Synthesis of 1′-fluorouracil nucleosides as potential antimetabolites

Article abstract of DOI:10.1016/j.tet.2006.07.101

The first synthesis of 1′-fluoronucleosides, which has long been synthetic targets as the potential antimetabolites, was achieved. Electrophilic fluorination of the 1′-position occurred to form an anomeric mixture of 1′-fluorouridine derivatives, when the

Full text of DOI:10.1016/j.tet.2006.07.101

Tetrahedron 62 (2006) 10011–10017
Synthesis of 10-fluorouracil nucleosides as potential
antimetabolites*
*
Tetsuya Kodama, Akira Matsuda and Satoshi Shuto
Graduate School of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
Received 7 June 2006; accepted 27 July 2006
Available online 23 August 2006
Abstract—The first synthesis of 10-fluoronucleosides, which has long been synthetic targets as the potential antimetabolites, was achieved.
Electrophilic fluorination of the 10-position occurred to form an anomeric mixture of 10-fluorouridine derivatives, when the lithium enolate,
prepared from 30,50-O-tetraisopropyldisiloxane-1,3-diyl (TIPDS)-protected 20-ketouridine (10) and LiHMDS, was treated with an electro-
philic fluorinating agent such as NFSI (13). Subsequent reduction of the 20-keto-moiety of the resulting b-nucleoside gave the protected
10-fluorouridine 16 and its arabino-type congener 17. Alternatively, nucleophilic fluorination was also successful. Thus, treatment of
20,30,50-tri-O-acetyl-10-phenylselenouridine (20) with DAST/NBS produced the 10-fluorouridine triacetate (21) and its a-anomer 22.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
arabinofuranosyladenine (ara A), which is readily inacti-
vated by adenosine deaminase in vivo.4
Antimetabolic nucleoside and nucleobase analogues are
essential for anticancer and antiviral chemotherapies. More
than 50 years ago, 5-fluorouracil (5-FU, 1, Fig. 1) was found
to be a very effective drug in cancer chemotherapy,2 and
5-FU and its prodrugs3 have become some of the most clin-
ically important anticancer drugs to date. As a result, the
introduction of fluorine atoms into nucleobases has been
studied extensively. For example, arabinofuranosyl-2-fluo-
roadenine (2) was identified as a metabolically stable conge-
ner of the clinically useful antileukemic and antiviral agent
In 1969, the naturally occurring 40-fluoronucleoside, nucleo-
cidine (3) was discovered, and its potent antitrypanosomal
activity was reported.5 This finding suggested that the intro-
duction of a fluorine atom into a sugar moiety of a nucleoside
could also be an effective strategy for the development of
potent antimetabolites, leading to the synthetic studies of flu-
orosugar nucleosides. Consequently, various sugar-modified
fluoronucleosides of biological importance were identified,
e.g., fluorination at the 20-position produced the anti-herpes
NH2
N
NH2
N
NH2
NH2
N
I
N
N
N
N
O
N
O
O
O
F
S
NH
N
O
N
F
N
N
O
HO
H2N
O
O
HO
HO
O
O
N
H
O
OH
F
F
F
F
1
H3C
F
5
OH
4
HO
2
HO
3
HO
HO
R
O
NH2
N
B
HO
O
N
N
NH
F
N
N
O
N
O
N
F
HO OH(H)
HO
HO
O
O
O
I
8: R = NH2
9: R = OH
B = nucleobase
7
6
HO OH
HO
F
Figure 1. Fluoronucleobases and fluoronucleosides.
*
See Ref. 1.
* Corresponding author. Tel./fax: +81 11 706 3769; e-mail: shu@pharm.hokudai.ac.jp
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.07.101

10012
T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
virus agent 20-deoxy-20-fluoroarabinosyl-5-iodocytosine
(FIAC, 4)6 and the potent anticancer drug gemcitabine (5).7
A potent anti-HIV agent 20,30-dideoxy-30-fluorothymidine
(6) also was developed,8 in which a fluorine atom was intro-
duced at the 30-position of thymidine. The 50-fluoro-40,50-
dehydroadenosine derivative 7 was shown to be a potent
mechanism-based inhibitor of adenosylhomocysteine hydro-
lase,9 an effective target enzyme for antiviral and antimalarial
chemotherapy. In recent years, the 20-fluoro-20-deoxynucleo-
sides 8 and 9 have been recognized as useful nucleoside units
for antisense and DNA/RNA aptamer studies,10 for which
their 50-triphosphates are now commercially available.
one of the most successful leaving groups in glycosida-
tion reactions.16 An anomeric fluoro substituent would
easily eliminate under the usual Lewis acidic glycosidation
conditions, and the proper glycosyl donor for the synthesis
of 10-fluoronucleosides would be unavailable. Consequently,
we decided to examine the synthesis of the 10-fluoronucleo-
sides starting from natural nucleosides.17
We recently developed an efficient method for functionaliz-
ing the 10-position of nucleosides to synthesize several 10-
branched nucleoside analogues of biological interest, which
is shown in Scheme 1.18,19 We found that the lithium enolate
III was formed when the 20-keto-nucleoside II was treated
with LiHMDS, and that the enolate III was trapped effec-
tively with PhSeCl to give stereoselectively the correspond-
ing 10-phenylselenenyl product IV. Further treatment of IV
with SmI2 produced the samarium enolate V, condensation
of which with an aldehyde formed stereoselectively the
corresponding 10-branched aldol product VI. With these
successful results in mind, we decided to investigate the syn-
thesis of the target 10-fluoronucleosides via both electro-
philic and nucleophilic approaches, as shown in Scheme 2.
