Sulfanyl radical promoted C4′-C5′ bond scission of 5′-oxo-3′,4′-didehydro-2′,3′-dideoxynucleosides

Article abstract of DOI:10.1039/b609995e

The synthesis of C5'-aldehydes 1a,b and preliminary results concerning their reactivity under radical conditions was analyzed. Treatment of C5'-aldehydes 4a, b under mildly basic conditions led to the formation of 3' 4' -didehydroaldehydes 1a, b and furfu

Full text of DOI:10.1039/b609995e

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Sulfanyl radical promoted C4ꢀ–C5ꢀ bond scission of
5ꢀ-oxo-3ꢀ,4ꢀ-didehydro-2ꢀ,3ꢀ-dideoxynucleosides†
Maria Luisa Navacchiaa and Pier Carlo Montevecchi*b
Received 12th July 2006, Accepted 4th September 2006
First published as an Advance Article on the web 12th September 2006
DOI: 10.1039/b609995e
Treatment of C5ꢀ-aldehydes 4a,b under mildly basic condi-
tions leads to the formation of 3ꢀ,4ꢀ-didehydroaldehydes 1a,b
and furfural 2. Sulfanyl radical addition to 1a,b eventually
gives rise to the lactones 11a,b through C4ꢀ–C5ꢀ bond scission
of the 1,2-dioxetane intermediates 10a,b.
5ꢀ-Oxo-derivatives represent an interesting class of nucleoside
analogs from both a synthetic1–3 and biological4,5 point of view.
The biological relevance of the 5ꢀ-oxonucleosides came to light
when C5ꢀ-aldehyde termini A were found as oxidatively damaged
products of DNA, deriving from strand breakage mediated
by natural or chemical agents, such as enediynes4 and metal
porphyrins.5
Pratviel et al.5a characterized the 3ꢀ,4ꢀ-unsaturated-5ꢀ-aldehydes
of 2ꢀ-deoxyadenosine and thymidine 1 among products derived
from metalloporphyrin-mediated DNA cleavage after thermal
treatment. These aldehydes have been shown to be the precur-
sors of furfural 2, which was identified as the eventual sugar
Scheme 1 Reagents and conditions: i: NCS, 2-mercaptoethanol; ii:
degradation product5b (Scheme 1). Earlier, the 3ꢀ,4ꢀ-unsaturated-
5ꢀ-aldehyde of thymidine had been suggested by Goldberg et al.4a
as a plausible intermediate for the formation of furfural 2 by nu-
clease/basic treatment of products derived from neocarzinostatin-
mediated DNA strand breakage (Scheme 1).
Mn-TMPyP/KHSO5; iii: enzyme digestion followed by basic treatment
at pH 12; iv: 90 ◦C.
unreported aldehyde 4a was obtained in 80% yield as a 10 : 90
mixture of 4a and its hydrated form 4aꢀ by silica gel column
chromatography of the reaction mixture. Similarly, the previously
reported,2b but neither isolated nor characterized, C5ꢀ-aldehyde
4b was obtained in 90% yield as an 80 : 20 mixture of 4b and its
hydrated form 4bꢀ.
The oxidation of 3b was also performed using a modified Swern
method, as reported by Matsuda.2b Work-up and flash column
chromatography led to a 70 : 30 mixture of 4b and 4bꢀ mixture in
75% yield.
The subsequent hydro-benzoyloxy elimination leading to 1a,b
was performed under mildly basic conditions by treatment of
a dichloromethane solution of the mixture of the appropriate
aldehyde 4a,b and its hydrated form 4aꢀ,bꢀ with 4 molar equivalents
of TEA at room temperature. In both cases the crude mixture was
chromatographed on a silica gel column to give the 3ꢀ,4ꢀ-didehydro-
5ꢀ-aldehydes 1a,b in excellent yield (85%).
