Photochemical generation of oligodeoxynucleotide containing a C4′-oxidized abasie site and its efficient amine modification: Dependence on structure and microenvironment

Article abstract of DOI:10.1021/jo702080r

(Chemical Equation Presented) Bleomycin-induced oxidative DNA damage under limited oxygen conditions results in the formation of the C4′-oxidized abasic site (1). We synthesized the oligodeoxynucleotides (ODN) 5, which contains 4′-o-nitrobenzyloxythymidine (3), and 6, which contains 2-nitrobenzyloxy-4′-methoxy-2′-deoxy-D-ribofuranoside (4) as the caged precursors of 7, an ODN containing 1, to study its reactivity with amines Photoirradiation of the single- and double-stranded 5 led to the formation of 7. Uncaging of the duplex was faster and the yield of 7 was higher with the double-stranded than with the single-stranded ODN. It was suggested that a low dielectric environment of the o-nitrobenzyloxy group in the minor groove of the duplex might accelerate the uncaging rate. Similarly, 6 and its duplex yielded 7 by photoirradiation However, the yields of 7 were lower than those of 5, and duplex formation slowed the uncaging rate. Reaction of the obtained 7 with an amine resulted in the formation of the lactam 2b in good yield in both single- and double-stranded forms, showing that amine modification of biomolecules by an ODN containing 1 is possible under physiologic conditions.

Full text of DOI:10.1021/jo702080r

Photochemical Generation of Oligodeoxynucleotide Containing a
C4′-Oxidized Abasic Site and Its Efficient Amine Modification:
Dependence on Structure and Microenvironment
Kazuteru Usui, Mariko Aso,* Mitsuhiro Fukuda, and Hiroshi Suemune*
Graduate School of Pharmaceutical Sciences, Kyushu UniVersity, Maidashi, Higashi-ku,
Fukuoka 812-8582, Japan
aso@phar.kyushu-u.ac.jp
ReceiVed September 22, 2007
Bleomycin-induced oxidative DNA damage under limited oxygen conditions results in the formation of
the C4′-oxidized abasic site (1). We synthesized the oligodeoxynucleotides (ODN) 5, which contains
4′-o-nitrobenzyloxythymidine (3), and 6, which contains 2-nitrobenzyloxy-4′-methoxy-2′-deoxy-D-
ribofuranoside (4), as the caged precursors of 7, an ODN containing 1, to study its reactivity with amines.
Photoirradiation of the single- and double-stranded 5 led to the formation of 7. Uncaging of the duplex
was faster and the yield of 7 was higher with the double-stranded than with the single-stranded ODN. It
was suggested that a low dielectric environment of the o-nitrobenzyloxy group in the minor groove of
the duplex might accelerate the uncaging rate. Similarly, 6 and its duplex yielded 7 by photoirradiation.
However, the yields of 7 were lower than those of 5, and duplex formation slowed the uncaging rate.
Reaction of the obtained 7 with an amine resulted in the formation of the lactam 2b in good yield in both
single- and double-stranded forms, showing that amine modification of biomolecules by an ODN containing
1 is possible under physiologic conditions.
Introduction
the reaction of an oligodeoxynucleotide (ODN) containing 1
with an amine under mild conditions (room temperature, pH
7).3,4 This indicates that the ODN containing 1 provides a
reactive keto aldehyde moiety and possibly modifies the amine-
containing biomolecules that interact with DNA. Lactam forma-
tion from ODN containing 1 proceeds without extra reagents,
such as NaBH4, as in the lysine modification via a Schiff base.
Bleomycin-induced oxidative DNA damage under limited
oxygen conditions results in the formation of the alkali-labile
C4′-oxidized abasic site (1) via the C4′ hydrogen abstraction
of 2′-deoxyribose and the subsequent oxidation of the resulting
C4′ radical to the C4′ cation (Scheme 1).1 At equilibrium, the
1,4-dihydroxytetrahydrofuran structure of 1 can exist in the open
form with a 1,4-dicarbonyl structure. Reactive aldehydes,
especially bifunctional ones, can be toxic as a result of their
reactivity with a DNA base and a protein.2 We found that an
unsaturated lactam (2a) and DNA fragments were formed in
(2) (a)Nechev, L. V.; Harris, C. M.; Harris, T. M. Chem. Res. Toxicol.
2000, 13, 431-435. (b) Xu, G.; Liu, Y.; Kansal, M. M.; Sayer, L. M. Chem.
Res. Toxicol. 1999, 12, 855-861. (c) Chen, L.-J.; Hecht, S. S.; Peterson,
L. A. Chem. Res. Toxicol. 1997, 10, 866-874.
(3) Aso, M.; Kondo, M.; Suemune, H.; Hecht, S. M. J. Am. Chem. Soc.
1999, 121, 9023-9033.
(4) Aso, M.; Ryuo, K.; Kondo, M.; Suemune, H. Chem. Pharm. Bull.
2000, 48, 1384-1386.
(1) Sugiyama, H.; Kilkuskie, R. E.; Hecht, S. M.; van der Marel, G. A.;
van Boom, J. H. J. Am. Chem. Soc. 1985, 107, 7765-7767.
10.1021/jo702080r CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/07/2007
J. Org. Chem. 2008, 73, 241-248
241

Usui et al.
