Design, synthesis and in vitro drug release investigation of new potential 5-FU prodrugs

Article abstract of DOI:10.1016/j.ejmech.2011.04.010

In order to identify new efficient prodrugs of 5-fluorouracil (5-FU) and to develop an original targeting approach using 2-fluoro-2-deoxyglucose (FDG) as a potential drug carrier, eight original 5-FU derivatives were synthesized: 5-FU was attached by the

Full text of DOI:10.1016/j.ejmech.2011.04.010

European Journal of Medicinal Chemistry 46 (2011) 2867e2879
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Design, synthesis and in vitro drug release investigation of new potential
5-FU prodrugs
,
,
a
,
,
Pierre Daumar a b, Caroline Decombat c , Jean-Michel Chezal b, Eric Debiton a b, Michel Madesclaire c,
*
,
,b
Pascal Coudert a b, Marie-Josèphe Galmier a
a Clermont Université, Université d’Auvergne, Imagerie moléculaire et thérapie vectorisée, BP 10448, F-63000 Clermont-Ferrand, France
b Inserm, UMR 990, F-63005 Clermont-Ferrand, France
c Clermont Université, Université d’Auvergne, BP 10448, F-63000 Clermont-Ferrand, France
a r t i c l e i n f o
a b s t r a c t
Article history:
In order to identify new efficient prodrugs of 5-fluorouracil (5-FU) and to develop an original targeting
approach using 2-fluoro-2-deoxyglucose (FDG) as a potential drug carrier, eight original 5-FU derivatives
were synthesized: 5-FU was attached by the N1 position of the pyrimidinic ring to the C1 position of the
FDG structure either by direct coupling (2a) or via various spacers (3, 6aec, 10b and 19). A new sensitive
high-performance liquid chromatography method was developed to simultaneously quantify 5-FU and
its derivatives in human plasma and other relevant media at physiological temperatures. Half-lives were
determined from the degradation profiles of these conjugates. Slow degradation of compounds 2a, 3, 10b
and 19 was observed in vitro at 37 ꢀC, but no 5-FU release was noticed. By contrast, the in vitro drug
release profiles of compounds 6aec followed pseudo-first-order kinetics, and 5-FU was found in all the
media. The antiproliferative activity of the eight compounds was assessed in vitro by a fluorometric assay
against two human solid cancer cell lines and one healthy cell line. A correlation was found between the
activities of the compounds and their ability to release 5-FU efficiently.
Received 17 December 2010
Received in revised form
23 March 2011
Accepted 3 April 2011
Available online 14 April 2011
Keywords:
5-Fluorouracil (5-FU)
2-Fluoro-2-deoxy-
Prodrug stability
HPLC
D-glucose derivatives
Chimaeric prodrugs
Anti-cancer prodrugs
Ó 2011 Elsevier Masson SAS. All rights reserved.
1. Introduction
Numerous derivatives have been synthesized to improve the
physicochemical, biopharmaceutical and pharmacokinetic proper-
5-Fluorouracil (5-FU; Fig. 1) was introduced as a rationally
designed anti-cancer agent more than fifty years ago [1]. It is widely
used in the treatment of solid malignant tumors such as colorectal
cancers. However, 5-FU displays various toxicities owing to its
nonspecific cytotoxicity for tumor cells, loses efficiency through
poor distribution to tumor sites, and is seriously limited in effec-
tiveness by drug resistance. Despite some improvements in
response rates achieved by combining 5-FU with newer chemo-
therapies such as irinotecan and oxaliplatin, new therapeutic
strategies that directly act on the 5-FU activity are still needed [2,3].
Over the years, a better understanding of its mechanism of
action has led to the development of several strategies for
increasing the anti-cancer activity of 5-FU: to decrease its rapid
metabolism, to increase its conversion into active metabolites, to
maintain high concentrations in both blood and tumor over long
periods of time, and more recently to identify predictive
biomarkers of response to 5-FU-based chemotherapy [4].
ties of 5-FU. Some of these have found a place in clinical practice,
their main benefit being for oral administration. For example,
capecitabine is an oral prodrug that allows higher intratumoral
levels of 5-FU to be reached through tumor-specific bioconversion
[5,6].
The control of drug level at the tumor site is a major factor on
which the efficiency of cancer chemotherapy directly depends
[7e11]. To obtain selective delivery of 5-FU to a tumor site, most
attempts have focused on the development of new drug delivery
systems based on passive targeting strategies [12]. To the best of
our knowledge, selective delivery of 5-FU toward tumor sites using
an active targeting strategy has never been reported, except with
the ADEPT strategy (antibody-directed enzyme prodrug therapy)
[13,14].
Active tumor-targeting strategies have been explored by our
team for several years using various vectors or carriers [15e17]. In
the course of this research, we recently chose to take advantage of
the well-documented, typical biochemical phenotype of invasive
solid tumors, namely the upregulation of glycolysis. Glycolysis-
mediated targeting using a glucose analog as an active drug carrier
is useful for the selective delivery of a drug to cancer cells, and
* Corresponding author. Tel.: þ33 4 7317 7993; fax: þ33 4 7317 7090.
E-mail address: caroline.decombat@u-clermont1.fr (C. Decombat).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.04.010

2868
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
Scheme 2. Synthesis of 3. Reagents and conditions: (i) 5-FUA, DCC, HOBt, DMF.
plasmatic proteins result in higher stability and resistance to
hydrolysis in biological media. Thus, Buur and Bundgaard [25]
reported that if arylcarbamoyl-5-fluorouracil derivatives display
very short biological half-lives, alkylcarbamoyl-5-fluorouracil have
biological half-live that might perfectly meet our requirements in
terms of stability for potential 5-FU prodrugs. Moreover, their
stability can be increased by varying the length of the alkyl chain.
Therefore, we synthesized compounds 6a, 6b and 6c, which display
alkyl chains with one, three and five methylene groups, respectively.
Finally, as a consequence of a more hindered environment of the
carbamoyl linkage, 1-(N-methylcarbamoyl)-5-fluorouracil deriva-
tives have been reported to have higher stability than their non-
methylated analogs [25]. This led us to design the two last
compoundsof ourset:derivatives 10b and19, whichpresentan alkyl
and an aryl spacer arm respectively. We tried to figure out how the
length of the aliphatic spacer arm would affect the stability of our
derivatives. Thus, two synthetic pathways leading to10a, which only
differs from 10b by the size of the spacer, were investigated. Despite
our efforts, this compound was not isolated but we chose to report
the synthesis of the intermediates.
Synthesis and full characterization of 5-FU derivatives are
described. Additionally, a new HPLC method to investigate both
their plasma and chemical stability was developed. The method
was used for the determination of the in vitro release profiles of 5-
FU from starting derivatives in three different media: Phosphate
Buffer Saline (PBS), human plasma and cell culture medium (MEM),
under physiological pH and temperature conditions, which gave
access to half-lives of the designed compounds. Finally their anti-
proliferative activity was assessed in vitro by a fluorometric assay
against two human solid cancer and one healthy cell lines.
Fig. 1. Structures of FDG, 5-FU, 5-FUA and HCFU.
numerous studies have already been undertaken in this research
area. We investigated an original tumor-targeting approach, using
the glucose analog 2-fluoro-2-deoxyglucose (FDG; Fig. 1) as a low
molecular weight drug carrier. This carbohydrate is known to be
trapped by primary and metastatic human tumors. Thus coupling an
FDG moiety with an anti-cancer agent could trap the glucoconjugate
inside the tumor cells until the active moiety is released. Following
this strategy, we have obtained promising results with FDG-coupled
chlorambucil derivatives [18,19].
Here, we designed a series of 5-FU glycoconjugates in which the
5-FU was attached by the N1 position of the pyrimidinic ring to the
C1 position of the FDG structure either by direct coupling or via
various spacers. As the 5-FU was extensively derivatized, we took
advantage of the rich related literature to adapt the stability of
our 5-FU derivatives to our practical purpose. The following
compounds were synthesized: direct-coupled derivatives 2a and
2b (Scheme 1), the 1-alkyl-5-fluorouracil derivative 3 (Scheme 2),
1-carbamoyl-5-fluorouracil derivatives 6aec (Scheme 3), and 1-(N-
methylcarbamoyl)-5-fluorouracil derivatives 10b (Scheme 5), and
19 (Scheme 6).
Indeed, galactopyranosyl-5-fluorouracil derivatives were pre-
sented in 1999 as interesting antitumor agents having lower
toxicities than the parent drug [20]. Moreover, the coupling step
between the carbohydrate and the pyrimidine takes place via
a simple and well described chemical reaction. This led us to first
investigate the direct coupling between FDG and 5-FU with the
synthesis of derivatives 2a and 2b.
Also, we synthesized the 1-alkyl-5-fluorouracil derivative 3.
According to the literature [21], this derivative should have anti-
carcinogenic activity and allow a slow 5-FU release.
2. Results and discussion
2.1. Chemistry
The synthesis of all the 5-FU derivatives was designed as
follows: first, functionalized FDG analogs 1a and 1b, which were
key intermediates, were synthesized as reported previously [18].
On the anomeric position of these precursors was grafted the
spacer arm, which was most often built in several steps. In a final
coupling step, the desired linking bond between 5-FU (N1 position)
or (5-fluorouracil-1-yl) acetic acid (5-FUA; Fig. 1), and the sugar
moiety was formed.
Finally, carbamoyl derivatives represent a deeply investigated
class of compounds, which is due to their suitable properties for oral
administration, as referred to before. This can be illustrated by the
clinical development of 1-hexylcarbamoyl-5-fluorouracil (HCFU;
Fig. 1), which was discovered in Japan in 1977. Since then, several
reports were published with regards to their physicochemical and
biological properties [22e27]. Interestingly, binding affinities to
2.1.1. Synthesis of 2a and 2b (Scheme 1)
We synthesized the glyconjugate 2a and its deacetylated analog
2b, for which the 5-FU was directly linked to the FDG moiety.
Compound 2a was prepared using a previously reported method
for nucleoside synthesis [28], starting from 5-FU and peracetylated
FDG 1a [18]. First, 5-FU reacts with hexamethyldisilazane (HMDS)
and trimethylchlorosilane (TCS) in dry acetonitrile under argon to
form the intermediate 2,4-bis(trimethylsilyl)-5-fluorouracil. This
species then reacts with 1a in the presence of Lewis acid SnCl4 to
form 2a as the
b form only. The stereochemistry was confirmed
Scheme 1. Synthesis of 2aeb. Reagents and conditions: (i) 5-FU, HMDS, TCS, CH3CN;
(ii) SnCl4; (iii) MeONa/MeOH.
from 1H NMR data: J1,2 ¼ 9 Hz ¼ Jax,ax [18]. Subsequent Zemplèn

