©1993 Oxford University Press Nucleic Acids Research, 1993, Vol. 21, No.

4 863-869

Endonuclease-sensitive DNA modifications induced by

acetone and acetophenone as photosensitizers
Bemd Epe1, Hedda Henzl, Waldemar Adam1 and Chantu R.Saha-Moller1
Institute of Pharmacology and Toxicology and institute of Organic Chemistry, University of Wurzburg,
Versbacher Str. 9, D-8700 Wurzburg, Germany

Received November 27, 1992; Revised and Accepted January 19, 1993

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

ABSTRACT
Repair endonucleases, viz. endonuclease III, by decomposition of lipid peroxides, e.g. under conditions of
formamidopyrimldlne-DNA glycosylase (FPG protein), oxidative stress (1—4).
endonuclease IV, exonuclease III and UV The triplet energy of many ketones is higher than that of DNA
endonuclease, were used to analyse the modifications bases; therefore, excited ketones can react with DNA by triplet
Induced In bacterlophage PM2 DNA by 333 nm laser energy transfer (reviewed in Refs. 5,6). This has been shown
Irradiation In the presence of acetone or acetophenone. to result in UV-type DNA damage, viz. the formation of
In addition to pyrlmldlne dlmers sensitive to UV cyclobutane pyrimidine dimers (7—9). Interestingly, (6-4)
endonuclease, 5,6-dihydropyrlmldlnes (sensitive to photoproducts, which make up approx. 25% of the modifications
endonuclease III) and base modifications sensitive to induced by direct excitation of DNA at 260 nm, are not formed
FPG protein were generated. The level of the last In the in parallel, possibly because these emanate from the singlet
case of acetone was 50% and in the case of excited DNA bases (10). DNA damage by energy transfer from
acetophenone 9% of the level of pyrlmidine dimers. excited acetone can occur under cellular conditions, since the
HPLC analysis of the bases excised by FPG protein phototoxicity induced by 313-nm radiation in the presence of
revealed that least some of them were 8-hydroxy- acetone in E.coli is subject to photoreactivation (11).
guanine (7,8-dihydro-8-oxoguanine). In the damage Besides energy transfer, other reactions of triplet excited
induced by direct excitation of DNA at 254 nm, which carbonyl compounds might generate DNA damage as well. Thus,
was analysed for comparison, the number of FPG the radical nature of the excited triplet states should allow electron
protein-sensitive base modifications was only 0.6% of or hydrogen transfer reactions, which could initiate radical-type
that of the pyrlmidine dimers. Mechanistic studies DNA damage. In addition, an energy transfer to molecular
demonstrated that the formation of FPG protein- oxygen could generate singlet oxygen, die genotoxicity of which
sensitive modifications did not involve singlet oxygen, is well established (12,13) To date, there is only limited
as the damage was not Increased in D2O as solvent. experimental evidence for these reactions. Thus, the formation
Hydroxyl radicals, superoxlde and H2O2 were also not of pyrimidine dimers from excited ketones is accompanied by
involved, since the relative number of single strand the generation of DNA strand breaks, and the ratio of the two
breaks and of sites of base loss (AP sites) was much types of modification varies with the availability of oxygen
lower than in the case of DNA damage Induced by (9,14,15). In addition, alkali-labile modifications of guanine
hydroxyl radicals and since the presence of SOD or residues were observed after exposure of DNA to UV-radiation
catalase had no effect on the extent of the damage. (313 nm) in the presence of acetone (16) and photodegradauon
However, the mechanism did involve an intermediate of 2'-deoxyguanosine has been observed under the same
that was much more efficiently quenched by azide Ions conditions (17).
than the triplet excited carbonyl compounds and which Here, we describe analysis by means of specific repair
was possibly a purine radical. Together, the data endonucleases of the DNA modifications induced by acetone and
Indicate that excited triplet carbonyl compounds react acetophenone in the presence of 330-nm laser radiation. The
with DNA not only by triplet-triplet energy transfer results indicate that the DNA damage consists not only of
yielding pyrimldine dimers, but also by electron transfer pyrimidine dimers, but also of a large number of modifications
yielding preferentially base modifications sensitive to recognized by FPG protein (formamidopyrimidine-DNA
FPG protein, which Include 8-hydroxyguanine. glycosylase), at least some of which are 8-hydroxyguanine
(7,8-dihydro-8-oxoguanine) residues.
INTRODUCTION
MATERIALS AND METHODS
The triplet excited states of carbonyl compounds such as acetone
or acetophenone are reactive, short-lived diradicals. In cellular Materials
systems, they can be generated not only by direct excitation of DNA from bacteriophage PM2 (PM2 DNA) was prepared
the ground state compounds at wavelengths between 300 and 360 according to the method of Salditt et al. (18). More than 95%
nm, but also in the dark by oxidation with certain enzymes and was in the supercoiled form, as determined by the method
864 Nucleic Acids Research, 1993, Vol. 21, No. 4

