Reductive electron transfer in phenothiazine-modified DNA is dependent on the base sequence

Article abstract of DOI:10.1002/chem.200401013

A new DNA assay has been designed, prepared and applied for the chemical investigation of reductive electron transfer through the DNA. It consists of 5-(10-methyl-phenothiazin-3-yl)-2′-deoxyuridine (Ptz-dU, 1) as the photoexcitable electron injector and 5

Full text of DOI:10.1002/chem.200401013

FULL PAPER
Reductive Electron Transfer in Phenothiazine-Modified DNA Is Dependent
on the Base Sequence
Clemens Wagner and Hans-Achim Wagenknecht*[a]
Abstract: A new DNA assay has been
designed, prepared and applied for the
chemical investigation of reductive
electron transfer through the DNA. It
consists of 5-(10-methyl-phenothiazin-
3-yl)-2’-deoxyuridine (Ptz-dU, 1) as the
photoexcitable electron injector and 5-
bromo-2’-deoxyuridine (Br-dU) as the
electron trap. The Ptz-dU-modified oli-
gonucleotides were synthesised by
means of a Suzuki–Miyaura cross-cou-
pling protocol and subsequent auto-
mated phosphoramidite chemistry. Br-
dU represents a kinetic electron trap,
since it undergoes a chemical modifica-
tion after its one-electron reduction
that can be analysed by piperidine-in-
duced strand cleavage. The quantifica-
tion of the strand cleavage yields from
irradiation experiments reveals impor-
tant information about the electron-
transfer efficiency. The performed
DNA studies focused on the base se-
quence dependence of the electron-
transfer efficiency with respect to the
proposal that CCÀ and TCÀ act as inter-
mediate electron carriers during elec-
tron hopping. From our observations it
became evident that excess-electron
transfer is highly sequence dependent
and occurs more efficiently over T–A
base pairs than over C–G base pairs.
Keywords:
transfer
oligonucleotides · phenothiazine
DNA
·
electron
·
electron transport
·
Introduction
standing of hole-transfer processes. As a result, the mecha-
nisms of hole transfer have been simply transferred to the
mechanistic description of reductive ET in DNA.[1] Accord-
ingly, a hopping mechanism was proposed over long distan-
ces involving the pyrimidine radical ions CCÀ and TCÀ as in-
termediate electron carriers.[6]
Until five years ago, most knowledge about the behaviour
of excess electrons in DNA came from g-pulse radiolysis
studies with randomly DNA-bound electron traps and sug-
In principal, DNA-mediated charge-transfer processes can
be regarded as either oxidative hole transfer or reductive
electron transfer reactions.[1] Both processes are in fact elec-
tron-transfer (ET) reactions; however, the differentiation is
necessary with respect to the molecular orbitals that are in-
volved. Since the interest in these processes has been driven
by its relevance to oxidative damage, which causes mutagen-
esis and carcinogenesis,[2] the photochemically induced oxi-
dation of DNA has attracted enormous research efforts over
the last two decades.[3] Several mechanisms have been eluci-
dated, such as the superexchange and the hopping mecha-
nism, and at least in part, have been experimentally veri-
fied.[3] On the other hand, reductive ET processes are cur-
rently used in DNA chip technology[4] and DNA nanotech-
nology.[5] Injection of an extra electron into DNA initiates a
type of charge transfer that is complementary to hole trans-
fer and entails migration of this electron. This topic has re-
mained considerably underdeveloped relative to the under-
gesting
a thermally activated, electron-hopping process
above 150 K.[7] The major disadvantage of this experimental
setup is that the electron injection and trapping does not
occur site-selectively. Most of the more recent photochemi-
cal studies have been analysed by chemical means by using
two different kinetic electron traps, which are a specially de-
signed T–T dimer or 5-bromo-2’-deoxyuridine (Br-dU). By
using a DNA assay consisting of an artificial DNA base with
a flavine structure as the photoexcitable electron donor and
a special T–T dimer as the electron trap, Carell et al. could
show that the amount of T–T dimer cleavage depends
rather weakly on the distance to the flavine group, indicat-
ing an electron-hopping process.[8] Giese et al. investigated
ET in DNA using a special T derivative that undergoes Nor-
rish I cleavage after irradiation, yielding an electron injec-
tion process onto the thymine.[9] Remarkably, they could
show that a single injected electron can cleave more than
[a] C. Wagner, Priv.-Doz. Dr. H.-A. Wagenknecht
Technical University Munich, Chemistry Department
Lichtenbergstr. 4, 85747 Garching (Germany)
Fax : (+49)89-289-13210
E-mail: wagenknecht@ch.tum.de
Chem. Eur. J. 2005, 11, 1871 – 1876
DOI: 10.1002/chem.200401013
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1871

one T–T dimer. Rokita et al. analysed the ET efficiencies
using a diaminonaphthalene derivative as the photoexcitable
charge donor and Br-dU as the electron trap.[10] Br-dU indu-
ces a piperidine-dependent strand cleavage upon a one-elec-
tron reduction.[11] It is important to point out that a signifi-
cant dependence of the ET efficiency on the intervening
DNA base sequence was reported in the latter studies. This
observation stands clearly in contrast to the previously men-
tioned T–T dimer experiments and could reflect differences
of the kinetic behaviour of the two different electron traps.