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2869
Scheme 3. Synthesis of 6aec. Reagents and conditions: (i) 1b, DCC, HOBt, DMF; (ii) H2 Pd/C, DCM/MeOH; (iii) 5-FU, triphosgene, pyridine.
deacetylation [29,30] of 2a using sodium methylate led to 2b in
the appropriate aliphatic primary amine 5aec to give the desired
quantitative yield.
compound 6aec, respectively. For this final step, the yields ranged
from 18 to 60%. The formation of the carbamoyl bond was
confirmed by 13C NMR, which showed the appearance of a new
signal at 145.6e150.8 ppm (NHCON).
2.1.2. Synthesis of 1-alkyl-5-fluorouracil derivative 3 (Scheme 2)
5-FUA which was obtained in two steps, using a previously
described general method [31], was a key intermediate for the
synthesis of the 1-alkyl-5-fluorouracil derivative 3. The reaction of
glucosamine 1b with 5-FUA, activated by the coupling agents N,N0-
dicyclohexylcarbodiimide (DCC) and hydroxybenzotriazole (HOBt),
led to the desired compound 3.
2.1.4. Synthesis of 1-(N-methylcarbamoyl)-5-fluorouracil
derivatives
2.1.4.1. Derivatives 10aeb and intermediates 8aeb, 9aeb and
12e14a (Schemes 4 and 5). Although compounds 10a and 10b are
the N-methylated analogs of derivatives 6a and 6b respectively,
their synthetic pathways were fully reconsidered.
2.1.3. Synthesis of 1-carbamoyl-5-fluorouracil derivatives 6aec and
intermediates 4, 5aec (Scheme 3)
Two synthetic routes leading to 10a were investigated
(Scheme 4). First, amine 9a was obtained in two steps from N-
methylated Z-glycine 7a [34]. Compound 7a was coupled with
glucosamine 1b using DCC/HOBt to give compound 8a, which was
further deprotected by hydrogenolysis to give amine 9a. Then,
a procedure adapted from Ozaki [22,24] was performed for the
derivatization of 5-FU toward 1-(N-methylcarbamoyl)-5-fluoro-
uracil derivatives: the suitable amine is condensed with tri-
phosgene to give the corresponding carbamoyl chloride, which can
either be isolated or used in situ if unstable, and then react with 5-
FU to give the 1-(N-methylcarbamoyl)-5-fluorouracil derivatives.
Unfortunately, this method was not suitable for the coupling of
amine 9a with 5-FU via a N-methylcarbamoyl bond. Therefore we
investigated the second five-step synthesis depicted on Scheme 4:
the salt 13 was initially obtained by esterification of acid 11 fol-
lowed by selective N-BOC deprotection of compound 12 [34]. Then,
the N-methylcarbamoyl bond to 5-FU was formed according to the
procedure described above in a 30% yield to give intermediate 14a.
At this point, two more steps were necessary to obtain the desired
To obtain the derivatives 6aec and form the carbamoyl bond, we
used one out of the three main synthetic methods described for the
synthesis of 1-carbamoyl-5-fluorouracil derivatives [22,24].
However, we preferred to replace phosgene by its solid synthetic
equivalent triphosgene, which is safer to handle [32].
The carbamoyl bond formation being the final step of the
syntheses, primary amines 5aec had to be obtained first. Syntheses
of amines 5a and 5b were previously described by our team [18],
and amine 5c was obtained in two steps, starting from 6-azido-
hexanoic acid (4), which was prepared according to a previously
described method [33]: the reaction of glucosamine 1b with 6-
azidohexanoic acid, activated by the coupling agents DCC and HOBt,
gave the azido compound 4. The subsequent hydrogenolysis of the
azide function was carried out with palladium on charcoal (Pd/C)
and gave the desired product 5c in good yield (82%).
To form the carbamoyl bond, 5-FU was condensed with tri-
phosgene in a basic medium at 0 ꢀC to form the unstable 1-chlor-
oformyl-5-fluorouracil species first, which then reacted in situ with
Scheme 4. Synthesis of 10a. Reagents and conditions: (i) 1b, DCC, HOBt, DCM; (ii) H2 Pd/C, MeOH/THF; (iii) triphosgene, NaHCO3, DCM; (iv) 5-FU, NaH, DMAc; (v) BnBr, DBU, CH3CN;
(vi) TFA, THF; (vii) H2 Pd/C, THF.

2870
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
Scheme 5. Synthesis of 10b. Reagents and conditions: (i) ClCO2Et, TEA, 1b, DCM; (ii) H2 Pd/C, THF/MeOH; (iii) triphosgene, NaHCO3, DCM; (iv) 5-FU, NaH, DMAc.
compound 10a: hydrolysis of benzyl ester 14a, and subsequent
coupling of acid 14b to glucosamine 1b. However, hydrogenolysis of
the ester 14a led to its acid analog 14b which was very unstable and
could not be isolated.
the loss of starting material and the formation of 5-FU or degraded
products was developed. For all compounds, the obtained chro-
matograms of the stability study were compared with those of the
media without 5-FU derivative. Only additional chromatographic
peaks were studied.
So we turned our efforts to the synthesis of a compound which
presents a longer spacer arm. According to this strategy, compound
10b was obtained in three steps, starting from derivative 7b
(Scheme 5). This latter compound was prepared from pyrrolidinone
in a multi-step synthesis, according to previous reports [35e37].
First, 7b was coupled with glucosamine 1b by a procedure which
used ethylchloroformate (ClCO2Et) and triethylamine (TEA). The
subsequent step, which led to secondary amine 9b, was the removal
of the benzyl carbamate N-protecting group by hydrogenolysis. The
final derivative 10b was obtained from secondary amine 9b
according to the N-methylcarbamoyl bond formation method
described above, which proved to be fairly efficient in this case.
2.2.1. Stability assessment of compounds 2a, 3, 10b and 19
Similar degradation profiles were observed for these
compounds, regardless of the media. Typical HPLC-UV chromato-
grams obtained from their stability study at 37 ꢀC in MEM are
presented on Fig. 2. These conditions were set to analyze 5-FU, 2a,
3, 10b and 19, which present retention times of 5.4, 18.1, 16.9, 17.6
and 19.7 min, respectively. The chromatographic profiles for these
compounds evolve in the same way in the course of time: intensity
of chromatographic peak of starting material decreased while two
or three new peaks appeared at shorter retention times (Table 1).
Nevertheless, the chromatograms obtained after 10 days showed
that the 5-FU was not released in the conditions of this study. In
order to characterize these degradation products (DP), the HPLC-
UV method developed was combined with MS detection. The
results obtained with compound 19 are presented in Fig. 3. The
mass spectra in negative mode showed [M ꢁ H]ꢁ peaks at m/z 568
(19), m/z 484 (DP2) and m/z 442 (DP3) for the chromatographic
peaks at 19.7 min, 16.2 min and 14.9 min respectively. The differ-
ence of 84 mass units between 19 and DP2 and a difference of 42
mass units between DP2 and DP3 suggest a loss of two or one
neutral CH2CO. These results could be consistent with a di- and tri-
deacetylation of compound 19. The results obtained with 10b are
presented in Fig. 4. For each identified compound the mass spectra
in positive mode showed the protonated compound [M þ H]þ and
sodium and potassium adducts [M þ Na]þ and [M þ K]þ. For the
compound 10b, the mass spectrum showed peaks at m/z 563, m/z
585 and m/z 601 for the chromatographic peak at 17.6 min. The
mass spectra showed [M þ H]þ at m/z 521, (DP1), m/z 479 (DP2)
and m/z 437 (DP3) for the chromatographic peaks at 16.4, 14.6 and
2.1.4.2. Derivative 19 and intermediates 16e18 (Scheme 6). Com-
pound 19 was obtained in four steps starting from arylamine 15,
which was synthesized using a procedure described by our team
[18]. The primary amine 15 was protected as a benzyl carbamate,
allowing us to carry out the subsequent monomethylation of the
compound 16 to form the derivative 17. After hydrogenolysis of
the benzyl carbamate protecting group, the secondary amine 18
was allowed to react in the final step of the synthesis to give
derivative 19.
2.2. In vitro stabilities of 5-FU derivatives
The eight final compounds 2a, 2b, 3, 6aec, 10b and 19 were
incubated at 37 ꢀC in three relevant media: Phosphate Buffer Saline
(PBS), cell culture medium (MEM) and human plasma. At appro-
priate intervals, samples were withdrawn and prepared as
described in experimental protocols.
A new sensitive high-
performance liquid chromatography method able to monitor both
Scheme 6. Synthesis of 19. Reagents and conditions: (i) ClCO2Bn, NaHCO3, H2O/acetone; (ii) NaH, MeI, THF; (iii) H2 Pd/C, THF/MeOH; (iv) triphosgene, NaHCO3, DCM; (v) 5-FU, NaH,
DMAc.

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2871
Table 1
Retention times of chromatographic peaks for each compound and its degraded
products (DP).
Compound
Retention time (min)
Parent product
DP1
DP2
DP3
2a
3
10b
19
18.1
16.9
17.6
19.7
ND
ND
16.4
ND
14.2a
13.5a
14.6
16.2
7.3
6.7a
13.4
14.9
ND: not detected.
a
not identified.
13.4 min respectively. The difference of 42 mass units between each
compound confirmed a loss of neutral CH2CO.
At this time, it was hypothesized that these degraded product
resulted from subsequent deacetylation reactions. To have this
hypothesis confirmed, the trideacetylated analog 2b of compound
2a was synthesized (Scheme 1) and analyzed in the same condi-
tions. The same retention time (7.3 min) was observed for 2b and
DP3, which confirms that the most hydrophilic degradation
product observed results from the trideacetylation of 2a.
2.2.2. Stability assessment of compounds 6a, 6b and 6c
Similar patterns were found for all the three potential prodrugs
tested in the three media. The chromatograms obtained in MEM at
37 ꢀC forcompounds6aec areshowninFig.5. Theretentiontimes for
6a, 6band6carerespectivelygivenat17.8min,17.9minand18.7min.
An exponential decay of 5-FU derivative concentration (Fig. 6)
accompanied by only an exponential increase in 5-FU concentration
was observed. The degradation profiles for compounds 6aec
appearedto follow pseudo-first-orderkineticsinthedifferentmedia.
From the individual disappearance profiles of the 5-FU derivatives,
the constants rate and the half-lives were estimated from plots of log
sample concentration versus time. Half-lives are reported in Table 2.
Compound 6a was pretty unstable in PBS and MEM at physio-
logical temperature and pH, as shown by the estimated half-lives
(t0.5: 2e3 min) obtained from the linear part of the kinetic plot.
In plasma, compound 6a also showed fast release of 5-FU under the
same conditions, but less rapidly though (t0.5 ¼ 6 min).
For 6b, the estimated t0.5 values were higher in all three media
(t0.5: 6e10 min). Half-life values indicated a lower rate of hydrolysis
compared to 6a.
The qualitative patterns of 6c hydrolysis were similar to that
observed for 6b, with the longest half-lives in the different media
(t0.5: 8e17 min).
According to these results, the release of 5-FU from each of these
compounds was slower in plasma than in pH 7.4 buffer or MEM,
which is very likely due to interactions with the plasmatic proteins.
Moreover, an increase in the length of the linker from one (6a) to
five carbons (6c) increases the half-life in the different media. For
example, t0.5 rose from 6 ꢂ 5 min to 17 ꢂ 3 min in plasma for
compound 6c. These data perfectly match the results previously
reported in the literature [25].
3. Biological activities
A preliminary evaluation of the antiproliferative activities of all
compounds was performed in vitro by a fluorometric assay (Resa-
zurin reduction test (RRT)), against two human solid cancer cell
lines (i.e. MCF 7: breast adenocarcinoma and PA 1: ovarian carci-
noma), and a healthy cell line (i.e. a human fibroblast primary
culture). Compounds 2a, 3, 10b and 19 were inactive against the
Fig. 2. (A) Typical HPLC-UV chromatograms monitored under the optimized condi-
tions of blank (MEM) and 5-FU at 100 mM. Typical HPLC-UV chromatograms from the
stability experiments in MEM at 37 ꢀC of compounds (B) 2a at 1 min, 24 h and 96 h, (C)
3 at 1 min, 96 h and 10 days, (D) 10b and (E) 19 at 1 min, 24 h and 48 h.
tested cell lines, showing IC50 superior to 50 mM, but compounds
6aec showed antiproliferative activities comparable to those
obtained with 5-FU (Table 3). These results suggest a strong