However, apart from the above-cited reports, the 3ꢀ,4ꢀ-
didehydro-2ꢀ,3ꢀ-dideoxy-5ꢀ-aldehydes 1 remain virtually uninvesti-
gated. Therefore, we considered that understanding their chemical
behaviour could be of interest to chemists and biochemists, in
order to elucidate the course of potential DNA damage. For
this purpose, we planned a synthetic route to the C5ꢀ-aldehydes
1 and started to study their chemical reactivity. In this paper
we report the synthesis of C5ꢀ-aldehydes 1a,b and preliminary
results concerning their reactivity under radical conditions. We
found that sulfanyl radical addition to the C3ꢀ atom under
oxygenated conditions eventually led to products deriving from
an unprecedented, and unexpected, C4ꢀ–C5ꢀ bond scission.
The 3ꢀ,4ꢀ-didehydro-2ꢀ,3ꢀ-dideoxy-5ꢀ-aldehydes 1a,b were syn-
thesized as follows. The O3ꢀ-benzoyl derivatives 3a6 and 3b2b
were prepared according to known procedures. The C5ꢀ-aldehydes
4a,b were obtained by oxidation of 3a,b through a DMSO-based
method using a modified Moffatt procedure.7 The previously
The C5ꢀ-aldehyde 1b could also be obtained in a one-pot process
in good yield (65% isolated yield) by oxidation of the adenosine
derivative 3b under Swern conditions,2b followed by final treatment
of the reaction mixture with 6 molar equivalents of TEA.
The hydro-benzoyloxy elimination reaction was found to be the
crucial step. In fact, the yield of the aldehydes 1a,b was strongly
dependent on the reaction time of the C5ꢀ-aldehydes 4a,b with
TEA. Prolonged reaction time caused a progressive decrease of
the yield of 1a,b with concomitant formation of furfural 2. The
formation of 1a,b and 2 vs. time was monitored by 1H NMR. As
aISOF, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, 40129,
Bologna, Italy
bDipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna,
Viale Risorgimento 4, 40136, Bologna, Italy. E-mail: montevec@
ms.fci.unibo.it
† Electronic supplementary information (ESI) available: NMR and
MS spectral data of compounds 1a,b, 4a,b, and 11a,b. See DOI:
10.1039/b609995e
3754 | Org. Biomol. Chem., 2006, 4, 3754–3756
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
shown in Fig. 1, the disappearance of starting 4a,b was complete
within 10 min, leading to 1a,b, and thereafter furfural 2 appeared
at the expense of the aldehyde 1a,b at a rate independent of the
nature of the base unit.
thymine as the major product, and a 50 : 50 3ꢀS/3ꢀR diastereomeric
mixture of the lactone 11a in 35% yield. Analogously, column
chromatography of the reaction mixture from1b furnished adenine
as major product, as well as a 70 : 30 3ꢀS/3ꢀR diastereomeric
mixture of the lactone 11b in 30% yield (Scheme 3). The absolute
configuration of the diastereomeric lactones 11a and 11b was
established on the basis of NOE experiments.
Fig. 1 Formation of products 1a,b and 2 vs. time from reaction of
C5ꢀ-aldehydes 4a,b with TEA.
From a mechanstic standpoint, the formation of furfural 2
from aldehydes 4a,b is easily explained through two consecutive
base-catalysed elimination (E2cb) reactions. The formation of the
carbanion/enolate 5a,b followed by elimination of the benzoate
ion leads to the aldehydes 1a,b. Further elimination of the H2ꢀ
proton gives the resonance-stabilised ion 6a,b, from which furfural
2 arises by loss of the base ion (Scheme 2). Interestingly, b-hydro-
acyloxy eliminations are well known under pyrolytic conditions,
whereas they are unprecedented under basic conditions.
Scheme 3 Reagents and conditions: i: PhSH, AIBN or O2, PhF, 82 ◦C.
The role played by dioxygen in the formation of 11a,b was
proved by reacting 1b with benzenethiol in refluxing fluorobenzene
in the presence of equimolar amounts of AIBN. Under these
hypoxic conditions the lactone 11b was formed in poor yields
(<5%), with adenine as the main reaction product.
