Novel 6-substituted uracil analogs as inhibitors of the angiogenic actions of thymidine phosphorylase

Article abstract of DOI:10.1016/S0006-2952(01)00783-3

Thymidine phosphorylase (TP) catalyzes the reversible phosphorolysis of thymidine and other pyrimidine 2′-deoxyribonucleosides. In addition, TP has been shown to possess angiogenic activity in a number of in vitro and in vivo assays, and its angiogenic activity has been linked to its catalytic activity. A series of 5- and 6-substituted uracil derivatives were synthesized and evaluated for their abilities to inhibit TP activity. Among the most active compounds was a 6-amino-substituted uracil analog, 6-(2-aminoethyl)amino-5-chlorouracil (AEAC), which was a competitive inhibitor with a K_i of 165 nM. The inhibitory activity of AEAC was selective for TP, as it did not inhibit purine nucleoside phosphorylase or uridine phosphorylase at concentrations up to 1 mM. Human recombinant TP induced human umbilical vein endothelial cell (HUVEC) migration in a modified Boyden chamber assay in vitro, and this action could be abrogated by the TP inhibitors. The actions of the inhibitors were specific for TP, as they had no effect on the chemotactic actions of vascular endothelial growth factor (VEGF). HUVEC migration was also induced when TP-transfected human colon and breast carcinoma cells were co-cultured in the Boyden chamber assay in place of the purified angiogenic factors, and a TP inhibitor blocked the tumor cell-mediated migration almost completely. These studies suggest that inhibitors of TP may be useful in pathological conditions that are dependent upon TP-driven angiogenesis.

Full text of DOI:10.1016/S0006-2952(01)00783-3

Biochemical Pharmacology 62 (2001) 1257–1263
Novel 6-substituted uracil analogs as inhibitors of the angiogenic
actions of thymidine phosphorylase
Robert S. Klein, Michelle Lenzi, Timothy H. Lim, Kylie A. Hotchkiss, Phyllis Wilson,
Edward L. Schwartz*
Department of Oncology, Albert Einstein Cancer Center, 111 East 210th St., Bronx, NY 10467, USA
Received 11 October 2000; accepted 11 April 2001
Abstract
Thymidine phosphorylase (TP) catalyzes the reversible phosphorolysis of thymidine and other pyrimidine 2Ј-deoxyribonucleosides. In
addition, TP has been shown to possess angiogenic activity in a number of in vitro and in vivo assays, and its angiogenic activity has been
linked to its catalytic activity. A series of 5- and 6-substituted uracil derivatives were synthesized and evaluated for their abilities to inhibit
TP activity. Among the most active compounds was a 6-amino-substituted uracil analog, 6-(2-aminoethyl)amino-5-chlorouracil (AEAC),
which was a competitive inhibitor with a K of 165 nM. The inhibitory activity of AEAC was selective for TP, as it did not inhibit purine
i
nucleoside phosphorylase or uridine phosphorylase at concentrations up to 1 mM. Human recombinant TP induced human umbilical vein
endothelial cell (HUVEC) migration in a modified Boyden chamber assay in vitro, and this action could be abrogated by the TP inhibitors.
The actions of the inhibitors were specific for TP, as they had no effect on the chemotactic actions of vascular endothelial growth factor
(VEGF). HUVEC migration was also induced when TP-transfected human colon and breast carcinoma cells were co-cultured in the Boyden
chamber assay in place of the purified angiogenic factors, and a TP inhibitor blocked the tumor cell-mediated migration almost completely.
These studies suggest that inhibitors of TP may be useful in pathological conditions that are dependent upon TP-driven angiogenesis.
©
2001 Elsevier Science Inc. All rights reserved.
Keywords: Thymidine phosphorylase; Angiogenesis; Endothelial cells
1
. Introduction
angiogenesis. TP was found to have a chemotactic effect on
endothelial cells and had angiogenic activity in several in
vivo assays [2–6]. When transfected into 3T3 or MCF7
carcinoma cells, the TP gene increased the vascularization
and growth of tumors growing in nude mice, but it had no
effect on the growth of the cells in vitro [6,7]. Western blot
analysis of primary human malignancies, including colon,
esophageal, gastric, breast, bladder, ovarian, and lung can-
cers, showed that TP expression was elevated up to 10-fold
in the tumors when compared with the corresponding non-
neoplastic regions of the same organs. In the majority of
these studies, higher TP expression was linked to increases
in angiogenesis, invasion, metastasis, and to shorter patient
survival [8–21]. In studies of GI cancers in which multiple
angiogenic factors were examined, TP was found to be an
independent prognostic marker of tumor aggressiveness
A number of polypeptide factors that are produced by
tumor epithelial and stromal cells and that promote angio-
genic activity have been identified, including VEGF, PD-
ECGF, FGF, TGF, and TNF-␣ [1]. The angiogenic factor
PD-ECGF has been found to be identical to human TP, an
enzyme that catalyzes the reversible conversion of thymi-
dine to thymine [2–5]. Several lines of evidence have es-
tablished the role of TP in experimental and clinical tumor
*
Corresponding author. Tel.: ϩ1-718-920-4015; fax: ϩ1-718-882-
4
464.
E-mail address: eschwart@aecom.yu.edu (E.L. Schwartz).