Table L Recognition spectrum of the repair endonucleases used in this study *

Repair endonuclease Sites of base loss (AP sites) Base modifications1'

regular0 l'-oxid. d 4'-oxid. e

FPG protein 8-hydroxyguanine, formamidopyrimidines

Endonuclease III 5,6-dihydropyrimidines
UV endonuclease cyclobutane pyrimidine photodimers
Endonuclease IV
Exonuclease III

* as identified to date; see Refe. 1 9 - 2 2 and, for reviews, Refs. 2 8 - 3 1 .

b
see Fig. 1 for chemical structures
c
unmodified desoxyribose moiety
d
desoxyribose oxidized in the 1' position
e
desoxyribose oxidized in the 4' position
r
recognition requires high enzyme concentrations (200 U/ml)

described below. Formamidopyrimidine-DNA glycosylase (FPG

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

protein) (19) was obtained from Dr S.Boiteux, Villejuif. KH-CHO

Endonuclease HI (20) was provided by Dr R.P.Cunningham,

Albany. Endonuclease IV (21) was a gift from Dr B.Demple,
Boston. UV endonuclease was partially purified from M. luteus drab.
Fapy-guanosine
(22). Exonuclease HI was purchased from Boehringer, 8-hydroxyguanoslne
(7,ft-dihydro-8-oxoguano8ine)
Mannheim, Germany. All repair endonucleases were tested for
their incision at reference modifications (i.e. thymine glycols NH,
induced by OsO4, AP sites induced by low pH, pyrimidine
dimers induced by UV254) under the applied assay conditions
(see below) to ensure that the correct substrate modifications are
fully recognized and no incision at non-substrate modifications <UU>.
takes place (see Ref. 23). NDPO2 (disodium 3,3'-(l,4-naphthyl- Fapy-adenoslne 5,6-dihydiothymldlne*
idene) dipropionate endoperoxide) was synthesized as described
by Nieuwint et al. (24) and was approx. 90% pure. Acetone was
NH,
obtained from Roth (Karlsruhe, Germany) and was more than
99.5% pure. Acetophenone was obtained from Aldrich OH
(Steinheim, FRG) and was 99% pure. Superoxide dismutase OOH
(SOD) was purchased from Sigma Chemie (Deisenhofen, I
dRib.
F.R.G.). Catalase was obtained from Serva (Heidelberg,
5,6-dihydrocytldLnes cyclobutane thymidine
F.R.G.). Both enzymes were tested for activity. dimer

DNA modification
Flgnre 1. Chemical structures of base modifications recognized by the repair
The exposure of PM2 DNA (10 /ig/ml) to all agents was carried endonucleases used in this study (see Table I).
out in phosphate buffer (5 mM KH2PO4, 50 mM NaCl, pH
7.4). The exposure to UV radiation (254 nm), to NDPO2 and
to xanthine in the presence of xanthine oxidase (10 units/ml) and the following repair endonuclease preparations: (i) exonuclease
Feflll) -EDTA (100 /*M) has been described previously (23,25). m , 300 U/ml in 20 mM Tris-HCl pH 8.0, 100 mM NaCl, 15
The exposure to UV radiation (333 nm) in the presence and mM CaCl2. (") FPG protein, 3 /tg/ml in BE,5 buffer (BE,
absence of acetone or acetophenone was carried out in a 4 x 4 buffer containing 15 mM EDTA), (iii) endonuclease HI, 40 ng/ml
mm quartz tube by means of an argon ion laser (Innova 100, in BE15 buffer and (iv) endonuclease IV, 10 U/ml in BE,5
Coherent; output 2.5 W in most experiments), the 333-nm buffer. The reactions were stopped by addition of 3 /tl 10%
spectral line of which was isolated with a prism and widened sodium dodecyl sulfate and the DNA applied to an agarose slab
to yield approx. 2.9 kW/m2. In some of the experiments, electrophoresis gel. After electrophoresis and staining with
sodium azide ( 0 - 6 mM), SOD (60 U/ml) or catalase (280 U/ml) ethidium bromide, the relative amounts of the supercoiled and
was added or H2O in the buffer was replaced by D2O. In the the nicked form of the DNA were determined using a
last case, the pD of the buffer was adjusted according to Srere fluorescence scanner (FTR20, Sigma Instruments, Berlin). From
et al. (26); the final isotope purity was greater than 95%. The these values, the average number of single strand breaks per DNA
DNA was precipitated by ethanol/sodium acetate and redissolved molecule produced either directly by the damaging agent or by
in BE,-buffer (20 mM Tris-HCl pH 7.5, 100 mM NaCl, 1 the subsequent enzymatic incision at the endonuclease-sensMve
mM EDTA) for damage analysis. modifications was calculated by means of a Poisson equation as
described previously (25,27).
DNA relaxation assay
An aliquot of 0.3 /ig of the modified DNA in 20 /J BE, buffer Modification of calf thymus DNA and HPLC analysis
was incubated for 30 min at 37 °C with 10 /d of buffer (for the The exposure of calf thymus DNA (50 ng in 100 /tl phosphate
determination of directly produced strand breaks) or of one of buffer) to acetone (2.7 M) plus 333 nm laser radiation (200
 Nucleic Acids Research, 1993, Vol. 21, No. 4 865

scatoptwnone (6 mil]

