Photo-oxidation of 6-thioguanine by UVA: The formation of addition products with low molecular weight thiol compounds

Article abstract of DOI:10.1111/j.1751-1097.2010.00771.x

The thiopurine, 6-thioguanine (6-TG) is present in the DNA of patients treated with the immunosuppressant and anticancer drugs azathioprine or mercaptopurine. The skin of these patients is selectively sensitive to UVA radiation-which comprises >90% of the UV light in incident sunlight-and they suffer high rates of skin cancer. UVA irradiation of DNA 6-TG produces DNA lesions that may contribute to the development of cancer. Antioxidants can protect 6-TG against UVA but 6-TG oxidation products may undergo further reactions. We characterize some of these reactions and show that addition products are formed between UVA-irradiated 6-TG and N-acetylcysteine and other low molecular weight thiol compounds including β-mercaptoethanol, cysteine and the cysteine-containing tripeptide glutathione (GSH). GSH is also adducted to 6-TG-containing oligodeoxynucleotides in an oxygen- and UVA-dependent nucleophilic displacement reaction that involves an intermediate oxidized 6-TG, guanine sulfonate (G^SO3). These photochemical reactions of 6-TG, particularly the formation of a covalent oligodeoxynucleotide-GSH complex, suggest that crosslinking of proteins or low molecular weight thiol compounds to DNA may be a previously unrecognized hazard in sunlight-exposed cells of thiopurine-treated patients.

Full text of DOI:10.1111/j.1751-1097.2010.00771.x

Photochemistry and Photobiology, 2010, 86: 1038–1045
Photo-oxidation of 6-Thioguanine by UVA: The Formation of Addition
Products with Low Molecular Weight Thiol Compounds
Xiaolin Ren1,2, Yao-Zhong Xu2 and Peter Karran*1
1Cancer Research UK London Research Institute, Clare Hall Laboratories, South Mimms, Herts, UK
2Department of Chemistry and Analytical Sciences, The Open University, Walton Hall, Milton Keynes, UK
Received 17 March 2010, accepted 14 May 2010, DOI: 10.1111 ⁄ j.1751-1097.2010.00771.x
Its ability to absorb energy from UVA—which comprises
>90% of incident solar UV radiation—distinguishes it from
ABSTRACT
The thiopurine, 6-thioguanine (6-TG) is present in the DNA of
patients treated with the immunosuppressant and anticancer
drugs azathioprine or mercaptopurine. The skin of these patients
is selectively sensitive to UVA radiation—which comprises
>90% of the UV light in incident sunlight—and they suffer
high rates of skin cancer. UVA irradiation of DNA 6-TG
produces DNA lesions that may contribute to the development of
cancer. Antioxidants can protect 6-TG against UVA but 6-TG
oxidation products may undergo further reactions. We charac-
terize some of these reactions and show that addition products
are formed between UVA-irradiated 6-TG and N-acetylcysteine
and other low molecular weight thiol compounds including
b-mercaptoethanol, cysteine and the cysteine-containing tripeptide
glutathione (GSH). GSH is also adducted to 6-TG-containing
oligodeoxynucleotides in an oxygen- and UVA-dependent nucle-
ophilic displacement reaction that involves an intermediate
oxidized 6-TG, guanine sulfonate (GSO3). These photochemical
reactions of 6-TG, particularly the formation of a covalent
oligodeoxynucleotide–GSH complex, suggest that crosslinking
of proteins or low molecular weight thiol compounds to DNA
may be a previously unrecognized hazard in sunlight-exposed
cells of thiopurine-treated patients.
the canonical DNA bases and the photochemical reactions of
DNA 6-TG constitute a potential danger to genome integrity.
Previous investigations of the photochemistry of 6-TG indi-
cated that this hazard is compounded by the liberation of
1
reactive oxygen species (ROS), principally singlet oxygen, O2
1
(4,5), when DNA 6-TG acts as a Type II photosensitizer. O2
can damage DNA and damaged DNA bases such as 8-oxo-7,
8-dihydroguanine (6) and DNA strand breaks (7) are well-
characterized reaction products. Proteins are also vulnerable
1
to damage by O2 (8) and UVA treatment of cells containing
DNA 6-TG causes a covalent crosslinking between the
subunits of the homotrimeric replication protein PCNA (9).
6-TG itself is a target for oxidation by photochemically
1
generated O2. Our previous studies (5,10) identified oxidized
forms of 6-TG, guanine sulfinate GSO2 and guanine sulfonate
GSO3 as significant UVA-induced 6-TG photoproducts. We
suggested that GSO2 and GSO3 were formed via an initial highly
reactive intermediate, guanine sulfenate GSO. When it is
irradiated in free solution, 6-TG can also dimerize to form
guanine-6-thioguanine, GSG. Some of these reactions are
likely to be of clinical importance and it is significant that the
skin of patients taking azathioprine is particularly sensitive to
UVA but not to UVB (11).
All three oxidized forms of 6-TG are potentially reactive. In
particular, the less oxidized GSO2 is likely to be vulnerable
to further oxidation to GSO3 by photochemically generated
ROS. We previously examined the impact of the antioxidant
ascorbate on the UVA-induced oxidation of 6-TG (10). This
widely distributed antioxidant confers significant protection
against the effects of UVA radiation and it prevented the
photochemical oxidation of 6-TG. N-acetylcysteine (NAC) is a
widely prescribed antioxidant that alleviates the symptoms of
several clinical conditions (12). Here, we have examined the
effects of NAC and other antioxidants on the photochemical
reactions of 6-TG. In contrast to ascorbate which provided
unequivocal protection against oxidation, the effects of NAC
and its derivative glutathione (GSH) are more complex. While
providing some protection against oxidation, these thiol-
containing compounds also react with photoactivated 6-TG.
