Synthesis and evaluation of 6-methylene-bridged uracil derivatives. Part 1: Discovery of novel orally active inhibitors of human thymidine phosphorylase

Article abstract of DOI:10.1016/j.bmc.2004.04.036

A series of novel 6-methylene-bridged uracil derivatives have been prepared as inhibitors of human thymidine phosphorylase (TP). To enhance the in vivo antitumor activity of fluorinated pyrimidine 2′-deoxyribonucleosides such as 2′-deoxy-5-(trifluoromethyl)uridine (F₃dThd), a potent TP inhibitor preventing their degradation to an inactive compound, has become a target of medicinal chemistry. We present here the synthesis and evaluation of novel human TP inhibitors. Introduction of an N-substituted aminomethyl side chain at the 6-position of 5-chlorouracil has improved water solubility and enhanced inhibitory activity compared with the known TP inhibitor, 6-amino-5-chlorouracil. Compound 42 was reasonably well absorbed in mice after oral administration. When combined with F₃dThd, compound 42 exerted its TP inhibitory potency by increasing the maximum plasma concentrations of the former as evidenced in experiments with monkeys. Positive changes in pharmacokinetic profile were accompanied by the enhanced in vivo antitumor activity of this combination when compared to F₃dThd alone, in mice bearing human tumor xenografts. Both biochemical and pharmacological effects appeared to fit the concept as anticipated.

Full text of DOI:10.1016/j.bmc.2004.04.036

Bioorganic & Medicinal Chemistry 12 (2004) 3431–3441
Synthesis and evaluation of 6-methylene-bridged uracil derivatives.
Part 1: Discovery of novel orally active inhibitors of human
thymidine phosphorylase
Shingo Yano,* Hideki Kazuno, Norihiko Suzuki, Tomohiro Emura, Konstanty Wierzba,
Jun-ichi Yamashita, Yukio Tada, Yuji Yamada, Masakazu Fukushima and Tetsuji Asao
Hanno Research Center, Taiho Pharmaceutical Co., Ltd, 1-27 Misugidai, Hanno-shi, Saitama 357-8527, Japan
Received 5 March 2004; revised 28 April 2004; accepted 29 April 2004
Abstract—A series of novel 6-methylene-bridged uracil derivatives have been prepared as inhibitors of human thymidine phos-
phorylase (TP). To enhance the in vivo antitumor activity of fluorinated pyrimidine 20-deoxyribonucleosides such as 20-deoxy-5-
(trifluoromethyl)uridine (F3dThd), a potent TP inhibitor preventing their degradation to an inactive compound, has become a target
of medicinal chemistry. We present here the synthesis and evaluation of novel human TP inhibitors. Introduction of an N-
substituted aminomethyl side chain at the 6-position of 5-chlorouracil has improved water solubility and enhanced inhibitory
activity compared with the known TP inhibitor, 6-amino-5-chlorouracil. Compound 42 was reasonably well absorbed in mice after
oral administration. When combined with F3dThd, compound 42 exerted its TP inhibitory potency by increasing the maximum
plasma concentrations of the former as evidenced in experiments with monkeys. Positive changes in pharmacokinetic profile were
accompanied by the enhanced in vivo antitumor activity of this combination when compared to F3dThd alone, in mice bearing
human tumor xenografts. Both biochemical and pharmacological effects appeared to fit the concept as anticipated.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
and thymidine phosphorylase (TP). The former, UP,
predominantly expressed in rodents such as mice and
rats, and the latter, TP, is a principal enzyme in
human.9;10 Therefore, the antitumor effects of fluori-
nated pyrimidine 20-deoxyribonucleosides in human can
be improved rather by an inhibitor of TP. Furthermore,
human TP has been identified with platelet derived
endothelial cell growth factor (PD-ECGF), known as an
endogenous angiogenic factor.11 Accordingly, an inhib-
itor of human TP may also inhibit the process of angi-
ogenesis, which is closely associated with the malignant
process of solid tumors and should bring supplementary
effects in the treatment of cancer. Reported early
examples of human or horse TP inhibitors are 6-am-
inouracil derivatives such as 6-aminothymine, 6-amino-
5-bromouracil, and 6-amino-5-chlorouracil (6A5CU,
3).12–15 In our study, 6A5CU showed a moderate
inhibitory effect against human TP and rat UP.16
6A5CU showed inhibitory activity against human TP-
induced angiogenesis in mouse model.15 Hence, we had
selected 6A5CU as a lead compound, considered as the
best inhibitor at that time. Baker’s group synthesized
several 6-anilino and 6-(1-naphthylmethylamino) deriv-
atives of uracil as potent inhibitors of E. coli TP.17;18
Fluorinated pyrimidine 20-deoxyribonucleosides pos-
sessing chemotherapeutic activity, such as 20-deoxy-5-
(trifluoromethyl)uridine (F3dThd) and 20-deoxy-5-flu-
orouridine (FdUrd), are known to have strong in vitro
and in vivo antitumor activities.1;2 However, F3dThd is
easily metabolized to an inactive compound, 5-(trifluo-
romethyl)uracil while FdUrd is metabolized to still
active 5-fluorouracil by pyrimidine nucleoside phos-
phorylases.3;4 In particular, F3dThd is rapidly converted
to an inactive pyrimidine base in several animal species
including mice, monkeys, and human.5;6 Therefore, none
of the above compounds have so far been able to pro-
vide satisfactory clinical antitumor effects.7;8 To prevent
such inactivation, several efforts have been made to
develop an inhibitor of pyrimidine nucleoside phos-
phorylases, represented by uridine phosphorylase (UP)
Keywords: Thymidine phosphorylase inhibitor; TPI; 20-Deoxy-5-(tri-
fluoromethyl)uridine (F3dThd); 6-Methylene-bridged uracil derivative.
* Corresponding author. Tel.: +81-42-972-8900; fax: +81-42-972-0034;
e-mail: syano@taiho.co.jp
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.04.036

3432
S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
These compounds were recognized as 6-aminouracil
derivatives, however, appeared to be insufficient in the
inhibition of mammalian TP.19 Woodman et al. have
reported that hydrophobic bonding region adjacent to
the active site of TP differed to a great extent among
animal species.20 Therefore, it was assumed that the
design of a novel potent inhibitor should be based on the
screening data obtained with human but not E. coli TP.
Several research groups explored a variety of novel 6-
aminouracil derivatives substituted at the 1-position or
the 6-amino function as potent human TP inhibi-
tors.21–23 In our previous paper,16 we reported on 5-
chloro-6-(2-iminopyrrolidine-1-yl)methyluracil hydro-
chloride (TPI, 1) as a novel specific inhibitor of human
TP potentiating the antitumor effect of F3dThd (Fig. 1).
2. Chemistry
Scheme 1 illustrates the synthesis of 6-substituted 5-
chlorouracil derivatives. 6A5CU was prepared from 6-
aminouracil (2) in a previously described manner.24
5-Chloro-6-(chloromethyl)uracil (5) was prepared by
chlorination of 6-(chloromethyl)uracil (4) with sulfuryl
chloride in acetic acid. 6-Aminomethyl-5-chlorouracil
(6) and its N-substituted derivatives 7–23 were obtained
by reacting 5 with a large excess of aqueous ammonia or
appropriate alkylamines in water. The methoxy ana-
logue 24 was prepared by treating 5 with NaOMe in
DMF. The synthesis of compounds 27, 28, and 31 is
illustrated in Scheme 2. 2-Amino-substituted ethyl
compounds 27 and 28 were prepared from 5-chloro-6-
In this paper, we describe the synthesis and biological
evaluation of a series of novel 6-methylene-bridged
uracil derivatives and disclose the full details of the
discovery process of the lead compounds of TPI. Several
derivatives have been synthesized to enhance the TP
inhibitory activity and water solubility, since these
properties should allow dosage reduction and improve
clinical efficacy. In designing an inhibitor, we have
focused on the hydrophobic bonding region that sur-
round the active site of human TP and interact with the
6-position of thymine. The selection of a final candidate
is based on the in vitro inhibition of TP, followed by a
reasonable pharmacokinetic (PK) profile and an ability
of increasing plasma concentration of F3dThd under in
vivo conditions, using mice. This study reveals that
compound 42 elevates the plasma level of F3dThd in
mice, but the most pronounced effects are shown in
monkeys. The in vivo antitumor activity of F3dThd
combined with 42 is evaluated in mice. The implication
of the improved PK profile of F3dThd after combined
treatment with 42 in the therapeutic outcome of the
former is clearly demonstrated.
