Intramolecular interactions in the triplet excited states of benzophenone-thymine dyads

Article abstract of DOI:10.1002/chem.200500345

Time-resolved and product studies on the synthesized dyads 1 and 2 have provided evidence that the benzophenone-to-thymine orientation strongly influences intramolecular photophysical and photochemical processes. The prevailing reaction mechanism has been established as a Paterno-Buechi cycloaddition to give oxetanes 3-6; however, the ability of benzophenone to achieve a formal hydrogen abstraction from the methyl group of thymidine has also been evidenced by the formation of photoproducts 7 and 8. These processes have been observed only in the case of the cisoid dyad 1. Adiabatic photochemical cycloreversion of the oxetane ring is achieved upon direct photolysis to give the starting dyad 1 in its excited triplet state. The photobiological implications of the above results are discussed with respect to benzophenone-photosensitized damage of thymidine.

Full text of DOI:10.1002/chem.200500345

FULL PAPER
DOI: 10.1002/chem.200500345
Intramolecular Interactions in the Triplet Excited States of
Benzophenone–Thymine Dyads
Noureddine Belmadoui,[a] Susana Encinas,[a] Maria J. Climent,[a] Salvador Gil,[b] and
Miguel A. Miranda*[a]
Abstract: Time-resolved and product
studies on the synthesized dyads 1 and
2 have provided evidence that the ben-
none to achieve a formal hydrogen ab-
straction from the methyl group of thy-
midine has also been evidenced by the
formation of photoproducts 7 and 8.
These processes have been observed
only in the case of the cisoid dyad 1.
Adiabatic photochemical cyclorever-
sion of the oxetane ring is achieved
upon direct photolysis to give the start-
ing dyad 1 in its excited triplet state.
The photobiological implications of the
above results are discussed with respect
zophenone-to-thymine
orientation
strongly influences intramolecular pho-
tophysical and photochemical process-
es. The prevailing reaction mechanism
has been established as a Paterno–
Büchi cycloaddition to give oxetanes
3–6; however, the ability of benzophe-
Keywords:
DNA
damage
·
oxetanes
·
photochemistry
·
photosensitization
spectroscopy
·
time-resolved
to
benzophenone-photosensitized
damage of thymidine.
Introduction
tions.[4–10] These adducts are considered to be formed from a
photochemical Paterno–Büchi type cycloaddition involving
the C5=C6 double bond of the 5’-pyrimidine and the C4 car-
bonyl group of the 3’-pyrimidine. The presumed oxetane is
thought to undergo fast ring-opening and isomerization to
the observed (6–4) photoproducts. In the last decade, it has
been discovered that a protein can effect the photoreversal
of (6–4) photoproducts.[11] Such a protein binds to the (6–4)
lesion in the dark and, upon absorption of a photon, repairs
the photoproduct to give the normal base forms.[12] Excita-
tion of the enzyme–substrate complex causes cycloreversion
by means of a photosensitized electron-transfer mechanism
analogous to that reported for the pyr<>pyr DNA photo-
lyase.[13,14] In this context, there are some examples in the lit-
erature on the synthesis of model systems designed to study
the mechanism of photosensitized reactions involving DNA
and proteins.[15]
At wavelengths longer than 290 nm, where DNA is not
absorbing, the bulk of the photobiological effects is mediat-
ed by photosensitizers.[16–18] For instance, a number of non-
steroidal anti-inflammatory drugs (NSAIDs) are known to
photosensitize DNA damage.[19–22] Hence, modified nucleo-
sides containing key substructures present in drugs and nu-
cleic acids can be relevant models to study the excited-state
interactions and the primary photophysical/photochemical
processes underlying drug-mediated photoreactions of DNA
and their photobiological consequences.[23] In particular,
they could help to gain some insight into the chemical
Exposure of living organisms to solar radiation may induce
lethal mutagenic and carcinogenic effects as a result of pho-
tochemical modifications of DNA. The cyclobutane pyrimi-
dine dimers pyr<>pyr are formed as the most abundant
mutagenic photoproducts.[1–3] Consequently, most research
on DNA photodamage, mutagenesis, and repair has focused
on the formation and repair of this particular type of
damage. More recent studies have identified the pyrimidine
(6–4) pyrimidone photoadducts, in which the 6-position of
one base is bonded to the 4-position of the adjacent one, as
very effective UV photoproducts that cause damaging muta-
[a] N. Belmadoui, Dr. S. Encinas, Dr. M. J. Climent,
Prof. Dr. M. A. Miranda
Instituto de Tecnología Química UPV-CSIC/Departamento de
Química
Universidad PolitØcnica de Valencia
Avda de los Naranjos s/n, 46022 Valencia (Spain)
Fax : (+34)963-877-809
E-mail: mmiranda@qim.upv.es
[b] Prof. Dr. S. Gil
Departamento de Química Orgµnica
Universidad de Valencia
Doctor Moliner 50, 46100 Burjassot, Valencia (Spain)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2006, 12, 553– 561
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
553

nature of the adducts formed in photosensitized DNA-drug
systems, which is poorly understood.
It will be shown that the predominating photoreaction
pathway in the case of dyad 1 is actually a Paterno–Büchi
cycloaddition to give a complex mixture of isomeric oxe-
tanes. The repair of this model lesion can be easily achieved
by direct photolysis that effects cycloreversion in a rare,
adiabatic process[46] leading to the ground-state thymine plus
the excited triplet state of the carbonyl moiety.
Among NSAIDs, ketoprofen (KP, 2-[3-benzoylphenyl]-
propionic acid) is a well-known benzophenone-derived pho-
tosensitizer. It produces both photoallergic and phototoxic
effects,[24–28] and its photosensitizing properties toward bio-
logical targets have been widely investigated.[29–33] Specifical-
ly, KP induces photooxidative DNA damage (including
strand breaks and base lesions) and photosensitizes the for-
mation of cyclobutane thymine dimers.[34–37] These photobio-
logical properties are attributable to the benzophenone
(BP) substructure, as the parent compound BP photosensi-
tizes similar DNA reactions.[29] Indeed, previous studies
have shown that there is a strong intermolecular interaction
between the excited BP chromophore and thymine deriva-
tives,[38–41] including the free thymidine nucleoside[42] and a
thymidine 5’-monophosphate.[43] Fast quenching of triplet
BP by thymine derivatives has been observed and attributed
to the occurrence of energy or electron transfer; however,
recent evidence has been obtained which suggests that this
quenching is mainly associated with a Paterno–Büchi cyclo-
addition leading to the formation of oxetanes.[44]
Results and Discussion
Photochemical reactivity ofdyads 1 and 2 : Steady-state irra-
diation of dyad 1 in acetonitrile, through Pyrex, led to the
formation of several photoproducts; in contrast, irradiation
of 2 gave only polymerization. The photoproducts were sep-
arated by semipreparative HPLC, and their structures were
unambiguously determined by a complete assignment of the
1H and 13C NMR signals by the use of a combination of H,H
(COSY) and H,C correlations (HSQC and HMBC). The
main photoproducts were found to be oxetanes 3–6 (com-
bined yield 52%). Minor amounts of cyclic compounds 7
and 8 (14%) were also obtained (Scheme 1).