We devised a scheme to accomplish the fluorination at the
10-position using the reaction of the enolate III or V with
electrophilic fluorinating agents. Alternatively, using the
10-phenylselenonucleoside VIII as the substrate, we also
planned to examine nucleophilic fluorination via oxidative
activation of the phenylseleno moiety.
As a result, a large number of fluoronucleobase and nucleo-
side analogues have been synthesized and biologically eval-
uated as potential antimetabolites. Thus, almost all of the
hydrogens attached to carbons in natural nucleobases and
nucleosides have been chemically replaced by fluorine
atoms, and, as a consequence, the anomeric 10-position
remaining the only site in nucleosides not fluorinated. It is
likely that, with the above success stories in mind, a number
of medicinal chemists thought that the eventual compounds,
namely the 10-fluoronucleosides, would make good synthetic
targets. However, they may have considered that the 10-
fluoronucleosides would be anticipated to be chemically un-
stable. As shown in Figure 2, the nucleosides might degrade;
the electronically highly negative fluorine atom would make
the 10-carbon reactive to nucleophiles, and also the fluorine
atom might promote elimination of the nucleobase because
of its conjugative electron-donating effect due to the un-
shared electrons on the fluorine atom. Therefore, one might
speculate that the 10-fluoronucleosides would be too unstable
to be synthesized, or assuming they could be synthesized,
that they would be too unstable to be isolated. In this report
we describe the first synthesis of 10-fluoronucleoside I
(Fig. 1),11 which has long been the synthetic target as poten-
tial antimetabolites.
B
O
O
B
O
O
TIPDS
LiHMDS
THF
TIPDS
OLi
O
O
O
II
III
TIPDS = i-Pr2SiOSii-Pr2
B = nucleobase
B
O
O
B
B
HO
HO
SmI2
O
B
HO
PhSeCl
O
O
SePh
F
TIPDS
–
F
F
O
O
IV
HO
X
HO
HO
X
X
B
A
C
B
O
O
O
B
O
X = H, OH
B = nucleobase
R
RCHO
TIPDS
TIPDS
OH
OSm3+
Figure 2. Possible resonance structures of 10-fluoronucleosides.
O
O
O
VI
V
Scheme 1.
2. Results and discussion
2.1. Synthetic plan
B
O
O
The synthesis of nucleoside analogues modified at the
anomeric 10-position12,13 is not as common as that of nucleo-
side derivatives modified at the other positions in the sugar
moiety, in spite of the biological importance of nucleoside
analogues.14 Most of the 10-modified nucleosides have
been synthesized via glycosidation reactions with sugars
with a proper anomeric substituent, even though these routes
were not stereoselective.15 However, the synthesis of the
10-fluoronucleosides via glycosidation between a nucleobase
and a sugar having an anomeric fluoro substituent would
seem improbable, since an anomeric fluoro group has been
F+ (electrophile)
F– (nucleophile)
(a)
F
III or V
TIPDS
O
O
VII
B
B
F
PgO
PgO
O
O
(b)
SePh
PgO OPg
VIII
PgO
OPg
IX
Scheme 2.

T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
Table 1. Fluorination of 10 at the 10-position via its enolatea
10013
2.2. Electrophilic fluorination
Entry N–F
Temp Solvent
Yield
Ratio
reagent (ꢁC)
(11+12 (120))b (b:a)
(%)
In recent years, a variety of N–F fluorination reagents, some
of which are shown in Figure 3, have been developed.20–23
These reagents are rather stable, easy to handle and have
been shown to be very effective in electrophilic fluorination
of various compounds. We examined fluorination of the lith-
ium or samarium enolate of 30,50-O-tetraisopropyldisil-
oxane-1,3-diyl (TIPDS)-protected 20-ketouridine (10), using
these N–F reagents (Scheme 3). The results are summarized
in Table 1.
1
13
13
13
13
13
13
13
14
15
15
ꢀ78 THF
ꢀ78 THF
87
72
1:2.4
2c
3
1:6.2
1:1.4
1:1.5
1:2.7
1:1.0
ꢀ78 Toluene–THF (4:1) 77
4
5
6
7
8
9
10
ꢀ78 Et2O–THF (4:1)
ꢀ78 DMA–THF (1:2)
ꢀ78 DMF–THF (1:2)
ꢀ40 DMF–THF (4:1)
85
88
57
58
52
69
57
1.4:1
rtd
DMF–THF (4:1)
1.9:1
2.0:1
3.3:1
ꢀ40 DMF–THF (4:1)
ꢀ40 DMF–THF (9:1)
F
N
Cl
BF4
O
S
O
S
a
–
N
–
To a solution of the enolate, prepared by treating 10 with LiHMDS
(2.1 equiv) at ꢀ78 ꢁC, was slowly added a solution of a N–F reagent in
the indicated solvent.
The a (a-nucleoside)/b (b-nucleoside) ratio was based on the isolated
yields.
KHMDS was used instead of LiHMDS.
The reaction did not proceed at ꢀ40 ꢁC.
