N-hydroxy-4-(4-chlorophenyl)thiazole-2(3H)-thione as a photochemical hydroxyl-radical source: Photochemistry and oxidative damage of DNA (strand breaks) and 2′-deoxyguanosine (8-oxodG formation)

Article abstract of DOI:10.1562/0031-8655(2000)072<0619:NHCTHT>2.0.CO;2

On irradiation of N-hydroxythiazole-2(3H)-thione 3 at 300 nm, the photoproducts disulfide 4, bisthiazole 5 and thiazole 6 are formed. During this photolysis, hydroxyl radicals are released, which have been detected by spin trapping with 5,5-dimethyl-1-pyrroline N-oxide (DMPO), coupled with electron paramagnetic resonance spectroscopy. In the presence of supercoiled pBR322 DNA, irradiation of thiazolethione 3 induces strand breaks through the photogenerated hydroxyl-radicals, as confirmed by control experiment with the hydroxyl-radical scavenger isopropanol. Singlet oxygen appears not to be involved, as attested by the lack of a D₂O isotope effect. During the photoreaction of thiazolethione 3 in the presence of 2′-deoxyguanosine (dG), the latter is photooxidized (ca 10% conversion after 2 h of irradiation) to the 7,8-dihydro-8-oxo-2′-deoxyguanosine as the main oxidation product. The dG conversion levels off after complete consumption of thiazolethione 3 and is suppressed by the addition of the hydroxyl-radical scavenger 2,6-di-tert-butylcresol or DMPO. Since the photoproducts 4-6 are ineffective as sensitizers for the photooxidation of dG and DNA, the hydroxyl radicals released in the photolysis of thiazolethione 3 are the oxidizing species of DNA and dG. These results suggest that the thiazolethione 3 may serve as a novel and effective photochemical hydroxyl-radical source for photobiological studies.

Full text of DOI:10.1562/0031-8655(2000)072<0619:NHCTHT>2.0.CO;2

-Hydroxy-4-(4-chlorophenyl)thiazole-2(3 )-thione as a Photochemical Hydroxyl-
N
H
radical Source: Photochemistry and Oxidative Damage of DNA (Strand Breaks)
and 2′-Deoxyguanosine (8-oxodG Formation)¶
Author(s): Waldemar Adam, Jens Hartung, Hideki Okamoto, Chantu R. Saha-Möller, and Kristina Špehar
Source: Photochemistry and Photobiology, 72(5):619-624.
Published By: American Society for Photobiology
DOI: http://dx.doi.org/10.1562/0031-8655(2000)072<0619:NHCTHT>2.0.CO;2
URL: http://www.bioone.org/doi/full/10.1562/0031-8655%282000%29072%3C0619%3ANHCTHT
%3E2.0.CO%3B2
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.

Photochemistry and Photobiology, 2000, 72(5): 619–624
N-Hydroxy-4-(4-chlorophenyl)thiazole-2(3H)-thione as a Photochemical
Hydroxyl-radical Source: Photochemistry and Oxidative Damage of DNA
(Strand Breaks) and 2؅-Deoxyguanosine (8-oxodG Formation)¶
ˇ
Waldemar Adam*, Jens Hartung, Hideki Okamoto, Chantu R. Saha-Mo¨ller and Kristina Spehar
Institut fu¨r Organische Chemie, Universita¨t Wu¨rzburg, Wu¨rzburg, Germany
Received 14 June 2000; accepted 17 August 2000
ABSTRACT
cals under mild conditions, e.g. in aqueous media at pH 7
on exposure to UVA light (photo-Fenton reagents [3,4]), are
needed to assess radical-induced oxidative damage of bio-
logical materials. Some photochemical sources of oxyl rad-
icals, which display DNA oxidizing activity (strand breaks
and base modification under relatively mild conditions),
have been reported during the last decade (3–13).
On irradiation of N-hydroxythiazole-2(3H)-thione 3 at
300 nm, the photoproducts disulfide 4, bisthiazole 5 and
thiazole 6 are formed. During this photolysis, hydroxyl
radicals are released, which have been detected by spin
trapping with 5,5-dimethyl-1-pyrroline N-oxide (DMPO),
coupled with electron paramagnetic resonance spectros-
copy. In the presence of supercoiled pBR322 DNA, ir-
radiation of thiazolethione 3 induces strand breaks
through the photogenerated hydroxyl-radicals, as con-
firmed by control experiment with the hydroxyl-radical
scavenger isopropanol. Singlet oxygen appears not to be
involved, as attested by the lack of a D2O isotope effect.
During the photoreaction of thiazolethione 3 in the pres-
N-hydroxy- and N-alkoxypyridinethiones 1 (Chart 1) have
served as convenient oxyl radical sources to assess oxidative
damage (6,7); indeed, the molecular mechanism of their pho-
tooxidative DNA damage has been studied thoroughly (14–
17). Although pyridinethiones 1 are readily available and
serve as efficient oxyl-radical sources, their DNA damage is
complicated: not only do the photoreleased oxyl radicals
cause strand breaks and base modifications, but the photo-
products of the pyridinethiones display significant photosen-
sitizing DNA damage (14). Recently, N-hydroxy-2(1H)-pyr-
idone 2 has been reported as a more selective photochemical
hydroxyl-radical source for oxidative DNA modification
(12,13); however, its efficacy for DNA photooxidation was
low (13). Therefore, it is highly desirable to develop a se-
lective and effective photochemical source of oxyl radicals
for the elucidation of the DNA damage caused by oxyl rad-
icals, but without the undesirable photosensitizing activity
by its photoproducts.