To gain a better understanding of this type of processes,
two thymidine-derived nucleosides (compounds 1 and 2)
have been prepared in the present work by covalently at-
taching a benzophenone derivative, namely (S)-ketoprofen,
to positions 5’ or 3’ of the sugar.
The regiochemistry of 3 and 4 became evident from the
13C chemical shifts of the bridgehead carbon atoms of the
oxetane ring (dCH
= 61.0/60.8 ppm and dC = 77.4/
77.9 ppm); they are essentially coincident with those report-
ed for related compounds.[41] In contrast, the regioisomeric
oxetanes 5 and 6 displayed characteristic 13C signals with
These dyads were designed as models to study the interac-
tion between excited drugs and nucleic acids. Furthermore,
they present a different spatial arrangement (transoid or
cisoid) of the BP chromophore relative to the thymine
(Thy) unit that could allow an investigation of the influence
of the benzophenone-to-thymine orientation on the photo-
physical and photochemical properties. Analogous model
systems based on S-KP have been used for the study of
drug-photosensitized lipid peroxidation,[45] although in this
case, S-KP was linked to a cyclohexa-1,4-diene moiety as a
source of the doubly allylic hydrogens present in polyunsa-
turated fatty acids.
Scheme 1. Photoproducts resulting upon irradiation of 1 in acetonitrile.
554
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 553– 561

Benzophenone–Thymine Dyads
FULL PAPER
Table 1. Selected proton NOE data for compound 3 in CDCl3.
Irradiation of:
NOE observed[%]:
ArH
H1’
H6
H5’
H5’
H4’
CH-CH3
H2’’
H2’
CH-CH3
0.61
0
0
0
0
0.52
0
2.54
0.12
0
ArH (d = 8.03ppm)
H1’ (d = 5.92 ppm)
H6 (d = 5.36 ppm)
H5’ (d = 4.70 ppm)
H5’ (d = 4.46 ppm)
H3’ (d = 4.40 ppm)
H4’ (d = 4.10 ppm)
CH-CH3 (d = 3.96 ppm)
H2’’ (d = 2.20 ppm)
H2’ (d = 1.80 ppm)
–
0
6.82
0.39
0.54
–
0
2.09
0
2.71
–
0.39
0
–
1.16
0
0
0.41
1.57
0
2.85
1.28
0
6.67
0.73
–
0.47
0
0
0
0
0
0.57
0
0.18
–
13.16
1.62
2.89
0
0
0
–
0
0
0
0
66.13
–
–
5.30
0
0
0
0
0
0
1.39
0
0
0
0
0
0
–
0
0
0.47
0.33
1.47
0
0
0
0
0
1.14
–
0
1.41
0.34
2.15
3.66
0
–
0
0
0
3.64
0
0
0
1.20
0
0
–
0
1.29
0
0
0
0.97
13.60
0.27
0
ꢁ
C CH3 (d = 1.77 ppm)
0
0
0
–
CH-CH3 (d = 1.61 ppm)
0
0
Table 2. Selected proton NOE data for compound 4 in CDCl3.
Irradiation of:
ꢁ
NOE observed[%]:
ArH
H1’
H6
H5’
H3’
H4’
CH-CH3
H2’’ + H2’
C CH3
CH-CH3
ArH (d = 8.08 ppm)
H1’ (d = 6.40 ppm)
H5’ (d = 4.85 ppm)
H6 (d = 4.74 ppm)
H5’ (d = 4.45 ppm)
H3’ (d = 4.38 ppm)
H4’ (d = 3.97 ppm)
CH-CH3 (d = 3.87 ppm)
H2’’ + H2’ (d = 2.20 ppm)
–
0
0
3.09
0
1.20
0
1.20
0
0
0.52
0
–
0
0
0
0
2.58
0
3.80
0
3.60
0
–
0
0
–
–
0.56
0
7.90
0.33
–
–
1.96
0
0.38
1.69
1.85
0
1.530
0
–
–
–
0
0.85
0.29
2.02
0.530
0
0
0
0
1.20
0.50
0
0
0
0
–
0
0.90
1.41
0
20.40
0
3.90
0
0
0
0
0
0
0
0
–
–
0
0
3.44
1.91
0.67
0
–
1.00
–
0
0
0
0.30
–
0
0.54
1.62
–
–
–
0
5.99
0.5
2.86
2.30
0
2.00
ꢁ
C CH3 (d = 1.75 ppm)
CH-CH3 (d = 1.58 ppm)
0
1.30
0
Table 3. Selected proton NOE data for compound 5 in CDCl3.
cycle. As a general rule, sizable
NOE effects were observed be-
tween protons at a distances of
less than 3–4 Š in the molecu-
lar models obtained upon
MOPAC optimization of the ge-
ometries (see the structures in
the Supporting Information,
S19–S22). Overall, these data
support the structural assign-
ment of the isolated oxetanes
as shown in Scheme 1.
Irradiation of:
NOE observed[%]:
H1’
H6
H5’
H4’
ArH (d = 8.07 ppm)
H1’ (d = 6.10 ppm)
H6 (d = 5.70 ppm)
H5’ (d = 4.68 ppm)
H3’ (d = 4.66 ppm)
H5’ (d = 4.33 ppm)
H4’ (d = 4.02 ppm)
CH-CH3 (d = 3.93 ppm)
H2’’ + H2’ (d = 2.53ppm)
CH-CH3 (d = 1.60 ppm)
0
–
0.09
0
0
0
0.57
0
2.98
0
5.10
0.56
–
0
1.28
0
2.57
0
0.80
–
0
0.14
0
3.90
0
1.75
–
0
1.38
0
–
3.42
2.06
0
1.59
0.34
0
0
0
0.62
0
0.80
ꢁ
C CH3 (d = 1.07 ppm)
0
0
The 1H NMR spectra of the
minor cyclic photoproducts 7
chemical shifts dCH = 79.8/81.9 ppm and dC = 57.8 ppm.[41]
This assignment is also supported by the known a- and b-ef-
fects of the oxygen atom on the 13C chemical shifts; detailed
estimations based on these effects are given in the Support-
ing Information S10–S15.
The stereochemistry of 3–6 was basically assigned by
means of NOE experiments (Tables 1–4). This assignment
was fully confirmed for oxetane 4 by X-ray crystallography.
For instance, upon irradiation of the proton of the oxetane
(H6) remarkable NOE enhancements for H1’ were only ob-
served for 3 and 5. Moreover, a clear NOE interaction was
found in the four compounds between H6 and the ortho-
proton of the benzene ring that is integrated into the macro-
and 8 showed the disappearance of the thymine methyl
group of 1 and the presence of two new diastereotopic
methylene protons between d = 3.3 and 3.6 ppm. In the
13C NMR spectra, the most salient feature was the lack of
the thymine methyl group and the ketone carbonyl, together
with the appearance of new signals corresponding to the
methylene carbon at dꢀ41 ppm and the quaternary carbon
at dꢀ79 ppm.
