4-Azido-2-pyrimidinone Nucleosides and Related Chemistry

Article abstract of DOI:10.1021/jo970761t

As a part of azide prodrug approach, we synthesized a 4-azido analog of ara-C (4) as a prodrug for ara-C. The compound 4 was obtained from 1-(β-D-arabinofuranosyl)uracil (1) in three steps. At pH 7.0 and 11.0, a loss of UV absorption of the compound 4 was observed resulting from a transformation that was proved by identifying the transformed product 5 by 1-D, and 2-D NMR as well as tandem mass spectral studies. In NMR studies, changes in the chemical shifts were observed at positions 5, 6, 1′, and 2′ between the compounds 4 and 5. A molecular peak at m/z 270.1 (MH⁺) was observed in the mass spectra of compounds 4 and the transformed product 5. A fragment at 180.2 was identified to be the compound 6, containing the 6,2′-anhydro linkage of compound 5. The X-ray analysis indicated that compound 4 exists as l-(β-D-arabinofuranosyl)tetrazolo[4,5-c]pyrimidin-2-one, with the azide moiety cyclized. To understand if the chemical instability of the nucleoside 4 was due to the arabino configuration of 2′-OH or due to the azido moiety, we also studied 1-(2,3-dideoxy-2-fluoro-β-D-arabinofuranosyl)tetrazolo[4,5-c]p3Timidin-2- one (11) and 4-azido-1-methyl-2-pyrimidinone (15). At pH 2.0 and 7.0, similar UV profiles were observed for compounds 11 and 15. However, at pH 11.0, λ_max shifted slowly to lower wavelength for both compounds 11 and 15. In a separate kinetic study, they were stable at pH 7.4 for up to 2.45 h. From the NMR and high-resolution mass spectral studies, it was concluded that in the presence of ammonium hydroxide, an addition of amine occurred at 6-position of compound 11. Thus, the stability profiles of compounds 4, 11, and 15 were different. The instability and the formation of 2′,6-anhydro bond in compound 4 in nonacidic media was due to the presence of 2′-OH in the arabino configuration and probably not due to the azide group.

Full text of DOI:10.1021/jo970761t

J . Org. Chem. 1997, 62, 7267-7271
7267
4-Azid o-2-p yr im id in on e Nu cleosid es a n d Rela ted Ch em istr y
Lakshmi P. Kotra,†,§ PingPing Wang,† Michael G. Bartlett,† Kirupa Shanmuganathan,†
Zusheng Xu,† So´crates Cavalcanti,† M. Gary Newton,‡ and Chung K. Chu*,†
Department of Medicinal Chemistry, College of Pharmacy, and Department of Chemistry,
The University of Georgia, Athens, Georgia 30602
Received April 29, 1997X
As a part of azide prodrug approach, we synthesized a 4-azido analog of ara-C (4) as a prodrug for
ara-C. The compound 4 was obtained from 1-(â-D-arabinofuranosyl)uracil (1) in three steps. At
pH 7.0 and 11.0, a loss of UV absorption of the compound 4 was observed resulting from a
transformation that was proved by identifying the transformed product 5 by 1-D, and 2-D NMR as
well as tandem mass spectral studies. In NMR studies, changes in the chemical shifts were observed
at positions 5, 6, 1′, and 2′ between the compounds 4 and 5. A molecular peak at m/ z 270.1 (MH+)
was observed in the mass spectra of compounds 4 and the transformed product 5. A fragment at
180.2 was identified to be the compound 6, containing the 6,2′-anhydro linkage of compound 5.
The X-ray analysis indicated that compound 4 exists as 1-(â-D-arabinofuranosyl)tetrazolo[4,5-
c]pyrimidin-2-one, with the azide moiety cyclized. To understand if the chemical instability of the
nucleoside 4 was due to the arabino configuration of 2′-OH or due to the azido moiety, we also
studied 1-(2,3-dideoxy-2-fluoro-â-D-arabinofuranosyl)tetrazolo[4,5-c]pyrimidin-2-one (11) and 4-azido-
1-methyl-2-pyrimidinone (15). At pH 2.0 and 7.0, similar UV profiles were observed for compounds
11 and 15. However, at pH 11.0, λmax shifted slowly to lower wavelength for both compounds 11
and 15. In a separate kinetic study, they were stable at pH 7.4 for up to 2.45 h. From the NMR
and high-resolution mass spectral studies, it was concluded that in the presence of ammonium
hydroxide, an addition of amine occurred at 6-position of compound 11. Thus, the stability profiles
of compounds 4, 11, and 15 were different. The instability and the formation of 2′,6-anhydro bond
in compound 4 in nonacidic media was due to the presence of 2′-OH in the arabino configuration
and probably not due to the azide group.
