxynucleosides. When the chromatogram from the digestion
mix was analysed to obtain the composition ratios of the
component nucleosides, the peak area for 7d was lower than
expected and an unidentified peak, with the UV absorption
characteristics of a nucleoside (retention time 9.10 min) was
also observed. It was initially thought that the unidentified peak
might be a partially digested dinucleotide, but its proportion in
the digestion mixture increased as the digestion time increased.
Given that the original cysteine nucleophile (6, retention time =
19.10 min) could also be detected in the digestion mix, the most
likely cause of this unidentified nucleoside was hydrolysis of the
8-S-cysteinyl derivative (7d) to give 8-oxo-2’-O-methylguanosine
(2b, Figure 1) and releasing (6). A standard of 8-oxo-2’-O-
methylguanosine was prepared31 and it was shown to have the
same retention time on HPLC as the unidentified nucleoside in
the digestion mixture, thus implying that 8-S-cysteinyl derivative
(7d) was breaking down in the digestion mixture.
clearly indicates the potential for using this reaction to label the
site of 8-nitroG lesions in DNA. However, whilst the hydrolysis of
the 8-S-cysteinyl guanosine derivative (7d) during the enzymatic
digestion was not evident when analysing and manipulating the
8-S-cysteinyl oligodeoxynucleotides, this reaction does suggest
that the labelling conditions might have to be optimised to
eliminate hydrolysis to the 8-oxo guanine nucleoside. The
observed hydrolytic breakdown of S-guanylated nucleosides
does suggest a potential additional source of 8-oxoguanosine
nucleosides/nucleotides in biological systems.
Conclusions
These studies reveal that the 8-nitro group exerts a significant
effect on the reactivity of the guanine base. In particular, the
reactivity of the guanine NH2 group towards electrophiles is
significantly reduced, whilst the reactivity of the NH group
appears to be enhanced, due to its lower pKa. The ease with
which the nitro group is replaced by thiol nucleophiles under mild
conditions and the high efficiency of this reaction show that sites
of guanine nitration in DNA can be fluorescently labelled using
suitably derivatised thiols. The derivatisation process leaves a
permanent marker of the original lesion that can be quantified.
By using the appropriate thiol (e.g. a thiol derivatised crown
ether) it should be relatively easy to adapt this labelling
technique so that the 8-nitroguanine base is converted into a
current-modulating marker that can be read through nanopore
sequencing.32
Table 2 Reverse-phase HPLC retention times (min) for nitro oligomers and
their reaction products with thiol nucleophiles. HPLC analysis conditions are
as described in the legend to Figure 3. Nitro-13-mer
= d(GCGTA
CGCATGCG); nitro-27-mer = d(GTGACTGGGAAAACCCTGGCGTTACCC);
nitro-37-mer = d(GCGCAACGCAATTAATGTGAGTTAGCTCACTCATTAG). G
= position of 8-nitro-2’-OMe-guanosine.
nitro-
oligomers
cysteinyl-
oligomers
8-fluoro
oligomers
Oligomer length
Finally, we have shown that reduction of 8-nitroguanosine
with sodium borohydride in aqueous solution gives 8-
aminoguanosine as the major product, but is also accompanied
by a denitration reaction to give guanosine. The formation of
13-mer
27-mer
37-mer
12.55
12.82
12.52
14.78
14.13
13.74
18.63
-
-
guanosine, in the presence of
a
hydride equivalent,
demonstrates that the direct repair of the 8-nitroguanine lesion is
a chemically feasible reaction.
These results significantly enhance our knowledge of the
chemical properties of this base modification and provide new
methods to detect and study this lesion.
Finally, we wanted to use a fluorescent thiol derivative in
order to demonstrate the labelling of the 8-nitroG lesion with a
readily quantifiable reporter group. We chose to use a cystamine
derivative of carboxyfluorescein (8), which had previously been
prepared as a mixture of regioisomers and used to label cell-
penetrating oligoguanidinium compounds.30 This fluorescent
thiol (8) was subsequently used to displace the nitro group in 8-
nitro-2’-O-methylguanosine (1d), using similar conditions to
those described for the preparation of (7d, Scheme 4). As the
fluorescent thiol (8) is a mixture of regioisomers, the resulting
fluorescein adduct formed with 8-nitro-2’-O-methylguanosine
(9d) was more difficult to isolate and characterise than the 8-S-
cysteinyl derivatives, but the identity of the product was
confirmed by high resolution mass spectrometry. The
fluorescent thiol (8) was also used to label the nitro 13-mer and
the resulting fluorescein-tagged 13-mer was isolated by HPLC
and identified by electrospray mass spectrometry.
