Synthesis of modified pyrimidine bases and positive impact of chemically reactive substituents on their in vitro antiproliferative activity

Article abstract of DOI:10.1016/j.ejmech.2008.06.002

The antiproliferative activity screening on human tumor cell lines of a series of modified uracil and cytosine bases as well as some corresponding acyclonucleosides, and comparison of structure-activity relationship revealed the importance of chemical rea

Full text of DOI:10.1016/j.ejmech.2008.06.002

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 44 (2009) 1172e1179
http://www.elsevier.com/locate/ejmech
Original article
Synthesis of modified pyrimidine bases and positive impact
of chemically reactive substituents on their in vitro
antiproliferative activity
a
Steffi Noll , Marijeta Kralj , Lidija Suman , Holger Stephan , Ivo Piantanida
b
b
a,
c,**
ˇ
*
a Institut fu¨r Radiopharmazie, Forschungszentrum Dresden-Rossendorf, Postfach 510119, 01314 Dresden, Germany
b
´
ꢀ
Division of Molecular Medicine, RuCer Bosˇkovic Institute, Bijenicka 54, Zagreb, Croatia
c
´
Division of Organic Chemistry and Biochemistry, RuCer Bosˇkovic Institute, Bijenicka 54, Zagreb, Croatia
ꢀ
Received 4 February 2008; received in revised form 26 May 2008; accepted 2 June 2008
Available online 11 June 2008
Abstract
The antiproliferative activity screening on human tumor cell lines of a series of modified uracil and cytosine bases as well as some
corresponding acyclonucleosides, and comparison of structureeactivity relationship revealed the importance of chemical reactivity of the sub-
stituent attached to the C5-position of uracil for the activity of studied compounds. Namely, the results obtained for the most active compounds,
5-(chloroacetylamino)uracil (2) and its acyclic sugar analogue 18, suggest that formation of a covalent bond between reactive substituent and
several possible targets within the thymidylate synthase mechanism (sulphur of the cysteine residue, basic part of the enzyme, N,N-methylene
tetrahydrofolate or its reactive iminium forms) is the most probable mode of action. In addition, novel C5-substituted uracil derivative 6 (5-[bis-
(2-p-methoxybenzylthioethyl)amine]acetylaminouracil) exhibited high antiproliferative activity against HeLa and MiaPaCa-2 cell lines, by an as
yet unknown mechanism.
Ó 2008 Elsevier Masson SAS. All rights reserved.
1. Introduction
antimetabolites based on the similarity of structure to the
naturally occurring pyrimidines and purines involved in the
biosynthesis of DNA. Novel compounds should interfere
with one or more biological applications of naturally occurring
analogues.
In general, nucleoside analogues are structurally, metaboli-
cally, and pharmacodynamically related agents that neverthe-
less have diverse biological actions and therapeutic effects.
This class of agents affects the structural integrity of DNA,
usually after incorporation during replication or DNA excision
repair synthesis, leading to stalled replication forks and chain
termination. One of the remarkable features that is still unex-
plained about nucleic acid antagonists, is how drugs with such
similar structural features, which share metabolic pathways
and elements of their mechanisms of action, show such diver-
sity in their clinical activities [1]. One of the most frequently
used approaches to new antitumor drugs is a design of
The cytostatic agent 5-fluorouracil (5-FU), one of the most
important agents for the treatment of colorectal, head and
neck, pancreatic and breast carcinomas, is an antimetabolite
exerting a cytotoxic effect primarily through inhibition of
thymidylate synthase (TS) [2]. In the cascade of enzymatic
conversions 5-fluorouracil related metabolites inhibit the
action of thymidylate synthase, the only de novo source of thy-
midylate, one of the essential constituents of DNA and conse-
quently lead to the thymineless death of cell [3]. Each of these
conversion steps is a target of intensive research with the aim
to improve bioactivity. Thus, a huge number of 5-FU ana-
logues were designed and studied with the aim of more
efficient blocking of the (b) step (ternary complex of enzyme,
substrate (ex-uracil) and coenzyme). Alternatively, antifolate
* Corresponding author. Tel.: þ49 351 2603091; fax: þ49 351 2603232.
** Corresponding author. Tel.: þ385 1 45 71 210; fax: þ385 1 46 80 195.
E-mail addresses: h.stephan@fzd.de (H. Stephan), pianta@irb.hr (I.
Piantanida).
0223-5234/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.06.002

S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
1173
O
inhibitors are achieving the same goal by interacting with the
coenzyme [4].
O
O
CH2
Cl
H
N
Cl
NH2
CO CH2 Cl
HN
HN
The higher activity of 5-FU derivatives toward tumor cells
in comparison to normal cells is based on higher division rates
of tumors as well as on the fact that enzymes in normal cells
degrade both uracil and 5-FU but in tumor cells that degrada-
tion process does not work [5]. However, other mechanisms
can also contribute to the cytostatic activity, as it is obvious
from the metabolism complexity of 5-FU and its prodrugs
[6]. For example, 5-FU is incorporated in DNA and RNA
and the maturation process and metabolism of RNA is influ-
enced [7].
In this work, we present the synthesis of a series of novel
uracil derivatives, differing in properties of C5-positioned
side chains, like chemical reactivity and bulkiness. Most of
these compounds were also prepared in a form of N1-
substituted acyclic uridine analogues, to explore the impact
of acyclic sugar analogues on biological activity. Furthermore,
several cytosine analogues were prepared and studied to inves-
tigate if the uracil moiety is essential for the biological activ-
ity, and in line with recently observed differences in biological
action between cytosine and uracil derivatives [7]. Bioactivity
of all novel compounds was probed in vitro by screening
cellular growth inhibition on several different human tumor
cell lines. Most potent compounds were additionally studied
by assessing their cell cycle perturbation impact.