Again, the stereochemistry was assigned by means of
NOE experiments (Tables 5 and 6). The main interaction
for discrimination purposes was found between the internal
methylene proton of 7 (d = 3.34 ppm) and the ortho-pro-
tons of the phenyl (C6H5) group at d = 7.57 ppm. The rela-
Chem. Eur. J. 2006, 12, 553– 561
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
555

M. A. Miranda et al.
Table 4. Selected proton NOE data for compound 6 in CDCl3.
Irradiation of:
ꢁ
NOE observed[%]:
ArH
H1’
H6
H3’
H2’
CH-CH3
C CH3
ArH (d = 8.57 ppm)
H1’ (d = 6.41 ppm)
H6 (d = 5.28 ppm)
H5’ (d = 4.55 ppm)
H4’ (d = 4.27 ppm)
H3’ (d = 4.20 ppm)
CH-CH3 (d = 4.10 ppm)
H5’ (d = 4.05 ppm)
H2’’ (d = 2.48 ppm)
H2’ (d = 1.72 ppm)
CH-CH3 (d = 1.61 ppm)
–
0
2.27
0
0
0
1.03
0
0
0
–
2.79
0
0
0
0
0
0
0
0
0
0.13–
0
0
0
0
0
1.68
0
0.18
0.01
1.830.97
0
1.23
0
0
3.07
0
0
0.47
0.84
0
2.22
–
–
0
0
0
1.21
0
0
0
0
2.04
0
0
7.57
–
0
0
0
0
0
1.28
0
0
0
–
0
0
0
0
0
0
0
0
0
–
0
0
0
ꢁ
C CH3 (d = 1.17 ppm)
0.50
Table 5. Selected proton NOE data for compound 7 in CD3OD.
been reported for thymine de-
rivatives.[38,41,44,47] The factors
governing regioselectivity in ox-
etane formation are not clear;
however, they are thought to in-
clude the stability of 1,4-biradi-
cal intermediates, steric effects,
and hydrogen-bonding interac-
tions. In the case of thymidine,
the N,O-acetal regioisomers
(analogous to 5 and 6) are the
major products by far. The en-
Irradiation of:
NOE observed[%]:
H1’
H6
C(H)H
C(H)H
CH-CH3
H2’’
H2’
ArH (d = 7.57 ppm, d)
(d = 7.16 ppm, s)
0
0
0
–
0
0
2.32
0
0
1.69
–
0
0
2.75
0
3.58
0
0
0
0
1.45
2.21
1.26
0
6.32
–
0
6.25
0
0
0
0
0
0
0
0
0
4.09
0
0
1.01
5.32
0
0
0
H6 (d = 6.54 ppm)
H1’ (d = 6.10 ppm)
C(H)H (d = 3.58 ppm)
C(H)H (d = 3.34 ppm)
H2’’ (d = 1.76 ppm)
H2’ (d = 0.77 ppm)
–
20.36
0
0
0
–
0
0
20.79
–
19.38
hanced formation of oxetanes with the O atom of the car-
bonyl linked to the C5 thymine position (as in 3 and 4) start-
ing from dyad 1 must be attributed to their lower ring
strain. Here, the rigidity of the resulting macrocycle seems
to play a key role.
On the other hand, the minor photoproducts 7 and 8 arise
as a result of hydrogen abstraction by the benzophenone
carbonyl from the methyl group of the thymine base. The
resulting biradical intermediate would collapse upon intra-
Table 6. Selected proton NOE data for compound 8 in CD3OD.
Irradiation of:
NOE observed[%]:
H1’
H6
H2’’
H2’
ArH (d = 7.30 ppm, s)
H6 (d = 6.67 ppm)
H1’ (d = 6.14 ppm)
H5’ (d = 4.31 ppm)
H2’’ (d = 2.12 ppm)
H2’ (d = 1.73ppm)
0
0
–
0
2.98
0
1.52
–
0
1.15
0
3.03
0
0
4.72
0
–
0
4.27
0
0
16.24
–
19.53
ꢁ
molecular coupling with C C bond formation. Hydrogen ab-
straction by benzophenone is well known.[48] Moreover,
some products of the reported benzophenone-mediated pho-
tosensitization of thymidine can be rationalized by a formal
hydrogen abstraction from the C5 methyl group [Eq. (1)].[42]
tive spatial arrangement of these protons, which explains the
observed NOE effect, is shown in the Supporting Informa-
tion (S23, top view). Unfortunately, the corresponding ex-
periment could not be performed with the other stereoisom-
er (8) because its two methylene protons (internal and exter-
3BP þ dThd ! BPHC þ dThdðꢁHÞ
C
ð1Þ
1
nal) do not appear as separate signals in the H NMR spec-
To sum up, the mechanisms taking place in the photoreac-
tions of dyad 1 are shown in Scheme 2.
trum. An interesting feature observed for compound 7 was
the upfield shift of its H2’ proton, which appeared at d =
0.77 ppm. In the minimized models (see Supporting Infor-
mation, S23(bottom view) and S24) it becomes clear that
this proton falls into the shielding region of the disubstituted
aromatic ring only in the case of 7. Accordingly, a clear
NOE interaction was observed between H2’ and the aromat-
ic singlet at d = 7.2–7.3ppm for compound 7, but not for its
stereoisomer 8.
Oxetanes 3–6 are clearly formed by means of an intramo-
lecular Paterno–Büchi cycloaddition between the carbonyl
group of the benzophenone and the double bond of the thy-
mine base. Intermolecular cycloadditions of this type have
Photophysics ofdyads 1 and 2 : Transient absorbance data
were recorded for the two dyads and compared with those
obtained for S-KP as a reference compound. All experi-
ments were performed in pure acetonitrile under an anaero-
bic atmosphere and with an excitation wavelength of
355 nm. In all cases, the typical triplet–triplet absorption
spectrum of benzophenone was obtained,[49] with a maxi-
mum at lꢀ530 nm (Figure 1a). However, significant differ-
ences were found in the lifetimes. Thus, decay of the triplet
state was markedly slower for S-KP (tKP = 1.3 ms) than for
556
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 553– 561

Benzophenone–Thymine Dyads
FULL PAPER
role in the intramolecular deac-
tivation of BP triplets by Thy in
the dyads. Such an energy
transfer should be evidenced
through the known triplet–trip-
let absorption of Thy at 370 nm,
which was not detected in our
experiments. Dimerization was
not observed, probably owing
to the inefficient intermolecular
reaction
between
small
amounts of triplet Thy and Thy
units of other nucleoside in the
ground state. In this context,
the use of a BP-functionalized
oligonucleotide with two (or
more) adjacent Thy bases could
increase the efficiency of this
Scheme 2. Simplified mechanisms of the intramolecular photoreaction between the two active moieties of
dyad 1.
process by an enhanced intramolecular quenching process.
Thus, it cannot be completely ruled out that a small frac-
tion of thermally activated BP triplets could achieve energy
transfer to Thy in very low quantum yields (below the detec-
tion limits of the laser flash photolysis technique). This reac-
tion pathway, although minor, could be of biological signifi-
cance.[42] Besides, the triplet energy of the Thy base in DNA
must be lower than that in the dyad; this would enhance the
prospects of triplet–triplet energy transfer (and hence cyclo-
butane dimer formation) in the biomacromolecule.