Abbreviations: AEAC, 6-(2-aminoethyl)amino-5-chlorouracil; CIMU,
-chloro-6-(1-imidazolyl-methyl) uracil; FGF, fibroblast growth factor;
5
HUVEC, human umbilical vein endothelial cell(s); PD-ECGF, platelet-
derived endothelial cell growth factor; PNP, purine nucleoside phosphor-
ylase; TGF, transforming growth factor; TNF-␣, tumor necrosis factor-␣;
TP, thymidine phosphorylase; UP, uridine phosphorylase; and VEGF,
vascular endothelial growth factor.
[18,19].
TP has a well documented role in the salvage biosynthe-
sis of pyrimidines. It also readily catabolizes and inactivates
a number of 5-substituted 2Ј-deoxyuridines, including sev-
0
006-2952/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved.
PII: S0006-2952(01)00783-3

1
258
R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
eral with antiviral or anticancer activities; therefore, the
identification of a TP inhibitor as a potential therapeutic
agent has been a medicinal chemistry objective for over 30
years [22–26]. Despite these efforts, only compounds with
modest potency, IC50 values Ն 30 ␮M, were identified
Table 1
Inhibition of TP activity by 5- and 6-substituted uracil analogs
[27–30]. Recent studies using site-directed mutagenesis
analysis suggested that the angiogenic activity of TP was
also linked to its enzymatic activity [7,27]. This has
prompted a renewed interest in the identification of TP
inhibitors to be used as potential anti-angiogenic agents, and
several new classes of TP inhibitors have been reported
Compound
R1
R2
IC (␮M)
50
1
2
3
4
5
6
Br
Cl
-NH2
-NHCH
-NHCH CH NH HCl
2 2 2
-NHCH CH OH
-NHCH CH OH
-NHCH CH CH OH
17
0.25
1.7
CH NH HCl
2 2
2
CH3
Cl
CH3
Cl
90
2
2
[28–31], including one class with IC50 values 1000-fold
Ͼ1000
650
2
2
lower than those previously described [30,31]. In this study,
we describe the synthesis of several new potent TP inhibi-
tors, and further demonstrate that these compounds block
the in vitro pro-angiogenic actions of both purified human
recombinant TP, and human tumor cells expressing high
levels of TP.
2
2
2
7
H
30
8
9
H
H
H
114
122
160
2
. Materials and methods
1
1
1
0
1
2
2
.1. Synthesis of 5,6-disubstituted and 6-substituted
uracil derivatives
Compounds 2 through 6 (Table 1) were obtained by
treatment of appropriately 5-substituted 6-chlorouracil de-
H
220
rivatives with an excess of the diamine or hydroxylamine
o
(
15- to 20-fold molar excess) at 100 without solvent for
H
H
275
1
–4 hr. Derivatives 7–12 were obtained by treatment of the
corresponding 5-substituted 6-chlorouracil derivatives with
the appropriate amine in 1.5- to 4-fold molar excess, using
the general methods previously described [32]. Reactions
13
Ͼ300
o
were conducted with the reactants in the neat at 200 (for
1
1) or in water at reflux (for 7–10 and 12).
The 6-chlorouracils used as starting materials for 2-12
1
4
5
Cl
0.30
500
were obtained as follows: the 5,6-dichlorouracil precursor
of 2, 4, and 6 was obtained by C-5 chlorination of commer-
cially available 6-chlorouracil with sulfuryl chloride in ace-
tic acid by a modification of a reported procedure [33].
1
CH3
16
CH3
Cl
200
15
6
-Chloro-5-methyluracil, which served as the precursor of 3
and 5 [34], was obtained by treatment of 5-methylbarbituric
acid [35] with POCl in concentrated H PO [36]. Deriva-
1
1
7
8
3
3
4
tives 13, 14, and 16-19 were prepared by treatment of
appropriately 5-substituted 6-chloromethyluracils with the
corresponding amine (in 4–10 molar excess in refluxing
methanol for 16 hr) or with the corresponding thiono de-
rivative (in 1.2 molar excess in refluxing methanol for 2 hr).
Commercially available 6-chloromethyluracil (used directly
as a synthetic precursor for 13 and 18) was chlorinated as
described above to give 5-chloro-6-chloromethyluracil, the
synthetic precursor for 14, 17, and 19. The thiosemicarba-
zone derivative of thymine-6-carboxaldehyde, compound
H
Ͼ300
1
9
Cl
300
alcohol with sodium borohydride followed by chlorination
with thionyl chloride as we reported previously [38].
All the amines used for the synthesis of 2-12 were com-
mercially available except for 4-benzyloxy-2-methylaniline
1
5, was prepared as described [37]. Thymine-6-carboxalde-
hyde was also used to prepare 5-methyl-6-chloromethylu-
racil (the synthetic precursor of 16) by reduction to the

R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
1259
(
for compound 7), which was prepared, using standard
methodologies, by benzylation of commercially available
-methyl-4-nitrophenol (heating to reflux in ethanol for 3 hr
with 0.05% trypsin/0.02% EDTA, and the cell number was
determined using a Coulter counter.