O
N
H
O
N
H
(a)
2
1
O
H
N
CO CH2 NH2
NH3
HN
2
O
N
H
(b)
3
H
N
O
NH boc
(c)
H
N
NH2
CH2
boc
CO CH2
N
H
2
HN
O
N
H
4
HC
COOH
CH
NH
boc
O
H
N
CO CH2
S
CH COOH
NH2
HN
2
O
N
H
(d)
5
S
N
CH2
CH2
OMe
OMe
S CH2
OMe
OMe
O
H
N
S
CO CH
N
2
HN
2
(e)
O
N
H
S CH2
6
Scheme 1. Synthesis of C5-substituted uracil derivatives 2e6. Reagents and
conditions: (a) ClCH2COCl, NaOH/H2O, 30 min, 0 ꢀC; 90 min, 22 ꢀC; recrys-
tallization from H2O (2: 62%); (b) 2, NH3(aq), 24 h, 22 ꢀC; Sephadex G10 (3:
81%); (c) 2, H2NCH2CH2NH-boc, H2O/MeOH, 1 h, 90 ꢀC, pressure; Sepha-
dex G10 (4: 76%); (d) 2, boc-cysteine, NaOH/EtOH, 2 h, 22 ꢀC; MPLC
(RP18, CH3CN/H2O ¼ 1/9) (5: 76%); (e) 2, NH4(SO4)2, hexamethyldisila-
zane, reflux, 3 h; bis(2-p-methoxybenzylthioethyl)-amine, (Et)3N, CH3CN,
90 ꢀC, 4 h; column chromatography (SiO2, CH2Cl2/MeOH ¼ 10/1) (6: 77%).
2. Synthesis
At first, uracil has been modified at the C5-position. Start-
ing from 5-aminouracil 1, the corresponding 5-(chloroacetyl)
aminouracil 2 e already known to show antiviral activity [8]
e was synthesized by treatment of chloroacetyl chloride in
1 M sodium hydroxide solution at 0 ꢀC. After acidification
with hydrochloric acid the product was obtained in high purity
by repetitive crystallization from water. Compound 2 was
converted into 5-(aminoacetyl)-aminouracil 3 with concen-
trated ammonia. The product was purified by column chroma-
tography on Sephadex G10 giving 3 in 81% yield. As shown in
Scheme 1, other uracil derivatives have been also accessible
by conversion of 2. Thus, heating 2 with N-boc-ethylenedi-
amine in methanol under pressure at 90 ꢀC led to 5-(N-boc-
ethylenediaminoacetyl)aminouracil 4. The reaction product
was separated by column chromatography on Sephadex G10.
Coupling of 2 with boc-protected cysteine was performed in
ethanol and 2 M sodium hydroxide. The protection group
was then removed by the treatment with 1 M hydrochloric
acid. Purification was carried out by MPLC yielding S-{2-
[(2,4-dioxo-1,2,3,4,-tetrahydropyrimidine-5-yl)amino]-2-oxoe
thyl}-D-cysteine 5. The 5-[bis-(2-p-methoxybenzylthioethyl)
amine]acetylaminouracil 6 was synthesized by direct conden-
sation of the persilylated 5-(chloroacetyl)aminouracil 2 with
bis-(2-p-methoxybenzylthioethyl)amine [9] in acetonitrile
followed by column chromatography on silica gel.
also investigated in vitro by screening cellular growth inhibi-
tion on several different human tumor cell lines. Briefly,
modification of the N1-position of the uracil molecule was
accomplished by direct condensation of the appropriate persi-
lylated base with the chloromethyl ether 1,3-dibenzyloxy-
2-chloro-methoxypropane and tetrabutylammonium iodide as
a catalyst. Thus the nucleic bases uracil 7, 5-hydroxyuracil
8, 5-aminouracil 1, 5-iodouracil 9, 6-methyluracil 10 and 5-
iodomethyluracil 11 were coupled to 1,3-dibenzyloxy-2-
chloro-methoxypropane in dry dimethyl formamide with
BzO
O
O
N
O
N
O
R1
R2
R1
R2
R1
R2
BzO
Cl
PdO
HN
HN
HN
cyclohexene
O
N
H
O
O
O
O
OBz
OH
OBz
OH
7. R1 = H,
8. R1 = OH, R2= H
1. R1 = NH2, R2 = H
9. R1 = I,
10. R1 = H, R2 = CH3
11. R1 = I, R2 = CH3
R2 = H
R1 = H,
R1 = OH, R2= H
R1 = NH2, R2 = H
R1 = I,
R1 = H, R2 = CH3
R1 = I, R2 = CH3
R2 = H
12. R1 = H,
13. R1 = OH, R2= H
14. R1 = NH2, R2 = H
15. R1 = I,
16. R1 = H, R2 = CH3
16. R1 = I, R2 = CH3
R2 = H
R2 = H
R2 = H
R2 = H
Scheme 2 shows the structure of some uridine derivatives
recently described as potential substrates of herpes simplex
virus type-1 thymidine kinase [10]. These compounds were
Scheme 2. Constitutions of uridine derivatives 12e17 [10].

1174
S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
R2
triethylamine yielding the O-benzylated precursors. Removal of
the benzyl protection groups with palladium oxide in cyclohex-
ene1-[(1,3-dihydroxy-2-propoxy)methyl]uracil (Acyclur) 12,
5-hydroxy-1-[(1,3-dihydroxy-2-propoxy)methyl]-uracil (Hydra-
cyclur) 13, 5-amino-1-[(1,3-dihydroxy-2-propoxy)methyl]uracil
(Amino-acyclur) 14, 5-iodo-1-[(1,3-dihydroxy-2-propoxy)-
methyl]-uracil (Iodacyclur) 15, 6-methyl-1-[(1,3-dihydroxy-2-
propoxy)methyl]uracil (Metacyclur) 16 and 5-iodo-6-methyl-1-
[(1,3-dihydroxy-2-propoxy)methyl]uracil (Iodmetacyclur) 17.
Of special interest is the chloroacetylated uridine derivative
18 (Scheme 3) which allows a direct comparison of biological
activity with the corresponding uracil compound 2. The prep-
aration of 18 was similar to 5-(chloroacetyl)aminouracil 2.