O
O
Cl
a
HN
HN
O
N
H
2
NH2
O
N
H
NH2
3; 6A5CU
O
O
Cl
b
HN
HN
Cl
Cl
O
N
H
4
O
N
H
5
c
c
O
O
Cl
Cl
HN
HN
NHR1
R2
O
N
H
O
N
H
6: R1 = H
15: R2 = N(CH3)2
7: R1 = CH3
16: R2 = N(CH2CH3)2
8: R1 = CH2CH3
17: R2 = azetidinyl
9: R1 = CH(CH3)2
10: R1 = CH2CH2CH3
11: R1 = (CH2)3OH
12: R1 = (CH2)4OH
13: R1 = CH2CH2NH2
14: R1 = (CH2)3NH2
18: R2 = pyrrolidinyl
19: R2 = piperidyl
20: R2 = hexamethyleneiminyl
21: R2 = morpholin-4-yl
22: R2 = 4-methylpiperazinyl
23: R2 = 2,5-dimethylpyrrolidinyl
O
O
Cl
HN
d
Cl
5
HN
OCH3
O
N
H
24
N
O
N
H
NH HCl
TPI; 1
Scheme 1. Reagents and conditions: (a) SO2Cl2, DMF; (b) SO2Cl2,
AcOH; (c) corresponding amine, H2O; (d) NaOMe, DMF.
Figure 1.
O
O
O
Cl
Cl
a
b
HN
HN
HN
O
N
H
Cl
O
N
H
Cl
O
N
H
R
25
26
27: R = N(CH3)2
28: R = pyrrolidyl
O
O
O
Cl
Cl
HN
HN
HN
c, a
b
OH
Cl
N
O
N
O
N
O
N
CH3
CH3
CH3
29
30
31
Scheme 2. Reagents and conditions: (a) SO2Cl2, AcOH; (b) dimethylamine or pyrrolidine, H2O; (c) SOCl2.

S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
3433
O
O
O
Cl
Cl
Cl
Cl
e
a
HN
HN
HN
N
N
N
HCl
N
N
N
Cl
O
N
H
O
N
H
O
N
H
32
36
5
b
d
c
O
O
Cl
HN
HN
O
N
NO2
N
N
Cl
O
N
H
33
O
N
H
N
HCl
HN
CH3
N
35
O
N
H
34
O
O
Cl
Cl
f
HN
HN
NH2
N
O
N
H
6
O
N
H
37
Scheme 3. Reagents and conditions: (a) N-acetylimidazole, MeOH; (b) 2-methylimidazole, KOH, H2O; (c) 4-nitroimidazole, KOH, H2O; (d) pyr-
azole, KOH, H2O; (e) 1H-1,2,4-triazole, KOH, H2O; (f) 2,5-dimethoxytetrahydrofuran, AcOH.
beyond its IC50 value. Despite the presence of some
pharmacological activity of 6A5CU, its water solubility
and potency of inhibiting human TP activity were not
sufficiently suitable to meet the criteria of good phar-
macologically active agents. The water solubility of
compound 6 was almost the same as that of 6A5CU
(0.13 vs 0.12 mg/mL). However, the hydrochloride salt
of 6 showed excellent solubility in water (>10 mg/mL).
In contrast, 6A5CU did not form a hydrochloride salt,
therefore excellent water solubility could not be
achieved. The sufficient water solubility is important for
the gastrointestinal absorption and is absolutely re-
quired for the intravenous drug formulation. Further
derivation of 6 was expected to result in an improvement
of the PK profile.
O
O
X
X
c
HN
HN
Cl
N
O
N
H
O
N
H
4: X = H
38: X = Br
39: X = I
41: X = H
42: X = Br
43: X = I
a
b
40: X = CH3
44: X = CH3
Scheme 4. Reagents and conditions: (a) NBS, DMF; (b) N-iodosuc-
cinimide, DMF; (c) pyrrolidine, H2O.
(2-chloroethyl)uracil (26), which was derived from 6-(2-
chloroethyl)uracil (25)25 by chlorination with sulfuryl
chloride. Compound 31 was prepared from the known
compound 2926 using a three-step sequence: (a) con-
version of the hydroxyl group into a chloride with
thionyl chloride, (b) chlorination at the 5-position with
sulfuryl chloride, and (c) substitution with pyrrolidine at
the allyl position.
Since 1-methyluracil, 1-methylthymine, and 1-ethylthy-
mine did not inhibit TP, Niedzwicki et al. predicted that
substitution at the 1-position of uracil with an alkyl
chain would not lead to create a potent TP inhibitor.31 It
indicated a low suitability of the analogues substituted
at the 1-position. Hence, we focused on the substitution
at the 6-position of uracil rather than the 1-position.
Scheme 3 illustrates the synthesis of compounds pos-
sessing a five-membered ring at the side chain. Com-
pound 32 was prepared by heating 5 with acetyl
imidazole27;28 in MeOH. Compounds 33–36 were pre-
pared by treating 5 with corresponding heterocycles in
the presence of KOH in water. Compound 37 was pre-
pared by treating 6 with 2,5-dimethoxytetrahydrofuran29
in acetic acid. Scheme 4 illustrates the synthesis of
5-substituted derivatives of compound 18. 5-Bromo-6-
(chloromethyl)uracil (38) and 6-chloromethyl-5-iodo-
uracil (39) were prepared by halogenation of 4 with small
excess of N-bromosuccinimide (NBS) or N-iodosuccini-
mide in DMF. 6-(Chloromethyl)thymine (40) was pre-
pared according to a previously reported procedure.30
They were converted to corresponding pyrrolidine ana-
logues 41–44 in the same manner as described for 18.
Table 1 shows the inhibitory effect of 6-methylene-
bridged uracil derivatives on human TP. Firstly, the in
vitro potency of 6 against human TP was almost the
same as that of 6A5CU (IC50, 15 lM). The potency of
N-methyl derivative 7, N-ethyl derivative 8, and N-iso-
propyl derivative 9 were close to that of 6. The inhibi-
tory effect was decreased in the case of N-propyl
derivative 10. The compounds 11–14 having a hydroxy
or an amino group at the end of alkyl side chains
appeared to be less effective in the case of introducing
longer methylene bonds. N,N-Dialkyl or N-substituted-
heterocyclic derivatives of 6 were also evaluated, and
dimethylamino derivative 15 showed the same potency
as 6, while diethylamino derivative 16 appeared to be
less potent. These observations led to the conclusion
that the hydrophobic region contacting the 6-position of
5-chlorouracil at the active site of human TP would not
exist, contrasting the results derived from E. coli TP,
3. Results and discussion
To develop the potent human TP inhibitor, its maxi-
mum plasma concentration (Cmax) value should be far

3434
S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
Table 1. Inhibitory effect of compounds 6–24, 27, 28, 32–37, and 41–44
on human TP
active site of human TP. Thirdly, 6-ethylene-bridged
5-chlorouracil derivatives 27 and 28 were evaluated.
Compound 28 containing pyrrolidine was found to be
less active than corresponding methylene-bridged
derivative 18. In contrast, compound 27 showed the
close inhibitory activity of compound 15 due to the
reduction of bulkiness. Fourthly, 1-methylated deriva-
tive 31 was confirmed to possess low TP inhibitory
activity (IC50 >100 lM) when compared with 18, and
this result was in agreement with previous prediction.
This result suggests two possibilities. Firstly, the con-
formational freedom of the 6-pyrrolidylmethyl chain
would be restricted by the N-methyl group, therefore the
preferred conformation would not be present at the
target site. Secondly, hydrogen donation from the 1-
position of 18 would play an important role for enzyme–
ligand interaction. In addition, compound 18 was more
water soluble (7.6 mg/mL) than 6A5CU. Based on these
results, we focused our study on the derivatives of 18.
O
X
HN
R
O
N
Y
a
Compd
R
X
Y
IC50
(lM)
6A5CU
6
15
23
NH2
NHCH3
NHCH2CH3
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
H
7
12
20
8
9
NHCH(CH3)2
NHCH2CH2CH3
NH(CH2)3OH
NH(CH2)4OH
NH(CH2)2NH2
NH(CH2)3NH2
N(CH3)2
32
>100
47
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
27
28
31
32
33
34
35
36
37
41
42
43
44
>100
81
>100
21
>100
N(CH2CH3)2
N-Azetidinyl
N-Pyrrolidinyl
N-Piperidyl
Among several five-membered aromatic heterocyclic
derivatives 32–37, only imidazole derivative 32 showed
good potency. In contrast to the results of 33 and 34, a
small substitution at the imidazole dramatically influ-
enced the ability to fit the active site of human TP.
2.6
2.2
120
N-Hexamethyleneiminyl
N-Morpholinyl
>100
>100
>100
>100
>100
25
N-(4-Methylpiperazinyl)
N-(2,5-Dimethylpyrrolidinyl)
OCH3
In the case of F3dThd and FdUrd, which are thought to
possess a time-dependent mode of action, it is assumed
that consecutive administration of these drugs would be
more effective than a single dose or intermittent
administration. Therefore, prolonged retention of those
drugs in plasma would be of great importance and
would be achieved by co-administration with human TP
inhibitors. To directly demonstrate the potency of these
compounds in vivo, small scale PK experiments were
performed in mice. The test compounds were adminis-
tered orally in different molar ratios together with
F3dThd. Table 2 shows the Cmax and the area under the
curve (AUC) values of each compound. Considering the
IC50 value of 6A5CU against human TP, its Cmax and
AUC0-6 at high dose such as 300 mg/kg in BALB/c mice
were not high enough to exert pharmacological effect
because of the limited absorption caused by its poor
water solubility. In contrast, compound 32 and the
hydrochloride salt of 18 had excellent water solubility
(>50 mg/mL). Hence, compounds 18 and 32 were
expected to have excellent PK profile. The Cmax of
compound 18 was much higher than that of 6A5CU,
even when administered at lower doses. The Cmax of
compound 32 was also higher than that of 6A5CU, but
lower than that of 18. Compound 18 expected to have
better pharmacological activity than compound 32
in vivo.