It has been reported that BP and KP photosensitize oxi-
dative damage to DNA. Although the guanine bases appear
to be the preferred sites for this damage, the free dThd nu-
cleoside has also been found to undergo BP-photosensitized
oxidation. Electron transfer from dThd to triplet BP, fol-
lowed by proton transfer, has been proposed as the operat-
ing mechanism. This would be equivalent to formal hydro-
gen abstraction for the thymine methyl group, with forma-
tion of an allylic radical as the key intermediate. Subsequent
trapping by oxygen accounts for some of the dThd oxidation
products. Although electron transfer from Thy to BP would
be thermodynamically feasible (DG = ꢁ30 kJmolꢁ1), for-
mation of the resulting radical ions was not observed in the
laser flash photolysis experiments with the dyads. However,
a significant amount of photoproducts arising from a formal
intramolecular hydrogen abstraction (such as 7 and 8) was
obtained; this strongly supports the mechanism proposed in
the literature[42] for the generation of oxidative dThd lesions.
Nonetheless, oxidation of the Thy sites in DNA by this
mechanism would be unlikely because the purine bases
(such as guanine) have much lower oxidation potentials and
would be much better electron donors toward the excited
BP chromophore.[37]
Figure 1. a) Transient spectrum obtained for dyad 1 in acetonitrile 35 ns
after the laser pulse at 355 nm. Similar spectra were obtained for dyad 2
and S-KP. b) Decays of the triplet signal at 530 nm for S-KP and dyads 1
and 2.
any of the dyads; this effect was specially noteworthy in the
case of 1 (t = 20 ns), as shown in Figure 1b.
The dramatically shorter lifetime of the excited triplet
state in the cisoid dyad 1 is in excellent agreement with its
higher reactivity, leading to the formation of oxetanes 3–6
and, to a lesser extent, to macrocycles 7 and 8. None of the
intermediate biradical species appearing in Scheme 2 were
observed (presumably too short-lived).
Mechanism ofbenzophenone-photosensitized DNA damage
at thymine sites: Formation of cyclobutane thymine dimers,
which is one of the major DNA lesions, has been reported
to occur upon BP and KP photosensitization.[36,37] The re-
quired triplet–triplet energy transfer, which would be ther-
modynamically disfavored, does not seem to play a major
Overall, the results obtained by means of time-resolved
and product studies on dyad 1 provide evidence that there is
indeed a strong intramolecular interaction between the BP
and Thy moieties in the triplet excited state. This is essen-
tially attributable to a Paterno–Büchi photoreaction, in
which the initial step is the formation of a new bond be-
Chem. Eur. J. 2006, 12, 553– 561
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
557

M. A. Miranda et al.
tween the excited carbonyl oxygen and one of the Thy ole-
finic carbons to give oxetanes 3–6 as final products. Such
products have been actually isolated and identified in the
preparative irradiations.
The question remains as to whether oxetane formation
upon photolysis of KP in the presence of DNA might still
be a possible reaction pathway. This possibility and its bio-
logical significance still need to be checked. In this context,
it is interesting to mention that oxetanes are believed to be
involved in the formation and photoenzymatic repair of the
DNA damage associated with pyrimidine (6–4) pyrimidone
photoproducts.[40] Here, photolyases play a key role by
acting as electron donors to achieve the reductive oxetane
cycloreversion to afford two repaired pyrimidine units.[40,50]
Adiabatic photochemical cycloreversion ofoxetanes : In a
preliminary communication, it was reported that the excited
triplet state of the carbonyl compound, together with the
ground state of the pyrimidine base, are generated upon
laser irradiation (l = 266 nm) of oxetanes derived from 1,3-
dimethylthymine.[46] This is a rare and interesting case of
adiabatic photochemical reaction whose mechanism needs
to be clarified in more detail.
Figure 2. Plot of the absorbance of the carbonyl band at 254 nm against
~
&
*
)
the irradiation time for a) oxetanes 3 ( ), BP-dThd ( ) and BP-DMT (
&
*
and, b) dyad 1 ( ) and BP in 2-propanol ( ).
To gain further understanding of this type of oxetane cy-
cloreversion, the direct photolysis of 3 (the major product
obtained upon irradiation of dyad 1) has been examined in
the present work. Here, the strain release associated with
cleavage of the macrocycle was expected to have a marked
influence on the process. Hence, the oxetanes obtained from
the intermolecular photochemical reaction of benzophenone
with thymidine (BP-dThd)[44] and 1,3-dimethylthymine (BP-
DMT)[46] were used for a comparison. All the experiments
were performed in acetonitrile/water (4:1 v/v).
an analogous experimental procedure was followed. In this
case, the reference was a solution of BP in 2-propanol, with
a limiting photoreduction quantum yield of 1.96. Both solu-
tions were photolyzed under an anaerobic atmosphere, and
the decrease of the carbonyl absorption at l = 254 nm was
followed. A photolysis quantum yield of 0.4 for dyad 1 was
obtained as described above (Figure 2b), also at low conver-
sions (<10%). Since 83% of all the photoproducts of 1 are
oxetanes (compounds 3–6, Scheme 1), the Paterno–Büchi
quantum yield would be ꢀ0.3.
As expected for an adiabatic cycloreversion, the transient
spectrum (Figure 3) resulting from laser flash photolysis (l
= 266 nm, 10 ns, 2–5 mJ per pulse) of oxetane 3 was identi-
cal to that obtained from 1 (Figure 1a), reliably assigned to
the well-characterized triplet–triplet absorption of benzo-
phenone (lmax = 530 nm). This process was nearly quantita-
tive, as indicated by the very high quantum yield (close to
unity), measured from the transient absorbance at l =
530 nm immediately after the laser pulse compared with
that of benzophenone.
In the case of BP-DMT, a lower efficiency for this adia-
batic process was previously reported (quantum yield
ꢀ0.4).[46] This is qualitatively consistent with the results
from preparative experiments (Figure 2a), in which 3 also
clearly reacted faster than BP-DMT. Thus, the strain release
associated with cleavage of the macrocyclic ring in 3 seems
to play an important role in enhancing the reaction.
Prior to irradiation, the UV/Vis absorption spectra of the
oxetanes showed a band in the l = 220–230 nm region. The
course of the photolysis was monitored by following the in-
crease of the UV absorption at l = 254 nm, which is typical
of the carbonyl compound. The cycloreversion quantum
yields were calculated from the slopes of the straight lines
obtained when the absorbance at l = 254 nm was plotted
against the irradiation time (Figure 2a); the values were
found to be 0.7 and 0.3for 3 and BP-dThd, respectively, ref-
erenced to BP-DMT (photolysis quantum yield = 0.5).[46]
Conversions were kept below 10% throughout the measure-
ments.