Only Lewis et al.[12] and our group[13] have focused on the
study of the dynamics of ET processes using either DNA
hairpins that have been capped by a stilbene diether deriva-
tive or pyrene-modified DNA duplexes, respectively.
All of the recent studies support conclusively the proposal
of a thermally activated, electron-hopping mechanism over
long distances involving CCÀ and TCÀ as intermediate electron
carriers. However, recently, using pyrene-modified nucleo-
sides as model compounds, we showed that proton transfer
interferes with ET, indicating that CCÀ cannot play a major
role as an intermediate electron carrier in DNA.[14] Herein,
we want to follow these results and elucidate the sequence
dependence of DNA-mediated ET. We present the synthesis
of 5-(10-methyl-phenothiazin-3-yl)-2’-deoxyuridine (Ptz-dU,
1) as a photoexcitable electron injector and the correspond-
ing phenothiazine-modified oligonucleotides. Together with
5-bromo-2’-deoxyuridine (Br-dU) as the kinetic electron ac-
ceptor in irradiation experiments, we focused on the ET effi-
ciency with respect to CCÀ and TCÀ as intermediate electron
carriers during electron hopping (Scheme 1).
most. However, our studies revealed a driving force
DG~0 eV,[14,15] which requires the potential E(dU/dUCÀ) to be
~À1.8 V. In this context, the measured value E(dU/dUCÀ)=
À1.1 V provided by Steenken et al.[17] is difficult to under-
stand and could reflect the result of a proton-coupled ET.
Thus it is likely, that the À1.1 V potential corresponds to
E(dU/dU(H)C).
In comparison to Py as the electron donor, the reduction
potential of phenothiazine (Ptz) in the excited state E(PtzC+/
Ptz*)=À2.0 V[18] should be more efficient for the photore-
duction of T and C within DNA. Ptz has only been used as a
charge acceptor to investigate the DNA-mediated oxidative
hole transfer.[19,20] In order to use Ptz as a charge donor for
reductive electron transfer, we synthesised Ptz-dU (1) and
incorporated it into oligonucleotides. By this synthetic ap-
proach we are able to exclusively photoinitiate a reductive
electron transfer, since the intramolecular electron transfer
in the Ptz-dU moiety can be regarded as an electron injec-
tion into the DNA preceeding the electron hopping.
Palladium-assisted routes to modified nucleosides have
been explored extensively.[21] For the preparation of Ptz-dU
(1), we applied a Suzuki–Miyaura cross-coupling protocol
that we developed recently for the synthesis of pyrene-modi-
fied nucleosides (Scheme 2).[22] In general, this type of palla-
dium-catalysed coupling has the advantage that it works
Scheme 1. Electron transfer in Ptz-dU-modified DNA: The electron is
trapped kinetically on the Br-dU group yielding a DNA lesion which can
be analysed by HPLC after treatment with piperidine (X–Y=T–A or C–
G, n=1–3).
Results and Discussion
Synthesis and ET properties of 5-(10-methyl-phenothiazin-3-
yl)-2’-deoxyuridine: Until now our group has focused mainly
on pyrene (Py) as the photoexcitable electron donor for the
investigation of reductive electron transfer in DNA.[13–15] Ac-
cording to our studies, Py in the excited state (Py*) is a sig-
nificantly weaker electron donor than calculated and, there-
fore, than expected. Combining the potentials E(PyC+/Py*)=
À1.8 V (vs NHE)[16] and E(dU/dUCÀ)=À1.1 V,[17] the driving
force DG for the ET process in Py–dU could be À0.6 eV at
Scheme 2. Synthesis of Ptz-dU (1) and the corresponding DNA building
block 8: a) Br2 (1 equiv), AcOH/NaOAc/CH2Cl2, 58C, 30 min; 67%; b) 4
(2 equiv), Et3N (3 equiv), [PdCl2(dppf)] (0.03 equiv), dioxane, reflux,
16 h; 57%; c) 5 (1.2 equiv), PdCl2(dppf) (0.1 equiv), NaOH (20 equiv),
THF/H2O/MeOH, reflux, 44 h; 34%; d) DMT-Cl (2.0 equiv), pyridine,
RT, 16 h; 52%; e) 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite
(2.6 equiv), Et3N (2.9 equiv), CH2Cl2, RT, 45 min; 95%.