N-hydroxy- or N-alkoxythiazole-2(3H)-thione derivatives
are well known to generate oxyl radicals for synthetic pur-
poses (18–22). In this context, it has recently been reported
that the modified thiazolethione chromophore 3 serves as an
improved and clean photochemical precursor for oxyl radical
(21,22). These facts encouraged us to examine the DNA-
photooxidizing activity of the thiazolethione 3; herein we
report the results of this investigation.
ence of 2
dized (ca 10% conversion after 2 h of irradiation) to the
7,8-dihydro-8-oxo-2 -deoxyguanosine as the main oxida-
؅-deoxyguanosine (dG), the latter is photooxi-
؅
tion product. The dG conversion levels off after complete
consumption of thiazolethione 3 and is suppressed by the
addition of the hydroxyl-radical scavenger 2,6-di-tert-bu-
tylcresol or DMPO. Since the photoproducts 4–6 are in-
effective as sensitizers for the photooxidation of dG and
DNA, the hydroxyl radicals released in the photolysis of
thiazolethione 3 are the oxidizing species of DNA and dG.
These results suggest that the thiazolethione 3 may serve
as a novel and effective photochemical hydroxyl-radical
source for photobiological studies.
INTRODUCTION
The elucidation of the molecular mechanism of the DNA
damage by reactive oxygen species is important to under-
stand the complex cellular events such as carcinogenesis and
mutagenesis through oxidative stress (1). Oxyl radicals (hy-
droxyl or alkoxyl) have been known to damage DNA
through base modification and strand breaks (2). Efficient
oxyl-radical sources, which generate selectively oxyl radi-
MATERIALS AND METHODS
Materials. Thiazolethione 3 was prepared by the method reported
previously (20,22) and recrystallized from isopropanol before use.
Supercoiled pBR322 DNA was purchased from Pharmasia Biotech
Europe GmbH (Freiburg, Germany) and 5,5-dimethyl-1-pyrorrine N-
oxide (DMPO)† was obtaiend from Fulka Chemie AG (Buchs, Swit-
¶Posted on the website on 6 September 2000.
*To whom correspondence should be addressed at: The Institut fu¨r
Organische Chemie, Universita¨t Wu¨rzburg, Am Hubland, D-
97074 Wu¨rzburg, Germany. Fax: 49-931-8884756;
e-mail: adam@chemie.uni-wuerzburg.de;
†Abbreviations: DMPO, 5,5-dimethyl-1-pyrroline N-oxide; dG, 2Ј-
deoxyguanosine; EPR, electron paramagnetic resonance; GRP,
guanidine-releasing products; 4-HO-8-oxodG, 4-hydroxy-8-oxo-
Internet: http://www-organik.chemie.uni-wuerzburg.de
᭧
2000 American Society for Photobiology 0031-8655/00
$5.00
ϩ0.00
4,8-dihydro-2Ј-deoxyguanosine; HPLC, high-performance liquid
619

620 Waldemar Adam et al.
Chart 1.
Scheme 1.
zerland). Ethidium bromide for biochemical use and boric acid were
obtained from Merck KGaA (Freiburg, Germany). Tris base was
purchased from Sigma–Aldrich Chemie GmbH (Deisenhofen, Ger-
many) and agarose was purchased from Serva Feinbiochemica
GmbH (Heiderberg, Germany). Preparative thin-layer chromatogra-
Electron paramagnetic resonance studies. A solution of thiazo-
lethione 3 (5 mM), and DMPO (90 mM) in water and 20% aceto-
˚
nitrile as cosolvent, was irradiated at 300 nm with 16 RPR-3000 A
lamps (300 nm, 21 W) in a Rayonet photoreactor at 0 C for 5 min
Њ
phy (TLC) separations were performed on Merck PLC plates (20
ϫ
20 cm2) coated with silica gel 60.
in a Pyrex vial, and the resulting mixture was analyzed by electron
paramagnetic resonance (EPR) spectroscopy (Bruker EPR 420 spec-
trophotometer).
Radiation sources. The irradiation experiments were carried out
in a Rayonet photoreactor (Southern New England Ultraviolet Com-
˚
Modification of pBR322 DNA. The reactions were carried out in
pany, Brandford, CT), equipped with 16 RPR-3000 A (300 nm, 21
˚
Eppendorf tubes with supercoiled pBR322 DNA (1
zolethione 3 (0.509 mM) in 5.0 mM KH2PO4 buffer at pH 7.4 and
10% acetonitrile as cosolvent. The samples (10 L final volume)
were prepared from stock solutions of pBR322 DNA (33.3 g/mL
␮g/mL) and thia-
W) or RPR-3500 A (350 nm, 24 W) lamps.