Recently, as a part of azide prodrug approach we
reported the synthesis and biotransformation of 6-azido-
9-(2,3-dideoxy-2-fluoro-â-D-arabinofuranosyl)purine (FAAd-
dP) as a prodrug for 9-(2,3-dideoxy-2-fluoro-â-D-arabino-
furanosyl)hypoxanthine (2′-F-ara-ddI).1 We also inves-
tigated 6-azido-(1-â-D-arabinofuranosyl)purine (6-AAP) as
a prodrug for 1-(â-D-arabinofuranosyl)adenine (ara-A)
and demonstrated that unlike ara-A, 6-AAP was stable
to adenosine deaminase, which generates ara-A both in
vitro and in vivo in mice.2 These azide prodrugs are
reduced to their amine analogs by the cytochrome P450-
NADPH reductase system after entering the biological
system to deliver the amine-containing drugs.
In continuation of our efforts to develop prodrugs for
various antitumor and antiviral agents to enhance the
pharmacokinetic properties, utilizing the azide prodrug
approach, we synthesized the 4-azido analog of ara-C (4,
Scheme 1) as a prodrug for 1-(â-D-arabinofuranosyl)cy-
tosine (ara-C). It was anticipated that the azide moiety
would be stable to cytidine deaminase but upon reduction
by the cytochrome P450-NADPH reductase system in
vivo, it would generate ara-C.
1 was synthesized from uridine, according to the pub-
lished procedure,3 which was protected by the treatment
with acetic anhydride, treated with 4-chlorophenyl di-
chlorophosphate and 1,2,4-triazole to give compound 2.
The triazolide 2 was treated with LiN3 in DMF at 50 °C
to obtain compound 3 in 27% yield from compound 1.
Deprotection of the acetyl groups of compound 3 using
basic conditions such as NH3/MeOH, NaOMe, or neutral
conditions (NaCN in ethanol) did not result in the desired
product but compounds of destruction. However, the
deprotection of acetyl groups was achieved by treatment
of compound 3 with 1 N HCl/MeOH at rt to afford
compound 4 in 72% yield. The compound 4 was charac-
terized by 1H NMR, 13C NMR, UV spectroscopy, and
elemental analysis. Interestingly, single crystal X-ray
crystallography of compound 44 (Figure 1, Table 1)
showed that the azido group at 4-position had cyclized
to form a tetrazole ring in conjugation with N3 of the
pyrimidine ring, forming 1-(â-D-arabinofuranosyl)tetra-
zolo[4,5-c]pyrimidin-2-one and such phenomenon was
also observed by other groups for similar systems.5
During in vitro biotransformation studies of compound
4 in the biological media, we failed to detect the prodrug
or its product by a HPLC assay. This prompted us to
study the chemical stability of the compound 4 at various
pHs, and the UV spectral pattern is shown in Table 2.
The compound 4 was obtained from 1-(â-D-arabino-
furanosyl)uracil (1) in three steps (Scheme 1). Compound
* Phone: (706) 542-5379. Fax: (706) 542-5381. E-mail: dchu@rx.
uga.edu.
The UV profiles of compound 4 in methanol and at pH
2.0 had similar patterns and did not show any changes
† Department of Medicinal Chemistry.
‡ Department of Chemistry.
§ Present address: Chemistry Department, Wayne State University,
Detroit, MI.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
(1) Koudriakova, T.; Manouilov, K. K.; Shanmuganathan, K.; Kotra,
L. P.; Boudinot, F. D.; Cretton-Scott, E.; Sommadossi, J .-P.; Schinazi,
R. F.; Chu, C. K. J . Med. Chem. 1996, 39, 4676-4681.
(2) Kotra, L. P.; Manouilov, K. K.; Cretton-Scott, E.; Sommadossi,
J . -P.; Boudinot, F. D.; Schinazi, R. F.; Chu, C. K. J . Med. Chem. 1996,
39, 5202-5207.
(3) Schinazi, R. F.; Chen, M. S.; Prusoff, W. H. J . Med. Chem. 1979,
22, 1273-1277.
(4) Primitive orthorhombic crystals with space group P212121 with
lattice parameters a ) 6.565(2) Å, b ) 13.108(3) Å, c ) 14.265(3) Å, V
) 1227.7(6) Å3.
(5) Napoli, L. D.; Mayol, L. J . Heterocycl. Chem. 1986, 23, 1401-
1403.