2872
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
100
50
A
(1) HPLC-UV
0
100
(2) HPLC-MS (TIC)
(3) HPLC-EIC 568
(4) HPLC-EIC 484
(5) HPLC-EIC 442
50
0
100
50
0
100
50
0
100
50
0
5
10
15
25
30 t (min)
20
568
15
B
[M-H]- 568
(1)
10
5
541
540
138
173
764
785
347
247
434
355
715
477
484
315
601 637
63
29
[M-H]- 484
20
15
10
5
(2)
456
249
245
741
485
578
641
147
91
714
423
442
779
198
257
62
321 353
100
[M-H]- 442
80
60
40
20
(3)
175
444
511
315 357 399
300 400
526 576
615
133
249
761
100
200
500
600
700
800
m/z
Fig. 3. Stability study of compound 19 at 37 ꢀC in MEM. (A) HPLC-UV-MS chromatograms of 19 and its degradation products at 48 h (1) HPLC-UV chromatogram; (2) HPLC-MS total
ion chromatogram (TIC) at m/z 50 to m/z 800; (3) Extracted ion current (EIC) chromatogram m/z 568 for 19 at 19.7 min; (4) EIC chromatogram m/z 484 for DP2 at 16.2 min; (5) EIC
chromatogram m/z 442 for DP3 at 14.9 min. (B) HPLC-ESI mass spectra of 19 (1) and its degradation products DP2 (2) and DP3 (3).
correlation between the biological activity of our compounds, and
their ability to release 5-FU efficiently.
purpose, we introduced various spacers and various types of
chemical functions on the 5-FU structure. A new HPLC-UV method
was developed to investigate the stability of the derivatives and
determine their half-lives from their degradation profiles in three
different media. In these in vitro conditions, 5-FU was not released
from the direct-coupled derivative 2a, the 1-alkyl-5-fluorouracil
derivative 3 or 1-(N-methylcarbamoyl)-5-fluorouracil derivatives
4. Conclusions
The synthesis of eight original FDG-coupled 5-FU derivatives is
reported. To adapt the stability of these derivatives to our practical

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2873
[M+H]+
563
100
80
(1)
60
[M+Na]+
585
40
+
601
[M+K]
20
59
59
433
389
391
602
100
100
346
331
680
291
503
702
171
168
243
559
521
750 798
+
[M+H]
10
8
6
4
2
(2)
[M+Na]+
543
559
[M+K]+
680
702
718
309
83
453
287
489
479
576
115
785
191
599 676
[M+H]+
13
10
[M+Na]+
(3)
59
501
409
371
349
5
[M+K]+
517
193
327
423
459
100
215
267
130
518
555
655
758
796
691
+
[M+Na]
100
80
[M+H]+
(4)
437
60
40
[M+K]+
609
588
460
482
307
20
59
87
151
239 272
327
413
662
697 741 792
556
m/z
Fig. 4. Stability study of compound 10b at 37 ꢀC in MEM. HPLC-ESI mass spectra of 10b (1) and its degradation products DP1 (2), DP2 (3) and DP3 (4).
10b and 19. On the other hand, 1-carbamoyl-5-fluorouracil deriv-
atives 6aec showed 5-FU release in the three studied media, with
estimated half-live values bounded by 2 and 17 min. Moreover,
these compounds and 5-FU showed similar antiproliferative
activities against the tested cell lines. Among the three 1-carba-
moyl-5-fluorouracil derivatives, 6c has the longest half-life, which
makes it the best candidate to be further in vivo investigated, and
potentially become an efficient 5-FU prodrug.
5.1.1. 1-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
5-fluorouracil, 2a
b-D-glucopyranosyl)-
Hexamethyldisilazane (HMDS) (0.26 mL, 1.20 mmol) and tri-
methylchlorosilane (TCS) (0.16 mL, 1.25 mmol) were successively
added to a solution of 5-FU (150 mg, 1.15 mmol) and sugar 1a
(350 mg, 1.00 mmol) in anhydrous acetonitrile (CH3CN) (10 mL)
under argon. The mixture was stirred at room temperature (rt) for
30 min and SnCl4 (0.16 mL, 1.40 mmol) was added. The solution was
heated at 45 ꢀC for 24 h and then diluted with ethyl acetate (EtOAc)
(20 mL), washed twice with water (10 mL), dried over MgSO4 and
concentrated under vacuum. The resulting residue was purified by
silica gel chromatography (cyclohexane e EtOAc, 5:5) to yield 2a
(244 mg, 0.58 mmol) as a white solid; mp 103 ꢀC; yield 58%.
(ꢁ)-ESI-MS: 418.8 [M ꢁ H]ꢁ.
5. Experimental protocols
5.1. Chemistry
Chemicals were purchased from commercial sources (Sigmae
Aldrich or Acros) and were used without further purification. 1H
and 13C NMR spectra were recorded on a Bruker AM-200 (4.5 T)
spectrometer operating at 200 and 50 MHz respectively. Chemi-
cal shifts are reported in parts per million relative to the inter-
nal tetramethylsilane standard for 1H and deuterated solvents
for 13C, which were purchased from C.E.A. Saclay (acetone-d6,
1H NMR (CDCl3):
d 9.95 (brs, 1H, 5-FU:NH-3); 7.48 (d, 1H,
J ¼ 5.7 Hz, 5-FU:H-6); 6.06 (d, 1H, J ¼ 9.2 Hz, H-1); 5.59 (td, 1H,
J ¼ 13.2 Hz, 9.2 Hz, H-3); 5.15 (t, J ¼ 9.2 Hz, 1H, H-4); 4.57 (td, 1H,
J ¼ 50.0 Hz, 9.2 Hz, H-2); 4.28 (dd, 1H, J ¼ 12.6 Hz, 4.9 Hz, H-6a);
4.18e4.03 (m, 2H, H-5, H-6b); 2.09, 2.08, 2.05 (3s, 9H, 3ꢃ CH3).
13C NMR (CDCl3):
d
170.61, 169.84, 169.73 (3ꢃ OCOCH3); 156.48
2
(5-FU:C-4, JCeF ¼ 26.7 Hz); 149.16 (5-FU:C-2); 141.24 (5-FU:C-5,
d
¼ 29.8 ppm; CDCl3,
d
¼ 77.2 ppm; DMSO-d6,
d
¼ 39.5 ppm).
1JCeF ¼ 239 Hz); 122.86 (5-FU:C-6, JCeF ¼ 34.0 Hz); 87.25 (C-2,
2
Melting points (mp) were measured on a Stuart Scientific SMP3
melting point apparatus and were not corrected. Analytical thin-
layer chromatography (TLC) was performed on SDS silica gel
60F254 precoated plates with detection by UV light and/or visuali-
zation with vanillin in sulfuric acid. Column chromatography was
1JCeF ¼ 193 Hz); 80.15 (C-1, 2JCeF ¼ 24.1 Hz); 74.87 (C-5); 72.68 (C-3,
2JCeF ¼ 19.0 Hz); 67.49 (C-4, 3JCeF ¼ 7.2 Hz); 61.58 (C-6); 20.68, 20.58,
20.52 (3ꢃ CH3).
5.1.2. 1-(2-Fluoro-2-deoxy-b-D-glucopyranosyl)-5-fluorouracil, 2b
performed with SDS silica gel 60 (Chromagel, 35e70 mm) using the
To a solution of acetylated compound 2a (218 mg, 0.52 mmol) in
dry methanol (MeOH) (5 mL) was added sodium methoxide
(MeONa) (46 mg, 0.86 mmol) under argon and the mixture was
stirred for 16 h at rt. After neutralization with IRC 50 Amberlite ion-
indicated solvent mixture expressed as volume/volume ratios.
Mass spectra were recorded on a TSQ7000 spectrometer (Thermo
Finnigan). Electrospray ionization mass spectrometry (ESI-MS) was
used in positive or negative mode.

