Convenient synthesis of nucleoside and isonucleoside analogues

Article abstract of DOI:10.1021/ol050496v

(Chemical Equation Presented) A very simple methodology to stereoselectively achieve tricyclic isonucleosides (nucleobase = thymine, uracil, and 5-fluoruracil) and 3′-C-branched nucleosides (nucleobase = theophylline) was performed by means of a DBU-media

Full text of DOI:10.1021/ol050496v

ORGANIC
LETTERS
2005
Vol. 7, No. 11
2161-2164
Convenient Synthesis of Nucleoside and
Isonucleoside Analogues
Luis AÄ
lvarez de Cienfuegos,† Antonio J. Mota,‡ and Rafael Robles*
Departamento de Qu´ımica Orga´nica, Facultad de Ciencias, UniVersidad de Granada,
Campus de FuentenueVa 18071, Granada, Spain
rrobles@ugr.es
Received March 8, 2005
ABSTRACT
A very simple methodology to stereoselectively achieve tricyclic isonucleosides (nucleobase
branched nucleosides (nucleobase
)
thymine, uracil, and 5-fluoruracil) and 3′-C-
)
theophylline) was performed by means of a DBU-mediated addition process using a readily available
2-bromo sugar. The mechanism for these transformations implies the loss of both substituents at C-2 and C-3 on the sugar moiety, and
although it seems that DBU is probably involved, its involvement has not yet been ascertained. Cytosine did not react under these conditions.
Nucleoside derivatives are the predominant molecular target
prodrugs for the clinical therapy of AIDS.1 Unfortunately,
several problems concerning the development of resistance
against these drugs,2 their cytotoxicity,3 and their glycosidic
bond cleavage4 or enzymatic deamination5 processes have
led to the synthesis of new families of these compounds. In
this manner, the synthesis of nucleoside analogues in which
the sugar and/or the heterocyclic moiety have been modified
has received much attention as a consequence of their general
biological activity and the potential use of such molecules
as antiviral and antineoplasic therapeutic agents.6 In particu-
lar, isomeric nucleosides (isonucleosides) have revealed
significant antiviral activities,7 and their syntheses are
currently receiving much attention.8 In the past few years,
we have developed new methods for the stereoselective
synthesis of nucleosides by means of furanoid glycals9 or
1,2-thiocarbonate sugars.10 New strategies have also been
performed concerning the synthesis of isonucleoside ana-
logues and seco-nucleosides through exocyclic methylene
furanoid-like sugars.11 In this study, we present a simple way
† Current address: Prof. A. Klibanov Laboratory, Department of
Chemistry, Massachusetts Institute of Technology, Room 56-570, 77
Massachusetts Avenue, Cambridge, MA 02139-4307.
‡ Current address: Laboratoire de Chimie Quantique, Institut Le Bel,
Universite´ Louis Pasteur, 4, rue Blaise Pascal, 67000-Strasbourg, France.
(1) (a) Mitsuya, H.; Yarchoan, R.; Broker, S. Science 1990, 249, 1533-
544. (b) Herdewijn, P. A. M. M. AntiViral Res. 1992, 19, 1-14.
(2) (a) St. Clair, M. H.; Martin, J. L.; Tudor-Williams, G.; Bach, M. C.;
Vavro, C. L.; King, D. M.; Kellam, P.; Kemp, S. D.; Larder, B. A. Science
1991, 253, 1557-1559. (b) De Clercq, E. Biochem. Pharmacol. 1994, 47,
155-169.
(6) (a) Herdewijn, P.; Balzarini, J.; De Clercq, E.; Pauwels, R.; Baba,
M.; Broder, S.; VanderHeghe, H. J. Med. Chem. 1987, 30, 1270-1278.
(b) Mitsuya, H.; Yarchoan, R.; Broder, S. Science 1990, 249, 1533-1544.
(c) De Clercq, E. Design of Anti-AIDS Drugs; Elsevier: New York, 1990.
(7) (a) Huryn, D. M.; Sluboski, B. C.; Tam, S. Y.; Weigele, M.; Sim, I.;
Anderson, B. D.; Mitsuya, H.; Broder, S. J. Med. Chem. 1992, 35, 2347-
2354. (b) Verheggen, I.; Van Aerschot, A.; Van Meervelt, L.; Rozenski, J.;
Wiebe, L.; Snoeck, R.; Andrei, G.; Balzarini, J.; Claes, P.; De Clercq, E.;
Herdewijn, P. J. Med. Chem. 1995, 38, 826-835. (c) Nair, V.; Jahnke, T.
S. Antimicrob. Agents Chemother. 1995, 39, 1017-1029. (d) Yu, H.-W.;
Zhang, L.-R.; Zhou, J.-C.; Ma, L.-T.; Zhang, L.-H. Bioorg. Med. Chem.
1996, 4, 609-614. (e) Wang, J.-F.; Yang, X.-D.; Zhang, L.-R.; Yang, Z.-
J.; Zhang, L.-H. Tetrahedron 2004, 60, 8535-8546.
(3) (a) Chen, C.-H.; Cheng, Y.-C. J. Biol. Chem. 1989, 264, 11934-
11937. (b) Chen, C.-H.; Valquez-Padua, M.; Cheng, Y.-C. Mol. Pharmacol.
1991, 39, 625-628.
(4) (a)York, J. L. J. Org. Chem. 1981, 46, 2171-2173. (b) Nair, V.;
Buenger, G. S. J. Org. Chem. 1990, 55, 3695-3697.
(5) (a) Bloch, A.; Robins, M. J.; McCarthy, J. R. J. Med. Chem. 1967,
10, 908-912. (b) Cooney, D. A.; Ahluwalia, G.; Mitsuya, H.; Fridland,
A.; Johnson, M.; Hao, Z.; Dajal, M.; Balzarini, J.; Broder, S.; Johns, D. G.
Biochem. Pharmacol. 1987, 36, 1765-1768. (c) Nair, V.; Sells, T. B.
Biochim. Biophys. Acta 1992, 1119, 201-204.
10.1021/ol050496v CCC: $30.25
© 2005 American Chemical Society
Published on Web 04/30/2005