These reactions result in the formation of novel addition
products via nucleophilic substitution. GSH also forms an
addition product with 6-TG in an oligodeoxynucleotide in
INTRODUCTION
The effectiveness of many therapeutic agents relies on their
ability to cause potentially cytotoxic DNA damage. The
anticancer and immunosuppressive thiopurines azathioprine
and 6-mercaptopurine are no exceptions in this regard. Their
efficacy relies partly on an ability to cause the accumulation of
6-thioguanine (6-TG) in the DNA of dividing cancer or
immune effector cells, and partly on the reactivity of DNA
6-TG (1). The thiopurine can accumulate quite extensively in
DNA without having a significant biological impact. DNA
6-TG is, however, considerably more reactive than any of the
canonical DNA bases and in vivo its relatively favorable
chemical methylation by S-adenosylmethionine (2) triggers a
chain of events that involves DNA mismatch repair and
culminates in cell death (3).
DNA 6-TG also has unusual optical properties. 6-TG is a
UVA chromophore with an absorbance maximum at 340 nm.
a reaction that also depends on UVA and O2. The
UVA-mediated formation of complex DNA lesions involving
*Corresponding author email: peter.karran@cancer.org.uk (Peter Karran)
ꢀ 2010 The Authors. JournalCompilation. The AmericanSocietyof Photobiology 0031-8655/10
1038

Photochemistry and Photobiology, 2010, 86 1039
to 80% for the next 10 min and maintained at 80% for the following
5 min.
thiol compounds has implications for the well-being of
sunlight-exposed cells of patients undergoing thiopurine
therapy.
HPLC program 2: Waters C18 column (Xterra, MS, 3.5 lm,
150 · 2.1 mm). Solvent C was kept constant at 5%. Solvent A was
increased from 0% to 20% during the first 10 min, increased to 80%
for the next 10 min and to 90% over the subsequent 3 min. Solvent A
was maintained at 90% for an additional 7 min.
MATERIALS AND METHODS
HPLC program 3: Waters dC18 column (Atlantis, 3 lm,
150 · 2.1 mm). Solvent C was kept constant at 5%. Solvent A was
increased from 0% to 20% during the first 10 min, then increased to
80% for the next 10 min and maintained at this concentration for the
following 5 min.
HPLC program 4: Waters C18 column (Xterra, MS, 3.5 lm,
150 · 2.1 mm). Solvent C was kept constant at 5%. Solvent A was
increased from 1% to 10% during the first 10 min, then increased to
40% for the next 10 min and maintained at 40% for an additional 5 min.
HPLC program 5: Waters dC18 column (Atlantis, 3 lm,
150 · 2.1 mm). Solvent C was kept constant at 5% for the first
23 min. It was reduced to 0% between 23 and 30 min. Solvent A was
increased from 0% to 20% during the first 10 min, to 80% during the
next 10 min and to 90% over the subsequent 3 min. It was maintained
at 90% for the following 7 min. The flow rate was increased from 0.2
to 0.25 mL min)1 over the first 23 min and maintained at this value for
the remainder of the elution time.
Chemicals. Sodium sulfide (Na2SÆ9H2O), NAC, cysteine, mercaptoeth-
anol, 6-TG, GSH, deuterium oxide, magnesium monoperoxyphthalate
(MMPP), and rose bengal were obtained from Sigma Aldrich (Dorset,
UK). Oligodeoxyribonucleotides were from Oligos Etc. Inc. (Wilson-
ville, OR). Guanine-6-sulfonate (GSO3), guanine-6-sulfinate (GSO2) and
the disulfide of thioguanine (GSSG) were prepared according to a
published protocol (13). 6-TG 2¢-deoxyribonucleoside (dGS) was
synthesized according to Fox et al. (14). Guanine-6-thioguanine
(GSG) was prepared as described previously (5).
Synthesis and characterization of 6-(2-b-hydroxyethyl)-thioguanine:
G
MSH. To a solution of 50 mg of potassium guanine sulfonate (GSO3) in
6 mL of 0.33 N NaOH, 200 lL of 15 M b-mercaptoethanol was added.
The mixture was left at 20ꢁC for 20 min after which analysis by HPLC
indicated complete disappearance of GSO3. The solution was left at )4ꢁC
for 2 days and the precipitate that formed was collected and character-
ized by absorbance, fluorescence and by 1H-NMR. UV and fluorescence
data are shown in Table 1. 1H-NMR (DMSO-d6): 7.879 (s,1H,8-H),
6.265 (s,2H,2-NH2), 4.163 (b,1H, OH), 3.613 (t,2H,a-H) and 3.349
(t, 2H, b-H). 13CMR (DMSO-d6):159.6 (4-C), 158.3 (2-C), 152.4 (8-C),
139.5 (6-C), 123.7 (5-C), 58.59 (a-C) and 41.13 (b-C).
UVA irradiation. UVA irradiation was carried out at a dose rate of
a UVH 254 lamp (UV Light Technology,
0.1 kJ m)2 s)1 using
Birmingham, UK. Wavelength range 320–400 nm, kmax 365 nm).