2874
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
Fig. 6. Degradation profiles of compounds (A) 6a, (B) 6b and (C) 6c in plasma (-),
MEM (:) and PBS (B).
evaporated to dryness. The residue was dissolved twice in cold
dichloromethane (DCM) (15 mL), to allow precipitation of residual
dicyclohexylurea, filtered again and evaporated to dryness. The
residue was purified by silica gel chromatography (DCM e MeOH,
96:4). The collected fractions were pooled, washed with a saturated
NaHCO3 solution and water, dried over MgSO4 and evaporated
under reduced pressure to give 3 (196 mg, 0.41 mmol) as a white
solid; mp 129 ꢀC; yield 33%.
Fig. 5. Typical HPLC-UV chromatograms from the stability experiments in MEM at
37 ꢀC of compounds (A) 6a at 1, 5 and 10 min, (B) 6b at 1, 15 and 30 min and (C) 6c at 1,
20 and 45 min.
exchange resin (Hþ), filtration and evaporation to dryness,
compound 2b (152 mg, 0.52 mmol) was obtained as a white solid.
mp 172 ꢀC; quantitative yield.
(þ)-ESI-MS: 478.3 [M þ H]þ, 500.3 [M þ Na]þ.1H NMR (acetone-
(ꢁ)-ESI-MS: 292.7 [M ꢁ H]ꢁ.
d6):
d
10.51 (se, 1H, 5-FU:NH-3); 8.42 (d, 1H, J ¼ 9.0 Hz, NHCOCH2);
1H NMR (DMSO-d6):
d
7.98 (d, 1H, J ¼ 7.0 Hz, 5-FU:H-6); 5.64 (d,
7.82 (d, 1H, J ¼ 6.5 Hz, 5-FU:H-6); 5.43 (m, 2H, H-1, H-3); 4.91 (t, 1H,
1H, J ¼ 8.9 Hz, H-1); 4.48 (td,1H, J ¼ 51.3 Hz, 8.9 Hz, H-2); 3.68e3.20
J ¼ 9.7 Hz, H-4); 4.48 (s, 2H, CH2); 4.36 (td, 1H, J ¼ 50.6 Hz,
(m, 5H, H-3, H-4, H-6a, H-6b, H-5).
13C NMR (DMSO-d6):
d
160.82 (5-FU:C-4, 2JCeF ¼ 21.7 Hz); 152.64
1
(5-FU:C-2); 141.44 (5-FU:C-5, JCeF ¼ 234 Hz); 124.35 (5-FU:C-6,
Table 2
2JCeF ¼ 34.5 Hz); 90.15 (C-2, JCeF
¼
183 Hz); 80.52 (C-1,
1
Degradation half-lives of 5-FU derivatives in different media, estimated according to
the first-order kinetic equation at 37 ꢀC (Mean Values ꢂ SEM, n ¼ 3).
2JCeF ¼ 22.4 Hz); 80.29 (C-5); 75.13 (C-3, 2JCeF ¼ 15.4 Hz); 69.73 (C-
3
4, JCeF ¼ 7.9 Hz); 61.07 (C-6).
Compound
Medium
Half-life (min)
6a
PBS
3 ꢂ 1
2
6 ꢂ 5
8 ꢂ 1
6
10 ꢂ 3
11 ꢂ 2
8
17 ꢂ 3
5.1.3. N-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
b
-D-glucopyranosyl)-
MEMa
Plasma
PBS
2-(5-fluorouracil-1-yl)acetamide, 3
6b
6c
Aminosugar 1b (382 mg,1.24 mmol) was dissolved in anhydrous
dimethylformamide (DMF) (20 mL) and (5-fluorouracil-1-yl)acetic
acid (5-FUA) (281 mg, 1.49 mmol), dicyclohexylcarbodiimide (DCC)
(307 mg, 1.49 mmol), and hydroxybenzotriazole (HOBt) (201 mg,
1.49 mmol) were added. The solution was stirred for 48 h at rt., the
dicyclohexylurea was filtered off, and the resulting solution was
MEMa
Plasma
PBS
MEMa
Plasma
a
n ¼ 1.

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2875
1
Table 3
5-FU:C-2); 141.75 (5-FU:C-5, JCeF ¼ 234 Hz); 123.17 (5-FU:C-6,
1
2JCeF
¼
37.8 Hz); 89.39 (C-2, JCeF
¼
187 Hz); 77.82 (C-1,
Antiproliferative activities of 5-FU and compounds 6a, 6b and 6c (IC50 in
Values ꢂ SEM, n ¼ 3).
mM; Mean
2JCeF ¼ 22.8 Hz); 73.88 (C-3, 2JCeF ¼ 18.6 Hz); 73.70 (C5); 68.80 (C-4,
3JCeF ¼ 7.3 Hz); 62.41 (C-6); 40.89 (CH2NHCO); 33.43 (CH2CONH);
28.46 (CH2CH2CH2); 20.37 (3ꢃ CH3).
Compounds
5-FU
6a
6b
6c
Fibroblast
MCF 7
PA 1
11 ꢂ 7
15 ꢂ 4
5 ꢂ 1
6.4 ꢂ 0.7
11 ꢂ 2
3 ꢂ 2
15 ꢂ 6
23 ꢂ 4
6 ꢂ 2
5.0 ꢂ 0.1
8.1 ꢂ 0.8
4 ꢂ 1
5.1.6. 6-Azido-N-(3,4,6-tri-O-acetyl-2-deoxy-2-fluoro-
glucopyranosyl)hexanamide, 4
b-D-
Compound 4 was prepared according to the procedure described
for compound 3, starting from 6-azidohexanoic acid (1.00 g,
6.36 mmol), DCC (1.57 g, 7.63 mmol), HOBt (1.03 g, 7.63 mmol), and
aminosugar 1b (2.34 g, 4.24 mmol) in DMF (45 mL). The mixture was
stirred for 18 h before work up. Compound 4 (913 mg, 2.05 mmol)
was obtained as a white powder. mp 84 ꢀC; yield 48%.
J ¼ 9.7 Hz, H-2); 4.17 (dd, 1H, J ¼ 12.6 Hz, 4.6 Hz, H-6a); 3.98 (m, 2H,
H-6b, H-5); 1.98, 1.97, 1.93 (3s, 9H, 3ꢃ CH3).
13C NMR (acetone-d6):
d
170.66, 170.14, 170.08 (3ꢃ OCOCH3);
2
168.28 (NHCOCH2); 158.04 (5-FU:C-4, JCeF ¼ 25.9 Hz); 150.55 (5-
(ꢁ)-ESI-MS: 444.9 [M ꢁ H]ꢁ.
1
FU:C-2); 140.92 (5-FU:C-5, JCeF ¼ 229 Hz); 131.04 (5-FU:C-6,
1H NMR (CDCl3):
d
6.99 (d,1H, J ¼ 9.2 Hz, NH); 5.46e5.31 (m, 2H,
1
2JCeF
¼
33.8 Hz); 89.52 (C-2, JCeF
¼
188 Hz); 78.16 (C-1,
H-1, H-3); 5.01 (t,1H, J ¼ 9.5 Hz, H-4); 4.39 (td,1H, J ¼ 50.2 Hz, 9.5 Hz,
H-2); 4.37 (dd,1H, J ¼ 12.6 Hz, 4.2 Hz, H-6a); 4.05 (dd,1H, J ¼ 12.6 Hz,
1.8 Hz, H-6b); 3.88 (ddd, 1H, J ¼ 12.6 Hz, 4.2 Hz, 1.8 Hz, H-5); 3.28 (t,
2H, J ¼ 7.6 Hz, CH2N3); 2.29 (t, 2H, J ¼ 7.2 Hz, CH2CONH); 2.08, 2.06,
2.05 (3s, 9H, 3ꢃ CH3); 1.73e1.62 (m, 4H, CH2CH2CH2CH2N3);
1.59e1.27 (m, 2H, CH2CH2N3).
2JCeF ¼ 23.0 Hz); 74.04 (C-5); 73.94 (C-3, 2JCeF ¼ 19.3 Hz); 68.85 (C-
3
4, JCeF ¼ 7.6 Hz); 62.54 (C-6); 50.51 (CH2); 20.58 (3ꢃ CH3).
5.1.4. 1-{N-[2-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
glucopyranosyl)amino-2-oxoethyl]carbamoyl}-5-fluorouracil, 6a
To an ice-cooled solution of 5-FU (89 mg, 0.69 mmol) in anhy-
drous pyridine (5 mL) was added triphosgene (204 mg, 0.69 mmol)
under argon. The solution was stirred for 1 h at 0 ꢀC, the amine 5a
(250 mg, 0.69 mmol) slowly added and the mixture was allowed to
reach rt. The mixture was then stirred for 15 h and evaporated to
dryness. The residue was taken up in DCM (4 mL) and the pyridine
hydrochloride was filtered off. The filtrate was washed twice with
a 6 N aqueous solution of hydrochloric acid (8 mL), dried over MgSO4
and concentrated under vacuum. The resulting residue was purified
by silica gel chromatography (cyclohexane e EtOAc, 2:8) to yield 6a
(210 mg, 0.40 mmol) as a white powder; mp 170 ꢀC; yield 59%.
(ꢁ)ESI-MS: 518.9 [M ꢁ H]ꢁ.
IR (KBr) cmꢁ1
: n 3342, 2099, 1746, 1681, 1228, 1034.
5.1.7. 3-Amino-N-(3,4,6-tri-O-acetyl-2-deoxy-2-fluoro-
b-D-
glucopyranosyl)hexanamide, 5c
To a suspension of 10% Pd/C (130 mg) in DCM/MeOH (60/30 mL)
was added the azido compound 4 (657 mg,1.47 mmol). The mixture
was hydrogenated at atmospheric pressure for 20 h, filtered
through celite, and concentrated to yield a residue, which was
taken up in diisopropyl ether (10 mL). Filtration of the precipitate
gave the amine 5c (505 mg, 1.20 mmol) as a yellow hygroscopic
powder; mp 163 ꢀC; yield 82%.
(þ)-ESI-MS: 421.0 [M þ H]þ.
1H NMR (DMSO-d6):
d
12.35 (s, 1H, 5-FU:NH-3); 9.49 (t, 1H,
1H NMR (CDCl3):
d
7.98 (d, 1H, J ¼ 8.9 Hz, NH); 5.42e5.29 (m, 2H,
J ¼ 5.4 Hz, CH2CONH); 9.05 (d, 1H, J ¼ 9.1 Hz, NHCON); 8.41 (d, 1H,
J ¼ 7.5 Hz, 5-FU:H-6); 5.68e5.46 (m, 2H, H-1, H-3); 4.87 (t, 1H,
J ¼ 9.2 Hz, H-4); 4.43 (td, 1H, J ¼ 50.5 Hz, J ¼ 9.2 Hz, H-2); 4.24e3.93
(m, 5H, H-6a, H-6b, H-5, CH2CONH); 2.03, 2.00,1.99 (3s, 9H, 3ꢃ CH3).
H-1, H-3); 5.01 (t, 1H, J ¼ 9.8 Hz, H-4); 4.52 (td, 1H, J ¼ 49.9 Hz,
9.8 Hz, H-2); 4.30 (dd, 1H, J ¼ 12.9 Hz, 4.8 Hz, H-6a); 4.02 (d, 1H,
J ¼ 12.9 Hz, H-6b); 3.84 (m, 1H, H-5); 2.94 (m, 2H, CH2NH2); 2.30
(m, 2H, CH2CONH); 2.06, 2.05, 2.02 (3s, 9H, 3ꢃ CH3); 1.87e1.23 (3m,
8H, CH2CH2CH2CH2NH2).
13C NMR (CF3CO2D):
d 170.70, 169.74, 169.35, 166.61 (CH2CONH,
2
3ꢃ OCOCH3); 150.77 (5-FU:C-4, JCeF ¼ 26.8 Hz); 145.61, 145.01
(NHCON, 5-FU:C-2); 135.69 (5-FU:C-5, 1JCeF ¼ 236 Hz); 118.65 (5-
5.1.8. 1-{N-[6-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
2
1
FU:C-6, JCeF ¼ 38.2 Hz); 82.27 (C-2, JCeF ¼ 191 Hz); 72.40 (C-1,
2JCeF ¼ 23.9 Hz); 68.89 (C-3, 2JCeF ¼ 19.8 Hz); 68.16 (C-5); 63.14 (C-
4); 57.02 (C-6); 38.53 (CH2NHCO); 13.65 (3ꢃ CH3).
glucopyranosyl)amino-6-oxohexyl]carbamoyl}-5-fluorouracil, 6c
Compound 6c was prepared according to the procedure
described for compound 6a, starting from 5-FU (51 mg, 0.39 mmol),
in anhydrous pyridine (3 mL), triphosgene (115 mg, 0.39 mmol) and
amine 5c (164 mg, 0.39 mmol). The mixture was stirred for 15 h
before work-up. Purification conditions similar to those used for
compound 6a lead to compound 6c (109 mg, 0.19 mmol), which was
obtained as a white powder; mp 134 ꢀC; yield 48%.
5.1.5. 1-{N-[4-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
glucopyranosyl)amino-4-oxobutyl]carbamoyl}-5-fluorouracil, 6b
Compound 6b was prepared according to the procedure
described for compound 6a, starting from 5-FU (83 mg, 0.64 mmol),
in anhydrous pyridine (4 mL), triphosgene (189 mg, 0.64 mmol) and
amine 5b (250 mg, 0.64 mmol). The mixture was stirred for 17 h
before work-up. Purification conditions similar to those used for
compound 6a lead to compound 6b (86 mg, 0.16 mmol), which was
obtained as a white powder; mp 90 ꢀC; yield 18%.
(ꢁ)-ESI-MS: 575.1 [M ꢁ H]ꢁ.
1H NMR (DMSO-d6):
d 12.27 (brs, 1H, 5-FU:NH-3); 9.12 (t, 1H,
J ¼ 5.6 Hz, NHCON); 8.78 (d, 1H, J ¼ 8.9 Hz, CH2CONH); 8.38 (d, 1H,
J ¼ 7.5 Hz, 5-FU:H-6); 5.75e5.39 (m, 2H, H-1, H-3); 4.85 (t, 1H,
J ¼ 9.0 Hz, H-4); 4.37 (td, 1H, J ¼ 50.3 Hz, J ¼ 9.0 Hz, H-2); 4.19e3.92
(m, 3H, H-5, H-6a, H-6b); 3.27 (m, 2H, CH2NH); 2.19e1.98 (m, 11H,
CH2CO, 3ꢃ CH3); 1.52 (m, 4H, 2ꢃ CH2); 1.27 (m, 2H, CH2).
(ꢁ)-ESI-MS: 546.9 [M ꢁ H]ꢁ.
1H NMR (acetone-d6):
d 11.00 (brs,1H, 5-FU:NH-3); 9.27 (brs,1H,
NHCON); 8.49 (d, 1H, J ¼ 7.4 Hz, 5-FU:H-6); 8.17 (d, 1H, J ¼ 9.3 Hz,
CH2CONH); 5.60e5.43 (m, 2H, H-1, H-3); 5.01 (t, 1H, J ¼ 9.4 Hz, H-
4); 4.45 (td, 1H, J ¼ 50.5 Hz, J ¼ 9.4 Hz, H-2); 4.28 (dd, 1H,
J ¼ 12.8 Hz, 5.0 Hz, H-6a); 4.13e4.05 (m, 2H, H-5, H6b); 3.48 (q, 2H,
J ¼ 6.8 Hz, CH2NH); 2.40 (t, 2H, J ¼ 7.2 Hz, NHCOCH2); 2.12e1.92 (m,
11H, CH2CH2CH2, 3ꢃ CH3).
13CNMR(DMSO-d6):
d
173.40,170.57,170.11,169.93(CH2CONH, 3ꢃ
OCOCH3); 157.58 (5-FU:C-4, 2JCeF ¼ 26.9 Hz); 150.79,150.06 (NHCON,
1
5-FU:C-2); 141.21 (5-FU:C-5, JCeF ¼ 232 Hz); 123.43 (5-FU:C-6,
2JCeF
¼
37.3 Hz); 89.04 (C-2, JCeF
¼
186 Hz); 77.00 (C-1,
1
2JCeF ¼ 22.5 Hz); 73.34 (C-3, JCeF ¼ 19.3 Hz); 72.46 (C-5); 68.32
(C-4); 62.18 (C-6); 35.87 (CH2CONH); 29.05, 26.34, 25.00
(CH2CH2CH2CH2NHCO); 21.02 (3ꢃ CH3). Signal of CH2NHCO in the
solvent signal.
2
13C NMR (acetone-d6):
d
172.91, 170.43, 169.88 (CH2CONH, 3ꢃ
OCOCH3); 157.10 (5-FU:C-4, 2JCeF ¼ 27.7 Hz); 151.17,150.50 (NHCON,

