New conformationally locked bicyclic N,O-nucleoside analogues of antiviral drugs

Article abstract of DOI:10.1016/j.bmcl.2004.11.048

In order to obtain rigidity within the sugar moiety of nucleosides, the bicyclic pyrimidine derivatives of N,O-isoxazolidines were designed and synthesized by using 1,3-dipolar cycloaddition of Δ¹- pyrrolidine-1-oxide and the appropriate vinyl-nucleobases.

Full text of DOI:10.1016/j.bmcl.2004.11.048

Bioorganic & Medicinal Chemistry Letters 15 (2005) 545–550
New conformationally locked bicyclic N,O-nucleoside analogues
of antiviral drugs
a,
Antonio Procopio, Stefano Alcaro, Antonio De Nino, Loredana Maiuolo,
a
b
b
*
a
Francesco Ortuso and Giovanni Sindona
b
a
Dipartimento di Scienze Farmaco-Biologiche, Universit a` della Magna Graecia, Complesso Nin `ı Barbieri,
8021 Roccelletta di Borgia (Cz), Italy
Dipartimento di Chimica, Universit a` della Calabria, Ponte Bucci, cubo 15C, 87036 Arcavacata di Rende (Cs), Italy
8
b
Received 21 October 2004; revised 17 November 2004; accepted 18 November 2004
Available online 23 December 2004
Abstract—In order to obtain rigidity within the sugar moiety of nucleosides, the bicyclic pyrimidine derivatives of N,O-isoxazoli-
1
dines were designed and synthesized by using 1,3-dipolar cycloaddition of D -pyrrolidine-1-oxide and the appropriate vinyl-
nucleobases.
ꢀ
2004 Elsevier Ltd. All rights reserved.
9
1. Introduction
side analogues has prompted the exploitation of
strategies aiming at locking the puckering of the fura-
nose ring into one of the two rotamers. Many anti-
HIV nucleoside analogues are, in fact, active when their
Modified nucleosides containing hydroxyl handles for
their enzymatic phosphorylation in vivo, such as AZT,
ddC, d4T, etc., are currently employed in the multi-drug
protocols adopted in therapeutic treatment of HIV
sugar ring conformational equilibria are centered
0
around the ( E, S-type) C 3 -exo conformation. Our ori-
3
1
infections.
ginal strategy for the formation of isoxazolidinyl nucle-
osides is based on the 1,3-dipolar cycloaddition of
2
A new line of research has been recently opened in this
field, which considers the replacement of the modified ri-
azomethinoxides on vinyl-nucleobases achievable by a
2
straightforward procedure.
,10
2
–5
bose with an isoxazolidine nucleus.
The structural feature of the bicyclic isoxazolidine
1
Preliminary in vivo tests have shown a promising bio-
logical activity for some of the members of this new fam-
ily of potential antiviral drugs. The antiviral activity
model (1, Fig. 1), obtained from D -pyrrolidine-1-oxide
and ethene, is represented by the geometric constraints
preventing cis–trans interconversion by nitrogen inver-
6
1
1
exhibited by nucleoside analogues has been correlated
to some extent to preferred conformations achieved by
the drug in the formation of the enzyme–inhibitor com-
plex, which temporarily inhibits the DNA strand prolif-
sion. Moreover, the presence of the second five-
membered ring fused to the isoxazolidine moiety induces
a restricted conformational mobility. If a nucleobase
7
eration. Unmodified nucleosides rings, as described by
8
the pseudorotation cycle, exist in either S-type (2 -
0
B
O
O
0
0
0
endo/3 -exo) or N-type (2 -exo/3 -endo) conformations.
The observation that reverse transcriptase is able to dis-
criminate between two conformationally locked nucleo-
N
N
(
2a-c)
(
1)
2
2
2
a; B= thymine
b; B= uracil
c; B= cytosine
Keywords: Nucleoside analogues; Vinylpyrimidine nucleobases; N,
O-Isoxazolidines.
*
Corresponding author. Fax: +39 0961391143; e-mail: procopio@
unicz.it
Figure 1.
0
960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.11.048

546
A. Procopio et al. / Bioorg. Med. Chem. Lett. 15 (2005) 545–550
was inserted in this molecular framework it could be ex-
pected that the modified nucleosides 2a–c show a pecu-
liar structure–function relationship similar to that
glycoside moiety. Solvent effects have been considered
1
3
using the GB/SA water method for all calculations
1
4
as implemented in MacroModel ver. 7.2. Each Monte
Carlo generated conformer has been energy minimized
9
already experienced with other nucleoside analogues.