Experimental Section
8-Nitroguanosine (1b) and 8-nitro-2’-O-methylguanosine (1d) were
prepared as previously described.8 2’3’5’-Tri-O-acetyl-8-nitroguanosine
(1c) was prepared as described by Saito et al.29(b) N-(tert-
butoxycarbonyl)cysteine ethyl ester (6) was prepared by esterification of
N-(tert-butoxycarbonyl)cystine.33
8-Oxo-2’-O-methylguanosine
was
prepared from 8-bromo-2’-O-methylguanosine8 following the procedure
of Nampalli and Kumar.31 5(6)-Carboxyfluorescein was purchased from
Sigma-Aldrich as a mixture of 5- and 6-carboxy regioisomers.
All RP-HPLC was performed using Gemini™ C18 columns (110 Å,
250 mm x 4.6 mm) purchased from Phenomenex. RP-HPLC analysis of
oligonucleotides was performed on an automated Gilson HPLC system
equipped with an autoinjector, a photodiode array detector and dual
hydraulic pump, operating with a flow rate 1 ml/min. Chromatograms
were recorded at 254 nm. Chromatographic data were controlled and
processed using UniPoint© Version 3.0. All oligonucleotides were purified
The mild conditions and clean reaction of the cysteine thiol
(6) with the nitro 13-mer (Figure 3) and longer sequences,
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
by RP-HPLC using
a
triethylammonium bicarbonate (TEAB) buffer
8-cysteinyl 13-mer = d(GCGTACG(tBu-Cys)CATGCG)
system. A 1M stock solution of TEAB was prepared by bubbling CO2
through a 1M aqueous solution of Et3N for 6 – 8 hours until the pH
measured 7.5 – 7.6. Stock solutions of TEAB were refrigerated until
needed. All UV measurements were taken on a PerkinElmer Lambda 25
spectrophotometer.
8-cysteinyl 27-mer = d(GTGACTGGG(tBu-Cys)AAAACCCTGGCG-
TTACCC)
8-cysteinyl 37-mer = d(GCGCAACGCAATTAATGTG(tBu-Cys)AGT-
TAGCTCACTCATTAG)
To a solution of the appropriate 8-nitro oligomer (3.0 OD (260nm) units)
dissolved in distilled H2O (100µl) was added 1M TRIS.HCl (pH 8.5,10µl)
and ethyl N-(tert-butoxycarbonyl)-cysteinate (6, 0.80mg, >100eq). The
resulting mixture was stirred at 37°C for 6 hours during which time the
progress of the reaction was monitored by RP-HPLC using a gradient of
0-26% acetonitrile in TEAB (0.1 M). Retention times for the
oligonucleotide substrates and products are presented in Table 2.
Oligonucleotide products were purified using the same gradient and the
fractions containing the products were evaporated 3 times with distilled
water in a vacuum centrifuge.
RP-HPLC analysis of 8-nitroguanosine reduction reactions were
performed on an Agilent 1260 Infinity system equipped with an
autoinjector,
a photodiode array detector and quaternary pump.
Chromatographic data were controlled and processed using Agilent
Open LAB Chemstation software. An elution gradient of 0-20%
acetonitrile in 0.1% aqueous acetic acid over 20 min, with a flow rate of
0.5 ml/min. Chromatograms were recorded at 254 nm.
LC-MS was performed on an Agilent 6530B accurate mass Q-TOF
G6530B mass spectrometer connected to an Agilent 1260 Infinity HPLC
system as described above, with a flow rate of 0.5 ml/min. LC-MS data
was obtained using a multimode ion source and processed using Agilent
Masshunter software. Spectra were recorded in negative ionisation mode.
8-Fluoroscein labelled oligomer
8-Fluoroscein 13-mer = d(GCGTACG(Fluro)CATGCG). To a solution of the
8-nitro 13-mer (3.1 OD units) dissolved in distilled H2O (100µl) was
added 1M TRIS.HCl (pH 8.5, 10µl) and 3’,6’-dihydroxy-3-oxo-N-(2-
sulfanylethyl)-3H-spiro[2-benzofuran-1,9’-xanthene]-5(6)-carboxamide (8,
1.7mg). The resulting mixture was stirred at 37°C for 24 hours during
which time the progress of the reaction was monitored by RP-HPLC
using a gradient of 0-26% acetonitrile in TEAB (0.1 M). Retention times
for the oligonucleotide substrates and products are presented in Table 2.