SCHEME 1. Formation of 1 and Its Reaction with Amine
Forming Lactam
the reactivity of an ODN containing 1 with amine. We believe
that the use of our caged nucleoside will be advantageous for
site-specific lysine modification of a targeted protein, because
irradiation of a caged ODN-protein complex was possible, in
which the interaction of the retained base moiety and the target
protein can be maintained. Thus, we prepared the ODN 5
containing 4′-o-nitrobenzyloxythymidine (3) (Chart 1). We also
designed 6, in which a caged sugar (4) is introduced, to compare
its uncaging efficiency with that of 5. Some caged precursors
of DNA and RNA containing an abasic site have an o-
nitrobenzyloxy group at the C1 position of the corresponding
sugars, and their photolysis results in the formation of abasic
sites containing oligonucleotide.6 The introduction of an o-
nitrobenzyloxy group at the C1 position of deoxyribose together
with a methoxy group at the C4 position appears to be another
promising design for the precursor of 1. An o-nitrobenzyloxy
ketal structure in 3 and an o-nitrobenzyloxy acetal structure in
4 may show different efficiency in the photolysis of 5 and 6. In
addition to the difference in the chemical structure, duplex
formation may place the o-nitrobenzyloxy group of 5 and 6 into
different environments, i.e., the o-nitrobenzyloxy group in the
minor groove in 5 and that between the stacked base pairs in 6,
thus affecting their photoreactivity. The sequence of 8, the
unmodified parent ODN of 5 and 6, contains the binding site
of the DnaA protein, which is involved in the initiation of
replication in Escherichia coli (Chart 1). Recently, the X-ray
crystal structure of a complex of the DNA binding domain
(domain IV) of the DnaA protein and the duplex 8 has been
reported.7 The results of X-ray crystallography suggest that one
of the lysine residues in domain IV (Lys-415) is located near
the phosphate group between C2 and T3. The open form of 7
may react with Lys-415 to modify it if the lysine residue comes
near the carbonyl functions during molecular vibration of the
protein conformation in solution. In this report, we describe the
synthesis of 6 using conventional phosphoramidite chemistry
and studies on the uncaging reactions of single- and double-
stranded forms of 5 and 6, along with the detailed kinetics. The
choice of a caged monomer and the structure of ODN affected
the efficiency of the uncaging reaction, and the duplex 5 yielded
7 most efficiently in this study. Subsequent amine treatment of
7 resulted in smooth lactam formation.
CHART 1. Caged Nucleoside 3, Caged Sugar 4, and ODNs
Containing DnaA Binding Site
In situ modification and mapping of the lysine residues in the
reaction of ODN containing 1 may be applicable in the
structure-function studies of DNA-binding proteins.
Results and Discussion
For such studies, the efficient synthesis of 1 in ODN and
quantitative studies of its lactam formation are necessary. Due
to its instability, especially under alkaline conditions, we planned
to synthesize ODN carrying a stable caged precursor that yields
1 by photolysis. Recently, Greenberg et al. have reported the
generation of 1 by photochemical cleavage of caged sugars,
which carry 3,4-dimethoxy-6-nitrobenzyloxy groups at the C1
and C4 positions of the 2′-deoxyribose.5 The 5′-O-silylated
O-methyl phosphoramidites prepared from the caged sugars
were incorporated into an ODN. Greenberg et al. then studied
the chemical stability and biological effects of ODN containing
1.5 In contrast, we designed a caged nucleoside, compound 3
(Chart 1), which retains the base moiety at the C1′position and
has an o-nitrobenzyloxy group at the C4′position to investigate
Synthesis of 6. The caged sugar 4 was synthesized using
essentially the same methods as for the synthesis of 38 via the
4,5-unsaturated 16 (Scheme 2). Koenigs-Knorr-type glycosi-
dation of 1-chloro-3,5-di-p-chlorobenzoyl-D-ribofuranose9 with
o-nitrobenzylalcohol in the presence of AgOTf gave the desired
â-anomer 13a (28%) and R-anomer 13b (47%). The use of
ZnCl2 as a catalyst increased the ratio of 13a/13b to 1.3/1 (69%).
Treatment of 13b with TMSOTf in CH3CN resulted in partial
epimerization at the C1 position to give 13a/13b in the ratio of
1/1.5 (68%). The stereochemistry of 13a was determined by
(6) (a) Peoc’h, D.; Meyer, A.; Imbach, J. -L.; Rayner, B. Tetrahedron
Lett. 1991, 32, 207-210. (b) Trzupek, J. D.; Sheppard, T. L. Org. Lett.
2005, 7, 1493-1496.
(7) Fujikawa, N.; Kurumizaka, H.; Nureki, O.; Terada, T.; Shirouzu, M.;
Katayama, T.; Yokoyama, S. Nucleic Acids Res. 2003, 31 (8), 2077-2086.
(8) Aso, M.; Usui, K.; Fukuda, M.; Kakihara, Y.; Goromaru, T.;
Suemune, H. Org. Lett. 2006, 8, 3183-3186.
(9) Ishido, R.; Kyomoro, H.; Nagase, T.; T katsuki, K.; Nakajima, C.;
Ooshida, H. Jpn. Kokai Tokkyo Koho JP 07224081 A 19950822 Heisei,
1995.
(5) (a) Kim, J.; Gil, J. M.; Greenberg, M. M. Angew. Chem., Int. Ed.
2003, 42, 5882-5885. (b) Greenberg, M. M.; Weledji, Y. N.; Kroeger, K.
M.; Kim, J.; Goodman, M. F. Biochemistry 2004, 43, 2656-2666. (c)
Greenberg, M. M.; Weledji, Y. N.; Kim, J.; Bales, B. C. Biochemistry 2004,
43, 8178-8183. (d) Kroeger, K. M.; Kim, J.; Goodman, M. F.; Greenberg,
M. M. Biochemistry 2004, 43, 13621-13629.
242 J. Org. Chem., Vol. 73, No. 1, 2008

Oligodeoxynucleotide with a C4′-Oxidized Abasic Site
SCHEME 2. Synthesis of Phosphoramidite 18 for ODN 6
SCHEME 3. Photolysis of 5 and 6 To Give 7
open dehydrated form (3849; calculated for 7 - H2O: 3847),
and sodium-added form (3888; calculated for 7 + Na: 3888)
(Supporting Information).
Studies of the decaging kinetics showed that the decay of 5
followed first-order kinetics in both water and phosphate buffer
with comparable rate constants of k ) 1.5 × 10-1 min-1 and
t1/2 ) 4.6 min (Figure 2a and Table 1). The yield of ODN 7 at
30 min was estimated to be about 70% based on a comparison
of the peak areas of 7 and 8 (Figure 2c).10 For preparation of
the duplex containing 1, Greenberg et al. hybridized an ODN
containing 1 under mild conditions (37 °C) because of its
instability at high temperature (above 55 °C).5a The direct
formation of the duplex containing 1 by photoirradiation of the
duplex caged ODN would be convenient if the reaction
proceeded efficiently.
Therefore, the photoreaction of the duplex 5 was studied next.
