Model studies of the (6-4) photoproduct photoreactivation: Efficient photosensitized splitting of thymine oxetane units by covalently linked tryptophan in high polarity solvents

Article abstract of DOI:10.1039/b514921e

Three covalently linked tryptophan-thymine oxetane compounds used as a model of the (6-4) photolyase-substrate complex have been prepared. Under 290 nm light, efficient splitting of the thymine oxetane with aromatic carbonyl compounds gives the thymine monomer and the corresponding carbonyl compounds by the covalently linked tryptophan via an intramolecular electron transfer, and exhibits a strong solvent dependence: the quantum yield (Φ) is ca. 0.1 in dioxane, and near 0.3 in water. Electron transfer from the excited tryptophan residue to the oxetane unit is the origin of fluorescence quenching of the tryptophan residue, and is more efficient in strong polar solvents. The splitting efficiency of the oxetane radical anion within the tryptophan ⁺-oxetane^- species is also solvent-dependent, ranging from ca. 0.2 in dioxane to near 0.35 in water. Thus, the back electron transfer reaction in the charge-separated species would be suppressed in water, but is still a main factor causing low splitting efficiencies in the tryptophan-oxetane systems. In contrast to the tryptophan-oxetane system, fast nonradiation processes are the main causes of low efficiency in the flavin-oxetane system. Hence, nonradiative processes of the excited FADH^-, rather than electron transfer to oxetane, may be an important factor for the low repair efficiency of (6-4) photolyase. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b514921e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Model studies of the (6–4) photoproduct photoreactivation: efficient
photosensitized splitting of thymine oxetane units by covalently
linked tryptophan in high polarity solvents†
Qin-Hua Song,*a Hong-Bo Wang,a Wen-Jian Tang,a Qing-Xiang Guoa and Shu-Qin Yub
Received 20th October 2005, Accepted 15th November 2005
First published as an Advance Article on the web 7th December 2005
DOI: 10.1039/b514921e
Three covalently linked tryptophan–thymine oxetane compounds used as a model of the (6–4)
photolyase–substrate complex have been prepared. Under 290 nm light, efficient splitting of the
thymine oxetane with aromatic carbonyl compounds gives the thymine monomer and the
corresponding carbonyl compounds by the covalently linked tryptophan via an intramolecular electron
transfer, and exhibits a strong solvent dependence: the quantum yield (U) is ca. 0.1 in dioxane, and near
0.3 in water. Electron transfer from the excited tryptophan residue to the oxetane unit is the origin of
fluorescence quenching of the tryptophan residue, and is more efficient in strong polar solvents. The
splitting efficiency of the oxetane radical anion within the tryptophan•+–oxetane•− species is also
solvent-dependent, ranging from ca. 0.2 in dioxane to near 0.35 in water. Thus, the back electron
transfer reaction in the charge-separated species would be suppressed in water, but is still a main
factor causing low splitting efficiencies in the tryptophan–oxetane systems. In contrast to the
tryptophan–oxetane system, fast nonradiation processes are the main causes of low efficiency in the
flavin–oxetane system. Hence, nonradiative processes of the excited FADH−, rather than electron
transfer to oxetane, may be an important factor for the low repair efficiency of (6–4) photolyase.
Introduction
The two major lesions in DNA induced by UV light (200–300 nm),
the cyclobutane pyrimidine dimers (CPDs) and the pyrimidine(6–
4) pyrimidone adducts ((6–4) photoproducts) (Fig. 1), are respon-
sible for the harmful effects of UV on organisms, such as growth
delay, mutagenesis and death, and constitute 70–80% and 20–30%
of the total photoproducts, respectively.1 The two photolesions can
be repaired through DNA photoreactivation catalyzed by CPD
photolyase and (6–4) photolyase, respectively.2
Although less readily formed than CPDs, (6–4) photoproducts
might actually be more effective at causing damaging mutations.3
The (6–4) photoproduct is thought to form in DNA as follows: [2 +
2] photoaddition (the so-called Paterno–Buchi reaction) of the C4
carbonyl (or amino) of the 3ꢀ thymine (cytosine) across the 5,6
double bond of the 5ꢀ thymine generate◦s an oxetane (or azetidine)
ring,2 which at temperatures above −80 C undergoes ring opening
by C4–O bond cleavage accompanied by a proton shift from N3
to generate the observed (6–4) photoproduct (Fig. 1).4
Fig. 1 Formation of the two major photoproducts in DNA under UV
light, CPDs and (6–4) photoproducts.
Whereas CPDs can be restored to their original forms by
simply breaking the C5–C5ꢀ and C6–C6ꢀ bonds, the breaking of
the C5–OH and C6–C4ꢀ bonds of (6–4) photoproducts would
not result in repair. The discovery of (6–4) photolyase and the
subsequent identification of structural and cofactor similarities of
(6–4) photolyase to CPD photolyase led to a proposal of a reaction
scheme very similar to that of CPD photolyase.5 In the model for
(6–4) photolyase, a critical step, in which (6–4) photolyase differs
from classical photolyase, is that upon binding to the substrate
the enzyme converts the open form of the (6–4) photoproduct
to the oxetane intermediate by a dark reaction. The next step is
photoinduced electron transfer (ET) to the oxetane intermediate
aDepartment of Chemistry, University of Science and Technology of China,
and State Key Laboratory of Elemento-Organic Chemistry, Nankai Uni-
versity, Hefei, 230026, Anhui, P. R. China. E-mail: qhsong@ustc.edu.cn;
Fax: +86-551-3601592; Tel: +86-551-3607524
bDepartment of Chemical Physics, University of Science and Technology of
China, Hefei, 230026, Anhui, P. R. China
† Electronic supplementary information (ESI) available: UV absorption
spectra and/or fluorescence emission spectra of compounds 2a, 3, 8b
and 10; the detail for the measurement of the splitting quantum yield;
copies of 1H and 13C NMR spectra of all new compounds. See DOI:
10.1039/b514921e
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 291–298 | 291
©

View Article Online
in a manner analogous to CPD photolyase, and spontaneous
splitting of the oxetane anion radical to restore two pyrimidines.2
To provide evidence for the ET mechanism, thymine oxetane
adducts with carbonyl compounds were prepared. Use of pho-
tosensitizers, laser flash photolysis and steady-state studies with
thymine oxetane adducts provided extensive support for the facile
cleavage of the oxetane anionic radical.6–8 The strongest support
for the ET model came from the investigation of photoinduced
cleavage of an oxetane ring covalently linked to flavin.7 In this
model system, it was found that only two-electron-reduced and
deprotonated flavin induced more efficient photosplitting of the
oxetane ring. This result further demonstrated the close mecha-
nistic similarities between CPD photolyase and (6–4) photolyase.