Amino-acyclur 14 was treated with chloroacetyl chloride in
aqueous sodium hydroxide at 0 ꢀC for 90 min. The purification
of the reaction product was carried out using reverse phase
MPLC.
Furthermore, some cytosine derivatives modified at the N1-
and N4-position have been investigated. Methylation of cyto-
sine 19 and 5-hydroxymethylcytosine 20 [11] by the treatment
with methyliodide and tetrabutylammonium hydroxide solu-
tion in DMF led smoothly to 1-methylcytosine 21 [12] and
1-methyl-5-hydroxymethylcytosine 22. The methylation of
4-thiouracil 23 under the same reaction conditions yielded 4-
methylthiouracil 24 [13]. By refluxing 24 with hydroxylamine
hydrochloride in dry ethanol followed by treatment with
concentrated ammonia, N4-hydroxycytosine 25 [14] was
obtained (Scheme 4).
R2
CH3I
R3
R3
N
N
(a)
O
N
H
O
N
R1
21. R1 = CH3, R2= NH2, R3 = H
22. R1 = CH3, R2= NH2, R3 = CH2OH
24. R1 = H, R2 = SCH3, R3 = H
19. R2 = NH2, R3 = H
20. R2 = NH2, R3 = CH2OH
23. R2 = SH, R3 = H
NHOH
N
NH2OH
24
(b)
O
N
H
25
Scheme 4. Synthesis of cyctosine derivatives 21, 22, 25 and 4-methylthiouracil
24. Reagents and conditions: (a) CH3I, [(C4H9)4N]OH, DMF, 2 h, 22 ꢀC;
recrystallization from EtOH (21: 86%); column chromatography (SiO2,
CH2Cl2/MeOH ¼ 1/1), (22: 82%); MPLC (RP18, CH3OH) (24: 84%); (b)
NH2OH, EtOH, 5 h, reflux; recrystallization from EtOH (25: 78%).
on HeLa, MiaPaCa-2 and SW620 cells but not on MCF-7
and H460 cells. On the contrary, compounds 2, 6 and 18
showed rather strong antiproliferative and cytotoxic activity
on the tested cell lines, comparable to 5-fluorouracil (5-FU).
The most prominent activity was seen on HeLa and Mia-
PaCa-2 cells by 2 and 6 (Fig. 1), which prompted us to
perform additional studies of the impact of 2 and 6 on the
cell cycle of HeLa cells. Besides, we additionally tested 2, 6
(Fig. 1), 18 and 5-FU (data not shown) for their inhibitory
activity on human skin keratinocytes HaCaT, as a model of
nontumorigenic cell line. Interestingly, compound 6 strongly
inhibited the growth (IC50 ¼ 9 mM), while 2 and 18 did not
(IC50 were >100 and ꢁ100, respectively). On the other hand
5-FU markedly inhibited the growth of HaCaT cells
(IC50 ¼ 0.4 mM).
A tosyl group was introduced into both cytosine 19 and
1-methylcytosine 21. Therefore, compounds 19 and 21 were
treated with p-toluenesulfonyl chloride in dry pyridine yield-
ing N1-( p-toluenesulfonyl)cytosine 26 [15,16] and 1-methyl-
N4-( p-toluenesulfonyl)cytosine 27. Pure products were
obtained by column chromatography (Scheme 5).
3. Biological results and discussion
In contrast to uracil derivatives, in general cytosine deriva-
tives did not show any growth inhibitory activity (22, 25, and
27). The only exceptions were 1-methylcytosine 21 causing
a weak antiproliferative effect exclusively on the HeLa cell
line and N1-tosylated cytosine 26, weakly inhibiting the
growth of MCF-7 cell line. Further experiments are needed
The tested compounds showed very different antiprolifera-
tive effects on the investigated panel of cell lines (Table 1 and
Fig. 1). Uracil derivatives revealed a much broader range of
cell growth inhibitory activity compared to cytosines. Com-
pounds 3, 12e17, 24 did not show any or very weak antipro-
liferative effect. Compounds 4 and 5 showed moderate growth
inhibitory activity mostly at the highest tested concentration
NH2
NHTs
H3C
SO2Cl
O
N
O
N
O
N
N
CH2
Cl
C
NH2
OH
NH
CO CH2 Cl
HN
Cl
HN
O
N
O
N
O
O
(a)
(a)
O
O
R1
R1
OH
OH
14
OH
18
19. R1 = H
26. R1 = H
21. R1 = CH3
27. R1 = CH3
Scheme 3. Synthesis of C5-chloroacetylated uridine derivative 18. Reagents
and conditions: (a) ClCH2COCl, NaOH/H2O, 30 min, 0 ꢀC; 90 min, 22 ꢀC;
MPLC (RP18, CH3CN/H2O ¼ 1/4) (18: 75%).
Scheme 5. Synthesis of tosylated cyctosine derivatives 26 and 27. Reagents
and conditions: (a) TsCl, pyridine, 6 h, 22 ꢀC; column chromatography
(SiO2, CH2Cl2/MeOH ¼ 50/1) (26: 90%, 27: 88%).

S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
1175
Table 1
Tumor cell growth inhibition presented as IC50 values (in mM)
S phase-active chemotherapeutic agent, with no activity
when cells are in G0 or G1. Treatment of cells with 5-FU
induces G1/S phase arrest. Therefore, we assessed the influ-
ence of the most active compounds 2 and 6 and 5-FU on the
cell cycle, as well as the possible induction of apoptosis on
the HeLa cell line, since the most prominent activity/selectiv-
ity was shown on this cell line. Interestingly, although the IC50
concentrations of these substances were quite similar, the cell
cycle influences showed a different pattern. Compound 2
induced strong G1/S phase arrest, accompanied with a reduc-
tion of cells in G2/M, but with no obvious apoptosis activation
(subG1 cells) (Fig. 2 and Table 2). Very similar, but somewhat
more pronounced influence on the cell cycle can be seen after
treatment with 5-FU, without significant difference between
10 mM (Table 2) and 50 mM concentrations (data not shown).
