Photooxidative damage of guanine in DG and DNA by the radicals derived from the α cleavage of the electronically excited carbonyl products generated in the thermolysis of alkoxymethyl-substituted dioxetanes and the photolysis of alkoxyacetones

Article abstract of DOI:10.1021/jo0056491

On thermolysis of the methoxy (MeO-TMD), tert-butoxy (^tBuO-TMD), and hydroxy (HO-TMD) derivatives of 3,3,4,4-tetramethyl-1,2-dioxetane (TMD) in the presence of dG and calf-thymus DNA, the guanine is oxidized considerably more efficiently than the parent TMD. The same trend in the oxidative reactivity is observed for the photolysis of the corresponding oxy-substituted ketones versus acetone. The oxidative reactivity order in the dioxetane thermolysis, as well as in the ketone photolysis, parallels the ability of the excited ketones to release radicals (determined by spin trapping with DMPO and EPR spectroscopy) upon α cleavage (Norrish-type-I reaction). In the presence of molecular oxygen, the carbon-centered radicals are scavenged to produce peroxyl radicals, which are proposed as the reactive species in the, oxidation of the guanine in dG and calf-thymus DNA.

Full text of DOI:10.1021/jo0056491

J . Org. Chem. 2001, 66, 597-604
597
P h otooxid a tive Da m a ge of Gu a n in e in DG a n d DNA by th e
Ra d ica ls Der ived fr om th e r Clea va ge of th e Electr on ica lly
Excited Ca r bon yl P r od u cts Gen er a ted in th e Th er m olysis of
Alk oxym eth yl-Su bstitu ted Dioxeta n es a n d th e P h otolysis of
Alk oxya ceton es
Waldemar Adam,* Markus A. Arnold, and Chantu R. Saha-Mo¨ller
Institut fu¨r Organische Chemie, Universita¨t Wu¨rzburg, Am Hubland, D-97074 Wu¨rzburg, Germany
adam@chemie.uni-wuerzburg.de
Received September 15, 2000
On thermolysis of the methoxy (MeO-TMD), tert-butoxy (tBu O-TMD), and hydroxy (HO-TMD)
derivatives of 3,3,4,4-tetramethyl-1,2-dioxetane (TMD) in the presence of d G and calf-thymus DNA,
the guanine is oxidized considerably more efficiently than the parent TMD. The same trend in the
oxidative reactivity is observed for the photolysis of the corresponding oxy-substituted ketones versus
acetone. The oxidative reactivity order in the dioxetane thermolysis, as well as in the ketone
photolysis, parallels the ability of the excited ketones to release radicals (determined by spin trapping
with DMP O and EPR spectroscopy) upon R cleavage (Norrish-type-I reaction). In the presence of
molecular oxygen, the carbon-centered radicals are scavenged to produce peroxyl radicals, which
are proposed as the reactive species in the oxidation of the guanine in d G and calf-thymus DNA.
In tr od u ction
age DNA in cell-free as well as in cellular systems has
been demonstrated.7,8
In oxidative stress, reactive oxygen species such as oxyl
radicals (hydroxyl, alkoxyl, and peroxyl), superoxide
radical anion, singlet oxygen, hydrogen peroxide or alkyl
peroxides are responsible for the damage of cellular
constituents through oxidation.1 The consequences of the
oxidative damage in the case of DNA may be mutagenesis
and carcinogenesis.2 As for the origin of these oxidants,
they are formed during oxygen metabolism3 or by expo-
sure to UV radiation (sunlight). The latter involves
electronically excited states, which may be generated by
excitation of endogenous sensitizers, but also by lipid
peroxidation (Russell mechanism).4
In our group, the oxidative damage of d G and DNA in
the thermolysis of 3-(hydroxymethyl)-3,4,4-trimethyl-1,2-
dioxetane (HO-TMD) has been studied intensively since
it is significantly more effective than the merely alkyl-
substituted TMD.9 The products detected in the oxidation
of calf-thymus DNA have been 7,8-dihydro-8-oxoguanine
(8-oxoGu a ) and guanidine-releasing products (GRP ),
e.g. oxazolone. The oxidative damage was attributed to
(5) (a) Kochevar, I. E.; Dunn, D. A. In Bioorganic Photochemistry;
Morrison, H., Ed.; J ohn Wiley & Sons: New York, 1990; Chapter 4,
pp 273-316. (b) Rahn, R. O. In DNA repair, a laboratory manual of
research procedures; Friedberg, E. C., Hanawalt, P. R., Eds.; Marcel
Dekker: New York, 1983; pp 75-85. (c) Lamola, A. A.; Yamane, Y.
Proc. Natl. Acad. Sci. U.S.A. 1967, 58, 443-446. (d) Epe, B.; Henzl,
H.; Adam, W.; Saha-Mo¨ller, C. R. Nucleic Acid Res. 1992, 21, 863-
869. (e) Zierenberg, B. E.; Kramer, D. M.; Geisert, M. G.; Kirste, R. G.
Photochem. Photobiol. 1971, 14, 515-520. (f) Charlier, M.; He´le`ne, C.
Photochem. Photobiol. 1972, 15, 527-536. (g) Hildenbrand, K.; Schulte-
Frohlinde, D. Int. J . Radiat. Biol. 1997, 71, 377-385. (h) Umlas, M.
E.; Franklin, W. A.; Chan, G. L.; Haseltine, W. A. Photochem. Photobiol.
1985, 42, 265-273. (i) Song, Q. H.; Yao, S. D.; Lin, N. Y. J . Photochem.
Photobiol. B: Biol. 1997, 40, 199-203. (j) Mennigmann, H.-D.; Wacker,
A. Photochem. Photobiol. 1970, 11, 2911-2916.
The induction of DNA damage by electronically excited
compounds directly (e.g., triplet ketones) is well estab-
lished.5 Since 1,2-dioxetanes produce electronically ex-
cited ketones upon thermolysis, they are ideally suited
to assess their excited-state reactivity without the need
of exposing the biological target directly to UV irradia-
tion. Another advantage is the selective generation of
triplet-excited ketones on thermal decomposition of di-
oxetanes.6 The propensity of various dioxetanes to dam-
(6) (a) Murphy, S.; Adam, W. J . Am. Chem. Soc. 1996, 118, 12916-
12921. (b) Turro, N. J .; Lechtken, P. J . Am. Chem. Soc. 1972, 94, 2886-
2888.
