Synthesis and Characterization of Oligonucleotides Containing the C4′-Oxidized Abasic Site Produced by Bleomycin and Other DNA Damaging Agents

Article abstract of DOI:10.1002/anie.200352102

Selective introduction of damage: The C4′-position of DNA serves as the site of attack for a number of DNA damaging agents, including bleomycin. The C4′-abasic site is a major product found in damaged DNA following abstraction of the hydrogen atom from th

Full text of DOI:10.1002/anie.200352102

Communications
Synthesis of Oligonucleotides
Synthesis and Characterization of
Oligonucleotides Containing the C4’-Oxidized
Abasic Site Produced by Bleomycin and Other
DNA Damaging Agents**
Jaeseung Kim, Jun Mo Gil, and Marc M. Greenberg*
Exposure of DNA to oxidative stress produces strand scission
and a variety of damaged nucleotides, which can be mutagenic
and/or cytotoxic. The determination of the effects of specific
lesions on nucleic acid structure, stability, and their inter-
action with polymerase and repair enzymes provide a
chemical basis for the biological effects of DNA damage.[1,2]
The execution of such investigations is greatly facilitated by
the preparation of oligonucleotides containing lesions at
defined sites.[3] Oxidized abasic site 1 results from formal
abstraction of the C4’-hydrogen atom of a nucleotide in DNA.
This alkali-labile lesion accounts for ꢀ 40% of the DNA
damage induced by the antitumor antibiotic bleomycin.[4] The
C4’-oxidized abasic lesion is also produced following expo-
sure of DNA to the neocarzinostatin chromophore, the
enediynes, g-radiolysis, and a variety of other damaging
agents.[5] This oxidized abasic lesion is believed to be
mutagenic and cytotoxic, and gives rise to deletions and
substitutions in bacteria and mammalian cells.[6] Investiga-
tions of 1 have relied upon its sequence nonspecific and
nonexclusive generation by the oxidizing agents mentioned
above. In contrast, studies concerning related abasic sites (2,
3) are facilitated by the accessibility of oligonucleotides
containing them at defined sites.[7–9] Herein we report the first
method for chemically synthesizing oligonucleotides that
contain 1 at a defined site, and preliminary investigation of
the lesion's effects on DNA.
The mechanism by which 1 is produced during DNA
oxidation by bleomycin was investigated, and information
regarding the source of oxygen at the C4 position was
obtained.[4d] To our knowledge there are no reports that
describe the ultimate stereochemistry of 1 in DNA, let alone
[*] Prof. Dr. M. M. Greenberg, J. Kim, J. M. Gil
Department of Chemistry
Johns Hopkins University
3400 N. Charles St., Baltimore, MD 21218 (USA)
Fax: (+1)410-516-8420
E-mail: mgreenberg@jhu.edu
[**] We are grateful for support of this research by the National
Institutes of General Medical Sciences (GM-63028).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5882
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200352102
Angew. Chem. Int. Ed. 2003, 42, 5882 –5885