Although the two classes of enzymes have similar structures, the
same chromophores, and the same basic reaction mechanism, an
important difference in repair efficiency exists between them, that
is, CPD photolyases repair CPDs with a uniformly high quantum
yield (0.7–0.98),2 and the quantum yield of (6–4) photolyases is
in the range of 0.05–0.11.5a,9 X-Ray crystal structures for both
CPD photolyase10 and photolyase–substrate complexes11,12 have
been obtained, and the mechanistic model for CPD photolyase
has well been established.2 However, study of (6–4) photolyase is
very insufficient, for example, the mechanism for the low repair
efficiency of (6–4) photolyase remains unknown.
In the CPD model systems, the photosensitized cleavage of the
CPD unit by a covalently linked chromophore had low efficiency
due to back electron transfer (BET), which competes with the
cleavage of the CPD radical anion within a charge-separated
species after photo-induced ET between the chromophore and
the CPD unit. The photosensitized cleavage of the CPD unit in
the model systems by a covalently linked chromophore such as
indole,13a arylamine,13b methoxybenzene,13c flavin14 or tryptophan15
has been extensively investigated. These studies offer insights into
the mechanistic features of CPD repair by DNA photolyase. In
contrast to CPD model systems, study of a covalently linked
oxetane–chromophore model is very rare.
In this work, we prepared three model systems (1–3), which are
composed of a thymine oxetane adduct with benzophenone or
benzaldehyde and a covalently linked tryptophan unit (Scheme 1).
The photosensitized cleavage of thymine oxetane units gives the
thymine monomer and the corresponding carbonyl compounds
under 290 nm light. Fluorescence quenching of tryptophan and
the splitting efficiency of the oxetane in these tryptophan–oxetane
systems were measured as functions of solvent polarity, and
revealed a strong solvent dependence.
Results and discussion
Synthesis of the model compounds 1–3
For the synthesis of the model compounds 1–3, 1-(carboxy-
methyl)thymine 614b,16 and its benzyl ester 7 were prepared from
thymine as shown in Scheme 2. The acetonitrile solution of
the benzyl ester and benzophenone was irradiated in a Pyrex
photochemical reactor (k > 290 nm) with a 300 W high-pressure
Hg lamp, and gave a thymine oxetane adduct 8a by the Paterno–
Buchi reaction. Using benzaldehyde instead of benzophenone, the
photochemical reaction would give 8b.
Tryptophan methyl ester (9)15 and 1-methyltryptophan methyl
ester (10)17 were prepared according to the literature methods
(Scheme 3). Compound 9 was synthesized with methanol in
the presence of SOCl2 in good yield. 1-Methyltryptophan was
prepared using tryptophan and iodomethane, then esterified with
methanol to give 10.
Scheme 1
Scheme 2 i) Chloroacetic acid, KOH, H2O, reflux; ii) benzyl alcohol, p-toluene sulfonic acid, benzene, reflux; iii) benzophenone or benzaldehyde,
hm > 290 nm, acetonitrile, room temp.
Scheme 3 i) Na, liquid NH3, Fe(NO3)3·9H2O, MeI; ii) SOCl2, CH3OH.
292 | Org. Biomol. Chem., 2006, 4, 291–298
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Scheme 4 i) Pd/C, H2, acetic acid, room temp., 8 h; ii) HOBt, TBTU, DMF, room temp., 8 h.
The model compounds 1–3 were prepared through con-
densation of the modified tryptophans 9 and 10 with N1-
(carboxylmethyl)thymine oxetanes obtained by hydrogenolytic
cleavage of the benzyl esters 8a and 8b, as shown in Scheme 4. The
condensation was performed in the presence of 1-hydroxy-1H-
benzotriazole (HOBt) in DMF for 8 h using 2-(1H-benzotriazole-
1-yl)-1,1,3,3-tetramethyluroniumtetrafluoroborate (TBTU) as a
coupling reagent, and generated compound 1 from 9 with 8a,
compound 2 from 10 with 8a and compound 3 from 9 with 8b.
Because racemic 8a or 8b was linked to L-tryptophan or N-methyl-
L-tryptophan, model compounds 1, 2 and 3 are diastereomeric.
The two diastereomers of model compounds 1, 2 and 3 were
separated by flash chromatography on silica gel. The two diastere-
omers of compound 2—2a and 2b—were characterized, and only
one diastereomer was characterized for compounds 1 and 3 (see
Experimental). The characterized diastereomers for compounds 1
and 3 were used in all experiments of analysis and measurements.
The splitting reaction is an intramolecular process because
splitting efficiencies of the model compounds are unchanging for
measurements at three concentrations—0.01 mM, 0.05 mM and
0.1 mM in methanol—within an experimental error of 5%. These
results ruled out the involvement of intermolecular photosensiti-
zation, i.e., the excited tryptophan residue of one model molecule
being responsible for the splitting of the oxetane unit of another
model molecule.
Fluorescence emission of model compounds 1–3
Fluorescence emission spectra of model compounds 1, 2a and
3 (symbol as Trp–Ox) in various solvents were measured at
room temperature on a fluorescence spectrometer. Fig. 3 shows
the fluorescence emission of compound 1 compared to the
fluorescence emission of the corresponding free tryptophan 9 in
methanol at room temperature. The fluorescence of the tryptophan
residue in compound 1 is much weaker than that of the free
tryptophan 9.
Photo-cleavage properties of compounds 1–3
The model compounds in an acetonitrile solution were irradiated
with a 290 nm light beam from a fluorescence spectrometer.