BF4
Cl
N
O O
Cl
N
F
BF4–
F
b
NFSI (13)
14
Selectfluor™ (15)
c
d
Figure 3. N–F fluorinating reagents.
presence of these additives (data not shown). We next exam-
ined the use of polar reaction solvents. N,N-dimethyl-
acetamide (DMA) was not as effective as a co-solvent,
where the undesired a-nucleoside was the major product
(entry 5). When DMF was used as a co-solvent, the ratio of
the b-isomer increased (entry 6, 57%, b/a¼1:1.0), while
an unidentified non-fluorinated product was generated to
decrease the total yield of the 10-fluorinated products. The
b-isomer was formed in preference to the a-isomer, when
the fluorination reaction was performed in a DMF/THF
(4:1) solvent (entry 7, 58%, b/a¼1.4:1). We next examined
other N–F fluorinating agents. Thus, the lithium enolate was
treated with N-fluoro-2,6-dichloropyridinium tetrafluoro-
borate (14)22 in DMF/THF. Although no fluorination pro-
ceeded at ꢀ40 ꢁC, the reaction at room temperature gave the
10-fluorinated products in 52% yield (entry 8, b/a¼1.9:1).
The desired b-selective result (b/a¼2.0:1) was observed
when 1-chloromethyl-4-fluoro-1,4-diazabicyclo[2.2.2]octane
bis(tetrafluoroborate) (F–TEDA–BF4, 15)23 was used as
the electrophile in DMF/THF (4:1) at ꢀ40 ꢁC to afford the
10-fluorinated products in 69% yield (entry 9). A similar reac-
tion with 15 performed in DMF/THF (9:1) at ꢀ40 ꢁC resulted
in further improvement of the b-selectivity (entry 10, 57%,
b/a¼3.3:1).
O
NH
N
O
O
O
1. LiHMDS
2. N-F reagent
TIPDS
O
(Table 1)
O
10
TIPDS = i-Pr2SiOSii-Pr2
O
F
O
O
O
NH
N
N
NH
N
TIPDS
O
OH
O
O
O
OH
12
O
F
TIPDS
'
O
O
11
F
O
O
O
NH
TIPDS
O
O
O
12
Scheme 3.
The lithium enolate of 10, prepared by treating 10 with
lithium hexamethyldisilazide (LiHMDS),19 was first treated
with a well-known N–F reagent, N-fluorobenzenesulfoni-
mide (NFSI, 13),21 as the electrophile at ꢀ78 ꢁC in THF.
The reaction successfully produced the expected 10-fluori-
nated 20-ketouridine derivatives in 87% yield as an anomeric
mixture of 11 and 12 (entry 1, b/a¼1:2.4). The a-nucleoside
12 was obtained mainly as the corresponding 20-hydrate 120
after purification by silica gel column chromatography.
When potassium hexamethyldisilazide (KHMDS) was used
as a base instead of LiHMDS, the percentage of the a-nucleo-
side 12 (120) increased (entry 2, b/a¼1:6.2). The effect of
solvent on the reaction was next investigated. The use of
the relatively non-polar toluene or Et2O, compared with
THF, as a co-reaction solvent somewhat increased the yield
of the desired b-anomer (entries 3 and 4). Therefore we
examined the reaction using MgBr2, Yb(OTf)3, or SmI3,
which have been known to be effective chelating agents,
as additives. However, fluorination did not proceed in the
Fluorination of the corresponding samarium enolate, pre-
pared from the 10-phenylseleno-20-ketouridine derivative
and SmI2,18 was examined with NFSI (13) as the fluorinating
electrophilic reagent in THF. Although the samarium eno-
late was more effective than the lithium enolate in the previ-
ous aldol-type condensation reaction with aldehydes as
electrophiles in terms of both yield and stereoselectivity,18
none of the fluorinated products was obtained in this case.
As described, the first fluorination at the 10-position of nucle-
osides was accomplished. We next tried the reduction of the
20-carbonyl moiety of the 10-fluoro-20-ketouridine derivative
11. After investigating various hydride reducing agents, for
example, NaBH4, NaBH4–CeCl3, DIBAL-H, and selectride,
we realized that reduction only proceeded successfully when
the b-nucleoside 11 was treated with DIBAL-H. However,
the reduction products could not be isolated probably owing
to their instability. Instead, uracil was quantitatively isolated

10014
T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
O
O
after silica gel column chromatography. Therefore, the
hydroxyl group resulting from the reduction products was
immediately protected without purification. When 11 was
treated with DIBAL-H at ꢀ78 ꢁC and an excess of Ac2O
and DMAP in THF, the expected 10-fluoro-20-O-acetylated
products were obtained in 68% yield as a diastereomeric
mixture at the 20-position (Scheme 4). The ratio of the
‘ribo’-type 16 to the ‘arabino’-type 17 was 1:4.24
NH
NH
N
N
1. Ac2O, DMAP, Et3N
2. TBAF, THF
O
O
AcO
O
O
O
3. Ac2O, DMAP, Et3N
SePh
SePh
TIPDS
91%
AcO
O
OAc
OH
Scheme 5.
O
the 5-bromo-10-fluoro-a-nucleoside 24 in 37% yield, where
the desired 10-b-nucleosidic product 21 was obtained in only
2% yield along with the corresponding a-nucleoside 22 in
5% yield. Ribonolactone tri-O-acetate (23) was produced
during this reaction, which was also produced in all of the
entries summarized in Table 2. When the reaction with
DAST/NBS was performed at lower temperature (ꢀ40 ꢁC),
thebrominationwas avoided thereby increasingthe10-fluoro-
products 21 and 22, where the 10-a-nucleoside was obtained
predominantly (21, 14% and 22, 56%, entry 2). A similar
reaction using EtCN as the solvent improved the yield of
the 10b-product (21, 29% and 22, 25%, entry 3).