Photochemistry of thiazolethione 3. Thiazolethione 3 (100 mg,
0.41 mmol) was dissolved in 100 mL of acetonitrile and the solution
␮
␮
was irradiated at 300 nm in a Rayonet photoreactor with 16 RPR-
˚
in 15.7 mM KH2PO4 buffer, pH 7.4) and thiazolethione 3 (5.07 mM
3000 A lamps (300 nm, 21 W) at 0ЊC for 10 min. The solvent was
in acetonitrile). The samples were irradiated at 350 nm with 12 RPR-
evaporated under reduced pressure (20ЊC, 18 Torr) and the residue
˚
3500 A lamps (350 nm, 24 W) in a Rayonet photoreactor for 5 min
at 0 C. The control experiments were carried out in the presence of
was extracted with ethyl acetate (ca 3 mL). The extract was con-
centrated and separated by preparative TLC (silica gel, 5:1 hexane :
ethyl acetate) to give disulfone 7 (Scheme 1) (19.0 mg, 40%, Rf 0–
0.24) and a mixture of disulfide 4, bisthiazole 5 and thiazole 6 (Rf
0.22–0.30). Subsequent preparative TLC separation of the latter mix-
ture (silica gel, 1:1 petroleum ether : dichloromethane) gave thiazole
6 (6.0 mg, 18%, Rf 0.080–0.20) and a 3:1 mixture of disulfide 4:
bisthiazole 5 (9.0 mg, 4: 18%, 5: 6%, Rf 0.22–0.30). The disulfide
4 was partly separated from the mixture of 4 and 5 by preparative
medium-pressure liquid chromatography (MPLC, silica gel, 5:1 di-
chloromethane : petroleum ether) and directly characterized, while
the bisthiazole 5 was characterized in the mixture by 1H nuclear
magnetic resonance (NMR) and mass spectroscopy.
Њ
isopropanol (10 vol%) or D2O (60 vol%) to assess the involvement
of hydroxyl radicals or singlet oxygen in the photolysis, respectively.
Detection of strand breaks. To the irradiated solutions was added
2.5
resulting mixture was transferred onto a 1% agarose gel, stained with
0.5 g/mL ethidium bromide. Electrophoresis was carried out in Tris
␮L of bromophenol gel-loading solution. An 8 ␮L aliquot of the
␮
buffer (18.0 mM Tris base, 18.0 mM boric acid and 10.0 mM ethy-
lenediaminetetraacetic acid at pH 8.0) at 78 V for ca 2 h in a Phar-
macia horizontal apparatus (GNA 100), equipped with a power sup-
ply (GPS 200/400). The fluorescent spots of the DNA were detected
by exposure to a UV transilluminator at 254 nm. Photographs of the
gels were taken with a Herolab camera E.A.S.Y. 429 K, which was
connected to a personal computer, equipped with Herolab E.A.S.Y.
software. The ratio of open-circular to the supercoiled DNA was
determined from the fluorescence intensities of the spots.
4: Colorless fine prisms, mp 156–158
(CDCl3) 7.80 (m, 4H), 7.50 (s, 2H), 7.38 (m, 4H); high-resolution
ЊC (lit. 155Њ
C [24]). 1H NMR
␦
mass spectroscopy (HRMS) (EI) m/e 451.9107, calcd for
C18H10Cl2N2S4 451.9104.
1
Photooxidation of 2
Typical procedure: the samples (final volume 200
0.509 mM of dG (25 L of a 4.07 mM aqueous solution) and 2.13
mM of thiazolethione 3 (30 L of a 14.2 mM solution in acetonitrile)
Ј-deoxyguanosine (dG) by thiazolethione 3.
5: H NMR (CDCl3)
␦ 7.72 (m, 4H), 7.44 (s, 2H), 7.34 (m, 4H);
␮L) contained
HRMS (EI) m/e 387.9660, calcd for C18H10Cl2N2S2 387.9662.
1
␮
6: Colorless prisms, mp 51–52.5
(CDCl3) 8.89 (d, 1H, J 2.0 Hz), 7.87 (m, 2H), 7.55 (d, 1H, J
2.0 Hz), 7.41 (m, 2H); MS (EI) m/e 195 (100%) (Mϩ).
7: Colorless fine needles, mp 227–229
C. 1H NMR (CDCl3)
(m, 4H), 7.41 (m, 4H), 6.95 (s, 2H); 13C NMR (CDCl3)
ЊC (lit. 51–53ЊC [25]). H NMR
␮
␦
ϭ
in 5 mM phosphate buffer (pH 7.0) and 15% acetonitrile as cosol-
ϭ
vent. The solutions were irradiated in Pyrex vials at 350 nm with
Њ
␦
7.78
˚
␦
107.7,
16 RPR-3500 A lamps (350 nm, 24 W) in a Rayonet photoreactor.
An 80 L aliquot of the photolysate was extracted with ethyl acetate
(2 200 L) and the aqueous phase was freeze-dried. The residue
was redissolved in 80 L of water and submitted to high-perfor-
127.7, 129.1, 130.0, 136.2, 142.6, 161.4; IR (KBr) 1105 (SO2), 1240
(NO), 1335 (SO2) cmϪ1; MS (EI) m/e 548 (37%) (Mϩ). Anal calcd
C18H10Cl2N2O6S4 (549.5): C, 39.35; H, 1.83; N, 5.10; S, 23.34.
Found: C, 39.26; H, 1.65; N, 4.93; S, 22.48.
␮
ϫ
␮
␮
mance liquid chromatography (HPLC) analysis.
Photoreaction of disulfide 4. A solution of disulfide 4 (27.8 mg,
0.0633 mmol) in a mixture of dioxane (1 mL) and water (0.3 mL)
HPLC analysis. The HPLC analytical system consisted of a Bis-
chof HPLC pump model 2200 (Bischof GmbH, Leonberg, Germa-
ny), equipped with a Rheodyne loop injector, model 7125 (Berkley,
was irradiated at 350 nm in a Rayonet photoreactor at 20
ЊC for 1 h.