Visible light irradiation. Visible light was delivered from a 1000 W
ozone-free xenon bulb emitting wavelengths between 240 and
1100 nm. Wavelengths less than 400 nm were removed by passage
through a polystyrene film.
Synthesis and characterization of glutathione-6-thioguanine: GGSH
.
To a solution of 10 mg of potassium guanine sulfonate in 3 mL of
2 N NaOH, 0.6 g GSH was added and the mixture incubated
overnight at 20ꢁC. The pH was then adjusted to 8.0 by the addition
of acetic acid and the total volume adjusted to 3.9 mL (equivalent
to 10 mM original sulfonate). Reverse phase HPLC analysis of the
product showed that all the sulfonate was converted to a novel
product GGSH. This was purified by HPLC on an Atlantis dC18
column 5 lm particle size (4.6 · 250 mm) eluted with metha-
nol ⁄ H2O ⁄ 0.1 M ammonium formate (pH 4.8) (10:9:1). Mass spectra
data were recorded by electrospray ionization mass spectrometry
RESULTS
Reactivity of oxidized 6-TG
Both GSO2 and GSO3 react with Na2S (Fig. 1). When a mixture
of GSO2 and GSO3 was treated with Na2S at 25ꢁC, GSO2 was
rapidly converted to 6-TG. Significant reaction had occurred
at 10 min and complete conversion of GSO2 to 6-TG was
observed following overnight incubation. Most of the GSO3
remains unchanged. This is compatible with our previous
observations that its conversion to 6-TG takes several days to
proceed to completion.
using
a Fisons Instruments (VG BioTech, Altrincham, UK).
Fluorescence emission and excitation spectra are obtained using a
FLUOROMAX-P spectrofluorometer (Jobin Yvon Horiba, Inc.).
UV–visible spectroscopy was performed using UVIKON-XL (Bio-
TeK Instruments) with Lab junior software.
Preparation of GSH adducted 13mer oligonucleotide. A 13 mer
oligonucleotide with the sequence 5¢-CTGCCAXGCACCA-3¢
(X = 6-TG) was oxidized under mild conditions using an established
protocol (5) to convert the 6-TG to GSO3. To the oxidized oligomer
(2.5OD) in 1 mL of 50 mM phosphate buffer pH 6.9, GSH was added
to a final concentration of 10 mM and the mixture incubated overnight
at 20ꢁC. Unreacted GSH was removed by a MicroSpin G-25 Spin
column. The product was characterized by reverse phase HPLC
analysis following acid depurination (0.1 N HCl, 70ꢁC, 60 min) or
digestion to deoxyribonucleosides by nuclease P1 and alkaline phos-
phatase as described previously (5).
Reverse phase HPLC. Reverse phase HPLC analysis was performed
using a Waters 2695 Alliance system equipped with photodiode array
and Waters 474 dual monochromator fluorescence detector. Solvent A
was methanol; Solvent B was water; Solvent C was 10 mM KH2PO4,
pH 6.5. The flow rate was constant at 0.2 mL min)1 unless otherwise
indicated.
Addition reactions of 6-TG photoproducts
N-acetylcysteine addition. We previously reported that the
antioxidant ascorbate protected 6-TG against photochemical
oxidation (10). A second antioxidant, NAC, also conferred
significant protection against photochemical destruction of
GSO2
GSO3
0.1
0.1
0min
0.1
HPLC program 1: Waters C18 column (Xterra, MS, 3.5 lm,
150 · 2.1 mm). Solvent C was kept constant at 20%. Solvent A was
increased from 0% to 20% during the first 10 min, then increased
10min
6-TG
16h
Table 1. Spectral properties of 6-thioguanine-adducts.
UV
Fluorescence
MS
0
5
0
5
0
5
Retention (min)
kmax
kmin
kex
kem
Calc.
Found
Figure 1. Reactions of GSO3 and GSO2 with Na2S. A mixture of GSO3
and GSO2 in aqueous solution (both 0.1 mM) was treated with 0.1 mM
Na2S at 25ꢁC for 10 min or overnight as indicated. Products were
separated by HPLC (Program 1) and eluates monitored at 320 nm.
Elution times: GSO3 2.5 min; GSO2 3.1 min; 6-TG 5.2 min.
GNAC
GMSH
GGSH
311.2
310.0
311.2
276.1
275.2
275.3
310
320
320
390
390
394
297.31
210.13
441.44
297.12
210.24
441.83

1040 Xiaolin Ren et al.
6-TG. The presence of NAC during irradiation significantly
increased the recovery of unchanged 6-TG. Figure 2 shows
that irradiation of a mixture of 6-TG and NAC also generated
a novel photoproduct that could be detected by its absorbance
at 320 nm (shown arrowed in Fig. 2) and which eluted at
around 6 min on HPLC. This compound was also highly
fluorescent (Table 1). It was identified by MS as 2-acetamido-
3-(2-amino-9H-purine-6-ylthio)propanoic acid (GNAC), an
addition product between NAC and a 6-TG photoproduct.
The MS, UV and fluorescence characteristics are presented in
Table 1 and the structure is shown in Fig. 2.
strated directly by performing the reaction in two stages. In the
first stage, 6-TG was irradiated to convert more than 90% of it
to photoproducts. When MSH was subsequently added to the
irradiated 6-TG and the products separated by HPLC,
significant amounts of 6-TG were recovered along with
MSH-TG (Fig. 3A, bottom panel). There was a concomitant
reduction in the recovery of both GSG and GSO2. The MS, UV
and fluorescence characteristics of this compound are given in
Table 1 and its structure is shown in Fig. 3B.