S0022-3263(97)00761-5 CCC: $14.00 © 1997 American Chemical Society

7268 J . Org. Chem., Vol. 62, No. 21, 1997
Kotra et al.
Ta ble 2. UV P r ofiles of Com p ou n d s 4, 11, a n d 15
Sch em e 1.a Syn th esis of Com p ou n d 4
compd
solvent
MeOH
UV (λmax)
4
268.5 nm (shoulder), 257.0 nm
H2O, pH 2.0
H2O, pH 7.0
H2O, pH 11.0
H2O, pH 2.0
H2O, pH 7.0
H2O, pH 11.0
MeOH
267.5 nm (shoulder), 255.5 nm
245.5 nm
239.0 nm
266.5 nm (shoulder), 252.0 nm
266.5 nm, 253.0 nm
256.0 nm
257.0
11
15
H2O, pH 2.0
H2O, pH 7.0
H2O, pH 11.0
256.5
256.0
248.0
Ta ble 3. 1H a n d 13C Ch em ica l Sh ifts (δ p p m ) for th e
Com p ou n d s 4, 5, 11, a n d 12
compd 4
compd 5
1H Signal
∆δ
2.2
0.8
-0.1
-1.9
-0.3
0.1
H-5
H-6
7.1
8.0
6.2
4.2
4.0
3.9
3.6
4.9
7.2
6.3
6.1
4.3
3.8
3.2
H-1′
H-2′
H-3′
H-4′
H-5′
0.4
13C Signal
C-2
C-4
C-5
C-6
C-1′
C-2′
C-3′
C-4′
C-5′
151.5
143.0
75.4
137.2
87.8
75.4
75.3
86.3
60.8
156.5
155.0
83.6
-5.0
-12.0
-8.2
50.7
-31.6
-29.8
0.8
86.5
119.4
105.2
74.5
86.6
60.7
-0.3
0.1
a
Reagents: (a) (i) Ac2O, pyridine, 0-4 °C, 16 h; (ii) chlorophen-
compd 11
compd 12
1H Signal
∆δ
yl dichlorophosphate, 1,2,4-triazole, pyridine, rt; (b) LiN3, anhyd
DMF, 50 °C, 3 h; (c) 1 N HCl/MeOH, rt, 26 h; (d) H2/PdCl2, MeOH,
50 psi, rt, 30 h.
H-5
H-6
7.18
8.11
6.29
6.23
6.31
5.94
0.95
1.80
0.35
0.28
0.23, 0.20
0.42
0.13
H-1′
H-2′
H-3′
H-4′
H-5′
NH2
5.48 (J H,F ) 54.6)
2.17, 2.57
4.26
5.20 (J H,F ) 55.1)
1.94, 2.37
3.84
3.62
3.49
6.09 (2 H)
13C Signal
C-2
C-4
C-5
C-6
C-1′
C-2′
C-3′
C-4′
C-5′
151.6
157.6
135.4
113.2
126.6
87.1
92.6
33.8
75.6
63.3
-6.0
143.1
93.0
136.8
86.8
91.7
32.5
79.4
62.8
7.7
-20.2
10.2
-0.3
-0.9
-1.3
3.8
F igu r e 1. ORTEP drawing of the X-ray structure of compound
4.
-0.5
provide basic environment and for the easy isolation of
the product.
Ta ble 1. Con for m a tion a l P a r a m eter s for Com p ou n d s
4 a n d 11
On the basis of the structure of compound 4 and
previously known literature,6,7 the structure of compound
5 was predicted to be as shown in Scheme 2. The proof
for the structure 5 was obtained from 1-D and 2-D NMR
and tandem mass spectral studies. Table 3 lists the
chemical shifts of various protons in compounds 4 and
5. Chemical shifts of the protons were assigned using
the 1-D proton and 2-D COSY spectra. Comparison of
the chemical shifts of protons in the compounds 4 and 5
showed that the major changes occurred for H-5 (2.2 ppm
upfield), H-6 (0.8 ppm upfield), and H-2′ (1.9 ppm
downfield). The other proton signals shifted less than
0.5 ppm either upfield or downfield.
torsion angles
pseudo-
rotation C3′-C4′-C5′-O5′ C2′-C1′-N1-C2 confor-
compd angle, P
(γ)
(ø)
mation
4
11
21.9°
159.4°
56.6°
175.0°
-155.6°
-132.2°
C3′-endo
C2′-endo
in the λmax for up to 2.45 h. However, at higher pHs (7.0
and 11.0), λmax shifted to lower wavelengths, indicating
a change in the base moiety of the compound 4. In order
to elucidate the structural changes in the compound 4
at higher pHs, it was treated with ammonium hydroxide
and the solvent was evaporated to dryness. It provided
a single product (later assigned to be compound 5) as
determined by NMR and mass spectra, and its UV profile
was similar to that observed in nonacidic media for
compound 4. Ammonium hydroxide was used merely to
(6) Fox, J . J .; Miller, N. C.; Cushley, R. J . Tetrahedron Lett. 1966,
4627-4634.