The solvent was evaporated under reduced pressure (20 C, 18 Torr)
Њ
CA). For the detection of 4-hydroxy-8-oxo-4,8-dihydro-2Ј-deoxy-
and the residue was chromatographed on silica gel by using a 1:1
mixture of ether : petroleum ether as eluent to give thiazole 6 (5.3
guanosine (4-HO-8-oxodG) and the photoconversion of thiazolethi-
one 3, Waters 994 photodiode array detector (Waters GmbH, Esch-
born, Germany) was used. Guanidine-releasing products (GRP)
were, after fluorescent derivatization (27), analyzed with Shimadzu
RF-551 spectrofluorometric detector (Bischoff GmBH, Leonberg,
Germany). In the determination of the dG conversion and 7,8-di-
mg, 43%) and 2-mercaptothiazole 8 (2.3 mg, 16%).
1
8: Pale yellow prisms, mp 207–210ЊC (lit. 210–212
Њ
C [26]). H
NMR (acetone-d6)
␦ 7.83 (m, 2H), 7.53 (m, 2H), 7.18 (s, 1H); MS
(EI) m/e 227 (100%) (Mϩ).
hydro-8-oxo-2Ј-deoxyguanosine (8-oxodG) formation, a SunChrom
SpectraFlow 600 photodiode array detector (SunChrom GmbH,
Friedrichsdorf, Germany) was connected in series with an ESA coul-
ochem model 5100A electrochemical detector, equipped with a high-
sensitive analytical cell model 5011 (ESA Inc. Bedford, MA). Con-
chromatography; HRMS, high-resolution mass spectroscopy;
MPLC, medium-pressure liquid chromatograpy; NMR, nuclear
magnetic resonance; 8-oxodG, 7,8-dihydro-8-oxo-2Ј-deoxyguano-
sine; Rf, retention factor; TLC, thin-layer chromatography; tr, re-
tention time; tris base, tris(hydroxymethyl)aminomethane.
version of thiazolethione 3 was followed by using a 250
ϫ 4.6 (i.d.)
mm Eurospher 100-C18 5 m column (Knauer GmbH, Berlin, Ger-
␮

Photochemistry and Photobiology, 2000, 72(5) 621
Figure 1. Absorption spectra of thiazolethione 3 in acetonitrile (—)
and in 5 mM phosphate buffer (pH 7.0), 10% acetonitrile (· · ·); the
open triangle (᭝) denotes the reported absorption maxima of the
tetraethylammonium salt of 3 (23).
Scheme 2.
(MPLC, silica gel), disulfide 4 and thiazole 6 (18%) were
isolated and identified (23,24) (Scheme 1)‡.
many), with a mixture of methanol and water (95:5) as eluent at a
flow rate 1 mL/min, detected at 310 nm (tR 3 min).
Although disulfide 4 was partly separated by MPLC from
a mixture of disulfide 4 and bisthiazole 5, the latter was not
separated from the disulfide 4. Thus, bisthiazole 5 was char-
acterized by 1H NMR and mass spectroscopy, and the yields
of disulfide 4 (18%) and bisthiazole 5 (6%) were determined
The quantification of dG and 8-oxodG was achieved on a 250
ϫ
4.6 (i.d.) mm Eurospher 100-C18, 7 m column, by elution with a
␮
mixture of 50 mM sodium citrate buffer (pH 5.0) and methanol (80:
20) at a flow rate of 1 mL/min. For the electrochemical detection of
8-oxodG, oxidation potential was set at
ϩ450 mV.
The GRP analysis (27) was performed as follows: a 100
quot of the photolysate was extracted with ethyl acetate (2
␮
ϫ
L ali-
200
1
by H NMR. When thiazolethione 3 was irradiated at 350
nm, the same products were observed as in the photolysis at
300 nm (data not shown). For the photolysis of thiazolethi-
one 3 in an aqueous medium (3:1 mixture of dioxane : wa-
ter), the disulfide 4 (6%), bisthiazole 5 (1%) and thiazole 6
(18%) were observed as photoproducts.
␮
L), and the aqueous phase was freeze-dried. The residue was re-
dissolved in 100
␮
L of water, and to the solution was added 38
L of an aqueous solution of 1,2-naphthoqui-
none-4-sulfonic acid (5 mg/mL). The mixture was kept at 65 C for
10 min in the dark and then acidified with 42 L of 1N HCl. The
resulting mixture was separated on a 250 46 mm (i.d.) Eurospher
100-C18, 5 m column with a mixture of methanol and 25 mM
ammonium formate (20:80) as eluent, at a flow rate of 1 mL/min
and detected spectrofluorometrically ( 355 nm, 405 nm).
␮L
of 1N NaOH and 20
␮
Њ
␮
ϫ
A mechanism is proposed for the formation of the pho-
toproducts derived from the thiazolethione 3 in Scheme 2.
On irradiation, thiazolethione 3 releases hydroxyl radicals
(vide infra) to give the thiyl radical A, and subsequent di-
merization of the thiyl radical A affords the disulfide 4
(22,28). Photodissociation of the C–S bond (28,29) in the
disulfide 4 may give the radicals B and C. Such a C–S bond
cleavage has been suggested in the photolysis of bis(2-ben-
zothiazolyl) disulfide (28). C–S bond fission may compete
efficiently with the expected S–S bond homolysis in disul-
fide 4, which produces the carbon-centered radical C; the
latter dimerizes to the bisthiazole 5 or abstracts a hydrogen
atom to afford thiazole 6. The disulfanyl radical B may also
generate the thiazolyl radical C by sulfur extrusion.