I
• UV">

1-

300 400 20 40 60 80 1000 2000

Radiation dose [kJ/mz]

Figure 2. DNA modifications induced in PNC DNA by various doses of 333 nm UV radiation in the presence or absence of acetone (2.7 M) or acetophenone
(6 mM): DNA single strand breaks (O) and sum of single strand breaks plus (i) sites sensitive to FPG protein ( • ) , (ii) UV endonuclease (A), (iii) endonuclease
m ( • ) , fiv) endonuclease IV (D) and (v) exonuclease m (A).

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

I •cttophtnont
• UV" 1 psfc*taf)

I UV ndo.

Acatoplwnona concentration [mM]

< o.o- acetone acetophe- uv'254 xanthlne111
/
333
+ UV none + UV 3 3 3 x-oxJFe
Figure 3. DNA modifications sensitive to (i) UV endonuclease (A) and (ii) FPG •ndonucl>ii»-»fiiltlv modlflcitlona DMA »Und bTMkl
protein ( • ) induced by 333 nm UV irradiation (29 kj/m2) in the presence of • FPOproMn ffl tndonucttf IV H 9in0l# SJJBJKJ brocks
various concentrations of acetophenone. • •ndonuolMM a • «<
D UV •ndonudMM

kJ/m2) was carried out as decribed above for PM2 DNA. The Figure 4. DNA damage profiles: various endonuclease-sensitive modifications
DNA was precipitated, redissolved in 100 y\ BE{ buffer and (first to sixth column) and single strand breaks (seventh column) observed in PM2
DNA after treatment with (i) acetone (2.6 M) plus U V " 3 (86 J/m2) (ii)
incubated with 5 /ig FPG protein for 2h at 37CC. The excised acetophenone (6 mM) plus UV533 (8.6 J/trr3); (iii) UV234 (8.7 J/m2) fiv) NDPO,
bases were separated from the DNA on an anion exchange in ^ O buffer (3.5 mM, 2 h, 37°Q and (v) ranthine (8 11M) in the presence
minicolumn (Quiagen, USA) and applied to a reversed-phase of xanthine oxidase and Fe(in)-EDTA (30 min, 20°Q.
HPLC column (LC-18S, Supelco, USA.; 25 cm; 5-/tm particle
size) equipped with a UV detector operating at 290 nm and eluted
with 50 mM phosphate buffer pH 5.5, which contained 5% damage, much lower concentrations of acetophenone than of
methanol, at a flow rate of 1.2 ml/min. When authentic acetone were needed; only few modifications are generated by
8-hydroxyguanine was added to unmodified DNA before the UV333 alone in the absence of any photosensitizer (Fig. 2,
incubation with FPG protein, the recovery was found to be more Table II).
than 80%. In Fig. 4, the data are replotted in the form of DNA damage
profiles, which show the ratios of the various types of
RESULTS modifications induced. For comparison, Fig. 4 also depicts DNA
damage profiles which were observed after (i) direct excitation
DNA damage profiles of DNA by 254 nm UV irradiation, (ii) exposure to singlet
Supercoiled PM2 DNA (104 bp) was exposed to 333 nm laser oxygen, generated by thermal decomposition of NDPO2, (32)
radiation in phosphate buffer, both in the presence and absence and (iii) exposure to hydroxyl radicals, generated by xanthine
of acetone or acetophenone. The irradiated DNA was analysed and xanthine oxidase in the presence of Fe(III)-EDTA. These
for strand breaks and for modifications sensitive to various repair damage profiles have been described earlier (23,25,33,34).
endonucleases (see Table I and Fig. 1). As shown in Fig. 2, all The data (Fig. 4) indicate that the DNA damage induced by
types of modification increased linearly with irradiation time. A irradiation in the presence of acetone or acetophenone consists
linear increase of the extent of damage was also observed when predominantly of base modifications; both strand breaks and sites
the acetophenone concentration was varied between 0.1 and 10 of base loss (AP sites, recognized by exonuclease IE or
mM at constant irradiation time (Fig. 3). From the best-fit straight endonuclease IV) are relatively rare. Among the base
lines, the numbers of modifications per radiation dose were modifications detected, pyrimidine dimers (recognized by the UV
calculated (Table II). To induce a comparable extent of DNA endonuclease from M.luteus) are most frequent. The number of
866 Nucleic Acids Research, 1993, Vol. 21, No. 4