CH2N(CH3)2
(N-Pyrrolidinyl)methyl
N-Pyrrolidinyl
N-Imidazolyl
82
>100
1.0
N-(2-Methylimidazolyl)
N-(4-Nitroimidazolyl)
N-Pyrazolyl
28
>100
>100
>100
>100
51
N-(1,2,4-Triazolyl)
N-Pyrrolyl
N-Pyrrolidinyl
N-Pyrrolidinyl
N-Pyrrolidinyl
N-Pyrrolidinyl
Br
I
0.51
1.3
3.8
CH3
a Inhibitory effect is reported as an IC50 value, indicating the concen-
tration of the compound required to inhibit enzyme activity by 50%.
A 0.6 mM concentration of [6-3H]dThd is used as the substrate for the
TP reactions. This assay tests at least three concentrations of each
compound.
which has a wide space to fit 1-naphthylmethylamino
substitution at the 6-position of 5-chlorouracil.18
Therefore, the reduction of bulkiness of substitution at
the 6-position of 5-chlorouracil was expected to improve
the inhibitory activity, which led to the synthesis of
small-sized cyclic amines. Azetidinyl and pyrrolidinyl
derivatives 17 and 18, respectively, appeared to better
match the active site of human TP and showed almost
10-fold stronger activity than 6. Further, introduction of
larger heterocyclic amines did not provide good results
(compounds 19–22). Furthermore, dimethylated deriv-
ative 23 dramatically diminished the inhibitory activity
when compared with 18, because of tight limitation of
the active site of human TP. Secondly, changing sub-
stitution at the 6-position of 5-chlorouracil from
CH2NHCH3 to CH2OCH3 greatly diminished the
activity. Hence, the nitrogen atom was essential to fit the
Finally, derivatives substituted at the 5-position of 18
were evaluated. Uracil derivative 41 was less active, but
the potency of iodo and methyl derivatives (43 and 44)
were close to that of 18. In contrast, 5-bromo derivative
42 was almost 4-fold more potent than 18, and its water
solubility was 4.6 mg/mL. In addition, the hydrochloride
salt of 42 had excellent water solubility (>50 mg/mL).
Hence, compound 42 was selected for further in vivo
biological evaluation. The Cmax of compound 42 was

S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
3435
Table 2. Oral absorption of compounds 18, 32, 42, and 6A5CU in mice after combined administration with F3dThd at different molar ratiosa
c
d
e
Compd
Dose (mmol/kg)
Molar ratiob
Cmax (lM)
Tmax (h)
AUC0-8 (lM h)
18
0.169
0.507
0.169
0.507
0.169
0.845
0.169
1.86
1
3
1
3
1
5
100
270
17
0.5
0.5
0.5
0.5
0.25
0.5
0.5
0.5
120
440
14
32
69
80
66
42
130
690
24f
79f
340
12
6A5CU
19
a Compounds 18, 32, or 42 with F3dThd (0.169 mmol/kg) dissolved in 0.5% hydroxypropylmethylcellulose (HPMC) were orally administered to ICR
mice (male, 6 weeks old, n ¼ 3). 6A5CU suspended in 0.5% HPMC was orally administered to BALB/c mice (male, 9 weeks old, n ¼ 3).
b Molar ratio of compounds based on F3dThd.
c Maximum plasma concentration after oral dosing.
d Time to Cmax
.
e Area under the concentration–time curve for 0–8 h after oral dosing.
f AUC for 0–6 h after oral dosing.
close to that of compound 18. We have previously re-
ported that phosphorolysis of F3dThd in monkey liver
and intestine was catalyzed mainly by TP reflecting a
typical pattern for human. On the other hand, UP was a
primary enzyme degrading F3dThd in mice, therefore
the effect of the inhibition of TP may not be easily de-
tected in mice.16 However, it seemed reasonable to
confirm positive PK changes in monkeys. Compound 42
(9.25 mg/kg) was administered orally with an equimolar
amount of F3dThd (10 mg/kg) in monkeys (n ¼ 3) to
evaluate the potency of elevating the plasma concen-
tration of F3dThd. Plasma F3dThd levels increased very
markedly, with a Tmax of 2 h, a Cmax of 14.1 lg/mL, and
an AUC value of 39.6 lg h/mL. The respective parame-
ters for F3dThd administration alone were a Tmax of 1 h,
a Cmax of 0.23 lg/mL, and an AUC value of 0.28 lg h/
mL.
vided evidence that the improved antitumor effect
in vivo was accompanied by elevated plasma and human
tumor tissue F3dThd levels due to the inhibition of
degradation by TP and by UP possibly.32 These results
suggest that compound 42 has sufficient in vivo activity
to enhance F3dThd plasma concentration, and therefore
to improve its antitumor activity.
4. Conclusion
We evaluated novel 6-methylene-bridged uracil deriva-
tives as potent human TP inhibitors. Starting from the
lead compound 6A5CU, prominent improvements, both
in in vitro activity (IC50, 30-fold more potent) and oral
absorption (Cmax, >6-fold more) in mice, have been
accomplished through introduction of a methylene-
bridge and pyrrolidine ring at the 6-position and opti-
mization of the substituent at the 5-position. Our study
on 6-methylene-bridged uracil derivatives has revealed
the tight limitation at the active site of human TP. The
substitution at the 6-position of 5-chlorouracil with
methylene-bridge was matched pyrrolidine and imidaz-
ole. This finding would significantly contribute to the
development of more potent inhibitors of human TP.
In addition, an antitumor experiment was performed to
prove the concept of enhancing the in vivo potency of
F3dThd by co-administration with compound 42. Table
3 shows the results of antitumor activity of orally
administered F3dThd alone and in combination with
equimolar 42 against the AZ-521 human stomach cancer
model in nude mice. F3dThd alone inhibited tumor
growth by 45.7%, and its effect was significantly
enhanced up to 73.5% by co-administration with com-
pound 42 (F3dThd alone; p ¼ 0:033, F3dThd with 42;
p ¼ 1:5 ꢀ 10ꢁ7, F3dThd alone vs F3dThd with 42;
p ¼ 0:10 by Dunnett’s t-test). The PK experiment pro-
Careful chemical design results in synthesis of inhibi-
tors with good water solubility and improved oral
absorption as shown in the case of 42. Compound 42
has enhanced plasma concentration of F3dThd and
Table 3. Antitumor effect and plasma level of F3dThd on human gastric cancer (AZ-521) bearing nude mice with or without compound 42a
Antitumor effect
RTVb
(mean SD)
Plasma level of F3dThdd
Group
N
IRc
(%)
N
Cmax
Tmax
(h)
AUC0-8
(lg h/mL)
(lg/mL)
Control
F3dThd alone
F3dThd+42
12
6
11.32 3.59
6.14 4.19
3.00 1.06
45.7
73.5
3
3
13.0
42.5
0.5
0.5
8.16
38.5
6
a A mixture of F3dThd (50 mg/kg) and compound 42 (46.3 mg/kg) or F3dThd (50 mg/kg) alone dissolved in 0.5% HPMC was administered orally once
a day for 2 weeks.
b RTV (relative tumor volume) was calculated as follows: RTV ¼ (mean tumor volume on day 15)/(mean tumor volume on day 0).
c IR (inhibition rate of tumor growth, %) ¼ [1)(mean RTV of drug-treated group)/(mean RTV of control group) · 100].
d Blood samples (heparin-plasma) were collected at 0.5, l, 2, 4, and 8 h after oral dosing.

3436
S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
potentiated in vivo antitumor activity of F3dThd, when
administered orally to mice. The excellent properties of
42 are particularly shown in PK experiments in mon-
keys, where plasma F3dThd levels were increased very
markedly by its co-administration with 42. These bio-
logical results have prompted us to select compound 42
as a candidate for further evaluation of biological pro-
files. Due to its good oral absorption, we expect com-
pound 42 to additionally inhibit the TP-induced
angiogenesis and as a result the process of metastasis in
cancer patients.
Anal. Calcd for C5H4Cl2N2O2: C, 30.80; H, 2.07; N,
14.37. Found: C, 30.85; H, 1.99; N, 14.41.
5.1.3. 6-(Aminomethyl)-5-chlorouracil (6). Compound 5
(10.0 g, 51.3 mmol) was suspended in 200 mL of 28%
aqueous NH4OH and stirred for 7 days at room tem-
perature in a sealed tube. An insoluble material was
collected by filtration, washed with water and MeOH to
give 6.08 g of a purple solid. The crude solid was dis-
solved in 45 mL of 1 M aqueous HCl at room temper-
ature. The reaction mixture was filtered to remove
insoluble material and the filtrate was concentrated
under a vacuum. The residue was recrystallized from
water and MeOH to give 4.35 g (40%) of a hydrochlo-
ride salt of 6 as white crystals: mp 220 ꢁC dec. 1H NMR
(DMSO-d6): d 3.96 (s, 2H), 8.60 (s, 2H), 11.55 (s, 1H),
11.79 (s, 1H). FAB-MS m=z: 210 (M)H)ꢁ, 174
(M)H)HCl)ꢁ. Anal. Calcd for C5H6ClN3O2ÆHCl: C,
28.32; H, 3.33; N, 19.82. Found: C, 27.98; H, 3.31; N,
19.68. The salt (1.00 g, 4.74 mmol) was treated with 28%
aqueous NH4OH, and the precipitate was collected by
filtration and washed with water and MeOH to give
5. Experimental
5.1. Chemistry
All melting points were determined on a Yanagimoto
micromelting point apparatus and were uncorrected.
