radical cations with water, which afforded products P3A and P5A
after subsequent strand cleavage (Scheme 3).12
But this is not the case as the data in Fig. 2 show. A possible
explanation is that the positive charge, injected into the adenine
is delocalized over adjacent A+T base pairs. According to this
suggestion, in well-organised sequences where the adjacent
base pairs of the DNA have nearly the same redox potentials, a
positive charge is not localised at one base pair but is distributed
over several base pairs. This is reminiscent of the ‘polaron’
picture for long distance charge transfer through DNA that was
discussed by G. B. Schuster13 and E. M. Conwell,14 or even of
the conductive band in a semiconductor.15 Further experiments
with this new assay will show how robust this charge
delocalisation is against changes in the A+T sequence.
First, we checked whether the efficiency of the charge
transfer towards the 5A-end is different from that towards the 3A-
end. It turned out that photolysis of 14a, where both sequences
between A·+ and the GGG units contain two A+T base pairs,
gave about the same amount of products P3A (55%) and P5A
(45%) (Fig. 2). In strand 14b one of the A+T sequences was
extended from 2 to 8 A+T base pairs. Nevertheless, we observed
nearly the same ratio of products P3A/P5A (Fig. 2). Thus, the
efficiency of the charge migration through the A+T sequences
alters very little depending on the number of the A+T base pairs
in these experiments.
Generous support by the Swiss National Science Foundation
and the Volkswagen Foundation are gratefully acknowledged.
If the positive charge hops reversibly between all adenines,
the charge transport via the stretch of 8 A+T base pairs of 14b
should be less efficient than via the stretch of 2 A+T base
pairs.2,4
Notes and references
1 B. Giese, J. Amaudrut, A.-K. Köhler, M. Spormann and S. Wessely,
Nature, 2001, 412, 318.
2 J. Jortner, M. Bixon, T. Langenbacher and M. E. Michel-Beyerle, Proc.
Natl. Acad. Sci. USA,, 1998, 95, 12759.
3 B. Giese and M. Spichty, ChemPhysChem, 2000, 1, 195; B. Giese and
A. Biland, Chem. Commun., 2002, 667.
4 E. Meggers, M. E. Michel-Beyerle and B. Giese, J. Am. Chem. Soc.,
1998, 120, 12950.
5 D. Crich and X. Hao, J. Org. Chem., 1999, 64, 4016.
6 The debenzylation of 3 required NBS because standard deprotection
methods like hydrogenation (Pd/C, H2) were unsuccessful. The
experimental details for the synthesis of modified nucleotides 7 and 10
will be described in a full paper. The configurations were confirmed by
comparing the 1H-NMR spectra of compound 8 with the analogous ester
nuclosides described in ref. 5. The a- and b-anomers are easily
distinguished by the chemical shifts of the OTBDMS group at C-3A.
7 M. Spormann and B. Giese, Synthesis, 2001, 2156.
8 E. Meggers, A. Dussy, T. Schäfer and B. Giese, Chem. Eur. J., 2000, 6,
485.
9 C. A. M. Seidel, A. Schulz and M. H. M. Sauer, J. Phys. Chem., 1996,
100, 5541; S. Steenken and S. V. Jovanovic, J. Am. Chem. Soc., 1997,
119, 617.
Scheme 3
10 Photolysis of nucleotide 10c was carried out following the procedure
described in ref. 7.
11 The syntheses of oligonucleotides were carried out on a DNA
synthesizer in 0.2 mmol scales (500 Å controlled pore glass support).
The standard method for 2-cyanoethylphosphoramidites was used,
except that the coupling of the modified nucleotide 7 was done manually
by removing the solid phase column from the machine and by passing
the phosphoramidite into the column via two syringes, one containing 7
in MeCN and the second syringe containing coupling solution. With this
modification there is no notable difference between the efficiency of
coupling for this modified amidite and non-modified ones. Workup was
done by standard procedures. The purity of all oligonucleotides was
controlled by reverse phase chromatography and MALDI-ToF MS.
Photolyses of the modified oligonucleotides were performed as
described in ref. 4. The sequences upstream (5A-direction) and
downstream (3A-direction) of the GGG units are ATATAATTTCG and
ATATTATGCGA, respectively.
12 For the strand cleavage at the sites of oxidised guanines 30 ml of the
probes were treated with 300 ml of 1 M piperidine at 90 °C for 30 min.
With this assay we could not detect cleavage at the adenines. By this
method the efficiencies of the charge transfer reactions are measured,
which not only depend upon the charge transfer rate but also upon the
rate of the irreversible trapping of the guanine radical cation (GGG·+) by
H2O. Only if this water trapping reaction is as fast or faster than the
endothermic electron transfer from A to GGG·+, these yields can be
correlated with charge transfer rate ratios. From our earlier experiments
(ref. 1) it can be deduced that this might actually be the case.
13 G. B. Schuster, Acc. Chem. Res., 2000, 33, 253.
Fig. 2 Histogram of denaturing polyacrylamide gels, obtained by subtrac-
tion of control experiments (irradiation of unmodified strands) from
irradiation experiments with the modified strand 14a (n = 1), and relative
yields of the strand cleavage products at the 5A- and 3A-GGG units for strands
14a (n = 1) and 14b (n = 4).
14 S. V. Rakhmanova and E. M. Conwell, J. Phys. Chem. B, 2001, 105,
2056.
15 D. Porath, A. Bezryadin, S. de Vries and C. Dekker, Nature, 2000, 403,
635.
CHEM. COMMUN., 2002, 2016–2017
2017