Table n . Endonuclease-sensitive modifications and single strand breaks induced in PM2 DNA by UV (333 nm) irradiation in the
presence and absence of photosensitizers

Photosensitizer Abs1 Modificationsb

FPG1 Endo IIId UVendo* Endo IV^ Exo OF SSBh

Acetone (2.7 M) 0.01 13.6±0.7 5.2 ±0.7 27.3 ±0.4 1.9±0.3 0.3 ±0.2 0.71 ±0.06
Acetophenone (6 mM) 0.10 26.5 ±1.3 6.3 ±1.4 291 ±23 2.1 ±0.3 1.5±0.3 2.1 ±0.1
None 0.00 0.07 ±0.02 0.01 ±0.02 0.40 ±0.04 0.04 ±0.03 0.01 ±0.03 0.027 ±0.008
1
absorbance [Ig (I</I)] per cm of the irradiated solution in comparison to the buffer alone.
b
number of modifications ( ± S.D.) per 104 base pairs and unit dose (J/mm2) calculated by linear regression from the data points
shown in Fig. 2.
c g
~ modifications sensitive to FPG protein (formamidopyrimidine-DNA glycosylase), endonuclease ID, UV endonuclease preparation
from M. luteus, endonuclease IV and exonuclease m, respectively
h
single strand breaks

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

1.0 1, . . .
1
'1 \
t
0.5
* i

•-
i
» UV«ndo. *
,1 1 10 100 1000 10000 • FPQpraWn

FPG protein concentration (ng/ml) an 1*-

H01 .01 10 100 0.0 0.1 0.2 0.3
Azlde c o n c e n t r a t i o n [mM]
Figure 5. Recognition by various concentrations of FPG protein of modifications
induced in PM2 DNA by exposure to acetone (2.6 M) plus UV 333 . Dam are
means of two experiments with UV doses of 86 and 172 J/m2. The number of
modifications recognized at 2 /ig/ml FPG protein is defined as 100%. Figure 6. Quenching by sodium azide of DNA modifications sensitive to FPG
protein ( • ) and to UV endonuclease (A) induced by acetone (2.6 M) plus UV333
(86 J/nr and 170 J/m2). Left panel: semilogarithmic plot of the relative number
base modifications sensitive to FPG protein is 50% of the number of DNA modifications versus the azide anion concentration (mM). Right panels:
Stem -VoLmer plots of the reciprocal of the relative number of DNA modifications
of pyrimidine dimers in the case of acetone, but only 9% in the versus the azide anion concentration.
case of acetophenone as photosensitizer. 5,6-Dihydropyrimidine
derivatives (recognized by endonuclease HI) are less frequent
(18% and 2% of the pyrimidine dimers in the case of acetone Table ID. Stem-Volmer quenching constants of azide anions for endonuclease-
and acetophenone, respectively). sensitive modifications
The presence of relatively large numbers of FPG-sensitive base
Damaging agent Quenching constants for
modifications in the photosensitizer-mediated DNA damage is modifications sensitive to1
in marked contrast to the damage induced by direct DNA exitation UVendo b FPG'
at 254 nm, where the number of FPG-sensitive modifications [mM" 1 ] [mM" 1
is only 0.6% of that of pyrimidine dimers (Fig. 4). On the other
hand, base modifications sensitive to FPG protein are the Acetone (2.7 M) plus UV333 0.14±0.02 38.4 ±0.9
NDPO^ (20 mM), 2 h, 37°C 1.44 ±0.08
predominant type of DNA damage detected after treatment with
singlet oxygen (generated by decomposition of NDPOj, Fig. 4). 1
calculated by linear regression from the data points shown in Figures 6 and 7
b
The damage profile induced by hydroxyl radicals (generated by UV endonuclease preparation from M.luteus
c
xanthine and xanthine oxidase in the presence of Fe(III)-EDTA) FPG protein (formamidopyrimidine-DNA glycosylase)
is characterized by large numbers of single strand breaks and
sites of base loss and is therefore completely different from the
damage produced by the photosensitizers (Fig. 4). The results demonstrate that recognition of these modifications
is saturated at 1000 ng/ml FPG protein too. Furthermore, die
Incision of base modifications by low concentrations of FPG enzyme concentration of 50 ng/ml at which 50% of the
protein modifications are recognized is similar to mat observed earlier
For the damage analysis described above, the chosen FPG protein for AP sites and the base modifications induced by singlet oxygen
concentration of 1 ;tg/ml has previously been demonstrated to (23).
be high enough to be, under die reaction conditions used, not
the limiting factor in the detection of either AP sites or the base Quenching by the azide anion
modifications induced by singlet oxygen (23). The recognition A comparison of the damage profiles shown in Fig. 4 gives rise
of the modifications induced by excited acetone with FPG protein to die conclusion that singlet oxygen could be responsible for
concentrations between 0.3 and 2000 ng/ml is shown in Fig. 5. the non-pyrimidine dimer part of the DNA modifications induced
 Nucleic Acids Research, 1993, Vol. 21, No. 4 867