Angewandte
Chemie
the stepwise mechanism by which it is formed from 4. A
plausible mechanism that was put forth suggests that initial
oxidation of the C4’ position activates hydrolysis of the
glycosidic bond (Scheme 1).[4c,d] The formation of 1 from the
Scheme 2. Synthesis of 1 in DNA.
Scheme 1. Proposed mechanism for the activation of the glycosidic
bond.
without exposing the bisacetal (7) to acid.[11] The o-nitro-
veratryl group is a member of the class of o-nitrobenzyl
photolabile protecting groups that are of general use in
synthesis, and oligonucleotide synthesis in particular.[12] It was
anticipated that oligonucleotides containing 7 would be
purified, stored, and used to generate 1 as needed.
acyclic precursor 5 should result in a mixture of diastereo-
mers.[4c,d] This expectation is consistent with NMR character-
ization of 2 in DNA, which also leads us to believe that the
stereoisomers of 1 should exist in equilibrium with each other
and small amounts of 5.[10] As there was no reason for us to
expect that 1 will exist in DNA as a single isomer, we saw no
reason to attempt to generate it from individual stereoisomers
of a stable precursor. While devising our approach, we also
took into account that chemical syntheses of biopolymers that
contain 2 and 3 are able to utilize the alkaline deprotection
conditions typically employed during oligonucleotide syn-
thesis by unmasking the alkali-labile lesions in a final
photochemical step.[7,8] These considerations led us to
design 6 as the phosphoramidite used during the automated
synthesis to ultimately synthesize 1 in DNA (Scheme 2).
Notable features of 6 include the o-nitroveratryl moiety
(ONV), which serves as the alkali-resistant phototrigger, and
the silyloxy protecting group for the primary alcohol. The
latter allows us to synthesize protected oligonucleotides
Phosphoramidite 6 was prepared from the 3’,5’-O-silyloxy
deoxyribose acetal 8 (Scheme 3). Cleavage of the methyl
acetal in the presence of 1,3-propane dithiol provided the
dithiane 9 with a C4-hydroxy group, which was then
oxidized.[13] Subsequent oxidative cleavage of the dithiane
ketone, 10, in the presence of 3,4-dimethoxy-6-nitrobenzyl
alcohol provided 11 as a mixture of four stereoisomers.[14]
Although separation of the isomers was not practical at this
step, the corresponding 1H NMR spectrum indicated that two
were formed in significantly greater amounts. Analytical
samples of the individual diastereomers were obtained upon
1
desilylation (12). H NMR spectroscopy was used to deter-
mine that 1R,4S-12 and 1S,4R-12 accounted for > 80% of the
mixture of cyclized compounds. In practice, the major
diastereomers were separated upon preparation of the cyclo-
dodecyloxy bis-trimethylsilyloxy silyl ethers (1R,4S- 1S,4R-
Scheme 3. a) Me2SiCl2, MeOH; b) tert-butyldiphenylsilylchloride (TBDPSCl), imadazole; c) propane dithiol, BF3·Et2O, CH2Cl2, À108C; d) oxalyl
chloride, dimethyl sulfoxide (DMSO), Et3N, À788C; e) N-bromosuccinimide, 3,4-dimethoxy-6-nitrobenzyl alcohol (NVOH), CH3CN:CH2Cl2 (2:1),
À108C; f) tetrabutylammonium fluoride (TBAF), THF; g)cyclododecyloxy-bis-trimethylsilyloxysilyl chloride (RCl), diisopropylamine, CH2Cl2 0–
258C; h) methyl N,N-diisopropylchlorophosphoramidite, diisopropylethylamine, CH2Cl2, 0–258C.
Angew. Chem. Int. Ed. 2003, 42, 5882 –5885
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5883

Communications
13), which were carried on to their respective phosphorami-
dites (1R,4S- 1S,4R-6).
The use of fluoride ions to unmask the primary hydroxy
group during synthesis required that oligonucleotides be
prepared on polystyrene solid-phase supports by using O-
methyl phosphoramidites.[11] Solid-phase oligonucleotide syn-
thesis was carried out in the 3’- to 5’-direction on 1 mmol scale
by using modified versions of previously reported cycles.[11]
With the exception of 6, all phosphoramidites are commer-
cially available, and were used to prepare oligonucleotides
containing as many as 30 nucleotides (14–16). 5’-O-Dime-
thoxytrityl phosphoramidites containing “fast deprotecting”
exocyclic amine protecting groups were used prior to
incorporating 6. This enabled us to monitor the progress of
the synthesis by measuring the amount of the dimethoxytrityl
cation released. Because of an absence of a similar UV-
absorbing chromophore, it was not possible to directly
measure the coupling of 6 or subsequent silylated phosphor-
amidites of the native nucleotides. Consequently, the coupling
efficiency of 1R,4S-6 was determined indirectly by comparing
the isolated yield of 14a to an otherwise identical oligonu-
cleotide prepared entirely with 5’-O-dimethoxytrityl phos-
phoramidites, which contained thymidine in place of 7. After
the resin loading had been taken into account, the yield of 14a
was 89%, which was as much as the control oligonucleotide,
and this yield can be used to estimate the lower limit for the
coupling yield of 6. No differences in coupling yields were
observed when a mixture of 1R,4S- and 1S,4R-6 was used, and
for simplicity oligonucleotide synthesis with this mixture is
currently the method of choice. All oligonucleotides were
deprotected in two steps (Scheme 2). Phosphate triester
demethylation by using Na2S2 preceded concentrated aque-
ous NH3 deprotection/cleavage. After deprotection, photo-
labile oligonucleotides, 14–16, were isolated by denaturing
polyacrylamide gel electrophoresis, and characterized by
ESIMS.[15]
Figure 1. HPLC analysis of (bottom) 14a (prior to photolysis) and
(top) 14b (after 20 min photolysis at lmax =350 nm; intensity:
4.2 mWcmÀ2 @ 360 nm). UV detection at 260 nm.
formation of the desired products, which were observed as a
mixture of the cyclic and acyclic (-H2O) forms.[15] The relative
amounts of the cyclic and acyclic forms varied from sample to
sample.
The stability of 1 in an oligonucleotide (5’-32P-15b)
towards cleavage (phosphate buffer (10 mm), pH 7.5; NaCl
(100 mm)) was measured as a function of temperature by
using denaturing gel electrophoresis (Figure 2). The C4’-
Figure 2. Arrhenius plot describing the cleavage of 1 in 15b.
oxidized abasic site (k = 2.46 ” 10À5 sÀ1,
t1/2 = 7.8 h) was
> 10 times as susceptible to cleavage than 3 under comparable
conditions (378C, pH 7.5), which was found to be 20 times
more reactive than a non-oxidized abasic site (2).[16] Despite
the relative lability of 1, these studies clearly indicate that 1
Photolytic liberation of 1 proved to be extremely efficient.
HPLC analysis of photolyzed 14 showed 100% conversion in
five minutes of irradiation to a single faster eluting product
that did not absorb at 360 nm (Figure 1). In all cases 100%
conversion of 7 was achieved during 20 min photolyses. Mass
spectral analysis of the photolysates of 14–16 verified
5884
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 5882 –5885