Prior to the photoreaction, UV melting studies were carried out
to determine the stability of the duplex 5. The Tm value indicated
that the duplex of 5 with 9 was stable (41.5 and 46.4 °C in
H2O and phosphate buffer, respectively, Table 2, entry 5) under
the photoreaction conditions employed (25 °C), although the
stability was slightly less than that of the unmodified 8:9 (Table
2, entry 1). The duplexes of 5 with the ODN containing base
other than adenine at Y (entries 6-8) were less stable than 5:9
(entry 5), indicating that the thymine in 3 formed a base pair
with the opposite adenine in the duplex. Detmer et al. studied
the impact of the C4′-alkylation of thymidine on the duplex
stability.11 The introduction of one modified thymine residue
with a methyl, ethyl, and isopropyl group at the C4′ position
into the duplex had little effect on its thermal stability and
conformation. The duplex 5 was not destabilized significantly,
and its CD spectrum was consistent with a B-type helix
(Supporting Information). Interestingly, the duplexes of 5 with
a mismatch were slightly more stable compared with the
unmodified duplexes containing the same mismatch [Table 2,
entries 6-8, ∆Tm′, Tm (8:10, 11, 12) - Tm (5:10, 11, 12)]. It is
known that a minor groove binder conjugated to DNA stabilizes
the X-ray analysis of 14, which was further converted into the
intermediate 16 via 15. Treatment of 16 with m-CPBA in MeOH
and subsequent solvolysis of the acetate group produced an
inseparable mixture of 4 and its C4 epimer (1:1). Protection of
the 5′-hydroxyl group as a DMTr ether facilitated the chro-
matographic separation of the diastereomers, and 17 was
obtained in 25% yield from 16. The phosphoramidite 18
prepared from 17 was incorporated into the 13-mer using the
conventional phosphoramidite method with an automated DNA
synthesizer. ODN synthesized in the trityl on mode was purified
by HPLC, and the DMTr group of the ODN was removed under
acidic conditions (15% AcOH, 0 °C, 3 h) to give 6 (Supporting
Information). The structure of 6 was confirmed by MALDI-
TOF MS ((m/z) 4016 (calculated for 6: 4015)).
Photoreaction of 5. The photoreaction of 58 was studied in
a 10 µM solution in water or phosphate buffer (10 mM, pH 7),
both containing 100 mM NaCl and 8 (3 µM), which was added
to the reaction mixture as an internal standard (Scheme 3, Figure
1a). HPLC analysis of the photoirradiated 5 (retention time at
26 min) at 365 nm indicated its smooth conversion to a new
product (19 min) in 30 min (the peak at 20 min is the internal
standard 8). The liberation of thymine (6 min) was also
confirmed by co-injection of the reaction mixture and thymine,
suggesting that the uncaging reaction occurred (Supporting
Information). The peak at 19 min was assigned to 7 on the basis
its conversion to the lactam as described in the last paragraph.8
The formation of 7 was also confirmed by analysis of the
MALDI-TOF MS spectrum of the photolyzed sample, which
showed peaks assigned to 7 (3866; calculated for 7: 3865), its
(10) HPLC analysis of the reaction mixture suggetsed that the peak areas
of 8 and 9, and 7-mer, 5′-CCT-GTG-G-3′, the internal standards, were
not changed during reactions.
(11) Detmer, I.; Summerer, D.; Marx, A. Chem. Commun. 2002, 2314-
2315.
J. Org. Chem, Vol. 73, No. 1, 2008 243

Usui et al.
FIGURE 2. Kinetics of the decaging of (a) 5 and (b) 6 in the single-
stranded (O) and duplex forms (b) in phosphate buffer. Plots of the ln
substrate fractions versus time show k values of 1.5 × 10-1 and 2.9 ×
10-1 min-1 for 5 and its duplex and 1.8 × 10-1 and 1.0 × 10-1 min-1
for 6 and its duplex, respectively. Yields of 7 formed by the uncaging
of (c) 5 and (d) 6 in the single-stranded (O) and duplex forms (b) in
phosphate buffer were calculated from peak areas of 7 based on a
comparison of the corresponding internal standard.
TABLE 1. Summary of Parameters of the Photoreaction of 5 and
6 in Phosphate Buffer.
5
6
ssa
dsb
ssa
dsb
k (min-1
krel
)
1.5 × 10-1
2.9 × 10-1
1.8 × 10-1
1.0 × 10-1
c
1
1.9
1.2
0.7
FIGURE 1. HPLC analysis of the photoirradiation of (a) 5 (retention
time at 26 min, with 8 as an internal standard at 20 min) and (b) the
duplex 5 (retention time at 26 min, with 9 at 24 min) at different time
intervals in phosphate buffer. The ODN 7 and thymine eluted at 19
and 6 min, respectively.
t
1/2 (min)
4.6
72 ( 1
2.4
82 ( 1
3.9
52 ( 3
6.9
60 ( 4
7 (%)
a Single stranded. b Double-stranded. c k/k5ss
TABLE 2. Tm Values of Duplexesa
duplexes, and the van der Waals interaction can be a factor in
stabilizing conjugated duplexes.12 The o-nitrobenzyloxy group
of 3 did not make a contribution to the stabilization of the full-
match duplex but to the duplex containing the mismatch. Upon
irradiation in water, the duplex 5:9 was photolyzed by first-
order kinetics with a rate constant of k ) 2.2 × 10-1 min-1
and t1/2 of 3.1 min. The decay of 5 and formation of 7 were
faster in the duplex than in the single-stranded form. In
phosphate buffer, the duplex 5:9 decayed even faster than in
d(5′-CCX-GTG-GAT-AAC-A-3′)
d(3′-GGY-CAC-CTA-TTG-T-5′)
Tm, (°C)
∆Tm (°C)d
∆Tm′ (°C)e
entry duplex (X:Y) H2Ob PBc
H2Ob
PBc
H2Ob
PBc
1
2
3
4
5
6
7
8
9
8:9 (T:A)
8:10 (T:G)
8:11 (T:T)
8:12 (T:C)
5:9 (3:A)
5:10 (3:G)
5:11 (3:T)
5:12 (3:C)
6:9 (4:A)
42.4 47.7
36.8 41.7
34.7 39.2
32.6 33.2
41.5 46.4
37.0 43.1
35.5 41.5
33.0 34.7
-5.6
-7.7
-6.0
-8.5
-9.8 -14.5
-0.9
-5.4
-6.9
-1.3 -0.9 -1.3
water with a rate constant of k ) 2.9 × 10-1 min-1 and t1/2
)
-4.6
-6.2
0.2
0.8
0.4
1.4
2.3
1.5
2.4 min, about 2-fold faster than the single-stranded form
(Figures 1b and 2a and Table 1). The yield of 7 from the duplex
5:9 was also higher in both water and phosphate buffer (82%
at 30 min based on the peak area of 910) than that from 5 (Figure
2c and Table 1).