to access tricyclic 2,2′-anhydro-3-deoxy-3-isonucleosides and
their corresponding open forms (3-deoxy-3-isonucleosides),
along with a 3′-C-branched nucleoside derivative incorporat-
ing two units of theophylline.
Our key starting material is the 2-bromo sugar 1 (Scheme
1), which is readily available from the 1,2-diol 2 through
Table 1. Yields Obtained for the Addition Process via
Scheme 2
entry
nucleobase used
compound formed (yield %)a
1
2
3
4
5
thymine (Ty)
uracil (U)
5-fluoruracil (5-FU)
cytosine (Cy)
theophylline (Tph)
5T (71)
5U (73)
5F (58)
5C (na)b
6 (50)c
Scheme 1. Synthesis of the 2-Bromo Sugar 1
a Optimized yield. b No reaction was observed. c Mixture of isomers
(∼9:1).
the sugar in a highly stereoselective fashion, with the loss
of both substituents at C-2 and C-3, the anomeric position
remaining unreacted. On the other hand, when theophylline
(Tph) was used as a nucleobase,16 a different final product
was observed: a 3′-C-theophyllinyl nucleoside (6).
The structure of compounds 5 has been unambiguously
secured by X-ray diffraction for the thymine derivative 5T
(CCDC 250080, see Supporting Information), which presents
unique NMR patterns clearly recognizable on all the other
tricyclic derivatives. Unfortunately, we have not achieved
good crystals for 6 (or its benzoylated derivative 6Bz) in
attempts to ascertain its structure. Nevertheless, after an
extensive NMR study, we can conclude that structural
differences from compounds 5 are due to the insertion of a
new nucleobase unit at the anomeric position, not being a
tricycle derivative. This is possible since theophylline does
the formation of the 1,2-thiocarbonate sugar 3 and the
furanoid glycal 4.12 The latter quickly reacts with NBS in
aqueous THF9a,13 to afford compound 1 in high yield.14
When the bromo derivative 1 was treated with pyrimidinic
bases [thymine (Ty), uracil (U), 5-fluoruracil (5-FU), and
cytosine (Cy)] in the presence of DBU, the corresponding
tricyclic sugars 5 were formed (as tricyclic isonucleosides
analogues),15 with cytosine being the only exception for
which we were not able to obtain the corresponding iso-
nucleoside derivative (Scheme 2, Table 1).
(8) (a) Jung, M. E.; Nichols, C. J. J. Org. Chem. 1998, 63, 347-355.
(b) D´ıaz, Y.; Bravo, F.; Castillo´n, S. J. Org. Chem. 1999, 64, 6508-6511.
(c) Bera, S.; Nair, V. Tetrahedron Lett. 2001, 42, 5813-5815. (d) Lei, Z.;
Zhang, L.; Zhang, L.-R.; Chen, J.; Min, J.-M.; Zhang, L.-H. Nucleic Acids
Res. 2001, 29, 1470-1475. (e) Chattopadhyay, A.; Salaskar, A. J. Chem.