O
NH
NH
N
DIBAL-H, THF
–78 °C
O
O
O
+
N
O
then
F
O
O
TIPDS
Ac2O, DMAP, THF
–78 °C—rt
68% (16:17 = 1:4)
F
TIPDS
O
OAc 16
O
O
11
O
O
NH
NR
N
O
N
O
O
O
O
O
F
TIPDS
OAc
17
OAc
OAc
TIPDS
O
O
18
O
NH
Scheme 4.
N
F
O
O
AcO
AcO
O
O
The stereochemistry of the compounds was confirmed from
NOE experiments of 17 (Fig. 4a). When the H-30 was irradi-
ated, an NOE was observed at the 6-H (1.0%), to indicate its
b-nucleoside structure. The 20-‘arabino’-type-configuration
was determined by irradiation of the H-20 to show an NOE at
the H-40 (3.9%).
F
N
NH
AcO
OAc
AcO
OAc
O
22
21
Table 2
20
F
O
AcO
AcO
O
O
O
NH
N
(a)
(b)
O
O
H
N
O
AcO
24
AcO
OAc
OAc
NH
O
F
1.0%
Br
AcO
O
O
H
O
N
23
N
O
H
H
O
Br
H
F
2.4%
Scheme 6.
OAc
24
AcO
: NOE
OAc
H
3.9%
H
17
The fluorination of 20 was next examined with AgF (entries
4 and 5). In these reactions, although the fluorination
proceeded, the undesired a-nucleoside 22 was formed pre-
dominantly. Treatment of 20 with XeF2 gave the desired
b-nucleoside 21 as the major product in 29% yield.
Figure 4. NOE data of 17 (a) and 24 (b).
2.3. Nucleophilic fluorination
We next tried nucleophilic fluorination at the 10-position
using the 10-phenylselenouridine derivative 20 as a substrate.
Since 1-phenylselenosugars have been effectively used as
glycosyl donors under oxidative conditions, we expected
that the 10-fluorination might occur by treating 10-phenyl-
selenonucleosides with a fluoro nucleophile under oxidative
conditions, such as DAST/NBS. The substrate 20 was read-
ily prepared from the 10-phenylselenouridine derivative 19,
which was obtained from uridine by the previously reported
method (Scheme 5).18
The stereochemistries of these products were confirmed by
NOE experiments as follows. In the experiments with 21
or 22, none of the NOE confirming the stereochemistry
was observed. However, as shown in Figure 4b, when the
Table 2. Fluorination of 20 at the 10-position
Entry Reagents (equiv)
Solvent Temp
Yielda
22 23
(%) (%) (%) (%)
21
24
1
2
DAST (2), NBS (2) CH2Cl2 rt
2
5
56
25
41
59
14
36
15
32
10
22
26
37
—
—
—
—
—
The results of the fluorination of 20 with nucleophilic fluo-
rinating agents are shown in Scheme 6 and Table 2. The
substrate 20 was first subjected to the reaction with DAST/
NBS conditions. Treatment of 20 with DAST (2 equiv)
and NBS (2 equiv) in CH2Cl2 at room temperature resulted
in the expected 10-fluorination. However, bromination at
the pyrimidine 5-position occurred concomitantly to give
DAST (2), NBS (2) CH2Cl2 ꢀ40 ꢁC 14
3
DAST (2), NBS (2) EtCN
ꢀ40 ꢁC 29
4b
5
AgF (2)
AgF (4)
XeF2 (1.2)
CH2Cl2 rt
CH2Cl2 rt
CH2Cl2 rt
7
9
29
6
a
Isolated yield.
The substrate 20 was recovered in 41% yield.
b

T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
10015
5.83 (dd, 1H, J¼8.3), 5.22 (dd, 1H, J¼8.7, 2.7), 4.23 (m,
1H), 4.19–4.14 (m, 2H), 1.14–1.01 (m, 28H); 13C NMR
(CDCl3, 125 MHz) d 196.38 (d, J¼25), 162.11, 148.9 (d,
J¼ca. 0), 137.65 (d, J¼13), 107.8 (d, J¼244), 103.4, 80.9,
71.0, 61.4, 17.4, 17.2, 17.2, 16.8, 16.8, 16.8, 16.7, 13.4,
13.0, 12.5, 12.4; FABHRMS calcd for C21H36FN2O7Si2
1
503.2045, found 503.2039 (MH+). 120: H NMR (CDCl3,
500 MHz) d 8.81 (br s, 1H), 7.74 (d, 1H, J¼8.3), 6.38 (s,
1H), 5.77 (dd, 1H, J¼8.3, 1.9), 4.51 (dd, 1H, J¼6.9, 1.5),
4.11–4.01 (m, 3H), 3.95 (d, 1H), 1.13–0.98 (m, 28H); 13C
NMR (CDCl3, 125 MHz) d 162.9, 151.5, 139.0, 116.0 (d,
J¼253), 102.5, 98.3 (d, J¼33), 83.5, 75.5, 62.9, 17.4, 17.3,
17.2, 17.0, 17.0, 16.7, 16.7, 13.3, 13.1, 12.7, 12.5;
FABHRMS calcd for C21H38FN2O8Si2 521.2151, found
521.2134 (MH+), calcd for C21H36FN2O7Si2 503.2045,
found 503.2060 (MH+ꢀH2O).