A simplified energy diagram which indicates that the
adiabatic photochemical cycloreversion of oxetane 3 is ener-
getically feasible is shown in Figure 4. The key intermediates
are supposed to be the 1,4 biradicals (singlet and triplet) de-
ꢁ
In a similar way, in order to determine the quantum yield
for the opposite process, which is the photolysis of dyad 1,
rived from C C bond scission. Cleavage of the triplet biradi-
cal to afford the ground state carbonyl chromophore ap-
558
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 553– 561

Benzophenone–Thymine Dyads
FULL PAPER
tion pathways for the ketoprofen triplet excited state only in
the case of the cisoid dyad 1.
In addition, unambiguous experimental evidence has been
obtained which supports the fact that the obtained macrocy-
clic oxetanes undergo an adiabatic photochemical cyclore-
version under direct photolysis to yield the starting dyad (1)
in its excited triplet state. This rare photochemical process is
more efficient than the retro-Paterno–Büchi reaction of the
oxetanes obtained from intermolecular photocycloaddition
between thymine derivatives and benzophenone.
Experimental Section
Materials: (S)-Ketoprofen [(S)-2-(3-benzoylphenyl)propionic acid, (S)-
KP], thymidine (dThd), and 1-[3-(dimethylamino)propyl]-3-ethylcarbo-
diimide hydrochloride (EDAC) were obtained from commercial sources.
Acetonitrile and methanol (HPLC grade) were used without further puri-
fication.
Analytical instrumentation: UV spectra were recorded on a UV/Vis scan-
ning spectrophotometer with a slit width of 5 nm. NMR spectra were re-
corded with a 300 MHz instrument. Bruker Avance400/500 spectrometers
were used for NOE and two-dimen-
Figure 3. a) Transient spectrum obtained 100 ns after laser irradiation of
oxetane 3 at 266 nm. b) Decay of the triplet signal at 530 nm.
sional NMR experiments; transient
NOE effects were recorded with
a
mixing time of 500 ms. For HPLC/MS
analyses, the API-ES positive ioniza-
tion was used for mass detection and
MeOH/H2O (50:50) as the eluent for
separation. The identity of the com-
pounds was confirmed by fast-atom
bombardment (FAB) and electronic
impact (EI) recorded in a high-resolu-
tion mass spectrometer (HRMS).
Nanosecond laser-flash photolysis: A
pulsed Nd:YAG laser was used for ex-
citation at 355 or 266 nm. The single
pulses had a duration of about 10 ns
and the energy was about 10 and 4 mJ
per pulse, respectively.
A
pulsed
xenon lamp was employed as the de-
tecting light source. The laser flash-
photolysis apparatus consisted of
pulsed laser, a Xe lamp, a monochro-
mator, and photomultiplier tube
a
Figure 4. Energy diagram for the adiabatic photochemical processes as cycloreversion of thymine oxetane
adduct 3 and cycloaddition of dyad 1.
a
(PMT) system. The output signal from
the oscilloscope was fed to a personal
computer. The substrate concentration
was about 1.5 mm and ꢀ0.1 mm for ir-
pears to be roughly twice as fast as its ring closure to the ox-
etane adducts.
radiation at 355 nm and 266 nm, respectively. All solutions were deaerat-
ed by bubbling nitrogen. These values ensured an absorbance of ꢀ0.2 in
the laser cell at the excitation wavelength.
Synthesis and characterization ofthe dyads : Compounds 1 and 2 were
obtained by condensation of the photosensitizing benzophenone-contain-
ing drug (S)-KP with dThd and a carbodiimide (EDAC) as the activating
agent. Both dyads were fully characterized by 1H NMR, 13C NMR, and
mass spectrometry (see the spectra in the Supporting Information, S2
and S3).
Conclusion
The benzophenone-to-thymine orientation has a strong in-
fluence on the intramolecular photosensitization of thymi-
dine by ketoprofen. The prevailing mechanism for such a
process is a Paterno–Büchi cycloaddition to afford oxetanes
3–6; however, the ability of benzophenone to achieve a
formal hydrogen abstraction from the methyl group of thy-
midine has also been evidenced through formation of photo-
products 7 and 8. These processes provide efficient deactiva-
Synthesis ofcompounds 1 and 2 : Thymidine (dThd) (0.30 g, 1.23 mmol)
and (S)-ketoprofen [(S)-KP] (0.27 g, 1.19 mmol) were dissolved in anhy-
drous pyridine (2 mL). EDAC (0.23g, 1.19 mmol) was slowly added at
08C under constant stirring. When the addition was complete, the reac-
tion mixture was kept at 08C for 2 h. After this time, the reaction mix-
ture was concentrated under reduced pressure to remove the pyridine.
The resulting viscous material was resuspended in CH2Cl2 (25 mL) and
Chem. Eur. J. 2006, 12, 553– 561
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
559

M. A. Miranda et al.
washed with an aqueous solution of NaHCO3 (1m) to remove the un-
reacted dThd and (S)-KP. The organic phase was dried with anhydrous
Na2SO4, and the solvent was evaporated to dryness under reduced pres-
sure. The residue was submitted to column chromatography (silica gel,
CHCl3/CH3OH 10:1, v/v) to afford the corresponding 5’-ester 1 (55%)
and 3’-ester 2 (16%).
S-Ketoprofen, 5’-ester with thymidine (1): Yield 55%; 1H NMR
(300 MHz, CDCl3): d = 8.57 (brs, 1H, NH), 7.85–7.10 (m, 10H, Ar-H +
HC=), 6.20 (t, J = 6.9 Hz, 1H, H1’), 4.55 (dd, J = 12.1, 4.0 Hz, 1H,
H5’), 4.22 (dd, J = 12.1, 3.3 Hz, 1H, H5’), 4.20–4.10 (m, 2H, H3’ + H4’),
3.85 (q, J = 7.1 Hz, 1H, CH-CH3), 3.02 (brs, 1H, OH), 2.25 (m, 1H,
60.8 (CH), 45.2 (CH), 37.1 (CH2), 22.3(CH 3), 17.8 ppm (CH3); HRMS
(FAB) calcd. for C26H26O7N2: 478.1740; found 478.1751.
Oxetane 5: Yield: 10%; 1H NMR (300 MHz, CDCl3): d = 8.07 (s, 1H,
Ar-H), 7.65–7.15 (m, 9H, Ar-H + NH), 6.10 (dd, J = 6.6, 4.2 Hz, 1H,
H1’), 5.70 (s, 1H, H6), 4.68 (dd, J = 12.3, 2.4 Hz, 1H, H5’), 4.66 (m, 1H,
H3’), 4.33 (dd, J = 12.3, 1.2 Hz, 1H, H5’), 4.02 (m, 1H, H4’), 3.93 (q, J
= 7.2 Hz, 1H, CH-CH3), 2.72 (brs, 1H, OH), 2.51 (m, 2H, H2’ + H2’’),
1.60 (d, J = 7.2 Hz, 3H, CH-CH3), 1.07 ppm (s, 3H, C CH3); 13C NMR
ꢁ
(75 MHz, CDCl3): d = 175.6 (C), 170.6 (C), 150.5 (C), 140.7 (C), 139.4
(C), 138.9 (C), 129.2 (CH), 128.5 (CH), 128.3 (CH), 127.8 (CH), 127.6
(CH), 126.7 (CH), 125.0 (CH), 87.7 (C), 83.8 (CH), 83.1 (CH), 79.8
(CH), 68.9 (CH), 63.7 (CH2), 57.8 (C), 47.5 (CH), 41.0 (CH2), 20.1 (CH3),
17.6 ppm (CH3); HRMS (EI) calcd. for C26H26O7N2: 478.1740; found
478.1708.