1872
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1871 – 1876

Electron Transfer Reactions
FULL PAPER
even in aqueous solutions and it tolerates the presence of
some unprotected functional groups.[23] The synthetic proce-
dure starts with the bromination of 10-methyl-phenothiazine
(2) by one equivalent of Br2 and subsequent Pd0-catalysed
coupling of the 3-bromo-10-methyl-phenothiazine (3) with
tetramethyl oxoborolane (4) according to the procedures by
Ebdrup and Mꢁller et al.[24] The palladium-catalysed
Suzuki–Miyaura coupling of the boronic acid ester 5 with
commercially available 5-iodo-2’-deoxyuridine (6) gives
Ptz-dU (1) in reasonable yield (34%).
The structure of Ptz-dU (1) was confirmed by different
spectroscopic techniques, including high-resolution mass
spectrometry and 2D NMR spectroscopy. Furthermore, the
Ptz-modified nucleoside 1 was characterised by UV-visible
absorption and steady-state fluorescence spectroscopy in
order to study its potential as a charge donor for ET in
DNA. The UV-visible spectrum of Ptz-dU (Figure 1, top)
ly published experiments using Py-modified nucleosides in
which back ET occurs partially into the Py excited state.
Preparation of Ptz-dU-modified oligonucleotides: For the
synthetic incorporation into oligonucleotides, the Ptz-modi-
fied uridine 1 was converted into the DMT-protected com-
pound 7 and then to the completely protected nucleoside 8
bearing the phosphoramidite group in the 3’-position
(Scheme 2). Standard procedures were applied for these two
synthetic steps. From the DNA building block 8, the Ptz-
modified oligonucleotides ss9–ss15 (ss=single strand) were
prepared by automated solid-phase synthesis on an Expe-
dite 8909 DNA synthesiser. Nearly quantitative coupling of
the monomer 8 was achieved with the standard coupling
time of 1.6 min. Additional to the Ptz-dU-modification, Br-
dU was inserted into the oligonucleotides by using the cor-
responding commercially available phosphoramidite. The
HPLC-purified oligonucleotides were identified by MALDI-
TOF mass spectrometry and quantified by their UV-visible
absorption.
Using the Ptz-modified oligonucleotides ss9–ss15, we pre-
pared the corresponding DNA duplexes 9–15 by slow cool-
Figure 1. Absorption and fluorescence spectra of Ptz-dU (1) and 10-
methyl phenothiazine (2) in MeOH/H2O=1:1. The emission was record-
ed at equal optical density (0.3) of 1 and 2 at the excitation wavelength
(340 nm).
ing together with 1.2 equivalents of the unmodified comple-
mentary strands. The synthesised Ptz-modified DNA duplex-
es 9–15 were subsequently characterised by their melting
temperatures Tm (Table 1) and their UV-visible absorption
spectra, which show clearly the presence of the Ptz chromo-
phore as a broad shoulder between 300 nm and 400 nm
(Figure 2).
As already mentioned, the focus in this manuscript was to
study the distance and base-sequence dependence of the ET
efficiency in Ptz-dU-modified DNA. Thus the sequences of
represents a combination of the absorption of the uridine
nucleoside (peak at ~260 nm) and the absorption of the Ptz
chromophore as a broad shoulder between 300 and 400 nm.
For the fluorescence spectroscopy measurements we chose
an excitation wavelength of 340 nm, a value which is clearly
in the absorption region of the Ptz moiety and avoids any
partial excitation of the uridine system. By using samples of
10-methyl-phenothiazine (2) and Ptz-dU (1) that were ad-
justed to the same optical density of 0.3 at 340 nm, a signifi-
cant difference in the emission quantum yield could be de-
tected (Figure 1, bottom). Remarkably, the Ptz-typical emis-
sion with its maximum at 445 nm is quenched completely in
case of Ptz-dU (1). This observation indicates an intramolec-
ular ET process, which together with back ET into the
ground state represents a nonradiative decay pathway for
the excited state. This result stands in contrast to our recent-
Table 1. Melting temperature (Tm) of the Ptz-modified DNA duplexes 9–
15 (2.5mm duplex, 10mm Na-Pi-buffer, 250mm NaCl, pH 7.0).
DNA
Tm [8C]
DNA
Tm [8C]
9
60
58
58
56
10
12
14
61
61
62
11
13
15
Chem. Eur. J. 2005, 11, 1871 – 1876
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1873

H.-A. Wagenknecht and C. Wagner
can be divided into three groups: the DNA duplexes 9 and
11 show comparable and high cleavage efficiency, the DNA
duplexes 13 and 10 are in the middle and the DNA duplexes
12 and 14 show the lowest cleavage efficiency. Interestingly,
the DNA duplexes with the intervening T–A base pairs (9,
11 and 13) show significantly higher cleavage efficiency rela-
tive to the DNA duplexes with the intervening C–G base
pairs (10, 12 and 14). It should be noted that the cleavage
efficiency of DNA 13 is comparable to that of DNA 10.