(7) Lamola, A. A.; Biophys. Res. Commun. 1971, 43, 898-899.
(8) (a) Adam, W.; Beinhauer, A.; Mosandl, T.; Saha-Mo¨ller, C. R.;
Vargas, F.; Epe, B.; Mu¨ller, E.; Schiffmann, D.; Wild, D. Environ.
Health Perspect. 1990, 88, 89-97. (b) Epe, B.; Mu¨ller, E.; Adam, W.;
Saha-Mo¨ller, C. R. Chem.-Biol. Int. 1992, 85, 265-281. (c) Emmert,
S.; Epe, B.; Saha-Mo¨ller, C. R.; Adam, W.; Ru¨nger, M. Photochem.
Photobiol. 1995, 61, 136-141. (d) Adam, W.; Epe, B.; Schiffmann, D.;
Vargas, F.; Wild, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 429-
431.
(9) (a) Adam, W.; Andler, S.; Nau, W.; Saha-Mo¨ller, C. R. J . Am.
Chem. Soc. 1998, 120, 3549-3559. (b) Adam, W.; Saha-Mo¨ller, C. R.;
Scho¨nberger, A. J . Am. Chem. Soc. 1996, 118, 9233-9238. (c) Adam,
W.; Saha-Mo¨ller, C. R.; Scho¨nberger, A. J . Am. Chem. Soc. 1997, 119,
719-723. (d) Adam, W.; Saha-Mo¨ller, C. R.; Scho¨nberger, A.; Berger,
M.; Cadet, J . Photochem. Photobiol. 1995, 62, 231-238.
* To whom correspondence should be addressed. Fax: +49 931
8884756. Internet: http://www-organik.chemie.uni-wuerzburg.de.
(1) Sies, H. In Oxidative Stress; Sies, H., Ed.; Academic Press:
London, 1985; Chapter 1, pp 1-10.
(2) (a) Ames, B. N. Science 1983, 221, 1256-1264. (b) Besaga, H. S.
Biochem. Cell. Biol. 1989, 68, 989-998. (c) Meneghini, R. Mutat. Res.
1988, 195, 215-230. (d) Cadenas, E. Annu. Rev. Biochem. 1989, 58,
79-110.
(3) (a) Halliwell, B. Mutat. Res. 1999, 443, 37-52. (b) Wang, D.;
Kreutzer, D. A.; Essigmann, J . M. Mutat. Res. 1998, 400, 99-115. (c)
Stogner, S. W.; Payne, D. K. Ann. Pharmacother. 1992, 26, 1554-1562.
(d) Marnett, L. J . Carcinogenesis 2000, 21, 361-370.
(4) Liu, Y.; Stolze, K.; Dadak, A.; Nohl, H. Photochem. Photobiol.
1997, 66, 443-449.
10.1021/jo0056491 CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/23/2000

598 J . Org. Chem., Vol. 66, No. 2, 2001
Adam et al.
Ta ble 1. Tr ip let-Excita tion P a r a m eter s of th e
Dioxeta n es a n d Rela tive Efficien cies of th e DG
Oxid a tion
Sch em e 1. Gen er a tion of Ra d ica ls in th e
Th er m olysis of 1,2-Dioxeta n es a n d in th e
P h otolysis of th e Cor r esp on d in g Keton es
rel efficiency of
d G oxidna
c
kdec
ΦT b
(%)
ET
P
dioxetane (10-6 s-1
)
(1017 s-1 L-1
)
dioxetane ketone
TMD
HO-TMD
4.3 ( 0.2 35 ( 3d
4.9 ( 1.2 14 ( 3d
9.1 ( 1.1
4.1 ( 0.9
3.5 ( 0.8
1.6 ( 0.3
0.1 ( 0.1 0.0 ( 0.1
1.0 ( 0.3 1.0 ( 0.4
1.5 ( 0.6 0.1 ( 0.1
1.2 ( 0.2 1.2 ( 0.1
MeO-TMD 3.4 ( 0.5 17 ( 4
tBuO-TMD 3.2 ( 0.3
8 ( 8
a
The values for HO-TMD and hydroxyacetone were set to
unity. Calculated according to eq 1 (T ) 37 °C). c Relative to HO-
TMD in CH3CN. Data from ref 12.
b
d
(1 mM) thermolysis were determined by monitoring the
decay of the chemiluminescence intensity at 50 °C in
CH3CN (Table 1). The decomposition rate constants (kdec
)
the radical species released upon R cleavage (Norrish type
I) of the initially formed excited ketone (Scheme 1). The
for the alkoxymethyl-substituted dioxetanes were deter-
mined by fitting the data to an exponential function (first-
order kinetics) and were found to be slightly smaller than
for the unsubstituted 3,3,4,4-tetramethyl-1,2-dioxetane
(TMD) or the hydroxymethyl-substituted (HO-TMD)
t
ones. For MeO-TMD and Bu O-TMD also the decay at
37 °C (5 mM) and 65 °C (1 mM) was measured to
calculate the activation parameters of these two diox-
etanes from these data (MeO-TMD: ∆Hq ) 24.6 ( 1.0
kcal/mol, ∆Sq ) 4.3 ( 3.2 cal/mol K, tBu O-TMD: ∆Hq )
24.4 ( 0.5 kcal/mol, ∆Sq ) -4.6 ( 1.6 cal/mol K, the data
for TMD and HO-TMD may be found in ref 12).