ꢁ
H2’’), 1.90 (s, 3H, =C CH3), 1.74 (m, 1H, H2’), 1.58 ppm (d, J = 7.1 Hz,
3H, CH-CH3); 13C NMR (75 MHz, CDCl3): d = 196.4 (C), 174.0 (C),
164.0 (C), 150.4 (C), 140.6 (C), 138.0 (C), 137.0 (C), 135.2 (CH), 132.8
(CH), 131.3 (CH), 129.9 (CH), 129.3 (CH), 128.6 (CH), 128.5 (CH),
128.3(CH), 110.9 (C), 85.1 (CH), 84.3(CH), 71.2 (CH), 64.3(CH 2), 45.4
(CH), 39.9 (CH2), 18.4 (CH3), 12.6 ppm (CH3); HRMS (EI) calcd for:
C26H26O7N2 478.1740; found 478.1750.
1
Oxetane 6: Yield: 9%; H NMR (300 MHz, CDCl3): d = 8.57 (s, 1H, Ar-
H), 7.65–7.20 (m, 9H, Ar-H + NH), 6.40 (t, J = 6.7 Hz, 1H, H1’), 5.28
(s, 1H, H6), 4.50 (dd, J = 12.8, 2.7 Hz, 1H, H5’), 4.29 (m, 1H, H4’), 4.20
(m, 1H, H3’), 4.10 (q, J = 6.9 Hz, 1H, CH-CH3), 4.05 (dd, J = 12.8,
7.5 Hz, 1H, H5’), 2.48 (m, 2H, H2’’ + OH), 1.72 (m, 1H, H2’), 1.61 (d, J
S-Ketoprofen, 3’-ester with thymidine (2): Yield 16%; 1H NMR
(300 MHz, CDCl3): d = 9.85 (brs, 1H, NH), 7.80–7.40 (m, 10H, Ar-H +
HC=), 6.20 (dd, J = 8.2, 6.1 Hz, 1H, H1’), 5.40 (m, 1H, H3’), 3.95–3.75
(m, 4H, H4’ + 2”H5’ + CH-CH3), 3.33 (brs, 1H, OH), 2.55–2.30 (m,
= 6.9 Hz, 3H, CH-CH3), 1.17 ppm (s, 3H, C CH3); 13C NMR (75 MHz,
ꢁ
CDCl3): d = 175.1 (C), 170.6 (C), 150.7 (C), 139.5 (C), 139.0 (C), 138.7
(C), 130.2 (CH), 129.3 (CH), 129.0 (CH), 128.5 (CH), 127.9 (CH), 126.3
(CH), 125.0 (CH), 86.8 (C), 84.8 (CH), 83.5 (CH), 81.9 (CH), 71.1 (CH),
ꢁ
2H, H2’ + H2’’), 1.85 (s, 3H, =C CH3), 1.56 ppm (d, J = 7.1 Hz, 3H,
CH-CH3); 13C NMR (75 MHz, CDCl3): d = 196.0 (C), 173.7 (C), 163.8
(C), 150.5 (C), 140.2 (C), 138.0 (C), 137.2 (C), 136.6 (CH), 131.3 (CH),
130.2 (CH), 129.9 (CH), 129.3 (CH), 129.0 (CH), 128.7 (CH), 128.4
64.3(CH ), 57.8 (C), 43.6 (CH), 40.6 (CH2), 16.7 (CH3), 14.0 ppm (CH3);
2
HRMS (EI) calcd for C26H26O7N2 478.1740; found 478.1753.
Compound 7: Yield: 6%; 1H NMR (300 MHz, CD3OD): d = 7.64–7.10
(m, 10H, Ar-H + NH), 6.54 (s, 1H, H6), 6.10 (dd, J = 9.3, 5.7 Hz, 1H,
H1’), 4.80 (dd, J = 12.9, 1.8 Hz, 1H, H5’), 3.95–3.90 (m, 3H, H3’ + H4’
+ H5’), 3.73 (q, J = 7.2 Hz, 1H, CH-CH3), 3.58 (d, J = 14.1 Hz, 1H,
CH2), 3.34 (d, J = 14.1 Hz, 1H, CH2), 1.76 (ddd, J = 13.5, 5.7, 1.5 Hz,
1H, H2’’), 1.47 (d, J = 7.2 Hz, 3H, CH-CH3), 0.77 ppm (ddd, J = 13.5,
9.3, 6.3 Hz, 1H, H2’); 13C NMR (75 MHz, CD3OD): d = 175.7 (C), 167.2
(C), 151.6 (C), 149.1 (C), 147.8 (C), 141.6 (C), 138.9 (CH), 130.3 (CH),
130.0 (CH), 129.2 (CH), 128.1 (CH), 127.3 (CH), 125.7 (CH), 124.9
(CH), 111.7 (C), 86.7 (CH), 84.7 (CH), 79.7 (C), 71.5 (CH), 64.6 (CH2),
46.7 (CH), 41.8 (CH2), 37.6 (CH2), 17.5 ppm (CH3); HRMS (EI) calcd
for C26H26O7N2 478.1740; found 478.1753; HRMS (FAB) calcd. for
C26H26O7N2 478.1740; found 478.1760.
Compound 8: Yield: 8%; 1H NMR (300 MHz, CD3OD): d = 7.60–7.10
(m, 10H, Ar-H + NH), 6.67 (s, 1H, H6), 6.14 (dd, J = 8.2, 6.0 Hz, 1H,
H1’), 4.31 (dd, J = 12.3, 7.5 Hz, 1H, H5’), 4.10–3.90 (m, 3H, H3’ + H4’
+ H5’), 3.80 (q, J = 6.9 Hz, 1H, CH-CH3), 3.40 (m, 2H, CH2), 2.12
(ddd, J = 13.8, 6.0, 2.7 Hz, 1H, H2’’), 1.73(ddd, J = 13.8, 8.2, 6.0 Hz,
1H, H2’) 1.39 ppm (d, J = 6.9 Hz, 3H, CH-CH3); 13C NMR (75 MHz,
CD3OD): d = 176.1 (C), 167.7 (C), 151.6 (C), 149.1 (C), 147.3(C), 142.0
(C), 140.0 (CH), 129.7 (CH), 129.0 (CH), 128.0 (CH), 127.4 (CH), 127.2
(CH), 126.2 (CH), 126.1 (CH), 111.7 (C), 85.6 (CH), 84.9 (CH), 79.6 (C),
71.5 (CH), 64.6 (CH2), 46.6 (CH), 39.9 (CH2), 39.6 (CH2), 18.3ppm
(CH3); HRMS (EI) calcd for C26H26O7N2 478.1740; found 478.1739.