3
with benzylbromide and potassium carbonate). The 4-ben-
zyloxy-2-methyl-nitrobenzene thus obtained was then re-
duced with iron powder in tetrahydrofuran/water at 0 under
2.4. Modified Boyden chamber migration assay
o
HUVEC (passages 1–4) were harvested using a cell
nitrogen to give the desired 4-benzyloxy-2-methylaniline.
All products were fully characterized by elemental anal-
ysis and H-NMR techniques.
dissociation solution (Sigma) and pelleted by centrifugation
(250 g for 5 min). Cells (10 ) were placed into transwell
5
1
inserts (Corning-Costar; 8 ␮m pore) precoated with fi-
bronectin (10 ␮g/mL). Inserts containing cells were placed
into a 24-well plate containing 0.7 mL M199 with 1% serum
2
.2. Assay of TP, UP, and PNP activities
o
and incubated (37 ) for 1 hr. A chemotactic stimulus, re-
The effects of the uracil analogs on purified recombinant
combinant human TP (100 ng/mL) or human VEGF (10
ng/mL; R&D Systems), was added to the lower wells in the
presence and absence of various concentrations of inhibi-
human TP activity were evaluated. In addition to the inhib-
itors, the assays contained 200 mM KH PO (pH 7.8), 0.2
2
4
3
o.
mM [5Ј- H]thymidine (50 mCi/mmol; Moravek Biochemi-
tors, and the cells were incubated for 5 hr at 37 At the end
cals), and 50 ng TP. Reactions were stopped after 20 min at
of the incubation period, the surface of the upper membrane
was swabbed with a cotton-tipped applicator to remove
non-migrating endothelial cells. Wells and inserts were
washed three times with PBS and stained for 30 min in
medium containing 10 ␮M Cell Tracker Green (Molecular
Probes). The cluster plate was washed three times with PBS,
and cells were fixed in 3.7% formaldehyde for 10 min and
mounted on microscope slides. For quantitation, five ran-
dom fields were photographed (100ϫ total magnification),
and the number of HUVEC in each field was counted.
Boyden chamber co-culture experiments utilized tumor
cells in place of the purified angiogenic factors; no other
angiogenic stimulus was added to the incubation. HT29
colon and MCF7 breast carcinoma cells that had been stably
transfected with a human TP cDNA (HT29/TPneo and
o
3
7 by the addition of 0.5 mL of ice-cold activated charcoal
in 5% trichloroacetic acid. After centrifugation, radioactiv-
ity in an aliquot of the supernatant was determined by liquid
scintillation counting.
UP activity was measured as previously described [39].
Reactions contained, in 0.1 mL, 50 mM Tris–HCl (pH 7.5),
1
mM potassium phosphate, 200 ␮M uridine, 1 ␮Ci
3
[5- H]uridine, 100 ng purified recombinant human UP (pro-
vided by Dr. G. Pizzorno), and increasing concentrations of
o,
inhibitor. After 30 min at 37 aliquots of the reactions were
analyzed on silica gel TLC plates using chloroform:meth-
anol:acetic acid (85:15:5, by vol.). Spots co-migrating with
authentic uracil and uridine were scraped, and the amount of
3
3
[
H]uracil formed from [ H]uridine was determined by liq-
5
uid scintillation counting. The effect of the TP inhibitors on
bovine PNP activity was assessed in a coupled spectropho-
tometric assay in which inosine was converted to hypoxan-
thine by PNP, and subsequently to uric acid by xanthine
oxidase [40]. Reactions contained 10 mM KH PO (pH
MCF7/TPneo) were used. Cells (2 ϫ 10 ) were added to the
lower wells and were allowed to attach for 24 hr. The
medium was then replaced with M199 medium containing
1% serum with and without inhibitor, combined with HU-
VEC-containing transwells, and migration over a 5-hr pe-
riod was analyzed as described above.
2
4
7
.7), 40 ␮M inosine, 19 ng PNP (Sigma), 0.01 U xanthine
oxidase (Sigma), and increasing concentrations of a uracil
analog. The change in absorbance of uric acid at 293 nm
was monitored continuously, and enzyme activity was cal-
3. Results and discussion
Ϫ1
Ϫ1
culated using an extinction coefficient of 12.9 mM cm
.3. Isolation of HUVEC
HUVEC were isolated from umbilical cords obtained
.
The ability of TP to inactivate pyrimidine 2Ј-de-
oxynucleosides of chemotherapeutic interest provided me-
dicinal chemists with an early incentive for developing TP
inhibitors. Consequently, there existed a rich background of
previous structure–activity relationships on which the cur-
rent studies were based [24]. One common theme that
emerged from a survey of the previous work on inhibition of
TP was that 5-substituted 6-aminouracil-based compounds
were effective inhibitors [22,23,32]. Thus, we explored this
class of TP inhibitors further. 5-Bromo-6-aminouracil was
synthesized and found to be a modestly active inhibitor of
human TP, with an IC50 of 17 ␮M (compound 1, Table 1).
This was in good agreement with a previous report of an IC50
of 30 ␮M for this compound when tested against horse liver
TP [22] and an IC50 of 15 ␮M for the inhibition of the human
2
less than 5 hr after delivery. After rinsing, veins were
incubated with collagenase (100 U/mL, Worthington) in
Medium 199 with Earle’s salts (GIBCO) for 10 min. Pri-
mary cultures were seeded into flasks precoated with 0.02%
(w/v) gelatin in PBS. Culture medium consisted of M199
containing 20% newborn calf serum, 5% pooled human
serum, 2 mM l-glutamine, 5 U/mL of penicillin, and 5
o
␮
g/mL of streptomycin. Cells were incubated at 37 in a
humidified incubator with 5% CO with a medium change
2
after 24 hr and every 2 days thereafter until confluent.