-9.4 -13.0
27.0 37.7 -15.4 -10.0
a Data were obtained from the average of three individual experiments.
Measurements were conducted with 3 µM duplex DNA in water or buffer,
both containing 0.1 M NaCl. Melting temperatures were obtained from the
maximum of the first derivative of the melting curve (A260 vs temperature).
b Measurements conducted in water. c Measurements condoucted in 10 mM
phosphate buffer (pH 7). d ∆Tm ) Tm - Tm (8:9). e ∆Tm′ ) Tm (8:9, 10,
11, 12) - Tm (5:9, 10, 11, 12).
Acceleration of the uncaging reaction of 5 was observed upon
duplex formation. In phosphate buffer, in which the Tm value
was higher than in water, the decaging reaction of the duplex
was faster than in water and the more stable duplex formation
resulted in a faster reaction. The o-nitrobenzyloxy group at the
C4′ position of 2′-deoxyribose may be located in the minor
groove of the duplex. We therefore focused on the microenvi-
ronment of the minor groove of the DNA duplex. It was reported
that the minor groove of the B-form DNA duplex is quite
nonpolar and exhibits a local dielectric constant of ca. 20 D,
which corresponds to that of 60% 1,4-dioxane in water.13 In
order to study the effects of the dielectric environment of the
minor groove on acceleration of the decaging reaction of the
(12) Kumar, S.; Reed, M. W.; Gamper, H. B., Jr.; Gorn, V. V.;
Lukhtanov, E. A.; Foti, M.; West, J.; Meyer, R. B., Jr.; Schweitzer, B. I.
Nucleic Acids Res. 1998, 26, 831-838.
244 J. Org. Chem., Vol. 73, No. 1, 2008

Oligodeoxynucleotide with a C4′-Oxidized Abasic Site
duplex 5, photoreaction of single-stranded 5 was conducted in
60% 1,4-dioxane in phosphate buffer. Interestingly, the uncaging
reaction proceeded with efficiency comparable to that of the
duplex in phosphate buffer (k ) 2.4 × 10-1 min-1, yield of 7
) 88%, Supporting Information). This result clearly indicates
that a low dielectric environment in the minor groove accelerates
the uncaging reaction of the duplex. Solvent dependence of the
photo cleavage of the o-nitrobenzyl linker was reported,14 and
photolysis of the veratryl-based linker was faster in 1,4-dioxane
than in aqueous solvents. The UV spectra of 3 were recorded
in phosphate buffer and buffer containing 60% dioxane. The
absorbance at 365 nm due to n-π* excitation did not show any
difference in the two buffers (data not shown). During photolysis
of the o-nitrobenzyl ether, the reaction rates of the aci-nitro
intermediate is also solvent dependent,15 and further studies are
required to determine which steps in the uncaging process are
affected by the dielectric environment during acceleration of
the uncaging of 5.
Photoreaction of 6. The photoreactions of both the single-
and double-stranded forms of 6 were studied in phosphate buffer
(Scheme 3). First, irradiation of 6 at 365 nm led to its decay
(Figure 3a). After 30 min of irradiation, the peak of 6 (retention
time at 40 min) completely disappeared and the main peak
appeared at 29 min16 along with several minor peaks. When
the photolyzed sample was analyzed 1 h after irradiation, a small
peak at 30 min disappeared and the area of the main peak
increased. The peak at 30 min was possibly an intermediate of
the uncaging reaction, although its structure was not identified.
For this reason, the photoirradiated samples were left to stand
for 60 min before analysis. MALDI-TOF MS measurement of
the photolyzed 6 showed a peak at the m/z values of 3865,
suggesting the formation of 7 (Supporting Information). Next,
the photolyzed 6 and 5 were coinjected (Figure 3b). The main
peak of the photolyzed 6 coeluted with 7 yielded by the
photolysis of 5, indicating that they are identical.
Kinetic studies of the uncaging of 6 were carried out based
on HPLC analysis. The photolysis of 6 proceeded even faster
than that of 5 (k ) 1.8 × 10-1 min-1, t1/2 ) 3.9 min, Figure 2b
and Table 1). The conversion of 6 to 7 appeared to be efficient
based on the results of HPLC analysis of the reaction mixture
(Figure 4a).6 However, the yield of 7 based on a comparison of
its peak area with an internal standard (7-mer, 5′-CCT-GTG-
G-3′,10 19 min) was lower than expected (52%), probably due
to formation of byproducts (Figure 2d and Table 1). Next, the
photoreaction of the duplex 6 was examined 60 min after
photoirradiation (Figure 4b). Photoreaction of the duplex 6
afforded 7 in 60% yield at 30 min (94% conversion). The CD
spectrum of the duplex 6 was consistent with a B-type helix
(Supporting Information), although the Tm value of 6 at 38 °C
in phosphate buffer was lower when compared with that of the
unmodified 8 and 5 (Table 2, entries 9, 1, and 5). Duplex
formation of 6 under photoreaction conditions also increased
the yield of 7 (60%, Figure 2d and Table 1). In contrast to the
photoreaction of 5, duplex formation decelerated the uncaging
reaction of 6, and the kinetic parameters indicated that the
reaction was 0.6-fold slower than that of the single-stranded 6
FIGURE 3. HPLC analysis after photoirradiation of 6. (a) Analysis
of 6 1 h after 30-min irradiation. (b) Co-injection of photolyzed 5 and
6.
(k ) 1.0 × 10-1 min-1, t1/2 ) 6.9 min, Figure 2b and Table 1).
Thus, the formation of 7 by photoirradiation of 6 proceeded,
but the yields of 7 were lower than with 5 in both the single-
and double-stranded forms due to the formation of byproducts.