␮
␭
ϭ
␭
ϭ
em
ex
RESULTS AND DISCUSSION
Absorption spectra of thiazolethione 3
The absorption properties of thiazolethione 3 were measured
in aqueous buffer and, for comparison, in acetonitrile. As
shown in Fig. 1, thiazolethione 3 shows absorption maxima
at 244 and 310 nm in acetonitrile, while in 5 mM phosphate
buffer (10% acetonitrile as cosolvent, pH 7.0) the first ab-
sorption band shifts to 300 nm. The absorption characteris-
tics of thiazolethione 3 in the phosphate buffer are similar
When the disulfide 4 was irradiated at 350 nm in a 3:1
to those reported for tetraethylammonium salt 3
Therefore, in aqueous phosphate buffer (pH 7.0), the con-
jugate base 3 of the thiazolethione 3 dominates in the acid–
Ј-NEt4 (22).
‡Disulfone 7 was isolated after thin-layer chromatographic (TLC)
(silica gel) separation of the crude photolysate and characterized
by spectroscopic data and elemental analysis. However, the di-
sulfone 7 was not detected in the photolysate of thiazolethione 3
before work-up and, thus, it is not a direct photoproduct. A control
experiment showed that when the thiazolethione 3 was placed
onto a preparative TLC plate (silica gel) and eluted with 2:5 ethyl
acetate : hexane without irradiation and even in the dark, the di-
sulfone 7 was observed.
Ј
base equilibrium.
Photochemistry of thiazolethione 3
Since it is important to know the photoproducts of the thia-
zolethione 3 in order to assess their individual photooxida-
tion activity (14), a detailed product study of the thiazole-
thione 3 photochemistry was conducted. Thiazolthione 3
(100 mg, 0.41 mmol) was irradiated in a Rayonet photo-
reactor (300 nm) at 0ЊC for 10 min in 100 mL of acetonitrile.
After separation of the photolysate by preparative TLC on
silica gel and medium-pressure liquid chromatography

622 Waldemar Adam et al.
Scheme 3.
mixture of dioxane : water, the thiazole 6 (43%) and 2-mer-
captothiazole 8 (16%) were obtained (Scheme 3). These re-
sults support the proposed mechanism in Scheme 2, in that
the observed photoproducts of thiazolethione 3 are formed
through the cleavage of the S–S and C–S bonds of its pri-
mary photoproduct, namely the disulfide 4.
Figure 2. EPR spectrum of the DMPO (89.5 mM) spin-trapping
adduct with the hydroxyl radical, formed in the photolysis of thia-
zolethione 3 (1 mM, irradiated at 300 nm) in H2O and 20 vol%
acetonitrile as cosolvent.
Detection of hydroxyl radicals on photolysis of
thiazolethione 3
formation by thiazolethione 3. As the photolysate of thia-
zolethione 3 (preirradiated for 10 min) induced insignificant
(Ͻ8% open-circular form) strand breaks on irradiation (data
In order to assess the propensity of thiazolethione 3 to gen-
erate hydroxyl radical on photolysis, a spin-trapping exper-
iment with DMPO was conducted. When a solution of 3 (1
mM) in a 80:20 mixture of water and acetonitrile was irra-
diated at 300 nm in the presence of DMPO (89.5 mM), EPR
not shown), the photoproducts of thiazolethione 3 have only
a minor effect on the induction of strand breaks. Further-
more, the expected thiyl radicals A (Scheme 2) are not re-
sponsible for the strand-break formation, since it has been
reported that sulfur-centered radicals do not cause significant
DNA damage (34). Therefore, it is concluded that the strand
breaks have resulted from the hydroxyl radicals, which have
been generated in the photolysis of thiazolethione 3.
spectroscopy revealed a characteristic 1:2:2:1 pattern (g
ϭ
2.0061, ϭ ␣H 14.6 G) for the DMPO adduct of hy-
␣
N
ϭ
droxyl radical (30) (Fig. 2). On irradiation of thiazolethione
3 at 350 nm, the same hydroxyl-radical spin adduct of
DMPO was detected as in the case of 300 nm. This confirms
unambiguously that thiazolethione 3 generates hydroxyl rad-
icals on photoexcitation in aqueous medium; however, the
expected sulfur-centered radical A (cf Scheme 2) was not
detected as a spin adduct of DMPO. It has been reported
(31) that the spin adduct of DMPO with 2-pyridylthiyl rad-
ical, formed in the photolysis of N-hydroxypyridinethione
1a, is too short-lived for EPR-spectral detection. By analogy,
we suspect that the spin adduct of DMPO with the thiyl
radical A also does not persist to be observed under the
present conditions.
Oxidation of dG in the photolysis of thiazolethione 3
In the photooxidation of dG (Chart 2) by thiazolethione 3,
moderate conversion of dG (up to 10% after 1 h of irradi-
ation at 350 nm in phosphate buffer [pH 7.0] and 15% ace-
tonitrile as cosolvent) was observed when a four-fold excess
of thiazolethione 3 versus dG was used. In this photoreac-
tion, 8-oxodG was obtained as the major oxidation product;
4-HO-8-oxodG, the typical Type-II product (27), was not
observed. Only trace amounts of oxazolone, the GRP, a
Type-I product (27), was detected. The product distribution
Strand-break formation in pBR322 DNA on photolysis
of thiazolethione 3
The photolysis of supercoiled pBR322 DNA (10 ␮g/mL in
5 mM KH2PO4 buffer at pH 7.4, 10% acetonitrile as cosol-
vent) was conducted in the presence of thiazolethione 3. As
shown in Fig. 3 (lane 1), this photolysis produced 30(Ϯ3)%
of the open-circular form of DNA. The blank (lane 4), i.e.