Primary cultures of HUVEC were harvested by incubation

1
260
R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
enzyme by the closely related 5-chloro-6-aminouracil de-
rivative [31]. The latter compound has also been shown to
have inhibitory activity when tested in TP-mediated angio-
genesis models [27]. Introduction of an ethylamino exten-
sion on the 6-amino position along with a 5-halogen sub-
stitution (compound 2) produced a greater than 65-fold
increase in potency for TP inhibition when compared with
5
-bromo-6-aminouracil (1). The inhibitory activity of 2 was
reduced 7-fold when a 5-methyl group was substituted for
the 5-chloro constituent (3), and 360- to 2600-fold when the
6
-aminoethylamino side chain was replaced with an amino-
ethanol substitution (4) or an amino-propanol substitution
6).
It is known from previous work that inhibition of TP is
(
very sensitive to both the precise substitution pattern of
different 6-aminouracils [22,23,32], and to the source of the
TP enzyme being examined, with some compounds found
to have up to 900-fold differences in inhibition of the
Escherichia coli enzyme compared with the rabbit liver
enzyme [23]. The basis for both of these differences likely
relates to variations in the hydrophobic regions that sur-
round the active sites of the enzyme and which interact with
the N-substituent groups. Because less is known of the
specificity of the human recombinant TP that we utilized,
we decided to reevaluate some of the previously identified
inhibitors that had been tested with TP from different
sources. Therefore, several of the 6-anilino and alkylamino
uracil derivatives identified earlier as the best inhibitors of
TP [22,23,38] were synthesized (compounds 7-12). They all
were found to have relatively modest inhibitory activity. A
more recent study [31] had reported that a number of 5-ha-
logeno-amino-substituted 6-methyluracil derivatives were
more potent inhibitors of TP than those described earlier. To
explore this possibility further, several of these were resyn-
thesized (14, 17, and 19) together with new derivatives
bearing different substitutions at position C-5 of the uracil
moiety (13, 15, 16, and 18) for study as potential anti-
angiogenic agents. Only one these, compound 14, was note-
worthy as a TP inhibitor, with an IC50 of 0.30 ␮M.
Fig. 1. Inhibition of TP, UP, and PNP activities by 5,6-substituted uracil
analogs. The effects of the uracil analogs CIMU (E‚ᮀ) and AEAC
(FŒ■) on TP (EF), UP (ᮀ■), and PNP (‚Œ) activities were evaluated.
In addition to the inhibitors at the indicated concentrations, TP assays
3
contained 200 mM KH PO (pH 7.8), 0.2 mM [5Ј- H]thymidine (50
2
4
o
mCi/mmol), and 50 ng recombinant human TP. Reactions (20 min at 37 )
were stopped with ice-cold activated charcoal in 5% trichloroacetic acid.
After centrifugation, radioactivity in an aliquot of the supernatant was
determined by liquid scintillation counting; TP activity in the absence of
inhibitor (100% value) was 222 nmol/mg/min. Data are means Ϯ SEM of
four experiments. UP reactions contained, in 0.1 mL, 50 mM Tris–HCl (pH
3
7
.5), 1 mM potassium phosphate, 200 ␮M uridine, 1 ␮Ci [5- H]uridine,
1
00 ng purified human UP, and the indicated concentrations of AEAC and
o
CIMU. After 30 min at 37 , aliquots of the reactions were analyzed by
silica gel TLC as described in “Materials and methods.” Radioactivity in
spots co-migrating with authentic uracil and uridine were determined by
liquid scintillation counting; activity in the absence of inhibitor was 3.92
␮mol/mg/min. Data are means Ϯ SEM of three experiments. PNP activity
was assessed in a coupled spectrophotometric assay in which inosine was
converted to uric acid by the combined actions of PNP and xanthine
oxidase. Reactions contained 10 mM KH PO (pH 7.7), 40 ␮M inosine, 19
2
4
ng PNP, 0.01 U xanthine oxidase, and increasing concentrations of AEAC
and CIMU. The change in absorbance of uric acid at 293 nm was moni-
tored continuously, and enzyme activity was calculated using an extinction
Ϫ1
Ϫ1
coefficient of 12.9 mM
cm . Data are from a single experiment;
activity in the absence of inhibitor was 102 ␮mol/mg/min.