When irradiated, the aromatic nitro group can abstract a
proximal hydrogen from a saturated carbon. Kotera reported
the efficient synthesis of the 2′-deoxyribonolactone lesion by
photoirradiation of ODN containing the 7-nitroindole nucleo-
side.17 The oxygen atom of the nitro group in 7-nitroindole is
sufficiently close to abstract the H-1′ atom (2.4 Å), and the
photoirradiation resulted in the generation of a C1′ radical that
eventually formed 2′-deoxyribonolactone. In the caged sugar
4, the H-1′ atom is present sufficiently close to the nitro group
to be abstracted, and its abstraction may compete with that of
the benzyl proton for the desired uncaging reaction and result
in lower yields (Supporting Information). In the duplex, the
nitrobenzyl group of 6 may be located between the stacking
base pairs. Deceleration of the uncaging rate in the duplex 6
might be attributed to quenching of the photoreaction of the
o-nitrobenzyl chromophore by the stacking base residues.
Lactam Formation from 5. Although the use of lysine
appeared to be appropriate as a model for modification of lysine
residues of proteins, preliminary results showed that the reaction
of lysine can give a mixture of R- and ꢀ-lactams under conditions
similar to those in the formation of 2b. 3-Amino-1,2-propanediol
and 2b are both water-soluble, which is appropriate for the
reaction and analysis. Thus, the reaction mixture of photolyzed
(13) Jin, R.; Breslauer, K. J. Proc. Natl. Acad. Sci. U.S.A. 1988, 85,
8938-8942.
(14) Holmes, C. P. J. Org. Chem. 1997, 62, 2370-2380.
(15) Schwo¨rer, M.; Wirz, J. HelV. Chim. Acta 2001, 84, 1441-1458.
(16) In Figures 3 and 4, the retention times of 5 and 7 were different
from those in Figure 1 because the photolyzed 5 and 6 were analyzed under
different HPLC conditions for better peak resolution (Experimental Section).
(17) Kotera, M.; Bourdat, A.-G.; Defrancq, E.; Lhomme, J. J. Am. Chem.
Soc. 1998, 120, 11810-11811.
J. Org. Chem, Vol. 73, No. 1, 2008 245

Usui et al.
FIGURE 5. (a) HPLC analysis of the reaction mixture of the
photolyzed duplex 5 and 3-amino-1,2-propanediol. The latam 2b, the
fragment 10-mer, 7, and 9 eluted at 11, 16, 18, and 23 min, respectively.
(b) Kinetics of the lactam formation from 7 in the single-stranded (O)
and duplex forms (b) in phosphate buffer. Plots of ln fraction substrate
present versus time give k ) 1.5 × 10-1 h-1 and t1/2 ) 4.6 h,
respectively. (c) Yields of the lactam 2b from 5 in the single stranded
(O) and duplex forms (b) in phosphate buffer.
FIGURE 4. HPLC analysis of the photoirradiation of (a) 6 (retention
time at 32 min, 7-mer as an internal standard and 7 eluted at 19 and 21
min, respectively) and (b) the duplex 6 (retention time at 38 min, 9
and 7 eluted at 35 and 26 min, respectively) at different time intervals.
HPLC conditions are described in Experimental Section. Yields of 7
were calculated based on (a) 7-mer and (b) 9.
areas with that of 9, were virtually equal (75% and 71%,
respectively, Figure 5c).
Conclusion
We synthesized the caged ODNs 5 and 6 using conventional
phosphoramidite chemistry and studied their photocleavage
reactions yielding ODN containing the C4′-oxidized abasic site
(1). The structure and environment where a caged group was
located affected the photoreactivity, and the direct photolysis
of the duplex 5 enabled efficient and convenient preparation of
the duplex containing 1. Lactam formation from 7 proceeded
efficiently with both the single- and double-stranded forms,
indicating the possibility of protein modification using ODNs
containing 1.
SCHEME 4. Lactam Formation from 7
5 was subjected to a reaction with an aqueous solution of
3-amino-1,2-propanediol (1000 equiv, pH adjusted to 7 with
AcOH in phosphate buffer). Heating the reaction mixture at
37 °C resulted in the conversion of 7 (18 min) to the lactam 2b
(11 min) and the fragment 10-mer (16 min) (Scheme 4 and
Figure 5a; the peak at 23 min is 9), the formations of which
were confirmed by co-injection of the reaction mixture with
authentic compounds and LC-MS for 2b (Supporting Informa-
tion).8 The efficiencies of lactam formation from 7 and its duplex
were comparable. The decrease in 7 followed first-order kinetics
(k ) 1.5 × 10-1 h-1 and t1/2 ) 4.6 h for 5 and duplex 5) (Figure
5b). The yields of lactam, determined by comparison of the peak
Experimental Section
2-Nitrobenzyloxy-3′-5′-O-(di-4-chlorobenzoyl)-2′-deoxy-D-ri-
bofuranoside (13a and 13b). Reaction with AgOTf. To a mixture
of 1-chloro-3,5-di-p-chlorobenzoyl-D-ribofuranose (6 g, 14.1 mmol),
o-nitrobenzyl alcohol (3.2 g, 21.1 mmol), and 3 Å molecular sieves
(7 g) in dry CH2Cl2 (200 mL) under an Ar atmosphere was added
AgOTf (10 g, 21.2 mmol) in dry THF (200 mL) dropwise during
45 min at -50 °C, and the mixture was stirred for an additional 5
h and then allowed to room temperature. After stirring for 12 h,
the mixture was filtered through Celite, which was washed with
CH2Cl2. The combined filtrates were washed with water and dried
over Na2SO4. The solvent was evaporated and the residue was
246 J. Org. Chem., Vol. 73, No. 1, 2008

Oligodeoxynucleotide with a C4′-Oxidized Abasic Site
purified by silica gel chromatography (ethyl acetate/n-hexane; 1:4)
to give 13a (2.2 g, 28%) and 13b (3.6 g, 47%) as oils, respectively.
2.26 (m, 1H), 2.08 (s, 3H); HRMS-FAB (m/z) calcd for C14H17-
INO6 422.0101 [M + H]+, found 422.0118.