On the basis of these experimental results, we can propose the
following mechanism for the formation of 11a,b. Benzenesulfanyl
radicals, produced by hydrogen atom abstraction either by cyano-
isopropyl radicals or dioxygen, add to the C3ꢀ carbon atom in
a regioselective manner. In principle, the resulting radical 7a,b
could react in two different ways: i) hydrogen atom abstraction
from benzenethiol; however, no hydro-sulfenylation adduct was
detected, not even when the reaction of 1b was repeated in the
presence of a large excess (5 molar equivalents) of benzenehiol;9
ii) trapping by dioxygen to give peroxyl radicals, from which the
hydroperoxide 8a,b can be formed by hydrogen atom abstraction
from benzenethiol. Actually, this reaction was expected, since
peroxyl radicals, and hydroperoxides, have been claimed as
intermediates in DNA strand breakage promoted by formation
of C4ꢀ-radicals under aerobic conditions and in the presence of
thiol. It has been reported that subsequent O–O bond scission
leads to oxy radicals, responsible for the formation of the observed
decomposition products.10
Scheme 2 Reagents and conditions: i: DCC, Cl2CHCOOH, DMSO, ii:
(COCl)2, DMSO, TEA, CH2Cl2; iii: TEA, CH2Cl2.
In order to explore the behaviour of the 3ꢀ,4ꢀ-unsaturated C5ꢀ-
aldehydes under radical stress we first reacted the aldehydes 1a,b
with benzenethiol under radical conditions. The reactions were
carried out8 with 1.2 molar equivalents of benzenethiol in a
sealed tube at 80 ◦C in the presence of dioxygen and in the
presence or absence of 0.2 molar equivalents of AIBN as initiator.
Complete disappearance of 1a,b occurred within 2 h. Silica gel
column chromatography of the reaction mixture from 1a furnished
We inferred that hydroperoxide 8a,b, in addition to the O–
O bond scission leading to the oxy radical 9a,b, and then to
the free base and unidentified sugar fragments,11 can give the 3-
hydroxy-1,2-dioxetane 10a,b through nucleophilic addition to the
a-carbonyl carbon atom.
1,2-Dioxetanes, an interesting class of chemiluminescent prod-
ucts, are known to undergo C–C and O–O bond scission leading
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3754–3756 | 3755
©

View Article Online
to two carbonyl compounds through a [2 + 2] retro-Diels–Alder
reaction.12 In this way, the lactones 11a,b can be easily formed
from 10a,b with concomitant elimination of formic acid. To our
knowledge this is the only example of a C4ꢀ–C5ꢀ bond scission
in nucleoside chemistry, even though intramolecular addition of
a hydroperoxide to the carbonyl carbon atom of an a-formyl
group, followed by fragmentation of the resulting 3-hydroxy-1,2-
dioxetane, was reported some time ago.13
It is worth noting that the lactone 11b, in contrast to 11a, was
formed with 70 : 30 stereoselectivity. This finding suggests that
benzenesulfanyl radicals approach the C3ꢀ–C4ꢀ double bond of 1a
nonstereoselectively, while they approach 1b preferentially from
the side opposite to the adenine base. This different behaviour
might be due to the greater steric hindrance of the adenine unit
with respect to the thymine.
5 (a) G. Pratviel, M. Pitie´, C. Perigaud, G. B. J. Gosselin and B. Meunier,
J. Chem. Soc., Chem. Commun., 1993, 149; (b) G. Pratviel, J. Bernadou
and B. Meunier, Angew. Chem., Int. Ed. Engl., 1995, 34, 746; G. Pratviel,
M. Pitie´, J. Bernadou and B. Meunier, Angew. Chem., Int. Ed. Engl.,
1991, 30, 702.
6 V. Gotor and F. Moris, Synthesis, 1992, 626.
7 J. A. Montgomery and H. J. Thomas, J. Org. Chem., 1981, 46,
594.