Two conclusions can be drawn from these studies: (a) the
replacement of the C-5 methyl group of thymine by chlorine
consistently resulted in significantly better inhibitors of TP
for this particular class of derivatives (compare 2 vs 3, and
most potent of the compounds tested, AEAC (2) and CIMU
(14) was directly compared, as shown in Fig. 1. Although
the two compounds had nearly identical inhibitory effects
on TP activity, they could be distinguished by their ability
to inhibit UP activity. CIMU inhibited UP activity with an
IC of 100 ␮M, a concentration approximately 300-fold
higher than required to inhibit TP, whereas AEAC remained
completely selective for TP inhibition up to a concentration
of 1 mM. Neither AEAC nor CIMU had an effect on PNP
activity up to a concentration of 0.1 mM. Further analysis of
AEAC with various concentrations of the substrate thymi-
dine demonstrated that AEAC was a competitive inhibitor
4
vs 5); and (b) potency increased with the replacement of
the aminoethanol group of compounds 4 and 5 by the
considerably more basic aminoethylamine in compounds 2
and 3, an observation that generally paralleled the increase
in potency seen with increasing basic strength of the func-
tionalities linked to the C-6 methylene group of the 5-ha-
logeno-amino-substituted 6-methyluracil derivatives [31].
Indeed, the observed potency of compounds 2 and 3 was
quite striking and suggests that future investigations should
focus on uracil analogs bearing a halogen at C-5 and other
strongly basic polyamine (or amidine) functionalities di-
rectly attached to C-6 of the uracil moiety via a C—N bond.
The inhibition of TP, UP, and PNP activities by the two
5
0
of TP with a calculated K of 165 nM (Fig. 2). Both TP and
i
VEGF induced HUVEC migration in a modified Boyden
chamber assay (Fig. 3), although only the migration induced
by TP was dependent upon the presence of thymidine (data

R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
1261
Fig. 3. Stimulation of HUVEC migration by TP and VEGF, and inhibition
by uracil analogs. Human endothelial cells, isolated as described in “Ma-
terials and methods,” were used in a modified Boyden chamber assay. Cells
5
(
10 ) were placed into fibronectin-coated transwell inserts, which were
then placed into 24-well plates in which the lower wells contained 0.7 mL
M199 with 1% serum. After 1 hr, recombinant human TP (200 ng/mL)
(left) or human VEGF (10 ng/mL) (right) was added to the lower wells.
Control wells (dashed line) had no added angiogenic factor. Wells also
contained the indicated concentrations (0–300 ␮M) of either CIMU (F and
single hatched bar) or AEAC (Πand double hatched bar). Cells were
Fig. 2. Double-reciprocal plot of TP inhibition by AEAC. Assays of TP
activity were carried out as described in Fig. 1 except that the concentration
of the substrate thymidine was varied between 33 and 200 ␮M. AEAC was
used at concentrations of 0 mM (E), 0.3 mM (■), and 0.9 mM (F). Data
are means of two experiments.
o
incubated for 5 hr at 37 , at which time non-migrating endothelial cells
were removed from the upper surface of the membrane. Cells were stained
with Cell Tracker Green, washed with PBS, fixed in formaldehyde, and
mounted on microscope slides. Migration was quantitated by counting five
random photographed (100ϫ total magnification) fields, and the number of
HUVEC in each field was counted. Data are means Ϯ SEM of three
experiments.
not shown). This confirmed that the effect of TP on endo-
thelial cells was dependent upon its catalytic activity. TP-
mediated HUVEC migration was inhibited by both CIMU
and AEAC in a concentration-dependent manner, with 50%
inhibition at 30 and 60 ␮M, respectively, and with complete
inhibition at concentrations of 200 ␮M (Fig. 3). In contrast,
neither compound had an inhibitory effect on VEGF-medi-
ated HUVEC migration, even at concentrations of 300 ␮M
level of TP expression and tumor angiogenesis and aggres-
siveness. The development of TP inhibitors would provide
a means to evaluate the extent to which TP overexpression
contributes to the early development, progression, and
maintenance of experimental and human cancers, particu-
larly in relation to other angiogenic factors whose expres-
(Fig. 3).
An elevated level of TP expression, when compared with
surrounding uninvolved tissue, is a common finding in a
wide range of human solid tumors. We have transfected
human colon and breast cancer cells with the TP cDNA as
a means of evaluating the angiogenic properties of TP-
overexpressing tumor cells. These cells were strong induc-
ers of HUVEC migration in the Boyden chamber assay
Table 2
Inhibition of HUVEC migration by a TP inhibitor in a co-culture assay
Tumor cells
CIMU
HUVEC cells/field
% Inhibition
None
Ϫ
Ϫ
ϩ
Ϫ
ϩ
11 Ϯ 1.7
265 Ϯ 32
69 Ϯ 9.4
115 Ϯ 9.0
28 Ϯ 2.6
HT29/TPneo
HT29/TPneo
MCF7/TPneo
MCF7/TPneo
77
84
(Table 2), and were 2- to 4-fold more chemotactic for
HUVEC than HT29 or MCF7 cells that had been transfected
with a control vector (data not shown). Consistent with its
inhibitory activities against purified TP, CIMU also strongly
inhibited HUVEC migration induced by TP-expressing tu-
mors cells (Table 2), an observation that supports further
investigation into the potential use of these inhibitors in
pathological conditions that are dependent upon TP-medi-
ated angiogenesis.
HUVEC migration was assessed in a modified Boyden chamber assay in
which human cancer cells were used in place of added angiogenic stimuli;
HT29 colon and MCF7 breast carcinoma cells that had been stably trans-
fected with a human TP cDNA (HT29/TPneo and MCF7/TPneo) were
5
used. Tumor cells (10 ) were added to the lower wells of a 24-well plate
and were allowed to attach for 24 hr. The medium was then replaced with
M199 medium with and without 0.3 mM CIMU, combined with HVUEC-
containing transwells, and migration over a 5-hr period was analyzed as
described in “Materials and methods.” Data are means Ϯ SEM of two
experiments, each done in triplicate.