Cysteine: addition and rearrangement. Cysteine also protected
6-TG against UVA-mediated oxidation. Inclusion of cysteine
(0–100 mM) during UVA irradiation of 6-TG (0.1 mM)
resulted in a concentration-dependent enhancement of 6-TG
recovery at the cost of reduced formation of GSO2, GSO3 and
GSG (Fig. 4A). At cysteine concentrations of 50 mM or above,
there was no detectable formation of GSG and the yield of
GSO2 was reduced by more than 80%. As with NAC and
MSH, cysteine reacted to form a novel photoproduct with a
longer retention time than 6-TG (Fig. 4B).
b-mercaptoethanol (MSH): addition involving oxidized 6-TG
intermediate. MSH had a similar effect to NAC on photo-
oxidation of 6-TG. MSH also reacted with irradiated 6-TG to
produce a new species that eluted late on HPLC (Fig. 3A).
This highly fluorescent compound was identified by mass
spectrometry as 6-(2-b-hydroxyethyl)-thioguanine, an addition
product of 6-TG and MSH (GMSH). GMSH was formed by
reaction between MSH and oxidized 6-TG. This was demon-
O
H
CH3
N
C
.015
OH
N
6-TG
.06
C
O
CH
.04
.02
0
6-TG
CH2
S
.04
.02
0
N
GSO3
NH2
N
H
N
2-acetamido-3-(2-amino-9H-purin-6-
0
ylthio)propanoic acid
0
4
8
0
4
8
0
4
8
(GNAC
)
Figure 2. Photochemical formation of a 6-TG-N-acetylcysteine addition product. An aqueous solution of 6-TG (0.1 mM) was unirradiated (left
panel) or UVA irradiated (30 kJ m)2) in the absence (middle panel) or presence (right panel) of N-acetylcysteine (10 mM). 6-TG products were
separated by HPLC (Program 1) immediately after irradiation and eluates monitored at 320 nm.
No treatment
B
A
6-TG
O H
C H
C H
S
GSG
2
2
GSO2
N
N
UVA
N
H
N
N H
2
UVA + MSH
6-(2β-hydroxylethyl)-thioguanine
(GMSH
UVA then MSH
)
0
10
20
Retention (min)
Figure 3. Photochemical reaction of 6-TG with 2-mercaptoethanol. (A) An aqueous solution of 6-TG (0.1 mM), unirradiated (—) or UVA
irradiated (30 kJ m)2) in the absence (– – ) or presence of 10 mM (
) 2-mercaptoethanol. In the bottom panel (– –) 10 mM 2-mercaptoethanol
24
Æ
Æ Æ Æ Æ Æ
Æ Æ
was added after irradiation and the mixture analyzed immediately. Products were separated by HPLC (Program 2) and absorbance monitoring at
320 nm. The novel reaction product is arrowed. (B) Suggested structure of the novel photochemical addition product.

Photochemistry and Photobiology, 2010, 86 1041
A
3
4
3
5
3
2
1
2
1
1
0
20 40 60 80 100
Cysteine (mM)
0
20 40 60 80 100
Cysteine (mM)
0
20 40 60 80 100
Cysteine (mM)
C
B
O
HO
NH
GSO2
C
GSG
CH
SH
6-TG
CH2
N
N
N
H
N
N H
2
0
4
8
12
16
Retention (min)
Figure 4. Photochemical reaction of 6-TG with cysteine. (A) Protection against oxidation. 6-TG was UVA irradiated with 30 kJ m)2 UVA in the
absence or presence of increasing concentrations of cysteine. Photoproducts were analyzed by HPLC (Program 3) and eluates monitored at 320 nm.
The relative amounts of photoproducts were determined from the integrated peak areas. (B) Formation of a novel photoproduct. 6-TG (0.1 mM)
was irradiated 30 kJ m)2 in the absence (upper panel) or presence of 100 mM cysteine. Products were separated by HPLC (Program 3) with
detection by A320. The novel photoproduct is arrowed. (C) Structure of the N-bonded 6-TG-cysteine photoproduct.
The nature of the novel photoproduct was investigated in
more detail. Like MSH and NAC, cysteine reacted with
preformed GSG. Immediately following the addition of cyste-
ine to GSG, a novel peak with a retention time of 11 min was
apparent following HPLC analysis. A minor peak was
observed to elute slightly later at 13.5 min. By 3 h, most of
the GSG had disappeared and the two novel peaks were of
approximately equal magnitude. After overnight incubation,
however, the later eluting species predominated. A small
amount of 6-TG was formed but this did not change
significantly over time. An example of these changes is shown
in Fig. S1. To explain these findings, we propose that the
initial product of nucleophilic displacement, an S-bonded
cysteine–guanine, is slowly converted via an internal rear-
rangement of the type previously described (15) to the
N-bonded product shown in Fig. 4C. The proposed reaction
mechanism is outlined in Fig. S1.
Glutathione: addition to oxidized 6-TG and oxidized DNA 6-TG.