1
5
with the force field MMFFs by means of 5000 itera-
tions of the Polak Ribiere conjugate gradient algorithm
adopting a convergence criterion equal to 0.01 kcal/
The synthesis of the three pyrimidine nucleosides 2a–c
was therefore designed to exploit the effect of the fused
ring on the isoxalidine moiety on the conformational
mobility of the new compounds compared with the
monocyclic analogues.
˚
A mol. In order to consider in the following analysis
only the most representative conformations, each gener-
ated structure has been compared to the others includ-
ing both geometric and energetic criteria. Conformers
with an internal energy difference lower than 0.1 kcal/
mol and with a root mean square deviation, computed
2
. Results and discussions
˚
on the heavy atomic coordinates, lower than 0.25 A
The design of isoxazolidine nucleosides 2a–c as novel
heterocyclic nucleoside analogues was based on the fol-
lowing considerations: (1) the use of rigid isoxazolidine
pyrrolidine bicyclic heterocycles as surrogates for con-
formationally restricted deoxyriboses; (2) the basicity
of the hydroxylamino nitrogen atom should confer en-
hanced stability to glycosidic linkage in acid media.
have been considered identical. The large average num-
ber of duplicate structures, in all cases higher than 147,
allowed us to estimate the conformational space of all
0
compounds widely explored founding both south (C2
0
0
0
endo–C3 exo) and north (C3 endo–C2 exo) conform-
ers. A structural comparison between them is reported
in Figures 2 and 3, respectively, as global minimum en-
ergy south and high energy north conformations of 2a–c
compounds.
In a comparative structural study performed on hexahy-
1
1
dropyrrolo[1,2-b]isoxazolidine it was concluded that a
fused pyrrolidine ring can impart significant rigidity to
the isoxazolidine portion on the molecule to the extent
that the resulting nucleoside analogues indeed show
identical conformations in the solid state and in solu-
tion. For this reason, we have conducted a conforma-
tional analysis for evaluating the reverse transcriptase
The conformer population has been evaluated by means
of the Boltzmann analysis by using the MOLINE ther-
1
6
modynamic module. With the purpose to validate
our conformational model, the 2a–c global minimum
energy conformers have been submitted to molecular
dynamic simulations. We performed two runs, respec-
tively, at 300 K (MD300) and at 500 K (MD500), com-
puted for 5 ns using a time step equal to 1.5 fs. During
these two calculations, v and m0–4 torsions, as defined
in Scheme 1, have been monitored.
(
RT) potential binding properties starting with conform-
ationally locked nucleoside probes whose structural fea-
tures in the solid state and in solution are virtually
identical.
We have designed the bicyclic nucleoside analogues 2a–
c, which are promising precursors for novel types of
therapeutic nucleoside derivatives. The sugar moiety in
The average value of v has been used to locate the posi-
tion of the base with respect to the glycoside moiety
while m0–4 to compute the phase angle of pseudorota-
8
2
a–c consists of a hexahydropyrrolo[1,2-b]isoxazolidine
tion P. In Table 1 are summarized the conformational
ring system, with the isoxazolidine part conformation-
ally restricted by the fused pyrrolidine ring. With the
aim to identify most stable conformations, the structural
properties of bicyclic pyrimidine derivatives 2a–c were
investigated by means of molecular modeling studies.
properties of 2a–c compounds, glycoside moieties ex-
pressed, in the Monte Carlo case, as sum of the Boltz-
mann population for each north (N) and south (S)
conformer and, in the molecular dynamics cases, as P.
The geometric properties of compounds 2a–c have been
evaluated considering these two descriptors that showed
a very low variability indicating, for all molecules, one
highly constrained conformation. Base rotamers, con-
trolled by the v torsion, have been widely explored dur-
ing molecular dynamics simulation of both (see Fig. 4).
12
All molecules have been built using the Maestro Linux
graphical user interface and submitted to 2000 steps of
Monte Carlo conformational search taking into account
the nucleobase rotation and the interconversion of the
Figure 2. Global minimum energy south conformers of 2a–c structures.