Immunohistochemical studies of certain human tumors
have shown a consistent positive correlation between the

1
262
R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
sion has also been shown to be correlated with increased
neovascularization. The possibility that a TP inhibitor could
have therapeutic utility as an anticancer drug is supported
by studies which found that elevated TP expression in-
creased the growth of tumor cells as xenografts in vivo
without altering their growth in vitro [6,7,30]. This en-
hancement of tumor growth in vivo was reversed when the
mice were treated with a TP inhibitor [27,30]. TP has also
been implicated as an angiogenic factor contributing to the
pathology of other diseases, including (a) rheumatoid arthri-
tis, where TP was found to be highly elevated in synovial
fluid, (b) psoriasis, in which increased TP expression was
seen in psoriatic lesions, and (c) gastric ulcers, where TP
was found to be elevated near gastric ulcer margins when
compared with uninvolved fundic and pyloric stomach [41–
istry 1992;31:12141–6.
3] Furukawa T, Yoshimura A, Sumizawa T, Haraguchi M, Akiyama S-I,
Fukui K, Ishizawa M, Yamada Y. Angiogenic factor. Nature 1992;
[
3
56:668.
[
4] Sumizawa T, Furukawa T, Haraguchi M, Yoshimura A, Takeysu A,
Ishizawa M, Yamada Y, Akiyama S-I. Thymidine phosphorylase
activity associated with platelet-derived endothelial cell growth fac-
tor. J Biochem (Tokyo) 1993;114:9–14.
5] Finnis C, Dodsworth N, Pollitt CE, Carr G, Sleep D. Thymidine
phosphorylase activity of platelet-derived endothelial cell growth
factor is responsible for endothelial cell mitogenicity. Eur J Biochem
1993;212:201–10.
[6] Ishikawa F, Miyazono K, Hellman U, Drexler H, Wernstedt C, Usuki
K, Takaku F, Risau W, Heldin C-H. Identification of angiogenic
activity and the cloning and expression of platelet-derived endothelial
cell growth factor. Nature 1989;338:557–62.
7] Moghaddam A, Zhang H-T, Fan T-PD, Hu D-E, Lees VC, Turley H,
Fox SB, Gatter KC, Harris AL, Bicknell R. Thymidine phosphorylase
is angiogenic and promotes tumor growth. Proc Natl Acad Sci USA
[
[
4
4]. Plasma TP levels were higher in intractable gastric
1
995;92:998–1002.
ulcer patients compared with normal individuals, duodenal
ulcer patients, and patients whose gastric ulcer had under-
gone significant resolution [44].
[
8] Fujimoto J, Sakaguchi H, Aoki I, Tamaya T. The value of platelet-
derived endothelial cell growth factor as a novel predictor of advance-
ment of uterine cervical cancers. Cancer Res 2000;60:3662–5.
An increasing appreciation of the critical role angiogen-
esis plays in tumors and in pathologies associated with
inflammatory responses has prompted the pursuit of a vari-
ety of different approaches to block neovascularization, and
a number of therapeutic agents are currently in clinical
trials. Approaches that target the actions of angiogenic fac-
tors have focused on the development of antibodies, small
molecules, or gene therapy to directly neutralize either the
angiogenic factor, or to block the receptors for the angio-
genic factors on endothelial cells [45–48]. TP is distinct
from other angiogenic factors in that, unlike classical factors
which bind to a cell surface receptor, TP exerts its actions
through its catalytic activity, the products of which are
presumed to be responsible for the angiogenic activities of
TP. Therefore, TP provides unique opportunities for the
design of angiogenesis inhibitors, and, in addition to their
therapeutic potential, these compounds should be useful in
evaluating the contribution of TP to the progression of a
number of pathological conditions.
[9] Yoshimura A, Kuwazuru Y, Furukawa T, Yoshida H, Yamada K,
Akiyama S. Purification and tissue distribution of human thymidine
phosphorylase; high expression in lymphocytes, reticulocytes and
tumors. Biochim Biophys Acta 1990;1034:107–13.
10] Touffet S, Gayet G, Samperez S, Jouan P. Demonstration of thymi-
dine phosphorylase activity in human healthy, adenomatous and can-
cerous prostate. Bull Cancer 1992;79:151–9.
[11] Peters GJ, van Groeningen CJ, Laurensse EJ, Pinedo HM. A com-
parison of 5-fluorouracil metabolism in human colorectal cancer and
colon mucosa. Cancer 1991;68:1903–9.
12] Reynolds K, Farzaneh F, Collins WP, Campbell S, Bourne TH,
Lawton F, Moghaddam A, Harris AL, Bicknell R. Association of
ovarian malignancy with expression of platelet-derived endothelial
cell growth factor. J Natl Cancer Inst 1994;86:1234–8.
13] O’Brien TS, Fox SB, Dickinson AJ, Turley H, Westwood M,
Moghaddam A, Gatter KC, Bicknell R, Harris AL. Expression of the
angiogenic factor thymidine phosphorylase/platelet-derived endothe-
lial cell growth factor in primary bladder cancers. Cancer Res 1996;
56:4799–804.