The proton nuclear magnetic resonance (1H NMR)
spectra were recorded on
a JEOL JNM-EX270
(270 MHz) spectrometer. Chemical shifts are given in d
values (ppm) using tetramethylsilane as the internal
standard. All J values are given in hertz. Mass spectra
were recorded on a JEOL JMS-SX102A spectrometer.
Elemental analyses were carried out with a Yanagimoto
C H N Corder MT-5. Reagents and solvents were ob-
tained from commercial sources and used without fur-
ther purification. Flash column chromatography was
performed with Merck silica gel 60 (art. 9385; 230–
400 mesh), and reaction progress was monitored by thin-
layer chromatography (TLC) analysis on silica gel 60
F254 plates (Merck). Visualization was done with UV
light (254 nm) or iodine. Yields were based on purified
material and were not optimized.
1
770 mg (93%) of 6 as a white solid: mp 227 ꢁC dec. H
NMR (DMSO-d6): d 3.61 (s, 2H). FAB-MS m=z: 174
(M)H)ꢁ. Anal. Calcd for C5H6ClN3O2ÆH2O: C, 31.02;
H, 4.17; N, 21.70. Found: C, 31.26; H, 4.16; N, 21.66.
5.1.4. 5-Chloro-6-[(methylamino)methyl]uracil (7). To a
suspension of 6.00 g (30.8 mmol) of 5 in 50 mL of water
was added 150 mL of 40% aqueous methylamine cooling
with an ice-bath and stirred for 4.5 h at room tempera-
ture in a sealed tube. The reaction mixture was con-
centrated under a vacuum. The residue was triturated
with MeOH, and the resulting precipitate was collected
by filtration and washed with MeOH to give 5.38 g
(92%) of 7 as a white powder: mp 197–199 ꢁC. 1H NMR
(DMSO-d6): d 2.27 (s, 3H), 3.61 (s, 2H). EI-MS (70 eV)
m=z: 189 (Mþ). Anal. Calcd for C6H8ClN3O2: C, 38.01;
H, 4.25; N, 22.16. Found: C, 37.62; H, 4.26; N, 21.94.
5.1.1. 6-Amino-5-chlorouracil (3). To a suspension of
25.0 g (197 mmol) of 6-aminouracil (2) in 155 mL of
DMF was added dropwise 16.1 mL (200 mmol) of sul-
furyl chloride over 1 h at room temperature. After 3.5 h
the reaction mixture was filtered to remove insoluble
material. To the filtrate was added ice/water (150 mL),
and the precipitate was collected by filtration and re-
crystallized from water to give 18.5 g (58%) of 3 as white
5.1.5. 5-Chloro-6-[(ethylamino)methyl]uracil (8). To a
suspension of 300 mg (1.54 mmol) of 5 in 3 mL of water
was added 5 mL of 70% aqueous ethylamine, cooling
with an ice-bath, and stirred for 2.5 h at room temper-
ature in a sealed tube. The reaction mixture was con-
1
crystals: mp >300 ꢁC. H NMR (DMSO-d6): d 6.47 (s,
2H), 10.37 (s, 1H), 10.67 (s, 1H). FAB-MS m=z: 160
(M)H)ꢁ. Anal. Calcd for C4H4ClN3O2Æ0.6H2O: C,
27.87; H, 3.04; N, 24.38. Found: C, 27.97; H, 2.74; N,
24.50.
centrated under
a vacuum. The precipitate was
triturated with EtOH, and the resulting precipitates were
collected by filtration and washed with EtOH to give
162 mg (52%) of 8 as a white powder: mp 189–191 ꢁC.
1H NMR (DMSO-d6): d 1.02 (t, 3H, J ¼ 7:1 Hz), 2.53
(q, 2H, J ¼ 7:1 Hz), 3.64 (s, 2H). FAB-MS m/z: 204
(M+H)þ. Anal. Calcd for C7H10ClN3O2Æ0.2H2O: C,
40.57; H, 5.06; N, 20.28. Found: C, 40.60; H, 5.11; N,
20.05.
5.1.2. 5-Chloro-6-(chloromethyl)uracil (5). To a suspen-
sion of 100 g (623 mmol) of 6-chloromethyluracil (4) in
500 mL of AcOH was added 75.0 mL (934 mmol) of
sulfuryl chloride at 50 ꢁC. The gas outlet was passed
through 4 M aqueous NaOH. After 2.5 h to the reaction
mixture was added ice/water (500 mL), and the precipi-
tate was collected by filtration and washed with water
and MeOH to give 101 g (83%) of 5 as a white powder:
5.1.6. 5-Chloro-6-[(isopropylamino)methyl]uracil (9). To
a solution of 1.50 g (25.4 mmol) of isopropylamine in
5 mL of water was added 500 mg (2.56 mmol) of 5 and
stirred for 11 h at room temperature. The reaction
1
mp 225 ꢁC dec. H NMR (DMSO-d6): d 4.46 (s, 2H),
11.57 (s, 1H), 11.71 (s, 1H). FAB-MS m=z: 195 (M+H)þ.

S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
3437
mixture was concentrated under a vacuum. The residue
was triturated with CHCl3, and the resulting precipitate
was collected by filtration and washed with 50% MeOH
to give 35 mg (6%) of 9 as a white solid: mp 185 ꢁC dec.
1H NMR (DMSO-d6): d 0.99 (d, 6H, J ¼ 6:1 Hz), 2.70
(sept, 1H, J ¼ 6:1 Hz), 3.64 (s, 2H). EI-MS (70 eV) m=z:
217 (Mþ). Anal. Calcd for C8H12ClN3O2Æ0.25H2O: C,
43.25; H, 5.67; N, 18.91. Found: C, 43.31; H, 5.80; N,
18.86.
C7H11ClN4O2Æ0.33H2O: C, 37.43; H, 5.23; N, 24.94.
Found: C, 37.44; H, 5.47; N, 24.94.
5.1.11. 6-{[(3-Aminopropyl)amino]methyl}-5-chlorouracil
(14). To a solution of 2.00 g (27.0 mmol) of trimethyl-
enediamine in 25 mL of water was added 500 mg
(2.56 mmol) of 5 and stirred for 6 days at room tem-
perature. The reaction mixture was concentrated under
a vacuum. The residue was triturated with EtOH, and
the resulting precipitate was collected by filtration and
washed with water to give 35 mg (6%) of 14 as a beige
5.1.7. 5-Chloro-6-[(propylamino)methyl]uracil (10). To a
solution of 1.00 g (16.9 mmol) of propylamine in 5 mL of
water was added 300 mg (1.54 mmol) of 5 and stirred for
13 h at room temperature. The reaction mixture was
concentrated under a vacuum. The residue was tritu-
rated with MeOH, and the resulting precipitate was
collected by filtration and washed with MeOH to give
1
powder: mp 135 ꢁC dec. H NMR (DMSO-d6): d 1.59
(tt, 2H, J ¼ 6:1, 6.1 Hz), 2.58 (t, 2H, J ¼ 6:1 Hz), 2.83 (t,
2H, J ¼ 6:1 Hz), 3.55 (s, 2H). FAB-MS m=z: 233
(M+H)þ. Anal. Calcd for C8H13ClN4O2Æ0.1H2O: C,
40.98; H, 5.67; N, 23.89. Found: C, 40.99; H, 5.89; N,
23.56.
1
251 mg (75%) of 10 as a white solid: mp 193–195 ꢁC. H
NMR (DMSO-d6): d 0.86 (t, 3H, J ¼ 7:3 Hz), 1.41 (tq,
1H, J ¼ 7:3, 7.3 Hz), 2.44 (t, 2H, J ¼ 7:3 Hz), 3.63 (s,
2H), 8.05 (1H, s). EI-MS (70 eV) m=z: 217 (Mþ). Anal.
Calcd for C8H12ClN3O2Æ0.2H2O: C, 43.43; H, 5.65; N,
18.99. Found: C, 43.60; H, 5.85; N, 18.98.
5.1.12. 6-[(Dimethylamino)methyl]-5-chlorouracil (15).
Compound 15 was prepared in 22% yield from 5 and
50% aqueous dimethylamine by a method similar to that
described for 7, as a yellow powder: mp 130 ꢁC dec. H
NMR (DMSO-d6): d 2.22 (s, 6H), 3.33 (s, 2H), 11.41 (s,
1H). EI-MS (70 eV) m=z: 203 (Mþ). Anal. Calcd for
C7H10ClN3O2: C, 41.29; H, 4.95; N, 20.64. Found: C,
41.25; H, 5.01; N, 20.42.