Angewandte
Chemie
dam, Z. Hantosi, M. M. Greenberg, J. Am. Chem. Soc. 2002, 124,
3263; c) S. Iwai, Chem. Eur. J. 2001, 7, 4343.
will be sufficiently long-lived to warrant attention by DNA
repair enzymes.
[4] a) J. Stubbe, J. W. Kozarich, Chem. Rev. 1987, 87, 1107; b) H.
Sugiyama, C. Xu, N. Murugesan, S. M. Hecht, J. Am. Chem. Soc.
1985, 107, 4104; c) L. E. Rabow, J. Stubbe, J. W. Kozarich, J. Am.
Chem. Soc. 1990, 112, 3196; d) L. E. Rabow, G. H. McGall, J.
Stubbe, J. W. Kozarich, J. Am. Chem. Soc. 1990, 112, 3203.
[5] For reviews see: a) C. von Sonntag, The Chemical Basis of
Radiation Biology, Taylor and Francis, London, 1987; b) M. M.
Greenberg in Comprehensive Natural Products Chemistry, Vol. 7
(Ed. E. T. Kool), Pergamon, Amstedam, 1999, p. 371; c) Z. Xi,
I. H. Goldberg in Comprehensive Natural Products Chemistry,
Vol. 7 (Ed. E. T. Kool), Pergamon, Amstedam, 1999, p. 553.
[6] a) R. A. O. Bennett, P. S. Swerdlow, L. F. Povirk, Biochemistry
1993, 32, 3188; b) L. F. Povirk, W. Houlgrave, Environ. Mol.
Mutagen. 1988, 11, 461; c) L. F. Povirk, I. H. Goldberg, Biochi-
mie 1987, 69, 8150.
[7] a) H. J. Lenox, C. P. McCoy, T. L. Sheppard, Org. Lett. 2001, 3,
2415; b) M. Kotera, A. Bourdat E. Defrancq, J. Lhomme, J. Am.
Chem. Soc. 1998, 120, 11810; c) C. Tronche, B. K. Goodman,
M. M. Greenberg, Chem. Biol. 1998, 5, 263; d) J. T. Hwang,
K. A. Tallman, M. M. Greenberg, Nucleic Acids Res. 1999, 27,
3805; e) B. K. Goodman, M. M. Greenberg, J. Org. Chem. 1996,
61, 2.
A preliminary characterization of the effect of 1 on duplex
thermal stability was carried out with the dodecamer 14b. The
lability of 1 at 558C led us to investigate the hybridization of
14b to a complement containing A opposite the lesion at
378C to avoid cleaving 1 during hybridization. The suitability
of the hybridization conditions was established by using the
tetrahydrofuran abasic site analogue, 17d. Identical Tm's were
obtained for 17d whether the duplex was hybridized at 378C
(30 min, followed by 2 h at 258C) or at 958C (5 min, followed
by slow cooling to 258C). Consequently, the mild conditions
(378C) were used for the hybridization of 14b, and the
temperature was not allowed to go above 558C during the
melting process. The Tm for 17b (37.18C, 2.2 mm) did not
change over three annealing/melting cycles.[15] Furthermore,
no decomposition of 14b subjected to the identical heating
conditions side-by-side with 17b was detected by HPLC.[15]
The Tm for the resulting duplex containing 1 (17b) was
comparable to that for 17d, as might be expected for the
structurally similar abasic sites. The higher Tm measured for
17c may reflect the stabilization of the duplex due to p-
stacking of the o-nitroveratryl groups.
[8] a) J. J. Vasseur, D. Peoc'h, B. Rayner, J. L. Imbach, Nucleosides
Nucleotides 1991, 10, 107; b) T. Horn, K. Downing, Y. Gee, M. S.
Urdea, Nucleosides Nucleotides 1991, 10, 299.