The HPLC chromatograms showed the simultaneous cleavage
of model compound 1, 2 or 3 into 4 or 5 and benzophenone
or benzaldehyde under 290 nm light. This was confirmed by
HPLC co-elution with standards. Fig. 2 shows a representative
set of HPLC chromatograms for model compound 2a irradiated
for various times. Obviously, the photochemical reaction of
model compound 2a, with a retention time of 11.3 min, to the
photosplitting product 5 and benzophenone with retention times
of 4.2 min and 8.9 min, respectively, is a clean conversion as no
other product was detectable.
Fig. 3 UV absorption spectra (solid line) and/or fluorescence emission
spectra (dashed line) (kex = 290 nm) of tryptophan methyl ester (9),
compounds 1 and 8a, in methanol.
The concentrations of the tryptophan unit for model com-
pounds 1, 2a and 3 and the free tryptophans 9 or 10 were controlled
within 0.05 for their absorbance at 290 nm. The fluorescence
intensity was further normalized with the absorbance. It was
observed that the fluorescence emission of the tryptophan unit
in model compounds 1–3 is quenched by the covalently linked
thymine oxetane, and the extent of fluorescence quenching [Q,
eqn (1), where F is fluorescence intensity] was expressed as follows:
Fig. 2 Typical HPLC chromatograms obtained from compound 2a in
acetonitrile (5 × 10−4 mol L−1, 3 mL) irradiated for 0, 1, 4, 7, 11 and
17 min by a HPLC instrument (HP Agilent 1100) with a C-18 reverse-phase
column and detected at 270 nm, methanol–water (70 : 30) as eluent.
Q = 1 − FTrp–Ox/FTrp
(1)
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 291–298 | 293
©

View Article Online
Table 1 Splitting quantum yield and degree of fluorescence quenching of
compounds 1 and 3 in various solvents
Quantum yields for cleavage of compounds 1–3
To measure the splitting quantum yields U, which are molecules
of oxetane split per photon absorbed, solutions (3 mL) of 1, 2
and 3 in various solvents were placed in quartz cuvettes (10 ×
10 mm) with a Teflon stopper, and then irradiated with 290 nm
light from a fluorescence spectrometer. The rates of oxetane unit
split were measured by monitoring the increase in absorbance
at 270 nm due to the regeneration of the 5,6-double bond of
pyrimidine 4 or 5 (symbol as Trp–Thy) and benzophenone or
benzaldehyde. The intensity of the light beam was measured by
ferrioxalate actinometry.19 Thus, the rate of photons absorbed
was obtained from the absorbance at 290 nm in terms of Beer’s
law. The observed quantum yields of oxetane splitting of 1,
2 and 3 were calculated according to U = (rate of oxetane
split)/(rate of photon absorbed), and listed in Tables 1, 2 and 3.
The splitting reactions of the model compounds reveal notable
solvent effects, and more efficient splitting reactions occur in high
polarity solvents, ranging from ca. 0.1 in dioxane to near 0.3 in
water.
There may exist different conformations between two diastere-
omers for each model compound (1, 2 or 3), which are covalently
linked between racemic oxetanes and an L-Trp residue, in solutions,
i.e., there are different relative orientations of the tryptophan
and oxetane rings. Thus, the different conformations of the
diastereomeric model compounds in solution could result in
different splitting efficiencies.
To assess the effect of differences in conformation on the
splitting quantum yield, we measured the quantum yield of the
two diastereomers of compound 2, 2a and 2b, as listed in Table 2.
It is obvious that the splitting quantum yield values for 2a and
2b are similar within an experimental error of 5%. Hence, the
conformational difference of 2a and 2b didn’t influence their
splitting quantum efficiencies.
Compound 1
Compound 3
Solvent
Q
U
φspl
Q
U
φs
H2O–CH3CN (95 : 5)
Acetonitrile
Methanol
THF
Dioxane
0.80
0.78
0.69
0.65
0.49
0.24
0.17
0.15
0.13
0.10
0.30
0.22
0.22
0.20
0.20
0.82
0.73
0.71
0.63
0.61
0.28
0.24
0.19
0.17
0.11
0.34
0.32
0.27
0.26
0.18
Table 2 Dependence of the efficiency on the solvent (er, dielectric
constants) for the two diastereomers of compound 2
2a
2b
Solvent
er
Q
U
φspl
U
H2O–CH3CN (95 : 5)
Acetonitrile
Methanol
THF
Dioxane
78.3a
37.8
32.7
7.6
0.80
0.67
0.68
0.51
0.47
0.26
0.17
0.16
0.13
0.11
0.33
0.26
0.24
0.26
0.23
0.27
0.18
0.17
0.12
0.11
2.2
a Dielectric constant of water.
The value of Q for compound 1 in methanol is 0.69 (Fig. 3). The
extent of fluorescence quenching is sensitive to the polarity of the
solvent. The values of Q increase with increasing polarity of the
solvents (Tables 1 and 2).
The fluorescence quenching of the tryptophan residue in the
model compounds is not a result of significant absorption of the
exciting light (290 nm) by the oxetane unit. Since the oxetanes
8a (Fig. 3) and 8b have no significant absorption at 290 nm, an
internal filter effect should not be significant.
Kim and Rose13b observed a two-fold higher splitting efficiency
of one over another diastereomer in a dimer–arylamine system
with a very short spacer. The large difference in splitting efficiency
may be a consequence of different BET rates within the diastere-
omeric charge-separated species. The stereoelectronic explanation
was suggested that interaction of the dimer radical anion and the
arylamine radical cation electronic system is different owing to
differences in the distances and angles between them.13b Obviously,
these differences in a system with a short spacer are more
remarkable than in a system with a long spacer, and the former
results in a notable difference in the quantum yield. In our model
systems with a longer spacer, a small difference in the distance
An intramolecular ET from the excited tryptophan to the
oxetane unit may be responsible for the fluorescence quenching.
On the basis of laser flash photolysis, fluorescence quenching,
and product analysis experiments, Falvey and coworkers6b demon-
strated that thymine oxetane adducts with aromatic carbonyl com-
pounds undergo a cycloreversion reaction upon photosensitizer
reductive ET reactions. The excited state oxidation potentials
(E*ox) for the photosensitizers range from −2.45 V to −3.32 V.6b
Since the value of E*ox for tryptophan is −2.78 V vs. SCE from
1.05 V vs. NHE18 for the oxidation potential of tryptophan, the
electron transfer reaction from the excited tryptophan to the
thymine oxetane unit is thermodynamically possible. Because there
is almost no overlap between the emission spectra of the free
tryptophans 9 and 10 and the absorption spectra of the oxetanes
8a and 8b, singlet–singlet energy transfer is an improbable pathway
of fluorescence quenching in the model compounds. Hence, the ET
reaction should be the reason for the fluorescence quenching in
the model compounds, and the degree of fluorescence quenching
Q should reflect the efficiency of the electron transfer reaction.