1
5.1.8. 5-Chloro-6-{[(3-hydroxypropyl)amino]methyl}ura-
cil (11). To a solution of 580 mg (7.72 mmol) of 3-ami-
no-1-propanol in 20 mL of water was added 500 mg
(2.56 mmol) of 5 and stirred for 19 h at room tempera-
ture. The reaction mixture was evaporated under a
vacuum. The residue was triturated with MeOH, and
the resulting precipitate was collected by filtration. The
crude solid was purified by flash column chromatogra-
phy (CHCl3–MeOH–Et3N, 20:1:0 to 1000:100:3) and
recrystallized from isopropanol to give 70 mg (12%) of
11 as pale yellow needles: mp 167–169 ꢁC. 1H NMR
(DMSO-d6): d 1.55 (tt, 2H, J ¼ 6:6, 6.6 Hz), 2.54 (t, 2H,
J ¼ 6:6 Hz), 3.44 (t, 2H, J ¼ 6:6 Hz), 3.63 (s, 2H). EI-
MS (70 eV) m=z: 233 (Mþ). Anal. Calcd for
C8H12ClN3O3: C, 41.12; H, 5.18; N, 17.98. Found: C,
41.12; H, 5.43; N, 18.05.
5.1.13. 6-[(Diethylamino)methyl]-5-chlorouracil (16).
Compound 16 was prepared in 33% yield from 5 and
diethylamine by a method similar to that described for
8, as a yellow powder: mp 155 ꢁC dec. 1H NMR
(DMSO-d6): d 0.97 (t, 6H, J ¼ 7:1 Hz), 2.55 (q, 4H,
J ¼ 7:1 Hz), 3.49 (s, 2H). FAB-MS m=z: 232 (M+H)þ.
Anal. Calcd for C9H14ClN3O2: C, 46.66; H, 6.09; N,
18.14. Found: C, 46.39; H, 6.28; N, 17.95.
5.1.14. 6-(Azetidinylmethyl)-5-chlorouracil (17). To a
suspension of 800 mg (4.10 mmol) of 5 in 15 mL of water
was added 713 mg (12.5 mmol) of azetidine and stirred
for 5 h at room temperature. The reaction mixture was
evaporated under a vacuum. The residue was triturated
with EtOH, and the resulting precipitate was collected
by filtration. The crude solid was purified by flash col-
umn chromatography (CHCl3–MeOH, 20:1) and tritu-
rated with AcOEt to give 355 mg (40%) of 17 as a white
5.1.9. 5-Chloro-6-{[(4-hydroxybutyl)amino]methyl}uracil
(12). Compound 12 was prepared in 2% yield from 5
and 4-amino-1-butanol by a method similar to that de-
scribed for 11, as white crystals: mp 164–166 ꢁC. 1H
NMR (DMSO-d6): d 1.38–1.48 (m, 4H), 2.46–2.51 (m,
2H), 3.38 (t, 2H, J ¼ 5:6 Hz), 3.63 (s, 2H). FAB-MS
m=z: 248 (M+H)þ. Anal. Calcd for C9H14ClN3O3: C,
43.64; H, 5.70; N, 16.97. Found: C, 43.57; H, 5.92; N,
16.72.
1
powder: mp 190–191 ꢁC. H NMR (DMSO-d6): d 1.98
(quintet, 2H, J ¼ 7:0 Hz), 3.27 (t, 4H, J ¼ 7:0 Hz), 3.46
(s, 2H), 11.23 (s, 1H). FAB-MS m=z: 216 (M+H)þ. Anal.
Calcd for C8H10ClN3O2: C, 44.56; H, 4.67; N, 19.49.
Found: C, 44.34; H, 4.72; N, 19.35.
5.1.10. 6-{[(2-Aminoethyl)amino]methyl}-5-chlorouracil
(13). Compound 13 was prepared in 70% yield from 5
and ethylenediamine by a method similar to that
described for 10, as a beige powder: mp 140 ꢁC dec.
1H NMR (DMSO-d6): d 1.38–1.48 (m, 4H), 2.46–
2.51 (m, 2H), 3.38 (t, 2H, J ¼ 5:6 Hz), 3.63 (s, 2H).
FAB-MS m=z: 219 (M+H)þ. Anal. Calcd for
5.1.15. 5-Chloro-6-(pyrrolidinylmethyl)uracil (18). Com-
pound 18 was prepared in 40% yield from 5 and pyr-
rolidine by a method similar to that described for 10, as
a white powder: mp 207–209 ꢁC. H NMR (DMSO-d6):
d 1.66–1.76 (m, 4H), 2.48–2.60 (m, 4H), 3.52 (s, 2H). EI-
MS (70 eV) m=z: 229 (Mþ). Anal. Calcd for
C9H12ClN3O2: C, 47.07; H, 5.27; N, 18.30. Found: C,
1

3438
S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
46.97; H, 5.36; N, 18.15. Compound 18 (460 mg,
2.00 mmol) was dissolved in 3 mL of 2 M aqueous HCl
at 80 ꢁC. To the reaction mixture was added 7 mL of
EtOH and allowed to stand at room temperature, and
the precipitate was collected by filtration to give 412 mg
(77%) of a hydrochloride salt of 18 as colorless needles:
MS m=z: 258 (M+H)þ. Anal. Calcd for C11H16ClN3O2:
C, 51.27; H, 6.26; N, 16.30. Found: C, 51.13; H, 6.41; N,
16.34.
5.1.21. 5-Chloro-6-(methoxymethyl)uracil (24). To a
solution of 500 mg (2.56 mmol) of 5 in 10 mL of DMF
was added 2.00 g of a 28% solution of NaOMe in MeOH
and stirred for 8 days at room temperature. The reaction
mixture was acidified with 10% HCl in MeOH, and the
precipitate was collected by filtration and washed with
water to give 25 mg (5%) of 24 as a white powder: mp
1
mp 255 ꢁC dec. H NMR (DMSO-d6): d 1.97 (s, 4H),
3.1–3.7 (m, 4H), 4.35 (s, 2H), 11.07 (s, 1H), 11.46 (s,
1H), 11.79 (s, 1H). FAB-MS m=z: 210 (M+H)þ, 230
(M+H)HCl)þ. Anal. Calcd for C9H12ClN3O2ÆHCl: C,
40.62; H, 4.92; N, 15.79. Found: C, 40.48; H, 4.94; N,
15.50.
1
>300 ꢁC. H NMR (DMSO-d6): d 3.31 (s, 3H), 4.31 (s,
2H), 11.16 (s, 1H), 11.58 (s, 1H). FAB-MS m=z: 191
(M+H)þ. Anal. Calcd for C6H7ClN2O3: C, 37.81; H,
3.70; N, 14.70. Found: C, 37.69; H, 3.69; N, 14.61.
5.1.16. 5-Chloro-6-(piperidylmethyl)uracil (19). Com-
pound 19 was prepared in 61% yield from 5 and piper-
idine by a method similar to that described for 10, as a
1
white powder: mp 195 ꢁC dec. H NMR (DMSO-d6): d
1.32–1.56 (m, 6H), 2.34–2.46 (m, 4H), 3.36 (s, 2H), 11.11
(s, 1H). EI-MS (70 eV) m=z: 243 (Mþ). Anal. Calcd for
C10H14ClN3O2: C, 49.29; H, 5.79; N, 17.24. Found: C,
49.37; H, 5.83; N, 17.15.
5.1.22. 5-Chloro-6-(chloroethyl)uracil (26). Compound
26 was prepared in 77% yield from 25 by a method
similar to that described for 5, as a white solid: mp
225 ꢁC dec. 1H NMR (DMSO-d6): d 3.01 (t, 2H,
J ¼ 6:9 Hz), 3.88 (t, 2H, J ¼ 6:9 Hz), 11.28 (s, 1H),
11.60 (s, 1H). EI-MS (70 eV) m=z: 208 (Mþ). Anal. Calcd
for C6H6Cl2N2O2Æ0.1AcOH: C, 34.63; H, 3.00; N, 13.03.
Found: C, 34.27; H, 3.02; N, 12.75.
5.1.17.
5-Chloro-6-(hexamethyleneiminylmethyl)uracil
(20). To a suspension of 800 mg (4.10 mmol) of 5 in
20 mL of water was added 1.00 g (10.1 mmol) of hexa-
methyleneimine and stirred for 2.5 h at room tempera-
ture. The precipitate was collected by filtration and
washed with water to give 495 mg (47%) of 20 as a white
5.1.23. 6-[2-(Dimethylamino)ethyl]-5-chlorouracil (27).
Compound 27 was prepared in 67% yield from 26 and
50% aqueous dimethylamine by a method similar to that
described for 7, as a white solid: mp 263 ꢁC dec.
1H NMR (CDCl3): d 2.36 (s, 6H), 2.71–2.82 (m, 4H).
FAB-MS m=z: 218 (M+H)þ. Anal. Calcd for
C8H12ClN3O2Æ0.1H2O: C, 43.79; H, 5.60; N, 19.15.