TaWe IV. Effects of SOD, catalase and D2O as solvent on the endonuclease-sensra've modifications
induced by acetone (2.7 M) plus UV333

Modifications Relative number of modifications (%)*-b

sensitive to SOD (60 U/ml) Catalase (280 U/ml) DjO

FPG protein 90±23 (4) 98±46 (4) 101 ±36 (4)

UV endonuclease 137 ±39 (4) 115±27 (4) 93 ±33 (4)
Endonuclease HI 91 ±37 (3) 60±20(3) 73±34(3)

* Number of modifications observed in H2O buffer without SOD or catalase defined as 100%
b
Values represent means ±S.D.; the number of determinations are given in parentheses

• OH-Qui m e d l . DMA
(40 pan!) • FPQ proMa
1.0
i V

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

0.5

FPQpreMn

= c
= .001 .01 10 100 0 2

i Azhte concentration [mil]

\J 10
Figure 7. Quenching by sodium azide of DNA modifications sensitive to FPG 10 0 5 10
protein induced by NDPOj (20 mM, 2 h, 37°Q. Left panel: semilogarithmic elutton tlm* (mln)
plot of the relative number of DNA modifications versus the azide anion
concentration (mM). Right panels: Stern-Volmer plots of the inverse of the
Figure 8. HPLC chromatograms of (i) authentic 8-hydroxyguanine (40 prool,
relative number of DNA modifications versus the azide anion concentration.
left panel); (ii) a preparation of bases excised by FPG protein (50 /tg/ml) from
calf thymus DNA (50 ^g) after irradiation with UV333 (200 kl/m2) in the
by near-UV radiation in the presence of acetone and presence of acetone (2.7 M) (central panel) and (iii) a control preparation, in
acetophenone. The following experiments were carried out to which FPG protein was omitted (right panel).
test this hypothesis.
The azide anion is known as an efficient quencher of singlet in D2O, as expected (data not shown). The large D2O effect was
oxygen (35). The influence of various concentrations of sodium also observed when the modification by NDPO2 was carried out
azide on the generation of DNA modifications by acetone plus in the presence of 20% acetone (data not shown). It must be
UV333 is shown in Fig. 6. The slopes of Stern-Volmer plots concluded that the damaging mechanism in the case of the
(inverse of the relative damage versus the azide concentration) photosensitizers does not involve singlet oxygen, H2O2 or
afford quenching constants which are given in Table HI. The data superoxide.
indicate that azide ions suppress the generation of FPG-sensitive
modifications by acetone plus UV333 300-fold more efficiently HPLC analysis of the FPG-sensitive base modifications
than the generation of the pyrimidine dimers. For comparison, The modified DNA bases excised by FPG protein from calf
the effect of azide ions on the singlet oxygen-mediated damage thymus DNA were analysed by HPLC equipped with a UV
induced by NDPO2 was also measured (Fig. 7). Unexpectedly, detector (290 nm). For DNA that had been exposed to acetone
the quenching constant for the FPG-sensitive modifications plus UV333 a peak was detected which co-eluted with
produced with NDPO2 was only one thirtieth of that for the 8-hydroxyguanine (7,8-dihydro-8-oxoguanine) (Fig. 8). No such
FPG-sensitive modifications produced with acetone and UV333 peak was observed with DNA incubated in the absence of FPG
(Table ffl). protein (Fig. 8). When unmodified DNA is incubated with FPG
protein, a small peak was detected which corresponds to approx.
Effects of superoxide and catalase and of DjO as solvent 2.5 pmol 8-hydroxyguanine per 50 /tg DNA (not shown). The
DNA damage mediated by singlet oxygen is expected to increase amount of 8-hydroxyguanine excised from DNA after exposure
in D2O as solvent, since the life-time of singlet oxygen is to acetone plus UV333 was only approx. 15% of the amount of
approx. ten-fold greater in D2O than in H2O (36,37). Decreased FPG-sensitive base modifications calculated on the assumption
damage in the presence of superoxide dismutase (SOD) and that: (0 the number of FPG-sensitive base modifications increases
catalase would indicate the involvement of superoxide or H2O2 linearly with the radiation dose even at high doses, (ii) relatively
in the damaging mechanism. As shown in Table IV, neither SOD high levels of damage (approx. 10 pyrimidine dimers and 5 FPG-
and catalase nor the solvent change to D2O had a significant sensitive modifications per 103 base pairs) do not prevent
effect (within the experimental error) on the generation of incision by FPG protein at all sensitive base modifications, (iii)
pyrimidine dimers and FPG-sensitive base modifications by the DNA concentration does not influence the number of
acetone plus UV330. In contrast, the number of FPG-sensitive modifications induced per mole DNA and (iv) calf thymus DNA
modifications generated by chemically generated singlet oxygen is as reactive as PM2 DNA. No attempts were made to verify
(thermal decomposition of NDPOJ was approx. ten-fold greater these assumptions.
868 Nucleic Acids Research, 1993, Vol. 21, No. 4