UVA irradiation of 6-TG in the presence of the cysteine-
containing tripeptide (glu-cys-gly; GSH) also generated a
photoaddition product (shown arrowed in Fig. 5A). The
6-TG
21kJ/m2
+ GSH
.03
.02
.01
A
B
C
670
470
270
70
GSO2
NH2
CH
O
C
CH2
OH
NH
CH2
CH2
HO
C
NH
CH
CH2
S
C
O
C
O
0
O
.02
42kJ/m2
+ GSH
N
N
.01
0
NH2
N
H
N
0
5
10
15
20
25
5
10
15
UVA Dose ( kJ m-2)
Retention (min)
Figure 5. Formation of a 6-TG-glutathione adduct. (A) A mixture of 0.1 mM 6-TG with GSH (10 mM) in water was UVA irradiated at the doses
shown and products were separated by HPLC (Program 1). The novel photoproduct (GGSH) is arrowed. (B) Oxygen dependence. A mixture of 6-
TG (0.1 mM) and GSH (10 mM) in water or D2O was irradiated with the UVA doses shown. Immediately after irradiation, photoproducts were
separated by HPLC (Program 4). The GGSH adduct was detected by absorbance at 310 nm and quantified by comparison to a standard curve based
on the authentic compound. D2O (r), H2O ( ), H2O with continuous helium purging (d). (C) The structure of GGSH
.

1042 Xiaolin Ren et al.
formation of this photoproduct was dependent on oxygen and
it was not observed when irradiation was carried out under
helium under conditions of rigorous oxygen exclusion. 1O2 was
involved in its formation and irradiation in the presence of
D2O increased photoproduct yield (Fig. 5B). The same com-
pound was generated by irradiation of 6-TG with visible light
in the presence of rose bengal (data not shown), confirming the
involvement of 1O2 in its formation. The absorbance and
fluorescence characteristics and MS of the addition product
identified it as 3-(2-amino-9H-purin-6-ylthio)-1-(carboxymeth-
7 h at 25ꢁC. In contrast, GGSH was not observed when the
generally more reactive GSO2 was incubated with GSH under
the same conditions. This is illustrated in Fig. 6C which shows
that when an equimolar (0.1 mM) mixture of GSO3 and GSO2
was treated overnight at 25ꢁC with a 50-fold excess of GSH,
GSO3 disappeared with concomitant formation of GGSH. The
amount of GSO2 recovered from the mixture was not signif-
icantly changed, however.
As a first step to examining whether GSH might react with
DNA 6-TG in cells exposed to UVA, we investigated
crosslinking of GSH to a 6-TG-containing oligonucleotide.
A [32P] end-labeled 13mer containing a single 6-TG: 5¢-
CTGCCAXGCACCA-3¢ (X = 6-TG) was first treated with
MMPP to convert 6-TG stoichiometrically to GSO3. The
oxidized oligomer was then incubated overnight with a large
excess of GSH and the products separated by PAGE.
Reaction products were quantified and Fig. 7A (and Fig. S2)
shows that under these reaction conditions, about 75% of the
input oligomer was converted to a form that migrated with a
mobility similar to that of a 14mer, consistent with an
increase in molecular weight associated with adducted GSH.
The structure of the adduct was confirmed directly. Follow-
ing acid depurination of the adducted oligomer, RP-HPLC
analysis revealed a peak of material eluting slightly later than
adenine. This material coeluted with authentic GGSH formed
from the reaction between synthetic GSO3 and GSH
(Fig. 7B). Similarly, when the same 6-TG 13mer was irradi-
ylamino)-1-oxopropan-2-ylamino)-5-oxopentanoic
acid
(GGSH, Fig. 5C). The authentic compound was synthesized
as described in Materials and Methods and a full character-
ization is presented in Table 1. The absence of a thione
structure causes a hypochromatic shift (ca 31 nm) and GGSH
has a kmax = 311 nm compared to 6-TG (kmax = 342 nm).
We note that its fluorescence might provide a particularly
sensitive means of GGSH detection.
The mechanism of formation of GGSH was investigated in
more detail. As 6-TG is unreactive toward GSH in the absence
1
of O2 and both GSO2 and GSO3 are products of O2-mediated
oxidation of 6-TG, we examined whether GSH reacted with
either GSO2 or GSO3. GGSH was formed by the reaction
between GSH and GSO3 but not GSO2. When GSO3 (0.1 mM)
was incubated for 7 h at 25ꢁC with increasing amounts of
GSH, it was converted to a single product that was detectable
by its A320 (Fig. 6A). The product had the same fluorescence
characteristics and the same retention time as GGSH on
RP-HPLC. Figure 6B shows that the half-time for the forma-
tion of GGSH from 0.1 mM GSO3 and 10 mM GSH is around
ated with UVA and digested to deoxynucleosides, dGSO2
,
dGSO3 and dGSG were all detected by their absorbance at
310 nm (Fig. 7Cii). Following overnight incubation with
GSO3
A
GSO3 GSO2
Adduct
no GSH
C
0.15
100:1
50:1
25:1
0.15
10:1
50:1
no GSH
5
10
15
Retention (min)
adduct
1000
800
600
400
B
0.15
50:1
400
800
1200
0
5
10
Incubation (min)
Retention (min)
Figure 6. Formation of GGSH from GSO3. (A) GSH concentration dependence. GSO3 (0.1 mM) was treated for 7 h at 25ꢁC with GSH at the molar
ratios shown. Reaction products were separated by HPLC (Program 1) and eluates monitored at 320 nm. The elution position (9.5 min) of GGSH
(Adduct) is arrowed. (B) Kinetics of formation. GSO3 (0.1 mM) was treated with 10 mM GSH. The formation of the addition product was
monitored at the times indicated by HPLC and quantified by comparison to a standard curve constructed from authentic GGSH. GSO3 was detected
and quantified by A325. GSO3 (d), GGSH
(
). (C) GSO3 and not GSO2 as an intermediate in GGSH formation. An equimolar mixture of GSO3 and
GSO2 (0.1 mM) was treated for 1 h (middle panel) or overnight (top and bottom panels) at 25ꢁC with GSH at the GSH:6-TG ratio indicated.