Thus considering the fact that strand degradation represents
the chemical result of the DNA-mediated ET process, it is
remarkable that just one intervening C–G base pair exhibits
a similar ET efficiency as three intervening T–A base pairs.
It is evident that in our assay T–A base pairs transport an
electron more efficiently than C–G base pairs.
Figure 2. UV-visible absorbance spectra of the Ptz-modified DNA du-
plexes 9–15 (2.5mm) in buffer (10mm Na-Pi, 250mm NaCl, pH 7).
duplexes 9–14 have been designed in such a way that the
Br-dU group as the electron acceptor is placed either two,
three or four base pairs away from the Br-dU group as the
photoexcitable electron donor. With respect to CCÀ and TCÀ
as the potential intermediate electron carriers during elec-
tron hopping, the intervening base pairs X–Y were chosen
to be either T–A or C–G. If proton transfer interferes with
electron transfer, we should observe significant differences
in the ET efficiencies between DNA set 9, 11 and 13 versus
DNA set 10, 12 and 14. The DNA 15 was prepared as a con-
trol duplex lacking the Br-dU group as a chemically reactive
electron trap.
Figure 3. Analysis of the strand cleavage experiments with DNA 9–14
(4mm) in buffer (10mm Na-Pi, 250mm NaCl, pH 7). Each experiment has
been repeated at least three times.
Irradiation experiments with Ptz-dU-modified DNA duplex-
es: As mentioned previously, Br-dU undergoes a chemical
modification after its one-electron reduction that can be an-
alysed by piperidine-induced strand cleavage (Scheme 1).[11]
Hence, the quantification of the strand cleavage yields im-
portant information about the ET efficiency.[10] Theoretical
studies showed that the electron affinity of Br-dU is signifi-
cantly higher than that of T.[25] However, based on reduction
potentials, Br-dU seems to be not a significantly better elec-
tron acceptor.[26] In conclusion, Br-dU represents more char-
acteristics of a kinetic than a thermodynamic electron trap.
The irradiation experiments of DNA duplexes 9–15 have
been performed in such a way that after the start of the ex-
periment, aliquots were collected every 5 min that were sub-
sequently treated with piperidine at elevated temperature
and finally analysed by HPLC. Each experiment had to be
repeated at least three times. A 75 W Xe lamp with a cut-
off filter (>305 nm) was used for these experiments in order
to selectively photoexcite the Ptz chromophore and to avoid
any degradation of the oligonucleotides by irradiation at
smaller wavelengths. No strand cleavages were observed
during the irradiation of DNA duplex 15. This experiment
represents an important control that any observed strand
cleavage in the DNA duplexes 9–14 can be assigned to the
presence of Br-dU as the electron acceptor. Indeed, strand
degradation can be observed during the irradiation of all
DNA duplexes 9–14, but the efficiencies of the strand cleav-
age show significant differences (Figure 3). These duplexes
Conclusion
The central motivation for this study was to elucidate the
role of pyrimidine bases during excess-electron transfer
through DNA. It was postulated that electron hopping in
DNA involves all base pairs (T–A and C–G) meaning that
both pyrimidine radical anions, CCÀ and TCÀ, play the role of
intermediate charge carriers,[6] although it was already
known that both radical anions exhibit a significantly differ-
ent basicity.[27] The DNA-mediated ET in the assay present-
ed herein was initiated by photoexcitation of Ptz-dU as the
electron donor and probed chemically with Br-dU as the
electron acceptor. Remarkably, from these strand cleavage
experiments it becomes clear that in our assay T–A base
pairs transport electrons more efficiently than C–G base
pairs (Scheme 3). This implies that CCÀ is not likely to play a
major role as an intermediate electron carrier.
This observation is supported by a number of recent pub-
lications:
1) As already mentioned in the introduction, Rokita ap-
plied aromatic amines as electron donors together with
Br-dU as an electron trap und showed that the ET effi-
ciency significantly depends on the intervening base se-
1874
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1871 – 1876

Electron Transfer Reactions
FULL PAPER
LCQ-ESI spectrometer. MALDI-TOF MS was performed in the analyti-
cal facility of the institute on a Bruker Biflex III spectrometer by using
3-hydroxypicolinic acid in aqueous ammonium citrate as the matrix. Ana-
lytical chromatography was performed on Merck silica gel 60 F254
plates. Flash chromatography was performed on Merck silica gel (40–
63 mm). C18-RP analytical and semipreparative HPLC columns (300 ꢂ)
were purchased from Supelco. Solvents were dried according to standard
procedures. All reactions were carried out under argon and protected
from light. Chemicals were purchased from Fluka, Lancaster and Aldrich
and were used without further purification. Spectroscopic measurements
were performed in quartz glass cuvettes (1 cm) and with Na-Pi buffer
(10mm). Absorption spectra and the melting temperatures (2.5mm
duplex, 250mm NaCl, 260 nm, 10–808C, interval 18C) were recorded on a
Varian Cary 100 spectrometer.