1
13a: Rf ) 0.39 (ethyl acetate/n-hexane; 1:5) × 2; H NMR (400
2-Nitrobenzyloxy-3′-O-acetyl-4′,5′-methylene-2′-deoxy-D-ribo-
furanoside (16). Compound 15 (0.92 g, 2.2 mmol) was heated with
DBU (0.97 mL, 6.5 mmol) in benzene under reflux for 5 h. After
the reaction was quenched with saturated aqueous NH4Cl solution,
the mixture was evaporated. The residue was diluted with ethyl
acetate and washed with water and dried over Na2SO4. The solvent
was evaporated and the residue was purified by silica gel chroma-
MHz, CDCl3) δ 8.02 (d, J ) 8.2 Hz, 1H), 7.94 (m, 4H), 7.60 (m,
2H), 7.43 (m, 5H), 5.66 (m, 1H), 5.52 (m, 1H), 5.09 (d, J ) 14.8
Hz, 1H), 4.92 (d, J ) 14.6 Hz, 1H), 4.63-4.43 (m, 3H), 2.72-
2.42 (m, 2H); HRMS-FAB (m/z) calcd for C26H22Cl2NO8 546.0722
[M + H]+, found 546.0737. 13b: Rf ) 0.32 (ethyl acetate/n-hexane;
1
1:5) × 2; H NMR (400 MHz, CDCl3) δ 8.07-7.30 (m, 12H),
1
tography (CH2Cl2/n-hexane; 1:1) to give 16 (0.62 g, 97%): H NMR
5.48 (m, 2H), 5.22 (d, J ) 14.8 Hz, 1H), 4.95 (d, J ) 15.0 Hz,
1H), 4.65-4.51 (m, 3H), 2.58-2.36 (m, 2H); HRMS-FAB (m/z)
calcd for C26H21Cl2NNaO8 568.0542 [M + Na]+, found 568.0565.
Reaction with ZnCl2. To a mixture of 1-chloro-3,5-di-p-chlo-
robenzoyl-D-ribofuranose (0.5 g, 1.16 mmol), 2-nitrobenzyl alcohol
(0.27 g, 1.76 mmol), NaHCO3 (0.13 g, 1.55 mmol), and 4 Å
molecular sieves (0.5 g) in dry CH2Cl2 (15 mL) was added a
solution of 1 M ZnCl2-Et2O (1.8 mL) at -50 °C under an Ar
atmosphere, and the mixture was stirred overnight at -50 °C. The
reaction mixture was diluted with CH2Cl2, and the mixture was
washed with saturated aqueous NaHCO3 solution and brine and
dried over Na2SO4. The solvent was evaporated under reduced
pressure, and the residue was purified by silica gel chromatography
(ethyl acetate/n-hexane; 1:4) to give 13a/13b, 1.3:1 (69%). Epimer-
ization of 13b. To a cooled (-35 °C) solution of 13b (100 mg,
0.18 mmol) in CH3CN (2 mL) was added TMSOTf (49 µL, 0.27
mmol) under an Ar atmosphere, and the mixture was stirred
overnight at -35 °C. After the reaction was quenched with saturated
aqueous NaHCO3 solution, the mixture was evaporated. The solvent
was evaporated under reduced pressure, and the residue was purified
by silica gel chromatography (ethyl acetate/n-hexane; 1:4) to give
13a/13b, 1:1.5 (68%).
2-Nitrobenzyloxy-2′-deoxy-D-ribofuranoside (14). To a suspen-
sion of 13a (3.4 g, 6.2 mmol) in MeOH (70 mL) was added
dropwise a solution of NaOMe (10 mL of a 1.85 M solution in
MeOH, 18.5 mmol) at room temperature, and the mixture was
stirred 3 h. After the reaction was quenched with solid NH4Cl (2
g), the mixture was evaporated. The residue was diluted with
AcOEt, washed with water, and dried over Na2SO4. The solvent
was evaporated, and the residue was purified by silica gel
chromatography (ethyl acetate/n-hexane; 3:1) to give 14 (1.7 g,
quant): 1H NMR (400 MHz, CDCl3) δ 8.06 (d, J ) 8.2 Hz, 1H),
7.65 (m, 2H), 7.46 (t, J ) 7.7 Hz, 1H), 5.38 (d, J ) 5.4 Hz, 1H),
5.14 (d, J ) 14.8 Hz, 1H), 4.92 (d, J ) 14.4 Hz, 1H), 4.58 (m,
1H), 4.09 (m, 1H), 3.70 (m, 2H), 2.43-2.17 (m, 3H). The structure
of 14 was determined by X-ray crystallography (Supporting
Information).
2-Nitrobenzyloxy-3′-O-acetyl-5′-iodo-2′-deoxy-D-ribofurano-
side (15). To a cooled (0 °C) solution of 14 (0.98 g, 3.6 mmol) in
DMF (20 mL), under an Ar atmosphere, was added dropwise over
5 min a solution of CH3P(OPh)3I (4.4 mmol in 10 mL DMF), and
the mixture was stirred for an additional 4 h. After addition of
saturated aqueous Na2S2O3 solution, the reaction mixture was
extracted with Et2O. The organic layer was washed with water and
brine, dried over Na2SO4, and concentrated under reduced pressure.
The residue was purified by silica gel chromatography (ethyl
acetate/n-hexane; 1:4 to 1:1) to give crude product (1.5 g). Next, a
mixture of the obtained crude product (1.5 g), dry pyridine (0.4
mL, 5.5 mmol), DMAP (40 mg, 0.3 mmol), and acetic anhydride
(0.52 mL, 5.5 mmol) in CH2Cl2 (10 mL) was stirred at room
temperature for 1 h. After addition of water, the reaction mixture
was extracted with CH2Cl2. The organic layer was washed with
saturated aqueous NaHCO3 solution, dried over Na2SO4, and
concentrated under reduced pressure. The residue was purified by
silica gel column chromatography (ethyl acetate/n-hexane; 1:3) to
give 15 (0.92 g, 60%): 1H NMR (400 MHz, CDCl3) δ 8.05 (d, J
) 8.1 Hz, 1H), 7.67 (m, 2H), 7.46 (m, 1H), 5.42 (dd, J ) 5.4, 2.0
Hz, 1H), 5.28 (m, 1H), 5.18 (d, J ) 14.7 Hz, 1H), 4.94 (d, J )
14.7 Hz, 1H), 4.31-4.26 (m, 1H), 3.33 (m, 2H), 2.57 (m, 1H),
(400 MHz, CDCl3) δ 8.10 (d, J ) 8.1 Hz, 1H), 7.73 (d, J ) 7.8
Hz, 1H), 7.65 (t, J ) 7.6 Hz, 1H), 7.46 (m, 1H), 5.80 (m, 1H),
5.55 (m, 1H), 5.16 (d, J ) 14.9 Hz, 1H), 5.02 (d, J ) 14.9 Hz,
1H), 4.50 (s, 1H), 4.25 (s, 1H), 2.53 (m, 1H), 2.22 (m, 1H), 2.09-
2.11 (m, 3H); HRMS-FAB (m/z) calcd for C14H16NO6 294.0978
[M + H]+, found 294.0959.