3.3. 2-O-Acetyl-1a-fluoro-3,5-O-(1,1,3,3-tetraisopropyl-
disiloxane-1,3-diyl)-b-D-ribo-pentofuranosyl]uracil (16)
and 1-[2-O-acetyl-1a-fluoro-3,5-O-(1,1,3,3-tetraisopro-
pyldisiloxane-1,3-diyl)-b-D-arabino-pentofuranosyl]-
uracil (17)
Figure 5. 1H NMR spectra of 21, 22, and 24.
H-40 of the 5-bromo-product 24 was irradiated, an NOE
was observed at the H-6 of the pyrimidine moiety to show
its a-nucleosidic structure. By comparing the 1H NMR
spectra of 21, 22, and 24, as shown in Figure 5, 21 and 22
were assigned as the b- and a-nucleosides, respectively,
since signals of the sugar moiety of 22 were superimposable
onto those of the a-nucleoside 24.
To a solution of 11 (75 mg, 0.15 mmol) in THF (3.0 mL) was
added a solution of DIBAL-H (1.0 M in hexane, 360 mL,
0.36 mmol) at the ꢀ78 ꢁC, and the mixture was stirred at
the same temperature for 10 min. After addition of a THF
solution (0.3 mL) containing Ac2O (2.0 M, 0.6 mmol) and
DMAP (0.2 M, 0.06 mmol), the resulting mixture was
warmed to room temperature and then partitioned between
AcOEt and H2O, and the aqueous layer was extracted with
CHCl3. The organic layers combined were dried (Na2SO4)
and evaporated. The residue was purified by column chro-
matography (Si2O, 20%, 25%, and 33% AcOEt in hexane)
to give a mixture of 16 and 17 (54 mg, 68%) as a pale yellow
solid. The diastereomers 16 (1.4 mg) and 17 (29 mg) were
obtained in a pure form from the mixture (45 mg) by prepar-
ative HPLC separation (YMC D-ODS-5-A, 250 mmꢂ
20 mm; 80% aqueous MeCN, 25 mL/min; 260 nm), fol-
lowed by column chromatography (neutral SiO2, 33%
AcOEt in hexane). 16: 1H NMR (CDCl3, 500 MHz) d 8.04
(br s, 1H, NH-3), 7.77 (d, 1H, H-6, J¼8.6), 5.95 (dd, 1H,
H-20, J¼5.7, 1.6), 5.72 (d, 1H, H-5, J¼8.6), 4.48 (dd, 1H,
H-30, J¼8.9, 5.7), 4.30 (ddd, H-40, 1H, J¼8.9, 2.4, 2.4),
4.18 (dd, 1H, H-50a, J¼13.5, 2.4), 4.02 (dd, 1H, H-50b,
J¼13.5, 2.4), 2.20 (s, 3H, 20-O-Ac), 1.09–0.94 (m, 28H,
TIPDS); 13C NMR (CDCl3, 125 MHz) d 168.7, 162.2,
148.6, 138.2 (d, J¼3.3), 115.5 (d, J¼248), 102.3, 83.6,
72.9 (d, J¼20), 67.6, 59.3, 20.5, 17.4, 17.2, 17.2, 16.9,
16.9, 16.8, 13.4, 13.0, 12.7, 12.7; FABHRMS calcd for
C23H39FN2O8Si2Na 569.2127, found 569.2132 (MNa+).
In conclusion, the first synthesis of 10-fluoronucleosides as
potential antimetabolites was achieved.25
3. Experimental
3.1. General
NMR chemical shifts are reported in parts per million down-
field from tetramethylsilane and J values are given in hertz.
The 1H NMR assignments reported for key compounds are
in agreement with 2D NMR spectra. Thin layer chromato-
graphy was done on Merck coated plates 60F254. Silica gel
chromatography was done on Merck silica gel 5715 or Kanto
Chemical silica gel 60 N (neutral). Reactions were carried
out under an argon atmosphere.
3.2. General procedure for the fluorination of 10 form-
ing 10-fluorouracil nucleosides 11 and 120 (Table 1)
A mixture of 10 (48 mg, 0.10 mmol) and LiHMDS (1.0 M
solution in THF, 210 mL, 0.21 mmol) in a solvent (2.0 mL)
was stirred at ꢀ78 ꢁC for 1 h. To the mixture, a solution of
an N–F fluorinating reagent (0.12 mmol) in the indicated
solvent (1.5 mL) was added at the indicated temperature,
and the resulting mixture was stirred at the indicated temper-
ature for 3 h. After neutralization with AcOH, the mixture
was partitioned between AcOEt and H2O, and the organic
layer was washed with brine, dried (Na2SO4), and evapo-
rated. The residue was purified by column chromatography
(SiO2, 25%, 33%, and 66% AcOEt in hexane) to give 11
1
17: H NMR (CDCl3, 500 MHz) d 8.15 (br s, 1H, NH-
3), 7.68 (d, 1H, H-6, J¼8.3), 5.88 (m, 1H, H-20), 5.71
(dd, 1H, H-5, J¼8.3, 2.4), 4.51 (m, 1H, H-30), 4.28 (m,
1H, H-40), 4.13 (dd, 1H, H-50a, J¼12.3, 3.6), 4.01 (dd,
1H, H-50b, J¼12.3, 9.7), 2.03 (s, 3H, 20-O-Ac), 1.08–
0.98 (m, 28H, TIPDS); NOE (400 MHz, CDCl3): irradiated
H-40 observed H-20 (3.9%), irradiated H-30 observed H-6
(1.0%); 13C NMR (CDCl3, 100 MHz) d 168.3, 162.8,
148.0, 138.8 (d, J¼1.7), 116.7 (d, J¼249), 101.9, 83.7,
81.0 (d, J¼46), 75.6, 61.7, 20.6, 17.6, 17.5, 17.4, 17.4,
17.0, 16.9, 16.9, 13.5, 13.3, 13.0, 12.5; FABHRMS calcd
for C23H39FN2O8Si2Na 569.2127, found 569.2133 (MNa+).