To determine the amount of triplet-excited ketones
formed during dioxetane thermolysis, the triplet quan-
tum yield (ΦT) was measured by chemiluminescence with
9,10-dibromoanthracene (DBA) as triplet-energy acceptor
(variation of the concentration of DBA at a constant
temperature and a constant dioxetane concentration);13
the data are given in Table 1. From the triplet quantum
yields (ΦT) and the decomposition rate constants (kdec at
37 °C), the triplet excitation flux (ETP) was calculated
carbon-centered radicals are trapped by molecular oxy-
gen, and the resulting peroxyl radicals are made respon-
sible for the observed DNA oxidation products. The
higher efficiency of HO-TMD to oxidize DNA compared
to TMD relates to the fact that the latter generates
triplet acetone, which is well established to undergo
reluctantly R cleavage10 relative to the triplet hydroxy-
acetone11 derived from HO-TMD.
according to the established relation in eq 1. The ET
p
The large difference in the capacity of HO-TMD versus
TMD to oxidize guanine by means of the in-situ-gener-
ated radicals raised the question how efficiently alkoxy-
substituted dioxetanes would cause such damage. For
this purpose, the alkoxy-substituted derivatives MeO-
parameter is a useful measure of the effective number
of triplet-excited species generated in the thermolysis of
the dioxetanes per unit time and volume.14 Significant
is the fact that the ET values (Table 1) for the alkoxy-
p
methyl-substituted dioxetanes were all smaller than that
for the unsubstituted TMD.
t
TMD and Bu O-TMD were chosen, both of which are
unknown, but readily prepared from HO-TMD by alkyl-
ation. Since these alkoxymethyl-substituted dioxetanes
release triplet-excited ketones, namely methoxyacetone
from MeO-TMD and tert-butoxyacetone from tBu O-
TMD; for comparison, it was important to examine the
oxidation of d G and DNA by these ketones on photo-
excitation. In this comparative study, we present the
results on the oxidation of d G and DNA in the thermoly-
sis of dioxetanes and the photolysis of the respective
ketones. The efficiency of the oxidative damage is cor-
related with the release of radicals (determined by spin
trapping with DMP O and EPR spectroscopy) through R
cleavage of the electronically excited ketones.
Rela tive Efficien cy of d G Oxid a tion . To assess the
relative efficiency of the various dioxetanes to oxidize d G,
the amount of d G consumed (% conversion) in the
thermolysis of the particular dioxetane (20.0 mM) with
d G (0.500 mM) was determined in a 9:1 mixture of
phosphate buffer (5.00 mM, pH 7.0) and acetonitrile at
Resu lts
(12) Saha-Mo¨ller, C. R.; Adam, W. In Comprehensive Heterocyclic
Chemistry Vol. 1B.; Katritzky, A. R., Rees, C. W., Scriven, E. F. V.;
Elsevier: Oxford, 1996; chapter 1.33, pp 1041-1082.
(13) Adam, W.; Zinner, K. In Chemical and Biological Generation
of Electronic Excited States; Adam, W., Cilento, G., Eds.; Academic
Press: New York, 1982; chapter 5, pp 153-189.
Dioxetan e Decom position Kin etics: Relative Tr ip-
let-Excita tion Efficien cy. The rates of the dioxetane
(10) Yip, R. W.; Dunston, J . M.; Cessna, A. J .; Sugamori, S. E. Trans.
Faraday. Soc. 1971, 67, 3149-3154.
(11) Steenken, S.; J aenicke-Zauner, W.; Schulte-Frohlinde, D. Pho-
tochem. Photobiol. 1975, 21, 21-26.
(14) Wilson, T.; Schaap, A. P. J . Am. Chem. Soc. 1971, 93, 4126-
4136.

Photooxidative Damage of Guanine in DG and DNA
J . Org. Chem., Vol. 66, No. 2, 2001 599
Ta ble 2. Absor p tion Ch a r a cter istics of th e Keton esa
b
CH3COCH2X, X )
λmax (nm)
ꢀmax (L/mol cm)
H
OH
265
270
269
268
8
8
11
19
OMe
OtBu
a
b
40 mM in 9:1 H2O:CH3CN, 20 °C. Error ( 10%.
50 °C and 15 h. For ease of comparison, the extent of d G
oxidation induced by the triplet-excited ketones gener-
ated in the dioxetane thermolysis was calculated relative
to that of HO-TMD, which was set to unity (Table 1). As
the data in Table 1 display, the alkoxymethyl-substituted
dioxetanes are slightly more efficient than HO-TMD in
oxidizing d G, but all oxymethyl-substituted dioxetanes
are as much as five to seven times more efficient than
the parent derivative TMD. Comparison of the relative
efficiency (% conversion) of the d G oxidation with the
relative efficiency of excited-state production (ETP) for the
various dioxetanes reveals that there is no direct cor-
F igu r e 1. Time profiles of the dependence of the d G (0.500
mM) conversion and dioxetane chemiluminescence in a 9:1
mixture of phosphate buffer (5.00 mM, pH 7.0) and aceto-
nitrile; [tBu O-TMD] ) 20.0 mM, after 1080 min (see arrow)
a further 1 equiv was added.
respondence (Table 1). Thus, while the ET values follow
P
the order TMD > HO-TMD > MeO-TMD > tBu O-TMD,
their efficiency of d G oxidation follows essentially the
opposite trend, i.e., TMD < HO-TMD ∼ MeO-TMD ∼
tBu O-TMD.
Sch em e 2. F or m a tion of Sp in Ad d u cts w ith
5,5-Dim eth yl-1-p yr r olin e N-Oxid e (DMP O) in th e
Th er m olysis of th e Dioxeta n es a n d P h otolysis of
th e Keton es
Since R cleavage has been proposed15 as the major
radical-generating process to account for the enhanced
d G oxidation in the thermolysis of HO-TMD compared
to TMD, the d G-oxidation experiments were also per-
formed by direct irradiation of the respective ketone
derived from the dioxetane thermolysis [20.0 mM ketone,
0.500 mM dG in a 9:1 mixture of phosphate buffer (5.00
mM, pH 7.0) and acetonitrile at 300 nm and 0 °C for 5
h]. The extent of d G conversion has been corrected for
the molar extinction coefficient of the ketone chro-
mophore to normalize the damage relative to the number
of excited species formed (Table 2).