A. Procopio et al. / Bioorg. Med. Chem. Lett. 15 (2005) 545–550
547
Figure 3. Energy minimized north conformers of 2a–c compounds. Relative energies, in kJ/mol, of these structures are: 2a 10.85, 2b 14.86 and 2c
0.77.
1
v0
onto the pyrimidine ring and that of the isoxazolidine
moiety.
v4
O
base
N
χ
v3
Taking into account the bicyclic structure of the glyco-
side moiety, the analysis of P clearly indicated the
v1
v2
0
dependence of this parameter on the C4 stereochemis-
try, which, in our compounds, was related to the ano-
Scheme 1. Notation of the rotatable bonds used for conformational
analysis.
0
meric carbon C2 configuration due to synthetic
pathway. The puckering interconversion was observed
only for not populated high energy conformations.
Table 1. Boltzmann probability in percent at room temperature for N
and S conformers and the average P values
In conclusion, compounds 2a–c showed quite similar
structural properties. The molecular complication of
the glycoside locked this moiety in a S conformation
widely reducing the probability of puckering intercon-
version also in extreme conditions (i.e., molecular
dynamics simulation at 500 K). This observation indi-
cates that these molecules have a strongly conserved
DNA-like shape and therefore they could be considered
for further optimization studies with the aim to design
new or more potent HIV reverse transcriptase
inhibitors.
Compd
Monte Carlo
MD300
MD500
N
S
P
P
2
2
2
a
b
c
5.44
5.53
4.10
94.56
94.47
95.90
178.89ꢁ
180.27ꢁ
179.00ꢁ
180.10ꢁ
178.83ꢁ
184.49ꢁ
The syn conformation was energetically less stable than
the anti one, likely due to the electrostatic and Van der
Waals repulsion between the sp oxygen in position 2
2
2a MD300
2b MD300
2c MD300
2a MD500
2b MD500
2c MD500
2
1
70
80
N
N
O
anti
O
N
9
0
N
0
O
N
syn
O
N
-
90
5
0
500
950
1400
1850
2300
2750
3200
3650
4100
4550
5000
picoseconds
Figure 4. v rotation during MD300 and MD500 simulations. anti- and syn-conformations are, respectively, depicted as white and grey areas.

548
A. Procopio et al. / Bioorg. Med. Chem. Lett. 15 (2005) 545–550
O
3
H C
O
B
toluene dry
110 ˚C
N
+
O-
N+
HN
B
5
4
H
6
3
NH
3
4a-c
2a-c
8%
1
N
2
O
O
NH
2
CH
3
O
B =
HN
N
2'
O
N
O
N
b
O
N
c
N
H
H
H
O
8
%
a
H
4
'
3'
H
Scheme 2. Synthetic route to 2a–c.
7
'
H
5'
H
1
1%
Based on molecular modeling the compounds 2a–c were
predicted to exist in a S-type conformation (correspond-
ing to E, C3 -exo in the ribo nucleoside analogues). In
3
order to confirm the assumed rigidity the modeled mol-
ecules were synthesized and a possible equilibrium be-
tween different conformational states were investigated
by analysis of proton NMR spectra.
6
'
0
H
H
H
0
0
Figure 5. (2 RS,4 SR)-1-(Hexahydropyrrolo[1,2-b]isoxazolyl) 5-methyl-
H-pyrimidine-2,4-dione (2a).
1
Any changes was observed for the relevant coupling
constants or chemical shifts between the low and high
temperature spectra indicating that the isoxazole ring
in these compounds does not appears to be involved in
a conformational N « S equilibrium in solution as is
the case for conventional nucleosides.
Here we describe a convenient first synthesis of 2a–c and
also discuss their conformation. The 1,3-cycloaddition
reaction of D -pyrrolidine-1-oxide 3 with N1-vinyl
pyrimidine derivatives 4a–c gave the bicyclic adducts
1
2
a–c in good yields (Scheme 2).
The configuration and conformation of the bicyclic
nucleosides 2a was evaluated by H NMR and NOE
1
The synthesized three modified nucleosides were prelim-
inarily screened in a cell-based system (VERO cells) for
their ability to inhibit HSV-1 replication (Fig. 6). In
addition, cytotoxicity was evaluated in parallel in an
MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-
phenyl)-2-(4-sulfophenyl)-2H-tetrazolium]-based assay
experiments. The spectrum evidenced the presence of a
NOE effect between the H-6 of the thymine moiety
0
and H-4 (8%/8%) indicating an anti conformation of
0
the thymidine ring and C4 -exo conformation of the
3
isoxazolidine ring, which resemble the E (C3 -exo) con-
0
(
Table 2).
formation in the b-D-ribo configurated nucleoside ana-
logues. In addition, mutual NOEs between H-4 and
0
0
0
0
One of these nucleosides (2b) displayed an interesting
inhibition of HSV-1 replication in the concentration
H-3 a, one of two hydrogen nuclei in 3 , (8%/8%) and
0
0
H-2 , and H-3 b, the other hydrogen nucleus in 3 ,
11%/11%) supported the assigned structure and
conformation.