Found: C, 43.83; H, 5.79; N, 18.84.
1
powder: mp 200 ꢁC dec. H NMR (DMSO-d6): d 1.51–
1.66 (m, 8H), 2.62–2.70 (m, 4H), 3.54 (s, 2H). FAB-MS
m=z: 258 (M+H)þ. Anal. Calcd for C11H16ClN3O2: C,
51.27; H, 6.26; N, 16.30. Found: C, 51.39; H, 6.50; N,
16.37.
5.1.18. 5-Chloro-6-(morpholin-4-ylmethyl)uracil (21).
Compound 21 was prepared in 78% yield from 5 and
morpholine by a method similar to that described for 20,
as a white solid: mp 245 ꢁC dec. 1H NMR (DMSO-d6): d
2.44–2.47 (m, 4H), 3.40 (s, 2H), 3.56–3.60 (m, 4H), 10.84
(s, 1H), 11.53 (s, 1H). EI-MS (70 eV) m=z: 245 (Mþ).
Anal. Calcd for C9H12ClN3O3Æ0.1H2O: C, 43.68; H,
4.97; N, 16.98. Found: C, 43.68; H, 4.81; N, 16.89.
5.1.24. 5-Chloro-6-(2-pyrrolidinylethyl)uracil (28). To a
suspension of 210 mg (1.0 mmol) of 26 in 5 mL of water
was added 0.42 mL (5.0 mmol) of pyrrolidine and stirred
for 7 days at room temperature. The reaction mixture
was evaporated under a vacuum. The residue was
purified by flash column chromatography (CHCl3–
MeOH, 10:1) and recrystallized from MeOH to give
1
22 mg (9%) of 28 as white crystals: mp 270 ꢁC dec. H
NMR (CDCl3): d 1.89 (quintet, 4H, J ¼ 3:3 Hz), 2.68 (t,
4H, J ¼ 3:3 Hz), 2.82 (m, 2H), 2.93 (m, 2H). FAB-MS
m=z: 244 (M+H)þ. Anal. Calcd for C10H14ClN3O2: C,
49.29; H, 5.79; N, 17.24. Found: C, 49.00; H, 6.02; N,
16.90.
5.1.19. 5-Chloro-6-[(4-methylpiperazinyl)methyl]uracil
(22). Compound 22 was prepared in 1% yield from 5
and 1-methylpiperazine by a method similar to that
1
described for 10, as a beige powder: mp 205 ꢁC dec. H
NMR (DMSO-d6): d 2.14 (s, 3H), 2.35 (s, 4H), 2.50 (s,
4H), 3.33 (s, 2H). FAB-MS m=z: 259 (M+H)þ. Anal.
Calcd for C10H15ClN4O2: C, 46.43; H, 5.84; N, 21.66.
Found: C, 46.44; H, 6.05; N, 21.53.
5.1.25. 5-Chloro-6-(chloromethyl)-1-methyluracil (30). A
solution of 430 mg (2.76 mmol) of 6-(hydroxymethyl)-1-
methyluracil (29) in 10 mL of thionyl chloride was
heated at 60 ꢁC for 2 h. The reaction mixture
was evaporated under a vacuum. The precipitate was
recrystallized from MeOH to give 270 mg (56%) of 6-
(chloromethyl)-1-methyluracil as beige crystals: mp 187–
188 ꢁC. H NMR (DMSO-d6): d 3.30 (s, 3H), 4.69 (s,
2H), 5.82 (s, 1H), 11.40 (s, 1H). FAB-MS m=z: 173
(M)H)ꢁ. Anal. Calcd for C6H7ClN2O2: C, 41.28; H,
5.1.20. 6-[(2,5-Dimethylpyrrolidinyl)methyl]-5-chloroura-
cil (23). Compound 23 was prepared in 9% yield from 5
and 2,5-dimethylpyrrolidine by a method similar to that
described for 20, as a pale yellow solid: mp 220 ꢁC dec.
1H NMR (DMSO-d6): d 1.00 (d, 6H, J ¼ 5:9 Hz), 1.38
(m, 2H), 1.83 (m, 2H), 2.67 (m, 2H), 3.59 (s, 2H). FAB-
1

S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
3439
4.04; N, 16.05. Found: C, 41.29; H, 4.03; N, 16.00. To a
solution of 190 mg (1.1 mmol) of 6-(chloromethyl)-1-
methyluracil in 1.5 mL of AcOH was added 0.1 mL
(1.3 mmol) of sulfuryl chloride and stirred for 1 h at
room temperature. The reaction mixture was evaporated
under a vacuum. The residue was purified by flash col-
umn chromatography (CHCl3–MeOH, 50:1) and re-
crystallized from MeOH to give 64 mg (28%) of 30 as
white crystals: mp 222–224 ꢁC. H NMR (DMSO-d6): d
3.40 (s, 3H), 4.84 (s, 2H), 12.00 (s, 1H). EI-MS (70 eV)
m=z: 208 (Mþ). Anal. Calcd for C6H6Cl2N2O2: C, 34.48;
H, 2.89; N, 13.40. Found: C, 34.43; H, 2.83; N, 13.36.
d6): d 5.18 (s, 2H), 7.95 (d, 1H, J ¼ 1:3 Hz), 8.43 (d, 1H,
J ¼ 1:3 Hz), 11.50 (s, 1H), 11.70 (s, 1H). FAB-MS m=z:
270 (M)H)ꢁ. Anal. Calcd for C8H6ClN5O4Æ1.25H2O: C,
32.67; H, 2.91; N, 23.81. Found: C, 32.87; H, 2.73; N,
23.60.
5.1.30. 5-Chloro-6-(pyrazolylmethyl)uracil (35). Com-
pound 35 was prepared in 17% yield from 5 and pyr-
azole by a method similar to that described for 33, as a
1
1
yellow solid: mp 220 ꢁC dec. H NMR (DMSO-d6): d
5.17 (s, 2H), 6.29 (dd, 1H, J ¼ 2:3, 1.5 Hz), 7.50 (d, 1H,
J ¼ 1:5 Hz), 7.86 (d, 1H, J ¼ 2:3 Hz), 11.53 (s, 1H),
11.68 (s, 1H). FAB-MS m=z: 227 (M+H)þ. Anal. Calcd
for C8H7ClN4O2: C, 42.40; H, 3.11; N, 24.72. Found: C,
42.01; H, 3.05; N, 24.39.
5.1.26.
5-Chloro-1-methyl-6-(pyrrolidinylmethyl)uracil
(31). To a suspension of 25 mg (0.12 mmol) of 30 in
0.5 mL of water was added 43 mg (0.60 mmol) of pyr-
rolidine and stirred for 14 h at room temperature. The
pH of the reaction mixture was adjusted to 5 by addition
of AcOH, and the precipitate was collected by filtration
and recrystallized from MeOH to give 12 mg (41%) of 31
as colorless needles: mp 222–224 ꢁC. 1H NMR (DMSO-
d6): d 1.70 (s, 4H), 2.57 (s, 4H), 3.43 (s, 3H), 3.74 (s, 2H),
11.81 (s, 1H). EI-MS (70 eV) m=z: 243 (Mþ). Anal. Calcd
for C10H14ClN3O2: C, 49.29; H, 5.79; N, 17.24. Found:
C, 49.28; H, 5.93; N, 17.26.
5.1.31. 5-Chloro-6-(1,2,4-triazolylmethyl)uracil (36). A
mixture of 300 mg (1.54 mmol) of 5 and 327 mg
(4.73 mmol) of 1H-1,2,4-triazole in 5 mL of 0.75 M
aqueous KOH was heated under reflux for 3 h. The
reaction mixture was allowed to stand at room tem-
perature and evaporated under a vacuum. The residue
was triturated with MeOH, and the resulting precipitate
was collected by filtration. The crude solid was purified
by flash column chromatography (CHCl3–MeOH, 10:1)
and triturated with CHCl3 to give 50 mg (14%) of 36 as a
5.1.27. 5-Chloro-6-(imidazolylmethyl)uracil hydrochlo-
ride (32). A mixture of 5.00 g (25.6 mmol) of 5 and
4.30 g (39.0 mmol) of N-acetylimidazole in 100 mL of
MeOH was heated under reflux for 2 days. The reaction
mixture was allowed to stand at room temperature, and
the precipitate was collected by filtration and washed
with 30 mL of 10% HCl in MeOH to give 4.32 g (64%) of
32 as a white powder: mp 244–246 ꢁC. 1H NMR
(DMSO-d6): d 5.40 (s, 2H), 7.74 (d, 1H, J ¼ 1:3 Hz),
7.82 (d, 1H, J ¼ 1:3 Hz), 9.29 (s, 1H), 11.74 (s, 1H),
11.78 (s, 1H). FAB-MS m=z: 227 (M+H)HCl)þ. Anal.
Calcd for C8H7ClN4O2ÆHClÆH2O: C, 34.18; H, 3.59; N,
19.93. Found: C, 34.27; H, 3.27; N, 19.85.
1
white powder: mp 235 ꢁC dec. H NMR (DMSO-d6): d
5.26 (s, 2H), 8.03 (s, 1H), 8.64 (s, 1H), 11.57 (s, 1H),
11.70 (s, 1H). FAB-MS m=z: 228 (M+H)þ. Anal. Calcd
for C7H6ClN5O2Æ0.2H2O: C, 36.36; H, 2.79; N, 30.29.