Products were separated by HPLC (Program 1) and the eluate monitored at 320 nm.

Photochemistry and Photobiology, 2010, 86 1043
C
i
B
dA
dC
dG
A260
A260
T/dGS
ii
A
A260
dGSO2
dGSG
A310
dGSO3
iii
GGSH
dGS
A310
A310
-
GSH
+
dGSO2
dGGSH
12
Retention (min)
iv
A310
5
10
15
20
Retention (min)
Figure 7. Glutathione binding to an oligonucleotide. (A) Adduct formation. 10 pmol of [32P]-end labeled 13mer oligonucleotide containing a single
6-TG (see Materials and Methods) was oxidized by treating with 50 nmol MMPP for 5 h. 8 pmol of the oxidized oligonucleotide was incubated
with GSH (1.25 lmol) overnight at 20ꢁC. Products were analyzed by SDS PAGE on a 20% sequencing gel. (B) Base analysis. The GSH adducted
13mer was acid depurinated and the released purines were analyzed by HPLC (Program 4). The region of the eluate that contained adenine is
shown. Top: hydrolysate of untreated 6-TG oligonucleotide. Middle: hydrolysate of GSH adducted oligonucleotide (arrowed in A). Bottom:
authentic GGSH prepared as described in Materials and Methods. (C) Deoxynucleoside analysis. 13mer oligonucloetides were digested with
nuclease P1 and alkaline phosphatase. Deoxynucleosides were separated by HPLC. (i) Untreated 13mer. Thymidine and 2-deoxy-6-thioguanine
(dGS) coelute. (ii) UVA irradiated (18 kJ m)2). (iii) As (ii) but incubated further overnight with 10 mM GSH. (iv) The product of overnight
incubation of a mixture of dGSO3 (0.1 mM) and GSH (10 mM).
GSH, dGSO2 remained unchanged whereas dGSO3 and dGSG
were replaced by a single novel species with a retention time
indistinguishable from that of the product of overnight
reaction between dGSO3 and GSH (Fig. 7Ciii,iv). These
findings are also consistent with the formation of a covalent
dGGSH adduct in the 13mer.
The reaction of GSG with GSH was confirmed in a separate
experiment in which authentic GSG, synthesized as described
previously (5), was incubated with GSH. This regenerated
6-TG and a product with the characteristic fluorescence and
retention time of GGSH (Fig. 8A, arrowed). In contrast, no
novel products were generated when G-SS-G, in which the two
guanines are joined by a conventional disulfide linkage, was
combined with GSH. In this case, GSH reduced G-SS-G to
6-TG as the single detectable product (Fig. 8B).
(18) and their skin is selectively sensitive to UVA radiation
(11). The photochemical reactions of 6-TG generate reactive
oxygen in the form of 1O2, a known source of DNA and
protein damage. We have previously demonstrated the UVA-
mediated oxidation of 6-TG to GSO2 and GSO3 and GSG
(5,10). GSO2 and GSO3 are also formed when DNA containing
6-TG is exposed to UVA. Both are powerful blocks to DNA
replication and transcription (5,19) and the latter is mutagenic
in Escherichia coli (20). The present study confirms that both
DNA photoproducts can undergo further reaction. GSO2 can
be oxidized and both GSO2 and GSO3 are converted back to
6-TG by Na2S. The reaction of GSO2 with Na2S is more rapid
than that of GSO3. Overall, our findings demonstrate a
significant susceptibility of these 6-TG photoproducts to either
reduction or the formation of addition compounds by nucle-
ophilic displacement reactions involving thiol compounds.
Thiol compounds are highly abundant in living cells and the
potential of oxidized forms of DNA 6-TG to react with
cellular constituents has implications for the biological effects
of combined DNA 6-TG.
DISCUSSION
Understanding the reactions of 6-TG is important because
patients undergoing treatment with the immunosuppressive
and anticancer agents azathioprine and 6-MP incorporate
measurable levels of this thiopurine into their DNA (16,17).
6-TG is more reactive than canonical DNA bases and the
UVA-mediated photochemical oxidation of DNA 6-TG is of
particular significance. Patients taking azathioprine have a
dramatically increased susceptibility to sunlight-related cancer
The reactions of 6-TG that we have investigated are
summarized in Fig. 9. The starting point for this study was
our previous observation that UVA irradiation of 6-TG
1
generates O2 which oxidizes it to GSO2 and GSO3. Ascorbate,
an acknowledged antioxidant, protects 6-TG against this
photo-oxidation (10). The present study indicates that another

1044 Xiaolin Ren et al.
G-SS-G
GSG
B
A
6-TG
GSH
6-TG
5
10
15
20
Retention (min)
0
5
10
15
Retention (min)
Figure 8. Reactions of dimeric 6-TG with GSH. (A) Guanine-6-thioguanine (G-S-G). GSG (0.1 mM in H2O) untreated (left panel) or immediately
after the addition of 10 mM GSH at 25ꢁC, was analyzed by HPLC (Program 1). Eluates were monitored at 320 nm. (B) GSSG (0.1 mM in H2O) was
treated and analyzed as in (A) above.