In conclusion, we have developed an efficient method for
synthesizing oligonucleotides that contain the C4’-oxidized
abasic site, 1, a common alkali-labile lesion. Despite the
lability of 1, the studies described above clearly indicate that
the lesion is amenable to routine handling necessary for
carrying out biochemical and biophysical characterization.
Specifically, the C4’-oxidized abasic site 1 is sufficiently stable
to enable examination of its interactions in a variety of
environments such as those that include polymerases and
DNA repair enzymes. The ability to store oligonucleotides
that contain a stable, convenient photochemical precursor for
the C4’-oxidized abasic site 6 in oligonucleotides will facilitate
such studies on the lesion.
Experimental procedures for the synthesis of 6, a descrip-
tion of oligonucleotide synthesis cycles, deprotection, photol-
ysis, melting experiments, and HPLC analysis conditions are
available in the Supporting Information. A table of the UV-
melting data and sample melt for 17b, HPLC traces of 14b
before and after 3 annealing/melting cycles, mass spectra of
14–16 are also available.
[9] a) G. R. Stuart, R. W. Chambers, Nucleic Acids Res. 1987, 15,
7451; b) M. Takeshita, C. N. Chang, F. Johnson, S. Will, A. P.
Grollman, J. Biol. Chem. 1987, 262, 10171.
[10] a) J. A. Wilde, P. H. Bolton, A. Mazumder, M. Manoharan, J. A.
Gerlt, J. Am. Chem. Soc. 1989, 111, 1894; b) M. Manoharan, S. C.
Ransom, A. Mazumder, J. A. Gerlt, J. A. Wilde, J. A. Withka,
P. H. Bolton, J. Am. Chem. Soc. 1988, 110, 1620.
[11] S. A. Scaringe, F. E. Wincott, M. H. Caruthers, J. Am. Chem. Soc.
1998, 120, 11820.
[12] a) G. H. McGall, A. D. Barone, M. Diggelmann, S. P. A. Fodor,
E. Gentalen, N. Ngo, J. Am. Chem. Soc. 1997, 119, 5081; b) M. C.
Pirrung, Angew. Chem. Int. Ed. 2002, 41, 1276; c) V. N. R. Pillai,
Synthesis 1980, 1.
[13] Y. Honda, A. Ori, G. Tsuchibashi, Chem. Lett. 1989, 1259.
[14] E. J. Corey, B. W. Erickson, J. Org. Chem. 1971, 36, 3553.
[15] See Supporting Information.
[16] a) Y. Roupioz, J. Lhomme, M. Kotera, J. Am. Chem. Soc. 2002,
124, 9129; b) S. Shibutani, M. Takeshita, A. P. Grollman, J. Biol.
Chem. 1997, 272, 13916.
Received: June 10, 2003
Revised: August 25, 2003 [Z52102]
Keywords: bioorganic chemistry · DNA damage ·
.
oligonucleotides
[1] a) H. Kamiya, Nucleic Acids Res. 2003, 31, 517; b) D. Wang,
D. A. Kreutzer, J. M. Essigmann, Mutat. Res. 1998, 400, 99.
[2] a) J. Cadet, C. D'Ham, T. Douki, J. P. Pouget, J. L. Ravanat, S.
Sauvaigo, Free Radical Res. 1999, 29, 541; b) M. Dizdaroglu, Free
Radical Res. 1998, 29, 551; c) L. Gros, M. Saparbaev, J. Laval,
Oncogene 2002, 21, 8905.
[3] a) J. Butenandt, L. T. Burgdorf, T. Carell, Synthesis 1999, 1085;
b) K. Haraguchi, M. O. Delaney, C. J. Wiederholt, A. Samban-
Angew. Chem. Int. Ed. 2003, 42, 5882 –5885
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5885