Oligonucleotide products were purified using the same gradient and the
fractions containing the products were evaporated 3 times with distilled
water in a vacuum centrifuge.
All NMR spectra were recorded on
a Bruker 400 MHz
spectrometer; 1H (400 MHz) and 13C (100 MHz). All chemical shifts are
reported in p.p.m. and coupling constants (J) in Hz. 13C spectra were H
1
decoupled. 1H and 13C chemical shifts are relative to an internal
standard of tetramethylsilane. Mass spectra of nucleosides and
chemically synthesised ODNs were recorded on a MicroMass LCT mass
spectrometer using electrospray ionisation (ES) and direct infusion
syringe pump sampling. All small molecules were dissolved in methanol.
Oligonucleotides were dissolved in water and analysed using negative
mode ES ionisation.
Table 3. Mass data for all the 13-mer oligonucleotides, obtained by ES
ionisation as described above. 13-mer sequence = d(GCGTACGCATGCG), G
= position of modification.
Acknowledgements
MM was supported by the Engineering and Physical Sciences
Research Council (EPSRC) (research grant EP/H002464/1) and
KJA and KRW were funded by EPSRC studentships. We would
like to thank the University of Liverpool MicroBioRefinery facility
for access to LC-MS and to Stephen Moss for LC-MS analysis.
We would also like to thank Alex Bretherton for performing
preliminary studies on the reduction of 8-nitroguanosine.
Calculated
exact mass
Measured
accurate mass
Oligonucleotide
8-NitroG-13mer
4050.7025
4250.7996
4438.7810
4050.6133
4250.6850
4438.7090
Cysteinyl 13-mer
Fluoroscein 13-mer
Keywords: Oligonucleotides; DNA modification, DNA lesion;
DNA repair
Oligodeoxynucleotide synthesis and purification
Synthesis and purification of 13- and 27- and 37-mer
oligodeoxynucleotides containing 8-nitro-2’-O-methyl-guanosine was
performed as previously described.8 All oligodeoxynucleotides containing
the 8-nitroguanine modification showed the characteristic absorption at
395 nm.
[1]
[2]
[3]
[4]
[5]
S. Steenken, S. V. Jovanovic, J. Am. Chem. Soc. 1997, 119, 617.
S. Kawanishi, Y. Hiraku, Y. Antioxid. Redox Signal. 2006, 8, 1047.
J. Cadet, T. Douki , J.-L. Ravanat, Acc. Chem. Res. 2008, 41, 1075.
V. Yermilov, J. Rubio, H. Ohshima, Febs. Lett. 1995, 376, 207.
S. Obeid, N. Blatter, R. Kranaster, A. Schnur, K. Diederichs, W. Welte
and A. Marx, EMBO J., 2010, 29, 1738.
[6]
[7]
N. Suzuki, M. Yasui, N. E. Geacintov, V. Shafirovich, S. Shibutani,
Biochemistry 2005, 44, 9238-9245.
Borohydride reduction of 8-nitroguanosine
K. Kaneko, T. Akuta, T. Sawa, H. W. Kim, S. Fujii, T. Okamoto, H.
Nakayama, H. Ohigashi, A. Murakami, T. Akaike, Cancer Letters 2008,
262, 239.
8-Nitroguanosine (1.0 mg, 3.05 µmoles) was dissolved in distilled water
(1 ml) with stirring. Sodium borohydride (5.8 mg, 0.153 mmoles, 50
equiv) was added, with additional portions (50 equiv) subsequently
added every 30 mins for 90 mins. Reaction monitored by HPLC, with
samples taken every 30 mins..
[8]
I. Bhamra, P. Compagnone-Post, I. A. O’Neil, L. Iwanejko, A. D. Bates,
R. Cosstick, Nucl. Acids Res. 2012, 40, 11126.
8-cysteinyl labelled oligomers
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
[9]
a) Y. Hiraku, Environ Health Prev. Med. 2010, 15, 63; b) Y. Hiraku, K.
Sakai, E. Shibata, M. Kamijima, N. Hisanaga, N. Ma, S. Kawanishi, M.
Murata, J. Occup. Health 2014, 56, 186-196.
[21] a) A. E. Pegg, M.E. Dolan, R.C. Moschel, Prog. Nucleic Acid Res. Mol.
Biol. 1995, 51, 167; b) S. Mitra, DNA Repair 2007, 6, 1064.
[22] L. A. Kaplan, J. Am. Chem. Soc., 1964, 86, 740.