DISCUSSION single strand breaks and AP sites, in contrast to the damage

induced by hydroxyl radicals, are infrequent lesions (Table IV,
The results confirm earlier findings that excited acetone and
Fig. 4). However, the reaction induced by excited acetone does
acetophenone induce UV-type damage in DNA, viz. die
involve an intermediate that is quenched by azide even better than
formation of cyclobutane pyrimidine dimers (7,9). Due to the
is singlet oxygen. This intermediate is not the triplet excited
higher absorption coefficient, acetophenone is much more
carbonyl compound, since the formation of pyrimidine dimers,
efficient at 333 ran as sensitizer than is acetone at equimolar
which are generated from the same species, is quenched much
concentrations; for solutions of equal absorbance, however, the
less efficiently (Table HI). Ratfier, die species might be a purine
yields of pyrimidine dimers would be expected to be similar (see
radical cation generated by electron transfer from the purine to
Table II)
the excited carbonyl compound (type I reaction, see Ref. 46).
In addition to pyrimidine dimers, other base modifications can Interestingly, this would imply that the formation of FPG-
be detected by means of repair endonucleases. Among these, base sensitive modifications by singlet oxygen does not involve the
modifications sensitive to FPG protein are most prominent. FPG- same intermediate, as the quenching constant in this case is
sensitive modifications have been shown to be generated in high significandy lower (Table IH). This is in agreement with a
yields by photosensitizers absorbing in the visible range of the mechanism suggested by Devasagayam et al. (41), in which
spectrum and by (chemically generated) singlet oxygen (23,38). singlet oxygen generates 8-hydroxyguanine via an unstable
Both in diese cases (39—41) and in the case of the acetone- endoperoxide intermediate.
mediated damage (Fig. 8), HPLC-analysis reveals that at least