The values of Q in Tables 1 and 2 show that ET reactions
from the excited tryptophan residue to the linked oxetane become
more efficient in high polarity solvents. With increasing polarity of
the solvent, decay processes caused by electron transfer would be
predominant, and fluorescence and nonradiative processes would
be suppressed.
Table 3 The splitting quantum yields of compound 3 in water–dioxane
binary solvent
Water : dioxane(%)
U
100a : 0
80 : 20
60 : 40
40 : 60
20 : 80
10 : 90
0 : 100
0.28
0.22
0.21
0.20
0.16
0.14
0.11
a Aqueous solution containing 5% acetonitrile.
294 | Org. Biomol. Chem., 2006, 4, 291–298
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
resulting from different angles between the donor and the acceptor
would give similar quantum yields.
The quantum yields are 2–3 times as high as the value from
the Trp–CPD system.15 The result is well in agreement with the
observation from a covalently linked flavin–oxetane system, i.e.,
the quantum yield of the flavin–oxetane system was 0.023, and 0.01
for a flavin–CPD system.7 The two-fold higher splitting quantum
yield of the oxetane system than the CPD system can be explained
by the splitting rate of the oxetane anion radical (>107 s−1)6b being
faster than that of the CPD radical anion (∼106 s−1).20 A fast
splitting of the oxetane radical anion can efficiently compete with
BET within the zwitterionic intermediate.
Furthermore, a solvent dependence was observed in solvent
mixtures of water–dioxane (Table 3). Upon addition of dioxane,
a gradual decrease of the quantum yield was observed until the
lowest U was reached in pure dioxane. The systematic reduction
of solvent polarity results in a significant decrease of the quantum
yield of the splitting reaction, ranging from 0.28 in aqueous
solution containing 5% acetonitrile to 0.11 in pure dioxane for
compound 3. This further demonstrates that an efficient splitting
reaction occurs in high polarity solvents.
The cleavage of compound 3 is more efficient than that
of compound 1. This is in agreement with the more efficient
photosensitized splitting of a free 1,3-dimethyluracil oxetane
adduct with benzaldehyde than with benzophenone.8 The splitting
quantum yields of compound 2 are similar to compound 1. If a
deprotonation of the tryptophan radical cation at N1 occurs prior
to the cleavage of the oxetane radical anion in the zwitterionic
intermediate, the BET would be strongly weakened in a diradical
anion, Trp•–Ox•−, to result in a high splitting efficiency. This also
confirmed that the tryptophan radical cation doesn’t deprotonate
before cleavage of the oxetane radical anion.15
The photophysical and photochemical processes of the model
compounds were illuminated with the simple mechanism shown
in Scheme 5. Under >290 nm light, a tryptophan unit in the
model compound absorbs a photon producing an excited state
of tryptophan. The excited tryptophan residue has relaxation
pathways as follows: fluorescence (kf), internal conversion and
intersystem crossing (together represented by knr), and electron
transfer to the linked oxetane (ket). These efficiencies sum to 1,
i.e. φf + φnr + φet = 1. The charge-separated species (Trp•+–Ox•−)
formed via electron transfer can undergo two processes, splitting
(φspl) or back electron transfer (φbet) resulting in an unproductive
reversal, and their efficiencies sum to 1. Among these processes,
ket and kspl contribute to the observed quantum yield of oxetane
splitting (U), i.e. U = φet × φspl.
In contrast to the Trp–Ox model systems, the splitting quantum
yield of the flavin–Ox model system is very low, U = 0.023.7
The change in the free energy of BET, DG◦, can be expressed as
follows:
−DG◦ = E◦ − E◦ − e2/er
(2)
ox
red
where E◦ and E◦ are redox potentials for chromophore and
ox
red
oxetane respectively. For the flavin–oxetane system, the BET is a
charge shift and no coulomb interaction occurs, and the coulomb
term for the Trp–Ox system can be neglected for highly polar
solvents such as water. Since Eox = 0.124 V21 for fully-reduced
flavin and 1.05 V (NHE)18 for tryptophan, the value of −DG◦ for
the Trp–Ox system is higher than that of the flavin–Ox system,
i.e. the rate of BET in Trp•+–Ox•− would be faster than that in
flavin•–Ox•−. However, the slow BET in the flavin–Ox system7 and
the flavin–CPD system14 didn’t result in higher splitting quantum
yields, and they are one order of magnitude lower than that of the
Trp–Ox system and the Trp–CPD system,15 respectively. Because
the fluorescence (kf) of fully-reduced flavin can be neglected,
the main processes competing with electron transfer ket are
nonradiative processes knr. Therefore, the main reason resulting
in low efficiency of flavin model systems should be knr rather
than kbet. This implies that the apoenzyme of CPD photolyase
nonconvalently binding the flavin cofactor can efficiently suppress
knr of the excited FADH−.
Besides, a fast kspl (>109 s−1 in the enzyme,22 and ∼106 s−1
for the model system20), which can efficiently compete with the
BET process, is also an important factor for high repair efficiency
of CPD photolyase. In contrast to CPD photolyase, inefficient
suppression of knr or slow kspl may be responsible for the low repair
efficiency of (6–4) photolyase. Cocrystal structures of photolyase–
DNA complexes and ultrafast spectroscopic studies on (6–4)
photolyase would be helpful for understanding the mechanistic
details of the photoreactivation, and might shed some light on the
reason for this major difference in the quantum yields of the two
types of photolyase.
Scheme 5
The efficiency of forward electron transfer (ket) can further be
discussed with a reasonable assumption, that is, the rate constants
for fluorescence and for the nonradiative relaxation pathways
are unaltered by attachment of the oxetane to tryptophan and
the electron transfer can occur in the excited model compounds.