On the other hand, 6 induced a significant accumulation in the
G1-phase of the cells at the expense of S phase population
with an impressive increase in the number of subG1 cells
(representing the apoptotic cells) after first 24 h, while during
the next 48 h, about 60% of cells died by apoptosis.
a
IC50 (mM)
Compound Cell lines
HeLa
4 ꢂ 2
>100
MiaPaCa-2 SW620
17 ꢂ 3 12 ꢂ 9
>100 >100
MCF-7
15 ꢂ 2
>100
H460
2
14 ꢂ 1
3
>100
>100
>100
4
50 ꢂ 46
50 ꢂ 47
80 ꢂ 18
3 ꢂ 2
50 ꢂ 44 >100
70 ꢂ 25 >100
70 ꢂ 25 ꢁ100
5
86 ꢂ 8
4 ꢂ 3
6
76 ꢂ 20
12
13
14
15
16
17
18
21
22
24
25
26
27
5-FUb
a
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
30 ꢂ 13 >100
>100
>100
73 ꢂ 26 >100
12 ꢂ 1
12 ꢂ 5
18 ꢂ 3
19 ꢂ 3
32 ꢂ 18
24 ꢂ 3
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
27 ꢂ 5
>100
15 ꢂ 2
4 ꢂ 1
10 ꢂ 3
9 ꢂ 2
2 ꢂ 0.7
IC50 e the concentration that causes 50% growth inhibition.
5. Discussion of the results
b
5-FU e 5-fluorouracil.
The tested compounds are pyrimidine analogues substituted
at different positions namely at C1, C4 and C5. Their antipro-
liferative activity depends considerably on the type of the sub-
stituents as well as the position on the pyrimidine derivatives.
However, among uracil derivatives (2e6, 12e17, and 24),
those which exhibit by far the highest activity are character-
ised by an amide linkage of the substituent attached to the
C5-position of the uracil moiety (R3, Scheme 4). It seems
that the antiproliferative activity is controlled by the chemical
reactivity of the substituent attached to the end of the amide
spacer; e.g., good leaving groups as chloride (2 and 18)
show high activity, while those which showed weak or no
activity have low- or non-reactive groups at the end of the
amide spacer (3, 4, and 5). The only exception is the high
activity of 6 (Scheme 1), although the p-methoxybenzylthio-
substituent on the end of the amide spacer is not expected to
be a good leaving group.
to explain the unique selectivity of 21 and 26 in cell growth
inhibition.
4. Cell cycle perturbations
Cell cycle and apoptosis are very important functional
parameters to assess the cellular metabolism, physiology and
pathology. The DNA content of the cell can provide a great
deal of information about the cell cycle, and consequently
the effect on the cell cycle of added stimuli, e.g., drug treat-
ment. For example, the major mechanism of action of nucleo-
side analogues is through incorporation into DNA, and thus
this class of agents shows specificity for cells in the S-phase.
A very similar influence of the cell cycle was already
described for 5-FU [17]. 5-FU is considered to be purely an
150
100
50
150
100
50
0
0
-50
-100
-50
-100
-9
-8
-7
-6
-5
-4
-3
-9
-8
log10 concentration of 2 (M)
-7
-6
-5
-4
-3
log10 concentration of 6 (M)
HeLa
SW 620
H 460
MCF-7
MiaPaCa-2
HaCaT
HeLa
SW 620
H 460
MCF-7
MiaPaCa-2
HaCaT
Fig. 1. Doseeresponse profiles for the most active compounds (2 and 6) tested on various human tumor cell lines in vitro. The cells were treated with the
compounds at different concentrations, and percentage of growth (PG) was calculated. Each point represents a mean value of four parallel samples in three
individual experiments.

1176
S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
Fig. 2. DNA histograms obtained by flow cytometry (see Section 7), and the percentages of cells in subG1 (apoptotic cells), G0/G1, S and G2/M cell cycle phase,
after the treatment of HeLa cells with 2 and 6 (at 2 ꢃ 10ꢄ5 M for 24 h, 48 h and 72 h).
Antiproliferative activity of 2 and its acyclic uridine
analogue 18 is comparable, which is in line with comparable
activity of 5-FU and its 20-deoxyribo derivatives [18]. Al-
though some studies have shown that changes in the sugar
part of uridine-like molecules can have strong biological
Table 2
Flow cytometric analysis of HeLa cells treated with 2 and 6 (c ¼ 20 mM) and
5-fluorouracil (5-FU, c ¼ 10 mM)
impact, this is not noticeable for 2 and 18, suggesting that
reactivity of the substituent at C5-position of uracil moiety
is mainly responsible for the observed activity.
Treatment
Cell cycle
phasea
HeLa
24 h
48 h
72 h
Control
SubG1
G0/G1
S
7 ꢂ 0.6
48 ꢂ 0.9
33 ꢂ 0.3
18 ꢂ 1
12 ꢂ 5
50 ꢂ 2
43 ꢂ 0.3b
6 ꢂ 1b
12 ꢂ 1
54 ꢂ 2
34 ꢂ 1
10 ꢂ 0.5
14 ꢂ 0.4
48 ꢂ 0.2b
45 ꢂ 1b
7 ꢂ 1b
24 ꢂ 4
If we consider thymidylate synthase (TS) as a possible
target for uracil derivatives, there are several possibilities by
which 2 and 18 can inhibit the enzyme. Upon positioning of
the uracil moiety into the binding site of TS, usually sulphur
of the cysteine residue undergoes Michael type addition to
the C6-position of uracil [18]. However, competitive binding
of the chemically reactive substituent attached to the C5-posi-
tion of uracil (2 and 18) to the same sulphur of the cysteine
residue is also likely to occur. To check this possibility, we
have studied this reaction by mixing one of the most active
compounds (2) with cysteine and have obtained in short reac-
tion time and under comparatively mild conditions the product
(5) in a quite high yield (Scheme 1). This experiment pointed
out that reaction of 2 with cysteine residue of TS is likely to
occur. Even more important, very low bioactivity of 5 (Table 1)
suggested a high stability of this particular carbonesulphur
bond under biologically relevant experimental conditions,
42 ꢂ 0.2
42 ꢂ 2
G2/M
15 ꢂ 1.8
2
6
SubG1
G0/G1
S
25 ꢂ 0.5
50 ꢂ 0.1b
48 ꢂ 1
G2/M
8 ꢂ 0.8b
SubG1
G0/G1
S
17 ꢂ 3.5b
56 ꢂ 2.4b
24 ꢂ 1.5b
20 ꢂ 1
10 ꢂ 3.5
60 ꢂ 2.4b
39 ꢂ 1b
1 ꢂ 1b
24 ꢂ 0.4b
59 ꢂ 1b
27 ꢂ 0.2b
13 ꢂ 1
12 ꢂ 2
59 ꢂ 1
41 ꢂ 1b
0 ꢂ 0b
58 ꢂ 2b
43 ꢂ 5
44 ꢂ 3
13 ꢂ 1
G2/M
5-FU
SubG1
G0/G1
S
26 ꢂ 2
57 ꢂ 3b
43 ꢂ 3
0 ꢂ 0b
G2/M
a
The results are shown as percentages of cell population in each cell cycle
phase. The experiment was repeated three times.