(7) Major, P. P.; Egan, E. M.; Herrick, D.; Kufe, D. W. Biochem.
Pharamcol. 1982, 31, 861-866.

4-Azido-2-pyrimidinone Nucleosides
J . Org. Chem., Vol. 62, No. 21, 1997 7269
Sch em e 2. Tr a n sfor m a tion of Com p ou n d 4 F or m in g th e 2′,6-An h yd r o Lin k a ge
Sch em e 3.a Syn th esis of Com p ou n d 11
a
Reagents: (a) Silylated uracil, DCE, reflux; (b) 4-chlorophenyl dichlorophosphate, 3-nitro-1,2,4-triazole, py, rt; (c) LiN3, DMF, rt; (d)
saturated NH3/MeOH, rt; (e) NH4OH; (f) H2, Pd/C, MeOH, 50 psi, rt, 15 h.
The chemical shifts of various carbon signals for the
anhydro linkage of compound 5) could only be obtained
compounds 4 and 5 are also shown in Table 3, and the
by the fragmentation of the C1′-O4′ bond and C2′-C3′
bond. In retrospect, the structure of the MS fragment 6
suggests the 2′,6-linkage in compound 5 and the proposed
transformation could occur as shown in Scheme 2.
Previously, it was known that ara-C forms a 2′,6-
anhydro linkage under alkaline conditions.6,7 When
ara-C was treated with 0.1 N NaOH at 60-70 °C, a
loss of UV absorption maxima was observed within 30
min.6 When 1-(â-D-arabinofuranosyl)-5-fluorocytosine
was treated under the same conditions, a loss of its UV
absorption maxima was observed and the structure of the
transformed compound contained the 2′,6-linkage. How-
ever, compounds containing 2′-OH in the ribo configu-
ration did not show loss of UV absorption under similar
conditions. This suggested that compounds containing
2′-OH in arabino configuration could undergo intramo-
lecular rearrangement, resulting in the formation of the
2′,6-anhydro linkage. In the case of compound 4, this
rearrangement even occurs at neutral pH (7.0) and under
physiological conditions (pH 7.4).
To understand if the chemical instability of the pyrim-
idine nucleoside 4 in the base moiety was due to the
arabino configuration of 2′-OH or due to the azido moiety,
we also studied the chemical stability of 1-(2,3-dideoxy-
2-fluoro-â-D-arabinofuranosyl)tetrazolo[4,5-c]pyrimidin-
2-one (11). Compound 11 was designed as a prodrug of
1-(2,3-dideoxy-2-fluoro-1-â-D-arabinofuranosyl)cytosine (2′-
F-ara-ddC) by utilizing the azide prodrug approach.8,9 The
compound 11 differs from compound 4 in that it has a
1
signals were assigned on the basis of H decoupled 13C,
DEPT-45, DEPT-90, DEPT-135, and heteronuclear COSY
(HETCOR) experiments. Major changes in the chemical
shifts of carbon signals were observed for C-4 (12.0 ppm
downfield), C-5 (8.2 ppm downfield), C-6 (50.7 ppm
upfield), C-1′ (31.6 ppm downfield), and C-2′ (29.8 ppm
1
downfield). Thus, the chemical shift changes of H and
13C signals between compounds 4 and 5 indicated that
the transformation of 4 involves the positions 4, 5, 6, 1′,
and 2′. This supports the structure of the predicted
compound 5. Further support for the structure of com-
pound 5 was obtained by high-resolution and tandem
mass spectral studies.
Mass spectra (liquid secondary ion mass spectroscopy,
LSIMS) of compounds 4 and 5 revealed a molecular peak
(MH+) at m/ z 270.1 for both compounds. The experi-
mentally determined empirical formulas for both com-
pounds were identical suggesting that compound 5 was
formed by a simple transformation from compound 4 at
higher pH. In the mass spectrum of compound 5, a major
fragment with m/ z 180.2 was observed along with the
molecular peak at m/ z 270.1 (MH+). However, in the
mass spectrum of compound 4, major fragments were at
m/ z 270.1 (MH+) and 138.2 (MH+ - arabinose), but no
significant fragment was observed at 180.2. From the
tandem and high-resolution mass spectral analysis, the
fragment at 180.2 was identified to be the MS fragment
6 (Scheme 2). The fragment 6 (containing the 2′,6-

7270 J . Org. Chem., Vol. 62, No. 21, 1997
Kotra et al.
which could possibly be at the 6-position (from the
changes in the NMR signals), from which the structure
was confirmed to be 12. The addition of the amine was
also confirmed by deuterium exchange studies and
subsequent MS/MS analysis.