1
as a white foam and 120 as a colorless oil. 11: H NMR
(CDCl3, 500 MHz) d 8.55 (br s, 1H), 7.52 (d, 1H, J¼8.3),

10016
T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
3.4. 1-[(1S)-2,3,5-Tri-O-acetyl-1a-phenylseleno-b-D-
ribo-pentofuranosyl]uracil (20)
168.9, 167.9, 162.1, 148.1, 138.7 (d, J¼2.2), 117.1 (d,
J¼248), 102.5, 80.2, 73.1 (d, J¼46), 68.7, 61.8, 20.6, 20.3,
20.2; UV lmax 247.7 nm (THF); FABHRMS calcd for
C15H17FN2NaO9 411.0816, found 411.0812 (MNa+). 23:
1H NMR (CDCl3, 500 MHz) d 5.72 (d, 1H, H-3, J¼6.1),
5.45 (d, 1H, H-2, J¼6.1), 4.73 (t, 1H, H-4, J¼3.1), 4.40
(dd, 1H, H-5a, J¼12.7, 3.1), 4.34 (dd, 1H, H-50b, J¼12.7,
A solution of 1918 (1.93 g, 3.00 mmol), Ac2O (0.85 mL,
9.0 mmol), Et3N (1.25 mL, 9.00 mmol), and DMAP
(73 mg, 0.60 mmol) in MeCN (28 mL) was stirred at room
temperature for 1 h. The resulting mixture was partitioned
between AcOEt and H2O, and the organic layer was washed
with aqueous saturated NaHCO3 and brine, dried (Na2SO4)
and evaporated. To a solution of the residue in THF (30 mL)
was added a mixture of TBAF (1.0 M in THF, 6.0 mL,
6.0 mmol) and AcOH (45 mL) at 0 ꢁC, and the resulting mix-
ture was stirred at the same temperature for 1 h and then
evaporated. The residue was purified by column chromato-
graphy (neutral SiO2, 5% MeOH in CHCl3) to give the desi-
lylated product as a yellow solid. A solution of the obtained
solid, Ac2O (0.88 mL, 9.3 mmol), Et3N (1.3 mL, 9.3 mmol),
and DMAP (75 mg, 0.60 mmol) in MeCN (30 mL) was
stirred at room temperature for 30 min. The resulting mix-
ture was partitioned between AcOEt and H2O, the organic
layer was washed with brine, dried (Na2SO4), and evapo-
rated. The residue was purified by column chromatography
(neutral SiO2, 50% AcOEt in hexane) to give 20 (1.44 g,
91%) as a white foam: 1H NMR (CDCl3, 500 MHz) d 8.48
(br s, 1H, NH-3), 7.51–7.26 (m, 5H, PhSe), 6.93 (d, 1H,
H-6, J¼8.5), 5.98 (d, 1H, H-20, J¼6.8), 5.24 (dd, 1H, H-30,
J¼8.0, 6.8), 5.20 (dd, 1H, H-5, J¼8.5, 2.1), 4.66 (ddd, 1H,
H-40, J¼8.0, 4.5, 2.6), 4.42 (dd, 1H, H-50a, J¼12.7, 2.6),
4.28 (dd, 1H, H-50b, J¼12.7, 4.5), 2.27, 2.09, 2.05 (each s,
each 3H, each Ac); 13C NMR (CDCl3, 125 MHz) d 170.2,
169.2, 168.3, 162.3, 148.2, 138.7, 130.1, 129.1, 124.9,
101.8, 100.2, 78.1, 75.1, 68.1, 61.3, 20.7, 20.4, 20.2;
FABHRMS calcd for C21H23N2O9Se 527.0569, found
527.0568 (MH+).
1
3.1), 2.17 (s, 3H, Ac), 2.14 (s, 6H, Acx2). 24: H NMR
(CDCl3, 400 MHz) d 8.86 (br s, 1H, NH-3), 8.10 (s, 1H,
H-6), 6.19 (dd, 1H, H-20, J¼4.5, 1.5), 5.66 (ddd, 1H, H-30,
J¼8.5, 4.5, 2.7), 4.65 (dd, 1H, H-50a, J¼12.8, 2.8), 4.57
(m, 1H, H-40), 4.14 (dd, 1H, H-50b, J¼12.8, 4.6), 2.14,
2.08, 2.05 (each s, each 3H, each Ac); NOE (400 MHz,
CDCl3): irradiated H-40 observed H-6 (2.4%); 13C NMR
(CDCl3, 125 MHz) d 170.4, 168.9, 167.8, 158.0, 147.2,
138.0 (d, J¼3), 116.9 (d, J¼250), 97.5 (d, J¼2), 80.5, 73.0
(d, J¼45), 68.5 (d, J¼2), 61.6, 20.6, 20.2, 20.2; FABHRMS
calcd for C15H17FBrN2O9 467.0102, found 467.0101 (MH+).