Analogous to the dioxetanes with HO-TMD as refer-
ence, the relative oxidation efficiency has been taken
versus hydroxyacetone as reference (set to unity). The
ketone photolysis data show (Table 1, last column) that
while the 1-hydroxy-2-propanone and 1-tert-butoxy-2-
propanone possess the same relative efficiency to oxidize
d G as the corresponding triplet-excited ketones formed
in the dioxetane thermolysis (Table 1, fifth column),
1-methoxy-2-propanone is much less efficient (only ca.
one tenth). Acetone, which does not generate radicals
upon photolysis, did not induce any d G oxidation.
Ch em ilu m in escen ce Mea su r em en ts d u r in g th e
d G Oxid a tion . To assess whether the electronically
excited species generated from the dioxetanes are in-
volved in the observed oxidation of d G, the latter was
to check whether the d G oxidation could be induced once
more. Indeed, an increase in d G oxidation (% conversion)
was observed. Thus, the time profile of the d G oxidation
parallels that of the formation of excited carbonyl species
in the thermolysis of the dioxetanes: The more excited
species that are formed, the more d G is oxidized.
Sp in -Tr a p p in g EP R Stu d ies in th e Dioxeta n e
Th er m olysis a n d Keton e P h otolysis. The R cleavage
of triplet-excited R-alkoxymethyl-substituted ketones to
generate radicals is a well-known photochemical reac-
tion.15 Since such excited ketones are formed in the
thermolysis of alkoxymethyl-substituted dioxetanes, spin-
trapping experiments with DMP O (Scheme 2) were
performed to provide spectroscopic evidence for the
intervention of radicals in the thermal decomposition of
the dioxetanes examined here. On thermolysis of all the
alkoxymethyl-substituted dioxetanes in aqueous solution
t
exposed to Bu O-TMD thermolysis at 50 °C and the
chemiluminescence emission measured [monitored con-
tinuously on a Mitchell-Hastings photometer (Figure 1)]
in parallel with the rate of d G oxidation (at each point
in Figure 1, a sample was taken and the d G conversion
determined by HPLC analysis). The decomposition rate
of the dioxetane was within the experimental error the
same in the presence [t1/2(50 °C, H2O/CH3CN 9:1) ) 740
( 100 min] and absence [t1/2(50 °C, CH3CN) ) 580 ( 100
min] of d G. After 1080 min (cf. arrow in Figure 1), a
further equivalent of dioxetane was added to the solution
(15) Steenken, S.; J aenicke-Zauner, W.; Schulte-Frohlinde, D. Pho-
tochem. Photobiol. 1975, 21, 21-26.

600 J . Org. Chem., Vol. 66, No. 2, 2001
Adam et al.
Ta ble 3. EP R Da ta for th e DMP O Ad d u cts F or m ed in
th e Th er m olysis of th e Dioxeta n es or P h otolysis of th e
Keton esa
substrate
TMD
DMPO adduct
DMPO-yl• b
g factor
RN (G) RH (G)
2.0056
2.0056
2.0056
2.0052
2.0053
2.0056
2.0056
2.0054
2.0054
2.0055
15.8
15.7
15.4
16.1
15.3
16.0
15.3
15.9
15.2
15.2
23.2
22.7
18.5
21.6
19.0
22.6
18.8
22.6
18.7
18.7
•
HO-TMD
HOCH2
Ac•
•
tBuO-TMD
MeO-TMD
CH3COCH2OH
tBuOCH2
Ac•
•
MeOCH2
Ac•
•
HOCH2
Ac•
CH3COCH2OtBu
Ac• c
a
Thermolysis of the dioxetanes (20 mM in 19:1 H2O/acetone)
at 37 °C, photolysis of the ketones (100 mM in 9:1 H2O/CH3CN)
b
at 300 nm in the Rayonet photoreactor, [DMP O] ) 45 mM. See
structures in the Introduction. c Besides acyl radicals, no other
carbon-centered radical adducts have been detected.
(10% acetonitrile as cosolvent) in the presence of DMP O,
two characteristic doublet-of-triplet EPR signals were
observed for the DMP O adducts of the acetyl (g ) 2.0055
( 0.0002, RN ) 15.3 ( 0.1 G, RH ) 18.7 ( 0.2 G) and a
carbon-centered radical (g ) 2.0054 ( 0.0002, RN ) 15.9
( 0.2 G, RH ) 22.1 ( 0.5 G, cf. Table 3). Without
dioxetane or the corresponding ketone, no EPR signals
were observed under the same experimental conditions
of the thermolysis (or photolysis). Analogous spectra have
been obtained upon irradiation of 1-hydroxy-2-propanone
and 1-tert-butoxy-2-propanone (data for acetyl and other
carbon-centered radicals see Table 3). The fact that the
acetyl radical has been detected in the R cleavage of the
triplet-excited carbonyl products, generated either in the
thermolysis of the dioxetanes or in the photolysis of the
corresponding ketones, implies that the other observed
carbon-centered radicals, should be the respective R-
alkoxymethyl species. The EPR parameters (g, RN, RH)
for these species in Table 3 are consistent with this
supposition. For the photolysis of 1-methoxy-2-pro-
panone, no radicals were detected in the spin-trapping
experiments.
Oxid a tion of d G. To assess the oxidation products of
d G in the thermolysis of the alkoxymethyl-substituted
dioxetanes, d G was treated with the dioxetanes at 50 °C
for 15 h. For this purpose, the (4R*)- and (4S*)-4-HO-8-
oxod G diastereomers and 8-oxod G were determined by
established HPLC methods16 and the guanidine-releasing
products (GR P ) were quantified by a fluorescence-
labeling HPLC assay with 1,2-naphthoquinone-4-sulfonic
acid).17 The results are shown in Figure 2 and exhibit
that the extent of d G oxidation (given as % convn)
decreases in the order MeO-TMD (71.7 ( 0.9%), HO-
TMD (49.0 ( 0.4%), tBu O-TMD (24.0 ( 0.2%) and TMD
(6.8 ( 0.2%). Except acetone, this order correlates within
the experimental error with the efficiency of the forma-
tion of triplet-excited carbonyl compounds in the ther-
molysis of the alkoxymethyl-substituted dioxetanes (Table
1).