Soc., Perkin Trans. 1 2002, 785-789. (f) Nair, V.; Chun, B.-K. ArkiVoc
2003, i, 9-21.
(9) (a) Robles, R.; Rodr´ıguez, C.; Izquierdo, I.; Plaza, M. T.; Mota, A.
Tetrahedron: Asymmetry 1997, 8, 2959-2965. (b) AÄ lvarez de Cienfuegos,
L.; Mota, A. J.; Rodr´ıguez, C.; Robles, R. Tetrahedron Lett. 2005, 46, 469-
473.
Scheme 2. Synthesis of Tricyclic Isonucleosides 5 and
3′-C-Branched Nucleoside 6
(10) (a) AÄ lvarez de Cienfuegos, L.; Rodr´ıguez, C.; Mota, A. J.; Robles,
R. Org. Lett. 2003, 5, 2743-2745. (b) Robles, R.; Rodr´ıguez, C.; AÄ lvarez
de Cienfuegos, L.; Mota, A. J. Tetrahedron: Asymmetry 2004, 15, 831-
838.
(11) Robles, R.; Izquierdo, I.; Rodr´ıguez, C.; Plaza, M. T.; Mota, A. J.;
AÄ lvarez de Cienfuegos, L. Tetrahedron: Asymmetry 2002, 13, 399-405.
(12) Robles D´ıaz, R.; Rodr´ıguez Melgarejo, C.; Izquierdo Cubero, I.;
Plaza Lo´pez-Espinosa, M. T. Carbohydr. Res. 1997, 300, 375-380.
(13) Robles-D´ıaz, R.; Rodr´ıguez Melgarejo, C.; Plaza Lo´pez-Espinosa,
M.-T.; Izquierdo Cubero, I. J. Org. Chem. 1994, 59, 7928-7929.
(14) Compound 4 (1 mmol) in THF (10 mL) was treated with NBS (1.1
mmol) and H2O (1 mL). After 10 min, TLC (ether) revealed that the reaction
was finished, showing a lower-running product. The mixture was concen-
trated under reduced pressure, and the residue was purified by column
chromatography (1:1 ether/hexane), yielding 1 (0.85 mmol, 85%) as a
mixture of anomers.
(15) To a solution of the corresponding nucleobase (Ty, U, 5-FU, or
Tph, 2 mmol) in dry DMF (15 mL) containing DBU (3.2 mmol) was added
1 (1 mmol), and the mixture was stirred for 10 min. After this time, TLC
(20:1 Cl2CH2/MeOH) showed that the reaction was finished (a lower-running
product appeared). An aqueous solution of KHSO4 (10%) was then added
until the pH became slightly acidic. The solution was evaporated under
reduced pressure, and the residue was extracted with Cl2CH2-H2O. The
organic layer was dried (MgSO4 anhydrous), filtered, and evaporated under
reduced pressure until dryness. The residue was purified by column
chromatography (20:1 Cl2CH2/MeOH) to yield respectively 5T, 5U, 5F,
and 6, the latter being a ∼9:1 mixture of isomers (see Table 1).
(16) Theophylline was chosen as a base because the absence of hydrogens
should prohibit the tricycle formation, as it has been experimentally proved.
It is noteworthy that nucleobases do not have a preference
for either the anomeric position or the C-2 position. These
positions are a priori highly sensitive to a possible nucleo-
philic attack. Conversely, they enter at the C-3 position on
2162
Org. Lett., Vol. 7, No. 11, 2005