2876
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
5.1.9. N-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
b
-
D
-glucopyranosyl)-
under vacuum to give the carbamoyl chloride of amine 13 (200 mg,
0.83 mmol) which was used without further purification.
2-(N0-benzyloxycarbonyl-N0-methyl)aminoethanamide, 8a
Compound 8a was prepared according to the procedure
described for compound 3, starting from aminosugar 1b (826 mg,
2.69 mmol), DCC (666 mg, 3.23 mmol) and HOBt (436 mg,
3.23 mmol) in anhydrous DCM (10 mL). The mixture was stirred for
24 h before work-up. The crude product was purified by silica gel
chromatography (cyclohexane e EtOAc, 4:6) to give compound 8a
(580 mg, 1.13 mmol) as a white powder; mp 116 ꢀC; yield 42%.
(ꢁ)-ESI-MS: 511.07 [M ꢁ H]ꢁ.
To a solution of 5-FU (108 mg, 0.83 mmol) in freshly distilled N,N-
dimethylacetamide (DMAc) (3.3 mL) was slowly added a 60%
dispersion of sodium hydride (NaH) in mineral oil (32 mg,
0.83 mmol) under argon. The mixture was heated at 30 ꢀC and the
carbamoyl chloride of amine 13 in DMAc (1 mL) was added dropwise.
The solution was stirred at 30 ꢀC for 15 h before evaporation. The
residue was taken up in DCM (4 mL), washed once with water, dried
over MgSO4 and concentrated under vacuum. The resulting residue
waspurified bysilica gel chromatography(cyclohexane e EtOAc, 2:8)
to give compound 14a (103 mg, 0.31 mmol) as a white powder; mp
140 ꢀC; overall yield 37%.
1H NMR (CDCl3):
d
7.34 (m, 5H, Harom.); 7.12 (d, 1H, J ¼ 8.9 Hz,
NH); 5.44e5.28 (m, 2H, H-1, H-3); 5.15 (s, 2H, NCOOCH2); 5.02 (t,
1H, J ¼ 9.4 Hz, H-4); 4.29 (dd, 1H, J ¼ 12.6 Hz, 4.4 Hz, H-6a); 4.27 (td,
1H, J ¼ 50.5 Hz, 9.4 Hz, H-2); 4.07e4.01 (m, 3H, CH2CONH, H-6b);
3.88 (dd, 1H, J ¼ 12.6 Hz, 4.4 Hz, H-5); 3.00 (s, 3H, NCH3); 2.08 (s,
3H, CH3); 2.04 (s, 6H, 2ꢃ CH3).
1H NMR (DMSO-d6):
d 11.00 (se, 1H, 5-FU:NH-3); 8.00, 7.89 (2d,
1H, J ¼ 6.4 Hz, 5-FU:H-6); 7.38 (m, 5H, Harom.); 5.20, 5.16 (2s, 2H,
COOCH2); 4.33, 4.29 (2s, 2H, CH2N); 3.03 (s, 3H, CH3N).
5.1.14. N-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
5.1.10. N-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
glucopyranosyl)-4-methylaminoethanamide, 9a
b-D-
glucopyranosyl)-4-(N0-benzyloxycarbonyl-N0-methyl)
aminobutanamide, 8b
N-Benzyloxycarbonyl (N-Cbz) protected derivative 8a (537 mg,
1.05 mmol) was added, with stirring, to a suspension of 10% Pd/C
(145 mg) in THF/MeOH (5/5 mL). The mixture was stirred at rt for
5 h, filtered through celite, and evaporated under reduced pressure
to give the expected compound 9a (330 mg, 0.87 mmol) as a gray
solid which was used without further purification; yield 83%.
(ꢁ)-ESI-MS: 378.97 [M ꢁ H]ꢁ.
A stirred solution of 7b (700 mg, 2.95 mmol) in DCM (20 mL) was
treated with TEA (0.50 mL, 3.54 mmol). The mixture was cooled to
0 ꢀC, treated with ClCO2Et (0.30 mL, 3.11 mmol), and stirred at rt for
30 min. Aminosugar 1b (824 mg, 2.68 mmol) in DCM (30 mL) was
then added, and the mixture was stirred for 20 h at rt. The mixture
was suspended in saturated aqueous Na2CO3 solution (60 mL) and
extracted three times with DCM (40 mL). The organic extracts were
combined, dried over MgSO4, and concentrated under reduced
pressure. The residue was purified by silica gel chromatography
(cyclohexane e EtOAc, 4:6), to give N-Cbz protected compound 8b
(326 mg, 0.60 mmol) as a white solid. mp 62 ꢀC; yield 22%.
(ꢁ)-ESI-MS: 539.0 [M ꢁ H]ꢁ.
1H NMR (CDCl3):
d
8.08 (d, 1H, J ¼ 9.8 Hz, NHCO); 5.46e5.30 (m,
2H, H-1, H-3); 5.07 (t, 1H, J ¼ 9.6 Hz, H-4); 4.36 (td, 1H, J ¼ 50.6 Hz,
9.6 Hz, H-2); 4.30 (dd, 1H, J ¼ 12.4 Hz, 4.5 Hz, H-6a); 4.08 (dd, 1H,
J ¼ 12.4 Hz, 2.1 Hz, H-6b); 3.86 (ddd, 1H, J ¼ 12.4 Hz, 4.5 Hz, 2.1 Hz,
H-5); 3.31 (d, 2H, J ¼ 4.4 Hz, CH2); 2.44 (s, 3H, NHCH3); 2.08, 2.07,
2.04 (3s, 9H, 3ꢃ CH3).
1H NMR (CDCl3):
d
8.16 (d,1H, J ¼ 8.7 Hz, NHCOCH2); 7.36 (m, 5H,
Harom.); 5.46e5.30 (m, 2H, H-1, H-3); 5.15 (s, 2H, NCOOCH2); 5.07 (t,
1H, J ¼ 9.5 Hz, H-4); 4.44 (td, 1H, J ¼ 50.3 Hz, 9.5 Hz, H-2); 4.30 (dd,
1H, J ¼ 12.5 Hz, 4.4 Hz, H-6a); 4.08 (dd,1H, J ¼ 12.5 Hz,1.7 Hz, H-6b);
3.85(m,1H, J ¼ 12.5 Hz, 4.4 Hz,1.7 Hz, H-5);3.58e3.16 (m, 2H, CH2N);
2.91 (s, 3H, NCH3); 2.20 (m, 2H, CH2CONH); 2.08, 2.06, 2.04 (3s, 9H,
3ꢃ CH3); 1.87 (m, 2H, CH2CH2CH2).
5.1.11. Benzyl 2-((tert-butoxycarbonyl)methylamino)acetate, 12
To a solution of 11 (1.44 g, 7.6 mmol) in CH3CN were added 1,8-
diaza-bicyclo[5.4.0]undec-7-ene (DBU) (1.1 mL, 7.6 mmol) and
benzyl bromide (BnBr) (0.9 mL, 7.30 mmol). After 5 h at 60 ꢀC, the
reaction mixture was allowed to cool to rt and evaporated to
dryness. The residue was diluted with EtOAc (60 mL) and washed
successively with a saturated solution of sodium bicarbonate
(NaHCO3) (50 mL), an aqueous potassium hydrogen sulfate solution
(1 N) (50 mL), and brine (40 mL). The organic layer was dried over
MgSO4 and concentrated under reduced pressure to give
compound 12 (1.73 g, 6.2 mmol) as a yellow oil; yield 81%.
5.1.15. N-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
b-D-
glucopyranosyl)-4-methylaminobutanamide, 9b
Compound 9b was prepared according to the procedure
described for compound 9a, starting from compound 8b (295 mg,
0.55 mmol) and 10% Pd/C (60 mg) in THF/MeOH (20/20 mL). The
mixture was stirred for 30 min before work-up. The crude product
was purified by silica gel chromatography (DCM e MeOH, 9:1) to
give compound 9b (200 mg, 0.49 mmol) as a yellow gum; yield 90%.
(þ)ESI-MS: 406.0 [M þ H]þ.
1H NMR (CDCl3):
d 7.34 (m, 5H, Harom.); 5.17, 5.15 (2s, 2H,
COOCH2); 4.01, 3.92 (2s, 2H, CH2N); 2.92, 2.91 (2s, 3H, NCH3); 1.46,
1.42, 1.37 (3s, 9H, 3ꢃ CH3).
1H NMR (CDCl3):
d 5.45e5.29 (m, 2H, H-1, H-3); 5.04 (t, 1H,
5.1.12. Benzyl 2-(methylamino)acetate, trifluoroacetic acid salt, 13
To a solution of 12 (2.00 g, 7.16 mmol) in THF (8 mL) was added
trifluoroacetic acid (TFA) (8 mL). The reaction mixture was stirred at
rt for 20 min, and evaporated to dryness to give compound 13
(755 mg, 2.72 mmol) as a yellow oil, which was used without
further purification; yield 38%.
J ¼ 9.6 Hz, H-4); 4.34 (td, 1H, J ¼ 50.5 Hz, 9.6 Hz, H-2); 4.31 (dd, 1H,
J ¼ 12.5 Hz, 4.2 Hz, H-6a); 4.07 (dd, 1H, J ¼ 12.5 Hz, 1.8 Hz, H-6b);
3.86 (m, 1H, H-5); 2.69 (t, 2H, J ¼ 6.2 Hz, CH2NH); 2.43e2.38 (m, 5H,
CH3, CH2CONH); 2.08, 2.07, 2.04 (3s, 9H, 3ꢃ CH3); 1.85 (m, 2H,
CH2CH2CH2).
1H NMR (CDCl3):
d 7.36 (s, 5H, Harom.); 5.17 (s, 2H, COOCH2); 3.41
(s, 2H, CH2N); 2.44 (s, 3H, CH3N).
5.1.16. 1-{N-[4-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
glucopyranosyl)amino-4-oxobutyl]-N0-methylcarbamoyl}-5-
fluorouracil, 10b
5.1.13. 1-[N-(2-Benzyloxy-2-oxoethyl)-N-methylcarbamoyl]-5-
fluorouracil, 14a
Compound 10b was prepared in two steps according to the
procedure described for compound 14a.
First, the carbamoyl chloride was formed starting from NaHCO3
(84 mg, 1.00 mmol), triphosgene (98 mg, 0.33 mmol) in dry DCM
(4 mL) and a solution of amine 9b (204 mg, 0.50 mmol) in DCM
(1.5 mL).
A solution of salt 13 (150 mg, 0.84 mmol) in DCM (1.5 mL) is added
dropwise to a slurry of NaHCO3 (141 mg,1.68 mmol) and triphosgene
(164mg, 0.55mmol)indryDCM(2mL)at10e15 ꢀCover10minunder
argon. The reaction mixturewasstirred atrtfor 1 h. The reaction mass
was then filtered off to remove NaCl and the filtrate was concentrated