Acknowledgements
This investigation was supported by a Grant-in-Aid for
Creative Scientific Research (13NP0401) from the Japan
Society for Promotion of Science. We are also grateful to
the Japan Society for Promotion of Sciences for support of
T.K. and to Ms. H. Matsumoto, A. Maeda, and S. Oka
(Center for Instrumental Analysis, Hokkaido University)
for technical assistance with NMR, MS, and elemental
analyses.
References and notes
1. This report constitutes Part 244 of nucleosides and nucleotides:
for Part 243, Mochiczuki, T.; Kondo, Y.; Abe, H.; Tovey, S. C.;
Dedos, S. G.; Taylor, C. W.; Paul, M.; Potter, B. V. L.; Matsuda,
S.; Shuto, S. J. Med. Chem., in press.
3.5. General procedure for the fluorination of 20 giving
10-fluorouridines 21 and 22 (Table 2)
A mixture of 20 (53 mg, 0.10 mmol) and reagent(s)
(0.20 mmol) in CH2Cl2 or EtCN (2 mL) was stirred at the
indicated temperature until 20 was disappeared on TLC.
The mixture was poured into TEAB buffer (0.1 M, pH 8.0)
at 0 ꢁC, and the whole was extracted with CHCl3. The extract
was washed with H2O, dried (Na2SO4), and evaporated, and
the residue was purified by column chromatography (neutral
SiO2, 40%, 45%, 50%, and 55% AcOEt in hexane) to give
21, 22, and 23 (and 24 in entry 1). 21 (colorless solid):
1H NMR (CDCl3, 500 MHz) d 8.97 (br s, 1H, NH-3), 7.54
(d, 1H, H-6, J¼8.5), 6.05 (dd, 1H, H-20, J¼12.4, 7.0), 5.76
(d, 1H, H-5, J¼8.5), 5.43 (dd, 1H, H-30, J¼6.9, 4.7), 4.61
(m, 1H, H-40), 4.52 (dd, 1H, H-50a, J¼12.4, 3.5), 4.28 (dd,
1H, H-50b, J¼12.4, 5.3), 2.14, 2.14, 2.11 (each s, each 3H,
each Ac); 13C NMR (CDCl3, 125 MHz) d 170.4, 169.6,
169.1, 162.7, 148.7 (d, J¼2.2), 138.3 (d, J¼13), 116.9 (d,
J¼252), 102.8, 82.0, 70.8 (d, J¼21), 69.2, 62.1, 20.7, 20.4,
20.2; UV lmax 248.6 nm (THF); FABHRMS calcd for
C15H17FN2NaO9 411.0816, found 411.0801 (MNa+). 22:
1H NMR (CDCl3, 500 MHz) d 8.52 (br s, 1H, NH-3), 7.79
(d, 1H, H-6, J¼8.3), 6.21 (dd, 1H, H-20, J¼4.5, 1.1), 5.78
(d, 1H, H-5, J¼8.3), 5.66 (ddd, 1H, H-30, J¼8.5, 4.4, 2.8),
4.62 (dd, 1H, H-50a, J¼12.7, 2.8), 4.54 (m, 1H, H-40), 4.13
(dd, 1H, H-50b, J¼12.7, 4.9), 2.12, 2.07, 2.05 (each s, each
3H, each Ac); 13C NMR (CDCl3, 125 MHz) d 170.4,
2. Heidelberger, C.; Duschinsky, R. Nature 1957, 179, 663–666.
3. (a) For examples, see: Fujii, S.; Ikenaka, K.; Fukushima, M.;
Shirasaka, T. Gann 1978, 69, 763–772; (b) Cook, A. F.;
Holman, M. J.; Kramer, P. J.; Trown, P. W. J. Med. Chem.
1979, 22, 1330–1335; (c) Miwa, M.; Ura, M.; Nishida, M.;
Sawada, N.; Ichikawa, T.; Mori, K.; Shimma, N.; Umeda, I.;
Ishitsuka, H. Eur. J. Cancer 1998, 34, 1274–1281 and refer-
ences therein.
4. Montgomery, J. A. Cancer Res. 1982, 42, 3911–3917.
5. Morton, G. O.; Lancaster, J. E.; VanLear, G. E.; Fulmor, W.;
Meryer, W. E. J. Am. Chem. Soc. 1969, 91, 1535–1537.
6. Watanabe, K. A.; Su, T.-L.; Klein, R. S.; Chu, C. K.; Matsuda,
A.; Chun, M. W.; Lopez, C.; Fox, J. J. J. Med. Chem. 1983, 26,
152–156.
7. Baker, C. H.; Banzon, J.; Bollinger, J. M.; Stubbe, J. J. Med.
Chem. 1991, 34, 1879–1884.
8. Hardewijn, P.; Balzarini, J.; De Clercq, E.; Pauwels, R.; Baba,
M.; Broder, S.; Vanderhaeghe, J. J. Med. Chem. 1987, 30,
1270–1278.