(
1
00
9
0
0
The absence of any coupling constant between H-4 of
0
80
7
the isoxazolidine ring and H-2 also indicates the confor-
0
mation of the isoxazolidine ring to be exclusively S-type.
So 2a was assigned as derived from an exo approach of
the dipolarophile to the nitrone furnishing almost exclu-
sively the 2 RS, 4 SR enantiomer where H-2 and H-4
are anti each other (Fig. 5). The constant reproducibility
of the 1,3-dipolar process and the homogeneity of the
H NMR data permit us to consider the extension of
these conclusions to the other two cycloadducts 2b and
2
0
1
2
4
8
µΜ
60
50
00 µΜ
00 µΜ
00 µΜ
00 µΜ
0
0
0
0
4
3
2
1
0
0
0
0
0
1
c.
2
a
2b
2c
For the purpose of detecting a possible equilibrium be-
tween different conformational states, proton NMR
spectra of thymine derivative 2a was recorded in
DMSO-d6 over the temperature range of 20–80 ꢁC.
Figure 6. Antiviral efficacy of 2a–c in VERO cells infected with HSV-1
in the presence of indicated concentration of the three compounds. The
results are expressed as mean values along with those of three
independent experiments.
Table 2. Percent growth of a VERO cell line exposed at four increasing concentrations (100, 200, 400, 800 lM)
Cell line
2a
400
ꢀ12
2b
2c
a
HSV-1 in VERO cells
100
200
ꢀ10
800
ꢀ55
100
ꢀ41
200
ꢀ55
400
ꢀ70
800
ꢀ83
100
ꢀ3
200
ꢀ3
400
ꢀ6
800
ꢀ67
ꢀ5
a
The negative value indicate the percent of cells killed.

A. Procopio et al. / Bioorg. Med. Chem. Lett. 15 (2005) 545–550
549
range 100 lM (41% of cells killed) to 500 lM (83% of
cells killed) (Table 2) and for the three compounds no
significant toxicity was detected in the treatment in the
MOLT-3 (human lymphoblastoid) and VERO (African
Green Monkey) cell lines assayed (CC50 > 1000 in every
case). Moreover, no apoptosis was detected.
by comparison of its spectral data with those reported
1
in literature.
7
4.2. General procedure 2a–c
6
A suspension of N1-vinyl pyrimidine derivatives 4a–c in
1
dry toluene (20 mM) was added to D -pyrrolidine-1-
oxide 3 (3.0 equiv) (prepared from its corresponding
hydroxylamine by HgO oxidation) and some grains of
hydroquinone as polymerization inhibitor. The reaction
mixture was stirred under reflux for 6–20 h (Table 3),
then the solution was concentrated under vacuum and
the crude mixture purified by flash-chromatography
(ethyl acetate/methanol = 9:1).
3
. Conclusions
Conformational analysis of nucleoside analogues 2a–c
by molecular modeling and NOE spectroscopy sup-
ported the assumption of rigid bicyclic structure in
which the isoxazolidine ring was locked in S-type
conformation.
0 0
.3. (2 RS,4 SR)-1-(Hexahydropyrrolo[1,2-b]isoxazolyl)
5-methyl-1H-pyrimidine-2,4-dione (2a)
4
The bicyclic N,O-nucleoside analogues 2a–c can be con-
sidered valuable structural probes for the evaluation of
RT binding affinity as well as useful templates for fur-
ther drug design investigations.
1
H NMR (300 MHz, DMSO-d ): d (ppm) 11.30 (s, 1H,
6
NH); 7.64 (s, 1H, H ); 6.16 (dd, 1H, H
0
0
, J
0
0
=
= 4.4); 3.84–3.73 (m, 1H, H ); 3.02–2.87
6
2
H2 H3 a
0
5.5, J
(m, 1H, H7 ); 2.70–2.53 (m, 2H, H ); 2.02–1.53
0
H2 H3 b
4
Furthermore, preliminary biological assays seem to re-
veal interesting chances to develop new potential antivi-
ral drugs starting from a locked nucleoside structures
resembling molecules 2a–c.