F igu r e 2. d G (0.500 mM) oxidation (upper) in the thermolysis
(37 °C for 15 h) of the substituted dioxetanes (20.0 mM) and
(lower) in the photolysis (300 nm at 0 °C for 5 h) of the
substituted ketones (20.0 mM) in a 9:1 mixture of phosphate
buffer (5.00 mM, pH 7.0) and acetonitrile.
products (GRP ). Note that the values for 8-oxod G in
Figure 2 have been multiplied by 100 and, thus, this
product is formed only in trace amounts. For the most
efficient dioxetane, HO-TMD, at most only 0.2%
8-oxod G has been detected.
The photolysis of the corresponding ketones in the
presence of d G displays similarities in the oxidation
efficacy to the dioxetane thermolysis in that the oxyfunc-
tionalized ketones are much more effective than acetone.
Thus, the oxidation of d G (% convn) decreases in the
order 1-tert-butoxy-2-propanone (43.4 ( 0.6%), 1-hydroxy-
2-propanone (13.5 ( 5.8%), 1-methoxy-2-propanone (1.8
( 0.2%), and acetone (<0.5%); the latter is ineffective for
the oxidation of d G. For all ketone photolyses, except the
photoinert acetone, the major oxidation product is of the
guanine-releasing type (i.e. GRP ), while 4-HO-8-oxod G
is formed in small amounts, and 8-oxod G only in traces.
Since the (4R*)- and (4S*)-4-HO-8-oxod G diastereo-
mers have been described in the literature18 as typical
type-II photoproducts, the d G oxidation was conducted
The main oxidation products are the (4R*)- and (4S*)-
4-HO-8-oxod G diastereomers and guanidine-releasing
(16) (a) Floyd, R. A.; Watson, J . J .; Wong, P. K.; Altmiller, D. H.;
Richard, R. C. Free Rad. Res. Commun. 1986, 1, 163-172. (b) Ravanat,
J .-L.; Douki, T.; Incardona, M. F.; Cadet, J . J . Liquid Chromatogr.
1993, 16, 3185-3202.
(17) Ravanat, J .-L.; Berger, M.; Benard, F.; Langlois, R.; Ouellet,
R.; van Lier, J . E.; Cadet, J . Photochem. Photobiol. 1992, 55, 809-
814.
(18) Adam, W.; Saha-Mo¨ller, C. R., Scho¨nberger, A. J . Am. Chem.
Soc. 1997, 119, 719-723.

Photooxidative Damage of Guanine in DG and DNA
J . Org. Chem., Vol. 66, No. 2, 2001 601
F igu r e 3. Oxidation products of CT DNA (100 mg/mL) for
the thermolysis of the dioxetanes in a 9:1 mixture of phosphate
buffer (5.00 mM, pH 7.0) and acetonitrile at 37 °C and 12 h.
F igu r e 4. Effect of radical scavengers on the oxidation of CT
DNA (100 mg/mL) for the Bu O-TMD (20.0 mM) thermolysis
(37 °C for 12 h) and for the ketone (20.0 mM) photolysis (300
nm) in a 9:1 mixture of phosphate buffer (5.00 mM, pH 7.0)
and acetonitrile.
t
in D2O, in which singlet oxygen has a 10-fold longer
lifetime.19 However, the amounts of the (4R*)- and (4S*)-
4-HO-8-oxod G diastereomers were the same within the
experimental error in D2O and H2O; thus, singlet oxygen
is not directly involved in these photooxidations of d G.
Addition of 2 mM BHT (2,6-Di-tert-butyl-4-methylphenol)
ucts down to 2.16 ( 0.27% for 8-oxoGu a and 0.19 (
0.03% for GRP . Similar results were obtained in the
photolysis of 1-tert-butyl-2-propanone in the presence of
calf-thymus DNA. The radical scavenger BHT inhibited
also efficiently the formation of oxidation products
[8-oxoGu a was reduced from 2.19 ( 0.28% (0.79 ( 0.14%
GRP ) to 0.87 ( 0.44% (0.15 ( 0.01% GRP )], as shown
in Figure 4. The main oxidation product was 8-oxoGu a ,
but also substantial amounts of guanidine-releasing
products (GRP ) were observed (8-oxoGu a /GRP ratio ca.
2:1). Photolysis of acetone was again ineffective in the
formation of DNA oxidation products (0.14 ( 0.12%
8-oxoGu a and 0.01 ( 0.02% GRP ).
t
to the d G oxidation experiment with Bu O-TMD caused
a reduction of d G conversion by 59% whereas 1 mM BHT
reduced the d G conversion by 28%.
Oxid a tion of Ca lf-Th ym u s DNA by th e Dioxeta n e
tBu O-TMD. Isolated calf-thymus DNA (0.100 mg/mL
which corresponds to a 62.5 µM guanine concentration)
was treated at 37 °C for 21 h with the alkoxymethyl-
t
substituted dioxetane Bu O-TMD to determine the gua-
nine oxidation products. For comparison, the dioxetanes
TMD (does not release radicals)9a and HO-TMD (gener-
ates carbon-centered radicals)9a have been also employed
(Figure 3). For all these dioxetanes, the 8-oxoguanine (8-
oxoGu a ) and guanidine-releasing products (GRP ) were
determined. Comparison of the yields of oxidation prod-
ucts for HO-TMD (1.60 ( 0.16% 8-oxoGu a , 1.60 ( 0.16%
Discu ssion
The similar qualitative trends, with the exception of
the methoxy derivatives (shall be discussed later), in the
oxidative reactivity toward d G for the photolysis of the
ketones (tBuO ∼ HO > H) and for the thermolysis of the
dioxetanes (tBuO ∼ HO > H) imply that electronically
excited ketones are responsible for the oxidation of d G
in both processes. Indeed, the time profile in Figure 1
reflects that the d G conversion parallels the chemilumi-
nescence intensity, which is taken as a measure of
excited-state formation during the thermolysis of tBu O-
TMD. For the dioxetane thermolysis, the trend in the
efficiency of d G oxidation (with the exception TMD)
parallels the triplet excitation flux (Table 1), while for
the ketone photolysis this qualitative trend coincides
reasonably well (notably acetone is out of line) with the
efficiency of singlet excitation (the molar extinction
coefficients in Table 2 are taken as measure). The
disparity in the present data, i.e., the inefficiency of
excited acetone (generated by direct photoexcitation or
by TMD photolysis) and methoxyacetone (generated
photochemically) to oxidize d G suggests that excited
states (singlet or triplet) are not directly involved in the
d G oxidation. Presumably other species generated by or
from the excited ketones serve as oxidants, whether
produced by photoexcitation of the ketones or by ther-
molysis of the dioxetanes.