The high-resolution mass spectrum (LSIMS) of com-
pound 11 showed only the molecular peak at m/ z
256.0857 (MH+, calcd 256.0846 corresponding to C9H11
-
N5O3F) and the MS/MS analysis of the molecular peak
showed the fragment peaks at m/ z 138.1 ([M - sugar
moiety]+), 118.8 ([M - base]+) and 110.1 ([138 - N2]+).
The high-resolution mass spectrum of compound 12
showed the molecular peak at 273.1123 (MH+, calcd
273.1111 corresponding to C9H14N6O3F), confirming the
addition of an amine moiety. The complete details of the
mass spectral and fragmentation analysis of compounds
4, 5, 11, and 12 will be published elsewhere.
F igu r e 2. ORTEP drawing of the X-ray structure of compund
11.
fluorine atom in the place of 2′-OH in the ara configu-
ration and lacks the 3′-OH moiety.
Compound 11 was synthesized from the bromo sugar
710 via the triazolide 9. The compound 7 was condensed
with silylated uracil under refluxing conditions to give
compound 8 in 75% yield. The compound 8 was treated
with 4-chlorophenyl dichlorophosphate and 3-nitro-1,2,4-
triazole in pyridine at rt to obtain the crude triazolide 9.
Treatment of the crude compound 9 with lithium azide
in DMF at rt to obtain compound 10 followed by the
deprotection of benzoyl group with saturated ammonia
in methanol afforded the desired compound 11. Single
crystal X-ray studies showed that the azide group at
4-position of compound 11 also forms a tetrazole ring with
N3 of the pyrimidine ring (Figure 2, Table 1), similar to
that in compound 4.11 The UV profiles of compound 11
at various pH are shown in Table 2.
At pH 2.0 and 7.0, UV profiles were similar for
compound 11 and there was no change in the λmax value.
However, at pH 11.0, λmax shifted to 256.0 nm. In a
separate kinetic study, compound 11 was stable at pH
7.4 and there was no change observed in the λmax and
the extinction coefficient, for up to 4 h. This indicated
that compound 11 was stable at physiological pH. To
investigate if the compound 11 can undergo any trans-
formations similar to compound 4, it was treated in a
similar fashion as compound 4 to obtain the transformed
product, which was later assigned to be compound 12.
In NMR studies, no major change was observed in the
chemical shifts of the sugar moiety. However, there were
major differences observed in the carbon and proton
chemical shifts of the base moiety, indicating that the
sugar moiety was not involved in the rearrangement. In
the proton spectrum (Table 3), H-5 and H-6 signals
showed a change of 0.95 and 1.80 ppm upfield, respec-
tively. In the 13C NMR spectrum (Table 3), major
changes in the chemical shifts between compound 11 and
12 were observed at C-2 (6.0 ppm downfield), C-4 (7.7
ppm upfield), C-5 (20.2 ppm downfield), and C-6 (10.2
ppm upfield), confirming that the base moiety was
involved in the transformation of compound 11 to 12.
Additionally, a broad singlet peak was observed at 6.09
in the proton spectrum of compound 12. High-resolution
mass spectral analysis suggested an amine addition,
To further confirm the structures and to investigate if
they could be reduced chemically, we performed hydro-
genation of the compounds 4 and 11 in the presence of
Pd/C or PdCl2, obtaining their corresponding amine
analogues, ara-C and 2′-F-ara-ddC, respectively.
We have also synthesized 4-azido-1-methyl-2-pyrimi-
dinone (15) as a model system to further support the
results obtained from compounds 4 and 11. The com-
pound 15 contains an azide moiety but no glycosidic bond.
The compound 15 was synthesized from 1-methyluracil
(13) via the triazolide intermediate 14. The UV profile
of the compound 15 is shown in Table 2. Compound 15
did not show any significant changes in the UV profile
at pH 2.0 and 7.0, indicating no change in the base moiety
under these experimental conditions. However at pH
11.0, the λmax shifted to 248.0 nm, similar to that of
compound 11. The above results, i.e., the UV profiles of
compounds 11 and 15 when compared to that of 4,
suggested that the instability/transformation of com-
pound 4 at pH 7.0 was not due to the azido group on the
pyrimidine ring (vide infra). While we did not perform
X-ray structure analysis on compound 15 to see if the
azide is cyclized; it is not critical at this point for our
analysis.