b
Statistically significant at p < 0.05.

S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
1177
otherwise if cysteine of 5 would be a good leaving group,
compound 5 should be as active as 2 and 18. Therefore, it is
possible that the sulphur of the cysteine residue of TS reacts
with 2 and 18 in two different ways: either by forming a cova-
lent bond at C6-position of uracil (common for enzymatic
TSeuracil interaction), or by reacting with the substituent
provided with a good leaving group. The latter reaction
product would immediately block the TS activity.
Another possibility (which does not exclude the above-
mentioned) is that upon binding of a cysteine residue of TS
at C6-position of uracil (common TS process) an enolate is
formed, which then attacks the coenzyme (N,N-methylene
tetrahydrofolate or some of its iminium forms), giving a ter-
nary complex of enzymeesubstrateecoenzyme [18]. In this
case, a reactive side chain of 2 and 18 would be very close
to the basic part on the enzyme which usually abstracts a pro-
ton from C5-position of the uracil moiety [18]. That should
lead to the covalent bond formation between the 2 (or 18)
reactive side chain and the basic part on the enzyme. Such a co-
valently blocked enzyme active site would result in a perma-
nently stabilized ternary complex of the enzyme, coenzyme
and compound. Thus, the coenzyme will be blocked in a sim-
ilar manner as in the case of 5-FU.
the hypothetical mechanism of thymidylate synthase as the
most probable mode of action. Compounds 2 and 18 are
similarly active to 5-fluorouracil (5-FU), therefore, the same
target (TS) along with several modes of action significantly
differing from the mechanism of 5-FU offer new and interest-
ing research path for increasing the activity and also the selec-
tivity of uracil derivatives aiming toward antitumor therapy.
Intriguingly, high antiproliferative activity of compound 6
toward HeLa and MiaPaCa-2 cell lines is comparable to the
activity of 5-fluorouracil (5-FU) but the mechanism of action
of 6, as well as biological targets in the cells are not known.
Because of that, compound 6 and its close analogues could
be considered as a promising alternative to the numerous
5-FU derivatives targeting TS pathway. Therefore, further,
more detailed studies of 6 in order to determine the action
pathways responsible for biological activity are of highest
interest.
7. Experimental
7.1. General remarks
1
Proton H NMR spectra were obtained on a Varian Inova-
As another possibility of a ternary complex blocking, the
reactive side chain of 2 and 18 could also alkylate N,N-meth-
ylene tetrahydrofolate or more likely some of its open, chem-
ically more reactive iminium forms. That is in line with our
results, showing that the most active compound 2 is able to al-
kylate in a high yield primary aliphatic nitrogens (giving for
example compound 4), and even secondary aliphatic nitrogen
atoms, giving compound 6 (Scheme 1), although in the latter
case our reaction conditions are not biologically relevant.
Compound 6 also showed high antiproliferative activity
toward some tumor cell lines (Table 1), comparable to the
activity of compound 2. Intriguingly, opposite to 2, compound
6 is less effective toward MCF-7 and H 460 cell lines. That
could be also correlated to the different results in cell cycle
perturbation experiments obtained for 2 and 6, respectively,
whereby 6 induced an impressive increase in the number of
apoptotic cells after first 24 h, while during the next 48 h,
about 60% of cells died by apoptosis. These results point to
a different mechanism of action of 6 compared to 2 and 18.
The thymidylate synthase (TS) is probably not the target of
6, which is highly cytotoxic especially to HeLa and
MiaPaCa-2, while other cell lines are rather resistant. How-
ever, to give any firm hypothesis about its mechanism of
action, further research is needed.
400 (400 MHz) instrument with DMSO-d6 as an internal
standard. The chemical shifts are given as d in parts per
million, the coupling constants as J in hertz. Elemental analy-
ses were carried out with a LECO CHNS 932 elemental
¨
analyzer. Melting points were determined on a BOETIUS
melting point apparatus and are uncorrected. ESI mass spectra
were recorded on a Quattro LC (Micromass Inc.). Chromato-
graphic purifications were performed using either Silica gel
60 (Merck), RP-18 material (LiChroprepÔ), or Sephadex
G10 (Pharmacia). A MPLC (middle-pressure-liquid-chroma-
tography) system (Knauer) with two columns (silica gel: Euro-
sil Baselect, 25 ꢃ 200 mm, Merck; RP-18 material, 30 ꢃ 480,
Kronlab) was used with pressures ranging from 2 to 5 bar.
7.1.1. Biological studies
7.1.1.1. Proliferation assay. The HeLa (cervical carcinoma),
MiaPaCa-2 (pancreatic carcinoma), SW620 (colon carci-
noma), MCF-7 (breast carcinoma), H460 (lung carcinoma)
(all purchased from American Type Culture Collection e
ATCC) and HaCaT (skin keratinocytes, kindly provided by
´
Prof. Sonja Radakovic, Department of Dermatology, Medical
University of Vienna, General Hospital, Vienna, Austria)
human cell lines were seeded into a series of standard
96-well microtiter plates on day 0, at 1 ꢃ 104e3 ꢃ 104 cells/
mL, depending on the doubling times of the specific cell lines.