the absence of thiazolethione 3, was ineffective (for conve-
nience of comparison, the blank value [30% open-circular
form] was set to zero). Addition of isopropanol, an efficient
hydroxyl-radical scavenger (32), inhibited efficiently strand-
break formation (3[Ϯ2]%, Fig. 3 [lane 2]). To assess whether
DNA strand breaks are induced by sensitized formation of
singlet oxygen, generated through photosensitization by thia-
zolethione 3 or its photoproducts, the photolysis was carried
out in D2O, in which the lifetime of singlet oxygen is about
10 times longer than in H2O (33). As shown in Fig. 3 (lane
3), D2O exhibited no significant effect in the strand-break
Figure 3. Gel-electrophoretic detection of photoinduced strand
breaks with pBR322 DNA (10
pH 7.4) and 10 vol% acetonitrile as cosolvent, irradiated at 350 nm
and 0 C in the presence of thiazolethione 3.
␮g/mL) in KH2PO4 buffer (5 mM,
Њ

Photochemistry and Photobiology, 2000, 72(5) 623
Figure 4. Conversion (%) of dG (ⅷ) and yield (%) of 8-oxodG (⅜)
in the photooxidation of dG (0.509 mM) by thiazolethione 3 (2.13
mM), irradiated at 350 nm in phosphate buffer (pH 7.0) and 15%
acetonitrile as cosolvent; the error bars represent at least two inde-
pendent experiments; the inset displays conversion of thiazolethione
3 (2.13 mM) during the irradiation at 350 nm in phosphate buffer
(pH 7.0) and 15% acetonitrile as cosolvent.
Chart 2.
suggests that neither the oxidation by singlet oxygen nor an
electron-transfer process is important in the photooxidation
of dG.
were conducted. The conversion of dG was suppressed mod-
erately in the presence of 5 mM of 2,6-di-tert-butylcresol
after 30 min of irradiation (2.7%, cf 4.2% in the absence of
any additive), while an excess amount of DMPO (895 mM)
inhibited, essentially completely, the dG conversion
Figure 4 shows the time profiles for the dG conversion
and the yield of 8-oxodG in the photolysis of thiazolethione
3, as well as the time dependence of the photodegradation
of thiazolethione 3 (Fig. 4, inset) for comparison. The con-
version of dG starts immediately (no induction period) and
levels off after 1 h of irradiation. The yield of 8-oxodG,
given relative to consumed dG, shows the same time depen-
dence as the dG conversion; as much as 5% (based on con-
sumed dG) was obtained after exposure for 2 h.
(Ͻ0.4%) and 8-oxodG formation (0%). Consequently, these
scavenging data corroborate that hydroxyl radicals are also
the main active species in the photooxidation of dG by thia-
zolethione 3.
Both time profiles, the one for the dG conversion and the
one for the 8-oxodG formation, run almost parallel to the
profile for the photoconsumption of thiazolethione 3 (see
Fig. 4, inset). Thus, after about 30 min, essentially all thia-
zolethione 3 is photolyzed, while beyond this time the con-
version of dG and the formation of 8-oxodG have leveled
off. This correspondence suggests that the photoproducts of
the thiazolethione 3 do not sensitize the photooxidation of
dG. To confirm this unequivocally, the authentic photoprod-
ucts of the thiazolethione 3 were tested for their efficacy of
dG photooxidation. When dG was irradiated (300 nm) with
the individual photoproducts (4 [0.05 mM], 3:1 mixture of
4 and 5 [4: 0.038 mM, 5: 0.013 mM] or 6 [0.05 mM]) in
phosphate buffer (pH 7.0) and 10% acetonitrile as cosolvent,
insignificant dG consumption and only a trace amount of 8-
oxodG were detected (0.1% for 4, 0.2% for 4 and 5, and
none for 6, based on the initial amount of dG). These results
confirm unequivocally that the photoproducts of thiazolethi-
one 3 are all ineffective in sensitizing the photooxidation of
dG. That sulfur-centered radicals are not involved in the dG
oxidation is substantiated by the fact that disulfide 4, which
generates thiyl radicals A on photoexcitation (Schemes 2 and
3), does not cause any significant dG conversion. Conse-
quently, the photooxidation of dG is caused by the hydroxyl
radicals released in the photolysis of thiazolethione 3.
To substantiate the involvement of hydroxyl radicals in
the dG photooxidation by thiazolethione 3, inhibition exper-
iments with the established radical scavenger 2,6-di-tert-bu-
tylcresol (35) and with the efficient radical trap DMPO (30)
CONCLUSIONS
The photooxidative damage of DNA (strand-break forma-
tion) and dG (8-oxodG formation) by thiazolethione 3, ex-
amined herein for the first time, is caused by the hydroxyl
radicals released upon irradiation. Spin-trapping and radical-
scavenging experiments confirm the generation of the hy-
droxyl radicals and their involvement in the photooxidation
of DNA and dG. The photoproducts of thiazolethione 3 are
ineffective in causing strand breaks and base oxidation, as
attested by control experiments.