[14] Relf M, LeJeune S, Scott PAE, Fox S, Smith K, Leek R, Moghaddam
A, Whitehouse R, Bicknell R, Harris AL. Expression of the angio-
genic factors vascular endothelial cell growth factor, acidic and basic
fibroblast growth factor, tumor growth factor ␤-1, platelet-derived
endothelial cell growth factor, placenta growth factor, and pleiotro-
phin in human primary breast cancer and its relation to angiogenesis.
Cancer Res 1997;57:963–9.
[
[
[
Acknowledgments
[
15] Ikeda N, Adachi M, Taki T, Huang C, Hashida H, Takabayashi A,
Sho M, Nakajima Y, Kanehiro H, Hisanaga M, Nakano H, Miyake M.
Prognostic significance of angiogenesis in human pancreatic cancer.
Br J Cancer 1999;79:1553–63.
We thank Drs. Vern Schramm and Giuseppe Pizzorno
for providing purine nucleoside phosphorylase and uridine
phosphorylase, respectively, and Greg Kicska for assistance
with the PNP assay. This work was supported by Grant
DAMD 17–96-1–6123 from the Department of Defense
Breast Cancer Research Program, Grants R01-CA54422
and P01-CA13330 from the National Cancer Institute, and
by a grant from the UJA-Federation of New York.
[
16] Takebayashi Y, Yamada K, Miyadera K, Sumizawa T, Furukawa T,
Kinoshita F, Aoki D, Okumura H, Yamada Y, Akiyama S-I, Aikou T.
The activity and expression of thymidine phosphorylase in human
solid tumors. Eur J Cancer 1996;32A:1227–32.
17] Luccioni C, Beaumatin J, Bardot V, Lefrancois D. Pyrimidine nucle-
otide metabolism in human colon carcinomas: comparison of normal
tissues, primary tumors and xenografts. Int J Cancer 1994;58:517–22.
18] Takabayashi Y, Akiyama S, Akiba S, Yamada K, Miyadera K, Sum-
izawa T, Yamada Y, Murata F, Aikou T. Clinicopathologic and
prognostic significance of an angiogenic factor, thymidine phosphor-
[
[
References
ylase, in human colorectal cancer.
J
Natl Cancer Inst
[
[
1] Folkman J, Shing Y. Angiogenesis. J Biol Chem 1992;267:10931–4.
2] Moghaddam A, Bicknell R. Expression of PD-ECGF factor in E. coli
and confirmation of its thymidine phosphorylase activity. Biochem-
1996;88:1110–7.
[19] Takahasi Y, Bucana CD, Liu W, Yoneda J, Kitadai Y, Cleary KR,
Ellis LM. Platelet-derived endothelial cell growth factor in human

R.S. Klein et al. / Biochemical Pharmacology 62 (2001) 1257–1263
1263
colon cancer angiogenesis: role of infiltrating cells. J Natl Cancer Inst
996;88:1146–51.
[32] Baker BR, Rzeszotarski W. Irreversible enzyme inhibitors. Thymi-
dine phosphorylase: on the nature and dimensions of the hydrophobic
bonding region. J Med Chem 1968;11:639–44.
[33] Gershon H, Dittmer K, Braun R. Pyrimidines. I. Some halogenated
mono-methylpyrimidines. J Org Chem 1961;26:1874–7.
[34] Nobuyuki F, Daiki I, Daiji I, Izumi M, Kimiko T. 6-Amino-5-
methyluracil derivative, Japanese Patent 1998; 10017555A.
[35] Dickey JB, Gray AR. Barbituric acid. In: Blatt AH, editor. Or-
ganic synthesis. Collective vol. II. London: John Wiley, 1943. p.
60.
[36] Gauri KK. Substituted pyrimidines as inhibitors of yeast alcohol
dehydrogenase. Arch Pharm (Weinheim) 1973;306:622–5.
[37] Cheng K-Y, Cheng CC. Pyrimidines. XX. A convenient preparation
of orotaldehyde, and thymine-6-carboxaldehyde. J Heterocyclic
Chem 1967;4:163–5.
[38] Klein RS, Fox JJ. Nucleosides. LXXVII. Synthesis of some 6-sub-
stituted uracils, and uridines by the Wittig reaction. J Org Chem
1972;37:4381–6.
1
[
20] Metzger R, Danenberg KD, Leichman CG, Salonga D, Schwartz EL,
Wadler S, Lenz HJ, Groshen S, Leichman L, Danenberg PV. High
basal level gene expression of thymidine phosphorylase (platelet-
derived endothelial cell growth factor) in colorectal tumors is asso-
ciated with non-response to 5-fluorouracil. Clin Cancer Res 1998;4:
2
371–6.
[
21] van Triest B, Pinedo HM, Blaauwgeers JLG, van Diest PJ, Schoen-
makers PS, Voorn DA, Smid K, Koekman K, Hoitsma HFW, Peters
GF. Prognostic role of thymidylate synthase, thymidine phosphory-
lase/platelet-derived endothelial cell growth factor, and proliferation
markers in colorectal tumors. Clin Cancer Res 2000;6:1063–72.
22] Langen P, Etzold G, Barwolff D, Preussel B. Inhibition of thymidine
phosphorylase by 6-aminothymine and derivatives of 6-aminouracil.