t
GRP ) versus Bu O-TMD (1.56 ( 0.01% 8-oxoGu a , 0.43
( 0.04% GRP ) discloses that both dioxetanes are ap-
proximately of equal oxidative reactivity toward DNA,
while TMD is inactive (20 mM: 0.17 ( 0.01% 8-oxoGu a ,
0.04 ( 0.03% GRP ). Again, this does not correspond with
the fact that the yield of triplet-excited ketone follows
t
the order TMD > HO-TMD > Bu O-TMD (Table 1).
With increasing concentration of tBu O-TMD it was found
that the amount of 8-oxoGu a increases as well, i.e., at
20 mM there are formed 2.58 ( 0.12% of 8-oxoGu a and
at 10 mM 1.56 ( 0.01% of 8-oxoGu a (Figure 3).
Radical-scavenging experiments (Figure 4) revealed
that in the presence of 2,6-di-tert-butyl-4-methylphenol
(BHT, 2 mM) or DMP O (2 mM) most of the DNA
oxidation was inhibited [8-oxoGu a drops from 3.20 (
0.20% (0.40 ( 0.07% GRP ) to 0.05 ( 0.04% (0.04 ( 0.03%
GRP ) with BHT and to 0.36 ( 0.02% (0.02 ( 0.02%
GRP ) with DMP O]. 2-Propanol (100 mM) was less
effective and reduced the amount of the oxidation prod-
(19) (a) Ogilby, P. R.; Foote, C. S. J . Am. Chem. Soc. 1983, 105, 5,
3423-3430. (b) Rodgers, M. A. J . J . Am. Chem. Soc. 1983, 105, 6201-
6205.

602 J . Org. Chem., Vol. 66, No. 2, 2001
Adam et al.
Sch em e 3. Nor r ish -Typ e-I a n d Typ e-II P h otoch em ica l Tr a n sfor m a tion s of Sin glet- a n d Tr ip let-Excited
Meth oxya ceton e
A possibility could be singlet oxygen (1O2), since it is
well established that excited ketones (generated by
photoexcitation of ketones or by chemiexcitation from
dioxetanes) afford this oxidant by energy transfer.20
However, this is unlikely to be a significant pathway
because control experiments in D2O for HO-TMD as well
firm this, since we have failed to detect spin-trapping
products in the photolysis of acetone and thermolysis of
TMD. In contrast, the oxyfunctionalized dioxetanes all
gave spin-trapped radicals with DMP O (Table 1) in
nearly the same efficiency and, thus, also nearly the same
t
extent of d G oxidation has been observed for Bu O-TMD,
t
as Bu O-TMD gave the same extent of d G oxidation as
HO-TMD and MeO-TMD. This correspondence is also
displayed for the ketone photolysis, except methoxy-
acetone. For this oxyfunctionalized ketone, no spin-
trapped radicals with DMP O were observed in its
photolysis and, correspondingly, only negligible d G con-
version was found (Table 1). This discord in the photo-
chemical reactivity of methoxyacetone requires mecha-
nistic rationalization.
in H2O. Since the lifetime of 1O2 in D2O is more than 10-
fold21 longer, the conversion of d G should have been
notably higher in this medium. Additionally, for HO-
TMD it was shown that this dioxetane is quite inefficient
in producing singlet oxygen,22 hardly sufficient to oxidize
d G to the extent observed in Figure 2.
In view of the photochemical property of appropriate
n,π*-excited ketones to undergo efficient R cleavage into
a pair of acyl and alkyl radicals (Scheme 1),23 these
reactive intermediates are likely to serve as oxidants,
especially in the presence of molecular oxygen. In aque-
ous media, sufficient dioxygen is present to trap such
radicals quite efficiently and generate oxygen-centered
derivatives (peroxyl, alkoxyl, hydroxyl), which effectively
oxidize d G.24 Indeed, experimental evidence for radicals
formed through R cleavage of the excited ketone (Scheme
2) has been provided through spin-trapping experiments
with DMP O and EPR-spectral detection (Table 1), as
well as the inhibition of d G and DNA oxidation (Figure
4) by BHT as radical scavenger. Certainly, the observed
qualitative trends in the d G oxidation by the photo-
excited ketones (tBuO ∼ HO > MeO ∼ H) and by the
dioxetanes (tBuO ∼ HO ∼ MeO > H) are in reasonable
accord with the propensity of radical release by the
excited ketones generated in the ketone photolysis (tBuO
∼ HO > MeO ∼ H) and in the dioxetane thermolysis
(tBuO ∼ HO ∼ MeO > H). Thus, we conclude that radical
formation from the excited ketones is responsible for the
d G oxidation; nevertheless, for both the ketone and the
dioxetane series, excited acetone is inefficient in oxidizing
d G. But it is an established fact that excited acetone
suffers negligible R cleavage;25 our present results con-
For R-methoxy-substituted ketones, the Norrish-Type-
II photoreaction may be expected as a competitive process
for R cleavage.26 Particularly for aliphatic ketones, this
process occurs efficiently from the singlet state.27,28 When
the ketone contains a â-oxygen atom, as is the case for
methoxyacetone, the Norrish-Type-II reaction is espe-
cially accelerated29 (AM1 calculations).30 Consequently,
the fact that no radicals have been detected by spin-
trapping in the direct irradiation of methoxyacetone and,
thus, no d G oxidation was found under these consitions,
implicates preferential Norrish-Type-II rather than R
cleavage (Scheme 3). Indeed, the expected acetone and
formaldehyde (its hydrate) as the photolysis products
were observed. In contrast, in the thermolysis of the
dioxetane MeO-TMD, in which the formation of the
triplet-excited methoxyacetone prevails, R cleavage domi-
nates, as manifested by the EPR-spectrally observed
spin-trapped radicals and the concomitant d G oxidation.