3-Bromo-10-methyl-phenothiazine (3) and 10-methyl-3-(4,4,5,5-tetra-
methyl [1,3,2]dioxaborolan-2-yl)-phenothiazine (5): These compounds
were synthesised according to procedures published by Ebdrup and
Mꢁller et al.[24] The analytical data were in agreement with the published
values.
5-(10-Methyl-phenothiazin-3-yl)-2’-deoxyuridine (1): 5-Iodo-2’-deoxyuri-
dine (6; 0.50 g, 1.41 mmol) was dissolved in THF/water (120 mL, 1:1).
Subsequently, 5 (0.58 g, 1.69 mmol), PdCl2(dppf) (1:1 complex with di-
chloromethane, 0.10 g, 0.14 mmol, 0.1 equiv), NaOH (1.13 g, 28.3 mmol,
20 equiv) and MeOH (50 mL) were added. The solution was saturated
with nitrogen at RT (30 min), refluxed for 44 h, neutralised with HCl
(1m), filtered through silica and extracted with EtOAc (4ꢃ30 mL). The
combined organic phase was dried over MgSO4 and concentrated to dry-
ness. The residue was purified by column chromatography on silica gel
(CH2Cl2/acetone 4:1, then EtOAc/MeOH 10:1) to give a yellow solid
(34% yield). Analytical HPLC ensured a purity of >99.5%. Rf =0.69
(EtOAc/MeOH/water 12:2:1); 1H NMR (500 MHz, [D6]DMSO): d=2.18
(m, 2H; 2’-H), 3.31 (s, 3H; CH3), 3.59 (m, 2H; 5’-H), 3.79 (m, 1H; 4’-H),
4.27 (m, 1H; 3’-H), 5.10 (br, 1H; 5-OH), 5.24 (br, 1H; 3-OH), 6.20 (t,
1H; 1’-H), 6.94 (m, 3H; Ptz-H), 7.14 (m, 1H; Ptz-H), 7.20 (m, 1H; Ptz-
H), 7.36 (m, 2H; Ptz-H), 8.12 ppm (s, 1H; 5-H); 13C NMR (125.8 MHz,
[D6]DMSO): d=162.59 (4-C), 150.31 (2-C), 145.58, 144.82, 137.62 (6-C),
128.29, 127.93, 127.63, 127.29, 126.46, 123.00, 122.31, 122.15, 115.07,
114.69, 112.83 (5-C), 87.99 (4’-C), 84.95 (1’-C), 70.65 (3’-C), 61.45 (5’-C),
39.90 (2’-C), 34.90 ppm (CH3); MS (ESI): m/z: 439 [M+], 440 [M++H],
879 [2M++H]; HRMS (MALDI): m/z calcd for C22H21N3O5S: 439.11263
[M++H+]; found: 439.1196; UV/Vis (MeOH/H2O 1:1, pH~6.5): nmax
(e)=260 nm (53200mÀ1 cmÀ1).
Scheme 3. Sequence dependence of the ET in Ptz-dU-modified DNA:
The ET occurs more efficiently over T–A base pairs than over C–G base
pairs, since proton-transfer (PT) interferes with ET.
quence.[10] The presence of C–G base pairs lowered the
ET efficiency significantly.
2) Sevilla employed EPR spectroscopy and showed that
proton transfer can slow down excess-electron transfer,
but does not stop it.[28]
3) The result fits into the interpretation of our spectroscop-
ic studies with Py-modified pyrimidine nucleosides as
models for ET in DNA.[14] Therein, the nonprotonated
radical anion of C could not be observed in aqueous so-
lution suggesting that the protonation of CCÀ by the com-
plementary DNA base G or the surrounding water mole-
cules will occur rapidly. Furthermore, we could show
that such protonation of CCÀ and deprotonation of C(H)C
can occur within several picoseconds. During this time
the hydrogen-bond interface can readjust and stabilise
the excess electron on the base by separating its spin
from its charge. Although these processes must be mi-
croscopically reversible they may ultimately terminate
electron migration in DNA due to the separation of spin
and charge.