9. McCarthy, J. R.; Jarvi, E. T.; Matthews, D. P.; Edwards, M. L.;
Prakash, N. J.; Bowlin, T. L.; Mehdi, S.; Sunkara, P. S.; Bey, P.
J. Am. Chem. Soc. 1989, 111, 1127–1128.
10. For examples, see: (a) Kawasaki, A.-M.; Casper, M.-D.; Freier,
S.-M.; Lesnik, E.-A.; Zounes, M.-C.; Cummins, l.-L.;
Gonzalez, C.; Cook, P.-D. J. Med. Chem. 1993, 36, 831–841;

T. Kodama et al. / Tetrahedron 62 (2006) 10011–10017
10017
(b) Ruckman, J.; Green, L. S.; Beeson, J.; Waugh, S.; Gillette,
W. L.; Henninger, D. D.; Claesson-Welsh, L.; Janjic, N. J. Biol.
Chem. 1998, 273, 20556–20557.
297–308; (f) Ono, A.; Dan, A.; Matsuda, A. Bioconjugate
Chem. 1993, 4, 499–508; (g) Yoshimura, Y.; Otter, B. A.;
Ueda, T.;Matsuda, A. Chem. Pharm. Bull. 1992, 40, 1761–1769.
16. Tamura, S.; Abe, H.; Matsuda, A.; Shuto, S. Angew. Chem., Int.
Ed. 2003, 42, 1021–1023 and references therein.
17. Synthesis of 10-modified nucleosides from natural nucleosides,
see: (a) Itoh, Y.; Haraguchi, K.; Tanaka, H.; Gen, E.; Miyasaka,
T. J. Org. Chem. 1995, 60, 656–662; (b) Yoshimura, Y.; Kano,
F.; Miyazaki, S.; Ashida, N.; Sakata, S.; Haraguchi, K.; Itoh, Y.;
Tanaka, H.; Miyasaka, T. Nucleosides Nucleotides 1996, 15,
305–324.
18. Kodama, T.; Shuto, S.; Ichikawa, S.; Matsuda, A. J. Org. Chem.
2002, 67, 7706–7715.
19. (a) Kodama, T.; Shuto, S.; Nomura, M.; Matsuda, A.
Tetrahedron Lett. 2000, 41, 3643–3646; (b) Kodama, T.;
Shuto, S.; Nomura, M.; Matsuda, A. Chem.—Eur. J. 2001, 7,
2332–2340.
20. (a) Lal, G.-S.; Pez, G.-P.; Syvret, R.-G. Chem. Rev. 1996, 96,
1737–1755; (b) Taylar, S.-D.; Kotoris, C.-C.; Hum, G.
Tetrahedron 1999, 55, 12431–12477.
21. Differding, E.; Ofner, H. Synlett 1991, 187–189.
22. Strekowski, L.; Kiselyov, A. S. Adv. Heterocycl. Chem. 1995,
62, 1–17.
23. Banks, R. E.; Mohialdin-Khaffaf, S. N.; Lal, G. S.; Sharif, I.;
Syvret, R. G. J. Chem. Soc., Chem. Commun. 1992, 595–596.
24. The reaction was subtle, where the 10-O-acetyluridine deriva-
tive 18 (Scheme 4) was sometimes obtained.
25. Although deprotection of the synthesized 10-fluoronucleosides
was investigated under various conditions, it was unsuccessful
probably due to the instability.
11. A part of this study has been published as a communication:
Kodama, T.; Shuto, S.; Matsuda, A. Tetrahedron Lett. 2006,
47, 4429–4432.
12. For examples, see: (a) Greenberg effectively used 10-modified
nucleoside to study the mechanism of oxidative damage on
DNA: Hwang, J.-T.; Greenberg, M. M. J. Am. Chem. Soc.
1999, 121, 4311–4315 and references cited therein; (b)
Chattopadhyaya synthesized a 10-branched C30-endo-locked
nucleoside as an useful antisense unit: Boon, E.-M.; Barton,
J.-K.; Pradeepkumer, P.-I.; Isaksson, J.; Petit, C.;
Chattopadhyaya, J. Angew. Chem., Int. Ed. 2002, 41, 3402–
3405 and references therein.
13. An antibiotic angustmycin C (10-hydroxymethyladenosine) has
been known to show antitumor activity, see: Y€untsen, H.;
Ohkuma, K.; Ishii, Y. J. Antibiot. 1956, 9A, 195–201.
14. Sukeda, M.; Shuto, S.; Sugimoto, I.; Ichikawa, S.; Matsuda, A.
J. Org. Chem. 2000, 65, 8988–8996 and references cited
therein.
15. For examples, see: (a) Mahmood, K.; Vasella, A.; Bernet, B.
Helv. Chem. Acta 1991, 74, 1555–1584; (b) Elliott, R. D.;
Niwas, S.; Riordan, J. M.; Montgomery, J. A.; Secrist, J. A.,
III. Nucleosides Nucleotides 1992, 11, 97–119; (c) Faivre-
Buet, V.; Grouiller, A.; Descotes, G. Nucleosides Nucleotides
1992, 11, 1651–1660; (d) Uteza, V.; Chen, G.-H.; Toui,
J. L.-Q.; Descotes, G.; Fenet, B.; Gouiller, G. Tetrahedron
Lett. 1993, 34, 8579; (e) Hayakawa, H.; Miyazawa, M.;
Tanaka, H.; Miyasaka, T. Nucleosides Nucleotides 1994, 13,