(25) Turro, N. J . In Modern Molecular Photochemistry; Benjamin/
Cummings Publishing Co.: Menlo Park, CA, 1978; pp 528-539.
(26) (a) Collins, P. M.; Gupta, P. J . Chem. Soc., Chem. Commun.
1969, 90-91. (b) Arnould, J . C.; Pete, J . P. Tetrahedron Lett. 1972,
13, 2415-2418. (c) Yates, P.; Szabo, A. G. Tetrahedron Lett. 1965, 6,
485-488.
(27) (a) Michael, J . L.; Noyes J r., W. A. J . Am. Chem. Soc. 1963, 85,
1027-1032. (b) Wagner, P. J .; Hammond, G. S. J . Am. Chem. Soc.
1965, 87, 4009-4011.
(28) (a) Dalton, J . C.; Dawes, K.; Turro, N. J .; Weiss, D. S.; Barltrop,
J . A.; Coyle, J . D. J . Am. Chem. Soc. 1971, 93, 7213-7221. (b) Wagner,
P. J .; Spoerke, R. W. J . Am. Chem. Soc. 1969, 91, 4437-4440.
(29) The Norrish-Type-II cleavage depends on the bond dissociation
energy of the γ-CH bond: For >95 kcal/mol, intersystem crossing is
faster than Norrish-Type II, while for <94 kcal/mol the Norrish-Type
II competes with intersystem crossing. Since a â-oxygen atom is
expected to lower γ-CH bond energy by ca. 9-12 kcal/mol (value
reported for CH2OH versus CH3; see: Benson, S. W. J . Chem. Educ.
1965, 42, 502-518, Kerr, J . A. Chem. Rev. 1966, 66, 465-500), a value
<90 kcal/mol is estimated for 1-methoxy-2-propanones (compared to
98 kcal/mol for 2-pentanone) and, thus, the Norrish-Type-II reaction
from the singlet state is likely.
(20) Wu, K. C.; Trozzolo, A. M. J . Photochem. 1979, 10, 407-410.
(21) (a) Rodgers, M. A. J . J . Am. Chem. Soc. 1983, 105, 6201-6205.
(b) Schmidt, R. J . Am. Chem. Soc. 1989, 111, 6983-6987. (c) Parker,
J . G.; Stanbro, W. D. J . Photochem. 1984, 25, 545-547.
(22) Briviba, K.; Saha-Mo¨ller, C. R.; Adam, W.; Sies, H. Biochem.
Mol. Biol. Int. 1996, 38, 647-651.
(23) (a) Paul, H.; Fischer, H. Helv. Chim. Acta 1973, 56, 1575-1594.
(b) Steenken, S.; J aenicke-Zauner, W.; Schulte-Frohlinde, D. Photo-
chem. Photobiol. 1975, 21, 21-26.
(24) (a) Maillard, B.; Ingld, K. U.; Scaiano, J . C. J . Am. Chem. Soc.
1983, 105, 5095-5099. (b) Brown, C. E.; Neville, A. G.; Rayner, D. M.;
Ingold, K. U.; Lusztyk, J . Aust. J . Chem. 1995, 48, 363-379. (c) von
Sonntag, C.; Schuchmann, H.-P. Angew. Chem., Int. Ed. Engl. 1991,
30, 1229-1253.
(30) Sengupta, D.; Bhattacharyya, A.; Sumathi, R.; Chandra, A. K.
J . Photochem. Photobiol. A: Chem. 1995, 86, 161-170.

Photooxidative Damage of Guanine in DG and DNA
J . Org. Chem., Vol. 66, No. 2, 2001 603
Sch em e 4. P ostu la ted Mech a n ism for th e Oxid a tion of DG by P er oxyl Ra d ica ls to th e 4-HO-8-oxod G a n d
GTP P r od u cts
Thus, it is concluded that in the direct photolysis experi-
ments of methoxyacetone, the observed Norrish-Type-II
cleavage proceeds from the singlet state, which competes
effectively with intersystem crossing and subsequent R
cleavage from the triplet state. For all the other photo-
excited ketones a similar R cleavage and oxidative
reactivity as for dioxetane thermolysis was observed.
These results make evident the advantage of dioxetanes
for photobiological studies in that their thermolysis allows
to generate selectively triplet-excited ketones without
trespassing the singlet state.
The major products in the radical induced oxidation
of d G, both for the ketone photolysis and dioxetane
thermolysis (Figure 2), are 4-HO-8-oxod G and the
guanidine-releasing products (GRP ), while 8-oxod G is
formed in only minor amounts. This is unusual, because
4-HO-8-oxod G has been assigned as a characteristic
singlet-oxygen photoproduct of d G.31 However, we have
ruled out that sufficient singlet oxygen is formed under
the present conditions to account for the large amounts
of observed 4-HO-8-oxod G. More likely, the 4-HO-8-
oxod G product is formed by oxidation of d G with radical
species, presumably by the peroxyl radicals generated in
the oxygenation of the carbon-centered radicals (Scheme
1), which are derived from the R cleavage of the elec-
tronically excited ketones. The production of 4-HO-8-
oxod G has also been observed in the oxidation of d G by
enzymatically generated peroxyl radicals.32 Since the C-8
and C-4 positions of guanine are known to be the major
sites of attack by hydroxyl radicals,33 presumably the
same two sites are also involved in the oxidation by
peroxyl radicals. We speculate (Scheme 4) that the initial
attack of a peroxyl radical at the C-8 position of d G would
generate a carbon-centered radical, which has two pos-
sibilities for further reaction: The first one is addition
of dioxygen,34 the second one formation of an oxaziridine
by loss of an alkoxyl radical, as known for peroxyl adducts
to simple alkenes.35 The latter oxaziridine intermediate
may rearrange rapidly to form 8-oxod G,36 while the
dioxygen adduct may be reduced to 4-HO-8-oxod G. We
conjecture that the peroxyl-radical adduct at the C-4 site
of d G has also at least two options: The oxirane
intermediate may be transformed by a peroxyl radical
to 4-HO-8-oxod G, while the dioxygen adduct leads
eventually by a complex sequence of transformations,
similar to those postulated in the Type-I oxidation of d G,
to the oxazolone (GRP ).