UVA
O
OH
N
OH
S
S
N
S
S
N
N
1O2
N
N
NH
NH2
N
NH
NH2
[O]
Na2S
N
N
H
N
H
NH2
NH2
N
N
H
N
H
N
1
1
3
2
6-TG
6-TG
[O]
6-TG
OH
NH
N
HN
N
N
O
O
S
N
α
β
RS-H
N
NH
NH2
N
S
N
H
N
N
NH2
GSG
6
H2N
4
SR
RS-H
RS-H
N
N
NH2
N
H
N
+
1
5
Figure 9. Reactions of 6-TG and its oxidation products. UVA irradiation of 6-TG
sulfenate (GSO . GSO can undergo further oxidation to guanine sulfinate (GSO2
generates 1O2 that oxidizes 6-TG to the unstable guanine
that can be rapidly converted back to 6-TG by Na2S or
)
)
. Reduced thiols (RS-H) can form addition products
further oxidized to guanine sulfonate (GSO3
)
with either GSO or GSO3. In the latter case
via a nucleophilic displacement of the SO3 group. GSO can also react with unchanged 6-TG to form guanine-6-guanine . This can also react with
reduced thiols to form an addition product and regenerate 6-TG.
widely used antioxidant, NAC, does not simply protect 6-TG
against oxidation and when NAC is present during UVA
irradiation of 6-TG, it reacts to generate a novel addition
product (GNAC) that we have identified as 2-acetamido-3-(2-
amino-9H-purine-6-ylthio)propanoic acid. The formation of
addition products is a characteristic feature of the interaction
between oxidized 6-TG and low molecular weight thiol com-
pounds and similar addition reactions occurred with MSH and
cysteine. Our findings suggest that photoactivated DNA 6-TG
might be reactive with abundant intracellular thiol compounds.
Consistent with this possibility, we observed the formation of
an addition product, GGSH, between the tripeptide GSH and 6-
TG. We were able to demonstrate the formation of GGSH
following UVA irradiation of a 6-TG-containing oligonucleo-
1
tide and show that it involves O2, a known product of 6-TG
photoactivation. This observation has significant implications
for the reactions of DNA 6-TG in a clinical setting. The DNA of
patients taking azathioprine contains a significant amount of 6-
TG (around 0.02% of guanine can be replaced by 6-TG [4],
corresponding to 105–106 6-TGs per genome). The intracellular
concentration of GSH in, for example human skin cells, can
also be very high—up to 10 mM (21).

Photochemistry and Photobiology, 2010, 86 1045
Our findings indicate that GSO2 is generally more easily
reduced to 6-TG than GSO3 whereas GSO3 is considerably more
susceptible to displacement by thiols. These properties are
illustrated by the relative ease of conversion from GSO2 to 6-TG
by Na2S and the preferential formation of GGSH from GSO3 in a
GSH ⁄ GSO2 ⁄ GSO3 mixture. GSO2 is rapidly converted to 6-TG
whereaswithGSO3,GGSH ismostlikelyproducedbynucleophilic
attack of the GSH thiol group on the C6 atom of GSO3. A similar
nucleophilic displacement via attack on the C6 atom probably
occurs with GSG and reaction with GSH again generates the
GGSH addition product. With the disulfide G-SS-G, however,
GSH acts simply as a reducing agent to split the disulfide bond
and to regenerate 6-TG. The photochemical reactions of 6-TG
and thiol compounds are summarized in Fig. 9.
2. Swann, P. F., T. R. Waters, D. C. Moulton, Y.-Z. Xu, M. Edwards
and R. Mace (1996) Role of postreplicative DNA mismatch repair
in the cytotoxic action of thioguanine. Science 273, 1109–1111.
3. Karran, P. (2007) Thiopurines, DNA damage, DNA repair and
therapy-related cancer. Br. Med. Bull. 79-80, 153–170.
4. O’Donovan, P., C. Perrett, X. Zhang, B. Montaner, Y.-Z. Xu, C.
A. Harwood, J. M. McGregor, S. L. Walker, F. Hanaoka and P.
Karran (2005) Azathioprine and UVA light generate mutagenic
oxidative DNA damage. Science 309, 1871–1874.
5. Zhang, X., G. Jeffs, X. Ren, P. O’Donovan, B. Montaner, C. M.
Perrett, P. Karran and Y.-Z. Xu (2006) Novel DNA lesions gen-
erated by the interaction between therapeutic thiopurines and
UVA light. DNA Repair 6, 344–354.
6. Cadet, J., T. Douki, J. L. Ravanat and P. D. Mascio (2009)
Sensitized formation of oxidatively generated damage to cellu-
lar DNA by UVA radiation. Photochem. Photobiol. Sci. 8, 903–
911.