Found: C, 36.74; H, 2.71; N, 29.96.
5.1.32. 5-Chloro-6-(pyrrolylmethyl)uracil (37). A mixture
of 500 mg (2.85 mmol) of 6 and 577 mg (4.37 mmol) of
2,5-dimethoxytetrahydrofuran in 8 mL of AcOH was
heated under reflux for 2 h. The reaction mixture was
allowed to room temperature and evaporated under a
vacuum. The residue was purified by flash column
chromatography (CHCl3–MeOH, 50:1) and triturated
with i-PrOH to give 155 mg (24%) of 37 as a beige
powder: mp 210 ꢁC dec. 1H NMR (DMSO-d6): d 4.92 (s,
2H), 6.04 (d, 2H, J ¼ 1:8 Hz), 6.87 (d, 2H, J ¼ 1:8 Hz),
11.61 (s, 2H). FAB-MS m=z: 226 (M+H)þ. Anal. Calcd
for C9H8ClN3O2Æ0.25H2O: C, 46.97; H, 3.72; N, 18.26.
Found: C, 47.26; H, 3.68; N, 17.87.
5.1.28.
5-Chloro-6-[(2-methylimidazolyl)methyl]uracil
hydrochloride (33). A mixture of 500 mg (2.56 mmol) of
5 and 631 mg (7.68 mmol) of 2-methylimidazole in
10 mL of 1 M aqueous KOH was heated at 80 ꢁC for 3 h.
The reaction mixture was allowed to stand at room
temperature and the pH was adjusted to 4 by addition of
6 M aqueous HCl. The precipitate was collected by fil-
tration and washed with water to give 50 mg (8%) of 33
1
as a beige powder: mp 240 ꢁC dec. H NMR (DMSO-
5.1.33. 5-Bromo-6-(chloromethyl)uracil (38). To a solu-
tion of 100 g (623 mmol) of 6-chloromethyluracil (4) in
1.0 L of DMF was added 122 g (685 mmol) of N-
bromosuccinimide cooling with an ice-bath and stirred
for 1 h. To the reaction mixture was added ice/water
(1.0 L), and the precipitate was collected by filtration
and washed with AcOH and water to give 126 g (84%) of
38 as a white powder: mp 245 ꢁC dec. 1H NMR (DMSO-
d6): d 4.47 (s, 2H), 11.61 (s, 1H), 11.66 (s, 1H). EI-MS
(70 eV) m=z: 240 (Mþ). Anal. Calcd for C5H4BrClN2O2:
C, 25.08; H, 1.68; N, 11.70. Found: C, 24.81; H, 1.67; N,
11.57.
d6): d 2.60 (s, 3H), 5.22 (s, 2H), 7.54 (d, 1H, J ¼ 2:0 Hz),
7.61 (d, 1H, J ¼ 2:0 Hz), 11.77 (s, 1H). FAB-MS
m=z: 241 (M+H)HCl)þ. Anal. Calcd for C9H9
ClN4O2ÆHClÆ1.25H2O: C, 36.08; H, 4.20; N, 18.71.
Found: C, 36.03; H, 3.94; N, 18.44.
5.1.29. 5-Chloro-6-[(4-nitroimidazolyl)methyl]uracil (34).
Compound 34 was prepared in 12% yield from 5 and 4-
nitroimidazole by a method similar to that described for
1
33, as a beige solid: mp 155–158 ꢁC. H NMR (DMSO-

3440
S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
5.1.34. 5-Bromo-6-(pyrrolidinylmethyl)uracil (42). To a
solution of 178 g (2.50 mol) of pyrrolidine in 1.0 L of
water was added 120 g (0.501 mol) of 38 and stirred for
45 min at 35 ꢁC. The pH of the reaction mixture was
adjusted to 7 by addition of 120 g (2.0 mol) of AcOH,
and the precipitate was collected by filtration and
washed with water and MeOH to give 112 g (82%) of 42
C10H15N3O2Æ0.2H2O: C, 56.43; H, 7.29; N, 19.74.
Found: C, 56.39; H, 7.36; N, 19.62.
5.2. Water solubility
Solubility of the compounds was determined in water
after shaking for 24 h at 23 ꢁC. The water-soluble con-
centration was determined by the HPLC method after
separation of soluble fraction.
1
as a white powder: mp 213–215 ꢁC. H NMR (DMSO-
d6): d 1.64–1.79 (m, 4H), 2.52–2.63 (m, 4H), 3.55 (s, 2H).
FAB-MS m=z: 274 (M+H)þ. Anal. Calcd for
C9H12BrN3O2: C, 39.44; H, 4.41; N, 15.33. Found: C,
39.54; H, 4.44; N, 15.49. Compound 42 (10.0 g,
36.5 mmol) was dissolved in 60 mL of 2 M aqueous HCl
at 80 ꢁC. To the reaction mixture was added 150 mL of
EtOH and allowed to stand at room temperature, and
the precipitate was collected by filtration to give 9.90 g
(87%) of a hydrochloride salt of 42 as colorless needles:
mp 255 ꢁC dec. 1H NMR (DMSO-d6): d 1.7–2.2 (m, 4H),
3.2–3.7 (m, 4H), 4.34 (s, 2H), 10.91 (s, 1H), 11.48 (s,
1H), 11.73 (s, 1H). FAB-MS m=z: 274 (M+H)HCl)þ.
Anal. Calcd for C9H12BrN3O2ÆHCl: C, 34.81; H, 4.22;
N, 13.53. Found: C, 34.76; H, 4.19; N, 13.30.
5.3. Biology
5.3.1. Assay of human TP activity. TP protein was par-
tially purified from human placenta, and TP activity was
measured by using [6-3H]dThd as the substrate accord-
ing to a previously described method.16 The reaction
mixture consisted of a total volume of 0.125 mL con-
taining 100 mM potassium phosphate buffer (pH 7.4),
0.6 mM of [6-3H]dThd containing 9.25 kBq of radio-
labeled compound, TP, and inhibitor solution, and the
mixture was incubated at 37 ꢁC for 5 min. The reaction
was terminated by heating at 100 ꢁC for 2 min in boiling
water, and the reaction mixture was centrifuged at 1900g
for 5 min. A 10 lL sample of the supernatant was
spotted on a silica gel TLC plate and developed with a
mixture of CHCl3, MeOH, and AcOH (17:3:1 by vol). A
mixture of 5 mM dThd and thymine was applied to the
plate as a visible marker before loading of the test
samples. The thymine spot was separated into a vial and
extracted with 4 M HCl (0.1 mL). After adding 10 mL of
scintillator, the radioactivity was measured.
5.1.35. 6-(Chloromethyl)-5-iodouracil (39). Compound
39 was prepared in 92% yield from 4 and N-iodosuc-
cinimide by a method similar to that described for 38, as
a white powder: mp 225 ꢁC dec. 1H NMR (DMSO-d6): d
4.49 (s, 2H), 11.52 (s, 1H), 11.58 (s, 1H). FAB-MS m=z:
285 (M)H)ꢁ. Anal. Calcd for C5H4ClIN2O2: C, 20.96;
H, 1.41; N, 9.78. Found: C, 21.10; H, 1.36; N, 9.87.
5.1.36. 5-Iodo-6-(pyrrolidinylmethyl)uracil (43). To a
solution of 740 mg (10.5 mol) of pyrrolidine in 10 mL of
water was added 1.0 g (3.5 mmol) of 39 and stirred for
20 h at room temperature. The reaction mixture was
evaporated under a vacuum, and the precipitate was
recrystallized from MeOH to give 195 mg (17%) of 43 as
5.3.2. Oral absorption in mice. Compound 3 (300 mg/kg)
suspended in 0.5% hydroxypropylmethylcellulose
(HPMC) was orally administered to BALB/c mice
(male, 9 weeks old, n ¼ 3). Compounds 18 (38.8 or
116.4 mg/kg), 32 (46.4 or 139.3 mg/kg), and 42 (46.3 or
231.4 mg/kg) with F3dThd (50 mg/kg) dissolved in 0.5%
HPMC were orally administered to ICR mice (male,
6 weeks old, n ¼ 3). Blood samples (heparinized plasma)
were collected at 0.5, l, 2, 4, and 8 h after administration.
The extraction of plasma samples was performed using
50% acetonitrile, 75% MeOH, or 10% trichloroacetic
acid added to the plasma (0.2 mL), and the precipitated
plasma proteins were removed by centrifugation
(3000 rpm, 10 min). These protein precipitating solvents
were used depending on the compound tested. The
compound concentrations in the supernatant were
measured by reverse-phase HPLC using a Chemcosorb
300-5C18 column (4.6 mm · 250 mm, Chemco Co., Ltd)
under the following chromatographic conditions:
mobile phase, 6% acetonitrile–potassium phosphate
buffer (pH ¼ 7.2); flow rate, 1.0 mL/min; detection, 272
or 300 nm; column temperature, room temperature.