Utilizing this assumption, it was deduced that the efficiency of the
electron transfer is equal to the degree of fluorescence quenching,
φet = Q.15 Thus, the calculated values of φspl were obtained from
φspl = U/Q, and were listed in Tables 1 and 2. It is obvious that the
splitting efficiencies of the oxetane radical anions are remarkably
dependent on solvent polarity. On the basis of the data of Q and φspl
in Tables 1 and 2, it might reasonably be expected that the electron
transfer (ket) and splitting (kspl) processes are accelerated in high
polarity solvents, or that fluorescence and BET are suppressed.
Experimental
General
1
Melting points are uncorrected. H and 13C NMR spectra were
measured with a Bruker AV 300 spectrometer operating at 300
and 75 MHz, respectively. The chemical shifts were referenced to
CHCl3 (d 77.16) in CDCl3 and DMSO (d 39.52) in [D6]DMSO for
13C NMR. IR spectra were recorded in KBr and measured in cm−1
on a Bruker Vector22 Infrared Spectrometer. Mass spectra were
obtained with a Micromass GCF TOF mass spectrometer. All
materials were obtained from commercial suppliers and were used
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 291–298 | 295
©

View Article Online
as received. Solvents of technical quality were distilled prior to use.
DMF was dried overnight with MgSO4 and distilled. Acetonitrile
and methanol were spectroscopic grade from commercial suppliers
and used without further purification.
2.0 mmol) and benzophenone (728 mg, 4.0 mmol) were dissolved
in acetonitrile (250 mL) and placed in a Pyrex reactor. Under a
nitrogen atmosphere, the solution was irradiated for 4 h with a
300 W high-pressure Hg lamp. The solvent in the reaction mixture
was removed by rotary evaporation. The residue was dissolved
with dichloromethane and subjected to column chromatography
(silica gel-H, petroleum ether–ethyl acetate 5 : 1) to yield 8a as a
Measurement of steady-state fluorescence emission
Fluorescence emission spectra were measured at room tem-
perature on a Perkin-Elmer Instruments LS55 Luminescence
Spectrometer. To determine the extent of fluorescence quenching,
the fluorescence intensity of 1, 2 or 3 was compared to that of
the corresponding tryptophan without an oxetane attached (8a
or 8b), that is Q = 1 − FTrpH–Ox /FTrpH . The concentrations of the
tryptophan residue of the Trp–Ox and the free tryptophan, 8a or
8b, were controlled within 0.05 for absorbance at 290 nm, and
the fluorescence intensities were normalized with the absorbances.
The wavelength of excitation was 290 nm, which is an absorption
peak of the model compounds resulting from the n → p* transition
of the indole unit.
white powder (196 mg, 21%). Mp 167–169 ◦C. mmax(KBr)/cm−1
=
1748m, 1699s, 1469m, 1199s; 1H NMR (300 MHz, CDCl3, TMS):
d = 1.71 (s, 3H, CH3), 3.55 (d, 1H, NCH2), 4.69 (s, 1H, CH), 4.75
(d, 1H, NCH2), 5.21 (s, 2H, OCH2), 7.29–7.38 (m, 15H, ArH);
13C NMR (75 MHz, CDCl3): d = 23.5 (CH3), 47.3 (NCH2), 66.4
(OCH2), 67.8 (NCH), 76.6 (CCH3), 91.2 (OC), 125.3, 126.0,
128.2, 128.4, 128.6, 128.7, 128.8, 128.9, 135.0, 138.5, 143.9, 151.0
(NCON), 168.2 (COO), 169.6 (NCO); TOFMS (CI) calcd for
(M + 1)+ C27H24N2O5: 457.1763, found 457.1760.
Benzyl-(6Z,8Z)-6-methyl-8-phenyl-7-oxa-3,5-dioxo-2,4-diaza-
bicyclo[4.2.0]octane-2-acetate 8b. Using benzaldehyde instead
of benzophenone, the same procedure was performed to give
compound 8b (130 mg, yield 17%) as white needles. Mp 182–
184 ◦C; mmax(KBr)/cm−1 = 1735s, 1697s, 1479m, 1206m; 1H NMR
(300 MHz, CDCl3, TMS): d = 1.81 (s, 3H, CH3), 3.96 (d, J =
6.6 Hz, 1H, NCH), 4.06 (m, 2H, NCH2), 5.14 (m, 2H, OCH2),
5.75 (d, J = 6.6 Hz, 1H, OCH), 7.29–7.36 (m, 10H, ArH), 7.59
(s, 1H, NH); 13C NMR (75 MHz, [D6]DMSO): d = 22.3 (CH3),
48.3 (NCH2), 64.2 (NCH), 66.3 (OCH2), 77.2 (CCH3), 85.6
(OCH), 126.3, 128.1, 128.2, 128.4, 128.5, 128.7, 135.5, 139.1,
151.4 (NCON), 168.9 (COO), 170.3 (NCO); TOFMS (CI) calcd
for (M + 1)+ C21H20N2O5: 381.1450, found 381.1456.
Model compound 1. A suspension of Pd/C catalyst (15 mg)
in AcOH (2 mL) was slowly added to a solution of 8a (100 mg,
0.22 mmol) in 6 mL of AcOH. The mixture was bubbled with
hydrogen and stirred for 8 h at room temperature. The reaction
mixture was filtered through a Celite pad. The filtrate was
evaporated to dryness in vacuo, and a solution of TBTU (141 mg,
0.44 mmol) and HOBt (59 mg, 0.44 mmol) in 8 mL of DMF
was added. After the reaction mixture was stirred for 30 min,
a solution of tryptophan methyl ester 9 (52 mg, 0.24 mmol) in
DMF (2 mL) was added, and stirred for 8 h at room temperature.