Although the N-hydroxypyridinethione 1a has been estab-
lished as effective and convenient photochemical hydroxyl-
radical source (6,7), its photoproducts significantly sensitize
the photooxidation of DNA (14). Evidently, thiazolethione
3 may serve as an advantageous alternative to N-hydroxy-
pyridinethione 1a for the photobiological studies, since it
generates hydroxyl radicals more selectively on photolysis,
without the complications of photosensitization. Also, thia-
zolethione 3 is more advantageous than pyridone 2 for pho-
tobiological studies because the former absorbs UVA light
(
␭ ϭ 350 nm) sufficiently to be photoreactive at this wave-
length, while the latter essentially does not absorb light at
330 nm (12). Indeed, the thiazolethione 3 is about two-
␭
Ͼ
fold more efficient in dG photooxidation than pyridone 2,
since 4 equiv. of thiazolethione 3 afforded 0.6% 8-oxodG
(based on starting amount of dG), while 10 equiv. of pyri-
done 2 gave maximally 0.8% of 8-oxodG (13).

624 Waldemar Adam et al.
Acknowledgements The present work was supported by the Deut-
sche Forschungsgemeinschaft (Sonderforshungsbereich 172 ‘‘Mo-
lekulare Mechanismen kanzerogener Prima¨nderungen’’) and Fonds
der Chemischen Industrie. H.O. thanks the Ministry of Education,
Science, Sports and Culture of Japan for a postdoctral fellowship.
droxyl radical-induced degradation of 2Ј-deoxyguanosine under
reducing conditions. J. Chem. Soc., Perkin Trans. 2, 1875–1880.
16. Aveline, B. M., I. E. Kochevar and R. W. Redmond (1996) N-
hydroxypyridine-2(1H)-thione: not a selective generator of hy-
droxyl radicals in aqueous solution. J. Am. Chem. Soc. 118,
289–290.
17. Aveline, B. M., I. E. Kochevar and R. W. Redmond (1996)
Photochemistry of the nonspecific hydroxyl radical generator,
N-hydroxypyridine-2(1H)-thione. J. Am. Chem. Soc. 118,
10 113–10 123.
18. Barton, D. H. R., D. Crich and G. Kretzschmar (1984) Forma-
tion of carbon-carbon bonds with radicals derived from the es-
ters of thiohydroxamic acids. Tetrahedron Lett. 25, 1055–1058.
19. Barton, D. H. R., D. Crich and P. Potier (1985) On the mech-
anism of the decarboxylative rearrangement of thiohydroxamic
esters. Tetrahedron Lett. 26, 5943–5946.
REFERENCES
1. Kasai, H. and S. Nishimura (1991) Formation of 8-hydroxy-
deoxyguanosine in DNA by oxygen radicals and its biological
significance. In Oxidative Stress: Oxidants and Antioxidants
(Edited by H. Sies), pp. 99–116. Academic Press, New York.
2. Burrows, C. J. and J. G. Muller (1998) Oxidative nucleobase
modifications leading to strand scission. Chem. Rev. 98, 1109–
1151.
3. Saito, I., M. Takayama and T. Matsuura, (1990) Phthalimides
hydroperoxides as efficient photochemical hydroxyl radical gen-
erators. A novel DNA-cleaving agent. J. Am. Chem. Soc. 112,
883–884.
20. Barton, D. H. R., D. Crich and G. Kretzschmar (1986) The
invention of new radical chain reactions. Part 9. Further radical
chemistry of thiohydroxamic esters; formation of carbon-carbon
Bonds. J. Chem. Soc., Perkin Trans. 1, 39–53.
4. Matsugo, S., S. Kawanishi, K. Yamamoto, H. Sugiura, T. Mat-
suura and I. Saito (1991) Bis(hydroxy)naphthaldiimide as a
‘‘photo-Fenton’’ reagent: sequence-specific photocleavage of
DNA. Angew. Chem. Int. Ed. Engl. 30, 1351–1353.
5. Epe, B., M. Ha¨rig, D. Ramaiah, H. Stopper, M. M. Abou-Er-
zahb, W. Adam and C. R. Saha-Mo¨ller (1993) DNA Damage
induced by furocoumarin hydroperoxides plus UV (360 nm).
Carcinogenesis 14, 2271–2276.
6. Adam, W., D. Ballmaier, B. Epe, G. N. Grimm and C. R. Saha-
Mo¨ller (1995) N-Hydroxypyridinethiones as photochemical hy-
droxyl radical source for oxidative DNA damage. Angew. Chem.
Int. Ed. Engl. 34, 2156–2158.
21. Hartung, J. and M. Schwarz (1997) Towards improved alkoxy
radical precursors—the synthesis of N-alkoxy-4-(p-chlorophen-
yl)thiazole-2(3H)-thiones. Synlett 848–850.
22. Hartung, J., M. Schwarz, I. Svoboda, H. Fueß and M. T. Duarte
(1999) A new generation of alkoxyl radical precursors—prep-
aration and properties of N-(alkoxy)-4-arylthiazole-2(3H)-thi-
ones. Eur. J. Org. Chem. 1275–1290.
23. Hartung, J., R. Kneuer, M. Schwarz, I. Svoboda and H. Fueß
(1999) On the selective O-alkylation of ambident nucleo-
philes—the synthesis of thiohydroxamic acid O-esters by phase-
transfer reactions. Eur. J. Org. Chem. 97–106.
¨
24. Runge, F., A. Jumar and F. Koehler (1965) Uber einige unsym-
merishe heterocyclische Disulfide. II. J. Prakt. Chem. 21, 39–
47.
7. Epe, B., D. Ballmaier, W. Adam, G. N. Grimm and C. R. Saha-
Mo¨ller (1996) Photolysis of N-hydroxypyridinethiones: a new
source of hydroxyl radicals for the direct damage of cell-free
and cellar DNA. Nucleic Acids Res. 24, 1625–1631.