In conclusion, the instability and 2′,6-anhydro bond
formation in compound 4 was due to the presence of 2′-
OH in the arabino configuration. The intramolecular
attack of the 2′-OH is faster and facilitated the trans-
formation to form the 2′,6-anhydro linkage. Due to the
lack of the 2′-OH, compounds 11 and 15 were relatively
more stable than the compound 4, especially at physi-
ological pH. However, compound 11 was not stable to
alkaline conditions like ara-C and exhibits susceptibility
to nucleophilic attack at C-6 position.
Exp er im en ta l Section
Melting points were determined on a MelTemp-II melting
point apparatus and are uncorrected. UV spectra were
recorded on a Beckman DU-7 or DU-650 spectrophotometer
either in methanol, buffer solution (pH 2.0, 7.0 or 11.0), or
phosphate-buffered saline (pH 7.4). Buffer powders were
purchased from Aldrich Chemical Co. NMR data were re-
corded on a Bruker-200 AC, Bruker-300 AM or Bruker-400
AMX spectrometer using DMSO-d6 as solvent unless otherwise
specified, and the chemical shifts are reported in ppm (δ).
Coupling constants (J ) are reported in hertz (Hz). The
abbreviations used are s (singlet), d (doublet), t (triplet), m
(multiplet), pd (pseudo doublet), pt (pseudo triplet), brs (broad
singlet), and dm (doublet of multiplet). Mass spectra were
recorded either on a Micromass AutoSpec high-resolution mass
spectrometer (LSIMS) or a Micromass Quatro II triple quad-
rupole mass spectrometer (ESI, MS/MS). Elemental analyses
(8) Martin, J . A.; Bushnell, D. J .; Duncan, I. B.; Dunsdon, S. J .; Hall,
M. J .; Machin, P. J .; Merrett, J . H.; Parkes, K. E. B.; Roberts, N. A.;
Thomas, G. J .; Galpin, S. A.; Kinchington, D. J . Med. Chem. 1990, 33,
2137-2145.
(9) Sterzycki, R. Z.; Ghazzouli, I.; Brankovan, V.; Martin, J . C.;
Mansuri, M. M. J . Med. Chem. 1990, 33, 2150-2157.
(10) Shanmuganathan, K.; Koudriakova, T.; Nampalli, S.; Du, J .;
Gallo, J . M.; Schinazi, R. F.; Chu, C. K. J . Med. Chem. 1994, 37, 821-
827.
(11) Monoclinic crystals with space group C2 with lattice parameters
a ) 20.186(4) Å, b ) 5.252(8) Å, c ) 10.354(2) Å, â ) 105.5(2)°, V )
1058(1) Å3.

4-Azido-2-pyrimidinone Nucleosides
J . Org. Chem., Vol. 62, No. 21, 1997 7271
were performed either by Atlantic Microlabs, Inc., Norcross,
GA, or by Galbraith Laboratories, Knoxville, TN. The stan-
dard workup procedure followed in the reactions, where
specified, was to wash the reaction mixture with an equal
volume of saturated NaHCO3 solution and brine and to dry
the organic layer (Na2SO4 or MgSO4).
DMF (30 mL), and LiN3 (1.5 g, 30.6 mmol) was added. The
reaction mixture was stirred at rt for 1 h and the solvent was
evaporated to dryness under high vacuum. The residue was
dissolved in ethyl acetate and the workup followed the
standard procedure. Then the organic layer was concentrated
to give compound 10 as a yellow solid (2.7 g, crude 98%), which
was treated with saturated ammonia in methanol (300 mL)
and stirred for 4 h at rt. Then the solvent was evaporated
and the residue was neutralized with AcOH (0.5 mL) in MeOH
(200 mL). The crude product was purified by silica gel column
chromatography (5% MeOH:CHCl3) to yield compound 11 (1.7
g, 90%) as a white solid: mp 133 °C; UV (H2O) λmax 252.0
(9322); 266.5 (sh, 8704) (pH 2.0); 253.0 (7981), 266.5 (shoulder,
1-(2,3,5-Tr i-O-a cet yl-â-D-a r a b in ofu r a n os-1-yl)-4-t r ia -
zolyl-2-p yr im id in on e (2). A solution of compound 13 (950
mg, 3.9 mmol) in pyridine (20 mL) was treated with acetic
anhydride (1.2 mL, 11.7 mmol) at 0 °C and stirred for 12 h.