(CH), 111.3(C), 86.6 (CH), 84.8 (CH), 75.2 (CH), 62.0 (CH ), 45.0 (CH),
2
36.0 (CH2), 18.3(CH 3), 12.1 ppm (CH3); HRMS (EI) calcd. for
C26H26O7N2: 478.1740; found 478.1721.
Steady-state photolysis of1 and 2 and characterization ofthe photoprod-
ucts
Steady-state photolysis of 1 and 2: A 2.63”10ꢁ3 m solution of the dyad (1
or 2) in acetonitrile was placed into a Pyrex tube surrounding a centrally
positioned quartz cooling jacket containing a 125 W medium-pressure Hg
lamp. The solution was degassed for 30 min with a stream of argon and
then irradiated for 5.5 h. After evaporation of the solvent, the residue
was submitted to reverse phase HPLC on a RP-18 column (water/aceto-
nitrile (40:60), isocratic solvent). Six photoproducts were separated in the
case of 1. All of them were fully characterized by 1H NMR, 13C NMR,
mass spectrometry and, in the case of oxetane 4, x-ray diffraction (see
the spectra in the Supporting Information, S4–S9). CCDC-266772 con-
tains the supplementary crystallographic data for this paper. These data
can be obtained free of charge from the Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Identification of photoproducts:
Oxetane 3: Yield 19%; 1H NMR (300 MHz, CDCl3): d = 8.03(s, 1H,
Ar-H), 7.40–6.90 (m, 9H, Ar-H + NH), 5.92 (dd, J = 9.4, 5.1 Hz, 1H,
H1’), 5.36 (s,1H, H6), 4.70 (dd, J = 12.7, 1.4 Hz, 1H, H5’), 4.46 (dd, J =
12.7, 1.9 Hz, 1H, H5’), 4.40 (m, 1H, H3’), 4.10 (m, 1H, H4’), 3.93 (q, J =
7.2 Hz, 1H, CH-CH3), 2.20 (m, 1H, H2’’), 2.05 (brs, 1H, OH), 1.80 (m,
The stereochemistry of oxetanes 3–6 and reduced cyclic photoproducts 7
and 8 was assigned by means of nuclear Overhauser effect (NOE) experi-
ꢁ
1H, H2’), 1.77 (s, 3H, C CH3), 1.61 ppm (d, J = 7.2 Hz, 3H, CH-CH3);
1
ments. On the other hand, the H and 13C signals were assigned by a com-
13C NMR (75 MHz, CDCl3): d = 173.1 (C), 170.0 (C), 150.6 (C), 144.4
(C), 139.4 (C), 137.6 (C), 129.1 (CH), 128.4 (CH), 128.3 (CH), 126.5
(CH), 126.2 (CH), 125.3(CH), 124.9 (CH), 92.0 (C), 84.6 (CH), 8.36
(CH), 77.2 (C, peak observed in the long-range HC correlation), 70.7
(CH), 63.4 (CH2), 61.0 (CH), 46.3(CH), 40.7 (CH 2), 22.9 (CH3),
19.1 ppm (CH3); HRMS (FAB) calcd. for C26H26O7N2: 478.1740; found
478.1750.
bination of H,H (COSY) and H,C (edited HSQC) correlations, along
with the NOE results. Quaternary carbons were assigned by long-range
H,C correlation (HMBC).
Determination ofthe photolysis quantum yields : Solutions (A254 = 0.4)
of the oxetanes 3, BP-dThd, and BP-DMT in acetonitrile/water (4:1)
under N2 were irradiated at l = 254 nm with a low-pressure Hg lamp.
The UV absorption of the carbonyl band at l = 254 nm was measured
after different irradiation times (0, 2, 3, 5, 7, and 10 s). The same proce-
dure was used for the irradiation of solutions (A350 ꢀ0.05) of dyad 1 or
equimolar mixtures of BP with dThd and BP with DMT in acetonitrile/
water (4:1) with a lamp emitting mainly at l = 350 nm (Gaussian distri-
bution).
Oxetane 4: Yield: 14%; 1H NMR (300 MHz, CDCl3): d = 8.08 (s, 1H,
Ar-H), 7.70–6.90 (m, 9H, Ar-H + NH), 6.38 (t, J = 7.5 Hz, 1H, H1’),
4.81 (dd, J = 12.3, 1.5 Hz, 1H, H5’), 4.74 (s, 1H, H6), 4.40 (dd, J = 12.3,
2.0 Hz, 1H, H5’), 4.35 (brs, 1H, H3’), 4.00–3.94 (m, 1H, H4’), 3.87 (q, J
= 7.2 Hz, 1H, CH-CH3), 2.20 (m, 2H, H2’ + H2’’), 1.80 (brs, 1H, OH),
1.75 (s, 3H, C CH3), 1.55 ppm (d, J = 7.2 Hz, 3H, CH-CH3); 13C NMR
ꢁ
(75 MHz, CDCl3): d = 173.6 (C), 168.8 (C), 151.2 (C), 145.8 (C), 139.4
(C), 138.6 (C), 128.9 (CH), 128.7 (CH), 128.2 (CH), 128.1 (CH), 125.1
(CH), 123.0 (CH), 121.8 (CH), 91.2 (C), 82.9 (CH), 82.0 (CH), 77.6 (C,
peak observed in the long-range HC correlation), 70.0 (CH), 61.3(CH 2),
The photolysis quantum yields were then obtained for each compound by
plotting the absorbance at 254 nm against the irradiation time, from the
slopes of each linear fitting. The quantum yield of 0.5 for the photolysis
of oxetane BP-DMT in acetonitrile was used as a standard.
560
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 553– 561

Benzophenone–Thymine Dyads
FULL PAPER
[27] J. G. Gould, M. G. Mercurio, C. A. Elmets, J. Am. Acad. Dermatol.
1995, 33, 551–573.
Acknowledgements
[28] F. Boscµ, G. Carganico, J. V. Castell, M. J. Gꢁmez-Lechꢁn, D.
Hernµndez, D. Mauleꢁn, L. A. Martínez, M. A. Miranda, J. Photo-
chem. Photobiol. B: Biol. 1995, 31, 133–138.
We thank the financial support given by the Spanish MCYT (Grant
CTQ 2004–03811), the Generalitat Valenciana (Grupos 03/082 and Proj-
ect GV04 A-349), and the UPV (Project PPI-06–03).
[29] F. Boscµ, M. A. Miranda, J. Photochem. Photobiol. B 1998, 43, 1–
26.
[30] F. Boscµ, M. L. Marín, M. A. Miranda, Photochem. Photobiol. 2001,
74, 637–655.
[31] F. Boscµ, M. A. Miranda, G. Carganico, D. Mauleon, Photochem.
Photobiol. 1994, 60, 96–101.
[32] A. Radschuweit, H. H. Ruttinger, P. Nuhn, W. Wohlrab, C. Huschka,
Photochem. Photobiol. 2001, 73, 119–127.
[1] J. Cadet, P. Vigny, in The Photochemistry of Nucleic Acids (Ed.: H.
Morrison), Wiley, New York, 1990, pp. 1–272.
[2] J.-S. Taylor, Pure Appl. Chem. 1995, 67, 183–190.
[3] J. Cadet, M. Berger, T. Douki, B. Morin, S. Raoul, J.-L. Ravanat, S.