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

some of the modifications are 8-hydroxyguanine residues. So far, The extent to which the endogenous formation of triplet excited
8-hydroxyguanine and formamidopyrimidines (imidazole ring- carbonyl compounds by enzymatic oxidation and lipid
opened purines) are the only base modifications known to be peroxidation could contribute to cellular DNA damage in the dark
recognized by FPG protein; however, other substrates cannot is difficult to estimate. On the other hand, there are indications
be excluded. A quantitative comparison of the yields of that the cytotoxic and mutagenic effects of near-UV irradiation
8-hydroxyguanine and FPG-sensitive modifications is problematic are mediated by (unknown) endogenous photosensitizers (for
as different damaging conditions had to be used for the two reviews, see Refs. 47,48). The miscoding properties of
determinations. The base modifications sensitive to FPG protein 8-hydroxyguanine have been demonstrated both under cellular
may possibly be identical with purine photoproducts which have and cell-free conditions (49-53). If triplet excited carbonyl
been found by others to be incised by an enzyme-containing compounds were directly involved in cellular DNA damage,
protein fraction from T4-infected E.coli after irradiation of DNA the results described here would predict diat mutations caused
with high UV doses at wavelengths between 260 and 300 nm (42). by 8-hydroxyguanine should occur in vivo. In particular, G-C
to T-A transversions should be detected in addition to the G-C
Besides base modifications sensitive to FPG protein, excited to A-T transitions caused by pyrimidine dimers. So far, the
acetone and acetophenone generate 5,6-dihydropyrimidines, mutations in the p53 tumor suppressor gene observed in human
which are sensitive to endonuclease HI. As in the case of the squamous cell carcinomas of the skin point exclusively to
FPG protein-sensitive modifications, their relative frequency pyrimidine dimers as the underlying DNA modifications (54).
compared to pyrimidine dimers is much higher than that observed Results wim other types of skin cancer or papillomas, in particular
after direct excitation of DNA by UV^» (Fig. 4; Table II). when induced by UVA (55) will help to clarify the role of DNA
Modifications sensitive to endonuclease HI have been observed damage induced by photosensitizers.
after treatment of DNA with high doses of UVC and UVB and
identified as pyrimidine hydrates (6-hydroxy-5,6-dihydro-
pyrimidines) (43—45). ACKNOWLEDGEMENTS
In contrast to the pyrimidine dimers, the FPG-sensitive
modifications cannot result from energy transfer (which gives We thank S.Boiteux for providing FPG protein, R.P.Cunningham
rise to excited DNA bases) as otherwise they should be generated for endonuclease IH and B.Demple for endonuclease IV. The
in similar yield when DNA is excited directly at 254 nm (Fig. 4). work was supported by the Wilhelm Sander-Stiftung and by the
Furthermore, the ratio of FPG-sensitive modifications and Deutsche Forschungsgemeinschaft (SFB 172).
pyrimidine dimers is five-fold higher for acetone than for
acetophenone (Table II, Fig. 4). It remains to be established
whether the higher energy of excited acetone or other parameters, REFERENCES
e.g. differing redox potentials, are responsible for this difference. 1. Cilento.G. (1984) Pure AppL Oiem., 56, 1179-1190.
Interestingly, in the damage induced by tetramethyl-l,2-di- 2. Cadenas.E. (1984) Photodian. Photobiol., 40, 823-830.
oxetane, a chemical source of triplet excited acetone, FPG- 3. Bander,WJ., Bohne.C, Cflento.G. and Dunfbrd.H.B. (1985) / BioL Chan.,
260, 10217-10225.
sensitive base modifications are even more frequent than 4. Sies, H. (1986) Angew. Chan. Int. Ed Engl., 25, 1058-1071
pyrimidine dimers (33). This indicates that the ratio of the two 5. Rahn.R.O. (1983) In Friedberg.E.C. and Hanawah,P.R. (eds.), DNA repair,
types of modification can be influenced by the ground-state a laboratory manual of research procedures, Marcel Dekker, New York,
molecules (e.g. undecomposed 1,2-dioxetane) that are present. Vol. 2, pp. 7 5 - 8 5 .
6. Kochevar.I.E., and Durm.D.A. 1990, In Morrison.H. (ed.), Bioorganic
As concluded from the lack of a D2O isotope effect (Table Photochemistry, John Wiley & Sons, New York, Vol. 1, pp- 273-316.
IV), singlet oxygen is not involved in the generation of FPG- 7. Lamola,A.A. and Yamane.Y. (1967) Proc. NatL Acad. Sd. U.S.A., 58,
sensitive base modifications by excited acetone. This is in marked 443-446.
contrast to the situation observed widi several photosensitizers 8. Ben-Ishai,R., Ben-Hur,E. and Homfeld.Y. (1968) Israel J. Chan., 6,
769-775.
which absorb visible light (12,38). Similarly, superoxide, H2O2 9. Rahn.R.O. (1979) Acta BioL MaL Germ., 38, 1225-1231.
and hydroxyl radicals are not involved, since superoxide 10 Mori.T., Matsunaga.T., Hirose.T. and Nikaido.O. (1988) Mutation Res.,
dismutase and catalase have no effect on the damage and since 194, 263-270.
 Nucleic Acids Research, 1993, Vol. 21, No. 4 869

11. Mennigmann,H.-D. and Wacker.A. (1970) Photochem. Photobiol., 11,

2911-2916.
12. Epe.B. (1991) Chenu-Biol Interactions, 80, 239-260.
13. Piette,J. (1991) J. Photochem. Photobiol B., 11, 241-260.
14. Zierenberg.B.E., Kramer.D.M., Geisert.M.G. and Kirste.R.G. (1971)
Photochem. Photobiol., 14, 515-520.
15. Charlier.M. and Helene,C (1972) Photochem. Photobiol., 15, 527-536.
16. Umlas.M.E., Franklin.W.A., Chan.G.L. and Haseltine.W.A. (1985)
Photochem. Photobiol., 42, 265-273.
17. Cadet J. and Berger.M. (1985) Photochem. Photobiol., 41, 113S.
18. Salditt,M., Braunstein.S.N., Camerini-Otero.R.D. and Franldin.R.M. (1972)
Virology, 48, 259-262.
19. Boiteux.S., O'Connor.T.R., Lederer.F., Gouyette.A. and LavaU. (1990)
J. BioL Otem. 265, 3916-3922.
20. Asahara,H., Wistort.P.M., BankJ.F., Bakerian,R.H. and Cunningham,R.P.
(1989) Biochemistry, 28, 4444-4449.
21. LevinJ.D., Johnson.A.W. and Danple.B. (1988) J. Biol. Chan. 263,
8066-8071.
22. Riazzudin.S. (1980) In Grossman.L. and Moldave.K. (eds.), Methods in
Enzymology, Academic Press, New York, Vol. 65, pp. 185-191.
23. MQller.E., Boiteux.S., Cunningham.R.P. and Epe.B., (1990) Nucleic Adds