2-Nitrobenzyloxy-4′-methoxy-5′-O-(4,4′-dimethoxytrityl)-2′-
deoxy-D-ribofuranoside (17) and C-4 Epimer of 17. To a solution
of 16 (2.6 g, 8.9 mmol) in dry MeOH (195 mL) under an Ar
atmosphere, was added dropwise a solution of m-CPBA (77%, 10
g, 44.6 mmol) in 72 mL of CH2Cl2 (predried with MgSO4) over
100 min. The reaction was stirred until consumption of the substrate
observed by TLC control (24 h). After addition of saturated aqueous
NaHCO3 solution until the pH has a value of 7, K2CO3 (3.7 g,
26.8 mmol) was added, and the mixture was stirred for 3 h. The
reaction mixture was concentrated under reduced pressure. The
residue was dissolved in AcOEt, washed with water, and dried over
Na2SO4. The solvent was evaporated and the residue was purified
by silica gel chromatography (ethyl acetate/n-hexane; 1:1 to ethyl
acetate) to give an inseparable mixture of 4 and its epimer (1.33
g). Next, a mixture of the obtained products (0.25 g, 0.82 mmol)
was coevaporated with anhydrous pyridine (twice) and dissolved
in dry pyridine (6.6 mL). DMTrCl (0.36 g, 1.07 mmol) was added
and the mixture was stirred overnight. The reaction mixture was
cooled to 0 °C, quenched by addition of water. The resulting
solution was extracted with ethyl acetate. The organic layer was
washed with water and brine, dried over Na2SO4, and concentrated
under reduced pressure. The residue was purified by silica gel
chromatography (ethyl acetate/n-hexane; 1:3 to 1:2) to give 17 (0.25
g, 25%) and C-4 epimer of 17 (0.25 g, 25%), respectively. 17: 1H
NMR (400 MHz, CDCl3) δ 8.03 (dd, J ) 8.1, 1.0 Hz, 1H), 7.67
(d, J ) 7.8 Hz, 1H), 7.54 (td, J ) 7.6, 1.0 Hz, 1H), 7.43-7.16 (m,
10H), 6.76 (m, 4H), 5.28 (d, J ) 5.1 Hz, 1H), 5.10 (d, J ) 15.1
Hz, 1H), 4.89 (d, J ) 15.1 Hz, 1H), 4.67 (dd, J) 9.4, 7.2 Hz, 1H),
3.76 (s, 6H), 3.23-3.21 (m, 5H), 2.37 (m, 1H), 2.13 (m, 1H);
HRMS-FAB (m/z) calcd for C34H35NO9 601.2312 [M]+, found
601.2335. C-4 epimer of 17: 1H NMR (400 MHz, CDCl3): δ 8.04
(d, J ) 7.6 Hz, 1H), 7.78 (d, J ) 7.8 Hz, 1H), 7.61 (t, J ) 7.3 Hz,
1H), 7.53-7.19 (m, 11H), 6.84 (m, 4H), 5.49 (t, J ) 5.4 Hz, 1H),
5.16 (d, J ) 15.4 Hz, 1H), 4.94 (d, J ) 15.1 Hz, 1H), 4.51 (d, J )
4.6 Hz, 1H), 3.76 (s, 6H), 3.58 (d, J ) 10.0 Hz, 1H), 3.18, (s, 3H),
3.09 (d, J ) 10.0 Hz, 1H) 2.43 (m, 1H), 2.25 (m, 1H); HRMS-
FAB (m/z) calcd for C34H35NO9 601.2312 [M]+, found 601.2286.
2-Nitrobenzyloxy-3′-O-(2-cyanoethyl N,N-diisopropylphos-
phoramisyl)-4′-methoxy-5′-O-(4,4′-dimethoxytrityl)-2′-deoxy-D-
ribofuranoside (18). Compound 17 (550 mg, 0.91 mmol) was
coevaporated with dry THF (3 times) and dissolved in dry CH2Cl2
(5 mL). 2-cyanoethyl N, N, N′, N′-tetraisopropylphosphorodiamidite
(0.76 mL, 2.39 mmol) was then added under an Ar atmosphere.
To the stirred solution was added dropwise a 0.9 M solution of
1H-tetrazole in CH3CN (1.2 mL, 1.09 mmol), and the mixture was
stirred 3 h at room temperature. The reaction mixture was diluted
with CH2Cl2 and the mixture was washed with saturated aqueous
NaHCO3 solution, brine and dried over Na2SO4. The solvent was
evaporated under reduced pressure and the residue was purified
by silica gel chromatography. Elution with ethyl acetate/n-hexane/
NEt3; 23.5:73.5:3 to give 18 (580 mg, 79%): 1H NMR (400 MHz,
CDCl3) δ 8.02 (d, J ) 8.1 Hz, 1H), 7.72 (dd, J ) 17.6, 7.7 Hz,
1H), 7.52 (dd, J ) 17.5, 8.1 Hz, 1H), 7.41 (m, 3H), 7.31 (m, 4H),
J. Org. Chem, Vol. 73, No. 1, 2008 247

Usui et al.
7.16 (m, 3H), 6.73 (m, 4H), 5.44 (dd, J ) 18.2, 5.2 Hz, 1H), 5.26
(m, 1H), 5.05 (d, 2H), 3.76 (s, 3H), 3.74 (s, 3H), 3.72-3.50 (m,
4H), 3.42 (dd, J ) 15.0, 9.7 Hz, 1H), 3.27 (s, 3H), 3.02 (m, 1H),
2.62 (m, 1H), 2.54-2.31 (m, 3H), 1.17 (m, 12H); HRMS-FAB (m/
z) calcd for C43H53N3O10P 802.3469 [M + H]+, found 802.3439.
Oligonucleotide Synthesis. The ODN 6 was synthesized using
standard solid-phase cyanoethyl phosphoramidite chemistry on a
DNA synthesizer. The oligomers were purified in DMT-on state
by HPLC on a reverse-phase COSMOSIL 5C18-AR-II column (20
mm i.d. × 250 mm) with a linear gradient of acetonitrile (5-40%
in 40 min) in 0.1 M triethylammonium acetate (pH 7). Purified
oligonucleotides were then detritylated by treatment with a solution
of 15% AcOH at 0 °C for 3 h to give 6. HPLC trace of detritylation
and MALDI-TOF MS spectrum of 6 are in Supporting Information.