Biochem Pharmacol 1967;16:1833–7.
[
[
[
[
23] Baker BR, Kelley JL. Irreversible enzyme inhibitors. 188. Inhibition
of mammalian thymidine phosphorylase.
971;14:812–6.
J
Med Chem
1
[39] Liu M, Cao D, Russell R, Handschumacher RE, Pizzorno G. Expres-
sion, characterization, and detection of human uridine phosphorylase,
and identification of variant uridine phosphorolytic activity in se-
lected human tumors. Cancer Res 1998;58:5418–24.
24] Niedzwicki JG, el Kounii MH, Chu SH, Cha S. Structure-activity
relationship of ligands of the pyrimidine nucleoside phosphorylases.
Biochem Pharmacol 1983;32:399–415.
25] Desgranges C, Razaka G, Rabaud M, Bricaud H, Balzarini J, De
Clercq E. Phosphorolysis of (E)-5-(2-bromovinyl)-2Ј-deoxyuridine
[40] Miles RW, Tyler PC, Furneaux RH, Bagdassarian CK, Schramm VL.
One-third-the-sites transition-state inhibitors for purine nucleoside
phosphorylase. Biochemistry 1998;37:8615–21.
(BVDU) and other 5-substituted-2Ј-deoxyuridines by purified human
thymidine phosphorylase and intact blood platelets. Biochem Phar-
macol 1983;32:3583–90.
26] Grancharov K, Mladenova J, Golovinsky E. Inhibition of uridine
phosphorylase by some pyrimidine derivatives. Biochem Pharmacol
[41] Takeuchi M, Otsuka T, Matsui N, Asai K, Horano T, Moriyama A,
Isobe I, Eksioglu YZ, Matukawa K, Kato T, Tada T. Aberrant pro-
duction of gliostatin/platelet-derived endothelial cell growth factor in
rheumatoid synovium. Arthritis Rheum 1994;37:662–72.
[42] Muro H, Waguri-Nagaya Y, Mukofujiwara Y, Iwahashi T, Otsuka T,
Matsui N, Moriyama A, Asai K, Kato T. Autocrine induction of
gliostatin/platelet-derived endothelial cell growth factor and GLS-
induced expression of metalloproteinases in rheumatoid arthritis sy-
noviocytes. Rheumatology 1999;38:1195–202.
[
1
991;41:1769–72.
[
27] Miyadera K, Sumizawa T, Haraguchi M, Yoshida H, Konstanty W,
Yamada Y, Akiyama S. Role of thymidine phosphorylase activity in
the angiogenic effect of platelet-derived endothelial cell growth fac-
tor/thymidine phosphorylase. Cancer Res 1995;55:1687–90.
[
[
[
28] Balzarini J, Esteban-Gamboa E, Esnouf R, Liekens S, Neyts J, De
Clercq E, Camarasa MJ, Perez-Perez MJ. 7-Deazaxanthine, a novel
prototype inhibitor of thymidine phosphorylase. FEBS Lett 1998;438:
[43] Creamer D, Jaggar R, Allen M, Bicknell R, Barker J. Overexpression
of the angiogenic factor platelet-derived endothelial cell growth fac-
tor/thymidine phosphorylase in psoriatic epidermis. Br J Dermatol
1997;137:852–5.
[44] Kusugai K, Joh T, Kataoka H, Sasaki M, Tada T, Asai K, Kato T, Itoh
M. Evidence for participation of gliostatin/platelet-derived endothe-
lial cell growth factor in gastric ulcer healing. Life Sci 1997;61:
1899–906.
[45] Cherrington JM, Strawn LM, Shawver LK. New paradigms for the
treatment of cancer: the role of anti-angiogenesis agents. Adv Cancer
Res 2000;79:1–38.
[46] Auerbach W, Auerbach R. Angiogenesis inhibition: a review. Phar-
macol Ther 1994;63:265–311.
9
1–5.
29] Esteban-Gamboa E, Balzarini J, Esnouf R, De Clercq E, Camarasa
MJ, Perez-Perez MJ. Design, synthesis, and enzymatic evaluation of
multisubstrate analogue inhibitors of Escherichia coli thymidine
phosphorylase. J Med Chem 2000;43:971–83.
30] Matsushita S, Nitanda T, Furukawa T, Sumizawa T, Tani A, Nishi-
moto K, Akiba S, Miyadera K, Fukushima M, Yamada Y, Yoshida H,
Kanzaki T, Akiyama S. The effect of a thymidine phosphorylase
inhibitor on angiogenesis and apoptosis in tumors. Cancer Res 1999;
5
9:1911–6.
[
31] Fukushima M, Suzuki N, Emura T, Yano S, Kazuno H, Tada Y,
Yamada Y, Asao T. Structure and activity of specific inhibitors of
thymidine phosphorylase to potentiate the function of antitumor 2Ј-
deoxyribonucleosides. Biochem Pharmacol 2000;59:1227–36.
[47] Ferrara N, Alitalo K. Clinical applications of angiogenic growth
factors and their inhibitors. Nat Med 1999;5:1359–64.
[48] Folkman J. Antiangiogenic gene therapy. Proc Natl Acad Sci USA
1998;95:9064–6.