The reaction mixture was diluted with 100 mL of water and
extracted with 200 mL of CHCl3 three times. The combined
organic layers were dried with MgSO4, filtered, and concentrated in
vacuo. The residual crude product was purified by silica gel column
chromatography (gel-H, CHCl3–petroleum ether 50 : 1) to give two
diastereomers, and the less polar isomer 1 as white powder (15 mg,
Measurements of splitting quantum yields of compounds 1–3
The 3 mL samples were irradiated with a 290 nm light from a
Shimadzu RF-5301PC spectrofluorophotomer. After certain time
intervals, the absorbance of the irradiated solutions was recorded
by a Lambda 45 UV–Vis spectrometer (Perkin-Elmer Instru-
ments). The quantum yields of splitting didn’t change with and
without N2-bubbling prior to irradiation within an experimental
error of 5%. Hence, the non-deaerated solution was employed in
all measurements of quantum yield. Since the photoproducts, the
pyrimidines and benzophenone or benzaldehyde, absorb light at
290 nm, in order to avoid competition of absorbing the irradiated
light between the model compounds and photoproducts, the
splitting extent of model compounds was controlled within 5%
in all the measurements.
1-(Carboxymethyl)thymine 6. 14b,16 Yield 61%. mmax(KBr)/cm−1
=
3026m, 1707s, 1664s, 1632s, 1419m, 1201s, 831m; 1H NMR
(300 MHz, [D6]DMSO, TMS): d = 1.76 (s, 3H, CH3), 4.37 (s,
2H, CH2), 7.50 (s, 1H, CH); 13C NMR (75 MHz, [D6]DMSO):
d = 11.9 (CH3), 48.5 (CH2), 108.5 (CCH3), 141.9 (CH), 151.0
(NCON), 164.3 (COOH), 169.6 (NCO).
Benzyl 1,2,3,4-tetrahydro-5-methyl-2,4-dioxopyrimidine-1-
acetate 7. 1-(Carboxymethyl)thymine 6 (1.8 g, 10.0 mmol) and
p-toluene sulfonic acid (0.8 g) were added to a solution of benzyl
alcohol (4 mL, 39 mmol) and benzene (30 mL), and heated to
reflux for 15 h; water was continuously removed from the reaction
mixture. The reaction mixture was concentrated in vacuo and the
precipitate was recrystallized twice from methanol to give 7 as
12%) was characterized. Mp 204–206 ◦C. mmax(KBr)/cm−1
=
3336m, 1728s, 1686s, 1470w, 1282w, 1211w; 1H NMR (300 MHz,
CDCl3, TMS): d = 1.58 (s, 3H, CH3), 3.37 (m, 3H, CHCH2 +
NCH2), 3.74 (s, 3H, OCH3), 4.40 (d, 1H, NCH2), 4.81 (s, 1H,
NCH), 4.87 (m, 1H, CHCH2), 6.40 (d, 1H, NHCH), 7.04–7.46
(m, 15H, ArH + Hindole), 8.36 (s, 1H, NH), 8.53 (s, 1H, NHindole);
13C NMR (75 MHz, CDCl3): d = 23.3 (CH3), 27.1 (CHCH2),
48.7 (NCH2), 52.7 (OCH3), 52.9 (NHCH), 66.2 (NCH), 76.1
(CCH3), 91.1 (OC), 108.7, 111.5, 118.3, 119.6, 122.1, 123.9, 125.4,
126.0, 127.4, 128.2, 128.3, 128.5, 128.7, 136.1, 138.6, 143.5, 151.7
(NCONH), 166.8 (COO), 169.7 (NHCO), 172.0 (CH2CONH);
TOFMS (CI) calcd for (M + 1)+ C32H30N4O6: 567.2244, found
567.2247.
colorless needles (2.2 g, 80%). Mp 179–181 ◦C. mmax(KBr)/cm−1
=
1738s, 1460s, 1224s; 1H NMR (300 MHz, CDCl3, TMS): d = 1.92
(s, 3H, CH3), 4.47 (s, 2H, NCH2), 5.22 (s, 2H, OCH2), 6.92 (s,
1H, CH), 7.34–7.41 (m, 5H, ArH), 8.37 (s, 1H, NH); 13C NMR
(75 MHz, [D6]DMSO): d = 11.8 (CH3), 48.5 (NCH2), 66.4 (OCH2),
108.7 (CCH3), 127.9, 128.2, 128.4, 135.5, 141.5, 151.0 (NCON),
164.3 (COO), 168.1 (NCO); TOFMS (CI) calcd for (M + 1)+
C14H14N2O4: 275.1032, found 275.1031.
Benzyl-Z-6-methyl-8,8-diphenyl-7-oxa-3,5-dioxo-2,4-diazabicyclo-
[4.2.0]octane-2-acetate 8a. The benzyl ester
7 (548 mg,
296 | Org. Biomol. Chem., 2006, 4, 291–298
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Model compound 2. Using 1-methyltryptophan methyl ester
hydrochloride instead of tryptophan methyl ester 9, the same
procedure as with synthesizing compound 1 was performed, and
two diastereomers 2a (24 mg, 19%) and 2b (19 mg, 15%) were
125.9, 127.0, 128.4, 136.1, 139.5, 151.2 (NCONH), 167.4 (COO),
170.2 (NHCO), 171.9 (CH2CONH); TOFMS (EI) calcd for (M+)
C26H26N4O6: 490.1852, found 490.1833.
obtained as white powders. 2a: mp 162–164 ◦C. mmax(KBr)/cm−1
=
1-(Carboxymethyl)thymine tryptophan methyl ester amide 4. 1-
(Carboxymethyl)thymine 6 (230 mg, 1.25 mmol) and an excess
of BOP (608 mg, 1.37 mmol) were dissolved in DMF (6 mL)
and stirred at room temperature for 30 min, and a solution of
tryptophan methyl ester 9 (272 mg, 1.25 mmol) in DMF (2 mL)
was added, and stirred for 5 h at room temperature. The reaction
mixture was purified by column chromatography on aluminium
oxide (100–200 mesh, ethyl acetate–methanol 20 : 1 → 3 : 1)
to give 4 as a white powder (134 mg, 28%). Mp 207–209 ◦C.