0
0
a
3
1
3
(m, 8H, H5 , H6 , H , CH3). C NMR (300 MHz,
0 0 0
7 b
DMSO-d ): d (ppm) 164.36; 150.95; 136.88; 109.29;
6
83.64; 64.95; 56.95; 43.31; 31.41; 24.13; 12.79. ESI/MS
m/z 260 (44) [M+Na] , 238 (100) [M+H] , 123 (50),
+
+
+
1
12 (26) [MꢀThy] . Anal. Calcd for C H N O : C,
1
1
15
3
3
4. Experimental
55.69; H, 6.37; N, 17.71. Found: C, 55.78; H, 6.32; N,
7.66.
1
Solvents and reagents were purified by standard proce-
dures and were distilled prior to use. H NMR spectra
were recorded at 300 MHz in DMSO-d using a tetra-
6
1
0
0
4.4. (2 RS,4 SR)-1-(Hexahydropyrrolo[1,2-b]isoxazolyl)
1H-pyrimidine-2,4-dione (2b)
methylsilane (TMS) as internal standard (Bruker ACP
00 MHz). Chemical shifts are given in ppm from
1
3
H NMR (300 MHz, DMSO-d ): d (ppm) 11.43 (s, 1H,
6
TMS and coupling constants in Hz. Microanalyses were
carried out with a Perkin–Elmer 240 analyzer.
NH); 8.02 (d, 1H, H , J
= 8.00); 6.33 (dd, 1H,
= 3.66); 5.82 (d, 1H, H5,
6
H6H5
H2
JH5H6 = 8.00); 4.00–3.89 (m, 1H, H4
1H, H7 ); 2.91–2.73 (m, 2H, H ); 2.20–1.97 (m, 5H,
0
, J
0
0
= 6.72, J
0
0
H2 H3 a
H2 H3 b
0
); 3.20–3.06 (m,
1
4
.1. D -Pyrrolidine-1-oxide (3)
0
0
3
a
1
3
H5
(ppm) 163.82; 150.97; 141.34; 101.58; 84.09; 64.80;
0
, H6
0
, H
0
). C NMR (300 MHz, DMSO-d ): d
7 b 6
NH OHÆHCl (8.83 g, 127.0 mmol) was added to a solu-
2
tion of 1,4-dibromobutane (10.0 g, 46.3 mmol) in dry
Et N (70 mL). The stirred solution was maintained
56.94; 43.70; 31.50; 24.15. ESI/MS m/z 224 (62)
+
+
[M+H] , 112 (67) [MꢀUra] , 84 (100), 68 (10), 56 (3).
3
under reflux for 1.5 h. The reaction mixture was then
cooled to room temperature and the solution separated
from the insoluble salts was diluted with ethylic ether
and filtered on Celite. The collected organic phases were
dried over Na SO , filtered and concentrated to give a
Anal. Calcd for C H N O : C, 53.80; H, 5.87; N,
3
1
0
13
3
18.82. Found: C, 53.68; H, 5.94; N, 18.88.
0
0
4.5. (2 RS,4 SR)-1-(Hexahydropyrrolo[1,2-b]isoxazolyl)
5-methyl-1H-4-amino-pyrimidine-2-one (2c)
2
4
crude mixture, which afforded the 1-hydroxypyrrolidine
in quantitative yield as a yellow oil pure enough to be
used in the oxidation step as shown by comparison of
1
H NMR (300 MHz, DMSO-d ): d (ppm) 7.78 (d, 1H,
6
H , J
= 7.70); 7.08 (d, 2H, NH , J
0
= 21.99);
= 2.75); 5.72
6
H6H5
2
gem
1
7
its spectral data with those reported in literature.