8. Adam, W., G. N. Grimm and C. R. Saha-Mo¨ller (1998) DNA
cleavage induced by alkoxyl radicals generated in the photolysis
of N-alkoxylpyridinethiones. Free Radic. Biol. Med. 24, 234–
238.
9. Adam, W., P. Groer, K. Mielke, C. R. Saha-Mo¨ller, R. Hutterer,
W. Kiefer, V. Nagel, F. W. Scheider, D. Ballmaier, Y. Schleger
and B. Epe (1997) Photochemical and photobiological studies
with acridine and phenanthridine hydroperoxides in cell-free
DNA. Photochem. Photobiol. 66, 26–33.
25. Erlenmeyer, H., J. Eckenstein, E. Sorkin and H. Meyer (1950)
¨
Uber die tuberkulostatische Wirkung von Derivaten der 3 iso-
meren Phenylthiazole. Helv. Chim. Acta 33, 1271–1276.
26. Vernin, G. and J. Metzger (1963) Synthe`ses d’aryl-4 thiazoles
substitue´s ou non en position-2. Bull. Soc. Chim. Fr. 2498–
2503.
27. Ravanat, J.-L., M. Berger, F. Bernard, R. Langlois, R. Ouellet,
J. E. van Lier and J. Cadet (1992) Phthalocyanine and naphthal-
ocyanine photosensitized oxidation of 2Ј-deoxyguanosine: dis-
tinct type I and type II products. Photochem. Photobiol. 55,
809–814.
28. Abdou, W. M., M. M. Sidky and H. Wamhoff (1987) Photo-
chemistry of pesticides, 12. On the photoconversion of 1,3-di-
hydro-2 H-benzimidazole-2-thione, 2(3H)-benzothiazole thione,
and 2-chlorobezothiazole. Z. Naturforsch. 42b, 1153–1158.
29. Gilbert, A. and J. Baggott (1991) Sulfur containing chromo-
phores. In Essentials of Molecular Photochemistry, pp. 477–
479. Blackwell Scientific Publications, Oxford, UK.
30. Finkelstein, E., G. M. Rosen and E. J. Rauckman (1980) Spin
trapping of superoxide and hydroxyl radicals: practical aspects.
Arch. Biochem. Biophys. 200, 1–16.
31. Reszka, K. J. and C. F. Chignell (1994) Photochemistry of 2-
mercaptopyridines. Part 1. An EPR and spin-trapping investi-
gation using 5,5-dimethyl-1-pyrroline N-oxide in aqueous so-
lutions. Photochem. Photobiol. 60, 442–449.
32. Neta, P. and R. H. Schuler (1975) Rate constants for the reaction
of OϪ radicals with organic substrates in aqueous and toluene
solution. J. Phys. Chem. 79, 1–6.
10. Adam, W., J. Cadet, F. Dall’Acqua, B. Epe, D. Ramaiah and
C. R. Saha-Mo¨ller (1995) Photosensitized formation of 8-hy-
droxy-2Ј-deoxyguanosine in salmon testes DNA by furocou-
marin hydroperoxides: a novel, intercalating ‘‘photo-Fenton re-
agent’’ for oxidative DNA damage. Angew. Chem. Int. Ed. Engl.
34, 107–110.
11. Adam, W., M. Berger, J. Cadet, F. Dall’Acqua, B. Epe, S.
Frank, D. Ramaiah, S. Raoul, C. R. Saha-Mo¨ller and D. Vedaldi
(1996) Photochemistry and photobiology of furocoumarin hy-
droperoxides derived from imperatorin: novel intercalating pho-
to-Fenton reagents for oxidative DNA modification by hydroxyl
radicals. Photochem. Photobiol. 63, 768–778.
12. Aveline, B. M., I. E. Kochevar and R. W. Redmond (1996)
Photochemistry of N-hydroxy-2(1H)-pyridone, a more selective
source of hydroxyl radicals than N-hydroxypyridine-2(1H)-thi-
one. J. Am. Chem. Soc. 118, 10 124–10 133.
13. Adam, W., S. Marquardt and C. R. Saha-Mo¨ller (1999) Oxi-
dative DNA damage in the photolysis of N-hydroxy-2-pyridone,
a specific hydroxyl-radical source. Photochem. Photobiol. 70,
287–291.
14. Adam, W., G. N. Grimm, S. Marquardt and C. R. Saha-Mo¨ller
(1999) Are pyridinethiones reliable photochemical oxyl-radical
sources for photobiological studies? The importance of second-
33. Rodgers, M. A. J. and P. T. Snowden (1982) Lifetime of O2
1
(
⌬
g) in liquid water as determined by time-resolved infrared
luminescence measurements. J. Am. Chem. Soc. 104, 5541–
5543.
34. Griffith, J. and J. A. Murphy (1992) Model studies for damage
to nucleic acids mediated by thiyl radicals. Tetrahedron 48,
5543–5556.
35. Muto, T., J. Tanaka, T. Miura and M. Kimura (1982) Iron-
catalyzed autooxidation of cholesterol in the presence of unsat-
urated long-chain fatty acid. Chem. Pharm. Bull. 30, 3172–
3177.
ary photolysis products in the guanine oxidation of 2Ј-deoxy-
guanosine and cell-free DNA. J. Am. Chem. Soc. 121, 1179–
1185.
15. Douki, T., S. Spinelli, J.-L. Ravanat and J. Cadet (1999) Hy-