5’-O-(4,4’-Dimethoxytrityl)-5-(10-methylphenothiazin-3-yl)-2’-deoxyuri-
dine (7): 4,4’-Dimethoxy-triphenylmethyl chloride (185 mg, 0.55 mmol)
was added to
a solution of 1 (120 mg, 0.27 mmol) in dry pyridine
(10 mL). The mixture was stirred overnight at RT. Subsequently, MeOH
(1 mL) was added. After 1 h at RT, the solution was concentrated to dry-
ness. The crude product was purified by flash chromatography (CH2Cl2/
acetone 4:1, then EtOAc/MeOH 10:1) yielding a pale yellow solid (52%
yield). Rf =0.85 (ethyl acetate/methanol/water 12:2:1); 1H NMR
(500 MHz, [D6]DMSO): d=2.22 (m, 1H; 2’-H), 2.36 (m, 1H; 2’-H), 3.12
(m, 1H; 5’-H), 3.18 (m, 1H; 5’-H), 3.25 (s, 3H; N-Me), 3.63 (s, 3H; O-
Me), 3.65 (s, 3H; OMe), 3.94 (m, 1H; 4’-H), 4.28 (m, 1H; 3’-H), 5.33 (br,
1H; 3’-OH), 6.22 (dd, J=6.7 Hz, 1H; 1’-H), 6.60 (d, J=8.5 Hz, 1H; Ptz),
6.73 (m, 4H; DMT), 6.95 (m, 2H; Ptz), 7.06–7.11 (m, 3H; Ptz), 7.14–7.22
(m, 8H; Ptz, DMT), 7.31 (m, 2H; DMT), 7.63 (s, 1H; 5-H), 11.56 ppm (s,
1H; N-H); 13C NMR (125.8 MHz, [D6]DMSO): d=162.10 (4-C), 157.98,
157.97, 149.82 (2-C), 145.08, 144.68, 144.39, 136.40 (6-C), 135.43, 135.21,
129.61, 127.74, 127.52, 127.43, 127.04, 126.74, 126.56, 126.28, 122.48,
121.91, 121.62, 114.51, 113.80, 113.07, 113.04, 85.78, 85.64 (4’-C), 84.79
(1’-C), 70.52 (3’-C), 63.84 (5’-C), 54.90 (O-CH3), 54.88 (O-CH3), 39.65 (2’-
C), 34.90 ppm (N-CH3); ESI-MS: m/z: 741 [M+], 764 [M++Na].
In summary, it is clear now that excess-electron transfer
through hopping is highly sequence dependent and occurs
faster and more efficiently over T–A base pairs than over
C–G base pairs.
Experimental Section
5’-O-(4,4’-Dimethoxy)trityl-5-(10-methylphenothiazin-3-yl)-2’-deoxyuri-
dine-3’-O-(2-cyanoethyl)-N,N-diisopropylphosphoramidite (8): Com-
pound 7 (150 mg, 0.2 mmol) was dissolved in dry CH2Cl2 (14 mL). Tri-
ethylamine (0.08 mL, 0.6 mmol) and 2-cyanoethyl-N,N-diisopropylchloro-
phosphoramidite (0.08 mL, 0.35 mmol) were added and the solution stir-
red for 1 h at RT. Ethanol (0.1 mL) was added and the mixture was
Materials and methods: 1H, 13C and the 2D NMR spectra were recorded
on a Bruker DMX500 spectrometer. NMR signals were assigned based
on 2D NMR measurements (DQF-COSY, HSQC). ESI mass spectra
were measured in the analytical facility of the institute on a Finnigan
Chem. Eur. J. 2005, 11, 1871 – 1876
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1875

H.-A. Wagenknecht and C. Wagner
poured into aqueous saturated NaHCO3 (50 mL) and extracted with
CH2Cl2. The organic layer was dried (Na2SO4) and evaporated yielding
the phosphoramidite 8 as a yellow solid (95%), which was used directly
for the oligonucleotide synthesis. ESI-MS: m/z: 941 [M+].
[4] a) G. Hartwich, D. J. Caruana, T. de Lumley-Woodyear, Y. Wu, C. N.
Campbell, A. Heller, J. Am. Chem. Soc. 1999, 121, 10803–10812;
b) E. M. Boon, D. Ceres, T. G. Drummond, M. G. Hill, J. K. Barton,
Nat. Biotechnol. 2000, 18, 1096–1100; c) E. M. Boon, J. E. Salas,
J. K. Barton, Nat. Biotechnol. 2002, 20, 282–286.
[5] e.g. a) C. Mao, W. Sun, Z. Shen, N. C. Seeman, Nature 1999, 397,
144–146; b) H. W. Fink, C. Schçnenberger, Nature 1999, 398, 407–
410; c) D. Porath, A. Bezryadin, S. de Vries, C. Dekker, Nature
2000, 403, 635–638; d) C. M. Niemeyer, M. Adler, Angew. Chem.
2002, 114, 3933–3937; Angew. Chem. Int. Ed. 2002, 41, 3779–3783.