mmax(KBr)/cm−1 = 3351s, 3311w, 1730s, 1688s, 1646s, 1354w,
1701s, 1475m, 1282m, 1213m; 1H NMR (300 MHz, CDCl3, TMS):
d = 1.56 (s, 3H, CH3), 3.35 (m, 3H, CHCH2 + NCH2), 3.70 (s,
3H, NCH3), 3.75 (s, 3H, OCH3), 4.43 (d, 1H, NCH2), 4.73 (s,
1H, NCH), 4.88 (m, 1H, CHCH2), 6.29 (d, 1H, NHCH), 6.84
(s, 1H, Hindole), 7.04–7.36 (m, 14H, ArH + Hindole); 13C NMR
(75 MHz, CDCl3): d = 23.3 (CH3), 27.3 (CHCH2), 32.8 (NCH3),
49.1 (NCH2), 52.7 (OCH3), 53.4 (NHCH), 66.2 (NCH), 76.5
(CCH3), 91.3 (OC), 107.9, 109.6, 118.6, 119.5, 122.1, 125.5, 126.1,
127.9, 128.2, 128.3, 128.4, 128.6, 128.8, 137.1, 138.8, 143.8, 151.0
(NCONH), 167.0 (COO), 169.4 (NHCO), 172.1 (CH2CONH);
TOFMS (CI) calcd for (M + 1)+ C33H32N4O6: 581.2400, found
581.2398. 2b: mp 120–122 ◦C. mmax(KBr)/cm−1 = 1709s, 1469m;
1H NMR (300 MHz, CDCl3, TMS): d = 1.38 (s, 3H, CH3), 3.31
(m, 3H, CHCH2 + NCH2), 3.71 (s, 6H, NCH3 + OCH3), 4.57 (d,
1H, NCH2), 4.86 (s, 1H, NCH), 4.89 (m, 1H, CHCH2), 6.35 (d, 1H,
NHCH), 6.79 (s, 1H, Hindole), 7.06–7.46 (m, 14H, ArH + Hindole);
13C NMR (75 MHz, CDCl3): d = 22.9 (CH3), 26.9 (CHCH2),
32.9 (NCH3), 49.0 (NCH2), 52.7 (OCH3), 53.1 (NHCH), 65.3
(NCH), 76.3 (CCH3), 91.3 (OC), 107.8, 109.8, 118.3, 119.8, 122.3,
125.5, 126.1, 127.7, 128.1, 128.3, 128.5, 128.6, 128.8, 137.1, 138.5,
143.6, 150.7 (NCONH), 167.3 (COO), 169.0 (NHCO), 171.9
(CH2CONH); TOFMS (EI) calcd for (M+) C33H32N4O6: 580.2322,
found 580.2319.
1
1221m; H NMR (300 MHz, [D6]DMSO, TMS): d = 1.73 (s,
3H, CH3), 3.09 (m, 2H, CHCH2), 3.57 (s, 3H, OCH3), 4.34 (m,
2H, NCH2), 4.53 (m, 1H, CHCH2), 6.97–7.50 (m, 6H, NCH +
Hindole), 8.68 (d, 1H, NHCH), 10.89 (s, 1H, NHindole), 11.26 (s,
1H, CNHC); 13C NMR (75 MHz, [D6]DMSO): d = 11.9 (CH3),
27.3 (CHCH2), 49.0 (NCH2), 51.9 (OCH3), 53.4 (NHCH), 108.0,
109.1, 111.5, 118.0, 118.5, 121.0, 123.7, 127.1, 136.0, 142.3, 150.9
(NCONH), 164.4 (COO), 167.1 (NHCO), 172.0 (CH2CONH);
TOFMS (CI) calcd for (M + 1)+ C19H20N4O5: 385.1512, found
385.1518.
1-(Carboxymethyl)thymine (1-methyltryptophan) methyl ester
amide 5. The same procedure as with the synthesis of 4 was
performed with 1-methyltryptophan methyl ester hydrochloride
and triethylamine instead of 9, and 5 was obtained as a white
powder (300 mg, 60%). Mp 196–198 ◦C. mmax(KBr)/cm−1 = 3435m,
1
Model compound 3. Sodium hydroxide (30 mg) and oxetane 8b
(190 mg, 0.5 mmol) were added to a solution of water (10 mL) and
methanol (10 mL), and stirred for 2 h at room temperature. The
reaction mixture was diluted with 100 mL of water and extracted
twice with 100 mL of ethyl acetate. Diluted hydrochloric acid
was dropped into the water layer to adjust the pH to 3. The
solution was extracted twice with 100 mL of ethyl acetate. The
organic phase was washed with water, dried with MgSO4, filtered
and concentrated in vacuo. Then a solution of TBTU (321 mg,
1.0 mmol) and HOBt (68 mg, 0.5 mmol) in 8 mL of DMF was
added. The reaction mixture was stirred for 30 min. After the
addition of a solution of tryptophan methyl ester 9 (131 mg,
0.6 mmol) in DMF (2 mL), the reaction mixture was stirred
for 8 h at room temperature. Then it was diluted with 100 mL
of water and extracted twice with 100 mL of ethyl acetate. The
combined organic layers were dried with MgSO4, filtered and
concentrated in vacuo. The residual crude product was purified
by silica gel column chromatography (gel-H, CHCl3–methanol
100 : 1). Two diastereomers were obtained, and the less polar
isomer 3◦as white powder (26 mg, 11%) was characterized. Mp
1740m, 1690s, 1666m, 1470w; H NMR (300 MHz, [D6]DMSO,
TMS): d = 1.74 (s, 3H, CH3), 3.09 (m, 2H, CHCH2), 3.58 (s,
3H, NCH3), 3.73 (s, 3H, OCH3), 4.33 (s, 2H, NCH2), 4.53 (m,
1H, CHCH2), 7.01–7.51 (m, 6H, NCH + Hindole), 8.67 (d, 1H,
NHCH), 11.26 (s, 1H, NH); 13C NMR (75 MHz, [D6]DMSO):
d = 11.8 (CH3), 27.0 (NCH3), 32.3 (CHCH2), 49.0 (NCH2), 51.8
(OCH3), 53.2 (NHCH), 107.9, 108.4, 109.6, 118.2, 118.5, 121.1,
127.4, 128.2, 136.5, 142.2, 150.8 (NCONH), 164.3 (COO), 167.0
(NHCO), 171.8 (CH2CONH); TOFMS (CI) calcd for (M + 1)+
C20H22N4O5: 399.1668, found 399.1667.
Acknowledgements
The authors are grateful for the financial support by the National
Natural Science Foundation of China (Grant no. 30470444,
20332020).