The solvent was evaporated and the resulting yellow syrup
was dissolved in EtOAc (50 mL). The organic layer was
subjected to the standard workup procedure and concentrated,
and the residue was dissolved in anhydrous pyridine (20 mL).
The solution was treated with 4-chlorophenyl dichlorophos-
phate (0.9 mL, 5.2 mmol) at 0 °C under argon and 3-nitro-
1,2,4-triazole (1.4 g, 11.9 mmol) was added. The reaction
mixture was stirred at rt for 24 h and the solvent was
evaporated. The residue was dissolved in CH2Cl2 and followed
the standard workup procedure. The solvent was evaporated
and the crude yellow compound 2 was used as such for the
next reaction.
1
7302) (pH 7.0); 256.0 (12 490); H NMR δ 8.11 (d, J ) 7.83,
1H, H-6), 7.18 (d, J ) 8.86, 1H, H-5), 6.29 (dd, J 1 ) 3.71, J 2
)
15.62, 1H, H-1′), 5.48 (dm, J ) 54.46, 1H, H-2′), 5.20 (brs, 1H,
5′-OH, D2O exchangeable), 4.26 (m, 1H, H-4′), 3.62 (m, 2H,
H-5′), 2.17, 2.57 (2m, 2H, H-3′); HRMS (LSMSI, m/ z) 256.0857
(calcd 256.0846). Anal. Calcd for C9H10N5O3F: C, 42.37; H,
3.95; N, 27.44; F, 7.45. Found: C, 42.59; H, 3.98; N, 27.30; F,
7.05.
1-Met h yl-4-(3-n it r o-1,2,4-t r ia zol-1-yl)-2-p yr im id in on e
(14). A solution of 1-methyluracil (13) (150 mg, 1.2 mmol) in
anhyd pyridine (10 mL) was cooled in an ice bath and treated
with 4-chlorophenyl dichlorophosphate (0.34 mL, 1.4 mmol)
followed by 3-nitro-1,2,4-triazole (410 mg, 3.6 mmol). The
reaction mixture was stirred at rt for 2 h and quenched with
methanol (1 mL). The solvent was evaporated and the residue
was dissolved in saturated NaHCO3 solution (20 mL). The
aqueous layer was extracted with CHCl3 (5 × 30 mL), and the
combined organic layers were dried (Na2SO4). The organic
layer was concentrated to obtain the crude compound 14.
4-Azid o-1-m eth yl-2-p yr im id in on e (15). A solution of the
crude compound 14 in DMF (10 mL) was treated with LiN3
(233 mg, 4.2 mmol) and stirred at 50 °C for 3 h. The reaction
mixture was then poured into water (30 mL) and extracted
with CHCl3 (5 × 40 mL). The combined organic layers were
concentrated and recrystallized from ethanol (5 mL) to give
the compound 15 as a white solid (93 mg, 52%): UV (H2O)
λmax 256.5 (10 481) (pH 2.0); 256.0 (10 958) (pH 7.0); 248.0
(13 253) (pH 11.0); 1H NMR δ 7.95(d, 1H, J ) 7.59, H-6), 7.09
(d, 1H, J ) 7.61, H-5), 3.59 (s, 3H, CH3). Anal. Calcd for
C5H5N5O: C, 39.74; H, 3.33; N, 46.34. Found: C, 39.66; H,
3.38; N, 46.23.
P r ep a r a tion of Com p ou n d s 5 a n d 12. Either compound
4 or 11 (100 mg) was dissolved in ammonium hydroxide
(compound 11 was warmed to dissolve completely) and the
solvent was evaporated to dryness. The transformation of
compound 4 took place in less than 5 min. The transformation
of compound 11 was slower and took 30 min to 1 h. The
residue was dissolved in water and freeze-dried to yield the
product 5 or 12.
Red u ction of Com p ou n d s 4 a n d 11. Either compound 4
or 11 (100 mg) was dissolved in MeOH (10 mL) and hydroge-
nated in the presence of Pd/C or PdCl2 (20 mg) at 55 psi at rt
for 15 h. The reaction mixture was filtered through a pad of
Celite and the solvent was evaporated. The crude mixture was
purified by prep TLC (20% MeOH:CHCl3) to give ara-C (21%)
or 2′-F-ara-ddC (28%), and the spectral data corresponds to
previously published reports.9,12
1-(2,3,5-Tr i-O-a cetyl-â-D-a r a bin ofu r a n os-1-yl)-4-a zid o-
2-p yr im id in on e (3). Crude compound 2 was dissolved in
dimethyl formamide (30 mL), and LiN3 (953 mg, 19.5 mmol)