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2877
The reaction mixture was stirred at rt for 20 h before work-up.
The resulting carbamoyl chloride of amine 9b (145 mg,
0.32 mmol) was a light yellow oil, used without further purification.
To a solution of 5-FU (40 mg, 0.31 mmol) in DMAc (2 mL) was
slowly added a 60% dispersion of NaH in mineral oil (12 mg,
0.31 mmol) under argon. The mixture was heated at 45 ꢀC and the
carbamoyl chloride of amine 9b (145 mg, 0.31 mmol) in DMAc
(1 mL) was added dropwise. The solution was stirred at 45 ꢀC for 8 h
and left to stand 12 h at rt before evaporation. The residue was
taken up in DCM (4 mL), washed once with water (6 mL), dried over
MgSO4 and concentrated under vacuum. The resulting residue was
purified by silica gel chromatography (DCM e MeOH, 95:5) to yield
10b (95 mg, 0.17 mmol) as a white solid; mp 81 ꢀC; overall yield
35%.
sodium thiosulphate (30 mL), and again water (30 mL), dried over
MgSO4, filtered and evaporated to dryness. The resulting residue
was purified by chromatography on alumina (cyclohexane e EtOAc,
6:4) to yield 17 (263 mg, 0.48 mmol) as a white gum; yield 45%.
1H NMR (CDCl3):
d
7.33 (m, 5H, Harom. Bn); nA ¼ 7.20, nB ¼ 7.05
(AA0BB0, JAB ¼ 8.8 Hz, 4Harom.); 7.05 (d, 2H, J ¼ 8.7 Hz, Harom.); 5.44
(td, 1H, J ¼ 14.6 Hz, 9.2 Hz, H-3); 5.17e5,02 (m, 4H, H-1, H-4,
COOCH2); 4,59 (td, 1H, J ¼ 50.4 Hz, 9.2 Hz, H-2); 4.32 (dd, 1H,
J ¼ 12.4 Hz, 5.3 Hz, H-6a); 4.17 (dd, 1H, J ¼ 12.4 Hz, 2.3 Hz, H-6b);
3.89 (ddd, 1H, J ¼ 12.4 Hz, 5.3 Hz, 2.3 Hz, H-5); 3.30 (s, 3H, CH3N);
2.13, 2.08, 2.07 (3s, 9H, 3ꢃ CH3).
5.1.19. 3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-1-(4-methylamino)
phenyl-b-D-glucopyranoside, 18
(þ)ESI-MS: 563.2 [M þ H]þ.
Compound 18 was prepared according to the procedure
described for compound 9a, starting from compound 17 (258 mg,
0.47 mmol) and 10% Pd/C (52 mg) in THF/MeOH (10/10 mL). The
mixture was stirred for 8 h before work-up. The crude product was
purified by silica gel chromatography (cyclohexane e EtOAc, 6:4) to
give compound 18 (112 mg, 0.27 mmol) as a yellow gum; yield 58%.
1H NMR (DMSO-d6):
d
8.86, 8.83 (2d, 1H, J ¼ 8.9 Hz, NHCO); 8.13,
8.08 (2d, 1H, J ¼ 6.3 Hz, 5-FU:H-6); 5.60e5.42 (m, 2H, H-1, H-3);
4.85 (t, 1H, J ¼ 9.7 Hz, H-4); 4.38 (td, 1H, J ¼ 50.5 Hz, 9.7 Hz, H-2);
4.18e3.92 (m, 3H, H-6a, H-6b, H-5); 2.96, 2.90 (2s, 3H, NCH3);
2.25e2.13 (m, 2H, CH2CONH); 2.08, 2.03, 1.98 (3s, 9H, 3ꢃ CH3);
1.96e1.77 (m, 2H, CH2CH2CH2); signal of CH2N in H2O signal.
1H NMR (CDCl3):
d
6.96 (d, 2H, J ¼ 8.6 Hz, Harom.); 6.58 (d, 2H,
13C NMR (CDCl3):
d
172.97 (NHCOCH2); 170.75, 170.07, 169.82
J ¼ 8.6 Hz, Harom.); 5.38 (td, 1H, J ¼ 14.7 Hz, 9.2 Hz, H-3); 5.08 (t,
1H, J ¼ 14.7 Hz, H-4); 4.96 (dd, 1H, J ¼ 7.6 Hz, 2.8 Hz, H-1); 4.52 (td,
1H, J ¼ 50.4 Hz, 9.2 Hz, H-2); 4.29 (dd, 1H, J ¼ 12.5 Hz, 5.2 Hz, H-6a);
4.13 (dd, 1H, J ¼ 12.5 Hz, 2.3 Hz, H-6b); 3.77 (ddd, 1H, J ¼ 12.5 Hz,
5.2 Hz, 2.3 Hz, H-5); 2.82 (s, 3H, CH3N); 2.10, 2.08, 2.04 (3s, 9H, 3ꢃ
CH3).
2
(3ꢃ OCOCH3); 156.99 (5-FU:C-4, JCeF ¼ 26.5 Hz); 150.81 (5-FU:C-
1
2); 148.19 (NCON); 141.11 (5-FU:C-5, JCeF ¼ 222 Hz); 125.53 (5-
2
1
FU:C-6, JCeF ¼ 34.4 Hz); 88.27 (C-2, JCeF ¼ 189 Hz); 77.16 (C-1,
2JCeF ¼ 24.1 Hz); 73.72 (C-5); 73.61 (C-3, 2JCeF ¼ 19.2 Hz); 68.14 (C-4,
3JCeF ¼ 7.3 Hz); 61.91 (C-6); 49.65 (CH2NCO); 37.52 (NCH3); 32.38
(NHCOCH2); 21.79 (CH2CH2CH2); 20.82, 20.78, 20.71 (3ꢃ OCOCH3).
5.1.20. 1-[N-4-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
5.1.17. [4-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-
glucopyranosyloxy)phenyl]benzyl carbamate, 16
b
-
D
-
glucopyranosyloxy)phenyl-N-methyl]carbamoyl-5-fluorouracil, 19
Compound 19 was prepared in two steps according to the
procedure described for compound 14a. First, the carbamoyl chlo-
ride was formed starting from 18 (100 mg, 0.24 mmol), NaHCO3
(41 mg, 0.48 mmol), triphosgene (108 mg, 0.37 mmol), and DCM
(4 mL). The mixture was stirred for 14 h before work-up to give the
carbamoyl chloride of arylamine 18 (91 mg, 0.19 mmol) used
without further purification. To a solution of 5-FU (37 mg,
0.29 mmol) in freshly distilled DMAc (2 mL) was slowly added a 60%
dispersion of NaH in mineral oil (12 mg, 0.29 mmol) under argon.
The mixture was heated at 45 ꢀC for 12 h and the carbamoyl
chloride of amine 18 in DMAc (1 mL) was added dropwise. The
mixture was stirred for 12 h before work-up to give a residue, which
was purified by silica gel chromatography (cyclohexane e EtOAc,
4:6) to give an oil. The latter was triturated in diisopropyl ether
(9 mL). Filtration of the precipitate gave compound 19 (39 mg,
0.068 mmol) as a beige powder; mp 84 ꢀC; overall yield 29%.
(ꢁ)-ESI-MS: 568.2 [M ꢁ H]ꢁ.
To a solution of 15 (713 mg, 1.79 mmol) in a mixture of water
(3 mL) and acetone (6 mL) was added NaHCO3 (316 mg, 3.76 mmol).
The solution was cooled to 0 ꢀC, treated with benzyl chloroformate
(ClCO2Bn) (1.84 mmol, 0.26 mL) dropwise, and stirred at rt for 14 h.
The reaction mixture was then poured onto ice (50 g) and the
precipitate formed was filtered off, washed with water (50 mL) and
dried. The resulting solid was purified by silica gel chromatography
(cyclohexane e EtOAc, 6:4) to give an oil, which was taken up in
diethyl ether. Filtration of the precipitate gave the amine 16
(586 mg, 1.10 mmol) as a white fluffy powder. mp 131 ꢀC; yield 61%.
(ꢁ)-ESI-MS: 531.0 [M ꢁ H]ꢁ.
1H NMR (CDCl3):
d
7.38 (m, 5H, Harom. Bn); nA ¼ 7.32, nB ¼ 7.02
(AA0BB0, JAB ¼ 8.9 Hz, 4Harom.); 6.68 (brs, 1H, NH); 5.39 (td, 1H,
J ¼ 14.4 Hz, 9.2 Hz, H-3); 5.19 (s, 2H, NHCOOCH2); 5.14e5.03 (m, 2H,
H-1, H-4); 4.55 (td, 1H, J ¼ 50.3 Hz, 9.2 Hz, H-2); 4.29 (dd, 1H,
J ¼ 12.3 Hz, 5.2 Hz, H-6a); 4.15 (dd, 1H, J ¼ 12.3 Hz, 2.1 Hz, H-6b);
3.84 (m, 1H, H-5); 2.11, 2.08, 2.05 (3s, 9H, 3ꢃ CH3).
1H NMR (CDCl3):
d
8.29 (se, 1H, NH); 7.47 (d, 1H, J ¼ 5.0 Hz, 5-
FU:H-6); 7.08 (m, 4H, Harom.); 5.43 (td, 1H, J ¼ 14.5 Hz, 8.8 Hz, H-3);
5.12 (m, 2H, H-1, H-4); 4.58 (td, 1H, J ¼ 50.6 Hz, 8.8 Hz, H-2);
4.34e4.11 (m, 2H, H-6a, H-6b); 3.90 (m, 1H, H-5); 3.43 (s, 3H,
CH3N); 2.12, 2.08, 2.07 (3s, 9H, 3ꢃ CH3).
5.1.18. N-[4-(3,4,6-Tri-O-acetyl-2-fluoro-2-deoxy-b-D-
glucopyranosyloxy)phenyl]-N-methyl-benzylcarbamate, 17
Compound 16 (570 mg, 1.07 mmol) and methyl iodide (MeI)
(67
m
L, 1.07 mmol) were dissolved in THF (8 mL) and the solution
13C NMR (CDCl3):
d
170.65, 170.10, 169.62 (3ꢃ OCOCH3); 156.69
2
was cooled to 0 ꢀC in a flask protected from moisture. A 60% NaH
dispersion in mineral oil (60 mg, 2.50 mmol) was then added
portionwise and cautiously with gentle stirring. The suspension
was stirred at rt for 5 h. A slight excess of MeI was added to
complete the reaction. EtOAc was then added (4 mL), followed by
water (4 mL), dropwise, to destroy the excess of NaH. The solution
was evaporated to dryness, and the oily residue partitioned
between ether (40 mL) and water (40 mL). The ether layer was
washed with aqueous saturated NaHCO3 solution (30 mL), and the
combined aqueous extracts acidified to pH 2 with 5 N aqueous HCI
solution. The product was extracted into EtOAc (50 mL). The
combined extracts were washed with water (35 mL), 5% aqueous
(5-FU:C-4, JCeF ¼ 27.0 Hz); 156.22 (Carom.); 156.00 (CH3NCON);
1
150.37 (5-FU:C-2); 140.73 (5-FU:C-5, JCeF ¼ 240 Hz); 136.78
(Carom.); 127.49 (Carom.); 125.03 (5-FU:C-6, 2JCeF ¼ 34.3 Hz); 127.49
2
1
(Carom.); 98.40 (C-1, JCeF ¼ 23.6 Hz); 89.10 (C-2, JCeF ¼ 191 Hz);
72.60 (C-3, 2JCeF ¼ 19.8 Hz); 72.22 (C-5); 67.99 (C-4, 3JCeF ¼ 7.1 Hz);
61.79 (C-6); 39.83 (NCH3); 20.86, 20.75, 20.65 (3ꢃ OCOCH3).
5.2. HPLC-UV analysis
A new HPLC-UV method was developed to simultaneously
analyze 5-FU and its derivatives. An HP 1050 (Agilent, palo Alto, CA,
USA) liquid chromatograph equipped with a 150 ꢃ 3.0 mm