Spinelli, Biol. Chem. 1997, 378, 1275–1286.
[4] J.-S. Taylor, Acc. Chem. Res. 1994, 27, 76–82.
[5] J. E. Cleaver, F. CortØs, D. Karentz, L. H. Lutze, W. F. Morgan,
A. N. Player, L. Vuksanovic, L. D. Mitchell, Photochem. Photobiol.
1988, 48, 41–49.
[33] J. Moser, Z. Sarabia, H. Minter, W. W. Lovell, G. M. J. Beijersber-
gen van Henegouwen, J. Photochem. Photobiol. B 2000, 58, 37–45.
[34] T. Artuso, J. Bernadou, B. Meunier, N. Paillous, Biochem. Pharma-
col. 1990, 39, 407–413.
[6] J.-S. Taylor, M. P. Cohrs, J. Am. Chem. Soc. 1987, 109, 2834–2835.
[7] P. E. M. Gibbs, B. J. Kilbey, S. K. Banerjee, C. W. Lawrence, J. Bac-
teriol. 1993, 175, 2607–2612.
[8] D. L. Mitchell, R. S. Nairn, Photochem. Photobiol. 1989, 49, 805–
819.
[9] C. A. Smith, M. Wang, N. Jiang, L. Che, X. Zhao, J.-S. Taylor, Bio-
chemistry 1996, 35, 4146–4154.
[10] A. Gentil, F. L. Page, A. Margot, C. W. Lawrence, A. Borden, A.
Sarasin, Nucleic Acids Res. 1996, 24, 1837–1840.
[35] T. Artuso, J. Bernadou, B. Meunier, J. Piette, N. Paillous, Photo-
chem. Photobiol. 1991, 54, 205–213.
[36] N. Chouini-Lalanne, M. Defais, N. Paillous, Biochem. Pharmacol.
1998, 55, 441–446.
[37] V. Lhiaubet, N. Paillous, N. Chouini-Lalanne, Photochem. Photobiol.
2001, 74, 670–678.
[38] I. von Wilucki, H. Matthaus, C. H. Krauch, Photochem. Photobiol.
1967, 6, 497–500.
[39] M. Charlier, C. HØlꢂne, Photochem. Photobiol. 1972, 15, 71–87.
[40] G. Prakash, D. E. Falvey, J. Am. Chem. Soc. 1995, 117, 11375–
11376.
[11] T. Todo, H. Takemori, H. Ryo, M. Ihara, T. Matsunaga, O. Nikaido,
K. Sato, T. Nomura, Nature 1993, 361, 371–374.
[12] T. Mizukoshi, K. Hitomi, T. Todo, S. Iwai, J. Am. Chem. Soc. 1998,
120, 10634–10642.
[41] K. Nakatani, T. Yoshida, I. Saito, J. Am. Chem. Soc. 2002, 124,
2118–2119.
[13] T. P. Begley, Acc. Chem. Res. 1994, 27, 394–401.
[14] A. Sancar, Biochemistry 1994, 33, 2–9.
[42] T. Delatour, T. Douki, C. DꢃHam, J. Cadet, J. Photochem. Photobiol.
B 1998, 44, 191–198.
[15] B. Morin, J. Cadet, J. Am. Chem. Soc. 1995, 117, 12408–12415.
[16] A. G. Bourdat, T. Douki, S. Frelon, D. Gasparutto, J. Cadet, J. Am.
Chem. Soc. 2000, 122, 4549–4556.
[17] J.-L. Ravanat, T. Douki, J. Cadet, J. Photochem. Photobiol. B 2001,
63, 88–102.
[18] T. Douki, J. Cadet, Int. J. Radiat. Biol. 1999, 75, 571–581.
[19] M. C. Marguery, N. Chouini-Lalanne, J. C. Ader, N. Paillous, Photo-
chem. Photobiol. 1998, 68, 679–684.
[43] I. G. Gut, P. D. Wood, R. W. Redmond, J. Am. Chem. Soc. 1996,
118, 2366–2373.
[44] S. Encinas, N. Belmadoui, M. J. Climent, S. Gil, M. A. Miranda,
Chem. Res. Toxicol. 2004, 17, 857–862.
[45] a) M. A. Miranda, L. A. Martínez, A. Samadi, F. Boscµ, I. M.
Morera, Chem. Commun. 2002, 280–281; b) F. Boscµ, I. Andreu,
I. M. Morera, A. Samadi, M. A. Miranda, Chem. Commun. 2003, 13,
1592–1593.
[20] C. Agapakis-Causse, F. Boscµ, J. V. Castell, D. Hernandez, M. L.
Marin, L. Marrot, M. A. Miranda, Photochem. Photobiol. 2000, 71,
499–505.
[46] A. Joseph, D. E. Falvey, J. Am. Chem. Soc. 2001, 123, 3145–3146.
[47] A. Joseph, G. Prakash, D. E. Falvey, J. Am. Chem. Soc. 2000, 122,
11219–11225.
[21] S. M. Starrs, R. J. H. Davies, Photochem. Photobiol. 2000, 72, 291–
297.
[22] G. Condorelli, L. L. Costanzo, G. De Guidi, S. Giuffrida, S. Sortino,
Photochem. Photobiol. 1995, 62, 155–161.
[48] a) G. Ciamician, P. Silber, Ber. Dtsch. Chem. Ges. 1900, 33, 2911–
2912; b) J. A. Bell, H. Linschitz, J. Am. Chem. Soc. 1963, 85, 528–
532; c) G. O. Schenck, H. D. Becker, K. H. Schulte-Elte, C. H.
Krauch, Chem. Ber. 1963, 96, 509–516.
[23] S. Encinas, M. J. Climent, S. Gil, U. O. Abrahamsson, J. Davidsson,
M. A. Miranda, ChemPhysChem 2004, 5, 1704–1709.
[24] P. D. Pigatto, N. Mozzanica, A. S. Bigardi, A. Legori, R. Valsecchi,
F. Cusano, A. Tosti, M. Guarrera, N. Balato, A. Sertoli, Contact Der-
matitis 1993, 29, 39–41.
[25] S. Ophaswongse, H. Maibach, Contact Dermatitis 1993, 29, 57–64.
[26] N. J. Neumann, E. Holzle, G. Plewig, T. Schwarz, R. G. Panizzon, R.
Breit, T. Ruzicka, P. Lehmann, J. Am. Acad. Dermatol. 2000, 42,
183–192.
[49] L. J. Martínez, C. Scaiano, J. Am. Chem. Soc. 1997, 119, 11066–
11070.
[50] a) Y. Wang, P. P. Gaspar, J. S. Taylor, J. Am. Chem. Soc. 2000, 122,
5510–5519; b) A. Sancar, Chem. Rev. 2003, 103, 6747–6753;
c) M. K. Cichon, S. Arnold, T. Carell, Angew. Chem. 2002, 114, 793–
796; Angew. Chem. Int. Ed. 2002, 41, 767–770.
Received: March 25, 2005
Published online: September 20, 2005
Chem. Eur. J. 2006, 12, 553– 561
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
561