6.09 (dd, 1H, H2
d, 1H, H , J
3.05–2.85 (m, 1H, H ); 2.72–2.54 (m, 1H, H ); 2.52–
0
, J
0
= 7.14, J
= 7.70); 3.74–3.61 (m, 1H, H );
0
0
H2 H3 a
H2 H3 b
0
(
5
H5H6
4
A 0.6 M solution of 1-hydroxypyrrolidine in CH Cl
2
0
0
2
7 a 3 a
was cooled at 0 ꢁC and yellow HgO (9.0 g, 41.5 mmol)
was added slowly. After 0.5 h, the mixture was warmed
to room temperature and reacted for 4 h as indicated by
TLC [CH Cl /CH OH = 9:1]. The crude mixture was
2.36 (m, 1H, H
0
b); 2.02–1.53 (m, 5H, H5
0
, H6
, H ).
0 0
7 b
3
1
3
C NMR (300 MHz, DMSO-d ): d (ppm) 166.25;
6
2
2
3
then diluted with CH Cl and filtered on Celite. The col-
2
2
Table 3. Cycloaddition reaction for 2a–c preparation
lected organic phases were dried over Na SO , filtered
4
and concentrated. The obtained crude products were
purified on silica gel by flash-chromatography
[CH Cl /CH OH = 9:1]. The product 3 was obtained
2 2 3
as pure compound with 89% yield and was identified
2
Nucleobase
Reaction time (h) Yield (%) Epimeric ratio (%)
Thymine (2a)
Uracil (2b)
Cytosine (2c) 20
6
95
65
80
98:2
90:10
98:2
17

550
A. Procopio et al. / Bioorg. Med. Chem. Lett. 15 (2005) 545–550
1
2
55.74; 141.57; 93.95; 84.65; 64.66; 56.98; 44.37; 31.47;
6. Chiacchio, U.; Corsaro, A.; Iannazzo, D.; Piperno, A.;
Pistara, V.; Rescifina, A.; Romeo, R.; Valveri, V.;
+
4.98. ESI/MS m/z 445 (12) [2MꢀH] , 223 (93)
+
+
Mastino, A.; Romeo, G. J. Med. Chem. 2003, 46,
3
[M+H] , 112 (100) [MꢀCyt] . Anal. Calcd for
C H N O : C, 54.04; H, 6.35; N, 25.21. Found: C,
5
696.
10 14 4 2
7
8
. Van Roey, P.; Salerno, J. M.; Chu, C. K.; Schinazi, R. F.
Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 3929.
. (a) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc.
3.91; H, 6.42; N, 25.29.
1972, 94, 8205; (b) Altona, C.; Sundaralingam, M. J. Am.
Chem. Soc. 1973, 95, 2333.
Acknowledgements
9
. (a) Bj o¨ rsne, M.; Szab o´ , T.; Samuelsson, B.; Classon, B.
Bioorg. Med. Chem. 1995, 3, 397; (b) Marquez, V. E.;
Ezzitouni, A.; Russ, P.; Siddiqui, M. A.; Ford, H., Jr.;
Feldman, R. J.; Mitsuya, H.; George, C.; Barchi, J. J., Jr.
J. Med. Chem. 1998, 120, 2780; (c) Olsen, A. G.;
Rajwanshi, V. K.; Nielsen, C.; Wengel, J. J. Chem. Soc.,
Perkin Trans. 1 2000, 3610; (d) Bhushan, R. G.; Vince, R.
Bioorg. Med. Chem. 2002, 10, 2325; (e) Kim, M. J.; Kim,
H. O.; Kim, H. D.; Kim, J. H.; Jeong, L. S.; Chun, M. W.
Bioorg. Med. Chem. Lett. 2003, 13, 3299; (f) Obika, S.;
Nanbu, D.; Hari, Y.; Morio, K. I.; In, Y.; Ispida, T.;
Imanishi, T. Tetrahedron Lett. 1997, 50, 8735; (g) Kværnø,
L.; Kumar, R.; Dahl, B. M.; Olsen, C. R.; Wengel, J. J.
Org. Chem. 2000, 65, 5161; (h) H a˚ kansson, A. E.;
Koshkin, A. A.; Sørensen, M. D.; Wengel, J. J. Org.
Chem. 2000, 65, 5167; (i) Shin, K. J.; Moon, H. R.;
George, C.; Marquez, V. E. J. Org. Chem. 2000, 65,
2172.
10. (a) Dalpozzo, R.; De Nino, A.; Maiuolo, L.; Procopio, A.;
De Munno, G.; Sindona, G. Tetrahedron 2001, 57, 4035;
(b) Dalpozzo, R.; De Nino, A.; Maiuolo, L.; Procopio, A.;
Romeo, R.; Sindona, G. Synthesis 2002, 172.