Test compounds were then added in 5-, 10-fold dilutions
(10ꢄ8e10ꢄ4 M) and incubated for a further 72 h. Stock solu-
tions were prepared in DMSO, (c ¼ 0.04 M), while working
dilutions were freshly prepared on the day of testing. The
solvent (DMSO) was also tested for eventual inhibitory activ-
ity by adjusting its concentration to be the same as in working
concentrations (DMSO concentration never exceeded 0.25%).
After 72 h of incubation the cell growth rate was evaluated by
6. Conclusions
Comparison of structureeactivity relationship in a series of
uracil derivatives revealed the importance of chemical activity
of the substituent attached to the C5-position of uracil with
regard to their antiproliferative activity on human tumor cell
lines. The results obtained by compound 2 and its acyclic
sugar analogue 18 pointed toward formation of a covalent
bond between reactive substituent and several targets within

1178
S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
performing the MTT assay, as described previously [19]. The
results were expressed as IC50, a concentration necessary for
50% of inhibition. The IC50 values for each compound were
calculated from doseeresponse curves using linear regression
analysis by fitting the test concentrations that give PG values
above and below the respective reference value (e.g., 50 for
IC50). Therefore, a ‘‘real’’ value for any of the response param-
eters is obtained only if at least one of the tested drug concen-
trations falls above, and likewise at least one falls below the
respective reference value. If however, for a given cell line
all of the tested concentrations produce PGs exceeding the
respective reference level of effect (e.g., PG value of 50),
then the highest tested concentration is assigned as the default
value, preceded by a ‘‘>’’ sign.
CH2), 8.09 (1H, s, C(6)H), 9.55 (1H, s, NHeCO), 10.72
(1H, br s, N(1)H), 11.53 (1H, br s, N(3)H); ESI-MS: m/z cal-
culated for C6H6N3O3Cl ([M þ H]þ): 204.58; found: 204.18.
7.2.3. 5-(Aminoacetyl)aminouracil (3)
A solution of 6-methyluracil 10 (520 mg, 2.55 mmol) in
concentrated ammonia (50 mL) was stirred at 22 ꢀC for 24
h. The ammonia was evaporated and the residue was dissolved
in water. Purification on a sephadex G10 column (25 mm ꢃ
100 mm) with water as eluent yielded 12 (381 mg, 81%) as
a white powder (Found: C, 39.27; H, 4.55; N, 30.13.
1
C6H8N4O3 requires C, 39.13; H, 4.38; N, 30.42%); H(400
MHz; DMSO-d6) 3.75 (2H, s, CH2), 7.48 (2H, br s, NH2),
8.06 (1H, s, C(6)H), 9.79 (1H, s, NHeCO), 10.77 (1H, br s,
N(1)H), 11.54 (1H, br s, N(3)H); ESI-MS: m/z calculated for
C6H8N4O3 ([M þ H]þ): 185.15; found: 185.12.
7.2. Cell cycle analysis
Cells were seeded (2 ꢃ 105 per well) in a 6-well plate. After
24 h the tested compounds were added at concentration of
50 mM. The attached cells were trypsinized, combined with
floating cells, washed with phosphate buffer saline (PBS)
and fixed in 70% ethanol 24, 48 and 72 h after the treatment
with compounds. Immediately before the analysis, the cells
were washed with PBS and stained with 2.5 mg/ml of propi-
dium iodide (PI) with the addition of 0.2 mg/ml of RNase A.
The stained cells were then analyzed with Becton Dickinson
FACSCalibur flow cytometer (20 000 counts were measured).
The percentage of the cells in each cell cycle phase was
determined using ModFit LTÔ software (Verity Software
House) based on the DNA histograms. As a minimum two
experiments were done in triplicates, and Student T-test
( p < 0.05) was used to measure the statistical significance.
7.2.4. 5-(N-Boc-ethylenediaminoacetyl)aminouracil (4)
To a suspension of 2 (120 mg, 0.6 mmol) in water (10 mL)
and methanol (3 mL), a solution of N-boc-ethylenediamine
(1.33 g, 8.3 mmol) in methanol (3 mL) was added and the
mixture was heated in a pressure cylinder under stirring at
90 ꢀC for 1 h. The solution was cooled down and the solvents
were evaporated. Then the residue was dissolved in water,
and purification on a sephadex G10 column with water as eluent
afforded product 4 (146.6 mg, 76%). (Found: C, 47.62; H, 6.57;
N, 21.26. C13H21N5O5 requires C, 47.70; H, 6.47; N, 21.39%);
1H (400 MHz; DMSO-d6) 1.35 (9H, s, 3 ꢃ Me), 2.53 (2H, m,
CH2eCH2), 2.97 (2H, m, CH2eCH2), 3.20 (2H, s, COe
CH2), 6.40 (1H, m, NHeCO), 6.76 (1H, m, CH2eNH ), 8.11
(1H, s, C(6)H), 9.32 (1H, s, C(5)NHeCO), 10.32 (1H, br s,
N(1)H), 11.10 (1H, br s, N(3)H); ESI-MS: m/z calculated for
C13H21N5O5 ([M þ H]þ): 328.34; found: 328.42.
7.2.1. Reagents
All solvents and reagents were purchased from commercial
sources and used without further purification. 5-Aminouracil
1, uracil 7, 5-hydroxyuracil 8, 5-iodouracil 9, 6-methyluracil
10, 5-iodomethyluracil 11, cytosine 19 and 4-thiouracil 23
were supplied by Serva. Uridine derivatives 12e17 [10], 5-
hydroxymethylcytosine 20 [11], 1-methylcytosine 21 [12],
4-methylthiouracil 24 [13], N4-hydroxycytosine 25 [14] and
N1-( p-toluenesulfonyl)-cytosine 26 [15,16] were prepared
according to the literature. Reactions were monitored by
thin-layer chromatography using TLC plates pre-coated with
silica gel 60F254 (Merck).