Our previous studies identified DNA GSO2 and GSO3 as
potentially biologically significant hazards produced by the
interaction between DNA 6-TG and UVA. They also provided
evidence that this photochemical interaction causes DNA
double-strand breaks, particularly in the vicinity of replication
forks (22). The results presented here indicate that another
hazard may arise as a result of the ability of DNA 6-TG
photoproducts to undergo further reaction. Low molecular
weight thiol compounds are extremely abundant in cells where
they comprise an important part of the defense against
oxidative stress. Reduced cysteine residues are present in
proteins and active cysteine residues are a key feature of many
enzymes. Reaction with a protein )SH group would generate a
covalent DNA–protein crosslink. If the cysteine were essential
for enzymatic activity, this attachment to DNA would
inactivate it. The easy formation of simple addition products
between biological thiols and photoactivated 6-TG suggests
that bulky DNA addition products may be a previously
unrecognized photochemical hazard. Covalently adducted low
molecular weight thiols or proteins might be difficult for the
cell’s replication, transcription and DNA repair machinery to
deal with. The formation of DNA–protein crosslinks in
particular may be a significant contributor to the highly toxic
interaction between DNA 6-TG and UVA.
7. Kumar, S. S., A. Ghosh, T. P. Devasagayam and P. S. Chauhan
(2000) Effect of vanillin on methylene blue plus light-induced
single-strand breaks in plasmid pBR322 DNA. Mutat. Res. 469,
207–214.
8. Shen, H. R., J. D. Spikes, P. Kopecekova and J. Kopecek (1996)
Photodynamic crosslinking of proteins. II. Photocrosslinking of a
model protein-ribonuclease A. J. Photochem. Photobiol. B, Biol.
35, 213–219.
9. Montaner, B., P. O’Donovan, O. Reelfs, C. M. Perrett, X. Zhang,
Y.-Z. Xu, X. Ren, P. Macpherson, D. Frith and P. Karran (2007)
Reactive oxygen-mediated damage to a human DNA replication
and repair protein. EMBO Rep. 8, 1074–1079.
10. Ren, X., F. Li, G. Jeffs, X. Zhang, Y.-Z. Xu and P. Karran (2010)
Guanine sulphinate is a major stable product of photochemical
oxidation of DNA 6-thioguanine by UVA irradiation. Nucleic
Acids Res. 38, 1832–1840.
11. Perrett, C. M., S. L. Walker, P. O’Donovan, J. Warwick, C. A.
Harwood, P. Karran and J. McGregor (2008) Azathioprine
treatment sensitizes human skin to ultraviolet A radiation. Br. J.
Dermatol. 159, 198–204.
12. Millea, P. J. (2009) N-acetylcysteine: Multiple clinical applica-
tions. Am. Fam. Physician 80, 265–269.
13. Doerr, I. L., I. Wempen, D. A. Clarke and J. J. Fox (1961)
Thiation of nucleosides. III. Oxidation of 6-mercaptopurines.
J. Org. Chem. 26, 3401–3409.
14. Fox, J. J., I. Wempen, A. Hampton and I. L. Doerr (1958)
Thiation of nucleosides. I. Synthesis of 2-amino-6-mercapto-9-
dihydrofuranosylpurine (thioguanosine) and related purine nu-
cleosides. J. Am. Chem. Soc. 80, 1669–1675.
15. Xu, Y.-Z. (1998) 6-Methylsulphoxypurine used for site-specific
and chemical cross-linking with cysteine and its peptide. Tetra-
hedron 54, 187–196.
16. Warren, D. J., A. Andersen and L. Slordal (1995) Quantitation of
6-thioguanine residues in peripheral blood leukocyte DNA
obtained from patients receiving 6-mercaptopurine-based main-
tenance therapy. Cancer Res. 55, 1670–1674.
Acknowledgements—This work was supported by Cancer Research
UK. We are grateful to Dr. Giuseppe Trigiante for NMR determi-
nations and to members of the EPSRC Mass Spectrometry Service
Centre, University of Swansea, for recording some of the mass
spectra.
17. Cuffari, C., D. Y. Li, J. Mahoney, J. Y. Barnes and T. M. Bayless
(2004) Peripheral blood mononuclear cell DNA 6-thioguanine
metabolite levels correlate with decreased interferon-c production
in patients with Crohn’s disease on AZA therapy. Dig. Dis. Sci. 49,
133–137.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online
version of this article:
Figure S1. 6-TG-cysteine addition product.
Figure S2. Glutathione binding to an oligonucleotide.
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting information sup-
plied by the authors. Any queries (other than missing material)
should be directed to the corresponding author for the article.
18. Penn, I. (2000) Post-transplant malignancy: The role of immu-
nosuppression. Drug Saf. 23, 101–113.
19. Brem, R., F. Li and P. Karran (2008) Reactive oxygen species
generated by thiopurine ⁄ UVA cause irreparable transcription-
blocking DNA lesions. Nucleic Acids Res. 37, 1951–1961.
20. Yuan, B. and Y. Wang (2008) Mutagenic and cytotoxic properties
of 6-thioguanine, S6-methylthioguanine and guanine-S6-sulfonic
acid. J. Biol. Chem. 283, 23665–23670.
21. Eiberger, W., B. Volkmer, R. Amouroux, C. Dherin, J. P. Radi-
cella and B. Epe (2008) Oxidative stress impairs the repair of
oxidative DNA base modifications in human skin fibroblasts and
melanoma cells. DNA Repair 7, 912–921.
REFERENCES
22. Brem, R., F. Li, B. Montaner, O. Reelfs and P. Karran (2010)
DNA breakage and cell cycle checkpoint abrogation induced by
a therapeutic thiopurine and UVA radiation. Oncogene. (In
press doi:10.1038/onc.2010.140)
1. Karran, P. and N. Attard (2008) Thiopurines in current medical
practice: Molecular mechanisms and contributions to therapy-
related cancer. Nat. Rev. Cancer 8, 24–36.