11. Perzanowski, H. P.; Al-Jaroudi, S. S.; Wazeer, I. M.; Ali,
Sk. A. Tetrahedron 1997, 53, 11869.
12. Schr o¨ dinger Inc., 1500 S.W. First Avenue, Suite 1180,
Portland, OR 97201-5815.
13. Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson,
T. J. Am. Chem. Soc. 1990, 112, 6127.
We thank MIUR (Ministero per lꢀUniversit a` e Ricerca)
for financial support (COFIN 2002).
References and notes
1
. (a) Larder, B. A.; Kellam, P.; Kemp, S. D. Nature 1993,
65, 451; (b) Collier, A. C.; Coombs, R. W.; Shoenfeld, D.
3
A.; Basset, R. L.; Timpone, J.; Baruch, A.; Jones, M.;
Facey, K.; Whitacre, C.; McAuliffe, V. J.; Friedman, H.
M.; Merigan, T. C.; Reichman, R. C.; Hooper, C.; Corey,
L. N. Engl. J. Med. 1996, 334, 1011; (c) Reijers, M. H. E.;
Weverling, G. J.; Jurriaans, S.; Wit, F. W. N. M.; Weigel,
H. M.; Kate, R. W. T.; Mulder, J. W.; Frissen, P. H. J.;
van Leeuwen, R.; Reiss, P.; Schuitemaker, H.; de Wolf, F.;
Lange, J. M. A. Lancet 1998, 352, 185; (d) Montaner, J. S.
G.; Hogg, R.; Raboud, J.; Harrigan, R.; OꢀShaughnessy,
M. Lancet 1998, 352, 1919; (e) van Praag, R. M.; Wit, F.
W.; Jurriaans, S.; de Wolf, F.; Prins, J. M.; Lange, J. M.
AIDS 2002, 16, 719.
2
. (a) Leggio, A.; Liguori, A.; Procopio, A.; Siciliano, C.;
Sindona, G. Tetrahedron Lett. 1996, 37, 1277; (b) Leggio,
A.; Liguori, A.; Procopio, A.; Siciliano, C.; Sindona, G.
Nucleos. Nucleot. 1997, 16, 1515; (c) Leggio, A.; Liguori,
A.; Mariuolo, L.; Napoli, A.; Procopio, A.; Siciliano, C.;
Sindona, G. J. Chem. Soc., Perkin Trans. 1 1997, 3097; (d)
Colacino, E.; Converso, A.; De Nino, A.; Leggio, A.;
Liguori, A.; Maiuolo, L.; Napoli, A.; Procopio, A.;
Siciliano, C.; Sindona, G. Nucleos. Nucleot. 1999, 18(4–
14. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.;
Liskamp, R.; Lipton, M.; Caufield, C.; Chang, G.;
Hendrickson, T.; Still, W. C. J. Comput. Chem. 1990, 11,
440.
5), 581.
3
. (a) Chiacchio, U.; Corsaro, A.; Iannazzo, D.; Piperno, A.;
Rescifina, A.; Romeo, R.; Romeo, G. Tetrahedron Lett.
15. Halgren, T. A. J. Comput. Chem. 1996, 17, 490.
16. Alcaro, S.; Gasparrini, F.; Incani, O.; Mecucci, S.; Misiti,
D.; Pierini, M.; Villani, C. J. Comput. Chem. 2000, 21,
515.
17. (a) Cordero, F. M.; Machetti, F.; DeCarlo, F.; Brandi, A.
Gazz. Chim. Ital. 1997, 127, 25; (b) Cicchi, S.; Corsi, M.;
Goti, A. J. Org. Chem. 1999, 64, 7243; (c) Murahashi, S.
I.; Mitsui, H.; Shiota, T.; Tsuda, T.; Watanabe, S. J. Org.
Chem. 1990, 55, 1736.
2001, 42, 1777; (b) Richichi, B.; Cicchi, S.; Chiacchio, U.;
Romeo, G.; Brandi, A. Tetrahedron 2003, 59, 5231.
. Merino, P. Curr. Med. Chem.-Anti. Infective Agents 2002,
4
5
1
, 389.
. (a) Pan, S.; Amankulor, N. M.; Zhao, K. Tetrahedron
998, 54, 6587; (b) Gi, H. J.; Xiang, Y.; Schinazi, R. F.;
Zhao, K. J. Org. Chem. 1997, 62, 88.
1