was added. The reaction mixture was stirred at 50 °C for 3 h
and dissolved in ethyl acetate (70 mL). The organic layer was
subjected to standard workup procedure and concentrated, and
the crude was purified by silica gel column chromatography
(15-50% EtOAc:Hx) to yield compound 3 (560 mg, 27% from
3) as white crystals: mp 160-161 °C; UV (MeOH) λmax 266.0
nm (sh), 255.0 nm, 203.0 nm; IR (KBr) 3125.80, 2984.26,
1743.87. Anal. Calcd for C15H17N5O8: C, 45.57; H, 4.33; N,
17.71. Found: C, 45.50; H, 4.34; N, 17.50.
1-(â-D-Ar a b in ofu r a n osyl)t et r a zolo[4,5-c]p yr im id in -2-
on e (4). Compound 3 (495 mg, 1.3 mmol) was treated with 1
N HCl in methanol (25 mL) and stirred for 26 h at rt. The
solvent was evaporated, the resulting white crystals were
washed with 2-propanol and dried under high vacuum to yield
compound 4 (261 mg, 72%): mp 123-126 °C; UV (MeOH) λmax
267.5 nm (shoulder), 257.5 nm, 206.0 nm; 267.5 nm (sh, 6141),
255.5 nm (6572) (pH 2.0); 245.5 nm (9537) (pH 7.0); 239.5 nm
(12 876) (pH 11.0); 1H NMR (DMSO-d6/D2O) δ 7.99 (d, J ) 7.7,
1H, H-6), 7.12 (d, J ) 7.96, 1H, H-5), 6.23 (d, J ) 4.12, 1H,
H-1′), 4.20 (t, J ) 3.6, 1H, H-2′), 3.98 (pt, J 1 ) 1.57, J 2 ) 3.16,
1H, H-3′), 3.90 (dd, J 1 ) 8.69, J 2 ) 5.20, 1H, H-4′), 3.67 (d, J
) 5.17, 2H, H-5′). Anal. Calcd for C9H11N5O5‚H2O: C, 37.63;
H, 4.56; N, 24.38. Found: C, 37.74; H, 4.56; N, 24.35.
1-(5-O-Ben zoyl-2,3-dideoxy-2-flu or o-â-D-ar abin ofu r an os-
1-yl)u r a cil (8). Uracil (3.2 g, 28.5 mmol) was treated with
1,1,1,3,3,3-hexamethyldisilazane (50 mL) under argon and
refluxed for 6 h. Excess solvent was removed under reduced
pressure and was treated with freshly prepared bromo sugar
710 in 1,2-dichloroethane (150 mL). The reaction mixture was
refluxed at 80 °C for 4 h and cooled to rt. The reaction mixture
was quenched with MeOH (10 mL) and concentrated. The
residue was dissolved in EtOAc and the workup followed the
standard procedure. The organic layer was concentrated and
purified on a silica column (2% MeOH:CHCl3) to obtain
compound 8 (2.4 g, 74%) as a major product: mp 171-172 °C;
UV (MeOH) λmax 260 nm. Anal. Calcd for C16H15N2O5F: C,
57.48; H, 4.52; N, 8.38; F, 5.68. Found: C, 56.96; H, 4.56; N,
8.15; F, 5.21.
P r ep a r a tion of Cr ysta ls. Either compound 4 or 11 was
dissolved in 1 mL of MeOH until the solution was saturated
and the solution was left at rt while the crystals grew. When
crystals of required size were seen, the solution was filtered,
the crystals were washed with cold MeOH and dried.
1-(2,3-Did eoxy-2-flu or o-â-D-a r a bin ofu r a n osyl)tetr a zo-
lo[4,5-c]p yr im id in -2-on e (11). A solution of compound 8 (2.6
g, 7.6 mmol) in pyridine (50 mL) was cooled in an ice bath
and 4-chlorophenyl dichlorophosphate (1.9 mL, 11.5 mmol) was
slowly added followed by 3-nitro-1,2,4-triazole (2.6 g, 22.9
mmol). The reaction mixture was stirred for 3 h and the
solvent was evaporated to dryness. The residue was dissolved
in chloroform and subjected to the standard workup procedure.
Upon concentration of the organic layer, a golden yellow foam
was obtained (9). The crude compound 9 was dissolved in
Ack n ow led gm en t. This work was supported by the
NIH grants AI 32351 and AI 25899.
J O970761T
(12) Roberts, W. K.; Dekker, C. A. J . Org. Chem. 1967, 32, 816-
817.