2878
P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
Phenomenex Luna PFP(2) reversed-phase column (Le Pecq, France),
particle size 3 m, poresize of 100 Å was used for the overallanalysis.
foreskin waste from a 6-year old Caucasian male and the cells used
in this work were from the seventh to twelfth passage of the
culture. Breast cancer adenocarcinoma MCF 7 and ovary terato-
carcinoma PA 1 human cell lines were purchased from the Euro-
pean Collection of Cell Cultures (ECACC; Salisbury, United
Kingdom). Stock cell cultures were maintained as monolayer in
75 cm2 culture flasks in GlutamaxÔ Eagle’s Minimum Essential
Medium with Earle’s salts (MEM; Invitrogen, Cergy-Pontoise,
France) supplemented with 10% fetal calf serum (Sigma, St Quen-
tin Fallavier, France), 1 mM sodium pyruvate (Invitrogen), 1X vita-
mins solution (Invitrogen), 1X non essential amino acids solution
m
The solvents used to prepare the mobile phase were 10 mM
ammonium formate buffer, adjusted to pH 3 with formic acid and
filtered through a 0.45 mm filter (Millipore, Saint-Quentin, Yvelines,
France) (component A) and acetonitrile (component B). The initial
mobile phase composition was 0% B, which was held for 6 min after
injection of sample. A linear gradient was then started in which the
proportion of B increased from 0% to 85% in 14 min. Eventually the
column was re-equilibrated at 0% B from 25 min to 35 min. The total
run time was 35 min. HPLC separations were performed at room
temperature at a flow rate of 0.4 mL/min and an injection volume of
(Invitrogen) and 4 mg/mL of gentamicine (Invitrogen). Cells were
10
mL. Detection of compounds was performed using a UV detector:
grown at 37 ꢀC in a humidified atmosphere containing 5% CO2.
at 264 nm from 0 min to 7 min and then at 258 nm.
5.6.2. Cell growth inhibition assay
5.3. Characterization of degradation products by HPLC-UV-MS
Cells were plated at a density of 5 ꢃ 103 cells per well in 96-well
microplates (NunclonÔ, Nunc, Roskilde, Denmark) in 150
mL of
The type of column and the mobile phase composition were the
same as described above. The mass spectral data were acquired by
electrospray ionization mass spectrometry (ESI-MS) on a TSQ7000
mass spectrometer from Thermo Finnigan (San José, CA, USA). The
detection was carried out by positive or negative electrospray
ionization and the source parameters were tuned as follows: spray
voltage, 4.5 kV; capillary temperature, 350 ꢀC; nitrogen sheath gas,
80 Psi; nitrogen auxiliary gas, 40 AU; argon collision pressure,
2.5 mTorr. The data acquisition was performed with the XcaliburÒ
1.1 software package from Thermo Finnigan. The mass range
scanned in MS runs was m/z 50e800. Each MS spectrum was
recorded by averaging 20 spectra.
culture medium and were allowed to adhere for 16 h before
treatment with the compound tested. Stock solution of each
compound was prepared in dimethylsulfoxide (DMSO) and kept
at ꢁ20 ꢀC until use. The percentage of DMSO was kept at 0.5% (v/v)
whatever the concentration tested. This percentage did not modify
cellular growth. Fifty microliters of a 4X solution in MEM were then
added and a 48 h continuous drug exposure protocol was used.
Then, the cytotoxic effect of compounds on tumor cells was
assessed by using Resazurin reduction test.
5.6.3. Resazurin reduction test
The resazurin reduction test (RRT) was carried out according to
the protocol described previously [38]. Briefly, plates were rinsed by
MS analyses were performed using total ion current (TIC)
chromatograms. Specific compounds (m/z corresponding to pro-
drugs and/or specified degradation products) were detected using
extracted ion current (EIC) chromatograms.
200
(Labsystems, Helsinki, Finland) and emptied by overturning on
absorbent toweling. Then,150 L of a 25 g/mL solution of resazurin
m
L PBS (37 ꢀC, Invitrogen) using a multichannel dispenser
m
m
in MEM without SVF or Phenol red was added in each well. The
plates were incubated 1 h at 37 ꢀC in a humidified atmosphere with
5% of CO2 for fluorescence development by living cells. Fluorescence
was then measured on the automated 96-well plate reader Fluo-
roskan Ascent FLÔ (Labsystems) using an excitation wavelength of
530 nm and an emission one of 590 nm. The fluorescence is
proportional to the number of living cells in the well and IC50 (drug
concentration required to decrease final cell population by 50%) was
calculated from the curve of concentration-dependent cell number
decrease, defined as the fluorescence in experimental wells as
a percentage of that in control wells, with blank values subtracted.
5.4. Sample preparation
The stability studies of potential prodrug in solution were con-
ducted at 37 ꢂ 2 ꢀC. A stock solution of each compound was
prepared in methanol or in a mixture of methanol and acetonitrile
(85:15, v:v) to give a final concentration of 10 mM. From this stock
solution, a 1:10 dilution was prepared in the relevant media (PBS
(Gibco), MEM or human plasma) in 7 mL glass tubes. The solution
was maintained at 37 ꢂ 2 ꢀC in an oven, and aliquots (500
mL) were
withdrawn at various time points, quenched with an equal volume
of ammonium formate buffer 10 mM pH 3 and vortexed for 1 min.
After centrifugation at 3200 r.p.m. for 10 min, an aliquot of super-
Acknowledgments
natant was filtered through a 0.45 mm PVDF filter (AIT, France) and
diluted (3:10) in 10 mM ammonium formate buffer. The mixture
was vortexed and analyzed by the chromatographic method
described above. The disappearance of the potential prodrug was
followed by HPLC-UV analysis. All the experiments were carried out
in triplicate unless otherwise stated.
The authors are grateful to M-J. Couret (Clermont Université,
Université d’Auvergne, BP 10448, F-63000 CLERMONT-FERRAND)
for HPLC-UV technical assistance.
References
5.5. Data analysis
[1] C. Heidelberger, N.K. Chaudhuri, P. Danneberg, D. Mooren, L. Griesbach,
R. Duschinsky, R.J. Schnitzer, E. Pleven, J. Scheiner, Fluorinated pyrimidines,
a new class of tumour-inhibitory compounds, Nature 179 (1957) 663e666.
[2] F. Tanaka, T. Fukuse, H. Wada, M. Fukushima, The history, mechanism and
clinical use of oral 5-fluorouracil derivative chemotherapeutic agents, Curr.
Pharm. Biotechnol. 1 (2000) 137e164.
The time-dependent decline in the concentrations of the
potential 5-FU derivatives after incubation in various media was
fitted using a first-order kinetics model, and their degradation half-
lives were estimated from the slopes of fitted lines.
[3] D.B. Longley, D.P. Harkin, P.G. Johnston, 5-fluorouracil: mechanisms of action
and clinical strategies, Nat. Rev. Cancer 3 (2003) 330e338.
[4] M. Malet-Martino, P. Jolimaitre, R. Martino, The prodrugs of 5-fluorouracil,
Curr. Med. Chem. Anticanc Agents 2 (2002) 267e310.
5.6. Biological assays
[5] M. Miwa, M. Ura, M. Nishida, N. Sawada, T. Ishikawa, K. Mori, N. Shimma,
I. Umeda, H. Ishitsuka, Design of a novel oral fluoropyrimidine carbamate,
capecitabine, which generates 5-fluorouracil selectively in tumours by
enzymes concentrated in human liver and cancer tissue, Eur. J. Cancer 34
(1998) 1274e1281.
5.6.1. Cell culture
Normal human fibroblasts were purchased from Promocell
(Heidelberg, Germany). This frozen culture was obtained from