7.2.5. S-{2-[(2,4-Dioxo-1,2,3,4,-tetrahydropyrimidine-5-
yl)amino]-2-oxoethyl}-D-cysteine (5)
Boc-cysteine (221.27 mg, 1 mmol) was added to a solution
of 2 (203.58 mg, 1 mmol) in ethanol (5 mL) and 2 M aqueous
sodium hydroxide (3 mL). The solution was stirred at 22 ꢀC
for 2 h and then acidified with 5 M hydrochloric acid to
remove the protection group. The reaction product was dried
at the rotary evaporator and the residue was dissolved in
acetonitrile/water 1:9. Purification using MPLC on an RP18
column with acetonitrile/water 1:9 as eluent yielded 5
(219.1 mg, 76%). (Found: C, 37.35; H, 4.11; N, 19.39; S,
11.03. C9H12N4O5S requires C, 37.50; H, 4.20; N, 19.43; S,
1
7.2.2. 5-(Chloroacetylamino)uracil (2) [8]
11.12%); H (400 MHz; DMSO-d6) 2.94 (2H, m, SeCH2),
5-Aminouracil 1 (6.5 g, 0.05 mol) was dissolved in 0.2 M
sodium hydroxide solution (50 mL). To this solution, chloroa-
cetyl chloride (5 mL, 0.06 mol) and 1 M aqueous sodium
hydroxide solution (70 mL) were added within 30 min while
stirring in an ice bath. Stirring was continued at 22 ꢀC for
90 min, and then the solution was acidified with 5 M hydro-
chloric acid. The yellow precipitated crystals were separated
and recrystallized from water giving 2 (6.4 g, 62%) (Found:
C, 35.57; H, 2.81; N, 20.76. C6H6N3O3Cl requires C, 35.40;
3.39 (2H, s, COeCH2), 3.44 (1H, m, CH), 3.42 (2H, m,
NH2), 8.06 (1H, s, C(6)H), 9.33 (1H, s, NHeCO), 10.68
(1H, s, N(1)H), 11.47 (1H, s, N(3)H); ESI-MS: m/z calculated
for C9H12N4O5S ([M þ H]þ): 289.28; found: 289.18.
7.2.6. 5-[Bis-(2-p-methoxybenzylthioethyl)amine]acetylami
nouracil (6)
Compound 2 (305.37 mg, 1.5 mmol) and ammonium
sulphate (30 mg) were dissolved in hexamethyldisilazane
(7 mL) and refluxed for 3 h. The clear reaction mixture was
1
H, 2.97; N, 20.64%); H (400 MHz; DMSO-d6) 4.34 (2H, s,

S. Noll et al. / European Journal of Medicinal Chemistry 44 (2009) 1172e1179
1179
dried at a rotary evaporator. The residue was dissolved in
acetonitrile (1.5 mL), treated with triethylamine (110 mL)
and bis(2-p-methoxybenzylthioethyl)amine [9] (270 mg, 0.7
mmol) dissolved in acetonitrile (1 mL). This solution was
stirred at 90 ꢀC for 4 h. After cooling, water (2 mL) was added
and the solvent was evaporated. The residue was dissolved in
dichloromethane/methanol 10:1 (5 mL) and purified on a silica
gel column with dichloromethane/methanol 10:1 as eluent
yielding 6 (629.1 mg, 77%) (Found: C, 57.18; H, 6.01; N,
10.11; S, 11.59. C26H32N4O5S2 requires C, 57.33; H, 5.92;
(22.5 mL) was stirred at 22 ꢀC for 6 h. The pyridine was evap-
orated and the residue was dissolved in acetic acid ethyl ester
and extracted twice with water. The organic layer was dried
with sodium sulfate, filtered and the solvent was evaporated.
The dark brown oil was dissolved in dichloromethane (5 mL)
and separated on a silica gel column with dichloromethane/
methanol 50:1 as eluent. Thus 27 (327.4 mg, 88%) of high
purity was obtained (Found: C, 51.48; H, 4.50; N, 15.14; S,
11.32. C12H13N3O3S requires C, 51.60; H, 4.69; N, 15.04; S,
1
11.48%); H (400 MHz; DMSO-d6) 2.35 (3H, s, Me), 3.26
1
N, 10.29; S, 11.77%); H (400 MHz; DMSO-d6) 2.45 (4H,
(3H, s, NeMe), 6.50 (1H, d, J 7.6 Hz, C(5)H), 7.34 (2H, d, J
8.0 Hz, Ph), 7.73 (2H, d, J 8.3 Hz, Ph), 7.81 (1H, d, J 7.6 Hz,
C(6)H); m/z (ESI) 279.31 (Mþ 280.26). ESI-MS: m/z calculated
for C12H13N3O3S ([M þ H]þ): 280.31; found: 280.26.
m, 2 ꢃ SeCH2), 2.59 (4H, m, 2 ꢃ NeCH2), 3.14 (2H, s,
COeCH2), 3.63 (4H, s, 2 ꢃ CH2), 3.69 (6H, 2 ꢃ OeMe),
6.81 (4H, d, J 8.0 Hz, 4 ꢃ Ph), 7.17 (4H, d, J 8.0 Hz, 4 ꢃ
Ph), 8.09 (1H, s, C(6)H), 9.12 (1H, s, NHeCO), 10.66 (1H,
br s, N(1)H), 11.52 (1H, br s, N(3)H); ESI-MS: m/z calculated
for C26H32N4O5S2 ([M þ H]þ): 545.68; found: 545.39.
Acknowledgements
7.2.7. 5-(Chloroacetyl)amino-1-[(1,3-dihydroxy-2-
propoxy)-methyl]uracil (18)
The authors are indebted to Brigitte Große for her careful
experimental assistance, and Madlen Matterna for ESI-MS
measurements. Support for this study was provided by Minis-
try of Science, Education and Sports of the Republic of
Croatia (Projects 098-0982914-2918 and 098-0982464-2514).