[6] B. Giese, Annu. Rev. Biochem. 2002, 71, 51–70.
[7] Z. Cai, M. D. Sevilla, Top. Curr. Chem. 2004, 237, 103–128.
[8] Most recently: C. Haas, K. Krꢅling, M. Cichon, N. Rahe, T. Carell,
Angew. Chem. 2004, 116, 1878–1880; Angew. Chem. Int. Ed. 2004,
43, 1842–1844.
[9] Most recently: B. Giese, B. Carl, T. Carl, T. Carell, C. Behrens, U.
Hennecke, O. Schiemann, E. Feresin, Angew. Chem. 2004, 116,
1884–1887; Angew. Chem. Int. Ed. 2004, 43, 1848–1851.
[10] Most recently: T. Ito, S. E. Rokita, Angew. Chem. 2004, 116, 1875–
1878; Angew. Chem. Int. Ed. 2004, 43, 1839–1842.
[11] a) E. Rivera, R. H. Schuler, J. Phys. Chem. 1983, 87, 3966–3971;
b) T. Chen, G. P. Cook, A. T. Koppisch, M. M. Greenberg, J. Am.
Chem. Soc. 2000, 122, 3861–3866.
[12] F. D. Lewis, X. Liu, S. E. Miller, R. T. Hayes, M. R. Wasielewski, J.
Am. Chem. Soc. 2002, 124, 11280–11281.
[13] N. Amann, E. Pandurski, T. Fiebig, H.-A. Wagenknecht, Chem. Eur.
J. 2002, 8, 4877–4883.
[14] a) R. Huber, T. Fiebig, H.-A. Wagenknecht, Chem. Commun. 2003,
1878–1879; b) M. Raytchev, E. Mayer, N. Amann, H.-A. Wagen-
knecht, T. Fiebig, ChemPhysChem 2004, 5, 706–712.
Preparation and characterisation of the oligonucleotides—general proce-
dure: The oligonucleotides were prepared on a Expedite 8909 DNA syn-
thesiser from Applied Biosystems (ABI) by standard phosphoramidite
protocols with chemicals and CPG (1 mol) from ABI and Glen Research.
After preparation, the trityl-off oligonucleotide was cleaved off the resin
and was deprotected by treatment with concentrated NH4OH at 608C for
10 h. The oligonucleotide was dried and purified by HPLC on a semipre-
parative RP-C18 column (300 ꢂ, Supelco) by using the following condi-
tions: A=NH4OAc buffer (50 mm), pH 6.5; B=MeCN; gradient=0–
15% B over 45 min. The oligonucleotides were lyophilised and quantified
by their absorbance at 260 nm[29] on a Varian Cary 100 spectrometer. Du-
plexes were formed by heating to 808C (10 min), followed by slow cool-
ing.
General procedure for the solid-phase synthesis of the phenothiazine-
modified oligonucleotides 9–15: The syntheses were performed on a
1 mol scale (CPG 1000 ꢂ, Glen Research) by using standard phosphor-
amidite protocols. Quantitative coupling of the building block 8 was ach-
ieved by using the minimal coupling time of 1.6 min. After preparation,
the trityl-off oligonucleotide was cleaved off the resin and was deprotect-
ed by treatment with concentrated NH4OH at RT for 36 h. The oligonu-
cleotide was dried and purified by HPLC on a semipreparative RP-C18
column (300 ꢂ, Supelco) using the following conditions: A=NH4OAc
buffer (50 mm), pH 6.5; B=MeCN; gradient=0–30 B over 45 min. The
oligonucleotides were lyophilised, quantified by their absorbance at
260 nm[29] and using e=53200 (260 nm) for 1. MS (MALDI): m/z calcd
for ss9: 5700 [M+]; found: 5711; m/z calcd for ss10: 5685 [M+]; found:
5704; m/z calcd for ss11 5700 [M+]; found: 5702; m/z calcd for ss12: 5670
[M+]; found: 5675; m/z calcd for ss13 5700 [M+]; found: 5705; m/z calcd
for ss14 5655 [M+]; found: 5663; m/z calcd for ss15: 5346 [M+]; found: 5345.
[15] T. Fiebig, A. Trifonov, H.-A. Wagenknecht, I. Buchvarov, unpub-
lished results.
+
[16] E(PyC /Py)=1.5 V (vs NHE) and E00 =3.25 eV, see: T. Kubota, K.
Kano, T. Konse, Bull. Chem. Soc. Jpn. 1987, 60, 3865–3877.
[17] S. Steenken, J. P. Telo, H. M. Novais, L. P. Candeias, J. Am. Chem.
Soc. 1992, 114, 4701–4709.