1
pale yellow needles: mp 178 ꢁC dec. H NMR (DMSO-
d6): d 1.64–1.79 (m, 4H), 2.49–2.57 (m, 4H), 3.57 (s, 2H),
11.36 (s, 1H). FAB-MS m=z: 322 (M+H)þ. Anal. Calcd
for C9H12IN3O2: C, 33.66; H, 3.77; N, 13.09. Found: C,
33.73; H, 3.89; N, 13.05.
5.1.37. 6-(Pyrrolidinylmethyl)uracil (41). Compound 41
was prepared in 9% yield from 4 and pyrrolidine by a
method similar to that described for 43, as pale yellow
1
crystals: mp 177–179 ꢁC. H NMR (DMSO-d6): d 1.68–
1.76 (m, 4H), 2.42–2.55 (m, 4H), 3.28 (s, 2H), 5.44 (s,
1H), 10.68 (s, 2H). FAB-MS m=z: 196 (M+H)þ. Anal.
Calcd for C9H13N3O2Æ0.8H2O: C, 51.57; H, 7.02; N,
20.04. Found: C, 51.59; H, 6.94; N, 19.73.
5.1.38. 6-(Pyrrolidinylmethyl)thymine (44). Compound
44 was prepared in 25% yield from 40 and pyrrolidine by
a method similar to that described for 43, as white
crystals: mp 196–198 ꢁC. H NMR (DMSO-d6): d 1.67–
1.75 (m, 4H), 1.78 (s, 3H), 2.45–2.55 (m, 4H), 3.38 (s,
2H). EI-MS (70 eV) m=z: 209 (Mþ). Anal. Calcd for
5.3.3. Oral absorption in cynomolgus monkeys. Com-
pound 42 (9.25 mg/kg) and F3dThd (10 mg/kg) dissolved
in 0.5% HPMC were orally administered to cynomolgus
monkeys (male and female, 2.90–3.65 kg, n ¼ 3). Blood
1

S. Yano et al. / Bioorg. Med. Chem. 12 (2004) 3431–3441
3441
8. Wilson, W. L.; Bisel, H. F.; Krementz, E. T.; Lien, R. C.;
Prohaska, J. V. Cancer Chemother. Rep. 1967, 51, 85.
9. el Kouni, M. H.; el Kouni, M. M.; Naguib, F. N. M.
Cancer Res. 1993, 53, 3687.
10. Maehara, Y.; Sakaguchi, Y.; Kusumoto, T.; Kusumoto,
H.; Sugimachi, K. J. Surg. Oncol. 1989, 42, 184.
11. Furukawa, T.; Yoshimura, A.; Sumizawa, T.; Haraguchi,
M.; Akiyama, S.; Fukui, K.; Ishizawa, M.; Yamada,
Y. Nature 1992, 356, 668.
12. Langen, P.; Etzold, G.; Barwolff, D.; Preussel, B. Biochem.
Pharmacol. 1967, 16, 1833.
13. Desgranges, C.; Razaka, G.; Rabaud, M.; Picard, P.;
Dupuch, F.; Bricaud, H. Biochem. Pharmacol. 1982, 31,
2755.
14. Desgranges, C.; Razaka, G.; Rabaud, M.; Bricaud, H.;
Balzarini, J.; De Clercq, E. Biochem. Pharmacol. 1983, 32,
3583.
15. Miyadera, K.; Sumizawa, T.; Haraguchi, M.; Yoshida, H.;
Konstanty, W.; Yamada, Y.; Akiyama, S. Cancer Res.
1995, 55, 1687.
16. Fukushima, M.; Suzuki, N.; Emura, T.; Yano, S.;
Kazuno, H.; Tada, Y.; Yamada, Y.; Asao, T. Biochem.
Pharmacol. 2000, 59, 1227.
17. Baker, B. R.; Rzeszotarski, W. J. Med. Chem. 1967, 10,
1109.
samples (heparinized plasma) were collected from a
forearm vein 0.5, l, 2, 4, and 8 h after administration.
The plasma concentrations of F3dThd were measured in
the same manner as previously described.16
5.3.4. Oral absorption and in vivo antitumor effect on
human gastric cancer (AZ-521) bearing nude mice.
Approximate 2 · 2 mm fragments of AZ-521 tumor
were implanted into right axillae of 8 weeks-old male
nude mice. When the tumor volume reached about
200 mm3 (day 0), a mixture of F3dThd (50 mg/kg) and
compound 42 (46.3 mg/kg) or F3dThd (50 mg/kg) alone
dissolved in 0.5% HPMC was administered orally once a
day for 2 weeks. Relative tumor volume (RTV) was
calculated as follows: RTV ¼ (mean tumor volume on
day 15)/(mean tumor volume on day 0). The antitumor
effects of F3dThd were estimated by the following
equation: inhibition rate of tumor growth (IR,
%) ¼ [1)(mean RTV of drug-treated group)/(mean RTV
of control group) · 100]. Blood samples (heparin-plas-
ma) were collected at 0.5, l, 2, 4, and 8 h after admin-
istration. The plasma concentrations of F3dThd were
measured in the same manner as previously described.16
18. Baker, B. R.; Hopkins, S. E. J. Med. Chem. 1970, 13, 87.
19. Baker, B. R.; Kelly, J. L. J. Med. Chem. 1971, 14, 812.
20. Woodman, P. W.; Sarrif, A. M.; Heidelberger, C.
Biochem. Pharmacol. 1980, 29, 1059.
21. Pan, B. C.; Chen, Z. H.; Chu, E.; Chu, M. Y. W.; Chu,
S. H. Nucleos. Nucleot. 1998, 17, 2367.
Acknowledgements
22. Focher, F.; Ubiali, D.; Pregnolato, M.; Zhi, C.; Gambino,
J.; Wright, G. E.; Spadari, S. J. Med. Chem. 2000, 43,
2601.
23. Klein, R. S.; Lenzi, M.; Lim, T. H.; Hotchkiss, K. A.;
Wilson, P.; Schwartz, E. L. Biochem. Pharmacol. 2001, 62,
1257.
24. Schroeder, E. F. U.S. Patent 3,080,364, 1963; Chem. Abstr.
1963, 59, 6420h.
25. Delia, T. J.; Munslow, W. D. J. Heterocycl. Chem. 1979,
16, 239.
We are grateful to Professor Kei Takeda of Hiroshima
University for giving us helpful advice.
We thank Dr. Kazuhiro Kobayashi, Mr. Tomio
Yamazaki, and Mr. Toshiharu Mutoh for performing
mass spectra and elemental analysis studies, and Mr.
Hideyuki Ohshimo and Mr. Masahiro Yamaguchi for
supporting the biochemical and pharmacological
experiments.
26. Burdi, D.; Begley, T. P. J. Am. Chem. Soc. 1991, 113,
7768.
27. Kamijo, T.; Yamamoto, R.; Harada, H.; Iizuka, K. Chem.
Pharm. Bull. 1982, 30, 4242.
References and notes
28. Kamijo, T.; Yamamoto, R.; Harada, H.; Iizuka, K. Chem.
Pharm. Bull. 1983, 31, 1213.
29. Blache, Y.; Gueiffier, A.; Chavignon, O.; Teulade, J. C.;
Milhavet, J. C.; Viols, H.; Chapat, J. P.; Dauphin, G.
J. Heterocycl. Chem. 1994, 31, 161.
30. Johnson, T. B.; Chernoff, L. H. J. Am. Chem. Soc. 1913,
35, 585.
1. Umeda, M.; Heidelberger, C. Cancer Res. 1968, 28,
2529.
2. Heidelberger, C.; Anderson, S. W. Cancer Res. 1964, 24,
1979.
3. Nakayama, C.; Wataya, Y.; Meyer, R. B.; Santi, D. V.;
Saneyoshi, M.; Ueda, T. J. Med. Chem. 1980, 23, 962.
4. Woodman, P. W.; Sarrif, A. M.; Heidelberger, C. Cancer
Res. 1980, 40, 507.
5. Rogers, W. I.; Hartman, A. C.; Palm, P. E.; Okstein, C.;
Kensler, C. J. Cancer Res. 1969, 29, 953.
6. Dexter, D. L.; Wolberg, W. H.; Ansfield, F. J.; Helson, L.;
Heidelberger, C. Cancer Res. 1972, 32, 247.
7. Ansfield, F. J.; Ramirez, G. Cancer Chemother. Rep. 1971,
55, 205.
31. Niedzwicki, J. G.; el Kouni, M. H.; Chu, S. H.; Cha, S.
Biochem. Pharmacol. 1983, 32, 399.
32. Yano, S.; Kazuno, H.; Sato, T.; Suzuki, N.; Emura, T.;
Wierzba, K.; Yamashita, J.; Tada, Y.; Yamada, Y.;
Fukushima, M.; Asao, T. Bioorg. Med. Chem., 2004, 12,
see doi: 10.1016/j.bmc.2004.04.046. The Rat UP IC50 value
of compound 42 was 14 lM. Compound 42 showed a high
enough Cmax value in mice to exercise the inhibition of UP
in vivo.