P. Daumar et al. / European Journal of Medicinal Chemistry 46 (2011) 2867e2879
2879
[6] C.M. Walko, C. Lindley, Capecitabine: a review, Clin. Ther. 27 (2005) 23e44.
[7] W.A. Denny, Prodrug strategies in cancer therapy, Eur. J. Med. Chem. 36
(2001) 577e595.
[22] S. Ozaki, Y. Ike, H. Mizuno, K. Ishikawa, H. Mori, 5-Fluorouracil derivatives. I. The
synthesis of 1-carbamoyl-5-fluorouracils, Bull. Chem. Soc. Jpn. 45 (1977)
1372e1375.
[8] T.M. Allen, Ligand-targeted therapeutics in anticancer therapy, Nat. Rev.
Cancer 2 (2002) 750e763.
[23] S. Ozaki, Synthesis and antitumor activity of 5-fluorouracil derivatives, Med.
Res. Rev. 16 (1996) 51e86.
[9] S. Jaracz, J. Chen, L.V. Kuznetsova, I. Ojima, Recent advances in tumor-targeting
anticancer drug conjugates, Bioorg. Med. Chem. 13 (2005) 5043e5054.
[10] J.T.C. Wojtyk, R. Goyan, E. Gudgin-Dickson, R. Pottier, Exploiting tumour
biology to develop novel drug delivery strategies for PDT, Med. Laser Appl. 21
(2006) 225e238.
[11] T. Ganesh, Improved biochemical strategies for targeted delivery of taxoids,
Bioorg. Med. Chem. 15 (2007) 3597e3623.
[12] Z. Liu, S. Rimmer, Synthesis and release of 5-fluorouracil from poly(N-
vinylpyrrolidinone) bearing 5-fluorouracil derivatives, J. Control Release 81
(2002) 91e99.
[13] R. Madec-Lougerstay, J.-C. Florent, C. Monneret, Synthesis of self-immolative
glucuronide spacers based on aminomethylcarbamate. Application to 5-
fluorouracil prodrugs for antibody-directed enzyme prodrug therapy, J. Chem.
Soc. Perkin Trans. 1 (1999) 1369e1376.
[14] R.M. Phelan, M. Ostermeier, C.A. Townsend, Design and synthesis of a beta-
lactamase activated 5-fluorouracil prodrug, Bioorg. Med. Chem. Lett. 19
(2009) 1261e1263.
[15] M. Rapp, I. Giraud, J.C. Maurizis, M.J. Galmier, J.C. Madelmont, Synthesis and
in vivo biodisposition of [14C]-quaternary ammonium-melphalan conjugate,
a potential cartilage-targeted alkylating drug, Bioconjug. Chem. 14 (2003)
500e506.
[16] M. Rapp, J.C. Maurizis, J. Papon, P. Labarre, T.D. Wu, A. Croisy, J.L. Guerquin-Kern,
J.C. Madelmont, E. Mounetou, A new O6-alkylguanine-DNA alkyltransferase
inhibitor associated with a nitrosourea (cystemustine) validates a strategy of
melanoma-targeted therapy in murine B16 and human-resistant M4Beu
melanoma xenograft models, J. Pharmacol. Exp. Ther. 326 (2008) 171e177.
[17] J.M. Chezal, J. Papon, P. Labarre, C. Lartigue, M.J. Galmier, C. Decombat,
O. Chavignon, J. Maublant, J.C. Teulade, J.C. Madelmont, N. Moins, Evaluation
of radiolabeled (hetero)aromatic analogues of N-(2-diethylaminoethyl)-4-
iodobenzamide for imaging and targeted radionuclide therapy of melanoma,
J. Med. Chem. 51 (2008) 3133e3144.
[18] B. Reux, V. Weber, M.J. Galmier, M. Borel, M. Madesclaire, J.C. Madelmont,
E. Debiton, P. Coudert, Synthesis and cytotoxic properties of new
fluorodeoxyglucose-coupled chlorambucil derivatives, Bioorg. Med. Chem. 16
(2008) 5004e5020.
[19] E. Miot-Noirault, B. Reux, E. Debiton, J.-C. Madelmont, J.-M. Chezal, P. Coudert,
V. Weber, Preclinical investigation of tolerance and antitumour activity of
new fluorodeoxyglucose-coupled chlorambucil alkylating agents, Invest. New
Drugs (2009) 1e10.
[24] S. Ozaki, X. Kong, Y. Watanabe, T. Hoshiko, T. Koga, T. Ogasawara, T. Takizawa,
H. Fujisawa, M. Iigo, A. Hoshi, 5-Fluorouracil derivatives. XXII. Synthesis and
antitumor activities of 1-carbamoyl-5-fluorouracils, Chem. Pharm. Bull. 45
(1997) 1372e1375.
[25] A. Buur, H. Bundgaard, Prodrugs of 5-fluorouracil. III. Hydrolysis kinetics
in aqueous solution and biological media, lipophilicity and solubility of
various 1-carbamoyl derivatives of 5-fluorouracil, Int. J. Pharm. 23 (1985)
209e222.
[26] Y. Suda, K. Shimidzu, M. Sumi, N. Oku, S. Kusumoto, T. Nadai, S. Yamashita,
The synthesis and in vitro and in vivo stability of 5-fluorouracil prodrugs
which possess serum albumin binding potency, Biol. Pharm. Bull. 16 (1993)
876e878.
[27] N. Oku, S. Yamashita, N. Sakuragi, K. Doi, S. Okada, K. Shimidzu, M. Sumi,
T. Nadai, S. Kusumoto, Y. Suda, Therapeutic efficacy of 5-fluorouracil prodrugs
using endogenous serum proteins as drug carriers: a new strategy in drug
delivery system, Biol. Pharm. Bull. 18 (1995) 181e184.
[28] H. Vorbrüggen, K. Krolikiewicz, B. Bennua, Nucleosides syntheses, XXII.
Nucleoside synthesis with trimethylsilyl triflate and perchlorate as catalysts,
Chem. Ber. 114 (1981) 1234e1255.
[29] G. Zemplén, Abbau der reduzierenden Biosen I.: Direk Konstitutions-
ermittlung der cellobiose, Ber. Dtsch. Chem. Ges. 59B (1926) 1254e1260.
[30] G. Zemplén, Abbau der reduzierenden Biosen, VII.: Konstitutions-Ermittlung
der Maltose, Ber. Dtsch. Chem. Ges. 60B (1927) 1555e1564.
[31] C. Ausin, J.A. Ortega, J. Robles, A. Grandas, E. Pedroso, Synthesis of amino- and
guanidino-G-clamp PNA monomers, Org. Lett. 4 (2002) 4073e4075.
[32] L. Cotarca, P. Delogu, A. Nardelli, V. Sunnjic, Bis(trichloromethyl) carbonate in
organic synthesis, Synthesis 5 (1996) 553e576.
[33] C. Grandjean, A. Boutonnier, C. Guerreiro, J.M. Fournier, L.A. Mulard, On the
preparation of carbohydrate-protein conjugates using the traceless Stau-
dinger ligation, J. Org. Chem. 70 (2005) 7123e7132.
[34] J.R. McDermott, N.L. Benoiton, N-Methylamino acids in peptide synthesis. II. A
new synthesis of N-Benzyloxycarbonyl, N-Methylamino acids, Can. J. Chem.
51 (1973) 1915e1919.
[35] D.A. Evans, S. Mito, D. Seidel, Scope and mechanism of enantioselective
Michael additions of 1,3-dicarbonyl compounds to nitroalkenes catalyzed by
nickel(II)-diamine complexes, J. Am. Chem. Soc. 129 (2007) 11583e11592.
[36] J.H. Shah, S. Izenwasser, B. Geter-Douglass, J.M. Witkin, A.H. Newman,
(þ/-)-(N-alkylamino)benzazepine analogs: novel dopamine D1 receptor
antagonists, J. Med. Chem. 38 (1995) 4284e4293.
[20] S.M. Courtney, Therapeutic compounds with pyrimidines base, PCT/GB96/
012519, 1999.
[21] P. Yin, M.-L. Hu, L.-C. Hu, Synthesis, structural characterization and anticarci-
nogenic activity of a new Gly-Gly dipeptide derivative: methyl 2-(2-(5-fluoro-
2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)acetate, J. Mol. Struct. 882
(2008) 75e79.
[37] Y. Aramaki, M. Seto, T. Okawa, T. Oda, N. Kanzaki, M. Shiraishi, Synthesis of 1-
benzothiepine and 1-benzazepine derivatives as orally active CCR5 antago-
nists, Chem. Pharm. Bull. 52 (2004) 254e258.
[38] M. Sassatelli, É Debiton, B. Aboab, M. Prudhomme, P. Moreau, Synthesis and
antiproliferative activities of indolin-2-one derivatives bearing amino acid
moieties, Eur. J. Med. Chem. 41 (2006) 709e716.