Compound 14 (462 mg, 2 mmol) was dissolved in 1 M aque-
ous sodium hydroxide (30 mL) and cooled down to 0 ꢀC. To
this solution, chloroacetyl chloride (2 mL) and 1 M aqueous
sodium hydroxide (25 mL) were added under stirring within
30 min. Stirring was continued at 22 ꢀC for 1 h. After acidifica-
tion with 5 M hydrochloric acid, the solvent was removed by
rotary evaporation and the residue was purified using MPLC
on an RP18 column with acetonitrile/water 1:4 as eluent giving
18 (461.1 mg, 75%) (Found: C, 39.18; H, 4.72; N, 13.49.
References
[1] D. Sampath, V.A. Rao, W. Plunkett, Oncogene 22 (2003) 9063e9074.
[2] D.B. Longley, D.P. Harkin, P.G. Johnston, Nat. Rev. Cancer 3 (2003)
330e338.
1
C10H14N3O6Cl requires C, 39.04; H, 4.59; N, 13.66%); H
(400 MHz; DMSO-d6) 3.41 (4H, m, 2 ꢃ CH2eOH), 3.51 (1H,
m, CH), 4.35 (2H, s, COeCH2), 4.61 (2H, t, J 5.2 Hz,
2 ꢃ OH), 5.20 (2H, s, NeCH2), 8.42 (1H, s, C(6)H), 9.63
(1H, s, NHeCO), 11.78 (1H, s, N(3)H); ESI-MS: m/z calculated
for C10H14N3O6Cl ([M þ H]þ): 308.69; found: 308.33.
[3] M. Goulian, B.M. Bleile, L.M. Dickey, R.H. Grafstrom, H.A. Ingraham,
S.A. Neynaber, M.S. Peterson, B.Y. Tseng, Adv. Exp. Med. Biol. 195
(1986) 89e95.
[4] (a) A. Gangjee, E. Elzein, M. Kothare, A. Vasudevan, Curr. Pharm. Des.
2 (1996) 263e280;
(b) E.M. Berman, L.M. Werbel, J. Med. Chem. 34 (1991) 479e485;
(c) R.C. Jackson, in: A.L. Jackman (Ed.), Antifolate Drugs in Cancer
Therapy, Humana Press, Totowa, 1999, pp. 1e12.
7.2.8. 1-Methyl-5-hydroxymethylcytosine (22)
[5] K.L. Mukherjee, C. Heidelberger, J. Biol. Chem. 235 (1960) 433e437.
[6] M. Malet-Martino, P. Jolimaitre, R. Martino, Curr. Med. Chem. Anti
Cancer Agents 2 (2002) 267e310.
Compound 22 was prepared starting from 5-hydroxyme-
thylcytosine 20 (1.41 g, 10 mmol) suspended in DMF (100
mL). To this suspension, 1 M methanolic tetrabutylammonium
hydroxide (10 mL) was added dropwise. After achieving
a clear solution, methyliodide (1.3 mL, 20 mmol) dissolved
in DMF (5 mL) was added within 10 min, and then the
mixture was stirred at 22 ꢀC for 2 h. The reaction mixture
was dried by rotary evaporation. Column chromatography on
silica gel with dichloromethane/methanol 1:1 as eluent yielded
22 (1.27 g,82%) as a white powder (Found: C, 46.61; H, 5.93;
N, 27.18. C6H9N3O2 requires C, 46.45; H, 5.85; N, 27.08%);
1H (400 MHz; DMSO-d6) 3.20 (3H, s, Me), 4.14 (2H, d, J
4.8 Hz, CH2), 4.98 (1H, t, J 5.2 Hz, OH), 7.53 (1H, s,
C(6)H); ESI-MS: m/z calculated for C6H9N3O2 ([M þ H]þ):
156.16; found: 156.24.
[7] H. Kazuno, Y. Shimamoto, H. Tsujimoto, M. Fukushima, A. Matsuda,
T. Sasakio, Oncol. Rep. 17 (2007) 1453e1460.
[8] G.H. Hitchings, R.L. Thomson, U.S. Patent 2,494,125, 1950.
[9] J.L. Corbin, K.F. Miller, N. Pariyadath, S. Wherland, A.E. Bruce,
E.I. Stiefel, Inorg. Chim. Acta 90 (1984) 41e51.
[10] M. Grote, St. Noll, B. Noll, B. Johannsen, W. Kraus, Can. J. Chem. 82
(2004) 513e523.
[11] T.L.V. Ulbricht, C.C. Price, J. Org. Chem. 21 (1956) 567e571.
[12] R.S. Hosmane, N.J. Leonhard, Synthesis (1981) 118e119.
[13] J.J. Fox, D. van Praag, J. Am. Chem. Soc. 82 (1960) 486e489.
[14] T.J. Delia, M.J. Olsen, G.B. Brown, J. Org. Chem. 30 (1965) 2766e2768.
ˇ
[15] B. Zinic, I. Krizmanic, D. Vikic-Topic, M. Zinic, Croat. Chem. Acta 74
ˇ
´
´
´
´
´
(1999) 957e966.
ˇ
[16] B. Kasnar, I. Krizmanic, M. Zinic, Nucleosides Nucleotides 16 (1997)
ˇ
´
´
1067e1071.
[17] J.L. Grem, Invest. New Drugs 18 (2000) 299e313.
[18] R.B. Silverman, The Organic Chemistry of Drug Design and Drug
Action, Elsevier Academic Press, New York, 2004, pp. 298e302.
7.2.9. 1-Methyl-N4-( p-toluenesulfonyl)cytosine (27)
A solution of 1-methylcytosine 21 (166.68 mg, 1.5 mmol)
and tosyl chloride (1.29 g, 6.75 mmol) in dry pyridine
ˇ
[19] M. Hranjec, M. Kralj, I. Piantanida, M. Sedic, L. Suman, K. Pavelic,
´
´
G. Karminski-Zamola, J. Med. Chem. 50 (2007) 5696e5711.