References
1
249–251 C. mmax(KBr)/cm−1 = 3392s, 1702s, 1478m, 1283m; H
1 J. Cadet and P. Vigny, in The Photochemistry of Nucleic Acids,
Bioorganic Photochemistry, ed. H. Morrison, Wiley Interscience, New
York, 1990, vol. 1, pp. 1–272; E. C. Friedberg, G. C. Walker and W.
Siede, DNA Repair and Mutagenesis, ASM Press, Washington DC,
1995; J.-S. Taylor, Pure Appl. Chem., 1995, 67, 183; J. Cadet, M. Berger,
T. Douki, B. Morin, S. Raoul, J.-L. Ravanat and S. Spinelli, J. Biol.
Chem., 1997, 378, 1275.
NMR (300 MHz, [D6]DMSO, TMS): d = 1.62 (s, 3H, CH3), 3.09
(m, 2H, CHCH2), 3.53 (s, 3H, OCH3), 3.85 (m, 2H, NCH2), 4.12
(d, J = 6.5 Hz, 1H, NCH), 4.49 (m, 1H, NHCH), 5.65 (d, J =
6.5 Hz, 1H, OCH), 6.91–7.09 (m, 3H, Hindole), 7.32–7.46 (m, 7H,
Hindole + benzene), 8.44 (d, 1H, NHCH), 10.87 (s, 1H, CNHC); 13C
NMR (75 MHz, [D6]DMSO): d = 22.4 (CH3), 27.0 (CHCH2),
48.8 (NCH2), 51.7 (OCH3), 53.2 (NHCH), 64.4 (NCH), 77.0
(CCH3), 85.2 (OCH), 109.1, 111.3, 117.9, 118.4, 120.9, 123.5,
2 A. Sancar, Chem. Rev., 2003, 103, 2203.
3 P. E. M. Gibbs, B. J. Kilbey, S. K. Banerjee and C. W. Lawrence,
J. Bacteriol., 1993, 175, 2607; J.-S. Taylor, Acc. Chem. Res., 1994, 27,
76.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 291–298 | 297
©

View Article Online
4 R. O. Rahn and J. L. Hosszu, Photochem. Photobiol., 1969, 10, 131.
5 (a) S.-T. Kim, K. Malhotra, C. A. Smith, J.-S. Taylor and A. Sancar,
J. Biol. Chem., 1994, 269, 8535; (b) A. Sancar, Science, 1996, 272, 48;
(c) T. Todo, H. Ryo, K. Yamamoto, H. Toh, T. Inui, H. Ayaki, T.
Nomura and M. Ikenaga, Science, 1996, 272, 109; (d) X. D. Zhao, J. Q.
Liu, D. S. Hsu, S. Y. Zhao, J.-S. Taylor and A. Sancar, J. Biol. Chem.,
1997, 272, 32580.
6 (a) A. Joseph and D. E. Falvey, J. Am. Chem. Soc., 1995, 117, 11375;
(b) A. Joseph, G. Prakash and D. E. Falvey, J. Am. Chem. Soc., 2000,
122, 11219; (c) A. Joseph and D. E. Falvey, Photochem. Photobiol. Sci.,
2002, 1, 632.
13 (a) S.-T. Kim, R. F. Hartman and S. D. Rose, Photochem. Photobiol.,
1990, 52, 789; (b) S.-T. Kim and S. D. Rose, J. Photochem. Photobiol.,
B, 1992, 12, 179; (c) D. G. Hartzfeld and S. D. Rose, J. Am. Chem. Soc.,
1993, 115, 850.
14 (a) R. Epple, E.-U. Wallenborn and T. Carell, J. Am. Chem. Soc., 1997,
119, 7440; (b) J. Bulenandt, R. Epple, E.-U. Wallenborn, A. P. M. Eker,
V. Gramlich and T. Carell, Chem.–Eur. J., 2000, 6, 62.
15 Q. H. Song, W. J. Tang, X. M. Hei, H. B. Wang, Q. X. Guo and S. Q.
Yu, Eur. J. Org. Chem., 2005, 1097.
16 A. S. Jones, P. Lewis and S. F. Withers, Tetrahedron, 1973, 29, 2293.
17 S. Yamada, T. Shioiri, T. Itaya, T. Hara and R. Matsueda, Chem. Pharm.
Bull., 1965, 13, 88; M. Cain, O. Campos, F. Guzman and J. M. Cook,
J. Am. Chem. Soc., 1983, 105, 907.
7 M. K. Cichon, S. Arnold and T. Carell, Angew. Chem., Int. Ed., 2002,
41, 767.
8 Q. H. Song, X. M. Hei, Z. X. Xu, X. Zhang and Q. X. Guo, Bioorg.
Chem., 2003, 31, 357.
9 K. Hitomi, S.-T. Kim, S. Iwai, N. Harima, E. Otoshi, M. Ikenaga and
T. Todo, J. Biol. Chem., 1997, 272, 32591.
18 M. R. DeFelippis, C. P. Murthy, F. Broitman, D. Weinraub, M. Faraggi
and M. H. Klapper, J. Phys. Chem., 1991, 95, 3416.
19 S. L. Murov, I. Carmichael and G. L. Hug, Handbook of Photochem-
istry, Marcel Dekker Inc., New York, 2nd edn, 1993.
10 H.-W. Park, S.-T. Kim, A. Sancar and J. Deisenhofer, Science, 1995,
268, 1866.
11 H. Komori, R. Masui, S. Kuramitsu, S. Yokoyama, T. Shibata, Y. Inoue
and K. Miki, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 13560.
12 A. Mees, T. Klar, P. Gnau, U. Hennecke, A. P. M. Eker, T. Carell and
L.-O. Essen, Science, 2004, 306, 1789.
20 L. I. Grossweiner, A. G. Kepka, R. Santus and J. A. Vigil, Int. J. Radiat.
Biol., 1974, 25, 521; S. R. Yeh and D. E. Falvey, J. Am. Chem. Soc.,
1991, 113, 8557.
21 R. F. Anderson, Biochim. Biophys. Acta, 1983, 722, 158.
22 S.-T. Kim, M. Volk, G. Rousseau, P. F. Heelis, A. Sancar and M.-E.
Michel-Beyerle, J. Am. Chem. Soc., 1994, 116, 3115.
298 | Org. Biomol. Chem., 2006, 4, 291–298
This journal is
The Royal Society of Chemistry 2006
©