UV-Melting Curve Measurements. The ODN 5, 6, or 8 was
hybridized with a 1:1 ratio (3 µM each) of 9 by heating in H2O or
10 mM phosphate buffer (pH 7), both containing 100 mM NaCl at
80 °C for 5 min and allowed to cool to room temperature slowly,
then cooled at 0 °C. Melting studies were carried out in 1 cm path
length quartz cells (under nitrogen atmosphere when temperature
was below 20 °C) on a UV-VISIBLE spectrometer equipped with
a thermoprogrammer and temperature controller. Absorbance was
monitored at 260 nm while temperature was changed at a rate of
0.5 °C min-1 from 0 to 80 °C. Melting temperatures were obtained
from the maximum of the first derivative of the melting curve versus
temperature. Hybridization of 5 and 6 with 10, 11, and 12 and Tm
measurements were carried out similarly.
Circular Dichroism (CD) Measurements. The circular dichro-
ism spectra were measured with a 0.1 cm path length quartz cell
using a spectropolarimeter between 340 and 220 nm in 10 mM
phosphate buffer (pH 7) containing 100 mM NaCl. DNA duplexes
were at 3 µM. Spectra were acquired every 1 nm with a bandwidth
setting of 1 nm at a speed of 50 nm min-1, averaging 10 scans.
CD spectra of the duplexes of 5:9 and 6:9 are in Supporting
Information.
Irradiation of Single- and Double-Stranded 5. Solutions of 5
(10 µM, containing 7 (3 µM)) as an internal standard and double-
stranded 5 (10 µM) were prepared in H2O (containing 100 mM
NaCl) or phosphate buffer (10 mM, pH 7, containing 100 mM
NaCl) and were purged with Ar (30 min). Aliquots of these
solutions (40 µL) were irradiated for 5, 10, 20, and 30 min,
respectively, in a quartz cuvette with a UV lamp (λmax ) 365 nm,
intensity 2.5 mW cm-2 at 364 nm) and were immediately analyzed
by HPLC using a COSMOSIL 5C18-AR-300 column (4.6 mm i.d.
× 250 mm). Solvent A: CH3CN. Solvent B: 50 mM ammonium
formate (pH 8). Gradient: initial, 96% B for 7 min, then linear to
14% A at 35 min. Detection was carried out at 260 nm. The yields
of 7 were the average of two experiments and determined by
comparison of its peak area with that of 8 or 9 based on their
extinction coefficients (105 M-1 cm-1): 7, 1.22; 8, 1.30; 9, 1.24.
HPLC analysis of the reaction mixture suggested that the peak areas
of 8 and 9 were not changed during the reactions. MALDI-TOF
MS spectrum of the photolyzed 5 is in Supporting Information.
Irradiation of Single-Stranded 5 in 60% 1,4-Dioxane. Solu-
tions of 5 (10 µM, containing 5-nat (3 µM)) as an internal standard
were prepared in 60% 1,4-dioxane in 10 mM phosphate buffer (pH
7) containing 100 mM NaCl and were purged with Ar (30 min).
Aliquots of these solutions (40 µL) were irradiated for 5, 10, and
20 min, respectively, in a quartz cuvette with a UV lamp (λmax
)
365 nm, intensity 2.5 mW cm-2 at 364 nm) and were evaporated
with a speedvac; the residues were dissolved in water (40 µL). A
volume of 10 µL of the sample was immediately analyzed by HPLC
similarly to the reactions in phosphate buffer. HPLC analysis of
the reaction mixture is in Supporting Information.
Irradiation of Single- and Double-Stranded 6. Solutions of 6
(10 µM, containing 7-mer, 5′-CCT-GTG-G-3′ (3µM), as an internal
standard and the double stranded 6 (10 µM) were prepared in
phosphate buffer (10 mM, pH 7, containing 100 mM NaCl) and
were purged with Ar (30 min). Aliquots of these solutions (40 µM)
were irradiated for 5, 10, 20, and 30 min, respectively, in a quartz
cuvette with a UV lamp (λmax ) 365 nm, intensity 2.5 mW cm-2
at 364 nm) and were analyzed by HPLC using a Waters XBridge
C
18 3.5 mm column (4.6 mm i.d. × 150 mm) for single stranded 6
and a COSMOSIL 5C18-AR-300 column (4.6 mm I.D. × 250 mm)
for the duplex 6. Solvent A: CH3CN. Solvent B: 50 mM
ammonium formate (pH 7.5-8). Gradient: initial, 96% B for 10
min, then linear to 10% A at 40 min. Detection was carried out at
260 nm. The yields of 7 were the average of two experiments and
determined by comparison of its peak area with that of 7-mer or 9
based on their extinction coefficients (105 M-1cm-1): 7-mer, 0.60.
MALDI-TOF MS spectrum of the photolyzed 6 is in Supporting
Information.
Acknowledgment. We thank Prof. Tsuyoshi Goromaru and
Ms. Yoshie Kakihara, Fukuyama University, for LC-MS
measurement. K.U. acknowledges support from JSPS research
fellowship.
Supporting Information Available: 1H NMR spectra of the
1
compounds 13a, 13b, 14, 15, 16, 17, and 18, H-1H COSY and
1H-1H NOESY spectra of 17, X-ray crystallographic data of 14
in CIF format, HPLC analyses of detrytilation of caged-ODN to
form 6, co-injection of the photolyzed 5 and thymine; irradiation
of 5 in 60% 1,4-dioxane, co-injection of the reaction mixture of
photolyzed 5 and 3-amino-1,2-propanediol with 2b and 10-mer,
MALDI-TOF MS of 6, photolyzed 5, and photolyzed 6, LC-MS
of the reaction mixture of photoirradiation of 5 followed by
treatment with 3-amino-1,2-propanediol and authentic 2b, circular
dichroism (CD) measurements, and ab initio calculation of 4. This
materialisavailablefreeofchargeviatheInternetathttp://pubs.acs.org.
JO702080R
248 J. Org. Chem., Vol. 73, No. 1, 2008