Strand cleaving experiments: Duplexes (4mm DNA, 10mm Na-Pi buffer,
250mm NaCl) were prepared by heating equimolar solutions of the single
strands to 808C for 10 min in the dark and subsequent slow cooling. The
measurements were performed in quartz glass cuvettes (1 cm). The fresh-
ly prepared duplexes were irradiated by an Xe lamp (75 W Xe lamp,
Oriel Instruments) and a cut-off filter (Andover Corporation, >305 nm).
Every 5 min aliquots (30 mL) of the sample solution (1000 mL) were
taken. Piperidine (3 mL) was added and the samples were heated to 908C
(30 min), lyophilised, dissolved in water (15 mL) and analysed by HPLC
(RP-C18, Supelco) under the following conditions: A=NH4OAc buffer
(50 mm), pH 6.5; B=MeCN; gradient=0–30% B over 45 min. The ob-
tained peaks were processed to give a ratio between unmodified and
modified ssDNA.
+
[18] E(PtzC /Ptz)=1.0 V (vs NHE), see: D. G. McCafferty, D. A. Friesen,
E. Danielson, C. G. Wall, M. J. Saderholm, B. W. Erickson, T. J.
Meyer, Proc. Natl. Acad. Sci. USA 1996, 93, 8200–8204; and E00
=
3.0 eV, see: D. J. Freed, L. R. Faulkner, J. Am. Chem. Soc. 1972, 94,
4790–4792.
[19] a) M. T. Tierney, M. Sykora, S. I. Khan, M. W. Grinstaff, J. Phys.
Chem. B 2000, 104, 7474–7576; b) M. T. Tierney, M. W. Grinstaff,
Org. Lett. 2000, 2, 3413–3416; c) S. A. N. Hashmi, X. Hu, C. E.
Immoos, S. J. Lee, M. W. Grinstaff, Org. Lett. 2002, 4, 4571–4574.
[20] a) K. Kawai, T. Takada, S. Tojo, T. Majima, J. Am. Chem. Soc. 2003,
125, 6842–6843; b) T. Takada, K. Kawai, X. Cai, A. Sugimoto, M.
Fujitsuka, T. Majima, J. Am. Chem. Soc. 2004, 126, 1125–1129.
[21] See recent review: L. A. Agrofoglio, I. Gillaizeau, Y. Saito, Chem.
Rev. 2003, 103, 1875–1916.
[22] E. Mayer, L. Valis, R. Huber, N. Amann, H.-A. Wagenknecht, Syn-
thesis 2003, 2335–2340.
[23] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft, the
Volkswagen-Stiftung and the Fonds der Chemischen Industrie. We are
grateful to Professor Horst Kessler, Technical University of Munich, for
the generous support and to Professor Michael Przybylski, University of
Konstanz, for measuring the HRMS.
[24] a) S. Ebdrup, J. Chem. Soc. Perkin Trans.
1 1998, 1147–1150;
b) C. S. Krꢅmer, T. J. Zimmermann, M. Sailer, T. J. J. Mꢁller, Synthe-
sis 2002, 1163–1170.
[25] a) S. D. Wetmore, R. J. Boyd, L. A. Eriksson, Chem. Phys. Lett.
2001, 343, 151–158; b) X. Li, M. D. Sevilla, L. Sanche, J. Am. Chem.
Soc. 2003, 125, 8916–8920.
[26] The reduction potential is only 0.06 eV lower than that of T: V. P.
Kadysh, Y. L. Kaminskii, L. N. Rumyantseva, V. L. Efimova, J. P.
Strandish, Khim. Geterotsikl. Soedin. 1992, 10, 1404–1408.
[27] S. Steenken, Free Radical Res. Commun. 1992, 16, 349–379.
[28] Z. Cai, X. Li, M. D. Sevilla, J. Phys. Chem. B 2002, 106, 2755–2762.
[29] J. D. Puglisi, I. Tinoco, Methods Enzymol. 1989, 180, 304–325.
[1] H.-A. Wagenknecht, Angew. Chem. 2003, 115, 2558–2565; Angew.
Chem. Int. Ed. 2003, 42, 2454–2460.
[2] a) P. OꢄNeill, E. M. Frieden, Adv. Radiat. Biol. 1993, 17, 53–120;
b) B. Armitage, Chem. Rev. 1998, 98, 1171–1200; c) C. J. Burrows,
J. G. Muller, Chem. Rev. 1998, 98, 1109–1151; d) D. Wang, D. A.
Kreutzer, J. M. Essigmann, Mutat. Res. 1998, 400, 99–115; e) S. Ka-
wanashi, Y. Hiraku, S. Oikawa, Mutat. Res. 2001, 488, 65–76.
[3] Top. Curr. Chem. 2004, 236 and 237, whole volumes.
Received: October 4, 2004
Published online: January 31, 2005
1876
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 1871 – 1876