Scheme 3. Proposed Mechanism to Obtain Derivatives 5
not allow the tricyclic formation (it has no hydrogens on
clarify it (see Supporting Information). As a result, an
attempted mechanism based on a cascade reaction, in which
DBU is actively involved, is proposed in Scheme 3. It should
be noted that DBU not only acts to catalyze the process but
also controls the stereochemistry of the addition. The
regiochemistry, however, is controlled by the intermediate
1b. Thus, the positive charge is necessarily placed on C-3
in order to avoid the vicinity of the δ+ anomeric carbon. All
these polar species are assumed to be significantly stabilized
by the solvent (DMF).
We have also carried out several modifications over the
tricyclic derivative 5T oriented to the synthesis of its
corresponding 3-deoxy-3-isonucleoside by regio- and ste-
reospecific ring-opening of the tricyclic system (Scheme 4).
In this manner, derivatives 7T, 8T, and 9T have been
obtained following common protocols (see Supporting
Information). It should be noted that the 7T f 8T transfor-
mation takes place with retention of the configuration on
C-2′, because it does not go through an SN2 process at this
position but instead goes through an OH- addition (alkaline
hydrolysis) at C-2 of the pyrimidine.18
In summary, we have developed a very simple methodol-
ogy to stereoselectively achieve tricyclic isonucleosides and
3′-C-branched nucleoside derivatives by means of a DBU-
mediated addition process using the readily available 2-bro-
mo sugar 1. The reaction conditions are soft, and we can
quickly provide the valuable title compounds in a few steps.
The stereochemistry of the tricyclic derivatives has been
ascertained by X-ray diffraction for the thymine derivative
and correlated to the other derivatives by finding similar
NMR patterns. The regio- and stereospecific ring-opening
of the tricyclic derivative 5T afforded the corresponding
3-deoxy-3-isonucleoside.
the nitrogens). The stereochemistry at C-3′ for the major
1
isomer in 6 can be deduced from the H NMR couplings,
since the key J values (in hertz) are JH2′,H3′ ) 6.0 (7.8 in
6Bz) and JH3′,H4′ ) 8.8, both indicative of an all-2,3,4-cis
arrangement. At the anomeric position, a JH1′,H2′ value of 1.5
Hz points to an R stereochemistry. These conclusions follow
the general behavior of the other bases employed. Neverthe-
less, with regard to the minor isomer, the 1H NMR coupling
constants are not conclusive, and it cannot be concluded that
this compound corresponds to the â anomer. In any case,
AM1 semiempirical calculations17 show that the R anomer
is thermodynamically more stable than the â isomer by 3.5
kcal/mol (see Supporting Information). Moreover, calcula-
tions also point out a theophylline unit bonding the sugar
by the sterically less hindered N-7 position (preferred by 14.7
kcal/mol according to AM1). The isomeric N-9 bonding
generates a bigger steric hindrance caused by the methyl
group at N-3 (see Supporting Information).
The mechanism involved in these transformations has not
been ascertained, and other experiments have been carried
out with derivatives chemically related with 1 in order to
Scheme 4. Chemical Modifications on the Tricyclic Sugar
Derivative 5T
Acknowledgment. We thank the Universidad de Granada
for the grant to L. AÄ lvarez de Cienfuegos and for providing
funds for this work, along with the Junta de Andaluc´ıa for
the research contract to A. J. Mota.
(17) Hyperchem, release 7.5; Hypercube, Inc.: Gainesville, FL.
(18) (a) Thedford, R.; Fleysher, M. H.; Hall, R. H. J. Med. Chem. 1965,
8, 486-491. (b) Robles, R.; Rodr´ıguez, C.; Izquierdo, I.; Plaza, M. T.; Mota,
A.; AÄ lvarez de Cienfuegos, L. Tetrahedron: Asymmetry 2000, 11, 3069-
3077.
Org. Lett., Vol. 7, No. 11, 2005
2163

Supporting Information Available: Complete physical
data (IR, RD, HMRS, and 1H and 13C NMR) of com-
pounds 1, 5T, 5U, 5F, 6, 6Bz, 7T, 8T, and 9T; the
X-ray structure for derivative 5T; AM1 structures, energies,
and Cartesian coordinates of the calculated structures;
additional experiments carried out in order to ascertain the
mechanism; and the protocols followed in the 5T f 9T
transformations. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL050496V
2164
Org. Lett., Vol. 7, No. 11, 2005