Downloaded from http://nar.oxfordjournals.org/ at University of Auckland on July 13, 2015

Res., 18, 5969-5973.
24. Nieuwint,A.W.M., AubryJ.M., Arwert.F., Kortbeek,H., Herzberg.S. and
Joenje.H. (1985) Free Rod. Res. Comms., 1, 1-9.
25. Epe, B., Mutzel, P. and Adam, W. (1988) Chenu-Biol. Interactions, 67,
149-165.
26. Srerc.P.A., Kosicki.G.W. and Lumry.R. (1961) Biochim. Biophys. Ada,
50, 184-185.
27. Epe,B., HeglerJ. and Wild.D. (1989) Carcinogenesis 10, 2019-2024.
28. Wallace.S.S. (1988) Environ. Mol. Mutagen. 12, 431-477.
29. Povirk.L.F. and Steighner.R. (1989) Mutation Res. 214, 13-22.
30. Lindahl.T. (1990) Mutation Res. 238, 305-311.
31. Doetsch.P.W. and Cunningham.R.P. (1990) Mutation Res. 236, 173-201.
32. Di Mascio.P. and Sies,H. (1989) J. Amer. Chan. Soc., I l l , 2909-2914
33. Epe.B., MQller.E., Adam.W. and Saha-M6Uer,C.R. (1992) Chent-BioL
Interactions, in press.
34. Epe,B., Pflaum.M., Hanng,M., HeglerJ. andRiidiger.H. (1992) Toxicology
Lett., in press.
35. Hall.R.D. and Chignell.C.F. (1987) Photochem. Photobiol., 45, 459-464.
36. Merkd.P.B., Nilsson.R. and Keams.D.R. (1972) / Amer. Chan. Soc., 94,
1030-1031.
37. Rodgers.M.A.J. and Snowden.P.T. (1982) J. Amer. Chem. Soc., 104,
5541-5543.
38. Epe.B., Pflaum.M. and Boiteux.S. (1992) Mutation Res., in press.
39. SchneklerJ.E., Price.S., Maidt.M.L., GutteridgeJ.M.C. and Floyd.R.A.
(1990) Nucleic Acids Res., 18, 631 -635.
40. Floyd,R.A., West,M.S., SchnekJerJ.E.. WatsonJJ. and Maidt.M.L. (1990)
Free Rod. Biol. Med., 9, 76.
41. Devasagayam.T.P.A., Steenken.S., Obendorf.M.S.W., Schulz.W.A. and
Sies.H. (1991) Biochemistry, 30 (1991) 6283-6289.
42. Gallagher.P.E. and Duker.NJ. (1989) Photochem. Photobiol., 49, 599-605.
43. Boorstein.R.J., Hilbert.T.P., CadeU-, Cunningham.R.P. and Teebor.G.W.
(1989) Biochemistry, 28, 6164-6170.
44. Boorstein.R.J., Hilbert.T.P., Cunningham.R.P. and Teebor.G.W. (1990)
Biochemistry 29, 10455-10460.
45. Weiss.R.B. and Duker.N.J. (1987) Photochem. Photobiol., 45, 763-768.
46. Foote.C.S. (1968) Science, 162, 963-970.
47. Eisenstark.A. (1987) Environ. Molec. Mutagenesis, 10, 317-337.
48. TyrreU.R.M. and Keyse.S.M. (1990) J. Photochem PhotMoL B,
4,349-361.
49. Wood.M.L., Dizdaroglu.M. Gajewslri.E. and EssigmannJ.M. (1990)
Biochemistry, 29, 7024-7032.
50. Moriya, M., Ou, C , Bodepudi, V., Johnson, F., Takeshita, M. and
Grollman, A.P. (1991) Mutation Res., 254, 281-288.
51. KkinJ.C, Bleeker.M.J., Saris.C.P., Roelem.H.C.P.F., Brugghe.H.F., van
den Elst.H., van der Marel.G.A., van BoomJ.H., WestraJ.G., Kriek.E.
and Berns.A.J.M. (1992) Nucleic Acids Res., 20, 4437-4443.
52. Cheng.K.C, Cahill.D.S., Kasai.H., Nishimura.S., and Loeb.L.A. (1992)
J. Biol. Chem., 267, 166-172.
53. Shflxaani.S., Takesrrita,M. and Grollman^V.P. (1991) Nature, 349,431-434.
54. BrashJJ.E., RuddphJ.A., SimonJ.A., Lin^., McKerma,GJ., Baden.H.P.,
Halperin.A.J. and PonteiU (1991) Proc. NalL, Acad. Sd. USA, 88,
10124-10128.
55. Kelfkens.G., de Gruijl.F.R. and van Leun J.C. (1991) Carcinogenesis, 12,
1377-1382.