8 Method A: A 10.0 mM solution of the appropriate C5ꢀ-aldehyde 1a or
1b (0.30 mmol), benzenethiol (0.60 mmol, 0.06 mL) and AIBN (50 mg,
0.30 mmol) in fluorobenzene (30 mL) was kept in a sealed tube in a
thermostated bath at 82 ◦C for 2 h. Method B: A 10.0 mM solution of
the appropriate C5ꢀ-aldehyde 1a or 1b (0.30 mmol) and AIBN (50 mg,
0.30 mmol) in fluorobenzene (30 mL) was refluxed under argon for
10 min, and then benzenethiol (0.60 mmol, 0.06 mL) was added. The
resulting solution was refluxed for 2 h. In both cases the solvent was
removed under reduced pressure, the residue was analysed by HPLC-
1
MS and H NMR and then chromatographed on a silica gel column
by gradual elution with hexane–ethyl acetate.
9 We might explain this finding by assuming that polar factors could
play a determining role, since both sulfanyl radicals and radicals 7a,b
are expected to be electrophilic in character. In fact, it is generally
assumed that the hydrogen atom transfer reaction is disfavored when
the attacking radical and the displaced radical show the same philicity.
See: B. P. Roberts and A. J. Steel, J. Chem. Soc., Perkin Trans.
2, 1994, 2155; B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1996,
2719.
10 W. K. Pogozelski and T. D. Tullius, Chem. Rev., 1998, 98, 1089.
11 As far as the formation of the free base is concerned, we cannot
exclude the possibility that the free base and the unidentified sugar
fragments arise from hydroperoxides 8 through initial O–O bond
scission in competition with the intramolecular nucleophilic addition
to the a-carbonyl atom. However, the finding that adenine is the
almost exclusive product when 1b was reacted under hypoxic conditions
suggested that another route is available for its formation. Since
both aldehydes 1a and 1b were found to be thermally stable, we
could infer that the free base might arise from radical 7a,b through
some decomposition reaction in competition with the trapping by
dioxygen.
This work was supported by the “Ministero della Ricerca
Scientifica e Tecnologica”, MURST (Rome).
References and notes
1 (a) M. L. Navacchia, C. Chatgilialoglu and P. C. Montevecchi, J. Org.
Chem., 2006, 71, 4445; (b) M. L. Navacchia, A. Manetto, P. C.
Montevecchi and C. Chatgilialoglu, Eur. J. Org. Chem., 2005, 4640;
(c) A. Romieu, D. Gasparutto and J. Cadet, J. Chem. Soc., Perkin Trans.
1, 1999, 1257; (d) T. Sugawara, B. A. Otter and T. Ueda, Tetrahedron
Lett., 1988, 29, 75.
2 (a) A. Kers, T. Szabo and J. Stawinski, J. Chem. Soc., Perkin Trans. 1,
1999, 2585; (b) S. Shuto, M. Kanaaki, S. Ichikawa, N. Minakawa and
A. Matsuda, J. Org. Chem., 1998, 63, 746; (c) B. Giese, P. Erdmann,
T. Schafer and U. Schwitter, Synthesis, 1994, 1310; (d) A. K. Saha, W.
Schairer, C. Waychunas, C. V. C. Prasad, M. Sardaro, D. A. Upson and
L. I. Kruse, Tetrahedron Lett., 1993, 38, 6017.
3 S. Kralikova, M. Budesinky, M. Masojidkova and I. Rosenberg,
Tetrahedron, 2006, 62, 4917.
4 (a) L. S. Kappen, I. H. Goldberg and J. M. Liesch, Proc. Natl. Acad.
Sci. U. S. A., 1982, 79, 744; (b) L. S. Kappen and I. H. Goldberg,
Biochemistry, 1983, 22, 4872.
12 M. Matsumoto, J. Photochem. Photobiol., C, 2004, 5, 27.
13 N. Duran, O. M. M. Faria Oliviera, M. Haun and G. Cilento, J. Chem.
Soc., Chem. Commun., 1977, 442.
3756 | Org. Biomol. Chem., 2006, 4, 3754–3756
This journal is
The Royal Society of Chemistry 2006
©