As for the 8-oxod G product, proposed to be generated
from d G by C-8 attack (Scheme 4), it is not expected to
survive the present oxidative conditions in view of its low
oxidation potential.37 Thus, we postulate electron transfer
from 8-oxod G to a peroxyl radical to generate the
corresponding radical cation. The latter is subsequently
transformed to 4-HO-8-oxod G and GRP (Scheme 4), for
(33) (a) Wetmore, S. D.; Boyd, R. J . J . Phys. Chem. B 1998, 102,
9332-9343. (b) Cadet, J .; Delatour, T.; Douki, T.; Gasparutto, D.;
Pouget, J .-P.; Ravanat, J .-L.; Sauvaigo, S. Mutat. Res. 1999, 424, 9-21.
(34) Cadet, J .; Berger, M.; Buchko, G. W.; J oshi, P. C.; Raoul, S.;
Ravanat, J .-L. J . Am. Chem. Soc. 1994, 116, 7403-7404.
(35) (a) Mayo, F. R. J . Am. Chem. Soc. 1958, 80, 2497-2500.
(36) Schmitz, E. Adv. Heterocycl. Chem. 1979, 24, 63-82.
(37) Steenken, S.; J ovanovic, S. V.; Bietti, M.; Bernhard, K. J . Am.
Chem. Soc. 2000, 122, 2373-2374.
(31) (a) Ravanat, J .-L.; Berger, M.; Benard, F.; Langlois, R.; Ouellet,
R.; van Lier, J . E.; Cadet, J . Photochem. Photobiol. 1992, 55, 809-
814. (b) Sheu, C.; Foote, C. S. J . Am. Chem. Soc. 1993, 115, 10446-
10447. (c) Sheu, C.; Foote, C. S. J . Am. Chem. Soc. 1995, 117, 474-
477.
(32) (a) Adam, W.; Kurz, A.; Saha-Mo¨ller, C. R. Chem. Res. Tox.,
submitted.

604 J . Org. Chem., Vol. 66, No. 2, 2001
Adam et al.
which there exist precedents.38,39 It is, therefore, not
surprising that only traces of 8-oxod G have been ob-
served under the present oxidative conditions. Indeed, a
control experiment with authentic 8-oxod G confirmed
that it was oxidized in the dioxetane thermolysis. Had
8-oxod G been formed in our oxidations, it would not
have accumulated and, thus, its intervention is uncertain.
In contrast to d G, the oxidation of DNA affords the
8-oxoGu a as the major product, but also GRP . As for
4-HO-8-oxoGu a , this oxidation product is difficult to
quantify in DNA, and to date no data for 4-HO-8-
oxoGu a in DNA have been reported. It is a general fact
that in the oxidation of DNA, the 8-oxoGu a is formed
in substantial amount, whereas in d G only in traces. For
example, in the hydroxyl-radical-induced oxidation of
DNA, the main oxidation product is 8-oxoGu a , but for
d G it is the GRP .33 This reactivity trend applies also in
the case of one-electron oxidants, which has been ra-
tionalized in terms of the intermediacy of the stabilized
guanine radical cation in the DNA matrix; subsequently,
the radical cation is hydrolyzed selectively to 8-oxoGu a
prior to deprotonation and formation of GRP .40 Since
under our conditions the 8-oxoGu a in DNA accumu-
lates but the 8-oxod G in d G is oxidized rapidly, the high
levels of 8-oxoGu a in DNA are due to its protection in
the DNA matrix against further oxidation. As an alter-
native to the above-mentioned electron-transfer process,
the 8-oxoGu a in DNA might be formed by the same
mechanism as already postulated for the formation of the
8-oxod G in the d G oxidation in Scheme 4, that is, by
the addition of a peroxyl radical to the C-8 position,
generation of an oxaziridine, and rearrangement of the
latter to 8-oxoGu a . Besides 8-oxoGu a , the large amount
of GRP is in accord with radical-type oxidation of DNA.41
Con clu sion
This study reveals that the major radical-induced
oxidation products of d G are GRP and 4-HO-8-oxod G,
while 8-oxod G is formed in only minor amounts. The
major oxidation product in DNA is 8-oxoGu a , but also
significant amounts of GRP are found. The observed
guanine oxidation in d G and in DNA is attributed to
radical species, in particular peroxyl radicals. The latter
are formed by dioxygen addition to the carbon-centered
radical species, which are generated upon R cleavage of
the excited ketones derived from photoexcitation (ketone
photolysis) or chemiexcitation (dioxetane thermolysis).
Ack n ow led gm en t. We are grateful for generous
funding by the Deutsche Forschungsgemeinschaft
(Sonderforschungsbereich 172: Molekulare Mechanis-
men kanzerogener Prima¨rvera¨nderungen) and the Fonds
der Chemischen Industrie.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the preparation of the key compounds and their
spectral data as well as the procedures for the quantitative
determination of the d G and DNA oxidation products. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(38) (a) Luo, W.; Muller, J . G.; Rachlin, E. M.; Burrows, C. J . Org.
Lett. 2000, 2, 613-616. (b) Hickerson, R. P.; Prat, F.; Muller, J . G.;
Foote, C. S.; Burrows, C. J . J . Am. Chem. Soc. 1999, 121, 9423-9428.
(39) Adam, W.; Saha-Mo¨ller, C. R.; Scho¨nberger, A. J . Am. Chem.
Soc 1996, 118, 9233-9238.
(40) (a) Hutter, M.; Clark, T. J . Am. Chem. Soc. 1996, 118, 7574-
7577. (b) Angelov, D.; Spassky, A.; Berger, M.; Cadet, J . J . Am. Chem.
Soc. 1997, 119, 11373-11380.
J O0056491
(41) Cadet, J .; Vigny, P. In Bioorganic Photochemistry; Morrison,
H., Ed.; J ohn Wiley & Sons: New York, 1990; Chapter 3, pp 184-
210.