[23] a) D. W. Lamson, P. Ulrich, R. O. Hutchins, J. Org. Chem. 1973, 38,
2928; b) V. Gold, A. Y. Miri, S. R. Robinson, J. Chem. Soc., Perkin
Trans. 2 1980, 243.
[10] Y. Hoki, M. Murata, Y. Hiraku, N. Ma, A. Matsumine, A. Uchida, S.
Kawanishi, Oncol. Rep. 2007, 18, 1165.
[11] T. Sawa, M. H. Zaki, T. Okamoto, T. Akuta, Y. Tokutomi, S.
K.Mitsuyama, H. Ihara, A. Kobayashi, M. Yamamoto, S. Fujii, H.
Arimoto, T. Akaike, Nat. Chem. Biol. 2007, 3, 727.
[24] A. N. Blinnikov and N. N. Makhova, Mendeleev Commun. 1999, 9, 13.
[25] a) V. Yermilov, J. Rubio, M. Becchi, M. D. Friesen, B. Pignatelli, H.
Ohshima, Carcinogenesis, 1995, 16, 2045; b) V. Yermilov, J. Rubio, H.
Ohshima, Febs Lett. 1995, 376, 207.
[12] V. Terzić, D. Padovani, V. Balland, I. Artaud, E. Galardon, Org. Biomol.
Chem. 2014, 12, 5360.
[13] M. A. Lewis, D. Graff, J. L. Bolton, J. A. Thompson, Chem. Res. Toxicol.
1996, 9, 1368.
[26] T. Sawa, M. Tatemichi, T. Akaike, A. Barbin, H. Ohshima, Free Radic.
Biol. Med. 2006, 40, 711.
[14] W. F. Veldhuyzen, Y.-F. Lam, S. E. Rokita, Chem. Res. Toxicol. 2001,
14, 1345.
[27] S. Dick, S. E. J. Bell, K. J. Alexander, I. A. O’Neil and R. Cosstick,
Chem. Eur. J., in press, DOI: 10.1002/chem.20170179
[15] T. Li, S. E. Rokita, J. Am. Chem. Soc. 1991, 113, 7771.
[16] S. E. Rokita, J. Yang, P. Pande, W. A. Greenberg, J. Org. Chem. 1997,
62, 3010.
[28] Y. Fuchi, S. Sasaki, Org. Lett. 2014, 16, 1760.
[29] a) Y. Ishima, H. Hoshino, T. Shinagawa, K. Watanabe, T. Akaike, T.
Sawa, U. Kragh-Hansen, T. Kai, H. Watanabe, T. Maruyama, M. Otagiri,
J. Pharm. Sci. 2012, 101, 3222; b) Y. Saito, H. Taguchi, S. Fujii, T.
Sawa, E. Kida, C. Kabuto, T. Akaike, H. Arimoto, Chem. Commun.
2008, 5984.
[17] J. Reinhard, W. E. Hull, C. W. Von Der Lieth, U. Eichhorn, H. C. Kliem,
B. Kaina, M. Wiessler, J. Med. Chem., 2001, 44, 4050.
[18] P. Pande, J. Shearer, J. Yang, W. A. Greenberg, S. E. Rokita, J. Am.
Chem. Soc. 1997, 121, 6773.
[30] Similar yields were obtained when glutathione was used as the
nucleophile.
[19] a) R. Singh, U. Manjunatha, H. I. Boshoff, Y. H. Ha, P. Niyomrattanakit,
R. Ledwidge, C. S. Dowd, I. Y. Lee, P. Kim, L. Zhang, Science 2008,
322, 1392; b) R. Tiwari, G. C. Moraski, V. Krchňák, P. A. Miller, M.
Colon-Martinez, E. Herrero, A. G. Oliver, M. J. Miller, J. Am. Chem. Soc.
2013, 135, 3539.
[31] S. Nampalli, S. Kumar, Bioorg. Med. Chem. Lett. 2000, 10, 1677.
[32] N. An, A. M. Fleming, H. S. White, C. J. Burrows, Proc. Natl. Acad. Sci.
USA, 2012, 109, 11504.
[33] J. Fernández-Carneado, M. Van Gool, V. Martos, S. Castel, P. Prados,
J. De Mendoza, E. Giralt, J. Am. Chem. Soc. 2005, 127, 869.
[20] A. Sancar, Chem. Rev. 2003, 103, 2203.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705541
Chemistry - A European Journal
FULL PAPER
This article is protected by copyright. All rights reserved.